Metabolic alterations and mitochondrial bioenergetic profile in HDAC4-driven tumorigenesis

Many tumors display a high rate of glycolysis and lactate secretion even in presence of ample oxygen concentrations and a reduced rate of aerobic respiration - the so called Warburg effect. Nevertheless, most tumor mitochondria are not defective in their ability to carry out oxidative phosphorylation. Previous experiments demonstrated that NIH-3T3 murine fibroblasts expressing the TM mutant of class IIa HDAC4, which has an almost total nuclear subcellular localization and therefore acts as a "super-repressive" form of HDAC4, are capable, unlike the WT counterpart, to grow in anchorage-independent way in-vitro and to form tumors in mice. The aim of this thesis is to characterize the tumoral metabolic phenotype of HDAC4TM-expressing cells. Although HDAC4TM-expressing cells had a proliferative advantage with respect to control cells, they did not lower the pH of the medium. In accordance, HDAC4TM cells lactate secretion, taken as a marker of glycolytic metabolism, was comparable with their counterpart cells and the inhibition of LDHa was able to discriminate between HDAC4TM- and H-RASG12V-expressing tumorigenic control cells which displayed a classic Warburg phenotype. However, HDAC4TM-expressing cells showed higher glucose shortage sensibility and a lower 2-DG IC50 compared to control cells, indicating a glycolysis dependance of HDAC4TM-expressing cells in between that of H-RASG12V-expressing cells and non-tumorigenic cells. In contrast to H-RASG12V cells, short-term blockade of glycolysis in HDAC4TM-expressing cells affected ATP production in a manner undistinguishable from control cells, while long-term glycolysis inhibition differentially impacted H-RASG12V and HDAC4TM tumorigenic potential. High resolution respirometry showed, in sharp contrast with H-RASG12V cells, a slight increase in basal mitochondrial oxygen consumption rate of HDAC4TM cells with respect to control cells, without affecting mitochondrial mass or membrane potential, that can be in part accounted for a slightly higher ETS complex I activity. OXPHOS inhibition did not cause a dramatic drop in intracellular ATP levels even though was sufficient to abolish HDAC4TM proliferative advantage with respect to cotrol cells. Surprisingly, however, OXPHOS inhibition revealed a less pronounced effect on HDAC4TM than on H-RASG12V tumorigenic potential. These results seem to support a model in which, despite a tumorigenic potential similar to H-RASG12V cells, the augmented glycolytic flux of HDAC4TM cells is not coupled to lactate secretion because of the preservation of mitochondrial compartment functionality but is speculatively used to increase the pool of metabolic intermediates needed for sustained cell proliferation.