A method for identifying a diagnosis and treatment target of hemophilic arthritis, the target and application thereof

Abstract

This invention discloses a method, targets, and applications for the diagnosis and treatment of hemophilic arthritis. The method includes: obtaining cartilage and synovial membrane samples from patients with hemophilic arthritis undergoing total knee arthroplasty; performing single-cell RNA sequencing, single-cell ATAC sequencing, and spatial transcriptomics analysis to determine the main cell types and key cell subpopulations constituting the microenvironment of hemophilic arthritis; identifying the molecular characteristics and differentiation processes of chondrocyte and synovial fibroblast subpopulations; measuring intercellular interactions in the knee joint microenvironment of the samples; and identifying specific regulatory transcription factors in hemophilic arthritis through paired epigenetic analysis to identify diagnostic and therapeutic targets for hemophilic arthritis. This method integrates single-cell multi-omics analysis to analyze the cellular characteristics of synovial and cartilage tissues and their interactions in the epigenetic pathogenic mechanism, identifying potential therapeutic targets.

Description

Technical Field

[0001] This invention relates to the field of bioengineering technology, and in particular to a method for identifying diagnostic and therapeutic targets for hemophilic arthritis, the targets themselves, and their applications. Background Technology

[0002] Hemophilia is a sex chromosome-related congenital clotting disorder caused by abnormal or absent function of plasma glycoproteins. For example, a deficiency in coagulation factor VIII (FVIII) is associated with hemophilia A, while a deficiency in coagulation factor IX (FIX) can lead to hemophilia B (Christmas disease). Hemophilic arthritis (HA) is characterized by long-term, recurrent spontaneous intra-articular bleeding, which gradually leads to irreversible clinical changes, primarily manifested as joint pain, progressive ankylosis, muscle atrophy, and severe osteoporosis.

[0003] As one of the most serious joint diseases, hemophilia (HA) has a well-documented pathophysiology, but the immunopathological mechanisms associated with intra-articular inflammation remain incompletely understood. The pathophysiology of HA involves two characteristic processes: synovial inflammation and cartilage degeneration. These processes may occur simultaneously and influence each other, but are not necessarily interdependent. Recurrent intra-articular hemorrhage induces synovial thickening (proliferation of fibroblasts and macrophages within the synovium) and angiogenesis and remodeling (production of vascular endothelial growth factor). The iron-rich synovial layer secretes thrombin, matrix metalloproteinases (MMPs), and various inflammatory cytokines, such as tumor necrosis factor-α (TNFα), interferon-γ (IFNγ), interleukin (IL)-1β, and IL-6. These factors directly induce nuclear factor-κB ligand receptor activator (RANKL), indirectly leading to cartilage degeneration. Proliferative synovium promotes angiogenesis and remodeling, thus exacerbating hemorrhage in a vicious cycle. The cellular and molecular mechanisms of the knee joint microenvironment in hemophilia remain to be explored.

[0004] While single-cell RNA sequencing can resolve the heterogeneity of transcriptional states in knee joint synovial cells and chondrocytes at high resolution, it struggles to reveal the spatial localization of lesions and the epigenetic mechanisms driving pathological changes. Meanwhile, spatial transcriptomics (ST) technology can identify gene expression levels in specific tissue structures and integrate RNA sequencing data for spatial mapping. As a single-cell epigenomic sequencing technology, scATAC-seq (single-cell transposase-accessible chromatin sequencing) can capture chromatin accessibility signals in target cells and analyze upstream regulatory states. In recent years, single-cell multi-omics technology has emerged as a powerful tool for precisely analyzing regulatory states. This technology can simultaneously analyze the interaction between epigenetic accessibility and the transcriptome within the same cell, thereby revealing the association between chromatin regulatory elements and target gene expression. Summary of the Invention

[0005] Objective: To overcome the shortcomings of existing technologies, this invention provides a method for identifying diagnostic and therapeutic targets for hemophilic arthritis, the targets themselves, and their applications. By integrating single-cell RNA sequencing (scRNA-seq), single-cell ATAC sequencing (scATAC-seq), and spatial transcriptomics technologies, this invention reveals the cellular characteristics of synovium and cartilage in human osteoarthritis (OA) and hemophilic arthritis (HA) and their interactions in epigenetic pathogenic mechanisms.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] In a first aspect, the present invention provides a method for identifying diagnostic and therapeutic targets for hemophilic arthritis, the targets themselves, and their applications, including: Cartilage and synovial tissue samples were obtained from patients with hemophilic arthritis and osteoarthritis during total knee arthroplasty; the cartilage and synovial tissue samples from the osteoarthritis patients served as the control group. Digestion was performed to obtain single-cell suspensions of cartilage tissue and synovial tissue. Single-cell RNA sequencing, single-cell ATAC sequencing, and spatial transcriptomics analysis were performed to identify the main cell types and key cell subpopulations constituting the microenvironment of hemophilic arthritis; among which, the key cell subpopulations include chondrocyte subpopulations and synovial fibroblast subpopulations. The molecular characteristics and differentiation process of chondrocyte subsets and synovial fibroblast subsets were identified, and the molecular characteristics of chondrocyte subsets and the developmental trajectory of synovial fibroblasts were obtained. The study measured the intercellular interactions in the knee joint microenvironment of the samples and identified specific regulatory transcription factors in hemophilic arthritis through paired epigenetic analysis, thereby identifying diagnostic and therapeutic targets for hemophilic arthritis.

[0008] In some embodiments, the method for identifying key cell subpopulations constituting the microenvironment of hemophilic arthritis includes: Single-cell RNA sequencing datasets, single-cell ATAC sequencing datasets, and synovial cavity transcriptome datasets were constructed based on single-cell suspensions of cartilage tissue and synovial tissue. The cell selection criteria for the single-cell RNA sequencing datasets were: fewer than 200 detected genes, fewer than 350 total transcripts, and a mitochondrial gene ratio of less than 20%. Single-cell RNA sequencing was performed based on a single-cell RNA sequencing dataset to obtain the main cell types and gene expression abundance in cartilage and synovial tissues of patients with osteoarthritis and hemophilic arthritis. Single-cell ATAC sequencing was performed using a single-cell ATAC sequencing dataset to identify the types of transcription factors enriched in cartilage tissue.

[0009] In some embodiments, the process of identifying the types of transcription factors enriched in cartilage tissue further includes: Differential peak analysis was performed on cartilage and synovial tissues from patients with osteoarthritis and hemophilic arthritis based on the single-cell ATAC sequencing dataset. The results were then superimposed on differentially expressed gene data from the single-cell RNA sequencing dataset to identify hemophilic arthritis-specific genes and transcription factors, and to construct a hemophilic arthritis regulatory network. Peak detection and motif analysis were performed based on the single-cell ATAC sequencing dataset, and peak-gene association analysis was performed on cartilage and synovial tissue samples, integrating the single-cell ATAC sequencing dataset and the single-cell RNA sequencing dataset. Chromatin accessibility maps of representative signaling pathways were constructed using single-cell ATAC sequencing to identify specific epigenetic differences in the knee joint microenvironment of patients with hemophilic arthritis; among them, the representative signaling pathways included VISFATIN and ANGPTL.

[0010] In some embodiments, the method for identifying the molecular characteristics of the chondrocyte subsets includes: UMAP analysis was performed on chondrocytes to obtain discrete chondrocyte subsets; Exosome secretion analysis was performed on chondrocyte subsets to identify chondrocyte subsets with high exosome secretion, which are the key chondrocyte subsets in hemophilic arthritis. GO analysis of specific differentially expressed genes was performed on the key chondrocyte subsets in hemophilic arthritis to obtain their molecular characteristics.

[0011] In some embodiments, the method for identifying the developmental trajectory of the synovial fibroblasts includes: Analysis of synovial fibroblasts yielded a subset of synovial fibroblasts; wherein, the subset of synovial fibroblasts includes a lining cell subset and a sub-lining cell subset; Gene expression abundance of each synovial fibroblast subset was detected to identify differentially expressed genes in the synovial fibroblast subsets; GO analysis was used to identify the enrichment sites of differentially expressed genes in the synovial fibroblast subsets. Exosome secretion analysis of synovial fibroblasts was performed to identify cell communication among different synovial fibroblast subsets; The developmental trajectory of synovial fibroblasts was obtained by analyzing RNA velocity and monocle2 technology.

