Elevated Fibronectin Levels in Profibrotic CD14+ Monocytes and CD14+ Macrophages in Systemic Sclerosis

Background Systemic sclerosis (SSc) is an autoimmune disease characterized by overproduction of extracellular matrix (ECM) and multiorgan fibrosis. Animal studies pointed to bone marrow-derived cells as a potential source of pathological ECM-producing cells in immunofibrotic disorders. So far, involvement of monocytes and macrophages in the fibrogenesis of SSc remains poorly understood. Methods and Results Immunohistochemistry analysis showed accumulation of CD14+ monocytes in the collagen-rich areas, as well as increased amount of alpha smooth muscle actin (αSMA)-positive fibroblasts, CD68+ and mannose-R+ macrophages in the heart and lungs of SSc patients. The full genome transcriptomics analyses of CD14+ blood monocytes revealed dysregulation in cytoskeleton rearrangement, ECM remodeling, including elevated FN1 (gene encoding fibronectin) expression and TGF-β signalling pathway in SSc patients. In addition, single cell RNA sequencing analysis of tissue-resident CD14+ pulmonary macrophages demonstrated activated profibrotic signature with the elevated FN1 expression in SSc patients with interstitial lung disease. Peripheral blood CD14+ monocytes obtained from either healthy subjects or SSc patients exposed to profibrotic treatment with profibrotic cytokines TGF-β, IL-4, IL-10, and IL-13 increased production of type I collagen, fibronectin, and αSMA. In addition, CD14+ monocytes co-cultured with dermal fibroblasts obtained from SSc patients or healthy individuals acquired a spindle shape and further enhanced production of profibrotic markers. Pharmacological blockade of the TGF-β signalling pathway with SD208 (TGF-β receptor type I inhibitor), SIS3 (Smad3 inhibitor) or (5Z)-7-oxozeaenol (TGF-β-activated kinase 1 inhibitor) ameliorated fibronectin levels and type I collagen secretion. Conclusions Our findings identified activated profibrotic signature with elevated production of profibrotic fibronectin in CD14+ monocytes and CD14+ pulmonary macrophages in SSc and highlighted the capability of CD14+ monocytes to acquire a profibrotic phenotype. Taking together, tissue-infiltrating CD14+ monocytes/macrophages can be considered as ECM producers in SSc pathogenesis.


