Meta-Analysis of Genome-Wide Gene Expression Differences in Onset and Maintenance Phases of Genetic Hypertension
Gene expression differences accompany both the onset and established phases of hypertension. By an integrated genome-transcriptome approach we performed a meta-analysis of data from 74 microarray experiments available on public databases to identify genes with altered expression in the kidney, adrenal, heart, and artery of spontaneously hypertensive and Lyon hypertensive rats. To identify genes responsible for the onset of hypertension we used a statistical approach that sought to eliminate expression differences that occur during maturation unrelated to hypertension. Based on this adjusted fold-difference statistic, we found 36 genes for which the expression differed between the prehypertensive phase and established hypertension. Genes having possible relevance to hypertension onset included Actn2, Ankrd1, ApoE, Cd36, Csrp3, Me1, Myl3, Nppa, Nppb, Pln, Postn, Spp1, Slc21a4, Slc22a2, Thbs4, and Tnni3. In established hypertension 102 genes exhibited altered expression after Bonferroni correction (P<0.05). These included Atp5o, Ech1, Fabp3, Gnb3, Ldhb, Myh6, Lpl, Pkkaca, Vegfb, Vcam1, and reduced nicotinamide-adenine dinucleotide dehydrogenases. Among the genes identified, there was an overrepresentation of gene ontology terms involved in energy production, fatty acid and lipid metabolism, oxidation, and transport. These could contribute to increases in reactive oxygen species. Our meta-analysis has revealed many new genes for which the expression is altered in hypertension, so pointing to novel potential causative, maintenance, and responsive mechanisms and pathways.
- spontaneously hypertensive rat
- Lyon hypertensive rat
- gene expression arrays
- reactive oxygen species
Essential hypertension (EH) is considered to arise from common risk alleles of small effect size,1 as well as rare risk alleles of large effect size. Most probably influence gene expression rather than structure of the encoded protein.2 These factors and the heterogeneity of EH have made identification of causative genes a challenge. Genetically pure strains of hypertensive rats are less likely to experience such difficulties. Here we performed a meta-analysis that combined existing data on global gene expression differences in relevant tissues of 2 models of genetic hypertension, the spontaneously hypertensive rat (SHR) and Lyon hypertensive (LH) rat, and controls to improve power and smooth over inconsistencies between published studies. This enabled a more robust identification of global expression differences in the onset and maintenance phases of genetic hypertension in these rat models and helped identify common molecular signatures of increased blood pressure (BP).
Data Collection and Organization
Data collection was based on steps described in the literature,3 taking particular care to obtain raw data. Global gene expression patterns, identified by microarray data sets, were available in public repositories (the Gene Expression Omnibus4 and the ArrayExpress5). To obtain insights into the onset of hypertension, data for prehypertensive (hypertensive strains before development of hypertension defined as <6 weeks of age6) were examined. The analyses were preformed on multiple tissue types and were confined to males because of paucity of female data. Inclusion criteria incorporated a suitable genetic rat model of hypertension without significant comorbidities or treatment and appropriate normotensive control. Tables S1 and S2 summarize the studies included,7–12⇓⇓⇓⇓⇓ and Table S3 summarizes the studies excluded and the reason for exclusion (see the online Data Supplement at http://hyper.ahajournals.org).
Maintenance Phase of Hypertension
Six separate studies were included in the “maintenance phase of hypertension” analysis, encompassing a total of 58 samples (30 hypertensive and 28 normotensive) from 2 Affymetrix platforms, RAE230A and RGU34A (Table S1). Platforms were independently preprocessed and normalized using robust multi-array (RMA) analysis.13 Between platforms, normalization was also performed with only the intersection of Entrez IDs14 from the 2 platforms being retained in the analysis. Differentially expressed genes were identified by a moderated 2-sample t test.15 Differentially expressed genes were selected based on their Bonferroni adjusted P value with a significance cutoff level of 0.05, as well as examining false discovery rate at the 5% level. The Gene Ontology (GO) Database16 was used to further interpret the differentially expressed gene data set and to identify overrepresented functional groups of genes.
Onset Phase of Hypertension
Two new studies relating to “juvenile” (<6 weeks) prehypertensive and age-matched normotensive samples were used in this part of the analysis, leading to 16 juvenile samples (8 prehypertensive and 8 normotensive) from Affymetrix platforms RAE230A and RGU34A (Table S2). Samples from the maintenance phase of hypertension section of the analysis were termed “adult.” The juvenile samples were preprocessed as described for the adults.
