A Prkd2 mutation causes elevated serum antibodies

Using the chemical germline mutagenN-ethyl-N-nitrosourea (ENU), we mutated approximately 29.2% of all mouse protein coding genes to a state of phenovariance and tested these mutations three times or more in the homozygous state in a screen for aberrant IgE levels in serum in response to immunization with alum-precipitated ovalbumin (OVA/alum) (20). In all, 125,073 coding/splicing changes, distributed among 65,374 mice from 2,285 pedigrees were tested for phenotypic effects.

A mutation inPrkd2, namedPurnama (20), was associated with increased serum concentrations of total and OVA-specific IgE after OVA/alum challenge (Fig. 1A,B). ThePrkd2^Pur mutation resulted in a tryptophan to arginine substitution at amino acid 807 (p.W807R) within the Prkd2 kinase domain. We recreated thePurnama mutation (Prkd2^W807R) in mice by utilizing CRISPR/Cas9 gene targeting.Prkd2^W807R/W807R mice exhibited excessive production of IgE in response to OVA/alum (Fig. S1A,B). Moreover, expression of Prkd2^W807R protein was significantly lower than that of wild-type Prkd2 when overexpressed in HEK293T cells (Fig. S1C). The IgE phenotype inPurnama mutants was not limited to the response to OVA/alum as they produced IgE in excess after immunization with another model allergen, papain (Fig. S2A).Prkd2 encodes an 875-amino acid serine/threonine kinase most highly expressed in the spleen, lymph node, thymus, and lung among those tissues examined (Fig. S3A). In the spleen, T cells and B cells expressed Prkd2, with higher levels of expression by T cells compared to B cells (Fig. S3B). We also generated a null allele ofPrkd2 (Prkd2^−/−) with CRISPR/Cas9 gene targeting, which recapitulated the elevated total and antigen-specific IgE production in heterozygous and homozygous mice (Fig. 1C,D,Fig. S2B,C). Further examination of total antibody levels in the serum ofPrkd2^−/− mice showed aberrantly elevated basal levels of IgE (Fig. 1C), IgA (Fig. 1E), IgM (Fig. 1F), and IgG1 (Fig. 1G), which were further increased following OVA/alum immunization (Fig. 1C-G). Total IgG2b was also slightly increased both before and after immunization inPrkd2^−/− mice (Fig. 1H). In contrast, serum concentrations of IgG2a (Fig. 1I), IgG2c (Fig. 1J), and IgG3 (Fig. 1K) were similar inPrkd2^+/−,Prkd2^−/−, and wild-type mice irrespective of immunization.

Major immune cell populations were present at normal frequencies in the spleens ofPrkd2^−/− mice (Fig. S3C-K), suggesting that their development occurs normally. To test whether the increased antibody production observed inPrkd2^−/− mice was due to the function of hematopoietic cells, we performed bone marrow (BM) transplantation. IrradiatedRag2^−/− recipient mice engrafted withPrkd2^−/− BM displayed elevated serum antibodies compared toRag2^−/− mice engrafted withPrkd2^+/+ BM, both before (IgA and IgM) and after (IgE, IgG1, IgA, and IgM) OVA/alum challenge (Fig. 1L-P), indicating that deficiency of Prkd2 in the hematopoietic compartment is sufficient to cause excessive antibody production. Using T cell receptor alpha knockout (Tcra^−/−) mice, we confirmed that except for IgM production, the altered antibody production inPrkd2^−/− mice was T cell-dependent (Fig. S3L-P). In summary, loss-of-function mutations inPrkd2 result in excessive T cell-dependent production of IgE, IgG1, and IgA.

Excessive cell autonomous TFH development occurs in Prkd2^−/− mice

IL-4, produced by both Th2 and TFH, induces the expression of activation-induced cytidine deaminase and subsequent antibody isotype switching to IgE and IgG1 (21,22). We found thatPrkd2^−/− CD4⁺ T cells produced significantly more IL-4 thanPrkd2^+/+ CD4⁺ T cells when stimulated with PMA/ionomycin or anti-CD3/CD28 antibodiesin vitro (Fig. 2A,B). In addition, flow cytometric analysis of cells fromPrkd2^−/− mice crossed toIl4 reporter mice that contain a bicistronic IRES-EGFP reporter cassette inserted in the endogenousIl4 locus (known as4-get mice) (23) showed greater percentages of GFP-expressing CD4⁺ T cells inPrkd2^−/− spleens compared toPrkd2^+/+ spleens (Fig. 2C). We therefore hypothesized that Prkd2 deficiency leads to the enhanced development of Th2 or TFH. In support of enhanced TFH development, flow cytometric quantification of TFH (CXCR5^highPD-1^highCD4⁺) in spleens and lymph nodes ofPrkd2^−/− mice revealed significantly increased frequencies and numbers compared to TFH inPrkd2^+/+ mice (Fig. 2D,Fig. S4A-C,Fig. S5). Consistent with the increased TFH numbers, frequencies and numbers of GC B cells and GC formation were also elevated inPrkd2^−/− mice, even without immunization (Fig. 2E,F,Fig. S4A,D,E). The frequency of plasma cells was also significantly higher in bone marrow fromPrkd2^−/− mice than inPrkd2^+/+ mice (Fig. 2G).

In contrast, expression of the Th2-inducing transcription factors GATA3 and STAT6 (24) was comparable inPrkd2^−/− andPrkd2^+/+ CD4⁺ T cells (Fig. S6A). Moreover, when CD4⁺ T cells were culturedin vitro under Th2-polarizing conditions, a smaller percentage ofPrkd2^−/− cells thanPrkd2^+/+ cells induced IL-4 expression after PMA/ionomycin stimulation (Fig. S6B). Indeed, there was an increased percentage of IL-4⁺ TFH and a reduced percentage of IL-4⁺ Th2 inPrkd2^−/− spleens compared toPrkd2^+/+ spleens, resulting in a greatly increased ratio of TFH to Th2 cells among IL-4-expressing CD4⁺ T cells inPrkd2^−/− spleens compared to that inPrkd2^+/+ spleens (Fig. 3A,B). These data indicate that an overabundance of TFH exist inPrkd2^−/− mice.

