Article available online at http://www.idealibrary.com on Microbial Pathogenesis 1999; 26: 111–119

PATHOGENESISMICROBIAL

Article No. mpat.1998.0257

Effect of lacto-N-neotetraose,asialoganglioside-GM1 and neuraminidase onadherence of otitis media-associatedserotypes of Streptococcus pneumoniae tochinchilla tracheal epitheliumH. H. Tong, M. A. McIver, L. M. Fisher & T. F. DeMaria∗

Division of Otologic Research, Department of Otolaryngology, College of Medicine,The Ohio State University, Columbus, Ohio, U.S.A.

(Received September 29, 1998; accepted in revised form November 19, 1998)

The adherence of Streptococcus pneumoniae (Spn) otitis media-associated serotypes 3, 6A and14 to ciliated chinchilla respiratory epithelium was investigated using a whole organ perfusiontechnique. We demonstrated that Spn adhere to chinchilla tracheal epithelium within 30 min andexhibit saturation kinetics indicating that the effect being observed is receptor mediated. Inhibitionof adherence was achieved by prior incubation of Spn with lacto-N-neotetraose (LNnT) or asialo-ganglioside GM1 (aGM1), recognized by glycoconjugate analogs of known Spn receptors. Neur-aminidase treatment of the tracheae increased Spn adherence in vitro and reversed the inhibitioneffect of LNnT suggesting that neuraminidase treatment resulted in an increase in the number ofavailable receptors for Spn. The chinchilla trachea organ perfusion culture system used in thisstudy imitates eustachian tube conditions more closely than isolated cell culture systems and is auseful model for investigating the role of Spn adherence in vitro in the pathogenesis of OM.

1999 Academic Press

Key words: Streptococcus pneumoniae, chinchilla, adherence, otitis media.

media (OM), pneumonia and other pneumo-Introductioncoccal infections, yet this process is poorly un-derstood. The regulation of adherence and the

It is widely believed that Streptococcus pneu- interrelationship between Spn virulence factorsmoniae (Spn) adherence to nasopharyngeal epi- that promote the carrier state, wherein Spn existsthelium is a prerequisite for induction of otitis for weeks in the upper airway without symp-

toms, has been the subject of numerous recentstudies. Spn are frequently isolated from the∗Corresponding author: Division of Otologic Research,nasopharynx of healthy individuals and in cer-The Ohio State University, Room 4331 UHC, 456 W. 10th

Ave., Columbus, Ohio 43210-1282, U.S.A. tain populations, nasopharyngeal carriage rates

0882–4010/99/020111+09 $30.00/0 1999 Academic Press

H. H. Tong et al.112

may exceed 70% [1]. Carriage of Spn is par-ticularly high in children of 1–2 years of ageand is related to predisposition to OM [2]. Themechanism whereby bacterial pathogens invadethe middle ear after colonizing the nasopharynxvia retrograde ascension of the eustachian tubehas not been elucidated.

A large number of studies have evaluatedadherence of Spn to various isolated cell typesor in cell culture systems in vitro. While thesereports provide important insights into the re-ceptors of Spn, cells harvested from tissues ofthe respiratory tract system are largely non-viable and are subject to day to day variability.We have previously utilized the chinchilla tra-chea organ perfusion culture technique to assessadherence of other OM pathogens [3]. The tra-chea is comprised of cuboidal as well as co-lumnar ciliated and non-ciliated cells and closely

9

6

15

Log CFU/ml

Log

CF

U a

dher

ent

bact

eria

/mm

tra

chea

5

3

2

7

4

6 8

approximates eustachian tube conditions [3].Figure 1. Saturation curve of adherence of Spn typeThe eustachian tube, however, is encased in bone 6A to chinchilla tracheal epithelium.and not amenable to organ culture. With the

trachea organ perfusion culture system, the bac-result is exposure of GlcNAcb1-4Gal, which isteria interact with a portion of a tubular organ,part of one of the Spn eukaryotic receptors firstwhich is organized, differentiated and containsdescribed by Andersson et al. on human pha-active ciliated epithelium almost identical to thatryngeal epithelial cells [9].which the organisms would encounter as part

