HDAC1 Controls CD8+ T Cell Homeostasis and Antiviral Response

Reversible lysine acetylation plays an important role in the regulation of T cell responses. HDAC1 has been shown to control peripheral T helper cells, however the role of HDAC1 in CD8+ T cell function remains elusive. By using conditional gene targeting approaches, we show that LckCre-mediated deletion of HDAC1 led to reduced numbers of thymocytes as well as peripheral T cells, and to an increased fraction of CD8+CD4– cells within the CD3/TCRβlo population, indicating that HDAC1 is essential for the efficient progression of immature CD8+CD4– cells to the DP stage. Moreover, CD44hi effector CD8+ T cells were enhanced in mice with a T cell-specific deletion of HDAC1 under homeostatic conditions and HDAC1-deficient CD44hi CD8+ T cells produced more IFNγ upon ex vivo PMA/ionomycin stimulation in comparison to wild-type cells. Naïve (CD44l°CD62L+) HDAC1-null CD8+ T cells displayed a normal proliferative response, produced similar amounts of IL-2 and TNFα, slightly enhanced amounts of IFNγ, and their in vivo cytotoxicity was normal in the absence of HDAC1. However, T cell-specific loss of HDAC1 led to a reduced anti-viral CD8+ T cell response upon LCMV infection and impaired expansion of virus-specific CD8+ T cells. Taken together, our data indicate that HDAC1 is required for the efficient generation of thymocytes and peripheral T cells, for proper CD8+ T cell homeostasis and for an efficient in vivo expansion and activation of CD8+ T cells in response to LCMV infection.


Introduction
Dynamic changes in histone acetylation patterns are mediated by the activity of histone acetyltransfereases (HATs) and histone deacetylases (HDACs) and are key events in the epigenetic regulation of gene expression. In addition, many non-histone targets of HATs/HDACs have been described and it has been demonstrated that reversible lysine acetylation can affect proteinprotein and protein-DNA interactions, protein stability and intracellular localization. This implies that lysine acetylation is an important post-translational modification regulating a variety of cellular pathways and thus broadening the functional role of HATs/HDACs beyond epigenetic gene regulation [1].
The application of HDAC inhibitors revealed a variety of T cell functions controlled by reversible lysine acetylation [2]. The mammalian HDAC family is sub-divided into 4 classes consisting of 18 members [3] and several HDAC family members have been implicated in the regulation of T cell development and differentiation [2,4]. The combined activity of HDAC1 and HDAC2 is essential for the progression of double-negative (DN) to doublepositive (DP) thymocytes [5,6]. HDAC7 regulates both positive and negative selection during T cell development [7][8][9] and class II HDACs (HDAC4, 5 and 10) have been implicated in the ThPOK-mediated silencing of the Cd8 gene loci during CD4 lineage differentiation [10]. HDACs have also been connected to the regulation of regulatory T cell function [11]. The activity of FoxP3 is regulated by acetylation [12] and it has been shown that HDAC7 and HDAC9 bind to FoxP3. This suggests that both HDAC7 and HDAC9 might regulate the activity of FoxP3 and Tregs. Moreover, HDAC6-or HDAC9-deficiency leads to increased Treg numbers and enhanced Treg function [13,14]. HDAC7 also controls CTL function, since HDAC7 function has been linked with the repression of key cytokines, cytokine receptors and adhesion molecules important for CTL function [15]. Further, it has also been shown that HDAC1 and HDAC2 are essential to prevent neoplastic transformation of immature T cells [5,6]. By using conditional gene targeting approaches, we previously showed that HDAC1 is a key regulator of Th2 cytokine responses [16]. Loss of HDAC1 (using the Cd4Cre delete strain) led to an increased inflammatory response in an in vivo allergic airway inflammation model and mice with HDAC1-deficient T cells displayed an increase in all clinical parameters of this Th2-type asthma model. This correlated with enhanced Th2 cytokine production of HDAC1-deficient T cells isolated from diseased mice. Although this study clearly demonstrated an important function for HDAC1 in peripheral T helper cells, the role of HDAC1 in CD8 + T cells as well as during earlier steps of T cell development has not been explored.
