| Nucleic Acids Research | Pages |
©1999 Oxford University Press |
Murine Stat2 is uncharacteristically divergent
Introduction
Materials And Methods
Cell culture
DNA
RNA
Antibodies
Protein analysis
Immunofluorescence
Results
Isolation of the murine Stat2 cDNA
Characterization of the murine Stat2 cDNA
Characterization of the murine Stat2 protein
The MuStat2 C-terminus contains a TAD
The C-termini of murine and human Stat2 bind distinct proteins
Discussion
Acknowledgements
References
Murine Stat2 is uncharacteristically divergent
Received July 12, 1999; Revised and Accepted August 27, 1999
DDBJ/EMBL/GenBank accession no. AF187231
ABSTRACT Characterization of the ability of human interferons (IFNs) to rapidly induce genes led to the identification of the first two members of the STAT (signal transducers and activators of transcription) family, Stat1 and Stat2. To study the unique role of this transcription factor in IFN signaling under more physiological conditions, murine Stat2 was isolated and found to be surprisingly divergent. This divergence was most striking in the C-terminal transcriptional activation domain. Studies on murine Stat2 indicate that it functions in IFN signaling. This includes IFN-[alpha]-dependent activation, nuclear translocation, DNA binding and activation of reporter genes. However, the profound divergence at the C-terminus suggests that murine Stat2 may have evolved to mediate some unique functions as well. To explore this possibility, proteins that interact with the C-termini of murine and human Stat2 were examined. These studies indicate that the murine and human C-termini interact with an overlapping, but distinct set of proteins.
INTRODUCTION
Type I interferons (IFN) ([alpha], [beta], [omega] and [tau]) are a group of related proteins that bind to a common receptor and exert important immunomodulatory, antiproliferative and antiviral effects. These responses are in large part achieved through the induction of genes by the JAK-STAT pathway (1-3). This entails the sequential ligand-dependent activation of two receptor-associated JAK kinases (Jak1 and Tyk2) and two STAT transcription factors (Stat1 and Stat2). Stat1 and Stat2 each become activated by the phosphorylation of a single tyrosine (4,5). These phosphotyrosyl residues then interact with the SH2 domain of the other STAT, leading to the formation of a Stat1:Stat2 heterodimer (6,7). This heterodimer translocates to the nucleus where it associates with p48 to form a stable SGF-3 complex on the IFN-stimulated response element [ISRE; (8,9)]. p48 provides the DNA binding domain and Stat2 the transcriptional activation domain in ISGF-3 (10-14). IFN-[alpha] can also transduce signals through a second transcription factor, a Stat1 homodimer, which is formed through a similar SH2 phosphotyrosyl interaction (6,15-17). Likewise, Stat1 homodimers translocate to the nucleus and bind DNA. However, in contrast to ISGF-3, Stat1 homodimers bind directly to a distinct element known as the IFN-[gamma] activation site [GAS; (18)]. This Stat1 homodimer also transduces important signals for IFN-[gamma] (19-21), suggesting that Stat2 is critical for defining the unique responses to type I IFNs (e.g. IFN-[alpha]).
Stat2 shares a number of conserved domains with other members of the STAT family. This includes an N-terminal coiled-coil domain, a `putative' DNA binding domain, an SH2 domain, a tyrosine activation domain, and a C-terminal transcriptional activation domain (TAD) (22,23). Stat2, however, also exhibits some additional unique features. It has the longest C-terminal domain, at least part of which functions as a transcriptional activation domain (see below). In contrast to all other STATs, Stat2 dimers [e.g. Stat1:Stat2 heterodimers and putative Stat2 homodimers; (7)] do not bind to members of the GAS family of enhancers. This raises the vexing question as to why Stat2 posses a conserved DNA binding domain.
Deletional studies, including ones on human Stat2, have mapped the STAT TAD to the C-terminus (12,14,24-26). As is the case with many transcription factors, the sequence of the TAD is not conserved between different STATs, suggesting that it may contribute to specificity. Naturally occurring STAT isoforms, in which the TAD is absent, have been reported for several STATs, but not Stat2 (14,25,27-30). This is consistent with the observation that ISGF-3 function is dependent on expression of full-length Stat2 (12). In contrast, an alternatively spliced isoform of Stat1, missing its TAD (i.e. Stat1[beta]), is fully functional in the ISGF-3 complex (14). While most other truncated STAT isoforms are generated by an RNA processing event (31), the truncated isoform of Stat5 is generated by a unique and developmentally regulated protein processing event (28).
In an effort to develop reagents to study Stat2 in the murine system, a murine cDNA was isolated and a specific antibody generated. Although the biological response to type I IFNs is conserved between man and mouse, murine Stat2 (MuStat2) is surprisingly divergent from its human counterpart. This contrasts with the significantly higher degree of identity found between human and porcine Stat2 (i.e. 84%; GenBank accession no. 2189980). Although a number of studies indicate that the function of MuStat2 is conserved overall, it appears to bind distinct nuclear proteins. These observations may provide some insight into the remarkable divergence of murine Stat2.
