ACC Plays A Dual Role As A Key Metabolic Node In Rewiring Cancer Metabolism To Survive Therapy Targeting The Warburg Effect

Background Adenine Monophosphate-activated Protein Kinase（AMPK） and Acetyl-Co A Carboxylaze（ACC） are the key regulators of energy homeostasis and associated metabolic effects, including glucose and lipid metabolism. AMPK is a central sensor of cellular metabolism. It is activated during situations of metabolic stress that lower the intracellular levels of ATP, a crucial energy-supplying molecule. ACC produces malonyl-Co A, an intermediate metabolite that functions as substrate for fatty acid synthesis and as negative regulator of fatty acid oxidation. ACCs are biotin-containing enzymes that catalyze the carboxylation of acetyl-Co A to form malonyl-Co A[5]. In humans, there are two ACC isoforms, ACC1 and ACC2, that are encoded by two separate genes[6]. Cancer metabolism takes very important effect in cancer occurance, invasion, metastasis and drug resistance, while AMPK and ACC are the key points in cancer metabolism. We previously reported that cetuximab, an EGFR-blocking antibody[7], inhibits cancer metabolism and reverses the Warburg effect in cancer cells. These findings indicate that a transient activation of AMPK is an early metabolic marker of cellular response to cetuximab and that high and sustained AMPK activity is an important mechanism by which cancer cells survive cetuximab treatment[4]. We also found increased levels of phosphorylated AMPK and phosphorylated ACC and total level of ACC in the tumors of patients after cetuximab treatment. We designed the experiment to further clarify（1） the ACC expression pattern in Head and Neck Squamous Cell Carcinoma（HNSCC） tissues and cells;（2）the cancer metabolic function and regulational mechanism of ACC at clinical, cellular and animal levels.Method 1. The Hypoxia inducible factor-1α（HIF-1α）-overexpressing plasmid and Acetyl-Co A Carboxylase（ACC）-overexpressing plasmid were constructed. Lentiviral stable transfection system was used to establish stably overexpressing HEK293_HIF-1α cell line. Western blot and Glucose consumption assay were used to verify the expression and the level of glucose consumption.2. The Acetyl-Co A Carboxylase（ACC）-overexpressing plasmid was constructed. Lentiviral stable transfection system was used to establish stably overexpressing HEK293 HIF-1α/ACC1_S79A and HEK293 HIF-1α/ACC2_S212A cell line. Western blot and Live/dead cell viablility assay were used to verify the expression of phosphorylated and total AMPK, ACC and the cell survival under energy stress（low glucose）. 3. Measurement the expression of AMPK and ACC in HNSCC cell lines by using the Western blot method. Measurement the sensitive ability of HNSCC cell lines to Cetuximab. Furthermore, test the level of m RNA of ACC1 and ACC2 in different HNSCC cell lines. Inhibition of synthesis of protein by using Cycloheximide to measure the degradation of ACC. 4. The ACC1_S79A and ACC2_S212A-overexpressing plasmids, ACC1-si RNA and ACC2-si RNA transfected HNSCC cell lines. Western blot, Live/dead cell viability assay, Cell proliferation assay and Apoptosis assay were used to verify the level of AMPK and ACC, cell survival and apoptosis before and after of Cetuximab treatment. 5. 5-Tetradecyloxy-2- furoic acid（TOFA）, an inhibitor of ACC and fatty acid synthesis, by using Western blot, flow cytometry, MTT and apoptosis assay measure the effect of Cetuximab treatment to HNSCC cell lines. 6. Xenograft mouse model was constructed. We inhibited the expression ACC by TOFA and then measureed the proliferation of the tumors through luciferase image. Ananylsis the expression of ACC and AMPK in different treatment groups. 7. The clinical data paraffin-fixed HNSCC samples were collected for retrospective study, the expression pattern of AMPK-T172 p, ACC-S79 p and ACC were analyzed by using immohistochemistry. The relationship between the expression levels and Cetuximab treatment was analyzed.The results 1. We established a model in HEK293 cells with increased demand on glucose following overexpression of a constitutively active（degradation-resistant） HIF-1α mutant, HIF-1α_P402A/P564 A and HIF-1α₃₅ODD. As expected, overexpression of the HIF-1α mutants led to a significant increase in glucose consumption by the cells due to the Warberg effect, which was accompanied by activation of AMPK（shown by increased AMPK T172 phosphorylation） and inhibition of ACC（shown by increased ACC S79 phosphorylation）. HEK293 cells overexpressing the HIF-1α mutants(HIF-1α_P402A/P564 A and HIF-1α₃₅ODD) were significantly more sensitive to low glucose culture（1m M, i.e., 0.18g/L） than control vector infected HEK293 cells. 2. ACC plays a critical role in protecting cells from low glucose-induced death of cells exhibiting an increased dependence on glucose following overexpression of HIF-1α mutant. Overexpression of ACC1_S79A and ACC2_S212A had no impact on impact on the activation of AMPK and inhibition of endogenous ACC caused by low glucose culture, indicating that ACC can protect cell death caused by low glucose culture independently from the impact of AMPK-mediated compensation to low glucose-induced energy stress. Moreover, knockdown of endogenous ACC1 or ACC2, particularly ACC1, resulted in more cell death than low glucose culture only in both HEK293 and HEK293 HIF-1α_P402A/P564 A cells. It is noteworthy that nearly 80% of cells were found were found dead overnight（16h） after knockdown of ACC1 in HEK293 HIF-1α_P402A/P564 A cells. 