GIMAP5 Deficiency Is Associated with Increased AKT Activity in T Lymphocytes

Long-term survival of T lymphocytes in quiescent state is essential to maintain their cell numbers in secondary lymphoid organs. In mice and in rats, the loss of functional GTPase of the immune associated nucleotide binding protein 5 (GIMAP5) causes peripheral T lymphopenia due to spontaneous death of T cells. The underlying mechanism responsible for the disruption of quiescence in Gimap5 deficient T cells remains largely unknown. In this study, we show that loss of functional Gimap5 results in increased basal activation of mammalian target of rapamycin (mTOR), independent of protein phosphatase 2A (PP2A) or AMP-activated protein kinase (AMPK). Our results suggest that the constitutive activation of the phosphoinositide 3-kinase (PI3K) pathway may be one of the consequences of the absence of functional GIMAP5.


Introduction
The GTPase of immune-associated protein (Gimap) || family is predominantly expressed in lymphocytes and cells of the hematopoietic system [1,2]. In diabetes prone Bio-Breeding (BB-DP) rats, the lyp allele arises from a frame-shift mutation within the Gimap5 gene that deletes 223 amino acids at the C-terminus [3,4]. The life span of T cells is reduced in the periphery of Gimap5 lyp/lyp rats resulting in a profound T lymphopenia in the secondary lymphoid organs [5][6][7]. Two independently generated lines of Gimap5 deficient mice also exhibit progressive loss of T cell populations [8,9]. Whereas the cell survival defect is confined to T cells in Gimap5 lyp/lyp rats, mice lacking Gimap5 show defects in various hematopoietic cell types including a breakdown of quiescence in hematopoietic stem cells [8][9][10]. Despite a decade of efforts by several groups, mechanisms underlying the pro-survival function of GIMAP5 remain unclear.

T cell preparation and stimulation
CD4 + T cells were used for all the experiments as the paucity of CD8 + T cells in Gimap5 mutant rats and mice precludes their isolation in sufficient numbers to carry out any experiment. Single-cell suspension from pooled lymph nodes (inguinal, axillary, brachial, mesenteric and cervical) of mice or rats were suspended in PBS supplemented with 2% FCS. CD4 + T cells were purified by negative selection using magnetic beads following manufacturer's instructions (Dynabeads, Life Technologies). T cells were stimulated by crosslinking with the anti-CD3 and or anti-CD28 mAb as required. Goat anti-hamster or anti-mouse antibodies were used as cross-linking agents for T cells from mice and rats, respectively. OT-II CD4 + T cells were stimulated with OVA323-339 peptide (0.5μM) presented by irradiated splenocytes that served as antigen-presenting cells (APC) for the indicated duration. In experiments using inhibitors, cells were pre-incubated in the presence or absence of the inhibitors in complete RPMI containing calcium for 30 min at 37°C. The vehicle control was PBS or DMSO that was diluted to the same concentration. The concentration of DMSO did not exceed 0.1%. Purified CD4 + T cells obtained from a single Gimap5 mutant rat or mouse are not sufficient to carry out analyses of signaling pathways. Therefore, lymph nodes were pooled from two Gimap5 sph/sph mice for every Gimap5 +/+ mouse and from three Gimap5 lyp/lyp rats for every control rat in any given experiment. Pooled samples were prepared for every repeat experiment. In all experiments, mice and rats were used at 4 weeks of age. Animals were sex-matched within a given experiment, even though both males and females were used.

PP2A activity assay
The PP2A protein phosphatase activity of CD4 + T cellular lysates was determined by measuring the generation of free phosphate from the threonine phospho-peptide using PP2A immunoprecipitation phosphatase assay kit. Cell lysates were prepared in lysis buffer (20 mM imidazole-HCL, 2 mM EDTA, 2 mM EGTA, PH7.0 with protease inhibitor cocktail). Five hundred micrograms of total cell lysate was immunoprecipitated with anti-PP2A antibody. PP2A activity in the immunoprecipitates was determined using 750 μM phospho-peptide as substrate. After 10 minutes of incubation at 30°C malachite dye was added, and free phosphate was measured by the relative absorbance at 650 nM.

