INFECTION AND IMMUNITY, Dec. 1992, p. 5030-50350019-9567/92/125030-06$02.00/0Copyright © 1992, American Society for Microbiology

Vol. 60, No. 12

Complement and Antibody Participation inOpsonophagocytosis of Type IV and V Group B Streptococci

MARY A. HALL,* MORVEN S. EDWARDS, AND CAROL J. BAKERSection of Infectious Diseases, Department of Pediatrics, Baylor College ofMedicine,

One Baylor Plaza, Houston, Texas 77030

Received 13 May 1992/Accepted 17 September 1992

Requirements for complement and antibody in neutrophil-mediated killing of serotype IV and V group Bstreptococci were investigated. Neutrophils from adults were tested in an opsonophagocytic assay with serafrom healthy adults, healthy newborns, and hypogammaglobulinemic, agammaglobulinemic, and C4-deficientpatients. For all serum sources, the bactericidal index for both serotypes exceeded 84% after 40 min ofincubation. Heat inactivation of sera ablated killing. Blockade of neutrophil receptor FcIII effected a maximumof 16% inhibition of opsonophagocytosis, and FcII receptor blockade demonstrated negligible inhibition. Whenneutrophil complement receptor 1 or 3 blockade was employed, the maximum inhibition detected was 26%.Simultaneous blockade of complement receptors 1 and 3 effected maximum inhibition levels of 25 and 65% forserotypes IV and V, respectively. Blockade of complement receptor 3 and neutrophil receptor FcIII inhibitedopsonophagocytosis by 56% for both serotypes. When serum complement concentrations were restricted,neutrophil-mediated killing diminished but was restored by the addition of hyperimmune rabbit antiserum.These findings suggest that complement and antibody are major participants in the opsonophagocytosis ofserotypes IV and V group B streptococci. A low prevalence of carriage or mediation of efficient phagocytosisby interactions of neutrophil complement and Fc receptors may contribute to the rarity of human infectionscaused by these two serotypes.

Group B streptococci (GBS) continue to be a major causeof serious infection in neonates and immunocompromisedadults. The type-specific capsular polysaccharide and itssialic acid component have been considered important viru-lence factors among GBS serotypes. The capsular polysac-charides Ia, Ib, II, and III and a protein antigen, c, have beenused to classify GBS strains into distinct serotypes. Approx-imately 5 to 10% of isolates from human sources, however,are considered nontypeable (14, 15). Recently, some isolatespreviously described as nontypeable were characterized asbelonging to antigenically distinct serotypes IV and V (10,20, 21). Like those of other GBS serotypes, type IV and Vcapsular polysaccharides consist of repeating polymers con-taining galactose, glucose, N-acetylglucosamine, and sialicacid (20, 21). Electron microscopy of type IV strains dem-onstrates a thin layer of scantily and irregularly distributedmaterial on the surface of the organism, whereas the type Vcapsule is regular in distribution and more dense (13, 17).Though isolated infrequently, type IV and V GBS do

cause invasive human infection and have been identifiedamong isolates from at least 15 countries (11). Knowledge ofhost defense mechanisms for response to disease due tothese serotypes is limited, however. The present study wasdesigned to investigate host factors possibly defining suscep-tibility to type IV and V GBS. Efficient phagocytosis of otherGBS serotypes by neutrophils (polymorphonuclear leuko-cytes [PMNL]) requires opsonization with complement andmay or may not depend on the presence of antibody to thecapsular polysaccharide (4-6). We examined complementand antibody participation and requirements for complementand Fc receptors in PMNL-mediated killing of these newserotypes.

* Corresponding author.

MATERIALS AND METHODS

Collection and preparation of sera. Sera were separatedfrom whole blood obtained by venipuncture and processedto preserve endogenous complement activity (18). Normalhuman serum (NHS) samples obtained from each of twohealthy adults (NHS1 and NHS2) and a pool obtained froman additional five healthy adults (NHS pool) were employed(Table 1). Hypogammaglobulinemic serum (HGS) was col-lected from a 12-year-old girl with common variable immu-nodeficiency (immunoglobulin G [IgG] level, 39 mg/dl).Agammaglobulinemic serum (AGS) was obtained from a9-year-old boy with severe combined immunodeficiency whohad an IgG level of 10 mg/dl. Neonatal serum samples wereobtained within 48 h of birth from eight healthy termneonates who had normal IgG levels. The 50% hemolyticcomplement levels for each of these sera were within thenormal range (300 to 600 U). Serum from a 6-year-old boywith C4 deficiency (C4D) had a slightly elevated IgG level(1,740 mg/dl) and a 50% hemolytic complement value of 0,which could be restored to normal by adding C4. Heatinactivation of complement activity in sera was carried out at56°C for 30 min.

