Search

Supplementary Tables 1-4, Figures 1-2 from Identification of Susceptibility Loci in a Mouse Model of KRASG12D-Driven Pancreatic Cancer

Simple Summary The hormone estrogen is well known for its role in promoting breast and ovarian cancer. Estrogen has the opposite effect on the liver, which it protects from cancer. We show that this protection requires Estrogen Receptor α, but not Estrogen Receptor β, and correlates with a female pattern of liver gene expression. Female expression in the liver requires Estrogen Receptor α expressed in non-liver cells. Surgical removal of the ovaries, which protect females from liver cancer at least in part through their production of estrogen, does not affect the female-specific liver gene expression pattern. Estrogen may therefore influence liver carcinogenesis through multiple independent mechanisms. We identify six genes that are strong candidates for mediating Esr1′s protection against liver cancer. Abstract Estrogen protects females from hepatocellular carcinoma (HCC). To determine whether this protection is mediated by classic estrogen receptors, we tested HCC susceptibility in estrogen receptor-deficient mice. In contrast to a previous study, we found that diethylnitrosamine induces hepatocarcinogenesis to a significantly greater extent when females lack Esr1, which encodes Estrogen Receptor-α. Relative to wild-type littermates, Esr1 knockout females developed 9-fold more tumors. Deficiency of Esr2, which encodes Estrogen Receptor-β, did not affect liver carcinogenesis in females. Using microarrays and QPCR to examine estrogen receptor effects on hepatic gene expression patterns, we found that germline Esr1 deficiency resulted in the masculinization of gene expression in the female liver. Six of the most dysregulated genes have previously been implicated in HCC. In contrast, Esr1 deletion specifically in hepatocytes of Esr1 conditional null female mice (in which Cre was expressed from the albumin promoter) resulted in the maintenance of female-specific liver gene expression. Wild-type adult females lacking ovarian estrogen due to ovariectomy, which is known to make females susceptible to HCC, also maintained female-specific expression in the liver of females. These studies indicate that Esr1 mediates liver cancer risk, and its control of sex-specific liver gene expression involves cells other than hepatocytes.

Hepatocellular carcinoma is increasingly important in the United States as the incidence rate rose over the last 30 years. C3HeB/FeJ mice serve as a unique model to study hepatocellular carcinoma tumorigenesis because they mimic human hepatocellular carcinoma with delayed onset, male gender bias, approximately 50% incidence, and susceptibility to tumorigenesis is mediated through multiple genetic loci. Because a human O6-methylguanine-DNA methyltransferase (hMGMT) transgene reduces spontaneous tumorigenesis in this model, we hypothesized that hMGMT would also protect from methylation-induced hepatocarcinogenesis. To test this hypothesis, wild-type and hMGMT transgenic C3HeB/FeJ male mice were treated with two monofunctional alkylating agents: diethylnitrosamine (DEN; 0.025 μmol/g body weight) on day 12 of life with evaluation for glucose-6-phosphatase-deficient (G6PD) foci at 16, 24, and 32 weeks or N-methyl-N-nitrosurea (MNU; 25 mg MNU/kg body weight) once monthly for 7 months starting at 3 months of age with evaluation for liver tumors at 12 to 15 months of age. No difference in abundance or size of G6PD foci was measured with DEN treatment. In contrast, it was unexpectedly found that MNU reduces liver tumor prevalence in wild-type and hMGMT transgenic mice despite increased tumor prevalence in other tissues. hMGMT and MNU protections were additive, suggesting that MNU protects through a different mechanism, perhaps through the cytotoxic N7-alkylguanine and N3-alkyladenine lesions which have low mutagenic potential compared with O6-alkylguanine lesions. Together, these results suggest that targeting the repair of cytotoxic lesions may be a good preventative for patients at high risk of developing hepatocellular carcinoma. Cancer Prev Res; 9(3); 245–52. ©2015 AACR.

Murine hepatocarcinogenesis requires growth hormone (GH). To determine if the GH-responsive transcription factor STAT5b (signal transducer and activator of transcription 5b) is also required, we compared the hepatic gene expression profiles of global Stat5b null mice to cancer-resistant mice mutant in the GH pathway—GH-deficient little and androgen receptor-null Tfm males. We found a high degree of overlap among Tfm, little, and Stat5b null males. The liver cancer susceptibility of global Stat5b null mice was assessed on three distinct genetic backgrounds: BALB/cJ (BALB), C57BL/6J (B6), and C3H/HeJ (C3H). The effect of Stat5b on hepatocarcinogenesis depended on the genetic background. B6 Stat5b null congenic males and females developed 2.4 times as many tumors as wild-type (WT) controls (P

We set out to better understand the signal transduction pathways that mediate liver tumor promotion by 2,3,7,8-tetrachlorodibenzo-p-dioxn ("dioxin"). To this end, we first employed congenic mice homozygous for either the Ahr(b1) or Ahr(d) alleles (encoding an aryl hydrocarbon receptor (AHR) with high or low binding affinity for dioxin, respectively) and demonstrated that hepatocellular tumor promotion in response to dioxin segregated with the Ahr locus. Once we had genetic evidence for the importance of AHR signaling, we then asked if tumor promotion by dioxin was influenced by "interleukin-1 (IL-1)-like" inflammatory cytokines. The importance of this question arose from our earlier observation that aspects of the acute hepatocellular toxicity of dioxin are dependent upon IL1-like cytokine signaling. To address this issue, we employed a triple knock-out (TKO) mouse model with null alleles at the loci encoding the three relevant receptors for tumor necrosis factors α and β and IL-1α and IL-1β (i.e., null alleles at the Tnfrsf1a, Tnfrsf1b, and Il-1r1 loci). The observation that TKO mice were resistant to the tumor promoting effects of dioxin in liver suggests that inflammatory cytokines play an important step in dioxin mediated liver tumor promotion in the mouse. Collectively, these data support the idea that the mechanism of dioxin acute hepatotoxicity and its activity as a promoter in a mouse two stage liver cancer model may be similar, i.e., tumor promotion by dioxin, like acute hepatotoxicity, are mediated by the linked action of two receptor systems, the AHR and the receptors for the "IL-1-like" cytokines.

Epidemiological studies indicate that sunlight exposure and vitamin D are each associated with a lower risk of colon cancer. The few controlled supplementation trials testing vitamin D in humans reported to date show conflicting results. We have used two genetic models of familial colon cancer, the Apc(Pirc/+) (Pirc) rat and the Apc(Min/+) (Min) mouse, to investigate the effect of 25-hydroxyvitamin D(3) [25(OH)D(3)] and two analogs of vitamin D hormone on colonic tumors. Longitudinal endoscopic monitoring allowed us to test the efficacy of these compounds in preventing newly arising colonic tumors and in affecting established colonic tumors. 25(OH)D(3) and two analogs of vitamin D hormone each failed to reduce tumor multiplicities or alter the growth patterns of colonic tumors in the Pirc rat or the Min mouse.

