Genetic Modifier Loci Linked to Intima Formation Induced by Low Flow in the Mouse Carotid
Objective— Previously we found dramatic strain-dependent differences in a low flow model of vascular remodeling. Specifically, intima formation in the left common carotid artery was ≈30-fold greater in SJL compared to C3HeB/Fe (C3H/F) mice. We hypothesized that a few genes control intima formation in response to low flow. A C3H/F and SJL backcross resulted in broad range of N2 intima phenotypes.
Methods and Results— Using genome-wide scan we identified two highly significant quantitative trait loci (QTLs) for intima, Im1 (intima modifier 1 locus) on chromosome 2 (Chr2; 77.6 cM, LOD=6.4), and Im2 on Chr11 (17 cM, LOD=5.3). One significant QTL Im3 was found on Chr18 (6 cM, LOD=3.0), and two suggestive QTLs (LOD=1.5 and 1.8) were identified on Chr7 and Chr17, respectively. Interestingly, the intima/media ratio trait mapped to the same QTLs as the intima trait. Haplotype mapping predicted 40 candidate genes. Six of these genes contained SNPs that differed between C3H/F and SJL.
Conclusions— We have successfully mapped 3 QTLs (Im1, Im2, and Im3) that are associated with carotid intima formation in response to low blood flow. These results may be important in identifying genes that influence carotid intima-media thickening and predict cardiovascular disease in humans.
Subclinical atherosclerosis is measured by intima-media thickening (IMT) in human carotid arteries, and carotid IMT is associated with increased cardiovascular risk.1 Mechanisms responsible for formation and progression of carotid IMT are unknown. Importantly, about 40% of the variability in the carotid IMT depends on family history in humans.2 A quantitative trait locus (QTL) has been identified for internal carotid IMT (LOD=3.4) on human chromosome (chr) 12.3 Intima has been proposed to be the “soil” for atherosclerosis, implying an important pathophysiologic role for this tissue.4 Only one genetic analysis of intima hyperplasia in animals, the aortic response to balloon injury in an intercross between BN and SHR rats, has been published.5 The authors identified several QTLs with the most significant intima hyperplasia QTL mapping to rat chr3. It is important to note that intima hyperplasia and atherosclerosis may be regulated by distinct genetic mechanisms in individual vascular beds as shown by a recent study in mouse that compared carotid versus aorta.6 These results suggest that carotid IMT is a complex trait likely determined by multiple environmental and genetic factors. Nonetheless, it has been proposed that insights into mechanisms responsible for human carotid IMT may be identified by genetic studies in animal models of atherosclerosis.7
Local hemodynamic factors (eg, shear stress) play an important role in carotid IMT. For example, carotid IMT is inversely related to carotid shear stress in healthy subjects: lower shear stress correlated with greater IMT values.8 Moreover, lower shear stress also correlated with progression of coronary atherosclerosis.9 An early study by Lindner’s group showed that complete cessation of blood flow in the carotid caused IMT that varied significantly in magnitude among 11 inbred mouse strains.11 Previously, we found dramatic strain-dependent differences in carotid IMT induced by low blood flow among 5 inbred strains of mice.10 The largest difference was in SJL/J (SJL) mice compared to C3HeB/FeJ (C3H/F) mice.12 We found that greater intima formation in SJL was associated with increased vascular inflammation and proliferation compared to C3H/F mice.12 Most recently we reported that arteries from SJL mice exhibit increased vascular oxidative stress, which may contribute to the basal endothelial dysfunction in this mouse strain.13 The goal of the present study was to use a genetic cross between C3H/F and SJL mice to map modifier genes that affect intima formation in response to low flow.
Materials and Methods
Animals and Surgeries
Parental strains (C3H/F and SJL) were purchased from Jackson Laboratory (Bar Harbor, Me) and bred internally (C3H/FxSJL[F1] and F1xSJL[N2]). Mice (10 to 12 weeks old) were used in accordance with the guidelines of the National Institutes of Health for the care and use of laboratory animals. All experimental protocols were approved by the University of Rochester Animal Care and Use Committee. We lost 4 of 142 total N2 because of surgery, similar to our previous reports.10,14
Tail snips were collected for genotyping 2 weeks after ligation. Instantly all animals were perfusion fixed with 10% paraformaldehyde in sodium phosphate buffer (pH 7.0) as described.14 Briefly, the left and right common carotid arteries from each mouse were harvested and embedded in paraffin. Cross-sections were stained with hematoxylin and eosin, and were analyzed using MCID image software (MCID Elite 6.0, Imaging Research Inc). Intima area of the ligated carotids from the C3H/FxSJL backcross mice are shown in supplemental Figure I (available online at http://atvb.ahajournals.org). Our findings demonstrate that variation of the intima areas are similar to that reported previously.10,14 We calculated volumes of each vessel compartment by using division numbers 2 to 9 from 10 initial divisions. We were successful in histological techniques (only 4/138 N2 samples were unusable). There were no differences in intima or intima/media ratio between males and females in all strains studied as we reported previously.10,14
All progeny from the N2 (F1xSJL) were screened using 80 microsatellite markers (ABI) spaced at approximately 18-cM intervals through the 20 mouse autosomes (http://www.cidr.jhmi.edu). DNA was visualized by fluorescent labeling using ABI Prism 377 sequencer (Applied Biosystems). Data were collected with GeneMapper software to determine genotypes.
