Molecular Cloning and Characterization of the Human AKT1 Promoter Uncovers Its Up-regulation by the Src/Stat3 Pathway*

Akt1, also known as protein kinase B (PKB) α, is frequently activated in human cancers and has been implicated in many cell processes by phosphorylation of downstream molecules. However, transcriptional regulation of Akt1 has not been documented. Here, we report the isolation and characterization of the human AKT1 promoter and demonstrate transcriptional up-regulation of AKT1 by the Src/Stat3 pathway. Protein and mRNA levels of AKT1 are elevated in cells expressing constitutively active Stat3 as well as in v-Src-transformed NIH3T3 cells. Knockdown of Stat3 reduces AKT1 expression induced by v-Src. Although the 4.2-kb region upstream of the transcription start site of the AKT1 promoter contains five putative Stat3-binding motifs, the promoter failed to be induced by Stat3 and/or Src. Further analysis reveals that major Stat3 response elements are located within exon 1 and intron 1 regions of the AKT1 gene, which is upstream of the AKT1 translation initiation site. In addition, ectopic expression of wild type AKT1 in Stat3-/- MEF cells largely rescues serum starvation-induced cell death. These findings indicate that the AKT1 promoter comprises exon 1 and intron 1, in addition to the sequence upstream of transcriptional start site. Our data further show that AKT1 is a direct target gene of Stat3 and contributes to Stat3 anti-apoptotic function.

Akt1, also known as protein kinase B (PKB) ␣, is frequently activated in human cancers and has been implicated in many cell processes by phosphorylation of downstream molecules. However, transcriptional regulation of Akt1 has not been documented. Here, we report the isolation and characterization of the human AKT1 promoter and demonstrate transcriptional up-regulation of AKT1 by the Src/Stat3 pathway. Protein and mRNA levels of AKT1 are elevated in cells expressing constitutively active Stat3 as well as in v-Src-transformed NIH3T3 cells. Knockdown of Stat3 reduces AKT1 expression induced by v-Src. Although the 4.2-kb region upstream of the transcription start site of the AKT1 promoter contains five putative Stat3-binding motifs, the promoter failed to be induced by Stat3 and/or Src. Further analysis reveals that major Stat3 response elements are located within exon 1 and intron 1 regions of the AKT1 gene, which is upstream of the AKT1 translation initiation site. In addition, ectopic expression of wild type AKT1 in Stat3 ؊/؊ MEF cells largely rescues serum starvation-induced cell death. These findings indicate that the AKT1 promoter comprises exon 1 and intron 1, in addition to the sequence upstream of transcriptional start site. Our data further show that AKT1 is a direct target gene of Stat3 and contributes to Stat3 anti-apoptotic function.
There are no significant differences between three members of Akt in terms of upstream regulators and downstream targets. However, several lines of evidence suggest that the biological/physiological functions of Akt1, Akt2, and Akt3 are different. First, there are different levels of activation and protein expression in various cell types between Akt1, Akt2, and Akt3 (5,(15)(16)(17)(18). Second, multimeric complexes formed by Akt proteins are restricted to individual members, for example, Tcl1b binds to Akt1 and Akt2 but not Akt3, indicating the need to maintain specificity in interactions with other signaling proteins (19). Third, although Akt1 and Akt2 proteins require membrane localization for activity, only Akt2 appears to accumulate in the cytoplasm during mitosis (20) and in the nucleus during muscle cell differentiation (21). In addition, microinjection of anti-Akt1 antibodies blocks cell cycle progression, whereas anti-Akt2 antibodies have no effect on cell cycle but attenuate muscle differentiation (22). Amplification of the Akt1 gene has been detected in a single gastric carcinoma cell line (23), whereas the Akt2 is amplified in different types of human tumors (16 -17, 24). Ectopic expression of wild type Akt2, but not Akt1 and Akt3, results in invasion and metastasis in human breast and ovarian cancer cells (25) and induces a malignant phenotype in mouse fibroblasts (20). Finally, knock-out mouse studies demonstrated distinct phenotypes between Akt1, Akt2, and Akt3. The mice deficient in the Akt2 are impaired in the ability of insulin to lower blood glucose because of defects in the action of the hormone on skeletal muscle and liver, similar in some important features to type 2 diabetes in human. In contrast, Akt1-deficient mice do not display a diabetic phenotype. The mice are viable but display impairment in organismal growth with smaller organs than wild type littermates (26 -28). In contrast, a recent report shows that Akt3 Ϫ/Ϫ knockout mice only exhibit a uniformly reduced brain size, affecting all major brain regions, suggesting a central role of Akt3 in postnatal development of the brain (29).
