Metabolic Disruption Early in Life is Associated With Latent Carcinogenic Activity of Dichloroacetic Acid in Mice

Abstract

Early-life environmental factors can influence later-life susceptibility to cancer. Recent evidence suggests that metabolic pathways may mediate this type of latency effect. Previously, we reported that short-term exposure to dichloroacetic acid (DCA) increased liver cancer in mice 84 weeks after exposure was stopped. Here, we evaluated time course dynamics for key events related to this effect. This study followed a stop-exposure design in which 28-day-old male B6C3F1 mice were given the following treatments in drinking water for up to 93 weeks: deionized water (dH2O, control); 3.5 g/l DCA continuously; or 3.5 g/l DCA for 4-52 weeks followed by dH2O. Effects were evaluated at eight interim time points. A short-term biomarker study was used to evaluate DCA effects at 6, 15, and 30 days. Liver tumor incidence was higher in all DCA treatment groups, including carcinomas in 82% of mice previously treated with DCA for only 4 weeks. Direct effects of DCA in the short-term study included decreased liver cell proliferation and marked mRNA changes related to mitochondrial dysfunction and altered cell metabolism. However, all observed short-term effects of DCA were ultimately reversible, and prior DCA treatment did not affect liver cell proliferation, apoptosis, necrosis, or DNA sequence variants with age. Key intermediate events resulting from transient DCA exposure do not fit classical cytotoxic, mitogenic, or genotoxic modes of action for carcinogenesis, suggesting a distinct mechanism associated with early-life metabolic disruption.

Keywords: carcinogenesis; dichloroacetic acid; early-life exposure; liver; metabolism.

Figure 1.

Experimental design for the chronic stop-exposure (A) and short-term continuous exposure (B) studies of dichloroacetic acid (DCA) in male B6C3F1 mice. A, Treatments included deionized water (dH₂O, control); continuous DCA at 3.5 g/l in the drinking water; or 3.5 g/l DCA for 4, 10, 26, or 52 weeks followed by control dH₂O for up to 93 weeks. Broken arrows indicate DCA administration, solid arrows indicate dH2O, and asterisks (*) indicate time points of sampling for histopathology. Groups are designated by DCA and dH2O treatment times separated by a vertical bar. B, Treatments included dH₂O control or 0.5, 1.0, 2.0, and 3.5 g/l DCA in the drinking water for up to 30 days. Broken arrows indicate DCA administration, solid arrows indicate dH₂O, and asterisks (*) indicate interim and final time points. Mice were 4 weeks of age at the start of both experiments.

Figure 2.

Time course effects of early-life DCA treatment on liver pathology. A, Incidence of hepatocellular (HC) adenoma or carcinoma over time. Asterisks (*) indicate a significantly greater incidence at P < 0.05 compared to concurrent dH₂O control group. B, Relative liver weight in dH₂O (white bars) and 10-week DCA (dark bars) groups over time. Arrow indicates stoppage of DCA treatment. Asterisks (*) indicate adjusted P < 0.05 compared to concurrent dH₂O control group. C, Images showing marked HC hypertrophy with cytoplasmic vacuolation (left) and necrosis (middle) after 10 weeks of DCA treatment compared to normal liver at 31 weeks from the same 10-week DCA group (right), showing complete reversal of morphologic changes. H&E stain, 40x objective magnification. D, E, Persistence of HC hypertrophy (D) and necrosis (E) following DCA exposure. Solid symbols indicate current DCA treatment. Asterisk (*) indicates adjusted P < 0.05 in incidence compared to concurrent dH₂O control group.

Figure 3.

Time course effects of DCA on liver cell proliferation in chronic and short-term studies. A, Cumulative incidence of proliferative (pre-neoplastic) hepatocellular foci in the chronic study at ≥31 weeks. Basophilic, eosinophilic, and clear cell indicate subtypes of foci; clear cell foci include “vacuolated” foci. Incidence data are provided in Supplementary Table S7. Note that the higher cumulative incidence in the 4-week DCA treatment group may be influenced by the lack of interim time points. B, Cell proliferation labeling index (LI) (%- positive cells) for current and prior DCA treatment in morphologically normal liver. C, Short-term dose effects of direct DCA exposure on liver cell proliferation LI determined by BrdU labeling index (%-positive cells). D, Representative images of BrdU labeling in liver sections from normal control and 3.5 g/l DCA groups at 6 days. Background staining with hematoxylin, 20x objective magnification. Full cell proliferation LI data sets are provided in Supplementary Tables S8 and S9. *adjusted P < 0.05 compared to Dh₂O Con group; ^ adjusted P < 0.05 for dose trend.

Figure 4.

Genomic effects of DCA following direct and prior exposure. A–C, RNA-seq profiles following treatment with 0, 1, 2 and 3.5 g/l DCA for 6 days. A, Principal component analysis of RPKM-normalized gene counts. Note clear separation of dose groups. B, Venn diagram showing overlap of significant DEGs for different DCA treatment groups with an FDR adjusted P value ≤ 0.05. Full DEG lists and RPKM-normalized counts are provided in Supplementary Table S11. C, Top canonical pathways in IPA shared across DCA treatment groups, ranked by -log(P value). A full list of canonical pathways is provided in Supplementary Table S12. D, E, Relative expression of select mRNA targets measured by RT qPCR. D, Time course effects following 10 weeks of DCA exposure. Bold labels on x-axis indicate week of study. E, Carry-over effects of prior DCA treatment at 78 weeks. Bold labels on x-axis indicate weeks of prior DCA exposure. Full RT-qPCR data sets are provided in Supplementary Table S15.

Figure 5.

Effects of current and prior treatment with DCA on DNA sequence variants in the liver. A, Frequency for each variant call (VC) at each locus as determined by DNA amplicon resequencing. Open circles represent mean frequency for each locus. B, Frequency of original quality-filtered VCs. Colored circles represent distinct VCs collated by sample. Open symbols represent mean variant frequency for each technical replicate (open diamonds and squares). C, Total number of variant calls in each sample (by group) presented as total (raw output, white bars), filtered for loci with parental heterozygosity (grey bars), and filtered for loci with parental heterozygosity and repeat breaks (black bars). D, Average number of variant calls (VCs) in dH2O control, prior DCA (10-week), and direct DCA groups at 26-week and 78-week time points. Left panel represents average VCs filtered by heterozygotes; right panel represents average VCs filtered by heterozygotes and breaks. Vertical bars indicate standard error of the mean. More detailed information is provided in Supplementary Table S16.