**2. Bone remodelling**

#### **2.1. Cellular and molecular mechanisms of bone remodelling**

Physiological remodelling of bone has proven to be a highly coordinated process constituted by bone resorption following bone formation, and it is necessary for damaged bone to be repaired, and to maintain mineral homeostasis (Mackie, Ahmed et al. 2008; Kular, Tickner et al. 2012; Thompson, Rubin et al. 2012). In addition to the traditional cells localized within the bone tissue (i.e. osteoclasts, osteoblasts, and osteocytes) necessary for completion of multiple bone remodelling cycles, several classes of immune cells have also been implicated in bone turnover and also bone disease states (Chen and Tuan 2008; Lories 2011; Mellis, Itzstein et al. 2011; Sims and Walsh 2012). The two processes of modelling and remodelling ensure proper development and maintenance of the skeletal system. The process of bone modelling is mandatory for growth and mechanically induced adaption of bone (Kular, Tickner et al. 2012; Thompson, Rubin et al. 2012), but it requires that the processes of bone formation and bone resorption (removal of "old" bone with micro-fractures) occur independently, and at distinct locations within the skeleton (Chang, Raggatt et al. 2008; Raggatt and Partridge 2010; Kular, Tickner et al. 2012). These tightly coordinated events require the synchronized activities of multiple cellular participants and compartments, in order to ensure that bone resorption and formation occur sequentially within the same anatomical locations to preserve bone mass with a functional geometry.

**Figure 1.** Schematic representation of a previously published microRNA profile distinguishing between osteoblasts and chondrocytes differentiated from human mesenchymal stem cells (hMSCs). Right hand frame: The six microRNA species have been proven to bind to and suppress the transcription of genes encoding nine factors deemed important for osteoblastogenesis, i.e. SATB2, HES-1, ETS-1, RUNX2, APC2, OSTERIX, VDR, LEF-1, and RNF11 (Marie 2008; Gorde‐ ladze, Djouad et al. 2009). Left hand frame: Interestingly, the microRNA species mir-24, and especially mir-149, may be regarded as important "switches" between the osteoblast and chondrocyte phenotypes, since they both positively suppress transcription factors (TFs), as predicted by using the MirnaViewer and the Sanger algorithms, necessary for osteoblastogenesis, and at the same time negatively affect factors (like ATF4, PIAS1, STAT6, and PPPIR6B) (Legendre, Dudhia et al. 2003; Komatsu, Mizusaki et al. 2004; Oh, Kido et al. 2007; Suzuki, Osumi et al. 2010), which serve as inhibitors of Sox9 expression, deemed as the most important TF initiating chondrocytogenesis and maintaining the chondrocyte phenotype.

#### **2.2. Cells being involved in bone remodelling**

2011; Sims and Walsh 2012). The two processes of modelling and remodelling ensure proper development and maintenance of the skeletal system. The process of bone modelling is mandatory for growth and mechanically induced adaption of bone (Kular, Tickner et al. 2012; Thompson, Rubin et al. 2012), but it requires that the processes of bone formation and bone resorption (removal of "old" bone with micro-fractures) occur independently, and at distinct locations within the skeleton (Chang, Raggatt et al. 2008; Raggatt and Partridge 2010; Kular, Tickner et al. 2012). These tightly coordinated events require the synchronized activities of multiple cellular participants and compartments, in order to ensure that bone resorption and formation occur sequentially within the same anatomical locations to preserve bone mass

**Figure 1.** Schematic representation of a previously published microRNA profile distinguishing between osteoblasts and chondrocytes differentiated from human mesenchymal stem cells (hMSCs). Right hand frame: The six microRNA species have been proven to bind to and suppress the transcription of genes encoding nine factors deemed important for osteoblastogenesis, i.e. SATB2, HES-1, ETS-1, RUNX2, APC2, OSTERIX, VDR, LEF-1, and RNF11 (Marie 2008; Gorde‐ ladze, Djouad et al. 2009). Left hand frame: Interestingly, the microRNA species mir-24, and especially mir-149, may be regarded as important "switches" between the osteoblast and chondrocyte phenotypes, since they both positively suppress transcription factors (TFs), as predicted by using the MirnaViewer and the Sanger algorithms, necessary for osteoblastogenesis, and at the same time negatively affect factors (like ATF4, PIAS1, STAT6, and PPPIR6B) (Legendre, Dudhia et al. 2003; Komatsu, Mizusaki et al. 2004; Oh, Kido et al. 2007; Suzuki, Osumi et al. 2010), which serve as inhibitors of Sox9 expression, deemed as the most important TF initiating chondrocytogenesis and maintaining the

with a functional geometry.

