Targeting antigens to CD180 but not CD40 programs immature and mature B cell subsets to become efficient antigen-presenting cells

Abstract

Targeting Ags to the CD180 receptor activates both B cells and dendritic cells (DCs) to become potent Ag-presenting cells. After inoculating mice with Ag conjugated to an αCD180 Ab, B cell receptors were rapidly internalized. Remarkably, all B cell subsets, including even transitional 1 (T1) B cells, were programed to process, present Ag and stimulate Ag-specific CD4⁺ T cells. Within 24-48 hours, Ag-specific B cells were detectable at T-B borders in the spleen; there they proliferated in a T cell-dependent manner and induced the maturation of T follicular helper (T_FH) cells. Remarkably immature B cells were sufficient for the maturation of T_FH cells after CD180 targeting: T_FH cells were induced in BAFFR^−/− mice (with only T1 B cells) and not in μMT mice (lacking all B cells) following CD180 targeting. Unlike CD180 targeting, CD40 targeting only induced DCs, but not B cells to become APCs and thus failed to efficiently induce T_FH cell maturation, resulting in slower and lower-affinity IgG Ab responses. CD180 targeting induces a unique program in Ag-specific B cells and is a novel strategy to induce Ag presentation in both DCs and B cells, especially immature B cells, and thus has the potential to produce a broad range of Ab specificities. This study highlights the ability of immature B cells to present Ag to and induce the maturation of cognate T_FH cells, providing insights towards vaccination of mature B cell deficient individuals and implications in treating autoimmune disorders.

Introduction

Targeting Ag directly to APCs is a highly efficient strategy to induce both humoral and cellular immunity (1). Ag targeting to CD180 (also called RP105), achieved by coupling an Ag directly to an αCD180 Ab (Ag-αCD180), has the advantage of targeting both dendritic cells (DCs) and B cells and of providing an adjuvant effect by activating CD180. Previously, we reported that this platform induces rapid and high affinity Ag-specific IgG responses, which are predominantly T cell-dependent (TD) (2). The direct conjugation of Ag to the αCD180 Ab is required to induce this Ab response; mice inoculated with αCD180 conjugated to the hapten 4-hydroxy-3-nitro-phenacetyl (NP) + free OVA generated Ab to NP and not OVA and vice versa (2). Furthermore, targeting Ag to CD180 can protect immunodeficient mice from a lethal virus infection. Mice deficient for the B cell activating factor receptor (BAFFR) lack mature B cells but do produce transitional 1 (T1) B cells (3). Vaccination of BAFFR^−/− mice with West Nile virus (WNV) E protein conjugated to αCD180 was sufficient to protect them from a subsequent lethal WNV challenge (4). Remarkably, the addition of an adjuvant was not required to induce protection. Conversely, μMT mice, which lack all B cells, were not protected from WNV infection, suggesting that the T1 B cells present in the BAFFR^−/− mice are utilized by the CD180 targeting vaccine to induce protection.

CD180 is a pattern recognition receptor, expressed on DCs, macrophages and B cells. While it is related to TLR family members, unlike TLRs it lacks a TIR domain and does not utilize MyD88 or TRIF for signal propagation (5). Instead, CD180 appears to act as a regulator of other signaling receptors, especially TLRs (5). B cells can be activated via stimulation by an αCD180 Ab, resulting in proliferation that is dependent on CD19 but not MyD88 (6, 7). Indeed, CD180 stimulation by Ab cross-linking on B cells activates signaling elements reminiscent of BCR signaling including Btk, Lyn and Vav (7–9). In addition, CD180 regulates B cell TLR signaling; while activation of B cells via CD180 does not require TLR4, CD180 expression is required for LPS-mediated activation of TLR4 (10). Furthermore, CD180 signaling synergizes to enhance the activation of several TLRs, including TLR-2, −4, −7 and −9 (11). Apparently, during Ag targeting, CD180 synergizes with BCR signaling to improve B cell activation and resulting Ab responses.

Here, we characterized the early events of B cell activation following Ag-αCD180 inoculation, comparing responses of different splenic B cell subsets. We demonstrate that CD180 targeted B cells, including immature T1 cells, are activated to become efficient APCs, which contribute to the development of robust TD humoral responses. Intriguingly, targeting Ags to CD180 induced an earlier and higher affinity Ag-specific IgG response than when Ags were targeted to CD40. Our data suggest that CD180-targeting, by capitalizing on the capabilities of newly formed B cells and the Ag-presenting capacity of multiple B cells subsets as well as DCs, may provide a novel avenue with which to effectively treat patients with immunodeficiencies or cancer.

Materials and Methods

Mice

C57BL/6, C57BL/6 knock-in Ly5.1 and B10.A-H2^a-H2-T18^a/SgSnJ mice were purchased from Jackson Labs (Bar Harbor, ME). B6.SJL-B1-8^hi knock-in Ly5.1 mice were a gift from Dr. Michel Nussenzweig (Rockefeller University, New York, NY) (12). Mature B cell deficient BAFFR^−/− mice were kindly provided by Dr. Klaus Rajewsky (Harvard Medical School, Boston, MA). B cell deficient μMT mice were a gift from Dr. David Rawlings (Seattle Children’s Research Institute, Seattle, WA). OT-II OVA-specific CD4⁺ TCR transgenic knock-in Ly5.1 mice were a gift from Dr. Michael Gerner (University of Washington, Seattle, WA). All strains, except B10.A were on a C57BL/6 background. All mice were age- and sex-matched for experiments and used at 8-12 weeks of age. Mice were housed in a specific pathogen free environment; all procedures were approved by the University of Washington Institutional Animal Care and Use Committee.

Ag-targeting constructs and adjuvants

The rat IgG_2a anti-CD180 (RP/14) was produced from a hybridoma (gift of Dr. Kensuke Miyake, University of Tokyo, Tokyo, Japan) or purchased from BioLegend (UltraLEAF purified). UltraLEAF purified anti-CD40 (1C10, rat IgG_2a) and rat IgG_2a isotype control (RTK2758) were purchased from BioLegend. NP-Ab conjugates were prepared by conjugation to the succinimidyl ester of NP-Osu (BioSearch) as previously described (2). Protein Ags OVA (Sigma) or HEL (Sigma) were conjugated to mAbs via thioether linkages as previously described (13). Free NP was removed by dialysis with molecular weight cut-off tubing of 8kDa (NP-Abs). Molar ratios of NP to mAb were confirmed by spectrophotometry and ranged from NP₅-Ab to NP₁₀-Ab. Molar ratios of OVA or HEL to mAb were quantitated by ELISA and ranged from Ag₁-Ab to Ag₂-Ab. All Ag-Ab constructs were filter sterilized (0.2μM) and stored at 4°C until use. The amount of conjugate administered references the weight of the αCD180 Ab, and in all experiments the moles of Ag inoculated were equivalent between all mice. Due to differences in conjugation ratios this resulted in slightly higher weights of both the Iso and αCD40 Ab in all inoculations. All inoculations were prepared to 200μL in pharmaceutical grade PBS and administered by retro-orbital inoculation. The TLR7 agonist R848 was obtained from Invivogen and, when used, mixed with the Ag-Ab preparation.

Adoptive transfers

Splenocytes from B1-8^hi mice were processed by mechanical disruption between frosted glass slides. B cells were isolated by magnetic bead negative selection (StemCell Technologies) and were >95% pure as determined by flow cytometry. 50×10^6 B cells/mL were labeled with 20μM CFSE (Invitrogen) at 37°C for 10 min and washed thoroughly. Labeled cells were transferred in 200μL i.v. 16-18 h prior to Ag-Ab inoculation. Male cells were never transferred into female mice.

CD4⁺ T cell depletion

Mice were inoculated with 100μg αCD4 (GK1.5, prepared in our lab or purchased from BioLegend) or 100μg Rat IgG_2b isotype control (BioLegend) i.p. 24 h prior to Ag-Ab inoculation and approximately 5 h prior to the adoptive transfer of B cells.