The transcriptional repressor Bcl6 plays a critical role in the development of TFH (25-30). Bcl6-deficient T cells fail to develop into TFH and sustain germinal center responses. UsingPrkd2^−/− mice conditionally lacking Bcl6 in CD4⁺ T cells (Prkd2^−/−Bcl6^fl/flCD4-Cre⁺), which have no TFH (Fig. S7A), we confirmed that the elevated frequencies of GC B cells (Fig. S7B) and increased serum concentrations of IgE (Fig. S7C,D), IgG1 (Fig. S7E), and IgA (Fig. S7F) inPrkd2^−/− mice were TFH-dependent. Interestingly, we observed that specific deletion ofBcl6 in wild-type CD4⁺ T cells resulted in elevated IgM production (Fig. S7G).

We next assessed the intrinsic proliferative potential ofPrkd2^−/− TFH by transferring a mixture containing equal numbers ofPrkd2^+/+ (CD45.1⁺) andPrkd2^−/− (CD45.2⁺) BM cells into lethally irradiatedRag2^−/− mice. As a result,Prkd2^−/− TFH expanded more robustly thanPrkd2^+/+ TFH, resulting in an increased frequency ofPrkd2^−/− TFH compared toPrkd2^+/+ TFH in the spleens of recipient mice (Fig. 3C). We observed a similar effect when naïve CD4⁺ T cells (CD3⁺CD4⁺CD62L^highCD44^low) were adoptively transferred intoRag2^−/− mice (Fig. 3D). These data indicate thatPrkd2^−/− TFH have a cell intrinsic proliferative advantage compared to wild-type TFH.

It was previously reported that CD4⁺ T cells from Prkd2 mutant mice, in which Ser707 and Ser711 were mutated to alanine within the Prkd2 kinase domain activation loop, exhibited reduced ability to produce IL-2 (17), a negative regulator of TFH development (31,32). However, we found that IL-2 production byPrkd2^−/− naïve CD4⁺ T cells was comparable to that ofPrkd2^+/+ naïve CD4⁺ T cells when stimulatedin vitro (Fig. S8A). Furthermore, frequencies of Th17 and Treg cells, Th cell subsets negatively or positively regulated by IL-2 (33), respectively, were unaffected or slightly increased inPrkd2^−/− mice (Fig. S8B,C). Therefore, neither IL-2 production nor IL-2-regulated immune cells were affected by Prkd2 deficiency.

Excessive GC B cell development in Prkd2^−/− mice is TFH-dependent

Bcl6 is also a key regulator of GC B cell development (34). Bcl6 mean fluorescence intensity (MFI) inPrkd2^−/− GC B cells was lower than that in the correspondingPrkd2^+/+ cells (Fig. S9A). On the other hand, consistent with increased numbers of GC B cells inPrkd2^−/− mice, frequencies of splenic Bcl6 expressing B cells were higher inPrkd2^−/− mice than inPrkd2^+/+ mice (Fig. S9B). Similar to GC B cells, Bcl6 MFI inPrkd2^−/− total B cells was lower than that in the correspondingPrkd2^+/+ cells (Fig. S9C). To test the effect of B cell-intrinsic Prkd2 deficiency on GC B cell development, we analyzed GC B cells inRag2^−/− mice engrafted with equal numbers ofPrkd2^+/+ (CD45.1⁺) andPrkd2^−/− (CD45.2⁺) BM cells. We found that in the resulting chimeras, frequencies ofPrkd2^−/− GC B cells were comparable to that ofPrkd2^+/+ GC B cells (Fig. S10A). We also generated bone marrow chimeras in whichPrkd2^−/− B cells developed in the presence ofPrkd2^+/+ TFH (Fig. S10B,C). We adoptively transferredPrkd2^+/+muMT BM (as a donor ofPrkd2^+/+ T cells) together with eitherPrkd2^+/+Tcra^−/− BM (Prkd2^+/+ B cell donor) orPrkd2^−/−Tcra^−/− BM (Prkd2^−/−B cell donor) into lethally irradiatedRag2^−/− mice, and analyzed GC B cells. In the presence ofPrkd2^+/+ T cells, the frequencies of GC B cells that developed were comparable betweenPrkd2^+/+ B cells andPrkd2^−/− B cells (Fig. S10D,E). Moreover, IgE, IgG1, and IgA were normally produced byPrkd2^−/− B cells with help fromPrkd2^+/+ TFH (Fig. S10F-I). Collectively, these findings strongly suggest that excessive differentiation ofPrkd2^−/− GC B cells is a consequence of Prkd2 deficiency in TFH, and not due to cell intrinsic defects ofPrkd2^−/− B cells.

Prkd2-dependent phosphorylation of Bcl6 limits Bcl6 nuclear translocation in CD4⁺ T cells

Bcl6 MFI inPrkd2^−/− TFH was slightly lower than inPrkd2^+/+ TFH (Fig. 4A). Consistent with increased numbers of TFH, frequencies of splenic Bcl6 expressing CD4⁺ T cells were higher inPrkd2^−/− mice than inPrkd2^+/+ mice (Fig. 4B). Bcl6 MFI in total CD4⁺ T cells was comparable betweenPrkd2^+/+ andPrkd2^−/− total CD4⁺ T cells (Fig. 4C). In both wild-type andPrkd2^−/− CD4⁺ T cells, Bcl6 was localized primarily in the nucleus (Fig. 4D). However, the amount of Bcl6 in nuclear fractions ofPrkd2^−/− CD4⁺ T cells was increased compared to that observed inPrkd2^+/+ CD4⁺ T cells. Conversely, when Bcl6 was co-transfected with Prkd2 in HEK293T cells, nuclear Bcl6 levels were reduced and cytoplasmic Bcl6 levels were increased (Fig. 4E). Notably, both mouse and human Prkd2 exhibited similar effects on the subcellular localization of Bcl6. These findings suggest that Prkd2 limits the nuclear translocation of Bcl6 in CD4⁺ T cells.

