A potent voltage-gated calcium channel inhibitor engineered from a nanobody targeted to auxiliary Ca_Vβ subunits

Essential revisions:

1) It is difficult to make sense of results presented in Figure 3B and C without more stage-setting. Specifically, the absence of an effect of nb.F3-Nedd4L on Ca_Vβ subunit expression runs counter to the expectation that Nedd4L, given its function as a ubiquitin ligase (and the fact that it binds directly to Ca_Vβ), might cause Ca_Vβ degradation. This becomes clear in subsequent Results sections, but it would be helpful to provide more context by setting up these earlier experiments as a test of whether Nedd4L acts broadly through degradation or via some other mechanism. The absence of an effect of nb.F3-Nedd4L on Ca_Vβ subunit expression would then support the sub-conclusion that nb.F3-Nedd4L acts through a mechanism independent of ubiquitin-dependent degradation. It may be a minor point, but there's no good reason to be coy about the mechanism if doing so might leave some readers scratching their heads.

We thank the reviewer for the constructive feedback. In the original manuscript, we prefaced the results presented in Figure 3 with a reference to our previous work which demonstrated that targeting the catalytic domain of the E3 ligase CHIP to YFP-tagged potassium or calcium channels using a GFP-nanobody selectively reduced their surface density without promoting degradation. In response to the reviewer’s concern we have changed the order of presentation in Figure 3 to better set the stage to communicate the results. Principally, we now present our examination of the impact of Ca_V-aβlator on Ca_Vβ expression first (original Figure 3A-C is now swapped with Figure 3D-F in the revised manuscript) and have added the following sentence:

“Given the classical role of E3 ubiquitin ligases in mediating degradation of target proteins, we first assessed if nb.F3-Nedd4L affected total Ca_Vβ expression (Figure 3A, B).”

2) A cartoon showing the overall mechanism of action of Ca_V-aβlation would be helpful.

We agree and have included such a cartoon in Figure 5H.

3) Interestingly, in addition to localizing to the plasma membrane, β1 and β3 subunits, but not β2 or β4 subunits, also showed substantial accumulation in nuclear membranes. Do the authors have any thoughts on what causes this differential distribution and how it might be exploited in the context of their system?

We agree with the reviewer that differential targeting of Ca_Vβ isoforms to distinct sub-cellular compartments is interesting. However, we wish to emphasize that these are exemplar images obtained from overexpressing Ca_Vβs in a heterologous system. While the approach was sufficient for us to confirm binding of nanobodies to Ca_Vβs inside cells, we believe it is prudent to not over-interpret the potential basis or functional relevance of the differential Ca_Vβ distribution from these data. While Ca_Vβ subunits have been reported to have differential distributions and binding partners, that is not the focus of our current work.

4) Specific concerns about Figure 5. The text states that the data in Figure 5 show that the infection of cardiomyocytes with F3-Nedd4L causes the ubiquitination of Ca_V1.2 and Ca_Vß (I agree) and causes the channel complex to re-locate away from dyads and to become associated with Rab7. The latter appears unconvincing: the co-localization with RyR2 in the control myocytes is not robust (Figure 5C) and the distribution pattern of Ca_V1.2 in the F3-Nedd4L infected cells is very different in Figure 5C (a scattered pattern, which is supposed to show movement away from dyads) and in Figure 5D (concentrated near the sides of the myocyte, which is supposed to show association with Rab7). These imaging data should be removed.

We thank the reviewer for this comment and have worked to make the data presented in Figure 5 clearer and more compelling. We have included magnifications of each image using a high-zoom box, that more clearly conveys the differential co-localization of α_1C/RyR2 and α_1C/Rab7 under the different conditions. This change has been described in the legend to Figure 5. Together with the quantitative Pearson’s correlation coefficient analyses, we believe these images affirm our finding that Ca_V-aβlator predominantly directs Ca_V1.2 to Rab7-positive late endosomes.

5) The data support the notion that the nanobody interacts with an epitope conserved among Ca_Vβs, thus allowing for a single construct with pan-Ca_V β binding. This is a nice feature for the first pass of experiments but much of the downstream utility/promise for the approach would require isoform specific nanobody development. While this is obviously theoretically possible, its notable that despite using two Ca_vβ isoforms for nanobody production, the best performing hit is nonselective. The authors could do better in pointing out this area of future optimization for the reader.

We agree with the reviewer that development of isoform-specific nanobodies would be important to realize the full potential of this approach. From our llama immunization we have obtained a number of nanobodies and are systematically working our way through to find potential isoform-specific variants. This is ongoing work that is beyond the scope of the current manuscript. We have added the following sentence to address this important future direction:

“However, this capability may be readily achieved with Ca_V-aβlators directed towards particular Ca_Vβ isoforms. A challenge to realize this possibility is the development of Ca_Vβ isoform-specific nanobodies which should be feasible given that there is sequence divergence among Ca_Vβs outside the conserved src homology 3 (SH3) and guanylate kinase (GK) domains.”

6) The results are presented in a way that mostly ignores complexities of likely importance. In particular, if this technique is to be used either therapeutically or as an investigative tool, one would want to know something about the time course of the effect and about whether it would be possible to regulate the extent of action of the nanobody complex. It would be a major undertaking to obtain such information. Nonetheless, it would useful to discuss these potential complications.

We concur. We have integrated these points into our Discussion:

“For this purpose, it may be desirable to generate Ca_V-aβlator versions whose time course and extent of action could be tuned by either a small molecule or light. Indeed, this is a focus of our ongoing work.”