Genetics of essential hypertension.

From the unimodal-bimodal controversy to

molecular technology
A Camussi and G Bianchi
Hypertension 1988, 12:620-628
doi: 10.1161/01.HYP.12.6.620
Hypertension is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX
72514
Copyright 1988 American Heart Association. All rights reserved. Print ISSN: 0194-911X. Online
ISSN: 1524-4563

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://hyper.ahajournals.org/content/12/6/620.citation

Subscriptions: Information about subscribing to Hypertension is online at

http://hyper.ahajournals.org//subscriptions/
Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters
Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax:
410-528-8550. E-mail:
journalpermissions@lww.com
Reprints: Information about reprints can be found online at
http://www.lww.com/reprints

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

Point of View

Genetics of Essential Hypertension

From the Unimodal-Bimodal Controversy to
Molecular Technology
ALESSANDRO CAMUSSI AND GIUSEPPE BIANCHI

HE aim of this brief review is to provide

researchers involved in investigating the
pathogenesis of essential hypertension with
a historical perspective of genetic studies on the
subject and to list some of the problems that should
be taken into account when considering the hypothesis that a given biochemical mechanism may be
involved in causing essential or genetic types of
hypertension. Because of obvious space limitations
and the wide variety of problems involved, it will
not be possible to provide a detailed, comprehensive review of all the published work on this topic.
However, we will try to focus on some common
beliefs that have been dominant in the literature and
continue to influence the thinking of many students
of hypertension, at least as indicated by discussions
at scientific meetings and articles in the specialist
journals. We will also briefly discuss the criteria for
allocating a given biochemical finding to its probable position in the chain of events leading from a
gene abnormality to increased blood pressure. Such
criteria appear to be missing in most of the published work on genetic rat models aimed at elucidating differences in biochemical systems of blood
pressure regulation between hypertensive strains
and their various normotensive controls.
Many genes coding for proteins involved in blood
pressure regulation (renin, atria! natriuretic factor,
mineralocorticoids, and adrenergic receptors)'-4 have
already been identified. In addition, molecular mechanisms that may affect the blood pressure level in
humans and experimental animals are under investigation. For these reasons, any discussion of the
genetic or primary mechanisms causing essential
hypertension must include attempts to find links
between these aspects at the different biological
levels of organizationfrom the ultimate phenotyp-

ical level (blood pressure) to the protein level (gene

products), and eventually, to the DNA structure.
The extension of this strategy to humans might be
possible if the complex interaction of genetic and
environmental influences were understood. Even if
the sophistication and the novelty of the techniques
discourage many students of hypertension, this
effort is justified by the demonstration that even
diseases with similar phenotypes and identical modes
of inheritance have turned out to be genetically
heterogeneous at both the DNA and primary gene
product levels.3-7 For all these reasons, our discussion of the genetic control of blood pressure will
examine the results obtained thus far with the
classic biometrical approach, which includes analysis of the frequency distribution of blood pressure
in the population and more sophisticated quantitative methods (studies of families, likelihood analysis of pedigree data), to determine whether the large
amount of data so furnished justifies a deeper investigation of the genetic control of essential hypertension and which genetic model best fits the experimental information. We will also address the
following question: If genetic factors play an important role in determining the variability of blood
pressure in the population, can a link with a given
genetically controlled protein polymorphism be identified, and to what extent are such polymorphisms
related to the variability of blood pressure?
The Biometrical Approach
When the blood pressure distribution of the human
population is considered, those persons with blood
pressure above a certain arbitrary limit are said to
have essential hypertension if other causes of hypertension can be ruled out. Systolic and diastolic
blood pressure levels are characteristics that can be
defined as quantitative; they are measured on a
continuous scale. It is clearly impossible to divide
the population values into discrete categories on the
basis of individual measurements. Nevertheless,
from the clinical standpoint a sharp distinction
exists: A portion of the population (a minority) is
defined as abnormal (hypertensive) and the remain-

From the Department of Genetics and Microbiology (A. Camussi)

and the Istituto di Scienze Mediche (G. Bianchi), University of
Milano, Milano, and the Istituto di Patologia Speciale Medica (G.
Bianchi), University of Sassari, Sassari, Italy.
Address for reprints: Prof. Giuseppe Bianchi, Istituto di Scienze
Mediche, Universita di Milano, Via F. Sforza, 35, 20122 Milano,
Italy.

620
Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

GENETICS OF ESSENTIAL HYPERTENSION/Camim/ and Bianchi

der as normal (normotensive) on the basis of blood
pressure level. This distinction does not depend on
any qualitative intrinsic characteristics of the trait,
but on the fact that persons with higher pressure
levels have a greater probability of having certain
serious health problems (e.g., coronary heart disease, stroke, and renal failure). Like height, weight,
or other morphological traits, an individual's blood
pressure is the end product of a large number of
morphological, physiological, and biochemical characteristics. These may, in turn, interact with factors
that determine the individual's environment, such
as dietary salt, stress, and exercise. Hence, blood
pressure can be described as a multifactorial trait
and, analogously, essential hypertension as a multifactorial disease or, more precisely, a multifactorial
risk factor. However, the major problem regarding
the pathogenesis of essential hypertension is to determine how much of the observed variability in the
blood pressure level is due to the presence of segregating genetic factors in the population, how many of
these factors exist, what the relative importance of
environmental or cultural influences is, and to what
extent individual genetic backgrounds interact.
The Unimodal-Bimodal Controversy

