FEMS Microbiology Letters 142 (1996) 191-194

High resolution DNA fingerprinting of Acinetobacter outbreak strains

Paul Janssen ‘, Lenie Dijkshoorn bj*

a Laboratorium voor Microbiologic, Universiteit Gent, B-9000 Gent, Belgium b Department of Medical Microbiology, L&den University Hospital, Building 1. L4-P, P. 0. Box 9600, 2300 RC Leiden. The NetherIan&

Received 27 June 1996; accepted 1 July 1996

Abstract

AFLP is a novel high resolution fingerprinting method that can be used to delineate intraspecific relationships among a

large variety of organisms, including bacteria. In the present study, this method was tested for its usefulness in the epidemiological typing of Acinetobucter strains. A total of 25 Acinetobacter strains originating from five hospital outbreaks in

three countries were used. Isolates from the same outbreak displayed identical banding patterns and each set of outbreak strains could be found in one particular AFLP cluster. These data are in good agreement with the results obtained by other

typing methods previously used on the same set of strains, indicating that AFLP analysis may be a valuable alternative in

epidemiological typing.

Keywords: Acinetobacter; AFL-; Genomic fingerprinting; Epidemiological typing

1. Introduction

Over the past few decades, strains of the genus

Acinetobacter have been reported to be increasingly associated with infection and epidemic spread in hos-

pital wards with severely ill patients [l]. The strains involved are frequently multiresistant to antibiotics

and can be difficult to eradicate [2,3]. Consequently, it is important to recognise early and to carefully monitor these strains in order to prevent or control their spread in hospitals. Various methods (reviewed

in [l]) have been used for the typing of acinetobac- ters from clinical specimens and outbreaks, including biotyping, phagetyping, antibiogram typing, protein

* Corresponding author. Tel.: +31 (71) 526 3931; Fax +31 (71) 524 8148.

electrophoretic profiling, ribotyping, plasmid typing, PCR-based DNA fingerprinting, and analysis of in-

frequently digested chromosomal DNA by pulsed-

field gel electrophoresis. However, genotypic and/or phenotypic differences between Acinetobacter strains

involved in outbreaks can be small and, with so many methods available, the value of each method for strain identification must be assessed systemati- cally. Criteria to be taken into account are reprodu-

cibility, discriminatory capacity, correlation with the epidemiology, and performance compared to other

typing methods. Other issues to be considered are rapidity, required skills, and overall costs.

AFLP@ is a novel high resolution genomic finger- printing method that can be used for a wide range of organisms. Essentially, it is a selective and highly standardized PCR-based method generating complex

0378-1097/96/$12.00 Copyright 0 1996 Federation of European Microbiological Societies. Published by Elsevier Science B.V.

PZZSO378-1097(96)00264-9

192 P. Janssen. L. DzjkshoornlFEMS Microbiology Letters 142 (1996) 191-194

but reproducible banding patterns [4] which can be

analysed by computer-assisted procedures. The ap-

plication of the AFLP method for the identification

and classification of bacteria, including members of

the genus Acinetobacter, has recently been reported

[5]. In addition, AFLP markers have been used to

distinguish outbreak from non-outbreak Acinetobac-

ter baumannii strains [6]. In the present study, we further tested AFLP@@ for

its ability to correlate DNA fingerprinting data with

the epidemiological origin of strains by using a set of 25 Acinetobacter outbreak strains collected from five

different hospitals in three countries. We have chosen

this particular set of strains because the resulting

AFLP data could then be verified by earlier pub-

lished typing data obtained on these strains during

a comparative evaluation of other typing methods,

including antibiogram typing, biotyping, cell enve- lope protein electrophoretic profiling, and ribotyping

171.

