The Transcription Factor HAND1 Is Involved in Cortical Bone Mass through the Regulation of Collagen Expression

Temporal and/or spatial alteration of collagen family gene expression results in bone defects. However, how collagen expression controls bone size remains largely unknown. The basic helix-loop-helix transcription factor HAND1 is expressed in developing long bones and is involved in their morphogenesis. To understand the functional role of HAND1 and collagen in the postnatal development of long bones, we overexpressed Hand1 in the osteochondroprogenitors of model mice and found that the bone volumes of cortical bones decreased in Hand1Tg/+;Twist2-Cre mice. Continuous Hand1 expression downregulated the gene expression of type I, V, and XI collagen in the diaphyses of long bones and was associated with decreased expression of Runx2 and Sp7/Osterix, encoding transcription factors involved in the transactivation of fibril-forming collagen genes. Members of the microRNA-196 family, which target the 3′ untranslated regions of COL1A1 and COL1A2, were significantly upregulated in Hand1Tg/+;Twist2-Cre mice. Mass spectrometry revealed that the expression ratios of alpha 1(XI), alpha 2(XI), and alpha 2(V) in the diaphysis increased during postnatal development in wild-type mice, which was delayed in Hand1Tg/+;Twist2-Cre mice. Our results demonstrate that HAND1 regulates bone size and morphology through osteochondroprogenitors, at least partially by suppressing postnatal expression of collagen fibrils in the cortical bones.


