Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates...

6
Polysaccharide structure dictates mechanism of adaptive immune response to glycoconjugate vaccines Ximei Sun a,b , Giuseppe Stefanetti a,c , Francesco Berti d , and Dennis L. Kasper a,1 a Department of Microbiology and Immunobiology, Harvard Medical School, Boston, MA 02115; b Graduate Program in Immunology, Harvard Medical School, Boston, MA 02115; c Department of Chemistry, University of Milan, 20133 Milan, Italy; and d Technical R&D, GSK Vaccines, 53100 Siena, Italy Contributed by Dennis L. Kasper, November 1, 2018 (sent for review September 25, 2018; reviewed by Peter R. Andreana and Moriya Tsuji) Glycoconjugate vaccines are among the most effective interven- tions for preventing several serious infectious diseases. Covalent linkage of the bacterial capsular polysaccharide to a carrier protein provides CD4 + T cells with epitopes that facilitate a memory re- sponse to the polysaccharide. Classically, the mechanism responsi- ble for antigen processing was thought to be similar to what was known for hapten-carrier conjugates: protease digestion of the carrier protein in the endosome and presentation of a resulting peptide to the T cell receptor on classical peptide-recognizing CD4 + T cells. Recently, an alternative mechanism has been shown to be responsible for the memory response to some glycoconjugates. Pro- cessing of both the protein and the polysaccharide creates glycopep- tides in the endosome of antigen-presenting cells. For presentation, the peptide portion of the glycopeptide is bound to MHCII, allowing the covalently linked glycan to activate carbohydrate-specific helper CD4 + T cells (Tcarbs). Herein, we assessed whether this same mech- anism applies to conjugates prepared from other capsular polysac- charides. All of the glycoconjugates tested induced Tcarb-dependent responses except that made with group C Neisseria meningitidis; in the latter case, only peptides generated from the carrier protein were critical for helper T cell recognition. Digestion of this acid- sensitive polysaccharide, a linear homopolymer of α(2 9)-linked sialic acid, to the size of the monomeric unit resulted in a dominant CD4 + T cell response to peptides in the context of MHCII. Our results show that different mechanisms of presentation, based on the struc- ture of the carbohydrate, are operative in response to different glycoconjugate vaccines. glycoconjugate | vaccine | Tcarb | antigen presentation | group C Neisseria meningitidis M any pathogenic bacteria express large-molecular-sized surface carbohydrates called capsular polysaccharides (CPSs). Beginning in the 1980s, CPSs of bacterial targets were coupled to carrier proteins to create effective glycoconjugate vaccines. These vaccines were more immunogenic than un- conjugated polysaccharides, especially in children under 2 y of age (1). Several glycoconjugate vaccines were developed and have played an enormous role in preventing infectious dis- eases caused by virulent pathogens, including vaccines against Haemophilus influenzae type b, Streptococcus pneumoniae, and Neisseria meningitidis (2, 3). On the basis of studies of hapten- carrier immune responses, it had been assumed that, in glyco- conjugate vaccines, the peptide processed from the carrier pro- tein is presented by the major histocompatibility class II (MHCII) molecule and that this signal plays a central role in activating CD4 + helper T cells. The peptide-recognizing T cells in turn help B cell maturation and the formation of immuno- logic memory (4). The original hypothesis was in part based on the failure of pure polysaccharides to elicit IgM-to-IgG class switching and substantial memory responses. Recently, we reported a different mechanism governing the immune responses to glycoconjugate vaccines, using the CPS of type III group B Streptococcus (GBSIII) as a model antigen. We showed that antigenic fragments of the polysaccharide are presented on the surface by MHCII molecules in the context of a covalently linked peptide from the carrier (5). These surface-presented CPSs are able to activate a subset of CD4 + T cells, designated carbohydrate- specific T cells (Tcarbs), which regulate the adaptive immune response to the GBSIII glycoconjugate (58). Similar mechanisms have now been shown to be responsible for T helper responses to glycoconjugates made with the type 3 S. pneumoniae poly- saccharide (Pn3P) (6). In the present study, we analyze the T cell response to gly- coconjugates of several other important pathogens, including conjugates made from Vi antigen of Salmonella Typhi (Vi), the CPS of type Ib group B streptococci (GBSIb), the CPS of H. influenzae type b (Hib), and the group C polysaccharide of N. meningitidis (MenC). We report that only MenC-specific IgG responses are not regulated by Tcarbs. This lack of Tcarb acti- vation is related to the structure of the polysaccharide. De- polymerization of MenC polysaccharide in the acidic environment of the endolysosome results in marked reduction in polysaccha- ride size to monomers, with a consequent failure to be recognized by T cells as an independent antigen. Given that there may be at least two mechanisms governing T cell responses to glyco- conjugates, a deeper understanding of factors influencing anti- gen processing and presentation as well as cooperation between T and B cells in response to glycoconjugate vaccination is a key factor to be considered in improving the design of the next- generation glycoconjugate vaccines. Significance Helper T cell responses to glycoconjugate vaccines are regu- lated through mechanisms dependent upon the structure of the polysaccharide. We show that three of the four important conjugate vaccines tested induced antibody responses regulated primarily by carbohydrate-recognizing helper T cells. However, the adaptive immune response to meningococcal group C (MenC) conjugate was restricted to peptide-recognizing helper T cells. We show that MenC is degraded to a monomeric sialic acid res- idue that cannot be recognized by T cell receptor as an inde- pendent antigen. The structure of the saccharide constitutes a critical factor in determining the processing and presentation of glycoconjugate vaccines. An understanding of the mechanisms underlying the immune responses to glycoconjugates will be crucial in the production of highly protective knowledge-based vaccines. Author contributions: X.S. and D.L.K. designed research; X.S. performed research; G.S. and F.B. contributed new reagents/analytic tools; X.S. and D.L.K. analyzed data; and X.S., G.S., F.B., and D.L.K. wrote the paper. Reviewers: P.R.A., University of Toledo; and M.T., Aaron Diamond AIDS Research Center. The authors declare no conflict of interest. Published under the PNAS license. See Commentary on page 14. 1 To whom correspondence should be addressed. Email: [email protected]. This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1816401115/-/DCSupplemental. Published online December 3, 2018. www.pnas.org/cgi/doi/10.1073/pnas.1816401115 PNAS | January 2, 2019 | vol. 116 | no. 1 | 193198 IMMUNOLOGY AND INFLAMMATION SEE COMMENTARY Downloaded by guest on June 18, 2020

Transcript of Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates...