Preparation of hyperimmune rabbit sera. New ZealandWhite rabbits were immunized with Formalin-killed type IVor V GBS by the method of McCarty and Lancefield (12).Postimmunization blood was collected and centrifuged at4°C, and serum was stored in aliquots at -70°C until use.

Isolation of neutrophils. PMNL were obtained fromhealthy adult volunteers. Citrate-phosphate-dextrose solu-tion (Abbott Laboratories, North Chicago, Ill.) was em-ployed as an anticoagulant, and dextran was used for sedi-mentation of fresh whole blood before purification overFicoll-Hypaque gradients as described previously (1). Iso-lated PMNL were resuspended in Dulbecco's phosphate-buffered saline (PBS; GIBCO, Grand Island, N.Y.) contain-

5030

on Decem

ber 6, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

OPSONINS FOR TYPE IV AND V GROUP B STREPTOCOCCI 5031

ing 0.2% dextrose (dPBS; pH 7.4) for a mean concentrationof 2 x 107 PMNL/ml. Leukocytes were counted manuallyfor each experiment; >95% purity and viability were con-

firmed by using Wright stain and trypan blue (both fromSigma Diagnostics, St. Louis, Mo.), respectively. Wheninhibition of neutrophil complement receptor 1 (CR1) was

involved, erythrocytes were lysed hypotonically (1) beforePMNL were added to the reaction mixture.

Bacteria. Type IV Czechoslovak National Collection ofType Cultures strain 1/82 (originally strain 3139; Rabinowitz,Jerusalem) was isolated from the bloodstream of a newbornwith sepsis (16). Type V Czechoslovak National Collectionof Type Cultures strain 10/84 (originally strain 1169-NT1;Wilkinson, Atlanta) also was isolated from an infant (22).These strains were provided by J. Jelinkova, Prague (11).Lyophilized bacteria were inoculated into Todd-Hewittbroth (Difco Laboratories, Detroit, Mich.), incubated at37°C overnight, aliquoted into sterile vials, and stored at-700C.Antibodies. Monoclonal antibodies (MAbs) to PMNL CR1

(CD35) and CR3 (CD11b/CD18) (Leu 15) were purchasedfrom Becton Dickinson (Mountain View, Calif.). The MAbMol, also directed to CR3, was obtained from Serotec,Bioproducts for Science, Inc. (Indianapolis, Ind.). MAbs toPMNL FcII (CD32) (IV.3) and FcIII (CD16) [3G8 and 3G8F(ab')2 fragment] receptors were purchased from Medarex,Inc. (West Lebanon, N.H.). MAb OKM1 (provided byFrank Schmalsteig, Galveston, Tex. [3]), directed to a lectin-dependent epitope of CR3, was purified from OKM1 murineascites fluid by gradient elution with a high-performanceliquid chromatography hydroxylapatite ion-exchange col-umn (Bio-Rad) (2). M1/70, a MAb directed to both iC3b-specific and lectin-like epitopes of CR3, was provided byDonald Anderson, Houston, Tex. (2, 18). Isotypic nonbind-ing MAbs MsIgGl, IgG2a, and IgG2b were purchased fromCoulter Immunology (Hialeah, Fla.) for use as controlsappropriate to each MAb employed. MAbs preserved insodium azide were dialyzed against 0.01 M PBS (pH 7.4)-0.2% bovine serum albumin (Armour Pharmaceutical Co.,Kankakee, Ill.) in a Spectra/Por microdialyzer (SpectrumMedical Industries, Inc., Los Angeles, Calif.). Concentra-tions of each MAb employed in experiments were deter-mined as saturating for 106 freshly isolated PMNL byimmunofluorescence flow cytometry (3). The concentrationsalso were confirmed as optimal functionally by performingdose-response experiments at concentrations ranging from0.6 to 20 p,g/ml in the opsonophagocytic assay. When MAbcombinations were tested, the final concentration of eachwas equal to its concentration when employed indepen-dently.OP assay. The opsonophagocytosis (OP) assay described

