Journal of Human Hypertension (1999) 13, 225–229  1999 Stockton Press. All rights reserved 0950-9240/99 $12.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE Genetic epidemiology of essential hypertension

I Gavras1, A Manolis1,2 and H Gavras1 1Hypertension and Atherosclerosis Section of the Department of Medicine, Boston University School of Medicine, Boston, MA, USA; 2Tzanio Hospital, Piraeus, Greece

This review article is intended to introduce the unini- heritable anthropometric, clinical or biochemical tiated clinician to the basic concepts, aims and early characteristics; and some applications of genetic epi- findings of the genetic epidemiology of hypertension. It demiologic techniques, such as linkage and association separates the rare monogenic ‘Mendelian’ hypertensive studies of certain gene polymorphisms with hyperten- disorders from the vast majority of patients with essen- sion using affected sibling pairs and large sibships or tial hypertension, which is a complex, polygenic, multi- wide genomic screens comparing affected and unaffec- factorial disorder resulting from interaction of several ted populations. Although so far there is no genotypic genes with each other and with the environment. It high- variation proven to be causally related to essential lights some clinical strategies used to enhance hypertension, its intermediate phenotypes or any of its searches for ‘candidates genes’, such as subgrouping complications, it is hoped that new, more efficient of populations into relatively homogenous groups or methods of genetic analysis will yield clinically mean- ‘intermediate phenotypes’ according to presumably ingful information.

Keywords: gene polymorphisms; candidate genes; monogenic disorders; intermediate phenotypes; polygenic multifactorial traits

Hypertensive phenotypes the interaction of several genes with each other and with the environment result in different phenotypes. Essential hypertension results from the interaction Hypertension is often associated with insulin of hereditary and environmental factors and is one resistance ranging from borderline fasting of the leading causes of premature cardiovascular hyperglycaemia/hyperinsulinaemia to clinically morbidity and mortality. Blood pressure levels are overt diabetes mellitus type II. The frequent coexist- highly correlated among family members, a fact ence of the two conditions has suggested a common attributable to either common genetic background or aetiologic background.4 If left untreated, hyperten- shared environment and lifestyle habits, or both. sion results in target organ complications, such as Studies of familial aggregation and rates of concord- congestive heart failure, stroke and renal failure, all ance among monozygotic and dizygotic twins sug- of which can be prevented to various extents by gest that genes may contribute as much as 40% of 1 optimal blood pressure control. Coronary artery dis- the risk to this disorder. High blood pressure, ease is also a frequent complication of both hyper- defined as systolic pressure у140 mm Hg and dia- у tension and diabetes, but appears to be less prevent- stolic 90 mm Hg, is characterised by a high but able by optimal blood pressure control.5 For these variable prevalence rate of 15–25% among popu- reasons, it is now accepted that arterial blood press- lations.2 But blood pressure does not have a bimodal ure elevation per se is only one of the factors con- distribution into normal and abnormal groups. tributing to target organ damage and does not con- Rather, it is a quantitative trait with wide pheno- sistently predict the occurrence or severity of typic variation and a continuous distribution, where hypertensive complications. In other words, hyper- an arbitrary cutoff line at some point of the Gaussian curve separates normal from abnormal. This indi- tension is too broad a phenotype. Over the years several efforts have been made to cates that hypertension is a complex trait, ie, a 5 phenotype that does not follow the classic Mendel- classify essential hypertensive into subtypes or ian rules of dominant or recessive inheritance ‘intermediate phenotypes,’ ie, subgroups homo- attributable to a single gene locus.3 More likely, it is geneous for one or more traits that might have diag- a polygenic and multifactorial disorder, in which nostic or prognostic significance. Various clinical or biochemical characteristics have been chosen in an effort to better define patients at risk of compli- cations. An early such effort was the classification Correspondence: Dr