Research Centre-Centre Hospitalier de lUniversité de Montréal (CHUM), Montreal, Quebec, Canada
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
functional genomics; angiotensin II receptor AT1B; aortic hyperplasia; vasoreactivity; congenic strains
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
As for a relationship between hypertension and hyperplasia, the initial evidence indicated that the cultured aortic smooth muscle cells (SMC) proliferated faster (i.e., hyperplasia) in the spontaneously hypertensive rats (SHR) than in the Wistar Kyoto rats (WKY) (21, 41). These findings suggested a link between hypertension and the SMC proliferation. Since then, a large body of evidence has been accumulated on their associations (5, 14, 21, 23, 27, 38). On the basis of these observations, one would expect that a higher SMC proliferation might correlate with a higher BP. Nevertheless, there was one genetic study reporting that hypertension was not associated with the growth of skin fibroblast in a F2 cross between SHR and WKY (20). This is not surprising, because an association between BP and the SMC growth in these two strains could simply be accidental, not causal, one way or the other.
To date, it has remained unclear whether there are quantitative train loci (QTLs) determining vascular SMC number and, if so, whether mechanisms determining hypertension and vascular hyperplasia could be separated. Our current goal is to discover QTLs controlling vascular SMC number regardless of, or unprejudiced by, association or dissociation with BP.
The power and utility of genetic analyses are evident in finding causal genes for vascular hyperplasia. The beauty of the genetic approach lies in its objectivity, i.e., one is not biased toward predicting an outcome based on a preconceived notion linking two associated phenomena, e.g., hypertension and hyperplasia. Nor are we biased toward the perception that hyperplasic alleles of a QTL, if it existed, had to come from a hypertensive strain, not from a normotensive strain. If hypertensive alleles of a QTL for BP could originate from a normotensive strain, and vice versa (2, 31), one could expect that SMC-increasing alleles of a QTL could originate from a normotensive strain.
Ever since the revelation of a QTL for BP in Dahl salt-sensitive (S) rats (9, 13), rat chromosome (Chr) 2 appears to play an important role in the development of hypertension in several of the hypertensive strains (1, 10, 11, 15, 18, 24, 33, 34, 36, 39, 40). In the process, a number of congenic strains have been developed that cover various regions of Chr 2 of S rats (16, 17). Taking advantage of these resources, some questions could be addressed regarding QTLs for hyperplasia. Specifically, 1) is there a QTL for vascular hyperplasia and particularly for aortic smooth muscle cell number (SMCN) in S rats? If so, does it act independently of blood pressure? And 2) if there is a QTL specific for SMCN, can any gene be supported as a candidate?
In the present studies, BPs of certain congenic strains were compared along with their respective indexes of vascular SMCN, vascular, cardiac, and renal weights, and internucleosomal DNA fragmentation, an indicator of apoptosis. In addition, when a candidate gene was found in the region containing the QTL for SMCN, sequencing would be conducted to detect significant mutations that could have the potential to alter the expression or/and function of the gene.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
To prove that BP C2QTL1 and BP C2QTL2 are separate genes, one would need to find a marker between D2Rat166 and Mme that is SS for both C2S.M1 and C2S.M6. For this purpose, we employed the rat genome database. Supercontigs located in the Chr 2 region between D2Rat166 and Mme were first identified by blasting them at the website http://www.ncbi.nlm.nih.gov/genome/seq/RnBlast.html. From such a supercontig, regions containing microsatellites were searched and, when found, were used to design markers for genotyping rats based on PCR. These new markers were designated with D2Chm prefixes, which represent CHUM. Consequently, a series of markers were generated. One particular marker designated D2Chm90 was polymorphic and located in the region of interest (Fig. 1).
BP Studies and Tissue Extractions
BP studies on the congenic strains were reported in detail previously (16, 17). In brief, male rats were weaned at 21 days of age, maintained on a low-salt diet (0.2% NaCl; Harlan Teklad 7034), and then fed a high-salt diet (2% NaCl; Harlan Teklad 94217) starting from 35 days of age until the end of the experiment. Telemetry probes were implanted when rats were 56 days old (i.e., after 3 wk of the high-salt diet), with their body weights between 250 and 320 g. The BP measurements lasted until the time of death.
Rats were killed by decapitation 30 days after the commencement of their BP measurements, i.e., at 14 wk of age. The organs of interest (see below) were removed, cleaned carefully from surrounding adventitial connective tissues and fat, blotted to remove excess blood, and weighed immediately. To minimize any potential inconsistency in the collection, one person was designated for harvesting one particular organ and also for the subsequent dissections into subsections of interest.