In this study we employed conditional gene targeting approaches to investigate the role of Hdac1 during early T cell development using the LckCre deleter strain. Moreover, we studied whether CD8 + T cell function and effector differentiation are regulated by HDAC1 under steady state conditions and during viral infection using Cd4Cre-mediated deletion of HDAC1. Although the relative distribution of DP and mature CD4SP and CD8SP cells as well as the expression of surface marker such as CD24, CD5 and CD69 was not changed, LckCre-mediated loss of HDAC1 led to reduced numbers of thymocytes as well as peripheral T cells. There was an increased fraction of CD8 + CD4cells within the CD3/TCRb lo population in Hdac1 f/f LckCre mice, indicating that HDAC1 is essential for the efficient progression of immature CD8SP cells to the DP stage. In addition, we observed that CD44 hi effector CD8 + T cells were enhanced in mice with a T cell-specific loss of HDAC1 under homeostatic condition and that Hdac1 f/f LckCre CD44 hi CD8 + T cells produced more IFNc upon ex vivo PMA/ ionomycin stimulation in comparison to wild-type cells. Naïve (CD44 l uCD62L + ) Hdac1 f/f Cd4Cre CD8 + T cells displayed a normal proliferative response upon in vitro anti-CD3/anti-CD28 stimulation, produced similar amount of IL-2 and TNFa, while IFNc production was slightly increased compared to Hdac1 f/f CD8 + T cells upon activation. Hdac1 f/f and Hdac1 f/f Cd4Cre mice showed similar cytotoxic activity in vivo, suggesting that CTL function is not dependent on HDAC1. The infection with LCMV revealed reduced numbers of MHC class I/peptide-gp33 tetramer-positive (tet-gp33 + ) CD8 + T cells in the absence of HDAC1 correlating with impaired cytokine expression upon re-stimulation of splenocytes and liver cell suspensions in comparison to WT mice. Taken together, our data indicate that HDAC1 is required for the efficient generation of thymocytes and peripheral T cells, for proper CD8 + T cell homeostasis and for an efficient in vivo expansion and activation of CD8 + T cells in response to LCMV infection.

Mice
Hdac1 f/f Cd4Cre mice were previously described [16]. PLZFluxoid (Plzf lu/lu ) mice were kindly provided by A. Bendelac (University of Chicago). P14 were obtained from the European Mouse Mutant Archive, IL-15-null mice were obtained from Taconic. All mice were bred and maintained in the preclinical research facility of the Medical University of Vienna. Food and water was provided to the animals ad libitum.

Purification and in vitro activation of CD8 + T cells
Purification of CD8 + T cells was carried out as previously described [17]. In brief, spleens and lymph nodes of 6-10 week old mice were isolated and single cell suspensions were generated using a nylon cell strainer. After red blood cell lysis, cells were negatively depleted via MACS and naive (CD44 l uCD62L + ) CD8 + T cells were sorted on a FACSAria cell sorter (BD). Cells were seeded onto 48 well cell culture plates previously coated with anti-CD3 and anti-CD28 in the presence of 20 U/ml rhIL2 for 48 h, followed by splitting and resting for 2 days in presence of IL-2. For intracellular stainings, cells were restimulated with PMA/ionomycin for 4 h, followed by stainings performed as described [16], followed by flow cytometric analysis (LSRII, BD Biosciences). For ELISAs, cells were reactivated with platebound anti-CD3 for 16 hours. Supernatants were collected and ELISAs were performed using ebioscience ''Ready-Set-Go'' ELISA kits according to manufacturer's instructions.

CFSE labeling
Naive CD8 + T cells were isolated as described above, labeled with 1 mM CFSE (ebioscience), seeded onto cell culture plates and activated for 48 hours as described above.