The shape of the overall distribution of blood

pressure levels was regarded with particular interest in the epidemiological surveys that dominated
the literature of essential hypertension during the
1950s and 1960s. Platt8-9 claimed the existence of a
bimodality in the distributions of blood pressure.
He attributed this to the segregation of a single
gene, even if there were large variations between
the measurements, but he thought that environmental conditions played little or no part in the genesis
of hypertension. This interpretation was in strong
opposition to the views of Pickering's group, I0-12
which supported the existence of unimodal distributions in the detailed, large surveys of the London
population. A unimodal distribution in a randommating population would suggest the existence of
many genetic factors segregating independently. An
extensive interpretation was that hypertension is
not a true disease, but that the risk arising from high
blood pressure is simply proportional to its entity,
environmental factors being responsible for the rate
and pattern of the rise with age.13
The debate between bimodalists and unimodalists
has now lost much of its importance, since the
shape of a distribution is not sufficient to prove the
existence of few or many segregating genes, at least
if the environmental conditions cannot be controlled, as in animal and plant experiments. In fact,
a bimodal distribution can be simulated in a polygenically controlled trait by a secondary effect
when the measured variable shows a tendency
toward self-enhancement; for instance, above certain levels of blood pressure, kidneys can be damaged, leading to further increases in blood pressure.M
On the other hand, bimodality cannot be observed

621

among closely overlapping distributions.13 Examples can be furnished by analyses of variability in

enzyme levels caused by various polymorphisms,
such as glutamate-pyruvate transaminase activities16
and red blood cell acid phosphatase activities.17 In
such cases, the combined distribution of enzymatic
levels in the population at large can be regarded as
unimodal and interpreted as multifactorial in origin,
even when the total population distribution is determined by the segregation of only two alleles. Finally,
a more careful reexamination of epidemiological
data from a large blood pressure survey in Bergen,
Norwaywhen the data were stratified according
to age and sexhas revealed, in an apparent unimodal distribution, the existence of two overlapping distributions with statistically different mean
values at different ages.18
Analysis of Aggregation in Families

A genetic hypothesis cannot be based exclusively

on the population distribution of the variable. Family data are also needed; in other words, the aggregation of the trait must be examined in relatives.
The aggregation of a phenotypic trait among persons who share genes in common (parents and
offspring, siblings) is a strong indication of the
existence of genetic control of the trait if the environment can be considered neutral. The study of
blood pressure levels in families has, however, no
simple genetic interpretation, because family members share a common environment as well as genes.
Many epidemiological data have been studied to
clarify the relation between probands and, generally, their first-degree relatives by means of different coefficients of resemblance (correlation and
regression coefficients). The most important studies
dealing with the quantitative genetic analysis of
blood pressure levels from large samples of the
population yielded different estimates of the degree
of association between parents and offspring and
between siblings (Table 1). The comparison of these
estimates is of interest. The correlation between
siblings was higher than that between parents and
offspring in the study of Miall et al.,19 lower in the
Framingham study,20 but similar in the Evans
County21 and Tecumseh studies.22 Furthermore, no
correlation between siblings was found in the black
population of Detroit,23 although the sample size
was small. Moreover, a reanalysis of the Detroit
data using more appropriate and sophisticated methods showed that a significant fraction of blood
pressure variation is caused by genetic factors.24
The complementary information furnished by the
analysis of the aggregation of blood pressure levels
of adopted children as compared with the blood
pressure of adoptive parents is included in the
Montreal Adoption Study.23
An analysis in twins can corroborate the interpretation of the genetic control of a quantitative trait.
Monozygotic twins represent the only model in
which genetically identical humans can be com-

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

622

HYPERTENSION

TABLE 1. Coefficients of Aggregation for Systolic Blood Pressure Among First-Degree Relatives
Correlation
Study and location

Child-parents

Mialletal.," South Wales, UK

0.24*
Havlik et al.,20 Framingham, MA,
0.34t
USA
Hayes et al.,21 Evans County,
0.13-0.26*
GA, USA
Longini et al.,22 Tecumseh, MI,
0.12-0.18*
USA
Shull et ah.,23 Detroit, MI, USA

Biron et al.,23 Montreal, Canada

Adopted children
0.09*
Natural children
0.32*
Correlation coefficient,
tlntraclass correlation coefficient.

Sib-sib
0.33*
0.18t
0.14-0.20*
0.17t
-0.02*
0.07*
0.28*

pared and where environmental effects may be

detected; in fact, any differences between them are
due to environmental influences. Comparison with
the trait in dizygotic twins, in whom both genetic
and environmental factors act, can quantify the
extent of genetic control, which is estimated to
represent as much as 82% of the variation in systolic blood pressure. 26 -" Some caution should be
observed because cultural influences are generally
more similar in twins than in siblings. An analysis of
the genetic control at the biometrical or phenotypical level based on the degree of association of the
trait in relatives has recently been reviewed.28-w All
the data revealed a stronger correlation between
genetically related persons than between spouses or
between adoptees and have so far strongly stressed
two major points. First, a familial aggregation of
blood pressure levels, and consequently of hypertension, exists even when large differences in different populations have been taken into account.
Second, the observed familiarity is due, in different
proportions, to shared genes and shared environment. However, a large amount of indetermination is
expected if the analysis is restricted to this approach.
For instance, a parent-sib correlation may be lower
than a sib-sib correlation for two reasons: a more
uniform environmental influence on siblings or the
presence of notable dominance effects of genes controlling the trait. Although a distinction between
these two causes is impossible at the population
level, they can greatly influence the interpretation of
the results. As discussed in detail by Cavalli-Sforza
and Bodmer,30 if a specific environmental effect on
siblings can be excluded, the estimate of the fraction
of blood pressure variance due to genetic factors will
increase. An interpretation could be that relatively
few genes are involved in the determination of blood
pressure, since a consistent dominance effect of
many genes seems unlikely.30 In the Montreal Adoption Study,31 however, the environmental effects
were carefully investigated and a differentiation was
made between a shared environment common to