previously (6) was employed; the bacterium/PMNL ratio inthe reaction mixture was approximately 3:1. Intracellularkilling of GBS by PMNL in the presence or absence of MAbwas quantitated. The reaction mixture contained 50 ,ul eachof PMNL and log-phase bacteria suspended in dPBS at

approximate concentrations of 106 cells. Serum with endog-enous complement represented 33% of the reaction mixturevolume, and dPBS or the saturating concentration of MAb indPBS completed a final volume of 300 ,ul. Controls for eachexperiment included reaction mixtures containing serum

from an adult with high OP activity, MAb in the absence ofserum, and tubes lacking either PMNL or serum. Aliquotswere plated for determination of bacterial counts before andafter incubation of the reaction mixture at 37°C with end-over-end rotation for 40 min. The bactericidal index was

calculated as the percent reduction of the initial inoculum inthe reaction mixture at 40 min. The inhibitory effect of MAbwas quantitated and expressed as percent inhibition ofPMNL-mediated killing. The results represent the means ofat least three experiments for each strain unless otherwiseindicated.Antibody determination. Serum antibody to serotype IV

GBS was determined by using an IgG-specific enzyme-linked immunosorbent assay (ELISA) based on the methodof Eisenstein et al. (7). Purified type IV polysaccharide(provided by Michael Wessels, Boston, Mass.) was bound topolystyrene Immulon 1 U-bottom microtiter plates (Dyna-tech Laboratories, Inc., Alexandria, Va.) with poly-L-lysine(Mr, 540,000) as described by Gray (9). A whole-cell ELISAwas carried out to determine antibody titers to type V GBS.Log-phase bacteria were resuspended in carbonate-bicar-bonate coating buffer (pH 9.6) for binding to wells ofmicrotiter plates. Serum dilutions were prepared in PBS-Tween 20. Alkaline phosphatase-conjugated goat anti-hu-man IgG or sheep anti-rabbit IgG (Sigma) was employed tobind human or hyperimmune rabbit serum (IRS), respec-tively. p-Nitrophenyl phosphate disodium substrate tablets(Sigma) were employed for color development, and theoptical density at 405 nm was read on an enzyme immunoas-say reader (model MR 5000; Dynatech Laboratories, Inc.,Chantilly, Va.).

Statistics. Data were analyzed by the Student unpaired ttest (two tailed) (8).

RESULTS

Preliminary experiments were performed to determine thekinetics of opsonophagocytosis of clinical isolates of sero-type IV and V GBS by PMNL from adults. Aliquots wereplated initially and after 10, 20, 40, and 80 min of incubation.At the 10- and 20-min intervals, the NHS pool with addedheat-inactivated IRS reduced the initial inoculum to a greaterextent than did other serum sources (Fig. 1 and 2). At 40min, the NHS pool and HGS also promoted 275% reductionin CFU per milliliter for both serotypes. The addition ofheat-inactivated IRS to the NHS pool further augmentedkilling of type IV GBS. After 80 min, HGS, the NHS pool,and the NHS pool plus heat-inactivated IRS promotedreduction of CFU approaching or exceeding 1 log1o unit(90%) for both serotypes. With heat-inactivation, the NHSpool and IRS effected no reduction in inoculum, and reactionmixtures lacking serum or PMNL permitted bacterialgrowth. The 40-min incubation interval was chosen to ex-amine the participation of complement and antibody inPMNL-mediated killing.

