Previous experiments have shown that STAT-induced STAT inhibitor-1 (SSI-1; also named suppressors of cytokine signaling-1 (SOCS-1) or Janus kinase binding protein) is predominantly expressed in lymphoid organs and functions in vitro as a negative regulator of cytokine signaling. To determine the function of SOCS-1 in vivo, we generated SSI-1 transgenic mice using the lck proximal promoter that drives transgene expression in T cell lineage. In thymocytes expressing SSI-1 transgene, tyrosine phosphorylation of STATs in response to cytokines such as IFN-γ, IL-6, and IL-7 was inhibited, suggesting that SSI-1 suppresses cytokine signaling in primary lymphocytes. In addition, lck-SSI-1 transgenic mice showed a reduction in the number of thymocytes as a result of the developmental blocking during triple-negative stage. They also exhibited a relative increase in the percentage of CD4+ T cells, a reduction in the number of γδ T cells, as well as the spontaneous activation and increased apoptosis of peripheral T cells. Thus, enforced expression of SSI-1 disturbs the development of thymocytes and the homeostasis of peripheral T cells. All these features of lck-SSI-1 transgenic mice strikingly resemble the phenotype of mice lacking common γ-chain or Janus kinase-3, suggesting that transgene-derived SSI-1 inhibits the functions of common γ-chain-using cytokines. Taken together, these results suggest that SSI-1 can also inhibit a wide variety of cytokines in vivo.
Cytokines regulate a variety of cellular functions, including growth, differentiation, and cell death. Until now, the mechanism of the cytokine signal transduction has been fairly well characterized. Upon cytokine binding to their receptor and subsequent receptor dimerization, receptor-associated Janus kinases (JAKs)3 are brought together and become activated. JAKs then phosphorylate tyrosine residues in the cytoplasmic domains of receptors that become the binding site for Src homology domain 2 (SH2) containing proteins including a family of STATs. JAKs also phosphorylate these proteins, leading to the activation of multiple signaling pathways, most importantly the JAK-STAT pathway (1, 2).
It has been well established that cytokines play an essential role in T cell development (3). In particular, cytokines that signal through common γ-chain (γc) are especially essential for the development of T cells as well as B cells. Common γ-chain is a shared component of receptors for IL-2, IL-4, IL-7, IL-9, and IL-15 and is constitutively and specifically associated with a member of JAK family, JAK3. The critical roles of γc and JAK3 in lymphoid development have been demonstrated by the findings that mutations in either of these genes cause SCID in humans. In addition, mice lacking γc or JAK3 show a similar immunodeficient phenotype (4, 5). Thymocytes in these mice feature developmental defects during the triple-negative (TN; CD3−CD4−CD8−) stage, and consequently their thymuses are strikingly small. Despite their unresponsiveness to mitogenic stimuli in vitro, peripheral T cells in these mice spontaneously express activation markers in vivo and show an age-dependent expansion in the periphery. In addition, these animals also exhibit a developmental defect in B cells, lack of γδ T cells, and NK cells (6, 7, 8, 9, 10).
These characteristics of γc−/− and JAK3−/− mice are likely to be related to the defective signal transduction of several γc-using cytokines such as IL-2, IL-7, and IL-15. For example, a defect in early T and B cell development and severe lymphocytopenia with lack of γδ T cells are observed in mice lacking IL-7 or IL-7Rα (11, 12, 13). Spontaneous activation and expansion of peripheral T cells are observed in mice lacking IL-2, IL-2Rα, or IL-2Rβ (14, 15, 16, 17). Lack of NK cells is observed in mice deficient in IL-15Rα (18).
