Research ArticleCELL BIOLOGY

The orphan nuclear receptor LRH-1/NR5a2 critically regulates T cell functions

See allHide authors and affiliations

Science Advances  17 Jul 2019:
Vol. 5, no. 7, eaav9732
DOI: 10.1126/sciadv.aav9732


LRH-1 (liver receptor homolog-1/NR5a2) is an orphan nuclear receptor, which regulates glucose and lipid metabolism, as well as intestinal inflammation via the transcriptional control of intestinal glucocorticoid synthesis. Predominantly expressed in epithelial cells, its expression and role in immune cells are presently enigmatic. LRH-1 was found to be induced in immature and mature T lymphocytes upon stimulation. T cell–specific deletion of LRH-1 causes a drastic loss of mature peripheral T cells. LRH-1–depleted CD4+ T cells exert strongly reduced activation-induced proliferation in vitro and in vivo and fail to mount immune responses against model antigens and to induce experimental intestinal inflammation. Similarly, LRH-1–deficient cytotoxic CD8+ T cells fail to control viral infections. This study describes a novel and critical role of LRH-1 in T cell maturation, functions, and immopathologies and proposes LRH-1 as an emerging pharmacological target in the treatment of T cell–mediated inflammatory diseases.


An interaction between dietary and microbial-derived lipids and the immune system has been suggested for a long time. Food- and microbiome-derived lipids regulate host metabolism and thereby obesity, which, in turn, promotes inflammation and associated diseases, such as type 2 diabetes and cancer [reviewed in (1) and (2)]. Best characterized is the role of short-chain fatty acids, such as butyrate and acetate, derived from bacterial digestion of food (3). However, also longer chain fatty acids and phospholipids are known to be taken up by the gastrointestinal tract and to regulate immunological processes [reviewed in (4)]. The underlying molecular mechanisms remain, so far, largely unexplored.

Lipid molecules are often bound and sensed by a large group of transcription factors, the nuclear receptor family. Nuclear receptors play important roles in various developmental, metabolic, and inflammatory processes. Lipid ligand binding promotes their transcriptional activation via coactivator recruitment [reviewed in (5)]. Best characterized is the group of steroid hormone receptors, which bind cholesterol derivatives, such as sex hormones, aldosterone, and glucocorticoids. In particular, the glucocorticoid receptor is known to exert potent immunoregulatory activities, providing the basis for the efficient therapeutic use of synthetic glucocorticoids in the treatment of inflammatory processes (6). Endogenous glucocorticoids are produced not only in the adrenal glands but also by extra-adrenal sources, such as the thymus, skin, lung, and intestine (7). Thus, we have previously shown that the intestinal epithelium is an important source of glucocorticoids and that intestinal glucocorticoids critically contribute to the regulation of local immune homeostasis (811).

The synthesis of immunoregulatory glucocorticoids in the intestinal epithelium is critically controlled by the nuclear receptor liver receptor homolog-1 (LRH-1/NR5a2) (9, 10), which belongs to the fushi tarazu family of transcription factors [reviewed in (12)]. Deletion of LRH-1 in the intestinal epithelium predisposes mice to development of intestinal inflammation due to the absence of local glucocorticoid synthesis (9). Furthermore, studies in patients with ulcerative colitis and Crohn’s disease revealed a clear negative correlation between LRH-1, intestinal glucocorticoids, and intestinal inflammation (9). LRH-1 is considered an orphan nuclear receptor, as no endogenous ligand has been identified so far. Various reports have, however, highlighted the fact that LRH-1 binds phospholipids, lipids, and lipophilic chemical compounds, resulting in transcriptional activation of the nuclear receptor (13, 14).

Although intestinal LRH-1 regulates immune cells via immunoregulatory glucocorticoids (10), little is currently known about the expression of LRH-1 in the hematopoietic system. Recently, a role of LRH-1 in the differentiation of macrophage subpopulations and the development of anti-infectious effector functions has been demonstrated (15). Similarly, we were able to demonstrate that LRH-1 contributes to the transcriptional control of Fas (CD95) ligand expression and associated effector functions in CD4+ T lymphocytes (16). However, the expression and role of LRH-1 in cells of the T cell lineage are currently unexplored and enigmatic. Here, we demonstrate a novel and direct role of LRH-1 in immune regulation. LRH-1 is expressed in murine thymocytes and mature T lymphocytes, and stimulation of cells further induces LRH-1 expression. CD4 promoter–driven deletion of LRH-1 at the CD4/CD8 double-positive stage in the thymus has only a minor impact on thymocyte maturation but severely reduces the number of peripheral T cells. While LRH-1–deficient T cells show normal signs of early T cell activation, they exert strongly reduced activation-induced proliferation in vitro and in vivo. LRH-1 deficiency severely impairs T cell responses and antibody production in vivo and abrogates CD4 T cell–mediated experimental colitis. Upon viral infection, LRH-1–deficient CD8+ T cells readily become activated and produce excessive amounts of cytokines yet fail to control virus expansion. This study demonstrates a critical role of LRH-1 in T cell maturation and the regulation of T cell homeostasis, T cell expansion, and effector functions. Furthermore, our study proposes LRH-1 as an emerging new target in the treatment of T cell–mediated inflammatory diseases.


LRH-1 deletion results in reduced mature T cells

Because of the lack of information concerning the expression of LRH-1 in hematopoietic cells, we first investigated its gene expression profile in different lymphatic tissues (fig. S1A) and immature (fig. S1B) and mature T cell subsets (fig. S1C). This revealed that Nr5a2 mRNA expression was low but detectable in all immature and mature T cell subsets of C57BL/6 wild-type mice (fig. S1, B and C) and in human peripheral blood mononuclear cells (PBMCs) (fig. S1D).

To address the role of LRH-1 in T lymphocytes, we next generated T cell–specific LRH-1–deficient mice using CD4 promoter–driven Cre recombinase expression [Nr5a2L2/L2 CD4-Cre, conditional knockout (cKO)]. As CD4 expression is first observed at the CD4+CD8+ stage of thymocyte maturation, we compared the thymi of cKO and control (L2/L2) animals but found no difference in size (fig. S2A), weight (fig. S2B), or cellularity (fig. S2C). Only a mild reduction of CD4+ and CD4+CD8+ and a corresponding increase in CD4CD8 thymocytes were observed (fig. S2D).

