Synergistic induction of nitric oxide synthase type II: In vitro effect of leptin and interferon-γ in human chondrocytes and ATDC5 chondrogenic cells




To study, in vitro, the effect of leptin (OB), alone or in combination with interferon-γ (IFNγ), on inducible nitric oxide synthase (iNOS) and NO production in human primary chondrocytes and in mouse embryonic chondrogenic ATDC5 cells.


Leptin receptor expression and iNOS messenger RNA expression were evaluated by reverse transcriptase–polymerase chain reaction. Then, iNOS activity was indirectly studied by measuring nitrite accumulation, using the Griess colorimetric reaction, in culture medium of human primary chondrocytes and ATDC5 cells.


ATDC5 mouse embryonic cells expressed functional OB receptor. Alone, neither OB nor IFNγ produced nitrite accumulation in culture medium. However, costimulation with OB and IFNγ resulted in dose-dependent up-regulation of the expression of iNOS and NO production in human primary chondrocytes and ATDC5 cells. Production of NO was blunted by the iNOS-specific inhibitors L-NG-nitroarginine methyl ester and aminoguanidine. In addition, the janus-activated kinase 2 (JAK2)–specific inhibitor Tyrphostin AG 490 completely blocked OB + IFNγ–driven up-regulation of iNOS and NO production.


Our data show for the first time a putative proinflammatory role of OB via iNOS induction and NO production. This occurs via activation of JAK2.

Differentiated chondrocytes, the predominant cell type in normal mature cartilage, synthesize a cartilage-specific extracellular matrix to maintain matrix integrity. In osteoarthritis (OA) and rheumatoid arthritis (RA), this homeostatic process is altered (1) and structural and biochemical changes occur, including degradation of the cartilage matrix, scarcity of extracellular matrix due to the loss of chondrocyte phenotype, and increased numbers of apoptotic chondrocytes (2).

Interferon-γ (IFNγ) and other proinflammatory cytokines can modify cartilage via production of nitric oxide (NO). In turn, NO has the potential to regulate several cartilage functions, including diminished synthesis of matrix proteins, apoptosis (2), and modulation of matrix metalloproteinases (3).

Human chondrocytes and established cell lines in vitro up-regulate inducible nitric oxide synthase (iNOS) in response to cytokines, yet this often requires a synergistic or costimulatory effect of 2 or more cytokines (4).

Leptin (OB), the 16-kd product of the OB gene involved in feeding behavior and energy homeostasis, is a regulatory pleiotropic factor in a variety of pathophysiologic processes, including immune system modulation (5), bone growth (6), and inflammation (7). In cartilage, the involvement of OB in roles in addition to its classic actions is further supported by the evidence that this hormone is a potent stimulator of bone growth in OB-deficient (ob/ob) mice (6). In addition, human articular chondrocytes express the functional long isoform (8) of the OB receptor (OB-RL), which suggests that OB could regulate the activity of these cells (9). Interestingly, the OB-RL is a member of the class I cytokine receptor superfamily, which transduces signals via the janus-activated kinase (JAK)/signal transducer and activator of transcription pathway. This pathway is also activated by several other cytokines, including IFNγ. To evaluate the effects of OB on NOS induction in chondrocytes, we have determined nitrite accumulation following stimulation with OB and IFNγ, alone or in combination, in human primary chondrocytes and in the ATDC5 mouse chondrogenic embryonic cell line cultured in vitro. This cell line is characterized by a well-defined sequence of differentiation (10) that goes from chondroprogenitors to fully differentiated hypertrophic chondrocytes and exhibits different phenotypes depending on the differentiating stage.

Since both the OB-R and the IFNγ receptor (IFNγ-R) transduce signals via JAK2 activation, and considering that this tyrosine kinase is involved in iNOS induction, the aim of the current study was to elucidate the role of JAK2 on OB + IFNγ–induced iNOS and NO production in the ATDC5 chondrogenic cell line.



Fetal bovine serum (FBS), tissue culture media, media supplements, recombinant human and mouse OB, recombinant human and mouse IFNγ, the JAK2-specific inhibitor Tyrphostin AG 490, L-NG-nitroarginine methyl ester (L-NAME), D-NG-nitroarginine methyl ester, aminoguanidine, and dexamethasone were purchased from Sigma (St. Louis, MO).

Cell culture.

ATDC5 cells were a kind gift from Dr. Agamemnon E. Grigoriadis (King's College, London Guy's Hospital, London, UK). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM)/Ham's F-12 medium supplemented with 5% FBS, 10 μg/ml recombinant human insulin, 10 μg/ml human transferring, 3 × 10−8M sodium selenite, and antibiotics (50 units/ml penicillin and 50 μg/ml streptomycin).

