Angiotensinogen Promoter

Sequence Variants in Essential Hypertension

Digna R. Velez, Mallikarjunrao Guruju,
Govindaiah Vinukonda, Alicia Prater, Ashok Kumar, and Scott M. Williams
Background: Essential hypertension is a complex
multifactorial disease caused by ill-dened genetic factors.
The angiotensinogen (AGT) gene has been implicated as a
risk factor in essential hypertension.
Methods: To assess the role of AGT in hypertension,
we evaluated two polymorphisms (A-6G and C-20A) in
the 5= region of the gene that have been shown to have a
role in transcriptional regulation. A total of 463 subjects
were studied: 243 African Americans (26 male and 34
female normotensives, 66 male and 117 female hyperten-
sives) and 220 whites (35 male and 60 female normoten-
sives, 55 male and 70 female hypertensives). African
American and white subjects were examined individually,
as signicant differences in allele and genotype frequen-
cies were observed between these two cohorts.
Results: White female hypertensives and normoten-
sives differed signicantly in genotype frequency at C-20A
(P .02). No other single site comparisons were signif-
icantly different between hypertensives and normotensives
in either the white or African American samples. Haplo-
type frequencies in white males also differed signicantly
between phenotypic classes (P .05). To evaluate the
data further, we assessed all polymorphic sites simulta-
neously by the examination of multisite interaction and
determined the single best genetic model for each pop-
ulation. A model that included both sites and gender
correctly predicted hypertension status in the white popu-
lation 59.1% of the time (P .039). The model generated
for the African American population was not signi-
cant.
Conclusions: Our results suggest that a complex set of
genetic factors interact with gender to predispose whites to
hypertension. Am J Hypertens 2006;19:12781285
2006 American Journal of Hypertension, Ltd.
Key Words: Genetics of hypertension, angiotensino-
gen, ethnic disparity.
H
uman essential hypertension is a multifactorial
disease caused by both genetic and environmen-
tal variables.
1,2
Hypertension can lead to multiple
outcomes, including myocardial infarction, heart failure,
stroke, and renal failure. Signicant ethnic differences in
the prevalence of hypertension exist among adults in the
US. The prevalence of hypertension among whites is 27%
(30% among men and 24% among women).
3
The preva-
lence of hypertension among African Americans is 44%
(43% among men and 45% among women).
3
The disparity
between the prevalence of hypertension in African-Amer-
icans and whites has lead many investigators to hypothe-
size that genetic variation in populations of African and
European descent may explain some of the disparity in this
disease. However, the cause of these differences remains
unclear, but it is likely that both genetic and environmental
factors may be involved in the population-specic distri-
butions of human essential hypertension.
Despite the effects of hypertension on health, its patho-
physiology remains unclear. Nonetheless, it is clear that
genetic factors contribute to the development of hyperten-
sion. Several studies have indicated that the portion of the
variance in blood pressure (BP) that is accounted for by
genetic factors (ie, heritability) is signicant. Data suggest
that the heritability of BP in whites ranges from 25%
for diastolic BP to 45% for systolic BP.
4
In African
Americans and Africans data indicate that hypertension
has a signicant genetic component, with estimates of
heritability from 43% to 68% for either systolic or dia-
stolic BP.
4,5
Received January 27, 2006. First decision May 9, 2006. Accepted May
11, 2006.
From the Center for Human Genetics Research, Vanderbilt Univer-
sity, Nashville, Tennessee; Department of Pathology, New York Medical
College, Valhalla, New York; and Division of Cardiovascular Medicine,
Department of Medicine, Vanderbilt University, Nashville, Tennessee.
Digna R. Velez and Mallikarjunrao Guruju contributed equally to the
work on this article.
This work was supported by research grants HL 49884 and 59547
from the National Heart, Lung, and Blood Institute and from Philip-
Morris Incorporated (to AK).
Address correspondence and reprint requests to Dr. Scott M.
Williams, Center for Human Genetics Research, 519 Light Hall,
Vanderbilt University, Nashville, TN 37232; e-mail: smwilliams@chgr.
mc.vanderbilt.edu
AJH 2006; 19:12781285
0895-7061/06/$32.00 2006 by the American Journal of Hypertension, Ltd.
doi:10.1016/j.amjhyper.2006.05.020 Published by Elsevier Inc.