When analyzing secondary lymphatic organs, we observed that the spleen of cKO animals was significantly reduced in size (Fig. 1A), organ weight (Fig. 1B), and cell numbers (Fig. 1C). Unexpectedly, CD8+ T cells were reduced by approximately 50%, and CD4+ lymphocytes were reduced by almost 80% in cKO spleens (Fig. 1D). In line with the significant loss of splenic CD3+ T cells (Fig. 1E), the relative number of B220+ B cells was increased (Fig. 1F) and no difference was observed for NK1.1+ (Fig. 1G) in cKO animals. The relative loss of CD4+ and CD8+ T cells was confirmed when analyzing total cell numbers (Fig. 1, H and I). Comparable changes in CD4+ and CD8+ T cell numbers were found in axial (fig. S3, A to F) and mesenteric lymph nodes (fig. S3, G to L).

Fig. 1 Deletion of Nr5a2 leads to loss of mature T lymphocytes.

(A) Representative picture of Nr5a2L2/L2 (L2/L2, ctrl) and Nr5a2L2/L2 CD4-Cre (cKO) spleens. ctrl, control. (B) Spleen organ weight relative to body weight (L2/L2, n = 17; cKO, n = 18 mice). (C) Spleen cellularity (n = 15). (D) Representative density plots of splenic CD4+ and CD8+ cells. The numbers indicate the percentage of cells. PE, phycoerythrin; FITC, fluorescein isothiocyanate. (E to G) Relative distribution of splenic CD3+ (E), B220+ (F), or NK1.1+ (G) CD3+ cells analyzed by flow cytometry (L2/L2, n = 11 mice; cKO, n = 12 mice per group). (H and I) Absolute numbers of splenic CD3+CD4+ and CD3+CD8+ T lymphocytes (L2/L2, n = 23; cKO, n = 17). (J) Representative hematoxylin and eosin (H&E) staining of L2/L2 or cKO spleens. (K) Representative immunohistology for CD3 (yellow) and B220 (blue) expression. (L) Representative staining for CD3ε (yellow), F4/80 (red), and 4′,6-diamidino-2-phenylindole (DAPI) (blue). Scale bars, 0.5 cm (A) and 300 μm (J to L). Mean values ± SD and individual values are shown in each graph. *P < 0.05 and ***P < 0.001. ns, not significant. Photo credit: Carina Seitz, University of Konstanz.

Because of the strong phenotype of LRH-1 deletion in mature T cell distribution, we next analyzed the splenic architecture. Histological analysis revealed obvious changes in the structure of white pulp follicles, which were more compact and with a lighter core, suggesting a reduced cell density (Fig. 1J). Immunohistological detection of B and T lymphocytes confirmed a reduced size of the T cell zone (Fig. 1K), while the overall distribution of B cell follicles and marginal zone macrophages (Fig. 1L) was not altered.

As CD4 promoter–driven deletion of LRH-1 had a stronger impact on numbers of peripheral CD4+ than CD8+ T cells, the question whether Cre-mediated deletion was less effective in CD8+ T cells arose. We thus used a Tomato-membrane green fluorescent protein (GFP) (mTmG) double-fluorescent Cre reporter mouse line, in which successful Cre-mediated recombination results in green fluorescence of cells (17). As expected, immature CD4CD8 thymocytes showed very low GFP levels and, hence, minimal LRH-1 deletion (fig. S4A), while the deletion was nearly complete in more mature CD4+CD8+ thymocytes (fig. S4B), as well as in thymic, splenic, axial, and mesenteric CD4+ and CD8+ subsets (fig. S4, C, D, and F to H). Microscopic analysis of spleen sections confirmed specific LRH-1 deletion (GFP+) in the T cell area (fig. S4E). Overall, the rate of GFP+ CD8+ T cells was slightly lower, however, with 80 to 90% still almost complete.

Sensitivity of LRH-1–deficient T cells to apoptosis induction

To understand the underlying cause of LRH-1 deletion-mediated loss in peripheral T cells, we investigated the apoptosis sensitivity of cKO T cells. Enhanced basal cell death was observed in cKO thymocytes (fig. S5, A to D) and splenocytes (fig. S6, A to D), while their response toward dexamethasone, etoposide, staurosporine, or plate-bound anti-CD3 promoting activation-induced cell death was comparable to the L2/L2 controls (figs. S5, A to D, and S6, A to D). Hence, increased stimulus-induced cell death does not appear to account for the selective loss of peripheral T cells in cKO mice.

As proliferating T cells are more sensitive to apoptosis-inducing agents, especially to activation-induced cell death (18), we addressed apoptosis induction in concanavalin A (ConA)–induced T cell blasts. Unexpectedly, cKO CD4+ cells and, to a lesser extent, CD8+ T cells failed to expand under these conditions and rather died, leading to approximately 80% basal cell death. This time-dependent increase in cell death of CD4+ cKO T cells was observed after ConA stimulation (fig. S6E) and anti-CD3/anti-CD28 treatment (fig. S6F).

LRH-1–deficient T cells are not anergic

Because of the increased cell death in response to mitogenic stimuli, we tested the hypothesis that LRH-1–deficient T cells fail to become activated after T cell receptor stimulation. While the activation induced up-regulation of the early activation markers CD69 and CD25 was slightly but significantly lower and slower in CD4+ and CD8+ cKO T cells, both responded to T cell activation by up-regulating both activation markers (Fig. 2, A and B).

Fig. 2 LRH-1–deficient T cells are not anergic.

(A and B) Up-regulation of the early activation marker CD69 (A) and CD25 (B) in unstimulated (unstim) or anti-CD3 (3 μg/ml) and anti-CD28 (1 μg/ml) (stim) CD4+ or CD8+ splenic T cells (n = 3 mice). (C and D) Production of interferon-γ (IFN-γ) (C) or interleukin-2 (IL-2) (D) in purified CD4+ or CD8+ splenic T cells. Cells were stimulated as indicated for 48 hours, and cytokine secretion was determined by enzyme-linked immunosorbent assay (ELISA) (n = 4 mice). Mean values ± SD of triplicates or quadriplicates of representative experiments are shown. *P < 0.05, **P < 0.01, and ***P < 0.001.

To further exclude anergy as a cause of reduced numbers of LRH-1–deficient T cells, we analyzed their capability of effector cytokine expression and secretion. Interferon-γ (IFN-γ) secretion was detected in purified CD4+ and CD8+ T cells of both cKO and L2/L2 mice with even much higher levels in cKO T cells (Fig. 2C). Similar results were obtained for interleukin-2 (IL-2) expression (Fig. 2D), confirming that LRH-1 deletion does not impair T cell activation.

