Skip to main content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Immunity. Author manuscript; available in PMC 2014 Oct 17.
Published in final edited form as:
Immunity. 2013 Oct 17; 39(4): 10.1016/j.immuni.2013.08.032.
Published online 2013 Oct 10. doi: 10.1016/j.immuni.2013.08.032
PMCID: PMC3808842
NIHMSID: NIHMS521661
PMID: 24120360

Epigenetic Modifications induced by the Transcriptional Repressor Blimp-1 regulate CD8+ T cell Memory Progression during Acute Virus Infection

Hyun Mu Shin,#1 Varun Kapoor,#1 Tianxia Guan,2 Susan M. Kaech,2 Raymond M. Welsh,1 and Leslie J. Berg1,*
Author information Copyright and License information Disclaimer

Associated Data

Supplementary Materials

SUMMARY

The transcription factor Blimp-1 regulates the overall accumulation of virus-specific CD8+ T cells during acute viral infections. We found that increased proliferation and survival of Blimp-1-deficient CD8+ T cells resulted from sustained expression of CD25 and CD27 and persistent cytokine responsiveness. Silencing of these genes reduced the Blimp-1-deficient CD8+ T cell response. Genome-wide Chromatin immunoprecipitation (ChIP) sequencing analysis identified Il2ra and Cd27 genes as direct targets of Blimp-1. At the peak of the anti-viral response, but not earlier, Blimp-1 recruited the histone modifying enzymes G9a and HDAC2 to the Il2ra and Cd27 loci, thereby repressing expression of these genes. In the absence of Blimp-1, the Il2ra and Cd27 genes exhibited enhanced histone H3-acetylation and reduced histone H3K9-trimethylation. These data elucidate a central mechanism by which Blimp-1 acts as an epigenetic regulator, enhancing the numbers of short-lived effector cells while suppressing the development of memory precursor CD8+ T cells.

INTRODUCTION

In response to a virus infection, CD8+ T cells proliferate and differentiate into effector cells that eradicate the pathogen. Upon viral clearance, homeostasis is restored and a stable population of virus-specific memory CD8+ T cells remains to protect against re-infection by that virus. The quality and quantity of the CD8+ T cell response during the initial phase of the primary response governs the frequency and function of long-lived CD8+ memory T cells (Obar and Lefrancois, 2010). For an optimal response, CD8+ T cells require at least three signals. These include antigenic stimulation through the T cell receptor (TCR), co-stimulation through receptors such as CD28, CD40, 4-1BB, CD27, ICOS and/or OX40, and cytokine stimulation via inflammatory cytokines (Duttagupta et al., 2009). The initial TCR engagement triggers the up-regulation of co-stimulatory molecules and cytokine receptors, which are critical for the clonal expansion and survival of the responding CD8+ T cells (Duttagupta et al., 2009). However, this population of CD8+ T cells is heterogeneous; the majority of effector cells die, while a small population survive and become memory cells (Obar and Lefrancois, 2010). Transcriptional profiling of effector and memory CD8+ T cells in both acute and chronic virus infection models has recently provided insight into the distinct gene expression programs characterizing distinct cell subsets (Doering et al., 2012). Nonetheless, the precise mechanisms by which these transcriptional programs are established and maintained during CD8+ T cell differentiation remain largely unknown.

During the past decade, numerous studies have shown that interleukin-2 (IL-2) plays an important role in regulating CD8+ T cell responses during the different stages of viral infection (Boyman and Sprent, 2012). In vivo administration of IL-2 during early stages of the viral response is detrimental to the survival of CD8+ T cells; however, IL-2 therapy during the contraction and memory stages of the response promotes CD8+ T cell survival (Blattman et al., 2003). Additional studies have indicated that both primary and secondary CD8+ T cell responses are impaired in the absence of IL-2 receptor signaling (Mitchell et al., 2010; Williams et al., 2006). CD25, a subunit of the IL-2 receptor is up-regulated by IL-2 in conjunction with TCR stimulation (Boyman and Sprent, 2012), and at early stages of the response to lymphocytic choriomeningitis virus (LCMV) infection, CD25 expression promotes the development of terminally-differentiated effector CD8+ T cells (Kalia et al., 2010). Nonetheless, the mechanism by which CD25 expression on CD8+ T cells is regulated over the course of the immune response has not been described.

Members of the tumor necrosis factor (TNF) superfamily also contribute to CD8+ T cell survival in vivo. CD27, a co-stimulatory molecule in the TNF receptor family binds to CD70 and promotes CD8+ T cell proliferation (Duttagupta et al., 2009). CD27-deficient mice have decreased primary and secondary CD8+ T cell responses to influenza, and CD27 has been shown to be important for generation of long term immunity to this infection (Hendriks et al., 2000). CD27 signals also promote survival of activated CD8+ T cells by preventing Fas-dependent apoptosis and by inducing the pro-survival factors Bcl2 and BclxL (Dolfi et al., 2008; Peperzak et al., 2010a). A recent study shows that CD27-CD70 interactions induce IL-2 production, thereby promoting clonal expansion of primed CD8+ T cells (Peperzak et al., 2010b). CD27 also marks CD8+ T cells with memory potential ((Hendriks et al., 2000); VK and RMW, unpublished observation). However, the mechanism by which CD27 expression on CD8+ T cells is regulated during the immune response is not understood.

Blimp-1 is a transcription factor known to regulate the terminal differentiation of numerous cell types (Martins and Calame, 2008). It drives the differentiation of CD8+ T cells towards the short-lived effector fate, and the absence of Blimp-1 leads to the development of memory precursor CD8+ T cells that are better at producing IL-2 (Kallies et al., 2009; Rutishauser et al., 2009). While Blimp-1 is known to regulate a number of target genes mediating plasma cell differentiation (Martins and Calame, 2008), only a few targets of Blimp-1 in CD8+ T cells have been identified. One recent study shows that Blimp-1 directly regulates expression of the DNA-binding inhibitor Id3, thereby contributing to the development of short-lived effector T cells (Ji et al., 2011). However, more global information on the targets of Blimp-1 in CD8+ T cells is currently lacking.

In previously characterized systems, Blimp-1 has been shown to mediate transcriptional repression by associating with histone modifying enzymes (Ancelin et al., 2006; Gyory et al., 2004; Smith et al., 2011; Su et al., 2009; Yu et al., 2000). As there is differential histone modification between memory and effector CD8+ T cell subsets (Araki et al., 2009), we considered whether Blimp-1 might function in T cells to regulate histone lysine modification. In osteosarcoma and B-cell lymphoma cell lines, Blimp-1 recruits the methyltransferase G9a and histone deacetylase HDAC1 and 2 to target genes, leading to repressive modifications such as histone H3-K9 trimethylation or H3 de-acetylation, respectively (Gyory et al., 2004; Kubicek et al., 2007; Yu et al., 2000). Whether any of these mechanisms are involved in the Blimp-1-mediated regulation of CD8+T cell differentiation is not known.

We show here that CD25 and CD27 expression is dysregulated in Prdm1−/− T cells responding to LCMV infection, that T cells expressing CD25 and CD27 preferentially survive during acute viral infection, that sustained CD25 or CD27 expression by retroviral transduction enhances T cell survival, and that silencing of CD25 and/or CD27 in Blimp-1-deficient CD8+ T cells decreases the magnitude of the virus-specific CD8+ T cell response. CD25 and CD27 were identified as direct targets of Blimp-1 using deep-sequencing analysis of Blimp-1-bound DNA targets in CD8+ T cells (ChIP-Seq). At the peak of CD8+ T cell expansion, but not early in infection, we show that Blimp-1 binds to regulatory regions of the CD25 and CD27 genes and recruits G9a or HDAC2 to promote repressive histone modifications at these loci. These data provide key insights into the mechanism by which Blimp-1 acts as an epigenetic regulator of target genes, thereby dictating the fate of CD8+ effector T cells.

RESULTS

Blimp-1 suppresses cytokine responsiveness of CD8+ T cells at the peak of the anti-viral response

To investigate the role of Blimp-1 in regulating CD8+ T cell responses during a virus infection, we used mice carrying a conditional allele of the Prdm1 gene in which exon 5 is flanked by loxP sites (Ohinata et al., 2005). This line was crossed to Cd4-cre+ transgenic mice, thereby deleting Prdm1 in all αβ T cells, and differs from those used previously to study the function of Blimp-1 in B and T lymphocytes (Martins et al., 2006; Piskurich et al., 2000). Hereafter, we will refer Prdm1flox/floxxCd4-cre+ mice as “Prdm1−/−”, Prdm1flox/+xCd4-cre+ mice as “Prdm1+/−”and CD4-cre+ littermate controls as “WT”. We did not detect any changes in the proportion of lymphocytes in various lymphoid organs (FigS1a), although naïve Prdm1−/− mice have a higher proportion of CD44hi CD4+ and CD8+ T cells (FigS1b), as reported (Kallies et al., 2006; Martins et al., 2006).

