Naïve and Innate Memory phenotype CD4⁺ T-cells have different requirements for active Itk for their development

Abstract

The Tec family kinase Itk regulates the development of conventional and innate CD8⁺ T cells, however little is known about the role of Itk in the development of CD4⁺ T cell lineages, although the role of Itk in the T cell activation and function is well defined. We show here that the Itk null mice have increased percentage of CD62L^LoCD44^Hi memory phenotype CD4⁺ T cells compared to WT mice. These cells arise directly in the thymus, express high levels of transcripts for the T-bet and IFNγ and are able to produce IFNγ directly ex vivo in response to stimulation. Itk deficiency greatly decreases the number of CD4⁺ T cells with CD62L^HiCD44^Lo naïve phenotype, but has no effect on the number of memory phenotype CD4⁺ T cells, indicating that the development of memory phenotype CD4⁺ T cells is Itk independent. We further show that the development of the naïve phenotype CD4⁺ T cells is dependent on active Itk kinase signals and can be rescued by expression of Itk specifically in T cells. Our data also show that the Itk is required for functional TCR signaling in these cells, but not for the innate function in response to IL-12/IL-18 or L. monocytogenes stimulation. These results indicate that CD62L^HiCD44^Lo “naïve” and CD62L^LoCD44^Hi “innate memory phenotype” CD4⁺ T cells may be independent populations that differ in their requirement for Itk’s signals for development. Our data also suggest that CD4⁺CD62L^LoCD44^Hi memory phenotype T cells have innate immune function.

Introduction

Mature CD4⁺ SP and CD8⁺ SP T cells arise from CD4⁺CD8⁺ DP T cell precursors in the thymus. The development of CD4⁺ and CD8⁺ T cells can be influenced by the strength and duration of signals received through the TCR in DP thymocytes (1, 2). In addition to conventional T cell lineages, DP thymocytes also give rise to some other lineages of mature T cells, such as T_reg and NKT cells (3-5), which are called non-conventional T cells. Studies of many knockout mice have identified proteins required for the T_reg and NKT cells, but less is known about signaling pathways leading to the specific development of CD4⁺ and CD8⁺ T cells.

Itk is the predominant Tec kinase expressed in T cells and is activated downstream of the T cell receptor (TCR) (6-12). Specifically, Itk seems to act as an amplifier of TCR signals, and is required for the full activation of PLCγ-1, Ca²⁺ mobilization, and activation of transcription factors such as NFAT, NF-κB, and AP-1. These transcription factors activate a number of genes, including cytokine and other genes involved in cytokine signaling, survival and differentiation (8-10, 13, 14). Thus, Itk can affect multiple processes important for T cell development, activation, and effector function (8). These affected processes include impaired positive selection (15, 16), altered CD4/CD8 commitment, defects in TCR induced proliferation, IL-2 production and reduced activation induced cell death (AICD) in the absence of Itk (8-10, 17).

We and others have recently shown that Itk is required for the development of conventional CD44^Lo “naïve” CD8⁺ T cells but not “innate memory phenotype” CD44^Hi CD8⁺ T cells (18-21). In the absence of Itk, CD8⁺ T cells resemble activated/memory cells, express memory markers, carry high levels of preformed messages for IFNγ and T-bet, and rapidly produce IFNγ ex vivo in response to stimulation (18, 19, 21). These CD8⁺ T cells developed as a result of interaction with MHC molecules expressed on haematopoietic cells in the thymus (18). All these properties suggest that these cells share properties with innate T cells such as H2-M3 specific CD8⁺ T cells, MAIT cells and CD1d specific NKT cells (5, 22-24). Although conventional CD8⁺ T cell development is abolished in the absence of Itk, the development of CD4⁺ T cell lineage seems to less affected (18, 19), suggesting that Itk may play different roles in the development of CD4⁺ and CD8⁺ T cell lineage. Here, we report that a higher percentage of CD4⁺ T cells in Itk null mice have a CD62L^LoCD44^Hi memory phenotype (MP) and show effector function ex vivo in response to stimulation. Itk deficiency greatly decreases the number of CD4⁺ T cells with naïve phenotype (NP), but has no effect on the numbers of the MP CD4⁺ T cells. We also specifically show that active Itk signaling is required for these effects, and that Itk is required for function through the TCR signaling of these cells, but not for the innate function in response to IL-12/IL-18 or L. monocytogenes stimulation. These results indicate that CD62L^HiCD44^Lo “naïve” and CD62L^LoCD44^Hi “memory” CD4⁺ T cells may include independent populations that differ in their requirement for Itk’s signals for development.

Materials and methods

Mice

WT and Itk^-/- mice used were 2 days to greater than 12 months of age and were kept in specific pathogen free conditions. Tg(Lck-ItkΔKin)Itk^-/- mice were generated previously in our lab and were backcrossed to the C57BL/6 background >10 generations (21). Tg(CD2-hItk)Itk^-/- mice were generated by cloning a human Itk cDNA into a transgenic expression cassette driven by the CD2 promoter and CD2 enhancer. These mice were backcrossed >5 generations. In both cases, the expression level of the transgene was roughly 30% of endogenous Itk as determined by quantitative RT-PCR. All mice were on the C57BL/6 background. All experiments were approved by the IACUC at Pennsylvania State University.