Impaired activation-induced proliferation in LRH-1–deficient T cells

As cKO T cells were neither hypersensitive to stimulus-induced apoptosis induction nor anergic, we hypothesized that LRH-1 deficiency directly affects T cell proliferation and expansion. Hence, we investigated the effect of mitogenic stimuli on Nr5a2 mRNA expression and found a significant increase in mouse splenocytes (Fig. 3A) and human PBMCs (Fig. 3B). Activation-induced Nr5a2 expression closely correlated with cell cycle progression, as monitored by the mRNA expression of cyclins (Fig. 3, C to E). As expected, expression of Myc, a direct regulator of the cell cycle machinery, was induced very early after activation (Fig. 3C). Nr5a2 expression was induced soon after stimulation with a peak at 3 to 6 hours (Fig. 3D), followed by cyclin E1 (Cdne1) (Fig. 3E), a transcriptional target of LRH-1 (19). We further observed increased LRH-1 promoter activation in response to mitogenic stimulation in Jurkat lT cells (Fig. 3F), whereas serum withdrawal decreased LRH-1 promoter activity (Fig. 3G).

Fig. 3 Nr5a2 expression is up-regulated upon stimulation.

(A) Nr5a2 mRNA expression in wild-type spleenocytes after 3 hours of phorbol myristate acetate (PMA) and ionomycin (iono) or ConA stimulation, normalized to unstimulated control (n = 3 mice per group). (B) NR5A2 mRNA expression in human PBMCs (hPBMCs) after stimulation with phytohemagglutinin (PHA-P) for 3 hours (n = 3 biological replicates). (C to E) Kinetics of (C) Myc, (D) Nr5a2, and (E) Cdne1 expression in splenocytes after anti-CD3/anti-CD28 stimulation normalized to unstimulated control (n = 3 mice). (F) Induction of NR5A2 promoter activity in Jurkat lT cells after stimulation with PMA and ionomycin. One representative experiment of five is shown (n = 3 technical replicates). (G) Effect of serum deprivation on NR5A2 promoter activity. One of three representative experiments is shown (n = 3 technical replicates). *P < 0.05, **P < 0.01, and ***P < 0.001.

We next tracked spleen cell division by carboxyfluorescein diacetate succinimidyl ester (CFSE) dilution assay and observed a prominent reduction in LRH-1–deficient CD4+ T cells, while proliferation was not reduced or delayed in the CD8+ subset (Fig. 4, A and B). To confirm a role of LRH-1 in T cell proliferation, specific pharmacological inhibitors were applied to mimic the loss of LRH-1 protein. While the LRH-1 inhibitor 3d2 (20) resulted in significant inhibition of cell division even for the lowest concentrations, the inactive control substance cd7 (20) did not (Fig. 4, C and D). 3d2 was also able to inhibit CD8+ T cell proliferation, although higher concentrations were necessary (Fig. 4D). An inhibition of T cell proliferation was also observed with another LRH-1 inhibitor, SR1848 (Fig. 4E) (21). Reduced activation-induced proliferation of LRH-1–deficient T cells was confirmed by 3H-thymidine incorporation of highly purified CD4+ and CD8+ T cells (Fig. 4F), suggesting that also in the absence of splenic accessory cells, CD8+ T cells depend on LRH-1. To understand the molecular basis of this reduced proliferation, we analyzed the expression of cell cycle–regulating target genes of LRH-1, i.e., c-Myc (Myc) and cyclin E1 (Cdne1) (19, 22). In agreement with the reduced proliferation observed in activated cKO T cells, a reduced induction Myc and Cdne1 expression was seen in CD4+ cKO T cells and reduced Myc expression in CD8+ T cells (Fig. 4, G and H).

Fig. 4 Impaired cell proliferation in LRH-1–deficient T cells.

(A and B) Cell proliferation in response to anti-CD3 ± anti-CD28 stimulation of splenic CD4+ and CD8+ T cells was analyzed by CFSE dilution after 72 hours. Mean values ± SD of three individual experiments are shown. (C and D) Spleen cells were stimulated with anti-CD3 and anti-CD28 in the presence or absence of the LRH-1 antagonist 3d2 or control substance cd7 at indicated concentrations, and proliferation was analyzed by CFSE dilution at 72 hours. Values were normalized to stimulated cells without inhibitors (n = 3 mice). (E) Splenocytes were treated with or without the LRH-1 inhibitor SR1848 and stimulated with anti-CD3/anti-CD28; cell proliferation was analyzed by CFSE dilution at 72 hours (n = 3 mice). (F) Activation-induced 3H-thymidine incorporation in purified CD4+ and CD8+ splenic T cells after 72 hours, normalized to unstimulated cells. Mean values ± SD of a representative experiment are shown (n = 5 experiments). (G and H) Activation-induced Myc (G) and Cdne1 (H) mRNA expression in purified CD4+ and CD8+ splenic T cells after 24 hours. *P < 0.05, **P < 0.01, and ***P < 0.001.

We next assessed the role of LRH-1 in homeostatic T cell expansion in vivo by adoptively transferring purified cKO or L2/L2 T cells into lymphopenic recipient mice. Analysis of spleen and mesenteric lymph nodes revealed major defects of CD4+ cKO T cell expansion and reduced repopulation of LRH-1–deficient CD8+ T cells (fig. S7, A and B). To further monitor activation-induced T cell expansion in vivo, we analyzed the ex vivo T cell restimulation of ovalbumin-immunized mice. While an antigen-dependent proliferation of T cells was detectable in immunized L2/L2 mice, no ex vivo proliferation was seen in T cells from cKO mice (fig. S8, A and B). In addition to the lack of ovalbumin-induced T cell expansion in vivo, a massive reduction of anti-ovalbumin immunoglobulins in the sera of cKO mice was observed, revealing defects in T cell–dependent B cell activation (fig. S8C).

Impaired colitis induction by LRH-1–deficient T cells

To test the consequence of impaired proliferation in a T cell–dependent pathogenesis model in vivo, we used the T cell transfer colitis model. In this model, naïve CD4+ CD45Rbhi T cells are transferred into lymphopenic Rag1−/− mice, where their uncontrolled activation by gut microbiota leads to subsequent colitis induction (Fig. 5A) (23, 24). While transfer of L2/L2 T cells resulted in a time-dependent loss of body weight (Fig. 5B) and massive colonic inflammation, cKO T cell transfer failed to promote body weight loss and caused significantly less colonic inflammation (Fig. 5, C to E). The reduced colitis induction of cKO donor cells observed was paralleled by diminished T cell expansion and repopulation of spleen and mesenteric lymph node (Fig. 5, F and G) and reduced IFN-γ expression in the colon of Rag1-deficient recipient mice (Fig. 5. H to J).

Fig. 5 Impaired colitis induction by LRH-1–deficient CD4+T cells.