Tetramer-specific cell enrichment

OVA-specific T cells were identified using OVA-IAb tetramers 2C and 3C, prepared as previously described (14). Tetramer-specific cells were enriched as previously described (15). Briefly, isolated splenocytes were incubated with the OVA-tetramer followed by anti-PE microbeads (Miltenyi Biotec). The suspension was run over magnetic LS columns (Miltenyi Biotec), washed multiple times and then eluted to enrich tetramer-specific cells. The bound fraction was stained for flow cytometry as below. Absolute cell counts were analyzed using AccuCheck counting beads (Invitrogen) as per the manufacturer’s instructions.

Flow cytometry

Red blood cells were lysed with RBC lysis buffer (Invitrogen) and isolated splenocytes were stained with Live/Dead Aqua (ThermoFisher), or Fixable Viability Dye eF780 (ThermoFisher) for 20 min at 4°C in the absence of FBS, to differentiate live and dead cells. Cells were subsequently stained for surface markers (Supplemental table 1) in the presence of F_c block (αCD16/32, BioLegend) and 2% FBS. Cells expressing NP-specific BCRs were detected by staining with NP-PE (BioSearch). HEL peptide-loaded MHC II was detected using a mAb clone C3H4 (16) (kindly provided by Dr. Ron Germain, NIAID, Bethesda, MD), which was biotinylated and detected using streptavidin-PE-TexasRed (BD). After washing, cells were fixed in 1% paraformaldehyde (PFA) and stored at 4°C until analysis. Cells were analyzed on an LSRII flow cytometer (BD) and data were analyzed using FlowJo (v.10, Tree Star). See Figure 1A for gating strategies.

Figure 1. The Ag-specific BCR is internalized rapidly following targeting to CD180.

B1-8^hi mice were inoculated with 50μg NP-Iso, NP-Iso + αCD180 or NP-αCD180. Three, six or 25 h later, spleens were harvested and processed for flow cytometry. (A) Representative flow plots of total live, single, B220⁺ cells three h after inoculation. The top panel demonstrates the gating strategy we used to identify different splenic B cell subsets. After gating out debris, doublets and dead cells, B220⁺ B cell subsets were evaluated using CD21 and CD24 expression. Transitional (T)1 B cells were defined as CD24^hi, CD21^lo; T2 as CD24^hi, CD21^mid; follicular (FO) as CD24^mid, CD21^mid; marginal zone (MZ) as CD24^hi, CD21^hi. The bottom panel demonstrates the change in NP-BCR expression following inoculation. (B) NP-BCR^hi expressing cells per spleen following inoculation in different B220⁺ splenic B cell subsets. Data is the mean ± SEM combined of two to three experiments per time point with n=4-8 per group. (C,D) In vitro internalization assay in the NP-specific BCR expressing K46 B cell line. Cells were incubated with fluorescently conjugated NP conjugates and internalization was evaluated using the ImageStream flow cytometer and Ideas software. Data is representative of three individual experiments. * p<0.05 and ** p<0.01 NP-Iso vs. NP-αCD180. † p<0.05, ‡ p<0.01 NP-Iso + αCD180 vs. NP-αCD180 by Kruskal-Wallis test with Dunn’s multiple comparison test (B) or 2-way ANOVA with Tukey’s multiple comparison test (D).

In-vitro internalization assay

The K46μM17 (K46) murine B cell lymphoma cell line expressing the B1-8 high affinity NP-specific BCR (17), was a gift from Dr. Louis B. Justement (University of Alabama, Birmingham AL). Cells were cultured in RPMI 1640 (GenClone) supplemented with 10% FBS (ThermoFisher), 1mM sodium pyruvate (GE), non-essential amino acids (GenClone), L-glutamate, penicillin, streptomycin (Corning), and 50μM β-mercaptoethanol. Alexa Fluor 647 was conjugated to NP-Iso and NP-αCD180 using a labeling kit (Invitrogen) according to the manufacturer’s instructions. Fluorescent Ag conjugates and αB220-eF450 (RA3-6B2, eBioscience) were incubated with K46 cells at 4°C for 20 min to allow for receptor binding. Cells were then incubated at 37°C to allow for BCR internalization, and subsequently fixed with 1% PFA to halt internalization. Cells were analyzed using the Imagestream X Mark II flow cytometer (Amnis), and data were analyzed using IDEAS software (Amnis). Internalization was quantified on focused, single cells as follows: an internal cellular zone was defined using an adaptive erode mask based on surface B220 staining and internalization was quantitated using the internal zone mask and the fluorescence of the Ag. Normalized internalization was calculated by dividing the percent of internalized cells at time × min over time 0 min.

APC isolation and Ex-vivo T cell stimulation

B6 mice (Ly5.2⁺) were inoculated with OVA-Ab conjugates. Splenocytes were processed to single cells as above. CD11c⁺ cells were isolated by magnetic bead positive selection (Miltenyi Biotec) according to manufacturer’s instructions. From the negative fraction, B cells were isolated by magnetic bead negative selection (StemCell Technologies) according to manufacturer’s instructions. B cells were then FACS sorted into T1 (B220⁺, CD24^hi, CD21^lo), FO (B220⁺, CD24^mid, CD21^mid) and MZ (B220⁺, CD24^hi, CD21^hi) subsets (as in Figure 1A, with smaller gates to limit subset spillover, as is a common sorting procedure), using an Aria flow cytometer (BD). Meanwhile, CD4⁺ T cells (Ly5.1⁺) were isolated from splenocytes from OT-II mice by magnetic bead negative selection (StemCell Technologies) and labeled with CFSE as above. 1×10⁵ CFSE-labeled CD4⁺ T cells were combined with 1×10⁵ APC (DC or B cell subset) and incubated in a 96-well plate for three days, following which cells were stained for surface markers and analyzed by flow cytometry as above.

Immunohistochemistry

Spleens were isolated from mice and cut in half. One half was fixed in 1% paraformaldehyde and then cryoprotected in 30% sucrose. The spleen half was embedded in OCT and frozen at −80°C; 6μm sections were cut and slides were stored at −80°C until use. Sections were allowed to thaw before equilibrating in PBS. Sections were washed with PBST (PBS containing 0.05% Tween-20), blocked for 30 min (PBST + 10μg/mL F_c blocking antibody + 10% normal goat serum) and then stained with B220-BV421, CD3-FITC, NP-PE and Ly6G-AF647 (in PBST + 3% BSA) for one h at room temperature. Following several PBST washes, glass slides were mounted over sections with ProLong Diamond (ThermoFisher) mounting medium. Confocal images were obtained using a 20x objective on a Nikon Eclipse Ti inverted microscope with a Nikon C2 confocal system. Images were processed using Nikon Elements AR v5.10.01.

ELISA and ELISPOT

NP-specific ELISA and ELISPOT assays were performed as described previously (2). Briefly, for ELISAs, sera were incubated over NiP-BSA coated polystyrene plates, detected using HRP-αIgG plus TMB substrate, and quantitated by comparison to a standard curve of known IgG concentrations. For ELISPOT assays, splenocytes or isolated bone marrow cells were incubated at 37°C overnight over NiP-BSA coated mixed cellulose ester membrane filter plates and IgG spots were detected by HRP-αIgG and AEC substrate.

Statistical analyses

Raw experimental data was analyzed using a Mann-Whitney test, a Kruskal-Wallis test with Dunn’s multiple comparison or a two-way ANOVA with Tukey’s multiple comparison test, using GraphPad Prism 7. Differences of p<0.05 were considered significant.

