Tomoko Nomoto

Expression Quantitative Trait Loci Analysis of 13 Genes in the Rat Prostate

Differential expression of mRNA among animal strains is one of the mechanisms for their diversity. cDNA microarray analysis of the prostates of BUF/Nac (BUF) and ACI/N (ACI) rats, which show different susceptibility to prostate cancers, found 195 differentially expressed genes. To identify loci that control differential expression of 13 genes with diverse expression levels, their expression levels were measured by quantitative RT-PCR in 89 backcross rats, and expression quantitative trait locus (eQTL) analysis was performed. Nine genes [Aldh1a1, Aldr1, Bmp6, Cdkn1a (p21), Cntn6, Ghr, Jund, Nupr1, and RT1-M3] were controlled by cis-acting loci. Cdkn1a, a cell cycle regulator and a candidate for a prostate cancer susceptibility gene, was mapped to its own locus and had polymorphisms, including a 119-bp insertion in the 5′ upstream region in BUF rats. Four genes (Kclr, Pbsn, Psat1, and Ptn) were controlled by trans-acting loci. Pbsn, a prostate-specific gene on chromosome X, was controlled by a QTL on chromosome 8. Depending upon which gene that we selected from the genes widely used for normalization (Actb, Gapd, or Ppia), different QTL were mapped for Kclr, Psat1, and Ptn. Normalization using Actb most appropriately explained the expression levels in a congenic strain for chromosome 3. eQTL analysis with precise measurement of expression levels and appropriate normalization was shown to be effective for mapping loci that control gene expression in vivo.

Cloning of the rat agouti gene and identification of the rat nonagouti mutation.