Introduction
Skeletal dysplasia (osteochondrodysplasia) occurs in approximately 1 in 5000 births and is a major cause of severe short stature [1]. The size of each bone collectively determines the height and figure of the skeleton. Bone is primarily composed of a fibril-forming collagen matrix, accounting for approximately 80% of the bone extracellular matrix (ECM) [2]. Mutations in the collagen gene cause various skeletal dysplasias characterized by long bone deformities and midface hypoplasia, highlighting the importance of collagens in bone development [3]. Collagen is composed of three alpha chains that form a triple-helical structure which is essential for fibril strength and, bone collagens contribute to both the quantity and quality of the bone tissue [4]. Heterozygous mutations in the human COL1A1 and COL1A2 genes have been identified in patients with Ehlers-Danlos syndrome arthrochalasia type I (Online Mendelian Inheritance in Man [OMIM] #130060) and II (OMIM #617821), osteogenesis imperfecta (OMIM #166200, #166210, #259420, #166220), and Caffey disease (OMIM #114000). Heterozygous mutations in COL5A1 and COL5A2 have been identified in patients with Ehlers-Danlos syndrome classic type I (OMIM #130000) and II (OMIM #130010). Mutations in COL11A1 and COL11A2 are linked to Stickler syndrome type II (OMIM #604841), fibrochondrogenesis 1 (OMIM #228520), fibrochondrogenesis 2 (OMIM #614524), and Marshall syndrome (OMIM #154780). Type I collagen is the most abundant component of the bone matrix, accounting for up to 95% of bone collagens [2,5]. The cortical bone is composed of packed collagen type I fibrils and is highly mineralized, adding strength and rigidity to long bones [6]. The ratio of type V and XI collagens is approximately 3% of type I collagen in mature bones [5]. Type V and XI collagens assemble with type I collagen in the perichondrium of bones and modulate the size and shape of the fibrils [7][8][9], as well as interfering with the process of mineralization, which determines the stiffness of bones [6].
The long bone consists of the bony diaphysis, which connects the two ends, and the epiphyses. Cortical bone is the compact, dense outer layer that covers the bones, while trabecular bone lies in the interior of the epiphyses. Long bones develop through endochondral ossification in which the hyaline cartilage is replaced by mineralized tissue in the ossification center located in the middle of the diaphysis. Osteoblasts are differentiated from pluripotent perichondrial progenitor cells in response to osteogenic signaling inputs from differentiated chondrocytes [10][11][12]. Osteoblasts then invade the calcified cartilage with vascular tissues, secrete collagens, and create mineralized bone collar or cortical bone in the endochondral bones [13]. Collagen is involved in the formation of a tissue scaffold and structural stabilization, and also provides a substrate for cell anchorage, binds with non-collagenous proteins, and regulates the bioavailability of cytokines and growth factors [14]. The transcription factors RUNX2 (Runt-related transcription factor 2) and SP7/Osterix are essential for osteoblast specification and subsequent ossification of the bones [15][16][17]. Runx2 is expressed in osteoblasts and perichondrial cells in the bone collar and in primary spongiosa in long bones, whereas Sp7/Osterix expression is restricted to osteoblasts [16,18]. Runx2 and Sp7/Osterix encode transcription factors which regulate the expression of fibril-forming collagen genes [18][19][20][21]. Runx2and Sp7/Osterix-deficient mice lack mineralization of endochondral and intramembranous bones due to the absence of functional osteoblasts [16,17]. In humans, RUNX2 haploinsufficiency causes cleidocranial dysplasia (OMIM #119600), characterized by bone and teeth anomalies, including delayed fontanel closure, hypoplastic clavicles, short stature, and supernumerary teeth [22][23][24]. Homozygous mutations in SP7/Osterix have been linked to type XII osteogenesis imperfecta (OMIM #613849) characterized by skeletal anomalies including repeated bone fractures, generalized osteoporosis, and mild bone deformities.
Basic helix-loop-helix (bHLH) transcription factors play essential roles during embryonic development. Hand1 and Hand2, which encode the bHLH transcription factor heart and neural crest derivatives expressed protein 1 (HAND1) and HAND2 respectively, are expressed in the developing limb primordium [25,26]. We previously reported that HAND1 and HAND2 act as negative regulators of intramembranous and endochondral ossification by inhibiting RUNX2 [25,27]. While conditional Hand1 knockout mice with Wnt1-Cre or Prrx1-Cre exhibit no observable mandibular or limb abnormalities [26,28], mice that conditionally overexpress Hand1 and Hand2 with Wnt1-Cre or Twist2-Cre exhibit shortened and malformed mandible and limbs [25,27], indicating that the expression level of HAND transcription factors is critical for bone development. The overexpression of HAND2 causes limb and heart defects in patients with 4q trisomy: dup (4)(q35. 2-q31.22) [29,30]. However, the role of HAND transcription factors in bone size and quality control during postnatal growth remain largely unknown.
Here, we demonstrate that HAND1 plays a role in the regulation of the temporal expression of collagens involved in the postnatal development of long bone defects. In Hand1-overexpressing mice, the expression levels of fibril-forming collagens, as well as Runx2 and Sp7/Osterix, the upstream genes of fibril-forming collagens, were significantly decreased. In addition, the expression levels of the members of the microRNA (miRNA)-196 family, which specifically target the 3 untranslated regions (UTRs) of COL1A1 and COL1A2, were significantly upregulated. Reduced expression of bone-related collagens in the cortical bones may result in long bone anomalies. These findings uncovered multifaceted roles of HAND1 in regulating bone size by negatively regulating the temporal expression of collagens in the cortical bone.

Overexpression of Hand1 Induces Developmental Defects in the Skeletal Bones
To investigate the role of HAND1 in the postnatal development of long bones, we examined mice that conditionally overexpressed Hand1 (Hand1 Tg/+ ;Twist2-Cre) driven by the Twist2-promoter [25,27]. During skeletal development, Twist2 promoter-driven Cre expression is detected in the chondrocytes of the growth plate and in the osteoblasts in the periosteum, perichondrium, and endosteum without leakage in tendons or interlimb flanks [31][32][33]. All Hand1-overexpressing mice (n = 54) exhibited hypoplastic ossification of the skeletal bones, including long bones, ribcages, and vertebral bodies at P1 ( Figure 1A), whereas Hand1 Tg/+ mice and Twist2-Cre mice were indistinguishable from their wild-type littermates. All (n = 54) Hand1-overexpressing mice also displayed preaxial polydactyly [25]. By P21, Hand1-overexpressing mice were severely dwarfed compared to littermate controls of the same sex [25]. The lengths of the femurs and radii of the Hand1-overexpressing mice decreased compared to that of the wild-type at P1 and P21 ( Figure 1B,C). Matrix metallopeptidase 13 (MMP13), encoded by MMP13, is a downstream target of RUNX2 [34,35] and is exclusively expressed in primary ossification centers of the skeletal bones [36][37][38]. Immunostaining of MMP13 showed that the osteogenesis domain was hypoplastic in the Hand1-overexpressing femurs ( Figure 1D). Micro-computed tomography (micro-CT) analysis demonstrated a reduction in the cortical bone volume ( Figure 1E). These findings suggest that Hand1 expression levels affect embryonic and postnatal development of skeletal bones and that the bone size alteration observed in Hand1-overexpressing mice continues during postnatal growth.

