© 2001 European Society of Cardiology
Application of industrial scale genomics to discovery of therapeutic targets in heart failure
a CuraGen Corporation 555 Long Wharf Drive, New Haven, CT 06511, USA
b COR Therapeutics Inc 256 E Grand Avenue, South San Francisco, CA 94080, USA
* Corresponding author. Tel.: +1-203-974-6297; fax: +1-203-401-3351. E-mail address: fmehraba{at}curagen.com (F. Mehraban).
 |
Abstract |
|---|
 Top Abstract 1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
In recent years intense activity in both academic and industrial sectors has provided a wealth of information on the human genome with an associated impressive increase in the number of novel gene sequences deposited in sequence data repositories and patent applications. This genomic industrial revolution has transformed the way in which drug target discovery is now approached. In this article we discuss how various differential gene expression (DGE) technologies are being utilized for cardiovascular disease (CVD) drug target discovery. Other approaches such as sequencing cDNA from cardiovascular derived tissues and cells coupled with bioinformatic sequence analysis are used with the aim of identifying novel gene sequences that may be exploited towards target discovery. Additional leverage from gene sequence information is obtained through identification of polymorphisms that may confer disease susceptibility and/or affect drug responsiveness. Pharmacogenomic studies are described wherein gene expression-based techniques are used to evaluate drug response and/or efficacy. Industrial-scale genomics supports and addresses not only novel target gene discovery but also the burgeoning issues in pharmaceutical and clinical cardiovascular medicine relative to polymorphic gene responses.
Key Words: Genomics • Expression profiling • Cardiovascular disease • Drug discovery
Received April 13, 2001; Revised May 30, 2001; Accepted August 14, 2001
|
1. Foreword |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
A major challenge facing researchers in the cardiovascular arena is deciphering the genetics of complex CVD traits. CVD results when genetic predisposition is juxtaposed on environmental risk factors. With approximately 3.5 billion bases of the human genome sequenced, the task of unraveling the meaning of the code and its portends to human disease remains a daunting undertaking. Given that the number of expressed genes in the human genome is estimated at approximately 35 000 [1,2], based on the assumption and method of calculation used, the task of deciphering the subset of these that may qualify as drug targets for CVD is a challenge that scientists will face in the next decade.
There are two complimentary approaches for derivation gene sequence. Sequencing cDNA derived from mRNA of a cell or tissue is the method of choice when only the genes expressed in that cell/tissue are of interest. It is estimated that in an animal any given tissue may express 5000–10 000 genes of the genome. Thus, cDNA sequencing provides the actual expression repertoire of the tissue of interest. On the other hand, sequencing the genomic DNA of an organism provides the sequence of both introns and exons as they occur on a chromosome. This genomic DNA sequence information has to be further analyzed by bioinformatic tools and techniques in order to derive the sequence of a putative expressed gene by a process termed in silico exon linking. As implied, this analysis provides an in silico prediction of a gene sequence. The accuracy of the predicted sequence depends on the predictive power of the analysis tools and techniques used, and the skill of the scientist in piecing together genomic information to derive actual expressed gene sequences. Further tissue profiling (for example by Northern blotting or polymerase chain reaction cDNA profiling) is then required in order to establish which tissue a predicted gene is expressed in.
Gene expression profiling and/or large scale sequencing of cDNAs derived from cardiovascular tissues using state-of-the-art technologies coupled with sophisticated bioinformatic gene analysis techniques are providing us with insight into complex genetic and biochemical pathways that present leads for development and validation as therapeutic targets. In the age of genomics, discovery of novel drug targets needs to incorporate and integrate gene expression data, gene sequence data, gene polymorphism data, as well as proteomic data, not only in the context of new target discovery, but also to support the development of products in the clinic.
As the name suggests, industrial scale genomics differs from the small laboratory scale set-up in that massive quantities of data are produced and analyzed on unprecedented, ever-increasing rates. This ‘production’ is process-dependent and requires a division of labor such that biology is fitted into an assembly process not unlike other manufacturing processes in other industries. The product here is gene sequence, gene expression data, and gene function data. Gene expression data are used as a lead to further biological evaluations, tissue expression, in vitro and in vivo activity assays, that leads to qualified targets. The validation of qualified targets in a disease context is yet another challenge for the pharmaceutical industry and involves additional drug candidate efficacy and safety evaluations.
