January 2003, 12:1 > Pharmacogenomics of hypertension.
Abstract  |  PDF (114 K)  |  References (55)  |  Permissions  |  View full-size inline images
Current Opinion in Nephrology and Hypertension:Volume 12(1)January 2003pp 61-70

Pharmacogenomics of hypertension

[Circulation and hemodynamics]

Cadman, Peter E.; O'Connor, Daniel T.

Department of Medicine and Center for Molecular Genetics, University of California, and VA San Diego Healthcare System, San Diego, California, USA

Correspondence to Daniel T. O'Connor MD, Department of Medicine and Center for Molecular Genetics, University of California at San Diego, and VASDHS (9111H), 3350 La Jolla Village Drive, San Diego, CA 92161, USA. Tel: +1 858 5528585, ext 7373; fax: +1 858 6426331; e-mail: doconnor@ucsd.edu

Abbreviations ACE: angiotensin converting enzyme AGT: angiotensinogen ATR1: angiotensin II receptor type 1 CYP: cytochrome P-450 I/D: insertion/deletion LVMI: left ventricular mass index NIH: National Institutes of Health PharmGKB: NIH Pharmacogenetics Knowledge Base SILVHIA: Swedish Irbesartan Left Ventricular Hypertrophy Investigation versus Atenolol SNP: single nucleotide polymorphism UCSD: University of California, San Diego

Abstract TOP

Purpose of review: The emerging field of pharmacogenomics has the potential to fundamentally change the management of essential hypertension, a common, perhaps polygenic syndrome characterized by substantial inter-individual variability in drug responsiveness. As understanding of sequence diversity in the human genome progresses, the prospect grows for tailoring the prescription of antihypertensive drugs to complement common genetic variations among individual patients, allowing optimization of blood pressure control and improved avoidance of drug side effects. Some principles of pharmacogenomics are presented here, along with a review of the most recent literature on genetic determinants of antihypertensive drug responses, with a preview of likely developments to come.

Recent findings: Polymorphisms at candidate pharmacodynamic loci (such as angiotensinogen, angiotensin converting enzyme, and the angiotensin II receptor) have already been shown to predict responses to such specific treatments as angiotensin converting enzyme inhibition and angiotensin II blockade. The National Institutes of Health have established a multi-institutional pharmacogenetics network and knowledge base, whose goals include understanding how common polymorphisms influence therapeutic responses to a variety of drugs, including antihypertensive agents.

Summary: The study of genetic determinants of drug responses, particularly at the pharmacodynamic (drug target/receptor and post-receptor) level, is likely to allow us to more precisely tailor therapy to the individual patient, as well as to promote the creation of novel therapies.

Introduction TOP

Hypertension is a common disorder, afflicting approximately 50 million people in the United States alone [1]. While the final phenotype of elevated blood pressure might be similar from patient to patient, underlying hereditary determinants of blood pressure elevation are likely to be polygenic and heterogeneous [2]. Over the years, it has become clear that essential hypertension is not a Mendelian single-gene disorder but rather a complex trait resulting from the interaction of environment and heredity. Alleles at many different loci are likely to contribute to the ultimate disease trait, and specific combinations of causative alleles might well vary from person to person [3].

Physicians tend to base their antihypertensive prescription decisions on personal experience as well as clinical guidelines such as the sixth consensus report by the Joint National Commission on Evaluation and Treatment of Hypertension [4]. These guidelines, however, are based on hypertension management in large populations of patients with unknown and likely differing heritable predispositions. The response of any particular patient to a specific medication is not easily predictable, and causes of inter-individual variation in blood pressure responsiveness to medications are largely unknown [5]. As a result, traditional hypertension management has consisted of a 'trial and error' approach, wherein patients are prescribed medication, reassessed, and then placed on different or additional drugs as needed. Overall adequacy of blood pressure control in clinics remains low (approximately 27% in the United States), intolerance of medication side effects is common, and polypharmacy is the norm [1,6].

In view of the recent achievements of the Human Genome Project [7], there is growing enthusiasm for applying the knowledge of genomic determinants of drug responsiveness to derive more 'personalized' medication regimens directed to the specific pathophysiology of each patient. Rather than empirically treating hypertension based on clinical experience and broad generalizations regarding such demographics as ethnicity or age [8-11], physicians eventually may perform genotyping of their patients, perhaps enabling identification of not only specific hereditary mechanisms of a patient's disease, but also more specific and rational therapies for that individual. Pharmacogenomics, therefore, is that branch of human genetics that focuses on the heritable determinants of drug responses. The terms 'pharmacogenomics' and 'pharmacogenetics' often have been used interchangeably in the literature, but the term pharmacogenomics alludes to the emerging possibilities of genome-wide approaches to understanding the effects of heredity on any phenotype, perhaps ultimately uncovering additive or even synergistic influences of multiple genes on drug effects. The power and promise of pharmacogenomics as a tool lies in its potential to predict a patient's response to available remedies prior to administration, and to a degree not previously possible.

