Extracellular RNA (exRNA) has been characterized as a molecular alarm signal upon cellular stress or tissue injury and to exert biological functions as a proinflammatory, prothrombotic, and vessel permeability–regulating factor. In this study, we investigated the contribution of exRNA and its antagonist RNase1 in a chronic inflammatory joint disease, rheumatoid arthritis (RA). Upon immunohistochemical inspection of RA, osteoarthritis (OA), and psoriatic arthritis synovium, exRNA was detectable only in the RA synovial lining layer, whereas extracellular DNA was detectable in various areas of synovial tissue. In vitro, exRNA (150–5000 nt) was released by RA synovial fibroblasts (RASF) under hypoxic conditions but not under normoxia or TNF-α treatment. RNase activity was increased in synovial fluid from RA and OA patients compared with psoriatic arthritis patients, whereas RNase activity of RASF and OASF cultures was not altered by hypoxia. Reduction of exRNA by RNase1 treatment decreased adhesion of RASF to cartilage, but it had no influence on their cell proliferation or adhesion to endothelial cells. In vivo, treatment with RNase1 reduced RASF invasion into coimplanted cartilage in the SCID mouse model of RA. We also analyzed the expression of neuropilins in synovial tissue and SF, as they may interact with vascular endothelial growth factor signaling and exRNA. The data support the concepts that the exRNA/RNase1 system participates in RA pathophysiology and that RASF are influenced by exRNA in a prodestructive manner.

As an endogenous alarm signal, eukaryotic extracellular RNA (exRNA) has been detected in the extracellular space of tissues, in plasma, cerebrospinal fluid, saliva, urine, and seminal fluid (15), and promotes immunoregulatory, procoagulatory, and vessel permeability–regulating functions (69). Moreover, following cell activation, exRNA was found enriched in supernatants of human leukocytes, epithelial cells, cardiomyocytes, and various tumor cells (1013).

exRNA consists predominantly of rRNA, together with mRNAs, microRNAs, small modified RNAs, and oligonucleotide RNAs (2, 3, 5, 12, 14, 15). exRNA, in association with microparticles or exosomes, can be released by cell death or actively by vital cells, especially under hypoxia (1, 10, 12, 13, 16, 17), in which case the underlying release mechanisms are unclear. The released exRNA is stabilized by binding to proteins or to cell surface molecules (9, 17) and may thus be resistant against degradation by RNases (13). Additionally, Argonaute proteins can protect not only cytosolic miRNA from RNase digestion (18) but may stabilize exRNA as well.

Certain cancer types are accompanied by systemic increase of specific exRNA, and it was proposed as a tumor marker in cancer diagnosis (11, 13). exRNA mediates tumor progression and atherogenesis by inducing TNF-α release and promoting macrophage polarization (19, 20). A role of exRNA in coagulation, vascular permeability, edema, and thrombus formation was also described (9, 21).

Based on the cell-damaging, proinflammatory functions of exRNA in vascular diseases, it appears plausible that exRNA may also contribute in an adverse manner in chronic inflammatory rheumatoid arthritis (RA). Chronic inflammation, synovial hyperplasia, and local hypoxia lead to synovial cell activation, causing severe joint damage. A central cell of joint destruction is the RA synovial fibroblast (RASF) (22). RASF are activated cells with increased production of proinflammatory factors, matrix-degrading enzymes, enhanced matrix adhesion, and cell migration (23). RASF are able to leave their local environment and migrate to distant cartilage (24, 25). The proinflammatory role of extracellular nucleic acids, specifically extracellular mitochondrial DNA and oxidized DNA, has been previously demonstrated in an arthritis model (26). Based on the proinflammatory activity of extracellular DNA (exDNA) due to tissue damage in RA, we investigated the role of the exRNA/RNase system in RA. Of note, the term exRNA is often used for actively released exRNA together with RNA from microparticles or exosomes. In the present study, exRNA denotes actively released exRNA, disregarding the effects of microparticle RNA release.

RA, osteoarthritis (OA), and psoriatic arthritis (PsA) synovium from joint replacement surgeries was collected (2730). Synovium was snap frozen or used for SF isolation. SF culture was performed as described previously (24, 25, 31) in DMEM containing 100 U/ml penicillin/streptomycin, 12.5 mM HEPES (all PAA Laboratories, Cölbe, Germany), and 10% FCS (Sigma-Aldrich, Taufkirchen, Germany) with 10% CO2 at 37°C. Dermal fibroblasts from RA patients (RADF) or healthy volunteers (normal dermal fibroblasts [NDF]) were isolated and cultured according to RASF; normal synovial fibroblasts (NSF) were from Cell Applications (San Diego, CA). Synovial fluid was aspirated from patients with knee joint effusions and stored at −80°C. Human varices were obtained from patients without rheumatic diseases who underwent vein stripping (cardiac and vascular surgery, Bad Nauheim, Germany). Endothelial cells (EC) were isolated according to a modified protocol (32). After washing, vessels were filled with collagenase H (1 mg/ml; Roche, Mannheim, Germany) for 1 h at 37°C. Detached cells were seeded onto rat tail collagen-coated plates. Purity of EC was confirmed by immunocytochemistry against CD31. EC were cultured to maximal passage 3 in DMEM, 20% FCS, 100 U/ml penicillin/streptomycin, 12.5 mM HEPES, and 100 μg/ml EC growth supplement (BD Biosciences, Heidelberg, Germany), 10% CO2 at 37°C.

Tissue collections were approved by local ethics committees. All patients gave written informed consent.

Five-micrometer cryosections from RA (n = 10), OA (n = 16), and PsA (n = 3) synovium and cultured RASF were fixed. Sections were stained using the SYTO RNASelect green fluorescent stain kit (Invitrogen, Karlsruhe, Germany), DAPI (Dako, Carpinteria, CA), and mounted with fluorescent mounting medium (Dako). Serial sections were stained with H&E (Sigma-Aldrich). SF or cryosections were incubated with goat anti-human RNase 1 (Polysciences, Hirschberg, Germany) for 2 h at room temperature, mouse anti-human neuropilin (NRP)-2 (R&D Systems, Wiesbaden, Germany) or mouse anti-human vimentin (Dako) Abs for 1 h at room temperature, or isotype controls. Histofine anti-mouse or anti-goat for a human tissue kit (BD Biosciences) was used for 30 min following incubation with a peroxidase substrate kit (3-amino-9-ethylcarbazole; Vector Laboratories, Grünberg, Germany) and hematoxylin.

Following staining, distribution of exRNA in the synovial tissue was quantified using the University of Texas Health Science Center at San Antonio ImageTool software, version 3.0. For each synovial tissue, merged RNA/DNA images and the DNA images, respectively, were used and transformed to grayscale images. DNA signals were subtracted from merged RNA/DNA signals to exclude cytoplasmic RNA signals. The total area of exRNA staining was calculated and values were used for further statistical evaluation.

