1
P2X7 receptor orchestrates multiple signalling pathways triggering
inflammation, autophagy and metabolic/trophic responses.
Running title: P2X7-activated Intracellular Pathways
Authors: Elisa Orioli*, Elena De Marchi*, Anna Lisa Giuliani*, Elena Adinolfi*§
Affiliations:
*Department of Morphology, Surgery and Experimental Medicine, Section of Experimental Pathology, Oncology and
Biology, University of Ferrara, Ferrara, Italy
§Corresponding author: Department of Morphology, Surgery and Experimental Medicine, Section of Experimental
Pathology, Oncology and Biology, University of Ferrara, Via Luigi Borsari, 46, 44121, Ferrara, Italy; Tel: +39 (0)532
455445; Fax: +39 (0)532 455351: Email: elena.adinolfi@unife.it
Abstract
P2X7 receptor is an ion channel activated by extracellular adenosine trisphosphate (eATP) that attracted increasing
attention for its role in immune reactions, neurobiology and oncology. As receptor for an extracellular ligand, P2X7
activates a series of intracellular signalling pathways mainly via alterations of the ion permeability, but also through
formation of a large unselective pore and direct interaction with other proteins. Here we wish to give an overview on the
main biochemical paths initiated by P2X7 activation by revising recent and established literature on P2X7-triggered
signalling cascades leading to cell death, inflammatory and immune response activation, proliferation and metabolism
modulation. We will focus on the well-known P2X7 inflammasome/NF-kB and pro-apoptotic networks but also cover
P2X7-activated emerging autophagic, pyroptotic and proliferative-oncogenic pathways, like beclin-1/LC3-II, caspase-11,
Akt and VEGF axes.
Keywords: P2X7, ATP, autophagy, pyroptosis, IL-1β, NLRP3, Akt, VEGF
Introduction
P2X7 is a cation channel that, upon engagement by its natural ligand extracellular ATP (eATP), mediates cellular influx
of sodium and calcium and efflux of potassium [1]. Moreover, thanks to the presence of its long intracellular tail and
following exposure to high (mM) agonist concentrations, P2X7 activates the opening of an unselective pore permeable
to solutes in the range of hundreds of Daltons [1-3]. This last feature has been associated to the absence of long-term
desensitization and receptor activated cytotoxicity [1]. The human p2rx7 gene is located at chromosome 12 (12q24.31)
2
at 130 kb from its homologous p2rx4 9826911 [4] and its transcription initiation site corresponds to adenine – 91 [5]. The
originally cloned rat, human and mouse P2X7 genes comprise 13 exons [1, 4, 6], nevertheless, different splice variants
have been identified including extra exons coding for both longer and shorter proteins [7-10]. Most of the P2X7 splice
variants truncated for the C terminal tail were either non-functional or showed limited agonist response and, with the
exception of the human P2X7B isoform, acted as dominant negative regulators of the full-length receptor. Conversely,
the human P2X7B variant, which is a functional ion channel but cannot form the large conductance pore, attracted
attention for its ability to positively modulate P2X7 activities, growth promotion included [11, 12]. More than 1500 human
P2X7 single nucleotide polymorphisms (SNPs) were identified, among which, a small number related to receptor’s
activity including 10 loss and 3 gain of function variants. P2X7 SNPs have been the focus of numerous studies trying to
exploit them as biomarkers in pathologies as different as infectious diseases, oncological conditions, metal illnesses and
many others (reviewed in [13-15]).
Insights on the secondary and tertiary structure of P2X7 receptor came from the publication of the crystallographic data
of the truncated zebra fish P2X4 followed by those of the human P2X3 and the truncated panda P2X7 receptor [16-19].
These studies allowed for prediction of a common P2X family trimeric organization where each of the three receptor’s
subunit includes a large extracellular loop, responsible for agonists and antagonists binding, two short transmembrane
domains, N and C termini, variable in length among receptor subtypes [14, 20]. Due to the similarity between P2X
monomer and a jumping dolphin, the receptor subunits domains have been named according to the body parts of this
animal: the extracellular region being composed by the head, the body, the dorsal fin and the flippers, while the
transmembrane helices include the tail and the fluke. Beta-sheets, organized in a β sandwich structure, form the body
region responsible for the interactions among the three subunits, which form a P2X channel [4, 17, 20]. The P2X7 subunits
have the longest C terminal domain of the P2X family, making P2X7 the highest molecular weight P2X protein accounting
for total 595 residues. No structural information is currently available on P2X7 C-terminal region although, nevertheless,
based on its sequence it was possible to predict putative binding sites, including one for lipopolysaccharide [21].
