Copyright Notice
This is a work of the US government and distributed under the terms of the Public Domain
Bookshelf ID: NBK558859
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Toxicology Program. NTP Technical Report on the Toxicology Studies of Bromodichloroacetic Acid (CASRN 71133-14-7) in F344/N Rats and B6C3F1/N Mice and Toxicology and Carcinogenesis Studies of Bromodichloroacetic Acid in F344/NTac Rats and B6C3F1/N Mice (Drinking Water Studies): Technical Report 583 [Internet]. Research Triangle Park (NC): National Toxicology Program; 2015 Oct.
NTP Technical Report on the Toxicology Studies of Bromodichloroacetic Acid (CASRN 71133-14-7) in F344/N Rats and B6C3F1/N Mice and Toxicology and Carcinogenesis Studies of Bromodichloroacetic Acid in F344/NTac Rats and B6C3F1/N Mice (Drinking Water Studies): Technical Report 583 [Internet].
Show detailsO.1. Introduction
The National Toxicology Program (NTP) bioassay of bromodichloroacetic acid in B6C3F1/N mice resulted in statistically significant increases in the incidences of hepatoblastoma in male mice and hepatocellular carcinoma in both male and female mice. Currently, very little information is available regarding bromodichloroacetic acid-induced mechanisms of toxicity and tumorigenesis. The study presented in this Appendix compares global gene profiles between laser capture microdissected hepatoblastoma, associated hepatocellular carcinoma, adjacent nontumor liver, and age-matched vehicle normal liver in order to identify global gene expression changes related to bromodichloroacetic acid exposure and mechanisms of hepatocarcinogenesis in bromodichloroacetic acid-exposed B6C3Fl/N mice.
The specific objectives of this study were to 1) identify significant differences in global gene expression between bromodichloroacetic acid-treated adjacent nontumor and vehicle control liver, 2) identify genomic alterations in bromodichloroacetic acid-treated hepatocellular carcinoma compared to its adjacent nontumor liver, and 3) identify genomic alterations in bromodichloroacetic acid-induced hepatoblastoma compared to hepatocellular carcinoma. The first objective addresses the chemical effect on the liver resulting from bromodichloroacetic acid exposure, and the second objective provides important information on the molecular characterization of hepatocellular carcinoma resulting from bromodichloroacetic acid exposure in the B6C3F1/N mouse. The third objective addresses the genomic alterations between bromodichloroacetic acid-treated hepatoblastoma compared to its associated hepatocellular carcinoma and molecular pathogenesis of mouse hepatoblastoma response in NTP rodent bioassays.
O.2. Materials and Methods
O.2.1. Animals and Tissue Sampling
Frozen tissues from spontaneous and treatment-related hepatocellular tumors were collected from the 2-year NTP bioassay of bromodichloroacetic acid and used for molecular biology analysis. The frozen sample selection was based on tumor size, i.e., when a tumor was at least 0.5 cm in diameter, one-half of that tumor was collected for fixation in 10% neutral buffered formalin, and the other corresponding half was flash frozen in liquid nitrogen. RNA isolated from frozen tissue was utilized for quantitative real-time PCR (qPCR) and microarray analysis. DNA isolated from either the frozen tissues or from formalin-fixed paraffin-embedded (FFPE) tissues was used for mutation analysis. In the current study, hepatoblastoma, hepatocellular carcinoma, and adjacent nontumor liver were sampled from bromodichloroacetic acid-treated B6C3F1/N male mice and normal liver from two male and four female age-matched vehicle control mice were collected at terminal sacrifice of the 2-year NTP bioassay. Histopathology was conducted on all the samples used in the molecular analysis. The samples from bromodichloroacetic acid-exposed and vehicle control B6C3F1/N mice are listed in Table O-1.
O.2.2. Laser Capture Microdissection, RNA Isolation, and Amplification
All the tissue samples considered for laser capture microdissection (LCM) were microscopically examined for the lack of autolysis, necrosis, and hemorrhage and the presence of adequate adjacent nontumor liver. Following histopathology, identification of samples that had a good representation of hepatoblastoma, hepatocellular carcinoma, and adjacent nontumor liver, frozen samples were obtained from the NTP frozen tissue repository, removed from cryovials, and embedded in OCT medium on dry ice. Following embedding, cryosections were prepared, stained with hematoxylin and eosin (H&E), and evaluated for adequate amounts of each tissue component (e.g., hepatoblastoma, hepatocellular carcinoma, and adjacent nontumor liver). Next, hepatoblastomas (N = 6), hepatocellular carcinomas (N = 6), adjacent nontumor liver (N = 6), and age-matched vehicle normal liver (N = 6) were LCMed from two to five serial 10 mm cryosections for microarray analysis (Table O-1). LCM was performed on the above tissues using MMI CellCut Plus (MMI, Haslett, MI). RNA extraction and whole genome amplification was performed using Ovation Pico WTA System V2 (NuGEN, San Carlos, CA) following manufacturer’s recommendation and RNA integrity was measured with Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA).
O.2.3. Microarray Hybridizations
Gene expression analysis was conducted on bromodichloroacetic acid-treated hepatoblastoma (N = 6), bromodichloroacetic acid-treated hepatocellular carcinoma (N = 6), adjacent nontumor liver (N = 6), and age-matched vehicle normal liver (N = 6) using Affymetrix Mouse Genome 430 2.0 GeneChip® arrays (Affymetrix, Santa Clara, CA). Total RNA (10 ng) was amplified as directed in the WT-Ovation Pico RNA Amplification System protocol and labeling with biotin following the Encore Biotin Module. Amplified biotin-aRNAs (5 µg) were fragmented and hybridized to each array for 18 hours at 45°C in a rotating hybridization. Array slides were stained with streptavidin/phycoerythrin utilizing a double-antibody staining procedure and then washed for antibody amplification following manufacturer’s protocol. Arrays were scanned in an Affymetrix Scanner 3000 and data was obtained using the GeneChip® Command Console Software (AGCC; Version 1.1) using the MAS5 algorithm to generate .CHP files.
