Differential regulation of IgA+ B cells in vitro by stromal cells from distinctive anatomical compartments

B cell development is regulated by stromal cells (SCs) that form a supportive microenvironment. These SCs along with other cell types produce cytokines, chemokines, and adhesion molecules that guide B cell commitment and differentiation. BM, spleen (Sp), and the gut lamina propria (LP) constitute distinctive anatomical compartments that support B cell differentiation. In order to characterize and compare the signals necessary to generate IgA+ B cells, we developed an in vitro system to co‐culture gut LP, BM, or Sp‐derived SCs with B lineage cells. Using this co‐culture system, we found that gut LP SCs promote IgA+ B cell accumulation through the production of soluble stimulatory factors. In contrast to gut LP SCs, BM and splenic SCs were found to impair IgA+ B cell accumulation in vitro. Taken together, these observations provide new insights into how SCs derived from different anatomical locations shape IgA+ B cell responses.


INTRODUCTION
About 80% of the antibody produced in mammals is IgA. The main function of IgA antibodies is to maintain homeostasis. In response to the colonization of commensal microbes, B cells in the gut-associated lymphoid tissue (GALT) produce large amounts of IgA, a dimeric secretory antibody responsible for the blockade of bacterial penetration.
IgA class switch recombination (CSR) and plasma cell differentiation in GALT is influenced by metabolic products and cytokines produced by activated epithelial cells, stromal cells (SCs), dendritic cells, B cells, and T cells within the microenvironment. [1][2][3][4][5] Among all these cell types, the contribution of SCs has until recently been largely neglected. SCs are nonhematopoietic cells that form a supportive matrix, which can also influence the immune response. 6 Therefore, understanding the interaction between SCs and lymphocytes becomes important. In this study, we focused on SCs derived from gut lamina propria (LP), BM, and spleen (Sp).
Abbreviations: APRIL, a proliferation-inducing ligand; BAFF, B-cell activating factor; CM, conditioned medium; CSR, class switch recombination; FM, fecal matter; GALT, gut-associated lymphoid tissue; IFN R, interferon receptor; LC-MS/MS, liquid chromatography-tandem mass spectrometry; LP, lamina propria; MCP-1, monocyte chemotactic protein-1; MSC, mesenchymal stromal cells; SC, stromal cell; Sp, spleen/splenic; VEGF, vascular endothelial growth factor germinal centers. However, several studies have found that BM is a site for B cell maturation and T-independent IgM plasma cell differentiation in response to blood-borne microbes. 18,19 On the other hand, the Sp is a secondary lymphoid organ responsible for the maturation, activation, and differentiation of B cells. 20 Within the white pulp, follicular dendritic cells, fibroblastic reticular cells, and marginal reticular cells are the most studied SC types. 20 After infection or immunization, Sp follicular B cells preferentially switch to IgG, and IgG + plasma cells are found in the red pulp. 21,22 However, to add to the complexity, in mice, both B1 cells and marginal zone B cells in the Sp contribute to IgA production in vitro. 23 The reason why IgA CSR does not predominantly take place within the BM and Sp is largely unknown. It could be possible that BM and Sp microenvironments actively inhibit IgA + plasma cell generation.
In short, SCS from different sites constitute distinctive microenvironments for differential IgA responses. Therefore, we established an in vitro model for analyzing the effects of SCs on IgA + B cells that were generated from BM-derived B cells. This model enabled us to study and compare the influence of stromal cells derived from the gut LP, BM, and Sp on IgA + B cell development. We found that SCs from distinctive anatomical compartments exert different effects on IgA + B cell generation.

Stromal cell preparation
Female C57BL/6 mice were sacrificed and small intestines were harvested. Detailed protocols for intestine digestion have been described in our previous published article. 24 Cells from the small intestine were To prepare the SCs for staining, SCs were treated with PBS with 1 mM EDTA for 5-10 min at room temperature (25 • C) on a shaking platform. Remaining adherent SCs were lifted by cell scrapers. SCs were stained following manufacturer's instructions.

