From cells to clinic: Single-cell transcriptomics shaping the future of orthopedics

Abstract

Single-cell RNA sequencing (scRNA-seq) technology hold significant potential for advancing orthopedic research. This review examines the impact of ScRNA-seq on the future development of orthopedic research and practice. In the study of osteoarthritis, scRNA-seq can finely characterize the changes in the subsets of chondrocytes and their role in disease progression. In rheumatoid arthritis, this technique reveals the complex heterogeneity and cell-to-cell interactions between fibroblasts and immune cells. ScRNA-seq offers insights into the heterogeneity of nucleus pulposus, annulus fibrosus, and endplate cells, providing a novel perspective on the pathological mechanisms of intervertebral disc degeneration. Single-cell analysis in osteosarcoma research has uncovered the complexity of the tumor microenvironment and mechanisms of immunosuppression. Through these studies, scRNA-seq enhances insights into disease pathogenesis and offers innoviate approaches for precision medicine and personalized treatment strategies.

The Translational Potential of this Article

This article systematically reviews the cellular heterogeneity, molecular mechanisms and immune microenvironment of orthopedic diseases (such as osteoarthritis, rheumatoid arthritis, intervertebral disc degeneration, osteosarcoma) by single-cell RNA sequencing (scRNA-seq), which provides a theoretical basis for accurate diagnosis, new therapeutic target discovery (such as TRPV1, CXCR4) and individualized treatment strategies. The combination of multi-omics and spatial transcriptome technology is expected to accelerate clinical translation and optimize the diagnosis and treatment system of orthopedic diseases.

Graphical abstract

1. Introduction

Orthopedic disorders affect the health of thousands of people and bring great stress to patients and their families. These conditions can severely restrict daily activities, and, in extreme cases, result in long-term disability, impacting the patient's work and social life. Individuals with osteoarthritis (OA) or rheumatoid arthritis (RA) may experience impaired mobility due to joint pain and stiffness, potentially requiring the use of wheelchairs or other assistive devices [1,2]. Spinal diseases such as intervertebral disc herniation may lead to persistent back pain and nerve compression, affecting patients' sleep quality and mental health [3]. These diseases impose a heavy burden on individuals and society. Researchers have used a variety of approaches to explore the molecular mechanisms of these diseases. Recent progress in molecular biology and genomics has enhanced the understanding of orthopedic disorders, enabling more precise diagnostic and treatment approaches.

Single-cell RNA sequencing (scRNA-seq) offers innovative methods for examining gene expression in individual cells [4,5]. This innovative technology has significantly advanced transcriptome analysis by offering exceptional precision and depth, enabling researchers to distinguish cell types and examine transcriptomic variations among cells [6]. ScRNA-seq aids in identifying novel biomarkers and enables personalized treatment by elucidating the intricate biological processes involved in the onset and progression of orthopedic disorders.

However, the application of scRNA-seq technology in orthopedic disorder research faces challenges, such as the limited number of bone tissue cells and the rigid bone matrix, complicating single-cell isolation. Osteoblasts and osteoclasts are prone to phenotypic changes under in vitro conditions, affecting the accuracy of data. Moreover, sequencing costs and challenges in acquiring human samples can impact the results’ accuracy and reproducibility. Overcoming these challenges necessitates enhanced sample processing methods and the creation of specialized bioinformatics tools [7].

In this article, we reviewed these orthopedic disorders orthopedic disorders through recent scRNA-seq findings, highlighting how this technology enhances our understanding of their pathophysiological processes and molecular mechanisms (Table 1).

Table 1.

Overview of scRNA-seq research in orthopedic disorders.

Author	Disease	Resource	Comparative groups	Number of effective cells	Distinct Cell Types Identified	Highlights
Ji et al. [12]	OA	Human articular cartilage	Different OA stages (n = 10)	1464	chondrocytes	It identified three new subsets of chondrocytes and found a potential transition among ProCs, preHTCs and HTCs.
Fan et al. [15]	OA	Human articular cartilage	Non-OA (n = 3) vs OA (n = 10)	135,896	chondrocytes	It highlighted InfC, preHTC, preFC and HTC as potential cell populations to target for therapy.
Sun et al. [18]	OA	Human articular cartilage	Non-OA (n = 3) vs OA (n = 3)	75,104	chondrocytes	METRNL + substypes is an early differentiated chondrocyte that may be protective for cartilage tissue renewal and regeneration.
Lv et al. [21]	OA	Human articular cartilage	Different OA cartilage areas (n = 6)	17638	chondrocytes	TRPV1 activation protects chondrocytes from ferroptosis and ameliorates OA progression by upregulating GPX4.
Li et al. [27]	OA	Human articular cartilage	Different OA stages (n = 3)	20839	chondrocytes	Microtubule stabilization is a promising therapeutic target for OA and cartilage injury.
Pu et al. [36]	OA	Human infrapatellar fat pad	Non-OA (n = 6) vs OA (n = 6)	49674	adipose stem and progenitor cells	Interstitial inflammatory fibroblasts modulated the activities of macrophages and T cells and regulated cartilage proliferation through the MK pathway
Sebastian et al. [28]	OA	Mice articular cartilage	Different OA stages (n = 10)	2490	chondrocytes	It found early molecular changes in OA chondrocytes versus normal chondrocytes.
Zhang F et al. [39]	RA	Human synovial tissues	Different RA stages (n = 70) vs OA (n = 9)	314,011	fibroblasts	It established six groups of cell type abundance phenotypes to predict treatment response.
Yan et al. [41]	RA	Human articular cartilage	Different OA cartilage areas (n = 8)	87542	chondrocytes	There is a macrophage polarization transition process in the immune-related macrophage chondrocyte subset in response to mechanical load。
Meng et al. [55]	RA	Human synovial tissues	Different RA stages (n = 70)	22925	fibroblasts	It identified the FGF10-FGFR1 pathway as a key signaling pathway in recurrent RA.
Binvignat et al. [61]	RA	Human peripheral blood	Non-RA (n = 18) vs RA (n = 18)	125,698	mononuclear cells	It observed an increase in CD4 T effector memory cells in patients with moderate-high disease activity.
Ling et al. [66]	IDD	Human NP tissues	Different IDD stages (n = 3)	36196	NP cells and immune cells	Macrophage polarization and NP progenitor cell metabolism may play a key role in the progression of IDD.
Han et al. [73]	IDD	Human NP tissues	Non-IDD (n = 1) vs IDD (n = 5)	30300	chondrocytes, endothelial cells and macrophages	A differentiated direction of chondrocytes interacts with macrophages and endothelial cells and has an inflammatory amplification effect.
Zhang Y et al. [70]	IDD	Human NP tissues	Non-IDD (n = 1) vs IDD (n = 5)	30300	chondrocytes	Genes associated with ferroptosis were significantly enriched in the mild IDD group
Shi et al. [76]	IDD	Human CEP tissues	Different IDD stages (n = 2)	8534	chondrocytes	Chondrocyte-macrophage interaction depends on FN1-a4b7 complex and FN1-a4b1 complex.
Swahn et al. [77]	IDD	Human AF and NP tissues	Non-IDD (n = 3) vs IDD (n = 10)	90834	chondrocytes and fibroblasts	THBS signaling drives fibrosis and tissue remodeling.
Zhou et al. [82]	OS	Human tumor tissues	Different OS stages (n = 11)	100987	osteoblastic OS cells, chondroblastic OS cells and osteoclastic cells	It revealed transdifferentiation between malignant osteoblasts and malignant chondrocytes.
Liu et al. [91]	OS	Human tumor tissues	Different OS stages (n = 18)	117964	T cells, myeloid cells and osteoblast cells	It revealed the pathway of osteoblast cell differentiation during lymph node metastasis. OS cells remodel the microenvironment of lymph nodes by interacting with bone marrow, cancer-associated fibroblasts, and NK/T cells.

