This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Corvol, P. Articles by Jeunemaitre, X. Search for Related Content PubMed PubMed Citation Articles by Corvol, P. Articles by Jeunemaitre, X.

Molecular Genetics of Human Hypertension: Role of Angiotensinogen¹

INSERM U36, Collège de France, 75005 Paris, France

	Abstract

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 

I. Introduction

II. Structure of the AGT Gene

III. Characteristics of AGT

A. AGT as the substrate for renin

B. AGT as a member of the serpin superfamily: other roles for AGT?

IV. Tissue and Cellular Distribution

V. Regulation

A. Hormonal regulation of AGT

B. Regulation of AGT in disease

C. Regulation of AGT gene transcription

VI. Molecular Genetics of AGT in Human Essential Hypertension

A. Strategies

B. Linkage studies

C. Association studies of AGT polymorphism and high blood pressure

D. Pregnancy-induced hypertension

E. M235T variants, plasma AGT concentration, and blood pressure

F. Association of AGT variants with the renal blood flow response to Ang II infusion

VII. Molecular Genetics of AGT and Other Diseases

A. AGT variants and coronary heart disease

B. AGT variants and complications of diabetes

C. AGT, obesity, and blood pressure

VIII. Effects of AGT Gene Variants and Gene Targeting on Ang I Production and Blood Pressure

A. Human studies

B. Genetic studies in rats and mice

C. Genetic hypertension in rats

D. Antisense AGT oligodeoxynucleotides

E. Transgenic animals

F. AGT gene duplication

IX. Conclusions and Perspectives

	I. Introduction

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
BLOOD pressure is regulated by a variety of mechanisms that involve the products of several genetic loci and a number of environmental factors. The heritable component of blood pressure has been documented in several studies of families and twins, and it is generally accepted that approximately 30% of the changes in blood pressure are attributable to genetic heritability and approximately 50% to environmental influences (1). Except for rare monogenic forms of hypertension such as glucocorticoid-suppressible hyperaldosteronism (2), Liddle’s syndrome (3), and apparent mineralocorticoid excess (4), little is known about the number of genes actually involved in human essential hypertension, their quantitative effect on blood pressure, their mode of transmission, or their interaction with other genes and environmental components.

Research on the molecular genetics of human hypertension developed less than 10 yr ago. Its goal was to identify the loci involved, to detect gene variants at the loci identified, to associate them with intermediate phenotypes, and ultimately to estimate their quantitative effects on blood pressure and their interaction with the principal environmental factors. The approach taken by several groups was to study candidate genes that may contribute to abnormal blood pressure because of their known effect on the cardiovascular system. The genes of the renin-angiotensin system have been most extensively studied because of the well documented role of this system in the control of blood pressure and in the pathogenesis of several forms of hypertension.

Renin acts on a single substrate, angiotensinogen (AGT), which is synthesized mainly by the liver and released into the circulation. In association with the Utah group, we first identified the role of AGT gene polymorphism in human essential hypertension (5). Several AGT gene variants were discovered, and one of them was found to be associated with an increase in plasma AGT and hypertension. A number of linkage and association studies have since examined several clinical situations and various populations. This review will discuss briefly the structure of the AGT gene, the principal characteristics of the protein, and its tissue distribution and regulation and will analyze the findings of genetic studies that support the role of AGT in human hypertension and certain other cardiovascular diseases.

	II. Structure of the AGT Gene

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
The human AGT cDNA is 1,455 nucleotides long and codes for a 485-amino acid protein (6). The AGT gene contains five exons and four introns, which span 13 kb (Fig. 1). The first exon (37 bp) codes for the 5'-untranslated region of the mRNA. There are two potential ATG sites, and the second exon codes for a signal peptide of 24 or 33 residues and the first 252 amino acids (59%) of the mature protein. Mature AGT contains 452 amino acid residues; the first ten amino acids correspond to angiotensin I (Ang I), and the other larger portion corresponds to des(Ang I)AGT. Exon 3 codes for 90 amino acids and exon 4 codes for 48 residues. Exon 5 contains a short coding sequence (62 amino acids), followed by a long 3'-untranslated sequence with two polyadenylation signals, accounting for the existence of two mRNA species differing in length by 200 nucleotides (7).

View larger version (16K): [in this window] [in a new window]   Figure 1. Schematic structure of the AGT gene, mRNAs, and proteins [Derived from Ref. 6.]

	III. Characteristics of AGT

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
Human AGT is a globular glycoprotein with a molecular mass between 55 and 65 kDa, depending on its state of glycosylation. It contains four putative N-linked glycosylation sites (Asn-X-Ser/Thr) that can be the origins of complex glycosylation chains. Although the role of this glycosylation process is not known in humans, the glycosylation status alters the AGT clearance rate in rats. The in vivo half-lives of two differently glycosylated forms of rat AGT are compatible with a two-compartment model and suggest that the more highly glycosylated form is secreted faster by the liver and eliminated more rapidly by the kidney than is the less glycosylated form (17). There are four cysteine residues in the human mature protein, three of which are also present in rat and mice AGT, but there are only two in sheep AGT. The arrangement of their disulfide bridges is not known.

A high molecular mass form of AGT is present in human plasma, especially in pregnant women, where it accounts for about 15% of the total AGT. High molecular mass AGT is the predominant form of the protein in amniotic fluid (18). It has been shown recently to be a covalent complex between AGT and two other proteins (19). A major part is formed by a 2:2 complex of AGT and the proform of a placental protein, the eosinophil major basic protein (proMBP); the proteins probably interact through disulfide bonds. A fraction of this complex seems to bind two molecules of complement C3dg in a 2:2:2 pattern to give a 300-kDa complex. The function of high molecular mass AGT is not known.

AGT is usually measured by an enzymatic assay for Ang I after its complete hydrolysis by excess renin. Direct immunoassays, using polyclonal and monoclonal antibodies against AGT, that measure both intact AGT and its inactive C-terminal part, des(Ang I)AGT, have also been developed (20). This technique accurately quantifies AGT in tissues such as the liver, which contains mainly enzymes capable of metabolizing Ang I, making the interpretation of the enzymatic assay difficult (21).

A. AGT as the substrate for renin
The rate-limiting step in the cascade of enzymatic events that make up the circulating renin-angiotensin system is the first step, the cleavage of AGT by renin. The plasma AGT concentration is in the micromolar range, whereas the plasma renin concentration is 1000 times lower. AGT is present in both the plasma and extracellular fluids and can be considered to be a "reservoir" for the action of renin. Another reservoir of AGT is the cerebrospinal fluid, which has the highest proportion of AGT (per mg of total protein) of all biological fluids (20, 22). Short-term regulation of the renin-angiotensin system does not seem to depend on changes in the plasma AGT concentration. For example, the rapid adaptation to changes in sodium intake is mediated by an abrupt change in the amount of renin released (which can vary more than 10-fold) by the renal juxtaglomerular (JG) cells (occurring within seconds or minutes), but changes in plasma AGT concentration are rather slow (occurring over hours or days).