[0012] In some embodiments, the method for determining intercellular interactions in the sample knee joint microenvironment includes: Based on the spatial transcriptome dataset, the spatial distribution of heterogeneous expression in synovial tissue was analyzed using 10x Visium ST. Unsupervised clustering analysis and UMAP spatial barcode point visualization were performed to obtain the clustering results of marker genes in synovial tissue samples of hemophilic arthritis and osteoarthritis. By mapping the clustering results of marker genes to spatial regions related to hemophilic arthritis, the distribution of marker genes in the synovial tissue of patients with hemophilic arthritis was obtained.

[0013] In some embodiments, the method for determining intercellular interactions in the sample knee joint microenvironment includes: performing exosome secretion analysis on synovial fibroblasts to identify cell communication between different synovial fibroblast subpopulations; To assess the intercellular interactions of chondrocytes, synovial fibroblasts, and macrophages in the knee joint microenvironment of the sample and identify targeted signaling pathways.

[0014] In a second aspect, the present invention provides a diagnostic and therapeutic target for hemophilic arthritis, wherein the diagnostic and therapeutic target is identified using the method described in the first aspect; the diagnostic and therapeutic target is NAMPT and ANGPTL4. Activation of NAMPT and ANGPTL4 exacerbates symptoms of hemophilic arthritis; anti-NAMPT antibody improves the progression of hemophilic arthritis by inhibiting synovial inflammation, and anti-NAMPT antibody and anti-ANGPTL4 antibody inhibit cartilage degeneration by reducing chondroitin loss.

[0015] Thirdly, the present invention provides the use of the diagnostic and therapeutic targets as described in the second aspect in the preparation of products for treating hemophilic arthritis.

[0016] Beneficial Effects: This invention integrates single-cell RNA sequencing, scATAC-seq, and spatial transcriptomics analysis to determine the spatial distribution and epigenetic status of expression heterogeneity constituting the microenvironment of hemophilic arthritis; it identifies the molecular characteristics and differentiation processes of synovial cell and chondrocyte subsets, and further focuses on the developmental trajectory of synovial fibroblasts to reveal the pathological fibrosis repair mechanism; it explores the interactions of major intercellular signal exchange to identify potential targets in the vicious cycle of the synovial-cartilage axis. Simultaneously, through paired epigenetic analysis, it identifies decisive regulatory transcription factors in hemophilic arthritis. This in-depth analysis of the association between the transcriptome and epigenome not only deepens the understanding of underlying molecular mechanisms but also provides valuable therapeutic insights for a wider range of hemorrhagic joint diseases. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the microenvironmental heterogeneity of the knee joint in patients with hemophilic arthritis and the interaction between fibroblasts and chondrocytes in an embodiment of the present invention.

[0018] Figure 2 This is a single-cell transcriptomics and chromatin accessibility map of hemophilic arthritis and osteoarthritis in the embodiments of the present invention. Figure 2 Figure A shows a schematic diagram of the detection and subsequent analysis process of single-cell RNA sequencing (scRNA-seq) and single-cell ATAC sequencing (scATAC) in knee osteoarthritis. Figure 2 In Figure B, there are two UMAP maps showing all scRNA-seq cells in the synovial tissue sample stained by cell type, stacked bar charts showing the contribution of each sample to each cell type (two images on the left), and two UMAP maps showing all scATAC-seq cells in the synovial tissue sample stained by inferred cell type (two images on the right). Figure 2 The middle C is a violin diagram of key marker genes used to identify synovial cell types; Figure 2 The middle D shows the UMAP diagram of all scRNA-seq cells in the cartilage tissue sample stained by cell type, the stacked bar chart of each sample's contribution to each cell type (two images on the left), and the UMAP diagram of all scATAC-seq cells in the cartilage tissue sample stained by inferred cell type, the stacked bar chart of each sample's contribution to each cell type (two images on the right). Figure 2 E in the middle is a violin diagram of key marker genes used to identify chondrocyte types; Figure 2 The middle image (F) shows the histological evaluation results of tissue sections of OA and HA cartilage (right side image) and synovium (left side image).

[0019] Figure 3 The figure shows the experimental results of single-cell outline and functional characteristics analysis of chondrocytes in an embodiment of the present invention. Figure 3 In the middle A section, there is a UMAP diagram of eight chondrocyte subsets. Figure 3 Figure B shows the proportional distribution of different cell subpopulations in the sample and tissue. Figure 3 The middle C is a UMAP diagram showing the expression and distribution of typical marker genes in chondrocyte subsets; Figure 3 Figure D shows the heatmap of marker genes for each chondrocyte population and the results of GO term enrichment analysis. Figure 3 E represents the RNA velocity map of chondrocyte subsets; Figure 3 The figure in the middle (F) shows the statistical results of the number of exosomes released by each chondrocyte subset; Figure 3 G represents the UMAP map of exosomes released from cartilage tissue; Figure 3 The figure in H represents the immunohistochemical results of marker genes for osteoarthritis and healthy chondrocyte subsets; Figure 3 Image I in the middle is a transmission electron microscope (TEM) image of chondrocyte mitochondria.

[0020] Figure 4 This figure shows the experimental results of single-cell lineage and functional characteristics analysis of fibroblast subpopulations in synovial tissue in an embodiment of the present invention. Figure 4A in the middle is a UMAP diagram of seven fibroblast subsets; Figure 4 In the middle B, there is a stacked diagram showing the distribution ratio of different cell subpopulations in the sample tissue. Figure 4 The middle C is a UMAP diagram showing the expression and distribution of typical markers in fibroblast subsets; Figure 4 Figure D shows the heatmap of marker genes in each fibroblast cluster and the results of GO term enrichment analysis. Figure 4 E in the middle is a violin diagram of gene expression in samples of hemophilic arthritis (left group) and osteoarthritis (right group); Figure 4 The middle F represents the fibroblast differentiation trajectory diagram. The upper image is colored according to pseudo-time, and the lower image is colored according to clustering. Figure 4 G represents the RNA velocity map of a fibroblast subset; Figure 4 H represents a pseudo-time heatmap of various genes involved in fibroblast development.

[0021] Figure 5 This is a spatial histological diagram of synovial heterogeneity in hemophilic arthritis according to an embodiment of the present invention. Figure 5 Image A shows H&E-stained sections of HA and OA synovial samples used for Visium detection and spatial visualization of speckle clustering (top left). Spatial visualization analysis of the expression levels of PRG4, APOE, SFRP2, THY1, DKK3, and MDK in synovial samples was performed using the Visium detection method (top right). The ratio of fibroblasts to macrophages and spatial cell communication in the synovial spatial transcriptome is shown at the bottom. Figure 5 Image B is a multicolor immunofluorescence staining diagram showing the distribution of different fibroblast subsets in the synovial sample; Figure 5 In the middle, C is a live-cell imaging image of mouse synovial tissue.

[0022] Figure 6 This is a diagram showing the experimental results of signal transduction pathways related to the interaction between cartilage and synovium in an embodiment of the present invention. Figure 6 In the middle, A is a heatmap of the relative intensity of all enriched signals (outputs and inputs) in the HA cluster; Figure 6 B in the diagram represents the relative intensity heatmap of all enriched signals (outputs and inputs) in the OA cluster; Figure 6 The diagram in C represents the interaction relationships among the ANGPTL, Visfatin, and MK signaling pathways in all HA cell clusters. Figure 6 The diagram in middle D shows the interaction relationships among the ANGPTL, Visfatin, and MK signaling pathways in all OA cell clusters.

[0023] Figure 7 This is a single-cell epigenome map of hemophilia-related knee osteoarthritis in an embodiment of the present invention. Figure 7 In the middle, A is a UMAP diagram showing all chondrocytes color-coded according to cell subtype; Figure 7In the middle B section, a UMAP diagram showing color coding of all fibroblasts according to cell subtypes is displayed. Figure 7 The middle C is a row-scaled heatmap of statistically significant distal peak-gene associations in cartilage. Each row in the right-hand plot represents gene expression, and the left-hand plot shows the correlation of distal peak accessibility. Figure 7 The middle D is a row-scaled heatmap of statistically significant distal peak-gene associations in the synovium. Each row in the right-hand plot represents gene expression, and the left-hand plot shows the correlation of distal peak accessibility. Figure 7 E represents the transcription factor binding motif with differential peak enrichment between HA and OA in chondrocytes; Figure 7 F represents the transcription factor binding motif with differential peak enrichment between HA and OA in fibroblasts; Figure 7 G represents the normalized footprint of Tn5 insertion around the binding sites of FOS (left panel) and EGR1 (right panel) transcription factors in scATAC-seq chondrocytes. Figure 7 In the middle H, the normalized footprint of Tn5 insertion around the binding sites of SMARCC1 (left panel) and JUNB (right panel) transcription factors in scATAC-seq fibroblasts is shown.

