Hox11 genes are required for regional patterning and integration of muscle, tendon and bone

Abstract

Development of the musculoskeletal system requires precise integration of muscles, tendons and bones. The molecular mechanisms involved in the differentiation of each of these tissues have been the focus of significant research; however, much less is known about how these tissues are integrated into a functional unit appropriate for each body position and role. Previous reports have demonstrated crucial roles for Hox genes in patterning the axial and limb skeleton. Loss of Hox11 paralogous gene function results in dramatic malformation of limb zeugopod skeletal elements, the radius/ulna and tibia/fibula, as well as transformation of the sacral region to a lumbar phenotype. Utilizing a Hoxa11eGFP knock-in allele, we show that Hox11 genes are expressed in the connective tissue fibroblasts of the outer perichondrium, tendons and muscle connective tissue of the zeugopod region throughout all stages of development. Hox11 genes are not expressed in differentiated cartilage or bone, or in vascular or muscle cells in these regions. Loss of Hox11 genes disrupts regional muscle and tendon patterning of the limb in addition to affecting skeletal patterning. The tendon and muscle defects in Hox11 mutants are independent of skeletal patterning events as disruption of tendon and muscle patterning is observed in Hox11 compound mutants that do not have a skeletal phenotype. Thus, Hox genes are not simply regulators of skeletal morphology as previously thought, but are key factors that regulate regional patterning and integration of the musculoskeletal system.

Keywords: Connective tissue; Hox genes; Limb development; Mouse; Musculoskeletal integration; Stromal cells.

Fig. 1.

Forelimb zeugopod skeletal elements are mispatterned in Hoxa11/d11 double mutants and Hoxa11eGFP expression is maintained in the zeugopod throughout forelimb development. (A,B) Alcian Blue- and Alizarin Red-stained skeletal preparations of E18.5 control (A) and Hoxa11/d11 double mutant (B) mouse forelimbs. (C-E,I-K) Whole-mount view of Hoxa11eGFP heterozygous embryo forelimb at E10.5 (C), E11.5 (D), E12.5 (E), E13.5 (I), E14.5 (J), and E18.5 (K). (F-H,L-N) Transverse sections through Hoxa11eGFP heterozygous forelimb co-stained with an antibody for the pre-chondrogenic marker Sox9 (red) at E10.5 (F), E11.5 (G), E12.5 (H), E13.5 (L), E14.5 (M), E18.5 (N). The initially broad and scattered expression of Hoxa11eGFP (C,F) becomes localized by E11.5 and is largely excluded from cartilage condensations from the earliest stages. Scale bars: 300 μm.

Fig. 2.

Hoxa11eGFP is expressed in the outer perichondrium, tendons and muscle connective tissue but excluded from chondrocytes, osteoblasts, muscle and endothelial cells. (A-C) Longitudinal sections through E14.5 forelimbs of Hoxa11eGFP heterozygous embryos co-labeled for osteoblast markers by in situ hybridization for Runx2 (brown; A), antibody staining for osterix (red; B,B′), or β-galactosidase staining for the Bmp2 ECR lacZ reporter (blue; C) shows that Hoxa11eGFP is not expressed in osteoblasts but in the adjacent cells of the outer perichondrium. B′ shows a higher magnification of the boxed region in B. GFP fluorescence was converted to purple in images A and C to allow color visualization. (D-O′) Transverse sections along the proximodistal axis of the limb in the zeugopod of Hoxa11eGFP heterozygous embryos. (D-F′) Co-staining with an antibody to the tendon marker Tnc shows that Hox11 genes are expressed in all zeugopod tendons. Some tendons are marked by arrows. F′ shows a higher magnification of the boxed region in F. (G-L) Hoxa11eGFP expression is observed in all cells of the tendon and surrounding cells of the tendon sheath through E18.5. (M-O′) Antibody staining for the muscle marker MF20 demonstrates that Hoxa11eGFP is not expressed in muscle cells but in the cells closely associated with the muscle masses. r, radius; t, tendon; u, ulna.

Fig. 3.

Hoxa11eGFP is expressed in muscle connective tissue but not endothelial cells or muscle cells. (A-I) Transverse sections through forelimb zeugopod of Hoxa11eGFP heterozygous embryos at E14.5. (A-C) Antibody staining with the muscle marker My32 shows that Hoxa11eGFP is not expressed in muscle cells. (D-F) Hoxa11eGFP is also excluded from the endothelial compartment marked with an antibody for PECAM. (G-I) Hoxa11eGFP displays significant co-expression with antibody staining to the muscle connective tissue marker Tcf4. Some Hoxa11eGFP, Tcf4 double-positive cells are marked by arrows. t, tendon.