Measurement of Oxygen consumption rates (OCR)
CD4 + T cells were stimulated overnight with plate-bound anti-CD3/CD28 and oxygen consumption rate (OCR) (in pmol/min) was measured using Seahorse XF-96 metabolic extracellular flux analyzer (Seahorse Bioscience). Briefly, CD4 + T cells were re-suspended in XF assay buffer and were plated onto Seahorse cell plates (6 × 10 5 cells per well). T cell attachment was enhanced by centrifugation. Cells were incubated at 37°C in a CO 2 -free incubator for one hour prior to measurements. Then OCR was measured under basal conditions and following addition of oligomycin (1 μM), FCCP (1.5 μM) and rotenone (100 nM) + antimycin A (1 μM).

ATP assay
The cellular levels of ATP were measured using a bioluminescence assay kit according to the manufacturer's instructions. Briefly, rat CD4 + T cells were untreated or stimulated with antibodies to CD3 and CD28 for the indicated times. The cells were collected and lysed in lysis buffer. The lysates were centrifuged at 13,000×g for 15 min at 4°C and supernatants were collected. ATP concentration was measured in 0.5 μg protein lysate. The luminescence was measured in a Luminometer by the absorbance at 560nM.

SDS-PAGE and Western blot
Cells were washed with cold PBS, harvested and re-suspended in SDS-PAGE sample buffer (50 mM Tris pH 6.8, 1% (w/v) SDS, 1 mM EDTA, 1 mM dithiothreitol). Equivalent amounts of proteins were separated in SDS-PAGE gels and transferred to polyvinylidene difluoride membranes. The blots were probed with primary antibodies followed by incubation with the appropriate HRP-conjugated secondary antibodies and developed with ECL reagent. After incubating in stripping solution (2% SDS, 62.5 mM Tris pH 6.8, 100 mM 2-mercaptoethanol) for 30 min at 55°C, the blots were blocked and re-probed for total proteins.

Statistical analysis
Data were analyzed using two-way ANOVA for statistical comparisons. p values less than 0.05 were considered statistically significant.

