Neovascularization in rheumatoid arthritis pannus is orchestrated by endothelial-pericyte signaling

Neovascularization in rheumatoid arthritis pannus is orchestrated by endothelial-pericyte signaling | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Neovascularization in rheumatoid arthritis pannus is orchestrated by endothelial-pericyte signaling Yoko Miura, Nadia Milad, Atsushi Usami, Shunichi Kosugi, Hideki Noguchi, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9546936/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Background Neovascularization is a defining feature of rheumatoid arthritis (RA) pannus, yet the molecular signaling patterns that govern this process remain incompletely understood. Here, we sought to characterize CD146⁺ populations in inflammatory synovium, including endothelial cells, pericytes, and their involvement in neovascularization. Methods Using collagen-induced chronic polyarthritis model in D1BC transgenic mice, bulk and single-cell RNA sequencing of synovial cells was performed, followed by integrative analyses with published single-cell RNA-seq datasets from human rheumatoid arthritis and lung cancer. Results Bulk RNA-seq revealed that CD146 expression stratifies endothelial and perivascular populations with angiogenic potential. Single-cell RNA sequencing of sorted CD146⁺ cells further classified endothelial subsets into Apln ⁺, Aplnr ⁺, Apln ⁺/ Aplnr ⁺, and double-negative groups, closely mirroring the aCap/gCap dichotomy described in pulmonary capillary repair. Ligand-receptor and gene ontology analyses demonstrated that these CD146⁺ endothelial subgroups engage in neovascularization but lack signatures associated with gas exchange, supporting their identity as repair-type vasculature. A second CD146⁺ population expressed type-2 pericyte markers, including Cspg4 , but lacked Acta2 and Pdgfrb , suggesting a transcriptionally distinct pericyte subset that interacts preferentially with Apln ⁺ and Aplnr ⁺ endothelial cell subsets. Immunohistochemistry confirmed the spatial association of NG2⁺ pericytes with CD31⁺ neovessels in pannus tissue. Conclusion Together, these findings reveal that RA pannus neovascularization proceeds through a pulmonary-like capillary repair program orchestrated by CD146⁺ Apln ⁺ lineage endothelial cells and NG2⁺ type-2 pericytes. This work establishes a unifying framework linking inflammation-driven angiogenesis with tissue repair vascular biology and identifies CD146⁺ subsets as potential targets for modulating pathological neovascularization in RA. rheumatoid arthritis synovium neovascularization pericyte endothelial cell CD146/Mcam pannus single-cell RNA sequencing (scRNA-seq) bulk RNA sequencing (bulk RNA-seq) vascular remodeling and regeneration Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Background Rheumatoid arthritis (RA) is a chronic inflammatory disease characterized by pannus formation in the joint due to hyperplasia of the synovial membrane accompanied by inflammation and bone erosion. Neovascularization is observed in the pannus early in disease and is known to influence disease progression [ 1 ]. This angiogenesis is thought to contribute to pathogenesis by facilitating inflammatory cell infiltration as well as increasing supply of oxygen and nutrients to pathologically activated remodeling cells. In fact, some research suggests that reducing neovascularization could be a therapeutic strategy in RA [ 2 ]. Endothelial cells and pericytes are the key cellular components of newly formed blood vessels found in arthritic joints as well as in tumor neovascularization, where endothelial cell populations are highly heterogeneous, encompassing distinct vascular-type and organ-specific characteristics. When neovascularization occurs during wound healing, endothelial cells form capillaries together with pericytes [ 3 ]. Generally, pericytes are composed of heterogeneous cells and possess the ability to undergo flexible transformation. Chondroitin sulfate proteoglycan 4 ( Cspg4 , encodes neuron-glial antigen 2, NG2) is the most widely recognized pericyte marker; on the other hand, Pdgfrb -negative pericytes were identified during skin development [ 4 ]. As a result, they differentiate into fibroblast-like or myofibroblast-like cells in response to various stimuli. Recent studies have revealed two distinct pericyte states as functional differentiation stages: type-1 pericytes and type-2 pericytes. Type-1 pericytes share gene expression patterns with myofibroblast, producing senescence-associated secretory phenotype (SASP) markers and expressing Acta2 , collagens such as Col1a1 , Tgfb1 (TGF-β), and Pdgfrb , similar to perivascular fibroblasts [ 5 – 7 ]. These cells are particularly important during stages of fibrosis following tissue damage. Meanwhile, type-2 pericytes play a regenerative role, expressing markers such as CD146, NG2 ( Cspg4 ), and fibroblast growth factor 5 ( Fgf5 ), while Pdgfrb expression is transiently reduced [ 8 , 9 ]. Also, other genes such as stanniocalcin 2 ( Stc2 ) and SRY-box transcription factors 2 and 10 ( Sox2 and Sox10 ) might be involved in the regenerative phase of pericytes during wound healing [ 10 , 11 ]. The contribution of pericytes and other cell types, such as fibroblasts and immune cells, in the formation of new vessels within arthritic joints have yet to be fully elucidated. In this study, synovial cells were isolated from the inflamed joints of a RA mouse model and analyzed using flow cytometry. We used CD146, or melanoma cell adhesion molecule (MCAM), surface expression to stratify cells since it is predominantly expressed by endothelial cells but also found in pericytes, smooth muscle cells, and some immune cells [ 4 , 12 ] and its expression has been linked to vascular regeneration, metastasis, and tumor progression in various types of cancer [ 13 ]. Next, we examined the specific roles of CD146 + cells through bulk RNA sequencing and single-cell RNA sequencing (scRNA-seq). We found that these CD146 + endothelial cells were involved in neovascularization and could be classified into subgroups based on Apln and apelin receptor ( Aplnr ) expression, vascular markers typically expressed in the lungs [ 14 ], where an intermediate population of Apln + / Aplnr + cells was observed, likely a transitory phenotype. Interaction analysis revealed that these cell populations participate in neovascularization reminiscent of vascular repair processes observed in the lung. These gene expression data were compared with previously reported gene ontology analyses of human joints, human lung cancer and mouse joint tissues and found to correlate well with neovascularization patterns in these studies. Overall, we conclude that Apln -positive endothelial cells and type-2 pericytes work together to promote neovascularization in the inflamed joint thus promoting pannus formation and remodeling in the context of polyarthritis. Interestingly, neovascularization in the pannus shares common features with pulmonary capillary repair, heavily relying on pericyte signaling, which may be useful as therapeutic targets in the development of treatments for arthritis and perhaps associated lung cancer. Methods Chronic polyarthritis mouse model A low dose of bovine collagen type II (bColII) was administered to induce chronic polyarthritis in D1BC mice as previously described [ 15 ]. Briefly, 7- to 8-week-old D1BC mice were housed in a pathogen-free animal care facility at Nagoya City University Medical School in accordance with institutional guidelines (animal ethics approval #21-033H02). Mice were anesthetized with isoflurane and administered subcutaneously at all 4 limbs (total 10 mg per mouse) bColII emulsified with an equal volume of complete Freund’s adjuvant (BD Biosciences, San Jose, CA). The same protocol was used for secondary immunization with bColII after 3 weeks from 1st immunization, except without adjuvant. Immunohistochemistry Normal and inflamed joints were fixed in 4% paraformaldehyde/PBS then decalcified in buffer containing 2.5% EDTA and 7% sucrose at 4℃ for one month. Paraffin sections (2 mm thickness) were deparaffinized, rehydrated, then underwent antigen retrieval in citrate buffer at 121ºC for 5 minutes. Following quenching (3% H 2 O 2 in methanol for 15 min) and blocking (10% goat serum in TBS-T for 20 min), slides were stained using the following primary antibodies diluted in 10% goat serum in TBS-T: rabbit anti-CD31 (1/1,500; ab28364; Abcam, Cambridge, UK), rabbit anti-MCAM/CD146 (1/500; #81701; Cell Signaling Technology, Danvers, MA, USA), rabbit anti-PDGFRβ (1/100; #3169; Cell Signaling Technology), rabbit anti-APLNR (1/500; bs-21310R; Bioss, Boston, MA, USA), anti-NG2 (1/100; AB5320; Merck Millipore, Burlington, MA, USA), and rabbit anti-vimentin (1/1,000; #5741; Cell Signaling Technology). Slides were then incubated with Histofine Simple Stain mouse anti-rabbit MAX-PO secondary antibodies (Nichirei Biosciences, Tokyo, Japan) and then Opal multiplex fluorescent immunohistochemistry system (Akoya Biosciences, Marlborough, MA, USA) were used according to the manufacturer’s instructions. All images were captured using a fluorescence microscope (BZ-X710; Keyence, Osaka, Japan). Pannus collection and bulk RNA sequencing (RNA-seq) Synovial cell isolation and flow cytometry was performed as previously described [ 15 ]. Briefly, approximately 1 x 10 5 cells of isolated synovial cells were blocked FcgR with anti-CD16/ CD32 antibody (BD Biosciences). Within the CD11b-negative fraction, cells were then sorted by fluorescence activated cell sorting (FACS) on a FACS Aria II (BD Biosciences) as follows: CD146 hi , CD146 mid , CD140a + , and double-negative. Indexed libraries were then sequenced using the NovaSeq platform (Illumina, San Diego, CA, USA) by Macrogen Inc (Seoul, South Korea) and can be found online via GEO accession number GSE311750. Single-cell RNA sequencing (scRNA-seq) Synovial cells obtained from 7 mice were pooled and FcgRs were blocked by anti-CD16/32 antibody (BD Biosciences) at 4℃ for 10 minutes. After washing by FACS buffer, synovial cells were collected positive selection using CD146-beads and MS column (Miltenyi Biotec). The scRNA-seq analysis was performed by ImmunoGeneTeqs Inc (Chiba, Japan) according to the TAS-seq protocol [ 16 ]. Raw sequencing data were deposited to the National Center for Biotechnology Information Gene Expression Omnibus and are accessible through Gene Expression Omnibus (series accession number GSE311751). Human scRNA-seq data for RA [ 17 ] and lung cancer [ 18 ] were obtained from the ArrayExpress site ( https://www.ebi.ac.uk/biostudies/arrayexpress ) at the European Bioinformatics Institute (EBI). Details of the bioinformatics protocol and other relevant details of the in silico analysis are described in the Supplementary Methods. Pannus collection and bulk RNA sequencing (RNA-seq) analysis After cells were collected from the pannus and washed with PBS containing 2% FBS, synovial cells were stained using the following primary antibodies: CD11b-PE (clone M1/70, BD Biosciences), podoplanin-BV421 (clone 8.1.1, BioLegend, San Diego, CA, USA), PDGFRα (CD140a)-APCvio770 (clone REA637, Miltenyi Biotec, NRW, Germany), CD146 (LSEC)-APC (clone ME-9F1, Miltenyi Biotec), PDGFRβ (CD140b)-PEvio770 (clone REA634, Miltenyi Biotec), and 7-aminoactinomycin D (7-AAD, BD Biosciences). These cells were analyzed using a FACSCanto II (BD Biosciences). Total RNA extracted from the collecting synovial cells was used in the ReliaPrep RNA Tissue Miniprep System (Promega, Madison, WI, USA) and total RNA concentration was calculated by Quant-IT RiboGreen (#R11490, Invitrogen, Waltham, MA, USA). To assess RNA integrity, samples were run on the TapeStation RNA Screentape (Agilent, Santa Clara, CA, USA). Only high-quality RNA preparations, with RIN greater than 7.0, were used for RNA library construction. The RNA isolated from each sample was used to construct sequencing libraries with the SMART-Seq™ mRNA kit (Takara Bio, Kusatsu, Japan), following the manufacturer's protocol. First-strand cDNA synthesis (from total RNA or cells) is primed by the 3’ SMART-Seq CDS Primer II A and uses the SMART-Seq v4 Oligonucleotide for template switching at the 5’ end of the transcript. Sequencing was performed using the Illumina NovaSeq 6000 (San Diego, CA). The first-strand cDNA selectively binds to SPRI beads leaving contaminants in solution which is removed by a magnetic separation. The beads are then directly used for PCR amplification. The Advantage 2 Polymerase Mix has been specially formulated for efficient and accurate amplification of cDNA templates by long-distance PCR. PCR-amplified cDNA is purified by immobilization on AMPure XP beads. The beads are then washed with 80% ethanol and cDNA is eluted with Elution Buffer. Prior to generating the final library for Illumina sequencing, the Covaris AFA system is used for controlled DNA shearing. The resulting DNA will be in the 200–400 bp size range. These cDNA fragments then go through an end repair process, the addition of a single ‘A’ base, and then ligation of the indexing adapters. The products are then purified and enriched with PCR to create the final cDNA library. The libraries were quantified using qPCR according to the qPCR Quantification Protocol Guide (KAPA Library Quantification kits for Illumina Sequencing platforms) and qualified using the Agilent Technologies 4200 TapeStation (Agilent Technologies, Waldbronn, Germany). Indexed libraries were then sequenced using the NovaSeq platform (Illumina, San Diego, CA, USA) by Macrogen Inc (Seoul, South Korea) and can be found online via GEO accession number GSE311750. Reverse transcription quantitative PCR (RT-qPCR) Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany) for pannus, lung, kidney, and liver tissues according to the manufacturer’s instructions. For quantitative PCR (qPCR), cDNA was synthesized using ReverTra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO, Osaka, Japan). The qPCR was performed using the PrimeTime Gene Expression Master Mix (Integrated DNA Technologies, Coralville, IA, USA). The relative expression of each gene was determined by an internal control using Hprt for each sample and expressed as a fold change of the control group. Single-cell RNA sequencing (scRNA-seq) analysis Briefly, hashtag-stained single nuclei suspension was pooled and approximately 20,000 cells were loaded onto the BD Rhapsody system (BD Biosciences). After loading unique cell barcode-immobilized magnetic beads, cells were lysed. Next, reverse transcription of polyA RNA and hashtag oligos derived from each single cell was performed according to the manufacturer’s instructions. Exonuclease I treatment and cDNA/hashtag amplification were performed as per the TAS-seq protocol. Sequencing libraries were prepared using the NEBNext Ultra II FS Library Prep Kit for Illumina (New England Biolabs, Ipswich, MA, USA) and quantified using the KAPA Library Quantification Kit (KAPA Biosystems, Wilmington, MA, USA) and a MultiNA system (Shimazu, Kyoto, Japan). Sequencing analysis was performed by a NovaSeq 6000 sequencer and a Novaseq S4 200 cycles v1.5 kit (Illumina) [Read1: 67 base-pair, Read2: 155 base-pair]. The resulting pair-end fastq files (R1: cell barcode reads, R2: RNA reads) of the TAS-Seq data were processed according to a previously reported method. Ensemble reference RNA (GRCh38 release-101) was used for mapping RNA reads and the known barcode sequence of each hashtag was used for mapping hashtag reads. Demultiplexing of each cell by hashtag read count was performed according to the method described previously. Bioinformatics analysis of scRNA-seq data Human scRNA-seq data for RA [ 17 ] and lung cancer [ 18 ] were obtained from the ArrayExpress site ( https://www.ebi.ac.uk/biostudies/arrayexpress ) at the European Bioinformatics Institute (EBI). The gene expression matrices were imported into Seurat (version 5.2.0), and quality control (QC) was performed to remove low-quality cells [ 16 ]. Cells meeting the following criteria were retained: a minimum of 200 and a maximum of 6000 genes per cell, a minimum of 500 and a maximum of 40,000 reads per cell to exclude empty droplets and potential doublets. To exclude potentially dying cells, cells showing mitochondrial gene content below 20% in the RA data or 5% in the lung cancer data were removed. The gene expression data were normalized using the SCTransform function of the sctransform package, specifying percent.mt for vars.to.regress and 3 for min_cells. Principal component analysis (PCA) was performed for dimensionality reduction, and cells were clustered using the FindNeighborsand function with the dims parameter set to 1:25 and the FindClusters function with the resolution parameter set to 1.5. The clusters were visualized using Uniform Manifold Approximation and Projection (UMAP) using the RunUMAP function setting the dims parameter to 1:25. Cell type annotation was performed using the celldex, SingleR, and SingleCellExperiment packages. The celldex human reference dataset was obtained using the R command (celldex, HumanPrimaryCellAtlasData) [ 19 , 20 ]. Each cell was annotated with the corresponding cell type in the celldex reference using the SingleR and as.SingleCellExperiment functions, and these cells were labelled as ‘singlr_labels’. Visualization of gene expression cells on UMAP and t-SNE We visualized cells expressing a specific gene on dimension reduction images (FIt-SNE for the mouse polyarthritis data or UMAP for the other data) using the Seurat FeaturePlot function with the option ‘max.cutoff=q90’ to limit the maximum color density to the 90% percentile of the expression value. To visualize cells co-expressing multiple genes on the dimension reduction images, we used the DimPlot function of Seurat. Specifically, we specified the cell list objects corresponding to each gene as the cells.highlight argument, and defined the color and size of the highlighted cells using the cols.highlight and sizes.highlight arguments. Biological theme comparison between cell clusters Differentially expressed genes (DEGs) in each cell cluster of the mouse polyarthritis data were identified using the Seurat FindAllMarkers function under the following conditions: the Wilcoxon rank-sum test, logfc.threshold = 0.25, and min.pct = 0.25). Gene ontology (GO) enrichment analysis was performed for the DEGs in each cell cluster using the R clusterProfiler package and the org.Mm.eg.db mouse GO database. The gene symbols of DEGs were converted to ENTREZ IDs, and enrichment analysis was performed using the compareCluster function with the options fun and organism set to ‘enrichWP’ and ‘Mus musculus’, respectively. (The output data was restricted to GO level 4 using the gofilter function. The output dot plots were created using the dotplot function and customized with the ggplot2 R package. Identification of Apln + and Aplnr + cell populations When counting the number of cells expressing a specific single or multiple genes (e.g., Apln and Aplnr ), cells with ≧ 0.1 expression values of the corresponding genes were targeted in human RA data, while cells with ≧ 1.0 values were targeted