Tumor cells can survive the hypoxia in the center of tumor masses by using glycolysis, rather than oxidative phosphorylation, to generate energy. Most tumor cells also stick to glycolysis when normally oxygenated. Valeria Fantin, Julie St-Pierre, and Philip Leder (Harvard Medical School, Boston, MA) now report that some tumor cells, despite their preference for glycolysis, nevertheless retain the ability to use oxidative phosphorylation.
The Boston team reached this conclusion by inhibiting lactate dehydrogenase A (LDH-A). This enzyme converts NADH and pyruvate, the products of glycolysis, into lactate and NAD+. The lactate is exported and the NAD+ used to keep glycolysis going.
Therefore, when LDH-A is shut off and oxygen is limited, NAD+ runs low and glycolysis alone cannot continue. When oxygen is around, however, the pyruvate can be converted to acetyl-CoA to feed the Krebs cycle, which in turn feeds oxidative phosphorylation to regenerate NAD+.
It was unclear whether tumor cells were capable of switching back to oxidative phosphorylation. It now seems that some are: some cells inhibited for LDH-A got their mitochondria going and grew almost as well as wild-type cells if oxygen levels were normal. The inhibited cells did poorly in low-oxygen and in tumor models, however.
A preference for glycolysis in normoxic conditions, even by cells perfectly capable of oxidative phosphorylation, is somewhat of a surprise given the greater efficiency (per glucose molecule) of oxidative phosphorylation. But in fact oxidative phosphorylation is slower than glycolysis. This difference may be critical for fast-growing tumor cells.
The bias toward glycolysis is probably driven by HIF-1α, which is induced by both hypoxia and oncogenes, and by oncogenes that increase glucose uptake. Normal cells rely far less on glycolysis and LDH-A, however, so anticancer therapies targeting LDH-A should be nontoxic. 
Reference:
