Activation of Mitogen-activated Protein Kinase Kinase (MKK) 3 and MKK6 by Type I Interferons*

There is accumulating evidence that the p38 MAP kinase pathway plays important roles in Type I interferon (IFN) signaling, but the mechanisms regulating p38 activation during engagement of the Type I IFN receptor remain to be defined. We sought to identify the events that lead to activation of the p38 MAP kinase in response to Type I IFNs. Our data demonstrate that treatment of sensitive cell lines with IFNα results in activation of both MAP kinase kinase 3 (MKK3) and MAP kinase kinase 6 (MKK6). Such IFN-inducible activation of MKK3 and MKK6 is essential for downstream phosphorylation and activation of the p38 MAP kinase, as shown by studies using mouse embryonic fibroblasts (MEFs) with targeted disruption of the Mkk3 and Mkk6 genes (MKK3–/– MKK6–/–). Similarly, IFN-dependent activation of the downstream effectors of p38, MAPKAPK-2 and MAPKAPK-3, is not detectable in cells lacking Mkk3 and Mkk6, demonstrating that the function of these MAP kinase kinases is required for full activation of the p38 pathway. To define the functional relevance of MKK3/6 engagement in Type I IFN signaling, IFN-inducible gene transcription was evaluated in the MKK3/MKK6 double knock-out cells. IFNα- and IFNβ-dependent transcription via either interferon-stimulated response element or IFNγ activated site elements was defective in MKK3 –/–/MKK6 –/– MEFs in luciferase reporter assays. In addition, IFN-dependent induction of two genes known to be of importance in the generation of IFN responses, Isg15 and Irf-9, was diminished in the absence of Mkk3 and Mkk6. The effects of Mkk3 and Mkk6 on IFN-dependent transcription were unrelated to any effects on the phosphorylation and activation of STAT proteins, indicating the presence of a STAT-independent mechanism. Altogether, our findings demonstrate that MKK3 and MKK6 are rapidly activated during engagement of the Type I IFN receptor and play important roles in Type I IFN signaling and the generation of IFN responses.

There is accumulating evidence that the p38 MAP kinase pathway plays important roles in Type I interferon (IFN) signaling, but the mechanisms regulating p38 activation during engagement of the Type I IFN receptor remain to be defined. We sought to identify the events that lead to activation of the p38 MAP kinase in response to Type I IFNs. Our data demonstrate that treatment of sensitive cell lines with IFN␣ results in activation of both MAP kinase kinase 3 (MKK3) and MAP kinase kinase 6 (MKK6). Such IFN-inducible activation of MKK3 and MKK6 is essential for downstream phosphorylation and activation of the p38 MAP kinase, as shown by studies using mouse embryonic fibroblasts (MEFs) with targeted disruption of the Mkk3 and Mkk6 genes (MKK3؊/؊ MKK6؊/؊). Similarly, IFN-dependent activation of the downstream effectors of p38, MAP-KAPK-2 and MAPKAPK-3, is not detectable in cells lacking Mkk3 and Mkk6, demonstrating that the function of these MAP kinase kinases is required for full activation of the p38 pathway. To define the functional relevance of MKK3/6 engagement in Type I IFN signaling, IFNinducible gene transcription was evaluated in the MKK3/MKK6 double knock-out cells. IFN␣-and IFN␤-dependent transcription via either interferon-stimulated response element or IFN␥ activated site elements was defective in MKK3 ؊/؊/MKK6 ؊/؊ MEFs in luciferase reporter assays. In addition, IFN-dependent induction of two genes known to be of importance in the generation of IFN responses, Isg15 and Irf-9, was diminished in the absence of Mkk3 and Mkk6. The effects of Mkk3 and Mkk6 on IFN-dependent transcription were unrelated to any effects on the phosphorylation and activation of STAT proteins, indicating the presence of a STAT-independent mechanism. Altogether, our findings demonstrate that MKK3 and MKK6 are rapidly activated during engagement of the Type I IFN receptor and play important roles in Type I IFN signaling and the generation of IFN responses.
There is accumulating evidence indicating that MAP kinase signaling pathways play important roles in IFN␣ and IFN␤ signaling (7). Specifically, engagement of the p38 MAP kinase pathway has been implicated in the regulation of interferondependent gene transcription via ISRE and GAS elements (22)(23)(24), whereas there is also evidence that the p38 pathway regulates Type I IFN-dependent induction of protein expression via post-transcriptional mechanisms (28). Importantly, it has been directly demonstrated that activation of the p38 and its downstream effectors is essential for the generation of the suppressive effects of Type I IFNs on normal and leukemic progenitors (25,26), as well as for the induction of antiviral responses (27).