INTRODUCTION
Systemic sclerosis (SSc) is an autoimmune disease, characterized by high morbidity and mortality and a significant reduction in quality of life. Microvascular damages, dysregulation of innate and adaptive immunity and multiorgan fibrosis are implicated in the pathophysiology of SSc (1). SSc is characterized by a patientto-patient variability in clinical manifestations, autoantibody profiles, extent of disease, treatment response and reduced survival rate (2). Changes in internal organ architecture, leading to pulmonary and cardiac complications and dysfunction remains the major causes of deaths among SSc patients (3).
Fibrogenesis is a multistage process, considered as the result of impaired tissue repair responses, in which abnormal production of cytokines, growth factors and angiogenic factors turn tissue homeostasis towards the excessive accumulation of extracellular matrix (ECM) (4). Fibrosis is usually an outcome of prolonged and exaggerated activation of fibroblasts, which differentiate into myofibroblasts (5). In contrast to physiological wound healing, in which myofibroblast are present only transiently, in fibrosis myofibroblasts become a permanent cellular component of the tissue (6).
Myofibroblasts are characterized by expression of alpha smooth muscle actin (aSMA) forming stress fibers and exhibit an increased capacity to synthesize collagens, fibronectin, and other ECM components. Fibronectin is a high-molecular weight glycoprotein of the extracellular matrix, which binds to integrins and other extracellular matrix proteins such as collagen, fibrin, and heparan sulfate proteoglycans (7). Increased deposition of fibronectin, paralleled to accumulated collagen, has been reported in the SSc skin (8,9).
The cellular source of myofibroblasts in wound healing and fibrotic lesions is still debatable. Although most evidence point to resident fibroblasts, other tissue-resident cell types or cells recruited from circulation have been shown to acquire a myofibroblast-like phenotype (10). In experimental models of lung and kidney fibrosis, bone marrow-derived fibrocytes have been shown to be recruited into injured or fibrotic tissues in response to chemokine signals, where they differentiate into fibroblasts or myofibroblasts (11). Lineage tracing experiments have provided convincing evidence that pericytes are an important source of myofibroblasts during renal fibrogenesis in animal models, and these cells might also be a source of myofibroblasts in SSc (12,13). During fibrogenesis, epithelial cells have been hypothesized to undergo epithelial-to-mesenchymal transition and transdifferentiate into myofibroblasts in response to TGFb and other profibrotic cytokines. Additionally, SSc tissues are chronically hypoxic, which can further promote the formation of myofibroblast-like cells (14). In addition to epithelial cells, endothelial cells have also been hypothesized to transdifferentiate into myofibroblasts through endothelialmesenchymal transition (15).
The primary mechanism underlying excessive fibrosis in SSc remains unknown; however, TGFb is one of the best-studied mediators regulating fibrotic processes, including fibroblast differentiation, ECM deposition and tissue contraction (16). Elevated levels of TGFb-regulated genes (cartilage oligomeric matrix protein, thrombospondin-1) were reported in the skin lesions from SSc patients (17). TGFb signalling regulates the expression of profibrotic genes mostly via the SMAD-dependent canonical pathway (18). However, the involvement of several non-canonical pathways in fibrotic processes was acknowledged. For example, activation of the ERK-MAPK signalling pathway leads to upregulation of type I collagen in SSc fibroblasts (19). Ultimately, inhibition of TGFb signalling has been addressed as a potential treatment strategy; however, conflicting results were reported following TGFb blockade (20).
Myeloid cells, including monocytes, have been shown to be essential regulators of fibrosis as producers of chemokines, inflammatory cytokines and growth factors (21,22). Our group has previously reported that the microenvironment of inflamed lung induces myeloid progenitors to acquire a myofibroblast phenotype (23). Similarly, we showed that heart-infiltrating myeloid cells served as myofibroblast progenitors in the model of post-inflammatory cardiac fibrosis (24). In the present study, we evaluated the differentiation potential of circulating CD14 + monocytes from healthy controls and SSc patients into a myofibroblast-like phenotype.

SSc Patients and Healthy Controls
Human blood samples and skin biopsies collection were approved by the local ethics committee of the Canton Zurich (KEK-ZH 515, PB-2016-02014, KEK-Nr 2018-01873). All study subjects provided written informed consent. SSc patients and healthy controls were recruited at the Department of Rheumatology of University Hospital Zurich. All the patients fulfilled the ACR/EULAR 2013 classification criteria for SSc.
The patients diagnosed with SSc by experts with Raynaud syndrome and a least one other SSc characteristic such as: SSc specific antibodies, SSc characteristic capillaroscopy changes and/or puffy fingers. The patients not fulfilling the ACR/ EULAR 2013 classification criteria for SSc were grouped as early-SSc. The patients fulfilling the criteria were further divided into lcSSc and dcSSc subgroups, according to Le Roy et al. (25). Detailed demographics and clinical characteristics of SSc patients are included in Table 1. Healthy control group was age-and gender matched (age mean=44.4 ± 10.8, female 17/ 20 (85%).
Data are presented as number (n)/total valid cases (N) (%). Disease duration was calculated as the difference between the date of the baseline visit and the date of the first non-Raynaud's symptom of the disease as reported by the patient. Pulmonary hypertension was judged on RHC. Active disease was defined as a score >3 by calculating European Scleroderma Study Group disease activity indices for systemic sclerosis proposed by Valentini. Immunosuppressive therapy was defined as treatment with corticosteroids (prednisone dose ≥10 mg/day or other dosage forms in equal dose) or any immunosuppressant.
Human endomyocardial biopsies were provided by the University Hospital Tubingen, Germany. Samples were obtained from SSc patients with cardiac involvement (inflammatory dilated cardiomyopathy) and controls (patients with healed myocarditis). Lung biopsies were provided by the Division of Rheumatology, Medical University of South Carolina, Charleston, USA. The samples were obtained from SSc-related interstitial lung disease (SSc-ILD). Sample tissue from downsized lung transplants from healthy individuals served as controls. The experiments with re-use of human material were approved by Swissethics (KEK-Nr 2019-00058, KEK-Nr 2018-01873) and were performed in conformity with the principles outlined in the Declaration of Helsinki. Transcriptomic analysis of an already published dataset of human lung tissues has been included. For this, the University of Pittsburgh Institutional Review Board approved ethics of use the human lung samples as previously described (26).

