Selective genotyping with epistasis can be utilized for a major quantitative trait locus mapping in hypertension in rats

Towards Precision Medicine for Hypertension: A Review of Genomic, Epigenomic, and Microbiomic Effects on Blood Pressure in Experimental Rat Models and Humans

Compelling evidence for the inherited nature of essential hypertension has led to extensive research in rats and humans. Rats have served as the primary model for research on the genetics of hypertension resulting in identification of genomic regions that are causally associated with hypertension. In more recent times, genome-wide studies in humans have also begun to improve our understanding of the inheritance of polygenic forms of hypertension. Based on the chronological progression of research into the genetics of hypertension as the "structural backbone," this review catalogs and discusses the rat and human genetic elements mapped and implicated in blood pressure regulation. Furthermore, the knowledge gained from these genetic studies that provide evidence to suggest that much of the genetic influence on hypertension residing within noncoding elements of our DNA and operating through pervasive epistasis or gene-gene interactions is highlighted. Lastly, perspectives on current thinking that the more complex "triad" of the genome, epigenome, and the microbiome operating to influence the inheritance of hypertension, is documented. Overall, the collective knowledge gained from rats and humans is disappointing in the sense that major hypertension-causing genes as targets for clinical management of essential hypertension may not be a clinical reality. On the other hand, the realization that the polygenic nature of hypertension prevents any single locus from being a relevant clinical target for all humans directs future studies on the genetics of hypertension towards an individualized genomic approach.

Identification of QTLs for seed and pod traits in soybean and analysis for additive effects and epistatic effects of QTLs among multiple environments

Soybean seed and pod traits are important yield components. Selection for high yield style in seed and pod along with agronomic traits is a goal of many soybean breeders. The intention of this study was to identify quantitative trait loci (QTL) underlying seed and pod traits in soybean among eleven environments in China. 147 recombinant inbred lines were advanced through single-seed-descent method. The population was derived from a cross between Charleston (an American high yield soybean cultivar) and DongNong594 (a Chinese high yield soybean cultivar). A total of 157 polymorphic simple sequence repeat markers were used to construct a genetic linkage map. The phenotypic data of seed and pod traits [number of one-seed pod, number of two-seed pod, number of three-seed pod, number of four-seed pod, number of (two plus three)-seed pod, number of (three plus four)-seed pod, seed weight per plant, number of pod per plant] were recorded in eleven environments. In the analysis of single environment, fourteen main effect QTLs were identified. In the conjoint analysis of multiple environments, twenty-four additive QTLs were identified, and additive QTLs by environments interactions (AE) were evaluated and analyzed at the same time among eleven environments; twenty-three pairs of epistatic QTLs were identified, and epistasis (additive by additive) by environments interactions (AAE) were also analyzed and evaluated among eleven environments. Comparing the results of identification between single environment mapping and multiple environments conjoint mapping, three main effect QTLs with positive additive values and another three main effect QTLs with negative additive values, had no interactions with all environments, supported that these QTLs could be used in molecular assistant breeding in the future. These different effect QTLs could supply a good foundation to the gene clone and molecular asisstant breeding of soybean seed and pod traits.

Isolation and high-throughput sequencing of two closely linked epistatic hypertension susceptibility loci with a panel of bicongenic strains

Interactions or epistasis between genetic factors may contribute to "missing heritability." While linkage analyses detect epistasis, defining the limits of the interacting segments poses a significant challenge especially when the interactions are between loci in close proximity. The goal of the present study was to isolate two such epistatic blood pressure (BP) loci on rat chromosome 5. A panel of S.LEW bicongenic strains along with the corresponding monocongenic strains was constructed. BP of each set comprising of one bicongenic and two corresponding monocongenic strains were determined along with the parental Salt-sensitive (S) strain. Epistasis was observed in one out of four sets of congenic strains, wherein systolic blood pressures (SBP) of the two monocongenic strains S.LEW(5)x6Bx9x5a and S.LEW(5)x6Bx9x5b were comparable to that of S, but the SBP of the bicongenic strain S.LEW(5)x6Bx9x5 (157 ± 4.3 mmHg) was significantly lower than that of S (196 ± 6.8 mmHg, P < 0.001). A two-way ANOVA indicated significant interactions between the LEW alleles at the two loci. The interacting loci were 2.02 Mb apart and located within genomic segments spanning 7.77 and 4.18 Mb containing 7,360 and 2,753 candidate variants, respectively. The current study demonstrates definitive evidence for epistasis and provides genetic tools for further dissection of the isolated epistatic BP loci.