O.2.4. Data Processing and Identification of Differentially Expressed Genes
Data processing and identification of differentially expressed genes was done as previously described in detail.168,169 Briefly, array fluorescent pixel intensity measurements were acquired, and gene expression data were normalized across all samples using the robust multiarray analysis (RMA) methodology.170 Using RMA-normalized data, for each probe set, pairwise comparisons were made using a bootstrap t-test while controlling the mixed directional false discovery rate (FDR).151 This methodology controls for the overall FDRs for multiple comparisons as well as directional errors when declaring a gene to be upregulated or downregulated. The level of significance was set at P < 0.05 and the mdFDR was set at 5%.
O.2.5. Bioinformatics Analyses
Partek Genomics Suite, version 6.6 (Partek Inc., St. Louis, MO), was used to perform PCA on the normalized data of hepatoblastoma, hepatocellular carcinoma, adjacent nontumor liver, and vehicle normal liver samples. PCA uses a linear transformation to reduce the dimension of the data from n probe sets to k principal components (PC). The first three PC that capture the majority of the variation in the data were used to visualize the spatial relationship of the hepatoblastoma, hepatocellular carcinoma, adjacent nontumor liver, and vehicle normal liver samples.
Ingenuity Pathway Analysis (IPA) 9.0, application build-220217, version-16542223 (Ingenuity systems Inc., Redwood City, CA), was used to evaluate the most statistically significant overrepresented canonical pathways. These canonical pathways are based on the Ingenuity knowledge base. The significant biological canonical pathways were derived from IPA and the statistical significance was assumed at a P < 0.001 (Fisher’s exact test).
O.2.6. Quantitative Real-Time PCR
RNA extraction and amplification were performed using Ovation Pico WTA System V2 (NuGEN, San Carlos, CA) following manufacturer’s recommendation. Total RNA (10 ng) was amplified using Ovation Pico WTA System and resulted in 6 µg of amplified cDNA. Relative quantitative gene expression levels were detected using real-time PCR with the ABI PRISM 7900HT Sequence Detection System (Life Technologies, Grand Island, NY) using SYBR green methodology. Primers were designed using Primer3Plus software171 to span exon-exon junctions with an annealing temperature of 60°C and amplification size of less than 150 bp.
Briefly, 25 ng of cDNA were added to a 25-µL PCR reaction to get a final concentration of 1.00 ng/µL of cDNA. Forward and reverse primer final concentrations were 100 nM in the SYBR green assay. The reactions were performed using the Power SYBR® Green PCR Master Mix (Life Technologies, Grand Island, NY). 18S was chosen as the endogenous control gene in our qPCR experiments. Relative quantification of gene expression changes was recorded after normalizing for 18S expression, computed by using the 2-∆∆CT method (user manual #2, ABI Prism 7700 SDS).
O.2.7. Comparative Mutation Analysis of H-ras and Ctnnb1 in Hepatoblastoma and Hepatocellular Carcinoma
To study the comparative mutation profile in hepatoblastoma and adjacent hepatocellular carcinoma, H-ras codon 61 and Ctnnb1 exon 2 and 3 mutation spectra were analyzed in 30 FFPE hepatoblastomas and adjacent hepatocellular carcinomas from the 2-year bromodichloroacetic acid mouse bioassay. Hepatoblastomas and adjacent hepatocellular carcinomas were LCMed as described above. DNA was isolated from LCM tissues using Ovation Pico WTA System V2 (NuGEN, San Carlos, CA) following the manufacturer’s recommendation. PCR amplification reaction and sequencing was done following previously published methods.102
O.3. Results
The first objective of this study was to identify significant differences in global gene expression between bromodichloroacetic acid-treated adjacent nontumor and vehicle control liver. Principal component analysis (PCA) clearly differentiated bromodichloroacetic acid-treated hepatoblastoma, hepatocellular carcinoma, adjacent nontumor liver, and vehicle normal liver based on their respective global gene expression profiles (Figure O-1, Figure O-2, and Figure O-3). Analysis of bromodichloroacetic acid-exposed adjacent nontumor liver compared to vehicle normal liver using Affymetrix Mouse Genome 430 2.0 GeneChip® arrays identified 841 probes (mapped to 790 genes) as differentially expressed. PCA captured 40.6% of data variation in all the genes and indicated distinct separation of the bromodichloroacetic acid-treated adjacent nontumor liver from vehicle normal liver (Figure O-1). When bromodichloroacetic acid-treated hepatocellular carcinoma was compared to adjacent nontumor liver, 1,237 probes (mapped to 1,087 genes) were identified as differentially expressed. Likewise, PCA captured 38.6% of data variation in all the genes between hepatocellular carcinoma, adjacent nontumor liver, and vehicle normal liver groups and there was distinct separation of hepatocellular carcinoma samples from adjacent nontumor liver and vehicle normal liver samples (Figure O-2). Analysis of hepatoblastoma compared to adjacent nontumor liver indicated 13,889 probes (mapped to 10,346 genes) as differentially expressed. PCA captured 47.4% of data variation in all the expressed genes and indicated a tight and distinct grouping of the hepatoblastomas from the associated hepatocellular carcinoma, adjacent nontumor liver, and vehicle normal liver suggesting marked differences in global gene expression between hepatoblastoma and other sample types (Figure O-3).
