provided for non-commercial research and …...the canadian journal of veterinary research is...

The Canadian Journal of Veterinary Research is published by Canadian Veterinary Medical Association. The attached copy is

furnished to readers for personal, non-commercial research and education use. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party Web sites are prohibited. Those who require further

information regarding reprints or archiving and manuscript policies are encouraged to contact [email protected].

Provided for non-commercial research and education use. Not for reproduction, distribution or commercial use.

Canadian Journal of Veterinary Research Revue Canadienne de Recherche Vétérinaire

Canadian Journal of Veterinary Research Revue Canadienne de Recherche Vétérinaire

ArticlesA 5-year study of the incidence and economic impact of variant infectious bursal disease viruses on broiler production in Saskatchewan, CanadaTara Zachar, Shelly Popowich, Bob Goodhope, Tennille Knezacek, Davor Ojkic, Philip Willson, Khawaja Ashfaque Ahmed, Susantha Gomis . . . . . . . . . . . . . . . . . . . . . . . .255

Detection and phylogenetic analysis of bovine papillomavirus in cutaneous warts in cattle in Tamaulipas, MexicoEdith Rojas-Anaya, Antonio Cantú-Covarrubias, José Francisco Morales Álvarez, Elizabeth Loza-Rubio . . . . . . . . . . . . . . . . . . . 262

Safety and efficacy of a novel European vaccine for porcine reproductive and respiratory virus in bred giltsMichael D . Piontkowski, Jeremy Kroll, Francois-Xavier Orveillon, Christian Kraft, Teresa Coll . . . . . . . . . . . . . . . . . . . . . . . . . . . 269

Efficient construction of Haemophilus parasuis mutants based on natural transformationJunxing Li, Xiufang Yuan, Lihua Xu, Lei Kang, Jun Jiang, Yicheng Wang . . . . . . . .281

Minimum dose, antigen content, and immunization duration of a trivalent vaccine of inactivated Haemophilus parasuis serovars 4, 5, and 12 against Glässer’s disease in pigsZhanqin Zhao, Huisheng Liu, Keshan Zhang, Qiao Xue, Kunpeng Chen, Yun Xue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

Evaluation of adipose-derived stromal vascular fraction from the lateral tailhead, inguinal region, and mesentery of horsesGarrett L . Metcalf, Scott R . McClure, Jesse M . Hostetter, Rudy F . Martinez, Chong Wang . . . . . . . . . . . . . . . . . . . . . . . . . 294

O c t o b e r / O c t o b r e 2 0 1 6 , V o l . 8 0 N o . 4table of contents continued on back cover/la table des matières se poursuit sur la couverture arrière

FOR PERSONAL USE ONLY

Published quarterly by the Canadian Veterinary Medical AssociationPublication trimestrielle de l’Association canadienne des médecins vétérinaires

C a n a d i a n J o u r n a l o fVe t e r i n a r y R e s e a r c h

Formerly CANADIAN JOURNAL OF COMPARATIVE MEDICINE

The journal publishes the results of original research in veterinary and comparative medicine. Manuscripts must be as concise as pos-sible, and the research described must represent a significant con-tribution to knowledge in veterinary medicine. Full-length papers, short communications, and review papers are welcome. All manu-scripts will be reviewed for scientific content and editorial accuracy.

Manuscripts must conform to the Instructions for Authors found on the CVMA website. Consultation of a recent issue of the journal to ensure that the manuscript conforms to current style is also recom-mended. Please use the website (www.canadianveterinarians.net) to submit articles. A publication charge of $90 per published page for CVMA Members and $130 per published page for Non-members, plus a cost of $15 per table, $20 per illustration, and $150 per color illustration is payable by contributors. Instructions for ordering reprints are sent with the galley proofs. Color illustrations will only be published if the author(s) agree to pay the extra cost.

R e v u e c a n a d i e n n e d er e c h e r c h e v é t é r i n a i r e

Remplace la REVUE CANADIENNE DE MÉDECINE COMPARÉE

La revue publie les résultats de recherches originales en médecine vétérinaire et comparative. Les manuscrits doivent être aussi brefs que possible et la recherche décrite doit apporter une contribution impor-tante à la médecine vétérinaire. Les exposés complets, les communi-cations brèves et les comptes rendus sont bienvenus. Nous véri fierons le fonds scientifique et la forme de tous les manuscrits.

Les manuscrits doivent être conformes aux directives aux auteurs qui se trouvent sur le site Web de l’ACMV. La consultation d’un récent numéro de la revue est aussi profitable, afin de s’assurer qu’un manuscrit se conforme au style courant. Veuillez utiliser le site Web (www.veterinairesaucanada.net) pour soumettre les articles. Les auteurs doivent acquitter des frais de 90 $ par page publiée pour les membres de l’ACMV et 130 $ par page publiée pour les non-membres, 15 $ pour chaque tableau, 20 $ pour chaque illustration, et 150 $ pour chaque illustration en couleur. Un formulaire de commande de tirés à part accompagne l’en voi des épreuves d’un manuscrit. Les illustrations en couleurs ne seront publiées que si les auteurs en assument les frais supplémentaires.

Established in 1937 ISSN 0830-9000

Address all correspondence to/Adresser toute correspondance à :Canadian Journal of Veterinary Research Revue canadienne de recherche vétérinaire339, rue Booth Street, Ottawa, Ontario K1R 7K1Tel./Tél. : (613) 236-1162 — Fax/Télécopieur : (613) 236-9681e-mail/Courriel : [email protected]© Canadian Veterinary Medical Association 2016

L’Association canadienne des médecins vétérinaires 2016

Typesetting/Typographie AN Design Communications www.an-design.ca

Canadian Journal of Veterinary ResearchRevue canadienne de recherche vétérinaireEditor — Rédacteur Éva Nagy, Guelph, OntarioAssociate Editor — Rédacteur adjoint Faizal Careem, Calgary, Alberta Deborah M. Haines, Saskatoon, SaskatchewanAssistant Editors — Assistants à la rédaction Serge Messier, Saint-Hyacinthe (Québec)Managing Editor — Directrice de la rédaction Heather Broughton, Ottawa, OntarioAssistant Managing Editor — Directrice adjointe à la rédaction Stella Wheatley, Ottawa, OntarioAdvertising Manager — Gérante de la publicité Laima Laffitte, Wendover, OntarioEditorial Coordinator/Coordonnatrice de la rédaction Linda Chow, Ottawa, Ontario

Abstracted and/or indexed in:/Résumé et/ou catalogué dans :Biological Abstracts Index VeterinariusChemical Abstracts Nutrition AbstractsCurrent Contents (Agric. Section) Science Citation IndexEMBASE/Excerpta Medica Veterinary BulletinIndex Medicus Derwent Veterinary Drug File

Subscription/Abonnement 2016Personal Subscription/ $165 $ (1 GST/HST as Abonnement personnel applicable; TPS/TVH en (Canada) vigueur)Personal Subscription/ $185 $ US/É.-U. Abonnement personnel (Foreign/Étranger)Institutional Subscription/ $215 $ (1 In Canada, GST/HST Abonnement institutionnel as applicable; Au Canada, TPS/TVH selon le cas)


Article

2016;80:255–261 The Canadian Journal of Veterinary Research 255

A 5-year study of the incidence and economic impact of variant infectious bursal disease viruses on broiler production in Saskatchewan, Canada

Tara Zachar, Shelly Popowich, Bob Goodhope, Tennille Knezacek, Davor Ojkic, Philip Willson, Khawaja Ashfaque Ahmed, Susantha Gomis

A b s t r a c tWhile the prevalence of infectious bursal disease virus (IBDV) on chicken farms in some provinces of Canada has been documented, the economic impact of variant IBDV infection on the broiler chicken industry in Saskatchewan has not. The objectives of this study were to identify the variant strains of IBDV circulating on Saskatchewan chicken farms and evaluate their economic impact on broiler production. Infection due to IBDV was detected in 43% of Saskatchewan chicken farms, with variant strains detected in infected birds closely related predominantly to NC171, 586, and Delaware-E. Infected flocks showed an IBDV antibody titer of 4236 geometric mean (GM), whereas an antibody titer of 157 GM was measured in uninfected flocks. Infected flocks had very low (0.06) bursa-to-body-weight (BBW) ratio (an indicator of immunity) compared to high BBW ratio (0.17) in uninfected flocks, which suggests a significant immunosuppression in the former. Flocks positive for IBDV had mean mortality of 8.6% and mean condemnation of 1.5%. In contrast, mean mortality in uninfected flocks was 6.1% and mean condemnation was 1.1%. The live market weight per grow area at 37 d of age was 29.3 kg/m2 in infected flocks and 34.0 kg/m2 in flocks without IBDV infection. Flock mortality and condemnation rate were positively correlated with IBDV infection, whereas low BBW ratio was inversely correlated, as expected. Overall, IBDV-infected flocks had higher mortality, bursal atrophy, poorer feed conversion ratio (FCR), and decreased meat production. Our data suggest that the broiler chicken industry in Saskatchewan loses 3.9 million kilograms of meat production per year due to variant IBDV strains.

R é s u m éBien que la prévalence du virus de la maladie infectieuse de la bourse (VMIB) sur les fermes de poulets dans quelques provinces canadiennes ait été documentée, l’impact économique d’infections par des variants du VMIB sur l’industrie du poulet à griller en Saskatchewan ne l’est pas. Les objectifs de la présente étude étaient d’identifier les souches variantes du VMIB circulant au sein des fermes de poulet de la Saskatchewan et d’évaluer leur impact économique sur la production de poulets à griller. L’infection due au VMIB a été détectée sur 43 % des fermes de poulet de la Saskatchewan, avec des souches variantes détectées chez des oiseaux infectés étant fortement apparentées en prédominance aux souches NC171, 586, et Delaware-E. Les troupeaux infectés présentaient une moyenne géométrique (MG) des titres d’anticorps contre le VMIB de 4236, alors que la MG des titres d’anticorps des oiseaux provenant de troupeaux non-infectés était de 157. Les troupeaux infectés avaient un très faible ratio (0,06) du poids bourse-poids corporel (BPC) (un indicateur de l’immunité) comparativement au ratio BPC élevé (0,17) dans les troupeaux non-infectés, ce qui suggère une immunosuppression significative dans les troupeaux infectés. Les troupeaux positifs au VMIB avaient un taux de mortalité moyen de 8,6 % et un taux de condamnation moyen de 1,5 %. À l’opposé, dans les troupeaux non-infectés le taux de mortalité moyen était de 6,1 % et le taux de condamnation moyen de 1,1 %. Le poids vif de marché à 37 j d’âge par surface de croissance était de 29,3 kg/m2 dans les troupeaux infectés et de 34,0 kg/m2 dans les troupeaux sans infection par VMIB. La mortalité dans le troupeau et le taux de condamnation étaient corrélés positivement avec l’infection par VMIB, alors qu’un faible ratio BPC était corrélé, tel qu’attendu, de manière inverse. De manière générale, les troupeaux infectés par le VMIB présentaient une mortalité plus élevée, une atrophie de la bourse, un mauvais ratio de conversion alimentaire, et une production de viande réduite. Nos données suggèrent que l’industrie du poulet à griller en Saskatchewan perd annuellement 3,9 millions de kilogrammes de production de viande à cause des souches variantes du VMIB.

(Traduit par Docteur Serge Messier)

Department of Veterinary Pathology, Western College of Veterinary Medicine (Zachar, Popowich, Goodhope, Ahmed, Gomis), Department of Animal and Poultry Science, College of Agriculture and Bioresources (Knezacek), and Canadian Centre for Health and Safety in Agriculture, University of Saskatchewan, Saskatoon, Saskatchewan (Willson); Animal Health Laboratory, University of Guelph, P.O. Box 3612, Guelph, Ontario (Ojkic).

Address all correspondence to Dr. Susantha Gomis; telephone: (306) 966-7299; fax: (306) 966-7439; e-mail: [email protected]

Received January 7, 2016. Accepted April 6, 2016.


256 The Canadian Journal of Veterinary Research 2000;64:0–00

I n t r o d u c t i o nInfectious bursal disease (IBD) is a highly contagious immuno-

suppressive disease that causes serious problems for the poultry industry worldwide (1). The disease is also known as Gumboro disease as it was first recognized in the United States near the town of Gumboro, Delaware (2). Infectious bursal disease is caused by the IBD virus (IBDV), a double-stranded ribonucleic acid (RNA) virus that belongs to the genus Avibirnavirus of the family Birnaviridae (3). The IBD virus exists in classical and variant forms that are antigeni-cally distinct (2). Emergence of variant IBDV strains has caused sub-stantial losses in the poultry industry. Most IBDV strains circulating in the United States (2) and Canada (4) are variant strains. Infection by IBDV can cause direct economic loss due to specific mortality (1,2). The indirect economic impact of IBD is profound, however, due to IBDV-induced immunosuppression that predisposes chickens to secondary infections by bacteria, viruses, and parasites, and results in increased mortality, growth retardation, and condemnation (5–7).

Infection by the IBD virus targets the bursa of Fabricius (BF) and results in severe immunosuppression due to loss of lymphocytes in the BF (8,9). There is strong evidence that surface immunoglobulin M-bearing B-lymphocytes are the major targets for IBDV (8,10). Variant IBDV strains (11,12) in neonatal chickens can escape from the maternal antibodies (MAb) against IBDV that are produced by broiler breeders immunized with standard or classic IBDV strain vaccines (7) and induce severe bursal damage (13–16). This results in profound immunosuppression and subclinical infections (16,17). Immunosuppression is greatest when infection occurs early post-hatch and is permanent because the damaged BF does not regenerate normal immune function. As the immunosuppression resulting from an IBDV infection is often the underlying cause of respiratory and enteric disease in chickens, as well as vaccination failures, IBD is an economically significant disease throughout the world (11). It may also precipitate a variety of other diseases or conditions that could play a role in airsacculitis, septicemia, toxemia, increasing condem-nations, and decreasing broiler growth rate (10,13).

Although variant IBDV infection has recently been confirmed on Canadian chicken farms (18), its economic impact on the Saskatchewan broiler chicken industry has not been studied. The objectives of this study were to identify the incidence of variant IBDV infection and to evaluate the associated economic losses in the broiler chicken industry in Saskatchewan, Canada.

M a t e r i a l s a n d m e t h o d s

Study designThis study was conducted over a 5-year period, with sample

collection occurring in 2007, 2009, and 2011. According to the Chicken Farmers of Saskatchewan, there were 63 broiler farms in the Saskatchewan broiler chicken industry from 2007 to 2011. In 2007, the incidence of IBDV was studied by histologic examination of the bursa of Fabricius (BF) and polymerase chain reaction (PCR). A follow-up survey was conducted in 2009 to obtain major economic parameters, such as farm and flock (birds kept in a barn on a poultry farm) information, feed conversion ratio (FCR), and stocking density,

in addition to histologic examination of BF, bursa-to-body-weight (BBW) ratio, and serology against IBDV. Another follow-up study was conducted in 2011 at those farms that were found to have been infected with IBDV in 2007 and 2009.

Incidence of variant IBDV in broiler farms in 2007In 2007, 177 flocks managed by 58 of 63 broiler farms (92%) were

chosen for this study. Bursas were collected from 40 randomly selected birds from farms with multiple barns (flocks) and from 20 birds from farms with only 1 barn (flock) at the time of processing for histologic examination throughout the 1-year period. Tissue sec-tions of bursae were fixed in 10% neutral-buffered formalin, embed-ded in paraffin, sectioned at a thickness of 5 mm, and stained with hematoxylin and eosin (H&E). A histologic score from 0 to 3 (0 = no visible lesions; 1 = mild, focal lymphoid depletion; 2 = moderate, multifocal lymphoid depletion; and 3 = moderate to severe multifo-cal lymphoid depletion) was assigned based on the tissue reactions of the BF (Figure 1). A bursal score was assigned to each bird and a cumulative score was determined per flock with the lowest score being 0 (no bursal atrophy) and the highest score being 60 (20 birds per flock with severe bursal atrophy). All farms with bursal atrophy and 10 farms with no bursal atrophy on histologic examination were further examined by collecting bursae from 10 randomly selected birds at 19 d of age for identifying IBDV by real-time polymerase chain reaction (RT-PCR) and nucleotide sequence analysis of the viral protein 2 (VP2) hypervariable region (19). This was carried out at the Animal Health Laboratory (AHL), University of Guelph, Ontario, as described previously (4). Chickens were vaccinated for IBDV 10 to 14 d post-hatch in 4 out of 58 farms (2 farms with IBDV infection and another 2 farms without IBDV infection) with Clonevac D-78 (Intervet Canada, Whitby, Ontario).

Economic impact of variant IBDV in broiler farms in 2009

In 2009, a follow-up survey was conducted to collect data from the same 58 farms (177 barns) on the following broiler production variables: number of barns (flocks) per farm; type of barn; age and flooring of barns; total number of birds placed; stocking density; downtime between flocks; vaccines administered; FCR; age at slaughter; market weight; total birds processed; and total flock mortality, including culled birds and total condemnations at the time of processing. For each barn, data were collected from 1 production cycle from April to June, 2009. In addition to collecting production data at all participating locations, additional studies were done on some farms. At the time of processing, sets of 20 birds were ran-domly selected from 52 of 177 barns for IBDV serology, BBW ratio, and histologic examination of BF.

Persistence of variant IBDV in broiler farms infected since 2007

In 2011, a follow-up serological study for IBDV was conducted in 82 of 177 barns (located on 25 of 58 farms) that were confirmed to have had IBDV infection in 2007 and 2009. Confirmation of IBDV in farms was defined as the presence of 2 out of 3 criteria associated with IBDV infection, namely IBDV titer, BBW ratio, or bursal atro-phy score. Infectious bursal disease (IBD) virus titers were assayed



using commercial enzyme-linked immunosorbent assay (ELISA) kits according to manufacturer’s instructions (IDEXX, Westbrook, Maine, USA) at the AHL, University of Guelph, Ontario.

Statistical analysisDescriptive statistics and correlations among observed variables

were calculated using Prism 5.0 (GraphPad Software, San Diego, California, USA). Geometric means (GM) were used to describe the central tendency of variables that were not normally distributed (antibody titer) and arithmetic means were used for the remaining variables. The indicators of infection and disease (cumulative his-topathology score of the BF, BBW ratio, and IBDV GM titer) were considered individually in order to determine the classification of farms. Wilcoxon Rank-Sum Test was used to demonstrate the sig-

nificance of differences in the central values of samples from each category of farm. Stepwise logistic regression was used to analyze the production variables [downtime between flocks, FCR, and meat production per square meter at 37 d of age (kg37/m2)] in flocks with or without IBDV infection using Statistix 7 (Analytical Software, Tallahassee, Florida, USA).

Re s u l t s

Incidence of variant IBDV in broiler farms in 2007The histologic study conducted in 2007 classified 33 of 58 farms

(57%) with minimal to no bursal damage and a mean cumulative score of 0.24 per farm. In contrast, 25 of 58 premises (43%) had

Figure 1. Histopathological appearance of bursa of Fabricius (BF) at 23: A — normal follicles of BF; B — mild bursal atrophy — follicular atro-phy of some follicles; C — moderate bursal atrophy — small lymphoid follicles and prominent interfollicular architecture; D — severe bursal atrophy — the entire BF architecture shows severe, diffuse loss of follicles.



moderate to severe bursal damage, with a mean cumulative score of 40.3 per farm (Figure 1). No variant IBDVs were isolated in 10 of 33 farms with minimal to no bursal atrophy on histologic examina-tion. In contrast, variant IBDV strains similar to NC171, 586, and Delaware E were isolated from 19-day-old broilers in 12 of the 25 farms with moderate to severe bursal atrophy (Figure 2).

Economic impact of variant IBDV in broiler farms in 2009

Production variables were collected from 58 broiler farms (177 barns) in Saskatchewan. Of 58 farms, 10 farms had a single barn, 15 farms had 2 barns, 18 farms had 3 barns, 10 farms had 4 barns, and the remaining 5 farms had 5 to 13 barns. The barns on 36 of 58 farms were $ 10 years old and 1 to 9 years old on 22 of 58 farms. Forty-seven of 58 farms had single-story barns, 5 of 58 farms had double-story barns, and the remaining 6 farms had both single- and double-story barns. Twenty-seven of 58 farms had concrete floors and the remaining farms had a combination of partial concrete, clay, sand, wood, or soil floors. The mean floor area of a barn was 4143 m2 and ranged from 743 to 14 864 m2. The median distance between processing plant and farm was 160 km, with a range of 2 to 411 km (Table I).

The mean placement on a farm was 69 048 birds, ranging from 13 000 to 217 700 birds. The mean bird density was 0.06 m2/bird, ranging from 0.04 to 0.09 m2/bird. The mean age at marketing was 37 d and ranged from 33 to 45 d. The mean market weight of a bird was 2.02 kg and ranged from 1.60 to 2.58 kg. The mean FCR was

1.83 and ranged from 1.59 to 2.22. The mean live bird production was 32.36 kg/m2 and ranged from 21.48 to 49.60 kg/m2. The mean mortality, including culled birds, was 7.04% and ranged from 1.26 to 20.7%. The mean BBW on a farm was 0.11% and ranged from 0.04% to 0.21%. The mean condemnation was 1.25% and ranged from 0.07% to 4.94%. Thirty-six of 58 farms maintained 9 d or longer downtime between flocks after cleaning and disinfection were completed. All broiler placements in Saskatchewan were supplied by 2 hatcheries and all broiler flocks were processed by 2 processing plants (Table I).

Infectious bursal disease (IBD) virus titers were positively cor-related with the cumulative histologic score of the BF (r = 0.71, P = 0.0001), flock mortality (r = 0.47, P = 0.02), condemnation rates (r = 0.44, P = 0.02) and negatively correlated with BBW ratio (r = 20.79, P = 0.000001). The condemnation rates at the time of processing were negatively correlated with the downtime between flocks (r = 20.31, P = 0.02) and BBW ratio (r = 20.53, P = 0.005). There is no significant correlation between broiler breeder IBDV vaccination against bursal atrophy, BBW ratio, or IBDV titer (r = 0.25, P = 0.25; r = 20.27, P = 0.19; and r = 0.30, P = 0.19, respectively).

Confirmation of IBDV in farms was defined as the presence of 2 of 3 criteria associated with IBDV infection, namely IBDV titer, BBW ratio, or bursal atrophy on histologic examination, as already described. Birds from farms with IBDV infection had a GM titer of 4236 (1578 to 7466) to IBDV. In contrast, farms with no IBDV infection had a GM titer of 157 (2 to 788) to IBDV. Farms with IBDV infection had a BBW ratio of 0.06 (0.04 to 0.08), while farms with no IBDV infection had a BBW ratio of 0.17 (0.13 to 0.21). Farms with

Figure 2. Phylogenetic analysis of variant IBDVs isolated in the Saskatchewan broiler chicken industry. A — Phylogenetic tree; B — Percent identity. Isolates similar to Del E = 13-010241-0001, 14-081981-0001; isolate similar to 586 = 13-010241-0002; isolates similar to NC171 = 13-004049-0001, 13-005381-0001, 14-034431-0001, 14-049186-0001, and 14-049186-0002.

A

B

Nucleotide substitution per 100 residues



IBDV infection had a cumulative histologic score of 49.58 (20.30 to 58.75), while farms with no IBDV infection had a cumulative his-tologic score of 2.11 (0.50 to 4.75). The FCR on farms with IBDV infection was 1.87, while the FCR was 1.81 in farms with no IBDV infection. The mean mortality was 8.6% (4.35% to 20.7%) in farms with IBDV infection and 6.1% (1.26% to 18.6%) on farms with no IBDV infection. The mean condemnation rate was 1.48% (0.25% to 4.94%) on farms with IBDV infection and 1.12% (0.07% to 4.6%) on farms with no IBDV infection. The live market weight per grow area (meat production) was 29.3 kg37/m2 on farms with IBDV infection and 34.0 kg37/m2 on farms with no IBDV infection (Table II).

On average, broiler producers repeat 6.5 production cycles per year. Based on our data (number of barns infected with IBDV, number of birds in barns infected with IBDV, difference in kg37/m2 meat production in barns with and without IBDV infection, and number of production cycles per year), the broiler chicken industry in Saskatchewan loses approximately 3.9 million kg per year due to “variant” IBDV infection. In 2014, this amount of chicken had a wholesale market value of over $13 million. (The wholesale price of chicken was $3.58/kg in 2014, according to the 2014 Annual Report of the Chicken Farmers of Canada.)

Persistence of variant IBDV in broiler farms infected since 2007

Of 58 farms studied, 10 out of 33 farms were without any sign of IBDV infection, i.e., no bursal atrophy, no IBDV detection, and low titers against IBDV in 2007 and 2009, and had an average IBDV GM titer of 42.6 (27 to 336) in 2011. In contrast, 23 out of 25 farms (92%) with a history of IBDV infection in 2007 and 2009 had an average IBDV GM titer of 4063 (1104 to 9412). Only 2 out of 25 farms (8%)

with IBDV infection in both 2007 and 2009 had a GM IBDV titer in 2011 that was classified as low (, 50). These 2 farms had a GM IBDV titer of 34.5 (26 to 43). Twenty-three of 25 farms (92%) maintained IBDV infection from 2007 to 2011.

D i s c u s s i o nThe economic impact of infectious bursal disease (IBD) on the

chicken industry is difficult to assess due to the complex nature of losses associated with this disease. In addition to direct losses, IBDV infection-induced immunodeficiency in chickens opens the door to other viral, bacterial, and parasitic infections, thus inflicting heavy indirect losses due to increased morbidity, mortality, and condemna-tion (6). Moreover, IBD viruses are resistant to many disinfectants and environmental factors. Once a poultry house becomes contami-nated with IBDV, viruses persist on the premises and disease tends to reappear in subsequent flocks (20). An effective IBDV prevention and control program must involve an effective surveillance strategy. This 5-year study was started in 2007 to investigate the prevalence and persistence of IBDV on Saskatchewan chicken farms and assess the financial impact on the chicken industry.

In 2007, histologic study identified 43% of farms with bursal atrophy, where RT-PCR and sequencing associated these 43% of farms with variant IBDV infection. The remaining 57% of premises showed no bursal atrophy and IBDV infection was not detected. Bursal atrophy, low BBW ratio, and high IBDV titers were noted in farms with 5 or more barns, hence bursal atrophy was noted in 93 of 177 barns or broiler flocks (53%) in 2007 (Table II).

In 2009, a positive correlation of IBDV antibody titer with bursal atrophy and a negative correlation of IBDV antibody titer with BBW

Table I. Characteristics of the Saskatchewan broiler chicken industry (2007 to 2011)

Description of the study TotalNumber of premises (farms) in Saskatchewan 63Number of premises (farms) studied 58Number of barns (flocks) 177Number of premises with barns over 10 y of age 36Description of premises (farms) Mean RangeDistance between farm and processor (km) 160 2 to 411Number of barns per premise 3 1 to 13Barn size m2 (ft2) 4143 (44 600) 743 to 14 864 (8000 to 160 000)Birds per m2 (ft2 per bird) 17.3 (0.64) 11.5 to 26.7 (0.40 to 0.94)Meat production (kg/m2) 32.36 21.48 to 49.60Number of birds placed on each farm 69 048 13 000 to 217 770Mortality including culls (%) 7.04 1.26 to 20.70Feed conversion ratio (FCR) (feed offered/weight gain) 1.83 1.59 to 2.22Number of birds processed 64 098 10 860 to 207 830Processed weight per farm (kg) 135 538 21 806 to 440 653Live weight at processing (kg) 2.02 1.60 to 2.58Condemnation rate (%) 1.25 0.07 to 4.94Cumulative histopathology score of bursa of Fabricius (BF) per farm 29.81 0.50 to 58.75Bursa-to-body-weight (BBW) ratio 0.11 0.04 to 0.21IBDV titer [geometric mean (GM)] 2137 2 to 7466Downtime between flocks (d) 12 1.5 to 22



ratio were observed. Bursa-to-body-weight (BBW) ratio was also negatively correlated with condemnation rates. Phylogenetic analysis of VP2 sequences (19) obtained from the viruses isolated from the BF of 19-day-old broilers revealed that the IBDV strains circulating on Saskatchewan chicken farms with bursal atrophy are variant strains closely related to NC171, 586, and Delaware-E (Figure 2).

Flocks at premises infected with variant IBDV in 2007 and 2009 were tested again for the presence of IBDV antibodies in 2011 and 92% of these premises were serologically positive. It appeared that IBDV infection persisted for at least 4 y in broiler farms once they were infected. Infectious bursal disease (IBD) virus infection was controlled in only 2 of 25 farms that had few (3 or 4) barns. This indicates that control of IBDV infection was manageable in small farms with few barns. Furthermore, the incidence of IBDV infection remained very low (10%; 1 of 10 farms) in farms with a single broiler barn. In contrast, farms with 4 or more barns had a much higher incidence (87%; 13 of 15 farms) of IBDV infection.

Farms with IBDV infection had higher IBDV titers, poorer FCR, higher mortality, more bursal atrophy, higher condemnation rates, and less meat production than premises without IBDV infection. Farms with no detectable IBDV infection had better economical parameters, including better feed conversion ratio (FCR), lower mortality, and decreased condemnation rates. The live market weight per grow area was 29.3 kg37/m2 in farms with IBDV infec-tion and 34.0 kg37/m2 in farms with no IBDV infection. Broiler farms in Saskatchewan follow 6.5 production cycles per year. Since IBDV

infection was noted in 53% of barns [25 of 58 (43%) of farms] in the Saskatchewan broiler industry, the projected total production loss was 3.9 million kg in barns infected with IBDV.

While the present data showing significant economic losses due to IBDV infection is in agreement with previous studies (21), it is difficult to determine whether all the losses were due to IBDV alone or there were other contributing factors related to management, environment, and feed. We have seen a reduced weight gain of 160 to 200 g per bird at 35 d old in commercial broilers after experimental infection with NC171 compared to control birds receiving no IBDV in our level-2 animal isolation facility (22).

Less downtime between flocks was correlated with increased condemnation rates. Infectious bursal disease (IBD) virus titer was not correlated with downtime, however, which reflects the per-sistence of IBDV in a broiler barn for a long duration, as reported previously (23). According to the survey data, the Saskatchewan industry practiced downtime of 1.5 to 22 d between flocks. It is likely that this duration of downtime is not sufficient to reduce IBDV load in a barn, especially when thorough cleaning and dis-infection are not practiced. Infectious bursal disease (IBD) virus is very resistant to most disinfectants and environmental factors and persists for months in contaminated poultry houses and for weeks in water, feed, and fecal droppings (23). Although barns were cleaned and disinfected after each cycle of birds in the study area, we have shown that IBDV infection was more prevalent in older barns and barns with no concrete floors probably because

Table II. Production variables and indicators of infection on farms with or without variant infectious bursal disease viruses (vIBDVs)

Without vIBDV With vIBDVDescription of flocks Total Percentagea Total Percentagea

Number of farms 33 (57%) 25 (43%)Number of barns 84 (47%) 93 (53%)Number of farms with barns . 10 y of age 19 (53%) 17 (47%) Without vIBDV With vIBDVIndicators of infection and diseasec Mean Range Mean Range P-valueCumulative histopathology score of the BFb 2.11 0.5 to 4.75 49.58 20.3 to 58.75 , 0.0001Bursa-to-body-weight (BBW) ratio (percentage) 0.17 0.13 to 0.21 0.06 0.04 to 0.08 , 0.0001IBDV (GM titer) 157 2 to 788 4 236 1578 to 7466 , 0.0001 Without vIBDV With vIBDVProduction variablesd Mean Range Mean Range Odds ratioDowntime between flocks (days) 12 2 to 22 11 4 to 20 NSFeed conversion ratio (FCR) 1.81 1.59 to 2.22 1.87 1.74 to 2.04 NS

Meat production (kg37/m2) 34.0 18.1 to 54.0 29.3 21.9 to 39.0 0.83 (CI .72 to 0.95)

P = 0.006a The number in parentheses indicates the portion of the grand total in each disease category.b Cumulative histopathology score from 2007 or mean of scores from 2007 and 2009 if farms were sampled in both years.c Since each of the indicators of infection and disease is considered individually in order to determine the classification of farms, Wilcoxon Rank-Sum Test is used to demonstrate the significance of differences in the central values of samples from each category of farm.d Stepwise logistic regression was used to analyze the production variables. The odds of a “vIBDV present” disease occurrence are 0.83 (83% of) less for a flock with a meat production increase of 1 kg37/m

2. The difference in the means (34.0 to 29.3 = 4.7 kg/m2) indicates the amount of loss due to the disease.BF — Bursa of Fabricius; GM — Geometric mean; NS — Not significant; CI — Confidence interval.



these types of facilities are more difficult to thoroughly clean and disinfect.

Based on our data, it is concluded that unique variant IBDV strains were prevalent on Saskatchewan farms and probably attributed to substantial economic losses of about 3.9 million kg per year in the broiler chicken industry in Saskatchewan. Furthermore, it may sug-gest that current broiler and broiler breeder vaccination programs are not effective for controlling variant IBDV infections. Studies to identify potential vaccine candidates to control variant IBDV in Saskatchewan are urgently needed.

A c k n o w l e d g m e n t sFinancial grants were provided by the Chicken Farmers of

Saskatchewan (Saskatchewan Chicken Industry Development Fund), Saskatchewan Agriculture Development Fund, and Natural Sciences and Engineering Research Council. Special thanks are extended to all broiler producers in Saskatchewan who participated in this study. Special thanks also to Sofina Foods Inc. (Lilydale Inc.), Wynyard, Saskatchewan and Prairie Pride Natural Foods Ltd., Saskatoon, Saskatchewan for helping with sample collection. Thanks to Drs. Azita Taghavi, Angela Oranchuk, Brenda Bryan, and Tara Funk for assisting with data collection and histologic examinations.

Re f e r e n c e s 1. van den Berg TP, Eterradossi N, Toquin D, Meulemans G.

Infectious bursal disease (Gumboro disease). Rev Sci Tech 2000; 19:509–543.

2. Saif YM, Fadly AM. Diseases of Poultry. 12th ed ed. Ames, Iowa: Blackwell Pub, 2008.

3. Jackwood DJ, Sommer-Wagner SE, Crossley BM, Stoute ST, Woolcock PR, Charlton BR. Identification and pathogenicity of a natural reassortant between a very virulent serotype 1 infectious bursal disease virus (IBDV) and a serotype 2 IBDV. Virology 2011 25;420:98–105.

4. Ojkic D, Martin E, Swinton J, Binnington B, Brash M. Genotyping of Canadian field strains of infectious bursal disease virus. Avian Pathology 2007;36:427–433.

5. Lee HJ, Kim JY, Kye SJ, Seul HJ, Jung SC, Choi KS. Efficient self-assembly and protective efficacy of infectious bursal disease virus-like particles by a recombinant baculovirus co-expressing precursor polyprotein and VP4. Virol J 2015;12:177.

6. Berg TP. Acute infectious bursal disease in poultry: A review. Avian Pathol 2000;29:175–194.

7. Geerligs HJ, Ons E, Boelm GJ, Vancraeynest D. Efficacy, safety, and interactions of a live infectious bursal disease virus vac-cine for chickens based on Strain IBD V877. Avian Dis 2015;59: 114–121.

8. Tsukamoto K, Tanimura N, Mase M, Imai K. Comparison of virus replication efficiency in lymphoid tissues among three infectious bursal disease virus strains. Avian Diseases 1995;39:844–852.

9. Petkov DI, Linnemann EG, Kapczynski DR, Sellers HS. Identification and characterization of two distinct bursal B-cell subpopulations following infectious bursal disease virus infec-tion of white leghorn chickens. Avian Diseases 2009;53:347–355.

10. Hoerr FJ. Clinical aspects of immunosuppression in poultry. Avian Diseases 2010;54:2–15.

11. Perozo F, Villegas P, Fernandez R, Cruz J, Pritchard N. Efficacy of single dose recombinant herpesvirus of turkey infectious bursal disease virus (IBDV) vaccination against a variant IBDV strain. Avian Diseases 2009;53:624–628.

12. De Herdt P, Ducatelle R, Uyttebroek E, Sneep A, Torbeyns R. Significance of infectious bursal disease serology in an inte-grated quality control program under European epidemiologic conditions. Avian Diseases [Research Support, Non-U.S. Gov’t]. 2000;44:611–617.

13. Gelb JJ, Jackwood DJ, Mundt E, et al. Characterization of infec-tious bursal disease viruses isolated in 2007 from Delmarva commercial broiler chickens. Avian Diseases 2012;56:82–89.

14. Abdel-Alim GA, Saif YM. Detection and persistence of infectious bursal disease virus in specific-pathogen-free and commercial broiler chickens. Avian Diseases 2001;45:646–654.

15. Dormitorio TV, Giambrone JJ, Guo K, Jackwood DJ. Molecular and phenotypic characterization of infectious bursal disease virus isolates (Caracterización molecular y fenotipica de aisla-mientos del virus de la enfermedad infecciosa de la bolsa). Avian Diseases 2007;51:597–600.

16. Jackwood DJ, Sommer-Wagner SE. Molecular epidemiology of infectious bursal disease viruses: Distribution and genetic analysis of newly emerging viruses in the United States. Avian Diseases 2005;49:220–226.

17. Icard AH, Sellers HS, Mundt E. Detection of infectious bursal disease virus isolates with unknown antigenic properties by reverse genetics. Avian Diseases 2008;52:590–598.

18. Ojkic D, Martin E, Swinton J, Binnington B, Brash M. Genotyping of Canadian field strains of infectious bursal disease virus. Avian Pathol 2007;36:427–433.

19. Letzel T, Coulibaly F, Rey FA, et al. Molecular and structural bases for the antigenicity of VP2 of infectious bursal disease virus. J Virol 2007;81:12827–12835.

20. Calnek BW. Diseases of Poultry. 10th ed. Ames, Iowa, USA: Iowa State University Press, 1997.

21. McIlroy SG, Goodall EA, McCracken RM. Economic effects of subclinical infectious bursal disease on broiler production. Avian Pathol 1989;18:465–480.

22. Kurukulasuriya S, Ahmed KA, Ojkic D, et al. Circulating strains of variant infectious bursal disease virus may pose a challenge for antibiotic-free chicken farming in Canada. Research in Veterinary Science 2016;108:54–59.

23. Toro H, Santen VLv, Hoerr FJ, Breedlove C. Effects of chicken anemia virus and infectious bursal disease virus in commercial chickens. Avian Diseases 2009;53:94–102.


Article


I n t r o d u c t i o nBovine papillomavirus (BPV) is a nonenveloped, icosahedral

virus with a double-stranded circular DNA genome of approxi-mately 8000 base pairs (bp) containing 5 or 6 open reading frames (ORFs) that are expressed early during infection and 2 ORFs that are expressed late during infection. Because BPV causes tumor-like lesions in the skin and mucosa (1), it is the carcinogenic virus most used as a study model in research pertaining to cattle. These viruses are divided into 29 genera according to the L1 gene sequence (2) and 12 subtypes, which include the genera Deltapapillomavirus, Xipapillomavirus, and Epsilonpapillomavirus, the last being a nonclas-sified BPV-7 (2–4). This classification is based on the lesion (i.e., warts and/or tumors in the skin and mucosa) or a tumor associa-tion (5). Subtypes 1 and 2 have been more widely studied than the

other types, and only a few studies have reported the prevalence of the other virus types. Most molecular epidemiologic studies have demonstrated coinfection by 2 or more BPV types in the same sample (6,7).

Papillomaviruses are classically described as epitheliotropic. The lesions are mainly mucocutaneous warts and are in large numbers as a result of highly productive infection. These warts are often visible in the skin of the scalp, tongue, teats, penis, oral cavity, and upper digestive tract (8). Furthermore, the virus has been detected in dif-ferent bodily fluids, such as semen, urine, blood, and milk (8). The presence of the virus has also been reported in the bladder, placenta, and lymphocytes, which may play roles as viral reservoirs (9–11).

To date, there are no reports of a predilection of any viral subtype for animals of a specific gender or age. Therefore, it is possible to find different types of lesions in a single animal. Although it is

Detection and phylogenetic analysis of bovine papillomavirus in cutaneous warts in cattle in Tamaulipas, Mexico

Edith Rojas-Anaya, Antonio Cantú-Covarrubias, José Francisco Morales Álvarez, Elizabeth Loza-Rubio

A b s t r a c tPapillomas occur more frequently in cattle than other domestic animals. The causal agent of bovine papillomatosis is a virus that belongs to the family Papillomaviridae. In Tamaulipas, Mexico, the virus is considered a serious problem and has impeded the export of cattle to the United States, resulting in serious economic losses. Owing to the lack of information regarding the subtypes of papillomaviruses that infect cattle in Mexico, the aim of this study was to determine the subtypes in Tamaulipas. Fifty-two warts were analyzed with the use of polymerase chain reaction (PCR) involving primers that amplify the E7 gene of bovine papillomavirus (BPV). The PCR products were sequenced to differentiate the BPV-1 and BPV-2 subtypes. The sequencing quality was determined with the use of MEGA 6.0 software. Comparison of the Tamaulipas sequences with those of known BPV types by means of the MUSCLE algorithm showed that 53% of the former were BPV-1 and 47% were BPV-2. The distribution of the 2 subtypes in the cattle was homogeneous. This study demonstrated the presence of BPV-1 and BPV-2 in cattle from Tamaulipas and constitutes the first molecular characterization of papillomas in Mexico.

R é s u m éLes papillomes sont rencontrés plus fréquemment chez les bovins que chez n’importe quelle autre espèce domestiques. L’agent causal de la papillomatose bovine est un virus appartenant à la famille Papillomaviridae. Dans l’état mexicain de Tamaulipas le virus est considéré comme un problème sérieux et a empêché l’exportation de bovin vers les États-Unis d’Amérique, causant ainsi des pertes économiques importantes. Étant donné le manque d’information concernant les sous-types de papillomavirus qui infectent les bovins au Mexique, l’objectif de l’étude était de déterminer les sous-types présents dans l’état de Tamaulipas. Cinquante-deux verrues ont été analysées par réaction d’amplification en chaine par la polymérase (ACP) à l’aide d’amorces amplifiant le gène E7 du papillomavirus bovin (PVB). Les produits de l’ACP ont été séquencés afin de différencier les sous-types PVB-1 et PVB-2. La qualité du séquençage fut déterminée à l’aide du logiciel MEGA 6.0. La comparaison des séquences obtenues pour l’état de Tamaulipas avec celles des types connus de PVB par l’algorithme MUSCLE a permis de démontrer que 53 % étaient des PVB-1 et 47 % de PVB-2. La distribution des deux sous-types chez les bovins était homogène. La présente étude démontre la présence de PVB-1 et PVB-2 chez les bovins de Tamaulipas et constitue le premier rapport sur la caractérisation moléculaire des papillomes au Mexique.


Departamento de Biotecnología en Salud Animal, Centro Nacional de Investigación Disciplinaria en Microbiología Animal, Instituto Nacional de Investigaciones Forestales Agrícolas y Pecuarias, Km 15.5, Carretera México Toluca Cuajimalpa, DF, México ZC 05110 (Rojas-Anaya, Morales Álvarez, Loza-Rubio); Departamento de Salud Animal y Epidemiología, Campo Experimental Las Huastecas, Instituto Nacional de Investigaciones Forestales Agrícolas y Pecuarias, Km. 55, Carretera Tampico-Mante, Villa Cuauhtémoc, Tamaulipas, México ZC 89610 (Cantú-Covarrubias).

Address all correspondence to Dr. Elizabeth Loza-Rubio; e-mail: [email protected]; [email protected]

Received November 9, 2015. Accepted April 13, 2016.



generally thought that papillomaviruses are species-specific, BPV has recently been detected in various livestock and bison (12), water buffaloes (13), giraffes, zebras, and antelopes (14), yaks (15), horses and donkeys (16,17), tapirs (18), and other species. Papillomas can disappear or progress to malignant tumors, and their progression is associated with some specific viral subtypes as well as expression of the early (E) genes of the virus (19). Interaction of the E5, E6, and E7 oncoproteins with the host DNA is associated with the progres-sion of a viral infection to a tumor. The subtypes BPV-1, BPV-2, and BPV-4, which express these genes, are associated with malignant tumors (20). The L1 gene target is widely used for the molecular detection of BPV (21) because it is conserved and present in all papil-lomaviruses. Because the E7 gene encodes an oncoprotein, this gene

was used in this study as an indicator of the viral subtypes causing tumor-associated papillomas and to identify animals in which the infection might progress to a malignant neoplasm.

Bovine papillomatosis is widely distributed throughout the world. No study conducted in Mexico has indicated which papil-lomavirus subtypes are present in the country’s herds, even though the presence of the disease is known. Veterinarians have observed virus-associated lesions in cattle in the major agricultural regions, including Tamaulipas, which ranks third in terms of the national pro-duction of cattle, accounting for 16.38% (22). In fact, this state is one of the most important localities for cattle exports to the United States, and in recent years this virus has become an animal health problem that has caused considerable economic losses. However, there are

Figure 1. Location of the 4 sampled municipalities in the state of Tamaulipas, Mexico.



no epidemiologic data for bovine papillomatosis in Tamaulipas, and the only information known is that the incidence of this disease has increased considerably in Tamaulipas over the last 7 y. Thus, there is a strong need to identify which BPV subtypes are circulating within this geographic area. The purpose of this study was there-fore to determine the subtypes of BPV that are present in cattle in Tamaulipas and to analyze them phylogenetically.

M a t e r i a l s a n d m e t h o d sThis study constituted a descriptive and cross-sectional investiga-

tion with respect to the prevalence of BPV infection in the population or subgroups of animals sampled within the population at a given time. Nonprobabilistic convenience sampling was done (23), and the data were analyzed with use of the SPSS descriptive statistics package, version 17.0 (SPSS, Chicago, Illinois, USA).

Sample collectionBiopsy samples of skin warts approximately 0.5 to 1 cm in diam-

eter were collected from different areas of the bodies of 52 female and male cattle chronically affected with cutaneous papillomatosis in 2 regions of Tamaulipas: in the north, the municipalities of Gustavo Díaz Ordaz and Reynosa; and in the southeast, Soto La Marina and Aldama (Figure 1). Some samples were fixed in formalin and embed-ded in paraffin by routine methods. Sections from 3 different types of lesions were stained with hematoxylin and eosin.

The study protocols were approved by the Institutional Animal Care and Use Committee of the Centro Nacional de Investigación Disciplinaria en Microbiología Animal (approval 002-2014). All efforts were made to minimize animal suffering and distress accord-ing to the guidelines of Mexican Regulation NOM-033-ZOO-1995 for the euthanasia of domestic and wild animals.