Antibodies and Flow Cytometry

Cells were incubated for half an hour at 4°C with antibodies in 100 μl PBS /2%FBS, followed by a two washes in PBS /2%FBS. The following antibodies were purchased from BD Pharmingen and used as suggested by the manufacturer (San Diego, California): anti-CD8α-FITC, CD44-Cychrome, CD122-PE, IFNγ-FITC, CCR7-PE, CD127-PE, NK1.1-PE, and BrdU-FITC. α-galactosyl Ceramide/CD1d tetramers were from the NIH tetramer Facility (Atlanta, GA). Anti-CD4-ECD was from Invitrogen (Carlsbad, California). Anti-CD62L-APC-Alexa-Fluo750 was purchased from eBioscience (San Diego, California). Cells were analyzed using a FC500 from Beckman Coulter (Fullerton, California).

Quantitative Real-Time PCR analysis

CD4⁺CD62L^Lo or CD4⁺CD62L^Hi T cells were sorted from the spleens of WT and Itk null mice using a Cytopeia Cell Sorter (Cytopeia Inc., Seattle, WA). Total RNA was prepared from sorted cells using RNease mini kit (Qiagen Sciences, Maryland). cDNA was generated using You Prime First-Strand beads (GE healthcare, Buckinghamshire, UK), and quantitative PCR was performed using primer/probe sets for IFNγ, T-bet and Eomesodermin (Applied BioSystems, Branchburg, New Jersey), with GAPDH as a housekeeping gene. Data was analyzed using the ΔΔ Comparative CT (threshold cycle) method and normalized to GAPDH and relative to a calibrator sample. The relative gene expression levels were then determined by comparing to the expression found in the WT CD4⁺CD62L^Lo populations, which were set as 1 or as indicated in the respective figure legends.

In vitro analysis of cytokine secretion

Splenocytes were stimulated with 50 ng/ml PMA/0.5 μM Ionomycin for 6 hours, or IL-12 (5 ng/ml, R&D Systems, Minneapolis, MN) and IL-18 (10 ng/ml, MBL, Watertown, MA) overnight, in the presence of Brefeldin A (10 μg/ml) and analyzed for intracellular IFNγ and cell surface CD4 and CD44 by flow cytometry.

BrdU Incorporation

Mice were treated with BrdU in drinking water (dissolved at 0.8 mg/ml) for 9 days, with mice given fresh BrdU-containing drinking water daily. Splenocytes were collected and stained for surface marker using appropriate antibodies, followed by washing in PBS and resuspending in ice-cold 0.15 M NaCl. The cells were fixed and permeabilized in ice-cold 95% ethanol for 30 min on ice, then washed with PBS and fixed again in 1% paraformaldehyde for 30 min at room temperature. To detect BrdU, the cells were spun down and resuspended in 1 ml DNase I solution (50 U/ml deoxyribonuclease I in 4.2 mM MgCl₂/0.15 M NaCl, pH=5) for 10 min at room temperature. The cells were then washed and resuspended in 100 μl of 1:10 dilution of anti-BrdU-FITC for 30 min at room temperature. The cells were washed and resuspended in 500 μl PBS and analyzed by flow cytometry. For turnover analysis, the mice were treated with BrdU-containing water for 9 days and then treated with normal water for indicated days, and then analyzed by the procedure described above.

Fetal Thymic Organ Culture (FTOC)

Fetal thymi were harvested from E16 pregnant females. Thymic lobes were placed in transwell plates (Costar) with 2 ml DMEM medium (containing 15% Fetal calf serum) for the indicated time periods.

Bone marrow chimeras

Bone marrow was isolated from femurs and tibia of Thy1.1 WT and Itk^-/- mice. 1×10⁷ cells were injected into lethally irradiated congenic WT (Thy1.2) mice. Mice were analyzed 8 weeks after reconstitution. To determine if Itk^-/- T cells can compete with WT T cells during development, a 1:1 mixture (5 × 10⁶ cells each) of bone marrow from Thy1.2/CD45.1 congenic WT and Thy1.2/CD45.2 congenic Itk^-/- mice were injected into irradiated Thy1.1/CD45.2 WT mice, followed by analysis of donor derived WT (Thy1.2/CD45.1) and Itk^-/- (Thy1.2/CD45.2) T cells 6 weeks after reconstitution.

Proliferation

Purified CD4⁺CD62L^Lo and CD4⁺CD62L^Hi T cells from WT and Itk^-/- mice were stimulated at 2×10⁵ cells/well in triplicate with 1 μg/ml anti-CD3 or 1 μg/ml anti-CD3 plus 1 μg/ml anti-CD28 for 3 days and proliferation was measured by ³H-thymidine incorporation over the final 18 hours.

