Dual role for pannexin 1 at synapses: regulating functional and morphological plasticity

Abstract

Pannexin 1 (PANX1) is an ion and metabolite membrane channel and scaffold protein enriched in synaptic compartments of neurons in the central nervous system. In addition to a well‐established link between PANX1 and synaptic plasticity, we recently identified a role for PANX1 in the regulation of dendritic spine stability. Notably, PANX1 and its interacting proteins are linked to neurological conditions involving dendritic spine loss. Understanding the dual role of PANX1 in synaptic function and morphology may help to shed light on these links. We explore potential mechanisms, including PANX1's interactions with postsynaptic receptors and cytoskeleton regulating proteins. Finally, we contextualize PANX1's dual role within neurological diseases involving dendritic spine and synapse dysfunction.

Abstract figure legend Pannexin 1 (PANX1) regulation of dendritic spines. The postnatal decrease in PANX1 expression could release key spine cytoskeleton‐regulating proteins enabling spine stabilization. Does the physiological decrease in neuronal PANX1 levels across brain development underlie dendritic spine maturation? Conversely, does the increased PANX1 expression and/or activity observed in inflammatory/injury contexts trigger pathological spine dysfunction, destabilization, and loss?

Introduction

Neurons communicate with one another at synapses, where the nerve terminals of one neuron pass signals to the postsynaptic compartments of another neuron. This communication between neurons forms the basis for brain circuitry. Synaptic plasticity refers to a range of processes that result in changes to the strength and efficiency of these synapses, and this is essential for brain network development and plasticity (reviewed in Appelbaum et al., ²⁰²³; Stampanoni Bassi et al., ²⁰¹⁹). In response to activity, synapses may exhibit functional plasticity, such as alterations in neurotransmitter release or receptor sensitivity, and morphological plasticity, including adjustments to the size and shape of the postsynaptic compartment. Dendritic spines are microscopic protrusions forming the postsynaptic compartments of excitatory synapses. The function and morphology of dendritic spines is therefore linked to excitatory synaptic strength, learning and memory (reviewed in Goto, ²⁰²²). Conversely, synapse dysfunction and dendritic spine structural aberrations are common to a variety of neurological conditions and are associated with cognitive decline (reviewed in Dejanovic et al., ²⁰²⁴). Importantly, the mechanisms underlying synapse function and dendritic spine structure in both physiological and pathophysiological circumstances remain incompletely understood.

The pannexin 1 (PANX1) channel and scaffold protein is enriched in the brain (Bruzzone et al., 2003; Ray et al., ²⁰⁰⁵) and is poised to act as a regulator of synaptic plasticity. Within the naïve rodent brain, PANX1 is expressed in both excitatory and inhibitory neurons (Loo et al., 2019; Ray et al., ²⁰⁰⁵; Vogt et al., ²⁰⁰⁵; Zhang et al., ²⁰¹⁴). Low levels are also present in glial cells (this is brain‐region dependent), as discussed elsewhere (reviewed in Sanchez‐Arias et al., ²⁰²¹). In neurons, PANX1 is enriched in synaptic compartments (Flores‐Muñoz et al., 2022; Sanchez‐Arias et al., ²⁰¹⁹) where it is found in close proximity to postsynaptic densities (Rangel‐Sandoval et al., 2024; Zoidl et al., ²⁰⁰⁷). Panx1 transcripts and PANX1 protein levels are high early in the postnatal period in rodents and sharply decline with age (Ray et al., 2005; Sanchez‐Arias et al., ²⁰¹⁹; Vogt et al., ²⁰⁰⁵), corresponding with critical periods of neuron and synapse development (reviewed in Farhy‐Tselnicker & Allen, ²⁰¹⁸). Moreover, PANX1 has strong ties to NMDA receptor activity (Patil et al., 2022; Rangel‐Sandoval et al., ²⁰²⁴; Thompson et al., ²⁰⁰⁸; Weilinger et al., ^{2012, 2016}) and neuronal cytoskeletal dynamics (Wicki‐Stordeur & Swayne, 2013; Flores‐Muñoz et al., ²⁰²²; Xu, Wicki‐Stordeur et al., ²⁰¹⁸), which are well known components of synaptic development and plasticity processes (Bellone & Nicoll, 2007; Franchini et al., ²⁰¹⁹; reviewed in Lei et al., ²⁰¹⁶).

PANX1 is a four‐transmembrane domain protein with intracellular N‐ and C‐termini that forms heptameric membrane channels (Deng et al., 2020; Jin et al., ²⁰²⁰; Michalski et al., ²⁰²⁰; Qu et al., ²⁰²⁰; Zhang et al., ²⁰²¹). These channels are permeable to small, elemental ions, like chloride, as well larger charged metabolites, like ATP and potentially lipids (Bialecki et al., 2020; reviewed in Hussain et al., ²⁰²⁴; Narahari et al., ²⁰²¹; reviewed in Navis et al., ²⁰²⁰), while also acting as a channel‐independent signalling hub via protein–protein interactions (Wicki‐Stordeur & Swayne, 2013; reviewed in Boyce et al., ²⁰¹⁴; DeLalio et al., ²⁰¹⁹; Frederiksen et al., ²⁰²³; Xiang et al., ²⁰²¹; Xu, Wicki‐Stordeur et al., ²⁰¹⁸). Recent cryo‐electron microscopy (cryo‐EM) structures of the PANX1 pore‐forming unit have provided insight into channel‐intrinsic mechanisms regulating PANX1 channel function (for example Hussain et al., ²⁰²⁴), yet there remain unanswered questions with respect to the range of substrate permeabilities, such as how the PANX1 pore might change to accommodate passage of larger molecules (reviewed in Mim et al., ²⁰²¹). The purported channel activation mechanisms are also varied, including receptor‐mediated, membrane stretch and caspase cleavage scenarios, which may be cell type and context dependent (Bao et al., 2004; Iglesias et al., ²⁰⁰⁸; Sandilos et al., ²⁰¹²). In terms of gating, recent structural studies demonstrate PANX1 can be regulated by intrapore lipids and the intracellular N‐terminal domain (Kuzuya et al., 2022). The intracellular C‐terminal domain can also control PANX1 gating (Dourado et al., 2014; Henze et al., ²⁰²⁴; Sandilos et al., ²⁰¹²); however, this has yet to be visualized in cryo‐EM studies due to the unstructured nature of this domain (Deng et al., 2020; Jin et al., ²⁰²⁰; Michalski et al., ²⁰²⁰; Qu et al., ²⁰²⁰; Zhang et al., ²⁰²¹). Phosphorylation by SRC kinase at sites on the intracellular loop and C‐terminal domains of mouse PANX1 activates the channel (DeLalio et al., 2019; Weilinger et al., ^{2016, 2012}); however, this regulation may not be conserved in humans (Ruan et al., 2024). Mounting evidence suggests the intracellular domains are also crucial for channel‐independent PANX1 scaffold/signalling hub functions through physical interactions with other proteins such as receptors, kinases and cytoskeletal regulators, many of which have known roles at the synapse (Frederiksen et al., 2023; Wicki‐Stordeur & Swayne, ²⁰¹³; Xu, Wicki‐Stordeur et al., ²⁰¹⁸; reviewed in O'Donnell & Penuela, ²⁰²³).