Intracellular HDAC1 and HDAC2 staining
Splenocytes were pre-incubated with Fc-block (BD Pharmingen) and stained with anti-CD8a, CD4 and TCRb antibodies. For immunodetection of HDAC1 and HDAC2, cells were fixed and permeabilized using Foxp3 Fixation/Permeabilization Concentrate and Diluent (eBioscience) according to the manufacturer's instruction. Subsequently, cells were blocked in 5% normal goat serum and then incubated with rabbit anti-mouse HDAC1 and mouse anti-mouse HDAC2 antibodies in permeabilization buffer (eBioscience) for 1 hour. Cells were washed with permeabilization buffer and incubated with Alexa Fluor 488-conjugated goat antirabbit IgG1 and biotinylated anti-mouse IgG1 antibodies, followed by a Streptavidin secondary staining.
cDNA synthesis and quantitative real-time PCR Cells were harvested with TRIZOL reagent (Invitrogen), and total RNA was isolated according the manufacturer's instructions. RNA was reversely transcribed using the iScript cDNA synthesis kit (Bio-Rad). Quantitative real-time PCR (qRTPCR) analysis was performed with the SuperScript III qPCR MasterMix (Invitrogen) on the CFX 96 Real-Time PCR detection system (Bio-Rad).
Generation of bone marrow chimeric mice BM transplantation was done as described [17]. Six to eight weeks after transplantation, the reconstituted mice were sacrificed and analyzed by flow cytometry (LSRII; BD Biosciences).
In vivo cytotoxic T lymphocyte assays CTL assays were performed as previously described [17]. In brief, Hdac1 f/f and Hdac1 f/f Cd4Cre mice were immunized by injecting 50 ml OVA peptide (SIINFEKL; 2 mg/ml in PBS)/ adjuvant (CpG ODN1668; 1 mg/ml) or 50 ml PBS as a control into the footpads. After 4 days, target cells (3610 7 ) were i.v. injected into the tail vein of immunized mice. Target cells consisted of a 1:1:1 mixture of splenocytes that were either pulsed without peptide, with irrelevant peptide (VYDFFVWL, derived from murine tyrosine-related protein-2; Bachem) or with relevant OVA peptide (SIINFEKL). Splenocytes (1610 7 cells in 1 ml PBS) were pulsed with peptide (10 mg/ml) at 37u for 2 hours. Subsequently, ''empty'' splenocytes, or splenocytes pulsed with irrelevant or relevant peptide were labeled with 0.05, 0.5, and 5 mM CFSE respectively. Eight hours after target cell injection, immunized Hdac1 f/f and Hdac1 f/f Cd4Cre mice were sacrificed and the percentage of CFSE low , CSFE med and CSFE hi target cells in the draining lymph node was determined by flow cytometry. Specific killing was calculated as follows: % of specific killing = [1-(% of CFSE hi cells/% of CFSE low cells)]6100.

LCMV infections: Cells and viruses
L929 cells were purchased from the European Collection of Cell Cultures. MC57 cells and lymphocyte choriomeningitis virus (LCMV) strain Armstrong were obtained from Dr. Zinkernagel (University of Zürich, Switzerland), the latter being propagated on L929 cells. Mice were infected intravenously with 200 pfu of LCMV Armstrong. LCMV virus titers were measured on MC57 cells using a focus-forming assay.
For peptide-specific cytokine production, 10 6 splenocytes were restimulated with gp33-41 peptide in the presence of Brefeldin A (5 mg/ml) for 5 hours at 37uC. Cells were stimulated with phorbolmyristateacetate (PMA, 50 ng/ml) and ionomycin (500 ng/ml; both purchased from Sigma) as positive control or left untreated as a negative control. For intracellular staining, restimulated cells were surface-stained and fixed with cytofix-cytoperm (BD Biosciences) for 20 min. Fixed cells were incubated at 4uC for 40 minutes with permeabilization buffer (2%FCS/0.5% Saponin/PBS) containing antibodies. Samples were analyzed by flow cytometry using a FACS Canto (Becton Dickinson). Gp33 peptide (KAVYNFATC) was purchased from Neosystem (Strasbourg, France).

Statistical analysis
Unless otherwise stated the data are presented as mean 6 SEM. GraphPad Prism software was used for data analysis and plotting. The P-values were calculated with two-tailed unpaired sample Student's t-test where significance levels were set as following: *, P,0.05; **, P,0.01; ***, P,0.001.

