As a defect in leptin action on body weight homeostasis has been reported for obese individuals despite high serum leptin concentrations, the current study was undertaken to determine whether chondrocytes from obese OA patients that are exposed to elevated leptin levels, exhibit an altered response to leptin.
Before examining the effect of BMI on chondrocyte responsiveness to leptin, we compared the gene expression pattern between obese and non obese OA patients. The RT-PCR analysis showed major changes between both groups with an elevated mRNA level of growth factors, aggrecan and TIMP-2 in chondrocytes from obese patients suggesting that the chondrocyte metabolic activity is increased with obesity. Such overexpression of TGFβ, IGF-1 and TIMP-2 has been previously reported in subcutaneous adipose tissue from obese individuals [17
] indicating that chondrocytes may also exhibit an obesity-related gene expression pattern. These findings suggest also that a pathological status may result in a similar pattern of expression of some factors in both chondrocytes and adipocytes. As obesity and OA are both associated with inflammation, the up-regulation of growth factors and MMP inhibitors may be an attempt at an adaptive response to an inflammatory environment. However, we failed to detect any difference in the expression of proinflammatory mediators between non-obese and obese OA patients, suggesting that the chronic low grade of inflammation associated with obesity rather than the OA-related joint inflammation is involved in the overexpression of TGFβ, IGF-1 and TIMP-2 in both chondrocytes and adipocytes.
The downstream targets of leptin identified in the current study are in agreement with those previously published because IGF-1, type 2 collagen and MMP-13 were overexpressed upon leptin stimulation [9
]. However, conflicting results were found for the effect of leptin on the expression of inflammatory mediators. Although Vuolteenaho and colleagues showed that leptin enhanced the expression of iNOS in human OA cartilage [13
], iNOS mRNA level remained unchanged in leptin-treated cells. The use of cartilage explants instead of isolated chondrocytes may explain the discrepancies because other authors were unable to find any effect of leptin on the expression of iNOS in cultured human primary chondrocytes [9
]. Similarly, we failed to detect any change in the expression of IL-1 whereas Simopoulou and colleagues reported a stimulatory effect of leptin on the production of the cytokine [9
]. In fact, human chondrocytes required long-term cultures to produce IL-1 suggesting an indirect leptin-mediated pathway.
One of the most relevant findings arising from this study is the BMI-dependent effect of leptin on the expression of the genes encoding TIMP-2 and MMP-13 in chondrocytes. Our results provided new insights on the role of leptin in matrix remodeling by regulating the balance between MMPs and TIMPs, which are the physiologic protein inhibitors of the degradative enzymes. Although TIMP-1 is highly expressed in cartilage and reduced in OA, we failed to detect any effect of leptin on its expression (data not shown). By contrast, mRNA levels for TIMP-2 was markedly increased in leptin-treated chondrocytes compared with control cells. However, this up-regulation of TIMP-2 was found in normal or overweight patients only, and decreased when the BMI of the patients increased. Besides, MMP-13 was overexpressed in leptin-stimulated chondrocytes obtained from obese OA patients, and this was shown from the lowest concentration of leptin. These BMI-dependent effects of leptin may change the degenerative process during OA. The leptin-induced TIMP-2 expression may delay cartilage destruction in non-obese OA patients while the adipokine may enhance cartilage damage in obese OA patients. It is worth noting that the BMI-dependent effect of leptin found for genes involved in matrix remodeling was not showed for COX-2, iNOS or IL-1 suggesting that the adipokine did not modulate the production of proinflammatory mediators in cartilage from obese patients.
The other interesting finding of the study is the influence of leptin concentration. Among the three concentrations used in the current study, two doses represented the in vivo
situation given that the synovial fluid levels of leptin in patients with OA range from 1 to 100 ng/ml [8
]. As the chondrocyte sensitivity to exogenous stimulating factor may be reduced when cells are isolated from the extracellular matrix, a higher leptin dosage was also tested [19
]. Most of the anabolic and catabolic genes were insensitive to the lowest dose of leptin regardless of BMI, MMP-13 being the single target of leptin at 20 ng/ml in obese patients. Beside, the relations between cell responsiveness and BMI indicated that chondrocytes collected from patients with low BMI displayed most responsiveness with 100 ng/ml of leptin. This dosage also induced the expression of genes encoding aggrecan, COX-2 and MMP-13 in obese patients. However, we were not able to find any significant correlation between cell responsiveness and BMI with these slight stimulatory effects of leptin at 100 ng/ml. The lack of a marked effect of leptin at 20 ng/ml also explains the lack of any association between chondrocyte response to the adipokine and BMI. The most striking effects of leptin in obese patients, that is the up-regulation of IGF-1, collagen type 2 and MMP-13, were found with the highest concentration of leptin. Chondrocytes from obese patients may already be so strongly exposed to elevated leptin levels within the joint that they are refractory to further stimulation. The attenuation of leptin sensitivity may result either from a downregulation of the leptin receptor or from impairments of the signal transduction process. In agreement with the data obtained by Simopoulou and colleagues [9
], we failed to detect any change in the expression of the receptor neither between normal or overweight and obese patients, nor upon stimulation of OA chondrocytes with 100 or 500 ng/ml of leptin (data not shown). The activation of the Janus kinase/STAT pathway has been investigated to determine whether the leptin-induced signal transduction process was not impaired [20
]. In the current study, the elevated concentration of leptin actually induced the phosphorylation of STAT 1α and -3 in chondrocytes. This STAT activation resulted in the expression of target genes in chondrocytes issued from obese patients but was unable to drive gene expression in cells collected from normal or overweight patients probably because of a negative feedback inhibition of leptin signaling. By contrast, the adipokine at 100 ng/ml mediated its effects through a STAT-independent pathway because this concentration of leptin failed to activate STAT 3. Other signaling pathways, such as ERK and phosphatidylinositol3-kinase, may be involved. Leptin was shown to regulate the differentiation of the ATDC5 chondrogenic cell line through activation of ERK1/2 [21
], and to activate the IRS-1/PI3K/Akt pathway to induce migration of human chondrosarcoma cells [22
]. The STAT-independent signal transduction process was, however, disrupted when chondrocytes from obese OA patients were stimulated with 100 ng/ml of leptin. As STAT 1 has been found to act as a negative transcriptional regulator [23
], leptin-induced STAT 1 phosphorylation in chondrocytes from obese patients may interfere with the transcriptional machinery and may lead to the loss of cell sensitivity to leptin. The responsiveness to the adipokine was then restored in chondrocytes from obese patients after activation of STAT 3 by high adipokine level. The relations found in the current study between BMI and the leptin-induced gene expression, which are negative for 100 ng/ml of leptin and positive for the highest leptin level, further support the finding that leptin may have the ability to differentially regulate the expression of target genes depending on both the BMI and the dose.