Gimap5-deficient T cells show constitutive mTORC1 activity
Despite the inability of CD4 + T cells from mice and rats with mutations in the Gimap5 gene to undergo robust proliferation in response to cognate antigens, as well as polyclonal stimulations such as cross-linking CD3 and CD28 or a combination of PMA and ionomycin, these cells were capable of differentiating into Th17 cells [31,36]. As differentiation into Th17 cells is associated with the activation of the mTORC1 complex [32][33][34], we assessed the mTORC1 pathway by analyzing the phosphorylation of its canonical substrates. Phosphorylation of 4EBP1 releases eIF4E to initiate cap-dependent translation [37]. Phosphorylation of S6K activates its kinase activity that results in the phosphorylation of ribosomal protein S6, a component of the 40S ribosome, which is one of the most characterized substrates of S6K [38]. Following crosslinking CD3/CD28 on the surface of CD4 + T cells, phosphorylation of the S6 peaked between 2 and 10 minutes in controls ( Fig 1A). To our surprise, we observed that even in the absence of any stimulation, S6 and 4EBP1 were heavily phosphorylated in CD4 + T cells from Gimap5 sph/ sph mice ( Fig 1A). It is possible that the polyclonality and plasticity of the T cell repertoire could have skewed the distribution of the T cell subsets that are present in the peripheral pool of Gimap5 sph/sph mice. Therefore, we assessed the expression of pS6 and p4EBP1 in CD4 + T cells from OTII transgenic control and Gimap5 sph/sph mice (OTII cells) following stimulation with OVA 323-339 presented by irradiated APCs. OTII cells from control mice showed a gradual increase in pS6 protein over a period of 3 hours. However, similar to polyclonal T cells, OTII cells from Gimap5 sph/sph mice showed elevated basal levels of expression of pS6 with minimal increase over time ( Fig 1B). The increased phosphorylation of S6 was not unique to Gimap5 sph/ sph mice as loss of functional GIMAP5 resulted in increased phosphorylation of S6 and 4EBP1 in Gimap5 lyp/lyp rats as well ( Fig 1C). Thirty minutes after stimulation with anti-CD3/CD28, the band corresponding to pS6 was reduced in both control and Gimap5 sph/sph T cells ( Fig 1A). These observations suggest that, even if pS6 is constitutively activated, pathways leading to TCR-induced de-phosphorylation of pS6 are not affected by Gimap5 deficiency.
As phosphorylation of 4EBP1 and S6 is mediated by signals from mTORC1, we addressed the role of mTORC1 in this phosphorylation. To this end, purified CD4 + T cells from control and Gimap5 sph/sph mice were pre-treated with rapamycin, an inhibitor of mTORC1 activity. Pre-treatment with rapamycin inhibited the constitutive and TCR-induced phosphorylation of S6 and 4EBP1 in CD4 + T cells from control and Gimap5 sph/sph mice (Fig 1D and 1E). Similarly, the constitutive phosphorylation of S6 and 4EBP4 observed in T cells from Gimap5 lyp/lyp rats was also diminished by treatment with rapamycin ( Fig 1F), indicating that the increased mTORC1 activity observed in CD4 + T cells, as a consequence of the loss of GIMAP5 functions, is comparable across species.
Activation of AMPK is not altered by Gimap5 mutation T cell quiescence also requires the suppression of the mTORC1 complex by signals from the LKB1/AMPK pathway [20]. In quiescent T cells catabolic LKB1/AMPK pathway promotes mitochondrial oxidative phosphorylation (OXPHOS) and autophagy to meet the energy requirements. Cytosolic Ca 2+ influx that follows TCR cross-linking can also transiently activate AMPK to increase the ATP production in anticipation of cell growth [39]. Therefore, we assessed the AMPK status in T cells at rest and following Ca 2+ influx from intracellular stores. We have shown previously that Ca 2+ release from the endoplasmic reticulum (ER), following TCR cross-linking, was not affected in T cells from Gimap5 lyp/lyp rats [29]. To mimic the TCRinduced Ca 2+ release from the ER we used thapsigargin (TG), a pharmacological inhibitor of the sarco/endoplasmic reticulum Ca 2+ -ATPase (SERCA) pump that is involved in the re-filling of the ER Ca 2+ stores. As the analysis of AMPK signaling was not consistent after stimulation through CD3/CD28, we used TG to induce Ca 2+ influx. As shown in Fig 2 we did not observe any difference between T cells from control and Gimap5 deficient rats or mice in the basal or Ca 2+ -induced phosphorylation of AMPK following the addition of TG. Furthermore, the pattern of phosphorylation of ACC, the downstream substrate of AMPK was also comparable between the control and Gimap5 mutants in both the rats and in mice. These results suggest that Gimap5 does not influence the activation of the AMPK pathway in T cells.
Energy requirements of quiescent cells are mostly met from catabolic processes involving mitochondrial oxidative respiration [40]. As Gimap5 mutation results in the loss of (A and C) CD4 + T cells from control and Gimap5 deficient mice (A) and rats (C) were left un-stimulated or stimulated with 5 μg/μL anti-CD3 or anti-CD3/CD28 for the indicated period. Phosphorylation and the total amount of the indicated proteins in the whole cell lysates were evaluated by Western blot. (B) CD4 + T cells from the OTII TCR-transgenic control and Gimap5 sph/sph mice were co-cultured with irradiated APC pulsed with indicated period were lysed and analyzed by Western blot. (D and F) CD4 + T cells from control and Gimap5 deficient mice (D) and rats (F) were treated with vehicle or 200 nM rapamycin for 30 min followed by TCR cross-linking at different time points. Cell lysates were examined by Western blot using the indicated antibodies. (E) CD4 + T cells from the OTII Gimap5 sph/sph and control mice were pretreated with vehicle or 200 nM rapamycin for 30 min before simulation with ova peptide for the indicated duration. Cell lysates were probed with specific antibodies. (A-F) Representative data from 3 independent experiments are shown. Histograms show densitometric data from 3 experiments. * p<0.05 control vs mutant cells. doi:10.1371/journal.pone.0139019.g001 Spontaneous AKT Activation in Gimap5-Deficient T Cells mitochondrial membrane potential over time and reduced ability of mitochondria to buffer cytosolic Ca 2+ [30], we assessed whether CD4 + T cells from Gimap5 sph/sph mice showed alterations in OXPHOS. We assessed mitochondrial respiration using an extracellular flux analyzer. Here, alterations in OXPHOS are reflected in the amount of oxygen consumed by the mitochondria, expressed as oxygen consumption rate (OCR). Again, we did not observe any difference in the OCR between T cells from control and Gimap5 sph/sph mice (Fig 3A). Similarly, mutation in Gimap5 did not affect the energy production in resting T cells or following cross-linking of CD3/ CD28 as measured by the total ATP content in these cells (Fig 3B). Taken together, these observations suggest that Gimap5 does not directly influence the energy content in T cells.

