Ketone body utilization drives tumor growth and metastasis

Abstract

We have previously proposed that catabolic fibroblasts generate mitochondrial fuels (such as ketone bodies) to promote the anabolic growth of human cancer cells and their metastasic dissemination. We have termed this new paradigm "two-compartment tumor metabolism." Here, we further tested this hypothesis by using a genetic approach. For this purpose, we generated hTERT-immortalized fibroblasts overexpressing the rate-limiting enzymes that promote ketone body production, namely BDH1 and HMGCS2. Similarly, we generated MDA-MB-231 human breast cancer cells overexpressing the key enzyme(s) that allow ketone body re-utilization, OXCT1/2 and ACAT1/2. Interestingly, our results directly show that ketogenic fibroblasts are catabolic and undergo autophagy, with a loss of caveolin-1 (Cav-1) protein expression. Moreover, ketogenic fibroblasts increase the mitochondrial mass and growth of adjacent breast cancer cells. However, most importantly, ketogenic fibroblasts also effectively promote tumor growth, without a significant increase in tumor angiogenesis. Finally, MDA-MB-231 cells overexpressing the enzyme(s) required for ketone re-utilization show dramatic increases in tumor growth and metastatic capacity. Our data provide the necessary genetic evidence that ketone body production and re-utilization drive tumor progression and metastasis. As such, ketone inhibitors should be designed as novel therapeutics to effectively treat advanced cancer patients, with tumor recurrence and metastatic disease. In summary, ketone bodies behave as onco-metabolites, and we directly show that the enzymes HMGCS2, ACAT1/2 and OXCT1/2 are bona fide metabolic oncogenes.

Figure 1. Fibroblasts overexpressing ketogenic enzymes show increased autophagy. hTERT fibroblasts overexpressing the ketogenic enzymes BDH1 and HMGCS2 were generated using a lentiviral approach. Western blot analysis was performed with antibodies against HMGCS2, BDH1, cathepsin B, beclin-1 and LC3B. Note that ketogenic fibroblasts overexpressing BDH1 and HMGCS2 show higher levels of the autophagy markers cathepsin B (active form), beclin-1 and LC3B-II (cleaved form) relative to control fibroblasts. Equal loading was assessed with β-actin.

Figure 2. Upon coculture with MCF7 cells, fibroblasts overexpressing BDH1 or HMGCS2 show Cav-1 downregulation. Fibroblasts overexpressing BDH1, HMGCS2 or empty vector control were co-cultured with MCF7 cells for 5 d. Then, cells were fixed and immunostained with anti-Cav-1 (red) and anti-K8–18 (green) antibodies. Nuclei were stained with DAPI (blue). Cav-1 staining is shown in the top panels to better appreciate that upon coculture with MCF7 cells, fibroblasts overexpressing BDH1 or HMGCS2 show downregulation of Cav-1, as compared with MCF7 cells cocultured with control fibroblasts. Original magnification, 40x.

Figure 3. Coculture with fibroblasts overexpressing BDH1 or HMGCS2 induces mitochondrial biogenesis in MCF7 cells. Fibroblasts overexpressing BDH1, HMGCS2 or empty vector control were co-cultured with MCF7 cells for 5 d. Then, cells were fixed and immunostained with anti-mitochondrial membrane (A) or HSP60 (B) antibodies. MCF7 cells were labeled using anti-K8–18 (green) antibodies. (A) Mitochondrial membrane staining (red) is shown in the top panels to better illustrate that fibroblasts overexpressing BDH1 or HMGCS2 induce mitochondrial biogenesis in MCF7 cells. (B) HSP60 staining (red) is shown in the top panels to better appreciate that fibroblasts overexpressing BDH1 or HMGCS2 promote the expression of the mitochondrial chaperon HSP60 in MCF7 cells. Original magnification, 63x.

