Hearts were isolated from pigs in a slaughter house (Bigard, Saint-Paul-sur-Ternoise, France) and dissected within up to 4 h to harvest aortic valvular cusps. Then, aortic cusp pieces of 25 mm² were incubated for 12 h at 4°C in a 30 mM HEPES solution [acid 4-(2-hydroxyethyl)-1piperazine ethane sulfonic, ref. c-40020, PromoCell], with or without 0.6% GA (ref. G5882, Sigma Aldrich).

Twenty four 4-week Sprague-Dawley male rats were used as hosts, after 1 week acclimation in the animal platform PLATANN of the University of Picardie Jules Verne, Amiens, France (agreement B80021009). Surgery was performed under general anesthetic (isoflurane gas). Rats were in prone position. Two incisions were made on the back of the host, then soft dissection with scissors permitted to obtain pouch for each incision. Aortic valvular cusps were rinsed in sterile 0.9% sodium chloride solution and then put inside of the pouch. The origin of the 2 pieces of valve cusps implanted in the same rats came from different pigs. The incision was then closed by separate points of prolen 5.0 and a local analgesic was injected all around the incision. This experimental project was authorized under the reference APAFIS#1945-2019022212357854 v4, in accordance with the European Directive 2010/63/EU.

The design of the experimental protocol is illustrated in Figure 1. Four groups of 6 rats (12 independent porcine aortic valve cusp pieces each time) were compared: rats transplanted with GA-free cusps or GA-fixed cusps and those transplanted with GA-fixed cusps and treated with 1 mg/kg/day SCH5217123 (a CXCR2 antagonist, ref. 331-10731-5, RayBiotech) injected locally subcutaneously (SC) or intraperitoneally (IP). This dose was shown to be effective in various animal models of inflammation (13). Fourteen days later, blood sampling was made by cardiac punction under the xyphoid and the cusps were explanted and rinsed with PBS. Using micro-surgical instruments, the porcine cusps were unfolded and spread, then cut in 3 parts to be analyzed by alizarin red staining, calcium assay, immunohistochemistry and Western-blotting.

The blood count was taken within one hour of the intracardiac puncture on an EDTA tube. The analysis was carried out on a XN-10 diagnostic system (Sysmex) and included red blood cells and white blood cells (lymphocytes, monocytes, neutrophils) counts.

Pieces of porcine aortic valve cusps intended for the histological analyzes were fixed overnight with paraformaldehyde (PFA 4%) and two other nights with a 20% sucrose solution the first night and 30% the second in order to dehydrate the tissues. Then, they were included in OCT (Optimal cutting temperature compound, ref. KMA-0100-00A, Cell Path), snap-frozen in isopentane and stored at −80°C. A few days later, tissues were cut in a cryostat (LEICA CM 1950) in slices 12 μm thick. Several slides of the same valve were prepared and stored at −80°C so that they could be used later for alizarin staining as well as for immunostaining. On the day of staining, the slides were taken out of the freezer and then dried for 30 min at room temperature. Alizarin red solution (40 nM and pH 4.2, ref. A5533-25G, Sigma Aldrich) was gently applied to the slices. The excess dye was then removed and dehydration was carried out by incubation in a 100% acetone solution for 1 min and then in an acetone-toluene solution (1:1) for 1 min as well. Finally, 2 min incubation in a toluene solution was carried out before mounting with the DPX solution (ref. 0652, Sigma Aldrich). Photos of the valves were taken using the optical microscope (LEICA DM 2500).

Pieces of porcine aortic valve cusps deposited in a 48-well plate were incubated overnight in a 0.6 N HCl solution, and then the content of tissue calcium was measured by the o-cresolphtalein complexone method. Briefly, a mixture containing 191 µg/ml of o-cresolphthalein and 2.33 mg/ml of 8-hydroxyquinoline with 0.5 M 2-Amino-2-methyl-1-propanol was prepared. The mixture was heated for 2 h at 65°C after covering it with parafilm to prevent evaporation. A volume of 25 µl of sample was taken from each well of the 48-well plate and then deposited in a new 96-well plate. A volume of 150 μl of the reaction mixture was added to each well. The plate was then shaken for 15 min at room temperature. Finally, the absorbance was read at 565 nm using a microplate reader and the amount of calcium was determined according to a standard curve of absorbance vs. calcium concentration. At the end of the assay, the valve pieces were dried at 37°C overnight to be weighed the next day on a precision balance. The calcium content of the valve cusp pieces was related to the tissue mass and, for each rat, the result of the 2 pieces of implanted valves was averaged.

Slices of porcine aortic valve cusps were unmasked with citrate for 15 min at 100°C, exposed to 0.3% H2O2 in PBS-1% BSA for 10 min at RT (to block endogenous peroxidase activity), and permeabilized for 1 h at RT with 0.15% Triton X-100 in PBS containing 1% BSA. Samples were then incubated overnight at 4°C with the primary antibodies anti-CD-3 (ref. A0452, Dako, dilution 1:100), anti-CD-11b (ref. ab8879, Abcam, dilution 1:1500), anti-CD-68 (ref. MCA341R, AbD Serotec, dilution 1:300), anti-CXCR2 (R&D Systems, MAB331, 1:50 dilution) or isotypes controls, in PBS with 1% BSA-0.1% Triton X-100. Samples were then rinsed three times in PBS with 1% BSA and incubated for 1 h at RT with the correspondent peroxidase-coupled secondary antibody (Cell signalling, 1:750 dilution) and thoroughly washed in PBS. Samples were then incubated with the peroxidase substrate from the Vector® NovaRED™ kit (Vector Labs, SK-4800) at RT until suitable staining developed. Finally, the coverslips were counterstained with haematoxylin solution and mounted on glass microscope slides using VectaMount™ mounting medium (Vector® H-5000). Sections were examined using an optical microscope (LEICA DM 2500). Quantitative analysis of the staining intensity was performed using ImageJ software.

Pieces of porcine aortic valve cusps were washed with PBS and then homogenized in NP-40 lysis buffer (25 mM Tris-HCl, 250 mM NaCl, 5 mM EDTA, and 1% SDS, pH 7.4) prepared with phosphatase and protease inhibitor cocktails, using gentleMACS tubes and a gentleMACS™ Dissociator. The supernatants were removed and the proteins quantified using a DC™ Protein Assay Kit (Bio-Rad) and diluted to equal concentrations. Protein samples were combined with 4X SDS sample buffer (8% SDS, 40% glycerol, 0.4% bromophenol blue, 5% β-mercaptoethanol, 240 mM Tris, pH 6.8), boiled 5 min at 95°C, and separated on 10% SDS-PAGE gels. After electrophoresis, proteins were transferred to nitrocellulose membranes (Ref. 10,600,002, Amersham,) using a Bio-Rad Trans-Blot Turbo Transfer System for 1 h. Membranes were blocked with 5% milk prepared in PBS 1%-Triton X-100 and then incubated with the primary antibodies anti-CXCR2 (Ref. MAB331, R&D Systems, 1:500 dilution), anti-CD68 (ref. MCA341R, AbD Serotec, dilution 1:300) or GAPDH (Ref. 2118, Cell signaling, 1:1,000 dilution) overnight at 4°C. After repeated washing, the membranes were incubated with the appropriate secondary antibody (dilution 1:1,000) in TBST with 5% milk for 1 h at room temperature and washed again. Then, the blots were developed using chemiluminescence reagent (Bio-Rad). The blots were viewed using a Chemidoc Imaging System (Bio-Rad) and the bands quantified using Image Lab software. The expression levels were corrected for loading differences by normalization against GAPDH.

Data are reported as the mean values ± SEM. Student's t-test and the Mann–Whitney test were used for the statistical analysis of two groups of parametric and non-parametric data, respectively. For comparison of more than two groups, one-way ANOVA and Kruskal Wallis tests were performed for parametric data and non-parametric data, respectively. P-values were corrected for multiple comparisons by the Tukey's post-hoc method. They were two-sided and considered statistically significant if P < 0.05. Analyses were done using GraphPad Prism software (Version 7.0; GraphPad, San Diego, CA).

Results of Alizarin red staining and calcium assay are reported in the Figures 2A,B respectively. They show a strong 14-day calcification of implanted GA-fixed cusps in comparison to implanted GA-free cusps (52.3 ± 11.8 vs. 0.76 ± 0.4 µg Ca/mg tissues, P < 0.0001). To verify that the calcification was indeed produced during the two weeks following implantation, a comparison of the calcification was carried out between the group of valves not implanted on day 1 and those implanted (in the control rats) after explantation on day 14. This comparison confirmed the significant calcification of implanted GA-fixed cusps compared to those not implanted (52.3 ± 11.8 vs. 0.25 ± 0.1 µg Ca/mg tissues, P < 0.0001). Furthermore, the results of alizarin staining and calcium assay showed a significant reduction in calcification of implanted GA-fixed cusps in the group of rats treated with SCH527123 IP or SCH527123 SC compared to the untreated rats (28.6 ± 9.9 vs. 52.3 ± 11.8 µg Ca/mg tissues, P = 0.0013 or 30.2 ± 16.4 vs. 52.3 ± 11.8 µg Ca/mg tissues, P = 0.0030, respectively).

We analyzed the blood cell count in the different groups of rats at sacrifice (Figure 3). Results showed no significant variation in the red blood count and hemoglobin levels between conditions, but revealed a significant higher total white blood count and lymphocyte and monocyte counts in rats transplanted with GA-fixed cusps than in those transplanted with GA-free cusps (15.4 ± 0.8 vs. 4.7 × 10⁹/L, P < 0.0001; 11.9 ± 1.9 vs. 3.8 ± 0.6 × 10⁹/L, P = 0.0009 and 1.2 ± 0.3 vs. 0.5 ± 0.2 × 10⁹/L, P = 0.0034, respectively). These effects were significantly prevented in rats treated with SCH527123 administered IP (8.1 vs. 15.4 × 10⁹/L, P = 0.0027; 6.8 ± 0.7 vs. 11.9 ± 1.9 × 10⁹/L, P = 0.0422 and 0.5 ± 0.1 vs. 1.2 ± 0.3 × 10⁹/L, P = 0.0458) or SCH527123 administered SC for the white blood count (7.6 ± 1.5 vs. 15.4 × 10⁹/L, P = 0.0015) and the lymphocyte count (6.4 ± 1.2 vs. 11.9 ± 1.9 × 10⁹/L, P = 0.0239). Conversely, there was no significant change in the rat blood count of neutrophils between conditions.

Representative images and quantitative results of immunohistochemistry analysis are provided in Figures 4A,B and the Supplementary Figure S1, respectively. They showed no significant difference in the staining of CD11b (a marker of neutrophils and monocytes) in GA-fixed cusps compared to GA-free cusps, after 14 days of subcutaneous implantation in rats (% of relative expression, 14.4 ± 9.2 vs. 9.3 ± 1.6 respectively, P = 0.5969). Conversely, a significant increase in the staining of CD3 (a marker of T-lymphocytes) and CD68 (a marker of macrophages) was shown in GA-fixed cusps compared to GA-free cusps (31 ± 11.9 vs. 12.7 ± 3.7 for CD3, P = 0.0072 and 28.1 ± 16.5 vs. 9.6 ± 5.4 for CD68, P = 0.0210, respectively), and these effects were significantly prevented in GA-fixed cusps explanted from rats treated with SCH527123 IP as compared to those explanted from non-treated rats (11.9 ± 8.1 vs. 31 for CD3, P = 0.0049 and 9.7 ± 3.3 vs. 28.1 ± 16.5 for CD68, P = 0.0213, respectively). In addition, CXCR2 staining was also increased in GA-fixed cusps compared to GA-free cusps (21.9 ± 6.3 vs. 6.6 ± 1.7, P < 0.0001, respectively), effects that were significantly reduced in the GA-fixed cusps explanted from rats treated with SCH527123 IP (10.6 ± 2.1 vs. 21.9 ± 6.3, P = 0.0006, respectively) or SC (13.2 ± 4.5 vs. 21.9 ± 6.3, P = 0.0076, respectively). Western blots also showed significant lower CD68 and CXCR2 expressions in GA-fixed cusps explanted from rats treated with SCH527123 IP as compared to GA-fixed cusps explanted from non-treated rats (in fold change to the GA-free cusps explanted from non-treated rats, 0.8 ± 0.1 vs. 1.5 ± 0.1, P = 0.0038 for CD68 and 0.9 ± 0.2 vs. 2.2 ± 0.4, P = 0.0177 for CXCR2, respectively) (Figure 4C).