Decreased FAM13B Expression Increases Atrial Fibrillation Susceptibility by Regulating Sodium Current and Calcium Handling

Figures

• 

  Figure 1. Identification of the AF Causal Gene and Candidate SNP (A) Partial regression plot showing the effect of the genome-wide association study (GWAS) single-nucleotide polymorphism (SNP) rs2040862 on left atrial appendage (LAA) expression of the FAM13B gene in African American subjects. Blue dots are subjects homozygous for the major allele, green dots are heterozygotes, and red dots are homozygous for the minor allele (r2 = 0.78 Pearson’s correlation, q value = 0.01). (B) Expression quantitative trait loci (eQTL) P values and linkage disequilibrium (LD) relationship (color scale) with the GWAS SNP rs2040862 in the region around FAM13B in African American subjects, identifying rs17171731 as the top eQTL SNP (P = 2.63 × 10−5). The gray-shaded area represents the longest FAM13B gene (transcribed from left to right) in Ensembl, but the LAA transcript starts ∼19 kb proximal to rs17171731. (C) Zoom in on FAM13B cis-eQTL for African American subjects with LD relationships for rs17171731 (top). The SNP rs17171731 occurs in a DNaseI hypersensitivity peak, whereas the atrial fibrillation (AF) GWAS SNP rs2040862 does not (bottom, Roadmap Epigenomics Project data for human fetal heart). (D)…

• 

  Figure 2. CRISPR Editing of rs17171731 (A) Sanger sequencing of complex insertion into rs17171731 site via nonhomologous end joining (NHEJ). Bold italic font, PAM sequence opposite strand; bold font, single guide RNA (sgRNA) target sequence; green highlight, rs17171731 reference allele; blue highlight, inserted base pairs. (B) Strategy to select for homology-directed repair (HDR) co-editing. An enhanced green fluorescent protein (eGFP) expression plasmid was mutated to include 2 stop codons and stably transfected into stem cells (top). Sequence of repaired eGFP after sgRNA and donor oligo transfection to correct stop codons and alter the PAM sequence to prevent re-editing (bottom). Gray highlight, sgRNA target sequence; bold italic font, PAM sequence; red highlight, Y40∗ and C49∗ stop codon mutations; underlined sequence, single-strand HDR repair oligo donor; green highlight, corrected eGFP sequence and altered PAM site. (C) Fluorescent activated cell sorting after eGFP and rs17171731 HDR co-editing, yielding 0.69% eGFP+ cells used for clonal growth. (D) Sanger sequencing of the rs17171731 region in the WT stem cells (top, homozygous reference allele) and co-edited cells (bottom,…

• 

  Figure 3. Functional Consequences of FAM13B Knockdown (A) Quantitative reverse transcriptase-polymerase chain reaction verification of small interfering RNA (siRNA) FAM13B knockdown in iCell cardiomyocytes normalized to ACTC1 messenger RNA (mRNA) relative to the scrambled (Scr) siRNA (mean ± SD, P < 0.001 by 2-tailed t-test, n = 4-6 biological replicates [each from technical triplicates] from 2 independent experiments, 1 Scr outlier was removed using the Grubbs test). (B) The strongest empirical network for FAM13B KD identified by Ingenuity Pathway analysis, which included the SCN2B gene encoding a subunit of the voltage-gated sodium channel. Green indicates repression and red indicates induction upon FAM13B knockdown, with the color saturation indicating the fold effect. (C) Representative sodium current (INa) traces and voltage-clamp protocol. A 100-ms pre-pulse to −120 mV was used from a holding potential of −60 mV. Depolarizing voltage steps from −80 to +70 mV were applied in 10-mV increments. (D) Summary INa-V plot for scramble (n = 20) vs FAM13B knockdown (n = 23) recordings. (E) Inactivation protocol and representative INa traces. As in (C), resting potential was −60 mV, with a…

