Genetic Dissection of a Blood Pressure Quantitative Trait Locus on Rat Chromosome 1 and Gene Expression Analysis Identifies SPON1 As a Novel Candidate Hypertension Gene
A region with a major effect on blood pressure (BP) is located on rat chromosome 1. We have previously isolated this region in reciprocal congenic strains (WKY.SHR-Sa and SHR.WKY-Sa) derived from a cross of the spontaneously hypertensive rat (SHR) with the Wistar-Kyoto rat (WKY) and shown that there are 2 distinct BP quantitative trait loci, BP1 and BP2, in this region. Sisa1, a congenic substrain from the SHR.WKY-Sa animals carrying an introgressed segment of 4.3Mb, contains BP1. Here, we report further dissection of BP1 by the creation of 2 new mutually exclusive congenic substrains (Sisa1a and Sisa1b) and interrogation of candidate genes by expression profiling and targeted transcript sequencing. Only 1 of the substrains (Sisa1a) continued to demonstrate a BP difference but with a reduced introgressed segment of 3Mb. Exonic sequencing of the 20 genes located in the Sisa1a region did not identify any major differences between SHR and WKY. However, microarray expression profiling of whole kidney samples and subsequent quantitative RT-PCR identified a single gene, Spon1 that exhibited significant differential expression between the WKY and SHR genotypes at both 6 and 24 weeks of age. Western blot analysis confirmed an increased level of the Spon1 gene product in SHR kidneys. Spon1 belongs to a family of genes with antiangiogenic properties. These findings justify further investigation of this novel positional candidate gene in BP control in hypertensive rat models and humans.
Blood pressure (BP) and hypertension exhibit a significant degree of genetic heritability.1 In the past 15 years, many genomic locations contributing to hypertension in rodent models have been mapped.2 Most of these regions, known as quantitative trait loci (QTLs), are quite large, typically 10 to >80 million base pairs (Mb). This has hampered the ability to identify the gene(s) responsible for the blood pressure effects underlying these QTLs. Constructing congenic strains has proved pivotal in reducing the genomic size of QTLs: the strategy creates animal strains where the native genomic background (eg, from a normotensive rat strain) remains unchanged except for an introgressed (replaced) specific genomic region of interest which is transferred from a contrasting strain (eg, from a hypertensive rat strain).2 The integration of genetic strategies with expression profiling and targeted sequencing3–6 has also proved a successful approach in accelerating the search for genes underlying various phenotypes related to cardiovascular disease. This integration strategy is based on the assumption that genes that are both differentially expressed or show structural variation and map to a disease-related QTL are likely to be involved in the pathophysiology of that disease.
In a previous study, we mapped a BP QTL region to rat chromosome 1 in F2 rats derived from a cross of the spontaneously hypertensive rat (SHR) with the Wistar-Kyoto rat (WKY). This QTL region was then isolated in reciprocal congenic strains WKY.SHR-Sa (WConSa) and SHR.WKY-Sa (SConSa) derived from these animals.7 In more recent work,8 we have dissected this region further by developing congenic substrains and demonstrated that at least 2 distinct QTLs (BP1 and BP2) affecting BP reside within this region of rat chromosome 1. The substrain Sisa1 (WKY.SHR-D1Rat420/D1Got161) contained BP1, the smallest introgressed segment (4.3Mb) that still demonstrated a BP effect, with BP2 mapping to a separate portion of the introgressed region captured within WConSa.8 In other work using our congenic strains, we have shown by cross-transplantation studies that the BP effect of these loci on rat chromosome 1 is mediated through the kidney9; WKY recipients of a WConSa kidney were found to have significantly higher BP at all time points when compared with recipients of a WKY kidney. The purpose of the present study was to carry out further genetic dissection to provide better physical resolution of the BP QTL in Sisa1 and to undertake both targeted transcript sequencing and microarray expression analyses to identify positional candidate genes.