In this study, we used trachea organ perfusionof the natural progression of OM pathogenesis.culture to evaluate the interaction of Spn typesTrachea organ perfusion culture affords an op-3, 6A and 14 major OM-associated serotypesportunity to evaluate adherence, adherence in-with chinchilla tracheal epithelium and invest-hibition and augmentation in a more controlled,igated the effect of lacto-N-neotetraose (LNnT)isolated fashion ex vivo, without the added vari-(Galb1-4GlcNAcb1-3Galb1-4Glc) and asialogan-able of the host’s inflammatory response.glioside GM1 (aGM1) (Galb1-3GalNAcb1-Bacterial factors that contribute to coloniza-4Galb1-4Glcb1-1Cer), two inhibitors of knowntion, the initial event in pneumococcal infection,Spn cell surface receptors on Spn adherence toare poorly defined. One Spn virulence factortrachea epithelium. Additionally, we invest-that potentially impacts colonization is neur-igated the ability of neuraminidase to enhanceaminidase, an enzyme that cleaves sialic acidadherence in vitro.from membrane glycoproteins. All clinical Spn

isolates examined to date have been shown toproduce neuraminidase [4] and a neuraminidase Resultsgene from Spn [NanA] has been cloned andsequenced [5]. Our laboratory and others have Adherence of Spn to chinchilla tracheal

epitheliumhypothesized that removal of sialic acid residuescould potentially impact significantly on thepathogenesis of Spn disease by exposing new With the culture conditions utilized, the

Spn inoculum typically achieved a density ofterminal carbohydrate residues that could serveas receptors for bacterial toxins, or adherence of 108 CFU/ml. Adherence assays were performed

with Spn at a concentration of 107 CFU/ml. Spnthe Spn cells [6]. Moreover, two previous reportsfrom our laboratory indicate that during Spn- adhered to chinchilla tracheal epithelium within

30 min of exposure. The saturation curve forinduced experimental OM in the chinchilla, ter-minal sialic acid residues are removed from the Spn 6A, whereby increasing the concentration

of bacteria, resulted in a maximum binding andsurface of the epithelium lining the lumen ofthe eustachian tube, presumably as a result of plateau of Spn adherence (Fig. 1). At a con-

centration of 107 CFU/ml, the trachea was halfthe production of neuraminidase [7, 8]. The

S. pneumoniae adherence to chinchilla tracheal epithelium 113

(P<0.05) of the control after preincubation withneuraminidase at a concentration of 0.05 units/ml and 0.1 units/ml, respectively. However, theeffect on adherence was diminished when thetracheae were exposed to a higher concentrationof neuraminidase (0.5 units/ml). Similarly, ad-herence of Spn 14 showed an increase of 203%of the control at a dose of 0.1 units/ml and adecrease to 100% of the control at 0.5 units/mlof neuraminidase. Adherence of Spn 3 increasedto 128%, 231% and 224% of the control at dosesof 0.05, 0.1 and 0.5 units/ml, respectively, andwas distinctive in that a decline in adherencewas not observed when this serotype was as-sayed with tracheas preincubated with 0.5 units/ml of neuraminidase. A 90 min preincubation ofthe tracheae with 0.5 units neuraminidase/mlwas required to reduce Spn adherence to 100%

0.6

700

0

0.0Neuraminidase (units/ml)

Adh

eren

ce (

% o

f co

ntr

ols)

0.5

200

600

500

400

300

100

0.1 0.2 0.3 0.4

*

of the control (data not shown).Figure 2. Adherence of Spn type 3, 6A and 14 tochinchilla tracheal epithelium pretreated with neur-aminidase. Data are expressed as percent of control The effect of lacto-N-tetraose (LNnT) onand represent a comparison of adherence from Spn adherence to chinchilla trachealtreated vs untreated trachea from the same animal. epitheliumEach data point represents the arithmetic mean andSEM derived from a minimum of three replicates.

