The first evidence of a genetic origin of cerebrovascular accidents was reported 30 years ago (1). Further evidence in favour of a hereditary contribution to this condition has accumulated during the last three decades. In particular, a familiar aggregation of stroke (2), as well as a higher incidence of stroke among monozygotic rather than dizygotic twins (3), has been reported. In addition, the existence of a few monogenetic disorders associated with stroke has been assessed. Finally, an animal model for spontaneous stroke (stroke-prone spontaneously hypertensive rat, SHRsp) was made available to the scientific community several years ago (4) and its pathophysiological characterization has been an attractive topic of research over the last few decades. In particular, the stroke predisposition of this strain was found to share several similarities with the human disease, including common risk factors such as hypertension and dietary components. In fact, a high salt dietary content exerts a causative role, whereas high potassium intake is a protective factor towards stroke in both rats and humans.
Thus, the robust amount of data provided by the epidemiological and basic research approaches had already introduced the concept of stroke as a complex, multifactorial and polygenic trait, comparable to all major cardiovascular diseases. However, only recently have the consistent advances of molecular biology and genetic techniques, along with the improvement of statistical approaches, allowed the identification of the genetic defects underlying monogenic syndromes of stroke as well as the first genetic dissection of stroke in the available animal model.
ROLE OF AN ANIMAL MODEL IN THE GENETIC DISSECTION OF STROKE AND ITS CONTRIBUTION TO UNDERSTANDING THE HUMAN DISEASE
A typical genetic linkage analysis approach was applied to F2 segregating populations obtained either from the stroke-prone SHR/stroke-resistant SHR mating or from the stroke-prone SHR/Wistar-Kyoto (WKY) mating. In the first study, occurrence of cerebrovascular accidents was the differentiating trait between two otherwise closely related strains, whereas in the second study the normotensive versus the hypertensive state was also taken into account. The aim of linkage studies in general is to identify whether any part of the diseased genome co- segregates with the occurrence of a specific phenotype within a familiar pedigree. This procedure requires extensive work on combined phenotyping and genotyping of each family component. With regard to stroke, the genome-wide scanning of the F2 segregating rat cohorts was performed through the use of a panel of randomly distributed polymorphic markers, whereas the chosen quantitative phenotype was either the stroke-proneness under a stroke-permissive dietary regimen (Japanesestyle diet, containing high sodium, low potassium, low protein content) (5), or the stroke sensitivity after middle cerebral artery occlusion (MCAO) (6). These studies led to the identification of genes directly involved in the predisposition to stroke of the SHRsp strain, in a blood pressure-independent manner. In particular, the linkage approach allowed the identification of different chromosomal areas containing genes for the two distinct stroke phenotypes of the stroke-prone rat. In fact, three chromosomal linkage areas (quantitative trait loci, QTL) were identified for stroke-proneness, explaining altogether one-third of the overall phenotype variance (5), and one linkage area was found for the stroke sensitivity after MCAO, accounting for 67% of the phenotype variance (6) (Table 1).
Thus, both studies confirmed what has long been suggested by the epidemiological and experimental evidence and they opened the way to the precise identification of the genetic factors contributing to stroke in the animal model and, hopefully, also in humans.
The possibility of extending information obtained from experimental models to humans, although still controversial, remains the most attractive goal of current cardiovascular genetic research and underlies all current efforts to identify genes responsible for cardiovascular traits such as stroke in suitable animal models. In fact, there is no doubt that the high degree of genetic homogeneity, the easy control of experimental conditions, and the optimal size of the animal populations created in the laboratory represent excellent advantages compared to the direct approach in humans. Moreover, significant support for the existence of a parallelism between animal models and humans has already been provided by a series of studies. Firstly, the genes encoding 11 beta-hydroxylase and aldosterone synthase, identified as positional candidates for salt-sensitive hypertension in the Dahl S rat, were also found to play a role in some forms of human hypertension. Secondly, mutations of the gene encoding alpha-adducin, a cytoskeleton protein, were found to be significantly related to hypertension both in the Milan rat and in the salt-susceptible type of human hypertension. Finally, a linkage area for hypertension in the rat on chromosome 10, also containing the angiotensin converting enzyme (ACE) gene, was found to be synthenic to a region of human chromosome 17 that was also linked to familial essential hypertension.