IPA comparison analysis of bromodichloroacetic acid-treated adjacent nontumor liver with vehicle normal liver indicated altered oncogenic, metabolic, and hepatic function-related pathways. IPA analysis showed that top biological functions that were altered in adjacent nontumor liver compared to vehicle normal liver included molecular pathways involved in cancer, cellular function and maintenance, and cell morphology (Figure O-4). Top toxicologic functions perturbed in adjacent nontumor liver included molecular pathways associated with liver necrosis/cell death, regeneration and proliferation (Figure O-5). Finally, top canonical pathways altered in adjacent nontumor liver included metabolic pathways such as cholesterol and retinol biosynthesis, and cancer pathways such as ephrin receptor, insulin receptor, and protein kinase A signaling (data not shown). Top genes differentially expressed in bromodichloroacetic acid-treated adjacent nontumor liver were involved in cell growth, cell proliferation, neoplasia, and transcriptional regulation (Table O-2).
In bromodichloroacetic acid-treated hepatocellular carcinoma, IPA analysis indicated dysregulation of a variety of metabolic and cancer related pathways. Overrepresented pathways altered in bromodichloroacetic acid-treated hepatocellular carcinoma included biological functions such as organismal function, cell movement, cancer signaling, cellular development, and cell growth and proliferation. Top overrepresented canonical pathways in hepatocellular carcinoma included fatty acid metabolism, xenobiotic signaling, cell cycle: G2/M DNA damage checkpoint regulation, aryl hydrocarbon receptor signaling, and apoptosis signaling (data not shown). In bromodichloroacetic acid-treated hepatocellular carcinoma, many of the top dysregulated cancer pathways included upregulation of oncogenes (Gpc3, Akr1c3, Plat, Itih5, Areg, Tff3, Afp) and downregulation of tumor suppressor genes (Dct, Gas1, Prlr, Socs2, Wnt5b) (Table O-3). Directionality of fold changes of a number of relevant genes observed on microarray were validated by qPCR (Table O-4).
Hepatoblastomas are characterized by dysregulation of Wnt/Ctnnb1 targets, embryonic development, and genomic imprinting. IPA analysis was performed to identify pathways dysregulated in hepatoblastoma compared to adjacent nontumor liver. Genes involved in Wnt/Ctnnb1 pathway signaling were dysregulated in hepatoblastoma, including upregulation of various Wnt signaling genes (Wnt9a, Wnt10a, Wnt7a), a gene involved in Wnt feedback-regulation (Axin2), positive effectors (Lef1, Dvl3), and Wnt antagonists (Dkk2, Wif1; Table O5). Secondly, there was significant upregulation of a number of genes involved in genomic imprinting (Igf2, Peg1, Peg10, Bex1, Meg3, H19, Ndn), which are typically expressed in fetal liver. Finally, genes related to embryonic stem targets such as Tbx1, Sox9, Suz12, and T were upregulated, suggesting alterations in stem cell programming (Table O-5).
NextBio meta-analysis software172 was used to identify the most concordant mouse microarray datasets in the curated literature compared to our hepatoblastoma microarray dataset. NextBio metaanalysis of publicly available gene expression databases indicated that hepatoblastomas in B6C3F1/N mice are genomically very similar to early embryonic mouse liver (E10.5 to E14.5).173,174
Hepatoblastomas show overrepresentation of pathways associated with carcinogenesis and stem/progenitor cell signaling compared to hepatocellular carcinoma. A comparison analysis between hepatoblastoma and hepatocellular carcinoma relative to adjacent nontumor liver was performed using IPA in order to identify significant differential gene expression changes between hepatoblastoma and hepatocellular carcinoma. Results of this comparison analysis indicated overrepresentation of 1) biologic functions including metabolic signaling pathways [LXR/FXR activation, fatty acid B-oxidation, xenobiotic metabolism signaling, FXR/RXR activation, triacylglycerol, and cholesterol biosynthesis (Figure O-6 and Figure O-7)]; 2) canonical pathways related to carcinogenesis [cell cycle control of chromosome replication, NRF2-mediated oxidative stress response, basal cell carcinoma signaling (Figure O8)], and a number of developmental and stem cell-related regulation pathways [mouse embryonic stem cell pluripotency, DNA methylation and transcriptional repression signaling, sonic hedgehog signaling (Figure O-9)] in hepatoblastoma compared to hepatocellular carcinoma. Select genes that were differentially expressed between hepatoblastoma and hepatocellular carcinoma were validated using qPCR (Table O-4).
Hepatoblastoma and hepatocellular carcinoma from B6C3F1/N mice do not share common mutations in Ctnnb1 and H-ras. In the present study, we screened for mutations in the “hot-spot” regions in mouse Ctnnb1 (exon 2, corresponding to exon 3 in human CTNNB1) and H-ras codon 61. Bromodichloroacetic acid-treated hepatoblastoma and adjacent hepatocellular carcinoma were LCMed from FFPE B6C3F1/N mouse livers (N = 30) to analyze the comparative mutation spectra of H-ras and Ctnnb1 genes in bromodichloroacetic acid-treated hepatoblastoma and adjacent hepatocellular carcinoma. Bromodichloroacetic acid-treated hepatoblastoma and hepatocellular carcinoma showed a relatively lower incidence of H-ras mutation and a higher incidence of Ctnnb1 mutation compared to historical spontaneous hepatocellular carcinoma (Table O6). Between bromodichloroacetic acid-treated hepatoblastoma and hepatocellular carcinoma, there was not much difference in the incidence of H-ras (7% and 13%, respectively) and Ctnnb1 mutations (23% and 10%, respectively) (Table O-6). Interestingly, for hepatoblastoma and its adjacent hepatocellular carcinoma, with the exception of one sample (LM122) the mutation spectra were different and there was no mutation overlap for H-ras and Ctnnb1 genes (Table O-7).
