Interleukin-1β up-regulation of Smad7 via NF-κB activation in human chondrocytes




We have previously shown that interleukin-1β (IL-1β) impairs transforming growth factor β (TGFβ) signaling through TGFβ receptor type II (TGFβRII) down-regulation and Smad7 up-regulation. This mechanism could account for the reduced responsiveness of osteoarthritic chondrocytes to TGFβ and the cartilage breakdown linked to this disease. The aim of this study was to investigate the molecular mechanism underlying the IL-1β–induced stimulation of Smad7 in human articular chondrocytes.


Human articular chondrocytes were treated with IL-1β in the presence of TGFβ1, pyrrolidine dithiocarbamate (a repressor of the NF-κB pathway), or cycloheximide. Then, steady-state messenger RNA and protein levels were estimated by real-time reverse transcription–polymerase chain reaction and immunocytology. In addition, transient transfections of p65 expression vector or p65-targeted short hairpin RNA were performed to define the effect of NF-κB on Smad7 expression.


TGFβRII overexpression restored the TGFβ response of human articular chondrocytes. However, this effect was transient, implying that a secondary mechanism was responsible for the alteration of the TGFβ response with long-term exposure to IL-1β. Moreover, IL-1β caused a late induction of the inhibitory Smad7. This effect was direct, since it did not require de novo synthesis. In addition, we established, by experiments with gain/loss of function, that the up-regulation of Smad7 by IL-1β is mediated through the NF-κB pathway, especially the p65 subunit.


These findings clarify the regulatory process of IL-1β on Smad7 expression. Understanding the molecular basis of IL-1β induction of Smad7 and the reduction of chondrocyte responsiveness to TGFβ provides new insights into the molecular mechanisms of osteoarthritis and may facilitate the identification of novel approaches for its treatment.

The maintenance of homeostasis of articular cartilage is crucial for the integrity of its structure and function. The balance of matrix turnover is regulated by catabolic and anabolic activity of the chondrocytes. In osteoarthritis (OA), an imbalance occurs in favor of matrix catabolism, which leads to the slow, but irreversible, degradation of cartilage.

Numerous studies have revealed that interleukin-1β (IL-1β) is the main mediator of cartilage destruction during the OA process (1). Indeed, this proinflammatory cytokine, whose levels are highly increased in OA synovial fluid, decreases extracellular matrix production, particularly that by cartilage-specific markers, such as type II collagen and aggrecan. In addition, it induces their degradation by stimulating both the synthesis and activation of metalloproteinase and by down-regulating the inhibitors of these enzymes. IL-1β modulates target gene transcription through different pathways. It activates, for example, the MAP kinases or NF-κB. Indeed, IL-1β induces the phosphorylation of IκB, triggering its degradation and the release of NF-κB factors, which translocate into the nucleus, where they regulate the transcription of target genes. NF-κB exists in the cytoplasm as a homodimer or heterodimer of variable subunits, the major form of which is composed of p50 and p65 (2).

In contrast, the anabolic activity of chondrocytes is classically thought to be maintained by factors such as insulin-like growth factor 1 or transforming growth factor β (TGFβ) (3). TGFβ initiates signaling through the ligand-dependent activation of transmembrane serine/threonine kinases, consisting of TGFβ receptor type I (TGFβRI) and TGFβRII. Then, the action of TGFβ is mediated intracellularly by receptor-associated Smad 2 (R-Smad2) and R-Smad3. Following phosphorylation by TGFβRI, Smad2 and Smad3 associate with Smad4. This complex migrates to the nucleus and participates in the transcriptional activation of target genes. In contrast, Smad7, an inhibitory Smad (I-Smad), acts as an inhibitor due to its interaction with ligand-activated TGFβRI, and it interferes with the phosphorylation of R-Smads, preventing nuclear translocation of the activated Smad complexes (4).

Our previous study indicated that IL-1β prevents the response of chondrocytes to TGFβ by down-regulating TGFβRII and up-regulating Smad7 (5). We have also previously demonstrated that IL-1β reduces TGFβRII expression through modulation of the Sp-1:Sp-3 ratio by NF-κB (6). In the present study, we provide a mechanism of Smad7 gene regulation. We found that IL-1β provokes the late induction of Smad7 by a mechanism that is independent of protein neosynthesis and involves activation of the NF-κB/p65 pathway.



Reagents were provided by Invitrogen (Cergy-Pontoise, France) unless indicated otherwise. Inhibitors (cycloheximide and pyrrolidine dithiocarbamate [PDTC], 10 μM and 100 nM, respectively), human recombinant IL-1β (Sigma-Aldrich, St. Quentin Fallavier, France), and TGFβ1 (R&D Systems, Lille, France) were resuspended in water, phosphate buffered saline (PBS)–bovine serum albumin (BSA), or PBS–HCl. Oligonucleotides were supplied by Eurogentec (Angers, France).