Adherence to Galb1-4GlcNacb1-3Gal, one of the∗ Significant difference P<0.05. —Χ—, Spn 3; —Β—,receptors for Spn that has been described forSpn 6A; —Α—, Spn 14.human nasopharyngeal epithelial cells, is in-hibited by LNnT, a natural oligosaccharide [9].Spn pretreated for 30 min with LNnT (1–140 lM)demonstrated a 20–65% dose dependent de-saturated. The saturation curves for Spn 3 and

14 exhibited kinetics comparable to Spn 6A (data crease in adherence. However, the dose for themaximum inhibition varied among these Spnnot shown).OM strains. Adherence of Spn 6A decreased toa maximum of 35% of control (P<0.05) at adose of 14 lM (Fig. 3). A maximum decrease ofThe effect of neuraminidase on the

adherence of Spn to chinchilla tracheal adherence to 44% of control at 28 lM for Spn 14(Fig. 3) and 38% of control for type 3 at 106 lMepithelium(Fig. 3) were observed. Maximum inhibition ofSpn type 3 required 7.5-fold more LNnT thanTo continue our previous studies evaluating Spn

extracellular products which might play a role that required for optimal inhibition of Spn 6A.Pretreatment of Spn with 28 mM lactose, a sugarin modulating adherence, we examined the role

of neuraminidase in enhancing adherence to which is known not to be active in diminishingSpn adherence [10] did not decrease adherence.the tracheal epithelium. Sialic acid, in various

linkages, is the terminal residue in the eukaryotic Spn pretreated with lactose adherence to thechinchilla was at a level of 100% of control valuescell glycocalyx and is the substrate for neur-

aminidase. In preliminary experiments, 30 min (data not shown).Finally, the LNnT-induced inhibition of Spnwas established as the optimal incubation time

for all strains and was thus used for all sub- 6A adherence was no longer evident when thetracheae were pretreated with neuraminidase.sequent incubations. Our data demonstrated

that the adherence of Spn types 3, 6A and 14 all The mean adherence of LNnT-treated Spn toneuraminidase-treated organ culture was 97%increased to greater than 100% of control after

preincubation of the trachea with neuraminidase of control values compared to the 65% decreasein adherence of LNnT-treated Spn to non-treatedat a concentration of 0.1 units/ml (Fig. 2). Ad-

herence of Spn 6A increased to 155% and 306% tracheae (data not shown).

H. H. Tong et al.114

0.30

180

00.05

Asialoganglioside GM1 (mM)A

dher

ence

(%

of

con

trol

)0.25

60

160

120

100

80

20

0.15

140

40

0.10 0.20160

180

0Lacto-N-Tetraose (µM)

Adh

eren

ce (

% o

f co

ntr

ol)

140

60

160

120

100

80

20

20 60 80 100

*

140

40

40 120

*

Figure 4. Effect of preincubation of Spn type 6A withFigure 3. Effect of preincubation of Spn type 3, 6Aand 14 with lacto-N-neotetraose on adherence to asialoganglioside GM1 on adherence to chinchilla

tracheal epithelium. Data are expressed as percentchinchilla tracheal epithelium. Data are expressed aspercent of control and each data point represent a of control and each data point represents a com-

parison of treated vs untreated Spn 6A on adherence.comparison of treated vs untreated Spn on adherence.Each data point represents the arithmetic mean and Each data point represents the arithmetic mean and

SEM derived from a minimum of three replicates.SEM derived from a minimum of three replicates.For each assay, experimental and control tracheae For each assay, experimental and control tracheae

were derived from the same animal.were derived from the same animal. ∗ Significantdifference P<0.05. Key as for Fig. 1.

additive effect on the inhibition of Spn 6A ad-herence. Under these experimental conditions,LNnT (14 lM) or aGM1 (0.13 mM) alone showedEffect of asialoganglioside GM-1 (aGM1)a 30% and 35% inhibitory effect respectively (Fig.on Spn 6A adherence to chinchilla tracheal5). The combined treatment with both inhibitorsepitheliumproduced a 60% reduction in adherence (Fig. 5).