Genes that were expressed differentially between the adult hypertensive samples and the juvenile prehypertensive samples were found by adjusting for age effects and avoiding direct nonage matched comparisons. This statistic was termed the adjusted fold difference (aFD) value. The aFD value removed the influence of age-specific differences between prehypertension and hypertension. Differentially expressed genes were selected based on an absolute aFD value >2.0. Overrepresentation GO analysis was performed on the set of differentially expressed genes.
We applied an integrated genome-transcriptome approach to identify those genes that could be involved in the pathogenesis of genetic hypertension, as suggested previously.17 More details involving the data collection and statistical analysis, including the preprocessing, differentially expressed selection, and aFD statistic, are available in the online Data Supplement.
Maintenance Phase of Hypertension
After Bonferroni correction (adjusted P<0.05), 102 genes exhibited expressions that differed significantly between hypertensive and normotensive samples (Table 1). (For more information see Table S4, available in the online Data Supplement). A less stringent multiple testing adjustment, by controlling for 5% false discovery rate, resulted in selection of 2170 genes (Table S5), including genes in the renin-angiotensin system (Agt, Agtr1, and Agtr1b) and genes for glutathione metabolism (Gstt1, Gstp1, and Gstm1). We decided to focus on those from the more conservative Bonferroni correction. Some of the differentially expressed genes were clustered at rat chromosome loci 1q21 (Ech1, Scn1b, and Sirt2), 1q36 (Cox6a2, Mapk3, and Tufm), 2q12 (Ak1, Ckmt2, and Homer1), 4q42 (Bcl2l13, Cosp7a, Gnb3, Tpi1, and Vhl), 7q34 (Cp, Cpt1b, Cyc1, Dgat1, Rpl3, and Syngr1), and 8q32 (Alas1, Cmtm6, Dync1li1, and Slc25a20; Table S4), all of which contain BP quantitative trait loci (QTLs) in rats.18 As verified in the Rat Genome Database,18 most (97%) of the genes identified were located in genomic regions suggested by QTL studies in rats as determinants of BP.
Among the differentially expressed genes, there was a significant overrepresentation of genes involved in energy production, such as generation of precursor metabolites and energy (GO:0006091), ATP (GO:0046034), oxidative phosphorylation (GO:0006119), ATP synthesis-coupled proton transport (GO:0015986), and mitochondrial transport (GO:0006839), among others. Pathways involving fatty acid and lipid metabolic processes were also overrepresented. Details, including adjusted P values, are shown in Table S6.
Onset Phase of Hypertension
Thirty six genes (Table 2) were found to be expressed differentially between hypertensive and prehypertensive samples based on an aFD value >2.0, where positive aFD values indicate higher expression in the hypertensive group and negative aFD values indicate higher expression in the prehypertensive group. This represents adjustment with age-matched controls (ie, genes that change with maturation were eliminated by our analysis). Most (97.2%) of the genes were located in QTL regions proposed previously as BP QTLs in rats18 (Table S7). The genes Nppa, Nppb, and Pla2g2a formed a chromosomal cluster at 5q36, a QTL identified previously for BP in rats. Among the differentially expressed genes, there was a significant overrepresentation of genes involved in circulatory system processes (GO:0003013), blood circulation (GO:0008015), and others. For details see Table S8.
This is the first meta-analysis (to our knowledge) of genome-wide gene expression in animal models of EH. It identifies genes for which the expression is altered in the maintenance, as well as the onset, phases of the condition. It is important to highlight that, with the exception of a few genes in specific studies, the genes reported here are largely new and have not been identified in each individual array study from which data were extracted. For this reason, these genes were not validated by quantitative real-time PCR in the initial studies, and the use of only validated genes instead of raw data for the analyses in the present meta-analysis would significantly reduce the size of the gene list obtained here, consequently compromising the ability to do a meta-analysis of genome-wide expression studies. The combination of biological meaning, results from previous association studies, as well as from genome-wide association studies (GWASs) and rat BP QTLs, and the use of statistical analyses that include adjustments by the conservative Bonferroni correction together support the validity of our findings. The genes Actn2, Cox6b2, and Csrp3 were seen in both the onset and maintenance phases of hypertension, and, in conjunction with the other genes below, are possible regulators of BP, thus meriting further study.