QTL analysis of soybean seed weight across multi-genetic backgrounds and environments

Seed weight, measured as mass per seed, is an important yield component of soybean and is generally positively correlated with seed yield (Burton et al, Crop Sci 27:1093, 1987). In previous reports, quantitative trait loci (QTL) associated with seed weight, were identified in single genetic background. The objective of the present study was to identify QTL and epistatic QTL underlying soybean seed weight in three RIL populations (with one common male parent 'Hefeng25') and across three different environments. Overall, 18, 11, and 17 seed weight QTL were identified in HC ('Hefeng25' × 'Conrad'), HM ('Hefeng25' × 'Maple Arrow'), and HB ('Hefeng25' × 'Bayfield') populations, respectively. The amount of phenotypic variation explained by a single QTL underlying seed weight was usually less than 10 %. The environment and background-independent QTL often had higher additive (a) effects. In contrast, the environment or background-dependent QTL were probably due to weak expression of QTL. QTL by environment interaction effects were in the opposite direction of a effects and/or epistasis effects. Four QTL and one QTL could be identified (2.0 < LOD < 9.06) in the HC and HB populations, respectively, across three environments (swHCA2-1, swHCC2-1, swHCD1b-1, swHCA2-2 (linked to Satt233, Satt424, Satt460, Satt428, respectively) and swHBA1-1(Satt449). Seven QTL could be identified in all three RIL populations in at least one location. Two QTL could be identified in the three RIL populations across three environments. These two QTL may have greater potential for use in marker-assisted selection of seed weight in soybean.

The statistical power of inclusive composite interval mapping in detecting digenic epistasis showing common F2 segregation ratios

Epistasis is a commonly observed genetic phenomenon and an important source of variation of complex traits, which could maintain additive variance and therefore assure the long-term genetic gain in breeding. Inclusive composite interval mapping (ICIM) is able to identify epistatic quantitative trait loci (QTLs) no matter whether the two interacting QTLs have any additive effects. In this article, we conducted a simulation study to evaluate detection power and false discovery rate (FDR) of ICIM epistatic mapping, by considering F2 and doubled haploid (DH) populations, different F2 segregation ratios and population sizes. Results indicated that estimations of QTL locations and effects were unbiased, and the detection power of epistatic mapping was largely affected by population size, heritability of epistasis, and the amount and distribution of genetic effects. When the same likelihood of odd (LOD) threshold was used, detection power of QTL was higher in F2 population than power in DH population; meanwhile FDR in F2 was also higher than that in DH. The increase of marker density from 10 cM to 5 cM led to similar detection power but higher FDR. In simulated populations, ICIM achieved better mapping results than multiple interval mapping (MIM) in estimation of QTL positions and effect. At the end, we gave epistatic mapping results of ICIM in one actual population in rice (Oryza sativa L.).

Possible association of tumor necrosis factor receptor 2 gene polymorphism with severe hypertension using the extreme discordant phenotype design

The tumor necrosis factor (TNF)-alpha pathway has a key role in regulating insulin resistance. TNF receptor 2 (TNFR2) is an emerging candidate gene for insulin resistance in essential hypertension. We examined the association of insulin resistance and enhanced TNF pathway with severe hypertension and the association of a microsatellite polymorphism of the TNFR2 gene with severe hypertension. Male severe essential hypertensive patients (HT) with the onset before 60 years of age and with genetic predispositions to hypertension were consecutively enrolled at our outpatient department (N=92). Normotensive men (NT) over 50 years of age were randomly registered from the participants in the annual health check program (N=78). Patients were selected as HT and NT who met stringent criteria for systolic/diastolic blood pressure (SBP/DBP) levels >or=180 and/or 110 mm Hg and <120/80 mm Hg, respectively. HT revealed significantly higher plasma insulin levels, C-reactive protein (CRP) and soluble fraction of TNFR2 concentrations (sTNFR2) than NT. A microsatellite polymorphism of the CA repeat in intron 4 of the TNFR2 gene was analyzed. The allele frequency of CA16 in HT differed significantly from that in NT (66/184 vs. 36/156, P=0.01 by chi(2) analysis). In HT, the CA16 carriers showed significantly higher SBP and plasma insulin levels and a higher tendency of sTNFR2 than did those without this allele. In NT, CA16 carriers revealed significantly higher sTNFR2 and CRP levels than did the CA16 non-carriers. These results suggest that the TNFR2 gene locus has a potential effect on developing severe hypertension through the augmented TNF pathway and insulin resistance.