The technical repertoire for assessment of differential gene expression has been steadily growing for the past decade. Present technology boasts a slew of techniques from lab-top individual small scale gene expression profiling such as subtractive hybridization, used mainly by solo academic laboratory researchers, to large high-throughput industrial scale sequencing such as the publicly funded Human Genome Project, and other public sequence and tissue expression repositories such as UniGene [3], Serial Analysis of Gene Expression (SAGE) [4], GenBank, European Molecular Biology Laboratory (EMBL), The Munich Information Center for Protein Sequences (MIPS) [5,6] , and Kyoto Encyclopedia of Genes and Genomes (KEGG databases) [7]. Celera Genomics (Rockville, MD), CuraGen (New Haven, CT), Incyte Pharmaceuticals (Palo Alto, CA), Human Genome Sciences (Rockville, MD), etc., exemplify additional notable proprietary private databases.
For the purpose of familiarizing the readers with terminology and names of some of the technologies and processes that are used in cardiovascular target discovery, and because the detailed description of the techniques are beyond the scope of this editorial, reference is made to a number of publications that describe these technologies [8–12]. However, an attempt is made to briefly outline some of them, and their use in CVD target discovery.
|
2. Differential gene expression (DGE) profiling: application to CVD gene discovery |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
DGE is a technique whereby a difference in the quantity of mRNA transcript for a gene or a collection of genes is measured in two tissues or cell samples that are under comparison. DGE technologies fall into two categories, ‘open’ architecture and ‘closed’ architecture systems; they differ in a crucial and important mode of operation. Open systems are not limited by a priori knowledge of differentially expressed gene sequence, for example GeneCalling® [9], SAGE [8], and Total Gene Expression Analysis (TOGA) [10,13]. By contrast, closed systems (e.g. spotted microarrays, DNA GeneChip® [11] are limited in power of detection of novel targets as they require gene sequence information prior to the start of profiling experiments. For more information on various glass chip and array-based technologies see Slabiak [12] and Marshall and Hodgson [14].
In the context of a diseased tissue or a cell-line exhibiting a certain phenotype of interest, much information may be gained by simply asking what genes are expressed in a diseased tissue or cell that are expressed at a lower or higher level in normal age-matched samples. In particular, the availability of suitable animal models of CVD and various in vitro cellular response models can be exploited using DGE profiling.
We used quantitative expression analysis (QEA) and GeneCalling to profile differentially expressed genes in several animal models of CVD, including a rat pressure overload-induced model [9], and Gs
transgenic mouse model [15,16].
GeneCalling is a rapid high throughput open-ended DGE technology [9]. Fig. 1 shows the basics of GeneCalling technology. Input RNA obtained from disease or normal tissue is converted to double-stranded cDNA and digested with 96 pairs of restriction enzymes. This is followed by ligation of unique primer-adapter sets and amplification by polymerase chain reaction (PCR). MegaBACE electrophoresis is used to separate and size the products. Two characterizing identifiers, i.e. electrophoretic mobility and the pair of restriction enzymes used to generate the fragments uniquely identifies each band. This unique identifier is used to query a database of in silico restriction enzyme pair digests of all known genes in that species. Using statistical criteria a GeneCall list is produced. A GeneCall is the probability of the band belonging to a known gene in a reference database. If a GeneCall is not produced for a particular band, the band is likely derived from a novel gene. In this case it is isolated and sequenced.
|
The results of a GeneCalling experiment are shown in Tables 1 and 2. Bands that are differentially expressed in Gs
hearts vs. wild-type age-matched hearts are GeneCalled by comparison of electrophoretic mobility (MegaBACE) and sequence specified by RE pair used. To confirm the GeneCall, the PCR reaction is repeated in the presence of excess unlabeled gene specific oligos in the reaction. Ablation of the electrophoretic peak corresponding to the band of interest indicates that the band was derived from that gene (also see Fig. 1 and legend). In this study we identified several potentially drugable and/or antibody targets that were differentially expressed in Gs hearts, including atrial natriuretic peptide, TGFβ, and uncoupling protein-2 [17].