Cell culture and treatments.

Human articular chondrocytes were prepared from the femoral heads obtained from 15 patients undergoing hip replacement surgery (ages 57–80 years; median 77 years), as previously described (5). Cells were not pooled from one patient to another. Chondrocytes were seeded at 4 × 104 cells/cm2 in 6-well plates and cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum (FCS), 100 IU/ml of penicillin, 100 μg/ml of streptomycin, and 0.25 μg/ml Fungizone, at a temperature of 37°C in an atmosphere of 5% CO2. Cells were cultured for a maximum duration of 12 days and were used as primary cultures. Chondrocyte cultures were treated with 1 ng/ml of human recombinant IL-1β in DMEM supplemented with 2% FCS.

Plasmids and transfection experiments.

Expression vectors (pSG5-p65, short hairpin RNA [shRNA]-p65, and shRNA-control) were obtained and used as previously described (5). Chondrocytes were transiently transfected with appropriate expression vector according to the calcium phosphate precipitation method. After overnight transfection, the culture medium was replaced with DMEM containing 2% FCS, the chondrocyte cultures were incubated with IL-1β (1 ng/ml), and appropriate treatments were performed.

RNA extraction and real-time reverse transcription–polymerase chain reaction.

Total RNA from primary human articular chondrocyte cultures was extracted using TRIzol. Following extraction, 1 μg of DNase I–treated RNA was reverse transcribed into complementary DNA (cDNA) in the presence of random hexamers (Applied Biosystems, Courtaboeuf, France) and Moloney murine leukemia virus reverse transcriptase. The reaction was performed at 37°C for 1 hour followed by a further 10-minute step at 95°C. Amplification of the generated cDNA was performed by real-time polymerase chain reaction in an Applied Biosystems SDS7000 apparatus. The relative level of messenger RNA (mRNA) was calculated using the 2math image method.

Immunofluorescence microscopy.

Cells were fixed in 4% paraformaldehyde for 15 minutes at room temperature and preincubated with PBS/0.2% BSA/0.05% saponin for 1 hour. Subsequently, the slides were incubated overnight at 4°C with rabbit polyclonal antibody against Smad7, then with a fluorescein isothiocyanate–conjugated anti-rabbit antibody for 1 hour. Thereafter, cells were incubated for 5 minutes with propidium iodide in order to stain the nuclei. Human articular chondrocytes were examined with a fluorescence microscope.

Statistical analysis.

All experiments were repeated with different donors at least 3 times, and the results were similar. Representative experiments are illustrated. Data are presented as the mean ± SD of triplicate determinations. Statistical significance was determined by Student's t-test. Differences were considered to be statistically significant at P < 0.05, P < 0.01, and P < 0.001.


Ectopic expression of TGFβRII and restoration of TGFβ-induced plasminogen activator inhibitor 1 (PAI-1) expression, but only for the brief incubation with IL-1β.

In a previous study (5), we demonstrated that IL-1β down-regulated the expression of TGFβRII. Therefore, when cells were preincubated with IL-1β, they were no longer able to respond to TGFβ, as judged by PAI-1 mRNA induction. Forced expression of TGFβRII in these cells restored their sensitivity to the factor. Nevertheless, when the duration of preincubation was longer (i.e., 48 hours instead of 1 hour), there was only partial restoration (Figure 1).

Figure 1.

Abrogation of the interleukin-1β (IL-1β)–induced loss of transforming growth factor β (TGFβ) responsiveness in human articular chondrocytes by overexpression of TGFβ receptor type II (TGFβRII). Chondrocytes were transfected with TGFβRII expression vector pCMV5-TGFβRII or with the insertless plasmid (5 μg). After 15 hours, the medium was replaced with Dulbecco's modified Eagle's medium plus 2% fetal calf serum for 24 hours. Cells were then incubated with IL-1β (1 ng/ml) for 1 hour (A) or for 48 hours (B) prior to the addition of TGFβ1 (2 ng/ml) for 4 hours. Total RNA was isolated and analyzed by real-time reverse transcription–polymerase chain reaction, using plasminogen activator inhibitor 1 (PAI-1) and GAPDH primers. Results shown represent the level of induction of PAI-1 mRNA by treatment with TGFβ over that of controls. Values are the mean and SD results from triplicate determinations. ∗∗∗ = P < 0.001; ∗∗ = P < 0.01 versus controls.

Late expression of Smad7 induced by IL-1β and independence of de novo protein synthesis.