Materials and Methods
Construction and Genotyping of Novel Congenic Substrains of Sisa1
The breeding strategy followed to produce the novel congenic substrains was similar to the method used to produce previous congenic strains8 and further details are given in the online data supplement available at http://circres.ahajournals.org. To determine the size of the introgressed segments in each new strain, animals were subsequently genotyped for all the available polymorphic markers mapped to the original Sisa1 region,8 as well as additional microsatellite markers identified from the latest Rat genome assembly (v3.4) using the simple tandem repeat track (located by the Tandem Repeat Finder program10) within the UCSC Genome browser (http://genome.ucsc.edu/). These markers are described in supplemental Table I in the online data supplement at http://circres.ahajournals.org. All genotyping was performed as described previously.8
BP was measured in male animals using the same tail plethysmography method we used previously in the mapping of the QTLs on rat chromosome 1 and analysis of the initial congenic strains.7,8 Further details are given in the online data supplement available at http://circres.ahajournals.org. All procedures were performed in accordance with our institutional guidelines.
For each gene, the longest transcript model according to the Ensembl database (http://www.ensembl.org/index.html) was chosen for targeted exon sequencing. All Ensembl genes were cross checked with the respective RefSeq (http://www.ncbi.nlm.nih.gov/RefSeq/) gene model and where differences were found these were only in the lengths of the 5′ and 3′ UTR sequences and not in the coding sequences. Primers were designed to span the exons. For genes expressed in the kidney sequencing was performed from cDNAs. Further details are given in the online data supplement available at http://circres.ahajournals.org. For genes not expressed in the kidney putative exons were amplified directly from genomic DNA and sequenced. All sequences were obtained from a minimum of 2 SHR and 2 WKY animals. Sequences were aligned and analyzed using Sequencher software (http://www.genecodes.com/).
Gene Expression Profiling and Integration With Genetic Mapping
Affymetrix GeneChip RAE230 expression analysis was used to look for differentially expressed probe sets between SHR, WKY, WConSa, SconSa, and Sisa1. Total RNAs extracted from whole kidneys of 5 separate rats from each strain at both 6 and 24 weeks of age were examined. Further details are given in the online data supplement. We identified all probe sets from the RAE230 chip set that mapped within the boundaries of the Chromosome 1 QTLs of interest from Affymetrix GeneChip NetAffx Analysis Centre database (http://www.affymetrix.com/analysis/index.affx) and Ensembl. Genes from the Sisa1 region that were not represented on the RAE230 microarray chip were investigated using semi-quantitative RT-PCR.
Validation of Microarray Gene Expression Data
To validate differences in mRNA expression detected on the microarray we assessed mRNA abundance by quantitative RT-PCR. PCR was performed by TaqMan assay using a 7900HT (Applied Biosystems). All reactions were performed in duplicate alongside negative controls. Gene expression was measured relative to the endogenous control β-2-microglobulin. Taqman assays were also used to quantify and compare Spon1 mRNA levels in aorta and hearts of SHR WKY and Sisa1 rats.
Proteins were extracted from 6 week and 24 week kidneys of WKY, SHR, and Sisa1 animals with RIPA buffer in the presence of protease inhibitors. Equal amounts of protein (15 μg) were electrophoresed on 8% polyacrylamide gels under denaturing conditions and transferred onto nitrocellulose membranes. Western blots were incubated with an anti-SPON1 polyclonal antibody (Abcam), followed by a horseradish peroxidase- (HRP-) conjugated anti-chicken secondary antibody. As a loading control, membranes were stripped and reprobed with an anti-β-actin monoclonal antibody (Abcam), followed by a horseradish peroxidase- (HRP-) conjugated anti-mouse secondary antibody. An enhanced chemiluminescence reaction resulted in visualization of the protein bands.
DNA Sequencing of Spon 1 Gene Promoter and Exon-Intron Junctions
Genomic DNA extracted from 2 WKY and 2 SHR rats was amplified by PCR using primers designed from rat genome sequence in the Ensembl database. The primers used for the promoter region mapped to the first 5 kbp of genomic sequence upstream of the rat Spon1 gene (Ensembl ID=ENSRNOG00000034303). The primers used for the exon-intron junctions of the Spon1 gene mapped to the flanking introns of each exon. Further details of these primers (supplemental Table III) and the PCR and sequencing protocols are given in the online data supplement. Sequences were aligned and analyzed using Sequencher software (http://www.genecodes.com/) and Spon1 data from Ensembl as a reference sequence. All polymorphisms were confirmed with DNAs from additional animals.
