Mice Deficient for N-ras: Impaired Antiviral Immune Response and T-Cell Function¹

Ras proteins are small, guanine-nucleotide-binding polypeptides that have a key role in the regulation of several cellular functions like proliferation, differentiation, or apoptosis. These proteins are involved in a significant percentage of human tumors and are a central point for many signal transduction pathways in the cell (1). Ras proteins can rapidly cycle between an inactive GDP-bound and an active GTP-bound state. Activated Ras proteins are then able to interact with downstream effector molecules and propagate the signal. Two different groups of molecules are involved in the regulation of Ras proteins: the guanine nucleotide exchange factors that promote the transition from the GDP-bound to the active GTP-bound state, and the GTPase-activating proteins that are able to inhibit Ras by stimulating the Ras GTPase activity.

Despite the similarity between the three mammalian ras isoforms (H-ras, K-ras, and N-ras), several differences have been found among them. Although for the first 80 amino acids N-, H-, and K-ras are identical, and for the next 80 they exhibit 85% identity between any pair of Ras isoforms, Ras proteins differ substantially in their COOH-terminal region with fewer than 15% conserved residues (2). This region is very important for the maturation and posttranslational modifications of the Ras products that are involved in the association of Ras proteins to the membranes and their subsequent activity (3, 4, 5).

Other functional differences between the Ras isoforms have been reported (1). For instance, the GTPase-activating protein NF1 has a 4-fold higher affinity for H-ras than for N-ras (6). Among the guanine nucleotide exchange factors, SmgGDS, Ras-GRF, and RasGRP2 activate differentially the three Ras isoforms (7, 8, 9). Regarding downstream ras effectors, in vitro assays also suggest differences in Ras isoform-dependent activation of PI3K, Raf-1, and Rac (10, 11). A specific role of K-ras and H-ras has been reported in the proliferation of human renal fibroblasts (12, 13), whereas N-ras signaling has been related recently in the cell survival of immortalized fibroblasts (14). Differences have also been found in the role of Ras proteins in tumorigenesis because different ras genes have been found mutated in different tumor types (15, 16). Finally, mice defective for the different ras isoforms exhibit different developmental phenotypes. N-ras and H-ras knockout mice develop and reproduce normally (17, 18, 19), whereas K-ras is essential for the embryonic development (20).

It has been demonstrated that, in lymphocytes, Ras plays an important role in the signaling pathways that activate cytokine gene induction and in the control of B-cell and T-cell development (21). Since the activation of Ras on T-cell activation was first demonstrated (22), several groups have confirmed the key role of Ras activation in antigen receptor-activated lymphocytes (23, 24, 25). In fact, the loss of Ras function prevents the normal activation of proliferation, cytokine production, and lymphocyte development that is induced by the recognition of the antigen (26, 27).

However, the specific roles, if any, of the different Ras isoforms in lymphocyte function are poorly understood. N-ras is the main ras gene found activated in human myeloid and lymphoid disorders (15, 16). Moreover, although transgenic mouse lines that express either the N-ras proto-oncogene or its oncogenic form show a high incidence of lymphomas (28), transgenic mice generated with the H-ras oncogene under the same promoter had a low or null incidence of those tumors (29, 30). Therefore, N-ras might play an important role in the regulation of lymphocyte function. To test this hypothesis, we investigated the specific role of N-ras in T-cell function and development using an N-ras-deficient mouse model.

Age-matched wild-type mice (N-ras+/+ mice from the same litters), and KO-N-ras mice (17) were housed under pathogen-free conditions. BALB/c mice were purchased from The Jackson Laboratory, Bar Harbor, ME. All of the work with mice conformed to guidelines approved by the Institutional Animal Care and Use Committee of New York University School of Medicine.