PP2A phosphatase activity is not altered in Gimap5-deficient T cells
The constitutive phosphorylation of S6 and 4BP1 in T cells from Gimap5 mutant mice and rats could also arise from suboptimal phosphatase activity. PP2A, a serine/threonine protein phosphatase, is the major phosphatase responsible for the de-phosphorylation of S6 and 4EBP1 [41]. An earlier study had shown that apoptosis induced by okadaic acid (OA), an inhibitor of the PP2A could be prevented by transfection with human GIMAP5 (previously called IAN5) gene [42]. As a corollary, we have also observed that Gimap5 lyp/lyp T cells lacking functional Gimap5 showed increased sensitivity to OA-induced apoptosis (S1 Fig). We confirmed the role of PP2A in dephosphorylating S6 and 4EBP1 by treating the cells with calyculin A, an inhibitor of PP2A [43]. As expected, treatment with calyculin A resulted in an increased phosphorylation of S6, 4EBP1 and AKT (S2 Fig). To determine whether the constitutive phosphorylation of S6 and 4EBP1 could be the result of decreased PP2A activity, we measured the activity of PP2A in the lysates of CD4 + T cells from Gimap5 lyp/lyp rats. Under steady state conditions, the activity of PP2A was comparable in CD4 + T cells from Gimap5 deficient and control rats (Fig 4). These observations suggest that the persistence of the phosphorylation of S6 and 4EBP1 cannot be attributed to decreased phosphatase activity, per se.
Gimap5 deficiency results in constitutively active AKT Cross-linking of TCR/CD3 complex results in the recruitment of PI3K to the CD3 complex, to generate phosphatidylinositol (3,4,5)-triphosphate (PIP3) at the plasma membrane [44]. Phosphoinositide-dependent kinase-1 (PDK1) and AKT are recruited by PIP3 through their pleckstrin homology (PH) domains leading to the phosphorylation of AKT at Thr308, followed by the phosphorylation of Ser473 by the mTORC2 complex [45]. Activated AKT phosphorylates the TSC1/TSC2 complex, resulting in the inhibition of its GTPase-activating protein (GAP) activity and releasing the RHEB GTPase to activate mTORC1 [46]. We assessed whether the increased phosphorylation of AKT following cross-linking of TCR/CD3 complex and CD28 could explain the increased mTOR activity. In CD4 + T cells from control mice, we observed that the pAKT peaked between 2 and 10 minutes after CD3/CD28 cross-linking, and returned to the basal level by 30 minutes (Fig 5A). In contrast, phosphorylation of AKT was detected even in the absence of stimulation in CD4 + T cells from Gimap5 sph/sph mice (Fig 5A). Crosslinking CD3/CD28 also increased the pAKT that returned to basal level within 30 minutes in Gimap5 sph/sph CD4 + T cells. But it was still higher than that observed in the control cells. Following activation of control OTII cells with peptide/APC, phosphorylation of AKT was evident at 5 minutes and was strong even at 180 minutes ( Fig 5B). However, in Gimap5 mutant OTII cells, pAKT was detected even in the absence of stimulation and showed minimal increase over time (Fig 5B). Increased basal AKT phosphorylation was also observed in CD4 + T cells from Gimap5 lyp/lyp rats (Fig 5C). These results strongly indicated that loss of functional GIMAP5 protein results in elevated basal phosphorylation of AKT in T cells. A previous report has shown that the expression of FOXO1, the down stream target of AKT, is reduced in CD4 + T cells from 4-week old Gimap5 sph/sph mice [31]. We observed that the phosphorylation of FOXO1 is comparable or lower in Gimap5 mutant CD4 + T cells even in the presence of hyperactivated Akt (Figs 5 and 6). Furthermore, to confirm that the activation of AKT was not an artifact, we determined the phosphorylation status of AKT and FOXO1 in the cells where mTOR was inhibited. While phosphorylation downstream of mTORC1 was inhibited following pre-treatment with rapamycin ( Fig 1D, 1E and 1F), phosphorylation of AKT and FOXO1 remained unchanged (Fig 5D and 5E). These results indicate that the constitutive activation of AKT likely underlies the activation of the mTORC1 pathway in T cells lacking functional GIMAP5 protein.
Constitutively activated PI3K is responsible for the increased pAKT in GIMAP5 deficient T cells Engagement of the TCR or IL-7R activates PI3K to generate PIP3 at the plasma membrane which results in the recruitment and activation of AKT [47]. To determine whether the phosphorylation of AKT was secondary to the activated state of PI3K, we pretreated the cells with different inhibitors of PI3K. LY294002 (LY) and CAL-101 markedly diminished the intensity of the pAKT in CD4 + T cells from control and Gimap5 sph/sph mice and Gimap5 lyp/lyp rats (Fig 6  and S3 Fig). Phosphorylation of AKT following activation of OTII cells with ova peptide was also completely prevented by CAL-101 (Fig 6D). More importantly, inhibition of PI3K greatly Spontaneous AKT Activation in Gimap5-Deficient T Cells diminished the amount of constitutive pAKT, indicating that aberrant activation of PI3K may be the underlying cause for the activation of the AKT/mTOR pathway. In addition, the level of pFOXO1, an mTORC1-independent substrate of pAKT was also reduced following treatment with PI3K inhibitors (Fig 6). The pattern of inhibition between LY and CAL-101 was comparable in control and mutant cells (Fig 6A versus 6C and Fig 6B versus 6D). While LY is pan PI3K inhibitor, CAL-101 is specific to PI3Kdelta isoform that is active in lymphocytes [44]. Thus, it is reasonable to conclude that the constitutive phosphorylation of AKT is essentially mediated by PI3K delta in the T cells from Gimap5 mutant mice.