The limiting role of AGT in the generation of plasma Ang I and therefore of plasma angiotensin II (Ang II), since angiotensin converting enzyme (ACE) is not rate limiting, is of great importance for evaluating its exact role in blood pressure regulation and hypertension. The affinity of renin for its substrate, its Michaelis-Menton constant (K_m), is about 1.25 ± 0.1 µmol/liter, more than 10-fold lower than the K_m for the homologous synthetic tetradecapeptide renin substrate (20.7 ± 7 µmol/liter) (23, 24). As the plasma AGT concentrations in rats and humans are about 1 µmol/liter, the plasma AGT would have to be 10-times greater for its hydrolysis to be a zero-order reaction. Hence the large amount of AGT in the plasma probably does not provide an excess of substrate for renin. Before genetic studies were undertaken, there was direct in vivo evidence for the limiting role of plasma AGT in Ang I generation from several independent sources. One was the decrease in blood pressure and PRA that occurs in rats after the injection of anti-AGT antibodies. The drop in blood pressure is dependent on the sodium balance and does not occur in bilaterally nephrectomized animals (25). The second was the increase in blood pressure that occurs after infusion of pure AGT into salt-depleted rats (26). These results suggest that the in vivo AGT concentration is most important for the activation of the renin system.

B. AGT as a member of the serpin superfamily: other roles for AGT?
The AGT gene belongs to the serpin (serine protease inhibitor) superfamily, which includes 1-antitrypsin, 1-antichymotrypsin, and antithrombin III. The structure of the AGT gene is similar to those of the other genes in terms of exon number, size, and splicing sites (27, 28). The structural similarity between AGT and these proteins (20% amino acid sequence identity) suggests that the genes of the serpin superfamily evolved from a primitive common ancestor some 500 million yr ago through a series of gene duplications, insertions, and deletions (28). AGT is probably one of the more distant members of the serpin superfamily and has probably lost its serine protease-inhibitory activity. Experiments conducted in this laboratory failed to show that AGT had any antiesterase activity when tested against various serine esterases (L. Wei, E. Clauser, and P. Corvol, unpublished observations).

AGT is an acute phase protein in the rat: its concentration increases markedly in response to inflammatory agents and tissue injury, but its function in these conditions is unknown (29, 30, 31). Its acute-phase response appears to be less marked in humans (32). Finally, the part of AGT remaining after renin cleavage, des(Ang I)AGT, has no known functions. Perhaps AGT [or des(Ang I)AGT] inhibits an unidentified proteolytic enzyme in these inflammatory states. AGT may stimulate the renin-angiotensin system locally in fever-induced vasodilatation, counteracting vasodilation and so helping to maintain blood pressure.

	IV. Tissue and Cellular Distribution

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
AGT is synthesized in the liver, and the plasma AGT concentration reflects mainly this synthesis. AGT is not stored in hepatocytes, but is constitutively secreted. Like other components of the renin-angiotensin system that are present in tissues involved in cardiovascular functions (33), AGT is synthesized in many tissues other than the liver. Even though these tissues produce less than in the liver, AGT appears to be an important component of the extravascular local renin-angiotensin system. Since the K_m of renin for AGT is relatively high, the rate of Ang I production by these systems is probably controlled by the local AGT concentration, rather than by the extravascular renin concentration. Other proteases in tissues may also hydrolyze AGT to Ang I (or even directly to Ang II) (34). The Ang II produced locally by these systems may have paracrine or autocrine effects. These tissue systems may well be regulated locally, independently of the circulating system.

The brain, large arteries, kidney, and adipose tissues are all established sites of AGT synthesis (35, 36, 37). In situ hybridization and immunohistochemistry studies have shown that glial cells are the most important source of AGT in the brain (38, 39, 40). Cultures of astrocytes from human brain tissue also produce AGT (41). It has been suggested that the AGT synthesized by glial cells is secreted and then taken up by neurons, thus forming a paracrine renin-angiotensin system (42). AGT is also present in the vascular walls, in the adventitia, and in the medial smooth muscle cell layer (43, 44, 45, 46). The principal AGT mRNA in the rat aorta seems to be in the brown adipose tissue (43). Thus the presence of highly vascularized adipose tissue around these vessels raises the possibility of a local angiotensin-generating system in which the adipose cells synthesize AGT. Ang II could then be generated by the actions of circulating renin and adventitial and/or endothelial ACE. There is, indeed, evidence for the local generation of angiotensin from AGT in isolated perfused rat blood vessels (47).

Most of the AGT mRNA in the heart is in the atria, and it is less abundant in the ventricles (35, 48). The presence of mRNA for the various components of the renin-angiotensin system, including Ang II receptors, in the heart is good evidence that the renin-angiotensin system has a physiological role in cardiac function. The amount of cardiac AGT appears to be hormonally regulated (35). Cardiac AGT mRNA is also increased in animals placed on a low-salt diet (49) and decreased in animals treated with ACE inhibitors (48).

AGT mRNA has been detected in the kidney (35, 50), mostly in proximal tubule cells (51). Changes in dietary sodium modulate the amount of AGT in the kidney: a low-salt diet increases the AGT mRNA in the proximal tubule cells (52), whereas the amount of AGT mRNA in the periaortic fat and liver remain unchanged. The renal AGT concentration, which also depends on circulating AGT (53), could be an important rate-limiting factor for the local generation of Ang I, at least during converting enzyme inhibition, and therefore a determining factor in the generation of renal angiotensin peptides and in the fine tuning of renal vascular resistance.

White adipose tissue, a rich source of AGT, is produced during the differentiation of preadipocytes to adipocytes (54). The recruitment of new fat cells could be enhanced by Ang II, via the release of prostaglandins that are adipogenic (55). The amount of AGT in white adipocytes in rats is influenced by their nutrition; it is reduced during fasting and increased with refeeding (56). The amounts of AGT protein and mRNA in retroperitoneal depots decline with age (57), suggesting that AGT is important for the growth of adipose tissue.

Finally, AGT mRNA has been detected in several other human tissues, such as the adrenal gland, where it is present in both the cortex and medulla (58). It could contribute, together with locally produced renin (59), to the paracrine generation of Ang II and the regulation of aldosterone secretion.

	V. Regulation

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
The plasma and tissue AGT concentrations both contribute directly to the circulating amounts of Ang I and Ang II. Thus, changes in these concentrations are important, as they can influence the degree of activation of the renin-angiotensin system. AGT is regulated by several hormonal factors, and some of the cis- and trans-regulatory elements involved in the regulation of its transcription have been identified.