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; IDD, intervertebral disc degeneration; OS, Osteosarcoma; NP, nucleus pulposus; AF, annulus fibrosus; CEP, cartilage endplate.

1.1. Methodology

This review offers insights into assessing the application of scRNA-seq in orthopedic disease research. Articles with full text in English published by October 2024 were eligible for inclusion in the review. This time allowed us to properly assess the practice in this field and the innovations that have changed over a considerable period of time. We conducted a comprehensive literature search utilizing databases such as PubMed, EMBASE, and Google Scholar. Searches incorporated keywords like ‘scRNA sequencing’ and ‘single-cell sequencing’ alongside terms such as ‘osteoarthritis’, ‘rheumatoid arthritis’, ‘intervertebral disc degeneration’, and ‘osteosarcoma’. To improve the search strategy, additional sources were identified by manually examining references cited in recent disease-specific reviews. Strict exclusion criteria were implemented, omitting independent abstracts, case reports, posters, and studies that were unpublished or lacked peer review. The review aims to include high-quality and reliable evidence by establishing these criteria. The review's scope was unrestricted regarding the number of studies included, aiming to collect extensive knowledge on the subject. The review encompassed various study designs, such as descriptive, animal model, cohort, and observational studies. The study encompasses both preclinical and clinical investigations, offering a comprehensive view of orthopedic disease research applications.

2. Results and discussion

2.1. Osteoarthritis

Osteoarthritis (OA) is a common joint disease influenced by age and multiple factors [8,9]. It is characterized by degeneration of articular cartilage [10]. The emergence of scRNA-seq offers unparalleled opportunities to uncover cartilage tissue cellular heterogeneity and map chondrocyte differentiation pathways.

ScRNA-seq enables high-resolution analysis of individual cell gene expression, revealing novel chondrocyte subpopulations not observable at the traditional population level. The study indicates that articular cartilage contains various chondrocyte subtypes, namely proliferative chondrocytes (ProCs), prehypertrophic chondrocytes (preHTCs), hypertrophic chondrocytes (HTCs), and fibrocartilage chondrocytes (FCs) [11]. Ji et al. [12] first published scRNA-seq studies on human OA cartilage, where they identified three new subsets: effector chondrocytes (ECs), regulatory chondrocytes (RegCs), and homeostatic chondrocytes (HomCs). ECs play a crucial role in energy supply by engaging in the TCA cycle, glycolysis, and lipid and amino acid metabolism. RegCs are linked to various signaling pathways, indicating their potential significance in OA progression. HomCs exhibited elevated expression of genes associated with cell cycle regulation, metabolism, and development, suggesting their potential role in influencing the circadian clock during OA progression. Because protective genes are highly expressed, these three populations are protective against OA progression. Subsequently, two undescribed chondrocyte populations were identified: reparative chondrocytes (RepC), and prefibrochondrocytes (preFC) [13]. The first type shows significant repair ability due to elevated expression of genes associated with extracellular matrix signaling and collagen fiber organization, whereas the second type is likely crucial in ERK signaling because of its high expression of fibroblast-related genes and IL11 [14].

Fan et al. [15] dentified two novel populations: the pre-inflammatory chondrocyte population (preInfC) and the inflammatory chondrocyte population (InfC). InfC are activated by MIF-CD74 ligand–receptor interaction during cartilage degeneration in OA knees. Using spatial transcriptomics analysis, Fan identified the articular surface as the most active region of transcriptional activity. Among these areas, many potential targets previously reported as candidates for OA treatment were detected. PRG4 functions as a lubricant, facilitating smooth joint movement [16]; while cartilage acidic protein 1 (CRTAC1) is proposed as a biomarker for early OA diagnosis and disease progression monitoring [17]. For OA patients, articular chondrocytes in the intermediate and deep zones are relatively quiescent. preHTCs and preFc are predominantly located in the articular surface and superficial regions of OA patients, indicating their potential role in enhancing OA treatment. The results of spatial transcriptomics analysis were highly consistent with the regions of activation of proinflammatory signaling pathways predicted by scRNA-seq, confirming the functional localization of these subsets in single-cell data (Fig. 1). Sun et al. [18] identified two novel chondrocyte subtypes: METRNL+ and PRG4+. METRNL+ is a common cartilage subtype in the normal population, representing an early differentiated chondrocyte that may contribute to cartilage tissue renewal and regeneration [19]. The OA group was characterized by the presence of PRG4+ and RegC-B cell subsets. The cartilage repair facilitated by these subtypes differs from the extracellular matrix renewal process of normal chondrocytes, potentially contributing to the degeneration of osteoarthritic cartilage tissue.