RNase activity was determined according to an established protocol (33) in supernatants (NSF, n = 6; RASF, n = 10; OASF, n = 6; PsA-SF, n = 4; NDF, n = 4; RADF, n = 3) and synovial fluids (RA, n = 14; OA, n = 5; PsA, n = 10). One hundred microliters of synovial fluid/supernatant was mixed with 50 μl of RNase buffer (50 mM Tris-HCl, 130 mM NaCl, 2 mM EDTA, 0.1 mg/ml acetylated BSA [pH 8]) containing 100 μl of polycytidylic acid solution (400 μg/ml) (all from Sigma-Aldrich) and incubated 5 min at 37°C. One hundred microliters of this solution was mixed with 250 μl of a solution containing 6% ice-cold perchloric acid, 20 mM lanthanum chloride, and 100 μl of fatty acid–free BSA (10 mg/ml) (all from Sigma-Aldrich), incubated 15 min on ice, centrifuged at 4°C for 15 min at 16,000 × g, and absorbance of supernatants was measured at 280 nm to determine the polycytidylic acid degradation. Samples were normalized to protein concentration measured using the BCA kit (Thermo Fisher Scientific, Bonn, Germany).

The DNA was isolated following the manufacturer’s instructions using the QIAamp DNA blood midi kit (Qiagen, Hilden, Germany) including a RNA digestion step with RNase (Qiagen). Total cellular RNA was isolated using the RNeasy midi kit (Qiagen) including DNA digestion according to the manufacturer’s instructions; nucleic acid concentration was evaluated by NanoDrop measurement (Peqlab Biotechnologie, Erlangen, Germany). For exRNA isolation, RASF were cultured in DMEM containing 100 U/ml penicillin/streptomycin, 12.5 mM HEPES, 10% Panexin NTA (PAN Biotech, Aidenbach, Germany) and 2 μl/ml RNase inhibitor (RiboGuard; Biozym Scientific, Hessisch-Oldendorf, Germany) under normoxia (21% oxygen) in the absence or presence of 15 ng/ml TNF-α or under hypoxia (2% oxygen) for 5 d. Every day, the supernatant was collected and replaced by fresh medium. Supernatants were centrifuged for 10 min at 1500 × g to remove detached cells or cellular debris. exDNA as well as exosomes and microparticles were excluded by passing the supernatants through DNA removing columns (total RNA kit; Peqlab Biotechnologie) using a centrifugation step for 1 min at 12,000 × g without former application of lysis buffer. This removes DNA by selective binding of DNA to the column silica matrix under specific buffer conditions and removes exosomes and microparticles through size exclusion. The resulting eluate was mixed with the same volume of cold ethanol (70%), and exRNA was isolated according to the manufacturer’s instructions using PerfectBind RNA columns (total RNA kit; Peqlab Biotechnologie), which selectively and reversibly bind RNA to the column silica matrix. The exRNA was eluted with sterile RNase-free deionized water. exRNA was characterized by a Agilent Bioanalyzer using the RNA 6000 Pico kit (Agilent Technologies, Waldbronn, Germany).

Jurkat cells were cultured in RPMI 1640 (PAA Laboratories), 100 U/ml penicillin/streptomycin, 12.5 mM HEPES, and 10% heat-inactivated FCS with 5% CO2 at 37°C. Total RNA was isolated using the RNeasy Midi kit including DNA digestion. The resulting RNA was transferred onto an RNase-free low melting agarose gel (Carl Roth, Karlsruhe, Germany), and electrophoresis was run at 110 V for 50 min. The gel was stained with ethidium bromide, and visible human rRNA bands were dissected and isolated using a QIAquick gel extraction kit (Qiagen) according to the manufacturer. Quantification was performed by NanoDrop measurement (Peqlab Biotechnologie).

RASF were stimulated with 10 μg of total RNA or DNA (n = 3) isolated from the respective patient or with 25 μg of rRNA (n = 2) from Jurkat cells compared with untreated RASF in DMEM containing 2% FCS for 24 h. To remove exRNA, RASF (n = 7) were similarly incubated. Afterwards, DNase1 (5 U/ml; Fermentas, St. Leon-Rot, Germany), RNase1 (5 U/ml; Fermentas), or saline solution (control) was added for 15 h. IL-6 was measured in supernatants by ELISA (R&D Systems).

RASF (n = 4) were stained with 2 μg/ml calcein-AM (Invitrogen) and allowed to adhere to unaffected regions of OA cartilage with/without RNase1 or DNase1 (5 U/ml). After 5 min, the cartilage was washed and adherent RASF (three representative areas) quantified.

RASF (n = 3 experimental replicates) were stained with calcein-AM, transferred onto EC monolayers, and incubated with/without DNase1 or RNase1 (1 μg/ml) for 1 h. After washing, adherent cells (four representative areas) were quantified.

RASF (n = 3 experimental replicates) were incubated with DMEM containing 2% FCS with/without 1 ng/ml fibroblast growth factor-2 (Sigma-Aldrich), RNase1 or DNase1 (1 μg/ml), or heparin (10 μg/ml) for 24 h. Proliferation was quantified (BrdU assay kit; Merck, Darmstadt, Germany).

Two micrograms of total RNA was mixed with 0.8 U/μl RNase inhibitor (Promega, Mannheim, Germany), 400 μM random hexamer primers (Roche), 10.8 mM 2′-deoxynucleoside 5′-triphosphate (Roche), 1.8 U/μl avian myeloblastosis virus reverse transcriptase, and 10 μl buffer (Promega) and included denaturation for 2 min at 70°C, cool down, reverse transcription for 15 min at 25°C, 5 min at 37°C, 60 min at 42°C, 30 min at 55°C, and 5 min at 70°C. Expression of vascular endothelial growth factor (VEGF) receptors (VEGF-R1, VEGF-R2, VEGF-R3) and the coreceptors NRP-1 and NRP-2 was measured by PCR using 3 μl cDNA, 10 μl Taq PCR Master mix (Qiagen), and primers (Supplemental Table I) with 0.25 μM each. The program included 5 min at 95°C and, subsequently, 30 cycles for 30 s at 95°C, then 55°C (VEGF-R1, VEGF-R2, and NRP-2), 53°C (VEGF-R3), or 50°C (NRP-1) for 1 min and 1 min 72°C. PCR product was analyzed by 2% agarose gel electrophoresis. cDNA of exRNA for real-time PCR was synthesized as described above. Two microliters of cDNA was mixed with 10 μl of ABsolute QPCR SYBR Green capillary mix (Thermo Fisher Scientific), 0.4 μl of 25 mM MgCl2 (Roche), and 18S primers (final pair concentration, 0.5 μM). The program included 15 min at 95°C, 40 cycles of 15 s at 95°C, 25 s at 60°C, and 30 s at 72°C. PCR was finished by melting curve analysis.

Experiments were approved by the local Animal Care and Use Committee (Germany). According to established protocols (24), Crl-scidBR mice (Charles River Laboratories) underwent implantation surgery. In brief, ipsilaterally, a gelatin matrix containing human cartilage with 1.5 × 105 RASF (34) and, contralaterally, cartilage in a gelatin matrix without RASF were s.c. implanted (24, 25). Animals received i.v. injections before implantation and three times weekly: group 1 (n = 10), 42 μg of RNase1/kg body weight; group 2 (n = 9), 42 μg of DNase1/kg body weight; group 3 (n = 9), saline solution. No change in weight, food intake, or behavior was observed during the course of the experiment. Following termination of the experiment after 45 d, implants were removed and snap frozen. Cartilage invasion by RASF was scored (five blinded researchers) (24, 25, 34).

