Four coding variants, which occurred in less than 0.3% of more than 18,000 control chromosomes (and less than 0.02% of more than 78,000 control chromosomes, if G844D is excepted), were identified in 178 LQT cases with no mutations in the 13 major LQTS genes. Among the genes associated with long QT syndrome, only
AKAP9 and
CALM2 are not included in the clinical exome of Illumina.
AKAP9 was associated with a single case of long QT syndrome [
22] and might rather be a QT interval modifier than a long QT causal gene [
23].
CALM2 mutations may result in neonatal bradycardia, cardiac arrest at a young age, a markedly prolonged QTc interval and eventually mild neurodevelopmental delay [
24,
25]. Only two patients of this
TRPM4 study were within the age range of
CALM2 mutations (Patients 2 and 4) but their QTc interval was never markedly prolonged as in
CALM2 mutations.
The
TRPM4 variants of this study are all heterozygous missense variants. Interestingly, one patients had in addition to LQTS, a conduction block (patient 1 had a Left Anterior Hemiblock) but he was 70 years old when the heart block appeared. This is consistent with Trpm4
−/− mice, which also present multilevel conduction blocks [
8] in addition to increased QT interval. The variant of patient 4 (p.G844D) was found in other cases with conduction blocks [
10,
21] or Brugada syndrome [
11]. Among these latter cases, one also had a prolonged QTc interval of 458 ms (patient 13 of reference 11).
In this study, comparing whole-cell currents recorded from WT and Val441Met and Arg499Trp
TRPM4 transfected HEK293 cells, showed densities significantly lower in both variants. The present study does not discriminate between a decrease in current density due to a modification of biophysical/regulatory properties of the ion channel, or a decreased protein expression as previously reported for other
TRPM4 mutants [
11]. However, there is a clear channel loss of function at the level of the membrane.
Until now, TRPM4 impact on cardiac activity was investigated only with regards to its contribution to the action potential (AP). According to its non-selective cation permeability (Na
+ and K
+), its opening is suspected to induce an outward repolarizing current in positive voltages but an inward depolarizing current in negative voltages. Since channel activity increases with membrane depolarization, it is suspected to be largely open at the upstroke and during the plateau of the AP. This is potentiated by the fact that TRPM4 is activated by internal Ca
2+, which increases during the AP. Altogether, one can predict that TRPM4 activation reduces phase 2 duration of AP and prolongs phase 3 counteracting repolarization. It seems that the effect on phase 2 is predominant, at least in mouse atria, since TRPM4 pharmacological inhibition by 9-phenanthrol reduces AP duration [
26]. Consistent with this result, Trpm4
−/− mice exhibit a shortened AP in atria, but also in ventricle [
26,
27]. Surprisingly, ECGs from Trpm4
−/− mice showed a prolonged QT interval, despite the reduction of ventricular AP duration [
8,
27]. The rationale for this counter-intuitive change might be that
Trpm4
−/−
mice also show an increased heart/body weight ratio, which is not due to cardiomyocyte hypertrophy but rather neonatal hyperplasia [
8]. QT duration depends on both ventricular AP duration and AP propagation within the ventricle. Thus, hyperplasia could increase the time for the excitation to reach the entire ventricles, and thus prolong the QT interval. Even if AP parameters are different between mice and human, the model of
Trpm4
−/− mice might be representative of heart disease related to
TRPM4 mutations in human. Indeed, mutations which induce loss of function may impact cardiac development as does
Trpm4 disruption in mice.
TRPM4 is not the first gene encoding a channel whose mutations result in LQT and developmental anomalies as it was already found in Andersen [
28] and Timothy [
29] syndromes. Unfortunately, contribution of the TRPM4 current to these phases of the AP was not evaluated yet in human. It was also not possible to estimate a putative cardiomyocyte hyperplasia in patients of this study. At least, it can be stated that none of the five TRPM4 variant carriers fulfilled the echographic criteria for hypertrophic cardiomyopathy.