Page 1: Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates mechanism of adaptive immune response to glycoconjugate vaccines Ximei Suna,b, Giuseppe

Polysaccharide structure dictates mechanism ofadaptive immune response to glycoconjugate vaccinesXimei Suna,b, Giuseppe Stefanettia,c, Francesco Bertid, and Dennis L. Kaspera,1

aDepartment of Microbiology and Immunobiology, Harvard Medical School, Boston, MA 02115; bGraduate Program in Immunology, Harvard MedicalSchool, Boston, MA 02115; cDepartment of Chemistry, University of Milan, 20133 Milan, Italy; and dTechnical R&D, GSK Vaccines, 53100 Siena, Italy

Contributed by Dennis L. Kasper, November 1, 2018 (sent for review September 25, 2018; reviewed by Peter R. Andreana and Moriya Tsuji)

Glycoconjugate vaccines are among the most effective interven-tions for preventing several serious infectious diseases. Covalentlinkage of the bacterial capsular polysaccharide to a carrier proteinprovides CD4+ T cells with epitopes that facilitate a memory re-sponse to the polysaccharide. Classically, the mechanism responsi-ble for antigen processing was thought to be similar to what wasknown for hapten-carrier conjugates: protease digestion of thecarrier protein in the endosome and presentation of a resultingpeptide to the T cell receptor on classical peptide-recognizing CD4+

T cells. Recently, an alternative mechanism has been shown to beresponsible for the memory response to some glycoconjugates. Pro-cessing of both the protein and the polysaccharide creates glycopep-tides in the endosome of antigen-presenting cells. For presentation,the peptide portion of the glycopeptide is bound to MHCII, allowingthe covalently linked glycan to activate carbohydrate-specific helperCD4+ T cells (Tcarbs). Herein, we assessed whether this same mech-anism applies to conjugates prepared from other capsular polysac-charides. All of the glycoconjugates tested induced Tcarb-dependentresponses except that made with group C Neisseria meningitidis; inthe latter case, only peptides generated from the carrier proteinwere critical for helper T cell recognition. Digestion of this acid-sensitive polysaccharide, a linear homopolymer of α(2 → 9)-linkedsialic acid, to the size of the monomeric unit resulted in a dominantCD4+ T cell response to peptides in the context of MHCII. Our resultsshow that different mechanisms of presentation, based on the struc-ture of the carbohydrate, are operative in response to differentglycoconjugate vaccines.

glycoconjugate | vaccine | Tcarb | antigen presentation | group C Neisseriameningitidis

Many pathogenic bacteria express large-molecular-sizedsurface carbohydrates called capsular polysaccharides

(CPSs). Beginning in the 1980s, CPSs of bacterial targets werecoupled to carrier proteins to create effective glycoconjugatevaccines. These vaccines were more immunogenic than un-conjugated polysaccharides, especially in children under 2 y ofage (1). Several glycoconjugate vaccines were developed andhave played an enormous role in preventing infectious dis-eases caused by virulent pathogens, including vaccines againstHaemophilus influenzae type b, Streptococcus pneumoniae, andNeisseria meningitidis (2, 3). On the basis of studies of hapten-carrier immune responses, it had been assumed that, in glyco-conjugate vaccines, the peptide processed from the carrier pro-tein is presented by the major histocompatibility class II(MHCII) molecule and that this signal plays a central role inactivating CD4+ helper T cells. The peptide-recognizing T cellsin turn help B cell maturation and the formation of immuno-logic memory (4). The original hypothesis was in part basedon the failure of pure polysaccharides to elicit IgM-to-IgGclass switching and substantial memory responses. Recently, wereported a different mechanism governing the immune responsesto glycoconjugate vaccines, using the CPS of type III group BStreptococcus (GBSIII) as a model antigen. We showed thatantigenic fragments of the polysaccharide are presented on thesurface by MHCII molecules in the context of a covalently linked

peptide from the carrier (5). These surface-presented CPSs areable to activate a subset of CD4+ T cells, designated carbohydrate-specific T cells (Tcarbs), which regulate the adaptive immuneresponse to the GBSIII glycoconjugate (5–8). Similar mechanismshave now been shown to be responsible for T helper responses toglycoconjugates made with the type 3 S. pneumoniae poly-saccharide (Pn3P) (6).In the present study, we analyze the T cell response to gly-

coconjugates of several other important pathogens, includingconjugates made from Vi antigen of Salmonella Typhi (Vi), theCPS of type Ib group B streptococci (GBSIb), the CPS of H.influenzae type b (Hib), and the group C polysaccharide of N.meningitidis (MenC). We report that only MenC-specific IgGresponses are not regulated by Tcarbs. This lack of Tcarb acti-vation is related to the structure of the polysaccharide. De-polymerization of MenC polysaccharide in the acidic environmentof the endolysosome results in marked reduction in polysaccha-ride size to monomers, with a consequent failure to be recognizedby T cells as an independent antigen. Given that there may beat least two mechanisms governing T cell responses to glyco-conjugates, a deeper understanding of factors influencing anti-gen processing and presentation as well as cooperation betweenT and B cells in response to glycoconjugate vaccination is a keyfactor to be considered in improving the design of the next-generation glycoconjugate vaccines.