VOL 12, No 6, DECEMBER 1988

parents and children (the so-called across-generation

environmental effects) and an environment shared
by children only (within-generation effects). For
instance, with regard to diastolic blood pressure,
30% of the phenotypical variability in siblings was
due to genes, 11% to across-generation environmental effects, and 20% to environmental effects shared
by the children alone.31
Blood pressure correlates with potential confounding traits such as age, sex, and anthropometric
characteristics, and this has also been observed for
many biochemical and physiological factors involved
in its control. The existence of such correlations can
lead to two different biases in the use of the aggregation of the blood pressure among genetically related
persons. First, groups of siblings and groups of
spouses will be more uniform for age than other
randomly selected persons. If the confounding variables are not considered in the analysis (a good
approach is the use of residuals after fitting a multiple
regression to the data), biased or distorted correlations can be the result of their joint variation instead
of the variation of the main trait. Second, the estimate of heritability coefficients from the association
coefficient is based on the assumption of random
mating in the study population. If blood pressure can
be influenced by concomitant variables, particularly
by anthropometric or behavioral characteristics that
can favor assortative mating, the basic assumption
for the unbiased estimation may not be upheld.
Analysis of Pedigree Data

For all of the reasons mentioned, a promising

strategy is to determine the relative contribution of
genes and of environmental factors directly from
pedigree data or, preferably, from data measured in
individual members of a family whose links in the
pedigree are known. The determination is made
possible by fitting a specific genetic model to family
data and by estimating its relative likelihood. Quantitative methods such as likelihood pedigree analysis and path analysis are commonly used in population genetics,32-33 but they have been employed
only recently in the study of the genetics of hypertension. In this procedure, each genetic model
(sporadic, monogenic, polygenic, or mixed with
different subdivisions) can be represented by a set
of genetic and environmental (familial or individual)
factors. A likelihood function is used to combine
the data from pedigrees with the parameters of the
model. The values of parameters that maximize the
function are chosen as representative of the model
and, consequently, of the distribution characteristics in the overall population. Different models are
compared with the sporadic one that does not
include genetic parameters, and the model with the
highest likelihood is chosen as that best fitting the
observed data. The most extensive applications of
the likelihood pedigree analysis can be found in the
Utah Survey34 and in the Montreal Adoption Study.31
The maximal likelihood analysis of pedigrees has

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

GENETICS OF ESSENTIAL HYPERTENSION/Camass/ and Bianchi

confirmed that a polygenic model is adequate to
explain the observed variability of blood pressure.
The Molecular Approach
Since it has been established that a substantial
portion of blood pressure variability is due to genetic
factors and that a polygenic model best explains this
variability, the question arises as to whether (and
eventually how) it would be possible to identify,
among the genes involved, the gene with the higher
relative effect. According to genetic theory, a true
difference does not exist between the so-called
polygenes and a major gene at the primary product
level and, consequently, between the genetic control of a quantitative and a qualitative trait. The
distinction is based only on the difficulty of dissecting the complex phenotypical expression into the
effects of individual genes that control the quantitative trait. One of the well-known examples of
phenotypical changes in the expression of a genotypical difference according to the biological level
of observation is provided by the phenylketonuria
syndrome.35 This syndrome is related to the high
level of phenylalanine in the plasma of affected
persons as a result of a mutation of the gene coding
for the enzyme responsible for the conversion of
phenylalanine into tyrosine. The distinction between
normal and affected persons is difficult to make if it
is based on a complex phenotypical trait such as
head size (Figure 1, bottom panel), because the
effects of other genes and of environment are added
(polygenic model). The contribution of the phenylketonuria gene to head size (the ultimate level in
this example) is obviously smaller as compared with
the total effect of other genes. The intelligence
quotient (IQ) distributions (see Figure 1, middle
panel) overlap slightly. In fact, the IQ is still highly

n
10
15 20
30

Percentages of phenylalanine in blood plasma, in milligrams

1 1 II I 1 ! 1

150

100

i i

>r

ni
50

20

Intelligence (Bind 10 scale)

'MI
330
310 290 270
Head sizelength plus breadth, in millimeters (corrected for sex)

FIGURE 1. Frequency distributions of different traits in

healthy persons and persons with phenylketonuria (shaded
area). See text for explanation. Adapted from Penrose.35

623

influenced by different genetic and environmental

factors. However, the effect of the phenylketonuria
gene is already clear. If we consider a level of
analysis closely related to the primary activity of
the gene (see Figure 1, top panel), the distinction
between normal and affected persons is clear, and
even if a certain variability is present, the effect of
the major gene is detectable in a qualitative way. If
we consider the phenylketonuria gene as only one
of the polygenes controlling head size, it acts at the
primary product level as a "normal" gene with a
clear qualitative effect.
Analysis of Biochemical and Physiological
Components of Blood Pressure or
Intermediate Phenotypes