478 Regenerative Medicine and Tissue Engineering

chondrocyte phenotype.

*Osteoblasts* - Osteoblasts are specialized bone-forming cells responsible for many of the crucial functions constituting bone remodelling (i.e. expression of osteoclastogenic factors, production of bone matrix proteins, and bone mineralization) (Karsenty 2008). The stage of differentiation acquired by the osteoblasts influences the functional contribution to and impact of these cells on *in vivo* bone remodelling. Immature pre-osteoblasts influence osteoclastogenesis, while mature osteoblasts are responsible for matrix production and mineralization. Osteoblasts develop from pluripotent hMSCs, which exhibit inherent potential to develop into adipocytes, myocytes, chondrocytes, and osteoblasts orchestrated by a series of defined transcription factors (TFs). Osteoblast differentiation has since long been reported to be controlled by RUNX2, but several other TFs are now recognized as equally important for osteoblastogenesis (Franceschi, Xiao et al. 2003). It has been published that 14 different TFs (and/or modulatory proteins), including SATB2, HES-1, ETS-1, LEF-1, SP7 (OSTERIX), RNF11, VDR, and APC2 (a positive regulator of β-catenin of the canonical WNT-pathway) define the differentiating potential of osteoblasts with a statistical extremely high probability (p = 2α < 5x10-13) (Gorde‐ ladze, Djouad et al. 2009).

**Figure 2. Putative interactions between transcription factors (TFs), microRNA species, and functional genes de‐ scribing:** (A). Hierarchical feed-forward regulation by MYC (Myc proto-oncogene protein) and JUN (transcription fac‐ tor AP-1) of the canonical WNT-system (important for osteoblastogenesis), including factors like SMAD3 (modulator of TCF/LEF), SFRP1 (secreted frizzled-related protein 1), CCND1 (Cyclin D1), LRP5 (low density lipoprotein receptor-re‐ lated protein 5), hsa-mir-339, hsa-mir-328, and hsa-mir-149 directly, while CCND2 (cyclin D2), GSK3B (glycogen syn‐

thase kinase 3β), NKD2 (naked cuticle 2), CTNNBIP1 (catenin β1), and WNT3 (wingless-type MMTV integration site family, member 3) are affected indirectly. **(B)**. Hierarchical feed-forward regulation by ETS1 (a stimulator of CCN2 = connective tissue growth factor, a mediator of TGFβ1-mediated matrix protein synthesis in osteoblasts) and SP3 by microRNA species including hsa-mir-339, hsa-mir-149, and hsa-mir-328. The TF RNF11, which has now received in‐ creasing attention as a major modulator of osteoblastogenesis (Gao, Ganss et al. 2005), is directly affected by hsamir-149. **(C)**. Hierarchical (and possibly superior) feed-forward regulation by Sox9 of STAT1- and EGR2-mediated transcription of the micro-RNA species mir-149 and mir-328. EGR2/KROX-2 (a transcription factor stimulated by WNT3) exerts a major impact on mir-149 and mir-328, which directly affect the transcription of genes like PRRX1, AKT3, TBX2, SLC14A1, NOTCH3, FKBP5, BDNF, and DLX3 (as predicted using the MirnaViewer and Sanger algorithms. Of these genes, PRRX1 is a regulator of OSTERIX, AKT3 and RUNX2 (Lu, Beck et al. 2011) where all are mutually in‐ volved in osteoblast differentiation, TBX2 is required for osteoblast proliferation, NOTCH3 inhibits osteoblast differen‐ tiation(Lin and Hankenson 2011), FKBP5 overexpression is known to ensure osteoblast differentiation from hMSCs, BDNF is involved in endochondral ossification (Yamashiro, Fukunaga et al. 2001; Marie 2008), and DLX3 is well known as an early inducer of osteoblast differentiation co-acting with RUNX2, DLX5, and MSX2 (Heining, Bhushan et al. 2011). This particular scheme of interactions emerged when combining an osteoblast microRNA micro-array with published osteoblast mRNA transcriptomes (Gordeladze 2011). The bioinformatics modulations are performed by us‐ ing the algorithm Mir@nt@n (Le Bechec, Portales-Casamar et al. 2011) with standard stringent conditions for putative modulations of gene expression. As the reader will notice, a few other microRNA species may also have been selected to be included amongst the six microRNAs constituting the present osteochondral signature.