B cell-SC co-culture
B lineage cells were harvested from the femurs and tibias of C56BL/6 mice and were isolated using a CD19 positive selection kit (Mojosort Biolegend, San Diego, California, United States). SCs were trypsinized with 0.25% trypsin and irradiated at 2000 cGy. A total of 20,000 irradiated SCs were plated in a co-culture system with 200,000 CD19selected BM cells in 500 L complete medium supplemented with a cocktail composed of the following factors: IL-21 (

Flow cytometry
Antibodies used for flow cytometry analysis included FITC-anti-IgA

Liquid chromatography-tandem mass spectrometry
Conditioned medium for mass spectrometry analysis was prepared by plating 500,000 stromal cells from gut LP, BM, and Sp in 10 mL     number of total IgA+B cells in culture × 1000 cells. Statistical significance was assessed by two-way ANOVA

Differentiation of IgA + B cells is optimized in the presence of gut LP-derived SCs
To test whether SCs from distinctive anatomical compartments exert different effects on IgA + B cell generation, we followed our published protocol that promotes the differentiation of IgA + B cells in an in vitro culture system. 24 This entailed seeding BM-derived CD19-selected B lineage cells into cultures containing IL-7, anti-CD40, TGF , IL-21, fecal matter, and either gut LP (from small intestines), BM-or Sp-derived SCs for 6 days. Gut LP SCs strongly enhanced the accumulation of IgA + B cells (Fig. 1A). However, SCs established from the BM or Sp failed to provide similar support (Figs. 1A). This finding was based both on total IgA + B cells detected on day 6 of cultures using flow cytometry (Fig. 1A) as well as in an ELISPOT assay measuring IgA secreting cells (Fig. 1B).
We next examined the differentiation and proliferation of the IgA + B cells generated in vitro. The differentiation status of IgA + B cells was assessed by Blimp-1 expression. Blimp-1 is a transcription factor that drives terminal differentiation of plasma cells. In order to more accurately assess Blimp-1 expression, we cultured CD19 + BM cells from Blimp-1 YFP mice. When CD19 + BM cells from Blimp1 YFP mice were co-cultured with SCs supplemented with factors, IgA + B cells were generated. Blimp-1 + IgA + B cells were detected when B cells were co-cultured with LP stroma, whereas they were almost absent when BM stroma or Sp stroma was present (Fig. 1C) (Fig. 1G). The amount of IgA antibodies secreted into culture was also different when IgA + B cells were co-cultured with different SCs (Fig. 1H). IgA + B cells arose from gut LP co-cultures produced IgA; however, the level of IgA secretion was not affected by the presence of SCs (statistically significant difference was found between gut LP and BM stroma; however, no statistical significance was detected when it was compared to no stroma control) (Fig. 1H). In other words, once B cells differentiated into IgA + B cells, they are programed to secrete IgA antibodies regardless of the microenvironment, with occasionally enhancement from gut LP stroma (Fig. 1H).
To further explore our culture conditions, we titrated each of the added factors and monitored the production of IgA + B cells ( Fig. 2A). IL-7 was added to promote pro-B and pre-B cell survival and proliferation. 24 IL-21 was added to promote B cell maturation and terminal differentiation. 24 TGF was added to enhance IgA CSR, while anti-CD40 was added to induce B cell activation. 24 Fecal matter (FM) was essential in priming gut LP SCs in promoting IgA + iNOS + cell development according to our previous published work, and this work also confirmed its role in priming gut LP SCs in promoting IgA + B cell accumulation. 24 We found that each of these factors enhanced the production of IgA + B cells but only in the presence of gut LP-derived SCs and other factors ( Fig. 2A). Again, BM-and Sp-derived SCs failed to induce a large accumulation of IgA + B cells, implying gut LP SCs secreted additional stimulatory factors ( Fig. 2A). To study the development and accumulation of IgA + B cells in vitro over time, we set up time-course experiments and analyzed the composition of the cocultures on a daily basis (Fig. 2B). This experiment demonstrated that IgA + B cells typically started to accumulate on day 5, and underwent an expansion during the subsequent days when co-cultured with gut LP stroma (Fig. 2B). However, an expansion of IgA + B cells was not detected when B cells were plated with BM or Sp stroma (Fig. 2B). cell lines using CD45, GP38, and CD31 markers indicated that NS451, PA6, LPSC1, LPSC4, LPSC5, and LPSC6 were CD45 − GP38 + CD31 − , while LPSC2 and LPSC3 were CD45 − GP38 + CD31 + . LPSC7 contained both CD45 − GP38 + CD31 − and CD45 − GP38 + CD31 + subsets (data not shown). Regarding IgA-promoting effects, there was considerable heterogeneity with some gut LP lines having a positive effect while others did not (Fig 2F). This finding indicated that the gut LP stroma indeed contained a heterogeneous population of cells with different IgA-promoting capacities, and some of which were similar to primary LP stroma.