CD14 + Monocytes Isolation and Differentiation
Blood samples were collected in EDTA tubes (BD Vacutainer) and processed within 24h. Peripheral blood mononuclear cells (PBMCs) were isolated by gradient centrifugation on cell separation medium (Lympholyte, Cedarlane), followed by magnetic-activated cell sorting for CD14 using human microbeads and AutoMACS Pro device (Miltenyi Biotec). Cells were cultured in DMEM low glucose medium (Sigma) with the addition of 10% Foetal Bovine Serum (FBS) and 1% penicillin/ streptomycin (both Gibco). To differentiate into myofibroblastlike cells, CD14 + monocytes were stimulated with 10 ng/ml TGFb (PeproTech), 10 ng/ml IL-4, 10 ng/ml IL-10 and 10 ng/ ml IL-13 (Immunotools) for 7 days. Reads were aligned to the genome using the STAR algorithm. Gene expression profiles were next calculated by the FeatureCount algorithm. For differentially expressed genes, the DeSEQ2 algorithm was used with a threshold of minimum 10 reads for a transcript to be considered as present. Pathway enrichment analysis of differentially expressed genes (p ≤ 0.01, |log 2 ratio| ≥0.5) was performed by the Metacore software, as described previously (27). Human explanted lung tissues were digested and scRNAsequencing was performed as previously described (26).

ScRNAseq Analysis of Each Sample
Raw count matrices produced from CellRanger were loaded from the dataset GSE128169. Empty droplets were distinguished from droplet-containing cells by using the emptyDrops function from the R package DropletUtils (28,29). Only droplets that obtained an FDR <= 0.001 were called as cells. Doublets were also discarded from further analysis using the R package scDblFinder. Low quality cells were identified based on the number of reads, genes and mitochondrial content using the R package scuttle. We excluded cells that were outliers according to at least one of the following thresholds: number of reads < 3 MADs, number of gene < 3 MADs, mitochondrial content > 3 MADs. Additionally, we discarded genes with < 1 UMI count in < 0.01 of the remaining cells. We used the SCTransform (30) method from the R package Seurat (31) to normalize and scale the data.

Integration of Multiple Samples
Multiple samples were combined using the Seurat integration workflow (32). After integration we reduced the dimensionality of the data using Principal Component Analysis (PCA). The main 30 principal components were used to perfor unsupervised clustering with a resolution value of 0.8. After clustering and visualization with Uniform Manifold Approximation and Projection (UMAP), cell populations were identified through examination of gene markers in the associated transcriptome.
Cells that had > 0 counts for the CD14 gene were called as CD14 + cells. All plots and downstream analysis were performed using Seurat.

RT-qPCR
For reverse transcription, 200-300ng of total RNA was used. The reaction was performed using MultiScribe reverse transcriptase (Thermo Fisher Scientific), random hexamers and RNAase inhibitor (both Roche). Subsequently, the qPCR reaction was performed using the SYBR green GoTaq qPCR master mix (Promega) on Agilent Stratagene Mx3005P qPCR instrument. Sequences of primers are listed in Supplementary Table 1.
Relative gene expression was calculated using the 2 -DDCt method. GAPDH was used as the reference gene.

