Human recombinant cytokine CXCL4 was purchased from Peprotech (Neuilly sur Seine, France). Human recombinant GM-CSF and M-CSF were obtained from Sanofi-Aventis (Montrouge, France) and Miltenyi Biotec SAS (Paris, France), respectively. Bleomycin sulfate (BLM), camptothecin, and cytochalasin D (CytoD) were purchased from Sigma-Aldrich (St-Quentin Fallavier, France).

Patients with SSc satisfied the 2013 ACR/EULAR classification criteria for SSc (16), and their clinical presentation was defined according to LeRoy et al. (17). They were consecutively included after written informed consent. This study was approved by the local ethics committees (Committees for Protection of Persons (CPP) Ouest-V France, CPP approval N°: 2019-A02611-56). Information regarding patients is resumed in Supplementary Tables 1 and 2 . Blood samples (PBMCs and sera) were collected from SSc patients at the Department of Internal Medicine and Clinical Immunology, Rennes Hospital, France, and those from healthy donors (HD) from Etablissement Français du Sang, Rennes, France.

Human macrophages were differentiated from peripheral blood monocytes. Buffy coats of healthy donors were provided by Etablissement Français du Sang (Rennes, France) and obtained after the written consent for the use of blood samples for experimental research. Peripheral blood mononuclear cells were obtained from blood buffy coats of HD or the blood of SSc patients through Ficoll gradient centrifugation. The monocytes, selected after a 1h adhesion step, were differentiated into macrophages (M0) during 6 days in RPMI 1640 medium GlutaMAX (Gibco, Life Technologies), containing 10% heat-inactivated fetal bovine serum (FBS, Lonza, Levallois-Perret, France), 2 mM L-glutamine, 20 IU/mL penicillin, 20 μg/mL streptomycin (ThermoFisher Scientific, Courtaboeuf, France) and 50 ng/ml of M-CSF or 400 IU/ml GM-CSF. The medium was replaced on day 4 and day 6. From day -6 to day 8, cells were maintained in a fresh medium containing 5% FBS either without GM-CSF (GM-M0) either 10 ng/ml M-CSF (M-M0) or supplemented with 1 µM CXCL4 for M4 polarization.

Male C57BL/6J mice weighing between 18 and 20 gr, used at 8 weeks of age, were acquired from Janvier Labs (Le Genest Saint Isle, France). The animals were all housed in similar autoclaved cages and fed food and water with identical housing conditions. They were maintained under a 12/12h light/dark cycle, with controlled room temperature and humidity. Experimental scleroderma-associated ILD was induced by daily intradermal injections of BLM solution (0.4 mg/kg in 100 µl) into the shaved back of mice (5 days a week) for 4 weeks. Serum and tissue biopsies were collected at the end of the experiment at day 28. Mice were randomly divided into 2 groups: intradermal injections of NaCl (n = 6), and intradermal injections of BLM dissolved in NaCl (n = 6). The number of groups and mice per group was pre-calculated depending on statistical power considerations based on the expected results on skin involvement and data from the literature. Animal studies were reviewed and approved by the Committee on the Ethics of Animal Experiments under the French Ministry of Higher Education and Research (#17011-2018100812449655). The study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals, EEC Council Directive 2010/63/EU.

Female C57BL/6J mice weighing between 18 and 20 gr, used at 8 weeks of age and purchased from Janvier Labs (Le Genest Saint Isle, France) were randomly divided into 2 groups: daily intradermal injections of 100 µl of PBS (n = 9) or HOCl (n = 8, one loss due to biopsy sample at week 3) as previously described (18). Serum and tissue biopsies were collected at the end of the experiment at day 42. Animal studies were reviewed and approved by the Committee on the Ethics of Animal Experiments under the French Ministry of Higher Education and Research (#17011–2018100812449655).