Human articular chondrocytes were a kind gift from Dr. J. Couceiro (University Institute of Orthopedics, University of Santiago de Compostela, Santiago de Compostela, Spain). They were obtained from patients who underwent autologous chondrocyte transplantation. Biopsy specimens were obtained from patients, ages 18–45 years, who had local cartilage damage in knee joints. Subsequently, biopsy specimens were enzymatically digested and cells were serially passaged to obtain a sufficient number of cells. Cells were used between the second and third passages. All patients gave informed consent for participation in the study.

Cell treatments and nitrite assay.

Cells (100,000) were cultured to 85–90% confluence in 24-well plates in complete medium. After 12 hours of starvation in serum-free DMEM/Ham's F-12 medium, cells were stimulated for 48 hours with OB (100, 200, and 400 nM) and IFNγ (1 ng/ml) alone or in combination. Forty-eight hours was the time selected for optimal observation according to our time-course experiments (data not shown). In our experimental set, signals of iNOS induction were detectable 18–24 hours following cytokine challenge, yet others have reported that NO synthesis occurs after a lag period of 6–8 hours and continues for at least 72 hours (11, 12). To test JAK2 and iNOS inhibitors, drugs were added simultaneously or 1 hour before cytokine stimulation.

Nitrite accumulation was measured in the culture medium by Griess reaction. Briefly, 100 μl of cell culture medium was mixed with 100 μl of Griess reagent (equal volumes of 1% [weight/volume] sulfanilamide in 5% [volume/volume] phosphoric acid and 0.1% [w/v] naphtylethylenediamine HCl), incubated at room temperature for 10 minutes, and then the absorbance at 550 nm was measured in a microplate reader (Titertek Multiscan; Labsystems, Helsinki, Finland). Fresh culture medium was used as blank in all the experiments. The amount of nitrite in the samples (in μM) was calculated from a sodium nitrite standard curve freshly prepared in culture medium.

RNA isolation and reverse transcriptase–polymerase chain reaction (RT-PCR).

RNA was isolated from cell culture by TRIzol LS, according to the manufacturer's instructions (Gibco BRL Life Technologies, Grand Island, NY). Total RNA (2–5 μg) was used to perform RT-PCR. Complementary DNA (cDNA) was synthesized using 200 units of Moloney murine leukemia virus (MMLV) reverse transcriptase (Gibco BRL) and 6 μl of dNTP mix (10 mM of each dNTP), 6 μl of first-strand buffer (250 mM Tris HCl, pH 8.3, 375 mM KCl, 15 mM MgCl2; Gibco BRL), 1.5 μl of 50 mM MgCl2, 0.17 μl of random hexamer solution (3 μg/μl; Gibco BRL), and 0.25 μl of RNAseOut (recombinant ribonuclease inhibitor, 40 units/μl; Gibco BRL), in a total volume of 30 μl. Reaction mixtures were incubated at 37°C for 50 minutes and at 42°C for 15 minutes. The RT reaction was terminated by heating at 95°C for 5 minutes, and the mixture was subsequently quick-chilled on ice.

Three microliters of the RT reaction mixture was used for PCR amplification. The amplification conditions were as follows: 5 μl of PCR buffer (200 mM Tris HCl, pH 8.4, and 500 mM KCl; Gibco BRL), 1.5 μl of 50 mM MgCl2, 4 μl of dNTP mix, 150 ng of mouse iNOS upstream primer 5′-CTCACTGGGACAGCACAGAA-3′, 150 ng of mouse iNOS downstream primer 5′-TGGTCAAACTCTTGGGGTTC-3′ (GenBank accession no. U43428), or mouse OB receptor long-form upstream primer 5′-TCTTCTGGAGCCTGAACCCATTTC-3′ and mouse OB receptor long-form downstream primer 5′-TTCTCACCAGAGGTCCCTAAACT-3′ (GenBank accession no. U58861), and 1.25 units of Taq DNA polymerase (Gibco BRL). The amplification conditions for mouse iNOS were as follows: denaturation at 95°C for 30 seconds, annealing at 60°C for 40 seconds, and extension at 72°C for 1 minute, while for mouse leptin receptor they were denaturation at 98°C for 30 seconds, annealing at 57°C for 30 seconds, and extension at 72°C for 1 minute. The 33-cycle amplification was completed with an additional step at 72°C for 10 minutes. The amplification was performed in an automatic thermal cycler (Mastercycler Gradient; Eppendorf, Madison, WI). To check the quality of messenger RNA (mRNA) in each sample, mouse OB-R and iNOS cDNA were amplified, together with mouse GAPDH. The PCR reaction generated a single 198-bp product for mouse iNOS, a single 647-bp product for the mouse OB-RL gene, and a single 376-bp amplicon for mouse GAPDH.

