Maternal helminth infection causes dysfunctional B cell development in male offspring

Abstract

Infections during pregnancy are known to trigger alterations in offspring immunity, often leading to increased disease susceptibility. Maternal helminth infections correlate with lower antibody titers to certain childhood immunizations and putative decreased vaccine efficacy. The mechanisms that underlie how maternal infection blunts offspring humoral responses are unclear. Using our murine model of maternal schistosomiasis, we found that maternal helminth infection decreases the germinal center response of all offspring to tetanus immunization. However, only male offspring have defects in memory B cell and long-lived plasma cell generation. We found this sex-specific aberration begins during B cell development within the bone marrow via alteration of the IL-7 niche and persists throughout antigenic activation in the germinal center in the periphery. Critically, these defects in males are cell intrinsic, persisting following adoptive transfer to control male pups. Together, these data show that maternal infections can alter both the bone marrow microenvironment and the development of B lymphocytes in a sex-specific manner. This is the first study to correlate defects in early life B cell development with ineffective antibody response after vaccination during maternal infection.

Introduction

Despite high vaccination coverage, infectious diseases remain prevalent around the globe, with vaccine efficacy differing across regions (1, 2). In some regions of Africa where parasitic infections are endemic, efficacy of common childhood vaccines is lower than in regions where parasitic infections are not as prevalent (3, 4). This suggests that the bystander immunosuppression driven by parasite infection (5) can impact humoral vaccine responses. Although causation has been observed in controlled murine models, due to the complexity of these interactions in humans, the magnitude of their impact in human populations and effective strategies to increase vaccine effectiveness in these regions are still unclear (6–8).

Schistosomiasis is a disease caused by a parasitic helminth that induces Th2 skewing during chronic infection. This is characterized by increased production of IL-4, a canonical Th2 cytokine, paired with a downregulation of IFNγ, a Th1 cytokine (9), both of which are necessary for the generation of a fully protective response to vaccination with most commercial vaccines (10, 11). While an inappropriate excess of either cytokine can prolong virus clearance (12), they can also decrease chronic, autoimmune-induced inflammation (13), indicating the role of the cytokine milieu in modulating immunosuppression.

Offspring born to mothers infected with parasites can also mirror these immunosuppressive outcomes, such as decreased responsiveness to antigens and protection from the development of some allergic and other chronic inflammatory diseases (14–16). Yet, contrary to primary patent infection, a decrease in Th2 cytokines, such as IL-4, is observed in mouse models of maternal schistosomiasis, due to epigenetic regulation of the transcription of these cytokines (14, 17). This suggests the immunosuppressive responses seen in offspring born to infected mothers is not due to excessive offspring Th2 cytokine production, but other unexplored mechanisms.

Maternal helminth infections, such as schistosomiasis, can impact offspring vaccine efficacy to common childhood vaccines (15, 16, 18). Maternal schistosomiasis has a prevalence rate varying between 1–60% across regions of Africa (19). In these studies, the children tested are not and have not been infected with schistosomiasis but have decreased vaccine efficacy (18), increasing the risk of a vaccine preventable disease outbreak in these communities. We have previously published with our murine model of maternal schistosomiasis, that after vaccination, pups from infected dams have transcriptional dysregulations of key B cell transcription factors, such as Ebf1 and Junb (17), restricting the germinal center reaction and decreasing vaccine titers.

An additional factor that has been known to cause differences in vaccine outcomes is biological sex. In both children and adults, biological females have a more robust humoral response to vaccination (20). Although sex hormone levels can have some effects on vaccine response (21–23), these differences are already evident during childhood when sex hormone levels are low. Therefore, other factors, such as gene regulation or maternal environment are likely responsible and have a long-lasting effect on vaccine outcomes.

In this study, we use this model of maternal schistosomiasis (17) to further understand how maternal infections impact offspring humoral immunity. First, we found that males born to Schistosoma mansoni infected mothers, and not females, have decreased humoral responsiveness to vaccination, and validated that this is comparable to human epidemiological data. This deficiency was due to a defect in steady state bone marrow B cell development and a sex-specific alteration in IL-7 receptor signaling. Additionally, central tolerance is also altered in a sex-specific manner, leading to an increase in receptor editing and unique clones in males from infected dams. We found these sex-specific defects to be B-cell intrinsic and to persist when transferred to control pups. This is the first study to show a link between defects in bone marrow B cell development and the generation of neutralizing antibody.

Materials and Methods

Mice

4get homozygous (Il4^tm1Lk, strain# 004190) mice and Rag1^−/− (C.129S7(B6)-Rag1^tm1Mom/J), strain# 003145) mice were obtained from The Jackson Laboratory and bred at the University of Utah animal facilities. Experiments were performed in strict accordance with the NIH guide for the Care and Use of Laboratory animals and institutional guidelines for animal care at the University of Utah under approved protocols (18–09001 and 21–07009). At six weeks of age, 4get homozygous females were infected with infected with ~35 Schistosoma mansoni cercaria, as previously described (17), or sham infected. Six weeks post S. mansoni or sham infection, females were paired to KN2 homozygous (Il4^tm1(CD2)Mmrs, a gift from Markus Mohrs (25)) males and remained paired until they reached humane endpoints. To generate 4getKN2 and KN2 homozygous littermates, 4getKN2 females were infected and mated to KN2 homozygous males as described above. Pups from mixed genotype litters were genotyped by flow cytometry before use.

In vivo treatments

Pups were weaned and genotyped (as necessary) at 28 days old. Steady state experiments were performed between 28–35 days of age. For immunization experiments, pups were immunized with 1/10^th the human dose of Tetanus/Diphtheria commercial vaccine (Grifols Therapeutics Inc (TDVAX) #13533013101) subcutaneously in the rear footpad. Mice were sacrificed at 14-days post-primary immunization or administered a secondary immunization at over 60 days post-primary immunization and sacrificed 3 days post-secondary immunization. For all experiments, littermates were used, and groups were split by sex and infection status of the mother.

Adoptive cell transfers

Steady state experiments were performed by harvesting whole bone marrow from male 4getKN2 pups from infected and control mothers as described above. Cell suspensions were pooled before counting using a hemocytometer. Three million cells were then injected intravenously into 4–6-week-old Rag1^−/− recipients. Recipients were bled weekly to check lymphocyte reconstitution. At 5 weeks post-transfer, recipients were euthanized, and cells were collected for flow cytometric analysis.

For immunization experiments, Rag1^−/− recipients were partially irradiated with 300 grays. The following day, donor cells were collected from bone marrow of males from infected and control mothers. Mature cells (CD3+CD4+CD8a+B220+CD11b+CD11c+Gr-1+Ter119+) were depleted by using the EasySepTM FITC Positive Selection Kit II (Stem Cell #17682) per manufacturer’s instruction. Non-labelled cells were then counted using a hemocytometer and 5×10⁵ cells were transferred intravenously into each Rag1^−/− male recipient mouse. Recipients were allowed 5 weeks to reconstitute before transferring 5×10⁵ T cells from wildtype 4getKN2 females. The following day, mice were immunized in the footpad with 1/10th the human dose of the commercial Tetanus/Diphtheria vaccine listed above. Fourteen days after immunization, mice were euthanized, and cells were harvested for analysis.