Despite the significant advances in our understanding of the contribution of MAP kinase pathways in the generation of IFN␣ responses, the mechanisms by which different Type I IFNs regulate activation of the p38 MAP kinase pathway are not known. There is some evidence that the small G-protein Rac1 is activated downstream of JAK kinases (24,26) and that such activation is essential for ultimate activation of the p38 MAP kinase (24). However, the kinases that directly phosphorylate and activate p38 are not known. In the present study we provide evidence that IFN␣ or IFN␤ treatment of sensitive cell lines results in activation of MAP kinase kinase 3 (MKK3) and MAP kinase kinase 6 (MKK6). Such activation is critical for activation of the p38 MAP kinase, as in cells with targeted disruption of both the Mkk3 and Mkk6 genes there is complete abrogation of the ability of Type I IFNs to induce p38 activation. Our data also establish that MKK3 and MKK6 are required for ISRE-and GAS-driven gene transcription in response to Type I interferons (IFN␣ and IFN␤). Type I IFNdependent tyrosine phosphorylation of STAT1 and STAT3, STAT1 phosphorylation on serine 727, and formation of STAT DNA-binding complexes are normally formed in the absence of MKK3/MKK6, establishing that these kinases do not modulate activation of STATs directly or indirectly via p38 MAP kinase activation. On the other hand, IFN-inducible activation of MAPKapK-2 is MKK3/6-dependent and interferon stimulated gene (ISG) expression is defective in MAPKAPK-2Ϫ/Ϫ MEFs, indicating that this kinase acts as a downstream effector for the MKK3/6-p38 MAP kinase pathway, to mediate transcriptional regulation of ISGs.
Cell Lysis, Immunoprecipitations, and Immunoblotting-Cells were stimulated with the indicated IFNs (10 4 IU/ml) for the indicated times, and lysed in phosphorylation lysis buffer as previously described (11)(12)(13)31). Immunoprecipitations and immunoblotting using an enhanced chemiluminescence (ECL) method were performed as previously described (11)(12)(13)31). Each immunoblot shown is representative of at least two independent experiments (Figs. 1 and 3). In most cases the blots represent three or more independent experiments for each panel (three for Figs. 4, 8, 9, C-F, and 11, A-D, and four independent experiments for Fig. 9, A and B), and all were highly reproducible.
Mobility Shift Assays-Actively growing cells were treated with IFN␣ for the indicated times. Ten g of nuclear extracts from IFN␣treated or untreated cells were analyzed using electrophoretic mobility shift assays (EMSAs), as described previously (18,22). A doublestranded oligodeoxynucleotide (ATTTCCCGTAAATCCC), representing a sis-inducing element (SIE) from the c-fos promoter was synthesized and used in the gel shift assays. A double-stranded oligodeoxynucleotide (CTGTTGGTTTCGTTTCCTCAGA), representing an ISRE element from the Isg-15 gene, was also synthesized and used to detect ISGF3 complexes.
In Vitro Kinase Assays-Cells were incubated in the presence or absence of IFN␣ for the indicated times at 37°C. The cells were then lysed with phosphorylation lysis buffer, equal amounts of total cell lysates were immunoprecipitated with antibodies against the indicated kinases or control non-immune rabbit immunoglobulin (IgG). In vitro kinase assays were subsequently carried out on the immunoprecipitates, essentially as previously described (22,27). Briefly, immunoprecipitated proteins were washed three times in phosphorylation lysis buffer containing 0.1% Triton and two times in kinase buffer ( Total cell lysates were resolved by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated forms of MKK3 and MKK6 on serines 189/207. B, the blot shown in A was stripped and re-probed with an antibody against MKK3. C, the same blot was stripped again and re-probed with an antibody against tubulin, to control for protein loading. D, Molt4 cells were incubated with or without human IFN␣ for the indicated times. Total cell lysates were resolved in SDS-PAGE and immunoblotted with an antibody against the phosphorylated form of MKK3/6 on serine 189/207. E, the blot shown in D was stripped and re-probed with an antibody against MKK3. F, the same blot shown in D and E was stripped again and re-probed with an antibody against tubulin, to control for protein loading.

FIG. 2. Activation of the MKK3 and MKK6 kinases by IFN␣.
KT-1 cells (panels A and C) or Molt4 cells (panels B, D, E, and F) were incubated in the presence or absence human IFN␣ for the indicated times. Total cell lysates were immunoprecipitated with antibodies against either MKK3 (A and B) or MKK6 (C, D, and E) or non-immune rabbit immunoglobulin (RIgG), as indicated. Subsequently, immune complex kinase assays were carried out on the immunoprecipitates using recombinant GST-p38␣ as an exogenous substrate. Proteins were resolved by SDS-PAGE and the phosphorylated form of p38␣ was detected by autoradiography. The Immobilon membrane from the experiment shown in E was subsequently immunoblotted with an anti-MKK6 antibody (F).
complex kinase assays were initiated by the addition of 30 l of kinase buffer containing 3 g of recombinant protein as a substrate and 10 Ci of [␥-32 P]ATP. Recombinant GST-p38␣ (Upstate Biotechnology) was used as a substrate for the MKK3 and MKK6 kinase assays. Hsp-25 protein (StressGen Laboratories) was used as substrate of MAPKAPK-2 and MAPKAPK-3. Reactions were incubated for 30 min at room temperature and were terminated by the addition of SDS sample buffer. Proteins were resolved by SDS-PAGE and phosphorylated substrate was detected by autoradiography. All in vitro kinase assay experiments were highly reproducible and each panel shown represents at least two to three independent experiments.