FITC-conjugated monoclonal antibodies specific for CD4, CD8, CD11b; phycoerythrin-conjugated monoclonal antibodies specific for CD8, CD69, CD25, F4/80, and CD44; a tri-color-conjugated monoclonal antibody specific for CD4; and unlabeled goat antihamster IgG (H+L) were purchased from Caltag. Phycoerythrin-conjugated monoclonal antibody specific for NK1.1 and unlabeled hamster antimouse CD3 and CD28 were purchased from PharMingen. Antibodies used for ras detection included anti-pan-ras Y13–259 (31), monoclonal antibodies for mouse N-ras (Santa Cruz Biotechnology) and for H-ras and K-ras (Oncogene Science). Antibodies for activated or total ERK,⁶ and activated AKT and JNK were from Promega.

For proliferation experiments, thymocytes that were isolated from mice were washed and resuspended in complete medium (RPMI 1640 plus 10% FBS, 100 units/ml penicillin, 100 μg/ml streptomycin, and 50 μm 2-mercaptoethanol). Thymocytes (2 × 10⁵/well) were incubated in 96-well plates and incubated for 72 h at 37°C in 5% CO₂ in air, with or without the following stimuli: plate-bound anti-CD3 (10 μg/ml) with or without plate-bound anti-CD28 (10 μg/ml), and PMA (100 ng/ml) plus ionomycin (500 ng/ml). Cells were pulsed during the last 10 h of incubation with [³H]thymidine, and the incorporation of radiolabel was measured by scintillation counting.

For apoptosis assays, thymocytes were incubated for 24 and 48 h in 6-well-plates (2 × 10⁶/well) with plate-bound anti-CD3 plus anti-CD28, PMA (100 ng/ml) plus ionomycin (500 ng/ml), or dexamethasone (1 μm). Apoptosis was monitored using an annexin V apoptosis detection kit (PharMingen).

To examine the expression of surface antigens, cells were washed and then resuspended in PBS containing 0.2% BSA. Saturating concentrations of antibodies were added, and the cells were incubated for 30 min at 4°C in the dark. Cells were washed three times and fixed in 0.5% paraformaldehyde in PBS for analysis on a FACScan flow cytometer (Becton Dickinson). Fluorescence data were analyzed using the CellQuest software. Dead cells and debris were excluded by characteristic forward and scatter profiles.

IL-2 production was measured in culture supernatants from stimulated thymocytes 20 h after activation by ELISA (R&D Systems).

For the survival experiments, 6–9-week-old mice were anesthetized and inoculated intranasally with different doses of influenza A virus (WSN strain) diluted in PBS. Mice were monitored daily, and survival was scored after 14 days. For comparison of subsets of immune cells and pathogenesis, mice were infected with 100 PFU of influenza virus and monitored at days 3, 5, 7, and 10 postinfection. Five mice of each genotype were sacrificed per day, and the left lung was removed for flow cytometric analysis to characterize the subsets of immune cells, and the right lung was used to carry out histological studies. In the last case, samples were fixed in 10% buffered formalin, embedded in paraffin, and 5-μm sections were stained with H&E and analyzed under the microscope.

For the generation of primary CTLs in vitro, responder cells (5 × 10⁵) from KO-N-ras and wild-type spleens were cultured in a total of 2 ml of complete RPMI 1640 with stimulator splenocytes (2.5 × 10⁵) from BALB/c mice that were previously irradiated (2000 rads). After 4 days, cells were harvested and CTL activity was determined using P815 cells as targets. For the NK activity assay, fresh splenocytes were used as effector cells, whereas YAC-1 cells were used as target cells. In vitro CTL activity and NK lysis were analyzed by incubating different numbers of effector cells with 2 × 10⁴ target cells for 4 h at 37°C. The cytotoxicity of both CTL and NK cells was measured using the CytoTox 96 nonradioactive cytotoxicity assay (Promega). All of the reactions were performed in quadruplicate.