DNA extractionA mechanical homogenizer (Qiagen, Hilden, Germany) was used

to extract DNA, which was purified with the QIAamp DNA Mini Kit (Qiagen) extraction kit according to the manufacturer’s recom-mended protocol with some modifications. Briefly, 100 mg of each wart sample in 200 mL of lysis buffer ATL and 20 mL of proteinase K was homogenized by vortexing. The mixture was incubated at 56°C for approximately 2 h. After the samples were centrifuged, 200 mL of buffer AL was added, and the samples were incubated at 70°C for 10 min. Ethanol (200 mL) was added, and the mixture was placed in a mini QIAamp spin column and centrifuged at 6000 3 g for 1 min. The column was washed with buffer AW1 and centrifuged as described previously. Finally, the DNA was eluted with 200 mL of elution buffer AE. The quantity and quality of the obtained DNA were determined spectrophotometrically. The DNA was then stored at 220°C until used.

Detection of the E7 geneThe E7 BPV gene (383 bp) was cloned as a positive control for

polymerase chain reaction (PCR). For detection of this gene the following primers were designed in house by amplifying a 203-bp fragment of the gene: E7BPV1, TAG GAA GCG AGG CWC AKA TA; and E7BPV−, CYC GMG GAC AAC ACA GG. The FastStart

PCR High-Fidelity PCR System (Roche, Mannheim, Germany) was used with the following: 5 mL of 103 PCR buffer with magnesium, 10 mM of each primer, 2.5 U of High-Fidelity Enzyme Blend, 200 mM of deoxynucleotide triphosphates, 2 mL of DNA, and water of molecular-biology grade. The samples were amplified in an iCycler (BioRad, Hercules, California, USA) with the following thermocy-cling protocol: 95°C for 5 min; 35 cycles of 95°C for 30 s, 57°C for 30 s, and 72°C for 30 s; and 72°C for 3 min. The amplification products were purified on 1.5% agarose gels with use of the QIAquick Gel Extraction Kit (Qiagen) according to the manufacturer’s instructions. The purified products were quantified and visualized on a 1.5% aga-rose gel to determine their quality and the purified DNA was then sent for sequencing to the Instituto de Biotecnología, Universidad Nacional Autónoma de México and the quality of sequences was evaluated with MEGA 6.0 software (24).

Phylogenetic analysisComparison and phylogenetic analyses were done with a 203-bp

fragment of the E7 gene. The nucleotide sequences of all BPV sub-types of different species and geographic origins obtained from GenBank (National Center for Biotechnology Information, Bethesda, Maryland, USA) were included in an alignment done with the algo-rithm MUSCLE (multiple-sequence comparison by log expectation), which involved calculating the distance between pairs of nucleotides (25). From the nucleotide sequences in this dataset, a phylogenetic tree was constructed by means of the maximum likelihood method with the JC 1 G 1 I model. A consensus tree was obtained from 1000 bootstrap replicates.

Data analysisThe case data were analyzed by nonparametric tests, with frequen-

cies used to determine the prevalence per herd and municipality, and a Chi-square test was conducted to determine statistical associations.

Re s u l t sThe prevalence of bovine papillomatosis in the cattle of

Tamaulipas (Table I) ranged from 2% to only 5% in the northeastern region, averaging 3% in Reynosa and 5% in Díaz Ordaz, whereas it ranged from 2% to 70% in the southeastern region, averaging 9% in Soto la Marina and 14% in Aldama. The virus had no predilection for

Table I. Prevalence of papillomatosis in breeding cattle in 2 regions of Tamaulipas, Mexico

Number (and %) of animals assessed Animals with chronic Region Total cutaneous papillomatosisNortheast Reynosa 220 6 (3) Díaz Ordaz 40 2 (5)

Southeast Aldama 764 111 (14) Soto La Marina 649 61 (9)



the gender or age of the animals, and the BPV-1 and BPV-2 subtypes were distributed homogeneously in the studied herds (Table II).

The animals presented a high degree of infection, showing several warts over the entire body. A cauliflower-like lesion was the most frequently found on different parts of the head, neck, flank, and foot. Histologic study of the cauliflower-like and flat warts showed altera-tions compatible with papillomavirus infection, such as hyperplasia of the epidermis, hyperkeratosis, and acanthosis (Figure 2).

The primers designed in-house to amplify the E7 gene were able to detect BPV subtypes 1 and 2 in the samples from 19 (36%) of the 52 animals: BPV-1 was detected in samples from 10 (53%) of

the 19 and BPV-2 in 9 (47%). Hyperplasia was present primarily in the animals infected with the BPV-2 subtype. The dendrogram obtained (Figure 3) indicated 2 different clades, corresponding to subtypes 1 and 2. It is noteworthy that the Mexican sequences clas-sified as subtype 1, which originated from the samples collected in Soto La Marina and Aldama, were grouped in a clade different from that containing the sequences classified as subtype 1 that were published in GenBank, which originated in Japan, the United States, Iraq, and Austria. No differences were found in the geographic distribution of the sequences grouped as subtype 2, which fell into 3 different subclades; these sequences corresponded to those of the samples obtained from Aldama, Soto La Marina, and Reynosa. The presence of both viral subtypes on the same farm was also dem-onstrated. The distance matrix showed little difference among the Mexican sequences but significant differences between the sequences of subtypes 1 and 2.

D i s c u s s i o nBefore this study, there was no published information regarding

the subtypes of BPV in any state of Mexico, despite the economic importance of bovine papillomatosis. We used PCR to assess the presence of BPV-1 and BPV-2 in samples of hyperplastic and mac-roscopic warts collected from chronically infected adult cattle in different herds in Tamaulipas, Mexico.

Of the 52 cattle evaluated by wart sampling, 19 (36%) were positive for 1 of 2 subtypes of BPV. Campo et al (26) detected BPV-2 DNA in natural tumors of immunosuppressed animals and demon-strated an association between BPV-2 and bovine bladder tumors. Borzacchiello et al (27) also demonstrated an association between

Table II. Features of the 19 cattle with wart samples positive for subtypes of bovine papillomavirus (BPV)

Age Lesion BPV Locality Gender Breed Herd (y) morphology Lesion location typeReynosa Female Charolais Beef 48 Cauliflower-like Flank 2

Díaz Ordaz Male Charolais Semental 16 Cauliflower-like Head, neck 1

Aldama Female Simbrah Beef 28 Cauliflower-like Head, neck 1 Female Simbrah Beef 29 Cauliflower-like Head, neck 2 Female Charolais/cebu Beef 14 Cauliflower-like Head, neck, foot 1 Male Simbrah Beef 16 Cauliflower-like Head, neck, flank 2 Female Simbrah Beef 14 Cauliflower-like Head, neck 2 Female Simbrah Beef 13 Cauliflower-like Head, neck 2 Female Simbrah Beef 11 Cauliflower-like Head, neck 2

Soto La Marina Female Charolais/cebu Beef 14 Cauliflower-like Head, neck, foot, abdomen 1 Female Charolais/cebu Beef 16 Cauliflower-like Head, neck, foot, foot 2 Female Charolais/cebu Beef 15 Cauliflower-like Head, neck, foot, foot 2 Female Simmental Beef 11 Cauliflower-like Head, neck, foot, foot 2 Female Simbrah Beef 40 Cauliflower-like Head, neck 1 Male Simbrah Beef 10 Cauliflower-like Head, neck, flank 1 Female Simbrah Beef 11 Cauliflower-like Head, neck, flank 1 Male Simbrah Beef 10 Cauliflower-like Head, neck, foot, flank 1 Female Simbrah Beef 11 Cauliflower-like Head, neck, foot, abdomen 1 Male Simbrah Beef 12 Cauliflower-like Head, neck, foot, flank 1

Figure 2. Hyperplasia of the epidermis, hyperkeratosis, and acanthosis in a wart sample, alterations compatible with papillomavirus infection. Hematoxylin and eosin; original magnification 1003.



BPV-2 and bovine bladder tumors and detected BPV-2 in samples with and without neoplastic change. Histopathological analysis of lesions is important for identifying intraepithelial tumors associated with oncogenic viruses such as BPV (28), and this analysis shows whether the virus has a predilection for anatomic areas related to a specific viral type. The histopathological findings reported in the literature for BPV include acanthosis, hyperkeratosis, parakeratosis, papillomatosis, and koilocytosis (29). We observed similar features in this study (Figure 2) and found that hyperplasia was present primarily in the animals infected with the BPV-2 subtype.

Both BPV-1 and BPV-2 have been reported to be associated with skin warts and fibropapillomas in cattle and sarcoids in horses. Owing to the presence of oncoproteins, BPV is genetically unstable.

The E5, E6, and E7 genes, which are related to the transformation of cells, are found in BPV-1. The function of each viral oncogene depends on the viral subtype. The E5 gene is highly conserved among the group of BPVs, is the main transforming gene (26), and induces cell transformation, proliferation, and survival (30). Not all oncogenes are present in all BPV subtypes, some subtypes being deficient in E6, regardless of how the infection occurs (31). In contrast, E7 is a zinc-binding protein, and E5 and E7 coexpression increases the transforming capacity of BPV (32). The oncogene E7 leads to cellular transformation, degradation of the pRb tumor sup-pressor, deregulation of the cell cycle, and tumorigenesis (29,33–35). We were interested to determine the presence of the E7 gene in the sampled warts because of this gene’s association with cellular

Figure 3. Maximum likelihood tree for bovine papillomavirus (BPV). The strains in blue are BPV-1, and the strains in red are BPV-2.



transformation. The E7 gene is important in oncogenesis, and inde-pendent of deletions or mutations its expression is related in some degree to tumor formation. Therefore, we selected this gene as the target in our molecular characterization study. It is more conserved and likely cannot be used to discriminate between some subtypes, whereas the E6 gene is not present in all subtypes.

The phylogenetic tree showed that the sequences of the detected BPV subtypes 1 and 2 were closest to those of North American strain type 1. Thus, these strains likely circulate between Mexico and the United States because of trade relations and may even infect wildlife species that can move freely across the border and have not been studied. The presence of both subtypes in different regions of Tamaulipas could indicate that the animals had been moved indis-criminately across the state. With respect to BPV-2, the tree showed that isolates from China and Mexico have a common ancestor (acces-sion no. M02219), but its origin is not specified in GenBank. This finding was corroborated by the distance matrix, which revealed a value less than 0.02 for the aforementioned isolate.

The main finding of this study is identification of the BPV sub-types in beef cattle with papillomatosis in Tamaulipas, Mexico. This identification, the first molecular characterization of BPV-1 and 2 in papillomas done in Mexico, is of epidemiologic importance because of the economic value related to the export of cattle from Tamaulipas to the United States: in 2014 alone, 138 094 cattle were exported (33). Biosecurity measures should be implemented in herd management to prevent or reduce the overall transmission of these subtypes among the animals.

A c k n o w l e d g m e n t sThis research was supported by the Fundación Produce-Tamaulipas

Project (grant 28-20120050) and Instituto Nacional de Investigaciones Forestales Agrícolas y Pecuarias (grant SIGI-94039336749).

Re f e r e n c e s1. Bocaneti F, Altamura G, Corteggio A, Velescu E, Roperto F,

Borzacchiello G. Bovine papillomavirus: New insights into an old disease. Transbound Emerg Dis 2016;63:14–23. Epub 24 March 2014.

2. Bernard HU, Burk RD, Chen Z, van Doorslaer K, Hausen H, de Villiers EM. Classification of papillomaviruses (PVs) based on 189 PV types and proposal of taxonomic amendments. Virology 2010;401:70–79.

3. Batista MV, Silva MA, Pontes NE, Freitas A. Molecular epidemi-ology of bovine papillomatosis and the identification of a puta-tive new virus type in Brazilian cattle. Vet J 2013;197:368–373.

4. Zhu W, Dong J, Shimizu E, et al. Characterization of novel bovine papillomavirus type 12 (BPV-12) causing epithelial papil-loma. Arch Virol 2012;157:85–91.

5. Borzacchiello G, Roperto F. Bovine papillomaviruses, papillomas and cancer in cattle. Vet Res 2008;39:45.

6. Carvalho CC, Batista MV, Silva MA, Balbino VQ, Freitas AC. Detection of bovine papillomavirus types, co-infection and a putative new BPV11 subtype in cattle. Transbound Emerg Dis 2012;59:441–447.

7. Schmitt M, Fiedler V, Müller M. Prevalence of BPV genotypes in a German cowshed determined by a novel multiplex BPV genotyping assay. J Virol Methods 2010;170:67–72.

8. Munday JS. Bovine papillomaviruses: A comparative review. Vet Pathol 2014;51:1063–1075.

9. Roperto S, Comazzi S, Ciusani E, et al. PBMCs are additional sites of productive infection of bovine papillomavirus type 2. J Gen Virol 2011;92:1787–1794.

10. Roperto S, Borzacchiello G, Esposito I, et al. Productive infection of bovine papillomavirus type 2 in the placenta of pregnant cows affected with urinary bladder tumors. PLoS One 2012;7:e33569. Epub 2012 Mar 27.

11. Roperto S, Russo V, Ozkul A, et al. Productive infection of bovine papillomavirus type 2 in the urothelial cells of naturally occur-ring urinary bladder tumors in cattle and water buffaloes. PLoS One 2013;8:e62227.

12. Tomita Y, Literak I, Ogawa T, Jin Z, Shirasawa H. Complete genomes and phylogenetic positions of bovine papillomavirus type 8 and a variant type from a European bison. Virus Genes 2007;35:243–249.

13. Maiolino P, Ozkul A, Sepici-Dincel A, et al. Bovine papilloma-virus type 2 infection and microscopic patterns of urothelial tumors of the urinary bladder in water buffaloes. Biomed Res Int 2013;2013:937918. Epub 2013 May 22.

14. Williams JH, van Dyk E, Nel PJ, et al. Pathology and immunohis-tochemistry of papillomavirus-associated cutaneous lesions in Cape mountain zebra, giraffe, sable antelope and African buffalo in Southern Africa. J S Afr Vet Assoc 2011;82:97–106.

15. Bam J, Kumar P, Leishangthem GD, Saikia A, Somvanshi R. Spontaneous cutaneous papillomatosis in yaks and detection and quantification of bovine papillomavirus-1 and -2. Transbound Emerg Dis 2013;60:475–480.

16. Lange CE, Tobler K, Ackermann M, Favrot C. Identification of two novel equine papillomavirus sequences suggests 3 genera in 1 cluster. Vet Microbiol 2011;149:85–90.

17. Nasir L, Brandt S. Papillomavirus associated diseases of the horse. Vet Microbiol 2013;167:159–167.

18. Kidney BA, Berrocal A. Sarcoids in two captive tapirs (Tapiricus bairdii): Clinical, pathological and molecular study. Vet Dermatol 2008;19:380–384. Epub 2008 Sep 24.

19. Munday JS, Thomson N, Dunowska M, Knight CG, Laurie RE, Hills S. Genomic characterisation of the feline sarcoid-associated papillomavirus and proposed classification as Bos taurus papil-lomavirus type 14. Vet Microbiol 2015;177:289–295.

20. Campo MS. Animal models of papilloma virus pathogenesis. Virus Res 2002;89:249–261.

21. Ogawa T, Tomita Y, Okada M, et al. Broad-spectrum detection of papillomaviruses in bovine teat papillomas and healthy teat skin. J Gen Virol 2004;85:2191–2197.

22. Ocupa Tamaulipas tercer lugar nacional en exportación de bovinos en pie. Benito Juárez, México: Secretaría de Agricultura, Ganadería, Desarrollo Rural, Pesca y Alimentación, 2015 Sept 23. Available from: http://www.sagarpa.gob.mx/Delegaciones/tamaulipas/Boletines/Paginas/2015B102.aspx Last accessed July 8, 2016.

23. Hernández R, Fernández C, Baptista P. Metodología de la inves-tigación. DF México, México: McGraw Hill, 2006:241–280.



24. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular Evolutionary Genetics Analysis Version 6.0. Mol Biol Evol 2013;30:2725–2729.

25. Edgar RC. MUSCLE: Multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 2004;32: 1792–1797.

26. Campo MS, Jarrett WFH, Barron R, O’Neil BW, Smith KT. Association of bovine papillomavirus type 2 and bracken fern with bladder cancer in cattle. Cancer Res 1992;52:6898–6904.

27. Borzacchiello G, Iovane G, Marcante ML, et al. Presence of bovine papillomavirus type 2 DNA and expression of the viral oncoprotein E5 in naturally occurring urinary bladder tumours in cows. J Gen Virol 2003;84:2921–2926.

28. Leto M, Santos-Júnior GF, Porro AM, Tomimori J. Human papil-lomavirus infection: Etiopathogenesis, molecular biology and clinical manifestations. An Bras Dermatol 2011;86:306–317.

29. Marins QS, Ferreira CE. Clinicopathologic and epithelial regres-sion study in cutaneous warts of bovines infected by papilloma-virus. Int Res J Pharm Pharmacol 2011;1:12–16.

30. Venuti A, Paolini F, Nasir L, et al. Papillomavirus E5: The small-est oncoprotein with many functions. Mol Cancer 2011;10:140.

31. Jackson ME, Pennie WD, McCaffery RE, Smith KT, Grindlay GJ, Campo MS. The B subgroup bovine papillomaviruses lack an identifiable E6 open reading frame. Mol Carcinog 1991;4:382–387.

32. Bohl J, Hull B, Vande Pol SB. Cooperative transformation and coexpression of bovine papillomavirus type 1 E5 and E7 pro-teins. J Virol 2001;75:513–521.

33. DeMasi J, Chao MC, Kumar AS, Howley PM. Bovine papiloma-virus E7 oncoprotein inhibits anoikis. J Virol 2007;81:9419–9425.

34. Freitas AC, Silva MAR, Jesus ALS, et al. Recent insights into bovine papillomavirus. Afr J Microbiol Res 2011;5:6004–6012.

35. Ocupa Tamaulipas primeros lugares en exportación de ganado bovino. Cd. Victoria, México: Gobierno del Estado de Tamaulipas, 2015 Jun 8. Available from: http://tamaulipas.gob.mx/2015/06/ocupa-tamaulipas-primeros-lugares-en-exporta cion-de-ganado-bovino/ Last accessed July 8, 2016.


Article


I n t r o d u c t i o nPorcine reproductive and respiratory syndrome virus (PRRSV)

can be devastating to breeding herds due to losses resulting from reproductive failure, i.e., reduced fertility, abortions, and premature farrowing (1–8), as well as delayed return to estrus (2,4). Piglets born to affected dams suffer from increased weakness at birth, decreased

growth rates, increased susceptibility to respiratory infections, and higher pre-weaning mortality (1,2,6,7,9). Vaccination before breeding is considered to be of value for decreasing these detrimental effects.

The 2 main types of commercially available vaccines, attenuated or modified live virus (MLV) and inactivated or killed virus (KV), confer various levels of protection. Several factors, including type of vaccine used (10), viral diversity (10,11), and timing (12,13), can

Safety and efficacy of a novel European vaccine for porcine reproductive and respiratory virus in bred gilts

Michael D. Piontkowski, Jeremy Kroll, Francois-Xavier Orveillon, Christian Kraft, Teresa Coll

A b s t r a c tPorcine reproductive and respiratory syndrome virus (PRRSV) can be devastating to commercial breeding operations. The objective of this study was to evaluate a novel European PRRSV vaccinal strain for safety and efficacy in bred gilts. In 2 experiments, 110 gilts were vaccinated intramuscularly and the vaccine was evaluated for safety and efficacy. Gilts in Experiment 1 were evaluated for local and systemic reactions and gilts in both experiments were observed for clinical signs of disease through farrow. In both experiments, piglet clinical observations, piglet average daily weight gain (ADWG), gilt serology [determined by enzyme-linked immunosorbent assay (ELISA)], gilt and piglet viremia [determined by quantitative real-time polymerase chain reaction (qPCR)], as well as piglet lung lesion scores and PRRS virus in lung tissue (qPCR) were determined. The vaccine was shown to be safe as there were no significant differences among groups in either experiment. Efficacy was established in Experiment 2 as both vaccinated groups were associated with desirable significant differences in percentage of gilts with abnormal clinical findings; gilt viral load post-challenge [day 125, day of farrowing (DOF), and DOF 1 13]; percentages of alive, healthy live, weak live, and mummified piglets per litter at farrowing and weaning; percentage of piglets per gilt that were positive for viremia; percentage of piglets per gilt with clinical disease; and piglet viral load on DOF. It was concluded that a vaccine formulated from the PRRSV modified live virus (MLV) strain 94881 is a safe and effective method of protection against the detrimental effects of virulent PRRSV infection in breeding female pigs.

R é s u m éLe virus du syndrome reproducteur et respiratoire porcin (VSRRP) peut être dévastateur pour les opérations de reproduction commerciales. L’objectif de la présente étude était d’évaluer l’innocuité et l’efficacité d’une nouvelle souche vaccinale européenne du VSRRP chez des cochettes saillies. Lors de deux expériences, 110 cochettes ont été vaccinées par voie intramusculaire afin d’évaluer l’innocuité et l’efficacité du vaccin. Les cochettes de l’Expérience 1 ont été évaluées pour la présence de réactions locales et systémiques et les cochettes dans les deux expériences ont été observées pour vérifier la présence de signes cliniques de maladie jusqu’au moment de la mise-bas. Lors des deux expériences on nota les observations cliniques des porcelets, le gain de poids quotidien moyen des porcelets (GMQ), les titres sérologiques des truies [déterminés par épreuve immuno-enzymatique (ELISA)], la virémie chez les cochettes et les porcelets [déterminées par réaction d’amplification en chaine par la polymérase quantitative (qACP)], ainsi que par les pointages de lésions pulmonaires des porcelets et la quantité de virus SRRP dans le tissu pulmonaire (qACP). Le vaccin s’est avéré sécuritaire et il n’y avait pas de différence significative entre les groupes dans les deux expériences. L’efficacité a été établie lors de l’Expérience 2 alors que les deux groupes vaccinés ont été associés à des différences significatives souhaitées dans le pourcentage de cochettes avec des trouvailles cliniques anormales; dans la charge virale post-challenge par cochette [jour 125, jour de la mise-bas (JMB), et JMB 1 13]; pourcentages de porcelets vivants, vivants en santé, vivants faibles, et momifiés par portée au moment de la mise-bas et au sevrage; pourcentages de porcelets par truie qui étaient positifs pour une virémie; pourcentage de porcelets par cochette avec une maladie clinique; et charge virale des porcelets au JMB. Il a été conclu qu’un vaccin formulé à partir de la souche 94881 du VSRRP vivant modifié est une méthode sécuritaire et efficace de protection contre les effets néfastes d’une infection par le VSRRP chez des porcs femelles de reproduction.


Boehringer Ingelheim Vetmedica Inc., 2621 North Belt Highway, St. Joseph, Missouri 64506, USA (Piontkowski, Kroll); Boehringer Ingelheim Vetmedica GmbH, Binger Street, 173, 55218 Ingelheim am Rhein, Germany (Orveillon); Boehringer Ingelheim Veterinary Research Center GmbH & Co. KG, Bemeroder Street, 31, 30559 Hannover, Germany (Kraft, Coll).

Address all correspondence to Dr. Jeremy Kroll; telephone: (515) 268-7412; e-mail: [email protected]

Received November 17, 2015. Accepted April 25, 2016.



directly affect the safety and efficacy of vaccinations. It is commonly accepted that inactivated vaccines are safe to administer to breed-ing animals (2,6,14) as they do not spread to other animals in the herd and do not induce reproductive failure in bred females (6) and are thus primarily recommended for immunizing animals used for reproduction.

Concerns have been raised about the safety of MLV vaccines due to the risk of vertical and horizontal transmission of vaccine virus (1,6,12,14–17) and the potential for reduced reproductive perfor-mance when administered during gestation (12). Despite safety concerns, MLV vaccines are commonly administered during gesta-tion as they have been reported to provide more protection against infection than inactivated vaccines when faced with heterologous challenge (11,16,17). It has been reported that inactivated vaccines did not prevent viremia and/or clinical signs of PRRS (2,6,17,18). Vaccination with an MLV vaccine, however, has been shown to pro-tect gilts from viremia and reduce congenital infection of piglets and pre- and post-natal death (2,6,16). In clinical trials, both attenuated (2,6,8,10,19) and inactivated (8,10,18) vaccines have been proven to provide effective protection against homologous challenge. Due to significant viral diversity, however, all challenge situations in the field could be considered heterologous (20).

Specifically, there are 2 main genotypes of PRRS virus, European (EU, Type 1) or North American (NA, Type 2) (3,8,15,19,21–25), which are further divided into multiple clusters (24). Since new strains continue to be found, including highly pathogenic strains (23), it is important that vaccines offer cross-protection against het-erologous strains. In addition to the type of vaccine used, timing of vaccine administration is also important.

The objective of this study was to evaluate the potential for safety and efficacy of a new PRRSV modified live virus (MLV) vaccine based on a novel strain [94881, US patent #8,765,142 B2, European Collection of Cell Cultures (ECACC) accession # ECACC 11012501 (parent strain) and ECACC 11012502 (attenuated strain)] (26) when administered to either bred gilts or to breeding-age gilts before conception and the subsequent heterologous challenge.


Animals and study designProtocols were reviewed and approved by the contract research

organization’s Institutional Animal Care and Use Committee before study initiation. Two experiments were conducted to establish the safety (Experiment 1) and efficacy (Experiment 2) of the test vac-cine in gilts. A total of 16 bred and 94 non-bred PRRSV-negative [enzyme-linked immunosorbent assay (ELISA) sample to positive (S/P) ratio of , 0.4], commercial mixed-breed, bred gilts were used for Experiment 1 and Experiment 2, respectively. For Experiment 1, a statistician randomly assigned gilts to either group 1A (n = 8) or 1B (n = 8) before day 0 and each group was then housed in 2 separate rooms. All gilts were 90 6 3 d of gestation at the time of first vaccination and were clinically healthy. On days 0 and 14, gilts in group 1B were administered the test vaccine at approximately 10 times (103) overdose, while gilts in group 1A received a placebo. Gilts in groups 1A and 1B were monitored for local and systemic

reactions, including serology and viremia testing, to establish vac-cine safety. All gilts in Experiment 1 subsequently farrowed on days 22 to 30 and piglets were monitored for number of piglets born live, dead (stillborn), weak, mummified, and crushed/mortality, as well as for viremia, average daily weight gain (ADWG), and lung pathology.

For Experiment 2, a statistician randomly assigned 94 healthy, non-bred, approximately 8-month-old, commercial mixed-breed gilts to 1 of 4 treatment groups (group 2A, n = 28; group 2B, n = 28; group 2C, n = 28; and group 2D, n = 10). Gilts in groups 2B and 2C were housed with their respective groups in 4 rooms (2 rooms per group) at the test facility for the duration of the trial, while gilts in groups 2A and 2D were housed together in a single room at an alternate facility. On day 0, approximately 28 d before breeding, gilts in groups 2A and 2D were administered a placebo product, while gilts in groups 2B and 2C were vaccinated with a low and high titer of the test vaccine, respectively. After vaccination, gilts were observed for local and systemic reactions. Gilts were tested for PRRSV serology and viremia, synchronized for breeding, artificially inseminated (AI), and evaluated for pregnancy by ultrasound. A total of 53 healthy bred gilts, [16 gilts each from groups 2A, 2B, and 2C and 5 gilts from group 2D (negative control group)] were randomly selected and subsequently enrolled into the challenge portion of the trial. Gilts in groups 2A to 2C were challenged with a heterologous [88.8% sequence homology at open reading frame (ORF) 5], Type 1 virulent strain of PRRSV and monitored for clinical signs of disease, including abortions and rectal temperatures. Gilts subsequently far-rowed on days 134 to 147 and piglets were monitored for number of piglets born live, dead (stillborn), weak, mummified, and crushed/mortality, as well as for viremia and ADWG. Additionally, all piglets that were either born dead or died before DOF 1 20 were necropsied and evaluated for lung pathology.

For the vaccination phase of Experiment 2, gilts in Groups 2B and 2C were housed by group in Biosafety Level 2 (BSL2) rooms at the test facility, while gilts in Groups 2A and 2D were housed together at an alternate facility until day 85 when the gilts were housed in rooms at the test facility. For Experiments 1 (entire trial) and 2 (beginning on day 85), bred gilts were housed in BSL2 rooms according to group and were placed in individual elevated crates. Crates were approximately 5 ft 3 7 ft in size, were equipped with a nipple waterer, feeder, and plastic slatted flooring that was elevated above the floor, and did not allow for nose-to-nose contact. For biosecurity purposes, each BSL2 room was separately ventilated with both HEPA filters and mechanical ventilation, ani-mal services staff were required to shower and don clean clothing before entering each room, and appropriate measures were taken to prevent accidental cross-contamination from vaccinates and/or challenged gilts to negative control gilts. All gilts were fed an age-appropriate, commercially available, non-medicated gestation or lactation ration (Heart of Iowa Cooperative, Roland, Iowa, USA) as appropriate for their condition. Food and water were available ad libitum.

Serology and viremia assaysBlood was collected by jugular venipuncture from gilts in

Experiments 1 and 2 on days 0, 14, DOF, and DOF 1 21 and on



days 0, 7, 14, 21, 56, 84, 118, 125, 132, DOF, DOF 1 7, DOF 1 13, and DOF 1 20, respectively. Venous whole blood was collected from piglets in Experiment 1 on the day of birth and DOF 1 21 and from piglets in Experiment 2 on the day of birth, DOF 1 7, DOF 1 13, and DOF 1 20, or when any piglet was found dead. Additionally, blood was collected from each mummified or stillborn piglet, but if this was not possible, thoracic or abdominal fluid was collected. Blood samples were processed for serum and fluids were aliquoted into appropriate tubes and held at either 2°C to 8°C or 270°C 6 10°C before testing. Samples held at 2°C to 8°C were tested for PRRSV anti-bodies at Boehringer Ingelheim Vetmedica Inc. Health Management Center, Ames, Iowa, USA (BIVI-Ames) using a commercially available ELISA kit (IDEXX Laboratories, Westbrook, Maine, USA). Results were reported as negative (ELISA S/P ratio of , 0.4) or positive (ELISA S/P ratio $ 0.4). Samples held at 270°C 6 10°C were tested for PRRSV ribonucleic acid (RNA) by quantitative real-time poly-merase chain reaction (qPCR; bioScreen GmbH, Münster, Germany). Results were reported as genome equivalent/mL (log10 GE/mL).

VaccinationGilts in both Experiment 1 and Experiment 2 were administered

either the test vaccine or the placebo at approximately 90 6 3 d of gestation and at 8 mo of age, respectively. In Experiment 1, gilts were vaccinated intramuscularly (IM) on the right side of the neck on day 0 and a repeat injection was administered on the left side of the neck on day 14. Gilts in group 1A were administered the placebo [phosphate-buffered saline (PBS)] and gilts in group 1B received a 103 overdose [approximately 1 3 108 50% Tissue Culture Infective Dose (TCID)50/dose] of the test vaccine (ReproCyc PRRS EU; Boehringer Ingelheim Vetmedica, St. Joseph, Missouri, USA). All gilts in Experiment 2 were vaccinated IM once on the right side of the neck. Specifically, gilts in group 2B (low titer) and 2C (high titer) were administered 2 mL of the test vaccine (ReproCyc PRRS EU; Boehringer Ingelheim Vetmedica) at the anticipated minimum immunizing dose (MID) of 1 3 102.43 (TCID)50/dose and at 1 3 103.90 TCID50/dose, respectively, while gilts in groups 2A and 2D received 2 mL of the placebo product (placebo matched product without PRRS 94881 MLV).

Post-vaccination observationsGilts in Experiment 1 were observed once daily for clinical signs of

disease including behavior (normal, recumbent, shivering, lethargic, or unconscious), respiration (normal, mild coughing, severe cough-ing, sneezing, abdominal breathing, or rapid respiration), digestion (normal, vomiting, diarrhea, and reduced or no appetite), and other (normal, hernia, thin, lame, edema around the eyes, etc.) on day 21 to their respective DOF 1 21. Similarly, gilts in Experiment 2 were observed once daily for clinical signs of disease from day 21 to 21 and at least 3 times weekly from day 22 to 115.

Injection site observationsFor gilts in Experiment 1, all injection sites were examined for

redness (none, slight, moderate, or severe), swelling (none, minimal, slight, moderate, or severe), heat, by feel, (normal, warm, or hot), and pain (absent or present during palpation) just before vaccina-tion and at 1 and 4 h post-vaccination on day 0 and once daily from

day 1 through day 14. All injection sites on the left side of the neck were monitored for signs of local reactions just before vaccination and at 1 and 4 h post-vaccination on day 14 and once daily from day 15 through day 28. Additionally, any gilt that exhibited an injection site reaction on the right and/or left side of the neck was observed until the reaction resolved or the end of the study 21 d post-farrow (DOF 1 21).

Estrus synchronization and breedingBeginning on day 8 and continuing through day 21, all gilts in

Experiment 2 were administered 6.8 mL of altrenogest (Matrix; Merck Animal Health, Whitehouse Station, New Jersey, USA) once daily on the feed for estrus synchronization purposes. On day 26 to 32, gilts were subsequently observed for estrus and were artificially inseminated a minimum of 2 times using semen obtained from a PRRS-negative boar farm upon determination of standing recep-tivity. Gilts were checked for pregnancy by ultrasound on day 84, approximately 55 6 3 d after breeding.

Animals removed from the trialNo animals were removed from Experiment 1, while 41 gilts

were excluded from Experiment 2 before challenge innoculation. Specifically, 5, 2, 3, and 1 gilt(s) from groups 2A, 2B, 2C, and 2D, respectively, did not display estrus after synchronization, were not bred, and were removed from the trial. After pregnancy was deter-mined, 16 gilts (2 from group 2A, 9 from group 2B, 4 from group 2C, and 1 from group 2D) were removed from the trial by day 89 as a result of lameness, not being pregnant, or late breeding. Additionally, 5, 1, and 5 gilts from groups 2A, 2B, and 2C, respectively, were ran-domly selected for removal by day 104, so that there was a total of 16 bred gilts in each of the challenged groups. Similarly, 3 gilts from group 2D were randomly selected for removal, so that the negative control group consisted of 5 bred gilts.

Challenge inoculationOn day 118, gilts in groups 2A, 2B, and 2C were challenged

intranasally (IN) with 4 mL (2 mL per nare) and IM with 2 mL of a heterologous Type-1 PRRSV. Briefly, on the day of the challenge, the challenge virus was thawed and diluted with minimum essential medium (MEM) to a targeted titer of 1 3 106 TCID50/6 mL dose. The challenge virus was isolated from an 8-week-old pig on a German farm with lungs infected by Type-1 PRRS that had previously shown signs of respiratory distress. Postmortem findings confirmed intersti-tial pneumonia and lung lesions suggestive of PRRSV. The challenge isolate was directly propagated on MA104 cells and a culture stock containing only Type-I PRRSV was developed for future challenge studies. Pre- and post-challenge titers were carried out and the challenge material was determined to have a mean titer of 1 3 106.30 TCID50/6 mL dose.

Post-challenge observationsGilts in Experiment 2 were observed once daily for clinical signs

of disease on study days 116 through DOF 1 20. Gilts were visually examined in crates and scored for clinical signs including respiration (0 = normal, 1 = panting/rapid respiration, 2 = dyspnea, or 3 = dead), behavior (0 = normal, 1 = mild-to-moderate lethargy, 2 = severely



lethargic or recumbent, or 3 = dead), and cough (0 = none, 1 = soft or intermittent, 2 = harsh or severe, repetitive, or 3 = dead). Observing personnel were blinded from treatment group assignment.

Rectal temperaturesRectal temperatures (°C) were collected from gilts in Experiment 1

once daily using a self-calibrating digital thermometer on days 21, 1 to 13, and 15 to 28. On vaccination days (days 0 and 14), rectal temperatures were recorded just before vaccination, as well as 4 h post-vaccination.

FarrowingGilts in Experiment 1 farrowed on day 22 to 30, while gilts in

Experiment 2 farrowed on days 134 to 147. On the day of parturi-tion, gilts in both experiments were administered oxytocin (Vet Tek, Blue Springs, Missouri, USA), according to label directions, to assist with farrowing as needed. On the day of birth, all live piglets in Experiments 1 and 2 were administered 1.0 mL of an iron injection (either Durvet, Blue Springs, Missouri, USA or Phoenix Scientific, St. Joseph, Missouri, USA) IM in the right ham to prevent iron defi-ciency anemia, as well as gentamicin (Sparhawk Laboratories, Lenexa, Kansas, USA), according to label directions, to prevent scours.

Reproductive performance of giltsFarrowing data was recorded for all gilts in Experiments 1 and 2.

On each gilt’s DOF (defined as the day the first piglet was deliv-ered), all piglets were classified into 1 of 5 categories: mummy, stillborn, weak, healthy, or crushed/mortality. A live piglet at birth was defined as any piglet that was healthy, weak, or crushed/mortality. The number of surviving piglets was determined on the last day of the trial (DOF 1 21 for Experiment 1 and DOF 1 20 for Experiment 2).

Observations of pigletsAll live piglets in both experiments were observed once daily for

clinical signs beginning on DOF 1 1 and continuing through to the end of the trial (DOF 1 1 to DOF 1 21 for Experiment 1 and DOF 1 1 to DOF 1 20 for Experiment 2). A daily total clinical observation score was determined as a summation of respiration, behavior, and cough scores, as previously described (post-challenge observations for gilts in Experiment 2).

Piglet average daily weight gainIndividual body weights (kg) of all piglets were collected on the

day of birth. Body weights were also collected on DOF 1 21 and DOF 1 20 for piglets in Experiment 1 and Experiment 2, respectively, or on the day a piglet was found dead. Average daily weight gain (ADWG) was determined for all surviving piglets from DOF through DOF 1 21 for Experiment 1 or DOF 1 20 for Experiment 2.

Postmortem examinationsAll gilts in Experiment 1 were euthanized by sedation and electro-

cution on DOF 1 21. If an injection site reaction (swelling) was still present, tissue samples were collected from the site and placed in a container with an appropriate amount of 10% formalin and submit-ted to the Iowa State University Veterinary Diagnostic Laboratory

(ISU VDL, Ames, Iowa, USA) for histopathologic examination (embedded in paraffin, sectioned, and stained with hematoxylin and eosin stain) by a board-certified veterinary pathologist who was blinded from animal treatments.

All surviving piglets in Experiment 1 were euthanized and nec-ropsied on DOF 1 21. The lungs were examined for gross lesions and the percent pathology for each lobe was recorded. Total lung lesion scores were determined for each pig using the European Union (EU) enzootic pneumonia formula. Specifically, total lung lesion scores were measured as a percentage of lung involvement calculated according to a weighting formula that accounts for the relative weight of each of the 7 lobes. The assessed percentage of lung lobe area with typical lesions was multiplied by the lobe fac-tor (left apical = 0.05, left cardiac = 0.06, left diaphragmatic = 0.29, right apical = 0.11, right cardiac = 0.10, right diaphragmatic = 0.34, and intermediate = 0.05) and the total weighted lung lesion score was then determined. Similarly, lung pathology was recorded for all stillborn piglets (excluding mummies), as well as piglets that died or were euthanized before study completion.

PRRSV quantitation in lung tissueFor all surviving piglets in Experiment 1 and all piglets dead

at delivery or that died or were euthanized before DOF 1 21 (Experiment 1) or DOF 1 20 (Experiment 2), 2 lung tissue samples were collected on DOF 1 21 or the day of death. One sample was placed in a Whirl-Pak bag (US Plastic Corporation, Lima, Ohio, USA) and the other in a container with 10% formalin. Samples in Whirl-Pak bags were stored at 270°C 6 10°C and then shipped to bioScreen GmbH for viremia testing for PRRSV RNA by qPCR as described by Revilla-Fernandez et al (27). Specifically, the 23 TaqMan Universal PCR Kit (Applied Biosystems, Foster City, California, USA) with AmpErase UNG (Applied Biosystems), the EU6-MGB: CTGTGAGAAAGCCCGGAC probe, and the primers EU6-343f-plus: GTRGAAAGTGCTGCAGGYCTCCA (sense) and EU6-462r-plus: CACGAGGCTCCGAAGYCCW (antisense) were used. Results were reported as log10 GE/mL for left and right/intermediate lung samples. Formalin-fixed samples were stored at room temperature for approximately 1 wk before being submitted to ISU VDL for embedding in paraffin blocks and subsequent storage, again at room temperature, for possible future testing.

Statistical analysesThe statistical analyses and data summaries were done using

SAS software, Version 8.2 (SAS Institute, Cary, North Carolina, USA). All data for both experiments were summarized descriptively (n = sample number, minimum, maximum, mean, median, interquar-tile range, or confidence interval) based on the type of variable. All data were analyzed assuming a completely random design structure and tests on differences were designed as 2-sided tests at a = 5%, with differences considered significant if P # 0.05.

The main objective of Experiment 1 was to compare the vaccinated group (group 1B) with the non-vaccinated group (group 1A), with primary variables of gilt reproductive performance and survival rate of offspring at weaning (DOF 1 21) and secondary, supportive parameters of systemic post-vaccination reactions, local injection site reactions, transplacental infection of piglets, and general health of



offspring during the suckling period (clinical observations, ADWG, and lung lesions).

Specifically, gilt reproductive performance and the survival rate of offspring, as well as gilt rectal temperatures, mean duration of injection site reactions, transplacental infection of piglets, and live piglets plus mummies and stillborn piglets that were PRRSV-positive by qPCR, viremia in piglets per litter, and piglet lung lesions, were analyzed using the Wilcoxon-Mann-Whitney test. Clinical observa-tions of gilts, proportion of gilts with an increase in rectal tempera-ture . 1.5°C compared to baseline days, proportion of gilts with an injection site reaction for at least 1 d, viremia post-vaccination in gilts, post-vaccination serology of gilts, and proportion of piglets per litter with a positive clinical observation score for at least 1 d were evaluated using Fisher’s exact test. Finally, mean daily rectal temperature of gilts, mean piglet body weight per litter on DOF, and mean piglet ADWG per litter from DOF through DOF 1 21 were compared using the analysis of variance (ANOVA) procedure.

The main objective of Experiment 2 was to compare the 2 vac-cinated groups (groups 2B and 2C) to the unvaccinated challenge control group (group 2A). Animals removed from the trial were included in their respective parameter of analysis until removal. Primary variables included proportions of live piglets on day of farrowing (DOF) and proportions of live piglets at 20 d of age (DOF 1 20), with supportive variables consisting of clinical obser-vations post-vaccination and post-challenge, serology, viremia, and

reproductive performance of the gilts, as well as viremia, clinical observations, and ADWG for the piglets.

Specifically, frequency tables were generated of gilts with at least 1 positive finding for clinical observations post-vaccination and post-challenge, for positive ELISA results, and for positive qPCR (qualitative analysis) of gilt viremia and differences between the challenge control and vaccine groups were tested using Fisher’s exact test. Gilt viremia data were also evaluated (quantitative analy-sis) separately for each day and positive versus negative (assigned values of 3.0 and 0.0 log10 GE/mL, respectively) were compared using the Wilcoxon-Mann-Whitney test. Differences in gilt reproduc-tive performance between groups were also tested by the Wilcoxon-Mann-Whitney test.

Piglet viremia data were evaluated qualitatively (calculated percentages of positive piglets per litter used as single values for comparison of groups) and quantitatively (calculated median qPCR values per litter used as single values for comparison of groups) for separate comparisons for each day using the Wilcoxon-Mann-Whitney test. Individual piglet qPCR data were used for summary statistics and viral load in lung samples from dead piglets before DOF 1 20 were evaluated descriptively only. Individual piglet ADWGs were calculated for between DOF and DOF 1 20 and dif-ferences among treatment groups were analyzed using the ANOVA procedure and subsequent t-tests. Least squares means of groups and differences between LS means with 95% confidence intervals

Figure 1. Mean rectal temperature of gilts in Experiment 1 after repeat vaccination with either the placebo (group 1A) or 103 overdose of the test vaccine (group 1B). Mean rectal temperatures (°C) in group 1A were significantly lower than in group 1B on day 0 (before vaccination), 6, and 14 1 4 h (P = 0.0003, P = 0.0220, and P = 0.0358, respectively). These differences were not biologically relevant (temperatures not above 40.0°C). No other significant differences were noted.* Significant difference between groups.

Study day

Gilt mean rectal temperatures°C

39.5

39.0

38.5

38.0

37.5

37.0

Group 1A

Group 1B

921 0

09 1

4h 1 2 3 4 5 6 7 8 9 1 0 11 12 13 14

149

1 4

h 15 16 17 18 19 20 21 22 23 24 25 26 27 28



(CI) were derived from the ANOVA. The analysis for DOF 1 20 and ADWG was repeated with weight at DOF as a covariate and weight data of piglets per gilt were summarized descriptively. The calculated percentages of piglets per litter with at least 1 positive finding for clinical observations on DOF 1 1 to DOF 1 20 were used as single values for comparisons among groups by the Wilcoxon-Mann-Whitney test and were analyzed assuming a completely random design structure.

Re s u l t s

Post-vaccination observationsIn Experiment 1, no gilt mortalities occurred in either group 1A

or 1B (placebo-vaccinated or repeat 103 overdose, respectively). From day 0 1 4 h to DOF 1 21, 50% of gilts in group 1A and 75% of gilts in group 1B exhibited at least 1 abnormal clinical finding for a minimum of 1 d (difference between groups was not significant; P = 0.6084). Specifically, 4 gilts in group 1A and 6 gilts in group 1B exhibited abnormal clinical findings. Two gilts had reduced or no appetite for 2 d and 1 gilt was lethargic for 1 d; 1 gilt had reduced or no appetite for 2 d and was lethargic for 2 consecutive d; 1 gilt had rapid respiration, reduced or no appetite for 1 d each, and was lethargic on 2 consecutive days; and 1 gilt had a small abscess on the

right side of the neck (noted before vaccination) for 28 d and other signs (no description other than farrowing) on 1 d. No analysis was conducted on gilts in Experiment 2 as no abnormalities were noted after vaccination.

Table I. Group summary of number of live piglets per litter at weaning. No significant differences were noted between groups 1A and 1B, while groups 2B and 2C had significantly higher numbers of live piglets at weaning than group 2A (P = 0.0063 and 0.0026, respectively)

Number of live piglets per litter at weaningGroup Minimum Maximum Median Mean P-value1Aa 4 12 9.5 8.6 1Ba 1 13 8.5 7.8

0.6937

2Ab 0 10 2.0 2.9 N/A2Bb 1 10 6.5 6.2 0.0063c

2Cb 2 12 7.5 6.9 0.0026c

2Db 9 14 10.0 10.8 N/Aa DOF 1 21.b DOF 1 20.c Compared to group 2A.DOF — Day of farrowing; N/A — Not applicable.

Figure 2. Summary of group reproductive performance results (number of piglets per litter) on day of farrowing (DOF) and summary by group of mean percentage of piglets per litter in each category at farrowing. There were significant differences between groups 2B and 2A for mean percentage of live, healthy live, and mummified piglets at birth (P = 0.0184, P = 0.0138, and P = 0.0190, respectively) and between groups 2C and 2A for mean percentage of live, healthy live, weak live, and mummified piglets.* Significant difference among groups compared to group 2A.** Negative control group; no analysis conducted.