Although hypertension is likely to be affected by com-
binations of many genes, it is possible that variants within
a single gene interact to signicantly increase or decrease
the risk of hypertension. For this reason, we undertook a
study to examine genetic variation in the renin-angiotensin
system (RAS), a potential pathway leading to hyperten-
sion. Many of the symptoms associated with the pheno-
typic expression of hypertension have been correlated with
activation of the RAS in different tissues of the cardiovas-
cular system.
6
Substantial evidence suggests that genes
involved in RAS are logical candidates to study the ge-
netic risk involved in essential hypertension.
7,8
The focus of the present study was on one candidate
gene, angiotensinogen (AGT). The AGT gene encodes the
precursor molecule for angiotensin II (Ang II), the effector
peptide of the RAS. Evidence suggests that production of
Ang II in local tissues involved in the regulation of BP is
central to developing hypertension.
9
As a result, the AGT
gene has been extensively studied as a candidate gene in
association studies.
7,10,11
Published studies have produced
an extensive list of potential variants in the AGT gene that
may affect BP.
12,13
In the present study, white and African-American indi-
viduals diagnosed with hypertension and normotensive
controls were examined at two promoter variants (A-6G
and C-20A) of the AGT gene, which have been previously
shown to affect either gene expression or BP.
1416
The
A6G polymorphism affects the basal rate of AGT tran-
scription with the 6A variant causing an increase in
transcription and plasma AGT. The C-20A polymorphism
suppresses a consensus sequence for an estrogen receptor
element and reduces estrogen-induced promoter activity.
17
The 20A variant is also associated with increased
plasma AGT level in white women.
16
In the present study we tested for differences in single
site allele and genotype frequencies at both positions in the
AGT promoter. We also tested for multisite effects on
phenotype. In our study we found a signicant multisite
model predictive of hypertension in whites that includes
C-20A, A-6G, and gender. No signicant effects were
observed in African Americans.
Methods
Subjects
All subjects were recruited from the outpatient department
of The State University of New York Health Science
Center at Brooklyn, NY, and Westchester Medical Center,
Valhalla, NY. All subjects gave informed consent before
participating in the study. The research protocol was ap-
proved by the Institutional Review Board at New York
Medical College, Valhalla, NY. Cases were diagnosed as
having been previously diagnosed with essential hyperten-
sion; most were on medication before enrollment in the
study. Hypertension was dened as systolic BP 140 mm
Hg, diastolic BP 90 mm Hg, or taking antihypertensive
therapy. Blood pressure was measured twice with the
subject seated with 5 min between measurements. The
normotensives (participants with systolic BP/diastolic BP
140/90 mm Hg and without a history of hypertension)
were recruited from the same population. All partici-
pants completed a questionnaire on personal medical
history and family history of hypertension. Four hundred
sixty-three subjects were recruited into the study (308
hypertensive subjects and 155 normotensive controls).
There were 243 African Americans and 220 whites. The
hypertensive class included 70 white women, 117 African-
American women, 55 white men, and 66 African-Ameri-
can men. The normotensive class included 60 white
women, 34 African-American women, 35 white men, and
26 African-American men. The average age of the hyper-
tensives was 64.1 years and the average age of the nor-
motensives was 55.6 years.
Ethnicity was self-dened. It should be noted that al-
though there may exist frequency differences between
ethnic groups for multiple candidate genes for hyperten-
sion, such differences may not reect underlying genetic
predispositions to the phenotype. Therefore, potential dif-
ferences in genetic predispositions were analyzed in the
two groups separately.
Genotyping
Venous blood was collected in EDTA tubes. Genomic
DNA was obtained using the Puregene DNA Isolation Kit
(Gentra, Minneapolis, MN). Genomic DNA was amplied
to produce a 111-bp fragment of the human angiotensino-
gen gene. The sequence of the forward primer was 5=-
CAGGCAGCCTGGGAACAGCTCCATCCC-3=, and the
sequence of the reverse primer was 5=-GCTTACCTTCTG-
CTGTAGTA-3=. Polymerase chain reaction (PCR) was
done in 30-L reactions, including 100 ng of genomic
DNA, 15 pmol of each primer, 3 L of 10buffer 15 mM
MgCl
2
, 500 mM KCl, 1% Triton X-100, and 100 mM
Tris-HCl at pH 9.0, 1.5 mM of each dNTPs, and 0.8 U of
Taq polymerase (Promega Biotec, Madison, WI). Ampli-
cations were done in a Perkin Elmer GeneAmp PCR
system 2400 thermal cycler for 35 cycles with denatur-
ation at 94C for 20 sec, annealing at 60C for 20 sec, and
extension at 72C for 20 sec. The nucleotide sequence of
the amplied DNA was conrmed by sequence analysis.
