Upregulation of ACSL1 in synovial macrophages promotes lipid peroxidation via the IκB/NF-κB pathway to accelerate osteoarthritis

Abstract

Background

Osteoarthritis (OA) is a globally prevalent degenerative joint disease, characterized by cartilage degradation and synovial inflammation. Increasing evidence suggests that macrophages in the synovium play a pivotal role in OA pathogenesis. Energy metabolism reprogramming has emerged as a key regulator of macrophage activation in inflammatory diseases. Long-chain fatty acid-CoA ligase 1 (ACSL1), an enzyme critical for lipid metabolism, has been implicated in various diseases. However, the specific mechanism by which ACSL1 regulates macrophage polarization and contributes to OA progression remains unclear.

Methods

In this study, we examined ACSL1 expression in the hyperplastic synovium of patients with knee OA and in a mouse model of OA induced by destabilization of the medial meniscus (DMM). We isolated bone marrow-derived macrophages (BMDMs) from C57 mice and transfected them with ACSL1 knockdown plasmids to assess the impact of ACSL1 on macrophage polarization and inflammatory cytokine release. We also investigated the effect of ACSL1 knockdown on cartilage degradation using BMDM supernatant in cartilage explant cultures. Intra-articular injection of AAV-shACSL1 was performed to evaluate its effect on OA progression in a trauma-induced mouse model. The expression of ACSL1, inflammatory cytokines (IL-1, IL-6, TNF-α), and lipopolysaccharide (LPS)-induced macrophage polarization markers (M1 and M2 markers) was assessed using qRT-PCR, Western blotting, and ELISA. Lipid peroxidation and the activation of the IκB/NF-κB signaling pathway were examined to elucidate the mechanism by which ACSL1 regulates inflammation.

Results

We observed increased ACSL1 expression in both the hyperplastic synovium of OA patients and the synovium of DMM-induced OA mice. Knockdown of ACSL1 in macrophages inhibited M1 polarization and reduced the release of key inflammatory cytokines, including IL-1, IL-6, and TNF-α. Furthermore, supernatants from ACSL1-knockdown BMDMs mitigated cartilage degradation in explant cultures. Intra-articular injection of AAV-shACSL1 reduced OA progression in a mouse model of trauma-induced OA. Mechanistically, ACSL1 knockdown alleviated LPS-induced inflammation by inhibiting lipid peroxidation and reducing the activation of the IκB/NF-κB pathway, a major regulator of inflammatory responses in macrophages.

Conclusions

ACSL1 plays a crucial role in regulating the inflammatory state of synovial macrophages in OA. By modulating macrophage polarization and lipid peroxidation, ACSL1 contributes to the progression of OA. Targeting ACSL1 could provide a novel therapeutic strategy for the prevention and treatment of OA.

The translational potential of this article

This study highlights the pivotal role of ACSL1 in regulating macrophage-mediated inflammation in OA. Targeting ACSL1 expression or its associated pathways could offer a new approach for modulating synovial macrophage activation and preventing cartilage degradation. These findings suggest that ACSL1 may serve as a potential therapeutic target for both the prevention and treatment of OA, particularly through strategies aimed at controlling lipid metabolism and inflammatory responses in the synovium.

Graphical abstract

1. Introduction

Osteoarthritis (OA) is a globally prevalent degenerative joint disease that causes pain and functional disability, leading to significant clinical and financial burdens [1]. OA is well recognized as a "whole-joint" disease, characterized by the progressive degeneration of articular cartilage, subchondral bone sclerosis, osteophyte formation, and synovial inflammation (synovitis) [2]. Previous studies have suggested that baseline synovitis, detected by magnetic resonance imaging (MRI), is linked to the progression of knee OA, typically observed as worsening Kellgren and Lawrence grade or joint space narrowing [3]. However, the precise role of synovitis in OA pathogenesis remains unclear. Synovitis is pathologically characterized by fibroblast-like synoviocyte (FLS) proliferation and macrophage recruitment, leading to synovial lining hyperplasia [4]. Macrophages represent 12–40 % of synovial immune cells and are involved in both inflammatory and resolution phases following tissue injury [5]. Notably, linear modeling demonstrated that the ratio of CD14⁺ macrophages to total macrophages is predictive of the Knee Injury and Osteoarthritis Outcome Score (KOOS) and WOMAC score in OA [6], highlighting the significant role of macrophages in OA progression. Furthermore, prior research has indicated that the balance between glycolysis and fatty acid oxidation in macrophages influences their inflammatory state [7]. However, the connection between macrophage energy metabolism and immune function regulation in OA has yet to be fully elucidated.

The long-chain acyl-CoA synthetase (ACSL) family comprises key enzymes involved in lipid metabolism, including fatty acid (FA) elongation, oxidative degradation, phospholipid biosynthesis, and protein acylation. ACSL1 is widely expressed in the liver, heart, adipose tissue, and muscle [8], where it functions as a key enzyme in FA uptake [9], FA oxidation [10], and triglyceride synthesis [11]. It exhibits broad substrate specificity, spanning 16- to 18-carbon saturated FAs and 16- to 20-carbon unsaturated FAs. ACSL1 regulates lipid β-oxidation or re-esterification depending on its subcellular localization, either in the mitochondria or the endoplasmic reticulum [12]. Moreover, ACSL1 has been shown to inhibit ferroptosis by mediating the incorporation of exogenous linolenic acid into cellular phospholipids [13]. Another study suggests that ACSL1 is crucial for lipid droplet formation in microglia associated with Alzheimer's disease [14]. However, the role of ACSL1 in modulating macrophage function within OA synovitis remains unclear.

Lipid peroxidation is a biologically significant free radical chain reaction [15], producing a variety of primary and secondary peroxidation products. Lipid hydroperoxides (LOOH) are the predominant primary products [55], and their decomposition generates reactive carbonyl species (RCS), including short-chain carbonyl derivatives and oxidized truncated phospholipids [55]. Short-chain carbonyl derivatives comprise a range of lipid aldehydes, such as α, β-unsaturated aldehydes, di-aldehydes, and ketoaldehydes [16]. Among them, the most studied are the α, β-unsaturated aldehyde 4-hydroxy-2-nonenal (4-HNE) and the dialdehyde malondialdehyde (MDA). 4-HNE and MDA have been shown to be associated with the pathogenesis of various diseases such as cardiovascular diseases [17], neurological diseases [18], and inflammatory bowel disease [19] by reacting with macromolecules (proteins, lipids, and nucleic acids) and regulating a variety of signaling pathways such as Nrf2, NF-κB, PKC, and ERK [20]. However, the specific role of macrophage lipid peroxidation in regulating synovial inflammation and OA progression remains unclear.

In this study, we observed increased ACSL1 expression in the hyperplastic synovium of OA, based on analyses of clinical specimens and the construction of a traumatic OA mouse model. Subsequent studies demonstrated that ACSL1 knockdown in macrophages significantly reduced LPS-induced inflammation and mitigated the promotion of chondrocyte catabolism by macrophage supernatant. Mechanistically, ACSL1 knockdown in primary bone marrow-derived macrophages markedly attenuated LPS-induced activation of the IκB/NF-κB signaling pathway and lipid peroxidation. Moreover, intra-articular injection of AAV-shACSL1 alleviated synovial inflammation and cartilage destruction in OA mice, suggesting that targeting ACSL1-mediated lipid peroxidation could represent a potential therapeutic strategy for OA.

The mechanistic, hypothesis-driven approach employed in this study underscores the critical importance of understanding the precise molecular pathways governing OA pathogenesis, particularly the relationship between lipid metabolism and macrophage function. By utilizing both in vitro and in vivo models, this study provides a clear mechanistic rationale for targeting ACSL1 in OA therapies. This approach facilitates the identification of ACSL1 as a key regulator of the inflammatory state of macrophages, which directly influences OA progression.

