Sex-Specific Differences in Extracellular Vesicle Protein Cargo in Synovial Fluid of Patients with Osteoarthritis

Women are at a significantly higher risk of developing osteoarthritis (OA) compared to males. The pathogenesis of osteoarthritis (OA) in women is poorly understood. Extracellular vesicles (EVs) have been shown to play an essential role in numerous signaling processes during the pathogenesis of age-related diseases via paracrine signaling. Molecular profiling of the synovial fluid-derived EVs cargo in women may help in the discovery of novel biomarkers and therapeutics for the treatment of OA in women. Previously, we reported that synovial fluid-derived EV miRNA cargo differs in a sex-specific manner. This study aims to characterize synovial fluid-derived EV protein cargo in OA patients. Our data showed sex-specific EVs protein content in OA. We found haptoglobin, orosomucoid, and ceruloplasmin significantly up-regulated, whereas apolipoprotein down-regulated in female OA EVs. In males, we discovered β-2-glycoprotein, and complement component 5 proteins significantly up-regulated and Spt-Ada-Gcn5 acetyltransferase (SAGA)-associated factor 29 down-regulated in male OA EVs. Database for Annotation, Visualization, and Integrated Discovery (DAVID) and QuickGO analysis revealed OA-specific protein involvement in several biological, molecular, and cellular pathways, specifically in inflammatory processes. In conclusion, synovial fluid EV protein content is altered in a sex-specific manner with OA, explaining the increased prevalence and severity of OA in women.


Introduction
Osteoarthritis (OA) of the knee joint is the most common form of arthritis and affects millions of people worldwide. It is the most common cause of disability among adults in the US, affecting approximately 27 million people in the US aged 25 and older, with increased prevalence in older individuals [1]. The prevalence is higher in women than men and increases after menopause [2].

Patient Samples
The ethical committee (Code: 657441-24) approved all methods, following the relevant guidelines and regulations of Augusta University. Discarded human synovial fluid waste samples used for this study were de-identified and did not require informed consent. The studies were completed with prior approval from the Augusta University Institutional Review Board (IRB). Knee joint synovial fluid from both healthy and osteoarthritic knees was obtained from patients undergoing arthrocentesis/total knee arthroplasty procedures. Donors with severe complications (diabetes, hypertension, HIV, and others) and synovial fluid contaminated with blood were excluded. The synovial fluid was transported to the laboratory and immediately used for exosome isolation after it was obtained from the operating room at the time of surgery.

Preparation of Exosome-Enriched Fractions
A step-wise centrifugation method was used to prepare exosome fractions [15,22,23]. To remove any cell debris, 1 mL of synovial fluid was diluted with 2 mL of phosphate-buffered saline (PBS) and briefly centrifuged at 3000 rpm for 20 min, followed by Total Exosome Isolation Reagent (Life Technologies, Carlsbad, CA, USA) to isolate exosomes as per manufacturer's protocol. This protocol involved initial precipitation followed by centrifugation. After centrifugation, pellets were dissolved in 200 µL of phosphate-buffered saline (PBS) as exosome-enriched fractions.