In comparable experiments, we found the at-tachment of Spn 6A to the chinchilla trachealepithelium was inhibited by prior incubation of Discussionthe Spn 6A with aGM1 for 30 min. AsialoGM1 isa known inhibitor of a second Spn carbohydratereceptor which contains the moiety GalNAcb1- A two-phase model describing pneumococcal

adherence has been proposed by Cundell et4Gal. The adhesion of Spn 6A decreased to 45%and 52% of the control at concentrations of al. [11]. Initially, Spn are believed to target an

anatomical niche, such as the nasopharynx, in0.13–0.25 mM of aGM1 (Fig. 4). The ten percentDMSO used to solubilize the aGM1 did not the host by binding to any of several surface

glycoconjugates, including Galb1-4GlcNAcb1-affect the viability of the Spn during the courseof these assays. 3Gal [12], GalNAcb1-4Gal [13] and GalNAcb1-

3Gal [11], which exist on normal, resting res-piratory epithelial or endothelial cells. The sec-ond phase of adherence, which is believed toThe combined effect of LNnT and aGM1

on Spn 6A adherence to chinchilla tracheal lead to invasion and disease, is induced bycytokine activation, and results in the de novoepitheliumexpression of new sugar specificities on the cy-tokine activated cells which results in a con-Incubation with LNnT, followed by a second

incubation with aGM1 for 30 min produced an comitant increase in Spn adherence. Activation

S. pneumoniae adherence to chinchilla tracheal epithelium 115

Neuraminidase is one of the potential Spnvirulence factors, examined as part of this study.Previous data suggest that neuraminidase mayplay a role in modulating adherence [17]. Pro-duction of neuraminidase is thought to con-tribute to the poor prognosis of pneumococcalmeningitis [18] and play a role in the patho-genesis of haemolytic-uremic syndrome [19].Spn neuraminidase has been detected in 78%of culture-positive human middle ear effusions(MEEs) from patients with acute OM and in 96%of Spn-positive MEEs from patients with chronicOME [20]. Moreover, two previous reports fromour laboratory indicate that during Spn-inducedexperimental OM in the chinchilla, terminalsialic acid residues are removed from the surfaceof the epithelium lining the lumen of the eus-tachian tube, presumably as a result of the pro-duction of neuraminidase [7, 8]. The result is

80

0LNnT

Spn6A treatment

Adh

eren

ce (

% o

f co

ntr

ol)

aGM1 LNnT + aGM1

50

70

60

40

30

20

10

exposure of GlcNAcB1-4Gal, which is part of oneFigure 5. Inhibition of adherence of Spn 6A chinchilla of the Spn eukaryotic receptors first describedtracheal epithelium preincubated with either lacto- by Andersson et al. [9]. In comparable studies,N-neotetraose or asialoganglioside GM1 alone or in LaMarco and co-workers demonstrated that thecombination. Data shown represents the mean±SEM direct inoculation of the chinchilla middle earfrom a minimum of three replicates. LNnT: Lacto-N- with 200 units of purified Spn neuraminidaseneotetraose; aGM1: asialoganglioside GM1.Ε, LNnT removed significant amounts of sialic acid from14 lM;Φ, aGM1 0.127 mM;∆, LNnT 14 lM+aGM1

middle ear mucosal epithelium [6]. These data0.127 mM.are consistent with a role for neuraminidase inthe pathogenesis of OM.

Data from the present study indicate thatof type II pneumocytes or vascular endothelialcells by IL-1 in vitro results in the expression of incubation of the trachea with neuraminidase

for 30 min significantly increases adherence to athe platelet-activating factor (PAF) receptors onthe cell surface and on increased adherence of maximum of 200% depending on the dose of

neuraminidase and the Spn strain being evalu-Spn [11].Spn adherence is a complex process and ad- ated. At the highest doses of neuraminidase the

effect was reversed which suggests excessivehesins or ligands that potentially contribute tothe regulation of adherence are actively being digestion by the enzyme resulting in a decline

of receptors. At 0.5 units/ml of neuraminidase,investigated. These include peptide permeases,choline binding proteins, opaque-transparent type 6A and 14 Spn exhibited decreased ad-

herence while Spn type 3 exhibited a decreasephenotypic variants, and the release of bacterialcell wall components which are responsible for adherence only after a 90 min incubation. The

reason for these differences is not clear butepithelial or endothelial cell activation [14, 15].Our data indicates that the chinchilla trachea presumably do not reflect the effect of type 3

polysaccharide capsule since previous data fromorgan culture model is ideally suited for screen-ing and evaluating adherence of Spn OM strains Andersson et al. indicate that adherence cap-

ability is independent of capsular type and thatin vitro. Our results show that all three Spnstrains adhere within 30 min of exposure to chin- both high and low adhering strains exist for all

