Regulation of Bone Formation by MicroRNAs

Previous work in our lab involved a laser capture microdissection approach to identify microRNAs (miRNAs) expressed in chondrocytes during long bone development (McAlinden, 2013). From this study, we identified both highly expressed miRNAs as well as differentially-expressed miRNAs in cells at distinct stages of chondrocyte differentiation within the developing growth plate. We are now pursuing specific miRNA candidates to determine their function in regulating both cartilage formation (see below) and bone formation. We found miR-181a and miR-138 to be more highly expressed in terminally-differentiated hypertrophic chondrocytes. This expression pattern suggested that these miRNAs may function to control not only chondrocyte differentiation, but also bone cell (osteoblast) differentiation given that hypertrophic cartilage serves as a template for endochondral bone formation.

We recently showed that miR-138 inhibits in vitro osteogenesis in monolayer and in 3D scaffold systems and that one of the mechanisms involved is via inhibition of RhoC and formation of the actin cytoskeleton (Zheng, 2018 ) (Fig. 1). On the other hand, the miR-181a/b cluster enhances osteogenesis, in part by increasing mitochondrial metabolism (Zheng 2019 ) (Fig. 2).

Figure 1: Effect of lentiviral (LV)-induced miR-138 over-expression on in vitro osteogenesis in monolayer (A: Alizarin Red staining) and in 3D scaffolds (B: microCT image). Phalloidin staining (C) shows that miR-138 inhibits the process of actin polymerization within the cells.
Figure 2:  Effects of lentiviral (LV)-induced over-expression of miR-181a/b on in vitro osteogenesis in monolayer (A: Alizarin Red staining) and 3D scaffolds (B: microCT image).  Seahorse technology was used to measure mitochondrial metabolism (C).  miR-181a/b increases the oxygen consumption rate of mitochondria.

Importantly, preliminary data shows that we can recapitulate the in vitro functions of these miRNAs in vivo in mouse models of bone fracture repair or heterotopic ossification (Figs. 3 and 4). Moving forward, we plan to thoroughly evaluate the in vivo effects of either activating or inhibiting miRNA function as a means to enhance bone fracture repair or inhibit/attenuate formation of heterotopic bone. We have also generated an in-house miR-181a/b conditional Rosa knock-in mouse to further evaluate the effects of miR-181a/b-1 over-expression on enhancing bone formation and repair. We also plan to further unravel the effects of these miRNAs on cell metabolism with the additional aim of identifying potential new therapeutic targeting strategies.

Figure 3: Local injection of LV-138 inhibits bone formation in a mouse model of trauma-induced heterotopic ossification (A-D) and in cortical bone defect model (E-H).  MicroCT images in B and D are of the regions containing heterotopic bone between the yellow lines at the calcaneus site shown in A and B, respectively.  Data is from 5 wk post HO induction. F and H show less bone repair 7 days following cortical bone injury due to miR-138 over-expression.  Data is compared to injection of a non-specific RNA control lentivirus (LV-NS).

Figure 4:  Over-expression of miR-181a/b appears to enhance endochondral ulnar fracture repair in mice.  14 days following fractual repair, microCT images reveal a larger mineralized repair callus formed in animals receiving local delivery of lentivirus expressing miR-181a/b-1 when compared to non-silencing (NS) control lentiviral injections.

A new avenue of research in this bone-focused project involves mitochondria-associated microRNAs (mito-miRs). The study of mito-miRs is still in its infancy and we don’t know anything about how they may function in regulating skeletal development or homeostasis. We recently carried out a miRNA microarray to identify miRNAs in the mitochondria of progenitor human mesenchymal stem/stromal cells (MSCs) as well as in MSCs during osteogenic induction. We plan to publish the findings from this array soon and anticipate that a number of exciting, fruitful projects will emerge from this work.