Maintenance Phase of Hypertension
Most genes in the maintenance phase of hypertension list have not been considered previously as genes for predisposition to hypertension. Their role is, however, supported by the higher protein levels found for some of them (Ak1,19 Atp5o,20 Hadha,20 and Ldhb21) in cardiac tissue of adult SHRs compared with Wistar Kyoto rats. There were no genes in common with those identified in human GWASs of EH and BP, although genes within the same families (Ak1, Aldh5a1, Ldhb, and Wdfy1; Table S9) were nevertheless identified. It should, however, be noted that human GWAS data account for only a small proportion of the likely genetic contribution to BP variation, although lowering the threshold for genome-wide significance would reveal more of the genes involved and, therefore, similarities with rat genes.22
Some of the genes that we list have been reported previously as candidate genes in other hypertension studies. These include Ech1,23 Vcam1,24 and Ucp325 in animal models. UCP3 polymorphisms have been associated with heart rate variability and family history of EH in young Japanese men.26 A direct association with EH has been described for the gene SDK1 in 2 independent Japanese samples.27 A meta-analysis of association studies of the polymorphism C825T in GNB3, involved in signal transduction, showed a small but significant increase in the risk for EH, especially after removal of heterogeneous studies.28 Variants of the gene LPL, of which the product is involved in the hydrolysis of triglycerides, were not only implicated in the development of EH but also in intermediate phenotypes associated with elevated plasma triglycerides levels.29 During hypertension, LPL activity also seems to be required for normal cardiac metabolic compensation.30 In patients newly diagnosed with EH, levels of FABP3, a marker of myocardial damage, were elevated and were correlated positively with arterial stiffness.31 In a Japanese population, the G allele of rs2279885 and C allele of rs2271072 in FABP3 were implicated as susceptibility loci for EH.32 The upregulation of Vegfb selectively increases angiogenesis in the ischemic myocardium,33 a known consequence of high BP. The genes Pdk1 and Pdk2 are also good candidates for hypertension because they are involved in the activation, by phosphorylation, of phosphatidylinositol 3-kinase/Akt, which has a role in vascular homeostasis and angiogenesis.34 Genes reported previously in hypertrophic cardiomyopathies, and which respond to the high BP by undergoing a change in expression, such as Csrp3, Myh6, and Prkaca, are also promising candidates (Table S10).
Changes in intracellular calcium concentration affect arterial smooth muscle contraction, so influencing BP.35 The genes Actn2 (GO:0005509), Csrp3 (GO:0006874), Dgkg (GO:0005509), Homer1 (GO:0051924, GO:0032236, and GO:0051928), Itpr1 (GO:0051480, GO:0006816, GO:0032469, and GO:0006874), and Pgm1 (GO:0055074) are all involved in calcium regulation or associated pathways and so represent new candidates for predisposition of hypertension.
The overrepresentation of GO terms for increased energy production and oxidative phosphorylation in hypertension are probably a result of the high expression of mitochondrial genes, such as ones for reduced nicotinamide-adenine dinucleotide dehydrogenases in complex I (Ndufa3, Ndufa8, Ndufb2, Ndufb4, Ndufb5, Ndufb6, Ndufb9, Ndufb10, and Ndufc1) of the electron transport chain involved in ATP production. Mitochondrial dysfunction in which there is increased electron leak from the electron transport chain will increase reactive oxygen species (ROS), known to be associated with hypertension,36 leading to a cycle of ROS-induced cell damage and oxidative stress.37 Mitochondrial dysfunction and increased ROS have been reported in failing myocardium37 and in the neurogenic component of hypertension in the SHR.38,39⇓
Onset Phase of Hypertension
The differential expression of certain genes identified here, either individually or in combination, could be the trigger for the initial increase in BP. When we compared the list of genes generated by our meta-analysis with GWAS results for EH and BP, the only locus in common was at ACTN2, whereas the genes NPPA and NPPB were close to a relevant locus (Table S9). Other evidence supports the importance of reduced natriuretic peptides A and B and gene expression in predisposition to hypertension. In a mouse model of hypertension, gene therapy increased plasma atrial natriuretic peptide–lowered BP.40 In humans, polymorphisms in natriuretic peptide A and natriuretic B genes are associated with natriuretic peptide levels and BP.41
A meta-analysis has shown that the ε4 allele of ApoE is associated with predisposition to EH.42 Deletion of ApoE increases systolic BP and cardiovascular disease risk.43 The gene Cd36 has been implicated in the risk of hypertension,44 atherosclerosis, and coronary heart disease.45 Expression of Spp1 was increased in peripheral blood and atherosclerotic lesions in aorta from EH patients.46 As for other genes that we have noted here, the product of Pln has been implicated in atherogenesis in humans,46 Egf encodes a vasoconstrictor,47 the Slc22a2 product affects renal dopamine and thereby sodium homeostasis and hypertension,48 Slc21a4 encodes a renal anion transporter, and Tnni3 serves to regulate sarcomeric function,49 where disorganization of sarcomeric structure has been observed in SHRs.50 Genes for hypertrophic cardiomyopathies (Ankrd1, Csrp3, Myl3, Pln, Postn, Thbs4, Tnni3, and Ttn) are also promising candidates (Table S10). Among the genes not thought previously to influence BP, Me1 is of interest because of its potential to increase NADPH production51 and thereby raise oxidative stress and ROS.