Micro-CT Analysis of Long Bones
Hand1 overexpression affects the morphology and length of long bones ( Figure 1). To assess the structure and mineral composition of long bones, we performed micro-CT scanning of the femurs from wild-type and Hand1-overexpressing mice at P1 and P21 and analyzed the trabecular and cortical bones. Micro-CT analysis demonstrated a significant reduction in bone mineral content (BMC), trabecular bone volume (BV), bone tissue mineral density (TMD; BMC/BV), the density measurement restricted to the calcified bone tissue only, and total volume of interest (TV) in the trabecular bones from P1 Hand1-overexpressing mice relative to their wild-type littermates (Figure 2A). In the cortical bones from P1 Hand1-overexpressing mice, TMD, BMC, and cortical BV were also significantly reduced ( Figure 2B). In contrast, the skeletal anomalies displayed by P21 Hand1-overexpressing mice were not significantly different from wild-type littermates with respect to trabecular bone volume (BV; Figure 2C); however, a significant reduction was noted in total volume (TV; Figure 2C). In the cortical bones from P21 Hand1-overexpressing mice, BMC and cortical BV were significantly reduced ( Figure 2D). These findings suggest that the coupling between bone formation and resorption is restored in the P21 Hand1-overexpressing mice and that Hand1 has an inhibitory function in the postnatal development of cortical bone volume, but not in the maintenance of the trabecular bone volume.  (D) Micro-CT analysis of cortical bones of femurs from wild-type and Hand1-overexpressing mice at P21 (n = 3 per genotype). Analysis of structural parameters of wild-type and Hand1-overexpressing mice indicated reduced cortical bone volume in the Hand1-overexpressing mice. * p < 0.05; ** p < 0.01; *** p < 0.001 (compared to the wild-type). TMD, bone tissue mineral density; BMC, bone mineral content; BV, bone volume; TV, total volume of interest; BV/TV, trabecular bone volume fraction; BMC/TV, volumetric bone mineral density.

Expression of Bone-Related Collagens in the Diaphyses
Since Hand1-overexpressing mice display a failure in cortical bone formation, we focused our attention on cortical bone-related collagens. Collagen genes are involved in bone anomalies in various genetic diseases in humans (Tables S1 and S2). Col1a1, Col1a2, or Col5a2 mutant mice exhibit a series of bone phenotypes, including decreased length of the long bones and thickness of the compact bones [39][40][41] (Table S3). When we investigated whether overexpression of Hand1 affects the expression of collagen genes in the cortical bones of neonatal Hand1-overexpressing mice, we found that the expression of Col5a2, encoding the α2(V)-collagen chain, was significantly affected, whereas the expression of other collagen genes was unaffected ( Figure 3A). In contrast, in P21 Hand1-overexpressing mice, the expression levels of Col1a1, Col1a2, Col5a2, Col11a1, and Col11a2 were downregulated, whereas the expression of Col5a1 barely changed ( Figure 3B). Runx2 and Sp7/Osterix encode transcription factors which regulate the expression of fibril-forming collagen genes [18][19][20][21]. Since the expression levels of collagen genes were downregulated in the Hand1-overexpressing mice ( Figure 3A,B), we investigated whether overexpression of Hand1 affects the expression of Runx2 and Sp7/Osterix in the diaphysis. As expected, the expression of Runx2 was significantly decreased in Hand1-overexpressing mice at P1 and P21 ( Figure 3A,B). In addition, the expression of Sp7/Osterix was significantly decreased in Hand1-overexpressing mice at P21 ( Figure 3B). These findings suggest that HAND1 regulates postnatal bone growth volume by regulating the temporal expression of cortical bone-related collagens.  Table S4. (D) Real-time quantitative PCR analysis of miR-196a and miR-196b in wild-type and Hand1-overexpressing mice (n = 3 per genotype) at P21. miR-196a was significantly upregulated in Hand1-overexpressing mice (Hand1 Mut). * p < 0.05 (compared to the wild-type).

