Metabolomic and proteomic biomarkers for III-V semiconductors: chemical-specific porphyrinurias and proteinurias.

A pressing need exists to develop and validate molecular biomarkers to assess the early effects of chemical agents, both individually and in mixtures. This is particularly true for new and chemically intensive industries such as the semiconductor industry. Previous studies from this laboratory and others have demonstrated element-specific alterations of the heme biosynthetic pathway for the III-V semiconductors gallium arsenide (GaAs) and indium arsenide (InAs) with attendant increased urinary excretion of specific heme precursors. These data represent an example of a metabolomic biomarker to assess chemical effects early, before clinical disease develops. Previous studies have demonstrated that the intratracheal or subcutaneous administration of GaAs and InAs particles to hamsters produces the induction of the major stress protein gene families in renal proximal tubule cells. This was monitored by 35-S methionine labeling of gene products followed by two-dimensional gel electrophoresis after exposure to InAs particles. The present studies examined whether these effects were associated with the development of compound-specific proteinuria after 10 or 30 days following subcutaneous injection of GaAs or InAs particles in hamsters. The results of these studies demonstrated the development of GaAs- and InAs-specific alterations in renal tubule cell protein expression patterns that varied at 10 and 30 days. At the 30-day point, cells in hamsters that received InAs particles showed marked attenuation of protein expression, suggesting inhibition of the stress protein response. These changes were associated with GaAs and InAs proteinuria patterns as monitored by two-dimensional gel electrophoresis and silver staining. The intensity of the protein excretion patterns increased between the 10- and 30-day points and was most pronounced for animals in the 30-day InAs treatment group. No overt morphologic signs of cell death were seen in renal tubule cells of these animals. Western blot analyses of the urines with antibodies to the 32-, 70-, and 90-kDa stress protein families did not show the presence of these molecules, indicating that these proteins were not excreted in the urine samples. These data suggest that the observed proteinuria patterns were not a result of cell death and that the observed chemical-specific proteinurias were produced before marked cellular toxicity. These findings suggest a hypothesis involving GaAs and InAs interference with stress protein chaperoning of reabsorbed proteins for proteosomic degradation and the probable chaperoning of damaged intracellular proteins from renal proximal tubule cells into the urinary filtrate. Overall, the results of these studies provide further information on the nephrotoxicity of these semiconductor compounds. They also suggest the use of two-dimensional gel electrophoresis with silver staining of urinary protein patterns as a potentially useful proteomic approach to renal damage early in relation to intracellular proteotoxicity in kidney tubule cells.