*Osteocytes* - During the course of bone formation, a sub-population of osteoblastic cells undergoes terminal differentiation and becomes engulfed within unmineralized osteoid. They are referred to as osteocytes (Palumbo 1986). Following the mineralization phase, these entombed cells form a network which extends throughout the mineralized bone tissue. Osteocytes respond to mechanical load (mechano-stimulation), and their network is believed to integrate the detection bone micro-damage (micro-fractures within the mineralized bone), which accumulates as a result of normal skeletal loading and fatigue (Verborgt, Tatton et al. 2002). Data have been obtained that support the idea that osteocytes initiate and direct the subsequent remodelling process that repairs damaged bone. Furthermore, the osteocytes have been hypothesized to serve as sensors for the nervous system connected to the skeleton, and for instance convey hypothalamus-derived nerve impulses driven by leptin receptors as a feedback regulatory loop controlling bone mass (i.e. bone turnover) within optimal limits (Takarada and Yoneda 2008; Thompson, Rubin et al. 2012; Motyl and Rosen 2012).

*T-cells and B-cells* - T-lymphocytes and B-lymphocytes serve as central players of the adaptive immune system facilitating the recognition and the destruction of pathogenic microorganisms. Mice, which are devoid of either B- or T-lymphocytes exhibit osteoporosis, indicating that these immune cells are routinely participating in the maintenance of bone homeostasis (Li, Toraldo et al. 2007). Mature B-cells produce ~50% of total bone marrow-derived OPG, which would contribute significantly to the restraint of osteoclast differentiation and activation during normal bone turnover. However, the role of T-cells in regulating bone remodelling during homeostasis is less clear. Based on data that T-cell-deficient CD40 knock-out and CD40L knockout mice are osteoporotic (Li, Toraldo et al. 2007), it has been proposed that T-cells work cooperatively with B-cells and enhance OPG production via CD40/CD40L co-stimulation (Raggatt and Partridge 2010)

*Osteomacs* - Osteomacs are defined as tissue macrophages residing on on or within the endosteal and/or periosteal surfaces. *In vitro*, osteomacs are required to ensure full functional differentiation of the osteoblast, i.e. the acquisition of mineralizing properties, which consti‐ tutes the late phase of the osteoblasts' life span. *In vivo*, osteomacs form a canopy over mature matrix-producing osteoblasts at sites of bone modelling, which serves as an ideal anatomical location from which to regulate such a process. A depletion of macrophages *in vivo* inevitably leads to a complete loss of endosteal osteomacs and associated osteoblasts, indicating that the osteomacs are needed to sustain osteoblast maturation and mineralizing property (Chang, Raggatt et al. 2008).

*Osteoclasts* - Osteoclasts are terminally differentiated myeloid derived cells, being adapted to excavate mineralized osteoid (bone matrix). These particular cells exhibit distinct morpholog‐ ical and phenotypic characteristics routinely used as functional markers. They include multinuclearity and expression of tartrate-resistant acid phosphatase (TRAP) and the calcito‐ nin (CT) receptor (Teitelbaum and Ross 2003). M-CSF (macrophage colony-stimulating factor) and RANK-L (receptor activator of NFκ-B ligand) serve as critical cytokines being required for the survival, expansion, and the differentiation of osteoclast precursors (hPBMCs = human peripheral blood monocytic cells) *in vitro* (Lacey, Timms et al. 1998*)*. The requirement of these cytokines for osteoclast induced bone resorption *in vivo*, has been demonstrated in animal model systems lacking functional M-CSF and RANK-L. The molecule OPG-2 (osteoproteger‐ in-2) serves as a soluble decoy receptor for RANK-L and a negative physiological regulator of osteoclastogenesis, i.e. loss of functional OPG in mice, results in osteoporosis, due to excessive induction (or differentiation) and activation of osteoclasts (Simonet, Lacey et al. 1997). The current paradigm asserts that the RANK-L/OPG expression ratio determines the degree of the osteoclasts' biological activity (i.e. their differentiation and function) (Hofbauer, Khosla et al. 2000). Several TFs, like FOS, MYC, AP-1 (JUN), PU.1, and CREB (Tondravi, McKercher et al. 1997) are deemed necessary to differentiate myeloid cells toward the osteoclast phenotype. MITF (microphthalmia-associated transcription factor) and NFATc1 (nuclear factor of activated T-cells, cytoplasmic 1), which is also important for hMSCs to develop into the osteoblastic lineage (Marie 2008) are also required for osteoclast formation and their expression of osteoclast-specific genes like TRAP (Luchin, Purdom et al. 2000), cathepsin K, and the CT receptor (Matsumoto, Kogawa et al. 2004; Hu, Sharma et al. 2007).