Even if the formal model of quantitative inheritance implies the existence of many genes that all
have similar and additive effects, a multifactorial
model can comprise some genes with a larger effect
on the level of the trait under study. The detection
of such a gene or genes is difficult indeed if the
analysis is restricted to the ultimate phenotypein
this context, blood pressure. The identification of
major physiological and biochemical components,
the variability of which must be analyzed in persons
drawn from a population or pedigree, can be of help
in understanding the genetic control of the ultimate
phenotypical level and the nature and degree of
interactions with the environmental effects. A basic
prerequisite should be that the variability of a
component trait or of the so-called intermediate
phenotypes is significantly associated with the variability of blood pressure in the population. The
numerous physiological and biochemical studies of
the pathogenesis of hypertension have furnished an
excellent basis for identifying important intermediate phenotypes. These characteristics, which are
generally investigated as biochemical or physiological indicators of hypertension, concern different
aspects of human physiology.36 To be of interest in
the genetic dissection of a complex trait, the mode
of inheritance of the intermediate phenotype must
be investigated. The detection of a simple genetic
control is more probable if the intermediate phenotype is close to the primary gene product. If we
know the genetic control of an intermediate phenotype, its relevance to the variability of blood pressure levels, and its relative frequency in different
pedigrees, it should be possible to predict the inheritance of blood pressure more accurately.
The dissection of a complex trait is shown in
Figure 2. In this scheme, the path connecting the
genes (at the bottom) to the ultimate levelblood
pressure (at the top)is represented by a number
of different intermediate levels that "dissect" the
complex trait; at each level one or more intermediate phenotypes can be determined. At the bottom,
the genetic information can be subdivided into
polygenes, the effects of which are not individually
measurable but that act at each level of dissection,

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

HYPERTENSION

624
PHEHOWtCAL LEVEL

TRAITS

whole body

VOL

12, No 6, DECEMBER 1988

2. Dissection of a complex trait at

different levels of biological organization. See
text for explanation. IPs = intermediate
phenotypes.
FIGURE

ENVIRONMENT

organs

ENVIRONMENT

cellular

ENVIRONMENT

subcellular

ENVIRONMENT

molecular

ENVIRONMENT
SEHETK lACtGMtNO

CAMLVDATE 6EKES

and candidate genes, the effects of which must be

determined. The paths connecting the intermediate
components (shown as dotted lines in Figure 2) and
the genetic interactions with regulatory effects
(shown as continuous lines) may be extremely complex. Environmental effects (cultural and stochastic) act independently at each level.
The analysis of the intermediate phenotypes can
produce useful information: an intermediate trait
that shows Mendelian inheritance and that can be
linked to the normal biochemical and physiological
regulation of blood pressure can act as a marker for
blood pressure. Of course, since blood pressure is
under polygenic control more than one marker trait
probably will be identified in the future. The information about simple intermediate phenotypes and
their link with blood pressure can be used to identify the candidate genes, their number, the frequency of alleles, and the mode of action (regulatory or structural). If the genes (measured genotype)
are known, then the final product (phenotypical
complex trait) can be predicted in individuals. The
so-called measured genotype approach of Sing and
colleagues37-38 follows these lines. Moreover, it
should be possible to distinguish between a true
multifactorial model (which emphasizes the role of
the interaction between multiple environmental characteristics and multiple genetic factors) and the
so-called multiple unilocus model (which relates the
high blood pressure to a mutation at any one of
many possible controlling loci.29
An Example of Analysis of Intermediate
Phenotypes in Rats
The model shown in Figure 2 allows us to incorporate the individual findings at the different biological levels in a logical sequence of events, going
from a possible gene difference to a blood pressure
abnormality. However, this approach to the genetic
mechanisms of human hypertension requires an
experimental and theoretical foundation in animal
models. For example, only after having demon-

strated that renal artery constriction causes hypertension in animals39 was it possible to detect the
renal causes of hypertension in humans. The same
may prove to be true for identifying the genetic
components of hypertension, although an enormously greater degree of complexity and sophistication will be necessary. The results accumulated
thus far by studying the problem in the Milan
hypertensive strain of rats (MHS) in comparison
with its normotensive control strain (MNS) may
help to illustrate this type of approach. These
studies are proceeding along the following lines:
1. Identification of one or more biochemical or physiological traits (intermediate phenotypes) that are relevant to the blood pressure differences between MHS
and MNS.
2. Identification of the protein polymorphisms underlying
a given biochemical or physiological trait.
3. Identification of possible DNA polymorphisms underlying the protein polymorphisms.
To assess the first point, an approach similar to
that suggested by Rapp28- * for evaluating the role
of 18-hydroxydeoxycorticosterone in Dahl rats or
of the vascular smooth muscle response to some
cations in spontaneously hypertensive rats was followed. To propose that a given trait may be relevant
in explaining a blood pressure difference between
two strains of rats, the following criteria must be
met, at least in part:
1. A difference in the trait between the two strains must
be demonstrated.
2. The trait must cosegregate with an increment of blood
pressure significantly different from zero in an F2
population obtained from the cross of F, hybrids
(MHS x MNS).
3. Some logical biochemical or physiological link must
exist between the trait and blood pressure.
These criteria were met for Na + -K + cotransport
and cell volume in erythrocytes from MHS, mainly
because several functional similarities exist between
the erythrocytes and renal tubular cells, which are