Basic concepts of drug action: pharmacokinetics and pharmacodynamics TOP

The central dogma on which pharmacogenomics is founded is that the patient's response to any pharmaceutical is influenced by variations in proteins encoded by the genome. The mechanisms by which these proteins determine drug effect may be divided broadly into the two categories: 'pharmacokinetic' and 'pharmacodynamic'.

Pharmacokinetic factors are processes that influence a drug's delivery to and arrival at its destination (i.e. its target or receptor); such factors include absorption, distribution, metabolism and excretion (Table 1).

Table 1

Table 1. Pharmacokinetic influences on drug action

By contrast, pharmacodynamic processes determine a drug's effect at or after it reaches its point of action. Hence, genetic approaches to pharmacodynamics focus on the genetic blueprints for such drug targets as enzymes or membrane receptors, as well as post-receptor signaling pathways to ultimate cellular effectors, and other intracellular mechanisms by which drugs produce changes in cell function. For example, the effects of the β1-adrenergic agonist isoproterenol at the sinus node on heart rate might be influenced by qualitative or quantitative variability in any of the gene products in the drug's signaling pathway (Table 2).

Table 2

Table 2. Pharmacodynamic pharmacogenomics of hypertension: antihypertensive drugs, potentially important pharmacodynamic candidate loci, and common genetic variations at those loci

Classical pharmacokinetic pharmacogenetics: phase I and phase II drug metabolism, 'one gene, many drugs' TOP

Historically, pharmacogenetic studies often focused on single gene polymorphisms (SNPs) that affected drug metabolism and demonstrated Mendelian heritability. Because of their unusually large, single gene effects on drug disposition, such traits were often readily recognized during plasma or urine drug assays in study populations. Investigators discovered that humans could be categorized as either 'fast metabolizers' or 'slow metabolizers' of not only specific drugs but also entire drug families.

An example of such a polymorphism is the hydroxylation of the adrenergic inhibitor debrisoquine, which is performed by the cytochrome P-450 (CYP) enzyme CYP2D6. By the 1980s investigators realized that an individual's ability to metabolize debrisoquine, in turn, predicted the metabolism of other medications processed by the same pathway [12-16]. Humans who were poor metabolizers of debrisoquine were found also to be inefficient in disposition of β-adrenergic antagonists (e.g. propranolol, metoprolol), perhaps resulting in additional β-blockade at comparable doses. The discovery of single gene determinants of drug metabolism formed the basis of classical pharmacogenetics [17], and the loci of importance in many of these cases were found to be enzymes of the CYP family, the enzyme superfamily integral to 'phase I' (oxidative/hydroxylating) drug metabolism [18-21]. About a half-dozen CYPs are involved in the majority of hydroxylations of drugs used therapeutically: CYP1A2, CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP3A4, CYP3A5, CYP3A7.

'Phase II' drug metabolizing enzymes catalyze the addition of substituents to drugs in such processes as acetylation, methylation, sulfation, or glucuronidation. An important example for hypertension is the N-acetylation of hydralazine by N-acetyltransferase 2 (NAT2); 'slow acetylators' may excrete the drug so poorly that the complication of drug-induced lupus erythematosis increases in frequency [22,23].

While variability in such phase I and II processes clearly influences the metabolism of anithypertensive medications still used today (e.g. β-adrenergic antagonist hydroxylation by CYP2D6; hydralazine acetylation by NAT2), genotyping at the responsible loci has not yet achieved widespread clinical use, in part because there are typically many polymorphisms at the responsible loci that can influence enzymatic activity. A case in point is CYP2D6, in which numerous inactivating (loss-of-function, such as premature truncation) polymorphisms have been described, while activating (gain-of-function, such as tandem duplication) polymorphisms have also been characterized [24]. Thus, many genotyping reactions might be required even to profile the function of a single locus such as CYP2D6.

In the case of many medications, drug metabolism, per se, might not be the primary determinant of efficacy. Indeed, drug responses in most individuals are likely to reflect the influence of multiple factors, both hereditary and environmental, rather than the effect of a single enzyme [25]. Furthermore, the example of antihypertensive β-adrenergic blockade reveals that, while β-adrenergic blockers may vary widely in their pharmacokinetic profiles (especially plasma t1/2), members within this class of antihypertensive medications differ little in their ability to lower blood pressure at conventional doses [26,27]. At higher drug doses, dose-response curves ultimately become flat, as maximal efficacy is approached within that class; at this point, plasma drug concentrations may no longer correlate with efficacy. In other words, drug response is in part attributable to events occurring after the arrival of a drug at its site of action [28]. For this reason, contemporary investigations into genetic predictors of drug efficacy will also focus on pharmacodynamic effects and candidate loci.

Pharmacodynamic pharmacogenomics: 'one drug, many genes', the role of single nucleotide polymorphisms TOP

Whereas pharmacokinetic factors determine to what extent a drug reaches its destination, pharmacodynamic determinants influence drug responses at the target/receptor, post-receptor signal transduction, and cellular effector levels. Growing knowledge of the biochemical and physiological mechanisms underlying common variations in blood pressure (e.g. autonomic, renin-angiotensin-aldosterone, and renal factors) and drug targets allows predictions of pharmacodynamic candidate genetic loci likely to be important (Table 2).

