The mammalian pannexin family comprises three distinct homologs: Panx1, Panx2, and Panx3. Genetic mapping reveals that both Panx1 and Panx3 reside on human chromosome 11, encoding proteins with molecular masses of 48 kDa and 45 kDa, respectively (Baranova et al., 2004). In contrast, Panx2 is localized to chromosome 22 and encodes a 70 kDa protein, making it the largest Panx family member characterized by an extended C-terminal domain (Baranova et al., 2004). Sequence analysis demonstrates limited structural conservation among Panx members, sharing only 50%–60% amino acid identity. However, structural conversation between different species is as high as 90% (Baranova et al., 2004). This evolutionary preservation underscores the functional importance of Panx proteins and validates their study in cross-species experimental models.

Panx proteins adopt a canonical four-transmembrane (4TM) topology, featuring two extracellular loops (ECL1/ECL2), cytoplasmic N- and C-termini, and an intracellular loop. Among the three family members, Panx1 has garnered the most attention, particularly for its ATP release during apoptotic cell clearance. ATP and UTP release (>80 nM) functions as a “find-me” signal that recruit phagocytes, enabling immunologically silent clearance of apoptotic cells (Chekeni et al., 2010; Elliott et al., 2009). Apoptotic stimuli trigger Panx1 channel activation through caspase3/7-mediated proteolytic cleavage at residues 376–379 of C-terminal domain (CTD) (Sandilos et al., 2012; Penuela et al., 2014a). This cleavage modification liberates a C-terminal autoinhibitory domain (CAD), relieving steric hindrance and inducing conformational changes in the N-terminal domain (NTD) to facilitate channel opening (Henze et al., 2024). Furthermore, hemichannels formed by truncated Panx1 (but not wild-type) permit Ca²⁺ influx that enables cell death without additional stimuli (Salgado et al., 2024). Importantly, functional analysis confirms that the Panx1 lacking NTD not only abolished functional channel opening but also exerted dominant-negative suppression of channel activity (Kuzuya et al., 2022). In addition, murine Panx2 was cleaved at aspartic acid (D) site Asp416 (equivalent to Asp415 of human Panx2) by caspase-3/7 activation (Sanchez-Pupo et al., 2022).

The ECL1/ECL2 of Panx1 constitute critical regulatory modules governing channel inhibition and ionic permeability (Michalski et al., 2020; Michalski and Kawate, 2016). Michalski and Kawate identified Tryptophan (W) site Trp74 in ECL1 as a key determinant of Panx1 sensitivity to classical inhibitors (CBX/PBN), with W74A mutations significantly attenuating drug-induced channel blockade (Michalski and Kawate, 2016). Subsequent investigations by the same team revealed that an inter-subunit cation-π interaction between Trp74 and Arg75 in ECL1 is essential for establishing Panx1’s ion selectivity profile (Michalski et al., 2020). And, substitutions at positions Ile247, Val258, and Phe262 in ECL2 were shown to selectively impair CBX-mediated inhibition (Michalski et al., 2020). These findings led to a hypothesis: rather than binding discrete amino acid motifs in ECL1, CBX exerts its inhibitory effect by stabilizing a closed-channel conformation through dynamic interactions bridging ECL1 and ECL2.

The ECLs serve as critical platforms for post-translational glycosylation that dictates subcellular trafficking of Panx proteins. Panx1 undergoes N-linked glycosylation at Asparagine (N) site Asn254 within ECL2, whereas Panx2 and Panx3 are modified at Asn86 (ECL1) and Asn71 within ECL1, respectively (Penuela et al., 2007; Sanchez-Pupo et al., 2018). Glycosylation is indispensable for plasma membrane localization of Panx1 and Panx3. Strikingly, Panx2 maintains constitutive localization to the endoplasmic reticulum (ER) and Golgi apparatus regardless of its glycosylation status (Penuela et al., 2007; Sanchez-Pupo et al., 2018).

Beyond glycosylation, Panx proteins undergo diverse post-translational modifications effecting channel gating and membrane trafficking, including phosphorylation (DeLalio et al., 2019; López et al., 2020; Medina et al., 2021; Metz and Elvers, 2022; López et al., 2021), deacetylation (Chiu et al., 2021), S-nitrosylation (Penuela et al., 2014a; Bunse et al., 2011; Lohman et al., 2012), ubiquitination and lipidation (D'Hondt et al., 2013). Lohman et al reported that S-nitrosylation of Panx1 via nitric oxide (NO) at cysteine (C) sites Cys40 and Cys346 inhibited Panx1 channel currents and ATP release (Lohman et al., 2012).