2. Materials and methods

2.1. Strains 3. Results and discussion

25 isolates from five different outbreaks were in-

vestigated. Each set of outbreak-associated strains

was previously assigned to a particular DNA group

(genomic species) based on DNA:DNA hybridisa- tion analysis [7]; isolates from three outbreaks, Ven-

lo (The Netherlands), Basildon (UK), and Newcastle (UK), belonged to A. baumannii (DNA group 2),

and isolates from the other two outbreaks, The Hague (The Netherlands) and Odense (Denmark), belonged to the unnamed DNA groups 3 and 13 sensu Tjernberg and Ursing [8]. All Acinetobacter

strains were grown aerobically on nutrient agar con- taining nutrient broth (Gibco BRL) supplemented

with 1.5% (w/v) Bacto agar (Difco). Details on the numerical designation of outbreak strains (Fig. 1) and epidemiological information on the outbreaks can be found in [7].

2.2. AFLP reactions

Of each strain, 1 ug of genomic DNA, prepared according to the method of Pitcher et al. [9], was

digested with 10 units of Hind111 and 10 units of

Tug1 restriction endonuclease. HindIII- and Tug1

halfsite-specific adaptors [5] were added to a final

concentration of 40 and 400 nM, respectively, and

ligated to the restriction fragments according to Vos

et al. [4]. The resulting template DNAs were precipi-

tated as described before [5]. Selective amplification

was achieved by the use of PCR primers HO1 and

TO5 [5]. Amplification procedures, PCR conditions, electrophoretic separation of PCR products, visuali-

zation of the banding patterns, and data acquisition have all been described in full detail [5]. Comparative

analysis of the AFLP patterns was facilitated by

using GelCompar@ software (Applied Maths, Kort-

rijk, Belgium). The unweighted-pair group method

using average linkages (UPGMA) [lo] was used to

cluster the AFLP patterns and a standard deviation

for each branch of the dendrogram was calculated.

These standard deviations are an indication for the significance and stability of the formed clusters and

are given as &values when discussing linkage levels

in the text.

Results of the cluster analysis depicted in a den-

drogram and the banding patterns of the isolates are

shown in Fig. 1. The AFLP pattern obtained on DNA of A. calcoaceticus type strain ATCC 23055T

was included at regular intervals in the gels and served as a reference track (not shown). Gels were

normalized and digital images combined by assigning

one particular reference track as a standard and alignment of all reference tracks versus this standard. At a cut-off value of 94% similarity, five clusters could be distinguished by AFLP@ and isolates with-

in each cluster clearly belonged to one particular outbreak, confirming the previously obtained typing

results [7]. AFLP profiles of the A. baumannii out- breaks from Venlo (The Netherlands) and Basildon (UK) were strikingly similar, with only two out of approx. 50 bands being different (as indicated in Fig. 1 by arrowheads), indicating that these strains may have derived from a common ancestral strain. The third outbreak of A. baumannii, Newcastle (UK), linked with the former two A. baumannii outbreaks at 67.0 + 1.2% similarity. The outbreaks from The

P. Janssen. L. DijkshoornlFEMS Microbiology Letters 142 (1996) 191-194

Vanlo (NL)

3242

E

I

Basildon (UK)

2 3371 1

. ,

E 3419

I

Odense (DK) 3417 3418 j ATCC 1944 1948 1947 1942 1945 I ATCC

17903’

Tl 18 H

lnU1

lague

3)

(W

193

Fig. 1. Digitized AFLP patterns of 25 Acinetobacter outbreak strains. Strain notation is as in Dijkshoom et al. [7]. The scale represents

the product-moment correlation coefficient (r) in % and the dendrogmm was constructed using the UPGMA clustering method [lo]. The

large arrow indicates the cut-off value of 94% similarity to define five clusters of outbreak strains. Small arrow heads indicate banding

pattern differences between the Venlo (The Netherlands) and Basildon (UK) outbreaks.

