Mmp13 expression is downregulated by ERα
To identify estrogen target genes that regulate osteoclastogenesis indirectly, i.e., via actions on cells of the mesenchymal lineage, we performed microarray analysis of GFP sorted Osx1 + cells without or with ERα, derived from the calvaria cells of ERαf/f;GFP:Osx1-Cre mice or GFP:Osx1-Cre controls. The highest up-regulated gene in ERα deleted mesenchymal/stromal cells encodes the matrix metalloproteinase 13 (Mmp13), as shown in the heat map (Fig. 1a) and in the volcano plot (Fig. 1b). The microarray findings from the GFP sorted ERα deleted Osx1 + calvaria cells were confirmed by qPCR and reproduced in cultures of ERα deleted Prrx1 + bone marrow stromal cells derived from ERαf/f;Prrx1-Cre mice (Fig. 1c). Moreover, the active form of MMP13 protein levels was 4-fold higher in the BM plasma of OVX C57BL/6 mice, as compared to sham controls (Fig. 1d). In line with our findings, ERα regulates the Mmp13 promoter activity in synoviocytes 14.
Mmp13 deletion in mesenchymal progenitors decrease the length of the femur and tibia
To elucidate the role of Mmp13 in bone homeostasis in vivo, we next generated mice with conditional deletion of Mmp13 in mesenchymal progenitors expressing Prrx1 (Mmp13ΔPrrx1) and used floxed mice (Mmp13f/f) as control. The Prrx1-cre transgene targets early limb bud and a subset of craniofacial mesenchymal stem cells. We did not detect a skull phenotype. All measurements were made in the femur and/or tibia. Please note that in the following description of the results, the p values from two-way ANOVA analysis are provided below each graph. The expression of the Mmp13 mRNA in femur and tibia shafts was dramatically decreased in the Mmp13ΔPrrx1 mice (Fig. 2a), establishing the effectiveness of the deletion. Body weight and uterine weight were not affected by the Mmp13 deletion (Fig. 2b-c). However, femur and tibia length was decreased Fig. 2d-e).
In contrast to the microarray data of Fig. 1, we did not detect a change of the mRNA levels of Mmp13 in the OVX Mmp13f/f or Mmp13ΔPrrx1 mice in the osteocyte-enriched bone marrow- devoid preparations we used for this measurement (Fig. 2a). As expected, OVX increased body weight in Mmp13f/f mice (Fig. 2b) and decreased uterine weight in Mmp13f/f and Mmp13ΔPrrx1 mice (Fig. 2c). Femur and tibia length was not affected by OVX in either genotype (Fig. 2d-e).
Mmp13 deletion increases cortical bone and attenuates the cortical bone loss caused by OVX
Mmp13 deletion in Prrx1 cells caused greater cortical thickness and cortical area in the femoral diaphysis as indicated by pairwise comparison between sham-operated Mmp13f/f and sham-operated Mmp13ΔPrrx1 mice (Fig. 3a-b). This effect was due to a smaller medullary area (Fig. 3c) while total area did not change (Fig. 3d). The greater cortical thickness with Mmp13 deletion was less marked in the distal metaphysis of the femur (Fig. 3e) and in the diaphysis of the tibia (Fig. 3f).
OVX of the Mmp13f/f control mice decreased cortical thickness at the femoral diaphysis and distal metaphysis as well as the tibia diaphysis (Fig. 3a, e-f). Consistent with our working hypothesis, the effects of OVX on cortical bone at the diaphysis and distal metaphysis of the femur and tibia diaphysis were attenuated in the Mmp13ΔPrrx1 mice (Fig. 3a, e-f). Together, these data suggest that Mmp13 deletion increases cortical bone mass in femur and tibia and prevents or attenuates the loss of cortical bone caused by estrogen deficiency.
