Dimethylarginine dimethylaminohydrolase 2, a newly identified mitochondrial protein modulating nitric oxide synthesis in normal human chondrocytes

Abstract

Objective

The mitochondrion is known to be important to chondrocyte survival. This study was undertaken to analyze protein expression profiles in chondrocyte mitochondria that are affected by interleukin-1β (IL-1β).

Methods

Normal human chondrocytes were isolated from knee cartilage obtained at autopsy from subjects with no history of joint disease. Cells were incubated for 48 hours with or without IL-1β (5 ng/ml). Proteins were separated by 2-dimensional electrophoresis and stained with Sypro Ruby, Coomassie brilliant blue, or silver. Qualitative and quantitative analyses were carried out using PDQuest software. Proteins were identified by mass spectrometry using matrix-assisted laser desorption ionization–time-of-flight/time-of-flight technology. The proteomic results were validated by real-time polymerase chain reaction, Western blotting, and microscopy. Nitric oxide (NO) was quantified using Griess reagent.

Results

Comparative analysis revealed differential expression of signal transduction proteins that regulate cytoskeleton, transcription, metabolic, and stress-related pathways. In total extracts, dimethylarginine dimethylaminohydrolase 2 (DDAH-2) did not show any change in expression after stimulation with IL-1β. However, in mitochondrial extracts, DDAH-2 expression was significantly increased after exposure to IL-1β. Conventional immunofluorescence and confocal microscopy revealed the presence of DDAH-2 in the mitochondria of IL-1β–stimulated chondrocytes. These results were reproducible in cartilage explants treated with IL-1β. In addition, we demonstrated that inhibition of the expression of DDAH-2, as well as interruption of its translocation to the mitochondria, reduced the NO production induced by IL-1β. DDAH-2 protein expression was higher in osteoarthritic (OA) cartilage than in normal cartilage.

Conclusion

In the present study, the presence of DDAH-2 in normal human chondrocytes and cartilage was identified for the first time. DDAH-2 could play an important role in IL-1β–induced NO production and in OA pathogenesis.

In most cells, the mitochondrion is an important organelle, because it plays a role in energy production and is a target and a generator of free radicals, as well as being a sensor of apoptotic/survival signals (1, 2). The chondrocyte is the only cell type present in mature cartilage and is responsible for cartilage repair and maintenance (3, 4). In recent years, links between mitochondria and chondrocyte function have been identified. The activity of the mitochondrial complexes succinate dehydrogenase (complex II) and ubiquinol cytochrome c reductase (complex III) is known to be down-regulated in osteoarthritis (OA) (5). In addition, nitric oxide (NO) seems to inhibit the activity of cytochrome c oxidase (complex IV) (6). Johnson et al have reported that a substantial decrease in mitochondrial ATP generation contributes to the pathogenesis of OA (7). Inflammatory cytokines, such as interleukin-1β (IL-1β) and tumor necrosis factor α (TNFα), inhibit mitochondrial complex I, affecting the oxidative phosphorylation process and mitochondrial membrane potential (8). The catabolic program induced by these proinflammatory stimuli is characterized by secretion of proteinases, suppression of matrix synthesis, and reduction in the number of chondrocytes (9, 10). Consequently, there are recently recognized important dimensions to the role of inflammation in OA progression (11).

In a previous report, we described the total proteomic profile of chondrocytes stimulated with IL-1β and TNFα, showing a clear up-regulation of metabolic pathways related to TNFα in IL-1β–stimulated cells (12). We also demonstrated that some proteins regulated by IL-1β had mitochondrial locations. There are, however, no previous reported studies of the exact effects of IL-1β on the mitochondrial proteome of normal human chondrocytes. The present study identifies, for the first time, the role of IL-1β in regulation of the mitochondrial proteome profile of normal human chondrocytes. Of particular interest is the finding that all of the proteins identified were up-regulated by IL-1β. We also identified a protein that was not previously recognized in chondrocytes, dimethylarginine dimethylaminohydrolase 2 (DDAH-2), and were able to show that IL-1β induced the translocation of DDAH-2 to the mitochondria and to demonstrate the role of DDAH-2 as a regulator of NO production. These findings may provide a better understanding of the participation of chondrocyte mitochondria in stress signaling and NO pathways and their relationship to the development of rheumatic diseases such as OA.

