The angiotensin II-dependent nuclear translocation of Stat1 is mediated by the Jak2 protein motif 231YRFRR.

In response to angiotensin II, Jak2 autophosphorylates and binds the angiotensin II AT(1) receptor. By studying a variety of Jak2 deletion proteins, we now show that the Jak2 protein motif (231)YRFRR is required for the co-association of this kinase with the AT(1) receptor. We also used a full-length Jak2 protein containing a (231)FAAAA amino acid substitution. Although this protein still autophosphorylated in response to angiotensin II, it did not co-associate with the AT(1) receptor. This uncoupling indicates that AT(1)/Jak2 co-association is not necessary for angiotensin II-induced Jak2 autophosphorylation and that Jak2 autophosphorylation per se is insufficient for AT(1) receptor co-association. In response to angiotensin II, the Jak2-(231)FAAAA mutant will tyrosine phosphorylate Stat1. However, in the absence of AT(1)/Jak2 co-association, Stat1 did not translocate into the cell nucleus and failed to mediate gene transcription. This notable result indicates that Stat1 tyrosine phosphorylation alone is insufficient for Stat1 nuclear translocation. In summary, we now show that, although Jak2-mediated tyrosine phosphorylation of Stat1 is independent of receptor co-association, Jak2-mediated recruitment of Stat1 to the AT(1) receptor is critical for Stat1 nuclear translocation and subsequent gene transcription.

Angiotensin II is the effector molecule of the renin-angiotensin system. It is vital for maintaining a wide variety of physiological responses, including salt and water balance, blood pressure, and vascular tone. These effects are transduced through a seven-transmembrane surface receptor called AT 1 (1). In addition to promoting the hydrolysis of heterotrimeric G proteins, activation of the AT 1 receptor by angiotensin II also results in the activation of several non-receptor tyrosine kinases, including Jak2 (2)(3)(4)(5).
Jak2 is a member of the Janus family of non-receptor tyrosine kinases that also includes Jak1, Jak3, and Tyk2. These proteins are ϳ130 kDa in mass and contain seven conserved Jak homology domains. Typically, Jak activation by a cytokine receptor leads to STAT 1 activation and thus transmission of a signal from the extracellular surface of the cell into the nucleus (6). Studies by our laboratory (2) and by Baker and co-workers (7)(8)(9) have shown that, similar to cytokines, activation of the AT 1 receptor by angiotensin II promotes 1) STAT tyrosine phosphorylation, 2) STAT nuclear translocation, 3) STAT DNA binding activity, and 4) STAT-dependent transcriptional activation.
Recently, our laboratory examined proximal signaling events that mediate Jak/STAT activation by the AT 1 receptor. Specifically, we demonstrated that the physical co-association of the AT 1 receptor with Jak2 is dependent on the AT 1 receptor motif 319 YIPP found within the carboxyl terminus of the receptor protein (10). Subsequently, we found that Jak2 must be catalytically active to associate with the AT 1 receptor, as inhibition of Jak2 kinase activity either by pharmacological means or by a dominant-negative Jak2 protein greatly reduces AT 1 /Jak2 co-association (11).
The studies in this report examined whether, in addition to a functional kinase domain, there are any other requirements for Jak2 co-association with the AT 1 receptor. Here we demonstrate that the Jak2 protein motif 231 YRFRR, located in the amino-terminal portion of the molecule, is also required for co-association with the AT 1 receptor. Conversion of the Jak2 protein motif from 231 YRFRR to 231 FAAAA functionally uncouples Jak2 autophosphorylation from AT 1 receptor binding, demonstrating that Jak2 autophosphorylation occurs independently of receptor co-association. When Jak2 fails to coassociate with the AT 1 receptor, a functional consequence is that Stat1 fails to translocate into the nucleus and mediate gene transcription. pRC-186, 5Ј-CATGATAATGCGGCCGCAATGAGAATAGCTAAGGAG-3Ј; pRC-211, 5Ј-CATGATAATGCGGCCGCAATGAAGTGCGTTCGAG-CGAAG-3Ј; pRC-221, 5Ј-CATGATAATGCGGCCGCAATGCACATTTT-AACCCGGAAG-3Ј; pRC-225, 5Ј-CATGATAATGCGGCCGCAATGCGG-AAGCGAATCAGGTACAG-3Ј; pRC-230, 5Ј-CATGATAATGCGGCCGC-AATGTACAGATTTCGCAGATTC-3Ј; and pRC-235, 5Ј-CATGATAATG-CGGCCGCAATGTTCATTCAGCAATTCAGTC-3Ј (the initiation methionine for each oligonucleotide is underlined). The bottom strand oligonucleotide for all constructs was 5Ј-GATACTGTCTGAGCGCACAG-TTTC-3Ј and utilized an internal AflII site. The pRC-240 plasmid was made by cutting pRC-ATD with AflII and closing with ligase. The pRC-FLAG-225, pRC-FLAG-235, and pRC-FLAG-240 constructs were made by placing the FLAG peptide sequence (DYKDDDDK) between the initiation methionine and the respective Jak2 sequence. Other than pRC-CTD, which lacks the Jak2 kinase domain, all deletion constructs were found to have readily detectable levels of tyrosine phosphorylation as measured by anti-phosphotyrosine Western blotting. The Jak2-Y231F and Jak2-231 FAAAA mutants were generated using the QuikChange site-directed mutagenesis system (Stratagene). All constructs were confirmed by DNA sequence analysis.