Protein Extraction and Immunoblotting
After the stimulation with cytokines, cells were washed once with ice-cold PBS, collected, centrifuged and lysed for 30 minutes on ice in RIPA buffer (Sigma) supplemented with proteases and phosphatases inhibitors (Roche). An equal amount of the protein was loaded and separated by SDS-PAGE electrophoresis, followed by wet transfer on the nitrocellulose membrane (GE Healthcare). The membrane was further incubated for 1 hour in blocking buffer (5% BSA in TBS-T). Further, membranes were probed overnight with primary antibodies, listed in Supplementary Table 2, in blocking buffer at 4°C. Further, membranes were incubated for 1 hour at room temperature with secondary HRP-conjugated antibodies. Signal was developed with ECL substrate (SuperSignal West Pico PLUS, Thermo Scientific) and acquired on the Fusion fx (Vilber) device.

ELISA
For the detection of human pro-collagen 1a1 DuoSet ELISA (RnD Systems) was used according to the manufacturer's protocol. Briefly, 96-well plates were coated with capture antibodies overnight in room temperature and further blocked with 2% BSA in phosphate-buffered saline (PBS) supplemented with 0.05%Tween 20. Between each step, plates were washed three times with PBS. The protein standards and samples were applied and incubated for 2 hours. Next, plates were incubated with biotin-conjugated detection antibodies for 2 hours and streptavidin-HRP for 30 minutes in room temperature. Signal was developed with TMB substrate (Thermo Scientific), and 450 nm absorbance was measured on a BioTEK HT plate reader (Tecan). Concentrations were calculated according to the respective standard curves. from Tocris Biosciences. To determine optimal non-toxic concentrations, we performed toxicity tests. CD14 + monocytes were incubated with 2-fold dilutions of inhibitors starting from 5 or 10 µM. Cells were incubated for 24h, stained with propidium iodide (Biolegend), and cytotoxicity was evaluated by flow cytometry. The highest non-toxic concentration was used in further experiments.

2D and 3D Co-Culture With Dermal Fibroblasts
To distinguish cells in both systems, monocytes were stained with Cell Trace Violet (Thermo Scientific), and fibroblasts were stained with CFSE (Biolegend) according to the manufacturer's protocol. For 2D co-cultures, fibroblasts were plated 24h prior to the addition of monocytes. Cells were cultured for 7 days and sorted using FACSAria III cell sorter.
For the 3D co-culture model, 3DProSeed ® hydrogel microtiter plate (Ectica Technologies) were used. Firstly, labelled fibroblasts were plated and allowed to penetrate hydrogels for 24h. Next, monocytes were added. Plates were incubated for 7 days, and anti-aSMA/phalloidin (both Sigma) staining was performed. Co-cultures were visualized using Leica SP8 confocal microscope.