Effect of antibody on OP. The reciprocal serum dilutionsproviding an optical density at 405 nm of 20.2 for the serumsources employed are shown in Table 1. As expected, IgGantibody with specificity for type IV or V GBS was notdetectable in AGS. In contrast, antibodies were detectableto a dilution of 105 in IRS, whereas intermediate levels ofnaturally acquired antibodies were found in the other humansera. The effect of antibody on the OP of type IV and V GBSthen was determined (Table 2). The two adult sera (NHS1and NHS2) and the NHS pool reduced the initial GBSinoculum by at least 84% + 11%. Healthy term neonate seracontaining total IgG and GBS antibody titers comparable tothose in adult sera had mean bactericidal indices (BIs) of89% + 13% for type IV GBS and 97% + 2% for type V GBS.HGS and AGS, the latter with essentially undetectable IgG(10 mg/dl), promoted BIs of 88% + 6% and 91% ± 6%,

VOL. 60, 1992

on Decem

ber 6, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

5032 HALL ET AL.

CFU/ml CFU/mi

0 20 40 60 80

Time (minutes)

FIG. 1. Kinetics of OP of type IV GBS by the NHS pool (0), theNHS pool with heat-inactivated IRS (*), and HGS (+). Controlsincluded heat-inactivated NHS pool (A), heat-inactivated IRS (v),no PMNL (*), and no serum (v). Results represent the means ofthree determinations for each interval. Standard deviations weres70 CFU/ml at 0, 10, and 20 min of incubation. At 40 and 80 min,standard deviations ranged from 1 to 358 and 4 to 1,127 CFU/ml,respectively.

respectively, for type IV GBS and 86% ± 6% and 84% ± 5%,respectively, for type V GBS. Reaction mixtures lackingserum or PMNL permitted bacterial growth.To better assess the role of antibody, diminishing concen-

trations of complement associated with various dilutions ofIRS were employed in the OP assay. The NHS pool wasemployed at 10, 5, and 1% as a serum source alone or withheat-inactivated IRS in the reaction mixture (Fig. 3). With10% NHS pool, IRS minimally augmented the BI. When 5%complement was employed, the effect was more pro-nounced; the mean BIs increased from 67% to 96% (type IV)and from 51% to 92% (type V), respectively. Results with 1%NHS pool were the most striking; IRS increased the BIsfrom 0% to 92% and 89% for type IV and V GBS, respec-tively. Heat-inactivated IRS employed alone as a serumsource did not kill either serotype. The optimal dilution ofIRS was 1:10 or 1:100, depending on the complement con-centration and GBS serotype. These results suggest thatantibody has an important role in PMNL-mediated killing oftype IV and V GBS when complement concentrations arediminished. However, when serum complement is present insufficient concentrations, killing can be mediated indepen-dently of antibody.Requirement for complement in OP. To determine whether

there was an absolute requirement of complement forPMNL-mediated killing of type IV or V GBS, NHS2 andHGS were subjected to heating at 56°C for 30 min. Thisablated bacterial killing, suggesting an absolute requirementof complement for opsonization and phagocytosis to occur.To assess the complement pathway used, C4D serum wasemployed. This serum promoted BIs of 91% ± 1% and 87%

0 20 40 60 80

Time (minutes)FIG. 2. Kinetics of OP of type V GBS by the NHS pool (0), the

NHS pool with heat-inactivated IRS (*), and HGS (+). Controlsincluded heat-inactivated NHS pool (A), heat-inactivated IRS (V),no PMNL (*), and no serum (v). Results represent the means ofthree determinations for each interval. Standard deviations werec70 CFU/ml at 0, 10, and 20 min of incubation. At 40 and 80 min,standard deviations were .250 CFU/ml.

+ 0% for type IV and V GBS, respectively (Table 3). Theseresults indicate that PMNL-mediated killing of types IV andV GBS may proceed via the alternative complement path-way when the classical complement pathway is not avail-able.

Role of complement and Fc receptors in OP. The effects ofPMNL complement and Fc receptor blockade by MAb weretested in the OP assay. Initial experiments determineddose-response relationships of inhibition of OP for eachMAb. To reveal optimum functional activity, MAbs inconcentrations ranging from 0.6 to 20 ,ug/ml in the reaction

TABLE 1. Serum sources employed

Specific IgG titer" ofSerum source CH50 Total IgG GBS type:

IV V

NHS1 398 838 4 400NHS2 451 1,200 4 400NHS pool 467 1,070 64 4,000HGS 626 39 <4 4AGS 352 10 <4 <4Neonatal serum 372 1,177 16 400C4D 0 1,740 4 400C4D plus C4 389 1,740 NAd NAIRS NA NA 100,000 100,000

a The normal range for 50% hemolytic complement (CH50) in adults is 300to 600 U.

b The normal range in adults is 850 to 1,460 mg/dl.I Results represent reciprocal serum dilutions yielding an optical density at

405 nm of .0.2 in the ELISA.d NA, not assessed.