Transient Cell Transfection-AT 1 /Jak2 co-association in BSC-40 cells was done using the vaccinia virus transfection/infection protocol (17,18). Cells were seeded in 100-mm dishes and transfected at nearconfluency with 10 g of pHA-AT 1 and 10 g of Jak2 expression vector in 20 l of Lipofectin. 4 h later, vaccinia virus clone vTF7-3 was added at a multiplicity of infection of 1.0 and incubated for 1 h. The medium was aspirated, and cells were incubated overnight in serum-containing medium and lysed 18 -20 h after infection. COS-7 cells were transiently transfected exactly as described (10).
Immunoprecipitation-To prepare lysates, cells were washed with 2 volumes of ice-cold phosphate-buffered saline containing 1 mM Na 3 VO 4 and lysed in 1.0 ml of ice-cold radioimmune precipitation assay buffer containing protease inhibitors (18). The samples were sonicated and incubated on ice for 1 h. Samples were spun at 12,000 ϫ g for 5 min at 4°C, and supernatants were normalized using the D c protein assay (Bio-Rad). Normalized lysates (ϳ400 g/ml) were immunoprecipitated for 2-16 h at 4°C with 2 g of antibody and 20 l of Protein A/G Plus-agarose beads (Santa Cruz Biotechnology). After centrifugation, protein complexes were washed three to five times with radioimmune precipitation assay buffer and resuspended in sample buffer. Bound proteins were boiled, separated by SDS-PAGE, and transferred onto nitrocellulose membranes. The immunoprecipitating anti-Jak2 pAb (HR758) and anti-HA mAb (12CA5) were from Santa Cruz Biotechnology and Roche Molecular Biochemicals, respectively. The anti-Tyr(P) mAb (PY20) and anti-Stat1 mAb (S21120) were from Transduction Laboratories.
Preparation of Nuclear Extracts-After treatment with angiotensin II, nuclear extracts were prepared using the high salt method of extraction (20 -22). Nuclear proteins were then dialyzed for 2 h at 4°C against a 500ϫ volume of solution containing 20 mM HEPES (pH 7.9 at 4°C), 20% glycerol, 100 mM KCl, 0.2 mM EDTA, 0.5 mM dithiothreitol, and protease inhibitors. Nuclear proteins were quantitated, and equal amounts were separated by SDS-PAGE. The samples were Westernblotted as indicated.
Luciferase Assay-100-mm dishes of 70% confluent COS-7 cell lines were transfected with the indicated DNA in 20 l of Lipofectin for 5 h. The cells were then trypsinized and seeded into 6-well plates at 4.5 ϫ 10 5 cells/well and allowed to attach overnight in serumcontaining medium. The following afternoon, the cells were washed and placed into serum-free Dulbecco's modified Eagle's medium containing 0.5% (w/v) bovine serum albumin. The following morning, the cells were stimulated with angiotensin II, and luciferase activity was measured from detergent extracts in the presence of ATP and luciferin using the reporter lysis buffer system (Promega) and a luminometer (Turner Designs Model 20/20). Luciferase values are reported as relative light units.