Bacterial infection

To analyze T cell secretion of cytokine ex vivo following infection, mice were infected with 2×10³ CFU L. monocytogenes for 24 hours, and splenocytes isolated and incubated in vitro with Brefeldin A (10 μg/ml) for a further 6 hours, followed by analysis of intracellular IFNγ as described previously (21).

Statistical analysis

Data was analyzed by Students’ t test, with a value of p<0.05 considered statistically significant.

Results

Increased percentages of CD4⁺CD62L^LoCD44^Hi T cells in Itk null mice

We and others have recently shown that Itk is required for the development of conventional CD44^Lo “naïve” CD8⁺ T but not a CD44^Hi “innate MP” population of CD8⁺ T cells (18-21). Similarly, Berg and co-workers have observed that there is a higher proportion of CD4⁺ T cells with a memory phenotype in mice lacking Itk (25). To determine if Itk affects the development of the CD4⁺ T cell lineage, we examined the CD4⁺ T cell lineages in the spleens of Itk^-/- and WT mice by characterizing the expression of surface maturation markers, CD62L and CD44. We found that Itk^-/- mice show a higher percentage of MP (for simplicity, we refer to cells carrying CD62L^Lo, CD44^Hi or CD62L^LoCD44^Hi as MP cells since similar results were observed using these markers) CD4⁺ T cells (Figure 1A). We also found that CD4⁺ T cells with MP were present in the youngest mice (less than one month) and the increase persists out to more than 12 month of age (Figure 1B). These data indicated that Itk regulates the development of CD4⁺ T cell lineage by altering the ratio of NP and MP CD4⁺ T cells.

Figure 1. Increased percentage of CD4⁺CD62L^LoCD44^Hi MP T cells in mice lacking Itk.

(a) Spleen cells from WT and Itk^-/- mice were stained for CD4, CD44, CD62L expression and analyzed by FACS. The percentages of CD4⁺CD62L^LoCD44^Hi and CD4⁺CD62L^Hi CD44^LoT cells in WT and Itk null mice are shown. (b) Splenocytes from WT and Itk^-/- mice were analyzed for CD4 and CD62L expression over the indicated time frame. The percentage of CD4⁺CD62L^Lo population is shown (n=4-5, *p<0.05). (c) WT and Itk^-/- mice were treated with BrdU containing water for 9 days, and splenocytes collected and analyzed for CD4 and CD62L expression along with analysis for BrdU. The percentage of BrdU⁺ cells on gated CD4⁺CD62L^Hi or CD4⁺CD62L^Lo cells were analyzed. (d) WT and Itk^-/- mice were treated with BrdU containing water for 9 days, and then placed on normal water for the indicated days prior to analysis of their splenocytes for BrdU incorporation in gated CD4⁺CD62L^Lo T cells as a percentage of total CD4⁺ T cells (n=3).

Homeostasis of naïve and MP CD4⁺ T cells does not contribute to the increased percentage of MP cells in Itk null mice

Maintenance of the size of the lymphocyte pool is critical for proper immune responses. When “naïve phenotype” (NP, for simplicity, we refer to cells carrying CD62L^Hi, CD44^Lo or CD62L^HiCD44^Lo as NP cells since similar results were observed using these markers) T cells are introduced into a lymphopenic compartment, they undergo homeostatic expansion, converting to a phenotype that resembles MP cells (26). It is possible that Itk null mice have altered T cell homeostasis, leading to increased percentage of MP T cells. To test this hypothesis, the turnover of CD4⁺ T cells in Itk^-/- mice and WT mice was determined by examining the CD4⁺ T cell incorporation of bromodeoxyuridine (BrdU). Itk^-/- and WT mice were fed with BrdU-containing water for 9 days, after which their CD4⁺ T cells in the spleen were analyzed for incorporation of BrdU. We found that the CD4⁺CD62L^Hi T cells incorporated little BrdU during this process, and CD4⁺CD62L^Hi cells from Itk^-/- mice incorporated slightly less BrdU compared to cells from WT mice (Figure 1C). We also found that amount of BrdU incorporated into CD4⁺CD62L^Lo cells from both Itk^-/- and WT mice were similar (Figure 1C), indicating that a similar percentage of Itk^-/- and WT CD4⁺CD62L^Lo cells were actively incorporating BrdU over this time period. These data suggest that Itk null T cells do not have increased proliferation in vivo. Indeed, transferring NP WT and Itk^-/- CD4⁺ T cells into RAG^-/- mice indicates that Itk^-/- T cells actually exhibit reduced homeostatic expansion in this lymphopenic environment (data not shown). To determine if these cells have altered turnover in vivo, we also performed pulse-chase experiments using BrdU labeling since the latter is not reused. Itk^-/- and WT mice were fed BrdU containing water for 9 days and then transferred to normal water to examine the rate of turnover of labeled MP CD4⁺ T cells. The data show that MP CD4⁺ T cells from both Itk^-/- and WT mice showed similar kinetics of decay (Figure 1D). Altogether, these data suggest that the homeostasis of NP and MP CD4⁺ T cells does not contribute to the altered ratio of NP and MP CD4⁺ T cells in Itk^-/- mice, although the homeostatic expansion and conversion of NP T cells to MP CD4⁺ T cells in a lymphopenic environment may be Itk dependent.