Recently, a new family of proteins has been identified and has been implicated in the regulation of cytokine signaling. These proteins are characterized by the presence of an SH2 domain and conserved C-terminal motifs (SC motifs; also referred to as the SOCS box) and variously called the STAT-induced STAT inhibitor (SSI) family (19, 20), the suppressor of cytokine signaling (SOCS) family (21, 22), or the cytokine-inducible SH2-containing protein (CIS) family (23, 24). Until now, eight members of this family have been identified, and their biological functions are currently being examined. Evidence to date suggests that at least three of them, CIS1, SSI-1/SOCS-1/JAB, and SSI-3/SOCS-3/CIS3, are involved in the negative feedback regulation in cytokine signaling (for a review, see Ref. 25).
Previous experiments in vitro have suggested that SSI-1, a member of this family, is a potent inhibitor of cytokine signaling. These experiments, using various cytokine-responsive cell lines, have shown that SSI-1 inhibits a wide variety of signal transduction of cytokines, such as IL-2, IL-3, IL-4, IL-6, leukemia inhibitory factor, IFNs, erythropoietin, thrombopoietin, growth hormone, and leptin (19, 21, 23, 26, 27, 28, 29). In addition, SSI-1 was shown in vitro to associate with all four members of JAKs (JAK1, JAK2, JAK3, and TYK2) and inhibit signaling by suppressing their kinase activities (19, 23, 30). These results suggest that SSI-1 can function as a general inhibitor of cytokine signaling by suppressing JAKs. On the other hand, interactions of SSI-1 with proteins other than JAKs have also been reported, such as its association with Tec and inhibition of IL-3 signaling (31) and its association with Grb-2 and Vav and inhibition of the proliferative function of stem cell factor (SCF) (32). Binding of SSI-1 to PYK2 (24) and insulin-like growth factor I receptor (33) in vitro was also reported, but their biological significance was not elucidated. These findings suggest that the functions of SSI-1 may not be limited to the inhibition of the JAK-STAT pathway.
To reveal the role of SSI-1 in vivo, SSI-1−/− mice have been generated (34, 35). Although the gross appearance of these mice is normal at birth, their growth is retarded with age. SSI-1−/− mice show progressive lymphocyte decrease due at least in part to accelerated apoptosis of lymphocytes. They also showed severe fatty degeneration in liver as well as less severe abnormalities in several organs, including the heart. In addition, all the animals died within 3 wk after birth. These results suggest that SSI-1 plays essential and nonredundant roles in normal neonatal development (34, 35).
More recently, two reports characterizing SSI-1−/− mice were published (36, 37). One report indicated that SSI-1−/− mice are hyper-responsive to exogenous as well as endogenous IFN-γ (36). The other one indicated that SSI-1 is expressed in developing thymocytes, and SSI-1 deficiency leads to the spontaneous activation of T cells (37). It is likely that both IFN-γ and T cells are critical to the disease and lethality of SSI-1−/− mice, since the inactivation of IFN-γ by introducing IFN-γ deficiency or the elimination of mature lymphocytes by introducing recombinase-activating gene 2 deficiency can prevent the phenotype of SSI-1−/− mice (36, 37). Thus, SSI-1 appears to have an essential role in the protection of mice from potentially toxic effects of IFN-γ and in the regulation of lymphocyte functions. However, considering the diverse inhibitory action of SSI-1 in vitro as well as the potential toxicity of other cytokines, such as IL-4 and TNF-α, whose overexpression can induce lymphocyte depletion and neonatal lethality in mice (38, 39), it remains to be elucidated whether signaling of cytokines other than IFN-γ is perturbed in the absence of SSI-1 in vivo.
Predominant expression of SSI-1 mRNA in lymphoid organs together with a profound lymphocyte decrease in SSI-1−/− mice suggest that SSI-1 has some essential roles in lymphocyte functions. However, functions of SSI-1 in lymphocytes in vivo have hardly been studied. For this purpose, we prepared lck-SSI-1 transgenic mice by using the lck proximal promoter that strongly drives transgene expression in T cell lineage, especially in thymocytes (40, 41). Consistent with the findings obtained in vitro, thymocytes overexpressing SSI-1 exhibited diminished activation of STATs in response to cytokines such as IL-4, IL-6, IL-7, and IFN-γ. Furthermore, lck-SSI-1 mice showed several defects in the development of T cells, which can be attributed to the impaired signaling through γc and JAK3. These results suggest that SSI-1 inhibits cytokine signaling in lymphocytes in vivo, as has been shown in various cell lines in vitro, and the main action of SSI-1 in lymphocytes is the suppression of the signaling of cytokines, including not only IFN-γ but also other cytokines, such as IL-7.