Consistent with previous studies (Rutishauser et al., 2009; Shin et al., 2009), there was a marked increase in both the number and proportion of CD8+ T cells in Prdm1−/− mice at days 7 and 14 following LCMV-Armstrong infection (Fig1a,b). CD44hi CD8+ T cells and LCMV-specific CD8+ T cells showed similar increases (Fig1a). Memory-precursor effector CD8+ T cells (MPEC; KLRG1loIL-7Rhi (Joshi et al., 2007)) were also increased in Prdm1−/− mice compared to WT at days 7 and 14 post-infection (Fig1c), consistent with previous data (Rutishauser et al., 2009). Deletion of Prdm1 in activated CD8+ T cells from Prdm1−/− mice was confirmed at day 7 and 14 post LCMV infection (FigS1c). Viral clearance in the spleen was normal in Prdm1−/− mice (FigS1d), indicating that the increased magnitude of the CD8+ T cell response to LCMV in Prdm1−/− mice was not due to impaired viral clearance. We also found that CD44hi CD8+ T cells from LCMV-infected Prdm1−/− mice were less apoptotic than those from WT mice at day 9 post-LCMV infection as shown by decreased TUNEL reactivity (Fig1d), in accord with increased expression of the pro-survival factor Bcl2 at day 7 post-infection (Fig1e). The transcription factor eomesodermin (EOMES) promotes persistence of memory CD8+ T cells (Banerjee et al., 2010). Consistent with this, we found that a higher proportion of virus-specific CD8+ T cells from the Prdm1−/− mice expressed EOMES compared to WT controls at days 7 and 14 post-infection (FigS1e).

An external file that holds a picture, illustration, etc. Object name is nihms-521661-f0001.jpg
Sustained cytokine responsiveness in Prdm1−/− CD8+ T cells following LCMV infection

(a) The total number of CD8+ T cells and frequencies of CD8+ CD44hi, GP33+ CD8+ and NP396+ CD8+ T cells at day 7 post-LCMV infection are shown for WT and Prdm1−/− mice.

(b) Proportion of total CD8+ T cells in LCMV-infected WT and Prdm1−/− mice at days 7 and 14 post-infection, relative to uninfected mice (D0). Histograms are representative of three independent experiments.

(c) NP396 tetramer-binding CD8+ T cells in WT and Prdm1−/− mice were analyzed for IL-7R and KLRG1 expression at days 7 and 14 post-LCMV infection. The numbers indicate the percentage of NP396 tetramer-binding CD8+ T cells in each subset. Data are representative of three independent experiments.

(d) Histograms show representative results of TUNEL staining in WT and Prdm1−/− CD44hi CD8+ T cells at day 9 post-LCMV infection. The graph shows a compilation of data from three independent experiments.

(e) NP396 tetramer-binding CD8+ T cells in WT and Prdm1−/− mice were analyzed for intracellular Bcl2 expression at days 7 and 14 post-LCMV infection. The numbers indicate the percentage of Bcl2+ T cells in each sample. Data are representative of three independent experiments.

(f,g) CD8+ T cells were isolated from (f) WT mice at day 5, 7 and 9 post-LCMV infection or from (g) WT, Prdm1+/− or Prdm1−/− mice at day 7 post-infection. Cells were labeled with CFSE and cultured for two days with or without the indicated cytokines. Numbers indicate the percentage of divided cells. Histograms are representative of three independent experiments.

Figure1 see also Figure S1.

Next, we tested the hypothesis that differential expression of cytokine receptors in WT versus Prdm1−/− cells contribute to altered T cell survival and/or proliferation. First, CD8+ T cells were isolated from WT mice infected with LCMV Armstrong at day 5 (early expansion), day 7 (peak expansion), and day 9 (contraction) post-infection, labeled with CFSE, and cultured in vitro for two days with a panel of cytokines (Fig1f). When isolated at day 5 post-infection, CD8+ T cells spontaneously proliferated in the absence of any cytokine; this response was dramatically diminished by day 9 post-infection. In addition, cells isolated at day 5 post-infection showed enhanced proliferation in response to exogenous IL-12, as well as to the common γ chain (γc)-family cytokines, IL-2, IL-4 and IL-15. This response was also attenuated by day 9 post-infection. In contrast, at day 7 post-infection Prdm1−/− cells had a dramatically enhanced proliferative response to the cytokines IL-2, IL-12, and IL-15 compared to Prdm1+/− and WT T cells (Fig1g).

As the cytokine responsiveness of T cells during virus infections can be regulated by the modulation of cytokine receptor expression (Kalia et al., 2010; Sarkar et al., 2008), we further examined the expression of CD25 (IL-2Rα), CD122 (IL-2Rβ or IL-15Rβ) and CD124 (IL-4Rα) on CD8+ T cells at various time points post-LCMV infection. Consistent with the functional data (Fig1f), CD25 and CD122 protein and mRNA expression peaked at day 5 post-infection and then diminished over time through day 9 (FigS1f,g). A similar pattern was seen for transcripts encoding the subunits of the IL-12 receptor, Il12rb1 and Il12rb2. In contrast, CD124 was expressed in the naïve CD8+ T cells, and both the protein and mRNA were down-regulated following LCMV infection, even at day 5 (FigS1f,g). Prdm1−/− CD8+ T cells exhibited increased expression of Il2ra mRNA and to a lesser extent Il12rb2 mRNA compared to WT control cells (FigS1h). Overall, these data suggest that decreased responsiveness to cytokines by WT CD8+ T cells is due to cytokine receptor down-regulation, and that Blimp-1 might function to suppress cytokine receptor expression during the anti-viral immune response.

Blimp-1 regulates CD25 expression in virus-specific CD8+ T cells during the anti-viral immune response

To test whether CD25 expression was dysregulated in CD8+ T cells lacking Blimp-1, we infected WT and Prdm1−/− mice with LCMV, and analyzed virus-specific CD8+ T cells at days 7-9 post-infection. At day 7, approximately twice as many Prdm1−/− T cells expressed CD25 as compared to WT T cells (Fig2a,b). This trend persisted through day 9 post-infection, although the overall proportion of CD25+ CD8+ T cells diminished dramatically after day 7. This function of Blimp-1 was intrinsic to the CD8+ T cells, as a similar difference in CD25 expression was seen in P14 T cell receptor (TCR) transgenic WT or Prdm1flox/floxxGzB-cre+ CD8+ T cells (in which Cre expression is under the control of the human Granzyme B promoter (Rutishauser et al., 2009)) responding to LCMV following adoptive transfer into WT congenic hosts (Fig2c). To further confirm the inverse correlation of Blimp-1 with CD25 expression, we infected Blimp-1-GFP mice, in which GFP expression is under the control of the Prdm1 gene (Kallies et al., 2006). At day 7 and 9 post-infection, a high proportion of CD44hi CD8+ T cells expressed Blimp-1; further, there was an increased proportion of CD25+ cells in the Blimp-1lo versus Blimp-1hi subset (Fig2d). These data together indicate that Blimp-1 represses CD25 expression at late stages during the anti-viral CD8+ T cell response.

An external file that holds a picture, illustration, etc. Object name is nihms-521661-f0002.jpg
Increased CD25 expression in Prdm1−/− CD8+ T cells during the anti-viral immune response

(a,b) CD8+ T cells from WT and Prdm1−/− mice were isolated at day 7-9 post-LCMV infection and stained with antibodies to CD25 plus GP33 or NP396 LCMV-specific tetramers. CD25 expression on tetramer-positive CD8+ T cells is shown (a) along with a compilation of data from three independent experiments (b). All error bars represent SEM.

(c) P14 TCR transgenic CD90.1+ Prdm1flox/floxxGzB-cre+ or Prdm1flox/floxxGzB-cre− (WT) littermate splenocytes were adoptively transferred into CD90.2+ congenic WT mice, followed by infection with LCMV. At day 9 post-infection, CD90.1+CD8+ T cells were analyzed for CD25 expression (left panels). On right, the percentages of CD25+CD90.1+CD8+ T cells from two independent experiments each with 4 to 5 mice are shown.