miR-196 Are Upregulated in Hand1-Overexpressing Long Bones
Since HAND1 significantly reduced Col1a1 and Col1a2 expression in the diaphysis, we speculated that the expression of miRNAs specific to type I collagen genes may be deregulated in the long bones of Hand1-overexpressing mice. To identify the miRNAs specific to type I collagen genes, putative miRNAs that target collagen alpha chains (Col1a1, Col1a2, Col5a1, Col5a2, Col11a1, and Col11a2) were identified (Table S4) using the miRNA prediction software TargetScan 7.2. By depicting the number of putative microRNAs (miRNAs) targeting collagen alpha chains [(COL1A1, COL1A2, COL5A1, COL5A2) and (COL1A1, COL1A2, COL11A1, COL11A2)] in the Venn diagram ( Figure 3C), we found two miRNAs, miR-196a and miR-196b, that commonly and specifically target COL1A1 and COL1A2 (Table S4). miR-196a is involved in the downregulation of type I collagen expression in scleroderma dermal fibroblasts [42]. The 3 UTRs of both Col1a1 and Col1a2 were predicted targets of miR-196a-5p and miR-196b-5p, but not the 3 UTR of other bone-related collagens ( Figure 3C, Table S4). When we examined whether miR-196a-5p and miR-196b-5p were deregulated in Hand1-overexpressing mice, we found that the expression of these miRNAs was upregulated in Hand1-overexpressing mice compared to wild-type littermates ( Figure 3D).

SDS-PAGE Analysis of Purified Collagen Samples
To further analyze whether HAND1 is functionally involved in the expression of collagens, thereby contributing to bone development, collagens were extracted and purified from the cortical bone from long bones at P7, P14, and P21. Equal amounts of collagen protein samples were loaded onto gels and analyzed by SDS-PAGE to evaluate the relative ratios of the collagen chains, not representing the absolute amount of each collagen type in bone tissue ( Figure 4A). The bands of type I collagen were assigned based on their migration patterns ( Figure 4A). Mass spectrometric analysis following in-gel trypsin digestion revealed that the trace bands above the alpha 1(I) chain band were type V and XI collagen alpha chains (Table S5; bands 1-3). Since type V and XI collagen chains were extracted first and more efficiently compared to type I collagen [43], the predominant extraction from bone by pepsin digestion resulted in detectable amounts of these minor collagens. The relative amounts of type V and XI collagens in the diaphyses increased with the maturation of long bones in the wild-type mice ( Figure 4A,B). In contrast, the increase in type V and XI collagens was delayed by the continuous overexpression of Hand1 in osteochondral progenitors ( Figure 4A,B). We also investigated the possibility of an abnormal posttranslational modification of type I collagen, which may be associated with a significant reduction in the size of cortical bones in Hand1-overexpressing mice at P21, but the analysis did not indicate any significant change in the modifications of proline and lysine residues compared to the wild-type (data not shown). Collectively, our findings suggest that Hand1 expression levels affect the postnatal bone volume of cortical bones through the temporal expression of bone-related collagens ( Figure 5).   [25,27]. RUNX2 binds its own promoter and establishes a positive autoregulatory loop [44]. RUNX2 also acts upstream of Col1a1, Col1a2, and Sp7/Osterix. SP7/Osterix in turn directly upregulates Col1a1, Col1a2, and Col11a2 by binding to Sp1 sites [18][19][20][21]45,46]. The expression levels of members of miR-196 family, which target the 3 UTR of Col1a1 and Col1a2 [42], are upregulated in Hand1-overexpressing mice. OSE, RUNX2-binding site; SP, Sp1 binding site.