'Pharmacogenomics' versus 'Pharmacogenetics' TOP

With the availability of newer and more rapid genome technologies, it becomes possible to study the contribution of many genetic loci to a drug response phenotype; hence the term 'pharmacogenomics' (that is, multi-locus or even genome-wide searches for genes affecting drug responses), rather than the more classical 'pharmacogenetics' (one gene→one drug response).

Single nucleotide polymorphisms TOP

In evaluating these loci, a useful genotypic focus is on the most common form of inter-individual genetic variation, SNPs. SNPs are variants in the DNA sequence, likely resulting from replication errors, in which one nucleotide is substituted for another, typically purine-for-purine (e.g. 'A' appearing instead of 'G') or pyrimidine-for-pyrimidine (C↔T) transitions [29,30]. By genomic standards, SNPs are quite common: SNPs with relatively high minor allele frequencies may occur as often as once every 1000 base pairs. In fact, one group has recently reported on the existence of some 1.42 million SNPs [31]. Thus, particular interest has been generated regarding the significance of SNPs at candidate genes such as those listed in Table 2.

SNPs might influence a patient's response to medication in a number of ways. A polymorphism occurring in the coding region of a gene (so-called 'coding SNP') can result in a qualitative alteration in protein structure (and hence function), if the coding SNP changes the amino acid specified by a codon (so-called 'nonsynonymous' or 'replacement' coding SNPs). It has been estimated that as many as ∼20-30% of nonsynonymous coding SNPs might result in altered protein function [32-34]. Even polymorphisms that occur in noncoding regions of DNA, however, may prove important, since such SNPs might occur in DNA motifs crucial for appropriate transcription, or correct RNA splicing, or efficient translation of messenger RNA. In addition, marker SNPs may occur in 'linkage disequilibrium' with (that is, near to) other functional, as-yet-unidentified alterations in genes, thereby representing potentially useful genetic markers for disease states or drug responses.

Though originally identified by gel-based sequencing, these SNP variants now can be detected more rapidly and inexpensively with high-throughput techniques, such as hybridization on chips or beads, or mass spectrometry. Such technical innovations should enable large-scale, clinically directed human studies (Fig. 1).

Figure 1

Figure 1. Basic approach to pharmacodynamic pharmacogenomic studies in human antihypertensive drug responses. Human subjects are both phenotyped in terms of response to drug (1) and genotyped at candidate pharmacodynamic loci (2). If a statistical association between drug response and genotype is established (3), subsequent patients with hypertension may be screened for such genetic variations (typically in the form of single nucleotide polymorphisms, SNPs) in order to determine optimal therapy (4).

Examples of previous clinical hypertension studies in pharmacodynamic pharmacogenomics: the renin-angiotensin-aldosterone system TOP

Since the early 1990s, much of the groundwork in the field of hypertensive pharmacodynamic pharmacogenomics has focused on genes that code for elements of the renin-angiotensin-aldosterone system. Two examples of polymorphisms that have been studied in patient populations and may ultimately prove clinically important are the angiotensin I converting enzyme (ACE) insertion/deletion and the angiotensinogen Met235Thr polymorphism.

In 1990, Rigat et al. [35] described the existence of a polymorphism at the ACE locus consisting of either the insertion or deletion (that is, presence or absence) of an ∼250 base pair Alu-repeat DNA segment in intron 16 of the gene. The authors found that individuals could be genotyped as possessing either two insertion alleles (I/I), an insertion and a deletion allele (I/D) or two deletion alleles (D/D) at the ACE locus, located on chromosome 17q23. In their study of 80 participants, serum ACE levels differed greatly as a function of genotype: 299±49, 393±66.8, or 494±88.3 μg/l for I/I, I/D and D/D genotypes, respectively (P<0.001 by ANOVA). In fact, genotype accounted for about 47% of the phenotypic variance in serum ACE levels among study participants. More recently, the Swedish Irbesartan Left Ventricular Hypertrophy Investigation versus Atenolol (SILVHIA) trial evaluated the ACE I/D polymorphism as a potential predictor of blood pressure response to treatment with either the angiotensin II receptor type 1 (ATR1) antagonist irbesartan or the β1 adrenergic antagonist atenolol [36]. The SILVHIA trial was a double-blind prospective study in which a total of 101 hypertensive adults with confirmed left ventricular hypertrophy were randomized to 3 months of treatment with either irbesartan or atenolol. All 101 participants were genotyped at the ACE locus. Within the irbesartan-treated group, there was a significantly greater drop in diastolic blood pressure among individuals possessing the I/I genotype as opposed to those with either the D/D or I/D genotypes: mean delta BP±SEM was -18±12 mmHg for I/I genotype, -6±9 mmHg for I/D genotypes, and -8±11 mmHg for D/D (P = 0.0096). Furthermore, no such association between genotype and blood pressure lowering was seen in the atenolol treated group, indicating the I/I genotype was specifically predictive of response to irbesartan.

