Independent Genetic Susceptibility to Cardiac Hypertrophy in Inherited Hypertension
Abstract—Cardiac hypertrophy is a common but not inevitable complication of hypertension. Variation in heart size in hypertensives may reflect independent genetic susceptibility to cardiac hypertrophy. Using an experimental genetic model, we determined the location of quantitative trait loci responsible for cardiac hypertrophy and/or hypertension. We studied 182 F2 male animals derived from a cross of the spontaneously hypertensive rat and normotensive Donryu rats. Direct mean arterial pressure (MAP) and left ventricular (LV) mass were measured at 20 weeks of age, and DNA was obtained for linkage analysis. The estimated heritability of MAP was 62% and for LV mass expressed per unit of body weight (relative LV mass) was 76%. We used 185 polymorphic markers, with an average intermarker distance of 12.3 centimorgans for a genome-wide scan in a representative subgroup of 46 animals to identify preliminary quantitative trait loci, which were then mapped in all 182 male F2 rats. Two loci showed logarithm of the odds scores of >4.0. One on chromosome 2, Lvm-1, was linked to relative LV mass but showed no evidence of linkage to MAP. Another locus on chromosome 1, Map-1, was linked to MAP. In the same region, a locus Lvm-2 was linked with relative LV mass. These data indicate the existence of a genetic locus on chromosome 2 of the spontaneously hypertensive rat that affects relative LV mass independently of blood pressure.
Hypertension is a significant risk factor for cardiovascular morbidity and mortality, including myocardial infarction, stroke, and cardiac hypertrophy. Cardiac hypertrophy, in particular LV hypertrophy, is itself a significant risk factor for cardiovascular disease.1 Like hypertension, cardiac hypertrophy has been shown to have a familial and presumably genetic basis.2
High blood pressure contributes to cardiac hypertrophy3 through hemodynamic load, but not all hypertensive patients develop cardiac hypertrophy,4 5 and in a substantial proportion of the population, LV hypertrophy has no obvious cardiovascular cause.6 Furthermore, certain growth factors such as angiotensin II may modulate LV mass independently of hemodynamic effects.3
Breeding studies using animal models have suggested that blood pressure and heart size may be determined by separate genetic factors. In the SHR, LV hypertrophy is present from an early age, when blood pressure is within the “normotensive” range.7 For the same level of blood pressure, adult SHR also have greater LV hypertrophy than other hypertensive rat models.8 In addition, certain treatments that reduce blood pressure to the same degree have quite different effects on SHR heart size.9
Crosses of hypertensive and normotensive strains have also suggested that blood pressure may not be the only determinant of heart size in the SHR. In a cross between SHR and the normotensive control strain DRY, it has been shown that the blood pressure of F1 animals are in the midparental range, but relative LV mass is closer to that of the DRY parent.10 Tanase et al11 showed that for different strains there is no correlation between blood pressure and LV mass. Pravenec et al12 demonstrated using recombinant inbred strains that there are possibly different genetic loci contributing to the blood pressure and LV mass.
Molecular genetics has been used to investigate phenotypically complex diseases and traits.13 14 Several genes have already been implicated in the regulation of blood pressure. The recent development of a genetic map for the rat15 offers the opportunity to examine the genetics of complex phenotypes in the rat and to identify QTLs. These approaches are powerful and have located several QTLs controlling individual phenotypes such as blood pressure13 14 or independent QTLs that influence associated phenotypes such as stroke16 and renal failure17 in hypertensive strains.
The aim of this study was to identify QTLs that influence heart size in the SHR. We used a cross of SHR and DRY in a genetic analysis of 182 male F2 rats using 185 markers from the rat genetic map.
Derivation of Experimental F2 Population
We studied 182 male F2 rats from an intercross of the SHR and DRY strains. These strains were chosen because of relatively large phenotypic and genotypic differences that provide the contrast required for informative linkage analysis. The SHR were derived from National Institutes of Health stock and maintained in an inbred colony for the last 18 years. The normotensive DRY strain was originally obtained in 1989 from Dr Tanase (Sankyo Co, Ltd) and was in its 64th generation at the time of mating for this study. The inbred status of both parental strains was confirmed by DNA minisatellite fingerprinting and by allozyme analysis.18 Male SHR were mated with female DRY to produce F1 rats. Eight F1 males and eight F1 females were intercrossed to produce an F2 population consisting of 182 males. All animals were fed standard laboratory rat chow (Norco Rat and Mouse Cubes) and had ad libitum access to drinking water. A 12-hour light/dark regimen was maintained throughout. All experimental protocols were approved by the Austin Hospital Animal Ethics Committee.