3. The expression of ACC is corresponding to the resistance of Cetuximab to HNSCC cells. Earlier work reported that cetuximab activated AMPK and inhibited HNSCC cell proliferation via downregulation of HIF-1α and inhibition of HIF-1α-regulated glycolysis. Here we further found there is a reverse correlation between high basal levels of T172-phosphorylated AMPK and S79-phosphorylated ACC and cellular sensitivity to cetuximab in a panel of HNSCC cell lines. This correlation was confirmed in two isogenic pairs of cells with acquired resistance to cetuximab following chronic exposure to cetuximab: HN5/HN5-R and Fa Du/Fa Du-R. We further found that overexpression of the constitutively active HIF-1α mutant HIF-1α-P402A/P564 A in HN5 cells potentiated cetuximab-induced AMPK activation and ACC inhibition and conferred significant resistance to cetuximab. 4. An important finding was that in the cetuximab-resistant cells, there was an increase not only in the level of S79-phosphorylated ACC1 but also in the level of total ACC. This increase in total ACC was not due to increased transcription, as indicated by our findings that the m RNA levels of ACC（both ACC1 and ACC2）,measured by real-time PCR, differed minimally between HN5 and HN5-R cells and were modestly lowered by cetuximab in HN5 cells but not in HN5-R cells and that the level of SREBP-1c, a key transcription factor for activating ACC expression, was similar between HN5 and HN5-R cells and was decreased by cetuximab in HN5 cells but not in HN5-R cells. Results similar to those in HN5-R cells were observed in UMSCC1 and MDA1986 cells, two other HNSCC cell lines resistant to cetuximab-induced growth inhibition. Interestingly, ACC was remarkably more stable in HN5-R cells than in HN5 cells, indicating that a post-transcriptional mechanism was responsible for the increase in total ACC in cetuximab-resistant cells. Knockdown of endogenous ACC1 or ACC2 by respective si RNAs did not impact the level of apoptosis induced by cetuximab in parental HN5 cells, but knockdown of endogenous ACC1 or ACC2, particularly knockdown of ACC1, significantly enhanced apoptosis induction in HN5-R cells. Similar results were observed in UMSCC1 cells. 5. Cetuximab resistant HNSCC cells were remarkably more sensitive than sensitive cells to treatment with TOFA（5-[tetradecyloxy]-2-furoic acid）, a cell-permeable allosteric inhibitor of ACC. UMSCC1 and MDA1986, two other HNSCC cell lines resistant to cetuximab-induced apoptosis, also demonstrated high sensitivity to TOFA. Moreover, chronic exposure of UMSCC1 cells to TOFA in culture resulted in a TOFA-resistant subline, UMSCC1-Tofa R, that was sensitive to cetuximab. TOFA plus cetuximab markedly increased apoptosis induction in HN5-R, UMSCC1, and MDA1986 cells compared to control or either single treatment. 6. Xenograft mouse model was constructed that TOFA plus cetuximab produced a significantly greater therapeutic effect, as shown by two-dimensional tumor measurements and by IVIS tumor imaging. TOFA was well tolerated: there was no noticeable loss of body weight of the mice treated with TOFA or TOFA plus cetuximab. Immunohistochemical evaluation of the tumor specimens showed higher levels of T172-phosphorylated AMPK, S79-phosphorylated ACC1, and total ACC in cetuximab-treated HN5-R xenografts than in control HN5-R xenografts. TOFA and TOFA plus cetuximab downregulated the level of total ACC without significantly affecting the levels of T172-phosphorylated AMPK and S79-phosphorylated ACC1,consistent with the findings from cell culture. The therapeutic strategy was repeated using UMSCC1 xenografts, and results showed a TOFA dose-dependent effect on the outcome of TOFA plus cetuximab. 7. Under current National Comprehensive Cancer Network guidelines, cetuximab is approved only for patients with metastatic HNSCC after failure of standard chemotherapy. Few patients with HNSCC who undergo a standard 8-week cetuximab treatment undergo surgery afterwards. However, we were able to collect tumor specimens from six patients who underwent post-cetuximab surgery, including one who underwent surgery both before and after cetuximab treatment. Tumor specimens from another 12 patients who received the same chemotherapy regimen without cetuximab and underwent post-chemotherapy surgery served as a control group. In the patient who underwent surgery both before and after cetuximab treatment（patient 6）, comparison of the surgical samples showed markedly higher levels of these markers after cetuximab treatment.Conclusions 1. ACC plays a dual role（through activation of fatty acid oxidation and maintenance of lipogenesis） indispensable for maintaining cell growth and proliferation following an extended period in energy stress. 2. ACC plays a critical role in rewiring the cell metabolism pattern from glycolysis-dependent in cetuximab-sensitive cells to lipogenesis-dependent with fatty acid oxidation in cetuximab-resistant cells. 3. Co-targeting ACC, which we found is a rational therapeutic strategy for treatment of cetuximab-resistant disease, could also improve the outcomes of other therapies targeting the Warburg effect.