2. Methods and materials

2.1. Human synovium

Normal synovium was obtained from individuals involved in road traffic accidents with no history of arthritic diseases (n = 6, aged 36.67 ± 4.50 years). Synovial tissue from osteoarthritis (OA) patients was collected during total knee replacement surgery (n = 12, aged 61.67 ± 9.74 years). Patient information, including ID, age, gender, and Body Mass Index (BMI) was presented in Appendix Table 1. After collection, the synovial tissue was categorized into moderate and severe OA groups based on Krenn's synovitis score and Kellgren–Lawrence grading from clinical X-ray imaging. All samples were initially divided into two portions: one for immunofluorescence imaging and the other for Western blot analysis. Patients with a history of malignancy, diabetes, or other severe diseases within the past 5 years were excluded. All patients provided informed consent for the use of their clinical information in scientific research. The study was approved by the Ethics Committee of the Third Affiliated Hospital of Southern Medical University (Approval number:2023-Ethic-116).

2.2. Animals

Ten-week-old male C57BL/6J (wild-type) mice were purchased from the Experimental Animal Centre of Southern Medical University (Guangzhou, China) and were raised to twelve weeks of age prior to surgery. The mice were randomized and assigned to groups using a random numbering system. This study was conducted in strict accordance with the Chinese Laboratory Animal Guidelines for Ethical Review of Animal Welfare (GB/T 35892-2018) for the protection of animal welfare, as well as the ARRIVE guidelines to ensure transparent reporting of experimental procedures, statistical methods, and data analysis. All animals were housed in cages without pathogens at a temperature of 22 ± 2 °C, humidity (50 ± 10 %), and a 12-h light/dark cycle with ad libitum access to food and water. Before surgery, the mice were anesthetized with pentobarbital via intraperitoneal injection. The medial menisco-tibial ligament anchoring the medial meniscus to the tibial plateau was transected to create a destabilization of the medial meniscus (DMM) OA model, as previously described [21]. Sham surgery was performed by opening and exposing the tissues of the right knee, followed by suturing the incision without interfering with the meniscus. All animal experiments were approved by the Ethics Committee of the Third Affiliated Hospital of Southern Medical University (Approval number:2022-Ethic-86).

2.4. Isolation of primary chondrocytes and bone marrow-derived macrophages (BMDMs)

Primary articular chondrocytes were isolated from the rib cartilage of 3-day-old C57BL/6 mice using a standard protocol with collagenase II (Sigma–Aldrich, St. Louis, MO, USA) in Dulbecco's modified Eagle's medium (DMEM, Gibco.). Primary chondrocytes were cultured in DMEM/F12 medium (Gibco, USA) with 10 % FBS and 1 % penicillin/streptomycin (Gibco). The medium was changed daily. Bone marrow-derived macrophages (BMDMs) were harvested from the bone marrow of 6-week-old male C57BL/6J mice as we described before [23]. Briefly, the femurs and tibias were separated and collected after the mice were sacrificed. The bone marrow cavities were exposed and flushed with complete DMEM (Gibco, Carlsbad, CA, USA) containing 10 % fetal bovine serum (FBS) (Gibco) and 1 % penicillin/streptomycin (Gibco). Red blood cells were then removed. The remaining cells were maintained in complete DMEM for 24 h. Non-adherent cells were centrifuged and collected, then plated in complete DMEM containing 10 % FBS with 25 ng/mL macrophage colony-stimulating factor (M-CSF, R&D Systems, Minneapolis, MN, USA) for 72 h. Validation of CD11b and F4/80 (macrophage-specific markers) expression in BMDMs by immunofluorescence staining to determine the purity of primary cells (Appendix Fig. 1A).

2.3. Articular injection in mice

To knock down the ACSL1 expression in the synovial tissue, an adeno-associated virus serotype 5 (AAV5) vector carrying shRNA targeting the mouse ACSL1 (AAV-shACSL1) was used (Hanbio Biotechnology Co. Ltd. Shanghai, China). Prior to injection, the mice were anesthetized with an intraperitoneal injection of 0.3 % sodium pentobarbital. AAV-shACSL1 (6 μl; 3.19 × 10^13 vg/ml) was intra-articular (IA) injected into the operated knees of DMM mice (commencing 1 week after DMM surgery) using a 33-gauge needle (Hamilton, Switzerland) as previously described [22]. IA injection of negative control AAV (AAV-shNC, 6 μl; 4.26 × 10^13 vg/ml) was used as a control.

2.5. Cell transfection and collection of culture supernatant

To knock down ACSL1 expression, BMDMs were transfected with a lentiviral vector targeting ACSL1 (shACSL1) (Hanbio Biotechnology Co. Ltd, Shanghai, China), The target sequences of these 3 pairs of shACSL1 have been listed below, and shACSL2 was selected for subsequent experiments. Briefly, BMDMs were seeded in 6-well plates at 50–70 % confluency and transfected with shACSL1 lentiviral particles using a multiplicity of infection (MOI) of 100. A non-targeting lentiviral shRNA (shNC) was used as a negative control. To overexpress ACSL1, macrophages were transfected with a overexpression plasmid carrying ACSL1 (OE-ACSL1; Hanbio Biotechnology Co., Ltd., Shanghai, China) using Lipofectamine 3000, while control cells were transfected with an pcDNA3.1-empty vector (OE-NC). The medium was replaced with fresh DMEM containing 10 % FBS after 24 h of incubation. BAY 11–7082 (2 μM,MedChemExpress, HY-13453) was added to the medium 0.5h before subsequent stimulation. Erastin (5 μM,MedChemExpress, HY-15763) was administered to the culture medium for 24 h to induce ferroptosis. Then the cells were stimulated with 500 ng/mL lipopolysaccharide (MedChemExpress, HY-D1056) for 24 h. The medium was changed after 24 h of stimulation, and the cells were incubated for another 48 h. The culture supernatant was collected into centrifuge tubes and centrifuged at 12,000 g to remove cellular debris. The centrifuged supernatant was aspirated and stored at −80 °C.

shACSL1 target sequence 1: GCTGATTGACATTCGGCAGTA

shACSL1 target sequence 2: CGACTTGTTGAAACTTGGGAA

shACSL1 target sequence 3: CCGAAGATCTTGCGATAATTT.

2.6. Cartilage explants

Three-week-old male C57 mice were euthanized, and femoral head cartilage explants were isolated. Cartilage explants were cultured in DMEM/F12 medium containing 10 % fetal bovine serum for 3 days and then further processed in 12-well plates. The medium was replaced with supernatant collected from LPS-stimulated macrophages, either transfected with shACSL1 or not. After 72 h of co-culture, the supernatant was discarded and the explants were fixed with 4 % paraformaldehyde.

2.7. Histology and immunofluorescence (IF) staining

Mouse knee joints and cartilage explants were fixed in 4 % paraformaldehyde for 12 h, decalcified with ethylenediaminetetraacetic acid (EDTA) for 21 days, then dehydrated and embedded in paraffin. Consecutive midsagittal sections (4 μm thick) were cut and stained with hematoxylin and eosin (H&E) or toluidine blue for morphological analysis. For IF staining, slides were deparaffinized, rehydrated, and rinsed with PBS. The slides were immersed in Tris–EDTA (pH 9.0) for 4 h at 95 °C in a water bath for antigen recovery. Slides were then blocked with 10 % normal bovine serum (Solarbio, Beijing, China) for 1 h at room temperature. The slides were then incubated with primary antibody at 4 °C overnight. The secondary antibody was applied for 1 h at room temperature, and then the IF slides were mounted with 4′,6-diamidino-2-phenylindole (DAPI, Thermo Fisher Scientific, Waltham, MA, USA). Antibodies used for IF staining were: rabbit anti-ACSL (Proteintech, 1:100, #13989-1-AP, CHINA), rabbit anti-MMP13 (Proteintech, 1:200, #18165-1-AP, CHINA), rabbit anti-COL2 (Abcam, 1:100, #ab34712, USA), mouse anti-F4/80 (Santa Cruz Biotechnology, 1:100, #sc-377009, USA), rabbit anti-ADAMTS5 (Invitrogen, 1:200, #PA1-1742, USA), rabbit anti-SLC7A11 (Proteintech, 1:100, #26864-1-AP, CHINA), rabbit anti-GPX4 (Abcam, 1:200, #ab125066, USA), rabbit anti-IκBα (Cell Signaling Technology, 1:200, #2859, USA), rabbit anti-phospho-NF-κB p65 (Cell Signaling Technology, 1:200, #3033, USA), rabbit anti-iNOS (Abcam, 1:200, #ab15323, USA), rabbit anti-CD206 (Proteintech, 1:100, #18704-1-AP, CHINA), species-matched horseradish peroxidase-conjugated secondary antibody (Jackson ImmunoResearch Laboratories), and species-matched Alexa-488- or Alexa-594-labeled secondary antibody (Invitrogen, USA).

