Masahide Shiozawa

Implications of circadian gene expression in kidney, liver and the effects of fasting on pharmacogenomic studies.

Pharmacogenomics offers the potential to define metabolic pathways and to provide increased knowledge of drug actions. We studied relative levels of gene expression in the rat using a microarray with 8448 rat UniGenes (1928 known genes, 6520 unknown ESTs) in the liver and kidney as a function of time of day and then of feeding regime, which are common variables in preclinical pharmacogenomic studies. We identified 597 genes, including several key metabolic pathways, whose relative expression levels are significantly affected by time of day: expression of some was further modified by feeding state. These would have sparked interest in a pharmacogenomic study. Our study demonstrates that two common variables in pharmacogenomic studies can have dramatic effects on gene expression. This study provides investigators with baseline information for both kidney and liver with respect to normal changes in gene expression influenced by time of day and feeding state. It also identifies 18 new genes that should be investigated for a role in circadian rhythms in peripheral tissues.

An integrated genetic linkage map of the laboratory rat

Abstract. The laboratory rat, Rattus novegicus, is a major model system for physiological and pathophysiological studies, and since 1966 more than 422,000 publications describe biological studies on the rat (NCBI/Medline). The rat is becoming an increasingly important genetic model for the study of specific diseases, as well as retaining its role as a major preclinical model system for pharmaceutical development. The initial genetic linkage map of the rat contained 432 genetic markers (Jacob et al. 1995) out of 1171 developed due to the relatively low polymorphism rate of the mapping cross used (SHR × BN) when compared to the interspecific crosses in the mouse. While the rat genome project continues to localize additional markers on the linkage map, and as of 11/97 more than 3,200 loci have been mapped. Current map construction is using two different crosses (SHRSP × BN and FHH × ACI) rather than the initial mapping cross. Consequently there is a need to provide integration among the different maps. We set out to develop an integrated map, as well as increase the number of markers on the rat genetic map.The crosses available for this analysis included the original mapping cross SHR × BN reciprocal F2 intercross (448 markers), a GH × BN intercross (205 markers), a SS/Mcw × BN intercross (235 markers), and a FHH/Eur × ACI/Hsd intercross (276 markers), which is also one of the new mapping crosses. Forty-six animals from each cross were genotyped with markers polymorphic for that cross. The maps appear to cover the vast majority of the rat genome. The availability of these additional markers should facilitate more complete whole genome scans in a greater number of strains and provide additional markers in specific genomic regions of interest.

Cloning and characterization of Xenopus laevis xSox7 cDNA.

A family of SRY-related genes has been termed SOX. We have isolated and sequenced a cDNA encoding a novel Sox protein from Xenopus laevis ovary. This cDNA contains an open reading frame (ORF) coding for 362 amino acids, which encompasses an HMG box and exhibits a strong (90%) identity to that of mouse Sox7; the cDNA was named xSox7 in this study. Northern analysis revealed that the xSox7 mRNA was 2.0 kb in length. Various adult frog tissues were tested by reverse transcription/polymerase chain reaction for xSox7 mRNA, and the results showed that xSox7 is expressed in a wide range of tissues. Furthermore, electrophoretic mobility shif assay indicated that recombinant xSox7 is capable of binding to AACAAT sequence.

XLS13A and XLS13B: SRY-related genes of Xenopus laevis

SRY-related cDNAs, XLS13A and XLS13B, have been isolated from Xenopus laevis ovary. The cDNAs encode polypeptides of 382 and 375 amino acids, respectively. Nucleotide sequences of the two cDNAs are highly homologous to each other. The type-A and type-B XLS13 proteins, and xSox13 reported previously share an identical high mobility group (HMG) box at the amino acid level, although they contain silent nucleotide alterations. The HMG box exhibits strong similarity (> 93% amino acid identity) to those of mouse Sox4/human SOX4 and chicken Sox11/human SOX11. The size of XLS13A/XLS13B mRNA was estimated to be 2.8 knt in Xenopus ovary by Northern analysis. Reverse transcription/polymerase chain reaction (RT/PCR) assay indicated that XLS13A and XLS13B mRNAs are present in various tissues of adult frog. The mRNAs of XLS13A and XLS13B of maternal origin found in unfertilized eggs disappear in the early stages of the Xenopus embryo. DNA-binding properties of the XLS13 HMG domain were examined by electrophoretic mobility shift assay (EMSA). The HMG domain preferentially binds to the canonical target sequence of SOX proteins, AACAAT, in vitro.