A. Hormonal regulation of AGT
A wide variety of in vivo and in vitro models have been used by investigators to demonstrate that glucocorticoids (especially dexamethasone), estrogens, Ang II, and thyroid hormones all stimulate the synthesis and release of AGT (see reviews in Ref. 23 and in Refs. 60–62). These hormones increase the amount of AGT mRNA in the rat liver (35, 44, 63, 64). The hormonal regulation of AGT mRNA has also been studied in both hepatic (65) and nonhepatic cell lines, such as pancreatic cells (66) and adipocytes (56). The effects of steroids are neutralized by antiglucocorticoids and antiestrogens; thyroidectomy decreases the plasma AGT concentration, and T₃ restores it. Nevertheless, the stimulation of AGT synthesis by Ang II is still a matter of debate. While Ang II has been consistently found to stimulate the synthesis of AGT by cells in culture, a study in rats casts considerable doubt on its tonic influence on AGT production in vivo (67). The fall in plasma AGT that occurs during chronic ACE inhibition is probably due to an increase in its consumption due to the rise in plasma renin and to a decrease in its synthesis (68), although other mechanisms are possible since ACE inhibition does not block only the renin-angiotensin system.

The plasma concentrations of AGT and estrogens increase in parallel during pregnancy (69). Oral contraceptive pills containing synthetic estrogens also cause a dose-dependent increase in plasma AGT (70), but there seems to be no direct relationship between the increase in plasma AGT and blood pressure. However, there were subtle changes in renal blood flow in women whose circulating AGT concentrations were increased by ingestion of oral synthetic estrogens (71). This stimulatory effect is not observed when the estrogen is given percutaneously (72), perhaps because estradiol does not accumulate in hepatic cells.

Assuming that the rate of Ang I formation at the usual plasma AGT concentration is one-half maximal, it appears logical to suspect that the increase in AGT produced by synthetic estrogen or glucocorticoid treatment plays a role in the pathophysiology of some secondary forms of hypertension, such as oral contraceptive-induced hypertension (73), or Cushing’s syndrome (74). Similarly, the 3–5 mm Hg increase in systolic blood pressure that occurred in women taking high doses of synthetic estrogens as contraceptives might have been due to activation of the renin system through an increase in plasma AGT (75).

The estrogen-induced increase in AGT is tissue-dependent. The effect of estrogen on the liver is important, but it does not stimulate AGT production in adipose tissue, and contradictory results have been reported in brain (40). Campbell et al. (76) showed that the estrogen-induced increases in plasma AGT are associated with increases in renal angiotensin peptides, despite reduced plasma renin and angiotensin peptide concentrations. Bilateral nephrectomy also leads to an increase in the hepatic synthesis and plasma concentration of AGT (77), but the manner in which it does so is thus far unknown (23, 78, 79). The additive effects of these various stimuli suggest that they trigger AGT synthesis through different transcriptional pathways.

The observations described above suggest that a chronic increase in plasma AGT might facilitate hypertension in predisposed individuals, who may also have an abnormally short feedback loop between Ang II and renin release. The increase in plasma Ang II after an increase in plasma AGT exerts a short negative feedback loop on renin release in the normal physiological state, and this limits the direct effect of changes in plasma AGT on blood pressure.

B. Regulation of AGT in disease
Several pathological conditions lead to increased AGT production, which contributes to the local generation of Ang II. The medial layer of an injured aorta contains increased amounts of AGT mRNA, as might be expected for a member of the serpin family, which suggests that AGT is involved in myointimal proliferation in response to vascular injury (80). Increased ACE activity and AGT synthesis also occur after experimental left ventricular hypertrophy (81), suggesting, along with other studies, that the cardiac renin system may help modulate the growth and hypertrophy of the heart. The change in renal hemodynamics caused by experimental heart failure leads to a specific increase in renal AGT mRNA, suggesting that it contributes to the activation of the intrarenal renin system (82).

C. Regulation of AGT gene transcription
AGT production is probably regulated mainly by controlling gene transcription, although there is evidence for posttranscriptional regulation, as when Ang II increases the stability of AGT mRNA (83). The cis- and trans elements that regulate the AGT gene have been recently reviewed by Brasier and Li (84), and only those most relevant to AGT genetics will be summarized here. Recent studies in mice indicate that the proximal promoter region of the mouse AGT gene (-96 to +22) is sufficient for expression of the AGT gene in mouse fibroblast cells during their differentiation into adipocytes (85). A liver-specific (AGF2) factor binds to the proximal promoter element (-96 to +52), and an ubiquitous nuclear factor (AGF3) binds to the core promoter element (-6 to +22). These two factors seem to act synergistically (85). The core promoter region of the human AGT gene in HepG2 cells has been analyzed by the same group (86). Electrophoretic mobility shift assays demonstrated that an ubiquitously expressed nuclear factor (AGT core promoter binding factor 1, AGCF1) bound to a region between positions -25 to -1, denoted AGCE1 (AGT core promoter element 1) (see Fig. 2), located between the TATA box and the transcription initiation site. AGCE1 appears to play a major role in activating AGT transcription, in particular by the downstream core elements. This region is probably more complex with several nuclear factors binding to its 5'- or 3'-side (87), which may be important for the general rate of transcription initiation and also for determining the pattern of AGT gene expression. These studies point to the importance of these regulatory regions where several natural variants of the AGT gene have been detected.

View larger version (9K): [in this window] [in a new window]   Figure 2. Localization of the missense mutations observed at the AGT gene. 1 = L10F; 2 = T104M; 3 =174M; 4 = L209I; 5 = L211R; 6 = M235T; 7 = Y248C; 8 = L359M; 9 = V388M; (GT)n microsatellite marker is located at approximately 5 kb downstream the 3'-end of the AGT gene.

	VI. Molecular Genetics of AGT in Human Essential Hypertension

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

A. Strategies
A number of strategies can be used for genetic analysis of complex diseases, such as hypertension, in which several genes and environmental factors act in concert to determine the phenotype (88). In addition to developing genetic models of high blood pressure, three categories of statistical analysis can be used in humans. One is linkage analysis, in which sibling pairs are analyzed, the second is association studies, in which large numbers of cases and controls are analyzed, and the third is linkage-association studies, in which the association between a genetic marker and the disease within families is examined. Each method has its advantages and disadvantages. Most of them have been used to establish the relationships between hypertension and the AGT gene. The following brief discussion will give the reader a clearer understanding of the studies performed thus far.