As with the ACE locus, investigators have wondered whether polymorphisms at the angiotensinogen (AGT) gene may be predictive of blood pressure drug responses. The angiotensinogen Met235Thr (M235T) variation represents a SNP in codon 235 within exon 2 of the AGT locus, wherein a thymine for cytosine substitution at nucleotide position 704 of the gene results in a threonine for methionine ('Thr for Met') amino acid substitution at codon 235 of the angiotensinogen molecule [37]. Individuals may be genotyped as 'Met/Met', 'Met/Thr' or 'Thr/Thr', depending on which version of the protein is encoded by each allele; carriers of the Thr allele (Met/Thr and Thr/Thr genotypes) tend to have higher plasma levels of angiotensinogen [37,38]. In the SILVHIA trial mentioned above, an association was not found between the Met235Thr polymorphism and blood pressure response to irbesartan. Other studies, however, have suggested this genetic variant may be important as a predictor of patient response to ACE inhibitor therapy. When Hingorani et al. [39] administered ACE inhibitor monotherapy to 125 hypertensive patients for 4 weeks, they found that the mean drop in systolic pressure was more than 5 mmHg greater (P = 0.004 for ANOVA comparison of genotypes) in patients with either the Thr/Thr or Thr/Met genotype than in those with the Met/Met genotype. Not all investigators, however, have found such an association between the Met235Thr polymorphism and response to ACE inhibition [40].

The year in review: recent studies of candidate loci TOP

The recently published literature on the pharmacogenomics of hypertension has focused largely on three different and clinically relevant areas of inquiry: salt-sensitive hypertension, genetic prediction of diuretic responsiveness, and prevention of hypertensive end-organ damage.

Polymorphisms predictive of salt-sensitivity of blood pressure TOP

Dietary salt restriction is a common recommendation made to patients with hypertension, and recent dietary studies have demonstrated that the average blood pressure within a population can be lowered significantly through reduction in sodium intake [41]. The response of any particular individual to salt restriction, however, is less predictable, and it has been estimated that only about 50% of patients will have a meaningful response to such maneuvers. Poch et al. [42•] recently reported on two polymorphisms that appear to be clinical predictors of salt-sensitive hypertension. One is the ACE I/D polymorphism already described. Another is the G534A polymorphism in the 11-β-hydroxysteroid dehydrogenase type 2 (HSD11B2) gene, a locus at which mutations can cause a rare form of salt-sensitive monogenic hypertension [43]. The HSD11B2 polymorphism studied by Poch et al. is characterized by an adenine for guanine substitution at nucleotide position 534 (G534A). The authors phenotyped 71 hypertensive patients as either 'salt-sensitive' (n = 35) or 'salt resistant' (n = 36) based on blood pressure response to a 14-day increase in dietary salt intake. They then genotyped participants at the ACE and HSD11B2 loci. The prevalence of salt-sensitive hypertension was significantly higher (P = 0.003) in patients with either the I/I (68%) or the I/D (59%) genotype at the ACE locus compared with D/D hypertensives (19%). Furthermore, the mean increase in systolic blood pressure following a dietary salt load was significantly greater in I/I and I/D hypertensives than in D/D hypertensives (mean delta BP±SEM: 8.3±2.1, 5.2±1.5, or 1.6±1.3 mmHg for I/I, I/D, or D/D individuals; P = 0.031). When the HSD11B2 gene G534A polymorphism was considered, individuals with the G/G genotype had a higher salt-induced rise in systolic blood pressure than those with the G/A genotype (mean delta BP±SEM: 5.4±1 for G/G versus -1.3±3.5 for G/A, in mmHg; P = 0.039). No hypertensives with the A/A genotype were available for comparison. While not strictly in the category of 'pharmacogenomics', observations from this study may one day influence the choice of first-line therapy for essential hypertension.

Polymorphisms that predict response to diuretic therapy TOP

Turner et al. [44•] reported on the C825T polymorphism at the G-protein β3 inhibitory subunit locus, which appears to predict patient responses to thiazide diuretics. Located at exon 10 of the gene encoding the inhibitory β3 isoform subunit of heterotrimeric G proteins, this SNP results in a shortened splice variant of the GNB3 protein. In studies of lymphocytes derived from white hypertensives, the protein encoded by the 825T allele (where a thymine appears rather than a cytosine at nucleotide position 825 of the gene) appears to display enhanced signal transduction and sodium-proton antiporter activity [45]. Turner et al. measured the C825T polymorphism in 387 adult hypertensives (197 black, 190 non-Hispanic white) who had undergone 4 weeks of hydrochlorothiazide monotherapy. Following genotyping, blood pressure response to diuretic therapy was compared in three genotypic groups (C/C homozygotes, C/T heterozygotes, and T/T homozygotes). When the three groups were compared within the pooled sample of blacks and whites, the mean reduction in systolic pressure with thiazide treatment was found to be significantly greater in T/T homozygotes and C/T heterozygotes compared with C/C hypertensives (16.3, 13.6, and 10.3 mmHg, respectively; P<0.001). The T allele appeared with higher frequency in blacks than whites (76.1 versus 28.9%, P<0.001). The authors, however, noted an association between the 825T allele and blood pressure reduction in all race-gender categories except black women.