#### **2.3. The process of bone remodelling**

thase kinase 3β), NKD2 (naked cuticle 2), CTNNBIP1 (catenin β1), and WNT3 (wingless-type MMTV integration site family, member 3) are affected indirectly. **(B)**. Hierarchical feed-forward regulation by ETS1 (a stimulator of CCN2 = connective tissue growth factor, a mediator of TGFβ1-mediated matrix protein synthesis in osteoblasts) and SP3 by microRNA species including hsa-mir-339, hsa-mir-149, and hsa-mir-328. The TF RNF11, which has now received in‐ creasing attention as a major modulator of osteoblastogenesis (Gao, Ganss et al. 2005), is directly affected by hsamir-149. **(C)**. Hierarchical (and possibly superior) feed-forward regulation by Sox9 of STAT1- and EGR2-mediated transcription of the micro-RNA species mir-149 and mir-328. EGR2/KROX-2 (a transcription factor stimulated by WNT3) exerts a major impact on mir-149 and mir-328, which directly affect the transcription of genes like PRRX1, AKT3, TBX2, SLC14A1, NOTCH3, FKBP5, BDNF, and DLX3 (as predicted using the MirnaViewer and Sanger algorithms. Of these genes, PRRX1 is a regulator of OSTERIX, AKT3 and RUNX2 (Lu, Beck et al. 2011) where all are mutually in‐ volved in osteoblast differentiation, TBX2 is required for osteoblast proliferation, NOTCH3 inhibits osteoblast differen‐ tiation(Lin and Hankenson 2011), FKBP5 overexpression is known to ensure osteoblast differentiation from hMSCs, BDNF is involved in endochondral ossification (Yamashiro, Fukunaga et al. 2001; Marie 2008), and DLX3 is well known as an early inducer of osteoblast differentiation co-acting with RUNX2, DLX5, and MSX2 (Heining, Bhushan et al. 2011). This particular scheme of interactions emerged when combining an osteoblast microRNA micro-array with published osteoblast mRNA transcriptomes (Gordeladze 2011). The bioinformatics modulations are performed by us‐ ing the algorithm Mir@nt@n (Le Bechec, Portales-Casamar et al. 2011) with standard stringent conditions for putative modulations of gene expression. As the reader will notice, a few other microRNA species may also have been selected

*Osteocytes* - During the course of bone formation, a sub-population of osteoblastic cells undergoes terminal differentiation and becomes engulfed within unmineralized osteoid. They are referred to as osteocytes (Palumbo 1986). Following the mineralization phase, these entombed cells form a network which extends throughout the mineralized bone tissue. Osteocytes respond to mechanical load (mechano-stimulation), and their network is believed to integrate the detection bone micro-damage (micro-fractures within the mineralized bone), which accumulates as a result of normal skeletal loading and fatigue (Verborgt, Tatton et al. 2002). Data have been obtained that support the idea that osteocytes initiate and direct the subsequent remodelling process that repairs damaged bone. Furthermore, the osteocytes have been hypothesized to serve as sensors for the nervous system connected to the skeleton, and for instance convey hypothalamus-derived nerve impulses driven by leptin receptors as a feedback regulatory loop controlling bone mass (i.e. bone turnover) within optimal limits

(Takarada and Yoneda 2008; Thompson, Rubin et al. 2012; Motyl and Rosen 2012).