Classic linkage studies, which test the cosegregation of a genetic marker and a trait, have been extremely useful for determining the genes involved in Mendelian traits. However, they require a specific genetic model, although some adjustments can be made. Use of the wrong model can lead to false positive or negative results, which makes the use of lod score analysis difficult in complex diseases that result from the interactions between several frequent susceptibility genes and the environment. Thus, allele-sharing methods, involving mainly sibling pairs, appear to be the best approach to testing the linkage between hypertension (taken here as qualitative trait) and genetic markers of a candidate gene (such as those of the renin-angiotensin system, and in this case AGT). This nonparametric approach does not require any a priori specification of the model for the inheritance of the trait and can be used on a single generation, thus avoiding some ambiguity about the age-related influence on blood pressure. The statistics are based on the number of alleles shared by the affected siblings compared with the random Mendelian distribution. While this is a robust method, it has relatively low statistical power and requires highly polymorphic markers and a large collection of well characterized sibling pairs. Such family collections have been identified so far in a few hypertension clinics. Another approach is to select sibling pairs contrasted for the trait (89). Although this strategy might appear more powerful, it is somewhat difficult to select individuals with low blood pressure values within a family in which hypertension is frequent.

The linkage analysis must be interpreted with caution regardless of which method is used. Samples that are too small, frequent disease-related alleles, and genetic heterogeneity are some of the reasons that might lead to false negative results. Conversely, a false estimation of the allele frequencies of a polymorphic marker, a low threshold of statistical significance, and a strategy of multiple testing of genes involved in the same hypertensive families can all lead to false positive results. These possibilities emphasize the need for using several approaches before declaring a locus to be the cause of a complex disease such as essential hypertension.

Association studies compare the frequencies of a marker (e.g., AGT gene polymorphism) in an affected population and in a control group (case-control study). Since it does not postulate any genetic model, this methodology is well suited to the detection of modest genetic effects within a complex and heterogeneous disease, such as high blood pressure (90). The statistical analysis is simple, based on a contingency table comparing the allele or the genotype frequencies. The probability of detecting an association between a candidate gene and the disease will depend on the allele frequency, the strength of its relationship with the disease, and the homogeneity of the population studied. A positive result may indicate a causal allele or an allele in linkage disequilibrium with the causal mutation. A negative result has almost no power, since it may be argued that other polymorphisms at the gene locus remain to be discovered. The main pitfall of this method is the possibility of false-positive results, mainly due to a population admixture and the failure to carefully select many well characterized cases and controls. As will be shown below, this problem is crucial for the M235T polymorphism of the AGT gene, whose allele frequency varies widely according to the subject’s ethnic origin. The use of an internal control within the family can provide a perfect match for ethnic ancestry. The haplotype relative risk method is based on a comparison of the genotype of the proband and the uninherited genotype (91). The transmission disequilibrium test is based on the frequency with which the associated allele transmitted from heterozygous parent to the affected proband (92).

Both linkage and association studies have been used to analyze the role of the AGT gene in essential hypertension and thus will be discussed separately.

B. Linkage studies
An extensive study of the potential role of the AGT gene in human essential hypertension was performed on two large series of hypertensive sibships yielding a total of 379 sibling pairs (Salt Lake City, UT, and Paris, France) (5). The highly polymorphic GT microsatellite located in the 3'-region of the AGT gene (11) and the powerful affected sibling pair methodology were used to obtain evidence of a genetic linkage between the AGT gene and hypertension in the whole set of families (5). A 17% excess of AGT allele sharing was found in severely hypertensive sibling pairs (subjects with a diastolic blood pressure above 100 mm Hg or being treated with two or more antihypertensive medications). While significant linkage was obtained in male pairs in both the Utah and Paris groups, no excess of shared AGT alleles was observed in female comparisons, suggesting the influence of an epistatic hormonal phenomenon. From these studies, it was estimated that mutations at the AGT locus might be a predisposing factor in at least 3–6% of hypertensive individuals younger than 60 yr of age. When the same methodology was used to analyze the same hypertensive sibling pairs from Utah or Paris, it showed no linkage with other genes of the renin system: renin (93), ACE (94), or Ang II type I receptor (95).

Two other linkage studies also indicate a relationship between the AGT locus and hypertension. Caulfield et al. (96) showed a strong linkage (25% excess of concordance) between the AGT gene locus and essential hypertension in a set of 63 British families. This linkage was also observed in the subgroup of patients with diastolic pressure above 100 mm Hg, but there was no difference among female-female pairs. However, there was no association between hypertension and the 174M or the 235T AGT variant in this population. The same group also found linkage and association of the AGT locus with high blood pressure in 63 affected sibling pairs of Africo-Caribbean origin (97), suggesting some similarity in the genetic basis of essential hypertension in populations of different ethnicity. The corroboration of these linkage studies, shown in Table 1, indicate that molecular variants of the AGT gene, such as M235T or those in linkage disequilibrium with this variant, are inherited predispositions to essential hypertension in humans.

The results of these studies must be interpreted with caution for several reasons. First, the frequency of the 235T (or 174 M) AGT allele varies from an ethnic group to another. The 235T allele is more frequent in the Asian (0.75) than in the Caucasian population (0.35) and is by far the predominant allele in the African population (up to 0.93 in Nigerians) (98, 99). As a consequence, positive results may arise from population admixture, and negative results in populations in which this allele is predominant may be due to the constraint of limited statistical power. Second, differences in the design of each study (especially interacting factors such as sex, age, body mass index) and the choice of the control group (which is critical in a case-control study) make any overall comparison difficult. For example, the negative study reported by Bennett et al. (100) deals with hypertensive subjects, both parents of which had histories of high blood pressure, which may lead to an aggregation of genes (100). Finally, most of the results reported to date have been obtained with relatively small numbers of patients, which can also generate false-negative or false-positive results. Because most studies linking AGT polymorphism and high blood pressure have clearly indicated the ethnicity of the populations, the results will be discussed accordingly.

1. Caucasians.
There are divergent results on the association between variants in AGT at position 235 (235T) or 174 (174M) and hypertension in Caucasians. Our original study (5) found that the M235T variant (Met Thr in amino acid position 235) was more frequent in hypertensive probands from Utah and Paris, especially in the more severe index cases (allele frequency: 0.50), than in controls (0.38). A subsequent study on 136 mild-to-moderate hypertensive subjects (101) also found that the frequency of the 235T allele was increased, although the increase was significant only for patients with a family history of hypertension. Schmidt et al. (102) also found a higher frequency of the 235T allele in subjects with hypertension, a family history of the disease, and early onset of hypertension. The effect of AGT gene polymorphism on blood pressure was studied in a genetic isolate, the Hutterite Brethren, in North America. Seven hundred forty one Hutterites were tested for an association between the systolic and diastolic blood pressure, and the M235T and T174M genotypes. There was a significant association between the resting systolic blood pressure and the 174M allele only in men (103): the AGT 174M allele accounted for 3% of the total variance in systolic blood pressure. A subsequent study showed an autosomal codominant effect of the AGT 174T allele on systolic blood pressure (104).

