Abstract 824: LKB1 loss in lung adenocarcinoma leads to global hypomethylation and altered FOXA binding

LKB1 is functionally lost in 30% of lung adenocarcinoma cases. There are no targeted therapies available for LKB1-loss tumors. Previous studies of LKB1-loss in a mouse model of pancreatic cancer suggested that LKB1 loss drives DNA hypermethylation and serine biogenesis. DNA methyltransferases (DNMTs) maintain CpG methylation, which suppresses gene transcription. The primary substrate for DNMTs is S-adenosyl Methionine (SAMe), and the maintenance of CpG methylation has been shown to be dependent on continued regeneration of SAMe. KRAS status is associated both with LKB1 loss and with a CpG island hypermethylation phenotype (CIMP). Using a previously published 16-gene signature for LKB1 loss, we set out to confirm whether the changes that occur in the mouse model of pancreatic cancer also occur in lung cancer. We applied our signature to samples in The Cancer Genome Atlas (TCGA) lung adenocarcinoma dataset (signature-positive samples are described as “LKB1-loss”), which allowed us to compare both gene expression data as well as methylation data from Illumina 450k microarrays. We then developed a cell-line model for LKB1-loss in order to study the mechanism by which LKB1 regulates DNA methylation. We observed that LKB1-loss is associated with widespread demethylation of the genome in TCGA samples. Approximately 14,000 CpG sites were significantly hypomethylated in LKB1-loss tumors, while only 200 were hypermethylated (adjusted p-value cutoff = 0.05). Furthermore, Student’s t-test showed that LKB1-loss samples have significantly lower average methylation (β-values), regardless of KRAS status. Using motif analysis, we found that demethylated CpG sites are significantly enriched for FOXA1/2/3 consensus binding sites. FOXA transcription factors are well known as pioneering transcription factors that mediate demethylation. We discovered in vitro that FOXA1 localization and turnover is dependent on LKB1 expression, specifically the activity of downstream kinase SIK. Rather than showing enhanced serine biogenesis, LKB1-loss tumors express less DNMT1 and LC-MS/MS analysis shows depletion of SAMe. Finally, in vitro models of LKB1-loss show an increased sensitivity to DNMT inhibitors and FOXA1 silencing. Together, these observations suggest that LKB1-loss lung adenocarcinoma patients have a distinct hypomethylation profile. This profile may make them uniquely sensitive to DNMT-inhibition in combination with other therapies, a hypothesis which we will test in future studies. Citation Format: Michael J. Koenig, Joseph Amann, Bernice Agana, Jacob Kaufman, Vicki Wysocki, Christopher Oakes, David P. Carbone. LKB1 loss in lung adenocarcinoma leads to global hypomethylation and altered FOXA binding [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 824.