P2X7 receptor activation regulates rapid unconventional export of transglutaminase-2

Abstract

Transglutaminases (denoted TG or TGM) are externalized from cells via an unknown unconventional secretory pathway. Here, we show for the first time that purinergic signaling regulates active secretion of TG2 (also known as TGM2), an enzyme with a pivotal role in stabilizing extracellular matrices and modulating cell-matrix interactions in tissue repair. Extracellular ATP promotes TG2 secretion by macrophages, and this can be blocked by a selective antagonist against the purinergic receptor P2X7 (P2X7R, also known as P2RX7). Introduction of functional P2X7R into HEK293 cells is sufficient to confer rapid, regulated TG2 export. By employing pharmacological agents, TG2 release could be separated from P2X7R-mediated microvesicle shedding. Neither Ca(2+) signaling alone nor membrane depolarization triggered TG2 secretion, which occurred only upon receptor membrane pore formation and without pannexin channel involvement. A gain-of-function mutation in P2X7R associated with autoimmune disease caused enhanced TG2 externalization from cells, and this correlated with increased pore activity. These results provide a mechanistic explanation for a link between active TG2 secretion and inflammatory responses, and aberrant enhanced TG2 activity in certain autoimmune conditions.

Keywords: Extracellular matrix stabilization; Innate immunity; P2X7 receptor; Purinergic signaling; Transglutaminase; Unconventional protein secretion.

Fig. 1.

P2X7R inhibition blocks TG2 secretion by macrophages. (A) Differentiated monocytes express TG2. THP-1 cells were differentiated for the indicated time with TPA and stimulated with LPS as indicated. Cell extracts were analyzed by western blotting for TG2 or β-tubulin as a loading control (*, non-specific reactivity). (B) TG2 export requires P2X7R activity. Differentiated THP-1 cells were pre-treated with vehicle or 5 μM P2X7R inhibitor A740003 for 10 min, then stimulated as indicated with BzATP for 10 min with or without inhibitor (inh, pulse). Cells were chased for 30 min in P2X7R agonist and antagonist-free medium. Collected medium from the pulse and chase (200 µl) were rendered cell-free by centrifugation, and analyzed for TG2 by western blotting. (C,D) P2X7R activation triggers TG2 secretion in macrophages derived from peripheral blood mononuclear cells. Macrophages were stimulated with BzATP and chased as in B, and collected medium from the of pulse and chase were analyzed for TG2 by western blotting alongside the cell lysates (C). The presence of 100 µM Ac-YVAD-CMK did not prevent externalization or cleavage of TG2, indicating a caspase-1-independent process (D).

Fig. 2.

P2X7R activation mediates TG2 externalization. (A) Analysis of TG2 secretion in HEK293 P2X7R cells. TG2-transfected cells were stimulated with BzATP or vehicle for indicated time (pulse), then incubated for 30 min in agonist-free medium (chase). TG2 secretion into cell-free supernatants was assessed by western blotting. (B) Inhibitor A740003 reversibly blocks P2X7R activation. P2X7R cells were incubated with Fluo-4-AM and 5 μM P2X7R inhibitor for 20 min prior to BzATP stimulation in the presence of inhibitor (top), washed with inhibitor-free medium for 5 min, and then re-stimulated with BzATP (bottom). The fluorescence (λ_ex, 488 nm; λ_em, 500–535 nm) change in individual cells was monitored by confocal microscopy (mean±s.e.m., n=30) (right). Optical sections of the same field before and 180 s after BzATP addition are shown (left). Scale bar: 25 µm. (C) P2X7R inhibitor (inh) blocks TG2 secretion. TG2-transfected P2X7R cells were pre-treated with P2X7R inhibitor or vehicle for 10 min before BzATP stimulation as indicated. TG2 release into medium was assessed as in A. (D,E) Cells release membrane-bound particles upon P2X7R activation. TG2 transfected P2X7R or parental cells were BzATP stimulated for 10 min, and chased in agonist-free medium. Conditioned media and cell lysate were analyzed by western blotting for TG2 and the microvesicle marker flotillin-2 (D) or, as a control, β-tubulin, IκBα and HMGB-1 (E).