Contemporarily, different proteins interacting with P2X7 C-terminal tail were identified either by immunoprecipitation
[22-25] or with a yeast two-hybrid strategy [26], including transmembrane, intracellular and heat shock proteins (HSP),
cytoskeletal elements and kinases, and the P2X4 receptor. Among these proteins, pannexin-1 plasma membrane
hemichannels have been claimed to mediate P2X7-dependent pore opening, IL-1β and ATP release [27, 28]. Many other
P2X7 interactors have been identified that are listed and classified in the P2X7 interactome database
(http://www.p2x7.co.uk/).
P2X7 ligands including agonists, antagonists and allosteric modulators are either present in nature or were synthetically
developed [14, 29, 30]. The P2X7 receptor shows low affinity (i.e. in the mM range) for eATP, while the best known,
albeit not completely selective, synthetic agonist is 2'(3')-O-(4-Benzoylbenzoyl) adenosine-5'-triphosphate (Bz-ATP).
Other naturally occurring molecules have been attributed a role as P2X7 agonists and positive allosteric modulators
including the cathelicidin derived peptide LL-37 [31, 32] and Alzheimer’s β amyloid 19299738 [33]. Several inhibitors,
antagonists or negative allosteric modulators have been developed, some of them administered to patients with a good
safety profile [34-36]. The classically used blockers include oxidized-ATP (oATP), brilliant blue G and KN62, which
have been recently substituted in in vitro and in vivo studies by more potent and selective antagonists/allosteric
modulators, such as AZ10606120, A740003 and its derivative A438079 [14, 37-39]. Further P2X7 positive allosteric
modulators, which have been shown to increase P2X7-dependent membrane permeability to large solutes, include drugs,
antibiotics and plant derived products, such as tenidap [40], polymixin B 15383600 [41] and ginsenosides [42].
3
P2X7-triggered intracellular signalling was attributed not only to ion fluxes and interacting proteins but also to activation
of downstream proteases, kinases and nuclear factors. The biophysical and structural properties of P2X7 and its role as
possible therapeutic target in inflammatory disorders and cancer have been recently reviewed [14, 43-47]. Here we wish
to give an overview of the main signalling pathways activated by P2X7 receptor, due to its channel-pore activity or via
interacting proteins. We describe the sequence of intracellular events initiated by P2X7 and leading to cell death,
autophagy and activation of inflammatory, metabolic and proliferative pathways.
P2X7 receptor activates death pathways leading to apoptosis, autophagy and pyroptosis
P2X7 is a molecule able to induce cell death via cell-specific downstream signalling events. Cytotoxic properties are
mainly due to its ability to form membrane pores permeable to large molecular weight molecules [48-50]. P2X7 pro-
apoptotic/necrotic activities have been extensively appraised elsewhere [51, 52]. Here we will briefly cover the most
salient mechanisms relating P2X7 to apoptotic cell death, besides to emerging evidence on autophagy and pyroptosis.
Among apoptotic-related alterations P2X7 pore opening associates with cell morphological changes like cell blebbing
and shrinkage, nuclear fragmentation and chromatin condensation [53-56] [Fig. 1]. Depending upon the cleaved caspase,
apoptosis can be initiated via the extrinsic (caspase-8) or the intrinsic (caspase-9) pathway, both triggered by P2X7 [56-
59]. Moreover, P2X7 is known to cause mitochondrial potential collapse and fragmentation followed by cytochrome c
release into the cytosol and ROS production [58, 60, 61]. Interestingly, this last mechanism is reversible upon brief (5-10
minutes) receptor activation. On the other hand, prolonged P2X7 opening triggers cell death via ROCK-1 [58].
Accordingly, the notion that P2X7 activation can be uncoupled from apoptosis and cytotoxicity has been confirmed by
increasing number of studies including those associating P2X7-dependent pore formation and autophagy [2, 62-65].
Autophagy is a self-degradative process indispensable for the equilibrium of energy sources and removal of misfolded
proteins or damaged organelles, involved in several pathological conditions [66] [Fig. 1]. A central player in autophagy
is the autophagosome that mediates lysosomal targeting and degradation of intracellular debris. Commonly used markers
of autophagy include beclin-1 activation, which triggers autophagosome aggregation, and LC3-II association to the
autophagosome. Sun and colleagues proposed P2X7-coupling to autophagy consequent to traumatic brain injury [67].
These authors suggested that oATP administration would prevent brain injury-associated cognitive and memory
impairments, by reducing beclin-1 levels, thus implying P2X7 as a positive regulator of the authophagic process [67]. On
the same line, Young and colleagues demonstrated that in DMD mdx dystrophic-mice myoblasts, P2X7 triggers an
autophagic flux starting with increased LC3-II aggregation, which co-localizes with the regions of membrane
permeabilization to the large solute Lucifer yellow, i.e. to the P2X7 pore formation sites. In WT and DMD mdx myoblasts
P2X7 activation causes increase of LC3-II that can be actively modulated by both upstream and downstream autophagy
inhibitors [62]. Moreover, dystrophic myoblasts showed increased LC3-II levels compared to the WT counterpart.