MATERIALS AND METHODS
Cell culture
293T, HeLa S3 and NIH 3T3 cells were acquired from ATCC. The 2fTGH and U6A cells were gifts from G. Stark (32). Primary mouse embryonic fibroblasts (PMEFs) were isolated as previously described (33). These cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (or 10% calf serum for the HeLa and NIH 3T3 cells) and penicillin/streptomycin (Life Technologies). For metabolic labeling, HeLa cells were grown in methionine-free DMEM (Life Technologies), supplemented with 5% dialyzed calf serum (Hyclone) and 200 µCi/ml [35S]methionine (1000 Ci/mmol Easy Tag; NEN) for 16 h.
DNA
A murine thymus cDNA library, a generous gift from M. Nussenzweig, was screened with a BglI-HindIII fragment (bp 1600-2039) of the human Stat2 cDNA under low stringency conditions (5× SSC at 52°C). The largest positive clone was subcloned and sequenced with multiple overlapping sense and antisense primers (each strand was sequenced at least twice). Expression constructs of murine and human Stat2 were prepared by cloning the corresponding cDNAs into the NotI and ApaI sites of RcCMV (Invitrogen). The MuStat2-Gal4 fusion proteins were prepared by cloning amino acids 725-925 of murine Stat2 (SacI-BamHI fragment) into pSG424 (34). Glutathione S-transferase (GST) fusion proteins encoding Stat2 C-termini were prepared by cloning PCR products amplified from a HuStat2 cDNA template (primers 5[prime]-CGC GGA TCC CTG AAA CAC AGG CTC ATT GGT-3[prime] and 5[prime]-CCG GAA TTC GCT AGA AGT CAG AAG GCA TCA A-3[prime], corresponding to amino acids 691-851) and a MuStat2 cDNA template (primers 5[prime]-CGC GGA TCC TTG AAA CAT AAA CTC ATT GTG-3[prime] and 5[prime]-CCG GAA TTC CTA GTC CTT AGA AGG TAT CAA-3[prime], corresponding to amino acids 690-925) into the BamHI and EcoRI sites of pGEX-2T (Pharmacia). All reading frames were confirmed by sequencing. The HuStat2-Gal4 fusion, which includes Stat2 amino acids 736-851, was a generous gift from D. Levy. The Gal4-luciferase and ISRE-luciferase reporter constructs have been described previously (34,35).
RNA
Total cellular RNA was prepared by guanidinium thiocyanate lysis (36). An aliquot of 1-3 µg of poly(A)+ RNA, prepared on oligo(dT) cellulose (Boehringer Mannheim; 37) was fractionated on 0.8% agarose-formaldehyde, transferred to nitrocellulose and hybridized with a MuStat2 cDNA 5[prime] probe (HincII-HincII, bp 786-1090) or 3[prime] probe (NheI-BamHI, bp 2524-2955). RNase protection analysis was carried out as described previously (38). Riboprobes were prepared with T7 and SP6 RNA polymerase (New England Biolabs) from linearized templates, generated by cloning the RsaI-HindIII (bp 1689-2013) or the SspI-PvuII (bp 2066-2302) fragment of MuStat2 into pBluescript KS II (Stratagene). [[alpha]-32P]UTP (5000 Ci/mmol; NEN) was incorporated to a final specific activity of 5 × 105 d.p.m./µg. These riboprobes were hybridized to total RNA for 16 h in 80% formamide at 45°C, and then digested with RNase T2 (Life Technologies) for >1 h at 37°C. The products were precipitated and resolved on a standard 6% acrylamide-8 M urea sequencing gel.
Antibodies
A polyclonal murine Stat2 specific antibody was generated by immunizing rabbits (BabCo) with a GST fusion protein with a HindIII-BamHI fragment (amino acids 672-925) of MuStat2, as described previously (31). The polyclonal human Stat2 antibody is directed against amino acids 713-785 (7) and the Stat1 antibody, which recognizes murine and human species, is directed against amino acids 715-750 (31). An anti-phosphotyrosine specific monoclonal antibody was purchased from Upstate Biotechnology Inc. (4G10).