Cell isolation for single cell suspension

For bone marrow harvesting, femurs were collected, and the tip connected to the knee was cut on a bias to expose the bone marrow. Bones were placed cut side down into a 200μL microcentrifuge tube with a hole in the bottom and ~50μL of Dulbecco’s Modified Eagle Medium (DMEM)(Corning #10–013-CV). This tube was then placed into a 1.5mL microcentrifuge tube and then pulsed in a microcentrifuge at ≥ 10,000 RPM for 15 seconds at 4°C. The 200μL tube with the flushed bones were then discarded. Erythrocytes were lysed using 1X lysis buffer (10X BD Pharm Lyse^™, BD Biosciences #555899, diluted to 1X) for 1 minute with mechanical disruption. The lysis was quenched using 1% fetal bovine serum (FBS)(Gibco #26140–079) in DMEM and centrifuged at 200xg for 5 minutes at 4° C. The supernatant was discarded and washed with 1X phosphate-buffered saline (PBS) before cell surface staining.

Bone marrow mesenchymal stromal cells (MSCs) were isolated and stained as previously described (84). In short, Bone marrow was flushed as described above. Then bones were chopped into small pieces and incubated with 1mg/mL of collagenase I (Sigma-Aldrich #SCR103) for half an hour at 37°C with shaking. After incubation, cells were flushed with PBS and washed before being passed through a cell strainer. Cells were then stained for flow cytometry as described below. MSCs are defined as Dump⁻B220⁻CD45⁻Sca-1⁺CD29⁺CD146⁺.

Popliteal lymph nodes (PLNs) and spleens were collected and processed as previously described (11). In brief, lymph nodes were collected, smashed through a 100μm cell strainer, and washed with 15mL of DMEM before staining. Spleen were harvested, smashed through a 100μm cell strainer, and washed before being lysed as described above, except for 5 minutes. Then cells were washed and stained as described.

Flow cytometry and antibodies

To stain cells for flow cytometry, cells were harvested as described above and resuspended in the appropriate volume of FACS buffer (2%FBS, 5mM EDTA in 1X PBS) as previously described (11, 17). The following antibodies from Invitrogen were used for cell surface staining: CD24 Super Bright 600 (M1/69), CD19 Super Bright 780 (1D3), CD19 PE-Cy5 (1D3), CD95 AF488 (15A7), CD11c FITC (N41B), CD3 FITC (17A2), Ter-119 FITC (TER119), B220 PerCP-Cy5.5 (RA36B2), Sca-1 APC (D7), IgM PE-Cy7 (II/41), CD249 PE (6C3), CD45 BV786 (30-F11). The following antibodies from Biolegend were used for cell surface staining: Zombie Red^™ Dye (for viability), Streptavidin APC/Fire^™ 750, CD127 BV711 (A7r34), GR-1 FITC (RB6–8C5), B220 PerCP-Cy5.5 (RA36B2), CD11b FITC (M1/70), CD79a APC (F11–172), CD179a PE (R3), CD38 AF700 (90), Ig light chain κ APC-Cy7 (RMK-45), Ig light chain λ PE (RML-42), Sca-1 APC/Fire^™ 750 (D7), CD146 PE (ME-9F1). The following antibodies from BD Biosciences were used for cell surface staining: IgD BV510 (11–26c.2a), CD43 BV650 (S7), GL7 PE (GL7), CD138 Biotinylated (281–2), IgG1 PE (A85–1), CD29 BV605 (HM β1–1).

For transcription factor staining, cells were fixed and permeabilized with the FoxP3 transcription factor staining kit (Invitrogen # 00–5523-00) per manufacture’s instruction, and subsequently stained with combinations the following transcription factor-specific antibodies from BD Biosciences (EBF1 PE (T26–818)), Invitrogen (Ki67 APC (SolA15), Phospho- SYK APC (moch1ct), Phospho-NFkB p65 PE (Ser529) (NFkBp65S529-H3) and Phospho-Stat3(Tyr705) eFluor450 (LUVNKLA)), and Biolegend (PAX5 PerCP-Cy5.5 (1H9)).

For cytokine staining, the BD Cytofix/Cytoperm^™ Plus kit (BD Biosciences #555028) per manufacturer’s instruction. For IL-7 staining, unconjugated goat anti-mouse IL-7 (R&D Systems #AF407) was incubated with cells for 30 minutes at room temperature after fixation and permeabilization. Cells were washed and then stained with donkey anti-goat IgG conjugated to AF647 (Life Technologies #A21447) for one hour at room temperature and washed twice before analysis.

Bone marrow B cells gating was done after first gating out mature, non-B cells (Dump+)(Viability^+/−CD3⁺CD4⁺CD8⁺CD11b⁺CD11c⁺Gr-1⁺Ter119⁺). Pre–pro-B cells (B220⁺CD19⁻), pro-B cells (B220⁺CD19⁺IgM⁻IgD⁻CD43⁺), large pre-B cells (B220⁺CD19⁺IgM⁻IgD⁻CD43⁺FSC^hi), small pre-B cells (B220⁺CD19⁺IgM⁻IgD⁻CD43⁺FSC^lo), immature B cells (B220⁺CD19⁺IgM⁺IgD⁻), and transitional B cells (B220⁺CD19⁺IgM⁻IgD⁺) gating was adopted previously published studies (29, 85, 86). Germinal center B cells are described as Viability⁻CD19⁺GL-7⁺FAS⁺, while antigen-specific or Tetanus-specific germinal center B cells are germinal B cells that are further gated using FITC conjugated Tetanus toxoid. Memory B cells are described as CD19⁺IgM⁻IgD⁻CD38⁺. Long lived plasma cells (LLPCs) are gated as IgD⁻CD19⁻CD138⁺ and plasmablasts are gated as IgD⁻CD19⁺ CD138⁺. All flow cytometry experiments were run on an Attune NxT (Invitrogen), and data were analyzed with FlowJo.

Tetanus toxoid fluorophore conjugation

Tetanus toxoid (List Labs #191B) was conjugated to using the Lightning Link ^® FITC conjugation kit (Abcam #ab102884) per manufacturer instruction. Ovalbumin conjugated to Biotin (Abcam #ab201795) was used in conjunction with the conjugated tetanus toxoid to control for nonspecific B cell binding. The decoy was incubated with streptavidin-BV786 (BD Biosciences #563858) for 10 minutes at RT before mixing with the conjugated tetanus toxoid antibody. Cells were stained with 0.01μL of conjugated tetanus toxoid for 20 minutes 4° C and washed twice with 1X PBS before surface antibody staining.

