Transcriptional regulation of the Drosophila raf proto-oncogene by Drosophila STAT during development and in immune response.

The Drosophila raf (D-raf) gene promoter contains a recognition consensus sequence for Drosophila STAT (D-STAT). By band mobility shift assay, we detected a factor binding to the D-STAT-recognition sequence in extracts of cultured Drosophila cells treated with vanadate peroxide. UV-cross-linking analyses suggested the size of the binding factor to be almost same as that of D-STAT. Furthermore, the binding activity was increased in cells cotransfected with HOP and D-STAT expression plasmids. These results strongly suggest that D-STAT binds to the D-STAT recognition sequence in the D-raf gene promoter. Transient luciferase expression assay using Schneider 2 cells indicated that the D-raf gene promoter is activated by D-STAT through the D-STAT-binding site. Furthermore, analyses with transgenic flies carrying Draf-lacZ fusion genes with and without mutations in the D-STAT-binding site pointed to an important role in D-raf gene promoter activity throughout development. We also found that the D-STAT-binding site is required for injury-induced activation of the D-raf gene promoter. Here we propose that D-STAT can participate in regulation of the mitogen-activated protein kinase cascade through D-raf gene activation.

The mitogen-activated protein kinase (MAPK) 1 cascade, activated in response to a variety of ligands, is highly conserved among eukaryotic organisms including yeast, Drosophila, and mammals (1)(2)(3). Raf, a constituent of the MAPK cascade, belonging to a family of serine/threonine protein kinases, acts as an important mediator of signals between upstream tyrosine kinases and downstream serine/threonine kinases in regulation of cell proliferation, differentiation, and development (4,5). D-raf, a Drosophila homolog of the human c-raf-1 gene, has been cloned, and mutants defective for this gene have been identified (6). Through analysis of the D-raf mutant phenotypes, it was found that D-raf functions in regulation of cell proliferation, as does mammalian c-raf-1, and in the determination of cell fates at embryonic termini (6 -8). D-raf is expressed throughout development in a wide range of tissues with high levels in tissues containing rapidly proliferating cells (7,8). Although multiple roles for D-raf in the regulation of cellular proliferation and differentiation have been demonstrated, little is known about the mechanisms controlling D-raf gene expression. In previous studies, we showed regulation by the DNA replication-related element (DRE)/DRE-binding factor (DREF) regulatory system (9), which appears to be of general importance for DNA replication- (10,11), cell cycle- (12), and proliferation-related (9) genes in Drosophila. Furthermore, the D-raf gene is probably another target of the Zerknü llt homeodomain protein-like DNA replication-related genes, such as DNA polymerase ␣ and proliferating cell nuclear antigen (9,13,14).
In addition to DRE, the D-raf gene promoter contains a Drosophila signal transducer and activator of transcription (D-STAT) recognition consensus sequence, 5Ј-TTCNNNGAA (15). The transcription factor STAT is known to be activated by the Janus kinase (JAK) in response to a variety of cytokines, growth factors, and interleukins in mammals (16,17). In Drosophila, a single JAK encoded by the gene hopscotch (hop) and STAT encoded by the gene marelle, also known as STAT92E or D-STAT, have been characterized (18 -20). Both have high homology with their mammalian counterparts (18 -20). Regarding D-STAT as an activator of transcription, only the pair rule gene even-skipped (eve) has been identified as a target (20).
In this study, we examined the role of D-STAT in regulation of the D-raf gene promoter. The obtained results indicate that the D-raf gene is a target of D-STAT activated by HOP and suggest the possibility of participation in regulation of the MAPK cascade.