Hague (The Netherlands) and Odense (Denmark),

belonging to genomic species 3 and 13TU, respec- tively, linked with the three outbreaks of A. bauman-

nii at 24.4 +2.8%. In addition, Acinetobacter type

strains CCUG 19096T, ATCC 19004T, and ATCC 17903T of DNA groups 2, 3, and 13TU, respectively,

were included in the numerical analysis and linked

with the various outbreaks at an average similarity

of approx. 60% (60.8 f 3.2, 57.lkO.8, and 58.1 f 1.4%, respectively). These linkage levels are

in perfect concordance with the inter- and intraspe- ciflc linkage levels obtained by AFLP@’ on 20 Acine-

tobacter strains representing the species A.

calcoaceticus, A. baumannii, and the unnamed

DNA groups 3 and 13TU within the A. calcoaceti- cus-A. baumannii complex [ 1 I].

Phenotypic methods to differentiate epidemic Aci- netobacter strains have often been found to be inade- quate [7,12,13] and as a rule, phenotypic results

should always be verified by genotypical characteri-

zation. Various molecular techniques have been used successfully to differentiate members of the A. cal-

coaceticus-A. baumannii complex and for the typing

of Acinetobacter outbreak strains, including ribotyp-

ing [ 141, pulsed-field gel electrophoresis [15,16], and PCR fingerprinting [17-19). Compared to these

methods, AFLP is perhaps more laborious and re-

quires special equipment. Therefore, it may not be considered as the first method of choice. However,

advantages of AFLP in comparison to other DNA fingerprinting methods are that the patterns are

highly reproducible and the patterns of 50 or more

clearly separated bands are well suited for analysis using computer software packages. These advantages allow to compare large numbers of strains applied to different gels on different occasions. In addition, the use of fluorescent dye-labeled primers and analysis of the PCR products on an automated DNA sequencer

194 P. Janssen. L. DijkshoornlFEMS Microbiology Letters 142 (1996) 191-194

would considerably speed up the AFLP method and

would make it less labor intensive. Combined with

its high discriminatory capacity, AFLP can play an important role in epidemiology, in particular as an

elaborate method to confirm preliminary screening

results.

Acknowledgments

We thank Peter Gerner-Smidt and Tyrone L. Pitt

for providing strains, and Kees Maquelin for excel-

lent technical assistance. P.J. was supported by con-

tract G.O.A. 91/96-2 of the Ministerie van de

Vlaamse Gemeenschap, Bestuur Wetenschappelijk

Onderzoek (Belgium), and L.D. was supported by the Stichting Medische Microbiologic Leiden (The

Netherlands).

References

Sl

VI

[31

[41

[51

161

171

Bergogne-BBrtzin, E. and Towner, K.J. (1996) Acinetobacter spp. as nosocomial pathogens: microbial, clinical, and epide-

miological features. Clin. Microbial. Rev. 9, 148-165.

Crombach, W.H.J, Dijkshoom, L., Van Noort-Klaassen, M.,

Niessen, J. and Van Knippenberg-Gordebeke, G. (1989) Con-

trol of an epidemic spread of a multi-resistant strain of Aci- netobacter calcoaceticus in a hospital. Intensive Care Med. 15,

166170.

Bergogne-Ben&in. E. (1995) The increasing significance of

outbreaks of Acinetobacter spp.: the need for control and

new agents. J. Hosp. Infect. 30, 441452.

Vos, P., Hogers, R., Bleeker, M., Reijans, M., Van de Lee, T.,

Homes, M., Freijters, A., Pot, J., Peleman, J., Kuiper, M. and

Zabeau, M. (1995) AFLP: a new concept for DNA finger-

printing. Nucleic Acids Res. 23, 44074414.

Janssen, P., Coopman, R., Huys, G., Swings, J., Bleeker, M.,

Vos, P., Zabeau, M. and Kersters, K. (1996) Evaluation of the

DNA fingerprinting method AFLP as a new tool in bacterial

taxonomy. Microbiology 142, in press.