Homeostatic analysis of CD8 + T cells in Hdac1 f/f Cd4Cre mice
We previously reported that there are no major alterations during T cell development or in the composition of peripheral T cell subsets in Hdac1 f/f Cd4Cre mice [16]. A careful re-analysis of peripheral T cell subsets revealed a slightly increased population of CD8 + T cells with a CD44 hi CD62L + memory-like phenotype ( Fig. 1A and 1B and Fig. S1A). To test whether the increase in CD44 hi CD62L + CD8 + T cells correlated also with an increase in IFNc production of CD44 hi HDAC1-null CD8 + T cells, splenocytes were isolated and stimulated ex vivo with PMA/ ionomycin. While the percentage of IFNc-producing cells within the total peripheral CD8 + T cell population was significantly enhanced in the absence of HDAC1 ( Fig. 1C and 1D), there was no difference in the percentage IFNc-producing cells within the CD44 hi CD8 + T cell subset (Fig. 1D). This indicates that CD44 hi effector CD8 + T cells are enhanced in Hdac1 f/f Cd4Cre mice, while IFNc production in CD8 effector T cells is not affected by the absence of HDAC1.
Next, we wanted to determine whether peripheral CD44 hi CD8 + T cells are enhanced due to T cell developmental defects that can be detected already at late stages during T cell maturation. There was no increase in CD44 hi cells within the mature CD8SP subsets and CD8SP cells showed a normal expression pattern of CD5, CD24 and CD69 ( Fig. S1B. Moreover, PMA/ionomycin stimulation did not indicate enhanced IFNc production ( Fig. S1C), suggesting that cytokine producing CD44 hi subsets are not enhanced within the CD8SP population.

T cell developmental defects in Hdac1 f/f LckCre mice
We previously showed that Cd4Cre mice had efficiently deleted Hdac1 alleles at the DP stage [16]. However, probably due to a slow turnover, HDAC1 protein is still detectable in DP and CD4SP thymocytes and was only fully absent in peripheral CD4 + and CD8 + T cells [16]. We confirmed these findings using intracellular HDAC1 staining approaches and detected some residual HDAC1 protein also in CD8SP thymocytes (Fig. S1D). This residual expression of HDAC1 protein might mask potential developmental defects within the CD8SP population in Hdac1 f/ f Cd4Cre mice. Therefore, Hdac1 f/f mice were crossed with the LckCre deleter strain and thymocyte development was assessed. While the relative percentage of DP cells and CD4SP thymocytes was similar between Hdac1 f/f and Hdac1 f/f LckCre mice ( Fig. 2A), the CD8SP subset was increased in Hdac1 f/f LckCre thymii ( Fig. 2A). Of note, Hdac1 f/f LckCre mice had reduced total thymocyte numbers accompanied with a reduction of DP and CD4SP cells, while CD8SP numbers were not reduced in Hdac1 f/ f LckCre mice (Fig. 2B). The assessment of CD3/TCRb expression revealed an increased fraction of CD3/TCRb lo cells within the CD8SP population in Hdac1 f/f LckCre mice (Fig. 2C). The Hdac1 f/f LckCre CD3/TCRb hi CD8SP population had a similar expression pattern of CD69, CD5 and CD24, and there were also similar percentages of CD44 hi cells within the CD8SP population (Fig. 2D). Intracellular staining revealed that HDAC1 protein was deleted in peripheral CD4 + and CD8 + T cells (Fig. 2E) and this was accompanied with a compensatory up-regulation of HDAC2 (Fig. 2E). During T cell development, HDAC1 was efficiently deleted already in DP thymocytes (Fig. 2F). To test whether the enhanced fraction of CD3/TCRb l uCD8 + cells is due to an increase in immature CD8SP thymocytes (ISP) or due to a failure of MHC class I-signaled DP thymocytes to differentiate into mature CD8SP thymocytes, we generated BM chimeric MHC class I-null mice. Hdac1 f/f or Hdac1 f/f LckCre BM cells were transferred into irradiated MHC class I-null recipient mice and T cell development was analyzed 8 weeks after the transfer. HDAC1-null BM chimeric MHC class I-null mice developed in addition to DP and CD4SP cells also a significantly enlarged fraction of CD8SP thymocytes (Fig. 2G). This indicates that the enhanced fraction of CD8SP cells in Hdac1 f/f LckCre mice is due to an increase in immature CD8SP thymocytes.
