As the name suggests, synovial fluid is the friction-minimising fluid from within the capsule of synovial joints. Synovial joints are the most common joint type in the human body, with examples including the hip, knee, and elbow. These joints have a propensity towards arthritides such as rheumatoid arthritis and osteoarthritis. As synovial fluid is the biofluid surrounding the tissues damaged/inflamed in these conditions, its composition in affected joints may be key to understanding pathogenesis and disease. In this search for answers, synovial fluid-derived extracellular vesicles (SF-EVs) have gained interest from researchers. EVs are nano-sized vesicles in which proteins, nucleic acids and small molecules from the cell of origin are encapsulated in a lipid bilayer. This cargo can be delivered to target cells, where it can have a variety of effects. These features make EVs attractive as biomarkers but also makes them vital for understanding disease pathogenesis, such as that of diseases of the synovial joints. In this article, we will discuss the composition of synovial fluid and its impact upon the isolation of EVs, before highlighting key SF-EV research.
The main cell type in synovial fluid are immune cells, which are thought to be a major source of SF-EVs.1 As with most inflammatory conditions, inflammation of the synovial joint is associated with a significant infiltration of leukocytes, making immune-derived EVs the main focus of SF-EV research.2 To enable effective, pure isolation of EVs, a basic understanding of the non-EV components of synovial fluid is essential.
Like cerebrospinal fluid, synovial fluid is an ultrafiltrate of plasma and has about one third of the total plasma protein concentration.3 Whilst protein content is comparatively low, the SF-EV concentration is unknown. This means that the EV-to-protein ratio cannot currently be compared to that of plasma – the biofluid in which most EV isolation methods have been optimised. What is known to be different to that of plasma is the viscosity of synovial fluid, which is about as viscous as raw egg white. Viscosity is the largest differentiating factor of synovial fluid when compared with other body fluids, as this viscosity is required for its lubricant properties within joints. This viscosity is mainly derived from the presence of the viscoelastic hyaluronan, which is present as a high molecular weight glycosaminoglycan polymer.4 Along with the proteoglycan lubricin, hyaluronan is secreted by fibroblast-like cells of the synovial membrane, which line the inward-facing side of the joint capsule.5 Both hyaluronan and lubricin contribute to the non-Newtonian viscosity of synovial fluid, which sees synovial fluid thinning under sheer force.6 Viscosity is also altered in disease.7 This has implications for EV isolation using methods affected by viscosity, such as ultracentrifugation (UC).
As eluded to above, one of the biggest considerations for isolating EVs from synovial fluid is hyaluronan, which has led to many studies including hyaluronidase in their workflows. However, the only reported impact of this is increasing 10,000 x g pellet particles and decreasing 100,000 x g pellet particles, whilst specifically increasing CD44+ EV detection by flow cytometry.8 SF-EVs may even be coated with hyaluronan.9 With only one study currently published on the matter, the impact of hyaluronidase on EV recovery has not yet reached consensus. There is, however, more evidence for the impact of EV isolation method on the quality of the EV isolate.
Foers et al. (2018) sought to compare conventional UC and density gradient UC (DG-UC) with size exclusion chromatography (SEC) for EV isolation from synovial fluid.10 UC and DG-UC similarly co-isolated large amounts of contaminant high density lipoproteins and had prevalent amorphous material present in transmission electron micrographs. With SEC, markers for high density lipoproteins were well separated from EV markers and electron micrographs were generally cleaner, meaning SEC was recommended as the best technique.10 Unfortunately, the SEC used in this study took a very long time to run (>150 minutes), making it unsuitable for high-throughput studies and unpalatable for even low-throughput research10. SEC using qEV columns can take as little as 15 minutes, making it much more scalable. The speediness of qEV columns was corroborated in a study by Chen et al. (2022) in which UC, precipitation and qEV SEC were compared. qEV columns were both the quickest and purest isolation method, removing most protein contamination.11 Unlike the long-running SEC columns used by Foers et al. (2018), qEV columns used by Chen et al. (2022) did not show fibrous protein webs in EV fractions. As such, not only does the literature suggests that SEC is the method of choice, but that qEV columns are preferable.10,11
Whilst several arthritides exist, the most prevalent are osteoarthritis and rheumatoid arthritis. Rheumatoid arthritis is described as an inflammatory arthritis and, whilst it is true that inflammation plays a greater role in the pathogenesis of rheumatoid arthritis, this does somewhat inaccurately diminish the role of inflammation in the perpetuation of osteoarthritis. Numerous studies into SF-EVs have compared the two, often with osteoarthritis used as a control of sorts due to ethical constraints collecting synovial fluid from healthy patients. As such, some perturbations in EVs in one or both conditions may be missed in these studies. Despite the drawbacks of the comparison between rheumatoid arthritis and osteoarthritis, several studies have found that there is a significantly higher number of EVs in rheumatoid arthritis as compared to osteoarthritis.12-14 This suggests that SF-EVs may be important in rheumatoid arthritis.
