Genetic variation in NCAM1 contributes to left ventricular wall thickness in hypertensive families

Abstract

Rationale

Left ventricular (LV) mass and related phenotypes are heritable, important predictors of cardiovascular disease, particularly in hypertensive individuals.

Objective

Identify genetic predictors of echocardiographic phenotypes in hypertensive families.

Methods & Results

A multi-stage genome-wide association study (GWAS) was conducted in hypertensive-ascertained African American families (HyperGEN, Stage I; GENOA, Stage II); findings were replicated in HyperGEN Caucasian families (Stage III). Echocardiograms were collected using a common protocol, and participants were genotyped with the Affymetrix Genome-Wide Human SNP 6.0 Array. In Stages I and II, 1258 and 989 African Americans, and Stage III 1316 Caucasians, were analyzed using mixed models adjusted for ancestry. Phenotypes included LV mass, LV internal dimension (LVID), wall thicknesses (posterior (PWT) and intraventricular septum (IVST)), and relative wall thickness (RWT). In Stage I, 5 single nucleotide polymorphisms (SNP) had P≤10⁻⁶. In Stage II, one SNP (rs1436109; NCAM1 intron 1) replicated with the same phenotype (PWT, P=0.025) in addition to RWT (P=0.032). In Stage III, rs1436109 was associated with RWT (P=5.47×10⁻⁴) and LVID (P=1.86×10⁻⁴). Fisher’s combined P-value for all stages was RWT=3.80×10⁻⁹, PWT=3.12×10⁻⁷, IVST=8.69×10⁻⁷, LV mass=2.52×10⁻³, and LVID=4.80×10⁻⁴.

Conclusions

This GWAS conducted in hypertensive families identified a variant in NCAM1 associated with LV wall thickness and RWT. NCAM is upregulated during the remodeling period of hypertrophy to heart failure in Dahl salt-sensitive rats. Our initial screening in hypertensive African-Americans may have provided the context for this novel locus.

Introduction

Increased left ventricular (LV) mass, termed LV hypertrophy, occurs in 33–43% of African Americans,¹ a prevalence double that of Caucasians.² African Americans also exhibit more concentric remodeling than Caucasians, a pattern of LV hypertrophy characterized by increased pressure load and risk of diastolic dysfunction.³ LV mass (LVM), and wall thickness phenotypes used in estimating LVM, are genetically determined and have high heritabilities,^5,6 particularly in African Americans.⁷

We describe results of a three-stage genome-wide association study (GWAS) for echocardiographically determined LVM, posterior wall thickness (PWT), intraventricular septal thickness (IVST), LV internal dimension (LVID), and relative wall thickness (RWT; 2*PWT/LVID) in two hypertensive-ascertained family studies. Stages I and II were done in African Americans from HyperGEN (Hypertension Genetic Epidemiology Network)⁸ and GENOA (Genetic Epidemiology Network of Atherosclerosis).^9,10 Stage III was done in HyperGEN Caucasian families.

Methods

Complete methods are presented in the Online Supplement. Eligible sibships had ≥2 siblings diagnosed with hypertension before age 60. Echocardiographic measures were collected using the same protocol in both studies. Genetic data for all stages were obtained via the Affymetrix 6.0 array for 1,264 African American and 1,358 Caucasian HyperGEN and 989 GENOA participants.

All five phenotypes were log-transformed. Principal components to adjust for population substructure were generated in Eigenstrat.¹¹ Analyses were conducted in two steps using GRAMMER method.¹² SNPs with a P-value<10⁻⁶ from Stage I were brought forward to Stage II; SNPs with a P-value <10⁻⁷ from Stage II were brought forward to Stage III. Fisher’s combined probability test was used to calculate the meta-analytic P-value.¹³

Results

Table 1 shows clinical and phenotypic data. All SNPs involved reported herein were in Hardy-Weinberg equilibrium. QQ plots from Stage I (HyperGEN African Americans) are in the Online Supplement Figure I. Stage I genome-wide association findings for associations where P ≤ 10⁻⁶, excluding results presented below, are shown in Online Supplement Table I. One of five SNPs from Stage I (rs1436109) replicated in Stage II for PWT (P-value=0.025) and RWT (P-value=0.032; Table 2); the SNP is intragenic in the neural cell adhesion molecule 1 (NCAM1) and the minor allele frequency was 2.3% in African Americans in both cohorts. The regional plot surrounding rs1436109 (Stage I) is shown in Figure 1; while only rs1436109 passed our pre-determined threshold, multiple SNPs within NCAM1 are associated with PWT. The only SNP tested in Stage III was rs1436109; the variant allele frequency was 11.6% and the association remained for RWT (P-value=5.47×10⁻⁴); rs1436109 was strongly associated with multiple phenotypes (i.e., Fishers P-values for three stages combined: RWT=3.80×10⁻⁹; PWT=3.12×10⁻⁷, IVST=8.69×10⁻⁷; LV mass=2.52×10⁻³ and LVID=4.80×10⁻⁴).