O.4. Discussion
Global gene expression profiling was used to identify genomic alterations in bromodichloroacetic acid-treated adjacent nontumor liver, hepatocellular carcinoma, and hepatoblastoma compared to vehicle normal livers. PCA analysis done on normalized gene expression data clearly distinguished adjacent nontumor liver from vehicle normal liver indicating distinct gene expression changes between groups related to bromodichloroacetic acid exposure. These data indicate that there are distinct genomic alterations at 2 years in bromodichloroacetic acid-treated liver that can be related to exposure. For example, when bromodichloroacetic acid-treated adjacent nontumor liver was compared to vehicle normal liver, a number of biological functions involved in cancer, cellular function and maintenance, cell death, and survival were perturbed in bromodichloroacetic acid-treated livers. Similarly, toxicologic functions perturbed in adjacent nontumor liver were associated with liver necrosis/cell death, liver proliferation, and liver damage. In the present study, bromodichloroacetic acid-treated adjacent nontumor liver were associated with changes in expression of genes involved in cell growth and proliferation (Lcn2, Lepr, Gpx3, Gas1), neoplasia (Socs3, Gsk3β, Id2, Fat4, Dpt). Specifically, genes involved in the process of tumorigenesis, like Mdm2 oncogene, were overexpressed in adjacent nontumor liver, which is also overexpressed in human hepatocellular carcinoma, and the overexpression of Mdm2 can result in excessive inactivation of tumor protein p53, diminishing its tumor suppressor function.175 Similarly, Ca2+-regulated actin-binding gene gelsolin (Gsn) was also found to be downregulated in adjacent nontumor liver compared to vehicle normal livers. Dysregulation of Gsn has been reported in a number of cancer types, including human colorectal, gastric, bladder, lung, prostate, and kidney, where gelsolin was downregulated, suggesting that it might act as a tumor suppressor.176 Although these changes are representative of gene expression changes between bromodichloroacetic acid-treated adjacent nontumor liver and vehicle normal liver, the influence of the adjacent tumor microenvironment and its influence on gene expression in adjacent nontumor liver cannot be ignored. Therefore, it is possible that some of the expression changes observed may be influenced by the presence of hepatocellular carcinoma adjacent to the adjacent nontumor liver and are likely not solely due to bromodichloroacetic acid toxicologic effects on the liver. This phenomenon of ‘field cancerization’ is defined as the presence of molecular changes associated with tumorigenesis in histologically normal tissues that are in the immediate vicinity of the tumor and could help to identify the potential molecular mechanisms responsible for hepatocarcinogenesis.177
The second objective of our study was to identify genomic alterations in bromodichloroacetic acid-treated hepatocellular carcinoma compared to its adjacent nontumor liver. PCA analysis showed clear separation of bromodichloroacetic acid-treated hepatocellular carcinoma from adjacent nontumor liver and vehicle normal livers, as expected when making nontumor to tumor comparisons in gene expression. Moreover, a number of genes were altered that were either classic oncogenes involved in hepatocarcinogenesis, or alterations in genes were suggestive of loss of normal hepatic function. For example, those genes involved in fatty acid metabolism and xenobiotic metabolism that were downregulated in bromodichloroacetic acid-treated hepatocellular carcinoma suggested hepatic dysfunction or reduced metabolic capability that can be seen in hepatic tumors. In terms of dysregulated cancer pathways, well known pathways involved in mouse hepatic carcinogenesis were altered including cell cycle G2/M checkpoint regulation, aryl hydrocarbon receptor signaling, and apoptosis regulation. More importantly, upstream tumor suppressor genes such as Tp53 and Cdnk2a were downregulated in hepatocellular carcinoma, while oncogenes such as Myc transcriptional regulators such as Hif1a, and growth factors like Vegfa, Fgf2, Tgfβ1 were upregulated in bromodichloroacetic acid-treated hepatocellular carcinoma. Loss of expression or mutations in Tp53 gene play an important role in hepatocellular carcinoma initiation or progression, depending on the context. Loss of expression of Tp53 with other cooperating events (oxidative stress, telomere erosion, DNA-damage signaling) have a more prominent role in hepatocellular carcinoma progression by facilitating continued proliferative potential, which could also contribute to genomic instability in hepatocellular carcinoma.178 MYC oncogene is a central mediator of human hepatocarcinogenesis and in mouse model of tumorigenesis wherein MYC activation is required for maintenance and expansion of transformed cells.179 Angiogenesis is an important event during the neoplastic process and is induced by the secretion of numerous angiogenic growth factors like vascular endothelial growth factor (VEGF), basic fibroblastic growth factor (FGF2), and transforming growth factor (TGF)-β1. In human hepatocellular carcinoma, increased expression of VEGF, FGF2 and TGFβ1 has been associated with poor differentiation, portal vein invasion, and poor prognosis.180,181 Similarly, many of the classic oncogenes like Afp, Gpc3, Tff3, Areg/Aregb, Scd2 and tumor suppressors like Dct, Socs2, Gas1, which are commonly dysregulated in human hepatocellular carcinoma, were also found to be similarly differentially expressed in bromodichloroacetic acid-treated hepatocellular carcinoma.