MapManager QTX20b was used for linkage analyses. To establish the whole genome linkage thresholds, permutation tests15 were run on our data at 1 cM intervals for 10 000 permutations. Suggestive (P=0.67), significant (P=0.05), and highly significant (P=0.001) levels were established as suggested.16
In Silico Haplotype Analysis
We evaluated the intima modifier QTLs in genomic regions that differ among 5 inbred strains (C3H/HeJ, C57Bl6/J, DBA/2J, FVB/NJ, SJL) that have been studied by our group.10 Haplotype maps of 30-Mb intervals on Chr2 and Chr11 were predicted using Perlegen Mouse SNP Browser (http://mouse.perlegen.com). Physical location of the genes is based on NCBI mouse build 37.1 (http://www.ncbi.nlm.nih.gov). Confirmation of the haplotype analysis was performed by evaluating the 40-Mb intervals on Chr2 and Chr11 of SNPs that differed between C3H/F and SJL mice obtained from Mouse Phenome Database located on The Jackson Laboratory web-site (http://www.jax.org).
Results are reported as mean±SEM. Statistical tests were done with JMP for MacIntosh. Comparison for 2 groups was performed using Student t test. Differences between 3 or more groups were analyzed by means of a repeated-measures 1-way ANOVA. The level of P<0.05 was regarded as significant.
The greatest differences in intima formation (≈30-fold) were found comparing SJL (highest) to C3H/F (lowest) among inbred mouse strains.10,12 Here we observed that both intima and intima/media ratio were significantly greater in SJL mice compared to C3H/F, F1s, and N2s (Figure 1). This result (especially the F1 value, which was not significantly different from C3H/F) suggests that variations in intima and intima/media ratio traits exhibited a dominance of recessive alleles in the C3H/FxSJL backcross (Figure 1A and 1B). After genotyping 134 N2 progeny we identified 2 highly significant QTLs on chr2 and chr11 that showed linkage to intima with LOD scores of 6.4 and 5.3, respectively (Figure 2A). There was 1 significant QTL on chr18 (LOD=2.9; P<0.05) and 2 suggestive QTLs located on chr7 and chr17, respectively (Figure 2A). Details on the significant and suggestive intima QTLs are listed in the Table. Interestingly, the most significant QTLs identified for intima trait colocalized with QTLs that control the intima/media ratio trait in the C3H/FxSJL backcross (Figure 2B). Intima/media ratio exhibited slightly higher LOD scores compared to the intima QTLs (Table).
Interval mapping identified Im1 (intima modifier 1) locus on chr2 with highest LOD at the marker D2Mit411.1, which we estimate accounts for 20% of intima formation in this cross (Figure 3A; Table). Im2 locus on chr11 peaked at marker D11Mit231.1 and accounts for 17% intima variation (Figure 3B). Finally, Im3 locus on the proximal part of chr18 (near D18Mit222.1 marker) accounts for 11% of the trait variation in the C3H/FxSJL backcross (Figure 3C). Likewise, we identified Imrm1 (intima/media ratio modifier 1), Imrm2, and Imrm3 loci on chr2, chr11, and chr18, respectively (Figure 3). The contribution of the 3 Imrm loci to the variation in the intima/media ratio trait was similar to that of the Im loci for intima in the C3H/FxSJL backcross (Table).
To show a relationship between phenotype and genotype, intima and intima/media ratio were evaluated against genotype at the significant loci. As expected, the N2 mice with a homozygous genotype (SJL/SJL) at the markers D2Mit411.1 and D11Mit231.1 exhibited greater intima and intima/media ratio values compared to heterozygous (C3H/SJL) littermates (Figure 4A and 4B). However, we found an opposite trend between genotypes at the D18Mit222.1 marker in relation to both traits (not shown).