Materials and Methods
Generation and genotyping of lck-SSI-1 transgenic mice
Murine SSI-1 cDNA was inserted into the BamHI site of the p1017 expression vector containing the lck proximal promoter (41). This construct was then injected into C57BL/6 blastocysts and implanted into pseudopregnant C57BL/6 mice. Founder mice were initially screened by PCR analysis of tail DNA using primers specific for lck proximal promoter (5′-ccagtcaggagcttgaatcc-3′) and SSI-1 (5′-gcagctcgaaaaggcagtcg-3′). For four of the founder mice (lck-2, -8, -11, and -12), the expression of the SSI-1 transgene was confirmed by Northern and Western blot analysis. Animals from both lck-2 and lck-11 lines, which showed fairly strong expression of SSI-1 transgene, exhibited a similar phenotype, and those from the lck-11 line were used for the experiments described below. All mice were kept in specific pathogen-free facilities.
Cell preparation and cell counts
Single-cell suspensions were obtained from thymuses, spleens, and mesenteric lymph nodes after having been passed through mesh filters. Spleen cells were also treated with Ack buffer (0.15 M NH4Cl, 1 mM KHCO3, and 0.1 mM Na2EDTA) to lyse RBC. The total number of cells was determined by microscopic observation of trypan blue-stained cells using hemocytometers.
Western blot analysis
Thymocytes (2.5 × 106) were stimulated with specified cytokines for 30 min or were left untreated. The cytokine concentrations were 50 ng/ml for murine IL-4 (PeproTech, London, U.K.), murine IL-6 (PharMingen, San Diego, CA), and murine IL-7 and murine IFN-γ (PeproTech). Cells were solubilized with ice-cold lysis buffer containing 1% Nonidet P-40, 0.1% sodium deoxycholate, 0.1% SDS, 10 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 mM Na3VO4, and 5 μg/ml aprotinin. Whole-cell lysates were separated on SDS-PAGE under reducing conditions and transferred to nitrocellulose membranes. Membranes were probed with following Abs: anti-SSI-1(1262B) (30), anti-phospho-STAT1(Y701) (Upstate Biotechnology, Lake Placid, NY), anti-phospho-STAT3(Tyr705), and anti-phoshpo-STAT5(Tyr694) (New England Biolabs, Beverly, MA) and then reprobed with Abs as follows: anti-STAT1, anti-STAT3 (Transduction Laboratory, Lexington, KY), and anti-STAT5b (this Ab also detects STAT5a; Santa Cruz Biotechnology, Santa Cruz, CA).
Flow cytometric analysis
Cells were stained with the following Abs: FITC-, PE-, or APC-conjugated anti-CD3e, -CD4, -CD8, -CD11, -CD25, -CD44, -CD62L, and -CD69; TCRβ; γδ-TCR; pan-NK; and Gr-1 (all purchased from PharMingen, San Diego, CA). Stained cells were analyzed on a FACSCalibur (Becton Dickinson, San Jose, CA) using CellQuest software (Becton Dickinson). Live lymphocytes were gated according to their forward and side scatter profiles.
In the case of annexin V staining, freshly isolated cells were stained with anti-CD3e-PE and anti-B220-APC (PharMingen), and washed twice with PBS. Cells were further stained with annexin V-FITC (MBL, Nagoya, Japan) according to the manufacturer’s instructions. These cells were analyzed by flow cytometry as described above.