MP CD4⁺ T cells carry preformed message for IFNγ and rapidly secrete this cytokine upon stimulation with P/I

Previously activated or memory CD4⁺ T cells have the ability to produce effector cytokines immediately ex vivo after stimulation. To determine if these MP CD4⁺ T cells exhibit effecter function ex vivo, we examined their ability to secrete IFNγ upon stimulation. We found that a large proportion of the CD4⁺CD44^Hi population produce IFNγ in response to PMA and Ionomycin stimulation, while the CD4⁺CD44^Lo population did not secrete any IFNγ during this period. CD4⁺CD44^Hi T cells from Itk^-/- mice behave similar to those from WT mice by rapid production of IFNγ, with a similar percentage of WT as well as Itk^-/- CD4⁺CD44^Hi T cells making this cytokine (Figure 2A). These results indicate that MP CD4⁺ T cells exhibit effector function to secrete effector cytokines ex vivo.

Figure 2. MP CD4⁺ rapidly secrete IFNγ upon stimulation and carry high levels of preformed message for IFNγ and T-bet.

(a) Splenocytes from WT and Itk^-/- mice were stimulated with PMA/Ionomycin followed by analysis for intracellular IFNγ in gated CD4⁺ T cells (data representative of at least 3 experiments with the same result). (b) CD4⁺CD62L^Lo and CD4⁺CD62L^Hi T cells from WT and Itk^-/- mice were sorted and mRNA for IFNγ and T-bet were analyzed by Q-RT-PCR. Data are corrected for GAPDH expression and expressed as fold over the WT CD4⁺CD62L^Hi populations, which was set at 1 (n=3, *p<0.05, there was no statistical difference between WT and Itk^-/- CD4⁺CD62L^Lo cells).

We and others have found that MP CD8⁺ T cells with innate function rapidly secrete IFNγ upon stimulation since they carry large amounts of preformed message for this cytokine as well as the IFNγ regulator T-bet (18, 19, 21). The ability of CD4⁺CD44^Hi but not CD4⁺CD44^Lo T cells to rapidly produce IFNγ when stimulated suggests that these two subsets may differ in the expression of preformed IFNγ message. We therefore analyzed mRNA from freshly isolated unstimulated CD4⁺CD62L^Lo and CD4⁺CD62L^Hi T cells from Itk^-/- and WT mice for preformed mRNA for IFNγ and T-bet by real-time quantitative RT-PCR. We found that CD4⁺CD62L^Lo T cells carry significantly higher levels of preformed message for IFNγ, as well as higher levels of the transcription factor T-bet compared to the CD4⁺CD62L^Hi T cells, although there was no difference between WT and Itk^-/- CD4⁺CD62L^Lo T cells (Figure 2B). Analysis of mRNA for Eomesodermin, a T-bet related transcription factor also revealed both WT and Itk^-/- CD4⁺CD62L^Lo cells expressed 3-4 fold more Eomesodermin than CD4⁺CD62L^Hi cells (data not shown). These data indicated that the MP CD4⁺ T cells have higher levels of T-bet and IFNγ transcripts, which may contribute to the ability of these cells to rapidly secrete IFNγ upon stimulation.

Phenotypic characterization of NP and MP CD4⁺ T cells in Itk^-/- mice

To further characterize these two T cell populations, we examined them for expression of a variety of surface markers. As shown in Figure 3, WT and Itk^-/- CD4⁺CD44^HiCD62L^Lo T cells expressed similar levels of CD122, CCR7 and CD127, suggesting that these two populations were the same in the two strains of mice. There were difference in expression of specific markers between NP and MP T cells in the expression of CD122, CCR7 and CD127. Of interest is that the MP, but not the NP subset also express low levels of NK1.1, the marker for NK and NKT cells, and Itk^-/- mice have a smaller percentage of these cells than WT mice as previously suggested (27, 28)(data not shown). The small percentage of the MP CD4⁺ T cells that are NK1.1 or α-GalCer/CD1d tetramer positive rule out the possibility that MP CD4⁺ T cells are NK or NKT cells, since these cells can also carry preformed message for IFNγ and rapidly secrete cytokine upon stimulation (28)(data not shown).

Figure 3. Surface phenotype of CD4⁺ CD62L^LoCD44^Hi and CD62L^HiCD44^Lo T cells from WT and Itk^-/- mice.

