Here, we show for the first time, the contribution of relaxin and also of β-estradiol to the degradation of TMJ fibrocartilage and (to a lesser extent) of knee articular cartilage in vivo, the lack of potentiation of these responses to relaxin by β-estradiol, and progesterone's prevention of matrix loss mediated by β-estradiol and/or relaxin. These findings suggest that, by modulating the remodeling of the ECM of cartilage, these hormones may play an important regulatory role in the normal and pathologic metabolism of cartilaginous tissues. In the latter scenario, it is conceivable that individuals with abnormal absolute or relative levels of one or more of these hormones or their receptors might incur progressive loss of matrix macromolecules, leading to joint disorders characterized by the degeneration of specific cartilages or fibrocartilages . Such potential hormone-mediated changes in the composition of the ECM can significantly impact the ability of joints to sustain and distribute mechanical loading and can also substantially affect the normal function and survival of cells in tissues [28–30].
There is a striking female preponderance for many types of joint diseases in general and for TMJ diseases in particular. TMJ disorder is an umbrella term describing a group of clinical signs and symptoms involving the masticatory musculature, the TMJ, and associated structures such as the fibrocartilaginous disc . TMJ disorders are distinguished from similar diseases of other joints by one specific epidemiologic difference, namely that, unlike many similar diseases of other joints that afflict postmenopausal women, TMJ diseases are observed primarily in women of reproductive age. On the basis of our findings, it is plausible that specific reproductive hormones target this highly fibrocartilaginous joint for degradative activity by upregulating particular MMPs that contribute to the loss of collagen and GAGs [5, 21]. In keeping with this postulate, our studies show that the responses of TMJ disc to relaxin, β-estradiol, and progesterone are more similar to those observed in the pubic symphysis than to those seen in the knee meniscus or articular cartilage, suggesting that the observed effects of these hormones are specific to cell type and tissue type. Additionally, the link between the relaxin- and β-estradiol-mediated induction of MMPs and the loss of collagen and GAGs in the TMJ disc by these hormones has been demonstrated previously in our tissue explant studies. However, the in vivo induction of MMPs by these hormones and the association between MMP induction and matrix loss remain to be established.
The reasons for the observed differences in responsiveness of the TMJ disc fibrocartilage, knee meniscus fibrocartilage, articular cartilage, and pubic symphysis to the hormones are not known. However, our recent work on identifying and quantifying estrogen receptor-α and -β and relaxin receptors LGR7 (leucine-rich repeat-containing, G protein-coupled receptor 7) and LGR8 may provide some insights into one potential reason for the observed differences. These findings show varied expression of these receptors between these tissues, supporting the conclusion that the more robust responses of the pubic symphysis and TMJ disc to relaxin and β-estradiol are possibly related to the presence of higher levels of receptors in these tissues than in the knee meniscus . The findings of the current study also raise the possibility that the distinct composition, organization, and biomechanical characteristics in different subtypes of cartilages  may influence the ability of systemic hormones to access each of the tissues, thereby influencing the net amount of matrix loss. These concepts need further study.
The similarities of our findings on the hormone-mediated changes in the collagen content of the pubic symphysis with those of previous investigators [12, 13] suggest that our model has several commonalities to animal models used previously and attest to its relevance for the purposes of this study. Samuel and colleagues , for example, showed that relaxin causes 64% (± 4%) and 68% (± 6%) decreases, respectively, in pubic symphysis collagen in unprimed and β-estradiol-primed ovariectomized non-pregnant rats. This compares with relaxin contributing to collagen loss of approximately 60% in unprimed and 80% in β-estradiol-primed rabbit pubic symphysis in the present study. Also, in agreement with our observations on the pubic symphysis, previous findings show that progesterone treatment rescues the collagen loss mediated by relaxin in estrogen-primed animals [12, 16]. Together, these observations indicate that the pubic symphysis served as an appropriate positive control for our experiments. In addition to confirming these previous observations on relaxin's contribution to collagen loss in the pubic symphysis, we also show new findings that β-estradiol and relaxin cause a slight, but statistically significant, loss of GAGs from this tissue.
Whereas a few previous studies have demonstrated other non-reproductive target sites, including the in vivo matrix turnover of lung alveolar tissues  and the in vitro modulation of tissue degradation in dermal fibroblasts  as well as fibrocartilaginous cells or tissues from joints [5, 21], our findings show for the first time that relaxin alters matrix composition of specific cartilages from non-reproductive sites in vivo. As with findings on reproductive tissues [9, 12, 16, 17, 22], we had previously demonstrated that the effect of relaxin on MMP induction is potentiated by β-estradiol in isolated fibrocartilaginous cells from the TMJ . In contrast, TMJ fibrocartilaginous explants  and the current in vivo findings show no potentiation by β-estradiol of relaxin's induction of MMPs or loss of collagen and GAGs. These differences in responses between cultured cells and tissue explants can likely be attributed to the differences in the behavior of isolated cells from cells in their natural matrix environment within the context of the explants and in the intact animal.
Our results also show a significant increase in systemic relaxin levels in all groups administered relaxin, with median concentrations ranging from a low of 37 pg/ml for rabbits receiving progesterone + relaxin to a high of 58 pg/ml for those administered β-estradiol + relaxin. This range of concentrations of relaxin in all groups of animals administered this hormone alone or in combination with other hormones is similar to that found systemically in cycling women [35–37]. Interestingly, our findings showed that β-estradiol priming enhances systemic relaxin concentrations, whereas priming with progesterone tended to diminish the serum levels of relaxin. These effects of hormone priming on systemic relaxin levels were reflected in the statistically greater fold increase of relaxin in β-estradiol + relaxin versus progesterone + relaxin groups (p < 0.01) (Figure 1a). This finding corresponds to some degree with the decreases in collagen and GAGs mediated by relaxin and β-estradiol + relaxin, but not by progesterone, progesterone + relaxin, or β-estradiol + progesterone + relaxin in the more responsive of the tissues studied, namely the TMJ disc and pubic symphysis. These results suggest that administration of relaxin using an implanted subcutaneous osmotic pump contributes to the increases in systemic relaxin concentration in which β-estradiol and progesterone may act as regulators of systemic relaxin concentration. However, further studies are required to clarify the mechanism by which β-estradiol, relaxin, and progesterone interact with each other to modulate systemic relaxin levels and subsequently maintain or disturb the homeostasis of the ECM in fibrocartilage.