(A) Scheme of transfer colitis induction in Rag1−/− recipient mice. (B) Colitis-induced loss of body weight (L2/L2 donor, n = 14; cKO donor, n = 9). (C to E) Representative histologies of colon samples from untreated Rag1−/− recipient mice (C) or after transfer of naïve L2/L2 (D) or cKO T cells (E). Top panels: Lower magnification. Bottom panels: Higher magnification. Scale bars, 150 μm. (F and G) Repopulation of spleen (F) or mesenteric lymph node (mLN) (G) by naïve L2/L2 or cKO T cells. (H to J) mRNA expression levels of the proinflammatory cytokines tumor necrosis factor–α (H), IL-1β (I), and IFN-γ (J) in colon samples or control recipients or recipients transferred with L2/L2 or cKO T cells. Mean ± SD and individual data points of n = 9 to 14 mice are shown. *P < 0.05, **P < 0.01, and ***P < 0.001.

Reduced regulatory function of LRH-1–deficient T cells

To further study the putative role of LRH-1 in regulatory T cells (Tregs), we conducted a colitis protection experiment. While naïve CD4+ CD45Rbhi T cells are able to infiltrate colonic tissue and induce colitis upon transfer into lymphopenic mice, cotransfer of CD4+ CD45Rblo T cells results in pronounced protection from colitis induction, mediated by Tregs in this T cell population (fig. S9A) (24, 25). Accordingly, the transfer of Ly5.1+ CD4+ CD45Rbhi T cells caused pronounced colitis, accompanied by body weight loss (fig. S9B) and colonic inflammation (fig. S9C), whereas the cotransfer of CD4+ CD45Rblo T cells from L2/L2 mice resulted in the restoration of body weight loss (fig. S9B) and reduced colonic inflammation (fig. S9D). In marked contrast, sort-purified cKO CD4+ CD45Rblo T cells were unable to attenuate disease development (fig. S9, B and E). Along these lines, reduced regulatory cKO donor T cell numbers were found in spleen (fig. S9F) and mesenteric lymph nodes (fig. S9G) of recipient mice, suggesting that the disturbed protective functions of cKO T cells resulted from reduced in vivo expansion. While L2/L2 Tregs were able to significantly reduce colonic Ifng mRNA expression and cKO T cells failed to do so (fig. S9H), there was no difference in Il1b expression (fig. S9I), but Treg populations from both L2/L2 and cKO mice promoted a significant reduction of colitis-induced Tnfa mRNA expression (fig. S9J). To exclude that the lack of protection by CD4+ CD45Rblo cKO T cells was caused by a reduced frequency of Treg in this population, we determined the percentage of splenic FoxP3+ CD4+ T cells in L2/L2 and cKO mice before transfer, revealing no difference (fig. S9K).

Impaired virus clearance by LRH-1–deficient cytotoxic T cells

The impact of LRH-1 deletion was more pronounced in regard to mature CD4+ T cell numbers, while CD8+ T lymphocytes were less affected. To investigate the effect of LRH-1 deletion on CD8+ T cell effector functions, we analyzed the antigen-specific T cell expansion and viral clearance after lymphocytic choriomeningitis virus (LCMV) infection, which is critically dependent on cytotoxic CD8+ T cells (26). Accordingly, LCMV infection resulted in increased numbers of splenic CD8αβ+ αβ T cell receptor–positive (TCRαβ+) cells in L2/L2 mice at day 6 after LCMV infection, indicative of a virus-specific T cell expansion. Although cKO mice had lower numbers of CD8αβ+ TCRαβ+ cells before infection, a clear virus-induced expansion was observed at days 6 and 8 after infection (Fig. 6A). Along with the virus-induced cytotoxic T cell expansion, an increase in CD69+ and CD25+ CD8+ T cells was observed at day 6 after infection; however, no difference was seen between L2/L2 and cKO mice (Fig. 6, B and C). Similarly, a comparable virus-induced increase of CD25+ CD8+ T cells in mesenteric lymph nodes was observed (fig. S10A). The expansion of virus-specific T cells was confirmed using intracellular staining of IFN-γ. While only ex vivo restimulation with the LCMV peptide gp33, but not the control peptide adn5, resulted in IFN-γ+ CD8+ T cells at days 6 and 8 after infection, no difference was seen between control and cKO mice (Fig. 6D). Similar observations were made when analyzing IFN-γ levels in serum samples (fig. S10B). Thus, LRH-1 deficiency does not appear to affect LCMV-induced CD8+ T cell activation and expansion.

Fig. 6 Impaired virus clearance by LRH-1–deficient cytotoxic T cells.

(A) Virus-induced expansion of CD8αβ+ TCRαβ+ T cells [0 days post infection (dpi), n = 11; 6 dpi, n = 9; 8 dpi, n = 3 mice]. (B and C) Up-regulation of activation marker CD69 (B) and CD25 (C) in CD8+ T cells after LCMV infection (0 dpi, n = 11; 6 dpi, n = 9; 8 dpi, n = 6 mice). (D) Quantification of IFN-γ–producing virus-specific CD8+ T cells after ex vivo restimulation with buffer control, Adn5 control peptide or GP33 peptide (0 dpi: n = 8; 6 dpi: n = 9; 8 dpi: n = 6 mice). (E to H) Virus titers after LCMV infection of L2/L2 or cKO mice in the spleen (E), liver (F), small intestine (G), and serum (H) (6 dpi, n = 9; 8 dpi, n = 6; 10 dpi, n = 7 mice). (I) Immunohistological detection of LCMV-infected cells in liver sections in L2/L2 or cKO mice at days 6 and day 10 after infection. VL4 LCMV nucleoprotein, red; nuclei, blue (DAPI). Representative pictures are shown. Scale bars, 150 μm. (J) mRNA expression of the cytotoxic T cell effector molecule perforin (Prf1) (0 dpi, n = 5; 6 dpi, n = 5; 8 dpi, n = 6; 10 dpi, n = 2 mice). (K) DNA fragmentation of GP33 or Adn5-loaded EL4 cells by virus-specific T cells, 8 days after LCMV infection, measured by loss of 3H-thymidine (n = 3 mice; representative experiment of two is shown). *P < 0.05, **P < 0.01, ***P < 0.001. E:T, effector:target.

An important effector function of virus-specific T cells is the killing of virus-infected cells and, thereby, virus clearance. Accordingly, a massive reduction of virus particles was seen in the spleen, liver, small intestine, and serum of LCMV-infected control mice at day 10 after infection (Fig. 6, E to H). However, while CD8+ cKO T cells appeared to become properly activated and expand after LCMV infection, they completely failed to control viral expansion, resulting in a high viral load in all tissues examined (Fig. 6, E to H). The lack of viral clearance was also obvious when detecting LCMV-infected cells in liver tissue using immunofluorescence. While control mice almost completely cleared the virus from liver tissue at day 10 after infection, cKO mice still contained high numbers of virus-producing hepatocytes (Fig. 6I). As LCMV clearance is critically dependent on perforin-mediated cytotoxicity (27), we analyzed Prf1 and Gzmb mRNA expression in the spleen of control and cKO mice, revealing an infection-induced increase in both mouse lines (Fig. 6J and fig. S10C). Furthermore, spleen cells and sorted CD8+ T cells from both virus-infected control and cKO mice killed LCMV peptide–presenting target cells equally well (Fig. 6K and fig. S10D). This indicates that impaired expression of cytotoxic effector molecules is not responsible for the inability of cKO T cells to clear LCMV.