MATERIALS AND METHODS

Cartilage procurement and processing.

This study was approved by the local ethics committee in Galicia, Spain. Normal human knee cartilage from autopsy subjects with no history of joint disease was provided by the Tissue Bank and the Autopsy Service at Complejo Hospitalario A Coruña. The specimens were from subjects ages 18–65 years at the time of death. Cartilage slices were removed from the condyles and treated with 0.5 mg/ml trypsin (Sigma-Aldrich) for 15 minutes at 37°C. The cartilage was incubated overnight in an orbital shaker at 37°C with 2 mg/ml clostridial collagenase (Sigma-Aldrich) in Dulbecco's modified Eagle's medium (DMEM; Gibco Life Technologies). The cells were then resuspended in fetal calf serum (FCS)–enriched DMEM and were used in the first passage. For tissue studies, pieces of cartilage 6 mm in diameter and 4 mm in height were cut from cartilage and stimulated in DMEM.

Primary chondrocyte culture and cell stimulation.

Chondrocytes were seeded in DMEM supplemented with 100 units/ml penicillin, 100 μg/ml streptomycin, 1% glutamine, and 10% FCS, in 162-cm² culture flasks for 2-dimensional (2-D) electrophoresis studies, 6-well plates for messenger RNA (mRNA) and unidimensional protein studies, chamber slides for immunofluorescence studies, and 96-well plates for NO assays. Culture flasks, plates, and slides were all from Costar. The cells were incubated in a humidified gas mixture containing 5% CO₂ in air at 37°C. When cells reached confluence, they were made quiescent by incubation for 48 hours in medium containing 0.5% FCS. Experiments were also performed without FCS. The cells were treated with IL-1β (5 ng/ml; Sigma-Aldrich) for predetermined amounts of time. Monensin (0.7 μl/ml; Biosciences) was used to inhibit protein transport, and homocysteine (10 μM; Sigma-Aldrich) was used to inhibit DDAH-2 expression.

Protein sample preparation and isolation of crude mitochondria.

Chondrocytes (3–5 × 10⁶) were recovered by trypsinization and resuspended in 10 mM NaCl, 1.5 mM CaCl₂, 10 mM Tris HCl (pH 7.5), 1 mM phenylmethylsulfonyl fluoride (PMSF; Sigma-Aldrich), and 10 μl protease inhibitor cocktail (Sigma-Aldrich). Ice-cold sucrose buffer (0.4 volume; 2.5×) was added. The homogenates were centrifuged at 1,200g for 10 minutes at 4°C, yielding a mitochondrial pellet and a cytosol supernatant fraction. The pellet was carefully resuspended in sucrose buffer (1×), and, to obtain the final mitochondrial pellet, the sample was diluted in a buffer containing 8.4M urea, 2.4M thiourea, 5% CHAPS, 1% carrier ampholytes, immobilized pH gradient buffer (pH 3–10 nonlinear; GE Healthcare), 0.4% Triton X-100, and 2 mM dithiothreitol.

Two-dimensional electrophoresis.

Seventy-five micrograms of protein from a pool of 4 donors was incubated with rehydration buffer (8.4M urea, 2M thiourea, 2% CHAPS, 0.5% carrier ampholytes, 1.2 % Destreak Reagent [GE Healthcare], and 0.002% bromphenol blue) and applied to 24-cm immobilized pH gradient buffer strips (pH 3–10 nonlinear). Gels from each pool were run in duplicate. Isoelectric focusing was performed for a total of 64,000V per hour in an IPGphor (GE Healthcare). Electrophoresis was carried out at 2.5W per gel for the first 30 minutes and then at 17W per gel until the dye reached the bottom of the gel. The 2-D electrophoresis gels were stained with Coomassie brilliant blue, silver nitrate, or Sypro Ruby (Molecular Probes–Invitrogen). Coomassie brilliant blue staining used 0.1% Brilliant Blue G-250 (Sigma-Aldrich) in 40% methanol with 10% acetic acid. For silver nitrate staining the gels were fixed overnight at 4°C in 40% ethanol with 10% acetic acid. Sensitization was achieved in 0.02% sodium thiosulfate (Fluka). After 2 washes with distilled water, gels were impregnated with 0.2% silver nitrate (Fluka) in 0.075% formalin for 60 minutes. The stain was developed in 3% potassium carbonate, 12.5 mg/liter sodium thiosulfate, and 0.025% formalin. The reaction was stopped by transferring the gels to 3% Trizma-base (Sigma-Aldrich) in 10% acetic acid. Sypro Ruby staining was performed according to the manufacturer's instructions.