Splenocytes from WT and Itk^-/- mice were stained for CD4, CD62L, CD44 and the indicated surface markers. FACS profiles shown are gated on CD62L^LoCD44^Hi and CD62L^HiCD44^Lo CD4⁺ T cells (percentages were not statistically different between the two types of cells from WT and Itk^-/- mice except NK1.1 and α-galCer/CD1d Tetramer staining in the CD62L^LoCD44^Hi cells (n=3). Data representative of 2 experiments with the same result).

MP CD4⁺ T cells develop in the thymus

Our data show that Itk^-/- mice have increased percentage of MP CD4⁺ T cells. One potential explanation for these results is that T cells that develop in the thymus migrated into a lymphopenia-like environment in the Itk^-/- mice, during which they proliferated and up-regulated CD44 and down-regulated CD62L. This could result in the finding of higher percentages of MP CD4⁺ T cells in periphery of Itk^-/- mice since a higher percentage of NP T cells would undergo lymphopenia-induced proliferation in these mice. However, we have already shown that Itk^-/- T cells do not undergo increased homeostatic expansion. Another potential explanation is that these MP CD4⁺ T cells develop in the thymus and migrate out into the periphery, and that in the absence of Itk, more of these cells develop, or alternatively, less NP phenotype cells develop in the thymus resulting in the observed increased percentage of these cells in these mice. We therefore wanted to determine if these cells originated in the thymus during T cell development. To examine this issue, we first analyzed CD4⁺ T cells in newborn mice from birth through the first 1 week. We found that MP CD4⁺ T cells were present at 2 days after birth in both WT and Itk^-/- mice (Figure 4A). Furthermore, increased percentages of CD4⁺CD44^hiCD62L^low was detected in Itk^-/- mice compared to WT controls. The percentage of MP CD4⁺ T cells decreased at 4 days and 7 days after birth in WT mice, while in Itk^-/- mice, this percentage also decreased but remained elevated compared to WT mice. This suggests that MP CD4⁺ T cells develop in the thymus. We also characterized the expression of specific surface maturation markers that identify these cells. We found that CD4⁺SP thymocytes in 6-8 week old Itk^-/- mice exhibited a higher percentage of CD44^Hi and CD122^Hi populations than cells in WT mice, which suggests that Itk^-/- mice contained higher percentage MP CD4⁺SP thymocytes than WT mice (Figure 4B). As previously reported, almost all CD8⁺SP thymocytes in Itk^-/- mice exhibit a MP (CD44^HiCD122^Hi) (18-21). To further confirm this, we evaluated fetal thymic organ cultures (FTOC) from WT and Itk^-/- mice where T cell development occurs in vitro, ruling out potential recirculation of already developed cells back into the thymus as would occur in the animal. Our results show that a higher percentage of CD4 SP T cells develop the CD44^HiCD122⁺ phenotype in Itk null FTOC than in the WT FTOC (Figure 4C).

Figure 4. MP CD4⁺ T cells develop in the thymus.

(a) Splenocytes from WT and Itk^-/- day 2, day 4 and day 7 old mice were stained for CD4, CD44 and CD62L and analyzed by FACS. The percentages of CD44^HiCD62L^Lo populations on gated CD4⁺ T cells were analyzed (n=4, *p<0.05). (b) Thymocytes harvested from 6-8 week old age-matched WT and Itk^-/- mice were stained for CD4, CD8 and CD44 and CD122 and analyzed by FACS. The histograms show gated CD4⁺ and CD8⁺ T cells. WT (solid line), Itk^-/-(dashed line), Isotype staining control (Shaded). Arrow indicates increased percentage of CD4 SP CD44^Hi and CD4 SP CD122^Hi cells in Itk null thymus. (c) Fetal Thymic Organ Cultures from embryonic day 16 of WT and Itk null mice incubated in vitro for the indicated days and analyzed for CD4 and CD8 (top panel) or for CD44 and CD122 expression in the CD4 single positive population (data representative of 3-4 mice).

These data suggest that MP CD4⁺ and CD8⁺ T cells compartment differ in some aspects of their development since the percentage of the CD4⁺ compartment that had this memory phenotype was lower than that seen in the CD8⁺ compartment, but they have similar phenotypes. These data also suggest that the increased percentage of MP CD4⁺ T cells observed in the absence of Itk reflects either enhanced development of these cells, or reduced development of NP CD4⁺ T cells.