(d) Blimp-1-GFP reporter mice were infected with LCMV, and CD44hi CD8+ T cells were analyzed at days 5, 7 and 9 post-infection for GFP expression (left panels). At right, Blimp-1lo and Blimp-1hi populations were analyzed for CD25 expression. The percentage ± SEM of CD25-positive cells is indicated on each plot. Data are representative of two independent experiments.

(e) WT mice were infected with LCMV and CD44hi CD8+ T cells were analyzed on days 5 and 9 post-infection for CD25 expression (left panels). At right, CD25lo and CD25hi populations were analyzed for IL-7R and CD62L expression. The percentages of positive cells are indicated on each plot. Data are representative of four independent experiments.

(f) P14 TCR transgenic CD45.1+ splenocytes were adoptively transferred into CD45.2+ congenic WT mice, followed by infection with LCMV. At day 7 post-infection, 1×106 CD25lo or CD25hi P14+ CD45.1+ CD8+ T cells were isolated by cell sorting and adoptively transferred into infection-matched CD45.2+ recipient mice. Eight days later, P14+ cells were assessed. Data shown include a compilation of two independent experiments and representative flow cytometry analysis.

Figure2 see also Figure S2.

Recent studies have demonstrated that CD25 expression during the early expansion phase of the anti-viral CD8+ T cell response is critical in regulating the cell fate, with CD25hi cells becoming short-lived effector cells (Kalia et al., 2010; Pipkin et al., 2010). Consistent with these data, we also observed that high CD25 expression at day 5 post- infection correlated with decreased IL-7R and CD62L expression (Fig2e); further, at this stage of the response, cells expressing higher Blimp-1 also expressed more CD25 (Fig2d) (Kalia et al., 2010). These data indicate that strong IL-2R signaling during the early stage of the anti-viral CD8+ T cell response promotes the development of short-lived effector cells. However, we considered whether the role of IL-2R signaling might be different at the later stage of the response, at which time CD8+ T cells are undergoing attrition. In support of this possibility, we found that at day 9 post-infection, CD25 expression on CD8+ T cells positively correlated with IL-7R and CD62L expression (Fig2e), indicating that the CD25hi population at this stage is enriched in memory precursor cells as compared to the CD25lo population. Analysis of virus-specific CD8+ T cells based on KLRG1 and IL-7R expression also showed higher expression of CD25 on both KLRG1hi IL-7Rlo and KLRG1lo IL-7Rhi populations from the Prdm1−/− mice compared to WT controls (FigS2a,b).

To test directly whether differences in CD25 expression on virus-specific CD8+ T cells are associated with differences in T cell survival, we sorted CD25hi and CD25lo P14 TCR transgenic CD8+ T cells from mice at day 7 post-LCMV infection. Sorted T cells were adoptively transferred into infection-matched congenic hosts, and the number of P14+ T cells was assessed at day 15 post-infection (8 days post-transfer). We found that a higher proportion and absolute number of virus-specific CD25hi CD8+ T cells survived when compared to the CD25lo transferred cells (Fig2f). These results indicate that, at day 7 post-infection, CD25 expression on virus-specific CD8+ T cells correlate with enhanced survival and/or persistence. To determine whether enhanced proliferation contributed to this effect, LCMV-infected mice were injected with BrdU at day 7 post-infection, and virus-specific CD8+ T cells were analyzed 12 hours later. A higher proportion of CD25hi T cells exhibited BrdU incorporation relative to the CD25lo subset, in WT and Prdm1−/− mice (FigS2c). Additionally, in response to ex vivo LCMV peptide stimulation at day 8 post-infection, a higher proportion of CD25hi CD8+ T cells produced interferon-γ (IFN-γ) and IL-2 compared to CD25lo CD8+ T cells (FigS2d), correlating high expression of CD25 with a functional memory T cell response. These data indicate that, following the peak of the CD8+ T cell response to LCMV, higher expression of CD25 correlates with enhanced survival of virus-specific T cells.

Blimp-1 regulates CD27 expression in virus-specific CD8+ T cells

CD27 is a co-stimulatory molecule that promotes CD8+ T cell proliferation and memory generation (Duttagupta et al., 2009; Hendriks et al., 2000). To test whether CD27 expression in CD8+ T cells is regulated by Blimp-1, we examined T cells from LCMV-infected WT and Prdm1−/− mice. At day 9 post-infection, there was a higher proportion of CD27+ virus-specific CD8+ T cells in the Prdm1−/− mice compared to controls (Fig3a), consistent with previous reports (Kallies et al., 2009; Rutishauser et al., 2009). This difference was visible at day 7.5 post-infection but increased in magnitude at days 8 and 9 (Fig3b). As was the case for CD25 expression (Fig2c), altered CD27 expression was due to a CD8+ T cell-intrinsic role for Blimp-1 (Fig3c), and was observed in both KLRG1hi IL-7Rlo and KLRG1lo IL-7Rhi populations (FigS3a,b). As a further correlation, we examined CD8+ T cells in LCMV-infected Blimp-1-GFP mice. At day 9 post-infection, Blimp-1hi cells expressed less CD27 than Blimp-1lo cells (Fig3d).

An external file that holds a picture, illustration, etc. Object name is nihms-521661-f0003.jpg
Increased CD27 expression in Prdm1−/− CD8+ T cells during the anti-viral immune response

(a,b) CD8+ T cells from WT and Prdm1−/− mice were isolated at day 9 post-LCMV infection and stained with antibodies to CD27 plus GP33 or NP396 LCMV-specific tetramers. CD27 expression on tetramer-positive CD8+ T cells is shown (a) along with a compilation of data from three experiments (b). All error bars represent SEM.

(c) P14 TCR transgenic CD90.1+ Prdm1flox/floxxGzB-cre+ or Prdm1flox/floxxGzB-cre− (WT) littermate splenocytes were adoptively transferred into CD90.2+ WT mice, followed by infection with LCMV. At day 9 post-infection, CD90.1+CD8+ T cells were analyzed for CD27 expression (left panels). At right, a compilation of data from two independent experiments each with 4 to 5 mice is shown.

(d) Blimp-1-GFP reporter mice were infected with LCMV, and CD44hi CD8+T cells were analyzed at day 9 post-infection for GFP expression (top panel). Below, Blimp-1lo and Blimp-1hi populations were analyzed for CD27 expression. The percentages of CD27-positive cells and the MFI of CD27 expression are indicated on each histogram. Data are representative of two independent experiments.

(e) WT mice were infected with LCMV and CD44hi CD8+ T cells were analyzed on day 9 post-infection for CD27 expression (left panel). At right, CD27lo (R1), CD27int (R2), and CD27hi (R3) populations were analyzed for IL-7R expression. The percentages of IL-7R-positive cells are indicated on each histogram. Data are representative of four independent experiments.

(f) P14 TCR transgenic CD45.1+ splenocytes were adoptively transferred into CD45.2+ WT mice, followed by infection with LCMV. At day 9 post-infection, 1×106 CD27lo or CD27hi P14+ CD45.1+ CD8+ T cells were isolated by cell sorting and were adoptively transferred into infection-matched CD45.2+ recipient mice. Seven days later, the absolute numbers and percentages of P14+ cells were determined. Data shown include a compilation of two independent experiments and representative flow cytometry analysis.

Figure3 see also Figure S3.

A previous study showed that CD27-expressing CD8+ T cells represent a functional memory cell pool (Hikono et al., 2007). In order to test whether the CD27+ CD8+ T cells present at day 9 post-LCMV infection represent a subset enriched in memory precursor cells, we examined IL-7R expression on cells with varying CD27 expression, and found that increasing amounts of IL-7R correlated with increased CD27 expression (Fig3e).

Virus-specific P14 CD27hi and CD27lo CD8+ T cells were next sorted from mice at day 9 post-LCMV infection, and adoptively transferred into infection-matched hosts. When analyzed 7 days later (day 16 post-infection),there was a higher proportion and absolute number of P14 CD8+T cells from the CD27hi subset compared to the CD27lo subset (Fig3f). These data together indicate that Blimp-1 normally represses CD27 expression at late stages of the anti-viral CD8+ T cell response and thereby contributes to the attrition of effector cells following virus clearance.