INFECT. IMMUN.

on Decem

ber 6, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

OPSONINS FOR TYPE IV AND V GROUP B STREPTOCOCCI 5033

TABLE 2. Effect of antibody on OP of type IV and V GBS

Mean BI (SD) of GBS type:Serum source

IV V

NHS1 97 (3) 96 (3)NHS2 90 (5) 97 (1)NHS pool 88 (8) 84 (11)HGS 88 (6) 86 (6)AGS 91 (6) 84 (5)Neonatal serum 89 (13) 97 (2)No serum 0 0NoPMNL 0 0

mixture were incubated with neutrophils from adults andbacteria opsonized with NHS1. The optimal concentrationsfor each MAb tested ranged from 2.5 to 10 ,ug/ml. Theseresults correlated with saturating concentrations demon-strated by immunofluorescence flow cytometry, and MAbswere employed as shown in Table 4.

Blockade ofPMNL CR1 by MAb CD35 effected 3% + 2%and 13% + 14% inhibition of OP for serotypes IV and VGBS, respectively (Table 4). CR3 blockade by Leu 15,M1/70, Mol, or OKM1 effected a maximum of 11% + 9%and 26% + 7% inhibition of intracellular killing for type IVand V GBS, respectively. MAb to PMNL FcRIII, 3G8,demonstrated a mean inhibition of7% ± 6% and 16% ± 14%,respectively, for serotypes IV and V. The F(ab')2 fragmentof 3G8 inhibited a mean of 3% + 4% and 5% ± 5%,respectively. MAb to PMNL FcRII, IV.3, mediated negligi-ble inhibition for both serotypes. Isotypic control MAbs didnot effect OP activity. These results suggest that blockade ofa single PMNL receptor is not sufficient to substantiallyimpede PMNL-mediated killing of type IV and V GBS. Inaddition to a low prevalence of carriage, efficient phagocy-tosis, mediated by interactions between neutrophil comple-ment and Fc receptors, may account for the rarity ofinfections caused by these serotypes in neonates.

Since all of the MAbs demonstrated minimal inhibitionindividually (<26%) for type IV and V GBS, a representative

Bactericidal Index M%loor-

80 LJso F

40 F

TABLE 3. Requirement for complement in OP of type IVand V GBS

Mean BI (SD) of GBS type:Serum source0

IV V

NHS2 90 (5) 97 (1)ANHS2 0 3 (6)HGS 88 (6) 86 (6)AHGS ob ob

C4D 91 (1) 87 (0)AC4D Ob ob

No serum 0 0NoPMNL 0 0

a A, heated for 30 min at 56'C.b Represents a single experiment.

type III GBS strain (M732) was used as a control todocument receptor inhibition function. Previously, Smith etal. (18) found that PMNL complement receptors have animportant role in PMNL-mediated killing of type III GBSwith MAbs at concentrations of 5 to 10 p,g/ml. CR1 blockadeeffected no significant difference in inhibition of OP betweentype III GBS and type IV or V GBS. However, inhibition ofkilling of type III by MAbs directed to CR3 epitopes wasconsistent with previous findings (maximum 77%) and sig-nificantly greater than that of type IV or V GBS (maximum26%; P c 0.01) (data not shown).Combination of neutrophil receptor blockade in OP. The

effect on inhibition of intracellular killing by PMNL withMAb combinations was assessed (Table 4). MAbs directedto CR3 (Leu 15 and M1/70) were tested in combination withMAb CD35 (directed to CR1). Both CD35 plus Leu 15 andCD35 plus M1/70 combinations resulted in significant in-creases in inhibition (P < 0.04) of OP of type IV GBS whencompared to inhibition demonstrated by individual MAbs.For type V GBS, the combination of CD35 plus Leu 15effected maximal inhibition of OP. The mean levels ofinteractive inhibition detected were 25% + 5% and 65% ±18% for CD35 plus Leu 15 combinations and 13% + 3% and

Bactericidal Index (S)100r

80 F

SO F

40 -

20 F

0

20-

0-1% NHS

Type IV GBS10% NHS 6% NHS

Type V GBS1% NHS

FIG. 3. Comparison of levels of neutrophil-mediated killing of type IV and V GBS, expressed as the BI, with 10, 5, and 1% NHS pool alone(-) and NHS pool with heat-inactivated IRS (a). Bars represent the means of duplicate determinations for each serotype.