Panx1 channel could be phosphorylated at tyrosine (Y) sites Tyr150 (Nouri-Nejad et al., 2021), Tyr198 (DeLalio et al., 2019) and Tyr308 (Weilinger et al., 2012) via Sarcoma-family kinase (SFK) and at serine(S) site Ser205 via Salt-inducible kinase (SIK) (Medina et al., 2021), and at T302/S328 via Protein kinase A (PKA) (López et al., 2020), and at Ser394 via Ca²⁺/calmodulin-dependent kinase II (CaMKII) (López et al., 2021). Mutation of Tyr150 prevented glycosylation and trafficking of Panx1 protein (Nouri-Nejad et al., 2021). Tyr198, Tyr308 and Ser205 phosphorylation of Panx1 regulated channel activation. Immunostaining of Tyr198 phosphorylation was enriched in the smooth muscle layer of arteries from hypertensive humans, suggestive a role in hypertensive vascular pathology (DeLalio et al., 2019). Tyr308 phosphorylation was mediated by anoxia-induced NMDA receptor activation in pyramidal neurons (Weilinger et al., 2012). Expressing Ser205A mutated Panx1 mice phenocopied the exacerbated airway inflammation in Panx1 ablation mice (Medina et al., 2021). Mutation of either T302 or S328 prevented Panx1 channel activation induced by the mechanical stretch (López et al., 2020).

However, the authenticity of Panx1 phosphorylation modification has been challenged. Ruan et al reported that the commercially available antibodies against phosphorylation sites of Tyr198 and Tyr308 at Panx1 were non-specific (Ruan et al., 2024). This critical finding necessitates re-evaluation of prior phosphorylation studies reliant solely on immunodetection approaches.

The Panx family exhibits distinct tissue distribution patterns with important functional implications. Panx1, the most widely distributed member among the Panx proteins, is expressed in almost every organ of the human body (Baranova et al., 2004). Panx2 is expressed in the nervous system, thymus, heart, lung, stomach, and spleen (Baranova et al., 2004; Le Vasseur et al., 2014). While Panx3 is found in the skin, ear, mammary gland, testis, blood vessels, small intestine and skeletal system (O'Donnell and Penuela, 2021).

Panx proteins are widely distributed in various cell types, such as epithelial cells, endothelial cells, fibroblast, astrocytes, monocytes, blood cells, platelets, osteoblast, etc. The Panx channels functions related to diseases in different cell types have been summarized systematically in Tables 1, 2.

While Panx proteins were historically characterized as plasma membrane-localized hemi-channels mediating autocrine/paracrine communication (Penuela et al., 2007; Penuela et al., 2009), emerging studies have shown their localization on organelle membranes. Panx1 located in cardiac mitochondria and modulated cardiac susceptibility to ischemia-reperfusion injury (Rusiecka et al., 2023). Panx2 was observed in ER and Golgi apparatus (Sanchez-Pupo et al., 2018), and the contact sites of ER and mitochondria, sensitizing cells to apoptotic stimuli (Le Vasseur et al., 2019). Panx3 was identified in ER, regulating osteoblast differentiation (Ishikawa et al., 2011), and Golgi apparatus, regulating vascular oxidative stress in endothelial cells (Wolpe et al., 2024).

Recent breakthroughs in Cryo-Electron microscopy (Cryo-EM) have resolved the heptameric architecture of Panx channels (Qu et al., 2020; Jin et al., 2020; Zhang et al., 2023; Tsuyama et al., 2025), revealing seven identical subunits arranged around a central pore with diameter constraints permitting selective permeation of molecules up to 1.5 kDa, including amino acids (Lazutkaite et al., 2017), nucleotides (Bao et al., 2004), and ions (Ma et al., 2012). The physiological functions of Panx channels in many organs have been well reviewed., including brain, skin, kidney, lung, heart, bladder, skeletal muscle, and digestive system, including liver and pancreas etc (Giaume et al., 2021; O'Donnell and Penuela, 2023; Abed et al., 2015; Badaoui and Chanson, 2023; Li et al., 2015; Andersson, 2015; Maes et al., 2015a; Maes et al., 2015b; Maes et al., 2014; Cigliola et al., 2015; Riquelme et al., 2013).