Protein analysis
Whole cell extracts were prepared from IFN-[alpha] A/D (1000 U/ml, 15 min; Pestka Research Labs) or IFN-[alpha]2 (1000 U/ml, 15 min; Hoffman-LaRoche) stimulated cells by lysing them in NP-40 buffer (0.5% NP-40, 0.4 mM PMSF, 1 mM DTT and 1 mM sodium orthovanadate) as described previously (15,27). Extracts were fractionated by 7% SDS-PAGE, either before or after immunoprecipitation (27), transferred to nitrocellulose, and immunoblotted with a 1:1000-fold dilution of primary antibody and then detected by ECL (Amersham; 27). For electrophoretic mobility shift assays (EMSAs), extracts were incubated with a radiolabeled ISG15 ISRE probe (5[prime]-gatcCTCGGGAAAGGGAAACCGAAACTGAAGCC-3[prime]) as reported previously (16,39). Supershift assays were performed by pre-incubating 3-4 µl of whole cell extracts with 0.6 µl of antibody at 4°C for 30 min prior to adding shift probe, in a standard shift reaction (16,39). For transient transfections, a total of 5 µg of DNA (per 10 cm dish) was introduced into 293T cells by the calcium phosphate/chloroquine method (40). 2fTGH, U6A and NIH 3T3 cells were transfected with a total of 20 µg of DNA (per 10 cm dish) for protein studies or 10 µg of DNA for reporter assays, by a standard calcium phosphate method (37). Reporter assays included 1 µg of TK-driven renilla luciferase reporter (Promega) for normalization in a dual luciferase assay (Turner luminometer). Transiently transfected cells were stimulated with IFN-[alpha] for 12 h, 24 h post-transfection. GST pull-down studies entailed incubating 10 µg (as determined by Coomassie blue staining) of agarose-bound (i.e. GSH-agarose; Sigma) GST alone (pGEX-2T; Pharmacia), GST-MuStat2 (amino acids 690-925) or GST-HuStat2 (amino acids 691-851) with ~600 µg of extracts prepared from metabolically labeled HeLa cells. Briefly, 5-10 × 107 labeled cells were washed in PBS, scraped into 2 ml of RSB and homogenized by 40 strokes in a dounce homogenizer (type `b' pestle) as previously described (31,41). Nuclear pellets were resuspended in 1 vol of buffer C (20 mM HEPES, pH 7.9, 420 mM NaCl, 20% glycerol, 1.5 mM MgCl2, 0.2 mM EDTA, 0.1 mM NaVO4, 0.2 mM PMSF and 1 mM DTT) diluted to 280 mM NaCl and incubated with the GST fusion proteins at 4°C for 16 h. Beads were washed in binding buffer (five times) and eluted by boiling in SDS-PAGE loading buffer. Eluted proteins were fractionated by 10% SDS-PAGE and identified by autoradiography after the gel was fluorography enhanced (EN3HANCE; Dupont).
Immunofluorescence
NIH 3T3 cells were seeded into 6-well plates containing coverslips at a density of 1 × 105/well and stimulated 24 h later with IFN-[alpha] A/D (1000 U/ml for 15 min). Cells were washed in 1× BWD (125 mM NaCl, 5 mM KCl, 5 mM glucose, 10 mM NaHCO3, 20 mM HEPES, 1 mM KH2PO4, pH 7.4), fixed in 3.7% formaldehyde for 20 min at 22°C, permeablized in 0.2% Triton (4 min) and then washed in 1× TBS (137 mM NaCl, 25 mM Tris, pH 7.4). Fixed cells were blocked in 10% milk in TBS (30 min) and stained with antibody diluted to a concentration of 1:100. Stained cells were visualized with a Cy-3-conjugated donkey anti-rabbit antibody (Jackson Immunochemicals) after excitation at 568 nm, at 400-fold magnification on a Zeiss LSM410 laser scanning confocal microscope.
RESULTS
Isolation of the murine Stat2 cDNA
The seven known human STATs consist of several conserved domains and a divergent C-terminal TAD. The conserved domains, which exhibit a high degree of identity between mouse and man (85-90%), include a multimerization (i.e. N-terminal) domain, coiled-coil domain, DNA binding domain and SH2 domain. To identify a murine Stat2, a cDNA fragment spanning the conserved SH2 domain of human Stat2 (HuStat2) was employed to screen a murine thymic cDNA library. The positive clone with the largest insert was characterized and sequenced. It contained an open reading frame of 2775 bp. Comparison of the conceptually translated 925 amino acid MuStat2 to HuStat2 (851 amino acids) revealed 76% identity over the first 712 amino acids, but complete divergence after this point (Fig. 1). In addition to being completely divergent in its C-terminus, MuStat2 is 75 amino acids longer. MuStat2 does, however, contain each of the characteristic STAT domains, including a putative activation tyrosine at amino acid 689 (22,23).
Figure 1. Amino acid sequence comparison of murine and human Stat2. Conceptual translation of the murine Stat2 cDNA sequence and the previously reported human Stat2 cDNA sequences are aligned to optimize sequence identity (Geneworks). Seventy-six percent of the amino acids between 1 and 712/713 are identical (enclosed in gray boxes). The putative boundaries of known structural domains, including the coiled-coil domain, DNA binding domain, linker domain, SH2 domain, tyrosine activation domain (*P) and the TAD are indicated (12,22,23). This sequence has been deposited in GenBank (accession no. AF187231). The first coding nucleotide represents base pair 1 in the nucleotide numbering scheme. Primers 1-5, employed for the analysis of the Stat2 gene, are indicated below the sequence with numbered arrows (1, 2020-GGG TGT TAC TAC CAG GAA-2038; 2, 2050-GAA GAA CAG AGG AAA TAT TTG AAA CA-2075; 3, 2158-GTG AAT GCA GAG CTC TTG-2175; 4, 2260-TAC TCA GTA TGG GGT CCA-2243; 5, 2269-CAA GTC CTG CTG GAG CCA-2286). Tandem repeat sequence motifs identified in murine Stat2 are denoted Mu-1-Mu-6, and those identified in human Stat2 are denoted Hu-1 and Hu-2 (see Discussion).