Live Healthy Weak Stillborn Mummy Crushed live live mortality

Mean percentage of piglets in each category on DOFM

ean

perc

enta

ge o

f pi

glet

s

100

90

80

70

60

50

40

30

20

10

0

Group 2A

Group 2B

Group 2C

Group 2D**

Group 1A

Group 1B



Injection site observationsIn Experiment 1, 1 gilt from each group exhibited an injection site

reaction for at least 1 d until day 14 (P = 1.0000), whereas no gilts in group 1A, but 7 out of 8 gilts in group 1B exhibited an injection site reaction of either redness, heat, and/or pain for at least 1 d from day 14 onwards (P = 0.0014). From day 0 1 1 h to day 14, gilts in groups 1A and 1B had median injection site reaction duration of 0.0 and 2.0 d, respectively. Swelling was measurable for only 1 gilt in group 1B (slight swelling on the right side on day 1 to day 4 and on day 17 and moderate swelling on the right side on days 18 to 23 and days 41 to 43), which was later determined to be an abscess. Swelling of the 1 gilt in group 1A (day 0 1 4 to day 1) and the 6 gilts in group 1B was found to be minimal.

Post-challenge observationsIn Experiment 2, percentages of total abnormal findings (a clinical

score . 0 for respiration, behavior, and/or cough indicating clinical disease) for at least 1 d for gilts in groups 2A (placebo-vaccinated and challenged), 2B [low-titer (MID) vaccinated and challenged], 2C (high-titer vaccinated and challenged), and 2D (placebo-vaccinated and not challenged) were 25.0%, 25.0%, 38.0%, and 60.0% from day 116 (2 d before challenge) to DOF 1 20, respectively. No sig-nificant differences in frequency of gilts positive for clinical disease were detected among vaccinated groups and the challenge control group during this time period (P $ 0.7043).

Rectal temperaturesIn Experiment 1, mean rectal temperature for gilts was within

the normal limits throughout the study and ranged from 37.9°C to 38.8°C in group 1A and from 38.0°C to 39.2°C in group 1B (Figure 1).

Reproductive performance of giltsNo significant differences were found between groups 1A and 1B

for median percentages of live piglets (healthy 1 weak 1 crushed/mortality) per litter at farrowing (89.0% and 91.7%, respectively; P = 1.0000). Median percentages of healthy live piglets per litter

(83.8% and 89.3%, respectively; P = 0.9262) and median percentages of weak live, stillborn, mummified, and crushed/mortality piglets per litter were 0.0% for all parameters for both groups (P $ 0.8825). There was no significant difference between groups for median percentages of live piglets at the time of weaning (100.0% and 90.5% for groups 1A and 1B, respectively; P = 0.2103).

Groups 2B and 2C had significantly higher percentages of live and healthy live piglets per litter (P # 0.0455) than group 2A. Group 2C had significantly lower percentages of weak live piglets per litter (P = 0.0024), groups 2B and 2C had significantly lower percentages of mummies per litter (P # 0.0190), and there were no significant dif-ferences among groups for the percentages of stillborn or crushed/mortality piglets per litter (P $ 0.1965). Median percentages of live piglets per litter at weaning were significantly higher in groups 2B and 2C than in group 2A (P = 0.0203 and P = 0.0022, respectively). The number of piglets born in each category per group is summa-rized in Figure 2, while the number of piglets weaned per group is shown in Table I.

Piglet clinical observationsThere was no significant difference between groups 1A and 1B for

median percentages of piglets per litter that were positive for clinical disease for at least 1 d (8.7% and 10.6%, respectively; P = 0.9814). In Experiment 2, median percentages of piglets per litter that were positive for clinical disease for at least 1 d were significantly lower in groups 2B and 2C than in group 2A (25.0% and 25.0% versus 100.0%, respectively; P # 0.0001).

Average daily weight gainFor piglets in both Experiments 1 and 2, there were no significant

differences in body weight among groups on DOF (P = 0.6766 and P $ 0.2972, respectively). Specifically, least squares mean body weights for piglets in groups 1A and 1B were 1.55 and 1.50 kg, respectively, and the mean body weight for piglets in groups 2A, 2B, 2C, and 2D were 1.34, 1.43, 1.40, and 1.39 kg, respectively. No significant difference in least squares means for ADWG was detected between groups 1A and 1B from DOF to DOF 1 21 (222 g/d and 226 g/d, respectively; P = 0.8760). Conversely, piglets in groups 2B and 2C had significantly higher least squares mean ADWG than those in group 2A from DOF to DOF 1 20 (194 g/d and 197 g/d versus 130 g/d, respectively; P # 0.0028) when weight on DOF was used as a covariate for analysis (Figure 3).

PRRSV viremia testingNo PRRSV RNA was detected in the serum of any gilt in either

experiment on day 0. In Experiment 1, all gilts in group 1A remained negative for the duration of the experiment. After vaccination (day 0), 2 out of 8 (25%) of gilts in group 1B were positive for PRRSV RNA by qPCR on day 14 (difference not significant; P = 0.4667), whereas no gilts in either group were positive on DOF and DOF 1 21. In Experiment 2, all gilts in both groups 2A and 2D remained negative until the day of challenge and for the duration of the experiment, except for 1 gilt in group 2D that was positive on DOF 1 7 (tested negative at all other time points).

After vaccination (day 0), 50% of gilts in groups 2B and 2C were PRRSV RNA positive, whereas 36% were PRRSV RNA positive on

Figure 3. Mean body weight of piglets in Experiments 1 and 2 on day of farrowing (DOF) and DOF 1 21 d (Experiment 1) or DOF 1 20 d (Experiment 2). There were no significant differences in mean piglet body weight between groups 1A and 1B in Experiment 1 and among groups 2A to 2D in Experiment 2 on DOF. Piglets in Groups 2B and 2C weighed significantly more than piglets in group 2A on DOF 1 20 (P # 0.0028).* Significant difference among groups compared to group 2A.** Negative control group; no analysis conducted.

DOF

DOF 1 21

DOF 1 20

Piglet mean body weights

1A 1B 2A 2B 2C 2D**

Group

Bod

y w

eigh

t (k

g)

7

6

5

4

3

2

1

0



day 7. From days 14 to 56, only 4% of gilts in group 2B remained positive, while up to 4% of gilts in group 2C were intermittently qPCR positive. All gilts were qPCR negative for PRRSV RNA on day 84 and day 118 (day of challenge). After the challenge, groups 2B and/or 2C had significantly lower percentages of positive gilts than group 2A on day 125 (7 days post challenge; DPC) (31.0% and 25.0% versus 100.0%, respectively; P # 0.0001), day 132 (14 DPC) [19.0% (P = 0.0290) and 31.0% (P = 0.1556) versus 63.0, respectively], DOF (25.0% and 6.0% versus 94.0%, respectively; P , 0.0002), and DOF 1 13 (0.0% and 6.0% versus 47.0%, respectively; P # 0.0155). No significant differences in percentages of positive gilts were detected between the vaccinated groups and the challenge control group on DOF 1 7 and DOF 1 20 (P $ 0.1719) (Figure 4).

No significant differences between groups in Experiment 1 were found in median percentages of piglets that were positive for PRRSV RNA on the day of farrowing. Specifically, median percentages of zero of pre-colostral blood samples from live piglets per litter from groups 1A and 1B were qPCR positive for PRRSV RNA on DOF (P = 0.4615). When blood/fluid samples from piglets born dead were included in the analysis, however, median percentages of positive piglets were 0.0% (group 1A) and 16.7% (group 1B) (P = 0.0769). Likewise, no significant differences were noted between live piglets in groups 1A and 1B on DOF 1 14 (0.0% and 34.6%, respectively; P = 0.0769) and DOF 1 21 (0.0% and 37.5%, respectively; P = 0.0769).

Conversely, significant differences in median percentages of piglets positive for PRRSV RNA were found on multiple occasions

when all piglets (alive or dead) born to gilts in Experiment 2 were evaluated. Specifically, groups 2B and 2C had significantly lower median percentages of positive piglets than group 2A (P = 0.0381 and P = 0.0018, respectively) on DOF. Similarly, groups 2B and 2C had significantly lower median percentages of positive piglets than group 2A on DOF 1 7 (P # 0.0293). On DOF 1 13, however, only group 2B had a significantly lower median percentage of positive piglets than group 2A (P = 0.0216), while the difference between groups 2C and 2A was not significant (P = 0.0860). No significant differences were detected among groups for percentages of positive piglets on DOF 1 20 (P $ 0.0614) (Table II).

PRRSV serologic testingAll gilts in Experiment 1 were seronegative for PRRSV by ELISA

testing on day 0. After vaccination, 75% of gilts in group 1B were shown to have seroconverted by day 14 and 100% by DOF. Gilts in group 1A remained seronegative throughout the study. In Experiment 2, all gilts were seronegative for PRRSV by ELISA test-ing on day 0 and day 7. All gilts in groups 2A and 2D remained seronegative up to and including the day of challenge (day 118) and until the conclusion of the trial (DOF 1 20), respectively. Conversely, on day 14, the percentage of PRRSV ELISA positive gilts in group 2B was 18% and in group 2C was 21.0%, which increased to 65.0% in 2B and 60.0% in 2C on day 56. On the day of challenge (day 118), percentages of seropositive gilts in groups 2B (56.0%) and 2C (50.0%) were decreasing. On day 125, percentages of seropositive

Figure 4. Summary of qPCR results for gilts (log10 GE/mL) from day 0 through day of farrowing (DOF) 1 20. In Experiment 2, gilts in groups 2B and 2C had significantly higher PRRS viral load than gilts in group 2A on day 7 (P = , 0.0001 and P = 0.0007, respectively). Results of qPCR were significantly lower for both groups on day 125, DOF, and DOF 1 13 d (P = # 0.0001, P = # 0.0002, and P = # 0.0155, respec-tively) and for group 2B on day 132 (P = 0.0230) compared to group 2A.* Significant difference among groups compared to group 2A.** Negative control group; no analysis conducted.qPCR — Quantitative real-time polymerase chain reaction.

Study day

Gilt mean percentage of positive qPCR results

Mea

n pe

rcen

tage

of

posi

tive

qP

CR

res

ults

100

90

80

70

60

50

40

30

20

10

0

Group 2A

Group 2B

Group 2C

Group 2D**

0 7 14 21 56 84 118 125 132 DOF DOF DOF DOF 17 113 120



gilts in groups 2B and 2C were higher than in group 2A (88.0% and 100.0% versus 6.0%, respectively). On day 132, all remaining gilts in groups 2A, 2B, and 2C were PRRSV seropositive through to the conclusion of the trial (DOF 1 20).

Postmortem examinationsOn DOF 1 21 (Experiment 1), persistent swelling in the right

neck region of 1 gilt in group 1B was evaluated. Necropsy and histopathologic examination revealed a well-encapsulated abscess approximately 3 cm in size. There were no other notable findings for any gilt in Experiment 1.

One piglet in group 2A died after blood collection and death was confirmed at necropsy to be due to a lacerated jugular vein and secondary blood loss. In Experiment 1, all piglets in group 1A that were necropsied on DOF 1 21 had no lung pathology in any lobe, whereas 3 out of 68 piglets in group 1B had minor lung lobe pathology (total lung lesion scores of # 1.85%; lesions described as very small areas of lung consolidation likely due to a past bacterial infection) and 1 piglet had 100% pathology in each lung lobe (total lung lesion score of 100%; lesions described as pneumonia and pleuritis likely due a bacterial infection). There were no significant differences among groups for median percentages of piglets per litter with lung lesion scores (P = 0.2000) and for median total lung lesion scores (P = 0.0531) (Table III).

PRRSV quantitation in lung tissueThere was no significant difference between groups 1A and 1B

in median percentages of piglets per litter that were positive for PRRSV RNA in the lung tissues by qPCR (0.0% and 10.0%, respec-tively; P = 0.0769). In Experiment 2, of the piglets that were either born dead, died, or were euthanized before DOF 1 20, the mean lung qPCR results were 4.676, 4.092, 3.547, and 0.000 log10 GE/mL for groups 2A, 2B, 2C, and 2D, respectively (Table IV) (no statistical analysis conducted).

D i s c u s s i o nVaccinating gilts and sows with a PRRS MLV vaccine is beneficial

for several reasons, including increased farrowing and weaning rates and decreased number of premature farrowings (1,4,7,8). Vaccination of dams also has positive effects on the piglets, including increased ADWG and decreased clinical signs (12). The effectiveness of a PRRS MLV vaccine in improving reproductive performance is vital because loss from piglet deaths, for whatever reason (mummified, stillborn, and/or disease-related), results in fewer piglets weaned and there-fore dramatically affects the economic productivity of the herd. A PRRS MLV vaccine must be safe to administer, however, in that it does not cause reproductive failure, local or systemic reactions, and negatively affect piglet survivability and growth performance.

In Experiment 1, primary criteria for evaluating safety included reproductive performance at farrowing and survival rate at wean-ing, with supportive parameters of local and/or systemic reactions to vaccination and transplacental infection, as well as general health of piglets (clinical observations and growth performance) during the suckling period.

In Experiment 2, effectiveness was primarily determined by evalu-ating reproductive performance at farrowing (number of live-born piglets) and the number of piglets at weaning. This was supported by evaluating gilts for clinical assessments post-vaccination and post-challenge, PRRSV serology, and viremia, as well as evaluation for total number of piglets, healthy live piglets, weak live piglets, mummies, stillborn piglets, and crushed/mortality piglets born per litter. Piglets were assessed for viremia, clinical observations, and ADWG. According to these criteria, the present data clearly demon-strated that vaccination of gilts with the novel PRRSV strain 94881 MLV vaccine (ReproCyc PRRS EU; Boehringer Ingelheim Vetmedica) is a safe option for preventing reproductive losses associated with the PRRS virus.

Specifically, PRRS-naive gilts that were given a repeat 103 over-dose did not exhibit any relevant differences in median percentages of live piglets born or live piglets at weaning compared to non-vaccinated placebo controls. During the suckling period, there were no significant differences between the vaccinated and non-vaccinated groups for systemic reactions to vaccination (post-vaccination observations), gilt viremia, and general health of piglets (clinical observations, least squares mean body weight at birth, and ADWG). There were no significant differences between groups for injection site observations after the initial 103 overdose and no biologically relevant differences in rectal temperatures after vaccination (did not exceed 40.0°C).

Table II. Group summary of percentages of piglets that were positive for PRRSV RNA as determined by qPCR. Groups 2B and 2C had significantly lower mean percentages of positive piglets than group 2A on day of farrowing (DOF) and DOF 1 7 (P # 0.0381 and P # 0.0175, respectively), while on DOF 1 13, group 2B had significantly lower percentages of positive piglets than group 2A (P = 0.0216)

Mean percentages of positive pigletsDay Group Mean P-valueDOF 2A 86.3 2B 58.1 0.0381a

2C 55.0 0.0018a

2D 0.0

DOF 1 7 2A 100.0 2B 76.6 0.0293a

2C 78.6 0.0175a

2D 0.0

DOF 1 13 2A 100.0 2B 75.4 0.0216a

2C 84.0 0.0860a

2D 0.0

DOF 1 20 2A 90.9 2B 75.3 0.0614a

2C 81.6 0.183a

2D 0.0a Compared to group 2A.DOF — Day of farrowing.



In Experiment 2, the vaccine also met the necessary requirements for efficacy as significantly higher percentages of live pigs were born and weaned in the vaccinated group than in the non-vaccinated chal-lenge control group. There were significant beneficial differences in 1 or both vaccinated groups compared to the challenge control group for percentages of healthy live piglets, weak live piglets, and mum-mified piglets born per litter. Results of gilt post-vaccination clinical observation were in alignment with the safety study, as there were no abnormal assessments for any group from day 1 through day 21 and no significant difference among groups for abnormal clinical assessments for at least 1 d from day 1 through day 113.

Similarly, there were no significant differences in gilt clinical observations post-challenge between the low-titer and high-titer vac-cinated groups compared to the placebo-vaccinated and challenged group. Interestingly enough, however, a higher percentage of gilts in the negative control group (placebo-vaccinated and not challenged) exhibited a clinical observation for at least 1 d from day 116 through DOF 1 20. Vaccinated gilts were also shown to have seroconverted as early as 14 d post-vaccination. Vaccinated gilts experienced a brief period of viremia after treatment was administered, as evidenced by positive gilts on day 7. Differences in viremia between vaccinated groups (groups 2B and 2C) were not significant from days 14 through 56, however, and all gilts were negative on day 84 and day 118.

Conversely, there were significantly lower percentages of qPCR-positive gilts in the vaccinated groups than in the challenge control group after the challenge [on days 125, 132 (low dose only), DOF, and DOF 1 13]. Piglets born to vaccinated gilts also exhibited favor-able results compared to piglets born to challenge control gilts, as 1 or both vaccinated groups had significantly lower percentages of piglets per litter that were positive for abnormal clinical findings for at least 1 d from DOF 1 1 through DOF 1 20, significantly higher

ADWG, and significantly lower percentages of piglets per litter that were viremic on DOF, DOF 1 7, and DOF 1 13. These results clearly indicate that even a 103 overdose of the vaccine was safe to adminis-ter during gestation and effectively prevented the detrimental effects of PRRSV infection in both gilts and piglets born to vaccinated gilts.

While a number of studies have been published describing the safety of NA-type PRRS modified live virus (MLV) vaccines in bred females, less information is available about the safety of EU-type MLV vaccines. One trial examined the safety of a single dose of a commercially available EU-type MLV PRRSV vaccine in gilts and sows that were in either early or late pregnancy and in lactating sows (7). In the trial, safety was established according to the lack of significant differences in local or systemic reactions between vacci-nated and non-vaccinated animals. Additionally, fluid and/or tissue samples collected from aborted piglets were negative for PRRSV RNA. These results are similar to the present safety evaluation data, as vaccinates exhibited neither significantly increased incidence of local or systemic reactions after the first vaccination and there were no significant differences between groups for piglet viremia on DOF, even though a 103 overdose of the vaccine had been administered.

A number of studies have examined the effects of the use of EU-type MLV vaccines in endemic situations. The effects of an EU-type MLV PRRS vaccine were evaluated in lactating sows and pregnant gilts and sows on 3 Polish farms with histories of chronic PRRSV infection (7). The vaccine was found to be beneficial as live piglets born and piglets weaned increased significantly compared to non-vaccinated animals. When the vaccine was administered to the remainder of the breeding females, regardless of pregnancy status, there were significant favorable differences in farrowing rate, live-born piglets, stillborns, weaned pigs, and abortions compared to pre-vaccination data for the same animals.

Table III. Summary of percentages of piglets per litter with lung lesions and total lung lesion scores per group. There were no significant differences between groups 1A and 1B in either category

Piglet lung lesions Group n Min Max Mean P-value% of piglets per litter 1A 8 0.0 0.0 0.00 1B 8 0.0 16.7 4.89 0.2000

Total lung lesion scores 1A 72 0.0 0.0 0.00 1B 68 0.0 100.0 1.52 0.0531

Table IV. Summary of qPCR results (log10 GE/mL) for PRRSV in lungs of piglets by group. In Experiment 2, although no statistical analysis was conducted to compare qPCR results of piglets that were born dead or died before day of farrowing (DOF) 1 20, the median and mean lung qPCR results were lower in dead piglets in groups 2B and 2C than those in group 2A

Piglet lung qPCRGroup N Min Max Median 95% CI Q Range Mean2A 141 0.00 7.95 5.140 4.810 5.390 2.990 4.6762B 79 0.00 7.45 4.780 3.000 5.260 2.620 4.0922C 75 0.00 6.84 4.220 3.000 5.100 5.620 3.5472D 4 0.00 0.00 0.000 0.000 0.000 0.000 0.000qPCR — Quantitative real-time polymerase chain reaction.



In a study conducted on an endemically PRRSV-infected farrow-to-finish farm in Greece, a group of gilts and sows were administered a single dose of an EU-type MLV vaccine at approximately 6 mo of age and 10 d after farrowing, respectively, while an additional group of gilts and sows were left untreated (1). The vaccinated group had significantly improved farrowing rates, higher numbers of live-born piglets, fewer dead or mummified piglets, and more weaned pigs (average of 0.7 piglets per litter) than the non-vaccinated group.

Another study described significantly improved farrowing rates, including higher numbers of live-born and fewer numbers of mum-mified and stillborn piglets per litter, when pigs on PRRSV-positive farms in Thailand were treated with either an EU-type or NA-type PRRSV MLV vaccine (4). Unfortunately, the data were reported as a generalization of vaccinates versus non-vaccinates (no groups based on timing of vaccination or type of PRRSV vaccination received). The authors did note, however, that vaccination improved fertility rate in gilts by a greater percentage (7.3%) than in sows (6.7%, 3.8%, 2.0%, and 0.4% in parity numbers 1, 2 to 3, 4 to 5, and $ 6, respectively).

While results similar to the present data [significantly improved reproductive performance at farrowing (more live and healthy live piglets born and fewer weak live piglets and mummies born) and more live pigs at weaning (average of 3.65 piglets per litter)] have been reported, these studies contrast directly with Experiment 2 as they did not involve an administered challenge of known virulence at a time of high susceptibility, the serological status of each animal before vaccination was not known, and the PRRSV strains present at the trial locations were not identified.

Only 1 study was found that examined the efficacy of 2 EU-type MLV PRRSV vaccines in gilts that also involved an administered challenge of PRRS virus. Using gilts vaccinated (IM) once at 24 d before AI and subsequently challenged with a heterologous strain, Scortti et al evaluated the effects of 2 different EU-type MLV PRRSV vaccines on gilts and their progeny (8). While no adverse reactions to either vaccine were noted, all vaccinated gilts except 1 developed vaccine-induced viremia. Neither vaccine provided complete protec-tion against challenge-induced viremia as the virus was isolated in the serum of approximately 40% of all vaccinated gilts. Conversely, reproductive performance significantly improved in both groups of vaccinated gilts, with more live piglets and fewer stillborn and mummified piglets born, and decreased incidence of transplacental infection of piglets after the challenge. Additionally, ADWG and piglet mortality rates were not significantly affected by congenital infection. While the reproductive results in this study are similar to the data presented here, only 50.0% of vaccinated gilts in group 2B and 36.0% of vaccinated gilts in group 2C became viremic as a result of vaccination. The majority of vaccinated gilts (69.0% of group 2B and 75.0% of 2C) were protected from challenge-induced viremia and, while congenital infection was noted for piglets born to vac-cinated gilts, piglet growth and survival rates were significantly improved by vaccination.

The data presented here has shown that this novel vaccine strain provides gilts with far better protection against PRRSV during pregnancy than inactivated vaccines. Inactivated EU-type vaccines have been shown to be protective in the face of a homologous PRRSV challenge, significantly increasing percentages of live and healthy piglets born (18) and decreasing duration of viremia and occurrence

of congenital infection, reducing percentage of mummified fetuses, and improving fetal survival (16). When a heterologous challenge was used, however, an inactivated EU-type vaccine failed to signifi-cantly protect gilts from clinical signs and viremia and decreased reproductive performance and transplacental infection (17).

These cited studies support the fact that the heterogeneity of the PRRS virus continues to make it difficult to develop a single-strain vaccine virus that is effective against the various strains of PRRSV in the field. It is difficult to directly compare the safety and efficacy of EU-type PRRS MLV vaccines in breeding females, due to the lim-ited nature of available data. Vaccination with a PRRS MLV vaccine, however, has generally been found to positively affect reproductive performance when used in PRRSV endemic situations. The present study provides evidence that a vaccine based on the novel PRRSV MLV strain 94881 is a safe and effective way to protect against the negative clinical and economic effects of PRRS viral infection when administered to gilts before breeding.

When the novel PRRS 94881 MLV vaccine, ReproCyc PRRS EU-type, was administered as a repeat 103 overdose (approximately 1 3 108 TCID50/6.0 mL) to bred gilts at approximately 90 d of ges-tation, there were no relevant differences for piglets in live-born, healthy live-born, general health during the suckling period, ADWG, live piglets at weaning, lung pathology, and viral load in lung tis-sue, and for gilts, in systemic reactions to vaccination and viremia. When a single dose of the test vaccine at various titer formulations was administered to breeding-age gilts 26 to 32 d before breeding, vaccinated gilts had more live born, healthy live-born, and weaned piglets per litter. The piglets had higher ADWG, fewer weak live and mummified piglets per litter, fewer piglets with abnormal clinical signs, fewer incidences of piglet viremia, no statistically relevant abnormal clinical assessments post-vaccination, positive serological responses as early as 14 d post-vaccination, and lower incidence of viremia following challenge.

In conclusion, A vaccine formulated from the PRRSV MLV strain 94881 has therefore been proven to be a safe and effective method of protection against the detrimental effects of virulent PRRSV infection in breeding female pigs.

A c k n o w l e d g m e n t sThis work was funded by Boehringer Ingelheim Vetmedica. The

authors thank Dr. Ryan Saltzman, Dr. Lyle Kesl, Dr. Stephan Pesch, Dr. Martin Vanselow, Mr. Rex Smiley, and Ms. Sarah Layton for technical assistance.

Re f e r e n c e s 1. Alexopoulos C, Kritas SK, Kyriakis CS, Tzika E, Kyriakis SC. Sow

performance in an endemically porcine reproductive and respira-tory syndrome (PRRS)-infected farm after sow vaccination with an attenuated PRRS vaccine. Vet Microbiol 2005;111:151–157.

2. Charerntantanakul W. Porcine reproductive and respiratory syndrome virus vaccines: Immunogenicity, efficacy and safety aspects. World J Virol 2012;1:23–30.

3. Kimman TG, Cornelissen LA, Moormann RJ, Rebel JMJ, Stockhofe-Zurwieden N. Challenges for porcine reproductive



and respiratory syndrome virus (PRRSV) vaccinology. Vaccine 2009;27:3704–3718.

4. Olanratmanee EO, Thanawongnuwech R, Kunavongkrit A, Tummaruk P. Reproductive performance of sows with and without PRRS modified live virus vaccination in PRRS-virus-seropositive herds. Trop Anim Health Prod 2014;46:1001–1007.

5. Osorio FA, Galeota JA, Nelson E, et al. Passive transfer of virus-specific antibodies confers protection against reproductive failure induced by a virulent strain of porcine reproductive and respiratory syndrome virus and establishes sterilizing immunity. Virol 2002;302:9–20.

6. Papatsiros, VG. Porcine respiratory and reproductive syndrome virus vaccinology: A review for commercial vaccines. Am J Anim Vet Sci 2012;7:149–158.

7. Pejsak Z, Markowska-Daniel I. Randomised, placebo-controlled trial of a live vaccine against porcine reproductive and respiratory syn-drome virus in sows on infected farms. Vet Rec 2006;158:475–478.

8. Scortti M, Prieto C, Simarro I, Castro JM. Reproductive per-formance of gilts following vaccination and subsequent heter-ologous challenge with European strains of porcine reproduc-tive and respiratory syndrome virus. Theriogenology 2006;66: 1884–1893.

9. Roca M, Gimeno M, Bruguera S, et al. Effects of challenge with a virulent genotype II strain of porcine reproductive and respira-tory syndrome virus on piglets vaccinated with an attenuated genotype I strain vaccine. Vet J 2012;193:92–96.

10. Scortti M, Prieto C, Álvarez E, Simarro I, Castro JM. Failure of an inactivated vaccine against porcine reproductive and respiratory syndrome to protect gilts against a heterologous challenge with PRRSV. Vet Rec 2007;161:809–813.

11. Mengeling WL, Lager KM, Vorwald AC, Clouser DF. Com-parative safety and efficacy of attenuated single-strain and multi-strain vaccines for porcine reproductive and respiratory syndrome. Vet Microbiol 2003;93:25–38.

12. Mengeling WL, Lager KM, Vorwald AC. Safety and efficacy of vaccination of pregnant gilts against porcine reproductive and respiratory syndrome. Am J Vet Res 1999;60:796–801.

13. Dewey CE, Wilson S, Buck P, Leyenaar JK. The reproductive performance of sows after PRRS vaccination depends on stage of gestation. Prev Vet Med 1999;40:233–241.

14. Scortti M, Prieto C, Martínez-Lobo FJ, Simarro I, Castro JM. Effects of two commercial European modified-live vaccines against porcine reproductive and respiratory syndrome viruses in pregnant gilts. Vet J 2006;172:506–514.

15. Martínez-Lobo FJ, Carrascosa de Lome L, Díez-Fuertes F, et al. Safety of porcine reproductive and respiratory syndrome modi-fied live virus (MLV) vaccine strains in a young pig infection model. Vet Res 2013;44:115–128.

16. Schelkopf A, Nerem J, Cowles B, Amodie D, Swalla R, Dee S. Reproductive, productivity, and mortality outcomes in late-gestation gilts and their litters following simulation of inad-

vertent exposure to a modified-live vaccine strain of porcine reproductive and respiratory syndrome (PRRS) virus. Vaccine 2014;32:4639–4643.

17. Karniychuk UU, Saha D, Vanhee M, et al. Impact of a novel inactivated PRRS virus vaccine on virus replication and virus-induced pathology in fetal implantation sites and fetuses upon challenge. Theriogenology 2012;78:1527–1537.

18. Plana-Durán J, Bastons M, Urniza A, Vayreda M, Vila X, Mañé H. Efficacy of an inactivated vaccine for prevention of reproduc-tive failure induced by porcine reproductive and respiratory syndrome virus. Vet Microbiol 1997;55:361–370.

19. Hu J, Zhang C. Porcine reproductive and respiratory syndrome virus vaccines: Current status and strategies to a universal vac-cine. Transbound Emerg Dis 2014;61:109–120.

20. Pileri E, Gibert E, Soldevila F, et al. Vaccination with a geno-type 1 modified live vaccine against porcine reproductive and respiratory syndrome virus significantly reduces viremia, viral shedding and transmission of the virus in a quasi-natural experi-mental model. Vet Microbiol 2015;175:7–16.

21. Geldhof MF, Vanhee M, Van Breedam W, Van Doorsselaere J, Karniychuk UU, Nauwunck HJ. Comparison of the efficacy of autogenous inactivated porcine reproductive and respiratory syndrome virus (PRRSV) vaccines with that of commercial vac-cines against homologous and heterologous challenges. BMC Vet Res 2012;8:182–197.

22. Han K, Seo HW, Park C, et al. Comparative virulence of repro-ductive diseases caused by type 1 (European-like) and type 2 (North American-like) porcine reproductive and respiratory syndrome virus in experimentally infected pregnant gilts. J Comp Path 2014;150:297–305.

23. Kiss I, Sámi L, Kecskeméti S, Hanada K. Genetic variation of the prevailing porcine respiratory and reproductive syndrome viruses occurring on a pig farm upon vaccination. Arch Virol 2006;151:2269–2276.

24. Labarque G, Van Reeth K, Nauwynck H, Drexler C, Van Gucht S, Pensaert M. Impact of genetic diversity of European-type porcine reproductive and respiratory syndrome virus strains on vaccine efficacy. Vaccine 2004;22:4183–4190.

25. Trus I, Bonckaert C, van der Meulen K, Nauwynck HJ. Efficacy of an attenuated European subtype 1 porcine reproductive and respiratory syndrome virus (PRRSV) vaccine in pigs upon chal-lenge with the East European subtype 3 PRRSV strain Lena. Vaccine 2014;32:2995–3003.

26. Burgard K, Kroll J, Layton SM, et al. United States Patent Number 8,765,142, United States Patent and Trademark Office, Alexandria, Virginia, 2014.

27. Revilla-Fernandex S, Wallner B, Truschner K, et al. The use of endogenous and exogenous reference RNAs for qualitative detection of PRRSV in porcine serum. J Virol Methods 2005;126: 21–30.


Article


I n t r o d u c t i o nHaemophilus parasuis is the causative agent of Glässer’s disease,

characterized by fibrinous polyserositis, arthritis, and meningi-tis. The bacterium is Gram-negative and a member of the family Pasteurellaceae. It can be a commensal bacterium colonized in the upper respiratory tract of clinically normal swine or a pathogen caus-ing severe systemic infection (1). Therefore, it is very important to differentiate between virulent and non-virulent isolates for diagnosis and control of the disease. Current research on virulence factors of H. parasuis is mainly based on comparison of high and low virulence strains, and an increasing number of putative virulence factors have been found (2,3). However, the function of these putative virulence factors is still to be determined due to a lack of a genetic manipula-tion method for construction of deletion mutant.

Construction of a genetic mutant is an efficient method for deter-mination of virulence factors. So far, 2 different systems have been

developed for transformation of H. parasuis including electropora-tion and natural transformation. A modified endogenous plasmid was used for the genetic manipulation of H. parasuis, and 15 out of 16 strains were transformable when the plasmid was pretreated with cell-free extracts (4). Electroporation efficiencies up to 105 were achieved when done with a native plasmid by growing H. parasuis at low temperature (5). A Tn5-based random mutagenesis method for use in H. parasuis was developed for preparation of mutant pools (6). However, there have been no reports of exogenous plasmid trans-formed into H. parasuis by electroporation. Hence, this method was not widely used due to difficulty in access to endogenous plasmid.

A natural transformation method was developed and mutants were constructed, but only 1/11 tested strains were naturally transformable. Moreover, the presence of specific DNA uptake signal sequence (USS) is required for natural transformation of H. parasuis, but the effect of the secondary messenger cyclic adenosine monophosphate (cAMP) on natural transformation is controversial

Efficient construction of Haemophilus parasuis mutants based on natural transformation

Junxing Li, Xiufang Yuan, Lihua Xu, Lei Kang, Jun Jiang, Yicheng Wang

A b s t r a c tStudies on virulence factors and pathogenecity of Haemophilus parasuis have long been hindered by a lack of a consistent system for genetic manipulation. In this study, competence was induced by transferring H. parasuis from rich medium to starvation medium media-IV (M-IV) and iscR gene deficient mutants of H. parasuis were generated efficiently. Transformation frequency varied from 4.1 3 1025 to 1.1 3 1028 when using circular plasmid, and increased to about 2- to 31-fold when transformed using linearized plasmid. Allele replacement occurred efficiently in 6 strains, which are transformable using both circular and linearized pTRU, but not in another 2 strains which could only be transformed using linearized plasmid. The iscR mutants were stable for at least 20 passages in vitro. Haemophilus parasuis strains vary extensively in natural transformation efficiency and the method established here allows for transformation of a larger spectrum of strains with an easily accessed plasmid. This provides important tools for genetic manipulation of H. parasuis.

R é s u m éLes études sur les facteurs de virulence et la pathogénicité d’Haemophilus parasuis ont longtemps été limitées à cause de l’absence d’un système constant de manipulations génétiques. Dans la présente étude, la compétence a été induite en transférant H. parasuis d’un milieu de culture riche au milieu pauvre M-IV et les H. parasuis mutants déficients pour le gène iscR étaient générés efficacement. La fréquence de transformation variait de 4,1 3 1025 à 1,1 3 1028 lors de l’utilisation d’un plasmide circulaire, et augmentait d’un facteur variant de 2 à 31 lorsque la transformation utilisait un plasmide linéarisé. Le remplacement d’allèle est survenu efficacement chez 6 souches, qui étaient transformables en utilisant le pTRU circulaire ou linéarisé, mais pas chez 2 autres souches qui ne pouvaient être transformées qu’en utilisant le plasmide linéarisé. Les mutants iscR étaient stables pendant au moins 20 passages in vitro. Les souches d’H. parasuis varient énormément dans leur efficacité de transformation naturelle et la méthode développée ici permet la transformation d’un plus large spectre de souches avec un plasmide facilement accessible. Ceci fournit d’importants outils pour la manipulation génétique d’H. parasuis.


Institute of Animal Husbandry and Veterinary Medicine, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China.

Address all correspondence to Professor Yicheng Wang; telephone: 86 571 8640 4121; fax: 86 571 8640 0836; e-mail: [email protected]

Received August 31, 2015. Accepted January 6, 2016.



(7,8). Recently, a successive markerless mutation system in H. para-suis through natural transformation was established, but only 1 of 6 screened strains was transformable (9). Searching for a consistent method to improve transformation efficiency for generation of H. parasuis mutants is still a challenge (2).

Natural competence has been extensively studied in Haemophilus influenzae, which is also a member of the family Pasteurellaceae. When H. influenzae was transferred from rich medium to starvation medium or from aerobic conditions to anaerobic conditions, compe-tence was developed (10,11). The addition of cAMP to exponentially

growing bacteria also induces competence (12). Transferring early exponentially growing bacteria to starvation medium also induces competence in Haemophilus parainfluenzae, Actinobacillus pleuropneu-moniae, and Gallibacterium anatis (13–15).

The objective of this study was to develop a consistent method for transformation and construction of mutant H. parasuis. Natural transformation of H. parasuis was conducted using 10 different field isolates and an iscR (a HTH-type transcriptional regulator at the 39 end of capsule gene cluster) deficient mutant was constructed in this study.

Table I. Bacterium strains and plasmids used in this study

Strain or plasmid Characteristics Source or isolation siteStrainsEscherichia coli DH5a F2, w 80dlacZ DM15, D(lacZYA 2argF )U169, deoR, recA1, endA1, hsdR17 (rK2, mK1), TaKaRa phoA, supE44 , l2, thi 21, gyrA96, relA1

H. parasuis ZJ0901 Serovar 4 clinical isolate, transformable Spleen ZJ1001 Serovar 13 clinical isolate, non-transformable Joint ZJ1004 Serovar 10 clinical isolate, transformable Trachea ZJ1017 Serovar 7 clinical isolate, transformable Joint ZJ1018 Serovar 13 clinical isolate, non-transformable Brain ZJ1112 Serovar 4 clinical isolate, transformable Pericardial effusion ZJ1208 Serovar 13 clinical isolate, transformable Bronchi ZJ1209 Serovar 4 clinical isolate, transformable Subcutaneous fluid ZJ1307 Serovar 7 clinical isolate, transformable Peritoneal effusion ZJ1404 Serovar 5 or 12 clinical isolate, transformable Joint

Plasmids pET 28a KanR Laboratory collection pUC 18 AmpR, suicide vector TaKaRa pTR AmpR, recombinant vector carrying about 970 bp homologous sequence upstream and This study downstream of iscR with kanamycin gene in between pTRU AmpR, recombinant vector carrying same gene with pRT except the addition of an USS This study in both homologous sequence

Figure 1. Schematic diagrams of the target region of the Haemophilus parasuis chromosome and of the recombinant plasmids pTRU and pTR.

chromosome

pUC18

P9

BamH I

M13-R

M13-F

Hind III

P6/P8

P1/P7

S2P4P2

S1 P3 P5

S1 P11

P12 S2 P10

chromosome

pUC18

973 bp upstream of iscR

973 bp upstream of iscR

iscR (447 bp)

kana ORF (816 bp)

974 bp downstream of iscR

974 bp downstream of iscR




Bacterial strains and culture conditionsThe H. parasuis strains used in this study are listed in Table I. All of

the strains were isolated from different farms in Zhejiang province, China. The serovar of the strains were determined by a newly devel-oped multiplex polymerase chain reaction (PCR) that distinguished between all previously described serovars except 5 and 12 (16). The strains were cultured at 37°C in brain heart infusion (BHI) or on Tryptic Soy Agar (TSA) supplemented with 0.025% nicotinamide adenine dinucleotide (NAD) and 2% bovine serum.

Plasmid constructionPrimers used in this study are listed in Table II, and a 9 bp USS

fragment of H. parasuis is present in primers 1 and 6. To construct a suicide vector for transformation and generate iscR mutants, a 973 bp fragment (containing BamH I site) upstream of the iscR ATG start codon and a 974 bp fragment (containing Hind III site) downstream of the iscR TAA stop codon were amplified from genome DNA of the strain ZJ1018; and a 816 bp fragment of kanamycin coding sequence was amplified from pET 28a vector. The 3 fragments were ligated by overlap PCR, as described (4), and cloned into pUC 18 vector (the pMB1 replicon carried in the plasmid allows it to replicate in E. coli but not H. parasuis) by BamH I and Hind III to form pTRU (Figure 1). Another plasmid containing the same gene with pTRU except the presence of the 9 bp USS fragment was constructed and designated as pTR (Figure 1). Plasmids were extracted from DH5a using a kit (Endofree Maxi Plasmid Kit; Tiangen, China), and DNA concentration was measured using spectrophotometer (NanoVue; GE Healthcare, Baie d’Urfé, Quebec).

Media-IV(M-IV) medium induction of competenceCompetence was induced by transferring H. parasuis from BHI

to M-IV medium (11). Haemophilus parasuis strains were retrieved

from 280°C and inoculated in BHI supplemented with NAD and bovine serum at 1:100 dilution. The bacteria were cultured at 37°C by rotating at 200 rounds per minute (rpm) in a shaker incubator. When the optical density at 600 nm (OD600) reaches 0.1,0.2, the bacteria were collected by centrifugation and washed once in equal volume of M-IV. The bacteria was resuspended in equal volume of M-IV and incubated at 37°C by rotating at 200 rpm for 100 min in a shaker incubator. Then cells were collected and resuspended in an equal volume of M-IV medium for the transformation assay.

Transformation procedureTransformation with closed-circular or linearized plasmid DNA

was done as described for H. influenzae with little modification (17). When transformed with closed-circular plasmid, 1 mg of circular recombinant plasmid pTR or pTRU was added to 1 mL of compe-tent cells, mixed gently by inverting the tube 5 to 6 times, and then incubated at 37°C for 30 min. Then 80% glycerol was added to a final concentration of 30% to 32%, and incubated at 25°C for 10 min after inverting the tube 5 to 6 times. The bacteria were collected and resuspended in BHI before plating on TSA plus NAD and serum (containing 25 mg/mL kanamycin) immediately at proper dilution. The TSA plates were incubated at 37°C for 24 to 72 h. Transformation frequencies were calculated by CFU mL21 on TSA with kanamycin divided by CFU mL21 on TSA without kanamycin.

When transformed with linearized plasmid, 1 mg of BamH I linearized pTRU was mixed with 1 mL of competent cells, and the mixture was incubated at 37°C for 15 min. Cells were harvested and resuspended in BHI before plating on TSA plus NAD and serum (containing 25 mg/mL kanamycin) immediately at proper dilution. The TSA plates were incubated at 37°C for 24 to 72 h, and transformation frequencies were calculated as described above.

The 10 randomly selected strains were also subjected to electro-poration and another natural transformation procedure with pTRU as described (4,8).

Figure 2. Effect of growth stage on development of competence in 3 dif-ferent strains. Haemophilus parasuis strains were grown in BHI and samples were taken and transferred to M-IV medium every 30 min during growth with OD600 ranging from 0.1 to 1.0. Competent cells induced by M-IV were transformed with 1 mg/mL pTRU in triplicate, and transforma-tion frequencies were calculated. The averages and standard deviations of 3 parallel tests are shown.

Table II. Primers used in this study

Primer name Primer sequence (59→39)P1 (BamH I) GAGGGATCCACCGCTTGTATGTGTTAGGCTTGCCAACGTATGCTP2 ATGGCTCATACTCAAACCTTTTAAAAATGCTCP3 AAGGTTTGAGTATGAGCCATATTCAACGGGAAACP4 GATTTTTATAAGATTAGAAAAACTCATCGAGCATCAAATGP5 GAGTTTTTCTAATCTTATAAAAATCTCTTAGTTP6 (Hind III) GAGAAGCTTACAAGCGGTGACTCGCCTTCAACAAAGTTAAAGCP7 (BamH I) GAGGGATCCATGTGTTAGGCTTGCCAACGTATGCTP8 (Hind III) GAGAAGCTTGACTCGCCTTCAACAAAGTTAAAGCP9 ATGAATGAAAATAATCATGATGAP10 TTAATGATGCGACCACTCAATTGAGTP11 ATGCCATGGGCAAATTAACTTCACGAGGACGP12 GCCCTCGAGATGGTGATGATCGTGGCAGS1 CGTGAAACTACACTTAAAGAAGTTGS2 CTAATTTAGCAATATCCACTAAACCTGM13-F TGTAAAACGACGGCCAGTM13-R CAGGAAACAGCTATGACCRestriction sites are underlined. USS are in italics.

1. 00E-03

1. 00E-04

1. 00E-05

1. 00E-06

1. 00E-07

1. 00E-08

1. 00E-09

ZJ1307ZJ0901ZJ1017

0 0.2 0.4 0.6 0.8 1 1.2OD600

Tran

sfor

mat

ion

freq

uenc

y



Transformation assaysThree H. parasuis strains were selected for optimizing transfor-

mation conditions. To determine the optimum growth stage for development of competence, H. parasuis competence was induced at various growth stages measured by OD600, and competent cells were transformed using pTRU. To determine the effect of USS sequence on transformation frequency, H. parasuis competence was induced when OD600 reached 0.1,0.2, and competent cells were transformed using pTR or pTRU. Comparison of transformation efficiency using pTR or pTRU was analyzed by one-way ANOVA using computer software (SPSS Statistics, version 17.0; IBM, USA), data is considered statistically significant when P # 0.05.

To measure the occurrence of natural competence among H. para-suis strains, 10 strains were selected randomly, and competence was induced at OD600 of 0.2, then transformed with circular or BamH I linearized pTRU.

Identification of iscR deficient mutants of H. parasuis

To check the reliability of homologous recombination in H. para-suis, 6 strains were transformed using circular or linearized pTRU, and genome DNA of 20 kanamycin resistant colonies were extracted by proteinase K digestion for each strain (18). Primers P3/P4, P11/P12 were used to check the insertion of kanamycin gene and dele-tion of iscR gene in genome DNA, and M13 primers were used to check the absence of pTRU in the bacteria. Replacement of iscR gene by the kanamycin gene in the chromosome was further confirmed by sequencing the PCR amplicon with primers P9 and P10, which are located at 1349 bp ahead of the upstream homologous arm and 278 bp behind the downstream homologous arm, respectively.

Re s u l t s

Competence development at different growth stage

Competence of H. parasuis strains were induced at various cell concentrations (OD600 0.1 to 1.0) and competent cells were transformed

using pTRU. Transformation frequency was highest at OD600 0.1 to 0.2, and dropped continuously with the increase in cell concentration (Figure 2). Extensive loss of competence development ability was observed when OD600 reached 0.8 to 1.0.

Effect of USS on natural transformation of H. parasuis

Two recombinant plasmid pTRU (containing 2 extra USS) and pTR were transformed into H. parasuis under the same conditions. Transformation efficiency of pTRU was slightly higher than that of pTR in ZJ0901 and ZJ1017, and lower in ZJ1307 (Figure 3). No sig-nificant difference was found between the 2 plasmids in transforma-tion efficiency (P . 0.05). Analysis of the pTR gene using computer software (MegAlign, DNASTAR Lasergene; DNASTAR, Madison, Wisconsin, USA) revealed no homologous sequence of any of the 2 reported USS sequences (7,8) and its complement sequence. The presence of the USS in pTRU did not facilitate the transformation of H. parasuis.