Fig. 1.

Cellular heterogeneity between OA patients and non-OA controls. (A) The cell fraction difference between patients with OA and non-OA controls. (B) The MIF expressed by other chondrocytes interacts with the CD74-CXCR2/4-CD44 complex expressed by InfCs and activates the MAPK signalling pathway, which increases the expression of the activator protein 1 (AP-1) transcription factor and further promotes the transcription of pro-inflammatory cytokines. (C) An integrative analysis of scRNA-seq data and bulk RNA-seq data demonstrates two major subtypes of patients with OA are stratified by preFC, resulting in an inflammatory-related subtype and a non-inflammatory-related subtype of OA. (D) An integrative analysis of scRNA-seq data and GWAS summary data shows that HTC is the key chondrocyte population that drives the susceptibility to the pathogenesis of OA. Reproduced from Fan Y et al. (2024) with permission from Annals of the Rheumatic Diseases.[15]

Qu et al. [20] first reported a subset of chondrocytes highly expressing SPP1, which has more aging properties and angiogenesis capacity. In animal models, SPP1 expression in cartilage exhibited spatial heterogeneity. Lv et al. [21] first identified siderocytic chondrocyte clusters in OA cartilage by single-cell sequencing. It was also revealed that TRPV1 protected chondrocytes from ferroptosis by partially upregulating GPX4, thereby significantly eliminating cartilage degeneration and ameliorating OA progression. The anti-ferroptosis effect of TRPV1 was confirmed by Gpx4 knockout mouse model, which was consistent with the functional prediction of TRPV1 highly expressed subsets in scRNA-seq, clarifying its therapeutic potential for OA. Their research team discovered that TRPV1 mitigates the progression of OA by inhibiting M1 macrophage polarization and reducing synovitis. Additionally, they identified a novel mechanism involving the local administration of the TRPV1 agonist, Capsaicin, for the treatment of OA in a rat model [22]. These findings suggest that TRPV1 is a novel therapeutic target reducing synovitis and alleviating OA. Guang et al. [23] discovered a novel chondrocyte subset, VIPER cluster 3. This key chondrocyte cluster was identified as crucial for cellular homeostasis and anti-aging. Animal studies confirmed that NDRG2, TSPYL2, JMJD6, and HMGB2 are significantly associated with the pathogenesis of OA [[24], [25], [26], [27]] and may crucially regulate chondrocyte cellular homeostasis and anti-aging processes. Li et al. [28] identified a CHI3L1+C subset distinguished by its regenerative capacity, stem cell potential, and active microtubule processes. MT stabilization enhances cartilage regeneration in rat models of cartilage injury by suppressing YAP activity. The discovery of the CHI3L1+C subpopulation and its link to MT stabilization suggests a possible mechanism for cartilage regeneration and chondrogenesis. MT stabilization plays a crucial role in cartilage repair, offering a promising therapeutic target for osteoarthritis (OA).

Sebastian et al. [29] initially identified nine distinct chondrocyte subtypes (Ucma^high, Cytl1^high, Chil1^high, Mef2c^high, Krt16^high, Tnfaip6^high, S100a4^high, Neat1^high, and divC chondrocyte clusters) in the articular cartilage of the healthy mouse knee joint. Additionally, Sebastian's investigation into mouse and human chondrocyte transcriptomes revealed that the Cytl1^high, Chil1^high, Krt16^high, S100a4^high, and divC clusters in mice correspond to human ECs, RegCs, ProCs, fibrocartilage chondrocytes, and cartilage progenitor cell clusters, respectively. The study identified early molecular alterations in osteoarthritis (OA) chondrocytes compared to normal chondrocytes, characterized by the up-regulation of Mmp3, Mmp13, Ptgs2, Inhba, Sfn, and IL11, and the down-regulation of Cytl1, Errfi1, and Il17b.Animal sequencing results have uncovered biological differences that enhance our understanding of chondrocyte diversity and function.

ScRNA-seq can reveal candidate therapeutic targets or biomarkers in different cell subsets. Deng et al. [30] suggested that RegCs are likely the primary cell type involved in necroptosis in OA, identifying four key necroptosis genes: RIPK3, CYBB, TRAF5 and HSP90AB1 [[31], [32], [33], [34]]. Fang et al. [35] highlighted the significance of programmed cell death-related genes in OA chondrocytes, confirming the pivotal role of CDKN1A, enriched in HomCs, in enhancing chondrogenesis through both in vivo and in vitro studies [36], and demonstrated its potential as a therapeutic target for OA.

By scRNA-seq of infrapatellar fat pad (IFP), Pu et al. [37] demonstrated that adipose stem and progenitor cells (ASPC) play a role in forming adipocytes and synovial fibroblasts (SLF). Among them, interstitial inflammatory fibroblasts (iiFBs) are a subset of ASPC with significant pro-inflammatory and proliferative characteristics. Cell communication analysis confirmed that iiFB interacted with chondrocytes through MDK molecules. iiFBs modulate macrophage and T cell activity within the IFP. Distinct from subcutaneous and visceral adipose tissue, iiFBs are a unique subset of ASPC within the IFP that influence cartilage proliferation via the MK pathway, impacting osteoarthritis (OA) progression.