in other data. The data were visualized by pie charts. LIANA analysis Cell-to-cell communication (CCC) between cell clusters in the mouse polyarthritis data was analyzed using the R package LIANA for ligand-receptor analysis framework, which calculates a consensus CCC rank using multiple resources and methods related to CCC. The Mouse Consensus was selected as resource using the select_resource function, and the mouse polyarthritis data object created by Seurat was processed with the liana_wrap and liana_aggregate functions. Heatmaps of ranked interactions for each cluster were generated using the liana_dotplot function. To generate a frequency plot of predicted interactions, predicted ligand-receptor pairs filtered for statistical significance with p-values < 0.05 were plotted using the heat_freq function. Statistical analysis Data were visualized and statistically analyzed using GraphPad Prism 10.5.0 (Boston, MA, USA). Histogram data is expressed as the mean ± standard error (SE). Data was not transformed and outlier exclusion was not required. One-way ANOVA tests followed by Tukey’s post-hoc correction for multiple comparisons and Dunnet’s test for parametric data were used when several independent groups were compared. All statistical analyses used two-tailed tests with a p-value of less than 0.05 considered significant. Results Neovascularization in the inflammatory synovium of mice with chronic polyarthritis D1BC mice develop chronic and severe joint polyarthritis following low-dose administration of bovine type II collagen, exhibiting characteristics closer to RA compared to conventional collagen-induced polyarthritis mouse model [ 15 ]. First, we used immunohistochemistry to visualize blood vessels, pericytes, and fibroblasts within the pannus of inflamed joints, where CD31 + blood vessels were observed, typically surrounded by Pdgfrβ + pericytes and distinct from vimentin + (Vim) fibroblasts (Fig. 1 A). In contrast, normal joints isolated from naïve mice lack obvious vascularization, exhibiting minimal staining in the joints for any of these markers. Expression profile of blood vessel-associated cells in the inflammatory synovium To further investigate neovascularization in the inflamed synovium, we examined various subtypes of vascular cells stratified by their expression of CD146 and CD140a (Pdgfrα), markers known to be upregulated in angiogenesis and constitutively expressed in perivascular cells such as pericytes [ 21 ]. Cells from the inflammatory synovium were cultured for less than three days (Figs. 1 B-C). FACS analysis revealed that CD11b-negative cells were divided into four distinct groups: CD146 high , CD146 mid , CD146 low CD140a high , and double-negative cells (Fig. 1 C-D). The majority of synovial cells were identified as CD11b-positive synovial macrophages (40%, Type A) and CD140a + fibroblasts (50%, Type B) (Figs. 1 C-E). Within the CD11b-negative population, CD146 high and CD146 mid cells represented 3.3 and 3.6% of total cells, respectively, indicating that vascular cells can be found in isolated synovial cells though in relatively small abundance. Next, bulk RNA-seq was used to characterize the transcriptomic profiles of CD146 high , CD146 mid , CD140a + , and double-negative populations. The CD146 high populations had the highest number of differentially expressed genes compared to the double-negative population. Upregulated genes in the CD146 high populations included several endothelial cell markers, such as Icam2 , Cd151 , Apln , and Ednrb (Fig. 1 F) [ 22 ]. On the other hand, CD146 mid and CD140a + cells showed similar expression patterns, primarily expressing pericyte markers Pdgfrb , Ang2 , and Acta2 (type-1 pericytes); osteo-chondrogenic markers Runx2 and Sox9 ; and fibroblast marker, Pdgfra (Fig. 1 F). Conversely, the double-negative population expressed few endothelial cell- or pericyte-associated genes; instead, they were found to express osteoclast and macrophage markers. Using RT-qPCR, we found that CD146 high cells expressed high levels of Cspg4 , while CD140a + cells showed high expression of Runx2 and Sox9 , a chondrocyte marker (Figs. 1 G-I) [ 5 , 9 , 22 ]. These results suggest that synovial cells isolated from pannus of inflammatory synovium, though predominantly composed of CD11b + cells and some osteochondrogenic cells [ 23 ], also include CD146 + endothelial cells, type-1 pericytes ( Pdgfrb + ), and type-2 pericytes ( Pdgfrb − ). Single-cell RNA-seq analysis of mouse synovial cells Bulk RNA-seq analysis of synovial cells revealed newly formed vessels contained type-1 and type-2 pericytes in CD146 high , CD146 mid , and CD140a + populations. To further characterize these cell types and assess their respective roles neovascularization, we performed single-cell RNA-seq analysis on synovial cells enriched using anti-CD146 antibody-conjugated ferrite beads. Cluster analysis revealed that CD146 + cell population is still small; however, it consists of distinct cell populations such as endothelial cells (cluster 14 or c14), and type-2 pericytes (c6, c11, and c19) (Figs. 2 A-B, Table 1 , and Supplementary Fig. 1, cluster annotation). We also observed fibroblast clusters (c3, c4, c5, c6, c8, c9, c11, c15, c16, c18, and c21), including type-1 pericytes, and macrophages (c0, c1, c2, c7, c10, c13, c19, and c20). Macrophage and pericyte subpopulations clustered with high intracluster correlation, whereas the CD146 + subpopulation was more distinct from these cells. Interestingly, endothelial cells (c14) showed no strong correlation with any other cluster. Combined with enrichment analysis, various cell populations were annotated (Figs. 2 C-H and Supplementary Fig. 1). We found that CD146 + cells do not express either Pdgfra nor Pdgfrb . Further trajectory analysis using Slingshot algorithm revealed that endothelial cells (c14) and one group of type-2 pericytes (c11) were terminal clusters and express matrix metalloproteinase, Mmp2 (Table 2 ). Conversely, the remaining type-2 pericytes (c6 and c19) were predicted to be immature (Figs. 2 I-K). Interestingly, c3 is distinct from the major population of fibroblasts likely due to a high prediction of immaturity; however, this population seems related to pericytes as it expresses Pdgfrb . Table 1 Endothelial cell markers detected in CD146 + / Pdgfrb − cells. Endothelial cell markers CD146 + / Pdgfrb − cells Interaction Cluster 14 Cluster 6 Cluster 11 Cluster 19 Angiopoietin 2, Angpt2 a + - - - Apelin, Apln b + - - - Apelin receptor, Aplnr b + - - - CD31, Pecam1 homophilic + - - - CD34, Cd34 + - - - Claudin 5, Cldn5 homophilic + - - - Endothelin 1, Edn1 + - - - Kinase insert domain receptor, Kdr c + - - - Protein C receptor, Procr + - - - P selectin, Selp + - - - TEK receptor tyrosine kinase (TIE2), Tek a + - - - Tyrosine kinase with Ig-like and EGF-like domains (TIE1), Tie1 a + - - - Vascular endothelial growth factor A, Vegfa c + - - - VE-cadherin 5, Cdh5 homophilic + - - - Vimentin, Vim + + + - Endothelial cell markers detected in CD146 + /Pdgfrb − cells are listed. Each cluster is annotated using tSNE data from scRNA-seq. Interaction types are indicated by matching letters (a–c), representing predicted ligand–receptor pairs. +: positive expression, -: negative expression Table 2 Type-2 pericyte markers detected in CD146 + / Pdgfrb − cells. Type-2 pericyte markers CD146 + / Pdgfrb − cells Cluster 14 Cluster 6 Cluster 11 Cluster 19 Chondroitin sulfate proteoglycan 4, neuron-glial antigen 2 (NG2), Cspg4 - + + + Class III β-tubulin, Tubb3 - + + + Fibroblast growth factor 5, Fgf5 - + + + Matrix metalloproteinase 2, Mmp2 - - + - Nerve growth factor receptor, Ngfr - + + + N-cadherin, Cdh2 - + + + Nestin, Nes - + + + Platelet-derived growth factor A, Pdgfa - + + + Platelet-derived growth factor B, Pdgfb + + + + Stanniocalcin 2, Stc2 - + + + SRY-box transcription factor 2, Sox2 - + + + SRY-box transcription factor 10, Sox10 - + + + Vascular endothelial growth factor A, Vegfa - + + + Type-2 pericyte markers detected in CD146 + /Pdgfrb − cells are listed. Each cluster is annotated using tSNE data from scRNA-seq. +: positive expression, -: negative expression Gene expression profile in CD146 + endothelial cells is conserved in rheumatoid arthritis and lung cancer Using scRNA-seq data from previous studies on rheumatoid arthritis (RA) and lung cancer [ 17 , 18 ], we found similar subpopulations of CD146 + endothelial cells in both human RA and lung cancer patients (Figs. 3 A-E). The gene expression characteristics of endothelial cells from our dataset has been summarized in Table 1 . Several angiogenesis-related ligand and receptor genes were found to expressed by endothelial cells, such as angiopoietin 2 ( Angpt2 ), tyrosine kinase with immunoglobulin-like and EGF-like domains ( Tie1 ), and TEK receptor tyrosine kinase ( Tie2 ) (Table 1 , interaction a); Apln and Aplnr (Table 1 , interaction b); and vascular endothelial growth factor A ( Vegfα ) and Kinase insert domain receptor ( Kdr ) (Table 1 , interaction c). Thus, neovascularization in the pannus may largely be driven by endothelial cell autocrine and/ or paracrine signaling. To further characterize these endothelial cells, qPCR was performed for Apln and Aplnr . Both genes were highly expressed in pannus from mouse polyarthritis, but were only faintly detectable in the kidney and the liver (Figs. 3 F-G). Furthermore, immunohistochemical data in mouse polyarthritis support overlapping signals between CD31, Aplnr, and CD146 (Figs. 3 H-I). CD146 + endothelial cells express neovascularization-related genes Overlap in Apln and Aplnr gene expression in scRNA-seq data revealed the presence of Apln + / Aplnr + cells within specific populations (Figs. 4 A-C). As previous studies have found that Apln -expressing Aplnr -positive gCap ( Apln + general capillary, gCap) exist as intermediate cells during the differentiation process from gCap to aerocyte (aCap) [ 3 ], we observed that Apln + / Aplnr + cells exhibiting characteristics similar to Apln + gCap exist in mouse polyarthritis model and human RA, but only 0.4% in lung cancer (Fig. 4 D). Since these cells play a role in neovascularization rather than gas exchange, we assessed the expression of neovascularization genes in this specific population (Figs. 4 E-G). In mouse polyarthritis, most Apln + / Aplnr + cells expressed Angpt2 , Cd34 , and Kdr , while 35% of Apln + / Aplnr + cells expressed two additional neovascularization-associated genes, Cldn5 and Procr , an expression pattern that corresponds well with observations in lung cancer. Unique transcriptomic profile of pericyte subpopulations Pdgfrb , though widely expressed in mature pericytes, was not found to be expressed in most CD146 + cells (including type-2 pericytes) (Figs. 5 A-B, Table 2 , and Supplementary Tables 1). Thus, CD146 − Pdgfrb + type-1 pericytes and CD146 + Pdgfrb − type-2 pericytes were mutually exclusive populations. Cells expressing Mcam (CD146 gene) were mainly found in c14, c6, c11, and c19, clusters which comprise the CD146 mid and CD146 high populations in the initial FACS analysis (Fig. 1 D) and were classified by Seurat cluster analysis as endothelial cells (c14) and type-2 pericytes (c6, c11, and c19) (Tables 1 – 2 ). In putative type-1 pericytes, only c3 expressed endothelin receptor type A ( Ednra ) and type B ( Ednrb ) (Supplementary Table 1). To probe further, we looked at other pericyte markers in these clusters and found that Fgf5 , in addition to Cspg4 , but not Pdgfrb , were highly expressed in these populations (Figs. 2 B, 5 C-D). In addition, we found that Stc2 , Sox2 , and Sox10 were elevated in c6, c11, and c19, genes predominantly associated with neovascularization and progenitor-like behavior at the site of injury (Figs. 5 E-G). To investigate the spatial localization of these pericyte-like cells in the inflamed joints, we used immunohistochemical staining for Ng2, CD31, and Pdgfrβ, which revealed that most CD31 + vessels were surrounded by Pdgfrβ + fibroblasts and some vessels were covered with Ng2 + cells in the same joint and little co-localization was noted (Fig. 5 H). Endothelial cells and pericytes drive pannus neovascularization signaling Next, we analyzed cell-cell interaction between CD146 + endothelial cells (c14) and type-2 pericytes (c6, c11, and c19) with the ligand-receptor prediction tool LIANA, which predicts interactions between each cell from scRNA-seq data (Fig. 6 A-B). Many ligands released by fibroblast and pericyte populations interact strongly with receptors expressed by endothelial cells (cluster 14), while macrophage-associated ligands only showed weak interactions with endothelial cell receptors (Figs. 6 C-F. and Supplementary Fig. 2). Focusing on CD146 + cells, delta like canonical notch ligand 4 ( Dll4 ) was found to be expressed in c14 and is known to interact with Notch4 receptor. Notch4 was expressed in c6, c11, and c19 and is a receptor involved in branch promotion and endothelial stabilization during angiogenesis [ 24 , 25 ]. Jagged 1 ( Jag1 ) was also expressed in c14 and has previously been shown to interact with Notch4 and block Dll4 binding (Table 3 ) [ 24 ]. Other genes involved in angiogenesis such as slit guidance ligand 3, Slit3 , was expressed in c6, c11, and c19 while its receptor, roundabout guidance receptor 4 ( Robo4 ) was expressed on c14 (Table 3 ) [ 26 ]. Also, pleiotrophin ( Ptn ) was found to be expressed by c6, c11, and c19, which may modulate excessive formation of microvasculature via binding to Kdr [ 27 ], expressed on c14 (Tables 1 – 3 and Supplementary Table 1). Furthermore, protein tyrosine phosphatase receptor type B ( Ptprb) , which negatively regulates Kdr by dephosphorylation, was also expressed on c14 [ 28 ]. Table 3 Molecular signature of endothelial cell and pericyte interactions in CD146 + / Pdgfrb − cells. Signaling molecules CD146 + / Pdgfrb − cells Interaction Hypoxia Cluster 14 Cluster 6 Cluster 11 Cluster 19 Delta-like canonical Notch ligand 4, Dll4 a * + - - - Notch 1, Notch1 a,b * + + + - Notch 4, Notch4 a,b + - - - Jagged 1, Jag1 b + + + + Roundabout guidance receptor 4, Robo4 c * + - - - Slit guidance ligand 3, Slit3 c - + + + Pleiotrophin, Ptn - + + + Protein tyrosine phosphatase receptor type B, Ptprb * + - - - Each cluster is annotated using tSNE data from scRNA-seq. Interaction types are indicated by matching letters (a–c), representing predicted ligand–receptor pairs. Hypoxia associated genes are marked with *. +: positive expression, -: negative expression However, based on the proximity of these pericyte-like cells to newly formed pannus vessels, we used cell-cell interaction data via LIANA analysis to look at which ligands expressed in c6, c11, and c19 may be signaling through receptors expressed on endothelial cells (c14). As expected, several of the ligand-receptor factors predicted were associated with neovascularization and interaction between type-1 and type-2 pericytes (Figs. 6 A-B, Supplementary Fig. 4). Discussion In this study, we characterized neovascularization in the arthritic joint using a mouse model of polyarthritis. A small number of CD146 + cells were identified within the inflammatory synovium. To elucidate the specific role these CD146⁺ cells play in neovascularization, bulk RNA-seq and scRNA-seq analyses were performed. CD146 + cells included various clusters of endothelial cells (c14) and type-2 pericytes (c6, c11, and c19). Although these cells express Cd146 and are involved in neovascularization based on interaction analysis, no Pdgfrb expression was observed, suggesting an intermediate, dedifferentiated state. Indeed, they exhibited gene expression patterns similar to transcriptomic profiles previously characterized in pulmonary vasculature during wound healing, including expression of Apln and Aplnr . Apelin-positive aCap are flattened endothelial cells that interact with alveolar type I epithelial cells, forming an ultrathin layer that facilitates gas exchange barrier in the lung while apelin receptor-positive gCap cells maintain vascular homeostasis, regulate permeability, and serve as progenitors for aCap during tissue repair [ 14 ]. Although typical lung aCap cells lack the expression of Vwf , Selp , and endothelin 1 ( Edn1 ), we observed that CD146 + endothelial cells in inflammatory synovium express Apln and/or Aplnr as well as Selp and Edn1 (Table 1 ). Furthermore, these cells also express Cldn5 and stem cell marker, Cd34 , indicating that cell lineage is directed toward neovascularization rather than gas-exchange [ 3 ]. Interestingly, we observed double-positive Apln -expressing gGap endothelial cells, which have previously been found to exhibit a stem cell-like phenotype in the lungs immediately after injury [ 14 ]. In line with this increase in neovascular signaling, we found that wound healing and hypoxia-related gene expression was detected, such as Apln , Edn1 , Cd34 , Vim , and Kdr [ 29 – 34 ]. Other hypoxia-related genes Angpt2 (c14), Pdgfb (c6, c11, c14, and c19), and Vegfa (c14) were also detected. It is possible that VEGF signaling might upregulate Dll4 and Notch4 expression (both in c14), which regulate extension and branching of capillary with Jagged1 ( Jag1 in c6, c11, c14, and c19) [ 35 ]. In addition, hypoxia disrupts the tight junction of endothelial cells via decreasing the protein levels of Cldn5, despite little effect on Cldn5 mRNA expression in c14 [ 36 ], thus affecting tight junctions between endothelial cells potentially contributing to edema and inflammation of the joint. In addition to endothelial cells, RNA-seq data revealed that CD146 + cells included pericytes. Our scRNA-seq data suggest that Pdgfrb -negative cells (c6, c11, and c19) represent type-2 pericytes, as these clusters were found to express type-2 pericyte-related genes such as Fgf5 , Nes , Tubb3 , Cspg4 , N-cadherin ( Cdh2 ) (Table 2 ). Also, Notch1 (c14, 6, 11, and 19) and Notch4 (c14) were detected, while Notch3 expression was not observed in any cluster following scRNA-seq analysis and only very low expression was noted using bulk RNA-seq (Table 3 )[ 37 ]. It is possible that this explains the lack of Pdgfrb expression since a lack of functional Notch3 is known to be associated with very low Pdgfrb expression and pericyte dysfunction [ 38 – 41 ]. Therefore, c6, 11, and 19 may fail to detect Pdgf signaling as they lack the appropriate receptor, which may affect their ability to associate with newly formed vessels leading to increased vascular permeability. This leakage into the synovial tissue likely contributes to pannus formation and joint swelling while promoting chronic inflammatory cell infiltration [ 42 ]. Other pericytes are classified as type-1 pericytes that express Pdgfrb ; however, most type-1 pericytes do not express Ednra or Ednrb except cluster 3 which was classified as immature type-1 pericytes (Table 3 ). Overall, this suggests that maintenance of vascular tension is insufficient and that the lack of functional, mature pericytes contributes to incomplete junction formation in new blood vessels and the development and persistence of joint edema [ 43 ]. It seems that endothelial cells and type-2 pericytes play synergistic roles in pannus