Fig. 4.

Forelimb zeugopod muscles are disrupted in Hoxa11/d11 double mutants whereas stylopod muscles are unaffected. (A-D) Transverse section through the stylopod of E14.5 control (A) and Hox11 double-mutant forelimbs (C) stained for differentiated muscle (My32 antibody) shows no difference in muscle pattern in this region in the absence of Hox11 genes. The zeugopod of the Hox11 mutant (D) displays severe patterning defects compared with control (B). (E-H) 3D reconstruction of serial sections stained with My32 antibody using Amira software. Numbers denote specific muscles listed in the table below. Many dorsal muscle groups are merged (10/11, 12/13/14) or absent (15) and several ventral muscle groups are absent (17, 19, 21). h, humerus; r, radius; u, ulna.

Fig. 5.

Tendon patterning is disrupted in the forelimb zeugopod of Hox11 double mutants. (A,F) Whole-mount in situ hybridization for the tendon marker scleraxis in control (A) and Hoxa11/d11 double mutant (F) forelimbs at E13.5 shows that tendon progenitors are normally specified in Hox11 mutants. (B-D,G-I) Transverse sections through the zeugopod of control (B-D) and Hoxa11/d11 (G-I) forelimbs at E14.5 from proximal (left) to distal (right) stained with an antibody for Tnc to detect tendons. The pattern of tendons in Hox11 mutants at this stage correlates with the altered muscle pattern in these animals. Numbers correspond to those shown in Fig. 4. (E,J) High magnification of Tnc-stained tendon in control (E) and Hox11 double mutant (J) shows disorganization of remaining tendons.

Fig. 6.

In the absence of Hox11 genes, tendon structure is abnormal and tendon sheath extracellular matrix is not present. (A-H) Transverse sections through the zeugopod of control (A,A′,C,E,G) and Hoxa11/d11 mutant (B,B′,D,F,H) forelimbs at E18.5. Hematoxylin and Eosin (H&E) staining (A-B′) shows abnormal tendon histology in Hox11 mutants. A′ and B′ show higher magnifications of the boxed regions in A and B, respectively. In situ hybridization for the tendon sheath marker Tppp3 (C,D) demonstrates presence of tendon sheath cells. Staining for components of the tendon extracellular matrix hyaluronic acid (HA) (red; E,F) and lubricin (green; G,H) is dramatically reduced surrounding the tendon of Hox11 mutants. In F and H, the tendon analogous to the one shown in E and G is outlined.

Fig. 7.

Tendons of Hox11 mutants have disorganized collagen fibers. (A,D) Sirius Red staining for collagen is reduced in Hox11 double-mutant (D) zeugopod tendons at E18.5 compared with control (A). Black arrowheads indicate three dorsal tendons. Insets show higher magnifications of boxed regions. (B,C,E,F) Transmission electron microscopy (TEM) images of collagen fibers in control (B) and Hox11 mutant (E) forelimb zeugopod tendon show disorganization of collagen in Hox11 mutant. Red arrowheads point to collagen fibers running parallel to the plane of section, opposite to the normal orientation. Synovial fluid-filled space surrounding the tendon in control sections (C) is not observed around tendons in Hox11 mutants (F). Yellow arrowheads indicate boundary of tendon fibroblasts. Blue arrowheads indicate boundary of tendon sheath cells.

Fig. 8.

Hox11 compound mutants display normal skeletal patterning, but muscle and tendon patterning is disrupted. (A) Alcian Blue- and Alizarin Red-stained skeletal preparation of an E18.5 Hox11 compound mutant (Hoxa11^+/-; Hoxd11^-/-) shows that the zeugopod skeletal elements are patterned normally when only one of the four Hox11 alleles is wild type. (B-D) Immunohistochemistry for differentiated muscle (My32 antibody) in transverse sections through the zeugopod of control (B) and Hox11 compound mutants (C,D). No separation is observed between the extensor digitorum communis and lateralis in compound mutants (red arrowheads) or between the extensor carpi radialis brevis and longus in Hoxa11^-/-;d11^-/+ (blue arrowheads). Ventral muscle groups are severely disorganized in compound mutants (C,D). (E-G) Transverse sections through the distal zeugopod of control (E) and Hox11 compound mutant (F,G) forelimbs at E14.5 stained with an antibody for Tnc to detect tendons shows disrupted tendon formation in Hox11 compound mutants compared with control. Numbers correspond to those shown in Fig. 4. Muscles and tendons were not numbered in compound mutants because precise identifications could not be made; asterisks indicate structures that could not be assigned. r, radius; u, ulna.