Significance

Helper T cell responses to glycoconjugate vaccines are regu-lated through mechanisms dependent upon the structure ofthe polysaccharide. We show that three of the four importantconjugate vaccines tested induced antibody responses regulatedprimarily by carbohydrate-recognizing helper T cells. However,the adaptive immune response to meningococcal group C (MenC)conjugate was restricted to peptide-recognizing helper T cells.We show that MenC is degraded to a monomeric sialic acid res-idue that cannot be recognized by T cell receptor as an inde-pendent antigen. The structure of the saccharide constitutes acritical factor in determining the processing and presentation ofglycoconjugate vaccines. An understanding of the mechanismsunderlying the immune responses to glycoconjugates will becrucial in the production of highly protective knowledge-basedvaccines.

Author contributions: X.S. and D.L.K. designed research; X.S. performed research; G.S. andF.B. contributed new reagents/analytic tools; X.S. and D.L.K. analyzed data; and X.S., G.S.,F.B., and D.L.K. wrote the paper.

Reviewers: P.R.A., University of Toledo; and M.T., Aaron Diamond AIDS Research Center.

The authors declare no conflict of interest.

Published under the PNAS license.

See Commentary on page 14.1To whom correspondence should be addressed. Email: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1816401115/-/DCSupplemental.

Published online December 3, 2018.

www.pnas.org/cgi/doi/10.1073/pnas.1816401115 PNAS | January 2, 2019 | vol. 116 | no. 1 | 193–198

IMMUNOLO

GYAND

INFLAMMATION

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

June

18,

202

0

Page 2: Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates mechanism of adaptive immune response to glycoconjugate vaccines Ximei Suna,b, Giuseppe

ResultsVarious Glycoconjugates Induce T Helper Cells Recognizing DifferentEpitopes. We had previously shown that a clear biomarker forTcarb-dependent responses was a polysaccharide-specific anti-body response after priming with a polysaccharide covalentlylinked to a carrier protein and boosting with the same poly-saccharide linked to a different and unrelated carrier protein. Toinvestigate the involvement of Tcarbs in the humoral immuneresponse to different glycoconjugates, we performed priming andboosting immunization experiments with glycoconjugates madewith Vi, GBSIb, Hib, and MenC CPSs. BALB/c mice wereprimed at the beginning of the experiment and boosted 14 d laterwith different antigen combinations. One week after the boost,serum levels of polysaccharide-specific IgG were determined.For Vi, GBSIb, and Hib glycoconjugates, boosting with a gly-coconjugate containing the same polysaccharide but a heterolo-gous carrier protein induced polysaccharide-specific IgG titers ofthe same magnitude as those seen after priming and boosting withglycoconjugates containing the same carrier (Fig. 1A). As reportedwith GBSIII and Pn3P (5, 6), this result supports a Tcarb-dependent mechanism for all three of these glycoconjugates.In marked contrast to the above results, priming and boosting

with MenC glycoconjugates containing heterologous carrierproteins induced significantly lower levels of MenC-specific IgGthan did primary and secondary immunization with either aMenC-OVA (ovalbumin) conjugate or a MenC-CRM197 (non-toxic mutant of diphtheria toxin) conjugate. This result wasconfirmed with a combination of different carriers, includingMenC-TT (tetanus toxoid), MenC-HEL (hen egg lysozyme),MenC-OVA, and MenC-CRM197 (Fig. 1B). We also found thatMenC-specific IgM levels were similar in groups immunized withdifferent conjugate combinations, a result suggesting a similarlevel of B cell activation independent of the carrier (SI Ap-pendix, Fig. S1A). To exclude the possibility that we were seeinga T cell-independent response, we treated mice with CD4-specificblocking antibody during the interval between priming andboosting with MenC-CRM197. The excellent booster IgG responseobserved in mice treated with isotype control antibody was sig-nificantly reduced in mice treated with anti-CD4 (SI Appendix,Fig. S1B). The failure of MenC conjugates prepared with differentcarrier proteins to induce a booster IgG response suggested thatperhaps these glycoconjugates were not able to induce Tcarbs andthat Tcarbs were not essential in regulating MenC polysaccharide-specific IgG responses.

Carrier-Specific T Cell-Mediated Adaptive Immune Response to MenCConjugates.On the basis of the previous results, we theorized thatcarrier protein/peptide-specific CD4+ T cells may be the onlyT cell subset essential to regulate the response to MenC conju-gate vaccines (9, 10). To further test this hypothesis, we primedmice with a combination of MenC-CRM197 plus unconjugatedOVA protein, and we boosted these mice with MenC-OVA. Wehypothesized that priming with this combination would yieldMenC-specific memory B cells resulting from MenC-CRM197 aswell as OVA-specific memory helper T cells generated from theOVA protein component. Interestingly, after boosting withMenC-OVA, mice primed with the mixture of MenC-CRM197and OVA protein had MenC IgG levels similar to those in miceprimed and boosted with MenC-OVA. The MenC IgG levelswere significantly lower in mice primed with MenC-CRM197without OVA (Fig. 2A and SI Appendix, Fig. S2 A and B). Thisresult suggests that in MenC glycoconjugates carrier protein-activated T cells are essential in inducing a humoral immuneresponse and antibody class switching.To further explore the requirements for cooperation between

B and T cells, we performed another series of immunizationexperiments. We first primed mice with MenC-CRM197 and then

boosted them with either MenC-CRM197 or MenC and CRM197protein (either alone or physically mixed but not conjugated).Booster IgG responses occurred only in mice that received the

Fig. 1. Requirement for helper T cell memory induction by either carbo-hydrate or peptide in several glycoconjugate vaccines. BALB/c mice wereprimed (day 0) and boosted (day 14) with different glycoconjugate combi-nations. Polysaccharide-specific IgG antibodies were measured by ELISA inserum obtained on day 21. (A) Tcarb-dependent IgG antibody induction toVi, GBSIb, and Hib CPSs. Serum titers are reported as the reciprocal dilutionthat results in an OD of 0.5 at 405 nm. (B) Tcarb-independent IgG antibodyinduction to MenC. MenC-specific IgG monoclonal antibody was used as astandard to determine the MenC IgG concentrations. n = 4–6 mice pergroup. All data are expressed as mean ± SEM values. *P ≤ 0.05; **P ≤ 0.01;***P ≤ 0.001; ****P ≤ 0.0001; ns, not significant.