neovascularization, where both cell types were found to be colocalized in the inflamed joint, as observed via histopathology. Further supporting these interactions, LIANA analysis revealed that CD146 + endothelial cells (c14) might associate directly via ligand/receptor interaction with c6, c11, and c19 within inflammatory synovium, especially under hypoxic conditions. Among the many binding partners between endothelial cells and type-2 pericytes listed in LIANA data of Fig. 6 and Table 3 , Robo4 expressed in CD146 + endothelial cell (c14) and its ligand, Slit3 , expressed in all type-2 pericytes are known to have similar effects as Ptn and Kdr interactions, inducing endothelial cell migration and tube formation during neovascularization [ 26 , 28 ]. Within endothelial cells, other interesting binding pairs found in our analysis are the juxtacrine interaction between Apln and Aplnr within endothelial cells and the interaction between Angpt2 , and Tek which is expressed exclusively in endothelial cells (c14). The latter molecular interaction has previously been shown to promote vascular sprouting along with VEGF (Table 1 ) [ 44 ]. Therefore, neovascularization processes orchestrated by both endothelial cells and pericytes represent an interesting potential therapeutic target in treating arthritis. Pericytes are generally considered a subset of mural cells; however, the present study focuses specifically on pericyte-like populations and does not encompass the full spectrum of mural cell subtypes such as vascular smooth muscle cells. Our characterization of CD146-positive cells in the synovium reveals new cell populations and suggests important interactions in the inflamed joint. However, there are some important caveats to the interpretation of these transcriptomic studies. Since only a small number of synovial cells could be collected from very small murine joint tissue, cells had to be cultured for three days prior to bulk RNA-seq and scRNA-seq analysis. During this period, the number of CD146 + cells increased slightly compared to direct FACS analysis of synovial cells. However, comparison of bulk RNA-seq and scRNA-seq data revealed that culture had little effect on gene expression profiles. Another limitation of our study is the reliance on RNA sequencing data, which does not fully reflect protein abundance and functional activity in the cell. To complement this, histological data analysis using immunostaining was performed. In some area of the pannus, typical neovascularization was observed, involving CD31 + endothelial cells and Ng2 + /Pdgfrb + pericytes (Fig. 6 H, yellow square). However, in another region, coexistence of Ng2 + /Pdgfrb − cells (likely type-2 pericytes) and CD31 + endothelial cells was also confirmed (Fig. 6 H, blue square). This suggests that Ng2 + /Pdgfrb − also colocalize with endothelial cells in the pannus. The formation of new blood vessels plays a key role in the development of joint arthritis and the persistence of inflammation in the synovium [ 45 ]. Based on our RNA sequencing data, hypoxic conditions in the joint cavity influence neovascularization through various signaling pathways primarily orchestrated by endothelial cells and pericytes. We found that CD146 + cell population expression patterns mirrored the vascular repair program normally activated in the lungs during injury and wound healing. However, it seems that neovascularization processes and interactions between endothelial cells and pericytes are insufficient, lacking certain markers required for proper junction formation. Thus, the commonalities between vascular regeneration and neovascularization within inflammatory pannus provide important insights for identifying anti-angiogenic therapeutic targets in RA and suggests alternative therapeutic targets aimed at modulating stromal-vascular interactions and reducing vascularization of arthritic tissue. Conclusions In summary, we provide a comprehensive characterization of CD146⁺ vascular cell populations in inflamed synovium using bulk and single-cell transcriptomic approaches. Our analysis identifies heterogeneous endothelial and pericyte states associated with neovascularization in inflammatory arthritis and suggests the presence of coordinated signaling interactions between these cell populations. While certain endothelial subsets share transcriptional features with capillary repair programs described in other tissues, their functional roles in synovial angiogenesis remain to be established. Importantly, our findings highlight the complexity of vascular remodeling in RA pannus and provide a resource for future studies aimed at dissecting endothelial-pericyte interactions in inflammatory settings. Given the descriptive nature of this study and limitations in cell isolation strategies, further functional and lineage-resolved analyses will be required to determine the precise contribution of these vascular cell subsets to disease progression and their potential as therapeutic targets. Abbreviations RA rheumatoid arthritis Cspg4 Chondroitin sulfate proteoglycan 4 NG2 neuron-glial antigen 2 SASP senescence-associated secretory phenotype Fgf5 fibroblast growth factor 5 Stc2 stanniocalcin 2 Sox2 SRY-box transcription factor 2 Sox10 SRY-box transcription factor 10 Mcam melanoma cell adhesion molecule (CD146) scRNA-seq single-cell RNA sequencing bulk RNA-seq bulk RNA sequencing Aplnr apelin receptor bColII bovine collagen type II FACS fluorescence-activated cell sorting Vim vimentin Angpt2 angiopoietin 2 Tie1 tyrosine kinase with immunoglobulin-like and EGF-like domains Tie2 TEK receptor tyrosine kinase Vegfα vascular endothelial growth factor A Kdr Kinase insert domain receptor gCap general capillary Ednra endothelin receptor type A Ednrb endothelin receptor type B Dll4 delta-like canonical Notch ligand 4 Jag1 Jagged 1 Robo4 roundabout guidance receptor 4 Ptprb protein tyrosine phosphatase receptor type B aCap aerocyte Edn1 endothelin 1 Cdh2 N-cadherin Declarations Ethics approval and consent to participate All mouse experiments were performed according to the rules and regulations of the Fundamental Guidelines for Proper Conduct of Animal Experiments and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education, Culture, Sports, Science, and Technology of Japan and were approved by the Committee on the Ethics of Animal Experiments of Nagoya City University. Consent for publication All authors have read and approved the final version of the manuscript and agree to its submission and publication. Availability of data and materials Data will be made available by the authors, upon reasonable request. Additionally, the series accession no. GSE311750 and GSE311751 for gene expression array is available. Competing interests None of the authors has any conflicts of interest, financial or otherwise, to disclose. Funding This work was funded by Grants-in Aid from the Ministry of Education, Culture, Sports, Science and Technology (MEXT)/ JSPS KAKENHI Grant Number 23K07879 and JSPS Short-term Fellowship PE25020. This work was supported by JSPS KAKENHI Grant Number JP22H04925 (PAGS). Authors’ Contributions Y.M., N.M., and S.K.(CA) conceived and the designed the research; Y.M., N.M, A.U., and S.K. (CA) performed experiments; Y.M., N.M, H.N., and S.K.(CA) analyzed the data; Y.M., N.M., S.K., H.N., and S.K.(CA) interpreted results of experiments; Y.M., N.M., S.K., and S.K(CA). prepared figures; Y.M., N.M., and S.K.(CA) drafted manuscript; Y.M., N.M., and S.K.(CA) edited and revised manuscript; Y.M., N.M., S.K., H.N. and S.K.(CA) approved the final version of manuscript. Acknowledgements We would like to thank Ayako Nitatoge, Satoko Ogawa, and Yoko Kanazawa for excellent technical support. References Liu M, Liu P, Li J, Huang Y, Wu R: Vascular synovial phenotype indicates poor response to JAK inhibitors in rheumatoid arthritis patients: a pilot study . PeerJ 2024, 12 :e18631. Lesturgie-Talarek M, Gonzalez V, Combier A, Thomas M, Boisson M, Poiroux L, Wanono S, Hecquet S, Carves S, Cauvet A et al : Inflammatory and angiogenic serum profile of refractory rheumatoid arthritis . Scientific reports 2025, 15 (1):7159. Godoy RS, Cober ND, Cook DP, McCourt E, Deng Y, Wang L, Schlosser K, Rowe K, Stewart DJ: Single-cell transcriptomic atlas of lung microvascular regeneration after targeted endothelial cell ablation . Elife 2023, 12 . Goss G, Rognoni E, Salameti V, Watt FM: Distinct Fibroblast Lineages Give Rise to NG2+ Pericyte Populations in Mouse Skin Development and Repair . Front Cell Dev Biol 2021, 9 :675080. Birbrair A, Zhang T, Wang ZM, Messi ML, Enikolopov GN, Mintz A, Delbono O: Role of pericytes in skeletal muscle regeneration and fat accumulation . Stem Cells Dev 2013, 22 (16):2298-2314. Schupp JC, Adams TS, Cosme C, Jr., Raredon MSB, Yuan Y, Omote N, Poli S, Chioccioli M, Rose KA, Manning EP et al : Integrated Single-Cell Atlas of Endothelial Cells of the Human Lung . Circulation 2021, 144 (4):286-302. Birbrair A, Zhang T, Files DC, Mannava S, Smith T, Wang ZM, Messi ML, Mintz A, Delbono O: Type-1 pericytes accumulate after tissue injury and produce collagen in an organ-dependent manner . Stem Cell Res Ther 2014, 5 (6):122. Chen J, Luo Y, Huang H, Wu S, Feng J, Zhang J, Yan X: CD146 is essential for PDGFRbeta-induced pericyte recruitment . Protein Cell 2018, 9 (8):743-747. Birbrair A, Zhang T, Wang ZM, Messi ML, Olson JD, Mintz A, Delbono O: Type-2 pericytes participate in normal and tumoral angiogenesis . Am J Physiol Cell Physiol 2014, 307 (1):C25-38. Loan A, Awaja N, Lui M, Syal C, Sun Y, Sarma SN, Chona R, Johnston WB, Cordova A, Saraf D et al : Single-cell profiling of brain pericyte heterogeneity following ischemic stroke unveils distinct pericyte subtype-targeted neural reprogramming potential and its underlying mechanisms . Theranostics 2024, 14 (16):6110-6137. Wang D, Wu F, Yuan H, Wang A, Kang GJ, Truong T, Chen L, McCallion AS, Gong X, Li S: Sox10(+) Cells Contribute to Vascular Development in Multiple Organs-Brief Report . Arteriosclerosis, thrombosis, and vascular biology 2017, 37 (9):1727-1731. Wang Z, Xu Q, Zhang N, Du X, Xu G, Yan X: CD146, from a melanoma cell adhesion molecule to a signaling receptor . Signal Transduct Target Ther 2020, 5 (1):148. Yan B, Lu Q, Gao T, Xiao K, Zong Q, Lv H, Lv G, Wang L, Liu C, Yang W et al : CD146 regulates the stemness and chemoresistance of hepatocellular carcinoma via JAG2-NOTCH signaling . Cell Death Dis 2025, 16 (1):150. Gillich A, Zhang F, Farmer CG, Travaglini KJ, Tan SY, Gu M, Zhou B, Feinstein JA, Krasnow MA, Metzger RJ: Capillary cell-type specialization in the alveolus . Nature 2020, 586 (7831):785-789. Miura Y, Isogai S, Maeda S, Kanazawa S: CTLA-4-Ig internalizes CD80 in fibroblast-like synoviocytes from chronic inflammatory arthritis mouse model . Scientific reports 2022, 12 (1):16363. Shichino S, Ueha S, Hashimoto S, Ogawa T, Aoki H, Wu B, Chen CY, Kitabatake M, Ouji-Sageshima N, Sawabata N et al : TAS-Seq is a robust and sensitive amplification method for bead-based scRNA-seq . Commun Biol 2022, 5 (1):602. Edalat SG, Gerber R, Houtman M, Luckgen J, Teixeira RL, Palacios Cisneros MDP, Pfanner T, Kuret T, Izanc N, Micheroli R et al : Molecular maps of synovial cells in inflammatory arthritis using an optimized synovial tissue dissociation protocol . iScience 2024, 27 (6):109707. Goveia J, Rohlenova K, Taverna F, Treps L, Conradi LC, Pircher A, Geldhof V, de Rooij L, Kalucka J, Sokol L et al : An Integrated Gene Expression Landscape Profiling Approach to Identify Lung Tumor Endothelial Cell Heterogeneity and Angiogenic Candidates . Cancer cell 2020, 37 (1):21-36 e13. Mabbott NA, Baillie JK, Brown H, Freeman TC, Hume DA: An expression atlas of human primary cells: inference of gene function from coexpression networks . BMC Genomics 2013, 14 :632. Aran D, Looney AP, Liu L, Wu E, Fong V, Hsu A, Chak S, Naikawadi RP, Wolters PJ, Abate AR et al : Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage . Nature immunology 2019, 20 (2):163-172. Crisan M, Yap S, Casteilla L, Chen CW, Corselli M, Park TS, Andriolo G, Sun B, Zheng B, Zhang L et al : A perivascular origin for mesenchymal stem cells in multiple human organs . Cell stem cell 2008, 3 (3):301-313. Zhang F, Jonsson AH, Nathan A, Millard N, Curtis M, Xiao Q, Gutierrez-Arcelus M, Apruzzese W, Watts GFM, Weisenfeld D et al : Deconstruction of rheumatoid arthritis synovium defines inflammatory subtypes . Nature 2023, 623 (7987):616-624. Miura Y, Ota S, Peterlin M, McDevitt G, Kanazawa S: A Subpopulation of Synovial Fibroblasts Leads to Osteochondrogenesis in a Mouse Model of Chronic Inflammatory Rheumatoid Arthritis . JBMR Plus 2019, 3 (6):e10132. Pedrosa AR, Trindade A, Fernandes AC, Carvalho C, Gigante J, Tavares AT, Dieguez-Hurtado R, Yagita H, Adams RH, Duarte A: Endothelial Jagged1 antagonizes Dll4 regulation of endothelial branching and promotes vascular maturation downstream of Dll4/Notch1 . Arteriosclerosis, thrombosis, and vascular biology 2015, 35 (5):1134-1146. Tefft JB, Bays JL, Lammers A, Kim S, Eyckmans J, Chen CS: Notch1 and Notch3 coordinate for pericyte-induced stabilization of vasculature . Am J Physiol Cell Physiol 2022, 322 (2):C185-C196. Zhang B, Dietrich UM, Geng JG, Bicknell R, Esko JD, Wang L: Repulsive axon guidance molecule Slit3 is a novel angiogenic factor . Blood 2009, 114 (19):4300-4309. Eltanahy AM, Koluib YA, Gonzales A: Pericytes: Intrinsic Transportation Engineers of the CNS Microcirculation . Front Physiol 2021, 12 :719701. Mellberg S, Dimberg A, Bahram F, Hayashi M, Rennel E, Ameur A, Westholm JO, Larsson E, Lindahl P, Cross MJ et al : Transcriptional profiling reveals a critical role for tyrosine phosphatase VE-PTP in regulation of VEGFR2 activity and endothelial cell morphogenesis . FASEB J 2009, 23 (5):1490-1502. Helker CS, Eberlein J, Wilhelm K, Sugino T, Malchow J, Schuermann A, Baumeister S, Kwon HB, Maischein HM, Potente M et al : Apelin signaling drives vascular endothelial cells toward a pro-angiogenic state . Elife 2020, 9 . Kidoya H, Naito H, Muramatsu F, Yamakawa D, Jia W, Ikawa M, Sonobe T, Tsuchimochi H, Shirai M, Adams RH et al : APJ Regulates Parallel Alignment of Arteries and Veins in the Skin . Dev Cell 2015, 33 (3):247-259. Eyries M, Siegfried G, Ciumas M, Montagne K, Agrapart M, Lebrin F, Soubrier F: Hypoxia-induced apelin expression regulates endothelial cell proliferation and regenerative angiogenesis . Circ Res 2008, 103 (4):432-440. Bartoszewski R, Moszynska A, Serocki M, Cabaj A, Polten A, Ochocka R, Dell'Italia L, Bartoszewska S, Kroliczewski J, Dabrowski M et al : Primary endothelial cell-specific regulation of hypoxia-inducible factor (HIF)-1 and HIF-2 and their target gene expression profiles during hypoxia . FASEB J 2019, 33 (7):7929-7941. Pugh CW, Ratcliffe PJ: Regulation of angiogenesis by hypoxia: role of the HIF system . Nature medicine 2003, 9 (6):677-684. Smith MH, Gao VR, Periyakoil PK, Kochen A, DiCarlo EF, Goodman SM, Norman TM, Donlin LT, Leslie CS, Rudensky AY: Drivers of heterogeneity in synovial fibroblasts in rheumatoid arthritis . Nature immunology 2023, 24 (7):1200-1210. Benedito R, Roca C, Sorensen I, Adams S, Gossler A, Fruttiger M, Adams RH: The notch ligands Dll4 and Jagged1 have opposing effects on angiogenesis . Cell 2009, 137 (6):1124-1135. Koto T, Takubo K, Ishida S, Shinoda H, Inoue M, Tsubota K, Okada Y, Ikeda E: Hypoxia disrupts the barrier function of neural blood vessels through changes in the expression of claudin-5 in endothelial cells . The American journal of pathology 2007, 170 (4):1389-1397. Wei K, Korsunsky I, Marshall JL, Gao A, Watts GFM, Major T, Croft AP, Watts J, Blazar PE, Lange JK et al : Notch signalling drives synovial fibroblast identity and arthritis pathology . Nature 2020, 582 (7811):259-264. Nadeem T, Bogue W, Bigit B, Cuervo H: Deficiency of Notch signaling in pericytes results in arteriovenous malformations . JCI Insight 2020, 5 (21). Kofler NM, Cuervo H, Uh MK, Murtomaki A, Kitajewski J: Combined deficiency of Notch1 and Notch3 causes pericyte dysfunction, models CADASIL, and results in arteriovenous malformations . Scientific reports 2015, 5 :16449. Domenga V, Fardoux P, Lacombe P, Monet M, Maciazek J, Krebs LT, Klonjkowski B, Berrou E, Mericskay M, Li Z et al : Notch3 is required for arterial identity and maturation of vascular smooth muscle cells . Genes Dev 2004, 18 (22):2730-2735. Mehes G, Tzankov A, Hebeda K, Anagnostopoulos I, Krenacs L, Bedekovics J: Platelet-derived growth factor receptor beta (PDGFRbeta) immunohistochemistry highlights activated bone marrow stroma and is potentially predictive for fibrosis progression in prefibrotic myeloproliferative neoplasia . Histopathology 2015, 67 (5):617-624. Lindahl P, Johansson BR, Leveen P, Betsholtz C: Pericyte loss and microaneurysm formation in PDGF-B-deficient mice . Science 1997, 277 (5323):242-245. Izquierdo E, Canete JD, Celis R, Santiago B, Usategui A, Sanmarti R, Del Rey MJ, Pablos JL: Immature blood vessels in rheumatoid synovium are selectively depleted in response to anti-TNF therapy . PloS one 2009, 4 (12):e8131. Lobov IB, Brooks PC, Lang RA: Angiopoietin-2 displays VEGF-dependent modulation of capillary structure and endothelial cell survival in vivo . Proceedings of the National Academy of Sciences of the United States of America 2002, 99 (17):11205-11210. Zhao F, Hu Z, Li G, Liu M, Huang Q, Ai K, Cai X: Angiogenesis in rheumatoid Arthritis: Pathological characterization, pathogenic mechanisms, and nano-targeted therapeutic strategies . Bioact Mater 2025, 50 :603-639. Additional Declarations No competing interests reported. Supplementary Files CD146PannusSuppFiguresART.docx CD146PannusSuppTableART.