Investigating the effect of genetic background on proteinuria and renal injury using two hypertensive strains

An earlier linkage analysis conducted on a population derived from the Dahl salt-sensitive hypertensive (S) and the spontaneously hypertensive rat (SHR) identified 10 genomic regions linked to several renal and/or cardiovascular traits. In particular, loci on rat chromosomes (RNO) 8 and 13 were linked to proteinuria, albuminuria, and renal damage. At both loci, the S allele was associated with increased proteinuria and renal damage. The current study aimed to confirm the linkage analysis and to evaluate the effect of genetic background on the ability of each locus (either RNO8 or RNO13) to exert a phenotypic difference when placed on a genetic background either susceptible (S rat) or resistant (SHR) to the development of renal disease. Congenic strains developed to transfer genomic segments from either RNO8 or RNO13 from the SHR onto the S genetic background [S.SHR(8) or S.SHR(13)] demonstrated significantly reduced proteinuria and improved renal function. Both congenic strains demonstrated significantly reduced glomerular and tubular injury, with renal interstitial fibrosis as the predominant pathological difference compared with the S. In contrast, transfer of RNO8 or RNO13 genomic regions from the S onto the resistant SHR genetic background [SHR.S(8) or SHR.S(13)] yielded no significant difference in proteinuria or glomerular, tubular, or interstitial injury compared with SHR. These findings demonstrate that genetic context plays a significant and important role in the phenotypic expression of genes influencing proteinuria on RNO8 and RNO13.

QTL, additive and epistatic effects for SCN resistance in PI 437654

PI 437654 is a unique accession because of its resistance to nearly all HG types (races) of soybean cyst nematode (Heterodera glycines Ichinohe; SCN). Objectives of this study were to confirm and refine the locations and gene action associated with SCN resistance previously discovered in PI 437654, and to identify new QTLs that may have been missed because of low coverage with genetic markers used in previous studies. Using 205 F(7:9) RILs and 276 SSR and AFLP molecular markers covering 2,406.5 cM of 20 linkage groups (LGs), we confirmed and refined the locations of major SCN resistance QTLs on LG-A2, -B1, and -G previously identified in PI 437654 or other resistant sources. We found that these major QTLs have epistatic effects among them or with other loci for SCN resistance. We also detected some new QTLs with additive or epistatic effects for SCN resistance to different HG types (races) on all LGs except LGs-B2 and -D1b. The QTL on LG-G was associated with resistance to HG types 2.5.7, 1.2.5.7, 0, and 2.7 (races 1, 2, 3, and 5), and it contributed a large proportion of the additive effects. The QTL on LG-A2 was associated with resistance to HG types 2.5.7 and 0 (races 1 and 3). The QTL on LG-B1, associated with resistance to HG types 2.5.7, 0, 2.7 (races 1, 3, and 5), was the similar QTL found in PI 90763 and PI 404198B. In addition to QTL on LGs-A2, -B1 and -G, a novel additive QTL associated with SCN resistance to HG types 0, 2.7, and 1.3.5.6.7 (race 3, 5, and 14) was identified on LG-I flanked by Sat_299 and Sat_189. Several minor QTLs on LGs-C1, D1a, H, and K were also found to be associated with SCN resistance. Confirmation of the new resistance QTL is underway by evaluating another RIL population with a different genetic background.

Selective genotyping and phenotyping strategies in a complex trait context

Selective genotyping and phenotyping strategies are used to lower the cost of quantitative trait locus studies. Their efficiency has been studied primarily in simplified contexts--when a single locus contributes to the phenotype, and when the residual error (phenotype conditional on the genotype) is normally distributed. It is unclear how these strategies will perform in the context of complex traits where multiple loci, possibly linked or epistatic, may contribute to the trait. We also do not know what genotyping strategies should be used for nonnormally distributed phenotypes. For time-to-event phenotypes there is the additional question of choosing follow-up time duration. We use an information perspective to examine these experimental design issues in the broader context of complex traits and make recommendations on their use.

