Correlations of Gene Expression with Blood Lead Levels in Children with Autism Compared to Typically Developing Controls

Yingfang Tian, Peter G. Green, Boryana Stamova, Irva Hertz-Picciotto, Isaac N. Pessah, Robin Hansen, Xiaowei Yang, Jeffrey P. Gregg, Paul Ashwood, Glen Jickling, Judy Van de Water, Frank R. Sharp

The objective of this study was to examine the correlation between gene expression and lead (Pb) levels in blood in children with autism (AU, n = 37) compared to typically developing controls (TD, n = 15). We postulated that, though lead levels did not differ between the groups, AU children might metabolize lead differently compared to TD children. RNA was isolated from blood and processed on Affymetrix microarrays. Separate analyses of covariance (ANCOVA) corrected for age and gender were performed for TD, AU, and all subjects (AU + TD). To reduce false positives, only genes that overlapped these three ANCOVAs were considered. Thus, 48 probe sets correlated with lead levels in both AU and TD subjects and were significantly different between the groups (p(Diagnosis x log₂Pb) < 0.05). These genes were related mainly to immune and inflammatory processes, including MHC Class II family members and CD74. A large number (n = 791) of probe sets correlated (P ≤ 0.05) with lead levels in TD but not in AU subjects; and many probe sets (n = 162) correlated (P ≤ 0.05) with lead levels in AU but not in TD subjects. Only 30 probe sets correlated (P ≤ 0.05) with lead levels in a similar manner in the AU and TD groups. These data show that AU and TD children display different associations between transcript levels and low levels of lead. We postulate that this may relate to the underlying genetic differences between the two groups, though other explanations cannot be excluded.