Discussion
In this study, by analyzing mouse models with Hand1 overexpression in osteochondral progenitors, we demonstrated that HAND1 regulates postnatal bone growth volume by regulating the temporal expression of cortical bone-related collagens. The following findings support this conclusion: (1) overexpression of Hand1 in osteochondral progenitors resulted in the reduction of cortical bone volume in long bones; (2) expression of bone-related collagen genes as well as Runx2 and Sp7/Osterix, the upstream genes of fibril-forming collagens, decreased in Hand1-overexpressing mice; (3) expression of the members of miR-196 family, which target the 3 UTR of Col1a1 and Col1a2, was upregulated in Hand1-overexpressing mice; and (4) the increase in the postnatal expression of type V and XI collagens was delayed in the cortical bones of Hand1-overexpressing mice. The slower increase in type V and XI collagens during the maturation of long bones may be involved in the bone phenotypes observed in Hand1-overexpressing mice. While the bone phenotype was less apparent, the difference in the expression levels of fibril-forming collagen genes, Runx2, and Sp7/Osterix was significant between the Hand1-overexpressing mice and wild-type mice with increased postnatal days. The feedback from signaling pathways of endochondral bones [15,47], such as the parathyroid hormone-like hormone/parathyroid hormone-related protein, Indian hedgehog, Wnt/β-catenin, and fibroblast growth factor pathways as well as regulation by cilium assembly and transcription factors, may influence the bone phenotype of Hand1-overexpressing mice. We previously reported that Indian hedgehog (Ihh) expression was downregulated in the femur epiphyses of Hand1-overexpressing embryos, at least in part through the RUNX2-Ihh axis [25]. Aberrant expression of Runx2 and Ihh may contribute to the abnormal development of cortical and trabecular bones. Since Hand1-overexpressing mice exhibit preaxial polydactyly [25] and a reduction in cortical bone volume, the HAND1 region (5q33.2) and regulatory region of the HAND1 gene might be the candidate regions involved in polydactyly and short stature in humans.

Type I Collagen Expression in Hand1-Overexpressing Mice
The ECM of bones is primarily composed of type I collagen [2]. Overexpression of Hand1 in mice affects the expression of type I collagen genes, Col1a1 and Col1a2, and results in a significant reduction in the cortical bone volume, suggesting that HAND1 acts as a regulator that determines the amount of type I collagen in long bones. Mice carrying Col1a1 point mutations in the donor splice site of intron 36 or exon 9 exhibit decreased length of long bones and compact bone thickness [48]. Mice carrying Col1a2 mutations in mouse models of human osteogenesis imperfecta also exhibit decreased bone volume, compact bone thickness, and abnormal compact bone morphology [41]. HAND1 regulates osteoblast differentiation by inhibiting transactivation of RUNX2 [25,27], which directly regulates its own expression in a positive autoregulatory loop [18]. RUNX2 is required for early stages of osteoblast differentiation and acts upstream of the fibril-forming collagens Col1a1 and Col1a2 [18,19]. Therefore, the expression of type I collagen genes may be regulated by HAND1 through the inhibition of RUNX2 transactivation. RUNX2 also acts upstream of SP7/Osterix [45], which in turn binds to the proximal promoters of Col1a1 and Col1a2 genes and upregulates these genes [20,21]. Since the 3 UTRs of Col1a1 and Col1a2 have conserved target sites of miR-196a [42], and the expression levels of the members of miR-196 family were upregulated in Hand1-overexpressing mice, this family of miRNAs may play a role in determining the volume of cortical bones by regulating the expression of type I collagen genes. Interestingly, while members of the miR-196 family target only Col1a1 and Col1a2, other bone-related collagen genes also have common miR-29 family-targeting sites in their 3 UTRs. Furthermore, this property is conserved between mice and humans. Our findings suggest that in addition to HAND1-mediated downregulation of Runx2 and Sp7/Osterix expression, HAND1-mediated upregulation of the members of miR-196 family may also contribute to the downregulation of type I collagen expression in the cortical bones.