FastANOVA: an Efficient Algorithm for Genome-Wide Association Study

Studying the association between quantitative phenotype (such as height or weight) and single nucleotide polymorphisms (SNPs) is an important problem in biology. To understand underlying mechanisms of complex phenotypes, it is often necessary to consider joint genetic effects across multiple SNPs. ANOVA (analysis of variance) test is routinely used in association study. Important findings from studying gene-gene (SNP-pair) interactions are appearing in the literature. However, the number of SNPs can be up to millions. Evaluating joint effects of SNPs is a challenging task even for SNP-pairs. Moreover, with large number of SNPs correlated, permutation procedure is preferred over simple Bonferroni correction for properly controlling family-wise error rate and retaining mapping power, which dramatically increases the computational cost of association study.In this paper, we study the problem of finding SNP-pairs that have significant associations with a given quantitative phenotype. We propose an efficient algorithm, FastANOVA, for performing ANOVA tests on SNP-pairs in a batch mode, which also supports large permutation test. We derive an upper bound of SNP-pair ANOVA test, which can be expressed as the sum of two terms. The first term is based on single-SNP ANOVA test. The second term is based on the SNPs and independent of any phenotype permutation. Furthermore, SNP-pairs can be organized into groups, each of which shares a common upper bound. This allows for maximum reuse of intermediate computation, efficient upper bound estimation, and effective SNP-pair pruning. Consequently, FastANOVA only needs to perform the ANOVA test on a small number of candidate SNP-pairs without the risk of missing any significant ones. Extensive experiments demonstrate that FastANOVA is orders of magnitude faster than the brute-force implementation of ANOVA tests on all SNP pairs.

Mapping QTLs with digenic epistasis under multiple environments and predicting heterosis based on QTL effects

Mixed linear model approach was proposed for mapping QTLs with the digenic epistasis and QTL by environment (QE) interaction as well as additive and dominant effects. Monte Carlo simulations indicated that the proposed method could provide unbiased estimations for both positions and genetic main effects of QTLs, as well as unbiased predictions for QE interaction effects. A method was suggested for predicting heterosis based on individual QTL effects. The immortalized F(2) (IF(2)) population constructed by random mating among RI or DH lines is appropriate for mapping QTLs with epistasis and their QE interaction. Based on the models and methodology proposed, we developed a QTL mapping software, QTLMapper 2.0 on the basis of QTLmapper 1.0, which is suitable for analyzing populations of DH, RIL, F(2) and IF(2). Data of thousand grain weight of IF(2) population with 240 lines derived from elite hybrid rice Shanyou 63 were analyzed as a worked example.

Evidence for a major gene affecting the transition from normoglycaemia to hyperglycaemia in Psammomys obesus

We investigated the mode of inheritance of nutritionally induced diabetes in the desert gerbil Psammomys obesus (sand rat), following transfer from low-energy (LE) to high-energy (HE) diet which induces hyperglycaemia. Psammomys selected for high or low blood glucose level were used as two parental lines. A first backcross generation (BC(1)) was formed by crossing F(1) males with females of the diabetes-prone line. The resulting 232 BC(1) progeny were assessed for blood glucose. All progeny were weaned at 3 weeks of age (week 0), and their weekly assessment of blood glucose levels proceeded until week 9 after weaning, with all progeny maintained on HE diet. At weeks 1 to 9 post weaning, a clear bimodal distribution statistically different from unimodal distribution of blood glucose was observed, normoglycaemic and hyperglycaemic at a 1:1 ratio. This ratio is expected at the first backcross generation for traits controlled by a single dominant gene. From week 0 (prior to the transfer to HE diet) till week 8, the hyperglycaemic individuals were significantly heavier (4--17%) than the normoglycaemic ones. The bimodal blood glucose distribution in BC(1) generation, with about equal frequencies in each mode, strongly suggests that a single major gene affects the transition from normo- to hyperglycaemia. The wide range of blood glucose values among the hyperglycaemic individuals (180 to 500 mg/dl) indicates that several genes and environmental factors influence the extent of hyperglycaemia. The diabetes-resistant allele appears to be dominant; the estimate for dominance ratio is 0.97.