IDENTIFICATION OF THE GENE ENCODING ATRIAL NATRIURETIC PEPTIDE AS A DIRECT GENETIC DETERMINANT OF STROKE IN RATS AND IN HUMANS
With regard to the investigation of the genetic basis of cerebrovascular accidents, we were able to add one more piece of evidence in favour of the important parallelism between animal models and humans. In fact, based on the result of the linkage study (5), we focused our attention on a positional candidate, i.e. the gene encoding atrial natriuretic peptide (ANP), an important regulator of cardiovascular functions (7). This peptide exerts natriuretic, diuretic and vasorelaxant properties and it is also expressed in the cerebral tissue. The gene encoding ANP was found to map at the peak of linkage of a stroke QTL identified on rat chromosome 5 in the stroke-prone model. Thus, this gene fulfilled all the criteria necessary to be considered an optimal candidate for the pathogenesis of cerebrovascular lesions in the SHRsp, within the context of our experimental conditions. In fact, through an extensive comparative analysis of its structure, regulation and function, we were able to identify several remarkable differences between the stroke-prone and the stroke-resistant rats (Table 2). Two structural mutations were found, one within the promoter region and one within the coding sequence, and they both resulted in functional consequences that could all be related to the pathogenesis of cerebrovascular lesions in the strokeprone model (8). In fact, a significantly different degree of ANP promoter activity was observed in the presence of a mutant regulatory enhancer site of the stroke-prone rat, and a different intra- and extracellular processing of the proANP precursor was observed depending on the presence of a Gly to Ser transposition in the SHRsp strain. The latter effect was associated with the presence of an extra peptide of 6.5 Kd and with greater levels of cyclic guanosine monophosphate (cGMP) production in the diseased strain (8). Furthermore, a marked difference in ANP expression between the two strains was found only in the brain (the target organ for stroke), but not in the heart, where most of the ANP synthesis takes place (8). Finally, an impaired vasorelaxant effect of ANP was shown in cerebral and peripheral vessels from SHRsp and not from stroke-resistant SHR, further suggesting the possible involvement of this gene in the predisposition to cerebrovascular lesions of the stroke-prone model.
Most importantly, when we attempted to identify the possible role of the gene encoding ANP in human stroke, through analysis of the largest cohort of cases and controls studied so far and collected in North America, the Physicians' Health Study, we were able to confirm a significant contribution of this gene to the human disease as well (9). Of note, its involvement appeared to be completely independent of all conventional risk factors for stroke, such as hypertension, diabetes and obesity. While this evidence awaits further confirmation, it certainly holds great promise for our future investigation of the genetic basis of cerebrovascular accidents. In addition, it should be noted that our findings supporting a strokerelated effect of ANP in the animal model and in humans agree with recent evidence showing ANP as a pro-apoptotic agent on both endothelial and myocardial cells.