(Raggatt and Partridge 2010)

480 Regenerative Medicine and Tissue Engineering

*T-cells and B-cells* - T-lymphocytes and B-lymphocytes serve as central players of the adaptive immune system facilitating the recognition and the destruction of pathogenic microorganisms. Mice, which are devoid of either B- or T-lymphocytes exhibit osteoporosis, indicating that these immune cells are routinely participating in the maintenance of bone homeostasis (Li, Toraldo et al. 2007). Mature B-cells produce ~50% of total bone marrow-derived OPG, which would contribute significantly to the restraint of osteoclast differentiation and activation during normal bone turnover. However, the role of T-cells in regulating bone remodelling during homeostasis is less clear. Based on data that T-cell-deficient CD40 knock-out and CD40L knockout mice are osteoporotic (Li, Toraldo et al. 2007), it has been proposed that T-cells work cooperatively with B-cells and enhance OPG production via CD40/CD40L co-stimulation

*Osteomacs* - Osteomacs are defined as tissue macrophages residing on on or within the endosteal and/or periosteal surfaces. *In vitro*, osteomacs are required to ensure full functional differentiation of the osteoblast, i.e. the acquisition of mineralizing properties, which consti‐

to be included amongst the six microRNAs constituting the present osteochondral signature.

Prior to activation, the resting bone surface is covered with bone-lining cells, including preosteoblasts intercalated with osteomacs. B-cells are present in the bone marrow and secrete OPG, which suppresses osteoclastogenesis (Raggatt and Partridge 2010).

*Activation*: The endocrine bone-remodelling signal PTH binds to the PTH receptor on preosteoblasts. Damage to the mineralized bone matrix results in localized osteocyte apoptosis, reducing the local TGF-β concentration and its inhibition of osteoclastogenesis (Juppner, Abou-Samra et al. 1991; Heino, Hentunen et al. 2002; Swarthout, D'Alonzo et al. 2002; Verborgt, Tatton et al. 2002; Aguirre, Plotkin et al. 2006; Bonewald 2007; Weitzmann and Pacifici 2007).

*Resorption*: In response to PTH signalling, monocyte chemotactic protein (MCP-1) is released from osteoblasts and recruits pre-osteoclasts to the bone surface. Additionally, osteoblast expression of OPG is diminished, and production of M-CSF and RANK-L is increased, in order to promote proliferation of osteoclast precursors (hPBMCs) and differentiation of mature


**Figure 3.** The impact of pre-microRNA and antago-microRNA species of the osteochondral microRNA signature on the differentiation of osteochondral cells from hMSCs. The following parameters were analysed: endogenous microRNA levels, and average alteration in various differentiation markers. MicroRNA levels and marker gene mRNAs (Collagens, Gli3, Sox9, Osteocalcin, and Runx2) were analysed using the Q-PCR method (Ambion). Estimations of GAG/DNA-ratio (chondrocytes only) were performed using colorimetric methods. Mineralized surface calculations in osteoblast cul‐ tures were achieved using the Alizarin red S staining method (Gordeladze, Noel et al. 2008)

osteoclasts. The mature osteoclastic cells become anchored to RGD-binding sites (Arg-Gly-Asp containing proteins), creating a localized micro-environment (i.e. a sealed zone) facilitating the degradation of mineralized bone matrix (Partridge, Jeffrey et al. 1987; Insogna, Sahni et al. 1997; Saftig, Hunziker et al. 1998; Burgess, Qian et al. 1999; McHugh, Hodivala-Dilke et al. 2000; Teitelbaum 2000; Ma, Cain et al. 2001; Yang, Chien et al. 2004; Li, Qin et al. 2007).

*Reversal*: So called reversal cells engulf and destroy demineralised, undigested collagen from the bone surface. Transition signals are generated for the purpose of halting bone resorption and to stimulate the process of bone formation(Tran Van, Vignery et al. 1982; Kraal, Rep et al. 1987; Heinemann, Siggelkow et al. 2000; Everts, Delaisse et al. 2002; Takahashi, Takahashi et al. 2004; Newby 2008).

*Formation*: Signalling molecules stimulating bone formation arise from degraded bone matrix, mature osteoclastic cells, and potentially also reversal cells. PTH-binding to and mechanostimulation of osteocytes reduce the expression and secretion of sclerostin (an inhibitor of BMP-binding to its receptor), allowing Wnt-stimulated bone formation to occur (Fermor and Skerry 1995; Harmey, Hessle et al. 2004; van Bezooijen, Roelen et al. 2004; Li, Zhang et al. 2005; Martin and Sims 2005; Murshed, Harmey et al. 2005; Zhao, Irie et al. 2006; Pederson, Ruan et al. 2008; Robling, Niziolek et al. 2008; Tang, Wu et al. 2009).

