Characterization of bone-derived chondrogenesis-stimulating activity on embryonic limb mesenchymal cells in vitro

Abstract. Demineralized bone matrix contains factors which stimulate chondrogenesis and osteogenesis in vivo. A water-soluble extract of bone has been shown to stimulate chondrogenesis in vitro in embryonic limb mesenchymal cells (Syftestad, Lucas & Caplan, 1985). The aim of this study was to analyse the cellular mechanism of the bone-derived chondrogenesis-stimulating activity, with particular attention on how normal requirements for chondrogenesis may be altered. The effects of bovine bone extract (BBE) on chondrogenesis in vitro were studied using micromass cultures of chick limb bud mesenchyme isolated from embryos at Hamburger-Hamilton (HH) stage 23/24, an experimental system which is capable of undergoing chondrogenic differentiation. Bovine diaphyseal long bones were demineralized and extracted with guanidine-HCl to prepare BBE (Syftestad & Caplan, 1984). High-density mesenchyme cultures (30 times 10⁶ cells/ml) were exposed to different doses of BBE (0–01-1-0 mg ml^-1) and chondrogenesis was quantified based on cartilage nodule number and [³⁵S]sulphate incorporation. BBE was tested on micromass cultures of varying plating densities (2–30 times 10⁶ cells/ml), on cultures of ‘young’ limb bud cells (HH stage 17/18), and on cultures enriched with chondroprogenitor cells obtained from subridge mesoderm. Since poly-L-lysine (PL) has recently been shown (San Antonio & Tuan, 1986) to promote chondrogensis, PL and BBE were introduced together in different doses, in the culture medium, to determine if their actions were synergistic. Our results show that BBE stimulates chondrogenesis in a dose-dependent manner and by a specific, direct action on the chondroprogenitor cells but not in normally non-chondrogenic, low density or ‘young’ limb bud cell cultures. The effects of PL and BBE are additive and these agents appear to act by separate mechanisms to stimulate chondrogenesis; PL primarily enhances nodule formation, and BBE appears to promote nodule growth.     

N-Cadherin expression and signaling in limb mesenchymal chondrogenesis: stimulation by poly-L-lysine

Cellular condensation is a requisite step in the initiation of mesenchymal chondrogenesis in the embryonic limb bud. We have previously shown that cellular condensation of limb chondroprogenitor mesenchymal cells is accompanied by elevated expression of N-cadherin during chondrogenesis both in vivo and in vitro. N-Cadherin-mediated cell-cell interaction is also functionally required for proper mesenchymal chondrogenesis both in vivo and in vitro. In this report, we have further analyzed the functional importance of N-cadherin in the cellular condensation-chondrogenesis pathway by examining N-cadherin expression and related activities in high density micromass cultures of chick limb mesenchymal cells in which chondrogenesis is being stimulated with the cationic polymer, poly-L-lysine (PL). The chondrogenesis-promoting action of PL is thought to involve the clustering of cells via ionic cross-linking, perhaps mimicking the action of an endogenous matrix component. Immunohistochemistry, immunoblotting, and Northern blot analysis all show that PL treatment results in a time-dependent increase in N-cadherin expression at both the protein and mRNA levels. In addition, inhibition of N-cadherin function with a neutralizing monoclonal antibody directed to its extracellular domain inhibits the chondrogenesis-stimulating effect of PL. PL treatment also alters the tyrosine-phosphorylation state of the N-cadherin associated signaling protein, beta-catenin. These results suggest that N-cadherin-mediated cell adhesion is a requisite regulatory component of the limb mesenchymal chondrogenic differentiation program, involving at least in part beta-catenin tyrosine phosphorylation as a signaling step.     

Polyionic regulation of cartilage development: promotion of chondrogenesis in vitro by polylysine is associated with altered glycosaminoglycan biosynthesis and distribution.

