Monosodium urate crystals induce oxidative stress in human synoviocytes

Background Gout is the most common inflammatory arthropathy of metabolic origin and it is characterized by intense inflammation, the underlying mechanisms of which are unknown. The aim of this study was to evaluate the oxidative stress in human fibroblast-like synoviocytes (FLS) exposed to monosodium urate (MSU) crystals, which trigger an inflammatory process. Methods Human FLS isolated from synovial tissue explants were stimulated with MSU crystals (75 μg/mL) for 24 h. Cellular viability was evaluated by crystal violet staining, apoptosis was assessed using Annexin V, and the cellular content of reactive oxygen species (ROS) and nitrogen species (RNS) (O2-, H2O2, NO) was assessed with image-based cytometry and fluorometric methods. In order to determine protein oxidation levels, protein carbonyls were detected through oxyblot analysis, and cell ultrastructural changes were assessed by transmission electron microscopy. Results The viability of FLS exposed to MSU crystals decreased by 30 % (P < 0.05), while apoptosis increased by 42 % (P = 0.01). FLS stimulated with MSU crystals exhibited a 2.1-fold increase in H2O2 content and a 1.5-fold increase in O2- and NO levels. Oxyblots revealed that the spots obtained from FLS protein lysates exposed to MSU crystals exhibited protein carbonyl immunoreactivity, which reflects the presence of oxidatively modified proteins. Concomitantly, MSU crystals triggered the induction of changes in the morphostructure of FLS, such as the thickening and discontinuity of the endoplasmic reticulum, and the formation of vacuoles and misfolded glycoproteins. Conclusions Our results prove that MSU crystals induce the release of ROS and RNS in FLS, subsequently oxidizing proteins and altering the cellular oxidative state of the endoplasmic reticulum, which results in FLS apoptosis. Electronic supplementary material The online version of this article (doi:10.1186/s13075-016-1012-3) contains supplementary material, which is available to authorized users.


Background
Gout is a uric acid (UA) metabolic disorder that promotes the formation and deposition of monosodium urate (MSU) crystals inside joints and periarticular soft tissues as a result of hyperuricemia. It is the most common inflammatory arthropathy in young men, and its prevalence is underestimated due to the long asymptomatic phase of the disease [1]. The global burden of gout is substantial and has increased in many parts of the world over the past 50 years [2]. Gout causes monocytic inflammatory cells to phagocytose MSU crystals. This induces the release of pro-inflammatory cytokines such as IL-8, IL-6, CCL2, interferon (IFN)-γ, and IL-1β by assembling and activating the NOD-like receptor pyrin containing 3 (NLRP3) inflammasome [3,4].
The deleterious effects of urate are primarily attributed to its ability to trigger the formation of reactive oxygen species (ROS) and activate NLRP3. However, these mechanisms have not yet been elucidated [5,6]. The activation of NADPH oxidase, xanthine oxidase, and nitric oxide synthase enzymes generates hydrogen peroxide (H 2 O 2 ), superoxide anion (O 2 -) and nitric oxide (NO), respectively. The interaction of these last two molecules promotes the generation of peroxynitrite (ONOO -), which in turn increases apoptosis, the degradation of connective tissues, and joint damage [7,8]. However, due to the complex interactions that take place within joints among various cell types, including neutrophils, macrophages, mast cells, endothelial cells and synovial fibroblasts, it is possible for synovial fibroblasts to play a role in modulating the inflammatory response to MSU crystals in patients with gout [9,10].
Previous studies have reported that endogenous ROS are overproduced during acute gout attacks, suggesting that oxidative stress (OS) contributes to acute gout attacks and to the painful and inflammatory responses that MSU crystals induce by currently unknown mechanisms [11]. The aim of this study was to evaluate the prooxidizing effect of MSU crystals in an in vitro model of crystal-induced inflammation. We focused on ROS and RNS associated with the generation of OS induced by MSU crystals in human-derived synovial membrane (SM) cells. We found that MSU crystals trigger an oxidative response and oxidize proteins, highlighting a possible mechanism underlying gout pathogenesis. The results shown here help explain how MSU crystals combined with ROS react with proteins of synoviocytes, increasing our understanding of the role of OS in the development of gout.