2.8. Grading of cartilage and synovium structure

Two blinded observers independently scored synovitis severity based on synovial lining cell layer enlargement, resident cell density, and inflammatory infiltration, using a 9-point scale (Appendix Table 2) where lower scores indicated moderate synovitis and higher scores indicated severe synovitis [24]. Cartilage damage was graded from 0 to 6 based on the OARSI scoring system, with evaluations including the hyaline-to-calcified cartilage layer thickness ratio [25]. In synovial analyses, lining cells were defined as the flat cells at the synovium's surface, while the secondary layer was characterized by thicker blood vessels and sparsely organized, spindle-shaped fibroblasts. The lining layer was delineated as the region of interest, and cell counts were calculated using the threshold method in ImageJ (version 1.53a). The ratio of immunostain-positive cells to DAPI-positive cells in the lining layer was calculated at 400× magnification. For immunofluorescence staining analysis, both intensity and positive cell counts were quantified using ImageJ (version 1.53a).

2.9. SA-β-Gal assay

Primary chondrocytes were co-cultured with supernatants from LPS-stimulated macrophages, with or without transfection with shACSL1. The SA-β-Gal staining kit (Biovision, Milpitas, CA, USA) was used according to the manufacturer's instructions. Senescent chondrocytes were dyed blue after overnight staining at 37 °C. The positive cells were counted in four randomly selected fields per treatment (n = 6).

2.10. Enzyme-linked immunosorbent assay

Supernatants from bone marrow-derived macrophages (BMDMs) were collected as previously described. All samples were centrifuged at 4500×g for 15 min. The Mouse IL-1, IL-6, and TNF-α Quantikine Kit (R&D Systems) was utilized to quantify the concentrations of IL-1, IL-6, and TNF-α in the supernatants.

2.11. BODIPY 581/591 C11 assay

BMDMs were inoculated into 6-well plates at a concentration of 1 × 10^6 cells/well for 24 h, and then incubated with 5 μmol/L BODIPY 581/591 C11 (Invitrogen, USA) for 30 min at 37 °C. Cells were washed twice with PBS. Cells were collected and analyzed for FITC fluorescence intensity by flow cytometry, and the ratio of green to red fluorescence was observed by microscopy.

2.12. MitoTracker and MitoSOX staining

Mitochondrial superoxide was detected using the fluorescent MitoSOX probe (Invitrogen) and the fluorescent MitoTracker probe (Yeasen, Shanghai, China). Cells were incubated in Hank's buffer with 2 μM MitoSOX (Red) and 2 μM MitoTracker (Green) for 30 min at 37 °C, washed with PBS twice, and assessed by confocal microscopy (Olympus FV1200).

2.13. Western blot analysis

BMDMs cultured in 6-well plates were lysed with 150 μL of radioimmunoprecipitation assay (RIPA) buffer containing protease and phosphatase inhibitors. Proteins were transferred to polyvinylidene difluoride membranes after electrophoresis. Membranes were further incubated for 14–16 h at 4 °C with primary antibodies diluted in 5 % BSA TBST. Then, the membranes were incubated with secondary antibodies for 1 h at room temperature. Target bands were visualized using FDbio-Dura ECL (FDbio Science, Hangzhou, China). Antibodies used for western blotting were: mouse anti-iNOS (Santa Cruz Biotechnology, 1:1000, #sc-7271, USA), rabbit anti-ACSL1 (Proteintech, 1:1000, #13989-1-AP, China), rabbit anti-TNFα (Abcam, 1:1000, #ab6671, USA), rabbit anti-MMP13 (Abcam, 1:1000, #ab39012, USA), rabbit anti-p16 (Proteintech, 1:1000, #10883-1-AP, China), rabbit anti-p21 (Proteintech, 1:1000, #10355-1-AP, China), rabbit anti-SOX9 (Abcam, 1:1000, #ab185966, USA), rabbit anti-p65 (Cell Signaling Technology, 1:1000, #8242, USA), rabbit anti-IκBα (Cell Signaling Technology, 1:1000, #9242, USA), and species-matched horseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA, USA).

2.14. The qRT-PCR

In this study, total RNA was extracted from primary murine bone marrow-derived macrophages (BMDM) using TRIzol reagent (Takara Bio Inc., Shiga, Japan) following standard procedures. The RNA concentration and purity were assessed using a NanoDrop spectrophotometer (Thermo Fisher Scientific), ensuring acceptable A260/A280 and A260/A230 ratios. Subsequently, 1 μg of total RNA was treated with a genomic DNA removal kit and reverse transcribed into cDNA using the 5 × HiScript II qRT SuperMix II kit (Vazyme Biotech, Nanjing, China). The reverse transcription reaction was performed at 37 °C for 15 min, followed by a final enzyme inactivation step at 85 °C for 5 s. Quantitative real-time PCR (qRT-PCR) was performed using ChamQ SYBR qPCR Master Mix (Vazyme), with each reaction containing 10 μL of 2 × Master Mix, 10 μM forward and reverse primers, and 600 ng of cDNA template. All reactions were conducted in triplicate on a StepOnePlus Real-Time PCR System (Applied Biosystems). The thermal cycling conditions were as follows: initial denaturation at 95 °C for 5 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 30 s. A melting curve analysis was performed at the end of the qRT-PCR to confirm primer specificity. Relative gene expression levels were calculated using the 2^−ΔΔCT2 method, with Gapdh used as the internal control to normalize expression levels across samples. Data are presented as the mean ± standard deviation (SD), and statistical analyses were performed using GraphPad Prism 8.4.3, with a p-value <0.05 considered statistically significant. Biological replicates were derived from independently cultured cells under identical conditions. The primer sequences for the target genes are listed in Appendix Table 3.

2.15. Data collection and analyses

Synovial high-throughput RNA-seq datasets GSE206678 (including 3 control rat synovium samples and 4 OA synovium samples) and the mouse BMDMs microarray GSE53986 (including 4 untreated samples and 4 LPS-treated replicates) were downloaded from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) between normal and OA synovium groups were identified using the “DESeq2” package in R (4.3.2), with an FDR-corrected p-value <0.05 and absolute log2 fold change (|log2FC| > 1). GO and KEGG pathway enrichments of DEGs were performed using the enrichKEGG function from the “ClusterProfiler” R package, with an FDR-corrected p-value threshold of 0.05, and visualized using the dotplot function from “ClusterProfiler.” A volcano plot was generated using the ggplot function from the “ggplot2” R package. A heatmap was created using the pheatmap function from the “pheatmap” R package. Gene set enrichment analysis (GSEA) was performed using the Molecular Signatures Database (MSigDB) to identify pathways involved in OA pathogenesis, and the relevant statistical information, including normalized enrichment scores (NES), FDR q-values, and p-values, for each pathway was presented in the figures.

2.16. Statistical analyses

Data are presented as mean ± SD. For comparisons between two groups, Shapiro–Wilk tests were performed to determine the appropriateness of parametric or non-parametric tests, followed by an unpaired Student's t-test if applicable. Additionally, for comparisons involving three or more groups, a multivariate statistical test was initially used to assess overall group differences. One-way analysis of variance (ANOVA) was performed when the data satisfied the assumptions of normality and homogeneity of variance. Following a significant ANOVA result, Tukey's Honest Significant Difference (HSD) post hoc test was applied for multiple pairwise comparisons. Moreover, eta squared (η²) was calculated for ANOVA, and Cohen's d was applied in independent samples t-tests to quantify effect sizes. P-values <0.05 were considered statistically significant.