Haralambos Gavras, Chief, Hypertension & into renin subgroups (ie, patients with high, normal Atherosclerosis Section, Boston University School of Medicine, 715 Albany Street, Boston, MA 02118, USA or low plasma renin activity in relation to their Received 15 October 1998; revised 16 November 1998; accepted blood pressure and state of sodium balance). It was 10 December 1998 shown in both retrospective and prospective surveys Genetic epidemiology of essential hypertension I Gavras et al 226 that high-renin patients are at increased risk of myo- partly responsible for these findings, it is felt that cardial infarction,6,7 a finding in agreement with the genetically determined characteristics play an equal demonstration that excess angiotensin II of exogen- or more important role.11–14 Differences in the bio- ous or endogenous origin can produce myocardial chemical and endocrine profiles of black hyperten- infarcts.8,9 sives suggest different pathophysiologic mech- A trait deemed to be useful for classification into anisms than in white hypertensives.15 A prime subgroups that would enhance subsequent genetic example of such a biological difference is the analysis, should be pathogenically meaningful and increased frequency and magnitude of salt sensi- show evidence of heritability. Because essential tivity (ie, magnitude of blood pressure rise in hypertension is a complex disorder, in which many response to a sodium load), which is encountered different genes at various loci are likely to be in about three-quarters of hypertensive black involved, it is expected that phenotypically different patients.16 Salt sensitivity is associated with dimin- subjects may present with different constellations of ished ability to excrete a sodium load, and hence gene mutations. Thus, searching for underlying gen- tendency to retain salt, which is attributable to a var- etic causes might be more likely to be fruitful in sub- iety of mechanisms, including alterations in cation populations classified according to intermediate transport across cell membranes, abnormal renal phenotypes. Presumably heritable clinical charac- haemodynamics, and decreased generation of intra- teristics, that have been used to subdivide the hyper- renal natriuretic substances such as kinins, atrial tensive population, include obesity (defined by the natriuretic peptides, prostaglandins and dopamine. body mass index), abdominal adiposity (defined by Black hypertensives also tend to exhibit differences the waist-to-hip ratio), early onset of hypertension in other endocrine characteristics, such as low (eg, before age 50 years), excessive severity of hyper- plasma renin activity, higher levels of endothelin,17 tension (eg, requiring triple drug regimen), hyperres- augmented alpha-adrenergic receptor reactivity, ponsiveness to adrenergic stimuli, salt sensitivity, higher levels of plasma vasopressin (AVP) and etc. Biochemical traits include elevated fasting glu- accentuated hypotensive response to an antivaso- cose and insulin levels, elevated circulating levels pressor AVP antagonist.18 It has also been proposed, of pressor hormones such as renin, catecholamines but not uniformly accepted, that black subjects have and vasopressin, altered characteristics of alpha- relative insulin resistance.15 Of particular interest is adrenergic receptors, decreased activity of the kal- the fact that some of these racial differences have likrein-kinin system, different hormonal responses been observed not only in black hypertensives but to salt loading, etc. Several of these traits have been also in black normotensives compared to age and shown (or are predicted) to coexist in the same indi- gender-matched Caucasians. For example, in a nor- vidual. For example, low-renin individuals are motensive population, lower levels of urinary kal- likely to have low kallikrein excretion and high cir- likrein were reported not only in black vs Caucasian culating vasopressin. Other biochemical parameters, adults19 but also in children.20 In fact, it is possible such as angiotensin-converting enzyme and angio- that low levels of kallikrein might be one reason for tensinogen levels, lipid profile, etc, have also been lack of a local renoprotective effect of kinins and reported to differ among hypertensive populations increased propensity of black hypertensives to renal and possibly contribute to cardiovascular risk. damage, even under good blood pressure control. A common intermediate phenotype is the so- There is strong clinical evidence suggesting that, called metabolic syndrome