Image acquisition and data analysis.

Silver- and Coomassie brilliant blue–stained gels were digitized using a densitometer (ImageScanner; GE Healthcare). Sypro Ruby–stained gels were digitized using a CCD camera (LAS 3000 imaging system; Fuji). Gel images were analyzed using PDQuest 7.3.1 software (Bio-Rad). Protein spots were quantified and characterized by their molecular mass and isoelectric point using bilinear interpolation between landmark features on each image previously calibrated with internal 2-D electrophoresis standards (Bio-Rad). Protein expression data from each gel were normalized for the total density present in the gel images.

Mass spectrometry (MS) analysis and database searches.

The protein spots of interest were excised from gels with up to 0.5 mg of protein. Briefly, the spots were washed, shrunk with 100% acetonitrile, and dried in a SpeedVac (Savant). The samples were digested using 12.5 ng/μl trypsin (Roche Diagnostics). Supernatants were collected and spotted onto a matrix-assisted laser desorption ionization (MALDI) target plate (98 × 2–spot Teflon-coated plates). Then, 3 mg/ml α-cyano-4-hydroxycinnamic acid matrix (Sigma-Aldrich) in 50% acetonitrile was added and allowed to air-dry. MALDI time-of-flight MS (MALDI-TOF MS) analyses and MS/MS sequencing analyses were carried out using the MALDI-TOF/TOF mass spectrometer 4800 Proteomics Analyzer (Applied Biosystems). The monoisotopic peptide mass fingerprinting data obtained from MALDI-TOF MS were used to search for protein candidates using Mascot (http://www.matrixscience.com). Searches were performed using the Swiss-Prot/TrEMBL (http://www.expasy.ch/sprot) and NCBI (http://www.ncbi.nlm.nih.gov) databases. Identifications were considered positive based on a score calculated from the number of matched peptides and the coverage of the theoretical sequences, with a mass accuracy of 50 ppm from internal calibration. In MALDI-TOF/TOF analysis, the amino acid sequence tag obtained from each peptide fragment was used for protein identification.

DDAH-2 mRNA analysis.

Messenger RNA from 5 × 10⁵ cells per condition was isolated with TRIzol reagent (Invitrogen), treated with Deoxyribonuclease I amplification grade (Invitrogen), and amplified with a Transcriptor First Strand cDNA Synthesis commercial kit (Roche Diagnostics). Polymerase chain reaction (PCR) analyses for DDAH-2 and the housekeeping gene porphobilinogen deaminase (PBGD) were conducted with the LightCycler 4800 SYBR Green I Master kit using the Real Time Light Cycler (Roche Diagnostics). The primers used were as follows: DDAH-2 sense 5′-GACTCCCTTCTCCACCAACTC-3′, antisense 5′-TTCTTGTTTCTTCACCTGTCTCC-3′; PBGD sense 5′-AGCTATGAAGGATGGGCAAC-3′, antisense 5′-TTGTATGCTATCTGAGCCGTCTA-3′. PCR data were analyzed using REST software (Qiagen), which provides statistical information for comparing groups with issues of reaction efficiency and reference gene normalization taken into account.

Western blotting.