Other studies have not found any association between the 235T variant and hypertension. Caulfield et al. (96) found no association between hypertension and the 235T AGT variant in their study of 63 British families. Similar results were obtained by Barley (105) for 64 cases and 74 controls. These two studies on UK individuals showed similar frequencies in hypertensives (0.50), but quite different frequencies in normotensives (0.40). Bennett et al. (100) found no association in 92 individuals, who were offspring of two hypertensive parents and who were ascertained through self-referral. Fornage et al. (106) found no association in a population-based selection of 104 hypertensives with mild hypertension and 195 matched normotensives, but there was a surprising difference in the frequency of the T235 allele in normotensive men (0.47) and women (0.36). More recently, Hingorami et al. (107) found no difference in the frequency of M235T in 223 hypertensive subjects and 187 normotensive individuals in East Anglia in the United Kingdom. A more powerful study performed in Finland by Kiema et al. (108) gave negative results with 508 mild hypertensives and 523 population-based controls.

In any case, four large studies recently showed a positive association between the M235T allele and essential hypertension. An association with severe hypertension, with a stronger relationship for men than for women, was found in a sample from the Framingham Heart Study and from the Atherosclerosis Risk in Community study (ARIC), when the effects of body mass index and triglycerides were taken into account. The proportions of cases attributable to the 235T allele were 8% in the ARIC population and 20% in the Framingham population (109). In a study testing a large number of AGT alleles, the frequency of the 235T allele was 0.47 in the 477 probands of hypertensive families and 0.38 (P = 0.004) in the 364 Caucasian controls (110). In a cross-sectional sample of 634 middle-aged subjects from the MONICA Augsburg cohort, Schunkert et al. (111) found that individuals carrying at least one copy of the T235 allele had higher systolic and diastolic blood pressures and were more likely to use antihypertensive drugs. Finally, another case-control study involving 802 hypertensive subjects and 658 Caucasian controls has shown a significant increase in the frequency of the T235 allele in men (0.46 vs. 0.40, P = 0.003) and in women whose hypertension was diagnosed before they were 45 yr of age (L. Tiret, H. Blanc, J. B. Ruivadets, D. Arveiler, G. Luc, X. Jeunemaitre, J. Tichet, C. Mallet, O. Poirier, P. F. Plovin, and F. Cambien, submitted). These studies emphasize the importance of sample size testing for susceptibility locus in a complex disease such as hypertension.

2. Japanese.
Fewer variations in the pathophysiology of hypertension and/or a more homogeneous genetic and environmental background may explain why the results reported for Japanese populations are more uniform. The frequency of the T235 allele is invariably high in the Japanese population, ranging from 0.65 (112) to 0.75 (113). An association between hypertension and molecular variants of AGT in the Japanese population has been found in several independent studies. All the studies reported so far show that the AGT variant 235T is more frequent in essential hypertensives than in normotensive controls (112, 113, 114, 115), except for one study that found a higher frequency of the 174M allele in hypertensives than in normotensives (116). Another study found the TT genotype of the AGT gene was a predictor of blood pressure in a subpopulation less than 50 yr old within a total of 347 Japanese subjects (117).

3. African and Americans of African origin.
Hypertension is more prevalent and more severe in African and African-Americans than in Caucasians. The genetic factors that contribute to this ethnic difference are not known. These hypertensive populations tend to be sensitive to sodium and have low circulating renin concentrations. The 235T allele is by far the most frequent in Africans (0.90) and in African- Americans (0.80) (118), which is consistent with an approximate 25% admixture of genes of European origin in the African-Americans (99). A weak association between blood pressure and AGT has been found in a community survey in Jamaica, but the effect was more marked in conjunction with the plasma ACE concentration (119). The frequency of the M235T variant has been evaluated in several studies, but it is difficult to rule out an association of this allele with hypertension (98) because of the high prevalence of this allele in these populations.

The frequent occurrence of the 235T allele in most human populations and a thorough haplotype analysis of ten diallelic variants of the AGT gene (110) suggests that the 235T variant of AGT is the ancestral form and the 235M allele is the neomorph. This is of interest when examining the development of the renin-angiotensin system across species, as the way in which living creatures adapted to their salt environment, by maintaining blood pressure and sodium balance through vasoconstriction and salt retention (120). Hominids were probably almost exclusively vegetarians several million years ago; therefore their diet contained very little sodium and a large amount of potassium, which can be a strong selective environment. That selection pressure could have favored salt retention mechanisms that had developed earlier in phylogeny, resulting in the selection of alleles giving optimal salt reabsorption. Since the 235T allele is associated with elevated circulating AGT that could influence the activity of the renin-angiotensin system, the AGT 235T allele could be one of the genes that fostered increased sodium reabsorption and increased blood pressure.

D. Pregnancy-induced hypertension
The above studies all suggest that a susceptible allele of the AGT gene is associated with increases in the plasma AGT concentration and in blood pressure. This effect could be more marked in conditions in which the synthesis of AGT is stimulated, such as in patients receiving oral estrogen or in pregnancy. Clinical studies have documented a familial tendency to develop preeclampsia, and familial studies have suggested that there are both genetic and environmental factors involved (121, 122).

Two reports initially indicated that the AGT locus played an important part in pregnancy-induced hypertension. Ward et al. (123) found a significant association between the AGT M235T variant and preeclampsia in both Caucasian and Japanese patients. Arngrimsson et al. (124) analyzed the allelic inheritance of the GT repeat in 52 sibling pairs of preeclamptic sisters and found a significant linkage between the AGT locus and preeclampsia in Icelandic and Scottish families. A more recent case-control study on 139 Japanese women with pregnancy-induced hypertension and 278 matched controls found a significant association with the 235T variant (125). However, two other limited studies in Caucasian women found no indication of cosegregation or association between preeclampsia and an AGT allele (126, 127). Thus, although the mechanisms proposed for preeclampsia and essential hypertension are different, a frequent variant of the AGT gene could predispose subjects to both diseases.