194 | www.pnas.org/cgi/doi/10.1073/pnas.1816401115 Sun et al.

Dow

nloa

ded

by g

uest

on

June

18,

202

0

Page 3: Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates mechanism of adaptive immune response to glycoconjugate vaccines Ximei Suna,b, Giuseppe

MenC-CRM197 conjugate for both primary and secondary im-munization. Similarly, priming of mice either with unconjugatedMenC or CRM197 protein alone or with a mixture of unconjugatedMenC and CRM197 protein did not support a robust secondaryIgG response upon boosting with the MenC-CRM197 conjugate(SI Appendix, Fig. S2C). This result suggested that a conjugate isrequired to stimulate MenC-specific memory B cells.To directly examine the contribution of carrier-specific T cells

to MenC-specific booster IgG responses, we performed adoptivetransfer experiments (Fig. 2B). Donor groups of BALB/c micewere immunized with either MenC-CRM197 or OVA, andsplenic and lymph node B and CD4+ T cells from each groupwere purified. Three groups of recipient mice all received B cellsfrom the MenC-CRM19–immunized donors. Groups A and Balso received CD4+ T cells from MenC-CRM197–immunizeddonors, while group C received CD4+ T cells from OVA-immunized donors. One day after adoptive transfer, group Arecipient mice were actively immunized with MenC-CRM197,

and groups B and C were immunized with MenC-OVA. Theresulting IgG and IgM antibody titers in the three groups areshown in Fig. 2C. Groups A and C had significantly higherMenC-specific IgG titers than group B, while MenC-specific IgMlevels were similar in the three groups. In a previously reportedexperiment supporting the role of Tcarbs in the immune re-sponse to glycoconjugates (6), Middleton et al. adoptively trans-ferred B cells and CD4+ T cells from mice immunized withPn3P-KLH (keyhole limpet hemocyanin) to recipient mice andthen immunized the recipient mice with either Pn3P-KLH orPn3P-OVA. The similar IgG levels in these two groups sup-ported a Tcarb-mediated response. In contrast to these resultswith the Pn3P conjugate, our observations with MenC glyco-conjugates suggested that a mechanism in which carrier protein/peptides are essential in recruiting T cell help is operative.

Conjugation Chemistry Is Not the Critical Factor in the Induction ofTcarbs. The structures of the GBSIII and MenC glycoconjugateswere very different, though in both the polysaccharides wereconjugated to carrier proteins through reductive amination.Whereas the GBSIII conjugate was cross-linked by oxidativecreation of aldehydes on a fraction (∼30%) of the side-chainterminal sialic acid residues of the repeating unit, the MenCconjugate was linked at a single site on the reducing end of thepolysaccharide (Fig. 3A). In addition, MenC depolymerization bymild acid hydrolysis and oxidation resulted in a size of 10–30 kDa, while GBSIII used for conjugation maintained its orig-inal size (>100 kDa). Thus, we hypothesized that conjugationchemistry might affect T cell recognition of carbohydrates. Toincrease the resemblance to the structure of a cross-linked gly-coconjugate, we conjugated MenC through the carboxylic acid to

Fig. 2. Carrier-specific T cells are essential in regulating the adaptive im-mune response to MenC conjugate. (A) BALB/c mice were primed (day 0)with MenC-CRM197 either alone or with unconjugated OVA protein, and allgroups were boosted (day 14) with MenC-OVA. MenC-specific IgG antibodieswere measured by ELISA in serum obtained on day 21. (B) Design of theadoptive transfer experiments. Recipient mice received B cells from MenC-CRM197-immunized mice and CD4+ T cells from mice immunized with eitherMenC-CRM197 (groups A and B) or OVA (group C). One day after adoptivetransfer, recipient mice were immunized with either MenC-CRM197 (group A)or MenC-OVA (groups B and C). (C) Concentrations of MenC-specific IgG andIgM in sera from recipient mice, measured by ELISA 7 d after immunization.n = 4 or 5 mice per group. All data are expressed as mean ± SEM values.****P ≤ 0.0001; ns, not significant.

Fig. 3. Effect of conjugate chemistry of MenC on selection of T helper cells.(A) Schematic representation of the structure of end-conjugated (Upper)and cross-linked (Lower) MenC glycoconjugates made by reductive amina-tion (Upper) and carbodiimide reaction (Lower), respectively. (B) Superdex200 elution profile of MenC-OVA from reductive amination (Left) or carbo-diimide reaction (Right). (C) Concentration of IgG antibody to MenC in BALB/cmice primed (day 0) and boosted (day 14) with cross-linked MenC glyco-conjugates containing either the same or a heterologous carrier protein, asmeasured by ELISA in serum obtained on day 21. n = 4 mice per group. Dataare expressed as mean ± SEM values. **P ≤ 0.01; ****P ≤ 0.0001.