Expression of Type V and XI Collagens in Hand1-Overexpressing Mice
Hand1-overexpressing mice exhibit decreased length of long bones with decreasing amounts of type V and XI collagen. Among the different members of fibrillar collagens, type V and XI collagens are minor but essential components of collagen fibrils by serving as templates for fibril polymerization of type I collagen [49]. Alpha 1(V), alpha 2(V), alpha 1(XI), and alpha 2(XI) chains accumulate in the collagen component of long bones at postnatal age [7]. Homozygous Col5a2 knockout mice produced structurally abnormal type V and I collagen fibrils, and Col5a2-deficient femurs grow at a slower rate than control bones, resulting in decreased bone size [6]. Homozygous Col11a1-mutant mice also exhibit skeletal defects, including decreased length of long bones, abnormal hindlimb morphology, and micromelia [50]. An increased amount of type V collagen is observed in gracile bone dysplasia (OMIM #602361) [51]. These findings suggest that appropriate amounts of type V and XI collagens are critical in the development and morphogenesis of long bones downstream of HAND1. In Hand1-overexpressing mice, the expression of Sp7/Osterix was significantly decreased. SP7/Osterix upregulates COL11A2 by binding to the GC-rich specific Sp1 binding site of the promoter [46]. Since the proximal promoters of Col5a2 and Col11a1 contain putative Sp1 binding sites [39,52], SP7/Osterix may also contribute in regulating the expression of type V and XI collagens in the cortical bones.
In summary, we found that the continuous expression of Hand1 in osteochondral progenitors resulted in decreased expression of type I, V, and XI collagens in the diaphyses of long bones. In Hand1-overexpressing mice, the expression of Runx2 and Sp7/Osterix, which encode transcription factors which regulate the expression of fibril-forming collagen genes, was significantly decreased. In addition, miR-196a-5p and miR-196b-5p, which target the 3 UTRs of COL1A1 and COL1A2, were upregulated in Hand1-overexpressing mice. Analysis of the fundamental relationship between the expression patterns of bone-related collagen genes responsible for bone structures will provide insights into how these collagen genes interact with factors involved in bone development. Investigation of the regulatory mechanisms of bone-related collagen gene expression by transcription factors and/or miRNAs may lead to new therapies for bone size and quality control.

Mice Conditionally Overexpressing Hand1
Hand1-overexpressing mice conditionally driven by Twist2-Cre (Hand1 Tg/+ ;Twist2-Cre) have been described previously [25,53]. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All animal experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee of the Tokyo Medical and Dental University (Permit Number: 0160215A, March 27, 2015).

Bone Staining, Histology, and Immunohistochemistry
Bone staining was performed using alizarin red and alcian blue as described previously [54]. Tissue samples for histology were fixed in 4% paraformaldehyde, decalcified, and embedded in paraffin, as described previously [25]. For MMP13 immunostaining, tissue sections were treated with 1mg/mL hyaluronidase (Sigma-Aldrich, St. Louis, MO, USA) at 37 • C for 45 min, and incubated with anti-MMP13 antibody (ab84594; Abcam, Cambridge, UK), followed by sequential treatment with components of the Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA, USA) and Immpact™ DAB peroxidase substrate (Vector Laboratories). The sections were counterstained with methyl green nuclear counterstain (Vector Laboratories).

Micro-Computed Tomography
Mineralized tissue formation was assessed by micro-computed tomography (micro-CT). Femurs were harvested at postnatal days 1 (P1) and P21 (n = 3 per group). Micro-CT images were scanned at a voltage of 75 kV and 140 µA in beam current, with filtration through a 0.1 mm brass plate, using an inspeXio SMX-100CT (Shimadzu, Kyoto, Japan). Scans were set at a pixel size of 512 × 512 and voxel size of 0.016 mm/voxel. The results were further analyzed using the TRI-3D-BON imaging system (Ratoc, Tokyo, Japan).