E. M235T variants, plasma AGT concentration, and blood pressure
Several lines of evidence indicate that the plasma AGT concentration is linked to the blood pressure. A remarkably high correlation between the concentration of the renin substrate and elevated blood pressure (r = 0.39, P < 0.0001) was first reported in a large study involving 574 subjects (128). The plasma AGT was also found to be elevated in hypertensives and the offspring of hypertensive parents compared with normotensives (129). In a study designed to assess genetic determinants of blood pressure by comparing the offspring of parents having high or low blood pressures ("four-corners" approach), Watt et al. (130) found that offspring of hypertensive patients had elevated plasma AGT levels compared to offspring of normotensive parents. Two cases of hypertension associated with hepatic cell tumors producing large amounts of AGT have been reported (131, 132). The above findings suggest that an increase in plasma AGT might elevate blood pressure and contribute to human essential hypertension.

There is evidence that the AGT genotype has a moderate effect on the plasma AGT concentration, as plasma AGT is about 20% higher in subjects (both men and women) carrying the T235 allele (5, 101). The effect associated with M235T appears to be codominant in females (5). Bloem et al. (99) also found that the plasma AGT concentrations of normotensive white American children carrying the 235TT genotype were about 13% higher than those with the 235M genotype. The effect of genotype remained significant after adjustment for body mass index, since the plasma AGT was highly correlated with body mass index. The same authors found that the mean plasma AGT concentration in African-American children was 19% higher than in Caucasians, but the association was not detected because the frequency of M235 was too low to observe an association. Indeed, when splitting the 235T allele with another AGT gene polymorphism in African-Americans, Pratt and co-workers (133) found a significant association with plasma AGT levels. Busjahn et al. (134) recently reported a trend to a codominant effect of the 235T AGT allele on the plasma AGT concentration in twins, although the variation in the concentration was too great for significance. Another recent study on a large sample of the MONICA Augsburg cohort also found a mild codominant and a significant increase in plasma AGT concentration according to the M235T variant (111). Because of the intra- and interassay variations in the plasma AGT measurement, and because the M235T variant only causes a 10–30% increase in plasma AGT, a large number of individuals are probably required to detect this relationship.

F. Association of AGT variants with the renal blood flow response to Ang II infusion
Williams and Hollenberg (135) have shown in a series of studies that nonmodulation is a trait in which renal, vascular, and adrenal zona glomerulosa responsiveness to angiotensin II is not modulated by a high Na⁺ intake. Up to 40% of essential hypertensive patients may have this trait; they have a reduced renal vascular response to angiotensin II when on a high-salt diet. Nonmodulators could, therefore, be a discrete subset of the essential hypertension population. This trait seems to be genetically inherited (136). The same group recently showed that the renal vascular response to Ang II of homozygous patients carrying the 235T genotype was blunted to that of patients with the heterozygous or homozygous M235 genotype (137). Obesity was also found to interact significantly with genotype and enhance the blunting of the renal vascular response. These results are in agreement with those obtained with AGT knock-out mice (see below) and suggest that patients carrying the 235T genotype produce more intrarenal Ang II, thereby contributing to a disturbed renal physiology.

	VII. Molecular Genetics of AGT and Other Diseases

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
A. AGT variants and coronary heart disease
Several studies have evaluated the relative risk of cardiovascular diseases conferred by gene variants of the renin-angiotensin system. Cambien et al. (138) reported in the "Etude Cas-Témoin de l’Infarctus du Myocarde" (ECTIM) an association between variants of the ACE gene and myocardial infarction in Caucasian patients considered at little risk of cardiovascular diseases according to their lipid profile and body mass index. The same case and control populations (630 patients who survived a myocardial infarction and 741 controls) were used to explore the role of the M235T and T174M AGT gene variants in the pathogenesis of coronary disease. Although they found no difference in the AGT genotype distributions in the cases and controls (139), carriers of the M174 allele were more often receiving antihypertensive treatment, and a greater proportion of them required several drugs. Three other studies have shown an association between the M235T AGT allele and coronary heart disease. A limited case-control study (82 cases and 160 controls) by Ishigami et al. (140) found a marginally higher frequency of the 235T allele in Japanese patients with coronary atherosclerosis (at least one coronary artery with more than 25% luminal diameter reduction); Katsuya et al. (141) reported that a group of 422 patients from New Zealand were at greater risk of coronary heart disease than were 406 matched controls: the 235T allele increased the risk of CHD 2- to 6-fold and the risk of myocardial infarction 3- to 4-fold, after adjustment for several other risk factors. Finally, Kamitani et al. (142) found that the 235TT genotype was more frequent in Japanese patients with myocardial infarction than in a control group and that the odds ratio for this genotype was further increased in patients carrying the ACE DD genotype. However, another large study designed to detect an association between variants in the renin-angiotensin system genes and the degree of coronary artery stenosis in Caucasians revealed no correlation with the ACE I/D polymorphism and M235T variants (143).

B. AGT variants and complications of diabetes
Degenerative retinopathy and nephropathy are frequent complications of insulin-dependent diabetes mellitus (IDDM). Several factors, such as the duration of IDDM, blood glucose control, and genetic factors, influence these complications. The genes of the renin-angiotensin system, in particular ACE (144, 145), may be implicated in diabetic nephropathy, since high glomerular concentrations of angiotensin II can favor glomerular hypertension and contribute to renal deterioration.

The relationship between AGT variants and diabetic nephropathy in patients with IDDM has been evaluated recently. A large study on Caucasian IDDM patients with (n = 195 patients) or without (n = 185 patients) nephropathy, who had been diabetic for about 26 yr found no difference in the M235T genotype distributions in the two groups (146). The patients with nephropathy who were homozygous for the 235T allele had a higher systolic blood pressure than those carrying the T235 M or the M235 M genotype. Similarly, another study on 380 patients who had had IDDM for 15–20 yr found that the 235T allele did not appear to increase susceptibility to nephropathy (147), although the 235T allele was more frequent among normotensive and hypertensive patients with microalbuminuria than in patients with normoalbuminuria. Another study on 423 Caucasian patients with type 1 diabetes and 663 with type 2 diabetes was also negative (148).

The absence of any association between the M235T variant and diabetic nephropathy does not definitively exclude a role for the AGT gene in this situation. A recent report found a positive correlation between the 235T variant and diabetic nephropathy in Irish patients with IDDM (149). The contributing effect of the M235T variant could be too small to be detected in most of the studies performed so far. A more recent study on 494 Caucasian IDDM patients found that the severity of renal involvement was associated with the ACE I/D variant (150). There was also a positive interaction between the ACE I/D and AGT M235T variants, suggesting that patients with both the ACE D and AGT 235T alleles were more at risk of developing nephropathy.