3. Results

3.1. Lipid metabolism and immune cell migration were altered in OA synovitis

First, we identified RNA sequencing data from MIA-induced OA in the GEO database (GSE206678). A total of 5884 differentially expressed genes (DEGs), including 2223 upregulated and 1837 downregulated genes, were identified from synovial samples of four KOA rats and three control rats (Fig. 1A). KEGG pathway analysis of the upregulated DEGs revealed significant participation in cellular behavior and tissue homeostasis, involving pathways such as “ECM-receptor interaction,” “Focal adhesion,” and “Glycosaminoglycan biosynthesis.” Additionally, several inflammation-related pathways were enriched, including “FcγR-mediated phagocytosis,” “Rheumatoid arthritis,” “PI3K-Akt signaling pathway,” “TNF signaling pathway,” and “NF-κB signaling pathway.” Interestingly, the "Lipid and atherosclerosis" pathway was also enriched (Fig. 1B). Downregulated DEGs were associated with significant changes in metabolic processes, such as "Thermogenesis," "Oxidative phosphorylation," and "Carbon metabolism." Pathways related to lipid metabolism, including "Fatty acid degradation" and "Fatty acid metabolism," were also enriched (Fig. 1C). Meanwhile, GO analysis revealed significant enrichment in pathways such as “actin binding,” “GTPase regulator activity,” “generation of precursor metabolites and energy,” “leukocyte migration,” “mitochondrial inner membrane,” and “mitochondrial matrix” (Fig. 1D–F). Correlation analysis of these GO terms showed a large cluster associated with mitochondrial functions (top left), including aerobic respiration, cellular respiration, and mitochondrial protein-containing complexes. The fatty acid metabolism cluster (bottom right) included terms such as fatty acid beta-oxidation and lipid oxidation, indicating involvement in lipid metabolism. An additional cluster related to the “Extracellular Matrix” and external encapsulating structure was also identified (top cluster) (Fig. 1G). Next, Gene Set Enrichment Analysis (GSEA) using the Molecular Signatures Database (MSigDB) revealed enrichment of the pathways "Fatty acid catabolic process" and "Response to lipid," suggesting alterations in lipid metabolism and signaling, indicative of potential metabolic disorders in OA. Furthermore, pathways related to "Monocyte chemotaxis" and "Leukocyte chemotaxis" were enriched, highlighting the roles of immune cell migration in OA synovitis. Altogether, the enrichment of pathways such as "Fatty acid catabolic process," "Response to lipid," "Monocyte chemotaxis," and "Leukocyte chemotaxis" indicates significant alterations in lipid metabolism and immune cell migration in OA synovitis (Fig. 1H). Additionally, GO terms related to mitochondrial respiration and lipid metabolism, such as “respiratory chain complex I″ and “fatty acid beta-oxidation using acyl-CoA dehydrogenase,” as well as terms related to monocyte chemotaxis, such as “regulation of monocyte chemotaxis,” were also enriched (Appendix Fig. 1B). These findings suggest that mitochondrial respiration and acyl-CoA dehydrogenase-related lipid metabolism are involved in the pathological process of inflammatory hyperplasia in the OA synovium.

Fig. 1.

Differential Gene Expression and Pathway Enrichment in OA Synovitis. (A) Heatmap showing the hierarchical clustering of differentially expressed genes (DEGs) between synovial samples from MIA-induced OA rats and control. (B) KEGG pathway enrichment analysis of upregulated DEGs. (C) KEGG pathway enrichment analysis of downregulated DEGs. (D–F) Gene Ontology (GO) analysis of significant DEGs. (G) Network plot illustrating the correlation of GO terms. (H) Gene Set Enrichment Analysis (GSEA) showing significantly enriched gene sets in the OA group compared to controls.

3.2. ACSL1 was upregulated in OA hyperplastic synovium

Our previous research has demonstrated that the M1 inflammatory polarization of macrophages in synovitis is crucial for regulating the progression of osteoarthritis [26,27]. In this study, we analyzed RNA-seq data (GSE53986) of LPS-stimulated bone marrow-derived macrophages (BMDMs) obtained from the NCBI database (Fig. 2A). KEGG and GO analyses revealed that LPS stimulation of macrophages activated classical pathways such as the "TNF signaling pathway", "NF-kappa B signaling pathway", and various infection-related pathways. Notably, the "Lipid and atherosclerosis" pathway also garnered our attention (Fig. 2B). By using Venn diagrams to find the intersection of several GO entries related to lipid metabolism, we found that the differential genes ACSL1, ACSL3, and ACSL4 are jointly involved in the regulation of several pathways (Fig. 2C). Among them, ACSL1 was the most significantly differentially expressed between the LPS-stimulated and control groups(Fig. 2D). To further elucidate the changes in the expression of ACSL1, ACSL3, and ACSL4 in synovial tissues from OA patients, we collected synovial tissues from OA and non-OA patients. Western blot analysis showed that the levels of inflammatory factors MMP13 and TNF-α were higher in the synovial tissues of OA patients, and the expression of ACSL1 was significantly increased in comparison with that of controls, while the expression levels of ACSL3 and ACSL4 showed no significant difference in the control group (Fig. 2E and F). Moreover, H&E staining and the Krenn synovitis scoring system were used to differentiate the severity of synovitis, OA was categorized into moderate and severe stages according to the Kellgren–Lawrence (K-L) score. Hematoxylin and eosin (H&E) staining indicated that synovial tissue from patients with moderate OA exhibited significantly more inflammatory hyperplasia and cell infiltration compared to controls. The hyperplasia and infiltration were even more pronounced in the synovium of patients with severe OA (Fig. 2G and H). Subsequently, immunofluorescence staining was performed to validate the expression of ACSL1 in synovial tissue. ACSL1 expression was significantly upregulated in the hyperplastic synovium of OA patients, with even more pronounced upregulation in the synovium of patients with severe OA (Fig. 2G and H). Furthermore, an OA model was constructed by performing destabilization of the medial meniscus (DMM) surgery on C57BL/6 mice, and specimens were collected at 4 and 8 weeks postoperatively. H&E staining results indicated that the mice showed more severe synovitis and higher Krenn synovitis scores at 8 weeks compared to 4 weeks after surgery (Fig. 2I and J). These findings indicate that in the hyperplastic synovial tissues of OA, the expression of ACSL1, a key protein regulating lipid metabolism in macrophages, is significantly increased.

Fig. 2.

ACSL1 was upregulated in OA hyperplastic synovium. (A) Heatmap showing hierarchical clustering of differentially expressed genes (DEGs) in Microarray Chip Data (GSE53986) of LPS-stimulated bone marrow-derived macrophages (BMDMs) compared to control. (B) KEGG pathway enrichment analysis highlighting pathways activated by LPS stimulation, including "TNF signaling pathway," "NF-kappa B signaling pathway," and "Lipid and atherosclerosis." (C) Venn diagram showing the overlap of GO terms related to lipid metabolism, identifying ACSL1, ACSL3, and ACSL4 as key regulators. (D) Violin plot displaying normalized expression levels of ACSL1, ACSL3, and ACSL4 in control and LPS-stimulated BMDMs. (E) Western blot analysis of ACSL1, ACSL3, ACSL4, TNF-α, and MMP13 in synovial tissues from severe OA and non-OA patients. n = 6 per group. (F) Quantification of protein expression levels of ACSL1, ACSL3, ACSL4, TNF-α, and MMP13 normalized to GAPDH. (G) Hematoxylin and eosin (HE) staining and immunofluorescence staining of ACSL1 in synovial tissues from control, moderate OA, and severe OA patients. n = 6 per group, Scale bar: 50 μm. (H) Krenn synovitis scores(η² = 0.70) and percentage of ACSL1 positive cells (η² = 0.82) in (G). (I) HE staining and immunofluorescence staining of ACSL1 in synovial tissues from control mice and destabilization of the medial meniscus (DMM) model mice at 4 and 8 weeks post-surgery. n = 6 per group, Scale bar: 50 μm. (J) Krenn synovitis scores (η² = 0.74) and percentage of ACSL1 positive cells (η² = 0.71) in (I). ∗P < 0.05, ∗∗P < 0.01, ns not significant. Data are shown as means ± SD. Statistical significance was determined by unpaired Student's t-test for two-group comparisons or one-way ANOVA for multiple groups.