For the detection of Ras proteins in thymus, proteins (400 μg/sample) were immunopurified using an anti-pan-ras immunoaffinity column, fractionated on 15% SDS polyacrylamide gels, and detected using antibodies specific for each isoform. For signaling and Ras activation experiments, thymocytes were isolated in RPMI and incubated at 37°C for 2 h. For TCR activation, 2 × 10⁷ cells were incubated for 30 min on ice in the presence or absence of anti-CD3 (10 μg/ml) an/or anti-CD28 (10 μg/ml). After washing them in cold PBS, cells were activated by resuspension for 1 min in 37°C PBS containing cross-linking mouse antihamster antibodies. Cells were lysed in 400 μl of lysis buffer [10% glycerol, 1% NP40, 50 mm Tris-HCl (pH 7.4), 200 mm NaCl, 2.5 mm MgCl₂, 1 mm phenylmethylsulfonyl fluoride, 1 μm leupeptin, 10 μg soybean trypsin inhibitor per ml and 0.1 μm aprotinin]. Lysates were incubated overnight at 4°C with Gluthatione-S-Transferase-Raf-RBD (Ras-binding domain) fusion protein (a gift from J. L. Bos, Utrecht University, Utrecht, the Netherlands) coupled to glutathione agarose beads to collect GTP-bound Ras. Beads were pelleted and washed three times with lysis buffer, and ras proteins were detected by Western blot. Active ERK, JNK, and AKT proteins were detected in the primary lysates in accordance with the manufacturer’s recommendations (Promega). Lysates were also probed with anti-ERK to confirm that comparable amounts of total protein were present.

To determine the role of N-ras in T-cell development, we analyzed thymocyte and splenocyte cellularity and subsets from wild-type and KO-N-ras mice by flow cytometric analysis. Although all of the thymocyte subsets are present in the KO-N-ras mice, the percentage of both single positive CD4 and CD8 thymocytes was lower among KO-N-ras cells than among wild-type cells (Table 1). This reduction was more apparent for CD8⁺ thymocytes, which showed a significant reduction of 25% of the average value found among wild-type thymocytes (P = 0.015). Interestingly, the population of splenocytes showed a similar result in the number of CD8⁺ cells (15% decrease), although in this case, differences did not reach a significant value (P = 0.089).

Because this reduced representation of single positive cells might result from ineffective positive selection, we characterized KO-N-ras for the surface marker CD69, which has been related with double-positive thymocytes undergoing positive selection (32). Interestingly, the percentage of CD69⁺ positive cells among the CD4/CD8 double positive thymocytes in the N-ras-deficient mice was significantly reduced by 42% as compared with their wild-type counterparts (Fig. 1 A).

We further investigated the influence of N-ras deficiency on CD4/CD8 maturation on thymocyte activation in vitro. After 72 h of incubation, wild-type thymocytes treated with anti-CD3 plus anti-CD28 showed a 2.16 ± 0.24-fold induction of CD8 single positive cells relative to the nontreated cells (Fig. 1 B), whereas KO-N-ras thymocytes showed a significantly weaker induction of CD8 single positive thymocytes (1.31 ± 0.12-fold induction; P < 0.01). Because significant differences were not detected in the case of CD4 single positive and CD4/CD8 double positive cells, these data support a specific role of N-ras in CD8 maturation on anti-CD3 plus anti-CD28 stimulation.

CD8⁺ T lymphocytes are crucial components of the cell-mediated immune response. The murine influenza pneumonia model is one of the better-defined experimental systems to study cell-mediated immune response in virus infections. CD8⁺ T cells rapidly increase from virtually undetectable in the naïve host to high numbers in the primary response to influenza virus infection (33) and play a critical role in viral clearance from influenza-infected mice (34). Because KO-N-ras thymocytes have low numbers of CD8⁺ single positive cells, we decided to test whether this alteration could affect the antiviral response in N-ras-deficient mice challenged with influenza virus. Although KO-N-ras were more sensitive than wild-type mice to 10,000 and 1,000 PFU of influenza virus (Fig. 2, A and B), differences were significant when 100 PFU of virus were used in the infection (P = 0.0014; Fig. 2 C). Although this dose was barely lethal for wild-type mice (90.5% survival), it clearly affected the viability of KO-N-ras mice (45.8% survival).