Exosome Protein Extraction, Digestion, and LC-MS/MS Analysis
Peptide digestion and mass spectrometry were performed as per our published method [24,25]. Initially, the exosome samples were dried overnight by lyophilization. Furthermore, 100 µL freshly made 50 mM ammonium bicarbonate buffer with 0.1% (w/v) RapiGest SF Surfactant (Waters), and 10 mM dithiothreitol were added into the sample tube to resuspend the exosomes and reduce the disulfide bonds at 60 • C for 30 min. The samples were then alkylated by iodoacetamide in the dark for 30 min, followed by digestion for 16 h using trypsin (Thermo Scientific #90057) at 37 • C. Trifluoroacetic acid was added to the sample tube to a final concentration of 0.1% (v/v) to stop digestion. The samples were then incubated at 37 • C for 40 min to cleavage the detergent. The samples were then centrifuged at 15,000× g for 5 min, and the supernatants were transferred into sample vials for LC-MS analysis.
Digested peptide samples were analyzed on an Orbitrap Fusion tribrid mass spectrometer (Thermo Scientific, New York, NY, USA), to which an Ultimate 3000 nano-UPLC system (Thermo Scientific) was connected [24,25]. Furthermore, 2 µL of peptide samples were first trapped on a Pepmap100 C18 peptide trap (5 µm, 0.3 × 5 mm) and then washed at 20 µL/min using 2% acetonitrile with 0.1% formic acid for 10 min [18,19]. Next, the cleaned peptides were washed off the trap and further separated on a Pepman 100 RSLC C18 column (2.0 µm, 75 µm × 150 mm) at 40 • C using a gradient of between 2% and 40% acetonitrile with 0.1% formic acid over 40 min at a flow rate of 300 nL/min [24,25]. LC-MS/MS analysis was performed using data-dependent acquisition in positive mode with the Orbitrap MS analyzer for precursor scans at 120,000 FWHM (full width at half maximum) from 300 to 1500 m/z, and the ion-trap MS analyzer for MS/MS scans at top-speed mode (3 s cycle time) [24,25]. Collision-induced dissociation method was used to fragment the precursor peptides with a normalized energy level of 30%. Raw MS and MS/MS spectrum for each sample were filtered and processed using the Proteome Discoverer software by Thermo Scientific (v1.4) and then submitted to SequestHT search algorithm against the Uniprot human database (10 ppm precursor ion mass tolerance: 10 ppm, product ion mass tolerance: 0.6 Da, static carbamidomethylation of +57.021 Da). The Percolator PSM validator algorithm was used for peptide spectrum matching validation and false discovery rate estimation.

Normalization, Statistical Analysis, and Pathway Analysis of Female-Specific Protein
The peptide spectrum match (PSM) count for each identified protein in the LC-MS/MS search results was used as a semi-quantitative measure for protein expression level. The PSM count for each protein in a specific sample was first normalized using the sum of the PSM counts for all proteins in that sample. Then, the mean PSM count for the three replicates in each group was calculated for each protein and further used for statistical analysis [24,25].
Protein content was compared between the OA vs. non-OA EVs. EdgeR R package was used to perform trimmed mean normalization (TMM), then the difference for protein expression between the groups (OA and non-OA) was analyzed. Proteins up-regulated or down-regulated with a p-value cutoff of 0.05 were considered differentially expressed for further analyses. Gene Ontology pathway analyses were conducted using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) [26,27] and QuickGO [28] on differentially expressed protein genes. Uniprot Knowledgebase (UniProtKB) protein descriptions and gene products were imported into DAVID and QuickGO for statistical analyses and GO term annotation based on integrated biological, molecular, and cellular pathways of the differentially expressed proteins.

EV Protein Cargo Differs Significantly between Male and Female Patients with OA
Previously, we characterized synovial fluid-derived EVs and reported that these EV's miRNA cargo alters in a gender-specific manner [15]. In this manuscript, we used the same EVs (previously characterized), and that is why we did not show the EV characterization data (please see Kolhe et al. 2017 [15]). In this study, we performed mass spectrometry analysis on synovial fluid-derived EVs proteins (male non-OA, n = 7 and OA, n = 7 and female non-OA, n = 8 and OA, n = 10). Previously, we have shown CD81, CD63, and Tsg101 markers on EVs using Western blot and CD9 using an immuno-gold label [15]. For this study, we used the same patient samples for mass spectrometry.
Based on mass spectrometry protein profiling, we identified multiple gender (male and female) specific differential proteins in OA and non-OA exosomes ( Table 1 and Figure 1). For example, we found haptoglobin, orosomucoid, and ceruloplasmin were significantly (p = 0.01) up-regulated, whereas apolipoprotein was down-regulated (p = 0.04) in female OA exosomes (Table 1a and Figure 2). In males, we discovered β-2-glycoprotein and complement component 5 significantly up-regulated (p = 0.003), and SAGA-associated factor 29 (p = 0.005) down-regulated in male OA exosomes (Table 1b and Figure 3). Our data, therefore, suggest that proteins carried by EVs in the synovial fluid are significantly altered with the osteoarthritic condition and are highly sex-specific.
Life 2020, 10, x FOR PEER REVIEW 4 of 13