The third objective of the current study was to evaluate global genomic profiles and mutation spectra of commonly altered oncogenes in hepatoblastoma compared to adjacent hepatocellular carcinoma in B6C3F1/N mice. Consistent with the histopathological findings, results of our global gene expression analysis show that mouse hepatoblastomas are markedly different and quite distinct from adjacent hepatocellular carcinomas and normal liver. In fact, it is quite remarkable to note that the spatial distribution of samples on the PCA plot suggests that there are more overlapping genes shared between hepatocellular carcinoma and normal liver than there are between hepatocellular carcinoma and hepatoblastoma (Figure O-3). Several biologic pathways observed as dysregulated in hepatoblastoma provide insight in terms of their origin, relationship to hepatocellular carcinoma. In mouse hepatoblastoma, there was overrepresentation of oncogenic signaling pathways including Wnt/Ctnnb1 target genes, hepatic metabolism targets, stem/pluripotent progenitor cell genes, and stem cell-related target genes (Table O-5). It is known that these pathways play a role in murine and human hepatocellular carcinoma,182-184 but surprisingly, these genes were not differentially expressed in adjacent hepatocellular carcinoma (Table O-5).
The Wnt/Ctnnb1 pathway is a master regulator of cell fate and proliferation during embryonic development, and it is essential for stem cell maintenance in a wide variety of tissues.182,185 In this study, we observed upregulation of Wnt-pathway related genes (Wnt9a, Wnt10a, Wnt7a), cancer target genes (cMyc, Ccnd1), genes involved in Wnt feedback-regulation (Axin2, Nkd1), positive effectors (Lef1, Dvl3), and Wnt antagonists (Dkk2, Dkk3, Wif1), and many of these genes expressed at high levels in hepatoblastoma tumors were targets of Lef1, a known Wnt/Ctnnb1 pathway transcription factor. Dysregulation of the Wnt/Ctnnb1/Lef1 pathway has also been observed in human cancer, and plays an important role in cancer stem cell biology,186 including maintenance of self-renewal and differentiation of cancer stem cells.187 Upregulation of downstream oncogene targets of Wnt signaling as seen in this study (cMyc, Ccnd1) have also been associated with human hepatoblastoma.183,188
Metabolic capacity and function of the embryonic liver is significantly decreased compared to the mature liver. A number of genes associated with normal hepatic metabolic function were significantly downregulated in hepatoblastoma compared to adjacent hepatocellular carcinoma, supporting this embryonal phenotype in hepatoblastoma. For example, there was downregulation of number of Nrf2 target genes coding for anti-oxidant enzymes, (Sod, Cat, Ho1, Gsr, Txn, Prdx1, Ftl), phase I and II xenobiotic enzymes (Fmo1, Ugt, Gst, Ephx1, Gclc, Nqo), xenobiotic transporter genes (Mrp2, Srb1), and chaperone and stress response (Hsp22, Hsp40, Hsp90, Clpp, Fkbp5, Herpud1) compared to hepatocellular carcinoma. Several other genes involved in LXR/RXR activation, LXR/FXR activation, fatty acid β-oxidation, FXR/RXR activation, triacylglycerol and cholesterol biosynthesis, and xenobiotic metabolism (Cyp2e1, Cyp2c8, Cyp2b6, Cyp1a2) were downregulated in mice hepatoblastoma compared to hepatocellular carcinoma, suggesting reduced metabolic function. Previous studies in human hepatoblastoma have documented reduced expression of CYP genes including CYP2C8, CYP3A4, CYP2C9.189 Downregulation or loss of function of these metabolic enzymes may suggest a metabolically inactive, embryonal origin of hepatoblastoma compared to hepatocellular carcinoma. Further, meta-analysis examination of our mouse hepatoblastoma dataset supported the hypothesized stem/multipotent nature of these tumors. Comparison of the current mouse hepatoblastoma dataset with curated datasets using NextBio meta-analysis software indicated significant concordance with early embryonic mouse liver (ED10.5 to 14.5).
Following the somatic mutation theory, carcinogenesis occurs in a step-wise manner through the accumulation of various mutations or deletions in oncogenes and tumor-suppressor genes, respectively.190 As such, a mutation that occurs during the transformation of a precursor lesion or early tumor should remain fixed within the genome throughout the progression to later stages of malignancy or metastasis. Therefore, mutation spectra between 30 hepatoblastomas and adjacent hepatocellular carcinomas were compared for H-ras and Ctnnb1 genes to see if they share common mutation spectra. Interestingly, in 11 hepatoblastoma and adjacent hepatocellular carcinoma samples that carried H-ras or Ctnnb1 mutations, the mutation spectra of these genes were different, and these tumors did not share common mutations with the exception of one sample. Current data are consistent with previous studies191 which show that the mutational spectrum is different in hepatoblastoma than that of adjacent hepatocellular tumors, suggesting that these tumors are distinct entities. This is consistent with our hypothesis that hepatoblastoma arises from a transformed stem or multipotent progenitor cell, rather than a transformed hepatocyte. However, more studies with greater sample sizes and other alternate testing modalities like RNA-Seq and exome sequencing are required before we can make more definitive conclusions. The current study has some limitations such as small sample size, samples from a single chemical study, and lack of frozen spontaneous hepatoblastoma samples. Moreover, factors such as tumor heterogeneity may have played a role in these results, as each tumor was sampled as a whole. Sampling multiple sites within each tumor may provide more information about mutation spectra and similarities or differences between hepatoblastoma and adjacent hepatocellular carcinoma as well as the possible cell of origin of these tumors.