Genetic background affects susceptibility to pancreatic ductal adenocarcinoma in the Ela-KRASG12D mouse model. In this model, KRAS oncogene expression is driven by an elastase promoter in acinar cells of the pancreas on an FVB/NTac (FVB) background [FVB-Tg(Ela-KRASG12D)] with the transgene carried on the Y chromosome. Through linkage analysis of crosses between the C57BL/6J (B6), BALB/cJ (BALB), and DBA/2J (D2) inbred strains of mice and resistant FVB-Tg(Ela-KRASG12D), we have identified six susceptibility loci that affect mean preinvasive lesion multiplicity. Markers on chromosome 2 segregated with high tumor multiplicity in all three strains; these loci were designated Prsq1-3 (pancreatic ras susceptibility quantitative trait loci 1-3; combined F2 and N2 LODW, 6.0, 4.1, and 2.7, respectively). Susceptibility loci on chromosome 4, designated Prsq4 and Prsq5, were identified in crosses between FVB transgenic mice and B6 or BALB mice (combined F2 and N2 LODW, 3.6 and 2.9, respectively). A marker on chromosome 12 segregated with tumor multiplicity in a BALB × FVB-Tg(Ela-KRASG12D) cross and was designated Prsq6 (LODW, ∼2.5). B6-Chr YFVB-Tg(Ela-KRASG12D) and BALB-Chr YFVB-Tg(Ela-KRASG12D) consomics, which carry the KRAS transgene on the FVB Y chromosome on an otherwise inbred B6 or BALB background, developed ∼4-fold (B6) and ∼10-fold (BALB) more lesions than FVB-Tg(Ela-KRASG12D) mice. By 12 months of age, 10% of BALB-Chr YFVB-Tg(Ela-KRASG12D) mice developed invasive carcinomas. Our findings provide evidence that regions of chromosomes 2, 4, and 12 influence the development and progression of pancreatic neoplasms initiated by an oncogenic allele of KRAS in mice. Cancer Res; 70(21); 8398–406. ©2010 AACR.

Sex hormones influence the susceptibility of inbred mice to liver cancer. C57BR/cdJ (BR) females are extremely susceptible to spontaneous and chemically induced liver tumors, in part due to a lack of protection against hepatocarcinogenesis normally offered by ovarian hormones. BR males are also moderately susceptible, and the susceptibility of both sexes of BR mice to liver tumors induced with N,N-diethylnitrosamine relative to the resistant C57BL/6J (B6) strain is caused by two loci designated Hcf1 and Hcf2 (hepatocarcinogenesis in females) located on chromosomes 17 and 1, respectively. The Hcf1 locus on chromosome 17 is the predominant modifier of liver cancer in BR mice. To validate the existence of this locus and investigate its potential interaction with Hcf2, congenic mice for each region were generated. Homozygosity for the B6.BR(D17Mit164-D17Mit2) region resulted in a 4-fold increase in liver tumor multiplicity in females and a 4.5-fold increase in males compared with B6 controls. A series of 16 recombinants covering the entire congenic region was developed to further narrow the area containing Hcf1. Susceptible heterozygous recombinants demonstrated a 3- to 7-fold effect in females and a 1.5- to 2-fold effect in males compared with B6 siblings. The effect in susceptible lines completely recapitulated the susceptibility of heterozygous full-length chromosome 17 congenics and furthermore narrowed the location of the Hcf1 locus to a single region of the chromosome from 30.05 to 35.83 Mb.

A cross-campus, cross-career stage and cross-disciplinary series of discussions at a large public university has produced a series of recommendations for addressing the problems confronting the biomedical research community in the US. DOI: http://dx.doi.org/10.7554/eLife.09305.001

In an effort to understand how genetics can influence individual sensitivity to environmentally induced disease, we performed a linkage analysis to identify murine loci in addition to the Ahr locus that influence the incidence of cleft palate and hydronephrosis in developing mice exposed to the pollutant 2,3,7,8-tetrachlorodibenzo-p-dioxin (dioxin). Administration of 64 microg/kg dioxin to C57BL/6J (B6) dams at embryonic day 9 (E9) led to palatal clefting and hydronephrosis in nearly 100% of embryos by E17. In contrast, similar exposure of CBA/J (CBA) dams led to cleft palate in only 8% and hydronephrosis in 69% of embryos. To determine the genetic basis for this strain-dependent sensitivity, linkage analyses on the progeny of a B6CBAF1 intercross and a CBAxB6CBAF1 backcross were performed. The incidences of cleft palate and hydronephrosis were assessed and genomic DNA from embryos was analyzed at informative simple sequence length polymorphism (SSLP) markers. One locus segregating with dioxin-induced cleft palate was identified (p0.01) and designated as chemically mediated teratogenesis number 1 (Cmt1). The Cmt1 locus is located on chromosome 3.

The C3H/HeJ (C3H) and CBA/J (CBA) mouse strains are classical mouse models of cancer susceptibility, exhibiting high risks for both spontaneous and chemically induced liver cancer. By analysis of backcrosses and intercrosses between C3H or CBA and resistant B6 mice, we have mapped a potent modifier of hepatocellular carcinoma development to distal chromosome 1, linked to the marker D1Mit33 with combined LODW scores of ∼5.9 (C3H) and 6.5 (CBA). We previously identified this region as one of two that modify susceptibility in the more distantly related C57BR/cdJ (BR) strain. Congenic B6.C3H(D1Mit5-D1Mit17) and B6.BR(D1Mit5-D1Mit17) mice developed significantly more liver tumors than B6 mice did (6- to 13-fold, P < 10−11, in males; 3- to 4-fold, P < 10−3, in females). Thus, distal chromosome 1 carries one or more genes that are sufficient to confer susceptibility to liver cancer.

The Chemical Pathology Study Section, Center for Scientific Review, NIH hosted a 1-day workshop in Ventura, California, in the winter of 2003. There were 32 invited participants including invited speakers. The workshop consisted of three sessions and focused on recent developments in our understanding of the molecular mechanisms of mutagenesis and DNA repair, namely, error-prone DNA polymerases, DNA repair models and DNA repair networks.

The circadian clock plays a significant role in many aspects of female reproductive biology, including estrous cycling, ovulation, embryonic implantation, onset of puberty, and parturition. In an effort to link cell-specific circadian clocks to their specific roles in female reproduction, we used the promoter that controls expression of Steroidogenic Factor-1 (SF1) to drive Cre-recombinase-mediated deletion of the brain muscle arnt-like 1 (Bmal1) gene, known to encode an essential component of the circadian clock (SF1-Bmal1(-/-)). The resultant SF1-Bmal1(-/-) females display embryonic implantation failure, which is rescued by progesterone supplementation, or bilateral or unilateral transplantation of wild-type ovaries into SF1-Bmal1(-/-) dams. The observation that the central clock, and many other peripheral clocks, are fully functional in this model allows the assignment of the implantation phenotype to the clock in ovarian steroidogenic cells and distinguishes it from more general circadian related systemic pathology (e.g., early onset arthropathy, premature aging, ovulation, late onset of puberty, and abnormal estrous cycle). Our ovarian transcriptome analysis reveals that deletion of ovarian Bmal1 disrupts expression of transcripts associated with the circadian machinery and also genes critical for regulation of progesterone production, such as steroidogenic acute regulatory factor (Star). Overall, these data provide a powerful model to probe the interlocking and synergistic network of the circadian clock and reproductive systems.