The present meta-analysis extends the power of the original microarray experiments (Table S2). Ours is distinct from the majority of meta-analyses that simply integrate and examine the same question as in the original studies. In the present analysis, specific care was taken to ensure that normalization was performed appropriately, and we validated our procedure by ensuring that the aFD values of known housekeeping genes were not altered (online Data Supplement). Furthermore, our analysis was successful in identifying biologically meaningful genes, such as ApoE, Cd36, Nppa, and Nppb.
There are differences in gene expression between tissues, strains, and ages that were not taken into account here. The literature contains, however, examples in which combining such data for different tissues52–54⇓⇓ and strains53 has led to successful outcomes. This helped support our decision to analyze SHRs and LH rats together to identify common molecular signatures of hypertension and so to get a better understanding of gene expression differences underlying hypertension in strains with different genetic backgrounds (in this way mimicking to some extent human EH). We found no apparent bias in results arising from the use of these different strains. Although some of the renin-angiotensin system genes were highlighted in the analysis using false discovery rate correction, the same did not happen when using Bonferroni correction, which is much more stringent, and these results were consistent when LH array data were removed from the analysis (data not shown). Also, one of the studies that we included compared global gene expression between SHRs and LH rats and found 215 differentially expressed genes to be in common between each strain,11 showing that the 2 different strains harbor common genes with altered expression in hypertension.
The use of public microarray data for the characterization of common pathophysiological features, irrespective of tissue, has, moreover, been validated in a meta-analysis, which showed that the molecular signature of a disease across different tissues was more prominent than the signature of individual tissue expressions across experiments measuring the same disease condition.54 Although it would be more precise to do a meta-analysis of each tissue separately, such subgroup analysis was not feasible in practice because of the small sample size within each subgroup. This is to be expected considering the high cost of microarray technology. So our decision to integrate samples from different tissues enabled us to detect the genes that we would otherwise have missed in subgroup analysis. We predicted that with appropriate normalization of the data sets and the conservative statistical correction of Bonferroni, many of the major genes involved would be similar between different tissues. Evidence from the literature, moreover, supports the combining of tissues for such studies.52–54⇓⇓ Added to this is the fact that most of the genes that we identified are known to influence cardiovascular parameters.
Most microarray results were available on databases. Nevertheless, there were some studies for which raw data were not available, as described in the online Data Supplement. In addition, some studies used controls that were not ideal, and these had to be excluded. Also, because of the complexity of hypertension, some of the genes identified here might be ones for which the expression is responding secondarily to the elevation in BP, rather than ones causing the BP rise. For this reason and for confirmation, further research will be required to validate the role of the genes identified here in the rat models and in EH patients. This challenging task is beyond the scope of the present study.
Our study highlights the potential value of a meta-analysis of available gene expression data from genetic models of hypertension for gene discovery in hypertension onset and maintenance. Various rat genes identified corresponded with genes, loci, and pathways arising from GWASs, BP QTL studies, association studies, and gene expression studies. The genes Actn2, Ankrd1, ApoE, Cd36, Csrp3, Me1, Myl3, Nppa, Nppb, Pln, Postn, Spp1, Slc21a4, Slc22a2, Thbs4, and Tnni3 were implicated in the onset phase of hypertension in the rat models studied, and the genes Atp5o, Ech1, Fabp3, Gnb3, Hadha, Ldhb, Myh6, Lpl, Pkkaca, Vegfb, Vcam1, and reduced nicotinamide-adenine dinucleotide dehydrogenase genes were altered in the maintenance phase of hypertension. Such gene expression differences can point to pathways, such as ones that result in increased ROS production, that might be involved. Our findings provide clues for follow-up research.