In conclusion, key differences in gene expression were observed between bromodichloroacetic acid-treated adjacent nontumor compared to vehicle normal liver involving genes mostly involved in metabolic pathways, cell death, cell growth and proliferation, and neoplasia, suggesting that bromodichloroacetic acid causes specific toxic and carcinogenic effects in the liver of exposed B6C3F1/N mice. Gene changes in bromodichloroacetic acid-treated adjacent nontumor liver are consistent with neoplastic signaling and may suggest microenvironmental changes preceding neoplastic transformation due to chemical treatment, or a type of ‘field cancerization’; however, given the fact that these bromodichloroacetic acid-treated nontumor tissues were isolated from sections adjacent to hepatocellular carcinoma, these gene expression changes may represent microenvironmental effects of the adjacent hepatocellular carcinoma on the histologically normal adjacent tissue. Although, it is interesting to consider that bromodichloroacetic acid exposure may be eliciting a ‘field cancerization’ situation in which chemical exposure is resulting in preneoplastic areas within regions adjacent to existing tumors,192,193 there was no evidence of dysplastic hepatic tissue in these areas, and the tumor microenvironment may be largely influencing cancer signaling in tumor-adjacent tissue.
Further, bromodichloroacetic acid-treated mouse hepatoblastomas are markedly different from adjacent hepatocellular carcinoma in terms of their morphology, global gene expression, and H-ras/Ctnnb1 mutation profiles. Mouse hepatoblastoma, similar to human hepatoblastoma, shares significant similarities in global gene expression, including dysregulation of genes involved in Wnt/β-catenin signaling, embryonic/stem cell pluripotency pathways, metabolic dysregulation, and expression of genomic imprinting genes. Furthermore, meta-analysis shows that mouse hepatoblastomas are very similar to early embryonic liver in terms of their gene expression profiles. These findings suggest that hepatoblastoma and hepatocellular carcinoma are very different entities, likely arising from the same hepatic linage, but hepatoblastoma arising as a result of transformation of a hepatic stem or multipotential progenitor cell. However, more studies are required to further understand the molecular tumorigenesis of hepatocellular carcinoma and hepatoblastoma.
Table O-1Hepatoblastomas, Hepatocellular Carcinomas, and Adjacent Nontumor Liver from Control and Treated Mice Used for Microarray Analysis in the Two-year Drinking Water Study of Bromodichloroacetic Acid
| Dose (mg/L) | Animal Number | Sex | Vehicle Normal Liver | BDCA-Treated Adjacent Nontumor | BDCA-Treated Hepatocellular Carcinoma | BDCA-Treated Hepatoblastoma |
|---|---|---|---|---|---|---|
| 0 | 446 | Female | Replicate 1 | − | − | − |
| 0 | 443 | Female | Replicate 2 | − | − | − |
| 0 | 17 | Male | Replicate 3 | − | − | − |
| 0 | 405 | Female | Replicate 4 | − | − | − |
| 0 | 424 | Female | Replicate 5 | − | − | − |
| 0 | 32 | Male | Replicate 6 | − | − | − |
| 1,000 | 343 | Male | − | Replicate 1 | Replicate 1 | Replicate 1 |
| 500 | 244 | Male | − | Replicate 2 | Replicate 2 | Replicate 2 |
| 1,000 | 366 | Male | − | Replicate 3 | Replicate 3 | Replicate 3 |
| 1,000 | 352 | Male | − | Replicate 4 | Replicate 4 | Replicate 4 |
| 1,000 | 304 | Male | − | Replicate 5 | Replicate 5 | Replicate 5 |
| 500 | 244 | Male | − | − | Replicate 6 | − |
| 500 | 636 | Male | − | Replicate 6 | − | Replicate 6 |
Table O-2Dysregulated Genes in Treated Adjacent Nontumor Livers Compared to Vehicle Normal Liver from B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
| Gene Symbol | Gene Name | Fold Change |
|---|---|---|
| Cell growth and proliferation | ||
| Lcn2 | Lipocalin 2 | 8.74 |
| Lepr | Leptin receptor | 6.32 |
| Gpx3 | Glutathione peroxidase 3 | 5.06 |
| Mt1H | Metallothionein 1H | 3.80 |
| Igfbp1 | Insulin-like growth factor binding protein 1 | 2.50 |
| Gas1 | Growth arrest-specific 1 | −3.50 |
| Neoplasia | ||
| Socs3 | Suppressor of cytokine signaling 3 | 3.60 |
| Ddit4 | DNA-damage-inducible transcript 4 | 3.30 |
| Itgb3 | Integrin, beta 3 | 2.90 |
| Gsk3ß | Glycogen synthase kinase 3 beta | 1.77 |
| Id2 | Inhibitor of DNA binding 2 | −2.40 |
| Stat1 | Signal transducer and activator of transcription 1 | −2.5 |
| Fzd8 | Frizzled family receptor 8 | −2.60 |
| Fat4 | FAT tumor suppressor homolog 4 | −3.42 |
| Clic5 | Chloride intracellular channel 5 | −4.30 |
| Dpt | Dermatopontin | −5.40 |
| Ca3 | Carbonic anhydrase III | −8.50 |
| Transcriptional regulators | ||
| Nolc1 | Nucleolar and coiled-body phosphoprotein 1 | 2.50 |
