The Long QT Syndrome (LQTS) is a multi-factorial disorder that predisposes to life-threatening cardiac arrhythmias. Both hereditary and acquired (mostly drug-induced) forms have been identified. However, recently, it has become clear that the interaction of multiple acquired and genetic factors (disease modifiers) play an important role in differentiating genotype into a continuous spectrum of clinical or subclinical phenotypes. The genotype-phenotype correlation thereby remains very unpredictable in asymptomatic patients, raising important concerns for clinical practice and also for drug development. Hence, a detailed analysis of clinical, genetic and 'in-vitro' functional and biophysical data remains important to further elucidate the mechanisms underlying this complex disease. We report how a single hERG mutation c.1039C>T (p.Pro347Ser) can give rise to two clinically distinct LQTS phenotypes. In one patient a reversible LQTS phenotype was found that was triggered by a combination of QT-prolonging compounds compatible with an acquired LQTS. In contrast, the same mutation also caused an overt congenital LQTS in a multi-generational family. Secondly, we investigated the structure-function relationship of the p.Pro347Ser-containing region within the hERG protein. Our data show that the charges of a KIKER cluster are important for channel gating through electrostatic interactions with the channel core. The introduction of more negative charges in this region (EIEEE instead of KIKER) resembled the functional phenotype of the p.Pro347Ser mutation. Likely, this pro-ser substitution alters the local protein flexibility, thus also the way the KIKER cluster modulates hERG function, which might be responsible for the LQTS effects in both cases. Hence, multi-level research provides a unique and differentiated insight into the hereditary and acquired pathogenic mechanisms of this intriguing multi-factorial disorder.