CLONING AND EXPRESSION OF XENOPUS LAEVIS XSOX12 CDNA

A family of SRY-related genes has been termed SOX. We have isolated and sequenced a cDNA encoding xSox12 from Xenopus laevis ovary. The cDNA contained an open reading frame (ORF) coding for 470 amino acids encompassing an HMG box characteristic of the SOX family, a leucine zipper motif and glutamine-rich segments. The size of the xSox12 mRNA was determined to be 3.0 knt by Northern analysis. The ovary was the most prominent in the expression of the Sox mRNA among the various tissues of adult frog as far as examined.

Mapping of the rat SM22 gene to chromosome 8q24: a candidate for high blood pressure and cardiac hypertrophy.

SM22 is a smooth muscle cell (SMC)-specific protein originally isolated from chicken gizzard (Lees-Miller et al. 1987). The gene encoding SM22 is mainly expressed in SMCs (Shanahan et al. 1993). Although SM22 is utilized as an SMC-specific marker and is suggested to play a role in smooth muscle contraction (Kobayashi et al. 1994), the function of this protein remains unknown. Smooth muscle contraction and proliferation have been considered to play a crucial role in the pathogenesis of high blood pressure. Interestingly, expression of the SM22 gene was also identified in cardiac and skeletal muscle cells only during embryogenesis (Li et al. 1996), suggesting a potential contribution of SM22 to cardiac development. In the past few years, several genetic mapping studies using various rat models have identified several quantitative trait loci (QTLs) for cardiovascular phenotypes, such as high blood pressure and cardiac hypertrophy (Pravenec et al. 1995; Schork et al. 1995). If the rat SM22 gene (Sm22) maps to the region containing one of these QTLs, Sm22 could be an attractive candidate gene responsible for high blood pressure or cardiac hypertrophy. Furthermore, this would be an important clue to elucidate the function of SM22. In this study, we mapped Sm22 by genetic mapping and fluorescence in situ hybridization (FISH). Using mouse cDNA for SM22 as a probe (Li et al. 1996), a rat genomic library was screened. This library was constructed from genomic DNA isolated from Fischer male rat liver and cloned into Lambda DASH vector (Stratagene Cloning Systems). From 1.8 x 106 plaques screened, three positive clones (XJ17-19) were isolated. Restriction enzyme mapping of these three clones revealed overlapping inserts. To generate genetic markers for Sm22, we screened XJ17 for simple sequence repeats (SSRs) by the method described previously (Koike et al. 1996). Clones containing an SSR [(CA)n repeat] were identified and sequenced. Then, primers flanking the (CA)n repeat were designed (D8Mcwl; forward primer: 5-ATGAGAACCTGATACCCCCC-3, reverse primer: 5-CCAAATGCCATGATGAAATG-3) by the computer program PRIMER (Lincoln et al. unpublished results) with the same criteria described previously (Jacob et al. 1991). Allele sizes for D8Mcw1 were determined for 48 rat strains (data not shown). To map Sm22, the 46 progeny of an F 2 intercross between the Spontaneously Hypertensive rat (SHR) and the Brown Norway rat (BN) were genotyped as previously described (Jacob et al. 1995). These SHR and BN strains were from Michal Pravenec in the Czech Republic: SHR/Cz and BN/Lx, respectively. After genotyping, linkage analysis was performed with the MAPMAKER computer package (Lander et al. 1987) with the same criteria as previously described (Jacob et al. 1995). As shown in Fig. 1, D8Mcwl mapped to the same position as D8Mgh8 on the rat Chromosome (Chr) 8. We have previously identified a blood pressure QTL between D8Mit3 and D8Mit5 (Schork et al. 1995). More interestingly, Sm22 appears to be in a region (between D8Mit5 and

Immunohistochemical study of localization of γ-glutamyl transpeptidase in the rat brain

Abstract γ-glutamyl transpeptidase (γ-GTP) is a membrane-bound enzyme which is known to play a crucial role in active transport of amino acids across membrane barriers. We prepared a monoclonal antibody recognizing specifically rat γ-GTP and investigated localization of the enzyme in the rat brain by immunohistochemistry with this antibody. The antigen was localized on the ependyma, epithelia of the choroid plexus and microvessels. More precise localization of γ-GTP was examined with immuno-electron microscopy. The antigen was recognized on the microvilli and cilia of the ependymal cells, microvilli of the choroid epithelial cells and luminal membranes of the vascular endothelial cells.