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 18 May, 2026 Reviewers agreed at journal 07 May, 2026 Reviewers agreed at journal 04 May, 2026 Reviewers invited by journal 04 May, 2026 Editor assigned by journal 03 May, 2026 Submission checks completed at journal 03 May, 2026 First submitted to journal 27 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9546936","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":636426180,"identity":"e6f1ee91-9197-4ab6-b5d5-a80116b93c22","order_by":0,"name":"Yoko Miura","email":"","orcid":"","institution":"Nagoya City University Graduate School of Medical Sciences","correspondingAuthor":false,"prefix":"","firstName":"Yoko","middleName":"","lastName":"Miura","suffix":""},{"id":636426182,"identity":"01e4da86-7a50-4937-90f4-a165cef77b1b","order_by":1,"name":"Nadia Milad","email":"","orcid":"","institution":"Nagoya City University Graduate School of Medical Sciences","correspondingAuthor":false,"prefix":"","firstName":"Nadia","middleName":"","lastName":"Milad","suffix":""},{"id":636426184,"identity":"d83d5f50-f4ff-4e82-9a16-d01e0e42d683","order_by":2,"name":"Atsushi Usami","email":"","orcid":"","institution":"Nagoya City University Graduate School of Medical Sciences","correspondingAuthor":false,"prefix":"","firstName":"Atsushi","middleName":"","lastName":"Usami","suffix":""},{"id":636426188,"identity":"ca95ece6-3781-4459-a71a-70ddc3a387fe","order_by":3,"name":"Shunichi Kosugi","email":"","orcid":"","institution":"Research Organization of Information and Systems","correspondingAuthor":false,"prefix":"","firstName":"Shunichi","middleName":"","lastName":"Kosugi","suffix":""},{"id":636426191,"identity":"47f45262-a386-4df2-8818-20b0ecf18cd6","order_by":4,"name":"Hideki Noguchi","email":"","orcid":"","institution":"Research Organization of Information and Systems","correspondingAuthor":false,"prefix":"","firstName":"Hideki","middleName":"","lastName":"Noguchi","suffix":""},{"id":636426192,"identity":"8e1165f1-7e80-4470-ac43-bb7fd88d69d8","order_by":5,"name":"Satoshi Kanazawa","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFElEQVRIiWNgGAWjYBACCTBZAcQH0GSY8Ws5Q7IWxjYsWnACyfazBz/+nLdNju8A78GPXyoOyzOwnzFg+FHDwG6OQ4s0T16yNO+228aSB/iSpWXOHDZs4MkxYOw5xsBs2YBdixxDjoE047bbiRsO8BhIS7YdZtx/g8eAgbeBgdkAh1Pl+N8Y//w553Y9UIvxb6AW+wYJHgPGv3i0SEvkmEnwNtxOMDjAYyb5se1wIkgLMz5bJGe8MbPmOXbbcOZhHjNrhjPpyQ08aQWHZY5J4PSLxPkc45s/am7L8x3vATIqrG0b2A9vfPimxiYZV4ghADDqmHkYmsFsoJMkkg0IagECxh8MdXCOHVFaRsEoGAWjYCQAAJ/8VyJovcuxAAAAAElFTkSuQmCC","orcid":"","institution":"Nagoya City University Graduate School of Medical Sciences","correspondingAuthor":true,"prefix":"","firstName":"Satoshi","middleName":"","lastName":"Kanazawa","suffix":""}],"badges":[],"createdAt":"2026-04-28 01:39:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9546936/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9546936/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109061547,"identity":"c9bea517-8f44-425a-a747-3b2fa8955feb","added_by":"auto","created_at":"2026-05-12 08:30:09","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1918135,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBlood vessels and CD146+ synovial cells in inflamed mouse joint tissue.\u003c/strong\u003e (A) Immunohistochemical staining of CD31 (green), PDGFRb (red), and vimentin (blue) in the inflamed and normal joints. Scale bars indicate 100 mm (white) and 20 mm (yellow). (B) Schematic diagram indicating the timeline for the induction of chronic inflammatory polyarthritis with bColII, followed by the culture of synovial cells obtained from inflamed joints. (C-D) Flow cytometric analysis of synovial cells from inflamed joints. Antibodies against CD11b, Pdpn, CD146, and CD140a were used. (E) The cell populations of CD146\u003csup\u003ehigh\u003c/sup\u003e, CD146\u003csup\u003emid\u003c/sup\u003e, CD140a\u003csup\u003e+\u003c/sup\u003e, and double-negative were calculated based on subpopulation delineated in Figure C. (F) Dot plots of CD146\u003csup\u003ehigh\u003c/sup\u003e, CD146\u003csup\u003emid\u003c/sup\u003e, CD140a, and negative subgroups for various pericyte and endothelial cell markers. (G-I) mRNA expression of \u003cem\u003eCspg4\u003c/em\u003e, \u003cem\u003eRunx2\u003c/em\u003e, and \u003cem\u003eSox9\u003c/em\u003e in whole pannus or subgroups from FACS (CD146\u003csup\u003ehigh\u003c/sup\u003e, CD146\u003csup\u003emid\u003c/sup\u003e, CD140a\u003csup\u003e+\u003c/sup\u003e, and double-negative) as measured via qPCR, expressed as fold change from housekeeping gene \u003cem\u003eHprt\u003c/em\u003e. Data are presented as the means±SE of four mice or six to seven mice (F-G), one-way ANOVA with Tukey (D) or Dunnet (F-G) * p\u0026lt;0.05 ** p\u0026lt;0.01 *** p\u0026lt;0.001 **** p\u0026lt;0.0001\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/472543c9a72127b09f2bc23b.png"},{"id":109061553,"identity":"9c228908-46be-4e13-af4d-f54d6cadb74a","added_by":"auto","created_at":"2026-05-12 08:30:13","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1324194,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCluster analysis of synovial cell scRNA-seq data.\u003c/strong\u003e(A-C) Flt-SNE and subgroup cluster analysis for scRNA-seq from joint cells. (D-H) Expression of \u003cem\u003ePdgfra\u003c/em\u003e, \u003cem\u003ePdgfrb\u003c/em\u003e, \u003cem\u003eMcam\u003c/em\u003e (CD146), \u003cem\u003ePecam1\u003c/em\u003e (CD31), and \u003cem\u003eEfnb2\u003c/em\u003e. (I-J) Trajectory analysis of cell subgroups based on clustering and scRNA-seq expression patterns. (K) Slingshot pseudotime analysis for each cluster.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/da9fd7a0a3b86ae9a269531a.png"},{"id":109061546,"identity":"04caace8-1912-407a-b51e-08fef0b330f9","added_by":"auto","created_at":"2026-05-12 08:30:09","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2668143,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eComparison of endothelial cell populations via scRNA-seq in mouse arthritis, human lung cancer and human arthritis.\u003c/strong\u003e(A-C) Flt-SNE plots and subgroup cluster analyses(umap data) for scRNA-seq data from mouse polyarthritis, human rheumatoid arthritis (RA), and lung cancer. (D-E) Expression comparisons between mouse polyarthritis and either human RA or lung cancer gene expression profiles. (F-G) Expression of \u003cem\u003eApln\u003c/em\u003e and \u003cem\u003eAplnr \u003c/em\u003ein mouse lung, pannus, kidney, and liver tissues as measured via qPCR, expressed as fold change from housekeeping gene \u003cem\u003eHprt\u003c/em\u003e. Data are presented as the means±SE of five to six mice, one-way ANOVA with Dunnet **** p\u0026lt;0.0001. (H) Immunohistochemical staining of CD31 (green), Aplnr (red), and Mcam or CD146 (blue) in the inflamed joint, scale bars indicate 100 mm (white) and 20 mm (yellow).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/2c467a0eb1839a3017fcda79.png"},{"id":109067772,"identity":"7b2c60c7-f93b-4f14-941e-7a034c78e2f2","added_by":"auto","created_at":"2026-05-12 10:00:42","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":642059,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eApelin and apelin receptor expression in mouse arthritis, human lung cancer and human arthritis.\u003c/strong\u003e (A-C) t-SNE plots for expression of \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e (blue), \u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e (red), \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e (green), and double-negative cells (gray) in the mouse polyarthritis, human rheumatoid arthritis (RA), and lung cancer, black outlines in each t-SNE plots indicate endothelial cell clusters. (D) Pie charts indicating the percentage of endothelial cells in each group: \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e, \u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e, \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e, and double-negative cells. (E-G) Heatmaps of the percentage of cells expressed \u003cem\u003eProcr\u003c/em\u003e, \u003cem\u003eKdr\u003c/em\u003e, \u003cem\u003eCldn5\u003c/em\u003e, \u003cem\u003eCd34\u003c/em\u003e, \u003cem\u003eAngpt2\u003c/em\u003e, and 5-genes of each endothelial cell subgroup in mouse polyarthritis, Human RA, and lung cancer.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/50192e4d730e76e0e9b250b7.png"},{"id":109061545,"identity":"7f792a7d-ff2c-4695-ae7b-eaef7ce444b2","added_by":"auto","created_at":"2026-05-12 08:30:07","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1237489,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCharacterization of pericyte populations in murine pannus tissue.\u003c/strong\u003e (A-B) Flow cytometric analysis of synovial cells from inflamed joints. (C-G) Pericyte and stem cell marker expression \u003cem\u003eFgf5\u003c/em\u003e, \u003cem\u003eCspg4\u003c/em\u003e, \u003cem\u003eStc2\u003c/em\u003e, \u003cem\u003eSox2\u003c/em\u003e, and \u003cem\u003eSox10\u003c/em\u003e. (H) Immunohistochemical staining of CD31 (green), PDGFRb(red), and NG2 (Blue) in the inflamed and normal joints. Scale bars indicate 100 mm (white) and 20 mm (yellow). Data are presented as the means±SE of three mice, one-way ANOVA with Tukey * p\u0026lt;0.05 ** p\u0026lt;0.01 *** p\u0026lt;0.001 **** p\u0026lt;0.0001\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/aa301b09490bc8d445d46cdf.png"},{"id":109061548,"identity":"41ceca00-6b52-47a7-bbcc-f6ed00da7ab0","added_by":"auto","created_at":"2026-05-12 08:30:09","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1307688,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eInteraction analysis between synovial cell clusters.\u003c/strong\u003e Heat map (A and B) and LIANA data (C to F) derived from scRNA-seq data. Vertical and Horizontal bars indicate cluster numbers (A) and the association via ligands and its receptors (B). LIANA analysis of type-2 pericytes [c6 (C), c11 (D), and c19 (E)] and endothelial cells [c14 (F)]. Each row shows the expression of each ligand within the cluster, with each ligand’s receptor represented by a vertical bar. Each cluster number is represented by a horizontal bar.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/407831cd93b86f215fd0d1c0.png"},{"id":109203840,"identity":"7b42f9c5-9617-4df0-a644-98feb009468a","added_by":"auto","created_at":"2026-05-13 14:48:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10035431,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/36a1cdd6-d647-4dc6-a3a9-cb5f2f04d418.pdf"},{"id":109061544,"identity":"2bdb6848-a1c3-4d83-b746-5b3192888e1b","added_by":"auto","created_at":"2026-05-12 08:30:07","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2928108,"visible":true,"origin":"","legend":"","description":"","filename":"CD146PannusSuppFiguresART.docx","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/5efce4d3dae3ebf169fb5bae.docx"},{"id":109061557,"identity":"85d9d3a1-f71c-474f-8fab-032252c0fb7f","added_by":"auto","created_at":"2026-05-12 08:30:18","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":22366,"visible":true,"origin":"","legend":"","description":"","filename":"CD146PannusSuppTableART.docx","url":"https://assets-eu.researchsquare.com/files/rs-9546936/v1/aefd0e868251e152c8e13323.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eNeovascularization in rheumatoid arthritis pannus is orchestrated by endothelial-pericyte signaling\u003c/p\u003e","fulltext":[{"header":"Background","content":"\u003cp\u003eRheumatoid arthritis (RA) is a chronic inflammatory disease characterized by pannus formation in the joint due to hyperplasia of the synovial membrane accompanied by inflammation and bone erosion. Neovascularization is observed in the pannus early in disease and is known to influence disease progression [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. This angiogenesis is thought to contribute to pathogenesis by facilitating inflammatory cell infiltration as well as increasing supply of oxygen and nutrients to pathologically activated remodeling cells. In fact, some research suggests that reducing neovascularization could be a therapeutic strategy in RA [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Endothelial cells and pericytes are the key cellular components of newly formed blood vessels found in arthritic joints as well as in tumor neovascularization, where endothelial cell populations are highly heterogeneous, encompassing distinct vascular-type and organ-specific characteristics.\u003c/p\u003e \u003cp\u003eWhen neovascularization occurs during wound healing, endothelial cells form capillaries together with pericytes [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Generally, pericytes are composed of heterogeneous cells and possess the ability to undergo flexible transformation. Chondroitin sulfate proteoglycan 4 (\u003cem\u003eCspg4\u003c/em\u003e, encodes neuron-glial antigen 2, NG2) is the most widely recognized pericyte marker; on the other hand, \u003cem\u003ePdgfrb\u003c/em\u003e-negative pericytes were identified during skin development [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. As a result, they differentiate into fibroblast-like or myofibroblast-like cells in response to various stimuli. Recent studies have revealed two distinct pericyte states as functional differentiation stages: type-1 pericytes and type-2 pericytes. Type-1 pericytes share gene expression patterns with myofibroblast, producing senescence-associated secretory phenotype (SASP) markers and expressing \u003cem\u003eActa2\u003c/em\u003e, collagens such as \u003cem\u003eCol1a1\u003c/em\u003e, \u003cem\u003eTgfb1\u003c/em\u003e (TGF-β), and \u003cem\u003ePdgfrb\u003c/em\u003e, similar to perivascular fibroblasts [\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. These cells are particularly important during stages of fibrosis following tissue damage. Meanwhile, type-2 pericytes play a regenerative role, expressing markers such as CD146, NG2 (\u003cem\u003eCspg4\u003c/em\u003e), and fibroblast growth factor 5 (\u003cem\u003eFgf5\u003c/em\u003e), while \u003cem\u003ePdgfrb\u003c/em\u003e expression is transiently reduced [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Also, other genes such as stanniocalcin 2 (\u003cem\u003eStc2\u003c/em\u003e) and SRY-box transcription factors 2 and 10 (\u003cem\u003eSox2\u003c/em\u003e and \u003cem\u003eSox10\u003c/em\u003e) might be involved in the regenerative phase of pericytes during wound healing [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. The contribution of pericytes and other cell types, such as fibroblasts and immune cells, in the formation of new vessels within arthritic joints have yet to be fully elucidated.\u003c/p\u003e \u003cp\u003eIn this study, synovial cells were isolated from the inflamed joints of a RA mouse model and analyzed using flow cytometry. We used CD146, or melanoma cell adhesion molecule (MCAM), surface expression to stratify cells since it is predominantly expressed by endothelial cells but also found in pericytes, smooth muscle cells, and some immune cells [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] and its expression has been linked to vascular regeneration, metastasis, and tumor progression in various types of cancer [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Next, we examined the specific roles of CD146\u003csup\u003e+\u003c/sup\u003e cells through bulk RNA sequencing and single-cell RNA sequencing (scRNA-seq). We found that these CD146\u003csup\u003e+\u003c/sup\u003e endothelial cells were involved in neovascularization and could be classified into subgroups based on \u003cem\u003eApln\u003c/em\u003e and apelin receptor (\u003cem\u003eAplnr\u003c/em\u003e) expression, vascular markers typically expressed in the lungs [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], where an intermediate population of \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e cells was observed, likely a transitory phenotype. Interaction analysis revealed that these cell populations participate in neovascularization reminiscent of vascular repair processes observed in the lung. These gene expression data were compared with previously reported gene ontology analyses of human joints, human lung cancer and mouse joint tissues and found to correlate well with neovascularization patterns in these studies. Overall, we conclude that \u003cem\u003eApln\u003c/em\u003e-positive endothelial cells and type-2 pericytes work together to promote neovascularization in the inflamed joint thus promoting pannus formation and remodeling in the context of polyarthritis. Interestingly, neovascularization in the pannus shares common features with pulmonary capillary repair, heavily relying on pericyte signaling, which may be useful as therapeutic targets in the development of treatments for arthritis and perhaps associated lung cancer.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eChronic polyarthritis mouse model\u003c/h2\u003e \u003cp\u003eA low dose of bovine collagen type II (bColII) was administered to induce chronic polyarthritis in D1BC mice as previously described [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Briefly, 7- to 8-week-old D1BC mice were housed in a pathogen-free animal care facility at Nagoya City University Medical School in accordance with institutional guidelines (animal ethics approval #21-033H02). Mice were anesthetized with isoflurane and administered subcutaneously at all 4 limbs (total 10 mg per mouse) bColII emulsified with an equal volume of complete Freund\u0026rsquo;s adjuvant (BD Biosciences, San Jose, CA). The same protocol was used for secondary immunization with bColII after 3 weeks from 1st immunization, except without adjuvant.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eImmunohistochemistry\u003c/h3\u003e\n\u003cp\u003eNormal and inflamed joints were fixed in 4% paraformaldehyde/PBS then decalcified in buffer containing 2.5% EDTA and 7% sucrose at 4℃ for one month. Paraffin sections (2 mm thickness) were deparaffinized, rehydrated, then underwent antigen retrieval in citrate buffer at 121\u0026ordm;C for 5 minutes. Following quenching (3% H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e in methanol for 15 min) and blocking (10% goat serum in TBS-T for 20 min), slides were stained using the following primary antibodies diluted in 10% goat serum in TBS-T: rabbit anti-CD31 (1/1,500; ab28364; Abcam, Cambridge, UK), rabbit anti-MCAM/CD146 (1/500; #81701; Cell Signaling Technology, Danvers, MA, USA), rabbit anti-PDGFRβ (1/100; #3169; Cell Signaling Technology), rabbit anti-APLNR (1/500; bs-21310R; Bioss, Boston, MA, USA), anti-NG2 (1/100; AB5320; Merck Millipore, Burlington, MA, USA), and rabbit anti-vimentin (1/1,000; #5741; Cell Signaling Technology). Slides were then incubated with Histofine Simple Stain mouse anti-rabbit MAX-PO secondary antibodies (Nichirei Biosciences, Tokyo, Japan) and then Opal multiplex fluorescent immunohistochemistry system (Akoya Biosciences, Marlborough, MA, USA) were used according to the manufacturer\u0026rsquo;s instructions. All images were captured using a fluorescence microscope (BZ-X710; Keyence, Osaka, Japan).\u003c/p\u003e\n\u003ch3\u003ePannus collection and bulk RNA sequencing (RNA-seq)\u003c/h3\u003e\n\u003cp\u003eSynovial cell isolation and flow cytometry was performed as previously described [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Briefly, approximately 1 x 10\u003csup\u003e5\u003c/sup\u003e cells of isolated synovial cells were blocked FcgR with anti-CD16/ CD32 antibody (BD Biosciences). Within the CD11b-negative fraction, cells were then sorted by fluorescence activated cell sorting (FACS) on a FACS Aria II (BD Biosciences) as follows: CD146\u003csup\u003ehi\u003c/sup\u003e, CD146\u003csup\u003emid\u003c/sup\u003e, CD140a\u003csup\u003e+\u003c/sup\u003e, and double-negative. Indexed libraries were then sequenced using the NovaSeq platform (Illumina, San Diego, CA, USA) by Macrogen Inc (Seoul, South Korea) and can be found online via GEO accession number GSE311750.\u003c/p\u003e\n\u003ch3\u003eSingle-cell RNA sequencing (scRNA-seq)\u003c/h3\u003e\n\u003cp\u003eSynovial cells obtained from 7 mice were pooled and FcgRs were blocked by anti-CD16/32 antibody (BD Biosciences) at 4℃ for 10 minutes. After washing by FACS buffer, synovial cells were collected positive selection using CD146-beads and MS column (Miltenyi Biotec). The scRNA-seq analysis was performed by ImmunoGeneTeqs Inc (Chiba, Japan) according to the TAS-seq protocol [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Raw sequencing data were deposited to the National Center for Biotechnology Information Gene Expression Omnibus and are accessible through Gene Expression Omnibus (series accession number GSE311751). Human scRNA-seq data for RA [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and lung cancer [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e] were obtained from the ArrayExpress site (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ebi.ac.uk/biostudies/arrayexpress\u003c/span\u003e\u003cspan address=\"https://www.ebi.ac.uk/biostudies/arrayexpress\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) at the European Bioinformatics Institute (EBI). Details of the bioinformatics protocol and other relevant details of the \u003cem\u003ein silico\u003c/em\u003e analysis are described in the Supplementary Methods.