Whereas Akt1 signaling has been extensively investigated, its transcriptional regulation remains largely unknown. In the present report, we have isolated and characterized the human AKT1 promoter. Multiple putative Stat3-binding sites reside within the promoter and major Stat3 response elements are identified in the exon 1 and intron 1 regions of the AKT1 gene. AKT1 is transcriptionally up-regulated by Stat3 and Src. The elevated expression level of AKT1 induced by Src is inhibited by dominant negative Stat3. Stat3 Ϫ/Ϫ mouse embryonic fibroblasts display lower levels of Akt1 and reintroduction of the AKT1 largely rescued cell death in response to serum starvation. Therefore, our results pro-vide the first evidence that AKT1 is a direct Stat3 target and mediates Stat3 survival signal.
Transcription Start Site Mapping of Human AKT1 Gene-For the analysis of the AKT1 transcription start site, human MCF7 mRNA was reverse transcribed at 55°C using SuperScript reverse transcriptase (Invitrogen) and a primer from the AKT1 exon 1-specific reverse complement oligonucleotide, 5Ј-TGACTTCTTTGACCCAGGCTGG-3Ј (31). Synthesized cDNAs were amplified by polymerase chain reaction using a series of forward primers specific for the DNA sequences within the 6,000 bp upstream of the translation start site and a reverse primer from the non-coding region of exon 1, and the products of these reactions were resolved by agarose gel electrophoresis.
Cloning and Analysis of Human AKT1 Promoter-To clone the 5Ј-flanking region of the human AKT1 gene, 4 cosmid clones were obtained by screening a human placenta genomic cosmid library (Stratagene) using the 5Ј non-coding region (a 520-bp fragment from ATG site) of the AKT1 as a probe. The cosmid clones were sequenced and compared with the human genome data base. Three clones were found to contain 12-28-kb DNA fragments upstream the translational start site. Two DNA fragments (Ϫ4293/ϩ1 and Ϫ4293/ϩ1888, Fig. 1C) were amplified with the GC-RICH PCR System (Roche) using a cosmid clone as template. The amplified DNA fragments, Ϫ4293/ϩ1 and Ϫ4293/ ϩ1888, were subcloned into the luciferase reporter vector pGL3 (Promega) at the KpnI/BglII site. Progressive deletion mutants of the pGL3-AKT1 promoter were created by PCR. The integrity of constructs was confirmed by DNA sequencing. The primers were used as follows: Luciferase Reporter Assay-NIH3T3 or HEK293 cells were cultured in 12-well plates and transiently transfected with AKT1-Luc, Src, and/or Stat3. The amount of DNA in each transfection was kept constant by the addition of empty vector. After 36 h of transfection, luciferase activity was measured using a luciferase assay reagent (Promega). Transfection efficiency was normalized by co-transfection with ␤-galactosidase expressing vector. The ␤-galactosidase activity was measured by using Galato-Light (Tropix). Luciferase activity was expressed as relative luciferase activity.