[0024] Figure 8 This diagram shows the experimental results of determining the TF gene regulatory network of the chondrocyte-fibroblast axis in hemophilia in this embodiment of the invention. Figure 8 A represents the specific transcription factor-gene regulatory network diagram of chondrocytes in hemophilic arthritis; Figure 8 B in the diagram represents the specific transcription factor-gene regulatory network of synovial fibroblasts. Figure 8 The middle C is a violin diagram showing the RNA expression levels of specific transcription factors in chondrocytes; Figure 8 D represents the RNA expression level of specific transcription factors in fibroblasts. (E) Accessibility of peaks highly correlated with NAMPT gene expression. (F) Accessibility of peaks highly correlated with ANGPTL4 gene expression.

[0025] Figure 9 In this embodiment of the invention, inhibiting ANGPTL4 and NAMPT can alleviate FVIII. - / - Figure showing the experimental results of arthritis progression after joint bleeding in mice. Figure 9 In the middle, A represents FVIII after different antibody treatments. - / - Representative images of morphological changes in mouse synovium and cartilage: the upper part shows the macroscopic appearance, the middle part shows the H&E staining, and the lower part shows the safranin / fast green staining. Figure 9 B represents a quantitative analysis of synovitis in HA mouse models injected with different antibodies; Figure 9 The graph in C represents the quantitative analysis results of the OARSI scores (n=6). Detailed Implementation

[0026] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use.

[0027] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may include different values. It should be noted that similar reference numerals and letters in the following figures denote similar items; therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0028] The present invention will be further described below with reference to the embodiments.

[0029] All statistical analyses and charts were generated using R software (version 3.6.3) and GraphPad Prism software (version 8.0). Continuous variables are expressed as mean ± standard deviation. Differences between groups were analyzed using the Mann-Whitney U test (two-tailed test), and differences in cell frequencies within each cluster were calculated using the chi-square test. A p-value < 0.05 was considered statistically significant.

[0030] After obtaining informed consent from all participants, fresh cartilage and synovial tissue samples were collected from patients with osteoarthritis (OA) and hemophilic arthritis (HA) during total knee arthroplasty. Hemophilic arthritis samples were derived from hemophilia A patients lacking factor VIII (FVIII), including 5 cartilage tissue samples and 3 synovial tissue samples. Additionally, 3 cartilage tissue samples and 3 synovial tissue samples were collected from the medial non-weight-bearing region of osteoarthritis patients as a control group for single-cell RNA sequencing analysis. Synovial tissue from one HA patient and one OA patient was used for ST analysis.

[0031] Tissue isolation and preparation methods include: Immediately place the cartilage and synovial tissue fragments removed during surgery on an ice-cold surface. TMTissue samples were fixed in Singleton tissue preservation solution. After thorough washing with Hanke balanced salt solution (HBSS), the tissue samples were placed in sCelLive. TM In the tissue separation solution (Singleron), use Singleron PythoN. TM The tissue was digested using the separation system and incubated at 37°C for 15 minutes. Cell clumps and other large residual particles were removed through a 40-micron sterile filter to purify the cell suspension. After centrifugation, the cell pellet was treated with GEXSCOPE. ® Red blood cells were resuspended in erythrocyte lysis buffer (RCLB, Singleton) to remove red blood cells.

[0032] Example 1: Single-cell transcriptome sequencing analyzes the microenvironment map of hemophilic arthritis.

[0033] To characterize the multicellular microenvironment of the hemophilia knee joint at the single-cell level, such as... Figure 2 As shown in Figure A, the dataset integrates 14 single-cell RNA sequencing (scRNA-seq) datasets (from 5 cartilage tissue samples and 3 synovial tissue samples in the HA group; 3 cartilage tissue samples and 3 synovial tissue samples in the OA group), 8 single-cell ATAC sequencing (scATAC-seq) datasets (from 2 cartilage tissue samples and 2 synovial tissue samples in the HA group; 2 cartilage tissue samples and 2 synovial tissue samples in the OA group), and 2 synovial cavity transcriptome datasets (from 1 synovial tissue sample in the HA group and 1 synovial tissue sample in the OA group).

[0034] The concentration is approximately 2 × 10 5 Inject SCOPE-chip with single-cell suspension of PBS (HyClone) at a concentration of cells / mL. TM Microfluidic chip, operating using Singleton Matrix ® Single-cell processing system. mRNA captured by barcode microspheres is reverse transcribed into cDNA and then amplified. Subsequently, according to GEXSCOPE... ® Single-cell RNA sequencing libraries were constructed using the Singleton single-cell RNA library construction kit. After library pooling, the final libraries were sequenced at 150 bp paired ends using an Illumina novaseq 6000 sequencer.

[0035] In the R (v4.2.1) environment, Seurat (v5.0.2) was used to read and process the raw gene-cell expression matrices of the single-cell RNA sequencing dataset. The Read10X function was used to read each sample matrix, and the CreateSeuratObject function was used to generate Seurat objects (with parameters set to min.cells=1 and min.features=150). To remove low-quality cells, quality control criteria were used to select cells that met the following conditions: fewer than 200 detected genes, fewer than 350 total transcripts, and a mitochondrial gene proportion of less than 20%. The mitochondrial gene proportion was calculated using the PercentageFeatureSet function. Figure 2 As shown in Figures B and D, after filtering for low-quality data, 53,287 cartilage tissue cells and 90,608 synovial tissue cells were obtained for subsequent analysis.

[0036] like Figure 2 As shown in Figure C, based on gene expression of typical cell markers, nine major cell types were detected in synovial tissue, including endothelial cells, fibroblasts, parietal cells, proliferating cells, B cells, T cells and NK cells, mast cells, mononuclear phagocytes, and plasmacytoid dendritic cells (pDCs).

[0037] Four main cell types were detected in cartilage tissue: chondrocytes, endothelial cells (ECs), fibroblasts, and immune cells (such as...). Figure 2 As shown in E, such as myeloid cells, T cells, NK cells, and a small number of mast cells and B cells.

[0038] UMAP technology was used to study the gene expression abundance of major cell types in cartilage and synovial tissue samples, enabling two-dimensional visualization of cell clustering, such as... Figure 2 As shown in Figures C and E, combined with the analysis of the total cell proportions in each group of disease samples, significant differences were found between the cartilage and synovial tissues of patients with hemophilic arthritis. Compared with the osteoarthritis group, the hemophilic arthritis group showed more pronounced inflammatory-related cells and immune infiltration in both cartilage and synovial tissues.

[0039] To delve deeper into the upstream regulatory state of transcription, eight matched scATAC-seq data samples were analyzed. After removing batch effects, low-quality cells, and duplicate cells, single-cell RNA sequencing data processing methods were employed. Marker genes were extracted based on gene accessibility for dimensionality reduction and preliminary cluster analysis. Based on common transcriptional and gene expression characteristics, the single-cell RNA sequencing data were mapped to single-cell ATAC sequencing results.

[0040] like Figure 2As shown in B and D, the obtained cell subpopulations and their proportions highly overlap with the previous single-cell RNA clustering results, indicating that the single-cell ATAC sequencing data completely preserves all major cell type categories. Figure 2 As shown in Figure F, the staining results of the tissue specimens revealed that, compared with OA, the synovium of HA exhibited more pronounced hyperplasia and hemosiderin deposition. HA cartilage was characterized by a sharp decrease in thickness, obvious fissures, and disordered cell distribution.

[0041] Example 2: Analysis of Single-Cell Lineage and Functional Characteristics of Cartilage Tissue Cells

[0042] Single-cell transcriptomic profiles of chondrocytes were determined from eight cartilage tissue samples from five patients with hemophilic arthritis (n=18,842 chondrocytes) and three patients with osteoarthritis (n=18,949 chondrocytes). A total of 37,791 chondrocytes were obtained after strict quality control. The OA group samples were taken from the intact non-weight-bearing cartilage of the lateral femoral condyle and served as a normal control group. Based on known marker genes, the characteristics of each cartilage cluster were annotated, and clusters were considered as homologous cartilage subgroups based on typical marker genes. The cartilage subsets were identified as follows: regulatory chondrocytes (RegC) (CHI3L1, CHI3L2), fibrochondrocytes (FC) (COL1A1, COL1A2, S100A4, PRG4), proliferative chondrocytes (HTC) (COL10A1, IBSP, JUN), homeostatic chondrocytes (HomC) (FOS, JUN, MMP3, RGS16), repair chondrocytes (RepC) (CILP, CILP2, COL2A1), and pre-proliferative chondrocytes (preHTC) (SOX9, COL9A3).