Drug class-specific prevention of hypertensive end organ damage TOP

Genetic polymorphisms soon may be used not only to predict the effect of medication on blood pressure in terms of mmHg but also in terms of risk reduction for cardiovascular morbidity and mortality. Another locus that appears linked to diuretic response is that of the cytoskeletal single transduction protein α-adducin, where the Gly460Trp variant has been associated with salt-sensitive hypertension and has been assessed recently as a potential tool in minimizing risk of myocardial infarction and stroke. Psaty et al. [46•] studied the α-adducin (ADD1) Gly460Trp polymorphism in a case control study of hypertensives enrolled in a health maintenance organization in Seattle, Washington. Among hypertensives who were carriers of the Trp460 allele (either one or two copies), diuretic therapy (which was defined as either use of a thiazide or loop diuretic) was associated with a significantly lower risk of myocardial infarction or stroke (odds ratio, 0.49; 95% confidence interval, 0.32-0.77) compared with other antihypertensive treatments. No such association with diuretic treatment was observed in patients with the wildtype genotype (Gly460 homozygotes). Furthermore, the apparent protective effect of diuretics in Trp460 allele carriers was independent of age, gender, ethnicity, diabetes or other cardiovascular risk factors. The polymorphism was noted to be quite prevalent (present in ∼1/3 of study participants), though the studied population included subjects from only two ethnicities (black and white).

Another complication of hypertension that remains an area of intense interest is left ventricular hypertrophy. Commonly found in patients with long-standing hypertension, left ventricular hypertrophy correlates with increased cardiovascular morbidity and mortality independent of other risk factors [47]. While the potential benefits of reversing the condition through medical therapy are still being evaluated in prospective trials, some data suggest that left ventricular hypertrophy regression is beneficial [48]. Kurland et al. [49•] have recently reported on polymorphisms in the angiotensinogen and ATR1 genes that appear to predict drug-specific regression in left ventricular mass with angiotensin II receptor blocker therapy. The authors analyzed genotypes at the angiotensin (Met235Thr and Thr174Met), ATR1 (A1166C) and ACE (I/D) genes, in 84 of the original 101 participants from the SILVHIA trial. All of the patients had confirmed left ventricular hypertrophy at initial enrollment and received 3 months of treatment with either irbesartan (n = 41) or atenolol (n = 43). After 3 months of treatment, patients were assessed for reduction in left ventricular mass index (LVMI) by echocardiography. Among patients treated with irbesartan, those who possessed the Thr/Met genotype at angiotensin codon 174 had a significantly greater reduction in LVMI than did individuals with the Thr/Thr genotype (delta LVMI of -23±31 versus 0.5±18 g/m2; P = 0.005). No patients with the relatively uncommon Met/Met genotype were available for comparison. The association of delta LVMI with Met174Thr genotype was not seen in the atenolol-treated group, indicating a drug-class-specific effect; the association also appeared independent of the degree of blood pressure reduction. A similar association between genotype and left ventricular hypertrophy reduction after irbesartan was observed in carriers of the AGT 235Thr or the ATR1 1166C alleles, though the mean reduction in left ventricular mass was not as pronounced as that seen in 174Met carriers.

Current efforts and future directions TOP

In the years to come, a number of centers will be likely to contribute to a growing understanding of the pharmacogenomics of hypertension. Within the USA, three programs of note include the Pharmacogenomic Knowledge Base (centered at Stanford University), the Autonomic Pharmacodynamic Pharmacodynamics project at the University of California at San Diego, and the Pharmacogenetics Network for cardiovascular Risk Therapy at the Lawrence Berkeley National Laboratory Center. All three of these programs receive support from the National Institutes of Health (NIH).

National Institutes of Health Pharmacogenetics Network and its Pharmacogenetics Knowledge Base TOP

As pharmacogenetic genotypes and phenotypes accumulate, there has been a growing demand among investigators for the development of cooperative scientific efforts and a centralized data library. To that end, the National Institutes of Health (NIH) Pharmacogenetics Research Network (http://http://www.nigms.nih.gov /pharmacogenetics) was created by the National Institute of General Medical Sciences, linking sites around the United States conducting human pharmacogenetic research in drug metabolism, cancer, pulmonary, and cardiovascular disease. The network, in turn, has created the Pharmacogenetics Knowledge Base (PharmGKB; http://http://www.PharmGKB.org ), which is managed by Stanford University, with support from several NIH institutes.

Since its creation in April 2000, the PharmGKB has developed into a multifunctional Web-based resource [50••,51,52]. The site acts a repository for submissions from the research community at large, providing a catalog of reported genotypes and their associated clinical (e.g. drug response), cellular, and molecular phenotypes. From its search engine interface, unregistered users may query the database on topics of interest. Visitors also can link to third-party websites and interactive bioinformatic/software tools such as dbSNP (http://http://www.ncbi.nlm.nih.gov /SNP/index.html), GeneTree (http://http://www.stats.ox.ac.uk /mathgen/software.html), and Cleaver (http://classify.stanford.edu/).

While publicly accessible, the clinical data within PharmGKB are maintained with careful regard for patient confidentiality, observing strict guidelines for exclusion of material that could be used for individual identification. In order to contribute to the database and gain access to restricted material, investigators must register as users online, providing contact information, an e-mail address, a unique user identifier, and a current project description. Once registered, users may submit genotypic or phenotypic data by either using Web-based citation forms or submitting data in XML format [53]. Once validated, submissions are entered into the knowledge base and made available to other registered users.

