Exertional heat injury and gene expression changes: a DNA microarray analysis study

This study examined gene expression changes associated with exertional heat injury (EHI) in vivo and compared these changes to in vitro heat shock responses previously reported by our laboratory. Peripheral blood mononuclear cell (PBMC) RNA was obtained from four male Marine recruits (ages 17-19 yr) who presented with symptoms consistent with EHI, core temperatures ranging from 39.3 to 42.5degreesC, and elevations in serum enzymes such as creatine kinase. Controls were age- and gender-matched Marines from whom samples were obtained before and several days after an intense field-training exercise in the heat ("The Crucible"). Expression analysis was performed on Affymetrix arrays (containing similar to12,600 sequences) from pooled samples obtained at three times for EHI group (at presentation, 2-3 h after cooling, and 1-2 days later) and compared with control values (average signals from two chips representing pre- and post-Crucible samples). After post hoc filtering, the analysis identified 361 transcripts that had twofold or greater increases in expression at one or more of the time points assayed and 331 transcripts that had twofold or greater decreases in expression. The affected transcripts included sequences previously shown to be heat-shock responsive in PBMCs in vitro (including both heat shock proteins and non-heat shock proteins), a number of sequences whose changes in expression had not previously been noted as a result of in vitro heat shock in PBMCs (including several interferon-induced sequences), and several nonspecific stress response genes (including ubiquitin C and dual-specificity phosphatase-1). We conclude that EHI produces a broad stress response that is detectable in PBMCs and that heat stress per se can only account for some of the observed changes in transcript expression. The molecular evidence from these patients is thus consistent with the hypothesis that EHI can result from cumulative effects of multiple adverse interacting stimuli.