EXPERIMENTAL PROCEDURES
Oligonucleotides-The sequences of double-stranded oligonucleotides containing the D-STAT recognition sequence in the D-raf gene (DrafSTATwt) and its two base-substituted derivatives (DrafSTAT-mut1 and DrafSTATmut2) were described previously (15). The oligonucleotide ODBS (optimum D-STAT-binding site) determined by sequence selection was prepared as described previously (20). The oligonucleotides eveS1 and eveS2 containing D-STAT-binding sites in the evenskipped (eve) gene stripe 3 promoter region were as described earlier (21). The following double-stranded oligonucleotides contained a 6-base pair linker sequence recognizable by BglII and BamHI and were chemically synthesized. The 2ϫODBSwt, 2ϫDrafSTATwt, and 2ϫDraf-* This work was supported in part by Korean Science and Engineering Foundation Grant 976-0500-004-2 (to M.-A. Y.) and grants-in-aid from the Ministry of Education, Science, Sports and Culture, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Plasmid Construction-A DNA fragment containing the D-raf promoter region (Ϫ663 to ϩ302) was isolated from the plasmid p5Ј-663DrafCAT (previously called p-878DrafCAT) (9) by digestion with SmaI and BstXI and blunt-ended. The fragment was then inserted into the SmaI site of the plasmid PGVB (Toyo Ink.) to create p5Ј-663Drafwtluc. The expression plasmids of D-STAT and HOP contain coding regions placed under control of the Drosophila actin 5C gene promoter derived from pPAC5C-PL (22). The oligonucleotide 2ϫODBSwt was inserted between SmaI and SacI sites of the plasmid TATA-PGVB (23) to create the plasmid 2ϫODBS-TATA-luc. The oligonucleotides 2ϫDrafSTATwt and 2ϫDrafSTATmut1 were inserted between SmaI and SacI sites of the plasmid TATA-PGVB to create 2ϫDrafSTATwt-TATA-luc and 2ϫDrafSTATmut1-TATA-luc, respectively. To construct p5Ј-663DrafSTATmut1-lacZ, a DNA fragment containing base-substituted mutations in the D-STAT-binding site of the D-raf gene promoter region was isolated from p5Ј-663DrafSTATmut1CAT by digestion with KpnI. Then the fragment was inserted into the KpnI site of the plasmid pCaSpeR-AUG-␤gal (24) having the P element. To construct the glutathione S-transferase (GST)-D-STAT-expressing plasmid, the polymerase chain reaction product containing full-length D-STAT (amino acids 1-761) was inserted between BamHI and NotI sites of pGEX-4T-2 (Amersham Pharmacia Biotech). This expression plasmid was used for large scale preparation of the D-STAT polypeptides, which were then applied as an antigen to raise antibodies. Extraction and purification of recombinant protein were carried out as described previously (25).
Antibody-GST-D-STAT was used to immunize mice as described previously (26). Antibody specific to D-STAT was affinity-purified from antiserum with GST-D-STAT-conjugated Sepharose after preabsorption with GST-conjugated Sepharose.
Cell Culture, DNA Transfection, and Assay of Luciferase-Drosophila Kc and Schneider 2 cells were grown at 25°C in M3 (BF) medium (27) supplemented with 2 and 10% fetal calf serum, respectively. Drosophila l (2)mbn hemocyte cells (28) were grown at 25°C in Schneider's medium (Sigma) supplemented with 10% fetal bovine serum. To induce immune reaction in l(2)mbn cells, culture medium containing LPS (a component of the outer membrane in Gram-negative bacteria) purchased from Sigma was added to cells. Transfection of various DNA mixtures in Kc and Schneider 2 cells was performed using the Cell-Fectin reagent (Life Technologies, Inc.). The luciferase assay was carried out by means of a PicaGene assay kit (Toyo Ink.) as described previously (29). Firefly luciferase activities were normalized to protein amounts determined by a Bio-Rad protein assay.
Treatment of Kc and Schneider 2 Cells with the Vanadate/Peroxide Mixture and the Band Mobility Shift Assay-To induce phosphorylated D-STAT protein, M3(BF) medium containing the indicated concentrations of sodium orthovanadate and hydrogen peroxide was preincubated for 15 min and then was added to growing cells. After incubation for 30 min, whole cell lysates were prepared by solubilizing cells for 10 min on ice in a buffer containing 20 mM Hepes (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 10 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 25% glycerol, and proteinase inhibitors containing 0.4 g of aprotinin, 0.4 g of leupeptin, 0.2 g of pepstatin A, 0.2 g of antipain, and 200 g of Pefabloc SC. After centrifugation at 10,000 ϫ g for 10 min at 4°C, the supernatant was divided into aliquots and stored at Ϫ80°C. Preparation of nuclear extracts was performed as described elsewhere (30). Band mobility shift assay was carried out as described earlier (10). In experiments with antibodies, Kc cell nuclear extracts were preincubated with antibody for 2 h on ice.