Characterization of the murine Stat2 cDNA
The unexpected divergence of our MuStat2 cDNA prompted several additional studies to confirm its identity. First, northern blotting studies were carried out to determine whether the conserved N-terminal and divergent C-terminal domains hybridize to the same size transcript. Unexpectedly, neither probe detected a transcript in preparations of total RNA. This contrasts with the more readily detectable levels of human Stat2 transcripts (C.Park and C.Schindler, unpublished observation). However, when poly(A)+ RNA was prepared from IFN-[alpha]-stimulated NIH 3T3 cells the N- and C-terminal probes both recognized the same ~4.5 kb message (see Fig. 2A). The transcript was not detected in unstimulated cells (data not shown), even though MuStat2 protein is readily detected in those cells (see below). These results suggest that there is a single major, but relatively rare, MuStat2 transcript.
Figure 2. Evaluation of the murine Stat2 transcript. (A) Northern blots with 5[prime] and 3[prime] probes. Aliquots of 1 (lanes 1 and 2) or 3 µg (lanes 3 and 4) of poly(A)+ selected RNA, prepared from NIH 3T3 and HeLa cells after stimulation with IFN-[alpha], was fractionated and sequentially hybridized with MuStat2 N-terminal (5[prime] probe, left panel) and C-terminal (3[prime] probe, right panel) probes. The relative mobility of the 18S and 28S rRNA bands are indicated. (B) RNase protection assay. Riboprobes A and B were hybridized with 20 µg of total NIH 3T3 RNA or tRNA, digested and gel fractionated. The sizes of fully digested probe A (324 bp) and probe B (236 bp) are indicated in the left margin. The position of undigested probe B and molecular weight markers are indicated in the right margin.
To exclude the possibility that there may be multiple similarly sized transcripts, a more sensitive RNase protection assay was carried out. Riboprobes were generated to portions of the murine cDNA representing either a conserved sequence (probe A, amino acids 564-671) or spanning the region of divergence (probe B, amino acids 689-767). Both riboprobes gave rise to the predicted `full-length' protected products when hybridized to RNA prepared from murine NIH 3T3 cells (Fig. 2B). It should be noted that these cells represent a different species than that from which the riboprobe was prepared. Probe B proved more difficult to digest, leading to a modest level of undigested product in both the control (i.e. tRNA) and experimental samples. Additional studies with riboprobes generated to a more C-terminal region also gave rise to fully protected products (data not shown). These studies indicate that the sequences between amino acids 564 and 767 are found on a single major transcript, and strongly argue against the possibility that the Stat2 gene gives rise to another significant murine Stat2 transcript (e.g. one with a C-terminus more homologous to HuStat2). Confirming this observation, both conserved and divergent sequences of MuStat2 map to the same genetic locus in a different strain of mouse (i.e. 129SVJ mice; data not shown).
Characterization of the murine Stat2 protein
Having provided strong evidence that the murine Stat2 gene gives rise to a single surprisingly divergent transcript, we proceeded to characterize the protein it encodes. An antibody, directed against both conserved and divergent amino acids (i.e. 672-925), was generated and evaluated on extracts prepared either from human or murine cell lines (see Fig. 3). A human cell line (i.e. 2fTGH) transfected with the murine Stat2 cDNA was included in the analysis (lane 9). The murine-specific antibody detected a single ~130 kDa protein in each of the wild-type strains represented by the murine cell lines examined. This larger molecular weight contrasts with the 113 kDa size of the human Stat2 protein, but is consistent with the larger open reading frame. Notably, this protein was absent in ES cells in which both alleles of the Stat2 gene had been targeted for deletion by homologous recombination (lane 4; C.Park and C.Schindler, manuscript in preparation). Moreover, the ~130 kDa endogenous murine Stat2 band is identical in size to the one detected in human cells transfected with the murine Stat2 cDNA (lane 9). Some additional minor bands, evident in PMEFs, are likely to represent proteolytic breakdown products, as their presence varies considerably between preparations. When this filter was reprobed with the HuStat2 antibody (Fig. 3A, lower panel), a distinct 113 kDa band was detected in each of the human cell lines. A minor band of ~100 kDa was detected in each of the murine extracts. However, this band was also found in Stat2 null cells and is therefore likely to represent a non-specific band.