3.3. The expression of ACSL1 in macrophages determines their inflammatory state

The expression of ACSL1 in macrophages is critical in determining their inflammatory state. To identify the specific cell types expressing ACSL1 in synovial tissue, we conducted immunofluorescence staining to co-localize CD68 (a macrophage marker) with ACSL1. Significant co-localization of both markers was observed in the lining layer of the synovium, with notably enhanced fluorescence intensity in the hyperplastic synovium of osteoarthritis (OA) (Fig. 3A and B), suggesting that ACSL1 is primarily expressed in synovial macrophages. However, immunofluorescence staining also revealed partial co-localization with CD55 (a marker of fibroblast lining cells), but no significant co-localization with CD34 (a marker of sublining fibroblasts) [28] (Appendix Fig. 1C), indicating that ACSL1 is expressed in multiple cell types within the synovium. Similarly, LPS-stimulated bone marrow-derived macrophages (BMDMs) in vitro exhibited increased pseudopod formation and elevated ACSL1 expression levels (Fig. 3C and D). Additionally, LPS stimulation induced the upregulation of iNOS and CD86 (markers of M1 polarization) and the downregulation of CD206 (a marker of M2 polarization) in primary macrophages, accompanied by increased ACSL1 expression (Fig. 3E and F). To further investigate the role of ACSL1 in macrophage polarization and inflammation in OA, we constructed lentiviral vectors carrying shRNA targeting ACSL1 (shACSL1-1- shACSL1-3). The effectiveness of ACSL1 knockdown in primary chondrocytes and BMDMs was confirmed by Western blot analysis. Among the shRNA constructs, shACSL1-2 and shACSL1-1 exhibited the most effective knockdown (Appendix Fig. 2A). Knocking down ACSL1 in macrophages by two pairs of shACSL1 (shACSL1-2 and shACSL1-1) significantly diminished pseudopod formation and reduced iNOS expression (Fig. 3G and H). Further qRT-PCR analysis demonstrated that shACSL1-2 and shACSL1-1 reduced the elevated mRNA levels of Nos2 and Cd86 in LPS-stimulated macrophages while also mitigating the LPS-induced decrease in Cd206 (Fig. 3I). Additionally, shACSL1-2 and shACSL1-1 decreased the production of inflammatory cytokines TNF-α, IL-1, and IL-6 at both the mRNA and protein levels in macrophages (Fig. 3J and K). For subsequent experiments, shACSL1-2 (referred to as shACSL1) was selected based on its efficient ACSL1 knockdown efficacy. In conclusion, these results indicate that ACSL1 plays a crucial role in regulating M1 polarization and inflammatory responses in macrophages.

Fig. 3.

ACSL1 regulates the inflammatory state of macrophages. (A) Immunofluorescence staining of synovial tissue showing co-localization of CD68 (macrophage marker, red) and ACSL1 (green) in control (Ctrl) and osteoarthritis (OA) samples. n = 6 per group, Scale bar: 50 μm. (B) Quantification of ACSL1-CD68 positive cells (Cohen's d = 2.02) in (A). (C) Immunofluorescence staining of ACSL1 and DAPI in bone marrow-derived macrophages (BMDMs) stimulated with LPS and control. n = 6 per group, Scale bar: 50 μm. (D) Quantification of fluorescence intensity of ACSL1 in (C). (E) Western blot analysis of ACSL1, CD206 (M2 marker), CD86 (M1 marker), and iNOS (M1 marker) in BMDMs stimulated with LPS and control. n = 6 per group. (F) Quantification of protein expression levels normalized to GAPDH in (E). (G) Immunofluorescence staining was performed to detect iNOS (red) and Phalloidin (green) in primary macrophages (top panel), and co-staining of ACSL1 (green) and DAPI (blue) (bottom panel) in control cells treated with shNC (Ctrl), LPS-treated cells with shNC (LPS), LPS + shACSL1-2-treated cells (shACSL1-2), and LPS + shACSL1-1-treated cells (shACSL1-1). n = 6 per group. Scale bar: 50 μm. (H) Quantification of iNOS fluorescence intensity in (G). (I) qRT-PCR analysis of mRNA expression levels of Nos2, Cd86, Cd206, and Acsl1 in macrophages treated with Ctrl, LPS, shACSL1-2, and shACSL1-1. (J) qRT-PCR analysis of mRNA expression levels of inflammatory cytokines Tnf-α, Il-1, and Il-6 in BMDMs treated with Ctrl, LPS, shACSL1-2, and shACSL1-1. (K) ELISA quantification of IL-1β, IL-6, and TNF-α levels in macrophage supernatants for Ctrl, LPS, and shACSL1-2. n = 6 per group. ∗P < 0.05, ∗∗P < 0.01, ns not significant. Data are shown as means ± SD. Statistical significance was determined by unpaired Student's t-test for two-group comparisons or one-way ANOVA for multiple groups.

3.4. The supernatant derived from macrophages with ACSL1 knockdown attenuates cartilage explant degeneration

To assess the effects of ACSL1 on macrophage programming and chondrocyte homeostasis, we utilized cartilage explants from 3-week-old C57BL/6 mice. The loss of toluidine blue staining, indicative of cartilage degeneration, was significantly reduced after co-culture with supernatants from LPS-stimulated ACSL1-knockdown BMDMs compared to controls. This was accompanied by increased expression of COL2 and decreased levels of ADAMTS5 and MMP13 in the cartilage (Fig. 4A and B). In vitro toluidine blue staining of primary chondrocytes revealed that ACSL1 knockdown alleviated the collagen degradation effects of the supernatant from inflammatory macrophages on chondrocytes. Furthermore, β-galactosidase staining indicated that ACSL1 knockdown alleviated the pro-senescent effects of the supernatant from inflammatory macrophages on chondrocytes. Additionally, in vitro studies revealed that the supernatant from ACSL1-knockdown BMDMs decreased the expression of γH2AX, a DNA damage marker, and reduced reactive oxygen species (ROS) induction in primary chondrocytes (Fig. 4C and D). Further qRT-PCR analysis showed that ACSL1 knockdown in inflammatory macrophages inhibited the expression of the catabolic genes Mmp3, Mmp13, and Adamts5, which had been induced in chondrocytes by inflammatory supernatant. Additionally, ACSL1 knockdown promoted the expression of the anabolic and differentiation factor Sox9 in chondrocytes, suggesting a shift towards a more anabolic state in the presence of reduced inflammatory signaling (Fig. 4E). Similarly, at the protein level, we observed that the supernatant from ACSL1-knockdown macrophages significantly reduced the expression of senescence markers P16 and P21 compared to the control group. Additionally, it decreased the levels of degradation markers ADAMTS5 and MMP13. Furthermore, the supernatant mitigated the reduction in synthesis markers SOX9 and COL2, induced by the inflammatory macrophage supernatant, indicating a protective effect on chondrocyte homeostasis (Fig. 4F and G). However, transfection of shACSL1 into primary chondrocytes in vitro had no significant effect on chondrocyte anabolism, catabolism, or senescence (Appendix Fig. 2B–E). Overall, these findings demonstrate that the loss of ACSL1 attenuates OA progression by downregulating the secretion of inflammatory cytokines from M1-polarized macrophages.

Fig. 4.

Supernatant from macrophages affected chondrocyte homeostasis. (A) Toluidine blue (TB) staining and immunofluorescence staining for COL2, ADAMTS5, and MMP13 in cartilage explants from 3-week-old C57BL/6 mice co-cultured with supernatants from shACSL1-BMDMs (shACSL1 transfection followed by addition of LPS stimulation) and control supernatants (shNC transfection with LPS stimulation). n = 6 per group. Scale bar: 50 μm. (B) Quantification of COL2, ADAMTS5, and MMP13 positive cells in (A). (C) TB staining, SA-β-galactosidase (SA-β-Gal) staining, and immunofluorescence staining for γH2AX in primary chondrocytes treated with supernatants from shACSL1 BMDMs and control. n = 6 per group. Scale bar: 20 μm, 200 μm. (D) Quantification of SA-β-Gal positive cells, γH2AX positive cells, and ROS fluorescence intensity in (C). (E) qRT-PCR analysis of mRNA expression levels of Mmp3, Mmp13, Sox9, and Adamts5 in primary chondrocytes treated with supernatants from shACSL1 BMDMs and control. (F) Western blot analysis of senescence markers (P16, P21), degradation markers (ADAMTS5, MMP13), and synthesis markers (SOX9, COL2) in primary chondrocytes treated with supernatants from shACSL1 BMDMs and control. n = 6 per group. (G) Quantification of protein expression levels normalized to GAPDH in (F). ∗P < 0.05, ∗∗P < 0.01, ns not significant. Data are shown as means ± SD. Statistical significance was determined by unpaired Student's t-test for two-group comparisons.

