Jhh$$$987p

Journal of Human Hypertension (2001) 15, 425–430
 2001 Nature Publishing Group All rights reserved 0950-9240/01 $15.00
www.nature.com/jhh
ORIGINAL ARTICLE
T594M and G442V polymorphisms of the
sodium channel ␤ subunit and
hypertension in a black population
YB Dong1,2, HD Zhu1, EH Baker1, GA Sagnella1, GA MacGregor1, ND Carter2,PD Wicks1,3, DG Cook3 and FP Cappuccio4Departments of 1Physiological Medicine, 2Child Health, 3Public Health Sciences, 4General Practice &Primary Care, St George’s Hospital Medical School, London, UK Polymorphisms of the epithelial sodium channel may
ing to increasing blood pressure and presence or
raise blood pressure by increasing renal sodium reab-
absence of drug therapy. The frequency of the 594M
sorption. This study examines frequency distributions
variant (heterozygotes and homozygotes) was 4.6%; the
and associations with hypertension of the T594M and of
frequency of the 442V variant was higher (27.0%). The
the G442V polymorphisms of the ␤ subunit of the epi-
frequency of the 594M variant increased with increasing
thelial sodium channel in a population-based sample.
blood pressure category (P ؍ 0.05) and was more com-
We studied a stratified random sample of 459 subjects
mon in hypertensives than normotensives. By contrast
(279 women), aged 40–59 years, of black African origin
the frequency of the 442V variant did not vary across
from general practices’ lists within a defined area of
increasing blood pressure categories (P ؍ 0.62). No
South London. All were first generation immigrants. The
gender difference was observed. Adjustment for age,
polymorphic variants were detected using single strand
sex and body mass index did not alter these findings.
conformational polymorphism technique (SSCP). The
These results suggest that the 594M variant may con-
prevalence of hypertension (BP у160 and/or 95 mm Hg
tribute to high blood pressure in black people of African
or on drug therapy) was 43%; of these, 76% were on
origin whereas the G442V polymorphism is unlikely to
drug therapy. The main analysis was carried out by
influence blood pressure in this population.
three ordered blood pressure categories (I to III) accord-
Journal of Human Hypertension (2001) 15, 425–430
Keywords: genetics; epidemiology; black people of African origin; sodium channel
Introduction
sensitive to changes in sodium intake.8 Previousstudies have shown that black people have a slower Hypertension is common in black populations of sodium excretion in response to intravenous sodium African descent living in urbanised environments.1 infusion suggesting the possibility of a defect in the It is a particular problem in Africans and Caribbeans control of renal sodium excretion. Moreover, there living in the United Kingdom1,2 with stroke and end- is now evidence of a strong heritable component of stage renal failure being the major causes of death salt-sensitivity in blacks.9 These observations there- and disability.3–6 In a recent population-based sur- fore suggest the possibility of a molecular defect in vey of cardiovascular risk factors in ethnic minority groups we found that nearly 50% of middle-aged Similarities in some phenotypic expressions black people (both West Africans and Afro- between hypertension in blacks and patients with Caribbeans) had high blood pressure requiring treat- Liddle’s syndrome,10 a monogenic form of hyperten- sion, have suggested that abnormalities of the distal High blood pressure in blacks is associated with tubular epithelial sodium channel—a major regu- low plasma renin activity, an index of sodium reten- lator of the overall control of sodium balance—may tion and volume expansion, and appears to be more underlie the development of high blood pressure inblacks.11 A number of variants of the sodium chan-nel ␤ subunit coding sequences have been identified Correspondence: Dr GA Sagnella (Physiological Medicine) or in subjects with hypertension.12–14 These lead to a Professor FP Cappuccio (General Practice and Primary Care), StGeorge’s Hospital Medical School, Cranmer Terrace, London single amino acid change rather than a major trunc- ation as seen in Liddle’s mutations and are much E-mail: g.sagnellaȰsghms.ac.uk or f.cappuccioȰsghms.ac.uk more frequent in blacks than in whites, especially Received 13 July 2000; revised and accepted 27 November 2000 the most commonly identified T594M and G442Vpolymorphisms.12 Sodium channel polymorphisms in a black population
In a previous case-control study we found that the them for checking. Anthropometry and blood press- 594M variant of the last exon (exon 12) of the ␤ sub- ure were taken with standardised methods as unit of the epithelial sodium channel was more fre- described elsewhere.7 Fasting venous blood was quent in hypertensive than normotensive blacks.15 taken in the seated position without stasis for serum In this study we also found that black people with biochemistry and plasma aldosterone measure- the 594M variant had suppressed plasma renin ment.7 A questionnaire was administered which activity when compared to wild-type individuals included personal medical history and drug treat- suggesting a state of corrected volume expansion.15 ment. After the interview, a complete 24-h urine col- It is not known, however, whether other sodium lection was obtained as previously described7 in 323 channel variants are also associated with high blood out of the 459 participants (70.4%). As the number pressure in blacks and whether the presence of two of participants on antihypertensive medication was or more sodium channel variants have a greater high (n = 149), the main analysis was carried out effect on sodium transport and blood pressure.