| Lrp1 | Low density lipoprotein receptor-related protein 1 | 2.50 |
| Pawr | PRKC, apoptosis, WT1, regulator | 2.0 |
| Meis1 | Meis homeobox 1 | −2.0 |
| Hlf | Hepatic leukemia factor | −2.0 |
Table O-3Top Dysregulated Genes Involved in Cancer Pathways in Treated Hepatocellular Carcinoma from B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
| Gene Symbol | Gene Name | Fold Change |
|---|---|---|
| Gpc3 | Glypican 3 | 39.4 |
| Akr1c3 | Aldo-keto reductase family 1, member C3 | 26.6 |
| Phgdh | Phosphoglycerate dehydrogenase | 23.3 |
| Plat | Plasminogen activator, tissue | 15.5 |
| Hsd3b1 | Hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 1 | 10.5 |
| Itih5 | Inter-alpha-trypsin inhibitor heavy chain family, member 5 | 10.4 |
| Areg/Aregb | Amphiregulin | 9.3 |
| Tff3 | Trefoil factor 3 (intestinal) | 8.9 |
| Scd2 | Stearoyl-Coenzyme A desaturase 2 | 8.4 |
| Igfbp3 | Insulin-like growth factor binding protein 3 | 7.9 |
| Afp | Alpha-fetoprotein | 7.8 |
| Ddr1 | Discoidin domain receptor tyrosine kinase 1 | 6.9 |
| Mep1a | Meprin A, alpha (PABA peptide hydrolase) | 6.1 |
| Dct | Dopachrome tautomerase | −15.2 |
| Cyp3a5 | Cytochrome P450, family 3A, polypeptide 5 | −10.2 |
| Cyp2c18 | Cytochrome P450, family 2C polypeptide 18 | −6.2 |
| Cyp2fF1 | Cytochrome P450, family 2F, polypeptide 1 | −7.6 |
| Prlr | Prolactin receptor | −7.8 |
| Gas1 | Growth arrest-specific 1 | −4.3 |
| Socs2 | Suppressor of cytokine signaling 2 | −2.3 |
| Mt1h | Metallothionein 1H | −2.3 |
| Wnt5b | wingless-type MMTV family, member 5B | −2.2 |
Table O-4Showing Validation of Microarray Gene Expression Changes Measured Using Real-time PCR in Treated Adjacent Nontumor Liver, Hepatocellular Carcinoma, and Hepatoblastoma from B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
| Gene Symbol | Gene Name | Microarray | qPCR |
|---|---|---|---|
| BDCA-Adjacent Nontumor Liver | |||
| Gpx3 | Glutathione peroxidase 3 | 5.0 | 27.5 |
| Cyp2b10 | Cytochrome P450, family 2, subfamily b, polypeptide 10 | −12.6 | −9.8 |
| Fat4 | FAT tumor suppressor homolog 4 | −3.5 | −13.3 |
| Dpt | Dermatopontin | −5.4 | −3.3 |
| Mt1 | Metallothionein 1 | 6.3 | 19.8 |
| Serpinb1a | Serine (or cysteine) peptidase inhibitor, clade B, member 1a | −16.2 | −24.4 |
| BDCA-Treated Hepatocellular Carcinoma | |||
| Cav1 | Caveolin 1 | 4.0 | 27.8 |
| Runx2 | Runt related transcription factor 2 | 3.4 | 8.5 |
| Prlr | Prolactin receptor | −6.1 | −5.4 |
| Gas1 | Growth arrest specific 1 | −4.3 | −7.2 |
| Pten | Phosphatase and tensin homolog | −1.5 | −1.4 |
| BDCA-Treated Hepatoblastoma | |||
| T | Brachyury | 262.0 | 6,300 |
| Bex1 | Brain expressed gene 1 | 166 | 7,300 |
| Igf2 | Insulin-like growth factor-2 | 59.7 | 712 |
| Lef1 | Lymphoid enhancer binding factor 1 | 47.4 | 455 |
| Wnt6 | Wingless-related MMTV integration site 6 | 60.6 | 1,033 |
| Wif1 | Wnt inhibitory factor 1 | 427.5 | 7,575.0 |
| Cyp2e1 | Cytochrome P450, family 2, subfamily e, poly1 | −212.6 | −654.1 |
Table O-5Ingenuity Pathway Analysis of β-catenin Target Genes and Genomic Imprinted Genes Dysregulated in Hepatoblastomas from B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
| Gene Symbol | Gene Name | Fold Change Hepatoblastoma | Fold Change Hepatocellular Carcinoma |
|---|---|---|---|
| β-Catenin/Wnt Target Genes Significantly Deregulated in Hepatoblastoma | |||
| Axin2 | Axin2 (conductin) | 20.3 | NS* |
| Dkk1 | Dickkopf homolog 1 | 33.0 | NS |
| Lef1 | Lymphoid enhancer-binding factor 1 | 43.1 | NS |
| Bmp4 | Bone morphogenetic protein 4 | 58.8 | NS |
| Dkk2 | Dickkopf 2 homolog | 8.6 | NS |
| Dvl3 | Dishevelled, dsh homolog 3 (Drosophila) | 2.0 | NS |
| Wnt 5, 6, 7, 9, 10 | Wingless-type MMTV integration site family | 10.0 to 90.0 | NS |
| Wif1 | WNT inhibitory factor 1 | 475.0 | NS |
| Stem-cell Related Targets | |||
| Tbx1 | T-box1 | 37.3 | NS |
| Sox9 | SRY (sex determining region Y)-box 9 | 3.2 | NS |
| Suz12 | Suppressor of zeste 12 homolog (Drosophila) | 3.0 | NS |
| T | T, brachyury homolog (mouse) | 268.0 | NS |
| Hepatic Targets | |||
| Hdac2 | Histone deacetylase 2 | 3.2 | NS |
| Cyp2e1 | Cytochrome P450, family 2, subfamily E, polypeptide 1 | −233.2 | NS |
| Cyp1a1 | Cytochrome P450, family 1, subfamily A, polypeptide 1 | −42.0 | NS |
| Upregulation of Genomic Imprinted Genes | |||
| Igf2 | Insulin-like growth factor 2 | 67.0 | 3.3 |
| Peg1 | Paternally expressed 1 | 5.7 | NS |
| Peg10 | Paternally expressed 10 | 3.5 | NS |
| Bex1 | Brain expressed, X-linked 1 | 166.0 | 44.0 |
| Meg3 | Maternally expressed 3 | 26.0 | NS |
| Ndn | Necdin homolog (mouse) | 6.0 | −1.23 |
| H19 | H19, imprinted maternally expressed transcript (non-protein coding) | 40.6 | NS |
- *
NS = No significant change.