Chondrocyte differentiation (hypertrophy) is one of the mechanisms of cartilage degradation in OA [38]. Previous studies have suggested that SRY-related HMG box 9 (SOX9) is a core regulator of chondrocyte differentiation and maintains the chondrocyte phenotype by activating key genes such as Col2a1 and Acan [39], and its decreased expression in OA is directly related to the imbalance of cartilage homeostasis [40]. Using scRNA-seq, Hata et al. [41] identified Serine incorporator 5 (Serinc5) as a specific marker gene of hypertrophic prechondrocytes and verified that Serinc5 inhibited cell proliferation and Col2a1 and Acan expression by inhibiting Sox9 transcriptional activity in primary chondrocytes. To regulate sequential chondrocyte differentiation from proliferation to hypertrophy, indicating that Serinc5 expression is epigenetically regulated during chondrocyte differentiation, revealing the direct impact of dynamic changes in SOX9 activity on cell state transition.

ScRNA-seq technology offers novel insights into osteoarthritis (OA) by uncovering the intricate cellular diversity within tissues. This precise analytical method identifies cell types and signaling pathways linked to disease progression, offering potential targets for new therapeutic strategies. With the continuous progress of technology and the optimization of data analysis methods, single-cell sequencing is expected to play a greater role in revealing the pathological mechanism of OA and guiding precision treatment. At the same time, its combination with other multi-omics technologies will further deepen the comprehensive understanding of OA.

2.2. Rheumatoid arthritis

Rheumatoid arthritis (RA) is an autoimmune disease characterized by joint inflammation and bone destruction [42]. In RA, the activation and proliferation of immune cells, such as T cells and macrophages, along with synovial fibroblasts, occur in the arthritic synovium, resulting in joint inflammation and bone destruction [43].

By scRNA-seq of RA synovium, Zhang F et al. [44] categorized the synovium based on cell type abundance into six distinct cell type abundance phenotypes (CTAPs): (1) endothelial cells, fibroblasts, and bone marrow cells (EFM); (2) fibroblasts (F); (3) T cells and fibroblasts (TF); (4) T and B cells (TB); (5) T cells and myeloid cells (TM); and (6) bone marrow cells (M). The CTAP paradigm offers a tissue classification framework that encompasses both broad cell type and detailed cell state diversity. This model could effectively prototype the classification of various tissue inflammations, including other immune-mediated diseases. Gaining a deeper understanding of tissue inflammation heterogeneity in RA and other autoimmune diseases could offer new insights into disease pathogenesis, identify novel therapeutic targets, and highlight essential components of precision medicine.

Research indicates that chondrocyte proliferation, apoptosis, and autophagy are linked to RA disease progression [45]. To explore the pathogenesis of RA, Yan et al. [46] conducted scRNA-seq on cartilage from both weight-bearing and non-weight-bearing femur regions in RA patients, discovering two novel immune chondrocyte subsets: inflammation-associated chondrocytes (IrC) and macrophage chondrocytes (MC). MC subsets were prevalent throughout the cartilage layer, particularly in the middle, exhibiting significant immune activity. Their primary roles included antigen processing and presentation, involvement in the MHC class II protein complex, and participation in immune receptor activity. HLA-related genes were highly expressed in MC subsets, and gene enrichment analysis showed that MC subsets were closely related to RA. IrC subsets are distributed throughout the cartilage layer and exhibit inflammatory properties such as immune cell adhesion, cytokine activity and receptors. IrC subsets highly express leukocyte chemokine genes that activate the NF-κB signaling pathway via nuclear factor-κB ligand (RANKL) [47,48], resulting in bone erosion and destruction. Research indicates that MC in weight-bearing regions is primarily associated with antigen processing, presentation, and MHC class II protein complex binding, aligning with the increased cartilage damage observed in RA patients in these areas. MC in the non-weight-bearing area was predominantly associated with cartilage development and extracellular matrix tissue, indicating potentially more active cartilage growth and development in RA patients. MCS in weight-bearing areas primarily initiate and sustain immune responses, enhancing antigen processing and presentation, akin to the functions of proinflammatory macrophages (M1). MCs in the non-weight-bearing zone are linked to connective tissue and cartilage development, potentially functioning like anti-inflammatory M2 macrophages by alleviating and repairing inflammation [49].

RA pathology typically involves synovial fibroblasts that cause joint damage by activating proinflammatory and tissue destruction pathways [50]. ScRNA-seq revealed the properties of RA synovial fibroblasts. Li et al. [51] identified POSTN as a crucial gene in RA, noting its expression level positively correlates with M2 macrophages during immune cell infiltration analysis. Wang et al. [52] found increased TIMP1 expression in RA synovial tissues and fibroblast-like synoviocytes (FLS). TIMP1 is a gene associated with ferroptosis [53] and holds promise as both a biomarker and therapeutic target for RA. He et al. [54] identified ferroptosis-related genes ENO1, GRN and PTGS2 [[55], [56], [57]], offering insights into the molecular mechanisms that regulate synovial ferroptosis and the immune microenvironment in RA. At the same time, He confirmed the significant role of the ENO1-ACO1 axis in regulating ferroptosis during RA development in animal models. Tagawa et al. [58] demonstrated the diversity of RA synovial fibroblasts and identified ARID5B as a suppressor of IL-6 production in RA. The RA risk allele within ARID5B's intron potentially increases IL-6 production, indicating ARID5B as a promising drug target for RA treatment. Torres et al. [59] identified mitochondrial-bound hexokinase HK2 as a crucial regulator of RA FLS phenotype, linking HK2-positive fibroblasts to genes like collagen alpha genes (COL1A1, COL3A1, etc.) and alpha actin-2 (ACTA2). The expression of these genes is reduced by dissociation of mitochondrial HK2. Dissociation of HK2 from mitochondria may become a new treatment for RA.

