Rebuttal from Francisco J. Arjona and Jeroen H. F. de Baaij

We fully agree with Funato and colleagues on the relevance of CNNM proteins for the maintenance of Mg2+ homeostasis (Funato et al. 2018). Both of our papers summarize the solid body of evidence demonstrating that CNNM2 is key for renal Mg2+ handling, while CNNM4 regulates intestinal Mg2+ transport. Nevertheless, we do not share the conclusions presented by Funato and colleagues suggesting that CNNMs are Na+/Mg2+ exchangers. Funato and colleagues mainly base their view on the use of the Magnesium-Green fluorescent probe in an immortalized cell line (HEK293) (Yamazaki et al. 2013; Hirata et al. 2014). The restrictions of the use fluorescent Mg2+ indicators, due to Ca2+ sensitivity and high dissociation constants, are widely described (Günther, 2006). Indeed, other groups using fluorescent Mg2+ probes could not show CNNM-induced Mg2+ fluxes or even measure Mg2+ influx against the Na+ gradient (Hardy et al. 2015; Sponder et al. 2016). Moreover, the Mg2+ efflux experiments of Funato and colleagues were reported to be performed by bathing the cells in a solution with 40 mM Mg2+, which will certainly surpass physiological intracellular Mg2+ concentrations (Yamazaki et al. 2013; Hirata et al. 2014). In contrast, when CNNM2 is overexpressed in HEK293 cells and Mg2+ efflux is measured under physiological conditions with the stable 25Mg2+ isotope, no Mg2+ efflux can be attributed to CNNM2 (Arjona et al. 2014). Furthermore, our group was the first to show CNNM-mediated Na+ influx that can be inhibited by a high extracellular Mg2+ concentration in 2011 (Stuiver et al. 2011). Based on these experiments we concluded that CNNM2 cannot act as Na+/Mg2+ exchanger because: (i) the inward Na+ current was also present when applying Mg2+-free pipette solution, (ii) the presumed exchanger did not function in the opposite direction, and (iii) the reversal potential that we measured was not in agreement with the theoretical reversal potential of a Na+/Mg2+ exchanger (Stuiver et al. 2011). Funato and colleagues repeated these electrophysiological experiments in similar conditions and found the same results (Yamazaki et al. 2013). Despite the discrepancies that we’ve mentioned, they find support in these experiments to claim Na+/Mg2+ exchange. In conclusion, the experimental evidence reported by Funato and colleagues does not support the notion that CNNM proteins are genuine Na+/Mg2+ exchangers, though they influence renal and intestinal Mg2+ transport. In line with our claim that CNNM proteins are not Na+/Mg2+ exchangers, Funato and colleagues write that: “it still remains uncertain whether they are genuine exchangers by themselves or cooperatively function with one or several further proteins” (Funato et al. 2018). In this regard, these authors allude to the need to clarify whether CNNM proteins may interact with the well-characterized Na+/Mg2+ exchangers of the SLC41 family (Kolisek et al. 2012; de Baaij et al. 2016).