Natural competence among H. parasuis strainsTen randomly selected field strains were transformed with circular

or linearized pTRU. It shows that 6 out of 10 strains could be trans-formed with both circular and linearized pTRU by using the M-IV procedure. Transformation frequency varies from 4.1 3 1025 to 1.1 3 1028 when using circular plasmid, and increases about 2- to 31-fold when transforming with linearized plasmid (Figure 4). Among the rest, 4 strains that could not be transformed with circular plasmid, 2 of them (ZJ1004 and ZJ1209) could be transformed using linearized pTRU at very low efficiency. Extensive difference in transformation efficiency was observed among strains (Figure 4).

Efficient targeted gene replacement of H. parasuis by natural transformation

The deletion of iscR and insertion of kanamycin gene were determined by PCR. It shows that the iscR gene was replaced by the kanamycin gene in the H. parasuis chromosome, and the iscR mutant shows genetic stability for at least 20 passages on TSA in the presence or absence of kanamycin (Figure 5). Twenty kanamycin resistant clones from each of the 6 strains transformed with circular

Figure 3. Effect of USS on transformation frequency in 3 different strains. Haemophilus parasuis were cultured in BHI medium, and transferred to M-IV medium when the culture OD600 reaches 0.1 to 0.2. Competent cells induced with M-IV were transformed using 1 mg/mL pTR (carrying the same gene sequence with pTRU except the absence of 2 USSs) or pTRU, and transformation frequencies were calculated. The averages and standard deviations of 3 independent experiments are shown.

Figure 4. Transformation frequencies in various strains. Ten strains of Haemophilus parasuis from different origins were cultured in BHI medium, and transferred to M-IV medium when the culture OD600 reaches 0.2. Competent cells induced with M-IV were transformed using 1 mg/mL circular or linearized pTRU, and transformation frequencies were calcu-lated. The averages and standard deviations of 3 independent experi-ments are shown.

1. 00E-04

1. 00E-05

1. 00E-06

1. 00E-07

1. 00E-08

1. 00E-03

1. 00E-04

1. 00E-05

1. 00E-06

1. 00E-07

1. 00E-08

ZJ14

04

ZJ09

01

ZJ10

01

ZJ10

04

ZJ10

17

ZJ10

18

ZJ11

12

ZJ12

08

ZJ12

09

ZJ13

07

pTRpTRU

Circular plasmidLinearized plasmid

ZJ1307 ZJ1017 ZJ0901

Tran

sfor

mat

ion

freq

uenc

y

Tran

sfor

mat

ion

freq

uenc

y



or linearized pTRU were checked, and over 90% of the clones were iscR mutant. Interestingly, targeted gene replacement only took place in the 6 strains that could be transformed using both circular and linearized plasmid, but not in strains (ZJ1004 and ZJ1209), which could be only transformed using linearized plasmid. In these 2 strains (ZJ1004 and ZJ1209), incomplete homologous recombina-tion may occur, and the kanamycin gene and iscR gene were both detected in the recombinants obtained from the linearized plasmid. In addition, the suicide of the pTRU in H. parasuis was confirmed by PCR with M13 primers (Figure 5).

D i s c u s s i o nAlthough increases in potential virulence factors have been found,

pathogenesis of H. parasuis is poorly understood. Identification of these factors involved in pathogenesis has been limited due to a lack of general methods for producing mutants. In this study, we found the applicability of the defined M-IV medium in inducing competence of H. parasuis, and developed a simple method for gen-erating mutants. The M-IV medium was first developed for inducing competence in H. influenzae (11). Competence development of H. parasuis was better at an early exponential growth stage (OD600 0.1 to 0.2), which is similar to the conditions known to be optimal for induction of competence in H. influenzae (17).

The DNA uptake signal sequence is highly over-represented in genomes of several members of Pasteurellaceae, and is evolutionary conserved (20). One USS signal sequence of H. parasuis, identical to that found in Actinobacillus pleuropneumoniae and Mannheimia haemo-lytica, was identified in the SH0165 genome (21). The effect of USS on natural transformation of H. parasuis has been confirmed previously, although H. parasuis strains may differ in requirements of the USS nucleotide sequence (7,8). But our result shows that the reported USS of H. parasuis has no effect on transformation efficiency of M-IV medium-induced competent cells. Similarly, it is believed that the USS does not create a practical barrier to efficient transformation in

Gallibacterium anatis (G. anatis), and plasmids with no USS sequences could be well transformed into the bacteria (15).

It is believed that the competence development of H. influenzae is due to a starvation response. In H. influenzae, competence is induced with the starvation medium and absolutely requires an increase in cAMP, an established indicator of nutritional stress, but excessive cAMP may interfere with competence induction (22). In H. parasuis, the effect of cAMP on natural transformation is contro-versial. Increasing transformation efficiency was observed when the concentration of cAMP supplementation rising from 0.05 to 8 mM, but no significant difference in transformation efficiency was observed in another report under similar cAMP concentration and transformation procedure (7,8). The genetic difference of strains used in a previous report may be responsible for the sensitivity of cAMP in inducing competence, and it had been shown that an icc deletion mutant of H. influenzae was much more sensitive to exogenous cAMP (22). We also tested the effect of cAMP on trans-formation efficiency in M-IV medium induced competent cells, and supplementation of 1 mM cAMP in either competence induction or transformation stage did not increase the transformation fre-quency. This may be because sufficient cAMP has been produced when cultured in starvation medium, extra cAMP is not necessary. It has been demonstrated that E. coli cAMP levels peak during the transition to starvation (23–25).

The variation in transformability was reported in several species. Only 18 of 31 serotype b isolates of H. influenzae were transformable, while 29 of 34 strains were transformable when the tested isolates were dominated by non-serotype b strains (26,27). In G. anatis, all 9 tested strains were transformable with chromosomal DNA or lin-earized gene disruption constructs, but only 6 of 9 were transform-able using circular plasmid (15). Ten randomly selected H. parasuis strains were subjected to natural transformation, and 6 of them were found transformable using both circular and linearized plasmid. Moreover, 2 of the remaining 4 strains could be transformed using linearized plasmid. The increase of transformation frequencies was observed when using linearized instead of the circular pTRU, and the increase of transformants generated by linear plasmid was also observed when using a different natural transformation procedure (9). Only 1 of the 10 strains (ZJ1208) were transformable when sub-jected to the previously reported transformation procedure, which similarly reported that 1 of 11 tested strains were transformable (8). Electroporation was used to transform H. parasuis, but no kanamy-cin resistant clone was observed under our experiment conditions.

In this study, 2 serovar 7 strains were transformable using both cir-cular and linearized plasmid; all 3 serovar 4 strains were transform-able with linearized plasmid, but only 2 of them were transformable using circular plasmid. Only 1 of the 3 serovar 13 strains could be transformed using both forms of plasmid, while the other 2 strains were non-transformable with both forms of the plasmid. Hence, natural competence of these strains shows no strict association with serovar, but it is impossible to carry on a statistical analysis due to the limited number of strains tested in this study. Analysis with more strains in each serotype is needed to determine the relationship between natural competence and serovar.

Study of H. parasuis virulence factors has long been hindered by problems in generating mutants. The homologous recombination

Figure 5. Identification of DiscR mutants. The PCR analysis of genomic DNA of the 2nd (P2) and 20th (P20) passages of DiscR mutants with wild type H. parasuis (HPs-WT) and pTRU as control. Lane 1—primers P3 and P4 were used to identify the insertion of kanamycin gene (816 bp) in chromosome. Lane 2—primers P11 and P12 were used to identify the deletion of iscR (447 bp) gene from chromosome. Lane 3—prim-ers S1 and S2 were used to check the replacement of iscR gene with kanamycin gene, and the size of PCR product increases from 1265 bp to 1634 bp after the replacement. Lane 4—primers M13-F and M-13-R were used to check the suicide of pTRU after homologous recombination took place. Lane 5—HP1F3/HP2F2 and HPRevx are H. parasuis species specific primers with amplicon size of approximately 1090 bp (19). Lane M—250 bp DNA ladder.

HPs-WT

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

pTRU

M bp4500300022501500

1000750

500

250

DiscR mutant (P2) DiscR mutant (P20)



was efficient in M-IV induced competent H. parasuis, and the iscR mutants were genetically stable. Moreover, another wza (capsu-lar export protein) mutant was also generated by the procedure established in this study in our laboratory (28). Hence, the method established here allows the transformation of a larger spectrum of strains with an easily accessed plasmid, which will facilitate the identification of virulence factors and the characterization of patho-genicity of H. parasuis.

A c k n o w l e d g m e n tThis study was financially supported by National Natural Science

Foundation of China (Grant No. 31201934).

Re f e r e n c e s 1. Aragon V, Segalés J, Oliveira S. Glässer’s disease. In: Zimmerman

JJ, Karriker LA, Ramirez A, Schwartz KJ, Stevenson GW, eds. Diseases of Swine. 10th ed, Ames, Iowa: Wiley-Blackwell, 2012:760–769.

2. Costa-Hurtado M, Aragon V. Advances in the quest for virulence factors of Haemophilus parasuis. Vet J 2013;198:571–576.

3. Zhang B, Tang C, Liao M, Yue H. Update on the pathogenesis of Haemophilus parasuis infection and virulence factors. Vet Microbiol 2014;168:1–7.

4. Chen L, Wu D, Cai X, et al. Electrotransformation of H. parasuis with in vitro modified DNA based on a novel shuttle vector. Vet Microbiol 2012;155:310–316.

5. Costa-Hurtado M, Galofré-Milà N, Martinez-Moliner V, Kiljunen S, Savilahti H, Aragon V. Enhancement of electro-transformation efficiency of Haemophilus parasuis. In: Proceedings of the 112th General Meeting of the American Society for Microbiology, San Francisco, California, USA, 16–19 June 2012. Poster 1612.

6. Luan SL, Chaudhuri RR, Peters SE, et al. Generation of a Tn5 transposon library in Haemophilus parasuis and analysis by transposon-directed insertion-site sequencing (TraDIS). Vet Microbiol 2013;166:558–566.

7. Bigas A, Garrido ME, de Rozas AM, Badiola I, Barbe J, Llagostera M. Development of a genetic manipulation system for Haemophilus parasuis. Vet Microbiol 2005;105:223–228.

8. Zhang B, Feng S, Xu C, et al. Serum resistance in Haemophilus parasuis SC096 strain requires outer membrane protein P2 expres-sion. FEMS Microbiol Lett 2012;326:109–115.

9. Zhang L, Li Y, Dai K, et al. Establishment of a successive mark-erless mutation system in Haemophilus parasuis through natural transformation. PLoS One 2015;10:e0127393.

10. Goodgal SH, Herriott RM. Studies on transformations of Hemophilus influenzae. I. Competence. J Gen Physiol 1961;44:1201–1227.

11. Herriott RM, Meyer EM, Vogt M. Defined nongrowth media for stage II development of competence in Haemophilus influenzae. J Bacteriol 1970;101:517–524.

12. Wise EM, Jr, Alexander SP, Powers M. Adenosine 39:59-cyclic monophosphate as a regulator of bacterial transformation. P Natl Acad Sci 1973;70:471–474.

13. Sisco KL, Smith HO. Sequence-specific DNA uptake in Haemophilus transformation. Proc Nati Acad Sci 1979;76:972–976.

14. Redfield RJ, Findlay WA, Bossé J, Kroll JS, Cameron AD, Nash JH. Evolution of competence and DNA uptake specificity in the Pasteurellaceae. BMC Evol Biol 2006;6:82.

15. Kristensen BM, Sinha S, Boyce JD, Bojesen AM, Mell JC, Redfield RJ. Natural transformation of Gallibacterium anatis. Appl Environ Microbiol 2012;78:4914–4922.

16. Howell KJ, Peters SE, Wang JH, et al. Development of a multiplex PCR for rapid molecular serotyping of Haemophilus parasuis. J Clin Microbiol 2015. doi:10.1128/JCM.01991–15.

17. Poje G, Redfield RJ. Transformation of Haemophilus influenzae. Methods Mol Med 2003;71:57–70.

18. De la Puente Redondo VA, Navas Méndez J, García del Blanco N, Ladrón Boronat N, Gutiérrez Martín CB, Rodríguez Ferri EF. Typing of Haemophilus parasuis strains by PCR-RFLP analysis of the tbpA gene. Vet Microbiol 2003;92:253–262.

19. Angen O, Oliveira S, Ahrens P, Svensmark B, Leser TD. Development of an improved species specific PCR test for detec-tion of Haemophilus parasuis. Vet Microbiol 2007;119:266–276.

20. Bakkali M, Chen TY, Lee HC, Redfield RJ. Evolutionary stability of DNA uptake signal sequences in the Pasteurellaceae. Proc Natl Acad Sci USA 2004;101:4513–4518.

21. Xu Z, Yue M, Zhou R, et al. Genomic characterization of Haemophilus parasuis SH0165, a highly virulent strain of serovar 5 prevalent in China. PLoS One 2011;6:e19631.

22. Macfadyen LP, Ma C, Redfield RJ. A 39, 59 cyclic AMP (cAMP) phosphodiesterase modulates cAMP levels and optimizes com-petence in Haemophilus influenzae Rd. J Bacteriol 1998;180:4401–4405.

23. Notley L, Ferenci T. Differential expression of mal genes under cAMP and endogenous inducer control in nutrient-stressed Escherichia coli. Mol Microbiol 1995;16:121–129.

24. Notley L, Ferenci T. Induction of RpoS-dependent functions in glucose-limited continuous culture: What level of nutrient limita-tion induces the stationary phase of Escherichia coli? J Bacteriol 1996;178:1465–1468.

25. Notley ML, Death A, Ferenci T. The relationship between exter-nal glucose concentration and cAMP levels inside Escherichia coli: Implications for models of phosphotransferase-mediated regulation of adenylate cyclase. Microbiol 1997;143:1909–1918.

26. Rowji P, Gromkova R, Koornhof H. Genetic transformation in encapsulated clinical isolates of Haemophilus influenzae type b. J Gen Microbiol 1989;135:2775–2782.

27. Maughan H, Redfield RJ. Extensive variation in natural compe-tence in Haemophilus influenzae. Evolution 2009;63:1852–1866.

28. Kang L, Li JX, Wang YC, et al. Construction and biological characteristic of Wza-deficient mutant of Haemophilus parasuis. J Agric Biotechnol 2015;23:806–815.


Article


I n t r o d u c t i o nHaemophilus parasuis, which belongs to the Pasteurellaceae family,

can cause severe infections of the upper respiratory tract, polysero-sitis, meningitis, and arthritis (Glässer’s disease) in pigs (1). To date, H. parasuis has mainly caused significant losses among piglets 4 to 12 wk old (2,3). Fifteen serovars of H. parasuis, which range from highly virulent to nonvirulent, have been described by immunodiffu-sion tests and multiplex polymerase chain reaction (PCR). Sequence variations within the capsule loci can distinguish all 15 serovars except serovars 5 and 12 (1,4). Moreover, the pathogenicity and prevalence of different H. parasuis serovars may vary among regions and over time within a region (5–8).

Vaccination is generally considered to be the most effective means of controlling Glässer’s disease, which normally has high mortality

and morbidity rates. Inactivated H. parasuis bacterin, which is used worldwide, can elicit efficient protection against challenge with a homologous serovar. However, disease control is limited because of the lack of effective vaccines against a broad spectrum of strains (9–11). Recently, subunit vaccines that comprise newly identi-fied protective antigens, such as recombinant transferrin-binding protein B (TbpB), outer membrane protein (OMP) formulations enriched with TbpB, 4 OMPs (OMP2, D15, PalA, and HPS-06257), or transferrin-binding protein A, have been proven to provide partial protection against H. parasuis challenge (9,10,12). All types of vac-cines usually provide protection against challenge with homologous serovars, but few studies have reported cross-protection (2,12,13). To date, vaccines against Glässer’s disease containing inactivated H. parasuis serovars 4 and 5, serovar 5, and serovars 1 and 6 are the primary commercial vaccines in China, the United States, and Spain,

Minimum dose, antigen content, and immunization duration of a trivalent vaccine of inactivated Haemophilus parasuis serovars 4, 5, and 12

against Glässer’s disease in pigsZhanqin Zhao, Huisheng Liu, Keshan Zhang, Qiao Xue, Kunpeng Chen, Yun Xue

A b s t r a c tThe objective of this study was to assess the minimum dose, antigen content, and immunization duration of a trivalent vaccine containing inactivated Haemophilus parasuis serovars 4, 5, and 12 and the Montanide GEL 01 PR adjuvant in piglets and pregnant sows. Our results demonstrated that the minimum vaccine dose was 2 mL per pig and the optimal antigen content 2.0 3 109, 1.0 3 109, and 1.0 3 109 colony-forming units/mL of serovars 4, 5, and 12, respectively. The vaccine provided effective protection 14 d after the 2nd vaccination, and the period of immune protection was 180 d (6 mo) after the 2nd vaccination. Maternal antibodies provided early protection for the piglets, and vaccinating the sows before farrowing helped to control disease and protected the piglets during lactation; the piglets were protected during the finishing period by being vaccinated during lactation. Our findings provide a basis for developing a commercial trivalent vaccine of inactivated H. parasuis serovars 4, 5, and 12 against Glässer’s disease.

R é s u m éL’objectif de la présente étude était d’évaluer la dose minimale, le contenu en antigène, et la durée d’immunisation d’un vaccin trivalent contenant les sérovars 4, 5, et 12 d’Haemophilus parasuis et l’adjuvant Montanide GEL 01 PR chez des porcelets et des truies gestantes. Nos résultats ont démontré que la dose minimale de vaccin était de 2 mL par porc et le contenu optimal en antigène était de 2,0 3 109, 1,0 3 109, and 1,0 3 109 unités formatrices de colonie/mL pour les sérovars 4, 5, et 12, respectivement. Le vaccin a fourni une protection efficace 14 j suite à la deuxième vaccination, et la période de protection immunitaire était de 180 j (6 mois) après la deuxième vaccination. Les anticorps maternels ont fourni une protection précoce aux porcelets, et la vaccination des truies avant la mise-bas aidait à maitriser la maladie et a protégé les porcelets durant la lactation; les porcelets étaient protégés durant la période de finition en étant vaccinés durant la lactation. Nos trouvailles fournissent des éléments de base pour développer un vaccin trivalent commercial contre la maladie de Glässer avec des souches inactivées d’H. parasuis sérovars 4, 5, et 12.


State Key Laboratory of Veterinary Etiological Biology, Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences, Lanzhou 730046, China (Zhao, Liu, Zhang); Laboratory of Veterinary Microbiology, College of Animal Science and Technology, Henan University of Science and Technology, Luoyang, China (Zhao, Chen, Qiao Xue); Laboratory of Medical Engineering, College of Medical Technology and Engineering, Henan University of Science and Technology, Luoyang, China (Yun Xue).

Zhanqin Zhao and Huisheng Liu contributed equally to this study.

Address all correspondence to Dr. Yun Xue; telephone: 186-13633799373; e-mail: [email protected]

Received April 3, 2016. Accepted July 6, 2016.



respectively. These vaccines play an important role in preventing and controlling Glässer’s disease (14).

In a previous study we screened the strongly immunogenic serovars 4, 5, and 12 as candidate vaccine strains and compared the effectiveness of vaccines containing inactivated H. parasuis serovars 4, 5, and 12 when administered with a variety of adjuvants [tra-ditional mineral oil, aluminum hydroxide, Montanide GEL 01 PR (a new adjuvant based on the dispersion of a high-molecular-weight polyacrylic polymer in water; SEPPIC, Shanghai, China), Montanide IMS 1313N VG, and Montanide ISA 760 VG]. We concluded that the Montanide GEL 01 PR should be used as a candidate adjuvant, deserving further study because a vaccine administered with this adjuvant was safe, generated high concentrations of antibodies, and exhibited 100% protection (15). The objectives of the present study were to assess the minimum dose, antigen content, and immuniza-tion duration of a trivalent vaccine containing inactivated H. parasuis serovars 4, 5, and 12 administered to pigs for the first time and to provide a basis for developing a commercial trivalent vaccine with these inactivated serovars to protect against Glässer’s disease.


AnimalsFemale Landrace 3 Large White piglets aged 3 to 4 wk old and

Landrace 3 Large White sows that were 8 to 9 wk pregnant were purchased from ShengPing Co., Ltd, Luoyang, China. A single nasal swab obtained from each pig was shown by PCR to be negative for H. parasuis as described previously (15). The pigs were determined to be serologically negative by a microagglutination test (MAT) developed in our laboratory. Briefly, 50 mL of sterile PBS was added to 96-well plates, along with 50 mL of test serum diluted 2-fold with sterile PBS. Into each well was mixed 50 mL of inactivated H. parasuis whole-cell antigen. The mixture was incubated at 37°C for 7 h and then at 4°C for 12 to 24 h. Titers $ 1:8 and # 1:4 were considered to be positive and negative, respectively.

All the pigs were housed in isolation in pens with a concrete floor. Feed and water were provided ad libitum throughout the study. All the animal procedures were conducted in compliance with the guide-lines of the Animal Care and Use Committee of Henan University of Science and Technology, Luoyang, China.

Isolates of H. parasuisIsolates of H. parasuis, serovars 4, 5, and 12, that were screened

by our laboratory (16) were used for vaccine production. The bac-teria were cultured on tryptic soy agar (TSA) supplemented with nicotinamide adenine dinucleotide (NAD), 10 mg/mL, and 5% fetal calf serum and incubated at 37°C for 48 h. The organisms were then washed off the TSA plates with sterile 15% skimmed milk powder and stored at −80°C until used.

Vaccine formulations and vaccination scheduleFour vaccines (F1 to F4) were compared. Briefly, H. parasuis

serovars 4, 5, and 12 were enumerated by plate counts. Then the bacterial cultures were inactivated by incubation with 0.3% form-aldehyde solution at 37°C for 48 h and pelleted by high-speed

centrifugation. The pellets were resuspended in sterile phosphate-buffered saline (PBS) as whole-cell antigens and subsequently for-mulated with Montanide GEL 01 PR adjuvant (kindly provided by SEPPIC) at an 85:15 ratio (H. parasuis:adjuvant) in accordance with the manufacturer’s instructions. The vaccines contained serovars 4, 5, and 12 at the following antigen concentrations: F1, 2.0 3 109, 1.0 3 109, and 1.0 3 109 colony-forming units (CFU)/mL, respec-tively; F2, 1.0 3 109, 5.0 3 108, and 5.0 3 108 CFU/mL, respectively; F3, 5.0 3 108, 2.5 3 108, and 2.5 3 108 CFU/mL, respectively; and F4, 2.5 3 108, 1.3 3 108, and 1.3 3 108 CFU/mL, respectively. The physical properties and sterility of the vaccines were described by the Veterinary Biological Products Procedures of the People’s Republic of China.

All the vaccines were administered intramuscularly twice (21 d apart) in the neck. The day of the 1st vaccination was designated as d 0. At 14 d after the 2nd vaccination all the piglets were challenged intraperitoneally with 9.0 3 109 CFU (15) of H. parasuis serovar 4 (approximately 3 mL), 5.2 3 109 CFU (15) of H. parasuis serovar 5 (approximately 2 mL), and 3.5 3 109 CFU (15) of H. parasuis serovar 12 (approximately 1.5 mL); the bacteria had been cultured in trypti-case soy broth supplemented with NAD, 10 mg/mL, and 5% fetal calf serum for 16 h. Control pigs were injected with 2 mL of sterile PBS on the same day that the other groups were vaccinated. Blood samples from the piglets were collected by puncture of the jugular vein.

Minimum vaccine dose for pigletsSixty piglets were randomly assigned to 12 groups (A1 to A12) of

5 animals each. The piglets in groups A1 to A3, A4 to A6, and A7 to A9 were vaccinated with 2, 1, and 0.5 mL, respectively, of the F1 vaccine. The piglets in groups A10 to A12 served as controls. Groups A1, A4, A7, and A10 were challenged with H. parasuis serovar 4; groups A2, A5, A8, and A11 were challenged with serovar 5; and groups A3, A6, A9, and A12 were challenged with serovar 12.

Blood samples from the piglets were collected before challenge (on d 35). The serum was stored at −80°C until the MAT analysis. The experiment was terminated 15 d after challenge; survivors were killed by intravenous administration of sodium pentobarbital. Piglets found moribund during the experiment were humanely killed in the same manner. All of the animals underwent necropsy; gross lesions were recorded (special attention was paid to the pleural, pericardial, and peritoneal cavities, the lungs, and the central nervous system), and samples were obtained for bacterial isolation.

Minimum vaccine dose for pregnant sowsTwenty pregnant sows were randomly assigned to 4 groups

(B1 to B4) of 5 animals each. The sows in groups B1, B2, and B3 were vaccinated with 2, 1, and 0.5 mL, respectively, of the F1 vaccine. The sows in groups B4 served as controls.

Blood samples from piglets born to the pregnant sows were col-lected at 1, 4, 7, 10, 15, 20, 25, 30, 35, 40, 45, and 50 d of age. Serum antibody titers were analyzed to assess dynamic changes in the H. parasuis antibody levels in these piglets. Fifteen piglets 21 to 22 d old were selected from each group of sows. Then 15 piglets were randomly assigned to 3 groups of 5 piglets each, and immediately challenged with H. parasuis serovars 4, 5, and 12, respectively. The procedures after challenge were as described.



Effect of vaccine antigen content on pigletsSeventy-five piglets were randomly assigned to 15 groups (C1 to

C15) of 5 animals each. The piglets in groups C1 to C3, C4 to C6, C7 to C9, and C10 to C12 were vaccinated with 2 mL of the F1, F2, F3, and F4 vaccines, respectively. The piglets in groups C13 to C15 served as controls. Groups C1, C4, C7, C10, and C13 were chal-lenged with serovar 4; groups C2, C5, C8, C11, and C14 were chal-lenged with serovar 5; and groups C3, C6, C9, C12, and C15 were challenged with serovar 12.

Blood samples from the piglets were collected before challenge (on d 35), and the serum was stored at −80°C until the MAT analysis. The procedures after challenge were as described.

Effect of vaccine antigen content on pregnant sows

Twenty-five pregnant sows were randomly assigned to 4 groups (D1 to D5) of 5 animals each. The sows in groups D1, D2, D3, and D4 were vaccinated with 2 mL of the F1, F2, F3, and F4 vaccines, respec-tively. The sows in groups D5 served as controls. The procedures on the piglets born to these sows were as described.

Immunization duration in pigletsOne hundred and fifty piglets were randomly assigned to 2 groups

(E1 and E2) of 75 animals each. The piglets in group E1 were vac-cinated with 2 mL of the F1 vaccine. The piglets in group E2 served as controls. Blood samples were collected on days 0 and 21 after the 1st vaccination and on days 7, 14, 28, 60, 90, 120, 150, 180, 210, and 240 after the 2nd vaccination. Serum titers of H. parasuis antibody were analyzed to assess dynamic changes in the levels.

Fifteen piglets were selected from each group at 14, 90, 180, and 210 d after the 2nd vaccination and then randomly assigned to 3 groups of 5 piglets each. The piglets were immediately challenged with H. parasuis serovars 4, 5, and 12, respectively. The procedures after challenge were as described.

Immunization duration in pregnant sowsTwelve pregnant sows were randomly assigned to 2 groups

(Z1 and Z2) of 6 animals each. The sows in group Z1 were vaccinated with 2 mL of the F1 vaccine. Group Z2 served as controls. Blood samples from the sows were collected on days 0 and 21 after the 1st vaccination and on days 7, 14, 28, 60, 90, 120, 150, 180, 210, and 240 after the 2nd vaccination. Serum titers of H. parasuis antibody were analyzed to assess dynamic changes in the levels.

Blood samples from piglets born to the pregnant sows were col-lected at 1, 4, 7, 10, 15, 20, 25, 30, 35, 40, 45, and 50 d of age. Serum titers of H. parasuis antibody were analyzed to assess dynamic changes in the levels. Fifteen piglets aged 21 to 22 d, 34 to 36 d, and 56 to 58 d were selected from each group and then randomly assigned to 3 groups of 5 piglets each and immediately challenged with H. parasuis serovars 4, 5 and 12, respectively. The procedures after challenge were as described.

Statistical analysisStatistical analysis was done by means of 1-way analysis of vari-

ance. A P-value , 0.05 was considered to be significant. The analysis was conducted with the use of SPSS Statistics for Windows, version 17.0 (SPSS, Chicago, Illinois, USA).

Re s u l t sAs Table I shows for the vaccine-dose experiments, among the

piglets the mean titers of antibody to serovars 4, 5, and 12 were significantly higher (P , 0.05) in groups A1 to A3 (vaccinated with 2 mL of the F1 vaccine) than in groups A4 to A6 and groups A7 to A9 (vaccinated with 1 mL and 0.5 mL of the F1 vaccine, respectively). The mean titers did not differ significantly between groups A4 to A6 and groups A7 to A9. No H. parasuis antibodies were observed in groups A10 to A12, the unvaccinated control piglets. After chal-lenge with H. parasuis serovars 4, 5, and 12, all the control piglets had

Table I. Protective efficacy of the F1 vaccinea against serovars 4, 5, and 12 of Haemophilus parasuis and mean titers of antibody before challenge in piglets that had received various doses of the vaccine

Protective efficacy Number of Vaccine (and mean antibody titerb)Piglet group piglets dose (mL) Serovar 4 Serovar 5 Serovar 12A7 to A9 15 0.5 60% (1:12) 60% (1:11) 60% (1:12)A4 to A6 15 1 100% (1:18) 80% (1:16) 100% (1:18)A1 to A3 15 2 100% (1:28) 100% (1:37) 100% (1:37)Controlsc (A10 to A12) 15 2 0% (—) 0% (—) 0% (—)a The vaccine, given as 2 intramuscular doses 21 d apart, contained serovars 4, 5, and 12 of H. parasuis at antigen contents of 2.0 3 109, 1.0 3 109, and 1.0 3 109 colony-forming units (CFU)/mL, respectively. The piglets were challenged intraperitoneally with H. parasuis 14 d after the second vaccination.b The mean antibody titers in groups A1 to A3 were significantly higher (P , 0.05) than those in the other experimental groups.c The piglets in the control group were injected with 2 mL of sterile phosphate-buffered saline the same days as the other groups were vaccinated. The dash indicates that no H. parasuis antibodies were detected.



signs of Glässer’s disease, including anorexia, psychiatric disorders, arthritis, convulsions, serofibrinous exudates, and polyserositis of varying severity within the pericardium, thoracic cavity, and enterocelia; these piglets began to die 6 h after challenge, and all had died by 72 h. Vaccination with 0.5 mL of the F1 vaccine provided 60% protection against challenge, whereas vaccination with 1 mL provided 100% protection except against serovar 5 (80%), and vac-cination with 2 mL provided 100% protection. The organism was isolated from tissues and organs of all of the dead experimental piglets, which had exhibited symptoms and pathological changes of Glässer’s disease in the upper respiratory tract similar to those in the controls. None of the surviving piglets exhibited similar symptoms or pathological changes.

Dynamic changes in the titers of maternal antibodies against H. parasuis in the piglets born to the pregnant sows are shown in Table II for the vaccine-dose experiments. The mean titers in the piglets born to the B1 and B2 groups (vaccinated with 2 mL and 1 mL of the F1 vaccine, respectively) were highest at day 20 and gradually decreased thereafter. However, the titers decreased faster in the B2 group than in the B1 group. The mean titer in the piglets born to the B3 group (vaccinated with 0.5 mL of the F1 vaccine) was highest at day 15 and then gradually decreased; however, the titers in these piglets were the lowest (# 1:16) and were maintained for a short time. The mean titer in the piglets born to the B1 group was not significantly different from that of the piglets born to the B2 group but was significantly higher (P , 0.05) than that of the piglets born to the B3 group. The mean titer in the piglets born to the B2 group was not significantly different from that of the B3 group. No mater-nal antibodies were observed in the controls (group B4). Among the 15 piglets 21 to 22 d old that were selected from each group and challenged with H. parasuis serovars 4, 5, and 12 the protective efficacy of the vaccine was the same as for groups A1 to A3, A4 to A6, A7 to A9, and A10 to A12.

In the antigen-content experiments the mean titers of antibody to serovars 4, 5, and 12 in the piglets (Table III) were significantly higher (P , 0.05) in groups C1 to C3 (vaccinated with the F1 vaccine) than in groups C4 to C6 (vaccinated with the F2 vaccine). The antibody titers in those groups were significantly higher (P , 0.05) than the titers in groups C7 to C9 and C10 to C12 (vaccinated with the F3 and F4 vaccines, respectively). The mean titers did not differ significantly between groups C7 to C9 and groups C10 to C12. No antibodies against H. parasuis serovars 4, 5, and 12 were observed in the control piglets (groups C13 to C15), all of which showed signs of Glässer’s disease after challenge and died by 72 h after challenge. The F1 vaccine provided 100% protection against challenge with H. parasuis serovars 4, 5, and 12, and the F2 vaccine provided 100% protection against challenge with H. parasuis serovars 5 and 12 but only 80% protection against serovar 4, whereas the F3 and F4 vaccines provided little pro-tection against the 3 serovars. The organism was isolated from tissues and organs of all of the dead experimental piglets, which exhibited symptoms and pathological changes of Glässer’s disease in the upper respiratory tract similar to those in the controls. None of the surviving piglets exhibited similar symptoms or pathological changes.

Dynamic changes in the maternal antibody titers of the piglets born to pregnant sows in groups D1, D2, and D3 (vaccinated with the F1, F2, and F3 vaccines, respectively) were similar to those of the piglets born to pregnant sows in groups B1, B2, and B3 (Table II), respectively. The piglets in group D4 (vaccinated with the F4 vaccine) had the lowest mean titer (# 1:8), which was maintained for only a short time. No maternal antibodies against H. parasuis serovars 4, 5, and 12 were observed in the piglets in the control group. Among the 15 piglets 21 to 22 d old that were selected from each group and challenged with H. parasuis serovars 4, 5, and 12 the protective efficacies of the vaccine for the piglets in the D1 to D4 groups were identical to those in the C1 to C3, C4 to C6, C7 to C9, and C10 to C12 groups (Table III), respectively.

Table II. Mean titers of maternal antibody to H. parasuis serovars 4, 5, and 12 in piglets born to pregnant sows that had received various doses of the F1 vaccine

Vaccine dose, mL (and sow group); mean antibody titera

Piglet age Serovar 4 Serovar 5 Serovar 12(days) 2 (B1) 1 (B2) 0.5 (B3) 0 (B4) 2 (B1) 1 (B2) 0.5 (B3) 0 (B4) 2 (B1) 1 (B2) 0.5 (B3) 0 (B4) 1 — — — — — — — — — — — — 4 1:4 1:4 1:2 — 1:4 1:4 1:2 — 1:4 1:4 1:2 — 7 1:8 1:8 1:4 — 1:8 1:4 1:8 — 1:4 1:8 1:8 — 10 1:16 1:16 1:8 — 1:16 1:16 1:8 — 1:16 1:16 1:8 — 15 1:32 1:32 1:8 — 1:32 1:32 1:16 — 1:32 1:32 1:16 — 20 1:32 1:32 1:8 — 1:64 1:64 1:8 — 1:32 1:32 1:8 — 25 1:32 1:32 1:4 — 1:32 1:16 1:8 — 1:16 1:16 1:8 — 30 1:16 1:16 1:2 — 1:16 1:16 1:4 — 1:16 1:16 1:4 — 35 1:16 1:16 — — 1:16 1:8 1:2 — 1:8 1:8 1:2 — 40 1:8 1:8 — — 1:8 1:4 — — 1:4 1:4 — — 45 1:4 1:2 — — 1:2 — — — 1:2 1:2 — — 50 — — — — — — — — — — — —a Antibody titers $ 1:8 and # 1:4 were deemed to be positive and negative, respectively. The mean titer in the piglets born to the B1 group was not significantly different from that of the piglets born to the B2 group but was significantly higher (P , 0.05) than that of the piglets born to the B3 group. The mean titers in the piglets born to the B2 and B3 groups did not differ significantly. The dash indicates that no H. parasuis antibodies were detected.



Dynamic changes in the titers of antibody to H. parasuis serovars 4, 5, and 12 in the piglets in the immunization-duration experiment are shown in Table IV. The mean titers gradually increased from the 1st vaccination to day 28 after the 2nd vaccination and then gradually decreased, disappearing by day 210. The piglets were chal-lenged with H. parasuis serovars 4, 5, and 12 at 14, 90, 180, and 210 d after the 2nd vaccination. One piglet challenged with H. parasuis serovar 4 died 10 d after challenge at d 210, whereas all of the other experimental piglets survived; all of the control piglets had signs of Glässer’s disease and died by 72 h after challenge. The organism was isolated from tissues and organs of the dead experimental piglets, which exhibited symptoms and pathological changes of Glässer’s disease in the upper respiratory tract similar to those in the control groups. None of the surviving piglets exhibited similar symptoms and pathological changes.

Dynamic changes in the titers of antibody to H. parasuis serovar 4, 5, and 12 in the pregnant sows are shown in Table V. The titers increased more slowly than those in the piglets represented in Table IV from the 1st vaccination to day 14 after the 2nd vaccination, but other changes were similar to those for the piglets. Dynamic changes in the titers of maternal antibody in the piglets born to preg-nant sows were identical to those in the B1 group (vaccinated with 2 mL of the F1 vaccine) as shown in Table II. All the piglets aged 21 to 22 d and 60% of those aged 34 to 36 d survived challenge with H. parasuis serovars 4, 5, and 12, whereas all of the piglets aged 56 to 58 d, as well as the control piglets, died after challenge. The organ-ism was isolated from tissues and organs of all of the dead piglets, which had signs of Glässer’s disease. None of the surviving piglets exhibited symptoms and pathological changes of Glässer’s disease in the upper respiratory tract, as occurred in the control group.

D i s c u s s i o nGlässer’s disease has high morbidity and mortality rates in pigs in

China and some European countries. Its prevention and control are major problems worldwide. To date, because of a lack of knowledge regarding the virulence factors and common protective antigens of H. parasuis, effective vaccines that provide cross-protection against all pathogenic serovars are not available (12). Additionally, no other commercially available vaccine types, such as subunit vaccines and genetically engineered attenuated H. parasuis vaccines, that are effec-tive against pathogenic H. parasuis serovars have been developed. Therefore, the development of commercially available multivalent vaccines containing inactivated field isolates is necessary to protect against Glässer’s disease.

In China, H. parasuis serovars 4 and 5 are the most prevalent, followed by serovars 14, 13, and 12 (5–8). A vaccine containing inactivated H. parasuis serovars 4 and 5 can reduce the severity of symptoms and the risk of death in pigs after challenge with serovars 13 and 14 but fail to protect them from infection with serovar 12 (17). Additionally, to the best of our knowledge, no com-mercially available vaccines against H. parasuis serovar 12 have been developed. Hence we assessed a trivalent vaccine containing inactivated H. parasuis serovars 4, 5, and 12, administered with the Montanide GEL 01 PR adjuvant, in this study.

It is well-known that H. parasuis mainly infects the upper respira-tory tract of pigs. In this study, classic symptoms of Glässer’s disease were observed in the upper respiratory tract of the unvaccinated pigs, whereas the vaccinated pigs were protected at the infection site (upper respiratory tract) even though an intraperitoneal chal-lenge was used.

Table III. Protective efficacy of the F1, F2, F3, and F4 vaccinesa against serovars 4, 5, and 12 of H. parasuis and mean antibody titers before challenge in piglets that had received 2 mL of vaccine

Protective efficacyVaccine type Number of (and mean antibody titerb)(and piglet group) piglets Serovar 4 Serovar 5 Serovar 12F1 (C1 to C3) 15 100% (1:28) 100% (1:28) 100% (1:37)F2 (C4 to C6) 15 80% (1:16) 100% (1:18) 100% (1:18)F3 (C7 to C9) 15 40% (1:12) 60% (1:11) 40% (1:11)F4 (C10 to C12) 15 0% (1:8) 0% (1:8) 20% (1:8)Controlsc (C13 to C15) 15 0% (—) 0% (—) 0% (—)a The vaccines were given as 2 intramuscular doses 21 d apart. They contained serovars 4, 5, and 12 of H. parasuis at the following antigen concentrations: F1, 2.0 3 109, 1.0 3 109, and 1.0 3 109 CFU/mL, respectively; F2, 1.0 3 109, 5.0 3 108, and 5.0 3 108 CFU/mL, respectively; F3, 5.0 3 108, 2.5 3 108, and 2.5 3 108 CFU/mL, respectively; and F4, 2.5 3 108, 1.3 3 108, and 1.3 3 108 CFU/mL, respectively. The piglets were challenged intraperitoneally with H. parasuis 14 d after the 2nd vaccination.b The mean antibody titers in groups C1 to C3 were significantly higher (P , 0.05) than those in the other experimental groups.c The piglets in the control group were injected with 2 mL of sterile phosphate-buffered saline the same days as the other groups were vaccinated. The dash indicates that no H. parasuis antibodies were detected.



Our experiments that aimed to determine the minimum vaccine dose for piglets and pregnant sows showed that the concentration of antibodies against H. parasuis serovars 4, 5, and 12 increased with increasing vaccine dose: the antibody titers induced by 2 mL of the F1 vaccine were significantly higher (P , 0.05) than the titers induced in the other groups of vaccinees, and 2 mL of the F1 vac-cine provided 100% protection against challenge with H. parasuis serovars 4, 5, and 12. These results revealed that the minimum vaccine dose was 2 mL per pig. The experiments that examined the effect of the antigen content on piglets and pregnant sows showed that the concentration of antibodies against H. parasuis serovars 4, 5, and 12 increased with increasing antigen content: the antibody titers induced by 2 mL of the F1 vaccine were significantly higher (P , 0.05) than the titers induced in the other groups of vaccinees, and 2 mL of the F1 vaccine provided 100% protection against chal-lenge with H. parasuis serovars 4, 5, and 12. These results suggest that the optimal antigen contents of H. parasuis serovars 4, 5, and 12 are 2.0 3 109, 1.0 3 109, and 1.0 3 109 CFU/mL, respectively.

The minimum vaccine dose and optimal antigen content of the trivalent vaccine containing inactivated H. parasuis serovars 4, 5, and 12 were used to assess the immunization duration in piglets and pregnant sows. In these experiments, antibody titers of 1:16 or greater were observed from 14 to 180 d after the 2nd vaccination, and all of the piglets survived challenge with H. parasuis serovars 4, 5, and 12 at 14, 90, and 180 d after the 2nd vaccination. These results indicate that the trivalent vaccine provided effective protection at d 14 after the 2nd vaccination and that the immune protection

lasted until 180 d (6 mo) after the 2nd vaccination. Our results also indicate that the 1st vaccination should be done in sows that have been pregnant for 8 to 9 wk, because high antibody concentrations could be detected in the vaccinated pregnant sows during farrowing. Additionally, sows should be given a booster injection 21 d later, because neonatal piglets could receive more maternal antibodies that allow them to survive without resulting disease while they produce an active immune response.

The maternal antibodies that are induced by commercially avail-able bacterin containing H. parasuis serovars 4 and 5 or containing H. parasuis serovars 2, 3, and 5 can provide early protection for pig-lets (18–20). The results obtained with the piglets born to vaccinated sows in this study support this view. Additionally, we detected maternal antibodies in neonatal piglets and dynamic changes in the antibody titers in vaccinated piglets aged 3 to 9 wk, as well as in pregnant sows. Our findings suggest that the maternal antibody titers were maintained for 6 wk (Table II) and therefore that vac-cinating sows before farrowing may help to control the disease and protect the piglets during lactation. However, the susceptibility to H. parasuis infection and disease increased in piglets after weaning because of the disappearance of maternal antibodies and the loss of passive protection. Therefore, it is important that piglets be inocu-lated with an effective vaccine before the disappearance of maternal antibodies. In this study the maternal antibody concentration was highest in the piglets at 20 d of age and then gradually decreased. Thus, we recommend that the 1st vaccination be done in piglets 3 to 4 wk old and that they be given a booster injection 21 d later, as

Table IV. Dynamic changes in the mean titers of antibody to H. parasuis serovars 4, 5, and 12 in 3 groups of 5 piglets each that received 2 mL of the F1 vaccinea (E1) and 15 controlsb (E2)

Piglet Day after first vaccination (0 or 21) or 2nd vaccination; mean antibody titerc

Serovar group 0 21 7 14 28 60 90 120 150 180 210 240 4 E1 — 1:8 1:16 1:32 1:64 1:64 1:32 1:32 1:16 1:16 1:8 1:4 5 E1 — 1:8 1:16 1:32 1:128 1:64 1:64 1:32 1:32 1:16 1:16 1:4 12 E1 — 1:8 1:16 1:32 1:128 1:64 1:64 1:32 1:32 1:16 1:16 1:4 E2 — — — — — — — — — — — —a The vaccine was given as 2 intramuscular doses 21 d apart.b The controls were injected with 2 mL of sterile phosphate-buffered saline the same days as the other piglets were vaccinated.c Antibody titers $ 1:8 and # 1:4 were deemed to be positive and negative, respectively. The dash indicates that no H. parasuis antibodies were detected.

Table V. Dynamic changes in the mean titers of antibody to H. parasuis serovars 4, 5, and 12 in pregnant sows, 6 of which received 2 mL of the F1 vaccinea (Z1) while the other 6 served as controlsb (Z2)

Sow Day after first vaccination (0 or 21) or 2nd vaccination; mean antibody titerc

Serovar group 0 21 7 14 28 60 90 120 150 180 210 240 4 Z1 — 1:4 1:8 1:16 1:64 1:64 1:32 1:32 1:16 1:16 1:8 1:4 5 Z1 — 1:4 1:8 1:16 1:128 1:64 1:64 1:32 1:32 1:16 1:16 1:4 12 Z1 — 1:4 1:8 1:16 1:128 1:64 1:64 1:32 1:16 1:16 1:16 1:4 Z2 — — — — — — — — — — — —a The vaccine was given as 2 intramuscular doses 21 d apart.b The controls were injected with 2 mL of sterile phosphate-buffered saline the same days as the other sows were vaccinated.c Antibody titers $ 1:8 and # 1:4 were deemed to be positive and negative, respectively. The dash indicates that no H. parasuis antibodies were detected.



this vaccination regimen provided protection against H. parasuis serovars 4, 5, and 12 for 180 d. Our results showed that the vacci-nated offspring of the vaccinated sows had longer-lasting serologic protection, which would help to prevent H. parasuis colonization.

Our findings provide new, detailed data regarding the vaccine optimization that is needed to guarantee the efficacy of a trivalent vaccine against Glässer’s disease in piglets and pregnant sows containing inactivated H. parasuis serovars 4, 5, and 12, and they provide a basis for developing a commercial vaccine. Meanwhile, the practical significance of our findings is that if H. parasuis infections are endemic in a swine-breeding facility it is necessary to vaccinate the sows before farrowing. The piglets could be protected during the finishing period if they are vaccinated during lactation. Field trials of the trivalent vaccine that we developed, administered with Montanide GEL 01 PR adjuvant, should be conducted on pig farms in the future.