Table O-6Incidence of H-ras and Ctnnb1 (β-catenin) Mutations in Treated Hepatoblastoma and Hepatocellular Carcinoma from B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
| H-ras | Ctnnb1 (β-catenin) | |||
|---|---|---|---|---|
| Hepatoblastoma | Hepatocellular Carcinoma | Hepatoblastoma | Hepatocellular Carcinoma | |
| Historical spontaneous | –a | 260/473 (55%)b | –a | 1/59 (2%)c |
| Chemical Exposed | 2/30 (7%) | 4/30 (13%) | 7/30 (23%) | 3/30 (10%) |
Table O-7Mutation Spectrum of H-ras and Ctnnb1 (β-catenin) in Treated B6C3F1/N Mouse Hepatoblastoma and Associated Hepatocellular Carcinoma from B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
| H-ras | Ctnnb1 (β-catenin) | |||
|---|---|---|---|---|
| Animal ID | Hepatoblastoma | Hepatocellular Carcinoma | Hepatoblastoma | Hepatocellular Carcinoma |
| LM101 | –a | 61 CAA → CGA | – | – |
| LM122 | – | 61 CAA → CGA | – 37 TCT → TTT | 34 GGA → GTA 37 TCT → TTT |
| HM343 | − | 61 CAA → CGA | 32 GAT → GTT | – |
| LM106 | − | 61 CAA → CTA | – | – |
| MM214 | 61 CAA → CTA | – | – | – |
| HM333 | 61 CAA → CTA | – | 52 CCT → CAT | – |
| LM129 | – | – | 32 GAT → AAT | – |
| MM236 | – | – | 32 GAT → GGT 33 TCT → TTT | 41 ACC → GCC |
| MM244 | – | – | 5 GCT → TCT | – |
| HM360 | – | – | 35 ATC → AGC | – |
| HM364 | – | – | – | 19–46 deletion |
- a
No mutations detected.
Figure O-1
Principal Component Analysis Comparing Global Gene Expression Profiles of Adjacent Nontumor Liver from Treated B6C3F1/N Mice and Vehicle Normal Liver from Control Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
Figure O-2
Principal Component Analysis Comparing Global Gene Expression Profiles of Hepatocellular Carcinomas (HCC) to Adjacent Nontumor Liver from Treated Mice and to Vehicle Normal Liver from Control B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
Figure O-3
Principal Component Analysis Showing Distinct Separation, Indicating Unique Gene Expression Profile, of Hepatoblastomas (HB) from Hepatocellular Carcinomas (HCC), Adjacent Nontumor Liver (AN) from Treated Mice, and Vehicle Normal Liver (VN) from Control B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
Figure O-4
Ingenuity Pathway Analysis™ of Differentially Expressed Genes Showing Top Biological Functions Perturbed in Adjacent Nontumor Liver from Treated B6C3F1/N Mice Compared to Vehicle Normal Liver from Control Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
These biological functions involved cancer, cellular function and maintenance, and cell death and survival.
Figure O-5
Ingenuity Pathway Analysis™ of Differently Expressed Genes Showing Top Toxicological Functions Perturbed in Adjacent Nontumor Liver from Treated B6C3F1/N Mice Compared to Vehicle Normal Liver from Control Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
These toxicological functions were associated with liver necrosis/cell death, liver proliferation, and liver damage.
Figure O-6
Ingenuity Pathway Analysis of Top Biological Pathways in Hepatoblastoma (Black Bars) Compared to Hepatocellular Carcinoma from Treated B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
The dotted line indicates the significance threshold of −log (p value).
Figure O-7
Ingenuity Pathway Analysis of Top Metabolic Pathways in Hepatoblastoma (Black Bars) Compared to Hepatocellular Carcinoma from Treated B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
The dotted line indicates the significance threshold of −log (p value).
Figure O-8
Ingenuity Pathway Analysis of Top Cancer Pathways in Hepatoblastoma (Black Bars) Compared to Hepatocellular Carcinoma from Treated B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
The dotted line indicates the significance threshold of −log (p value).
Figure O-9
Ingenuity Pathway Analysis of Top Developmental and Stem Cell Related Pathways in Hepatoblastoma (Black Bars) Compared to Hepatocellular Carcinoma from Treated B6C3F1/N Mice in the Two-year Drinking Water Study of Bromodichloroacetic Acid
The dotted line indicates the significance threshold of −log (p value).
- Transcriptomic and Mutational Analysis of Bromodichloroacetic Acid Treated Nontu...Transcriptomic and Mutational Analysis of Bromodichloroacetic Acid Treated Nontumor Liver and Hepatocellular Tumors in B6C3F1/N Mice - NTP Technical Report on the Toxicology Studies of Bromodichloroacetic Acid (CASRN 71133-14-7) in F344/N Rats and B6C3F1/N Mice and Toxicology and Carcinogenesis Studies of Bromodichloroacetic Acid in F344/NTac Rats and B6C3F1/N Mice (Drinking Water Studies)
Your browsing activity is empty.
Activity recording is turned off.
See more...