To compare the mean BP of each strain, a one-way ANOVA and Bonferroni multiple comparison test was performed. For microarray expression analysis, the MAS 5.0 statistical algorithm (Affymetrix) was used to obtain an absolute expression level metric for each probe set. All subsequent data analysis was undertaken using Genespring (v7.2, Silicon Genetics). Data were normalized per chip using the distributions of all genes. In addition we applied a per gene normalization using the median of each gene. Statistical significance of expression changes across the strains was computed within GeneSpring using a 1-way ANOVA test (P<0.01) on an expressed gene list. This expressed gene list was derived by filtering for probe sets that were called “Present” in 3 of the 25 samples (5 replicates per strain) from the same age group. The parameter tested being the 5 strains with a parametric test (assume variances equal) and the Benjamini and Hochberg procedure to control false discovery rates. The Student-Newman-Keuls Post Hoc test was used to indicate the strain pairs where statistical differences occur. For the quantitative RT-PCR experiments, δCT values were averaged across biological replicates. The δδCT values for each strain were then expressed as fold change using the 2−δδCT method. Statistical analysis of quantitative RT-PCR and Spon1 protein analysis was undertaken using the Students t test with a significance level of P=0.05.
This data set has been deposited in the ArrayExpress microarray data public repository (http://www.ebi.ac.uk/arrayexpress/) with the accession number E-TABM-45 (release date January 1, 2006).
Dissection of the Sisa1 Congenic Region
Two mutually exclusive substrains SHR.WKY-D1Got158/D1Got161 (Sisa1a) and SHR.WKY-D1Rat420/D1Rat278 (Sisa1b) were produced that had inherited different segments of the Sisa1 congenic region (Figure 1). The original Sisa1 congenic region was effectively dissected into 2 regions: Sisa1a (3 Mb in length) containing 20 genes and Sisa1b (1.1 Mb) containing 12 genes (supplemental Table II). All the genes from the Sisa1 region have been captured in these 2 substrains apart from a novel gene [ENSRNOG00000026767] which is positioned in a gap between the Sisa1a and Sisa1b regions. Furthermore, with additional markers (supplemental Table I), uncertainty as to the position of the external boundaries of the congenic regions (Figure 1) has been reduced, to a minimal size (<60kb) that excludes the residence of any genes in these regions.
BP in Congenic Substrains
The mean BP measurements of Sisa1a and Sisa1b animals at 16 and 20 weeks are shown in Figure 2 as well as the BPs of Sisa1 and SHR animals studied at the same time. As expected from previous results, Sisa1 animals had a significantly lower BP compared with SHR animals with differences of ≈11mmHg and 9mmHg at 16 and 20 weeks respectively (P<0.001). The Sisa1a substrain animals also had a significantly lower BP compared with SHR (P<0.001 with differences of ≈8mmHg and ≈6.5mmHg respectively at the 2 ages) and exhibited a BP phenotype that was not significantly different to that of Sisa1. However, the Sisa1b animals showed the converse, with an average BP similar to that of SHR but significantly different in comparison to Sisa1 animals (P<0.001).
Exon Sequencing of Genes Located in Sisa1a Region
The 20 genes in the Sisa1a region include 17 genes with functional annotation and 3 other (novel) genes (Table). Of the known genes, 16 could be amplified from WKY and SHR kidney RNA and cDNA sequencing was used to compare WKY and SHR sequences. The parathyroid hormone precursor was not expressed in the kidney and was not sequenced. No detectable transcripts were observed from the 3 novel genes in the Sisa1a region either, but the putative exons of these genes were amplified and sequenced from genomic DNA. In total, almost 30kb of exonic sequence was compared between the strains.
