Fibrotic pathways and fibroblast-like synoviocyte phenotypes in osteoarthritis

Abstract

Osteoarthritis (OA) is the most common form of arthritis, characterized by osteophyte formation, cartilage degradation, and structural and cellular alterations of the synovial membrane. Activated fibroblast-like synoviocytes (FLS) of the synovial membrane have been identified as key drivers, secreting humoral mediators that maintain inflammatory processes, proteases that cause cartilage and bone destruction, and factors that drive fibrotic processes. In normal tissue repair, fibrotic processes are terminated after the damage has been repaired. In fibrosis, tissue remodeling and wound healing are exaggerated and prolonged. Various stressors, including aging, joint instability, and inflammation, lead to structural damage of the joint and micro lesions within the synovial tissue. One result is the reduced production of synovial fluid (lubricants), which reduces the lubricity of the cartilage areas, leading to cartilage damage. In the synovial tissue, a wound-healing cascade is initiated by activating macrophages, Th2 cells, and FLS. The latter can be divided into two major populations. The destructive thymocyte differentiation antigen (THY)1^─ phenotype is restricted to the synovial lining layer. In contrast, the THY1⁺ phenotype of the sublining layer is classified as an invasive one with immune effector function driving synovitis. The exact mechanisms involved in the transition of fibroblasts into a myofibroblast-like phenotype that drives fibrosis remain unclear. The review provides an overview of the phenotypes and spatial distribution of FLS in the synovial membrane of OA, describes the mechanisms of fibroblast into myofibroblast activation, and the metabolic alterations of myofibroblast-like cells.

Keywords: TGF-β; fibroblast to myofibroblast transition; fibrosis; inflammation; mechanical stress; metabolism; osteoarthritis; senescence.

Figure 1

Synovial joint architecture and stressors contributing to osteoarthritis (OA). Schematic representation of healthy joint physiology (left) and OA pathology (right). The joint comprises the femur and the tibia bone, covered with hyaline articular cartilage. To ensure smooth movement, the joint cavity is filled with synovial fluid, which also supplies the avascular cartilage. From the outside, the joint is enclosed by the joint capsule. This consists of an outer fibrous membrane and the inner synovial membrane. OA is characterized by cartilage degradation, bone erosion and bone cysts as well as synovial fibrosis and synovitis. Stress factors contributing to OA include mechanical overload, injury, inflammation, obesity, and aging. Figure was created with BioRender.com.

Figure 2

Synovial membrane architecture of the healthy (left) and osteoarthritic joint (right). The healthy synovium comprises a thin lining layer with barrier-forming CX₃CR1⁺ TREM2⁺ MERTK⁺ resident macrophages (type A synoviocytes) and CD55⁺ PRG4⁺ THY1⁻ fibroblasts (type B fibroblast-like synoviocytes). The lining layer separates the synovial cavity from the tissue. The sublining layer hosts various fibroblast and macrophage populations, adipocytes, and blood vessels. During osteoarthritis, the integrity of the barrier, maintained by tight junctions of resident macrophages, is disrupted in the lining layer. Figure was created with BioRender.com.

Figure 3

Synovial fibroblast subpopulations are depicted graphically, illustrating surface markers, transcriptional profiles, and functions of distinct subsets within an arthritic joint. The identification of these main subsets involved scRNA-seq studies, mass spectrometry, and histological evaluations. However, it is crucial to emphasize the presence of distinct activation states in human arthritic joints. Therefore, drawing from comprehensive analyses detailed in the text, we propose the existence of an additional lining and sublining layer myofibroblast-like phenotype. Broadly, lining layer FLS lacking thymocyte differentiation antigen 1 (THY1) expression play a pivotal role in lubrication, with this subset being more prominent in osteoarthritis (OA) compared to rheumatoid arthritis (RA). In contrast, sublining layer FLS expressing THY1 expand during inflammation and exhibit different spatial distributions within the joint. This illustration is derived from human data and the figure was created using BioRender.com.

Figure 4

Localization of α-SMA positive cells within the synovial membrane of patients with osteoarthritis. Representative image for PRG4 (yellow), α-SMA (magenta), THY1 (cyan), and DAPI (gray). Scale bars indicate 100 µm (78).

Figure 5

The fibroblast to myofibroblast transition. Schematic overview of stressors and corresponding phenotype alterations (left), and downstream intracellular processes (right) that contribute to myofibroblast differentiation and ultimately lead to synovial fibrosis. Fibroblasts are activated by various types of stimuli such as mechanical and oxidative stress, which initiate different intracellular processes. Initially the proto-myofibroblast develops followed by the myofibroblast. However, activation can sometimes be reversed or lead to apoptosis of the myofibroblasts. At the cellular level, the transition of fibroblast to myofibroblast leads to a pronounced increase in intracellular stress fibers containing alpha-smooth muscle actin (α-SMA) and the expression of collagen 1, extra domain A fibronectin (ED-A-FN), and extracellular matrix remodeling enzymes. Moreover, they produce cytokines such as transforming growth factor (TGF)-β, vascular endothelial growth factor (VEGF), connective tissue growth factor (CTGF), interleukin (IL)-1, IL-6, and IL-8 and are in close contact with their environment via integrins. Figure was created with BioRender.com.

Figure 6

TGF-β signaling via canonical and non-canonical pathways. Cell surface TGF-β type II receptor binds soluble active TGF-β, which causes association and phosphorylation of TGF-β type I receptor. TGF-β type I receptor/ALK5 activation can induce the canonical phosphorylation of Smad2/3 signaling or non-canonical signaling via non-Smad pathways by phosphorylation of, e.g., ERK1/2. Alternatively, TGF-β type I receptor/ALK1 activation can induce the non-canonical phosphorylation of Smad1/5/8. Phosphorylated Smad2/3 and Smad1/5/8 form complexes with Smad4, which translocate into the nucleus and regulate target gene expression. Inhibitory Smad6 and Smad7 can repress canonical or non-canonical signaling via Smad. Figure was created with BioRender.com.

Figure 7

Schematic overview of the different mechanical forces that physiologically act on the joint during locomotion. Figure was created with BioRender.com with elements from (148).

Figure 8

Metabolic shifts upon transition from fibroblast activation and expansion to senescent myofibroblasts. This transition involves an enhanced glycolysis, pentose phosphate pathway (PPP), fatty acid oxidation (FAO), lipolysis, and glutaminolysis as key processes in fibroblast activation and expansion. Quiescent myofibroblasts instead use glycolysis to generate acetyl-CoA fueling the TCA and mitochondrial oxidative phosphorylation to generate ATP and to feed lipid and fatty acid synthesis (FAS). Figure was created with BioRender.com.