[0043] Further UMAP analysis of chondrocytes yielded eight discrete chondrocyte subsets, named Cho1-Cho8 groups. Figure 3 The distribution of A in osteoarthritis (OA) and hemophilic joint (HA) differs significantly. Figure 3(Cho1, B). The Cho1 chondrocyte subset is highly homologous to effector chondrocytes (ECs) and highly expresses the genes CYTL1, TF, CLEC3A, FRZB, and CHAD. The high expression of CHI3L1 and CHI3L2 in the Cho2 subset indicates that they correspond to regulatory chondrocytes (RegCs). The Cho3 subset highly expresses genes related to extracellular matrix signaling and collagen fibrillary tissue, such as COL3A1, COMP, CILP2, and CILP, suggesting its repair capacity as a reparative chondrocyte (RepC). The Cho4 subset is more enriched in HA cartilage tissue, and its highly expressed genes are related to mitochondrial electron transport and hydrogen peroxide response, including MT-ND1 / 2 / 3 / 4 / 5 / 6, MT-ATP6 / 8, MT-CO1 / 2 / 3, PDE4D, and MALAT1. The Cho5 subset highly expresses proliferation markers such as IBSP, COL10A1, and SPP1 and is enriched in the HA cluster. The Cho6 subset corresponds to homeostatic chondrocytes (HomCs) and highly expresses the JUN, RGS16, and FOS genes, which regulate the cellular homeostatic response to external stimuli. The Cho8 subset highly expresses PRG4, S100A4, TMSB4X, and ribosomal protein-coding genes, which both promote chondrocyte proliferation and are associated with the vascular system. Figure 3 (C and D in the middle).

[0044] like Figure 3 As shown in Figure E, chondrocytes differentiate into fibrotic hypertrophy in hemophilic arthritis.

[0045] Compared with osteoarthritis chondrocytes, the proportion of exosomes released by the Ch4 and Ch5 subsets in the cartilage tissue of hemophilic arthritis is higher, indicating that exosomes secreted by these two specific subsets play a key role in the pathogenesis of hemophilic arthritis. Figure 3 (C1F and C2G). Based on the GO analysis results of specific differentially expressed genes, the Ch4 cluster was associated with cellular waste degradation, autophagy, and lysosomal-related activities; while the Ch5 cluster was enriched in extracellular matrix and structural support-related pathways.

[0046] The distribution of chondrocyte subsets in the aforementioned single-cell RNA sequencing study was characterized by immunohistochemical detection of highly expressed genes. For example... Figure 3 As shown in Figure H, HTCs are distributed throughout the entire stratum of HA cartilage tissue, which may stem from excessive damage to cartilage morphology; while in OA samples, HTCs are only distributed in the superficial and deep layers of cartilage tissue; mitochondrial-associated chondrocytes are expressed only in the superficial layer of OA cartilage tissue, but are widely overexpressed in HA cartilage. Figure 3 As shown in Figure I, transmission electron microscopy (TEM) confirmed that eosinophilic chondrocytes exhibit mitochondrial matrix thinning and vacuolation morphology.

[0047] Example 3: Analysis of Single-Cell Lineage and Functional Characteristics of Fibroblast Subpopulations in Synovial Tissue

[0048] Synovial fibroblasts regulate tissue homeostasis, coordinate inflammatory responses, and mediate tissue damage. Anatomically, the normal synovium consists of two layers: the basal layer and the lining layer, although the boundary between them is not clearly defined. Based on the expression characteristics of surface markers PRG4+ and THY1+ (CD90+), fibroid synovial cells (FLS) can be divided into the inner lining layer and the lower lining layer. In a healthy state, the inner lining layer of the synovium is in direct contact with the synovial fluid and is one to two cell layers thick. CX3CR1+ tissue-resident macrophages are distributed in the inner lining layer. They constitute the internal immunophysical barrier and secrete hyaluronic acid and lubricin.

[0049] Synovial fibroblasts were regrouped based on specific marker genes, identifying two lining cell subpopulations and five sub-lining cell subpopulations. Figure 4 (AC). Fb2 and Fb3 were both identified as lining synovial fibroblasts. Fb3 cells showed a higher tendency to express inflammatory chemokines (CXCL1, IRF1) and metabolic stress-related genes (KLF4 / KLF2, NR4A1 / 2). Sublining FLS cells in the Fb4 subgroup expressed lipid regulation-related genes (LRP1B, APOE, and APOC1) and signaling pathways regulating the cytoskeleton (KAZN, MALAT1). These genes may be highly expressed in synovium subjected to repeated hemorrhage stimulation, thereby driving synovial fibrosis and inflammatory responses. Fb1 sublaminar fibroblasts expressed genes such as CD34, C3, and MFAP5, which are associated with vascular structure in the perivascular interstitium beneath the lining, and also expressed genes related to immune inflammatory processes and matrix memory. Fb6 sublaminar fibroblasts play an important role in angiogenesis and remodeling. Fb7 sublining fibroblasts express genes associated with cartilage degradation (MDK, MMP2, and COL1), which are involved in the degradation of bone, cartilage, and extracellular matrix, and are upregulated during tissue repair and remodeling. Figure 4 (D).

[0050] Further investigation was conducted into the fibrotic and invasive behavior of myofibroblast-associated fibroblast (FLS) in hemophilic arthritis. Fibroblast activation protein-α (FAPα, encoded by the FAP gene) and FSP1 (encoded by the S100A4 gene) mark active fibroblasts or myofibroblast precursors, while αSMA (encoded by the ACTA2 gene) marks mature myofibroblasts. Subsynovial fibroblasts in hemophilic arthritis express higher levels of myofibroblast markers, including ACTA2, FAP, and COL1A1. Therefore, it is proposed that Fb4, Fb6, and Fb7 belong to myofibroblasts and are enriched during synovial fibrosis in hemophilic arthritis. Figure 4(E). This indicates that hemophiliac synovium involves a unique pathological fibrosis process, associated with lipid metabolism, oxidative stress, and cartilage degradation. Given the active extracellular matrix remodeling capacity of SFRP2+ synovial fibroblasts (Fb2), the abundance of genes related to matrix remodeling was analyzed, including multiple collagen genes (COL1A1 and COL1A2), proteoglycans (VCAN), and secreted growth factor (ANGPTL2).

[0051] GO analysis revealed that differentially expressed genes (DEGs) were primarily enriched in pathways related to extracellular matrix, collagen fibers, and wound repair. GO analysis also indicated that Fb4 is associated with collagen-containing extracellular matrix, chloride-coated endosome membranes, and amide binding. Genes expressing Fb2 were associated with phagocytosis, complement activation recognition, IgG immunoglobulin complex binding, and immunoglobulin receptor binding.

[0052] Exosome analysis was performed on synovial tissue to elucidate intercellular communication among subpopulations. Single-cell feature matrices of cartilage and synovium were constructed using the `seurat_to_genes` and `scmappr_and_pathway_analysis` functions from the `scmappr` package, respectively, to obtain scores for differentially expressed genes in specific cell clusters. Enrichment analysis was performed using the `comparecluster` function from the `clusterprofiler` package in R to compare the functional differences of specific genes in different cell clusters.

[0053] Healthy joint synovial tissue secretes significantly more exosomes than degenerative arthritis tissue, with Fb4 being the main secretory subset.

[0054] We combined RNA velocity analysis with Monocle2 technology to map the stemness and differentiation trajectories of synovial fibroblasts.

[0055] Methods: CytoTRACE v0.3.3 was used to predict developmental differentiation potential by calculating gene expression counts in individual cells from single-cell RNA sequencing data. RNA velocity, based on the relative abundance of newly generated (unspliced) and mature (spliced) mRNA, estimated gene splicing and degradation rates, thus enabling reliable prediction of single-cell developmental lineages and cell dynamics. BAM files for each sample were processed using Velocyto (v0.17.17) to generate LOOM files containing information on unspliced ​​and spliced ​​transcripts. Velocyto analysis used GENCODE v38 (gencode.v38.annotation.gtf) as a gene annotation reference and filtered using the repetitive sequence mask file hg38_rmsk.gtf to reduce interference from repetitive sequence regions. Each sample was processed independently, and the generated LOOM files were stored in a specified output directory according to the sample barcode index. Unless otherwise specified, all parameters used Velocyto default values. Subsequently, loompy (v3.0.8) was used to merge multiple sample loom files, with the "Accession" key ensuring the uniqueness of the merged sample information. The generated AnnData objects were then analyzed using scvelo (v0.3.3) and Scanpy (v1.11.0). First, data standardization and high-variability gene filtering were performed (scv.pp.filter_and_normalise, default parameters: min_shared_counts=20, n_top_genes=2000). Then, the neighboring cell matrix was calculated (scv.pp.moments, default parameters: n_pcs=30, n_neighbours=30), laying the foundation for subsequent velocity estimation. RNA velocity calculation used a randomized mode (scv.tl.velocity, mode="stochastic", other parameters default), and an inter-cell velocity graph was constructed using velocity_graph (scv.tl.velocity_graph, default parameter mode="stochastic").

