The formation of an effective humoral immune response is dependent on intricate interactions between B cells and other immune cell types. These interactions are enabled by the precise localization of these various immune cells in their respective niches (1). The spleen is a highly compartmentalized organ, and its intact histological structure is a prerequisite for the establishment of an effective humoral immune response (2, 3). The positioning of follicular B cells and other immune cells within the spleen is a highly dynamic process orchestrated by a complex network of chemokines and cytokines, but also by other secreted molecules like sphingosine-1-phosphate (S1P) (1, 4–9). S1P is a bioactive lipid that exhibits a steep gradient between high vascular concentrations and low presence in the surrounding interstitium (10). S1P binds to a family of five membrane bound S1P receptors S1PR_1–5. While S1PR_1–3 are nearly ubiquitously expressed, S1PR₄ expression is restricted to cells of hematopoietic origin (11). By blocking S1P signaling with a functional antagonist of S1P receptors S1PR₁ and S1PR_3–5, Han et al. revealed a role of S1P signaling in the germinal center reaction (12). However, the identity of the S1P receptor mediating this effect remains unknown.

In a previous study, we have reported a reduced migration of LPS stimulated peritoneal B cells from the abdominal cavity into splenic follicles (13). S1PR₄ expression is a common feature of various splenic B cell populations. In the present study, we sought to characterize the impact of S1PR₄-mediated signaling on the splenic architecture and the development of the B cell response during systemic infections.

Follicular B cells and MZ B cells express high levels of S1PR₁ and lower levels of S1PR₃ and S1PR₄, while S1PR₂ expression is very low and S1PR₅ expression absent (4, 29). Compared to follicular B cells, S1PR₂ expression is up-regulated in germinal center B cells (7). While distinct functions have been attributed to the S1P receptors S1PR₁ – S1PR₃ in splenic follicular B cells and MZ B cell biology, the role of S1PR₄-mediated signaling in the spleen remains obscure (7, 27, 30).

In the present manuscript, we show that the size of splenic primary follicles was significantly reduced in s 1 p r 4 − / − mice compared to wt animals, while the cumulative size of the B220⁺ B cell area was similar in both strains. Interestingly, mice deficient for the chemokine receptor CXCR5 also show structural aberrances of primary follicle structure (31, 32). Adoptively transferred CXCR5-deficient B cells fail to enter the splenic primary follicle (31). We have previously shown that, although S1PR₄-mediated signaling induces chemotactic migration in peritoneal B1a and B1B cells, splenic B cells show no chemotactic response to S1P gradients (33, 34). We therefore examined whether S1PR₄-mediated signaling modifies CXCR5-mediated chemotaxis. Indeed, S1PR₄ deficiency resulted in a significantly reduced chemotactic response of splenic B cells to CXCL13 gradients, while S1P alone did not induce a chemotactic response in these cells. These results clearly indicate that S1PR₄-mediated signaling modifies the CXCR5-mediated chemotactic response of follicular B cells, possibly resulting in the observed structural aberration in the spleen of s 1 p r 4 − / − mice. Interestingly, interference with CXCL12- and CXCL13-induced chemotaxis has also been shown for another S1P receptor, S1PR₂, in germinal center B cells in vitro (7). S1PR₂-mediated signaling negatively modified CXCL13-induced chemotaxis, while S1PR₄ increased CXCL13-induced chemotaxis in splenic B cells in our experimental setting (7). These data suggest that modulation of CXC-chemokine induced chemotaxis is a common mechanism in the S1P-mediated control of B cell migration.

Transitional type 1 B cells are the link between the bone marrow and the spleen, where they represent the initial B cell immigrants (35, 36). Interestingly, inhibition of S1P-mediated signaling has been shown to increase transitional B cell numbers in the peripheral blood of multiple sclerosis patients (37). In our S1PR₄ deficient mouse model the number of transitional B cells in the spleen is significantly reduced. The mechanisms regulating the immigration of transitional B cells from blood to the spleen are not well understood. Our results suggest that S1PR₄ may be involved in this process, and its deficiency may result in increased peripheral transitional B cell numbers by inhibiting entry into the spleen. Further research is required in order to verify this hypothesis.

Numbers of pro-/pre-B cells are reduced in the bone marrow of s 1 p r 4 − / − animals. CXCL12 signaling is required for the development of pre-pro-B and pro-B cells, probably by regulating the positioning of these cells in their appropriate niches within the bone marrow (38). B cell precursors are significantly reduced in bone marrow of chimeric wildtype mice that have been reconstituted with CXCR4 deficient fetal cells (39). Moreover, pre-B cells were increased in the blood of CXCR4 deficient animals, indicating that the retention of B cell precursors within the bone marrow is dependent on CXCR4 (40). In splenic T cells, S1P has been shown to act in synergy with CXCL12 (41). Further experiments should investigate whether similar mechanisms are involved in the reduction of pre-/pro-B cell numbers in s 1 p r 4 − / − mice.

Marginal zone (MZ) B cells physiologically shuttle between the MZ and the follicle, thus ensuring the effective transfer of antigens between these two localizations (4, 8). Only about 55% of MZ B cells are localized in the MZ, while 45% can be found inside the follicle (27). The respective proportions are determined by the expression levels of S1PR₁ and CXCR5 (27). But also S1PR₃-mediated signaling has been shown to impact MZ B cell shuttling (27). The data shown in this manuscript demonstrates that S1PR₄-mediated signaling does not affect the proportion of MZ B cells in the MZ and the follicular area. However, it remains to be investigated whether the exchange rate, which has been shown to reach 20% per hour (27), is affected by S1P signaling via S1PR₄. In vitro imaging in the intact spleen in our S1PR₄ deficient mouse model will answer this question in the future. At the same time, total MZ B cell numbers were significantly increased in naïve s 1 p r 4 − / − animals the FACS analysis of the whole spleen. Whether these changes are the result of altered migrational processes, an impact on the proliferative capacity of MZ B cells or MZ B cell survival remains to be shown in further experiments. S1PR₄ has been shown to affect cell proliferation in S1PR₄ overexpressing T cell lines (42). Experiments performed in our own lab showed no influence of S1PR₄ mediated signaling on the proliferation of peritoneal B cells (data not shown). Further research is required to clarify the role of S1PR₄ mediated signaling in MZ B cell proliferation and survival.

Only moderate micro-anatomical differences exist in s 1 p r 4 − / − animals under steady-state-conditions and we wondered whether strong B cell activation would aggravate these histopathological changes. Abdominal sepsis is a clinically relevant disease state resulting in such a strong B cell activation (23). While the primary T cell dependent B cell response against specific antigens is severely impaired in septic mice, a strong unspecific increase of serum IgM and IgG levels as well as a pronounced germinal center reaction within the spleen can be observed in murine models of sepsis (24, 43). The main source of the IgM and IgG secreting cells in this model was the spleen (23). In the murine CASP model used herein, S1PR₄ deficiency resulted in a significantly reduced GC formation in the septic spleen. Participation of S1P-mediated signaling in the development of the germinal center response has been previously reported by Han et al., who observed a reduced germinal center formation in mice after fingolimod treatment (12). However, the identity of the receptors mediating this effect has not yet been dissected. Disorganized development of GCs in the spleen has also been reported in the absence of CXCR4 expression in B cells (31, 44, 45). Both CXCL12 and CXCL13 have been implicated in the regulation of GC organization in dark and light zone (45). The histopathological features of s 1 p r 4 − / − spleens are reminiscent of those seen in mice with a B cell specific conditional CDC42 knock-out, showing GCs of reduced size (46). CDC42 is part of the S1PR₄ associated intracellular signaling cascade in lymphocytes (47). According to the recycling hypothesis, the development of an effective GC reaction depends on multiple rounds of B cell circulation between light and dark zone of the GC (25). These iterative rounds of mutation and selection in their respective zones are necessary for a stepwise optimization of B cell affinities towards the locally presented antigen, giving rise to affinity-matured memory and plasma cells (26). Our findings suggest that S1PR₄ impacts the chemotactic control of B cell circulation resulting in structural aberrances of GC development in s 1 p r 4 − / − animals. However, further experiments are required to examine the hypothesis that S1PR₄ mediated signaling modulates not only CXCL13 induced chemotaxis but also that of other chemokines, e.g. CXCL12.

