Contrasting Ionic Mechanisms of Impaired Conduction in FHF1 and FHF2 Deficient Hearts

Proper function of the cardiac sodium channel is essential for normal membrane excitability and conduction in the heart. Fibroblast growth factor homologous factors (FHFs) 1–4 bind to the C-terminal domains of voltage gated sodium channels, regulating channel trafficking and gating properties.¹ FHF2 is the predominant family member expressed in the mouse ventricle, with knockout mice displaying cardiac conduction disease.² In contrast, FHF1 is the predominant family member expressed in mouse atria (Figure A-D) and human atrial and ventricular myocardium³ (Figure E). Clinically, mutations in FHF1 have been linked to both idiopathic ventricular tachycardia⁴ and Brugada syndrome⁵. Furthermore, reductions in FHF1 expression have been observed in diseased left atrial and left ventricle tissue³, while FHF2–4 levels are not significantly changed (Figure E). Thus, dysregulated FHF1 activity may play a mechanistic role in both heritable and acquired arrhythmic disorders. Accordingly, using complementary in vivo murine models and human induced pluripotent stem cell derived cardiomyocytes (iPSC-CMs), we sought to determine the distinct consequences of FHF1 deficiency on cardiac electrophysiology.

Figure. FHF1 and FHF2 have distinct effects on cardiac sodium channel function.

(A–D) Fhf1 is highly enriched in the mouse atria. (A) Pseudo-colored RNA-Scope of a wildtype heart showing Fhf1 is predominantly expressed in the atria. (B) Higher magnification views of selected left atrial and left ventricular regions (n = 1) (C) Fhf1 is highly enriched in purified atrial myocytes compared to ventricular myocytes (n = 3 atria; n = 10 ventricles). (D) Fhf1-4 transcript abundance in murine heart by cardiac chamber (n = 3 atria; n = 10 ventricles). (E) FHF1 is the predominant FHF in the human heart. FHF1–FHF4 transcript abundance in human cardiac chambers under normal and heart failure conditions (n = 9–12 patient samples per cardiac chamber/condition).³ (F) Increased temperature leads to conduction slowing in Fhf1^KO hearts. P-wave (top panel) and QRS duration (bottom panel) plotted against temperatures at 37°C or 43°C. Fhf1^KO mice demonstrated P-wave prolongation at elevated temperatures; QRS duration was unchanged. (Fhf1^KO n = 10; male:female 5:5; Fhf1^WT n = 11; male:female 6:5). (G, H) Measurement of sodium current (I_Na) in Fhf1^KO and Fhf2^KO murine atrial myocytes. (G) I_Na density as a function of voltage. Fhf1^KO atrial myocytes have significantly reduced peak I_Na density compared to Fhf1^WT and Fhf2^KO cells. (H) Cardiomyocyte voltage-gated sodium channel V_1/2 steady-state inactivation. Available I_Na expressed as fraction of maximal I_Na. Voltage dependence of inactivation does not significantly differ between Fhf1^KO, Fhf2^KO, Fhf1^WT in atrial myocytes. (n = 10–12 cells/genotype). (I, J) Measurement of I_Na in human iPSC-CMs. (I) I_Na density as a function of voltage. FHF1^KO and FHF1,2^KO iPSC-CMs have significantly reduced peak I_Na density. FHF2^KO iPSC-CMs do not differ from Fhf1^WT iPSC-CMs. (J) FHF2^KO and FHF1,2^KO iPSC-CMs show a hyperpolarizing shift in inactivation compared to FHF1^KO and Fhf1^WT iPSC-CMs (n = 14–16 cells/genotype). *p-value<0.05, **p-value<0.01, ***p-value<0.001, ****p-value<0.0001. Where normally distributed, Student’s t-test or one-way ANOVA used with Tukey’s test for individual comparisons performed. Otherwise, non-parametric testing performed using Wilcoxon signed-rank test for comparisons between paired samples; Mann–Whitney U test used for comparison of independent variables. RA, right atrium; LA, left atrium; Ao, aorta; RV, right ventricle; LV, left ventricle; HFrEF, heart failure with reduced ejection fraction

All protocols conformed to the Association for the Assessment and Accreditation of Laboratory Animal Care and the NYU Grossman School of Medicine Animal Care and Use Committee. The data that support the findings of this study are available from the corresponding author on reasonable request.