In agreement with the reduced numbers of thymocytes, the percentage and numbers of peripheral T cells were reduced in Hdac1 f/f LckCre mice, however the CD4/CD8 T cell ratio was normal ( Fig. 3A and 3B). Even though CD44 hi cells were not enhanced within the thymic CD3 hi CD8SP population (Fig. 2D), an increase in peripheral CD44 hi CD62L + CD8 + T cells was observed in Hdac1 f/f LckCre mice ( Fig. 3C and 3D), similar to the observations in Hdac1 f/f Cd4Cre mice. The CD44 hi CD62L + CD8 + T cell fraction in Hdac1 f/f LckCre mice (approx. 35%, Fig. 3D) was even more enhanced compared to Hdac1 f/f Cd4Cre mice (approx. 21%, Fig. 1A), indicating that the LckCre-mediated earlier loss of HDAC1 in comparison to Cd4Cre-mediated deletion has a stronger effect on peripheral T cell homeostasis. Ex vivo PMA/ ionomycin stimulations of peripheral T cells revealed that the percentage of IFNc-producing CD44 hi cells was increased in Hdac1 f/f LckCre mice ( Fig. 3E and 3F), indicating altered peripheral CD8 + T cell subset composition. Peripheral CD4 + T cell homeostasis was also altered in Hdac1 f/f LckCre mice, since there was a mild reduction of naïve CD44 l uCD62L + T cells within the HDAC1/2-null CD4 + T cell population ( Fig. S2A and S2B).

HDAC1-null CD8 + T cells display enhanced cytokine expression
HDAC1 is a crucial modulator of Th2 cytokine expression in CD4 + T cells [16]. To study whether HDAC1 controls the extent of cytokine expression in CD8 + T cells, we sorted naïve (CD44 l uCD62L + ) wild-type and HDAC1-null CD8 + T cells and activated them with anti-CD3/anti-CD28. In order to exclude any alterations in CD8 + T cells due to potential developmental defects upon early LckCre-mediated loss of HDAC1, CD8 + T cells were isolated from Hdac1 f/f and Hdac1 f/f Cd4Cre mice. Wild-type and HDAC1-null CD8 + T cells showed a similar proliferation rate upon activation (Fig. 4A) and produced similar levels of IL-2 (Fig. 4B) and TNFa (Fig. 4C), while IFNc production HDAC1null CD8 + T cells was slightly enhanced (Fig. 4C).
Next, we analyzed whether HDAC1 regulates CTL effector function. Runx3, T-bet and Eomesodermin are key transcriptional regulators of CTL differentiation and function [19][20][21][22][23][24]. Runx3d and Tbx21 (encoding for T-bet) were expressed at similar levels in HDAC1-null CD8 + T cells in comparison to wild-type CD8 + T  cells, while Eomes was slightly up-regulated in the absence of HDAC1 (Fig. S3). Upon activation, HDAC1-null CD8 + T cells expressed wild-type levels of GzmB (encoding for Granzyme B), while Prf1 (encoding for Perforin) expression was slightly enhanced in the absence of HDAC1 (Fig. 4D). To test whether the up-regulation of Perforin correlated with increased CTL activity in the absence of HDAC1, in vivo CTL assays were performed. This revealed that Hdac1 f/f and Hdac1 f/f Cd4Cre mice had a similar CTL activity in vivo (Fig. 4E). Together, this indicates that HDAC1 has only a minor role in the in vitro differentiation of CD8 effector T cells.

Loss of HDAC1 leads to reduced proliferation and expansion of gp33-specific CD8 + T cells in response to LCMV infection
To study the in vivo CD8 + T cell response in the absence of HDAC1 in response to a pathogen, Hdac1 f/f and Hdac1 f/f Cd4Cre were i.v. infected with LCMV (Armstrong). On day 8, the spleen and liver of infected mice were isolated and the percentage of virus-specific CD8 + T cells was determined using MHC class I tetramers specific for the viral peptides GP33 (tet-gp33). This revealed reduced numbers of tet-gp33 + CD8 + T cells in the liver and spleen in the absence of HDAC1 (Fig. 5A), which correlated with reduced percentages of INFc + and TNFa + splenic and reduced TNFa + liver HDAC1-null CD8 + T cells in comparison to wild-type CD8 + T cells upon re-stimulation of splenocytes and liver cell suspensions with gp33 (Fig. 5B). A similar tendency of reduced tet-gp33 + CD8 + T cells and reduced IFNc production was also seen on day 6 post-infection (Fig. S4). Interestingly, splenic HDAC1-null CD4 + T cells responses were also reduced upon restimulation with the MHC class II-specific peptide gp64-80 (P13) in comparison to wild-type CD4 + T cells, indicated by reduced percentages of INFc + and TNFa + CD4 + T cells in the spleen but not liver of LCMV-infected Hdac1 f/f Cd4Cre mice (Fig. 5C).