The whole heart was then dissected into the left ventricle plus the septum, and into the right ventricle. The weight of each section was then recorded and corrected for the body weight of the respective animal. The dissection was performed according to a previously published procedure (29, 37).
Measurement of Aortic Cross-Sectional Areas
The following procedure was done, essentially according to a previously published protocol (5). The thoracic aorta was isolated from the diaphragm to above the first intercostal artery and cleaned of adherent adventitial tissue and fat. A 3-mm-long ring of aorta was cut between the third and fourth intercostal arteries. The aortic rings were immerse fixed in 4% parformaldehyde overnight and processed according to routine histological procedures for morphological examinations in paraffin-embedded tissues. The fixation of aortas was done in none-pressure-controlled conditions. The aortic endothelium of the remaining aorta segments was removed by scrubbing the intimal surface with a cotton-tipped applicator. The aortic media was immediately snap frozen in liquid nitrogen and kept at 80°C until further processing for DNA extraction.
The medial cross-sectional area was evaluated in 5-µm-thick, hematoxylin-eosine-stained sections of aorta. Photomicrographs of the aortic sections were taken at x400 magnification, digitalized, and analyzed using the public domain NIH Image program 1.61 (developed at the National Institutes of Health and available at http://rsb.info.nih.gov/nih-image/). Two independent measurements of each tissue section were conducted to ensure the reproducibility of the image analysis.
Determination of Vascular SMCN
The following was done essentially according to a procedure previously published (5). Briefly, three consecutive sections (5-µm thickness) were obtained from aortic rings sectioned perpendicularly to the longitudinal axis of the vessel. Tissue sections were stained with hematoxylin, photographed, and printed at a final magnification of x400. In the top section, an area, ad, of vessel wall was delineated by two parallel lines approximately perpendicular to the wall edges. A disector was defined as the three-dimensional probe bounded by ad and the top surface of the "bottom" section. The disector has volume vd = ad x hd, where ad is the area of the disector and hd is the height of the disector. The term hd is obtained by the following equation: hd = (s l) x t, where s is the number of serial sections and t the average section thickness. Within ad, the number of nucleus profiles, nt, was determined, and in the subsequent sections, each of the nt nuclei was followed and marked. In the final section, the number of nuclei still present was determined, nb. The number nd = (nt nb) is then the number of "downward-pointing" nucleus ends within the disector when counting from top to bottom. Binucleate SMC account for <0.5% of the SMC in the thoracic aorta of hypertensive rats (30). Therefore, on the assumption that each cell contains only one nucleus, an estimate of cell numerical density, Nv, is given by Nd = nd/vd. The number of cells per unit vessel length then was estimated from a1 x Nv, where a1 is the medial cross-sectional area determined with the image analyzer as described above.
DNA Content and Apoptosis
The remaining aortic media was pulverized in liquid nitrogen. The total tissue DNA was extracted and quantified as described previously (5). To quantify the degree of internucleosomal DNA fragmentation, a hallmark of apoptosis (programmed cell death), the 3'-OH ends in the extracted DNA were radiolabeled using [32P]dNTP and terminal deoxynucleotidyl transferase. DNA was separated by gel electrophoresis and blotted, and radioactivity incorporated in the small fragments (1501,500 bps) was quantified using a PhosphoImager as described previously (5).
Vasoreactivity Studies
Vasoactive responses to angiotensin II (ANG II) and phenylephrine (PE) were examined as described previously (26) in rings of thoracic aorta isolated from S and C2S.M2 rats implanted or not with a telemetry probe. Briefly, rats were killed by exposure to CO2, and freshly isolated vessels without adventitial adipose and connective tissue or endothelium were cut into 3-mm-long rings to be placed at 37°C in organ chambers filled with oxygenated Krebs solution containing, in mM, 11 dextrose, 117.5 NaCl, 1.18 MgSO4, 1.2 KH2PO4, 25 NaHCO3, 4.7 KCl, and 2.5 CaCl2. Isometric contractions were measured with the use of isometric force transducers (Harvard Apparatus) and a digitalized data acquisition system (Model MP100, Biopac System; Harvard). Rings were equilibrated under a passive tension of 1 g and then challenged with 40 mmol/l KCl to document receptor-independent contractility. Tissues were also stimulated with cumulative concentrations of PE (1100 mmol/l) and ANG II (300 pmol/l to 300 µmol/l). Separate sets of rings were used to measure contractile responses in the presence or absence of nitro-L-arginine methyl ester (L-NAME; 100 µmol/l). The integrity of the endothelium was evaluated on the plateau of PE-induced maximal contraction by measurement of the relaxant response to acetylcholine (Ach; 1 µmol/l). In a subset of tissues, ANG II-induced contractile effect was also evaluated in the presence of losartan (75 nmol/l).