Sun et al. PNAS | January 2, 2019 | vol. 116 | no. 1 | 195

IMMUNOLO

GYAND

INFLAMMATION

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

June

18,

202

0

Page 4: Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates mechanism of adaptive immune response to glycoconjugate vaccines Ximei Suna,b, Giuseppe

derivatized carrier protein treated with adipic acid dihydrazide(ADH) linker (SI Appendix, Fig. S3) (11). This method activateda fraction of repeating units along the chain of MenC withoutreducing its molecular size (>100 kDa). The cross-linked MenCconjugate was significantly larger than the end-linked MenCconjugate, as shown by size exclusion chromatography (Fig. 3B).We performed priming and boosting immunization experi-

ments with the cross-linked MenC glycoconjugates containingeither the same or different carrier proteins. Mice primed andboosted with MenC glycoconjugates containing the same carrierprotein—either MenC-ADH-CRM197 or MenC-ADH-TT—hadstrong booster MenC-specific IgG responses. However, primingand boosting with MenC conjugates containing different carrierproteins induced significantly lower titers of MenC-specific IgG(Fig. 3C). This result was similar to that observed with MenClinked at a single reducing end of the polysaccharide to either thesame or different proteins (Fig. 1B). Therefore, conjugationchemistry does not seem to explain the failure of MenC glyco-conjugates to induce Tcarbs.

Processing and Presentation of MenC Polysaccharide. We previouslyreported that, following immunization with a glycoconjugatevaccine, a depolymerized form of GBSIII (∼10 kDa) bound to anMHCII binding peptide can be presented and the carbohydraterecognized by the T cell receptor (TCR) (5). T cell recognition ofcarbohydrates was also reported for zwitterionic polysaccharides(12), which presumably bind directly to MHCII by electrostaticinteractions. In this study, we speculated that MenC poly-saccharide might fail to be presented in the context of MHCII onthe antigen-presenting cell (APC) surface. One possible un-derlying reason could be the sensitivity to acidic hydrolysis ofMenC, which is a linear homopolymer of α(2 → 9)-linked sialicacid. It is possible that the ketosidic linkage between sialic acidrepeating units on MenC is hydrolyzed to an extremely small sizein the endolysosome and therefore cannot be presented to orrecognized by TCRs.We conducted a flow cytometry analysis to assess this possibility

by determining whether antigenically active MenC carbohydratescan be presented on the APC surface. Bone marrow-deriveddendritic cells (BMDCs) from wild-type mice were incubated withunconjugated or conjugated GBSIII or MenC for 18 h. The cellswere collected and stained with a monoclonal antibody specific forGBSIII or MenC at 4 °C and then incubated with a fluorophore-conjugated secondary antibody. Consistent with previously pub-lished results, the control saccharide (GBSIII) was detected on thesurface of the cells incubated with the GBSIII conjugate. How-ever, no BMDC surface presentation of MenC was detected afterincubation of the cells either with unconjugated MenC or withconjugated MenC-TT or MenC-CRM197 (Fig. 4A). Thus, pro-cessed MenC carbohydrate was not loaded onto the APC surface,was altered structurally, or was depolymerized to a very small size.Any of these possibilities might explain why the MenC antigenicepitope could not be detected on the surface of the APC by themonoclonal antibody and why it could not be recognized by theTCR, with consequent Tcarb induction.We then examined MenC polysaccharide processing in APCs.

CPSs are taken up into the APC endosome and are depoly-merized by oxidative agents such as reactive oxygen species andreactive nitrogen species (5, 12, 13) into smaller carbohydrates,generally ∼10 kDa in size. One possible explanation for lack ofMenC detection on the APC surface is that MenC is digestedthrough oxidative depolymerization or acidic hydrolysis to anextremely small size so that it cannot be recognized by the TCR.Accordingly, we assessed the size of the MenC saccharide withinthe APC endolysosome. We incubated radioactively labeledMenC {[3H]-MenC} with Raji B cells for 18 h, then isolated andlysed the endolysosome.We ran the lysate on a size-exclusion HPLCcolumn and found that MenC was substantially depolymerized to

a size similar to that of a monomer of sialic acid (Fig. 4B). Weperformed parallel experiments with a control polysaccharide—polysaccharide A (PSA) from Bacteroides fragilis—that isknown to be processed to a size of ∼15 kDa (12). We comparedthe molecular size of the endosomally depolymerized MenC tothat of similarly processed [3H]-PSA on Aquagel-OH 20 (SI Ap-pendix, Fig. S4) and found that the size of the processed saccha-rides differed substantially: PSA was at least 7 kDa, and MenCappeared to be much smaller at ∼200 Da. To further define themechanisms involved in the depolymerization of MenC, weincubated Raji B cells with [3H]-MenC in the presence of eithera reactive oxygen species inhibitor (4-OH TEMPO) or anATPase proton pump inhibitor (bafilomycin A1). Interestingly,

Fig. 4. Processing and presentation of MenC. (A) Flow cytometry analysis ofBMDCs after incubation (18 h) with unconjugated GBSIII or GBSIII-TT (Left) orwith unconjugated MenC, MenC-TT, or MenC-CRM197 (Right) followed bysurface staining with monoclonal antibody to GBSIII or MenC. (B) Elutionprofile of lysates of Raji B cell endolysosomes after 18 h of incubation with[3H]-MenC on a size exclusion HPLC column: ProSEC 300S (MW range, 1,500–800,000) or Aquagel-OH 20 (MW range, 100–20,000). (C) Raji B cells weretreated with either 4-OH TEMPO (a superoxide inhibitor) or bafilomycin A1(BFA, an endosomal acidification inhibitor) for 1 h before incubation with[3H]-MenC. The molecular size distributions of the endolysosomal lysateswere analyzed on a size exclusion column (ProSEC 300S). SA, sialic acid.

196 | www.pnas.org/cgi/doi/10.1073/pnas.1816401115 Sun et al.

Dow

nloa

ded

by g

uest

on

June

18,

202

0

Page 5: Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates mechanism of adaptive immune response to glycoconjugate vaccines Ximei Suna,b, Giuseppe

less depolymerization of the polysaccharide was detected in theendolysosome after incubation with either inhibitor. In bafilomycin-treated cells, a small peak was seen at the void volume of thecolumn that was not observed with 4-OH TEMPO treatment.When treated with a combination of the two inhibitors, a sim-ilar upward size shift was observed (Fig. 4C). This result sug-gested that both oxidative depolymerization and acidic hydrolysiscontribute to MenC depolymerization, with acid hydrolysis per-haps being somewhat more important.