Negative controls consisted of omitting the RT reaction mixture for each sample and amplifying samples of RT reaction mixture without MMLV. To exclude a competitive amplification of genomic DNA, RT-PCR was performed on mouse genomic DNA. The identity of the amplimer for iNOS and OB-RL was confirmed by performing RT-PCR together with positive controls. PCR products were separated on 1.5% agarose gel, stained with ethidium bromide, examined with ultraviolet light, and visualized with a Typhoon 9410 documentation system (Amersham Pharmacia Biotech, Little Chalfont, UK).

Statistical analysis.

All results are expressed as the mean ± SEM of at least 3 independent experiments, each with at least 3 independent observations. Statistical analysis was performed by analysis of variance followed by the Student-Newman-Keuls test or Bonferroni adjustment. P values less than 0.05 were considered significant.


As shown in Figure 1, a single 647-bp product for a mouse OB-RL gene was obtained (lanes 1 and 2) by amplifying chondrocyte mRNA with specific primers for OB-RL. The quality of RNA was ensured by amplifying the mouse GAPDH gene (lanes 3 and 4), and positive control for OB-R was obtained by amplifying mRNA from the J774.1 mouse macrophage cell line.

Figure 1.

Reverse transcriptase–polymerase chain reaction (PCR) analysis of the expression of the functional long isoform of the leptin receptor (OB-RL) in cultured ATDC5 cells. PCR products were contained in lanes 1 and 2; lanes 3 and 4 reflect runs with the GAPDH internal control to ensure quality of RNA. Lane 5 is a positive control for OB-RL amplified from mouse J774-A1 macrophage mRNA. MW = molecular weight marker.

In supernatants from human articular chondrocytes stimulated for 48 hours, neither OB nor IFNγ alone was able to induce nitrite accumulation (Figure 2A). In contrast, costimulation with OB and IFNγ significantly increased nitrite accumulation in a dose-dependent manner. This effect was also observed in the mouse chondrogenic ATDC5 cell line (Figure 2B). This result was further reproduced on fully differentiated ATDC5 cells after 21 differentiation days (data not shown).

Figure 2.

A, Nitrite production in stimulated and unstimulated human primary chondrocytes. Cells were starved 8 hours before stimulation in a serum-free medium. Stimulation (48 hours) was carried out under serum-free conditions. B, Nitrite production in stimulated and unstimulated ATDC5 mouse chondrogenic cell line. Cells were starved 8 hours before stimulation in serum-free medium. Stimulation (48 hours) was carried out under serum-free conditions. C, Effect of inducible nitric oxide synthase inhibitors on nitrite production in ATDC5 cells. Stimulation (48 hours) was carried out under serum-free conditions. L-NG-nitroarginine methyl ester (L-NAME) and aminoguanidine were added simultaneously with cytokines. Values are the mean and SEM. ∗ = P < 0.05; ∗∗∗ = P < 0.001, versus unstimulated control, interferon (IFN) alone, and leptin (OB) alone. ### = P < 0.001 versus OB (400 nM) + IFNγ (1 ng/ml).

To confirm whether NO formation was produced via iNOS, ATDC5 cells were incubated for 48 hours in the presence of the iNOS inhibitor aminoguanidine (1 mM), added simultaneously with the cytokines. As shown in Figure 2C, aminoguanidine completely inhibited nitrite accumulation in the culture supernatant induced by costimulation. Similar results were obtained in the presence of L-NAME, another iNOS inhibitor.

The role of JAK2 in nitrite production was investigated in ATDC5 cells by using Tyrphostin AG 490. Tyrphostin (10 μM), administered 1 hour before costimulation, prevented the nitrite accumulation induced by OB + IFNγ (Figure 3, top). This was confirmed by RT-PCR analysis, showing the presence of iNOS mRNA in OB + IFNγ–costimulated cells and the absence in costimulated cells in the presence of Tyrphostin (Figure 3, bottom). Cell vitality, evaluated at the end of the experiment using trypan blue staining, was >95%. In addition, iNOS induced by costimulation was inhibited by dexamethasone, a classic inhibitor of de novo iNOS synthesis that does not affect the activity of the preexpressed or constitutive enzyme.

Figure 3.