Given that we have previously shown that IL-1β up-regulates Smad7 protein after 12 hours of incubation (5), we hypothesized that IL-1β decreases the responsiveness of chondrocytes to TGFβ by down-regulating TGFβRII and subsequently up-regulating Smad7. We investigated the effect of IL-1β on Smad7 mRNA expression in human articular chondrocytes in a time-dependent manner (Figure 2A). IL-1β was found to significantly stimulate Smad7 expression. However, this up-regulation was seen only from 12 hours of incubation up to 96 hours. This late effect was consistent with the results shown in Figure 1, which implied that a secondary mechanism was responsible for the alteration of the TGFβ response with long-term exposure to IL-1β.

Figure 2.

Up-regulation of Smad7 expression in human articular chondrocytes after 12 hours of incubation with interleukin-1β (IL-1β). A, Levels of Smad7 mRNA in chondrocytes incubated with IL-1β for different time periods. Chondrocytes were cultured for 5–6 days in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS) and then incubated in DMEM plus 2% FCS for 24 hours. IL-1β (1 ng/ml) or vehicle (control) was added for the indicated times, and Smad7 expression was analyzed by real-time reverse transcription–polymerase chain reaction (RT-PCR). The modulation of Smad7 mRNA expression was normalized against the GAPDH signal. B, Levels of Smad7 mRNA in chondrocytes incubated in the presence or absence of IL-1β or cycloheximide (CHX). Subconfluent chondrocytes were incubated for 24 hours with DMEM plus 2% FCS and then treated with cycloheximide for 30 minutes before the addition of IL-1β (1 ng/ml). Cultures were incubated for an additional 12-hour period, and Smad7 expression was analyzed by real-time RT-PCR. Values in A and B are the mean and SD results from triplicate determinations. ∗∗ = P < 0.01; ∗∗∗ = P < 0.001, versus controls. C, Cellular localization of Smad7 protein in chondrocytes, as determined by immunofluorescence staining (see Materials and Methods for details). Cells were treated with IL-1β for the indicated times (original magnification × 64).

We reproduced the experiment in the presence of cycloheximide. The translation inhibitor did not counteract the stimulatory effect of IL-1β on Smad7 mRNA levels (Figure 2B), which indicates that this up-regulation is independent of de novo protein synthesis.

Immunostaining of human articular chondrocytes revealed that the Smad7 protein staining become stronger upon IL-1β treatment for 24 hours, showing the concordance of the effect of both the mRNA and protein levels (Figure 2C).

Involvement of the NF-κB/p65 pathway in IL-1β regulation of Smad7.

Since IL-1β is known to exert numerous effects on cartilage metabolism through the NF-κB pathway, we speculated about the putative involvement of NF-κB in the induction of Smad7 by IL-1β. First, we attempted to specifically inhibit this pathway by using the chemical inhibitor PDTC. This inhibitor was able to block the induction of Smad7 expression by IL-1β (Figure 3A), which suggests that IL-1β up-regulates Smad7 through activation of NF-κB.

Figure 3.

Involvement of the NF-κB/p65 pathway in the up-regulation of Smad7 induced by interleukin-1β (IL-1β) in human articular chondrocytes. A, Levels of Smad7 mRNA in chondrocytes incubated in the presence or absence of IL-1β or pyrrolidine dithiocarbamate (PDTC). Subconfluent chondrocytes were treated with IL-1β (1 ng/ml) for 48 hours in the presence or absence of the NF-κB inhibitor PDTC. Smad7 expression was then analyzed by real-time reverse transcription–polymerase chain reaction (RT-PCR). B, Levels of Smad7 mRNA in chondrocytes incubated in the presence or absence of IL-1β or short hairpin RNA (shRNA)–p65 vector. Chondrocytes were transfected overnight with sh-control or shRNA-p65 vectors and placed in Dulbecco's modified Eagle's medium (DMEM) plus 2% fetal calf serum (FCS) in the presence of IL-1β (1 ng/ml) for 48 hours. Smad7 expression was then analyzed by real-time RT-PCR. C, Levels of Smad7 mRNA in chondrocytes treated with different concentrations of pSG5-p65 vector. Chondrocytes were transfected overnight with insertless pSG5 vector and various amounts of pSG5-p65 (up to 10 μg) and placed in DMEM plus 2% FCS for 24 hours. Smad7 expression was then analyzed by real-time RT-PCR. Values are the mean and SD results from triplicate determinations. ∗∗ = P < 0.01; ∗∗∗ = P < 0.001, versus controls. NS = not significant.

Given that p65 is a classic activator subunit of NF-κB and that it regulates TGFβRII expression upon IL-1β treatment, we examined its involvement in Smad7 expression. To address this question, we used different approaches to the in vitro gain/loss of p65 function. First, in order to target p65 mRNA silencing, transient transfection of human articular chondrocytes with shRNA-p65, which induced an important decrease in p65 expression (data not shown), was sufficient to abrogate IL-1β induction of Smad7 expression (Figure 3B). Moreover, overexpression of the p65 subunit in human articular chondrocytes resulted in a dose-dependent increase in Smad7 expression, as compared with control cells transfected with the insertless plasmid (Figure 3C).