X. This consists of a con- although salt-sensitivity tends to increase with age, stellation of clinical and biochemical character- it has a large heritable component.21 The high preva- istics, including hypertension, abdominal adiposity lence of salt-sensitivity described in African Amer- (‘apple shape’), hyperlipidaemia, hyperglycaemia icans appears to be also present in native Africans, and hyperinsulinaemia.4 Each one of these charac- as suggested by clinical observations in South Africa teristics is an independent coronary risk factor and (Dr Naiker, personal communication). At this time their coexistence multiplies the risk. Although its it is unknown which traits make black subjects more prevalence increases with age, this syndrome prone to hypertension and to hypertensive appears to have a clear genetic component, as the nephrosclerosis. If genetic studies could identify coexistence of hypertension and diabetes is twice as specific characteristics relevant to the pathogenesis common in African Americans and three times of hypertension or its complications in this popu- as common in Mexican Americans compared to lation, preventive treatments could be developed whites.10 which address these characteristics (eg, enzymatic enhancement of renal kinin production in an effort Hypertension in black people to reduce the excessive rates of hypertension-related end-stage renal disease). The prevalence of essential hypertension in black people is at least twice that of whites.2 It is charac- Molecular genetics of essential terised by an earlier onset, more severe course and increased susceptibility to some hypertension- hypertension related complications, such as left ventricular Although rare pedigrees have been discovered hypertrophy, stroke and end-stage renal disease in which display Mendelian forms of hypertension, the comparison to white hypertensives.11–13 Although essential hypertension encountered in over 95% of lack of optimal blood pressure control due to socio- the hypertensive population is unlikely to be economic, dietary and behavioural factors may be attributable to a single ‘hypertension gene’. It is Genetic epidemiology of essential hypertension I Gavras et al 227 more likely that a number of genes, whose products failed to demonstrate linkage by sib-pair analysis.45 participate in the pathophysiology of hypertension Polymorphisms in the insulin receptor gene have by influencing various biological factors, would also been associated with hypertension.46 account for the genetic variability of blood pressure In a series of studies from Europe it was reported levels. Such ‘candidate genes’ include those enco- that genetic variants in the ACE gene are important ding for various components of the renin-angioten- in determining plasma concentrations of ACE.47,48 sin-aldosterone system, the kallikrein-kinin system, Subsequently, the same investigators observed that the cation-transport systems regulating flux across one of these variants (which is determined by a cell membranes, the sympathetic nervous system, 287 bp deletion) was significantly more frequent in various vasoactive autacoids, neurotransmitters, etc. patients who had sustained a myocardial infarct;49,50 Animal models have provided valuable insights moreover, the association between high levels of into the genetics of hypertension in humans. For ACE and coronary events was strongest in people example, in strains of Dahl-S and spontaneously who would otherwise be classified as low-risk by the hypertensive rats (believed to be the experimental usual clinical criteria. Another team of investigators counterpart of human essential hypertension), found a higher frequency of homozygotes for the investigators have linked blood pressure to the deletion-allele in subjects with cardiac hypertro- genes for renin, angiotensin-converting enzyme phy,51 which is itself an independent coronary risk (ACE) and atrial natriuretic peptide receptor.22–25 factor. Even if these findings are verified in other However, several linkage and association studies of populations, they would not indicate whether the human renin and ACE loci have given negative excess ACE is a causative factor or merely a marker results,26–31 except for one study in a black popu- of coronary risk, and, if causative, whether its effects lation, which found a possible association between are due to abundant formation of angiotensin II or blood pressure and a restriction fragment length to excessive degradation of bradykinin.52 These data polymorphism (RFLP) in the renin gene.32 The SA