Enzyme-linked immunosorbent assay (ELISA)

Anti-tetanus specific IgG1 endpoint titers were determined by enzyme-Linked immunosorbent assay (ELISA) using the mAb X56 (BD) and Immulon 4HB plates (Thermo Fisher Scientific) as previously described (17). Avidity ELISA were performed by following the above ELISA protocol with an additional 15-minute agitated incubation with 0.5M ammonium thiocyanate for treated wells or 15-minute incubation with 1X PBS for control wells before secondary staining as described in (87). Human data are an unpublished timepoint split by sex of the child from a human maternal schistosomiasis trial published in 2018 by Ondigo et al (18).

Tetanus neutralization assays

Modified tetanus neutralization assays(88) were performed by first coating an Immulon 4 HBX 96 well plate (Fisher Scientific #14–245-153) with 10μg/mL of trisialoganglioside-GT_1b from bovine brain (Sigma-Aldrich # G3767–1MG) in ethanol overnight. The next day, the plate was blocked with 1% BSA for two hours at room temperature. Meanwhile, serum from immunized pups were incubated with tetanus toxin for two hours at 37°C. After flicking off the blocking solution, the serum incubated with tetanus was added to the plate and let to sit at room temperature for two hours. After washing, anti-tetanus toxoid rabbit serum (VWR # BOSSBS-11772R) was added to the plate for two hours at room temperature. After washing, anti-rabbit IgG HRP (Abcam # ab6721) was added to the plate and left to incubate for two hours at 37°C. Finally, SuperAqua Blue ELISA substrate (Thermo Fisher Scientific #00–4203-56) was added to each well before reading plate at 405nm with spectrophotometer.

Bone marrow B cell cultures

Bone marrow cells were isolated as described above, under sterile conditions. All B cells were isolated from bone marrow using the Dynabeads^™ Mouse Pan B (Invitrogen #11441D) as per manufacture’s instruction. After isolation, cells were stained with CellTrace^™ Violet (Invitrogen #C34557) following manufacturers’ instruction, except staining at a cell density of 500,000 cells/mL. After a 20-minute incubation at 37°C, cells were washed and then plated into 6 well tissue cultured treated dishes at 2 million cells per well in 2 mL of culture media (RPMI1640, 10% FCS, 2mM L-glutamine and 1 IU/mL Pen-Strep). Each sample was incubated either with media alone or 10ng/mL of recombinant mouse IL-7 (Irvine Scientific #200–03-10UG). Forty-eight hours after culture, cells were harvested by washing wells with cold PBS before washing and staining for flow cytometric analysis.

V(D)J scRNASeq

Single cell suspensions were isolated from tissue as described above then sorted using either a FACSAria or a Sony MA900 cell sorter. Cells were processed as previously described (17). In brief, sorted cells (DAPI-B220⁺CD19⁺IgM⁺IgD^+/−) were pelleted by centrifugation and resuspended in 1% BSA (bovine serum albumin) in 1X PBS and counted by hemocytometer and resuspended to a concentration of 1,200 cells/μL to run on the 10x platform. Paired-end RNASeq (125 cycles) was performed via an Agilent HiSeq next-generation sequencer.

V(D)J scRNASeq analysis

Raw data from scRNASeq experiments in this manuscript can be found in the NCBI’s Gene Expression Omnibus database (GSE202403). Sequencing reads were processed by using a 10X Genomics CellRanger pipeline and further analyzed using R Studio. Prior to analysis, low quality cells (greater than 15% mitochondrial gene representation and/or fewer than 200 and/or more than 6000 genes per cell) were filtered out before clustering and differential expression testing (DEGs) by using the Seurat package (89–91). The biological identities of cell clusters were annotated by referencing a published scRNASeq data set (36) and by surveying known transcriptional cell markers in the scRNASeq data set. DotPlots were made using Seurat. V(D)J data sets were analyzed using 10X Loupe V(D)J browser, scRepertoire (92) and VDJView (93).

Computer software versions

R_4.2.1, dbplyr_2.2.1, devtools_2.4.4, dplyr_1.0.1.0, edgeR_3.39.6, ggplot2_3.3.6, ggrepel_0.9.1, HDF5Array_1.25.2, hdf5r_1.3.5, RColorBrewer_1.1–3, rhdf5_2.41.1, scRepertoire_1.7.2, sesame_1.15.8, Seurat_ 4.1.1, shiny_1.7.2, and wesanderson_0.3.6 were used.

Statistical Analysis

For scRNASeq experiments, p-values and adjusted p-values were calculated while doing differential gene expression analysis, which uses a non-parametric Wilcoxon rank sum test. All other statistical analyses were done using GraphPad Prism v8.0. For multiple comparisons, data was analyzed by analysis of variance (ANOVA) with Turkey’s multiple comparisons test. p-values ≤ 0.05 were considered statistically significant and is denoted in figures.

Results

Males born to schistosome infected dams have decreased tetanus/diphtheria vaccine efficacy

To determine if decreased vaccine efficacy in offspring born to schistosome infected mothers (18) is sex specific, we used our previously published maternal Schistosoma mansoni infection murine model. 4get homozygous females (24), were either sham-infected or infected with a low dose of S. mansoni. After the start of the egg laying phase females were paired to KN2 homozygous males resulting in dual IL-4 reporter pups (25). At 28–35 days old, aged-matched 4getKN2 pups from sham-infected (uninfected or control) and schistosome-infected dams were immunized with a commercial tetanus/diphtheria (Td) vaccine (Fig. 1A). We have previously shown that mice born to infected mothers have lower frequencies of germinal center (GC) B cells, memory B cells, and bulk plasma cells (17). It was hypothesized that these differences in humoral immunity were due to a defect in IL-4 producing cells within the draining lymph node after immunization. However, after splitting both the frequency of GC B cells (Fig. S1A) and IL-4 secreting CD4+ T cells (Fig. S1B) by sex, we see a sex-specific difference in GC B cells in males that is not mirrored by the IL-4 producing T cells. This suggests the GC B cell defect in males from infected mothers is independent of IL-4 secreting T cells within the draining lymph node. Although we decoupled the GC phenotype from T cell IL-4 production, we continued to utilize the 4getKN2 model due to increases in successful pregnancies and deliveries compared to infected C57BL6/J females and increased litter sizes compared to infected BALB/c females.

Figure 1:

Maternal schistosomiasis causes a sex-specific defect in the immunization response to Tetanus toxin. (A) Model of experimental outline for immunization experiments. (B) Flow cytometric gating strategy for antigen-specific germinal center B cells. (C) Frequency of antigen-specific germinal center B cells. Representative of >3 separate litters per group. (D) Flow cytometric gating strategy for bulk memory B cells. (E) Frequency of bulk memory B cells. Representative of >3 separate litters per group. (F) Anti-tetanus IgG1 titers from pups from 3 days post-secondary challenge response, as measured by ELISA. Representative of >3 separate litters per group. (G) Secondary analysis of serum antibody titers from a previously collected and published data set of children in Kenya born to schistosome infected and control mothers using published cut off values. (H) Tetanus neutralizing assay using serum. Data are representative of >2 separate litters per group. All statistics determined by two-way ANOVA and p-values are denoted on graphs.