Results

Targeting antigen to CD180 induces rapid BCR internalization

CD180 Ag targeting induces an expansion of follicular (FO) and immature (T1/T2) subsets while marginal zone (MZ) B cells decline one day after immunization (2). CD180 Ag targeting also induces T1 B cells in BAFFR^−/− mice to promote protective immune responses (4). Given these findings we decided to characterize early phenotypic changes in different splenic B cell subsets induced by CD180-targeting. B1-8^hi mice, with a transgenic BCR that recognizes the hapten NP on 15-20% of splenic B cells, were immunized with NP conjugated to αCD180 (NP-αCD180) or an isotype control (NP-Iso). We examined the expression of the NP-specific BCR by staining cells with an NP-PE conjugate. The expression of CD21 and CD24 on B220⁺ cells were used to define splenic B cell subsets as seen in Figure 1A and as previously described (18). There was no change to the total number of splenic B cell subsets 3 to 6 hours after inoculation, while there was a modest increase in T2 and FO B cells and a decrease in total MZ B cells by 24 hours (Fig. S1A) as previously reported (2). However, three hours after inoculation the number of cells expressing the NP-specific BCR decreased significantly; this corresponded with an increase in IgM⁻ B cells, suggesting that the BCR was being internalized following interaction with its cognate Ag (Fig. 1A bottom panel). Interestingly, when NP was conjugated to αCD180, the NP-BCR was internalized more rapidly than when NP was conjugated to the isotype control (Fig. 1A,B). By three hours after inoculation, the frequency of NP-specific B220⁺ cells in mice that had received NP-Iso had decreased about 7-8-fold (e.g., from 13% to 1.7%, Fig. 1A) and many B cells still expressed high levels of the NP-BCR. However, in mice inoculated with NP-αCD180, the frequency of NP-BCR⁺ cells dropped more than 100-fold, and the few detectable NP-BCR⁺ cells had very low BCR expression levels (Fig. 1A). This rapid internalization occurred in all splenic B cell subsets including immature T1 and T2 B cells (Fig. 1B, S1B). We did not observe a corresponding decrease in the expression of BCR associated proteins such as CD21 or CD23 (Fig. S1C). By six hours after inoculation, the expression of the NP-BCR reached its nadir in both immunization groups. However, while the NP-BCR had returned to the surface by 24 hours post-inoculation in NP-Iso immunized mice, it was slower to be re-expressed on FO, T1 and T2 B cells from NP-αCD180 treated mice. MZ B cells from mice that received NP-αCD180 still had not re-expressed the NP-BCR after 24 hours. Given the overall reduction of MZ B cells by 24 hours (Fig. S1A) it is possible that Ag-specific MZ B cells migrated from the spleen. Nonetheless, the NP-specific BCR is internalized more rapidly on all the subsets as compared to NP-Iso inoculation. Interestingly, direct conjugation of the Ag to αCD180 was required for the faster internalization rate, as mice immunized with NP-Iso + αCD180 demonstrated similar BCR internalization kinetics as mice immunized with NP-Iso (Fig. 1B).

In order to rule out the possibility that NP in the immunization was blocking the binding of the NP-PE stain, we confirmed the decrease of NP-BCR expression by staining for Ig λ light chain associated with the NP-BCR (12). As expected, the expression of Ig λ decreased with similar kinetics as NP-binding BCR following NP-αCD180 inoculation (Fig. S1D). We also tested for the expression of the NP-BCR in permeabilized cells and, as expected, found that the BCR was being internalized rather than shed; similar numbers of NP-binding cells were detected in mock and immunized mice (Fig. S1E). To further assess our in vivo findings, we developed an in vitro internalization assay using the K46μM17 B cell line that expresses the B1-8 NP-specific BCR (17). AF647-NP-Iso and AF647-NP-αCD180 were incubated with K46 cells and allowed to internalize. Cells were then visualized at different time points using the ImageStream flow cytometer (Fig. 1C,D). The red Ag moved from the cell surface to the cytoplasm within three minutes after treatment with NP-αCD180. This process took at least five minutes if the Ag was attached to the isotype Ab (NP-Iso) even in the presence of unconjugated αCD180 (NP-Iso + αCD180). The results are consistent with the differential kinetics of BCR internalization observed in vivo.

B cells present Ag and stimulate Ag-specific T cells following CD180 targeting

Given the adjusted rate of BCR-Ag internalization when Ag was attached to αCD180, we hypothesized that Ag processing within B cells might also be altered, thus resulting in differential abilities to present Ag. To address potential changes in B cell activation, we first examined levels of MHC II and CD86 on the surface of NP-BCR⁻ or NP-BCR⁺ B cells 24 hours after immunization of B1-8^hi mice (as in Fig. 1). Given that NP-BCR expression is variable at this time point depending on the inoculation and subset (Fig. 1B), we decided to quantify NP-BCR⁺ cells based on their expression of NP-BCR⁺ or reduction of surface IgM (see Fig. S2). When mice were inoculated with either NP-Iso or NP-αCD180, NP-specific B cells had elevated levels of MHC II on all the B cell subsets compared to NP-BCR⁻ cells (Fig. 2A). With the exception of FO B cells, inoculation of NP-Iso + αCD180 induced only moderate increases in MHC II expression regardless of BCR specificity. NP-BCR⁺ B cells also showed increased expression of CD86 following BCR-engagement, and additionally, CD86 levels were significantly increased on B cells from mice inoculated with NP-αCD180 as compared to NP-Iso +/− αCD180 (Fig. 2A). Remarkably, NP-αCD180 inoculation even induced increased expression of both MHC II and CD86 on NP-specific immature T1 B cells.

Figure 2. Targeting Ags to CD180 results in Ag processing and presentation by B cells.

(A) B1-8^hi mice were inoculated with 50μg NP-Iso, NP-Iso + αCD180 or NP-αCD180. 24 h later, the expression of MHC II and CD86 was evaluated on NP-BCR⁻ and NP-BCR⁺ splenic B cells by flow cytometry. Data is the mean ± SEM representative of two independent experiments with n=3-4 per group. (B) B10.A mice were inoculated with 35μg HEL-Iso, HEL-Iso + αCD180 or HEL-αCD180. Three or 24 h later HEL-peptide bound-MHC II I-A^k, represented as peptide bound MHC II over total MHC II expression (HEL-I-A^k/I-A^k MFI) was evaluated on splenic B cell subsets and conventional (c)DCs by flow cytometry. cDCs were defined as B220⁻, CD3⁻, NK1.1⁻, CD11b^−/lo, CD11c^hi. Data is the mean ± SEM representative of two independent experiments per time point with n=3-4 per group. (C) C57BL/6 mice were inoculated with 35μg OVA-Iso, OVA-Iso + αCD180 or OVA-αCD180. Three or 24 h later DCs were isolated from spleen by positive bead selection, B cells were isolated by negative bead enrichment and B cell subsets were sorted by FACS. 1×10⁵ APCs were co-cultured with 1×10^5 CFSE-labeled OT-II T cells. Three days later, T cell proliferation was evaluated as CFSE dilution by flow cytometry. Data is the mean ± SEM combined of three experiments per time point with n=1 (Mock, OVA-Iso) or n=6 (OVA-Iso + αCD180 and OVA-αCD180). * p<0.05, ** p<0.01, ***p<0.001, ****p<0.0001 by 2-way ANOVA with Sidak’s multiple comparison test (A), Kruskal-Wallis test with Dunn’s multiple comparison test (B) or Mann-Whitney test (C).

To assess the ability of different B cell subsets to present Ag, we utilized two different assays; identification of peptide/MHC II complexes on the surface of B cells, and a functional readout of Ag-presentation via the stimulation of Ag-specific CD4⁺ T cells. B10.A mice (MHC II I-A^k) were immunized with hen egg lysozyme (HEL) conjugated to αCD180 or the isotype control Ab. HEL peptide (aa 46-61) expressed in association with MHC II I-A^k was quantitated by staining cells with the C4H3 monoclonal Ab (16). To account for changes in total MHC II expression levels following inoculation and any potential non-specificity of the C4H3 antibody, we evaluated Ag presentation by determining the ratio of Ag-bound MHC II to total MHC II (Fig. 2B). HEL peptide-MHC II complexes were expressed on all B cell subsets, including T1 B cells, as early as three hours following HEL-αCD180 inoculation but not after HEL-Iso +/− αCD180 inoculation; expression was maintained for at least 24 hours.