We examined the possible interaction between Prkd2 and Bcl6 by co-transfecting tagged versions of both proteins into HEK293T cells. Prkd2 interacted with Bcl6 in reciprocal coimmunoprecipitation experiments (Fig. 4F,G).In vitro pull-down experiments using purified tagged recombinant proteins supported a direct interaction between Prkd2 and Bcl6 (Fig. 4H). We next examined the phosphorylation state of Bcl6 in the presence of Prkd2 in HEK293T cells using phosphate-affinity (Phos-tag) PAGE, in which the phosphorylated form of Bcl6 can be separated as a slower migrating band from non-phosphorylated Bcl6. A phosphorylated form of Bcl6 was observed when it was co-expressed with wild-type Prkd2, but not when it was co-expressed with a kinase-inactive form of Prkd2 lacking the kinase domain (Fig. 4I). Furthermore, kinase-inactive Prkd2 failed to suppress nuclear translocation of Bcl6 (Fig. 4J). These data suggest that an interaction between Bcl6 and Prkd2 leads to Bcl6 phosphorylation, either directly by Prkd2 or via a Prkd2 kinase-dependent event, thereby limiting Bcl6 access to the nucleus.

Collectively, our data indicate that Prkd2 deficiency derestricts Bcl6 nuclear translocation in CD4⁺ T cells resulting in excessive cell autonomous TFH development, and consequently leading to increased GC formation and activation/proliferation of B cells. These TFH, by virtue of their increased numbers and their increased propensity to produce IL-4 when stimulated, are likely responsible for excessive IL-4 levels and ultimately excessive IgE, IgG1, and IgA production inPrkd2^−/− mice.

Bcl6 downregulates Prkd2 in CD4⁺ T cells

TFH expand in response to antigenic challenge, for example during infection, and we hypothesized that Prkd2 levels or activity may be dynamically regulated to permit full expansion of TFH during an immune response. As expected to support such TFH development, both frequencies of Bcl6-expressing CD4⁺ T cells and Bcl6 MFI in CD4⁺ T cells were increased in response to immunization of wild-type mice with OVA/alum (Fig. 5A-C). At the same time (6 days after immunization), Prkd2 expression was decreased in CD4⁺ T cells from immunized mice, both at the protein and mRNA levels (Fig. 5C). In addition, we found that splenic TFH isolated from wild-type mice displayed low levels of Prkd2 expression concomitant with high levels of Bcl6 expression relative to non-TFH (Fig. 5D,Fig. S11). Based on these data and evidence from transcription profiling experiments thatPrkd2 is a target of Bcl6 (35), we hypothesized that Bcl6 repressesPrkd2 expression in CD4⁺ T cells in response to immunization. Consistent with this hypothesis, neither Prkd2 mRNA nor protein levels were downregulated in CD4⁺ T cells isolated fromPrkd2^+/+Bcl6^fl/flCD4-Cre⁺ mice challenged with OVA/alum (Fig. 5E). These data suggest that Bcl6 represses Prkd2 expression in CD4⁺ T cells following immunization to promote the efficient expansion of TFH.

We were interested in the molecular signals that trigger the Bcl6-mediated downregulation ofPrkd2 in CD4⁺ T cells after immunization. Among various cytokines applied to naïve CD4⁺ T cells, we found that IL-12, which is known to promote Th1 differentiation (36), suppressed Prkd2 protein and mRNA expression in wild-type CD4⁺ T cells (Fig. S12A). Previous studies have demonstrated that IL-12 upregulates Bcl6 in CD4⁺ T cells (37,38); consistent with this, Bcl6 expression was enhanced in naïve CD4⁺ T cells culturedin vitro with IL-12 (Fig. S12B). We therefore hypothesized that IL-12-driven Bcl6 expression downregulatesPrkd2 in CD4⁺ T cells. As expected, both frequencies of Bcl6-expressing CD4⁺ T cells and Bcl6 MFI in total CD4⁺ T cells were impaired inIl12a^−/− mice (Fig. S12C,D). Bcl6 MFI was also lower in TFH fromIl12a^−/− mice than in TFH fromIl12a^+/+ mice (Fig. S12E). Consistently,Il12a^−/− mice exhibited reduced frequencies of TFH and GC B cells compared to wild-type mice (Fig. S12F,G). Moreover, IgE production was diminished inIl12a^−/− mice compared to wild-type mice (Fig. S12H). However, we found that Prkd2 expression decreased in CD4⁺ T cells in response to immunization ofIl12a^−/− mice with OVA/alum, possibly because of remaining Bcl6 expression inIl12a^−/− CD4⁺ T cells (Fig. S12I). These data suggest the importance of other factor(s), either alone or in addition to IL-12, for the Bcl6-mediated downregulation of Prkd2 in CD4⁺ T cells that leads to TFH development following immunization.

Older Prkd2^−/− mice develop anti-DNA antibodies

The consequence of increases in TFH numbers and/or hypergammaglobulinemia is often autoimmunity (14).PRKD2 has been identified in genome-wide association studies as a candidate gene in a risk region for primary sclerosing cholangitis (54,55), an autoimmune disease characterized by inflammation of the bile ducts in the liver. PSC patients produce elevated levels of anti-nuclear antibodies (56). In support of these reports, we detected higher titers of anti-double-stranded DNA (dsDNA) antibodies that developed with increasing age inPrkd2^−/− mice compared to wild-type mice (Fig. 6A-D). In addition, lymph nodes were enlarged inPrkd2^−/− mice, which is another sign of autoimmune disease (Fig. S14A). Bronchus-associated lymphoid tissues (BALT) were spontaneously formed inPrkd2^−/− lungs (Fig. S14B,C). Although the function of BALT is incompletely understood, BALTs are frequently observed in patients with various autoimmune disorders or allergies (Fig. S14D,E). The data presented in this paper provide a plausible mechanistic explanation for these findings, namely that excessive TFH may drive autoantibody production in human and mice with deficiencies ofPRKD2.