3.5. Knocking down ACSL1 in macrophages alleviates experimental osteoarthritis

To further clarify the regulatory role of ACSL1 in OA, an OA model was created by performing destabilization of the medial meniscus (DMM) surgery on C57BL/6 mice. Intra-articular injections of AAV-shACSL1 were subsequently administered to knock down ACSL1 in vivo (Fig. 5A). Results showed that eight weeks after DMM surgery, mice exhibited significantly greater cartilage destruction compared to four weeks post-DMM surgery (Appendix Fig. 3A and B) and the control group, with higher OARSI scores and a marked decrease in the ratio of hyaline cartilage to calcified cartilage. Additionally, mice showed synovial hyperplasia and inflammatory infiltration at eight weeks post-DMM surgery, and AAV-shACSL1 attenuated the inflammatory infiltration (Fig. 5B and C, Appendix Fig. 3A and B). Furthermore, the Von Frey mechanical pain threshold assay showed that intra-articular injection of AAV-ACSL1 alleviated the reduction in pain threshold during OA progression (Appendix Fig. 3C). DMM surgery promoted catabolism and inhibited anabolism in chondrocytes, as evidenced by increased ADAMTS5 and MMP13 expression and reduced COL2 expression. No significant difference in cartilage ACSL1 expression was observed at 8 weeks after DMM surgery compared to controls; however, shACSL1 was effectively knocked down in chondrocytes (Appendix Fig. 2F and G). AAV-shACSL1 mitigated the elevated catabolism in OA and restored anabolic activity (Fig. 5D and F, Appendix Fig. 3D and E). Synovial ACSL1 expression significantly increased with inflammatory proliferation of the synovial tissue, but AAV injection into the joint cavity significantly suppressed ACSL1 expression. Furthermore, synovial tissues expressed higher levels of iNOS (a macrophage M1 polarization marker) and lower levels of CD206 at 8 weeks post-DMM surgery, but intra-articular injection of AAV-shACSL1 reduced iNOS expression in the hyperplastic synovium (Fig. 5E and G, Appendix Fig. 3F and G). These results indicate that intra-articular injection of AAV-shACSL1 to knock down ACSL1 in synovial tissue alleviates cartilage destruction, synovial inflammation, and catabolic processes in chondrocytes in OA.

Fig. 5.

Knockdown of ACSL1 in macrophages alleviates experimental OA. (A) Schematic representation of the experimental design. (B) Safranin O/Fast Green and Hematoxylin and Eosin (H&E) staining of cartilage and synovial tissue from sham, DMM 8 weeks (DMM 8W), and AAV-shACSL1 groups. n = 6 per group. Scale bar: 200 μm, 100 μm, 50 μm. (C) Quantification of OARSI scores(η² = 0.89), HC/CC ratios(η² = 0.84), and Krenn synovitis scores(η² = 0.82) in (B). (D) Immunofluorescence staining of ADAMTS5, COL2, and MMP13 in cartilage tissue from sham, DMM 8W, and AAV-shACSL1 groups. Scale bar: 100 μm. n = 6 per group. (E) Immunofluorescence staining of ACSL1, iNOS, and CD206 in synovial tissue from sham, DMM 8W, and AAV-shACSL1 groups. n = 6 per group. Scale bar: 50 μm. (F) Quantification of ADAMTS5(η² = 0.83), COL2(η² = 0.85), and MMP13(η² = 0.81) positive cells from (D). (G) Quantification of ACSL1(η² = 0.94), iNOS (η² = 0.88), and CD206(η² = 0.82) positive cells from (E). ∗P < 0.05, ∗∗P < 0.01, ns not significant. Data are shown as means ± SD. Statistical significance was determined using one-way ANOVA followed by Tukey's post-hoc test for multiple comparisons where necessary.

3.6. Knockdown of ACSL1 inhibits the inflammatory polarization of macrophages by reducing lipid peroxidation

To investigate the potential mechanisms by which ACSL1 regulates macrophage inflammatory polarization, we analyzed the MIA-induced microarray sequencing dataset of OA synovial membranes (GSE206678) and the high-throughput sequencing dataset of LPS-stimulated primary BMDMs (GSE53986). GO terms related to lipid metabolism were significantly enriched in the GSE206678 dataset (Fig. 1B), while GO terms related to inflammation-regulated signaling pathways and lipid metabolism were also enriched in the GSE53986 dataset (Fig. 2B). Therefore, we examined the changes in lipid peroxidation during macrophage inflammatory polarization, a process closely linked to ACSL1 function. Mitochondria are the primary organelles responsible for the production of reactive oxygen species (ROS) within cells. MitoTracker was used to track mitochondrial location, and MitoSOX detected mitochondrial oxidative stress. LPS stimulation of primary BMDMs significantly increased MitoSOX fluorescence intensity and its distribution compared to the control group. However, intracellular knockdown of ACSL1 markedly reduced MitoSOX fluorescence intensity, indicating that ACSL1 knockdown can alleviate LPS-induced ROS production in macrophages (Fig. 6A and B). Additionally, BODIPY 581/591 C11 was used to label oxidized lipids within macrophages, and shACSL1 significantly mitigated the production and accumulation of oxidized fatty acids induced by LPS in primary BMDMs, as evidenced by reduced FITC green fluorescence intensity in the cytoplasm (Fig. 6C and D). Flow cytometry analysis further revealed that LPS stimulation caused a rightward shift in the FITC-A fluorescence absorption peak, indicating a significant increase in the accumulation of oxidized lipids within inflammatory macrophages. Knocking down ACSL1 reduced the LPS-induced accumulation of oxidized lipids (Fig. 6E and F). To confirm the regulatory role of ACSL1 in macrophage lipid peroxidation in vivo, PTGS2, a marker of lipid peroxidation, was examined and found to be significantly upregulated in the hyperplastic synovium of OA. Intra-articular administration of AAV-shACSL1 significantly attenuated PTGS2 expression levels. Furthermore, the antioxidant proteins SLC7A11 and GPX4, which were significantly downregulated in OA synovium, showed partial restoration after ACSL1 knockdown (Fig. 6G and J). In addition, shACSL1 inhibited the LPS-induced increase in NRF2 and the decrease in GPX4 at the protein level (Fig. 6H and I). These findings suggest that ACSL1 plays a critical role in lipid peroxidation in macrophages in response to inflammatory stimuli.

Fig. 6.

ACSL1 regulates lipid peroxidation in macrophages. (A) Immunofluorescence staining showing MitoTracker (green) and MitoSOX (red) in primary BMDMs treated with control (Ctrl), LPS, and LPS with ACSL1 knockdown (shACSL1). n = 6 per group. Scale bar: 20 μm. (B) Quantification of MitoSOX mean fluorescence intensity (MFI) in (A). (C) Immunofluorescence staining with BODIPY 581/591 C11 (TRITC channel for non-oxidized lipids, FITC channel for oxidized lipids) in primary BMDMs treated with Ctrl, LPS, and LPS + shACSL1. n = 6 per group. Scale bar: 50 μm. (D) Quantification of FITC channel MFI in (C). (E) Flow cytometry analysis of FITC-A fluorescence absorption peak in primary BMDMs treated with Ctrl, LPS, and LPS + shACSL1. n = 6 per group. (F) Quantification of FITC-A MFI in (E). (G) Immunofluorescence staining of PTGS2, SLC7A11, and GPX4 in synovial tissue from sham, DMM 8W, and shACSL1 groups. n = 6 per group. Scale bar: 50 μm. (H) Western blot analysis of ACSL1, NRF2, and GPX4 in primary macrophages treated with Ctrl, LPS, and LPS + shACSL1. n = 6 per group. (I) Quantification of protein expression levels normalized to GAPDH in (H). (J) Quantification of PTGS2(η² = 0.75), SLC7A11(η² = 0.87), and GPX4(η² = 0.84) positive cells in (G). ∗P < 0.05, ∗∗P < 0.01, ns not significant. Data are shown as means ± SD. Statistical significance was determined using one-way ANOVA followed by Tukey's post-hoc test for multiple comparisons where necessary.