Northern and Western Blot Analysis-Northern blot analysis of total cellular RNA was performed according to standard procedures. RNA was extracted using the RNeasy purification kits (Qiagen Inc.). Total RNA was electrophoresed in 1.0% formaldehyde-agarose gels, transferred to Duralon-UV TM membrane (Stratagene), and then hybridized with randomly primed ␣-32 P-labeled cDNA probes for AKT1. Membranes were exposed to autoradiography and the mRNA levels were visualized and quantified using PhosphorImager analysis (Amersham Biosciences). Western blot analysis was performed as described previously (32). Briefly, the cells were lysed with RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 5 g/ml aprotinin, and 5 g/ml leupeptin), separated in SDS-PAGE, and immunoblotted with appropriate antibodies as indicated in the figure legends.
Chromatin Immunoprecipitation (ChIP) and EMSA Assay-ChIP assay was performed essentially as previously described (33). Solubilized chromatin was prepared from a total of 2 ϫ 10 7 asynchronously growing HEK293 cells that were transfected with wild type Stat3 and v-Src. The chromatin solution was diluted 10-fold with ChIP dilution buffer (1.1% Triton X-100, 1.2 mM EDTA, 167 mM NaCl, 16.7 mM Tris-HCl, pH 8.1, 0.01% SDS, protease inhibitors), and precleared with protein-A beads blocked with 2 g of sheared salmon sperm DNA and preimmune serum. The precleared chromatin solution was divided and utilized in immunoprecipitation assays with either an anti-Stat3 antibody or an anti-actin antibody. Following wash, the antibody-protein-DNA complex was eluted from the beads by resuspending the pellets in 1% SDS, 0.1 M NaHCO 3 at room temperature for 20 min. After cross-link, protein and RNA were removed by incubation with 10 g of proteinase K and 10 g of RNase A at 42°C for 3 h. Purified DNA was subjected to PCR with primers specific for 13 putative Stat3-binding sites within the AKT1 promoter. The sequences of the PCR primers used are as follows: region 1 forward (Ϫ4293), 5Ј-CTTCGTGAACATTAACGACAGGG-  25,2005 electrophoresis and visualized by BioImage. EMSA was performed as previously described (35).

Cloning of the Human AKT1 Promoter Revealed Multiple Putative
Stat3-binding Sites-To analyze the transcriptional regulation of the serine/threonine kinase AKT1, we cloned the 5Ј-flanking region of the human AKT1 gene. Sequence analyses revealed that the AKT1 gene consists of 12 exons. The exon 1 is 1.4 kb and is located within the 5Ј-untranslated region. The translation initiation site ATG resides within exon 2 (Fig. 1A). The transcription start site, which was determined by 5Ј-rapid amplification of cDNA ends PCR, lies 1,888 bp upstream of the translation start site. A putative TATA box was identified 7 bp upstream of the transcriptional start site. Transcription element analyses 3 of 6,181 bp upstream of the translation initiation site of the AKT1 gene revealed multiple binding sites for Stat3, NFB, AP1, and GC box within the these regions (Fig. 1B). The transcription factor that has the most binding sites in the AKT1 promoter is Stat3 (12 putative Stat3-binding sites: Ϫ4230/Ϫ4223, Ϫ4143/Ϫ4136, Ϫ4053/Ϫ4046, Ϫ3287/Ϫ3278, Ϫ2851/Ϫ2844, ϩ407/ϩ414, ϩ439/ϩ448, ϩ483/ϩ491, ϩ978/ϩ986, ϩ1047/ϩ1055, ϩ1324/ϩ1331, and ϩ1443/ϩ1450). Consensus sequences of the Stat3-binding site are TT(N) 4 -6 AA (37). These observations suggest that the AKT1 gene could be regulated by Stat3 at the transcriptional level.