Determination of Phenotypes
Arterial pressure was measured by the direct intra-arterial method in conscious animals at the age of 20 weeks when adult levels of blood pressure are achieved. Body weight was measured, and the rats were anesthetized briefly (typically <15 minutes) with methohexitone (Boehringer, 50 mg/kg IP) for insertion of polyethylene catheters (PE-50) into the left carotid artery. The catheters were exteriorized in the interscapular region, and the rats were allowed to recover in individual cages overnight with ad libitum access to food and water. The following morning (between 9:00 am and 11:00 am) rats remained in their own cages, and a blood pressure transducer (model DPT-3003-S, Peter Von Berg, Munich, Germany) was attached to the arterial catheter. Transducer signal was preamplified through a Grass model 7C preamplifier before analog-digital signal conversion (Analog Digital Instruments) for storage and off-line analysis of data. Once the animals were resting quietly and the blood pressure had stabilized, MAP was recorded for 30 minutes with continuous sampling. The 30-minute average of these readings was used to estimate MAP.
Rats were killed with overdose of pentobarbitone (Boehringer, 100 mg/kg IP). The body length of each rat was obtained by lying the rat straight on its back and measuring to the nearest millimeter from the tip of the nose to the base of the tail. Ventricular mass was determined by removing the whole heart, excising the atria, and dissecting the RV wall from the left ventricle and interventricular septum. Ventricles were blotted dry of blood before weighing to the nearest milligram, and relative LV mass was calculated by dividing the LV mass by body weight. Relative RV mass was calculated in a similar manner. The weight of both left ventricles and right ventricles was also divided by BL2 to provide a correction in relation to body surface area.
DNA Extraction, Purification, and Amplification
Testes, spleen, and liver were collected for DNA analysis. DNA was extracted by standard methods.18 Approximately 200 mg of tissue was added to 5 mL of lysis solution (0.1 mol/L Tris-HCl, pH 8.5, 0.005 mol/L EDTA, 0.5% SDS, 0.2 mol/L NaCl). Proteinase K was added to a final concentration of 300 μg/mL. This tissue sample was incubated at 37°C with shaking overnight. The DNA was subsequently phenol extracted and then ethanol precipitated before recovery by centrifugation at 2000g. The resulting DNA pellet was resuspended in 200 μL of water. For genotyping, the DNA was diluted to 5 ng/μL.
The DNA samples from the F2 progeny of the SHR×DRY intercross were genotyped by PCR amplification with radiolabeled forward (5′) primer (using T4 kinase) and an unlabeled reverse (3′) primer. Primer concentration for PCR was 100 nmol/L and was used to amplify 20 ng of template DNA in a 10-μL reaction. Amplification conditions were as follows: initial denaturation at 92°C for 3 minutes, followed by 27 cycles of 92°C for 1 minute, 55°C for 2 minutes, and 72°C for 3 minutes. PCR products were visualized by running on 6% polyacrylamide gels at 85 W per gel for 3 hours, wrapping the gels in cellophane wrap, and exposing directly to x-ray film.
The genetic linkage map for this cross was constructed by genotyping SHR and DRY parents with 643 markers, of which 362 (or 56%) were polymorphic. On the basis of existing maps, we selected 185 markers to provide a genomic map with an expected average intermarker distance of approximately 10 cM. Of the 185 markers, 184 were SSLPs and the remaining marker was an RFLP.18
A subgroup of 46 animals, the number of samples able to be run on a single polyacrylamide gel, was chosen for the initial genome-wide scan. The particular animals were selected on the basis of calculated expected LOD scores to provide a representative sample covering the entire range for both relative LV mass and MAP phenotypes.19 In brief, rats were sampled from the upper and lower values from the distribution of MAP and relative LV mass. This approach maximized genetic contrast and potential linkage information.19 The complete genome-wide study of this group allowed us to build a genetic linkage map for this cross and to provide preliminary QTLs for the traits MAP and relative LV mass with greater efficiency than scanning all markers in all F2 rats. The specificity and sensitivity of the approach has been validated previously.19
The map was constructed from the linkage data obtained using MAPMAKER.20 Any errors flagged were checked, and double recombinants were confirmed. The final order was tested using the “RIPPLE” command, which alters local order and checks for integrity of the linkage group. The order of markers was also compared with the latest genetic map (Reference 1515 and unpublished observations [H. J. Jacob, 1997]).