osteoclasts. The mature osteoclastic cells become anchored to RGD-binding sites (Arg-Gly-Asp containing proteins), creating a localized micro-environment (i.e. a sealed zone) facilitating the degradation of mineralized bone matrix (Partridge, Jeffrey et al. 1987; Insogna, Sahni et al. 1997; Saftig, Hunziker et al. 1998; Burgess, Qian et al. 1999; McHugh, Hodivala-Dilke et al. 2000; Teitelbaum 2000; Ma, Cain et al. 2001; Yang, Chien et al. 2004; Li, Qin et al. 2007).

tures were achieved using the Alizarin red S staining method (Gordeladze, Noel et al. 2008)

**Figure 3.** The impact of pre-microRNA and antago-microRNA species of the osteochondral microRNA signature on the differentiation of osteochondral cells from hMSCs. The following parameters were analysed: endogenous microRNA levels, and average alteration in various differentiation markers. MicroRNA levels and marker gene mRNAs (Collagens, Gli3, Sox9, Osteocalcin, and Runx2) were analysed using the Q-PCR method (Ambion). Estimations of GAG/DNA-ratio (chondrocytes only) were performed using colorimetric methods. Mineralized surface calculations in osteoblast cul‐

*Reversal*: So called reversal cells engulf and destroy demineralised, undigested collagen from the bone surface. Transition signals are generated for the purpose of halting bone resorption and to stimulate the process of bone formation(Tran Van, Vignery et al. 1982; Kraal, Rep et al. 1987; Heinemann, Siggelkow et al. 2000; Everts, Delaisse et al. 2002; Takahashi, Takahashi et

*Formation*: Signalling molecules stimulating bone formation arise from degraded bone matrix, mature osteoclastic cells, and potentially also reversal cells. PTH-binding to and mechanostimulation of osteocytes reduce the expression and secretion of sclerostin (an inhibitor of BMP-binding to its receptor), allowing Wnt-stimulated bone formation to occur (Fermor and Skerry 1995; Harmey, Hessle et al. 2004; van Bezooijen, Roelen et al. 2004; Li, Zhang et al. 2005; Martin and Sims 2005; Murshed, Harmey et al. 2005; Zhao, Irie et al. 2006; Pederson,

Ruan et al. 2008; Robling, Niziolek et al. 2008; Tang, Wu et al. 2009).

al. 2004; Newby 2008).

482 Regenerative Medicine and Tissue Engineering

**Figure 4. A. Co-culture of chondrocytes in alginate beads (or osteoblasts in 2D culture) in standard medium (Gordeladze 2011).** Osteochondral cells were differentiated from hMSCs in appropriate medium for 14 days. Then, a dentine slice, with close to 100% enriched peripheral blood monocytes (PBMCs) had been differentiated for 14 days with RANKL-L and M-CSF, was placed adjacent to the osteochondral cells, and both cell types were co-cultured for 7 days. At the end of the experiment, the number of TRAP-positive cells was counted, and the eroded dentine surface (%) was estimated in a reflecting light microscope. **B**. Co-culture of chondrocytes in alginate beads (or osteoblasts in 2D culture) in standard medium (Gordeladze 2011). The subject experimental setting was also applied to differentiat‐ ing osteoblast. Here, we used the cytokine mixture as described in (B), and antago-microRNAs counteracting the os‐ teochondral signature. Variables analysed were much like those for the experiment with chondrocytes. However, here we measured mineralizing surface (Alizarin red S coloration) induced by the osteoblasts in a 2D culture.**C**. Co-culture of chondrocytes in alginate beads (or osteoblasts in 2D culture) in standard medium (Gordeladze 2011). The above mentioned experimental setting was used with chondrocytes differentiated from hMSCs in alginate beads in the ab‐ sence or presence of cytokines (IL-1, IL-6, IL-17, and TNFα), pre-microRNAs of all microRNA species of the osteochon‐ dral signature, or the set of cytokines along with the pre-microRNAs. Parameters analysed were: Q-PCR of transcription factors (TFs) and marker genes expressed by the chondrocytes, clinical score of alginate beads containing chondrocytes embedded in newly synthesized proteoglycan matrix, the number of Trap-positive osteoclasts, and the eroded dentine surface (%).

*Termination*: Sclerostin expression then fades or ceases, and the process of bone formation comes to a halt. The newly deposited osteoid is subsequently mineralized, the bone surfaces return to their resting state, This is when bone-lining cells intercalate with osteomacs, and the remodelling cycle is concluded (Raggatt and Partridge 2010).