Immature B cells are not known to stimulate Ag-specific T cells. However, given the induction of CD86 expression on and Ag presentation by T1 B cells following CD180 Ag targeting, we tested whether these B cells could also stimulate Ag-specific T cells. C57BL/6 (B6) mice were inoculated with OVA-Ab conjugates; splenic CD11c⁺ DCs and B cells were then isolated; and T1, FO and MZ B cell subsets were sorted by FACS. The B cells or DCs were incubated with CFSE labeled OT-II T cells specific for OVA. As expected, DCs from OVA-αCD180 inoculated mice obtained at both three and 24 hours post inoculation were robust stimulators of OVA-specific T cells (93-96% of CD4⁺ T cells had diluted CFSE; Fig. 2C). All B cell subsets tested were also able to stimulate T cells, to varying levels. At three hours post immunization, FO and MZ cells demonstrated equivalent T cell stimulatory capacities, while by 24 hours FO B cells had a slightly reduced capacity to stimulate T cells compared to MZ B cells. Intriguingly, T1 B cells were capable of stimulating OVA-specific T cells at both time points, although to a lesser extent than the mature B cell subsets. As expected, B cells or DCs from mice inoculated with OVA-Iso were unable to stimulate proliferation in OVA-specific T cells. Intriguingly, APCs from mice that had received OVA-Iso + αCD180 were similarly unable to activate Ag-specific T cells. These data demonstrate that when Ag is directly conjugated to αCD180, both immature and mature B cells are capable of presenting Ag and stimulating Ag-specific T cells.

CD180 targeting induces Ag-specific B cells to migrate to the splenic T cell zone

The ability of B cells to present Ag and activate Ag-specific T cells in vitro suggested that CD180 targeting may program B cells to migrate to the T cell zone in order to interact with cognate T cells. To test this hypothesis, we first examined the surface expression of CXCR5, the chemokine receptor that binds to CXCL13, guiding cells to B cell follicles (19). All Ag-specific FO B cells that encountered Ag decreased their CXCR5 expression 24 hours after inoculation (Fig. 3A), consistent with reported events following BCR stimulation (19). However, the levels of CXCR5 on Ag-specific FO B cells from mice inoculated with NP-αCD180 were significantly lower than Ag-specific FO B cells from mice inoculated with NP-Iso +/− αCD180. This decrease in CXCR5 expression suggested that FO B cells were gaining the ability to exit B cell follicles and might migrate to T cell zones.

Figure 3. CD180 Ag targeting induces Ag-specific B cells to migrate to the T-B border region of the spleen.

(A) B1-8^hi mice were inoculated with 50μg NP-Iso, NP-Iso + αCD180 or NP-αCD180. 24 h later, the expression of CXCR5 was evaluated on NP-BCR⁻ and NP-BCR⁺ splenic B cells by flow cytometry. Data is the mean ± SEM representative of two independent experiments with n=3-4 per group. * p<0.05, ** p<0.01, ***p<0.001, ****p<0.0001 by 2-way ANOVA with Sidak’s multiple comparison test. (B-E) B cells, enriched from B1-8^hi mice, were transferred into WT mice, which were rested overnight and then inoculated with 50μg NP-Iso + αCD180 or NP-αCD180. Spleens were isolated from uninoculated mice (Day 0) or inoculated mice every day for three days. Spleens were cryopreserved, sectioned and stained for NP-BCR, B220, CD3 and Ly6G. Images were acquired by confocal microscopy at 20x objective and are representative of 5 images taken from two sections and two mice per group. (D) Total NP-specific B cells were counted at the border region between the B220 expressing B cell follicle and the CD3 expressing T cell zone from a total of 20 images from 4 sections representing 2 different mice per group.

To address this possibility, we isolated B cells from B1-8^hi mice and adoptively transferred them into WT mice, rested the mice for approximately 18 hours, and then inoculated them with NP-Iso + αCD180 or NP-αCD180. We isolated spleens every day after immunization for three days to examine B cell migration by confocal microscopy (Fig. 3B-D). NP-specific B cells at the time of inoculation (~18 hours after transfer) were present in B220⁺ B cell follicles and in extra-follicular regions, defined by presence of Ly6G⁺ neutrophils (Fig. 3B). One day following NP-αCD180 inoculation, NP-specific B cells could be found at the border between the B cell follicles and the T cell zone, as defined by the presence of CD3⁺ cells (Fig. 3C top panel). More NP-specific B cells were identified at the T/B border 1 day following NP- αCD180 inoculation than at day 0 or following NP-Iso + αCD180 inoculation (Fig. 3D), suggesting that NP-specific B cells trafficked to the T/B border following CD180 targeting. By day 2 after NP-αCD180 inoculation (Fig. 3C, middle panel), the numbers of NP-specific B cells had clearly increased, and they remained near T-B borders even at day three (Fig. 3C, bottom panel). By contrast, three days after inoculation of mice with NP-Iso + αCD180, very few NP-specific B cells could be found, and those that were detected were located in follicles or extra-follicular spaces (Fig. 3E). Taken together these data indicate that CD180 targeting induced B cells to present Ag, migrate to T-B borders and proliferate, possibly in a T-cell dependent manner.

Ag-specific B cells, with the help of T cells, proliferate rapidly following CD180 targeting.

To address further which B cell subsets proliferate in response to CD180 targeting, we adoptively transferred CFSE-labeled B1-8^hi B cells into B6 mice one day prior to NP-αCD180 inoculation. NP-specific B cells underwent several cell divisions within three days following NP-αCD180 inoculation (Fig. 4A). In contrast, unlike in mock control mice, Ag-specific B cells were no longer detected in mice that had received NP-Iso or NP-Iso + αCD180 (Fig. 4A). In addition, non-specific B cells proliferated to some degree in response to both NP- αCD180 and NP-Iso + αCD180 inoculation. As both the numbers of non-specific proliferating B cells and level of proliferation were the same following both inoculations (Fig. S3A,B), this non-specific B cell proliferation was likely due to the activation of CD180, independent of other signals such as the BCR. Indeed, the division index, the average number of cell divisions that a cell underwent, was significantly higher in Ag-specific B cells following NP-αCD180 inoculation than non-specific B cells (Fig. S3C), showing that Ag-specific cells were more likely to begin to proliferate and undergo more rounds of proliferation than non-specific B cells. These data suggest that CD180 and BCR signaling synergizes in Ag-specific cells to boost proliferation and/or there are additional signals received by the Ag-specific cells that promote activation. The Ag-specific proliferating B cells induced by NP-αCD180 comprised a single population that straddled the FO and T1 gates (Fig. 4B). The overall number of T1 B cells increased after CD180 targeting (Fig. 4C), which is likely reflective of their proliferation (Fig. 4B) but may also be due to increased survival. These data highlight the unique ability of CD180 targeting to activate newly formed Ag-specific B cells.

Figure 4. CD180 targeting induces rapid, T cell-dependent Ag-specific B cell proliferation.

Splenic B cells were isolated from B1-8^hi (Ly5.1⁺) mice, labeled with CFSE and transferred into C57BL/6 mice. The next day the mice were inoculated with 50μg NP-Iso, NP-Iso + αCD180 or NP-αCD180 and spleens were harvested three days later for analysis by flow cytometry. (A) Representative flow plots of B220⁺, Ly5.1⁺ cells, and data (mean ± SEM) of CFSE diluted, NP-BCR⁺ cells per spleen. (B) Representative flow plots of B220⁺, Ly5.1⁺ cells showing their CD21, CD24 along the axes and either CFSE or NP-BCR expression as a heat map overlay. (C) Total B220⁺, Ly5.1⁺ B cell subsets from the spleen. (D) C57BL/6 mice were inoculated with anti-CD4 or an isotype control Ab 5h before B cell adoptive transfer and 24h before NP-Ab inoculation (as above). Three days later B220⁺, Ly5.1⁺, NP-BCR⁺, CFSE diluted cells were examined in the spleen by flow cytometry. Data is mean ± SEM combined from two independent experiments, n=5 per group. * p<0.05, ** p<0.01 by Kruskal-Wallis test with Dunn’s multiple comparison test.

