Jak and STAT activation by type I IFN
The ligand mediated bridging of the IFN receptor chains brings into close proximity their intracellular domains and the associated Jaks, i.e. Tyk2 and Jak1. Based on studies with the Tyk2-deficient 11,1 cell line reconstituted with a kinase inactive Tyk2 mutant (Tyk2 K930R), a temporal order of activation was proposed, whereby Jak1 is activated first and then it activates Tyk2 by transphosphorylation. However, Tyk2 ÔbasalÕ kinase activity is still needed for full Jak1 activation. Interestingly, IFN!/∀ has the potential to activate all the STATs present in a given cell. However, the ÒclassicalÓ response to IFN!/∀ refers to activation of STAT1 and STAT2, binding of IRF9 to forme the trimeric transcriptional complex known as ISGF3 (IFN-stimulated gene factor 3). ISGF3 binds to ISRE elements in the promoter of target genes. IFN!/∀ can also induce the activation of STAT1 and STAT3 homodimers as well as STAT1:STAT3 heterodimers, which all bind to GAS containing sequences.
Numerous studies were undertaken to try to determine the exact mechanism and order of STAT activation upon IFN treatment. I will summarize schematically the principal results obtained by different groups and propose a general model of IFN!/∀-induced STAT activation. These informations were gathered from a large number of published work performed on cell lines either stably or transiently transfected, in in vitro pulldown assays, with endogenous or chimeric receptors.
Type I IFN and SLE
Systemic lupus erythematosus (SLE) is a multisystem autoimmune disease characterized by tissue damage resulting from the deposits of immune complexes and the presence of anti-nuclear antibodies. Several sets of data suggest an important pathogenic role for IFNs in SLE (Banchereau and Pascual, 2006; Crow, 2007). IFN! is present in particularly high levels in SLE patient serum. Therapeutic administration of IFN! to patients with viral infection or malignancy occasionally results in induction of typical lupus antibodies and, in some cases, clinical lupus. This indicated that, given the appropriate genetic background and perhaps in the setting of concurrent stimuli, SLE could be induced or maintained by IFN. Using microarray technology several groups have demonstrated that IFN-induced genes are among the most prominent observed in peripheral blood cells of lupus patients. Additional data have proposed that a key pathogenic event in SLE can be a break in peripheral tolerance mechanisms after activation of myeloid dendritic cells in response to IFN/∀. Murine studies have also supported a role for type I IFN in SLE. For example, both New Zealand Black (NZB) and B6/lpr lupus-susceptible mice deficient in functional IFN receptor show significantly less severe manifestations of autoimmunity as well as decreased renal disease and improved survival. One significant question that remains to be answered is wheter the overexpression of IFN is the primary abnormality contributing to development of disease or is it produced only after antibodies and immune complexes have formed. Available data suggest that IFN can act both at the initiating as well as later steps.
Genetic contributions to variability among individuals in production and signaling of IFN have been suggested by recent investigations. A study performed to identify single nucleotide polymorphisms (SNPs) in a group of IFN pathway genes found statistically significant associations of IRF5 (IFN-response factor 5) and Tyk2 SNPs with a diagnosis of SLE (Graham et al., 2007; Sigurdsson et al., 2005). IRF5 is a gene encoding a transcriptional factor that has been implicated in TLR (Toll-like receptor) signaling and IFN production. Two SNPs have been found in Tyk2 leading to a V362F substitution in the N-terminus and to I684S in the KL domain. The functional consequence of these Tyk2 and IRF5 variants remains to be determined.
Cytokines utilizing the common IL-10R2: IL-10, IL-22, IL-26 and IFN
The IL-10 cytokine family comprises several viral and cellular homologs, including IL-19, IL-20, IL-22, IL-24 and IL-26. IL-28/29 (IFN) have been classified within this family. However, given their antiviral activity, which is by definition the common feature of all IFNs, it seems more appropriate to classify IFN as a distinct family (type III IFNs). The receptors for all of these cytokines are made of components belonging to the class II cytokine receptor family.