Establishment of Transgenic Flies and Quantitative Measurement of ␤-Galactosidase Activity in Extracts-
To establish transgenic flies carrying p5Ј-663DrafSTATmut1-lacZ, P element-mediated germ line transformation was carried out as described earlier (33,34). Quantitative measurements of ␤-galactosidase activity in larval extracts were carried out as described previously (34). In this study, the transgenic flies carrying p5Ј-663Drafwt-lacZ (previously called p5Ј-1103Draf-lacZ) (35) were used as a positive control. To correct for endogenous ␤-galactosidase activity, extracts from the wild-type strain (white) were included in each experiment, and this background reading was subtracted from readings obtained with each transformant line.
Treatment of Embryos with Vanadate/Peroxide Mixture, Injury Experiments of Larvae, and X-gal Staining of Larval Fat Bodies-Embryos carrying one copy of p5Ј-663Drafwt-lacZ or p5Ј-663DrafSTAT-mut1-lacZ fusion genes were collected 5 h after mating of the transgenic lines with host strain w, dechorionated, and treated with octane. They were then exposed to 1 mM sodium orthovanadate and 2 mM hydrogen peroxide for 30 min at 25°C, washed in phosphate-buffered saline buffer, and further incubated for 3 h at 25°C. Quantitative measurements of ␤-galactosidase activity were carried out as described above. Injury experiments were performed by pricking third instar larvae 72 Ϯ 4 h after egg laying, with a fine needle (34). Histochemical analyses of lacZ expression were conducted as described earlier (34 -36).

Location of a Potential D-STAT Recognition Sequence in the D-raf Gene
Promoter-Since the transcription initiation site of the D-raf gene was uncertain, we determined the site by primer extension analysis. 2 The mapped transcription initiation site is 215 base pairs upstream from the reported putative cap site (9), predicted from the consensus signal sequence for transcription initiation of Drosophila mRNA (37). In the region between Ϫ82 and Ϫ74 with respect to the newly determined transcription initiation site, we found the nucleotide sequence 5Ј-TTCGCG-GAA (15) that perfectly matches the D-STAT recognition consensus sequence, 5Ј-TTCNNNGAA (20) (Fig. 1). Since the sequence is located close to the transcription initiation site, we considered that D-STAT may be involved in regulation of the D-raf gene promoter. We designated this putative D-STAT recognition sequence the Draf-STAT site. As noted previously (9), DRE is located in the region between Ϫ155 and Ϫ142 ( Fig. 1).
Detection of Factors Binding to the ODBS and Draf-STAT Site in Vanadate/Peroxide-treated Cells-It has been noted that activated D-STAT protein in Drosophila cells treated with vanadate/peroxide mixture can bind to oligonucleotide ODBS with high affinity (20). It is well known that vanadate/peroxide inhibits tyrosine phosphatases and maintains D-STAT in an active phosphorylated form (20,38). In a band mobility shift assay using extracts of Kc cells treated with vanadate/peroxide, the binding activity to oligonucleotide ODBS proved greater in vanadate/peroxide-treated Kc cell extracts than that in vanadate-treated cell extracts ( Fig. 2A). The DNA-protein complex was diminished effectively by adding unlabeled oligonucleotides ODBS, eveS1, and eveS2 as competitors (Fig. 2B). Effective competition was also observed with the addition of the DrafSTATwt oligonucleotide, while oligonucleotide DrafSTAT-mut1 ( Fig. 1) did not compete for binding (Fig. 2B). From the extent of competition, it appears that the Draf-STAT and eve gene D-STAT-binding sites have comparable affinity to the vanadate/peroxide inducible factor, probably a D-STAT protein. Essentially similar results were obtained with Schneider 2 cell extracts (data not shown).
To determine if the shifted band represents a ODBS-D-STAT complex, anti-D-STAT antibody was added to the binding re-action with vanadate/peroxide-treated Kc cell nuclear extracts. As shown in Fig. 2C, the addition of the anti-D-STAT antibody supershifted the DNA-protein complex, and this supershifted complex formation was effectively diminished by adding unlabeled oligonucleotides ODBS and DrafSTATwt as competitors.