Immunohistochemistry, Imaging and Quantification
Collected tissue samples were washed in PBS and fixed for 16 hours in 4% paraformaldehyde in PBS. Next, tissues were rinsed in distilled water and transferred to 50% ethanol. Biopsies were then dehydrated (three incubations in 80% ethanol for 1 hour, three incubations in 96% ethanol for 1 hour, two incubations in 100% ethanol for 1 hour). Tissues were cleared twice in xylene for 1 hour and subsequently incubated twice in a 56°C paraffin bath for 3 hours. After paraffin embedding, 4 mm thick sections were placed on Superfrost Plus slides (Thermo Scientific) and dried overnight. Lung sections were cut at a thickness of 4 mm and stained with hematoxylin and eosin (HE) for analysis of the lung architecture and the presence of cellular infiltrates, and with Picrosirius Red to detect collagen deposition using standard protocols (33).
For immunohistochemistry sections were deparaffinized in xylene for 10 minutes (3 times) and rehydrated by sequential incubations in ethanol solutions (100%, 100%, 96% and 80%) for 3 minutes each. Sections were eventually washed for 5 minutes in distilled water. Antigen retrieval was performed in citrate buffer ( Full slide images were acquired using a Zeiss Axio Scan Z1 slide scanner. ImageJ software was used for relative quantification of the signal in the sections. For CD45, mannose-R, CD86, iNOS and arginase-1 staining analysis, the "HDAB" plug-in was applied, while for Picrosirius Red and a-SMA "Fast Red, Fast Blue" plug-in was used. Deconvoluted images were used to calculate the area of the nuclear staining and the area of the specific signal. The value for each section was calculated as the ratio of the specific signal to nuclei signal.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism 8 software. Data distribution was calculated using the Shapiro-Wilk test. Normally distributed data were presented as mean ± standard deviation and analyzed by unpaired two-tailed parametric t-test. For non-normally distributed data, unpaired non-parametric Mann-Whitney U test was used, and medians were presented. For comparisons of more than two groups, twoway ANOVA test with multiple comparisons (normally distributed data) and Kruskal-Wallis test with multiple comparisons (non-normally distributed data) were used. Differences were considered statistically significant for p<0.05. n refers to the number of biological replicates.
Additional methods are described in the Supplementary Material.

Infiltration of CD14 + Monocytes Into the Organs Co-Localizes With Fibrosis in SSc
In line with data published previously (34), we observed increased infiltration of CD14 + cells into the heart and lungs of SSc patients compared to control tissues ( Figures 1A, B). Moreover, CD14 signal co-localized with collagen-rich lesions in different SSc tissues ( Figure 1C). This data suggests that monocytes may participate in fibrotic processes in many organs during pathological tissue remodeling in SSc. Furthermore, heart and lungs of SSc patients showed the increased levels of aSMA + (myo)fibroblasts, CD68 + pro-inflammatory macrophages and mannose-R + (CD206) alternatively activated macrophages (Figures 2A-C) that reflects inflammatory and fibrotic microenvironment in different SSc tissues.

Activation of Profibrotic Pathways in CD14 + Blood Monocytes in SSc
We compared the transcriptome profiles of the total pool of SSc CD14 + monocytes in relation to CD14 + monocytes obtained from healthy controls (HC). Pathway analysis revealed  enrichment of SSc and fibrotic-associated processes, such as "TGF-beta signalling pathway", "Toll-like Receptor Signalling Pathway", "Angiogenesis" and "VEGF signalling pathway" in SSc CD14 + monocytes ( Figure 3A). At the transcription level, we observed dysregulated expression of essential profibrotic genes: TGFBR2, MMP9, JUN, COL9A3, COL18A1, FN1 and WNT5B (Figures 3B, C). Further, we demonstrated that SSc CD14 ++ CD16monocyte population showed significantly higher FN1 expression than SSc CD14 + CD16 + and CD14l ow CD16 + monocyte populations ( Figure 3D). Of note, we did not notice any difference in FN1 expression between these three monocyte populations in HC monocytes ( Figure 3D). Importantly, SSc CD14 ++ CD16and CD14 + CD16 + monocyte populations revealed profibrotic features by expressing upregulated FN1 levels compared to HC monocytes ( Figure 3D). This result indicated an activated, profibrotic phenotype of SSc CD14 + monocytes in the circulation, which might predispose them to acquire pathogenic phenotype in the fibrotic tissue.
Importantly, in all macrophage clusters in SSc lungs, we observed upregulated expression of FN1 in CD14 + cells (Figures 4D, E).