Both male DBA/2J and C3H/HeJ mice are highly susceptible to hepatocarcinogenesis induced by experimental treatment with N,N-diethylnitrosamine (DEN) relative to male C57BL/6J mice. While C3H/HeJ mice carry multiple sensitivity loci, designated Hcs (hepatocarcinogen sensitivity), our previous study indicated that the susceptibility of DBA/2J mice results from the combined effects of multiple sensitivity loci and two major resistance loci, Hcr-I and -2 (hepatocarcinogen resistance). We proposed that BXD-15 recombinant inbred mice, which are extremely resistant to DEN-induced hepatocarcinogenesis, may carry the Hcr loci from the parental DBA/2J mice, but few, if any, of the multiple sensitivity loci. Conversely, the extremely sensitive BXD-11 recombinant inbred mice may carry most of the multiple sensitivity loci of the DBA/2J parents, but neither of the major resistance loci. In order to confirm our genetic model for hepatocarcinogenesis in DBA/2J mice and to evaluate the phenotypic effects of the Hcr loci on the Hcs loci of C3H/HeJ mice, we characterized hepatocarcinogen sensitivities of F 1 mice generated from the crosses involving BXD-11, BXD-15, C3H/HeJ and C57BL/6J strains. When male mice were initiated with DEN at 12 days of age and liver tumors were enumerated at 32 weeks of age, (BXD-15 X BXD-11)F 1 mice had one sixth the number of liver tumors observed in (C57BL/6JxBXD-11)F 1 mice, consistent with our previous conclusion that DBA/2J mice possess hepatocarcinogen resistance genes in spite of their high susceptibility to DEN. Significantly, (C57BL/6J X C3H/HeJ)F 1 mice also had a 2.3-fold greater number of liver tumors and 5.5-fold higher total volume of initiated lesions per liver as compared with (BXD-15XC3H/HeJ)F 1 mice, indicating that the hepatocarcinogen resistance genes inherited by BXD-15 mice are capable of suppressing the Hcs phenotype. Thus the Hcr loci may influence a wide variety of hepatocarcinogen sensitivity loci and be able to act as general resistance loci for chemical hepatocarcinogenesis. Stereological analysis of initiated hepatocellular lesions with glucose 6-phosphatase deficiency revealed that the resistance genes largely influence the promotion stage of hepatocarcinogenesis.

The construction in the 1990s of high density genetic maps of the mouse and rat based on simple sequence length polymorphisms led to an explosion of activity directed toward the identification of quantitative trait loci (QTLs) that control a broad array of normal and abnormal biology. More than 3,900 mouse and nearly 1,000 rat QTLs have been mapped by linkage analysis in studies of, among others, behavior, bone morphogenesis, cardiovascular function, and metabolism, as well as diseases including arthritis, diabetes, and cancer [1], [2]. The development of cancer is a complex, multi-step process that begins with a genetic or epigenetic event in a normal cell (initiation), followed by expansion and evolution of the initiated cells during the promotion stage, and culminating with the acquisition of malignant phenotypes, including invasiveness and metastatic potential, during tumor progression [3]. That complexity has prompted many investigators, including Hunter and colleagues (whose work is presented in this issue, [4]), to pursue the identification of cancer modifier genes, QTLs that alter cancer development in rodents. Compelling motivations for this work include the expectation that the genes underlying the QTLs will provide paradigms for understanding genetic variation in human cancer risk, that identification of the relevant genes will yield insights into pathways critical for carcinogenesis, and that these genes and the pathways they represent will provide novel targets for intervention to prevent or treat cancer. Nearly 250 QTLs that modify cancer risk or pathogenesis have been mapped in mice or rats [1], [2], [5]. However, despite this wealth of genetic information, only a small handful of these QTLs have been identified at the molecular level, as specific genes or non-coding elements. The paucity of molecular identifications applies more broadly, and Flint et al. estimated that, by 2005, less than 1% of rodent QTLs had been carried to the level of the gene [6]. The slow accrual of gene identifications is a consequence of the long path from QTL to gene (Figure 1). The starting point for most cancer QTL studies is the observation of significant variation in cancer risk among inbred strains. For virtually any tissue site, large (up to 100-fold) differences in the incidence or multiplicity of spontaneous or induced tumors may be found in the literature going back to the development of inbred strains early in the last century. Linkage analysis of segregating backcrosses or intercrosses between a pair of inbred strains with divergent cancer phenotypes may lead to the identification of one or more QTLs that control, for example, tumor incidence, multiplicity, latency [7], [8], or, as in the work by Hunter and colleagues (Faraji et al., this issue [4]), metastatic potential. Other experimental designs, including crosses between congenic or chromosome substitution strains, may be used to increase the power to detect QTLs. An intrinsic limitation of this approach is that, owing to the quantitative, variable phenotype, the precision for mapping QTLs is typically low; even with large crosses and a high density of genetic markers, the resulting 1.5 LOD support interval may be 20 cM (around 40 Mb) and contain hundreds of genes. Mapping QTLs to higher resolution requires the time and resources to produce congenic lines that carry a limited interval of the high (or low) risk donor strain's genome on the genetic background of the other strain, followed by phenotypic analysis of recombinant lines derived from that congenic. This fine mapping may yield intervals of the order of one to a few megabases, with one to 40 potential candidate genes. A caveat to this approach is that genetic complexity, with multiple sub-intervals contributing to the phenotype, has been observed for cancer QTLs more often than not, expanding the hunt for the causative genes. Prioritization of candidates within the interval may be based on sequence analysis, taking advantage of the high density SNP maps available for a large number of strains or the recent whole-genome assemblies available for a handful of strains [9], [10]. Depending on knowledge of the site of action of the QTL (e.g., whether it is cell-autonomous or acts indirectly), gene expression analysis by microarray may also be used to prioritize candidates. The “gold standard” for proof that a particular candidate is the causative gene by transgenesis or allelic substitution by homologous recombination has been achieved in only a few cases, but analysis of gene knockout strains or demonstration of specific genetic or epigenetic alterations in the orthologue in human tumors has more often provided a weight of evidence in favor of a particular candidate. Figure 1 From QTL to causative gene. The fact that metastases account for most cancer-related deaths led Hunter and colleagues to pursue QTLs that controlled the risk for metastasis in a transgenic, Polyoma-middle T (PyMT) model for breast cancer in mice, largely following the path depicted in Figure 1. More than a decade ago, they demonstrated the presence of a metastasis susceptibility gene on chromosome 9 in crosses between NZB and PyMT-FVB mice [7] and validated the existence of this modifier in chromosome substitution strains [8]. Faraji et al. [4] now describe the development and analysis of congenic mouse lines carrying various segments of proximal chromosome 9 from NZB mice on an FVB genetic background, allowing them to narrow the interval for the susceptibility QTL to a 21 megabase region. They used haplotype, DNA sequence, and gene expression data to prioritize the list of candidates and describe biological studies of one of them, Cadm1, in the present paper. Cadm1, also known as Tslc1 (Tumor suppressor in lung cancer 1), is an immunoglobulin superfamily cell adhesion molecule. Based on their studies, the authors hypothesize that Cadm1 expression suppresses metastasis by sensitizing tumor cells to elimination via immune surveillance. Their studies also demonstrate the complexities of identifying QTL genes. Despite the over-expression of Cadm1 in NZB relative to FVB mice and the fact that the NZB chromosome 9 interval enhances metastasis, they found that ectopic expression of this candidate gene suppresses metastasis and that high expression of Cadm1 in tumors is associated with improved survival in women with breast cancer. Thus, one or more additional modifier genes within the interval likely lead to the phenotype of enhanced metastasis in the congenic mice and remain unidentified. The work by Hunter and colleagues represents a substantial investment in time and financial resources, much of which involves mouse breeding and maintenance. The current economic climate makes the launching of projects using similar approaches difficult. Fortunately, over the decade since the inception of the project, new rodent resources and methods have been developed making future analysis of complex traits in rodents more efficient in both time and resources. The Faraji et al. [4] study began with extensive mapping of loci associated with susceptibility to metastasis, which took many years. In the near future, similar work could be done in a single series of phenotyping experiments using new mouse resources such as the Collaborative Cross (CC) [11] or the Diversity Outbred (DO) populations of mice [12]. In addition, once a QTL is fine-mapped, finding the causative genetic element will be facilitated by open access to mouse/rat whole genome sequences for multiple strains [10]. Likewise, validation of candidate genes will no longer require time-consuming production of knockout (KO) mice from engineered embryonic stem cells but can be obtained from various repositories generated from the mouse KOMP project [13], [14]. While the KOMP resources provide a wide variety of mutants, most are on a C57BL/6 background. Often phenotypes must be evaluated on other genetic backgrounds, which requires six to ten generations of backcrossing over a period of one or more years. Alternative approaches are now available for mice and rats that use Zn finger nucleases or TALENS to knockout or replace genes [15]. These technologies are able to produce mouse or rat KO founders in less than 3 months on almost any genetic background. The future for the use of rodent models to unravel the genetic complexity of common disease is expanding due to the intriguing published studies in this area, such as that by Hunter and colleagues, coupled with emerging new powerful genomic technologies and animal resources.

