Receptor 2-Deficient Mice Polysaccharide in C3- and Complement Streptococcal Type III Capsular Impaired Antibody Response To

Impaired Antibody Response to Group B Streptococcal Type III Capsular Polysaccharide in C3- and Complement Receptor 2-Deficient Mice This information is current as of September 24, 2021. Olga Pozdnyakova, Hilde-Kari Guttormsen, Farah N. Lalani, Michael C. Carroll and Dennis L. Kasper J Immunol 2003; 170:84-90; ; doi: 10.4049/jimmunol.170.1.84 http://www.jimmunol.org/content/170/1/84 Downloaded from

References This article cites 38 articles, 16 of which you can access for free at:

http://www.jimmunol.org/content/170/1/84.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 24, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2003 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Impaired Antibody Response to Group B Streptococcal Type III Capsular Polysaccharide in C3- and Complement Receptor 2-Deﬁcient Mice1

Olga Pozdnyakova,2* Hilde-Kari Guttormsen,2‡ Farah N. Lalani,‡ Michael C. Carroll,* and Dennis L. Kasper3†‡

Group B Streptococcus (GBS) is the foremost bacterial cause of serious neonatal infections. Protective immunity to GBS is mediated by specific Abs to the organism’s capsular polysaccharide Ags. To examine the role of complement in the humoral immune response to type III GBS capsular polysaccharide (III-PS), mice deficient in C3 or in CD21/CD35 (i.e., complement receptors 1 and 2; CR1/CR2) were immunized with III-PS. Mice deficient in C3 or Cr2 had an impaired primary immune response to III-PS. The defective response was characterized by low IgM levels and the lack of an isotype switch from IgM to IgG Ab Downloaded from production. Compared with wild-type mice, C3- and Cr2-deficient mice exhibited decreased uptake of III-PS by follicular dendritic cells within the germinal centers and impaired localization of III-PS to the marginal zone B cells. Complement-dependent uptake of capsular polysaccharide by marginal zone B cells appears necessary for an effective immune response to III-PS. The normal immune response in wild-type mice may require localization of polysaccharide to marginal zone B cells with subsequent transfer of the Ag to follicular dendritic cells. The Journal of Immunology, 2003, 170: 84Ð90. http://www.jimmunol.org/ espite advances in antimicrobial therapy, infections with the help of T cells, but the presence of T cell factors enhances the extracellular encapsulated bacteria such as Neisseria IgG response. meningitidis, Streptococcus pneumoniae, and group B Despite the clinical significance of bacterial polysaccharides, the D 4 Streptococcus (GBS) remain important clinical problems. GBS is mechanisms underlying immune responses to polysaccharide Ags the most common cause of serious infections in newborns and have not yet been fully elucidated. It has been suggested that the young infants in most of the developed world, with a 5–8% case- cells within the splenic marginal zone (MZ) play an important role fatality ratio in the United States (1). Nine serotypes of GBS have in responses to TI-2 Ags. The MZ is a specialized compartment been identified on the basis of distinct capsular polysaccharides. containing a subpopulation of B lymphocytes, i.e., MZ B cells, by guest on September 24, 2021 Type III is the serotype most frequently associated with meningitis dendritic cells, and MZ macrophages (5). One hypothesis impli- Ͼ and with infant infections of late onset ( 7 days after birth). cates MZ B cells in the induction of humoral responses to TI-2 Ab responses in the spleen play an essential role in host defense Ags. MZ B cells are distinguished from follicular B cells by cell against encapsulated bacteria (2). Accordingly, patients with ana- surface phenotype (e.g., IgMhigh, CD21high, IgDlow, CD23low). tomic or functional asplenia are highly susceptible to infections Lane et al. (6) reported that the immune response of recovering with these organisms (3, 4). Protection against neonatal type III x-irradiated mice to TI-2 Ags correlated with the reappearance of GBS infections is associated with the presence of Abs specific for B cells in the MZ. Furthermore, responsiveness to polysaccharide the type III GBS capsular polysaccharide (III-PS). Pure polysac- Ag develops in children at ϳ2 years of age—a timing that coin- charides are characterized as thymus-independent type 2 (TI-2) cides with the maturation of MZ B cells (7, 8). A recent study with Ags. These Ags can induce a primary immune response without Pyk-2-deficient mice lacking MZ B cells showed a reduction in IgM and IgG responses to trinitrophenyl (TNP)-Ficoll (a model TI-2 Ag) as well as abrogated Ag localization to MZ B cells. *Departments of Pathology and Pediatrics, Center for Blood Research, †Department However, studies of CD19-deficient mice, which also have re- of Microbiology and Molecular Genetics, Harvard Medical School, and ‡Channing duced numbers of MZ B cells, have revealed normal-if not ele- Laboratory, Department of Medicine, Brigham and Women’s Hospital, Harvard Med- ical School, Boston, MA 02115 vated-responses to haptenated Ficoll (9, 10). An alternative hy- Received for publication May 24, 2002. Accepted for publication October 23, 2002. pothesis is that MZ macrophages are essential in early uptake of TI-2 Ags. Macrophages within the MZ take-up and retain carbo- The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance hydrate macromolecules such as Ficoll (11). However, the impor- with 18 U.S.C. Section 1734 solely to indicate this fact. tance of macrophage uptake remains unclear because in vivo elim- 1 This work was supported by National Institutes of Health Grants 1R01 AI36389 (to ination of MZ macrophages appears not to alter the response to M.C.C.) and AI23339 (a Merit Award to D.L.K.). TNP-Ficoll (12). 2 O.P. and H.-K.G. contributed equally. Binding of Ficoll by MZ B cells in naive wild-type (WT) mice 3 Address correspondence and reprint requests to Dr. Dennis L. Kasper, Channing has been shown to be dependent on C3 and on CD21/CD35 (com- Laboratory, 181 Longwood Avenue, Boston, MA 02115. E-mail address: [email protected] plement receptors (CR) 1 and 2; CR1/CR2) (13), whereas uptake 4 Abbreviations used in this paper: GBS, group B Streptococcus; III-PS, type III GBS by MZ macrophages takes place independent of C3 (3). Despite capsular polysaccharide; TI-2, thymus-independent type 2; MZ, marginal zone; TNP, this differential complement dependence for Ag uptake by MZ trinitrophenyl; WT, wild type; CR, complement receptor; CVF, cobra venom factor; FDC, follicular dendritic cell; tg, transgenic; HSA, human serum albumin; ASC, cells, a role for complement in the humoral response to TI-2 Ags Ab-secreting cell; GC, germinal center; PNA, peanut lectin agglutinin. remains controversial. Early studies by Pepys (14) showed that