Top, Effect of janus-activated kinase 2 inhibitor Tyrphostin AG 490 on nitrite production in stimulated and unstimulated ATDC5 mouse chondrogenic cell line. Cells were starved 8 hours before stimulation in serum-free medium. Tyrphostin AG 490 was added 1 hour before cytokine challenge. Values are the mean and SEM. ∗∗∗ = P < 0.001 versus unstimulated control, interferon (IFN) alone, and leptin (OB) alone. ### = P < 0.001 versus OB (400 nM) + IFNγ (1 ng/ml). Bottom, Reverse transcriptase–polymerase chain reaction analysis of the expression of the inducible nitric oxide synthase (iNOS) in cultured ATDC5 cells. Lane 1, Unstimulated control; lane 2, IFNγ; lane 3, OB; lane 4, OB + IFNγ; lane 5, OB + IFNγ + Tyrphostin AG 490; lane 6, OB + IFNγ + dexamethasone; lanes 7–12, GAPDH internal controls.


In the present work, we have investigated the effects of OB and IFNγ on iNOS induction and NO production by chondrocytes. To the best of our knowledge, this is the first study to show that OB, acting synergistically with IFNγ, can induce iNOS in chondrocytes, leading to the production of NO by a mechanism involving the activation of JAK2. There is sufficient evidence to show that NO influences cartilage degeneration by altering collagen and proteoglycan synthesis and by leading to chondrocyte apoptosis. Several previous studies have shown high concentrations of NO in arthritic synovium, articular cartilage, and synovial fluid, derived in part from chondrocytes. In vitro, articular OA cartilage is able to produce large amounts of NO in the absence of cytokines (13), and this production is synergistically or additively enhanced by cytokines (5). In our work, NO production required costimulation with OB and IFNγ. The source of IFNγ in the joint under pathologic conditions has been investigated. In RA and OA, IFNγ is produced by synovial and cartilage cells. It is also produced by a high percentage of CD4+ (Th1) lymphocytes in the synovial fluid in RA patients and, in synergy with interleukin-1 (IL-1), stimulates the production of IL-6, NO, and prostaglandin E2 (14). Under our experimental conditions, the effect, if any, of IFN alone is negligible.

The proinflammatory role of OB has been previously documented. In antigen-induced arthritis, OB regulates both humoral and cell-mediated immune responses (15). Furthermore, OB increases during acute inflammation (8, 16), and OB-deficient mice exhibited a large decrease in antimicrobial activity of both macrophages and stroma vascular fraction cells of adipose tissue. Finally, it is well known that markers of inflammation, such as C-reactive protein, IL-6, tumor necrosis factor α, and OB, are elevated in obese individuals. In this report, we present evidence showing that OB up-regulates iNOS, as do different cytokines, suggesting that the hormone could directly participate in the damage that occurs in joints. The source of OB has not been addressed yet. De Bari et al (17) have recently demonstrated that under appropriate culture conditions, synovial membrane–derived cells are induced to differentiate into chondrocytes, osteocytes, and adipocytes. Notwithstanding, whether adipose tissue in synovium and/or periarticular tissues could produce OB under normal or abnormal circumstances has yet to be demonstrated.

The intracellular signaling events in iNOS induction are not completely defined, particularly those evoked by new members of the cytokine superfamily, such as OB. Since both OB-R and IFNγ-R transduce signals via JAK2 activation, the observation that Tyrphostin AG 490 restrained iNOS induction indicates that tyrosine kinase activity is involved in the early events leading to enzyme expression. Activation of the JAK2 signaling pathways represents a crucial step in iNOS induction in several cellular targets. It is likely that JAK2 activation serves to phosphorylate the signal transducers and activators of transcription (STATs) when cytokine receptors lack intrinsic kinase activity, as in OB-R and IFNγ-R. Activated STATs form dimers, translocate to the nucleus, and bind to response elements to induce transcription. Therefore, it is possible that STATs are not the only transcription factors involved in iNOS induction elicited by OB + IFNγ-R costimuli; other factors, such as nuclear factor κB, among others, could also be involved. Furthermore, synthesis/induction of other factors that then induce iNOS could also occur.

In conclusion, there are potentially significant physiologic and physiopathologic aspects of iNOS induction triggered by OB and IFNγ in chondrocytes. Therefore, elucidation of the molecular mechanisms by which this induction is triggered is of major importance, and may have implications for the design of treatment strategies and for the accurate definition of the events that regulate the inflammatory response. One significant result gleaned from this investigation is that recently discovered hormones, such as OB, may function as cytokines and play important physiologic roles in promoting inflammatory responses.


The authors are very grateful to Dr. M. Silva (Institute of Orthopedics, University of Santiago de Compostela, Santiago de Compostela, Spain) for helpful advice and for excellent assistance.