DiscussionAntibodies to CPSs mediate protection against encapsulatedbacteria (2). Several highly effective glycoconjugate vaccineshave been created using a hapten-carrier protein conjugationstrategy (14, 15). It has become standard practice to couple CPSsfrom bacterial targets to T cell-dependent carrier proteins tocreate glycoconjugate vaccines (16–19). Immunization with gly-coconjugates, as opposed to pure polysaccharides, elicits T cellhelp for B cells that produce IgG antibodies to the polysaccharidecomponent (2, 4). In addition to inducing polysaccharide-specificIgM-to-IgG switching, glycoconjugate immunization elicits both Bcell and T cell memory responses (2).Other than polysaccharides with a zwitterionic charge motif

(e.g., PSA of B. fragilis or the polysaccharide of type 1 S. pneu-moniae), which bind directly to MHCII through electrostatic in-teractions, most bacterial polysaccharides fail to bind to MHCIIand therefore are not presented to the TCR. As a result, mostpure polysaccharides induce immune responses that are T cellindependent. The traditional explanation for the mechanism bywhich glycoconjugates induce humoral immune responses is thatthe carrier protein portion of the conjugate activates CD4+ T cellsto help carbohydrate-specific B cells produce long-lasting IgGantibodies through both cognate and cytokine-mediated interac-tions (2, 4). The classical hypothesis of immune activation byglycoconjugate vaccines suggests that only peptides generatedfrom the polysaccharide-linked carrier protein can be presented toand recognized by T cells. This view, however, ignores the syn-thetic linkage of carbohydrates to proteins by strong covalentbonds that are unlikely to be broken within the endosome. Thus,we previously raised the possibility of glycopeptide presentation toT cells. We considered whether T cells could recognize carbohy-drates linked to another molecule (e.g., a peptide) whose bindingto MHCII allows carbohydrate presentation on the APC surface.Our earlier work uncovered a key feature of the cellular and

molecular mechanisms underlying adaptive immune responsesmediated by some glycoconjugate vaccines (5). We showed thatglycoconjugate immunization induces CD4+ T cells, designatedTcarbs, that recognize only the carbohydrate portion of theglycoconjugate vaccine. Upon endosomal uptake by APCs, aGBSIII glycoconjugate undergoes depolymerization, yielding aglycan of reduced size (∼10 kDa) that is chemically bound to apeptide (glycanp-peptide) fragment. Glycanp-peptide is displayedon the surface of APCs in the context of MHCII to the CD4+

T cells. We successfully generated T cell clones and validated theexistence of T cells that recognize only the processed carbohy-drate portion of the glycoconjugate—i.e., Tcarbs (8). Thesefindings suggested that Tcarbs contribute to the protection in-duced by the GBSIII glycoconjugate vaccine. Similar mecha-nisms involving Tcarb responses were recently reported for type3 S. pneumoniae glycoconjugates (6).In the present study, we sought to determine whether the

mechanisms involved in processing and presenting GBSIII andPn3P glycoconjugates are also present for glycoconjugates ofother bacterial polysaccharides. We show that a Tcarb-dependentresponse is induced by glycoconjugates made with the Vi poly-saccharide of Salmonella Typhi, the type b polysaccharide of H.influenzae, and the type Ib polysaccharide of group B Streptococ-cus. In contrast, we found that a glycoconjugate made with the

group C polysaccharide of N. meningitidis induces carrier-specifichelper T cells, not Tcarbs. Active immunization with variousvaccine constructs and adoptive transfer experiments clearlyshowed that carrier peptide-specific CD4+ T cells are sufficient toinduce adaptive immune antibody responses to the MenC conju-gate. Given that the covalent linkage between the linking sialicacid residue and the lysine group on the protein generated duringchemical conjugation would not be broken down in the endoly-sosome, it is likely that some processed sugars fromMenC are stillpresented on the surface along with the conjugated peptide.However, our findings indicate that these sugars do not constitutean antigenic epitope and do not sufficiently mask the peptide inthe MHCII binding cleft to prevent its recognition; thus they failto induce Tcarb helper responses. We did show that the MenCpolysaccharide is digested to a reduced size comparable to that ofa single sialic acid residue in the endosome. Although these sialicacids linked to the MHCII binding peptide are too small to fit inthe TCR pocket independent of the peptide, the attachment ofthe monosaccharide could modify the TCR specificity to thepeptide. Detailed studies on CD4+ T cell recognition of glyco-peptides showed that the TCR makes specific contact with boththe sugar moiety and peptide residues (20–22). In the case of theMenC conjugate, it is possible that peptide modified with a singlesialic acid might initiate the activation of a novel T cell subset thatcollaborates as helper T cells with the dominant peptide-specificT cells.N. meningitidis is a major cause of bacterial meningitis

worldwide, especially in the African meningitis belt, and has ahigh associated mortality. MenC conjugate vaccine has beenshown to be safe and immunogenic and to be capable of priminginfants, toddlers, young children, and adults for immunologicmemory (23–26). However, preliminary surveillance data inEngland and Wales suggest a waning of effectiveness from 1 yafter three-dose priming in infancy (27). A better understandingof the mechanisms involved in the immune response to MenCglycoconjugate immunization could lead to better vaccines withimproved efficacy. Clearly, more than one mechanism is responsi-ble for the induction of immune responses to glycoconjugates.It is reasonable to assume that glycoconjugate vaccine design andscheduling might be optimized if we had a fuller understandingof the mechanisms underlying this “decision-making” process. Weshow that, to induce more sustainable memory responses withMenC glycoconjugates, booster immunization is more effectivewhen the polysaccharide is linked to the same carrier proteinrather than to a heterologous protein. Our findings support arecent controlled trial conducted in the United Kingdom andMalta, wherein boosting children with Hib-MenC-TT vaccine af-ter priming them with a single MenC-TT dose in infancy resultedin a more robust bactericidal antibody response than boosting withMenC-CRM197; the response elicited by the MenC boost with thesame carrier protein used for priming resulted in antibodies per-sisting at 24 mo of age (28).Overall, our study suggests that different mechanisms are in-

volved in immune responses to immunization with differentglycoconjugates and that the structure of the polysaccharide iscritical to the mechanisms used by APCs to present antigen toCD4+ T cells. An understanding of these differences is an im-portant factor in the specific design of each glycoconjugate andoptimization of the vaccination schedule.