Stat3 Increases AKT1 Expression at mRNA and Protein Levels-To directly demonstrate whether the AKT1 is transcriptionally regulated by Stat3, MCF-10A cells were infected with adenovirus expressing constitutively active Stat3 (Stat3C) and dominant negative Stat3. The cells infected with adeno-green fluorescent protein vector were used as control. Northern blot analysis showed that constitutively active Stat3C up-regulates the AKT1 (Fig. 2A). Furthermore, immunoblotting studies revealed an elevated protein level of AKT1 in cells treated with constitutively active Stat3C (Fig. 2B). Expression of dominant negative Stat3 slightly inhibited the mRNA and protein levels of AKT1 (Fig. 2, A and B).
As Stat3 is strongly activated by Src kinase (30), we next examined if the expression level of AKT1 is elevated in v-Src-transformed NIH3T3 cells. As shown in Fig. 2, C and D, both protein and mRNA levels of AKT1 were significantly increased in v-Src-transformed NIH3T3 cells. Furthermore, blockage of Stat3 with antisense RNA considerably reduced v-Src-induced AKT1. Moreover, up-regulation of AKT1 was also detected in the human breast cancer cell line MDA-MB-468, which exhibits constitutively active Src and Stat3, but not in MDA-MB453, which does not. Inhibition of Src/Stat3 by Src inhibitor (PD180970) or Stat3-siRNA reduced AKT1 protein level in MDA-MB-468 (Figs. 2, E and F). In addition, conditional knock-out of the Stat3 gene decreased Akt1 expression in mouse embryonic fibroblasts (Fig. 6A). Based on these data, we concluded that Akt1 is a downstream target of Stat3.
Stat3 Transactivates the AKT1 Promoter-We further examined whether the AKT1 promoter is regulated by Stat3. Luciferase reporter assay revealed that pGL3-AKT1Ϫ4293/ϩ1, which contains 5 putative Stat3-binding sites, was not stimulated by wild type or constitutively active Stat3, even in combination of Stat3 with v-Src (Fig. 3, A and B). However, ectopic expression of constitutively active Stat3 significantly induced the pGL3-Akt1Ϫ4293/ϩ1888 activity (Fig. 3C), suggesting that major Stat3 response elements reside in the region of the AKT1 gene between the transcription start site and translation initiation site, which contains exon 1 and intron 1 (Fig. 1B).
Src Induces AKT1 Promoter Activity through Stat3-As v-Src-transformed NIH3T3 cells express high levels of AKT1 that were attenuated by knockdown Stat3 (Fig. 2), we assumed that the AKT1 promoter should be induced by Src through Stat3. To test this hypothesis, a luciferase reporter assay was performed with NIH3T3 cells transfected with v-Src, pGL3-AKT1Ϫ4293/ϩ1888, and wild type Stat3. As shown in Fig.  4A, expression of v-Src alone induces the reporter activity about 1-fold, whereas co-expression of v-Src and Stat3 significantly stimulated the AKT1 promoter activity. Furthermore, wild type Stat3 enhances v-Srcinduced AKT1 promoter in a dose-dependent manner (Fig. 4B). Notably, expression of dominant negative Stat3 considerably reduces the AKT1 promoter activity induced by v-Src (Fig. 4C). These data indicate that Stat3 mediates Src-stimulated AKT1 promoter activity.
Stat3 Response Elements Are Primarily Located within the Exon 1/Intron 1 Region-As shown in Fig. 1C, 12 putative Stat3 binding motifs were identified within the ϳ6.0-kb region of the upstream translation initiation site of the AKT1 gene. Src and Stat3 induced pGL3-AKT1Ϫ4293/ϩ1888 but not pGL3-AKT1Ϫ4293/ϩ1 activity (Fig. 3), implying that the Stat3 response elements reside in a region between the transcription start site and translation initiation site. To test this, we created a series of deletion mutants of the AKT1 promoter. Reporter assay showed that pGL3-AKT1Ϫ325/ϩ1888 was significantly induced by coexpression of v-Src/Stat3 (Figs. 3C and 5A), whereas deletion of a region between Ϫ4293 and Ϫ3172, i.e. pGL3-AKT1Ϫ3172/ϩ1888 considerably reduced the promoter activity. Furthermore, pGL3-AKT1Ϫ3172/ϩ595 failed to respond to v-Src/Stat3 (Fig. 5A), indicating that the major Stat3 response elements exist in exon 1/intron 1 regions, where there are 7 putative Stat3 binding motifs, as well as in a distal region of the promoter. Notably, the results also suggest the presence of a repression factor binding site(s) within the Ϫ3172/Ϫ325 region because the promoter activity is significantly increased by deletion of this region as revealed by comparison of the activity between Ϫ3172/ ϩ1888 and Ϫ325/ϩ1888 (Fig. 5, A and B).