The use of animal models in the study of complex disease: all else is never equal or why do so many human studies fail to replicate animal findings?

The study of the genetics of complex human disease has met with limited success. Many findings with candidate genes fail to replicate despite seemingly overwhelming physiological data implicating the genes. In contrast, animal model studies of the same genes and disease models usually have more consistent results. We propose that one important reason for this is the ability to control genetic background in animal studies. The fact that controlling genetic background can produce more consistent results suggests that the failure to replicate human findings in the same diseases is due to variation in interacting genes. Hence, the contrasting nature of the findings from the different study designs indicates the importance of non-additive genetic effects on human disease. We discuss these issues and some methodological approaches that can detect multilocus effects, using hypertension as a model disease. This article contains supplementary material, which may be viewed at the BioEssays website at http://www.interscience.wiley.com/jpages/0265-9247/suppmat/index.html.

Experimental gene interaction studies with SERT mutant mice as models for human polygenic and epistatic traits and disorders

Current evidence indicates that virtually all neuropsychiatric disorders, like many other common medical disorders, are genetically complex, with combined influences from multiple interacting genes, as well as from the environment. However, additive or epistatic gene interactions have proved quite difficult to detect and evaluate in human studies. Mouse phenotypes, including behaviors and drug responses, can provide relevant models for human disorders. Studies of gene-gene interactions in mice could thus help efforts to understand the molecular genetic bases of complex human disorders. The serotonin transporter (SERT, 5-HTT, SLC6A4) provides a relevant model for studying such interactions for several reasons: human variants in SERT have been associated with several neuropsychiatric and other medical disorders and quantitative traits; SERT blockers are effective treatments for a number of neuropsychiatric disorders; there is a good initial understanding of the phenotypic features of heterozygous and homozygous SERT knockout mice; and there is an expanding understanding of the interactions between variations in SERT expression and variations in the expression of a number of other genes of interest for neuropsychiatry and neuropharmacology. This paper provides examples of experimentally-obtained interactions between quantitative variations in SERT gene expression and variations in the expression of five other mouse genes: DAT, NET, MAOA, 5-HT(1B) and BDNF. In humans, all six of these genes possess polymorphisms that have been independently investigated as candidates for neuropsychiatric and other disorders in a total of > 500 reports. In the experimental studies in mice reviewed here, gene-gene interactions resulted in either synergistic, antagonistic (including 'rescue' or 'complementation') or more complex, quantitative alterations. These were identified in comparisons of the behavioral, physiological and neurochemical phenotypes of wildtype mice vs. mice with single allele or single gene targeted disruptions and mice with partial or complete disruptions of multiple genes. Several of the descriptive phenotypes could be best understood on the basis of intermediate, quantitative alterations such as brain serotonin differences. We discuss the ways in which these interactions could provide models for studies of gene-gene interactions in complex human neuropsychiatric and other disorders to which SERT may contribute, including developmental disorders, obesity, polysubstance abuse and others.

Harnessing the mouse to unravel the genetics of human disease

Complex traits, i.e. those with multiple genetic and environmental determinants, represent the greatest challenge for genetic analysis, largely due to the difficulty of isolating the effects of any one gene amid the noise of other genetic and environmental influences. Methods exist for detecting and mapping the Quantitative Trait Loci (QTLs) that influence complex traits. However, once mapped, gene identification commonly involves reduction of focus to single candidate genes or isolated chromosomal regions. To reach the next level in unraveling the genetics of human disease will require moving beyond the focus on one gene at a time, to explorations of pleiotropism, epistasis and environment-dependency of genetic effects. Genetic interactions and unique environmental features must be as carefully scrutinized as are single gene effects. No one genetic approach is likely to possess all the necessary features for comprehensive analysis of a complex disease. Rather, the entire arsenal of behavioral genomic and other approaches will be needed, such as random mutagenesis, QTL analyses, transgenic and knockout models, viral mediated gene transfer, pharmacological analyses, gene expression assays, antisense approaches and importantly, revitalization of classical genetic methods. In our view, classical breeding designs are currently underutilized, and will shorten the distance to the target of understanding the complex genetic and environmental interactions associated with disease. We assert that unique combinations of classical approaches with current behavioral and molecular genomic approaches will more rapidly advance the field.