Results are shown as mean ± SEM. An unpaired Student t test was performed using Graphpad Prism and Microsoft Excel. Differences were considered statistically significant with a p value <0.05.

exRNA was detectable within the RA synovial lining layer and in the intercellular synovial compartments of the lining layer (Fig. 1). The strongest exRNA signal was observed outside the outermost cells of the whole RA lining layer (Fig. 1), whereas exRNA was only detectable in limited areas of the OA lining layer. In PsA, exRNA signals were always colocalized with DNA staining, whereby exRNA was absent in intercellular areas (Fig. 1). Besides, exDNA could be detected in different areas of the synovial tissues (Fig. 1, Supplemental Fig. 1). Quantification of synovial exRNA signals revealed a significant increase in RA (n = 6, 77.97 ± 26.33 arbitrary units [AU]) versus OA (n = 10, 14.44 ± 5.16 AU, p = 0.009, Fig. 2A). In PsA (n = 3), only very low exRNA signals were detectable (1.82 ± 1.04 AU) (Fig. 2A).

FIGURE 1.

Distribution of exRNA in the synovial tissues from patients with inflammatory joint diseases. (A) Synovial tissue sections from RA (n = 6), OA (n = 10), and PsA (n = 3) patients were stained with H&E to depict the tissue architecture of synovial membranes. Serial sections from the respective RA, OA, and PsA patients were stained to visualize distribution of DNA [blue (B)] and RNA [green (C)]. Merged images are indicated in low (D) and high (E) magnification in the same row; the arrows point to exRNA. Scale bars, 100 μm (A–D), 50 μm (E). LL, lining layer, SL, sublining.

FIGURE 1.

Distribution of exRNA in the synovial tissues from patients with inflammatory joint diseases. (A) Synovial tissue sections from RA (n = 6), OA (n = 10), and PsA (n = 3) patients were stained with H&E to depict the tissue architecture of synovial membranes. Serial sections from the respective RA, OA, and PsA patients were stained to visualize distribution of DNA [blue (B)] and RNA [green (C)]. Merged images are indicated in low (D) and high (E) magnification in the same row; the arrows point to exRNA. Scale bars, 100 μm (A–D), 50 μm (E). LL, lining layer, SL, sublining.

Close modal
FIGURE 2.

Analysis of exRNA and RNase in joints from patients with different arthritic diseases. (A) The intensity of exRNA staining in the synovium specimen from patients with RA (n = 6), OA (n = 10), and PsA (n = 3) was carried out by using the University of Texas Health Science Center at San Antonio ImageTool software. Fluorescent images were transformed to grayscale images and DNA images were subtracted from DNA/RNA merged images. The area of the resulting signal was quantified. (B) RNase activity of synovial fluid from patients with RA (n = 14), OA (n = 5), or PsA (n = 10) was quantified (mean ± SEM). Total protein content was determined and samples were normalized to the protein concentration. (C) Total protein concentration in the respective synovial fluids from patients with RA, OA, or PsA was determined (mean ± SEM). (D) Synovial tissue sections from patients with RA (n = 18), OA (n = 7), or PsA (n = 2) were stained for RNase1 (red staining indicated by arrows) and nuclei were stained blue with hematoxylin. Scale bars, 50 μm. *p < 0.05. LL, lining layer; ns, not significant; SL, sublining.

FIGURE 2.

Analysis of exRNA and RNase in joints from patients with different arthritic diseases. (A) The intensity of exRNA staining in the synovium specimen from patients with RA (n = 6), OA (n = 10), and PsA (n = 3) was carried out by using the University of Texas Health Science Center at San Antonio ImageTool software. Fluorescent images were transformed to grayscale images and DNA images were subtracted from DNA/RNA merged images. The area of the resulting signal was quantified. (B) RNase activity of synovial fluid from patients with RA (n = 14), OA (n = 5), or PsA (n = 10) was quantified (mean ± SEM). Total protein content was determined and samples were normalized to the protein concentration. (C) Total protein concentration in the respective synovial fluids from patients with RA, OA, or PsA was determined (mean ± SEM). (D) Synovial tissue sections from patients with RA (n = 18), OA (n = 7), or PsA (n = 2) were stained for RNase1 (red staining indicated by arrows) and nuclei were stained blue with hematoxylin. Scale bars, 50 μm. *p < 0.05. LL, lining layer; ns, not significant; SL, sublining.

Close modal

RNase activity in synovial fluids was significantly increased in RA (31.6 ± 4.5 U/mg protein, p = 0.023) and OA (27.0 ± 4.5 U/mg protein, p = 0.033) versus PsA (18.2 ± 2.3 U/mg protein). Differences were not significant between OA and RA groups (p = 0.569; Fig. 2B). Total protein concentration in synovial fluids was comparable in all groups (RA, 30.4 ± 2.8 mg/ml; OA, 33.6 ± 2.3 mg/ml; PsA, 38.2 ± 2.9 mg/ml) (Fig. 2C). RNase1 was detectable in the lining layer of all patients (Fig. 2D).

In vitro, the release of exRNA by cultured RASF was characterized (Fig. 3A). In contrast to normoxic conditions (21% oxygen; Fig. 3B) or TNF-α stimulation under normoxia (Fig. 3C), hypoxia (2% oxygen) led to a detectable exRNA release by RASF. Size analysis of exRNA (n = 3) revealed RNA signals between 100 and 5000 nt length and peaks at ∼150 and 4200 nt (Fig. 3D–F). The 18S rRNA was present in the exRNA of RASF analyzed by quantitative PCR (data not shown). It has previously been shown by Ng et al. (35) that apoptosis of RASF was not altered by hypoxia. In contrast to RASF, OASF did not secrete exRNA under any of the conditions (Supplemental Fig. 2).

FIGURE 3.

Analysis of exRNA from RASF and RNase activity in vitro. (A) The standard size distribution of an RNA ladder (Agilent Technologies) is indicated as comparison. Cell culture supernatants from RASF were analyzed for exRNA by capillary electrophoresis for size distribution of RNA species. RASF were cultured under normoxia (B) or under normoxia with 15 ng/ml TNF-α (C) or under hypoxia (2% oxygen) using three RASF populations (DF). (D–F) Left arrow, ~150 nucleotides; right arrow, ~4200 nucleotides. (G) RNase activity in the supernatants of cultured SF from patients with RA (n = 10), OA (n = 6), or PsA (n = 4) was quantified and compared with samples from healthy volunteers (NSF, n = 6), NDF (n = 4), or RADF (n = 3) (mean ± SEM) under normoxia. (H) RNase activity of cultured RASF and OASF was quantified under normoxia and hypoxia.

FIGURE 3.

Analysis of exRNA from RASF and RNase activity in vitro. (A) The standard size distribution of an RNA ladder (Agilent Technologies) is indicated as comparison. Cell culture supernatants from RASF were analyzed for exRNA by capillary electrophoresis for size distribution of RNA species. RASF were cultured under normoxia (B) or under normoxia with 15 ng/ml TNF-α (C) or under hypoxia (2% oxygen) using three RASF populations (DF). (D–F) Left arrow, ~150 nucleotides; right arrow, ~4200 nucleotides. (G) RNase activity in the supernatants of cultured SF from patients with RA (n = 10), OA (n = 6), or PsA (n = 4) was quantified and compared with samples from healthy volunteers (NSF, n = 6), NDF (n = 4), or RADF (n = 3) (mean ± SEM) under normoxia. (H) RNase activity of cultured RASF and OASF was quantified under normoxia and hypoxia.