Restriction enzyme SfaNI cleaves the amplied fragment
when nucleotide A was present at 20 and restriction en-
zyme EcoO109I cleaves the fragment when nucleotide C
is present at 20. Similarly, restriction enzyme BstNI
recognizes a site when nucleotide A is present at 6 and
EagI cleaves the DNA when nucleotide G is present at the
6 site. After restriction enzyme treatment, the reaction
mixture was analyzed by 3% metaphor gel electrophoresis.
Statistical Analysis
Single site allele frequency analyses were performed using
Tools for Population Genetic Analysis (TFPGA, version
1.3 available at http://bioweb.usu.edu/mpmbio/index.htm),
1279 AJHDecember 2006VOL. 19, NO. 12 ANGIOTENSINOGEN GENETIC VARIATION AND HYPERTENSION
based on Raymond and Rousset.
18
Genotype distributions
at single sites were compared between hypertensives and
normotensives using the program R X C at the same
website, and based on the Metropolitan method.
18
Hardy-
Weinberg analyses were also performed using TFPGA.
Statistical signicance was determined using Fishers
exact tests.
Haplotype Analysis
Haplotype analyses to test for linkage disequilibrium and
determine haplotype frequencies were performed using
FASTEH
19
and Powermarker software. Both use an EM
algorithm to determine haplotype frequencies when phase
is unknown and measure linkage disequilibrium. Power-
marker software is available at http://www.powermarker.
net. The Powermarker haplotype trend analysis was also
performed. This is a regression analysis testing haplotype-
trait association.
20
This approach can be applied to both
quantitative traits and dichotomous traits. The test for
association uses an F test.
20
Multifactor Analysis
Multisite analysis was performed using multifactor dimen-
sionality reduction available at www.epistasis.org.
21
Mul-
tifactor dimensionality reduction collapses all of the
genetic data into two categories, high and low risk. It does
this by exhaustively searching all single site and multisite
combinations of the data and then categorizing each mul-
tisite genotype cell into either high risk or low risk on the
basis of the ratio of cases to controls in each cell. It selects
one genetic model, either single or multisite, that most
successfully predicts phenotype. Cross-validation and test-
ing accuracy maximization are used to choose the single
best model. Testing accuracy is the proportion of individ-
uals classied correctly. The model with maximum testing
accuracy and cross-validation consistency is chosen as the
best model. The testing accuracy is representative of the
generalizabilty of the model.
Statistical signicance was determined only for the
single best model. Statistical signicance is determined
empirically by permuting the hypertensive and normoten-
sive labels 1000 times. Permutation testing is used to
derive a null distribution under no association with phe-
notype, that is, to determine the probability of chance
observations. Statistical signicance was determined by
comparing testing accuracy of the observed data to the
distribution under the null distribution of no association.
The null hypothesis of no association was rejected if the
measured testing accuracy was within the upper tail of the
permuted distribution (a testing accuracy in the upper
Table 1. African American and white allele and genotype comparisons
Site Genotype P
AGT-6 AA AG GG Allele Genotype
Hypertensive
AA Males 52 12 2 .001 .001
CS Males 15 33 7
Normotensive
AA Males 18 8 0 .001 .001
CS Males 7 22 6
Hypertensive
AA Females 86 30 1 .001 .001
CS Females 21 32 17
Normotensive
AA Females 28 6 0 .001 .001
CS Females 13 40 7
Site Genotype P
AGT-20 AA AC CC Allele Genotype
Hypertensive
AA Males 53 11 2 .27 .04
CS Males 37 18 0
Normotensive
AA Males 17 7 2 .67 .05
CS Males 17 18 0
Hypertensive
AA Females 86 29 2 .21 .41
CS Females 57 13 0
Normotensive
AA Females 22 12 0 .52 .47
CS Females 44 16 0
AA African Americans; CS whites.