Thymocyte proliferation assay
For proliferation assays, 4 × 105 thymocytes/well in 96-well plates were cultured for 4 days in 100 μl of RPMI 1620 containing 5% FCS (Life Technologies, Tokyo, Japan), 2-ME (Nacalai Tesque, Kyoto, Japan), penicillin-G and streptomycin, and stimulated with 10 μg/ml plate-coated anti-CD3 (2C11; PharMingen) as well as with or without cytokines as indicated. The concentration used for cytokines were 20 ng/ml for murine IL-2, 20 ng/ml for murine IL-4, and 50 ng/ml for murine IL-7 (all purchased from PeproTech). Cell proliferation was determined by cell counting kit (Dojin Laboratories, Kumamoto, Japan) according to manufacturer’s instructions, and the results were shown as the absorbance at 450 nm (OD450).
Generation of lck-SSI-1 transgenic mice
To identify the in vivo function of SSI-1 in lymphocytes, we used the lck proximal promoter to express SSI-1 in T cell lineage of transgenic mice (Fig. 1⇓A).
SSI-1 transgene expression in thymus was confirmed by Northern analysis (data not shown) as well as Western blot analysis. As shown in Fig. 1⇑B, Western blot analysis showed highly expressed SSI-1 protein in lck-SSI-1 thymocytes, but not in splenocytes, which is consistent with previous findings showing that the transgene expression driven by the lck proximal promoter is strong in thymocytes, but declines in mature T cells (40).
SSI-1 inhibits the activation of STATs in primary lymphocytes
Using lck-SSI-1 thymocytes, we first checked the function of SSI-1 in cytokine signaling of lymphocytes. When thymocytes from lck-SSI-1 mice and their littermates were stimulated with several cytokines, tyrosine phosphorylation of STATs in response to IL-4, IL-6, IL-7, or IFN-γ was inhibited in lck-SSI-1 thymocytes (Fig. 1⇑C and data not shown). Thus, SSI-1 was found to suppress cytokine signaling in primary lymphocytes.
lck-SSI-1 mice show a decrease in the number of T cells
lck-SSI-1 mice appeared healthy and normal, except for their markedly small thymuses and slightly atrophic spleens. The cellularity of thymuses and spleens were 6 and 1.5 times fewer than those of the wild-type littermates, respectively (Fig. 1⇑D). Furthermore, the number of splenic T cells, estimated on the basis of the percentage of CD3-positive splenocytes, was reduced by 10-fold, even more severely reduced than the thymocytes. Thus, enforced expression of SSI-1 in T cell lineage results in a selective decrease in T cells.
Early thymocyte development is impaired in lck-SSI-1 mice
To characterize the reduction in the number of thymocytes, we analyzed lck-SSI-1 thymocytes by flow cytometry. The expression profiles of CD4 and CD8 were normal in lck-SSI-1 mice (Fig. 2⇓A) except for a slight increase in the CD4:CD8 ratio (Fig. 2⇓B). This indicates that the absolute number of lck-SSI-1 thymocytes had already been reduced at the CD4−CD8− stage. We therefore focused on immature thymocytes gating on the TN (CD3−CD4−CD8−) fraction. TN thymocytes can be divided into four subsets according to their expression of CD25 and CD44, and their maturation sequence is as follows: CD44+CD25− (pro-T1), CD44+CD25+ (pro-T2), CD44−CD25+ (pro-T3), and CD44−CD25− (pro-T4) (3). In lck-SSI-1 TN thymocytes, a decrease in the percentages of pro-T3 and pro-T4 thymocytes was seen as well as an increase in the percentages of earlier subsets, namely pro-T1 and pro-T2 cells (Fig. 2⇓C). This finding suggests that early thymocyte development is partially blocked around the pro-T2 to pro-T3 transition in lck-SSI-1 mice. In addition, the percentage of γδ T cells, which are known to be generated from pro-T2 cells (3), was reduced in lck-SSI-1 thymocytes, although the percentage of αβ T cells was comparable to that in control littermates (Fig. 2⇓D).