The expression and role of LRH-1 has been well described in various endodermal tissues, while considerably less is known about its expression and role in other tissues, specifically the hematopoietic system. The first indirect report investigating the LRH-1 expression in pancreatic tumors, where they accidentally observed LRH-1–positive tumor-infiltrating leukocytes, came from Benod and colleagues (28). More recently, Lefèvre et al. (15) described LRH-1 expression in macrophages and a role of LRH-1 in cytokine-induced differentiation. Our own studies revealed a role of LRH-1 in the transcriptional control of Fas ligand expression in CD4+ T cells (16). However, besides these few studies, the expression and role of LRH-1 in the hematopoietic system is currently unexplored. One of the underlying reasons seems to be that steady-state levels are extremely low; thus, LRH-1 expression and function in immune cells had been largely ignored. We observed that expression levels in unstimulated thymocytes and mature T cells are about 100 to 1000 times lower than those observed in the intestine and the liver (fig. S1, A to D). However, LRH-1 expression in T cells appears to be rather dynamic. Activation of T cells by mitogenic stimuli strongly induces LRH-1 promoter activity, expression, and function (Fig. 3, A, B, D, and F). LRH-1 induction by mitogenic stimuli correlates well with the up-regulation of cell cycle–regulating genes, i.e., c-Myc and cyclin E1 (Fig. 3, C to E). This is in line with the established role of LRH-1 in the transcriptional control of these cyclin genes and the regulation of proliferation of intestinal epithelial stem cells and tumor cells (19). Supporting a suggested role of LRH-1 in cell cycle progression, we observed a profound inhibition of activation-induced proliferation when LRH-1 was either genetically deleted or pharmacologically inhibited (Fig. 4, A to F). Reduced proliferation was observed not only in vitro but also during homeostatic expansion in vivo (fig. S7, A and B) and the pathogenesis of experimental colitis (Fig. 5, A to G). This proliferation deficiency is unlikely a consequence of improper T cell activation, as LRH-1–deficient T cells readily up-regulated early activation markers (Fig. 2, A and B), and levels of IL-2 and IFN-γ produced often exceeded those of control T cells (Fig. 2, C and D). Given the previously described role of LRH-1 in the transcriptional regulation of cell cycle–related genes, e.g., cyclin D1 and E1 and c-Myc, in the liver and intestinal epithelial cells (19), the lack of LRH-1 very likely leads to defects in activation-induced proliferation in T cells due to reduced expression of these LRH-1 target genes. We have seen that cyclin E1 (Cdne1) and c-Myc (Myc) are expressed at reduced levels in LRH-1–deficient T cells (Fig. 4, G and H).

Unexpectedly, we observed quite substantial differences between the impact of LRH-1 deletion on CD4+ versus CD8+ T cells (Fig. 1, D, H, and I). While LRH-1 deletion resulted in a threefold reduction of mature peripheral CD4+ T cells, CD8+ T cells were only reduced by half (Fig 1, H and I). Similarly, activation-induced proliferation of CD8+ T cells appeared to be less dependent on LRH-1, as in a mixed spleen cell population CD8+ T cells divided normally in response to TCR stimulation, while CD4+ were severely impaired (Fig. 4, A and B). During homeostatic T cell expansion in vivo, LRH-1–deficient CD8+ T cells barely differed from wild-type T cells, while LRH-1–deficient CD4+ T cells were not able to compete with wild-type T cells in repopulating Rag1−/− recipients (fig. S7, A and B). In addition, during LCMV infection, a normal expansion of LRH-1–deficient virus-specific CD8+ T cells was observed, although they started at a lower level (Fig. 6A). As the CD4 promoter becomes activated already at the double-positive stage during thymic maturation, the LRH-1 gene is already deleted at this stage of T cell development (fig. S4, A to D). As CD4 promoter activity drops in mature CD8+ T cells, an insufficient duration of Cre expression and associated failure to completely delete the LRH-1 gene could account for the reduced impact of LRH-1 deletion in CD8+ T cells. However, the use of the mTmG double-fluorescent reporter mouse revealed efficient Cre-mediated conversion of tdTomato-positive to GFP-positive cells in both CD4+ and CD8+ T cell populations, although Cre-mediated excision appeared to be slightly more efficient in CD4+ T cells (fig. S4, F to H). Nonetheless, it is unlikely that incomplete LRH-1 deletion accounts for the differences in proliferation observed, as the vast majority of LRH-1–deficient CD8+ T cells were also GFP+, indicating a sufficient phase of active Cre recombinase during their development. The difference in LRH-1 dependency between CD4+ and CD8+ T cells may be rather related to the presence or absence of compensatory signals. LRH-1–deficient CD8+ T cells showed proliferation defects only when highly purified (Fig. 4F) but not when stimulated in the presence of antigen-presenting cells (Fig. 4, B and D). Thus, costimulatory signals distinct from B7-CD28 interaction may provide compensatory signals, enabling LRH-1–independent proliferation.