The findings from the transcript sequencing experiments are summarized in Table 1. There was no detectable sequence variation between the SHR and WKY transcripts in the majority of the genes. Synonymous single nucleotide polymorphisms (SNPs) were found in K+ channel protein, R-RAS2 and 1 of the novel genes ENSRNOG00000025393. A synonymous and nonsynonymous SNP was discovered in the RSB-11-77 gene, with the latter resulting in a methionine to valine substitution in the SHR. The cytoskeleton associated-2 gene also had a nonsynonymous SNP, causing an arginine to histidine change at residue 142 in the SHR. Interestingly, for the CASL gene both the WKY and SHR strains produced an identical truncated open reading frame when compared with the reference (Brown-Norway) sequence in the Ensembl database. There was also a polymorphism in the 3′UTR region of this gene. For the Spon1 gene we sequenced the entire known 5′ and 3′ UTR sequences. Spon1 had 3 synonymous SNPs and a polymorphism in the 3′UTR region.
Gene Expression Analysis
To identify strong positional candidates from the microarray expression profiling we used a strict test regime, the genes had to show significant and consistent differential expression between the SHR and WKY genotypes across all 5 strains and at both ages. Sixteen of the Sisa1a genes were represented on the microarray chip and for those genes not represented on the chip, semi-quantitative RT-PCR was used to look for expression differences. Where renal expression was observed in these RT-PCR experiments, no interstrain differences were observed.
The only gene from the Sisa1a region to pass all the microarray tests for significance and maintained differential expression across all the strains and at both ages was Spon1 (Ensembl identifier ENSRNOG00000034303), otherwise known as F-Spondin. The expression profile of Spon1 across the strains was validated by quantitative RT-PCR (Figure 3a). At both 6 and 24 weeks of age, relative Spon1 mRNA levels in the kidney of the WKY rat were significantly lower compared with the SHR. Spon1 mRNA levels were also significantly lower in Sisa1 animals compared with SHR; at 6 weeks of age, Spon1 mRNA level in Sisa1 animals were intermediate between SHR and WKY but at 24 weeks the Sisa1 Spon1 mRNA level was similar to that of WKY.
During the microarray analysis, data relating to BP2 (the second BP QTL captured within the introgressed segment of WConSa but distinct from the QTL region of Sisa1)8 was also examined. In total, 853 of the 1273 genes within the WConSa region were represented on the microarray chip and of these, 3 further genes in addition to Spon1 were identified as being differentially expressed using the criteria described above; Homer-2 (Ensembl identifier ENSRNOG00000019297), XP_215009 (Ensembl identifier ENSRNOG00000018042) and Thumpd1 (Ensembl identifier ENSRNOG00000014946). Of these, Thumpd1 maps to BP2. Differential expression of these genes was also confirmed by quantitative RT-PCR (data not shown).
Expression of Spon1 in Kidneys of Sisa1a and Sisa1b Subcongenic Strains
In additional experiments, we analyzed Spon1 mRNA levels in kidneys of Sisa1a and Sisa1b subcongenic strains. Contemperarous SHR, WKY, and Sisa1 animals were also studied. The data for 6 week animals are shown in Figure 3b. Sisa1a which still carries the WKY Spon1 allele showed an expression level similar to WKY and Sisa1 whereas in Sisa1b the level had increased back to the SHR level. Similar results were obtained at 24 weeks (data not shown).
Expression of Spon1 in Vascular Tissues
We also examined Spon1 expression in the heart and aorta (Figure 4). Both tissues expressed Spon1. Comparing Ct and δ Ct values from the quantitative PCR across the strains and in different tissues indicated that Spon1 mRNA expression levels are similar in kidney and heart, and approximately 4-fold lower in aorta. Importantly, both vascular tissues showed the same inter-strain profile of Spon1 expression as the kidney, with a lower level in Sisa1 compared with the SHR supporting the genetic regulation of Spon1 expression.