• 

  Figure 4. FAM13B Localization and Recombinant Protein (A–D) FAM13B-GFP fusion protein expression in iCell cardiomyocytes green fluorescent protein (GFP) fluorescence showing expression in a sarcomere pattern and on the plasma membrane (A). (B) α-actinin immunofluorescence (red) showing Z-disc localization with the transfected cell surrounded by nontransfected cells. (C) Hoechst 33342 staining of nuclei. (D) Merged image, showing FAM13B-GFP expression overlaps with or is adjacent to the Z-disc marker protein α-actinin; ×60 objective lens, scale bar shows 20 μm. (E) rhFAM13B purification. 6 Histidine (6xHis)-maltose binding protein (MBP)-hFAM13 fusion protein in bacterial lysate was eluted from an amylose column (left panel), cleaved with 6xHis-Tobacco Etch Virus (TEV) protease releasing the rhFAM13B from the 6xHis-MBP affinity tags (center panel), and purified by flow through in the Ni resin column (right panel). (F) Rho GTPase-activating protein (RhoGAP) activity assay (in singleton assay) showing background GTPase activity of the p50 RhoGAP alone (striped bar) and for each of the 4 Rho proteins alone (white bars), whereas p50 RhoGAP increased the GTPase activity of 3 of 4 Rho proteins…

• 

  Figure 5. Heart Structure Function and Basal ECG of WT and Fam13b KO mice (A) Ejection fraction (left) and left ventricular mass (right) in male wild-type (WT) (n = 5) and knockout (KO) (n = 8) mice (means shown, Student’s t-test). (B) Heart (left) and heart weight normalized to tibial length (right side) in male (top, 14 WT and 14 KO) and female (bottom, 7 WT and 4 KO) mice (means shown, Student’s t-test). (C) Gross appearance of representative perfused and fixed hearts ex vivo (Males, left; females, right, WT, top; KO, bottom). (D) Four-chamber fixed section stained with hematoxylin-eosin (H&E) and Masson’s Trichrome. (E) Basal electrocardiogram (ECG) readings in male (left, 17 WT and 19 KO) and female (right, 11 WT and 9 KO) WT (black symbols) and KO (red symbols) mice, with each symbol representing a biological replicate. Data were analyzed by Mann-Whitney tests, with median values shown (∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001). P = duration of P-wave; PR = duration of PR interval; QRS = duration of QRS peak; QT = duration of QT interval; QTc = duration of QT interval corrected for a standard heart cycle; RR = duration of heart cycle via RR interval.

• 

  Figure 6. Intracardiac Pacing-Induced Arrhythmia (A) Representative surface electrocardiogram (ECG) and atrial intracardiac electrogram from a female Fam13b knockout (KO) mouse showing burst pacing followed by an occurrence of atrial fibrillation (AF) with RR variability and P-wave disruption (SR, sinus rhythm phase with normal P waves and double-headed blue arrows showing the regular RR interval; Stimulus, showing the programmed electrical stimulation; AF, the atrial fibrillation phase with rapid and chaotic atrial activation with lack of regular P waves, the double-headed red arrows showing RR interval variability). AF events stopped spontaneously followed by normal SR. (B) Poincare plot showing representative RR variability in a single female Fam13b KO before (black symbols) and after pacing-induced AF (red symbols), with the median RR variability 18 times larger in AF (P < 0.001, Mann-Whitney test). (C, D) Representative surface ECGs and intracardiac recordings from 2 male Fam13b KO mice showing instances of ventricular arrhythmia due to premature ventricular contractions (PVC).

• 

  Figure 7. Pacing-Induced AF and VA (A) Susceptibility to atrial fibrillation (AF) induction per mouse for male and female wild-type (WT) (white bars) and Fam13b knockout (KO) (red bars) mice under basal and isoproterenol (IsoP) conditions. (B) Susceptibility to AF per trial for all mice in each group. (C) Susceptibility to ventricular arrhythmia (VA) per mouse for male and female WT and Fam13b KO mice under basal and IsoP conditions, as well as in both sexes and conditions combined. For all plots, (N/N) represents the number in arrhythmia/the total number. The numbers in arrhythmia vs not in arrhythmia were analyzed using Fisher’s exact test, with the relative risk (for KO vs WT) and P values shown below each graph with P values <0.05 shown in bold.