Genes that affect a person's chance of developing hepatocellular carcinoma (HCC), as BRCA1 and BRCA2 affect a person's chance of developing breast or ovarian cancer, have been difficult to detect. The vast majority of liver cancers can be attributed to Hepatitis B or C virus infection, aflatoxin exposure, or alcoholic cirrhosis, alone or in combination (Montalto et al., 2002). This high background of predisposing environmental factors makes identifying less penetrant genetic contributors more difficult. Familial patterns of susceptibility to liver cancer independent of environmental factors have helped identify a few monogenic metabolic syndromes (e.g., hemochromatosis; Dragani, 2010). However, the analysis of liver tumors points to a variety of other genes that affect liver tumor development. The patterns of chromosome gain and loss in liver cancers worldwide reveal several regions that are gained or lost in up to 86% of tumors, including gains of 1q, 6p, 8q, and 20q, and losses of 1p, 4q, 6q, 8p, 13q, 16q, and 17p (Lau and Guan, 2005; Chochi et al., 2009; Zhang et al., 2010). Chromosome analyses combined with genome-wide association studies have revealed candidate HCC modifier genes for some of these regions, such as PAPSS1 on chromosome 4q and HCAP1 on chromosome 17p (Wan et al., 2004; Shih et al., 2009).

(1994). Genetic Control of Hepatocarcinogenesis In C3H Mice. Drug Metabolism Reviews: Vol. 26, No. 1-2, pp. 201-208.

Background Hub proteins are connected through binding interactions to many other proteins. Smad3, a mediator of signal transduction induced by transforming growth factor beta (TGF-β), serves as a hub protein for over 50 protein-protein interactions. Different cellular responses mediated by Smad3 are the product of cell-type and context dependent Smad3-nucleated protein complexes acting in concert. Our hypothesis is that perturbation of this spectrum of protein complexes by mutation of single protein-binding hot-spots on Smad3 will have distinct consequences on Smad3-mediated responses. Methodology/Principal Findings We mutated 28 amino acids on the surface of the Smad3 MH2 domain and identified 22 Smad3 variants with reduced binding to subsets of 17 Smad3-binding proteins including Smad4, SARA, Ski, Smurf2 and SIP1. Mutations defective in binding to Smad4, e.g., D408H, or defective in nucleocytoplasmic shuttling, e.g., W406A, were compromised in modulating the expression levels of a Smad3-dependent reporter gene or six endogenous Smad3-responsive genes: Mmp9, IL11, Tnfaip6, Fermt1, Olfm2 and Wnt11. However, the Smad3 mutants Y226A, Y297A, W326A, K341A, and E267A had distinct differences on TGF-β signaling. For example, K341A and Y226A both reduced the Smad3-mediated activation of the reporter gene by ∼50% but K341A only reduced the TGF-β inducibilty of Olfm2 in contrast to Y226A which reduced the TGF-β inducibility of all six endogenous genes as severely as the W406A mutation. E267A had increased protein binding but reduced TGF-β inducibility because it caused higher basal levels of expression. Y297A had increased TGF-β inducibility because it caused lower Smad3-induced basal levels of gene expression. Conclusions/Significance Mutations in protein binding hot-spots on Smad3 reduced the binding to different subsets of interacting proteins and caused a range of quantitative changes in the expression of genes induced by Smad3. This approach should be useful for unraveling which Smad3 protein complexes are critical for specific biological responses.

Using a mouse predisposed to neoplasia by a germ line mutation in Apc (Apc(Min)), we tested whether induced hyperplasia is sufficient to increase intestinal tumor multiplicity or size in the intestine. We found that hyperplasia in the jejunum correlated with a significant increase in tumor multiplicity. However, tumor multiplicity was unchanged in the hyperplastic colon. This result indicates that even an intestine predisposed to neoplasia can, in certain regions including the colon, accommodate net increased cell growth without developing more neoplasms. Where hyperplasia correlated with increased tumor multiplicity, it did not increase the size or net growth of established tumors. This result suggests that the event linking hyperplasia and neoplasia in the jejunum is tumor establishment. Two novel observations arose in our study: the multiple intestinal neoplasia (Min) mutation partially suppressed both mitosis and transforming growth factor alpha-induced hyperplasia throughout the intestine; and zinc treatment alone increased tumor multiplicity in the duodenum of Min mice.