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

GENETICS OF ESSENTIAL HYPERTENSION/Camwss/ and Bianchi

more directly involved in causing blood pressure
differences between MHS and MNS. Therefore, the
dissection of these two traits proceeded in order to
evaluate whether a protein polymorphism was
responsible for the differences between the two
strains. The results of these studies showed that the
cause of the difference in cell volume and Na+-K+
cotransport probably was localized in the network
of membrane skeleton proteins. Cross-immunization
experiments between MHS and MNS with the use
of these proteins demonstrated an immunochemical
difference between MHS and MNS of a 105kilodalton membrane skeleton protein. An erythrocyte complementary DNA (cDNA) library was
screened with the antibody against this protein, and
a cDNA probe coding for a 31-kilodalton protein
was isolated. When injected into rabbit, this protein
stimulated the production of an antibody that bound
to the natural 105-kilodalton protein. Further studies are in progress to isolate the entire gene coding
for the 105-kilodalton protein and to evaluate its
possible polymorphism in the two rat strains.
Table 2 illustrates the most significant findings
obtained in these two rat strains at the different
biological levels, together with other gene products

625

that may be considered candidates for causing other

biochemical or physiological differences between
MHS and MNS.41-57 When the links in sequencegene to protein to biochemical functions to physiological function to organ function to blood pressure
can be proved to be true and the relative weights to
each passage can be assessed in a given rat model of
genetic hypertension, then we should have the
theoretical and experimental background to
approach this problem in human pedigrees, provided that the environmental factors can be evaluated as well in humans as in rats. The key question
is how to prove that a given biochemical or physiological sequence is linked to a given gene and is
really the cause, or an important part of the cause,
of the blood pressure differences between the two
strains of rats. The most straightforward approach
to this question is certainly the use of transgenic
animals.58 These are animals that have foreign DNA
integrated into their germ line by experimental
procedures. The most common methods of introducing foreign DNA are direct injection into the
embryo or transfection by retro viral vectors. After
introduction, the foreign DNA integrates into the
host DNA and responds to the host environment. If

TABLE 2. Major Differences Between MHS and MNS According to the Level of Biological Organization
Phenotypical (or
biological) level
Whole body
Organ
Cellular
Subcellular

Biochemical

Molecular

DNA

Intermediate phenotypes*
Kidney cross-transplantation
H2O and Na metabolism
Kidney function in whole animal
Isolated kidney
Structure and function of
Erythrocytest
Renal tubular cells
Inside-out erythrocyte vesicles (without
membrane skeleton)
Reseated erythrocyte ghosts (with membrane
skeleton)
Luminal renal membrane
Basolateral renal membrane
Na+-K+ cotransport in inside-out erythrocyte
vesicles, erythrocytes, and luminal rena]
membrane
Na+-H+ countertransport in luminal renal
membrane
Calpain activity in erythrocytes and renal tubular
cells
105-kilodalton protein in membrane skeleton of
erythrocytes or 78-kilodalton protein in rena]
membrane
32- to 34-kilodalton protein in renal membrane
Calpain inhibitor in erythrocytes and renal tubular
cells
Isolation of the entire gene coding for the
105-kilodalton protein in membrane skeleton of
erythrocytes is in progress

Experimental findings
Pressor effect | in kidney of MHS41-
Na retention in MHS43
GFR and Na tubular reabsorption f in MHS*
GFR, Na tubular reabsorption, and O2
consumption f in MHS43- <*
Cell volumes or Na content | in MHS44- "
Cell volume or Na content \ in MHS47-
Na transport is equal between strains50-3I
Volume 1 in MHS30
Na transport t in MHS"-
Na transport f in MHS (unpublished data)
AH these activities are faster in MHS48- 3| - M
(and unpublished data)

Immunochemical difference between MNS and

MHS33-
Difference in isoelectric pattern56
Concentrations 10 times I in MHS33-M
Unpublished data

GFR = glomerular filtration rate; | or 1 = higher or lower in MHS than in MNS.

By type of experiment performed.
tThe characteristics of erythrocytes and blood pressure are highly correlated in F2 hybrids obtained by crossing the F, (MHS x MNS)
hybrids.37 Moreover, bone marrow transplantation between MHS or MNS in irradiated F, hybrids has shown that the origin of the
functional difference between erythrocytes of MHS and MNS is in stem cells.37

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

626

HYPERTENSION

integration occurs at an early stage in a germ cell

precursor, the foreign DNA will be passed to future
progeny. In this way, a difference between the two
strains of rats in a given DNA sequence coding for
a protein of interest may be tested as a possible
cause of a biochemical or physiological sequence of
events leading to hypertension (or to part of it). To
be successful, this experiment must be performed
between strains of rats whose genomes are similar
except for the DNA sequence of interest.
The need for this condition to be met is clear from
Figure 2. In fact, only if the polygenes forming the
genetic background, apart from those regulating blood
pressure, are identical between the two strains can
we test the effect of a given candidate gene. Such
strains of hypertensive and normotensive animals
currently are not available. However, strains of rats
with a reasonable approximation to the ideal situation can be selected. For instance, since normotension is partly dominant over hypertension in MHS
and MNS, a substrain of congenic normotensive rats
is being selected by repeatedly backcrossing normotensive hybrids to MHS. Successive generations
of rats obtained from consecutive backcrosses showed
blood pressure values ranging from normotension to
hypertension. The number of backcrosses needed to
reach a true congenic strain is critical because of the
possibility of linkage between the genes of interest.
In fact, the closer the associated genes, the higher
will be the number of backcrosses needed to separate them.59 For independent genes, however, seven
or eight backcross cycles should be sufficient to
attain more than 99% coisogenicity between the
congenic normotensive strain and the MHS strain.
At that point, virtually only the genes affecting blood
pressure should differ between them.
If a DNA polymorphism can be detected between
the two strains, the relative gene may be a suitable
candidate for a transgenic experiment. However,
the present methods used in transgenic animals
must be improved before the application of this
approach to understanding the genetic basis of
hypertension can become practical. First, transgenic techniques also need to be developed in rats
(the majority of studies have used mice). Second, a
precise targeting of the foreign DNA into the host
genome should be accomplished, along with the
capability of "switching off" the endogenous gene.
At present, the expression, if any, of the foreign
DNA may occur at different times with different
control mechanisms, as compared with the natural
genomic situation.
Future Directions
These approaches involving transgenic animals
are expected to be available soon. However, previously published results and future trends in
research in animal models or humans that are
likely to contribute to the identification of candidates, either as gene products or as DNA segments,