Conflict of interest statementNone of the authors has any financial or personal relationships

that could inappropriately influence or bias the content of the paper.

A c k n o w l e d g m e n t sWe thank Xiaojian Xi for his excellent technical assistance and

excellent animal care. This study was supported by the Open Funds of State Key Laboratory of Veterinary Etiological Biology (grant SKLVEB2013KFKT009), the National Natural Science Foundation of China (grants 31302106 and 31001051), and the Research and Development Foundation of Henan University of Science and Technology (grant 2015ZDCXY04).

Re f e r e n c e s 1. Martinez-Martinez S, Frandoloso R, Rodriguez Ferri EF, et al.

Immunoproteomic analysis of the protective response obtained with subunit and commercial vaccines against Glässer’s disease in pigs. Vet Immunol Immunopathol 2013;151:235–247.

2. Brockmeier SL, Loving CL, Mullins MA, et al. Virulence, transmission, and heterologous protection of four isolates of Haemophilus parasuis. Clin Vaccine Immunol 2013;20:1466–1472.

3. Brockmeier SL, Register KB, Kuehn JS, et al. Virulence and draft genome sequence overview of multiple strains of the swine pathogen Haemophilus parasuis. PloS One 2014;9:e103787.

4. Howell KJ, Peters SE, Wang J, et al. Development of a multiplex PCR assay for rapid molecular serotyping of Haemophilus para-suis. J Clin Microbiol 2015;53:3812–3821. Epub 2015 Sep 30.

5. Cai X, Chen H, Blackall PJ, et al. Serological characterization of Haemophilus parasuis isolates from China. Vet Microbiol 2005; 111:231–236.

6. Zhang J, Xu C, Guo L, et al. Prevalence and characterization of genotypic diversity of Haemophilus parasuis isolates from south-ern China. Can J Vet Res 2012;76:224–229.

7. Chen S, Chu Y, Zhao P, et al. Development of a recombinant OppA-based indirect hemagglutination test for the detection of antibodies against Haemophilus parasuis. Acta Trop 2015;148:8–12.

8. Li M, Song S, Yang D, Li C, Li G. Identification of secreted proteins as novel antigenic vaccine candidates of Haemophilus parasuis serovar 5. Vaccine 2015;33:1695–1701.

9. Martin de la Fuente AJ, Gutiérrez Martin CB, Pérez Martínez C, Garcia Iglesias MJ, Tejerina F, Rodríguez Ferri EF. Effect of dif-ferent vaccine formulations on the development of Glässer’s disease induced in pigs by experimental Haemophilus parasuis infection. J Comp Pathol 2009;140:169–176.

10. Hu M, Zhang Y, Xie F, et al. Protection of piglets by a Haemophilus parasuis ghost vaccine against homologous challenge. Clin Vacc Immunol 2013;20:795–802.

11. Murtaugh MP. Advances in swine immunology help move vac-cine technology forward. Vet Immunol Immunopathol 2014; 159:202–207. Epub 2014 Feb 23.

12. Huang X, Li Y, Fu Y, et al. Cross-protective efficacy of recom-binant transferrin-binding protein A of Haemophilus parasuis in guinea pigs. Clin Vaccine Immunol 2013;20:912–919.

13. Frandoloso R, Martínez-Martínez S, Yubero S, Rodríguez-Ferri EF, Gutiérrez-Martín CB. New insights in cellular immune response in colostrum-deprived pigs after vaccination with subunit and commercial vaccines against Glässer’s disease. Cell Immunol 2012;277:74–82.

14. Yuan F, Fu S, Hu J, et al. Evaluation of recombinant proteins of Haemophilus parasuis strain SH0165 as vaccine candidates in a mouse model. Res Vet Sci 2012;93:51–56.

15. Xue Q, Zhao ZQ, Liu HS, Chen KP, Xue Y, Wang L. First com-parison of adjuvant for trivalent inactivated Haemophilus para-suis serovars 4, 5 and 12 vaccines against Glässer’s disease. Vet Immunol Immunopathol 2015;168:153–158.

16. Xu HJ. Screen of Inactivated Haemophilus parasuis Candidate Vaccine [MD dissertation]. Luoyang, China: Henan University of Science and Technology, 2013.

17. Cai XW. Isolation and Characterization of Haemophilus parasuis and Development of Its Diagnostic Method and Inactivated Bacterin [PhD dissertation]. Hubei, China: Huazhong Agricultural University, 2006.

18. Nielsen R. Pathogenicity and immunity studies of Haemophilus parasuis serotypes. Acta Vet Scand 1993;34:193–198.

19. Solano-Aguilar GI, Pijoan C, Rapp-Gabrielson V, Collins J, Carvalho LF, Winkelman N. Protective role of maternal antibod-ies against Haemophilus parasuis infection. Am J Vet Res 1999; 1:81–87.

20. Baumann G, Bilkei G. Effect of vaccinating sows and their piglets on the development of Glässer’s disease induced by a virulent strain of Haemophilus parasuis serovar 5. Vet Rec 2002;1:18–21.


Article


Evaluation of adipose-derived stromal vascular fraction from the lateral tailhead, inguinal region, and mesentery of horses

Garrett L. Metcalf, Scott R. McClure, Jesse M. Hostetter, Rudy F. Martinez, Chong Wang

A b s t r a c tUse of mesenchymal stem cells (MSCs) found in the stromal vascular fraction (SVF) of equine adipose tissue has promising applications for regenerative therapies. The most commonly used source of equine adipose tissue is the subcutaneous tailhead. The objective of this study was to compare 3 adipose depot sites in horses and determine the viability and cellular yield, capillary density, gene expression for selected markers, and colony-forming unit fibroblasts (CFU-Fs) in adipose tissue taken from these sites. Adipose tissue was excised from the area lateral to the tailhead, the inguinal region, and the small colon mesentery of 6 horses. Lipoaspirate was also collected from the area lateral to the tailhead. Stromal vascular fraction (SVF) was prepared in duplicate from the 3 different adipose tissue depots. The total nucleated and dead cell counts was determined manually using a hemocytometer and percent viability was calculated. Mass and volume of adipose were determined in order to calculate density and factor-VIII immunohistochemical staining was used to determine vascular density in the excisional adipose tissue samples from each horse. Quantitative polymerase chain reaction (qPCR) was used to quantify gene expression for selected cellular markers from each site. There were significant differences in viability, yield of nucleated cells/gram of adipose tissue, vascular density, gene expression, and CFU-Fs among adipose depots. Adipose from the mesentery yielded the highest number of nucleated cells/gram of tissue and the highest vascular density and percentage of CFU-Fs. In the horse, both the anatomical site of collection and the method of tissue collection significantly impact the yield and composition of cells in the SVF. Further study is needed to assess whether one adipose source is superior for harvesting mesenchymal stem cells (MSCs) and whether the differences among sources are clinically relevant for in-vivo treatment of musculoskeletal injuries in horses.

R é s u m éL’utilisation de cellules souches mésenchymateuses (CSMs) retrouvées dans la fraction du stroma vasculaire (FSV) du tissu adipeux équin a des applications prometteuses pour les thérapies régénératrices. La source la plus fréquemment utilisée de tissu adipeux équin est le tissu sous-cutané de la base de la queue. L’objectif de la présente étude était de comparer trois sites de dépôts adipeux chez le cheval et de déterminer la viabilité et la récolte cellulaire, la densité capillaire, l’expression génique de marqueurs sélectionnés, et le nombre de fibroblastes formateur des colonies (FFC) dans le tissu adipeux prélevés de ces sites. Le tissu adipeux a été excisé de la région latérale à la base de la queue, de la région inguinale, et du mésentère du petit colon de six chevaux. Des aspirations de lipide ont également été prélevées de la région latérale de la base de la queue. La FSV a été préparée en duplicata à partir de chacun des trois dépôts différents de tissu adipeux. Les dénombrements totaux des cellules nucléées et mortes ont été déterminés manuellement à l’aide d’un hémocytomètre et le pourcentage de viabilité calculé. La masse et le volume de tissu adipeux ont été déterminés afin de calculer la densité et la coloration par immunohistochimie du facteur VIII a été utilisée afin de déterminer la densité vasculaire dans les échantillons de tissu adipeux excisé de chaque cheval. Une réaction d’amplification en chaine par la polymérase quantitative (ACPq) a été utilisée pour quantifier l’expression génique pour des marqueurs cellulaires sélectionnés de chaque site. Il y avait des différences significatives dans la viabilité, le rendement de cellules nucléées/gramme de tissu adipeux, la densité vasculaire, l’expression génique, et les FFCs entre les dépôts adipeux. Le tissu adipeux provenant du mésentère a généré le plus grand nombre de cellules nucléées/gramme de tissu et la plus haute densité vasculaire et pourcentage de FFCs. Chez le cheval, le site anatomique de prélèvement et la méthode de prélèvement du tissu ont un impact significatif sur le rendement et la composition cellulaire dans la FSV. Des études additionnelles sont requises pour évaluer si une source de tissu adipeux est supérieure pour récolter des cellules souches mésenchymateuses et si les différences entre les sources sont cliniquement pertinentes pour le traitement in vivo de blessures morpho-squelettiques chez les chevaux.


Department of Veterinary Clinical Sciences (Metcalf, McClure), Department of Pathology (Hostetter), Department of Diagnostics and Production Animal Medicine and Department of Statistics (Wang), Iowa State University, Ames, Iowa, USA; InGeneron Inc., 8205 El Rio, Houston, Texas, USA (Martinez).

Address all correspondence to Dr. Scott McClure; telephone: (515) 231-4680; e-mail: [email protected]

Conflict of interest Rudy Martinez, who completed the polymerase chain reaction (PCR) and colony-forming unit fibroblast (CFU-F) assay for this study, is employed by InGeneron. The authors who completed the remainder of the study and compiled the data have no financial interests or holdings in InGeneron.

Received April 11, 2016. Accepted May 19, 2016.



I n t r o d u c t i o nRegenerative cell-based therapy is a promising tool for treating

musculoskeletal injuries in equine athletes and has the potential to regenerate or modulate healing of tissue. Superficial digital flexor tendonitis is an example of a common injury in the horse industry with poor healing, long convalescence periods, loss of earnings, and high reinjury rates either in the same or contralateral limb (1–3). Mesenchymal stem cells (MSCs) have potential for treating mus-culoskeletal injuries in horses in a wide variety of disciplines (4–7).

There are numerous tissues that contain MSCs, but the most prac-tical and widely used tissues are bone marrow and adipose. There are also many techniques for obtaining MSCs. The most commonly used techniques are culture-expanded bone marrow and adipose-derived stem cells in the stromal vascular fraction (SVF), which contains mesenchymal stem cells and a variety of nucleated cells that can contribute to tissue healing (8–10). Since there are relatively few MSCs in bone marrow, culture expansion is indicated to obtain the desired number of MSCs for implantation (8,10,11). There are data that support MSCs derived from bone marrow for specific applica-tions (12). When the SVF is used for immediate implantation, there is potential for high MSC yield.

Data indicate that up to 222 times more MSCs are found in adipose tissue than in equal amounts of bone marrow in the horse and in humans up to 500 times more MSCs are found in adipose than in bone marrow (8). Because adipose tissue has a relatively high density of MSCs, fresh SVF can be used for immediate implantation without expansion, which is an attractive feature of SVF regenerative therapy (13,14). A number of factors must be considered when choosing the source and technique, including timing of implantation and time required to obtain cells, harvest location, patient morbidity, and efficacy. These factors affect the clinician’s selection of cultured MSCs or fresh SVF. Regardless of the source, the objective is to maximize the MSC harvest.

When bone marrow is used, it has been shown that different harvest locations vary in nucleated cell counts (15). It is not known, however, if this is also true for adipose harvest sites. Adipose is an abundant source of mesenchymal stem cells, with depots in the horse found lateral to the tailhead, the inguinal region, neck, omentum, and mesentery. Historically, the harvest site of choice for adipose-derived mesenchymal stem cells (MSCs) or stromal vascular fraction (SVF) in horses is lateral to the tailhead (16–18). Fat lateral to the tailhead varies in vascular density, which has in turn been shown to correlate with MSC colony-forming units (CFUs) in vitro (16). There is limited information, however, on the optimal location for harvesting adipose to obtain the SVF in horses and other species.

Certain clinical situations can dictate the harvest site of adipose tissues, including cosmetic appearance, convenience, adequate adi-pose depot for collection, and positioning of a horse that is already under general anesthesia. The concern with using alternative sites for adipose harvesting is the quality and quantity of cell yield. It has been found in humans that not all fat depot sites exhibit equivalent characteristics and there are differences in surface marker expression, doubling time, and SVF yield among subcutaneous, omental, and intrathoracic fat depot sites (19). Additionally, use of lipoaspiration can speed the adipose recovery process and result in a very small

incision lateral to the tailhead, which improves cosmetic outcome and decreases potential morbidity, although the lipoaspiration pro-cedure itself could affect cell yield and viability (18).

The objective of this study was to evaluate the nucleated cell frac-tions in adipose tissue obtained from the tailhead (excisional and lipoaspirate), inguinal region, and mesentery of the same horse and compare the viability and cellular yield, tissue and vascular density, gene expression for selected markers, and CFU-F in this tissue. We hypothesized that adipose depot sites with higher vascular density would yield a higher concentration of nucleated cells than sites with lower vascular density.


AnimalsSix healthy horses were used in this study, as determined by

physical examination and known health histories. It was required that the horses not have any prior history or evidence of intra-abdominal surgery or asymmetry over the gluteal or tailhead region. The median age was 3 y (1 to 15 y) and mean 6 standard deviation (SD) weight was 482 6 86.1 kg (318 to 559 kg). Breed distribution was 5 Quarter Horses and 1 Missouri Fox Trotter. All horses had been euthanized for reasons not associated with this study. Euthanasia procedures were approved by the Iowa State University Institutional Animal Care and Use Committee, which complies with the American Veterinary and Medical Association (AVMA) Guidelines for the Euthanasia of Animals.

Collection of adipose tissueFollowing euthanasia, sites over the tailhead, inguinal area, and

abdomen were immediately clipped and aseptically prepared. Enough adipose tissue to fill a 50-mL conical tube was harvested from the area lateral to the tailhead, the inguinal region, and the small colon mesentery by surgical excision. Additionally, adipose tissue was collected by lipoaspiration from the contralateral tailhead by making a 1-cm stab incision with a #15 scalpel blade 8 cm from midline adjacent to the tailhead. The adipose tissue was aspirated with a lipoaspirate cannula (3-mm Disposable Cannula; Shippert Medical Technologies, Centennial, Colorado, USA) and a 60-mL luer lock syringe (VacLok; Merit Medical Systems, South Jordan, Utah, USA), without the use of tumescent fluid, to collect approximately 10 to 15 mL of adipose tissue.

Excisional adipose tissue samples from each location were divided into 2 portions, 1 for immunohistochemical (IHC) analysis and the other for cell isolation. The adipose for IHC was fixed in 10% buff-ered formalin. The lipoaspirate was evaluated for cell isolation only and not used for density or IHC evaluation.

Nucleated cell isolation and countingThe weight of each sample for cell isolation was determined

before enzymatic processing. The lipoaspirates were processed in aliquots of approximately 10 g using a commercially available system (ARC System; InGeneron, Houston, Texas, USA) to isolate SVF cells from the adipose tissue. Lipoaspirates were incubated at 37°C and agitated in the processing unit with a proprietary blend



of proteolytic enzymes (Matrase reagent; InGeneron) for 30 min to dissociate the tissue and release the nucleated cells, which were then separated from the solid tissue by filtration (100 mm filter, Steriflip; EMD Millipore, Bilerica, Massachusetts, USA). The excisional adi-pose tissue samples used for cell isolation were minced with Mayo scissors and processed in approximately 10-gram aliquots. Similar to lipoaspirates, the minced excisional adipose samples were processed with Matrase reagent (InGeneron) at 37°C, with the exception that processing time was 60 min, following the manufacturer’s recom-mendation for excisional adipose tissue.

After processing and filtration, all samples were treated similarly by adding approximately 25 mL of Lactated Ringer’s solution (LRS) to reach a total volume of 50 mL to dilute enzymes and achieve bal-ance and centrifuged for 10 min at 600 3 g. The cells were washed twice by centrifugation and resuspension of the cell pellet in 40 mL of LRS each time. The final cell pellet was resuspended in 5 mL of LRS. Samples from each tissue site for each horse were run in duplicate.

To obtain nucleated cell counts, cells were stained with fluores-cent nucleic acid stain (SYTO 13; Life Technologies/Thermo Fisher Scientific, Waltham, Massachusetts, USA) using a 1:1 solution and then counted using a hemocytometer and standard light micro-scope at 403 magnification with a portable imaging cytometer (Bioscope; InGeneron). Non-viable cells were quantified by prepar-ing a 1:1 solution of the cell suspension from each sample in 0.4% Trypan Blue solution. Non-viable cells were counted using a hemo-cytometer under light microscopy at 403 magnification. Nucleated cell viability was calculated using the following formula: (Nucleated Cells − Dead Cells)/Nucleated Cells 3 100.

The remaining cells not used for counting were centrifuged at 600 3 g for 10 min and the supernatant removed. The final stromal vascular fraction (SVF) was suspended in 2 mL of LRS and aliquots were mixed with an equal volume of Trizol reagent (TRIzol; Life Technologies/Thermo Fisher Scientific) to lyse cells and preserve the ribonucleic acid (RNA) and then frozen at −80°C until gene expression was determined or cryopreserved in cryopreservation media containing 5% (v/v) dimethyl sulfoxide (DMSO) for later analyses in cell culture (CryoStor CS 5; BioLife Solutions, Bothell, Washington, USA).

Adipose tissue densityThe mass of each sample was determined using a digital elec-

tronic scale (PB303; Mettler-Toledo, Columbus, Ohio, USA) and the samples were then submerged in a known volume of deionized water at a temperature of 37°C in a graduated cylinder. A stylet from a 20-gauge, 8-in spinal needle was used to submerge the adipose tissue and this was used each time for consistency. The volume of water displaced by the sample was used to determine the density of the sample, using the formula Density = Mass/Volume.

ImmunohistochemistryThe fixed adipose tissues were embedded in paraffin and rou-

tinely sectioned at ~5 to 8 mm, then mounted on glass slides for IHC factor-VIII staining, which was done manually. A series of degrading alcohol steps was used to dehydrate the sections. The blocking step with 10% goat serum was carried out before incubation with primary antibody. The slides were incubated for 120 min with 1:100 rabbit

anti-human polyclonal factor-VIII antibodies (Factor VIII; Dako North America, Carpenteria, California, USA), which have been shown to cross-react with equine tissue. After incubation, slides were rinsed and anti-rabbit mouse immunoglobulin G (IgG) secondary antibody (HK638-UK; BioGenex, Fremont, California, USA) was applied and allowed to incubate for 15 min. Horseradish peroxidase (Life Technologies/Thermo Fisher Scientific) was applied to tissue sections for 15 min. NovaRED peroxidase (Vector Laboratories, Burlingame, California, USA) was applied to the sections for 5 min and the sections were then washed with deionized water and counterstained with hematoxylin for 2 min. The factor VIII-stained adipose tissue was evaluated under light microscopy by 1 evalu-ator who was blinded (GM). The factor VIII-stained capillaries were manually counted at 403 power in 3 fields of view that were randomly selected.

Quantitative PCRThe relative gene expression of MSC surface markers CD44, CD73,

CD90, CD105, CD146, CD166, hematopoietic markers CD34 and CD45, endothelial cell marker CD31, along with genes associated with adipogenic (PPARG2), osteogenic (OSTEOCALCIN) lineages, as well as type-1 collagen (Col 1A2) was measured in total RNA from each adipose source.

For each condition tested, 2 replicates of total RNA were isolated from freshly isolated cells and measured using spectrophotometry (Nanodrop ND-100; Thermo Fisher Scientific, Wilmington, Delaware, USA). First strand complementary deoxyribonucleic acid (cDNA) synthesis (First Strand Kit; Qiagen, Chatsworth, California, USA) was carried out using 0.500 mg of total RNA, according to the manufacturer’s instructions. All individual samples were analyzed in duplicate. Each plate contained both a no-reverse transcriptase (RT) control and no-template control and the gene used as constitu-tive baseline was b-actin. Quantitative PCR was carried out using a thermal cycler (iCycler; Bio-Rad, Hercules, California, USA) and a complete amplification mix (Green Supermix; Bio-Rad), with primer sequences listed in Appendix A (Sigma Genosys; Sigma-Aldrich, St. Louis, Missouri, USA).

Colony-forming unit fibroblast assayThe assay for colony-forming unit fibroblasts (CFU-Fs) was

adapted from an established method (20). Cells were plated and CFU-Fs were induced after first passage. For each adipose sam-ple, cells from the first passage were plated at 3 cell densities: 250 000 cells, 50 000 cells, and 10 000 cells per 35-mm diameter well and repeated twice for each density. Cells were incubated for 14 d in standard growth media with 10% horse serum. Media were changed twice weekly. The cells were fixed with 4% formalin, stained with hematoxylin for 10 min, washed with phosphate-buffered saline (PBS), and pictures were obtained from 5 randomly chosen sections in each well for counts using a 2.53 objective on a light microscope. Two different experienced investigators, blinded to cell source, counted all visible colonies individually. A colony-forming unit was defined as a cluster containing at least 10 fibroblast-like, fusiform cells. After counting, the density of 10 000 cells per well was selected for further evaluation. The CFU-Fs were expressed as CFU-F% per 1 3 106 adherent cells.



Statistical analysisData were entered into a spreadsheet (Microsoft Office Excel;

Microsoft, Redmond, Washington, USA) and imported into a sta-tistical software package for analysis (SAS; SAS Institute, Cary, North Carolina, USA). Responses were analyzed using linear-mixed models, with location and age as the fixed effects and horse as the random effect. For variables found to be significant in the model, comparisons among locations were assessed using an F-test, fol-lowed by Tukey pairwise comparisons. The qPCR data were normal-ized to b-actin and quantitated with the delta-delta cycle threshold (CT) method. There were 3 groups for comparison of density and immunohistochemistry (IHC), because the physical material of the lipoaspiration did not permit determination of these variables. The CFU-F data were evaluated with a 1-way analysis of variance (ANOVA) using tissue source as the main effect and a Tukey’s post-hoc test. A P of # 0.05 was considered significant.

Re s u l t sThe source of the adipose resulted in significantly different nucle-

ated cells/gram of tissue {P = 0.0001 [mean values 6 standard error (SE): lipoaspirate = 761 667 6 227 348; tailhead = 1 900 000 6 227 348; mesentery = 2 608 333 6 227 348; inguinal = 992 333 6 227 348]} and percent viable [P = 0.0150 (lipoaspirate = 86.8 6 3.3; tailhead = 92.5 6 3.3; mesentery = 96.8 6 3.3, inguinal = 91.1 6 3.3)], but not for tissue density (g/cm3) [P = 0.0666 (tailhead = 0.94 6 0.008; mes-entery = 0.92 6 0.008; inguinal = 0.938 6 0.008)] (Figure 1). The IHC

Table I. The relative expression of each gene for the stromal vascular fraction (SVF) from each of the 4 sources

Gene Lipoaspiration Tailhead Mesentery Inguinal P-valueCD44 4.7a 3.66b,c 4.06a,c 3.78b,c 0.0095CD73 7.02a 5.75b 5.21b 5.05b , 0.0001CD90 4.35a 2.87b,c 3.14b 2.25c , 0.0001CD105 4.16a 4.32a 4.89b,c 4.61a,c 0.007CD146 4.5a 4.85a 5.79b,c 5.09a,c 0.0041CD166 6.26 6.26 6.23 5.75 0.423CD34 2.97a 2.91a 3.40a,b 2.38a,c 0.0067CD45 6.65a 5.93a 5.06b 5.58a 0.0006CD31 2.05 2.29 2.88 2.08 0.1764PPARG2 4.92 4.48 5.17 4.36 0.1934OSTEO 11.11 11.73 11.29 10.78 0.4789Col 1A2 2.07a 0.57b 3.7c 2.12a , 0.0001The qPCR data were normalized to b-actin and quantitated with the delta-delta cycle threshold (CT) method. As the data in the table are the expression of the gene relative to b-actin, for the CD44 lipoaspirate sample, the expression was 4.7 times that of the housekeeping gene b-actin.P-value in each row shows that the significance for the linear-mixed model and of the values with different superscript letters in the same row are significantly different based on Tukey pairwise comparison. Significance was set at P # 0.05. The superscript letters denote significance.

Figure 1. The mean (6 SE) yield of nucleated cells/gram of tissue (A) and percent viability (B) for lipoaspirate, tailhead, inguinal, and mes-enteric adipose. The density (C) and factor-VIII immunohistochemistry (IHC) (D) were obtained for tailhead, inguinal, and mesenteric adipose only because the lipoaspiration from the tailhead was represented by the solid sample from the tailhead and lipoaspiration was not physically conducive to obtaining these data. Columns with different superscript letters denote a significant difference.

Lipoaspirate Tailhead Mesentery Inguinal


Tailhead Mesentery Inguinal

Tailhead Mesentery Inguinal

Nuc

leat

ed c

ells

/gP

erce

ntg/

mL

Mea

n ca

pilla

ries

/hpf

Nucleated cells/gram

Viability

Density

Factor VIII IHC

3 500 000

3 000 000

2 500 000

2 000 000

1 500 000

1 000 000

500 000

0

100

95

90

85

80

75

0.95

0.94

0.93

0.92

0.91

0.90

0.89

30

25

20

15

10

5

0

a

a

b

b,c

a,ba,b

a,c

b

A

B

C

D

a

a

a

aa

b



showed a significantly different capillary density (capillaries/hpf) among locations, as shown by factor-VIII staining [P , 0.0001 (tail-head = 20.6 6 1.56; mesentery = 23.6 6 1.56; inguinal = 10.6 6 1.56)] (Figure 1).

For the qPCR outcome data, adipose source was signifi-cantly related to differences found for CD 44 (P = 0.0095), CD 45 (P = 0.0006), CD 73 (P # 0.0001), CD 90 (P # 0.0001), CD 105 (P = 0.007), CD 146 (P = 0.0041), Col1A2 (P # 0.0001), and CD 34 (P = 0.0067) (Table I). Osteocalcin was the only outcome variable significantly related to age (P = 0.048). The mean osteocalcin expres-sion of the 15-year-old horse was 10.2 compared to 11.5 for all the remaining samples.

The CFU-F assay also resulted in significantly different frequen-cies of CFU-F formation among the various adipose sources. The mean (6 SD) CFU-F% for the liposuction was 7.0% (6 0.11) and for the mesentery was 8.59% (6 0.11), both of which were significantly higher than the excisional adipose from the tailhead (2.42% 6 0.04) and inguinal (2.11% 6 0.06) locations (Figure 2).

A subjective finding not included as an outcome parameter but identified during the study was that adipose obtained by liposuc-tion and from the mesentery could easily be filtered after enzymatic processing, whereas excisional adipose from the tailhead occasion-ally required a second filter because of clogging from fibrous tissue. All the inguinal samples required multiple filters, which suggests a denser connective tissue matrix.

D i s c u s s i o nThe data obtained in this study confirmed our hypothesis that the

more vascular, dense adipose tissue would yield a higher concen-tration of nucleated cells. We found that there are significant differ-ences among locations for nucleated cell yield per gram of adipose, viability, gene expression, and CFU-F%. We also found that fewer capillaries, as identified by factor-VIII IHC in the inguinal adipose, corresponded to the lowest yield of nucleated cells/gram, which is similar to results of previous studies (16).

The mesenteric adipose tissue provided the highest nucleated cells/gram of tissue and viability, which corresponds with the IHC vascular density findings among the tissue sites. The nucleated cells/

gram in the excisional adipose from the tailhead were statistically similar to those in the mesentery and were significantly higher in both sites than in the lipoaspirate from the tailhead. The viability and yield of the lipoaspirated adipose tissue from the tailhead were the lowest of the 4 groups. Physical destruction during the lipoaspi-ration process can contribute to lower recovery and viability. It is well-established for human subcutaneous adipose tissue that harvest method can affect cell yield (21). The lower yield associated with mechanical trituration of adipose tissue caused by lipoaspiration has been previously described (18) and was similarly seen in this study.

We chose not to use tumescent fluid in this study because its anal-gesic properties were unnecessary, as the tissues were harvested from euthanized horses and we could accurately determine nucleated cells/gram of tissue. This could cause decreased yield and viability, however, as the tumescent fluid acts as a lubricant and buffer from physical damage during the aspiration process and speeds up the harvest of the adipose, which decreases the tissue damage created by the motion of the cannula. In a study comparing adipose collection techniques in humans, however, liposuction with tumescent fluid and adipose collected by resection were not significantly different in terms of viability and cellular yield of MSCs, while doubling times were significantly longer with ultrasound-assisted liposuction (22).

Tissue density was determined because it was initially thought that adipose with a greater percentage of adipose vacuoles and therefore a lower density would contain less vasculature and fewer nucleated cells. We found the opposite in that the tissue with the higher density had lower nucleated cells/gram because it was higher in dense fibrous connective tissue. Although the differences in densities were not significant, the mesenteric adipose tissue was much more easily digested and had less fibrous tissue. The lipoaspirated adipose tissue from the tailhead behaved similarly to the mesenteric adipose tissue during digestion and filtration. There was more connective tissue within the inguinal adipose during its harvest, mincing, and filtering. This was also evident histologically and corresponded with the higher density of the inguinal tissue. It is intuitive to think that the inguinal adipose tissue generated lower nucleated cells/gram due to the fibrous connective tissue, which would lower cellular yield.

It is known that most of the MSCs in bone marrow are in the ini-tial segment of the collection and a small percentage of nucleated cells are MSCs (15,23). An advantage of any of the adipose sources is that there can be large deposits with a consistent yield of cells/gram of tissue that do not decline like bone marrow sources. In order to gain more MSCs from adipose, more adipose can be harvested and processed. The results of this study showed a similar cell yield for the lipoaspirate from the tailhead (7.65 3 105) as in a previous study (7.9 3 105) (18). The previous study demonstrated a 23-times increase in adherent cells and 50 times as many CFU-Fs compared to an equal number of nucleated cells derived from bone marrow.

The value of adipose-derived stromal vascular fraction (SVF) for regenerative medicine is emphasizing the need to further character-ize the SVF. It is recognized that there are changes in surface markers between the SFV and cultured adipose-derived stem cells (24). The data presented here are the first data obtained from equine adipose and provide information about surface markers on the equine SVF that can be used in future studies. Differences in gene expression

Figure 2. The mean (6 SD) for the colony-forming unit fibroblasts (CFU-Fs) from each source. The data are presented as CFU-F% per 1 3 106 adherent cells. Cells collected from the mesentery and by lipoaspiration had a 3 to 4 times greater amount of CFU-Fs than cells isolated from the tailhead and inguinal adipose depots. Columns with different superscript letters denote a significant difference.


Per

cent

Percent CFU-F

12

10

8

6

4

2

0

a

b

a

b



were found among the sources. The SVF from all tissue sources expressed CD73, CD90, and CD105 cell surface markers, which according to the International Society for Cellular Therapy, denotes the presence of MSCs (25). Although there is no specific surface marker group for equine stem cells, in those used for this study, the lipoaspirate sample showed the significantly highest expression for CD44, CD73, and CD90. This indicates that, despite somewhat lower cell yields, there may be a higher percentage of MSCs in the SVF obtained by lipoaspiration. The CFU-F data also demonstrated that there was a higher percentage of CFU-Fs in the lipoaspirate and the mesenteric samples.

Since these cells are from the SVF with multiple cell types pres-ent, we expected expression of hematopoietic markers (25). The hematopoietic markers were highest in the tailhead (CD45) and the mesentery (CD34) and lower in the inguinal region, similar to the IHC factor-VIII staining for capillaries. The niche concept, where stem cells are located close to the vasculature, is consistent with our findings (26,27). As the perivascular environment may provide the conditions that maintain the population of undifferentiated MSCs, the SVFs with the maximal MSC yield are likely to contain numerous cells positive for CD34 and CD45.

It is difficult to compare our results for gene expression with other studies of equine SVF because of the variation in isolation methods and cellular surface markers being studied. Some of the many types of cells present in the SVF can also stimulate healing, similar to MSCs. The stromal vascular fraction contains fibroblasts, endothelial cells, leukocytes, and very small embryonic-like cells with additional cytokine production and healing capabilities that are not found in culture-expanded stem cells (28–30). In addition to the cells being transplanted when using the SVF, a recent study showed the ability of the SVF to induce or enhance MSC migration and proliferation in vitro (14). It was speculated that mediators released from the SVF resulted in the marked migration of the MSCs in the culture.

The objective of this study was to make a relative comparison of adipose depots for harvesting SVF in the horse using qPCR to determine the relative gene expression. While immunophenotyp-ing of cells by flow cytometry could be complementary to the data obtained in this study, quantitative PCR was used in this study for a number of reasons. In studies where both methods have been used, outcomes have been similar (31,32). As there is no uniformly accepted definitive phenotype or surface marker for isolating MSCs from uncultured samples, they have been phenotypically identified retrospectively (33). Cell surface markers change with culture and at this time we do not have a composite phenotype of freshly isolated MSCs or a single marker that solely identifies native MSCs (34). In 1 study in which fresh bone marrow aspirate was followed over time, gene expression followed the same pattern as cellular protein expression at each time point (31). There are relatively few antibodies that have been validated in the horse and the relative expression of qPCR data has provided a viable comparison when antibodies have failed to bind equine cells (20,35).

Quantitative PCR has been used previously for cell surface mark-ers of MSCs from multiple equine tissues (32,35–37). The expression of the surface markers CD90 and CD105 as characteristic of equine MSCs has also been determined by qPCR (37). In the heterogenous

population of the SVF, the relative expression of the genes evaluated provides information about differences in the cell populations at the different locations. Because of potential issues with cellular debris and the multiple cell types present in fresh SVF, the question of the validity for multiple antibodies in the horse, and the lack of previous data using flow cytometry in fresh SVF of the horse, qPCR provided the most valuable information for this study.

The CFU-F is the standard for defining the number of progenitor cells in the SVF (24). The frequency of CFU-F from human adipose ranges from 1% to 10%. Our data from the horse fall into this range, with inguinal SVF at 2.11% and mesenteric SVF at 8.9%. The SVF from the mesentery and lipoaspirate was statistically similar and both were higher than that from the solid adipose from the tailhead and inguinal region. The mesenteric adipose had the highest cell yield, viability, and factor-VIII staining and the lowest density of CFU-Fs.

We selected locations that would represent abundant adipose depots and flexibility in harvesting during different recumbencies. The mesentery was included as a harvest site because many ath-letic horses have limited subcutaneous deposits. Furthermore, the other adipose tissues were of similar subcutaneous origin and the abdomen represented a source that may be expected to be different based on location and gross appearance. Mesenteric adipose can be harvested with the horse in dorsal recumbency during a surgical procedure when it is elected to obtain a SVF or it can be obtained in other horses with limited invasiveness by laparoscopy.

This study was designed to compare autogenous cell yield from different locations within a horse to determine if there is an optimum source. We did not select horses based on age, but there was a horse that was 15 y old in the study. No relationship was found between cell yield and age. It has been shown that age can affect the number of MSCs obtained from bone marrow, as the ilium is found to deplete with age in equids, which makes it in an unreliable source of bone-marrow collection in older horses. In humans, the CFU-F numbers also significantly decrease in older age groups (38,39).

A potential limitation of this study is that the multipotency of the mesenchymal stem cells (MSCs) present in the stromal vascular fraction (SVF) was not evaluated. As previous studies verified the multipotency of adipose-derived MSCs (18), however, it was not pursued in this study. Additionally, a larger group of horses with a more extensive age range could allow the effect of age on MSC yield to be further evaluated.

In this study, significant differences were found in viability, cel-lular yield, vascular density, gene expression, and CFU-F among adipose tissue depots in the horse. The clinical application of this information indicates that there are multiple sources of adipose in horses that can be used to collect SVF. Further work is needed to determine whether the differences among adipose sources observed in this study are clinically relevant for in-vivo treatment of musculo-skeletal injuries in horses.

A c k n o w l e d g m e n t sThe authors thank Lisa Leake, Dr. Michael Coleman, and Maria

P. Almario for their technical support. This study was supported in kind by Ingeneron, Inc.



Re f e r e n c e s 1. Dyson SJ. Medical management of superficial digital flexor ten-

donitis: A comparative study in 219 horses (1992–2000). Equine Vet J 2004;36:415–419.

2. Avella CS, Ely ER, Verheyen KL, Price JS, Wood JL, Smith RK. Ultrasonographic assessment of the superficial digital flexor tendons of National Hunt racehorses in training over two racing seasons. Equine Vet J 2009;41:449–454.

3. Ross MW, Genovese RL, Dyson SJ, Jorgensen JS. Superficial digital flexor tendonitis. In: Ross MW, Dyson SJ, eds. Diagnosis and Management of Lameness in the Horse. 2nd ed. St. Louis, Missouri: Saunders Elsevier, 2011:706–726.

4. Pacini S, Spinabella S, Trombi L, et al. Suspension of bone marrow-derived undifferentiated mesenchymal stromal cells for repair of superficial digital flexor tendon in race horses. Tissue Eng 2007;13:2949–2955.

5. Smith RK. Mesenchymal stem cell therapy for equine tendinopa-thy. Disabil Rehabil 2008;30:1752–1758.

6. Godwin EE, Young NJ, Dudhia J, Beamish IC, Smith RK. Implan-tation of bone marrow-derived mesenchymal stem cells demon-strates improved outcome in horses with overstrain injury of the superficial digital flexor tendon. Equine Vet J 2012;44:25–32.

7. Renzi S, Ricco S, Dotti S, et al. Autologous bone marrow mesen-chymal stromal cells for regeneration of injured equine ligaments and tendons: A clinical report. Res Vet Sci 2013;95:272–277.

8. Fraser JK, Wulur I, Alfonso Z, Hedrick MH. Fat tissue: An underappreciated source of stem cells for biotechnology. Trends Biotechnol 2006;24:150–154.

9. Riordan NH, Ichim TE, Min WP, et al. Non-expanded adipose stromal vascular fraction cell therapy for multiple sclerosis. J Transl Med 2009;7:1–9.

10. Gentile P, Orlandi A, Scioli MG, Di Pasquali C, Bocchini I, Cervelli V. Concise review: Adipose-derived stromal vascular fraction cells and platelet-rich plasma: Basic and clinical impli-cations for tissue engineering therapies in regenerative surgery. Stem Cells Transl Med 2012;1:230–236.

11. Strem BM, Hicok KC, Zhu M, et al. Multipotential differentiation of adipose tissue-derived stem cells. Keio J Med 2005;54:132–141.

12. Frisbie DD, Smith RKW. Clinical update on the use of mesen-chymal stem cells in equine orthopaedics. Equine Vet J 2010;42: 86–89.

13. Strioga M, Viswanathan S, Darinskas A, Slaby O, Michalek J. Same or not the same? Comparison of adipose tissue-derived versus bone marrow-derived mesenchymal stem and stromal cells. Stem Cells Dev 2010;21:2724–2752.

14. Kol A, Walker NJ, Galuppo LD, et al. Autologous point-of-care cellular therapies variably induce equine mesenchymal stem cell migration, proliferation and cytokine expression. Equine Vet J 2013;45:193–198.

15. Adams MK, Goodrich LR, Rao S, et al. Equine bone marrow-derived mesenchymal stromal cells (BMDMSCs) from the ilium and sternum: Are there differences? Equine Vet J 2012;45:372–375.

16. Meirelles LS, Sand TT, Harman RJ, Lennon RJ, Caplan A. MSC frequency correlates with blood vessel density in equine adipose tissue. Tissue Eng Part A 2009;15:221–229.

17. Burk J, Ribitsch I, Gittel C, et al. Growth and differentiation characteristics of equine mesenchymal stromal cells derived from different sources. Vet J 2013;195:98–106.

18. Bruno I, Martinez R, Sanchez A, Friddle C, McClure SR. Char-acterization of nucleated cells from equine adipose tissue and bone marrow aspirate processed for point-of-care use. J Equine Vet Sci 2014;34:1118–1127.

19. Russo V, Yu C, Belliveau P, Hamilton A, Flynn LE. Comparison of human adipose-derived stem cells isolated from subcutaneous, omental, and intrathoracic adipose tissue depots for regenerative applications. Stem Cells Transl Med 2014;3:206–217.

20. Braun J, Hack A, Weis-Klemm M, et al. Evaluation of the osteo-genic and chondrogenic differentiation capacities of equine adipose tissue-derived mesenchymal stem cells. Am J Vet Res 2010;71:1228–1236.

21. Banyard D, Salibian A, Widgerow AD, Evans GRD. Implications for human adipose-derived stem cells in plastic surgery. J Cell Mol Med 2015;19:21–30.

Appendix A. The primer sequences utilized in this study.

Gene Forward 59 Reverse 59

CD44 CGCAGGGGTATTCCATGTG CCCTTCTATGAACCCGTACCTGCD73 GGGATTGTTGGATACACTTCAAAAG GCTGCAACGCAGTGATTTCACD90 CATCGCTCTCCTGCTGACA GGTGTGGCGGTGGTATTCTCCD105 CGACGCCAATCACAACATAC GTAGCCAGGGATGTTCCTCTCCD146 GTGTGGAAATCAGGTGTTTGGC CATCCGTTGTCTCTTCCTCCATCCD166 GAACACGATGAGGCAGACGAG CAACGACGATTCCCACGATGCD34 GCACAAGTCTGACCTGAGCA GGAATAGTGCTGGTGGCTCTCCD45 CTGATGTGATGATTGCTGCTC GGCTTCCTGGTCTCCACTCCD31 TCTAGAACGGAAGGCTCCCT TGGGAGCAGGGCAGGTTCAPPARG2 GACATCAAGCCCTTCACTACTG CTGAATAATAAGGTGGGGATGCOSTEO CAGTGAGGTGGTGAAGAGATTC CTCACACACCTCCCTCCTGCOL2A1 AGCAGGAATTTGGTGTGGAC TCTGCCCAGTTCAGGTCTCTb-actin CCAGCACGATGAAGATCAAG GTGGACAATGAGGCCAGAATGAPDH CCAGAACATCATCCCTGCTT CGTATTTGGCAGCTTTCTCC



22. Oedayrajsingh-Varma MJ, van Ham SM, Knippenberg M, et al. Adipose tissue-derived mesenchymal stem cell yield and growth characteristics are affected by the tissue-harvesting procedure. Cytotherapy 2006;8:166–177.

23. Caplan AL. The mesengenic process. Clin Plast Surg 1994;21: 429–435.

24. Bourin P, Bunnell BA, Casteilla L, et al. Stromal cells from the adipose tissue-derived stromal vascular fraction and cul-ture expanded adipose tissue-derived stromal/stem cells: A joint statement of the International Federation for Adipose Therapeutics (IFATS) and Science and the International Society for Cellular Therapy (ISCT). Cytotherapy 2013;15:641–648.

25. Dominic M, Le Blanc K, Mueller I, et al. Minimal criteria for defining multipotent mesenchymal stem cells. The International Society for Cellular Therapy position statement. Cytotherapy 2006;8:315–317.

26. Kolf CM, Cho E, Tuan RS. Mesenchymal stromal cells. Biology of adult mesenchymal stem cells: Regulation of niche, self-renewal and differentiation. Arthritis Res Ther 2007;9:204.

27. Stewart MC, Steward AA. Mesenchymal stem cells: Character-istics, sources, and mechanisms of action. Vet Clin North Am Equine Pract 2011;27:243–261.

28. Xu H, Barns GT, Yang Q, et al. Chronic inflammation in fat plays a crucial role in development of obesity-related insulin resistance. J Clin Invest 2003;112:1821–1830.

29. Ratajczak MZ, Zuba-Surma EK, Wysoczynski M, Ratajczak J, Kucia M. Very small embryonic-like stem cells: Characterization, developmental origin, and biological significance. Exp Hematol 2008;36:742–751.

30. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW. Obesity is associated with macrophage accumula-tion in adipose tissue. J Clin Invest 2003;112:1796–1808.

31. Radcliffe CH, Flaminio M, Fortier LA. Temporal analysis of equine bone marrow aspirate during establishment of puta-tive mesenchymal progenitor cell populations. Stem Cells Dev 2010;19:269–282.

32. Mohanty N, Gulati BR, Kumar R, et al. Immunophenotypic characterization and tenogenic differentiation of mesenchymal stromal cells isolated from equine umbilical cord blood. In Vitro Cell Dev Biol Anim 2014;50:538–548.

33. Liu Z, Zhuge Y, Velazquez OC. Trafficking and differentiation of mesenchymal stem cells. J Cell Biochem 2009;106:984–991.

34. Baer PC. Adipose-derived mesenchymal stromal/stem cells: An update on their phenotype in vivo and in vitro. World J Stem Cells 2015;6:256–265.

35. Ranera B, Ordovás L, Lyahyai J, et al. Comparative study of equine bone marrow and adipose tissue-derived mesenchymal stromal cells. Equine Vet J 2012;44:33–42.

36. Lange-Consiglio A, Corradetti B, Rutigliano L, Cremonesi F, Bizzaro D. In vitro studies of horse umbilical cord matrix-derived cells: From characterization to labeling for magnetic resonance imaging. Open Tissue Eng Regen Med J 2011;4: 120–133.

37. Raabe O, Shell K, Würtz A, Reich CM, Wenisch S, Arnhold S. Further insights into the characterization of equine adipose tissue-derived mesenchymal stem cells. Vet Res Commun 2011; 35:355–365.

38. Stolzing J, Jones E, McGonagle D, Scutt A. Age-related changes in human bone marrow-derived mesenchymal stem cells: Consequences for cell therapies. Mech Ageing Dev 2008;129: 163–173.

39. Delling U, Lindner K, Ribitsch I, Jülke H, Brehm W. Comparison of bone marrow aspiration at the sternum and the tuber coxae in middle-aged horses. Can J Vet Res 2012;76:52–56.


Article


I n t r o d u c t i o nThe development of hepatic fibrosis is an important event in the

progression of liver disease and has been shown to be a negative prognostic factor in humans with chronic hepatitis (1). Currently, the only way to diagnose hepatic fibrosis in dogs is by histological assessment of a liver biopsy specimen. Liver biopsy is invasive, there

is a risk of hemorrhage, and as only a small portion of the organ is evaluated (2); this technique is susceptible to sampling variation (3,4). Because of these disadvantages, serum markers of hepatic fibrosis have been developed for use in humans (5).