Table 1.

Phenotypic values for HyperGEN and GENOA.

Characteristic	Stage I Sample HyperGEN African Americans Mean (SD) or % N=1258	Stage II sample GENOA African Americans Mean (SD) or % N=989	Stage III Sample HyperGEN Caucasians Mean (SD) or % N=1316
Age, years	45.25 (13.32)	62.75 (9.38)	49.59 (13.93)
Females, %	66.8	70.2	51.06
Diabetic, %	16.8	31.1	9.2
Body mass index, kg/m2	32.53 (8.02)	31.59 (6.32)	29.40 (6.11)
Systolic blood pressure, mm Hg	129.15 (22.12)	138.29 (20.97)	123.33 (19.17)
Diastolic blood pressure, mm Hg	73.92 (11.59)	79.50 (10.65)	70.69 (10.02)
Hypertension, %	72.31	80.40	55.78
LV mass (indexed to height2.7), g/m2.7	42.21 (12.42)	39.38 (10.65)	37.54 (11.59)
LV Hypertrophy, %	25.94	15.77	14.92
Diastolic posterior wall thickness, cm	0.86 (0.13)	0.82 (0.12)	0.81 (0.15)
Diastolic intraventricular septal thickness, cm	0.93 (0.15)	0.90 (0.13)	0.88 (0.14)
Diastolic LV internal diameter, cm	5.16 (0.53)	5.18 (0.50)	5.16 (0.50)

LV, left ventricular

Table 2.

P-values for SNP rs1436109 in HyperGEN and GENOA samples.

Phenotype*	Stage I: HyperGEN African American	Stage II: GENOA African American	Stages I+II Meta- analysis†	Stage III: HyperGEN Caucasian	Stages I+II+III Meta- analysis†
LVM	2.92×10−3	1.56×10−1	3.96×10−3	8.88×10−2	2.52×10−3
RWT	6.22×10−7	3.24×10−2	3.77×10−7	5.47×10−4	3.80×10−9
PWT	3.60×10−7	2.52×10−2	1.77×10−7	1.49×10−1	3.12×10−7
LVID	4.90×10−1	9.83×10−1	8.34×10−1	1.86×10−4	4.80×10−3
IVST	1.78×10−7	5.29×10−2	1.83×10−7	4.47×10−1	8.69×10−7

^*Phenotypes adjusted for age, age², sex, center, and principal components 1–30; Chr, chromosome;

^†Fisher’s combined probability; SNP, single-nucleotide polymorphism; LVM, left ventricular mass; RWT, relative wall thickness; PWT, diastolic posterior wall thickness; LVID, left ventricular interal dimension; IVST intraventricular septal thickness.

Figure 1.

Genome-wide association findings for the log of diastolic posterior wall thickness adjusted for age, age², sex, center, and principal components 1–30 in 1 Mb region surrounding single-nucleotide polymorphism (SNP) rs1436109 in HyperGEN African Americans. Points are colored according to the level of linkage disequilibrium of the each SNP with the rs1436109; r² estimates generated using 1000 Genomes June 2010 Yoruban data. Hack marks at the top reflect SNP density. The bottom panel indicates gene location; exons are indicated with a wider band. This figure is based on output generated by LocusZoom.¹⁴

Discussion

This is the first GWAS of echocardiographic phenotypes in hypertensive African Americans and Caucasians. We identified a novel locus, NCAM1, which explained 1.96% of variation in PWT in our discovery cohort. This same variant demonstrated consistent associations in the combined analysis of all three hypertensive-ascertained cohorts with other measures of the LV (IVST, LVID, RWT) and mass (LVM/ht^2.7).

NCAM1 is an abundant glycoprotein involved in cell-cell and cell-matrix adhesion¹⁵ that mediates adhesive interactions between cells in the nervous system and muscle.¹⁶ Kimura et al.¹⁷ demonstrated in pulmonary-hypertension-induced RV hypertrophy an increase in regional anatomical hyperinnervation that lead to hyperfunctioning of cardiac sympathetic nerves, downregulation of neuronal cellular function and an induction of fetal gene expression in sympathetic neurons. They suggested that “this phenomenon was explained by rejuvenation of cardiac sympathetic nerves as well as the hypertrophic cardiomyocytes, which also showed the fetal form gene expression.” The results from this GWAS indicate NCAM1 is particularly relevant for increased LV wall thickness in the context of pressure overload (i.e, hypertension), a condition highly prevalent in African Americans. NCAM1 is also upregulated in myocytes in or adjacent to infarct regions and during remodeling in rat induced heart failure.^{18, 19} Although NCAM1 may have no known direct functional involvement with our phenotypes, it is directly connected through biological interactions with numerous genes, many of which are expressed in cardiac tissue or have been shown to be functionally involved with cardiac phenotypes (Figure 2).