Peripheral T cells show an activated/memory phenotype in lck-SSI-1 mice
We next analyzed the peripheral lymphocytes of lck-SSI-1 mice. As shown in Fig. 3⇓, FACS analysis of lck-SSI-1 splenocytes and lymph node cells revealed a severe decrease in peripheral T cells. In addition, peripheral T cells of lck-SSI-1 mice showed an apparently higher CD4:CD8 ratio than control cells (Fig. 3⇓D). On the other hand, the development of B cells in lck-SSI-1 mice was normal, as judged by the expression of CD23, IgM, and IgD (data not shown), and their number was also normal, which is consistent with T cell-specific expression of SSI-1 transgene under the control of the lck proximal promoter.
Forward scatter analysis by flow cytometry revealed that splenic T cells of lck-SSI-1 mice were slightly larger than those of their control littermates (data not shown). We next analyzed the expression of T cell activation markers, CD44, CD62L, and CD69. As shown in Fig. 4⇓, increased proportions of both CD4 and CD8 splenic T cells from lck-SSI-1 mice showed an up-regulation of CD44 and CD69 (data not shown) and a down-regulation of CD62L, which are characteristics of activated/memory T cells. These results suggest that peripheral T cells in lck-SSI-1 mice are activated spontaneously in vivo.
Increased apoptosis of peripheral T cells in lck-SSI-1 mice
Despite the enhanced activation of peripheral T cells, the number of T cells in the periphery was severely reduced, as shown in Fig. 1⇑D. This decrease in peripheral T cells was also observed in older (1 year of age) lck-SSI-1 mice. To examine the cause of this reduction, we analyzed the apoptosis of lymphocytes by staining freshly prepared splenocytes with annexin V, which detects early apoptotic cells by binding to phosphatidylserine on the cell surface. As shown in Fig. 5⇓, annexin V-positive CD3+ T cells were increased in lck-SSI-1 splenocytes. This increase seemed to be specific to T cells, as the percentage of annexin V-positive B cells was not increased in lck-SSI-1 mice. Thus, lck-SSI-1 mice show an increase in apoptosis of peripheral T cells.
It has been reported that activated T cells also up-regulate the expression of the death receptor, Fas, and its ligand, Fas ligand (FasL) that leads to their apoptosis known as activation-induced cell death (AICD) (42). We therefore next examined the expression of Fas and FasL in splenic T cells of lck-SSI-1 mice (Fig. 6⇓). Consistent with the increase in the percentage of activated T cells, the expression of Fas was slightly up-regulated in lck-SSI-1 splenic T cells. In addition, we found that FasL expression in lck-SSI-1 T cells was up-regulated compared with that in littermates. These results suggest that the increased apoptosis of lck-SSI-1 T cells is partly mediated by the Fas-FasL systems.
The phenotype of lck-SSI-1 mice strikingly resembles that of mice lacking γc or JAK3
Previous reports using gene-targeting technique have revealed a wide variety of molecules that play essential roles in T cell development. Among them, the phenotype of lck-SSI-1 mice strikingly resembles that of T cells lacking γc or JAK3. Both γc−/− and JAK3−/− mice show an indistinguishable phenotype in T cells that is characterized by a defective thymocyte development with a fairly normal expression of CD4 and CD8 (except for a slight increase in the CD4:CD8 ratio), a partial developmental block around the pro-T2 and pro-T3 transition, a lack of γδ T cells and spontaneous activation, as well as increased apoptosis of peripheral T cells (6, 7, 8, 9, 10, 43, 44, 45, 46, 47, 48, 49). All these characteristics was also seen in lck-SSI-1 mice, suggesting that cytokines that signal through γc and JAK3 were inhibited in T cells of lck-SSI-1 mice. This idea is further supported by the results shown in Fig. 7⇓, demonstrating that lck-SSI-1 thymocytes were hyporesponsive not only to IL-7, but also to other γc- and JAK3-using cytokines, such as IL-2 and IL-4.