Luciferase Reporter Assays-Cells were transfected with a ␤-galactosidase expression vector and either an ISRE-luciferase construct or a luciferase reporter gene containing eight GAS elements linked to a minimal prolactin promoter (8ϫ GAS) using the Superfect transfection reagent as per the manufacturer's recommended procedure (Qiagen). The ISRE-luciferase construct (22) included the wild-type Isg15 ISRE (TCGGGAAAGGGAAACCGAAACTGAAGCC) cloned via cohesive ends into the BamHI site of the pZtkLuc vector, and was provided by Dr. Richard Pine (Public Health Research Institute, New York, NY). The 8ϫ GAS construct (34) was kindly provided by Dr. Christofer Glass (University of California San Diego, San Diego, CA). 48 h after trans-fection, triplicate cultures were either left untreated or treated with IFN␣ or IFN␤ as indicated.
Quantitative RT-PCR (TaqMan)-Cells were treated with IFN␣ for 6 h. RNA was isolated using the RNeasy kit (Qiagen). 1 g of total RNA was reverse transcribed to cDNA using the OmniScript RT kit and Oligo(dT) primer (Qiagen). Real-time reverse transcriptase-PCR of the mouse Isg15 and Irf-9 genes was carried out by an ABI 7900 Sequence Detection System (Applied Biosystems, TaqMan Assays on Demand) (62) using FAM-labeled probes and primers. Relative quantitation of mRNA levels was plotted as -fold increase compared with untreated cells. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used for normalization (63). Briefly, ⌬C T values (target gene C T minus Gapdh C T ) for each triplicate sample were averaged, and ⌬⌬C T was calculated as previously described. mRNA amplification was determined by the formula 2 Ϫ⌬⌬CT (64).
Chromatin Immunoprecipitation (ChIP) Assays-ChIP assays were performed essentially as previously described (27,32,33). ChIP DNA was used as a template for PCR with the Isg15 gene ISRE forward primer (FP) (5Ј-gggacctaaaggtgaagatg-3Ј) and reverse primer (RP) (5Јcctcatagatgttgctgtgg-3Ј), which amplifies a 270-bp fragment of the murine Isg15 gene ISRE. Equal amounts of total cell lysates (100 g/ lane) were analyzed by SDS-PAGE, and immunoblotted with an antibody that recognizes the phosphorylated form of MKK3 on serine 189 or the phosphorylated form of MKK6 on serine 207. B, the blot shown in A was re-probed with an anti-tubulin antibody, as control for protein loading. C, the same blot shown in A and B was stripped and re-probed with an anti-MKK6 antibody. D, similar experiment to the one shown in A-C, except that after SDS-PAGE analysis, anti-MKK3 immunoblotting was performed. Total cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of p38 on Thr 180 and Tyr 182 . B, the blot shown in A was stripped and re-probed with an antibody against p38, to control for protein loading. C, the indicated MEFs were incubated for 30 min in the presence or absence of anisomycin, as indicated. Total cell lysates were resolved by SDS-PAGE and immunoblotted with an anti-phospho-p38 antibody. D, the blot shown in C was stripped and re-probed with an antibody against p38, to control for protein loading. E, MKK6Ϫ/Ϫ MEFs and MKK3Ϫ/Ϫ MK-K6Ϫ/Ϫ MEFs were incubated for 30 min in the presence or absence of mouse IFN␣ as indicated. Total cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of p38 on Thr 180 and Tyr 182 . F, the blot shown in E was stripped and re-probed with an antibody against p38, to control for protein loading.

RESULTS
We initially examined whether treatment of sensitive cell lines with IFN␣ results in phosphorylation of the MKK3 and/or MKK6 kinases. KT-1 or Molt-4 cells were incubated for different times in the presence or absence of IFN␣, and cell lysates were resolved by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated forms of MKK3 on serine 189 and/or MKK6 on serine 207. IFN␣ treatment resulted in strong phosphorylation of MKK3/6 in both Molt-4 and KT-1 cells (Fig. 1, A and D), suggesting that these kinases act as substrates for an upstream IFN-regulated kinase(s) and may be activated to participate in the generation of Type I IFN signals. In subsequent studies, we directly examined whether the kinase domains of MKK3 and MKK6 are activated in an IFN␣-dependent manner. Cells were treated with IFN␣, cell lysates were immunoprecipitated with anti-MKK3 or anti-MKK6 antibodies, and in vitro kinase assays were carried out on the immunoprecipitates using recombinant p38 as an exogenous substrate. IFN␣ treatment resulted in induction of the kinase domains of both MKK3 (Fig. 2, A and B) and MKK6 (Fig.  2, C and D), as reflected by the phosphorylation of exogenous p38 in the in vitro kinase assays. Thus, IFN␣ induces activation of both MKK3 and MKK6, suggesting that these kinases are engaged in IFN signaling and regulate activation of downstream cellular pathways.