In human patients with hepatic fibrosis, an increased rate of extracellular matrix turnover results in matrix components being released into the bloodstream (5). Hyaluronic acid (HA) is a highly

Evaluation of hyaluronic acid, procollagen type III N-terminal peptide, and tissue inhibitor of matrix metalloproteinase-1 as serum markers

of canine hepatic fibrosisJonathan A. Lidbury, Aline Rodrigues Hoffmann, Joanna K. Fry, Jan S. Suchodolski, Jörg M. Steiner

A b s t r a c tThe only way to diagnose hepatic fibrosis in dogs is by histological assessment of a liver biopsy specimen. As this technique is invasive and susceptible to sampling variation, serum biomarkers are used to detect hepatic fibrosis in humans. The objective of this study was to assess the utility of hyaluronic acid (HA), procollagen type III N-terminal peptide (PIIINP), and tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) as serum markers of canine hepatic fibrosis. Serum samples were collected from 47 dogs with histologically confirmed hepatobiliary disease and 24 healthy dogs in order to measure concentrations of HA, PIIINP, and TIMP-1. Hepatic fibrosis was staged using a 5-point scoring scheme. There was no correlation between serum concentrations of HA or PIIINP and the severity of hepatic fibrosis. There was a negative correlation between serum concentration of TIMP-1 and the severity of hepatic fibrosis (rs = 20.33; P = 0.036). It was not possible to use serum concentrations of HA, PIIINP, or TIMP-1 to discriminate between dogs with absent-to-moderate hepatic fibrosis and those with marked-to-very-marked fibrosis. The results of this study do not support the utility of measuring serum concentrations of HA, PIIINP, or TIMP-1 for diagnosing canine hepatic fibrosis. Further studies are needed to support this finding.

R é s u m éLe seul moyen de diagnostiquer la fibrose hépatique chez les chiens est par évaluation histologique d’un échantillon de biopsie hépatique. Étant donné que cette technique est invasive et sujette à variation dans l’échantillonnage, chez l’humain des marqueurs sériques sont utilisés pour détecter la fibrose hépatique. L’objectif de la présente étude était d’évaluer l’utilité de l’acide hyaluronique (AH), du peptide N-terminal du pro-collagène de type III (PNPIII), et de l’inhibiteur tissulaire de la métalloprotéinase-1 matricielle (ITMP-1) comme marqueurs sériques de la fibrose hépatique canine. Des échantillons de sérums ont été récoltés de 47 chiens avec une pathologie hépatobiliaire confirmée histologiquement et 24 chiens en santé afin de mesurer les concentrations d’AH, PNPIII, et ITMP-1. La fibrose hépatique a été catégorisée en utilisant un schéma de pointage en cinq points. Il n’y avait pas de corrélation entre les concentrations sériques d’AH ou de PNPIII et la sévérité de fibrose hépatique. Il y avait une corrélation négative entre la concentration sérique d’ITMP-1 et la sévérité de la fibrose hépatique (rs = 20,33; P = 0,036). Il n’était pas possible d’utiliser les concentrations sériques d’AH, de PNPIII, ou d’ITMP-1 afin de différencier les chiens avec une fibrose hépatique absente à modérée de ceux avec une fibrose marquée à très marquée. Les résultats de la présente étude ne permettent pas de justifier l’utilité de mesurer les concentrations sériques d’AH, de PNPIII, ou d’ITMP-1 pour le diagnostic de la fibrose hépatique canine. Des études supplémentaires sont requises pour soutenir cette trouvaille.


Department of Small Animal Clinical Sciences (Lidbury, Suchodolski, Steiner), Department of Veterinary Pathobiology (Hoffmann), College of Veterinary Medicine & Biomedical Sciences, Texas A&M University, 4474 TAMU, College Station, Texas 77843, USA; Gulf Coast Veterinary Specialists, 1111 West Loop South, Houston, Texas 77027, USA (Fry).

Address all correspondence to Dr. Jonathan Lidbury; telephone: (979) 845-2351; fax: (979) 862 2864, e-mail: [email protected]

Dr. Fry’s current address is Nashville Veterinary Specialists and Animal Emergency, 2971 Sidco Drive, Nashville, Tennessee 73204, USA.

Preliminary results of this study were presented as abstracts at the 2013 European College of Veterinary Internal Medicine Conference in Liverpool, September 4 to 6, 2013 and at the 2014 American College of Veterinary Internal Medicine Forum in Nashville, June 5 to 7, 2014.

Some of the dogs enrolled in this study were part of a separate study of novel inflammatory markers of hepatic disease, the manuscript of which is currently under review with another journal. Craig SM, Fry JK, Rodrigues Hoffmann A, Manino P, Heilmann RM, Suchodolski JS, Steiner JM, Hottinger HA, Hunter SL, Lidbury JA. Serum C-reactive protein and S100A12 concentrations in dogs with hepatic disease. J Small Anim Pract 2016 Jun 7. doi: 10.1111/jsap.12504. [Epub ahead of print] PubMed PMID: 27271454.\ This study forms part of Dr. Lidbury’s PhD thesis at the College of Veterinary Medicine & Biomedical Sciences, Texas A&M University, 4474 TAMU, College Station, Texas.

Received January 12, 2016. Accepted June 28, 2016.



evolutionarily conserved glycosaminoglycan component of the extracellular matrix (6) that is used as a serum biomarker of hepatic fibrosis (5). In a study of humans with chronic hepatitis-C, the sensi-tivity and specificity of HA for distinguishing between patients with extensive fibrosis and those with milder fibrosis were 86% and 70%, respectively (7). It has previously been shown that HA increased in dogs with hepatic disease, including cirrhosis (8,9) and congenital portosystemic shunts (CPSS) (10).

In a study of human patients with chronic hepatitis-C, measure-ment of serum concentrations of procollagen type III N-terminal peptide (PIIINP) had a sensitivity of 92% and a specificity of 76% for differentiating between patients with extensive fibrosis and those with milder fibrosis (7). Serum concentrations of PIIINP were increased in growing dogs, but not in dogs with chronic broncho-pulmonary disease (11). Serum concentrations of PIIINP have not previously been assessed in dogs with hepatobiliary disease.

Tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) is a protein that inhibits the action of matrix metalloproteinase-1, thus slowing the degradation of extracellular matrix during fibrosis (12). In a study of human patients with chronic hepatitis-C, measure-ment of serum concentration of TIMP-1 had a sensitivity of 75% and a specificity of 70% for differentiating between patients with extensive fibrosis and those with milder fibrosis (7). To the authors’ knowledge, serum concentrations of TIMP-1 have not previously been assessed in dogs with hepatobiliary disease.

We hypothesized that serum concentrations of HA, PIIINP, and TIMP-1 would be positively correlated with the histological stage of hepatic fibrosis and would allow the discrimination between dogs with absent-to-moderate hepatic fibrosis and those with marked-to-very-marked fibrosis. The main objectives of this initial exploratory study were to determine if there is a correlation between serum con-centrations of these analytes and the histological severity of hepatic fibrosis and to determine if these markers allow the discrimination between dogs with absent-to-moderate hepatic fibrosis and those with marked-to-very-marked fibrosis. The secondary objectives of the study were to compare serum concentrations of HA, PIIINP, and TIMP-1 among dogs with different types of hepatobiliary disease and healthy dogs, as well as to partially validate a nonspecies-specific serum HA assay for use with canine serum.


AnimalsDogs over 1 y of age (over 2 y for large and giant breeds) with

histologically confirmed hepatobiliary disease diagnosed at Gulf Coast Veterinary Specialists or Texas A&M University Veterinary Medical Teaching Hospital from March 1, 2011 to February 28, 2013 were enrolled in this prospective observational study. This age cutoff was used to avoid enrolling growing dogs. Diagnosis of hepatobili-ary disease was based on a combination of clinical signs, laboratory testing, diagnostic imaging findings, histological evaluation of a liver biopsy specimen, and in some cases, findings after surgical exploration of the abdominal cavity.

These dogs were divided into 4 groups: i) chronic hepatitis; ii) hepatic neoplasia, which could be primary or secondary;

iii) congenital portosystemic shunt (CPSS); and iv) other hepatobi-liary diseases, including vacuolar hepatopathy, nodular regenera-tion, and gallbladder mucocele. When possible, an extra liver biopsy specimen was collected from the dogs with hepatobiliary disease for evaluating the severity of fibrosis.

Healthy staff-owned dogs over 1 y of age (over 2 y of age for large and giant breeds) were enrolled at the Texas A&M University Veterinary Medical Teaching Hospital. The health of these dogs was assessed by an owner questionnaire, physical examination, complete blood (cell) count (CBC), serum biochemistry profile, and measure-ment of serum concentration of pancreas-specific lipase (Spec cPL Test; IDEXX Laboratories, Westbrook, Maine, USA). Dogs with clinically important abnormalities were excluded from the study.

The study was approved by the Texas A&M University Institutional Animal Use Committee. Informed owner consent was given before the dogs were enrolled in the study.

Serum samplesAt the time of liver biopsy or immediately before euthanasia and

necropsy, 3 to 5 mL of blood were collected from the dogs by jugular venipuncture and placed into sterile anticoagulant-free tubes. Once a firm blood clot had formed, the blood was centrifuged at 1300 3 g for 15 min to separate the serum from the red blood cells. The serum was stored at −80°C until analysis.

AssaysThe assays were conducted in batches containing samples from

healthy and diseased dogs. Samples were run in duplicate using materials, including standards and quality controls, provided by the respective manufacturers according to their instructions. Serum concentrations of HA were measured with a commercially avail-able enzyme-linked immunosorbent assay (ELISA) marketed for use in humans and other animals (Hyaluronan ELISA Kit; Echelon Biosciences, Salt Lake City, Utah, USA). This assay has previously been used for measuring HA in canine serum (13–16). Assay preci-sion was assessed by calculating the intra-assay coefficients of varia-tion (%CV) for 3 samples (low, medium, and high concentration) run 6 times on the same ELISA plate. Repeatability was assessed by cal-culating the inter-assay %CV for 3 samples (low, medium, and high concentration) run 7 times on different days. We assessed analytical accuracy by calculating observed-to-expected ratios (%) when equal volumes of 2 of 4 canine serum samples were mixed in every possible combination. Assay linearity was assessed by calculating observed-to-expected ratios (%) when 5 samples were serially diluted, 1:2, 1:4, 1:8, and 1:16, with sample buffer. Serum concentrations of PIIINP were measured with a human radioimmunoassay (UniQ PIIINP RIA; Orion Diagnostica, Espoo, Finland) that has previously been validated for use with canine serum (11). Serum concentrations of TIMP-1 were measured using a commercially available canine ELISA (TIMP-1 ELISA; USCN Life Science, China) that the manufacturer has validated for use with canine serum.

Histological assessmentLiver biopsies were fixed in neutral buffered formalin, processed

for routine histology, and embedded in paraffin. Serial sections of liver were stained with hematoxylin and eosin and picrosirius red



(Picrosirius Red Stain Kit; Polysciences, Warrington, Pennsylvania, USA). Hepatic fibrosis was staged by a board-certified veterinary anatomic pathologist (ARH) using a 5-point scoring scheme, i.e., 0 — absent, 1 — mild, 2 — moderate, 3 — marked, and 4 — very marked. This scheme is an adaptation of the Ishak scheme and was devised by Drs. van den Ingh, Grinwis, and Rothuizen at the University of Utrecht and has previously been used to assess hepatic fibrosis in dogs (17). Information about the clinical history or serum concentrations of extracellular matrix components was not provided to the pathologist during the scoring process.

Statistical methodsThe distribution of continuous data was assessed using the

Kolmogorov-Smirnov test and by visual inspection of frequency histograms. Non-parametric data were expressed as the median (minimum−maximum). Serum concentrations of HA, PIIINP, and HA were compared among the groups of dogs using the Kruskal-Wallis test, followed by post-testing with Dunn’s test as needed. Comparisons of continuous variables between 2 groups of dogs were made using Mann-Whitney U-tests. The relationships between serum HA, PIIINP, and TIMP-1 concentrations and the hepatic fibrosis stage were assessed using Spearman’s rank correlation (rs). A statistical software package was used for these calculations (GraphPad Prism 5; GraphPad Software, LaJolla, California, USA). Statistical significance was set as P , 0.05.

Re s u l t sForty-seven dogs with hepatobiliary disease were enrolled in the

study. The median age of the dogs was 10 y (range: 1 to 17 y). Twenty were neutered males (43%), 3 were intact males (6%), 22 were spayed females (47%), and 2 were intact females (4%). The following breeds were represented: 3 Labrador retrievers (6%), 3 miniature schnauzers (6%), 3 Yorkshire terriers (7%), 2 Chihuahuas (4%), and 16 mixed-breed dogs (32%). Twenty dogs (43%) had chronic hepatitis, 17 (36%) had hepatobiliary neoplasia (13 with hepatocellular carcinoma/adenoma, 2 with hemangiosarcoma, 1 with cholangiocarcinoma, and 1 with lymphoma), 4 dogs (9%) had CPSS, and 6 (13%) had other hepatobiliary disease (2 dogs with gall bladder mucoceles, 2 with nodular hyperplasia, and 1 with portal vein hypoplasia and portal hypertension, and 1 with acute hepatitis). Twenty-four healthy dogs were enrolled in the study. The median age of these dogs was 3 y (range: 1 to 13 y). Eight were neutered males (33%), 11 were spayed females (46%), and 5 were intact females (21%). The following breeds were represented: 3 miniature schnauzers (13%), 2 German shepherds (8%), 2 hounds (8%), 2 Australian shepherds (8%), and 3 mixed-breed dogs (13%).

All the hepatic samples were deemed to be adequate for evalua-tion. Hepatic fibrosis stage scores were assigned to the dogs as fol-lows: 6 dogs (13%) had no fibrosis, 10 dogs (21%) had mild, 14 dogs (30%) had moderate, 8 dogs (17%) had marked, and 9 dogs (19%) had very marked fibrosis.

The intra-assay %CV for the HA ELISA was 6.2%, 1.9%, and 12.2% for low, medium, and high concentration samples, respectively. The inter-assay variability %CV for the assay was 15.1%, 13.4%, and 15.3% for low, medium, and high concentration samples, respec-

tively. The mean (6 standard deviation; minimum−maximum) observed-to-expected ratio for spiking recovery experiments was 97.2% (6 7.2%; range: 89.5% to 110.9%). The mean observed-to-expected ratio for the dilutional parallelism experiments was 96.0% (6 16.9%; range: 75.4% to 129.3%).

Serum concentrations of HA were measured in 45 dogs with hepa-tobiliary disease (96%) and in 24 healthy dogs (100%). There was no correlation between serum concentration of HA and the sever-ity of hepatic fibrosis (rs = 0.24; P = 0.114; Figure 1). There was no significant difference in serum concentrations of HA between dogs with absent-to-moderate hepatic fibrosis and those with marked-to-very-marked fibrosis (P = 0.598; Table I). Serum concentrations of HA were higher in healthy dogs (median: 201 ng/mL; range: 84 to 1464 mg/mL) than in dogs with neoplasia (median: 126 ng/mL; range: 82 to 1532 ng/mL; P , 0.05; Table I). There were no other significant differences in serum concentrations of HA among the groups of dogs (Table II).

Serum concentrations of PIIINP were measured in 39 dogs with hepatobiliary disease (83%) and in 24 healthy dogs (100%). There was no correlation between serum concentration of PIIINP and the severity of hepatic fibrosis (rs = 0.12; P = 0.472; Figure 2). There was no significant difference in serum concentrations of PIIINP between dogs with absent-to-moderate hepatic fibrosis and those with marked-to-very-marked fibrosis (Table II; P = 0.707). There were no significant differences in serum concentrations of PIIINP among the groups of dogs (P = 0.440; Table I).

Serum concentrations of TIMP-1 were measured in 41 dogs with hepatobiliary disease (87%) and in 24 healthy dogs (100%). There was a negative correlation between serum concentration of TIMP-1 and the severity of hepatic fibrosis (rs = 20.33; P = 0.036; Figure 3). There was no significant difference in serum concentrations of TIMP-1 between dogs with absent-to-moderate hepatic fibrosis and those with marked-to-very-marked fibrosis (P = 0.124; Table I). There were

Figure 1. There was no correlation between serum concentrations of hyaluronic acid (HA) and the stages of hepatic fibrosis in dogs (rs = 0.24; P = 0.114).



no significant differences in serum concentrations of TIMP-1 among the groups of dogs (P = 0.139; Table II; Figure 4). Post-hoc analysis revealed that there was no significant difference in serum concen-tration of TIMP-1 between dogs with hepatic neoplasia (median: 45 ng/mL; range: 6 to 100 ng/mL) and dogs with non-neoplastic hepatobiliary disease (median: 22 ng/mL; range: 5 to 156 ng/mL; P , 0.1 but . 0.05). Post-hoc analysis showed that dogs with primary hepatobiliary neoplasia (hepatocellular adenomas, hepatocellular carcinomas, and cholangiocarcinomas) had higher serum concentra-tions of TIMP-1 than those with tumors and those with neoplasia secondarily affecting the liver (lymphoma and hemangiosarcoma), with median concentrations of 47 ng/mL (range: 25 to 100 ng/mL) and 6 ng/mL (range: 6 to 12 mg/mL), respectively (P = 0.0115).

D i s c u s s i o nThe HA ELISA that we used in this study had acceptable precision,

accuracy, and linearity for measuring canine serum concentrations of HA. The inter-assay %CVs of 15.1%, 13.4%, and 15.3% were slightly higher than desirable, which suggests suboptimal repeatability. We do not think this is likely to have affected our conclusions, however, as there was no trend for serum concentrations of HA to be related to the severity of hepatic fibrosis. The relevance of the finding that dogs with hepatic neoplasia had lower serum concentrations of HA than

healthy dogs is unknown and neither is the reason that this occurred. There were only small differences in the median concentration of HA between the groups, however, and it may therefore not be clinically important. We are not aware of a similar relationship in humans with hepatic neoplasia. There was no relationship between the severity of hepatic fibrosis and serum concentrations of HA in the dogs enrolled in this study. This is in contrast to studies in humans in whom HA has been shown to be a useful marker of hepatic fibrosis (5,7) and a previous study in dogs in which those with hepatic cirrhosis had higher serum concentrations of HA than those with non-cirrhotic hepatic disease, those with non-hepatic disease, or healthy dogs (8).

Another study also showed that serum concentrations of HA were higher in dogs with advanced hepatic fibrosis than in those with mild fibrosis (9). The aforementioned studies in dogs used a different assay than the one we used. Previous studies using the ELISA that we used in our study, however, found that Chinese Shar pei dogs with cutaneous mucinosis have higher serum concentrations of HA than healthy dogs or Chinese Shar peis without cutaneous mucino-sis (13–16), which suggests that this assay is capable of detecting elevated serum concentrations of HA in dogs. The median serum concentrations of HA in 5 healthy dogs reported in 1 of these stud-ies was 244 ng/mL (range: 166 to 302 ng/mL), which is similar to that of the 23 healthy dogs from our study with 201 ng/mL (range: 84 to 1464 ng/mL) (15). Another possible reason for the discrepancy

Table I. Serum concentrations of extracellular matrix components in dogs at different stages of hepatic fibrosis

Absent-to-moderate Marked-to-very- Analyte fibrosis marked fibrosis P-valueHA concentration 1310 ng/mL 156 ng/mL 0.598median (min−max) (50 to 1532) (50 to 3360)

PIIINP concentration 8.2 mg/L 9.8 mg/L 0.707median (min−max) (3.3 to 50.0) (3.1 to 38.6)

TIMP-1 concentration 30 ng/mL 19 ng/mL 0.124median (min−max) (5 to 100) (6 to 59) HA — Hyaluronic acid; PIIINP — Procollagen type III N-terminal peptide; TIMP-1 — Tissue inhibitor of matrix metalloproteinase-1.

Table II. Serum concentrations of extracellular matrix components in healthy dogs and dogs with various types of hepatobiliary disease

Congenital Other Chronic Hepatobiliary portosystemic hepatobiliary P-value Analyte Healthy hepatitis neoplasia shunt disease (Kruskall-Wallis)HA concentration 201 ng/mL* 154 ng/mL 126 ng/mL* 171 ng/mL 126 ng/mL 0.038median (min−max) (84 to 1464) (50 to 3360) (82 to 1532) (70 to 576) (106 to 304)

PIIINP concentration 10.0 mg/L 8.2 mg/L 8.1 mg/L 12.5 mg/L 10.9 mg/L 0.440median (min−max) (3.8 to 48.2) (3.1 to 38.6) (3.3 to 35.3) (6.0 to 16.3) (3.3 to 22.7)

TIMP-1 concentration 20 ng/mL 21 ng/mL 45 ng/mL 26 ng/mL 23 ng/mL 0.139median (min−max) (5 to 100) (6 to 156) (6 to 100) (13 to 34) (5 to 63) * P , 0.05 for post-testing between groups.HA — Hyaluronic acid; PIIINP — Procollagen type III N-terminal peptide; TIMP-1 — Tissue inhibitor of matrix metalloproteinase-1.



in results between our study and these previous studies on dogs with hepatic disease (8,9) is that our study contained relatively few dogs with marked or very marked hepatic fibrosis and potentially serum concentrations of HA only increase in dogs with very marked fibrosis (n = 9).

There was no difference in serum concentrations of PIIINP among healthy dogs, those with chronic hepatitis, CPSS, hepatobiliary neo-plasia, or other hepatobiliary disease. Additionally, no relationship was observed between the severity of hepatic fibrosis and serum

concentrations of PIIINP. In humans, serum concentrations of PIIINP have been shown to be useful in distinguishing between patients with no or mild fibrosis and those with moderate or severe fibrosis (5,7). It is not known why serum concentrations of PIIINP were not increased in these dogs with hepatic fibrosis. One possible explana-tion is that this protein is not leaked into the bloodstream of dogs with hepatic fibrosis. Type III collagen was increased in the livers of dogs with chronic hepatitis (18), which suggests that this form of collagen is important in canine hepatic fibrosis. It is interesting that a previous study did not find that serum concentrations of PIIINP were increased in dogs with chronic bronchopulmonary disease, but that concentrations in bronchoalveolar lavage fluid were increased (11). In humans, PIIINP has also been shown to have a moderately strong positive correlation with the severity grade of hepatic necroin-flammatory activity (19). We therefore cannot rule out that this was a confounding factor that obscured the presence of a true relationship between PIIINP and the severity of hepatic fibrosis in our study.

Upon visual inspection of the data, the median serum concentra-tion of TIMP-1 in dogs with hepatic neoplasia (median: 45 ng/mL;

Figure 2. There was no correlation between serum concentrations of procollagen type III N-terminal peptide (PIIINP) and the stages of hepatic fibrosis in dogs (rs = 0.12; P = 0.472).

Figure 3. There was a weak negative correlation between serum concen-trations of tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) and the stages of hepatic fibrosis in dogs (rs = 20.33; P = 0.036).

Figure 4. Serum concentrations of tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) in healthy dogs and dogs with hepato-biliary disease. The solid horizontal bars represent the median serum concentration of TIMP-1 for that group. For the group of dogs with hepatobiliary neoplasia, 3 had metastatic disease (marked with stars), 2 had hemangiosarcoma, 1 had lymphoma, and 12 had hepatocellular adenoma/ carcinoma (marked with squares). CPSS — Congenital porto-systemic shunt.



range: 6 to 100 ng/mL) was higher than that of healthy dogs (median: 20; range: 5 to 100 ng/mL) or a combined group of dogs with non-neoplastic hepatobiliary disease formed post-hoc (median: 22; range: 5 to 156 ng/mL), but the differences were not statistically significant. This was possibly due to type-II error, as the sample size in each group was relatively small. Additionally, dogs with primary liver neoplasia (hepatocellular adenoma/carcinoma and cholangio-carcinoma) had significantly higher serum concentrations of TIMP-1 than those with lymphoma and hemangiosarcoma.

Humans with a variety of tumors, including hepatic metastases (20) and pulmonary neoplasms (21), have been shown to have increased serum concentrations of TIMP-1. In 1 study of humans with hepatic metastases, higher serum concentrations of TIMP-1 were shown to be a poor prognostic factor (20). Dogs with spontaneously occurring mammary tumors have been shown to have a relatively low tissue TIMP-1 activity compared to rats with induced mammary tumors (22), but to our knowledge, serum concentrations of TIMP-1 have not previously been reported in dogs with hepatobiliary neoplasia. It is important to emphasize, however, that it was not hypothesized a priori that serum concentrations of TIMP-1 would be higher in dogs with primary hepatic neoplasia than in metastatic tumors. This means that we cannot draw any definitive conclusions from this finding. Further studies in a larger group of dogs with hepatic neoplasia are needed to confirm or refute our findings and to assess serum concen-trations of TIMP-1 in dogs with other types of neoplasia.

There was a weak negative correlation between serum con-centration of TIMP-1 and the severity of fibrosis. As there was no significant difference in the fibrosis stages assigned to dogs with hepatobiliary neoplasia and those with other hepatobiliary diseases, it is unlikely that the non-significant trend for increased serum concentrations of TIMP-1 observed in dogs with neoplasia acted as a confounding factor, which explains this unexpected negative correlation. There was no difference in serum concentra-tions of TIMP-1 between dogs with absent-to-moderate fibrosis and those with marked-to-very-marked fibrosis. These findings contradict those in studies of humans in which TIMP-1 is a use-ful marker of hepatic fibrosis and is positively correlated with its severity (5,7). The reason for this difference between the 2 species is not known.

It is important to discuss several limitations of this work. Firstly, as there were relatively few dogs (9 out of 48; 19%) that were assigned a score of very marked hepatic fibrosis, it is possible that this was why we failed to find a relationship between any of the extracellular matrix components and the severity of hepatic fibrosis because they are only increased in dogs with very marked fibrosis. Even if this was the case, however, the utility of these markers would be limited if they can only distinguish between dogs with mild fibrosis and those with very marked fibrosis. The sample sizes in some of the groups were small and this may have led to type-II error when evaluating differences in the analytes between these groups. Additionally, we cannot completely exclude the possibility that some of the dogs that were enrolled in this study had subclinical disease that was causing fibrosis of another organ, therefore increasing their serum concentration of extracellular matrix components and act-ing as a confounding factor. The group of dogs with hepatobiliary disease in this study, however, would be similar to the population

of dogs in which these markers would be used if they were shown to be valuable. It is therefore important to recognize their lack of apparent diagnostic utility in this population.

In conclusion, serum concentrations of HA, PIIINP, and TIMP-1 were not able to discriminate between dogs with and without hepatic fibrosis. Therefore, the results of this study do not support the utility of measuring serum concentrations of HA, PIIINP, or TIMP-1 for diagnosing canine hepatic fibrosis. Further studies are needed to confirm this finding.

Conflict of interestNone of the authors has any financial or personal relationships

that could inappropriately influence or bias the content of this paper.

Re f e r e n c e s 1. Everhart JE, Wright EC, Goodman ZD, et al. Prognostic value

of Ishak fibrosis stage: Findings from the hepatitis C antiviral long-term treatment against cirrhosis trial. Hepatology 2010; 51:585–594.

2. Petre SL, McClaran JK, Bergman PJ, Monette S. Safety and effi-cacy of laparoscopic hepatic biopsy in dogs: 80 cases (2004–2009). J Am Vet Med Assoc 2012;240:181–185.

3. Bedossa P, Dargere D, Paradis V. Sampling variability of liver fibrosis in chronic hepatitis C. Hepatology 2003;38:1449–1457.

4. Kemp SD, Zimmerman KL, Panciera DL, Monroe WE, Leib MS. Histopathologic variation between liver lobes in dogs. J Vet Intern Med 2015;29:58–62.

5. Baranova A, Lal P, Birerdinc A, Younossi ZM. Non-invasive markers for hepatic fibrosis. BMC Gastroenterol 2011;11:91.

6. Chen WY, Abatangelo G. Functions of hyaluronan in wound repair. Wound Repair Regen 1999;7:79–89.

7. Leroy V, Monier F, Bottari S, et al. Circulating matrix metal-loproteinases 1, 2, 9 and their inhibitors TIMP-1 and TIMP-2 as serum markers of liver fibrosis in patients with chronic hepatitis C: Comparison with PIIINP and hyaluronic acid. Am J Gastroenterol 2004;99:271–279.

8. Kanemoto H, Ohno K, Sakai M, et al. Blood hyaluronic acid as a marker for canine cirrhosis. J Vet Med Sci 2009;71:1251–1254.

9. Glinska-Suchocka K, Orlowska A, Spu·zak J, Jankowski M, Kubiak K. Suitability of using serum hialuronic acid concentra-tions in the diagnosis of canine liver fibrosis. Pol J Vet Sci 2015; 18:873–878.

10. Seki M, Asano K, Sakai M, et al. Serum hyaluronic acid in dogs with congenital portosystemic shunts. J Small Anim Pract 2010; 51:260–263.

11. Schuller S, Valentin S, Remy B, et al. Analytical, physiologic, and clinical validation of a radioimmunoassay for measurement of procollagen type III amino terminal propeptide in serum and bronchoalveolar lavage fluid obtained from dogs. Am J Vet Res 2006;67:749–755.

12. Arthur MJ, Mann DA, Iredale JP. Tissue inhibitors of metallopro-teinases, hepatic stellate cells and liver fibrosis. J Gastroenterol Hepatol 1998;13 Suppl:S33–38.

13. Grützner N, Heilmann RM, Cranford SM, Holzenburg A, Suchodolski JS, Steiner JM. Inflammatory, immunological,



and intestinal disease biomarkers in Chinese Shar-Pei dogs with marked hypocobalaminemia. J Vet Diagn Invest 2015;27: 31–40.

14. Olsson M, Meadows JR, Truvé K, et al. A novel unstable dupli-cation upstream of HAS2 predisposes to a breed-defining skin phenotype and a periodic fever syndrome in Chinese Shar-Pei dogs. PLoS Genet 2011;7:e1001332.

15. Zanna G, Fondevila D, Bardagi M, Docampo MJ, Bassois A, Ferrer L. Cutaneous mucinosis in shar-pei dogs is due to hyal-uronic acid deposition and is associated with high levels of hyaluronic acid in serum. Vet Dermatol 2008;19:314–318.

16. Docampo MJ, Zanna G, Fondevila D, et al. Increased HAS2-driven hyaluronic acid synthesis in shar-pei dogs with heredi-tary cutaneous hyaluronosis (mucinosis). Vet Dermatol 2011;22: 535–545.

17. Fieten H, Biourge VC, Watson AL, Leegwater PA, van den Ingh TS, Rothuizen J. Nutritional management of inherited copper-associated hepatitis in the Labrador retriever. Vet J 2014;199: 429–433.

18. Spee B, Arends B, van den Ingh TS, et al. Transforming growth factor beta-1 signalling in canine hepatic diseases: New models for human fibrotic liver pathologies. Liver Int 2006;26:716–725.

19. Holoman J, Glasa J, Galbavy S, et al. Serum markers of liver fibrogenesis, and liver histology findings in patients with chronic liver diseases. Bratisl Lek Listy 2002;103:70–75.

20. Bunatova K, Pesta M, Kulda V, et al. Plasma TIMP1 level is a prognostic factor in patients with liver metastases. Anticancer Res 2012;32:4601–4606.

21. Ylisirnio S, Hoyhtya M, Turpeenniemi-Hujanen T. Serum matrix metalloproteinases -2, -9 and tissue inhibitors of metallopro-teinases -1, -2 in lung cancer-TIMP-1 as a prognostic marker. Anticancer Res 2000;20:1311–1316.

22. Kawai K, Uetsuka K, Doi K, Nakayama H. The activity of matrix metalloproteinases (MMPS) and tissue inhibitors of metallopro-teinases (TIMPs) in mammary tumors of dogs and rats. J Vet Med Sci 2006;68:105–111.


Article


I n t r o d u c t i o nAdipose tissue is an abundant, less expensive, and safe bodily tis-

sue to study. It can be isolated in large quantities by minimally inva-sive liposuction. Immature pluripotent cells are important targets for tissue engineering, regenerative medicine, and gene therapy (1). Adipose-derived stem cells (ADSCs) are the richest sources of mul-tipotent cells and can be easily cultivated and expanded. These cells pose less psychological and physical impact to the donor than bone marrow cells (BMCs). The phenotype and expansive capacity of ADSCs are stable, without spontaneous re-differentiation and shift of cellular markers after dozens of passages (2). They have a high capacity to differentiate and self-renew within and across lineage barriers (3). Adipose-derived stem cells can be differentiated into osteogenic, adipogenic, myogenic, chondrogenic, and neurogenic lineages (3–5). Adipose-derived stem cells also have low immuno-genicity, which offers great promise for use in regenerative medicine (6–8). Other research shows ADSCs represent a practical, abundant, and appealing source for cell replacement of donor tissue (9).

Currently, the shortage of available donors and graft rejection hamper widespread transplantation. Stem cell therapy is a cur-rent research topic and holds great promise for disease treatment. Adipose-derived stem cells can grow rapidly in vitro. They are avail-able for harvesting in large numbers using a less invasive operation. In this research, we describe the isolation and culture procedures of pig ADSCs, and demonstrate that ADSCs are able to differentiate into many cells types.


Isolation, culture, and purification of ADSCsAnimal experiments were performed in accordance with the

guidelines established by the Institutional Animal Care and Use Committee at Chinese Academy of Agricultural Sciences. Adipose tissues were harvested from the visceral (omental region) of pigs under aseptic conditions. Briefly, the extracellular matrix was dissoci-ated with 0.1% (m/v) type I collagenase (Sigma) and the harvested

Identification and characterization of pig adipose-derived progenitor cellsShuang Zhang, Chunyu Bai, Dong Zheng, Yuhua Gao, Yanan Fan, Lu Li, Weijun Guan, Yuehui Ma

A b s t r a c tAdipose-derived stem cells (ADSCs) are multipotent, and can be differentiated into many cell types in vitro. In this study, tissues from pigs were chosen to identify and characterize ADSCs. Primary ADSCs were sub-cultured to passage 28. The surface markers of ADSCs: CD29, CD71, CD73, CD90, and CD166 were detected by reverse-transcription polymerase chain reaction assays and the markers CD29, CD44, CD105, and vimentin were detected by immunofluorescence. Growth curves and the capacity of clone-forming were performed to test the proliferation of ADSCs. Karyotype analysis showed that ADSCs cultured in vitro were genetically stable. To assess the differentiation capacity of the ADSCs, cells were induced to differentiate into osteoblasts, adipocytes, epithelial cells, neural cells, and hepatocyte-like cells. The results suggest that ADSCs from pigs showed similar biological characteristics with those separated from other species, and their multi-lineage differentiation shows potential as an application for cellular therapy in an animal model.

R é s u m éLes cellules souches dérivées du tissu adipeux (CSDAs) sont pluripotentes, et peuvent se différencier en plusieurs types de cellules in vitro. Dans la présente étude, des tissus d’origine porcine ont été choisis afin d’identifier et de caractériser les CSDAs. Des CSDAs primaires ont été sous-cultivées jusqu’au passage 28. Les marqueurs de surface des CSDAs suivants : CD29, CD71, CD73, CD90, et CD166 ont été détectés par réaction d’amplification en chaîne par la polymérase utilisant la transcriptase réverse et les marqueurs CD29, CD44, CD105, et vimentine ont été détectés par immunofluorescence. Des courbes de croissance et la capacité à former des clones ont été effectuées afin de tester la prolifération des CSDAs. L’analyse du caryotype a montré que les CSDAs cultivées in vitro sont génétiquement stables. Afin d’évaluer la capacité de différentiation des CSDAs, des cellules ont été induites à se différencier en ostéoblastes, adipocytes, cellules épithéliales, cellules neurales, et des cellules apparentées aux hépatocytes. Les résultats suggèrent que les CSDAs porcines ont montré des caractéristiques biologiques similaires à celles de d’autres espèces, et que leur différentiation en plusieurs lignées montre un potentiel pour une application de thérapie cellulaire dans un modèle animal.


College of Wildlife Resources, Northeast Forestry University, Harbin, P. R. China (Zhang, Zheng, Gao, Fan); Institute of Animal Science, Chinese Academy of Agricultural Sciences, Beijing, P.R. China (Bai, Gao, Fan, Li, Guan, Ma).

Address all correspondence to Yuehui Ma; e-mail: [email protected]

Received October 18, 2015. Accepted December 22, 2015.



pellet was re-suspended in complete medium (10). At 70% to 80% confluence, the cells were passaged with 0.125% trypsin. Generally, after 3 to 4 passages, the cells were homogenous.

Identification of ADSCsSurface marker detection. Cells were incubated in the following

antibodies: i) mouse anti-pig CD29 (1:100; Abcam); ii) rat anti-pig

CD44 (1:100; Abcam); iii) mouse anti-pig CD105 (1:100; Abcam); and iv) mouse anti-pig Vimentin (1:100; Abcam). The cells were then incubated by FITC-conjugated goat anti-mouse immunoglobulin and goat anti-rat immunoglobulin (bioss) and examined using a Nikon TE-2000-E confocal microscope.

RNA isolation and RT-PCR. RNA was extracted from the cells from passages 4, 14, 20, and 26 using Trizolreagent (Invitrogen).

Table I. Primer information for adipose-derived stem cell identification

Product Temperature Gene name Primer sequences Cireles length (bp) (°C)CD29 F 59 TGATTGCTGGTTCTACTTCACA 39 30 302 58 R 59 TTCCCTCATACTTCGGATTGAC 39

CD71 F 59 CATGCAAACTGGTGTCCTCA 39 30 210 59 R 59 TCTGGGCAAGGTTCAATAGG 39

CD73 F 59 AACCCACCTTCCAAAGAGGT 39 30 293 60 R 59 GTGCCATCCAGGTAGACGAT 39

CD90 F 59 GGCATCGCTCTCTTGCTAAC 39 30 250 60 R 59 GGACCTTGATGTCGTACTTGC 39

CD166 F 59 CCATCACGGTTCACTATTTGG 39 30 213 58 R 59 GCAGAGCAGTTTCACAGACG 39

CD31 F 59 GCCCATTTCCTACCAACTTTTA 39 30 318 60 R 59 GGCTTGTTCTCCTTTTCTTTGT 39

COL1al F 59 GGAATGAAGGGACACAGAGG 39 30 207 63 R 59 AGCAGCACCAGTAGCACCAT 39

RUNX2 F 59 GAGAGAGGGAGAGAGAGAGAGGA 39 30 206 57 R 59 TGTTTGTGAGGCGAATGAAG 39

OPN F 59 TCCCAAAGTCAGCCAAGAAT 39 30 262 60 R 59 TTGTTTCCAGAGCCACACAC 39

LPL F 59 TGACAAAGATGCCCAGTGCT 39 30 208 59 R 59 TTTCCCAAAGGACAGAATGC 39

PPAR-g F 59 CCAGGTTTGCTGAATGTGAA 39 30 216 58 R 59 TAGGAGTGGGTGAAGGCTCA 39

ALB F 59 GCCTCTTGTGGATGAGCCTA 39 30 241 62 R 59 GTTCAGGACCAGGGACAGAT 39

AFP F 59 GCCAAAGTGGAGAGGAAAGA 39 30 143 60 R 59 AGCAGCCAAAGAAAGAATCG 39

MAP2 F 59 TCAAGTCAACAGGCATCAGC 30 398 58 R 59 TTTTCCTCTTGCCCCTTTTT

NF F 59 TGGAAGTAGTGGTCGGAAGA 30 460 56 R 59 GGGAACTGGAGGGTAGAT

CK18 F 59 TGCAAATACTGTGGACAATGC 30 375 60 R 59 TCCTCTCGGTTCTTCTGAGC

E-Cadherin F 59 CCCAACACTTCTCCCTTCAC 39 30 281 60 R 59 CACCCTTCTCCTCCGAATAA 39

GAPDH F 59 ACCCAGAAGACTGTGGATGG 39 30 285 60 R 59 CTCAGTGTAGCCCAGGATGC 39



Figure 1. Morphology and characteristics of adipose-derived stem cells (ADSCs). A — Morphology of primary cultured and sub-cultured adipose-derived progenitor cells. (a) Primary cells after culture for 48 h. Many cells started to adhere. (b) ADSCs exhibited a fibroblast-like morphology and expanded easily. (c) ADSCs grew to 90% confluence about 7 days later. (d) The ADSCs of passage 28 displayed cell senescence. Scale bar = 50 mm. B — Characteristics of ADSCs surface antigens at different passages. Immunofluorescence showing CD29-, CD44-, CD105- and vimentin-positive cells. Scale bar = 50 mm. C — RT-PCR analysis showed that ADSCs at different passages express CD29, CD71, CD73, CD90, and CD166, while they didnin-positive cells Lanes 1 CD29, 2 CD71, 3 CD73, 4 CD90, 5 CD166, 6 CD31, 7 GAPDH. GAPDH served as the internal control.

A

B 1 2 3 4 5 6 7P4

P14

P20

P26

C DAPI Fluorescence Merged



We used the TaKaRa RNA PCR Kit (AMV) Ver. 3.0 for reverse-transcription polymerase chain reaction (RT-PCR). The cDNAs were amplified by Emerald Amp Max PCR (TaKaRa). The specific primers are listed in Table I. The PCR products were visualized by 2% agarose gel electrophoresis.

Growth kineticsTo assess growth dynamics, ADSCs at different passages were

seeded in triplicate in 24-well plates (1 3 104 cells/well) (11). The population doubling time (PDT) was calculated as follows:

PDT = (t 2 t0) log2/(logNt 2 logN0),

Where: t0 = start time of culture, t = termination time of culture, N0 = initial number of cells in culture, and Nt = the final number of cells in culture.

Colony-forming cell assayCells from passages 4, 12, and 27 were seeded in 24-well micro-

plates at a density of 1 3 104 cells/well, and numbers of colony-forming units (CFU) were counted to calculate the colony-forming rate, which is formulated as CFU numbers/starting cell number per 24-well 3 100%.

Karyotype analysisThe karyotype of P8 cells was analyzed. Cells were subjected to

hypotonic treatment and fixed, and the chromosome numbers were counted from 100 spreads under an oil immersion objective upon Giemsa staining.

Adipogenic differentiation and confirmationAdipose-derived stem cells from passage 16 were plated onto

60-mm dishes. When the cells confluenced to 70%, the medium was changed to adipocyte-inducing medium. Cells from the control group were fed with DMEM medium. After 10 d, the induced cells were fixed with 4% (m/v) paraformaldehyde and stained with Oil-Red O for 30 min to visualize lipid droplets. The adipogenic specific genes LPL and PPARg were detected by RT-PCR.

Osteogenic differentiation and confirmationThe ADSCs from passage 18 were plated onto 60-mm dishes for

osteogenic differentiation. The cells were incubated with osteogenic-inducing medium when they confluenced to 80%. Cells from the control group were fed with DMEM medium. The cells were fixed with 4% paraformaldehyde (m/v) at 10 and 22 d after induction. These cells were also stained with alizarin to visualize the mineral-ized matrix. The osteogenic-specific genes OPN, COL1, RUNX2 were detected by RT-PCR.

Hepatic differentiation and confirmationTo induce hepatogenic differentiation of ADSCs, passage 13 cells

were incubated on a 60-mm dish. When the cells confluenced to 50%, the medium was changed to hepatic-inducing medium. After 15 d, the cells were fixed with 4% paraformaldehyde (m/v) and stained with periodic acid Schiff to visualize the glycogen deposits. The hepatogenic specific genes AFP and ALB were detected by RT-PCR.

Epithelial cell differentiation and confirmationThe ADSCs from passage 8 were incubated on a 6-well plate. After

the cells confluenced to 50%, the medium was changed to L-DMEM medium containing 10% FBS (v/v), 1% B27, 1% glutamine, 1% streptomycin, 10 ng/mL bFGF, and 15 ng/mL EGF. After 10 d, the

Table II. Medium information for adipose-derived stem cells

Complete medium DMEM (Gibco), 10% (v/v) FBS (Biochrom), 10 ng/mL basic fibroblast growth factor (Peprotech), 2 mM L-glutamine, 1% B-27 (m/v) (Gibco), and 104 IU/mL penicillin/streptomycin

Adipocytes inducing L-DMEM, 10% FBS (v/v), 1% streptomycin, medium 0.5 mmol/l 3-isobutyl-1-methylxanthine

(IBMX), 10 mg/L INS, 1.0 mmol/L dexamethasone, 200 mmol/L indomethacin

Osteogenic inducing L-DMEM, 10% FBS (v/v), 100 nM medium dexamethasone, 250 mM L-ascorbic acid-2-

phosphate and 10 mM b-glycerophosphate

Hepatic inducing which contained L-DMEM medium, 5% FBS medium (v/v), 20 ng/mL FGF-4, 20 ng/mL HGF,

40 nmol/mL dexamethasone and ITS

Neurogenic induction L-DMEM, 2% B27, 1% glutamine, 40 ng/mL medium-1 bFGF, 20 ng/mL EGF

Neurogenic induction L-DMEM, 2% B27, 1% glutamine, 40 ng/mL medium-2 bFGF, 20 ng/mL EGF, 1% N2, 50 mg/mL Vc,

10 ng/mL GDNF

Figure 2. Growth curves of the adipose-derived stem cells. The growth curves of P4, P14, P20, and P26 were all typically sigmoidal, with cell density reflected by the vertical axis. Population doubling time calculated from the growth curve was 32 h.

1 2 3 4 5 6 7 8

7

6

5

4

3

2

1

0

P7

P12

P27

Time (day)

Cel

l den

sity

(3 1

04 )



cells were conformed with immunofluorescent assays of epithelial markers PCNA and CK18. The epitherial specific genes CK18 and E-Carderin were detected by RT-PCR.

Neurogenic differentiation of ADSCs and confirmation

The ADSCs from passage 9 were plated onto a 6-well plate. After the cells confluenced to 70%, the medium was changed to a neuro-genic-induction medium. The induction was divided into 2 steps: i) the cells were incubated with the neurogenic induction medium-1 for 7 d; ii) the cells were then incubated with the neurogenic induc-tion medium-2 for 7 d. The cells were conformed with immuno-fluorescent assays of neural markers MAP2, GFAP, and b-tubulin. The neural-specific genes NF and map2 were detected by RT-PCR.

Re s u l t s

Isolation, culture, and morphological observationPrimary cells isolated from adipose tissue adhered to plates and

began to elongate after 48 h (Figure 1A-a). After about 3 d, the cells exhibited a fibroblast-like morphology (Figure 1A-b). Cells expanded rapidly and confluenced to 90% 10 d later (Figure 1A-c). Cells were cultured up to passage 28 with most cells showing signs of senes-cence, such as vacuolization and karyopyknosis (Figure 1A-d).

Identification of the ADSCsSpecific surface antigen markers of ADSCs were detected by

immunofluorescence and RT-PCR. Results of immunofluorescence

Figure 3. Colony-forming cell assay. Colony-forming units of P7, P12, and P27 adipose-derived progenitor cells were counted, which indicated that colony-forming rates decreased but did not disappear with increasing passage number. (a), (b), and (c) are colony-forming rates of P7, P12, and P27 respectively, (d) is the bar chart of colony-forming rates for different passages of adipose-derived progenitor cells.