3.7. ACSL1 regulates lipid peroxidation through the activation of the IκB/NF-κB pathway

To clarify how ACSL1 regulates lipid peroxidation, we analyzed microarray data from LPS-stimulated macrophages, revealing significant activation of the IκB/NF-κB signaling pathway in both the "TNF signaling pathway" and "lipid and atherosclerosis" entries. Differential expression of the IκB/NF-κB signaling pathways in these entries suggests that IκB/NF-κB may be a key pathway for LPS-stimulated lipid peroxidation in macrophages (Appendix Fig. 4A–C). Immunofluorescence staining revealed that phosphorylation of IκBα (p-IκBα) significantly increased in the synovial tissues of moderate and severe OA mice compared to controls (Fig. 7A and B). Additionally, silencing ACSL1 via AAV-shACSL1 significantly reduced p-IκBα-positive cells in DMM-induced OA mice (Fig. 7C and D). Western blot analysis revealed that LPS treatment significantly activated the NF-κB signaling pathway and suppressed GPX4 expression in BMDMs, effects that were significantly attenuated by ACSL1 knockdown (shACSL1). Conversely, erastin (5 μM, 24 h), a known ferroptosis inducer, enhanced IκBα and p65 phosphorylation and partially restored GPX4 expression (Fig. 7E). Immunofluorescence analysis further confirmed that ACSL1 knockdown (shACSL1) attenuated LPS-induced phosphorylation of p65 (p-p65) in macrophages (Fig. 7F). Mechanistic studies demonstrated that plasmid-mediated ACSL1 overexpression alone did not significantly affect NF-κB activation (p-p65/p-IκBα), M1 polarization markers (iNOS expression), or oxidative stress indicators (GPX4 levels). However, LPS stimulation combined with ACSL1 overexpression synergistically enhanced NF-κB activation, promoted M1 polarization, and exacerbated lipid peroxidation. The NF-κB inhibitor BAY 11–7082 (BAY11) effectively attenuated these amplified effects (Fig. 7G). qRT-PCR analysis demonstrated that ACSL1 overexpression potentiated LPS-induced pro-inflammatory cytokine (IL-1β, IL-6, TNF-α) expression, effects that were significantly attenuated by BAY11 treatment (Fig. 7H). Taken together, our findings demonstrate that ACSL1 potentiates macrophage IκB/NF-κB signaling activation via lipid peroxidation-dependent mechanisms, thereby driving inflammatory polarization and subsequent cytokine secretion in LPS-stimulated macrophages.

Fig. 7.

ACSL1 regulates lipid peroxidation through the activation of the IκB/NF-κB pathway. (A) Immunofluorescence staining for phosphorylated IκBα (p-IκBα, green) in synovial tissue from control (Ctrl), moderate OA, and severe OA mice. n = 6 per group. Scale bar: 50 μm. (B) Quantification of p-IκBα positive cells(η² = 0.69) in (A). (C) Immunofluorescence staining for p-IκBα (red) in synovial tissue from sham, DMM 8 weeks (DMM 8W), and AAV-shACSL1-treated mice. n = 6 per group. Scale bar: 50 μm. (D) Quantification of p-IκBα positive cells(η² = 0.75) in (C). (E) Western blot analysis of ACSL1, GPX4, p-IκBα, IκBα, p-p65, and p65 in primary macrophages treated with Ctrl, LPS, LPS + shACSL1, and LPS + shACSL1+Erastin. (F)Immunofluorescence staining for phosphorylated p65 (p-p65, green) in primary BMDMs treated with shNC (Ctrl), LPS with shNC (LPS), and LPS with ACSL1 knockdown (shACSL1). Scale bar: 10 μm. (G) Western blot analysis of ACSL1, iNOS, GPX4, p-p65, p65, p-IκBα, and IκBα in primary BMDSs treated with OE-NC, LPS, OE-ACSL1, LPS + OE-ACSL1, LPS + BAY11-7082, and LPS + OE-ACSL1+ BAY11-7082. (H) qRT-PCR analysis of mRNA expression levels of Il-1, Il-6, and Tnf-α in primary chondrocytes treated with OE-NC, LPS, OE-ACSL1, LPS + OE-ACSL1, LPS + BAY11-7082, and LPS + OE-ACSL1+ BAY11-7082. ∗P < 0.05, ∗∗P < 0.01, ns not significant. Data are shown as means ± SD. Statistical significance was determined using one-way ANOVA followed by Tukey's post-hoc test for multiple comparisons where necessary.

4. Discussion

This study revealed that ACSL1 expression was abnormally elevated in hyperplastic synovial tissues of OA. In vitro ACSL1 knockdown significantly reduced macrophage M1 polarization under inflammatory conditions, while trauma-induced OA in mouse knees was alleviated by joint injection of AAV-shACSL1. Further experiments showed that ACSL1 predominantly regulated macrophage lipid peroxidation, influencing the inflammatory state of macrophages. ACSL1 facilitates the secretion of inflammatory factors through activation of the IκB/NF-κB signaling pathway. These findings indicate that ACSL1 regulates lipid metabolism disorders and inflammatory proliferation in the OA synovium (Fig. 8).

Fig. 8.

Recent studies have highlighted the critical role of synovial macrophages as major inflammatory cells in OA pathogenesis [29]. The progression of synovitis is associated with increased cartilage damage [30]. Synovitis scoring systems based on microscopic histological features have been developed, such as the Krenn's score [24], which assesses synovial hyperplasia, stromal cell activation, and inflammatory infiltration. Radiographic progression, including the development of erosions in hand OA and accelerated knee OA, is associated with synovitis [31,32]. Despite these associations, the role of synovitis in OA progression remains unclear. Matrix-assisted laser desorption ionization mass spectrometry imaging (MALDI-MSI) revealed significant alterations in the spatial distribution of fatty acids (particularly arachidonic acid), glycerophospholipids (including phosphatidylcholine), and sphingolipids within the synovial tissues of OA patients [33]. Comprehensive analysis of fatty acid profiles in OA synovial tissues demonstrated distinct expression patterns between synovial tissues and synovial fluid compartments [34,35]. A recent preprint study (bioRxiv, 2023) by Ouyang and colleagues elucidated a novel mechanism whereby senescent fibroblast subpopulations promote cartilage degradation in OA through paracrine secretion of saturated fatty acids [36]. Single-cell RNA sequencing analysis of infrapatellar fat pads conducted by Ding et al. further delineated the pivotal role of the lipid metabolism regulator APOE in orchestrating fibroblast-chondrocyte crosstalk and modulating macrophage–chondrocyte interactions. [37]. Furthermore, transcriptomic profiling of 70 OA and 36 control synovial specimens identified significant enrichment of lipid metabolism-related Gene Ontology (GO) terms in specific synovial cell subpopulations. Notably, in a high-fat diet (HFD)-induced rat OA model, they observed concurrent synovial inflammatory lipolysis and epithelial–mesenchymal transition, accompanied by enhanced proliferation, invasive capacity, and pro-inflammatory/pro-metabolic activities in synovial fibroblasts [38]. Collectively, these findings implicate dysregulated lipid metabolism in synovial tissues as a potential modulator of OA progression. Our analysis of rat knee OA synovial RNA sequencing from the GEO database revealed significant metabolic reprogramming, with marked dysregulation of mitochondrial metabolic pathways, as well as marked activation of pro-inflammatory signaling cascades. Nevertheless, several limitations should be acknowledged regarding the translational relevance of rat synovial tissue sequencing data, as the complex immune-metabolic regulatory effects mediated by monosodium iodoacetate (MIA)-induced joint injury may not completely recapitulate the multifaceted pathophysiology of human OA progression. We also analyzed microarray sequencing of inflammatory macrophages, identifying significant lipid metabolism disorders and differential expression of ACSL1, a key lipid metabolism regulator. Despite differences in substrate specificity, subcellular localization, and tissue distribution, all ACSL isozymes convert free long-chain fatty acids into fatty acyl-CoA esters, crucial in lipid biosynthesis and fatty acid degradation [39]. Beatty et al. validated the effectiveness of ACSL1 in regulating ferroptosis in breast cancer [13]. Another study demonstrated that ACSL1 deficiency critically mediates alcohol-induced lipotoxicity and cell death in alcohol-associated liver disease. These findings underscore the pivotal role of ACSL1 in cell death through its regulation of cellular lipid metabolism. However, the regulatory role of lipid metabolism disorders in macrophage populations in OA synovial inflammation remains insufficiently explored. Our findings confirmed increased ACSL1 expression in inflammatory hyperplastic synovium, and that ACSL1 knockdown alleviated LPS-induced inflammatory polarization, implicating its role in regulating macrophage inflammatory responses. While we found that ACSL1 was predominantly expressed in macrophages, the fluorescence co-localization of ACSL1 with CD55 in synovial membranes also suggests its multiple regulatory roles in the OA synovium, warranting further investigation into the specific mechanisms through which ACSL1 regulates synovial inflammatory hyperplasia. Nonetheless, a statistically significant age difference between the control group (healthy individuals) and the OA group was observed. Although OA predominantly affects older adults, this age discrepancy could serve as a potential confounding factor, as age is known to influence both inflammatory responses and disease progression in OA. Despite efforts to closely match the groups, this limitation should be considered when interpreting the results, as age-related variations in immune cell function, metabolic changes, and tissue repair mechanisms may contribute to the observed differences. Future studies, involving larger sample sizes and age-matched cohorts, are needed to further clarify the role of these factors in OA pathogenesis.