A protective antibody response begins with GC formation and the emergence of high affinity antigen-specific cells from the GC reaction. Using fluorescently conjugated tetanus toxoid, we identified antigen-specific GC B cells (Fig. 1B, S1C). We found that both males and females from schistosome infected mothers have decreased frequency of antigen-specific B cells within the GC (Fig. 1C). This validated our previous work, which suggested that a lower total GC response decreased the number of antigen-specific cells within the GC (17). Although this result is not sex-specific, we found that key products of the GC reaction, memory B cells (Fig. 1C) and long-lived plasma cells (LLPCs) (Fig. S1D) are impacted in a sex-specific manner. Males from infected mothers have decreased memory B cell and LLPC frequencies, while plasmablasts (Fig. S1E) are unchanged. Interestingly, there is a slight increase in plasmablasts in females (Fig. S1E). During the primary vaccine response, plasmablasts are not derived from the GC response (26), indicating that the entirety of the antigen-specific response may not be altered in males from schistosome infected dams, but that the GC response in males from infected dams is particularly hindered.

Finally, to test the efficacy of primary response to vaccination, serum anti-tetanus toxoid IgG1 antibody titers were measured. Males from infected mothers had a decrease in antibody titer (Fig. S1F), consistent with the decrease seen in antibody secreting LLPCs. These antibodies do not have a difference in avidity (Fig. S1G) or non-specific binding (Fig. S1H). Males from infected mothers also do not have a decrease in anti-tetanus toxoid IgM antibodies (Fig. S1I), again indicating that the dysfunction occurs at the GC level. Overall, this illustrates males born to schistosome infected dams have decreased primary tetanus-specific vaccine response via defective a GC reaction.

Although the primary vaccine response was altered, the secondary, or recall response, can be more indicative of how effective the first vaccination was at priming immunity. To assess the recall response, pups from control and infected dams were immunized as above and rested for >60 days before secondary immunization. Three-days post-secondary vaccination, serum was taken, and anti-tetanus toxoid IgG1 titers were measured. Paralleling the primary response, the anti-tetanus toxoid IgG1 recall response was only decreased in males from schistosome infected dams (Fig. 1F). We then wanted to compare our findings to epidemiological data by utilizing data provided from a human maternal infection study done in Kenya (18). Although this study was not powered to look at the effect of biological sex of the offspring on vaccine induced immunity, we found that male children born to schistosome-infected women trend towards lowered tetanus-specific antibody titers at 18 months of age (Fig. 1G). Together, these data show that maternal schistosomiasis decreases vaccine-induced tetanus-specific antibody titers in male offspring.

To test the functionality of tetanus-specific antibodies, we performed tetanus toxin neutralization assays. We found that, consistent with current vaccine literature (27), males have increased antigen neutralization capability compared to females (Fig. 1H). Additionally, we found males from infected mothers have diminished neutralizing capability compared to males from control mothers, while females from infected mothers have the least capacity to neutralize tetanus toxin (Fig. 1H). Directly comparing the neutralizing capacity of littermate females and males, both sexes of offspring born to infected mothers produce less neutralizing antibody than their counterparts born to control mothers. In summary, although both males and females from schistosome-infected dams have decreased antigen-specific GC B cells, only males show a defect in the GC reaction. This results in lower numbers of antigen specific memory B cells, antibody secreting plasma cells and neutralizing antibodies, likely reducing vaccine efficacy compared to both control offspring and littermate females.

B cell development in the bone marrow effects peripheral B cell availability

Previously, we have linked steady state peripheral B cell numbers to diminished GC response (17). It is unclear if B cell development, which occurs in the bone marrow, is also failing, or if the deficit begins in the periphery after immunization. To begin this investigation, we first wanted to establish if there was a correlation between mature bone marrow B cells and naïve B cell frequencies in the periphery of mice at steady state. We focused on the popliteal lymph node (PLN), which is the draining lymph node after footpad immunization used in this study. We found a positive correlation between the frequencies of bone marrow and peripheral popliteal B cells at steady state (Fig. 2A). Furthermore, we found males from schistosome-infected mothers have significantly fewer naïve B cells in the periphery compared to males from control dams at steady state (Fig. 2B). To determine where the peripheral B cell deficiency begins, developing bone marrow B cell populations were assessed between pups from control and schistosome-infected mothers (Fig. S2). The frequency of pre-pro B cells, or the first stage of B cell lineage commitment in the bone marrow (28, 29), is not significantly altered between either sex comparing maternal infection status (Fig. 2C). Similarly, both pro- and bulk pre-B cells, the subsequent developmental steps before mature B cell receptor (BCR) expression and B cell selection, are not altered (Fig. 2E, 2F). Following light chain rearrangement, a mature BCR is present on the surface of these B cells that has yet to undergo selection. These are termed immature B cells (Fig. S3A). This is where the sex-specific difference in B cell development begins. Males from infected dams have a lower frequency of immature (Fig. 2H) and transitional/ mature (T1) B cells (Fig. S3B), without overall lower hematopoietic cellularity in the bone marrow (Fig. S3C). Overall, this shows that maternal schistosomiasis causes a sex-specific alteration in offspring B cell development, leading to a decrease in peripheral B cells.

Figure 2:

Defects in bone marrow B cell development affects peripheral B cell number. (A) Correlation plot of the frequency of T1 B cells in the bone marrow compared to the frequency of CD19⁺IgM⁺IgD⁺ B cells in popliteal lymph nodes. Each dot represents a single mouse. Experiment was performed two separate times, representative of four separate litters. (B) Quantification of number of CD19⁺IgM⁺IgD⁺ naïve B cells in popliteal lymph nodes at steady state determined by flow cytometry. Representative of 2 litters per group. (C) Representative flow plot showing gating strategy for pre-pro B cells in the bone marrow (left) with frequency quantification (right). Representative of >10 litters per group. (D) Representative flow cytometry plots showing gating strategy for pro-B cells, bulk pre-B cells, small pre-B cells, and large-pre B cells. (E) Frequency of pro-B cells, representative of >10 litters per group, determined by flow cytometry. (F) Frequency of bulk pre-B cells, representative of >10 litters per group, determined by flow cytometry. (G) Frequency of immature B cells, representative of >10 litters per group, determined by flow cytometry.

