Liver kinase B1 (LKB1), encoded by the gene STK11, is a serine/threonine kinase that directly phosphorylates and activates a family of 14 downstream kinases known as the AMP-activated protein kinase (AMPK)-related protein kinases (AMPKRs). LKB1 is a tumor suppressor gene inactivated in the cancer predisposition disorder Peutz-Jeghers syndrome as well as in 20% of all human lung cancers. LKB1 plays critical roles in a number of developmental, physiological, and pathogenic processes through regulation of metabolism, cell growth, cell polarity and autophagy. LKB1 accomplishes these functions through its activation of the AMPKRs and thus delineation of which AMPKRs are critical for particular processes is an area of active interest and study.

After decades of connections to cancer and metabolic diseases, our initial studies into LKB1 and the AMPKRs uncovered an unexpected role for AMPK, a metabolic regulator in cells, in lineage specification during embryogenesis. Embryonic stem cells lacking AMPK are incapable of forming normal chimeric embryoid bodies (EBs) upon differentiation. These AMPK-null EBs expressed lower levels of the transcription factor Tfeb, a master transcriptional regulator of lysosomes, resulting in diminished endolysosomal function and a defective endoderm. These results collectively uncovered a mechanism by which LKB1 through AMPK acts as a novel regulator of cell fate determination during differentiation.

In parallel efforts to explore LKB1-dependent functions of AMPK and its related AMPKRs, we discovered that detachment of cells from their substratum is a universal promoter of LKB1-driven AMPKR activation and downstream substrate phosphorylation. Using an enrichment technique based on antibodies that capture a phosphorylated substrate motif targeted by AMPKRs for phosphorylation, we performed quantitative phosphoproteomic screens on a panel of detached mammalian cells that expressed LKB1, but were null for pairs of AMPKRs. We identified several known substrates of AMPKRs which underwent detachment-induced, LKB1-dependent increases in phosphorylation, thus validating our approach. We additionally identified several novel candidate proteins containing an optimal AMPKR motif which underwent an LKB1-dependent increase in phosphorylation. We validated six of these candidate proteins, including RASAL2, RalGAPα, SIPA1-Like proteins, MISP, BAIAP2L1 and MTSS2 as novel AMPKR substrates. Collectively, these results reveal unexpected effectors of LKB1 function and shed light on specific roles these AMPKRs play in cells.

Mutations in the Kirsten ras oncogene homolog from the mammalian ras gene family (KRAS) are often observed in patients suffering from several forms of cancers. These activating mutations in KRAS are suggested to be implicated in approximately 30% of non-small cell lung cancers (NSCLC). Among these patients with activating KRAS mutations, the most prevalent co-mutation occurs with the tumor suppressor serine/threonine kinase 11 (STK11), also known as Liver kinase B1 (LKB1). Co-mutations in KRAS and LKB1 make the tumors hyperaggressive and modulates the sensitivity to various therapeutics. The function of LKB1 in vivo is diverse, and is central to cellular metabolism, polarity, and growth. LKB1 phosphorylates and activates 14 downstream kinases known as the AMPK-related kinases (AMPKRs), which upon several known and unknown stimuli, control these above-mentioned cellular processes. While several seminal findings have been made regarding KRAS-LKB1 (KL) driven NSCLC, one of the most significant findings showed loss of the salt-inducible kinase family (SIKs), specifically SIK1 and SIK3, enhanced KRAS driven tumor growth, and approached that of the loss of LKB1. Here we show how loss of all three Sik family members in Kras-driven (KS) genetically engineered murine models of NSCLC leads to enhanced tumorigenesis, and most closely resembles the KL tumor growth phenotype. We further characterize distinct biological overlaps between KL and KS tumors, to show to what extend the Sik’s mediate Lkb1-dependent tumor biology and tumor suppression. Transcriptional analysis of these tumors revealed expression differences in several well characterized pathways, which changed in common in both KL and KS tumors. Ultimately we aimed to target several enriched pathways or individual genetic targets using an in vivo Cas9 screen, in order to uncover LKB1/SIK dependencies that can serve as future potential therapeutic or pharmacological targets for patients suffering from LKB1 mutations in NSLCLC.