Given the rapid Ag-presentation by B cells and their ability to stimulate Ag-specific T cells in vitro after CD180 targeting, we tested whether the in vivo Ag-specific B cell proliferation observed three days after Ag targeting was dependent on T cells. We transferred CFSE-labeled B1-8^hi B cells into B6 mice that previously had been inoculated with an αCD4 mAb to deplete CD4⁺ T cells, or an isotype control mAb. The numbers of Ag-specific B cells that proliferated in the absence of CD4⁺ T cells was significantly reduced compared to the isotype control-treated mice (Fig. 4D). This suggests that the CD180 targeted B cells, not only are capable of stimulating Ag-specific T cells, but that these T cells in turn are then required for Ag-specific B cell proliferation.

CD180 targeted B cells contribute to the rapid development of Ag-specific T follicular helper cells.

Given that Ag-specific B cell proliferation three days after CD180 targeting was TD, we next investigated if these T-B interactions led to T cell activation and maturation as well. Three days following inoculation of mice with OVA-Iso + αCD180 or OVA-αCD180, OVA-specific CD4⁺ T cells were analyzed using tetramer enrichment. Total splenocytes were incubated with a class II OVA tetramer and then run over a magnetic column to enrich for Ag-specific CD4⁺ T cells. The resulting fraction was analyzed by flow cytometry for the expression of CXCR5 and PD-1 to identify Ag-specific T follicular helper (T_FH) cells. As seen in figure 5A, there was a significant increase in OVA-specific CXCR5⁺, PD-1⁺ T_FH cells three days after OVA-αCD180 inoculation as compared to the mock or unconjugated control. Thus, at the same time that Ag-specific B cells proliferate following CD180 targeting, Ag-specific CD4⁺ T cells are programmed to mature into T_FH cells. Given that we had observed CD180-targeted B cells presenting Ag to cognate T cells in vitro, we next investigated whether B cells were required for T_FH cell maturation. Following inoculation of NP-αCD180, μMT mice (lacking all B cells) failed to form mature T_FH cells by three days (Fig. 5B,C). Intriguingly, there was not a significant difference in either the total number of T_FH cells per spleen or the T_FH cells as a percent of the total OVA-specific population in BAFFR^−/− mice (with only T1 B cells) as compared to WT controls following CD180 targeting (Fig. 5C). Furthermore, there was less of a difference in the T_FH as percent of tetramer⁺ cells in WT and BAFFR^−/− mice (mean 43.38 vs. 37.25) (Fig. 5C, right panel) compared to the total tetramer⁺ T_FH cell number in the spleen (mean 30.02 vs. 18.10) (Fig. 5C, left panel). This would suggest that, while the mature B cells in the WT mice contribute to the maturation of T_FH cells and thus there are more total T_FH in the WT mice, immature B cells have equivalent capacity to mature B cells in providing maturation signals to cognate T cells, resulting in more equal levels as measured by percent of total Ag-specific T cells. Overall, these data suggest that the rapid maturation of T_FH cells following CD180 targeting requires B cells and that immature B cells are sufficient for this response.

Figure 5. CD180 targeting induces rapid T_FH cell maturation.

(A) C57BL/6 (WT) mice were inoculated with 25μg OVA-Iso + αCD180 or OVA-αCD180; 3d later OVA-tetramer-specific cells were isolated from splenocytes by magnetic bead enrichment. Total B220⁻, CD11c⁻, CD11b⁻, CD3⁺, CD8⁻, CD4⁺, CD44⁺, tetramer⁺ cells per spleen were analyzed by flow cytometry. (B,C) WT, BAFFR^−/− and μMT mice were inoculated with 25μg OVA-αCD180 and T_FH cells were analyzed by flow cytometry 3d later following OVA-tetramer enrichment. (B) Representative flow plots of CXCR5 and PD-1 expression on the CD44⁺, tetramer⁺ populations. (C) Total Tetramer⁺, CXCR5⁺, PD-1⁺ T_FH cells per spleen (left) and T_FH cells as a % of CD44⁺ Tetramer⁺ cells (right). Data is mean ± SEM combined from three independent experiments (A,C). n=6 (Mock), 9 (OVA-Iso + αCD180, OVA-αCD180 in BAFFR^−/−, OVA-αCD180 in μMT) or 12(OVA-αCD180 in WT) per group. * p<0.05, ** p<0.01 by Kruskal-Wallis test with Dunn’s multiple comparison test. T_FH: T follicular helper cell.

In contrast to CD180-targeted B cells, Ag-specific B cells do no proliferate following CD40 targeting.

The findings above show that targeting Ag to CD180 induces rapid Ag-presentation, movement and proliferation of Ag-specific B cells, and that these responses are dependent on the direct conjugation of Ag to αCD180. We next asked whether these responses were unique to CD180 targeting, or if we could recapitulate the results by targeting Ag to a different B cell activating receptor, CD40. We selected CD40 as a target since like CD180, it is expressed on both B cell and non-B cell APCs, and because anti-CD40 alone can induce DCs to become efficient APCs (20).

We first compared the ability of CD40 targeting vs. CD180 targeting to induce early Ag-specific B cell proliferation (Fig. 6). B6 mice were inoculated with CFSE-labeled Ly5-1⁺ B1-8^hi B cells and then either NP-αCD180 or NP-αCD40; three days later spleens were isolated and CFSE dilution was evaluated. By examining the fluorescent intensity of CFSE as a heatmap overlay of the Ly5.1⁺ B cell’s CD21 and CD24 expression (Fig. 6A), we observed that multiple B cell subsets proliferate in response to CD40 targeting, as evident by the green, low-intensity CFSE signal. By contrast, CD180 targeting induced proliferation in a more restrictive population, straddling the FO/T1 gates (Fig. 6A). Furthermore, CD180 targeting induced proliferation of T1 B cells (Fig. 6B) and the proportion of surviving transferred B cells that were T1 B cells was significantly greater than following NP-Iso inoculation (Fig. 6B). This was not observed following CD40 targeting; there was no change in the proportion of T1 B cells following NP-αCD40 inoculation compared to NP-Iso inoculation. To address further whether this was phenomenon was restricted to CD180 targeting, we tested the ability of R848, a TLR7 agonist, to improve T1 B cell survival; TLR7 is expressed in T1 B cells and TLR7 stimulation induces proliferation of T1 B cells (21). We did not observe significant changes in the T1 B cell population following NP-Iso + R848 inoculation compared to NP-Iso alone (Fig. S3D), highlighting the unique capacity of CD180 targeting to activate this immature subset.

Figure 6. B cell proliferation differs as a result of CD40 vs. CD180 targeting.

Splenic B cells were isolated from B1-8^hi (Ly5.1⁺) mice, labeled with CFSE and transferred into C57BL/6 mice. The next day the mice were inoculated with 50μg NP-Iso, NP-αCD40 or NP-αCD180 and spleens were harvested three days later for analysis by flow cytometry. (A) Representative flow plots of B220⁺, Ly5.1⁺ cells showing their CD21, CD24 expression along the axes and CFSE expression as a heatmap overlay. (B) Splenic B cell subsets as a percent of total B220⁺, Ly5.1⁺ cells. Data is mean ± SEM combined from two experiments, n=6 per group. (C) Representative flow plots of B220⁺, Ly5.1⁺ cells showing their CD21, CD24 along the axes and NP-BCR expression as a heat map overlay. (D) CFSE diluted NP-BCR⁻ (left) or NP-BCR⁺ (right) B220⁺, Ly5.1⁺ cells per spleen. (E) Division index of NP-BCR⁻ (left) or NP-BCR⁺ (right) cells. Data (D,F) is mean ± SEM combined from three experiments, n=9 per group. * p<0.05, *** p<0.001 by Kruskal-Wallis test with Dunn’s multiple comparison test (B), or Mann-Whitney test (D,E).