complement depletion with cobra venom factor (CVF) had no ef- ELISPOT assay for III-PS-specific Ab-secreting cells (ASCs) fect on the primary response to polyvinylpyrrolidone 360. Studies The frequency of ASCs in spleens was determined by ELISPOT assay. In by Markham et al. (15) demonstrated that complement depletion brief, 24-well plates (Falcon, San Diego, CA) were coated overnight with with CVF abrogated the Ab response to pneumococcal type 14 III-PS conjugated to HSA (III-HSA; 5 ␮g/ml). Wells were blocked with polysaccharide but did not affect the response to the sialic acid- 1% BSA in PBS, splenocytes diluted in DMEM (Life Technologies, Grand ϫ 6 ϫ 6 ϫ 6 containing III-PS. In contrast, studies by Pryjma and Humphrey Island, NY) were added (5 10 ,1 10 ,or0.5 10 cells per well), and preparations were incubated for 12–18hat37°C. Plates were washed (16) identified a reduced immune response to pneumococcal type in 0.1% BSA, then coated for 4 h at room temperature with goat anti-mouse 3 polysaccharide in CVF-treated mice. Moreover, Griffioen et al. IgM or IgG conjugated to alkaline phosphatase (Sigma-Aldrich). Plates (17) documented enhanced Ab responses to pneumococcal poly- were developed with X-phosphate/5-bromo-4-chloro-indolyl-phosphate saccharide upon conjugation to C3d. Humans deficient in comple- (BCIP; Boehringer Mannheim, Indianapolis, IN) in 2-amino-2-methyl-1- propanol (AMP; Sigma-Aldrich) agarose solution. Spots were counted with ment exhibit increased susceptibility to infections by encapsulated a Leica microscope (Deerfield, IL). bacteria, possibly because of impaired humoral immunity and lack of effector function (18). Much of our knowledge about humoral immunity to TI-2 Ags comes from studies conducted with nonphysiological Ags. To investigate the role of complement in humoral immunity to a clinically relevant polysaccharide, we immunized mice deficient in C3 or CD21/CD35 with purified III-PS. These mice had impaired IgM

and IgG responses to III-PS; the impairment was characterized by Downloaded from a negligible uptake of Ag by follicular dendritic cells (FDCs) and MZ B cells. These observations suggest that humoral responses to a clinically relevant TI-2 Ag-containing sialic acid depend on complement-tagged Ag trapping and processing by MZ B cells.

Materials and Methods http://www.jimmunol.org/ Mice WT, C3-deficient (C3null), and CD21/CD35-deficient (Cr2null) mice were maintained on a C57BL6/129Sv mixed background. T cell-deficient (CD3⑀ transgenic (tg)) tg mice and WT controls were maintained on a C3H background. All mice were used at 6–10 wk of age. Mice were housed in a specific pathogen-free barrier animal facility at Harvard Medical School (Boston, MA).

Preparation of III-PS and conjugates by guest on September 24, 2021 Type III GBS strain M781 was the source of the puriﬁed polysaccharide. III-PS was conjugated to human serum albumin (HSA; Sigma-Aldrich, St. Louis, MO) and to biotin hydralazine (Pierce, Rockford, IL) by reductive amination (19). The conjugated polysaccharide retained the speciﬁcity of the pure polysaccharide, as shown by ELISA inhibition.

Immunization protocol On day 0, mice were injected i.p. with 8 ␮g of III-PS in saline or with saline only. Mice were bled 3 days before and 11 days after immunization. Levels of type III-speciﬁc IgM and IgG were quantiﬁed by ELISA. To assess III-PS localization in spleens, mice were injected i.v. with 50 ␮gof biotinylated III-PS at different time points before harvest (15 min, 30 min, 1 h, 2 h, or 16 h). Harvested spleens were either snap-frozen for cryosec- tioning or mechanically dissociated in preparation for FACS analysis and/or ELISPOT).

Reconstitution with III-PS-speciﬁc IgG and IgM Sera from naive mice and from mice immunized i.p. with three 2-␮g doses of III-PS covalently linked to tetanus toxoid (III-PS-TT absorbed to 0.5 mg