Given the range of channel‐dependent and ‐independent functions of PANX1, it is not surprising that PANX1 is implicated in a variety of neurological phenomena including neural progenitor cell maintenance, neurite outgrowth, spine stability and maturation, synaptic plasticity, neuronal network formation, and learning and memory (Ardiles et al., 2014; Flores‐Muñoz et al., ²⁰²²; Gajardo et al., ²⁰¹⁸; Prochnow et al., ²⁰¹²; Sanchez‐Arias et al., ²⁰¹⁹; Sanchez‐Arias, Candlish et al., ²⁰²⁰; Wicki‐Stordeur & Swayne, ²⁰¹³; Wicki‐Stordeur et al., ²⁰¹²; Wicki‐Stordeur et al., ²⁰¹⁶; Xu, Wicki‐Stordeur et al., ²⁰¹⁸). Work from others and us suggests this is likely due, in part, to several key interactions with components of the spine cytoskeleton (Wicki‐Stordeur & Swayne, 2013; Xu, Wicki‐Stordeur et al., ²⁰¹⁸; Flores‐Muñoz et al., ²⁰²²; Frederiksen et al., ²⁰²³; Figure 1A ). Moreover, PANX1 is linked to several neurological conditions associated with synapse dysfunction and loss, including Alzheimer's disease (AD), stroke and epilepsy (Frederiksen et al., 2023; (reviewed in Sanchez‐Arias et al., ²⁰²¹; Yeung et al., ²⁰²⁰; however, the underlying mechanisms are unclear. In the current review, we examine what is known about PANX1's dual role in synaptic function and structure, in both physiological and pathophysiological contexts.

Figure 1. PANX1 regulates dendritic spine stability.

A, PANX1 expression drops precipitously across development in the rodent cortex, corresponding to the peak period of dendritic spine development, and Panx1 KO or PANX1 block causes precocious dendritic spine stabilization. We hypothesize that PANX1 acts by sequestering key dendritic spine‐stabilizing proteins (ARP3 of the ARP2/3 complex, CRMP2), and the developmental decrease in PANX1 levels releases these proteins, promoting cytoskeletal remodelling and dendritic spine stabilization. B, in inflammation, activated glia and infiltrating immune cells in the brain release proinflammatory cytokines, such as TNFα. PANX1 levels and activity increase in CNS inflammatory conditions where dendritic spine dysfunction and loss occur. We postulate that TNFα increases neuronal PANX1 levels, which promotes re‐emergence of development‐like spine plasticity/instability leading to spine loss.

PANX1 in functional synaptic plasticity

Functional plasticity refers to activity‐dependent refinement of synaptic strength, often through modifications to neurotransmitter release, receptor expression and/or receptor sensitivity (reviewed in Goto, ²⁰²²; Magee & Grienberger, ²⁰²⁰). Recent works support a role for PANX1 as a modulator of synaptic transmission, network ensemble formation and synaptic functional plasticity (Ardiles et al., 2014; García‐Rojas et al., ²⁰²³; Prochnow et al., ²⁰¹²; Sanchez‐Arias et al., ²⁰¹⁹). Considering that most of these studies involve the hippocampal region, PANX1's role in other brain regions remains relatively unexplored.

Several recent studies have uncovered the influence of PANX1 on neuronal membrane properties and synaptic transmission. Hippocampal CA1 pyramidal neurons in adult global Panx1 KO mice exhibited enhanced excitability, with increased input resistance and excitatory threshold compared to wild‐type controls (Flores‐Muñoz et al., 2022; Prochnow et al., ²⁰¹²; Südkamp et al., ²⁰²¹). Panx1 KO neurons also showed indications of heightened release probability and larger readily releasable vesicle pools compared to wild‐type (Flores‐Muñoz et al., 2022). Electron microscopy analysis confirmed Panx1 KO mice had more synaptic vesicles per bouton as well as more docked vesicles per active zone in these neurons. Current injection in Panx1 KO neurons resulted in impaired spike frequency adaptation with higher action potential frequency compared to wild‐type cells. The action potentials in Panx1 KO neurons were overall shorter with an increased peak compared to wild‐type (Südkamp et al., 2021). However, spontaneous synaptic transmission appeared overall comparable between Panx1 KO and wild‐type pyramidal cells (Flores‐Muñoz et al., 2022). PANX1 also regulates inhibitory transmission to some extent in the hippocampus. Acute inhibition of PANX1 decreased both evoked and spontaneous GABA signalling onto pyramidal neurons; however, inhibitory transmission was comparable between global Panx1 KO and wild‐type mice (García‐Rojas et al., 2023), suggesting compensatory mechanisms were at play. At the network level, Fluo‐4 imaging of spontaneous Ca²⁺ transients in primary cortical neuron cultures revealed increases in both overall number of network ensembles (groups of co‐active neurons) and number of cells per ensemble in global Panx1 KO cultures compared to controls (Sanchez‐Arias et al., 2019). Moreover, we found higher baseline Ca²⁺ levels and altered Ca²⁺ transients in the Panx1 KO cells. Abnormal neuronal network responses were also recently indicated in Panx1‐null zebrafish optic tectum and pallium; however, these experiments were performed under conditions of cellular stress (Zoidl et al., 2024).

The role for PANX1 in neuronal function becomes more apparent under conditions required to elicit functional synaptic plasticity. Early work with adult global Panx1 KO mice demonstrated heightened and longer‐lasting long‐term potentiation (LTP) in hippocampal CA1 neurons in response to high frequency stimulation when compared to wild‐type controls (Ardiles et al., 2014; Prochnow et al., ²⁰¹²). Moreover, long‐term depression (LTD) was virtually abolished in this same region, with Panx1 KO animals instead showing slight potentiation following a low‐frequency stimulation paradigm (Ardiles et al., 2014). Wild‐type neurons treated with the PANX1 blocker probenecid demonstrated a more rapid return to baseline in LTD experiments compared to untreated control cells. Notably, these effects were not observed in juvenile animals (1 month old), indicating an age‐dependent role for PANX1. In rat primary hippocampal cultures, PANX1 and P2X7 purinergic receptor interplay regulated presynaptic strength and density (measured by vesicular glutamate transporter fluorescence intensity and bouton counts) (Rafael et al., 2020). PANX1 also increased the excitation to inhibition ratio within the hippocampus (García‐Rojas et al., 2023), which may underlie its overall role in modifying the threshold for induction of hippocampal functional plasticity.