Statistical Analysis
In BP analyses, repeated measures ANOVA, followed by Dunnetts test (which permits a correction for multiple comparisons and sample sizes), was used to compare a parameter between two groups as presented previously (16, 17). During the comparison, ANOVA was first used to analyze the data to see whether there was any difference among the groups. If the difference was significant, then the Dunnetts test followed, to see which group was different and the level of significance compared with the S strain. If the difference was not significant, then no Dunnetts test followed. In functional vasoreactivity studies, data from S and C2S.M2 were compared by unpaired Students t-test using Prism 4.0 (GraphPad Software).
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The most informative strain is C2S.M2 (Fig. 1), which had no effect on BP (Ref. 17 and summarized in Fig. 1, bottom) and had the same cardiac, left ventricular, and renal weights as S (Fig. 2, A and B) but showed higher SMCN (Table 1). In fact, aortic SMC counts of both C2S.M1 and C2S.M2 were higher than that of the S strain (P < 0.039). In contrast, SMC counts of the C2S.M6 and C2S.M strains were not different from that of the S strain (Table 1).
|
|
Figure 2 also shows comparisons of cardiac and renal weights corrected for body weights among the strains. Kidney weights in C2S.M1, C2S.M6, and C2S.M were lower (P < 0.03) than in S rats, whereas those of C2S.M2 were not different from those in S rats (Fig. 2, A and B). Cardiac and left ventricular weights were only significantly lowered in C2S.M and C2S.M6 (P < 0.03) but not in C2S.M1 strains (P > 0.10). Table 2 shows assessments of DNA content and DNA fragmentation indexes in both cardiac ventricles and aortas.
|
Angiotensin Receptor AT1B (Agtr1b) Gene as a Candidate for the SMCN QTL
In previous pharmacological studies, we observed that AT1 and not AT2 receptors were involved in the SMC growth (5, 38), implying that AT1 receptors might be important in mediating the SMC proliferation. There are two types of AT1 receptors, AT1A and AT1B, encoded by Agtr1a and Agtr1b genes, respectively. In our congenic strains, the Agtr1a gene on Chr 17 (8) was the same as in S (10, 16, 17), but the Agtr1b on Chr 2 (Fig. 1) was different in that it contained alleles from the Milan normotensive strain (MNS; Fig. 1). Because it is located in the chromosome region harboring the SMCN QTL (Fig. 1), Agtr1b could be conceived as a candidate gene for the SMCN QTL. We reasoned that to qualify Agtr1b to be the candidate gene, it had to contain significant mutations that could potentially result in differences either in its function or in its level of expression between S and congenic strains trapping the SMCN QTL.
Mutational evidence supporting Agtr1b as a candidate gene of the SMCN QTL.
Indeed, when Agtr1b genes were cloned and sequenced, several nucleotide mutations were detected in the coding region. Two of these mutations were of special interest in that they were found in MNS, C2S.M, C2S.M1, and C2S.M2 but were different from those in C2S.M6 and S, which had the same nucleotides. These two nucleotides resulted in significant changes in amino acids. One T-to-C change at nucleotide position 6 caused a change of (aliphatic) isoleucine to (hydroxyl) threonine at amino acid position 2. Another A-to-G change at nucleotide position 118 resulted in a significant change of (sulfuric) methionine to (aliphatic) valine at amino acid position 20. The nucleotide positions were numbered from the sequence coding for the first amino acid, as given previously (12). These mutations support functional implications of Agtr1b as the candidate for the SMCN QTL (Fig. 1), since C2S.M, C2S.M1, and C2S.M2 possess MNS alleles at Agtr1b (i.e., C at position 6 and G at position 118), whereas C2S.M6 possesses S alleles at Agtr1b (i.e., T at position 6 and A at position 118).
Functional evidence supporting Agtr1b as a candidate gene of the SMCN QTL.
If the mutations found in Agtr1b were truly significant, one would expect to observe functional differences between the MNS Agtr1b and S Agtr1b genes.