To determine whether the altered CD8 + T cell responses in Hdac1 f/f Cd4Cre mice are due to CD8 + T cell-intrinsic alteration, we crossed Hdac1 f/f Cd4Cre mice with P14 TCR transgenic mice that express TCRs (V b 8 and V a 2) specific for gp33 in the context of MHC class I [25]. Hdac1 f/f Cd4Cre,P14 T cells did not display any obvious developmental alterations and the fraction of V a 2 + T cells was similar between Hdac1 f/f and Hdac1 f/f Cd4Cre P14 tg mice (data not shown). Hdac1 f/f and Hdac1 f/f Cd4Cre,P14 CD8 + T cells were adoptively transferred into C57BL/6 mice. The next day, recipient mice were infected with 200 pfu LCMV Armstrong and on day 7 the anti-LCMV-specific CD8 + T cell response was determined upon ex vivo re-stimulation with gp33 peptide. In line with the observations in LCMV-infected Hdac1 f/f Cd4Cre mice, HDAC1-null P14 CD8 + T cells displayed a significant reduction of V b 8 + /V a 2 + CD8 + T cells (Fig. 6A), suggesting impaired expansion of HDAC1-null CD8 + T cells. Moreover, HDAC1-null P14 CD8 + T cells produced less IFNc and TNFa in comparison to transferred HDAC1-sufficient P14 CD8 + T cells (Fig. 6B). In addition, the HDAC1-null P14 CD8 + T cell population displayed lower percentages of CD44 hi and CD62L lo cells, again indicating reduced activation (Fig. 6C). To better evaluate the proliferation/ expansion properties of HDAC1-null CD8 + T cells upon in vivo infection, CFSE-labeled Hdac1 f/f and Hdac1 f/f Cd4Cre P14 CD8 + T cells (CD45.2 + ) were adoptively transferred into C57BL/6 (CD45.1 + ) mice. The next day, recipient mice were infected with LCMV and on day 5 the proliferation history of transferred P14 CD8 + T cells was determined. This revealed reduced proliferation of HDAC1-null P14 CD8 + T cells (Fig. 6D) as well as severely reduced IFNc production upon gp33 re-stimulation (Fig. 6E). Collectively, these data indicate that HDAC1 is required for an efficient in vivo expansion and activation of CD8 + T cells in response to LCMV infection.

Discussion
In this study we investigated how loss of HDAC1 affects the differentiation of effector CD8 + T cells. Our data indicate that HDAC1 is required for the efficient generation of thymocytes and peripheral T cells, for proper CD8 + T cell homeostasis and for an efficient in vivo expansion and activation of CD8 + T cells in response to LCMV infection. Thus, our study reveals an immune regulatory function of HDAC1 in CD8 lineage T cells.
We observed an increase in CD8 + T cells with a CD44 hi CD62L + effector/memory phenotype in the absence of HDAC1. Although HDAC1 was efficiently deleted in peripheral CD4 + and CD8 + T cells using the Cd4Cre delete strain [16], HDAC1 protein was still detectable in CD4SP thymocytes [16] as well as in CD8SP cells, most likely due to a slow turnover. The slowly disappearing HDAC1 protein expression in Hdac1 f/ f Cd4Cre mice makes it difficult to dissect developmental alterations during late stages of T cell development from peripheral defects. However, HDAC1 protein was absent in CD4SP and CD8SP thymocytes upon LckCre-mediated earlystage deletion. This correlated with an enhanced fraction of CD44 hi CD62L + CD8 + T cells in comparison to Hdac1 f/f Cd4Cre mice, indicating that efficient loss of HDAC1 in CD8SP thymocytes leads to developmental alterations resulting in enhanced CD44 hi subsets. However, TCRb hi HDAC1-null CD8SP thymocytes displayed a similar expression pattern of maturation markers, suggesting that positive selection is not severely affected in the absence of HDAC1.