RESULTS
The Jak2 Protein Motif 231 YRFRR Is Required for AT 1 Receptor Binding in Vivo-To determine regions of Jak2 that are required for co-association with the AT1 receptor, we utilized an in vivo vaccinia virus-based expression system that exploits two aspects of Jak2 protein expression. First, we previously demonstrated that overexpression of wild-type Jak2 in COS-7 cells results in AT 1 /Jak2 co-association independent of angiotensin II treatment (11). The expressed Jak2 is activated not by angiotensin II, but rather by oligomerization and autophosphorylation of the highly expressed protein. Although the COS-7 expression system worked well for wild-type Jak2, we (12,13,15,16) and others (23) observed that some Jak2 deletion proteins are poorly expressed. To overcome this, BSC-40 cells (a vaccinia virus permissive cell line) are transfected with an expression vector containing the Jak2 cDNA under the control of the T7 promoter. The cells are then infected with a vaccinia virus that produces T7 polymerase. This system greatly enhances Jak2 protein expression and has been used to examine various aspects of Jak2 signal transduction (12,13,15,16,23). This system was used to overexpress various Jak2 deletion constructs along with an HA-tagged AT 1 receptor cDNA (pHA-AT 1 ). As with the COS-7 cell system, Jak2 is activated by oligomerization/autophosphorylation of the highly expressed protein, which in turn binds the AT 1 receptor. In vivo AT 1 /Jak2 co-association is measured by immunoprecipitating with anti-HA mAb and blotting with anti-Jak2 pAb.
Prior to assessing co-immunoprecipitation of Jak2 with the AT 1 receptor, several controls were performed. First, we demonstrated that overexpression of Jak2 required the presence of both the transfected cDNA and infection of the recombinant vaccinia virus, as omission of either one resulted in the loss of overexpression (data not shown). Second, we confirmed that the vaccinia-derived Jak2 protein was post-translationally modified by tyrosine phosphorylation and was catalytically active (data not shown).
To demonstrate specificity of the co-immunoprecipitation system, BSC-40 cells were transfected with plasmids encoding wild-type Jak2 (pRC-WT) and the HA-tagged AT 1 receptor under the control of the T7 promoter. The cells were then infected with vaccinia virus clone vTF7-3, which produces T7 polymerase. Lysates were prepared 18 -20 h after infection. The lysates were then immunoprecipitated with anti-HA mAb and blotted with anti-Jak2 pAb (Fig. 1A). AT 1 /Jak2 co-association was seen only when the Jak2 plasmid was present (lane 3 versus lanes 1 and 2). Furthermore, omission of either the HA-tagged AT 1 receptor plasmid (lane 4) or the immunoprecipitating anti-HA mAb (lane 5) resulted in specific loss of Jak2 co-association, indicating that neither was Jak2 immunoprecipitated nonspecifically by anti-HA mAb nor was Jak2 binding the Protein A/G-agarose nonspecifically.
Using this system, we then coexpressed the HA-tagged AT 1 receptor with several Jak2 deletion constructs. A schematic representation of these constructs is shown as Fig. 1B. Initially, expression vectors encoding wild-type Jak2, an aminoterminal deletion (ATD), an internal deletion (PKD), or a Cterminal deletion (CTD) were utilized. As shown in Fig. 2A (upper panel), both wild-type Jak2 and the pseudo-kinase deletion (PKD; ⌬523-746) co-associated with the receptor. However, the amino-terminal deletion (ATD; ⌬1-240) and the Cterminal deletion (CTD; ⌬1000 -1129) did not. We confirmed that all Jak2 proteins were expressed by immunoprecipitating and Western blotting aliquots from these same samples with anti-Jak2 pAb ( Fig. 2A, lower panel). We believe the inability of the CTD mutant to bind the receptor is due to the loss of the kinase domain since previous work demonstrated that inhibi-tion of Jak2 kinase activity either by pharmacological means or by rendering the kinase domain inactive by a dominant-negative mutation prevents AT 1 /Jak2 co-association (11).
To further analyze the amino terminus of Jak2 as being important for binding the AT 1 receptor, we tested the hypothesis that the initial 240 amino acids of Jak2 are required for efficient co-association with the AT 1 receptor. For these experiments, we used the AflII deletion construct (AFL; ⌬251-473). As shown in Fig. 2B (upper panel), the AFL mutant was able to co-associate with the HA-tagged AT 1 receptor in vivo. In Fig.  2B (lower panel), we confirmed expression of the individually transfected Jak2 proteins. Although these data indicate that amino acids 251-473 are not required for binding the AT 1 receptor, they lead to the hypothesis that amino acids 1-240 of Jak2 are specifically required for binding the AT 1 receptor in vivo.