using three ordered blood pressure (BP) categories: The G442V polymorphism, a mutation of exon 8 I (BP Ͻ140 mm Hg and Ͻ90 mm Hg and not on drug of the sodium channel ␤ subunit, is of particular therapy); II (BP у140 and Ͻ160 mm Hg or у90 and interest as it was previously identified in 10 out of Ͻ95 mm Hg and not on drug therapy); III (BP у160 50 black hypertensives of African origin and in only one out of 475 hypertensive whites.12,16 The specific therapy). Further analyses using measured blood objectives of the present study, therefore, were to pressure values were carried out in people not on extend our previous work on the T594M and to regular drug therapy for hypertension.
investigate the association with blood pressureacross a range of increasing blood pressure levels of Genetic analysis
both the T594M and the G442V polymorphisms ofthe ␤ subunit of the epithelial sodium channel in a Genomic DNA was isolated from whole blood as large population-based sample of middle-aged black previously described using Nucleon BACC DNA people of African origin living in London.
extraction kit.17 The 594M and 442V variants weredetected using polymerase chain reaction (PCR) andsingle strand conformational polymorphism (SSCP) Subjects and methods
analysis. PCR and SSCP conditions for detection of Population
the 594M variant were as previously established.15,18For this variant, a subset of the individuals in Group The population sample was obtained as previously I (80%) was included in our previous clinic-based described.2,7 In brief, the Wandsworth Heart & comparison,15 however, for the present study, these Stroke Study2,7 is a population-based cross-sectional were re-genotyped as part of the whole population- survey of 1577 men and women aged 40–59 years based sample. For detection of the 442V variant, obtained from age and sex-registers of general prac- 100 ng of genomic DNA was amplified using a sense titioners in a defined area of South London. Field- primer (5Ј-CTCTTGGCCGCCTTTCTG-3Ј) and an work was undertaken from March 1994 to July 1996.
antisense primer (5Ј-ATGCCTGCCCGCTGCTGTGC- Ethnic group was recorded at the time of the inter- 3Ј) in a total volume of 25 ␮l reaction mixture con- view, based on the answers to a combination of taining 15 pmol of each primer, 200 ␮M dNTPs, 1.5 questions including country of birth, language, reli- mM MgCl2 and 0.25 U Redhot DNA polymerase gion, history of migration and parental country of (Advanced Biotechnologies). After an initial denat- birth.7 All black participants of African origin (both uration at 94°C for 5 min, amplification was carried West Africans and Afro-Caribbeans) were first gener- out by 35 cycles at 94°C (1 min), 62.5°C (1 min) and ation immigrants. The study protocol was approved 72°C (30 s), and a final extension at 72°C for 5 min.
by the local Ethics Committee. All participants gave For detection of the variants by SSCP, 4 ␮l aliquots their informed consent to participate. The T594M of each PCR product were denatured by addition of polymorphism was determined in 458 (83.4%) and 8 ␮l denaturing solution (94% formamide, 10 mM that of the G442V polymorphism in 459 (83.6%) out NaOH, 0.25% bromophenol blue) and heating to of the 549 (279 women) black people who took part 94°C for 5 min, followed by rapid cooling on ice.
in this study. The baseline characteristics of those Samples were separated at 4°C by electrophoresis at not genotyped were comparable to those who were 250V for 10 h on 0.8 × MDE (Flowgen) gels prepared in 0.6 × TBE and run in 0.6 × TBE running buffer.
The DNA was visualised by silver staining. The Protocol
ability of the SSCP method to detect the G442Vpolymorphism was confirmed by direct sequencing Participants attended a dedicated screening unit at during the development and validation of this SSCP St George’s Hospital between 08.00 and 12.00, after method. For this, PCR products were purified with an overnight fast. They were asked to refrain from spin columns (QIAGEN) and subjected to direct smoking and from taking vigorous exercise for at sequence analysis of both strands employing a dye least 1 h before the visit and to bring all drugs with terminator kit on an ABI 377 automated sequencer.