|
|
Another open-ended DGE method, SAGE, has been successfully employed to address gene expression differences in endothelial cells after treatment with an atherogenic stimulus [18]. In SAGE, a short tag of 10 bp is used to uniquely identify each cDNA transcript derived from a specific mRNA, these tags are concatenated into a single DNA molecule and sequenced. This allows direct identification of a large number of sequence tags that are representative of the mRNA pool being profiled [8]. By theoretical considerations, a 10-bp sequence of nucleotides is sufficiently unique that it can identify a transcript with a high level of certainty from others expressed in that tissue.
Spotted DNA arrays and gene chips are proving to be useful tools for the detection of gene expression differences in normal vs. diseased tissue. Spotted DNA arrays are made by spotting or ‘printing’ DNA sequences onto nylon or glass surfaces. On the other hand, gene chips are usually made by direct synthesis of DNA sequences onto activated glass surfaces or silicon chips by photolithography [11,14]. These are hybridization-based technologies that provide gene expression information by hybridization of an unknown nucleotide sequence in a sample to a known gene sequence or fragment immobilized on a support. Hybridization results are usually detected by fluorescence and the information is digitally analyzed to provide quantitative gene expression data. With the present state of annotation of the human genome, more than 20% of the known genes can be surveyed in a single profiling experiment using gene chips or spotted DNA microarrays. This approach was used to assess DGE in a rat model of myocardial infarction [19]. In the latter study a total of 7000 cDNAs corresponding to approximately 4200 genes were spotted on a microarray, and approximately 400 were found to be differentially modulated in post-ischemic heart tissue. Of note, the microarray approach was also successful in identifying the Tangier Disease mutation (ABC1 transporter defect) where nearly 10 000 human cDNAs were arrayed [20]. As noted above, the power of this technique is limited to detection of known genes, and known gene variants, and does not offer opportunity to discover novel targets, unless the approach is coupled with the generation of an expressed sequence database that provides proprietary novel sequence information for generation of arrays (discussed below). As annotation of the draft human genomic sequences progresses more and more genes that are likely expressed by cardiovascular tissues may be profiled on arrays and chips. A notable contribution is a listing of more than 40 000 ESTs derived from cDNA library sequencing of cardiovascular tissue has been published by Hwang et al. [21]. This compendium is a valuable reference resource for all researchers in academia and private industry alike. However, it should be noted that this is a list of genes expressed in cardiovascular tissue and should not be confused with genes that are differentially expressed in a disease context.
A variety of DNA arrays and chips using radioactive or fluorescent detection are available for disease gene profiling experiments. Examples are those offered by Genome Systems (St. Louis, MO), Clontech, (Palo Alto, CA), Incyte Pharmaceuticals Palo Alto, CA), and Affymetrix (Santa Clara, CA), Synteni (Fremont, CA), Motorola Inc. (Schaumburg, IL), Rosetta Inpharmatics Inc. (Kirkland, WA), etc. However, in the context of CVD a more direct and fruitful approach would be to triage the expressed human genome in terms of a minimal set of genes that can be arrayed that are expected to be involved in CVD.
|
3. Generation of an expressed sequence database for discovery of novel cardiovascular genes |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
Many genes suspected of involvement in CVD pathogenesis are likely derived from two types of cells, namely platelets and endothelial cells. These genes are of major interest in CVD drug target discovery because they are involved in thrombosis, atherosclerosis, and angiogenesis. To address this specific sub-set of CVD genes we generated a database of expressed gene sequences derived from these cell types.