C. AGT, obesity, and blood pressure
AGT is quite abundant in adipose tissue, and the local synthesis of AGT could be part of a paracrine system influencing the release of Ang II in the vicinity of the vascular smooth muscle cells. A person’s nutritional status could influence the blood supply to adipose tissue and thus affect vascular resistance and blood pressure in obese individuals. The plasma AGT concentration is in fact strongly correlated with blood pressure during weight loss (151). A study of 67 obese Japanese women showed that those who were hypertensive had a higher intraabdominal/subcutaneous fat index than the normotensives (152). The same study showed a strong significant correlation between this fat index and systolic (r = 0.62) and diastolic (r = 0.63) blood pressures, regardless of age or body mass index. The association between intraabdominal fat accumulation and hypertension raises the question of whether AGT is involved in the pathogenesis of hypertension in obese individuals. The previously mentioned ECTIM study (139) found that the M174T AGT variant was associated with hypertension only in lean individuals. Subjects with body mass index less than 26 kg/m² who carried the M174 allele were 2–4 times more likely to have high blood pressure than homozygotes for the T174 allele, whereas there was no association in subjects with a body mass index greater than 26 kg/m². These results suggest that the production and regulation of AGT in adipose tissue interact with blood pressure.

	VIII. Effects of AGT Gene Variants and Gene Targeting on Ang I Production and Blood Pressure

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
The evidence incriminating AGT in human hypertension remains statistical and raises the question of whether the 235T AGT allele is itself functional or whether it is a marker of another closely linked variant. Other approaches are, therefore, needed to describe the precise influence of AGT on Ang production and blood pressure control.

A. Human studies
Two rare missense mutations of the AGT gene have been described recently. These provide a unique opportunity to study in vitro and in vivo the consequences of these modifications of the AGT protein structure (Fig. 3). A mutation at the site (Leu₁₀-Val₁₁) where AGT is cleaved by renin was recently found in a patient with preeclampsia (153): a phenylalanine residue replaced the leucine residue. This resulted in a 2-fold increase in the catalytic efficiency of the renin acting on the corresponding human mutated AGT produced by transient expression. ACE also had more than a 2-fold greater catalytic efficiency for the angiotensin decapeptide with phenylalanine at residue 10 than for natural Ang I. These kinetic differences may significantly affect the production of Ang II and alter the function of the systemic or local renin systems in this form of pregnancy-induced hypertension. The same mutation has also been found in two other unrelated hypertensive patients screened for variants of AGT because of a severe and familial form of hypertension. The wild type Ang I and the mutated Ang I (Ang I-Phe₁₀) were measured separately by specific RIAs in one affected patient. The basal and captopril-stimulated plasma concentrations of Ang I-Phe₁₀ were elevated, showing increased production of angiotensin from the mutated AGT allele (X. Jeunemaitre, T. Guyene, M. Azizi, J. Menard, and P. Corvol, unpublished).

View larger version (9K): [in this window] [in a new window]   Figure 3. Polymorphisms of the promoter region of the human AGT gene. The three polymorphisms are in strong linkage disequilibrium with M235T. The G-6A substitution has been found to be associated with an increased AGT expression. [Derived from X. Jeunemaitre et al. (5); K. Yanai et al. (86), and I. Inoue (87).]

There is no evidence that the common 235T variant directly affects the function, secretion, or metabolism of AGT, although the replacement of a methionine by a threonine residue is not neutral. Cohen et al. (155) used multiple monoclonal antibodies and two immunometric assays to show that there was an epitopic change that allowed plasma samples homozygous and heterozygous for this allele to be readily discriminated. However, the plasma AGT 235T variant had the same K_m for renin as AGT 235M (T. Guyene, J. Menard, and P. Corvol, unpublished results).

The 235T variant could be a marker for a putative, as yet unknown, molecular variant(s) that directly mediate predisposition to hypertension. Haplotypes have been generated in a large number of hypertensive French and Japanese subjects to resolve this question (110). Both groups had an AGT gene with a G A substitution at position -6 upstream of the initial transcription site, which was found to be in almost complete linkage disequilibrium with the 235T allele (Fig. 4). Tests of promoter activity and DNA binding studies with nuclear proteins from a human hepatoma cell line showed that this nucleotide substitution may be the functional variant, as it affects the basal transcription rate of AGT, which could explain the association of the M235T variant with the plasma AGT concentration (87).

View larger version (7K): [in this window] [in a new window]   Figure 4. Linkage disequilibrium between the M235T polymorphism in exon 2 of the AGT gene and the G(-6)A polymorphism in the promoter region of the gene.

C. Genetic hypertension in rats
Lodwick et al. (156) showed a cosegregation of the rat AGT locus with a specific, modest increase in pulse pressure in F2 rats derived from a cross between spontaneously hypertensive rats (SHR) and normotensive Wistar Kyoto rats (WKY). This locus accounted for approximately 20% of the genetic variance in this phenotype. The sequences of the AGT mRNA and plasma AGT protein in hypertensive and normotensive rats were similar, but the AGT mRNA concentrations of the two strains were different. It has also been reported that the AGT gene locus does not cosegregate with blood pressure in F2 rats derived from a cross between stroke-prone spontaneously hypertensive (SHR-SP) and WKY rats (157) and that there is a missense mutation in the SHR-SP AGT gene (158). Thus, the findings in SHR and SHR-SP appear to differ from those in humans. There are also differences in a number of high blood pressure loci identified in hypertensive rats that are not linked to hypertension in humans, such as the renin gene (93, 159) and the SA locus (160, 161). Clearly, disease-relevant genetic loci are not necessarily the same in man and rats.

D. Antisense AGT oligodeoxynucleotides
The role of AGT in regulating blood pressure in rats has also been recently evaluated using antisense oligodeoxynucleotides targeted at AGT mRNA. Pedrazzini et al. (162) generated a transgenic mouse line carrying an inducible antisense AGT gene, and showed a transient decrease in AGT synthesis in these animals. Injection of liposome-encapsulated antisense AGT oligodeoxynucleotides decrease plasma Ang II and blood pressure in SHR (163, 164).

E. Transgenic animals
The role of AGT in blood pressure regulation can also be explored by creating transgenic animals that overproduce AGT. One difficulty lies in the species specificity of the renin AGT reaction. Rat renin hydrolyzes a Leu-Leu bond in rat AGT, in constrast to the Leu-Val bond in human AGT. Because of this species specificity, mice or rats producing human renin or human AGT do not develop hypertension. Kimura et al. (165) generated transgenic mice that contained the entire rat AGT gene including 1.6 kb of 5'-flanking sequence. These animals developed hypertension, especially the males, and AGT was overproduced in the liver and brain. The synthesis of AGT in the brain seemed to be a prerequisite for the development of a hypertensive phenotype. Other transgenic animals have been generated by Ohkubo et al. (166), who introduced the rat renin gene, the rat AGT gene, or both into mice. The genes were under the control of the mouse metallothionine I gene promoter. A similar chimeric renin-angiotensin system was constructed by cross-mating separate lines of transgenic mice carrying either the human renin or AGT genes (167). Both models showed captopril-sensitive hypertension only in the transgenic mice carrying both transgenes. These experiments show the species specificity of the renin-AGT reaction in vivo confirming the in vitro findings (168). Although these transgenic experiments clearly demonstrate that overexpression of the renin and AGT genes leads to increased blood pressure, their relevance to the pathogenesis of human hypertension is questionable.