For the subsequent functional study, C2S.M2 was chosen because it carries a smaller chromosome interval of MNS than C2S.M1 (Fig. 1). In our work, contractile responses to ANG II were documented in rings of aorta isolated from C2S.M2 and S rats. The analysis indicated that contractile responses were not different, whether a telemetry probe had been implanted (n = 6/strain) or not (n = 67/strain), and the data were therefore pooled for each strain. As shown in Table 3, contractile responses to ANG II were significantly reduced in C2S.M2 rats compared with S rats, without a significant change in apparent receptor affinity. The contractions induced by ANG II were AT1 receptor dependent, since they were significantly inhibited (>90%) in the presence of losartan (data not shown). In contrast, there was no interstrain difference in the contractile responses to phenylephrine (1-adrenergic agonist) or KCl (receptor-independent contractile stimulus) (Table 3). Similar conclusions were reached with tissues challenged in the presence or absence of the nitric oxide synthase inhibitor L-NAME (Table 3). Thus the hyperplasic aorta from C2S.M2 rats exhibited a selective decrease in ANG II-induced AT1 receptor-dependent contractility compared with S rats.
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
A QTL for the Vascular SMCN Functions Independently of QTLs for Hypertension
C2S.M1 had a lower BP than S (Fig. 1 and Ref. 17) but had more aortic SMCN (Table 1). In contrast, C2S.M2 had the same BP as S (Fig. 1 and Ref. 17) but showed SMCN similar to those of C2S.M1 (Table 1). This fact clearly indicates that these genetic determinants for BP and aortic SMCN act independently of each other. On examination of the chromosome interval on the map, the QTL for the aortic SMCN should be present in the chromosome fragment shared between C2S.M1 and C2S.M2, which is between Prlr and D2Rat303 markers (Fig. 1). In contrast, BP C2QTL1 exists in the segment not shared between C2S.M1 and C2S.M2, i.e., between the D2Rat303 and Mme markers (Fig. 1). It is noted that the QTL for the vascular SMCN acts independently also of cardiac, left ventricular, and renal hypertrophy, because C2S.M2 had the same cardiac, left ventricular, and renal masses as S (Fig. 2, A and B).
Using two congenic strains, a stringent genetic testing, we found that aortic SMCN and BP have different genetic, and possibly physiological, bases. Although MNS rats per se were not directly analyzed for SMCN, C2S.M6 served effectively as a negative control, because it contained a small segment from the MNS strain not including the region delineated by Prlr and D2Rat303 markers (Fig. 1). Because C2S.M6 has no hyperplastic effect, whereas C2S.M1 and C2S.M2 do (Table 1), the effects to increase SMCN in C2S.M1 and C2S.M2 have to be attributed to the QTL residing in the Prlr/D2Rat303 segment, not to any MNS genetic background. The MNS rat would not be a good control, because it differs greatly from the S genome on the rest of the chromosomes (19, 25) and because it gives no information on the effects of genome changes during the congenic construction (7).
Because SMCNs were slightly different (Table 1) between C2S.M1 (170 ± 11) and C2S.M2 (134 ± 21), a possibility cannot be ruled out that there could be another QTL for SMCN in the region of BP C2QTL1 (Fig. 1). In that case, the same QTL could be responsible for both BP and SMCN.
It is noteworthy that the alleles to increase SMCN come from MNS, which is normotensive. Nonetheless, it is not surprising that MNS QTL alleles cause SMCN to increase, whereas S QTL alleles decreased SMCN. In an independent study, Lewis rats, a normotensive strain, actually carry BP-increasing alleles compared with S at a QTL on Chr 3, although the overall BP of the Lewis strain is considerably lower than that of the hypertensive S strain (31). This fact indicated that the BP-increasing QTL alleles in Lewis were hidden and were only unveiled by isolating the region harboring it from the rest of the Lewis genome by congenic strains. Therefore, our present and other studies further demonstrated the power and necessity of genetic dissections of individual QTLs using congenic strains.
The detection of BP C2QTL1 added further proof supporting the power of the congenic approach. In linkage analysis, there was no indication that a QTL existed in the region around C2QTL1 (9, 12). It is only after the isolation of a segment by use of a congenic strain (C2S.M1) that BP C2QTL1 was found (17). This phenomenon is due, probably, to epistatic gene-gene interactions (6, 7) among other things and indicates that the effect of the genetic background impacted by other QTLs affecting the same phenotype may need to be removed for the manifestation of a QTL for SMCN (Fig. 1).
