David Geffen School of Medicine at UCLA, Los Angeles, California 90095
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
quantitative trait loci; genome scan
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Twin studies, population studies and epidemiological studies have long suggested etiologic roles for genetic factors in the pathogenesis of essential hypertension. Although recent progress has been made in identifying rare mutations that cause Mendelian forms of altered BP control in affected families (15, 16), unfortunately, the genetic components of the common form of essential hypertension have proved less tractable. Encouragingly, several novel human hypertension loci have recently been identified. These include HYT1 [Online Mendelian Inheritance in Man (OMIM) ID 603918] on human chromosome 17q2122 (3, 12), HYT2 (OMIM 604329) on human chromosome 15q (33, 34), and HYT3 (OMIM 607329) on human chromosome 2p2524 (2, 36). However, the genetic complexity of human populations and the variable environmental influences on human essential hypertension imposed significant experimental challenges in identifying the causative gene(s) of human essential hypertension. Development of suitable animal models of these human essential hypertension loci may provide the necessary tools to help define the nature of the genetic alterations underlying human essential hypertension.
Capitalizing on the well-developed genetic and genomic tools available for laboratory strains of inbred mice and on the ability to vigorously control diet and environmental conditions, we have developed and characterized a new F2 intercross model of differential BP regulation in mice. Whole genome quantitative trait loci (QTL) mapping studies of these mice have led us to identify a set of novel genetic loci that are significantly involved in BP regulation in young adult mice fed a normal diet. Interestingly, one of these mouse BP controlling loci is located in a genomic region homologous to the human genomic interval containing HYT1, a major confirmed human hypertension locus (3, 12).
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Phenotyping.
Systolic BPs were measured using a validated tail-cuff BP measuring system (model 229; IITC, Woodland Hills, CA). Tail-cuff BP measurement in trained mice has been shown to be reproducible and to correlate well with mean arterial pressures in unrestrained, unanesthetized mice (13). A set of training and measurement protocols were developed and strictly followed to ensure reliable and reproducible tail-cuff BP measurements.
In preliminary studies, we compared the tail-cuff BP of A/J and B6 mice measured at 31°C (the minimal temperature necessary to establish a constant blood flow through the mouse tail) and at 38°C. Our results indicated that the BPs of A/J mice at 31°C and 38°C were 103 ± 6 and 100 ± 10 mmHg, respectively (not significant). In contrast, the BPs of B6 mice at 31°C and 38°C were 134 ± 10 and 115 ± 10 mmHg, respectively (P < 0.02). These results demonstrate that the BP of B6 mice is significantly dependent on the animals core temperature. To minimize the effect of heat stress, mice were warmed in their own uncovered cage within a constant 31°C temperature chamber for a period of 1525 min before and during the BP measurement. To minimize the diurnal variation in BP measurements, all mice were kept under a strict 6 AM/6 PM, 12:12-h light/dark cycle, and all BP measurements were acquired between 1 PM and 5 PM. To minimize age-related variations, BPs were obtained from mice that were between 10 and 14 wk of age. To minimize operator-dependent measurement errors, each mouses BPs were measured on two separate days by two different individuals blinded to each others results. When these two separate measurements did not agree within <5 mmHg, additional independent measurements were obtained.
With suitably trained and cooperative mice, successive BP readings were reproducible to within 2 mmHg. Of all the mice studied, the BPs of 28 mice still did not agree within 7 mmHg after 4 or 5 additional measurements. Data from these mice were not used. All animal protocols were approved by UCLAs Chancellors Animal Research Committee.
Genotyping.