To test this hypothesis, progressive amino-terminal dele-tions of the AFL mutant were made at amino acids 68, 120, 168, and 240. Using the same system, BSC-40 cells were cotransfected with pHA-AT 1 and the indicated Jak2 deletion expression vectors and then infected with vaccinia virus. AT 1 /Jak2 co-association was measured by immunoprecipitating the cell lysates with anti-HA mAb and Western blotting with anti-Jak2 pAb (Fig. 2C, upper panel). We observed that the AFL, 68, 120, and 168 deletion proteins all bound the receptor. However, the 240 deletion did not. We confirmed expression of all the Jak2 deletion proteins by immunoprecipitating and blotting equal aliquots from these same samples with anti-Jak2 pAb (Fig. 2C,  lower panel). These results indicate that the AT 1 receptorbinding region is located between amino acids 168 and 240. Jak2 deletion molecules were then generated at amino acids 186, 211, 221, and 240. Another round of co-association was performed using the same vaccinia virus-based in vivo assay. We found that Jak2 deletions 168, 186, 211, and 221 all coassociated with the HA-tagged AT 1 receptor, whereas the 240 mutant did not (Fig. 2D, upper panel). As before, we confirmed expression of all Jak2 deletions by immunoprecipitating and Western blotting equal aliquots from these same samples with anti-Jak2 pAb (Fig. 2D, lower panel). These data indicate that an important AT 1 -binding region of Jak2 is located between amino acids 221 and 240. Deletion mutants were then made at amino acids 225, 230, and 235, and AT 1 co-association was assessed as described above (Fig. 2E, upper panel). This revealed that deletion proteins 221, 225, and 230 co-associated with the AT 1 receptor, whereas Jak2 deletion proteins 235 and 240 did not. Expression of all Jak2 proteins was confirmed by immunoprecipitating and Western blotting equal aliquots of these samples with anti-Jak2 pAb (Fig. 2E, lower panel).
These data suggest that one region of Jak2 that facilitates co-immunoprecipitation with the AT 1 receptor is located between amino acids 231 and 235. One issue that concerned us, however, was the appearance of a lower band of ϳ68 kDa that was observed with many of the Jak2 deletion constructs. The protein appeared to be Jak2-related, as it was immunoreactive against anti-Jak2 pAb and was never seen when cells were transfected with the empty vector control. To better understand the nature of this protein, we generated three additional Jak2 deletion constructs: pRC-FLAG-225, pRC-FLAG-235, and pRC-FLAG-240. These molecules were modified by placing the FLAG peptide sequence (DYKDDDDK) between the initiator methionine and the respective Jak2 sequence. This allowed us to Western blot these proteins with either anti-Jak2 pAb or anti-FLAG mAb. These studies demonstrated that the lower bands were a truncated form of Jak2 that lacked the aminoterminal portion of the molecule; loss of the amino-terminal FLAG peptide sequence resulted in a shorter protein that was no longer immunoreactive with the anti-FLAG antibody, but was reactive with the anti-Jak2 antibody (data not shown). Consistent with our previous results, coexpression of these three constructs with the HA-tagged AT 1 receptor demonstrated that pRC-FLAG-225 was able to coprecipitate with the HA-tagged AT 1 receptor, whereas pRC-FLAG-235 and pRC-FLAG-240 were not (data not shown).
In summary, these data confirm our previously published work (11) demonstrating that Jak2 must possess a functional kinase domain to bind the AT 1 receptor; deletion of the kinase domain abrogates AT 1 /Jak2 co-association. In addition, the data demonstrate for the first time that Jak2 must possess amino acids 231-235, encoding the 231 YRFRR protein motif, to coprecipitate with the AT 1 receptor in vivo; constructs lacking 231 YRFRR fail to bind the AT 1 receptor.
A Jak2 Molecule Containing a 231 FAAAA Mutation in Place of 231 YRFRR Fails to Bind the AT 1 Receptor, but Autophospho-  1-4) or control IgG (lane 5) and then Western-blotted with anti-Jak2 pAb to assess AT 1 /Jak2 co-association. Shown is one of three independent experiments. B, shown is a summary of Jak2 deletion analysis studies and co-association with the HA-tagged AT 1 receptor. The schematic represents full-length Jak2 and the seven respective Jak homology domains (1)(2)(3)(4)(5)(6)(7). Also shown are the deletions generated and a table summarizing whether each was able to co-associate with the HA-tagged AT 1 receptor in vivo. Co-IP, co-immunoprecipitate.
