Pore mutation N617D in the skeletal muscle DHPR blocks Ca²⁺ influx due to atypical high-affinity Ca²⁺ binding

Essential revisions:

Two main areas were identified as points needed to be addressed by the authors to strengthen the conclusions of the work.

1. The conclusions that the inward/outward currents observed in Figure 2 are not mediated by N617D channels, while the inward Li+ currents in the absence of Ca²⁺ in Figure 3 are mediated by N617D channels should be directly verified with the use of a DHP antagonist.

In agreement with the reviewers’ request, we performed additional patch-clamp experiments in the presence of 10 µM of the 1,4-DHP antagonist nifedipine in the external recording solution to directly verify that the inward / outward currents in Figure 2 are not mediated by DHPR(N617D) channels, whereas the inward Li⁺ currents in the absence of Ca²⁺ in Figure 3 are mediated by DHPR(N617D). Both our claims could be verified by these experiments and according text blocks have been added to the Results (lines 189-198; 226-231; 235-237) and Material and Methods (lines 490-493) sections. Additionally, a new figure panel (Figure 2c) and a new figure (Figure 4) with the corresponding legends have been included.

2. The final section of the paper presents a conceptual model for how the apparent increase in Ca binding affinity severely reduces Ca flux (i.e. "block") and at the same time Ca occupancy of the EEEE motif preserves selectivity over monovalent cations. The key concept is that N617D enhances the Ca binding at the DCS site (0.4 μm affinity) which lies in the pore vestibule external to EEEE. It is proposed that this high-affinity binding at DCS impedes Ca transition to the EEEE site. A potential problem with this model is that if the DCS → EEEE translocation of Ca is too slow to allow "double occupancy" of EEEE by Ca (and thereby reduce EEEE affinity to mM range and promote flux) then how does the EEEE site get occupied with single Ca? The on-rate for initial block of Li current when adding Ca would be very slow. Quantitative simulation would be very helpful to demonstrate it is possible for the DCS → EEEE translocation of Ca to be slow enough to prevent a detectable Ca conductance, and yet at the same time be sufficiently fast to establish Ca block of Li current in a reasonable amount of time.

In order to address these queries, we performed molecular dynamics (MD) simulations in combination with an enhanced sampling technique (described in the Materials and methods section of the revised manuscript; lines 501-542) to estimate the differences in free energy profiles of Ca²⁺ ions passing through the selectivity filters of wt DHPR and mutant DHPR(N617D). We found that in wt DHPR immediately after the equilibration phase, two Ca²⁺ ions compete for the EEEE locus. Consequently, Ca^{2+ –} Ca²⁺ repulsion occurs in this doubly occupied EEEE locus, so that the first of the two Ca²⁺ ions is pushed towards the cytosolic side. Moreover, metadynamics simulations capturing the passage of Ca²⁺ ions through the pore revealed a significantly smaller energy barrier for the wt DHPR selectivity filter compared to mutant DHPR(N617D).

MD simulations on mutant DHPR(N617D) revealed a different picture. Already 1 ns after the equilibration phase, we find one Ca²⁺ ion stabilized in the EEEE locus and the other bound to the DCS locus. This shows that already in a very short equilibration phase, the Ca²⁺ ions reach and occupy the EEEE and DCS loci, and since ions cannot pass though the selectivity filter regions anymore, this result in complete occlusion of the pore. This additional stabilization at the DCS locus that a Ca²⁺ ion experiences while moving towards the cytosolic side is also reflected in the ~ 8-times higher energy barrier compared to wt DHPR. The slow detachment and thus slow migration of the Ca²⁺ ions towards the cytosolic side, as observed in our simulation study, suggests occlusion of the pore without the force applied during simulations to pull the Ca²⁺ ions through the pore.

Observations from the simulation experiments are presented in a new figure (Figure 6, with 2 videos as figure supplement) and the Discussion section (lines 356-365, and 373-383).

As a consequence of these new insights, we removed the sentence “Additionally, undisturbed continuous high affinity binding of Ca²⁺ to the EEEE locus results in complete occlusion of the DHPR(N617D) pore, even for monovalent cations, as demonstrated above” (lines 435-437 of the original manuscript) from the Discussion section. Misleadingly it suggested that divalent and monovalent cations can only be blocked at the continuously Ca²⁺- occupied EEEE locus in mutant DHPR(N617D), a statement for which no experimental evidence exists. Considering the tight dimension of the pore around the DCS locus, it is very likely that cations are blocked already at the Ca²⁺- occupied DCS locus. This mechanism would allow a very fast on-rate for the initial block of Li⁺ currents when adding Ca²⁺ ions to the external solution, independent of the speed of the DCS → EEEE translocation, which can be slow enough to prevent detectable Ca²⁺ conductance. Detailed, quantitative MD simulations with reliable time scales and interaction energies to predict if the DHPR(N617D) channel pore can be blocked via the Ca²⁺- occupied DCS locus is a challenging aim, but considering the massive need of computing resources would require an independent study.