illustrate the problems with attributing a causal role gene has also shown some promise as a candidate to genetic variants identified via statistical analysis gene for hypertension. This gene was first identified of affected versus normal populations. Various stra- as a candidate for rat hypertension, as it showed tegies are employed to further dissect these findings. hyperexpression in the kidneys of spontaneously In recent years, a small subset of patients that used hypertensive rats compared to the normotensive to be classified as essential hypertensives, have now control strain.33 Subsequent genetic analysis has been found to have familial hypertension with Men- shown that the SA genotype conveys a significant delian patterns of inheritance, that is attributable to effect on blood pressure in rats and the SA gene was a specific single genetic mutation. The first such associated with hypertension in a study of Japanese pedigree was a family with glucocorticoid-remedi- hypertensive patients,34 but not Caucasian hyperten- able hypertension, a rare type of mineralocorticoid sives.35 Likewise, a genetic mutation of the ␣-addu- excess caused by a chimaeric 11-␤-hydroxy- cin gene (a substance involved in ion transport and lase/aldosterone synthase gene.53 Other such rare signal transduction in cells) was found to be associa- single gene mutations have been found to cause the ted with hypertension in a strain of hypertensive syndrome of apparent mineralocorticoid excess,54 rats36 and in an Italian hypertensive population37 Liddle’s syndrome of pseudoaldosteronism,55 Gor- but not in a Scottish population.38 don’s syndrome of pseudohypoaldosteronism56 and Linkage and association studies between variants a syndrome characterised by hypertension, short of the angiotensinogen (AGT) gene and hypertension stature, brachydactyly and excessive propensity to have produced contradictory results.39–43 Linkage strokes.57 Additional such ‘Mendelian pedigrees’ between polymorphism of the AGT gene and human may be discovered in the future, but so far they have hypertension was initially reported in several differ- not helped explain the mechanisms of essential ent geographically separated populations.39–40 Two hypertension. Indeed, several investigators have specific variants in the AGT gene were found to cor- explored a possible role of mutations or variants of relate with higher circulating AGT levels and to be the epithelial sodium channel (similar to those caus- significantly associated with hypertension in one ing Liddle’s syndrome) in the pathogenesis of essen- study that compared allele frequencies of hyperten- tial hypertension, but found no such evidence.58,59 sives and normotensives from France and Utah, What they did find was significant racial differences USA39 and in another from Japan.40 However, in a in the distribution of these variants59,60 which again similar analysis of hypertensives and normotensives underlines the risks of pooling populations from dif- in the UK no association was found between either ferent genetic backgrounds.3 one of these AGT variants and hypertension.41 Fur- In conclusion, so far there is no genotype proven thermore, a repeat analysis of pooled large numbers to be causally related to essential hypertension or to of subjects from several European populations failed one of its intermediate phenotypes. At most, it could to confirm the original findings.42 In African Amer- be stated that some of the polymorphisms described icans the frequency of the M235T variant was far above may confer increased susceptibility to hyper- greater than that of Caucasians but was not associa- tension or its complications in certain populations. ted with hypertension.43 Interestingly, this AGT Further progress is expected via continuing use of variant was reported to be statistically associated molecular genetic techniques, such as genomic with pre-eclampsia.44 Another study found an screens for detection of additional polymorphisms association between a polymorphism of the angiot- in other candidate genes, identification of new gen- ensin II receptor gene (AT1) and hypertension, but etic markers and systematic genetic mapping of spe- Genetic epidemiology of essential hypertension I Gavras et al 228 cific chromosomal regions. When the Human Gen- likrein concentration in childhood: relation to blood ome project is completed, it is expected to provide pressure, race, and urinary electrolytes. Am J Epide- a list of all genes located in each chromosomal miol 1976; 104: 124–132. region and hence further facilitate these efforts. 