Mice

Eight- to ten-week old pure C57BL/6J background males purchased from The Jackson Laboratory were mutagenized with N-ethyl-N-nitrosourea (ENU) as described previously (20). Mutagenized G0 males were bred to C57BL/6J females, and the resulting G1 males were crossed to C57BL/6J females to produce G2 mice. G2 females were backcrossed to their G1 sires to yield G3 mice, which were screened for phenotypes. Whole-exome sequencing and mapping were performed as described (20). C57BL/6.SJL (CD45.1; #002014) (57),Rag2^−/− (#008449) (58),Tcra^−/− (#002116) (59),muMT (#002288) (60),Bcl6^fl/fl (#023727) (61),Il12a^−/− (#002692) (62),NZB/BINJ (#000684) (63),NZW/LacJ, (#001058) (64) andCD4-Cre transgenic mice (#017336) (65) were purchased from The Jackson Laboratory.4-get reporter mice were purchased from genOway.Prkd2^−/−/4-get,Prkd2^−/−Tcra^−/−,Prkd2^−/−Bcl6^fl/flCD4-Cre⁺, and NZB/NZW F1 hybrid mice were generated by intercrossing mouse strains. Mice were housed in specific pathogen–free conditions at the University of Texas Southwestern Medical Center and all experimental procedures were performed in accordance with institutionally approved protocols.

Generation of Prkd2^W807R and Prkd2 null mice using CRISPR/Cas9 system

Female C57BL/6J mice were superovulated by injection of 6.5 U pregnant mare serum gonadotropin (PMSG; Millipore), followed by injection of 6.5 U human chorionic gonadotropin (hCG; Sigma-Aldrich) 48 hours later. The superovulated mice were subsequently mated overnight with C57BL/6J male mice. The following day, fertilized eggs were collected from the oviducts andin vitro–transcribed Cas9 mRNA (50 ng/μl) and Prkd2 small base-pairing guide RNA (50 ng/μl; Prkd2: 5’- TCTCAGCCACCCATGGTTAC -3’) was injected into the cytoplasm or pronucleus of the embryos. For generation of thePrkd2^W807R mutation that recapitulated the original Purnama allele, the following homology directed repair template was also injected: atcaacaacctgttgcaggtgaagatgcgcaagcgctacagcgtggacaagtctctcagccacccaAGGttacaAgtgacgtagggagggggcctagaggaggcggcagagctaggtctctaattggctgggtgagtggg (affected codon TGG>AGG, and a silent mutation caG>caA designed to remove the sgRNA target site are indicated with uppercase font). The injected embryos were cultured in M16 medium (Sigma-Aldrich) at 37°C in 5% CO₂. For the production of mutant mice, two-cell stage embryos were transferred into the ampulla of the oviduct (10–20 embryos per oviduct) of pseudo-pregnant Hsd:ICR (CD-1) female mice (Harlan Laboratories). Sequencing showed that the null allele contained a 7 bp deletion of sequence TACAGGT (the final 5 bp at the 3’ end of exon 17 and the first 2 bp at the 5’ end of intron 17); no Prkd2 protein expression was detected in T cells fromPrkd2^−/− mice (Fig. 4D).

Immunization and ELISA analysis of serum immunoglobulins

Mice were immunized i.p. with aluminum hydroxide absorbed ovalbumin (OVA/alum; 200 μg; Invivogen) or papain (500 μg; Sigma). Unless otherwise noted, sera were harvested before immunization on day 0 and on day 10 post-immunization. For G3 mice, sera were harvested on day 14 after immunization. For analysis of T cell or B cell populations, unimmunized mice or immunized mice on the indicated day after immunization were sacrificed and spleen or lymph node cells were isolated for flow cytometric analysis.

Blood was collected in MiniCollect Tubes (Mercedes Medical) and centrifuged at 1,500 ×g to separate the serum. To measure serum antibody isotype levels (IgE, IgA, IgM, IgG1, IgG2a, IgG2b, IgG2c, IgG3), freshly isolated serum was subjected to sandwich Enzyme-Linked ImmunoSorbent Assay (ELISA) analysis. ELISA kits for IgE, IgA, IgM, IgG2c, and IgG3 were purchased from Invitrogen. ELISA kits for IgG1, IgG2a, and IgG2b were purchased from Bethyl. ELISA analysis was performed according to the manufacturer's instructions. For ELISA analysis of antigen-specific IgE, or double-stranded DNA-specific IgG, Nunc MaxiSorp flat-bottom 96-well microplates (Thermo Fisher Scientific) were coated with 10-20 μg/mL OVA, 10 μg/mL Papain, or 50 μg/mL ultra pure calf thymus DNA solution (Thermo Fisher Scientific) at 4°C overnight. Plates were washed four times with washing buffer (0.05% (v/v) Tween-20 in PBS) using a BioTek microplate washer then blocked with 1% (v/v) BSA in PBS for 1 hour at room temperature. Serum samples were added to the prepared ELISA plates. After 2 hours incubation, the plates were washed four times with washing buffer and then incubated with HRP-conjugated goat anti-mouse IgE, HRP-conjugated goat anti-mouse IgG, HRP-conjugated goat anti-mouse IgA, or HRP-conjugated goat anti-mouse IgM (SouthernBiotech) for 30-45 minutes at room temperature. Plates were washed four times with washing buffer then developed with SureBlue TMB Microwell Peroxidase Substrate and TMB Stop Solution (KPL). Absorbance was measured at 450 nm on a Synergy Neo2 Plate Reader (BioTek).

Bone marrow transplantation and naïve CD4⁺ T cell adoptive transfer

To generate bone marrow chimeras,Rag2^−/− mice were lethally irradiated with 13 Gy via gamma radiation (X-RAD 320, Precision X-ray Inc.). The mice were given intravenous injection of 4 × 106 bone marrow cells derived from the tibia and femurs of the respective donors. For 2 weeks post-engraftment, mice were maintained on antibiotics. To generate mixed bone marrow chimeras the same procedure was performed, except that equal numbers of bone marrow cells from each of the donor genotypes were mixed (2 × 10⁶ cells from each genotype) and transferred into irradiatedRag2^−/− mice intravenously. 8-12 weeks after bone marrow engraftment, the chimeras were euthanized to assess TFH and GC B cell development in spleen by flow cytometry, or antibody responses to OVA/alum immunization. Chimerism was assessed using congenic CD45 markers. Splenic T cells and B cells were isolated fromRag2^−/− mice engrafted withmuMT andTcra^−/− bone marrows. Cells were lysed and subjected to immunoblot analysis to examine their Prkd2 level.