Genomic DNA were prepared using a standard proteinase-K-SDS digestion, ethanol precipitation procedure. DNA were adjusted to 25 µg/ml in 10 mM Tris, pH 8.0. All genotyping were carried out at the National Heart, Lung, and Blood Institute (NHLBI)-sponsored Mammalian Genotyping Service at the Marshfield Clinic. Genotyping was carried out with a set of 111 microsatellite markers selected from the MIT mouse genome map (6). These markers span the mouse genome at an average distance of 15 cM. The markers used were D1Mit211, 303, 49, 217, 102, 36, 165, 362; D2Mit81, 296, 192, 75, 274, 285, 266; D3Mit151, 212, 215, 45; D4Mit149, 1, 111, 166, 203, 251; D5Mit267, 113, 10, 240, 136, 30, 223; D6Mit86, 223, 188, 149, 15; D7Mit74, 228, 232, 37, 329, 44, 259; D8Mit64, 205, 249, 112, 245; D9Mit206, 207, 196, 182, 82; D10Mit123, 183, 198, 42, 150, 14; D11Mit227, 231, 242, 212, 124, 180, 214; D12Mit182, 153, 172, 158, 194, 134; D13Mit16, 198, 139, 202, 213, 78; D14Mit44, 141, 203, 194, 165, D15Mit175, 85, 63, 107, 35; D16Mit132, 154, 4, 71; D17Mit164, 52, 180, 93, 155; D17Mit164, 52, 180, 93, 155; D18Mit19, 60, 152, 162, D19Mit68, 40, 91, 71 and DXMit124, 141, 119, 79, and 135.
Statistical analyses.
QTL mapping was carried out using MapManager QTX version b17. ANOVA and other statistical analyses were carried using Statview 5.0 and GraphPad Prism 3.0.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Blood pressure in (A/J x B6)F1 mice.
At 10 wk of age, the BPs of male and female (A/J x B6)F1 are 103 ± 10 and 108 ± 5 mmHg, respectively. These results showed that the BP of (A/J x B6)F1 mice resembled A/J more than B6 mice.
Blood pressure in (A/J x B6)F2 mice.
The systolic BPs of mice in an n = 763 (A/J x B6)F2 intercross is shown in Fig. 1 and summarized in Table 2. We found that the BP phenotypes in the (A/J x B6)F2 intercross can differ by as much as 70 mmHg among the F2 progeny mice. This variability in the systolic BP phenotype is consistent with the notion that BP is a polygenic trait. The polygenic nature of systolic BP phenotype in (A/J x B6)F2 cross is also supported by the fact that the F2 BPs did not segregated into A/J-like and B6-like bimodal distributions. Finally, we observed that the BP variation is similar between male and female F2 progenies. These results were confirmed in a second n = 758 (A/J x B6)F2 intercrosses (Table 2 and Fig. 2).
|
|
|
For the first cross consisting of 763 (A/J x B6)F2s, 107 mice with BP >127 mmHg and 118 mice with BP <103 mmHg were genotyped with a set of 110 markers spanning the mouse genome at 15-cM intervals. Whole genome QTL analyses of the genotype data using the MapManager QTX program (17, 18) identified one significant BP QTL on MMU7 (LOD = 3.9) and at least four suggestive BP QTLs (2.3 < LOD < 3.3) on MMU1 (LOD = 2.7), MMU4 (LOD = 2.4), MMU9 (LOD = 2.7), and MMU11 (LOD = 3.2).
For the second cross consisting of 758 (A/J x B6)F2 mice, 126 mice with BP >128 mmHg and 114 mice with BP <106 mmHg were genotyped with the same set of 110 markers. The genotype data from this second set of samples were combined with the genotype data from the first set and analyzed with MapManager QTX. By increasing the number of sample genotyped from 225 to 465, the statistical power of our sample was increased. As a result, the significant QTL on MMU7 was independently confirmed with an LOD score of 5.4. The suggestive QTLs on MMU1, MMU4, and MMU11 were also confirmed with highly significant LOD scores of 6.8, 9.8, and 6.3, respectively. However, the suggestive QTL on MMU9 failed to be confirmed. We named the BP QTLs on MMU1, MMU4, MMU7, and MMU11 as Abbp1, Abbp2, Abbp3, and Abbp4. The LOD score plots of these BP loci are shown in Fig. 3.
|
Using the interaction feature of MapManager QTX and the recommended 1.0e-5 threshold for detecting interacting loci, we detected a pair of interacting BP QTLs involving D7Mit228 (Abbp3) and D17Mit180 with a likelihood ratio statistic (LRS) of 38.3. By itself D7Mit228 had an LRS of 16.7, and by itself D17Mit180 had an LRS score of 4.3. Permutation analyses of the data indicated that 95% genome-wide significant threshold for a pair of interacting QTLs is achieved at LRS = 37.