Figure 4. Coculture with fibroblasts overexpressing BDH1 or HMGCS2 increases the MCF7 cell population. GFP (+) MCF7 cells were grown in co-culture with fibroblasts harboring BDH1, HMGCS2 or the empty vector control for 5 d. Then, cells were trypsinized and analyzed by flow cytometry using a 488 nm laser, to identify the GFP (+) MCF7 cell population. Fibroblasts were identified as the GFP (-) cell population. (A) Traces of the relative percentage of MCF7-GFP (+) cells and GFP (-) fibroblasts. (B) Histogram representation of the relative percentage of MCF7 cells and fibroblasts. The presence of fibroblasts overexpressing BDH1 and HMGCS2 increases the proportion of MCF7 cells by > 2-fold. *p < 0.0002.

Figure 5. Fibroblasts overexpressing HMGCS2 promote tumor growth of MDA-MB-231 breast cancer cells, without an increased vascularity. (A) To evaluate the tumor-promoting properties of HMGCS2 fibroblasts, we used a xenograft model. HMGCS2 or LV-105 control fibroblasts were co-injected with MDA-MB-231 breast cancer cells into the flanks of athymic nude mice. After 4.5 weeks tumors were analyzed. Note that tumor weight and volume are increased by 2.8-fold in tumors derived from HMGCS2 overexpressing fibroblasts, compared with the LV-105 empty vector control. p-values are as indicated. n = 10. (B) Tumor angiogenesis. To evaluate if HMGCS2 fibroblasts promote tumor growth by increasing tumor angiogenesis, frozen tumor sections were analyzed by CD31 immunostaining. Quantification of the number of CD31 (+) vessels per field indicates that vessel density is not increased in tumors overexpressing HMGCS2, relative to control tumors.

Figure 6. Generation of MDA-MB-231 cells overexpressing enzymes for ketone utilization. MDA-MB-231 breast cancer cells overexpressing ACAT1, ACAT2, OXCT1, OXCT2 or LV-105 empty vector control were generated using a lentiviral approach. Western blot analysis was performed with isoform-specific antibodies to confirm the overexpression of ACAT1 (A), ACAT2 (B), OXCT1 (C), OXCT2 (D), relative to empty vector control cells. Equal loading was assessed with β-actin.

Figure 7. Overexpression of ACAT1/2 and OXCT1/2 in breast cancer cells promotes tumor growth. (A) Tumor growth. A xenograft model employing MDA-MB-231 breast cancer cells overexpressing an empty vector (Lv-105), ACAT1/2 or OXCT1/2 were injected into the flanks of athymic nude mice. Tumor weights and volumes were measured 4 weeks post-injection. Note that MDA-MB-231 cells overexpressing ACAT1/2 and OXCT1/2 greatly promote tumor growth, resulting in a 2-fold increase in tumor weight and a 3-fold increase in tumor volume, respectively, relative to control cells. (B) Tumor angiogenesis. Tumor frozen sections were cut and immunostained with anti-CD31 antibodies and vascular density (number of vessels per field) was quantified. Note that ACAT1/2 overexpression leads to a 20% increase in vessel density, while no significant change was seen in xenografts with MDA-MB-231 overexpressing OXCT1/2.

Figure 8. Overexpression of ACAT1/2 in breast cancer cells promotes lung metastasis. We used a lung metastasis assay model of MDA-MB-231 cells overexpressing empty vector (Lv-105), ACAT1/2 or OXCT1/2 injected into the tail vein of athymic nude mice. After 8 weeks, lungs were insuflated and lung metastases were scored. Note that MDA-MB-231 cells overexpressing ACAT1/2 greatly promote lung metastasis as compared with control cells and OXCT1/2-overexpressing cells.

Figure 9. Ketones and two-compartment tumor metabolism. Cancer-associated fibroblasts are catabolic and undergo autophagy/mitophagy. This results in ketone body production, via HGMCS2 and BDH isotorms, driving the production of 3-hydroxy-butyrate in the tumor stroma. These ketone bodies are then transferred from stromal fibroblasts to adjacent cancer cells, via mono-carboxylate transporters (MCT1/4). Then, in cancer cells, stromally derived ketone bodies are converted to acetyl-CoA, using the metabolic enzymes OXCT1/2 and ACAT1/2. After conversion to acetyl-CoA, energy is generated via the TCA cycle and oxidative mitochondrial metabolism (OXPHOS).