Materials and MethodsMice. Six-week-old female BALB/c mice were purchased from Taconic Bio-sciences. All mouse experiments were approved by the Harvard Medical AreaStanding Committee on Animals (Animal Protocol Is00000636).

Antigens. Purified MenC, the MenC-CRM197 conjugate, Vi polysaccharide, theVi-CRM197 conjugate, Hib polysaccharide, the Hib-CRM197 conjugate, andCRM197 protein were obtained from GSK Vaccines and GSK Vaccines Institute

Sun et al. PNAS | January 2, 2019 | vol. 116 | no. 1 | 197

IMMUNOLO

GYAND

INFLAMMATION

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

June

18,

202

0

Page 6: Polysaccharide structure dictates mechanism of adaptive ... · Polysaccharide structure dictates mechanism of adaptive immune response to glycoconjugate vaccines Ximei Suna,b, Giuseppe

for Global Health. The Hib-OMPC (outer membrane protein complex) con-jugate was purchased from Merck. GBSIb was isolated and purified fromtype Ib group B Streptococcus. Two conjugation methods were applied:cross-linked GBSIb and end-linked MenC conjugates were made throughreductive amination, as previously described (8); cross-linked MenC and Viconjugates were made using EDC [1-ethyl-3-(3-dimethylaminopropyl)carbodiimide], with an ADH linker to couple the carboxylic acid groupsfrom the protein and the carbohydrate.

Immunizations and Antibody Responses. Groups of BALB/c mice were primedon day 0 and boosted on day 14 by i.p. injection of the antigen of interest (4–6 μg as polysaccharide content) in PBS mixed with 0.5 mg of alum hydroxidegel adjuvant. At least four mice were immunized in each experiment. Micewere bled from the tail vein 1 wk after boosting immunization. Levels ofcarbohydrate-specific antibodies in the serum were determined by solid-phase ELISA, as previously described (29).

Adoptive Transfer. Groups of donor BALB/c mice were primed and boostedwith 4 μg of MenC-CRM197 as saccharide content given i.p. at 3-wk intervals.Mice were killed 5 d after boosting immunization. CD4+ T cells were isolatedfrom spleens and lymph nodes of mice immunized with either MenC-CRM197

or OVA and negatively selected with a mouse CD4+ T cell isolation kit (130-104-454, Miltenyi Biotech). B cells from mice immunized with MenC-CRM197

were isolated with a mouse B cell isolation kit (130-104-443, Miltenyi Bio-tech). CD4+ T cells (107) from mice immunized with MenC-CRM197 or OVAand B cells (107) from MenC-CRM197–immunized donors were adoptivelytransferred to recipient mice. The recipient mice were immunized 1 d afteradoptive transfer with MenC-CRM197 or MenC-OVA.

Antigen Presentation by BMDCs. BMDCs from wild-type mice were incubatedwith antigen (GBSIII or the GBSIII-TT conjugate, MenC or theMenC-CRM197 or

MenC-TT conjugate) for 18 h at 37 °C. After incubation, cells were collected,washed five times with PBS, and labeled at 4 °C first with either GBSIII-specific monoclonal antibody or MenC-specific monoclonal antibody andthen with a fluorophore-labeled secondary antibody. Surface staining wasassessed by flow cytometry (MACSQuant Analyzer).

Cell Fractionation and in Vitro Processing Assays. Raji B cells (108) were cul-tured in the presence of 1 mg of [3H]-MenC for 18 h at 37 °C. Cells were thenwashed five times with PBS to remove unreacted [3H]-MenC. Cell lysis wasperformed by passage of cells through a 27-gauge needle in 250 mM sucrosewith 10 mM Tris·HCl, pH 7.5. Differential centrifugation was then used tofractionate the lysed cells into endolysosome and cell-membrane fractions,as previously described (5, 12, 30). Endolysosomal fractions were solubilizedby boiling for 20 min in 1% SDS and analyzed by size exclusion chroma-tography on either Agilent ProSEC 300S [MW (molecular weight) range,1,500–800,000] or Aquagel-OH 20 (MW range, 100–20,000) with an UltiMate3000 system.

Statistical Analysis. Statistical significance was determined with the ordinaryone-way ANOVA, using GraphPad Prism 7.0c. Data with P values of ≤0.05were considered statistically significant (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001;****P ≤ 0.0001).

ACKNOWLEDGMENTS. We thank Dr. Lok-To Sham and Dr. Thomas G.Bernhardt for assistance with the HPLC system and GSK Vaccines and GSKVaccines Institute for Global Health, both part of a group of companies, forkindly providing several important reagents. This work was supported byGrants 5R01AI089915 and 5U19AI109764 from the National Institute ofAllergy and Infectious Diseases and by funding from the European Union’sHorizon 2020 Research and Innovation Programme under Marie SkłodowskaCurie Grant Agreement 661138.

1. Astronomo RD, Burton DR (2010) Carbohydrate vaccines: Developing sweet solutionsto sticky situations? Nat Rev Drug Discov 9:308–324.

2. Avci FY, Kasper DL (2010) How bacterial carbohydrates influence the adaptive im-mune system. Annu Rev Immunol 28:107–130.

3. Kelly DF, Pollard AJ, Moxon ER (2005) Immunological memory: The role of B cells inlong-term protection against invasive bacterial pathogens. JAMA 294:3019–3023.