This increased sensitivity to influenza infection could be explained by the impaired cytolytic activity of KO-N-ras CTLs and NK cells. To characterize the CTL activity, a mixed lymphocyte reaction (MLR) assay was performed as described in “Materials and Methods.” After 4 days of cocultivation with irradiated BALB/c splenocytes, cellularity and splenocyte subsets were analyzed, and the stimulated lymphocytes were used as effector cells in a standard cytolysis assay. No significant differences were detected in the cytolytic activity of activated splenocytes (Fig. 3,A). Interestingly, there was a 25% decrease in the number of total cells among the KO-N-ras samples (data not shown). Analysis of thymocyte subsets showed that this decrease was more pronounced for CD8⁺ cells (32.9%) compared with CD4⁺ cells (16.8%). NK cells are part of the innate immune response that is activated in the earlier stages of viral infections to control virus spread until the adaptive immune response is fully activated. NK cell cytolytic activity was assessed by incubating freshly prepared splenocytes with YAC-1 target cells. As Fig. 3 A shows, there were no differences in the cytolytic activity between KO-N-ras and wild-type NK cells.

To better determine the role of N-ras in the different stages of the antiviral response, mice of each genotype were infected with 100 PFU of influenza virus and analyzed at days 3, 5, 7, and 10 postinfection. As Table 2 shows, we characterized, not only the CD8⁺ subset, but also the subsets of other immune cells that play an important role in this antiviral response: CD4⁺, NK, granulocytes, and macrophages. There were no significant differences between wild-type and KO-N-ras CD4⁺ cells. However, the pattern for the other studied subsets was different between KO-N-ras and wild-type mice. At day 5 postinfection, cells from KO-N-ras showed significantly lower cell numbers of both CD8⁺ and NK⁺ subsets than cells from wild-type infected lungs. Regarding granulocyte levels, KO-N-ras cells showed a delayed profile in comparison with wild-type mice. However, this trend did not reach statistical significance. Finally, in the case of macrophages, KO-N-ras mice showed lower levels than wild-type samples at days 5 and 7 (P = 0.05 and P = 0.01, respectively).

At day 5 postinfection, the total number of cells harvested from infected lungs was significantly lower among KO-N-ras mice (26% average reduction, P < 0.05; Table 2). In fact, all of the subsets of immune cells showed lower numbers in KO-N-ras mice than in wild-type mice. This average reduction (21% for CD4⁺ cells, 23% for macrophages, 24% for granulocytes, 27% for NK cells, and 36% for CD8⁺ cells) is suggesting an impaired proliferation of KO-N-ras cells in the early stages postinfection. Corroborating this hypothesis, PI staining of the total cells harvested from KO-N-ras-infected lungs showed a significant reduction of cells in S phase at day 5 postinfection (40% average reduction, P < 0.05).

We also monitored pulmonary histopathology in the same infected mice. Again, clear differences were found by this comparative study (Fig. 3 B). Both KO-N-ras and wild-type mice showed bronchiolitis characterized by high numbers of lymphocytes. However, this inflammatory response was stronger among wild-type mice than among KO-N-ras mice. Whereas wild-type lungs showed a diffuse inflammatory process, 38% of the studied KO-N-ras lungs showed a focal bronchiolitis. Moreover, interstitial inflammation (pneumonitis) was also found more frequently among wild-type mice than in KO-N-ras mice (40 versus 14%, respectively). Finally, at late stages of the infection diffuse alveolar damage was also found in KO-N-ras mice, whereas wild-type mice lungs were characterized by repair processes.

In the previous experiments, we determined that KO-N-ras T-cells were defective in proliferation in vivo. To characterize the molecular mechanisms responsible for this phenotype, we decided to test in vitro the proliferative response of KO-N-ras thymocytes. Significant differences between wild-type and KO-N-ras mice were observed when thymocytes were stimulated with anti-CD3, anti-CD3 plus anti-CD28, or PMA plus ionomycin (Fig. 4 A). However, whereas the reduction in the case of PMA plus ionomycin was only 1.43-fold, a higher reduction was observed for anti-CD3 stimulation (4.50-fold) and CD3 plus CD28 stimulation (4.05-fold).