To define the roles of MKK3 and MKK6 in IFN signaling, we used MEFs with targeted disruption of the Mkk3 gene, or both the Mkk3 and MKK6 genes (31). We initially examined whether mouse IFN␣ induces phosphorylation/activation of MKK3/6 in parental MEFs. As shown in Fig. 3, IFN␣-inducible phosphorylation of MKK3/6 was clearly detectable in wild-type MEFs (Fig. 3, A and B), whereas in single MKK3 knock-out MEFs there was residual phosphorylation detected in the anti-Ser 189/ 207 immunoblots, reflecting the presence of MKK6 (Fig. 3, A  and C). On the other hand, in double knock-out MKK3Ϫ/Ϫ MKK6Ϫ/Ϫ MEFs there were no bands corresponding to the phosphorylated forms of either MKK3 and/or MKK6, consistent with the lack of expression of both proteins (Fig. 3, A, C, and D).
To examine whether MKK3 and/or MKK6 are required for Type I IFN-dependent activation of p38, the various MEFs were subsequently incubated with mouse IFN␣, the cells were lysed, and total cell lysates were analyzed by SDS-PAGE and immunoblotted with an anti-phospho-38 antibody that recognizes the phosphorylated form of p38 on threonine 180 and tyrosine 182. As expected (22,24,27), IFN␣ induced strong phosphorylation of p38 in wild-type MEFs that express both MKK3 and MKK6 (Fig. 4, A and B). IFN-dependent phosphorylation was also detectable in the MKK3Ϫ/Ϫ single knock-out MEFs (Fig. 4, A and B) and MKK6Ϫ/Ϫ knock-out MEFs (Fig. 4, E and F). However, in cells lacking expression of both MKK3 and MKK6 there was complete lack of phosphorylation of p38 by IFN␣ (Fig. 4, A, B, E, and F), whereas the levels of p38 protein expression were similar to the levels of the protein in wild-type MEFs (Fig. 4, A and B). Similarly, the induction of phosphorylation of p38 in response to anisomycin was defective in MKK3Ϫ/Ϫ MKK6Ϫ/Ϫ MEFs (Fig. 4, C and D), but intact in wild-type MEFs and MKK3Ϫ/Ϫ MEFs (Fig. 4, C and D), indicating that MKK6 and MKK3 play redundant roles on the phosphorylation of p38 in response to chemical stress. Altogether, these findings established that MKK3 and MKK6 play important roles on the activation of the p38 MAP kinase pathway by IFN␣.
Previous studies using pharmacological inhibitors of p38 or p38␣ knock-out cells have established that the function of the p38 MAP kinase pathway is essential for IFN-dependent gene transcription via ISRE or GAS elements (22,27). Our findings using MKK3Ϫ/Ϫ MKK6Ϫ/Ϫ MEFs established that these kinases are required for IFN-inducible p38 activation, raising the possibility that the function of MKK3/6 may be required for IFN-dependent transcriptional activation. To address this issue, wild-type MEFs, or MKK3/6 double knock-out MEFs were FIG. 8. The function of MKK3 and MKK6 is not required for IFN␣-dependent serine or tyrosine phosphorylation of STAT1 or STAT3. A, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out mouse embryonic fibroblasts were incubated in the absence or presence of mouse IFN␣ for the indicated times. The cells were lysed and equal amounts of total lysates were analyzed by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of STAT1 on tyrosine 701. B, the blot shown in A was stripped and re-probed with an anti-STAT1 antibody to control for loading. C, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out mouse embryonic fibroblasts were incubated in the absence or presence of mouse IFN␣, at the indicated times. The cells were lysed, and equal amounts of total cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of STAT1 on serine 727. D, the blot shown in C was stripped and re-probed with an anti-STAT1 antibody to control for loading. E, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out mouse embryonic fibroblasts were incubated in the absence or presence of mouse IFN␣, at the indicated times. The cells were lysed and equal amounts of total cell lysates were analyzed by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of STAT3 on tyrosine 705. F, the blot shown in E was stripped and re-probed with an anti-STAT3 antibody to control for loading. transfected with an ISRE-luciferase construct. The cells were subsequently treated with mouse IFN␣ or mouse IFN␤, and luciferase activity was measured. IFN␣-and IFN␤-dependent gene transcription via ISRE elements was defective in the absence of MKK3/6 (Fig. 5, A and B). Similarly, in reporter assays using an 8ϫ GAS luciferase construct, we found that IFN-inducible gene transcription via GAS elements was defective in the absence of MKK3 and MKK6 (Fig. 6, A and B). These studies strongly suggested that the function of these MAP kinase kinases might be essential for the transcription of genes that have ISRE and/or GAS elements in their promoters. To directly address this issue, we studied the requirement of MKK3 and MKK6 on the expression of mRNA for Isg15 and Irf-9, two IFN-sensitive genes that play important roles in the generation of IFN responses (64 -66). As expected, treatment of wild-type MEFs with mouse IFN␣ resulted in strong induction of Isg15 and Irf-9 gene expression (Fig. 7, A and B). Such induction was partially defective in single MKK3Ϫ/Ϫ or MKK6Ϫ/Ϫ MEFs (Fig. 7, A and B) and was completely abrogated in MKK3/6 double knock-out MEFs (Fig. 7, A and B). The defective expression of these genes was similar to that seen in STAT1 knock-out cells as compared with parental MEFs (Fig.  7, C and D). Taken altogether, these studies establish that in addition to the function of the STAT pathway, activation of MKK3/6 by the Type I IFN receptor is required for transcriptional activation of ISGs.