1280 AJHDecember 2006VOL. 19, NO. 12 ANGIOTENSINOGEN GENETIC VARIATION AND HYPERTENSION
5.0% was considered signicant). Permutation testing cor-
rects for chance ndings that may result from multifactor
dimensionality reductions exhaustive search of all combi-
nations of variables. High risk genotypes for hyperten-
sion were dened as those that have a ratio adjusted to
the population examined. This ratio was calculated as the
number of cases divided by the number of controls in the
study groups. The rationale for establishing such a thresh-
old is to establish a baseline measurement or null hypoth-
esis of the expected numbers in each multifactor cell. The
thresholds for the present analyses were 3.05 for African
Americans and 1.32 for whites. Any ratio larger than this
baseline is a deviation from the expected and can be
viewed as high risk. This means that when more hyper-
tensives than normotensives are observed than expected,
high risk status is assigned. The analysis of the African-
American population was also performed both with over-
sampling and undersampling of the population to generate
a balanced dataset with a hypertensive/normotensive ratio
of 1.0 (D. R. Velez, S. M. Williams, in prep.). The odds
ratio (OR) for the best multifactor model was calculated
by categorizing high risk genotypes as exposed and low
risk genotypes as nonexposed.
Results
There were signicant allelic and genotypic differences
between ethnic groups. The A-6G single nucleotide poly-
morphism (SNP) differed signicantly between African
Americans and whites with respect to both allele and
genotype frequencies regardless of gender or phenotypic
class (Table 1). At C-20A genotype frequency differences
existed between men from the two groups regardless of
phenotype (hypertensive men, P .04 and normotensive
men, P .05) (Table 1). These results demonstrate that the
two ethnic groups are not genetically homogeneous.
Therefore, all subsequent analyses were conducted sepa-
rately for African Americans and whites.
In whites there were signicant allele and genotype
frequency differences between normotensive men and
Table 2.
A. White single-site analysis of malefemale difference within phenotypes
Site Genotype
Freq. G
P
AGT-6 AA AG GG Allele Genotype HWE
1
Hypertensive
Males 15 33 7 0.43 .50* .19* .17
Females 21 32 17 0.47 .49
Total 36 65 24 0.45 .09 .15 .72
Normotensive
Males 7 22 6 0.49 .65* .78* .19
Females 13 40 7 0.45 .01
Total 20 62 13 0.46 .004
AGT-20 AA AC CC Freq. C Allele Genotype HWE
Hypertensive
Males 37 18 0 0.16 .12* .10* .32
Females 57 13 0 0.09 1.00
Total 94 31 0 0.12 .13 .09 .21
Normotensive
Males 17 18 0 0.26 .05* .03* .08
Females 44 16 0 0.13 .58
Total 61 34 0 0.18 .04
B. Comparisons of allele and genotype frequencies between hypertensive and normotensives
Site
P
Allele Genotype
AGT-6
Males .27 .43
Females .25 .04
AGT-20
Males .14 .12
Females .34 .30
* Indicates comparison between men and women within phenotype class at site; Indicates comparison between hypertensive and
normotensive populations at site; Indicates comparison between hypertensive and normotensive men; Indicates comparison between
hypertensive and normotensive women.
1
Hardy Weinberg equilibrium.
1281 AJHDecember 2006VOL. 19, NO. 12 ANGIOTENSINOGEN GENETIC VARIATION AND HYPERTENSION
women for allele frequency (P .05) and genotype fre-
quency (P .03) at C-20A (Table 2A). However, analysis
of African Americans revealed no signicant differences
in allele or genotype frequency between the male and
female subjects in either phenotypic group (Table 3A).
Single site analyses of allele and genotype frequency
differences between the normotensives and hyperten-
sives in the white population revealed only one signif-
icant difference. An association was seen at A-6G
between hypertensive and normotensive women for ge-
notype (P .04; Table 2B). No signicant allele or
genotype associations were seen in the African-Ameri-
can population (Table 3B).
The normotensive white women deviated from Hardy-
Weinberg equilibrium at the A-6G site with an excess of
heterozygotes (P .01) (Table 2A). Hypertensive white
women did not deviate from Hardy-Weinberg equilibrium at
the A-6G site. No deviations from Hardy-Weinberg equilib-
rium were seen in African Americans in either phenotypic
class. Signicant linkage disequilibrium existed in white men
and women between A-6G and C-20A regardless of pheno-
type. The pairwise linkage disequilibrium coefcients (D=)
are all strong (D= 1.0, P .001) for whites (Table 4).
African-American hypertensive women showed signicant
pairwise linkage disequilibrium (D= 1.0, P .01). Afri-
can-American men of both phenotypes and African-Ameri-
can normotensive women did not demonstrate signicant D=
values (Table 4). These results are consistent with expected
higher values of linkage disequilibrium in whites as com-
pared to African Americans.