After stimulation, cells (5 × 10⁵/well) were lysed in 0.2M Tris HCl (pH 6.8) containing 2% sodium dodecyl sulfate, 20% glycerol, 1 μg/ml cocktail inhibitor, and 1 mM PMSF. Protein concentrations were determined using the BCA assay (Pierce). We also hybridized each membrane with mouse anti-human anti–α-tubulin or α-ATPase (1:2,500; Sigma-Aldrich) in an LAS 3000 image analyzer. Quantitative changes in band intensities were evaluated with ImageQuant 5.2 software (GE Healthcare). Pathway Studio 7.0 software (Ariadne Genomics) was used to study interactions among the identified proteins, as well as their roles.

DDAH-2 detection by immunofluorescence.

Chondrocytes were seeded at 5 × 10⁴ per well. To localize the mitochondria, Mitotracker solution was added according to the instructions of the manufacturer (Invitrogen). The cells were then incubated with 100 μl of anti-human DDAH-2 (1:250) for 1 hour at room temperature. Phycoerythrin (PE)–labeled goat anti-rabbit antibody (1:20) was added, with further incubation in the dark for 30 minutes at room temperature. Nuclear staining was performed with DAPI (2 μg/ml). Cells were visualized by fluorescence microscopy with a Leica DMSL microscope connected to a Leica DC100 digital camera or to a Nikon A1R confocal microscope. NIS-Elements software (Nikon) was used for microscopy and photography. To perform immunofluorescence studies on tissue, samples were cut into 4-μm serial sections using a cryostat at –30°C. The sections were then incubated with primary anti–DDAH-2 antibody (1:250) and PE-conjugated goat anti-rabbit antibody (1:20). Nuclear staining was performed with DAPI, and the samples were examined with a microscope connected to a Leica DC100 digital camera.

Statistical analysis.

Assays of samples from individual donors were performed in duplicate. The mean and SEM were calculated. SPSS version 15.0 was used to perform analysis of variance and Tukey tests. In the proteomic analysis, normalization tools and the statistical package from PDQuest software were used. P values less than 0.05 were considered significant.

RESULTS

Differential mitochondrial proteomic profiles of human chondrocytes exposed to IL-1β and chondrocytes not exposed to IL-1β.

The mitochondrial proteomic profiles of chondrocytes were determined after 48-hour incubation of cells with IL-1β at 5 ng/ml. For both control cultures and cultures with IL-1β treatment, mitochondrial protein extracts from 4 donors were pooled and experiments were performed in duplicate under each experimental condition. After duplicate gels were run under each condition, PDQuest software detected 360 common spots that matched in both conditions (control and IL-β–exposed). (2-D electrophoresis maps are available from the corresponding author upon request). Of these matched spots, 6 were differentially regulated between the non–IL-1β–exposed control and IL-1β–exposed conditions, showing different patterns of protein expression. Using MS we identified these 6 spots as corresponding to 6 different proteins. Biologic functions of the identified proteins include signal transduction, stress and defense, transcription, protein synthesis and turnover, metabolism, and cellular organization (Table 1).

Table 1. Proteins exhibiting significantly altered expression in normal human chondrocytes from 4 subjects after interleukin 1β (IL-1β) exposure, as determined by PDQuest analysis*

* SAP18 = Sin3A-associated protein; 18 kd. IL-1β = interleukin-1β. † Calculated from the number of matched peptides and the coverage of the theoretical sequences.
Swiss-Prot accession no.	O00422	O95865	P23381	P52895	P07355	P04179
Protein name	Histone deacetylase complex subunit SAP18	Dimethylarginine dimethylamino-hydrolase 2	Tryptophanyl–transfer RNA synthetase	Aldo-keto reductase family 1 member C2	Annexin A2	Mitochondrial superoxide dismutase
Cellular role	Signal transduction	Stress and defense	Transcription, synthesis, turnover	Metabolism	Cellular organization	Stress and defense
Location	Nucleus	Mitochondrion, cytosol	Cytosol	Cytosol	Mitochondrion	Mitochondrion
Mr
Experimental	44.2	35.6	46.1	39.3	37.4	24.0
Expected	17.6	29.9	53.5	37.1	38.9	24.9
pI
Experimental	5.6	5.7	6.2	7.4	7.5	8.0
Expected	9.4	5.7	5.8	7.1	7.6	8.3
% coverage of theoretical sequence	38	33	32	25	63	43
Score†	56	84	99	59	250	98
No. of identified peptide sequences	6	10	11	6	25	9
Ratio of expression, IL-1β exposure vs. basal	4.23	2.59	2.19	3.40	8.13	4.10