Large pre-B cells from males from schistosome-infected mothers are less responsive to IL-7 signaling

Although we determined that the sex-specific drop in B cell numbers begins at the immature B cell stage, it is unknown if this was caused by an inability to transition to immature B cells from pre-B cells, or if there is a defect in BCR expression, which occurs at this stage. To begin to understand this, we examined both large and small pre-B cells and their ability to proliferate and survive to continue developing into mature B cells. First, we focused on large pre-B cells, which is the proliferative stage of pre-B cell development after Ig heavy chain rearrangement (30). We found that by splitting the bulk pre-B cell population by size, there was no difference in large pre-B cell frequency (Fig. 3A), but we observed a decrease in small pre-B cell frequency only in males from infected dams (Fig. 3B), indicating a possible defect in proliferation in large pre-B cells.

Figure 3:

Maternal schistosomiasis decreases IL-7 available in the bone marrow for B cell development. (A) Frequency of large pre-B cells, representative of >10 litters per group, determined by flow cytometry. (B) Frequency of small pre-B cells, representative of >10 litters per group, determined by flow cytometry. (C) Histogram of IL-7 staining from stromal cells of bone marrow from pups from S. mansoni infected mothers and control mothers. Black peak represents cells stained with secondary anti-goat IgG-APC antibody without IL-7 primary stain. Histogram representative of >2 litters per group. (D) Adjusted MFI of IL-7 staining on bone marrow mesenchymal stromal cells (MSCs), determined by flow cytometry. MSCs gated as Dump-B220-CD45-Sca-1+CD29+CD146+. Adjusted MFI determined by averaging the MFI of IL-7 from MSC gate of males and females from uninfected mothers, then dividing all MFIs by average MFI to achieve consistency between experiments. Representative of >4 litters per group. (E) Adjusted MFI (as calculated above) of IL-7R on large pre-B cells on bone marrow as determine by flow cytometry, Representative of >4 litters per group. (F) Frequency of EBF-1+ large pre-B cells in bone marrow of mice born to control and infected dams. Representative of >3 litters per group. (G) Frequency of Ki-67+ large pre-B cells, determined by flow cytometry. Representative of >4 litters per group. (H) CellTrace^™ violet staining on large pre-B cells cultured with IL-7 for two days. (I) Frequency of CellTrace^™ high (undivided) large pre-B cells comparing paired unstimulated cultures with cultures with added IL-7. Representative of 3 separate experiments. (J) Frequency of undivided large pre-B cells (CellTrace^™ Violet^hi) after two-day culture of bone marrow B cells with IL-7. (K) Paired frequencies of large pre-B cells pre and post-culture with IL-7 by flow cytometric analysis of culture aliquots. (L) Number of differentially expressed genes (DEGs) between groups from V(D)J single cell RNA sequencing of large pre-B cells cluster. (M) Heatmap of gene expression from large pre-B cell cluster of V(D)J single cell RNA sequencing. Asterisk above gene name denotes a significant DEG between males from infected mothers compared to males from uninfected mothers. Statistics from flow cytometry calculated by two-way ANOVA. Statistics from RNA sequencing calculated by Seurat package(90) using non-parametric Wilcoxon rank sum test.

Large pre-B cell proliferation is dependent on the response to IL-7 being secreted from bone marrow mesenchymal stromal cells (MSCs) (31–33). To determine if the alteration in large pre-B cell frequencies in pups from infected mothers is due to cytokine availability, we measured the production of IL-7 by MSCs using flow cytometry (Fig, 3C). We found both males and females from schistosome-infected dams, had lower levels of IL-7 measured by MFI (Fig. 3D). Although the amount of IL-7 receptor (IL-7R) on the surface of large pre-B cells is unchanged (Fig. 3E), downstream signaling of IL-7R, including the transcription of EBF-1 (34), is only reduced in males from infected mothers, and not their female littermates (Fig. 3F). While both males and females from schistosome-infected dams have decreased IL-7, only males show a significant decrease in Ki-67 (Fig. 3F), a marker of proliferation that is highly expressed after IL-7R signaling (31, 32, 34). This indicates a decrease in large pre-B cell proliferation, possibly due to lower IL-7 availability.

To establish if large pre-B cells from males born to infected mothers have an altered responsiveness to IL-7, bone marrow B cells were isolated and cultured with and without IL-7. CellTrace^™ Violet staining was used to assess their capacity to proliferate with equivalent levels of IL-7 (Fig. 3H), where high CellTrace^™ Violet levels indicate undivided cells. All groups underwent significant proliferation of large pre-B cells when IL-7 was added to the cultures (Fig. 3I). When comparing the proliferative capacity of large pre-B cells, we found large pre-B cells from males from infected mothers and not females were less responsive to IL-7 and proliferated less (Fig. 3J), similar to the ex vivo data. Importantly, the overall frequency of large pre-B cells in vitro was not significantly changed (Fig. 3K). This, in addition to an absence of preferred media for differentiation and an incubation period of less than 3 days (35), indicates that the CellTrace^™ signal from large pre-B cells is from proliferation and not cell differentiation. These data demonstrate that males from infected dams not only have lower IL-7 secretion from MSCs, but also that large pre-B cells are less responsive to IL-7 signaling as measured by EBF-1 expression, and proliferation.

Downstream of IL-7R signaling is the transcription of genes necessary for B cell proliferation and survival, such as EBF-1 (Fig 3F, (31). To evaluate these transcriptional changes within large pre-B cells, single cell V(D)J RNA sequencing was performed on bone marrow B cells (Fig. S4A, S4B). Large pre-B cells were subset within the dataset before analysis based on published cell markers (36). Comparing differences in differentially expressed genes (DEGs) between groups, the biggest difference is between males from uninfected mothers and males from infected mothers (217 genes) (Fig. 3L). Surprisingly, although there is less IL-7, there were no DEGs between females from control and infected dams on the transcript level (Fig. 3L), suggesting that the IL-7R signaling cascade is transcriptionally equivalent. Interestingly, males from infected mothers seem to differ the most from each group (Fig. 3L), indicating a unique transcriptional regulation in these large pre-B cells. One transcript necessary for large pre-B cell development and downstream of IL-7 signaling is Tcf3 (37), encoding for the protein E2A. Males from schistosome-infected mothers have decreased Tcf3 transcript in large pre-B cells (Fig. 3M). Additional genes downstream of IL-7R signaling, such as Ebf1 and Ccnd3, which are important for large pre-B cell proliferation (31), are also decreased in males from infected mothers. Yet, other genes downstream of IL-7R signaling, such as Stat5b and Il7r are not altered in males from infected mothers (Fig. 3M). Overall, these data demonstrate that while maternal schistosome infection causes a decrease in IL-7 secretion by MSCs in all littermates, only males have associated reductions in downstream transcripts necessary for B cell development and large pre-B cell proliferation.