Defects in CD4+ T cell proliferation correlated well with the inability of LRH-1–deficient CD45RBhi CD4+ T cells to promote inflammation in the transfer model of colitis (Fig. 5, A to H). Similarly, the LRH-1–deficient Tregs in the CD45RBlo CD4+ T cell population also failed to expand in vivo and to mediate protection from colitis induction (fig. S9, A to G). This inability to prevent transfer colitis is not due to a generally reduced level of Tregs, as the percentage of FoxP3+ Tregs among splenic CD4+ T cells was similar in control and LRH-1–deficient mice (fig. S9K). Thus, this rather suggests that in vivo proliferation of CD4+ Tregs is also critically controlled by LRH-1. A somewhat different situation was observed in the LCMV infection model. Although T cell–specific LRH-1–deficient mice had generally lower numbers of CD8+ T cells before infection, LCMV promoted a rapid expansion of virus-specific T cells even in the absence of LRH-1, as seen by a general increase in CD8+ TCRαβ+ T cells in the spleen (Fig. 6A) and lymph nodes (fig. S10A) and comparable numbers of IFN-γ–producing CD8+ T cells upon ex vivo restimulation with the LCMV-specific peptide gp33 (Fig. 6D). Also along these lines, comparable production of cytokines and an increase in activation marker–positive (CD25 and CD69) CD8+ T cells were observed (Fig. 6, B and C, and fig. S10B). In marked contrast, LRH-1–deficient cytotoxic T cells failed to control viral spreading. While in control animals LCMV titers were reduced at day 10 after infection by several logs and barely detectable, viral titers in mice with defective LRH-1 expression remained at a high level (Fig. 6, E to I). Given that virus-specific T cells appeared to expand normally, this difference is difficult to reconcile. An obvious guess would be that LRH-1–deficient CD8+ T cells fail to express cytotoxic effector molecules. However, gene expression analysis revealed that granzyme B and perforin are, at least on an mRNA level, induced upon viral infection and equally expressed in control and cKO mice (Fig. 6J and fig. S10C). While we have recently identified the cytotoxic effector molecule Fas ligand as a target gene of LRH-1 (16), it is unlikely that impaired Fas ligand expression could account for inefficient viral clearance, as perforin (27) but not Fas ligand (29) is critical for controlling LCMV spreading. Furthermore, we confirmed that T cells from both virus-infected control and cKO mice killed LCMV peptide–presenting target cells equally well, suggesting normal cytotoxic effector functions in both mouse lines.

Thus, alternative mechanisms have to be considered for the inefficiency of cKO mice to control LCMV. A possible explanation is that CD8+ cells expand but cannot reach critical numbers to efficiently eliminate LCMV. Another possibility is that cKO T cells are more readily exhausted and thereby cannot control viral expansion. Alternatively, an efficient cytotoxic T cell response may also depend on proper T helper cell responses, which is likely impaired in cKO mice with reduced proliferative potential of CD4+ T cells. In support of this hypothesis, it was reported that CD4+ T cell help was required for effective CD8+ T cell–mediated resolution of LCMV infection (30). Thus, T helper cells have been shown to critically orchestrate expansion and recruitment of cytotoxic CD8+ T cells [reviewed in (31)].

An interesting aspect of our study is its potential translational application. Although no endogenous ligands for LRH-1 have been identified yet, a number of pharmacological inhibitors with high specificity have been developed. The fact that transient inhibition of LRH-1 with two inhibitors that act in different mechanisms [3d2 (Fig. 4, C and D) (20) and SR1848 (Fig. 4, E and F) (21)] was able to confirm that the proliferation-regulating role of LRH-1 in T cells and the relatively low expression of LRH-1 in comparison to the liver and intestine may open an interesting therapeutic window. The use of LRH-1 inhibitors in relatively low doses could permit the efficient inhibition of T cell mediated immunopathologies while not affecting vital functions of other tissues. While studying the role of LRH-1 in the transcriptional control of Fas ligand, we were already able to provide proof of principle for this concept. Injection of the lectin ConA into mice leads to rapid T cell activation and Fas ligand–dependent damage of the liver, which was strongly reduced after administration of 3d2 (16). The liver seems to tolerate 3d2 relatively well, as only a minimal increase in 3d2-induced serum transaminases was observed.

In summary, we have described a novel role for LRH-1 in T cell development, proliferation, and effector functions. Furthermore, our study proposes LRH-1 as an emerging new target in the treatment of T cell–mediated inflammatory diseases.



Animals were housed in individually ventilated cages at the central animal facility of the University of Konstanz. Male and female mice aged 7 to 15 weeks on C57BL/6 background carrying the leukocyte marker CD45.2 were used for all in vitro and in vivo experiments unless otherwise stated. For mRNA quantitation experiments, wild-type C57BL/6 mice were used. Nr5a2L2/L2 (L2/L2) mice (9) were bred with CD4-Cre transgenic mice (32) to obtain Nr5a2L2/L2 CD4-Cre (cKO) animals and kept heterozygous for Cre expression. The Cre-reporter mTmG transgenic mice were provided by U. Koch (École Polytechnique Fédérale de Lausanne) and crossed with Nr5a2L2/L2 CD4-Cre to obtain Nr5a2L2/L2 mTmG CD4-Cre mice. Throughout all experiments, cKO animals were compared to their corresponding floxed littermate controls. Ly5.1 mice expressed the leukocyte marker CD45 isoform 1. Rag1−/− (provided by M. Basler, University of Konstanz) mice were bred on a Ly5.1 background.

Cell culture and reagents

All experiments were performed in 96-well round-bottom plates (Greiner) in technical triplicates with 200,000 cells per well in complete culture medium unless otherwise stated. Cell culture media and supplements were purchased from Sigma-Aldrich, if not otherwise indicated.

Human leukemic Jurkat cells (the American Type Culture Collection) with stably transfected SV40 large T antigen (lT) were maintained in RPMI 1640 medium supplemented with 5% fetal calf serum (FCS), 2 mM l-glutamine, and gentamycin (20 μg/ml) at 37°C and 5% CO2. Cells have been routinely tested for the absence of mycoplasma. Mouse T cells were cultured in RPMI 1640 medium supplemented with 10% FCS, 2 mM l-glutamine, gentamycin (20 μg/ml ), and 50 μM β-mercaptoethanol.

Luciferase reporter assay

The 1.5-kb human LRH-1 promoter luciferase reporter and control constructs have been described previously (33). Jurkat lT cells were transfected by nucleofection (Amaxa) (16), seeded at a concentration of 3 × 105 cells/ml in 12-well plates, and cultured overnight before treatment with phorbol myristate acetate (PMA) (50 ng/ml) and ionomycin (1.1 μg/ml ) (both from Enzo Life Sciences) or serum withdrawal for further 24 hours. Luciferase assays were carried out as described previously (34). Chemiluminescence was measured using an Infinite 200 PRO (Tecan) microplate reader. pCMV β-galactosidase was used to correct for transfection efficiencies. To test endogenous LRH-1 activity, Jurkat lT cells were transfected with a luciferase reporter construct containing 5× LRH-1 response elements derived from the small heterodimer partner promoter (35).

T cell isolation and activation

Single-cell suspensions from the spleen, thymus, and lymph nodes were prepared by manual disruption between frosted glass slides. Human PBMCs were prepared by Ficoll density centrifugation.

For Nr5a2 mRNA quantitation experiments, freshly isolated splenocytes were stimulated with 5 μg/ml ConA (Sigma-Aldrich) or PMA (5 ng/ml) and ionomycin (200 ng/ml) for 3 hours and immediately frozen in appropriate reagents for subsequent mRNA analysis. Alternatively, cells were activated with anti-CD3ε antibody (3 μg/ml) (clone 145-2C11) and precoated on 96-well flat bottom plates in 50 mM tris (pH 9) and soluble purified anti-CD28 antibody (1 μg/ml) (BioLegend). ConA-activated mouse splenocytes were prepared as a source of T cell blasts and cultured as described (16). Human PBMC were treated with phytohemagglutinin (5 μg/ml) (Sigma-Aldrich) for 3 hours before RNA isolation.