21 Svetkey LP, McKeown SP, Wilson A. Heritability of salt sensitivity in black Americans. Hypertension 1996; 28: 854–858. References 22 Kurtz TW et al. Cosegregation of the renin allele of the 1 Ward R. Familial aggregation and genetic epidemi- spontaneously hypertensive rat with an increase blood ology of blood pressure. In: Laragh JH, Brenner BM pressure. J Clin Invest 1990; 85: 1328–1332. (eds). Hypertension: pathophysiology, diagnosis and 23 Hilbert P et al. Chromosomal mapping of two genetic management. Vol 1. Raven Press: New York, 1990, pp loci associated with blood pressure regulation in her- 81–100. editary hypertension in rats. Nature 1991; 353: 521– 2 Burt VL et al. Trends in the prevalence, awareness, 529. treatment, and control of hypertension in the adult US 24 Jacob HJ et al. Genetic mapping of a gene causing population. Hypertension 1995; 26: 60–69. hypertension in the strokeprone SHR. Cell 1991; 67: 3 Lander ES, Schork NJ. Genetic dissection of complex 213–224. traits. Science 1994; 265: 2037–2048. 25 Deng Y, Rapp J. Cosegregation of blood pressure with 4 Reaven GM. Role of insulin resistance in human dis- angiotensin converting enzyme and atrial natriuretic ease. Diabetes 1988; 37: 1595–1607. peptide receptor genes using Dahl salt-sensitive rats. 5 MacMahon SW et al. The effects of drug treatment for Nature Genet 1992; 4: 267–272. hypertension on morbidity and mortality from cardio- 26 Naftilan AJ et al. A lack of genetic linkage of renin vascular disease: a review of randomized controlled gene restriction fragment length polymorphisms with trials. Progr Cardiovasc Dis 1986; 29 (Suppl 1): 99–118. human hypertension. Hypertension 1989; 14: 219–224. 6 Brunner HR et al. Essential hypertension. Renin and 27 Soubrier F et al. Similar Frequencies of renin gene aldosterone, heart attack and stroke. N Engl J Med restriction fragment length polymorphisms in hyper- 1972; 286: 441–449. tensive and normotensive subjects. Hypertension 7 Alderman MH et al. Association of the renin-sodium 1990; 16: 712–717. profile with risk of myocardial infarction in patients 28 Jeunemaitre X et al. Sib-pair linkage analysis of renin with hypertension. N Engl J Med 1991; 324: 1098– gene haplotypes in human essential hypertension. 1104. Hum Genet 1992; 88: 301–306. 8 Gavras H et al. Acute renal failure, tubular necrosis 29 Jeunemaitre X et al. Absence of linkage between the and myocardial infarction induced in the rabbit by angiotensin-converting enzyme locus and human intravenous angiotensin II. Lancet 1971; II: 19–22. essential hypertension. Nature Genet 1992; 1: 72–75. 9 Gavras H et al. Angiotensin and norepinephrine- 30 Schmidt S et al. Polymorphism of the angiotensin con- induced myocardial lesions: experimental and clinical verting enzyme gene is apparently not related to high studies in rabbit and man. Am Heart J 1975; 89: 321– blood pressure; Dutch hypertension and offspring 332. study. J Hypertens 1993; 11: 345–348. 10 National High Blood Pressure Education Program 31 Harrap SB et al. The angiotensin I converting enzyme Working Group Report on Hypertension in Diabetes. gene and predisposition to high blood pressure. Hypertension 1994; 23: 145–158. Hypertension 1993; 21: 445–460. 11 Wallin JD. Hypertension in black patients. Epidemiol- 32 Barley J et al. Renin and atrial natriuretic peptide ogic and pathophysiologic considerations. J Clin Hyp- restriction fragment length polymorphisms; associ- ertens 1986; 1: 7–12. ation with ethnicity and blood pressure. J Hypertens 12 Dustan HP, Curtis JJ, Luke RG, Rostand SG. Systemic 1991; 9: 993–996. hypertension and the kidney in black patients. Am J 33 Iwai N, Inagami T. Identification of a candidate gene Cardiol 1987; 60: 731–771. responsible for the high blood pressure of spon- 13 Whittle JC, Whelton PK, Seidler AJ, Klag MJ. Does taneously hypertension rats. J Hypertens 1992; 10: racial variation in risk factors explain black-white dif- 1155–1157. ferences in the incidence of hypertension end-stage 34 Iwai N et al. Human SA gene locus is a candidate locus renal disease? Arch Intern Med 1991; 151: 1359–1364. for essential hypertension. Hypertension 1994; 23: 14 Gillum RF. Pathophysiology of hypertension in blacks 375–380. and whites. A review of the basic of racial blood press- 35 Nabika T et al. The SA gene and human essential ure differences. Hypertension 1979; 1: 468–475. hypertension. Hypertension 1994; 24: 374 (Abstract). 15 Falkner B. Differences in blacks and whites with ␣ essential hypertension: biochemistry and endocrine. 