Stromal-derived soluble factors play a role in the regulation of IgA + B cells in vitro
We next wanted to determine whether soluble factors secreted by SCs or contact-dependent mechanisms were responsible for the accumulation of IgA + B cells in vitro. We determined the number of gut LP SCs needed to elicit IgA-promoting effects by titrating the gut LP SCs. Conditioned medium (CM) from the same number of SCs was provided to the starting B cell population and we found similar levels of IgA + B cell production, indicating soluble factors played a major role in the regulation of IgA + B cells (Fig. 3A). CM from 10,000 gut LP SCs was enough to induce a large number of IgA + B cells accumulated in vitro with no significant difference detected with higher number of SCs plated (Fig. 3A). A total of 20,000 SCs were plated for all co-culture experiments to achieve a 1:10 ratio of 1 stromal cell to 10 input B cells.
To examine the roles of soluble factors from different stromal environment, we utilized a dual chamber culture system. While optimizing the dual chamber experiment, we found that the addition of Cytodex

BAFF stimulates the accumulation of IgA + B cells in vitro
We were interested in elucidating the role of BAFF in the co-culture system, as it is an essential positive regulator of IgA + B cell development. To examine the effect of BAFF in culture, we performed a BAFF neutralization assay (Fig. 4A). The addition of anti-BAFF blocking antibodies decreased the number of IgA + B cells recovered from B cellgut LP stroma co-cultures, indicating that BAFF was one of the major contributors of IgA + B cell development (Fig. 4A).
Since BAFF played a role in the accumulation of IgA + B cells in vitro, we wondered if the addition of BAFF to cultures with BM or Sp stroma increased the number of IgA + B cells. Interestingly, the addition of BAFF resulted in an increase in IgA + B cells in cultures without stroma but not in cultures containing either BM or Sp stroma (Fig. 4B). To elucidate the mechanism behind this antagonistic effect, we examined B cell proliferation. We found that the cocktail of factors (containing IL-7, IL-21, TGF , CD40, and FM) with BAFF alone (and no stroma) was sufficient to induce proliferating IgA + B cells, but in the presence of BM or splenic stroma, proliferating IgA + B cells were not increased (Fig. 4C). Flow plots of Fig. 4C demonstrated that the addition of BAFF increased the percentage of both total proliferating cells and proliferating IgA + B cells in control wells without SCs (Fig. 4D). However, in the presence of BM or Sp SCs, such increase was not observed, suggesting that BM and Sp stroma exerted a negative influence on the proliferation of BAFF-treated IgA + B cells (Fig. 4D). Such inhibitory effect could be either directly impacting on IgA + B cells responding to BAFF, or indirectly via BAFF neutralization. It led us to investigate the presence of inhibitory factors using mass spectrometry.