No interacting loci were detected for Abbp1, Abbp2, and Abbp4, indicating that these three BP loci act independently to exert their effect on BP and their individual effects are not subjected to epistatic genetic interactions from other loci. These findings are important for future attempts to physically refine these loci to identify the underlying genes. This is because phenotypes of primary acting loci will continue to breed true, whereas the phenotypes of epistatic loci may be lost upon segregating the epistatistically interacting alleles into separate congenic and subcongenic strains.
Phenotypic effects of Abbps.
The contributions to the BP phenotype of the A/J and the B6 alleles of each Abbp were determined by the ANOVA statistics test using the Statview 5.0 Program. From the results illustrated in Fig. 4, it can be seen that homozygosity for the B6 alleles of Abbp1, Abbp2, and Abbp4 are associated with an increase of 712 mmHg in BP relative to homozygosity for the A/J alleles of these loci. In contrast, homozygosity for the B6 alleles of Abbp3 is associated with a decrease of 15 mmHg in BP relative to homozygosity for the A/J alleles of these loci. The fact that most of the B6 alleles of the Abbps are associated with increases in the BP phenotypes is consistent with the observation that B6 mice have higher BP than A/J mice.
|
Using comparative genomics tools from both the Mouse Genome Database (MGD) web site (http://www.informatics.jax.org), from the Rat Genomic Database web site (http://rgd.mcw.edu), and from the Human/Mouse homology web site (http://www.ncbi.nlm.nih.gov/Homology/), we have determined the most likely rat (RNO) and human (HSA) genomic region that is homologous to the 2025 cM genomic interval containing the peak LOD score for each of the Abbps. These results are tabulated in Table 3.
|
Genomic intervals for the Abbps.
Lander and Botstein (14) proposed that the region within 1 LOD score unit of the maximum LOD can be used to represent the 96.8% confidence interval for strong QTL. Using this criteria, the LOD score plot for Abbp1 in Fig. 3 indicated that the peak LOD score is located at 35 cM on MMU1 and that the 1 LOD confidence interval for Abbp1 spanned 25 cM from 23 cM to 48 cM on MMU1. By placing Aox1, a marker gene mapped to 23 cM on MMU1 and Ncl, a marker gene located at 48 cM on MMU1 from the MGD mouse genetic map onto the February 2002 release of the public mouse genome sequence (http://genome.cse.ucsc.edu/), this 25-cM interval is estimated to be
28 Mbp in length and contains
120 known genes. Similarly, the 1 LOD confidence interval for Abbp2 spanned
12 cM on MMU4, is
15 Mbp long, and contains
55 known genes. The 1 LOD confidence interval for Abbp3 spanned 20 cM on MMU7, is
40 Mbp in length, and contains
190 known genes. The 1 LOD confidence interval for Abbp4 spanned 10 cM on MMU11, is
14 Mbp in length, and contains
200 known genes.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Numerous transgenic and gene-targeted mice have demonstrated the utility of the laboratory mouse in studying the role of defined single gene alterations on the BP phenotype (27). However, compared with the established role of the rat in studies of polygenic BP controls, only a few polygenic mouse models of differential BP controls have been reported. Sugiyama et al. (30) detected two significant BP loci on MMU15 and MMU7 in a (BALB/cJ x CBA/CaJ) F2 intercross involving 207 males. The two parental strains studied had BPs differing by only 8 mmHg. In contrast, BPs in F2 mice ranged from 80 mmHg to 124 mmHg, a difference of 44 mmHg. This study illustrated that genetic loci with substantial effects on normal BP homeostasis can be uncovered by studying progenies from inbred mouse strains with similar BPs.
In a salt-induced hypertension model involving 250 male backcross offspring from F1 mice produced from the salt-sensitive B6 and the non-salt-sensitive A/J inbred mouse strain, Sugiyama et al. (29) detected six significant salt-induced hypertension loci. Five of these loci are concordant with hypertension loci in rats, and four of these loci were concordant with hypertension loci in humans.
Using an eight-way cross design involving the inbred LP/J, SJL/J, BALB/cJ, C57BL/6J, 129/J, CBA/J, RF/J, and BDP/J mouse strains and nearly 50 generations of inbreeding, Schlager and Sides (25) developed and characterized several inbred mouse strains with either spontaneously elevated or decreased systolic BPs. Genetic mapping studies using these unique hypertensive mouse strains identified major significant systolic BP loci on MMU10 and on MMU13 in addition to suggestive systolic BP loci on MMU2, MMU6, MMU8, and MMU18 (32).