Methods
This study was approved by the Research Committee of the Instituto Nacional de Rehabilitacion (Ref.02/13) of Mexico and was carried out according to the principles of the Helsinki declaration. Written informed consent was obtained from all patients.

MSU preparation
MSU crystals were synthesized by uric acid (UA) crystallization according to the method described by Denko and Whitehouse [12] and modified by Scanu et al. [10]. MSU crystals were characterized by polarized light microscopy and scanning electron microscopy (SEM) based on the crystallographic characteristics birefringence, size, and morphology [13], and were sterilized at 180°C for 2 h. The absence of microbial contaminants was confirmed by culturing for microorganisms, and the crystals were determined to be bacterial endotoxin-free by Limulus amebocyte cell-lysate assay (Sigma-Aldrich).

Isolation and cell culture of fibro-synoviocytes
A primary culture of synoviocytes was obtained via mechanic-enzymatic breakdown of SM collected from patients with osteoarthritis (OA) (n = 5) during knee joint replacement. Synoviocytes were isolated from tissue explants following digestion with collagenase type IA (1 mg/mL) (Gibco, Life Technologies) for 2 h with mixing at 37°C. Cells were seeded in T25 flasks at a density of 250,000/flask until confluence. The cells were cultured in DMEM-F12 supplemented with 10 % fetal bovine serum and 1 % penicillin-streptomycin (Gibco, Life Technologies), and they were incubated in a controlled CO 2 atmosphere at a regulated temperature. At confluence, cells were harvested (TrypLE Express, Gibco, Life Technologies) and seeded into new flasks that kept synoviocytes from different patients separated.
For the experiments, cells were used at the third or fourth passage.

Phenotyping of fibroblast-like synoviocytes analysis Characterization by qRT-PCR
Upon the third passage, fibroblast-like synoviocytes (FLS) phenotype was determined by assessing the expression of the uridine diphosphate glucose dehydrogenase gene (UGDH), and CD14 gene was used for macrophagelike synoviocytes by qRT-PCR. Total RNA from each patient was extracted by the Trizol method [14]. The qRT-PCR technique was performed by amplifying primers (Additional file 1) in a Rotor-Gene Q thermocycler (Qiagen), according to the commercial kit RT 2 First Strand Kit from Qiagen. The results were normalized to the housekeeping GAPDH gene and relative quantification was performed through REST-09 software (Relative Expression Tool software 2009). After amplification, a melting assay was performed to confirm the specific size of the products of each gene.
Total protein was obtained from the culture of sinoviocytes. Analysis of the protein content was performed by WB according to Serratos et al. [15] Normalization was performed with Beta-actin antibody from Sigma (A3854). Blots were revealed using Immobilon Western Chemiluminescent HRP Substrate (Millipore Corporation, USA). The blots were scanned with an Amersham Imager 600 RGB (GE) and densitometry was analyzed using ImageQuant TL 8.1 software.

Characterization by flow cytometry
To evaluate surface markers associated with fibroblasts and macrophage, a flow cytometry (FC) assay was performed according to Landa-Solís C et al. [16] Cells were marked with monoclonal antibodies PE-conjugated CD166 and PE-conjugated CD14 from BD PharMigenTM (San Diego, CA, USA). Data were collected through a BD FACSCalibur flow cytometer and analyzed with Cell-QuestTM PRO software (Becton-Dickinson).
Cell stimulation, viability, and apoptosis FLS were treated for 24 h with MSU crystals at 0, 60, 75, 80, and 100 μg/mL. Cell viability was assessed by the crystal violet method [17] after MSU crystal cell stimulation. Based on these results, only one concentration was used for all subsequent tests. FLS apoptosis was assessed by FC detection of annexin V using a commercial kit (Annexin V Alexa Fluor 488 from Molecular Probes). Treatment with 100 μM H 2 O 2 for 30 minutes was used as positive control for oxidation because an increase in H 2 O 2 formation is associated with inflammation and fast OS induction in cells [18]. Unstimulated cells were used as negative control.