RNA isolation and real time quantitative polymerase chain reaction

Tissue RNA was isolated using peqGOLD TriFast (Peqlab) according to the manufacturer’s instructions. For sorted cells, the ReliaPrep RNA Mini Kit (Promega) was used. Complementary DNA (cDNA) was prepared using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative polymerase chain reaction (PCR) was performed on a StepOnePlus Instrument using SYBR Green PCR Master Mix (both from Applied Biosystems). Primers were designed to span exon-exon junctions. Gene expression was normalized to Actb or GAPDH, respectively. The following primers were used: Nr5a2, 5′-TTGAGTGGGCCAGGAGTAGT-3′ (forward) and 5′-ACGCGACTTCTGTGTGTGAG-3′ (reverse); Actb, 5′-TATTGGCAACGAGCG GTTCC-3′ (forward) and 5′-GCACTGTGTTGGCATAGAGG-3′ (reverse); Myc, 5′-CTAGTGCTGCATGAGGAGAC-3′ (forward) and 5′-TTTGCCTCTTCTCCACAGAC-3′ (reverse); Cdne1, 5′-TTCTGCAGCGTCATCCTCTC-3′ (forward) and 5′-TGTGCCAAGTAGAACGTCTC-3′ (reverse); Tnfa, 5′-TAGCCCACGTCGTAGCAAAC-3′ (forward) and 5′-ACAAGGTACAACCCATCGGC-3′ (reverse); Il1b, 5′-TGCCACCTTTTGACAGTGATG-3′ (forward) and 5′-ATGTGCTGCTGCGAGATTTG-3′ (reverse); Ifng, 5′-CGGCACAGTCATTGAAAGCC-3′ (forward) and 5′-TGTCACCATCCTTTTGCCAGT-3′ (reverse); Prf1, 5′-TCTTGGTGGGACTTCAGCTTT-3′ (forward) and 5′-TCCATACACCTGGCACGAAC-3′ (reverse); Gzmb, 5′-GCTGCTAAAGCTGAAGAGTAAGG-3′ (forward) and 5′-TCACATTGACATTGCGCCTG-3′ (reverse); human NR5A2, 5′-GGGCAACAAGTGGACTATTC-3′ (forward) and 5′-CCAGCTGGAAGTTTTCAAGG-3′ (reverse); and human GAPDH, 5′-ATGGAGAAGGCTGGGGCTCA-3′ (forward) and 5′-AGTGATGGCATGGACTGTGGTCAT-3′ (reverse).

Flow cytometry and cell sorting

Single-cell suspensions of freshly isolated T cells were stained with one or more of the following monoclonal antibodies: anti–CD3ε-APC (anaphase-promoting complex) (145-2C11), anti–CD8-PE (phycoerythrin) (53-6.7), anti–CD45.1-BV421 (A20), anti–CDCD45.2-APC (104), anti–CD4-FITC (fluorescein isothiocyanate) (RM4-5), and anti–CD8α–PerCP-Cy5.5 (53-6.7) from BD Pharmingen; anti–CD4-Cy5 (GK1.5), anti–CD25-APC (PC61.5), anti–F4/80-APC (BM8), anti–NK1.1-PE (PK136), anti–CD45R (B220)–PerCP-Cy5.5 (RA3-6B2), anti–CD4-PE (GK1.5), anti–CD8β-FITC (eBioH35-17.2), anti–FoxP3-PE (FJK-16s), and anti–TCRβ-PE-Cy7 (H57-597) from eBioscience; anti–IFN-γ–FITC (XMG1.2) and anti–CD69-PE (H1.2F3) from BioLegend; and anti–CD3ε-FITC (145-2C11; purified from culture supernatant); anti-VL4 antibody was a gift from M. Basler (University of Konstanz). The following polyclonal secondary antibodies were used: goat anti-rat immunoglobulin G (IgG)–Alexa Fluor 568 (Invitrogen) and goat anti-hamster IgG-FITC cocktail (G70-204/G94-56, BD Pharmingen). Flow cytometry was performed on an LSRFortessa (BD Biosciences). For cell sorting, an Aria III (BD Biosciences) instrument was used. Data were analyzed using FlowJo software (TreeStar).

For intracellular cytokine staining, cells were restimulated in vitro with gp33 peptide (1 μg/ml) (KAVYNFATC) or adn5 control peptide (SGPSNTPPEI; both from ProImmune) and incubated with brefeldin A (10 μg/ml) (LKT Labs) for 5 hours. Staining (anti–CD8-PE and anti–IFN-γ–FITC) was performed as described before (36). For detection of Tregs, splenocytes were surface-stained with anti–CD4-FITC before they were stained intracellularly with anti–FoxP3-PE using the Fix/Perm Kit (BioLegend).

Histology and immunohistochemistry

Mouse colons and small intestinal segments were swiss-rolled and fixed in 10% formalin, followed by paraffin embedding. Tissue sections (4 μm) were stained with hematoxylin and eosin for histological analysis using standard procedures.

For immunohistochemistry, organs were immediately embedded in O.C.T. (Tissue-Tek) cryosection medium and shock frozen on dry ice. Sections (7 μm) were cut, and samples were fixed with fresh acetone for 1 min, treated with anti-CD16/32 Fc-block, stained with appropriate antibodies, and detected with Alexa Fluor 568– and FITC-conjugated secondary antibodies. mTmG cryosections were directly analyzed without further processing. All sections were embedded in 4′,6-diamidino-2-phenylindole containing Fluoroshield (Sigma-Aldrich) and analyzed on an Axio Observer Z1 (Zeiss) microscope. Fiji software (37) was used for image processing.

Cell death assays

Isolated T cells were cultured in the presence or absence of dexamethasone (Sigma-Aldrich), etoposide (Enzo Life Sciences), staurosporine (Sigma-Aldrich), plate bound anti-CD3 antibody, or solvent control (dimethyl sulfoxide or ethanol) for 8 hours. Cells were counterstained with anti–CD4-Cy5 and anti–CD8-PE in binding buffer and analyzed for annexin-V positivity, as described (38).

Proliferation assays

Lymphocyte proliferation was monitored using CFSE (Sigma-Aldrich) dilution, as described (39). Cells were counterstained with anti–CD4-Cy5 and anti–CD8-PE and analyzed by flow cytometry. In some experiments, the pharmacological LRH-1 inhibitor 3d2 (20), control substance cd7 (20) (both synthesized by ChemBridge Corporation), or SR1848 (21) were added 30 min before T cell activation.