AlOH2) were separated into IgG- and IgM-containing fractions over a protein G column using an Immunopure (G) IgG Purification kit (Pierce) per instructions from the manufacturer. The fractions were desalted over a PD10 column (Amersham Pharmacia Biotech, Piscataway, NJ) and cali- brated with PBS (pH 7.4). The fractions from immunized mice contained 5.9 and 4.3 ␮g/ml of III-PS-specific IgG and IgM, respectively; those frac- FIGURE 1. The Ab response to III-PS is dependent on C3 and Cr2. ␮ tions from naive mice contained 0.005 and 0.023 g/ml, respectively. WT Primary IgM (a) and IgG (b) responses to III-PS were significantly lower null ␮ and C3 mice received an i.v. injection of IgG or IgM (200 l each) from in mice deficient in C3 or in CD21/CD35 than in WT mice 11 days after immunized and naive mice 2 h before injection of biotinylated III-PS. immunization. ELISA data from one representative experiment are shown; Values of p Ͻ 0.05. Horizontal ,ء .a total of five experiments were performed ELISA for III-PS-specific Ab bars indicate means. The frequency of IgM ASCs (c) was significantly reduced Levels of III-PS-specific Abs were quantitated by ELISA, as described in the spleens of C3- and Cr2-deficient mice and in those of saline-injected WT previously (20). The limits of IgM and IgG Ab detection were 20 and 5 mice. Mice were immunized with 8 ␮g of III-PS, and splenocytes were col- ng/ml, respectively. lected for an ELISPOT assay on day 16 after immunization. 86 COMPLEMENT-DEPENDENT HUMORAL RESPONSE TO GBS POLYSACCHARIDE

Immunofluorescence analysis of splenic sections were harvested on day 16 after immunization, and single-cell sus- For immunofluorescence analysis, 7-mm cryosections of spleen were cut, pensions were analyzed in an ELISPOT assay (Fig. 1c). Consistent fixed, and stained as described previously (21). At least 10 sections per with the low Ab levels, only negligible numbers of IgM ASCs spleen were analyzed. Biotinylated Ag was detected with fluorescence- were observed in C3null, Cr2null, and saline-injected WT mice as labeled streptavidin (streptavidin-Cy-Chrome, BD PharMingen (San Di- opposed to immunized WT mice; specific values (per 5 ϫ 106) ego, CA); or streptavidin-Alexa 568 Fluor, Molecular Probes (Eugene, were 56 Ϯ 22 for immunized WT mice, 5 Ϯ 1 for immunized OR). The following Abs were used for staining: rat anti-mouse FDC-M1, null Ϯ null Ϯ anti-mouse IgM-Cy5 (clone 341.12), peanut lectin agglutinin (PNA)-FITC C3 mice, 7 3 for immunized Cr2 mice, and 4 0.5 for (EY Laboratories, San Mateo, CA), anti-mouse CD3-PE (BD PharMin- saline-injected WT mice ( p Ͻ 0.01 for all groups). Negligible gen), anti-mouse IgD-FITC (Southern Biotechnology Associates, Birming- numbers of IgG ASCs were observed in all four groups (data not ham, AL), anti-mouse IgM-PE (BD PharMingen), and anti-human C3d- shown). FITC (DAKO, Carpinteria, CA). Pictures were taken with a Bio-Rad MRC confocal microscope (Hercules, CA) using Bio-Rad Radiance 2000 software. Germinal center (GC) response and complement-mediated FACS analysis retention of III-PS within GCs Single-cell suspensions of splenocytes were stained with rat anti-mouse Humoral responses are dependent on histologically defined areas Abs CD11b-PE, CD24-PE, and CD23-FITC (BD PharMingen). III-PS bi- within splenic follicles called GCs. Although GC formation is usu- otin was detected with streptavidin-allophycocyanin (Molecular Probes). ally described in immune responses to thymus-dependent Ags, re- Cells were analyzed with a FACSCalibur (BD Biosciences, Mountain cent studies have demonstrated that GCs can also develop in re- View, CA) flow cytometer. sponse to TI Ags such as dextran B512 (22, 23). To investigate GC Downloaded from Statistical analysis formation in response to III-PS, we harvested spleens and prepared splenic sections on day 13 after primary immunization. Although Statistical analysis was performed with Prism software, version 2.0b the number of splenic follicles did not differ among the four (GraphPad Software, San Diego, CA). The Student t test for unequal vari- ϩ ances was used to compare means, and the ␹2 test was used to compare groups, the number of GCs (PNA clusters) was significantly frequencies for different experimental groups. A two-tailed p value of lower in immunized Cr2-deficient mice and saline-injected WT Ͻ0.05 was considered significant.

mice than in immunized WT mice (Table I). Notably, PNA-positive http://www.jimmunol.org/ GCs were observed in C3-deficient mice despite an impaired Ab re- Results and Discussion null null sponse and an absence of specific Ag staining in the follicles. The Primary Ab response to III-PS in WT, C3 , and Cr2 mice GCs probably represent an ongoing response to environmental Ags. To analyze the role of complement in primary Ab responses to To examine whether III-PS localization within GCs is depen- III-PS, we immunized WT, C3null, and Cr2null mice i.p. with 8 ␮g dent on C3 and Cr2, we injected immunized mice with biotinylated of III-PS and collected serum samples at multiple time points after III-PS and harvested their spleens 16 h later (Fig. 2 and Table I). immunization. III-PS-specific Ig were quantified by ELISA (Fig. 1, Immunofluorescence analysis of splenic sections revealed a high a and b). Before day 11 no significant IgM response was observed percentage of III-PS-positive GCs in immunized WT mice (47% Ϯ in either complement-deficient or WT mice (data not shown). 10%). In contrast, a negligible number of splenic follicles stained by guest on September 24, 2021 However, the day 11 IgM response of immune WT mice to the positive for the capsular polysaccharide in immunized deficient polysaccharide Ag was significantly greater than that of saline- mice and saline-injected WT mice. As expected, III-PS colocalized injected WT mice (6493 Ϯ 2418 ng/ml vs 190 Ϯ 25 ng/ml; p Ͻ with M1ϩ FDCs within GCs of immunized WT mice (Fig. 2, a and 0.05). By contrast, the primary IgM response was impaired in the b). However, no III-PS was detected in GCs of immunized defi- deficient mice. Notably, IgM Ab levels in deficient mice were in a cient mice or in saline-injected WT controls (Fig. 2, cÐh, and range similar to that measured in saline-injected WT mice (311 Ϯ Table I) despite the presence of a network of FDCs in each group. 43 ng/ml for C3null mice and 80 Ϯ 18 ng/ml for Cr2null mice; Fig. Localization of III-PS to GCs does not in itself define specificity of 1a). The IgG isotype switch observed in immunized WT mice GC B cells, it does, however, provide a relative measure of the (366 Ϯ 173 ng/ml) did not occur in immunized C3null mice (4 Ϯ humoral response to the specific Ag. 0.3 ng/ml), immunized Cr2null mice (5 Ϯ 3 ng/ml), or saline-in- To examine the importance of T cells in humoral and GC re- jected WT mice (6 Ϯ 1 ng/ml) ( p Ͻ 0.01 for all groups; Fig. 1b). sponses to III-PS, we immunized CD3⑀ tg mice, which are defi- To further examine the nature of the defective response in de- cient in T and NK cells, with III-PS (24). CD3⑀ tg mice had normal ficient mice, we analyzed the frequency of IgM ASCs. Spleens IgM responses; mean anti-III-PS IgM levels Ϯ SEM were 1000 Ϯ