CD25 and CD27 expression enhances the survival of CD8+ T cells

To test whether persistent CD25 or CD27 expression has a functional role in promoting CD8+ T cell survival during an anti-viral immune response, we used retroviral gene transfer to constitutively express CD25 or CD27 in LCMV-specific T cells. P14 TCR transgenic CD8+ T cells were stimulated in vitro with LCMV-GP33 peptide, infected with retroviruses, and then transferred into LCMV-infected (day 1 post-infection) recipient mice (Fig4a). Expression of hCD2 indicated that there was a similar transduction efficiency for each of the retrovirus (RV) constructs; furthermore, retroviral transduction led to increased CD25 or CD27 expression compared to mock RV controls (Fig4b). When analyzed at days 8, 11 and 14 post-infection, LCMV-specific P14+ cells expressing either CD25 or CD27 were increased in proportion relative to cells transduced with the empty RV (mock RV; Fig4c,d). These results indicate that persistent CD25 and CD27 expression promotes enhanced expansion of the T cells rather than effecting contraction, thereby resulting in increased survival of CD8+ T cells.

An external file that holds a picture, illustration, etc. Object name is nihms-521661-f0004.jpg
Sustained CD25 and CD27 expression extends the survival of CD8+ T cells

(a) Splenocytes from P14 TCR transgenic mice were stimulated with GP33 peptide for 1 day, and then transduced with retroviruses expressing CD25 or CD27, or an empty virus (mock RV) as a negative control. 2×105 splenocytes were adoptively transferred into recipient mice one day following LCMV infection, and splenocytes from infected recipient mice were analyzed on days 8, 11 and 14 post-infection.

(b) A subset of transduced splenocytes was cultured in vitro for 2 days, and the transduction efficiency assessed by staining for human CD2 (hCD2). At right, enforced CD25 or CD27 expression on hCD2+ cells was assessed.

(c,d) Representative plots (c) and a compilation of data (d) from two independent experiments showing hCD2+ CD8+ T cells at days 8, 11 and 14 post-LCMV infection in recipient mice receiving cells transduced with the indicated retrovirus.

(e) P14 TCR transgenic CD90.1+ Prdm1flox/floxxGzB-cre+ splenocytes were adoptively transferred into CD90.2+ WT mice, followed by infection with LCMV. At day 5 post-infection, CD90.1+CD8+ T cells isolated by cell sorting were stimulated with αCD3 and αCD28 for 24 hours and then transduced with retroviruses expressing short hairpin RNAs to silencing CD25 (shCD25), shCD27, shCD25 and shCD27, or scrambled, as a negative control. Retrovirally-transduced cells were adoptively transferred into day 6 infection-matched CD90.2+ recipient mice.

(f) A subset of transduced CD90.1+CD8+ T cells was cultured in vitro for 2 days, and the transduction efficiency assessed by GFP fluorescence.

(g,h) Three days following transfer, the percentages ± SEM of GFP+CD90.1+CD8+ cells were determined; dot-plots show a representative experiment (g, top panel) and a compilation of data from three independent experiments is shown as a scatter plot (h). A representative example of CD25 and CD27 expression on GFP+CD90.1+CD8+ cells is shown, along with the percentages ± SEM from three independent experiments compiled (g, lower panels).

To address whether CD25 or CD27 expression accounts for the enhanced survival of Prdm1−/− CD8+ T cells during an anti-viral immune response, we utilized shRNA silencing to diminish expression of these proteins in LCMV-specific Prdm1−/− T cells. For these experiments, adoptively transferred P14+ CD90.1+ Prdm1flox/floxxGzB-cre+ splenocytes were activated in vivo by infecting recipient mice with LCMV. At day 5 post-infection, P14+ T cells were isolated by cell sorting, stimulated in vitro with anti-CD3 and anti-CD28 antibodies for 24 hours, infected with retroviruses, and then transferred into LCMV-infected (day 6 post-infection) recipient mice (Fig4e). As shown, we achieved similar transduction efficiencies for each of the RV constructs (Fig4f); in addition, each short hairpin (sh) RNA was able to reduce expression of the targeted gene (Fig4g). When analyzed at day 9 post-infection, Prdm1−/− CD8+ T cells transduced with shCD25 or shCD27 were reduced in proportion compared to cells transduced with the scrambled shRNA control (Fig4g,h). Moreover, silencing of both CD25 and CD27 further decreased the frequency of GFP+ CD8+ T cells (Fig4g,h). These results indicate that the increased magnitude and survival of CD8+ T cells in Prdm1−/− mice was in part, due to sustained expression of CD25 and CD27.

Genome-wide ChIP-Seq analysis identifies Il2ra and Cd27 as direct targets of Blimp-1 in CD8+ T cells

Transcriptional regulation by Blimp-1 has been extensively studied in B cells, leading to the identification of Myc, Pax5, Bcl6, Ciita, Spib, Id3, and Prdm1 as direct Blimp-1 target genes (Martins and Calame, 2008). Substantially less is known about the direct targets of Blimp-1 in T cells, particularly CD8+ T cells. To address whether Il2ra or Cd27 might be direct targets of Blimp-1, we performed genome-wide ChIP-Seq analysis. Naïve CD8+ T cells from OT-I TCR transgenic Rag1−/− mice were stimulated for three days in vitro, in the presence of IL-2 to up-regulate Blimp-1 protein (Gong and Malek, 2007) (Fig5a). Chromatin was immunoprecipitated with anti-Blimp-1 antibody and DNA was subjected to deep sequencing to identify the genomic locations of Blimp-1 binding sites. The majority of sites identified were located within introns or in regions distal to a gene (49% and 29%, respectively); a minority of binding sites were found in promoter regions (Fig5b). Using the Ingenuity pathway database, we classified 10847 Blimp-1 target genes, and found that the majority of known genes were in the groups of enzymes, transcriptional regulators, transporters and kinases (FigS4). To address the functional importance of specific Blimp-1 target genes in CD8+ T cells, we chose to focus on cytokine and other transmembrane receptors (Fig5c and Table S1), as these are known to be critical in regulating T cell function and differentiation. Consistent with previous studies (Ji et al., 2011; Rutishauser et al., 2009), this analysis identified Prdm1, Id3, Bcl6, Pax5, Myc and Ciita genes as having Blimp-1 binding sites in CD8+ T cells (Fig5d). In addition we identified Il2ra, Cd27, Il2rb, Sell (Cd62l), Eomes and Ccr6 as Blimp-1 targets (Fig5e). In CD4+ T cells, Il2 is a downstream target of Blimp-1 (Martins et al., 2008); however, we did not detect Blimp-1 binding to the Il2 locus in CD8+ T cells.

An external file that holds a picture, illustration, etc. Object name is nihms-521661-f0005.jpg
Genome-wide identification of Blimp-1 binding sites in CD8+ T cells

(a) Naïve CD8+ T cells isolated from OT-I+Rag1−/− mice were stimulated with plate bound αCD3 and αCD28 and IL-2 (50ng/ml) for 3 days, followed by Blimp-1 ChIP-Seq.

(b) Frequency (pie chart) of Blimp-1 binding sites localized to each region of target genes in the mouse genome (mm9). TSS, transcription start site; TES, transcription end site.

(c) Functional grouping of genes related to secreted proteins and their cognate receptors with Blimp-1 binding sites within 10kb of the transcription start site.

(d, e) Identification of Blimp-1 binding sites on known (d) and unknown (e) target genes. The scale bars indicate the relative kbp scale on each gene, and the numbers on the right display the magnitude of sequence enrichment on a log2 scale. Peaks were identified with Cisgenome 2.0 and all Blimp-1 binding site enrichment data are shown in the ChIP-Seq tracks.

(f) WT and Prdm1−/− mice were infected with LCMV, and CD44hi CD8+ T cells were isolated at day 7 post-infection. As a control, naïve CD8+ T cells were isolated from uninfected mice (D0). ChIP assays were performed with antibodies to Blimp-1 or mouse IgG. Each ChIP eluate was amplified by Q-PCR for the indicated regions of the Il2ra gene (PCR amplicons 1 and 2; left two panels) or the Cd27 gene (PCR amplicons 1, 2, and 3; right three panels). Amplicon 2 of Il2ra and amplicon 3 of Cd27 indicate non-specific regions where Blimp-1 binding was not detected by ChIP-seq. Data shown include a compilation of three independent experiments and error bars represent SEM.

Figure5 see also Figure S4 and Table S1.

The ChIP-Seq analysis identified Blimp-1 binding sites in the first intron of the Il2ra gene and at two distal sites within 10kb of the Cd27 transcription start site (Fig5e). When using activated cells from LCMV-infected mice at day 7 post-infection, we found that Blimp-1 bound to the Il2ra and Cd27 gene loci in WT CD8+ T cells (Fig5f). As expected, there was no binding of Blimp-1 to any of these sites in Prdm1−/− CD8+ T cells, nor was Blimp-1 binding detected at non-specific regions (Amp2 in Il2ra and Amp3 in Cd27). These data support the conclusion that Blimp-1 binding to the Il2ra and Cd27 loci represses expression of CD25 and CD27, respectively, during the anti-viral CD8+ T cell response.