Since the vanadate/peroxide-inducible factor binding to ODBS was competitively blocked by the DrafSTATwt oligonucleotide, a band mobility shift assay was carried out again using the latter as a probe. As observed with ODBS, binding activity was enhanced in cells treated with increasing concentrations of vanadate/peroxide (Fig. 3A). The DNA-protein complex disappeared on adding unlabeled DrafSTATwt oligonucleotide as a competitor but not the mutant oligonucleotide carrying a base-substitution in the D-STAT core binding sequence, DrafSTATmut1, even when added in excess (Fig. 3B). The oligonucleotide DrafSTATmut2 (Fig. 1) carrying a basesubstitution in a noncritical sequence for binding of D-STAT competed effectively (Fig. 3B). These results taken together indicate that the vanadate/peroxide-activated D-STAT protein has affinity for ODBS, eveS1, eveS2, and Draf-STAT sites.
Size Determination for the Polypeptide(s) Binding to the Draf-STAT Site by UV-cross-linking Analysis-To determine the size of the polypeptide(s) binding to the Draf-STAT site, a UV-cross-linking assay was carried out using the UVDraf-STAT oligonucleotide as a probe and extracts of vanadate/ peroxide-treated cells. As shown in Fig. 4, polypeptides at around 88 kDa were specifically cross-linked with the probe. A lesser amount of radiolabeled polypeptides was observed on adding unlabeled DrafSTATwt oligonucleotides as competitors. However, the DrafSTATmut1 oligonucleotides did not compete. Essentially the same results were obtained with different cell lines, Kc (Fig. 4A) and Schneider 2 (Fig. 4B). Since the molecular weight of D-STAT protein is reported to be 87,500 and the D-STAT protein migrates to a position corresponding to 88 kDa on SDS-polyacrylamide gel electrophoresis (20), it is again very likely that the protein bound to the probe is the D-STAT protein.
Increase in Binding to the D-STAT Recognition Sequence in D-STAT and HOP Transiently Expressed Cells-To confirm that the factor binding to the D-STAT recognition sequence is activated by HOP, we performed a band mobility shift assay using D-STAT and HOP transient expression cell extracts. The DNA-protein complex formed with oligonucleotide ODBS in- creased effectively upon D-STAT and HOP coexpression (Fig. 5A, lane 4) but was diminished when unlabeled oligonucleotides, ODBS, eveS1, eveS2, or DrafSTATwt were added as competitors. Little competition was observed when the Draf-STATmut1 oligonucleotide was added (Fig. 5B). From these results, it is likely that D-STAT activated by HOP has high affinity for the Draf-STAT site and D-STAT-recognition sites in the eve gene and ODBS.
D-STAT and HOP Enhance D-raf Gene Promoter Activity-Since D-STAT very likely binds to the Draf-STAT site, we conducted the following experiments to test whether D-STAT activated by HOP can regulate activity of the D-raf gene promoter. Expression of the HOP protein had little effect on the D-raf gene promoter activity (Fig. 6A). However, expression of D-STAT protein gave an enhanced transcriptional signal possibly because of endogenous HOP expression (Fig. 6A). When D-STAT and HOP protein were expressed simultaneously, the p5Ј-663Drafwt-luc produced the strongest transcriptional signal (Fig. 6A). These results indicate that D-raf gene promoter activity is dependent on the amounts and ratio of D-STAT to HOP.
For further confirmation of the Draf-STAT sequence as a target of the activated D-STAT, three luciferase reporter constructs were prepared in which transcription of the luciferase genes was driven by the Drosophila metallothionein gene basal promoter carrying two copies of the ODBS wild type sequences (2ϫODBS), the Draf-STAT wild type (2ϫDrafSTATwt), or its mutant type (2ϫDrafSTATmut1). The HOP and D-STAT expression alone induced transcriptional activation of 2ϫDraf-STATwt-TATA-luc by 2-fold (Fig. 6B). When both HOP and D-STAT were expressed in Schneider 2 cells, the promoter activity of 2ϫDrafSTATwt-TATA-luc increased much more effectively (17-fold) (Fig. 6B). In contrast, little activation by HOP and D-STAT expression was observed with 2ϫDrafSTAT-mut1-TATA-luc (Fig. 6B). Thus, the Draf-STAT sequence functions as a target of the D-STAT protein, which appears to be activated by HOP protein.