Combining scRNA-seq and spatial transcriptomics, Meng et al. [60] conducted sequencing and analysis of synovial tissue samples from patients experiencing RA relapse and those in remission. Meng categorized FLS into four subsets: FBL-CD55, CXCL12+ sublining fibroblasts (FBS-CXCL12), MFAP5+ sublining fibroblasts (FBS-MFAP5), and VEGFA + sublining fibroblasts (FBS-VEGFA). The observation of a high proportion of FBL-CD55 in patients with recurrent RA suggests that intimal fibroblasts proliferate significantly in these patients. Meng identified that FGF10-FGFR1 signaling in intimal FLS is potentially crucial in the pathogenesis of recurrent RA. Inhibition of FGFR1 may slow inflammatory erosion of joints and this possibility was confirmed in the rat model [61,62]. Targeting FGF10-FGFR1 signaling may provide a new therapeutic opportunity for patients with recurrent RA. Meng's study also found that the IFN-α response pathway was highly expressed in lining FLS from RA patients with recurrent. This pathway is thought to have a potential pathogenic role in RA [63]. The study could offer novel perspectives for creating therapeutic targets for RA.

Immune cell infiltration is one of the important features of RA [64]. With the help of scRNA-seq, Ding et al. [65] demonstrated that in RA, T cells interact directly with immune cells in the RA microenvironment via paracrine signals and primarily engage with other cells through the MIF signaling pathway. Using a combination of machine learning and Mendelian randomization approaches, eight T-cell-associated diagnostic features were identified. Among them, ICOS, IL6ST and PPP1CB are identified as risk factors for RA, whereas GADD45A, CD3D, SLFN5, PIP4K2A and MIER1 are recognized as protective factors. Based on these features, a T cell-related diagnostic model was constructed. This study provides an important reference for the immunotherapy strategy of RA. To investigate the role of monocytes in RA, Binvignat et al. [66] conducted scRNA-seq on peripheral blood mononuclear cell samples from participants. The study noted a rise in CD4 T effector memory cells among patients with moderate to high disease activity, while those with low disease activity or in remission exhibited a reduction in nonclassical monocytes.

Biesemannet al. [67] utilized scRNA-seq and analysis in the development of new drugs for treating RA. They developed a model of TNF and IL-6 interactions, demonstrating their synergistic effects on RA FLS and T cells. The study identified fibroblasts and T cell subsets as key disease-driven cell types targeted by anti-TNF/IL-6 therapies in RA. In vitro and in vivo studies demonstrated that the bispecific anti-TNF/IL-6 nanoantibody compound or combination therapy outperformed monospecific treatment. Huynh et al. [68] conducted scRNA-seq analysis on drug-treated tissues and examined transgenic mice administered JAK inhibitors. They demonstrated that JAK inhibitors influence gene expression in synovial fibroblasts and macrophages in RA. JAK inhibitors disrupt the interaction between RA synovial fibroblasts and macrophages by targeting oncostatin M (OSM) signaling. The OSM-driven synovial macrophage-fibroblast loop is a crucial driver of RA and an important drug target in vivo.

ScRNA-seq has revealed the unique functions and interactions of different cell types such as synovial fibroblasts and T cells in RA (Fig. 2), providing insights into cellular heterogeneity and the disease microenvironment. It lays the foundation for the discovery of new therapeutic targets. It is foreseeable that scRNA-seq will continue to play a role in RA research in the future, and by combining with other analytical techniques, this technology will provide a more comprehensive perspective for a deeper understanding of the pathogenesis of RA.

Fig. 2.

Interaction of synovial fibroblasts with various types of immune cells. Adapted from Cheng et al. (2021) with permission from Front Immunol, with modifications for clarity [101].

2.3. Intervertebral disc degeneration

Intervertebral disc degeneration (IDD) is the main cause of low back pain [69]. The intervertebral disc (IVD) includes the nucleus pulposus (NP) originating from the notochord, the annulus fibrosus (AF), and the endplate (EP) originating from the sclerotium [70]. ScRNA-seq identifies new cell subsets in (IVD and elucidates their functions, enhancing our understanding of IDD pathology and offering fresh perspectives on therapeutic approaches.

Ling et al. [71] studied NP organization in human intervertebral discs at various stages of degeneration. The study categorized nucleus pulposus cells (NPC) into six types: metabolic homeostatic (Met NPC), adhesive (Adh NPC), inflammatory response (IR NPC), endoplasmic reticulum stress (ERS NPC), fibrocartilaginous (Fc NPC), and progenitor cells marked by CD70 and CD82 (Pro NPC). In the advanced stages of IDD, IR NPC and Fc NPC constituted a significant portion of NPC. Ling identified diverse immune cells such as M1 and M2 macrophages, T cells, myeloid progenitor cells, and neutrophils. Further analysis identified notable interactions between macrophages and Pro NPC through macrophage migration inhibitory factor (MIF) and NF-kB signaling [72,73] during IDD progression. Macrophages exhibited dynamic polarization between M1 and M2 subtypes throughout IDD progression. Functional enrichment analysis indicated that this polarization significantly influenced cell metabolism, particularly in the regulation of Pro NPC.

Chondrocytes are essential for the function of the IVD. A novel class of regulatory chondrocytes, enriched with growth factors, has been discovered in the intervertebral discs (IVDs) of healthy individuals. These chondrocytes regulate the chondrogenic pathway, contributing to the homeostasis of chondroid ECM [74]. Zhang Y et al. [75] identified new subsets of chondrocytes in the NP by scRNA-seq. The study identified seven chondrocyte subsets: fibrochondrocyte progenitor cells (FCP), chondroprogenitor cells (CPC), homeostatic chondrocytes (HomCs), and four newly discovered subsets labeled C1-C4. The C3 subset was exclusively identified in patient samples, with its highly expressed genes linked to inflammatory immune responses and extracellular matrix remodeling. The study identified a significant enrichment of ferroptosis-associated genes in the mild IDD group and noted alterations in the expression levels of ferroptosis markers FTL and HO-1 [76,77] in the rat model. The findings indicate a strong association between ferroptosis and the development of IDD. The study by Han et al. [78] similarly identified three traditionally defined and four novel subsets of chondrocytes. The study indicates that CPCs are plentiful in healthy discs but diminish as degeneration progresses, exhibiting significant proliferative ability and energy use. This suggests a novel approach for treating degenerative disc diseases with NP-derived progenitor cells. Gene enrichment analysis revealed a significant increase in C1 and C3 in the IDD group, accompanied by elevated levels of chemokines (e.g., CXCL8, CXCL2) and matrix-degrading enzymes (e.g., MMP2, MMP3, MMP13), all linked to the inflammatory response [79,80] In addition, the results also suggest that the differentiation of chondrocytes into fate2 is the main factor leading to IDD, but the specific mechanism of this process still needs to be further studied.

