Gpr97 is essential for the follicular versus marginal zone B-lymphocyte fate decision

Abstract

Gpr97 is an orphan adhesion GPCR and is highly conserved among species. Up to now, its physiological function remains largely unknown. Here, we show that Gpr97 deficiency results in an extensive reduction in B220(+) lymphocytes in mice. More intensive analyses reveal an expanded marginal zone but a decreased follicular B-cell population in Gpr97(-/-)spleen, which displays disorganized architecture characterized by diffuse, irregular B-cell areas and the absence of discrete perifollicular marginal and mantle zones. In vivo functional studies reveal that the mutant mice could generate antibody responses to T cell-dependent and independent antigens, albeit enhanced response to the former and weakened response to the latter. By screening for the molecular events involved in the observed phenotypes, we found that lambda 5 expression is downregulated and its upstream inhibitor Aiolos is increased in the spleen of mutant mice, accompanied by significantly enhanced phosphorylation and nuclear translocation of cAMP response element-binding protein. Interestingly, increased constitutive Nf-κb p50/p65 expression and activity were observed in Gpr97(-/-) spleen, implicating a crucial role of Gpr97 in regulating Nf-κb activity. These findings uncover a novel biological function of Gpr97 in regulating B-cell development, implying Gpr97 as a potential therapeutic target for treatment of immunological disorders.

Figure 1

Targeted disruption of Gpr97 in mice. (a) Schematic representation of Gpr97 gene KO strategy. The targeting vector was designed to delete exon 1 harboring ATG codon and exon 2. Genomic regions amplified by PCR for genotyping are indicated by arrows. Exons are indicated as black boxes. (b) PCR analysis of ES cell clones. Genomic DNA extracted from ES clones was amplified using 5′-external and 3′-external primers as shown in panel (a). Homologous recombination events yielded a 6.5-kb or 3.8-kb fragment, respectively. (c) PCR genotyping of progenies from heterozygous matings. WT (+/+), heterozygous (+/−) and homozygous (−/−) mice were identified by PCR amplification of the fragments specific for either Gpr97 WT allele (944-bp) or the mutant allele (730-bp). (d) RT-PCR analysis for Gpr97 mRNA in BM of WT, Gpr97^+/− and Gpr97^−/− mice. The Gpr97-specific 192-bp product is absent in Gpr97^−/− and reduced in Gpr97^+/− mice. β-actin was used as a loading control. (e) Immunoblotting of BM protein samples with antibodies against Gpr97 and Gapdh shows absence of Gpr97 in mutant mice

Figure 2

Absence of Gpr97 impairs B lymphopoiesis in BM, spleen (SP) and peripheral blood (PB). (a) Splenocytes, BM and PB cells were recovered from 12-week-old WT and age-, sex-matched KO mice (n=7 for each). The cells were labeled with monoclonal antibodies as indicated and subjected to flow cytometric analysis. The numbers in each quadrant indicate the percentages of cell populations. (b) The relative numbers of B220⁺, CD3⁺ and Gr-1⁺ cells were expressed as mean±S.E.M. (n=7). Individual fractions defined in (a) were calculated. *P<0.05; **P<0.01

Figure 3

Analysis of subpopulations of B cells from WT and KO mice. (a) BM cells from WT and Gpr97-KO mice (n=8 for each) were labeled for B220 and CD43 (upper panels), or B220, IgM, IgD (lower panels), and gated on total lymphocytes or B220⁺ cells, respectively. Numbers represent the percentages of cells in the plot that fall within the different regions, corresponding to different stages of B-cell differentiation. (b) The relative and absolute numbers of CD43⁺B220^int (pro-B), CD43⁻B220^int (pre-B/immature B), CD43⁻B220^hi (mature B), B220⁺IgM⁺IgD⁻ (immature B), B220⁺IgM⁺IgD⁺ (mature B) in BM were expressed as mean±S.E.M. (n=8). Individual fractions defined in (a) were calculated. (c) Flow cytometry profiles of the spleen of WT and KO mice labeled for B220, IgM, CD21 (upper panels), or B220, CD21, CD23 (lower panels), and gated on B220⁺ cells. Numbers represent the percentages of cells in the plot that fall within the different regions, corresponding to different stages of B cell differentiation. (d) The bar charts summarize the relative and absolute numbers of B220⁺CD21^low/−IgM^high (T1), B220⁺CD21^highIgM^high (T2), B220⁺CD21^intIgM^int (mature B, M), B220⁺CD21^highCD23^low/− (MZ B), B220⁺CD21^intCD23^high (FO B) in the spleen. Individual fractions defined in (c) were used for calculations. Data show the mean±S.E.M. The differences between WT and KO mice were examined for statistical significance that was indicated as asterisks: *P<0.05; **P<0.01; ***P<0.001