We have shown previously that the difference between C57BL/6J and C3H/HeJ male mice in their susceptibilities to chemically-induced liver tumors results predominantly from an allelic difference at the Hepatocarcinogen sensitivity (Hcs) locus. This locus modulates the rate of growth of preneoplastic liver lesions and may also play a role in the turnover of normal hepatocytes in the adult liver. To define further the growth regulatory pathway of which the Hcs gene is a component, we asked whether the expression of the Hcs gene would modulate the response of preneoplastic liver lesions to the physiologic growth stimulus generated by a two-thirds hepatectomy. Twelve-day-old male and female C57BL/6J and C3H/HeJ mice were injected with 0.5 mumols N-ethyl-N-nitrosourea/g body weight. At six weeks of age half the animals received a two-thirds hepatectomy. Groups of animals were killed between 14 and 44 weeks of age for analysis of glucose-6-phosphatase (G6Pase)-deficient foci and hepatic tumors. The partial hepatectomy induced a regenerative response that caused both the G6Pase-deficient foci and the surrounding, histochemically normal hepatocytes to undergo several rapid rounds of division. As a result, the G6Pase-deficient foci were larger in the hepatectomized animals than in the sham operated controls. The foci in the non-hepatectomized C57BL/6J male mice grew more slowly than in the C3H/HeJ male mice [volume doubling time (Td) = 2.9 +/- 0.1 weeks and 2.0 +/- 0.6 weeks, respectively]. Following partial hepatectomy, the G6Pase-deficient foci in the C57BL/6J male mice maintained a significantly higher growth rate (Td = 2.2 +/- 0.3 weeks) than the foci in the sham operated C57BL/6J male mice. The partial hepatectomy did not have any long term effect on the growth rate of the G6Pase-deficient foci in the C3H/HeJ male mice nor in female mice of either strain. At 32 weeks of age, the mean liver tumor multiplicity for hepatectomized C57BL/6J male mice was approximately 5.3-fold greater than that for sham operated animals. In contrast, a two-thirds hepatectomy resulted in a 60% reduction in the number of liver tumors in C3H/HeJ male mice relative to sham operated mice. These data demonstrate that partial hepatectomy can act as a promoter of hepatocarcinogenesis in C57BL/6J male mice but not C3H/HeJ male mice. We propose that the Hcs gene and partial hepatectomy may promote hepatocarcinogenesis through the same pathway of growth regulation.

Inbred strains of mice vary in their frequency of liver tumors initiated by a mutation in the Hras1 (H-ras) proto-oncogene. We sequenced 4.5 kb of the Hras1 gene on distal Chr 7 in a diverse set of 12 commonly used laboratory inbred strains of mice and detected no sequence variation to account for strain-specific differences in Hras1 mutation prevalence. Furthermore, the Hras1 sequence is essentially monoallelic for an ancestral gene derived from the M. m. domesticus species. To determine if the monoallelism and associated low rate of polymorphism are unique to Hras1 or representative of the general chromosomal locale, we extended the sequence analysis to 12 genes in the final 8 Mb of distal Chr 7. A region of at least 2.5 Mb that encompasses several genes, including Hras1 and the H19/Igf2 loci, demonstrates virtually no sequence variation. The 12 inbred strains share one dominant haplotype derived from the M. m. domesticus allele. Chromosomal regions flanking the monoallelic segment exhibit a significantly higher rate of variation and multiple haplotypes, a majority of which are attributed to M. m. domesticus or M. m. musculus ancestry. Electronic supplementary material The online version of this article (doi:10.1007/s00335-007-9061-1) contains supplementary material, which is available to authorized users.

The aryl hydrocarbon receptor (AHR) is a vertebrate protein that mediates the toxic and adaptive responses to dioxins and related environmental pollutants. In an effort to better understand the details of this signal transduction pathway, we employed the yeast S. cerevisiae as a model system. Through the use of arrayed yeast strains harboring ordered deletions of open reading frames, we determined that 54 out of the 4,507 yeast genes examined significantly influence AHR signal transduction. In an effort to describe the relationship between these modifying genes, we constructed a network map based upon their known protein and genetic interactions. Monte Carlo simulations demonstrated that this network represented a description of AHR signaling that was distinct from those generated by random chance. The network map was then explored with a number of computational and experimental annotations. These analyses revealed that the AHR signaling pathway is defined by at least five distinct signaling steps that are regulated by functional modules of interacting modifiers. These modules can be described as mediating receptor folding, nuclear translocation, transcriptional activation, receptor level, and a previously undescribed nuclear step related to the receptor's Per–Arnt–Sim domain., 4500 yeast deletion mutants, transformed with a reporter construct that is sensitive to dioxin, were used to create a protein-interactive map that outlines five distinct steps required for dioxin-mediated cell signaling

Female C57BR/cdJ (BR) mice are unusually susceptible to spontaneous and chemically induced hepatocarcinogenesis relative to females of other inbred strains, in part because they are insensitive to the inhibitory effects of ovarian hormones on liver tumor development; BR males are intermediate among strains in their sensitivity. C57BL/6J (B6) male and female mice are relatively resistant among inbred strains. Linkage analysis of crosses between BR and resistant B6 mice identified two loci, on Chromosomes 17 and 1, that accounted for the high susceptibility of BR mice to hepatocarcinogenesis. To determine whether the increased susceptibility of BR relative to B6 mice is intrinsic to the target hepatocytes or is the result of local or systemic differences in milieu, we determined the strain of origin of tumors that arose in BR--B6 aggregation chimeras. Chimeras were treated at 12 days of age with N,N-diethylnitrosamine, and individual tumors were dissected from 15 males at 32 weeks and from 7 females at 50 weeks of age. DNA was prepared from each tumor, and quantitative PCR assays were used to determine the strain of origin for each tumor. The overall contribution of each strain to non-neoplastic liver was determined using the PCR assay and through analysis of the relative amount of glucose phosphate isomerase activity associated with the BR and B6 electrophoretic variants; the median contribution of B6 cells to non-neoplastic liver was 50%. A majority (91%) of the 230 tumors analyzed from both sexes was derived from the BR donor, indicating that the net overall effect of BR susceptibility genes is cell autonomous.

Publisher Summary The potential applications of predictive statistical models in toxicology based on gene expression measurements are numerous. For example, short-term studies measuring gene expression could be used to predict long-term toxicity studies like those still performed by the National Toxicology Program (NTP). Other short-term gene expression studies could be used to predict which chemicals would be teratogenic or cause more subtle developmental changes after human prenatal exposure. In either case, the application of microarray analysis and predictive statistical models has the potential to be extremely useful from both economic and human health perspectives. The application of predictive statistical models to chemically induced gene expression is the next logical step in the developing field of toxicogenomics. The development of these models may eventually open the door to a new era of toxicological testing where relatively short and inexpensive microarray studies may allow the assessment of the human health risks associated with a previously untested chemical. However, the accuracy and applicability of these models are highly dependent of the quality of the training sets used in their development. As the public gene expression database grows, more chemicals may be added to training the models and those models will become more predictive.

