Na+, kidney, hypertension and genes: lessons from rats.

Abstract

Dahl was the first investigator to use rats to demonstrate that genetic and familial factors are involved in the pressor effect of a high salt diet. He subsequently showed that, in this rat strain (carrying his name), the pressor effect of salt is somehow mediated by the kidney since transplantation of this organ could ‘transplant’ the salt sensitivity [1]. Rapp further developed the Dahl rat strain into inbred and congenic strains, which enabled study of the genetic aspects of the relationship between salt, hypertension and the kidney. These studies led to the identification of gene polymorphisms or quantitative trait loci (QTLs) involved in blood pressure regulation [2]. During the pregenetic era, alterations of tubular Na+ reabsorption, vascular tone, renin–angiotensin–aldosterone system (RAAS) and sympathetic activity have been identified as the most important mechanisms underlying secondary or primary forms of arterial hypertension, both in animals and humans. However, studies on the role of each of these mechanisms have been hampered by the complex network of reciprocal interactions among them. New genetic studies, both in rat models and in humans, have detected gene polymorphisms that affect one or more of these mechanisms, but these gene effects are certainly subjected to the same complex inter-relationships. Traditional ‘epidemiology’ attempts to define the context favouring or inhibiting the ‘pathological’ effect of a given environmental factor. Similarly, ‘genetics’ should define the role of a given gene by performing appropriate empirical observations in the field to measure the environmental or genetic factors (context) ‘modulating’ the effect of that gene. Of course, definition of the context is much easier in animals than in humans. Taken together with the obvious possibility of performing the appropriate experiments, this makes the ‘lessons’ from animal studies indispensable in our understanding of human complexity [3]. However, the crucial question is which type of lessons? Although the evolutionary divergence between rats and humans took place approximately 40 million years ago [4], most of the fundamental mechanisms regulating body sodium, renal function and blood pressure in mammals had already been achieved in rats at the level of sophistication present in humans. Most of hypertension triggering mechanisms work similarly in the two species. For these reasons, observations on rat models, both at pathophysiological and genetic levels, may help to organize current genetic knowledge in an integrated manner [3]. This may provide the theoretical framework needed to propose a hierarchical relationship, if any exists, among the different environmental or genetic factors. For example, it is known that alterations of body sodium, via diet, renal artery constriction or kidney transplantation, involve nearly all four mechanisms indicated above, in spite of the uniqueness of the initial triggering factor. Thus, renal Na+ retention may be increased by a variety of mechanisms, including activation of the sympathetic nervous system [5]. In turn, a primary alteration of kidney function leading to sodium retention may activate the sympathetic nervous system [6]. Similar inter-relationships must also exist among genes affecting these different functions. Therefore, only by studying the appropriate time-dependent changes in animals can this hierarchical relationship be elucidated. Indeed, such an approach has detected a temporal pattern of differential gene expression between salt-sensitive and salt-resistant Dahl rats [7]. By integrating pathophysiological and genetic knowledge, it may be possible to propose a distinction between genetic mechanisms that trigger real initial events and those that ‘modulate’ or modify the effect of the primary one. Why are these studies important? Unfortunately, the current genetic guidelines for defining causation of a single gene in a complex disease such as hypertension are unable to account for genetic or environmental interactions. Attempts to circumvent this deficiency by increasing the stringency of the statistical criteria indicating genetic linkages or associations have so far produced mixed results [8,9]. Biological interactions can only be inferred from empirical observations and not from deductive mathematical logic [9,10]. Considering the relationship between genes, Na+, kidney and hypertension, two important groups of established facts should be considered as a starting point for any discussion. (i) In the three genetic animal models of hypertension, where kidney cross-transplantation has been carried out, hypertension travels with the kidney [1,11,12]. Data consistent with these rat findings have also been collected in humans [13]. (ii) In all the monogenic forms of human hypertension, an alteration in tubular Na+ reabsorption has been found [14] To these facts, two further observations may be added. (i) More than 50 years of pathophysiological studies aimed at elucidating the complex inter-relationship between Na+-kidney and blood pressure regulation have demonstrated that some mechanisms are more important than others in the long-term regulation of blood pressure. The work of Guyton et al. [15] has clearly shown that the renal pressure–natriuresis mechanism must be reset to produce a permanent increase in blood pressure. Such resetting may occur either because of some intrinsic kidney mechanisms (genetic or acquired from spontaneous or experimental renal injury) or an extrinsic one (i.e. mineralocorticoids hormones, sympathetic activation, etc.). (ii) A variety of genetic–molecular mechanisms regulating the constitutive Na+ reabsorption capacity of the tubular cells has been suggested to be involved in genetic forms of hypertension [16]; furthermore, in a recent review [17], the primacy of abnormalities in tubular reabsorption over those of glomerular filtration has been proposed to explain the kidney damage associated with diabetes and the consequent increase in blood pressure. For all these reasons, study of the mechanisms at point (ii), together with those involved in the regulation of renal blood flow and glomerular function, is of paramount importance when distinguishing primary genetic mechanisms working at the kidney level from secondary ones that modulate their effects. However, the opposite may also be true: the pressor effect of the primary genetic polymorphisms affecting mineralocorticoid secretion by the adrenals is certainly modulated by the genetic polymorphisms regulating tubular Na+ transport. By the accomplishment of sequencing of the entire rat genome [18], most of the genes involved in tubular sodium reabsorption along the nephron may become known. Together with our knowledge of the biochemical machinery of tubular cells, this may facilitate the grouping of the gene polymorphisms affecting the constitutive tubular cell capacity, which is crucial for blood pressure regulation. Admittedly, the same variation in tubular function can be achieved through different