Transcriptomic and genetic profiling in a spontaneous non-human primate model of hypertrophic cardiomyopathy and sudden cardiac death

Fig. 1. Gross pathologic images of rhesus macaque left ventricles. Representative gross pathologic images of a four-year-old male (A), a six-year-old female (B), and a 12-year-old female (C) rhesus macaque affected with severe HCM, and a nine-year-old cardiovascularly normal female control (D) rhesus macaque are displayed (n = 4). Hearts were cross-sectionally cut midway between the apex and base of the heart at the level of the papillary muscles in the LV. Individuals A and C were obtained following a spontaneous SCD event. Note severe LVH and complete luminal obliteration in the HCM-affected individuals. Scale bar = 1 cm. RVPW = right ventricular posterior wall, IVS = interventricular septum, LVPW = left ventricular posterior wall, HCM = hypertrophic cardiomyopathy, LV = left ventricle, SCD = sudden cardiac death, LVH = left ventricular hypertrophy.

Fig. 2. RNA-Seq principal component analysis plot of rhesus macaque LV tissue. Plotted are the results of similarity within and between groups from a principal component analysis (PCA) of five HCM-affected (red) and seven HCM-unaffected (blue) female (circle) and male (triangle) rhesus macaque gene expression profiles from RNA-Seq (n = 12). PC = principal component.

Fig. 3. Volcano plot of DEGs in HCM-affected LV tissue. Illustrated is the global expression profile of upregulated (red), downregulated (blue), and not significant (grey) genes in the LV posterior wall tissues of five HCM-affected and seven HCM-unaffected rhesus macaques (n = 12). Gene names of the 20 top-most statistically significant DEGs are displayed. HCM = hypertrophic cardiomyopathy, DEGs = differentially expressed gene(s).

Fig. 4. Summary of total identified and shared rhesus macaque and pediatric and adult human HCM DEGs. The total number of identified down- (blue) and upregulated (red) LV DEGs in HCM-affected rhesus macaques identified via RNA-Seq are presented (A). Venn diagrams of shared HCM DEGs between rhesus macaque and pediatric human (PH; solid color) and adult human (AH; dot-patterned) from previously reported datasets are displayed (B and C, respectively). LV = left ventricle, DEGs = differentially expressed genes, PH = pediatric human, AH = adult human, RNA-Seq = RNA-Sequencing, HCM = hypertrophic cardiomyopathy.

Fig. 5. Top 20 enriched GO and KEGG pathway terms for LVPW DEGs between HCM-affected and -unaffected rhesus macaques. Enrichment plots for gene ontology (GO) biological process (BP) term analyses are presented for down- and upregulated DEGs (A and B, respectively). Enrichment plots for GO molecular function (MF) term analyses are presented for down- (C) and upregulated (D) DEGs. Enrichment plot for GO cellular component (CC) terms for upregulated DEGs is presented (E). KEGG30–32 pathway enrichment plots are provided for down- and upregulated DEGs (F and G, respectively). The y-axis on each graph lists identified enriched terms, whereas the gene ratio is denoted on the x-axis. The size of each individual point depicts the total number of genes binned to a given enriched term; high or low PAdjusted-values are represented by red or blue coloring, respectively. Enrichment terms were obtained from the gene expression profile data of five and seven HCM-affected and -unaffected rhesus macaques, respectively (n = 12). GO = gene ontology, LVPW = left ventricular posterior wall, DEG = differentially expressed gene, HCM = hypertrophic cardiomyopathy, BP = biological process, MF = molecular…

Fig. 6. Manhattan plot of global genomic PAllelic-value < 0.05 variants. The total PAllelic-value < 0.05 variants (n = 2,005,816) of 15 HCM-affected and 21 cardiovascularly normal geriatric rhesus macaques are plotted along their corresponding genomic position (x-axis; n = 36). The − log10(PAllelic-value) are plotted along the y-axis, where the blue- and red-dashed lines represent PAllelic-value = 0.00001 (− log10[0.00001] = 5) and PAllelic-value = 2.49E-8 (− log10[2.49E-8] = 7.60) significance thresholds, respectively.

Fig. 7. Polygenic modeling of HCM status in rhesus macaques. Summary of hyper parameters (pve, pge and n_gamma) of five independent runs of BSLMM indicating model convergence among runs (A). Polygenic score distributions of the control (blue) and affected (red) groups (B). Accuracy summary of leave-one-out cross validation (C). Bar graphs represent the number of individuals in each group, and colors indicate the proportion of individuals correctly or incorrectly (blue or red, respectively) predicted to be affected or unaffected by HCM. Polygenic modeling was conducted utilizing the whole-genome sequencing data of 15 and 21 HCM-affected and -unaffected rhesus macaques (n = 36). Random (alpha), fixed (beta), and sparse effect (gamma/PIP) from the five runs were averaged, then polygenic scores (PGS) from the identified eight intergenic and seven gene variants with PIP > 0 (QSOX1 [1], ENSMMUG00000055166 [1], SLC26A3 [1], MYO3A [1], TMEM135 [1], COX10 [1], MAPK4 [1]; were estimated for each animal. Polygenic scores were calculated as: the effect size for each variant: beta × gamma + alpha, multiplied by the number of variant alleles, then summed among sites. pve = percent variation explained,…