Close modal

In vitro, different fibroblast types under normoxia released RNases into the culture medium as shown by varying RNase activities in the supernatants: NSF, 1.70 ± 0.31 (n = 6); RASF, 1.47 ± 0.19 (n = 10); OASF, 2.12 ± 0.15 (n = 6); PsA-SF, 1.73 ± 0.38 (n = 4); NDF, 1.20 ± 0.3 (n = 4); RADF, 0.67 ± 0.09 (n = 3). RNase activity was significantly reduced in RASF as compared with OASF (p = 0.037) (Fig. 3G). However, hypoxia did not change the release of RNase1 by RASF or OASF in comparison with normoxic conditions (Fig. 3H).

NRP-1 and NRP-2 mRNAs were strongly expressed in all RASF (n = 6) and OASF (n = 3) as well as in human venous EC (HVE) (Fig. 4A), whereas VEGFR-1/-2/-3 were only detectable in HVE but not in SF (Fig. 4A). NRP-2 was present on cells within the synovial RA and OA lining layer, colocalizing with vimentin in the lining layer but not the sublining (Fig. 4B). NRP-2 was also detectable in cultured RASF (Fig. 4B).

FIGURE 4.

Expression of potential exRNA binding partners on synovial fibroblasts and in the synovial tissue. (A) Cytoplasmatic RNA from RASF (n = 6), OASF (n = 6), and primary HVE as positive control (n = 1) was evaluated. cDNA generation with primer pairs for VEGF-1, VEGF-2, VEGF-3, NRP-1, and NRP-2 was determined using RT-PCR. (B) RA (n = 6) and OA (n = 2) synovial tissue as well as cultured RASF were stained immunohistochemically for NRP-2 (red) and vimentin (brown). Nuclei were stained with hematoxylin to define the tissue structure. Arrows indicate NRP-2 signals in the lining layer. Scale bars, 50 μm. LL, lining layer; SL, sublining.

FIGURE 4.

Expression of potential exRNA binding partners on synovial fibroblasts and in the synovial tissue. (A) Cytoplasmatic RNA from RASF (n = 6), OASF (n = 6), and primary HVE as positive control (n = 1) was evaluated. cDNA generation with primer pairs for VEGF-1, VEGF-2, VEGF-3, NRP-1, and NRP-2 was determined using RT-PCR. (B) RA (n = 6) and OA (n = 2) synovial tissue as well as cultured RASF were stained immunohistochemically for NRP-2 (red) and vimentin (brown). Nuclei were stained with hematoxylin to define the tissue structure. Arrows indicate NRP-2 signals in the lining layer. Scale bars, 50 μm. LL, lining layer; SL, sublining.

Close modal

Stimulation of RASF with isolated total RNA of the same patient or total DNA, respectively, slightly increased IL-6 secretion (RNA, 135 ± 40%; DNA, 156 ± 35%) (data not shown). However, this effect was not statistically significant. Stimulation with human rRNA, the most abundant intracellular RNA/exRNA with tumor-reducing properties in mice (36), did not change IL-6 release compared with control (rRNA, 104 ± 16%) (data not shown). By addition of nucleases to remove endogenous extracellular nucleic acids, IL-6 secretion was slightly reduced when adding RNase (77 ± 14% of control) or DNase (87 ± 13% of control), but the changes were statistically not significant (data not shown).

RASF adhesion to extracellular matrix or cells is essential for RASF migration and cartilage invasion (24, 25). RASF adhesion to cartilage was significantly reduced by RNase (cells/visual field, 11.9 ± 2.9 versus 34.5 ± 4.2 for control, p = 0.001) (Fig. 5A), whereas DNase had no significant influence (30.3 ± 1.2 versus 34.5 ± 4.2 for control) (Fig. 5A). Adhesion of RASF to EC was neither affected by RNase or DNase (Fig. 5B) nor was cell proliferation of cultured RASF (passage 4 or 10), whereas heparin reduced cell proliferation (37) independently of the duration of RASF cultivation (Fig. 5C).

FIGURE 5.

(A) RNase reduces significantly the adhesion of RASF (n = 3) to cartilage matrix in contrast to DNase. (B) RNase and DNase treatment had no effect on RASF adhesion to EC (n = 3). (C) RASF proliferation of an early (4) and late passage (10) was not affected by the presence of RNase1 and DNase1. FGF as proliferation inducer and heparin as proliferation inhibitor were used as controls. *p < 0.05. ns, not significant.

FIGURE 5.

(A) RNase reduces significantly the adhesion of RASF (n = 3) to cartilage matrix in contrast to DNase. (B) RNase and DNase treatment had no effect on RASF adhesion to EC (n = 3). (C) RASF proliferation of an early (4) and late passage (10) was not affected by the presence of RNase1 and DNase1. FGF as proliferation inducer and heparin as proliferation inhibitor were used as controls. *p < 0.05. ns, not significant.

Close modal

In the SCID mouse model (Fig. 6A), RASF cartilage invasion on the ipsilateral site was significantly reduced after RNase and DNase treatment (control, 1.92 ± 0.24; RNase1, 1.22 ± 0.21, p = 0.037; DNase, 0.80 ± 0.19, p = 0.002) (Fig. 6B, 6C). In contrast, cartilage invasion on the contralateral site was significantly reduced by DNase but not by RNase1 (control, 1.51 ± 0.28; RNase1, 1.28 ± 0.39, p = 0.625; DNase, 0.84 ± 0.18, p = 0.042) (Fig. 6B, 6D).

FIGURE 6.

Influence of nuclease treatment on cartilage invasion of RASF in the SCID mouse model of RA. (A) The experimental procedures of the SCID mouse model of RA with nuclease application in comparison with physiological saline solution (NaCl) as a control are shown. (B) Representative H&E-stained implant sections from tissues of differently treated mice as indicated are shown; arrows mark RASF invasion zones. (C) RASF invasion was significantly reduced at the ipsilateral implantation site (cartilage coimplanted with RASF) under i.v. application of RNase but not at the contralateral implantation site. (D) Intravenous application of DNase results in a significant reduction of RASF invasion at the ipsilateral implantation site (cartilage coimplanted with RASF) as well as at the contralateral site (implantation of cartilage without RASF). Control, n = 9 animals; RNase, n = 10 animals; DNase, n = 9 animals. Scale bars, 100 μm. NaCl, saline solution. *p < 0.05. ns, not significant.

FIGURE 6.

Influence of nuclease treatment on cartilage invasion of RASF in the SCID mouse model of RA. (A) The experimental procedures of the SCID mouse model of RA with nuclease application in comparison with physiological saline solution (NaCl) as a control are shown. (B) Representative H&E-stained implant sections from tissues of differently treated mice as indicated are shown; arrows mark RASF invasion zones. (C) RASF invasion was significantly reduced at the ipsilateral implantation site (cartilage coimplanted with RASF) under i.v. application of RNase but not at the contralateral implantation site. (D) Intravenous application of DNase results in a significant reduction of RASF invasion at the ipsilateral implantation site (cartilage coimplanted with RASF) as well as at the contralateral site (implantation of cartilage without RASF). Control, n = 9 animals; RNase, n = 10 animals; DNase, n = 9 animals. Scale bars, 100 μm. NaCl, saline solution. *p < 0.05. ns, not significant.