VOL 12, No 6, DECEMBER 1988

involved in blood pressure regulation may be grouped

into two categories.
First, genes coding for proteins already known as
blood pressure regulators, such as renin, angiotensinogen, mineralocorticoid and glucocorticoid receptors, kallikrein, and kininogen, have already been
described, while genes coding for other proteins of
this type are probably under investigation. The
study of polymorphisms both at DNA and protein
levels, either in human pedigrees with hypertensive persons or in strains of rats with genetic
hypertension, may provide information on their
possible primary involvement in the determination
of the set point of blood pressure. The previously
mentioned biometrical methods (measured genotypes) can quantify the relative importance of
these molecular differences as compared with the
remaining genetic background.
Second, extensive measurements of intermediate
phenotypes in human pedigrees, families, or strains
of rats have already been carried out. These measurements regard blood cell volume, ion concentration, ion transport across the blood cell membrane
mediated by different pathways (Na+-K+ pump,
Na+-K+ cotransport, Li+-Na+ countertransport, Na+H+ countertransport),60-62 and many other biochemical or physiological functions. These intermediate
phenotypes are being measured either because of a
logical link between them and some classic pressor
mechanisms (such as the release of neurotransmitters, the sodium or calcium concentration in vascular smooth muscle cells, or the renal retention of
sodium) or because they showed distinctive values
in hypertensive subjects (or rats) as compared with
values in their appropriate controls. To be relevant
to the understanding of the genetic mechanisms
according to the scheme shown in Figure 2, these
studies should include a combination of as many
biochemical and physiological tests as possible performed on all the individual subjects of the same
pedigree or in different rat strains. In fact, we need
to describe a substantial portion of the sequence
illustrated in Figure 2 and Table 2 in persons
belonging to the same pedigree (or rat strain) to be
confident that a given genetic link or interaction is
at work in that particular pedigree or rat strain. Of
course, the closer the intermediate phenotype is to
the DNA level, the lower the influence of other
concomitant genetic and environmental factors is
expected to be. Hypertension is so common and
heterogeneous that it would not be surprising _to
find, even in a single large pedigree, different molecular mechanisms with individual genetic characteristics and polymorphisms.
Acknowledgments
The authors gratefully acknowledge Mr. Barry R. Barber and
Dr. Rawley Foulkes for revising the English of the manuscript
and Mrs. Pinuccia Protasoni for secretarial assistance.