An expressed sequence database is a database containing cDNA sequence information (with or without associated annotation) generated by sequencing cDNA fragments of reverse-transcribed mRNA. These sequence fragments are bioinformatically ‘assembled’ together to form gene nucleotide sequences from which the protein coding frames can be predicted. The assembled cDNA sequences represent a sampling of the total expression repertoire of a given tissue. Because the level of mRNA expression of various genes varies from several thousand copies to a few copied per cell, the technical prowess in generating such a database is of utmost importance in order to produce a truly diverse database that represents an accurate mRNA profile of the tissue used.
We used SeqCallingTM technology [22] to generate an expressed sequence database comprising human platelet and EC sequences assembled together with public sequences to form consensus extensions. The process is depicted in Fig. 2a. SeqCalling uses a similar chemistry to GeneCalling; cDNA fragments generated by RE digestion are fractionated in a desired size range, cloned and sequenced. Sequences that cluster together according to pre-defined parameters are assembled to form sequence assemblies that comprise a representative sequence of the components according to PHRED and PHRAP methods [23–25].
|
PHRED and PHRAP are informatic algorithmic computer programs. PHRED is a base-calling program that gives a score to a base at a given position in a sequence trace according to certain criteria. The user can use this PHRED score as a guide to how certain a base call is at a given position in a DNA sequence. PHRAP is s a computer program that examines input DNA sequences and assembles like DNA sequences according to preset criteria and algorithms to produce a consensus sequence that is representative of all the input sequences.
SeqCalling has been shown to produce sequences biased towards the protein-coding region of genes (Fig. 2b). The sequence information can be used to create microarrays and/or DNA chips such as to facilitate further profiling or gene expression for a variety of CVD and drug responsiveness research, and also CVD markers, helping develop potential cardiovascular drug candidates faster and more cost effectively than ever before.
|
4. Expression pharmacogenomics (ExPg) as a tool for assessment of drug efficacy and/or toxicity |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
A genomic-based technique that is rapidly gaining recognition as a tool for drug safety and efficacy evaluation is ExPg. As the name implies, ExPg is the study of the effect of a drug on gene expression.
An essential and costly step in bringing a drug candidate to the market is the long process of safety and toxicity evaluation associated with obtaining regulatory approval.
ExPg can be used to assess drug toxicity and thus avoid costly drug withdrawals from the market following unexpected toxicity effects. For example, the anti-obesity appetite suppressant drugs dexfenfluramine (DXF) and fenfluramine (FEN) were launched in the US for management of obesity, and were soon withdrawn because of suspected cardiotoxicity and valvulopathy induced in users. These drugs are selective serotonin reuptake inhibitors (SSRI) and appear to have toxic effects on cardiac valves associated with the presence of 5HT2 receptor sub-type on this tissue [26].
ExPg was used to assess the gene expression profile in rat heart tissue after dosing with DXF, FEN and positive (dihhydroergotamine and sumatriptan) and negative control compounds (fluoxetine and sibutramine). Nineteen thousand gene fragments derived from dosed rat cardiac tissue were assayed for differential expression by GeneCalling [27]. This study showed that DXF and FEN induce gene responses similar to known toxic SSRI compounds dihhydroergotamine and sumatriptan. It is likely that in many cases the toxicity associated with newly-derived compounds of pharmaceutic promise may be largely predicted by using ExPg.
ExPg may also be used to assess the efficacy of a drug by comparison of gene expression profiles of known efficacy-related marker pathways such as metabolic and/or signaling pathways. In this approach expression profiles of candidate efficacy marker genes are assessed in animals dosed with the drug under investigation and compared with that of related compounds of known efficacy. Therefore, ExPg is likely to prove a valuable and indispensable tool for the pharmaceutical industry by helping to streamline the process of candidate drug efficacy and/or toxicity assessments and to enable these drug development hurdles to be overcome in a shorter time than previously achievable.
|
5. Sequence polymorphisms (SP) and single nucleotide polymorphisms (SNPs) |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
Often a disease gene arises through either a sequence alteration leading to a change in mRNA expression level or a sequence change manifesting in an alteration in the activity of the resulting protein. There are two types of sequence polymorphisms. SNPs are single base mutations leading, in some cases, to an alteration in functionally important amino acids in the protein. Another type is Insertion/Deletion (InDel) polymorphism. InDels, as implied, are insertions or deletions of whole stretches of sequence giving rise to allelic variants. Polymorphisms are scattered throughout the human genome with a high frequency (approx. every 300–500 bp), and occur in intronic and exonic sequences with as yet unknown rates. In undertaking the procurement of gene polymorphism databases it is therefore important to sequence not only the exonic protein coding regions of the gene but also in the exon/intron boundaries where gene regulatory elements are localized, for example, sequences flanking the 5' most exon. The clustering of various polymorphisms in quantitative trait loci may give rise to haplotypes that contribute to the manifestation of disease phenotypes.