In mice, the sex difference in susceptibility to hepatocarcinogenesis results from the tumor promoting activity of testosterone and from the inhibition of tumor promotion by ovarian hormones. We investigated the role of growth hormone in the sex-dependent regulation of susceptibility, because sex hormones are known to regulate the temporal pattern of growth hormone secretion and subsequent sex differences in liver gene expression. We found that in both males and females, wild-type mice developed significantly more tumors than growth hormone-deficient, C57BL/6J-lit/lit (B6-lit/lit) mutant mice following perinatal treatment with the carcinogen N,N-diethylnitrosamine (DEN). B6 wild-type males developed 36-59-fold more liver tumors per animal than age matched B6-lit/lit males and wild-type females developed 11-fold more tumors than B6-lit/lit females. We bred the little mutation onto the more susceptible C57BR/cdJ (BR) and C3H/HeJ (C3H) strains to assess the effect of growth hormone deficiency on hepatocarcinogenesis on additional genetic backgrounds. Growth hormone deficiency suppressed liver tumor development to

Hormonal and genetic factors strongly influence the susceptibility of inbred mice to hepatocarcinogenesis. Female C57BR/cdJ (BR) mice are extremely susceptible to liver tumor induction relative to other strains because they are genetically insensitive to the inhibition of hepatocarcinogenesis by ovarian hormones. To determine the genetic basis for the sensitivity of BR mice relative to resistant C57BL/6J (B6) mice, we treated 12-day-old B6BRF1 x B6 and B6BRF1 x B6BRF1 (F2) animals with N,N-diethylnitrosamine (0.1 micromol/g of body weight) and enumerated liver tumors at 32 weeks of age in males and at 50 weeks in females. Genomic DNA samples from backcross and F2 mice were analyzed for 70 informative simple sequence length polymorphism markers. Genetic markers on chromosome 17 (D17Mit21) and chromosome 1 (D1Mit33) cosegregated with high tumor multiplicity in both sexes. Together, these loci [designated Hcf1 and Hcf2 (Hepatocarcinogenesis in females), respectively] account for virtually all of the difference in sensitivity between BR and B6 mice. The Hcf1 locus accounts for a majority of the higher susceptibility of BR mice of both sexes. Backcross female mice heterozygous at both loci (33 +/- 23 tumors per mouse) and at Hcf1 only (17 +/- 18) were 15- and 8-fold more sensitive, respectively, than mice homozygous for the B6 alleles at Hcf1 and Hcf2 (2.2 +/- 3.9). In backcross male mice, the double heterozygotes (35 +/- 22) and Hcf1 heterozygotes (28 +/- 12) were 5.4- and 4.3-fold more sensitive than mice homozygous for B6 alleles at both loci (6.5 +/- 5.4).

In order to study the interaction of genetic and hormonal factors during murine hepatocarcinogenesis, we compared the number of liver tumors induced by treatment of 12-day-old mice with N,N-diethylnitrosamine (DEN) (0.05 mumol/g body wt) in intact mice and animals gonadectomized at 8 weeks of age from the three inbred strains, C3H/HeJ (C3H), C57BL/6J (B6), and C57BR/cdJ (BR). At 50 weeks of age, the mean liver tumor multiplicity in intact BR females was 28 +/- 13, while that for intact female C3H and B6 mice was 1.4 +/- 4.7 and 0.5 +/- 1.0, respectively. In ovariectomized mice, the yield of liver tumors was approximately 8-fold higher than in intact C3H (10.3 +/- 7.5) and B6 (4.1 +/- 6.6) females. Only a slight increase (35 +/- 14) was seen in ovariectomized BR females compared to intact BR females. Castration resulted in lower mean tumor multiplicities at 32 weeks of age in the males of all three strains. Intact male C3H, B6, and BR mice had mean liver tumor multiplicities of 61 +/- 34, 7.4 +/- 13, and 26 +/- 18, respectively, while the mean tumor multiplicities in castrated C3H, B6, and BR mice were 24 +/- 14, 0.5 +/- 0.9, and 6.1 +/- 10 tumors per mouse, respectively. The apparent rate of growth of glucose-6-phosphatase-deficient, preneoplastic foci in DEN-treated BR females was significantly higher than in B6 females. The growth rates of hepatic foci in BR and B6 males were similar but foci in BR males were 5-fold more numerous than in B6 males. The high sensitivity of BR females may be due, at least in part, to the failure of ovarian hormones to inhibit the growth of preneoplastic foci and the subsequent development of liver tumors. Since BR males had a larger number of hepatic foci, it is likely that androgens increase the rate of focus formation in BR males.

The hepatocarcinogen sensitivity (Hcs) loci were originally identified as determinants of the approximately 50-fold higher susceptibility of male C3H/HeJ (C3H) mice to perinatally induced hepatocarcinogenesis relative to male C57BL/6J (B6) mice. These two inbred strains also differ in other phenotypes related to hepatocarcinogenesis, including their incidences of spontaneous liver tumors and the properties of neoplastic hepatic lesions. To test the hypothesis that the Hcs loci also influence these phenotypes, we characterized male mice from B6, C3H, and nine BXH recombinant inbred (a) strains for spontaneous liver tumor development, the frequency of activating mutations in tumors, and the presence of cytoplasmic inclusions in preneoplastic lesions. By comparing these results to the relative susceptibilities of the parental and Rl strains to N, N-diethylnitrosamine (a) -and N-ethyl-N-nitrosourea-induced hepatocarcinogenesis in preweanling male mice, we concluded that the C3H alleles of the Hcs loci also positively influence the spontaneous development of liver tumors in male animals. While strain-dependent differences in the frequency of Ha-ras-1 activation in DEN-initiated liver tumors were observed, this phenotype was not correlated with susceptibility to liver tumor induction. The formation of eosinophilic intracytoplasmic inclusion bodies observed specifically in B6 liver tumors, which has been suggested to be associated with the resistance of this strain to hepatocarcinogenesis, also segregated independently of the Hcs loci. © 1995 Wiley- Liss, Inc.

C3H/HeJ(C3H) mice are approximately 50-fold more susceptible to liver-tumor induction than C57BL/6J(B6) mice. This difference in susceptibility is a consequence of allelic differences in hepatocarcinogen sensitivity(Hcs) genes that control the growth of preneoplastic lesions in the liver. We have shown previously that these two strains differ in their responses to partial hepatectomy, which acts as a promoter of hepatocarcinogenesis in B6 mice but not in C3H mice. To determine whether there are also strain-specific differences in normal regulation of hepatic growth, we compared liver regeneration in C3H and B6 mice at the levels of DNA synthesis and gene expression. Partial hepatectomy induced a cascade of controlled events resulting in the regeneration of the liver to its original mass 11 d after surgery. We observed a two-fold greater level of DNA synthesis in C3H mice relative to B6 mice during the first peak of DNA synthesis, which occurred 35 h after hepatectomy in both strains. While the c-fos transcript was readily induced in both strains, there was a reduction in the expression of the late response genes E2F1 and dihydrofolate reductase in the livers of B6 mice when compared with the expression of these transcripts in the livers of C3H mice. The differential regulation of E2F1 between B6 and C3H mice may indicate that the Hcs genes and E2F1 function in the same signal-transduction pathway of normal growth control.© 1995 Wiley-Liss, Inc

Phenomenal progress has been made during the last decade in the molecular identification of human “cancer genes”, such as those resulting in the heritable development of retinoblastoma (1), Wilm’s tumor (2), and familial adenomatous polyposis (3). These rare genetic diseases are inherited with very high penetrance, making cancer development in carriers of the mutant genes a virtual certainty. Because of this high penetrance and the rarity of the cancers (or extreme nature of the phenotype), it has been possible to identify these genes through formal genetic studies in affected families. In contrast, human genetic studies have been less effective in identifying “risk-modifier” genes that may influence the susceptibility of individuals to the development of more common malignancies. Even if such genes resulted in 5- to 50-fold increases in cancer risk, their recognition would be hampered by a relatively low penetrance and the large variation in environmental factors that influence cancer development in humans. In spite of these difficulties, King and coworkers (4, 5) have recently demonstrated the existence of a locus that increases the risk for breast cancer in humans by tenfold and have mapped the gene to human chromosome 17q21.