Spn serotypes examined [2]. The reversal of thechilla tracheal epithelial tissues. The saturationcurve kinetics indicated that the effect being LNnT-induced inhibition of the adherence of

Spn 6A to the tracheae previously treated withobserved is receptor mediated and based onadherence to a specific, albeit limited, number neuraminidase suggests that neuraminidase ex-

poses additional receptors on the tracheal epi-of receptors as opposed to any number of non-specific mechanisms such as net surface charge, thelium for Spn. Their specificity can not be

determined from the present study. Spn 3 mayhydrophobic or lipophilic forces, and fibro-nectin-like cell adhesin molecules [16]. simply possess more ligands for the additional

H. H. Tong et al.116

receptors being exposed by the neuraminidase receptors are contributing to adherence in thismodel as well.treatment. These data indicate that the trachea

organ culture adherence model is also useful The use of oligosaccharides that block Spnadherence represents exciting new avenues forfor studying the effect of various enzymes and

exoproducts on adherence. the development of prevention as well as treat-ment modalities. A recent report by Idanpaan-The results from the present study also dem-

onstrate that the chinchilla trachea organ culture Heikkila et al. indicates that the administrationof LNnT intratracheally dramatically decreasedis amenable for studying the inhibition of ad-

herence with receptor analogues. Two different the Spn present in the lungs and conferred pro-tection against bacteraemia in rabbits. Moreover,glycolipids have been identified as receptors for

Spn adhesins. The proposed binding sites are this glycoconjugate when given therapeutically24 h after infection ameliorated pneumoniae andglycosphingolipids containing the disaccharides

GlcNAcb1-3Gal in one and GalNAcb1-4Gal in bacteraemia in the rabbit model [21].In conclusion, we have demonstrated that thethe other. The inhibitory effects observed on Spn

adherence to the chinchilla tracheal epithelium trachea organ culture adherence model is usefulfor studying the effect of various enzymes andby LNnT and aGM1 in the present study confirm

that these receptors are indeed present on the exoproducts on adherence of Spn OM isolates.Our data corroborates our previous findings thattracheal epithelium tissue and confirm pre-

viously published reports that these glyco- neuraminidase enhances the adherence of Spnto upper respiratory tract epithelium and thatconjugates inhibit the adherence of Spn to

individual desquamated human oropharyngeal adherence can be blocked with glycoconjugatesanalogues of Spn epithelial cell receptors. Theor nasopharyngeal cells [9], or lung and endo-

thelial cells in vitro [21–24]. Data obtained from maximum inhibitory effect of LNnT on Spnadherence to chinchilla trachea epitheliumthe trachea organ cultures are analogous to that

from previous reports with normal or ‘resting varied among different Spn serotypes examined.Studies are currently in progress in our in vivocells’. An attempt was not made in this study

to activate the trachea organ cultures with cyto- chinchilla model of OM to determine whetherSpn adherence can be blocked with various gly-kines, which results in a shift in pneumococcal

adherence to new receptors, particularly the PAF coconjugates so as to reduce colonization andsubsequent infection of the middle ear.receptor on cells in culture [25].

Our study showed that the inhibitory effect ofLNnT varied among the three OM Spn capsulartypes. The doses of LNnT for maximum in-

Materials and methodshibition of adherence were 14 lM, 28 lM and106 lM for types 6A, 14 and 3 respectively.

AnimalsAs mentioned earlier, previous reports indicatethere is no correlation between Spn capsule type

Healthy adult chinchillas (Chinchilla lanigera)and adherence capability and that poor as wellweighing 400–600 g were used as the source ofas high adherence capacity strains exist amongtracheae. The animals were killed with a lethalall the Spn capsular types tested [13]. Therefore,dose of xylazine and ketamine (Rompun, Haver-capsule type alone probably does not accountLockhart, Shawnee, KS: and Vetalar, Parke-for these differences. One possible explanationDavis, Morris Plains, NJ, respectively).for this variation between strains could be quan-

titative differences in the relative amounts ofthe Spn ligand or adhesin inhibited by LNnT.The reason for less inhibition at higher doses of BacteriaLNnT is unknown. However, it is clear thateach Spn serotype has a different optimal con- Spn type 6A (EF3114 was kindly provided by