Analysis of F-Spondin Protein Levels
F-Spondin is proteolytically cleaved in vivo to yield a 60kDa fragment containing the reelin and spondin domains, and a 50kDa fragment containing the 6 C-terminal thrombospondin-type 1 repeat (TSR) domains. The anti-F-Spondin antibody used in these experiments specifically recognizes the TSR domains and thus detected a 50kDa protein in rat kidney lysates (Figure 5). When measured relative to β-actin, F-Spondin protein levels were significantly higher (1.7-fold) in 6 week SHR than in 6 week old WKY animals, but were not significantly different between WKY and Sisa1 kidney samples. A similar trend was observed in 24-week kidney samples, where F-Spondin expression was 3.6-fold higher in SHR than in WKY. However, F-Spondin levels were 1.8-fold higher in Sisa1 than in WKY although significantly lower (1.8-fold) when compared with SHR animals at 24 weeks.
Comparison of Spon1 Promoter and Exon-Intron Junctions of SHR and WKY
Three single nucleotide polymorphisms (SNPs) were identified between WKY and SHR in the first 5 kbp of sequence upstream of Spon1: a G/A polymorphism (WKY/SHR) at position −2009 relative to the translation start site, a G/A polymorphism at position −4183, and a T/C polymorphism at position −4639. Using the VISTA Genome browser, the 5 kbp of rat sequence upstream of Spon1 was aligned with orthologous regions in human and mouse. This analysis identified several conserved regions (70% sequence conservation over a 100bp window), 1 of which contains the SNP at position −4183 (supplemental Figure I). A comparative sequence analysis, of all exon-intron junctions for the 19 exons of the Spon1 gene showed only 2 intron polymorphisms (supplementary Table III) between SHR and WKY that were 30 and 58 bases from their respective intron-exon junctions. Neither of these created an alternative splice site according to the known mammalian consensus sequences at the 5′ and 3′ splice sites.
We have used a combination of creating congenic strains with higher resolution introgressed segments, strong prior evidence from transplant experiments that the QTL effect resides in the kidney,9 and targeted transcript sequencing and microarray gene expression analysis, to identify putative hypertension candidate genes underlying a BP QTL on rat chromosome 1.
There are several notable findings from the work. The fact that the BP effect seen in Sisa1 could be dissected out with retention of the QTL in the Sisa1a strain and reversion of the phenotype in the Sisa1b strain to the SHR phenotype provides strong evidence that the BP effect is directly because of genomic alterations in the Sisa1a region. We investigated the possibility that the effect may be because of a structural variation in one of the proteins coded by the genes located in Sisa1a. However, this search identified only 2 nonsynonymous changes: one in RSB-11-77 that caused a methionine to valine change at position 186 in the SHR and the other in the Cytoskeleton associated-2 gene leading to an arginine to histidine substitution at position 142. Both types of changes have been reported to alter function in other proteins.11,12 However the respective changes do not occur in any known structural or functional motifs in these genes.
Our most interesting findings came from the microarray gene expression analysis. For this, we set strict criteria whereby the candidate genes not only had to show consistent genotype-related expression differences across several strains but also exhibit differential expression at two chronological points of 6 and 24 weeks of age. This was to identify changes that were likely to be genetically determined and to exclude changes occurring only at the later age, which could be secondary to blood pressure differences. The focus on renal expression was as a result of our previous work, where we observed that the major proportion of the BP effect from the loci on rat chromosome 1 is mediated through the kidney.9 The microarray studies highlighted Spon1 as a gene of particular interest from the Sisa1 region; it was the only gene that showed genotype-related differences in expression at both ages. Thus, compared with WKY, mRNA levels of Spon1 were higher not only in SHR but also in WConSa, which exhibits a SHR genotype for the chromosomal region in which Spon1 resides. In contrast, Spon1 expression levels in the SConSa and Sisa1 congenic strains, which contain the WKY genotype for Spon1 in an SHR background, were similar to the WKY strain. Furthermore, the Sisa1 subcongenic strains (Sisa1a and Sisa1b) showed differences in expression corresponding to the Spon1 allele they carried. Moreover, the Spon1-encoded F-Spondin protein was expressed at a significantly higher level in the kidney of SHR than WKY and Sisa1 rats at both 6 and 24 weeks of age. The findings as a whole provide persuasive evidence that Spon1 expression is genetically regulated and tracks with BP.