Despite the robust proliferation of B cells following CD40 targeting (Fig. 6A), very few of the proliferating B cells were Ag-specific (Fig. 6C). Indeed, CD180 but not CD40 targeting induced vigorous proliferation in Ag-specific B cells, while CD40 targeting induced more robust proliferation in non-specific B cells (Fig. 6D,E). These data suggest that CD180 signaling, in contrast to CD40 signaling, is able to synergize with BCR signaling to program Ag-specific B cell activation. Furthermore, CD40 signaling alone is a stronger inducer of B cell activation than CD180 alone, demonstrating a greater chance of off-target activation following CD40 targeting.

CD40 targeting does not replicate the induction of T-B interactions observed following CD180 targeting.

We next compared the ability of B cell subsets and DCs to stimulate Ag-specific CD4⁺ T cells following CD40 vs. CD180 targeting. B6 mice were inoculated with OVA-αCD40 or OVA-αCD180; 24 hours later splenic DCs and B cell subsets were isolated and incubated with CFSE-labeled OT-II T cells. Unlike CD180 targeting, inoculating mice with OVA-αCD40 did not activate either T1 or FO B cells to stimulate Ag-specific T cell proliferation (Fig. 7A). CD40-targeted MZ B cells induced very low levels of T cell proliferation, significantly lower than CD180 targeted MZ B cells. While CD11c⁺ DCs from CD40 targeted mice could induce Ag-specific CD4 T cell proliferation, they were significantly less effective compared to CD180-targeted DCs. More OT-II T cells stimulated by CD180 targeted DCs divided and once they started proliferating, underwent more cell divisions than OT-II T cells stimulated by CD40 targeted DCs (Fig. 7B). Next, we compared the ability of CD40 targeting vs. CD180 targeting to induce the maturation of T_FH cells. B6 mice were inoculated with either OVA-αCD40 or OVA-αCD180. Three days later OVA-specific T cells, following tetramer enrichment of splenocytes, were examined by flow cytometry. Compared to CD40 targeting, there was an increase in the total number of CD4⁺, CD44⁺, tetramer⁺ cells in the spleens of mice following CD180 targeting, but this was not significant (Fig. 7C, left). However, the phenotype of these CD4⁺ T cells was different; we found that there was a significant increase in the numbers of OVA-specific, CXCR5⁺, PD-1⁺ T_FH cells in mice that had received OVA-αCD180 compared to OVA-αCD40 (Fig. 7C, right).

Figure 7. CD40 targeting does not recapitulate key events following CD180 targeting.

(A,B) C57BL/6 mice were inoculated with 35μg OVA-αCD40 or OVA-αCD180; 24h later DCs were isolated from spleen by positive bead selection, B cells were isolated by negative bead enrichment and B cell subsets were sorted by FACS. 1×10⁵ APCs were co-cultured with 1×10⁵ CFSE-labeled OT-II T cells. Three days later, T cell proliferation was evaluated as CFSE dilution by flow cytometry. The division index (B, left) is the average number of divisions that a cell has undergone. Data is the mean ± SEM combined of three experiments with n=4 (OVA-αCD40) or n=6 (OVA-αCD180). (C) C57BL/6 mice were inoculated with 25μg OVA-αCD40 or OVA-αCD180; 3d later OVA-tetramer-specific cells were isolated from splenocytes by magnetic bead enrichment. Total CD44⁺, Tetramer⁺ (left) and Tetramer⁺, CXCR5⁺, PD-1⁺ T_FH cells (right) were analyzed by flow cytometry. Data is mean ± SEM combined from three independent experiments, n=12 per group. (D,E) C57BL/6 mice were inoculated with 50μg NP-Iso, NP-αCD40 or NP-αCD180. Once a week, mice were bled and NP-specific IgG (D, left) and the ratio of high affinity NP-specific IgG (NP2) to total NP-specific IgG (NP20) (D, right) was quantitated by ELISA. Nine weeks after inoculation, spleens and bone marrow were harvested and NP-specific IgG secreting antibody secreting cells (ASCs) were evaluated by ELISPOT (E). Data is mean ± SEM combined from two independent experiments, n=8 per group. * p<0.05, ** p<0.01, ***p<0.001 by Mann-Whitney test (A, B, C, D right), Kruskal-Wallis test with Dunn’s multiple comparison test (E) or 2-way ANOVA with Sidak’s multiple comparison test (F left).

Given the differences in both B cell and T cell activation following CD40 vs. CD180 targeting, we next compared the ability of these two platforms to induce Ag-specific IgG responses. Both the magnitude and kinetics of NP-specific IgG Ab responses were different after targeting Ag to CD40 vs. CD180 (Fig. 7D, left). Following inoculation of mice with NP-αCD180, NP-specific IgG production peaked at day 7 (mean 7325 μg/mL serum), while the peak response following CD40 targeting with NP-αCD40 occurred at day 14 and was about 10-fold lower (mean 748 μg/mL serum). Furthermore, the NP-specific IgG titers decreased more rapidly following CD40 targeting than CD180 targeting; by 49 days post inoculation, the titers were about 2-fold higher in mice that had received NP-αCD180 compared to NP-αCD40. To assess relative affinity of the Abs, we measured the binding of IgG to BSA with low levels of conjugated NP (NP2), which will bind to high affinity Abs, and compared the binding of IgG to BSA conjugated to high levels of NP (NP20) which will bind Ab of both high and low affinity. CD180 targeting was also better at inducing affinity maturation than targeting to CD40. At days 14 and 21 following inoculation, the NP2:NP20 ratio was greater than 2-fold higher in mice that had received NP-αCD180 versus NP-αCD40 (Fig. 7D, right), indicating a greater proportion of high affinity Ab in these mice. Finally, we examined the levels of NP-specific IgG Ab secreting cells (ASCs) in the spleen and bone marrow (BM) following targeting of Ag to CD40 or CD180. Nine weeks after inoculation with NP-αCD180, there were on average 590 NP-specific IgG ASCs in the spleen, compared with more than 10-fold fewer ASCs (mean 45 NP-specific IgG ASCs) in mice inoculated with NP-αCD40 (Fig. 7E, left). Similarly, there were significantly more NP-specific ASCs in the BM of mice inoculated with NP-αCD180 compared with NP-αCD40 (18.9 vs. 2.1 per 10^6 BM cells; Fig. 7E, right). Collectively, these data demonstrate that Ag targeting to CD180 induced a more rapid, robust, high affinity and long-lasting Ag-specific Ab response than targeting Ag to CD40.

Discussion

Herein we have demonstrated how Ag-αCD180 robustly activates multiple B cell subsets to induce a rapid, high affinity and long-lasting Ag-specific IgG Ab response. Immediately following inoculation of Ag-αCD180, the Ag-specific BCR is rapidly internalized on all B cell subsets. Subsequently, Ag is processed and presented on MHC II, and the co-stimulatory ligand CD86 is induced. These Ag-specific B cells then migrate to T-B borders within a day or two and proliferate in a T cell-dependent manner. Ag-specific T cells receive signals leading to their maturation into T_FH cells; a process that requires B cells and that immature B cells are sufficient to induce. The initial clonal expansion of B cells includes in part an immature B cell population that most likely seeds the germinal centers (GCs) evident by day 7 (2). Intriguingly, targeting to CD180 induced an earlier and more robust Ag-specific Ab response than targeting to CD40.