FROM STROKE QUANTITATIVE TRAIT LOCI TO SPECIFIC GENES: STRATEGIES OF POSITIONAL CLONING
Based on the above discussion, it appears that the discovery of QTLs represents only the first step towards the positional cloning of the gene(s) responsible for the phenotype of interest. Analysis of candidate genes mapping within QTLs, such as our investigation of the ANP gene for one of the stroke QTLs identified in the strokeprone/ stroke-resistant SHR F2 intercross, is one of the possible follow-up strategies. As documented by our own experience, this approach certainly provides a major contribution in supporting or excluding potential candidates through the demonstration or the exclusion of the existence of gene abnormalities, respectively. However, parallel additional approaches are usually needed in order to reduce the time necessary for final identification of the gene hidden inside the chromosomal region of linkage. In particular, once a QTL has been identified, it becomes mandatory to start a more precise refinement of the chromosomal area of linkage through the generation of congenic strains. This process allows the isolation of the desired QTL from the donor strain (either the diseased or the control strain) into the genetic background of the recipient opposite strain. The resultant genetic manipulation will prove or disprove a specific pathogenetic role of the gene located inside the QTL. In addition, in parallel to the generation of congenic lines, a fine saturated map of the region of interest is necessary in order to progressively narrow the area containing the disease-related gene and, thus, in order to generate congenic sublines carrying smaller and smaller pieces of the original QTL and still showing the disease phenotype. Only the smallest region, of approximately 1 cM, can be first subjected to physical mapping with the help of artificial genomic libraries, and then to a search for the expressed sequences contained within the region. The latter need to be compared among diseased and non-diseased individuals, i.e. stroke-prone and stroke-resistant rats.
Thus, based on the results obtained by the two existing linkage studies, the generation of congenic rat strains carrying QTLs for both stroke proneness and stroke sensitivity is currently in progress. Preliminary observations, obtained at an intermediate number of generations aimed at the introgression of a chromosome 1/QTL for stroke proneness into a stroke-resistant genetic background, support a pathogenetic role for the gene contained within this area (10). In fact, these congenic rats show a higher frequency of cerebrovascular events, compared to the stroke-resistant parental strain, together with a marked impairment of growth as it is constantly observed in the stroke-prone parental rat. Therefore, the latter evidence justifies the laborious and time-consuming work needed to establish the congenic lines first and then to carry out the positional cloning of the gene.
When at least some of the genetic factors responsible for common forms of stroke are assessed, through the extensive genetic work currently ongoing in animals and in humans, the identification of individuals carrying either variant or wild-type allelic configurations for specific 'risk' genes will become an important tool in the early prevention of cerebrovascular accidents in predisposed subjects. Furthermore, through our understanding of the fine mechanisms dependent on gene mutations and underlying the disease pathogenesis, we will be able to design targeted, specific therapeutic approaches to reduce the risks of stroke.
Acknowledgement: This work was supported by an IRCCS grant to M.V.; the Italian Telethon grant n.E825 and by the Eurhypgen Concerted Action of the European Community.
1. Marshall J. Familial incidence of cerebrovascular disease. J Med Genet
2. Welin L, Svardsudd K, Wilhelmsen L, Larsson B, Tibblin G. Analysis of risk factors for stroke
in a cohort of men born in 1913. N Engl J Med
3. Brass LM, Isaacsohn JL, Merikangas KR, Robinette CD. A study of twins and stroke
4. Okamoto K, Yamori Y, Nagaoka A. Establishment of the strokeprone spontaneously hypertensive rat (SHR). Circ Res
1974;33/34 (suppl I):I43-53.
5. Rubattu S, Volpe M, Kreutz R, Ganten U, Ganten D, Lindpaintner K. Chromosomal mapping of quantitative trait loci contributing to stroke
in a rat model of complex human disease. Nat Genet
6. Jeffs B, Clark JS, Anderson NH, et al. Sensitivity to cerebral ischemic insult in a rat model of stroke
is determined by a single genetic
locus. Nat Genet
7. Levin ER, Gardner DG, Samson WK. Natriuretic peptides. N Engl J Med
8. Rubattu S, Lee MA, De Paolis P, et al. Altered structure, regulation and function of the gene encoding atrial natriuretic peptide
in the stroke
-prone spontaneously hypertensive rat. Circ Res
9. Rubattu S, Ridker PM, Stampfer JM, Volpe M, Hennekens CH, Lindpaintner K. The gene encoding atrial natriuretic peptide
and the risk of human stroke
10. Rubattu S, Ganten U, Volpe M, Lindpaintner K. Increased incidence of stroke
in animals carrying the SHRSP-derived stroke
-related locus, STR1. Circulation