Because these differences could be caused by a differential induction of apoptosis on T-cell activation, we analyzed this parameter using annexin V as an apoptotic marker. Thymocytes were cultured in the presence of CD3 plus CD28, dexamethasone, or PMA plus ionomycin, and 24 h later, were assayed for apoptosis. As Fig. 4 B shows, T-cell activation via either TCR (CD3 plus CD28), or downstream of the TCR (PMA plus ionomycin) induced the same apoptotic ratios in wild-type and KO-N-ras cells. Likewise, no differences were found in the case of the induction of apoptosis by a ras-independent pathway (dexamethasone treatment).

Production of IL-2 is one of the most important and is one of the earliest events in T-cell activation in which Ras has been implicated (35). As shown in Fig. 4 C, N-ras-deficient thymocytes, stimulated with anti-CD3 plus anti-CD28 or PMA plus ionomycin, showed lower levels of IL-2 than did wild-type samples. However, whereas in N-ras-deficient cells stimulated with PMA plus ionomycin, the reduction of IL-2 production was not significant (19.3%), the same cells showed a significant reduction (90.7%) of IL-2 production when they were treated with anti-CD3 plus anti-CD28 (P < 0.05).

IL-2 promotes T-cell proliferation by binding to a high-affinity receptor composed of three transmembrane proteins (α, β, γ_c chains). Because the α chain is undetectable on resting T cells, and its expression is correlated with their proliferative responses, we analyzed its expression in stimulated wild-type and KO-N-ras thymocytes. No differences were detected in either group for the expression of IL-2Rα chain in CD4⁺ stimulated thymocytes (Fig. 4,D). Interestingly, for CD8⁺ cells, KO-N-ras mice showed significantly lower levels of IL-2Rα chain when cells were incubated with CD3 plus CD28, whereas no differences were detected when CD8⁺ cells were treated with PMA plus ionomycin (Fig. 4 D).

The alterations described for KO-N-ras mice in this work could be attributable either to decreased levels of total Ras or to the lack of an N-ras-specific function. To test the first possibility, we determined the levels of total Ras in thymi from KO-N-ras and wild-type mice. No significant differences were found between KO-N-ras and wild-type mice for the levels of total Ras (Fig. 5,A, top panel). We also compared the levels of each Ras isoform between KO-N-ras and wild-type samples (Fig. 5 A, bottom panels). Interestingly, K- and H-ras levels were increased in KO-N-ras thymi, suggesting an up-regulation of H-ras and K-ras in the thymus of N-ras-deficient mice to keep constant the levels of total Ras.

The fact that the total levels of Ras protein are not significantly altered in KO-N-ras thymi is consistent with a specific role of N-ras in T-cell function. Therefore, we analyzed the activation of the different Ras isoforms in stimulated mouse thymocytes. In wild-type thymocytes, anti-CD3 plus anti-CD28 treatment induced the activation of the three Ras isoforms as follows: N-ras > K-ras > H-ras (Fig. 5). Interestingly, the pattern of activation for H-ras and K-ras in KO-N-ras thymocytes on CD3 plus CD28 stimulation was stronger than that found among wild-type thymocytes (Fig. 5). These results suggest that there is no rescue of the N-ras-deficiency-associated phenotype by increased activation of either of the other two isoforms.

Our data indicate that N-ras performs a function in mouse thymocytes that cannot be accomplished by other Ras isoforms, even though they are equally activated on TCR stimulation. Therefore, the lack of N-ras activation could imply a defective downstream TCR signaling that is responsible for the impaired T-cell functions found in KO-N-ras thymocytes. To test the specific effect of N-ras in the downstream Ras pathways, we analyzed mitogen-activated protein kinase (MAPK), JNK, and AKT activation in mouse primary thymocytes. When wild-type cells were stimulated with anti-CD3 plus anti-CD28, all of these molecules were activated, whereas a different kinetics of activation for these three molecules was observed in KO-N-ras thymocytes (Fig. 6). Although the pattern of ERK2 activation was similar in wild-type and KO-N-ras thymocytes, a lower activation was observed in KO-N-ras-activated cells (32 and 27% attenuated signal, 2 and 5 min, respectively, postactivation). In the case of the AKT activation, wild-type and KO-N-ras thymocytes showed similar levels of activation, but a shorter duration of AKT activation was observed in N-ras-deficient cells. The maximum values for activated-AKT in wild-type cells were observed 5 min after the activation of the cells, whereas it was 2 min postactivation for the N-ras-deficient cells. Interestingly, abnormally high levels of steady-state activated JNK were detected in KO-N-ras thymocytes. In addition, no JNK activation was observed after the treatment with anti-CD3 plus anti-CD28, suggesting an inhibition of the activation of this protein by TCR cross-linking in N-ras-deficient cells.