It is also of note that the C2S.M strain included the chromosome segment contained in C2S.M1 and C2S.M2 (Fig. 1), and yet, C2S.M did show hyperplasia (Table 1). This phenomenon can be explained by the following. There could be another QTL or a modifying gene present in the segment in C2S.M that is lacking in C2S.M1, i.e., between D2Rat166 and Adh, that would have a hypoplasic effect. When combined, the overall effect would be hypoplasic. There was an example of this epistatic interaction between two QTLs. When BP-lowering alleles at one QTL were experimentally put together with BP-raising alleles at another QTL, the combined effect was the same as for the BP-lowering alleles on Chr 3 (31). This fact indicates that these QTLs do not act alone, but in a hierarchical relationship with each other in controlling the overall BP. The same principle could be applied to the genetic determinants of SMCN.
As to the mechanisms for increasing SMCN, there are two possibilities. The QTL could either favor the SMC proliferation or reduce the SMC death. The present data showing an inverse relationship between DNA fragmentation and SMCN (Table 2) suggest a possible role for reduced apoptosis. Defining the relative contribution of these cellular mechanisms, however, is worthy of further investigations.
Agtr1b is a Candidate Gene for an SMCN QTL
It is noted that Agtr1b is present in the region containing the QTL for SMCN (Fig. 1). Interestingly, we detected two nucleotide differences leading to significant changes in amino acids in congenic strains possessing MNS Agtr1b alleles (i.e., C2S.M, C2S.M1, and C2S.M2), and that possessing the S Agtr1b allele (i.e., C2S.M6; see RESULTS). The presence of these mutations in Agtr1b corresponding to the location of the QTL provides a possible mechanism for increased SMCN, because the pro-growth and anti-apoptotic effects of AT1 receptors on SMC are well known (4).
The mutations found in Agtr1b could have functional roles in controlling SMC proliferation. As an initial investigation of this issue, we chose to measure contractile responses to ANG II, since this is a proximal marker of receptor function. Surprisingly, we observed that responses to ANG II were reduced in C2S.M2 vs. S rats. We speculate that a prior AT1 receptor activation in vivo may have resulted in receptor desensitization in the aortic tissue isolated from C2S.M2 rats. Such a negative effect of in vivo receptor activation on the vascular sensitivity ex vivo seems reminiscent of hypertensive human and rat arteries, where an enhanced local expression of endothelin resulted in a selective attenuation of contractile responses to endothelin in tissues stimulated ex vivo (35). Alternatively, the altered responses to ANG II may be an epiphenomenon. In vitro studies with cultured SMCs are needed to better define the functional correlates of the mutations in Agtr1b and its significance for the SMC proliferation.
Another way to validate whether Agtr1b would remain a good candidate gene or not for the SMCN QTL is by fine congenic mapping. If Agtr1b is the SMCN QTL, no matter how small the region harboring the QTL is minimized to (e.g., 100200 kb), Agtr1b should always be included in the same region, and this congenic substrain should also have an effect on SMCN. Alternatively, Agtr1b could simply be a marker indicating the approximate location of the QTL for SMCN. The inducible form of the nitric oxide synthase (Nos2) is a case in point. Nos2 was located in a broad region containing a QTL for BP initially, but a fine mapping ruled it out as a QTL because a congenic strain harboring it did not show a BP effect (32). Furthermore, fine congenic mapping will also test the possibility that there are multiple QTLs for SMCN present in the interval between Prlr and D2Rat303.
In summary, a major finding in the present work is that there is a QTL(s) for vascular SMCN that acts independently of hypertension. This discovery has laid the foundation for the identification of the QTL and will lead to revelations of its underlying physiological mechanisms regulating vascular remodeling independently of BP. Agtr1b is a candidate gene worthy of further investigation, using fine mapping and in vitro functional assays, to determine whether it is the SMCN QTL. By use of the same strategy of fine congenic mapping and comparative congenic mapping of BP QTLs, the QTL(s) for SMCN can and will be localized to a smaller interval for candidate cloning (i.e., Agtr1b) or, if not, then positional cloning.
![]() |
GRANTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
FOOTNOTES |
---|
Address for reprint requests and other correspondence: A. Deng, Research Centre-Centre Hospitalier de lUniversité de Montréal (CHUM), 7-132 Pavillon Jeanne Mance, 3840, rue St. Urbain, Montréal, Québec, H2W 1T8, Canada (e-mail: alan.deng{at}umontreal.ca).
10.1152/physiolgenomics.00063.2004.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Visit Other APS Journals Online |