Altered CD4⁺ lineage development in the absence of Itk is intrinsic to bone marrow-derived cells

To better understand whether the altered development of Itk^-/- CD4⁺ T cells was due to defects intrinsic to the developing T cells, we generated bone marrow chimeric mice in which WT and Itk^-/- bone marrow was injected into lethally irradiated WT congenic mice. After reconstitution, the percentage of CD4⁺CD44^HiCD62L^Lo T cells in the spleen was clearly higher in mice reconstitution with Itk^-/- bone marrow compared to those reconstituted with WT bone marrow (Figure 5A). However, when we compared the number of CD4⁺CD44^HiCD62L^Lo T cells in the spleen, we found similar numbers regardless of whether the mice received WT or Itk^-/- bone marrow (Figure 5B). To further determine if CD4⁺CD44^HiCD62L^Lo MP T cells are indeed able to develop independently of Itk expression, we performed competitive mixed bone marrow chimera analyses to determine if Itk null cells cam effectively compete with WT cells in the same host for development to these two cell populations. Our results confirm that while development of Itk null CD4⁺CD44^LoCD62L^Hi NP T cells were reduced compared to their WT counterparts, development of CD4⁺CD44^HiCD62L^Lo MP T cells was not affected and equal numbers of WT and Itk^-/- cells developed (Figure 5C&D). This indicates that Itk is not required for the development of CD4⁺CD44^HiCD62L^Lo MP T cells, but is required for the development of CD4⁺CD44^LoCD62L^Hi NP T cells, and that the increased percentage observed in the absence of Itk is due to reduced development of the latter population. These data also suggest that these two populations of T cells are distinct and have distinct requirements for their development.

Figure 5. Altered CD4⁺ T cell development in the absence of Itk is intrinsic to bone marrow-derived cells.

(a) Bone marrow from Thy1.1 congenic WT and Itk^-/- mice were injected into irradiated Thy1.2 WT mice, followed by analysis 8 weeks after reconstitution. The percentages of CD4⁺CD44^HiCD62L^Lo and CD4⁺CD44^LoCD62L^Hi populations of gated Thy1.1⁺ (donor) cells are shown (data representative of 3 mice with the same result). (b) The numbers of donor derived Thy1.1⁺ CD4⁺CD44^HiCD62L^Lo and CD4⁺CD44^LoCD62L^Hi T cells were determined and plotted (n=3, p<0.05). (c) A 1:1 mixture of bone marrow from Thy1.2/CD45.1 congenic WT and Thy1.2/CD45.2 congenic Itk^-/- mice were injected into irradiated Thy1.1 WT mice, followed by analysis of donor derived T cells 6 weeks after reconstitution. The percentages of CD44^Hi populations of gated donor WT and Itk null CD4⁺ cells are shown (data representative of 7 mice with the same result). (d) The numbers of donor derived WT or Itk null CD4⁺CD44^HiCD62L^Lo and CD4⁺CD44^LoCD62L^Hi T cells were determined and plotted (n=3, repeated twice, *p<0.05).

Tec kinase activity is required for the presence of NP CD4 ⁺ T cells, but not for the MP CD4⁺ T cells

Our data shows that Itk regulates the development of CD4⁺ T cells by changing the ratio of NP and MP CD4⁺ T cells, and suggest that these represent unique and separate populations of CD4⁺ T cells. Our data also shows that their development is intrinsic to bone marrow-derived cells. To further determine if this process was intrinsic to the T cells and if the Tec kinase signaling was involved, we examined the percentage and absolute numbers of these two CD4⁺ T cell lineages in transgenic mice carrying Itk expressed in a T cell specific manner (driven by the CD2 promoter, Tg(CD2-hItk)Itk^-/- mice), as well as transgenic mice carrying a mutant Itk lacking its kinase also expressed in a T cell specific manner (Tg(Lck-ItkΔKin)Itk^-/- mice, (21)). Tg(CD2-hItk)Itk^-/- mice expressed low levels of Itk under the CD2 promoter (approximately 25-30%), however this was sufficient to significantly rescue the development of NP phenotype CD4⁺ T cells, but did not have any effect on the numbers of MP CD4⁺ T cells (Figure 6A, B). By contrast, analysis of the NP and MP CD4⁺ T cells in Tg(Lck-ItkΔKin)Itk^-/- mice carrying the mutant Itk lacking its kinase domain instead of WT Itk revealed that these mice had NP and MP CD4⁺ T cells populations similar to those seen in Itk null mice (Figure 6). These data indicate that active signaling by Itk enhances the development of NP CD4⁺ T cells, but has little effect on the development of MP CD4⁺ T cells.

Figure 6. The development of MP CD4⁺ but not NP CD4⁺ T cells is independent of active Itk signaling.

(a) Spleens from WT, Itk^-/-, Tg(CD2-Itk_tg)Itk^-/-, and Tg(Lck-ItkΔKin)Itk^-/- mice were stained for CD4, CD44, CD62L expression and analyzed by FACS, gating on CD4⁺ populations. The percentages of CD4⁺CD62L^LoCD44^Hi and CD4⁺CD62L^HiCD44^Lo populations are shown (data representative of 6 mice with the same result). (b) Total numbers of splenic CD4⁺CD62L^LoCD44^Hi and CD4⁺CD62L^Hi CD44^Lo T cell populations from the same mice as Figure 6a. (n=3, *p<0.05 vs. Itk^-/- CD4⁺CD62L^Hi CD44^Lo T cells; **p<0.05 vs. Itk^-/- CD4⁺CD62L^Hi CD44^Lo T cells).