Antagonistic transmodulation between IL-1β and TGFβ has been well documented, particularly in the context of tissue homeostasis. For example, TGFβ antagonizes numerous effects of IL-1β and decreases IL-1 receptor expression (7). In contrast, IL-1β is able to interfere with the TGFβ system (8) and particularly with TGFβRII (5). Thus, we recently showed that IL-1β abrogates the TGFβ response in articular cartilage through down-regulation of TGFβRII by a mechanism involving the NF-κB pathway and transcription factors Sp-1/Sp-3 (6). However, we report herein an additional mechanism by which IL-1β antagonizes the response of chondrocytes to TGFβ. We propose that the control of Smad7 expression via NF-κB might provide a subsequent mechanism whereby IL-1β counteracts TGFβ signaling.

In the present study, we found that ectopic expression of TGFβRII was able to totally abrogate the loss of the TGFβ response induced by IL-1β, but only for short incubation times, and that this cytokine stimulated Smad7 expression after longer exposure times (>12 hours). These data suggest that IL-1β counteracts TGFβ signaling through 2 complementary systems: first, by down-regulating TGFβRII and second, by stimulating inhibitory Smad7 expression. Increasing the expression of Smad7 probably promotes its association with TGFβRI, reducing receptor accessibility for R-Smads, Smad2/3 phosphorylation, and nuclear translocation, and therefore target gene regulation. It has been observed, for example, that Smad7 overexpression totally prevented TGFβ-induced proteoglycan synthesis in chondrocytes on both the mRNA and protein levels and that it completely antagonized the effects of TGFβ on proliferation (9).

Thus, regulation of I-Smads, associated with a decrease in TGFβRII, interferes with the TGFβ process and might therefore be responsible for the lack of anabolic synthetic activity of chondrocytes observed in the upper zones of OA cartilage. However, no correlation between the localization of Smad7 in OA or healthy cartilage and the expression of the major anabolic genes of articular chondrocytes has yet been reported (10).

We also demonstrated that the effects of IL-1β on Smad7 expression in human articular chondrocytes were mediated through the NF-κB pathway. This study confirms the cross-talk between TGFβ/Smad and NF-κB, which has recently been proposed by Ishida et al (11) with regard to skin wound sites. Their observations indicate that the absence of IL-1 receptor antagonist (IL-1Ra) induces an aberrant NF-κB activation and an increase in Smad7 proteins in IL-1Ra–knockout mice as compared with wild-type mice. Conversely, Smad7 has been reported to regulate the NF-κB pathway: it is able to block the TGFβ-induced phosphorylation of IκBα, resulting in a decrease in NF-κB DNA binding (12). However, inhibition of Smad7 increased IκBα expression and reduced accumulation of the NF-κB p65 subunit in the nucleus of lamina propria mononuclear cells from patients with gut inflammation, indicating that Smad7 could also act as an NF-κB activator in some conditions (13).

Finally, we showed that, in contrast to TGFβRII regulation, Smad7 stimulation by IL-1β did not require protein neosynthesis. This conclusion might appear contradictory to the data obtained when the p65 subunit was silenced. However, upon IL-1β stimulation, NF-κB accumulates from 2 pathways: activation of preexisting protein (i.e., degradation of IκB) and neosynthesis by transcription/translation. Cycloheximide inhibits only the second pathway, and NF-κB is still produced through inactivation of IκB, which explains why IL-1β is still effective even without de novo translation.

The hypothesis that IL-1β activates the NF-κB/p65 pathway, which could directly enhance Smad7 transcription, is strengthened by recent studies showing that NF-κB/RelA inhibits TGFβ/Smad signaling in 3T3 cells by stimulating Smad7 expression (14), but the hypothesis is inconsistent with the findings of the study by Nagarajan et al (15), who reported that the mouse Smad7 promoter is down-regulated by NF-κB. This latter observation is probably due to the divergence of sequences from one species to another.

In conclusion, we describe a new mechanism of IL-1β–induced suppression of TGFβ signaling by the NF-κB/p65–dependent pathway in human chondrocytes. These data support the concept that IL-1β, via NF-κB, interferes with TGFβ signaling through indirect down-regulation of TGFβRII and subsequent direct up-regulation of Smad7.


Dr. Boumédiene 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 design. Baugé, Leclercq, Boumédiene.

Acquisition of data. Baugé, Attia.

Analysis and interpretation of data. Baugé, Galera, Boumédiene.

Manuscript preparation. Baugé, Pujol, Boumédiene.

Statistical analysis. Baugé, Attia.