VOL. 60, 1992

on Decem

ber 6, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

5034 HALL ET AL.

TABLE 4. Requirement for complement in OP of type IVand V GBS

Mean (SD) % inhibition ofMAb (concn [(Lg/ml]) initial BIa of GBS type:

IV V

CD35 (2.5) 3 (2)b.C 13 (14/b'Leu 15 (10) 11 (9)bd 26 (7)PaM1/70 (10) 5 (5)C 4 (4)Mol (2.5) 4 (3) 3 (3)OKM1 (10) 5 (5) 1 (1)3G8 (5) 7 (6)d 16 (14)IV.3 (5) 1 (1) 2 (2)CD35 (2.5) plus Leu 15 (10) 25 (5)C 65 (18)CD35 (2.5) plus M1/70 (10) 13 (3)e 46 (25)3G8 (5) plus Leu 15 (10) 56 (17) 56 (11)MsIgGC (8) 4 (4) 5 (5)

a The type IV initial BI was 97% (3%), and the type V initial BI was 96%(3%).

b p c 0.04 compared with data from CD35 plus Leu 15.c P c 0.02 compared with data from CD35 plus M1/70.d p C 0.03 compared with data from 3G8 plus Leu 15.1 P c 0.05 compared with data from 3G8 plus Leu 15.f MsIgG, isotypic murine MAb controls.

46% ± 25% for CD35 plus M1/70 combinations for types IVand V, respectively. PMNL CR3 plus FcRIII blockade wastested in combination using Leu 15 and 3G8. Levels ofinhibition of killing were 56% ± 17% and 56% ± 11%,respectively, for type IV and V GBS and were significantlygreater than those detected when these MAbs were em-ployed alone (P c 0.03). On the basis of these results,multiple deficiencies in complement protein production orPMNL receptor upregulation could explain enhanced viru-lence of GBS types IV and V in some newborns.

DISCUSSION

Knowledge of host factors explaining the susceptibility ofneonates to type IV and V GBS infections is limited. Ourexperiments evaluated the role of complement and antibodyin PMNL-mediated killing of these recently defined sero-types. The objective was to define the relative contributionsand interactions of these opsonins. A number of conclusionsmay be drawn from our results. First, complement appearsto have a prominent role in the OP of type IV and V GBS. Allsera with intact complement, including HGS and AGS,promoted significant PMNL-mediated killing after 40 min ofincubation. Anitibody titers in AGS were undetectable byELISA when compared to the other serum sources, but OPactivities were the same among the various sera. Heatinactivation of sera, regardless of source, ablated killing.Thus, complement is required for OP and appears to becapable of acting independently of antibody in promotingneutrophil-mediated killing of type IV and V GBS.The role of antibody appears to be of some significance.

Kinetically, heat-inactivated IRS increased OP activity pro-moted by the NHS pool at a rate faster than that promotedby the NHS pool alone or by HGS. Antibody appears to beeven more important, however, when complement is lim-ited. Limiting the NHS pool to 5 or 1% concentrationresulted in poor or no OP activity. Killing was restored bythe addition of heat-inactivated, type-specific IRS. Thissuggests that antibody, presumably in very high titers,efficiently promotes killing of type IV and V GBS whencomplement is available in very low concentrations.

