Osteoarthritis (OA) has a complex aetiology that includes genetic, hormonal, metabolic, and biomechanical factors. Specific risk factors such as age, sex, race, trauma, obesity, genetic, and oestrogen and nutritional deficiencies have been identified [1
]. Oxytocin (OT) is a pituitary hormone that regulates the function of target organs and modulates a wide range of behaviours, such as social recognition, love, and fear [6
]. OT receptors are present in the peripheral tissues, including adipose, muscle, subchondral bone, testis, ovary, heart, lung, and vascular tissues [10
Plasma OT levels were lower in ovariectomized mice and rats than in control animals, and were found to be lower in postmenopausal women who developed osteoporosis than in their healthy counterparts [17
]. Daily subcutaneous OT injection prevented and alleviated bone loss in ovariectomized mice by enhancing bone microarchitecture and biomechanical strength and by reducing marrow adiposity [17
]. Additionally, OT receptor knockout mice developed osteoporosis [19
]. OT acts on muscle and subchondral bone, which are important structures in OA physiopathology [20
], although the underlying mechanism is not well understood [16
]. The subchondral modifications and osteoblast actions in OA are well known [21
]. Indeed, our group previously showed that OT can inhibit the differentiation of adipocytes and stimulate that of osteoblasts [17
]. Data on the relationship between OT and OA are scarce, and a previous study suggested a protective action of OT on cartilage degradation [16
This present study investigated the role of OT in OA aetiology. Using two human stem cell lines, we aimed to analyze the role of OT on chondrocyte formation in vitro and explore the effect of OT in a rodent model in vivo. Finally, we analyzed the existence of a correlation, if any, between OT and the development of AO in an animal model and a human cohort.
OT acts on muscle and subchondral bone, which are important structures in the physiopathology of OA. In this study, we demonstrated a possible association between OT and OA. Our in vitro results show that OT stimulates chondrogenesis through the OT receptor expressed by chondrocytes ([16
] and data not shown). We previously reported that OT inhibits the differentiation of adipocytes but induces that of osteoblasts. Moreover, subcutaneous OT administration reversed bone loss in ovariectomized mice—an animal model of menopause—by enhancing bone microarchitecture and biomechanical strength, and reducing marrow adiposity [17
The relationship between adipose tissue and OA has been widely studied [25
]. Adipose tissue synthesizes and releases adipokines that modulate bone metabolism by directly or indirectly regulating bone formation and resorption. Adipocytes also express OT receptor; signalling through these receptors induces lipolysis [26
]. Systemic administration of OT has been shown to affect appetite, body weight gain, glucose homeostasis, and lipid metabolism in animal models [26
]. In this present study, we confirmed that OT affects adipose tissue, which was demonstrated by the fact that OT-treated rats gained less body weight than untreated controls.
Oestrogen deficiency plays an important role in the pathophysiology of OA [27
]. OT promotes osteoblastogenesis in both hMADS and hBMS cells [17
], which can explain the involvement of OT in OA. During OA development, activation of catabolic enzymes, including matrix metalloproteinases and aggrecanases, leads to softening of the articular cartilage and low-grade inflammation characterized by the release of the cytokines tumour necrosis factor-α and IL-1. In turn, this phenomenon likely further induces the expression of catabolic enzymes, such as ADAMTS-4 [29
]. We found that OT attenuated the effects of IL-1β, as shown by the reduction in ADAMTS-4
mRNA transcript levels. Thus, OT may modulate chondrogenesis, at least partially, by abrogating the deleterious effects of IL-1β and attenuating inflammation. Moreover, our in vitro results, which showed the positive effect of OT treatment on various genes involved in chondrogenesis, and those of a recent study showing that OT controls chondrocyte matrix degradation through metalloproteinases [16
] strongly support the role of OT in the pathophysiology of OA.
Remodelling of the subchondral bone in OA is mediated by the activities of osteoclasts and osteoblasts [31
]. Although the association between subchondral bone and cartilage in OA is not fully understood, several mechanisms have been proposed that include microdamage repair, increased vascularity stimulated by angiogenic factors, and enhanced bone–cartilage crosstalk through an increased number of subchondral plate pores [32
]. In animal models, osteoblasts of OT knockout mice exhibit reduced mineralization activity and downregulated expression of genes related to osteoblast differentiation [11
]. However, OT stimulates the differentiation of osteoblasts into a mineralizing phenotype by inducing the upregulation of BMP-2, which in turn controls the expression of Schnurri-2 and -3, Osterix, and activating transcription factor-4 [19
OT has dual effects on osteoclasts, as it is known to stimulate osteoclast formation both directly through the activation of nuclear factor-κB and mitogen-activated protein kinase signalling as well as indirectly through the upregulation of receptor activator of nuclear factor-κΒ ligand [11
]. Furthermore, OT inhibits bone resorption by mature osteoclasts by triggering cytosolic Ca2+
release and nitric oxide (NO) synthesis [19
]. It is therefore possible that in OA, a disease with complex physiopathology that is still not completely elucidated, OT has a beneficial action [10
] likely via effects on oestrogens, but also has harmful activities that may include stimulating NO synthesis or upregulating BMP-2. Although BMPs are involved in all phases of chondrogenesis, which affect chondrocyte differentiation and cartilage anabolism [12
], recent studies have shown that BMPs can also have harmful effects on articular cartilage [15
It is known that there are different osteoarthritic phenotypes and that the risk factors will differ according to the targeted joint, which might explain the lack of effect found in our animal model that may have not been the most suitable. Perhaps a model of metabolic osteoarthritis might have been more appropriate. Alternatively, the duration of rat treatments (28 days) might not have been sufficient, as 8 weeks can be necessary to normalize body weight and osteopenia in mice [17
]. In our human model, the absence of differences in OT rate according to disease severity may reflect the fact that all of the subjects came from the ADEM cohort, which was focused on severe digital OA. However, the use of subjects who all exhibited advanced osteoarthritis lesions makes it impossible to look for a correlation between the OT rate and severity of osteoarthritis, even though it can show a potential action of OT.
In conclusion, we found that OT stimulates chondrogenesis and that women with digital OA have low circulating levels of this hormone. These findings provide a basis for analysing the mode of action of OT in this disease and may aid in the development of more efficient therapies.