To determine whether Stat3 could directly bind to the Stat3-binding site within the AKT1 promoter in vivo and to further define the Stat3 response elements in the promoter, we carried out ChIP assay, which detects specific genomic DNA sequences that are associated with a particular transcription factors in intact cells. HEK293 cells were transfected with wild type Stat3 and v-Src and immunoprecipitated with a Stat3 antibody. The Stat3 bound chromatin was subjected to PCR using oligonucleotide primers that amplify regions spanning each Stat3-binding site within the AKT1 promoter. As show in Fig. 5C, the anti-Stat3 antibody pulled down four Stat3-binding sites (Ϫ4230/Ϫ4223, ϩ978/ ϩ986, ϩ1324/ϩ1331, and ϩ1443/ϩ1450 (SB1, SB7, SB9, and SB10)). In contrast, immunoprecipitation with an irrelevant antibody (anti-actin) resulted in the absence of bands in these sites. These results indicate that Stat3 directly binds to the AKT1 promoter. By mutation of Stat3 binding consensus sequences (TT 3 GG) in AKT1Ϫ325/ϩ1888 that is highly induced by Src/Stat3 and Stat3C (Fig. 5, A and B), we further demonstrated that the Stat3-binding sites (SB6, SB9, and SB10) within the exon 1 and intron 1 regions were required for Stat3 transactivation of the AKT1 promoter (Fig. 5D). Moreover, EMSA revealed that Stat3 is capable of binding in vitro to DNA oligonucleotides corresponding to the four Stat3 SIE/GAS binding sites identified in the Akt1 promoter (Fig.  5E). Unlike SIE positive control, however, the supershift was not detected in these four Stat3-binding sites within the promoter (Fig. 5E  and data not shown), which is also observed in other Stat3-induced promoters, such as vascular endothelial growth factor, Bcl-x L , and c-Myc (34,38,39).
AKT1 Mediates Stat3 Function-Both Stat3 and AKT1 play an essential role in cell survival (40 -45). Because AKT1 is a direct target of Stat3, we reasoned that AKT1 could mediate Stat3 function. To test this hypothesis, conditional knock-out of Stat3 MEF cells were infected with Cre and adenovirus expressing wild type AKT1 (Fig. 6A). Cell survival was evaluated after serum withdrawal for 24 and 48 h. Triple experiments showed that Cre-infected Stat3 MEF increased cell death approximately 30%. Ectopic expression of wild type AKT1 largely rescued knockdown Stat3-induced cell death (Fig.  6B). It is noted that the Akt1 protein level was considerably decreased in Stat3 knock-out MEFs, further suggesting the critical role of Stat3 in transcriptional regulation of the AKT1 gene. To further examine the effects of Akt1 on the Stat3 cell survival signal, constitutively active Stat3-presenting breast cancer cell line MDA-MB-468 was treated with Stat3-siRNA or Akt1-siRNA as well as the co-transfection of siRNA-Stat3 and HA-Akt1 (Fig. 6C). Tunel assay revealed that knockdown of either Stat3 or Akt1 induced cell death about 45-50% in response to serum starvation. However, reintroduction of Akt1 largely inhibits the apoptosis resulted from knockdown of Stat3 (Fig. 6D).