Critically, the key events described above are dependent on the physical connection between the Ag and the αCD180 Ab, as the unconjugated control (Ag-Iso + αCD180) failed to induce Ag presentation, Ag-specific T cell activation or Ag-specific B cell proliferation. These findings are somewhat counterintuitive, as Ag-specific B cells exposed to Ag-Iso + αCD180 should still receive signals through both the BCR and CD180. The divergence in response and ultimate inability of separated Ag and αCD180 Ab to induce a robust Ab response is most likely due to differences between Ag-αCD180 vs. Ag + αCD180 in Ag internalization and subsequent processing and presentation. After encountering an Ag, the BCR induces a signaling cascade that includes the phosphorylation and activation of downstream kinases as well as the adaptor protein 2 (AP-2) complex, which is critical for the initiation of clathrin-mediated endocytosis (22). AP-2 binds to the BCR via an YxxΦ motif on the Igαβ subunit, Φ being a bulky hydrophobic amino acid such as leucine or isoleucine (23). Intriguingly, the short cytoplasmic tail of CD180 contains a conserved YxxI sequence (5); thus, this sequence could potentially provide an additional partner for AP-2 binding and thereby facilitate the more rapid BCR internalization observed following Ag-αCD180 inoculation (Fig. 1). Activation of Lyn downstream of the BCR directly phosphorylates the clathrin heavy chain promoting the coupling of clathrin to actin, while other elements of BCR signaling, such as Vav, Bam32 and Btk, are also critical for endocytosis (22, 24). The Lyn and Btk kinases are also activated following the ligation of CD180 on the surface of B cells (5). Thus, it is conceivable that the binding of both the BCR and CD180 through the Ag-αCD180 complex enhances internalization by bringing together critical signaling partners. Further studies are required to assess this possibility.

Once the B cells have acquired the Ag via direct delivery from Ag-αCD180, they downregulate CXCR5, travel to B-T borders within one day and rapidly begin to proliferate (Fig. 3). This initial proliferation requires T cells (Fig. 4), and may well be independent of DCs. In support of this hypothesis, we previously demonstrated that the majority of the Ag-specific IgG response 10 days after Ag-αCD180 inoculation requires the expression of CD180 on B cells but not DCs (2). Scandella and colleagues observed a similar phenomenon following vesicular stomatitis virus infection, wherein vesicular stomatitis virus-specific B cells traveled to T-B borders rapidly, proliferated and produced neutralizing Abs in mice that had been depleted of CD11c⁺ DCs (25). Furthermore, it has been demonstrated that naïve CD4⁺ T cells are predominantly activated by Ag-presenting B cells and not DCs following virus-like-particle vaccination (26). Regardless, both B cells and DCs from CD180 targeted mice are capable of presenting Ag to cognate T cells. The consequences of Ag presentation by B cells are multifaceted and distinct from those of DCs. To start with, the peptide/MHC II complexes derived from BCR-associated Ag acquisition are predominantly of the rare M1 MHC II conformation, defined by the Ia.2 epitope (27). This MHC epitope is associated with the robust activation of CD4⁺ T cells (28). It is tempting to speculate, therefore, that CD4⁺ T cells activated by B cells are qualitatively different than those stimulated by DCs. This may be one potential explanation for the higher affinity Abs following Ag-αCD180 inoculation than Ag-αCD40, given the ability of CD180 targeting but not CD40 targeting to induce B cells to present Ag to T cells.

The differential effects of B cell- versus DC-derived T cell activation are not fully understood. The combined APC action of B cells and DCs leads to more robust proliferation and cytokine production by CD4⁺ T cells than when DCs or B cells alone are capable of presenting Ag (29). Furthermore, B cells are required for the maturation of T_FH cells (30). Indeed, while the initial upregulation of CXCR5 on CD4⁺ cells relies on DC MHC II expression, only the combined APC function of both DCs and B cells is sufficient to induce fully mature CXCR5⁺, PD-1^hi T_FH cells following Ag + alum stimulation (31). This finding corresponds with our data, in which we observed Ag-specific CXCR5⁺, PD-1⁺ T_FH cells in WT mice but not B cell deficient μMT mice following Ag-αCD180 inoculation (Fig. 5). It is likely that the combined Ag-presentation and expression of T cell co-stimulatory molecules such as CD86 following CD180 targeting (Fig. 2) are driving this reaction. Indeed, prolonged T/B interactions at the follicle border which are necessary for both the production of extrafollicular plasmablasts and the seeding of GCs are dependent on cognate interactions, suggesting that the B cell MHC II/TCR interaction is critical (32, 33). This is partly due to TCR-dependent activation of ICOS on the T cell, which is critical for T_FH cell development and maintenance (32). However, it is also possible that CD180 antigen targeting influences the cytokine profile of Ag-specific B cells which may influence the maturation of T_FH cells. Additional studies have demonstrated that while initial CD4⁺ T cell activation and expansion may rely on DCs, the APC function of B cells is critical to the formation of memory T cell populations (34). Furthermore, BCR-related Ag-presentation by B cells can break CD4⁺ T cell tolerance (26), a critical step for many immunotherapeutic vaccination strategies including the treatment of patients chronically infected with hepatitis B virus (35). Thus, the combined APC action of both B cells and DCs following Ag-αCD180 inoculation highlight a strength of this vaccine strategy in producing robust and sustained adaptive immune responses.

The effect of targeting Ag to APCs has been explored for many years. The majority of strategies have been to target DCs through a number of different receptors (36–39). Many of these platforms excel at inducing Ag-specific CD8⁺ T cells, but few are able to elicit robust TD Ab responses without the addition of an adjuvant (37). For example, targeting to DEC-205 is only able to activate Ag-specific CD4⁺ T cells and induce Ag-specific IgG in combination with additional signals such as αCD40 (37). Indeed Ag targeting to DEC-205 without an adjuvant induced Ag-specific tolerance (1, 40). Directing Ag to the DC receptor DCIR2 induces DCs to hand-off Ag to and activate B cells, which in turn become potent APCs, but are unable to promote GC formation without the addition of a TLR7- or TLR9-based adjuvant (41). Park and colleagues demonstrated that targeting influenza Ag to Clec9A is able to protect mice from death but not severe disease following a lethal influenza challenge; boosting mice with Ag + CpG before challenge lessened morbidity significantly (42). By comparison, our studies show that a single immunization targeting Ag to CD180 without an adjuvant is sufficient to induce robust TD dependent Ag-specific IgG responses and protective Ab responses (4), a regimen with the potential to lessen off-target or adverse events.

While many studies have focused on targeting to CD40 and CD40-based adjuvants, since CD40 signaling can robustly activate APCs, there are some concerns of adverse off-target effects with this platform, such as cytokine release syndrome (43). Adding to this concern, our data reveal that targeting to CD40 does not efficiently activate Ag-specific B cells to become APCs, but rather causes robust non-specific B cell proliferation (Fig. 6D). In contrast, CD180 targeting effectively and specifically activated Ag-specific B cells to become APCs. This represents a fundamental difference between these two strategies of targeting innate immune programing (via CD180) vs. activating adaptive programming (via CD40) (Figure 8). CD180 targeting programs Ag-specific B cells to seek cognate T cell interactions, while CD40 targeting itself mimics T cell help, thus removing the checkpoint of Ag-specific T cell help. Therefore, the early induction of Ag-presentation in B cells by CD180 targeting results in rapid maturation of T_FH cells, not observed following CD40 targeting (Fig. 7C). This may explain the more rapid and higher affinity Ag-specific IgG produced in response to CD180 targeting compared to CD40 (Fig. 7D). In addition, the generation of long-lived plasma cells (LLPCs) relies on T_FH interactions in GCs (44). The lack of LLPCs in the BM following CD40 targeting (Figure 7E) suggest that GCs may not form or are severely limited. Although CD40 targeting induced moderate affinity maturation, this may not necessarily require formation of GCs, as several studies have found evidence of GC-independent somatic hypermutation (45–47). Furthermore, isotype-switched plasmablasts can form without any specific Th subset, including T_FH, but require a CD40 signal (48). Thus, mimicking CD40-CD40L interactions following CD40 targeting, while capable of inducing isotype-switched plasmablasts, do not fully recapitulate a cognate interaction and are therefore not sufficient to drive a robust GC reaction and LLPC formation. The reduction in LLPCs in CD40-targeted mice compared to CD180-targeted mice suggest that long-term term B cell responses are more robust in response to CD180. Indeed we have previously demonstrated secondary Ab responses following CD180 targeting are much higher than Ag + Alum inoculation (2). However, following a prime-boost vaccination scheme utilizing a replication-deficient adenovirus expressing CD40L and respiratory syncytial virus L, Ag did lead to a boost in nAbs and a reduction in virus titer post virus challenge (49). Further studies are required to define more fully the differences in secondary B cell responses would be following CD180 vs. CD40 targeting.