However, some details and new research contents still need to be supplemented here. Panx1 channel is also involved in the development of lactational mammary gland (Stewart et al., 2016) and oocyte (Kordowitzki et al., 2021). In pancreatic islet cells, Panx1 was found to be expressed in β and δ cells, and Panx2 in β, α, and δ cells, while Panx3 was absent in pancreas (Scemes et al., 2008; Berchtold et al., 2017; Miranda et al., 2022). Functional studies in INS-1E β -cells demonstrated Panx1 and Purinergic Receptor P2X 7 (P2X7) coupling drives glucose-induced ATP release, stimulating Ca²⁺ transients, cell proliferation and insulin secretion (Tozzi et al., 2018). Panx2 deficiency rendered pancreatic β cells susceptible to cytokine-induced apoptosis in vitro through hyperactivation of iNOS/STATs pathways and impaired glucose tolerance in vivo (Berchtold et al., 2017). Latest research shows that Panx2 deficiency can disrupt visual pathways and cause eye defects in zebrafish (Shanbhag et al., 2025). Panx3 primarily participates in physiological processes including cartilage/bone development, remodeling, and regeneration, as well as pathological conditions such as obesity, osteoarthritis, and degenerative disc disease (O'Donnell and Penuela, 2021; Ishikawa and Yamada, 2017). Emerging evidence reveals that Panx3 collaborates with the Purinergic Receptor P2X 4 (P2RX4) channel to sustain serum antibody titers and promote the survival of bone marrow plasma cells (Ishikawa et al., 2024).

Accumulating data indicate potential functional redundancy may exist among the Panx isoforms. In the skin, the expression level of the Panx3 was significantly upregulated in Panx1 KO mice (Penuela et al., 2014b). Additionally, both Panx1 and Panx3 are capable of recruiting monocytes (Chekeni et al., 2010; Pillon et al., 2014). A study on cerebral ischemic stroke revealed that Panx1 gene KO had no effect on the disease, while dual Panx1 and Panx2 ablation significantly alleviated the obstruction phenomenon (Bargiotas et al., 2011). In cutaneous squamous cell carcinoma (cSCC), Panx1 and Panx3 exhibit divergent tumor-modulatory roles. Panx1 drives tumor progression through ubiquitous overexpression that accelerates malignant proliferation and invasion, whereas Panx3 exerts tumor-suppressive effects (O'Donnell, 2024).

Emerging evidence delineates the functional pleiotropy of Panx channels in tumorigenesis, with isoforms-specific roles spanning tumor-intrinsic programs.

Panx1 channel, particularly in endothelial cells, proximal tubules, and cardiomyocytes, promotes sterile inflammation by releasing ATP. This drives cytokine/chemokine expression, NLRP3 inflammasome activation, and neutrophil infiltration across diverse ischemic and nonischemic organ injuries (Sharma et al., 2018; Jankowski et al., 2018; Yi et al., 2022; Pavelec et al., 2024). During fibrosis, Panx1 promotes macrophage recruitment (Crespo Yanguas et al., 2018). In allergic responses, Panx1 primarily mediates extracellular adenosine accumulation, thereby influencing adaptive immunity by regulating Treg-Teff cell crosstalk (Medina et al., 2021).

Inhibiting Panx1 activity or knock down its expression suppresses expression of proinflammatory cytokine, chemokine and leukocyte adhesion molecule while reducing neutrophils infiltration or macrophage recruitment in ischemic and nonischemic organ injuries, conversely, it aggravates allergy severity.

Panx1 is proposing as a valid immune-related therapeutic target for cancer due to regulating immune cells infiltration especially neutrophils. Tumor associated neutrophils and macrophages are abundantly in tumor-immune microenvironment serve as immunosuppressive functions (Kumar et al., 2016). Neutrophils release their chromatin extracellular and assemble into web-like structural neutrophil extracellular Traps (NETs). NETs involve in immune protection, inflammatory (Papayannopoulos, 2018) and auto-immune diseases (Lee et al., 2017) and cancer (Adrover et al., 2023).