IL-10 is a cytokine with broad anti-inflammatory properties that result from its ability to inhibit the function of both macrophages and dendritc cells, including their production of pro-30 inflammatory cytokines (O’Garra and Vieira, 2007). IL-10 has also been reported to suppress the differentiation of both Th1 and Th2 cells. A number of genes in all these cells are inhibited to some extent by IL-10 mediated STAT3 signaling, but the mechanisms responsible for the suppressive activities of IL-10 are still unclear. IL-10 is a noncovalent symmetric homodimer that binds to a receptor complex consisting of IL-10R1 and IL-10R2 chains, associated to Jak1 and Tyk2, respectively. While IL-10R2 is expressed on a broad range of cells, the expression IL-10R1 is restricted to leukocytes. The major downstream effector of IL-10 signaling is STAT3, which is recruited on phosphorylated tyrosine residues of IL-10R1 upon IL-10 treatment. In addition to STAT3, STAT1 and STAT5 can also be activated in response to IL-10, but their precise role in IL-10 signaling remains unclear (Dumoutier and Renauld, 2002). The IL-10R2 cytoplasmic domain lacks tyrosine residues. IL-22 and IL-26 are IL-10 Ð related cytokines preferentially produced by Th17 and Th1 cells, respectively (Donnelly et al., 2004; Liang et al., 2006; Zenewicz et al., 2007). Their receptors contain the shared IL-10R2 subunit together with a specific IL-22R1 (for IL-22) or IL-20R1 (for IL-26) chain, both of which activate Jak1. Both IL-22 and IL-26 activate strongly STAT3, and to a lesser extent STAT1. In addition, IL-22 can activate ERK and p38 signaling pathways (Lejeune et al., 2002). In contrast to the IL-10R, the receptors of IL-22 and IL-26 have only been found on nonhematopoietic tissues such as colon, liver, lung and skin.
Type III IFNs are a novel family of antiviral cytokines. They comprise IFN1 (IL-29), IFN2 (IL-28A) and IFN3 (IL-28B) (Uze and Monneron, 2007). The IFN receptor consists of a specific IFNLR1 chain (also termed IL28R) and the shared IL-10R2. Type III IFNs can induce the phosphorylation of STAT1, STAT2, STAT3 and STAT5 in lung carcinoma and colorectal adenocarcinoma cell lines (Dumoutier et al., 2003; Kotenko et al., 2003) and STAT4 in BW5147 T lymphoma cells (Dumoutier et al., 2004). Type III IFNs are the only cytokines other than IFN/∀ that can induce activation of STAT2 and the formation of ISGF3 complex. However, while IFNLR1 was shown to be expressed in most organs, it is certainly not expressed in all cell types and appears to be subtly regulated. The restricted expression pattern of IFNLR1 and the ensuing restricted IFN responsiveness is a major difference with the type I IFN system.
Physiopathological consequences of Tyk2 deficiency
The functional implication of Tyk2 in signaling pathways activated by the various cytokines mentioned above has been confirmed by studies performed in Tyk2 knock-out mice and by identification of a Tyk2-deficient patient.
Tyk2 knock-out mice
Initial studies reported a surprisingly mild phenotype of Tyk2 knock-out mice. Tyk2-/- mice are viable and fertile, with no overt development abnormalities, but with increased susceptibility to vaccinia virus and LCMV (lymphocytic choriomeningitis virus). IFN! response is impaired, but not abolished, whereas IL-12-driven Th1 differentiation is completely absent. In contrast to the finding in 11,1 cells (see above), IFNAR1 was normally expressed in MEFs lacking Tyk2. The absence of Tyk2 had no effect on Tpo, IL-6, LIF, IL-3 and G-CSF and, surprisingly, on IL-10 responses. Further characterization of Tyk2-/- mice revealed a critical role of Tyk2 in mediating LPS-induced endotoxin shock. A role of Tyk2 in allergic airway hypersensitivity was also described. Subsequently, Tyk2 was shown to be critical for optimal IL-10 Ð mediated signaling. Thus, Tyk2 likely plays a regulatory role broadly impacting innate and adaptive phases of both Th1- and Th2-mediated immune responses. In term of STAT activation, the phosphorylation of STAT3 was the most affected in Tyk2-/- mice, both in response to IL-12 and IFN!/∀, suggesting a preferential interplay between Tyk2 and STAT3.