Given functional synaptic plasticity is crucial for learning and memory (reviewed in Goto, ²⁰²²; Magee & Grienberger, ²⁰²⁰), it is not surprising that PANX1 deficient animals exhibited abnormalities in behaviours associated with these functions. In short, global Panx1 KO mice showed deficits in novel object recognition tasks and had indications of slight memory impairment (Prochnow et al., 2012). Subsequent work found comparable results between wild‐type and global Panx1 KO mice in the Morris water maze spatial learning paradigm; however, Panx1 KO animals demonstrated significant defects in ensuing reversal learning (Gajardo et al., 2018). Both novel object recognition and reversal learning in these contexts require LTD‐mediated plasticity (Griffiths et al., 2008), which is strongly attenuated in Panx1 KO hippocampus.

Several putative mechanisms underlie the role of PANX1 in regulating synaptic transmission and functional plasticity. In terms of channel‐dependent effects, PANX1 is a well‐known adenosine triphosphate (ATP) release channel (Bao et al., 2004; Sandilos et al., ²⁰¹²). Within the nervous system, ATP and its metabolite adenosine are both potent signalling molecules, binding to purinergic (P2Y and P2X) and adenosine (A1–3) receptors, respectively (reviewed in Shigetomi et al., ²⁰²³). Prochnow et al. (2012) initially found that the heightened LTP observed in global Panx1 KO mouse hippocampus could be normalized through exogenous addition of adenosine. The authors suggested physiological release of ATP via PANX1 channels, and its subsequent degradation to adenosine, activated presynaptic A1 receptors, which in turn limited glutamate release. Another pathway involving PANX1 release of ATP was identified in the retina between photoreceptors and horizontal cells (Vroman et al., 2014). In this system, ATP released via PANX1 was broken down into adenosine monophosphate (AMP), phosphate groups and protons, creating a slightly acidic extracellular environment that inhibited Ca²⁺ entry and glutamate release from the photoreceptors. Hyperpolarization of the horizontal cells decreased PANX1 activity, alkalinized the synaptic cleft, and promoted Ca²⁺ uptake and glutamate release from the photoreceptors. Further investigation is required to ascertain whether this represents a common PANX1‐based regulatory mechanism across brain regions and cell types. Evidence from peripheral synapses (neuromuscular junction) also implicates PANX1‐mediated ATP release in inhibitory presynaptic purinergic signalling (Bogacheva et al., 2022; Miteva et al., ^{2017, 2018, 2020}), although this mechanism requires further validation in CNS synapses.

Others have implicated PANX1 in regulation of synaptic endocannabinoid neurotransmitter levels. Bialecki et al. (2020) found postsynaptic PANX1 buffered levels of the lipid mediator anandamide under physiological conditions in the adult hippocampus. Anandamide concentrations were increased following acute inhibition of PANX1, which increased glutamate release through presynaptic TRPV1 channel activation. Blocking NMDA receptor ligand binding or SRC kinase activity inhibited this process, suggesting a physiological role for the NMDA receptor–SRC–PANX1 signalling complex, originally implicated in neuronal death during excitotoxicity (Weilinger et al., 2016, 2012), in regulating synaptic endocannabinoid levels. Supporting this, García‐Rojas et al. (2023) made similar observations while investigating PANX1's influence on inhibitory neurotransmission in adult CA1 hippocampus. In their experiments, PANX1 block increased synaptic endocannabinoid concentrations, activating presynaptic CB1 receptors and cAMP/protein kinase A signalling, thereby decreasing GABA release. In contrast to the findings from the facilitated glutamate release paradigm (Bialecki et al., 2020), the role of PANX1 in regulating inhibitory neurotransmission was independent of NMDA receptors. Whether PANX1 itself directly fluxes endocannabinoids, such as anandamide, or indirectly mediates their passage across membranes and/or their intracellular degradation remains unclear.

To gain further insight into PANX1's role at the synapse, we performed an in‐depth bioinformatics analysis of a neural PANX1–protein interaction network (Frederiksen et al., 2023) identified in Neuro‐2a (N2a) cells overexpressing enhanced green fluorescent protein (EGFP)‐tagged murine PANX1 (Wicki‐Stordeur & Swayne, 2013; Xu, Wicki‐Stordeur et al., ²⁰¹⁸). PANX1–protein interaction partners classified under the Gene Ontology cellular component term ‘synapse’ were further stratified into functionally linked protein clusters. These protein clusters were connected to (1) gene expression and translation, (2) cytoskeleton organization, (3) cell communication and its regulation, and (4) vesicle‐mediated transport. Although most of the PANX1–protein interaction partners identified in our initial screening have yet to be further validated (but see Wicki‐Stordeur & Swayne, ²⁰¹³; Xu, Wicki‐Stordeur et al., ²⁰¹⁸), this large‐scale pathway analysis underscores the relevance of a PANX1–cytoskeleton connection and provides novel perspectives on how PANX1 may fit into synaptic functions.

Notably, PANX1 impacts expression of key synaptic molecules. Transcript levels of metabotropic glutamate receptor 4 (Grm4) were increased specifically in adult global Panx1 KO whole hippocampal tissue compared to control (Prochnow et al., 2012); however, this was not quantified at the protein level. The authors also note this expression difference was not present in very young (postnatal day 8) Panx1 KO mice. Hippocampal synaptosomes and whole hippocampal tissue from adult global Panx1 KO mice showed increased synaptophysin (presynaptic protein), postsynaptic density protein 95 (PSD‐95), and synapse‐associated protein (SAP)102 (postsynaptic proteins) expression compared to wild‐type controls (Flores‐Muñoz et al., 2022). We similarly found that synaptosomes from juvenile global Panx1 KO cortex had heightened levels of PSD‐95, as well as increased GLUA1 and GLUN2A glutamate receptors compared to wild‐type controls (Sanchez‐Arias et al., 2019). Higher GLUN2A levels were also noted in PSD‐enriched membrane fractions from adult global Panx1 KO hippocampus, as well as heightened GLUN2B in synaptosomes from the same tissue, compared to wild‐type controls (Gajardo et al., 2018). Building on this, the authors determined that the effects of Panx1 KO on functional synaptic plasticity within the adult mouse hippocampus (heightened LTP and reduced LTD) resulted from changes to GLUN2A and GLUN2B subunit contribution to NMDA receptor function.