While genotyping approaches to CVD have not been widely implemented on a large scale, several studies addressing polymorphisms relative to CVD have been reported that focused on sequence variations within genes known to be involved in atherosclerosis and hypertension. A study of 9.7 kb of the lipoprotein lipase gene in 71 individuals revealed 88 variations within the sequenced interval, of which 79 were SNPs and the remainder were InDels [28]. Perhaps the most comprehensive attempt of genotype SNP mapping for a candidate CVD susceptibility gene is that reported for ApoE gene where 5.5 kb spanning the entire ApoE genomic region was sequenced in 77 individuals resulting in identification of 21 SNPs and one InDel [29]. Oligonucleotide ligation assays were performed on 20 SNP sites on more than 2000 individuals followed by scoring of the relative allele frequencies which proved to be generally consistent with that found in the core sample. Angiotensin converting enzyme (ACE) was also found to be polymorphic with 78 varying sites over a 24-kb locus [30]. Although there is some controversy surrounding the contribution of polymorphisms in this gene to CVD, there appears to be enough evidence to suggest that this polymorphism may be associated with some forms of heart disease [30–41]. The demonstration of InDel polymorphisms in ACE and associated controversies underscores the importance of genomic research, in particular, SNP-related gene mapping, and may be viewed as one of the many challenges facing both academic and industry researchers in coming years.
These studies, although not industrial scale genotyping, emphasize the predictive potential of polymorphism haplotypes in CVD. The challenge that the genomic industry faces is scanning the whole human genome for polymorphisms and relating the sequence information to specific disease susceptibilities or drug responsiveness. The genomic industry is increasingly cognizant of the importance that polymorphisms play in manifestation of disease. Much information needs to be analyzed to produce a genetic map of CVD disease polymorphic haplotypes. Several genomic companies including CuraGen, Sequenom (San Diego, CA), Genaissance Pharmaceuticals (New Haven, CT), and Gemini Genomics (Cambridge, UK) are presently engaged in industrial scale analysis of disease-related SNPs for disease genotyping. Of note an analysis of SNPs in the glucocorticoid receptor gene was found to be predictive of propensity to hypertension [42,43].
Genetic polymorphisms are also often a source of individual patient variation of response to drugs because they impact the activity of drug metabolizing enzymes or, for example, the crucial structural features in drug receptors, ion channel, transporter, etc. [44]. The efficacy of a pharmacologic agent may be predicted with a good deal of certainty if knowledge of the interacting proteins’ underlying genetic polymorphism is known. Relative to this, it should be noted that chips on which certain SNPs or other genetic predispositions can be assayed (i.e. diagnostics) are the subject of continuing research. This is because the molecular genotype for many diseases and/or drug responses is not as yet adequately defined. This technology would need to be used in conjunction with sustained strong research in identifying genes and defects that are associated with a phenotype or a drug response in patient cohorts. In short, industry can provide the chips for SNP analysis, but the database of sequences to query is largely insufficient and thus this provision is for the moment highly restricted. It may be many years before genetic anomalies are screened in this way. As industry makes more headway in mapping these polymorphisms, we should be able to construct a map of human gene SNP haplotypes that are predictive of certain CVD conditions such as atherosclerosis, thrombosis, and dislipidemia. Similarly, such a knowledge base when correlated with drug response parameters will help physicians prescribe the more efficacious drugs for their patients.
|
6. Limitations of industrial genomic research data |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
As with any technology, there are limitations to industrial scale approaches. Chip-based technologies are inherently prone to errors and need careful calibration, and batch-to-batch production quality control. A second criticism leveled on approaches that quantify cDNA is that these technologies provide expression data regarding mRNA availability. These mRNA profiling data do not address whether a protein is made, the rate at which it is made and degraded, and how secondary structure modifications affect its biologic activity. Because of these limitations, protein expression and other proteomic-based data are required to complement and complete the gene expression data.