[0056] Monocle2 (v2.30.1) describes cell developmental lineages and gene expression differences using a "pseudo-temporal" dimension. To correct for sequencing depth differences, estimateSizeFactors is used to estimate the size factor for each cell, while the estimateDispersions function is used to assess gene expression dispersion to identify highly variable genes. Subsequently, differential expression analysis is performed for each cell subpopulation using Seurat's FindAllMarkers. To achieve dimensionality reduction, the DDRTree method (reduceDimension, max_components = 2) is used to embed cells into a two-dimensional space. Then, pseudo-temporal ordering (orderCells) is applied to each cell type to generate pseudo-temporal trajectories. In pseudo-temporal differential gene analysis, significant genes (q-value < 0.01, |avg_log2FC| > 0.75) are screened, and the expression dynamics of genes on the pseudo-temporal axis are modeled using the differentialGeneTest function. Cluster analysis is performed on significant genes (num_clusters = 8), generating pseudo-temporal heatmaps and compiling a gene clustering table to record expression patterns.

[0057] Based on the CytoTrace score, the developmental trajectory was observed to primarily originate from the Fb2, Fb3 (lining fibroblasts), and Fb5 subsets (these subsets highly express S100A4), while fibrotic sublining FLS (Fb4, Fb6, and Fb7) occupied the endpoint of the developmental trajectory. The overall branching trajectory revealed the differentiation relationship from lining FLS to sublining FLS, accompanied by pathological fibrosis. Figure 4 The results of RNA velocity analysis showed a similar trend (in the middle F). Figure 4 (G).

[0058] Further investigation is needed into the regulatory functions of transcription factors during FLS development. For example... Figure 4 As shown in Figure H, the expression of cell cycle and proliferation regulatory genes (CCNE1, FOSL1) and transcription factor and epigenetic regulatory genes (SOX4, ZNF524) was upregulated, and the expression of metabolic and stress response genes (CRYM, NDUFAF1, TRIB1) and cell adhesion and migration related genes (FSCN1, LAMA5) also showed an upregulated trend.

[0059] Example 4: Specific spatial distribution of heterogeneous expression in synovial tissue

[0060] 10x Visium ST analysis was used to characterize local structures and microenvironments, elucidating the interactions between cells at different spatial locations. HA synovial tissue samples contained 2182 loci, 1210 median genes, and 2370 median UMIs. OA synovial tissue samples contained 2717 loci, 1185 median genes, and 2073 median UMIs. After unsupervised cluster analysis and UMAP spatial barcode visualization, HA and OA synovial tissue samples exhibited 9 and 3 independent clusters, respectively.

[0061] The association between hemophilia-related spatial regions and single-cell populations was analyzed to determine the distribution of major cell types in the synovium. Results showed that synovial hyperplasia was mainly characterized by macrophage infiltration, with enrichment of marker genes such as PRG4, HMOX1, and IFI30 in this region. In contrast, OA samples showed a homogeneous mixture of lining fibroblasts and a small number of macrophages, mainly enriched with marker genes COMP, A2M, MALAT1, CHI3L2, and SPARCL1. The HA synovial hyperplasia zone was mainly composed of sublining FLS cells and macrophages, while the OA synovium showed a homogeneous distribution. Figure 5 (A)

[0062] Furthermore, multicolor immunofluorescence (mIF) technology was used to characterize the heterogeneity of synovial cell subsets at the spatial level and to evaluate previously identified structural cell and fibroblast markers.

[0063] Mice were divided into three groups: a control group, an OA group, and a HA group, with six mice in each group. The control group consisted of eight-week-old male C57BL / 6J (wild-type) mice. The six mice in the OA group underwent anterior cruciate ligament transection four weeks prior. The HA group consisted of eight-week-old male F8 mice. - / - Mice were subjected to acupuncture model intervention.

[0064] One day prior to surgery, 25 μg of F4 / 80 (647) and 25 μg of PDGFRα (FITC) were injected via the tail vein. 25 μg of CD31 was injected again before imaging, and in vivo imaging was performed 2 hours later. Animals were anesthetized with tribromoethanol (20 ml / kg) via intraperitoneal injection. Hair was thoroughly removed from the legs and knee area of ​​mice using depilatory cream. The knee area was cleaned with iodine solution to prepare for surgical exposure. Mice were fixed in the observation mold. The skin was carefully incised using surgical scissors and forceps, and the connective tissue around the knee joint was separated. The observation area was wiped with PBS to fully expose the knee joint. The observation board was mounted on the observation site, ensuring a tight fit to the knee joint and maintaining a clear field of view. The instrument (IVM in vivo imaging system: IVIM Technology IVM-CMS3, Korea) was set to confocal mode and equipped with a 40x objective lens. The distance between the objective lens and the observation board was adjusted before imaging. The laser intensity was set to 20, the receiver intensity to 500, the average frame rate to 30, and the imaging depth to 200 μm.

[0065] like Figure 5 As shown in B and C, OA synovial samples exhibit typical characteristics of synovial hyperplasia and synovial vegetation formation. Increased inflammatory and immune-related cell types and proportions were also observed in the synovial hyperplasia region of HA, indicating a more complex microenvironment in hemophilic knee joints. SFPR2+ / THY1+ FLS cells in the deep epithelial synovium form discrete perivascular structures in the deep sublining layer, especially forming a distinct perivascular zone near lymphocyte aggregation areas. APOE+ FLSs are distributed among fibroblasts in the outer layer of the lining layer, possibly due to their initial exposure to oxidative stress in the joint microenvironment. Conversely, DKK3+ FLSs are widely distributed in the loose connective tissue of the lining layer. This suggests that the hyperplastic synovial region in HA is rich in fibrotic myofibroblasts of the lining layer, providing spatial proximity for their interaction with cartilage. Cell distribution in the synovium of hemophilic gene mice was observed using in vivo imaging techniques. Under normal conditions, fibroblasts and macrophages are sparsely and uniformly distributed; osteoarthritis (OA) is characterized by fibroblast proliferation. In the HA synovium, fibroblasts and macrophages proliferate significantly and form clusters.

[0066] Example 5: Cellular interaction between cartilage and synovium in the knee joint microenvironment of HA and OA

[0067] The development of hemophilic arthritis is a complex, multi-step process influenced by various factors. The importance of intercellular communication in disease evolution has received widespread attention. To elucidate the intercellular interactions within the knee joint microenvironment of osteoarthritis (OA) and hemophilic arthritis (HA) samples, we revealed potential intercellular interactions based on single-cell ligand-receptor gene expression.

[0068] The associations among chondrocytes, fibroblasts, and macrophages were assessed using the R package CellChat. All analyses were performed in an R (v4.2.1) environment using CellChat (v4.0.0). Seurat objects were created using the CreateSeuratObject() function (parameters: min.cells = 1, min.features = 150), normalized using NormaliseData() (method = 'LogNormalise', scale.factor = 10000), and 2000 highly variable genes were identified using FindVariableFeatures() (selection method = 'vst', number of features = 2000). Subsequently, the computeCommunProbPathway() function integrated the communication strength of each level of the signaling pathway, while aggregateNet() constructed the overall intercellular communication network. To assess signal flow and the strength of interactions between different cell types, netAnalysis_computeCentrality() calculates node centrality metrics (including degree centrality, proximity centrality, and betweenness centrality) to describe the criticality of signals in the network.