The observation that CD44 hi cells are enhanced in Hdac1 f/ f Cd4Cre and Hdac1 f/f LckCre mice raises the question about the cellular nature of HDAC1-null CD44 hi T cells. CD44 hi T cells represent a heterogeneous population consisting of effector/ memory T cells known to be dependent on IL-15 [26,27]. In addition, CD44 hi subsets contain also innate-like T cells that are able to rapidly produce cytokines in response to activation, and the development of these cells is dependent on PLZF [28]. Preliminary results with Hdac1 f/f LckCre mice either carrying the luxoid mutation of PLZF (Plzf lu/lu ), which leads to an inactivated form of PLZF [29] or with IL-15-null mice [27] indicated that there was a similar relative reduction in the percentage of CD44 hi cells in the absence of either PLZF or of IL-15 between Hdac1 f/f and Hdac1 f/ f LckCre mice (Fig. S5). Thus, it is unlikely that the enhanced CD44 hi population in Hdac1 f/f LckCre mice is due to the preferential development/expansion of a PLZF-or an IL-15dependent CD8 + T cell subset.
Hdac1 f/f LckCre mice had reduced numbers of total thymocytes as well as peripheral T cells. This is in contrast to a previous study, which reported similar numbers of thymocytes in Hdac1 f/f LckCre mice in comparison to wild-type mice [6]. However, the other study did not investigate T cell development in much detail and peripheral T cell subsets were not analyzed. Finally, we observed that the population of TCRb lo immature CD8 single-positive thymocytes was increased among Hdac1 f/f LckCre thymocytes. Together, our data strongly indicate that LckCre-mediated deletion of HDAC1 affects T cell development and peripheral T cell homeostasis.
Another finding of our study is the observation that in vitro CTL differentiation and in vivo CTL activity is largely unaffected in the absence of HDAC1. We observed a compensatory upregulation of HDAC2 in CD8 + T cells similar to reports for other cell lineages such as CD4 + T cells [16], B cells [30], during erythrocyte-megakaryocyte differentiation [31] and in non-hematopoietic cells [32][33][34][35]. Thus it is likely that this closely related class I HDAC family member partially compensates for loss of HDAC1 and thus masks potential functional consequences of HDAC1 deletion. However, HDAC1 was essential for a proper anti-LCMV response in vivo. LCMV-infected Hdac1 f/f Cd4Cre mice displayed reduced numbers of virus-specific tet-gp33 + CD8 + T cells as well as reduced cytokine production upon re-stimulation, although the number of virus-specific tet-np396 + T cells and the clearance of LCMV virus was not altered (data not shown). This suggests that HDAC1-null CD8 + T cells might react differentially to the signaling strength/quality of different viral-peptide-MHC complexes. The reduced anti-gp33 + CD8 + T cell response was not due to impaired CD4 + T cell responses in the absence of HDAC1, since Hdac1 f/f Cd4Cre P14 TCR transgenic CD8 + T cells that were adoptively transferred into wild-type mice displayed a similar reduction in the percentage of anti-viral gp33-specific CD8 + T cells as well as cytokine production in comparison to transferred wild-type P14 CD8 + T cells. Moreover, CFSE labeling experiments combined with adoptive transfer approaches indicated reduced proliferation, suggesting that virus-specific tet-gp33 + T cells are reduced due to impaired expansion rather than increased apoptosis. Collectively, these data indicate that HDAC1 is required for an efficient in vivo expansion and activation of CD8 + T cells in response to LCMV infection. . On day 8, spleen and liver were isolated and cells were analyzed. The percentage of viral-specific CD8 + T cells was determined using MHC class I tetramers specific for the viral peptides gp33 (tet-gp33). Diagram shows the percentage of tet-gp33 + CD8 + T cell populations isolated from spleen and liver of Hdac1 f/f and Hdac1 f/f Cd4Cre mice. Mean 6 SD is shown (n = 3, analyzed in 1 experiment). (B) Mice