EV Protein Cargo Differs Significantly Between Male and Female Patients with OA
Previously, we characterized synovial fluid-derived EVs and reported that these EV's miRNA cargo alters in a gender-specific manner [15]. In this manuscript, we used the same EVs (previously characterized), and that is why we did not show the EV characterization data (please see Kolhe et al. 2017 [15]). In this study, we performed mass spectrometry analysis on synovial fluid-derived EVs proteins (male non-OA, n = 7 and OA, n = 7 and female non-OA, n = 8 and OA, n = 10). Previously, we have shown CD81, CD63, and Tsg101 markers on EVs using Western blot and CD9 using an immuno-gold label [15]. For this study, we used the same patient samples for mass spectrometry.
Based on mass spectrometry protein profiling, we identified multiple gender (male and female) specific differential proteins in OA and non-OA exosomes ( Table 1 and Figure 1). For example, we found haptoglobin, orosomucoid, and ceruloplasmin were significantly (p = 0.01) up-regulated, whereas apolipoprotein was down-regulated (p = 0.04) in female OA exosomes (Table 1a and Figure  2). In males, we discovered β-2-glycoprotein and complement component 5 significantly upregulated (p = 0.003), and SAGA-associated factor 29 (p = 0.005) down-regulated in male OA exosomes (Table 1b and Figure 3). Our data, therefore, suggest that proteins carried by EVs in the synovial fluid are significantly altered with the osteoarthritic condition and are highly sex-specific. Differences between normal and OA patients were examined using Student's t-test, and only those that were significantly different at the p value 0.05 level were selected.

DAVID and QuickGO Analysis of Differentially Expressed Proteins
To analyze the functions and effects of the differentially expressed proteins on various pathways, Database for Annotation, Visualization, and Integrated Discovery (DAVID) and Quick GO annotation analyses were performed. DAVID annotation analysis revealed extensive involvement of these proteins in endopeptidase and hydrolase activities, vesicle transport and receptor-mediated

DAVID and QuickGO Analysis of Differentially Expressed Proteins
To analyze the functions and effects of the differentially expressed proteins on various pathways, Database for Annotation, Visualization, and Integrated Discovery (DAVID) and Quick GO annotation analyses were performed. DAVID annotation analysis revealed extensive involvement of these proteins in endopeptidase and hydrolase activities, vesicle transport and receptor-mediated endocytosis, and a robust, defensive immune response to stress and stimulus in the OA exosomes of synovial fluid in women. Several proteins are involved, and their expression up-regulated several folds in each of the processes, the most enriched being blood microparticle involvement, endocytosis, and endopeptidase and hydrolase activity. The details of the DAVID annotation analysis are shown in the Table 2. QuickGO annotation analysis revealed extensive involvement of these proteins in multiple biological (lipid transport, regulation of immune system process, lipoprotein metabolic process), molecular (hydrolase activity, ATP-dependent helicase activity, nucleic acid binding), and cellular pathways (extracellular region, nucleus, nucleoplasm) in OA exosomes. Furthermore, these proteins were shown to be primarily involved in regulating the immune system process in women with OA (Table 3). We also performed DAVID and QuickGO analysis on male EVs cargo proteins but did not find significant changes (Supplementary Tables S1 and S2).