Itk is required for TCR induced but not innate signal induced elaboration of MP CD4⁺ effector function

Itk regulates signals emanating from the T cell receptor (6-12). To determine if the MP CD4⁺ T cells require Itk for their proliferation through the TCR, we analyzed purified CD4⁺CD62L^Lo and CD4⁺CD62L^Hi T cells for proliferative responses to anti-CD3 or anti-CD3/28 stimulation. The results show that both CD4⁺CD62L^low and CD4⁺CD62L^hi populations from Itk^-/- mice had less proliferation in response to anti-CD3 and anti-CD3/28 stimulation (Figure 7A), which indicated that both populations of cells are dependent on Itk for TCR induced proliferation. Similar results were found when we examined IFNγ secretion, with only WT CD4⁺CD62L^low but not Itk^-/- CD4⁺CD62L^low cells making this cytokine, although at much lower levels than seen with figure 2a (data not shown).

Figure 7. TCR activation, but not innate activation, of MP CD4⁺ TCR is Itk dependent.

(a) CD4⁺CD62L^Lo and CD4⁺CD62L^Hi T cells from WT or Itk^-/- mice were stimulated with anti-CD3 and anti-CD3/28 for 3 days. Thymidine uptake was determined as a measure of proliferation in the last 18 hrs (n=3, *p<0.05). (b) Splenocytes from WT and Itk^-/- mice were stimulated with IL-12 and IL-18 followed by analysis for intracellular IFNγ on gated CD4⁺ T cells. Percentages shown indicate those CD4⁺ T cells secreting IFNγ upon stimulation (data representative of 6 mice with the same result). (c) WT and Itk^-/- mice were infected with 2×10³ CFU L. monocytogenes. Twenty-four hrs later, splenocytes were harvested and CD4⁺CD44^Hi T cells were analyzed for intracellular IFNγ. Percentages shown indicate those CD4⁺ T cells secreting IFNγ upon infection (data representative of 6 mice with the same result).

A role for MP CD8⁺ T cells in the early innate response following infection with L. monocytogenes has been reported (29-31). Infection of macrophages with L. monocytogenes results in the secretion of IL-12 and IL-18, which together can induce the rapid secretion of IFNγ from MP CD8⁺ T cells. We have shown that MP CD8⁺ T cells respond to IL-12/IL-18 stimulation by rapidly secreting IFNγ, and can also respond to infection with L. monocytogenes by rapidly secreting IFNγ (21). We therefore determined if MP CD4⁺ T cells could also respond to IL-12/IL-18 stimulation to secrete IFNγ. Analysis of IFNγ production in cells from WT and Itk^-/- mice revealed that MP CD4⁺ T cells produced significant levels of IFNγ, and similar percentages of WT and Itk null CD4⁺CD44^hi cells responded (Figure 7B, note that PMA/Ionomycin stimulated cells, shown in Figure 2A, reveal similar responses). Analysis of MP CD4⁺ T cells revealed that these cells (and a similar percentage in WT and Itk^-/- mice) could also rapidly secrete this IFNγ during infection with L. monocytogenes (Figure 7C). These data suggest that Itk is required for TCR stimulation of MP CD4⁺ T cells, but not for their development or elaboration of innate immune function.

Discussion

In this report, we show that development of a population of CD4⁺ T cells that carry memory markers CD44^HiCD62L^Lo are independent of Itk expression, while the “naïve” population of CD4⁺ T cells, CD44^LoCD62L^Hi, are dependent on Itk for their development. Our data also suggest that these MP CD4⁺ cells develop very early in the thymus, and are not dependent on active Itk kinase mediated signals. These MP CD4⁺ T cells carry preformed message for IFNγ and T-bet, and rapidly secrete this cytokine upon stimulation with IL-12 and IL-18. More importantly, this population rapidly secretes IFNγ upon infection with L. monocytogenes, suggesting that they may participate in the early “innate” immune response. These data suggest that these cells represent an apparently separate subpopulation of CD4⁺ cells from those with the “naïve” phenotype.

The data in this paper provides compelling evidence to support the existence of two independent subpopulations of CD4⁺ T cells: CD62L^HiCD44^Lo which are an apparent “naïve” phenotype and CD62L^LoCD44^Hi, which are an apparent “memory” phenotype. We use the term apparent as these markers have traditionally been used to refer to these two populations, however, it is clear that they include distinct populations that have different requirements for development. The CD4⁺CD62L^HiCD44^Lo T cells resemble the NP T cell population, and lack preformed message for IFNγ and T-bet, and do not rapidly secrete IFNγ upon stimulation with P/I, IL-12/IL-18 or Listeria infection. This population is dependent on Itk expression and activity for its development. By contrast, the CD4⁺CD62L^LoCD44^Hi T cell population develops in an Itk independent manner. Our data also shows that the development of these two populations of CD4⁺ T cells is T cell intrinsic since, since Itk null bone marrow also gave rise to these two populations in a cell intrinsic manner when transferred into WT mice. The independence of these two cell populations is supported by the fact that they appear very early in T cell development. These two populations of cells were not significantly different in their expression of TCR Vβ 3, 5, 6 and 8, suggesting that they are not oligoclonally selected or expanded in vivo (data not shown), as in the case of NKT cells (32).