Phenotypic similarity between lck-SSI-1 mice and mice lacking γc or JAK3
In this study we demonstrated that enforced expression of SSI-1 in the T cell lineage results in impaired development of thymocytes as well as enhanced activation and apoptosis of peripheral T cells. Furthermore, this phenotype of lck-SSI-1 mice is quite similar to that of mice lacking γc or JAK3, strongly suggesting that SSI-1 can inhibit in vivo the signaling of cytokines that signal through γc and JAK3. This supposition is supported by the finding that lck-SSI-1 thymocytes showed hyporesponsiveness to γc- and JAK3-using cytokines such as IL-2, IL-4, and IL-7. (Figs. 1⇑C and 7) and is in agreement with the previous findings in vitro that SSI-1 inhibits various cytokine signaling by suppressing JAKs (25).
Recently, SSI-1 was shown to have an inhibitory role in signaling through c-Kit, a receptor for SCF (32). SCF signaling is also known to play an essential role in early thymocyte development (3). However, the phenotype of lck-SSI-1 mice is not attributable to the defect in this SCF/c-Kit pathway, since c-Kit is expressed only in pro-T1 and pro-T2 thymocytes, and mice encoding mutant c-Kit (w/w) showed a severe reduction, not a relative increase, in these most immature thymocytes (50). Thus, as denoted above, it is likely that developmental defects in lck-SSI-1 thymocytes are essentially attributable to the suppression of γc-JAK3 signaling pathways. Further studies are required, however, to rule out the possibility that defects in other signaling pathways were partly implicated in the phenotype of lck-SSI-1 mice.
Phenotypic difference between lck-SSI-1 mice and mice lacking γc or JAK3
As well as the similarities, there were several phenotypic differences between lck-SSI-1 mice and mice lacking γc or JAK3 (45, 46, 47, 48, 49). In thymus, the decrease in the number of thymocytes of lck-SSI-1 mice (∼6-fold) was less dramatic than that observed in these knockouts (>10-fold). In addition, NK cells, which were absent in γc−/− and JAK3−/− mice, were not reduced in lck-SSI-1 mice. In the periphery, although lck-SSI-1 T cells showed activated/memory phenotype, age-dependent accumulation of CD4+ T cells, which was commonly observed in γc−/− and JAK3−/− mice, was not observed in lck-SSI-1 mice.
All these differences appear to be explained by the characteristics of the lck proximal promoter. Previous reports suggested that the lck proximal promoter is not efficient for the transgene expression in most immature thymocytes, such as common lymphoid progenitors and pro-T1 cells (51). Therefore, it is conceivable that thymocytes at their most immature stage in lck-SSI-1 mice are less affected by SSI-1 transgene expression, resulting in a less severe reduction in the number of lck-SSI-1 mice. This hypothesis also explains the presence of NK cells in lck-SSI-1 mice because NK cells are known to develop from most immature thymocytes. On the other hand, it is known that the expression driven by the lck proximal promoter declines in peripheral T cells (40). In line with this, the expression of the SSI-1 transgene in the periphery was not detected by Northern (data not shown) or Western (Fig. 1⇑B) analysis, but only by RT-PCR analysis (data not shown). Thus, it is highly possible that lck-SSI-1 T cells gradually lose transgene expression after emigration from the thymus and recover their ability to respond to cytokines in the periphery. In the case of γc−/− and JAK3−/− mice, lack of IL-2 signaling appears to cause the peripheral expansion of activated T cells, since IL-2 is critical in sensitizing T cells to AICD (42). Indeed, it was previously reported that superantigen-induced deletion of peripheral T cells is impaired in γc−/− mice as a result of defective FasL expression in activated T cells (52). In this point of view, it seems significant that lck-SSI-1 splenic T cells showed increased expression of Fas and FasL (Fig. 6⇑). Thus, in lck-SSI-1 mice, although emigrants from thymus became activated in the periphery, these T cells appear to be deleted by Fas-mediated AICD after they become responsive to cytokines such as IL-2.