Previously we reported that maintenance of lumen area correlated best with carotid stenosis (measured by percentage of intima+media area/external lamina area).10 To estimate the pathophysiological contribution of the studied traits we plotted intima or intima/media ratio against stenosis in N2 animals. There were significant correlations between intima and stenosis, as well as between intima/media ratio and stenosis (Figure 4C and 4D). Taken together, these results suggest that intima and intima/media phenotypes are highly related and driven by the same pathogenetic mechanisms. Moreover, we believe that common genes are located within the QTLs for intima and intima/media ratio traits.
Previously we found dramatic variation in carotid intima formation with higher values in SJL and FVB/NJ mice compared to C3H/HeJ, C57Bl/6J, and DBA/2J mice.10 Similarly, SJL and FVB/NJ had the greatest intima after cessation of flow in the carotid compared to other mouse strains.11 We hypothesized that allelic variations at the QTLs identified on chr2 and chr11 occur within haplotype blocks that are shared by SJL and FVB/NJ, but differ from the other strains. Using Perlegen Mouse SNP browser we identified 2 high-priority regions of different haplotype blocks residing in the regions of Im1 and Im2 loci (boxes, Figure 5A and 5B). The yellow blocks show regions where C3H/HeJ shares alleles with C57Bl/6J and DBA/2J (low intima strains), but differs from FVB/NJ and SJL mice (high intima strains). The blue blocks show regions where the SJL allele is uniquely shared with FVB/NJ. Analysis of the 2 high-priority 1.4-Mb haplotype blocks uncovered 40 known or predicted genes, which represent candidate genes for flow-induced intima formation (supplemental Tables I and II). Because C3H/F, the inbred substrain we used in the cross, is not available on the Perlegen dataset we also obtained informative SNPs within the identified 40-Mb intervals on Chr2 and Chr11 using C3H/F and SJL mice from the Mouse Phenome Database. When we cross-referenced these SNPs with the 40 candidate genes listed in supplemental Tables I and II, we identified 6 genes (Dsn1, Src, Ctnnbl1, 2610036D13Rik, 2610304G08Rik, Spred2), in which SNPs differed between C3H/F and SJL strains. Two important caveats of these analyses are the relatively low density of informative SNPs among the selected inbred strains, and potential differences in the pathophysiology of intima formation between SJL and FVB/NJ mice.12,13
The major finding of the present study is identification for the first time of 3 QTLs and associated genetic variation that contribute significantly to intima formation in the mouse carotid in response to low blood flow. Based on our previous studies of intima formation among 5 inbred mouse strains,10 we hypothesized that a few genes would play major roles in intima formation. Here we discuss several candidate genes that reside within the 3 QTLs based on the following approach. First, by using a traditional genetic approach (C3H/FxSJL backcross) we identified 3 QTLs that linked to intima formation in response to low blood flow. Second, the phenotype–genotype relationship supported our hypothesis that genetic variation in the SJL strain contributes dominant susceptibility alleles to intima formation in this model that may be highly related to arterial stenosis. Third, by analyzing the ancestral haplotype regions that are similar in strains with large intima and differ from strains with small intima, we were able to refine the Im1 and Im2 loci to 40 candidate genes. Fourth, by evaluating genetic variations between parental C3H/F and SJL strains within the QTLs, we identified sequence differences in six genes among the 40 candidate genes identified by haplotype mapping.
Understanding the pathophysiologic mechanisms that contribute to intima formation has been a long-lasting problem in vascular biology with significant relevance to clinical cardiovascular disease.17 Carotid IMT is a complex trait that is likely controlled by a large number of environmental and genetic factors. Recent genetic analyses of intima hyperplasia in the abdominal aorta in response to balloon injury identified several QTLs in an intercross between BN and SHR rats.5 Interestingly, the most significant intima hyperplasia QTL mapped to rat chr3, which is syntenic to mouse chr2. This finding suggests that genes underlying Im1 locus linked to flow-induced intima formation in mice may also control intima proliferation in response to balloon injury in rats.