To generate mixed naïve CD4⁺ T cell chimeras,Rag2^−/− mice were lethally irradiated with 13 Gy via gamma radiation as described above. Splenic naïve CD4⁺ T cells were isolated. Equal numbers ofPrkd2^+/+ andPrkd2^−/− naïve CD4⁺ T cells were mixed (1.5 × 10⁶ cells from each genotype) and transferred into irradiatedRag2^−/− mice intravenously. 7 days after naïve CD4⁺ T cell transfer, the chimeras were euthanized to assess TFH development in spleen by flow cytometry. Chimerism was assessed using congenic CD45 markers.

Flow cytometry and cell sorting

The following antibodies were used: FITC CD3ε (145-2C11), BV786 CD4 (RM4-5), BV421 CD5 (53-7.3), BV 510 CD8α (53-6.7), BV711 CD11c (HL3), PE CD19 (1D3), Alexa Fluor 700 CD19 (1D3), BUV395 CD19 (1D3), BV421 CD23 (B3B4), APC CD43 (S7), PE-CF594 CD44 (IM7), Alexa Fluor 700 CD45R (RA3-6B2), PE-CF594 CD45.2 (104), BV510 CD95 (Jo2), PE Bcl6 (K112-91), PE RORγT (Q31-378), Alexa Fluor 647 Foxp3 (MF23), FITC IgM (Ⅱ/41), PerCP/Cy5.5 IgM (RMM-1), Alexa Fluor 647 T and B cell activation antigen (GL7), PE F4/80 (BM8.1), PE-Cy7 CD62L (MEL-14), and PE Mouse IgG1, κ isotype control were purchased from BD Bioscience; APC/Cy7 CD3e (145-2C11), BV605 CD11b (M1/70), PerCP/Cy5.5 CD21/CD35 (7E9), PE/Cy7 CD45.1 (A20), FITC CD45.1 (A20), PE CD138 (281-2), APC CXCR5 (L138D7), BV421 PD-1 (29F.1A12), BV421 IL-2 (JES6-5H4), PE IL-4 (11B11), APC/Cy7 IgD (11-26c.2a), and BV650 NK1.1 (PK136) were purchased from Biolegend; PE CD3e (145-2C11) was purchased from Invitrogen. Splenocytes or lymph node cells were incubated with purified anti-mouse CD16/CD32 (2.4G2; TONBO) in PBS containing 5% FCS for blocking on ice. The cells were then stained using a 1:100 dilution of the indicated fluorescence dye-labeled antibodies. Intracellular staining for cytokine was performed after using BD Cytofix/Cytoperm (BD Biosciences) according to the manufacturer’s instructions. Intracellular staining for Bcl6, RORγT, and Foxp3 was performed after using Mouse Foxp3 Buffer Set (BD Biosciences) according to the manufacturer’s instructions. The stained cells were analyzed using a BD LSRFortessa cell analyzer (BD Biosciences).

Splenic CD4⁺ T cells were stained with anti-CD3, anti-CD4, anti-CXCR5, and anti-PD-1 as described above. CXCR5^lowPD-1^low or CXCR5^highPD-1^high CD4⁺ T cells were then sorted using BD FACSAria IIU (BD Biosciences).

Isolation of primary immune cells

Splenic immune cells were isolated with Naïve CD4⁺ T Cell Isolation Kit (Miltenyi Biotech), CD4⁺ T Cell Isolation Kit (Miltenyi Biotech), CD8a⁺ T Cell Isolation Kit (Miltenyi Biotech), Pan B Cell Isolation Kit (Miltenyi Biotech), CD11b MicroBeads (Miltenyi Biotech), CD11c MicroBeads (Miltenyi Biotech), CD49b (DX5) MicroBeads (Miltenyi Biotech), Neutrophil Isolation Kit (Miltenyi Biotech) according to manufacturer’s instructions. Purities were over 95% in all experiments tested by flow cytometry.

T cell stimulation in vitro

Splenic CD4⁺ T cells or naïve CD4⁺ T cells were isolated as described above. Cells were stimulated in vitro with Leukocyte Activation Cocktail with BD GolgiPlug (PMA/ionomycin) (BD Biosciences), Dynabeads Mouse T-Activator CD3/CD28 (Thermo Fisher Scientific), or plate bound anti-TCRβ (10 μg/ml) and anti-CD28 (1 μg/ml) for 12 to 72 hours according to the manufacturer’s instructions. IL-4 level in supernatant was analyzed by mouse IL-4 ELISA kit (Thermo Fisher Scientific).

Differentiation of Th subsets in vitro

Naïve CD4⁺ T cells isolated from spleens were cultured in 24 well plates coated with anti-TCRβ (10μg/ml) and anti-CD28 (1 μg/ml). For Th0 condition, cells were incubated in the presence of IL-2 (20 ng/ml). For Th1 condition, cells were incubated with IL-2 (20 ng/ml), IL-12 (10 ng/ml), and with or without anti-IL-4 (5 μg/ml). For Th2 condition, cells were incubated with IL-2 (20 ng/ml), IL-4 (10 ng/ml), and anti-IFN-γ (5 μg/ml). Medium was refreshed every two days and cells were harvested on day 4-6. In vitro differentiated Th2 cells were used on day 6 for stimulation with PMA/ionomycin, except that stimulation was for 6 hours. Cells were stained with PE IL-4 (11B11; Biolegend) and analyzed by flow cytometry.