These observations indicated that AKT, JNK, and ERK proteins showed an abnormal pattern of activation in thymocytes deficient for N-ras. Because these three molecules, located downstream of Ras, play a critical role in T-cell activation, survival, and differentiation, their abnormal activation pattern on TCR cross-linking in N-ras-deficient mice is consistent with the defective T-cell function found in these animals and helps to define the functions specifically mediated by N-ras.

Considerable evidence supports the fact that Ras proteins play a crucial role in the normal function and development of T cells. The loss of Ras function has been related to defective activation of proliferation, cytokine production, and lymphocyte development as well as with the induction of T-cell anergy (26, 27, 36). However, the differential role of Ras isoforms in T-cell function remains unexplored. In this study, using mice carrying a null mutation in the N-ras gene, we provide evidence for an important role of N-ras in T-cell immune function, early T-cell activation, and thymocyte development.

We have detected a significant reduction for the single positive CD8 thymocytes harvested either from fresh N-ras-deficient thymi or from thymocytes stimulated through the TCR. Therefore, N-ras could be a crucial Ras isoform in the signaling related to the positive selection of CD8 cells, and/or it could be involved indirectly in the development of this type of T-cells. The lack of IL-2 production that we have detected among stimulated KO-N-ras thymocytes supports the second hypothesis. Although IL-2 is not produced by CD8 single positive cells, these cells present receptors for this cytokine. Therefore, the low CD8 numbers among KO-N-ras mice might be attributable more to a defective signaling in the IL-2-secretor cells rather than to an abnormal signaling in CD8 cells themselves.

We have also demonstrated that thymocytes deficient for N-ras proliferate less than their wild-type counterparts. It is interesting to point out that this phenomenon was not caused by an increase in the induction of apoptosis, and it could be the reason why KO-N-ras mice show a lower thymus cellularity in comparison with wild-type mice (Table 1). The defective proliferation found among N-ras defective thymocytes in vitro is probably caused by a defect in the first stages of T-cell activation, because the production of IL-2 is especially low up to 20 h after activation. The low levels of IL-2Rα synthesis in KO-N-ras thymocytes on in vitro stimulation also favor a defective early activation among these cells. De novo synthesis of the α chain of the IL2-R complex is one of the first events associated with T-cell activation. In this work, we have demonstrated that the IL-2Rα chain is underexpressed in KO-N-ras thymocytes that are stimulated with anti-CD3 plus anti-CD28, but it is not underexpressed when cells are stimulated with PMA plus ionomycin. Therefore, N-ras seems to be a crucial molecule for T-cell activation and proliferation induced by TCR transactivation.

One of the most important consequences of the deficiency of N-ras that we have identified is related to the T-cell immune function. It is well known that CD4⁺ and CD8⁺ T cells are critical components of the cell-mediated immune response. This is the type of immune response triggered by viral infections. Virus-specific CD4⁺ T-cell response contributes to viral clearance by producing IL-2, which facilitates CD8⁺ T-cell activation and expansion, and by secreting IFN-γ and tumor necrosis factor α (37). Using a murine influenza pneumonia model, Eichelberger et al. (34) have demonstrated that CD8⁺ T cells play a critical role in the primary response to virus infection. Our in vivo studies showed an altered antiviral immune response in N-ras-deficient mice on influenza virus infection. KO-N-ras mice were more sensitive than wild-type mice to influenza virus infection. Interestingly, at a lower viral dose, the survival differences were higher for both groups. These results suggest that T cells are more sensitive to N-ras deficiency under low external insults or stimuli, which is similar to real-life viral infections.