All of the identified proteins were up-regulated with exposure to IL-1β, compared with control conditions. The greatest increase was in annexin A2 (ANXA2), which increased 8.13-fold after treatment. This protein plays a role in the reorganization of the cytoskeleton. Other proteins whose levels were increased by exposure to IL-1β were superoxide dismutase 2 (SODMn) (4.10-fold), aldo-keto reductase family 1 member C2 (AKR1C2) (3.40-fold), and tryptophanyl–transfer RNA synthetase (SYW), which regulates tryptophan formation (2.19-fold).

Interestingly, we found 2 proteins, DDAH-2 and histone deacetylase complex subunit SAP18 (SAP18), that have not previously been identified in cartilage or in chondrocytes. SAP18 and DDAH-2 were increased 4.23-fold and 2.59-fold, respectively, by exposure to IL-1β.

Validation experiments.

To verify the results obtained by 2-D electrophoresis and MS, we evaluated whether one of the proteins modulated by IL-1β was detectable by other techniques. Total protein extracts of IL-1β–treated cells were separated by Western blotting, but DDAH-2 did not show any variation (Figure 1A). However, using mitochondrial extracts, we observed a significant increase (mean ± SEM 4.13 ± 1.41–fold) (P < 0.05) in the protein level of DDAH-2 with IL-1β exposure (P < 0.05)(Figure 1B). To confirm the mitochondrial location of DDAH-2 we performed immunofluorescence and confocal microscopy studies. Figure 2A shows that DDAH-2 was more evident around the nuclei in the non–IL-1β–exposed controls, whereas in IL-1β–treated cells the protein was located within mitochondria, as shown by Mitotracker staining. The presence of DDAH-2, specific to mitochondria, was confirmed using confocal microscopy (Figure 2B).

Figure 1. Results of experiments performed to verify, at the protein level, the findings obtained by 2-dimensional electrophoresis and mass spectrometry. A, Western blots showing dimethylarginine dimethylaminohydrolase 2 (DDAH-2) protein in normal human knee articular chondrocytes treated with interleukin-1β (IL-1β) (5 ng/ml, 48 hours). DDAH-2 was measured in total cell extracts and in mitochondrial (mit.) extracts. Results shown are representative of 4 experiments with total extracts and 10 experiments with mitochondrial extracts. B, Quantification of DDAH-2 protein expression in mitochondrial extracts. Chondrocytes from 10 different donors were used. Values are the mean ± SEM. ∗ = P < 0.05 versus basal level.

Figure 2. Localization of DDAH-2 protein, as determined by conventional immunofluorescence (A) and confocal immunofluorescence (B). Chondrocytes were left unstimulated or were stimulated for 48 hours with IL-1β (5 ng/ml). Left panels show the cells after incubation with anti–DDAH-2 and staining with DAPI. Middle panels show localization of mitochondria after staining with Mitotracker (Mitot) and DAPI. Red fluorescence represents abundant DDAH-2, blue fluorescence shows nuclei stained with DAPI, and green fluorescence represents mitochondria. Right panels are overlays of the images obtained with anti–DDAH-2 plus DAPI and with Mitotracker plus DAPI. These results appear as yellow fluorescence when there is colocalization of DDAH-2 staining and mitochondrial staining, as is evident in IL-1β–treated cells. Results shown are representative of findings in chondrocytes from 3 different donors. See Figure 1 for other definitions.

We also tested whether IL-1β induced the transport of DDAH-2 to the mitochondria. Cells were incubated for 24 hours with monensin (a general inhibitor of protein transport) (0.7 μl/ml), with monensin plus IL-1β (5 ng/ml), or with IL-1β alone. As shown in Figure 3, monensin alone produced no change in DDAH-2 localization, but when it was administered in combination with IL-1β, the level of DDAH-2 in the mitochondrial compartments was reduced compared to the levels induced by IL-1β alone. Interestingly, the location of DDAH-2 in the mitochondria was also demonstrated in cartilage slices from normal human donors after treatment with IL-1β at 5 ng/ml for 48 hours (Figure 4A).