Close modal

exRNA is present in different body fluids and tissues, has proinflammatory function, and is proposed to be associated with cytokine release and cellular activation (6). A role of exRNA as an immunomodulatory factor in the pathophysiology of RA has not yet been considered. However, increased levels of exDNA due to increased tissue destruction were demonstrated in RA synovial fluids previously (38), and injection of mitochondrial and oxidized DNA into murine knees induced RA-like synovitis (26).

In our experiments, the presence of exRNA and RNase within RA joints could be demonstrated. exRNA in RA synovium was detectable, especially in the lining layer, whereas exDNA, found to be present in RA synovial fluid (38), was not detectable in the lining layer but was in other areas of the synovial tissue. We therefore propose that exDNA does not accumulate in the lining layer. It was described that DNA-containing neutrophil extracellular traps, derived from dying neutrophils, are increased in the synovium (39) and blood of RA patients (40). In this study, we confirmed the presence of exDNA in RA patients.

In OA synovium, the exRNA was limited to small areas of the lining layer in contrast to RA. Hypoxia-inducible factor 1α as low oxygen signal was previously described to be increased in the RA and OA lining layer (41, 42). exRNA was higher in RA than OA synovium, which could be due to a reduced partial oxygen pressure in the synovial fluid of RA versus OA patients (43). Hypoxia in the lining layer could be required to induce exRNA release. In PsA, RNA signals were not detectable outside of nuclei. However, hypoxia represented by hypoxia-inducible factor 1α expression was described to be present in the PsA lining layer (44). The difference for RA may be due to the lower thickness of the lining layer in PsA compared with RA or that exRNA release may additionally be induced by other factors. Furthermore, inflammation in PsA is located in synovium but also at sites of enthesis (30), a compartment not evaluated in this study owing to the unavailability of enthesis-containing tissue.

Reasons for digestion-resistant exRNA are stabilization through binding to protein/lipoprotein complexes or sequestration of exRNA within lipid microvesicles (17, 18, 45, 46). Cytoplasmatic RNAs were not visible after tissue staining due to quick mRNA transport to ribosomes followed by cytoplasmatic RNA degradation. Therefore, in this approach staining of intracellular RNAs was solely discernable in colocalization with nuclear DNA. The absence of exRNA staining in some tissue regions could be due to an increased level of extracellular RNases, particularly RNase1.

The observed increase of RNase1 in RA synovium and increased RNase activity in RA synovial fluids is in line with a previous study identifying increased RNase5 levels in RA synovial fluids (47). However, the existence of exRNAs as proinflammatory factors was not uncovered yet, such that mechanistic relationships remained elusive at that time. Increased synovial RNase activity might be the consequence of increased synovial exRNA concentrations, reflecting an anti-inflammatory attempt to reduce proinflammatory effects of exRNA. Similar compensatory mechanisms were observed in RA, for example regarding the secretion of anti-inflammatory factors (e.g., IL-10) (48, 49). However, their levels are not sufficient to confine a proinflammatory milieu in RA. This may also apply to the exRNA/RNase balance in RA.

Therefore, we analyzed whether cultured RASF secrete exRNA and RNases, and besides baselines production, exRNA release was particularly elevated under hypoxia. The size of exRNA ranged from 150 to 5000 nt and included 18S rRNA. RASF cultures under normoxia (with/without TNF-α) reduced exRNA secretion to nondetectable levels. Secretion of exRNA by RASF under hypoxia is in line with observations that hypoxia is a hallmark of inflamed RA joints (50). Significantly increased RNase levels in RA versus PsA synovial fluids were not detectable in cultured cells. The increased RNase release may result from high numbers of fibroblasts or other cells in the hyperplastic RA synovial lining layer. Of note, all synovial fluid samples were collected during joint aspiration from actively inflamed joints (including OA); however, protein concentrations in synovial fluids do not necessarily reflect the situation within the living tissue. Therefore, under pathophysiological conditions in RA not only the individual levels of exRNA and RNase have to be considered, but the overall exRNA/RNase balance as well.

Because of limited amounts of authentic exRNA from body fluids, isolated cellular RNA that has been documented to be equivalent in functional activities was used for in vitro experiments. The concentrations of extracellular nucleic acids in cellular experiments were selected according to another study (26), which documented the proinflammatory response of exDNA in an arthritis animal model. However, in the present study using RASF, we could not confirm this proinflammatory effect of exDNA in vitro. We looked at IL-6 as a central proinflammatory factor in RA together with TNF-α. Only small, nonsignificant changes of IL-6 secretion after treatment of RASF with cellular RNA and DNA were observed, possibly due to a reduced stability of the prepared RNA in comparison with actively released exRNA. Therefore, we reduced basal exRNA levels via RNase treatment to examine whether this would reduce basal IL-6 secretion. However, this did not significantly change IL-6 secretion of RASF either.

As an interesting function of the exRNA/RNase system, RNase exposure significantly reduced the adhesion of RASF to cartilage but not to EC, indicating that cell–matrix but not cell–cell interactions were altered by exRNA. These aspects are important for RASF migration and attachment, especially with regard to potential modulation of pathologic RASF long-distance migration through the vasculature (24, 25). As cartilage adhesion and IL-6 release by RASF was altered by exRNA in vitro, the functional effects were confirmed in the SCID mouse model of RA.

Soucek et al. (51) demonstrated in animal models that administered RNase was detectable after 1 d of i.v. application in the blood and in organs. In our experiments, DNase administration reduced invasiveness of migratory and nonmigratory RASF. exDNA is most likely present at both implantation sites due to cartilage destruction and local cell death similar to results previously shown for extracellular mitochondrial DNA and oxidized DNA (26). The DNase administered i.v. into SCID mice might have an effect on local cells, which interact with RASF, thus affecting their invasion. Additionally, RASF actively invade and degrade cartilage matrix, thereby inducing the damage of chondrocytes, which in turn may release DNA due to apoptosis, which can be observed in implants as well as in human tissues (data not shown). Therefore, the disturbance or modulation of cell–cell and cell–matrix interactions may be the reason for the dramatic effect of DNase on RASF invasion in vivo, whereas there was no effect of DNase in vitro.

RNase treatment in mice revealed a cartilage-protective effect at the ipsilateral site in the arthritis model. However, RASF still migrated and invaded the contralateral cartilage, indicating that long-distance migration of RASF was not affected by exRNA, consistent with our findings on RASF-to-EC adhesion in vitro. Therefore, migrating and nonmigrating RASF probably represent distinct subpopulations that respond differently to variations in the exRNA/RNase system.

Unlike exDNA, which was detectable particularly at the cartilage border of the explants (Supplemental Fig. 3), RNA signals could only be detected intracellularly or at most at a very low level around blood vessels (Supplemental Fig. 3). Therefore, RNase treatment is most likely less effective in SCID mice than is DNase treatment. This difference to the stained human tissue (Fig. 1), where both exDNA and exRNA could be detected, may be due to the use of isolated synovial fibroblasts and the relatively short incubation period of 45 d. Hence, the isolated synovial fibroblasts are not capable of producing detectable amounts exRNA in the explants within that time frame.