rylates in Response to Angiotensin II-The vaccinia virus deletion analysis studies presented in Figs. 1 and 2 were performed without angiotensin II treatment. To examine the role of the Jak2 231 YRFRR motif in ligand-mediated binding to the AT 1 receptor, we generated two additional Jak2 molecules. The first was a full-length Jak2 in which the 231 YRFRR motif was converted to 231 FAAAA (Jak2-231 FAAAA). The second was a fulllength Jak2 containing a single Y231F point mutation (Jak2-Y231F). To evaluate the Jak2-231 FAAAA construct, COS-7 cells were transiently cotransfected with pHA-AT 1 and this plasmid. The cells were then treated with angiotensin II, and ligand-dependent AT 1 /Jak2 co-association was measured. We found that the Jak2-231 FAAAA mutant did not co-associate with the AT 1 receptor in response to angiotensin II (Fig. 3A, upper panel). We confirmed equal expression of both wild-type and Jak2-231 FAAAA mutant proteins by immunoprecipitating and Western blotting equal aliquots from these samples with anti-Jak2 pAb (Fig. 3A, lower panel). In contrast to this result, the Y231F mutant bound the AT 1 receptor in a manner that was virtually identical to wild-type Jak2, indicating that in the context of the 231 YRFRR motif, tyrosine 231 appears to provide a structural contribution rather than a phosphotyrosine binding residue (data not shown).
To determine whether the Jak2-231 FAAAA mutant autophosphorylates in response to angiotensin II, COS-7 cells were transfected as described for Fig. 3A, but the lysates were then FIG. 2. AT 1 /Jak2 co-association requires the Jak2 protein motif 231 YRFRR. BSC-40 cells were transfected with the indicated plasmids and infected with vaccinia virus clone vTF7-3. Lysates were prepared, and half of each lysate was immunoprecipitated (IP) with anti-HA mAb and then Western-blotted (immunoblotted (IB)) with anti-Jak2 pAb to assess AT 1 /Jak2 co-association. The remaining half was immunoprecipitated and Western-blotted with anti-Jak2 pAb to assess Jak2 protein expression. Shown is one of three (A-C) or four (D and E) independent experiments for each construct. Constructs lacking either the kinase domain or 231 YRFRR failed to co-associate with the HA-tagged AT 1 receptor in vivo. WT, wild-type Jak2. immunoprecipitated with anti-Jak2 pAb. The samples were first blotted with anti-Tyr(P) mAb to measure the angiotensin II-dependent tyrosine phosphorylation of Jak2 (Fig. 3B, upper  panel). These data clearly show that both the wild-type and 231 FAAAA mutant proteins were tyrosine-phosphorylated in response to angiotensin II. They also show that, although Jak2-231 FAAAA was unable to bind the receptor in response to ligand, it was certainly capable of tyrosine autophosphorylation. We confirmed equal precipitation of Jak2 protein by blotting these same samples with anti-Jak2 pAb (Fig. 3B, lower  panel). When COS-7 cells were transfected with an empty vector control in place of Jak2-231 FAAAA, we observed virtually no tyrosine phosphorylation of Jak2 in response to angiotensin II (data not shown). This is because non-transfected COS-7 cells have very little endogenous Jak2. Thus, the response seen in cells transfected with Jak2-231 FAAAA is not due to endogenous (wild-type) Jak2, but is due to the transfected construct. The experiments in Fig. 3 indicate that, although Jak2-231 FAAAA was capable of tyrosine autophosphorylation, it was unable to bind the AT 1 receptor in response to angiotensin II.
A Jak2 Molecule Containing a 231 FAAAA Mutation in Place of 231 YRFRR Mediates the Angiotensin II-dependent Tyrosine Phosphorylation of Stat1-We then measured the angiotensin II-dependent tyrosine phosphorylation of Stat1. This protein is a substrate of Jak2 and is tyrosine-phosphorylated in response to angiotensin II (2, 6). COS-7 cells were transfected with plasmids encoding the AT 1 receptor and an empty vector control, wild-type Jak2, or Jak2-231 FAAAA. 2 days later, the cells were treated with angiotensin II, and the resulting lysates were immunoprecipitated with anti-Tyr(P) mAb and blotted with anti-Stat1 mAb (Fig. 4A). In the absence of transfected Jak2 expression vector, no Stat1 phosphorylation was observed in this system. In contrast, addition of wild-type Jak2 plasmid resulted in the angiotensin II-dependent tyrosine phosphorylation of Stat1. When the Jak2-231 FAAAA mutant protein was transfected into cells, we found that it phosphorylated Stat1 in a manner that was identical to that of wild-type Jak2.