N-ras deficiency is affecting not only CD8⁺ T-cell immune function but also the complex antiviral response mediated by these cells. A low production of IL-2 in infected lungs, as we observed in CD3 plus CD28-stimulated thymocytes, could imply a lower activation of several types of immune cells involved in the antiviral response. The most common defect was the reduction in cell numbers at day 5 postinfection. This hypothesis is reinforced by the fact that in vitro cytolitic activity of CTLs and NKs is normal among KO-N-ras mice. However, the possibility of a defective cytolitic activity in vivo among N-ras-deficient mice cannot be completely ruled out.

Recently, N-ras has been related to cell survival of immortalized fibroblasts (14). Moreover, the KO-N-ras immortalized fibroblasts showed a high susceptibility to the induction of apoptosis on tumor necrosis factor α and FAS treatment. In that system, it was demonstrated that N-ras-defective fibroblasts showed undetectable levels of steady-state activated AKT, and that N-ras promotes cell survival by down-regulation of JNK and p38, whereas ERK activation remained intact (14, 38). In our work, using a different cell type (primary thymocytes) we have found alterations in the activation of AKT, JNK, and ERK. In other words, both survival and proliferative pathways located downstream of Ras are affected in this cell type. It is interesting to note that, in both fibroblasts and thymocytes, neither K-ras nor H-ras were able to replace N-ras. On the other hand, in primary thymocytes, the lack of N-ras did not increase either the apoptotic rate of thymocytes treated with different mitogens or the levels of activated JNK in stimulated thymocytes. The different results obtained in fibroblasts and thymocytes defective for N-ras could be explained with the different signaling pathways studied in both cases (apoptosis/survival versus proliferation/differentiation) and/or the molecular context that is characteristic of each one of these cellular types.

In this work, we have observed that, all of the defective characteristics found in KO-N-ras thymocytes were always more pronounced when cells were stimulated through the TCR than when they were activated by a TCR-independent stimulus. These results indicate that N-ras is a key downstream molecule in TCR signaling. Because the levels of total Ras in N-ras-deficient T-cells are the same as those found in wild-type T cells, an explanation for the key role of N-ras in T-cell function could be its molecular specificity. In fact, although we have found an increase in the activation of both K-ras and H-ras in KO-N-ras thymocytes on TCR activation, this overactivation was not able to rescue the phenotype observed in N-ras-deficient thymocytes. Therefore, some molecule(s) (mainly guanine nucleotide exchange factors) that are involved in TCR signaling could bind and activate preferentially N-ras, and/or specific downstream effectors could be activated in an isoform-specific fashion by N-ras and not by K- or H-ras. Another possible explanation for the specific role of N-ras in T-cell function and development could be a differential microlocalization of N-ras with respect to the other ras isoforms.

In summary, we have demonstrated that neither K-ras nor H-ras are able to completely replace the specific role of N-ras in development of CD8 thymocytes, thymocyte proliferation, and antiviral immune response. Moreover, the results presented here indicate that although K-ras and, in a lesser manner, H-ras, as well as N-ras, can be activated by TCR stimulation, the presence of N-ras is crucial to reach T-cell activation thresholds and a proper downstream Ras signaling, especially under low antigen stimuli, which should be the more physiological situation. Our study of KO-N-ras mice reveals the crucial role of this Ras isoform in T-cell function and thymocyte development and opens the way to dissect the signal transduction pathways specifically induced by N-ras in thymocytes. Furthermore, given the key role of oncogenic N-ras in hematopoietic tumors, the conclusions derived from this work will be an important contribution to understand the molecular pathogenesis of T-cell tumorigenesis, and they will be very useful for future therapeutic strategies for these types of tumors based on the inactivation of oncogenic N-ras.

We thank Drs. Alan Frey, Sasa Radoja, Ramon Gimeno, and David Levy for their multiple suggestions, helpful discussions, and critical comments on the manuscript. We also thank John Hirst for his help with flow cytometric analysis. We are grateful to David Levy (New York University, New York, NY) and Johannes Bos (Utrecht University, Utrecht, the Netherlands) for providing reagents. We also thank Raju Kucherlapati for generously embarking on the construction of the N-ras knockout mice.