[0069] Using heatmap visualization technology, the incoming and outgoing signal networks are presented. Figure 6 (A and B). Ch1, Ch2, and Ch5 were observed to primarily act as receptor cells, while Fb1 and Fb4 primarily acted as sender cells. Notably, in hemophilic knee osteoarthritis, the TGFβ, ANGPTL, MK, MIF, CXCL, FGF, and CCL signaling pathways exhibited stronger and more abundant interactions. Figure 6 (A). In the ANGPTL signaling pathway, ANGPTL2 / 4 are the major ligands in cartilage, while ITGA5, ITGB1, SDC2, and SDC4 are the major receptors. This pathway suggests the process of vascular endothelial injury and regeneration, and also drives synovial fibrosis. Fibroblasts were found to influence chondrocyte type primarily through the MDK-(LRP1 + SDC2) ligand-receptor pair. Fibroblasts were found to play a major sending cell role in the MDK pathway. Figure 6 (C)

[0070] Analysis of MDK signaling pathway gene expression across all cell types revealed that MDK expression was highest in fibroblasts, while its receptor genes were almost unexpressed in the synovium. This indicates that fibroblasts primarily interact with other cells through the MDK signaling pathway, with minimal impact on themselves. Therefore, it is speculated that fibroblasts may influence chondrocyte function by secreting MDK. MDK may promote fibroblast proliferation and extracellular matrix (ECM) deposition, accelerating the fibrotic repair process of joints after hemorrhage. In the VISFATIN signaling pathway, NAMPT is the major ligand for chondrocytes type 1 and 3, while ITGA5, ITGB1, and INSR are the major receptors for fibroblasts type 1, 2, and 4, respectively. Figure 6 (CD). This suggests that abnormal energy metabolism in cartilage tissue may drive an inflammatory state in fibroblasts.

[0071] In summary, the data indicate that a vicious positive feedback loop exists between the states of chondrocytes and fibroblasts.

[0072] Example 6: Chromatin Accessibility Analysis in Hemophilic Arthritis

[0073] To further explore the role of key upstream transcription factors in the pathogenesis of hemophilic arthritis, single-cell ATAC sequencing analysis was performed on chromatin accessibility lineages of hemophilic arthritis and osteoarthritis patient samples.

[0074] Single-cell ATAC-seq library construction: First, collected tissue and lysis buffer were added to a 1.5 ml centrifuge tube and homogenized manually 15 times using a Dounce mortar and pestle. Then, ATAC nuclear separation buffer was added and mixed, followed by double filtration through 70 μm and 40 μm cell filters to remove impurities. The nuclei were precipitated by centrifugation at 4°C and 500 × g for 5 minutes and resuspended in pre-chilled PBS. Utilizing the Tn5 transposase properties, fragments containing sequencing ligation sequences were introduced into chromatin-accessible regions. After transposition, the products were captured using scATAC magnetic beads. ATAC library construction was completed after PCR amplification, product purification, and Qubit quality testing.

[0075] scATAC processing and cluster analysis: The `createArrowFiles()` function generates Arrow files with parameters `minTSS = 4` and `minFrags = 1000` to ensure basic quality control for transcription start site enrichment and fragment counting. This function also excludes sex chromosome and mitochondrial genome sequences (`excludeChr = c('chrM', "chrY", 'chrX')`). Duplex sequence scores are calculated using `addDoubletScores()`, and a UMAP-based k-nearest neighbor algorithm (`k=20`, `nTrials=5`) is used to identify potential duplex sequences. After duplex sequence filtering, an `ArchRProject` object is constructed for subsequent dimensionality reduction and cluster analysis. In the quality control phase, `plotFragmentSizes()` and `plotTSSEnrichment()` are used to evaluate fragment length distribution and transcription start site enrichment, generating sample-level quality control visualizations. Based on the feature matrix of TileMatrix, iterative latent semantic indexing (Iterative LSI) dimensionality reduction was performed using the addIterativeLSI() function (4 iterations, 50,000 features, retaining the first 20 dimensions). To mitigate batch effects among samples, Harmony batch correction was further performed using the addHarmony() function (with Sample as the grouping variable). Cluster analysis was performed using the Seurat-based addClusters() function with a resolution of 1. The results were visualized as a confusion matrix heatmap, showing the distribution relationship between samples and clusters. Subsequently, based on the Harmony and LSI dimensionality reduction results, UMAP and t-SNE (addUMAP() and addTSNE()) were used for visualization. To identify the marker genes for each cluster, the Wilcoxon rank-sum test was performed on GeneScoreMatrix using getMarkerFeatures(). Genes were screened using a threshold of FDR ≤ 0.05 and Log2FC ≥ 1, and a list of marker genes for each cluster was output. To achieve multi-omics integration, scATAC-seq data was mapped and integrated with matched single-cell transcriptomes (Seurat objects seRNA) using addGeneIntegrationMatrix(). Based on previously clustered and annotated chondrocyte single-cell ATAC-seq data, population-based peak detection and motif enrichment analysis were performed using ArchR (v1.0.2) in the R environment. This method aims to further elucidate the differences in chromatin accessibility regions and transcription factor binding patterns in chondrocytes under different pathological conditions.

[0076] Before peak detection, a pseudo-batch coverage file was generated using the `addGroupCoverages()` function based on the "group" parameter to ensure sufficient signal for subsequent peak identification in all cell subpopulations. The genome reference version used was BSgenome.Hsapiens.UCSC.hg38, and peak localization was performed using MACS2 (v2.2.7.1). A reproducible peak set was generated within each Cgroup using the `addReproduciblePeakSet()` function, with the default parameter `groupBy="group"` and `pathToMacs2` pointing to a predefined path. Two pseudo-reproducible samples were generated for each subpopulation, with each group containing 40-500 cells and a peak count capped at 150,000. The generated peak sets were saved and loaded as a new object (projHeme4), and then a peak matrix was constructed using `addPeakMatrix()` for differential analysis.

[0077] Transcription factor binding site annotation was performed using the `addMotifAnnotations()` function, integrating motif annotations from the cis-BP database (motifSet='cisbp'). After generating the peak-motif correspondence matrix, the genomic locations (sequence name, start site, and end site) of all peaks were extracted and saved as a `pSet_peaks.tsv` file for subsequent motif matching verification. The motif matching matrix was obtained using the `getMatches()` function. Using the `GenomicRanges` package, local overlap analysis was performed on specific chromosomal regions (such as multiple segments of chromosome 11) to identify potentially binding transcription factors within differentially open regions, including TFAP2C, KLF4, EGR1, and SP1. For motif enrichment analysis of differentially open peaks, the `peakAnnoEnrichment()` function was used to detect the enrichment of transcription factors in upregulated peaks (Log2FC ≥ 0.5) and downregulated peaks (Log2FC ≤ -0.5), using a uniform threshold of FDR ≤ 0.1. Enrichment results were sorted in descending order of -log10(FDR) value. Scatter plots and annotation plots of the top 30 enriched transcription factors were generated using ggplot2 (v3.5.1). The enriched transcription factor data were saved as motifsUp.tsv and motifsDo.tsv files, and corresponding visualization plots were exported to show the global trend of motif enrichment in the HA and OA groups. For key transcription factors, plotGroups() was used to display the motif bias score between the HA and OA groups, and plotBrowserTrack() was used to generate corresponding accessibility browser track plots. The final results were uniformly output using plotPDF() to adapt to the visualization needs of different analysis levels.

[0078] Cartilage and synovial membrane samples were analyzed using single-cell ATAC-seq with ArchR software. Cell types were located and annotated based on single-cell RNA sequencing data. Six cell types were annotated for chondrocytes, and five cell types were annotated for synovial fibroblasts. Figure 7 (Chinese AD).

[0079] Transcription factors enriched in HA cartilage: The FOS family (FOSL1, FOSL2, JUNB, JUND) are widely involved in stress response, cell differentiation, and cartilage repair. In hemophilia patients, upregulation of these transcription factors may be involved in inflammation regulation and cartilage regeneration during cartilage repair. The BACH family (BACH1, BACH2) is associated with oxidative stress and anti-inflammatory responses. Hemophilia patients experience elevated oxidative stress levels due to repeated bleeding injuries, and upregulation of the BACH family may help address these stresses. In contrast, OA cartilage-specific transcription factors exhibit distinctly different functional characteristics. PAX family transcription factors (PAX3, PAX5) are widely involved in cartilage development and stem cell maintenance. In osteoarthritis, downregulation of these factors may indicate weakened chondrocyte differentiation and repair capacity. MLL family transcription factors are involved in pluripotent stem cell differentiation and gene rearrangement. Downregulation of MLL in osteoarthritis may lead to decreased cartilage repair capacity. ZBTB4 is associated with cell proliferation and immune responses. In osteoarthritis, its downregulation may affect cartilage repair and cell regeneration. Figure 7 (E).

[0080] Hyaline cartilage synovial tissue is mainly enriched with FOS family (FOS, FOSL1, FOSL2, FOSB), JUN family (JUN, JUNB, JUND), RUNX family (RUNX1, RUNX2, RUNX3) and NFE2 family (NFE2, NFE2L2) transcription factors, which are involved in acute injury repair, inflammatory response and tissue fibrosis.