The present meta-analysis suggests, moreover, that animal models of EH may reveal genes exhibiting transcriptional differences that could in part be relevant to human EH. Microarray and other molecular studies in humans will be required to confirm our findings. Because of the heterogeneity of EH, we expect that large sample sizes and carefully chosen controls will be needed for such research. A particular challenge will be obtaining sufficient tissue sample numbers and types from a large enough number of EH patients and controls to show significant results. Although the kidney will be a prime tissue to study, obtaining samples from human kidneys and other tissues presents a challenge. New technologies may, therefore, be required to extend the findings to humans.
Sources of Funding
This work was supported in part by Australian Research Council grant DP0770395 (to Y.H.J.Y.), an Endeavour International Postgraduate Research Scholarship (to F.Z.M.), and an Australian Postgraduate Award (to A.E.C).
F.Z.M. and A.E.C. contributed equally to this work.
- Received April 23, 2010.
- Revision received May 16, 2010.
- Accepted June 3, 2010.
- ↵Delles C, McBride MW, Graham D, Padmanabhan S, Dominiczak AF. Genetics of hypertension: from experimental animals to humans [Published ahead of print December 24, 2009]. Biochim Biophys Acta. PMID: 20035862.
- ↵Stranger BE, Forrest MS, Dunning M, Ingle CE, Beazley C, Thorne N, Redon R, Bird CP, de Grassi A, Lee C, Tyler-Smith C, Carter N, Scherer SW, Tavare S, Deloukas P, Hurles ME, Dermitzakis ET. Relative impact of nucleotide and copy number variation on gene expression phenotypes. Science. 2007; 315: 848–853.
- ↵Ramasamy A, Mondry A, Holmes CC, Altman DG. Key issues in conducting a meta-analysis of gene expression microarray datasets. PLoS Med. 2008; 5: e184.
- ↵Barrett T, Troup DB, Wilhite SE, Ledoux P, Rudnev D, Evangelista C, Kim IF, Soboleva A, Tomashevsky M, Marshall KA, Phillippy KH, Sherman PM, Muertter RN, Edgar R. NCBI GEO: archive for high-throughput functional genomic data. Nucleic Acids Res. 2009; 37: D885–D890.
- ↵Parkinson H, Kapushesky M, Kolesnikov N, Rustici G, Shojatalab M, Abeygunawardena N, Berube H, Dylag M, Emam I, Farne A, Holloway E, Lukk M, Malone J, Mani R, Pilicheva E, Rayner TF, Rezwan F, Sharma A, Williams E, Bradley XZ, Adamusiak T, Brandizi M, Burdett T, Coulson R, Krestyaninova M, Kurnosov P, Maguire E, Neogi SG, Rocca-Serra P, Sansone SA, Sklyar N, Zhao M, Sarkans U, Brazma A. ArrayExpress update: from an archive of functional genomics experiments to the atlas of gene expression. Nucleic Acids Res. 2009; 37: D868–D872.
- ↵Lozinskaya IM, Cox RH. Effects of age on Ca2+ currents in small mesenteric artery myocytes from Wistar-Kyoto and spontaneously hypertensive rats. Hypertension. 1997; 29: 1329–1336.
- ↵Grayson TH, Ohms SJ, Brackenbury TD, Meaney KR, Peng K, Pittelkow YE, Wilson SR, Sandow SL, Hill CE. Vascular microarray profiling in two models of hypertension identifies caveolin-1, Rgs2 and Rgs5 as antihypertensive targets BMC Genomics. 2007; 8: 404.
- ↵Rysa J, Leskinen H, Ilves M, Ruskoaho H. Distinct upregulation of extracellular matrix genes in transition from hypertrophy to hypertensive heart failure. Hypertension. 2005; 45: 927–933.
- ↵Westhoff TH, Scheid S, Tolle M, Kaynak B, Schmidt S, Zidek W, Sperling S, van der Giet M. A physiogenomic approach to study the regulation of blood pressure. Physiol Genomics. 2005; 23: 46–53.