Table I. III-PS retention within the GCs of micea

No. Spleen No. III-PSϩ No. III-PSϩ GCs Genotype Follicles No. GCs GCs ϫ 100%/ GCs

WT-C57BL6/129Sv (n ϭ 10) 45.7 Ϯ 3.3 21.7 Ϯ 2.9 10.6 Ϯ 2.6 47.1 Ϯ 10.0 C3null-C57BL6/129Sv (n ϭ 8) 37.9 Ϯ 2.6 21.8 Ϯ 4.1 0 0 Cr2null-C57BL6/129Sv (n ϭ 6) 50.0 Ϯ 5.3 4.0 Ϯ 1.0b 00 WT-C57BL6/129Sv saline (n ϭ 10) 31.8 Ϯ 2.5 9.5 Ϯ 1.4b 00 CD3␧ tg/C3H (n ϭ 10) 29.8 Ϯ 2.7 0.4 Ϯ 0.3b 00 WT/C3H (n ϭ 9) 32.3 Ϯ 1.4 12.6 Ϯ 1.7 3.5 Ϯ 1.0 27.8 Ϯ 5.4

a Immunohistochemical analysis of splenic sections prepared from WT, C3null, and Cr2null mice maintained on a mixed C57BL6/129Sv background and of CD3␧ tg and WT mice maintained on a C3H background. Spleens were harvested from immune and control mice 16 h after i.v. injection of 50 ␮g of biotinylated III-PS. Sections were analyzed, and the mean Ϯ SEM values are presented. Follicles and GCs were identiﬁed by anti-IgM and PNA staining, respectively. Ag was detected by biotin-streptavidin binding. No signiﬁcant difference in spleen size among mouse strains (as indicated by the total number of spleen follicles) was observed. b p Ͻ 0.01 vs immunized WT mice. The Journal of Immunology 87

Localization and deposition of III-PS in the MZ The splenic MZ is an important site in defense against encapsulated bacteria. Detection and uptake of particulate Ags are enhanced by the macrophages lining the MZ sinuses (12). Moreover, the MZ region includes B cells that bear specific receptors for bacterial Ags, including bacterium-specific carbohydrate Ags (26). To investigate which cell types in the MZ of the spleen are re- sponsible for the uptake of III-PS and to determine whether clear- ance is complement-dependent, we injected biotinylated III-PS i.v. into immune and nonimmune mice of each group and harvested the spleens at various time points. Ag uptake in the MZ was as- sessed by confocal microscopy and FACS analysis. Analysis of splenic sections from nonimmune mice identified III-PS staining within the MZ of all three groups (Fig. 3, a, d, and g). Ag deposition colocalized with CD11bϩ cells in naive mice, a result sug- gesting uptake by macrophages (data not shown). In immune WT mice, III-PS localized to MZ B cells and FDCs within GCs (Fig. 3b). A significant finding was the lack of Ag staining in splenic Downloaded from sections from immunized C3- and Cr2-deficient mice (Fig. 3, e and h). These results suggest that Ag uptake by MZ B cells is dependent on C3 and CD21/CD35 in immune mice, whereas uptake of III-PS by macrophages in naive mice is complement-independent. To determine whether III-PS detected on MZ B cells of immune mice correlates with complement deposition, we stained splenic http://www.jimmunol.org/ sections with anti-C3d, III-PS-biotin, and anti-IgM (Fig. 3, c, f, and i). As predicted, C3 protein and III-PS colocalized to FDCs within GCs and MZ B cells in immune WT mice (Fig. 3c). By compar- ison, low levels of C3 staining were evident in the sinuses and blood vessels in spleens from Cr2-deficient mice, but no C3 was detectable within the splenic follicles (Fig. 3i). No C3 protein or III-PS was found in spleens of C3-deficient mice (Fig. 3f). To further examine colocalization of III-PS and C3 protein on