The agouti locus is one of the classic coat color loci of laboratory animals such as mice and rats (Doolittle et al. 1996; RATMAP 2000). In the mouse, the agouti gene encodes a 131-amino acid paracrine signaling molecule whose pulsatile expression produces a characteristic pattern of banded pigment in individual hairs (Bultman et al. 1992; Miller et al. 1993; Michaud et al. 1994). Gain-of-function mutation of mouse agouti, such as lethal-yellow (A) and viable-yellow ( A) result in pleiotropic effects, including obesity, diabetes, increased tumor susceptibility, and premature infertility (Duhl et al. 1994; Michaud et al. 1994). Loss-of-function mutation of agouti, such as nonagouti ( a), results in a plain black coat on the back and belly. On the other hand, in the rat, the agouti gene has not been cloned yet. The molecular bases for other rat color mutations, including albino, beige, brown, cinnamon, fawn, nonagouti, mahogany and pinked-eye dilution, have not been identified yet (Prieur and Meyers 1984; Yoshida et al. 1987; Nishimura et al. 1989; Dekker-Ohno et al. 1993; Ohno et al. 1996; Robinson 1998), in spite of the fact that these coat color mutations have been used as important genetic resources. In this study, we describe the determination of the rat agouti cDNA sequence and the identification of the rat nonagouti mutation. To identify rat expressed sequence tags (ESTs) showing homology with the mouse agouti cDNA, we screened the dbEST by a BLAST search with the sequence of mouse agouti (accession number L06941). Two rat ESTs, UI-R-C2P-RK-D-09-0-UI and UI-R-C3-TP-C-03-0-UI, were identified, and both originated from the 12-day embryos of nonagouti Sprague-Dawley (SD) rats. We purchased these two EST clones from Research Genetics (Huntsville, Ala.) and sequenced one of them (UI-R-C3-TP-C-03-0-UI). The most 58 sequences (pos. from 1 to 65) of the insert showed significant homology with the ventral-specific first exon of the mouse agouti gene (accession number L76476), and the most 3 8 sequence also showed significant homology with the mouse agouti cDNA sequence. In order to amplify the full-length cDNA of the rat agouti gene, we designed PCR primers as follows; 5 8-CCA GTC TGA GTC CTT GAG CC-38 and 58-ACT GTA TTC TGA TTT TAG CCT CCA-38. We performed RT-PCRs using cDNAs from the ventral skin of 5-day neonatal ACI (agouti) rats, and obtained∼800-bp fragments. Combining the data from the direct sequencing with the data from the EST clone, we were able to determine a 716-bp sequence of the rat wild-type agouti cDNA (Fig. 1), which encompassed the complete coding sequence (CDS) with 131 amino acids and a 33-bp poly-A tail (accession number AB045587). The deduced rat agouti polypeptide shared 123 amino acids with the mouse agouti protein and had all of the known biologically active sites, such as a signal sequence, an N-linked glycosylation site, a lysine-rich basic domain, and C-terminal cysteine residues (Bultman et al. 1992; Perry et al. 1996). We further obtained the exon-intron boundaries of the agouti gene by amplifying introns by PCR. The PCR primers were designed based on the positions of introns deduced from the sequences of the mouse agouti gene (Wilson et al. 1995), and PCR was performed with the genomic DNAs from the ACI rat. Approximately 1.8-kb and 3.0-kb fragments were obtained from introns 2 and 3, respectively. However, no fragment was obtained from intron 1. We determined the nucleotide sequences of the splice donor and acceptor sites of the two introns (accession numbers AB045588, AB045589, and AB045590). To search for the causative mutation for the nonagouti phenotype, we first examined mRNA expression levels of agouti in the skin of agouti ACI and WKAH rats and nonagouti BN rats. A probe derived from the CDS of the rat agouti gene was used to detect the transcripts in the neonatal (5d) skin of the three strains of rats on Northern blots. A 0.8-kb transcript was detected in ACI and WKAH rats, while almost no transcript was detected in BN rats (Fig. 2). Quantities of the agouti transcripts normalized to those ofb-actin were measured by using an Imaging plate and a computerized image display system BAS2000 (Fuji Film, Tokyo, Japan). Compared with the ACI rat, expression levels of the transcripts in WKAH and BN rats were calculated as 54% and 15%, respectively. Because WKAH showed the agouti coat color phenotype, we initially hypothesized that the nonagouti coat color phenotype was caused only when the agouti transcript was markedly reduced. In order to identify a causative mutation for the marked reduction of the agouti transcript in nonagouti rats, we determined the nucleotide sequences of the entire cDNA and exon-intron boundaries, using agouti ACI and nonagouti BN rats. The full-length cDNAs were amplified from the neonatal (5d) ventral skin of both strains. Direct sequencing of the cDNA revealed that the BN rat had a 19-bp deletion which was located at nucleotide positions 202 to 220 of the rat agouti cDNA (Fig. 1). The deletion was expected to result in a frame-shift from the 36th amino acid of the rat agouti and produce a stop codon at the next 12th amino acid. The resultant truncated polypeptide completely lacks the N-glycosylation site, the highly basic domain, and the C-terminus cysteine residues, suggesting the loss-of-function of agouti in the rat strains carrying the 19-bp deletion. In addition to the deletion , a G to A base substitution was found at the 5th position of the splice donor site of putative intron 3, which appeared to disrupt the consensus donor sequence (GTAAGT or GTGAGT) (Alberts et al. 1994). To clarify which mutation is responsible for the nonagouti phenotype, correlation of the presence/absence of the respective mutation with the coat color phenotype was examined. A complete correlation between the genotype and the phenotype in the nine nonagouti strains (BN, BUF, DON, F344, SD, SHR, TM, WKY, WTC) and six agouti strains (ACI, IS, LEA, LEC, NER, WKAH) (Festing 1998) was observed in the case of the 19-bp deletion (Fig. 3). On the other hand, in the case of the splice site mutation, the correlation was not perfect. The mutant A nucleotide was present in the agouti WKAH strain, while the correlation was good in the rest of the strains (Fig. 3). Therefore, based on the predicted effect Correspondence to: T. Ushijima; E-mail: tushijim@ncc.go.jp Mammalian Genome 12, 469–471 (2001). DOI: 10.1007/s003350020010