Figure 3. Effect of monensin (Mon), a general inhibitor of protein transport, on IL-1β–induced transport of DDAH-2 into mitochondria. Normal human knee articular chondrocytes were incubated for 24 hours under control conditions or with monensin (10 μM) alone, IL-1β (5 ng/ml) alone, or monensin and IL-1β. Red fluorescence represents DDAH-2, blue fluorescence shows nuclei stained with DAPI, and green fluorescence represents mitochondria (stained with Mitotracker [Mitot]). Reduced yellow fluorescence in cells incubated with monensin and IL-1β and stained with Mitotracker, anti–DDAH-2, and DAPI indicates reduced levels of DDAH-2 in mitochondria. Results shown are representative of findings in chondrocytes from 3 different donors. See Figure 1 for other definitions. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131.

Figure 4. DDAH-2 in normal and osteoarthritic (OA) human knee articular cartilage. A, Presence of DDAH-2 protein, as determined by conventional immunofluorescence, in unstimulated (left) and IL-1β–stimulated (48 hours) (right) cartilage explants. B, DDAH-2 immunofluorescence in normal cartilage (left) and OA cartilage (right). In A and B, red fluorescence represents DDAH-2 protein, blue fluorescence shows nuclei stained with DAPI, and green fluorescence represents mitochondria. Results shown are representative of findings in chondrocytes from 3 different donors. C, Quantification of DDAH-2 levels in normal and OA cartilage. Values are the mean ± SEM of 3 experiments. ∗ = P < 0.01 versus control. See Figure 1 for other definitions. Color figure can be viewed in the online issue, which is available at http://online library.wiley.com/journal/10.1002/(ISSN)1529-0131.

In order to investigate whether DDAH-2 is expressed under disease conditions, we carried out immunohistologic experiments in normal and OA cartilage. In these studies, levels of DDAH-2 protein were found to be higher in OA cartilage than in normal cartilage (Figures 4B and C). DDAH-2 distribution analysis showed that DDAH-2 levels were higher in the deep zone of OA cartilage than in superficial-zone cartilage (mean ± SEM 72 ± 0.24% DDAH-2–positive cells versus 44 ± 0.14%) (data not shown).

DDAH-2, NO production, and mitochondria.

To study the potential role of DDAH-2 in regulating NO production in chondrocytes, as it does in many other cell types, we evaluated the effect of homocysteine, an inhibitor of DDAH-2 expression, in NO release. In these experiments, chondrocytes were first costimulated for 24 hours with homocysteine (10 μM) and IL-1β (5 ng/ml) or with IL-1β alone. Our results confirmed that homocysteine reduced the expression of DDAH-2 mRNA in IL-1β—stimulated chondrocytes, as it does in other cell types (mean ± SEM ratio of DDAH-2 with 10 μM homocysteine versus with 5 ng/ml IL-1β [set at 1] 0.61 ± 0.18; n = 4) (P < 0.05) (Figure 5A).

Figure 5. Effect of DDAH-2 on nitric oxide (NO) synthesis. A, Expression of DDAH-2 mRNA after treatment of normal human knee articular chondrocytes with IL-1β (5 ng/ml) alone or in combination with homocysteine (Hcys) (10 μM) for 24 hours. B, NO production, measured by Griess reaction, in cells left untreated or treated for 24 hours with homocysteine alone, IL-1β alone, or homocysteine and IL-1β. C, NO release by IL-1β–stimulated chondrocytes with or without treatment with monensin (Mones) to inhibit translocation of DDAH-2 into the mitochondria. Values are the mean ± SEM of duplicate experiments using chondrocytes from 4 different donors. & = P < 0.05 versus IL-1β alone; ∗ = P < 0.05 versus basal level. See Figure 1 for other definitions.