Dr B. Andersson, Department of Clinical Im-centration of LNnT that is required for maximuminhibition. Our data suggests that both receptors munology, University of Goteborg, Sweden)

have been described previously [9]. Spn type 3exist in tracheal epithelium and adhesins forthese receptors are present on Spn cell surfaces. was originally obtained from Dr. G. S. Giebink,

Department of Pediatrics, University of Min-The fact that the combination inhibition ex-periments with type 6A reduced adherence by nesota and has also been previously described

[26]. Spn type 14 was obtained from the Centeronly 60% might suggest that PAF or other Spn

S. pneumoniae adherence to chinchilla tracheal epithelium 117

for Disease Control, Atlanta, GA. Mid log-phase average tracheal organ culture length in mil-limeters.cultures were prepared from chocolate agar

plate subcultures by inoculating Todd-Hewittbroth (Sigma, St. Louis, MO) with bacteria ob-tained by washing the plates with 5 ml of Dul- Adherence of Spn to neuraminidase-

treated chinchilla tracheabecco’s phosphate-buffered saline (D-PBS). Aftera 2.5 h incubation, the broth cultures were cent-rifuged at 3500 g for 22 min, washed twice and For adherence studies with neuraminidase-

treated tracheae, one half trachea was incubatedresuspended in D-PBS. The concentration of Spnwas determined by standard colony plate count. with D-PBS containing neuraminidase

(0.05–0.5 U/ml, Sigma, St. Louis, MO) for 30 minThe colonial morphology (opaque or trans-parent) of each strain was originally determined at 37°C. The other half of the trachea incubated

in D-PBS without neuraminidase served as aby Dr. Jeffrey Weiser, Children’s Hospital ofPhiladelphia, and confirmed prior to assay ac- control. The trachea was rinsed three times with

D-PBS to remove neuraminidase and then in-cording to the method established by Weiser etal. [27]. Type 6A was predominantly transparent; cubated with 2 ml of a Spn suspension con-

taining 5×107 CFU/ml for 15 to 90 min at roomtype 14 uniformly transparent, and type 3 pre-dominantly opaque. temperature. Each trachea was removed from

the chamber and CFU/mm trachea were thendetermined as described above. Adherence inthe treated half trachea was expressed as percentof controls not treated with neuraminidase. EachTracheal organ cultureexperiment was performed in triplicate.

The trachea organ cultures were prepared asreported in detail previously [3]. Briefly, tracheaewere excised aseptically and kept in prewarmed, Adherence of Spn treated with

oligosaccharides specific for Spn epithelialsterile, minimal essential medium with Earle’ssalts (MEM; M.A. Bioproducts, Walkersville, cell receptorsM.D.). The tracheae were trimmed of extraneousmuscle and connective tissue, sectioned into two To evaluate the ability of two glycoconjugates

known to inhibit Spn adherence, Spnequal halves of approximately 8–10 mm length,and mounted in the central trough of the per- (5×107 CFU/ml) were preincubated for 30 min

at 37°C in a water bath with shaking with eitherfusion chamber with 1.5% agarose in MEM (w/v), so that fluid placed in one well of the chamber lacto-N-neutetraose (Sigma, St. Louis, MO),

which contains the moiety GlcNacb1-3Gal, at acould flow through the tracheal lumen into theadjoining well. The organ cultures were placed final concentration of 1–140 lM in D-PBS or

aGM1 (Sigma Chemical, St. Louis, MO), whichon a rocker platform (Bellco) adjusted to 6.5 cpmand equilibrated prior to the assay for 30 min in contains the moiety GalNacb1-4Gal, at a final

concentration of 0.06–0.25 mM in D-PBS con-a controlled atmosphere chamber (37°C, 5% CO2,Bellco, Vineland, NJ). After equilibration, bac- taining 10% dimethyl sulfoxide (DMSO) and 1%

bovine serum albumin (BSA). A third series ofterial suspensions or test reagents were sub-stituted for equilibration medium. For routine assays served as a negative control and consisted

of Spn incubated with lactose (Sigma, St. Louis,adherence assays, the organ cultures were in-cubated in the controlled atmosphere chamber MO) 28 mM as originally used by Andersson et

al. [13]. After treatment the bacteria were washedfor 30 min with rocking. The bacterial sus-pension was removed, and the trachea was twice, resuspended to a concentration of