Next, the contribution of the interactions between neutro-phil complement and Fc receptors in the OP of type IV andV GBS appears to be substantial. Inhibition by MAb di-rected to PMNL CR1, CR3, FcRII, or FcRIII alone wasminimal. However, when CR1 and CR3 blockade werecombined, inhibition of OP increased significantly. CR3 andFcRIII combinations also significantly reduced bacterialkilling compared with that with blockade of a single PMNLreceptor type. These findings suggest that blockade of asingle PMNL receptor is not sufficient to impede PMNL-mediated killing of type IV and V GBS. Efficient neutrophilcomplement and Fc receptor interactions may account forthe rarity in neonates of infections caused by these sero-types. Likewise, enhanced virulence of these serotypes insome newborns may be explained by deficiencies in PMNLreceptor regulation.We also observed that type IV and V GBS may be

opsonized efficiently via the alternative complement path-way. Serum deficient in C4 supported efficient neutrophil-mediated killing of both organisms. This finding is consistentwith but does not prove the importance of the alternativecomplement pathway for opsonization of type IV and V GBSin the setting of specific antibody deficiency.

Finally, type IV and V GBS have opsonic requirementsthat are distinct from those of other well-defined GBSserotypes. Antibody and complement both play integralroles in OP of type III GBS (6), the most common serotypecausing invasive GBS disease in newborns. Type III strainsrequire antibody with specificity for capsular polysaccharidefor optimal opsonization. In contrast, type Ia GBS strainsmay employ the classical complement pathway for PMNL-mediated killing in an antibody-independent fashion (4).T'ype IV and V GBS appear similar to type Ia GBS in thatcomplement alone may promote killing. However, type IVand V GBS are similar to both of these major serotypes inthat heat-inactivated serum does not support opsonization.In addition, the extent of inhibition of BI seen with mono-clonal antibodies alone or in combination is lower for typeIV and V GBS than that observed for serotype Ia or III GBS(18). In particular, only the combination of 3G8 and Leu 15promoted substantial inhibition for type IV GBS. However,examination of additional clinical strains of types IV and Vwill be necessary to confirm these observations.

Since all of the sera used for these experiments promotedefficient PMNL-mediated killing of type IV and V GBS,these organisms presumably are removed easily from thebloodstream, even in neonates. Complement appears tohave an essential role in opsonization and phagocytosis ofthese two serotypes. Interactive neutrophil complement andFc receptor mediation may contribute to the rarity of type IVand V GBS infections in neonates. Conversely, low preva-lence of carriage may account for the infrequent occurrenceof neonatal disease due to these serotypes. Since the expres-sion of neutrophil receptors in immature hosts may bediminished relative to that in adults (19), enhanced suscep-tibility of some hosts to type IV or V GBS might best beexplained by deficiencies in complement function or de-creased receptor interactions.

ACKNOWLEDGMENTS

This work was supported in part by Public Health Service grantAI19800 from the National Institute of Allergy and InfectiousDiseases.We thank Robin Schroeder for secretarial assistance and Marcia

Rench and Melissa Hickman for technical assistance.

INFECr. IMMUN.

on Decem

ber 6, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

OPSONINS FOR TYPE IV AND V GROUP B STREPTOCOCCI 5035

REFERENCES1. Anderson, D. C., B. J. Hughes, and C. W. Smith. 1981. Abnor-

mal mobility of neonatal polymorphonuclear leukocytes. J.Clin. Invest. 68:863-874.

2. Anderson, D. C., L. J. Miller, F. C. Schmalsteig, R. Rothlein,and T. A. Springer. 1986. Contributions of the Mac-i glycopro-tein family to adherence-dependent granulocyte functions:structure-function assessments employing subunit-specificmonoclonal antibodies. J. Immunol. 137:15-27.

3. Anderson, D. C., F. C. Schmalsteig, M. J. Finegold, B. J.Hughes, R. Rothlein, L. J. Miller, S. Kohl, M. F. Tosi, R. L.Jacobs, T. C. Waldrop, A. S. Goldman, W. T. Shearer, and T. A.Springer. 1985. The severe and moderate phenotypes of herita-ble Mac-1, LFA-1 deficiency: their quantitative definition andrelation to leukocyte dysfunction and clinical features. J. Infect.Dis. 152:668-689.

4. Baker, C. J., M. S. Edwards, B. J. Webb, and D. L. Kasper.1982. Antibody-independent classical pathway-mediated op-sonophagocytosis of type Ia, group B Streptococcus. J. Clin.Invest. 69:394-404.

5. Baker, C. J., B. J. Webb, D. L. Kasper, and M. S. Edwards.1986. The role of complement and antibody in opsonophago-cytosis of type II group B streptococci. J. Infect. Dis. 154:47-54.