Profibrotic Cytokine Stimulation Induces Differentiation of Monocytes Into Myofibroblast-Like Cells
Bone marrow-derived cells have been proposed as one of the sources of myofibroblasts. Therefore, next, we evaluated the potential of CD14 + monocytes to differentiate into myofibroblast-like cells in response to profibrotic stimulation with TGFb, IL-4, IL-10 and IL-13 cytokines. We observed the induction of profibrotic gene expression including ACTA2, COL1A1 and FN1 ( Figure 5A). Additionally, stimulated monocyte-derived cells secreted ECM components: type I collagen and fibronectin (Figures 5B, C). Time-dependent profibrotic stimulation with TGFb, IL-4, IL-10 and IL-13 cytokines revealed that CD14 + monocytes upregulated profibrotic gene expression already after 24h and sustain these levels up to 7 days ( Figure 5D). Of note, we did not observe any difference in the differentiation potential between CD14 + monocytes obtained from SSc patients and healthy controls.  We hypothesized that the microenvironment might play a determinant role in the profibrotic response of CD14 + monocytes. To test this hypothesis, we co-cultured CD14 + monocytes with side-matched dermal fibroblast from SSc patients and HCs for 7 days. To analyze gene expression, cell types were separated using FACS sorting. We observed that co-culture with dermal fibroblasts induced expression of profibrotic genes ACTA2, COL1A1 and FN1 in both HC and SSc monocytes. Importantly, we noticed a significant induction of gene expression by SSc fibroblasts ( Figure 6A). Further, we aimed to observe phenotypic changes of CD14 + monocytes after co-cultures in 2D and 3D models. Monocytes, which were sorted out after the 2D co-culture and re-plated in a chamber slide, were stained with phalloidin and anti-aSMA antibody. Compared to monocytes cultured alone, cells after the co-culture acquired spindle shape, however, they failed to form stress fibers ( Figure 6B). Similar characteristics were observed for monocytes in 3D co-culture ( Figure 6C). Taken together, the fibroblast microenvironment induced a change in monocyte gene expression and morphology towards myofibroblast-like cells. Representative images of immunofluorescence staining of CD14 + monocytes with and without co-culture with fibroblasts (blue-DAPI nuclear staining, red-Phalloidin staining, green-a-SMA staining). (C) Representative images of immunofluorescence staining of HC and SSc CD14 + monocytes cocultured with fibroblasts in 3D hydrogel model (blue-monocytes stained with Cell Trace Violet, green-fibroblasts stained with CFSE, red-a-SMA staining, grey-Phalloidin staining). ns, not significant.

TGFb Is a Key Regulator of Profibrotic Gene Expression in Monocyte-Derived Myofibroblast-Like Cells
Fibrotic processes might be driven by various stimuli; however, TGFb is regarded as a pivotal mediator during SSc. To analyze the role of TGFb signalling in monocyte-to-myofibroblast-like cell differentiation, we inhibited selected TGFb downstream pathways in CD14 + monocytes treated with the profibrotic cytokine cocktail (TGFb, IL-4, IL-10 and IL-13). Treatment with the TGFBR1 kinase inhibitor SD-208 completely abolished expression of the ECM components collagen 1 and fibronectin on both mRNA and protein levels ( Figure 7). Interestingly, TGFb-induced expression of COL1A1 was suppressed by the inhibitor of canonical SMAD-dependent pathway SIS3, but not by inhibition of TAK1-dependent noncanonical pathway with 5z-7-oxozeaenol (OXO). On the other hand, blocking of canonical and non-canonical TGFb downstream pathways was sufficient to decrease expression and secretion of fibronectin ( Figure 7B).