Interestingly, in this model P2X7-dependent autophagy was not mediated by calcium but was subordinated to receptor’s
interaction with HSP90 [62], previously shown to be part of the P2X7 membrane complex [22, 23]. In dystrophic
myoblasts, HSP90 inhibition with geldanamycin caused significantly reduced P2X7 pore formation leading to decreased
autophagy [62]. Geldanamycin-dependent reduced P2X7 pore opening seems to be myoblast specific as it was not
reproduced neither in macrophages [62] nor in P2X7-transfected HEK293 cells [23, 68]. In this last model, geldanamycin
treatment rather caused increased P2X7 activity via reduction of the HSP90 tyrosine phosphorylated form [23]. These,
apparently contrasting, data may reflect, at least in part, the cellular context involved and, therefore, could be helpful in
4
designing dystrophy targeted P2X7 treatments [23, 62]. These may prove beneficial also in light of recent reports linking
P2X7 blockade or ablation to reduction of dystrophic symptoms in mdx mice [69-71]. Increasing evidence associates
autophagy to an efficacious defence against intracellular pathogens such as mycobacteria [72, 73]. P2X7 plays a role also
in this phenomenon by favouring the elimination of infected macrophages and mycobacteria via autophagolysosome
formation [74]. However, while P2X7 generally increases the first phases of autophagy favouring the formation of
autophagosomes, in some models the receptor also leads to lysosomal function impairment thus causing a blockade of
the overall autophagic flux [75].
P2X7 also triggers pyroptosis, a form of cell death occurring in immune cells via activation of inflammatory caspases
[76, 77]. This is not surprising since, in monocyte/macrophage cell lineage, both P2X7 activation and pyroptosis cause
blebbing, death, IL-1β and IL-18 secretion [52, 78-81]. Pyroptosis requires two particular stimuli, PAMPs (pathogen
associated molecular patterns) and DAMPs (damage associated molecular patterns) [78, 82]. ATP is a recognized DAMP
that, through P2X7, in dendritic cells and macrophages, mediates the activation of the main pyroptotic caspase, i.e.
caspase-1 [80, 83] [Fig. 1]. As above mentioned caspase-1 is not the only caspase required for the initiation of pyroptosis,
as one other member of the inflammatory caspases, including caspase-11 in mice and caspase-4 and caspase-5 in humans
are necessary [81]. Yang and colleagues have recently shown that P2X7 and pannexin-1 are required for LPS-mediated
pyroptosis involving caspase-11 [76]. In LPS treated bone marrow derived macrophages caspase-11 partially cleaves
pannexin-1 causing P2X7 activation via eATP release leading to NLRP3 assembly, caspase-1 activation, IL-1β release,
gasdermin D cleavage and pyroptosis [76]. These data are also supported by reduced sensitivity to IL-1β-dependent
endotoxic shock of P2X7, pannexin-1 or caspase-11 null mice [76].
P2X7-related inflammatory pathways
Once established that P2X7 sustains the maturation and release of IL-1β via K
+
efflux [83] it was soon evident that the
receptor was involved in the activation of the IL-1β maturation platform: the NLRP3 inflammasome [84] [Fig. 2].
Inflammasomes are high molecular weight protein complexes formed in the cytosolic compartment in response to PAMPs
or DAMPs, constituted of a sensor and an adaptor molecule plus caspase-1 [85]. The synthetic process of immature, full
length, pro-IL-1β originates from the recognition of invading organisms-derived PAMPs. Once synthesised, pro-IL-1β
undergoes a proteolytic cleavage by caspase-1 to generate a mature active cytokine. The activity of caspase-1, in turn,
depends on assembly of the NLRP3 inflammasome, which converts pro-caspase-1 into a functioning enzyme [86-88].
The most common mechanism of inflammasome activation is the change in the microenvironment ionic composition with
the decrease in K
+
concentration [89, 90]. Indeed, in 2008 Rubartelli and colleagues first associated P2X7-dependent K
+
efflux with NLRP3 inflammasome activation, demonstrating that different danger signals can induce ATP secretion that,
in turn, interacting with P2X7, causes IL-1β and IL-18 maturation [84]. This notion has been confirmed in several
following papers (revised in [85]) demonstrating that P2X7 triggers NLRP3 inflammasome aggregation and function in
macrophages, dendritic and peripheral blood mononuclear cells, neutrophils, microglia and possibly in astrocytes [91-
95]. Depending upon cell type, P2X7 inflammasome interaction mediates secretion not only of IL-1β and IL-18 but also
of other pro-inflammatory cytokines and alarmins such as IL-6 and IL-1α [91, 96].