To better establish a time course for angiotensin II-mediated Stat1 tyrosine phosphorylation and to measure Stat1 tyrosine phosphorylation by an alternate protocol, transfected COS-7 cells were treated with angiotensin II for periods of 0 -30 min. A reciprocal protocol was then performed whereby the lysates were immunoprecipitated with anti-Stat1 mAb and Westernblotted with anti-phospho-Stat1 pAb (Fig. 4B, upper panel). This blotting antibody is specific for phosphorylated Stat1 protein at tyrosine 701, a site of phosphorylation that correlates with Stat1 activation (24,25). These data showed that Jak2-231 FAAAA phosphorylated Stat1 over a time course that was nearly identical to that of wild-type Jak2, with peak Stat1 tyrosine phosphorylation occurring between 3 and 6 min after angiotensin II treatment for both groups. Furthermore, the alternate protocol produced a result that was similar to that in Fig. 4A. Collectively, these results indicate that Jak2-231 FAAAA is not only capable of autophosphorylation, but is also able to phosphorylate a substrate in response to angiotensin II.
A Jak2 Molecule Containing a 231 FAAAA Mutation in Place of 231 YRFRR Fails to Mediate the Angiotensin II-dependent Nuclear Translocation of Stat1-Given that Jak2-231 FAAAA is able to mediate Stat1 tyrosine phosphorylation, we next investigated whether it could mediate Stat1 nuclear translocation. COS-7 cells were transfected with plasmids encoding the AT 1 receptor and either wild-type Jak2 or Jak2-231 FAAAA. After angiotensin II treatment, nuclear proteins were isolated and then separated via SDS-PAGE. After transfer onto nitrocellulose membranes, Stat1 nuclear translocation was measured by Western blotting the samples with anti-Stat1 mAb (Fig. 5A). In response to angiotensin II, Stat1 was found only in the nucleus of cells transfected with wild-type Jak2 plasmid, but not in that FIG. 3. Conversion of 231 YRFRR to 231 FAAAA inhibits the angiotensin II-dependent co-association of Jak2 with the AT 1 receptor, but not Jak2 tyrosine autophosphorylation. COS-7 cells were transfected with 10 g of HA-tagged AT 1 receptor and either 5 g of wild-type Jak2 (Jak2 WT) or 12.5 g of Jak2-231 FAAAA. The cells were treated with 10 Ϫ7 M angiotensin II (Ang II) for the indicated times, and lysates were prepared. A, lysates were immunoprecipitated (IP) with anti-HA mAb and immunoblotted (IB) with anti-Jak2 pAb to assess AT 1 /Jak2 co-association (upper panel). Equal aliquots from these samples were immunoprecipitated and blotted with anti-Jak2 pAb to assess Jak2 expression (lower panel). B, lysates were immunoprecipitated with anti-Jak2 pAb and then blotted with either anti-Tyr(P) mAb to assess Jak2 tyrosine autophosphorylation (upper panel) or anti-Jak2 pAb to measure Jak2 expression (lower panel). Shown is one of three independent experiments for each. FIG. 4. Conversion of 231 YRFRR to 231 FAAAA does not inhibit the angiotensin II-dependent tyrosine phosphorylation of Stat1. COS-7 cells were transfected with 10 g of HA-tagged AT 1 receptor and either 5 g of wild-type Jak2 (Jak2 WT) or 12.5 g of Jak2-231 FAAAA. The cells were treated with 10 Ϫ7 M angiotensin II (Ang II) for the indicated times, and lysates were prepared to assess Stat1 tyrosine phosphorylation. A, lysates were immunoprecipitated with anti-Tyr(P) mAb and Western-blotted (immunoblotted (IB)) with anti-Stat1 mAb. B, lysates were immunoprecipitated with anti-Stat1 mAb and Westernblotted either with anti-phospho-Stat1 pAb to assess Stat1 tyrosine phosphorylation (upper panel) or with anti-Stat1 mAb to assess equal precipitation of Stat1 protein (lower panel). Shown is one of three independent experiments for each. of cells transfected with Jak2-231 FAAAA. To establish a better time course of Stat1 nuclear translocation and to determine the efficiency and specificity of the nuclear extraction procedure, COS-7 cells were transfected exactly as described for Fig. 5A, but were then treated with 10 Ϫ7 M angiotensin II for 0, 10, 20, 30, 45, and 60 min. First, the samples were Western-blotted with anti-Stat1 mAb to assess the time course of Stat1 nuclear translocation (Fig. 5B, upper panel). This result reconfirmed that Stat1 translocated into the nucleus of cells that received wild-type Jak2 plasmid and angiotensin II, but not into that of cells that received Jak2-231 FAAAA, regardless of angiotensin II treatment. Analysis of three independent experiments indicated that the accumulation of Stat1 in the nucleus of cells transfected with wild-type Jak2 peaked ϳ20 -30 min after angiotensin II addition, whereas cells that received Jak2-231 FAAAA failed to mediate ligand-dependent Stat1 nuclear translocation. To determine the efficiency and specificity of the nuclear extraction procedure, the membrane was stripped and Western-blotted with an antibody directed against the SV40 large T antigen (Fig. 5B, middle panel). The large T protein is highly abundant in COS-7 cells and is found only in the nucleus of these cells (26,27). The results indicated that the samples were of nuclear origin as determined by the amount of large T protein found in each sample relative to a COS-7 whole cell lysate sample. To demonstrate that the nuclear preparations were devoid of cytoplasmic contamination, the membrane was blotted with an antibody directed against the cytoplasmic protein marker Eps15 (Fig. 5B, bottom panel) (28,29). The lack of Eps15 in the nuclear extracts indicated that the Stat1 was of nuclear origin and not from cytoplasmic contamination.