40

35

30

25

20

15

10

5

0 P7 P12 P27

Passage of cells

Col

ony-

form

ing

units

P7

P12

P27

Figure 4. Chromosomes at metaphase (left) and karyotype (right) of pig adipose-derived MSCs (10003). The pig chromosome number was 2n = 38. The sex chromosome type was XY.

1 2 3 4 5

6 7

8 9 10 11 12 XY

13 14 15 16 17 18



Figure 5. Multi-potential of adipose-derived stem cells. (a) After incubation in adipogenic for 15 days, the cells stained with oil-red (1,2) and 3 was for control. (b) RT-PCR showed the expression of adipogenic specific genes, such as LPL and PPAR-g. 1 for the inducted group, 2 for the control group. (c) After incubation in osteogenic for 22 days, the cells stained with alizarin red (1,2) and 3 was for control. (d) RT-PCR showed the expression of osteoblast specific genes, such as osteopontin (opn), collage type 1 (COL1) and RUNX2. 1 for the inducted group, 2 for the control group. (e) After

1

2

LPL PPAR-g GAPDH

1

2

GAPDH OPN COL1 RUNX2

1

2

AFB ALP GAPDH



incubation in hepatogenic for 15 days, the cells stained with periodic Acid-Schiff stain (1,2) and 3 was for control. (f) RT-PCR showed the expression of hepatogenic specific genes, such as AFB and ALP. 1 for the inducted group, 2 for the control group. (g) The morphology of the induced ADSCs and undifferentiated ADSCS. (h) RT-PCR showed the expression of epithelial cells specific genes include CJ18 and E-Carderin. 1. induced group; 2. control group. (i) Immonofluorescence assay of the induced ADSCs. 1 — Phase contrast; 2 — DAPI; 3 — CK18; 4 — Merged.

1

2

3

1

2

Phase contrast DAPI immunofluenscence Merged

Map2 NF GAPDH

Induced Induced Control Control

CK18 E-C GAPDH



staining demonstrated that the ADSCs were CD29, CD44, CD105, and Vimentin positive (Figure 2). The RT-PCR indicated that the ADSCs expressed CD29, CD71, CD73, CD105, and CD166 but didn’t express endothelial marker CD31 (Figure 1C).

Growth kineticsGrowth and proliferation of ADSCs were similar at P4, P14, P20,

and P26 according to the growth curves analysis (Figure 3). After a latency phase of 1 to 2 d, cell growth entered the logarithmic phase, and reached the plateau phase at approximately day 8. The PDT was 28, 31, 32, and 36 h for P4, P14, P20, and P26, respectively.

Colony-forming cell assayColony formation was observed by microscopy after 6 d. Colony-

forming levels were 37 6 0.2%, 30.6 6 0.1%, and 24 6 0.3% for passages 4, 12, and 27, respectively, demonstrating the capacity of cultured ADSCs for self-renewal (Figure 4).

Karyotype analysisThe diploid chromosome number of pig ADSCs was 2n = 38, con-

sisting a pair of sex chromosomes per gender. There was no normal diploid chromosome ploidy missing or broken (Figure 5).

Adipogenic differentiation of the ADPCsThe adipogenic potential was evaluated by differentiation of post-

confluent ADSCs. The obvious lipid droplet appeared after 10 d of induction, and it was in the cytoplasm. The lipid droplet was stained strongly with Oil O Red through fatty acids, and presented as bright red inclusions (Figure 5a). The expression of adipocyte specific mark-ers LPL and PPARg was tested by RT-PCR for 10-day differentiated cells. The undifferentiated cells were tested as controls. The results are shown in Figure 5b.

Osteogenic differentiation of the ADPCsMorphological changes were used as evidence for osteogenic

differentiation during the induction. At the end of the 22-day induction period, calcium crystals appeared in the cytoplasm and the differentiation was confirmed by alizarin staining for calcium (Figure 5c). The expression of osteoblast specific markers OPN, COL1, and RUNX2 was tested by RT-PCR for the 10-day and 22-day differentiated cells. The undifferentiated cells were tested as controls. The results are shown in Figure 5d.

Hepatic differentiation of the ADSCsMorphological changes in cultured ADSCs at passage 13 were

observed. Along with the differentiation, the cells at passage 13 lost their fibroblastic morphology gradually and became a flatter and broader in shape; at the 15th day of induction, these cells formed a polygonal shape. The ability of glycogen-storage at differentiated cells was formed using PAS staining (Figure 5e) AFB, a mature functional hepatocyte-specific marker and AFP, a marker of imma-ture hepatocytes, were tested by RT-PCR. Both differentiated cells expressed the AFB and ALP, but the undifferentiated cells were not expressed (Figure 5f).

Epithelial cells differentiation of the ADSCsDuring the induction, many morphological changes appeared

in the cultured cells. The cells at passage 6 lost their fibroblastic morphology gradually and became flatter and broader in shape; at the 13th day of differentiation, the cells appeared cobblestone in shape. Induction of 13-day cells was further confirmed by immu-nofluorescence stain for CK18 (Figure 5g). CK18 is a specific anti-body for epithelial cells. Specific surface markers were evaluated by RT-PCR and the undifferentiated cells were used as controls (Figure 5i).

Neurogenic differentiation of ADSCsThe ADSCs were pre-induced for 7 d, after which spindle-shaped

cells began to contract with an irregular form. Following induction, cell bodies further contracted and became round, triangular, or cone-shaped with multipolar processes. The processes continued to grow with many branches forming and cone-like terminal expan-sions were observed. A number of cells demonstrated very long processes, which appeared similar to the long axon of neurons. Immunofluorescence staining results showed that the MAP2, GFAP, and b-tubulin markers of neural cells were expressed in the differen-tiated cells (Figure 5j). The RT-PCR showed that the induced group cells were positive for NF and map2.

D i s c u s s i o nIn this study, ADSCs were successfully isolated from the visceral

omental region of pigs. The tissues were washed with PBS 8 times and digested through 1% I type collagenase for 1 h. The cells were cultured in a medium that contained 10% FBS, 1% B27, 1% gluta-mine, 1% streptomycin, and 10 ng/mL bFGF. Because bFGF can promote mitosis and stimulate anabolism, it is important to the growth of ADSCs and B27 is a secondary growth factor that pro-vides nutrients for ADSCs. The medium can guarantee ADSCs in the best state of growth. There were few cells that had adhered 48 h after inoculation. The cells first confluence to 80% was 10 d after inoculation. During the first 10 d, the medium was changed every 2 d. Later, cells were passaged stably over 2 d on average. At the 26th passage of ADSCs, the characteristics of aging appeared and the times of passage were prolonged by 4 d. This was maybe due to the cells themselves, several times the enzyme digestion, or the medium. Further research is required to discover the specific reasons for this.

There were no specific surface markers used to identify and screen for MSCs. So RT-PCR, immunofluorescent staining, and flow cytometry were used to test for the surface markers of ADSCs. The results showed ADSCs highly expressed CD29, CD44, CD71, CD73, CD90, CD166, and vimentin, weekly expressed CD105, and were unexpressed CD31.

We explored the ability of clone formation for ADSCs at passages 4, 12, and 26. Results showed that ADSCs have a strong ability of clone formation in a good growth state. In order to detect whether there was any variation in the proliferation of ADSCs, we conducted karyotype analysis. Results demonstrated there was no chromosome missing or broken in the ADSCs.



In order to study the pluripotency of ADSCs, we differentiated the cells into ectoderm, mesoderm, and endoderm. We differenti-ated the ADSCs into epithelial and neurogenic cells for ectoderm. The results of cell differentiation were verified by cell morphology, RT-PCR, and immunofluorescence stain. For epithelial cell mor-phology, the control groups were typical fibroblast-like shape and the induced group cells changed into a cobblestone-like shape. For RT-PCR, the control groups didn’t express epithelial-cell specific markers CK18 and E-cadherin, while the induced group cells highly expressed CK18 and E-cadherin. For immunofluorescence staining, the induced cells were positive to the antibody CK18, which is a marker for epithelial cells. For the neurogenic cells, the control group cells still retained a fibroblast-like shape, while the induced group cells changed into a typical neuronlike morphology and long axons were observed. Results of immunofluorescence staining showed the induced group cells were positive for GFAP, MAP2, and b-tubulin. The cells were then differentiated into adipocytes and osteoblasts for mesoderm. And we detected the results by cell morphology, staining and RT-PCR. For adipogenic differentiation, the induced cells appeared as obvious fat drops in the cytoplasm, expressed the adipocytes specific markers LPL and PPARg, and the fat drops stained red with Oil-Red O. For osteogenic differentiation, the induced cells had an oval-like shape with many aggregations. Along with the increase of aggregations, calcium nodules were emerging in the cytoplasm. The induced cells were stained with alizarin, and with Oil-Red O and expressed the osteoblast specific markers OPN, COL1, and RUNX2. For hepatic differentiation, the induced cells changed from a fibroblast-like shape to a polygonal or round shape. And RT-PCR results showed the induced cells expressed mRNA for ALB and AFP, while these were not expressed in control groups. We evaluated the efficiency of differentiation by PAS staining and the induced cells were dyed purple in the cytoplasm. Results showed that ADSCs have the ability of crossing layer differentiation. Though ADSCs can differentiate into 3 layers, the function of induced cells was not tested; therefore, more research is needed in this capacity.

Adipose tissue is an abundant source of mesenchymal stem cells, which have shown promise in the field of regenerative medicine. Various clinical trials have shown the regenerative capability of adipose-derived stem cells in subspecialties of medical fields such as plastic, orthopedic, oral maxillofacial, and cardiac surgeries. Breast reconstruction and augmentation trials have been reported by Yoshimura et al (11–12). Adipose-derived stem cells were first used to stimulate craniofacial bone repair in calvarial defects (13). These stem cells have been used to heal chronic fistulas in Crohn’s disease (14–15).

Myocardial infarction is a life-threatening medical emergency, which is the greatest cause of death in developing countries. Until now, myocardial infarction is mainly dealt with through drugs and surgery, which can have many side effects on patients. Adipose tissue taken from around the heart has shown that ADSCs can differentiate into myocardial cells in vitro. Stem cell therapy is an effective way to cure many diseases, so it is significant if ADSCs derived from vis-ceral (omental region) could be differentiated into myocardial cells and be used to induce cells to cure myocardial infarction.

In conclusion, ADSCs were isolated from visceral omental region of the pig, and the self-renewal ability and differential potential

were evaluated in vitro. The present study illustrates the potential application of adipose tissue as an adult stem cell source for regen-erative therapies.

Re f e r e n c e s 1. Ogawa R, Hanawa H, Hirai Y, et al. Potential utility of adipose-

derived stromal cells (ASCs) as a carrier for ex vivo gene therapy — Gene transfer efficiency of ASCs. Molecular Therapy 2005; 11:S412.

2. Shen J-F, Sugawara A, Yamashita J, Ogura H, Sato S. Dedifferen-tiated fat cells: An alternative source of adult multipotent cells from the adipose tissues. Int J Oral Sci 2011;3:117–124.

3. Jones PM, Courtney ML, Burns CJ, Persaud SJ. Cell-based treat-ments for diabetes. Drug Discov Today 2008;13:888–893.

4. Urbán VS, Kiss J, Kovács J, et al. Mesenchymal stem cells coop-erate with bone marrow cells in therapy of diabetes. Stem Cells 2008;26:244–253.

5. Karnieli O, Izhar-Prato Y, Bulvik S, Efrat S. Generation of insulin-producing cells from human bone marrow mesenchymal stem cells by genetic manipulation. Stem Cells 2007;25:2837–2844.

6. Gu LH, Zhang TT, Li Y, Yan HJ, Qi H, Li FR. Immunogenicity of allogeneic mesenchymal stem cells transplanted via different routes in diabetic rats. Cell Mol Immunol 2014;12:444–455.

7. Kim I, Bang SI, Lee SK, Park SY, Kim M, Ha H. Clinical impli-cation of allogenic implantation of adipogenic differentiated adipose-derived stem cells. Stem Cells Transl Med 2014;3: 1312–1321.

8. Lohan P, Coleman CM, Murphy JM, Griffin MD, Ritter T, Ryan AE. Changes in immunological profile of allogeneic mesen-chymal stem cells after differentiation: Should we be concerned? Stem Cell Res Ther 2014;5:99.

9. Zhu Y, Liu T, Song K, Fan X, Ma X, Cui Z. Adipose-derived stem cell: A better stem cell than BMSC. Cell Research 2008;18:s165.

10. Gong X, Hou L, Bai C, et al. Isolation and biological characteris-tics of chicken adipose-derived progenitor cells. DNA Cell Biol 2011;30:453–460.

11. Yoshimura K, Sato K, Aoi N, Kurita M, Hirohi T, Harii K. Cell-assisted lipotransfer for cosmetic breast augmentation: Supportive use of adipose-derived stem/stromal cells. Aesthetic Plast Surg 2008;32:48–55.

12. Yoshimura K, Asano Y, Aoi N, et al. Progenitor-enriched adipose tissue transplantation as rescue for breast implant complications. Breast J 2010;16:169–175.

13. Lendeckel S, Jödicke A, Christophis P, et al. Autologous stem cells (adipose) and fibrin glue used to treat widespread traumatic calvarial defects: Case report. J Craniomaxillofac Surg 2004; 32:370–373.

14. García-Olmo D, García-Arranz M, Herreros D, Pascual I, Peiro C, Rodríguez-Montes JA. A phase I clinical trial of the treatment of Crohn’s fistula by adipose mesenchymal stem cell transplanta-tion. Dis Colon Rectum 2005;48:1416–1423.

15. Garcia-Olmo D, Garcia-Arranz M, Herreros D. Expanded adi-pose-derived stem cells for the treatment of complex perianal fistula including Crohn’s disease. Expert Opin Biol Ther 2008; 8:1417–1423.


Article


I n t r o d u c t i o nBreast tumors are the most common tumor in female dogs (1,2).

They have no breed predisposition and mainly affect senile animals and those not castrated before the first heat (1). Dogs may be used as the natural surrogate model for the evaluation of tumors in the human female breast because of similarities between the clinical, pathological, epidemiological, and developmental characteristics of tumors in dogs and humans (3).

Several systems have been proposed to assess the prognosis of these animals. One such system is the TNM, where: T — describes

the tumor size; N — lymph node metastasis; and M — distant metas-tasis (4). There are studies showing a strong correlation between the tumor size and malignancy (5) with the presence of the metastases in the lymph nodes and the survival of the affected animal (6,7). In human female patients, the TNM system is still widely used as a predictive factor for mammary carcinomas, although more modern methods such as biomarkers are available (8).

Another method that can be used for the prognostic evaluation of tumors is histological grading (9). Grade III tumors have been associated with a low survival time (10,11). It has also been estab-lished, with regard to the histological classification of the tumors,

Clinical staging in bitches with mammary tumors: Influence of type and histological grade

Lígia F. Gundim, Camila P. de Araújo, William T. Blanca, Ednaldo C. Guimarães, Alessandra A. Medeiros

A b s t r a c tBreast tumors are the most common tumors in dogs and the study of disease prognostic factors is important for establishing the appropriate treatment protocols. The purpose of this study was to clinically stage mammary tumors of bitches and correlate the stages with histological type and grade. The tumors of 63 dogs were clinically staged based on the findings of tumor sizing, lymph node evaluation, and radiographic examination. After surgical excision, the tumors were classified histologically and graded. The relationship between the tumor grade, stage, and histological type was evaluated using a binomial test. Stage I tumors were the most numerous (31.75%), followed by tumors at stages II, III, IV, and V. Animals with histological grade I carcinomas presented stage I, II, or III tumors more frequently and stage IV and V tumors less frequently. The number of animals with simple carcinomas that were at stage I of the disease was greater than that at stage V. Carcinomas in the mixed tumors were less aggressive; however, the small number of animals in stage V of the disease made any statistical association impossible. The complex carcinomas presented with the invasion of the lymph nodes and less cellular differentiation in a larger number of animals than did simple carcinomas. Histological grading proved to be the best parameter for the prognostic evaluation of the breast carcinomas.

R é s u m éLes tumeurs mammaires sont les tumeurs les plus fréquentes chez les chiens et l’étude des facteurs de pronostic de la maladie est importante afin d’établir les protocoles de traitement appropriés. Le but de cette étude était de déterminer le stade clinique des tumeurs mammaires de chiennes et de corréler les stades au type histologique et le grade. Les tumeurs de 63 chiens ont été classées cliniquement en se basant sur la taille de la tumeur, l’évaluation des nœuds lymphatiques, et l’examen radiographique. Après excision chirurgicale, les tumeurs ont été classées histologiquement et un grade attribué. La relation entre le grade de la tumeur, le stade, et le type histologique a été évaluée en utilisant un test binomial. Les tumeurs de Stade I étaient les plus nombreuses (31,75 %), suivies des tumeurs des stades II, III, IV, et V. Les animaux avec un carcinome de grade I histologiquement présentaient des tumeurs de stade I, II, ou III plus fréquemment et des tumeurs de stade IV et V moins fréquemment. Le nombre d’animaux avec un carcinome simple qui était au stade I de la maladie était plus grand que ceux au stade V. Les carcinomes dans les tumeurs mixtes étaient moins agressifs; toutefois, le petit nombre d’animaux au stade V de la maladie rendait impossible toute association statistique. Les carcinomes complexes se présentaient avec une invasion des nœuds lymphatiques et moins de différenciation cellulaire dans un plus grand nombre d’animaux que les carcinomes simples. Le pointage histologique s’est avéré être le meilleur paramètre pour l’évaluation du pronostic des carcinomes mammaires.


Laboratory of Veterinary Pathology, Faculty of Animal Medicine, Federal University of Uberlândia, Av. Mato Grosso, 3289, Bloco 2S. Campus Umuarama. Uberlândia, MG. Brazil (Gundim, de Araújo, Blanca, Medeiros); Faculty of Mathematics, Federal University of Uberlândia, Av. João Naves de Ávila, 2121. Campus Santa Mônica. Uberlândia, MG. Brazil (Guimarães).

Address all correspondence to Dr. Lígia F. Gundim; e-mail: [email protected]

Received December 11, 2015. Accepted April 4, 2016.



that complex carcinomas exhibit less aggressive behavior than the simple carcinomas. Of the simple carcinomas, anaplastic carcinoma is the most aggressive, followed by solid papillary, and tubule car-cinomas (1).

Owing to the paucity of studies in dogs assessing histological gradation, histological typing, and staging as prognostic factors, this study aimed to examine the clinical staging of breast malignancies in bitches and correlate the stages of tumors with their histological type and grading.


SamplingThe study included 63 dogs with mammary tumors that visited

the Veterinary Hospital of the Federal University of Uberlândia vol-untarily. Inclusion criteria for the selection of animals were: bitches with malignant mammary neoplasia, no pre-existing diseases, avail-ability of documented medical history, and previous histopathologi-cal and radiological examinations. All dogs included in the study were intact (not spayed).

Clinical stagingIn order to determine the clinical stage of the tumors, chest X-rays

were performed at the ventrodorsal and right and left laterolateral positions. The animals also underwent a complete physical exami-nation, including the careful palpation of the mammary glands and verification of the general clinical condition of the animal. During the clinical examination, both the mammary chains and the regional lymph nodes were explored by examining the number of affected mammary glands, the presentation of tumor masses, the tumor size, and the surface of the lymph nodes involved.

All dogs underwent surgical removal of breast and regional lymph nodes. Fragments of these tissues were collected and subjected to histopathological evaluation. Samples were fixed in 10% buffered formalin and subjected to histological processing (12), followed by the preparation of the histopathological slides stained with hematoxylin-eosin (H&E) stain.

Using data on tumor sizes, radiological findings, and assessment of the presence of the metastatic lymph nodes by histopathology, the tumors of the bitches were staged according to the TNM system (Table I) (4).

Histological classificationBreast tumors were histologically classified according the method

of Cassali et al (2). Neoplasms were grouped according to their histogenesis into complex carcinomas, simple carcinomas (tubular, papillary, solid, and anaplastic carcinomas), and mixed tumors. The invasive lobular carcinomas, inflammatory carcinomas, carcinosar-comas, carcinomas in situ, osteosarcomas, and fibrosarcomas were classified as other types of tumors.

Histologic gradingHistological grading of breast tumors was performed by 2 observ-

ers according to the method of Elston and Ellis (9), in which the tumors were classified as being grade I, II, or III, on the basis of the formation of the tubules, nuclear pleomorphism, and mitosis number.

Statistical analysisThe relationship between the histological grade of tumors, their

clinical staging, and their histological type, was evaluated using the binomial test for comparing 2 proportions, with a P = 0.05 signifi-cance value. The statistical analyses were performed using the Action program, version 2.9 (Software Action. Estatcamp-Consultoria em estatística e qualidade, São Carlos — SP, Brasil.)

Re s u l t sAnimals aged 8 to 14 years were the most affected (77.19%).

Mongrel dogs showed the highest prevalence of breast tumors (44.26%), followed by poodles with 19.67%. No dogs were castrated prior to the removal of breast tumors, although some were castrated at the time of tumor resection. The distribution of bitches according to the results of the clinical staging is shown in Table I.

The most frequent tumor type was tubular carcinoma, followed by carcinoma in mixed tumors (Table III). Among the animals

Table I. Staging system for breast tumors, according Owen (4)

Regional Distant Stage Size metastasis metastasisStage I T1 N0 M0Stage II T2 N0 M0Stage III T3 N0 M0Stage IV Any size N1 M0Stage V Any size N0 or N1 M1T1 — # 3 cm; T2 — 3 to 5 cm; T3 — $ 5 cm; N0 — without lymph node metastasis; N1 — metastasis in lymph node present; M0 — absence of distant metastases; M1 — presence of distant metastases.

Table II. Number of dogs according to the clinical stage

Clinical stage Number Frequency (%)Stage I 20 31.75Stage II 12 19.05Stage III 12 19.05Stage IV 14 22.22Stage V 5 7.93

Table III. Number of tumors according to histological classification

Histological classification Number Frequency (%)Tubular carcinoma 17 26.56Carcinoma in mixed tumor 15 23.44Solid carcinoma 11 17.19Complex carcinoma 7 10.94Papilar carcinoma 6 9.37Others 8 12.50



with simple carcinomas, most were classified as stage I, with only 3 animals presenting lung metastases (stage V) and 10 animals exhib-iting lymph node metastases (stage IV). It was observed that dogs with simple carcinomas were most frequently classified as stage I and less frequently as stage V (P = 0.026). None of the animals in the mixed tumor group were classified as stage V and only one was classified as stage IV. However, there were no significant differences in the frequency of bitches among the clinical stages for this histo-logical type (P = 0.07) (Table IV).

With regard to the histologic grading of the tumors, 60 carcinoma samples were graded; tumors diagnosed as osteosarcomas and fibrosarcomas were excluded since carcinomas only were evaluated for histological grading. Among the tumors classified as simple carcinomas, 40% (14/35) were classified as grade I, 50% (18/36) as grade II, and 11.1% (4/36) as grade III. Among the complex carci-nomas, 85.7% (6/7) were classified as grade I and 14.28% (1/7) as grade II. Among the carcinomas in the mixed tumor group, 68.8% (11/16) were classified as grade I and 31.3% (5/16) as grade II. It was observed that the animals with the histological grade I presented stage I, II, or III tumors more frequently, and stage IV (P = 0.01) and stage V (P = 0.0009) tumors less frequently (Table V).

D i s c u s s i o nPrevious studies have reported tumor metastases in less than 50%

of dogs, which corresponds with the findings of this study. A retro-spective study of 99 dogs in the United States reported 25% of the animals at stage I, 12% at stage II, and 15% at stage III of the disease (13). However, a study in Brazil that examined 36 dogs described 22.2% of the animals as stage I and 75% as stage II of the disease; none of the animals were classified as stage III (14).

Given that the differentiation between stages I, II, and III is based only on the tumor size, another study conducted in Brazil reported that 35% of malignant tumors were smaller than 5.0 cm (11), while other studies have reported that neoplasms measured less than 3.0 cm [48.81% (15) and 44.7% (16), respectively].

A study involving human female subjects reported that 55.3% of patients presented tumors smaller than 3 cm and that the tumor diameter negatively affected the chance of survival despite resection (7). The smaller tumor size in these female subjects might have been due to early diagnosis. A similar correlation was observed herein which might be due, in part, to the owners seeking medical attention for swelling in the breasts sooner than usual.

With regard to the bitches diagnosed with the lymph node inva-sion in the present study (29.7%), previous studies conducted in

Portugal and Colombia showed similar frequencies of occurrence [298% (16) and 30% (17), respectively]. Another study, however, had classified only 15.46% of the animals as stage IV (13).

The frequency of distant metastases in the present study was 7.9%. Previous studies have also reported similarly low frequencies of bitches at this stage of the disease; Ribas et al (18) and Gomes et al (17) reported that 8.3% and 16.7% of dogs were diagnosed with pulmonary metastasis, respectively. However, a negative radiological finding does not prove the absence of metastases, since the detection of tumors is challenging owing to their smaller sizes (19). Moreover, highly accurate diagnostic methods such as magnetic resonance imaging (MRI) and positron emission tomography (PET) scans are rarely practiced in veterinary medicine. Another reason for the fewer number of dogs at stage V of the disease is the fact that these animals are often already at advanced stages of the disease without having received any surgical treatment.

With regard to the frequency of tumors according to histological type, previous studies have reported that tubular carcinoma is the most common (20,21), which corresponds to the findings in this study. However, the incidences of other types of tumors reported by different studies vary. The different histological classification systems used for the canine mammary tumors makes it difficult to compare the results of the various studies. The solid carcinoma has been reported as the second most frequent type of carcinoma in a previous study (20); however, another study reported the complex carcinoma as being the second most frequent (21). Carcinoma in mixed tumors also appears as one of the more frequent types of carcinomas (22), and in the present study, it was found to be the third most frequent.

The simple carcinomas were found to have the highest frequency of occurrence in this study when the tumors were grouped on the basis of their histological origin. The simple carcinomas have also been reported as the most frequent type of tumors in studies con-ducted in Japan (23), Brazil (24), and Iran (21).

Several authors have reported that simple carcinomas are more aggressive and have a worse prognosis than complex carcinomas (25,26). Therefore, the clinical staging of simple carcinomas in the present study was unexpected and can be attributed to the fewer number of the animals at the stages II, III, and V.

However, the staging of the carcinomas in mixed tumors in this study revealed the lower aggressiveness of these tumors, as has pre-viously been described by Cassali et al (26), who also reported that these tumors have a better prognosis compared to other carcinomas. However, in the present study, the statistical corroboration of this fact was hindered because none of the animals with carcinomas in mixed tumors were at stage V.

Table IV. Frequency of histological types of breast carcinomas according to the clinical stage

StageHistological type I II III IV VSimple carcinoma 13 5 5 10 3Complex carcinoma 1 1 1 3 1Carcinoma in mixed tumor 5 5 5 1 0Others 1 1 1 0 1

Table V. Frequency of histological grade of breast carcinomas according to the clinical stage

Grade Stage I Stage II Stage III Stage IV Stage V1 11 6 5 3 12 8 5 3 9 33 1 0 2 2 1



With regard to the complex carcinomas, we did not expect a large number of the animals to present lymph node invasions (stage IV) since Rasotto et al (27) had previously shown that 66.6% of the simple carcinomas in their study had metastasized to the lymph nodes and in 81.01% of the cases, invasion of the lymphatic vessels was observed; whereas, among the complex carcinomas, only 11.11% had metastasized to the lymph nodes and 20% showed lymphatic invasion.

The time-evolution of the tumor might affect the occurrence of the metastases. The larger tumors usually have a longer develop-ment duration, which increases the likelihood of metastasis (28). Unfortunately, the determination of the duration of tumor progres-sion depends on the owner’s observation of the condition and is therefore often unclear.

In the histological grading findings, complex carcinomas showed less differentiation compared to simple carcinomas, as has been reported previously (27,10). The degree of differentiation of simple carcinomas is a matter of debate; it has been reported that this histo-logical type of carcinoma has no association with the degree of differ-entiation (27) and that their classification as grade III (10) is frequent. The discrepancies between the results of the various studies might be because of the subjectivity of the gradation method, which can vary according to the reviewer. Another hypothesis is that a greater number of tubular tumors were classified as carcinomas in this study since tubule formation is one of the parameters considered for gradation, and tubular carcinomas tend to attain lower scores in the evaluation.

Histological grade is a factor that can influence the clinical staging of a tumor. A study in Slovenia reported that 64.71% of animals had stage I and grade I tumors, a result that is comparable to what we presented in this study (29). Similarly, a study conducted in Madrid described a significant association between the histological grade and the clinical staging of tumors (30). Other previous studies have also correlated the type and histological grade of the tumors with the overall survival rate (16).

A limitation of the present study was the fewer number of ani-mals at stage V of the disease. Although mixed tumor carcinomas typically show less aggression, fewer metastases, and lower cell differentiation, whereas simple carcinomas are more aggressive, this association could not be proven statistically because of the fewer number of animals at stage V. It can, however, be concluded that the histological grading proved to be the best parameter for the prognostic evaluation of breast carcinomas.

Re f e r e n c e s 1. Misdorp, W. Tumors of the mammary gland. In: Meuten, DJ, ed.

Tumors in Domestic Animals. 4th ed. Ames, lowa: Iowa State Pr, 2002:575–606.

2. Cassali GD, Lavalle GE, de Nardi AB, et al. Consensus for the diagnosis, prognosis and treatment of canine mammary tumors. Braz J Vet Pathol 2011;4:153–180.

3. Cassali GD. Comparative mammary oncology: Canine model. BMC proceedings 2013;7:1.

4. Owen LN. TNM Classification of Tumors in Domestic Animals, 1st ed. Geneva, Switzerland: World Health Organization, 1980: 1–53.

5. Sorenmo KU, Kristiansen VM, Cofone MA, et al. Canine mam-mary gland tumours; a histological continuum from benign to malignant; clinical and histopathological evidence. Vet Comp Oncol 2009;7:162–172.

6. de Araújo MR, Campos LC, Ferreira E, Cassali GD. Quantitation of the regional lymph node metastatic burden and prognosis in malignant mammary tumors of dogs. J Vet Intern Med 2015; 29:1360–1367.

7. Dong G, Wang D, Liang X, et al. Factors related to survival rates for breast cancer patients. Int J Clin Exp Med 2014;7:3719–3724.

8. Orucevic A, Chen J, McLoughlin JM, Heidel RE, Panella T, Bell J. Is the TNM staging system for breast cancer still relevant in the era of biomarkers and mmerging personalized medicine for breast cancer — An institution’s 10-year experience. Breast J 2015;21:147–154.

9. Elston CW, Ellis IO. Assessment of histological grade. In: Elston CW, Ellis IO, eds. Systemic Pathology. The Breast. 3rd ed. London: Churchill Livingstone, 1998:365–384.

10. Karayannopoulou M, Kaldrymidou E, Constantinidis TC, Dessiris A. Histological grading and prognosis in dogs with mammary carcinomas: Application of a human grading method. J Comp Path 2005;133:246–252.

11. Estrela-Lima A, Araújo MSS, Costa-Neto JM, et al. Immunophe-notypic features of tumor infiltrating lymphocytes from mam-mary carcinomas in female dogs associated with prognostic factors and survival rates. BMC Cancer 2010;10:256.

12. de Tolosa EMC, Rodrigues CJ, Behmer OA, Neto AGF. Manual de Técnicas Para Histologia Normal E Patológica. 1st ed. Barueri, Brazil: Manole, 2003:1–341.

13. Philibert JC, Snyder PW, Glickman N, Glickman LT, Knapp DW, Waters DJ. Influence of host factors on survival in dogs with malignant mammary gland tumors. J Vet Intern Med 2003;17: 102–106.

14. Rêgo MAS, dos Santos FL, Rabelo SSA, de Almeida EL, de Lima ER, Menezes MM. Caracterização do perfil clínico, laboratorial, citológico e histopatológico em cadelas portadoras de tumor de mama submetidas à mastectomia [Master Thesis]. Recife, Brazil: Universidade Federal Rural de Pernambuco, 2012.

15. Kristiansen VM, Nødtvedt A, Breen AM, et al. Effect of ovario-hysterectomy at the time of tumor removal in dogs with benign mammary tumors and hyperplastic lesions: A randomized controlled clinical trial. J Vet Intern Med 2013;27:935–942.

16. Santos AA, Lopes CC, Ribeiro JR, et al. Identification of prognos-tic factors in canine mammary malignant tumours: A multivari-able survival study. BMC Vet Res 2013;9:1.

17. Gómez BJ, Ramírez, MR, Maldonado JE. Presence of lung metas-tases in bitches affected by malignant mammary neoplasms in Medellin (Colombia). Rev MVZ Córdoba 2012;17:2983–2990.

18. Ribas CR, Dornbusch PT, de Faria MR, Wouk AFPF, Cirio SM. Alterações clínicas relevantes em cadelas com neoplasias mamárias estadiadas. Arch Vet Sci 2012;17:60–68.

19. Cavalcanti MF, Cassali GD. Fatores prognósticos no diagnóstico clínico e histopatológico dos tumores de mama em cadelas — Revisão. Clín Vet 2006;61:56–64.

20. Salas Y, Márquez A, Diaz D, Romero L. Epidemiological study of mammary tumors in female dogs diagnosed during the



period 2002–2012: A growing animal health problem. PLoS ONE 2015;10:1.

21. Tavasoly A, Golshahi H, Rezaie A, Farhadi M. Classification and grading of canine malignant mammary tumors. Vet Res Forum 2013;4:25–30.

22. Cintra PP, de Paula CAT, Calazans SG, de Souza JL, Magalhães GM. Reclassificação e determinação do tipo histológico predomi-nante em neoplasias mamárias caninas do hospital veterinário da universidade de franca nos anos de 2010 a 2012. Encicélopdia Biosfera 2014;10:2606–2615.

23. Yoshida K, Yoshida S, Choisunirachon N, et al. The relationship between clinicopathological features and expression of epithelial and mesenchymal markers in spontaneous canine mammary gland tumors. J Vet Med Sci 2014;76:1321–1327.

24. Oliveira Filho JC, Kommers GD, Masuda EK, et al. Estudo retro-spectivo de 1.647 tumores mamários em cães. Braz J Vet Res 2010; 30:177–185.

25. Cassali GD, Lavalle GE, Ferreira E, et al. Consensus for the diagnosis, prognosis and treatment of canine mammary tumors — 2013. Braz J Vet Pathol 2014;7:38–69.

26. Cassali GD, Bertagnolli AC, Ferreira E, Damasceno KA, Gamba CO, Campos CB. Canine mammary mixed tumours: A review. Veterinary Medicine International 2012;2012:1–7.

27. Rasotto R, Zappulli V, Castagnaro M, Goldschmidt MH. A retrospective study of those histopathologic parameters pre-dictive of invasion of the lymphatic system by canine mammary carcinomas. Vet Pathol 2012;49:330–340.

28. Matos AJ, Baptista CS, Gärtner MF, Rutteman GR. Prognostic studies of canine and feline mammary tumours: The need for standardized procedures. Vet J 2012;193:24–31.

29. Cerovšek M, Plavec T, Zrimšek P, Pogacnik M, Zabavnik J. Clinicopathological survey of 56 canine malignant mammary tumours in Slovenia — Prognostic value of clinical stage and histological grade. Slov Vet Res 2013;50:93–102.

30. Peña L, De Andrés PJ, Clemente M, Cuesta P, Pérez-Alenza MD. Prognostic value of histological grading in noninflamma-tory canine mammary carcinomas in a prospective study with two-year follow-up: Relationship with clinical and histological characteristics. Vet Pathol 2013;50:94–105.


Article


I n t r o d u c t i o nAcepromazine is the phenothiazine derivative most commonly

used for sedation in dogs. The drug has been found to possess antagonistic action at dopaminergic receptors within the brain; this may explain the mechanism by which acepromazine induces seda-tion (1). Based on extensive data on the use of acepromazine in dogs, the degree of sedation following acepromazine administration in canine patients is mild to moderate (2–4).

It has been reported that acepromazine is devoid of analgesic properties (1). Therefore, acepromazine is commonly administered

in combination with opioid analgesics to provide sedation and anal-gesia, which facilitates handling of dogs for placement of venous catheters, preparation for surgery, and diagnostic procedures. In addition to providing analgesia, there is evidence that an opioid analgesic enhances the degree of sedation induced by acepromazine in dogs. In a previous study, methadone, morphine, butorphanol, and tramadol were compared and methadone appeared to be the most effective for this purpose (3).

Tramadol has been classified as an opioid analgesic although it was found to provide analgesia by opioid and non-opioid mecha-nisms (5). In humans, its opioid analgesic properties were associated

Tramadol does not enhance sedation induced by acepromazine in dogsEduardo R. Monteiro, Renan B. Lobo, Juarez S. Nunes Jr, Julia P.P. Rangel, Flavia S. Bitti

A b s t r a c tThe sedative effect of acepromazine combined with 2 doses of tramadol [3 and 5 mg/kg body weight (BW)] was compared with the sedative effect of acepromazine alone in dogs and the effects of each sedative protocol on cardiorespiratory variables were examined. This was a prospective, randomized, blinded, crossover study. Each of 6 dogs received 3 treatments at 1-week intervals. During all anesthetic episodes, dogs received 0.05 mg/kg BW acepromazine. Approximately 25 min later, dogs were given physiological saline (control) or tramadol [3 mg/kg BW (TR3) or 5 mg/kg BW (TR5)]. All drugs were administered intravenously. Variables evaluated included heart rate (HR), respiratory rate (RR), systolic, mean, and diastolic blood pressures (SAP, MAP, and DAP), and sedation [by use of a simple descriptive scale (SDS, range: 0 to 3) and a numeric rating scale (NRS, range: 0 to 10)]. Variables were recorded 25 min after acepromazine and for 80 min after saline or tramadol. Acepromazine administration resulted in mild sedation in most dogs and decreased RR, SAP, MAP, and DAP in all treatments. Tramadol administration did not significantly increase SDS or NRS scores compared to acepromazine alone. The only exception to this rule was observed at 20 min after TR3, when NRS was higher in this group than in the control treatment. Administration of tramadol (TR3 and TR5) decreased HR. Under the conditions of this study, sedation induced by acepromazine with tramadol was similar to that of acepromazine alone. The main adverse effects of the combination were a decrease in blood pressure and HR, without clinical significance.

R é s u m éL’effet sédatif de l’acépromazine combiné à deux doses de tramadol [3 et 5 mg/kg de poids corporel (PC)] a été comparé à l’effet sédatif de l’acépromazine seul chez des chiens et les effets de chaque protocole de sédation sur des variables cardio-respiratoires ont été examinés. Il s’agissait d’une étude prospective croisée, randomisée, réalisée à l’aveugle. Chacun des six chiens a reçu trois traitements à des intervalles de 1 semaine. Durant tous les épisodes anesthétiques, les chiens ont reçu 0,05 mg/kg PC d’acépromazine. Environ 25 min plus tard, les chiens ont reçu de la saline physiologique (témoin) ou du tramadol [3 mg/kg PC (TR3) ou 5 mg/kg PC (TR5)]. Toutes les drogues étaient administrées par voie intraveineuse. Les variables évaluées incluaient le rythme cardiaque (RC), le rythme respiratoire (RR), les pressions sanguines systolique, moyenne, et diastolique (PSS, PSM, et PSD), et la sédation [en utilisant une échelle descriptive simple (EDS, écart : 0 à 3) et une échelle de gradation numérique (EGN, écart : 0 à 10)]. Les variables ont été enregistrées 25 min après l’acépromazine et pendant 80 min après l’administration de saline ou de tramadol. L’administration d’acépromazine a résulté en une légère sédation chez la plupart des chiens et on nota une diminution de RR, PSS, PSM, et PSD avec tous les traitements. L’administration de tramadol ne fit pas augmenter de manière significative les pointages EDS et EGN lorsque comparée à l’acépromazine seul. La seule exception à cette règle a été observée à 20 min après TR3, alors que l’EGN était plus élevée dans ce groupe comparativement au témoin. L’administration de tramadol (TR3 et TR5) entraîna une diminution du RC. Dans les conditions de la présente étude, la sédation induite par l’acépromazine avec du tramadol était similaire à celle de l’acépromazine seul. Les principaux effets adverses de la combinaison étaient une diminution de la pression sanguine et du RC, mais sans signification clinique.


School of Veterinary Medicine, University of Vila Velha, Vila Velha, ES, Brazil 29102-920 (Monteiro, Lobo, Nunes Jr, Rangel, Bitti)

Address all correspondence to Dr. Eduardo R. Monteiro; telephone: 55-51-96006922; fax: 55-51-33087305; e-mail: [email protected]

Received January 28, 2016. Accepted March 28, 2016.



with the production of the active metabolite O-desmethyltramadol (M1) (6), which was found to possess greater affinity for opiate recep-tors than the parent drug tramadol (7). Drowsiness has been reported as an adverse effect after tramadol administration in humans (8).

There is conflicting evidence about the sedative effect of tramadol in dogs. In one study, a dose-related sedative effect was observed in dogs administered 1, 2, or 4 mg/kg body weight (BW) tramadol by intravenous (IV) injection (9), whereas no sedation was evidenced after IV tramadol (approximately 4 mg/kg BW) to dogs in another study (10). When combined with acepromazine, tramadol appeared to be less effective in enhancing the degree of sedation induced by the phenothiazine than methadone, morphine, and butorphanol (3). However, the dose of tramadol (2 mg/kg BW) was not equipotent to the dose of other opioids and this may have accounted for the low efficacy of tramadol in this later study (3). The present study aimed to compare the sedative effect of acepromazine combined with 2 different doses of tramadol [3 mg/kg BW (TR3) or 5 mg/kg BW (TR5)] to the sedative effect of acepromazine alone in dogs. We hypothesized that tramadol would enhance sedation induced by acepromazine. A second objective of this study was to evaluate the effects of each sedative protocol on cardiorespiratory variables of dogs.

M a t e r i a l s a n d m e t h o d sThe present study was initiated after approval by the institutional

Animal Research Ethical Committee (protocol 324-2014). Six healthy adult crossbreed dogs (4 male and 2 female) were used. Average weight of the dogs was 18.7 6 3.1 kg (mean 6 SD). Healthy status

was confirmed by physical examination and laboratory evaluation [complete blood (cell) count (CBC) and biochemistry profile].

This was a prospective, randomized, blinded, crossover study. Each dog received 3 treatments on different occasions, with 1-week washout intervals. During all anesthetic episodes, dogs were admin-istered acepromazine (Acepran 0.2%; Vetnil, Louveira, São Paulo, Brazil), 0.05 mg/kg BW, IV. Approximately 25 min after the adminis-tration of acepromazine, physiological saline (control) IV or tramadol IV (Cloridrato de tramadol 50 mg/mL; Hipolabor Farmacêutica, Belo Horizonte, Minas Gerais, Brazil) at TR3 or TR5, was given.

Dogs were fasted for 12 h prior to each experiment, but had free access to water. A 20-gauge catheter was introduced into a cephalic vein and the dogs were acclimated for 20 min to the laboratory environment. Baseline values for each variable were then recorded with the dogs gently restrained in lateral recumbency. Heart rate (HR) was measured with a stethoscope and respiratory rate (RR) was counted by observing chest wall movements. An oscillometric device (PetMap Classic; Ramsey Medical, Tampa, Florida, USA) was used for indirect measurement of systolic, mean, and diastolic blood pressures (SAP, MAP, and DAP, respectively). The blood pressure cuff was positioned proximal to the carpus in the nondependent limb and the cuff size was chosen according to the manufacturer’s directions. For every time point, 5 consecutive measurements of blood pressure were performed and averaged.

Sedation was scored by use of a simple descriptive scale (SDS) and a numeric rating scale (NRS). The SDS ranged from 0 to 3 as follows: 0 — no sedation; 1 — mild sedation, less alert but still active; 2 — moderate sedation, drowsy, recumbent but can walk; or 3 — intense sedation, very drowsy, unable to walk (11,12). The NRS ranged from

Table I. Medians (interquartile range) simple descriptive scale (SDS, range: 0 to 3) and numeric rating scale (NRS, range: 0 to 10) sedation scores in 6 dogs. Sedation was assessed at 25 min after administration of acepromazine, 0.05 mg/kg body weight (BW), IV (time point ACP) and at 20, 40, 60, and 80 min after administration of physiological saline (control) IV, 3 mg/kg BW tramadol (TR3) or 5 mg/kg BW tramadol (TR5) IV

ACP 20 40 60 80SDS Control 1.0 1.0 1.0 1.0 1.0 (1.0 to 1.3) (1.0 to 1.3) (1.0 to 1.3) (1.0 to 1.3) (0.0 to 1.3)

TR3 1.0 1.0 1.5 1.5 1.0 (1.0 to 2.0) (1.0 to 2.0) (1.0 to 2.0) (1.0 to 2.0) (1.0 to 2.0)

TR5 1.5 2.0 1.0 1.0 1.0 (1.0 to 2.0) (1.0 to 2.0) (1.0 to 2.0) (1.0 to 2.0) (0.8 to 2.0)

NRS Control 2.5 2.5 2.0 1.0 1.0 (2.0 to 4.3) (1.8 to 4.0) (1.0 to 2.9) (1.0 to 2.5) (0.0 to 2.3)a

TR3 4.0 4.5 4.0 3.3 2.8 (3.9 to 5.3) (3.8 to 6.3)b (2.5 to 5.3) (1.8 to 5.0) (1.4 to 3.3)a

TR5 4.0 4.3 3.0 2.0 1.0 (3.4 to 4.3) (3.8 to 5.3) (2.0 to 4.0) (1.8 to 4.0) (0.8 to 3.3)a

a Significant difference compared to time point ACP.b Significant difference compared to the control treatment (P , 0.05).



0 to 10, where 0 represented no sedation and 10 represented the most sedation possible (12). For the NRS, only whole numbers could be selected. A single observer, unaware of the treatment administered, was responsible for scoring SDS and NRS values. This person was familiar with both scoring systems. Dogs were initially observed without interaction with the observer. Thereafter, objective variables (HR, RR, SAP, MAP, and DAP) were recorded. Finally, the observer encouraged the dogs to stand and walk. Based on noninteractive and interactive behaviors, the observer recorded SDS and NRS scores.

After recording baseline variables, the dogs were administered acepromazine through the cephalic catheter. Within 20 to 25 min after acepromazine administration, all variables were recorded (time point ACP). Thereafter, physiological saline (control), TR3, or TR5 was administered over 1 min through the cephalic catheter and all variables were reassessed at 20 min intervals for 80 min (time points 20, 40, 60, and 80).

Sample size calculation was performed using G*Power for Windows Version 3.1.6 (Heinrich Heine Universität Düsseldorf, Germany) and was based on sedation scores reported for dogs in previous studies (2–3). Results of the power analysis indicated 6 dogs would be necessary to detect 1.0 point differences between groups for SDS scores and 3.0 point differences between groups for NRS scores with a power of 80% at 5% level of significance.