DISCUSSION
Alterations of AKT1 at the DNA level have been reported in a single gastric cancer (23). However, a number of tumors exhibit elevated levels of mRNA, protein, and/or kinase of AKT1 (46), implicating that AKT1 is regulated at the transcriptional, translational, and/or post-translational levels. Post-translational regulation of AKT1 has well been doc-umented (47). In this report, we cloned the human AKT1 promoter and demonstrated a number of transcription factor binding sites within the promoter. Notably, 12 putative Stat3 binding motifs were identified in the promoter, 4 of which were shown to directly bind to Stat3 in vivo and in vitro as revealed by ChIP and EMSA assays. The promoter activity is significantly induced by constitutively active Stat3. Furthermore, we demonstrated that ectopic expression of constitutively active Stat3 or v-Src significantly induces mRNA and protein levels of AKT1. Knockdown Stat3 decreased Akt1 expression in v-Src-transformed cells and mouse embryonic fibroblasts. Reintroduction of the AKT1 rescued the cell death in Stat3 Ϫ/Ϫ MEFs. We have noted that constitutively active, but not wild type Stat3 induces AKT1 promoter activity (Fig. 4A). However, knock-out Stat3 in MEFs and breast cancer cells significantly reduced Akt1 expression (Fig. 6, A and C). These results suggest that the basal level of Stat3, which could distribute to the cytoplasm and the nucleus, is required for expression of the normal level of its target genes, such as AKT1. Ectopic expression of wild type Stat3 may not increase the nuclear fraction of Stat3 and only constitutively active Stat3 could translocate into the nucleus and transactivate its target gene(s). This notion is supported by the observation that Stat3 is persistently activated in many human cancers and transformed cell lines and that only the activated Stat3 transform cells in cell cultures and induce tumor formation in nude mice (48,49). Akt and Stat3 pathways play an important role in cell processes associated with tumorigenesis such as cell survival, growth, and angiogenesis (50 -54). Cross-talk between these 2 pathways, however, has not been documented. Therefore, the data presented in this study provide the first evidence of the AKT1 gene as a direct downstream target of Stat3. Recently, we have also observed that inhibition of the AKT1 pathway induces hypoxia inducible factor-1␣ and that knockdown AKT1 largely abrogates the hypoxia inducible factor-1␣ expression stimulated by Stat3 (55). These findings further support that Stat3 targets Akt to exert its function.
It has been shown that Src is a key molecule for activation of the Akt pathway. We and others (11,56) have previously demonstrated that constitutively active Src induces Akt kinase activity through phosphatidylinositol 3-kinase. There is also evidence showing that Src directly binds to Akt and activates Akt through phosphorylation of Akt at tyrosine 315 (57). Inhibition of Src reduces Akt activation by growth factor(s). In addition, Src mediates estrogen/estrogen receptor and androgen/androgen receptor activation of Akt (58 -61). In the present report, we present evidence of Src regulation of Akt at the transcriptional level through activation of Stat3 (Fig. 6E). It is noted that the Akt1 promoter activity induced by Src is considerably lower than that by expression of constitutively active Stat3C (Figs. [3][4][5], suggesting that a repressive molecule(s) toward the Akt1 promoter is regulated by Src kinase.
In summary, we have isolated and characterized the AKT1 promoter. The promoter sequence analysis demonstrates Stat3 transcriptional regulation of the AKT1. Furthermore, we have shown that Src/Stat3 induces Akt1 expression through directly binding to the promoter. Blocking Stat3 by antisense or genetic knock-out significantly decreases Akt1 expression. Ectopic expression of AKT1 rescues the cell survival phenotype from Stat3 Ϫ/Ϫ MEF cells. These findings are important for several reasons. First, they provide a mechanistic understanding of regulation of the AKT1 at the transcriptional level. Second, a direct link between AKT1 and Stat3 pathways has now been established. Finally, pharmacological inhibition of the AKT pathway may have anti-growth effects in tumor cells with activation of Stat3 or vice versa.