Dijkshoom, L., Aucken, H.M., Gerner-Smidt, P., Janssen, P.,

Kaufmann, M.E., Garaizar, J., Ursing, J. and Pitt, T.J. (1996)

Comparison of outbreak and nonoutbreak Acinetobacter bnu- mannii strains by genotypic and phenotypic methods. J. Clin.

Microbial. 34, 1519-1525.

Dijkshoom, L., Aucken, H.M., Gemer-Smidt, P., Kaufmann,

M.E., Ursing, J. and Pitt, T.L. (1993) Correlation of the typ-

ing methods for Acinetobacter isolates from hospital out-

breaks. J. Clin. Microbial. 31, 702-705.

[8] Tjernberg, I. and Ursing, J. (1989) Clinical strains of Acineto- batter classified by DNA-DNA hybridization. APMIS 97,

595605.

[9] Pitcher, D.G., Saunders, N.A. and Owen, R.J. (1989) Rapid

extraction of bacterial genomic DNA with guanidium thiocya-

nate. Lett. Appl. Microbial. 8, 151-156.

]lO] Sneath, P.H.A. and Sokal, R.R. (1973) The principles and

practice of numerical classification. Numerical Taxonomy,

W.H. Freeman, San Francisco, CA.

[ll] Dijkshoom, L. (1996) Acinetobacter Microbiology. In:

Acinetobacter (Bergogne-Btrezin,, E., Joly-Guillou, M.L. and

Towner, K.J., Eds.), p. 63, CRC Press, Boca Raton, FL.

[12] Kropec, A. Htibner, J. and Daschner, F.D. (1993) Compari-

son of three typing methods in hospital outbreaks of Acineto- batter calcoaceticus infection. J. Hosp. Infect. 23, 133-141.

[13] Seifert, H., Schulze, A., Baginsky, R. and Pulverer, G. (1994)

Comparison of four different typing methods for epidemiolo-

gical typing of Acinetobacter baumannii. J. Clin. Microbial. 32,

18161819.

[14] Gemer-Smidt, P. (1992) Ribotyping of the Acinetobacter cal- coaceticus-Acinetobacter baumannii complex. J. Clin. Micro-

biol. 30, 268&2685.

[15] Struelens, M.J., Carlier, E., Maes, N., Serruys, E., Quint,

W.G.V. and Van Belkum, A. (1993) Nosocomial colonization

and infection with multiresistant Acinetobacter bawnannii: outbreak delineation using DNA macrorestriction analysis

and PCR-fingerprinting. J. Hosp. Infect. 25, 15-32.

[16] Seifert, H. and Gerner-Smidt, P. (1995) Comparison of ribo-

typing and pulsed-field gel electrophoresis for molecular typ-

ing of Acinetobacter isolates, J. Clin. Microbial. 33, 1402-

1407.

[17] Graser, Y., Klare, I., Halle, E., Gantenberg, R., Buchholz, P.,

Jacobi, H.D., Presber, W. and Schonian, G. (1993) Epidemio-

logical study of an Acinetobacter buumunnii outbreak by using

polymerase chain reaction fingerprinting. J. Clin. Microbial.

3 1, 2417-2420.

[18] Reboli, A.C., Houston, E.D., Monteforte, J.S., Wood, CA.

and Hamill, R.J. (1994) Discrimination of epidemic and

sporadic isolates of Acinetobacter baumannii by repetitive ele-

ment PCR-mediated DNA-fingerprinting. J. Clin. Microbial.

32, 263552640.

[19] Snelling, A.M., Gemer-Smidt, P., Hawkey, P.M., Heritage, J.,

Pamell, P., Porter, C., Bodenham, A.R. and Inglis, T. (1996)

Validation of use of whole-cell repetitive extragenic palindro-

mic sequence-based PCR (REP-PCR) for typing strains be-

longing to the Acinetobacter calcoaceticus-Acinetobacter bau- mannii complex and application of the method to the

investigation of a hospital outbreak. J. Clin. Microbial. 34,

1193-1202.