Collectively, these data show that, although Jak2-231 FAAAA is capable of autophosphorylation and tyrosine phosphorylating Stat1, its failure to co-associate with the AT 1 receptor results in the failure of Stat1 to translocate into the nucleus.
This result is reminiscent of the growth hormone-mediated activation of Stat3; recruitment of Stat3 to the growth hormone receptor is critically dependent on Jak2 kinase (30).
A Jak2 Molecule Containing a 231 FAAAA Mutation in Place of 231 YRFRR Fails to Mediate Stat1-dependent Gene Transcription-To assess the functional impact of the Jak2-231 FAAAA mutation, we chose to investigate angiotensin II-dependent, Stat1-mediated gene transcription. For this assay, COS-7 cells were transfected as described for Fig. 5, but we also included a luciferase reporter plasmid containing a Stat1-binding, sisinducible element. This plasmid contains a tandem repeat of a minimal DNA enhancer element, the thymidine kinase TATAcontaining promoter, and the firefly luciferase cDNA (31). Each copy of the DNA enhancer contains a sis-inducible element, a serum response element, and an AP-1-binding site. We have previously demonstrated that this plasmid is a good indicator of Jak/STAT-mediated gene transcription (12,16). The transfected cells were then treated with angiotensin II for varying times, and luciferase activity was measured (Fig. 6). Transfection of the Jak2-231 FAAAA mutant resulted in substantially reduced angiotensin II-mediated luciferase activity with respect to wild-type Jak2 controls. The greatest difference in luciferase values between the two groups occurred 24 h after angiotensin II addition (p Ͻ 0.00001; n ϭ 6 for each point).
These data indicate that the failure of Jak2-231 FAAAA to facilitate Stat1 nuclear translocation also impairs Stat1-dependent transcriptional activation.

DISCUSSION
This work is significant for three fundamental reasons. First, we have identified a minimal region of Jak2 ( 231 YRFRR) that is required for Jak2 co-association with a cell-surface receptor. Second, the generation of the Jak2-231 FAAAA mutant functionally uncoupled ligand-mediated Jak2 autophosphorylation from AT 1 receptor co-association. Third, the work suggests that when Jak2 fails to bind the AT 1 receptor, one functional consequence is that Stat1 does not translocate into the nucleus and therefore fails to mediate gene transcription.
In regard to 231 YRFRR-mediated receptor co-association, exactly how 231 YRFRR mediates the binding of Jak2 with the AT 1 receptor is not fully understood. Our data suggest that it is probably a structural motif rather than a simple phosphotyrosine interaction in that the Jak2-Y231F mutant bound the AT 1 receptor, whereas the Jak2-231 FAAAA mutant did not. Consistent with our observation is a report describing a naturally occurring Jak3 mutation from a patient with autosomal severe combined immunodeficiency; a single Y100C amino acid substitution prevents co-association of Jak3 with the common ␥-chain of the interleukin-2 receptor (32). Similar to our results, the authors of this report concluded that Tyr 100 does not mediate a phosphotyrosine-dependent interaction, but rather contributes to a larger structural motif, as the Y100C mutation disrupts receptor co-association, whereas a Y100F mutation does not. The amino acid sequence immediately downstream of Tyr 100 is 100 YRLRF, and this bears some homology to the 231 YRFRR motif described in this report.