MZ B cells, we prepared suspensions of splenic cells from immu- by guest on September 24, 2021 nized WT and deficient mice and subjected these cells to FACS analysis (Fig. 3j and Table II). MZ B cells were identified as positive for B220 and CD24 (heat-stable Ag) and negative for CD23. No reduction in the number of MZ B cells was apparent in C3- and FIGURE 2. Retention of III-PS within GCs and colocalization with Cr2-deficient mice (data not shown). Analysis of MZ B cells har- FDCs in WT mice. a and b, III-PS was bound on FDCs within the GCs of vested at various time points indicated increasing uptake of Ag immune WT mice 16 h after injection with biotinylated Ag. cÐh, Despite over the 120-min period after i.v. Ag injection (Fig. 3j and Table the normal appearance of FDCs in Cr2null,C3null, and saline-injected WT II). The frequency of III-PS-positive MZ B cells was significantly mice, no Ag staining was evident within the GCs. Three-color immunofluorescence analysis was performed on splenic cryosections from WT, greater in WT mice than in deficient mice at each time point ex- C3-deficient, and Cr2-deficient mice. B cell follicles are stained blue amined. Numbers of III-PS-positive MZ B cells were negligible in (IgM), FDCs are green (FDC-M1), and III-PS is red. FDC-positive areas saline-injected WT mice (data not shown). correspond to GCs. Colocalization of green and red appears yellow on a The frequency of C3 binding by MZ B cells increased after two-color image (a), and colocalization of all three colors appears white on injection of Ag in WT mice (Table II). However, a background a three-color image (b). Insets in a and b indicate a one-color image of level of C3-positive MZ B cells was observed in saline-injected III-PS within the GC. Results are representative of at least three of the WT controls. This result suggests that complexes of C3 protein and experiments summarized in Table I. Ags (probably environmental Ags) are taken up constitutively by MZ B cells. C3 detection was judged to be specific because neg- 300 ng/ml for WT mice and 1500 Ϯ 400 ng/ml for CD3⑀ tg mice. ligible numbers of C3-positive cells were observed in C3-deficient In contrast, anti-III-PS IgG levels were negligible in CD3⑀ tg mice mice. Uptake of III-PS and C3 is essentially Cr2-dependent, al- (9 Ϯ 0.5 ng/ml vs 263 Ϯ 84 ng/ml for WT mice; p Ͻ 0.01). The though some C3 and III-PS binding took place in the absence of absence of GCs in T and NK cell-deficient mice was consistent Cr2. Expression of other complement receptors (e.g., crry/p65) with an impaired IgG response (13 Ϯ 2 and 0.4 Ϯ 0.3 GCs for WT that bind C3b (27) or residual CD21/CD35 (28) may explain the and CD3⑀ tg mice, respectively; p Ͻ 0.01). III-PS deposition was low levels of C3 on Cr2null MZ B cells. The results of splenic cell not observed in the splenic follicles of CD3⑀ tg mice, whereas WT analysis by confocal microscopy and FACS strongly suggest that mice had detectable III-PS in 27.8% Ϯ 5.4% of GCs. Therefore, III-PS decorated with activated complement products efficiently the presence of T cells enhances IgG production and the GC re- localizes to FDCs and MZ B cells by binding through CD21/CD35 sponse, even though III-PS is considered a TI Ag. We cannot rule complement receptors. out a role for NK cells in response to III-PS, as they can modulate An alternative explanation for the lack of III-PS uptake by MZ TI-2 responses (25). In contrast, the IgM response to III-PS ap- B cells in C3- and Cr2-deficient mice is the failure of both strains pears to be T- and NK-cell independent. to produce significant levels of III-PS-specific Abs (Fig. 1). To 88 COMPLEMENT-DEPENDENT HUMORAL RESPONSE TO GBS POLYSACCHARIDE

FIGURE 3. Localization of III-PS Ag in spleens of nonimmune and immune mice injected with biotinylated III-PS 2 h before harvest. III-PS localizes to the outer MZ in nonimmune mice (a, d, and g). Binding is independent of complement. The Ag localizes to the GC in immunized WT mice (b) and colocalizes with

C3d and IgM (c). In contrast, III-PS is not detected Downloaded from within splenic follicles of immunized C3null mice (e and f)orCr2null mice (h and i). IgMhigh, IgDlow MZ B cells stain red; the MZ is indicated by double-headed arrows. IgMlow, IgDhigh follicular B cells appear Ag localized within the ,ء .green, and III-PS is blue GC of a WT mouse (b). j, III-PS binding is detected on B220ϩ CD24ϩCD23Ϫ MZ B cells (ﬁlled rectangular) http://www.jimmunol.org/ of III-PS immune WT mice. III-PS binding on immunized C3null or Cr2null MZ B cells was negligible (see Table II). The amount of bound Ag increases with time. A histogram illustrates III-PS binding by MZ B cells in a representative immune WT mouse at the indicated time points. by guest on September 24, 2021

determine whether Ab alone or both Ab and complement are re- III-PS-specific Abs in reconstituted WT mice. Alternatively, ac- quired for opsonization of III-PS, naive WT and C3null mice were tively immunized mice developed an ongoing response, whereas given an i.v. injection of 1.2 ␮g of IgG or 1 ␮g of IgM anti-III-PS passively immunized mice received only one dose of III-PS-spe- Abs before receiving an injection of biotinylated III-PS. The levels cific Abs. Another explanation is the possible clonal expansion of of III-PS-specific Abs detected in sera at the time of harvest were III-PS-specific B cells in the MZ of immune mice. Evidence of clonal similar in WT and C3null mice (data not shown). FACS analysis expansion of MZ B cells in humans was recently reported (29). performed on splenic cells from the two strains revealed uptake of In summary, we have demonstrated that the humoral response to III-PS by MZ B cells from WT mice (4.2 and 4.9% for III-PS- the clinically relevant, sialic acid-containing capsular polysaccha- specific IgG and IgM, respectively) but not by those from C3null ride Ag of GBS type III is complement-dependent. Notably, mice mice (0.7 and 0.8%, respectively) 2 h after i.v. Ag injection (Fig. deficient in C3 or CD21/CD35 have impaired responses charac- 4). Therefore, uptake of III-PS by MZ B cells requires both spe- terized by low IgM levels and negligible ASC numbers. No isotype cific Abs and C3. Uptake of III-PS by MZ B cells from WT and switch to IgG Abs takes place in these deficient mice. In naive C3null mice that were given IgG and IgM fractions from naive sera mice, III-PS is rapidly taken up by macrophages within the splenic was Ͻ1% (data not shown). Notably, the level of uptake in naive MZ in a complement-independent manner. By contrast, in immune WT mice reconstituted with III-PS specific Abs was lower than mice, III-PS is localized to the surface of MZ B cells by a mech- that in immunized WT mice (4.2 (IgG naive mice) and 4.9% (IgM anism that is dependent on both C3 and CD21/CD35. The lack of naive mice) vs 18.4% (III-PS immunized)). One possible expla- III-PS bound to MZ B cells in C3- and Cr2-deficient mice corre- nation for the differences could be the reduced levels of circulating lates with impaired localization of Ag to FDCs within GCs. The Journal of Immunology 89