Decreased surrogate light chain expression on large pre-B cells does not alter receptor signaling during maternal schistosomiasis

Large pre-B cells can be characterized by their dependency on IL-7 signaling or pre-B cell receptor (pre-BCR) signaling (36). Earlier stages of pre-B cell development rely on IL-7 signaling from stromal cells to proliferate. As these cells migrate away from MSCs, and available IL-7 is reduced, they become dependent on the successful rearrangement of the heavy chain of the pre-BCR to survive and proliferate (38, 39). Since we found a proliferative defect in large pre-B cells, we wanted to know if this was solely due to reduced IL-7 signaling, or if pre-BCR signaling was also defective.

The pre-BCR is composed of two chains; the heavy chain, comprised of the rearranged Igμ, and the surrogate light chain, made of the proteins VpreB and λ5 (40) (41). Generally, the role of the pre-BCR is to assess the functionality of the Ig heavy chain, while also triggering further B cell maturation steps after successful immunoglobulin heavy chain rearrangement (42). Males from S. mansoni infected dams had lower VpreB on the cell surface, while there was no difference in expression in females from infected mothers (Fig. 4A).

Figure 4:

Pre-BCR signaling is not reduced due to maternal infection. (A) Adjusted MFI of VpreB on surface of large pre-B cells measured by flow cytometry. Adjusted MFI determined by gating large pre-B cells and then measuring the MFI of VpreB. Then, MFI values from pups from control mothers was averaged for each experiment and MFI values were divided by averaged control MFI for consistency between experiments. Representative of >4 litters per group. (B) Adjusted MFI (as described above) of phospho-SYK gated on large pre-B cells, determined by flow cytometry. Representative of >5 litters per group. (C) Adjusted MFI (described above) of PAX5 gated on large pre-B cells, determined by flow cytometry. Representative of >3 litters per group. All statistics calculated by two-way ANOVA.

Signaling through both the pre-BCR and mature BCR occurs through the transmembrane signaling molecules Igα and Igβ (43, 44). Downstream of this signaling is the phosphorylation of SYK (45), leading to the phosphorylation of STAT3 and STAT5 (46). In turn, these induce the transcription of genes necessary for Ig light chain rearrangement (47, 48). Although there is lower VpreB on the surface of large pre-B cells in males from infected dams, there is not a defect in the downstream phosphorylation of SYK in males or females from infected dams (Fig. 4B). Furthermore, there was no difference in protein expression of PAX5 (Fig. 4C), which is induced by the nuclear translocation of phosphorylated STAT5 (49) and is important for Igκ light chain rearrangement (50). These data indicate that maternal schistosome infection does not alter offspring pre-BCR signaling, and the decrease in small pre-B cells is a result of proliferative defects in large pre-B cells.

Ig light chain preference is skewed in males from schistosome infected mothers

Following light chain rearrangement, a mature BCR is present on the surface of immature B cells that has yet to undergo selection. We found that similar to the large pre-B cell phenotype, these immature B cells from males from schistosome infected mothers also have a decrease in EBF-1 expression, while female littermates retain normal expression levels (Fig. 5A). Because EBF-1 is essential for normal B cell development and function, we examined the transcriptional profile of these cells to determine if they had aberrant gene expression, We found males born to S. mansoni infected dams had an increase in transcripts related to activation after BCR signaling, such as Cd69 (51), Nfkb1 (52), and Fos (53) (Fig 5B). We also found an increase in phospho-NFκB in immature B cells in males from infected mothers (Fig. 5C).

Figure 5:

Inappropriate immature B cell activation during B cell development in males from S. mansoni infected mothers skews light chain preference. (A) Frequency of EBF1+ immature B cells determined by flow cytometry. Representative of >5 litters per group. (B) Dot plot from V(D)J single cell RNA sequencing showing canonical activation genes (Cd69, Nfkb1, and Fos) as listed below. Size of dot shows percent of cluster expressing those genes. Color represents average expression of each gene. White stars indicate adjusted p-value >0.0001when compared to all other groups by non-parametric Wilcoxon rank sum test. (C) Adjusted MFI of phospho-NFκB by flow cytometry from immature B cells. Representative of >3 litters per group. (D) Representative flow cytometric plot of Igκ and Igλ light chains on immature B cells in bone marrow. (E) Frequency of Igλ positive immature B cells by flow cytometry. Each group representative of >3 litters. (F) Frequency of Igλ positive T1 B cells by flow cytometry. Each group representative of >3 litters. (G) Frequency of Igλ positive naive B cells in spleen by flow cytometry. Each group representative of >3 litters. (H) Frequency of Igλ positive naïve B cells in PLN by flow cytometry. Each group representative of >3 litters. Flow cytometric statistics calculated by two-way ANOVA.

To determine how premature BCR activation affected selection, we stained for Ig light chain κ and λ (Fig. 5D). Igκ rearrangement occurs first during immature B cell development. During B cell central tolerance, the mature BCR is exposed to self-antigen to test reactivity. If the cell becomes activated, as is indicated in immature B cells in males from schistosome infected mothers, they are capable of undergoing receptor editing, allowing for the rearrangement of the Igλ chain (43, 54, 55). While there is no increase in Igλ usage in immature B cells between groups (Fig. 5E), there is an increase in usage in males from infected dams in T1 B cells (Fig. 5F). This is mirrored by Igκ gene usage (Fig. S4). This is carried out into the periphery, where males from infected mothers, and not female littermates, have an increase in Igλ usage in both the PLN (Fig. 5G) and spleen (Fig. 5H). With receptor editing and Igλ usage only occurring after central tolerance failure, these data suggest males from infected dams have an increased requirement for receptor editing.

Males from S. mansoni infected mothers have an increase in unique BCR clonotypes

To determine the effect of maternal infection on the outcome of recombination, clonotypes of immature and T1 B cells were examined by single cell V(D)J sequencing (Fig. S5). When comparing males from infected mothers to males from uninfected mothers, males from infected mothers have an expansion of unique clones, shown on the Y-axis (Fig. 6A). These unique clonotypes are not due solely to the environment during maternal infection because when comparing female littermates from infected mothers to females from uninfected mothers, most of their clonotypes are shared (Fig. 6B). Additionally, when directly comparing females and males from infected mothers with the same maternal environment, there is an increase in unique clonotypes in the males (Fig. 6C), suggesting that these clones are sex specific.

Figure 6:

Males from S. mansoni infected mothers have unique BCR clonotypes based on light chain usage and increased apoptotic signature genes. (A) ScatterPlots generated by scRepertoire comparing clonotypes of immature B cells between males from S. mansoni and control mothers and (B) females from S. mansoni and control mothers and (C) males and females from infected mothers. Single cell V(D)J sequencing was performed on B220⁺CD19⁺IgM⁺IgD^+/− bone marrow from pups from S. mansoni infected and control mothers. Each group represents at least 3 mice. Class groups clonotype based on if they are single or expanded and for which sample. Total n shows how many sequenced BCRs are represented by each dot, corresponding to dot size. (D) Heatmaps of Igκ gene pairings from single cell V(D)J sequencing. Scales represent number of times pairing was sequenced. Each group representative of pool of 3–4 mice. (E) Gene usage charts from VDJView from each sample group from single cell V(D)J sequencing showing λ gene pairings between variable and joining regions. White arrows on males from infected mothers indicate pairings not present in other samples. (F) Single clones from sequencing preformed in F were subset before using Seurat to calculate differentially expressed genes and p-values. Transcripts analyzed are listed to the left of the dots. p-values listed represent p-value between males from infected mothers compared to other three groups. Average expression and percent expression are the same as explained above.