Fig. 3.

Membrane blebs induced by P2X7R activation contain TG2. (A) P2X7R signaling induces rapid membrane blebbing. Fluo-4-AM-loaded P2X7R cells were stimulated with BzATP while acquiring fluorescence and phase-contrast images by real-time microscopy to visualize morphological changes and Ca²⁺ signaling simultaneously (top). Membrane blebs are indicated by arrows. ATP stimulation of parental cells induces oscillating Ca²⁺ signals but no overt morphological changes (bottom). Scale bar: 25 µm. (B,C) TG2 redistributes into membrane blebs. To confirm export of tagged TG2, TG2- (wild-type, WT) or TG2–GFP-expressing P2X7R cells were stimulated with 100 µM BzATP for 10 min, chased for 30 min in agonist-free medium, followed by analysis of conditioned media and cell extracts for TG2 by western blotting (B). To localize GFP-tagged TG2 during BzATP stimulation, real-time confocal microscopy was employed. Genesis of a membrane bleb is depicted (arrows), with an optical section of GFP fluorescence overlaid onto phase-contrast images to correlate morphological changes with changes in TG2 distribution (C). Scale bar: 10 µm.

Fig. 4.

P2X7R-mediated TG2 export is not due to microvesicle release. (A,B) Analysis of vesicle release by nanoparticle tracking. TG2- or mock-transfected P2X7R cells were stimulated with BzATP for 10 min, chased for 30 min in agonist-free medium, and conditioned media were analyzed for nanoparticles using light scattering in combination with particle tracking (Nanosight). Particle distribution and total particle concentration is shown (mean±s.e.m.; n=5) (A). Particles were broadly assigned to one of four fractions based on volume: representing exosomes (∼60 nm; ≤80 nm diameter), microvesicles (∼145 nm; 81–262 nm), larger vesicles (∼335 nm; 263–425 nm) and aggregates or membrane blebs (≥426 nm) (B). (C,D) Analysis of isolated microvesicles for TG2. Cell-free medium (S1) from BzATP- or control-treated cells was subjected to differential centrifugation (P, pellet; S, supernatant): in C, 3000 g twice (P2, P3), 10,000 g (P4), and 100,000 g (P5, S5), and in D, 3000 g followed by separation of microvesicles (MV) on a sucrose cushion. Fractions were analyzed for TG2 by western blotting.

Fig. 5.

Extracellular Ca²⁺ regulates TG2 secretion. (A) P2X7R-mediated TG2 export at different [Ca²⁺]_ex. P2X7R cells expressing TG2 were stimulated with BzATP for 10 min in medium containing 0.9 or 2.2 mM Ca²⁺ or in Ca²⁺-free medium, and chased for 30 min in respective media without BzATP. Conditioned media were analyzed by western blotting for TG2 and flotillin-2. (B) TG2 catalytic activity is not required for P2X7R-mediated export. P2X7R cells expressing TG2 or the TG2 C277S mutant were stimulated with BzATP in medium containing 0.9 or 2.2 mM Ca²⁺ and TG2 export was assessed as above. (C,D) [Ca²⁺]_ex regulates P2X7R activity. P2X7R or parental cells were stimulated with BzATP, as indicated, in PSS containing YO-PRO1 and different concentrations of Ca²⁺. To determine YO-PRO1 uptake by cells after BzATP application, changes in well-specific fluorescence (λ_ex, 480-10 nm; λ_em, 520-10 nm) were monitored over time. A representative experiment of dye uptake in Ca²⁺-free PSS is shown as mean±s.e.m. of two wells (C). In D, the initial rates of YO-PRO1 uptake at different [Ca²⁺]_ex in response to 300 μM BzATP are given (mean±s.e.m.; n=2).

Fig. 6.