Discussion
Long-term survival of T lymphocytes in a quiescent state is essential to maintain their cell numbers in the peripheral circulation and in secondary lymphoid organs. Even though cytokines, T cell antigen receptor and various signaling pathways are implicated in the maintenance of T cell survival, the processes by which these signaling pathways are integrated are not yet well understood. In BB-DP rats, T lymphopenia, that arises due to the recessive mutation in the lyp allele, promotes abortive homeostatic expansion of T cells resulting in autoimmune diabetes [6] (reviewed in [7]). As a consequence of this mutation, single positive thymocytes and mature T lymphocytes undergo spontaneous apoptosis [48,49]. Twenty years after its discovery, the mutation was mapped to the Gimap5 locus (formerly Ian5) leading to the discovery of a new family of proteins [3,4]. A GTPase domain in the N-terminus that can potentially hydrolyze GTP characterizes GIMAP proteins. Some of these possess a C-terminal hydrophobic trans-membrane domain, which targets GIMAP5 to the endo-lysosomal compartment [50]. While in rats the functional consequence of Gimap5 mutation is restricted to the T cell lineage, in mice it extends to other hematopoietic cells and prevents the survival of hematopoietic stem cells [8,9].
In Gimap5 lyp/lyp rats the frame-shift mutation truncates the coding sequence at amino acid 109 such that the putative polypeptide product lacks both a functional GTPase domain and the C-terminal membrane anchor. [3,4]. In Gimap5 sph/sph mice a point mutation disrupts the functional GTPase domain. While the mRNA is expressed, the protein is not detected, probably due to its lack of stability [9]. These observations suggest that the GTPase domain of GIMAP5 protein may be functional. Studies with GIMAP2 indicate that the GTPase domain in GIMAP2, 7 and 5 can bind GTP and can potentially hydrolyze GTP when they form dimers [51]. GIMAP5 has comparable functions, at least in T cells, in both rats and mice as observed in this study where we have carried out most of the experiments in both Gimap5 mutant rats and mice.
Most of the GIMAP proteins are expressed in the cells of the hematopoietic system [52]. However, the mechanisms by which GIMAP proteins promote the survival of T and B lymphocytes are far from well understood. In mice, GIMAP5 has been reported to interact with the members of the Bcl-2 family [2,10]. However, we did not observe co-localization of Bcl-2 with rat GIMAP5 in primary rat T lymphocytes or following overexpression [35]. Our observations rather suggest that GIMAP5 may be involved in the integration of the survival signals generated through the TCR by regulating the calcium homeostasis in T lymphocytes [29]. In fact the Spontaneous AKT Activation in Gimap5-Deficient T Cells TCR-induced release of Ca 2+ from the ER stores is normal in GIMAP5 deficient T cells suggesting that certain aspects of the proximal TCR signaling machinery is intact in these cells [29]. The reduced Ca 2+ influx may possibly be attributable to the inability of the mitochondria to move on microtubules to buffer the cytosolic Ca 2+ in the absence of functional GIMAP5 protein [30]. A similar function has been recently reported for GIMAP4 that interacts with actin and tubulin for cytokine secretion in CD4 + T cells [53]. These observations collectively raise the possibility that the survival defect in these Gimap5 mutant T cells can also, in part, be due to defects in vesicular traffic.