genetic molecular mechanisms. Work on rats must be paralleled by appropriate validations in humans, either by pathophysiological or genetic studies [3]. Due to the well-established role of epistatic interactions among loci in determining the blood pressure effect of a given gene, the genetic network where the effect of a given gene is studied is of paramount importance. All the rat models used comprise inbred strains obtained by dozens of brother–sister matings. Together with the need to overcome inbreeding degeneration, this procedure certainly creates an artificial genetic network that may amplify or blunt single genes at work in a freely mating population. Moreover, the blood pressure measuring procedure per se (catheter or cuff) may affect the genotype–blood pressure relationship [19]. Finally, when an inbred strain is selected for high or normal blood pressure, we simply pick-up and concentrate in one strain a number of genes that increase or decrease blood pressure levels in the respective genetic networks of the two strains. These peculiar characteristics of genetic rat models may account for the rather modest correspondence between rat and human genetic data. To date, the QTLs suggested to be associated with blood pressure are 179 in rats [2,20,21] (http://rgd.mcw.edu) and 76 in humans [22–26]. Because many of these QTLs are overlapping across the chromosomes, the real number is certainly lower (approximately one-third). Eight relevant chromosomal traits harbouring QTL for blood pressure are syntenic in rats and human (1q, 2q, 5q, 7q, 8q, 9q, 15q, 17q) [20–26]. Both genome-wide linkage analysis and candidate gene studies have allowed the identification of several putative classes of genes implicated in blood pressure regulation among receptors (angiotensin, glucocorticoid, dopamine, adrenergic receptors), structural proteins (adducin, Gβ3 protein), hormone or vasoactive substances (renin, angiotensinogen, angiotensin-converting enzyme, 11β-hydroxylase, aldosterone synthase) and other genes (SA, WNK1, WNK4) [23,27]. For example, adducin polymorphisms have been detected both in rat models (MHS) and in humans. The ‘hypertensive’ gene variants of both species produces similar biochemical, cellular and renal function abnormalities in rats and an appropriate subset of patients [28]. Although an association with blood pressure is demonstrated in rats, mixed results have been obtained in humans when the general population of hypertensives is considered. However, when appropriate genetic interactions or the biological characteristics of the patients are taken into account, more consistent results are obtained across the different studies [9,28]. For these reasons, it is important to define the biochemical, cellular or pathophysiological similarities, or differences, occurring between the two strains of rats and between hypertensive and normotensive human subjects to evaluate to what extent the genetic findings in rats are relevant for a subset of patients with the peculiar biological and pathophysiological characteristics closer to those of rat models [3]. In this issue of the journal, Lighthall et al. [29] describe a variation in the expression of genes involved in NO synthesis at level of the kidney. These findings belong to a promising research line, which, in particular, has been developed by Cowley et al. [30], who rather convincingly demonstrated that NO production in the renal medulla of Dahl rats is crucial for medullary blood flow regulation and overall tubular Na+ reabsorption and related blood pressure changes. In turn, the NO reduction of NaCl reabsorption is counter-regulated by the production of reactive oxygen species (ROS) which increases NaCl reabsorption [31]. These two systems may be selectively modulated within the kidney with appropriate manoeuvres producing consensual changes in tubular reabsorption and blood pressure [30,32,33]. Therefore, it is reasonable to postulate that the biological effect of the polymorphisms of the genes involved in the regulation of these systems may also undergo the same type of crosstalk. This interaction must also include those genes that directly affect the constitutive capacity of tubular cells to transport Na+, such as Na+ channels with their modulatory protein (NEDD4, WNK4, Skf, etc.), adducin or the RAAS system. In turn, NO and ROS renal production or the activity of the RAAS are also affected by variation in body Na+ content. From these and other data, a complex network of feedback loops emerges, interacting reciprocally and collectively regulating blood pressure, body sodium and renal Na+ excretion. This network may ‘blunt’ or ‘magnify’ the changes in blood pressure associated with variations in Na+ intake. Hypertension may develop because either some ‘major gene', or a combination of genes, shifts the set point of this network in the presence of a critical amount of dietary sodium. In view of the long series of mixed results obtained with candidate genes to date [9,34], the latter possibility is more likely. Moreover, besides the genetic network and dietary Na+ [35], other factors, such as gender [36,37], body mass index [38] and age, may affect the effect size of a given gene. However, the claim that a candidate gene must somehow be associated with hypertension across different populations clearly contrasts with the well-established notion that hypertension is a very heterogeneous condition in terms of genetic and pathophysiological mechanisms, response to therapy and organ complications [9,34,39]. We clearly face a dilemma. On the one hand, we need to measure as many genotypes and environmental factors as possible to define the appropriate genetic and environmental background. On the other hand, power considerations limit the number of these factors, even for rather large population sizes. From the practical point of view, it is impossible to estimate the genotype effect or interaction of all the genes involved in these networks. Could pharmacogenomics contribute to the dissection of this complexity? The strategy of selectively blocking one mechanism with a drug to define its role within a complex network has been applied successfully for decades in biology and pathophysiology. The availability of selective RAAS blockers has provided much greater insight regarding the role of RAAS in hypertension and cardiovascular pathophysiology than the measurement of individual RAAS components in various tissues or body fluids. Similarly, in carriers (either humans or rats) of a given genotype, if we are able to find a selective therapeutic advantage of a drug that interferes with the sequence of events triggered by the genotype(s), we may provide an important argument in favour of the role of that genotype(s) over the others, in that particular context. Data along this line have already been collected [39]. In conclusion, the lesson from rats may be very useful in dissecting the genetic complexity of primary hypertension, provided that all the modulator and/or confounding factors at work in the two species are adequately taken into account.