For platelet isolation apyrase (Grade II from potato, #A7646, Sigma-Aldrich) and prostacyclin (#P5515, Calbiochem) were used. Platelets were activated with collagen-related peptide (CRP; Richard Farndale, University of Cambridge, Cambridge, United Kingdom), adenosine diphosphate (ADP; #A2754, Sigma-Aldrich), the thromboxane A2 analogue U46619 (U46; #1932, Tocris), PAR4 peptide (PAR4; St. Louis, Missouri, MO, United States), thrombin (Thr; #10602400001, Roche Diagnostics, Germany), lipopolysaccharide (LPS, #L4524, Sigma-Aldrich) or tumor necrosis factor alpha (TNFα, #300-01A, Preprotech). For immunoblotting antibodies targeting Pannexin-1 (Panx1, #91137S, Cell Signaling, 1:1,000) and phosphorylated Panx1 (pPanx1 Tyr¹⁹⁸, #ABN1681, Merck, 1:1,000), glyceraldehyde 3-phosphate dehydrogenase (GAPDH, #2118S, Cell Signaling, 1:1,000) and horseradish peroxidase (HRP)-linked anti-rabbit secondary antibodies (#7074S, Cell Signaling, 1:2,500) were used. For flow cytometric analysis fluorophore conjugated antibodies labelling P-selectin (CD62P, Wug. E9-FITC, #D200, Emfret Analytics, 1:10), active integrin αIIbβ3 (JON/A-PE, #D200; Emfret Analytics, 1:10), GPIbα (CD42b, Xia. G5-PE, #M040-2 Emfret Analytics, 1:10), GPVI (JAQ1-FITC, #M011-1, Emfret Analytics, 1:10), integrin α5 (CD49e, Tap. A12-FITC, # M080-1, Emfret Analytics, 1:10), integrin β3 (CD61, Luc. H11-FITC, #M031-1, Emfret Analytics, 1:10) Ly6G (1A8-APC, #127614, BD Biosciences, 1:30) and phosphatidylserine (PS) exposure (CyTM 5 Annexin V, #559934, BD Bioseciences, 1:10) were used. Immunofluorescence staining was performed using protein blocking solution (#X0909, Dako) antibodies against GPIbα (CD42b, #M042-0, Emfret Analytics, 1:50) and Ly6G (#551459; BD Pharming, 1:100), biotinylated secondary antibodies (#BA-9400, Vector, 1:200), streptavidine FluorTM 660 conjugates (#50-4317-80, Thermo Fisher Scientific, 1:20) and DAPI (#10236276001, Roche, 1:3,000) was used. MHEC5-T cells were cultivated in DMEM high glucose cell culture medium (#41965062, Gibco).

Fresh citrate-anticoagulated blood (BD-Vacutainer^®; Becton, Dickinson and Company; #367714) was collected from abdominal aortic aneurysm (AAA) patients before open surgery or endovascular aortic repair (EVAR). Blood samples of healthy volunteers with an age ≥60 years served as age-matched controls (AMCs). All AAA patients and healthy volunteers provided their written informed consent to participate in this study according to the Ethics Committee of the University Clinic of Duesseldorf, Germany (2018-140-kFogU, study number: 2018064710; biobank study number: 2018-222_1 and MELENA study: 2018-248-FmB, study number: 2018114854). This study was conducted according to the Declaration of Helsinki and the International Council for Harmonization Guidelines on Good Clinical Practice. The procedure of this study was also approved by the Ethics Committee of the University Clinic of Duesseldorf, Germany.

Citrate-anticoagulated blood (BD-Vacutainer^®; Becton, Dickinson and Company; #367714) was centrifuged at 200 × g for 10 min at room temperature (RT). After centrifugation, the platelet-rich plasma (PRP) was collected and added to PBS pH 6.5 (supplemented with 2.5 U/mL apyrase and 1 μM PGI₂) in a volumetric ratio of 1:1, followed by a centrifugation step at 1,000 × g for 6 min. The platelet pellet was resuspended in Tyrode´s buffer pH 7.4 (containing 140 mM NaCl₂, 2.8 mM KCl, 12 mM NaHCO₃, 0.5 mM Na₂HPO₄, 5.5 mM Glucose). The platelet cell number was measured using a hematology analyzer (Sysmex - KX21N, Norderstedt, Germany) and adjusted for the following experiments. For human plasma preparation, fresh citrate-anticoagulated blood was centrifuged for 10 min at 1,500 × g at 4°C. The platelet free plasma was collected and stored at −70°C until use.

Human platelets were isolated as described in 4.3. Isolated platelets (40 × 10⁶) were stimulated with ADP (10 µM), respectively CRP (0.1 or 1 μg/mL) for 10 min at 37°C. Platelet stimulation was terminated by centrifugation at 800 × g at 4°C for 5 min. Platelet releasates were collected and analyzed for soluble Panx1 via ELISA. The platelet pellet was lysed by adding 1x human lysis buffer (100 mM Tris-HCl, 725 mM NaCl₂, 20 mM EDTA, 5% TritonX-100, complete protease inhibitor). Lysis was performed for 15 min at 4°C. Afterwards, 6x Laemmli buffer was added to the platelet lysates, followed by a denaturation step at 95°C for 5 min. The platelet lysates were stored at −70°C until use. For immunoblotting the platelet lysates were separated by a SDS-polyacrylamide gel electrophoresis (12% SDS-polyacrylamide gel) and electro-transferred onto a nitrocellulose blotting membrane (GE Healthcare Life sciences). After blotting, the membrane was blocked with 5% non-fat dry milk in TBST (Tris-buffered saline, containing 0.1% Tween20) for 1h. The membrane was incubated with primary antibodies against Panx1 (#91137S, Cell Signaling, 1:1,000), respectively phospho Panx1 Tyr¹⁹⁸ (#ABN1681, Merck, 1:1,000) at 4°C overnight and with an appropriate HRP-conjugated secondary antibody (#7074S, Cell Signaling, 1:2,500) for 1h at RT. GAPDH (#2118S, Cell Signaling, 1:1,000) served as loading control. Bands were visualized by chemiluminescence detection reagents (BioRad, #1705061) and imaged using a FusionFX Chemiluminescence Imager System (Vilber). Quantification of immune-reactive band intensities was performed by using the FUSION FX7 software (Vilber).

Pathogen-free Panx1 ^fl/fl mice were provided by Dr. Brant Isakson (University of Virginia, Charlottesville, VA, United States) and cross-bread with PF4-Cre mice, that were purchased from the Jackson Laboratory (C57BL/6-Tg [Pf4-cre] Q3Rsko/J), to generate megakaryocyte/platelet specific deletion of Panx1 as described earlier (Metz and Elvers, 2022).

The PPE surgery was performed only in male mice aged 10-12 weeks. All other experiments were conducted with male and female mice with an age of 2–4 months. Mice were maintained in an environmentally controlled room at 22 ± 1°C with a 12 h day-night cycle. Mice were housed in Macrolon cages type III with ad libitum access to food (standard chow diet) and water. All animal experiments were conducted according to the Declaration of Helsinki and approved by the Ethics Committee of the State Ministry of Agriculture, Nutrition and Forestry State of North Rhine-Westphalia, Germany (Reference number: AZ 81-02.05.40.21.041; AZ 81-02.4.2018. A409).

The porcine pancreatic elastase (PPE) perfusion model represents an experimental in vivo model to investigate AAA formation and progression in mice by the infusion of porcine pancreatic elastase into the abdominal aorta. Therefore, mice were anesthetized with 2%–3% isoflurane. Additionally, all mice received a locally subcutaneous (s.c.) injection of buprenorphine (0.1 mg/kg body weight) 3 min prior surgery. The anesthesia was continuously monitored throughout the whole procedure. After laparotomy, an aortotomy above the iliac bifurcation was conducted following the temporal ligation of the proximal and distal infrarenal aorta. Afterwards a catheter was inserted at the distal part of the previously conducted aortotomy, to perfuse the infrarenal aortic segment for 5 min with sterile isotonic saline (NaCl₂, 0.9%) containing type I porcine pancreatic elastase (2.5 U/mL, Sigma- Aldrich) at 120 mmHg. After perfusion, the catheter was removed and the aortotomy was closed. After surgery all mice received subcutaneous injection of buprenorphine (0.1 mg/kg body weight) every 6 h within the daylight phase for the following 3 days. Additionally, all operated mice received buprenorphine (0.3 μg/mL in H₂O) in the drinking water for a time period of 3 days after surgery. Animals were euthanized to collect aortic tissue and blood samples at day 3, 7, 14, respectively 28 after PPE surgery. The PPE surgeries were conducted by Dr. med. Joscha Mulorz (Department of Vascular- and Endovascular Surgery, University Hospital Duesseldorf) and Julia Odendahl (Department of Cardiology, Pulmonology and Vascular Medicine, University Hospital Duesseldorf; S1-Project TRR259).

For monitoring of AAA development and progression after PPE surgery, the dilation of the abdominal aorta was analyzed by measuring the aortic diameter within the aneurysm segment via Ultrasound. Ultrasound measurements were performed prior surgery (baseline) and at day 3, 7, 14, 21, and 28 after PPE. For ultrasound imaging, all mice were anesthetized with 2%–3% isoflurane and positioned on a heating plate at 37°C. All ultrasound measurements were conducted using a Vevo 2,100^® High-Resolution In Vivo Micro-Imaging System (VisualSonics). Inner aortic diameter and aortic wall thickness were measured using a standardized imaging algorithm with longitudinal B-Mode imaging during the systolic phase.

Mice were kept under constant anesthesia with isoflurane and were euthanized by terminal heart puncture followed by cervical dislocation. For platelet and plasma preparation blood samples were collected. After opening the thorax, the vascular system was rinsed with cold heparin solution (20 U/mL; 4°C) to avoid clot formation. Therefore, the heart was punctured with a butterfly cannula at the apex cordis of the left ventricle. The right atrium was incised and approximately 20 mL of heparin solution were perfused under constant pressure through the vascular system. Afterwards organs, including the thoracic and abdominal aorta, were removed and fixed for 24 h in 4% paraformaldehyde (PFA, #P087.5, Carl Roth) at 4°C. After fixation, the aortic tissue was stepwise dehydrated in ethanol followed by an incubation in Roti®Histol (Carl Roth) for 12 h. Afterwards the aortic tissue was neutralized and embedded in paraffin (#Carl Roth).

Murine platelet isolation was conducted as described elsewhere (Donner et al., 2020). Briefly, blood was collected in 300 µL heparin solution (20 U/mL in PBS). Cell counts were determined using a hematology analyzer (Sysmex—KX21N, Norderstedt). The whole blood samples were centrifuged at 250x g for 5 min at RT. Platelet-rich plasma (PRP) was obtained by centrifugation at 50 × g for 6 min at RT. The PRP was washed twice in Tyrode’s buffer (136 mM NaCl, 0.4 mM Na2HPO4, 2.7 mM KCl, 12 mM NaHCO3, 0.1% glucose, 0.35% bovine serum albumin (BSA), pH 7.4), containing apyrase (0.02 U/mL) and prostacyclin (0.5 μM) by centrifugation at 650 × g for 5 min at RT. For experimental use, the isolated platelets were resuspended in Tyrode’s buffer supplemented with 1 mM CaCl₂ (in absence of apyrase and prostacyclin). For plasma preparation whole blood was collected in heparin coated Microvettes^® (Microvette^® 500 Lithium heparin gel, #20.1346.100, Sarstedt) and centrifuged at 10,000 × g for 5 min at 4°C. Plasma samples were stored at −70°C until use.

For the preparation of platelet releasates, murine platelets were isolated (40 × 10⁶) as described above and stimulated with CRP (0.1 or 1 μg/mL) or thrombin (0.1 U/mL) for 10 min at 37°C. After stimulation sample were centrifuged at 650 × g for 5 min at 4°C. The platelet releasates were collected and stored at −70°C before use.

The MHEC5-T (Mouse heart endothelial cell clone 5) cells were obtained from H. Langer, University Clinic Tübingen, Germany and cultured at 37°C in a humidified atmosphere with 5% CO₂ in DMEM high glucose cell culture medium (#41965062, Gibco), containing 10% fetal calf serum (FCS; Life technologies), 1% penicillin-streptomycin (10,000 U/ml penicillin; 10,000 μg/ml streptomycin; Life technologies) and 200 mM l-glutamine (Life technologies). All following steps were conducted under sterile conditions. MHEC5-T cells were seeded in a density of 2.5 × 10⁴ cells/well into a 48-well plate and incubated with releasates of activated, respectively resting platelets for 3 h at 37°C (5% CO₂). As controls, MHEC5-T were incubated with the indicated platelet agonists (negative control). Stimulation with 100 ng/mL TNFα served as positive control. After incubation, the cell culture supernatants were collected and cells were lysed using RLT Buffer from the RNeasy Mini Kit (#74106; Qiagen).

RNA isolation of platelet releasate treated MHEC5-T cells was performed using the RNeasy Mini Kit (#74106; Qiagen). The isolation was conducted according to the manufacturer’s instructions. Afterwards RNA purity and concentration was analyzed using a spectro-photometer (BioPhotometer D30, Eppendorf).

First, a DNA digestion was performed using DNAse I (#4716728001, Roche) to remove genomic DNA. Therefore, all samples were supplemented with DNAse I and incubated at 37°C for 30 min. Reaction was stopped by incubating all samples at 75°C for 10 min. Afterwards, cDNA synthesis was conducted by using the ImProm-IITM Reverse Transcription System (#A3800, Promega), following the manufacturer’s instructions. For cDNA synthesis, a total amount of 100 ng RNA was used.

For analysing the endothelial gene expression of P-selectin and E-selectin after platelet releasate treatment, a quantitative real-time PCR was performed using the Fast Sybr Green Master Mix (#4385612, Thermo Fisher) by following the manufacturer’s protocol. The following primer were used: P-selectin (for: CAT​CTG​GTT​CAG​TGC​TTT​GAT​CT; rev: ACC​CGT​GAG​TTA​TTC​CAT​GAG​T) and E-selectin (for. ATG​CCT​CGC​GCT​TTC​TCT​C; rev: GTA​GTC​CCG​CTG​ACA​GTA​TGC). GAPDH (for: GGT​GAA​GGT​CGG​TGT​GAA​CG; rev: CTC​GCT​CCT​GGA​AGA​TGG​TG) served as housekeeping gene. The evaluation was performed according to the ΔΔCt method. Data were normalized to WT resting.

Paraffin embedded aortic tissue of the aneurysm segment 14 days after PPE was sliced in 5 µm sections using an automatic microtome (Microm HM355, Thermo Fisher Scientific). First, tissue sections were deparaffinised and hydrated. For antigen unmasking the tissue sections were heated up in citrate buffer (pH 6.0) for 10 min at 300 W. Afterwards, the tissue sections were blocked for 1 h at RT with protein blocking solution (#X0909, Dako). After washing with PBS the aortic tissue sections were specifically stained with primary antibodies for platelet GPIbα (CD42b #M042-0, Emfret Analytics, 1:50), respectively for neutrophils (Ly6G #551459; BD Pharming, 1:100). Incubation with primary antibodies was conducted overnight at 4°C. Specific IgG primary antibodies served as controls. After washing with PBS the aortic tissue sections were incubated with a biotinylated secondary antibody (#BA-9400, Vector, 1:200) for 1h at RT. Hereafter, the tissue sections were incubated with a streptavidine FluorTM 660 conjugate (#50-4317-80, Thermo Fisher Scientific, 1:20) for 30 min at RT. Additionally, all sections were stained with DAPI (4′,6-Diamidine-2′-phenylindole dihydrochloride, #10236276001, Roche, 1:3,000) to visualize cell nuclei within the aortic tissue. The stained tissue sections were embedded with mounting medium (#S3023, Dako) and stored at 4°C until imaging. All images were generated using an Axio Observer. D1 microscope (Zeiss).

Flow cytometry was performed as described elsewhere (Klatt et al., 2018). Briefly, heparinized murine whole blood was washed three times with Tyrode’s buffer by centrifugation at 650 g for 5 min. After washing the whole blood samples were diluted in Tyrode’s buffer containing 1 mM CaCl₂. For the analysis of platelet activation, samples were stimulated with indicated agonists and specifically labeled with antibodies against P-selectin (CD62P, Wug. E9-FITC, #D200, Emfret Analytics) and active integrin αIIbβ3 (JON/A-PE, #D200; Emfret Analytics) in a ratio of 1:10 for 15 min at 37°C. Reaction was stopped by adding 300 µL of PBS to all samples. To analyze glycoprotein (GP) expression, washed whole blood was labeled for GPIbα (CD42b, Xia. G5-PE, #M040-2 Emfret Analytics), GPVI (JAQ1-FITC, #M011-1, Emfret Analytics), integrin α5 (CD49e, Tap. A12-FITC, # M080-1, Emfret Analytics), or integrin β3 (CD61, Luc. H11-FITC, #M031-1, Emfret Analytics) in a ratio of 1:10 for 15 min at RT. For the analysis of platelet-neutrophil aggregate formation, washed whole blood was stimulated with the indicated agonists and labeled for the platelet marker GPIbα (CD42b, Xia. G5-PE, #M040-2 Emfret Analytics) and for the neutrophil marker Ly6G (1A8-APC, #127614, BD Biosciences) for 15 min at 37°C. For phosphatidylserine (PS)-exposure measurements, washed murine whole blood was diluted with binding buffer (containing 10 µM HEPES, 140 µM NaCl, 2.5 mM CaCl2, pH 7.4). Samples were stimulated with the indicated agonists and labelled with a PS detecting antibody (CyTM 5 Annexin V, #559934, BD Biosciences) for 15 min at RT. GPIbα (CD42b) was used as platelet specific marker. All samples were analyzed using a FACSCalibur flow cytometer (BD Biosciences).

All ELISA analysis were performed according to the manufacturer’s instructions. For quantification of Panx1 plasma concentration in AAA patients a human Panx1 ELISA (#39097, Signalway Antibody) was performed. For plasma analysis in mice, an IL-1β (#DY401, R&D Systems), a MMP9 (#MMpT90, R&D Systems) and a MMP2 (#MMP200, R&D Systems) ELISA were used.

Data are presented as arithmetic means ± SEM (Standard error of mean), statistical analysis was performed using GraphPad Prism 8 (version 8.4.3). Statistical differences were determined using a two-way ANOVA with a Sidak’s multiple comparison post hoc test, an unpaired multiple t-test or an unpaired student’s t-test. Significant differences are indicated by asterisks (*** p < 0.001; ** p < 0.01; * p < 0.05).

First, we analyzed the concentration of soluble Panx1 in the plasma of AAA patients. As shown in Figure 1A, increased amounts of soluble Panx1 were detected in patient’s plasma (Figure 1A). No alterations have been detected in the releasates of stimulated platelets using ADP and collagen-related peptide (CRP) to activate platelets and to induce the deletion of Panx1 from the platelet surface (Figure 1B). However, a significant increase in soluble Panx1 was only detected after platelet stimulation with 1 μg/mL CRP using platelets from AAA patients but not from healthy controls (Figure 1B). Next, we investigated the phosphorylation of Panx1 at Tyr¹⁹⁸ as a marker for Panx1 activation. The activation of platelets from AAA patients resulted in elevated phosphorylation of Panx1 at Tyr¹⁹⁸ only after stimulation with 1 μg/mL CRP but not with ADP or low dose CRP (0.1 μm/mL) (Figures 1C, D). However, reduced tyrosine phosphorylation was detected in platelets from AAA patients compared to healthy controls. While, total levels of Panx1 were comparable between both groups (Figure 1E).

The results from AAA patients prompted us to analyze the impact of platelet Panx1 in further detail. We therefore induced experimental murine AAA utilizing the PPE model. We found that genetic loss of platelet-specific Panx1 did not alter diameter progression of AAA over a 28-day time course as shown by ultrasound tracking (Figures 2A–C). However, the incidence to develop AAA in mice was reduced in platelet-specific Panx1 knock-out mice (Figure 2D). Further, our data demonstrated reduced survival of Panx1 ^fl/fl -Pf4-Cre ⁺ mice after PPE surgery compared to controls (Figure 2E) potentially due to prolonged hemostasis in these mice (Molica et al., 2019). In line with unaltered aortic diameter expansion, we detected comparable aortic wall thickness between platelet-specific Panx1 knock-out and control mice as well as unaltered body weight (Figure 2F, Supplementary Figure S1). Furthermore, the analysis of MMPs in the plasma of mice using specific ELISAs revealed that MMP-9 levels of platelet-specific Panx1 knock-out mice were only reduced by trend (Supplementary Figure S1D) while MMP2 levels were only significantly increased in control but not in Panx1 knock-out mice at 14 days after PPE infusion of the mouse aorta (Supplementary Figure S1E). All these results suggest unaltered extracellular matrix (ECM) remodeling in these mice.

In a next step, we analyzed platelet activation after elastase infusion into the aortic wall using platelets from PPE mice at different time points after surgery (Figure 3). We did not detect major alterations in integrin αIIbβ3 (fibrinogen receptor) activation or P-selectin exposure as marker for degranulation as determined by flow cytometry (Supplementary Figure S2). In contrast, decreased Annexin-V binding as marker for pro-coagulant activity was detected at day 7 post PPE surgery following high doses of CRP that stimulates the major collagen receptor GPVI (Supplementary Figure S2D). At 14 days after PPE surgery, we also detected decreased P-selectin exposure and active integrin αIIbβ3 in response to high doses of CRP and PAR4 peptide that stimulates the thrombin receptor PAR4 (Figure 3A). Defects in P-selectin exposure were even more pronounced after 28 days post-surgery, while defects in integrin activation have only been detected after stimulation of platelets with low dose of CRP (Figure 3B). Pro-coagulant activity was reduced with high dose of CRP but unaltered after thrombin stimulation of platelets (Figure 3B). However, GP exposure at the platelet surface was unaltered over the whole observation period of 28 days (Figures 3A, B, Supplementary Figure S2C). In addition, we measured blood cell counts in naïve and PPE mice at different time points after surgery. Platelet and RBC counts were reduced on day 14 post-surgery while no alterations have been detected in mean platelet volume (MPV), WBC count and neutrophil-platelet conjugates after stimulation of platelets with either CRP or thrombin (Supplementary Figure S3).

To analyze the role of platelet Panx1 in inflammation, we analyzed Panx1 deficient platelets and their potential to form conjugates with neutrophils (Figures 4A, B). First, we analyzed the formation of platelet-neutrophil conjugates in the presence of LPS or TNF-α in vitro (Figure 4A). We detected significantly reduced platelet-neutrophil conjugates after stimulation using flow cytometry (Figure 4A). In a next step, we analyzed the formation of platelet-neutrophil conjugates in PPE mice at different time points. As shown in Figure 4B, reduced binding of neutrophils to platelets was detected after 14 days post-surgery. Reduced formation of platelet-neutrophil aggregates in platelet-specific Panx1 deficient mice resulted in reduced migration of neutrophils into the aortic wall of these mice 14 days after surgery (Figure 4C, Supplementary Figure S5). Reduced migration of neutrophils was accompanied by reduced migration of platelets into aortic tissue 14 days after surgery that might be due to reduced platelet activation at this time point (Figure 3A, Figure 4D). However, the plasma level of the acute phase cytokine IL-1β was only reduced by trend in platelet-specific Panx1 knock-out mice (Supplementary Figure S4A).

Platelet-mediated recruitment of inflammatory cells to inflamed tissue can be modulated by activated endothelial cells (Rainger et al., 2015). To analyze, if platelet Panx1 modulates the activation of endothelial cells, we analyzed platelet-induced activation of MHEC5-T cells (endothelial cell line) in vitro. To this end, platelets were activated with CRP or thrombin and the supernatant was collected. Afterwards, we incubated MHEC5-T cells with platelet releasates of naïve Panx1 deficient and control mice for 3 h and analyzed gene expression of P-selectin and E-selectin in MHEC-5 T cells (Figures 5A, B). As shown in Figure 5A, the expression of P-selectin in MHEC5-T cells was only altered with resting but not with activated Panx1 deficient platelets (Figure 5A). In contrast, a strong reduction in E-selectin gene expression was detected when MHEC5-T cells were incubated with the supernatant of CRP-activated Panx1 deficient platelets compared to controls (Figure 5B). These results suggest that Panx1 in platelets is important for the activation of endothelial cells as reflected by gene expression of adhesion molecules such as P- and E-selectin (Figures 5A, B). To analyze plasma levels of soluble P-selectin and E-selectin in PPE mice, we used platelet-specific Panx1 deficient mice and the respective controls. As shown in Figures 5C, D, no alterations were detected 14 days after PPE infusion of mouse aorta between groups (Figures 5C, D).