107 CFU/ml in D-PBS and added to the chamberwashed three times with 2.0 ml aliquots of sterilePBS before recovery of the tracheae from the for adherence assay with half of an excised

trachea. The other half-trachea served as theagarose plug. The organs were then immersedin 1.0 ml D-PBS and homogenized with the Ten- control and was incubated with Spn, pre-

incubated with diluent without any glyco-broeck tissue grinder (Wheaton, Millville, NJ).Dilutions of the homogenate in duplicate were conjugate. Lacto-N-neotetraose inhibition assays

were conducted with all three strains. Inhibitionplated onto chocolate agar, and colonies werecounted after appropriate incubation. Data are studies with aGM1 were conducted only with

type 6A. To investigate the combined effect ofexpressed as colony forming units (CFU) per

H. H. Tong et al.118

adherence of Haemophilus influenzae to chinchilla trach-lacto-N-neotetraose and aGM1 on the inhibitioneal epithelium. Am J Otolaryngol 1988; 9: 127–34.of Spn adherence, Spn 6A were also treated in

4 Kelly RT, Grieff D, Farmer S. Neuraminidase activityD-PBS with aGM1 at 0.13–0.25 mM for 30 min in Diplococcus pneumoniae. J Bact 1965; 91: 601–3.at 37°C, washed once, then incubated with lacto- 5 Camara M, Boulnois GJ, Andrew PW, Mitchell TJ. A

neuraminidase from Streptococcus pneumoniae has theN-neotetraose (14 lM) for another 30 min atfeatures of a surface protein. Infect Immun 1994; 62:37°C. The adherence assay was then performed3688–95.as described above. In separate control ex-

6 LaMarco KL, Diven WF, Glew RH. Experimental al-periments, Spn were incubated with diluent con- teration of chinchilla middle ear mucosae by bacterialtaining 10% DMSO and 1% BSA. All assays were neuraminidase. Ann Otol Rhinol Laryngol 1986; 95: 304–8.

7 Linder TE, Lim DJ, DeMaria TF. Changes in the structureconducted in triplicate.of the cell surface carbohydrates of the chinchilla tu-Finally, preliminary data were generatedbotympanum following Streptococcus pneumoniae-in-using neuraminidase-treated tracheae to deter-duced otitis media. Microb Pathog 1992; 13: 293–303.mine the effect of neuraminidase on LNnT-in- 8 Linder TE, Daniels RL, Lim DJ, DeMaria TF. Effect of

duced inhibition in vitro. In four separate assays intranasal inoculation of Streptococcus pneumoniae on thestructure of the surface carbohydrates of the chinchillatracheae were incubated with 0.1 units/ml ofeustachian tube and middle ear mucosa. Microb Pathogneuraminidase or with diluent prior to per-1994; 16: 435–41.forming the adherence assay with lacto-N-neo-

9 Andersson B, Dahmen J, Frejd T et al. Identification oftetraose treated Spn 6A. an active disaccharide unit of a glycoconjugate receptorfor pneumococci attaching to human pharyngeal epi-thelial cells. J Exp Med 1983; 158: 559–70.

10 Barthelson R, Mobasseri A, Zopf D, Simon PM. Ad-herence of Streptococcus pneumoniae to respiratory epi-Statistical analysisthelial cells is inhibited by sialylated oligosaccharides.Infect Immun 1998; 66: 1439–44.All results are expressed as arithmetic means 11 Cundell DR, Masure HR, Tuomanen EI. The molecular

and SEM from a minimum of three adherence basis of pneumococcal infection: a hypothesis. Clin Infectassays per data point. Observations are assumed Dis 1995; 21 (Supplement 3): S204–12.

12 Andersson B, Gray BM, Dillon HC, Bahrmand A, Svan-to be approximately normally distributed andborg-Eden C. Role of adherence of Streptococcus pneu-differences between treatment groups of andmoniae in acute otitis media. Pediatr Infect Dis 1988; 7:controls assay were evaluated by Student’s t- 476–80.

test. P<0.05 was selected as the level of sig- 13 Krivan HC, Roberts DD, Ginsburg V. Many pulmonarynificance for all analyses. pathogenic bacteria bind specifically to the carbo-

hydrate sequence GalNAcb1-4Gal found in some glyco-lipids. Proc Natl Acad Sci USA 1988; 85: 6157–61.