Discussion
The synovial fluid consists of secretory products from synovium, cartilage, and other components of articular joints. Several recent studies have analyzed the synovial fluid for miRNAs, cytokines, and proteins to better understand the pathophysiological status of OA [13][14][15]29]. Recently, we reported that synovial fluid-derived EVs carry specific miRNAs in osteoarthritis patients, and importantly, these miRNAs were gender-specific. EVs are 40-100 nm in diameter and contain various components, including proteins, lipids, and miRNA [16]. Previously, we also characterized the synovial fluid-derived EVs of OA and non-OA patients [15]. We found that EVs isolated from the synovial fluid are round-shaped vesicles,~100 (±10) nm diameter size ranges, with no change in concentration between OA and non-OA patients [15]. We also found expression of CD9 (using electron microscopy immuno-gold staining) and Tsg101, CD63, and CD81 (using Western blot), which are the markers for EVs [15].
It has been previously reported that EV protein plays a vital role in various age-related and degenerative diseases [30][31][32][33]. Considering the critical role of EV protein cargoes in the progression of various age-related diseases, it is clinically relevant to analyze EVs' protein content of OA synovial fluid. We hypothesized that synovial fluid-derived EV protein content not only differs in OA condition, compared to non-OA, but also differs in a sex-specific manner. We analyzed the composition of EV protein cargo using mass spectrophotometry in the synovial fluid of males and females with OA compared with the controls.
Our mass spectrophotometry data showed that numbers of proteins were differentially present in the EV cargo of OA samples. Interestingly, as expected, female EV protein profiles were completely different to males'. In females, we found that haptoglobin, orosomucoid, and ceruloplasmin, were significantly up-regulated and apolipoprotein L down-regulated. In contrast, in males, β-2-glycoprotein and complement component 5 proteins were up-regulated and SAGA-associated factor 29 significantly down-regulated in OA EVs. To the best of our knowledge, our study is the first to characterize EV protein cargo from the synovial fluid of non-OA and OA patients in a sex-specific manner.
Our data indicate that these different EV protein cargoes can help to develop gender-specific biomarkers for diagnosis and understanding of OA pathophysiology. For example, the haptoglobin, which is elevated in female synovial fluid EV cargo, is known to be dysregulated in several other female-related pathological conditions [34][35][36][37]. Haptoglobin is increased during periods of inflammation and has been identified as an acute-phase glycoprotein [38]. Haptoglobin has been shown to have increased production during inflammation via IL-6 and TNF-alpha and serves as a modulator of inflammation [38,39]. Haptoglobin fragments are detectable in OA serum, indicating an alteration in protein pattern during OA [40,41]. Haptoglobin has also been shown to correlate with OA severity, with increased levels in the synovial fluid of patients with more symptomatic OA [42]. Elevated levels of haptoglobin correlate with an increased inflammatory state of cytokines, and cell-mediated inflammation [34][35][36][37]. Another important acute-phase protein, orosomucoid (also known as AGP-1), only up-regulated significantly in female synovial fluid-derived EVs cargo. This protein has been dysregulated in numerous disease processes, including liver disease, cancer, HIV, and other inflammatory conditions [43][44][45]. Orosomucoid has also been shown to interact with TLR-4 and CD14 and modulate immune response [45][46][47]. Another study has shown that orosomucoid is down-regulated in the presence of the female hormone estrogen via an estrogen receptor-dependent pathway [48]. This indicates that these proteins play an important role in gender-specific immunomodulation of synovial health. It is possible that differential EV protein cargo in female synovial fluid may be due to a decline in estrogen levels. It is well known that declining estrogen levels are inversely related to increased incidence and severity of OA in females [49][50][51][52]. We previously demonstrated that estrogen inhibitor treatment alters miRNA cargo in female synovial fibroblast cells [15]. Surprisingly, we did not find significant changes in pro-inflammatory cytokines (such as IL-1, IL-6, and TNF-a), which play an important role in OA progression.