[0081] Repeated bleeding and tissue damage caused by hemophilia lead to activation of synovial fibroblasts, which in turn initiate repair mechanisms, fibrosis, and immune regulation by upregulating the expression of these factors. NFIC and NFIL3 are often associated with immune responses and chronic inflammation in osteoarthritis. In osteoarthritis, these factors may participate in chronic inflammatory responses and articular cartilage degeneration by regulating the immune function of synovial fibroblasts. Transcription factors enriched in synovial fibroblasts are mainly associated with chronic inflammation, immune regulation, cartilage degeneration, and repair processes. Figure 7 Unlike acute injury repair in HA, transcription factors enriched in OA are more closely associated with long-term chronic inflammation, immune system regulation, and degeneration and remodeling of cartilage and synovium.

[0082] Example 7: Analysis of Specific Regulatory Networks

[0083] Based on scATAC-seq data from cartilage and synovial tissues, differential peak analysis was performed on HA and OA, and the results were superimposed on scRNA-seq differentially expressed gene data. If a gene is upregulated in HA and present in the scATAC-seq differential peaks, it is defined as an HA-specific gene; if a transcription factor is upregulated in HA and its corresponding motif is enriched in the differential peaks, it is defined as an HA-specific transcription factor. By identifying HA-specific target genes and transcription factors using the above methods, a HA regulatory network specific to cartilage and synovial tissues was constructed.

[0084] To elucidate the differences in chromatin accessibility and transcriptional regulatory characteristics of chondrocytes and synovial fibroblasts in patients with hemophilic arthritis (HA) and osteoarthritis (OA), multilayer ensemble analysis was conducted based on single-cell ATAC-seq and single-cell RNA-seq data to construct an HA-specific regulatory network. First, for each tissue type, differentially accessible peaks calculated by ArchR (v1.0.2) and differentially expressed genes (DEGs) identified by Seurat analysis were extracted. Co-localized genes upregulated in both omics datasets were screened, centered on the HA upregulation signal. Specifically, the differential annotation peak files (including promoter region annotation genes) processed by ChIPseeker and the corresponding single-cell transcriptome differential expression results files were read. Gene names were uniformly converted to uppercase and spaces removed to ensure cross-omics matching accuracy. Subsequently, gene sets significantly upregulated in single-cell RNA sequencing analysis and annotated as "HA" clusters were identified. The intersection of this gene set with the differentially accessible peak annotated genes was performed to obtain a set of HA group-specific candidate target genes with co-expression and co-accessibility characteristics.

[0085] To further characterize the cis-regulatory elements of these co-localized genes, differential peak regions were extracted, and standard BED files were generated for subsequent sequence analysis. Peak region sequences were extracted from the human reference genome hg38 using bedtools (v2.30.0), and motif scans were performed on the differential peak regions of cartilage and synovial samples. Pattern enrichment analysis was performed using the FIMO tool (v5.3.0) in the MEME suite in conjunction with the JASPAR2024 pattern database (combined_matrices_445470.meme). Default scan parameters included the FIMO built-in background model, single-sequence pattern scanning, and no parallelization. All FIMO outputs were then merged and filtered, retaining only significant sites and quantifying motif frequencies to derive the potential transcription factor enrichment information within each differentially enriched peak region. At the integration level, the motif enrichment results were cross-validated using single-cell RNA sequencing differential expression data, identifying transcription factors upregulated in the HA group and exhibiting significant motif enrichment in open chromatin regions, which were defined as HA-specific regulatory factors. Simultaneously, downstream genes present in the differential peaks and upregulated in the HA group were defined as HA-specific target genes. This strategy systematically identified potential disease-specific transcriptional regulatory relationships within cartilage and synovial tissues.

[0086] In the network visualization and construction process, the R packages igraph (v2.0.2) and ggraph (v2.2.1) were used to model and draw the regulatory relationships between transcription factors (TFs) and target genes. First, an edge relationship matrix (TF-Gene) was generated based on the intersection of motif analysis and differentially expressed genes, and a node information table was created, marking transcription factors and target genes as different node types. A directed network graph was constructed using ggraph's force-directed layout algorithm (layout="kk"), with arrows indicating the regulatory direction (default arrow type="closed", angle=15, length=0.1 inches). Node colors distinguished functional categories: red represented ordinary nodes, and blue represented high-confidence key regulatory factors or target genes. The node size was set to 2 for transcription factors and 1 for target genes, which could be adjusted using scale_size_continuous(range=c(2,4)). Transcription factors were represented by squares (shape=22), and target genes by circles (shape=21). The final network graph was exported as an image file for storage.

[0087] Based on scRNA expression matrices and transcription factors from the AnimalTFDB23 database, pyscenic (v0.11.0) and the SCENIC R toolkit were used to construct transcription factor networks from synovial and cartilage samples. Both GENIE3 and GRNBoost2 employed regression tree methods to output regulatory networks based on the co-expression patterns of regulatory factors and direct / indirect target genes. First, GRNBoost2 predicted the regulatory network based on the co-expression relationships between regulatory factors and target genes. Then, the AUCell method was used to quantify the activity of regulatory factors in specific cells from single-cell RNA sequencing data.

[0088] like Figure 8 As shown in the AD diagram, in hemophilic arthritis, JUNB and FOS may promote cartilage degradation and inflammatory responses by activating the expression of inflammatory mediators and cytokines. JUNB and FOS, as members of the AP-1 transcription factor family, participate in the regulation of inflammatory responses, cell proliferation, and apoptosis. In hemophilic arthritis, the upregulation of KLF10 may be associated with chondrocyte apoptosis and cartilage matrix degradation, participating in the regulation of multiple pathways such as TGF-β and Wnt / β-catenin. In hemophilic arthritis, EGR1 may play a role by promoting inflammatory responses and repairing cell damage. EGR1 exacerbates joint inflammation by activating various inflammatory factors such as IL-1β and TNF-α, thereby promoting cartilage degradation and pathological changes. Although EGR1 also promotes degenerative changes in cartilage in osteoarthritis, its role in osteoarthritis is more focused on regulating chronic cartilage degradation and repair processes compared to hemophilic arthritis. Single-cell RNA sequencing (SCENIC) analysis of chondrocytes also identified the specific transcription factors JUNB, FOS, KLF10, and EGR1. Repeated hemorrhage leads to iron deposition and oxidative stress in the synovial microenvironment, which can induce CEBPD upregulation, promoting the production of large amounts of IL-6, IL-8, and CXCL8 by fibroblasts, thereby enhancing the synovial inflammatory response. This stimulates fibroblast proliferation, leading to synovial tissue thickening and fibrosis. Simultaneously, this process promotes the secretion of matrix metalloproteinases (MMPs) from fibroblasts into the extracellular matrix (ECM), resulting in cartilage degradation. STAT1 in HA fibroblasts promotes the expression of CXCL10 and IL-6 by fibroblasts through the IFN signaling pathway, enhancing the inflammatory response and activating synovial fibroblasts. Single-cell RNA sequencing (scRNA-seq) analysis of synovial fibroblasts by SCENIC also identified CEBPD and STAT1 as specific transcription factors.

[0089] Peak detection and motif analysis were performed based on scATAC data, and peak-gene association analysis was conducted on cartilage and synovial samples using the Circeo tool in ArchR, integrating scATAC and scRNA data. Combining this with the vicious cycle observed in cartilage-synovial axis cell communication, chromatin accessibility maps of representative signaling pathways such as VISFATIN and ANGPTL were further explored using scATAC sequencing to elucidate specific epigenetic differences in the hemophilic arthritis microenvironment. As shown in Figures E and F, the ligand NAMPT in chondrocytes and the ligand ANGPTL4 in fibroblasts were both significantly activated in hemophilic arthritis, driving their transcription and significantly upregulating their expression.

[0090] Example 8: Inhibition of ANGPTL4 and NAMPT can alleviate FVIII - / - Progress in a mouse model of hemophilic arthritis induced by acupuncture

[0091] Data from Example 7 indicate that NAMPT and ANGPTL4 play an exacerbating role in the pathogenesis of hemophilic arthritis. Further research is needed to determine whether blocking their function with corresponding antibodies can alleviate the progression of hemophilic arthritis.

[0092] FVIII gene knockout hemophilia (FVIII) - / - Mice were purchased from the Jackson Laboratory in Bar Harbor, Maine. Animal experiments for this study were approved by the Animal Use Committee of the First Affiliated Hospital of Soochow University. All mice were housed in an SPF-grade animal laboratory.