For 3H-thymidine incorporation assays, cells were pulsed with 0.5 μCi 3H-thymidine (Hartmann Analytic) per well for 18 hours and harvested using a semiautomated cell harvester (Packard). 3H-thymidine integration was determined by scintillation counting. For cell cycle profile analysis, wild-type splenocytes were cultured for indicated time points before nuclear DNA content was analyzed as described earlier (40).

Analysis of cytokine expression and serum immunoglobulin titers

Secretion of murine IL-2 and IFN-γ was quantified by enzyme-linked immunosorbent assay (ELISA) using matched antibody pairs [anti-mouse IFN-γ (RA-6A2/XMG1.2) and anti-mouse IL-2 (JES6-1A12/JES6-5H4)]. All reagents were purchased from BioLegend and used according to the manufacturer’s instructions.

For the determination of anti-ovalbumin–specific IgG antibodies in mouse serum, polystyrene microtiter plates (Nunc) were coated with ovalbumin (100 μg/ml) and horseradish-conjugated goat anti-mouse IgG (Jackson ImmunoResearch) was used for detection.

Ovalbumin immunization and in vitro restimulation

Mice were subcutaneously injected with 100 μg of ovalbumin emulsified in incomplete Freund’s adjuvant (Invivogen). After 14 days, the spleen and serum were isolated for further analysis of antibody production by ELISA and proliferation of antigen-specific T cells by in vitro restimulation and 3H-thymidine incorporation.

Adoptive transfer studies

Isolated splenocytes were sorted accordingly on the basis of CD4 and CD8α staining. A minimal purity of 98% was confirmed, and appropriate cell populations were mixed in a 1:1 ratio. A total of 2 × 106 cells were intraperitoneally injected into Rag1−/− recipient mice. Animals were euthanized 21 days after injection, and secondary lymphatic organs were analyzed for T cell numbers, subsets, and origin (CD45.1 versus CD45.2).

Transfer colitis and protection

Splenocytes were sorted on the basis of CD4-PE and CD45Rb-FITC staining for CD4+ CD45RBhi and CD4+ CD45RBlo, as described previously (23). A minimum purity of 98% was ensured. For the induction of transfer colitis, Rag1−/− received 0.5 × 106 CD45Rbhi cells from either L2/L2 or cKO mice by intraperitoneal injection. For transfer colitis protection experiments, Rag1−/− animals were intraperitoneally injected with 0.5 × 106 CD4+ CD45Rbhi cells from Ly5.1 donor mice and 0.25 × 106 CD4+ CD45Rblo T cells from either L2/L2 or cKO mice. In both experiments, body weight was recorded to monitor disease progression and estimate colitis severity. Animals were euthanized once most of the mice reached a critical weight loss of 20%. Cells from secondary lymphatic organs were isolated, stained for CD4 and CD45.1, and analyzed by flow cytometry. CD4+ CD45.1 cells were considered as repopulating donor T cells.

LCMV infection and virus-specific T cell cytotoxicity

Mice were intravenously injected with 2.5 × 105 plaque-forming units of LCMV strain WE in phosphate-buffered saline and euthanized at day 6, 8, or 10, respectively. For characterization and intracellular staining, T cells were isolated from spleen and mesenteric lymph node and stained as described above. Virus titers were analyzed in different organs using plaque assay, as described previously (41). For the detection of virus-specific cytotoxic T cells, EL4 thymoma cells were labeled with 3H-thymidine (10 μCi/ml) overnight and pretreated with gp33 or adn5 peptide. Effector cells were mixed with 2 × 104 target cells in triplicate at the indicated effector:target ratio for 16 hours before the 3H-thymidin content was determined by scintillation counting. DNA fragmentation was calculated as previously described (16).

Study approval

All animal experiments were performed in accordance with German animal experimentation regulations approved by the Ethics Review Board of the Regierungspräsidium Freiburg. The G*Power software was used to predetermine sample size of all animal experiments.


Values reported are means ± SD. Statistical significance was determined by unpaired two-tailed Student’s t test using Prism (version 6.0b, GraphPad). Where indicated, one-way analysis of variance (ANOVA) with Bonferroni’s multiple comparisons test was used. P values are indicated as follows: *P < 0.05, **P < 0.01, and ***P < 0.001. Confidence interval was set to 95%.


Supplementary material for this article is available at

Fig. S1. Nr5a2 expression in immature and mature T lymphocytes.

Fig. S2. Minor impact of Nr5a2 deletion on immature T lymphocytes.

Fig. S3. Reduced mature T cells in axial and mesenteric lymph nodes of LRH-1–deficient mice.

Fig. S4. Efficiency of LRH-1 deletion in immature and mature T cells.

Fig. S5. Apoptosis induction in LRH-1–deficient thymocytes.

Fig. S6. Apoptosis induction in LRH-1–deficient splenocytes.

Fig. S7. Homeostatic expansion is impaired in LRH-1–deficient T cells.

Fig. S8. Impaired ovalbumin-induced T cell expansion and antibody production in LRH-1–deficient mice.

Fig. S9. Reduced Tregs function of LRH-1–deficient T cells.

Fig. S10. Activation and cytotoxicity of LRH-1–deficient CD8+ T cells.

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.


Acknowledgments: We thank A. Glöckner, J. Negrassus, and D. Eichbichler for expert technical help; A. Holz and H. Henseleit for animal caretaking; M. Basler for the VL4-antibody; U. Koch and F. Radtke for the mTmG reporter mouse; and M. Groetrupp for advice; S. Martin for stimulating discussions and suggestions. Funding: This work was supported by grants from the German Science Foundation (DFG) (BR 3369/4-1, INST 38/500-1, and INST 38/498-1) to T.B. C.S. was supported by a fellowship from the graduate school RTG1331 (supported by the DFG). J.H. was supported by a fellowship from the Chinese Scholarship Council, and S.M. was supported by a Carl Zeiss fellowship. Author contributions: C.S., A.-L.G., J.H., T.S., C.R., P.G.-B., E.D., T.S.P., S.M., C.Sch. and L.D. conducted experiments. C.S., T.S.P., and C.R. sorted cells. K.S. and N.C. provided valuable reagents and advised on the manuscript. C.S. and T.B. wrote the manuscript. T.B. supervised the project and finalized the manuscript. Competing interest: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. The floxed LRH-1 mouse line can be provided by K.S. pending on scientific review and a completed material transfer agreement. Request for this mouse line should be submitted to K.S. ( Additional data related to this paper may be requested from the authors.

Stay Connected to Science Advances

Navigate This Article