Shi et al. [81] conducted scRNA-seq on IVD endplate cells from IDD patients, identifying nine cell types: chondroblast (CB), regulatory chondrocytes (RegC), homeostatic chondrocytes (HomC), pre-hypertrophic chondrocytes (pre-HTC), fibrocartilage chondrocytes (FC), proliferative chondrocytes (ProC), hypertrophic chondrocytes (HTC), T cells/natural killer cells (T/NK), and macrophages (MA). Further analysis showed that chondroblasts (CB) may regulate the expression of DCN through EGR1 to participate in chondrogenesis. Chondrocytes contribute to internal homeostasis and ECM regulation, altering the endplate microenvironment and significantly impacting IDD. Research indicates a strong association between macrophages and chondrocytes, mediated by FN1-a4b7 and FN1-a4b1 complexes. Considering the essential functions of chondrocytes and immune cells in IDD, future treatments might focus on altering chondrocyte types and modulating immune cells to improve disease outcomes.

Using scRNA-seq of AF and NP tissue from normal and IDD patients, investigators found that fibrochondrocyte subsets (FC-1 and FC-2) were significantly increased in diseased NP [82]. Further analysis showed that FC-1 and FC-2 were closely related to thrombospondins (THBS). THBS1 gene expression was significantly upregulated in diseased NP compared to healthy tissues. Research indicated that THBS1 facilitates fibrosis via both the TGFβ pathway and an independent mechanism [83]. The data indicate that THBS signaling significantly contributes to disease progression, fibrosis, and tissue remodeling, key characteristics of degenerative NP. In sequencing AF tissue, they identified a significant expansion of a new disease-associated subpopulation, termed disease-associated chondrocytes (Dacs), and speculated that these cells most likely arose from stem cells. In addition, disease-associated chondrocytes and fibrochondrocyte subsets, which are considered to be major transmitters of THBS signaling, were found to be enhanced in both THBS1 gene expression and protein abundance in AF. Collectively, these findings reinforce the critical role of THBS signaling in IDD pathogenesis.

ScRNA-seq technology has offered significant insights into IDD research. It reveals the differentiation and regulation mechanisms of EP cells, NP cells and AF cells, identifies specific cell subsets, and elaborates the dynamic changes of these cells during disc degeneration (Fig. 3). In addition, scRNA-seq has unraveled the roles of various cells, such as chondrocytes and macrophages, in the process of degeneration, providing new ideas for potential therapeutic targets and intervention strategies. However, this technique also has some limitations, such as high cost, complexity of data analysis, and lack of spatial background information. Integrating scRNA-seq with bioinformatics technologies will enhance our comprehension of IVD cell heterogeneity, cell interactions, and identify key regulators in IDD development.

Fig. 3.

Heterogeneity and intercellular crosstalk in intervertebral disc degeneration. Adapted from Wang et al. (2023) with permission from iScience, with modifications for clarity [102].

2.4. Osteosarcoma

Osteosarcoma (OS) is a malignant bone tumor primarily affecting children and adolescents [84,85]. OS is believed to develop from primitive mesenchymal-derived osteoblasts and typically manifests in rapidly growing bones [86]. The highly aggressive nature of OS makes the treatment a great challenge. ScRNA-seq analysis enhances our understanding of the mechanisms driving OS development and progression.

Zhou et al. [87] divided osteoblastic malignant cells into six subsets based on comprehensive gene expression profiles through scRNA-seq data analysis. Two proliferative subsets exhibited elevated expression of genes linked to the S and G2/M phases. The remaining four subsets exhibited increased activity in signaling pathways associated with angiogenesis, MYC, IFN-α, KRAS, TP53, and other distinctive gene sets. Notably, several genes are overexpressed in the Myc, mTORC1, hypoxia, and oxidative phosphorylation signaling pathways in cases of pulmonary metastasis or recurrent osteoblastic OS. These pathways could significantly contribute to OS chemoresistance and recurrence. Furthermore, their findings revealed a significant upregulation of genes associated with osteoblast differentiation, ossification, bone morphogenesis, as well as histone methylation and acetylation during the transdifferentiation of chondroblastic tumor cells. This indicates that epigenetic changes could significantly influence the transdifferentiation of these cells (Fig. 4).

Fig. 4.

Overview of the crosstalk network between clusters in osteosarcoma [103].

TRAIL, a cytokine, induces apoptosis. TRAIL selectively triggers apoptosis in tumor cells without affecting normal cells. Feng et al. [88] identified that endothelial cells in OS tissues express multiple TRAIL receptors, suggesting that targeting these cells and angiogenesis could be effective for TRIAL-mediated OS therapy.