\u003c/p\u003e\n\u003ch3\u003ePannus collection and bulk RNA sequencing (RNA-seq) analysis\u003c/h3\u003e\n\u003cp\u003eAfter cells were collected from the pannus and washed with PBS containing 2% FBS, synovial cells were stained using the following primary antibodies: CD11b-PE (clone M1/70, BD Biosciences), podoplanin-BV421 (clone 8.1.1, BioLegend, San Diego, CA, USA), PDGFRα (CD140a)-APCvio770 (clone REA637, Miltenyi Biotec, NRW, Germany), CD146 (LSEC)-APC (clone ME-9F1, Miltenyi Biotec), PDGFRβ (CD140b)-PEvio770 (clone REA634, Miltenyi Biotec), and 7-aminoactinomycin D (7-AAD, BD Biosciences). These cells were analyzed using a FACSCanto II (BD Biosciences).\u003c/p\u003e \u003cp\u003eTotal RNA extracted from the collecting synovial cells was used in the ReliaPrep RNA Tissue Miniprep System (Promega, Madison, WI, USA) and total RNA concentration was calculated by Quant-IT RiboGreen (#R11490, Invitrogen, Waltham, MA, USA). To assess RNA integrity, samples were run on the TapeStation RNA Screentape (Agilent, Santa Clara, CA, USA). Only high-quality RNA preparations, with RIN greater than 7.0, were used for RNA library construction. The RNA isolated from each sample was used to construct sequencing libraries with the SMART-Seq\u0026trade; mRNA kit (Takara Bio, Kusatsu, Japan), following the manufacturer's protocol. First-strand cDNA synthesis (from total RNA or cells) is primed by the 3\u0026rsquo; SMART-Seq CDS Primer II A and uses the SMART-Seq v4 Oligonucleotide for template switching at the 5\u0026rsquo; end of the transcript. Sequencing was performed using the Illumina NovaSeq 6000 (San Diego, CA). The first-strand cDNA selectively binds to SPRI beads leaving contaminants in solution which is removed by a magnetic separation. The beads are then directly used for PCR amplification. The Advantage 2 Polymerase Mix has been specially formulated for efficient and accurate amplification of cDNA templates by long-distance PCR. PCR-amplified cDNA is purified by immobilization on AMPure XP beads. The beads are then washed with 80% ethanol and cDNA is eluted with Elution Buffer. Prior to generating the final library for Illumina sequencing, the Covaris AFA system is used for controlled DNA shearing. The resulting DNA will be in the 200\u0026ndash;400 bp size range. These cDNA fragments then go through an end repair process, the addition of a single \u0026lsquo;A\u0026rsquo; base, and then ligation of the indexing adapters. The products are then purified and enriched with PCR to create the final cDNA library. The libraries were quantified using qPCR according to the qPCR Quantification Protocol Guide (KAPA Library Quantification kits for Illumina Sequencing platforms) and qualified using the Agilent Technologies 4200 TapeStation (Agilent Technologies, Waldbronn, Germany). Indexed libraries were then sequenced using the NovaSeq platform (Illumina, San Diego, CA, USA) by Macrogen Inc (Seoul, South Korea) and can be found online via GEO accession number GSE311750.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eReverse transcription quantitative PCR (RT-qPCR)\u003c/h2\u003e \u003cp\u003eTotal RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany) for pannus, lung, kidney, and liver tissues according to the manufacturer\u0026rsquo;s instructions. For quantitative PCR (qPCR), cDNA was synthesized using ReverTra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO, Osaka, Japan). The qPCR was performed using the PrimeTime Gene Expression Master Mix (Integrated DNA Technologies, Coralville, IA, USA). The relative expression of each gene was determined by an internal control using \u003cem\u003eHprt\u003c/em\u003e for each sample and expressed as a fold change of the control group.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSingle-cell RNA sequencing (scRNA-seq) analysis\u003c/h3\u003e\n\u003cp\u003eBriefly, hashtag-stained single nuclei suspension was pooled and approximately 20,000 cells were loaded onto the BD Rhapsody system (BD Biosciences). After loading unique cell barcode-immobilized magnetic beads, cells were lysed. Next, reverse transcription of polyA RNA and hashtag oligos derived from each single cell was performed according to the manufacturer\u0026rsquo;s instructions. Exonuclease I treatment and cDNA/hashtag amplification were performed as per the TAS-seq protocol. Sequencing libraries were prepared using the NEBNext Ultra II FS Library Prep Kit for Illumina (New England Biolabs, Ipswich, MA, USA) and quantified using the KAPA Library Quantification Kit (KAPA Biosystems, Wilmington, MA, USA) and a MultiNA system (Shimazu, Kyoto, Japan). Sequencing analysis was performed by a NovaSeq 6000 sequencer and a Novaseq S4 200 cycles v1.5 kit (Illumina) [Read1: 67 base-pair, Read2: 155 base-pair]. The resulting pair-end fastq files (R1: cell barcode reads, R2: RNA reads) of the TAS-Seq data were processed according to a previously reported method. Ensemble reference RNA (GRCh38 release-101) was used for mapping RNA reads and the known barcode sequence of each hashtag was used for mapping hashtag reads. Demultiplexing of each cell by hashtag read count was performed according to the method described previously.\u003c/p\u003e\n\u003ch3\u003eBioinformatics analysis of scRNA-seq data\u003c/h3\u003e\n\u003cp\u003eHuman scRNA-seq data for RA [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and lung cancer [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e] were obtained from the ArrayExpress site (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ebi.ac.uk/biostudies/arrayexpress\u003c/span\u003e\u003cspan address=\"https://www.ebi.ac.uk/biostudies/arrayexpress\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) at the European Bioinformatics Institute (EBI). The gene expression matrices were imported into Seurat (version 5.2.0), and quality control (QC) was performed to remove low-quality cells [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Cells meeting the following criteria were retained: a minimum of \u003cem\u003e200\u003c/em\u003e and a maximum of \u003cem\u003e6000\u003c/em\u003e genes per cell, a minimum of 500 and a maximum of 40,000 reads per cell to exclude empty droplets and potential doublets. To exclude potentially dying cells, cells showing mitochondrial gene content below \u003cem\u003e20%\u003c/em\u003e in the RA data or 5% in the lung cancer data were removed. The gene expression data were normalized using the SCTransform function of the sctransform package, specifying percent.mt for vars.to.regress and 3 for min_cells. Principal component analysis (PCA) was performed for dimensionality reduction, and cells were clustered using the FindNeighborsand function with the dims parameter set to 1:25 and the FindClusters function with the resolution parameter set to 1.5. The clusters were visualized using Uniform Manifold Approximation and Projection (UMAP) using the RunUMAP function setting the dims parameter to 1:25. Cell type annotation was performed using the celldex, SingleR, and SingleCellExperiment packages. The celldex human reference dataset was obtained using the R command (celldex, HumanPrimaryCellAtlasData) [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Each cell was annotated with the corresponding cell type in the celldex reference using the SingleR and as.SingleCellExperiment functions, and these cells were labelled as \u0026lsquo;singlr_labels\u0026rsquo;.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eVisualization of gene expression cells on UMAP and t-SNE\u003c/h2\u003e \u003cp\u003eWe visualized cells expressing a specific gene on dimension reduction images (FIt-SNE for the mouse polyarthritis data or UMAP for the other data) using the Seurat FeaturePlot function with the option \u0026lsquo;max.cutoff=q90\u0026rsquo; to limit the maximum color density to the 90% percentile of the expression value. To visualize cells co-expressing multiple genes on the dimension reduction images, we used the DimPlot function of Seurat. Specifically, we specified the cell list objects corresponding to each gene as the cells.highlight argument, and defined the color and size of the highlighted cells using the cols.highlight and sizes.highlight arguments.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eBiological theme comparison between cell clusters\u003c/h2\u003e \u003cp\u003eDifferentially expressed genes (DEGs) in each cell cluster of the mouse polyarthritis data were identified using the Seurat FindAllMarkers function under the following conditions: the Wilcoxon rank-sum test, logfc.threshold\u0026thinsp;=\u0026thinsp;0.25, and min.pct\u0026thinsp;=\u0026thinsp;0.25). Gene ontology (GO) enrichment analysis was performed for the DEGs in each cell cluster using the R clusterProfiler package and the org.Mm.eg.db mouse GO database. The gene symbols of DEGs were converted to ENTREZ IDs, and enrichment analysis was performed using the compareCluster function with the options fun and organism set to \u0026lsquo;enrichWP\u0026rsquo; and \u0026lsquo;Mus musculus\u0026rsquo;, respectively. (The output data was restricted to GO level 4 using the gofilter function. The output dot plots were created using the dotplot function and customized with the ggplot2 R package.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of Apln\u003csup\u003e+\u003c/sup\u003e and Aplnr\u003csup\u003e+\u003c/sup\u003e cell populations\u003c/h2\u003e \u003cp\u003eWhen counting the number of cells expressing a specific single or multiple genes (e.g., \u003cem\u003eApln\u003c/em\u003e and \u003cem\u003eAplnr\u003c/em\u003e), cells with ≧\u0026thinsp;0.1 expression values of the corresponding genes were targeted in human RA data, while cells with ≧\u0026thinsp;1.0 values were targeted in other data. The data were visualized by pie charts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eLIANA analysis\u003c/h2\u003e \u003cp\u003eCell-to-cell communication (CCC) between cell clusters in the mouse polyarthritis data was analyzed using the R package LIANA for ligand-receptor analysis framework, which calculates a consensus CCC rank using multiple resources and methods related to CCC. The Mouse Consensus was selected as resource using the select_resource function, and the mouse polyarthritis data object created by Seurat was processed with the liana_wrap and liana_aggregate functions. Heatmaps of ranked interactions for each cluster were generated using the liana_dotplot function. To generate a frequency plot of predicted interactions, predicted ligand-receptor pairs filtered for statistical significance with p-values\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were plotted using the heat_freq function.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eData were visualized and statistically analyzed using GraphPad Prism 10.5.0 (Boston, MA, USA). Histogram data is expressed as the mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard error (SE). Data was not transformed and outlier exclusion was not required. One-way ANOVA tests followed by Tukey\u0026rsquo;s post-hoc correction for multiple comparisons and Dunnet\u0026rsquo;s test for parametric data were used when several independent groups were compared. All statistical analyses used two-tailed tests with a p-value of less than 0.05 considered significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eNeovascularization in the inflammatory synovium of mice with chronic polyarthritis\u003c/h2\u003e \u003cp\u003eD1BC mice develop chronic and severe joint polyarthritis following low-dose administration of bovine type II collagen, exhibiting characteristics closer to RA compared to conventional collagen-induced polyarthritis mouse model [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. First, we used immunohistochemistry to visualize blood vessels, pericytes, and fibroblasts within the pannus of inflamed joints, where CD31\u003csup\u003e+\u003c/sup\u003e blood vessels were observed, typically surrounded by Pdgfrβ\u003csup\u003e+\u003c/sup\u003e pericytes and distinct from vimentin\u003csup\u003e+\u003c/sup\u003e (Vim) fibroblasts (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). In contrast, normal joints isolated from na\u0026iuml;ve mice lack obvious vascularization, exhibiting minimal staining in the joints for any of these markers.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eExpression profile of blood vessel-associated cells in the inflammatory synovium\u003c/h2\u003e \u003cp\u003eTo further investigate neovascularization in the inflamed synovium, we examined various subtypes of vascular cells stratified by their expression of CD146 and CD140a (Pdgfrα), markers known to be upregulated in angiogenesis and constitutively expressed in perivascular cells such as pericytes [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Cells from the inflammatory synovium were cultured for less than three days (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB-C). FACS analysis revealed that CD11b-negative cells were divided into four distinct groups: CD146\u003csup\u003ehigh\u003c/sup\u003e, CD146\u003csup\u003emid\u003c/sup\u003e, CD146\u003csup\u003elow\u003c/sup\u003eCD140a\u003csup\u003ehigh\u003c/sup\u003e, and double-negative cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC-D). The majority of synovial cells were identified as CD11b-positive synovial macrophages (40%, Type A) and CD140a\u003csup\u003e+\u003c/sup\u003e fibroblasts (50%, Type B) (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC-E). Within the CD11b-negative population, CD146\u003csup\u003ehigh\u003c/sup\u003e and CD146\u003csup\u003emid\u003c/sup\u003e cells represented 3.3 and 3.6% of total cells, respectively, indicating that vascular cells can be found in isolated synovial cells though in relatively small abundance.\u003c/p\u003e \u003cp\u003eNext, bulk RNA-seq was used to characterize the transcriptomic profiles of CD146\u003csup\u003ehigh\u003c/sup\u003e, CD146\u003csup\u003emid\u003c/sup\u003e, CD140a\u003csup\u003e+\u003c/sup\u003e, and double-negative populations. The CD146\u003csup\u003ehigh\u003c/sup\u003e populations had the highest number of differentially expressed genes compared to the double-negative population. Upregulated genes in the CD146\u003csup\u003ehigh\u003c/sup\u003e populations included several endothelial cell markers, such as \u003cem\u003eIcam2\u003c/em\u003e, \u003cem\u003eCd151\u003c/em\u003e, \u003cem\u003eApln\u003c/em\u003e, and \u003cem\u003eEdnrb\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. On the other hand, CD146\u003csup\u003emid\u003c/sup\u003e and CD140a\u003csup\u003e+\u003c/sup\u003e cells showed similar expression patterns, primarily expressing pericyte markers \u003cem\u003ePdgfrb\u003c/em\u003e, \u003cem\u003eAng2\u003c/em\u003e, and \u003cem\u003eActa2\u003c/em\u003e (type-1 pericytes); osteo-chondrogenic markers \u003cem\u003eRunx2\u003c/em\u003e and \u003cem\u003eSox9\u003c/em\u003e; and fibroblast marker, \u003cem\u003ePdgfra\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF). Conversely, the double-negative population expressed few endothelial cell- or pericyte-associated genes; instead, they were found to express osteoclast and macrophage markers. Using RT-qPCR, we found that CD146\u003csup\u003ehigh\u003c/sup\u003e cells expressed high levels of \u003cem\u003eCspg4\u003c/em\u003e, while CD140a\u003csup\u003e+\u003c/sup\u003e cells showed high expression of \u003cem\u003eRunx2\u003c/em\u003e and \u003cem\u003eSox9\u003c/em\u003e, a chondrocyte marker (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eG-I) [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. These results suggest that synovial cells isolated from pannus of inflammatory synovium, though predominantly composed of CD11b\u003csup\u003e+\u003c/sup\u003e cells and some osteochondrogenic cells [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], also include CD146\u003csup\u003e+\u003c/sup\u003e endothelial cells, type-1 pericytes (\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e), and type-2 pericytes (\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eSingle-cell RNA-seq analysis of mouse synovial cells\u003c/h2\u003e \u003cp\u003eBulk RNA-seq analysis of synovial cells revealed newly formed vessels contained type-1 and type-2 pericytes in CD146\u003csup\u003ehigh\u003c/sup\u003e, CD146\u003csup\u003emid\u003c/sup\u003e, and CD140a\u003csup\u003e+\u003c/sup\u003e populations. To further characterize these cell types and assess their respective roles neovascularization, we performed single-cell RNA-seq analysis on synovial cells enriched using anti-CD146 antibody-conjugated ferrite beads. Cluster analysis revealed that CD146\u003csup\u003e+\u003c/sup\u003e cell population is still small; however, it consists of distinct cell populations such as endothelial cells (cluster 14 or c14), and type-2 pericytes (c6, c11, and c19) (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA-B, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, and Supplementary Fig.\u0026nbsp;1, cluster annotation). We also observed fibroblast clusters (c3, c4, c5, c6, c8, c9, c11, c15, c16, c18, and c21), including type-1 pericytes, and macrophages (c0, c1, c2, c7, c10, c13, c19, and c20). Macrophage and pericyte subpopulations clustered with high intracluster correlation, whereas the CD146\u003csup\u003e+\u003c/sup\u003e subpopulation was more distinct from these cells. Interestingly, endothelial cells (c14) showed no strong correlation with any other cluster. Combined with enrichment analysis, various cell populations were annotated (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC-H and Supplementary Fig.\u0026nbsp;1). We found that CD146\u003csup\u003e+\u003c/sup\u003e cells do not express either \u003cem\u003ePdgfra\u003c/em\u003e nor \u003cem\u003ePdgfrb\u003c/em\u003e. Further trajectory analysis using Slingshot algorithm revealed that endothelial cells (c14) and one group of type-2 pericytes (c11) were terminal clusters and express matrix metalloproteinase, \u003cem\u003eMmp2\u003c/em\u003e (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Conversely, the remaining type-2 pericytes (c6 and c19) were predicted to be immature (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eI-K). Interestingly, c3 is distinct from the major population of fibroblasts likely due to a high prediction of immaturity; however, this population seems related to pericytes as it expresses \u003cem\u003ePdgfrb\u003c/em\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eEndothelial cell markers detected in CD146\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e cells.