Three main mechanisms have been proposed to explain P2X7-dependent trigger of NLRP3 inflammasome: K
+
efflux,
ROS production and via direct interaction. Although the requirement of K
+
efflux is well established [77, 97, 98], little is
5
known about the mechanism by which K
+
concentration drop is translated into intracellular signals. Recent studies have
partially elucidated this question suggesting that NEK7 kinase will sense the ATP-induced K
+
decrease and bind to
NLRP3 inflammasome regulating its function [99, 100]. In addition, Yaron and colleagues have recently suggested that
P2X7-dependent K
+
efflux would be leading mitochondrial ROS generation, via organellar Ca
2+
influx, causing
inflammasome activation [101]. P2X7-dependent ROS production via NADPH activation is an established data [102-
104] that has been shown to be dependent upon the MEK-ERK signalling network [104] to cause IL-1β secretion [103,
105, 106] and autophagy activation [107]. P2X7 engagement by ATP also determines uncoupling of the thioredoxin
(TRX)/TRX reductase (TRXR) system eliciting ROS-dependent inflammasome recruitment [108]. In this context, the
P2X7/P2X4 complex plays a controversial role as P2X4 has been claimed to both up-modulate and decrease ROS
formation and inflammasome activation through interaction with P2X7 [106, 107]. Finally, emerging evidence suggests
that P2X7 could directly interact with components of different inflammasomes including NLRP2, ASC and NLRP3 [95,
109, 110]. The presence of P2X7 receptor in the same protein complex with NLRP2 and ASC will explain an
unconventional IL-1β maturation in non-immune cells such astrocytes [95]. In a more conventional situation, such as
peritoneal macrophages, P2X7-NLRP3 interaction drives NLRP3 association to the interior leaflet of plasma membrane,
via discrete Ca
2+
concentration increase. Calcium influx also induces the progressive accumulation of P2X7 at the
NLRP3/P2X7 membrane complex, finally resulting in a striking rise of the two proteins in a defined plasmalemmal region
[109]. Despite interaction, NLRP3/P2X7 crosstalk is complicated by an inverse correlation of their expression. Indeed,
P2X7 null mice express enhanced NLRP3 levels [109] whereas, on the contrary, NLRP3 KO animals over-express P2X7
[110]. Interestingly, this inverse relation was also found in chronic lymphocytic leukemia patients, which while over-
expressing P2X7, showed dramatically reduced NLRP3 levels that correlated with increased cell survival/proliferation
[110]. Hematologic malignancies are not the only pathologic situation exacerbated or induced by the P2X7-
inflammasome axis. Indeed, a growing body of literature reports a role for P2X7/NLRP3 and related proteins in several
pathological conditions including response to infective agents [111-114], depressive and gestational disorders [115-119],
epilepsy [120], fibrosis [121], osteonecrosis [122], liver, kidney and brain injury [123-126].
The relevance of P2X7 is not limited to IL-1β maturation and secretion but also extended to activation of NF-kB, the
main nuclear factor responsible for pro-IL-1β synthesis. NF-kB is a transcription factor covering a pivotal role in
modulating cytokine expression in response to different stimuli such as TNF-α, IL-1β and TLRs agonists, relevant in cell
growth and carcinogenesis, too [127-129]. The five members of the mammalian NF-kB family are present in unstimulated
cells as homo- or hetero-dimers bound to IkB family proteins [130]. Binding to IkB prevents the NF-kB/IkB complex
from translocation to the nucleus, thus maintaining NF-kB inactive. On the contrary, free NF-kB dimers are able to
translocate to the nucleus leading to a robust proinflammatory gene expression [131-133]. P2X7 activation by eATP has
been included among various stimuli of the NF-kB pathway for a long time [134]. Activated P2X7 was indeed, found to
induce NF-kB pathway activation resulting in selective DNA binding of NF-kB subunit p65 in microglia, osteoclasts and
osteoblasts [134-136]. A following study suggested P2X7 as responsible for NF-kB activation through the MyD88
pathway [137] since MyD88 silencing almost abolished P2X7-induced NF-kB activation. Co-immunoprecipitation
experiments suggested a direct P2X7/MyD88 interaction, which was lost upon P2X7 C-terminal region truncation [137].
Of interest, P2X7 antagonist administration in murine models of inflammatory diseases as colitis and liver fibrosis,
resulted in striking reduction of symptoms and NF-kB levels [138, 139]. Moreover, different pro-inflammatory agents
were shown to increase P2X7/NF-kB levels [140, 141] while anti-inflammatory substances exerted an opposite response