Regulatory T cells (Tregs) have been found to have a significant immunosuppressive function in osteosarcoma (OS). The study by Cheng et al. [89] demonstrated significant Treg infiltration in OS samples, alongside elevated activation of oxidative phosphorylation, angiogenesis, and mTORC1 pathways. In addition, CXCL signaling was also significantly enhanced in these Tregs. CXCL signaling triggers cell chemotaxis through binding to CXCR receptors, leading to chemokine production. These chemokines play key roles in maintaining tumor growth, promoting angiogenesis, and evaying immune surveillance [90]. Research indicated that osteoblasts and endothelial cells recruit Tregs via the CXCL12/CXCR4 axis, and osteoblasts, endothelial cells, along with myeloid cells, enhance Treg proliferation through the TGFB1/CXCR4 axis. These findings suggested that Tregs may regulate OS tumor cell growth and invasion through interactions between TGFB1, CXCL12, and CXCR4 signaling. Consequently, CXCR4 is regarded as a significant target for OS treatment.

Studies have shown that mature regulatory dendritic cells (mregDCs) may shape the immunosuppressive microenvironment of OS by recruiting Tregs. Using single-cell sequencing technology, researchers found that mregDCs actively recruit Tregs through the secretion of chemokines CCL17, CCL19, and CCL22. Furthermore, inhibitory receptor ligands, such as CD274-PDCD1 and PVR-TIGIT, are employed to directly suppress T cell activity, thereby contributing to the establishment of an immunosuppressive microenvironment. Additionally, tumor cells enhance immune evasion by down-regulating major histocompatibility complex class I (MHC-I) molecules and overexpressing CD24-Siglec10 ″don't eat me" signals, which inhibit the phagocytosis and antigen presentation capabilities of macrophages [91]. The role of tumor-expressed CD24 in facilitating immune evasion through its interaction with Siglec10 has been substantiated [92,93]. Collectively, these intercellular communication networks culminate in the formation of an immune-tolerant microenvironment, promote tumor progression, and provide a theoretical foundation for immunotherapy strategies targeting myeloid cells.

Researchers have used scRNA-seq technology to explore the regulatory role of non-immune cells, such as fibroblasts, in the tumor microenvironment and to reveal their impact on recurrent osteosarcoma. Huang et al. [94] identified a significant enrichment of cancer-associated fibroblasts (CAFs) in the epithelial–mesenchymal transition (EMT) pathway within recurrent osteosarcoma. Lysyl oxidase (LOX) contributes to VEGF induction, HIF-1α activation, and the enhancement of angiogenesis and EMT in various tumors [95]. Huang found LOX is highly expressed in CAFs, which may lead to EMT process and poor prognosis of osteosarcoma. The study demonstrated the variability in recurrent OS and identified a novel mechanism by which CAFs regulate EMT in OS through LOX. Targeting LOX in CAFs shows promising potential in remodeling the TME and treating recurrent OS.

OS is prone to metastasis, and the prognosis of patients with metastasis is extremely poor [96]. Liu et al. [97] employed scRNA-seq to investigate the molecular mechanisms funderlying lymphatic metastasis in OS. Single-cell sequencing of para-cancer tissues, primary tumors, and lymph node samples revealed that osteoblasts were continuously activated during tumor progression and were the predominant malignant cells in OS. Osteoblasts are categorized into seven subpopulations, with cluster 6 identified as crucial for lymph node metastasis. Comprehensive bioinformatics analysis and animal studies have confirmed the pivotal role of the ETS2/IBSP signaling axis in OS lymph node metastasis, offering new targets and strategies for inhibiting metastasis. Liu's study observed a notable decrease in T cells within metastatic lymph nodes, indicating that tumor cells might promote their own survival by diminishing T cell presence. Liu's spatial transcriptome analysis revealed that two tumor-associated macrophage subsets, SPP1 Mac and SELENOP Mac, aggregate near osteoblasts and, along with fibroblasts, create a barrier that hinders T cell infiltration, fostering a microenvironment conducive to tumor growth. At the same time, endothelial cells are dispersed between fibroblasts and osteoblasts to provide nutritional support for tumor growth. The spatial transcriptome analysis and immunohistochemistry results further verified the enrichment of SPP1 protein at the tumor edge, and confirmed the formation mechanism of immunosuppressive microenvironment predicted by scRNA-seq. These findings have important implications for understanding the metastatic mechanisms of OS and designing novel therapeutic strategies.

ScRNA-seq technology has shown great potential in OS research, promoting a deeper understanding of this complex tumor. Through scRNA-seq, researchers have revealed the significant influence of a variety of cells on the immune microenvironment, and elucidated the mechanism of immune evasion and its regulatory effect on OS progression. Through these studies, scientists are able to dissect the expression patterns of specific genes, provide novel insights into molecular mechanisms such as tumor cell apoptosis, better delineate the intratumor heterogeneity and immunosuppression characteristics of OS, and decode its metastasis mechanism. This technology has also helped to construct OS gene panels as prognostic models to assess prognostic risk [98]. In conclusion, scRNA-seq is a powerful tool for uncovering the molecular mechanisms and cell interactions in OS, advancing personalized treatment and precision medicine.

3. Conclusion

ScRNA-seq technology provides important help in the study of disease mechanisms. It can identify and resolve the multiple cell types and their heterogeneity involved in the pathogenesis of arthritis, in-depth analysis of the function of specific cell subsets, and reveal the complex network of cell–cell interactions. ScRNA-seq elucidates inflammatory signaling pathways and gene expression alterations during disease progression, examines spatiotemporal disease dynamics, and offers crucial data for identifying therapeutic targets and analyzing drug mechanisms. These insights may enhance understanding of disease mechanisms and aid in developing effective treatments.

Although ScRNA-seq technology has shown great potential, it still faces some problems and deficiencies. The high cost and technical requirements of this technique limit its wide application. In addition, single-cell sequencing cannot provide information on the spatial distribution of cells in tissues, which poses a limitation to understanding the disease microenvironment. Limitations in sample sources and cross-platform standardization issues may also affect the reproducibility and consistency of the findings. Therefore, although single-cell sequencing provides a new perspective for disease research, its widespread application and translation of research results still need to overcome many challenges.