Sodium channel polymorphisms in a black population
YB Dong et al
To prevent observer bias, those who carried out the separately, with OR of 3.93 (95% CI 0.69–29.1) and genotyping (YBD and HDZ) were unaware of sample 1.68 (0.49–5.82)), respectively when comparing origin and all gels were cross-checked by a separate hypertensives (category III) with the other two groups (I+II) combined. Amongst Caribbeans (n =289) the frequency of the 594M variant was 5.1% innormotensives (I + II) vs 6.1% in hypertensives (III).
Statistical methods
Amongst West Africans (n = 170) the frequencies Allele frequencies were tested for Hardy–Weinberg were 0% vs 6.7% for normotensives and hyperten- equilibrium using chi-square tests to compare sives, respectively. However, due to the small num- observed against expected frequencies with df=1.
bers, this sub-group analysis has limited statistical Odds ratios (OR) and exact 95% confidence inter- power. The analysis of the 11 untreated individuals vals (95% CI) were calculated using Epi-Info. Chi- out of the 21 possessing the 594M variant did not square tests for trend in proportions with 1 degree show any significant difference in weight, body of freedom were used to assess the relationship mass index, serum electrolytes, creatinine and between genotype and blood pressure category.19 plasma aldosterone compared to those untreated Further analyses allowing for confounders were car- without the 594M variant (Table 3). Twenty-four ried out using logistic regression with the SPSS-PC hour urinary sodium excretion also did not differ 8.0, in which hypertensives (category III) were com- between hypertensives (n = 145) and normotensives pared with normotensives (categories I+II).
(n = 178) (168 ± 68 vs 167 ± 65 mmol/24 h, mean ±s.d., P = 0.88) or between those with (n = 18) and without (n = 305) the 594M variant (177 ± 74 vs 166 Demographic and other characteristics of the whole The frequency of the G442V polymorphism was population sample by gender, adjusted for age are much higher than that of the T594M. The genotype given in Table 1. Men (n = 180) were slightly older (GG, GV, VV) frequencies were 72.9%, 24.8% and than women (n = 279) (52.0 ± 5.6 [mean ± s.d.] vs 2.2% with allele frequencies of 85.4% and 14.6%, 50.5 ± 5.8 years; P = 0.006). The prevalence of hyper- respectively. However, in contrast with the fre- tension (defined as category III) was 43% (196/459) quency of the 594M mutation, the frequency of the and was comparable in men and women (P = 0.98).
442V mutation was comparable in each blood press- The frequency of the 594M variant (heterozygotes ure category (Table 2). Amongst those not on treat- plus homozygotes) was 4.6%. The genotype (TT, ment, there were no significant differences in TM, MM) frequencies were 95.4%, 4.4% and 0.2% anthropometry, blood pressure and biochemistry with allele frequencies of 97.6% and 2.4%, respect- according to G442V polymorphism (Table 4).
ively. The 594M variant was found in 21 individ- Adjustment for age, sex and body mass index did uals, only one of whom was homozygous. The fre- not alter the pattern of the associations with hyper- quency of the variant increased with increasing tension of either the T594M or the G442V polymor- blood pressure category (chi-square for trend, P = 0.05) and was more common in the hypertensive Both variants were in Hardy–Weinberg equilib- rium and were not in linkage disequilibrium (counts Higher frequencies of the M allele amongst hyper- for genotype combinations were: TT/GG = 314; tensives was also seen in men and women analysed TT/GV = 112; TT/VV = 10; TM/GG = 18; TM/GV =2; TM/VV = 0; MM/GG = 1; MM/GV = 0; MM/VV =0). To examine the possibility of an interaction Table 1 Characteristics of men and women of black African ori-
between these two variants, the presence of hyper- tension was analysed by the possible genotype com-binations. Because of the small numbers in some subgroups only three genotype groups were ident- ified for analysis according to prevalence of hyper- tension (category III). In these, hypertension was present in 40.1% (126/314) in those with TT/GG genotype, 45.9% (56/122) in those with the 442V variant (TT/GV+TT/VV) and 68.4% (13/19) in those with the 594M variant (TM/GG+MM/GG) (chi- square = 6.49, P = 0.039). These results are consist- ent with hypertension being more common in those with the 594M variant but do not suggest any sig- nificant interaction between these two variants.
