HIV-related fatigue is multi-causal in origin and potentially related to mitochondrial dysfunction caused by toxicity from nucleoside reverse transcriptase inhibitor (NRTI) antiretroviral therapy. CD14+ cells are undifferentiated macrophages, vulnerable to HIV infection, and easily accessible for gene expression experiments in a purified cell population. We utilized a novel mitochondrially-specific gene expression microarray to assess mitochondrial and nuclear genes in CD14+ cells of low- and high-fatigued, NRTI-treated HIV/AIDS patients (n=5 each). Novel Bayesian and liquid association network methods identified 33 genes predictive of low versus high fatigue and 32 genes predictive of healthy versus HIV infection. Sulfotransferase 2B1 (SULT2B1) is relevant to both the cholesterol and testosterone pathway, and like several inner mitochondrial membrane genes also identified, predictive of fatigue status, while outer mitochondrial membrane genes were predictive of HIV status. A surprising finding was that adenylate cyclase 2 (ADCY2) was a predictor of both HIV and fatigue; it had the highest Kendall’s Tau association value in the HIV group, but in reverse, the lowest Tau value in the fatigue group. Assaying CD14+ cells may provide an alternative to muscle biopsy and a minimally invasive procedure to evaluate patient mitochondrial function, and Bayesian and network tools are useful to identify the link between subjective symptom perceptions and underlying biologically pathways. Fatigue status in HIV patients treated with NRTIs was found to be linked to RNA expression differences related to mitochondrial function. Additional studies are needed to confirm the relevance of our findings in CD14+ cells in other tissues (e. g. skeletal muscle) and to understand the significance of key genes such as SULT2B and ADCY2 in fatigue and HIV disease.

The design was a comparison of HIV high fatigue to HIV low fatigue. Low fatigue was determined as 3-7 on the Revised Piper Fatigue Score and High fatigue was a score of 7 or greater. Five HIV negative control samples were used to compare normal fatigue and non-disease status.