To further investigate the role of these proteins in signaling pathways, we performed bioinformatics analysis using DAVID and QuickGO analysis on these differentially regulated proteins. The female EV protein cargo demonstrated a complex web of pathway connections in which these proteins interact to elicit several cellular, molecular, and biological processes. These proteins are involved in endopeptidase and hydrolase activity, immune receptor-mediated endocytosis, receptor-mediated phagocytosis, immune system regulation, and response to stress and activation. Previously, we reported that female synovial fluid-derived EV miRNA cargo also affects similar signaling pathways [15]; specifically, the immune system and response to stress signaling pathways. MicroRNA cargo from our previous study [15] and protein EV cargo from this study showed strong evidence that the immune system plays a massive role in the development of OA in females (Figure 4). These findings indicate a complex interplay of dysregulated synovial fluid-derived EV cargos (miRNAs and proteins) in female OA.
Life 2020, 10, x FOR PEER REVIEW 9 of 13 modulator of inflammation [38,39]. Haptoglobin fragments are detectable in OA serum, indicating an alteration in protein pattern during OA [40,41]. Haptoglobin has also been shown to correlate with OA severity, with increased levels in the synovial fluid of patients with more symptomatic OA [42]. Elevated levels of haptoglobin correlate with an increased inflammatory state of cytokines, and cellmediated inflammation [34][35][36][37]. Another important acute-phase protein, orosomucoid (also known as AGP-1), only up-regulated significantly in female synovial fluid-derived EVs cargo. This protein has been dysregulated in numerous disease processes, including liver disease, cancer, HIV, and other inflammatory conditions [43][44][45]. Orosomucoid has also been shown to interact with TLR-4 and CD14 and modulate immune response [45][46][47]. Another study has shown that orosomucoid is down-regulated in the presence of the female hormone estrogen via an estrogen receptor-dependent pathway [48]. This indicates that these proteins play an important role in gender-specific immunomodulation of synovial health. It is possible that differential EV protein cargo in female synovial fluid may be due to a decline in estrogen levels. It is well known that declining estrogen levels are inversely related to increased incidence and severity of OA in females [49][50][51][52]. We previously demonstrated that estrogen inhibitor treatment alters miRNA cargo in female synovial fibroblast cells [15]. Surprisingly, we did not find significant changes in pro-inflammatory cytokines (such as IL-1, IL-6, and TNF-a), which play an important role in OA progression.
To further investigate the role of these proteins in signaling pathways, we performed bioinformatics analysis using DAVID and QuickGO analysis on these differentially regulated proteins. The female EV protein cargo demonstrated a complex web of pathway connections in which these proteins interact to elicit several cellular, molecular, and biological processes. These proteins are involved in endopeptidase and hydrolase activity, immune receptor-mediated endocytosis, receptor-mediated phagocytosis, immune system regulation, and response to stress and activation. Previously, we reported that female synovial fluid-derived EV miRNA cargo also affects similar signaling pathways [15]; specifically, the immune system and response to stress signaling pathways. MicroRNA cargo from our previous study [15] and protein EV cargo from this study showed strong evidence that the immune system plays a massive role in the development of OA in females ( Figure  4). These findings indicate a complex interplay of dysregulated synovial fluid-derived EV cargos (miRNAs and proteins) in female OA. Previous studies have shown the significance of EVs in the pathogenesis of OA and their potential to use them as markers for disease severity and progression [18,53,54]. Our data strongly agrees with published literature but also suggests that female and male OA should be investigated separately to get more accurate and valuable information. Our data indicate that EV-derived protein plays a vital role in Life 2020, 10, 337 10 of 13 female OA and its pathogenesis, helping to explain the increased prevalence and severity of OA in women. Further in-depth in vitro and in vivo studies should be investigated, as prevention, screening, and treatment can be personalized based on gender for more clinical successful treatment of OA.
Our study does have limitations, the most important being that we had a small sample size. Larger sample sizes are needed, as well as various stages of OA in women and men should be investigated. To conclude, ours is the first study to show male and female-specific protein EV profiling in OA. Exosomes and their contents are vital to understanding the pathophysiology of OA in a gender-specific manner. These proteins contribute to several female-specific biologic, molecular, and cellular pathways and may help to explain the increased prevalence and severity of OA in women.