Publisher Summary This chapter focuses on the genetic regulation of liver cancer development in mice. Understanding the mechanism of the action of specific genes that control hepatocarcinogenesis in rodents provides insights into the underlying biology of carcinogenesis and paradigms for understanding the action of risk-modifier genes in humans. Genetic approaches to understand the nature and regulation of the event associated stages of carcinogenesis include the comparative studies of inbred animal strains, linkage analysis to identify cancer susceptibility genes, and the analysis of carcinogenesis in animals that carry specific mutations. The studies of both spontaneous and chemically induced tumors in mice and rats have demonstrated that there is considerable variation among inbred strains in their susceptibilities to carcinogenesis. The cancer susceptibility genes that determine the differences in susceptibility between inbred strains can be identified through a “reverse genetic” approach. A wide variety of agents, including microsomal enzyme inducers, hypolipidemic agents, and steroid hormones, act as the promoters of hepatocarcinogenesis in rats and mice. Chronic treatment with phenobarbital (PB), a strong promoter of hepatocarcinogenesis in rat, resulted in significant increases in the number of preneoplastic and neoplastic hepatic lesions in DEN-treated C3H/He and DBA/2, but not C57BL/6 mice. Recent studies have provided clues to the molecular basis for the high susceptibility of C3H hepatocytes to in vitro transformation. The karyotypic analysis of cell lines derived from C3H hepatocytes revealed that all of the lines contained the monosomy or partial deletion of chromosome 4. Based on the results, it is proposed that this region of chromosome 4 contained a suppressor gene, designated Lci (liver cell immortalization) for which the C57BL/6 allele was more active than the C3H allele.

E2F has been implicated in growth control because of its association with the retinoblastoma protein and the presence of E2F binding sites in the promoters of several growth-regulated genes. Proteins that bind to an E2F site have been cloned from human and mouse cells. However, these two proteins (human E2F1 and mouse DP-1) are quite different in sequence. We have now cloned a mouse cDNA encoding a protein 86% identical to the human E2F1 protein. The mouse E2F1 cDNA encodes a 430-amino-acid protein with a predicted molecular weight of 46,322 and detects mRNAs of 2.7 and 2.2 kb. Using primers complementary to sequences in the mouse E2F1 3' untranslated region, we mapped the mouse E2F1 gene to chromosome 2, near the Agouti and c-src loci. To understand the role of the different E2F family members in the growth of mouse NIH 3T3 cells, we examined the levels of E2F1 and DP-1 mRNAs in different stages of the cell cycle. Since the levels of E2F1 but not DP-1 mRNA correlated with changes in transcription from the dhfr promoter, we examined whether E2F1 could activate various growth-regulated promoters. We found that E2F1 could activate some (dhfr, thymidine kinase, and DNA polymerase alpha) but not all (thymidylate synthase, cad, and c-myc) of these promoters. On the basis of changes in levels of E2F1 and its ability to transactivate growth-regulated promoters, we propose that E2F1 may mediate growth factor-initiated signal transduction.

Although expression of the enzyme gamma-glutamyl transpeptidase (GGT) is the most common phenotypic marker of preneoplastic foci in the livers of carcinogen-treated rats, it is not generally expressed in mouse liver tumors or hepatic foci. However, several carcinogens, including safrole and ortho-azoaminotoluene (OAT), have been reported to induce GGT-positive foci in mice. We asked whether safrole and OAT induce GGT expression in preneoplastic foci or if these compounds select for a distinct set of lesions that can be identified by their GGT-positive phenotype. We treated 12-day-old male and female C57BL/6J mice with N,N-diethylnitrosamine (DEN) (0.20 mumol/g body wt) to initiate hepatocarcinogenesis. From 6 to 24 weeks of age, during the promotion phase of hepatocarcinogenesis, groups of mice were treated with 3,4,5,3',4',5'-hexabromobiphenyl (HBB), safrole or OAT. Additional groups of female mice were ovariectomized at 6 weeks of age with or without subsequent chronic treatment with testosterone. All the animals were killed at 24 weeks of age and serial liver sections were stained for glucose-6-phosphatase (G6Pase) or GGT. Both testosterone and HBB were strong promoters of the development of G6Pase-deficient foci. No GGT-positive foci were observed in animals treated with these agents or with DEN alone. In mice fed safrole or OAT during the promotion period, female mice developed more G6Pase-deficient foci than male mice, and GGT-positive foci were observed. Analysis of serial sections revealed that the G6Pase-deficient foci and the GGT-positive foci were independent populations. The relative number of these two classes of foci varied according to the treatment regimen. In females fed safrole, 7% of the foci in the liver were GGT-positive while in female mice fed OAT, 45% were GGT-positive. In all groups of mice in which we observed GGT-positive foci and in ovariectomized female mice, we noted a third independent population of foci which demonstrated significantly increased expression of G6Pase relative to surrounding normal liver. These data indicate that different treatments during the promotion stage of hepatocarcinogenesis in the mouse may give rise to distinct populations of preneoplastic lesions. Further studies of the molecular events giving rise to these distinct lesions will provide insights into the multiple pathways that result in hepatocarcinogenesis.

The PPCD1 mouse, a spontaneous mutant that arose in our mouse colony, is characterized by an enlarged anterior chamber resulting from metaplasia of the corneal endothelium and blockage of the iridocorneal angle by epithelialized corneal endothelial cells. The presence of stratified multilayered corneal endothelial cells with abnormal patterns of cytokeratin expression are remarkably similar to those observed in human posterior polymorphous corneal dystrophy (PPCD) and the sporadic condition, iridocorneal endothelial syndrome. Affected eyes exhibit epithelialized corneal endothelial cells, with inappropriate cytokeratin expression and proliferation over the iridocorneal angle and posterior cornea. We have termed this the "mouse PPCD1" phenotype and mapped the mouse locus for this phenotype, designated "Ppcd1", to a 6.1 Mbp interval on Chromosome 2, which is syntenic to the human Chromosome 20 PPCD1 interval. Inheritance of the mouse PPCD1 phenotype is autosomal dominant, with complete penetrance on the sensitive DBA/2J background and decreased penetrance on the C57BL/6J background. Comparative genome hybridization has identified a hemizygous 78 Kbp duplication in the mapped interval. The endpoints of the duplication are located in positions that disrupt the genes Csrp2bp and 6330439K17Rik and lead to duplication of the pseudogene LOC100043552. Quantitative reverse transcriptase-PCR indicates that expression levels of Csrp2bp and 6330439K17Rik are decreased in eyes of PPCD1 mice. Based on the observations of decreased gene expression levels, association with ZEB1-related pathways, and the report of corneal opacities in Csrp2bp(tm1a(KOMP)Wtsi) heterozygotes and embryonic lethality in nulls, we postulate that duplication of the 78 Kbp segment leading to haploinsufficiency of Csrp2bp is responsible for the mouse PPCD1 phenotype. Similarly, CSRP2BP haploinsufficiency may lead to human PPCD.