TG2 export is independent of K⁺ efflux and membrane depolarization. (A) Calmidazolium (calm) blocks flotillin-2 but not TG2 release. TG2-transfected P2X7R cells were pre-treated for 10 min and then stimulated with BzATP in medium containing 1 µM calmidazolium or vehicle. Cells were chased in agonist-free medium, and conditioned media analyzed by western blotting for TG2 and flotillin-2. (B) Calmidazolium does not affect P2X7R-dependent ‘membrane pore’ formation. P2X7R cells were pre-treated with calmidazolium, P2X7R inhibitor A740003 or vehicle for 10 min prior to stimulation with 100 µM BzATP in the presence of respective inhibitors or carrier in PSS containing YO-PRO1 and 0.9 mM Ca²⁺. Dye uptake was monitored over time. Results are shown as mean±s.e.m. of two wells, and are representative of three independent experiments. (C) Calmidazolium ameliorates the large rise in [Ca²⁺]_i. Fluo-4-AM-loaded P2X7R cells were pre-treated with calmidazolium, P2X7R inhibitor or vehicle for 20 min prior to stimulation with 100 μM BzATP in the presence of inhibitors or carrier. Fluorescence change (λ_ex, 485-12 nm; λ_em, 520-10 nm) relative to control in response to agonist treatment was monitored (mean±s.e.m. of eight replicate wells).

Fig. 7.

P2X7R-mediated membrane pore formation is required for TG2 externalization. (A) P2X7R-mediated pore formation is pannexin independent. P2X7R cells were pre-treated with ¹⁰Panx or trovafloxacin (Trova) as indicated, and then stimulated with BzATP in PSS with respective inhibitors, YO-PRO1 and 0.9 mM Ca²⁺. Results are given as initial rates of dye uptake relative to control (mean±s.e.m.; n=3). Pannexin inhibitors did not affect dye uptake, neither at limiting nor saturating agonist concentration. (B,C) Characterization of expression of mutant P2X7Rs. Extracts of cells stably expressing wild-type (wt), A348T or P451L P2X7R, or the P2X7R variant B (varB) were analyzed by western blotting with antibodies against the P2X7R extracellular domain and β-tubulin, as a loading control (B). Membrane localization of receptor was confirmed by immunocytochemistry (C; compare to Fig. S2B). Images reflect an optical section acquired by confocal microscopy. Scale bar: 12.5 µm. (D,E) Pore formation is enhanced in cells expressing P2X7R A348T. YO-PRO1 uptake following stimulation of cells with 100 µM BzATP is shown as mean±s.e.m. (n=3) fluorescence (D). Comparison of initial rate of YO-PRO1 uptake for P2X7R-A348T- and P451L-expressing cells highlights increased pore activity for P2X7R A348T but unchanged ligand regulation (E). Results are mean±s.d. (n=2). (F–H) TG2 export correlates with receptor pore activity. TG2-transfected cells expressing P2X7R variants were stimulated with BzATP for 10 min, and chased in agonist-free medium. Conditioned media were analyzed by western blotting for TG2 (F), and results (mean±s.e.m., n=3) quantified by densitometry (G). Note, cell lysates confirm comparable TG2 expression levels in different cell lines (F). For thioredoxin-1 (Trx) detection, media (P2X7R cells) were analyzed by western blotting after separation in 16% SDS-PAGE Tricine gels (H).

Fig. 8.

Mechanism controlling TG2 export. Schematic showing different events occurring upon P2X7R activation by ATP. (A) Ion channel activity triggers intracellular signaling that results in actin reorganization and microvesicle shedding. However, these microvesicles do not contain TG2. (B) Coupling between P2X7R and pannexin-1 triggers hemichannel pore opening. TG2 secretion is unaffected by blocking pannexin-1 channels. (C) P2X7R itself can form a membrane pore through conformational changes and, possibly, receptor oligomerization in a process that involves the extended intracellular C-terminal sequence. TG2 secretion is associated with this membrane pore activity but independent of ion channel function, and occurs in conjunction with thioredoxin-1 (Trx) externalization. As thioredoxin can reactivate TG2 functionally blocked in an oxidized state, this might ensure that externalized TG2 has transamidation activity. Flot2, flotillin-2.