Most SF-EV studies have focused on identifying the origin of EVs focusing primarily on immune cell markers. These studies have shown that there is an enrichment in rheumatoid arthritis for EVs from fibroblasts13, T regulatory cells12, CD4+ T helper cells15, CD8+ cytotoxic T cells1,15, monocytes15,16, platelets17, and both granulocytes in general15 and neutrophils specifically12,13,16. In contrast, osteoarthritis is enriched for EVs derived from activated T cells18, macrophages18, B cells18 and natural killer cells18. But what could these immune-derived EVs be doing?
The proteins in SF-EVs from people with rheumatoid arthritis are enriched for complement and coagulation cascade proteins, and those involved in chemotaxis.14 One study19 found that SF-EVs from rheumatoid arthritis patients cause fibroblast-like cells of the synovial membrane to produce higher amounts of cytokines and other proteins with known roles in osteoarthritis and rheumatoid arthritis.20-22 When comparing the SF-EV protein changes in rheumatoid arthritis vs osteoarthritis between two studies13,14 (Figure 1), the top rheumatoid arthritis upregulated EV protein present in both datasets was the rheumatoid arthritis biomarker S100A9.23,24 Another protein upregulated in rheumatoid arthritis SF-EVs is vimentin (Figure 1). The citrullinated form of vimentin is present in SF-EVs25 and is an important autoantigen in a subset of rheumatoid arthritis patients.26,27 Other proteins on this list (Figure 1) have also been shown to be involved or altered in rheumatoid arthritis. It is possible, then that SF-EVs in rheumatoid arthritis perpetuate inflammation.
Furthermore, these SF-EVs in rheumatoid arthritis patients also contain miRNAs known to target cytokine pathways and are highly enriched for the anti-inflammatory pregnancy zone protein, suggesting an involvement in modulating inflammation.14,28 This dichotomy of pro- and anti-inflammatory properties could co-exist in the same individual EVs, or could be separated in different subpopulations of EVs from different cells of origin. It has been shown that medium sized SF-EVs from people with osteoarthritis represented the majority of B cell-derived EVs, whilst large and small EVs were more enriched for macrophage-derived EVs, suggesting that bias in EV size between studies may alter the proportion of different EV types in isolates.18 Studies in the field currently use a large range of EV isolation techniques, likely contributing to some of the lack of concordance seen between studies by isolating different populations. This could be improved by the adoption of a standardised procedure shown to perform well in synovial fluid separation, such as qEV columns.
In order to move the field towards producing clinical benefits, several key questions must be answered.
1. Are EVs displaying immune cell markers produced by these immune cells? When an EV is positive for an immune cell marker, the assumption is that the immune cell in question is the origin of the EV. However, it is not yet certain if this is true or whether these EVs originate from other cells aberrantly expressing these proteins as part of the joint pathology.
2. Are we missing potentially vital biomarkers or treatment candidates by lacking healthy controls? Nearly all the studies looking at SF-EVs treat osteoarthritis as control samples due to ethical constraints. However, immune dysregulation is clearly present in osteoarthritis joints.
3. Will standardisation of SF-EV isolation methods improve concordance between studies? There is currently substantial discordance between the main contributing EV source cells in different studies, indicating the need for larger cohorts, standardised methods and systematic reviews.
Studies of SF-EVs have pointed to a potential inflammatory (and/or anti-inflammatory) role, but the impact of SF-EVs on their surrounding tissues has yet to be adequately determined. Could these tiny vesicles hold the key to new therapeutic targets for inflammatory joint conditions?