β-Catenin mutations and nuclear accumulation during progression of rat stomach adenocarcinomas

Aberrant Wnt/β‐catenin signaling caused by mutations in exon 3 of the β‐catenin gene has been identified in a number of human malignancies, including stomach cancer. However, studies of mutation frequency have yielded conflicting results, and timing during progression remains largely unknown. In this study, we utilized an animal model to address this question. A total of 20 ACI male rats were treated with N‐methyl‐N′‐nitro‐N‐nitrosoguanidine (MNNG) in the drinking water and 22 induced differentiated adenocarcinomas were histopathologically and immunohistochemically evaluated for β‐catenin localization. Fourteen tumors (63.6%) that showed homogeneous low‐grade morphology, preserving cell polarity, were found to harbor β‐catenin protein on the cell membranes (M). Eight tumors exhibited regions of high‐grade morphology among areas with low‐grade morphology, and they were characterized by denser cell growth and loss of cell polarity. Among these 8 tumors, 4 (18.2%) showed cytoplasmic localization (C) of β‐catenin in small regions. The remaining 4 tumors (18.2%) contained more dysplastic regions that displayed nuclear (N) β‐catenin staining. Analysis of DNA obtained by microdissection demonstrated that all of 4 regions with C staining and 20 with M staining, as well as 17 samples of surrounding normal mucosa (S) had wild‐type β‐catenin. In contrast all of 3 regions with N staining featured mutations (3 of 3=100%; N vs. C, P<0.05; N vs. M and N vs. S, P<0.001, Fishers exact test) in exon 3, at glycine 34, threonine 41, and serine 45, which affected phosphorylation sites. In conclusion, β‐catenin mutations appear to be associated with the late progression stage of adenocarcinoma development in rat stomach carcinogenesis, in contrast to the case of colorectal cancers, in which mutations appear to occur in the early stages.

Whole-genome analyses of loss of heterozygosity and methylation analysis of four tumor-suppressor genes in N-methyl-N'-nitro-N-nitrosoguanidine-induced rat stomach carcinomas.

N‐Methyl‐N′‐nitro‐N‐nitrosoguanidine (MNNG)‐induced rat stomach carcinomas are considered to be a good model for differentiated‐type human stomach carcinomas. However, as for their molecular basis, only infrequent mutations of Catnb (β‐catenin) and Trp53 (p53) have been observed. Here, we carried out a whole‐genome analysis of loss of heterozygosity (LOH) using 21 stomach carcinomas induced by MNNG in F1 hybrids of ACI and BUF rats, and also analyzed promoter methylation of four tumor‐suppressor genes. LOH analysis was performed using 130 polymorphic markers covering rat chromosomes 1–20 with an average interval of 20 Mbp. Despite adapting conditions so that LOH could be detected with up to a 50% contamination of stromal cells, no LOH was detected at any loci. CpG islands in putative promoter regions of four tumor‐suppressor genes, Cdh1 (E‐cadherin), Cdkn2a (p16), Mlh1, and Rassf1a, were analyzed by methylation‐specific polymerase chain reaction (PCR). However, no methylation was detected. In contrast, the promoter region of Pgc (pepsinogen C), which lacks a CpG island, was methylated in all 21‐cancer samples. These results indicated that LOH spanning a chromosomal region larger than 30–40 Mbp or silencing of Cdh1, Cdkn2a, Mlh1, and Rassf1a, was not involved in MNNG‐induced rat stomach carcinomas. The search for other genes involved in these carcinomas needs to be continued. (Cancer Sci 2005; 96: 409  – 413)

Linkage and Microarray Analyses of Susceptibility Genes in ACI/Seg Rats: A Model for Prostate Cancers in the Aged