Mass spectrometric analysis of stromal cell secreted factors
We demonstrated that soluble factors play a role both in the stimulation and inhibition of IgA + B cell accumulation in vitro. Specifically, BAFF is a major contributor to the stimulatory effect. However, the addition of anti-BAFF antibodies even at high concentrations could not fully ablate the induction of IgA + B cells, suggesting that other factors besides BAFF might play a role (Fig. 4A). We also determined that BM and Sp stroma secreted soluble inhibitory factors. As a result, a proteomic analysis to identify these factors is necessary.
In order to screen for other possible soluble factors, we performed a LC-MS/MS analysis on the supernatant of SCs (Figs. 5A and B).
Approximately 1593 proteins were quantified using the LC-MS/MS method, and a large proportion (420/1593) of the detected proteins was found across all the 3 SC subsets (Fig. 5A). BM and Sp stroma shared a more similar secretion pattern than gut LP stroma (Fig. 5B).
Among all the differentially secreted proteins, a group of proteins from the complement system (C1s, C1ra, C4b) was highly expressed by gut  (Fig. 5C). MCP-1 itself was not sufficient to induce IgA + B cell accumulation since the addition of recombinant MCP-1 protein to B cell cultures showed no significant inductive effect (Fig. 5C).

DISCUSSION
Our previous published work has detailed an important role for gut LP SCs in the generation of IgA + iNOS + plasma cells. 24  The addition of anti-BAFF neutralization antibodies greatly reduces IgA + B cell accumulation, but does not completely ablate it, indicating that other factors besides BAFF are involved. MCP-1 is also one of the contributors to IgA + B cell accumulation in vitro. MCP-1, a chemokine that recruits monocytes and dendritic cells to sites of inflammation, is secreted by macrophages, dendritic cells, and mesenchymal SCs. 30,31 In this study, we have found that MCP-1 is necessary but not sufficient to induce IgA response in vitro. MCP-1 may act through priming B cells to be more sensitive to other pro-IgA stimulating factors or localizing B cells to a microenvironment favorable to plasma cell differentiation, but such mechanism demands further experiments to verify. such as sBCMA or sTACI as they were found in physiological conditions to neutralize BAFF and APRIL. [27][28][29] It has been shown in both in vitro and in vivo studies that both BM and Sp sites promote plasma cell survival and differentiation. 21,34 In particular, BM niche promotes long-term survival of IgA + plasma cells. 34,35 As a result, there are positive regulators in both BM and Sp microenvironments. However, in this in vitro setting, we propose that the inhibitory effect by BM or Sp stroma on IgA + B cell generation dominates.
Inevitably, this in vitro system has some caveats. First, gut LP, BM, and splenic primary stroma contains a heterogeneous population of SCs, and our in vitro system might not fully recapitulate the contribution and interaction of all SCs in vivo. Our data confirmed previous publications that freshly isolated gut stroma contain four populations: CD45 − GP38 + CD31 + lymphatic endothelial cells, CD45 − GP38 + CD31 − fibroblastic reticular cells, CD45 − GP38 − CD31 + blood endothelial cells, and CD45 − GP38 − CD31 − double negative cells. 36 However, gut LP SCs after culture become mostly fibroblastic reticular cells. In regards to BM and Sp SCs, as suggested before, we might miss the IgA-promoting component of BM and Sp SCs in vitro. Nonetheless, an in vitro system permits a more detailed analysis of SC and IgA + B cell interaction, enabling screening methods for identifying any stimulatory or inhibitory molecules. Second, analysis of secreted factors is based on irradiated SCs. There is evidence that irradiation can change the secretion pattern of cells, but most of the changes are on stromal cell growth, such as VEGF levels and platelet-derived growth factor (PDGF)-AA and PDGF-AB/BB levels. 37 Third, there might be macrophages present in the CD19-selection and stromal preparation. However, the percentage of non-B cells and CD45 + SCs in culture was minimal and thus negligible.
Altogether our findings indicate that the interaction between IgA + B cells and SCs is dynamic in vitro. We propose that IgA + B cell devel-