Blimp-1 is associated with histone modifying enzymes G9a and HDAC2 in CD8+ T cells

To address the mechanism by which Blimp-1 mediates transcriptional repression in primary CD8+ T cells, we first performed co-immunoprecipitation assays on T cells activated in vitro. Since Blimp-1 expression in T lymphocytes requires a combination of TCR plus cytokine stimulation (Gong and Malek, 2007), we compared cells activated through the TCR for three days in the presence or absence of a cytokine cocktail containing IL-2, IL-4, and IL-12. Blimp-1 was then immunoprecipitated, and potential binding partners were identified by immunoblotting with antibodies to the histone methyltransferases, G9a or EZH2, or to the histone acetyltransferases, HDAC1 or HDAC2. As shown in Fig6a, Blimp-1 associated with G9a and HDAC2 but not with HDAC1 or EZH2. Similarly, immunoprecipitates of HDAC2 or G9a co-precipitated Blimp-1, but, interestingly, G9a was not associated with HDAC2, indicating that the Blimp-1:HDAC2 and the Blimp-1:G9a complexes are distinct. To confirm these interactions in a more physiological setting, a single cell-based proximity ligation assay was performed. This assay reveals protein-protein interactions when two protein-antibody complexes are in sufficiently close proximity. Interactions between Blimp-1 and G9a, Blimp-1 and HDAC2, or Blimp-1 and EZH2 were analyzed in situ in activated CD8+ T cells by intracellular labeling with each antibody (Fig6b). These data demonstrate that Blimp-1 and G9a, or Blimp-1 and HDAC2, are within 16 nm of each other in intact nuclei of activated CD8+ T cells, thereby supporting a role for Blimp-1 in recruiting these repressive chromatin modifying enzymes to Blimp-1 target genes in CD8+ T cells.

An external file that holds a picture, illustration, etc. Object name is nihms-521661-f0006.jpg
Blimp-1 recruits G9a and HDAC2 in activated CD8+ T cells

(a) WT CD8+ T cells were stimulated in vitro with αCD3 and αCD28 for 2 days with (+) or without (−) a cocktail of cytokines containing IL-2, IL-4 and IL-12. Lysates were immunoprecipitated with αBlimp-1 (left top panels, right top panel and middle panel), αHDAC2 (left bottom panels), or αG9a (right bottom panels) antibodies, followed by immunoblotting (IB) with the indicated antibodies. Data are representative of three independent experiments. WCL, whole cell lysate.

(b) Stimulated CD8+ T cells were stained with αHDAC2 and either a mouse IgG control (ISO) or αBlimp-1 (left upper panels), with αG9a and a mouse IgG control or αBlimp-1 (right upper panels) or with αEZH2 and a mouse IgG control or αBlimp-1 (left lower panels), followed by the Duolink proximity ligation assay. Samples were counter-stained for nuclei (blue; DAPI). Yellow signals demonstrate close proximity of the two proteins. The graph (lower right panel) is a compilation of data from three independent experiments and error bars represent SEM.

Blimp-1 promotes repressive chromatin modifications by recruiting histone modifying enzymes to target genes

To examine whether Blimp-1 mediates the recruitment of histone modifying enzymes G9a or HDAC2 to Blimp-1 binding sites, we isolated CD8+ T cells from WT or Prdm1−/− mice at day 7 post-LCMV infection. Chromatin immunoprecipitation (ChIP) assays revealed that both G9a and HDAC2 were present at the Il2ra and Cd27 gene loci in WT CD8+ T cells at day 7 post-infection (Fig7a,b). We did not detect binding of G9a, or HDAC2 to any of these sites in Prdm1−/− CD8+ T cells, indicating that Blimp-1 is required for the recruitment of G9a and HDAC2 to these genes. To determine whether Blimp-1 forms a complex with either G9a or HDAC2 at these loci, sequential ChIP assays (ChIP-reChIP) were performed with CD8+ T cells from WT or Prdm1−/− mice at day 7 post-LCMV infection. First, a primary ChIP with anti-Blimp-1 antibody was performed; following elution of these complexes from the primary antibody, secondary immunoprecipitations with anti-G9a or anti-HDAC2 antibodies were performed (Fig7c). These data indicated that Blimp-1 is in close proximity to G9a and HDAC2 on the Il2ra gene locus. On the Cd27 gene, Blimp-1 was tightly associated with HDAC2, but association with G9a was weaker at the distal site and absent from the proximal region. Together, these findings provide evidence that Blimp-1 recruits G9a and HDAC2 to target genes in primary T cells activated in vivo.

An external file that holds a picture, illustration, etc. Object name is nihms-521661-f0007.jpg
Blimp-1 regulates histone modifications at the Il2ra and Cd27 gene loci in CD8+ T cells at day 7 post-LCMV infection

(a, b) WT and Prdm1−/− mice were infected with LCMV, and CD44hi CD8+ T cells were isolated at day 7 post-infection. As a control, naïve CD8+ T cells were isolated from uninfected mice (D0). ChIP assays were performed with antibodies to G9a, or HDAC2. Each ChIP eluate was amplified by Q-PCR for the indicated region (PCR amplicon1) of the Il2ra gene or for each of two regions (PCR amplicon1 and amplicon2) of the Cd27 gene. Data shown include a compilation of three independent experiments.

(c) ChIP eluates from αBlimp-1 immunoprecipitates were re-precipitated with αG9a, αHDAC2 or rabbit IgG (isotype control; ISO) antibodies. Secondary ChIP eluates were amplified by Q-PCR for the indicated regions of the Il2ra and Cd27 genes. Data shown include a compilation of three independent experiments.

(d,e) WT and Prdm1−/− mice were infected with LCMV, and CD44hi CD8+ cells were isolated at day 7 post-infection. ChIP assays were performed with antibodies to modified histone H3 as indicated. ChIP eluates were amplified by Q-PCR for the indicated regions of the Il2ra and Cd27 genes. Two permissive modifications of histone H3 (acetylated H3, H3Ac; tri-methylated lysine 4; H3K4me3) and three repressive modifications (di-methylated lysine 9, H3Krme2; tri-methylated lysine 9, H3K9me3; tri-methylated lysine 27, H3K27me3) were analyzed. Iso, isotype control antibody. Data shown include a compilation of three independent experiments.

(f) The mRNA expression of Il2ra and Prdm1 in CD25hi or CD25lo CD44hi CD8+ T cells sorted at day 5 post-infection were analyzed by quantitative RT-PCR. Data shown include a compilation of three independent experiments.

(g-i) CD25hi CD44hi CD8+ T cells or CD27hi CD44hi CD8+ T cells were sorted at day 5 post-infection. ChIP assays were performed with antibodies to Blimp-1, G9a, HDAC2 or control IgG, and with antibodies to modified histone H3 as indicated. For direct comparison, ChIP assays were performed with antibodies to Blimp-1 or control IgG on CD44hi CD8+ T cells sorted at day 7 post-infection (g,h; D7*). Each ChIP eluate was amplified by Q-PCR for the indicated regions of the Il2ra and Cd27 genes. Each graph is a compilation of data from three independent experiments and error bars represent SEM.

To determine whether this recruitment is associated with changes in histone modifications at these loci, we examined histone H3 acetylation and methylation on the Il2ra and Cd27 genes by ChIP. Histone H3 acetylation (H3Ac) and trimethylation on lysine 4 (H3K4me3) are indicative of permissive chromatin states, as they correlate with gene expression (Weng et al., 2012). In contrast, histone H3 methylation on lysine 9 (H3K9me2 or H3K9me3) and on lysine 27 (H3K27me3) mark repressive chromatin states that correlate with gene silencing (Kouzarides, 2007). CD8+ T cells isolated from Prdm1−/− mice at day 7 post-LCMV infection had increased amounts of H3Ac and H3K4me3, and reduced amounts of H3K9me2, H3K9me3 and H3K27me3, at the Il2ra gene locus compared to CD8+ T cells from WT control mice (Fig7d). Analysis of the Cd27 locus revealed that Prdm1−/− CD8+ T cells also showed an increase in permissive modifications and a reduction in repressive modifications compared to WT cells, although these correlations were not uniformly observed at both of the regions analyzed (Fig7e). Overall, these chromatin states correlate with the increased CD25 and CD27 expression seen in Prdm1−/− versus WT CD8+ T cells from LCMV-infected mice.