National Institutes of Health pharmacogenetics network investigations of blood pressure or the circulation TOP

The NIH Pharmacogenetics Network currently supports investigations into the pharmacogenetics of hypertension at both the University of California, San Diego and at the Lawrence Berkeley National Laboratory Center. These two programs differ somewhat in focus.

University of California, San Diego center: 'Autonomic Pharmacodynamic Pharmacogenomics' TOP

One target of the NIH Pharmacogenetics Research Network is improved understanding of how heredity governs human autonomic cardiovascular drug responses. Centered at the University of California, San Diego (UCSD), this initiative seeks to identify polymorphisms that influence autonomic drug responses, with a focus on pharmacodynamic determinants; the pharmacodynamic focus is enabled by emphasis on parenteral agonist effects in such regional circulatory beds as pulmonary, renal, forearm, and hand. To examine human subject medication responses at the level of drug receptors and post-receptor intracellular effectors, phenotypes are determined under controlled circumstances designed to avoid such confounders as pharmacokinetic variables (e.g. how well the patient absorbs or metabolizes the administered drug) or the influence of the baroreflex on sympathetic tone. Thus, the central strategy of the project will be to examine local vascular responses to drugs, allowing for predominantly pharmacodynamic phenotypes that can be used to stratify genotypes.

Candidate pharmacodynamic genetic loci for the autonomic drug responses are based on likely sites or events within drug responsive cells: drug targets (e.g. receptors), post-receptor signal transduction molecules, and ultimate cellular effectors (Fig. 2). Following autonomic circulatory phenotyping, individuals are SNP-genotyped at candidate loci, not only to perform simple SNP allelic associations, but also for haplotype inference and haplotype associations with drug-response phenotypes. The functional significance of phenotype-associated SNPs will be tested by both in-vitro and in-vivo methods. To confirm the predictive value of identified SNPs or SNP haplotypes, genomic DNA samples from patients in major US interventional hypertension trials can be genotyped and the blood pressure responses stratified by particular SNPs [54,55].

Figure 2

Figure 2. 'Autonomic Pharmacodynamic Pharmacogenomics'. initiative centered at the University of California, San Diego. Candidate gene single nucleotide polymorphism strategy within drug target cells: receptors, postreceptor signal transduction, and ultimate cellular effectors.

Lawrence Berkeley National Laboratory Center: 'Pharmacogenetics Network for Cardiovascular Risk therapy' TOP

This center focuses its efforts on two cardiovascular risk factor treatments: ACE inhibition (with ramipril) for blood pressure lowering in patients with hypertension, and HMG coenzyme A reductase inhibition (with atorvastatin) for plasma lipid lowering in patients with dyslipidemia. SNPs and SNP haplotypes are determined for approximately 50 genes involved in the metabolism or action of these drugs on their target phenotypes.

Future questions for investigation and resolution TOP

If the technologies of pharmacogenomics are to be adopted into the routine treatment of essential hypertension, these advances will be subject to the same economic considerations that govern healthcare in general. Questions of cost effectiveness, accuracy, and efficiency will need to be addressed before the 'promise' of these techniques can be transformed into 'practice'. Presently, the cost of multi-locus genotyping per individual patient remains high, and the necessary resources are available only within academic medical centers and commercial research and development laboratories. Present turn-around times of weeks to months for genotype results fall short of the desired 'quick answer' that would allow clinicians to make decisions at the bedside. Finally, if pharmacogenetic tools are developed for wide-scale application in hypertension management, long-term, prospective, randomized trials still will need to be performed to verify that such interventions, designed to make health care more rational, actually convey some practical benefits to patients, perhaps in the form of fewer physician office visits, less frequent side effects, or lower overall cost of care.

Conclusion TOP

Choosing the most effective blood pressure medication for an individual patient with hypertension has long been an imperfect endeavor, in which physicians use past experience and the various generalizations regarding ethnicity and age to guide therapy. The emerging field of pharmacogenomics may change how this common disease is approached therapeutically. As understanding of hypertension genetics progresses, so does the potential to apply mechanism-based therapies that match the specific pathophysiology of each individual patient. NIH-sponsored projects directed toward these goals include the PharmGKB and the Autonomic Pharmacodynamic Pharmacogenomics program at UCSD.

Acknowledgements TOP

The authors would like to acknowledge the Pharmacogenetics Research Network of the National Institutes of Health, and the PharmGKB. Pharmacogenomics work in San Diego was supported by the Department of Veterans Affairs, and by NIH grants HL69758 and RR17458.