There was a signicant association between hyperten-
sion status and haplotype frequency for white men (P
.05; Table 5). No other haplotype frequency differences
were observed between phenotype classes (Table 5). There
was a difference in the haplotype frequency distribution of
whites and African Americans, with the most common
haplotype for whites being A-A (frequency ranging from
0.23 to 0.44) and G-A (ranging from 0.43 to 0.55),
whereas in African Americans it was A-A (ranging from
0.65 to 0.77) (Table 5).
Table 3.
A. African American single-site analysis of malefemale difference within phenotypes
Site Genotype
Freq. G
P
AGT-6 AA AG GG Allele Genotype HWE
Hypertensive
Males 52 12 2 0.12 .75* .28* .23
Females 86 30 1 0.14 .69
Total 138 42 3 0.13 .75 1.00 1.00
Normotensive
Males 18 8 0 0.15 .39* .37* 1.00
Females 28 6 0 0.09 1.00
Total 46 14 0 0.12 1.00
AGT-20 AA AC CC Freq. C Allele Genotype HWE
Hypertensive
Males 53 11 2 0.11 .52* .39* .18
Females 86 29 2 0.14 1.00
Total 139 40 4 0.13 .15 .17 .52
Normotensive
Males 17 7 2 0.22 .65* .30* .29
Females 22 12 0 0.18 .56
Total 39 19 2 0.19 1.00
B. Comparisons of allele and genotype frequencies between hypertensive and normotensives
Site
P
Allele Genotype
AGT-6
Males .62 .32
Females .40 .52
AGT-20
Males .10 .25
Females .43 .37
* Indicates comparison between men and women of populations at site; Indicates comparison between hypertensive and normotensive
populations at site; Indicates comparison between hypertensive and normotensive men; Indicates comparison between hypertensive and
normotensive women.
1282 AJHDecember 2006VOL. 19, NO. 12 ANGIOTENSINOGEN GENETIC VARIATION AND HYPERTENSION
In addition to single site and haplotype analyses, mul-
tisite analysis was also performed to determine whether
there are interactions that associate with hypertension.
Gender was also included in the multifactor dimensional-
ity reduction analysis as a contributing variable in the
model because AGT C-20A is a putative estrogen recep-
tor-binding site. Our analysis found a signicant multisite
model for whites. The model included A-6G, C-20A, and
gender in the white population. The average testing accu-
racy was 59.1% for the white sample (P .039) (Fig. 1).
The best 1, 2, and 3 locus models identied by the mul-
tifactor dimensionality reduction method in the white data-
set are seen in Table 6. The African Americans did not
show a signicant model.
The model found by multifactor dimensionality reduc-
tion may provide a way to assess genetic risk of hyper-
tension in white populations. The OR of the multifactor
high risk genotypes to the low risk genotypes is 2.21 (95%
condence intervals 1.283.82, P .001) for the white
population. The results support a multilocus genetic model
with gender as a signicant factor. The model shows no
uniform effect for a single genotype class except 6AA,
which appears to be uniformly high risk.
Discussion
Racial differences were identied between whites and
African Americans at the A-6G and C-20A polymor-
phisms for both genotype and allele frequencies (Table 1).
Differences in the haplotype frequencies between the two
ethnic groups were also signicant (Table 5). These nd-
ings required that analyses be performed separately for the
two groups. In addition, we found signicant differences
in genotype or allele frequencies between men and women.
Our study identied one single site association between
hypertension and polymorphisms at the A-6G position in
white women (P .04 for genotypic associations uncor-
rected for multiple comparisons). The fact that it was not
Table 4. Haplotype pairwise disequilibriumstatistics
Population
Pairwise multiallelic
linkage disequilibrium
statistic (D=) P
AA Females
HT 1.00 .01
NT 0.26 .16
AA Males
HT 1.00 .13
NT 0.49 .44
CS Females
HT 1.00 .001
NT 1.00 .001
CS Males
HT 1.00 .001
NT 1.00 .001
AA African Americans; CS whites; HT hypertensive; NT
normotensive.
Table 5. Haplotype trend regression analysis
Haplotype (-6 -20) Normotensive Hypertensive Combined P*
White women
G-A 0.55 0.47 0.51 .3
A-C 0.12 0.09 0.11
A-A 0.33 0.44 0.38
White men
G-A 0.51 0.43 0.46 .05
A-C 0.26 0.16 0.2
A-A 0.23 0.41 0.34
African-American women
G-A 0.05 0.14 0.12 .37
A-C 0.14 0.14 0.14
A-A 0.77 0.72 0.73
African American men
G-A 0.14 0.12 0.13 .14
A-C 0.19 0.11 0.14
A-A 0.65 0.77 0.73
* Comparing hypertensives and normotensives.