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eEndothelial cell markers\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c6\" namest=\"c2\"\u003e \u003cp\u003eCD146\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e cells\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eInteraction\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCluster\u003c/p\u003e \u003cp\u003e14\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCluster\u003c/p\u003e \u003cp\u003e6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCluster 11\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCluster 19\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAngiopoietin 2, \u003cem\u003eAngpt2\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eApelin, \u003cem\u003eApln\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eApelin receptor, \u003cem\u003eAplnr\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCD31, \u003cem\u003ePecam1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehomophilic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCD34, \u003cem\u003eCd34\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClaudin 5, \u003cem\u003eCldn5\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehomophilic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEndothelin 1, \u003cem\u003eEdn1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKinase insert domain receptor, \u003cem\u003eKdr\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ec\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProtein C receptor, \u003cem\u003eProcr\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eP selectin, \u003cem\u003eSelp\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTEK receptor tyrosine kinase (TIE2), \u003cem\u003eTek\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTyrosine kinase with Ig-like and EGF-like domains\u003c/p\u003e \u003cp\u003e(TIE1), \u003cem\u003eTie1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVascular endothelial growth factor A, \u003cem\u003eVegfa\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ec\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVE-cadherin 5, \u003cem\u003eCdh5\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ehomophilic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVimentin, \u003cem\u003eVim\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003eEndothelial cell markers detected in CD146\u003csup\u003e+\u003c/sup\u003e/Pdgfrb\u003csup\u003e\u0026minus;\u003c/sup\u003e cells are listed. Each cluster is annotated using tSNE data from scRNA-seq.\u0026nbsp;Interaction types are indicated by matching letters (a\u0026ndash;c), representing predicted ligand\u0026ndash;receptor pairs. +: positive expression, -: negative expression\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eType-2 pericyte markers detected in CD146\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e cells.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eType-2 pericyte markers\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003eCD146\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e cells\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCluster\u003c/p\u003e \u003cp\u003e14\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCluster\u003c/p\u003e \u003cp\u003e6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCluster\u003c/p\u003e \u003cp\u003e11\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCluster\u003c/p\u003e \u003cp\u003e19\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eChondroitin sulfate proteoglycan 4, neuron-glial antigen 2 (NG2), \u003cem\u003eCspg4\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClass III β-tubulin, \u003cem\u003eTubb3\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFibroblast growth factor 5, \u003cem\u003eFgf5\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMatrix metalloproteinase 2, \u003cem\u003eMmp2\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNerve growth factor receptor, \u003cem\u003eNgfr\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN-cadherin, \u003cem\u003eCdh2\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNestin, \u003cem\u003eNes\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePlatelet-derived growth factor A, \u003cem\u003ePdgfa\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePlatelet-derived growth factor B, \u003cem\u003ePdgfb\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStanniocalcin 2, \u003cem\u003eStc2\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSRY-box transcription factor 2, \u003cem\u003eSox2\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSRY-box transcription factor 10, \u003cem\u003eSox10\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVascular endothelial growth factor A, \u003cem\u003eVegfa\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003eType-2 pericyte markers detected in CD146\u003csup\u003e+\u003c/sup\u003e/Pdgfrb\u003csup\u003e\u0026minus;\u003c/sup\u003e cells are listed. Each cluster is annotated using tSNE data from scRNA-seq.\u0026nbsp;+: positive expression, -: negative expression\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eGene expression profile in CD146\u003csup\u003e+\u003c/sup\u003e endothelial cells is conserved in rheumatoid arthritis and lung cancer\u003c/h2\u003e \u003cp\u003eUsing scRNA-seq data from previous studies on rheumatoid arthritis (RA) and lung cancer [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], we found similar subpopulations of CD146\u003csup\u003e+\u003c/sup\u003e endothelial cells in both human RA and lung cancer patients (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-E). The gene expression characteristics of endothelial cells from our dataset has been summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Several angiogenesis-related ligand and receptor genes were found to expressed by endothelial cells, such as angiopoietin 2 (\u003cem\u003eAngpt2\u003c/em\u003e), tyrosine kinase with immunoglobulin-like and EGF-like domains (\u003cem\u003eTie1\u003c/em\u003e), and TEK receptor tyrosine kinase (\u003cem\u003eTie2\u003c/em\u003e) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, interaction a); \u003cem\u003eApln\u003c/em\u003e and \u003cem\u003eAplnr\u003c/em\u003e (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, interaction b); and vascular endothelial growth factor A (\u003cem\u003eVegfα\u003c/em\u003e) and Kinase insert domain receptor (\u003cem\u003eKdr\u003c/em\u003e) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, interaction c). Thus, neovascularization in the pannus may largely be driven by endothelial cell autocrine and/ or paracrine signaling. To further characterize these endothelial cells, qPCR was performed for \u003cem\u003eApln\u003c/em\u003e and \u003cem\u003eAplnr\u003c/em\u003e. Both genes were highly expressed in pannus from mouse polyarthritis, but were only faintly detectable in the kidney and the liver (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF-G). Furthermore, immunohistochemical data in mouse polyarthritis support overlapping signals between CD31, Aplnr, and CD146 (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eH-I).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eCD146\u003csup\u003e+\u003c/sup\u003e endothelial cells express neovascularization-related genes\u003c/h2\u003e \u003cp\u003eOverlap in \u003cem\u003eApln\u003c/em\u003e and \u003cem\u003eAplnr\u003c/em\u003e gene expression in scRNA-seq data revealed the presence of \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e cells within specific populations (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA-C). As previous studies have found that \u003cem\u003eApln\u003c/em\u003e-expressing \u003cem\u003eAplnr\u003c/em\u003e-positive gCap (\u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e general capillary, gCap) exist as intermediate cells during the differentiation process from gCap to aerocyte (aCap) [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], we observed that \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e cells exhibiting characteristics similar to \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e gCap exist in mouse polyarthritis model and human RA, but only 0.4% in lung cancer (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). Since these cells play a role in neovascularization rather than gas exchange, we assessed the expression of neovascularization genes in this specific population (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE-G). In mouse polyarthritis, most \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e cells expressed \u003cem\u003eAngpt2\u003c/em\u003e, \u003cem\u003eCd34\u003c/em\u003e, and \u003cem\u003eKdr\u003c/em\u003e, while 35% of \u003cem\u003eApln\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003eAplnr\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e cells expressed two additional neovascularization-associated genes, \u003cem\u003eCldn5\u003c/em\u003e and \u003cem\u003eProcr\u003c/em\u003e, an expression pattern that corresponds well with observations in lung cancer.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eUnique transcriptomic profile of pericyte subpopulations\u003c/h2\u003e \u003cp\u003e\u003cem\u003ePdgfrb\u003c/em\u003e, though widely expressed in mature pericytes, was not found to be expressed in most CD146\u003csup\u003e+\u003c/sup\u003e cells (including type-2 pericytes) (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA-B, Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, and Supplementary Tables\u0026nbsp;1). Thus, CD146\u003csup\u003e\u0026minus;\u003c/sup\u003e\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e+\u003c/sup\u003e type-1 pericytes and CD146\u003csup\u003e+\u003c/sup\u003e\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e type-2 pericytes were mutually exclusive populations. Cells expressing \u003cem\u003eMcam\u003c/em\u003e (CD146 gene) were mainly found in c14, c6, c11, and c19, clusters which comprise the CD146\u003csup\u003emid\u003c/sup\u003e and CD146\u003csup\u003ehigh\u003c/sup\u003e populations in the initial FACS analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD) and were classified by Seurat cluster analysis as endothelial cells (c14) and type-2 pericytes (c6, c11, and c19) (Tables\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). In putative type-1 pericytes, only c3 expressed endothelin receptor type A (\u003cem\u003eEdnra\u003c/em\u003e) and type B (\u003cem\u003eEdnrb\u003c/em\u003e) (Supplementary Table\u0026nbsp;1). To probe further, we looked at other pericyte markers in these clusters and found that \u003cem\u003eFgf5\u003c/em\u003e, in addition to \u003cem\u003eCspg4\u003c/em\u003e, but not \u003cem\u003ePdgfrb\u003c/em\u003e, were highly expressed in these populations (Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB, \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC-D). In addition, we found that \u003cem\u003eStc2\u003c/em\u003e, \u003cem\u003eSox2\u003c/em\u003e, and \u003cem\u003eSox10\u003c/em\u003e were elevated in c6, c11, and c19, genes predominantly associated with neovascularization and progenitor-like behavior at the site of injury (Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eE-G). To investigate the spatial localization of these pericyte-like cells in the inflamed joints, we used immunohistochemical staining for Ng2, CD31, and Pdgfrβ, which revealed that most CD31\u003csup\u003e+\u003c/sup\u003e vessels were surrounded by Pdgfrβ\u003csup\u003e+\u003c/sup\u003e fibroblasts and some vessels were covered with Ng2\u003csup\u003e+\u003c/sup\u003e cells in the same joint and little co-localization was noted (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eH).\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eEndothelial cells and pericytes drive pannus neovascularization signaling\u003c/h2\u003e \u003cp\u003eNext, we analyzed cell-cell interaction between CD146\u003csup\u003e+\u003c/sup\u003e endothelial cells (c14) and type-2 pericytes (c6, c11, and c19) with the ligand-receptor prediction tool LIANA, which predicts interactions between each cell from scRNA-seq data (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA-B). Many ligands released by fibroblast and pericyte populations interact strongly with receptors expressed by endothelial cells (cluster 14), while macrophage-associated ligands only showed weak interactions with endothelial cell receptors (Figs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC-F. and Supplementary Fig.\u0026nbsp;2). Focusing on CD146\u003csup\u003e+\u003c/sup\u003e cells, delta like canonical notch ligand 4 (\u003cem\u003eDll4\u003c/em\u003e) was found to be expressed in c14 and is known to interact with Notch4 receptor. \u003cem\u003eNotch4\u003c/em\u003e was expressed in c6, c11, and c19 and is a receptor involved in branch promotion and endothelial stabilization during angiogenesis [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Jagged 1 (\u003cem\u003eJag1\u003c/em\u003e) was also expressed in c14 and has previously been shown to interact with Notch4 and block Dll4 binding (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Other genes involved in angiogenesis such as slit guidance ligand 3, \u003cem\u003eSlit3\u003c/em\u003e, was expressed in c6, c11, and c19 while its receptor, roundabout guidance receptor 4 (\u003cem\u003eRobo4\u003c/em\u003e) was expressed on c14 (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Also, pleiotrophin (\u003cem\u003ePtn\u003c/em\u003e) was found to be expressed by c6, c11, and c19, which may modulate excessive formation of microvasculature via binding to Kdr [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], expressed on c14 (Tables\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and Supplementary Table\u0026nbsp;1). Furthermore, protein tyrosine phosphatase receptor type B (\u003cem\u003ePtprb)\u003c/em\u003e, which negatively regulates Kdr by dephosphorylation, was also expressed on c14 [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMolecular signature of endothelial cell and pericyte interactions in CD146\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e cells.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSignaling molecules\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003eCD146\u003csup\u003e+\u003c/sup\u003e/\u003cem\u003ePdgfrb\u003c/em\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e cells\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eInteraction\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHypoxia\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCluster 14\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCluster 6\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCluster 11\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eCluster 19\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDelta-like canonical Notch ligand 4, \u003cem\u003eDll4\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ea\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNotch 1, \u003cem\u003eNotch1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ea,b\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNotch 4, \u003cem\u003eNotch4\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ea,b\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eJagged 1, \u003cem\u003eJag1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoundabout guidance receptor 4, \u003cem\u003eRobo4\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ec\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSlit guidance ligand 3, \u003cem\u003eSlit3\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ec\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePleiotrophin, \u003cem\u003ePtn\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProtein tyrosine phosphatase receptor type B, \u003cem\u003ePtprb\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e+\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e \u003cp\u003eEach cluster is annotated using tSNE data from scRNA-seq.\u0026nbsp;Interaction types are indicated by matching letters (a\u0026ndash;c), representing predicted ligand\u0026ndash;receptor pairs. Hypoxia associated genes are marked with *. +: positive expression, -: negative expression\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eHowever, based on the proximity of these pericyte-like cells to newly formed pannus vessels, we used cell-cell interaction data via LIANA analysis to look at which ligands expressed in c6, c11, and c19 may be signaling through receptors expressed on endothelial cells (c14). As expected, several of the ligand-receptor factors predicted were associated with neovascularization and interaction between type-1 and type-2 pericytes (Figs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA-B, Supplementary Fig.\u0026nbsp;4).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we characterized neovascularization in the arthritic joint using a mouse model of polyarthritis. A small number of CD146\u003csup\u003e+\u003c/sup\u003e cells were identified within the inflammatory synovium. To elucidate the specific role these CD146⁺ cells play in neovascularization, bulk RNA-seq and scRNA-seq analyses were performed. CD146\u003csup\u003e+\u003c/sup\u003e cells included various clusters of endothelial cells (c14) and type-2 pericytes (c6, c11, and c19). Although these cells express \u003cem\u003eCd146\u003c/em\u003e and are involved in neovascularization based on interaction analysis, no \u003cem\u003ePdgfrb\u003c/em\u003e expression was observed, suggesting an intermediate, dedifferentiated state. Indeed, they exhibited gene expression patterns similar to transcriptomic profiles previously characterized in pulmonary vasculature during wound healing, including expression of \u003cem\u003eApln\u003c/em\u003e and \u003cem\u003eAplnr\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eApelin-positive aCap are flattened endothelial cells that interact with alveolar type I epithelial cells, forming an ultrathin layer that facilitates gas exchange barrier in the lung while apelin receptor-positive gCap cells maintain vascular homeostasis, regulate permeability, and serve as progenitors for aCap during tissue repair [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Although typical lung aCap cells lack the expression of \u003cem\u003eVwf\u003c/em\u003e, \u003cem\u003eSelp\u003c/em\u003e, and endothelin 1 (\u003cem\u003eEdn1\u003c/em\u003e), we observed that CD146\u003csup\u003e+\u003c/sup\u003e endothelial cells in inflammatory synovium express \u003cem\u003eApln\u003c/em\u003e and/or \u003cem\u003eAplnr\u003c/em\u003e as well as \u003cem\u003eSelp\u003c/em\u003e and \u003cem\u003eEdn1\u003c/em\u003e (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Furthermore, these cells also express \u003cem\u003eCldn5\u003c/em\u003e and stem cell marker, \u003cem\u003eCd34\u003c/em\u003e, indicating that cell lineage is directed toward neovascularization rather than gas-exchange [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Interestingly, we observed double-positive \u003cem\u003eApln\u003c/em\u003e-expressing gGap endothelial cells, which have previously been found to exhibit a stem cell-like phenotype in the lungs immediately after injury [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. In line with this increase in neovascular signaling, we found that wound healing and hypoxia-related gene expression was detected, such as \u003cem\u003eApln\u003c/em\u003e, \u003cem\u003eEdn1\u003c/em\u003e, \u003cem\u003eCd34\u003c/em\u003e, \u003cem\u003eVim\u003c/em\u003e, and \u003cem\u003eKdr\u003c/em\u003e [\u003cspan additionalcitationids=\"CR30 CR31 CR32 CR33\" citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Other hypoxia-related genes \u003cem\u003eAngpt2\u003c/em\u003e (c14), \u003cem\u003ePdgfb\u003c/em\u003e (c6, c11, c14, and c19), and \u003cem\u003eVegfa\u003c/em\u003e (c14) were also detected. It is possible that VEGF signaling might upregulate \u003cem\u003eDll4\u003c/em\u003e and \u003cem\u003eNotch4\u003c/em\u003e expression (both in c14), which regulate extension and branching of capillary with Jagged1 (\u003cem\u003eJag1\u003c/em\u003e in c6, c11, c14, and c19) [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. In addition, hypoxia disrupts the tight junction of endothelial cells via decreasing the protein levels of Cldn5, despite little effect on \u003cem\u003eCldn5\u003c/em\u003e mRNA expression in c14 [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e], thus affecting tight junctions between endothelial cells potentially contributing to edema and inflammation of the joint.