were infected as described in A. On day 8, spleens and livers were isolated and cell suspensions were re-stimulated with gp33 peptide for 5 hours. IFNc and TNFa expression was determined by intracellular cytokine staining. The percentage of INFc + or of TNFa + producing CD8 + T cells is shown (mean 6 SEM; n = 3, analyzed in 1 experiment). (C) Mice were infected as described in A. On day 8, spleen and liver were isolated and cell suspensions were re-stimulated with the MHC class II-specific peptide gp64-80 (P13) for 5 hours. IFNc and TNFa expression was determined by intracellular cytokine staining. The percentage of INFc + and TNFa + producing CD4 + T cells is shown (mean 6 SEM; n = 3, analyzed in 1 experiment). (A-C) Statistical analysis was performed using a two-tailed non-paired Student's t test. The P-values were defined as following: *, P,0.05; **, P,0.01; ***, P,0.001. doi:10.1371/journal.pone.0110576.g005  4 h and IFNc production was analyzed by intracellular cytokine staining. Mean 6 SEM is shown (n = 4, performed in 2 independent experiments). Cells were gated on CD8SP thymocytes. (D) Intracellular HDAC1 levels in Hdac1 f/f (solid grey) and Hdac1 f/f Cd4Cre (black line) CD8SP thymocytes. Data shown are representative of 2 independent experiments. (A, C) Statistical analysis was performed using a two-tailed non-paired Student's t test. The P-values were defined as following: *, P,0.05; **, P, 0.01; ***, P,0.001; n.s., not significant. (TIF) Figure S2 The distribution of naïve and memory CD4 + T cells in Hdac1 f/f LckCre mice. (A) CD44/CD62L expression on peripheral Hdac1 f/f and Hdac1 f/f Cd4LckCre CD4 + T cells. Numbers in the plots indicate the percentage of cells in the respective quadrants. (B) The summary of the percentage of CD44 l uCD62L + , CD44 hi CD62L + and CD44 hi CD62L -Hdac1 f/f (n = 6) and Hdac1 f/f LckCre (n = 7) CD4 + T cells is shown (mean 6 SEM; performed in 4 independent experiments). Statistical analysis was performed using a two-tailed and non-paired Student's t test. The P-values were defined as following: *, P, 0.05; **, P,0.01; ***, P,0.001; n.s., not significant. (TIF) Figure S3 Expression of key transcription factors in Hdac1 f/f Cd4Cre CD8 + T cells. Naïve Hdac1 f/f and Hdac1 f/ f Cd4Cre CD8 + T cells were stimulated with anti-CD3/anti-CD28 for 48 hours. Cells were split 1:2 on day 2, cultured for 2 additional days and re-stimulated with anti-CD3 overnight. The expression of Tbx21, Runx3d and Eomes was assessed by qRTPCR before (''resting'') and after (''react.'') overnight restimulation with anti-CD3. Expression was normalized to Hprt1 expression (mean 6 SEM; n = 3; performed in 3 independent experiments). Statistical analysis was performed using a two-tailed and non-paired Student's t test. The P-values were defined as following: *, P,0.05. (TIF) Figure S4 Hdac1 f/f Cd4Cre mice show impaired CD8 + T cell responses upon LCMV infection. (A) Hdac1 f/f and Hdac1 f/f Cd4Cre mice were infected i.v. with 200 pfu LCMV (Armstrong). On day 6, spleen and liver were isolated and cells were analyzed. The percentage of viral-specific CD8 + T cells was determined using MHC class I tetramers specific for the viral peptides gp33 (tet-gp33). Diagram shows the percentage of tet-gp33 + CD8 + T cell populations isolated from spleen and liver of Hdac1 f/f and Hdac1 f/f Cd4Cre mice. Mean 6 SD is shown (n = 4, analyzed in 1 experiment). (B) Mice were infected as described in A. On day 6, spleens and livers were isolated and cell suspensions were re-stimulated with gp33 peptide for 5 hours. IFNc and TNFa expression was determined by intracellular cytokine staining. The percentage of INFc + or of TNFa + producing CD8 + T cells is shown (mean 6 SEM; n = 4, analyzed in 1 experiment, except for IFNc + in the spleen where n = 3). (A, B) Statistical analysis was performed using a two-tailed non-paired Student's t test. The Pvalues were defined as following: *, P,0.05; **, P,0.01; ***, P, 0.001. (TIF)