Discussion
Results are expressed as age-adjusted means (s.e.). an = 177 andn = 276 for men and women, respectively. bn = 169 and n = 270 In the present study we examined the association between two polymorphisms (T594M and G442V) in Sodium channel polymorphisms in a black population
Table 2 Association between the T594M and the G442V polymorphisms and blood pressure categories
aI (BP Ͻ 140 mm Hg and Ͻ90 mm Hg and not on drug therapy); II (BP у140 and Ͻ160 mm Hg or у90 and Ͻ95 mm Hg and not ondrug therapy); III (BP у160 mm Hg and/or у95 mm Hg or being on drug therapy). bChi-square for trend = 3.83, df = 1; P = 0.05). cChi-square for trend = 0.24, df = 1; P = 0.62.
Table 3 Characteristics by T594M genotype for untreated participants
Results are expressed as age- and sex-adjusted means (95% CI). ageometric mean. bn = 293 and cn = 285 in TT group.
Table 4 Characteristics by G442V genotype for untreated participants
Results are expressed as age- and sex-adjusted means (95% CI). ageometric mean. bn = 224, cn = 219 in GG group and n = 79 inGV+VV group.
the ␤ subunit of the epithelial sodium channel and the problems of selection bias and differential mis- high blood pressure in black people of African ori- classification which are potential causes of false gin. Our study has several important aspects in positive results in clinic-based case-control studies addressing genetic variations within a population.
The sample has been taken within the same geo- The present work makes two important contri- graphical area thereby mitigating the potential butions: it provides population-based frequency effects of differences in environmental background.
estimates of the T594M and the G442V polymor- It examines first generation immigrants with both phisms according to increasing blood pressure level parents born in the country of origin and belonging and it also examines whether the presence of both to the same ethnic background, thus reducing the the 442V and the 594M variants have a greater effect potential impact arising from an unknown degree of on blood pressure then when either is present alone.
admixture. Moreover, the design overcomes many of The G442V variant affects exon 8 of the sodium Sodium channel polymorphisms in a black population
YB Dong et al
channel ␤ subunit and it results in a single amino expansion than other black hypertensive people acid change with substitution of valine (GTC) for without this variant but further studies are required glycine (GGC) at amino acid 442. Exon 8 encodes a to elucidate the mechanisms whereby such suscepti- segment of the extracellular loop and changes in this bility is expressed and its potential therapeutic region would not be predicted to increase sodium channel activity. Indeed, while the 442V variant wasmore common than the 594M variant amongst black Acknowledgements
people, overall there were no significant differencesin the frequency of the 442V variant according to The Wandsworth Heart & Stroke Study has received blood pressure grouping. Moreover, there did not support from the Wandsworth Health Authority, SW appear to be any positive interaction in relation to Thames Regional Health Authority, NHS R&D Direc- high blood pressure between the 594M and the 442V torate, British Heart Foundation, British Diabetic variants, although the relatively small numbers and Association and The Stroke Association. Studies of the absence of some haplotypes, clearly limits the the genetics of the epithelial sodium channel are interpretation of the results regarding the presence supported by the British Heart Foundation. We of any interaction between these variants.
thank MJ Rothwell for technical assistance for the The observation of a significant trend of increas- DNA extraction. GAS, GAM, NDC, DGC and FPC are ing prevalence of the 594M variant across increasing members of the St George’s Cardiovascular Research blood pressure categories is consistent with our pre- Group. The study was presented at the Ninth Euro- vious case-control study of hypertensives compared pean Meeting on Hypertension, Milan, 1999.
to normotensives and suggests that its presence maycontribute to raise blood pressure across the range.
References
Two studies in black people have not shown anassociation between the T594M polymorphism and 1 Cappuccio FP. Ethnicity and cardiovascular risk: vari- hypertension. In a study of African Americans, the ations in people of African ancestry and South Asian overall frequency of the variant was comparable to origin. J Hum Hypertens 1997; 11: 571–576.
2 Cappuccio FP, Cook DG, Atkinson RW, Strazzullo P.
our own estimate but it was not more common Prevalence, detection and management of cardiovascu- amongst hypertensives.20 Furthermore, a linkage lar risk factors in different ethnic groups in South analysis of 63 affected sibling pairs in the Caribbean London. Heart 1997; 78: 555–563.
failed to show a linkage of the epithelial sodium 3 Balarajan R. Ethnic differences in mortality from channel genes to hypertension.21 The reasons for ischaemic heart disease and cardiovascular disease in these differences are still not clear but they may be England and Wales. BMJ 1991; 302: 560–564.
explained by the greater degree of genetic admixture 4 Roderick PJ et al. Population need for renal replace- of African Americans compared to first generation ment therapy in Thames regions: ethnic dimension.
immigrants to the UK and by the lack of statistical BMJ 1994; 309: 1111–1114.