In subsequent studies, we examined whether targeted deletion of the Mkk3 and Mkk6 genes has any effects on the phosphorylation and activation of IFN-regulated STAT proteins, and if so, whether such effects account for the defective IFNdependent gene transcription in MKK3/6Ϫ/Ϫ cells. Phosphorylation of STAT1 on tyrosine 701 is required for the interaction of STAT1 with STAT2 and the subsequent formation of the mature ISGF3 complex (STAT1-STAT2-IRF9) that translocates to the nucleus and binds to ISRE elements to initiate transcription (2)(3)(4)(5). Similarly, phosphorylation on tyrosine 701 in STAT1 or on tyrosine 705 in STAT3 is required for the formation of STAT1:1 homodimers or STAT1:3 heterodimers that bind to GAS elements in the promoters of certain interferon-regulated genes (2)(3)(4)(5)7). On the other hand, phosphorylation on serine 727 in the C terminus of STAT1 is essential for full transcriptional activation of the protein, but does not modify its binding to ISRE or GAS elements (35,36). As shown in Fig. 8, A and B, IFN␣ treatment induced tyrosine phosphorylation of STAT1 in all different MEFs, including MKK3/6ϩ/ϩ, MKK3Ϫ/Ϫ MKK6ϩ/ϩ, and MKK3/6Ϫ/Ϫ MEFs (Fig. 8B). Phosphorylation of STAT1 on serine 727 was also intact in MKK3Ϫ/Ϫ and MKK3/6Ϫ/Ϫ MEFs (Fig. 8, C and D), indicating that these MKK kinases do not regulate downstream signals required for phosphorylation of the STAT1 protein. In a similar manner, lack of MKK3 expression or of both MKK3 and MKK6 had no effect on the IFN-dependent phosphorylation of STAT3 FIG. 9. Stress-induced phosphorylation of STAT1 does not require MKK3 or MKK6. A, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out MEFs were incubated for 20 min in the absence or presence of 0.5 M NaCl, as indicated. The cells were lysed and equal amounts of total cell lysates were resolved by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of STAT1 on tyrosine 701. B, the blot shown in A was stripped and re-probed with an anti-STAT1 antibody, to control for loading. C, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out MEFs were incubated for 20 min in the absence or presence of 0.5 M NaCl, as indicated. The cells were lysed, and equal amounts of total cell lysates were resolved by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of STAT1 on serine 727. D, the blot shown in C was stripped and re-probed with an anti-STAT1 antibody to control for loading. E, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out MEFs were incubated for 30 min in the absence or presence of anisomycin, as indicated. The cells were lysed and equal amounts of total lysates were resolved by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of STAT1 on tyrosine 701. F, the blot shown in E was stripped and re-probed with an anti-STAT1 antibody to control for loading. G, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out MEFs were incubated for 30 min in the absence or presence of anisomycin, as indicated. The cells were lysed, and equal amounts of total cell lysates were resolved by SDS-PAGE and immunoblotted with an antibody that recognizes the phosphorylated form of STAT1 on serine 727. H, the blot shown in G was stripped and re-probed with an anti-STAT1 antibody to control for loading. on tyrosine 705 (Fig. 8, E and F), further establishing that the requirement of MKK3/6 for IFN-dependent gene transcription via GAS elements is unrelated to effects on the phosphorylation of STAT proteins.