Protein oxidation
After derivatization using 2, 4-dinitrophenylhydrazine (DNPH), the protein oxidation products were identified by scanning carbonyl groups with the OxyblotTM Protein Oxidation Detection Kit (Millipore Inc.) according to the manufacturer's instructions. Image detection was performed with two methods: a conventional chemiluminescent detection and a fluorescence method using ECL Plex goat-alfa-rabbit IgG-Cy5 (GE, Healthcare), a 630 nm excitation filter and a 670 nm emission filter. The images were scanned with Amersham Imager 600 RGB (GE, Healthcare), and analyzed using ImageQuant TL 8.1 software.

Morphostructural characterization by transmission electron microscopy
The FLS were fixed with 2.5 % glutaraldehyde, treated with 1 % osmium tetroxide, and dehydrated with alcohol and propylene oxide. The samples were embedded in an epoxy resin and polymerized at 60°C for 24 h. Sections were cut 80-90 nm thick, and stained with 4 % uranyl acetate and lead citrate. The cells were then analyzed under a transmission electron microscope (TEM; Philips, model Tecnai 10) equipped with a Mega View II digital camera. A voltage of 80 kV was employed.

Statistical analysis
Each experiment was performed at least three times with the sample from each patient in independent experiments. Mean values were statistically analyzed with GraphPad Prism v. 6.0 using variance analysis, followed by the one-way post hoc Dunnett test. P < 0.05 was considered statistically significant.

MSU crystallization
As evidenced by polarized light microscopy and SEM, chemically synthesized MSU crystals exhibited a characteristic needle-shaped negative birefringence, a 5-40 μm range size, and triclinic structure [19]. In addition, two blinded experts reported similar morphological characteristics in synthetic MSU crystals and those obtained from the synovial fluid of patients during an acute gouty attack, and they were unable to differentiate between the two types of crystals (data not shown) (see Additional file 2).

FLS characterization
FLS isolated during SM biopsies expressed the PDH4 protein, UDGH gene and CD166 membrane receptor, as assessed by WB, IFA, qRT-PCR and FC, respectively (data not shown) (see Additional files 3, 4, and 5).

Assessment of cell viability and apoptosis
Cell cultures exposed to MSU crystals at a concentration of 75 μg/mL maintained a viability of 77 % ± 0.50, while a concentration of 100 μg/mL resulted in a viability of 54.82 % ± 0.46 compared to unstimulated FLS (P ≤ 0.05) (Fig. 1a). FLS exposed to 75 μg/mL MSU had invagination of crystals in the cytoplasm and cellular stress (Fig. 1b).
Adding MSU crystals at 75 μg/mL or H 2 O 2 at 100 μM induced apoptosis in 42 and 45 % of cells, respectively, compared to 21 % cell apoptosis observed in unstimulated cultures. This increment was significant (P ≤ 0.05) (Fig. 1c). Based on these results, we used a concentration of 75 μg/mL MSU for all subsequent experiments.

Evaluation of oxidative stress
MSU crystals induced a 2.5-fold increase in intracellular production of H 2 O 2 in comparison to untreated FLS (P < 0.05), and a 2.1-fold increase compared to a positive control (Fig. 2a-c and g). Similarly, stimulating FLS with crystals for 24 h or with H 2 O 2 for 30 minutes yielded 1.5-fold and 1.8-fold increases in O 2 -, respectively, compared to untreated FLS (Fig. 2d-f and g). Finally, only a 1.5-fold increase in NO was observed in FLS upon exposure to MSU crystals (Fig. 3).
Analysis of oxidized proteins FLS exposed to MSU crystals had a protein oxidation pattern similar to the one observed in H 2 O 2 -treated cells, in comparison to unstimulated FLS (Fig. 4a, b).

Ultrastructural analysis
Untreated FLS were analyzed under transmission electron microscopy (TEM). The cells exhibited irregular nuclei containing loose chromatin; the mitochondria, rough endoplasmic reticulum (ER) and vacuoles were distributed homogeneously throughout the cytoplasm; and cytoplasmic prolongations were observed (Fig. 5a, b).
In addition to the presence of intracellular crystals, MSU crystal-stimulated FLS had an increased number of vacuoles and a reduction of ER (Fig. 5c). Furthermore, aggregates of misfolded glycoproteins (MP) were evident in the lumen (Fig. 5d). These aggregates were also observed to a lesser extent in the FLS incubated with H 2 O 2 , although few cytoplasmic prolongations were present (Fig. 5e, f ).