Cell culture, plasmids, and transfections

HEK293T cells were cultured at 37°C in 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM, Life Technology) containing 10% (vol/vol) FBS and 1% antibiotics (Life Technology). Full-length human or mouse Bcl6, full-length human Prkd2, or human kinase domain-deleted Prkd2 were cloned into pCMV HA_N vector (Addgene). These plasmids were deposited in Addgene along with maps and sequences (Addgene ID; 137850, 137851, 138410, and 138411). pcDNA3.1(+)-N-DYK containing full-length human or mouse Prkd2 or full-length human Bcl6 were purchased from Genescript. pcDNA3.1(+)-N-DYK containing human kinase domain-deleted Prkd2 were purchased from Genescript. Tryptophan 807 within mouse Prkd2 was replaced into Arginine by using Q5 site-directed mutagenesis kit (New England Biolabs) according to manufacturer’s instruction and cloned into pcDNA3.1(+)-N-DYK. Transfection of plasmids was carried out using Lipofectamine 2000 (Life Technologies) according to the manufacturer’s instructions. In some experiments, cells were treated with increasing concentration (0, 1, 5, 10 μM) of MG132 (Millipore) for 12 hours.

Immunoprecipitation and pulldown assay

HEK293T cells were harvested between 36 and 48 hours post-transfection and lysed with NP-40 lysis buffer (25 mM Tris–HCl pH 8.0, 150 mM NaCl, 1% (vol/vol) Nonidet P-40, and protease inhibitors) for 45 min at 4°C. Immunoprecipitation was performed using anti-FLAG M2 magnetic beads (Sigma). Whole cell lysates and immunoprecipitates were immunoblotted using FLAG M2 antibody (Sigma) and HA antibody (Sigma).

FLAG-tagged human Bcl6 was transfected into HEK293T cells. 36 hours after transfection, FLAG-tagged human Bcl6 was immunoprecipitated using anti-FLAG M2 magnetic beads. Beads were then incubated with recombinant human GST-tagged Prkd2 (Thermo Fisher Scientific) for 2 hours at 4°C. Beads were washed three times with NP-40 lysis buffer. Beads were incubated with 3× FLAG peptide (Sigma) for 30 minutes and eluted proteins were subjected to immunoblot analysis.

Immunoblot analysis

Cells were lysed with either NP-40 lysis buffer (25 mM Tris–HCl pH 8.0, 150 mM NaCl, 1% (vol/vol) Nonidet P-40, and protease inhibitors) or SDS lysis buffer (50 mM Tris-HCl pH 6.8, 1% (vol/vol) SDS, and 10 % (vol/vol) glycerol). Protein concentrations were normalized by performing BCA protein assay (Pierce). Lysates were subjected to gel electrophoresis using NuPAGE 4-12% Bis-Tris gels (Life Technologies). After electrophoresis, proteins were transferred to nitrocellulose membranes (Bio-Rad). For Phos-tag PAGE, cell lysates were subjected to gel electrophoresis using precast polyacrylamide gels containing Phos-tag (SuperSep Phos-tag; Wako). After electrophoresis, proteins were transferred to polyvinylidene difluoride (PVDF) membranes using a Mini Trans-Blot Cell (BioRad).

Membranes were blocked with Tris-buffered saline containing 0.1% (v/v) Tween-20 and 5% (w/v) non-fat dry milk (LabScientific) for 1-2 hours at room temperature. Proteins were detected by incubating membranes with the following antibodies: anti-Prkd2 (EPR1495Y) (Abcam); anti-α-Tubulin (DM1A), anti-β-Actin (13E5), anti-GATA3 (D13C9), anti-Histone H3 (3H1), anti-STAT6 (D3H4), HRP-conjugated anti-Rabbit IgG, and HRP-conjugated anti-Mouse IgG (Cell Signaling Technology); anti-Bcl6 (K112-91) (BD Biosciences). The chemiluminescence signal was developed using Super Signal West Pico Chemiluminescent Substrate or Super Signal West Dura Extended Duration Substrate kit (Thermo Fisher Scientific) and detected with a G:Box Chemi XX6 system (Syngene). Signal intensity of bands were quantified by ImageJ.

Histology

Spleens were harvested, embedded in Tissue-Tek O.C.T. (Sakura; Finetek) and frozen at −80°C. Cryostat sections (8 μm in thickness) were prepared, air-dried, and fixed in ice-cold acetone for 5 minutes. Sections were blocked with 3% BSA-containing PBS (wt/vol) for 30 minutes. Sections were further blocked using Avidin/Biotin Blocking Kit (Vector labs) and stained with PNA-biotin (Vector) and FITC anti-IgD (Biolegend). Biotinylated antibodies were detected with streptavidin, Alexa Fluor 546 conjugated antibody (Thermo). Sections were mounted with ProLong Gold antifade reagent (Invitrogen). Images were captured with a Zeiss AxioImager M1 microscope. Lungs were fixed with 4% Paraformaldehyde (Santa Cruz) for 24 hours at 4°C and then embedded in paraffin. Sections were stained with hematoxylin and eosin (HE). Images were captured with a Leica Application Suite V4 (Leica).

Fractionation analysis

Subcellular compartments were fractionated using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific) according to the manufacturer’s instructions.

Quantitative RT-PCR

Total RNA was isolated using the Quick-RNA MiniPrep Kit (Zymo Research). cDNA fragments were reverse transcribed using SuperScript III First-Strand Synthesis SuperMix for qRT-PCR (Thermo Fisher Scientific) according to the manufacturer’s instructions. qPCR was performed using TaqMan probes forPrkd2 (Thermo Fisher Scientific; Mm00626821_m1) orGAPDH (Thermo Fisher Scientific; Mm99999915_g1). Fluorescence from the TaqMan probe for thePrkd2 orGAPDH RNA gene was detected using a StepOnePlus Real-Time PCR System (Applied Biosystems). Relative standard curve method was used for the quantification.

Statistical Analysis

The statistical significance of differences between groups was determined using the indicated statistical tests and GraphPad Prism software. Data are expressed as means ± SD. Differences withP values ≥ 0.05 were considered to be not significant (NS). All differences withP values < 0.05 were considered significant.P values are denoted by *P < 0.05, **P < 0.01, ***P < 0.001.