We confirmed the presence of NRPs as coreceptors for VEGF receptors on RASF. It was previously shown that NRP-1 is highly expressed on RASF, and antiapoptotic signals of VEGF-A165 were transduced by NRP-1 in RASF (52). Owing to binding of exRNA to VEGF, the interaction of RNA/VEGF complexes with VEGF-R2 and NRP-1 is increased, resulting in elevated vascular permeability (8). In RA and OA synovium, NRP-2, which is known to interact with VEGF and thus affects VEGF-induced signaling, was expressed in the lining layer. Whether this interaction also involves exRNA, as in the case of NRP-1, remains to be shown.

Another interaction partner of exRNA could be TNF-α–converting enzyme (TACE), which is expressed by synovial cells (53). It was recently demonstrated by our group that exRNA promotes TNF-α release via TACE activation on macrophages, resulting in a global proinflammatory burst, because TNF-α subsequently provokes the release of exRNA (19). In the present study, exRNA released by RASF increased TACE activity on synovial macrophages, and the enhanced level of released TNF-α may in turn affect synovial cells. The nature of the cellular interaction of exRNA that could involve TLR to some extent remains to be elucidated in the context of RA.

Taken together, the presence and function of exRNA at vulnerable sites is an unexpected aspect in the pathogenesis of RA. exRNA is released under hypoxia by RASF and accumulates in the lining layer, where it acts as a direct stimulus on RASF, particularly in the active zone of rheumatoid joint destruction. Indirectly, exRNA induces RASF to promote TNF-α release from macrophages. The different responses of RASF to the exRNA/RNase system reveal new insights into the cellular pathophysiology that may lead to novel therapeutic approaches.

We acknowledge Sabrina Brückmann, Simone Benninghoff, Bärbel Fühler, Juri Schklarenko, Sakine Simsekyilmaz, and Dimitry Grün for their excellent technical assistance. We thank Dr. Grit Krumbholz for her support during rRNA isolation.

This work was supported in part by research grants from the German Ministry of Education and Research as part of the project Immunological Memory in Health and Disease, German Research Foundation Grants MU 1383/14-1 and NE 1174/6-1, the local State Offensive for Development of Scientific and Economic Excellence program Medical RNomics (to K.T.P. and S.F.), and by the Excellence Cluster Cardio-Pulmonary System. H.A.C.-F. is funded by a Startup Grant of the Excellence Cluster Cardio-Pulmonary System from the German Research Foundation (Bonn, Germany) and the Peter und Traudl Engelhorn-Stiftung (Weilheim, Germany). Part of the work by H.A.C.-F. and K.T.P. was supported by the Russian Government program for competitive growth of Kazan Federal University, Kazan (Russian Federation).

The online version of this article contains supplemental material.

Abbreviations used in this article:

     
  • AU

    arbitrary unit

  •  
  • EC

    endothelial cell

  •  
  • exDNA

    extracellular DNA

  •  
  • exRNA

    extracellular RNA

  •  
  • HVE

    human venous endothelial cell

  •  
  • NDF

    normal dermal fibroblast

  •  
  • NRP

    neuropilin

  •  
  • NSF

    normal synovial fibroblast

  •  
  • OA

    osteoarthritis

  •  
  • PsA

    psoriatic arthritis

  •  
  • RA

    rheumatoid arthritis

  •  
  • RADF

    dermal fibroblasts from RA patient

  •  
  • RASF

    rheumatoid arthritis synovial fibroblast

  •  
  • SF

    synovial fibroblast

  •  
  • TACE

    TNF-α–converting enzyme

  •  
  • VEGF

    vascular endothelial growth factor.