The development of cartilage nodules in cultures of chick limb bud mesenchyme (Hamburger-Hamilton stages 23/24) is significantly promoted when the culture medium is supplemented with (poly-L-lysine (PL) (M(r) greater than or equal to 14K) (San Antonio and Tuan, 1986. Dev. Biol. 115: 313). Here we present findings consistent with the hypothesis that PL may promote chondrogenesis by interacting electrostatically with sulfated glycosaminoglycans (GAGs): (1) poly-L-ornithine, poly-L-histidine, poly-D,L-lysine, and lysine-containing heteropolypeptides stimulate chondrogenesis in proportion to their contents of cationic residues; (2) the effects of PL are diminished when limb mesenchyme cultures are supplemented with exogenous GAGs, including heparin, dermatan sulfate, and chondroitin sulfate; (3) in high density cultures of limb bud mesenchyme, the release of sulfated macromolecules, but not of proteins in general, into the culture medium was significantly inhibited by PL (398K M(r)) treatment, and a net increase in total GAG content of the PL-treated cultures was observed; and (4) in monolayer cultures of cells derived from other chick embryonic tissues, including liver, skeletal muscle, and calvaria, PL treatment promoted the cell layer-associated retention of sulfated GAG. These effects were not observed using the nonstimulatory, low M(r) PL (4K). Based on the above findings and those from previous studies, it is proposed that PL may promote chondrogenesis by interacting electrostatically with cartilage GAGs, thus trapping the extracellular matrix around the newly emerging cartilage nodules and thereby stabilizing their growth and differentiation.     

Sulfoxide stimulation of chondrogenesis in limb mesenchyme is accompanied by an increase in type II collagen enhancer activity

We have utilized a modification of the limb bud mesenchyme micromass culture system to screen compounds that might stimulate chondrogenesis. Two compounds in the sulfoxide family (methylphenylsulfoxide and p-chlorophenyl methyl sulfoxide) were stimulatory at 10(-2) M and 10(-3) M, respectively; whereas other sulfoxides and organic solvents were not active at these concentrations. In addition, specific growth factors (basic FGF, IGF-I, IGF-II) were not chondroinductive at concentrations that are active in other cell systems. Both sulfoxide compounds stimulated cartilage nodule formation, [35S]sulfate incorporation, and activity of the regulatory sequences of the collagen II gene. In contrast, transforming growth factor beta-1 (10 ng/ml) stimulated sulfate incorporation but produced only a diffuse deposition of cartilage matrix and reduced the ability of the cells to utilize the regulatory sequences of the collagen II gene. The sulfoxides appear to promote the differentiation of limb bud cells to chondrocytes and thus exhibit chondroinductive activity.     

Effect of TGF-beta 1, TGF-beta 2, and bFGF on chick cartilage and muscle cell differentiation

In insulin containing defined medium TGF-beta 1, TGF-beta 2, and bFGF all stimulate chondrogenic differentiation in high-density micromass cultures of distal limb bud mesenchyme cells of chick embryos. In addition bFGF inhibits myogenic differentiation, while TGF-beta 1 and TGF-beta 2 appear to have no effect. TGF-beta 1 and bFGF together act additively to enhance chondrogenesis, while TGF-beta blocks the bFGF inhibitory action on myoblasts, thus allowing them to differentiate. In the absence of insulin, the inhibitory effect of bFGF on muscle cell differentiation is reduced; cartilage differentiation in the presence of the above growth factors is also slightly reduced.     

Cell-cell interaction by mouse limb cells during in vitro chondrogenesis: Analysis of the brachypod mutation

This paper contains observations and experiments which collectively demonstrate a requirement for cell-cell interactions among limb bud mesenchyme cells during chondrogenic differentiation. Limb bud cells isolated from brachypodism^H (bp^H) and wild-type mouse embyros between Thieler stage 16–17 and midstage 21 were compared with respect to their abilities to undergo chondrogenic differentiation in high-density micromass cultures. Nodules formed by dissociated Day 12 (stage 20) bp^H limb bud cells have been reported previously to be abnormally reduced in size and number, and delayed in formation. We corroborate these results, but find that bp^H cultures prepared from earlier-stage limb buds (between stages 16–17 and early stage 21) are progressively more like wild-type cultures. Stage 16–17 bp^H cultures at 72 hr actually contain normal numbers of and size nodules, while stage 18 bp^H cultures are intermediate between stages 16–17 and stage 21 in nodule formation. On the other hand, we also find that the initial rate of aggregate formation is normal even in bp^H cultures prepared from stage 20 cultures in which nodule formation is not normal. Preparation of cultures composed primarily of early stage 21 bp^H limb bud cells mixed with small quantities (e.g., 5%) of stage 16–17 wild-type limb bud cells showed significant increases in cartilage nodule formation over control cultures composed only of early stage 21 bp^H cells. Greater proportions of wild-type cells obtained from embryos older than stages 16–17 were required for the same degree of normalization, supporting the hypothesis that a specific cell type, whose proportion decreases normally in the limb bud over time, is required to increase in vitro chondrogenesis by bp^H cells. Additionally, cultures containing stage 23 chick limb cells and early stage 21 bp^H cells at a ratio of 1:20 contained wild-type levels of nodules per square millimeter of culture. Thus, bp^H cells appear to respond to chondrogenic inductive signals from normal limb mesenchyme cells. In order to test for the ability of bp^H limb bud mesenchyme to induce chondrogenesis, stage 16–17 bp^H and wild-type limb bud cells, which form identical numbers of aggregates and nodules in culture, were each mixed with early stage 21 bp^H cells at ratios of 1:20, 1:10, and 1:3. Although low proportions of wild-type stage 17 cells significantly increased the number of aggregates and nodules in these mixed cultures, low proportions of bp^H stage 16–17 cells did not. It is, therefore, concluded that the primary defect of the bp^H mutation is likely to reside in the reduced ability of a specific mesenchyme cell subpopulation to provide an inductive stimulus for chondrogenesis.     

