ATP-release pannexin channels are gated by lysophospholipids

Authors’ Response (27 January 2025)

Revised preprint

GENERAL ASSESSMENT

The revised manuscript by Henze et al. presents a novel and significant contribution to the field, demonstrating that lysophospholipids (LPC) act as endogenous activators of pannexin channels. The study provides compelling evidence that LPC activation of PANX1 and PANX2 channels facilitates the release of signaling molecules critical for immune responses, particularly in the context of inflammation and inflammasome activation. The removal of the cryo-EM data, addressing concerns about the binding site and mechanism of LPC activation, is a thoughtful revision that enhances the manuscript's focus. While the authors have addressed many of the previous critiques, key mechanistic questions remain regarding channel permeability and the specificity of LPC-induced metabolite release.

RECOMMENDATIONS

Essential revisions:

1. While the study suggests that LPC-induced PANX1 activation results in the release of metabolites via the pannexin channel pore, no direct evidence is provided to confirm this. It is possible that the release occurs through alternative pathways or that detected metabolites are by-products of other permeating substances. Additional experiments or a more thorough discussion of these possibilities would enhance the manuscript’s rigor.

We thank the reviewers for pointing this out. While the overlap between our current study and the published secretomics studies by Medina et al. supports the direct release of these metabolites, we agree that our experiments do not rule out the possibility of an alternative pathway. We have updated the discussion to acknowledge this limitation.

1. The manuscript lacks data on the ion selectivity of LPC-activated PANX1, an important aspect for understanding the channel’s permeability profile. A comparison of the selectivity of LPC-induced PANX1 currents to those activated by other stimuli, such as C-terminal cleavage or depolarization, would clarify whether LPC induces a unique or comparable open state. The reviewers appreciate the technical challenges of these experiments, however, thus a more thorough discussion of the uncertainties may be appropriate.

We agree that characterizing the channel’s permeability profile is essential. However, as the reviewers noted, prolonged or repeated lysophospholipid (LPC) perfusion often destabilizes the patch, making it difficult to apply conventional reversal potential analysis. Additionally, the application of voltage ramps would inevitably activate Panx1 channels, further complicating the interpretation of the results.

As a workaround, we analysed whole-cell current density before and after LPC treatment at -60 mV using different buffers containing various anions and cations. Both Panx1 and Panx2 channels produced significantly larger currents in NaCl or NMDG-Cl compared to NaGluconate, indicating that LPC-activated channels are more selective for anions under these conditions. Interestingly, currents in NMDG-Cl were slightly smaller than in NaCl, suggesting that NMDG may have an inhibitory effect on these channels activated by LPC. While we acknowledge that this analysis does not directly compare ion selectivity within the same patch, the almost negligible current observed in NaGluconate strongly suggests that small anion conductance through both Panx1 and Panx2 channels is greater than cation conductance. We have included these findings in the new Fig. 1.

Another important aspect of Panx1 channels is their ability to allow the permeation of cationic molecules, such as YOPRO-1, when the C-terminus is cleaved. In our experiments, we demonstrate that LPC-activated full-length Panx1 channels are permeable to both anions (e.g., ATP and Cl) and cations (e.g., YOPRO-1). Although the precise mechanism underlying ion selectivity remains to be elucidated, our data support the conclusion that LPC can facilitate the release of multiple signalling molecules through full-length Panx1.

1. The YO-PRO-1 uptake observed in the absence of LPC contradicts prior findings (e.g., Bayliss et al., eLife, 2021) that full-length PANX1 does not release large molecules like ATP or YO-PRO-1 under similar conditions. This discrepancy raises concerns about the interpretation of the proteoliposome experiments. Addressing this by ion selectivity experiments, or discussing differences in experimental conditions and reconstitution protocols, could provide clarity as well as strengthen the conclusions.

We are also aware of the discrepancy and currently lack a clear understanding of the underlying mechanism. However, there are substantial differences between the two experimental setups that may account for the divergent results.

First, we used human Panx1 tagged in the flexible intracellular loop, whereas the Bayliss group used frog Panx1 tagged with GFP at the C-terminus. This difference in tagging and species may have contributed to variations in basal activity. Second, the lipid compositions used for reconstitution were significantly different. In our experiments, we used 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPG), and sphingomyelin, while the Bayliss group employed a mixture of 70% brain phosphatidylcholine, 15% total brain lipid extract, 14% cholesterol, and 1% phosphatidylinositol 4,5-bisphosphate. Given that the function of many ion channels is heavily influenced by lipid composition, these differences could have contributed to the observed discrepancy. Regardless, our functional reconstitution experiments clearly demonstrate that LPC stimulates YOPRO-1 uptake in a dose-dependent manner, which forms the foundation of our interpretation. We have included this discussion in the revised manuscript. discrepancy.

Optional suggestions:

1. The manuscript refers to experiments conducted in human monocytes, but the actual cell line used was THP-1, a human monocytic leukemia cell line. This should be made clearer in the text to avoid confusion.

We are confused by this comment. Our manuscript states that we used "phorbol 12-myristate 13-acetate (PMA)-differentiated human THP-1 monocytes" which should clearly indicate which cells we used in the study.

1. The behavior of LPC at concentrations exceeding its Critical Micelle Concentration (CMC) (4–8 μM for 16:0 Lyso-PC) should be considered. The authors should discuss whether micelle formation affects the observed channel activation and how this might influence the interpretation of the results.

Our dose-response experiments, shown in Supplementary Fig. 3, suggest that CMC does not appear to affect their activity. While we are eager to understand the delivery method and action mechanisms of lysophospholipids, such studies are beyond the scope of the current work.

(This is a response to peer review conducted by Biophysics Colab on version 2 of this preprint.)