FIG. 1. Summary of two-site genotype combinations and sex dif-
ferences associated with high risk and low risk for white hyperten-
sive individuals. Each multifactorial cell is high risk (dark gray) or
low risk (light gray). For each multifactorial combination, hypo-
thetical distributions of cases (left bar in cell) and controls (right
bar in cell) are shown. Each cell represents a multisite genotype,
the genotype is labeled on the figure. F women; M men. There
is evidence of epistasis or genegene interaction.
1283 AJHDecember 2006VOL. 19, NO. 12 ANGIOTENSINOGEN GENETIC VARIATION AND HYPERTENSION
found in men is interesting, as it suggests that different
genetic factors explain the differences in the prevalence of
hypertension in men and women. This hypothesis is fur-
ther supported by the differences found in the multisite
analysis, as a difference was found between men and
women at the 6AG and 20AA interacting genotypes.
Although we failed to nd any signicant differences
among African-American hypertensives and normoten-
sives, this may be due to the small number of African-
American normotensives in our sample.
Our results are in agreement with some previous studies
of AGT variants and hypertension, but not others.
2226
One coding AGT variant (M235T) has previously been
associated with both hypertension and altered sodium han-
dling in an Italian cohort,
24
and data suggest that M235T
is in strong linkage disequilibrium with A-6G in at least
some populations.
23
However, salt sensitivity is reportedly
more prevalent in African Americans than in whites and
we failed to identify an association in our African-Amer-
ican sample despite previous ndings suggesting an asso-
ciation with A-6G in African Americans.
22
The failure to
replicate ndings is not uncommon in the genetic analysis
of hypertension, and it has been hypothesized that this is
due to failure to account for other important variables.
27
This hypothesis may also explain why the C-20A site and
gender associated with hypertension in whites but not in
African Americans in our study, despite there being strong
evidence that it is an estrogen receptor element.
Our analyses also failed to identify strong haplotype
associations with BP phenotype (P .05 in white men
uncorrected for multiple comparisons). However, in an
analysis of multilocus genotypes with multifactor dimen-
sionality reduction we found a signicant model that
included both variants and gender. This analysis supports
the role that multiple AGT variants play a role in the
regulation of BP. Importantly, it also suggests the advan-
tage of using genotype analyses over allele or haplotype
analyses. Although allele or haplotype analysis may be
statistically more powerful in identifying associations
because of the greater number of alleles as compared to
the number genotypes,
28
the genotype is likely the most
important level of biological signicance because individ-
uals are made up of genotypes that operate together.
The multisite analysis supports the hypothesis that sin-
gle site association may not provide the best explanation
of an observed phenotype. The multifactor dimensionality
reduction procedure searches through all possible dimen-
sions, from the best single site model to the best two-site
model. Multifactor dimensionality reduction found that the
single-site genotype model was not as strong as the two-
site model that incorporated gender. Instead, complex ep-
istatic models may provide a more biologically relevant
explanation, including interactions of variants within a
gene.
2
The multifactor dimensionality reduction analysis
was also used to generate OR to determine the level of
increased risk for the high risk genotypes. There is a
2.21-fold increase in risk for hypertension in whites with
any of the high risk genotypegender combinations rela-
tive to any of the other genotypegender combinations.
This OR is statistically signicant and provides a more
typical epidemiologic interpretation of the genetic model,
suggesting that the multisite model does signicantly in-
crease disease risk. The level of risk increase is relatively
high for hypertension compared to most analyses using a
single site, where the OR are usually below 2 and almost
always below 3.
29,30
Our ndings suggest that a genetic
model for hypertension may require multiple site analyses.
The fact that gender and C-20A were incorporated into
the multisite genetic model supports the hypothesized role
of C-20A being in an estrogen receptor element. Studies
have shown that the A variant at the C-20A promoter
region binds to the estrogen receptor, and reporter con-
structs with this variant can transactivate AGT expression
upon co-transfection of an expression vector containing
estrogen receptor coding sequences.
17
In addition, Arp-1,
a homolog of COUP-TFII a steroid-thyroid hormone re-
ceptor, binds to a sequence at the transcription initiation
site of the human angiotensinogen gene and downregu-
lates the estrogen-induced promoter activity when the A
allele at C-20A is present.