Statistical analyses were performed by use of a computer soft-ware (Prism 5.0; GraphPad Software, La Jolla, California, USA). Data distribution was analyzed by the Kolmogorov-Smirnov test. For normally distributed variables (HR, RR, SAP, MAP, and DAP), comparisons between treatments were analyzed by a 2-way repeated measures analysis of variance (ANOVA) with time and treatment as factors. When a significant difference between treatments was iden-tified, a Bonferroni correction for multiple comparisons was used to determine what treatments differed. A 1-way repeated measures

Figure 1. Distribution of sedation scores subjectively assessed by use of the simple descriptive scale (SDS) in 6 dogs at 25 min after 0.05 mg/kg body weight (BW) acepromazine (time point ACP) and at 20, 40, and 60 min after administration of physiological saline (Control), 3 mg/kg BW tramadol (TR3) or 5 mg/kg BW tramadol (TR5). All drugs were administered intravenously. The SDS ranged from 0 to 3 where 0 — no sedation; 1 — mild sedation; 2 — moderate sedation; and 3 — intense sedation.



ANOVA was performed to detect differences over time within each treatment. If a significant difference was detected, a Dunnett’s test for multiple comparisons was used to compare all time points with time point ACP. For comparisons between treatments and over time in sedation scores, a Friedman test was performed and post hoc analysis was conducted by use of the Dunn’s test for multiple comparisons. Differences were considered significant if P , 0.05.

Re s u l t sAll 6 dogs completed the study. Acepromazine administration

resulted in mild sedation in most dogs (Table I, Figure 1). Heart rate did not change significantly after acepromazine but a decrease was observed in SAP, MAP, DAP, and RR in all treatments at time point ACP (Table II).

Administration of the experimental treatment (time points 20 to 80 min) did not significantly increase SDS or NRS scores compared to time point ACP. The only significant difference between treatments in sedation scores was observed at 20 min, when NRS was higher in the TR3 treatment compared to the control treatment. At 80 min, NRS scores decreased in all treatment groups compared to the values at time point ACP (Table I). The distribution of SDS scores in each treatment is summarized in Figure 1.

Blood pressure and RR did not differ among treatments through-out the study. In addition, SAP, MAP, DAP, and RR did not change significantly after administration of the experimental treatment. The only exception to this rule was observed at time point 40, when DAP was significantly higher in TR5 compared to the value at time point ACP. Administration of tramadol (TR3 and TR5) resulted in significant decreases in HR from 20 to 80 min and the value for the TR3 treatment was significantly lower compared to the control treat-ment at 40 min (Table II).

D i s c u s s i o nThe present study revealed that combining TR3 or TR5 with

0.05 mg/kg BW acepromazine failed to enhance the degree of sedation induced by acepromazine. Mild to moderate sedation was observed after administration of acepromazine alone or the combina-tion acepromazine-tramadol.

On the basis of SDS scores, the findings of this study are in agreement with previous studies that, in dogs, sedation following acepromazine alone ranges from mild to moderate (3,13). After intramuscular administration of acepromazine in dogs, peak seda-tive effect appears to occur within 15 to 30 min (2,4). In the present study, SDS and NRS scores assessed from 20 to 80 min in the control

Table II. Mean 6 SD heart rate (HR), systolic, mean and diastolic blood pressures (SAP, MAP, and DAP), and respiratory rate (RR) in 6 dogs before administration of any drug (baseline, BL), at 25 min after administration of acepromazine (time point ACP), 0.05 mg/kg body weight (BW), IV and at 20, 40, 60, and 80 min after administration of physiological saline (control) IV, 3 mg/kg BW tramadol (TR3) or 5 mg/kg BW tramadol (TR5), IV

BL ACP 20 40 60 80HR (beats/min) Control 97 6 12 93 6 20 90 6 21 86 6 16 77 6 11a 81 6 10 TR3 85 6 10 94 6 12 76 6 12a 66 6 8a,b 71 6 13a 69 6 9a

TR5 97 6 11 97 6 22 74 6 6a 74 6 9a 76 6 13a 73 6 8a

SAP (mmHg) Control 127 6 17a 106 6 11 107 6 20 109 6 20 109 6 14 112 6 14 TR3 130 6 13a 110 6 13 109 6 14 106 6 10 109 6 10 107 6 9 TR5 136 6 16a 110 6 16 111 6 14 113 6 12 112 6 15 111 6 8

MAP (mmHg) Control 88 6 12a 73 6 7 75 6 11 78 6 13 76 6 13 79 6 9 TR3 88 6 12a 76 6 7 78 6 11 75 6 8 78 6 9 77 6 5 TR5 94 6 11a 77 6 9 81 6 8 83 6 11 79 6 12 80 6 8

DAP (mmHg) Control 67 6 11a 55 6 5 59 6 8 61 6 11 60 6 12 62 6 7 TR3 65 6 10a 58 6 6 61 6 10 59 6 9 62 6 7 61 6 4 TR5 72 6 7a 58 6 5 65 6 5 67 6 9a 61 6 10 65 6 8

RR (breaths/min) Control 33 6 9a 21 6 8 17 6 5 19 6 5 19 6 5 19 6 4 TR3 43 6 20a 23 6 8 18 6 6 16 6 3 17 6 3 17 6 3 TR5 35 6 10a 22 6 6 17 6 2 17 6 4 18 6 6 18 6 7a Significant difference compared to time point ACP.b Significant difference compared to the control treatment (P , 0.05).



treatment were not significantly greater than at time point ACP (approximately 25 min after acepromazine administration). These results suggest that peak sedative effect after IV acepromazine occurs within 25 min after administration of the drug. If sedation was assessed more frequently in this study, it would be possible to determine more precisely the period of time between IV aceproma-zine administration and peak sedative effect.

When an opioid analgesic is administered in combination with acepromazine, sedation is expected to be greater than after acepromazine alone, but there is conflicting evidence. In one study, intramuscular acepromazine-methadone (0.05 mg/kg BW and 0.5 mg/kg BW, respectively) resulted in non-significantly increase in sedation score compared to 0.1 mg/kg BW acepromazine (2). In another study, sedation induced by acepromazine (0.05 mg/kg BW) was significantly improved when morphine (0.5 mg/kg BW), metha-done (0.5 mg/kg BW) or butorphanol (0.2 mg/kg BW) were admin-istered 15 min after acepromazine via IV injection (3). Conversely, the combination of acepromazine (0.05 mg/kg BW) with hydro-morphone (0.1 mg/kg BW) failed to induce greater sedation than acepromazine alone (4). The discrepancies between studies are prob-ably related to differences on the scoring systems used to subjectively assess sedation, routes of drug administration and pharmacological characteristics of opioid analgesics used.

In a previous study in dogs, sedation assessed after combining acepromazine (0.05 mg/kg BW) with tramadol (2 mg/kg BW) was only mildly increased and for a short period (15 min), compared to sedation assessed after acepromazine alone (3). In the present study, it was hypothesized that higher doses of tramadol (TR3 or TR5), combined with acepromazine, would result in consistently greater sedation scores than acepromazine alone. No significant differences were observed in SDS scores between the control treatment and the TR3 and TR5 treatments. In addition, no significant difference in NRS scores were detected between the control and TR5 treatment and NRS scores in the TR3 treatment were significantly higher than in the control treatment at a single time point (20 min) only. These findings indicate there is no clinically important enhancement in sedation induced by acepromazine when doses of tramadol as high as TR5 are combined with phenothiazine.

The inefficacy of tramadol in enhancing the degree of sedation induced by acepromazine in dogs may be related to its pharmaco-kinetic and pharmacodynamic characteristics. Sedation following opioid administration is attributable to agonistic action at opioid receptors at brain sites (5). In cloned human opioid receptors, a greater affinity for opioid receptors was found for the M1 metabolite than for the parent drug tramadol (7). It has been reported that dogs can only produce low levels of M1 and that this metabolite has a fast elimination phase in this species (9). Therefore, sedation may be impaired in dogs because of its inability to produce significant amounts of M1. This statement is supported by results of a previous study where administration of tramadol orally (11 mg/kg BW) or by IV (4.4 mg/kg BW), did not result in detectable signs of sedation, whereas administration of M1 resulted in mild sedation in dogs (10).

Acepromazine administration did not change HR in the present study. Heart rate measurements from 20 to 80 min in the TR3 and TR5 groups were typical of opioid analgesics in dogs. A sustained decrease of 25% to 30% in HR was observed in both groups com-

pared with time point ACP. The negative chronotropic effect of opi-oid analgesics is related to their actions within the medulla oblongata increasing vagal tone (14). Although the effect of opioids on HR is dose-related, there is a plateau after which further increases in dose does not result in more bradycardia (15). In agreement with this previous study, our findings revealed the dose of tramadol did not influence the magnitude of the decrease in HR.

A decrease in SAP, MAP, and DAP was observed after aceproma-zine administration in all treatment groups. The effect of aceproma-zine on blood pressure is thought to result from antagonistic action at alpha-adrenergic receptors within vascular beds resulting in vasodilation (16). Moreover, in one study in dogs, acepromazine administration resulted in a decrease in stroke volume, which could also reduce blood pressure via a decrease in cardiac output (17). Tramadol administration did not have any influence on blood pres-sure such that SAP, MAP, and DAP values from 20 to 80 min were not significantly lower than at time point ACP.

Both acepromazine and opioids can reduce RR in dogs (15,17). The respiratory effects of acepromazine are associated with its sedative effect and relieving of anxiety (17). Despite causing a reduc-tion in RR, acepromazine did not change blood gases in dogs (17). Administration of opioid analgesics such as fentanyl, result in dose-related respiratory depression in dogs that are awake, as represented by decreased pH and PaO2 and increased PaCO2 (15). Different from other opioid analgesics with high affinity for m receptors, tramadol was found to cause less respiratory depression than morphine in humans (18) and in dogs, there was no change in RR after IV injec-tion of tramadol at doses as high as 4 mg/kg BW (9). Finding that RR decreased below baseline in all treatment groups at time point ACP and did not change any further after administration of tramadol suggests that acepromazine was the main drug responsible for the decrease in RR in this study.

An indirect rather than a direct method was used for measuring blood pressure, which is a limitation of this study. Oscillometric devices have been used in dogs (19–21), and 1 study reported an acceptable accuracy (21). However, agreement with invasive blood pressure was poor in 2 other studies (19,20). Despite this limitation, in the present study it was aimed to assess the behavior of blood pressure over time and not to report absolute values.

Under the conditions of this study, sedation induced by aceproma-zine and tramadol was similar to that of acepromazine alone. The main adverse effects of the combination were a decrease in blood pressure and HR, without clinical significance.

Re f e r e n c e s1. Rankin DC. Sedatives and tranquilizers. In: Grimm KA, Lamont

LA, Tranquilli WJ, Greene SA, Robertson SA, eds. Veterinary Anesthesia and Analgesia. Ames, Iowa: Wiley Blackwell, 2015: 196–206.

2. Monteiro ER, Figueroa CDN, Choma JC, Campagnol D, Bettini CM. Effects of methadone, alone or in combination with acepromazine or xylazine, on sedation and physiologic values in dogs. Vet Anaesth Analg 2008;35:519–527.

3. Monteiro ER, Junior AR, Assis HMQ, Campagnol D, Quitzan JG. Comparative study on the sedative effects of morphine,



methadone, butorphanol or tramadol, in combination with acepromazine, in dogs. Vet Anaesth Analg 2009;36:25–33.

4. Hofmeister EH, Chandler MJ, Read MR. Effects of acepromazine, hydromorphone, or an acepromazine-hydromorphone combina-tion on the degree of sedation in clinically normal dogs. J Am Vet Med Assoc 2010;237:1155–1159.

5. KuKanich B, Wiese AJ. Opioids. In: Grimm KA, Lamont LA, Tranquilli WJ, Greene SA, Robertson SA, eds. Veterinary Anesthesia and Analgesia. Ames, Iowa: Wiley Blackwell, 2015: 207–226.

6. Poulsen L, Arendt-Nielsen L, Brøsen K, Sindrup SH. The hypo-algesic effect of tramadol in relation to CYP2D6. Clin Pharmacol Ther 1996;60:636–644.

7. Lai J, Ma SW, Porreca F, Raffa RB. Tramadol, M1 metabolite and enantiomer affinities for cloned human opioid receptors expressed in transfected HN9.10 neuroblastoma cells. Eur J Pharmacol 1996;316:369–372.

8. Vazzana M, Andreani T, Fangueiro J, et al. Tramadol hydrochlo-ride: Pharmacokinetics, pharmacodynamics, adverse side effects, co-administration of drugs and new drug delivery systems. Biomed Pharmacother 2015;70:234–238.

9. McMillan CJ, Livingston A, Clark CR, et al. Pharmacokinetics of intravenous tramadol in dogs. Can J Vet Res 2008;72:325–331.

10. KuKanich B, Papich MG. Pharmacokinetics of tramadol and the metabolite O-desmethyltramadol in dogs. J Vet Pharmacol Ther 2004;27:239–246.

11. Valverde A, Cantwell S, Hernández J, Brotherson C. Effects of acepromazine on the incidence of vomiting associated with opioid administration in dogs. Vet Anaesth Analg 2004;31:40–45.

12. Monteiro ER, Nunes-Junior JS, Bressan TF. Randomized clini-cal trial of the effects of a combination of acepromazine with morphine and midazolam on sedation, cardiovascular variables and the propofol dose requirements for induction of anesthesia in dogs. Vet J 2014;200:157–161.

13. Gomes VH, Monteiro ER, Dias RS, de Oliveira RLS, da Silva MFA, Coelho K. Comparison of the sedative effects of morphine, meperidine or fentanyl, in combination with acepromazine, in dogs. Ciência Rural 2011;41:1411–1416.

14. Laubie M, Schmitt H, Vincent M. Vagal bradycardia produced by microinjections of morphine-like drugs into the nucleus ambig-uus in anaesthetized dogs. Eur J Pharmacol 1979;59:287–291.

15. Arndt JO, Mikat M, Parasher C. Fentanyl’s analgesic, respira-tory, and cardiovascular actions in relation to dose and plasma concentration in unanesthetized dogs. Anesthesiology 1984;61: 355–361.

16. Ludders JW, Reitan JA, Martucci R, Fung DL, Steffey EP. Blood pressure response to phenylephrine infusion in halothane-anesthetized dogs given acetylpromazine maleate. Am J Vet Res 1983;44:996–999.

17. Stepien RL, Bonagura JD, Bednarski RM, Muir WW, 3rd. Car-diorespiratory effects of acepromazine maleate and buprenor-phine hydrochloride in clinically normal dogs. Am J Vet Res 1995;56:78–84.

18. Vickers MD, O’Flaherty D, Szekely SM, Read M, Yoshizumi J. Tramadol: Pain relief by an opioid without depression of respira-tion. Anaesthesia 1992;47:291–296.

19. Shih A, Robertson S, Vigani A, da Cunha A, Pablo L, Bandt C. Evaluation of an indirect oscillometric blood pressure monitor in normotensive and hypotensive anesthetized dogs. J Vet Emerg Crit Care (San Antonio) 2010;20:313–318.

20. Acierno MJ, Fauth E, Mitchell MA, da Cunha A. Measuring the level of agreement between directly measured blood pres-sure and pressure readings obtained with a veterinary-specific oscillometric unit in anesthetized dogs. J Vet Emerg Crit Care (San Antonio) 2013;23:37–40.

21. Vachon C, Belanger MC, Burns PM. Evaluation of oscillometric and Doppler ultrasonic devices for blood pressure measurements in anesthetized and conscious dogs. Res Vet Sci 2014;97:111–117.


Article


I n t r o d u c t i o nHorses and donkeys commonly suffer from ocular injuries, a con-

sequence of the anatomical size and positioning of the eyes accompa-nied by their inborn “fight or flight” response, which creates a need for further investigation of topical ophthalmic anesthetics in order to select the most effective medication for each case. The cornea is the most sensitive tissue in the body (1). Its sensitivity has been

documented by use of an esthesiometer (2,3). The Cochet-Bonnet corneal esthesiometer (CBA) has become the most commonly used esthesiometer in clinical and research settings (1,4). The filament of the CBA becomes increasingly more rigid as it is shortened and exerts an increased pressure on the corneal surface. The corneal touch threshold (CTT) is the minimal amount of corneal stimulation resulting in a consistent blink reflex and is measured by the use of a corneal esthesiometer (5,6). Corneal sensitivity has been measured

Degree of corneal anesthesia after topical application of 0.4% oxybuprocaine ophthalmic solution in normal equids

Erika Little, Kathy Yvorchuk-St. Jean, William Little, Fortune Sithole, Guy St. Jean

A b s t r a c tOxybuprocaine hydrochloride ophthalmic solution has been widely used off-label in horses and donkeys, despite lack of data demonstrating efficacy and safety in these species. The objective of this study was to assess anesthetic efficacy of 0.4% oxybuprocaine hydrochloride ophthalmic solution in horses (n = 5) and donkeys (n = 24) and compare the effects with 0.5% proparacaine hydrochloride ophthalmic solution. The baseline corneal touch threshold (CTT) was measured with a Cochet-Bonnet esthesiometer. Donkeys (n = 12) and horses (n = 5) in group A received sterile ophthalmic solutions 0.4% oxybuprocaine with fluorescein (also termed benoxinate with fluorescein, abbreviated as ben 1 flu) instilled in one eye and 0.9% sterile sodium chloride solution (NaCl) with fluorescein (Na 1 flu) in the contralateral eye. Donkeys (n = 12) and horses (n = 5) in group B received sterile ophthalmic solutions (ben 1 flu) in one eye and 0.5% proparacaine with fluorescein (prop 1 flu) in the contralateral eye. The CTT was measured at 1 and 5 min post-application and at 5-minute intervals until 75 min after treatment. The CTT changes over time differed significantly between oxybuprocaine-treated and control eyes (P , 0.001). The CTT continued to decrease throughout the duration of the study when compared with baseline values. No statistically significant difference in onset, depth, or duration of corneal anesthesia was found between oxybuprocaine and proparacaine treated eyes during the time of the study. Interestingly, horses were shown to have a significantly more sensitive cornea than donkeys (P = 0.002). Oxybuprocaine and proparacaine reduced corneal sensitivity in donkeys and horses. No local irritation was observed with 0.4% oxybuprocaine.

R é s u m éLa solution ophtalmique d’hydrochlorure d’oxybuprocaïne a été utilisée extensivement en dérogation chez les chevaux et les ânes, malgré le manque de données démontrant son efficacité et son innocuité chez ces espèces. L’objectif de la présente étude était d’évaluer l’efficacité anesthétique d’une solution ophtalmique d’hydrochlorure d’oxybuprocaïne 0,4 % chez des chevaux (n = 5) et des ânes (n = 24) et comparer les effets avec une solution ophtalmique d’hydochlorure de proparacaïne 0,5 %. La valeur de base du seuil de contact cornéen (SCT) a été mesurée à l’aide d’un esthésiomètre Cochet-Bonnet. Les ânes (n = 12) et chevaux (n = 5) du groupe A ont reçu une solution ophtalmique stérile d’oxybuprocaïne 0,4 % avec de la fluorescéine (également appelée benoxinate avec fluorescéine, abrévié ben 1 flu) dans un œil et une solution stérile de chlorure de sodium 0,9 % (NaCl) avec de la fluorescéine (Na 1 flu) dans l’œil contra-latéral. Les ânes (n = 12) et chevaux (n = 5) du groupe B ont reçu les solutions ophtalmiques stériles de (ben 1 flu) dans un œil et de la propacaïne 0,5 % avec de la fluorescéine (prop 1 flu) dans l’œil contra-latéral. Le SCT a été mesuré à 1 et 5 min post-application et à des intervalles de 5 min jusqu’à 75 min après le traitement. Les changements dans le temps du SCT différaient de manière significative entre les yeux traités à l’oxybuprocaïne et les témoins (P , 0,001). Le SCT continua de diminuer tout au long de la durée de l’étude lorsque comparé aux valeurs de base. Aucune différence significative dans le début, la profondeur, ou la durée de l’anesthésie cornéenne ne fut trouvée entre les yeux traités à l’oxybuprocaïne et la proparacaïne durant la durée de l’étude. De manière intéressante, les chevaux avaient une cornée significativement plus sensible que les ânes (P = 0,002). L’oxybuprocaïne et la proparacaïne ont réduit la sensibilité cornéenne chez les ânes et les chevaux. Aucune irritation locale ne fut observée avec l’oxybuprocaïne 0,4 %.


Clinical Sciences, Ross University School of Veterinary Medicine, Saint Kitts and Nevis, West Indies NJ 08830.

Address all correspondence to Dr. Erika Little; telephone: 732-898-0185; e-mail: [email protected]

Received June 8, 2016. Accepted June 28, 2016.



using an esthesiometer in horses (7–9) as well as many other species (9); however, no studies measuring corneal sensitivity were found for donkeys. In previous studies mean corneal sensitivity measured by CTT using a CBA in healthy adult horses has been reported to range from 2.12 1 0.62 cm to 5.01 1 0.61 cm (6–8,10–12). The central portion of the equine cornea is the most sensitive as determined by previous studies (7,8).

Commonly used topical ophthalmic anesthetics include: 0.5% proparacaine, 0.5% tetracaine, and 0.4% oxybuprocaine (benoxinate); proparacaine and tetracaine solutions being the most widely used in veterinary ophthalmology (4). Ocular sensitivity manifested by symptoms of acute conjunctival hyperemia, chemosis, and nictitans protrusion has been reported in dogs following topical application of tetracaine (4,13). In dogs, 0.4% oxybuprocaine hydrochloride (benoxinate) is appropriately suited as a topical ocular anesthetic agent that can be used in many clinical settings with few risks of conjunctival changes compared to tetracaine solution (14,15). Its anesthetic effects are comparable in depth and duration to that of proparacaine solution (16,17). In the United States, proparacaine solution is a common topical anesthetic solution used in veterinary practice (4,18). In the dog, studies have shown that significant anesthetic effects on the cornea can be measured for 45 min after the instillation of proparacaine (15) as well as oxybuprocaine solutions (14). The use of proracaine has been evaluated in many veterinary species and has also been proven to be a safe and effective corneal anesthetic agent in cats (16), dogs (9), and horses (6,11).

The rationale for this study was to determine the degree of cor-neal anesthesia provided by 0.4% oxybuprocaine with fluorescein (ben 1 flu) ophthalmic solution compared to 0.5% proparacaine with fluorescein (prop 1 flu) solution in donkeys and horses in addition to evaluating ben 1 flu for adverse effects in horses and donkeys. This study also allowed for comparison of response of ben 1 flu solution between donkeys and horses. The hypothesis was that a single topical application of ben 1 flu or prop 1 flu solution would provide similar anesthetic depth, but ben 1 flu would elicit longer anesthetic duration compared to 0.5% proparacaine solution in clinically normal donkeys and horses. It was anticipated that both donkeys and horses would respond similarly to the topical applica-tion of ben 1 flu solution.


AnimalsTwenty-four healthy adult castrated male Abyssinian donkeys

ranging from 4 to 12 y of age (mean: 6.8 y, SD 1\2 2.47 y) and 5 healthy adult horses (3 geldings and 2 mares) ranging from 10 to 32 y in age (mean:L 21.6, SD 1\2 8.1 y) were used in the study. Horses were mixed breed Quarter Horse and Thoroughbred.

Donkeys were randomly divided into 2 groups of 12 for the study. Each horse was used in both group A and group B with a 2-week washout period between treatments. All animals were university-owned and were housed on-pasture throughout the duration of the study. Animals were brought into the same treatment room for the study days to ensure constant ambient temperature, relative humid-ity, and elimination of environmental artifacts. All animals were considered healthy based on physical examination findings. Only ophthalmologically normal donkeys and horses, demonstrating a Schirmer Tear Test (Schirmer tear test strips; Schering-Plough Animal Health, Union, New Jersey, USA) value . 10 mm/min, absence of corneal fluorescein stain (Bio-Glo fluorescein sodium ophthalmic strips; HUB Pharmaceuticals, Rancho Cucamonga, California, USA) retention, and absence of active corneal or adnexal disease were included in the study. Animals were monitored for adverse effects including blepharospasm, epiphora, and chemosis throughout the study and the investigators, upon completion of the procedure, per-formed fluorescein staining of the cornea. Blepharospasm, epiphora, and conjunctival hyperemia were graded on a scale from 0 to 3 for each of the criteria evaluated. Grade 0 was considered normal, grade 1 mildly affected, grade 2 moderately affected, and grade 3 severely affected.

ProceduresAll animals in the study were randomly allocated to treatment

order. For each animal the baseline CTT was determined in both eyes by use of a Cochet-Bonnet esthesiometer (Cochet-Bonnet esthesiom-eter; Luneau Ophtalmologie, Chartres, France) immediately prior to topical administration of solutions. All measurements were obtained with minimal restraint and without the use of sedation.

One investigator applied CBA to the cornea while another investi-gator visualized the application of the esthesiometer while wearing head loupes with a 43 magnification.

All solutions were drawn up in a sterile fashion using a new sterile 25-gauge needle and 1-mL syringe. A new sterile fluorescein strip was used to create the saline with fluorescein and the proparacaine with fluorescein. In order to administer the treatments, 0.2 mL of the selected drug solution was drawn into a 1-mL syringe via a 25-gauge needle. The needle was broken off at the hub and the drug solution was gently sprayed onto the corneal surface (6,11). The same bottle of each drug solution was used throughout the study period and stored according to the manufacturer’s guidelines. In group A, CTT results were compared after topical administration of 0.9% sterile sodium chloride solution (NaCl) with fluorescein (Na 1 flu) solu-tion in 1 eye and topical administration of ben 1 flu (Fluorescein sodium and benoxinate hydrochloride ophthalmic solution, USP 0.25%/0.4%; Bausch & Lomb, Tampa, Florida, USA) in the contra-lateral eye. Corneal touch threshold measurements were obtained immediately after baseline for Group A, 0.2 mL of ben 1 flu was

Table I. Esthesiometer conversion

Nylon length (mm) 60 55 50 45 40 35 30 25 20 15 10 5

Mean values of 0.4 0.5 0.55 0.7 0.8 1 1.4 1.8 2.8 5.1 10.3 15.9 pressures (mg/mm2)



administered in 1 eye and 0.2 mL of Na 1 flu was administered in the contralateral eye. Investigators were masked as to which solu-tion was being placed in which eye. In group B, CTT results were compared after a topical administration of ben 1 flu was instilled in one eye and a topical administration of prop 1 flu solution was instilled in the contralateral eye. For all animals, CTT was measured at time 0 (baseline), at 1 min, and 5 min, continuing every 5 min until 75 min after medications had been administered. Following data collection, the cornea was re-stained with fluorescein dye to ensure no corneal damage was sustained. The protocol used for this study was approved by the Institutional Animal Care and Use Committee of the Ross University School of Veterinary Medicine.

Measurement of corneal sensitivityThe CBA was used to measure the sensitivity of the central por-

tion of the cornea. It has a 0.12 mm diameter flexible monofilament nylon thread which has a range of 5 to 60 mm in length to control the intensity of corneal pressure applied by the monofilament. The length of the nylon filament corresponds to pressures of 11 to 200 mg/0.0113 mm2, respectively. The CBA is calibrated by the manufacturer and the length of exposed filament determines the pressure exerted on the cornea. The CTT is the mean filament length in mm at which a consistent blink response (3 out of 5 stimulations) is elicited. The CTT readings in mm can be converted to applied force

measurements using mg per S where S is 0.0113 mm2 of sectional area of the filament. The conversion of measurements made with the esthesiometer (Luneau SAS, Chartres Cedex, France) is shown in Table I. To determine corneal sensitivity, the filament of the esthesi-ometer was advanced toward the globe and applied perpendicular to the central portion of the cornea. The filament was pressed to the cor-nea until slight deflection of the filament was evident. The filament was initially applied at 60 mm (maximum length) and decreased in 5 mm increments until a consistent blink was elicited. Investigators wore 43 binocular loupes to improve accuracy of detection of fila-ment bending. The CTT, or corneal sensitivity, was recorded as the length of the esthesiometer filament that induced a blink reflex for at least 3 of 5 stimulations for a specified filament length (14). Two investigators observed each treatment and concurred with all results.

TreatmentsUsing sterile technique, fluorescein stain was added to the 0.5%

proparacaine solution (Proparacaine hydrochloride ophthalmic solu-tion; Akorn, Lake Forest, Illinois, USA) and to the 0.9% NaCl by a licensed veterinary technician to create a color consistent with that of the ben 1 flu solution. This was done to mask investigators as to which solution was placed in the eye. After obtaining the baseline CTT for each eye, ben 1 flu solution, prop 1 flu solution, or Na 1 flu was applied to the eye. For Group A, ben 1 flu was instilled in 1 eye

Figure 1. Donkeys: 0.4% oxybuprocaine versus 0.5% proparacaine.

Time (minutes)

CTT

(cm

)

Mean CTT Oxybuprocaine

Mean CTT Proparacaine

6

5.5

5

4.5

4

3.5

3

2.5

2

1.5

1

0.5

00 1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

Figure 2. Donkeys: 0.9% sodium chloride (NaCl) versus 0.4% oxybuprocaine.

Time (minutes)

CTT

(cm

)

Mean CTT saline eye

Mean CTT oxybuprocaine eye

6

5.5

5

4.5

4

3.5

3

2.5

2

1.5

1

0.5

00 1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

Figure 3. Horses: 0.4% oxybuprocaine versus 0.5% proparacaine.

Time (minutes)

CTT

(cm

)

Mean CTT Oxybuprocaine

Mean CTT Proparacaine

65.5

54.5

43.5

32.5

21.5

10.5

00 1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

Figure 4. Horses: 0.9% sodium chloride (NaCl) versus 0.4% oxybuprocaine.

Time (minutes)

CTT

(cm

)Mean CTT saline eye

Mean CTT oxybuprocaine eye

6

5.5

5

4.5

4

3.5

3

2.5

2

1.5

1

0.5

00 1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75



and Na 1 flu was instilled in the contralateral eye. For Group B, ben 1 flu was instilled in one eye and prop 1 flu solution was instilled in the contralateral eye. For each eye in all animals, the CTT value was determined at 1 min, at 5 min, and at 5-minute intervals until 75 min following administration of solution (Figures 1 to 4).

Evaluation of potential adverse effectsThroughout the duration of the study, animals were monitored for

evidence of blepharospasm, conjunctival hyperemia, and epiphora. Immediately after CTT measurements were obtained, all corneas were stained with fluorescein stain and evaluated with an oph-thalmoscope (Ophthalmoscope; Welch Allyn, Skaneateles Falls, New York, USA) to identify any injury to the cornea during the procedure.

Statistical analysisRepeated measures analysis of variance (ANOVA) models were

run to account for the clustering of CTT measurements of one eye over time for each of the treatments. Each ANOVA model (for each treatment comparison) had CTT as the outcome and time, treatment, and animal identification as factors. For each treatment in a treat-ment comparison group (in both horses and donkeys) a univariate ANOVA model was run to compare mean CTT measured at time 0 (baseline) and time 1 (1 min after treatment). A P , 0.05 was con-sidered significant.

Statistical software (Stata statistical software, Version 13; Stata Corp, College Station, Texas, USA) was used to perform the ANOVA analysis.

Re s u l t sThere was no evidence of conjunctival hyperemia, blepharospasm,

or epiphora noted at any time throughout the study (Grade 0 out of 3). All eyes were evaluated for corneal injury immediately after data collection using fluorescein staining and ophthalmoscope evaluation.

Minimal adverse ocular effects in this experimental protocol were noted in a single donkey. This animal developed a mild superficial corneal abrasion associated with fractious behavior resulting in the handle of the esthesiometer instrument abrading the cornea. The corneal abrasion was treated topically with an ophthalmic triple antibiotic ointment (Neomycin and Polymyxin B sulfates and baci-tracin zinc ophthalmic ointment; Akorn, Lake Forest, Illinois, USA) every 6 h after the procedure. No stain uptake was noted by 36 h post-abrasion, at which time treatment was discontinued. The CTT data from this animal were eliminated from statistical analysis due to the development of the corneal abnormality.

Oxybuprocaine (benoxinate) compared to salineOxybuprocaine (benoxinate) solution significantly decreased

CTT in both donkeys (P , 0.001) and horses (P , 0.001) versus the controls (Na 1 flu). Means 1/2 SD, with associated P-values, are included in Table I. Significant differences (P , 0.01) were deter-mined by ANOVA for mean CTT values between time 0 and 1 min, as well as all other time points post-administration of ben 1 flu, while no significant differences were noted between any of the post-

administration times. The ANOVA models in both species showed no significant difference in mean CTT between baseline and any time points, nor within post-administration times for the control groups (P = 0.184 donkeys and P = 0.854 horses). In this study, sev-eral donkeys and horses were noted to reach an individual CTT of 0 mm (no blink response at a filament length of 5.5 mm) for several min with ben 1 flu.

Oxybuprocaine (benoxinate) compared to Proparacaine

Compared with baseline values, CTT was significantly decreased (P , 0.01) by 1 min after treatment in the ben 1 flu and prop 1 flu treated eyes for both donkeys and horses. Donkeys showed no significant difference (P = 0.60) between ben 1 flu treated eyes and prop 1 flu treated eyes at any time point. There was a statistically significant difference (P = 0.001) between the 2 anesthetic solutions in horses. There were no significant differences in the CTT between donkeys and horses relating to ben 1 flu (P = 0.0983) or prop 1 flu (P = 0.2435). This study was unable to assess duration of medica-tions as no animal had returned to baseline values by the end of the evaluation period.

Donkeys compared to horsesHorses were shown to have a significantly more sensitive cornea

(P = 0.002) than donkeys at baseline (Mean CTT = 5.4 1/2 0.68 cm and 4.22 1/2 1.09 cm respectively). The CTT was not significantly different between donkeys and horses receiving ben 1 flu solution (P = 0.0983) or prop 1 flu solution (P = 0.2435).

Figures plotting the means over time for each group are included, and were generated using a commercial spreadsheet program (Excel; Microsoft Office, https://products.office.com/en-us/excel).

D i s c u s s i o nThis study assessed and compared efficacy of corneal anesthesia

after application of ben 1 flu and prop 1 flu ophthalmic solutions to normal eyes of horses and donkeys. To the author’s knowledge, this is the first study to evaluate the efficacy of topical ophthalmic administration of 0.4% oxybuprocaine (benoxinate) hydrochloride and compare efficacy between ben 1 flu and prop 1 flu solutions in the eyes of healthy donkeys and horses. In this study ben 1 flu solution significantly decreased CTT in donkeys (P , 0.001) and in horses (P = 0.0001) versus the controls. In donkeys, no significant difference was noted between ben 1 flu treated eyes and prop 1 flu treated eyes (P = 0.60); however in horses, a statistically significant difference was noted (P = 0.001). In a previous study, the duration of corneal anesthesia in horses using proparacaine solution was 25 min, with maximal corneal desensitization achieved by 5 min (11). In the current study, several donkeys and horses were noted to reach an individual CTT of 0 mm by 1 min post-administration, remaining for several min with both ben 1 flu and prop 1 flu solutions. The reason for this finding is unknown but it is possible that behavioral attributes might partially account for the difference. Although unknown, it is possible that the horses and donkeys in this study might be considered more tolerant as they are handled and examined daily by faculty and students. The reason for the apparent difference



in corneal sensitivity between the donkey and horse is unknown. It is possible that a species difference in sensitivity exists. The authors were unable to identify a study examining the differences in sensitiv-ity between these 2 species; therefore, this presents an opportunity for further investigation.

Additional considerations include repeated administration of topical anesthetic as well as volume instilled in each treatment. In a previous study in horses, tetracaine hydrochloride was found to significantly increase the depth and length of corneal anesthesia when administered twice (1 min between treatments) (12). A similar experiment in dogs proved that 1 drop of 0.5% proparacaine hydro-chloride ophthalmic solution applied twice with 1 min between doses resulted in significantly increased depth and duration of cor-neal anesthesia compared with a single drop dosage (15). Although there was a significant reduction in corneal sensation with a single application of ben 1 flu solution, further studies to evaluate whether or not more complete corneal anesthesia can be achieved with addi-tional applications of the solution (i.e., a second topical administra-tion 1 min after initial treatment) are warranted.

The impact of volume instilled within the eye was not evaluated in this study. The volume instilled, which is approximately equal to 4 drops of solution, is greater than the volume used in studies evaluating the onset and duration of proparacaine solution in cats and dogs (11,15). However, this volume represents the amount typi-cally instilled in clinical application in horses (11).

The Cochet-Bonnet esthesiometer is reliable for measuring cor-neal sensitivity in normal eyes of horses (1,4), and with exception of one donkey whose behavior created mild injury, resulted in no significant ophthalmic trauma in this study. The application of the Cochet-Bonnet esthesiometer is relatively simple and safe to perform. Limitations of the esthesiometer include subjective interpretation of the response, variations in technique as well as a lack of standardiza-tion of deflection pressure (1,6). Variations in humidity and ambient temperature can also alter results by affecting the rigidity of the esthesiometer’s filament (1). Corneal touch threshold measurement was agreed upon by 2 investigators as they concurrently observed the animals’ responses, requiring the animal to blink 3 times out of 5 consecutive applications of the Cochet-Bonnet esthesiometer. This procedure was used in attempt to limit the subjective evaluation of the positive response in the CTT measurements. Standardization of environmental factors was achieved by having all study subjects assessed in the same air-conditioned treatment room. One limitation of this study is that CTT measurements ended prior to all animals returning to baseline values. This precluded comparison of dura-tion of effect between proparacaine and oxybuprocaine. Further studies conducted until all animals return to baseline CTT values are necessary for comparison of these medications as well as to definitively determine the duration of topical anesthesia provided by the ben 1 flu solution.

Though the use of the Cochet-Bonnet esthesiometer is a reliable technique for measuring corneal sensitivity in veterinary studies, alternatively, a non-contact corneal esthesiometer (NCCE) such as the modified Belmonte esthesiometer, could have been used to evaluate corneal sensitivity. These devices use controlled pulses of air of varying pressures to stimulate the cornea. The advantage of a NCCE over the CBA is that a large, continuous range of stimulus

intensities can be produced (19,20). While air-puff esthesiometers are commonly used in research and clinical evaluations in human medicine, the filament-based esthesiometer is routinely used in veterinary research. As few to no studies could be found for such evaluation in equids, for comparison basis, the authors of this study chose to use the filament-based esthesiometer.

This study assessed animals with a normal, healthy cornea and did not evaluate animals with corneal disease such as ulcerations. It has been shown that repeated applications of topical anesthetic agents might cause injury, local irritation, and corneal epithelial toxicity (4,17). These were not observed at any point during this study. Corneal ulceration and perforation may occur with repetitive or prolonged use and is therefore discouraged (4). Future studies need to be conducted to determine the efficacy and duration of effect of oxybuprocaine hydrochloride solution in donkeys and horses with corneal disease. Use of the medications and the application of the Cochet-Bonnet esthesiometer resulted in no long-term damage in healthy horse and donkey eyes for the length of this study. The medications used are combination products normally used in oph-thalmologic examinations. The techniques and medications described are relatively simple and safe to perform and administer.

In conclusion, the use of ben 1 flu solution is effective and would pose minimal risk to donkeys and horses with normal corneas.

A c k n o w l e d g m e n tFunded with an intramural grant from Ross University School of

Veterinary Medicine.

Re f e r e n c e s1. Millodot M. A review of research on the sensitivity of the cornea.

Ophthalmic and Physiological Optics 1984;4:305–318.2. Krein SR, Lindsey JC, Blaze CA, Wetmore LA. Evaluation of risk

factors, including fluconazole administration, for prolonged anesthetic recovery times in horses undergoing general anes-thesia for ocular surgery: 81 cases (2006–2013). J Am Vet Med Assoc 2014;244:577–581.

3. Braz LG, Módolo NS, do Nascimento P, Jr, et al. Perioperative cardiac arrest: A study of 53,718 anaesthetics over 9 yr from a Brazilian teaching hospital. Br J Anaesth 2006;96:569–575.

4. Herring IP. Mydriatics/cycloplegics, anesthetics, and tears substitutes and stimulators. In: Gelatt KN, ed. Veterinary Ophthalmology, 5th ed. Hoboken, New Jersey: Wiley-Blackwell, 2013:423–434.

5. Johnston GM, Eastment JK, Wood JLN, Taylor PM. The confi-dential enquiry into perioperative equine fatalities (CEPEF): Mortality results of phases 1 and 2. Vet Anaesth Analg 2002;29: 159–170.

6. Pucket JD, Allbaugh RA, Rankin AJ, Ou Z, Bello NM. Com-parison of efficacy and duration of effect on corneal sensitivity among anesthetic agents following ocular administration in clinically normal horses. Am J Vet Res 2013;74:459–464.

7. Brooks DE, Clark CK, Lester GD. Cochet-Bonnet aesthesiometer-determined corneal sensitivity in neonatal foals and adult horses. Vet Ophthalmol 2000;3:133–137.



8. Kaps S, Richter M, Spiess BM. Corneal esthesiometry in the healthy horse. Vet Ophthalmol 2003;6:151–155.

9. Wieser B, Tichy A, Nell B. Correlation between corneal sensitiv-ity and quantity of reflex tearing in cows, horses, goats, sheep, dogs, cats, rabbits, and guinea pigs. Vet Ophthalmol 2013;16: 251-262.

10. Millodot M. Objective measurement of corneal sensitivity. Acta Ophthalmol 1973;51:325–334.

11. Kalf KL, Utter ME, Wotman KL. Evaluation of duration of cor-neal anesthesia induced with ophthalmic 0.5% proparacaine hydrochloride by use of a Cochet-Bonnet aesthesiometer in clinically normal horses. Am J Vet Res 2008;69:1655–1658.

12. Monclin SJ, Farnir F, Grauwels M. Duration of corneal anaesthe-sia following multiple doses and two concentrations of tetracaine hydrochloride eyedrops on the normal equine cornea. Equine Vet J 2011;43:69–73.

13. Koch SA, Rubin LF. Ocular sensitivity of dogs to topical tetra-caine HCl. J Am Vet Med Assoc 1969;154:15–16.

14. Douet JY, Michel J, Regnier A. Degree and duration of corneal anesthesia after topical application of 0.4% oxybuprocaine

hydrochloride ophthalmic solution in ophthalmically normal dogs. Am J Vet Res 2013;74:1321–1326.

15. Herring IP, Bobofchak MA, Landry MP, Ward DL. Duration of effect and effect of multiple doses of topical ophthalmic 0.5% proparacaine hydrochloride in clinically normal dogs. Am J Vet Res 2005;66:77–80.

16. Binder DR, Herring IP. Duration of corneal anesthesia following topical administration of 0.5% proparacaine hydrochloride solu-tion in clinically normal cats. Am J Vet Res 2006;67:1780–1782.

17. McGee HT, Fraunfelder FW. Toxicities of topical ophthalmic anesthetics. Expert Opin Drug Saf 2007;6:637–640.

18. Stiles J, Krohne S, Rankin A, Chang M. The efficacy of 0.5% proparacaine stored at room temperature. Vet Ophthalmol 2001;4:205–207.

19. Murphy PJ, Patel S, Marshall J. A new non-contact corneal aes-thesiometer (NCCA). Ophthalmic Physiol Opt 1996;16:101–107.

20. Stapleton F, Tan ME, Papas EB, et al. Corneal and conjunctival sensitivity to air stimuli. Br J Ophthalmol 2004;88:1547–1551.


Produits et services Supplies and services

RENOUVELEZ VOTRE ABONNEMENT

DÈS MAINTENANT!

RENEW YOUR SUBSCRIPTION

NOW!

Four issues of the Canadian Journal of Veterinary Research for 2016 (online)

Quatre numéros de la Revue canadienne de recherche vétérinaire pour 2016

Subscription/Abonnement 2016Personal Subscription/ $165 $ (1 GST/HST as Abonnement personnel applicable; TPS/TVH en (Canada) vigueur)

Personal Subscription/ $185 $ US/É.-U. Abonnement personnel (Foreign/Étranger)

Institutional Subscription/ $215 $ (1 In Canada, GST/HST Abonnement institutionnel as applicable; Au Canada, TPS/TVH selon le cas)

Contact/Communiquez avec : Linda Chow Editorial Coordinator/Coordonnatrice de la rédactionCanadian Journal of Veterinary ResearchRevue canadienne de recherche vétérinaire339 rue Booth Street Ottawa, Ontario K1R 7K1Canada

Tel./Tél. : (613) 236-1162, x117Fax/Télécopieur : (613) 236-9681E-mail/Courriel : [email protected]

w w w . c a n a d i a n v e t e r i n a r i a n s . n e t

w w w . v e t e r i n a i r e s a u c a n a d a . n e t

Tables of contentsComing events

Article abstractsClassified advertising

Subscription informationInstructions for authors

Tables des matières Événements à venirRésumés d’articles Annonces classées Information sur l’abonnement Directives à l’intention des auteurs

Canadian Journal of Veterinary Research

available online now!

Articles are now easily accessible over the Internet at

PubMed Centralwww .pubmedcentral .com

Reaching Canada’s VeterinariansGet your message into

The Canadian Veterinary JournalFor more information contact:Laima LaffitteAdvertising ManagerTel.: (613) 673-2659Fax: (613) 673-2462e-mail: [email protected]


Evaluation of hyaluronic acid, procollagen type III N-terminal peptide, and tissue inhibitor of matrix metalloproteinase-1 as serum markers of canine hepatic fibrosis Jonathan A . Lidbury, Aline Rodrigues Hoffmann, Joanna K . Fry, Jan S . Suchodolski, Jörg M . Steiner . . . . . . . . . . . . . . . . . . . . . . . . 302

Identification and characterization of pig adipose-derived progenitor cells Shuang Zhang, Chunyu Bai, Dong Zheng, Yuhua Gao, Yanan Fan, Lu Li, Weijun Guan, Yuehui Ma . . . . . . . . . . . . . . . . 309

Clinical staging in bitches with mammary tumors: Influence of type and histological grade Lígia F . Gundim, Camila P . de Araújo, William T . Blanca, Ednaldo C . Guimarães, Alessandra A . Medeiros . . . . . . . . . . . . . . . . . .318

Tramadol does not enhance sedation induced by acepromazine in dogsEduardo R . Monteiro, Renan B . Lobo, Juarez S . Nunes Jr, Julia P .P . Rangel, Flavia S . Bitti . . . . . . . . . . . . . . . . . . . . . . . . . .323

Degree of corneal anesthesia after topical application of 0.4% oxybuprocaine ophthalmic solution in normal equidsEricka Little, Kathy Yvorchuk-St . Jean, William Little, Fortune Sithole, Guy St . Jean . . . . . . . . . . . . . . . . . . . . . . . . . . 329


provided for non-commercial research and …...the canadian journal of veterinary research is...

Documents