A novel and lethal de novo LQT-3 mutation in a newborn with distinct molecular pharmacology and therapeutic response

Abstract

Background: SCN5A encodes the alpha-subunit (Na(v)1.5) of the principle Na(+) channel in the human heart. Genetic lesions in SCN5A can cause congenital long QT syndrome (LQTS) variant 3 (LQT-3) in adults by disrupting inactivation of the Na(v)1.5 channel. Pharmacological targeting of mutation-altered Na(+) channels has proven promising in developing a gene-specific therapeutic strategy to manage specifically this LQTS variant. SCN5A mutations that cause similar channel dysfunction may also contribute to sudden infant death syndrome (SIDS) and other arrhythmias in newborns, but the prevalence, impact, and therapeutic management of SCN5A mutations may be distinct in infants compared with adults.

Methods and results: Here, in a multidisciplinary approach, we report a de novo SCN5A mutation (F1473C) discovered in a newborn presenting with extreme QT prolongation and differential responses to the Na(+) channel blockers flecainide and mexiletine. Our goal was to determine the Na(+) channel phenotype caused by this severe mutation and to determine whether distinct effects of different Na(+) channel blockers on mutant channel activity provide a mechanistic understanding of the distinct therapeutic responsiveness of the mutation carrier. Sequence analysis of the proband revealed the novel missense SCN5A mutation (F1473C) and a common variant in KCNH2 (K897T). Patch clamp analysis of HEK 293 cells transiently transfected with wild-type or mutant Na(+) channels revealed significant changes in channel biophysics, all contributing to the proband's phenotype as predicted by in silico modeling. Furthermore, subtle differences in drug action were detected in correcting mutant channel activity that, together with both the known genetic background and age of the patient, contribute to the distinct therapeutic responses observed clinically.

Significance: The results of our study provide further evidence of the grave vulnerability of newborns to Na(+) channel defects and suggest that both genetic background and age are particularly important in developing a mutation-specific therapeutic personalized approach to manage disorders in the young.

Figure 1. Clinical and genetic profile of the proband.

A. Baseline ECG recording reporting a QTc of approximately 825 ms. Note 2∶1 AV block. B. Recording after infusion with IV lidocaine (see text). QTc is approximately 450 and 2∶1 block is absent. C. Chromatograms of DNA sequences: the proband but neither the mother nor father has a 4418 T = >G in SCN5A resulting in a substitution of cytsteine for phenylalanine at amino acid 1473 in Na_v1.5.

Figure 2. The F1473C mutation increases non-inactivating late current.

Whole-cell wild type (WT, A) and mutant (F1473C, B) Na⁺ channel currents at −10 mV (200 ms pulses applied from a −100 mV holding potential at 0.5 Hz). Shown are averaged raw TTX sensitive traces (WT, n = 11; F1743C, n = 31). C. Averaged WT and F1473C currents normalized to peak current and superimposed at high (left) and low (inset) gain (WT n = 11, F1473C n = 31). D. Bar graph representing mean +/− S.E.M. percentage of current remaining at 200 ms with respect to peak current for both WT and F1473C (WT n = 11, F1473C n = 31); * p<<0.001.

Figure 3. The F1473C mutation does not affect the voltage dependence of Na⁺ channel activation.

A. Mean current voltage relationships recorded in low external sodium solution (Methods S1) in response to 25 ms voltage pulses from −80 mV to +75 mV (5 mV increments) for WT (n = 13, open triangles) and F1473C (n = 7, closed squares). B. Mean activation curves obtained from data in A, for WT (open) and F1473C channels (filled) show no affect of the F1473C mutation on the voltage dependence of activation (see text for details).

Figure 4. The F1473C mutation has multiple biophysical consequences.

A. The F1473C mutation causes an 8.8±1.6 mV depolarizing shift in the steady state availability curve. Data shown for WT (n = 9, open triangles) and F1473C (filled squares, n = 25) channels. B. Mutation at F1473C speeds the time course of recovery from inactivation as measured at −100 mV following a 50 ms pulse to −10 mV. WT (open triangles, n = 5) and F1473C (closed squares, n = 5). C. Normalized WT (grey) and F1473C (black) currents in response to a ramp in voltage from −120 mV to +40 mV (see Methods S1) over the course of 1s reveal F1473C-enchanced inward current over a broad range of voltage. Shown are averaged TTX-sensitive currents for WT (light grey, n = 4) and F1473C (black, n = 11) channels. Note the inverted bell-shaped response of WT channels to this protocol reveals “window” of current for WT channels (see text for details).

Figure 5. Relative sensitivity of peak and late F1473C channel current to three different Na⁺ channel blockers.

A–C. Averaged normalized currents from F1473C channels evoked by 200 ms depolarizing currents to −10 mV at 0.5Hz in the presence and absence of 50 µM mexiletine (A, n = 5), 50 µM ranolazine (B, n = 15), and 10 µM flecainide (C, n = 5). Traces in the presence of each drug are shown in grey and indicated by an arrow in the inset. Drug-free traces are black. Note that the units and presentation reflect normalization to peak current under drug-free conditions. D. Bar graph summarizes fraction of peak (grey bars) and late current (black bars) blocked by each drug. Each drug inhibited late current preferentially over peak current (* p<<0.05). Inhibition of peak and late current was not significantly different among all drugs tested.

Figure 6. Mexiletine and Ranolaziine, but not Flecainde, correct mutation-induced voltage shift in steady state inactivation.

Steady state inactivation was measured with a 5s conditioning pulse (see Methods S1) followed by a brief −10 mV test pulse. A. Flecainide (n = 4) did not significantly alter V_1/2 or slope factor of the channel availability curve. B–C. Mexiletine (n = 5) and Ranolazine (n = 9) each caused significant hyperpolarizing shifts in the voltage-dependence of steady-state inactivation (Table 1).

Figure 7. Mexiletine and ranolazine inhibit F1473C channel activity during repolarization.

Inward currents were measured in response to a pulse to +20 mV for 100 ms followed by repolarization to −100 mV at −1.6 V/sec (negative ramp) (see Methods S1). A. F1473C channels, relative to WT channels, show a very large increase in non-inactivating current for the duration of the voltage pulse to +20 mV and also cause a large surge of inward current during the negative voltage ramp. B–C. Mexiletine (B) and ranolazine (C) inhibit the mutation-induced repolarizing inward current (Mexiletine block of repolarization phase peak current = 82.1+/−0.2%, n = 7; Ranolazine block of repolarization phase peak current = 83.1+/−0.2%, n = 4).

Figure 8. Simulation of the impact of the F1473C mutation and its modulation by mexiletine on the ventricular action potential.

A. Shown are the computed action potentials for the F1473C mutant (black trace) as compared to WT (blue trace) at 1Hz. Also shown is the contribution of F1473C late current alone to action potential prolongation, without changing the channel steady state availability (red trace). B. The simulations show the drug markedly rescues the channel function reducing the F1473C APD from 588 ms to 386 ms.