Lastly, because industrial research is by its very nature a capital-based, highly competitive undertaking, the results of research are usually not immediately available to the scientific community to access and build upon. However, the results usually do become available in patents and patent co-operation treaty publications in European countries. While accessibility of genomic data may be a limiting factor in some cases, there is ample publicly available data to enable researchers to address the immediate issues pertaining to CVD candidate drug discovery.
It should also be pointed out that the scientific community can access sequence data and other information published in patents and use the data for research purposes only, as long as the work does not infringe the commercial rights of the industry.
|
7. Concluding remarks |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
Sequencing of cardiovascular tissues working in tandem with various DGE technologies is a powerful and indispensable tool for the future of drug target discovery in CVD. Assignment of function and biologic pathways to novel genes and known genes in a novel context, ultimately, needs to utilize proteomic technologies such as the Yeast-Two-Hybrid system to map the protein interactions [45], and thence to derive a map of the human proteome. These and other emerging technologies together with gene mapping techniques should provide researchers the tools to fine-map CVD susceptibility genes and ultimately produce a multi-locus physical map of CVD risk genes and markers in the context of the human genome.
|
Acknowledgements |
|---|
We wish to thank CuraGen Corporation SeqCalling Bioinformatic group for the data shown in Fig. 2b and Drs Peter Lomedico, Elizabeth Walker, Cynthia Green, and Bonnie Gould-Rothberg of CuraGen Corporation, and Dr James Topper of COR Therapeutics Inc. for their comments.
|
References |
|---|
|
TopAbstract1. Foreword2. Differential gene expression...3. Generation of an...4. Expression pharmacogenomics...5. Sequence polymorphisms (SP)...6. Limitations of industrial...7. Concluding remarksReferences |
|---|
- Ewing B., Green P. Analysis of expressed sequence tags indicates 35,000 human genes. Nat Genet (2000) 25:232–234.[CrossRef][Web of Science][Medline]
- Liang F., Holt I., Pertea G., Karamycheva S., Salzberg S.L., Quackenbush J. Gene index analysis of the human genome estimates approximately 120,000 genes. Nat Genet (2000) 25:239–240.[CrossRef][Web of Science][Medline]
- http://www.ncbi.nlm.nih.gov/unigene/.
- http://www.ncbi.nlm.nih.gov/sage/.
- Baker W., van den Broek A., Camon E., et al. The EMBL nucleotide sequence database. Nucleic Acids Res (2000) 28:19–23.
[Abstract/Free Full Text] - Mewes H.W., Heumann K., Kaps A., et al. MIPS: a database for genomes and protein sequences. Nucleic Acids Res (1999) 27:44–48.
[Abstract/Free Full Text] - http://www.genome.ad.ip/kegg.html.
- Velculescu V.E., Zhang L., Vogelstein B., Kinzler K.W. Serial analysis of gene expression. Science (1995) 270:484–487.
[Abstract/Free Full Text] - Shimkets R.A., Lowe D.G., Tai J.T., et al. Gene expression analysis by transcript profiling coupled to a gene database query. Nat Biotechnol (1999) 17:798–803.[CrossRef][Web of Science][Medline]
- Sutcliffe J.G., Foye P.E., Erlander M.G., et al. TOGA: an automated parsing technology for analyzing expression of nearly all genes. Proc Natl Acad Sci USA (2000) 97:1976–1981.
[Abstract/Free Full Text] - Pease A.C., Solas D., Sullivan E.J., Cronin M.T., Holmes C.P., Fodor S.P. Light-generated oligonucleotide arrays for rapid DNA sequence analysis. Proc Natl Acad Sci USA (1994) 91:5022–5026.