1
Dinger
M. E.
,
Mercer
T. R.
,
Mattick
J. S.
.
2008
.
RNAs as extracellular signaling molecules.
J. Mol. Endocrinol.
40
:
151
159
.
2
Formosa
A.
,
Piro
M. C.
,
Docimo
R.
,
Maturo
P.
,
Sollecito
D. R.
,
Kalimutho
M.
,
Sancesario
G.
,
Barlattani
A.
,
Melino
G.
,
Candi
E.
,
Bernardini
S.
.
2011
.
Salivary miRNAome profiling uncovers epithelial and proliferative miRNAs with differential expression across dentition stages.
Cell Cycle
10
:
3359
3368
.
3
Hanke
M.
,
Hoefig
K.
,
Merz
H.
,
Feller
A. C.
,
Kausch
I.
,
Jocham
D.
,
Warnecke
J. M.
,
Sczakiel
G.
.
2010
.
A robust methodology to study urine microRNA as tumor marker: microRNA-126 and microRNA-182 are related to urinary bladder cancer.
Urol. Oncol.
28
:
655
661
.
4
Pogue
A. I.
,
Hill
J. M.
,
Lukiw
W. J.
.
2014
.
MicroRNA (miRNA): sequence and stability, viroid-like properties, and disease association in the CNS.
Brain Res.
1584
:
73
79
.
5
Zubakov
D.
,
Boersma
A. W.
,
Choi
Y.
,
van Kuijk
P. F.
,
Wiemer
E. A.
,
Kayser
M.
.
2010
.
MicroRNA markers for forensic body fluid identification obtained from microarray screening and quantitative RT-PCR confirmation.
Int. J. Legal Med.
124
:
217
226
.
6
Deindl
E.
,
Fischer
S.
,
Preissner
K. T.
.
2009
.
New directions in inflammation and immunity: the multi-functional role of the extracellular RNA/RNase system.
Indian J. Biochem. Biophys.
46
:
461
466
.
7
Fischer
S.
,
Gerriets
T.
,
Wessels
C.
,
Walberer
M.
,
Kostin
S.
,
Stolz
E.
,
Zheleva
K.
,
Hocke
A.
,
Hippenstiel
S.
,
Preissner
K. T.
.
2007
.
Extracellular RNA mediates endothelial-cell permeability via vascular endothelial growth factor.
Blood
110
:
2457
2465
.
8
Fischer
S.
,
Nishio
M.
,
Peters
S. C.
,
Tschernatsch
M.
,
Walberer
M.
,
Weidemann
S.
,
Heidenreich
R.
,
Couraud
P. O.
,
Weksler
B. B.
,
Romero
I. A.
, et al
.
2009
.
Signaling mechanism of extracellular RNA in endothelial cells.
FASEB J.
23
:
2100
2109
.
9
Nakazawa
F.
,
Kannemeier
C.
,
Shibamiya
A.
,
Song
Y.
,
Tzima
E.
,
Schubert
U.
,
Koyama
T.
,
Niepmann
M.
,
Trusheim
H.
,
Engelmann
B.
,
Preissner
K. T.
.
2005
.
Extracellular RNA is a natural cofactor for the (auto-)activation of factor VII-activating protease (FSAP).
Biochem. J.
385
:
831
838
.
10
Böttcher
K.
,
Wenzel
A.
,
Warnecke
J. M.
.
2006
.
Investigation of the origin of extracellular RNA in human cell culture.
Ann. N. Y. Acad. Sci.
1075
:
50
56
.
11
Pigati
L.
,
Yaddanapudi
S. C.
,
Iyengar
R.
,
Kim
D. J.
,
Hearn
S. A.
,
Danforth
D.
,
Hastings
M. L.
,
Duelli
D. M.
.
2010
.
Selective release of microRNA species from normal and malignant mammary epithelial cells.
PLoS One
5
:
e13515
.
12
Wissler
J. H.
2004
.
Extracellular and circulating redox- and metalloregulated eRNA and eRNP: copper ion-structured RNA cytokines (angiotropin ribokines) and bioaptamer targets imparting RNA chaperone and novel biofunctions to S100-EF-hand and disease-associated proteins.
Ann. N. Y. Acad. Sci.
1022
:
163
184
.
13
Zhou
H.
,
Xu
W.
,
Qian
H.
,
Yin
Q.
,
Zhu
W.
,
Yan
Y.
.
2008
.
Circulating RNA as a novel tumor marker: an in vitro study of the origins and characteristics of extracellular RNA.
Cancer Lett.
259
:
50
60
.
14
El-Hefnawy
T.
,
Raja
S.
,
Kelly
L.
,
Bigbee
W. L.
,
Kirkwood
J. M.
,
Luketich
J. D.
,
Godfrey
T. E.
.
2004
.
Characterization of amplifiable, circulating RNA in plasma and its potential as a tool for cancer diagnostics.
Clin. Chem.
50
:
564
573
.
15
Bar
M.
,
Wyman
S. K.
,
Fritz
B. R.
,
Qi
J.
,
Garg
K. S.
,
Parkin
R. K.
,
Kroh
E. M.
,
Bendoraite
A.
,
Mitchell
P. S.
,
Nelson
A. M.
, et al
.
2008
.
MicroRNA discovery and profiling in human embryonic stem cells by deep sequencing of small RNA libraries.
Stem Cells
26
:
2496
2505
.
16
Sloane
J. A.
,
Blitz
D.
,
Margolin
Z.
,
Vartanian
T.
.
2010
.
A clear and present danger: endogenous ligands of Toll-like receptors.
Neuromolecular Med.
12
:
149
163
.
17
Wang
K.
,
Zhang
S.
,
Weber
J.
,
Baxter
D.
,
Galas
D. J.
.
2010
.
Export of microRNAs and microRNA-protective protein by mammalian cells.
Nucleic Acids Res.
38
:
7248
7259
.
18
Winter
J.
,
Diederichs
S.
.
2011
.
Argonaute proteins regulate microRNA stability: increased microRNA abundance by Argonaute proteins is due to microRNA stabilization.
RNA Biol.
8
:
1149
1157
.
19
Fischer
S.
,
Gesierich
S.
,
Griemert
B.
,
Schänzer
A.
,
Acker
T.
,
Augustin
H. G.
,
Olsson
A. K.
,
Preissner
K. T.
.
2013
.
Extracellular RNA liberates tumor necrosis factor-α to promote tumor cell trafficking and progression.
Cancer Res.
73
:
5080
5089
.
20
Simsekyilmaz
S.
,
Cabrera-Fuentes
H. A.
,
Meiler
S.
,
Kostin
S.
,
Baumer
Y.
,
Liehn
E. A.
,
Weber
C.
,
Boisvert
W. A.
,
Preissner
K. T.
,
Zernecke
A.
.
2014
.
Role of extracellular RNA in atherosclerotic plaque formation in mice.
Circulation
129
:
598
606
.
21
Kannemeier
C.
,
Shibamiya
A.
,
Nakazawa
F.
,
Trusheim
H.
,
Ruppert
C.
,
Markart
P.
,
Song
Y.
,
Tzima
E.
,
Kennerknecht
E.
,
Niepmann
M.
, et al
.
2007
.
Extracellular RNA constitutes a natural procoagulant cofactor in blood coagulation.
Proc. Natl. Acad. Sci. USA
104
:
6388
6393
.
22
Bartok
B.
,
Firestein
G. S.
.
2010
.
Fibroblast-like synoviocytes: key effector cells in rheumatoid arthritis.
Immunol. Rev.
233
:
233
255
.
23
Neumann
E.
,
Lefèvre
S.
,
Zimmermann
B.
,
Gay
S.
,
Müller-Ladner
U.
.
2010
.
Rheumatoid arthritis progression mediated by activated synovial fibroblasts.
Trends Mol. Med.
16
:
458
468
.
24
Lefèvre
S.
,
Knedla
A.
,
Tennie
C.
,
Kampmann
A.
,
Wunrau
C.
,
Dinser
R.
,
Korb
A.
,
Schnäker
E. M.
,
Tarner
I. H.
,
Robbins
P. D.
, et al
.
2009
.
Synovial fibroblasts spread rheumatoid arthritis to unaffected joints.
Nat. Med.
15
:
1414
1420
.
25
Neumann
E.
,
Lefèvre
S.
,
Zimmermann
B.
,
Geyer
M.
,
Lehr
A.
,
Umscheid
T.
,
Schönburg
M.
,
Rehart
S.
,
Müller-Ladner
U.
.
2010
.
Migratory potential of rheumatoid arthritis synovial fibroblasts: additional perspectives.
Cell Cycle
9
:
2286
2291
.
26
Collins
L. V.
,
Hajizadeh
S.
,
Holme
E.
,
Jonsson
I. M.
,
Tarkowski
A.
.
2004
.
Endogenously oxidized mitochondrial DNA induces in vivo and in vitro inflammatory responses.
J. Leukoc. Biol.
75
:
995
1000
.
27
Aletaha
D.
,
Neogi
T.
,
Silman
A. J.
,
Funovits
J.
,
Felson
D. T.
,
Bingham
C. O.
 III
,
Birnbaum
N. S.
,
Burmester
G. R.
,
Bykerk
V. P.
,
Cohen
M. D.
, et al
.
2010
.