At day 5 post-infection Blimp-1 and CD25 were highly expressed, and, in contrast to day 9 post-infection, their expression positively correlated with each other (Fig2d and Fig7f). ChIP analysis of the Il2ra locus in CD25hi CD8+ T cells isolated at day 5 post-LCMV infection showed high amounts of histone H3 acetylation in the absence of repressive modifications, along with an absence of Blimp-1, G9a, or HDAC2 binding (Fig7g). Blimp-1 binding was still detected at one site on Cd27 at day 5-post-infection, but in this case, Blimp-1 binding was not sufficient to recruit G9a or HDAC2 (Fig7h), consistent with the lack of repressive modifications and high amounts of histone H3 acetylation at this locus (Fig7i). Together, these findings demonstrate that Blimp-1 induces a repressive chromatin state at the Il2ra and Cd27 genes in CD8+ T cells at the peak of the response to LCMV infection, but not at earlier stages.

DISCUSSION

Prdm1−/− CD8+ T cells have an enhanced capacity to differentiate into memory cells (Kallies et al., 2009; Rutishauser et al., 2009). In this report, we have identified an important mechanism by which Blimp-1 association with histone modifying enzymes suppresses CD25 and CD27 expression to regulate the overall magnitude of the virus-specific memory CD8+ T cell response. Upon virus infection and T cell activation, co-stimulatory signals, along with the inflammatory cytokine milieu, are key components leading to the generation of CD8+ effector and memory-precursor T cells that together promote virus clearance and provide long-term protection against re-infection. After virus clearance, CD8+ T cells down-regulate co-stimulatory receptor and cytokine receptor expression (Dolfi et al., 2008; Kalia et al., 2010). Following this process, the vast majority of the virus-specific effector population undergoes apoptosis (Razvi et al., 1995), most likely due to cytokine deprivation and the lack of co-stimulatory signals. Our data indicate that CD25 (or CD27) is impacting the small proportion of effector cells capable of long-term survival, and provides an important signal to promote the survival of cells in this specific subset that are expressing the receptor. We propose that Blimp-1 normally functions to down-regulate cytokine receptor expression, thereby promoting the death of effector T cells. This mechanism would function in addition to regulation of proliferation and survival genes by Blimp-1, as proposed previously (Martins et al., 2008).

Over the past decade, numerous studies have investigated the role of IL-2 signaling in CD8+ T cell activation and differentiation in response to virus infections (Boyman and Sprent, 2012). These studies show that CD25 expression and the strength of IL-2R signaling during T cell priming influences the relative generation of effector versus memory CD8+ T cells. Furthermore, in vivo administration of IL-2 or IL-15 during T cell contraction increases the size of the CD8+ memory pool. These data prompted us to investigate whether IL-2 receptor expression was regulated by distinct mechanisms at different stages of the anti-viral immune response, and further, whether the function of IL-2R signaling might also vary between T cell priming and T cell contraction. CD25 expression on CD8+ T cells is rapidly induced during the early stage of viral infection, mediated by a combination of TCR and IL-2R signaling (Boyman and Sprent, 2012). During this early stage, cells expressing CD25 are destined to be short-lived effectors, in keeping with their high Blimp-1 expression (Kalia et al., 2010; Rutishauser et al., 2009). Our data indicate that, at this time point, Blimp-1 expression is not able to override the positive signals promoting CD25 expression. In contrast, by day 7 post-infection, CD25 expression is nearly absent on virus-specific cells, a change that is Blimp-1-dependent. We propose that following viral clearance and the cessation of IL-2 production, CD25 expression becomes susceptible to regulation by Blimp-1. It has been shown previously that Prdm1−/− mice consistently maintain higher proportions and numbers of memory CD8+ T cells (Rutishauser et al., 2009). Although we have not directly looked at later time points post-viral infection, the transcription factor EOMES was upregulated in the Prdm1−/− mice, consistent with our ChIP-Seq analysis. Prdm1−/− mice were also more responsive to IL-15 stimulation. Therefore, it is likely that the EOMES and IL-15 axis is promoting the homeostatic proliferation and survival of the increased memory T cell pool in the Prdm1−/− mice.

Here we confirmed that CD25 and CD27 expression was significantly altered in Prdm1−/− CD8+ T cells during the course of LCMV infection. Blimp-1 directly bound to the Il2ra gene locus, was required for recruitment of HDAC2 and G9a to this gene, and was required for the repressive chromatin modifications seen at the Il2ra gene locus in CD8+ T cells from day-7 LCMV-infected mice. The transcriptional regulation of the Il2ra gene in response to TCR stimulation plus IL-2 has been well-studied (Malek, 2008). Several transcription factors orchestrate this response, and four major regulatory regions have been identified (Kim et al., 2001; Malek, 2008). The first intron of the gene contains binding sites for signal transducer and activator of transcription-5a (STAT5a), STAT5b and high-mobility group protein-1 (HMG-I (Y)), and binding of these factors occurs in response to IL-2 (Kim et al., 2001). Interestingly, it is this region that also contains the Blimp-1 binding site identified in our ChIP-Seq analysis and verified by our Blimp-1 ChIP assays, further supporting a model in which Blimp-1-mediated repression of Il2ra is antagonistic to the positive signals up-regulating Il2ra transcription.

Unlike CD25, the transcriptional regulation of CD27 is not well defined in T cells. While present on naïve T cells, TCR stimulation leads to further up-regulation of CD27 (de Jong et al., 1991). The mechanisms leading to CD27 down-regulation are even more poorly characterized. While ligation with CD70 or stimulation with phorbol myrsitate acetate (PMA) down-regulate CD27 expression (Hintzen et al., 1994; Nolte et al., 2005), the factors mediating this event are not known. Our work has provided evidence that Blimp-1 plays a key role in CD27 down-regulation following the peak of the anti-viral CD8+ T cell response. As with the Il2ra locus, Blimp-1 binds to Cd27 and recruits the histone modifying enzymes, HDAC2 and G9a, leading to repressive modifications at this locus during CD8+ T cell contraction.

The histone modifying enzymes, G9a,has recently been a topic of investigation in lymphocytes (Thomas et al., 2008). While the development of lymphocytes in G9a-deficient mice was found to be normal, amounts of H3K9me2 are greatly reduced, leading to aberrantly low usage of Igλ L chains in B cells (Thomas et al., 2008); in addition, G9a has been shown to regulate H3K9me3 in vivo (Collins and Cheng, 2010). More recently, G9a-deficient CD4+ T cells were shown to be impaired in T helper-2 (Th2) cell-associated cytokine production and exhibited increased IL-17A and IFN-γ production, demonstrating that G9a regulates genes involved in lineage specification of CD4+ T cells (Lehnertz et al., 2010). Our data indicate that in virus-specific CD8+ T cells, Blimp-1 recruits HDAC2 and G9a to critical target genes that determine the fate of these T cells during the contraction of the immune response. While it remains possible that Blimp-1 recruits additional chromatin modifying enzymes not examined in our studies, our data clearly demonstrate that Blimp-1 plays a key role in regulating epigenetic marks that impact the balance of short-lived effector versus memory precursor CD8+ T cells responding to acute virus challenge.

EXPERIMENTAL PROCEDURES

Mice

Mice were bred and housed in specific pathogen free conditions at the University of Massachusetts Medical School (UMMS) in accordance with the guidelines of the Institutional Animal Care and Use Committee of UMMS (IACUC). OT-I TCR transgenic Rag1−/− mice were purchased from Taconic. Wild type C57BL/6 (CD90.2+ CD45.2+) mice were purchased from Jackson Laboratories and CD4-Cre mice were a gift from Dr. Joonsoo Kang at UMMS. Prdm1−/− mice have been described previously (Ohinata et al., 2005). Unless otherwise indicated, C57BL/6 CD4-Cre+ transgenic mice were used as wild-type (WT) controls. For some experiments, P14+ Prdm1fl/fl mice were crossed to Granzyme B-Cre+ transgenic mice to delete Blimp-1 in activated CD8+ T cells as previously described (Rutishauser et al., 2009).