17
High levels of estrogen coun-
teract any suppressive effects of Arp-1 on the promoter
activity of the AGT gene.
17
These data may help to ex-
plain the way gender and sex hormones affect the RAS and
ultimately BP regulation.
References
1. Williams SM, Addy JH, Phillips JA III, Dai M, Kpodonu J, Afful J,
Jackson H, Joseph K, Eason F, Murray MM, Epperson P, Aduonum
A, Wong LJ, Jose PA, Felder RA: Combinations of variations in
multiple genes are associated with hypertension. Hypertension
2000;36:26.
2. Williams SM, Ritchie MD, Phillips JA, Dawson E, Prince M,
Dzhura E, Willis A, Semenya A, Summar M, White BC, Addy JH,
Kpodonu J, Wong L-J, Felder RA, Jose PA, Moore JH: Multilocus
analysis of hypertension: a hierarchical approach. Hum Heredity
2004;57:2838.
Table 6. Summary of white MDR results
No. of factors Best model for each interaction Prediction accuracy
Cross validation
consistency
1 [AGT-20] 0.546 8/10
2 [AGT-6 - AGT-20] 0.559 7/10
3 [AGT-6 - AGT-20 - Sex] 0.591 10/10
1284 AJHDecember 2006VOL. 19, NO. 12 ANGIOTENSINOGEN GENETIC VARIATION AND HYPERTENSION
3. Cooper RS, Wolf-Maier K, Luke A, Adeyemo A, Banegas JR,
Forrester T, Giampaoli S, Joffres M, Kastarinen M, Primatesta P,
Stegmayr B, Thamm M: An international comparative study of
blood pressure in populations of European vs. African descent.
BMC Med 2005;3:2.
4. Gu C, Borecki I, Gagnon J, Bouchard C, Leon AS, Skinner JS,
Wilmore JH, Rao DC: Familial resemblance for resting blood pres-
sure with particular reference to racial differences: preliminary anal-
yses from the HERITAGE Family Study. Hum Biol 1998;70:7790.
5. Rotimi CN, Cooper RS, Cao G, Ogunbiyi O, Ladipo M, Owoaje E,
Ward R: Maximum-likelihood generalized heritability estimate for
blood pressure in Nigerian families. Hypertension 1999;33:874
878.
6. Corvol P, Persu A, Gimenez-Roqueplo AP, Jeunemaitre X: Seven
lessons from two candidate genes in human essential hypertension:
angiotensinogen and epithelial sodium channel. Hypertension 1999;
33:13241331.
7. Jeunemaitre X, Soubrier F, Kotelevtsev YV, Lifton RP, Williams
CS, Charru A, Hunt SC, Hopkins PN, Williams RR, Lalouel JM,
Corvol P: Molecular basis of human hypertension: role of angio-
tensinogen. Cell 1992;71:169180.
8. Tsai CT, Fallin D, Chiang FT, Hwang JJ, Lai LP, Hsu KL, Tseng
CD, Liau CS, Tseng YZ: Angiotensinogen gene haplotype and
hypertension: interaction with ACE gene I allele. Hypertension
2003;41:915.
9. Muller DN, Fiebeler A, Park JK, Dechend R, Luft FC: Angiotensin
II and endothelin induce inammation and thereby promote hyper-
tension-induced end-organ damage. Clin Nephrol 2003;60(Suppl 1):
S2S12.
10. Bloem LJ, Manatunga AK, Tewksbury DA, Pratt JH: The serum
angiotensinogen concentration and variants of the angiotensinogen
gene in white and black children. J Clin Invest 1995;95:948953.
11. Ishigami T, Umemura S, Tamura K, Hibi K, Nyui N, Kihara M,
Yabana M, Watanabe Y, Sumida Y, Nagahara T, Ochiai H, Ishii M:
Essential hypertension and 5= upstream core promoter region of
human angiotensinogen gene. Hypertension 1997;30:13251330.
12. Jain S, Tang X, Narayanan CS, Agarwal Y, Peterson SM, Brown
CD, Ott J, Kumar A: Angiotensinogen gene polymorphism at 217
affects basal promoter activity and is associated with hypertension in
African-Americans. J Biol Chem 2002;277:3688936896.
13. Jain S, Li Y, Patil S, Kumar A: A single-nucleotide polymorphism
in human angiotensinogen gene is associated with essential hyper-
tension and affects glucocorticoid induced promoter activity. J Mol
Med 2005;83:121131.