Distinct functions of BMP4 and GDF5 in the regulation of chondrogenesis

Bone morphogenetic protein 4 (BMP4) and growth/differentiation factor 5 (GDF5) are closely related protein family members and regulate early cartilage patterning and differentiation. In this study, we compared the functional outcome of their actions systematically at various stages of chondrogenesis in mouse embryonic limb bud mesenchyme grown in micromass cultures. Overall, both growth factors enhanced cartilage growth and differentiation in these cultures. Uniquely, BMP4 not only accelerated the formation and maturation of cartilaginous nodules, but also induced internodular mesenchymal cells to express cartilage differentiation markers. On the other hand, GDF5 increased the number of prechondrogenic mesenchymal cell condensation and cartilaginous nodules, without altering the overall pattern of differentiation. In addition, GDF5 caused a more sustained elevated expression level of Sox9 relative to that associated with BMP4. BMP4 accelerated chondrocyte maturation throughout the cultures and sustained an elevated level of Col10 expression, whereas GDF5 caused a transient increase in Col10 expression. Taken together, we conclude that BMP4 is instructive to chondrogenesis and induces mesenchymal cells toward the chondrogenic lineage. Furthermore, BMP4 accelerates the progression of cartilage differentiation to maturation. GDF5 enhances cartilage formation by promoting chondroprogenitor cell aggregation, and amplifying the responses of cartilage differentiation markers. These differences may serve to fine-tune the normal cartilage differentiation program, and can be exploited for the molecular manipulation in biomimetics.     

Cell sorting and chondrogenic aggregate formation in micromass culture

A fundamental feature of cartilage differentiation in the developing limb is the formation of a prechondrogenic cell condensation. An apparently similar process of prechondrogenic cell aggregation occurs in micromass cultures of limb bud mesenchyme with the formation of cellular aggregates which often differentiate into cartilage nodules. We have investigated the process of aggregate formation in micromass culture using chimaeric mixtures of potentially chondrogenic and nonchondrogenic cell types. Two systems were studied: mixtures of distal and proximal limb mesenchyme cells and mixtures of distal limb cells with avian tendon fibroblasts. In both cases cultures of varying proportions of each cell type have been prepared. The results demonstrate that aggregate formation in vitro is the consequence of a cell sorting process which can involve prechondrogenic cells of widely different spatial origins within the developing limb. This contrasts with in vivo prechondrogenic condensation in which there is no evidence of cell sorting (Searls, R.L. (1967), J. Exp. Zool. 166, 39-50). However, our findings do indicate that cell surface differences occur in apparently undifferentiated limb mesenchyme. The results also suggest that mesenchymal cell aggregates must achieve a threshold size before chondrogenesis can proceed. In addition, the results show that under some culture conditions nonchondrogenic cells will form aggregates.     