ACI/Seg (ACI) rats develop prostate cancers spontaneously with aging, similar to humans. Here, to identify genes involved in prostate cancer susceptibility, we did linkage analysis and oligonucleotide microarray analysis. Linkage analysis was done using 118 effective rats, and prostate cancer susceptibility 1 (Pcs1), whose ACI allele dominantly induced prostate cancers, was mapped on chromosome 19 [logarithm of odds (LOD) score of 5.0]. PC resistance 1 (Pcr1), whose ACI allele dominantly and paradoxically suppressed the size of prostate cancers, was mapped on chromosome 2 (LOD score of 5.0). When linkage analysis was done in 51 rats with single or no macroscopic testicular tumors, which had larger prostates and higher testosterone levels than those with bilateral testicular tumors, Pcs2 and Pcr2 were mapped on chromosomes 20 and 1, respectively. By oligonucleotide microarray analysis with 8,800 probe sets and confirmation by quantitative reverse transcription-PCR, only two genes within these four loci were found to be differentially expressed >1.8-fold. Membrane metalloendopeptidase (Mme), known to inhibit androgen-independent growth of prostate cancers, on Pcr1 was expressed 2.0- to 5.5-fold higher in the ACI prostate, in accordance with its paradoxical effect. Cdkn1a on Pcs2 was expressed 1.5- to 4.5-fold lower in the ACI prostate. Additionally, genes responsible for testicular tumors and unilateral renal agenesis were mapped on chromosomes 11 and 14, respectively. These results showed that prostate cancer susceptibility of ACI rats involves at least four loci, and suggested Mme and Cdkn1a as candidates for Pcr1 and Pcs2.

Insertional mutation of the Attractin gene in the black tremor hamster

The hamster black tremor (bt) mutation induces a black coat color and a defective myelination in the central nervous system (CNS) that manifests as a tremor. On the other hand, loss-of-function mutations of the Attractin (Atrn) gene, such as Atrnmg, Atrnmg-L, and Atrnmg-3J in mice, and Atrnzi in rats, induce both darkening of coat color and hypomyelination and vacuolation in the CNS. The close resemblance of the mutant phenotypes led us to postulate that the bt/bt hamster also might harbor a mutation in Atrn. Here, we cloned the hamster Atrn cDNA and identified bt as a loss-of-function mutation of Atrn. While the human and rat Atrn genes encode both membrane- and secreted-type proteins, the hamster Atrn gene encoded only membrane-type protein with 1,427 amino acids, as in the case of the mouse. Hamster Attractin protein had 93.6%, 96.8%, and 96.8% identities with human, rat, and mouse membrane-type Attractin. In the brain of the bt/bt hamster, aberrant transcripts with more than three size species were observed, and the most predominant transcript encoded the truncated Attractin without transmembrane domain. In the Atrn gene of bt/bt hamster, an ∼10-kb DNA fragment, which had 557-bp direct repeats in both ends and was flanked by the identical 6-bp target duplication sequences, was inserted into exon 24. In addition, the insertion was cosegregated with neurodegeneration in the CNS of 50 intercross progeny. These results indicated that the hamster bt mutation was the ∼10-kb retrotransposon-like insertion into the Atrn gene, which resulted in aberrant transcripts. The bt/bt hamster will provide a useful tool for further understanding of the pleiotropic functions of Attractin.

Predisposition to lung tumorigenesis.