Discussion
The current study revealed that MSU crystals are able to decrease cell viability through apoptosis induction in FLS. Although the definitive mechanism for MSUinduced apoptosis has not been established, it has been demonstrated that MSU crystals exert different apoptotic effects depending on the cell type interacting with the crystals. While some studies have reported that MSU crystals inhibit neutrophil apoptosis [20], others have shown that they do not induce any change in the percentage of apoptosis for osteoblast-like cells [21]. Recently, MSU crystals have been shown to promote renal The arrows indicate the intracellular vacuoles of MSU crystals. c Apoptosis is revealed by Annexin V detection (yellow arrows) in synoviocytes exposed to MSU crystals and H 2 O 2 (100 μM). Additionally, columns show quantification of the apoptotic cells by flow cytometry. Values are expressed as the mean ± standard deviation *P < 0.05 vs control cell apoptosis through a mechanism involving ROS generation [22]. However, no data were available on their influence on FLS. According to one report, apoptosis is induced in chondrocytes isolated from patients with RA following stimulation with MSU crystals [23]. The association of this apoptotic state with the loss of cartilage repair and regeneration capacity could highlight a link between FLS apoptosis and the tissue damage observed in gouty patients. Moreover, the relationship between the increment of ROS and NO and the loss of FLS viability caused by MSU crystals is consistent with published findings [24][25][26].
In addition, we established that crystal-exposed FLS produced H 2 O 2 , O 2 and, to a lesser extent, NO, promoting a state of cellular oxidation. One mechanism involved in ROS production is the NADPH oxidase system in THP-1 cells stimulated with MSU crystals [27]. This mechanism of ROS generation has also been shown in FLS from patients with OA and RA that were exposed to TNF-α and IL-1β, exhibiting a heightened state of cellular oxidation [28]. Our experiments proved, via an increase in ROS/RNS, that MSU crystals activated an oxidative state in FLS. The increase in H 2 O 2 observed in FLS exposed to MSU crystals for 24 h is similar to that reported for FLS stimulated with advanced oxidation proteins products; a threefold to eightfold increase in H 2 O 2 was observed compared to unstimulated control cells [29]. This suggests that MSU crystal-mediated ROS overproduction in FLS is involved in the disturbance of homeostasis within the joint microenvironment, which can damage all cellular components, including DNA, lipids and proteins [30]. However, proteins are possibly the most immediate vehicle for inflicting oxidative damage on cells because they are often catalysts. Therefore, we assessed the influence of ROS in oxidized protein content of FLS affected by MSU crystals.
The impact of ROS on the proteins of FLS with MSU crystals was clearly seen on images because there were more spots and with higher intensities than in control cells, indicating increased carbonyl content. While there are no reports that can be directly compared to our data, accumulation of protein carbonyls [31] has been observed in some rheumatic diseases (including RA and psoriasis), but it is known that exposure of proteins to ROS leads to denaturalization, loss of function, crosslinking, aggregation, and fragmentation. Under these conditions, it is suggested that accumulation of some compounds in the joint, like glycosaminoglycans and hyaluronic acid, cause damage by reducing joint viscosity [32]. However, there are no studies of the underlying mechanism. We suspected that the increase in OS might be contributing to synovial cell damage altering the functional and structural integrity. Therefore, we visualized OS-induced ultrastructural changes triggered by MSU-crystals in gout. In our model, we observed an increase in rough ER and in the presence of MP aggregates due to cellular stress in the FLS. These findings are similar to those described for synoviocytes exposed to an adjuvant used for treating arthritis (i.e., a reduction of the Golgi apparatus, mitochondria and ER [33]), and to the ones describing the appearance of vacuoles in FLS cytoplasm due to the internalization of particles. In addition, intracellular lysosomes and other cytoplasmatic formations were found [34], and these morphological changes suggest the induction of autophagy in the cells [35].