DISCUSSION
Our data demonstrated that human peripheral blood circulating CD14 + monocytes exhibited profibrotic phenotype in SSc patients. Similarly, several reports indicated fibroblast-like/ circulating Collagen I-producing CD14 + monocytes or CD14fibrocytes as profibrotic cell progeny in SSc (35). Likewise, the alternatively activated macrophages derived from circulating CD14 + monocytes have also been considered as profibrotic cells in SSc (36). SSc patients have higher levels of circulating CD34 + CD14 + cells, Collagen-I-producing CD14 + monocytes and CD163 + monocytes (36). Importantly, monocyte count could be incorporated into the clinical assessment of patients with fibro-proliferative disorders (37), including SSc patients with interstitial lung disease (ILD). Accordingly, the elevated numbers of Collagen-I-producing fibrocytes and profibrotic monocytes in SSc patients were associated with ILD progression and implicated in the pathogenesis of SSc-ILD (36). Circulating Collagen-I-producing fibrocytes have also been coupled with a growing repertoire of human diseases including renal fibrosis (38), cirrhosis (39), nephrogenic systemic fibrosis (40), pulmonary fibrosis (41)(42)(43), and asthma (44). It has been reported that the fibronectin splice variantfibronectin extra domain A (FnEDA) is an endogenous TLR4 ligand. Importantly, elevated levels of FnEDA were demonstrated in the serum and skin lesions from SSc patients and in mouse model of cutaneous fibrosis (45). In chronic inflammatory diseases and during the healing phase of acute inflammatory reactions, the alternatively activated M(IL-4) macrophages showed significantly upregulated levels of fibronectin that suggest an active role of fibronectin + macrophages in the ECM deposition and tissue remodeling (46). Furthermore, the alveolar macrophages from lungs of SSc-ILD patients displayed elevated levels of fibronectin (47). These reports are in line with our results, which presented the elevated fibronectin levels in circulating CD14 + monocytes and CD14 + pulmonary macrophages in SSc patients and highlighted the capability of CD14 + monocytes to acquire a profibrotic phenotype. Of note, monocyte-to-macrophage differentiation has been linked with the increase production of fibronectin (48). We, therefore, concluded that tissue-infiltrating CD14 + monocytes/macrophages can be considered as ECM producers in the pathogenesis of fibrosis in SSc. All these data point to fibronectin as a potential target in therapeutic strategy in SSc.
Monocyte infiltration into the tissue affects the initiation of fibrotic processes in the skin and internal organs in SSc (49)(50)(51). Our data clearly presented infiltrating CD14 + monocytes in the collagen-rich area in SSc-ILD lungs, myocardium of SSc patients with inflammatory dilated cardiomyopathy (iDCM) and in SSc skin (27). Increased numbers of spindle-shaped CD14 + CD34 + collagen-I + cells were found in the lungs of SSc-ILD patients (36). The bone marrow origin of fibroblasts or myofibroblasts in different tissues under homeostasis and disease condition has been debatable for many years now. In vivo data, with the use of cutaneous wound mouse model, confirmed that at least a part (15%-20%) of the spindle-shaped dermal fibroblasts were bone marrow origin (52). In contrast, another report suggested that bone marrow-derived progenitors contributed to the inflammatory cell pool infiltrating the wound area; however, they did not differentiate into dermal (myo) fibroblasts at the wound site (53,54). In the bleomycininduced skin fibrosis model, a significant number of CD45positive collagen-producing cells of bone marrow origin contributed to collagen production during dermal fibrogenesis but not under homeostasis, indicating bone marrow originated cells as a cell source for (myo)fibroblasts under disease condition (55). Similarly, in bleomycin-induced lung fibrosis mouse model bone marrow-derived progenitors gave rise to collagenproducing cells but not to aSMA + myofiboblasts in the inflamed/fibrotic lungs (56,57). Oppositely, in the renal fibrosis the inflammatory macrophages, characterized by alternatively activated macrophage markers, acquired fibroblast phenotype by collagen I and aSMA expression and actively contributed to the fibrogenesis (58).
Nevertheless, the transition of bone marrow-derived monocytes or macrophages (mainly alternatively activated macrophages) into functional myofibroblasts has been mediated by canonical SMAD2/3-dependent or non-canonical TAK-1-dependent TGFb signal (24,(59)(60)(61). Accordingly, our results confirmed that both canonical and non-canonical TGFb pathways were important to obtain the fibroblast-like phenotype of circulating CD14 + monocytes, indicating the broad treatment options.