MDMs were treated for 1 h with 5 μM of cytochalasin D (CytoD), an actin polymerization inhibitor that prevents cytoskeletal remodeling, and serves as a negative control for phagocytosis. The cells were incubated in the presence of fluorescent latex microspheres (ratio 10:1) (Fluoresbrite™ Plain YG 1.0 Micron Microsphere, Polysciences, Warrington, USA) for 45 min at 37°C or at 4°C (negative control). After several washing with PBS to eliminate the non-phagocytosed beads, MDMs were detached in the presence of Accutase^® cell detachment solution. The fluorescence emitted at 525 nm by the cells having achieved phagocytosis was quantified by flow cytometry on an LSR II cytometer with FACSDiva software (BD Biosciences). The results were expressed in % of phagocytosis, calculated as follows: % cells fluorescent (37°C) - % cells fluorescent (4°C).

After cell washing and detachment using Accutase^® cell detachment solution (BioLegend, Paris, France), MDMs were stained with Fixable Viability Stain 780 (BD Biosciences, Le Pont de Claix, France) for 10 min at room temperature. MDMs were first blocked in PBS supplemented with 2% FBS solution containing FcR blocking reagent (Miltenyi Biotec SAS) for 10 min at room temperature to avoid nonspecific binding, and then re-suspended and incubated with specific antibodies or appropriate isotype controls for 30 min at 4°C. Cells were washed with PBS, collected by centrifugation (2500 rpm for 5 min) and then analyzed on a LSR II cytometer and FlowLogicTM software (Miltenyi Biotec SAS). The quantification of efferocytosis receptors was performed using the following antibodies: FITC anti-CD36, FITC anti-ITGβ5, PE anti-MERTK (BioLegend), APC anti-CD36L1/SR-B1 (Miltenyi Biotec SAS), and their respective isotype control, as recommended by BD Biosciences (Le Pont de Claix, France). Results were expressed as the mean ratio of median fluorescence intensity (MFI) calculated as follows: MFI (mAb of interest)/MFI (isotype control mAb).

Total RNA was extracted from human cells and mouse tissue with a Nucleospin RNA extraction Kit (Macherey-Nagel, Hoerdt, France) according to the manufacturer’s instructions. RNA concentrations were measured by spectrofluorimetry using a NanoDrop 1000 (Thermo Fisher Scientific) and reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). Quantitative PCR (qPCR) assays were next performed using the fluorescent dye SYBR Green methodology and a CFX384 Real-Time PCR detector (Bio-Rad Laboratories, Marnes-la-Coquette, France). Human and mouse predesigned KiCqStart^® SYBR^® Green primers for gene expression analysis were purchased from Sigma-Aldrich. The specificity of amplified genes was evaluated using the comparative cycle threshold method (CFX Manager Software). The Mean of Cq values was used to normalize the expression of the steady-state target mRNA levels to those of the 18S ribosomal protein, using the 2(−ΔΔCq) method.

Levels of IL-6, TNFα, CCL18, and CCL22 secreted in MDM culture media were quantified using specific Duoset ELISA kits from R&D Systems, Bio-Techne, France: IL-6: DY206; for TNFα: DY210; for CCL18: DY394; and for CCL22: DY336). Levels of S100A8 in mouse lung extract were also quantified by ELISA (R&D Systems, DY8596-05). Levels of mouse sera or human plasmatic CXCL4 were also quantified by ELISA (R&D Systems, human: DY795, mouse: DY595).

Data were presented as means ± standard error on the mean (SEM). Comparison between more than 2 groups was performed by repeated measure analysis of variance (ANOVA) followed by Dunnett’s or Newman–Keuls multiple comparison post hoc test. Depending on the conditions, Student’s paired-or unpaired t-tests were used to compare two groups. A p-value < 0.05 was considered significant. Data analyses were performed with GraphPad Prism 5.0 software (GraphPad Software, La Jolla, CA, USA).