Real-Time Quantitative PCR
Bony diaphyses were harvested, frozen, wrapped in foil, and ground into a powder using a mortar and pestle containing liquid nitrogen. Total RNA containing the miRNA fraction was extracted using TRIzol (Thermo Fisher Scientific, Waltham, MA, USA). Real-time quantitative PCR (qRT-PCR) was performed as described previously [25] and the relative expression of the target genes was normalized to β-actin. For each sample, three replicates were run for each gene. Primer sequences used for qRT-PCR are listed in Table S6.
The miRNA fraction was extracted from total RNA using the miRNeasy Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Putative miRNAs that target the collagen family (Col1a1, Col1a2, Col5a1, Col5a2, Col11a1, and Col11a2) were identified using the miRNA prediction software TargetScan 7.2 (http://www.targetscan.org) and are shown in Table S4. qRT-PCR of the miRNAs was performed using specifically designed stem-loop primers for mature miRNA analysis (GeneCopoeia, Rockville, MD, USA), following the manufacturer's protocol. cDNA was synthesized from 2 µg of total RNA and specific qRT-PCR experiments for miRNAs were carried out using All-in-One miRNA qRT-PCR Detection Kit (GeneCopoeia). Amplification and detection were performed using the StepOne Plus System (Thermo Fisher Scientific). Each PCR reaction was run in triplicate. The endogenous control, U6 (GeneCopoeia) was used for normalization, and the relative expression of miRNAs was calculated using the 2 -∆∆Ct method.

Extraction and Purification of Cortical Bone Collagens
The diaphyses of the long bones were dissected from Hand1-overexpressing mice at P7, P14, and P21 and pooled prior to analysis. The bones were demineralized in 0.5 M EDTA (pH 7.8) for 3 d at 4 • C, and the soft tissue and bone marrow were removed from the diaphysis after cutting off the epiphyses. The demineralized bones were treated with 5 mg/mL pepsin (Sigma-Aldrich) in 0.5 M acetic acid for 3 d at 20 • C, and the extracted collagens were purified using salt precipitation (1 M NaCl) and isoelectric precipitation (pH 8.0).

Protein Identification Using in-Gel Digestion Followed by Mass Spectrometry
Protein identification was performed using in-gel digestion as described previously [55]. The relative expression of collagen chains purified from the cortical bones from wild-type and Hand1-overexpressing mice was evaluated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using a 5% gel under non-reducing conditions. After staining with Coomassie Brilliant Blue R-250, the SDS-PAGE gel was scanned and the band intensity was measured by densitometric analysis using Multi Gauge version 3.0 (Fujifim, Tokyo, Japan). Protein bands of purified collagen samples were excised and digested in-gel with trypsin (Promega, Madison, WI, USA) at 37 • C for 16 h. The tryptic digests were analyzed by liquid chromatography-mass spectrometry on a maXis II quadrupole time-of-flight mass spectrometer (Bruker Daltonics, Bremen, Germany) coupled to a Shimadzu Prominence UFLC-XR system (Shimadzu, Kyoto, Japan) with chromatographic separation using an Ascentis Express C18 HPLC column (2.7 µm particle size, L × I.D. 150 mm × 2.1 mm; Supelco, Bellefonte, PA, USA) as described previously [56]. A database search was performed against the UniProtKB/Swiss-Prot database (release 2018_05) for Mus musculus species (16970 protein entries) using ProteinPilot software 4.0 (AB Sciex, Foster City, CA, USA).

Statistical Analysis
Calculations and statistical analyses were performed using Microsoft Office Excel 2004 (Microsoft Corporation, Redmond, WA, USA). All data were expressed as mean ± standard error of mean (S.E.M). A two-tailed Student's t-test was performed to compare two groups of independent samples and a normal distribution was assumed. Results with p < 0.05 were considered as statistically significant.
Supplementary Materials: Supplementary materials can be found at http://www.mdpi.com/1422-0067/21/22/ 8638/s1. Table S1. Skeletal disorders induced by mutations in type I collagen genes in humans. Table S2. Skeletal disorders induced by mutations in types V and XI collagen genes in humans. Table S3. Skeletal phenotypes induced by mutations in cortical bone-related collagen genes in mice. Table S4. MicroRNAs that are predicted to target cortical bone-related collagen genes. Table S5. Summary of the proteins identified in bands that decreased in Hand1-overexpressing mice. Table S6. Primer sequences for real-time quantitative PCR.