Based on the analysis of whole transcriptome sequencing data, the expression level of Panx1 was related to the immune infiltration, affecting the infiltration of neutrophils, tumor-associated fibroblasts, macrophages, myeloid-derived dendritic cells (MDSCs), and monocytes in PDAC (Bao et al., 2021). Analysis of bulk RNA sequencing data in breast cancer indicated that Panx1 was highly expressed in tissue of basal-like breast cancer, positively correlating with the expression of ENTPD1 (CD39)/NT5E (CD73) in the tumor microenvironment, and showing a significant correlation with tumor-associated neutrophils infiltration (Chen et al., 2022). Mukai et al. demonstrated that Spermidine released by Panx1 channel in apoptotic tumor cells could mediate a new immune-escape mechanism via inducing neutrophil extracellular traps (NETs) and shaping the immune rejection microenvironment (Mukai et al., 2022). Damage Associated Molecular Patterns (DAMPs) refer to endogenous molecules released by dead cells, which are endogenous danger signals. A risk assessment model for PDAC immunotherapy based on DAMPs molecules found that PDAC patients with high DAMPs expression had worse prognosis and immune status. Panx1 is one of the DAMPs molecules, activating the NOD1/NF-κB signaling pathway through ATP release, promoting tumor progression and immune regulation (Wu et al., 2024). In Low-grade glioma (LGG), the expression of Panx2 was negatively correlated with the infiltrating levels of macrophages, dendritic cells, and CD4 T cells (Xu et al., 2021).

Notably, whole-transcriptome analyses identified Panx1 as a ferroptosis- associated genes whose dysregulation correlated with drug resistance, tumor immune infiltration, and tumor stemness in lung adenocarcinoma (LUAD) (Zhang A. et al., 2021; Ren et al., 2021). In LUAD, Panx1 on cancer cells was transcriptionally regulated by p53 protein and assisted to IL-6 production and secretion via ATP/P2 receptors/Ca2+/AKT/NF-κB signaling axis. Secreted IL-6 could promote the polarization of macrophages into M2-like phenotype which enhancing pro-metastatic effects on cancer cells (Phan et al., 2023).

As previously mentioned, Panx channels are involved in inflammatory and cancer processes and playing important functions. Clinical suppression of Panx channels function may be able to assist in disease treatment. Currently, Panx channel inhibitors mainly target the Panx1 channel, and there are no reported inhibitors for Panx2 or Panx3, which is a significant issue that needs to be addressed in future. Koval et al systematically reviewed Panx1 channel blocking reagents, focusing on chemical reagents and small molecular mimetic peptides. Additionally, for clinical drugs, they compared the advantages, the specificity and feasibility of clinical applications (Koval et al., 2023).

At present, the commonly used Panx1 inhibitors for researchers are Probenecid (PBN), Carbenoxolone (CBX), and 10Panx. Firstly, PBN is an FDA-approved drug for the treatment of gout and has been found to be one of the effective drugs blocking Panx1 channel (Silverman et al., 2008). It is also reported as an inhibitor of Urate Transporter 1(URAT1/SLC22A11) (Shin et al., 2011; Ebert et al., 2014), ABCC4/MRP4 (Ventimiglia et al., 2015), Organic Anion Transporters(OATs) (Granados et al., 2022), Organic cation transporters (OCTs) and P2X7 (García-R et al., 2023). Secondly, CBX, historically utilized for gastric ulcer treatment (Henman, 1970), is currently serves as a broad-spectrum inhibitor of both Panx and Connexin channels (Davidson et al., 1986; Bruzzone et al., 2005). Thirdly, 10Panx is a 10-peptide fragment targeting the W74 to Y83 segment of the ECL1 of Panx1 protein, capable of specifically inhibiting Panx1 function (Pelegrin and Surprenant, 2006). 10Panx is a good study tool for the Panx channel function research, but it cannot be used for clinical treatment. Finally, Spironolactone is an effective anti-hypertensive drug and is also one of the inhibitors of Panx1 channel (Good et al., 2018b). Critically, all known pharmacological blockers of Panx1 channels lack selectivity, necessitating the development of novel targeted inhibitors. Moreover, current Panx1 blockers require high concentrations in vitro or doses in vivo to achieve efficacy, which may elicit significant off-target effects (Koval et al., 2023). Particularly, CBX is excluded from contemporary first-line regimens for gastric ulcer management (Kavitt et al., 2019).