SSI-1 inhibits downstream signaling of JAKs
Previous in vitro analyses of CIS1, the first member of the SSI family, have suggested that CIS1 inhibits the function of STAT5 by masking the STAT5 binding site of cytokine receptors. A recent report has shown that CIS1 transgenic mice exhibit a phenotype similar to that of mice lacking STAT5a/b, suggesting that CIS1 specifically inhibits STAT5 also in vivo (53). In contrast, the phenotype of lck-SSI-1 mice is not attributable only to the inhibition of a single member of STATs, since none of the known knockout mice of STATs (including STAT1, STAT3, STAT4, STAT5a/b, and STAT6) shows an apparent defect in T cell development (54, 55, 56, 57, 58, 59, 60, 61, 62). This is in line with the previous findings in vitro that SSI-1 associates not with cytokine receptors but with JAKs and inhibits their kinase activity.
Judging from the observations that SSI-1 inhibits JAKs in vitro, close similarity between lck-SSI-1 mice and mice lacking γc or JAK3 suggests that the function of JAK3 is inhibited in lck-SSI-1 mice. It should, however, be noted that SSI-1 is not a specific inhibitor of JAK3, because the phosphorylation of STATs in response to IL-6 and IFN-γ, which require neither JAK3 nor γc for their signal transduction, was inhibited in lck-SSI-1 thymocytes. In addition, we cannot rule out the possibility that the phenotype of lck-SSI-1 mice may be attributed entirely to the suppression of JAK1, since JAK1 constitutively associates with several cytokine receptor systems, including IL-2R, IL-4R, and IL-7R, and JAK1−/− mice, although less defined than JAK3−/− mice, also exhibit a defect in early lymphoid development (63). This hypothesis can also explain the finding that the phosphorylation of STATs in response to IL-6 and IFN-γ was inhibited in lck-SSI-1 thymocytes. Thus, although the phenotype of lck-SSI-1 mice can be attributed to the inhibition of JAKs, more information on JAK1−/− mice and further studies are required to specify the precise targets of SSI-1’s action in lck-SSI-1 mice.
Functions of SSI-1 in lymphocytes
Lines of previous evidence suggest that SSI-1 is important for the regulation of lymphocyte functions. SSI-1 mRNA is predominantly expressed in lymphoid organs such as thymus and spleen (19, 21), and T cells as well as B cells themselves express SSI-1 in vivo (35). However, the function of SSI-1 in primary lymphocytes has not been fully studied to date. We have shown here that SSI-1 can efficiently inhibit cytokine signaling in primary lymphocytes also. Moreover, as the phenotype of lck-SSI-1 mice can essentially be attributed to the suppression of JAKs, the inhibition of cytokine signaling appears to be the most essential function of SSI-1 in lymphocytes.
Recently, it was reported that SSI-1 deficiency leads to a reduction in the number of lymphocytes (34, 35) and a spontaneous activation of T cells (37). Interestingly, as we showed here, enforced expression of SSI-1 in T cells also ended in T cell reduction and activation. These results indicate that inadequate expression of SSI-1 in T cells, both too little and too much, results in a similar perturbation of their development and homeostasis. In particular, loss of SSI-1 in lymphocytes appears to be critical, since mice reconstituted with SSI-1−/− lymphocytes exhibit lethality similar to that in SSI-1−/− mice (37). Thus, tightly regulated expression of SSI-1 in T cells is likely to be required to balance the effect of cytokines in vivo.
