Electrical synaptic transmission requires a postsynaptic scaffolding protein

Only the following revisions need to be implemented:

1. The conclusion that ZO1 plays a role in gap junction regulation/function is not well supported by the data presented here as the authors cannot rule out that fact that gap junctions simply fail to form in ZO1 mutants. The authors should revise their conclusions in results and discussion to tone down the claim.

We appreciate the Reviewers’ concern and suggestions. We realize that we were not sufficiently clear when describing our perspective on the possible functional roles of ZO1 at Mauthner cell mixed synapses. Here we address the use of ‘function’ and ‘regulation’ in the text:

As stated in the first sentence of the Discussion section, we propose that ZO1 ‘is essential for the formation and function of electrical synapses’, not only function. Why? As pointed out by Reviewer 1, ZO1 could have a role in the formation of gap junctions. However, the existence of abortive synaptic structures in ZO1b-/- mutant fish (see Figure 1H-J) indicates that the neurons are attempting to construct and stabilize gap junction channels at the membrane, suggesting the existence of junctional structures. This finding raises the possibility that ZO1 could be functionally required for efficient insertion of channels at gap junctions. Such a possibility is consistent with previous data at these synapses (Flores et al. 2012) and with the proposed role of ZO1 at Cx43-containing gap junctions (Rhett and Gourdie, 2012). Finally, as a key scaffold, ZO1 might serve various functions, including both the formation and maintenance of functional channels, as well as potentially scaffolding other interactors such as kinases/phosphatases. Given all these possibilities, we cautiously chose to argue that ZO1 is likely to be involved in the ‘formation and function’ of electrical synapses.

We only discuss ‘regulation’ on within the Discussion, where we contrast differences in the role of ZO1 at Cx43 vs. Cx36-related gap junctions. In this section, we discuss potential regulatory roles of ZO1 at electrical synapses that were suggested in previous papers and tried to put our findings in the context of these speculations. However, we realize now that the last sentence of the paragraph could be seen as too strong and have now toned it down.

2. In Figure 2, the authors quantitated ZO1 localization to CEs or M/CoLo and showed that the amount of the localized ZO1 was halved in the connexin mutants. This means that the ZO1 localization to CEs or M/CoLo at least in part depends on connexins. This result should be carefully discussed in the Discussion section. The finding was described in the summarizing sentence on the page 8, "1) ZO1b localizes to putative electrical synaptic sites largely independent of Connexin proteins", but was not carefully described in the other sentences. The following sentences should be modified by weakening the asserting phrases in the main texts.

P8

"ZO1b localizes to the electrical synapse independent of Cxs"

"ZO1 does not require the presence of channel-forming proteins to localize at neuronal GJs during synaptogenesis"

P21

"In particular, the fact that ZO1 localizes to sites of synaptic contact independent of the Connexins"

We appreciate the suggestions and have clarified the indicated phrases. We have added an additional discussion of the reciprocal interaction to the Discussion.

3. In Figure 5A and 5B, the size of the circle markers is too large. They mask other data points in the back and make data distribution unclear. Set the marker size smaller.

Thank you, figure is updated.

4. Figure S7 does not exist. It seems a typo: Figure S6 instead of S7.

Thank you, all supplemental figure instances updated to eLife style.

5. P26: "The M-cell input resistance was estimated by applying a hyperpolarizing-current step of -1 nA and 20 ms in duration and measuring the voltage deflection caused, followed by derivation of resistance with Ohm's law."

Describe whether the authors subtracted pipette resistance from the calculated value.

As described in the Results and Methods sections, electrophysiological recordings were performed under the current clamp configuration, which allows compensation of the electrode resistance with the ‘bridge’ balance. This compensation was done automatically using the feature provided by the 700A Multiclamp amplifier followed by manual adjustment, if needed, for accurate compensation. Reference to bridge compensation is now included in the methods section.

6. Please include membrane potential recordings in all figures. Do any of the treatments substantially alter Vm? Why are membrane potentials missing everywhere?

As requested, we now include the values of membrane potential for the illustrated recordings in their corresponding figure legends. They were not previously reported for stylistic reasons, following the custom of the Pereda laboratory. The averaged values of resting potential for wild type and mutant zebrafish are reported in Table 1.

Finally, bath application of MA significantly altered the resting potential of the Mauthner cells in wild type zebrafish, averaging -80.2 ± 0.7 mV in control and -84 ± 1 mV after MA application (p = 0.04, n = 5). In contrast, bath application of CNQX/DAP5 did not alter the resting potential of the Mauthner cell in wt, tjp1b -/-, gjd2a -/- or gjd1a -/- zebrafish. Following the reviewers’ suggestion, we have now added reference to these observations (see electrophysiology results section).

7. The Discussion has a paragraph that is more than 2 pages. Please break that paragraph, and perhaps look at the Discussion and use headings to highlight what the authors think the essential issues are.

Thank you for the suggestion, the Discussion has been updated with headings and edited for clarity/brevity.