Aim1: To examine the role of HSF1 in sustaining tissue overgrowth driven by oncogenic PI3K/AKT signaling. Our preliminary results show that the PI3K/AKT signaling cascade is required for activation of the HSR/PSR by heat shock (HS) in MEFs and for constitutive HSF1 activation in malignant cells. Importantly, AKT physically interacts with HSF1. Furthermore, AKT phosphorylates HSF1 at Ser230, and expression of the constitutively active AKT1 or loss of the tumor suppressor PTEN is sufficient to activate HSF1. By contrast, AKT inhibitors block HSF1 Ser230 phosphorylation and its DNA binding to HSP gene promoters. Furthermore, constitutively active PI3K/AKT signaling causes overgrowth or enlargement of both brains and livers in mice, conditions similar to megalencephaly and hepatomegaly in humans respectively, leading to rapid postnatal death. Importantly, simultaneous deletion of Hsf1 in both tissues impedes overgrowth and prolongs animal survival. Moreover, Hsf1 deletion also markedly impedes the liver overgrowth in mice deficient for Pten, a tumor suppressor negatively regulating PI3K activity, prolonging their survival. Our results further show that constitutively active PI3K/AKT disrupts proteostasis and induces proteotoxic stress, which is markedly heightened by Hsf1 deficiency. Based on these preliminary results, we plan to interrogate: 1) whether HSF1 is a new physiological substrate for AKT; 2) whether and how HSF1 suppresses proteotoxic stress induced by constitutive activation of PI3K/AKT signaling and thereby promotes tissue overgrowth in vivo; and 3) the molecular mechanisms underlying disrupted proteostasis in overgrown tissues.
Aim 2 : To examine the role of HSF1 in promoting lipid metabolism and protein lipidation. Our previous studies revealed that HSF1 is a physiological substrate for AMPK, a key cellular metabolic sensor, and that the AMPK-mediated Ser121 phosphorylation negatively regulates HSF1 activation. Now, our preliminary results using HSF1 deletion constructs deficient for transcriptional activity show that, just like the wild-type HSF1, they interact with AMPK and suppress AMPK Thr172 phosphorylation, a modification key to its activation, indicating a transcription-independent mechanism of action of HSF1. Conversely, Hsf1 deficiency causes AMPK activation, which is blocked by the AMPK inhibitor. Interestingly, our results show that HSF1 can be co-precipitated with both AMPK and LKB1, revealing a LKB1-AMPK-HSF1 protein complex. Furthermore, in human kidney and breast cancer samples higher HSF1 mRNA levels are inversely correlated with AMPK Thr172 phosphorylation, congruent with the results of our mechanistic studies. Our preliminary data show that Hsf1 deficiency and enhanced HSF1 expression result in diminished and heightened cellular lipid content, respectively, suggesting that HSF1 promotes lipogenesis to support malignancy. Strikingly, Hsf1-deficient mice display markedly reduced whole-body fat mass. Importantly, these effects of HSF1 on cellular lipid content and body fat mass can be markedly rescued by either AMPK inhibitors or siRNA-mediated AMPK knockdown, suggesting that the lipogenic effect of HSF1 is largely mediated via AMPK suppression. At the molecular level, HSF1 deficiency causes inactivation of SREBP1c, a key transcription factor controlling lipogenic gene expression, in addition to inactivation of ACC. Cholesterol is an important lipid implicated in many key cellular processes, including membrane composition, signaling transduction, and synthesis of steroid hormones. Congruent with diminished cellular lipid content, our results reveal a markedly reduced cellular cholesterol level caused by HSF1 deficiency, which is rescued by AMPK inhibition. Based on these preliminary results, we plan to investigate: 1) the molecular mechanisms underlying AMPK suppression by HSF1; 2) whether HSF1 promotes cholesteroylation of sonic hedgehog (SHH) proteins and supports SHH signaling; and 3) whether HSF1 promotes lipid metabolism and SHH cholesteroylation in xenografted human melanoma models.
| Su, Kuo-Hui; Dai, Chengkai (2017) mTORC1 senses stresses: Coupling stress to proteostasis. Bioessays 39: |