As previous studies have implicated the p38 MAP kinase pathway in the regulation of serine phosphorylation of STAT1 in response to some (37,38), but not all (24), stress signals, we examined whether activation of MKK3 and/or MKK6 is essential for stress-induced phosphorylation of STAT1. Consistent with our previous findings demonstrating that osmotic stressinduced phosphorylation of STAT1 is p38-independent (24), we found that NaCl induced tyrosine 701 (Fig. 9, A and B) or serine 727 (Fig. 9, C and D) phosphorylation of STAT1 in the absence of MKK3/6. Similarly, anisomycin-induced phosphorylation of STAT1 was MKK3/6-independent (Fig. 9, E, F, G, and H).
Altogether, our findings established that MKK3 and/or MKK6 kinases do not mediate IFN-inducible STAT phosphorylation. However, these studies did not exclude the possibility that MKK3 and/or MKK6 may mediate activation of other, yet unknown, factors that may associate with STATS and regulate their binding to the promoters of ISGs. To address this possibility, studies were performed in which the induction of ISGF3 and SIF DNA binding complexes by mouse IFN␣ was examined by gel shift assays in cells that expressed MKK3/6 or lacked MKK3 and/or MKK6. As shown in Fig. 10, the formation of ISGF3 complexes (Fig. 10A) and SIF complexes (Fig. 10B) was clearly inducible by IFN␣ in both MKK3/6Ϫ/Ϫ and MKK3/ 6ϩ/ϩ MEFs. To define whether binding of STAT1 to ISRE elements in vivo is MKK3/6-dependent, ChIP assays were performed. STAT1 was found to be present in a complex that binds to ISRE elements in the promoter of Isg15 gene, independently of expression of MKK3 and/or MKK6 (Fig. 10C), establishing that these kinases do not affect binding of STAT1 to the promoters of ISGs in vivo. We also considered the possibility that activation of STAT proteins may indirectly exhibit regulatory effects on the activation of the p38 pathway by IFN␣. To address this, we compared the phosphorylation/activation of the p38 MAP kinase in MEFs with targeted disruption of the Stat1 gene (STAT1Ϫ/Ϫ) to parental STAT1ϩ/ϩ MEFs. As shown in Fig. 11, IFN␣-dependent phosphorylation of p38 was inducible in both STAT1ϩ/ϩ (Fig. 11, A and B) and STAT1Ϫ/Ϫ (Fig. 11, C and D) MEFs. Similarly, the activation of the downstream effector of p38, MAPKAPK-2, was clearly inducible in both STAT1ϩ/ϩ and STAT1Ϫ/Ϫ MEFs (Fig. 11, E and F).
Taken together, our studies established that MKK3 and/or MKK6 are required for p38 activation and IFN-dependent gene transcription, without modifying activation of the classic IFNregulated STAT pathway. However, the downstream effectors of the MKK3/6/p38 pathway and the precise events that mediate its regulatory effects on IFN-dependent transcriptional ac- FIG. 10. Formation of ISGF3 and SIE complexes and binding of STAT1 to ISRE elements does not require the function of MKK3 or MKK6. A, MKK3/6 WT and MKK3/6 double knockout MEFs were treated with mouse IFN␣, as indicated. Nuclear extracts were treated with 40,000 cpm of a 32 P-labeled ISRE, and complexes were resolved by native gel electrophoresis and visualized by autoradiography. Supershifts with an anti-STAT1 antibody or control rabbit IgG (RIgG) were performed as indicated. B, MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out MEFs were treated with mouse IFN␣, as indicated. Nuclear extracts were treated with 40,000 cpm of a 32 P-labeled SIE, and complexes were resolved by native gel electrophoresis and visualized by autoradiography. Supershifts with an anti-STAT1 antibody or control rabbit IgG were performed as indicated. C, binding of STAT1 to the promoter of Isg15 gene is MKK3/6-independent. MKK3/6 WT, MKK3Ϫ/Ϫ, and MKK3/6 double knock-out MEFs were incubated for 60 min in the absence or presence of mouse IFN␣, as indicated. ChIPs were performed with either an anti-STAT1 antibody or control non-immune rabbit immunoglobulin (RIgG), as indicated. The precipitated chromatin was analyzed using primers specific for an Isg15-ISRE sequence. Input chromatin for the different cell lines is shown in lanes 4, 8, and 12. tivation are not known. In experiments in which the activation of p38-regulated kinases was examined in MKK3/6 knock-out MEFs, we found that IFN-dependent activation of the kinase MAPKAPK-2 requires MKK3 and/or MKK6 activity (Fig. 12, A  and B). Similar results were obtained when activation of the related kinase, MAPKAPK-3, in response to IFN␣ was examined (Fig. 12, C and D). To determine whether MAPKAPK-2 is a downstream mediator of the regulatory effects of the MKK3/ 6/p38 cascade on IFN gene transcription, we examined the inducibility of the Isg15 and Irf-9 genes in MEFs from mice with targeted disruption of the MAPKAPK-2 gene (39). MAP-KAPK-2ϩ/ϩ and MAPKAPK-2Ϫ/Ϫ MEFs were treated with mouse IFN␣ and expression of Isg15 and Irf-9 genes was determined using quantitative real time RT-PCR. As shown in Fig. 13, the IFN-dependent gene expression was defective in the absence of MAPKAPK-2, demonstrating that this kinase is a downstream mediator of the transcriptional regulatory effects of the MKK3/6/p38 signaling pathway. DISCUSSION There is a plethora of evidence indicating that non-STAT pathways play critical roles in interferon signaling and the generation of the biological effects of interferons (4 -8). Multiple signaling cascades are activated during engagement of the Type I IFN receptor, in a manner consistent with the pleiotropic biological activities of these cytokines. Among the signaling elements regulated by Type I IFNs is p38, a kinase known to play important roles in the induction of stress responses. The function of the p38 pathway is critical for a variety of normal cellular functions, including regulation of apoptosis and cellcycle progression, gene transcription, cytokine production, and cell differentiation (reviewed in . In addition, the p38 pathway is a component of various pathophysiologic responses (43) and its abnormal activation in various disease settings contributes to the pathogenesis of certain hematologic malignancies and bone marrow failure syndromes (43,44).