Chondrogenesis of limb bud mesenchyme in vitro: stimulation by cations

To analyze the nature of cell-cell interactions in chondrogenesis, two cations that influence these interactions, calcium and poly-L-lysine (PL), were tested for their effects on chondrogenesis in vitro. High density cultures of chick limb bud mesenchyme (Hamilton-Hamburger stages 23/24), were exposed to culture media containing calcium (0.6-3.3 mM) or PL (1-10 micrograms/ml). Both cations stimulated chondrogenesis in a dose-dependent manner, and also promoted cartilage formation in normally non-chondrogenic, low cell density cultures. Chondrogenesis was assayed based on cartilage nodule number, [35S]sulfate incorporation, and expression of type II collagen as detected by immunohistochemistry. The calcium effect was not mimicked by other divalent cations (Cd, Co, Ni, Mg, Mn, and Sr). The effect of PL was dependent on its Mr (greater than or equal to 14K) and charge, and was mimicked by poly-D-lysine but not by lysine or other analogs of PL or lysine (epsilon-amino caproic acid, lysozyme, poly-L-arginine, and spermidine). Calcium and PL probably act by different mechanisms since their effects were additive, and required their presence on different days of culture: calcium acted on Day 1, and PL on Day 2. It is proposed that calcium may play a role in the cell aggregation phase of chondrogenesis whereas PL, or a naturally occurring polypeptide of similar nature, may promote chondrogenesis by crosslinking specific anionic components of the cell surface or extracellular matrix.     

Exogenous glycosaminoglycans modulate chondrogenesis, cyclic AMP level and cell growth in limb bud mesenchyme cultures

Effects of hyaluronate, heparin and chondroitin-6-sulfate were studied on micromass cultures of chick limb bud mesenchyme (Hamburger and Hamilton stages 23-24). Histochemical, electron microscopical, biochemical and radiochemical investigations of day 4 cultures revealed dose-dependent inhibitory effects of these glycosaminoglycans on chondrogenesis, cyclic AMP level and growth of cells. In addition, hyaluronate with 100 micrograms/ml dose caused a displacement of newly formed proteoglycan from cultures into the medium. It is supposed that exogenous glycosaminoglycans influence ionic equilibrium in the immediate vicinity of cells and disturb the organization of the prechondrogenic extracellular matrix resulting in alterations of cell membrane--cytoskeleton associations. These alterations may provoke a reduction in cyclic AMP level and DNA synthesis. It is suggested that a reduction in cyclic AMP level preceding the expression of cartilage phenotype results in the inhibition of chondrogenesis.     

Evidence for histogenic interactions during in vitro limb chondrogenesis

The requirement for homotypic cell interaction was studied by making chimeric micromass cultures containing various proportions of chick and quail limb mesenchyme. Cultures made from limb mesenchyme from embryos of Hamburger and Hamilton stages 23–24 produce large clumps of cartilage cells, identified by the accumulation of an extracellular matrix which stains with alcian blue at pH 1 and by the ability of cells to take up ³⁵SO₄ rapidly, as demonstrated autoradiographically. Dissociated mesenchyme from stage 19 embryos did not produce cartilage in micromass cultures, but only precartilage cell aggregates. Micromass cultures prepared from mixtures of mesenchyme cells obtained from stage 19 and stages 23–24 embryos contained decreasing numbers of cartilage nodules as the proportion of stage 19-derived mesenchyme increased. At the same time the number of aggregates was not affected. When the ratio of stage 19- to stage 24-derived cells was 3:1 or greater, no nodules were detected. The actual number of cells from each stage was verified by using mixtures of quail and chick cells, which are microscopically distinguishable. Additional evidence suggests that the stage 19-derived mesenchyme inhibits chondrogenesis by passively preventing stage 24-derived cells from interacting. The results presented are consistent with the suggestions that (1) homotypic cell interaction plays a role in limb chondrogenesis and (2) the capacity to interact in the required manner is acquired after the embryos have reached stage 19. These phenomena might be involved in the normal histogenesis of cartilage tissue.     

Involvement of Frzb-1 in mesenchymal condensation and cartilage differentiation in the chick limb bud

In developing limb bud, mesenchymal cells form cellular aggregates called "mesenchymal condensations". These condensations show the prepattern of skeletal elements of the limb prior to cartilage differentiation. Roles of various signaling molecules in chondrogenesis in the limb bud have been reported. One group of signaling factors includes the Wnt proteins, which have been shown to have an inhibitory effect on chondrogenesis in the limb bud. Therefore, regulation of Wnt activity may be important in regulating cartilage differentiation. Here we show that Frzb-1, which encodes a secreted frizzled-related protein that can bind to Wnt proteins and can antagonize the activity of some Wnts, is expressed in the developing limb bud. At early stages of limb development, Frzb-1 is expressed in the ventral core mesenchyme of the limb bud, and later Frzb-1 expression becomes restricted to the central core region where mesenchymal condensations occur. At these stages, a chondrogenic marker gene, aggrecan, is not yet expressed. As limb development proceeds, expression of Frzb-1 is detected in cartilage primordial cells, although ultimately Frzb-1 expression is down-regulated. Similar results were obtained in the recombinant limb bud, which was constructed from dissociated and re-aggregated mesenchymal cells and an ectodermal jacket with the apical ectodermal ridge. In addition, Frzb-1 expression preceded aggrecan expression in micromass cultures. These results suggest that Frzb-1 has a role in condensation formation and cartilage differentiation by regulating Wnt activity in the limb bud.     