One explanation for PANX1's role in modifying the synaptic proteome, and thus synaptic function, may lie in its ability to regulate the cytoskeleton (reviewed in Boyce et al., ²⁰¹⁴; O'Donnell & Penuela, ²⁰²³). PANX1 interacts with several cytoskeleton‐associated proteins (Frederiksen et al., 2023; Wicki‐Stordeur & Swayne, ²⁰¹³; Xu, Wicki‐Stordeur et al., ²⁰¹⁸; reviewed in O'Donnell & Penuela, ²⁰²³). Evidence suggests it modulates actin filament formation and/or stability in the brain, possibly by controlling RHO GTPase activity (Flores‐Muñoz et al., 2022), and can also influence microtubule dynamics (Xu, Wicki‐Stordeur et al., 2018). Tight control of cytoskeletal stability and remodelling underlies synaptic plasticity, in part through modulating the contingent of receptors and channels at the postsynaptic membrane (reviewed in Gentile et al., ²⁰²²; Parato & Bartolini, ²⁰²¹). For example, the cortical actin cytoskeleton is responsible for localization and function of AMPA and NMDA receptors, and disruptions impair plasticity processes crucial for learning and memory (reviewed in Dutta et al., ²⁰²¹).

PANX1 and morphological plasticity

Neuron and synapse morphology is intimately tied to function (reviewed in Gentile et al., ²⁰²²). Morphological plasticity refers to the ability of synapses to change their shape and structure in response to neural activity. These morphological adaptations may include altered synaptic density, curvature and size of synaptic elements (reviewed in Goto, ²⁰²²; Magee & Grienberger, ²⁰²⁰).

PANX1 is emerging as a novel regulator of neuronal morphology. In foundational work, we investigated the direct impact of PANX1 disruption (pharmacological block with probenecid or Panx1 knockdown) on neurite outgrowth in otherwise untreated N2a cells, a murine neural crest‐derived cell line, and primary neural precursor cells (Wicki‐Stordeur & Swayne, 2013). This work revealed that PANX1 inhibits neurite outgrowth and/or stability. Relatedly, disruption of PANX1 also directly impacted cell morphology and differentiation in melanocytes, another neural crest‐derived cell (Penuela et al., 2012). Panx1 KO N2a cells undergoing long‐term treatment with retinoic acid possessed fewer neurites than wild‐type or PANX1 overexpressing cells (Xing et al., 2024). It is important to note that these constitutive Panx1 KO N2a cells were driven to differentiate with exogenous stimuli, whereas the previous work (Wicki‐Stordeur & Swayne, 2013) studied the direct impact of PANX1 disruption without exogenous stimuli. It is reasonable to speculate that differences in outcomes of PANX1 disruption in these N2a cell studies arise from the divergent experimental design and point to potential interference or crosstalk between PANX1 signalling pathways and retinoic acid signalling pathways regulating neurite outgrowth and/or stability. Consistent with a direct role for PANX1 in the inhibition of neurite outgrowth and/or stability, probenecid treatment increased dendritic length and branching in apical and basal dendrites of adult hippocampal CA1 neurons (Flores‐Muñoz et al., 2020). These findings were recapitulated in global Panx1 KO mice compared to wild‐type animals, and in Panx1 KO primary hippocampal neuron cultures, which were rescued by transient PANX1–EGFP expression (Flores‐Muñoz et al., 2022). In the peripheral nervous system, global Panx1 KO increased axon fibre cross sectional area in mouse sciatic nerves, and elevated axonal outgrowth length and density in dorsal root ganglion explants (Horton et al., 2017). Yet, others noted decreased neurite branching and dendritic length in dissociated global Panx1 KO dorsal root ganglion neurons in vitro compared to controls, with similar results in a neuron‐specific Panx1 KO paradigm (Xing et al., 2024).

PANX1 also regulates synapse and dendritic spine structure. We showed that global Panx1 KO enhanced spine density in both primary cortical neuron cultures and juvenile mouse pyramidal cells within layer 5 somatosensory cortex (Sanchez‐Arias et al., 2019). This increase in spine density was recapitulated in juvenile cortical excitatory neuron‐specific Panx1 KO mice (Sanchez‐Arias et al., 2019), suggesting a neuron‐specific role for PANX1 in regulating dendritic spine development in the cortex. Further examination revealed Panx1 KO stabilized nascent spines (called ‘protrusions') in days in vitro (DIV)10 immature primary cortical neuron cultures by decreasing spine turnover and movement, and these metrics could be rescued by transient expression of PANX1–EGFP (Sanchez‐Arias et al., 2019). In mature global Panx1 KO mice, hippocampal CA1 neurons exhibited significantly longer spines with a higher proportion of morphologically mature spines compared to control animals (Flores‐Muñoz et al., 2022). Electron microscopy analyses further revealed an increase in multi‐contact compound synapses in Panx1 KO neurons compared to controls. Notably, no difference was noted in overall dendritic spine density in these mature Panx1 KO mice; however, probenecid treatment in wild‐type mice did increase spine density (Flores‐Muñoz et al., 2020). Supporting these findings, DIV14–15 primary hippocampal neuron cultures from Panx1 KO animals also had increased spine length and proportion of morphologically mature spines compared to control cultures, and these metrics were rescued by transient PANX1–EGFP expression (Flores‐Muñoz et al., 2022). There was no significant effect of Panx1 KO on dendritic spine density in this experimental paradigm, implying that either the PANX1‐dependent spine density changes observed by us (Sanchez‐Arias et al., 2019; Sanchez‐Arias, Candlish et al., ²⁰²⁰) in the juvenile brain and immature primary cultures are normalized throughout development, or may be region specific (cortex versus hippocampus). Regardless, the evidence suggests PANX1 is a regulator of neuron, synapse and dendritic spine morphology, normally acting to prevent precocious spine stabilization and maturation as well as increases to dendrite length and complexity.