\u003c/p\u003e \u003cp\u003eIn addition to endothelial cells, RNA-seq data revealed that CD146\u003csup\u003e+\u003c/sup\u003e cells included pericytes. Our scRNA-seq data suggest that \u003cem\u003ePdgfrb\u003c/em\u003e-negative cells (c6, c11, and c19) represent type-2 pericytes, as these clusters were found to express type-2 pericyte-related genes such as \u003cem\u003eFgf5\u003c/em\u003e, \u003cem\u003eNes\u003c/em\u003e, \u003cem\u003eTubb3\u003c/em\u003e, \u003cem\u003eCspg4\u003c/em\u003e, N-cadherin (\u003cem\u003eCdh2\u003c/em\u003e) (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Also, \u003cem\u003eNotch1\u003c/em\u003e (c14, 6, 11, and 19) and \u003cem\u003eNotch4\u003c/em\u003e (c14) were detected, while \u003cem\u003eNotch3\u003c/em\u003e expression was not observed in any cluster following scRNA-seq analysis and only very low expression was noted using bulk RNA-seq (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e)[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. It is possible that this explains the lack of \u003cem\u003ePdgfrb\u003c/em\u003e expression since a lack of functional Notch3 is known to be associated with very low \u003cem\u003ePdgfrb\u003c/em\u003e expression and pericyte dysfunction [\u003cspan additionalcitationids=\"CR39 CR40\" citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. Therefore, c6, 11, and 19 may fail to detect Pdgf signaling as they lack the appropriate receptor, which may affect their ability to associate with newly formed vessels leading to increased vascular permeability. This leakage into the synovial tissue likely contributes to pannus formation and joint swelling while promoting chronic inflammatory cell infiltration [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Other pericytes are classified as type-1 pericytes that express \u003cem\u003ePdgfrb\u003c/em\u003e; however, most type-1 pericytes do not express \u003cem\u003eEdnra\u003c/em\u003e or \u003cem\u003eEdnrb\u003c/em\u003e except cluster 3 which was classified as immature type-1 pericytes (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Overall, this suggests that maintenance of vascular tension is insufficient and that the lack of functional, mature pericytes contributes to incomplete junction formation in new blood vessels and the development and persistence of joint edema [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIt seems that endothelial cells and type-2 pericytes play synergistic roles in pannus neovascularization, where both cell types were found to be colocalized in the inflamed joint, as observed via histopathology. Further supporting these interactions, LIANA analysis revealed that CD146\u003csup\u003e+\u003c/sup\u003e endothelial cells (c14) might associate directly via ligand/receptor interaction with c6, c11, and c19 within inflammatory synovium, especially under hypoxic conditions. Among the many binding partners between endothelial cells and type-2 pericytes listed in LIANA data of Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e and Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, \u003cem\u003eRobo4\u003c/em\u003e expressed in CD146\u003csup\u003e+\u003c/sup\u003e endothelial cell (c14) and its ligand, \u003cem\u003eSlit3\u003c/em\u003e, expressed in all type-2 pericytes are known to have similar effects as \u003cem\u003ePtn\u003c/em\u003e and \u003cem\u003eKdr\u003c/em\u003e interactions, inducing endothelial cell migration and tube formation during neovascularization [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Within endothelial cells, other interesting binding pairs found in our analysis are the juxtacrine interaction between \u003cem\u003eApln\u003c/em\u003e and \u003cem\u003eAplnr\u003c/em\u003e within endothelial cells and the interaction between \u003cem\u003eAngpt2\u003c/em\u003e, and \u003cem\u003eTek\u003c/em\u003e which is expressed exclusively in endothelial cells (c14). The latter molecular interaction has previously been shown to promote vascular sprouting along with VEGF (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]. Therefore, neovascularization processes orchestrated by both endothelial cells and pericytes represent an interesting potential therapeutic target in treating arthritis.\u003c/p\u003e \u003cp\u003ePericytes are generally considered a subset of mural cells; however, the present study focuses specifically on pericyte-like populations and does not encompass the full spectrum of mural cell subtypes such as vascular smooth muscle cells. Our characterization of CD146-positive cells in the synovium reveals new cell populations and suggests important interactions in the inflamed joint. However, there are some important caveats to the interpretation of these transcriptomic studies. Since only a small number of synovial cells could be collected from very small murine joint tissue, cells had to be cultured for three days prior to bulk RNA-seq and scRNA-seq analysis. During this period, the number of CD146\u003csup\u003e+\u003c/sup\u003e cells increased slightly compared to direct FACS analysis of synovial cells. However, comparison of bulk RNA-seq and scRNA-seq data revealed that culture had little effect on gene expression profiles. Another limitation of our study is the reliance on RNA sequencing data, which does not fully reflect protein abundance and functional activity in the cell. To complement this, histological data analysis using immunostaining was performed. In some area of the pannus, typical neovascularization was observed, involving CD31\u003csup\u003e+\u003c/sup\u003e endothelial cells and Ng2\u003csup\u003e+\u003c/sup\u003e/Pdgfrb\u003csup\u003e+\u003c/sup\u003e pericytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eH, yellow square). However, in another region, coexistence of Ng2\u003csup\u003e+\u003c/sup\u003e/Pdgfrb\u003csup\u003e\u0026minus;\u003c/sup\u003e cells (likely type-2 pericytes) and CD31\u003csup\u003e+\u003c/sup\u003e endothelial cells was also confirmed (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eH, blue square). This suggests that Ng2\u003csup\u003e+\u003c/sup\u003e/Pdgfrb\u003csup\u003e\u0026minus;\u003c/sup\u003e also colocalize with endothelial cells in the pannus.\u003c/p\u003e \u003cp\u003eThe formation of new blood vessels plays a key role in the development of joint arthritis and the persistence of inflammation in the synovium [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Based on our RNA sequencing data, hypoxic conditions in the joint cavity influence neovascularization through various signaling pathways primarily orchestrated by endothelial cells and pericytes. We found that CD146\u003csup\u003e+\u003c/sup\u003e cell population expression patterns mirrored the vascular repair program normally activated in the lungs during injury and wound healing. However, it seems that neovascularization processes and interactions between endothelial cells and pericytes are insufficient, lacking certain markers required for proper junction formation. Thus, the commonalities between vascular regeneration and neovascularization within inflammatory pannus provide important insights for identifying anti-angiogenic therapeutic targets in RA and suggests alternative therapeutic targets aimed at modulating stromal-vascular interactions and reducing vascularization of arthritic tissue.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn summary, we provide a comprehensive characterization of CD146⁺ vascular cell populations in inflamed synovium using bulk and single-cell transcriptomic approaches. Our analysis identifies heterogeneous endothelial and pericyte states associated with neovascularization in inflammatory arthritis and suggests the presence of coordinated signaling interactions between these cell populations. While certain endothelial subsets share transcriptional features with capillary repair programs described in other tissues, their functional roles in synovial angiogenesis remain to be established. Importantly, our findings highlight the complexity of vascular remodeling in RA pannus and provide a resource for future studies aimed at dissecting endothelial-pericyte interactions in inflammatory settings. Given the descriptive nature of this study and limitations in cell isolation strategies, further functional and lineage-resolved analyses will be required to determine the precise contribution of these vascular cell subsets to disease progression and their potential as therapeutic targets.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003erheumatoid arthritis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCspg4\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eChondroitin sulfate proteoglycan 4\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNG2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eneuron-glial antigen 2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSASP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003esenescence-associated secretory phenotype\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eFgf5\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003efibroblast growth factor 5\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eStc2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003estanniocalcin 2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSox2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSRY-box transcription factor 2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSox10\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSRY-box transcription factor 10\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMcam\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003emelanoma cell adhesion molecule (CD146)\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003escRNA-seq\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003esingle-cell RNA sequencing\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ebulk RNA-seq\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ebulk RNA sequencing\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAplnr\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eapelin receptor\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ebColII\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ebovine collagen type II\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eFACS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003efluorescence-activated cell sorting\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eVim\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003evimentin\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eAngpt2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eangiopoietin 2\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eTie1\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003etyrosine kinase with immunoglobulin-like and EGF-like domains\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eTie2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eTEK receptor tyrosine kinase\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eVegfα\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003evascular endothelial growth factor A\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eKdr\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eKinase insert domain receptor\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003egCap\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003egeneral capillary\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEdnra\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eendothelin receptor type A\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEdnrb\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eendothelin receptor type B\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDll4\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003edelta-like canonical Notch ligand 4\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eJag1\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eJagged 1\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRobo4\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eroundabout guidance receptor 4\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePtprb\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eprotein tyrosine phosphatase receptor type B\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eaCap\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eaerocyte\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEdn1\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eendothelin 1\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCdh2\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eN-cadherin\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll mouse experiments were performed according to the rules and regulations of the Fundamental Guidelines for Proper Conduct of Animal Experiments and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education, Culture, Sports, Science, and Technology of Japan and\u0026nbsp;were approved by the Committee on the Ethics of Animal Experiments of Nagoya City University.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have read and approved the final version of the manuscript and agree to its submission and publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData will be made available by the authors, upon reasonable request. Additionally, the\u0026nbsp;series accession no. GSE311750 and GSE311751 for gene expression array is available.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNone of the authors has any conflicts of interest, financial or otherwise, to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was funded by Grants-in Aid from the Ministry of Education, Culture, Sports, Science and Technology (MEXT)/ JSPS KAKENHI Grant Number 23K07879 and JSPS Short-term Fellowship PE25020. This work was supported by JSPS KAKENHI Grant Number JP22H04925 (PAGS).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors’ Contributions\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eY.M., N.M., and S.K.(CA) conceived and the designed the research; Y.M., N.M, A.U., and S.K. (CA) performed experiments; Y.M., N.M, H.N., and S.K.(CA) analyzed the data; Y.M., N.M., S.K., H.N., and S.K.(CA) interpreted results of experiments; Y.M., N.M., S.K., and S.K(CA). prepared figures; Y.M., N.M., and S.K.(CA) drafted manuscript; Y.M., N.M., and S.K.(CA) edited and revised manuscript; Y.M., N.M., S.K., H.N. and S.K.(CA) approved the final version of manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to thank Ayako Nitatoge, Satoko Ogawa, and Yoko Kanazawa for excellent technical support.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eLiu M, Liu P, Li J, Huang Y, Wu R: \u003cstrong\u003eVascular synovial phenotype indicates poor response to JAK inhibitors in rheumatoid arthritis patients: a pilot study\u003c/strong\u003e. \u003cem\u003ePeerJ\u0026nbsp;\u003c/em\u003e2024, \u003cstrong\u003e12\u003c/strong\u003e:e18631.\u003c/li\u003e\n \u003cli\u003eLesturgie-Talarek M, Gonzalez V, Combier A, Thomas M, Boisson M, Poiroux L, Wanono S, Hecquet S, Carves S, Cauvet A\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eInflammatory and angiogenic serum profile of refractory rheumatoid arthritis\u003c/strong\u003e. \u003cem\u003eScientific reports\u0026nbsp;\u003c/em\u003e2025, \u003cstrong\u003e15\u003c/strong\u003e(1):7159.\u003c/li\u003e\n \u003cli\u003eGodoy RS, Cober ND, Cook DP, McCourt E, Deng Y, Wang L, Schlosser K, Rowe K, Stewart DJ: \u003cstrong\u003eSingle-cell transcriptomic atlas of lung microvascular regeneration after targeted endothelial cell ablation\u003c/strong\u003e. \u003cem\u003eElife\u0026nbsp;\u003c/em\u003e2023, \u003cstrong\u003e12\u003c/strong\u003e.\u003c/li\u003e\n \u003cli\u003eGoss G, Rognoni E, Salameti V, Watt FM: \u003cstrong\u003eDistinct Fibroblast Lineages Give Rise to NG2+ Pericyte Populations in Mouse Skin Development and Repair\u003c/strong\u003e. \u003cem\u003eFront Cell Dev Biol\u0026nbsp;\u003c/em\u003e2021, \u003cstrong\u003e9\u003c/strong\u003e:675080.\u003c/li\u003e\n \u003cli\u003eBirbrair A, Zhang T, Wang ZM, Messi ML, Enikolopov GN, Mintz A, Delbono O: \u003cstrong\u003eRole of pericytes in skeletal muscle regeneration and fat accumulation\u003c/strong\u003e. \u003cem\u003eStem Cells Dev\u0026nbsp;\u003c/em\u003e2013, \u003cstrong\u003e22\u003c/strong\u003e(16):2298-2314.\u003c/li\u003e\n \u003cli\u003eSchupp JC, Adams TS, Cosme C, Jr., Raredon MSB, Yuan Y, Omote N, Poli S, Chioccioli M, Rose KA, Manning EP\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eIntegrated Single-Cell Atlas of Endothelial Cells of the Human Lung\u003c/strong\u003e. \u003cem\u003eCirculation\u0026nbsp;\u003c/em\u003e2021, \u003cstrong\u003e144\u003c/strong\u003e(4):286-302.\u003c/li\u003e\n \u003cli\u003eBirbrair A, Zhang T, Files DC, Mannava S, Smith T, Wang ZM, Messi ML, Mintz A, Delbono O: \u003cstrong\u003eType-1 pericytes accumulate after tissue injury and produce collagen in an organ-dependent manner\u003c/strong\u003e. \u003cem\u003eStem Cell Res Ther\u0026nbsp;\u003c/em\u003e2014, \u003cstrong\u003e5\u003c/strong\u003e(6):122.\u003c/li\u003e\n \u003cli\u003eChen J, Luo Y, Huang H, Wu S, Feng J, Zhang J, Yan X: \u003cstrong\u003eCD146 is essential for PDGFRbeta-induced pericyte recruitment\u003c/strong\u003e. \u003cem\u003eProtein Cell\u0026nbsp;\u003c/em\u003e2018, \u003cstrong\u003e9\u003c/strong\u003e(8):743-747.\u003c/li\u003e\n \u003cli\u003eBirbrair A, Zhang T, Wang ZM, Messi ML, Olson JD, Mintz A, Delbono O: \u003cstrong\u003eType-2 pericytes participate in normal and tumoral angiogenesis\u003c/strong\u003e. \u003cem\u003eAm J Physiol Cell Physiol\u0026nbsp;\u003c/em\u003e2014, \u003cstrong\u003e307\u003c/strong\u003e(1):C25-38.