Spon1 has direct orthologues in many species including mouse and human, indicating that it is a highly conserved gene in evolutionary terms. The Spon1 orthologue in human has been termed “vascular smooth muscle cell growth-promoting factor” and shown to inhibit angiogenesis via the blockade of integrin receptors on human vascular endothelial cells.13 This is consistent with a large body of research demonstrating that the thrombospondins (proteins with the thrombospondin domain), to which F-spondin belongs, have an antiangiogenic activity.14 Interestingly, it has been demonstrated that the SHR strain has a decreased angiogenic potential compared with the WKY rats and this observation, together with other evidence, has led to the hypothesis that hypertension in the SHR could be caused by an impaired capacity for vascular growth from an early developmental stage.15 In this context, our finding of differential expression of Spon1 in vascular tissue which appears to be genetically regulated is notable. Indeed, it has been shown that the narrowed lumen of distal afferent arterioles in the SHR kidney at 7 weeks is related to later development of increased blood pressure.16 Therefore, it is possible that Spon1 overexpression (leading to increased F-Spondin protein levels) could affect blood pressure in the SHR by influencing blood vessel formation in the kidney of the developing animal.
The genetic mechanism underlying the differential expression of Spon1 between SHR and WKY genotypes remains to be elucidated. Our extensive sequencing of the promoter region of Spon1 has identified three single nucleotide polymorphisms (SNPs) between the strains. These provide targets for further analysis as does the SNP identified in the 3′UTR region (Table 1). Also, to elucidate whether there were Spon1 differential splice variants between the SHR and WKY strains, we sequenced the exon-intron junctions of all 19 exons and did not detect any splice site variants (supplemental Table III).
Although not a primary aim of this study, the microarray data also allowed us to look for genes that could underlie the other BP QTL, BP2, in this region of chromosome 1.8 Thumpd1, located within this region, showed a consistent genotype-related pattern of expression in the various parental and congenic strains that would make it a putative candidate gene for the second BP QTL. Interestingly, another study6 has recently also confirmed expression differences in Thumpd1 between hypertensive and normotensive strains and suggested that this is a candidate gene for hypertension. Functional annotation on this gene is sparse. The encoded protein contains an ATP/GTP-binding site motif as well as a THUMP domain that is involved in RNA-binding. We also identified 2 other genes that are located in the larger WConSa congenic region that showed expression differences between the strains; the first, Homer-2, encodes a protein that binds to the cytoplasmic regions of metabotropic glutamate receptors mGluR1 and mGluR5 and appears to aid the coupling of surface receptors to cause intracellular calcium release in neurons.17 The second gene, XP—215009, has little functional information attached to it. However, neither maps to one of the narrowed BP QTL regions. In addition, we observed the differential expression of the Sa gene, which has previously received considerable attention as a hypertension candidate gene18 between the SHR and WKY genotypes of the 5 strains examined. However, recent research19 using gene targeting has shown that the Sa gene is unlikely to play a role in blood pressure regulation.
Although our approach has successfully identified Spon1 as a strong positional candidate gene for one of the BP QTLs on rat chromosome 1, further work is clearly required to confirm its candidacy, including examining the effect on BP and other phenotypes of altering its level of expression through gene modification techniques. In assessing the present findings, limitations of our interrogation of the Sisa1a region need to be borne in mind. Specifically, the microarray chips used in this study, and previous studies like this, are not capable of picking up differential expression of alternative spliceforms. Therefore, currently we cannot exclude the possibility that an undetected difference in level of a specific splice variant in one of the Sisa1a genes apart from Spon1 is responsible for the BP effect. Second, we have focused on the genes located in the region. Increasing evidence is emerging that intergenic regions may influence phenotypic traits. If further interrogation of Spon1 does not explain the BP QTL effect, then these possibilities will need to be considered.
In summary, through a combination of creating novel congenic strains with higher resolution introgressed segments, targeted transcript sequencing and microarray gene expression profiling, we have identified Spon1 as a novel strong candidate gene underlying a major BP QTL on rat chromosome 1. If further studies confirm its role in BP regulation, then examining its role in BP homeostasis in humans and risk of hypertension will be of great interest.
