MATERIALS AND METHODS

Mice

STR/ort mice were purchased from Harlan Winkelmann, Borchen, Germany. A total of 60 male and 62 female STR/ort mice were analysed and were killed at two different times, five months apart. The first cohort included 55, the second 67 mice. At each time, between two and four mice of each sex and cohort were killed. BALB/c mice aged 2 and 12 months served as controls.

Histology

Both knees of each mouse were prepared, thin sections were cut in an anterior to posterior direction and analysed for osteoarthritic changes by haematoxylin and eosin staining. The specimens were graded as follows: grade 0, healthy cartilage; grade 1, irregularities of the cartilaginous surface; grade 2, fissural ulcerations; grade 3, complete loss of articular cartilage with the subchondral bone being exposed. Both knees of two to four male and female STR/ort mice in each age group and cohort were analysed. The degradations of the medial and lateral tibial plateaux and femoral condyles were graded (grades 0–3), and the grades added up for both knees. The maximum score possible for each mouse is thus 24. Means were calculated for the various age and sex groups.

Immunohistochemistry

Cytokine expression of articular chondrocytes was assessed in the knees of 2, 3.5, and 10.5 month old STR/ort mice. Two month old BALB/c mice were used as controls. Immunohistochemical staining was performed using monoclonal antibodies recognising IL6, TNFα, TGFβ1, IL1β, IL4, and IL10. Three to four STR/ort mice were analysed in each age and sex group, the total numbers of articular chondrocytes in each thin section were counted, and the percentage of positive cells calculated. Staining with the primary monoclonal antibody omitted served as a control.

Statistical analysis

A Mann-Whitney test was performed to calculate the differences between the medians of cytokine producing chondrocytes in mice of different ages and sexes. All knee joints analysed were considered individually in this calculation.

RESULTS

Cartilage degeneration starts earlier and is more severe in male STR/ort mice

Whereas in female mice the onset of OA did not begin before 4.5 months of age, the first signs of OA started between 2 and 3.5 months of age in the males, and more severe grades of joint destruction were reached (fig 1A). In general the medial sides of the knee joints were more severely affected than the lateral ones. Interestingly, both of the male cohorts showed almost identical courses of disease, whereas the degree of cartilage degeneration differed in the two female cohorts. BALB/c mice only developed mild osteoarthritic changes and only at the age of 12 months.

Articular cytokine expression in male and female mice shows different patterns

The most dramatic differences were seen for IL1β, IL4, and TGFβ1 at 2 months of age. Whereas the articular chondrocytes of male BALB/c and STR/ort mice showed only mild expression of TGFβ1, IL1β, and IL4, chondrocytes from age matched female mice showed a marked increase (fig 1B). In STR/ort mice, the differences between males and females was significant, with p values of 0.006 for TGFβ1 and of 0.042 for IL4. The difference for IL1β did not quite reach significance (table 1). Figure 2 shows examples of staining performed with TGFβ1- and IL1β-specific antibodies.

At 3.5 months of age, the articular cytokine expression slowly declined in both sexes, with the exception of IL4 in the males (fig 1B). Interestingly, there was a statistically significant increase in TNFα positive chondrocytes in male STR/ort mice at 3.5 months of age, with a corresponding p value of 0.026 (table 1). At 10.5 months of age, all the cytokines analysed were reduced, and the reduction ranged between 1.4- and 8.2-fold (fig 1B). No positive cells were seen in any of the control stainings. With the exception of IL1β in the 2 month old male mice, the cytokine expression was more pronounced on the lateral sides of the joints.

DISCUSSION

Two unexpected findings emerge from this study, the first one being the increased expression of IL1β, TGFβ1, and IL4 in young female mice. A previous study showing that TGFβ1 combined with IL4 is chondroprotective in in vitro experiments⁴ is in line with our observation in STR/ort mice. We therefore suggest that an early peak of TGFβ1 and IL4 in the female mice mediates the protective effect, preventing and/or delaying the degenerative changes observed in the males. The correlation between increased cytokine expression and reduced cartilage degeneration at the lateral sides of the joints supports this idea (table 1). Even though our results support the importance of cytokines for cartilage metabolism, the role of IL1β remains obscure. On the one hand, IL1β is among the three cytokines up regulated in young STR/ort females being protected from OA. On the other, IL1β is known as a potent mediator of cartilage degradation.³ However, TGFβ1 combined with IL4 has in vitro been shown to be chondroprotective.⁴ We therefore suggest that the early peak of TGFβ1 and IL4 protects the cartilage of female mice and thereby contributes to the reduced severity of OA.

The differences in articular cytokine expression between male and female STR/ort mice are also reflected in BALB/c mice which do not develop OA. However, the varying incidence and severity of OA in STR/ort mice as well as the sexual dimorphism already indicate a multifactorial disease. Quite likely, the BALB/c background lacks the genetic predisposition responsible for the induction of OA in STR/ort mice.

Our results suggest that articular cytokines play a part in mediating the sexual dimorphism in the OA of STR/ort mice. At 2 months of age, sexual maturity is at its peak in the females and an impact of hormones on the cytokine expression is intriguing. Two of the most prominent hormones regulating cartilage generation and regeneration are oestrogens and leptin, and receptors for both of them are expressed on articular chondrocytes.^8,9 Likewise, bone metabolism has been shown to be regulated by oestrogens and here, the effect is thought to be mediated by IL1β.^8,10 IL1β in articular chondrocytes has been described as stimulating the expression of both TGFβ1 and IL4.¹¹ It is thus feasible that in the 2 month old female STR/ort mice, strong hormonal signalling results in increased IL1β levels, leading to an increase in articular TGFβ1 and IL4. Likewise, for human OA, the rise in incidence and severity among postmenopausal female patients has been explained by the withdrawal of oestrogens, resulting in the loss of a cartilage protective effect.^8,12 In STR/ort mice a study with ovariectomised females suggests that oestrogens have only a minor role in the protection from OA, but larger numbers of animals will be required to reach firm conclusions.¹³ As of today, nothing is known about the role of leptin or of male sex hormones in murine OA.

The second interesting finding is the absence of articular cytokines in the old STR/ort mice. Our data are in agreement with previous data showing a marked decrease of articular cytokines in old mice at a late stage of the disease.¹⁴ Our results for the cytokine expression of 8–9 month old mice differ from those published previously.¹⁴ However, the published STR/ort colony was only mildly affected by OA, whereas in our mice all the 8–9 month old STR/ort males showed clear signs of OA. We conclude that the rapid decline of cytokine expression is connected with the rapid progression of OA in the males.

Collectively, our data focus attention on the role of hormones in regulating articular cytokine expression and thus regulating cartilage metabolism. Understanding the correlation between oestrogens, leptin and, possibly, additional hormones and the expression of IL1β, TGFβ1, and IL4 may further our understanding of degenerative bone and joint diseases.