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

GENETICS OF ESSENTIAL HYPERTENSION/Camss/ and Bianchi

References
1. Murakami K, Hirose S, Miyazaki H, et al. Complementary
DNA sequences of renin: state of art review. Hypertension
1984;6(suppl l):l-95-I-100
2. Yamanaka M, Greenberg B, Johnson L, et al. Cloning and
sequence analysis of the cDNA for the rat atrial natriuretic
factor precursor. Nature 1984;309:719-722
3. Arriza JL, Weinberger C, Cerelli G, et al. Cloning of human
mineralocorticoid receptor complementary DNA: structural
and functional kinship with the glucocorticoid receptor.
Science 1987^237:268-275
4. Dixon RAF, Kobilka BK, Strader DJ, et al. Cloning of the
gene and cDNA for mammalian beta-adrenergic receptor
and homology with rhodopsin. Nature 1986^321:75-79
5. Ng WG, Rae TF, Donnell GN. Disorders of carbohydrate
metabolism. In: Emery AEH, Rimoin DL, eds. Principles
and practice of medical genetics. Edinburgh: Churchill Livingstone, 1983:1267-1285
6. Antonarakis SE, Kazazian HH, Orkin SH. DNA polymorphism and molecular pathology of human globin gene cluster. Hum Genet 1985;69:1-14
7. Seegmiller JE. Disorders of purine and pyrimidine metabolism. In: Emery AEH, Rimoin DL, eds. Principles and
practice of medical genetics. Edinburgh: Churchill Livingstone, 1983:1286-1305
8. Platt R. Heredity in hypertension. Q J Med 1948,17:111-132
9. Platt R. Heredity in hypertension. Lancet 1963;l:899-904
10. Hamilton M, Pickering GW, Roberts JAF, Sowry GSC. The
aetiology of essential hypertension: I. The arterial pressure
in the general population. Clin Sci 1954;13:11-35
11. Hamilton M, Pickering GW, Roberts JAF, Sowry GSC. The
aetiology of essential hypertension: II. Scores for arterial
blood pressure adjusted for differences in age and sex. Clin
Sci 1954,13:37-49
12. Hamilton M, Pickering GW, Roberts JAF, Sowry GSC. The
aetiology of essential hypertension: IV. The role of inheritance. Clin Sci 1954;13:273-274
13. Oldham PD, Pickering G, Roberts JAF, Sowry GS. The
nature of essential hypertension. Lancet 1960;l:1085-1093
14. Wilson C, Byron FB. The vicious circle in chronic Blight's
disease: experimental evidence from the hypertensive rat. Q
JMed 1941;10:65-93
15. Harris H, Smith CAB. The sib-sib age on onset correlation
among individuals suffering from a hereditary syndrome produced by more than one gene. Ann Eugen 1948;14:3O9-318
16. Vogel F, Motulski AG. Human genetics: problems and
approaches. Berlin; Heidelberg: Springer-Verlag, 1986:174
17. Harris H. Enzyme polymorphisms in man. Proc R Soc Biol
1966;164:298-310
18. McManus IC. Bimodality of blood pressure levels. Stat Med
1983;2:253-258
19. Miall WE, Heneage P, Kholsa T, Lovell HG, Moore P.
Factors influencing the degree of resemblance in arterial
pressure of close relatives. Clin Sci 1967^33:271-283
20. Havlik RJ, Garrison RJ, Feinleib M, Kannel WB, Castelli
WP, McNamara PM. Blood pressure aggregation in families.
Am J Epidemiol 1979,110:304-312
21. Hayes CG, Tyroler HA, Cassel JC. Family aggregation of
blood pressure in Evans County, Georgia. Arch Intern Med
1971;128:965-975
22. Longini IM, Higgins MW, Minton PC, Moll PP, Keller JB.
Environmental and genetic sources of familiar aggregation of
blood pressure in Tecumseh, Michigan. Am J Epidemiol
1984;120:131-144
23. Shull WJ, Harburg E, Erfurt JC, Sehork MA, Rice R. A
family set method for estimating heredity and stress: II.
Preliminary results of the genetic methodology in a pilot
survey of Negro blood pressure, Detroit, 1966-67. J Chronic
Dis 197023:83-92
24. Moll PP, Harburg E, Burns TL, Sehork MA, Ozgoren F.
Heredity, stress and blood pressure, a family set approach:
the Detroit project revisited. J Chronic Dis 1983;36:317-328

627

25. Biron P, Mongeau JG, Bertrand D. Familial aggregation of

blood pressure in 588 adopted children. Can Med Assoc J
1976;115:773-775
26. Vanderholen R, Brewer G, Honeyman MS, Morrison J,
Hobler SN. A study of hypertension in twins. Am Heart J
1970;79:454-457
27. Feinleb M, Garrison RJ, Fabsitz R, et al. The NHLBI twin
study of cardiovascular risk factors: methodology and summary of results. Am J Epidemiol 1977;106:284-295
28. Rapp JP. Genetics of experimental and human hypertension.
In: Genest J, Kuchel O, Hamet P, Cantin M. eds. Hypertension: physiopathology and treatment. 2nd ed. New York:
McGraw-Hill, 1983:582-598
29. Sing CF, Boerwinkle E, Turner ST. Genetics of primary
hypertension. Clin Exp Hypertens [A] 1986;8:623-651
30. Cavalli-Sforza LL, Bodmer WF. The genetics of human
populations. New York: Freeman, 1973:535-536
31. Mongeau J, Biron P, Sing CF. The influence of genetics and
household environment upon the variability of normal blood
pressure: the Montreal Adoption Study. Clin Exp Hypertens
[A] 1986;8:653-660
32. Elston RC, Stewart J. A general model for the genetic
analysis of pedigree data. Hum Hered 1971^1:523-526
33. Hasstedt SJ. A mixed model likelihood approximation in a
large pedigree. Comput Biomed Res 1982; 15:295-298
34. Williams RR, Dadone M, Hunt SC, et al. The genetic
epidemiology of hypertension: a review of past studies and
current results for 948 persons in 48 Utah pedigrees. In: Rao
DC, Elston RC, Kuller LH, et al., eds. Genetic epidemiology of coronary heart disease: past, present and future. New
York: Alan R. Liss, 1984:419-444
35. Penrose LS. Measurement of pleiotropic effects in phenylketonuria. Ann Eugen 1952;16:120124
36. Williams RR, Hunt SC, Dadone MM, Smith JB, Ash KO,
Kuida H. Cation flux and other possible biological markers
of genetically predisposed hypertension. In: Filler LJ, Lauer
RM, eds. Children's blood pressure, report of the 88th Ross
Conference on Pediatric Research. Columbus, OH: Ross
Laboratories, 1985:100-123
37. Boerwinkle E, Chakraborty R, Sing CF. The use of measured genotype information in the analysis of quantitative
phenotypes in man: I. Models and analytical methods. Ann
Hum Genet 1986^0:181-194
38. Sing C, Boerwinkle E, Moll PP, Templeton AR. Characterizing genes affecting quantitative traits in humans. In: Weir
BS, Eisen EJ, Goodman MN, Nankong J, eds. Proceedings
of the Second International Conference of Quantitative
Genetics. Sunderland, MA: Sinaver Associates, 1988:250-269
39. Goldblatt H, Lynch J, Hanzal RF, Summerville WW. Studies on experimental hypertension: production of persistent
elevation of systolic blood pressure by means of renal
ischemia. J Exp Med 1934;59:347-379
40. Rapp JP. A genetic locus (Hyp-2) controlling vascular smooth
muscle response in spontaneously hypertensive rats. Hypertension 1982;4:459-467
41. Bianchi G, Fox U, Di Francesco GF, Bardi U, Radice M.
The hypertensive role of the kidney in spontaneously hypertensive rats. Clin Sci Mol Med 1973;45:135s-139s
42. Bianchi G, Fox U, Di Francesco GF, Giovannetti AM, Pagetti
D. Blood pressure changes produced by kidney crosstransplantation between spontaneously hypertensive rats and
normotensive rats. Clin Sci Mol Med 1974;47:435-448
43. Bianchi G, Baer PG, Fox U, Duzzi L, Pagetti D, Giovannetti
AM. Changes in renin, water balance, and sodium balance
during development of high blood pressure in genetically
hypertensive rats. Circ Res 1975;36, 37(suppl I):I-153-I-161
44. Ferrari P, Cusi D, Barber BR, et al. Erythrocyte membrane
and renal function in relation to hypertension in rats of the
Milan hypertensive strain. Clin Sci 1982;63:61s-64s
45. Salvati P, Ferrario RG, Parenti P, Bianchi G. Renal function
of isolated perfused kidneys from hypertensive (MHS) and
normotensive (MNS) rats of the Milan strain: role of calcium. J Hypertens 1987^:31-38