4. Mitchison NA (2004) T-cell-B-cell cooperation. Nat Rev Immunol 4:308–312.5. Avci FY, Li X, Tsuji M, Kasper DL (2011) A mechanism for glycoconjugate vaccine ac-

tivation of the adaptive immune system and its implications for vaccine design. NatMed 17:1602–1609.

6. Middleton DR, Sun L, Paschall AV, Avci FY (2017) T cell-mediated humoral immuneresponses to type 3 capsular polysaccharide of Streptococcus pneumoniae. J Immunol199:598–603.

7. Lai Z, Schreiber JR (2009) Antigen processing of glycoconjugate vaccines; the poly-saccharide portion of the pneumococcal CRM(197) conjugate vaccine co-localizes withMHC II on the antigen processing cell surface. Vaccine 27:3137–3144.

8. Avci FY, Li X, Tsuji M, Kasper DL (2012) Isolation of carbohydrate-specific CD4(+) T cellclones from mice after stimulation by two model glycoconjugate vaccines. Nat Protoc7:2180–2192.

9. Katz DH, Paul WE, Goidl EA, Benacerraf B (1970) Carrier function in anti-hapten im-mune responses. I. Enhancement of primary and secondary anti-hapten antibodyresponses by carrier preimmunization. J Exp Med 132:261–282.

10. Paul WE, Katz DH, Goidl EA, Benacerraf B (1970) Carrier function in anti-haptenimmune responses. II. Specific properties of carrier cells capable of enhancing anti-hapten antibody responses. J Exp Med 132:283–299.

11. Zou W, Jennings HJ (2009) Preparation of glycoconjugate vaccines. Carbohydrate-Based Vaccines and Immunotherapies, eds Guo Z, Boons G-J (John Wiley & Sons,Hoboken, NJ), pp 55–88.

12. Cobb BA, Wang Q, Tzianabos AO, Kasper DL (2004) Polysaccharide processing andpresentation by the MHCII pathway. Cell 117:677–687.

13. Duan J, Avci FY, Kasper DL (2008) Microbial carbohydrate depolymerization byantigen-presenting cells: Deamination prior to presentation by the MHCII pathway.Proc Natl Acad Sci USA 105:5183–5188.

14. Mitchison NA (1971) The carrier effect in the secondary response to hapten-proteinconjugates. I. Measurement of the effect with transferred cells and objections to thelocal environment hypothesis. Eur J Immunol 1:10–17.

15. Mitchison NA (1971) The carrier effect in the secondary response to hapten-proteinconjugates. II. Cellular cooperation. Eur J Immunol 1:18–27.

16. Beuvery EC, van Rossum F, Nagel J (1982) Comparison of the induction ofimmunoglobulin M and G antibodies in mice with purified pneumococcal type 3 and

meningococcal group C polysaccharides and their protein conjugates. Infect Immun37:15–22.

17. Schneerson R, Barrera O, Sutton A, Robbins JB (1980) Preparation, characterization,and immunogenicity of Haemophilus influenzae type b polysaccharide-protein con-jugates. J Exp Med 152:361–376.

18. Wessels MR, et al. (1993) Stimulation of protective antibodies against type Ia and Ibgroup B streptococci by a type Ia polysaccharide-tetanus toxoid conjugate vaccine.Infect Immun 61:4760–4766.

19. Weintraub A (2003) Immunology of bacterial polysaccharide antigens. Carbohydr Res338:2539–2547.

20. Deck MB, Sjölin P, Unanue ER, Kihlberg J (1999) MHC-restricted, glycopeptide-specificT cells show specificity for both carbohydrate and peptide residues. J Immunol 162:4740–4744.

21. Deck B, Elofsson M, Kihlberg J, Unanue ER (1995) Specificity of glycopeptide-specificT cells. J Immunol 155:1074–1078.

22. Harding CV, Kihlberg J, Elofsson M, Magnusson G, Unanue ER (1993) Glycopeptidesbind MHC molecules and elicit specific T cell responses. J Immunol 151:2419–2425.

23. Khatami A, et al. (2012) Persistence of antibody response following a booster dose ofHib-MenC-TT glycoconjugate vaccine to five years: A follow-up study. Pediatr InfectDis J 31:1069–1073.

24. Snape MD, et al. (2006) Serogroup C meningococcal glycoconjugate vaccine in ado-lescents: Persistence of bactericidal antibodies and kinetics of the immune responseto a booster vaccine more than 3 years after immunization. Clin Infect Dis 43:1387–1394.

25. Snape MD, Pollard AJ (2005) Meningococcal polysaccharide-protein conjugate vac-cines. Lancet Infect Dis 5:21–30.

26. Micoli F, et al. (2013) Development of a glycoconjugate vaccine to prevent meningitisin Africa caused by meningococcal serogroup X. Proc Natl Acad Sci USA 110:19077–19082.

27. Trotter CL, Andrews NJ, Kaczmarski EB, Miller E, Ramsay ME (2004) Effectiveness ofmeningococcal serogroup C conjugate vaccine 4 years after introduction. Lancet 364:365–367.

28. Pace D, et al. (2015) Immunogenicity of reduced dose priming schedules of serogroupC meningococcal conjugate vaccine followed by booster at 12 months in infants:Open label randomised controlled trial. BMJ 350:h1554.

29. Guttormsen HK, et al. (1999) Cognate stimulatory B-cell-T-cell interactions are criticalfor T-cell help recruited by glycoconjugate vaccines. Infect Immun 67:6375–6384.

30. Schröter CJ, et al. (1999) A rapid method to separate endosomes from lysosomalcontents using differential centrifugation and hypotonic lysis of lysosomes. J ImmunolMethods 227:161–168.

198 | www.pnas.org/cgi/doi/10.1073/pnas.1816401115 Sun et al.

Dow

nloa

ded

by g

uest

on

June

18,

202

0