T cell Isolation and adoptive transfer

2×105 P14+ CD45.1+ splenocytes were adoptively transferred into CD45.2+ recipient mice, followed by infection with LCMV. At day 7 post-infection, total CD8+ T cells were isolated with CD8 T Cell Isolation Kit (Miltenyi Biotec) and then CD25lo or CD25hi P14+ CD45.1+ CD8+ T cells were further sorted on a FACS Aria. Sorted 1×106 cells were adoptively transferred into infection-matched CD45.2+ congenic WT mice. For CD27, 1×106 CD27lo or CD27hi P14+ CD45.1+ CD8+ T cells at day 9 post-infection were sorted on a FACS Aria and adoptively transferred into infection-matched CD45.2+ congenic WT mice. Recipient mice were analyzed at day 8 post-transfer for CD25, or day 7 post-transfer for CD27. The absolute numbers and percentages of P14+ cells were determined by staining with α-CD8, α-CD44 and α-CD45.1 antibodies.

Proximity ligation assay (PLA) and confocal microscopy

CD8+ T cells were cyto-spun onto positively-charged microscope slides (Fisher, 12-550-20) and washed with cold PBS twice, followed by fixation with 4% paraformaldehyde at 25°C for 10 min. Fixed cells were washed with PBS twice and permeabilized in 0.5% Triton X-100/PBS at 4°C for 6 min, followed by washing with 70% EtOH. After blocking samples in the PLA blocking solution for 30 min at 37°C, samples were incubated at 4°C overnight with αHDAC2 and either a mouse IgG control or αBlimp-1 (1:100 dilution), with αG9a and a mouse IgG control or αBlimp-1 (1:100 dilution) or with αEZH2 (1:100 dilution) and a mouse IgG control or αBlimp-1 (1:100 dilution), followed by the Duolink proximity ligation assay according to the manufacturer's instructions. Samples were counter-stained for nuclei (blue; DAPI). The signals (Red) from each pair of PLA probes were detected using laser-scanning confocal microscopy (Leica TCS SP5 II) with a 63x phase contrast oil immersion objective (numerical aperture=1.3). The nuclei images were captured using the UV laser. Duolink Detection kit 613 (LNK-90133-30), Duolink PLA probe Mouse Plus (LNK-90701-30) and Duolink PLA probe Rabbit Minus (LNK-90602-30) were purchased from Olink bioscience.

CFSE Labelling and in vitro T cells stimulation

CD8+ T cells were isolated from WT mice at day 5, 7 and 9 post-LCMV infection, or from WT, Prdm1+/− or Prdm1−/− mice at day 7 post-infection. Cells were labeled with CFSE as previously described (Shi et al., 2009) and were cultured for two days with or without the cytokines indicated. IL-1β, IL-6, IL-12, IL-18, IL-21 and IL-23 were purchased from R&D. IL-2 and IL-4 were purchased from BD Biosciences. IFNγ, IL-7 and IL-15 were purchased from PeproTech, INC. IFNβ was purchased from PBL Interferon Source. Cytokine concentrations were as follows: IFNβ (200U), IFNγ (200U), IL-1β (50ng/ml), IL-2 (40ng/ml), IL-4 (40ng/ml), IL-6 (40ng/ml), IL-7 (20ng/ml), IL-12 (20ng/ml), IL-15 (20ng/ml), IL-18 (20ng/ml), IL-21 (20ng/ml), IL-23 (40ng/ml).

Statistical Analysis

The statistical difference between samples was analyzed with unpaired Student's t test. All error bars in this report represent standard error of the mean (SEM).

Supplementary Material

02

ACKNOWLEDGMENTS

We thank A Tarakhovsky for Prdm1 floxed mice, SL Nutt for Blimp-1-GFP mice and RT Woodland for the contribution of mice. We thank members of the Berg and Welsh labs for technical assistance and helpful discussions. We thank OH Cho and R Rohatgi for helpful discussions and critical reading of the manuscript. We also thank Flow Cytometry Core Lab and Deep Sequencing Core facility at UMMS, Worcester. This work was supported by NIH grants AI084987 (LJB), AI017672 (RMW) and AI081675 (RMW), AI074699 (SMK) and HHMI (SMK).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Accession numbers

Blimp-1 ChIP-Seq data have been deposited in the The Gene Expression Omnibus, accession number GSE48358.

The authors have no financial conflicts of interest.