As we previously described, MDMs from patients with SSc have significantly decreased capacities to eliminate apoptotic cells (n=19) when compared to MDMs from healthy donors (HD) (n=27) ( Figure 1A ) (4). We also confirmed that CXCL4 serum levels were significantly higher in SSc patients (n=150) than those of HD (n=42) ( Figure 1B ) (12). The clinical characteristics of patients with SSc in Figures 1A and B are reported in Supplementary Tables 1 and 2 , respectively. To evaluate the impact of soluble factors from the serum of SSc patients on the efferocytosis capacities of macrophages, we exposed human MDMs isolated from HD to plasma from HD or SSc patients and then exposed MDMs to CFSE⁺ apoptotic Jurkat cells to compare their phagocytic index. MDMs exposed to HD and SSc plasma showed a significant decrease in efferocytosis when compared to MDMs exposed to a control medium and MDM exposed to SSc plasma patients exhibited a significant decrease in efferocytosis properties when compared to MDM exposed to HD plasma ( Figure 1C ). Altogether, these data strongly suggested that serum elements like soluble CXCL4 may contribute to defect of efferocytosis.

As the expression of the chemokine CXCL4 was previously correlated with the fibrotic complications of the disease (12), we explored its expression in two mouse models of SSc-ILD induced by intradermal injections of repeated BLM or HOCl in which lung fibrosis has been validated (18, 19). In the HOCl model, serum levels of CXCL4 were significantly elevated when compared to the saline (NaCl) group, while in the BLM model, CXCL4 levels showed a moderate increase that did not reach statistical significance ( Figure 2A ). The mRNA expressions of CXCL4 in lung tissues were significantly increased in both BLM- and HOCl-treated mice. Such inductions were accompanied by a significant overexpression of some M4 macrophagic markers such as S100A8 and MMP7, with a greater induction in the lung of BLM-mice as MMP7 mRNA overexpression was not significant in HOCl-treated mice ( Figure 2B ). Protein expression of S100A8 in the lung was also significantly increased in BLM-exposed mice ( Figure 2C ). To evaluate whether the presence of an M4 macrophage phenotype was associated with an efferocytosis defect, we administered CFSE⁺ apoptotic Jurkat cells to NaCl- or BLM-treated mice. Our results demonstrated that the percentage of AMs (Gr1^Int and CD11b^Int) that ingested CFSE⁺ apoptotic Jurkat cells and the fluorescence intensity of CFSE per macrophage were both significantly decreased in the BAL of BLM-induced SSc mice when compared to the control group ( Figure 2D ). Altogether, these results strongly suggested that SSc-ILD mice presented at the systemic and the lung tissue levels a M4 phenotype and demonstrated that AMs have a defect of efferocytosis.

To better understand the effects of CXCL4 on MDM efferocytosis capacities, we characterized human M4-MDMs functions in vitro and explored their phenotype as compared to previous descriptions (15) ( Figure 3A ). Microscopic observation of MDMs in culture revealed that M4-MDMs had less spindles than M0-MDMs and they were rounded, showing similarities with GM-CSF-differentiated MDMs (GM-MDMs) ( Figure 3B ). Among the M4 markers previously identified (20), we found a significant increase of MMP7 mRNA expression in comparison to M0-MDMs, whereas CD163 mRNA expression was reduced in M4-MDMs, without reaching statistical significance. The mRNA expression of S100A8 and the M2a/M4 markers, CCL18 and CCL22 were found unchanged when compared to M0-MDMs ( Figure 3C ). Flow cytometry analysis of membrane markers showed that M4-MDMs tend to have reduced expression of CD163, that was previously found downregulated in M4 macrophages (21), and of CD86, an M1-marker when compared to M0-MDMs whereas the expression of CD206, a M2a marker, was found unchanged ( Supplementary Figure 1 ). M4-MDMs also showed a pro-inflammatory phenotype with a high secretion of cytokines such as IL-6 and TNF-α as compared to M0-MDMs. Interestingly, M4-MDMs also secreted higher levels of the pro-fibrotic M2 markers CCL18 and CCL22 as compared to M0-MDMs ( Figure 3D ). Altogether, our results suggest a specific phenotype of CXCL4-induced M4-MDMs, with mixed pro-inflammatory and pro-fibrotic features.