Localization of the rat genes encoding glucagon, glucagon receptor, and insulin receptor, candidates for diabetes mellitus susceptibility loci

~D6partement de Biologic Mol6culaire, Universit6 Libre de Bruxelles, Rue des Chevaux, 67, B-1640 Rhode-St-Gen~se, Belgium aLaboratoire de Chimie Biologique, Facult6 de M6decine, Universit6 Libre de Bruxelles, B-1070 Bruxelles, Belgium 3Cardiovascular Research Center, Massachusetts General Hospital-East, 149 13 th Street, 4 t~ Floor, Charlestown, Massachusetts 02129, USA 4Department of Physiology, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, Wisconsin 53226-0509, USA

Gene structure and chromosomal mapping of the rat smooth muscle calponin gene

Abstract. Smooth muscle cells (SMC) express a battery of lineage-restricted genes whose encoded proteins impart the unique contractile phenotype that characterizes this muscle type. While the encoded function of many SMC-restricted genes has been extensively analyzed, less is known about their position within the genome and the regulatory factors governing their transcription. In this report, we define the gene structure, 5′ promoter analysis, and chromosomal mapping of the rat smooth muscle calponin (CnnI) gene. The rat CnnI gene is comprised of seven exons spanning approximately 8 kb of genomic sequence. The intron-exon boundaries of the rat CnnI gene match precisely those in human and mouse. Primer extension and RNase protection assays indicate two major transcription start positions (tsp). Comparative sequence analysis of the 5′ promoter region reveals several conserved cis regulatory elements, including a TA-rich element within 30 nt of the tsp that could be a recognition site for TATA-binding protein and two CCAAT boxes. Transient and stable transfection studies support the hypothesis that distal regulatory elements confer SMC-restricted expression of CnnI. Finally, using an F2 intercross, we have mapped the rat CnnI gene to the telomeric end of Chromosome (Chr) 8. These studies provide additional information relating to the control of CnnI gene expression and provide a platform to begin assessing the potential linkage of CnnI to spontaneous and experimental disease phenotypes in rats.

Localization of rat genes in the nitric oxide signaling pathway: candidates for the pathogenesis of complex diseases