Mouse inbred strains with inherited predisposition and resistance to lung cancer provide an essential tool for the dissection of the genetics of this complex disease. We have previously mapped a major locus (Pulmonary adenoma susceptibility 1, Pas1) affecting inherited predisposition to lung cancer in mice on chromosome 6, near Kras2. Appropriate crosses that include susceptible mice (Pas1(s)) provide a model system for identifying loci that can modify the lung cancer predisposition phenotype caused by Pas1. Using this approach we have mapped the Pulmonary adenoma resistance 1 (Par1) locus that behaves like a modifier locus of Pas1. More recently, we mapped additional lung tumor resistance loci (Par2, and Par4), and a locus specifically involved with lung tumor progression (Papg1). The mapping of Pas1 in mice stimulated us to test the possible association of genetic markers located in the homologous human region (12p12) with risk and prognosis of lung adenocarcinomas in man. In the Italian population, we carried out an association study by genotyping lung adenocarcinoma patients and healthy controls for genetic markers located in the putative region of interest. Homozygosity of the A2 allele at a Kras2/RsaI polymorphism, and allele 2 at a VNTR polymorphism in the PTHLH gene showed borderline statistically significant associations with lung cancer risk. Furthermore, the same alleles were significantly associated with tumor prognosis. Studies on association were then performed in the Japanese and in European populations. In the Japanese population, the KRAS2/RsaI marker was significantly associated with prognosis of lung adenocarcinoma, whereas the European study did not confirm this association. Our results may provide evidence for the existence of the human PAS1 locus, suggesting that the mouse model of inherited predisposition to lung tumorigenesis is predictive of a human genetic mechanism of susceptibility to lung cancer.

Differential expression of genes related to levels of mucosal cell proliferation among multiple rat strains by using oligonucleotide microarrays.

Induction levels of cell proliferation, in response to gastric mucosal damage by N-methyl-N′-nitro-N-nitrosoguanidine (MNNG), are different among rat strains and correlate with susceptibility to MNNG-induced gastric carcinogenesis. Here, we used oligonucleotide microarrays to search for genes that show expression levels accordant with the extents of cell proliferation among six rat strains. Expression levels of 8,800 probe sets were analyzed in the pylorus of ACI, LEW, WKY (strains with strong cell proliferation), F344, (ACI × BUF)F1, and BUF rats (strains with weak cell proliferation) after 2-week MNNG treatment. No genes showed complete accordance, and 22 genes showed accordance with one or two exceptions. After confirmation by quantitative RT-PCR, four genes—cellular retinoic acid-binding protein II (Crabp2), fatty acid binding protein 1 (Fabp1), progastricsin (pepsinogen C, Pgc), and UDP-glucuronosyltransferase 2 family member 5 (Ugt2b5)—were found to show good accordance with only one exception. Crabp2, Fabp1, and Ugt2b5 were differentially expressed between ACI and BUF rats both before and after MNNG treatment. Although Crabp2 had been identified as one of the 16 genes differentially expressed between ACI and BUF rats with cDNA-RDA, Fabp1 and Ugt2b5 were newly identified in this study. All three genes are known to be involved in retinoic acid-mediated signaling and could be involved in the control of differential induction of cell proliferation.

Linkage disequilibrium pattern in the L‐myc gene in Italian and Japanese non‐small‐cell lung‐cancer patients

Italian and Japanese non‐small‐cell lung‐cancer patients were genotyped for an intragenic L‐myc EcoRI restriction site polymorphism previously reported to be associated with lung‐tumor prognosis in Asian populations but not in Caucasians. Screening of the L‐myc sequence in Italian samples allowed identification of 2 additional 3′‐UTR SNPs, located 2.3–3.0 kb from the EcoRI polymorphism, but no coding polymorphism was found. No significant association was found between any of the 3 SNPs and lung‐tumor prognosis in Italian patients, consistent with the reported difference between Caucasian and Asian populations. Moreover, the newly discovered polymorphisms in the Italian group were not present in Japanese patients. Significant LD between EcoRI and the 2 other SNPs was detected in the Italian population, whereas no significant LD between the 2 3′‐UTR markers was detected despite their close proximity (0.7 kb). Thus, the disparate conclusions about the role of L‐myc polymorphism in tumor prognosis among different populations may rest in population‐specific LD between the functional gene and the L‐myc polymorphism.