After having established that homocysteine reduced DDAH-2 expression, we evaluated the effect on NO production. After 24 hours of incubation with IL-1β, the amount of NO was 179.75 ± 32.03% (mean ± SEM; n = 4) of that in control chondrocytes (P < 0.05). However, when homocysteine was added to the IL-1β treatment, the amount of NO release was significantly reduced compared with that observed with IL-1β alone (84.02 ± 16.70% of control; n = 4) (P < 0.05 versus IL-1β alone) (Figure 5B). We then tested whether blocking the transport of DDAH-2 into the mitochondria could also affect production of NO. Cotreatment with monensin (0.7 μl/ml) and IL-1β (5 ng/ml) for 24 hours reduced the level of NO release to 62.68 ± 17.56% of control (P < 0.05 versus IL-1β alone) (Figure 5C). These results demonstrate the role of DDAH-2 in the regulation of the NO production mediated by IL-1β in chondrocytes.

DISCUSSION

The present study is, to our knowledge, the first to compare the proteome of mitochondrial proteins in unstimulated and IL-1β–stimulated chondrocytes. We found 6 proteins, AKR1C2, ANXA2, SAP18, SODMn, SYW, and DDAH-2, that were up-regulated after 48 hours of treatment with IL-1β. After identifying these proteins we observed that most of them are active in metabolic and stress-related pathways, indicating that IL-1β induced a high level of activity in the cell. Using Pathway Studio 7.0 software, we studied the role of some of these proteins in cellular pathways, as well as in certain disease states (additional information available from the corresponding author upon request).

AKR1C2 has a key role in the metabolism of steroid hormones and prostaglandin E synthesis, and it confers resistance to certain tumor cells in lung, prostate, and breast cancer, regulating survival cascades (13–15). ANXA2 is important in the remodeling of the cytoskeleton and in the cell proliferation process, because of its role as a vesicular transporter in membrane–membrane or membrane–cytoskeleton connections. It can bind CD44, one of the principal cellular receptors of hyaluronic acid, in the formation of compartments rich in cholesterol (16). Histone deacetylases, such as SAP18, have been shown to regulate matrix metalloproteinases and aggrecanases in cartilage, and inhibitors of histone deacetylases could block degradation of the extracellular matrix and inhibit the expression of these enzymes mediated by proinflammatory enzymes (17).

SODMn is a protein related to stress signals; it eliminates O₂⁻ and releases H₂O₂ (18). It is important to note that the mitochondrion is one of the principal producers of reactive oxygen species (ROS). This is particularly relevant with regard to diseases related to aging, in which accumulation of ROS is very high (19, 20). Superoxide can damage DNA, inducing mutations, and also reacts with such molecules as NO, forming peroxynitrite (ONOO⁻), which initiates DNA strand breakage and modifications of purines and pyrimidines (21, 22). A decrease in the expression of extracellular SODMn has been observed in OA chondrocytes (23, 24), and our group has obtained similar results using mitochondrial fractions of OA and normal chondrocytes in culture (25). In contrast, other groups have found increased levels of SODMn in synovial fluid from patients with rheumatoid arthritis (26), and, in accordance with those results, we showed in a previous study that SODMn in chondrocytes was increased after IL-1β incubation (12). It seems that in a short-term model of acute inflammation, an increased level of SODMn is a key factor for moderating ROS levels. SYW regulates ERK, Akt, and endothelial NOS (eNOS) activation pathways that are associated with angiogenesis, cytoskeletal reorganization, and sheer stress–responsive gene expression.

From all identified proteins we found DDAH-2, a stress-related protein, to be increased by IL-1β; this protein has not been previously identified in chondrocytes. In other cell types, DDAH-2 has the ability to inhibit asymmetric dimethylarginine (ADMA), a natural inhibitor of NO synthase (NOS). The DDAH family of proteins is formed by two isoforms, DDAH-1 and DDAH-2, both of which utilize ADMA as their substrate and increase NOS activity; however, their regulation, cellular location, and tissue specificity are very different (27). DDAH-2 is the most prevalent isoform in blood vessels and endothelium and is important for embryonic development (27). NOS enzymes convert L-arginine to L-citrulline, producing NO. Thus, the increase in NOS activity leads to an increase in NO levels, a characteristic phenomenon observed after exposure to IL-1β (28, 29) and in cartilage and synovial fluid in many articular diseases. DDAH-2 has also been linked to other diseases, such as hypertension, cerebral hemorrhage, cardiac disease, and diabetes (27, 30–32).