Tyk2 has also been implicated in the surveillance of B lymphoid tumors (Stoiber et al., 2004). It was shown that Tyk2-/- mice develop Abelson-induced B lymphoid leukemia/lymphoma with a higher incidence and shortened latency compared with control mice. The cell-autonomous properties of transformed cells were unaltered, but the high susceptibility of Tyk2-/- mice resulted from an impaired tumor surveillance. The increased rate of leukemia/lymphoma formation was linked to a decreased in vitro cytotoxic capacity of Tyk2-/-NK cells toward tumor-derived cells. The role of Tyk2 in tumor surveillance was confirmed in a model of E-Myc transgenic mice where B lymphomas arise slowly and require an additional genetic alteration or hit, shown to be either overexpression of the anti-apoptotic protein Bcl-2 or alteration of p53-dependent signaling (Sexl, 2007). In the case of Bcl-2 overexpression as the second hit, the disease latency was found to be shorter in Tyk2-/-animals. However, Tyk2 deficiency did not have any impact on tumors that displayed impaired p53 signaling. These data suggest an important role for Tyk2 in tumor surveillance.
Human Tyk2 deficiency
Recently, a single patient with Tyk2 deficiency has been described (Minegishi et al., 2006). The patient exhibited broader and more profound immunological defects than could be anticipated from studies of Tyk2-/- mice. The signaling defects resulted in a complex clinical picture that included Hyper IgE Syndrome (HIES) and susceptibility to multiple infectious pathogens. The patient T lymphocytes displayed impaired response to IFN!/∀, IL-12, IL-23, IL-10, and, unexpectedly, to IL-6. The response to IFN-∃ was reduced compared with a healthy individual, presumably due to lower amount of STAT1. The patient’s increased susceptibility to viral infections pointed out the critical role of Tyk2 in transmitting signals from the IFN!/∀ receptor. IFN!/∀-induced phosphorylation of Jak1, STAT1, STAT2, STAT3 and STAT4 was completely abrogated in patientÕs T cells. Furthermore, the surface expression of IFNAR1 was decreased, confirming the role of Tyk2 in the proper trafficking of IFNAR1. The patient also suffered from atypical mycobacterial infections as a consequence of deficits in the IL-12 and IFN∃ axis. Lack of IL-12 signaling resulted in impaired Th1 differentiation and IFN∃ production. Concomitantly, and in concert with the absence of suppressive IL-10 signaling, Th polarization was skewed toward the Th2 phenotype, presenting exaggerated in vitro Th2 differentiation with increased production of IL-5 and IL- 13. This led to another feature of the patientÕs disease – atopic dermatitis and elevated level of IgE. The patient cells showed also abrogated IL-23 signaling, but the significance of this impairement on Th17 maintenance was not examined. Another difference between humans and mice is the responsiveness to IL-6. While BMM (bone marrow macrophages) and EF from Tyk2-/- mice displayed normal IL-6 induced signaling, PBMCs from the Tyk2 deficient patient showed a partial Tyk2-dependence of IL-6 response. What is the exact role of Tyk2 in IL-6 signaling, and whether Tyk2 plays a role in the response to other cytokine utilizing the gp130 chain, remains to be elucidated.
In conclusion, these results establish the critical role of Tyk2 in humans. However, the clinical picture presented by this patient does not seem very common. In fact, the authors point out that this genetic lesion may be the determinant for only a subset of patients with autosomal-recessive (AR) HIES. In most sporadic and autosomal-dominant (AD) cases, these clinical manifestations are part of a multisystem disorder including abnormalities of the soft tissue, skeletal and dental systems. Interestingly, dominant-negative mutations of STAT3 have been found recently at the origin of sporadic and AD HIES (Holland et al., 2007; Minegishi et al., 2007; Renner et al., 2007). Defects in Tyk2 signaling seem to coincide well with observed physiological defects in STAT3 activation. However, the consequences of STAT3 impaired function seem not to be restricted to the immune system, as observed for Tyk2, but extend to the whole organism. These findings confirm once more the preferential rapport of Tyk2 with STAT3.