We secondly evaluated the phagocytic and efferocytic functions of M4-MDMs. The capacities of M4-MDMs to phagocyte fluorescent pHrodo+ apoptotic cells, determined by real-time fluorescence microscopy, were significantly reduced. GM-MDMs and M0-MDMs pretreated with cytochalasin D (cytoD), a drug inhibiting phagocytosis via the impairment of actin depolymerization and cytoskeleton remodeling (22), also showed decreased efferocytosis capacities ( Figure 3E ). By flow cytometry, we showed that M4-MDMs were also less efficient in engulfing FITC-beads than M0-MDMs and that their phagocytic abilities were similar to those of GM-MDMs and MDMs exposed to cytoD ( Figure 3F ). To better understand such an alteration of efferocytosis in CXCL4-induced MDMs, we evaluated by flow cytometry the membrane expression of several receptors involved in the recognition of apoptotic cells. As previously described (23), the expression of CD36 was significantly reduced in M4-MDMs in comparison to M0-MDMs. The expression of the CD36L1 and MERTK also tended to decrease in M4-MDMs but without reaching statistical significance whereas ITGβ5 expression did not differ between M4-MDMs and M0-MDMs ( Figure 3G ). Altogether, these results suggested that the moderate decrease in CD36 expression could not fully explain the reduced efferocytosis capacities of M4-MDMs and that other mechanisms could be involved.

To explain the lower abilities of M4-MDMs to accumulate apoptotic cells in phagolysosomes ( Figure 3E ), we explored the LC3-associated phagocytosis (LAP), an internalization process utilizing some proteins of the autophagy machinery. LC3 recruitment and conversion are involved in the canonical and non-classical autophagy pathways. It is well known that the recruitment of LC3 relies on the formation and elongation of the phagophore which is slow during autophagy unlike the LAP process characterized by a quick LC3 recruitment that is performed in less than 15 min after apoptotic cell exposure and that reaches a maximum after one hour (24, 25). M4-MDMs were thus exposed to PKH67⁺ apoptotic Jurkat cells for 45 min before cellular immunostaining of LC3, and LAPosome quantification by fluorescence microscopy. LAPosomes are characterized by circular recruitment of the LC3 protein surrounding the apoptotic Jurkat cells ( Figure 4A ), or IgG-beads ( Figure 4B ), as opposed to phagosomes without peri-phagosomal reinforcement of LC3 staining. While we confirmed the decreased phagocytic capacities of CXCL4-induced MDMs to engulf apoptotic Jurkat cells ( Figure 4C ) and IgG-beads ( Figure 4D ) in comparison to resting M0-MDMs, LAPosome proportion among phagosomes was found to be higher in M4-MDMs as compared to M0-MDMs after a 45-minute efferocytosis of apoptotic Jurkat cells ( Figure 4E ) or of IgG-beads ( Figure 4F ). LAPosomes only represented a moderate proportion of the total number of phagosomes, LAP being a dynamic and fleeting process. MDMs also showed a diffuse cytoplasmic and nuclear fluorescence of LC3, the protein being mainly in its soluble form LC3-I with a nuclear reservoir (26).

To explain the higher LAPosome proportion in M4-MDMs, we next evaluated the expression of autophagy-associated proteins involved in LAP both in vivo and in vitro. We analyzed Rubicon and NOX2 mRNA expression, both known to promote LAP. These proteins are part of a complex that also includes BECLIN1. Additionally, we analyzed the mRNA expression of the ATG (autophagy-related) proteins ATG4 participating in the conversion of LC3-I to its conjugated form LC3-II, and ATG5 involved in LC3-II integration into the phagosome to form LAPosomes. mRNA expression of LC3A, ATG4, ATG5, BECLIN1, NOX2, Rubicon, and p62 were all significantly increased in the lung of BLM-SSc mice in comparison to controls ( Figure 5A ). By contrast, only BECLIN1 and ATG4 mRNA expression were upregulated in SSc-MDMs when compared to HD-MDMs ( Figure 5B ), and only BECLIN1 in M4-MDMs compared to M0-MDMs ( Figure 5C ). The clinical characteristics of patients with SSc of Figure 5B are reported in Supplementary Table 1 . Altogether, our data suggest that, while phagocytosis was less efficient, LAP seemed increased in vivo in the fibrotic lungs of BLM-mice as well as in vitro in SSc-MDMs, with enhanced LAP in M4-MDMs in vitro.