[0093] In this embodiment, mice were divided into a disease-free model group (Sham), an HA model group (HA), a NAMPT treatment group (HA+Anti-NAMPT), an ANGPTL4 treatment group (HA+Anti-ANGPTL4), and an ANGPTL4 and NAMPT combined treatment group (HA+Anti-NAMPT+Anti-ANGPTL4), with 8 mice in each group.

[0094] When establishing the mouse HA model, 3-month-old male FVIII mice were used. - / - Mice underwent acupuncture at the right knee. On day 0, a 30G needle was used to puncture the right knee to induce bleeding.

[0095] Treatment regimen: Every two weeks following acupuncture, mice were injected intra-articularly into the right knee joint cavity with NAMPT (A22044, ABclonal), ANGPTL4 (A2011, ABclonal), or a mixture of both antibodies (200 ng each, injection volume 5 μL). The disease-free model group was injected with PBS. All mice were sacrificed at week 8 after acupuncture, and tissue samples from the right knee joint were collected for histological analysis and subsequent studies.

[0096] The synovial and cartilage tissue samples removed during surgery were immediately placed in 4% paraformaldehyde fixative. After paraffin embedding, 4 μm thick sections were cut. H&E staining was used to reveal the morphological characteristics of the cartilage and synovial tissues, and safranin / fast green staining was used to detect the degree of cartilage degeneration. Imaging scans were performed using a PANNORAMIC MIDI (3DHISTECH) scanner.

[0097] like Figure 9 As shown in Figure A, acupuncture-induced knee hemorrhage in mice manifested as significant swelling and bleeding, and the symptoms were significantly relieved after treatment with a combination of NAMPT and ANGPTL4 antibodies. Figure 9 As shown in Figures B and C, anti-NAMPT antibody treatment significantly inhibited synovial inflammation in mice, resulting in a marked decrease in synovitis scores. Following combined treatment with NAMPT and ANGPTL4 dual-target antibodies, as... Figure 9 Safranin A / Fast Green staining showed reduced loss of chondroitin proteoglycans and inhibited cartilage degeneration.

[0098] In summary, these findings suggest that intra-articular injection of anti-ANGPTL4 and anti-NAMPT antibodies can alleviate synovial inflammation and cartilage degeneration, thereby improving the progression of hemophilic arthritis.

[0099] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying diagnostic and therapeutic targets in hemophilic arthritis, characterized in that, include: Cartilage and synovial tissue samples were obtained from patients with hemophilic arthritis and osteoarthritis during total knee arthroplasty; the cartilage and synovial tissue samples from the osteoarthritis patients served as the control group. Digestion was performed to obtain single-cell suspensions of cartilage tissue and synovial tissue. Single-cell RNA sequencing, single-cell ATAC sequencing, and spatial transcriptomics analysis were performed to identify the main cell types and key cell subpopulations constituting the microenvironment of hemophilic arthritis; among which, the key cell subpopulations include chondrocyte subpopulations and synovial fibroblast subpopulations. The molecular characteristics and differentiation process of chondrocyte subsets and synovial fibroblast subsets were identified, and the molecular characteristics of chondrocyte subsets and the developmental trajectory of synovial fibroblasts were obtained. The study measured the intercellular interactions in the knee joint microenvironment of the samples and identified specific regulatory transcription factors in hemophilic arthritis through paired epigenetic analysis, thereby identifying diagnostic and therapeutic targets for hemophilic arthritis.

2. The method for identifying diagnostic and therapeutic targets for hemophilic arthritis according to claim 1, characterized in that, The method for identifying key cell subsets that constitute the microenvironment of hemophilic arthritis includes: Single-cell RNA sequencing datasets, single-cell ATAC sequencing datasets, and synovial cavity transcriptome datasets were constructed based on single-cell suspensions of cartilage tissue and synovial tissue. The cell selection criteria for the single-cell RNA sequencing datasets were: fewer than 200 detected genes, fewer than 350 total transcripts, and a mitochondrial gene ratio of less than 20%. Single-cell RNA sequencing was performed based on a single-cell RNA sequencing dataset to obtain the main cell types and gene expression abundance in cartilage and synovial tissues of patients with osteoarthritis and hemophilic arthritis. Single-cell ATAC sequencing was performed using a single-cell ATAC sequencing dataset to identify the types of transcription factors enriched in cartilage tissue.

3. The method for identifying diagnostic and therapeutic targets for hemophilic arthritis according to claim 2, characterized in that, The process of identifying the types of transcription factors enriched in cartilage tissue also includes: Differential peak analysis was performed on cartilage and synovial tissues from patients with osteoarthritis and hemophilic arthritis based on the single-cell ATAC sequencing dataset. The results were then superimposed on differentially expressed gene data from the single-cell RNA sequencing dataset to identify hemophilic arthritis-specific genes and transcription factors, and to construct a hemophilic arthritis regulatory network. Peak detection and motif analysis were performed based on the single-cell ATAC sequencing dataset, and peak-gene association analysis was performed on cartilage and synovial tissue samples, integrating the single-cell ATAC sequencing dataset and the single-cell RNA sequencing dataset. Chromatin accessibility maps of representative signaling pathways were constructed using single-cell ATAC sequencing to identify specific epigenetic differences in the knee joint microenvironment of patients with hemophilic arthritis; among them, the representative signaling pathways included VISFATIN and ANGPTL.

4. The method for identifying diagnostic and therapeutic targets for hemophilic arthritis according to claim 1, characterized in that, The methods for identifying the molecular characteristics of the chondrocyte subsets include: UMAP analysis was performed on chondrocytes to obtain discrete chondrocyte subsets; Exosome secretion analysis was performed on chondrocyte subsets to identify chondrocyte subsets with high exosome secretion, which are the key chondrocyte subsets in hemophilic arthritis. GO analysis of specific differentially expressed genes was performed on the key chondrocyte subsets in hemophilic arthritis to obtain their molecular characteristics.

5. The method for identifying diagnostic and therapeutic targets for hemophilic arthritis according to claim 1 or 4, characterized in that, The methods for identifying the developmental trajectory of the synovial fibroblasts include: Analysis of synovial fibroblasts yielded a subset of synovial fibroblasts; wherein, the subset of synovial fibroblasts includes a lining cell subset and a sub-lining cell subset; Gene expression abundance of each synovial fibroblast subset was detected to identify differentially expressed genes in the synovial fibroblast subsets; GO analysis was used to identify the enrichment sites of differentially expressed genes in the synovial fibroblast subsets. Exosome secretion analysis of synovial fibroblasts was performed to identify cell communication among different synovial fibroblast subsets; The developmental trajectory of synovial fibroblasts was obtained by analyzing RNA velocity and monocle2 technology.

6. The method for identifying diagnostic and therapeutic targets for hemophilic arthritis according to claim 1, characterized in that, The method for determining intercellular interactions in the knee joint microenvironment of the sample includes: Based on the spatial transcriptome dataset, the spatial distribution of heterogeneous expression in synovial tissue was analyzed using 10x Visium ST. Unsupervised clustering analysis and UMAP spatial barcode point visualization were performed to obtain the clustering results of marker genes in synovial tissue samples of hemophilic arthritis and osteoarthritis. By mapping the clustering results of marker genes to spatial regions related to hemophilic arthritis, the distribution of marker genes in the synovial tissue of patients with hemophilic arthritis was obtained.

7. The method for identifying diagnostic and therapeutic targets for hemophilic arthritis according to claim 1, characterized in that, The method for determining the intercellular interactions in the knee joint microenvironment of the sample includes: performing exosome secretion analysis on synovial fibroblasts to identify cell communication between different synovial fibroblast subpopulations; To assess the intercellular interactions of chondrocytes, synovial fibroblasts, and macrophages in the knee joint microenvironment of the sample and identify targeted signaling pathways.

8. A diagnostic and therapeutic target for hemophilic arthritis, characterized in that, The diagnostic and therapeutic targets are identified using the method described in any one of claims 1-7; The diagnostic and therapeutic targets are NAMPT and ANGPTL4; Activation of NAMPT and ANGPTL4 exacerbates symptoms of hemophilic arthritis; anti-NAMPT antibody improves the progression of hemophilic arthritis by inhibiting synovial inflammation, and anti-NAMPT antibody and anti-ANGPTL4 antibody inhibit cartilage degeneration by reducing chondroitin loss.

9. The use of the diagnostic and therapeutic targets as described in claim 8 in the preparation of products for treating hemophilic arthritis.

Images