A surprising finding of the present study was the demonstration of the effect of IL-1β in transport of DDAH-2 to the mitochondria, a localization not previously described in any cell type. In cultured human endothelial cells, DDAH-1 has been associated with the cytosol and the nucleus, but DDAH-2 had been found only in the cytosol (33). However, in a recently reported study, DDAH-2 was found in the nuclei of rat vascular smooth muscle cells (27). In Western blot analyses of total cell extracts as well as mRNA analyses (results not shown), we found no difference between control and IL-1β–exposed conditions. However, assays using purified mitochondrial proteins and conventional and confocal microscopy confirmed the results from 2-D electrophoresis in mitochondrial extracts, showing that DDAH-2 was found only in the mitochondria of IL-1β–treated cells. This effect was also observed in cartilage explants from human donors. Therefore, we tested whether DDAH-2 is translocated from the cytosol to the mitochondria by the action of IL-1β. We used an inhibitor of general protein transport, monensin (34, 35), to block the transport of DDAH-2 with IL-1β treatment. Interestingly, results similar to those obtained in IL-1–stimulated normal cartilage were found in OA cartilage, supporting the notion that NO and DDAH-2 have a role in OA.

After we ascertained that the protein migrated to the mitochondria, we examined the role of DDAH-2 in regulating NO production in chondrocytes, as it does in other cell types. To test this, we reduced the expression of DDAH-2 using homocysteine, as previously described (36, 37). As expected, treatment with 10 μM homocysteine reduced DDAH-2 mRNA expression in chondrocytes, but it also inhibited the production of NO induced by the cytokine. This finding demonstrates how DDAH-2 relates to NO production in IL-1β–stimulated cells. The family of NOS enzymes represents a complex group of proteins with two described isoforms in chondrocytes: a constitutive enzyme (eNOS) and an inducible isoform (inducible NOS [iNOS]) (38). DDAH-2 has been reported to interact with eNOS in the cytosol of endothelial cells and with iNOS in vascular smooth muscle cells (29). The isoform that is up-regulated by lipopolysaccharide, IL-1, or TNF in chondrocytes is iNOS, suggesting that DDAH-2 may directly or indirectly increase iNOS activity. We tested whether inhibition of the translocation of DDAH-2 into mitochondria also had an effect on NO production. We used monensin to block this transport, as described above, and found that levels of NO released to the supernatants were also reduced, indicating the important actions of DDAH-2 in chondrocyte mitochondria.

Further evidence of the relationship of DDAH-2 and mitochondria comes from studies showing that DDAH-2 gene expression is down-regulated by coupling factor 6, a mitochondrial protein of the ATP synthase complex, again establishing the relationship among the pathways of mitochondria, ROS, and NO (39, 40). It is well known that NO reduces the activity of mitochondrial complex IV in chondrocytes and that many other factors affect the mitochondrial respiratory chain, inducing inflammation processes (6, 41). NO can also interact with other ROS, producing ONOO⁻ and inducing a cascade of negative ROS-mediated mitochondrial and cellular effects, including DNA damage, inflammatory pathway activation, induction of the expression of apoptotic proteins, and cellular senescence. These combined results indicate the importance of DDAH-2 in the regulation of NO-related pathways, particularly in relation to chondrocyte mitochondria.

In conclusion, in this study of normal human knee chondrocytes we have identified some mitochondrial proteins that are linked to the inflammatory response and up-regulated by IL-1β exposure. In particular, DDAH-2 could represent a new focus in the study of NO production in chondrocytes, particularly in relation to the mitochondrial ROS-related events that could be involved in aging-related rheumatic diseases such as OA.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Blanco had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Cillero-Pastor, Mateos, Fernández-López, Oreiro, Ruiz-Romero, Blanco.

Acquisition of data. Cillero-Pastor, Mateos, Fernández-López, Oreiro, Ruiz-Romero, Blanco.

Analysis and interpretation of data. Cillero-Pastor, Blanco.