Role of the Draf-STAT Site in Transcriptional Activation of the D-raf Gene during Development-To confirm a role of the Draf-STAT site for D-raf gene promoter activity in vivo, transgenic Drosophila were used. Previously, we established three independent transgenic fly lines carrying a wild type D-raf gene promoter-lacZ fusion gene (p5Ј-663Drafwt-lacZ) (35). In the present study, four independent transformant lines carrying the p5Ј-663DrafSTATmut1-lacZ fusion gene that has a mutation in the D-STAT core binding sequence were established. As shown in Fig. 7, mutation in the Draf-STAT site resulted in extensive reduction of lacZ expression throughout development.
The Draf-STAT Site Is Required for Vanadate/Peroxide-inducible Transcriptional Activation of the D-raf Gene in Living Flies-To confirm whether expression of the D-raf gene is regulated by the activated D-STAT in living flies, the dechorion-ated and permeabilized embryos were incubated in vanadate/ peroxide mixture. Transcriptional activation of the p5Ј-663Drafwt-lacZ transgene by vanadate/peroxide treatment is shown in Fig. 8. The level of ␤-galactosidase activity of the vanadate/peroxide-treated embryos bearing one copy of the p5Ј-663Drafwt-lacZ fusion gene was 2.2-fold higher than that of untreated embryos. In embryos bearing one copy of p5Ј-663DrafSTATmut1-lacZ fusion gene, the expression of lacZ was little induced by vanadate/peroxide treatment. Thus, it was confirmed in living embryos that the Draf-STAT site is required for vanadate/peroxide-inducible activation of the D-raf gene promoter.
Role of the Draf-STAT Site in Activation of the D-raf Gene Promoter in the Immune Response-The hemocytes play an important role in the events that lead to the activation of the immune response. Drosophila l(2)mbn hematocyte cell line is derived from larval hemocytes of the mutant lethal (2) malignant blood neoplasm (l(2)mbn) (39), in which the diptericin and cecropin genes are rapidly induced by the addition of LPS (40,41). Mammalian STAT has been implicated in a variety of immunity-related signaling pathway, including the response to proinflammatory cytokines (42).
Previously, we reported that the D-raf gene promoter was activated by injury or bacterial challenge (34). To investigate whether the Draf-STAT site is involved in this activation, the band mobility shift assay was carried out with nuclear extracts of the l(2)mbn cells treated with LPS and the DrafSTATwt oligonucleotides as a probe. Specificity of binding was confirmed in competition with DrafSTATwt and its base-substituted derivative DrafSTATmut1. A 2-fold increase of the DNAprotein complex was detected in l(2)mbn nuclear extracts 1 h after LPS treatment (Fig. 9), indicating that D-STAT is activated in response to treatment with LPS.
To further examine the role of the Draf-STAT site in the immune response in vivo, larvae carrying one copy of the p5Ј-663Drafwt-lacZ fusion gene or p5Ј-663DrafSTATmut1-lacZ fusion gene were injured with fine needles, and after 3 h, corresponding to the time of maximum induction of the D-raf gene in larvae (34), quantitative analysis of ␤-galactosidase activity in total crude extracts of the larvae was carried out. After injury, the level of ␤-galactosidase activity in larvae carrying one copy of the p5Ј-663Drafwt-lacZ fusion gene was induced 1.7-fold, but that in larvae carrying the p5Ј-663DrafSTATmut1-lacZ gene did not change as shown in Fig. 10A.
Since activation of the D-raf gene promoter after injury was most prominent in the larval fat body, known to be involved in innate immunity, serving as the functional homologue of the vertebrate liver (34), expression of the Draf-lacZ fusion gene in this tissue was examined by histochemical staining. lacZ staining signals in larval fat bodies carrying the p5Ј-663Drafwt-lacZ after injury were dramatically increased, whereas lacZ expression was not induced in p5Ј-663DrafSTATmut1-lacZ transgenic larval fat bodies (Fig. 10B).

DISCUSSION
The D-raf gene promoter contains a D-STAT recognition sequence in the region between Ϫ82 and Ϫ74 (Fig. 1). In this study, we identified a factor binding to the Draf-STAT site as having almost the same molecular weight as D-STAT and showed that the Draf-STAT site plays an important role in activation of the D-raf gene promoter by HOP/D-STAT. We also observed that the Draf-STAT site is required for D-raf gene promoter activation during development and in response to injury. These results lead us to suggest that the D-raf gene is a target of activated D-STAT protein through which the latter can participate in regulation of the MAPK cascade.
