Background: Long-term use of nucleoside analogue reverse transcriptase inhibitors (NRTIs) is associated with lactic acidosis, a rare but potentially fatal side effect of highly active antiretroviral therapy (HAART). Recent clinical data further suggest an association of lactic acidosis with the use of protease inhibitors (PIs) and non-nucleoside analogue reverse-transcriptase inhibitors (NNRTIs). Although HAART-associated lactic acidosis is hypothesized as a result of mtDNA depletion and respiratory chain dysfunction by NRTIs, the effects of combination therapy have not been assessed. Therefore, we hypothesize that combination therapy causes uncoupling of mitochondrial respiration from oxidative phosphorylation leading to increased glycolysis and thereby increased lactate production.

Methods: To test our hypothesis, we treated HepG2 cells derived from human hepatoma cell line with or without d4T, ddI, SQV, RTV, and NVP at physiologically relevant concentrations. Cells were treated for up to 15 days and analyzed for lactate production (Sigma kit), mitochondrial membrane potential (JC-1 assay), and mitochondrial mass (nonyl acrydine orange staining). Uncoupling of mitochondrial respiration was analyzed using semi-quantitative RT-PCR and primers specific for the UCP-2 gene.

Results: Antiretroviral combination treatment of HepG2 cells significantly increased lactate production by 145%–200%, when compared to control or single antiretroviral treatment. Elevation of lactate levels was seen as early as 9 days after treatment and was inversely correlated with mitochondrial membrane potential. However, mitochondrial mass and cardiolipin content remained unchanged. UCP2 mRNA levels were significantly increased in HepG2 cells treated with antiretrovirals when compared to controls.

Conclusions: Mitochondrial uncoupling proteins are known to alter membrane potential and oxygen consumption and are proposed to play a major role in energy expenditure. Our results indicate that ART-associated uncoupling of oxidative phosphorylation is probably involved in HAART-associated lactic acidosis and hepatotoxicity. These results indicate that not only NRTIs, but also PIs and NNRTIs, can affect mitochondrial function. We are currently investigating similar mechanisms in primary human liver cells. Understanding the mechanisms will provide safer drug design as well as assist clinicians in determining the best combination therapy with fewer adverse side effects.