Both FDC and Tfh cells are essential for the development of an effective GC reaction (26, 48). Avancena et al. reported that a predetermined number of FDC clusters determine the maximum number of GCs that may develop after antigenic stimulation. Our results clearly show that the number of FDC clusters is similar in wt and s 1 p r 4 − / − animals, thus excluding an influence of S1PR₄ deficiency on the number of FDC clusters. However, since the size of FDC clusters was significantly reduced in s 1 p r 4 − / − animals, further experiments involving conditional knock-out models of S1PR₄ in either B cells or FDC are required to determine the respective contribution of S1PR₄ expression in these two cell types for the development of the GC reaction. Tfh cells are responsible for both initiation and maintenance of the GC reaction. Their numbers showed no differences in s 1 p r 4 − / − and wt animals, therefore excluding an influence of S1PR₄ deficiency on Tfh accumulation as the source for the reduced GC size in the spleens of septic s 1 p r 4 − / − animals.

Although the s 1 p r 4 − / − genotype is characterized by a compromised development of splenic GCs, antibody production after polymicrobial intra-abdominal challenge resulted in only minor differences between s 1 p r 4 − / − and wt animals. In CASP-induced sepsis, the spleen is the main source of antibody production (23). However, CXCR5 deficient mice also display a compelling disturbance of GC formation, while antibody formation upon immunization with T cell dependent antigens is normal (31). In our model, it remains to be determined whether normal antibody levels in s 1 p r 4 − / − animals are due to extrasplenic antibody formation or whether the formation of these predominantly low-affinity antibodies is mainly due to the activity of short-lived plasma cells and not due to affinity maturated plasma cells originating in the splenic follicles (49, 50). In contrast, after immunization with SRBC, production of SRBC-specific IgM antibodies was slightly but significantly reduced in s 1 p r 4 − / − animals after the second immunization cycle. This reflects the disturbance of T cell dependent antibody production in s 1 p r 4 − / − animals, which requires specific interaction between B and T cells. The implication of S1P signaling in the development of an effective GC response and the production of high affinity antibodies has been previously elegantly revealed using the functional S1P antagonist fingolimod (12). Fingolimod binds to four out of five S1PRs (S1PR₁, S1PR₃ – S1PR₅). However, since fingolimod is an unspecific inhibitor of S1P signaling, it remains to be shown whether the morphological changes in GC development secondary to S1PR₄ deficiency affect also the affinity maturation during the antibody response

The S1P-S1PR axis is a complex signaling system. Cells usually express several S1PRs on their surface. The difference in signaling through various S1PR resides in variations of G-protein coupling of the different S1PRs (51). B cells express S1PR₁ and S1PR_3–4, GC B cells additionally S1PR₂ (4, 7, 29). All of these receptors share down-stream signaling via G_i proteins (51). Thus, we cannot formally exclude that some effects of S1PR₄ deficiency may be masked by compensatory signaling via other S1P receptors. Further experiments using receptor specific antagonists in vitro are required to exclude this possibility. In order to better differentiate B cell intrinsic from B cell extrinsic effects of S1PR₄ mediated signaling and their mechanistic implication in the generation of the morphological and functional alterations reported in this manuscript, further experiments using conditional B cell specific S1PR₄ knock-out models should be performed.

In summary, S1PR₄-deficient mice exhibit a regular lymphatic architecture with only slight alterations in follicle size. During systemic infectious challenge reduced germinal center formation was observed. However, this did not translate into a relevant impairment of RBC specific antibody formation in our model. These results indicate a role for S1PR₄ expression during the development of the GC reaction and warrant further research into consequences of impaired S1PR₄ signaling.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Landesamt für Landwirtschaft, Lebensmittelsicherheit u. Fischerei Mecklenburg Vorpommern (LALLF M-V) and by the Landesamt für Gesundheit und Soziales Berlin (LAGeSo). Written informed consent was obtained from the owners for the participation of their animals in this study.

Conceptualization: JR and TS. Methodology: TS, AK and JR. All authors partook in the experiments. Software: JR, AK. Validation: JR and TS. Formal analysis: JR and TS. Writing original draft: JR. Writing review and editing: TS, AK, CH, MH and FL. Visualization: JR. Supervision: TS. All authors contributed to the article and approved the submitted version.

This work was financed by an University Research Fund (Forschung und Lehre) of the Faculty of Medicine of the University of Greifswald. J.R., M.H. and F.L. were financially supported by a scholarship from the Gerhard Domagk programe of the University Medicine Greifswald.

We would like to thank Antje Janetzko for dedicated technical and logistical support.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.