5 Raleigh VS. Diabetes and hypertension in Britain’s eth- nic minorities: implications for the future of renal ser- The 594M variant affects the last exon of the C- vices. BMJ 1997; 314: 209–213.
terminal of the ␤ subunit of the epithelial sodium 6 Stewart JA et al. Ethnic differences in incidence of channel and results in a single amino acid change stroke: prospective study with stroke register. BMJ with substitution of methionine (ATG) for threonine 1999; 318: 967–971.
(ACG) at amino acid 594. The threonine residue at 7 Cappuccio FP, Cook DG, Atkinson RW, Wicks PD. The position 594 in the ␤ subunit is a potential consen- Wandsworth Heart and Stroke Study. A population- sus target site for phosphorylation by protein kinase based survey of cardiovascular risk factors in different C (PKC) which downregulates sodium channels.
ethnic groups. Methods and baseline findings. Nutr Therefore the 594M variant could increase sodium Metab Cardiovasc Dis 1998; 8: 371–385.
channel activity by causing affected channels to 8 He FJ, Markandu ND, Sagnella GA, MacGregor GA Importance of the renin system in determining blood become insensitive to negative regulation by PKC.22 pressure fall with salt restriction in black and white In this study there were no significant differences hypertensives. Hypertension 1998; 32: 820–824.
in plasma electrolytes and aldosterone according to 9 Svetkey LP, McKeown SP, Wilson AF. Heritability of T594M genotype although these comparisons may salt sensitivity in black Americans. Hypertension have limited statistical power due to the relatively 1996; 28: 854 –858.
small sample size. However, plasma aldosterone is 10 Shimkets RA et al. Liddle’s syndrome: heritable not a sensitive marker of volume status and the human hypertension caused by mutations in the beta effect on sodium balance remains a possibility. In subunit of the epithelial sodium channel. Cell 1994; fact, there is some evidence that the 594M variant 79: 407– 414.
may influence sodium balance in vivo as in our pre- 11 Schild L et al. A mutation in the epithelial sodium channel causing Liddle disease increases channel vious study15 plasma renin activity was significantly activity in the Xenopus laevis oocyte expression sys- lower in hypertensive black people with the 594M tem. Proc Natl Acad Sci USA 1995; 92: 5699–5703.
variant than in hypertensive blacks without this 12 Persu A et al. Genetic analysis of the beta subunit of variant. This suggests that individuals with the the epithelial Na+ channel in essential hypertension.
594M variant may have a greater degree of volume Hypertension 1998; 32: 129–137.
Sodium channel polymorphisms in a black population
13 Chang H, Fujita T. Lack of mutations in epithelial 18 Dong YB et al. Use of single-strand conformation poly- sodium channel ␤ subunit gene in human subjects morphism analysis to screen for the T594M sodium with hypertension. J Hypertens 1996; 14: 1417–1419.
channel mutation. Biochem Soc Trans 1998; 26: s393.
14 Dong YB et al. Mutations in the ␤ subunit of the epi- 19 Armitage P, Berry G. Statistical Methods in Medical thelial sodium channel in a London Caucasian stroke Research (2nd edn). Blackwell Scientific Publications: population. Hypertension 1997; 30: 1001 (Abstract).
15 Baker EH et al. Association of hypertension with 20 Su YR et al. A novel variant of the beta-subunit of the T594M mutation of ␤ sub-unit of epithelial sodium amiloride-sensitive sodium channel in African Amer- channels in black people resident in London. Lancet icans. J Am Soc Nephrol 1996; 7: 2543–2549.
1998; 351: 1388–1392.
21 Munroe PB et al. Absence of linkage of the epithelial 16 Dong YB et al. Sodium channel ␤ subunit G442V sodium channel to hypertension in black Caribbeans.
mutation in a London black hypertensive population.
Am J Hypertens 1998; 11: 942–945.
Hypertension 1998; 32: 808 (Abstract).
22 Cui Y et al. Loss of protein kinase C inhibition in the 17 Sagnella GA et al. A population study of ethnic vari- beta-T594M variant of the amiloride-sensitive Na+ ations in the angiotensin converting enzyme I/D poly- channel. Proc Natl Acad Sci USA 1997; 94: 9962–
morphism. Relationship with gender, hypertension and impaired glucose metabolism. J Hypertens 1999;
17: 657–664.

Source: http://www.warwick.ac.uk/fac/med/staff/cappuccio/publications/jhumhypert01whss.pdf