Despite the lymphopenia and the defective proliferative response following TCR stimulation, Gimap5-deficient T cells do differentiate into Th17 cells and contribute to tissue pathologies [9,36,54]. T cell-specific deletion of regulatory associated protein of mTOR (RAPTOR), the mTOR binding partner that also binds p70S6k and 4E-BP1, prevents the down-regulation of Gfi1, a negative regulator of Th17 differentiation [55]. Our observations show that this skewing towards Th17 differentiation in Gimap5-deficient T can be attributed to the increased mTORC1 signaling (Fig 1). As a consequence, Gimap5-deficient T cells show a decrease in the Treg population, which could also aggravate the observed pathologies [9,56]. However, it is not clear if the observed phenotypes are secondary to lymphopenia. Competitive reconstitution leads to the loss of the mutant T cells over time in rats, making it difficult to address the role of lymphopenia [49]. However, reconstitution with Gimap5-sufficient cells greatly reduced the development of autoimmune diabetes in the rat model and the intestinal lesions in the mouse model [7,9], suggesting that the manifestation of the pathologies associated with the Gimap5 mutation are secondary to the lymphopenia-induced activation.
Quiescence in T cells also requires the repression of mTORC1 pathway by the LKB1/AMPK axis that promotes catabolic pathways involving mitochondrial OXPHOS [20]. The activation of the mTORC1 pathway is not due to defects in AMPK activation as seen from the equivalent activation of AMPK and the mitochondrial OXPHOS in Gimap5 deficient T cells (Figs 2 and  3). Rather, Gimap5-deficient T cells show spontaneous activation of AKT, which then activates the mTORC1 pathway through the suppression of the TSC complex (Fig 5). The constitutive activation of AKT appears to be specific to T lymphocytes as it is not observed in B cells, even though their survival is also affected in the absence of GIMAP5 [9]. Thus, activation of the mTORC1 pathway can be attributed to the spontaneous activation of the PI3K/AKT axis.
T cells require two signals to sustain T cell activation, cytokine production and cell proliferation [57]. The primary signal through the TCR needs to be accompanied by a second signal through the co-stimulatory molecules. Signals through the TCR complex alone, in the absence of co-stimulation, result in the development of anergic T cells that cannot easily respond to robust re-stimulation through the TCR [58]. Recently, two groups have reported that spontaneous gain of function mutations in PI3K resulted in profound immunodeficiency, as these T cells were unable to mount an efficient immune response to cross-linked anti-CD3/CD28 stimulation [59,60]. Similarly, here we show that as a consequence of the mutation in Gimap5, the PI3K/AKT axis is constitutively activated in the absence of spontaneous activation of TCRmediated proximal signaling pathways. It has been shown that mutations in PI3K can alter the signaling threshold during T cell development [61]. However analysis of the phenotype of thymocytes in Gimap5 mutant rats and mice suggest that GIMAP5 is required for T cell survival Spontaneous AKT Activation in Gimap5-Deficient T Cells during later stages of T cell development in the thymus and in the periphery. We hope that the dissection of the signaling pathways in GIMAP5-deficient cells will eventually contribute to the understanding of the mechanisms through which GIMAP5 promotes the survival of lymphocytes.
Supporting Information S1 Fig. Gimap5 lyp/lyp T cells are more sensitive to OA induced apoptosis. Total lymphocytes from control and Gimap5 lyp/lyp rat were cultured with or without 200 nm OA for 8 h. The apoptosis in gated CD4 + lymphocytes was analyzed by annexin V staining using flow cytometry. The number of cells recovered from cultures of lymphocytes Gimap5 lyp/lyp rats diminishes with time. As a consequence only few CD4 + cells can be acquired in the live gate. Representative data from 3 independent experiments are shown.