After the map was constructed, QTLs affecting phenotypes were mapped relative to genotypes using the MAPMAKER/QTL software package.20 21 Briefly, the program calculates the most likely phenotypic effect having genotypes SHR/SHR, SHR/DRY, or DRY/DRY at a putative QTL and then calculates a LOD score reflecting the strength of evidence for the existence of the QTL and the proportion of the phenotypic variance explained. On the basis of these preliminary linkage analyses, we genotyped the entire F2 population with markers flanking the putative QTLs. LOD scores of more than 4.0 were accepted as indicating significant linkage.22 We estimated genetic variance in the F2 cross by subtracting environment/error variance derived from the F0 and F1 generations from the total variance (data not shown). This method was used to calculate the percentage of genetic variance accounted for by particular QTLs.
Data are expressed as mean±SEM unless stated otherwise. Statistical analysis of the effects of genotypes on phenotypes was performed using multivariate ANOVA. MAP and relative LV mass were entered as dependent variables and the genotypes as factors in these analyses. These analyses allowed for testing interaction (epistasis) between QTLs. Where no higher-order interaction was detected, the effects of QTLs on MAP or relative LV mass were assessed by univariate ANOVA. To allow for covariance between the phenotypes, in particular possible effects of MAP on relative LV mass, we included the Roy-Bargman Step-down F test in the multivariate analysis. This procedure adjusted relative LV mass for covariance with MAP before testing for effects of particular loci. In addition, at one locus (Lvm-2), we assessed the possible independent effect of genotype on relative LV mass by obtaining the studentized residuals of MAP regressed on Lvm-2, which were then regressed on relative LV mass. The resulting studentized residuals of this regression were used in ANOVA to examine the effects of the Lvm-2 locus. All statistical analyses were undertaken using SPSS version 6.1.23 24
Significant differences were observed for MAP and relative LV mass between male SHR, DRY, and F1 rats (Table 1⇓). There were no significant differences in body weight between SHR and DRY at 20 weeks of age. From the observed variances, we estimated that the degree of genetic determination for MAP was 61.7%, and for relative LV mass it was 75.9%.
Coverage of the Genetic Linkage Map
Table 2⇓ shows the genomic coverage of the polymorphic markers. The selected genetic markers gave a genome coverage, on average, of 1 marker every 12.3 cM, ranging between 1.1 cM and 45 cM. It should be noted that the maximum distance between any locus and a marker is half of the maximum distance between markers. Approximately 13% of markers were separated by more than 25 cM. The majority of these markers were on the X chromosome and in regions where markers are as yet unavailable for this particular cross.
Lvm-1, a Locus Responsible for Relative LV Mass Independently of MAP
We found significant evidence of a locus on chromosome 2 (LOD 4.3) located around D2Mgh15 with a 2.0 LOD support interval of approximately 25 cM (Fig 1⇓). This locus accounted for 22.4% of the total variation in relative LV mass and 29.5% of the genetic variance. Fig 1⇓, in addition to showing the LOD plot for LV mass, reveals that MAP is not linked to this locus on chromosome 2 in our cross.
The average value for relative LV mass for rats homozygous for the SHR alleles at D2Mgh15 was approximately 7% greater than rats homozygous for the DRY alleles (Table 3⇓; multivariate Pillais statistic=.085; approximate F=3.36; P=.01; univariate test for relative LV mass, F=6.76; P=.002). The relative LV mass of heterozygous rats was similar to that of rats homozygous for the DRY alleles, suggesting that the mode of inheritance of increased LV mass in relation to this locus is recessive. There were no significant differences in MAP between rats grouped according to genotypes at the D2Mgh15 locus (Table 2⇑; univariate test for MAP, F=0.90; P=.408). We refer to this locus as Lvm-1 for LV mass-1.