We used a shuttle vector based on the Epstein-Barr virus origin of plasmid replication (oriP) to determine the types of mutations induced by depurination in human cells. Plasmid DNA was incubated at pH 2 at 40 degrees C for various times to induce up to 20 apurinic (AP) sites per 9.7-kb plasmid and electroporated into lymphoblastoid cells derived from either a normal individual or an ataxia telangiectasia patient. After replication of the vector in the human cells, plasmid DNA was isolated and analyzed for mutations induced in the plasmid-encoded herpes simplex virus type 1-thymidine kinase gene. Both the frequencies and types of mutations induced by depurination were essentially identical for normal and ataxia telangiectasia cells. The mutant frequency at 20 AP sites/plasmid was 10-fold to 13-fold greater than that observed for untreated DNA. Deletion and frameshift events accounted for 46-55% of the mutants induced by depurination. The induced deletions were relatively small (median size, 100-150 bp) and characterized by short (1-5 bp) regions of sequence homology at the endpoints. These mutations and the frameshifts, a majority of which occurred in runs of identical nucleotides, are consistent with a model involving AP-site-induced template dislocation during DNA synthesis. A broad spectrum of base-substitution mutations, which accounted for 19-36% of the induced mutants, was observed. The apparent preference for insertion opposite AP sites in human cells was G (43-55%) greater than A approximately C (18-21%) greater than T (9-14%). Our results in human cells contrast markedly with those published previously for the mutational specificity of AP sites in Escherichia coli, in which a large majority of the mutants resulted from insertion of an A opposite the abasic site.

Expression of recessive mutant phenotypes can occur by a number of different mechanisms. Inactivation of the wild-type allele by base-substitution mutations, frameshift mutations or small deletions occurs at both hemizygous and heterozygous cellular loci, while other events, such as chromosome level rearrangements, may not be detected at hemizygous loci because of inviability of the resulting mutants. In order to assess the relative contribution of each type of mutational event, we isolated a human lymphoblastoid cell line that is heterozygous at the adenine phosphoribosyltransferase (aprt) locus. The mutation rate for the expression of the mutant phenotype (aprt(+/-)----aprt-/-) was 1.3 x 10(-5)/cell/generation. Molecular analysis of the DNA from 26 mutant clones revealed that 19% had undergone deletion of the entire wild-type allele. The aprt heterozygote carries a mutation in the coding sequence of the gene that results in the loss of a restriction site. Analysis of aprt-/- mutants for this restriction fragment length difference revealed that 23% of the mutants contained point mutations or small (less than 100 bp) deletions. The remainder of the mutants (58%) resulted from reduction to homozygosity of the mutant allele. We suggest that, as in tumor cells in vivo, reduction to homozygosity is a major mechanism for the expression of recessive mutant phenotypes in cultured human cells.

The frequency and pattern of mutations at codon 61 of the c-Ha-ras gene have been analyzed in 195 liver tumors and 132 precancerous liver lesions from various rodent strains with differing susceptibility to hepatocarcinogenesis. By using the polymerase chain reaction and allele-specific oligonucleotide hybridization, C----A transversions at the first base and A----T transversions or A----G transitions at the second base of c-Ha-ras codon 61 were detected in 20-60% of spontaneous or carcinogen-induced liver tumors of the C3H/He, CBA, CF1, and B6C3F1 mouse strains, which are highly susceptible to hepatocarcinogenesis. No such mutations, however, could be found in any of the 31 liver tumors of the insensitive C57BL/6J and BALB/c mouse strains or in any of the 35 liver tumors of the comparatively resistant Wistar rat. Further analyses of c-Ha-ras codon 12 mutations in liver tumors from the three insensitive rodent strains also failed to give any positive results. In early precancerous liver lesions, c-Ha-ras codon 61 mutations were found in 13-14% of lesions of the sensitive C3H/He and B6C3F1 mouse strains but not in any of the 34 lesions of the insensitive C57BL/6J mouse. Taken together, our results indicate a close correlation between the mutational activation of the c-Ha-ras gene in liver tumors of the different rodent strains and their susceptibility to hepatocarcinogenesis, whereby the mutations appear to provide a selective growth advantage, leading to a clonal expansion of the mutated liver cell population, only in livers of sensitive but not of insensitive strains.

The number of apurinic/apyrimidinic (AP) sites in supercoiled SV40 deoxyribonucleic acid (DNA) after treatment with several electrophilic mutagens was quantitated by electrophoretic analysis of the DNA after cleavage of the phosphodiester bonds adjacent to AP sites by a specific endonuclease. The compounds studied, in order of increasing yields of AP sites obtained on incubation with the DNA for 5 h at 37 degrees C, were dimethylcarbamoyl chloride, ethyl methanesulfonate, N-ethyl-N-nitrosourea, 2-(N-acetoxyacetylamino)fluorene, beta-propiolactone, N-methyl-N-nitrosourea, methyl methanesulfonate, 1'-acetoxyestragole, 4-(N-acetoxyacetylamino)stilbene, (+/-)-7 beta, 8 alpha-dihydroxy-9 alpha, 10 alpha-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene, N-(benzoyloxy)-N-methyl-4-aminoazobenzene, and 1-pyrenyloxirane. After a 5-h incubation at 37 degrees C and extraction of unreacted compound, further incubation at 70 degrees C generally increased the yield of AP sites; an exception was N-(benzoyloxy)-N-methyl-4-aminoazobenzene-reacted DNA. Except for DNA treated with N-ethyl-N-nitrosourea and N-methyl-N-nitrosourea, which are known to bind to a significant extent to DNA phosphates, the number of alkali-labile lesions in the treated DNA was similar to the number of AP sites. For the compounds studied there was no direct correlation between the number of AP sites produced and missense mutagenic activity, as measured in Salmonella typhimurium strain TA100.

Approximately 20% of a dose of estragole, a naturally occurring flavoring agent, was excreted in the urine of outbred male CD -1 mice as a conjuage (presumably the glucuronide) of 1'-hydroxyestragole, Estragole and its 1'-hydroxy metabolite caused significant increases in the incidences of hepatocellular carcinomas in male CD-1 mice that received the compounds by sc injection at 1-22 days of age. Estragole induced hepatocellular carcinomas by 15 months in 23 and 39% of the mice that received total doses of 4.4 and 5.2 mumoles, respectively, and lived to an age of 12 months or more. Of the 12-month survivors given a total dose of 4.4 mumoles of 1'-hydroxyestragole, 70% developed hepatocellular carcinomas; the incidence in mice that received only the vehicle (trioctanoin) was 12%. Multiple tumors ocurred in 5, 28, 64, and 0%, respectively, of the mice in each of these 4 groups. Of the mice given a total dose of 4.4 mumoles of 1'-hydroxysafrole, 59developed hepatocellular carcinomas; 39% of the mice bore multiple liver tumors. As previsously demonstrated for 1'-acetoxysafrole, 1'-acetoxyestragole and 1'-acetoxy-1-allyl-4-methoxynaphthalene reacted nonenzymatically with guanosine and inosine to form adducts. These electrophilic esters were strongly mutagenic for the Salmonella typhimurium missense mutant TA100. 1'-Acetoxyallybenzene had little or no activity in either of these tests. Attempts to demonstrate liver-mediated mutagenicty for 1'-hydroxysafrole and 1'-hydroxyestragole in the bacterial test system were unsuccessful.