- ↵Cerutti C, Kurdi M, Bricca G, Hodroj W, Paultre C, Randon J, Gustin MP. Transcriptional alterations in the left ventricle of three hypertensive rat models. Physiol Genomics. 2006; 27: 295–308.
- ↵Clemitson JR, Dixon RJ, Haines S, Bingham AJ, Patel BR, Hall L, Lo M, Sassard J, Charchar FJ, Samani NJ. 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. Circ Res. 2007; 100: 992–999.
- ↵Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics. 2003; 4: 249–264.
- ↵Maglott D, Ostell J, Pruitt KD, Tatusova T. Entrez Gene: gene-centered information at NCBI. Nucleic Acids Res. 2007; 35: D26–D31.
- ↵Smyth GK. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004; 3: article 3.
- ↵Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G. Gene ontology: tool for the unification of biology–the Gene Ontology Consortium. Nat Genet. 2000; 25: 25–29.
- ↵Pravenec M, Wallace C, Aitman TJ, Kurtz TW. Gene expression profiling in hypertension research: a critical perspective. Hypertension. 2003; 41: 3–8.
- ↵Twigger SN, Shimoyama M, Bromberg S, Kwitek AE, Jacob HJ. The Rat Genome Database, update 2007: easing the path from disease to data and back again. Nucleic Acids Res. 2007; 35: D658–D662.
- ↵Graham D, McBride MW, Gaasenbeek M, Gilday K, Beattie E, Miller WH, McClure JD, Polke JM, Montezano A, Touyz RM, Dominiczak AF. Candidate genes that determine response to salt in the stroke-prone spontaneously hypertensive rat: congenic analysis. Hypertension. 2007; 50: 1134–1141.
- ↵Oguri M, Kato K, Yokoi K, Yoshida T, Watanabe S, Metoki N, Yoshida H, Satoh K, Aoyagi Y, Nozawa Y, Yamada Y. Assessment of a polymorphism of SDK1 with hypertension in Japanese Individuals. Am J Hypertens. 2010; 23: 70–77.
- ↵Chen P, Jou YS, Fann CS, Chen JW, Chung CM, Lin CY, Wu SY, Kang MJ, Chen YC, Jong YS, Lo HM, Kang CS, Chen CC, Chang HC, Huang NK, Wu YL, Pan WH. Lipoprotein lipase variants associated with an endophenotype of hypertension: hypertension combined with elevated triglycerides. Hum Mutat. 2009; 30: 49–55.
- ↵Yamashita H, Bharadwaj KG, Ikeda S, Park TS, Goldberg IJ. Cardiac metabolic compensation to hypertension requires lipoprotein lipase. Am J Physiol Endocrinol Metab. 2008; 295: E705–E713.
- ↵Gedikli O, Ozturk S, Yilmaz H, Baykan M, Kiris A, Durmus I, Baltaci D, Karahan C, Celik S. Relationship between arterial stiffness and myocardial damage in patients with newly diagnosed essential hypertension. Am J Hypertens. 2008; 21: 989–993.
- ↵Ueno T, Soma M, Tabara Y, Tokunaga K, Tahira K, Fukuda N, Matsumoto K, Nakayama T, Katsuya T, Ogihara T, Makita Y, Hata A, Yamada M, Takahashi N, Hirawa N, Umemura S, Miki T. Association between fatty acid binding protein 3 gene variants and essential hypertension in humans. Am J Hypertens. 2008; 21: 691–695.
- ↵Li X, Tjwa M, Van Hove I, Enholm B, Neven E, Paavonen K, Jeltsch M, Juan TD, Sievers RE, Chorianopoulos E, Wada H, Vanwildemeersch M, Noel A, Foidart JM, Springer ML, von Degenfeld G, Dewerchin M, Blau HM, Alitalo K, Eriksson U, Carmeliet P, Moons L. Reevaluation of the role of VEGF-B suggests a restricted role in the revascularization of the ischemic myocardium. Arterioscler Thromb Vasc Biol. 2008; 28: 1614–1620.
- ↵Shiojima I, Walsh K. Role of Akt signaling in vascular homeostasis and angiogenesis. Circ Res. 2002; 90: 1243–1250.