The pathophysiology of SSc is closely related to the activated TGFb-dependent pathway. The levels of a latent, not-active form of TGFb is similar in SSc and healthy serum samples. Active TGFb serum levels were significantly higher in SSc, mainly in dcSSc, patients and correlated with clinical manifestations (digital ulcers, lung fibrosis, positive antitopoisomerase I and higher modified Rodnan score) (62). These results indicate TGFb as a potential marker of fibrotic and vascular involvement in SSc on one side, and on the other side, it reflects an altered microenvironment, which may predispose circulating monocytes towards an activated and profibrotic phenotype. Accordingly, we used this cytokine, among others, to activate circulating monocytes towards a profibrotic phenotype. As expected, CD14 + monocytes stimulated with TGFb, IL-4, IL-10 and IL-13 or co-cultured in 2D and 3D model with human dermal fibroblasts were able to adopt a functional myofibroblastlike phenotype that may indicate these cells as one of the cellular sources of profibrotic cells upon their entrance into the tissues. The previous report indeed showed that GM-CSF, IL-4 and endothelin-1 alone or in combination induced myofibroblastlike differentiation of circulating SSc and healthy monocytes (63). Additionally, the profibrotic phenotype of circulating CD14 + monocytes from SSc-ILD patients was confirmed by the expression of CD163 and boosted secretion of CCL18 and IL-10 in response to pro-inflammatory stimuli (36). However, as in our stimulation conditions, monocytes from SSc patients showed no differences in acquiring more pronounced profibrotic phenotype compared to monocytes from healthy donors. Therefore, we assume that under inflammatory conditions (i.e. in SSc) more circulating already activated monocytes are recruited to the injury site, where they play a profibrotic role, either by the transition into fibroblast-like cells or by the production of profibrotic stimuli. In line with this hypothesis, Bhandari et al. demonstrated that soluble factors in the local milieu are crucial for pro-fibrotic activation of SSc macrophages (measured as secretion of CCL2, IL-6, TGFb), arisen from circulating monocytes either stimulated with SSc sera or conditioned media from indirect co-culture with dermal SSc fibroblasts (64). These results point to monocytes and monocytederived macrophages as probable key players in fibrogenesis in SSc, suggesting targeted cell therapeutic options as feasible and favorable in SSc.
Besides a known and active role of IL-6 and TGF-b signalling in the pathogenesis of SSc, there is increasing evidence that also Th-2 cytokines: IL-4 and IL-13, are involved in the pathology of SSc (65,66). Both cytokines relate to profibrotic responses, including connection with increased expression of novel myofibroblast marker: periostin, a matricellular protein important in fibrogenesis (67). Based on our data, the combination of TGFb, IL-4, IL-10 and IL-13 was sufficient to obtain a fibroblast-like phenotype by circulating CD14 + monocytes, indicating that IL-4/IL-13 targeted therapy might be an attractive option against fibrogenesis in SSc (68). Indeed, a randomized, double-blind, placebo-controlled, 24-week, phase II, proof-of-concept study (trial registration number: NCT02921971) with romilkimab (a bispecific immunoglobulin-G4 antibody that binds and neutralizes IL-4/IL-13; SAR156597) in early dcSSc patients revealed a beneficial effect of this treatment on skin disease changes, i.e. statistically significant decrease in mRSS from baseline to week 24 versus placebo (69).
Given the alerted activation status of circulating CD14 + monocytes and their ability to accumulate in the affected organs in SSc, we believe that further research should focus on finding even more targeted therapies, which may invert the activation status of monocytes in the circulation.

DATA AVAILABILITY STATEMENT
The RNAseq datasets for this study can be found in the Gene Expression Omnibus under the accession numbers: GSE157840, GSE128169.

ETHICS STATEMENT
Human blood samples and skin biopsies collection were approved by the local ethics committee of the Canton Zurich (KEK-ZH 515, PB-2016-02014, KEK-Nr 2018-01873). All study subjects provided written informed consent. The experiments with re-use of human material were approved by Swissethics (KEK-Nr 2019-00058, KEK-Nr 2018-01873) and were performed in conformity with the principles outlined in the Declaration of Helsinki. The animal study was reviewed and approved by Commission on Animal Experimentation of the Canton Zurich.