What might be the molecular mechanism(s) underlying PANX1's modulation of neuron and synapse structure? Insights from us and others point towards a major role for PANX1 in cytoskeletal regulation. Dynamic changes to both actin microfilament and microtubule cytoskeletons are crucial for neuron and synapse structure (reviewed in Gentile et al., ²⁰²²; Parato & Bartolini, ²⁰²¹). PANX1 also physically interacts with components of both actin and microtubule cytoskeletons, including β‐actin, actin related protein (ARP)3, and collapsin response mediator protein (CRMP2) (Bhalla‐Gehi et al., 2010; Frederiksen et al., ²⁰²³; Wicki‐Stordeur & Swayne, ²⁰¹³; Xu, Wicki‐Stordeur et al., ²⁰¹⁸; reviewed in O'Donnell & Penuela, ²⁰²³). Furthermore, our recent analysis of a neural PANX1 protein interactome uncovered additional putative cytoskeletal interaction partners (Frederiksen et al., 2023). ARP3 is a component of the ARP2/3 macromolecular complex that binds to and promotes actin microfilament polymerization (Mullins et al., 1998; Welch et al., ¹⁹⁹⁷; reviewed in Zheng et al., ²⁰²³), and is involved in regulating dendritic spine maturation (Spence et al., 2016). Recent work from global PANX1 KO mice demonstrated increased filamentous actin in dendrites and spines of primary hippocampal neurons and in whole hippocampus slices compared to controls (Flores‐Muñoz et al., 2022). The authors also found significantly increased expression of several actin‐binding proteins and RHO GTPase family members, RAC1 and RHO‐A, in PANX1 KO hippocampus compared to control. Closer examination revealed while PANX1 KO increased levels of active RAC1, expression of active RHO‐A was virtually abolished. This implies that neural PANX1 is intimately involved in the balance of RHO GTPase activity (Flores‐Muñoz et al., 2022), which acts as a master regulator of actin cytoskeleton within the cell (reviewed in Bement et al., ²⁰²⁴). CRMP2 is a microtubule binding protein that promotes microtubule polymerization and stabilization (reviewed in Gu & Ihara, ²⁰⁰⁰; Niwa et al., ²⁰¹⁷) and plays a part in dendritic spine formation and plasticity (Zhang et al., 2018, 2020). Our data indicate that blocking PANX1, which promoted neurite outgrowth (Wicki‐Stordeur & Swayne, 2013), also increased microtubule polymerization and stabilization, and disrupted the PANX1‐CRMP2 interaction (Xu, Wicki‐Stordeur et al., 2018). Taken together, it is likely that neuronal PANX1 functions, at least in part, by reducing cytoskeletal stability, and decreases in PANX1 expression and/or function, as observed across neurodevelopment, allow for actin and microtubule polymerization and dendritic spine stabilization (Fig. 1A ).

PANX1 in neurological conditions involving synaptic dysfunction and spine loss

A common theme across neurological conditions associated with impaired cognition is abnormal synaptic function and structure (reviewed in Dejanovic et al., ²⁰²⁴; Devinsky et al., ²⁰¹⁸; Elendu et al., ²⁰²³). Given the emerging roles for PANX1 in synaptic plasticity outlined above, it is reasonable to speculate that PANX1 may play a part in these pathology‐associated synaptic defects.

Several lines of evidence indicate PANX1's expression and/or function is impacted by common pathology‐associated triggers; this could be linked to spine instability and loss in neurological conditions in which PANX1 has been implicated. For instance, in endothelial cells, PANX1 levels and activity were upregulated following exposure to tumour necrosis factor α (TNFα; Yang et al., ²⁰²⁰), a pro‐inflammatory cytokine that can modulate synapse function and structure (reviewed in Baht et al., ²⁰¹⁸). Within the brain, TNFα is increased in a variety of neurological conditions and injuries associated with inflammation (reviewed in Heir & Stellwagen, ²⁰²⁰, Sanchez‐Arias, Wicki‐Stordeur et al., 2020). PANX1 may also be upregulated in neurons in response to heightened TNFα. Given increased PANX1 expression acts to destabilize dendritic spines (Sanchez‐Arias et al., 2019), increases in PANX1 expression associated with injury and/or inflammation could contribute to spine instability and loss associated with a number of neurological conditions, ranging from neurodevelopmental disorders to neurodegenerative diseases (Fig. 1B ).

Additional signalling pathways common to neurological conditions and injury associated with spine instability and/or loss also modulate PANX1 activity. Importantly, PANX1's function is intimately linked to that of NMDA receptors (Patil et al., 2022; Rangel‐Sandoval et al., ²⁰²⁴; Thompson et al., ²⁰⁰⁸; Weilinger et al., ^{2016, 2012}), which exhibit hyperactivity due to glutamate build‐up in CNS pathologies. This mechanism is two pronged, as physiological PANX1–NMDA receptor signalling is necessary for synaptic plasticity (Bialecki et al., 2020), yet hyperactivity of the signalling complex may underlie aberrant synaptic transmission leading to epileptiform activity, anoxic depolarizations and cell death (Thompson et al., 2008, 2006; Weilinger et al., ^{2016, 2012}). Moreover, neuronal swelling enhanced neuronal PANX1 activity downstream of increased reactive oxygen species (ROS) generation (Weilinger et al., 2023). Intriguingly, PANX1 played a dual role here, both contributing to swelling‐induced neuronal death and promoting protective microglial contacts. Increases in ROS levels are common to physiological ageing and a range of neurological conditions (reviewed in Zhou et al., ²⁰²²), representing a shared mechanism through which overactivated PANX1 may result in synaptic dysfunction. Finally, in apoptotic cells, caspase cleavage of the PANX1 C‐terminal tail resulted in uncontrolled channel opening, leading to immune cell recruitment and cell death (Chekeni et al., 2010; Sandilos et al., ²⁰¹²). Of note, caspases also play key roles in the brain beyond cell death, such as neurite outgrowth, dendritic spine pruning and functional synaptic plasticity, and therefore caspase regulation of PANX1 could be important to consider in these contexts (reviewed in Espinosa‐Oliva et al., ²⁰¹⁹; Li & Sheng, ²⁰¹²).

Taken together, rising evidence implicates PANX1 expression and/or function in a variety of neurological conditions associated with synapse and dendritic spine aberrations. In this section, we focus on AD, epilepsy, and cerebral stroke, which represent pathophysiological states with robust alterations in synaptic plasticity and in which evidence suggests PANX1 plays a pathological role (reviewed in Sanchez‐Arias et al., ²⁰²¹; Yeung et al., ²⁰²⁰).

PANX1 and Alzheimer's disease

Synaptic failure is an early cause of cognitive decline and memory dysfunction in AD and is directly correlated with the loss of dendritic spines (reviewed in Meftah & Gan, ²⁰²³; Tzioras et al., ²⁰²³). In fact, markedly reduced dendritic spine density is a hallmark of AD post‐mortem human brain and mouse model tissue and is recapitulated in primary culture models (Boros et al., 2017; reviewed in Tzioras et al., ²⁰²³). While the exact mechanisms underlying synapse dysfunction and spine loss in AD remain obscure, several AD‐associated molecular stimuli such as soluble amyloid β (Aβ) peptides (reviewed in Karisetty et al., 2020), ROS (reviewed in Tönnies & Trushina, ²⁰¹⁷) and inflammatory signalling molecules (reviewed in Heir & Stellwagen, ²⁰²⁰; Meissner et al., ²⁰¹⁵) may play a role.