Functions of SSI-1 in cytokine signaling
It has been shown that SSI-1 in vitro suppresses a wide variety of cytokine signaling by inhibiting JAKs. Subsequent generation of SSI-1−/− mice revealed that SSI-1 deficiency leads to a complex and fatal disease characterized by progressive lymphocyte depletion, fatty degeneration of liver, and neonatal lethality within 3 wk after birth (34, 35). Surprisingly, despite the diverse action of SSI-1 in vitro, recent reports have shown that introducing IFN-γ deficiency can essentially prevent the disease and lethality of SSI-1−/− mice (36, 37). Thus, these results suggest the possibility that SSI-1 is a specific inhibitor of IFN-γ.
However, lines of evidence argue against this supposition. As described earlier, several groups have shown in various experiments in vitro that SSI-1 binds to JAKs and thereby inhibits their kinase activity (30, 64, 65). IFN-γ-specific action cannot be explained by this well-established inhibitory mechanism of SSI-1, since JAK1 and JAK2, which function as downstream kinases of IFN-γ signaling, have nonredundant roles for signaling of several cytokines other than IFN-γ (63, 66, 67). Therefore, inhibiting JAK1 or JAK2 results in the suppression of not only IFN-γ, but other cytokines also, such as IL-6 or erythropoietin. In addition to this, our present study of transgenic mice clearly suggests that SSI-1 in vivo can inhibit the signaling of several cytokines including not only IFN-γ but also others, such as IL-7.
Although our transgenic model may lead to the overestimation of SSI-1’s functions, several findings in SSI-1−/− mice suggest that SSI-1 is not specific to IFN-γ signaling. It was previously demonstrated that SSI-1−/− thymocytes proliferate more vigorously than SSI-1+/+ thymocytes in response to anti-CD3 plus IL-2 or IL-4. In line with this, SSI-1−/− thymocytes showed markedly sustained tyrosine phosphorylation of STAT6 in response to IL-4 (34). Recently, it was shown that SSI-1−/− splenocytes exhibit enhanced proliferation in response to IL-2 alone, suggesting their hyper-responsiveness to IL-2 (37). This hyper-responsiveness to cytokines such as IL-2 may be a reason for the spontaneous activation of SSI-1−/− T cells in vivo. In addition, interestingly, our preliminary data showed that SSI-1/STAT6 double-knockout mice were partially rescued from the disease and the neonatal lethality seen in mice lacking SSI-1 alone (our unpublished observations). As STAT6 is an essential molecule for IL-4 and IL-13 signaling (60, 61, 62, 68), this result suggests that IL-4, IL-13, or both have an additional or synergistic effect on the toxicity of IFN-γ in SSI-1−/− mice. This result also raises the possibility that the complex action of other cytokines, which might be masked by the devastating toxicity of IFN-γ, is partly implicated in the phenotype of SSI-1−/− mice. Taken together, we propose here that SSI-1 has the potential to inhibit the signaling of multiple cytokines in vivo. Further studies are currently underway to elucidate the diversity of SSI-1 function in vivo.
We thank Dr. R. M. Perlmutter for providing p1017 vector, K. Sato for excellent maintenance of mice, Dr. H. Fujiwara for helpful discussion, and R. Harada, N. Kameoka, and M. Tagami for secretarial assistance.
↵1 This work was supported by grants from the Ministry of Education, Science, and Culture, Japan, and the Osaka Foundation for Promotion of Clinical Immunology.
↵2 Address correspondence and reprint requests to Dr. Tadamitsu Kishimoto, Osaka University, 1-1 Yamada-oka, Suita City, Osaka 565-0871, Japan. E-mail address:
↵3 Abbreviations used in this paper: JAK, Janus kinase; AICD, activation-induced cell death; APC, allophycocyanin; CIS, cytokine-inducible SH2-containing protein; CD62L, CD62 ligand; γc, common γ-chain; FasL, Fas ligand; SCF, stem cell factor; SH2, Src homology domain 2; SOCS, suppressor of cytokine signaling; SSI, STAT-induced STAT inhibitor; TN, triple negative.
- Received February 29, 2000.
- Accepted May 30, 2000.
- Copyright © 2000 by The American Association of Immunologists