14 Tuomanen EI. The biology of pneumococcal infection.Pediatr Res 1997; 42: 253–8.

15 Rosenow C, Ryan P, Weiser JN et al. Contribution ofAcknowledgementsnovel choline-binding proteins to adherence, col-onization and immunogenicity of Streptococcus pneu-moniae. Mol Microbiol 1997; 25: 819–29.This study was supported, in part, by a grant from

16 Beachey EH. Bacterial adherence: adhesin-receptorthe NIDCD/NIH, R01 DC03105-02. The authors wishinteractions mediating the attachment of bacteria toto thank Carrie Schreiner and Tonya Mikes for manu- mucosal surfaces. J Infect Dis 1981; 143: 325–45.

script preparation. 17 Kelly RT, Farmer S, Grieff D. Neuraminidase activitiesof clinical isolates of Diplococcus pneumoniae. J Bact 1967;94: 272–3.

18 O’Toole RD, Goode L, Howe C. Neuraminidase activityin bacterial meningitis. J Clin Invest 1971; 50: 979–85.

19 Alon U, Adler SP, Chan JCM. Hemolytic-uremic syn-Referencesdrome associated with Streptococcus pneumoniae. Reportof a case and a review of the literature. AJDC 1984; 138:496–9.1 Riley ID, Douglas RM. An epidemiological approach

to pneumococcal disease. Rev Infect Dis 1981; 3: 233–45. 20 Diven WF, Doyle WJ, Vietmeier B. Hydrolytic enzymesin otitis media pathogenesis. Ann Otol Rhinol Laryngol2 Andersson B, Eriksson B, Falsen E et al. Adhesion of

Streptococcus pneumoniae to human pharyngeal epi- 1988; 132: 6–9.21 Idanpaan-Heikkila I, Simon PM, Zopf D et al. Oligo-thelial cell in vitro: differences in adhesive capacity

among strains isolated from subjects with otitis media, saccharides interfere with the establishment and pro-gression of experimental pneumococcal pneumonia. Jsepticemia, or meningitis or from healthy carriers. Infect

Immun 1981; 32: 311–17. Infect Dis 1997; 176: 704–12.22 Andersson V, Rosen I, Hakansson A, et al. Antibodies3 Bakaletz LO, Hoepf TM, DeMaria TF, Lim DJ. The

effect of antecedent influenza A virus infection of the to pneumococcal polysaccharides in human milk: lack

S. pneumoniae adherence to chinchilla tracheal epithelium 119

of relationship to colonization and acute otitis media. activated human cells by the receptor platelet-activatingfactor. Nature 1995; 377: 435.Pediatr Infect Dis J 1996; 15: 498–507.

23 Cundell DR, Tuomanen EI. Receptor specificity of ad- 26 Watanbe N, DeMaria TF, Lewis DM, Mogi G, Lim DJ.Experimental otitis media in chinchillas–II: comparisonherence of Streptococcus pneumoniae to human type-II

pneumocytes and vascular endothelial cell in vitro. of the middle ear immune responses to Streptococcuspneumoniae types 3 and 23. Ann Otol Rhinol LaryngolMicrob Pathog 1994; 17: 361–74.

24 Cundell DR, Pearce BJ, Sandros J, Naughton AM, Ma- 1982; 93 (91 Suppl.): 9–16.27 Weiser JN, Markiewicz Z, Tuomanen EI, Wani JH. Re-sure HR. Peptide permeases from Streptococcus pneu-

moniae affect adherence to eucaryotic cells. Infect Immun lationship between phase variation in colony mor-phology, intrastrain variation in cell wall physiology,1995; 63: 2493–8.

25 Cundell DR, Gerard NP, Gerard C, Idanpaan-Heikkila and nasopharyngeal colonization by Streptococcus pneu-moniae. Infect Immun 1996; 64: 2240–5.I, Tuomanen EI. Streptococcus pneumoniae anchor to