Because of Ig light chain skewing, we looked at light chain gene usage. Looking at Igκ pairings, it becomes evident that males from infected dams have an increase in unique pairings and gene usage, as indicated by the decrease of zero counts (tan color, Fig. 6D). Additionally, when looking at Igλ pairings, males from infected dams have an increase in unique pairings (Iglv2 to Iglj3 and Iglv1 to Iglj2), as indicated by white arrows, that are not present in other groups (Fig. 6E).

Although it is unclear if these pairings are functional, cells with these unique clonotypes have a high expression of Casp9, Casp8, and Casp7 (Fig. 6F). After unsuccessful BCR crosslinking, such as after failed selection, cell death can be mediated by the Caspase-8 pathway (56). This induces the formation of the apoptosome, which results in the activation of Caspases 7 and 9 (57, 58), effectively inducing apoptosis. Non-expanded clones from male pups born to infected mothers had significantly higher expression of Casp8, Casp9, and Casp7 compared to non-expanded clones from other groups, indicating a higher rate of cell death in these non-expanded clones. Overall, males from infected mothers have alternative light chain pairings, leading to an increase of unique clonotypes and a higher rate of cell death.

Decrease in vaccine efficacy during maternal schistosomiasis is due to cell intrinsic B cell defects that begin in the bone marrow

To determine if the steady state decrease in B cells was dependent on the bone marrow microenvironment or if it was due to cell intrinsic factors, we performed adoptive cell transfers. Whole bone marrow from male pups from S. mansoni infected and control dams were transferred into Rag1^−/− mice, which do not have mature B or T cells. Mice were rested for 5 weeks to allow for reconstitution and then bone marrow and PLNs were harvested for flow cytometry (Fig. 7A). Resembling the phenotype in pups from infected and uninfected dams, the frequencies of immature, T1, and peripheral B cells is significantly lower in the mice that received bone marrow from males from S. mansoni infected mothers (Fig. 7B–D). Since the stromal cell environment is equal between mice that received bone marrow from each group, it can be concluded that the decrease in B cell frequency is due to cell intrinsic signaling defects.

Figure 7:

Maternal schistosomiasis induced defects in B cell number and immunization induced humoral immunity are cell intrinsic and begin in the bone marrow. (A) Model schematic for steady state adoptive cell transfer. (B) Frequency of immature and (C) T1 B cells in the bone marrow of Rag1^−/− mice after receiving bone marrow from 4getKN2 males from control and infected mothers. Experiments performed 3 times, representing >6 litters. (D) Frequency of CD19⁺IgM⁺IgD⁺ B cells in lymph nodes of Rag1^−/− recipients. (E) Schematic for adoptive transfer and immunization model. Flow cytometric frequencies of (F) GC B cells and (G) antigen-specific GC B cells from draining lymph node after adoptively transferring cells from males from control and infected mothers to Rag1^−/− recipients before immunizing with Td and harvesting 14 days post-immunization. Statistics calculated by student’s t-test. (H) Tetanus neutralizing assay (left) and area under the curve (right) from serum of Rag1^−/− recipients. Experiments repeated three independent times and represents >3 litters from infected mothers.

To establish if aberrant bone marrow B cell development in males could directly affect vaccine efficacy, we transferred only B220⁺ B cells from male bone marrow into irradiated Rag1^−/− mice. At 5 weeks post-transfer, splenic CD4⁺ T cells from control 4getKN2 mice were transferred to Rag1^−/− recipient mice so that all groups had an equivalent T cell pool. The following day, mice were immunized with Td, and the draining PLNs were harvested 14 days post-immunization, along with serum for antibody analysis (Fig. 7E). Mirroring the previous data from intact mice, mice that received B cells from male pups from infected mothers have a decrease in GC B cell frequency (Fig. 7F) and a decrease in the frequency of antigen-specific B cells within the GC (Fig. 7G) compared to mice that received control B cells. Finally, we found that antibodies made in mice given B cells from males from infected dams have decreased capacity to neutralize tetanus toxin (Fig. 7H), similar to intact male mice. Together, this demonstrates that a cell intrinsic defect in developing B cells induced by maternal infection can directly alter protective immune parameters in male offspring.

Discussion

Although it is well established that maternal parasitic infections can alter offspring immunity (14–17, 59–63), their direct effect on offspring B cell development is still unknown. Here, we find that males from S. mansoni infected mothers have a decrease in antigen-specific germinal center B cells, leading to a decrease in anti-tetanus toxoid IgG1 titers and lower toxin neutralization ability (Fig. 1). Primary parasitic infections can negatively impact vaccine effectiveness (64) and previous work extended this to maternal infections, where the offspring does not have an active infection (15, 65, 66). Maternal humoral immunity has previously been shown to impact offspring GC responses in the context of maternal vaccine-induced immune responses, where high maternal antibody titers can restrict offspring plasma cell generation and limit GC expansion to homologous immunization (67). Our model differs because the dams in this study are not immunized, but infected, and the pups are challenged with a heterologous commercial vaccine, suggesting a unique mechanism restricting GC function in pups from infected mothers. Importantly, we find that males from helminth infected mothers have a unique decrease in both antibody production and neutralizing function that suggests a decrease in vaccine efficacy, while females do not have a change in the overall titer (Fig. 1H). This suggests that the biological sex of the offspring can alter the susceptibility to infection induced modulation of immune function. This has been shown previously with offspring psychosis during maternal infection, where males have increased incidence and severity compared to their female littermates (68). Although the specific mechanism is unknown, it is proposed that there are differences in in utero micro-environments between male and female placentas, and that these differences can cause sex-specific developmental changes that persist into childhood. This is unexplored in the context of maternal infections and will be the focus of future work.