\u003c/li\u003e\n \u003cli\u003eLoan A, Awaja N, Lui M, Syal C, Sun Y, Sarma SN, Chona R, Johnston WB, Cordova A, Saraf D\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eSingle-cell profiling of brain pericyte heterogeneity following ischemic stroke unveils distinct pericyte subtype-targeted neural reprogramming potential and its underlying mechanisms\u003c/strong\u003e. \u003cem\u003eTheranostics\u0026nbsp;\u003c/em\u003e2024, \u003cstrong\u003e14\u003c/strong\u003e(16):6110-6137.\u003c/li\u003e\n \u003cli\u003eWang D, Wu F, Yuan H, Wang A, Kang GJ, Truong T, Chen L, McCallion AS, Gong X, Li S: \u003cstrong\u003eSox10(+) Cells Contribute to Vascular Development in Multiple Organs-Brief Report\u003c/strong\u003e. \u003cem\u003eArteriosclerosis, thrombosis, and vascular biology\u0026nbsp;\u003c/em\u003e2017, \u003cstrong\u003e37\u003c/strong\u003e(9):1727-1731.\u003c/li\u003e\n \u003cli\u003eWang Z, Xu Q, Zhang N, Du X, Xu G, Yan X: \u003cstrong\u003eCD146, from a melanoma cell adhesion molecule to a signaling receptor\u003c/strong\u003e. \u003cem\u003eSignal Transduct Target Ther\u0026nbsp;\u003c/em\u003e2020, \u003cstrong\u003e5\u003c/strong\u003e(1):148.\u003c/li\u003e\n \u003cli\u003eYan B, Lu Q, Gao T, Xiao K, Zong Q, Lv H, Lv G, Wang L, Liu C, Yang W\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eCD146 regulates the stemness and chemoresistance of hepatocellular carcinoma via JAG2-NOTCH signaling\u003c/strong\u003e. \u003cem\u003eCell Death Dis\u0026nbsp;\u003c/em\u003e2025, \u003cstrong\u003e16\u003c/strong\u003e(1):150.\u003c/li\u003e\n \u003cli\u003eGillich A, Zhang F, Farmer CG, Travaglini KJ, Tan SY, Gu M, Zhou B, Feinstein JA, Krasnow MA, Metzger RJ: \u003cstrong\u003eCapillary cell-type specialization in the alveolus\u003c/strong\u003e. \u003cem\u003eNature\u0026nbsp;\u003c/em\u003e2020, \u003cstrong\u003e586\u003c/strong\u003e(7831):785-789.\u003c/li\u003e\n \u003cli\u003eMiura Y, Isogai S, Maeda S, Kanazawa S: \u003cstrong\u003eCTLA-4-Ig internalizes CD80 in fibroblast-like synoviocytes from chronic inflammatory arthritis mouse model\u003c/strong\u003e. \u003cem\u003eScientific reports\u0026nbsp;\u003c/em\u003e2022, \u003cstrong\u003e12\u003c/strong\u003e(1):16363.\u003c/li\u003e\n \u003cli\u003eShichino S, Ueha S, Hashimoto S, Ogawa T, Aoki H, Wu B, Chen CY, Kitabatake M, Ouji-Sageshima N, Sawabata N\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eTAS-Seq is a robust and sensitive amplification method for bead-based scRNA-seq\u003c/strong\u003e. \u003cem\u003eCommun Biol\u0026nbsp;\u003c/em\u003e2022, \u003cstrong\u003e5\u003c/strong\u003e(1):602.\u003c/li\u003e\n \u003cli\u003eEdalat SG, Gerber R, Houtman M, Luckgen J, Teixeira RL, Palacios Cisneros MDP, Pfanner T, Kuret T, Izanc N, Micheroli R\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eMolecular maps of synovial cells in inflammatory arthritis using an optimized synovial tissue dissociation protocol\u003c/strong\u003e. \u003cem\u003eiScience\u0026nbsp;\u003c/em\u003e2024, \u003cstrong\u003e27\u003c/strong\u003e(6):109707.\u003c/li\u003e\n \u003cli\u003eGoveia J, Rohlenova K, Taverna F, Treps L, Conradi LC, Pircher A, Geldhof V, de Rooij L, Kalucka J, Sokol L\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eAn Integrated Gene Expression Landscape Profiling Approach to Identify Lung Tumor Endothelial Cell Heterogeneity and Angiogenic Candidates\u003c/strong\u003e. \u003cem\u003eCancer cell\u0026nbsp;\u003c/em\u003e2020, \u003cstrong\u003e37\u003c/strong\u003e(1):21-36 e13.\u003c/li\u003e\n \u003cli\u003eMabbott NA, Baillie JK, Brown H, Freeman TC, Hume DA: \u003cstrong\u003eAn expression atlas of human primary cells: inference of gene function from coexpression networks\u003c/strong\u003e. \u003cem\u003eBMC Genomics\u0026nbsp;\u003c/em\u003e2013, \u003cstrong\u003e14\u003c/strong\u003e:632.\u003c/li\u003e\n \u003cli\u003eAran D, Looney AP, Liu L, Wu E, Fong V, Hsu A, Chak S, Naikawadi RP, Wolters PJ, Abate AR\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eReference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage\u003c/strong\u003e. \u003cem\u003eNature immunology\u0026nbsp;\u003c/em\u003e2019, \u003cstrong\u003e20\u003c/strong\u003e(2):163-172.\u003c/li\u003e\n \u003cli\u003eCrisan M, Yap S, Casteilla L, Chen CW, Corselli M, Park TS, Andriolo G, Sun B, Zheng B, Zhang L\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eA perivascular origin for mesenchymal stem cells in multiple human organs\u003c/strong\u003e. \u003cem\u003eCell stem cell\u0026nbsp;\u003c/em\u003e2008, \u003cstrong\u003e3\u003c/strong\u003e(3):301-313.\u003c/li\u003e\n \u003cli\u003eZhang F, Jonsson AH, Nathan A, Millard N, Curtis M, Xiao Q, Gutierrez-Arcelus M, Apruzzese W, Watts GFM, Weisenfeld D\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eDeconstruction of rheumatoid arthritis synovium defines inflammatory subtypes\u003c/strong\u003e. \u003cem\u003eNature\u0026nbsp;\u003c/em\u003e2023, \u003cstrong\u003e623\u003c/strong\u003e(7987):616-624.\u003c/li\u003e\n \u003cli\u003eMiura Y, Ota S, Peterlin M, McDevitt G, Kanazawa S: \u003cstrong\u003eA Subpopulation of Synovial Fibroblasts Leads to Osteochondrogenesis in a Mouse Model of Chronic Inflammatory Rheumatoid Arthritis\u003c/strong\u003e. \u003cem\u003eJBMR Plus\u0026nbsp;\u003c/em\u003e2019, \u003cstrong\u003e3\u003c/strong\u003e(6):e10132.\u003c/li\u003e\n \u003cli\u003ePedrosa AR, Trindade A, Fernandes AC, Carvalho C, Gigante J, Tavares AT, Dieguez-Hurtado R, Yagita H, Adams RH, Duarte A: \u003cstrong\u003eEndothelial Jagged1 antagonizes Dll4 regulation of endothelial branching and promotes vascular maturation downstream of Dll4/Notch1\u003c/strong\u003e. \u003cem\u003eArteriosclerosis, thrombosis, and vascular biology\u0026nbsp;\u003c/em\u003e2015, \u003cstrong\u003e35\u003c/strong\u003e(5):1134-1146.\u003c/li\u003e\n \u003cli\u003eTefft JB, Bays JL, Lammers A, Kim S, Eyckmans J, Chen CS: \u003cstrong\u003eNotch1 and Notch3 coordinate for pericyte-induced stabilization of vasculature\u003c/strong\u003e. \u003cem\u003eAm J Physiol Cell Physiol\u0026nbsp;\u003c/em\u003e2022, \u003cstrong\u003e322\u003c/strong\u003e(2):C185-C196.\u003c/li\u003e\n \u003cli\u003eZhang B, Dietrich UM, Geng JG, Bicknell R, Esko JD, Wang L: \u003cstrong\u003eRepulsive axon guidance molecule Slit3 is a novel angiogenic factor\u003c/strong\u003e. \u003cem\u003eBlood\u0026nbsp;\u003c/em\u003e2009, \u003cstrong\u003e114\u003c/strong\u003e(19):4300-4309.\u003c/li\u003e\n \u003cli\u003eEltanahy AM, Koluib YA, Gonzales A: \u003cstrong\u003ePericytes: Intrinsic Transportation Engineers of the CNS Microcirculation\u003c/strong\u003e. \u003cem\u003eFront Physiol\u0026nbsp;\u003c/em\u003e2021, \u003cstrong\u003e12\u003c/strong\u003e:719701.\u003c/li\u003e\n \u003cli\u003eMellberg S, Dimberg A, Bahram F, Hayashi M, Rennel E, Ameur A, Westholm JO, Larsson E, Lindahl P, Cross MJ\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eTranscriptional profiling reveals a critical role for tyrosine phosphatase VE-PTP in regulation of VEGFR2 activity and endothelial cell morphogenesis\u003c/strong\u003e. \u003cem\u003eFASEB J\u0026nbsp;\u003c/em\u003e2009, \u003cstrong\u003e23\u003c/strong\u003e(5):1490-1502.\u003c/li\u003e\n \u003cli\u003eHelker CS, Eberlein J, Wilhelm K, Sugino T, Malchow J, Schuermann A, Baumeister S, Kwon HB, Maischein HM, Potente M\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eApelin signaling drives vascular endothelial cells toward a pro-angiogenic state\u003c/strong\u003e. \u003cem\u003eElife\u0026nbsp;\u003c/em\u003e2020, \u003cstrong\u003e9\u003c/strong\u003e.\u003c/li\u003e\n \u003cli\u003eKidoya H, Naito H, Muramatsu F, Yamakawa D, Jia W, Ikawa M, Sonobe T, Tsuchimochi H, Shirai M, Adams RH\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eAPJ Regulates Parallel Alignment of Arteries and Veins in the Skin\u003c/strong\u003e. \u003cem\u003eDev Cell\u0026nbsp;\u003c/em\u003e2015, \u003cstrong\u003e33\u003c/strong\u003e(3):247-259.\u003c/li\u003e\n \u003cli\u003eEyries M, Siegfried G, Ciumas M, Montagne K, Agrapart M, Lebrin F, Soubrier F: \u003cstrong\u003eHypoxia-induced apelin expression regulates endothelial cell proliferation and regenerative angiogenesis\u003c/strong\u003e. \u003cem\u003eCirc Res\u0026nbsp;\u003c/em\u003e2008, \u003cstrong\u003e103\u003c/strong\u003e(4):432-440.\u003c/li\u003e\n \u003cli\u003eBartoszewski R, Moszynska A, Serocki M, Cabaj A, Polten A, Ochocka R, Dell\u0026apos;Italia L, Bartoszewska S, Kroliczewski J, Dabrowski M\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003ePrimary endothelial cell-specific regulation of hypoxia-inducible factor (HIF)-1 and HIF-2 and their target gene expression profiles during hypoxia\u003c/strong\u003e. \u003cem\u003eFASEB J\u0026nbsp;\u003c/em\u003e2019, \u003cstrong\u003e33\u003c/strong\u003e(7):7929-7941.\u003c/li\u003e\n \u003cli\u003ePugh CW, Ratcliffe PJ: \u003cstrong\u003eRegulation of angiogenesis by hypoxia: role of the HIF system\u003c/strong\u003e. \u003cem\u003eNature medicine\u0026nbsp;\u003c/em\u003e2003, \u003cstrong\u003e9\u003c/strong\u003e(6):677-684.\u003c/li\u003e\n \u003cli\u003eSmith MH, Gao VR, Periyakoil PK, Kochen A, DiCarlo EF, Goodman SM, Norman TM, Donlin LT, Leslie CS, Rudensky AY: \u003cstrong\u003eDrivers of heterogeneity in synovial fibroblasts in rheumatoid arthritis\u003c/strong\u003e. \u003cem\u003eNature immunology\u0026nbsp;\u003c/em\u003e2023, \u003cstrong\u003e24\u003c/strong\u003e(7):1200-1210.\u003c/li\u003e\n \u003cli\u003eBenedito R, Roca C, Sorensen I, Adams S, Gossler A, Fruttiger M, Adams RH: \u003cstrong\u003eThe notch ligands Dll4 and Jagged1 have opposing effects on angiogenesis\u003c/strong\u003e. \u003cem\u003eCell\u0026nbsp;\u003c/em\u003e2009, \u003cstrong\u003e137\u003c/strong\u003e(6):1124-1135.\u003c/li\u003e\n \u003cli\u003eKoto T, Takubo K, Ishida S, Shinoda H, Inoue M, Tsubota K, Okada Y, Ikeda E: \u003cstrong\u003eHypoxia disrupts the barrier function of neural blood vessels through changes in the expression of claudin-5 in endothelial cells\u003c/strong\u003e. \u003cem\u003eThe American journal of pathology\u0026nbsp;\u003c/em\u003e2007, \u003cstrong\u003e170\u003c/strong\u003e(4):1389-1397.\u003c/li\u003e\n \u003cli\u003eWei K, Korsunsky I, Marshall JL, Gao A, Watts GFM, Major T, Croft AP, Watts J, Blazar PE, Lange JK\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eNotch signalling drives synovial fibroblast identity and arthritis pathology\u003c/strong\u003e. \u003cem\u003eNature\u0026nbsp;\u003c/em\u003e2020, \u003cstrong\u003e582\u003c/strong\u003e(7811):259-264.\u003c/li\u003e\n \u003cli\u003eNadeem T, Bogue W, Bigit B, Cuervo H: \u003cstrong\u003eDeficiency of Notch signaling in pericytes results in arteriovenous malformations\u003c/strong\u003e. \u003cem\u003eJCI Insight\u0026nbsp;\u003c/em\u003e2020, \u003cstrong\u003e5\u003c/strong\u003e(21).\u003c/li\u003e\n \u003cli\u003eKofler NM, Cuervo H, Uh MK, Murtomaki A, Kitajewski J: \u003cstrong\u003eCombined deficiency of Notch1 and Notch3 causes pericyte dysfunction, models CADASIL, and results in arteriovenous malformations\u003c/strong\u003e. \u003cem\u003eScientific reports\u0026nbsp;\u003c/em\u003e2015, \u003cstrong\u003e5\u003c/strong\u003e:16449.\u003c/li\u003e\n \u003cli\u003eDomenga V, Fardoux P, Lacombe P, Monet M, Maciazek J, Krebs LT, Klonjkowski B, Berrou E, Mericskay M, Li Z\u003cem\u003e\u0026nbsp;et al\u003c/em\u003e: \u003cstrong\u003eNotch3 is required for arterial identity and maturation of vascular smooth muscle cells\u003c/strong\u003e. \u003cem\u003eGenes Dev\u0026nbsp;\u003c/em\u003e2004, \u003cstrong\u003e18\u003c/strong\u003e(22):2730-2735.\u003c/li\u003e\n \u003cli\u003eMehes G, Tzankov A, Hebeda K, Anagnostopoulos I, Krenacs L, Bedekovics J: \u003cstrong\u003ePlatelet-derived growth factor receptor beta (PDGFRbeta) immunohistochemistry highlights activated bone marrow stroma and is potentially predictive for fibrosis progression in prefibrotic myeloproliferative neoplasia\u003c/strong\u003e. \u003cem\u003eHistopathology\u0026nbsp;\u003c/em\u003e2015, \u003cstrong\u003e67\u003c/strong\u003e(5):617-624.\u003c/li\u003e\n \u003cli\u003eLindahl P, Johansson BR, Leveen P, Betsholtz C: \u003cstrong\u003ePericyte loss and microaneurysm formation in PDGF-B-deficient mice\u003c/strong\u003e. \u003cem\u003eScience\u0026nbsp;\u003c/em\u003e1997, \u003cstrong\u003e277\u003c/strong\u003e(5323):242-245.\u003c/li\u003e\n \u003cli\u003eIzquierdo E, Canete JD, Celis R, Santiago B, Usategui A, Sanmarti R, Del Rey MJ, Pablos JL: \u003cstrong\u003eImmature blood vessels in rheumatoid synovium are selectively depleted in response to anti-TNF therapy\u003c/strong\u003e. \u003cem\u003ePloS one\u0026nbsp;\u003c/em\u003e2009, \u003cstrong\u003e4\u003c/strong\u003e(12):e8131.\u003c/li\u003e\n \u003cli\u003eLobov IB, Brooks PC, Lang RA: \u003cstrong\u003eAngiopoietin-2 displays VEGF-dependent modulation of capillary structure and endothelial cell survival in vivo\u003c/strong\u003e. \u003cem\u003eProceedings of the National Academy of Sciences of the United States of America\u0026nbsp;\u003c/em\u003e2002, \u003cstrong\u003e99\u003c/strong\u003e(17):11205-11210.\u003c/li\u003e\n \u003cli\u003eZhao F, Hu Z, Li G, Liu M, Huang Q, Ai K, Cai X: \u003cstrong\u003eAngiogenesis in rheumatoid Arthritis: Pathological characterization, pathogenic mechanisms, and nano-targeted therapeutic strategies\u003c/strong\u003e. \u003cem\u003eBioact Mater\u0026nbsp;\u003c/em\u003e2025, \u003cstrong\u003e50\u003c/strong\u003e:603-639.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"arthritis-research-and-therapy","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"arrt","sideBox":"Learn more about [Arthritis Research \u0026 Therapy](http://arthritis-research.biomedcentral.com/)","snPcode":"13075","submissionUrl":"https://submission.nature.com/new-submission/13075/3","title":"Arthritis Research \u0026 Therapy","twitterHandle":"@ArthritisRes","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"rheumatoid arthritis, synovium, neovascularization, pericyte, endothelial cell, CD146/Mcam, pannus, single-cell RNA sequencing (scRNA-seq), bulk RNA sequencing (bulk RNA-seq), vascular remodeling and regeneration","lastPublishedDoi":"10.21203/rs.3.rs-9546936/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9546936/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eNeovascularization is a defining feature of rheumatoid arthritis (RA) pannus, yet the molecular signaling patterns that govern this process remain incompletely understood. Here, we sought to characterize CD146⁺ populations in inflammatory synovium, including endothelial cells, pericytes, and their involvement in neovascularization.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eUsing collagen-induced chronic polyarthritis model in D1BC transgenic mice, bulk and single-cell RNA sequencing of synovial cells was performed, followed by integrative analyses with published single-cell RNA-seq datasets from human rheumatoid arthritis and lung cancer.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eBulk RNA-seq revealed that CD146 expression stratifies endothelial and perivascular populations with angiogenic potential. Single-cell RNA sequencing of sorted CD146⁺ cells further classified endothelial subsets into \u003cem\u003eApln\u003c/em\u003e⁺, \u003cem\u003eAplnr\u003c/em\u003e⁺, \u003cem\u003eApln\u003c/em\u003e⁺/\u003cem\u003eAplnr\u003c/em\u003e⁺, and double-negative groups, closely mirroring the aCap/gCap dichotomy described in pulmonary capillary repair. Ligand-receptor and gene ontology analyses demonstrated that these CD146⁺ endothelial subgroups engage in neovascularization but lack signatures associated with gas exchange, supporting their identity as repair-type vasculature. A second CD146⁺ population expressed type-2 pericyte markers, including \u003cem\u003eCspg4\u003c/em\u003e, but lacked \u003cem\u003eActa2\u003c/em\u003e and \u003cem\u003ePdgfrb\u003c/em\u003e, suggesting a transcriptionally distinct pericyte subset that interacts preferentially with \u003cem\u003eApln\u003c/em\u003e⁺ and \u003cem\u003eAplnr\u003c/em\u003e⁺ endothelial cell subsets. Immunohistochemistry confirmed the spatial association of NG2⁺ pericytes with CD31⁺ neovessels in pannus tissue.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eTogether, these findings reveal that RA pannus neovascularization proceeds through a pulmonary-like capillary repair program orchestrated by CD146⁺ \u003cem\u003eApln\u003c/em\u003e⁺ lineage endothelial cells and NG2⁺ type-2 pericytes. This work establishes a unifying framework linking inflammation-driven angiogenesis with tissue repair vascular biology and identifies CD146⁺ subsets as potential targets for modulating pathological neovascularization in RA.\u003c/p\u003e","manuscriptTitle":"Neovascularization in rheumatoid arthritis pannus is orchestrated by endothelial-pericyte signaling","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-12 08:29:11","doi":"10.21203/rs.3.rs-9546936/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-18T14:35:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"256551933584583434936316925966865622166","date":"2026-05-07T19:42:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"232413398840753253897576892359406726684","date":"2026-05-04T10:44:54+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-04T07:55:35+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-04T02:59:00+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-04T02:58:43+00:00","index":"","fulltext":""},{"type":"submitted","content":"Arthritis Research \u0026 Therapy","date":"2026-04-28T01:26:58+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"arthritis-research-and-therapy","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"arrt","sideBox":"Learn more about [Arthritis Research \u0026 Therapy](http://arthritis-research.biomedcentral.com/)","snPcode":"13075","submissionUrl":"https://submission.nature.com/new-submission/13075/3","title":"Arthritis Research \u0026 Therapy","twitterHandle":"@ArthritisRes","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"59f496be-27dc-429c-bf5e-14465ddbcfe4","owner":[],"postedDate":"May 12th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-18T14:35:10+00:00","index":13,"fulltext":""},{"type":"reviewerAgreed","content":"256551933584583434936316925966865622166","date":"2026-05-07T19:42:09+00:00","index":12,"fulltext":""},{"type":"reviewerAgreed","content":"232413398840753253897576892359406726684","date":"2026-05-04T10:44:54+00:00","index":11,"fulltext":""},{"type":"reviewersInvited","content":"5","date":"2026-05-04T07:55:35+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-04T02:59:00+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-04T02:58:43+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-12T08:29:11+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-12 08:29:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9546936","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9546936","identity":"rs-9546936","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}