At day 8, there were statistically significant decreases in the ratio of villous/crypt areas at 170 and 520 mg/L SDD (Fig. 8). At day 91, the villous/crypt ratio was significantly altered at 520 mg/L (Fig. 8). Functional analyses using DAVID and IPA at day 8, revealed the enrichment of the same molecular and cellular functions between non-overlapping differentially expressed Rigosertib genes at ≤ 60 mg/L and ≥ 170 mg/L SDD (1295 and 4176 unique genes, respectively, |fold change| > 1.4, P1(t) > 0.95). Over-represented functions included

RNA processing, cell cycle, cell death, cell morphology, and cytoskeleton (data not shown). Similar functional analysis at day 91 identified a total of 3954 genes at ≤ 170 mg/L and 1110 genes expressed only at 520 mg/L SDD (|fold change| > 1.4, P1(t) > 0.95) with overlapping functions related to cell cycle, cellular function and maintenance and post-translational modifications (not shown). This is the first paper to report the genome-wide gene expression effects of Cr(VI), in the form of SDD, on the mouse small intestine and phenotypically

associate differential gene expression to complementary histopathology, biochemical analyses, and tissue dosimetry. SDD elicited dose-dependent differential gene expression in the duodenum and jejunum. Dose–response analysis indicates most changes occur between 14 mg/L SDD (76 differentially expressed genes at 91 days) and 60 mg/L SDD (1857 differentially expressed genes at 91 days), with little differential Fludarabine expression below 4 mg/L SDD. Quantitative dose–response modeling of gene expression changes indicated that responses RO4929097 manufacturer to SDD were similar in both intestinal segments at both time points. The median EC50 values at day 8 and day 91 in the duodenum and jejunum ranged from 39 to 55 mg/L SDD, whereas

the BMDL values at day 91 were 56 and 49 mg/L SDD in the duodenum and jejunum, respectively. Dose-dependent gene expression and associated functions are consistent with SDD concentrations that elicited phenotypic effects (e.g. cytoplasmic vacuolization) described in Thompson et al. (2011b). Taken together with no evidence of focal proliferation or neoplastic lesions in two 90-day drinking water studies (NTP, 2007 and Thompson et al., 2011b) despite clear signs of Cr(VI)-induced tissue injury (Fig. 8), it is highly plausible that Cr(VI)-induced tumorigenicity is the result of constant tissue damage and compensatory crypt epithelial cell proliferation. SDD-elicited intestinal differential gene expression may also be partially due to Cr(III) that is likely present at high concentrations following the bolus reduction of Cr(VI) at the high SDD concentrations. Although not as bioavailable due to passive uptake (Dayan and Paine, 2001), Cr(III) may alter carbohydrate/insulin signaling, and lipid metabolism pathways (Vincent, 2004).