Panx1 channel inhibitors play a significant role in the treatment of brain, liver, lung, and kidney injuries. Firstly, following brain ischemia, PBN pretreatment could inhibit reperfusion induced CA1 neuron death. Furthermore, administering PBN continuously for 7 days after reperfusion protected nerve cells from I/R damage (Wei R. et al., 2015). In vitro treatment of isolated posterior cerebral arteries with CBX or Spironolactone significantly dampened the development of brain muscle-derived muscle tone, similar to the phenomenon that application with the Apyrase, an ATP hydrolyzing enzyme (Good et al., 2018a). In vivo administration of 10Panx reduced the volume of brain infarction and alleviated neurological deficits as well as vascular inflammatory damage (Wei et al., 2020).

Secondly, administration of 10Panx in vivo inhibited liver damage caused by excessive use of acetaminophen (APAP), suppressed the recruitment of neutrophils by reducing Panx1 function, improved oxidative state of liver, and promoted tissue regeneration (Maes et al., 2017).

Thirdly, in mice asthma induced by OVA, ATP released by Panx1 worse inflammatory response. Inhibiting Panx1 function with 10Panx reduced airway hyperreactivity in lung (Arzola-Martínez et al., 2021). In addition, pretreatment of CBX or PBN in vivo protected lung tissue from I/R-induced tissue edema, vascular permeability changes, inflammatory factor release, as well as neutrophils infiltration and activation (Sharma et al., 2018).

Fourth, CBX pretreatment protected kidney from I/R injury (Jankowski et al., 2018). Furthermore, Fan Yin et al. supposed Panx1 facilitated acute kidney injury (AKI) via ATP/NLRP3/Caspase11/IL1β signaling axis (Yi et al., 2022). In another research, CBX inhibited Panx1-mediated NLRP3 activation and the release of IL-6, TNF-α, and IL-1β, contributing to the inhibition of renal sepsis injury (Huang et al., 2020).

The application of CBX, PBN, or 10Panx in vitro or in vivo can inhibit the proliferation, migration, invasion, and tumor formation of cancer cells, and even alter their sensitivity to chemotherapy drugs. Firstly, in colorectal cancer, in vitro experiments showed that PBN and 10Panx significantly inhibited the proliferation of colon cancer cells. In vivo experiments indicated that intraperitoneal injection of PBN in mice promoted tumor necrosis and inhibited subcutaneous tumor proliferation (Fierro-Arenas et al., 2024). In PDAC, intratumoral injection 10Panx inhibited UQCRC1 overexpressed tumor growth (Wang et al., 2020).

The application of 10Panx in vitro inhibited the function of Panx1 in platelets, and its culture medium hindered tumor growth, migration, and invasion of co-cultured PDAC cells. In addition, the P-selectin glycoprotein ligand-1 (PSGL-1) is a chemical molecule specifically band to the CD62P on the platelets. Chemically synthesized PSGL-1 with CBX produced PC63435, a specific Panx1 inhibitor band to platelets, blocked the Panx1/IL-1β pathway, suppressed PDAC tumor invasion and metastasis in vivo (Li et al., 2023).

In breast cancer, the truncated form of Panx1 (1–89) promoted the survival of metastatic breast cancer cells and retained sensitivity to CBX, 10Panx, and PBN (Furlow et al., 2015). PBN also blocked the Panx1-mediated EMT signaling pathway, preventing the invasive and metastatic characteristics of breast cancer cells (Jalaleddine et al., 2019). CBX and PBN reduced the invasion and migration capabilities of testicular cancer Tcam-2 cells (Sun Y. Y. et al., 2020) and inhibited cancer cell growth and migration in melanoma (Freeman et al., 2019). In squamous cell carcinoma, PBN and Spironolactone attenuated cell proliferation and migration (O'Donnell, 2024).

In addition, CBX and PBN preferred autophagy in the I-10/DDP drug-resistant cell line of testicular cancer, indicating that blocking Panx1 channel function sensitized tumor cells to DPP treatment (Yuan et al., 2022).