With respect to the second point, when COS-7 cells are transfected with Jak2-231 FAAAA and then treated with angiotensin II, Jak2-231 FAAAA autophosphorylates, but does not co-associate with the AT 1 receptor. This observation suggests that ligand-mediated Jak2 autophosphorylation has therefore been uncoupled from AT 1 receptor co-association. Since the Jak2-231 FAAAA mutant can be activated independently of receptor co-association, one interpretation is that Jak2 autophosphorylation temporally precedes receptor co-association. The current findings are important in that they strongly suggest that AT 1 /Jak2 co-association is not necessary for angiotensin FIG. 5. Conversion of 231 YRFRR to 231 FAAAA inhibits the angiotensin II-dependent nuclear translocation of Stat1. COS-7 cells were transfected exactly as described in the legend to Fig. 4. A, after treatment with 10 Ϫ7 M angiotensin II (Ang II) for the indicated times, nuclear proteins were prepared and electrophoresed via SDS-PAGE. The proteins were transferred onto a nitrocellulose membrane and blotted with anti-Stat1 mAb to assess Stat1 nuclear translocation. B, COS-7 cells were treated with 10 Ϫ7 M angiotensin II for longer time periods. Nuclear extracts were prepared from each time point and then subjected to SDS-PAGE along with a whole cell lysate (WCL) from non-transfected COS-7 cells. The proteins were transferred onto a nitrocellulose membrane and blotted as indicated. Shown is one of three independent experiments. Jak2 WT, wild-type Jak2.
II-induced Jak2 activation and that Jak2 activation, although necessary, is not sufficient for its interaction with the AT 1 receptor (11).
Perhaps most significant is the observation that when Jak2 fails to bind the AT 1 receptor, Stat1 does not translocate into the nucleus and therefore fails to mediate gene transcription. The data in Fig. 4 demonstrate that Stat1 tyrosine phosphorylation alone is insufficient for Stat1 nuclear translocation, and the data in Fig. 5 indicate that failure of Jak2 to bind the AT 1 receptor correlates with markedly decreased Stat1 nuclear translocation.
Our finding that Stat1 tyrosine phosphorylation alone is insufficient for Stat1 nuclear localization has precedent in other published studies. Herrington et al. (33) reported that a Stat5b protein with a small mutation in the DNA-binding region was defective in nuclear location despite its ability to be tyrosine-phosphorylated and to dimerize in response to ligand. Similarly, Strehlow and Schindler (34) reported a modified Stat1 protein that was phosphorylated on tyrosine and that dimerized, but was unable to be transported to the nucleus. Although these reports described a phenotype similar to what we observed with Jak2-231 FAAAA, in both studies, the phenotype was a direct result of a STAT structural mutation. Nonetheless, these studies support our observation by providing other examples of STAT proteins that are phosphorylated on tyrosine and that dimerize, but are unable to transport to the nucleus.
We recently reported that, in response to angiotensin II, Jak2 not only serves as a Stat1 kinase, but also acts as a molecular bridge in recruiting Stat1 to the AT 1 receptor (35). One interpretation of the Jak2-231 FAAAA data presented here is that recruitment of Stat1 to the AT 1 receptor by wild-type Jak2 further modifies Stat1 in a manner that promotes its nuclear translocation. The exact modification(s) that occur at the AT 1 receptor and whether this phenomenon is applicable to other STAT family members is under intense investigation. The concept that Jak2 acts as a molecular bridge in recruiting STAT proteins to a cell-surface receptor is not without precedent. In the case of growth hormone, Jak2 acts as a scaffold in recruiting Stat3 to the growth hormone receptor (30). The present data are significant in that they extend this observa-tion by demonstrating that when Jak2 fails to act as a scaffold in recruiting Stat1 to the AT 1 receptor, one consequence is that Stat1-mediated gene transcription is severely compromised.
In summary, this work provides novel insight into the mechanism of Jak2 activation and receptor co-association. Specifically, it defines a region of Jak2 that is required for receptor co-association and advances our understanding of the temporal sequences that occur when the Jak/STAT signaling pathway is activated. Most importantly, this work establishes a correlation between the loss of a Jak2 signaling complex at the level of a cell-surface receptor with the loss of transcriptional regulation within the nucleus. FAAAA inhibits angiotensin II-mediated, Stat1-dependent transcriptional activation. COS-7 cells were transfected with 10 g of HA-AT 1 , 10 g of luciferase reporter plasmid, and either 5 g of wild-type Jak2 (f) or 12.5 g of Jak2-231 FAAAA (q). After angiotensin II treatment, luciferase activity was measured from detergent-soluble extracts. Values are expressed as the mean Ϯ S.D. (n ϭ 6 for each time point). *, p Ͻ 0.0001; **, p Ͻ 0.001; ***, p Ͻ 0.00001 (Student's t test). Shown is one of three independent experiments.