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011

628

HYPERTENSION

46. Salvati P, Pinciroli GP, Bianchi G. Renal function of isolated

perfused kidneys from hypertensive (MHS) and normotensive (MNS) rats of the Milan strain at different ages. J
Hypertens 1984;2(suppl 3):351-353
47. Ferrari P, Nussdorfer G, Torielli L, et al. Renal function and
cell characteristics during the development of essential and
genetic hypertension. In: Hofman A, Grobbee DE,
Schalekamp MADH, eds. Early pathogenesis of primary
hypertension. Amsterdam: Elsevier Science Publishers, 1987:
111-123
48. Ferrari P, Ferrandi M, Torielli L, Canessa M, Bianchi G.
Relationship between erythrocyte volume and sodium transport in the Milan hypertensive rat and age-dependent changes.
J Hypertens 1987 ;5:199-206
49. Beck F, Bianchi G, Dorge A, Rick R, Schramm M, Thurau
K. Sodium and potassium concentration of renal cortical
cells in two animal models of primary arterial hypertension.
J Hypertens 1987;l(suppl 2):38-39
50. Ferrari P, Torielli L, Ferrandi M, Cirillo M, Bianchi G.
Volumes and Na+ transports in intact red blood cells,
resealed ghosts and inside-out vesicles of Milan hypertensive rats. J Hypertens 1986;4(suppl 6):S379-S381
51. Parenti P, Hanozet GM, Bianchi G. Sodium and glucose
transport across renal brush border membranes of Milan
hypertensive rats. Hypertension 1986;8:932-939
52. Hanozet G, Parenti P, Salvati P. Presence of a potential
sensitive Na transport across renal, brush border membrane
vesicles from rats of the Milan hypertensive strain. Biochim
Biophys Acta 1985;819:179-186
53. Pontremoli S, Melloni E, Salamino F, et al. Decreased level of
calpain inhibitor activity in kidney from Milan hypertensive
rats. Biochem Biophys Res Commun 1987;145:1287-1294

VOL 12, No 6, DECEMBER 1988

54. Pontremoli S, Melloni E, Salamino F, et al. Decreased level

of calpain inhibitor activity in red blood cells from Milan
hypertensive rats. Biochem Biophys Res Commun 1986;138:
1370-1377
55. Ferrandi M, Salardi S, Modica R, Bianchi G. Antigenic
properties of erythrocyte membranes from a hypertensive
strain of rats. Ann NY Acad Sci 1987^04:584-587
56. Parenti P, Modica R, Salardi S, Righetti PG, Gianazza E,
Bianchi G. Alteration of protein composition in kidney brush
border membrane vesicles from hypertensive rats [Abstract].
In: Murer H, Rossier B, Warnock D, eds. Proceedings of the
Tenth ISN Congress Meeting: Structure, function and regulation of membrane transport proteins (Stansstad, Switzerland, August 3-6, 1987)
57. Bianchi G, Ferrari P, Trizio D, et al. Red blood cell abnormalities and spontaneous hypertension in the rat: a genetically
determined link. Hypertension 1985;7:319-325
58. Palmiter RD, Brinster RC. Transgenic mice. Cell 1985;41:
343-345
59. Green EL. Mating systems. In: Green EL, ed. Biology of the
laboratory mouse. New York: McGraw-Hill, 1966:11-22
60. Postnov YV, Orlov SN. Ion transport across plasma membranes in primary hypertension. Physiol Rev 1985;65:904-945
61. Bianchi G. Ion transport across blood cell membrane in
essential hypertension. In: Robertson JB, ed. Current opinion in cardiology; vol 1. London: Gower Medical, 1986:
634-639
62. Baudouin-Lagros M, Cloix JF, Crabos M, et al. Membrane
markers of arterial hypertension. In: Zanchetti A, Tarazi
RC, eds. Pathophysiology of hypertension: regulatory mechanisms. Amsterdam: Elsevier Science, 1986:670-686 (Handbook of hypertension; vol 8)
(Hypertension 12: 620-628, 1988)

Downloaded from http://hyper.ahajournals.org/ by guest on November 28, 2011