14. Inoue I, Nakajima T, Williams CS, Quackenbush J, Puryear R,
Powers M, Cheng T, Ludwig EH, Sharma AM, Hata A, Jeunemaitre
X, Lalouel JM: A nucleotide substitution in the promoter of
human angiotensinogen is associated with essential hypertension
and affects basal transcription in vitro. J Clin Invest 1997;99:
17861797.
15. Zhao YY, Zhou J, Narayanan CS, Cui Y, Kumar A: Role of C/A
polymorphism at 20 on the expression of human angiotensinogen
gene. Hypertension 1999;33:108115.
16. Morgan L, Crawshaw S, Baker PN, Broughton PF, Kalsheker N:
Polymorphism in oestrogen response element associated with vari-
ation in plasma angiotensinogen concentrations in healthy pregnant
women. J Hypertens 2000;18:553557.
17. Narayanan CS, Cui Y, Zhao YY, Zhou J, Kumar A: Orphan receptor
Arp-1 binds to the nucleotide sequence located between TATA box
and transcriptional initiation site of the human angiotensinogen gene
and reduces estrogen induced promoter activity. Mol Cell Endocri-
nol 1999;148:7986.
18. Raymond M, Rousset F: An exact test for population differentiation.
Evolution 1995;49:12801283.
19. Zhao JH, Sham PC: Faster haplotype frequency estimation using
unrelated subjects. Hum Heredity 2002;53:3641.
20. Zaykin DV, Westfall PH, Young SS, Karnoub MA, Wagner MJ,
Ehm MG: Testing association of statistically inferred haplotypes
with discrete and continuous traits in samples of unrelated individ-
uals. Hum Heredity 2002;53:7991.
21. Ritchie MD, Hahn LW, Roodi N, Bailey LR, Dupont WD, Parl FF,
Moore JH: Multifactor-dimensionality reduction reveals high-order
interactions among estrogen-metabolism genes in sporadic breast
cancer. Am J Hum Genet 2001;69:138147.
22. Rasmussen-Torvik LJ, North KE, Gu CC, Lewis CE, Wilk JB,
Chakravarti A, Chang YP, Miller MB, Li N, Devereux RB, Arnett
DK: A population association study of angiotensinogen polymor-
phisms and haplotypes with left ventricular phenotypes. Hyperten-
sion 2005;46:12941299.
23. Jeunemaitre X, Inoue I, Williams C, Charru A, Tichet J, Powers M,
Sharma AM, Gimenez-Roqueplo AP, Hata A, Corvol P, Lalouel
JM: Haplotypes of angiotensinogen in essential hypertension. Am J
Hum Genet 1997;60:14481460.
24. Siani A, Russo P, Paolo CF, Iacone R, Venezia A, Russo O, Barba
G, Iacoviello L, Strazzullo P: Combination of renin-angiotensin
system polymorphisms is associated with altered renal sodium han-
dling and hypertension. Hypertension 2004;43:598602.
25. Wu X, Luke A, Rieder M, Lee K, Toth EJ, Nickerson D, Zhu X,
Kan D, Cooper RS: An association study of angiotensinogen poly-
morphisms with serum level and hypertension in an African-Amer-
ican population. J Hypertens 2003;21:18471852.
26. Zhu X, Chang YP, Yan D, Weder A, Cooper R, Luke A, Kan D,
Chakravarti A: Associations between hypertension and genes in the
renin-angiotensin system. Hypertension 2003;41:10271934.
27. Williams SM, Haines JL, Moore JH: The use of animal models in
the study of complex disease: all else is never equal or why do so
many human studies fail to replicate animal ndings? Bioessays
2004;26:170179.
28. Sasieni PD: From genotypes to genes: doubling the sample size.
Biometrics 1997;53:12531261.
29. OShaughnessy KM: The genetics of essential hypertension. Br J
Clin Pharmacol 2001;51:511.
30. Sethi AA, Nordestgaard BG, Tybjaerg-Hansen A: Angiotensinogen
gene polymorphism, plasma angiotensinogen, and risk of hyperten-
sion and ischemic heart disease: a meta-analysis. Arterioscler
Thromb Vasc Biol 2003;23:12691275.
1285 AJHDecember 2006VOL. 19, NO. 12 ANGIOTENSINOGEN GENETIC VARIATION AND HYPERTENSION
Reproducedwith permission of thecopyright owner. Further reproductionprohibited without permission.