Table of contents :
LIST OF ABREVIATIONS
LIST OF FIGURES
1. CYTOKINE SIGNALING
1.1. !-helical cytokines and their receptors
1.1.1. Erythropoietin and interferon !: two representative !-helical cytokines
1.1.2. Class I and class II receptors
1.2. The Jak/STAT signaling pathway
1.2.1. The Jak tyrosine kinase family
1.2.2. The STAT transcription factors
1.3. Signal termination/downmodulation
1.3.1. The SOCS family of cytokine-signaling repressors, the PIAS family of STAT inhibitors and Jak-STAT
1.3.2. Cytokine response modulation through availability of signaling components
2. TYK2 IN CYTOKINE SIGNALING
2.1. Type I IFNs
2.1.1. The type I IFN receptor
2.1.2. Jak and STAT activation by type I IFN
2.1.3. Tyk2 as a chaperone
2.1.4. Type I IFN and SLE
2.2. Cytokines utilizing the common IL-10R2: IL-10, IL-22, IL-26 and IFN »
2.3. Cytokines utilizing IL-12R#1: IL-12 and IL-23
2.4. Physiopathological consequences of Tyk2 deficiency
2.4.1. Tyk2 knock-out mice
2.4.2. Human Tyk2 deficiency
3. STRUCTURE/FUNCTION ORGANIZATION OF THE JAK KINASES
3.1. The N-terminal region: FERM and SH2-like domains
3.1.1. A new Tyk2 interacting protein: Jakmip1
3.2. The kinase-like domain as a sensor of ligand binding
3.2.1. Jak2V617F in Polycythemia vera
3.3. The tyrosine kinase domain
The role of Pot1 in IFN! signaling
The Tyk2V678F mutant in a homo- vs heterodimeric receptor context
The effect of the P1104A on Tyk2 activity
MATERIALS AND METHODS
siRNA and plasmids
SDS-PAGE and Western blot
Tyk2 immunoprecipitation and in vitro kinase assay
Luciferase reporter assay
POT1: A NEW TYK2 INTERACTING PROTEIN
Database analyses of Pot1 mRNA transcripts
Mapping of Pot1 transcripts
Subcellular localisation of the murine Pot1
Functional studies of Pot1
Identification of Pot1 interactors by yeast two-hybrid screen
The role of GIT1 in IFN! signaling.
V678F, an activating mutation of Tyk2
Tyk2V678F basal phosphorylation in vivo and in vitro
The V678F mutant leads to basal STAT3 phosphorylation but normal IFN! induced signaling
Analysis of the Tyk2V678F mutant placed in a homodimeric receptor complex
Equivalent basal STAT3 phosphorylation level in 11,1 and EpoR/R1 clones
Analysis of the Tyk2P1104A mutant
Impaired in vivo auto/transphosphorylation of Tyk2 P1104A
Tyk2 P1104A rescues IFN! signaling
Tyk2P1104A cannot auto/transphosphorylate itself in vitro
Pot1 isoforms and localization
Pot1 is not implicated in the IFN!-induced transcriptional response in 293T cells
Pot1 interacting proteins
The regulatory role of the KL domain
Tyk2 loss-of-function mutations
Tyk2V678F: a gain-of-function phenotype
Tyk2V678F has no effect on IFN!-induced signaling, but leads to basal STAT3 phosphorylation
Preferential Tyk2-STAT3 interaction
The prerequisite of a homodimeric receptor for Jak2V617F-mediated transformation
Tyk2V678F placed in a homodimeric receptor context confers ligand hypersensitivity
The effect of Tyk2V678F on STAT3 basal phosphorylation is not linear
A general model of IFN-induced STAT activation
Jak2V617F- and Tyk2V678F-mediated STAT5 activation
Physiological consequences of constitutively active Tyk2