- ↵Ide T, Tsutsui H, Kinugawa S, Utsumi H, Kang D, Hattori N, Uchida K, Arimura K, Egashira K, Takeshita A. Mitochondrial electron transport complex I is a potential source of oxygen free radicals in the failing myocardium. Circ Res. 1999; 85: 357–363.
- ↵Lopez-Campistrous A, Hao L, Xiang W, Ton D, Semchuk P, Sander J, Ellison MJ, Fernandez-Patron C. Mitochondrial dysfunction in the hypertensive rat brain: respiratory complexes exhibit assembly defects in hypertension. Hypertension. 2008; 51: 412–419.
- ↵Chan SH, Wu KL, Chang AY, Tai MH, Chan JY. Oxidative impairment of mitochondrial electron transport chain complexes in rostral ventrolateral medulla contributes to neurogenic hypertension. Hypertension. 2009; 53: 217–227.
- ↵Schillinger KJ, Tsai SY, Taffet GE, Reddy AK, Marian AJ, Entman ML, Oka K, Chan L, O'Malley BW. Regulatable atrial natriuretic peptide gene therapy for hypertension. Proc Natl Acad Sci U S A. 2005; 102: 13789–13794.
- ↵Newton-Cheh C, Larson MG, Vasan RS, Levy D, Bloch KD, Surti A, Guiducci C, Kathiresan S, Benjamin EJ, Struck J, Morgenthaler NG, Bergmann A, Blankenberg S, Kee F, Nilsson P, Yin X, Peltonen L, Vartiainen E, Salomaa V, Hirschhorn JN, Melander O, Wang TJ. Association of common variants in NPPA and NPPB with circulating natriuretic peptides and blood pressure. Nat Genet. 2009; 41: 348–353.
- ↵Yang R, Powell-Braxton L, Ogaoawara AK, Dybdal N, Bunting S, Ohneda O, Jin H. Hypertension and endothelial dysfunction in apolipoprotein E knockout mice. Arterioscler Thromb Vasc Biol. 1999; 19: 2762–2768.
- ↵Pravenec M, Churchill PC, Churchill MC, Viklicky O, Kazdova L, Aitman TJ, Petretto E, Hubner N, Wallace CA, Zimdahl H, Zidek V, Landa V, Dunbar J, Bidani A, Griffin K, Qi N, Maxova M, Kren V, Mlejnek P, Wang J, Kurtz TW. Identification of renal Cd36 as a determinant of blood pressure and risk for hypertension. Nat Genet. 2008; 40: 952–954.
- ↵Timofeeva AV, Goriunova LE, Khaspekov GL, Il'inskaia OP, Sirotkin VN, Andreeva ER, Tararak EM, Bulkina OS, Buza VV, Britareva VV, Karpov Iu A, Bibilashvili R. Comparative transcriptome analysis of human aorta atherosclerotic lesions and peripheral blood leukocytes from essential hypertension patients. Kardiologiia. 2009; 49: 27–38.
- ↵Kim J, Lee CK, Park HJ, Kim HJ, So HH, Lee KS, Lee HM, Roh HY, Choi WS, Park TK, Kim B. Epidermal growth factor induces vasoconstriction through the phosphatidylinositol 3-kinase-mediated mitogen-activated protein kinase pathway in hypertensive rats. J Pharmacol Sci. 2006; 101: 135–143.
- ↵MacDonald MJ. Feasibility of a mitochondrial pyruvate malate shuttle in pancreatic islets: further implication of cytosolic NADPH in insulin secretion. J Biol Chem. 1995; 270: 20051–20058.
- ↵Zahn JM, Poosala S, Owen AB, Ingram DK, Lustig A, Carter A, Weeraratna AT, Taub DD, Gorospe M, Mazan-Mamczarz K, Lakatta EG, Boheler KR, Xu X, Mattson MP, Falco G, Ko MS, Schlessinger D, Firman J, Kummerfeld SK, Wood WH III, Zonderman AB, Kim SK, Becker KG. AGEMAP: a gene expression database for aging in mice. PLoS Genet. 2007; 3: e201.
- ↵de Magalhaes JP, Curado J, Church GM. Meta-analysis of age-related gene expression profiles identifies common signatures of aging. Bioinformatics. 2009; 25: 875–881.
- ↵Dudley JT, Tibshirani R, Deshpande T, Butte AJ. Disease signatures are robust across tissues and experiments. Mol Syst Biol. 2009; 5: 307.