Figure 3. Characterization of murine Stat2 protein. (A) MuStat2 and HuStat2 have different molecular weights. Whole cell extracts, prepared from several cell lines, were fractionated by SDS-PAGE and sequentially immunoblotted (Blot) with human Stat2 (HuSt2)- or murine Stat2 (MuSt2)-specific antibodies as indicated. Baculovirus recombinant HuStat2 (rHuSt2; 7) and 2fTGH cells transfected with MuStat2 (2fTGH/MuSt2) are in lanes 1 and 9, respectively. Human (Hu) and murine (Mu) cell lines are indicated. ES (embryonic stem) cells are from the 129SVJ strain, PMEFs are from a cross between the 129SVJ and C57Bl/6J strains, NIH 3T3 cells are fibroblasts from the NIH/Swiss strain, M12 cells are B cells (a spontaneous lymphoma) from the BALB/c strain. Molecular weight standards (in kDa) are indicated in the right margin and the identity of Stat2 in the left margin. (B) Murine Stat2 is tyrosine phosphorylated in response to IFN-[alpha]. Extracts were prepared from HeLa (human, Hu) or NIH 3T3 (murine, Mu) cells before and after stimulation with IFN-[alpha] (t = 15 min). Stat2 immunoprecipitated (IP) with murine- or human-specific antisera and sequentially immunoblotted (Blot) with antiphosphotyrosine (pTyr; Upstate Biotechnology Inc.) or human Stat2 (HuSt2)- or murine Stat2 (MuSt2)-specific antibodies as indicated. Recombinant HuStat2 (rHuSt2) is in lane 1. Molecular weight standards (in kDa) are indicated in the right margin.
To determine whether the endogenous MuStat2 protein can become activated (i.e. tyrosine phosphorylated), extracts were prepared from NIH 3T3 cells before and after stimulation with IFN-[alpha] (Fig. 3B). When these extracts were immunoprecipitated with the murine-specific antibody, MuStat2 was found to be tyrosine phosphorylated in stimulated human cells (lanes 6 and 7). Likewise, control preparations of HuStat2 were tyrosine phosphorylated in IFN-[alpha] stimulated cells (Fig. 3B, lanes 4 and 5). Again, both antibodies were species specific. These studies demonstrate that endogenous MuStat2 is specifically tyrosine phosphorylated in response to stimulation with IFN-[alpha], and provide additional evidence that it is bona fide murine Stat2.
Next, the ability of these proteins to participate in the formation of ISGF-3 was examined. Extracts were prepared from 2fTGH (human) cells expressing MuStat2 before and after stimulation with IFN-[alpha]. Control extracts were prepared from IFN-[alpha]-stimulated 2fTGH or NIH 3T3 (murine) cells. Consistent with the larger size of murine Stat2, murine ISGF-3 (MuISGF-3) exhibits a notably slower mobility than human ISGF-3 (HuISFG-3; Fig. 4A and B). Moreover, expression of MuStat2 in 2fTGH cells (Fig. 4A) leads to the induction of a new `larger' DNA binding complex, which co-migrates with MuISGF-3. A faster migrating HuISGF-3 complex is also evident in these extracts. Antibody addition (i.e. `supershift') studies confirm that the faster migrating complex is composed of human Stat1 and human Stat2, whereas the slower migrating complex is composed of human Stat1 and murine Stat2. Under these conditions, the HuStat2 antibody does exhibit some modest reactivity to MuStat2. Similar results were obtained when native MuISGF-3 and HuISGF-3 were evaluated by antibody addition (see Fig. 4B).
Figure 4. Murine Stat2 forms a distinct DNA binding complex. (A) Recombinant murine Stat2 is a component of ISGF-3. Whole cell extracts were prepared from 2fTGH cells transfected with a murine Stat2 cDNA (2fTGH/MuSt2 cDNA) before (-) and after (+) stimulation with HuIFN-[alpha] (t = 15 min) as indicated. DNA binding activity was evaluated with an ISG15 ISRE probe either before (PreI.) or after the addition of antibodies specific for murine Stat2 ([alpha]-MuSt2), Stat1 ([alpha]-St1), or human ([alpha]-HuSt2) as indicated. The mobility of murine (Mu) and human (Hu) ISGF-3 is indicated in the left margin. (B) Endogenous murine Stat2 is a component of ISGF-3. Whole cell extracts were prepared from HeLa (Hu) cells or murine splenocytes (Mu) before (Unstim.) and after stimulation with IFN-[alpha] (t = 15 min) as indicated. DNA binding activity was evaluated as detailed in (A). The mobility of murine (Mu) and human (Hu) ISGF-3 is indicated in the left margin.