Mmp13 deletion increases trabecular bone but does not affect the loss of bone caused by OVX in this compartment
Trabecular bone volume was higher in both the distal femur and the proximal tibia by approximately 3.7- and 3-fold respectively, in sham-operated Mmp13ΔPrrx1 mice as compared to sham-operated Mmp13f/f controls (Fig. 4a-e). This difference was due to higher trabecular number (Fig. 4b) and thickness (Fig. 4c); and was mirrored by lower trabecular separation (Fig. 4d). In several mice, trabecular bone extended to the midshaft (see lower left micro-CT image in Fig. 3a).
As seen before 15,16, at six-months of age estrogen sufficient female mice have very little trabecular bone mass remaining at the distal femur (Fig. 4a). There was no discernible effect of the OVX at this site in Mmp13f/f mice (Fig. 4a-d). However, we observed a loss of trabecular bone mass in both the femur and tibia of the OVX Mmp13ΔPrrx1 mice (Fig. 4a, e). Collectively, these data indicate that Mmp13 deletion increases trabecular bone mass but does not prevent the loss of trabecular bone caused by estrogen deficiency.
Mmp13 deletion decreases osteoclast number in cortical bone
To elucidate the cellular mechanism by which the Mmp13 deletion increased cortical bone volume, we performed histomorphometric analysis of the endocortical surface of the femur. Mmp13 deletion caused an approximately 50% reduction in osteoclast number and surface (Fig. 5a-c), as indicated by pairwise comparison between sham-operated Mmp13f/f and sham-operated Mmp13ΔPrrx1 mice. The MMP13 deletion had no effect on mineral apposition rate (MAR), mineralized surfaces (MS) or bone formation rate (BFR) (Fig. 5d-g). These findings suggest that a decrease of osteoclast number and thereby resorption are responsible for the increase of cortical bone.
OVX of both Mmp13f/f and Mmp13ΔPrrx1 mice resulted in the expected increase in osteoclast number and surface (Fig. 5a-b), while MAR, MS and BFR were not affected (Fig. 5d-f). Surprisingly, in the OVX Mmp13ΔPrrx1 mice the increase in osteoclast number and surface was greater (5.5-fold) as compared to the OVX-induced increase in the Mmp13f/f mice (2-fold).
Mmp13 deletion increases whole-bone strength of the femur
It has been previously reported that Mmp13-/- mice have increased cortical bone fragility 17. To examine bone strength in female Mmp13ΔPrrx1 mice we performed three-point bending of the femur (Fig. 6a). Despite the thicker cortices in female Mmp13ΔPrrx1 mice, the moment of inertia was not different from the controls (Fig. 6b). Nonetheless, the yield load, peak load, and stiffness were all higher in Mmp13ΔPrrx1 mice (Fig. 6c). With respect to the estimated material properties, female Mmp13ΔPrrx1 mice had increased yield stress and ultimate stress but the same modulus as compared to control mice (Fig. 6d). Material density determined by micro-CT was also not different (Fig. 6e). Therefore, and in contrast to a previous report, deletion of Mmp13 led to an increase in bone structural and material properties.
Similar to females, Mmp13 deletion in males increases trabecular bone mass and whole-bone strength of the femur, but has no effect on cortical bone
The bone phenotype of sex steroid sufficient male Mmp13ΔPrrx1 and Mmp13f/f mice was analyzed at 4 and 6 month of age. Body weight was not affected by the Mmp13 deletion at either age (Fig. 7a), but femur length decreased at 4 and 6 months of age (Fig. 7b), as it did in females. In difference to females, cortical thickness in male mice was not affected by the Mmp13 deletion (Fig. 7c). Trabecular bone volume increased 1.4- and 1.6-fold at 4 and 6 months, respectively (Fig. 7d), but this increase was lower compared to the one we observed in Mmp13ΔPrrx1 female mice. The increased trabecular bone volume was associated with an increase in the number of trabeculae and a decrease in trabecular separation (Fig. 7e, g), but no change in trabecular thickness (Fig. 7f).
Finally, similar to females, three-point bending of the femur in male Mmp13ΔPrrx1 mice revealed higher bone strength including an increase in stiffness and modulus (Fig. 7h-j), with no change in material density (Fig. 7k); but unlike females the moment of inertia was decreased in males (Fig. 7h).