36 Cusi D et al. -Adducin polymorphism in primary Hypertension 1990; 15: 681–686. hypertension: linkage and association studies – 16 Luft FC, Grim C, Fineberg N, Weinberger M. Effects relationship to salt sensitivity. Lancet 1997; 349: of volume expansion and contraction in normotensive 1353–1358. whites, blacks and subjects of different ages. Circu- 37 Bianchi G et al. Two point mutations within the addu- lation 1979; 59: 653–660. cin genes are involved in blood pressure variation. 17 Ergul S, Parish DC, Puett D, Ergul A. Racial differences Proc Natl Acad Sci USA 1994; 91: 3999–4003. ␣ in plasma endothelin-1 concentrations in individuals 38 Kamitani A et al. Human -adducin gene, blood press- with essential hypertension. Hypertension 1996; 28: ure, and sodium metabolism. Hypertension 1998; 32: 652–655. 138–143. 18 Bakris G et al. Role of vasopressin in essential hyper- 39 Jeunemaitre X et al. Molecular basis of human hyper- tension: racial differences. J Hypertens 1997; 15: tension: role of angiotensinogen. Cell 1992; 71: 169– 545–550. 180. 19 Levy SB, Lilley JJ, Frigon RP, Stone RA. Urinary kal- 40 Hata A et al. Angiotensinogen is a risk factor for essen- likrein and plasma renin activity as determinants of tial hypertension in Japan. J Clin Invest 1994; 93: renal blood flow. J Clin Invest 1977; 60: 129–138. 1285–1287. 20 Zinner SH et al. Familial aggregation of urinary kal- 41 Caulfield M et al. Linkage of the angiotensinogen gene Genetic epidemiology of essential hypertension I Gavras et al 229 to essential hypertension. N Engl J Med 1994; 330: polymorphism of the angiotensin-converting-enzyme 1629–1623. gene and left ventricular hypertrophy. N Engl J Med 42 Brand E et al. Evaluation of the angiotensinogen locus 1994; 330: 1634–1638. in human essential hypertension: a European study. 52 Gavras H. Angiotensin-converting enzyme inhibition Hypertension 1998; 31: 725–729. and the heart. Hypertension 1994; 23: 813–818. 43 Rotimi C et al. Angiotensinogen gene in human hyper- 53 Lifton RP et al. A chimaeric 11-hydroxylase/ tension. Lack of an association of the 235T allele aldosteronesynthase gene causes glucocorticoid- among African Americans. Hypertension 1994; 24: remediable aldosteronism and human hypertension. 591–594. Nature 1992; 355: 262–265. 44 Ward K et al. A molecular variant of angiotensinogen 54 Mune T et al. Human hypertension cused by ␤ associated with preeclampsia. Nature Genet 1993; 4: mutations in the kidney isozyme of 11 -hydroxys- 59–61. teroid dehydrogenase. Nature Genet 1995; 10: 394– 45 Bonnardeaux A et al. Angiotensin II Type 1 receptor 399. gene polymorphisms in human essential hypertension. 55 Shimkets RA et al. Liddle’s syndrome: heritable Hypertension 1994; 24: 63–69. human hypertension caused by mutations in the beta 46 Ying LH, Zee RY, Griffiths LR, Morris BJ. Association subunit of the epithelial sodium channel. Cell 1994; 79: 407–414. of a RFLP for the insulin receptor gene, but not insulin, 56 Mansfield TA et al. Linkage of familial hyperkalemia with essential hypertension. Biochem Biophys Res and hypertension, pseudohypoaldosteronism Type II, Commun 1991; 181: 486–492. to 1q3–42 and 17p11–q21. Hypertension 1997; 30 (Part 47 Cambien F et al. Familial resemblance of plasma angi- 1): 475 (Abstract). otensin-converting enzyme level: The Nancy study. 57 Schuster H et al. Severe autosomal dominant hyper- Am J Genet 1988; 43: 774–780. tension and brachydactyly in a unique Turkish kin- 48 Rigat B et al. An insertion/deletion polymorphism in dred maps to human chromosome 12. Nature Genet the angiotensin I-converting enzyme gene accounting 1996; 13: 98–100. for half the variance of serum enzyme levels. J Clin 58 Melander O et al. Mutations and variants of the epi- Invest 1990; 86: 1343–1346. thelial sodium channel gene in Liddle’s Syndrome and 49 Cambien F et al. Deletion polymorphism in the gene primary hypertension. Hypertension 1998; 31: 1118– for angiotensin-converting enzyme is a potent risk fac- 1124. tor for myocardial infarction. Nature 1992; 359: 641– 59 Persu A et al. Genetic analysis of the ␤ subunit of the 644. epithelial Na+ channel in essential hypertension. 50 Tiret L et al. Deletion polymorphism in angiotensin- Hypertension 1998; 32: 129–137. converting enzyme gene associated with parental his- 60 Baker EH et al. Association of hypertension with tory of myocardial infarction. Lancet 1993; 341: 991– T594M mutation in beta subunit of epithelial sodium 992. channels in black people resident in London. Lancet 51 Schunkert H et al. Association between a deletion 1998; 351: 1388–1392.