Figure 4

Histopathological and immunohistochemical analysis of the spleen of WT and KO mice (n=4, each group). (a) and (b) Spleen sections from 12-week-old WT and KO mice were stained with H&E. Perifollicular MZs were indicated by arrows. Scale bar, 100 μm. (c) and (d) Immunohistochemistry of spleen sections from (a) and (b) were stained with B cell-specific mAb B220. Scale bar, 200 μm. (e) and (f) Histological section of Gpr97-KO spleen showed an increased population of IgM⁺ cells (labeled in red with anti-IgM Ab) in MZ. The green signals represent FITC-labeled anti-MOMA-1, which is specific for metallophilic macrophages that separate the MZ from the follicles. Scale bar, 200 μm. (g) and (h) Mice were killed 1 week after secondary immunization with IFA-OVA. Sections of spleen were labeled with PNA (brown). Hematoxylin counterstain was used. Scale bar, 200 μm

Figure 5

Serum Ig levels, TI and TD antigen responses in WT and KO mice. (a) IgM, IgG1, IgG2a, IgG2b, IgG3 and IgA levels in the sera of 3-month-old WT (n=8) and KO (n=8) mice were evaluated by ELISA. Serum levels of IgM (b) and IgG3 (c) anti-DNP antibodies were measured before (t₀) and 14 days after (t₁₄) immunization with the TI antigen DNP-Ficoll in WT and KO mice. Serum levels of IgM (d) and IgG1 (e) anti-DNP antibodies were measured before (t₀), at the time of secondary immunization (t₁₄) and 7 days after secondary immunization (t₂₁) with the TD antigen DNP-KLH in WT and KO mice. Data show the mean±S.E.M. The differences between WT and KO mice were examined for statistical significance that was indicated as asterisks: *P<0.05; **P<0.01; ***P<0.001

Figure 6

Gpr97 deficiency leads to downregulation of lambda 5. (a) The expression levels of the genes related to B-cell development were evaluated by real-time qPCR in BM cells. There was a reduction in lambda 5 mRNA level in KO mice as compared with WT mice (n=8 for each group). (b) Real-time qPCR analysis of the expression levels of genes in splenocytes. Lambda 5 mRNA level was reduced in KO mice as compared with WT mice (n=8 for each group). (c) Staining of BM cells from WT and KO mice (n=7 for each group) with anti-B220 and anti-lambda 5 antibodies. (d) Relative numbers of lambda 5⁺ cell were presented as mean±S.E.M. (n=7). (e) Lambda 5 reporter construct and Gpr97 expression vector were co-transfected into Hela cells. Relative luciferase activity was monitored in triplicate 48 h post transfection. The differences between two groups were examined for statistical significance that was indicated as asterisks in panels (a), (d) and (e): **P<0.01; ***P<0.001

Figure 7

pCREB and Aiolos levels were increased in the spleen of Gpr97-KO mice. (a) Real-time qPCR analysis of Aiolos expression levels in splenocytes (n=5 for each group). Data show the mean±S.D. (b) Western blot images representing the expressions of pCREB and total CREB in total cell lysates and nuclear fractions from spleen. (c) No difference was observed in serum BAFF level between WT and KO mice (n=5, each group). Data show the mean±S.E.M. (d) Immunoblots of NF-κB1/p105/p50 and p65 in whole-cell lysates, cytoplasmic and nuclear fractions from splenocytes. (e) Notch2 protein expression was analyzed in spleen by western blot. The differences between WT and KO mice were examined for statistical significance that was indicated as asterisks: ***P<0.001