These populations also carry different cell surface markers that separate them phenotypically. Some of these markers can be clearly tied to their function, such as the expression of CD122 in CD4⁺CD44^Hi T cells, which allows responsiveness to IL-15 (33-35). Indeed, IL-15 has been shown to be critical for the expansion of similar populations of CD4⁺CD62L^LoCD44^Hi “memory” phenotype T cells, as well as in generating effective memory T cells following exposure to antigen (36, 37). These CD4⁺CD62L^LoCD44^Hi T cells also express lower levels of the chemokine receptor CCR7, suggesting that their trafficking may different from the CD4⁺CD62L^HiCD44^Lo “naïve” population which express higher levels of this receptor. In addition to their phenotypic characteristics, CD4⁺CD62L^LoCD44^Hi T cells are distinct from CD4⁺CD62L^HiCD44^Lo T cells in their effector functions, the CD4⁺CD62L^LoCD44^Hi T cells resemble previously activated/memory T cells and produce IFNγ directly ex vivo in response to stimulation. These features are shared by some nonconventional T cell lineages that can develop effector function prior to antigen encounter such as NKT cells (38). We note that it is unlikely that the cytokine secretion response that we observe is due to NKT cells, since these latter cells comprise at most 10% of this MP CD4⁺ T cell population, but we get up to 45% of these MP CD4⁺ T cells secreting IFNγ upon stimulation. In addition, the Itk^-/- have reduced percentage and numbers of NKT cells (data not shown, (27, 28)). On the basis of these characteristics, CD62L^LoCD44^Hi CD4⁺ T cells should likely be included amongst these types of innate T cells, and we suggest the term “innate memory phenotype” CD4⁺ T cells.

The CD4⁺CD62L^LoCD44^Hi T cells with innate function are also distinct from the CD4⁺CD62L^HiCD44^Lo T cells in that they have different intracellular signaling requirements for development. As we show here, Itk deficiency greatly decreases the number of CD4⁺CD62L^HiCD44^Lo T cells, while having no effect on the number of CD4⁺CD62L^LoCD44^Hi T cells, suggesting that Itk is not required for the development of the CD4⁺CD62L^LoCD44^Hi T cells. Although the development of CD4⁺CD62L^LoCD44^Hi T cells is Itk independent, our data also shows however, that Itk is still required for functional TCR signaling in these cells. Thus development and functional TCR activation of these cells have different signaling requirements.

We and others have previously reported that the development of “conventional” CD8⁺CD44^LoCD122^Lo T cells is abolished in the absence of Itk (18-21). Our data shows that the development of a similar population of CD4⁺ T cells is also affected in the absence of Itk, suggesting that Itk plays a critical role in the development of both CD8⁺ as well as CD4⁺ T cells that have a “naïve” CD44^Lo phenotype. We have also shown that the MP CD8⁺ T cells observed in the absence of Itk share the same properties with a population in normal WT mice, and more importantly can function in an innate manner to rapidly secrete IFNγ during infection with Listeria (21). Together, our data suggest that the “memory” phenotype CD44^Hi CD8⁺ and CD4⁺ T cell populations that develop in an Itk independent manner both seem to be able to function innately.

CD62L and CD44 are used as markers to distinguish naïve and memory T cell in many studies (39). The CD62L^LoCD44^Hi CD4⁺ T cells described in this paper are defined as MP T cells and these cells can arise spontaneously in normal mice, which is different from the antigen-specific memory T cells generated by antigen administration (37). It is assumed that memory T cells having this phenotype are generated by antigen activation, expansion and differentiation, and that MP T cells found in normal mice reflect the fact that T cells are exposed to various environmental antigens leading to the development of these memory T cells (36, 37). However, our data suggest that at least some T cells with the CD4⁺CD62L^LoCD44^Hi MP pool can be generated during thymic development and exist very early in the life of the animal. Our data therefore suggest that some or perhaps most CD4⁺CD62L^LoCD44^Hi MP T cells are not descendants of naïve T cells that have responded to foreign antigens, but are a unique population of T cells that have innate function and behave like traditional memory T cells. It is possible that this T cell population develops in order to rapidly respond to antigen or innate signals until naïve T cells can differentiate and participate in the immune response.

Abbreviations Used

AICD

activation induced cell death

DP

Double Positive

Itk

Interleukin-2 Inducible T cell kinase

MAIT

Mucosal Associated Invariant T cells

MP

Memory Phenotype

NP

Naïve Phenotype

P/I

PMA and Ionomycin

PLCγ-1

phospholipase-Cγ1

Rlk/Txk

Resting Lymphocyte Kinase

SP

Single Positive

T-bet

T-box expressed in T cells