Because low flow is associated with increased risk for atherosclerosis, there is a supposition that genes responsible for flow-induced vascular remodeling will overlap with genes that influence atherosclerosis. However, it should be noted that the QTLs for carotid atherosclerosis did not overlap with those for aortic lesions in a genetic cross between C3H and B6 strains on the ApoE−/− background.6 There was also no association between carotid lesions and plasma lipids. The authors identified 1 significant QTL, named Cath1 (25 cM, LOD=4.5) on chr12, and 4 suggestive QTLs on chromosomes 1, 5, 6, and 11. Notably, 1 suggestive QTL that controls aortic lesions overlaps with Im2 on chr11 linking intima formation to low flow (Figure 2). Thus our data are consistent with the concept that low flow may be a primary stimulus for both intima and atherosclerotic lesion formation. In keeping with this, a new model of flow-dependent remodeling using an extravascular cast showed a significant role for low flow in carotid lesion development and atherosclerosis progression in ApoE−/− mice.18 The relevance of our genetic analysis is highlighted by the fact that both intima and intima/media ratio traits significantly correlated with stenosis (Figure 4C and 4D). Recently, the Lsq-1 locus on chr7 in mice was related to the vascular response to hindlimb ischemia, a finding that may have relevance for peripheral arterial disease.19 Coincidentally, one of the suggestive intima QTLs (LOD=1.5) in the present study also mapped to the distal part of chr7. The overlap of a flow-induced intima locus with a peripheral arterial disease locus suggests another clinically relevant phenotype that may involve the same genes, which contribute to the intima trait. Evaluations of site-specific vessel geometry, flow, and plaque volume have enabled accurate correlations between local shear stress and plaque progression in human coronary arteries.9 Recent follow-up studies strongly suggested that atherosclerosis progression in minimally diseased coronary segments occurred almost exclusively in areas of low shear stress.20 Thus, our low flow Im loci may provide insights into genes that control human atherosclerosis progression.
There are 2 genetic studies that evaluated carotid IMT in humans. A linkage analysis was conducted to localize QTLs influencing carotid IMT.3 Significant linkage to IMT in the internal portion of the carotid was found on human chr12 (LOD=3.4). Similar to animal models of atherosclerosis, O'Donnell et al reported subclinical atherosclerosis in multiple arterial beds in the Framingham Heart Study Offspring cohort recently.21 The authors performed whole-genome association studies and found 11 SNPs that were associated with subclinical atherosclerosis with P<10−5. They identified candidate genes ABI2 for internal carotid IMT and PCSK2 for common carotid IMT, whereas modest associations were found for the previously reported candidate genes (NOS3 and ESRI).
Using an in silico approach we narrowed the Im loci to 6 genes (Dsn1, Src, Ctnnbl1, 2610036D13Rik, 2610304G08Rik, Spred2) from 40 candidates listed in the supplemental Tables I and II. Among the 4 genes that are expressed proteins each of them could potentially affect intima formation: (1) Src, the Rous sarcoma oncogene is a known growth mediator22; (2) Dsn1, the MIND kinetochore complex component is required for proper chromosome alignment and cell cycle progression23; (3) Ctnnbl1, catenin (cadherin-associated protein) βlike 1 protein is involved in cell apoptosis24; and (4) Spred2, sprouty-related protein, belongs to a new protein family harboring a conserved N-terminal EVH1 domain, which is related to the VASP (vasodilator-stimulated phosphoprotein) EVH1 domain (Enabled/VASP homology 1 domain) and a C-terminal sprouty-related (SPR) domain, typical for the Drosophila sprouty proteins.25 In Drosophila, sprouty inhibits signaling by fibroblast growth factor and epidermal growth factor by preventing phosphorylation of Raf kinase and blocking the MAP kinase pathway.26 Future studies will be required to determine the specific roles of the candidate genes for the dramatic vascular phenotypes in SJL mice compared to other strains of mice.12,13
Our study suggested several candidate genes for intima formation (based on sequence variation between parental strains) that will require multiple approaches to characterize their roles in intima formation. For example, genetic interactions among the causal genes within the Im loci with other genes involved in intima formation can be studied by evaluation of mRNA and protein expression profiles in carotids in response to low blood flow. Based on one published expression profile of flow-induced remodeling in rat mesenteric arteries,27 we anticipate many changes in mRNA and protein expression. Specifically these authors found that alteration of flow in resistance arteries modified expression of ≈5% of genes by >2-fold.
In summary, this study for the first time identified several QTLs that link to intima formation in response to low blood flow. Intima formation in response to flow is a complex trait, likely regulated by multiple modifier genes. The present study has shown 3 genomic regions (and 6 candidate genes) that may regulate intima formation in response to low flow. These genes are candidates to play a role in many clinical settings including vascular injury, carotid atherosclerotic disease (IMT), and peripheral arterial disease. While the present in silico mapping approach is very powerful, future studies will be required to investigate these candidate genes and interactions within the Im loci in an integrative manner.
We thank Amy Mohan for help with mice husbandry, Sarah Mack for help with tissue processing, and Dr Stephen Welle, Andrew Cardillo, and Michelle Zanche from Functional Genomic Center (University of Rochester, NY) for help with genotyping.
Sources of Funding
This study was supported by NIH HL-62826 to B.C.B.