• 1.Crotty S, T Follicular Helper Cell Biology: A Decade of Discovery and Diseases. Immunity50, 1132–1148 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 2.Roco JAet al. , Class-Switch Recombination Occurs Infrequently in Germinal Centers. Immunity51, 337–350.e337 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 3.Kawabe Tet al. , The immune responses in CD40-deficient mice: impaired immunoglobulin class switching and germinal center formation. Immunity1, 167–178 (1994). [DOI] [PubMed] [Google Scholar]
• 4.Fahey LMet al. , Viral persistence redirects CD4 T cell differentiation toward T follicular helper cells. J Exp Med208, 987–999 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 5.Cubas RAet al. , Inadequate T follicular cell help impairs B cell immunity during HIV infection. Nat Med19, 494–499 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 6.Bentebibel SEet al. , Induction of ICOS+CXCR3+CXCR5+ TH cells correlates with antibody responses to influenza vaccination. Sci Transl Med5, 176ra132 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 7.Duan Zet al. , Genetic polymorphisms of CXCR5 and CXCL13 are associated with non-responsiveness to the hepatitis B vaccine. Vaccine32, 5316–5322 (2014). [DOI] [PubMed] [Google Scholar]
• 8.Simpson Net al. , Expansion of circulating T cells resembling follicular helper T cells is a fixed phenotype that identifies a subset of severe systemic lupus erythematosus. Arthritis Rheum62, 234–244 (2010). [DOI] [PubMed] [Google Scholar]
• 9.Vinuesa CGet al. , A RING-type ubiquitin ligase family member required to repress follicular helper T cells and autoimmunity. Nature435, 452–458 (2005). [DOI] [PubMed] [Google Scholar]
• 10.Harada Yet al. , The 3' enhancer CNS2 is a critical regulator of interleukin-4-mediated humoral immunity in follicular helper T cells. Immunity36, 188–200 (2012). [DOI] [PubMed] [Google Scholar]
• 11.Kobayashi T, Iijima K, Dent AL, Kita H, Follicular helper T cells mediate IgE antibody response to airborne allergens. J Allergy Clin Immunol139, 300–313 e307 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 12.Yao Y, Wang ZC, Yu D, Liu Z, Role of allergen-specific T-follicular helper cells in immunotherapy. Curr Opin Allergy Clin Immunol18, 495–501 (2018). [DOI] [PubMed] [Google Scholar]
• 13.Crotty S, T follicular helper cell differentiation, function, and roles in disease. Immunity41, 529–542 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 14.Vinuesa CG, Linterman MA, Yu D, MacLennan IC, Follicular Helper T Cells. Annu Rev Immunol34, 335–368 (2016). [DOI] [PubMed] [Google Scholar]
• 15.Roy A, Ye J, Deng F, Wang QJ, Protein kinase D signaling in cancer: A friend or foe?Biochimica et biophysica acta. Reviews on cancer1868, 283–294 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 16.Azoitei N, Cobbaut M, Becher A, Van Lint J, Seufferlein T, Protein kinase D2: a versatile player in cancer biology. Oncogene37, 1263–1278 (2018). [DOI] [PubMed] [Google Scholar]
• 17.Matthews SAet al. , Unique functions for protein kinase D1 and protein kinase D2 in mammalian cells. Biochem J432, 153–163 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Navarro MNet al. , Protein kinase D2 has a restricted but critical role in T-cell antigen receptor signalling in mature T-cells. Biochem J442, 649–659 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 19.Navarro MN, Feijoo-Carnero C, Arandilla AG, Trost M, Cantrell DA, Protein kinase D2 is a digital amplifier of T cell receptor-stimulated diacylglycerol signaling in naive CD8(+) T cells. Science signaling7, ra99 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 20.Wang Tet al. , Real-time resolution of point mutations that cause phenovariance in mice. Proc Natl Acad Sci U S A112, E440–449 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 21.Kim RJet al. , IL-4-induced AID expression and its relevance to IgA class switch recombination. Biochem Biophys Res Commun361, 398–403 (2007). [DOI] [PubMed] [Google Scholar]
• 22.Kubo M, T follicular helper and TH2 cells in allergic responses. Allergol Int66, 377–381 (2017). [DOI] [PubMed] [Google Scholar]
• 23.Mohrs K, Wakil AE, Killeen N, Locksley RM, Mohrs M, A two-step process for cytokine production revealed by IL-4 dual-reporter mice. Immunity23, 419–429 (2005). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 24.Walker JA, McKenzie ANJ, TH2 cell development and function. Nat Rev Immunol18, 121–133 (2018). [DOI] [PubMed] [Google Scholar]
• 25.Nurieva RIet al. , Bcl6 mediates the development of T follicular helper cells. Science325, 1001–1005 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 26.Liu Xet al. , Bcl6 expression specifies the T follicular helper cell program in vivo. J Exp Med209, 1841–1852, S1841-1824 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 27.Yu Det al. , The transcriptional repressor Bcl-6 directs T follicular helper cell lineage commitment. Immunity31, 457–468 (2009). [DOI] [PubMed] [Google Scholar]
• 28.Johnston RJet al. , Bcl6 and Blimp-1 are reciprocal and antagonistic regulators of T follicular helper cell differentiation. Science325, 1006–1010 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 29.Hatzi Ket al. , BCL6 orchestrates Tfh cell differentiation via multiple distinct mechanisms. J Exp Med212, 539–553 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 30.Weinstein JSet al. , TFH cells progressively differentiate to regulate the germinal center response. Nat Immunol17, 1197–1205 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 31.Ballesteros-Tato Aet al. , Interleukin-2 inhibits germinal center formation by limiting T follicular helper cell differentiation. Immunity36, 847–856 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 32.DiToro Det al. , Differential IL-2 expression defines developmental fates of follicular versus nonfollicular helper T cells. Science361, (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 33.Kim CH, Regulation of FoxP3 regulatory T cells and Th17 cells by retinoids. Clin Dev Immunol2008, 416910 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 34.Basso K, Dalla-Favera R, BCL6: master regulator of the germinal center reaction and key oncogene in B cell lymphomagenesis. Adv Immunol105, 193–210 (2010). [DOI] [PubMed] [Google Scholar]