Regulation of Cartilage Formation by microRNAs

In addition to a function in regulating bone formation, we also find that miR-138 and miR-181a/b affects chondrocyte differentiation too. While miR-138 inhibits osteogenesis (Fig. 1) it does not inhibit overt chondrogenesis. However, miR-138 appears to prevent formation of large, terminally differentiated hypertrophic chondrocytes and formation of type X collagen (a protein specifically generated by hypertrophic chondrocytes). On the other hand, miR-181a/b apparently promotes hypertrophic chondrocyte differentiation (Fig. 5). These findings are important in the context of cartilage and osteochondral tissue engineering and osteoarthritis. For example, it is desirable for articular cartilage engineering to prevent formation of hypertrophic chondrocytes since these cells are associated with catabolic events in osteoarthritis. However, promoting hypertrophic differentiation may be beneficial for endochondral bone formation that requires a hypertrophic cartilage template. We are now designing new tissue engineering approaches in collaboration with Dr. Farsh Guilak to attempt to engineer osteochondral tissue consisting of a layer of articular cartilage and underlying bone tissue via scaffold technology and miRNA-mediated progenitor cell differentiation. Other projects will involve modulating miRNA function in mouse models of osteoarthritis to attempt to inhibit or attenuate progression of this disease.

Figure 5:  Effects of miR-181a/b or miR-138 over-expression on in vitro chondrogenesis.  Compared to non-silencing (NS) RNA control (A), Safranin-O staining showed increased hypertrophic-shaped cells (B) and increased collagen type X (COL 10) matrix staining (E) in LV-181a/b cultures.  miR-138 suppressed type X collagen protein production.

Long Non-Coding RNAs in Regulating Chondrogenesis

We have identified, via a collaborative effort with the Guilak Lab, long non-coding RNAs (lncRNAs) that appear to regulate chondrogenesis. We first identified these lncRNA candidates from an RNA-Seq study where mRNA and lncRNA expression was determined during human MSC chondrogenesis. We are currently studying two lncRNAs (LOC105370526 and ZFHX4-AS1): suppression of these lncRNAs via shRNA technology appears to inhibit in vitro chondrogenesis (Fig. 6). Studies are underway to determine the effects of over-expressing these lncRNAs during chondrogenesis and to determine the mechanism by which these non-coding RNAs function.

Figure 6: Knockdown of IncRNAs, LOC105370526 and ZFHX4-AS1, results in inhibition of in vitro human MSC chondrogenesis.  Following 6 weeks of chondrogenic induction, Safranin-O staining of pellet tissue sections shows reduction in pellet size and, in the case of LOC knock-down, there was an inhibition of proteoglycan matrix formation.

Generation of Osteochondral Tissue Constructs

The goal of this project, in collaboration with Dr. Farsh Guilak’s lab, is to engineer native-like human osteochondral tissue by modulating non-coding microRNA (miRNA) function in skeletal progenitor cells embedded within 3D woven PCL scaffolds. This work is clinically important given that there is an unmet need to treat articular cartilage lesions, a condition that can lead to osteoarthritis over time.

Preliminary data (Fig. 1 below) shows that lentivirus (LV) over-expressing miR-181a/b appears to enhance hypertrophic chondrocyte differentiation in skeletal progenitor cell chondrogenesis assays. This is shown by an increase in chondrocyte size (a typical feature of hypertrophic chondrocytes; B) and synthesis of a collagen type X extracellular matrix (E) when compared to cultures transduced with control non-silencing (NS) RNA (A, D). On the other hand, over-expression of miR-138 does not have apparent effects on cell shape (C), but there is a decrease in type X collagen expression (F).

We now propose to follow on from these findings to attempt to generate osteochondral tissue in 3D scaffolds containing a layer of articular-like cartilage (induced by miR-138) and a bottom layer of hypertrophic-like cartilage that will provide an appropriate template for bone formation when placed in osteogenic medium. Fig. 2 below depicts what we aim to achieve. Published work from the Guilak lab has already displayed the efficacy of their PCL scaffolds in permitting lentiviral binding, and to support the generation of a cartilage matrix by cells embedded within these structures. This work received a 3.0 percentile score from the NIH in October 2019 and we are excited to begin this work soon.

Print Friendly, PDF & Email