The differentiation of normal and muscle-free distal chick limb bud mesenchyme in micromass culture

Distal chick wing bud mesenchyme from stages 19 to 27 embryos has been grown in micromass culture. The behavior of cultures comprising mesenchyme located within 350 microns of the apical ectodermal ridge (distal zone mesenchyme) was compared to that of cultures of the immediately proximal mesenchyme (subdistal zone cultures). In cultures of the distal mesenchyme from stages 21-24 limbs, all of the cells stained immunocytochemically for type II collagen within 3 days, indicating ubiquitous chondrogenic differentiation. At stage 19 and 20, this behavior was only observed in cultures of the distal most 50-100 microns of the limb bud mesenchyme. Between stages 25 and 27, distal zone cultures failed to become entirely chondrogenic. At all stages, subdistal zone cultures always contained substantial areas of nonchondrogenic cells. The different behavior observed between distal zone and corresponding subdistal zone cultures appears to be a consequence of the presence of somite-derived presumptive muscle cells in the latter, since no such difference was observed in analagous cultures prepared from muscle-free wing buds. The high capacity of the distal zone for cartilage differentiation supports a view of pattern formation in which inhibition of cartilage is an important component. However, its consistent behavior in vitro indicates that micromass cultures do not reflect the in vivo differences between the distal zones at different stages. The subdistal region retains a high capacity of cartilage differentiation and the observed behavior in micromass reflects interactions with a different cell population.     

High density micromass cultures of embryonic limb bud mesenchymal cells: An in vitro model of endochondral skeletal development

Summary To study the mechanisms regulating endochondral skeletal development, we examined the characteristics of long-term, high density micromass cultures of embryonic chicken limb bud mesenchymal cells. By culture Day 3, these cells underwent distinct chondrogenesis, evidenced by cellular condensation to form large nodules exhibiting cartilage-like morphology and extracellular matrix. By Day 14, extensive cellular hypertrophy was seen in the core of the nodules, accompanied by increased alkaline phosphatase activity, and the limitation of cellular proliferation to the periphery of the nodules and to internodular areas. By Day 14, matrix calcification was detected by alizarin red staining, and calcium incorporation increased as a function of culture time up to 2 to 3 wk and then decreased. X-ray probe elemental analysis detected the presence of hydroxyapatite. Analogous to growth cartilage developing in vivo, these cultures also exhibited time-dependent apoptosis, on the basis of DNA fragmentation detected in situ by terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate (dUTP) nick end labeling (TUNEL), ultrastructural nuclear morphology, and the appearance of internucleosomal DNA degradation. These findings showed that cellular differentiation, maturation, hypertrophy, calcification, and apoptosis occurred sequentially in the embryonic limb mesenchyme micromass cultures and indicate their utility as a convenient in vitro model to investigate the regulatory mechanisms of endochondral ossification.     

Phorbol esters inhibit chondrogenesis in limb mesenchyme by mechanisms independent of PGE2 or cyclic AMP1

Effects of the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) on chondrogenesis and concentrations of prostaglandin E2 (PGE2) and cyclic AMP (cAMP) were investigated in micromass cultures of chick limb mesenchyme derived from the distal tip of stage 25 limb buds. TPA completely inhibited chondrogenesis during the first 4 days of culture; however, a few small cartilage nodules formed by day 6. Relative to control cultures, both PGE2 and cAMP concentrations were altered by TPA treatment during the 6-day period of cell culture. Concentrations of both compounds increased in control cells during the first 24 h of culture and then declined during the remaining 5 days. In TPA-treated cells both PGE2 and cAMP levels increased progressively during the 6 days of days of cell culture, each being elevated at day 6 by twofold over control cells. The results suggest the presence of regulatory pathways important in chondrogenesis which occur independent of those initiated by PGE2 and the cAMP system.