A growing body of evidence links PANX1 expression and function to AD‐associated stimuli and their pathological impacts at the synapse. The hippocampus of the APP/PS1 mouse model of AD exhibited heightened PANX1 expression levels and activity between 3 and 12 months of age compared to wild‐type controls (Flores‐Muñoz et al., 2020). These animals showed characteristic impairments in functional synaptic plasticity (LTP and LTD), reduced dendritic length and branching, and decreased dendritic spine density at 6 months of age. Acute block of PANX1 with probenecid in these APP/PS1 mice rescued synaptic plasticity deficits, and restored dendrite structure and dendritic spine densities to wild‐type levels. The underlying mechanism could involve over‐activation of the tau‐kinase p38 mitogen activated protein kinase downstream of increased PANX1 activity in the AD‐model mice. Similarly, global Panx1 KO mice were resistant to oligomeric Aβ‐induced deficits in hippocampal function, namely diminished LTP and impaired spike frequency adaptation (Südkamp et al., 2021). In vitro, neuronal PANX1 may be indirectly activated by Aβ‐induced glial cell release of glutamate and ATP, and ensuing activation of NMDA and P2X7 receptors (Orellana et al., 2011). In this context, PANX1 activity was correlated with increased neuronal death. ROS, which contribute to both normal ageing processes and an array of neurodegenerative conditions such as AD (reviewed in Zhou et al., ²⁰²²), were also found to open neuronal PANX1 channels in juvenile rat and mouse hippocampal neurons (Weilinger et al., 2023). Notably, PANX1 activity influenced generation of ROS in several cell types (reviewed in Xu, Chen et al., ²⁰¹⁸), suggesting there may exist a positive feedback loop between PANX1 expression/function and ROS generation during AD progression. A recent report described heightened PANX1 expression and activity in cultured human umbilical vein endothelial cells exposed to TNFα (Yang et al., 2020), a proinflammatory cytokine upregulated in AD (reviewed in Heir & Stellwagen, ²⁰²⁰), which may contribute to dendritic spine loss. Whether neuronal PANX1 is impacted by TNFα remains unknown. Notably, brain mast cells, which secrete various proinflammatory cytokines, including TNFα, are stimulated to release these cytokines in a PANX1‐dependent manner following Aβ exposure (Harcha et al., 2015). Together this implies there could exist a positive feedback loop in inflammatory conditions in which TNFα upregulates PANX1 and increased PANX1 activity promotes further cytokine release.

PANX1 in epilepsy

Epilepsy is characterized by aberrant and excessive neuronal activity leading to recurrent seizures. The underlying mechanisms are complex and not yet fully elucidated (extensively covered elsewhere; for example, reviewed in Devinsky et al., ²⁰¹⁸). In this section we focus on aberrant synaptic plasticity in the context of epilepsy and seizure models in relation to the emerging role of PANX1 in synaptic function and structure. One common theme in epilepsy is an abnormal excitation to inhibition balance (reviewed in van van Hugte et al., ²⁰²³), although whether this is a cause or consequence remains unknown. Morphological abnormalities are also frequent in both human epilepsy patients and animal seizure models (reviewed in Kumari & Brewster, ²⁰²⁴). While at the synapse level this most commonly presents as reduced dendritic spine density, there are also reports of synaptogenesis and increased spine density, changes in spine shape, or a dynamic combination (reviewed in Jean et al., ²⁰²³; Kumari & Brewster, ²⁰²⁴). Together, these suggest a re‐emergence of neuroplasticity processes normally seen during development, to re‐establish circuitry lost to epileptiform activity (reviewed in Jean et al., ²⁰²³; Kumari & Brewster, ²⁰²⁴).

A large body of work has implicated PANX1 expression and function in epilepsy and seizure models and was examined in recent comprehensive reviews (for example Aquilino et al., ²⁰¹⁹; Yeung et al., ²⁰²⁰). To briefly summarize, brain tissue from human epilepsy patients and from rodent seizure models exhibit increased PANX1 expression levels. PANX1 function was implicated in NMDA receptor‐induced epileptiform activity in the rodent hippocampus (Thompson et al., 2008), and PANX1 may also be activated by other seizure‐associated stimuli, such as increased extracellular K⁺ and intracellular Ca²⁺ (reviewed in Aquilino et al., ²⁰¹⁹). Accordingly, PANX1 inhibition reduced ictal discharge frequency and duration in human epilepsy brain tissues (Dossi et al., 2018), and PANX1 block or deletion decreased seizures or seizure‐like activity in a variety of animal models (reviewed in Aquilino et al., ^{2020, 2019}; Whyte‐Fagundes et al., ²⁰²²).

Given these observations, it is likely that PANX1 contributes to the synaptic abnormalities commonly observed in epilepsy and seizure models. Therefore, it would be logical to explore targeting PANX1 to reduce epileptiform activity, as well as to modulate synapse function and remodelling.

PANX1 in stroke

Stroke results from compromised blood flow to a given brain region and, in survivors, is associated with varying severity of cognitive dysfunction (reviewed in Elendu et al., ²⁰²³). In neurons, the decreased supply of oxygen and nutrients causes glutamate build‐up and anoxic depolarizations leading to primarily Ca²⁺‐mediated excitotoxicity and cell death. In addition, many surviving neurons are damaged and/or exposed to a variety of inflammatory cytokines and other deleterious compounds, resulting in aberrant synaptic transmission and synapse loss (reviewed in Li et al., ²⁰²⁴). In the weeks and months following stroke, developmental plasticity processes are re‐activated to facilitate neuronal circuit repair and recovery (reviewed in Joy & Carmichael, ²⁰²¹).

Shortly after its discovery, PANX1 was implicated in neuronal death following stroke. This subject has been extensively reviewed elsewhere (for example, Yeung et al., ²⁰²⁰). Briefly, using acutely isolated hippocampal neurons and/or acute hippocampal slices exposed to stroke‐like stimuli (oxygen–glucose deprivation or anoxia), PANX1 was identified as the channel mediating anoxic depolarizations leading to neuron death (Thompson et al., 2006; Weilinger et al., ²⁰¹²). Knocking out or blocking PANX1 was neuroprotective, reducing cell death and/or infarct size in several stroke paradigms (Cisneros‐Mejorado et al., 2015; Dvoriantchikova et al., ²⁰¹²; Xiong et al., ²⁰¹⁴). Notably, in one study, knock out of both Panx1 and Panx2 provided maximal protective effects (Bargiotas et al., 2011). Differences in stroke model (permanent vs. transient occlusion–reperfusion), and/or sex differences in the role of PANX1 in the permanent occlusion model (Freitas‐Andrade et al., 2017) may underlie these disparate findings.

Several mechanisms could underlie increased PANX1 activity following stroke. For instance, increased metabotropic activity of NMDA receptors following stroke activates SRC kinase, which in turn phosphorylates PANX1 and increases its activity (Weilinger et al., 2012). Inhibiting the NMDA receptor–SRC–PANX1 signalling complex provided neuroprotective effects in vitro and in vivo (Weilinger et al., 2016). PANX1 may also be activated by other components of the ischaemic environment such as increased neuronal swelling and ROS (Weilinger et al., 2023), reactive nitrogen species (Zhang et al., 2008), pro‐inflammatory cytokines like TNFα (Yang et al., 2020) and effector caspase activity (Chekeni et al., 2010).

