Analysis of human colorectal cancer (CRC) and mouse models of CRC reveals decreased expression of enzymes that mediate short chain fatty acid (SCFA) metabolism, including ACSS2. Unlike in many other cancers, low ACSS2 expression is associated with CRC progression, and predicts poor disease-free survival. ACSS2 generates cytosolic and nuclear acetyl CoA from acetate for de novo lipid synthesis and drives locus-specific histone acetylation to promote gene expression. Experimentally increasing ACSS2 in human CRC cells limits clonogenicity and growth in soft agar, consistent with ACSS2 restraining tumorigenesis. Colon tumors develop in a unique metabolic environment, in which gut bacteria generate high levels of SCFAs, including acetate and butyrate. Normal colon epithelia rely on bacterial SCFAs, primarily butyrate, for their energy needs. As with ACSS2, butyrate metabolic enzymes, including ACSM3 and ACADS, are downregulated in CRC, and low expression correlates with poor survival. We show that increasing ACSM3 expression reduces proliferation in human CRC cell lines, and that mice with a spontaneous deletion of ACADS are more susceptible to carcinogen induced colon adenomas. Thus, our preliminary data support a model in which ACSS2 and butyrate metabolic enzymes limit tumorigenesis specifically in the colon. Our overall hypothesis is that low ACSS2 expression promotes CRC via decreased histone acetylation and expression of its target genes, including those that encode butyrate metabolism. This facilitates a switch from SCFA metabolism to glycolysis in the tumor. However, it also creates a cancer-specific vulnerability, whereby CRC cells with low ACSS2 are impaired in their ability generate acetyl CoA for histone acetylation and de novo lipid synthesis. We propose to exploit this tumor-specific vulnerability therapeutically by targeting acetyl CoA synthesis together with nuclear processes that depend on ACSS2-mediated histone acetylation. Three Aims will address our overall hypothesis: 1) We will use mouse models, human CRC cell lines and xenografts to test if ACSS2 and ACADS restrain both primary colon tumors and metastases. 2) We will test if ACSS2 limits CRC by controlling locus-specific histone acetylation, and if ACSS2 drives butyrate metabolic gene expression to increase butyrate use and inhibit glycolysis. 3) We will test the hypothesis that low ACSS2 levels sensitize colon tumors to therapies targeting chromatin-templated functions that depend on ACSS2. This work will elucidate a novel tumor restraint function for ACSS2 in CRC, determine how ACSS2 controls metabolism, and identify new therapeutic approaches for CRC.

This work will determine how the enzymes that mediate short chain fatty acid metabolism limit colon cancer progression, and demonstrate that their reduced expression drives tumorigenesis. We will test the possibility that lower enzyme expression specifically in colon tumor cells creates a therapeutic vulnerability by altering both metabolic and gene expression programs. We will exploit this vulnerability in colon cancer to develop novel combinatorial therapeutic approaches that target metabolism and gene expression.