2010 rheumatoid arthritis classification criteria: an American College of Rheumatology/European League against Rheumatism collaborative initiative.
Ann. Rheum. Dis.
69
:
1580
1588
.
28
Altman
R.
,
Asch
E.
,
Bloch
D.
,
Bole
G.
,
Borenstein
D.
,
Brandt
K.
,
Christy
W.
,
Cooke
T. D.
,
Greenwald
R.
,
Hochberg
M.
, et al
Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association
.
1986
.
Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee.
Arthritis Rheum.
29
:
1039
1049
.
29
Arnett
F. C.
,
Edworthy
S. M.
,
Bloch
D. A.
,
McShane
D. J.
,
Fries
J. F.
,
Cooper
N. S.
,
Healey
L. A.
,
Kaplan
S. R.
,
Liang
M. H.
,
Luthra
H. S.
, et al
.
1988
.
The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis.
Arthritis Rheum.
31
:
315
324
.
30
Taylor
W.
,
Gladman
D.
,
Helliwell
P.
,
Marchesoni
A.
,
Mease
P.
,
Mielants
H.
CASPAR Study Group
.
2006
.
Classification criteria for psoriatic arthritis: development of new criteria from a large international study.
Arthritis Rheum.
54
:
2665
2673
.
31
Müller-Ladner
U.
,
Kriegsmann
J.
,
Franklin
B. N.
,
Matsumoto
S.
,
Geiler
T.
,
Gay
R. E.
,
Gay
S.
.
1996
.
Synovial fibroblasts of patients with rheumatoid arthritis attach to and invade normal human cartilage when engrafted into SCID mice.
Am. J. Pathol.
149
:
1607
1615
.
32
Baudin
B.
,
Bruneel
A.
,
Bosselut
N.
,
Vaubourdolle
M.
.
2007
.
A protocol for isolation and culture of human umbilical vein endothelial cells.
Nat. Protoc.
2
:
481
485
.
33
Zimmermann
S. B.
,
Sandeen
G.
.
1965
.
A sensitive assay for pancreatic ribonuclease.
Anal. Biochem.
10
:
444
449
.
34
Judex
M.
,
Neumann
E.
,
Fleck
M.
,
Pap
T.
,
Mountz
D. J.
,
Gay
R. E.
,
Schölmerich
J.
,
Nishioka
K.
,
Gay
S.
,
Muller-Ladner
U.
.
2001
.
“Inverse wrap”: an improved implantation technique for virus transduced synovial fibroblasts in the SCID-mouse model for rheumatoid arthritis.
Mod. Rheumatol.
11
:
145
150
.
35
Ng
C. T.
,
Biniecka
M.
,
Kennedy
A.
,
McCormick
J.
,
Fitzgerald
O.
,
Bresnihan
B.
,
Buggy
D.
,
Taylor
C. T.
,
O’Sullivan
J.
,
Fearon
U.
,
Veale
D. J.
.
2010
.
Synovial tissue hypoxia and inflammation in vivo.
Ann. Rheum. Dis.
69
:
1389
1395
.
36
Lavelle
S. M.
,
MhicIomhair
M.
.
2003
.
Inhibition of foreign-body sarcoma by mammalian RNA?
Technol. Health Care
11
:
143
147
.
37
Garg
H. G.
,
Mrabat
H.
,
Yu
L.
,
Hales
C. A.
,
Li
B.
,
Moore
C. N.
,
Zhang
F.
,
Linhardt
R. J.
.
2011
.
Anti-proliferative effects of O-acyl-low-molecular-weight heparin derivatives on bovine pulmonary artery smooth muscle cells.
Glycoconj. J.
28
:
419
426
.
38
Hajizadeh
S.
,
DeGroot
J.
,
TeKoppele
J. M.
,
Tarkowski
A.
,
Collins
L. V.
.
2003
.
Extracellular mitochondrial DNA and oxidatively damaged DNA in synovial fluid of patients with rheumatoid arthritis.
Arthritis Res. Ther.
5
:
R234
R240
.
39
Khandpur
R.
,
Carmona-Rivera
C.
,
Vivekanandan-Giri
A.
,
Gizinski
A.
,
Yalavarthi
S.
,
Knight
J. S.
,
Friday
S.
,
Li
S.
,
Patel
R. M.
,
Subramanian
V.
, et al
.
2013
.
NETs are a source of citrullinated autoantigens and stimulate inflammatory responses in rheumatoid arthritis.
Sci. Transl. Med.
5
:
178ra40
.
40
Dwivedi
N.
,
Radic
M.
.
2014
.
Citrullination of autoantigens implicates NETosis in the induction of autoimmunity.
Ann. Rheum. Dis.
73
:
483
491
.
41
Giatromanolaki
A.
,
Sivridis
E.
,
Maltezos
E.
,
Athanassou
N.
,
Papazoglou
D.
,
Gatter
K. C.
,
Harris
A. L.
,
Koukourakis
M. I.
.
2003
.
Upregulated hypoxia inducible factor-1α and -2α pathway in rheumatoid arthritis and osteoarthritis.
Arthritis Res. Ther.
5
:
R193
R201
.
42
Hollander
A. P.
,
Corke
K. P.
,
Freemont
A. J.
,
Lewis
C. E.
.
2001
.
Expression of hypoxia-inducible factor 1α by macrophages in the rheumatoid synovium: implications for targeting of therapeutic genes to the inflamed joint.
Arthritis Rheum.
44
:
1540
1544
.
43
Taylor
P. C.
,
Sivakumar
B.
.
2005
.
Hypoxia and angiogenesis in rheumatoid arthritis.
Curr. Opin. Rheumatol.
17
:
293
298
.
44
Yamamoto
T.
2013
.
Angiogenic and inflammatory properties of psoriatic arthritis.
ISRN Dermatol.
2013
:
630620
.
45
Sisco
K. L.
2001
.
Is RNA in serum bound to nucleoprotein complexes?
Clin. Chem.
47
:
1744
1745
.
46
Tsui
N. B.
,
Ng
E. K.
,
Lo
Y. M.
.
2002
.
Stability of endogenous and added RNA in blood specimens, serum, and plasma.
Clin. Chem.
48
:
1647
1653
.
47
Lioté
F.
,
Champy
R.
,
Moenner
M.
,
Boval-Boizard
B.
,
Badet
J.
.
2003
.
Elevated angiogenin levels in synovial fluid from patients with inflammatory arthritis and secretion of angiogenin by cultured synovial fibroblasts.
Clin. Exp. Immunol.
132
:
163
168
.
48
Deleuran
B. W.
,
Chu
C. Q.
,
Field
M.
,
Brennan
F. M.
,
Katsikis
P.
,
Feldmann
M.
,
Maini
R. N.
.
1992
.
Localization of interleukin-1α, type 1 interleukin-1 receptor and interleukin-1 receptor antagonist in the synovial membrane and cartilage/pannus junction in rheumatoid arthritis.
Br. J. Rheumatol.
31
:
801
809
.
49
Katsikis
P. D.
,
Chu
C. Q.
,
Brennan
F. M.
,
Maini
R. N.
,
Feldmann
M.
.
1994
.
Immunoregulatory role of interleukin 10 in rheumatoid arthritis.
J. Exp. Med.
179
:
1517
1527
.
50
Sivakumar
B.
,
Akhavani
M. A.
,
Winlove
C. P.
,
Taylor
P. C.
,
Paleolog
E. M.
,
Kang
N.
.
2008
.
Synovial hypoxia as a cause of tendon rupture in rheumatoid arthritis.
J. Hand Surg. Am.
33
:
49
58
.
51
Soucek
J.
,
Poucková
P.
,
Strohalm
J.
,
Plocová
D.
,
Hlousková
D.
,
Zadinová
M.
,
Ulbrich
K.
.
2002
.
Poly[N-(2-hydroxypropyl)methacrylamide] conjugates of bovine pancreatic ribonuclease (RNase A) inhibit growth of human melanoma in nude mice.
J. Drug Target.
10
:
175
183
.
52
Kim
W. U.
,
Kang
S. S.
,
Yoo
S. A.
,
Hong
K. H.
,
Bae
D. G.
,
Lee
M. S.
,
Hong
S. W.
,
Chae
C. B.
,
Cho
C. S.
.
2006
.
Interaction of vascular endothelial growth factor 165 with neuropilin-1 protects rheumatoid synoviocytes from apoptotic death by regulating Bcl-2 expression and Bax translocation.
J. Immunol.
177
:
5727
5735
.
53
Charbonneau
M.
,
Harper
K.
,
Grondin
F.
,
Pelmus
M.
,
McDonald
P. P.
,
Dubois
C. M.
.
2007
.
Hypoxia-inducible factor mediates hypoxic and tumor necrosis factor α-induced increases in tumor necrosis factor-α converting enzyme/ADAM17 expression by synovial cells.
J. Biol. Chem.
282
:
33714
33724
.

The authors have no financial conflicts of interest.

Supplementary data