Figure 8. Comparing CD180 vs. CD40 targeting.

(1) Upon interaction with a CD180 targeting construct, cognate T1 and FO B cells internalize, process and present Ag on MHCII along with concomitant CD86 expression. FO B cells also downregulate CXCR5 and migrate to the T-B border. It is unclear whether T1 B cells also migrate towards the T cell zone. (2) Once at the T-B border, CD180 targeted activated cognate B cells require cognate interactions and begin to proliferate. (3) These clonal offspring seed germinal centers. (4) Cognate T cells that interact with CD180 targeted B cells, either mature or immature, expand and rapidly mature into T_FH with upregulation of CXCR5 and PD-1. (5) The end result of these interactions is the production of high affinity Ag-specific IgG and long-lived plasma cells. (6) Non-cognate B cells that interact with the Ag-αCD180 are capable of processing and presenting Ag. However, they do not receive a signal through the BCR and therefore do not downregulate CXCR5 and therefore do not travel to the T-B border. These cells undergo limited proliferation due to CD180 signaling. (7) By contrast, CD40 targeted cognate B cells do not present Ag and are not programmed to seek out cognate T cell interactions. There is limited evidence of Ag-specific B cell proliferation and (8) CD40 targeted B cells do not interact with cognate T cells, thus by day 3 T_FH cell maturation is not evident. (9) The result of CD40 targeting is the production of moderate affinity Ag-specific IgG and short-lived plasma cells which may form independently of germinal centers. (10) Non-cognate B cells, however, robustly proliferate due to CD40 signaling.

Another important feature of CD180 targeting is its ability to target and activate immature or newly formed B cells. Immature B cells migrating from the BM first travel to the spleen as T1 B cells (50, 51). Typically, T1 B cells are programmed to undergo apoptosis upon BCR stimulation; indeed this may be a mode of continued negative selection within the spleen (52). However, T1 B cells can be activated by other signals, including TLR ligation or IL-4 stimulation (53–56). Furthermore, BCR-stimulated T1 B cells can be rescued from apoptosis with the addition of certain signals such as IL-4 or CD40 (53, 57). CD180 appears to have a similar effect on T1 B cells given the expansion of this population observed following BCR + CD180 stimulation in the CD180 targeted B cells (Fig. 4C). The TLR-related functions of T1 B cells are largely T-cell independent; there is little evidence of T1 B cells directly interacting with T cells. Indeed, Chung and colleagues reported that T1 B cells can present Ag on MHC II, but that they fail to upregulate CD86 and therefore are poor stimulators of Ag-specific T cells (58). Ag-αCD180 activation of T1 B cells appears to bypass this limitation, as T1 B cells not only up-regulate CD86 and MHC II but also are quite potent stimulators of Ag-specific T cells. Intriguingly, after Ag-αCD180 inoculation, we observed mature T_FH cells in BAFFR^−/− mice but not μMT mice, suggesting, for the first time, that even newly formed B cells are able capable of providing maturation signals to developing T_FH cells. Furthermore, the initial clonal B cell expansion that occurs following CD180 targeting is of a select population of cells at the T1-FO interface (Fig. 4B), a population not observed following CD40 targeting (Fig. 6A). Our previous data suggests that these are, in fact, activated immature B cells. B cells from BAFFR^−/− mice, which in the absence of BAFF signaling fail to mature beyond the T1 stage, are responsive to Ag-αCD180 vaccination, forming Ag-specific plasma cells. This response results in Ag-specific IgG production that is sufficient to protect mice from a lethal viral challenge (4).

The ability to achieve protective immunity through targeting immature B cells demonstrates a significant advantage of this vaccination platform in situations where the B cell repertoire is immature. For example, the B cell compartment of neonates has yet to fully mature thus contributing to reduced humoral responses in infants and young children following vaccination (59). In addition, patients receiving B cell depletion therapies, such as Rituximab for rheumatoid arthritis alleviation, respond poorly to influenza virus and pneumococcal vaccination (60, 61). Having a vaccine platform that is capable of acting on the first newly formed B cells to emerge from the BM may be a successful strategy for protecting these immunocompromised individuals. A potential concern of CD180 targeting transitional B cells is the activation of bystander B cells. Although in the B1-8^Hi system this bystander activation was limited (Fig. 2, S3), and certainly was lower following CD180 targeting compared to CD40 (Fig. 6), these reactions might be more evident in an un-biased system. However, given that CD180 activation renders B cells very sensitive to BCR-induced apoptosis, much more so than CD40 activated B cells (65), we believe that potential adverse events due to bystander CD180 activation would be limited.

Another advantage of targeting immature B cells is the potential range of BCR reactivity within these cells. Sequences of the light chain in IgHμ transgenic mice demonstrated a more restrictive Ig repertoire on mature B cells compared to immature B cells (66). Consistently, it has been estimated that, while selection of B cells begins in the BM, approximately 1/3 of peripheral tolerance occurs at the late transitional stage within the periphery (67). Furthermore, a study of circulating human B cells identified a unique heavy chain repertoire in transitional B cells as compared to both immature BM and naïve mature B cells (68), suggesting that both positive and negative selective pressures may be acting upon peripheral newly formed B cells. Some of this diversification may occur in the gut; studies in germ-free mice reveal that the mucosal commensal reactive Ig repertoire is defined within T1 B cells (69). Similarly, T2 B cells within human gut-associated lymphoid tissue have been found to be activated and express mutated IGHV genes (70). Thus, interaction between T1/T2 B cells with commensal antigens drives diversification of the Ig repertoire. Indeed, T1 B cells, unlike FO B cells, constitutively express activation-induced cytamine deaminase and undergo somatic hypermutation in BM (21, 54, 55, 71). Thus, T1 B cells are attractive targets in the hunt to generate methods to induce broadly neutralizing Abs by vaccination. It has been demonstrated, for example, that germline targeting of human naïve B cells can induce the production of broadly neutralizing Abs against HIV (72). Therefore, the combined ability to activate B cells as APCs and to target immature B cells via CD180 Ag targeting make this vaccine platform ideal in forming robust humoral and cellular immunity especially in immunodeficient or immunosuppressed individuals or with the goal of generating broadly neutralizing Abs.

Key points.

• CD180 antigen targeting facilitates antigen presentation by B cells to CD4⁺ T cells.

• Immature B cells are sufficient for T follicular helper cell maturation.

• Compared to CD180, CD40 targeting induces slower and lower-affinity IgG responses.

Abbreviations

AP-2

Adaptor protein 2

ASC

Ab secreting cell

B6

C57BL/6

BAFF(R)

B cell activating factor (receptor)

BM

Bone marrow

DC

Dendritic cell

FO

Follicular

GC

Germinal center

Iso

Isotype

LLPC

Long-lived plasma cell

MZ

Marginal zone

NP

4-hyrody-3-nitrophenacetyl

T(1/2)

Transitional (1/2)

TD

T-cell-dependent

T_FH

T follicular helper cell

WNVE

West Nile Virus E protein