Figure 2.

Interactions of NCAM1 generated with Ingenuity Pathways Analysis (Ingenuity® Systems, www.ingenuity.com). Blue lines are protein-protein interactions. Gray line interactions coded with letter. Red genes are expressed in cardiac tissue.

The NCAM1 SNP demonstrating the strongest association is in an 80 kilobase linkage disequilibrium block in intron 1. NCAM1 contains a large intron 1, and while the functional or regulatory significance of this SNP is unknown, intron 1 is known to contain more regulatory regions than other introns.²⁰ There are at least four transcription factor binding sites (TFBS) within a 4 kilobases of rs1436109 that are conserved in humans, mouse, and rat, and two potential TFBS contain our SNP. There are regulatory elements within 6 Kb of rs1436109 as identified by DNase genomic footprinting and DNase I hypersenstitivity clusters, histone modifications, enhancer and promoter associated histone markers, or transcription factor ChIP seq. Future work will require characterization of the functional relevance of NCAM1 for pressure overload LV hypertrophy.

Strengths and Limitations

This is the first GWAS of echocardiographic traits initially conducted initially in hypertensive African Americans, a group at particularly high risk of LVH and subsequent cardiovascular risk. We found one locus with strong statistical evidence in separate measures of wall thickness (IVST and PWT) and RWT. Given that hypertension is a potent “environmental” contributor to LVH, it is plausible that hypertensive-prone families in HyperGEN contributed to our ability to detect genes that are susceptible to the effects of blood pressure on cardiac structure. HyperGEN and GENOA had identical echocardiography protocols and readings were performed at a common reading center. We implemented a three-stage design to reduce Type I error. The family-based design reduces the potential for population stratification and we corrected for admixture. One limitation is the low power to detect a significant Stage II finding: we had 65% power to detect the observed effect and allele frequency assuming a P-value of 0.05. An additional limitation is the decreased genomic coverage of GWAS chips for African populations which makes replication and genome-wide significant P-values more difficult to achieve.^{21, 22} Given these limitations, the consistent, significant association of rs1436109 with PWT and correlated phenotypes is even more interesting.

Conclusions

We identified a novel locus that appears to be relevant for LV hypertrophy in hypertensive individuals. While it is unlikely that the allelic effects observed in this study will be useful in the immediate future as clinical predictors of LVH in individual patients, NCAM1 contributed to the burden of LV hypertrophy in these hypertensive African Americans and Caucasians. Further genetic studies to identify specific causal variants and functional studies can illuminate the mechanisms by which NCAM1 influences LV wall thickness.

Novelty and Significance.

What is Known?

• Cardiac hypertrophy is the common response of the heart to pathological stimuli, such as hypertension and obesity.

• Left ventricular (LV) mass and related echocardiographically determined traits are important predictors of cardiovascular outcomes including mortality, particularly in African Americans.

• Genetic factors are important determinants of LV mass and related cardiac traits.

• LV mass and other indices of cardiac hypertrophy are complex traits, determined in part by the cumulative effects of a very large number of genetic variants.

• Genome-wide association studies are currently the method of choice for identifying common genetic variants that influence expression of complex traits.

What New Information Does This Article Contribute?

• This study identified a variant in intron 1 of the NCAM1 gene that was associated with LV wall thicknessand relative wall thickness.

• The discovery of this variant occurred in hypertensive African American families, and replicated in Caucasian hypertensive families.

Elucidation of the genetic basis of LV hypertrophy may increase our understanding of the mechanisms involved in its pathogenesis. We identified a variant in the NCAM1 gene that accounts for some variation in LV wall thickness and relative wall thickness in cohorts of hypertensive individuals. Previous rat model studies have shown that NCAM1 is upregulated in myocytes in or adjacent to infarct regions and during cardiac remodeling in induced heart failure. The current study provides independent epidemiological evidence that suggests that NCAM may be biologically relevant for variation in cardiac structure. It is unlikely that the genetic association observed in this study will be immediately useful as a clinical predictor of LV hypertrophy in individual patients. Nevertheless, these findings provide a rationale for further genetic and functional studies to elucidate the mechanisms by which NCAM1 influences the expression of cardiac hypertrophy.

Non-standard Abbreviations and Acronyms

GWAS

genome-wide association study

HGNC

Human Genome Organisation Gene Nomenclature Committee

HWE

Hardy-Weinberg Equilibrium

IVST

intraventricular septal thickness

kb

kilobase i.e., a 1000 base segment of DNA

LV

left ventricular

LVID

left ventricular internal dimension

LVM

left ventricular mass

NCAM1

neural cell adhesion molecule 1 gene

NCBI

National Center for Biotechnology Information

PWT

posterior wall thickness

SNP

single-nucleotide polymorphism

TFBS

transcription factor binding site