Table II. III-PS and C3d binding to MZ B cells in immunized micea

% of Positive Cells

III-PS binding C3 binding

Experiment No. Genotype No Ag 30 min 120 min No Ag 30 min 120 min

1: C57BL6/129Sv WT 0.3 15.3 72.1 28.9 91.9 89.9 Cr2null 0.1 0.1 0.3 18.7 25.0 25.9

2: C57BL6/129Sv WT 0.3 17.3 67.2 15.5 87.0 81.7 C3null 0.1 0.2 0.2 0.1 0.1 0.1 Cr2null 0.3 0.1 0.1 8.3 9.4 16.9

3: C57BL6/129Sv WT 0.2 3.5 5.2 21.8 35.8 38.3 WT ND 6.1 10.3 ND 33.1 44.2 C3null 0.1 0.4 0.3 0.1 0.1 0.1 C3null ND 0.2 0.3 ND 0.1 0.1 Cr2null 0.2 0.1 0.2 15.6 20.7 16.8 Cr2null ND 0.2 0.2 ND 18.2 14.6

4: C57BL6 WT 0.4 3.7 9.8 ND ND ND Downloaded from C3null 0.2 0.2 0.1 ND ND ND Cr2null 0.1 0.2 0.3 ND ND ND

5: C57BL6 WT 0.3 4.9 12.3 ND ND ND C3null 0.2 0.6 0.4 ND ND ND Cr2null 0.2 0.2 0.1 ND ND ND

a Data represent the frequency (%) of MZ B cells binding III-PS or C3 based on FACS analysis. Splenic cells were harvested http://www.jimmunol.org/ 30 or 120 min after i.v. injection of biotinylated III-PS into immunized WT and C3null or Cr2null mice and analyzed. III-PS and C3 protein were detected on the surface of CD24ϩCD23Ϫ MZ B cells. A total of ﬁve different experiments were performed. III-PS binding was dependent on C3 and Cr2. The amount of Ag and C3d bound to MZ B cells increased with time. Differences in Ag and C3 binding between experiments can be explained in part by differences between lots of biotinylated III-PS. In addition, some background binding of C3 to other complement receptors is possible in Cr2null mice.

Whether MZ B cells are required for transferring III-PS to FDCs in WT mice, but only negligible uptake by MZ B cells in mice is undetermined from these results. One possible pathway is that deﬁcient in C3 or CD21/CD35. In contrast, we found that III-PS is