There is now strong evidence that activation of the p38 MAP kinase by Type I IFNs is important for the generation of their FIG. 11. Type I IFN-dependent activation of p38 and MAP-KAPK-2 is STAT-independent. A-D, STAT1ϩ/ϩ (A and B) and STAT1Ϫ/Ϫ (C and D) MEFs were incubated with mouse IFN␣ for the indicated times. Total cell lysates were resolved by SDS-PAGE and immunoblotted with an antibody against the phosphorylated form of p38 (A and C). The same blots were stripped and re-probed with an anti-p38 antibody to control for loading. E, STAT1ϩ/ϩ and STAT1Ϫ/Ϫ MEFs were incubated for 30 min with or without mouse IFN␣, as indicated. Total cell lysates were immunoprecipitated with an anti-MAPKAPK-2 antibody or control non-immune rabbit immunoglobulin (RIgG) as indicated, and immunoprecipitated proteins were subjected to in vitro kinase assays using Hsp25 as an exogenous substrate. Proteins were resolved by SDS-PAGE and phosphorylated Hsp25 was detected by autoradiography. F, the Immobilon membrane from the experiment shown in E was probed with an anti-MAPKAPK-2 antibody, to control for protein loading. double knock-out MEFs were incubated for 60 min with or without mouse IFN␣, as indicated. Total cell lysates were immunoprecipitated with an anti-MAPKAPK-2 antibody or control non-immune rabbit immunoglobulin (RIgG) as indicated, and immunoprecipitated proteins were subjected to in vitro kinase assays using Hsp25 as an exogenous substrate. Proteins were resolved by SDS-PAGE and phosphorylated Hsp25 was detected by autoradiography. B, the Immobilon membrane from the experiment shown in A was probed with an anti-MAPKAPK-2 antibody, to control for protein loading. C, MKK3/6 WT and MKK3/6 double knock-out MEFs were treated for 60 min in the absence or presence of mouse IFN␣, as indicated. Cell lysates were immunoprecipitated with an anti-MAPKAPK-3 antibody or control non-immune rabbit immunoglobulin as indicated, and immunoprecipitated proteins were subjected to in vitro kinase assays, using Hsp25 as an exogenous substrate. Proteins were resolved by SDS-PAGE and phosphorylated Hsp25 was detected by autoradiography. D, the Immobilon membrane from the experiment shown in C was probed with an anti-MAPKAPK-3 antibody, to control for protein loading.
biological effects, including antiviral responses (23,25), antiproliferative effects on normal erythroid and myeloid hematopoietic progenitors (26), and suppression of growth of leukemic cell lines and primary leukemic CFU-GM progenitors (25). The primary mechanism by which the p38 pathway mediates the generation of such IFN-induced functional responses appears to be regulation of transcription of interferon-sensitive genes (22,24,27). Previous studies using pharmacological inhibitors of p38 or dominant-negative p38 mutants have demonstrated that IFN␣-dependent gene transcription via either ISRE or GAS elements requires the function of p38 (22)(23)(24). Moreover, recent studies using cells with targeted disruption of the p38␣ gene have established an important role for this isoform in IFN␣ signaling (27).