Another property of STATs is their ligand-dependent nuclear translocation. To determine whether endogenous MuStat2 translocates to the nucleus in response to stimulation with IFN-[alpha], immunofluorescence studies were carried out. NIH 3T3 cells were stained with the murine Stat2 and control Stat1 antibodies before and after stimulation with IFN-[alpha] (Fig. 5). As previously reported for HuStat2 (15), MuStat2 is predominantly cytoplasmic in unstimulated cells (Fig. 5C). Following IFN-[alpha] treatment, there is a dramatic redistribution of MuStat2 to the nucleus (Fig. 5D). Control experiments with an antibody that recognizes murine Stat1 provided similar results (Fig. 5A and B). As previously reported, some Stat1 was found in the nucleus of unstimulated cells (Fig. 5A; 15,19). This may reflect the role of Stat1 in transducing other signals (42). The studies in this section demonstrate that the protein encoded by the MuStat2 cDNA is activated by IFN-[alpha].
Figure 5. Murine Stat2 translocates into the nucleus in response to IFN-[alpha]. NIH 3T3 cells, before (A and C) or after (B and D) stimulation with IFN-[alpha] A/D (t = 15 min), were stained with either a Stat1 ([alpha]-Stat1, A and B)- or murine Stat2 ([alpha]-Stat2, C and D)-specific antibodies. Cells were visualized using a Cy-3-conjugated secondary donkey anti-rabbit antibody and visualized after excitation at 568 nm by a Zeiss LSM410 laser scanning confocal microscope.
The MuStat2 C-terminus contains a TAD
The C-terminus of HuStat2 has been shown to encode a TAD that is required for ISGF-3 function (12). Since ISGF-3 appears to be critical in mediating the biological response to type I IFNs in both the murine and human systems, the divergent C-terminus of MuStat2 should be expected to encode a TAD as well. To test this, the C-terminal 201 amino acids of MuStat2 were fused to the DNA binding domain of transcriptionally inactive GAL4 (34). The ability of this chimeric transcription factor to drive expression of a GAL4-driven reporter was then assayed in NIH 3T3 cells (24,25). As shown in Figure 6A, the MuStat2-GAL4 fusion promoted a robust ~50-fold induction of the luciferase reporter. This compared favorably with the transcriptional activation potential of the control VP16-GAL4 fusion (24). Likewise, a HuStat2-GAL4 fusion construct that included the C-terminal 116 amino acids, known to encode a TAD [D.Levy, personal communication; (12)], also potently induced the luciferase reporter (~30-fold).
Figure 6. Murine Stat2 is a transcriptional activator. (A) Stat2-GAL4 fusions promote induction of a luciferase reporter gene. NIH 3T3 cells were co-transfected with a Gal4-luciferase reporter and the indicated Gal4 fusion constructs (Gal4, empty vector; Gal4-HuStat2, Gal4 fused to the C-terminus of human Stat2; Gal4-MuStat2, Gal4 fused to the C-terminus of murine Stat2; Gal4-VP16, Gal4 fused to the VP16 TAD). After 24 h, extracts were prepared and luciferase activity determined. Data represent the averages of four experiments. (B) Murine and human Stat2 rescue the ability of U6A cells to respond to IFN-[alpha]. U6A cells were co-transfected with an ISRE-driven luciferase reporter gene and either empty vector (RcCMV), a human Stat2 expression construct (HuStat2), or a murine Stat2 expression construct (MuStat2). Twenty-four hours after transfection, the cells were stimulated with IFN-[alpha] (t = 12 h) and then evaluated for luciferase activity. Data represent the averages of four experiments.
To evaluate the transcriptional activation potential of MuStat2 under more physiological conditions, Stat2-defective human U6A cells (32) were reconstituted with either MuStat2 or HuStat2 cDNAs. The ability of these cells to respond to IFN-[alpha] was then evaluated by induction of a well-characterized ISRE-driven luciferase reporter (35). As has been previously demonstrated (32), HuStat2 fully restored the ability of U6A cells to promote the IFN-[alpha]-dependent induction of the ISRE-driven reporter gene (~7-fold above unstimulated cells). Consistent with the preceding GAL4 fusion assay (Fig. 6A) and EMSA (Fig. 4), MuStat2 was also able to fully reconstitute the ability of U6A cells to respond to IFN-[alpha] (also ~7-fold above unstimulated cells; Fig. 6B). In contrast, U6A cells transfected with a MuStat2 construct missing the most C-terminal 44 amino acids were unable to respond to IFN-[alpha] (data not shown). These studies indicate that despite its sequence divergence, MuStat2 encodes a potent TAD that is functionally analogous to the one encoded by its human counterpart.