Michael Kaiser for help with the exon-intron sequencing.
Sources of Funding
The work was supported by the Wellcome Trust Functional Genomics Initiative in Cardiovascular Genetics (Grant Ref 066780). N.J.S. holds a British Heart Foundation Chair of Cardiology.
↵*These authors contributed equally to this work.
Original received July 25, 2006; revision received February 16, 2007; accepted February 20, 2007.
Lifton RP. Molecular genetics of human blood pressure variation. Science. 1996; 272: 676–680.
Rapp JP. Genetic analysis of inherited hypertension in the rat. Physiol Rev. 2000; 80: 135–172.
Aitman TJ, Glazier AM, Wallace CA, Cooper LD, Norsworthy PJ, Wahid FN, Al-Majali KM, Trembling PM, Mann CJ, Shoulders CC, Graf D, St Lezin E, Kurtz TW, Kren V, Pravenec M, Ibrahimi A, Abumrad NA, Stanton LW, Scott J. Identification of Cd36 (Fat) as an insulin-resistance gene causing defective fatty acid and glucose metabolism in hypertensive rats. Nat Genet. 1999; 21: 76–83.
Yagil C, Hubner N, Monti J, Schulz H, Sapojnikov M, Luft FC, Ganten D, Yagil Y. Identification of hypertension-related genes through an integrated genomic-transcriptomic approach. Circ Res. 2005; 96: 617–625.
Hinojos CA, Boerwinkle E, Fornage M, Doris PA. Combined genealogical, mapping, and expression approaches to identify spontaneously hypertensive rat hypertension candidate genes. Hypertension. 2005; 45: 698–704.
Hubner N, Wallace CA, Zimdahl H, Petretto E, Schulz H, Maciver F, Mueller M, Hummel O, Monti J, Zidek V, Musilova A, Kren V, Causton H, Game L, Born G, Schmidt S, Muller A, Cook SA, Kurtz TW, Whittaker J, Pravenec M, Aitman TJ. Integrated transcriptional profiling and linkage analysis for identification of genes underlying disease. Nat Genet. 2005; 37: 243–253.
Frantz SA, Kaiser M, Gardiner SM, Gauguier D, Vincent M, Thompson JR, Bennett T, Samani NJ. Successful isolation of a rat chromosome 1 blood pressure quantitative trait locus in reciprocal congenic strains. Hypertension. 1998; 32: 639–646.
Frantz S, Clemitson JR, Bihoreau MT, Gauguier D, Samani NJ. Genetic dissection of region around the Sa gene on rat chromosome 1: evidence for multiple loci affecting blood pressure. Hypertension. 2001; 38: 216–221.
Clemitson JR, Pratt JR, Frantz S, Sacks S, Samani NJ. Kidney specificity of rat chromosome 1 blood pressure quantitative trait locus region. Hypertension. 2002; 40: 292–297.
Benson G. Tandem repeats finder: a program to analyze DNA sequences. Nucleic Acids Res. 1999; 27: 573–580.
Geffner ME, Su F, Ross NS, Hershman JM, Van Dop C, Menke JB, Hao E, Stanzak RK, Eaton T, Samuels HH, et al. An arginine to histidine mutation in codon 311 of the C-erbA beta gene results in a mutant thyroid hormone receptor that does not mediate a dominant negative phenotype. J Clin Invest. 1993; 91: 538–546.
Norrelund H, Christensen KL, Samani NJ, Kimber P, Mulvany MJ, Korsgaard N. Early narrowed afferent arteriole is a contributor to the development of hypertension. Hypertension. 1994; 24: 301–308.
Kato A, Ozawa F, Saitoh Y, Fukazawa Y, Sugiyama H, Inokuchi K. Novel members of the Vesl/Homer family of PDZ proteins that bind metabotropic glutamate receptors. J Biol Chem. 1998; 273: 23969–23975.
Walsh V, Somody L, Farrell A, Zhang B, Brown J, Pritchard C, Vincent M, Samani NJ. Analysis of the role of the SA gene in blood pressure regulation by gene targeting. Hypertension. 2003; 41: 1212–1218.