While the role for PANX1 in stroke has largely centred on neuronal death, it is likely PANX1 also plays a role in modulating stroke‐triggered changes in synapses. Given its emerging importance in functional and morphological synaptic plasticity, heightened PANX1 function may contribute to aberrant synaptic transmission and/or dendritic spine destabilization and loss in surviving neurons in the peri‐infarct region. On the other hand, PANX1 could play a role in the heightened plasticity processes typically observed during post‐stroke recovery, a period associated with circuit remodelling and recovery (reviewed in Joy & Carmichael, ²⁰²¹). Such complex spatio‐temporal interplay should be considered in the development of therapeutic strategies.

Discussion and future directions

PANX1 has dual functionality at the synapse, influencing both functional and morphological plasticity. Importantly, the relationship between functional and morphological synaptic plasticity is dynamic, bidirectional and highly integrative: functional changes precede and precipitate structural changes, which in turn can stabilize and sustain functional alterations for learning and memory (reviewed in Appelbaum et al., ²⁰²³). Here, we first discuss synaptic transmission and functional synaptic plasticity, in which PANX1 appears to dampen neuronal excitability and modulate the threshold for achieving functional plasticity. We then analyse morphological plasticity, where PANX1 functions to limit dendritic length and complexity, as well as destabilize dendritic spines. We finally outline connections between PANX1 and example CNS pathologies exhibiting synaptic dysregulation, in which PANX1 expression and/or activity are upregulated and implicated in disease progression.

Given PANX1's expression is dramatically higher in developmental compared to mature time points in the rodent brain (Sanchez‐Arias et al., 2019; Vogt et al., ²⁰⁰⁵), it is intriguing that the relative impact of PANX1 in hippocampal functional synaptic plasticity, based on Panx1 KO data, follows an opposing pattern. Panx1 KO increased LTP and abolished LTD in adult mouse hippocampal CA1 region but had no effect in juvenile hippocampus (Ardiles et al., 2014). This may be explained by changes in PANX1's localization, activation dynamics and/or protein interaction partners. For example, PANX1 functionally interacts with NMDA receptors and physically associates with the obligatory subunit GLUN1 (Bialecki et al., 2020; Patil et al., ²⁰²²; Thompson et al., ²⁰⁰⁸; Weilinger et al., ^{2016, 2012}). The expression and integration of the auxiliary NMDA receptor subunits (GLUN2A–D, GLUN3A–B) varies based on cell type, location and developmental stage, and is a key contributor to functional diversity across NMDA receptors (reviewed in Hansen et al., ²⁰²¹). How this impacts the PANX1–NMDA receptor signalling complex is unknown, but could feasibly alter PANX1's influence on synaptic plasticity in a region‐ and time‐specific manner.

This also more generally raises the issue of brain region‐specific roles for PANX1. Most PANX1–synaptic plasticity‐related studies were performed in adult rodent hippocampus, with hippocampus‐derived primary neuronal cultures, or acutely isolated hippocampal neurons (e.g. Flores‐Muñoz et al., ²⁰²²; García‐Rojas et al., ²⁰²³; Prochnow et al., ²⁰¹²). While it is likely that at least some components of PANX1's synaptic function are conserved across brain regions, its role in synaptic plasticity in regions other than hippocampus should be examined in future studies.

Further consideration of the role of PANX1 in different cell‐types in healthy and disease contexts across the lifespan is also warranted. For example, Panx1 is relatively enriched in neurons and low to absent in glial cells in the cortex under naïve conditions (Loo et al., 2019; Vogt et al., ²⁰⁰⁵; Zappalà et al., ²⁰⁰⁶; Zhang et al., ²⁰¹⁴; reviewed in Sanchez‐Arias et al., ²⁰²¹), yet glia from other regions (i.e. hippocampus, spinal cord) seem to express PANX1 (Boassa et al., 2014; Rigato et al., ²⁰¹²). Moreover, there are some reports of PANX1 expression in cultured glia (Huang et al., 2007; Iglesias et al., ²⁰⁰⁹; Pelegrin & Surprenant, ²⁰⁰⁶). Given cultured glia are substantially changed from those in vivo, with respect to morphology and gene expression (Lattke et al., 2021; Li et al., ²⁰²¹), these data must be interpreted with care. Future use of tools such as RNAScope could monitor cell type‐specific changes in expression across development and disease states within the context of various experimental paradigms. Additionally, comprehensive human expression data for PANX1 are lacking. The emergence of cell‐type specific spatio‐temporal transcriptomic and proteomic (e.g. Brain RNA‐Seq, https://brainrnaseq.org) data from rodent and humans will be instrumental in directing the pannexin field in future studies.

Several potential mechanisms contribute to PANX1 regulation of synaptic plasticity. These include (1) inhibition of neurite stability through sequestration of cytoskeletal regulating proteins and (2) modulation of NMDA receptor activity and associated buffering of endocannabinoid signalling. Our in‐depth bioinformatics analysis of a neural PANX1–protein interactome network additionally implicated PANX1 in (1) gene expression and translation, (2) cytoskeleton organization, (3) cell communication and its regulation, and (4) vesicle‐mediated transport within the synapse (Frederiksen et al., 2023).

Notably, the well‐known role of PANX1 as an ATP release channel in various tissues and cell types (Bao et al., 2004; Sandilos et al., ²⁰¹²) is conspicuously lacking in PANX1–synapse‐associated literature from the CNS (yet see peripheral synapse associations: Bogacheva et al., ²⁰²²; Miteva et al., ^{2017, 2018, 2020}). ATP and its metabolite adenosine are key regulators of synaptic transmission, functional plasticity and neuronal circuit function (reviewed in Shigetomi et al., ²⁰²³), and their signalling pathways are implicated in various neuropathological processes (reviewed in Sanchez‐Arias et al., ²⁰²¹). Within the brain, purinergic signalling can be particularly difficult to probe given the various release modalities from multiple cell types, and an array of responding receptor types. Extracellular ATP activates ionotropic P2X and metabotropic P2Y receptor families in neurons and glial cells to elicit myriad effects. This ATP is also rapidly degraded into adenosine, which activates the metabotropic P1 family of receptors, generally dampening synaptic communication. In one report, adding exogenous adenosine rescued functional plasticity defects in Panx1 KO mouse hippocampus (Prochnow et al., 2012); however, this was not reported in subsequent studies. Notably, activation of purinergic signalling pathways triggers cytoskeletal remodelling in other cell types (Degagné et al., 2013; Flores‐Muñoz et al., ²⁰²²; Goldman et al., ²⁰¹³) and cytoskeletal dynamics are a key dictator of neuronal morphology (reviewed in Parato and Bartolini, ²⁰²¹; Gentile et al., ²⁰²²). Yet, to our knowledge, the contribution of PANX1‐mediated ATP release in synaptic morphological plasticity remains unknown. By combining cell‐type‐specific genetic manipulations with the growing toolset for probing extracellular ATP dynamics and signalling (summarized in Shigetomi et al., ²⁰²³), this could be specifically assessed in future studies.