Interestingly, we found that in our murine model, females from infected mothers have an increase in plasmablasts (Fig. S1E), which are independent of the GC reaction. While there is a trend of lower frequencies of LLPCs in the females (Fig. S1D), and there is a significant decrease in antigen-specific GC B cells (Fig. 1C), there is not a change in antibody titers, as there is in males (Fig. 1F, S1F). Moreover, upon secondary challenge, females from infected mothers have an increase in anti-tetanus toxin IgG1 titers (Fig. 1F) and a trending decrease in tetanus toxin neutralization capability (Fig. 1H). Generally, males are known to have increased antigen neutralizing capacity, while females have higher antibody titers (27, 69, 70). This can be seen as a compensatory mechanism; where because there are more antibodies available in females, there is not as much pressure for them to neutralize antigen. Instead, females could produce additional antibodies that can readily aggregate or block antigenic binding (71), rendering them functional for anti-pathogen immunity, but not neutralization. Meanwhile, males have lower antibody titers but an increase in neutralization capacity, likely due to the necessity of antigen neutralization over aggregation. However, males from infected mothers not only have lower anti-tetanus toxin IgG1 antibody titers than males from uninfected mothers, but they also have less ability to neutralize tetanus toxin. This suggests that males from schistosome infected mothers do not have the compensatory mechanism that may exist in females, so that vaccine efficacy is likely diminished from maternal infection. If, as suggested by the data from the Kenyan cohort, this is true in humans, this increases the risk of an outbreak of a vaccine-preventable disease, especially among young male children.

This is the first study to connect modulation of developing B cells in the bone marrow to peripheral B cell functionality and immunization induced humoral immune responses. We demonstrate that a decrease in developing B cells in the bone marrow causes a decrease in the peripheral B cell pool that is long lived (Fig. 2A, 2B, S5), suggesting a steady turnover of naïve peripheral B cells, rather than an accumulation to overcome a shortage of cells during early life. Starting at the large pre-B cells stage, both males and females from infected mothers have a decrease in MSC-derived IL-7 (Fig. 3). It has previously been shown that maternal infection/inflammation can alter offspring production of IL-4 (14, 17), our current data extends this knowledge to a key homeostatic cytokine IL-7. Since offspring of both sexes are affected, we can postulate that the decrease in production of IL-7 by MSCs is driven by maternal factors and not offspring sex. While this current work has not determined the mechanism underlying this reduction, we hypothesize that it may be epigenetically regulated the way that IL-4 (14) is in this system, exploration of this possibility will be the subject of future work.

While IL-7 levels in the stromal cells of the bone marrow are decreased in both males and females, portions of the downstream signaling cascade are only affected in males (Fig. 3). There is no documentation of a sex-specific difference in IL-7 signaling or a difference in threshold of IL-7 for signaling via different transcriptional regulators in B cells the literature. However, females from schistosome infected mothers seem to be able to compensate for a decreased amount of stromal cell secreted IL-7, as shown by no difference in either IL-7-induced or any other transcripts in large pre-B cells (Fig. 3L) and normal levels of the key downstream protein EBF-1 (Fig. 3F). Meanwhile, males from infected mothers have a marked decrease in EBF-1 expression (Fig. 3F), and in vivo (Fig. 3G) and in vitro (Fig. 3J) proliferation. This supports the possibility that males from schistosome infected mothers have a differential threshold of IL-7 needed to induce downstream B cell proliferation and differentiation signaling cascades compared to female littermates.

We also found that large pre-B cells from males from schistosome infected mothers have lower surface expression of VpreB (Fig. 4A). But like what we see in females with IL-7R signaling, there seems to either be a compensatory mechanism or the level of reduction in VpreB does not cross the threshold required to disrupt normal downstream signaling events (Fig. 4B, C). There is very little known about differential signaling sensitivity through the pre-BCR and sex-based differences in pre-BCR and mature BCR signaling, but it can be speculated that males have decreased surrogate light chain signaling sensitivity. This would allow for a lower signaling threshold and lower surrogate light chain surface expression in males to trigger similar downstream events, as seen in females. More work is needed to determine sex-specific threshold requirements in this critical B cell developmental step.

Immature B cells are decreased in males but not females from infected dams (Fig. 2G) and have an unusual upregulation of immune activation genes (Fig. 5). Although there is little information about cell activation during selection, it has been shown that naïve B cells that have markers of activation can contribute to an autoimmune phenotype (72–74). Additionally, the increased skewing towards receptor editing, Igλ chain usage, and unique clone pairings in males from infected mothers (Fig. 5E–H, Fig. 6) can be hallmarks of autoimmune disease (55, 75). While there have recently been studies looking into maternal inflammation and the subsequent triggering of the onset of offspring autoimmune disease (76, 77), this link has not been studied in helminth infections. While primary helminth infections have been shown to be protective against autoimmune disease, due to generation of a suppressive anti-inflammatory response (4, 13, 78–82), it is unclear if children born to schistosome infected mothers have a skewed immune response towards inflammation or towards immunosuppression before primary schistosome infection.

The purpose of the primary immunoglobulin repertoire is to recognize a wide breadth of antigens. Although it is assumed that an increase in epitope diversity during the generation and development of adaptive immune cells would be advantageous during a primary infection or immunization, our study shows that an increase in B cell epitope diversity in the bone marrow does not correlate to increased vaccine efficacy. Using V(D)J sequencing, we found males from infected mothers have the most diverse BCR repertoire in the bone marrow (Fig. 6A–D). Yet, when immunized with Td, males from schistosome infected mothers have a decreased germinal center response and tetanus toxin neutralization ability (Fig. 1). We attribute the increase in epitope diversity in males from infected mothers to an increase in single clones. Since we find an increase in caspase transcripts (Fig 6F), we believe that these clones, which represent a part of the emerging repertoire, do not successfully complete selection and tolerance, and likely undergo apoptosis within the bone marrow. This would then lead to a decrease in both cell numbers and BCR diversity in the periphery and a restriction of the initial pool of BCRs available to respond to antigenic challenges such as immunization. Our data demonstrates that males have both a decreased antigen specific germinal center response and decreased neutralizing capacity, suggesting that this alteration in repertoire is functionally relevant. Two possible reasons these unique clones might not survive are: they could be autoreactive or anergic and are selected out, or there could be a cell intrinsic restriction of recombination, such as reduced expression of EBF-1 or STAT-5 (83). Since B cells have a second chance at receptor rearrangement in the germinal center reaction, future work will focus on tracking clones from the naïve bone marrow pool through the germinal canter reaction to understand if/which of these clones are dying in the bone marrow versus failing germinal enter selection in the periphery and their ability to lead to neutralizing antibody.

Overall, this is the first study to examine bone marrow B cell development during maternal infection and directly connect modulation of development to reduced efficacy of early childhood vaccines. While there are still mechanistic questions about the maternal and in utero microenvironmental drivers of modulation of the bone marrow niche, this study provides important insight into how developmental conditions can affect mature cell immune function, while also beginning to uncover possible differences in cytokine signaling thresholds that are sex specific. Future studies will be focused on the role of the maternal cytokine environment in offspring B cell development and possible modulation of autoimmune disease onset.

Key Findings.

• Pups from infected mothers have decreased IL-7 secretion in the bone marrow

• Males from infected mothers have a sex-specific defect in the B cell development

• Defects in male B cell development during maternal infection alter vaccine responses

Glossary

GC

germinal center

Td

tetanus/diphtheria commercial vaccine

BCR

B cell receptor

DEG

differentially expressed gene