The C-termini of murine and human Stat2 bind distinct proteins
TADs are believed to promote transcription by interacting with co-activators and the basal transcriptional machinery. To determine whether the divergent C-termini of murine and human Stat2 bind to a similar or differing set of nuclear proteins, GST pull-down studies were carried out. GST fusions with the C-termini of HuStat2 (i.e. amino acids 691-851) and MuStat2 (i.e. amino acids 690-925) were incubated with nuclear extracts prepared from metabolically labeled (i.e. [35S]methionine) NIH 3T3 and HeLa cells. For both cell types, several categories of interacting proteins were identified: those that bound both the C-termini equivalently; those that bound specifically with the murine target; and those that bound specifically with the human target. Some differences in the proteins that bound each of the GST fusion proteins were observed when HeLa and NIH 3T3 cell extracts were compared. Moreover, IFN-[alpha] stimulation did not affect the pattern of binding (data not shown). A representative experiment with extracts prepared from unstimulated HeLa cells, which always provided a stronger signal (i.e. because they grow at a significantly higher density), is presented in Figure 7. Consistent with the functional equivalence between the C-termini of murine and human Stat2 (see above), the majority of proteins that interact with these C-termini are the same. However, proteins with molecular weights of ~250 and ~52 kDa preferentially bound the human GST fusion protein, and proteins of ~60, ~30 and ~22 kDa preferentially bound the murine fusion protein. These observations suggest that although these divergent Stat2 C-termini bind to many similar proteins, they also exhibit distinct binding preferences. It is intriguing to speculate that a need for Stat2 to bind other species-specific proteins may have provided the evolutionary pressure for these sequences to diverge.
Figure 7. The divergent C-termini of murine and human Stat2 bind similar proteins. Metabolically labeled ([35S]methionine) nuclear extracts from 5-10 × 107 HeLa cells were incubated with either immobilized human GST-Stat2 (GST-HuSt2) or murine GST-Stat2 (GST-MuSt2) fusion proteins. The bound proteins were recovered, fractionated by SDS-PAGE and visualized by autoradiography. Molecular weight standards (in kDa) are indicated in the left margin and species specifically bound proteins (Mu, murine; Hu, human) in the right margin.
DISCUSSION
Studies over the past 30 years have highlighted the important role IFNs play in providing protection against viral infections in both the murine and human system (1,43). This response is dependent on the induction of genes. Characterization of the ability of human type I IFNs to induce target genes led to the identification of the first two members of the STAT family of transcription factors (31,44,45). Subsequent studies have identified seven mammalian members of the STAT family, which transduce vital signals for all cytokines (2,3,46). Although STAT family members vary in sequence, murine and human homologs exhibit a high degree of sequence homology. In fact, the similarity between the sequences and chromosomal locations of STATs in these two species has led to the hypothesis that multiple duplications of a single primordial STAT gene gave rise to the entire family of seven genes (47).
Until recently, IFN-induced STAT signals have been characterized predominantly in the human system. However, in mice, which offer a number of advantages in the study of signaling cascades, IFNs stimulate an analogous biological response. Recently for example, deletion of the IFN-[alpha] receptor and Stat1 gene in mice have provided important insights into IFN biology (20,21,48). In an effort to develop reagents that would permit the analogous biochemical and genetic characterization of Stat2, murine Stat2 was identified. Unexpectedly, MuStat2 is significantly divergent. Several experimental approaches evaluating nucleic acids prepared from several distinct murine strains (e.g. northern blotting, RNase protection and gene mapping) demonstrate that this cDNA encodes the predominant, and most likely only, murine Stat2 species. An alternative Stat2 splice isoform that is missing part of the SH2 domain was not detected in our studies (49). Studies on the murine Stat2 protein, outlined below, provide further support for this conclusion.
Consistent with its larger open reading frame, the murine Stat2 gene encodes a protein that is notably larger than its human counterpart (i.e. 130 versus 113 kDa). In response to type I IFNs, murine Stat2 becomes tyrosine phosphorylated, associates with Stat1, translocates to the nucleus, and participates in the formation of ISGF-3. As anticipated, when Tyr689, which corresponds to the tyrosine known to mediate the activation of HuStat2 [i.e. Y690; (5)] is mutated to Phe, MuStat2 fails to become activated (J.Valens and C.Schindler, unpublished observation). Additionally, studies in 2fTGH cells confirm that murine Stat2 is responsible for the significantly slower mobility of murine ISGF-3 in DNA binding studies.
Biological responses to type I IFNs, including the activation of target genes, are well conserved between the murine and human species. Studies in human cells have demonstrated that the C-terminus of Stat2 provides the TAD for ISGF-3-regulated genes. Likewise, GAL4 fusion experiments and studies in U6A cells demonstrate that the divergent murine Stat2 C-terminus encodes a potent TAD (Fig. 6). Consistent with studies on human Stat2 (12), experiments with C-terminally truncated MuStat2 (i.e. deletion of the terminal 44 amino acids), or with a chimeric Stat2 where the murine and human C-termini have been exchanged at amino acid 672, indicate that the murine C-terminal residues are critical for transcriptional activation (C.Park and C.Schindler, unpublished observation). Closer inspection of the murine Stat2 sequence reveals that much of the divergence can be attributed to six tandem copies of a novel 16 amino acid Leu/Pro-rich repeat (APQVLLEP·APQVQLEP) between amino acids 756 and 851 (see Fig. 1, repeats Mu-1-Mu-6). This repeat appears to represent a less well conserved eight amino acid tandem duplication. This is particularly evident in the most divergent repeat 5 (denoted Mu-5a and Mu-5b). HuStat2 does not encode such a repeat motif, but there are two shorter (i.e. 10-11 amino acid) and less well conserved, Leu/Pro-rich motifs between amino acids 714 and 737 (see Fig. 1, repeats Hu-1 and Hu-2). The role of these repeat motifs in transcriptional activation is not clear, since deletional studies have mapped the TAD distal to these repeats. Perhaps these repeats mediate the interaction with species-specific proteins important for some aspects of Stat2 biology (see below).