In the context of neurological conditions, PANX1 regulates a spectrum of pathological cellular processes ranging from aberrations in synaptic plasticity through to epileptiform activity, anoxic depolarizations, and cell death (Dossi et al., 2018; Thompson et al., ²⁰⁰⁶; Weilinger et al., ²⁰¹²; reviewed in Yeung et al., ²⁰²⁰). Gradients in soluble stimuli that regulate PANX1 expression levels and/or activity could form the basis for increased levels of PANX1 function. For example, PANX1 expression is upregulated both ipsilateral and contralateral to a stroke in the rodent brain (Freitas‐Andrade et al., 2017). However, TNFα, which activates and increases the expression levels of PANX1 in endothelial cells (Yang et al., 2020), is higher in the peri‐infarct region due to its release by activated glia and infiltrating immune cells in this area shortly following stroke (reviewed in Xue et al., ²⁰²²). Moreover, PANX1 can be activated through NMDA receptor function (Thompson et al., 2008, 2006; Weilinger et al., ^{2016, 2012}), and NMDA receptors are hyperactive in the peri‐infarct region where oxygen and nutrient depletion causes acute glutamate build‐up in the hours and days following stroke (reviewed in Joy & Carmichael, ²⁰²¹). These could create a PANX1 expression and activation gradient across space and time that mediates neuronal death at the infarct, synaptic dysfunction and spine destabilization in the peri‐infarct, and the re‐emergence of developmental plasticity processes at later time points.

Our recent bioinformatics analysis of a neural PANX1 protein interactome provided additional insights regarding PANX1's disease relevance and potential pathways underlying PANX1's role in these conditions (Frederiksen et al., 2023). Several PANX1 interacting proteins overlapped with susceptibility genes for schizophrenia and autism spectrum disorders, and to a lesser extent with susceptibility genes for AD and Parkinson's disease (PD). N‐Ethylmaleimide sensitive factor (NSF), a vesicle‐fusing ATPase, was a commonality between PANX1 interactors and PD susceptibility genes. Notably, NSF is characterized by the Gene Ontology cellular component term ‘synapse’. Expression of dominant negative Nsf2 in Drosophila disrupted the neuromuscular junction and synapse development (Stewart et al., 2002, 2005). In rat primary hippocampal neuron cultures, NSF regulated synaptic AMPA receptor levels (Hu & Hsueh, 2024; Huang et al., ²⁰⁰⁵), and NSF knockdown inhibited dendritic spine enlargement by modulating synaptic actin cytoskeleton (Hu & Hsueh, 2024). In addition to PD, NSF dysfunction was implicated in an infantile epileptic disorder (Suzuki et al., 2019) and in stroke (Yuan et al., 2021). The PANX1–NSF association warrants future study to characterize its role in physiological and pathological synapse function.

While the role of PANX1 in synaptic plasticity is emerging, so too are many avenues for future study. For instance, are the molecular changes to the synaptic proteome in Panx1 KO neurons (Flores‐Muñoz et al., 2022; Gajardo et al., ²⁰¹⁸; Prochnow et al., ²⁰¹²; Sanchez‐Arias et al., ²⁰¹⁹) simply the response of compensatory mechanisms, or does PANX1 itself play an active role in protein synthesis, trafficking and/or stability? Two lines of evidence suggest the latter option is plausible: (1) PANX1 modulates cytoskeleton stability, which is important for forming and maintaining the synaptic protein complement, and (2) our analysis of neural PANX1 protein interaction partners unveiled a subset involved in gene synthesis and translation. Secondly, while PANX1 expression and/or activity is heightened in many CNS pathologies exhibiting aberrant synaptic plasticity, does PANX1 directly regulate pathological synaptic changes? Whether PANX1 plays a dual role in both synaptic dysfunction and cell death as well as neuroplastic repair processes within a given CNS condition remains to be seen. There also exist limited human data on the role of PANX1 in CNS health and disease. Although the PANX1 sequence possesses regions of high inter‐species variability (Dourado et al., 2014), there is some indication that the role of PANX1 as an inhibitor of spine formation may be evolutionarily conserved: an invertebrate analogue of PANX1 (Unc‐9/Innexin) regulates spatial arrangement of GABAergic synapses in C. elegans (Hendi et al., 2022). There are sporadic reports of human PANX1 variants in the literature, such as one associated with multisystem dysfunction including intellectual disability (Shao et al., 2016), and others associated with polymicrogyria (Akula et al., 2023). It is critical to advance our understanding of PANX1 in human neurobiology. To do this, a multifaceted approach will be necessary, combining resources and techniques such as comprehensive spatio‐temporal human transcriptomics and/or proteomics data, advances in genome sequencing, and application of human model systems such as induced pluripotent stem cell‐derived neural cultures.

Additional information

Competing interests

The authors declare no conflict of interest.

Author Contributions

L.A.S., L.E.W.S., and A.C.M. conceptualized, wrote, and revised the manuscript. A.V.A., A.C.M., and L.A.S. created the figure. All authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

This work was supported by grants from the Alzheimer Society Research Program and the Brain Canada Foundation [application #0000000064; this Project has been made possible by the Canada Brain Research Fund (CBRF), an innovative arrangement between the Government of Canada (through Health Canada) and Brain Canada Foundation, and Alzheimer's Society of Canada] as well as the Canadian Institutes of Health Research [PJT 189953 and PJT 185887] awarded to L.A.S.

Supporting information

Peer Review History

Biographies

Adriana Casillas Martinez is an MSc student in the Swayne Lab investigating the role of pannexin 1 (PANX1) in dendritic spine stability in the context of neuroinflammation associated with Alzheimer's disease.

Leigh Wicki‐Stordeur obtained her PhD in Neuroscience in the Swayne Lab in 2015, where the primary focus of her thesis research was the role of PANX1 in ventricular zone neurodevelopment. She is now a Research Associate in the Swayne Lab.

Annika Ariano is also an MSc student in the Swayne Lab and studies the regulation of PANX1 expression during dendritic spine development.

Leigh Anne Swayne obtained her PhD in Neuroscience in 2005 and started her independent research laboratory at the University of Victoria in 2011. She is a cell biologist studying the role of ion channels and their protein–protein interactions in neuronal and cardiomyocyte development and disease. Her expertise on pannexins in neurobiology is internationally recognized.