References and recommended reading TOP

Papers of particular interest, published within the annual period of review, have been highlighted as: TOP

• of special interest TOP

•• of outstanding interest TOP

1 Burt VL, Whelton P, Roccella EJ, et al. Prevalence of hypertension in the US adult population: results from the Third National Health and Nutrition Examination Survey, 1988-1991. Hypertension 1995; 25:305-313.
2 Sharma P, Fatibene J, Ferraro F, et al. A genome-wide search for susceptibility loci to human essential hypertension. Hypertension 2000; 35:1291-1296.
3 Gavras I, Manolis A, Gavras H. Genetic epidemiology of essential hypertension. J Hum Hypertens 1999; 13:225-229.
4 The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Arch Intern Med 1997; 157:2413-2446.
5 Chapman AB, Schwartz GL, Boerwinkle E, Turner ST. Predictors of antihypertensive response to a standard dose of hydrochlorothiazide for essential hypertension. Kidney Int 2002; 61:1047-1055.
6 Cohen JS. Adverse drug effects, compliance, and initial doses of antihypertensive drugs recommended by the Joint National Committee vs. the Physician's Desk Reference. Arch Intern Med 2001; 161:880-885.
7 Venter JC, Adams MD, Myers EW, et al. The sequence of the human genome. Science 2001; 291:1304-1351.
8 Buhler FR, Hulthen L, Kiowski W, Bolli P. Greater antihypertensive efficacy of the calcium channel inhibitor verapamil in older and low renin patients. Clin Sci 1982; 63:439s-442s.
9 Cubeddu LX, Aranda J, Singh B, et al. A comparison of verapamil and propranolol for the initial treatment of hypertension: racial differences in response. JAMA 1986; 256:2514-2521.
10 Saunders E, Weir MR, Kong BW, et al. A comparison of the efficacy and safety of a beta-blocker, a calcium channel blocker, and a converting enzyme inhibitor in hypertensive blacks. Arch Intern Med 1990; 150:1707-1713.
11 Preston RA, Materson BJ, Reda DJ, et al. Age-race subgroup compared with renin profile as predictors of blood pressure response to antihypertensive therapy. JAMA 1998; 280:1168-1172.
12 Gonzalez F, Skoda R, Kimura S, et al. Characterization of the common genetic defect in humans deficient in debrisoquine metabolism. Nature 1988; 331:442-446.
13 Mahgoub A, Idle J, Dring L, et al. Polymorphic hydroxylation of debrisoquine in man. Lancet 1977; 2:584-586.
14 Broly F, Gaedigk A, Heim M, et al. Debrisoquine hydroxylase genotype and phenotype. DNA Cell Biol 1991; 10:545-558.
15 Alvan G, von Bahr C, Seidman P, Sjoqvist F. High plasma concentrations of β-receptor blocking drugs and deficient debrisoquine hydroxylation. Lancet 1982; 1:333.
16 Dahl M, Johansson I, Palmertz MP, et al. Analysis of the CYP2D6 gene in relation to debrisoquine and desipramine hydroxylation in a Swedish population. Clin Pharmacol Ther 1992; 52:12-17.
17 Weber W. Human drug-metabolizing enzyme variants. In: Borrow M, Harper PS, Motulsky AG, Scriver C, editors. Pharmacogenetics. 1st ed. New York: Oxford University Press; 1997. pp. 131-216.
18 Eichelbaum M, Spannbrucker N, Steincke B, Dengler H. Defective N-oxidation of sparteine in man: a new pharmacogenetics defect. Eur J Clin Pharmacol 1979; 16:183-187.
19 Lennard M, Silas J, Freestone S, et al. Oxidation phenotype: a major determinant of metoprolol metabolism and response. N Engl J Med 1982; 307:1558-1560.
20 Broly F, Marez D, Sabbagh N, et al. An efficient strategy for the detection of known and new mutations of the CYP2D6 gene using single strand conformation polymorphism analysis. Pharmacogenetics 1995; 5:373-384.
21 Evans WE, Relling MV. Pharmacogenomics: translating functional genomics into rational therapeutics. Science 1999; 286:487-491.
22 Uetrecht JP, Woosley RL. Acetylator phenotype and lupus erythematosus. Clin Pharmacokinet 1981; 6:118-134.
23 Hofstra AH. Metabolism of hydralazine: relevance to drug-induced lupus. Drug Metab Rev 1994; 26:485-505.
24 Meyer UA, Zanger UM. Molecular mechanisms of genetic polymorphisms of drug metabolism. Annu Rev Pharmacol Toxicol 1997; 37:269-296.
25 Turner ST, Schwartz GL, Chapman AB, et al. Antihypertensive pharmacogenetics: getting the right drug into the right patient. J Hypertens 2001; 19:1-11.
26 Materson B, Reda D, Cushman W, et al. Single-drug therapy for hypertension in men: a comparison of six antihypertensive agents with placebo. N Engl J Med 1993; 328:914-921.
27 Materson B, Reda D, Preston R, et al. Response to a second single antihypertensive agent used as monotherapy for hypertension after failure of the initial drug. Arch Intern Med 1991; 151:1413-1423.
28 Israili Z. Correlation of pharmacological effects with plasma levels of antihypertensive drugs in man. Ann Rev Pharmacol Ther 1979; 54:235-241.
29 Doris P. Hypertension genetics, single nucleotide polymorphisms, and the common disease: common variant hypothesis. Hypertension 2001; 39:323-331.