• 35.Liu Xet al. , Genome-wide Analysis Identifies Bcl6-Controlled Regulatory Networks during T Follicular Helper Cell Differentiation. Cell Rep14, 1735–1747 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 36.Szabo SJet al. , A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell100, 655–669 (2000). [DOI] [PubMed] [Google Scholar]
• 37.Nakayamada Set al. , Early Th1 cell differentiation is marked by a Tfh cell-like transition. Immunity35, 919–931 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 38.Schmitt Net al. , IL-12 receptor beta1 deficiency alters in vivo T follicular helper cell response in humans. Blood121, 3375–3385 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 39.Kim TH, Jung SH, Cho KH, Interlinked mutual inhibitory positive feedbacks induce robust cellular memory effects. FEBS Lett581, 4899–4904 (2007). [DOI] [PubMed] [Google Scholar]
• 40.Jacob F, Monod J, Genetic regulatory mechanisms in the synthesis of proteins. J Mol Biol3, 318–356 (1961). [DOI] [PubMed] [Google Scholar]
• 41.Oppenheim AB, Kobiler O, Stavans J, Court DL, Adhya S, Switches in bacteriophage lambda development. Annu Rev Genet39, 409–429 (2005). [DOI] [PubMed] [Google Scholar]
• 42.Ptashne M, Lambda's switch: lessons from a module swap. Curr Biol16, R459–462 (2006). [DOI] [PubMed] [Google Scholar]
• 43.Yuan JS, Kousis PC, Suliman S, Visan I, Guidos CJ, Functions of notch signaling in the immune system: consensus and controversies. Annu Rev Immunol28, 343–365 (2010). [DOI] [PubMed] [Google Scholar]
• 44.Perry JA, Kornbluth S, Cdc25 and Wee1: analogous opposites?Cell Div2, 12 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 45.Chen KCet al. , Kinetic analysis of a molecular model of the budding yeast cell cycle. Mol Biol Cell11, 369–391 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 46.Weinmann AS, Regulatory mechanisms that control T-follicular helper and T-helper 1 cell flexibility. Immunol Cell Biol92, 34–39 (2014). [DOI] [PubMed] [Google Scholar]
• 47.Schmitt N, Liu Y, Bentebibel SE, Ueno H, Molecular Mechanisms Regulating T Helper 1 versus T Follicular Helper Cell Differentiation in Humans. Cell Rep16, 1082–1095 (2016). [DOI] [PubMed] [Google Scholar]
• 48.McKee ASet al. , Alum induces innate immune responses through macrophage and mast cell sensors, but these sensors are not required for alum to act as an adjuvant for specific immunity. J Immunol183, 4403–4414 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 49.Miwa Det al. , Protein kinase D2 and heat shock protein 90 beta are required for BCL6-associated zinc finger protein mRNA stabilization induced by vascular endothelial growth factor-A. Angiogenesis16, 675–688 (2013). [DOI] [PubMed] [Google Scholar]
• 50.Ohnuki Het al. , BAZF, a novel component of cullin3-based E3 ligase complex, mediates VEGFR and Notch cross-signaling in angiogenesis. Blood119, 2688–2698 (2012). [DOI] [PubMed] [Google Scholar]
• 51.Okabe Set al. , BAZF, a novel Bcl6 homolog, functions as a transcriptional repressor. Molecular and cellular biology18, 4235–4244 (1998). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 52.Takenaga Met al. , Bcl6-dependent transcriptional repression by BAZF. Biochem Biophys Res Commun303, 600–608 (2003). [DOI] [PubMed] [Google Scholar]
• 53.Takamori Met al. , BAZF is required for activation of naive CD4 T cells by TCR triggering. International immunology16, 1439–1449 (2004). [DOI] [PubMed] [Google Scholar]
• 54.Ji SGet al. , Genome-wide association study of primary sclerosing cholangitis identifies new risk loci and quantifies the genetic relationship with inflammatory bowel disease. Nat Genet49, 269–273 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 55.Karlsen TH, Folseraas T, Thorburn D, Vesterhus M, Primary sclerosing cholangitis - a comprehensive review. J Hepatol67, 1298–1323 (2017). [DOI] [PubMed] [Google Scholar]
• 56.Hov JR, Boberg KM, Karlsen TH, Autoantibodies in primary sclerosing cholangitis. World J Gastroenterol14, 3781–3791 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 57.Janowska-Wieczorek Aet al. , Platelet-derived microparticles bind to hematopoietic stem/progenitor cells and enhance their engraftment. Blood98, 3143–3149 (2001). [DOI] [PubMed] [Google Scholar]
• 58.Hao Z, Rajewsky K, Homeostasis of peripheral B cells in the absence of B cell influx from the bone marrow. J Exp Med194, 1151–1164 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 59.Mombaerts Pet al. , Mutations in T-cell antigen receptor genes alpha and beta block thymocyte development at different stages. Nature360, 225–231 (1992). [DOI] [PubMed] [Google Scholar]
• 60.Kitamura D, Roes J, Kuhn R, Rajewsky K, A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin mu chain gene. Nature350, 423–426 (1991). [DOI] [PubMed] [Google Scholar]
• 61.Hollister Ket al. , Insights into the role of Bcl6 in follicular Th cells using a new conditional mutant mouse model. J Immunol191, 3705–3711 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 62.Mattner Fet al. , Genetically resistant mice lacking interleukin-12 are susceptible to infection with Leishmania major and mount a polarized Th2 cell response. Eur J Immunol26, 1553–1559 (1996). [DOI] [PubMed] [Google Scholar]
• 63.Paigen B, Morrow A, Brandon C, Mitchell D, Holmes P, Variation in susceptibility to atherosclerosis among inbred strains of mice. Atherosclerosis57, 65–73 (1985). [DOI] [PubMed] [Google Scholar]
• 64.Depis F, Kwon HK, Mathis D, Benoist C, Unstable FoxP3+ T regulatory cells in NZW mice. Proc Natl Acad Sci U S A113, 1345–1350 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 65.Sawada S, Scarborough JD, Killeen N, Littman DR, A lineage-specific transcriptional silencer regulates CD4 gene expression during T lymphocyte development. Cell77, 917–929 (1994). [DOI] [PubMed] [Google Scholar]

Supplementary Materials