Despite the established role of the p38 MAP kinase in the generation of IFN-dependent responses, the precise signaling events that regulate its activation in IFN-sensitive cells have not been elucidated. It has been previously shown that the small GTPase Rac1, a known upstream effector of p38 (45)(46)(47), is activated in an IFN␣-dependent manner in human cells lines, and that overexpression of a dominant-negative Rac1 mutant blocks IFN-dependent p38 activation (25). There is also some evidence that the IFN-dependent activation of the Rac1/ p38 pathway requires upstream tyrosine kinase activity, presumably that of JAK kinases (24). The activation of p38 in a Rac1-dependent manner has suggested that an IFN-inducible kinase cascade that ultimately regulates p38 activity is engaged downstream of Rac1, but the identity of the IFN-activated MAP kinase kinases (MKK) has been unknown. In the present study, we provide the first evidence that the protein kinases MKK3 and MKK6 are activated in an IFN␣-inducible manner in sensitive human cell lines. Our data establish that these kinases are phosphorylated on serines 189/207 and that their kinase domains are induced. They also show that the IFN␣-dependent activation of the p38 (p38␣) kinase and gene transcription via either ISRE or GAS elements are disrupted in mouse embryonic fibroblasts lacking MKK3 and MKK6. In addition, using quantitative real time RT-PCR, we establish that the expression of two ISGs known to mediate IFN responses, Isg15 and Irf-9, depends on the presence of MKK3 and MKK6. Altogether, our findings reveal a critical requirement for these kinases in IFN-transcriptional regulation, via engagement of the p38 MAP kinase pathway.
Previous studies have demonstrated that MKK3 and MKK6 regulate p38 activation in response to stress (48 -51). The mechanisms by which MKK3 and MKK6 regulate p38 activation involves phosphorylation of the p38 kinase on both threonine and tyrosine residues within a Thr-Gly-Tyr motif present in the kinase subdomain VIII (40 -42, 52, 53), resulting in activation of the catalytic domain of p38. It should be pointed out that in addition to MKK3 and MKK6, another member of this family of kinases, MKK4, is known to phosphorylate and activate p38 (48,54,55). However, in contrast to MKK3/6, MKK4 is a dual specificity kinase and, in addition to the p38 MAP kinase, phosphorylates and activates JNK (40 -42, 52, 53). Interestingly, MKK4 is also activated by IFN␣ in human T-cell lines downstream of protein kinase C- (56), suggesting that this kinase may also regulate IFN-dependent activation of p38. However, our studies using MKK3/MKK6 double knockout MEFs demonstrate that MKK4 cannot compensate in the absence of MKK3/6. Thus, the primary role of MKK4 in IFN signaling is not p38 activation, but it may be instead required for IFN-dependent activation of JNK (57). This selective use of MKK3 and MKK6 for activation of p38 by Type I IFNs appears to be similar to the one seen in the case of tumor necrosis factor-␣ (30). As in the case of IFN␣, tumor necrosis factor-␣ cannot activate p38 in MKK3Ϫ/Ϫ MKK6Ϫ/Ϫ MEFs (30), although it is known that tumor necrosis factor-␣ activates MKK4 (58 -60). This is in contrast to the role of MKK4 on UV radiation-induced activation of p38, in which case MKK4 compensates for the absence of MKK3 and/or MKK6 (30). Thus, differences in the utilization and targeting of MAP kinase kinases between certain cytokines and stress signals exist. The functional purpose of such distinct patterns of MKK utilization remains unknown at this time and remains to be defined in future studies.
Interestingly, there is some evidence that MKK3 and MKK6 may exhibit relative specificity toward the p38 isoforms that they activate, with MKK6 functioning as a common activator for p38␣, p38␤, and p38␥, whereas MKK3 has been reported to activate p38␣ and p38␥, but not p38␤ (61). Type I IFNs are known to activate the ␣ isoform of p38 (22)(23)(24) and it is likely that the ␤, ␥, and ␦ isoforms are also activated, although this needs to be directly established in future studies. Independently of the regulation of activation of distinct p38 isoforms in response to IFNs, MKK3 and MKK6 are essential for IFN-dependent gene transcription, and therefore, the generation of Type I IFN-biological responses. These regulatory effects on ISG-transcriptional regulation are unrelated to any effects on the STAT pathway. Lack of MKK3 and MKK6 expression does not affect the IFNor stress-inducible phosphorylation of STAT1 on serine 727, establishing that there is no regulation of a STAT1 serine kinase downstream of MKK3/MKK6. Similarly, the binding of various IFN-activated complexes to the promoters of ISGs is intact in the absence of MKK3 and MKK6, further establishing that events distinct from STAT activation mediate the transcriptional regulatory effects of the MKK3/6/p38 pathway. One element that appears to participate in the signaling cascade that mediates p38-dependent transcriptional activation of ISGs is the kinase MAPKAPK-2, as evidenced by our studies using MAPKAPK-2 knock-out MEFs. The precise mechanisms by which MAPKAPK-2 mediates such effects remains to be resolved, whereas further work is necessary to define the functional contribution of other upstream and downstream effector kinases of the p38 pathway in the interferon system.