We first determined the effects of Fhf1 deficiency on cardiac conduction parameters using electrocardiography (ECG). Adult Fhf1 knockout (Fhf1^KO) mice did not show differences in ECG parameters under baseline conditions. Previously, we showed that Fhf2 knockout (Fhf2^KO) mice demonstrate atrial and ventricular conduction abnormalities when challenged with hyperthermic stress.² At elevated temperatures, Fhf1^KO mice demonstrated P-wave prolongation while Fhf1^WT mice showed no change (Figure F upper panel; Fhf1^KO 18.24ms [17.12, 18.59] versus Fhf1^WT 16.31ms [15.87, 16.83]). No atrial arrhythmias were observed. Fhf1^KO mice did not exhibit QRS prolongation with temperature elevation (Figure F lower panel; Fhf1^KO 11.89ms [11.60, 13.11] versus Fhf1^WT 11.67ms [11.08, 12.10]). The susceptibility to atrial conduction slowing in both Fhf1^KO and Fhf2^KO mice suggest non-redundant functions of these two Fhf family members on sodium channel behavior in atrial myocytes.

To explore distinct regulatory functions of FHF1 and FHF2 on the cardiac sodium channel in atrial myocytes, we performed whole cell patch clamp. Fhf1^KO atrial myocytes showed diminished peak sodium current (I_Na) density at −40mV compared to Fhf1^WT and Fhf2^KO atrial myocytes. There was no difference in peak I_Na density between Fhf1^WT and Fhf2^KO atrial myocytes (Figure G; Fhf1^KO 32.07pA/pF ± 1.80 vs Fhf1^WT 43.36 pA/pF ± 2.97 vs Fhf2^KO 41.26 pA/pF ± 2.32; n = 10–12 cells/genotype). Fhf2^KO atrial myocytes demonstrated a nonsignificant hyperpolarizing shift in V_1/2 steady state inactivation compared to Fhf1^KO and Fhf1^WT atrial myocytes, reflecting the lower abundance of Fhf2 in atrial compared to ventricular myocytes.⁴ (Figure H; Fhf1^KO −89.40mV ± 1.23 vs Fhf1^WT −89.34 mV ± 1.06 vs Fhf2^KO −92.58mV ± 1.13; n = 10 cells/genotype).

To investigate whether functional differences of FHF1 and FHF2 are conserved in human cardiomyocytes, we generated iPSC-CMs engineered with single and double knockout of FHF1 and FHF2. iPSC-CMs express both FHF1 and FHF2 prior to genetic modification (data not shown). Loss of FHF1 led to significantly reduced peak I_Na density at −25mV compared to FHF^WT and FHF2^KO, with no additional reduction in peak I_Na density in FHF1,2^KO iPSC-CMs (Figure I; FHF1^KO 77.90 pA/pF ± 11.5 vs FHF2^KO 146.5 pA/pF ± 20.2 vs FHF1,2^KO 73.86 pA/pF ± 7.57 FHF^WT 160.7 pA/pF ± 20.4; n = 14–16 cells/genotype). Loss of FHF2 led to a hyperpolarizing shift in V_1/2 steady state inactivation compared to FHF^WT and FHF1^KO, with no additional shift in inactivation in FHF1,2^KO iPSC-CMs (Figure J; FHF1^KO −78.0 mV ± 0.86 vs FHF2^KO −92.9 mV ± 1.21 vs FHF1,2^KO −93.1 mV ± 0.85 vs FHF^WT −80.8 mV ± 1.08; n = 14–16 cells/genotype). Absence of synergistic effects on I_Na in FHF1,2^KO iPSC-CMs indicates that FHF1 and FHF2 have functionally distinct effects on the cardiac sodium channel in both mouse and human cardiomyocytes.

In summary, our data indicate that FHF1 and FHF2 are key sodium channel regulatory proteins that influence excitability and conduction through different ionic mechanisms. FHF1 reduces sodium conductance through its effects on peak I_Na density, whereas FHF2 exerts its major effects on conduction through changes in sodium channel inactivation. These results indicate that species-specific differences in cardiac sodium channel function are influenced by which FHF family member is dominant. Further, given that FHF1 expression is significantly reduced in failing human atrial and ventricular tissue, our data suggest that FHF1 deficiency, through its effects on sodium currents, may be clinically operative and contribute to conduction slowing and the heightened arrhythmia burden observed in diseased hearts.³

Nonstandard Abbreviations and Acronyms.

FHF

fibroblast growth factor homologous factor

FGF

fibroblast growth factor

iPSC-CM

induced pluripotent stem cell derived cardiomyocytes