6. Edwards, M. S., A. Nicholson-Weller, C. J. Baker, and D. L.Kasper. 1980. The role of specific antibody in alternative com-plement pathway-mediated opsonophagocytosis of type III,group B streptococcus. J. Exp. Med. 151:1275-1287.

7. Eisenstein, T. K., B. J. Cueninck, D. Resavy, G. D. Shockman,R. B. Carey, and R. M. Swenson. 1983. Quantitative determina-tion in human sera of vaccine-induced antibody to type-specificpolysaccharides of group B streptococci using an enzyme-linkedimmunosorbent assay. J. Infect. Dis. 147:847-856.

8. Glantz, S. A. 1987. Primer of biostatistics, 2nd ed. McGraw-HillBook Co., New York.

9. Gray, B. M. 1979. ELISA methodology for polysaccharideantigens: protein coupling of polysaccharides for adsorption toplastic tubes. J. Immunol. Methods 28:187-192.

10. Jelhnkova, J., and J. Motlova. 1985. Proposed revision of no-menclature of group B Streptococcus (GBS) serotypes, p.51-52. In Y. Kimura, S. Kotani, and Y. Shiokawa (ed.), Recentadvances in streptococci and streptococcal diseases. ReedBooks, Berkshire, England.

11. Jelinkova, J., and J. Motlova. 1985. Worldwide distribution oftwo new serotypes of group B streptococci: type IV andprovisional type V. J. Clin. Microbiol. 21:361-362.

12. McCarty, M., and R. C. Lancefield. 1955. Variation in thegroup-specific carbohydrate of group A streptococci. J. Exp.Med. 102:11-17.

13. Molinari, A., C. Von Hunoistein, G. Donelli, S. Paradisi, G.Arancia, and G. Orefici. 1987. Effects of some capsular compo-nents on pathogenicity of type IV and provisional type V groupB streptococci. FEMS Microbiol. Lett. 41:69-72.

14. Motlova, J., M. Wagner, and J. Jelinkova. 1986. A search fornew group-B streptococcal serotypes. J. Med. Microbiol. 22:101-105.

15. Perch, B., E. Kjems, and J. Henrichsen. 1979. New serotypes ofgroup B streptococci isolated from human sources. J. Clin.Microbiol. 10:109-110.

16. Rabinowitz, S., M. Ferne, and S. T. Halfon. 1980. Group Bstreptococci: different serotypes associated with clinical infec-tions. Lab. Pract. 29:733-734.

17. Ryc, M., J. Jelinkova, J. Motlova, and M. Wagner. 1988.Immuno-electronmicroscopic demonstration of capsules ongroup-B streptococci of new serotypes and type candidates. J.Med. Microbiol. 25:147-149.

18. Smith, C. L., C. J. Baker, D. C. Anderson, and M. S. Edwards.1990. Role of complement receptors in opsonophagocytosis ofgroup B streptococci by adult and neonatal neutrophils. J.Infect. Dis. 162:489-495.

19. Smith, J. B., D. E. Campbell, A. Ludomirsky, R. A. Polin, S. D.Douglas, B. Garty, and M. C. Harris. 1990. Expression of thecomplement receptors CR1 and CR3 and the type III Fcreceptor on neutrophils from newborn infants and from fetuseswith Rh disease. Pediatr. Res. 28:120-125.

20. Wessels, M. R., V.-J. Benedf, H. J. Jennings, F. Michon, J. L.DiFabio, and D. L. Kasper. 1989. Isolation and characterizationof type IV group B Streptococcus capsular polysaccharide.Infect. Immun. 57:1089-1094.

21. Wessels, M. R., J. L. DiFabio, V.-J. Benedi, D. L. Kasper, F.Michon, J. Brisson, J. Jelinkova, and H. J. Jennings. 1991.Structural determination and immunochemical characterizationof the type V group B Streptococcus capsular polysaccharide. J.Biol. Chem. 266:6714-6719.

22. Wilkinson, H. W. 1977. Nontypable group B streptococci iso-lated from human sources. J. Clin. Microbiol. 6:183-184.

VOL. 60, 1992

on Decem

ber 6, 2020 by guesthttp://iai.asm

.org/D

ownloaded from