The three groups defined by genotypes at the D2Mgh15 locus showed significant differences in body weight, with lowest values in those homozygous for the SHR alleles (Table 3⇑). The same tendency was seen for BL2, but these differences were not significant. The significance of these differences in terms of inheritance is uncertain because these phenotypes do not differ between SHR and DRY parents. Despite having the lowest body weight, the highest average (although not statistically significant) raw LV mass was seen in F2 rats homozygous for the SHR D2Mgh15 allele (Table 3⇑). When raw LV mass was corrected for its relationship with body weight in ANOVA, there remained a significant relationship (P=.01) between the D2Mgh15 genotypes and LV mass. LV mass expressed per BL2 was greatest (Table 3⇑, P=.06) in rats homozygous for the SHR D2Mgh15 allele. No significant differences were observed between the D2Mgh15 genotypes for relative RV mass or RV mass expressed per BL2 (Table 3⇑).
Loci Affecting Relative LV Mass and MAP
We found another locus associated with both relative LV mass and MAP. This QTL was linked to relative LV mass with a LOD score of 4.6 near D1Mit3 (Fig 2⇓). This locus, designated Lvm-2, accounts for 18.6% of the total variance of relative LV mass and 24.5% of the genetic variation. Rats homozygous for SHR at this locus had relative LV masses of approximately 8% greater than those homozygous for the DRY alleles (Table 4⇓; multivariate Pillais statistic=.172; approximate F=7.12; P<.0001; univariate test for relative LV mass, F=8.87; P<.0001). The relative LV mass of heterozygous rats was similar to that of rats homozygous for the SHR alleles, suggesting a dominant mode of inheritance for increased relative LV mass.
However, MAP was also linked to chromosome 1 (Table 4⇑; univariate test for MAP, F=8.41; P<.0001), making it difficult to determine whether relative LV mass and MAP are under separate genetic control. The Roy-Bargman Step-down F test indicated that a significant effect of Lvm-2 on relative LV mass remained after adjustment for covariation in MAP (F=6.23, P=.003). The multiple regression technique (see “Methods”) also revealed evidence for a significant independent effect of Lvm-2 on relative LV mass (F=5.82, P=.004).
For MAP, there appear to be two distinct loci, each with a LOD score >4.0 on chromosome 1 (Fig 2⇑). However, when we adjusted for one locus, the other disappears, suggesting a single locus responsible for MAP. We have designated this locus Map-1. This locus accounts for 21.2% of the total variance and 34.4% of the genetic variance in MAP. Rats homozygous for the SHR alleles have MAP approximately 10 mm Hg higher (P=.004) than those homozygous for the DRY alleles (Table 4⇑). The MAP of rats heterozygous at this locus was slightly higher than the midparental value, suggesting that the mode of inheritance for increased blood pressure is additive. No differences were observed between the D1Mit3 genotype groups for body weight, BL2, relative RV mass, or RV mass expressed per BL2 (Table 4⇑).
Developmental and pharmacological studies of SHR7 8 9 have revealed a dissociation between blood pressure and LV hypertrophy. Breeding experiments10 11 12 also suggest that there may be different determinants of heart size and blood pressure in mature rats.
Using a genome-wide search, we found evidence of a QTL designated Lvm-1 on chromosome 2, which influences relative LV mass and not blood pressure in adult animals. At this locus, the SHR genotype is associated with increased relative LV mass in a recessive pattern, such that F2 animals homozygous for the SHR allele for Lvm-1 have significantly greater relative LV mass than rats homozygous or heterozygous for the DRY allele. The results of linkage mapping and multivariate statistical analysis show no significant effect of the Lvm-1 QTL on MAP.
The phenotype relative LV mass was selected because (1) it accounts for individual differences in body size that may confound comparisons of raw LV mass, (2) it is a phenotype that shows a significant difference between the parental strains and a high degree of genetic determination, and (3) the association with increased risk of cardiovascular disease in epidemiological studies has been based on measurements of LV mass that are corrected for body size.1
In our F2 population, rats that were homozygous for the SHR allele at Lvm-1 were approximately 20 g lighter than those homozygous for the DRY allele. This difference in body weight contributes mathematically to the difference in relative LV mass. The biological reason for differences in body weight are not evident. It seems unlikely that this locus determines body size because no such differences exist between parental SHR and DRY. Therefore, we conclude that for a given body weight, the SHR allele at the Lvm-1 locus is associated with increased size of the left ventricle and that this difference is not associated with blood pressure. The Lvm-1 QTL accounts for about one quarter of the genetic variance in relative LV mass in this cross.