MZ B cells bind complexes of C3 and Ag and transport them into preferentially taken up by MZ macrophages in naive mice. The by guest on September 24, 2021 the follicular zone. mechanism of uptake is unclear but does not depend on comple- In a recent report, Guinamard et al. (13) observed that the hu- ment, as no differences in Ag binding were observed among WT, moral response to TNP-Ficoll was dependent on an intact MZ C3null, and Cr2null mice after i.v. injection of III-PS. One possible compartment and the presence of C3 and Cr2. Injection of naive explanation for the difference in the handling of the two Ags is that mice with TNP-Ficoll Ag resulted in rapid uptake by MZ B cells Ficoll, which is not a bacterial Ag, may directly activate complement, leading to covalent attachment of C3b and deposition on MZ B cells. Importantly, III-PS includes a terminal sialic group that can inhibit complement activation and, therefore, III-PS may be bound by carbohydrate recognition receptors (e.g., mannose receptors) expressed constitutively by MZ macrophages (30). In a previous study, Markham et al. (15) reported that transient depletion of C3 by CVF had no significant effect on the day 5 IgM response to III-PS in CFA injected s.c. Moreover, they reported that 5 days following immunization, specific IgM synthesis was not dependent on an intact spleen and that no III-PS specific splenic ASCs were identified. One explanation for the apparent differences between this earlier study and our current results is that depletion of circulating C3 with CVF is not likely to affect local production of C3 by myeloid cells within the lymphoid compartment. Recent studies have demonstrated that local C3 synthesis by myeloid cells within the spleen and peripheral lymph nodes is sufficient to enhance the humoral response to protein Ags irrespec- tive of the C3 levels in the blood (31, 32). FIGURE 4. Reconstitution of III-PS binding to MZ B cells in WT mice One model that could account for our current observations is by III-PS-specific IgG and IgM. FACS analysis was performed on splenic that phagocytic cells within the splenic MZ and other secondary cells harvested 2 h after i.v. injection of biotinylated III-PS into WT and lymphoid tissues bind Ag via carbohydrate recognition, such as C3null mice. Two hours before Ag injection, mice were given III-PS-specific IgG or IgM i.v. Immunized WT mice were used as controls. III-PS was de- mannose receptor or mannan binding protein, and transport it into tected on the surface of MZ B cells. Two separate different experiments were the B cell follicles, as proposed by Martinez-Pomares et al. (33). performed. Injection of III-PS-specific Abs into WT mice partially restored Alternatively, myeloid dendritic cells bind III-PS and “present” the their ability to bind III-PS. The binding was C3-dependent. Ag to B cells within the MZ, leading to plasmablast formation and 90 COMPLEMENT-DEPENDENT HUMORAL RESPONSE TO GBS POLYSACCHARIDE specific IgM release (34–36). In this scenario, the presence of independent type 2 antigens: modulation of the cells with specific antibody. Eur. specific Ab overcomes the inhibitory effect of sialic acid and ac- J. Immunol. 19:675. 13. Guinamard, R., M. Okigaki, J. Schlessinger, and J. V. Ravetch. 2000. Absence of tivates classical pathway complement, resulting in C3 attachment marginal zone B cells in Pyk-2-deficient mice defines their role in the humoral and more efficient uptake by MZ B cells. The high frequency with response. Nat. Immunol. 1:31. 14. Pepys, M. B. 1974. Role of complement in induction of antibody production in which MZ B cells bind C3-coated complexes of III-PS in WT vivo: effect of cobra factor and other C3-reactive agents on thymus-dependent immune mice suggests an efficient process of Ag localization to the and thymus-independent antibody responses. J. Exp. Med. 140:126. MZ compartment. MZ B cells express higher levels of CD21 than 15. Markham, R. B., A. Nicholson-Weller, G. Schiffman, and D. L. Kasper. 1982. The presence of sialic acid on two related bacterial polysaccharides determines follicular B cells and might be expected to preferentially bind C3d- the site of the primary immune response and the effect of complement depletion III-PS complexes. Because we find a high frequency of Ag-C3d- on the response in mice. J. Immunol. 128:2731. positive B cells, it is unlikely that only III-PS-specific B cells bind 16. Pryjma, J., and J. H. Humphrey. 1975. Prolonged C3 depletion by cobra venom factor in thymus-deprived mice and its implication for the role of C3 as an C3d complexes (Fig. 3j). However, efficient localization to this essential second signal for B-cell triggering. Immunology 28:569. compartment would enhance specific interaction with cognate B 17. Griffioen, A. W., G. T. Rijkers, P. Janssens-Korpela, and B. J. Zegers. 1991. Pneumococcal polysaccharides complexed with C3d bind to human B lympho- cells and recent studies have reported clonal expansion of Ag- cytes via complement receptor type 2. Infect. Immun. 59:1839. activated MZ B cells (29). An alternative mechanism for comple- 18. Bird, P., and P. J. Lachmann. 1988. The regulation of IgG subclass production in ment enhancement in the current model is that C3d-coated III-PS man: low serum IgG4 in inherited deficiencies of the classical pathway of C3 activation. Eur. J. Immunol. 18:1217. coligates the CD21/CD19/CD81 coreceptor and B cell receptor on 19. Wessels, M. R., L. C. Paoletti, D. L. Kasper, J. L. DiFabio, F. Michon, K. Holme, MZ B cells, thereby lowering the threshold for activation of B cells and H. J. Jennings. 1990. Immunogenicity in animals of a polysaccharide-protein in a manner similar to previous observations using protein Ags conjugate vaccine against type III group B Streptococcus. J. Clin. Invest. 86: 1428.

null Downloaded from (37). Recent work by Cariappa et al. (38) suggests that Cr2 20. Guttormsen, H. K., C. J. Baker, M. S. Edwards, L. C. Paoletti, and D. L. Kasper. mice have increased numbers of MZ B cells. The results reported 1996. Quantitative determination of antibodies to type III group B streptococcal polysaccharide. J. Infect. Dis. 173:142. herein demonstrate similar MZ B cell numbers in all experimental 21. Ahearn, J. M., M. B. Fischer, D. Croix, S. Goerg, M. Ma, J. Xia, X. Zhou, groups. Although the generation and localization of MZ B cells R. G. Howard, T. L. Rothstein, and M. C. Carroll. 1996. Disruption of the Cr2 appear to be independent of CD21/CD35, MZ B cells lacking locus results in a reduction in B-1a cells and in an impaired B cell response to T-dependent antigen. Immunity 4:251. CD21/CD35 clearly cannot capture complement-coated III-PS. 22. Wang, D., S. M. Wells, A. M. Stall, and E. A. Kabat. 1994. Reaction of germinal ␣ In conclusion, the data presented in this study show that TI centers in the T-cell-independent response to the bacterial polysaccharide (1- http://www.jimmunol.org/ Ͼ responses to III-PS are complement-dependent. They further sug- 6)dextran. Proc. Natl. Acad. Sci. USA 91:2502. 23. Sverremark, E., and C. Fernandez. 1998. Role of T cells and germinal center gest that complement-tagged III-PS is trapped via CD21/CD35 on formation in the generation of immune responses to the thymus-independent car- MZ B cells, which, in turn, transport the Ag to FDCs within GCs. bohydrate dextran B512. J. Immunol. 161:4646. 24. Wang, B., C. Biron, J. She, K. Higgins, M. J. Sunshine, E. Lacy, N. Lonberg, and The proximity of MZ B cells to marginal sinuses and the high- C. Terhorst. 1994. A block in both early T lymphocyte and natural killer cell level expression of CD21/CD35 by these cells appear to emphasize development in transgenic mice with high-copy numbers of the human CD3E the importance of their function as sentinel cells for recognizing gene. Proc. Natl. Acad. Sci. USA 91:9402. 25. Vos, Q., C. M. Snapper, and J. J. Mond. 1999. Heterogeneity in the ability of systemic bacterial infections. cytotoxic murine NK cell clones to enhance Ig secretion in vitro. Int. Immunol. 11:159.