STAT sequence comparisons have determined that Stat2 contains the largest C-terminus, far more than appears to be required for an active TAD. In fact, only the most C-terminal residues are likely to represent the TAD that associates with the basal transcriptional apparatus as well as an association with transcriptional activators [e.g. YY-1, p300/CBP and MCM5; (50-53)]. The remaining residues (i.e. those more amino proximal) may mediate other, potentially more species-specific protein-protein interactions. Consistent with the possibility that murine and human Stat2 evolved to mediate interaction(s) with additional proteins that are unique to each species, murine and human Stat2 C-terminal GST fusion proteins bind to overlapping, but distinct sets of nuclear proteins. These observations support a model in which the need for murine and human Stat2 to interact with species-specific proteins provided the evolutionary pressure for these sequences to rapidly diverge. However, despite these differences, murine and human Stat2 mediate a conserved biological response to type I IFNs.
ACKNOWLEDGEMENTS
We would like to thank S. Li and M. Azam for their assistance in the isolation of Stat2 genomic clones and the generation of the GST fusion protein employed to generate the MuStat2 antibody. We would also like to acknowledge Anne-Marie Pablo for excellent technical assistance and George Stark for generously providing U6A cells. These studies were supported by the National Institute of Health and the Medical Scientist Training Program (C.Y.P.).
REFERENCES
*To whom correspondence should be addressed. Tel: +1 212 305 5380; Fax: +1 212 543 0063; Email: cws4{at}columbia.edu
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Vlasakova, Z. Novakova, L. Rossmeislova, M. Kahle, P. Hozak, and Z. Hodny Histone deacetylase inhibitors suppress IFN{alpha}-induced up-regulation of promyelocytic leukemia protein Blood, February 15, 2007; 109(4): 1373 - 1380. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bhattacharjee, B. Xu, D. A. Frank, G. M. Feldman, and M. K. Cathcart Monocyte 15-Lipoxygenase Expression Is Regulated by a Novel Cytosolic Signaling Complex with Protein Kinase C {delta} and Tyrosine-Phosphorylated Stat3 J. Immunol., September 15, 2006; 177(6): 3771 - 3781. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Lo, R. M. Brazas, and M. J. Holtzman Respiratory Syncytial Virus Nonstructural Proteins NS1 and NS2 Mediate Inhibition of Stat2 Expression and Alpha/Beta Interferon Responsiveness J. Virol., July 15, 2005; 79(14): 9315 - 9319. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Zimmermann, M. Trilling, M. Wagner, M. Wilborn, I. Bubic, S. Jonjic, U. Koszinowski, and H. Hengel A cytomegaloviral protein reveals a dual role for STAT2 in IFN-{gamma} signaling and antiviral responses J. Exp. Med., May 16, 2005; 201(10): 1543 - 1553. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Nusinzon and C. M. Horvath Interferon-stimulated transcription and innate antiviral immunity require deacetylase activity and histone deacetylase 1 PNAS, December 9, 2003; 100(25): 14742 - 14747. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Kraus, J. F. Lau, J.-P. Parisien, and C. M. Horvath A Hybrid IRF9-STAT2 Protein Recapitulates Interferon-stimulated Gene Expression and Antiviral Response J. Biol. Chem., April 4, 2003; 278(15): 13033 - 13038. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. Lau, I. Nusinzon, D. Burakov, L. P. Freedman, and C. M. Horvath Role of Metazoan Mediator Proteins in Interferon-Responsive Transcription Mol. Cell. Biol., January 15, 2003; 23(2): 620 - 628. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-P. Parisien, J. F. Lau, and C. M. Horvath STAT2 Acts as a Host Range Determinant for Species-Specific Paramyxovirus Interferon Antagonism and Simian Virus 5 Replication J. Virol., June 5, 2002; 76(13): 6435 - 6441. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Papin, P. Duquesnoy, C. Cazeneuve, J. Pantel, M. Coppey-Moisan, C. Dargemont, and S. Amselem Alternative splicing at the MEFV locus involved in familial Mediterranean fever regulates translocation of the marenostrin/pyrin protein to the nucleus Hum. Mol. Genet., December 1, 2000; 9(20): 3001 - 3009. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||