Several authors have reported in other crosses that loci on chromosome 2 affect blood pressure in SHR12 and SHR-SP.25 Although the exact physical relationship between these loci and the Lvm-1 locus is unknown, they do not appear to be related. Although the Lvm-1 locus was not associated with blood pressure differences in this cross, the genetic background may have negated such a phenotypic effect. In the DRY×SHR cross, we have reported an association between blood pressure and the NGF locus on chromosome 2.18 However, Lvm-1 is a considerable distance from the NGF gene and, as we have reported previously,18 the NGF locus shows no association with variation in LV mass. The preliminary genome scan confirmed the absence of linkage between NGF and relative LV mass locus, and although it suggested linkage with MAP, we did not perform a thorough multipoint analysis around the NGF gene.
We also found a region on chromosome 1 that influences both blood pressure (Map-1) and relative LV mass (Lvm-2). The coincidence of inheritance of traits at this locus may have several explanations. It may indicate the effects of a single gene in this region that influences blood pressure, and relative LV mass follows as a physiological effect, ie, cardiac hypertrophy resulting from hypertension. Alternatively, one gene may influence both traits directly, or there may be two genes in this region that influence blood pressure and relative LV mass independently. Our statistical analyses to adjust for the effects of MAP on relative LV mass and for the effects of Lvm-2 on MAP provided results that were consistent with a significant independent effect of Lvm-2 on relative LV mass. In addition, this locus seems to be associated with differences in RV mass, which is independent of systemic arterial pressure.
A recent study16 of SHR-SP found that the inheritance of stroke was linked to a QTL on chromosome 1 designated STR1. This locus maps to the same region as Lvm-2 and is also centered around D1Mit3. The effects of STR1 were also independent of blood pressure. Stroke in the SHR-SP is associated with structural abnormalities in the vasculature.16 26 It is possible that genes exerting effects on cardiovascular growth and structure reside in this region of chromosome 1 and may affect the functional structure of the heart and vessels. A recent study of the inheritance of LV mass in recombinant inbred animals suggested a locus on chromosome 17 affected heart size such that the SHR allele was associated with low heart weight.12
Our findings will lead to a more focused search for mutation in candidate genes that might explain the results of our linkage analysis. However, it should also be possible, using existing markers, to compare the cardiac cellular physiology and structure in F2 animals with contrasting genetic predisposition to cardiac hypertrophy. Furthermore, it should also be possible to selectively breed animals that improve genetic and physiological understanding of cardiac hypertrophy.27
Selected Abbreviations and Acronyms
|BL2||=||body length squared|
|LOD||=||logarithm of the odds|
|Lvm-1||=||left ventricular mass-1 locus|
|Lvm-2||=||left ventricular mass-2 locus|
|MAP||=||mean arterial pressure|
|Map-1||=||mean arterial pressure-1 locus|
|NGF||=||nerve growth factor|
|PCR||=||polymerase chain reaction|
|QTL||=||quantitative trait locus|
|RFLP||=||restriction length fragment polymorphism|
|SHR||=||spontaneously hypertensive rat(s)|
|SHR-SP||=||stroke-prone spontaneously hypertensive rat(s)|
|SSLP||=||simple sequence length polymorphism|
Dr Innes is supported in part by a grant from the National Heart, Lung, and Blood Institute (1U10HL154508). Dr Jacob was supported in part by grants from the National Heart, Lung, and Blood Institute (1U10HL154508), National Institute of Diabetes, Digestive and Kidney Diseases (5RO1DK46612), National Center for Research Resources (5RO1RR08888), and a sponsored research program by Bristol-Myers Squibb. Dr McLaughlin was supported in part by the Stanley M. Sarnoff Foundation. This research is supported in part by the National Health and Medical Research Council of Australia. We thank Dr George Koike for assistance with the MAPMAKER/QTL computer package.
- Received June 30, 1997.
- Revision received August 8, 1997.