An interesting model of hypertension in pregnant mice has been recently reported: Takimoto et al. (169) observed that transgenic female mice producing human AGT developed a transient elevation in blood pressure in late pregnancy when they were mated with transgenic males producing human renin. Blood pressure returned to normal after delivery, renal histology showed glomerulopathy lesions, and the placenta was very abnormal. These abnormalities are also found in human pregnancy-induced hypertension. However, this result was not observed with the reverse combination of female renin transgenic mice and male AGT transgenic mice. This combination produced pups that were hypertensive, but no hypertension developed in the mothers, indicating the role of placental renin in the development of maternal hyper-tension.

F. AGT gene duplication
Finally, a fourth approach has been used by Smithies and co-workers (170, 171) to quantitatively evaluate the possibility that genetically determined elevated AGT concentrations are a predisposing factor for hypertension. The original strategy was to engineer mice having genetically determined high plasma concentrations of AGT, thereby mimicking the increased plasma AGT in familial hypertension. Targeted gene disruption and duplication were used to generate mice having one, two, three, or four copies of the AGT gene. Plasma AGT increased progessively, but not linearly, with the AGT gene copy number. The one-copy animals had plasma AGT concentrations that were about 35% of the normal value, and the three-copy animals had concentrations that were 124% of normal. Blood pressure also increased with the copy number; it was approximately 122 for the one-copy mice, 129 for the two-copy (wild type) mice, and 138 mm Hg for the three-copy animals (171). Renal blood flow varied inversely as the number of AGT copies. These results demonstrate directly that small increases in plasma AGT, of the same order of magnitude as those associated with the human AGT M235T variant, can influence the fine control of renal vascular resistance and blood pressure. These elegant studies also show that varying the number of copies of genes is probably the best way of assessing the roles of genes in a complex quantitative trait, such as high blood pressure.

	IX. Conclusions and Perspectives

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 
AGT has long been considered to be simply a plasma "reservoir" upon which renin acts. The molecular genetic studies reviewed here strongly suggest that the AGT locus is involved in essential hypertension and also in some cases of pregnancy-induced hypertension. Experimental data from transgenic rodents and targeted AGT mice that have allowed the effect of the AGT gene to be quantified also support the human data. The results indicate that a molecular variant of AGT (235T/-6A) is associated with an increased rate of AGT gene transcription, which could result, in turn, in small increases in plasma and tissue AGT concentrations. Because the plasma, and especially the tissue, AGT concentration is rate limiting for Ang II generation, this genetically chronic overstimulation of the renin system favors kidney sodium reabsorption by the kidney, vascular hypertrophy and(or) increased sympathetic nervous system activity and predisposes the carrier to the development of common cardiovascular diseases. This effect may be more marked in the presence of other predisposing genes and/or deleterious environmental factors.

The role of AGT in hypertension and cardiovascular diseases can be summarized as follows:

1. AGT variants appear to have a modest effect on blood pressure in the whole population, which probably explains some of the positive and negative findings with the M235T variant. The percentage of hypertension attributable to AGT is difficult to quantify but does not appear to exceed approximately 10%. The relative risk of cardiovascular disease conferred by this gene is probably not greater than 20–30%.

2. The effect of AGT variants is probably modulated by a variety of interacting genes, such as the ACE gene (109), and environmental factors such as salt intake and obesity. This type of susceptibility allele will only be useful if it results in a great relative risk of developing hypertension. Thus, defining the subset of individuals in whom it may play a more important role, such as pregnancy-induced hypertension, will be one of the important tasks for the near future.

3. Local increases in AGT in tissues involved in blood pressure regulation (such as the kidney and the brain) may well be important for Ang II generation. This will have to be investigated in patients with different AGT genotypes. The small increase in plasma AGT should have only a minor effect on Ang II generation and should be counteracted by the negative feedback of this peptide on renin secretion.

4. The response to antihypertensive agents, especially those that block the renin system, needs to be evaluated in patients classified according to their genotype. A preliminary investigation on a limited number of hypertensive patients classified according to their AGT or ACE phenotypes (172) provided no evidence for such association. Large, prospective studies, with randomization of patients according to their genotype, are needed to obtain meaningful information.

	Acknowledgments

 
The English text was checked by Dr. Owen Parkes.

	Footnotes

 
Address reprint requests to: P. Corvol, M.D., INSERM U36, 3 rue d’Ulm, 75005 Paris, France.

¹ This work was supported by Grants from INSERM, the Collège de France, Bristol-Myers Squibb, the Association Claude Bernard, and the Association Naturalia and Biologia.

	References

Top Abstract I. Introduction II. Structure of the... III. Characteristics of AGT IV. Tissue and Cellular... V. Regulation VI. Molecular Genetics of... VII. Molecular Genetics of... VIII. Effects of AGT... IX. Conclusions and Perspectives References

 

1. Ward R 1990 Familial aggregation and genetic epidemiology of blood pressure. In: Laragh JH, Brenner BM (eds) Hypertension: Pathophysiology, Diagnosis and Management. Raven Press Ltd, New York, pp 81–100
2. Lifton RP, Dluhy RG, Powers M, Rich GM, Cook S, Ulick S, Lalouel J-M 1992 A chimaeric 11ß-hydroxylase/aldosterone synthase gene causes glucocorticoid-remediable aldosteronism and human hypertension. Nature 355:262–265[CrossRef][Medline]
3. Shimkets RA, Warnock DG, Bositis CM, Nelson-Williams C, Hansson JH, Shambelan M, Gill JJ, Ulick S, Milora RV, Findling JW, Canessa C, Rossier BC, Lifton RP 1994 Liddle’s syndrome: heritable human hypertension caused by mutations in the ß subunit of the epithelial sodium channel. Cell 79:407–414[CrossRef][Medline]
4. Mune T, Rogerson FM, Nikkila H, Agarwal AK, White PC 1995 Human hypertension caused by mutations in the kidney isozyme of 11ß-hydroxysteroid dehydrogenase. Nat Genet 10:394–399[CrossRef][Medline]
5. Jeunemaitre X, Soubrier F, Kotelevtsev YV, Lifton RP, Williams CS, Charru A, Hunt SC, Hopkins PN, Williams RR, Lalouel J-M, Corvol P 1992 Molecular basis of human hypertension. Role of angiotensinogen. Cell 71:169–180[CrossRef][Medline]