[Abstract/Free Full Text] - Slabiak T. Hybridization array technologies http://www.bio.com/articles/dnachip/dnachip.html.
- Green C.D., Simons J.F., Taillon B.E., Lewin D.A. Open systems: panoramic views of gene expression. J Immunol Methods (2001) 250:67–79.[CrossRef][Web of Science][Medline]
- Marshall A., Hodgson J. DNA chips: An array of possibilities. Nat Biotechnol (1998) 16:27–31.[CrossRef][Web of Science][Medline]
- Iwase M., Uechi M., Vatner D.E., et al. Cardiomyopathy induced by cardiac Gs alpha overexpression. Am J Physiol (1997) 272:H585–H589.[Web of Science][Medline]
- Geng Y.J., Ishikawa Y., Vatner D.E., et al. Apoptosis of cardiac myocytes in Gs alpha transgenic mice. Circ Res (1999) 84:34–42.
[Abstract/Free Full Text] - Gausin V, Tomlinson JE, Mehraban F et al. GeneCalling yields new insights into mechanism of cardiomyopathy in mice overexpressing Gs in the heart. American Heart Association, 2001; Anaheim, CA, Circulation, 2001.
- de Waard V., van den Berg B.M., Veken J., Schultz-Heienbrok R., Pannekoek H., van Zonneveld A.J. Serial analysis of gene expression to assess the endothelial cell response to an atherogenic stimulus. Gene (1999) 226:1–8.[CrossRef][Web of Science][Medline]
- Stanton L.W., Garrard L.J., Damm D., et al. Altered patterns of gene expression in response to myocardial infarction. Circ Res (2000) 86:939–945.
[Abstract/Free Full Text] - Lawn R.M., Wade D.P., Garvin M.R., et al. The Tangier disease gene product ABC1 controls the cellular apolipoprotein-mediated lipid removal pathway. J Clin Invest (1999) 104:R25–R31.[Web of Science][Medline]
- Hwang D.M., Dempsey A.A., Wang R.-X., et al. A genome-based resource for molecular cardiovascular medicine: Toward a compendium of cardiovascular genes. Circulation (1997) 96:4146–4203.
[Abstract/Free Full Text] - Rothberg JM, McKenna MP, Predki PF, Windemuth A, Shimkets RA. Inventors; method of identifying nucleic acids. Patent WO 040757, 2000.
- Ewing B., Hillier L., Wendl M.C., Green P. Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res (1998) 8:175–185.
[Abstract/Free Full Text] - Ewing B., Green P. Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res (1998) 8:186–194.
[Abstract/Free Full Text] - Green P. http://www.genome.washington.edu/uwgc/analysistools/phrap.htm.
- Fitzgerald L.W., Burn T.C., Brown B.S., et al. Possible role of valvular serotonin 5-HT(2B) receptors in the cardiopathy associated with fenfluramine. Mol Pharmacol (2000) 57:75–81.
[Abstract/Free Full Text] - Gould Rothberg B.E. The use of animal models in expression pharmacogenomic analysis. Pharmacogenomics J (2001) 1:48–58.[Medline]
- Nickerson D.A., Taylor S.L., Weiss K.M., et al. DNA sequence diversity in a 9.7-kb region of the human lipoprotein lipase gene. Nat Genet (1998) 19:233–240.[CrossRef][Web of Science][Medline]
- Nickerson D.A., Taylor S.L., Fullerton S.M., et al. Sequence diversity and large-scale typing of SNPs in human apolipoprotein E gene. Genome Res (2000) 10:1532–1545.
[Abstract/Free Full Text] - Rieder M.J., Taylor S.L., Clark A.G., Nickerson D.A. Sequence variation in the human angiotensin converting enzyme. Nat Genet (1999) 22:59–62.[CrossRef][Web of Science][Medline]
- Lachurie M.L., Azizi M., Guyene T.T., Alhenc Gelas F., Menard J. Angiotensin-converting enzyme gene polymorphism has no influence on the circulating renin–angiotensin–aldosterone system or blood pressure in normotensive subjects. Circulation (1995) 91:2933–2942.