Bone and cartilage tissues have a lower cell density compared to other tissues, necessitating larger samples for scRNA-seq to obtain adequate RNA quantity and quality. Due to their increased stiffness compared to other tissues, extracting single cells from bone and cartilage, particularly bone, is more challenging. When using collagenase to digest tissue after dissection, bone samples require longer digestion times and more steps. ScRNA-seq technology currently mainly provides single-cell level data of gene expression, which cannot directly reveal the spatial distribution of cells in bone tissue and their interaction with the surrounding microenvironment. Orthopedic disorders involve intricate cell–cell and cell–matrix interactions, and the absence of spatial information can hinder a comprehensive understanding of these interactions and disease mechanisms.

To enhance the use of single-cell sequencing and spatial transcriptome technology in orthopedic disease research, we should focus on the following areas.

Developing specific sample processing technologies: Specific sample processing and cell separation technologies should be developed according to the particularity of bone tissue, such as optimizing extracellular matrix removal methods, to improve cell separation efficiency and cell viability and ensure high-quality single-cell suspension. This can improve the success rate of single-cell sequencing and the reliability of the data.

Integrating multi-omics approaches: Single-cell sequencing, and spatial transcriptomics enhance the spatial context of cells, elucidating their precise locations within bone tissue and interactions with the surrounding environment. For instance, Chen et al. [99] utilized spatial transcriptomics technology to determine that NP progenitor cells are predominantly located within the outer subpopulation of the NP. Their study also clarified that Tie2 serves as a marker for vascular endothelial cells rather than NP progenitor cells, thereby rectifying previous misconceptions derived from in vitro culture or single-cell isolation methods. Additionally, Zhou et al. [100] constructed a spatio-temporal and single-cell transcriptome map detailing the embryonic development of human IVD. Their research further identified that notochordal SPP1 is associated with both the formation and degeneration of IVD. This single-cell transcriptome map is anticipated to facilitate the development of strategies aimed at preventing and regenerating IVD degeneration. Concurrently, the principal findings of scRNA-seq should be corroborated through the integration of additional omics datasets and the application of orthogonal methodologies, including proteomics and functional assays. To augment the translational applicability of scRNA-seq results, future research should focus on the incorporation of multi-omics data alongside clinical sample analysis.

Construction of large-scale sample bank: To establish a large-scale and high-quality sample bank of patients with orthopedic disorders, covering samples of different ages, genders, disease types and processes, so as to improve the representativeness and universality of the research results. These biobanks can provide the basis for multicenter studies and facilitate data sharing and standardization.

Develop precision analysis tools: Design and develop bioinformatics tools and algorithms specifically for bone tissue data analysis, especially algorithms that can handle complex extracellular matrix and rare cell populations, to improve the accuracy of data analysis and the efficiency of extracting biological significance.

Interdisciplinary cooperation: Strengthen the cooperation among orthopedics, molecular biology, computational biology and engineering, and combine the expertise of each field to promote the innovation and optimization of technology. Such interdisciplinary cooperation can accelerate the development and application of new technologies and improve the translation efficiency of research results.

Clinical translation research should promptly apply scRNA-seq technology findings to clinical practice, particularly in disease diagnosis, therapeutic target discovery, and personalized treatment plan design. Meng et al. [60] designed a specific FGFR1 inhibitor based on single-cell data to significantly reduce joint erosion in a rat model of RA. Lv et al. [21] similarly validated the ability of TRPV1 to treat OA in a gene knockout mouse model based on single-cell data. Cheng et al. [89] combined the conclusions of single-cell sequencing analysis to construct a prognostic model of OS based on the relevant characteristics of Tregs and clinical information. These cases demonstrate the potential of scRNA-seq technology for clinical translation. In future research, the implementation of adaptive trials should be considered to expedite the translation of single-cell discoveries into clinical applications.

Currently, numerous challenges remain in the study of orthopedic diseases. Notably, the dynamic heterogeneity of cell subsets within complex tissues, such as articular cartilage and the bone tumor microenvironment, and the mechanisms underlying their functional transformations during disease progression have yet to be fully elucidated. The identification and characterization of rare cell populations, including progenitor and transitional cells, along with their regulatory networks, remain insufficiently understood. Furthermore, there is an urgent need to elucidate the spatiotemporal characteristics and dynamic evolution of cell-to-cell interactions within the disease microenvironment. Future research should integrate multi-omics data, including spatial transcriptomics, epigenomics, and proteomics, to construct single-cell precision molecular maps of orthopedic diseases. This approach is essential for advancing precision diagnostic and therapeutic strategies.

ScRNA-seq has become an influential method, transforming our comprehension of orthopaedic systemic disease by offering detailed insights into cellular diversity and gene expression patterns. Its use in orthopedic degenerative diseases and bone tumors aids in identifying novel cell types, disease-specific gene expression alterations, and potential therapeutic targets. Despite challenges and limitations, scRNA-seq holds promising potential for clinical translation, offering a precise approach to orthopedic disease research and practice. scRNA-seq uncovers the intricate cellular landscape and molecular mechanisms of these diseases, offering a crucial resource for advancing targeted therapeutic interventions. Advancements in sequencing and analysis technology are anticipated to enhance the role of scRNA-seq in orthopedic research and patient prognosis.

Author contributions

Q.W. collected information and wrote the original manuscript. Q.W. and M.H. conducted formal analysis and collected information. J.J.G. conducted formal analysis, provided guidance and funding. Q.W., M.H. and J.J.G. edited the manuscript. All authors read and approved the final manuscript.

Availability of data and materials

No unique dataset was generated during preparation of this manuscript.

Consent for publication

Not applicable.

Funding

This work was supported by National Nature Science Foundation of China (grant number 62475181), Jiangsu Province Science and Technology Innovation Support Plan Project (grant number BZ2022051), China–Europe Sports Medicine Belt and Road Joint Laboratory, Ministry of Education of PRC (grant number 2023297), and Key Research Project of Higher Education Teaching Reform of Soochow University (grant number 2023–12).

The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Declaration of competing interest

The authors declare that they have no competing interests.