REFERENCES

  • Ancelin K, Lange UC, Hajkova P, Schneider R, Bannister AJ, Kouzarides T, Surani MA. Blimp1 associates with Prmt5 and directs histone arginine methylation in mouse germ cells. Nat Cell Biol. 2006;8:623–630. [PubMed] [Google Scholar]
  • Araki Y, Wang Z, Zang C, Wood WH, 3rd, Schones D, Cui K, Roh TY, Lhotsky B, Wersto RP, Peng W, et al. Genome-wide analysis of histone methylation reveals chromatin state-based regulation of gene transcription and function of memory CD8+ T cells. Immunity. 2009;30:912–925. [PMC free article] [PubMed] [Google Scholar]
  • Banerjee A, Gordon SM, Intlekofer AM, Paley MA, Mooney EC, Lindsten T, Wherry EJ, Reiner SL. Cutting edge: The transcription factor eomesodermin enables CD8+ T cells to compete for the memory cell niche. J Immunol. 2010;185:4988–4992. [PMC free article] [PubMed] [Google Scholar]
  • Blattman JN, Grayson JM, Wherry EJ, Kaech SM, Smith KA, Ahmed R. Therapeutic use of IL-2 to enhance antiviral T-cell responses in vivo. Nat Med. 2003;9:540–547. [PubMed] [Google Scholar]
  • Boyman O, Sprent J. The role of interleukin-2 during homeostasis and activation of the immune system. Nat Rev Immunol. 2012;12:180–190. [PubMed] [Google Scholar]
  • Collins R, Cheng X. A case study in cross-talk: the histone lysine methyltransferases G9a and GLP. Nucleic acids research. 2010;38:3503–3511. [PMC free article] [PubMed] [Google Scholar]
  • de Jong R, Loenen WA, Brouwer M, van Emmerik L, de Vries EF, Borst J, van Lier RA. Regulation of expression of CD27, a T cell-specific member of a novel family of membrane receptors. J Immunol. 1991;146:2488–2494. [PubMed] [Google Scholar]
  • Doering TA, Crawford A, Angelosanto JM, Paley MA, Ziegler CG, Wherry EJ. Network analysis reveals centrally connected genes and pathways involved in CD8+ T cell exhaustion versus memory. Immunity. 2012;37:1130–1144. [PMC free article] [PubMed] [Google Scholar]
  • Dolfi DV, Boesteanu AC, Petrovas C, Xia D, Butz EA, Katsikis PD. Late signals from CD27 prevent Fas-dependent apoptosis of primary CD8+ T cells. J Immunol. 2008;180:2912–2921. [PMC free article] [PubMed] [Google Scholar]
  • Duttagupta PA, Boesteanu AC, Katsikis PD. Costimulation signals for memory CD8+ T cells during viral infections. Crit Rev Immunol. 2009;29:469–486. [PMC free article] [PubMed] [Google Scholar]
  • Gong D, Malek TR. Cytokine-dependent Blimp-1 expression in activated T cells inhibits IL-2 production. J Immunol. 2007;178:242–252. [PubMed] [Google Scholar]
  • Gyory I, Wu J, Fejer G, Seto E, Wright KL. PRDI-BF1 recruits the histone H3 methyltransferase G9a in transcriptional silencing. Nat Immunol. 2004;5:299–308. [PubMed] [Google Scholar]
  • Hendriks J, Gravestein LA, Tesselaar K, van Lier RA, Schumacher TN, Borst J. CD27 is required for generation and long-term maintenance of T cell immunity. Nat Immunol. 2000;1:433–440. [PubMed] [Google Scholar]
  • Hikono H, Kohlmeier JE, Takamura S, Wittmer ST, Roberts AD, Woodland DL. Activation phenotype, rather than central- or effector-memory phenotype, predicts the recall efficacy of memory CD8+ T cells. J Exp Med. 2007;204:1625–1636. [PMC free article] [PubMed] [Google Scholar]
  • Hintzen RQ, Lens SM, Beckmann MP, Goodwin RG, Lynch D, van Lier RA. Characterization of the human CD27 ligand, a novel member of the TNF gene family. J Immunol. 1994;152:1762–1773. [PubMed] [Google Scholar]
  • Ji Y, Pos Z, Rao M, Klebanoff CA, Yu Z, Sukumar M, Reger RN, Palmer DC, Borman ZA, Muranski P, et al. Repression of the DNA-binding inhibitor Id3 by Blimp-1 limits the formation of memory CD8+ T cells. Nat Immunol. 2011;12:1230–1237. [PMC free article] [PubMed] [Google Scholar]
  • Joshi NS, Cui W, Chandele A, Lee HK, Urso DR, Hagman J, Gapin L, Kaech SM. Inflammation directs memory precursor and short-lived effector CD8(+) T cell fates via the graded expression of T-bet transcription factor. Immunity. 2007;27:281–295. [PMC free article] [PubMed] [Google Scholar]
  • Kalia V, Sarkar S, Subramaniam S, Haining WN, Smith KA, Ahmed R. Prolonged interleukin-2Ralpha expression on virus-specific CD8+ T cells favors terminal-effector differentiation in vivo. Immunity. 2010;32:91–103. [PubMed] [Google Scholar]
  • Kallies A, Hawkins ED, Belz GT, Metcalf D, Hommel M, Corcoran LM, Hodgkin PD, Nutt SL. Transcriptional repressor Blimp-1 is essential for T cell homeostasis and self-tolerance. Nat Immunol. 2006;7:466–474. [PubMed] [Google Scholar]
  • Kallies A, Xin A, Belz GT, Nutt SL. Blimp-1 transcription factor is required for the differentiation of effector CD8(+) T cells and memory responses. Immunity. 2009;31:283–295. [PubMed] [Google Scholar]
  • Kim HP, Kelly J, Leonard WJ. The basis for IL-2-induced IL-2 receptor alpha chain gene regulation: importance of two widely separated IL-2 response elements. Immunity. 2001;15:159–172. [PubMed] [Google Scholar]
  • Kouzarides T. Chromatin modifications and their function. Cell. 2007;128:693–705. [PubMed] [Google Scholar]
  • Kubicek S, O'Sullivan RJ, August EM, Hickey ER, Zhang Q, Teodoro ML, Rea S, Mechtler K, Kowalski JA, Homon CA, et al. Reversal of H3K9me2 by a small-molecule inhibitor for the G9a histone methyltransferase. Mol Cell. 2007;25:473–481. [PubMed] [Google Scholar]
  • Lehnertz B, Northrop JP, Antignano F, Burrows K, Hadidi S, Mullaly SC, Rossi FM, Zaph C. Activating and inhibitory functions for the histone lysine methyltransferase G9a in T helper cell differentiation and function. J Exp Med. 2010;207:915–922. [PMC free article] [PubMed] [Google Scholar]
  • Malek TR. The biology of interleukin-2. Annu Rev Immunol. 2008;26:453–479. [PubMed] [Google Scholar]
  • Martins G, Calame K. Regulation and functions of Blimp-1 in T and B lymphocytes. Annu Rev Immunol. 2008;26:133–169. [PubMed] [Google Scholar]
  • Martins GA, Cimmino L, Liao J, Magnusdottir E, Calame K. Blimp-1 directly represses Il2 and the Il2 activator Fos, attenuating T cell proliferation and survival. J Exp Med. 2008;205:1959–1965. [PMC free article] [PubMed] [Google Scholar]
  • Martins GA, Cimmino L, Shapiro-Shelef M, Szabolcs M, Herron A, Magnusdottir E, Calame K. Transcriptional repressor Blimp-1 regulates T cell homeostasis and function. Nat Immunol. 2006;7:457–465. [PubMed] [Google Scholar]
  • Mitchell DM, Ravkov EV, Williams MA. Distinct roles for IL-2 and IL-15 in the differentiation and survival of CD8+ effector and memory T cells. J Immunol. 2010;184:6719–6730. [PMC free article] [PubMed] [Google Scholar]
  • Nolte MA, Arens R, van Os R, van Oosterwijk M, Hooibrink B, van Lier RA, van Oers MH. Immune activation modulates hematopoiesis through interactions between CD27 and CD70. Nat Immunol. 2005;6:412–418. [PubMed] [Google Scholar]
  • Obar JJ, Lefrancois L. Memory CD8+ T cell differentiation. Ann N Y Acad Sci. 2010;1183:251–266. [PMC free article] [PubMed] [Google Scholar]
  • Ohinata Y, Payer B, O'Carroll D, Ancelin K, Ono Y, Sano M, Barton SC, Obukhanych T, Nussenzweig M, Tarakhovsky A, et al. Blimp1 is a critical determinant of the germ cell lineage in mice. Nature. 2005;436:207–213. [PubMed] [Google Scholar]
  • Peperzak V, Veraar EA, Keller AM, Xiao Y, Borst J. The Pim kinase pathway contributes to survival signaling in primed CD8+ T cells upon CD27 costimulation. J Immunol. 2010a;185:6670–6678. [PubMed] [Google Scholar]
  • Peperzak V, Xiao Y, Veraar EA, Borst J. CD27 sustains survival of CTLs in virus-infected nonlymphoid tissue in mice by inducing autocrine IL-2 production. J Clin Invest. 2010b;120:168–178. [PMC free article] [PubMed] [Google Scholar]
  • Pipkin ME, Sacks JA, Cruz-Guilloty F, Lichtenheld MG, Bevan MJ, Rao A. Interleukin-2 and inflammation induce distinct transcriptional programs that promote the differentiation of effector cytolytic T cells. Immunity. 2010;32:79–90. [PMC free article] [PubMed] [Google Scholar]
  • Piskurich JF, Lin KI, Lin Y, Wang Y, Ting JP, Calame K. BLIMP-I mediates extinction of major histocompatibility class II transactivator expression in plasma cells. Nat Immunol. 2000;1:526–532. [PubMed] [Google Scholar]
  • Razvi ES, Jiang Z, Woda BA, Welsh RM. Lymphocyte apoptosis during the silencing of the immune response to acute viral infections in normal, lpr, and Bcl-2-transgenic mice. Am J Pathol. 1995;147:79–91. [PMC free article] [PubMed] [Google Scholar]
  • Rutishauser RL, Martins GA, Kalachikov S, Chandele A, Parish IA, Meffre E, Jacob J, Calame K, Kaech SM. Transcriptional repressor Blimp-1 promotes CD8(+) T cell terminal differentiation and represses the acquisition of central memory T cell properties. Immunity. 2009;31:296–308. [PMC free article] [PubMed] [Google Scholar]
  • Sarkar S, Kalia V, Haining WN, Konieczny BT, Subramaniam S, Ahmed R. Functional and genomic profiling of effector CD8 T cell subsets with distinct memory fates. J Exp Med. 2008;205:625–640. [PMC free article] [PubMed] [Google Scholar]
  • Shin H, Blackburn SD, Intlekofer AM, Kao C, Angelosanto JM, Reiner SL, Wherry EJ. A role for the transcriptional repressor Blimp-1 in CD8(+) T cell exhaustion during chronic viral infection. Immunity. 2009;31:309–320. [PMC free article] [PubMed] [Google Scholar]
  • Smith MA, Wright G, Wu J, Tailor P, Ozato K, Chen X, Wei S, Piskurich JF, Ting JP, Wright KL. Positive regulatory domain I (PRDM1) and IRF8/PU.1 counter-regulate MHC class II transactivator (CIITA) expression during dendritic cell maturation. J Biol Chem. 2011;286:7893–7904. [PMC free article] [PubMed] [Google Scholar]
  • Su ST, Ying HY, Chiu YK, Lin FR, Chen MY, Lin KI. Involvement of histone demethylase LSD1 in Blimp-1-mediated gene repression during plasma cell differentiation. Mol Cell Biol. 2009;29:1421–1431. [PMC free article] [PubMed] [Google Scholar]
  • Thomas LR, Miyashita H, Cobb RM, Pierce S, Tachibana M, Hobeika E, Reth M, Shinkai Y, Oltz EM. Functional analysis of histone methyltransferase g9a in B and T lymphocytes. J Immunol. 2008;181:485–493. [PMC free article] [PubMed] [Google Scholar]
  • Weng NP, Araki Y, Subedi K. The molecular basis of the memory T cell response: differential gene expression and its epigenetic regulation. Nat Rev Immunol. 2012;12:306–315. [PMC free article] [PubMed] [Google Scholar]
  • Williams MA, Tyznik AJ, Bevan MJ. Interleukin-2 signals during priming are required for secondary expansion of CD8+ memory T cells. Nature. 2006;441:890–893. [PMC free article] [PubMed] [Google Scholar]
  • Yu J, Angelin-Duclos C, Greenwood J, Liao J, Calame K. Transcriptional repression by blimp-1 (PRDI-BF1) involves recruitment of histone deacetylase. Mol Cell Biol. 2000;20:2592–2603. [PMC free article] [PubMed] [Google Scholar]