[Abstract/Free Full Text] - Ishigami T., Iwamoto T., Tamura K., et al. Angiotensin I converting enzyme (ACE) gene polymorphism and essential hypertension in Japan. Ethnic difference of ACE genotype. Am J Hypertens (1995) 8:95–97.[CrossRef][Web of Science][Medline]
- Danser A.H., Schalekamp M.A., Bax W.A., et al. Angiotensin-converting enzyme in the human heart. Effect of the deletion/insertion polymorphism. Circulation (1995) 92:1387–1388.
[Abstract/Free Full Text] - Cambien F., Evans A. Angiotensin I converting enzyme gene polymorphism and coronary heart disease. Eur Heart J (1995) 16(Suppl. K):13–22.
[Abstract/Free Full Text] - Evans A.E., Poirier O., Kee F., et al. Polymorphisms of the angiotensin-converting-enzyme gene in subjects who die from coronary heart disease. Q J Med (1994) 87:211–214.[Web of Science][Medline]
- Cambien F., Costerousse O., Tiret L., et al. Plasma level and gene polymorphism of angiotensin-converting enzyme in relation to myocardial infarction. Circulation (1994) 90:669–676.
[Abstract/Free Full Text] - Cambien F., Poirier O., Lecerf L., et al. Deletion polymorphism in the gene for angiotensin-converting enzyme is a potent risk factor for myocardial infarction. Nature (1992) 359:641–644.[CrossRef][Medline]
- Rigat B., Hubert C., Alhenc Gelas F., Cambien F., Corvol P., Soubrier F. An insertion/deletion polymorphism in the angiotensin I-converting enzyme gene accounting for half the variance of serum enzyme levels. J Clin Invest (1990) 86:1343–1346.[Web of Science][Medline]
- Fujimura T., Yokota M., Kato S., et al. Lack of association of angiotensin converting enzyme gene polymorphism or serum enzyme activity with coronary artery disease in Japanese subjects. Am J Hypertens (1997) 10:1384–1390.[Web of Science][Medline]
- Winkelmann B.R., Nauck M., Klein B., et al. Deletion polymorphism of the angiotensin I-converting enzyme gene is associated with increased plasma angiotensin-converting enzyme activity but not with increased risk for myocardial infarction and coronary artery disease. Ann Intern Med (1996) 125:19–25.
[Abstract/Free Full Text] - Villard E., Tiret L., Visvikis S., Rakotovao R., Cambien F., Soubrier F. Identification of new polymorphisms of the angiotensin I-converting enzyme (ACE) gene, and study of their relationship to plasma ACE levels by two-QTL segregation-linkage analysis. Am J Hum Genet (1996) 58:1268–1278.[Web of Science][Medline]
- Lin R.C., Wang W.Y., Morris B.J. Association and linkage analyses of glucocorticoid receptor gene markers in essential hypertension. Hypertension (1999) 34:1186–1192.
[Abstract/Free Full Text] - Griffiths LR, Rutherford S, Morris BJ, inventors; Polymorphisms in a microsatellite region of a glucocorticoid receptor gene patent US06156510. 2000.
- Evans W.E., Relling M.V. Pharmacogenomics: translating functional genomics into rational therapeutics. Science (1999) 286:487–491.
[Abstract/Free Full Text] - Uetz P., Giot L., Cagney G., et al. A comprehensive analysis of protein–protein interactions in Saccharomyces cerevisiae. Nature (2000) 403:623–627.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  V. Dzau Risk assessment in cardiovascular disease: from traditional risk factors to genomics Eur. Heart J. Suppl., August 1, 2003; 5(suppl_F): F48 - F55. [Abstract] [PDF]
|
|
|
|
 |
|
 S. M. Wasserman, F. Mehraban, L. G. Komuves, R.-B. Yang, J. E. Tomlinson, Y. Zhang, F. Spriggs, and J. N. Topper Gene expression profile of human endothelial cells exposed to sustained fluid shear stress Physiol Genomics, December 26, 2002; 12(1): 13 - 23. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||