26. Martin, F., A. M. Oliver, and J. F. Kearney. 2001. Marginal zone and B1 B cells by guest on September 24, 2021 Acknowledgments unite in the early response against T-independent blood-borne particulate anti- We thank Robert A. Barrington and Elahna Paul for comments on this gens. Immunity 14:617. manuscript, Barbara G. Reinap for purification and oxidation of the poly- 27. Molina, H., W. Wong, T. Kinoshita, C. Brenner, S. Foley, and V. M. Holers. 1992. Distinct receptor and regulatory properties of recombinant mouse comple- saccharide, and Valerie M. Brostrom for fractionation of mouse Abs. ment receptor 1 (CR1) and Crry, the two genetic homologues of human CR1. J. Exp. Med. 175:121. 28. Hasegawa, M., M. Fujimoto, J. C. Poe, D. A. Steeber, and T. F. Tedder. 2001. References CD19 can regulate B lymphocyte signal transduction independent of complement 1. Baker, C. J., M. A. Rench, M. S. Edwards, R. J. Carpenter, B. M. Hays, and activation. J. Immunol. 167:3190. D. L. Kasper. 1988. Immunization of pregnant women with a polysaccharide 29. Tierens, A., J. Delabie, L. Michiels, P. Vandenberghe, and C. De Wolf-Peeters. vaccine of group B Streptococcus. N. Engl. J. Med. 319:1180. 1999. Marginal-zone B cells in the human lymph node and spleen show somatic 2. Timens, W., A. Boes, T. Rozeboom-Uiterwijk, and S. Poppema. 1989. Immatu- hypermutations and display clonal expansion. Blood 93:226. rity of the human splenic marginal zone in infancy: possible contribution to the 30. Stahl, P. D., and R. A. Ezekowitz. 1998. The mannose receptor is a pattern deficient infant immune response. J. Immunol. 143:3200. recognition receptor involved in host defense. Curr. Opin. Immunol. 10:50. 3. van den Eertwegh, A. J., J. D. Laman, M. M. Schellekens, W. J. Boersma, and 31. Fischer, M. B., M. Ma, N. C. Hsu, and M. C. Carroll. 1998. Local synthesis of E. Claassen. 1992. Complement-mediated follicular localization of T-indepen- C3 within the splenic lymphoid compartment can reconstitute the impaired im- dent type-2 antigens: the role of marginal zone macrophages revisited. Eur. J. Im- mune response in C3-deficient mice. J. Immunol. 160:2619. munol. 22:719. 32. Verschoor, A., M. A. Brockman, D. M. Knipe, and M. C. Carroll. 2001. Cutting 4. Brown, A. R. 1992. Immunological functions of splenic B-lymphocytes. Crit. edge: myeloid complement c3 enhances the humoral response to peripheral viral Rev. Immunol. 11:395. infection. J. Immunol. 167:2446. 5. Morse, H. C., 3rd, J. F. Kearney, P. G. Isaacson, M. Carroll, T. N. Fredrickson, 33. Martinez-Pomares, L., M. Kosco-Vilbois, E. Darley, P. Tree, S. Herren, J. Bon- and E. S. Jaffe. 2001. Cells of the marginal zone-origins, function and neoplasia. nefoy, and S. Gordon. 1996. Fc chimeric protein containing the cysteine-rich Leuk. Res. 25:169. domain of the murine mannose receptor binds to macrophages from splenic mar- 6. Lane, P. J., D. Gray, S. Oldfield, and I. C. MacLennan. 1986. Differences in the ginal zone and lymph node subcapsular sinus and to germinal centers. J. Exp. recruitment of virgin B cells into antibody responses to thymus-dependent and Med. 184:1927. thymus-independent type-2 antigens. Eur. J. Immunol. 16:1569. 34. Colino, J., Y. Shen, and C. Snapper. 2002. Dendritic cells pulsed with intact 7. Timens, W. 1991. The human spleen and the immune system: not just another Streptococcus pneumoniae elicit both protein- and polysaccharide-specific im- lymphoid organ. Res. Immunol. 142:316. munoglobulin isotype responses in vivo through distinct mechanisms. J. Exp. 8. Lee, C. J., S. D. Banks, and J. P. Li. 1991. Virulence, immunity, and vaccine Med. 195:1. related to Streptococcus pneumoniae. Crit. Rev. Microbiol. 18:89. 35. Garcia De Vinuesa, C., A. Gulbranson-Judge, M. Khan, P. O’Leary, 9. Rickert, R. C., K. Rajewsky, and J. Roes. 1995. Impairment of T-cell-dependent M. Cascalho, M. Wabl, G. Klaus, M. Owen, and I. MacLennan. 1999. Dendritic B-cell responses and B-1 cell development in CD19-deficient mice. Nature 376: cells associated with plasmablast survival. Eur. J. Immunol. 29:3712. 352. 36. Balazs, M., F. Martin, T. Zhou, and J. Kearney. 2002. Blood dendritic cells 10. Sato, S., D. A. Steeber, and T. F. Tedder. 1995. The CD19 signal transduction interact with splenic marginal zone B cells to initiate T-independent immune molecule is a response regulator of B-lymphocyte differentiation. Proc. Natl. responses. Immunity 17:341. Acad. Sci. USA 92:11558. 37. Carter, R. H., and D. T. Fearon. 1992. CD19: lowering the threshold for antigen 11. Humphrey, J. H. 1985. Splenic macrophages: antigen presenting cells for T1–2 receptor stimulation of B lymphocytes. Science 256:105. antigens. Immunol. Lett. 11:149. 38. Cariappa, A., M. Tang, C. Parng, E. Nebelitskiy, M. Carroll, K. Georgopoulos, 12. Kraal, G., H. Ter Hart, C. Meelhuizen, G. Venneker, and E. Claassen. 1989. and S. Pillai. 2001. The follicular versus marginal zone B lymphocyte cell fate Marginal zone macrophages and their role in the immune response against T- decision is regulated by Aiolos, Btk, and CD21. Immunity 14:603.