- Accepted October 15, 1997.
Kannel WB, Gordon T, Castelli WP, Margolis JR. Electrocardiographic left ventricular hypertrophy and risk of coronary heart disease: the Framingham study. Ann Intern Med. 1970;72:813–822.
Harrap SB, Dominiczak AF, Fraser R, Lever AF, Morton JJ, Foy CJ, Watt GC. Plasma angiotensin II, predisposition to hypertension, left ventricular size in healthy young adults. Circulation. 1996;93:1148–1154.
Tingleff J, Munch M, Jakobsen TJ, Torp-Pedersen C, Olsen ME, Jensen KH, Jorgensen T, Kirchoff M. Prevalence of left ventricular hypertrophy in a hypertensive population Eur Heart J.. 1996;17:143–149.
Kannel WB, Dannenberg AL, Levy D. Population implications of electrocardiographic left ventricular hypertrophy. Am J Cardiol. 1987;6D:85I–93I.
Folkow B. The structural cardiovascular factor in primary hypertension: pressure dependence and genetic reinforcement. J Hypertens. 1986;4(suppl 3):S51–S56.
Folkow B. Physiological aspects of primary hypertension. Physiol Rev. 1982;62:347–504.
Harrap SB, Mitchell GA, Norman TL. Is cardiovascular hypertrophy in young SHR the result of a recessive gene? In: Sassard J, ed. Genetic Hypertension Colloque. Montrouge, France: INSERM/John Libbey Eurotext Ltd; 1992;218:341–343.
Tanase H, Yamori Y, Hansen CT, Lovenberg W. Heart size in inbred strains of rats, part 1. Genetic determination of development of cardiovascular enlargement in rats. Hypertension. 1982;4:864–872.
Pravenec M, Gauguier D, Schott J-J, Buard J, Kren V, Bila V, Szpirer C, Szpirer J, Wang J-M, Huang H, St Lezin E, Spence MA, Flodman P, Printz M, Lathrop GM, Vergnaud G, Kurtz TW. Mapping of quantitative trait loci for blood pressure and cardiac mass in the rat by genome scanning of recombinant inbred strains. J Clin Invest. 1995;96:1973–1978.
Hilbert P, Lindpaintner K, Beckmann JS, Serikawa T, Soubrier F, Dubay C, Cartwright P, De Gouyon B, Julier C, Takahasi S, Vincent M, Ganten D, Georges M, Lathrop GM. Chromosomal mapping of two genetic loci associated with blood pressure regulation in hereditary hypertensive rats. Nature. 1991;353:521–529.
Jacob HJ, Brown DM, Bunker RK, Daly MJ, Dzau VJ, Goodman A, Koike G, Kren V, Kurtz T, Lernmark A, Levan G, Mao Y-P, Petterson A, Pravenec M, Simon JS, Szpirer C, Troillet MR, Weiner ES, Lander ES. Genetic linkage map of the laboratory rat, Rattus norvegicus. Nat Genet. 1995;9:63–69.
McLaughlin MG, Jacob HJ. A targeted genotyping strategy for complex diseases and multiple phenotypes. Genetics. In press.
Lander ES, Botstein D. Mapping Mendelian factors underlying quantitative traits using RFLP analysis. Genetics. 1989;121:185–189.
SPSS Base System User’s Guide. Chicago, Ill: MJ Norusis/SPSS, Inc; 1990:244–301.
SPSS Advanced Statistics User’s Guide. Chicago, Ill: MJ Norusis/SPSS, Inc; 1990:70–119.
Clark JS, Jeffs B, Davidson AO, Lee WK, Anderson NH, Bihoreau MT, Brosnan MJ, Devlin AM, Kelman AW, Lindpaintner K, Dominiczak AF. Quantitative trait loci in genetically hypertensive rats: possible sex specificity. Hypertension. 1996;28:898–906.
Dominiczak AF, Devlin AM, Lee WK, Anderson NH, Bohr DF, Reid JL. Vascular smooth muscle polyploidy and cardiac hypertrophy in genetic hypertension. Hypertension. 1996;27:752–759.
Harrap SB, Jones EF, Datodi SR, Norman TL. The development of a new line of rats with genetic cardiac hypertrophy. J Hypertens. 1994;12(suppl 3):S166. Abstract.