In the embryos homozygous for null alleles of D-raf, the embryogenesis proceeds normally depending on the maternal D-raf activity. However, the proliferation in imaginal discs and other proliferating tissues is severely affected during larval stages, indicating that the zygotic D-raf activity is essential after hatching (6,7). It is remarkable that expression of D-raf during larval stages, especially in early stages, is largely dependent on the transcriptional activation through the D-STATbinding site (Fig. 7). It has been reported that mutants defective in hop or D-STAT showed the similar proliferation defects (19,43). These observations suggest that the HOP/D-STAT signaling pathway contributes to the D-raf gene activation at significant stages of development.
Recently, the mosquito (Anopheles gambiae) STAT gene was cloned (44). It is noteworthy that bacterial challenge results in nuclear translocation of A. gambiae STAT protein in mosquito fat bodies and induction of DNA binding activity that recognizes an A. gambiae STAT target site (44). Furthermore, in vitro treatment with vanadate/peroxide enhances translocation of A. gambiae STAT into the nucleus in midgut epithelial cells (44). These observations provide evidence of direct participation of the STAT pathway in the immune response in insects and are in agreement with our conclusion that the Draf-STAT site is required for D-raf gene promoter activation after injury, which is supported by the results from the band mobility shift assays with the LPS-treated l(2)mbn cells (Fig. 9) and the injury experiment using transgenic flies (Fig. 10). In addition, it is well known that some STAT family members in mammals participate in the immune response (45) so that a function in immunity appears to have been conserved during evolution in mammals and insects.
In mammals, several lines of evidence pointing to cross-talk between JAK/STAT, the mammalian homologues of Drosophila HOP/D-STAT, and MAPK pathways have been reported. First, MAPKs activate some STAT family members, STAT1a, STAT3, and STAT4, by direct phosphorylation on the serine residue of STAT proteins, participating in the mechanism by which interferon stimulated ligand-induced early response genes (46,47). MAPKs also directly phosphorylate the interferon-␣/␤ receptor and stimulate the JAK/STAT pathway (48). On the other hand, it was recently shown that epidermal growth factor receptor is phosphorylated by Jak2 in signaling by growth hormone, thereby providing docking sites for Grb2 and activating MAPKs and their target gene expression, independently of the intrinsic tyrosine kinase activity of epidermal growth factor receptor (49). These reports suggest that the two signaling pathways are activated by mutual stimulation. Direct activation of the raf gene by STAT has not been reported. However, the human A-raf-1 gene promoter contains putative STAT binding sites (31). Although further analyses are necessary to generalize our findings to include mammals, the present study, Embryos bearing one copy of p5Ј-663Drafwt-lacZ or p5Ј-663DrafSTATmut1-lacZ were dechorionated, permeabilized, and incubated in the vanadate/peroxide mixture. To correct for endogenous ␤-galactosidase activity, the wild type (white) value was subtracted. The ␤-galactosidase activity of the vanadate/ peroxide-untreated p5Ј-663Drafwt-lacZ was scored as 1. Relative levels of expression are shown as the mean Ϯ S.E. of six independent experiments. A, third instar larvae having one copy of p5Ј-663Drafwt-lacZ or p5Ј-663DrafSTATmut1-lacZ fusion gene were sacrificed 3 h after injury, and crude extracts were prepared as described under "Experimental Procedures." The ␤-galactosidase activity of uninjured p5Ј-663Drafwt-lacZ was scored as 1. Relative levels of expression shown are mean Ϯ S.E. values from six independent experiments. B, histochemical staining of ␤-galactosidase activity in fat bodies of third instar larvae having one copy of p5Ј-663Drafwt-lacZ or p5Ј-663DrafSTATmut1-lacZ fusion gene. The fat bodies were dissected 3 h after injury and stained. a, uninjured larva carrying p5Ј-663Drafwt-lacZ; b, injured larva carrying p5Ј-663Drafwt-lacZ; c, uninjured larva carrying p5Ј-663DrafSTATmut1-lacZ; d, injured larva carrying p5Ј-663DrafSTATmut1-lacZ. for the first time, suggests a novel mechanism by which STAT can participate in regulation of the MAPK cascade through raf gene activation.