30 Wang DG, Fan JB, Siao CJ, et al. Large-scale identification, mapping, and genotyping of single-nucleotide polymorphisms in the human genome. Science 1998; 280:1077-1082.
31 Sachidanadam R, Weissman D, Schmidt SC, et al. A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms. Nature 2001; 409:928-933.
32 Sunyaev S. Ramensky V, Koch I, et al. Prediction of deleterious human alleles. Hum Mol Genet 2001; 10:591-597.
33 Chasman D, Adams RM. Predicting the functional consequences of non-synonymous single nucleotide polymorphisms: structure-based assessment of amino acid variation. J Mol Biol 2001; 307:683-706.
34 Ng PC, Heikoff S. Accounting for human polymorphisms predicted to affect protein function. Genome Res 2002; 12:436-446.
35 Rigat B, Hubert C, Alhenc-Gelas F, et al. 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.
36 Kurland L, Melhus H, Karlsson J, et al. Angiotensin converting enzyme gene polymorphism predicts blood pressure response to angiotensin II receptor type 1 antagonist treatment in hypertensive patients. J Hypertens 2001; 19:1783-1787.
37 Bloem LJ, Manatunga A, Tewkesbury DA, Pratt JH. The serum angiotensinogen concentration and variants of the angiotensinogen gene in black and white children. J Clin Invest 1995; 95:948-953.
38 Jeunemaitre X, Soubrier F, Koeletsev YV, Lifton RP, et al. Molecular basis of human hypertension: role of angiotensinogen. Cell 1992; 71:169-180.
39 Hingorani AD, Jia H, Stevens PA, et al. Renin-angiotensin system gene polymorphisms influence blood pressure and the response to angiotensin converting enzyme inhibition. J Hypertens 1995; 13:1602-1609.
40 Dudley C, Keavney B, Casadei B, et al. Prediction of patient responses to antihypertensive drugs using genetic polymorphisms: investigation of renin-angiotensin system genes. J Hypertens 1996; 14:259-262.
41 Sacks FM, Svetkey LP, Vollmer WM, et al. Effects on blood pressure of reduced dietary sodium and dietary approaches to stop hypertension (DASH) diet. N Engl J Med 2001; 344:3-10.
42• Poch E, Gonzalez D, Giner V, et al. Molecular basis of salt sensitivity in human hypertension. Hypertension 2001; 38:1204-1209. Genotypes at the HSD11B2 and ACE loci predicted salt sensitivity in patients with high blood pressure.
43 Lifton RP. Molecular genetics of blood pressure variation. Science 1996; 272:676-680.
44• Turner ST, Schwartz GL, Chapman AB, Boerwinkle E. C825T polymorphism of the G protein β3-subunit and antihypertensive response to a thiazide diuretic. Hypertension 2001; 37:739-743.
45 Siffert W, Naber C, Walla M, Ritz E. Enhanced G protein activation in immortalized lymphoblasts from patients with essential hypertension. J Clin Invest 1995; 96:759-766.
46• Psaty BM, Smith NL, Heckbert SR, et al. Diuretic therapy, the α-adducin gene variant, and the risk of myocardial infarction or stroke in persons with treated hypertension. JAMA 2002; 287:1680-1689.
47 Levy D, Garrison RJ, Savage DD, et al. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990; 322:1561-1566.
48 Verdecchia P, Schillaci G, Borgioni C, et al. Prognostic significance of serial changes in left ventricular mass in essential hypertension. Circulation 1998; 97:48-54.
49• Kurland L, Melhus H, Karlsson J, et al. Polymorphisms in the angiotensinogen and angiotensin II type 1 receptor gene are related to change in left ventricular mass during antihypertensive treatment: results from the Swedish Irbesartan Left Ventricular Hypertrophy Investigation versus Atenolol (SILVHIA). J Hypertens 2002; 20:657-663.
50•• Hewett M, Oliver DE, Rubin DL, et al. PharmGKB: the Pharmacogenetics Knowledge Base. Nucl Acids Res 2002; 30:163-165.
51 Klein TE, Chang JT, Cho MK, et al. Integrating genotype and phenotype information: an overview of the PharmGKB project. Pharmacogenomics J 2001; 1:167-170.
52 Oliver DE, Rubin DL, Stuart JM, et al. Ontology development for a pharmacogenetics knowledge base. Pac Symp Biocomput 2002; 7:65-76.
53 Rubin DL, Hewett M, Oliver DE, et al. Automating data acquisition into ontologies from pharmacogenetics relational data sources using declarative object definitions and XML. Pac Symp Biocomput 2002; 7:88:-99.
54 Agodoa LY, Appel L, Bakris GL, et al. Effect of ramipril vs. amlodipine on renal outcomes in hypertensive nephrosclerosis: a randomized controlled trial. JAMA 2001; 285:2719-2728.
55 Davis BR, Cutler JA, Gordon DJ, et al. Rationale and design for the antihypertensive and lipid-lowering treatment to prevent heart attack trial (ALLHAT). Am J Hypertens 1996; 9:342-360.

hypertension; blood pressure; polymorphism; heredity; gene; antihypertensive; pharmacogenomics; pharmacogenetics; single nucleotide polymorphism; pharmacodynamics

© 2003 Lippincott Williams & Wilkins, Inc.