Nitric oxide (NO) has critical regulatory function in multiple cell types. In the vasculature, NO is a vasodilator but also modulates smooth muscle cell proliferation (Schmidt and Walter 1994) and apoptosis (Pollman et al. 1996). In the nervous system, NO serves as a neurotransmitter, whereas, in endocrine tissues, NO has important roles in the regulation of hormone release, such as insulin and renin (Schmidt and Walter 1994). Alterations in NO production and responsiveness could contribute to the pathogenesis of several complex human diseases, including hypertension, diabetes mellitus, and renal failure. Binding of NO to soluble guanylate cyclase (sGC) stimulates the production of cGMP, an intracellular second messenger, which is thought to mediate many of the biological functions of NO. sGC is a heterodimer composed of a andb subunits, and there are two of each isoform encoded in the rat genome, a1 a2, b1, andb2 (GUCY1A1, GUCY1A2, GUCY1B1, and GUCY1B2, respectively; Wong and Garbers 1992). The biological effects of cGMP are mediated, in part, by cGMP-dependent protein kinase (PRKG). There are two PRKG genes encoded in the mammalian genome, one of which, type I (PRKG1), has two isoforms which are the products of alternative splicing of the first exon (Ørstavik et al. 1997). In the last decade, genetic mapping studies for complex traits have identified several quantitative trait loci (QTLs) with various rat strains as animal models for human diseases. With respect to NO synthesis, genes for three isoforms of nitric oxide synthase (NOS), neuronal ( Nos1), inducible (Nos2), and endothelial ( Nos3), were mapped, and possibilities for candidate genes were addressed (Deng et al. 1995; Deng and Rapp 1995, 1997; Hu ̈bner et al. 1995). In this study, since the pathway identification is a principal strategy for finding genes contributing to complex human traits, genetic markers for rat genes involved in NO responsiveness, Gucy1b1and Prkg1, were developed. Comparative mapping between rat and human was performed by radiation hybrid (RH) mapping and fluorescence in situ hybridization (FISH) studies. These genes mapped to regions containing QTLs linked to hypertension, non-insulin-dependent diabetes mellitus, and renal failure, suggesting that these genes are attractive candidates contributing to the pathogenesis of these diseases (Deng and Rapp 1992; Harris et al. 1995; Schork et al. 1995; Galli et al. 1996; Gauguier et al. 1996; Brown et al. 1996). To generate genetic markers for P kg1andGucy1b1,genomic fragments containing a (CA/GT)n repeat within/near these genes were identified with the two-step strategy. First, using cDNAs for PRKG1 and GUCY1B1, rat genomic libraries (Stratagene for Prkg1 and Clontech forGucy1b1) were screened to isolate genomic fragments containing these genes. A cDNA for PRKGI was generated by reverse transcription followed by polymerase chain reaction (RT-PCR) from rat brain mRNA with degenerate primers (forward primer: 58-CCGAATTCAGGAGCATGGGCACCYTGCG-38; reverse primer: 5 8-CCGGATCCTTTATRAGATCCTTGGA-38) corresponding to amino acids shared by human and bovine PRKG1s (Sandberg et al. 1989; Wolfe et al. 1989). A cDNA for GUCY1B1 was obtained from Dr. Masaki Nakane (Nakane et al. 1990). Second, these genomic clones were screened with (CA)15 and (GT)15 oligonucleotides to isolate genomic fragments containing a (CA/GT)n repeat followed by sequencing. Then, primers flanking a (CA/GT)n repeat within/near these genes were designed by the computer program PRIMER (Lincoln, unpublished results) with the same criteria described previously (Jacob et al. 1991): D1Mgh25 (Prkg1): forward primer 58-CTTCACCACTAATAACTAACCC-38, reverse primer 5 8-AGAGGAGTGGAAGTTGGG-38; D2Mgh17(Gucy1b1): forward primer 58-AAAGCGACAGAGGAATGTTCA-38, reverse primer 5 8GGAATTCAGATGGGCTCAGA-38). Allele sizes for these markers were determined for 12 rat strains (Jacob et al. 1995), identifying that these markers are informative between the Spontaneously Hypertensive rat (SHR) and the Brown Norway rat (BN, data not shown). Then, the 46 progeny of an F 2 intercross between SHR and BN were genotyped as previously described (Jacob et al. 1995). These SHR and BN strains were derived from the colonies of Dr. Michal Pravenec in the Czech Republic: SHR/Cz and BN/ Lx, respectively. After genotyping, linkage analysis was performed with the MAPMAKER computer package (Lander et al. 1987) with the same criteria as previously described (Jacob et al. 1995). As shown in Fig. 1A,Prkg1 (D1Mgh25) mapped between D1Mit6 and D1Mit7 on the rat Chr 1. Interestingly, it has been demonstrated that this region contains QTLs for non-insulindependent diabetes mellitus (NIDDM) (Galli et al. 1996; Gauguier et al. 1996) and renal failure (Brown et al. 1996). Since NO is known to modulate pancreatic insulin secretion and renal homeostasis (Schmidt and Walter 1994), abnormalities in NO responsiveness could contribute to the pathogeneses of NIDDM and renal failure. Therefore, Prkg1is a shared candidate for the pathogenesis of NIDDM and renal failure. Previously, the cytogenetic position of the human PRKG1 was identified (Ørstavik et al. 1992) at the human Chr 10q11.2. To determine the precise position of PRKG1 in the human genome, we designed PCR primers (forward primer: 5 8-TTACTATGGTACAGAAACTGGGC-38, reverse primer: 5 8-ATCATGAAACTTCTAGCACTGGT-38) based on the published sequence of the PRKGI cDNA (Sandberg et al. 1989), and the RH mapping was carried out with Gene Bridge 4 panel. PRKG1 mapped to 4.4 Correspondence to: G. Koike in Milwaukee. Mammalian Genome 10, 71–73 (1999).