To develop a predominantly genetic model of differential BP regulation in mice, we have carried out a series of initial exploratory and subsequent confirmatory genetic mapping studies of BP-regulating loci in a F2 population produced from intercrossing F1 mice from mating the A/J and B6 inbred mouse strains. In our studies, all mice were kept under standard conditions and fed a standard diet containing 6% fat. In addition, all mice were phenotyped between 1014 wk of age. Thus the variations in the BP phenotype in these F2 mice are primarily due to differences in the genomic composition they inherited from their A/J or B6 grandparents. As a result, the BP loci detected in these studies reflect loci that are associated with genetic differences between the A/J and B6 mouse strains with environmental influences experimentally controlled to a minimum.
Blood pressure is a complex trait that is determined by the combined influences of genetic and environmental factors. It is not unreasonable to speculate that the phenotypes of BP loci detected in genetic studies can be modulated by environmental factors. Thus it will be important to determine how or whether the likely candidate-environmental factors such as diet and age will interact with these Abbp loci to modulate BP. However, it is interesting to note that Abbp1 on MMU1 and Abbp2 on MMU4 mapped to the same 2025 cM genomic intervals as the recently reported salt-induced hypertension loci Bpq1 and Bpq3, respectively (29). Bpq1 and Bpq3 were detected in a male only B6.A/J F1 x B6 backcross design in which the animals were fed a high-salt diet (29).
Abbp4 on MMU11 is of particular interest because this genomic interval is syntenic to confirmed BP loci in human (3, 12) and rat (9, 11). Inspection of the Abbp4 genomic interval in the draft mouse genome sequence (31) revealed the following candidate genes which could potentially play important roles in BP modulation. These include Slc4a1, anion exchanger 1 (AE1) (1); CA-RP X, carbonic anhydrase related protein 10 (21); Cacnb1, ß1-subunit of a L-type voltage-dependent calcium channel (23); Psa, puromycin-sensitive aminopeptidase (10); and Igfbp4, insulin-like growth factor binding protein 4 (35).
Whole genome quantitative trait mapping studies produce statistical evidences for the presence of and for the approximate genomic location of genetic factors associated with the trait under investigation (4). To physically confirm these statistical results and to provide physical specimens to carry out additional genetic studies to refine the mapped loci, the creation of specific congenic mouse strains carrying the respective genomic interval are required (22). The construction of a panel of consomic mouse strains in which each of the 20 chromosomes from the parental inbred B6 mice was replaced with the corresponding chromosome from the A/J mice has been described (20). We have obtained and established local breeding colonies of B6.A/J strains consomic for A/J chromosomes 1, 4, 7, 11, and 17. Preliminary tail-cuff BP measurements confirmed the BP lowering effects of the donor A/J chromosomes 1, 4, and 11 (data not shown). These consomic mouse strains will allow us to investigate whether diet and age will have any effects on the BP phenotype associated with replacing each of the Abbp1-, Abbp2-, or Abbp4-containing B6 chromosomes with the corresponding chromosome from A/J. In addition, these consomic mouse strains will serve as the ideal starting strains for refining the genomic intervals of Abbp1, Abbp2, and Abbp4 to less than 1 cM by serial backcrossing to inbred B6 mice. With the available and soon to be completed genomic sequences, single nucleotide polymorphism (SNP) resources for both B6 and for A/J mice and with whole genome expression profiling tools, it should be feasible to identify the genes and to elucidate the nature of the allelic differences responsible for the 715 mmHg differential BP associated with the B6 vs. the A/J alleles of the Abbps identified in this study.
![]() |
GRANTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
FOOTNOTES |
---|
Address for reprint requests and other correspondence: D. D. L. Woo, Division of Nephrology, David Geffen School of Medicine at UCLA, 7-155 Factor Bldg, 10833 Le Conte Ave., Los Angeles, CA 90095 (E-mail: dwoo{at}mednet.ucla.edu).
10.1152/physiolgenomics.00027.2003.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|