Our laboratory recently identified the autophagy initiating kinase ULK1 as a novel AMPK substrate, which directly connects the highly conserved AMPK-mediated energy sensing pathway to the regulation of autophagy. Genetic deletion of ULK1 led to an accumulation of mitochondria with abnormal morphology, supporting a role for ULK1 kinase activity in the maintenance of mitochondrial homeostasis, yet few mammalian ULK1 substrates have been identified beyond those involved in the canonical initiation of autophagy. Here we have identified and validated the E3 ligase Parkin, a known mediator of mitochondrial quality control and the most commonly mutated gene in autosomal recessive early-onset Parkinson’s disease, as a direct ULK1 substrate. ULK1-mediated phosphorylation of Parkin was required for maximal Parkin activity in multiple assays of Parkin function. These data reveal an important role for ULK1-mediated Parkin phosphorylation in the regulation of mitochondrial homeostasis, as this modification seems to prime Parkin for maximal activity following depolarization. Further studies using mouse models have indicated that Parkin is phosphorylated in vivo in mouse livers and primary hepatocytes in response to activators of the AMPK/ULK1 pathway, including Metformin, the most widely prescribed drug for the treatment of Type II diabetes. These data suggest the potential for profound in vivo relevance for the regulation of Parkin function downstream of the AMPK/ULK1 signaling cascade.

AMP-activated protein kinase (AMPK) is a master regulator of metabolism. AMPK senses when cellular (and organismal) energy levels are low, and orchestrates adaptive changes to metabolic pathways, cellular processes, and gene expression programs to restore energy balance in response. Due its central homeostatic function, AMPK represents an important signaling mode in many contexts of health and disease, and an intriguing target for therapeutic intervention in many metabolic diseases.Elucidating specific downstream effects of AMPK and its molecular effectors is key for understanding the role of AMPK activation in different health contexts. While many functions of AMPK have been identified, particularly cytosolically, AMPK’s function in the nuclear compartment has remained elusive. With growing evidence that AMPK has important effects on gene expression, this work aims to illuminate nuclear AMPK activity: Here we examine its nuclear localization, nuclear targets, and transcription factor effectors of AMPK activation. Using a nuclear proteomics approach, we found that AMPKα1 shuttles into the nucleus following activation. We evaluated the upstream signals of this effect and found that an LKB1-dependent pathway is involved in response to both allosteric activation and mitochondrial stress. We performed a bioinformatics screen of transcription factors to identify potential direct targets of AMPK in the nucleus and validated the novel substrate SRF and several AMPK-SRF-dependent genes. Further, we optimized a biochemical approach for the identification of additional nuclear substrates. In addition to these protein-level studies, we perform transcript-level analyses. Major known effectors of AMPK-dependent transcription are the transcription factors TFEB and TFE3 (TFEB/3). Previous work has shown that transcription factors induced by TFEB/3 can be critical adaptive mediators of transcription downstream of AMPK activation. We searched for other possible key secondary effectors of transcriptional cascades downstream of AMPK-TFEB/3 by gene expression analysis and identified five transcription factors of interest which behave as AMPK-dependent TFEB/3 target genes: CREBRF, CREBL2, KLF9, KLF11, and KLF15. This combination of biochemical approaches, proteomic strategies, gene editing, RNA-sequencing, and bioinformatics analysis has culminated in the characterization of multiple aspects of nuclear AMPK activity, the identification of a novel nuclear substrate, and insights into AMPK-mediated transcriptional rewiring.

LKB1 is a serine/threonine kinase and tumor suppressor inactivated in 20% of lung cancers. Patients bearing mutant LKB1 have limited response to currently available therapies. Because LKB1 is known to activate a family of 14 kinases, understanding the critical targets of AMPK-related (AMPKR) kinases in lung cancer is critical to develop new therapeutics for LKB1-mutant lung cancers. I identify several novel substrates of the AMPKRs, including multiple small GTPase activating proteins (GAPs) family members, which mediate the tumor suppressive function of LKB1 in mutant KRAS genetically engineered mouse models (GEMMs) and lung cancer cell lines. While all candidate GAPs provided some tumor suppressive function within the lung, the RasGAP protein RAS protein activator like 2 (Rasal2) demonstrated the strongest activation and appears to phenocopy LKB1’s tumor suppressive function in vivo. I identified Rasal2 as a novel salt-inducible kinase (SIK) 1 and 3 substrate that is specifically phosphorylated at two serine sites, serine 56 and serine 89, in an LKB1-dependent manner. Collectively these findings reveal new LKB1-dependent signaling pathways could reveal new potential therapeutic avenues for this patient population. Additionally, I collaborated with Lillian Eichner to identify that HDAC3, a downstream substrate of the SIKs, plays a critical role in KRAS-mutant lung cancer. HDAC3 appears to direct and enhance the transcription effects of NKX2-1, a lung cancer lineage transcription factor, to mediate expression of a set of downstream target genes. Specifically, FGFR1 was identified as a critical new target of HDAC3. Using this, we were able to identify transcriptional cassettes that become hyperactivated in KRAS:LKB1 (KL) double mutant cells that develop resistance to MEK inhibitors such as trametinib, which can be reversed by an HDAC1/3 inhibitor entinostat. The combination of trametinib and entinostat shows therapeutic benefit in the KRAS:LKB1 double mutant GEMM.