Original received August 25, 2008; final version accepted October 10, 2008.
Fox CS, Polak JF, Chazaro I, Cupples A, Wolf PA, D'Agostino RA, O'Donnell CJ. Genetic and environmental contributions to atherosclerosis phenotypes in men and women: heritability of carotid intima-media thickness in the Framingham heart study. Stroke. 2003; 34: 397–401.
Schwartz SM, deBlois D, O'Brien ER. The intima. Soil for atherosclerosis and restenosis. Circ Res. 1995; 77: 445–465.
Nestor AL, Cicila GT, Karol SE, Langenderfer KM, Hollopeter SL, Allison DC. Linkage analysis of neointimal hyperplasia and vascular wall transformation after balloon angioplasty. Physiol Genomics. 2006; 25: 286–293.
Li Q, Li Y, Zhang Z, Gilbert TR, Matsumoto AH, Dobrin SE, Shi W. Quantitative trait locus analysis of carotid atherosclerosis in an intercross between C57BL/6 and C3H apolipoprotein E-deficient mice. Stroke. 2008; 39: 166–173.
Carallo C, Irace C, Pujia A, De Franceschi MS, Crescenzo A, Motti C, Cortese C, Mattioli PL, Gnasso A. Evaluation of common carotid hemodynamic forces. Relations with wall thickening. Hypertension. 1999; 34: 217–221.
Stone PH, Coskun AU, Kinlay S, Clark ME, Sonka M, Wahle A, Ilegbusi OJ, Yeghiazarians Y, Popma JJ, Orav J, Kuntz RE, Feldman CL. Effect of endothelial shear stress on the progression of coronary artery disease, vascular remodeling, and in-stent restenosis in humans. In vivo 6-month follow-up study. Circulation. 2003; 108: 438–444.
Korshunov VA, Berk BC. Strain-dependent vascular remodeling: the “Glagov phenomenon” is genetically determined. Circulation. 2004; 110: 220–226.
Korshunov VA, Nikonenko TA, Tkachuk VA, Brooks A, Berk BC. Interleukin-18 and macrophage migration inhibitory factor are associated with increased carotid intima-media thickening. Arterioscler Thromb Vasc Biol. 2006; 26: 295–300.
Korshunov VA, Berk BC. Flow-induced vascular remodeling in the mouse: a model for carotid intima-media thickening. Arterioscler Thromb Vasc Biol. 2003; 23: 2185–2191.
Doerge RW, Churchill GA. Permutation tests for multiple loci affecting a quantitative character. Genetics. 1996; 142: 285–294.
Healy DA, Zierler RE, Nicholls SC, Clowes AW, Primozich JF, Bergelin RO, Strandness DE Jr. Long-term follow-up and clinical outcome of carotid restenosis. J Vasc Surg. 1989; 10: 662–668,discussion 668–669.
Cheng C, Tempel D, van Haperen R, van der Baan A, Grosveld F, Daemen MJ, Krams R, de Crom R. Atherosclerotic lesion size and vulnerability are determined by patterns of fluid shear stress. Circulation. 2006; 113: 2744–2753.
Dokun AO, Keum S, Hazarika S, Li Y, Lamonte GM, Wheeler F, Marchuk DA, Annex BH. A quantitative trait locus (LSq-1) on mouse chromosome 7 is linked to the absence of tissue loss after surgical hindlimb ischemia. Circulation. 2008; 117: 1207–1215.
Stone PH, Coskun AU, Kinlay S, Popma JJ, Sonka M, Wahle A, Yeghiazarians Y, Maynard C, Kuntz RE, Feldman CL. Regions of low endothelial shear stress are the sites where coronary plaque progresses and vascular remodelling occurs in humans: an in vivo serial study. Eur Heart J. 2007; 28: 705–710.
O'Donnell CJ, Cupples LA, D'Agostino RB, Fox CS, Hoffmann U, Hwang SJ, Ingellson E, Liu C, Murabito JM, Polak JF, Wolf PA, Demissie S. Genome-wide association study for subclinical atherosclerosis in major arterial territories in the NHLBI’s Framingham Heart Study. BMC Med Genet. 2007; 8 Suppl 1: S4.
Ishida M, Ishida T, Thomas S, Berk BC. Activation of extracellular signal-regulated kinases (ERK1/2) by angiotensin II is dependent on c-Src in vascular smooth muscle cells. Circ Res. 1998; 82: 7–12.
Kline SL, Cheeseman IM, Hori T, Fukagawa T, Desai A. The human Mis12 complex is required for kinetochore assembly and proper chromosome segregation. J Cell Biol. 2006; 173: 9–17.