The role of synovial macrophages in driving the progression of OA has been extensively studied [40]. Single-cell sequencing of OA synovium with 10,640 cells identified pro-inflammatory macrophages as a key component of the HLA-DRA + subpopulation [41]. As a result, targeting pro-inflammatory macrophage infiltration into the synovium has been suggested as a therapeutic strategy [42,43]. However, the mechanisms through which macrophages regulate localized inflammatory proliferation remain poorly understood. During OA development, chondrocytes undergo transcriptional changes to produce MMPs [44], cartilage fragments, aggrecan, fibronectin, and other DAMPs, which in turn stimulate synoviocytes to adopt a chronic inflammatory profile [45,46]. Studies indicate that pro-inflammatory macrophages in OA synovium exhibit upregulated production of matrix metalloproteinases (MMP3, MMP13, and MMP9) and aggrecanases (ADAMTS4 and ADAMTS5), contributing to articular degeneration [47]. Additionally, the secretion of pro-inflammatory cytokines such as IL-1β, IL-6, and TNF-α promotes destructive processes in chondrocytes and mesenchymal cells, downregulating the synthesis of type II collagen and aggrecan, thereby limiting chondrogenesis [48]. Our study demonstrated that ACSL1 knockdown alleviated the elevated secretion of inflammatory cytokines, including IL-1, IL-6, and TNF-α, following LPS administration. Furthermore, ACSL1 knockdown significantly attenuated the pro-catabolic and pro-senescence effects of LPS-stimulated macrophage-conditioned media on both cartilage explants and primary chondrocytes. Importantly, we demonstrated that direct ACSL1 knockdown in primary chondrocytes did not significantly alter their anabolic/catabolic balance or senescence status. These findings suggest that ACSL1 primarily contributes to OA pathogenesis by modulating synovial inflammation and pro-inflammatory factor secretion rather than through direct effects on articular cartilage. These results suggest that increased synovial macrophage ACSL1 expression mediates pro-inflammatory effects that play a key role in accelerating cartilage degeneration.

Oxidative stress [49], chronic low-grade inflammation [50], decreased chondrocyte proliferation and extracellular matrix degradation [51] are considered major contributors to the pathogenesis of OA. Lipid peroxidation, a specific form of lipid oxidation, has attracted increasing attention due to its involvement in various pathological conditions [52]. Lipid peroxidation occurs through three mechanisms: (1) enzymatic oxidation, (2) non-enzymatic free radical-mediated oxidation, and (3) non-enzymatic free radical-independent oxidation [53]. Free radicals can attack lipids containing carbon–carbon double bonds, particularly polyunsaturated fatty acids (PUFAs), generating lipid peroxidation products. Lipid hydroperoxides (LOOH) are primary products that degrade into secondary aldehydes, including malondialdehyde (MDA), hexanal, and 4-hydroxy-2-nonenal (4-HNE) [54]. In this study, we demonstrate that ACSL1 functions as a critical regulator of lipid peroxidation in macrophages. ACSL1 knockdown markedly attenuated LPS-induced generation of lipid peroxidation products, mitochondrial ROS accumulation, and total lipid peroxidation levels. Notably, the ferroptosis inducer Erastin restored lipid peroxidation and reactivated NF-κB signaling, indicating that ACSL1 orchestrates LPS-induced inflammatory responses via lipid peroxidation-mediated regulation of NF-κB activation. Importantly, ACSL1 overexpression alone failed to trigger macrophage NF-κB activation (phospho-p65/phospho-IκBα), M1 polarization (iNOS), or lipid peroxidation (GPX4). These findings demonstrate that ACSL1 is insufficient to initiate inflammation without upstream NF-κB activation. Previous studies have also demonstrated that ACSL1 expression in macrophages is dependent on the induction of the interferon-β (TRIF1) signaling pathway by toll-like receptor 4 (TLR4) and its bridging protein TIR structural domain [55]. However, overexpression of ACSL1 in hepatocytes triggers intracellular triglyceride accumulation but does not promote fatty acid oxidation [56]. We hypothesize that compensatory mechanisms mediated by other ACSL family members inhibit NF-κB signaling activation in response to ACSL1 overexpression alone, providing a foundation for future mechanistic investigations. Furthermore, LPS stimulation induced a synergistic increase in NF-κB activity, M1 polarization, and lipid peroxidation in ACSL1-overexpressing macrophages. The NF-κB inhibitor BAY 11–7082 (BAY11) attenuated these effects, confirming that ACSL1-mediated pro-inflammatory and lipid peroxidation regulation depends on active NF-κB signaling and functions through a positive feedback loop. Quantitative RT-PCR analysis showed that ACSL1 overexpression enhanced LPS-induced expression of pro-inflammatory cytokines (IL-1β, IL-6, TNF-α), which was markedly reduced by NF-κB inhibition. Intriguingly, BAY11 partially inhibited LPS-induced ACSL1 upregulation, suggesting that ACSL1 may be a transcriptional target of NF-κB, though this requires further validation. Furthermore, intra-articular injection of AAV-shACSL1 significantly altered the expression of SLC7A11, PTGS2, and GPX4, indicating that ACSL1 mitigates synovial inflammatory hyperplasia through regulation of lipid peroxidation in macrophages.

In conclusion, our findings indicate that ACSL1 functions as a key regulator in synovial M1-polarized macrophages. ACSL1 alone is insufficient to initiate inflammation or lipid peroxidation responses, but rather functions as a downstream effector and amplifier of NF-κB signaling, regulating the pro-inflammatory effects of inflammatory stimuli on macrophages. Increased ACSL1 expression promotes lipid peroxidation in macrophages and elevates the production of inflammatory cytokines, driving synovial inflammatory hyperplasia and disrupting chondrocyte homeostasis, ultimately contributing to cartilage degradation in OA. ACSL1 represents a promising therapeutic target for the treatment of OA.

Credit author statement

Zihao Yao, Daozhang Cai and Haiyan Zhang conceived and designed the study; Yanhui Li, Changgui Peng, Xuming Li, and Junyu Jin performed the experiments; and Jianying Pan analyzed the data; and wrote and revised the manuscript.

Declaration of competing interest

A conflict of interest occurs when an individual's objectivity is potentially compromised by a desire for financial gain, prominence, professional advancement or a successful outcome. The Editors of the Journal of Orthopaedic Translation strive to ensure that what is published in the Journal is as balanced, objective and evidence-based as possible. Since it can be difficult to distinguish between an actual conflict of interest and a perceived conflict of interest, the Journal requires authors to disclose all and any potential conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: