Miltenyi Biotec GmbH
Originally published online as doi:10.1189/jlb.0708402 on March 31, 2009

Published online before print March 31, 2009
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Figure 1
Right arrow All Versions of this Article:
jlb.0708402v1
85/6/919    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Vries, T. J.
Right arrow Articles by Everts, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Vries, T. J.
Right arrow Articles by Everts, V.
(Journal of Leukocyte Biology. 2009;85:919-927.)
© 2009 by Society for Leukocyte Biology

Myeloid blasts are the mouse bone marrow cells prone to differentiate into osteoclasts

Teun J. de Vries*,{dagger},1, Ton Schoenmaker*,{dagger}, Berend Hooibrink{ddagger}, Pieter J. M. Leenen§ and Vincent Everts*,{dagger}

Departments of
* Periodontology and
{dagger} Oral Cell Biology, Academic Centre for Dentistry Amsterdam (ACTA), University of Amsterdam and VU University Amsterdam, Research Institute MOVE, and
{ddagger} Department of Cell Biology and Histology, Academic Medical Center (AMC), University of Amsterdam, The Netherlands; and
§ Department of Immunology, Erasmus MC, University Medical Center, Rotterdam, The Netherlands

1 Correspondence: Experimental Periodontology, c.o. Oral Cell Biology, Academic Centre for Dentistry Amsterdam (ACTA), University of Amsterdam and VU University Amsterdam, Van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands. E-mail: teun.devries{at}vumc.nl

ABSTRACT

Cells of the myeloid lineage at various stages of maturity can differentiate into multinucleated osteoclasts. Yet, it is unclear which developmental stages of this lineage are more prone to become osteoclasts than others. We investigated the osteoclastogenic potential of three successive stages of myeloid development isolated from mouse bone marrow. Early blasts (CD31hi/Ly-6C–), myeloid blasts (CD31+/Ly-6C+), and monocytes (CD31–/Ly-6Chi), as well as unfractionated marrow cells, were cultured in the presence of M-CSF and receptor activator of NF-{kappa}B ligand (RANKL), and the differentiation toward multinucleated cells and their capacity to resorb bone was assessed. Myeloid blasts developed rapidly into multinucleated cells; in only 4 days, maximal numbers were reached, whereas the other fractions required 8 days to reach maximal numbers. Bone resorption was observed after 6 (myeloid blasts and monocyte-derived osteoclasts) and 8 (early blast-derived osteoclasts) days. This difference in kinetics in osteoclast-forming capacity was confirmed by the analysis of osteoclast-related genes. In addition, the myeloid blast fraction proved to be most sensitive to M-CSF and RANKL, as assessed with a colony-forming assay. Our results show that osteoclasts can develop from all stages of myeloid differentiation, but myeloid blasts are equipped to do so within a short period of time.

Key Words: osteoclastogenesis • CD31 • Ly-6C • heterogeneity

INTRODUCTION

Since 1975, it is recognized that multinucleated osteoclasts are cells that form from monocyte/macrophage-lineage cells [1 2 3 4 ]. It appeared that osteoclasts could be derived from a number of cell types belonging to the myeloid differentiation pathway. Osteoclasts can be generated not only from bone marrow precursors [4 ] but also from blood-borne monocytes [5 , 6 ], splenocytes [2 ], peritoneal macrophages [5 ], and dendritic cells (DC) [7 , 8 ]. In the context of bone remodeling, some of these sources of osteoclast precursors may not be physiological, as osteoclasts are formed in the proximity of the bone surface, in an environment relatively rich in hematopoietic stem cells. In this context, little is known about what cell type of myeloid differentiation is most primed for osteoclast differentiation. Attempts to characterize osteoclast precursor cells in bone marrow have used fractionation with markers CD11b and Gr-1/Ly-6G, which are also expressed on granulocytes [9 ]. CD11b may not be the ideal marker to isolate osteoclast precursors, as two studies showed that CD11b+ and CD11b– cells are able to differentiate into osteoclasts [9 , 10 ].

To determine what cell type of the myeloid differentiation is prone to differentiate into osteoclasts, we fractionated mouse bone marrow into phenotypically defined, successive stages of myeloid development using CD31 (ER-MP12) and Ly-6C (ER-MP20) expression profiles [11 , 12 ]. These fractions encompass early blasts (CD31hi/Ly-6C–), myeloid blasts (CD31+/Ly-6C+), and monocytes (CD31–/Ly-6Chi) and have been shown to be the only fractions responsive to M-CSF [11 ] and GM-CSF [12 ]. Such stimulation leads to the development of macrophages and DC, respectively. Macrophage as well as DC maturation followsthe sequence of early blasts, myeloid blasts, monocytes, and subsequently, macrophages [11 ] or DC [12 ], resulting in increased expression of macrophage markers CD11b/macrophage-antigen 1 and F4/80 [11 ] or DC markers MHC class II and CD11c [12 ], respectively. In various studies, the present distinction of myeloid precursor stages based on CD31/Ly-6C expression was validated and used to investigate myeloid development under different experimental conditions [11 ]. Using these markers, bone marrow can be fractionated accurately, enabling the isolation of cell fractions relatively devoid of lymphocytic or granulocytic cells [11 ], whereas such cell types were possible contaminants in studies where CD11–/Gr-1– [9 ] or CD11b–/lo/CD45R– [10 ] cells were studied as osteoclast precursors.

Here, we have investigated whether any of these defined, successive myeloid differentiation stages distinguish themselves as a primary source for osteoclast differentiation.

MATERIALS AND METHODS

Bone marrow isolation
Permission for animal experiments described here was obtained from the Animal Welfare Committee of the VU University (Amsterdam, The Netherlands). Bone marrow from 6-week-old male C57BL/6J mice (Harlan, Horst, The Netherlands) was isolated from tibiae after crushing with a mortar. The isolation procedure was described elsewhere in detail [13 ]. The cell suspension was aspirated through a 21-gauge needle and filtered over a 100-µm pore-size cell-strainer filter (Falcon/Becton Dickinson, Franklin Lakes, NJ, USA). Until plating, all cell suspensions were kept on ice.

Immunofluorescence labeling, flow cytometry, and sorting
All immunofluorescent labelings and washes took place in PBS containing 1% BSA. Bone marrow cell suspensions were spun down (200 g, 10 min, 4°C) and incubated for 30 min in 25 µl biotinylated ER-MP12, recognizing CD31 [11 ]/106 cells, which were washed once and incubated in 25 µl/106 cells FACS buffer containing FITC-conjugated ER-MP20, recognizing Ly-6C [11 ] and streptavidin-PE conjugate (Becton Dickinson, San Jose, CA, USA; 10 µl/106 cells) for 30 min. Cells were washed and recovered in culture medium. Cells were sieved through 50 µm filters (filcons, Becton Dickinson) before cell sorting. Early blasts (CD31hi/Ly-6C–), myeloid blasts (CD31+/Ly-6C+), and monocytes (CD31–/Ly-6Chi) were sorted at 3 x 107 cells/h on FACSAria (Becton Dickinson). A typical profile is shown (see Fig. 1 ). Similar numbers of tartrate-resistant acid phosphatase (TRACP)-positive, multinucleated cells were found after a culture period of 6 days (unlabeled: 46±12; labeled and sorted: 50±13, obtained from 1.0x105 bone marrow cells; mean±SEM; n=6 mice), indicating that cell labeling and sorting did not affect osteoclastogenesis.


Figure 1
View larger version (40K):
[in this window]
[in a new window]

  		Figure 1. Two-color flow cytometric analysis of murine bone marrow. Cells were labeled with ER-MP12 (CD31) and ER-MP20 (Ly-6C). Percentages of cells found per gated area are indicated (n=4 mice from two experiments; mean±SD). Bold, Subpopulations highly enriched in early blasts (1), myeloid blasts (4), and monocytes (6). Other fractions mainly contain lymphocytes (2), erythroid blasts (3), or granulocytes (5) as shown in ref. [11 ].

Osteoclastogenesis
Sorted cells were plated in 96-well flat-bottom tissue culture-treated plates (Costar, Cambridge, MA, USA) at a density of 1.5 x 104 cells/well in 150 µl culture medium containing 30 ng/ml recombinant murine (rm)M-CSF (R&D Systems, Minneapolis, MN, USA), with or without 20 ng/ml rm-receptor activator of NF-{kappa}B ligand (RANKL; RANKL-TEC, R&D Systems). In other experiments, cells were seeded on 650 µm thick bovine cortical bone slices. Culture media were replaced every 3 days. At the end of the culture period, wells were washed with PBS and fixed in 4% PBS-buffered formaldehyde and stored at 4°C (used for TRACP staining) or in water (used for bone resorption) or dissolved in RNA lysis buffer from the RNeasy Mini Kit (Qiagen, Hilden, Germany) and stored at –80ºC (for RNA isolation).

In some experiments, cells were cultured for 3 days in the presence of M-CSF and RANKL, detached with cell dissociation solution (Sigma-Aldrich, St. Louis, MO, USA), and subsequently, recultured at a density of 104 cells for another 24 h in the presence of the cytokines.

Bone resorption
Sorted cells (1.5x104) were cultured on bone slices with M-CSF and RANKL. Bone resorption was assessed as described before [13 ]; resorption pits were visualized by using Coomassie Brilliant blue. The surface areas of individual resorption pits were measured using Image-Pro Plus software (MediaCybernetics, Silver Spring, MD, USA).

RNA analysis and real-time quantitative PCR (qPCR)
RNA isolation and real-time qPCR were performed as described in detail [14 ]. Real-time PCR primers were designed using the Primer Express software, Version 2.0 (Applied Biosystems, Foster City, CA, USA; Table 1 ). To avoid amplification of genomic DNA, each amplicon spanned at least one intron. For the external standard curve used in the PCR reactions, cDNA was used from bone marrow, which was cultured for 4 days with M-CSF + RANKL. The PCR reactions of the different amplicons had equal efficiencies.


View this table:
[in this window]
[in a new window]

  		Table 1. qPCR Primers Used

PBGD was used as the housekeeping gene [15 ]. Expression of this gene was not affected by the experimental conditions. Samples were normalized for the expression of PBGD by calculating the {Delta} comparative threshold (Ct; CtPBGD–Ctgene of interest), and expression of the different genes is expressed as 2–({Delta}Ct).

CFU assay
To determine the number of cells per fraction that proliferated in the presence of M-CSF or M-CSF + RANKL, 2 x 104 total bone marrow cells or 103 early blasts, myeloid blasts, or monocytes were seeded in 1 ml 1% methocult medium (M3134, Stem Cell Technologies, Vancouver, Canada), supplemented with 60 ng/ml M-CSF and 40 ng/ml RANKL [13 ]. The number of colonies (>10 cells/colony) was assessed after 8, 11, and 14 days of culturing. Size of colonies was measured at 11 days using a computerized XY tablet and Image-Pro Plus software (MediaCybernetics). At the end of the culture period, methocult was washed gently for several times with water, upon which a TRACP staining was performed. Colonies were scored as TRACP-negative or as colonies containing TRACP-positive cells.

[3H]Thymidine labeling
To determine the effect of M-CSF or the combination of M-CSF + RANKL on proliferation at the time-point where multinucleation was first observed, unfractionated bone marrow cells, early blasts, myeloid blasts, and monocytes were cultured in the presence of 30 ng/ml M-CSF or 30 ng/ml M-CSF + 20 ng/ml RANKL, and after 3 days, [3H]thymidine (7.4 MBq/well, Amersham Biosciences, Uppsala, Sweden) was added to the culture for 24 h. Excess label was washed away, and incorporation was measured.

Statistical analysis
Bone marrow cell sorts were from three mice per experiment. All data shown were from two pooled experiments (n=6). The Kruskal-Wallis nonparametric variance test (one-way ANOVA) followed by Tukey-Kramer’s multiple comparison test was used when multiple comparisons were made. Differences between groups were considered significant at P < 0.05 (two-tailed).

RESULTS

Earliest osteoclast formation from myeloid blasts
Based on the CD31 and Ly-6C antigen expression, six bone marrow fractions (Fig. 1) were isolated and cultured with M-CSF and RANKL for 4 or 6 days. No TRACP-positive, multinucleated cells were obtained from Fractions 2 (enriched in lymphoid cells), 3 (enriched in erythroid cells), or 5 (enriched in granulocytes) at any time-point analyzed (data not shown). TRACP-positive, multinucleated cells were only found to originate from Fractions 1 (enriched in early blasts), 4 (enriched in myeloid blasts), and 6 (enriched in monocytes).

Formation of osteoclast-like cells proved to be cell density-dependent (Fig. 2B ). Cells (15x103)/fraction/well as seeding density were used throughout the remaining analyses. To investigate putative differences in kinetics of osteoclast formation potential between the various fractions, M-CSF/RANKL-stimulated cultures were analyzed at different time-points (Fig. 2) . At 2 days of culture, all cells deriving from early blasts were TRACP-negative, whereas TRACP-positive, mononuclear cells were detectable primarily in myeloid blast cultures and to a lesser extent, in monocyte cultures. No osteoclast-like cells could be detected during the first 3 days in any of the cultures. However, highest numbers of osteoclast-like cells were formed already at Day 4 from myeloid blasts (Fig. 2 A, C, and F) . At least three times more osteoclast-like cells were formed from this fraction compared with the other fractions (P<0.001 myeloid blast vs. bone marrow or early blasts or monocytes, Fig. 2C ). Formation of TRACP-positive, multinucleated cells from myeloid blasts peaked at 4 days (Fig. 2F) , followed by a smaller peak at 8 days. This secondary peak is possibly a result of fusion of myeloid blast progeny with smaller multinucleated cells (three to five nuclei), as particularly this latter category dropped between 4 and 8 days (106±17 vs. 37±5; P=0.0004; n=6; mean±SD). As a result, larger osteoclast-like cells (greater than five nuclei) appeared at the expense of smaller osteoclasts at Day 8. For the other fractions, the highest numbers of osteoclast-like cells were present at Day 8 in total bone marrow (Fig. 2D) , early blast (Fig. 2E) , and monocyte cultures (Fig. 2G) . Extraordinary large, multinucleated cells with a diameter of almost 1 mm and containing many nuclei arose after prolonged culture (10 days) of myeloid blasts and monocytes but not in early blast cultures (Fig. 2A , lower panel).


Figure 2
View larger version (39K):
[in this window]
[in a new window]

  		Figure 2. Osteoclast formation from myeloid bone marrow fractions on tissue-culture plates. Unfractionated bone marrow, early blasts, myeloid blasts, and monocytes were cultured with M-CSF and RANKL for 4–10 days in 96-well plates. (A) Examples of TRACP-positive, multinucleated cells (greater than two nuclei) formed after 4 days (upper panel) and 10 days (lower panel) cultured on tissue-culture plates. Original bars = 100 µm. (B) Formation of TRACP-positive, multinucleated cells proved to increase with increasing seeding densities, 3.8, 7.5, 15, and 30 x 103 cells/well, which were plated in the presence of M-CSF and RANKL, and multinucleated cells (MNCs) were assessed at 4 days. From here, 15 x 103 cells were plated in the remaining experiments. eb, Early blasts; mb, myeloid blasts; mo, monocytes. (C) At least three times more TRACP-positive, multinucleated cells formed from myeloid blasts at 4 days. ***, P < 0.001. (D–G) Maximal numbers of osteoclasts formed after 4 days (myeloid blasts, F) or 8 days (total bone marrow, D; early blasts, E; monocytes, G). *, P < 0.05; **, P < 0.01. (H and I) Presence of TRACP+ multinucleated cells at 4 days (greater than two nuclei, H; greater than five nuclei, I) in M-CSF- and RANKL-treated cultures, where cells were replated in the same density at 3 days. Number of TRACP+ multinucleated cells was largest (all cells with greater than two nuclei, H) or exclusively (all cells with greater than five nuclei, I) in replated myeloid blasts. N.S., Not significant. Numbers of TRACP-positive, multinucleated cells (mean±SEM) are from two experiments (C–G) or one experiment (H and I); n = 3 per experiment.

Differences in proliferation among the three myeloid stages could account for the observed differences in osteoclast formation. To study whether M-CSF + RANKL-responsive cells from the three fractions are prone to fuse and to correct for differences in proliferation, cells were collected prior to the onset of cell fusion at 3 days of culture and were replated at identical densities (104 cells). Cells were cultured further for 24 h. Multinucleated, TRACP-positive cells were only observed in myeloid blast and monocyte cultures (Fig. 2 H and I) with 2.5x more multinucleated cells in myeloid blast cultures (Fig. 2H) . Larger multinucleated cells containing greater than five nuclei were seen exclusively in myeloid blast cultures (Fig. 2I) .

Osteoclast formation on cortical bone slices took longer than on plastic. First evidence for the presence of osteoclast-like cells was seen after 6 days of culture (Fig. 3 A and B-F) . Again, at early time-points, more osteoclast-like cells were formed from myeloid blasts than from any other cell fraction. At Day 6, at least three times more osteoclast-like cells were formed from myeloid blasts compared with total bone marrow or early blasts (P<0.001) or monocytes (P<0.01). In cultures of unfractionated bone marrow (Fig. 3C ), the highest number of osteoclasts was present at Day 8 and for monocytes, at Days 6–8 (Fig. 3F) . Between Days 8 and 10, significantly more osteoclast-like cells were present in cultures from early blasts than at earlier time-points (Fig. 3D) . No extraordinary large multinucleated cells were seen when cultured for a prolonged time on bone (Fig. 3A , lower panel).


Figure 3
View larger version (37K):
[in this window]
[in a new window]

  		Figure 3. Osteoclast formation from myeloid bone marrow fractions on cortical bone. Unfractionated bone marrow, early blasts, myeloid blasts, and monocytes were cultured with M-CSF and RANKL for 4–10 days on cortical bone slices. (A) Examples of TRACP-positive, multinucleated cells (greater than two nuclei) formed after 6 days (upper panel) and 10 days (lower panel) on bone slices. Original bar = 100 µm. Arrows indicate examples of TRACP-positive, multinucleated cells (greater than five nuclei) formed after 6 days (upper panel) and 10 days (lower panel) cultured on cortical bone slices. (B) At least three times more TRACP-positive, multinucleated cells formed from myeloid blasts at 6 days. **, P < 0.01; ***, P < 0.001. (C–F) Highest numbers of osteoclasts formed after 6 days (myeloid blasts, E), 8 days (total bone marrow, C; monocytes, F), or 10 days (early blasts, D). *, P < 0.05. Numbers of TRACP-positive, multinucleated cells (mean±SEM) are from two experiments (n=3 per experiment).

Earlier and more bone resorption from myeloid blast-derived osteoclasts
Osteoclastic activity was assessed by analyzing bone resorption after 6, 8, and 10 days of culturing. At 6 days of culture, resorption pits were present in bone slices, on which myeloid blasts or monocytes were cultured (Fig. 3A , upper panel). First signs of osteoclastic activity from early blast-derived osteoclasts were observed after 8 days (Fig. 4 A , middle panel, B, and C), in agreement with the immature nature of the progenitor cells. Initial differences between early blast-derived osteoclasts and myeloid blast- or monocyte-derived osteoclasts in percentage of bone resorption and pit size disappeared over time (Fig. 4A , bottom panel, B, and C). However, at Days 6 and 8, the percentage of resorbed bone (P<0.01) was significantly higher for the myeloid blast fraction compared with any other fraction.


Figure 4
View larger version (38K):
[in this window]
[in a new window]

  		Figure 4. Bone resorption by myeloid precursor cell-derived osteoclasts. Total bone marrow, early blasts, myeloid blasts, and monocytes were cultured with M-CSF and RANKL for 6–10 days on cortical bone slices, upon which resorption pits were visualized. (A) Examples of bone resorption pits by early blast-, myeloid blast-, and monocyte-derived osteoclasts after 6 (top panel), 8 (middle panel), and 10 days (bottom panel) culture. Original bar = 100 µm. (B) Percentage of bone resorbed over time. After 6 and 8 days culture, highest bone resorption was observed in myeloid blasts cultured with M-CSF and RANKL. **, P < 0.01; ***, P < 0.001. (C) Pit size over time. After 8 days, the pit size of myeloid blast- and monocyte-derived osteoclasts was larger than pit size from osteoclasts derived from early blasts and total bone marrow. Percentage bone resorbed and pit size (mean±SEM) are from two experiments; n = 3 per experiment; *, P < 0.05.

Pit size of myeloid-derived and monocyte-derived osteoclasts was significantly larger (P<0.05, Fig. 4C ) than pit size from early blast-derived osteoclasts.

Distinct expression pattern of osteoclast-related genes per myeloid precursor cell type
Baseline characteristics c-Fms and RANK
c-Fms expression at t = 0 differed significantly between the fractions (relative expression unfractionated bone marrow 0.79±0.16; early blasts 1.9±0.3; myeloid blasts 8.3±1.1; monocytes 17.1±3.4). RANK expression at t = 0 did not differ between the cell fractions. However, expression of RANK was higher for each of the myeloid fractions compared with unfractionated bone marrow.

Osteoclast gene expression over time
We analyzed the expression of osteoclastogenesis-related genes c-Fms and RANK (Fig. 5 A and B) , osteoclast-related genes NFATc1, DC-STAMP, TRACP, calcitonin receptor (Fig. 5 C-F) , β3 integrin subunit (Supplemental Fig. 1A), and cathepsin K (Supplemental Fig. 1B) over time for all myeloid-lineage fractions upon stimulation with M-CSF and RANKL. In general, these cultures resulted in a time-dependent, increased expression of osteoclast-related genes (Fig. 5 C-F) for all cell fractions. Except for monocyte marker F4/80 and the β5 integrin unit, expression of which is reciprocally regulated with the β3 integrin during osteoclast maturation [16 , 17 ], all osteoclastogenesis (RANK and c-Fms, Fig. 4 A and B ) and osteoclast-related genes (Fig. 5 D-F) were up-regulated, and highest expression was reached at t = 4 days for DC-STAMP, TRACP, and calcitonin receptor (Fig. 5 D-F) . Expression decreased after an initial rise (NFATc1, Fig. 4C ) at t = 3 (myeloid blasts) or t = 4 days (bone marrow, early blasts, and monocytes).


Figure 5
View larger version (28K):
[in this window]
[in a new window]

  		Figure 5. Kinetic profile of gene expression by myeloid precursor cell populations stimulated with M-CSF and RANKL. Relative expression of (A) c-Fms, (B) RANK, (C) NFAT1c, (D) DC-STAMP, (E) TRACP, (F) calcitonin receptor, (G) F4/80, and (H) β5 integrin subunit was assessed at t = 0 and at 1, 2, 3, 4, 6, and 8 days of culturing with M-CSF and RANKL. Expression was normalized for PBGD, expression of which was stable throughout culturing for all cell fractions. Results (mean±SEM) obtained from two experiments (n=3/experiment) are shown.

For monocytes, myeloid blasts, and bone marrow, an initial increase of expression of the typical monocyte-lineage marker F4/80 (Fig. 5G ) and the β5 integrin unit (Fig. 5H) was noted after 1 day of culture with M-CSF and RANKL. This was followed by a decrease of expression at Days 2–4. Peak expression of these mRNAs in early blast-derived cells was only after 6 days of culture (Fig. 5 G and H) . The latter finding appears to coincide with the lower degree of differentiation of this fraction of cells.

Expression of osteoclast-related genes per cell fraction
Unfractionated bone marrow and each of the three cell fractions displayed a characteristic expression pattern of osteoclast-related genes during culture, correlating with the kinetics of osteoclast development from the various fractions. Expression was strikingly similar among the osteoclast-related genes TRACP, calcitonin receptor, β3 integrin, and cathepsin K (Fig. 5 C-F ; Supplemental Fig. 1, A and B).

Myeloid blast fraction contains highest number of colony-forming cells
Enhanced osteoclast formation by the myeloid blast cell fraction could mean that this fraction contains more proliferating cells or cells that proliferate faster. We investigated this possibility by studying the colony-forming capacity of the various fractions in a viscous methylcellulose-containing medium supplemented with M-CSF or the combination of M-CSF and RANKL [13 , 18 ]. In principle, the number of colonies reflects the number of cells that proliferate under the influence of the growth factors, and the size of the colonies possibly reflects proliferation rate per osteoclast precursor cell.

Examples of colonies are shown in Figure 6A . No colonies grew in the absence of M-CSF. Similar numbers of colonies were found between M-CSF and M-CSF + RANKL for the different bone marrow cell fractions studied. Significantly more colonies were obtained from the myeloid blast fraction compared with early blasts and monocytes (Fig. 6B) . In general, colonies emerged from 6 days onward and were compact and had a more or less round shape. However, colonies formed from early blasts (in the presence of M-CSF and RANKL) were densely packed in the center of the colony, whereas the cells at the periphery migrated away from the center (Fig. 6A , lower panel). Consequently, colony size was increased significantly in early blasts under these conditions (Fig. 6C) . Average colony size obtained from the other three cell fractions was similar in M-CSF- and M-CSF + RANKL-stimulated cultures, although the average size of colonies differed considerably; for instance, myeloid blast-derived colonies were 2.5 times larger than monocyte-derived colonies (Fig. 6A) . Statistical analyses revealed no significant differences in colony size. We assessed the responsiveness to RANKL in the M-CSF + RANKL-cultured colonies by analyzing the extent of TRACP-positive colonies. At 8 days, there were more colonies containing TRACP-positive cells derived from early blast and myeloid blasts than from monocytes (Fig. 6 E and F) . Percentages of colonies containing TRACP-positive cells leveled at 11 days at ~60% for unfractionated bone marrow and all fractions.


Figure 6
View larger version (42K):
[in this window]
[in a new window]

  		Figure 6. Effect of M-CSF or M-CSF + RANKL on colony formation of myeloid precursor cell populations in methylcellulose and [3H]thymidine incorporation in adherent cultures. (A) Appearance of the colonies formed after 11 days of culturing in M-CSF-containing medium (upper panel) or M-CSF + RANKL-containing medium (lower panel). Note the wavered appearance of early blast-derived colonies grown with M-CSF and RANKL. Original bar = 100 µm. (B) Number of colonies (defined as >10 cells) grown in M-CSF + RANKL-containing medium was highest from the myeloid blast population (P<0.01 compared with early blasts, and P<0.05 compared with monocytes). Compared with unfractionated bone marrow, significantly more colonies were formed from all cell populations (P<0.001). *, P < 0.05; **, P < 0.01. (C) Size of colonies formed in the presence of M-CSF (C) or M-CSF + RANKL (D). (E) Micrograph of two colonies stained for TRACP activity after 8 days incubation with M-CSF and RANKL. Arrows indicate TRACP-positive cells present in one colony; the other colony is TRACP-negative. (F) Percentage of TRACP-positive colonies after 8 and 14 days. Significantly less TRACP-positive colonies were found in monocyte-derived colonies at 8 days. Results (mean±SEM) obtained from two experiments (n=5 mice in total) are shown. (G) Twenty-four [3H]thymidine incorporation between Days 3 and 4 in myeloid-lineage cells cultured without cytokines (–), with M-CSF, or with M-CSF + RANKL between Days 3 and 4 at the onset of multinucleation (M-CSF + RANKL). Incorporation was negligible in cultures without cytokines and decreased significantly in myeloid blast and monocyte cultures cultured with M-CSF + RANKL. *, P < 0.05.

Divergent effects of M-CSF versus M-CSF + RANKL on [3H]thymidine incorporation
As similar numbers of colonies were obtained with M-CSF or M-CSF + RANKL, we studied the effect of these two culture conditions on proliferation of the different cell fractions between 3 and 4 days of culturing in adherent plates. This is the time-frame crucial for the initial formation of multinucleated cells on plastic, as no multinucleated cells were seen at Day 3, whereas at Day 4, the first multinucleated cells are formed. Hardly any proliferation took place in the absence of cytokines. No differences in proliferation were found between M-CSF- and M-CSF + RANKL-treated, unfractionated bone marrow and early blasts. However, proliferation was significantly lower in myeloid blast and monocyte cultures stimulated with M-CSF + RANKL compared with M-CSF alone (Fig. 6G) . Thus, at the onset of cell fusion, RANKL affects mainly proliferation of osteoclast precursor cells in the latter two cell fractions and not in early blast and total bone marrow cultures.

DISCUSSION

Multinucleated osteoclasts are formed in close proximity to the bone surface. Immature cells from bone marrow, also containing osteoclast progenitor cells, are in close vicinity to cells of the osteoblast lineage. In the present study, we demonstrate that murine osteoclasts are formed exclusively from bone marrow populations that contain cells of the myeloid lineage: from early blasts, myeloid blasts, and monocytes. Following stimulation with M-CSF and RANKL, each of these cell types followed its own distinctive route of osteoclast formation (summarized in Fig. 7 ). Particularly, one cell population from the myeloid differentiation route, the myeloid blasts, proved to form osteoclasts within a relatively short period of time.


Figure 7
View larger version (22K):
[in this window]
[in a new window]

  		Figure 7. Myeloid blasts are the mouse bone marrow cells prone to differentiate into osteoclasts. Summarizing scheme showing the differences in kinetics of osteoclast formation from three successive myeloid precursor cell populations. Myeloid blasts develop earliest into osteoclasts, on plastic (left panel) and on cortical bone slices (right panel). Onset and maximal bone resorption follow the same sequence (myeloid blasts, monocytes, early blast) as osteoclast formation on bone.

The fractionation of the different bone marrow cell populations was done by making use of CD31 and Ly-6C. These markers were used previously to monitor the differentiation of macrophages [11 ] and DC [12 ]. In both studies, it was shown that early blasts have the highest proliferative potential but require more time to differentiate than more mature stages. These studies further showed that differentiation of macrophages and DC follows the consecutive differentiation route: early blast, myeloid blast, monocyte, and eventually, macrophage [11 ] or DC [12 ]. A major source of osteoclast precursors is considered to be blood monocytes [5 , 19 ]. Conceivably, the blood-borne monocyte seems to be the recruited cell type for osteoclast formation at sites devoid of bone marrow, such as at eroding joints in rheumatoid arthritis. However, two recent reports support the concept that osteoclast development may deviate from the alleged mainstream of mature monocyte/macrophage development. Jacquin et al. [10 ] have shown that mouse bone marrow-derived osteoclasts form with highest efficiency from CD11b–/low cells, and monocytes typically have a CD11bhi phenotype. Clanchy et al. [20 ] have shown that human blood-derived osteoclasts form with high frequencies from proliferating, immature monocytes. Multinucleated giant cells also form at higher frequencies from fresh monocytes than from monocytes differentiated for some days toward macrophages [21 ]. Although osteoclasts can be generated from all three stages of myeloid differentiation in mouse bone marrow, our study revealed that myeloid blasts rather than monocytes are the candidate cells from which osteoclasts are readily formed. Hence, preferred osteoclast formation seems to deviate from the classical maturation sequence early blasts -> myeloid blasts -> monocytes and ultimately, mature end-stage. If this sequence were apparent first, osteoclasts would have been detected from monocytes. Therefore, our results suggest that osteoclast differentiation in the bone marrow follows a sequence different from macrophage or DC differentiation: Osteoclasts are preferentially formed from a less-mature stage, being the myeloid blasts. Compared with monocytes, this stage contains more M-CSF + RANKL-responding cells, as detected with the CFU assay. However, in an attempt to minimize the effect of proliferation, we assessed osteoclast formation by replating M-CSF- and RANKL-responsive cells from all cell types at 3 days, 1 day before fusion occurs. Also in this experiment, myeloid blasts gave rise to more multinucleated cells than early blasts and monocytes.

qPCR analysis further revealed that expression of NFATc1, an important transcription factor for osteoclast-related genes [22 ], precedes expression of osteoclast-related genes in myeloid blast-derived cells and peaks at 3 days of culturing with M-CSF and RANKL, 1 day earlier than observed in the other cell fractions (Fig. 5) . In parallel with the drop in osteoclast numbers after Day 4, mRNA expression of osteoclast-related genes TRACP, calcitonin receptor, and cathepsin K is highest at Day 4 and typically drops in the myeloid blast-derived cultures, whereas expression of these genes peaks later in case of early blast-derived osteoclasts or reaches a plateau phase in case of monocyte-derived osteoclasts.

In line with the immature nature of the precursor cells, osteoclast differentiation from early blasts took longer than differentiation from monocytes (Fig. 3 D and F) , resulting in a delay in osteoclastic activity (Fig. 4 B and C) and expression of osteoclast-related genes (Fig. 5) . Proliferation of early blasts was not inhibited by addition of RANKL, compared with M-CSF alone, whereas myeloid blast and monocyte proliferation were decreased under these conditions between 3 and 4 days of culturing (Fig. 6D) , possibly indicating that the M-CSF + RANKL-cultured early blasts were not ready yet to differentiate into osteoclasts at that time-point. Myeloid blasts and monocytes with decreased proliferation when cultured with M-CSF and RANKL may have acquired the state of the relatively quiescent ostoclast precursor [23 ]. Especially cyclines, which are pivotal for proliferating cells, were shown to be down-regulated by RANKL in cultures at the time-point where the formation of syncytiae is about to start [23 ]. The intriguing wavered colony appearance generated in the presence of M-CSF and RANKL, as noted previously by Arai et al. [24 ] (Fig. 6A) , was only observed in early blast-derived colonies. This is another indication that early blasts respond differently to M-CSF and RANKL and implies that cells derived from these early blasts have the unique capacity to migrate within the viscous solution, possibly by expressing the appropriate adhesion molecules.

We showed that each of the three developmental stages of the myeloid lineage has its own characteristic osteoclast differentiation pathway. This suggests that the organism is flexible with regard to use of different cell types for osteoclast recruitment, as suggested recently by Rivollier et al. [7 ]. Their study showed that osteoclasts can also form from immature stages of DC, the cells closest to macrophages in hematopoietic development [7 ]. Recently, it was shown that distinct subsets of blood-borne monocytes exist, where fates were shown in rapid responses such as in inflammatory reactions or in slower reactions by contributing to the pool of resident macrophages in noninflamed tissues [25 , 27 ]. Analogous to this dichotomy of monocytes in the circulation, different paces of osteoclast recruitment within the bone marrow cavity could be envisaged (see also Fig. 7 ). In this respect, myeloid blasts could be the fast responders and early blasts and monocytes the reserve pool of osteoclast precursors. It is also feasible that the monocyte has acquired features of the F4/80-positive osteal tissue macrophage (OsteoMac), which possibly has a role related to anabolic functions, supporting bone formation by osteoblasts. Thus, the role of OsteoMacs and possibly monocytes as osteoclast precursor may be rather limited [26 , 28 ].

Heterogeneity of macrophages has long been recognized. Different microenvironments can give rise to macrophage diversity (reviewed in refs. [29 , 30 ]). In previous studies, we were the first to show that osteoclasts are also functionally distinct at different skeletal sites: Calvarial osteoclasts use different proteases and express higher levels of TRACP than long bone osteoclasts [31 , 32 ]. Such heterogeneity among osteoclasts might be explained by existence of various precursor cells giving rise to different types of osteoclasts. More complex possibilities of the formation of osteoclasts can be proposed by assuming heterogenic fusions of diverse progenitors (e.g., early blasts and monocytes), a principle shown to exist in vivo for heterogeneous tumor/osteoclast precursor syncitiae [33 ]. In principle, heterogeneous fusion of the various precursors from the classical monocyte-lineage pathway alone or together with the recently discovered precursors from the DC-derived osteoclast pathway [34 , 35 ] could give rise to a great diversity of osteoclasts, each of them equipped for the appropriate function at a desired skeletal site.

ACKNOWLEDGEMENTS

The study was supported by a grant from the Royal Academy of Sciences (KNAW) to T. J. d. V.

Received July 6, 2008; revised March 3, 2009; accepted March 3, 2009.

REFERENCES

    1
  1. Walker, D. G. (1975) Bone resorption restored in osteopetrotic mice by transplants of normal bone marrow and spleen cells Science 190,784-785[Abstract/Free Full Text]
    2. 2
  2. Marks, S. C., Jr, Walker, D. G. (1981) The hematogenous origin of osteoclasts: experimental evidence from osteopetrotic (microphthalmic) mice treated with spleen cells from beige mouse donors Am. J. Anat. 161,1-10[CrossRef][Medline]
    3. 3
  3. Burger, E. H., Van der Meer, J. W., van de Gevel, J. S., Gribnau, J. C., Thesingh, G. W., Van Furth, R. (1982) In vitro formation of osteoclasts from long-term cultures of bone marrow mononuclear phagocytes J. Exp. Med. 156,1604-1614[Abstract/Free Full Text]
    4. 4
  4. Scheven, B. A., Visser, J. W., Nijweide, P. J. (1986) In vitro osteoclast generation from different bone marrow fractions, including a highly enriched hematopoietic stem cell population Nature 321,79-81[CrossRef][Medline]
    5. 5
  5. Quinn, J. M. W., Neale, S., McGee, J. O. D., Athanasou, N. A. (1998) Human osteoclast formation from blood monocytes, peritoneal macrophages, and bone marrow cells Calcif. Tissue Int. 62,527-531[CrossRef][Medline]
    6. 6
  6. Fujikawa, Y., Quinn, J. M., Sabokbar, A., McGee, J. O., Athanasou, N. A. (1996) The human osteoclast precursor circulates in the monocyte fraction Endocrinology 137,4058-4060[Abstract]
    7. 7
  7. Rivollier, A., Mazzorana, M., Tebib, J., Piperno, M., Aitsiselmi, T., Rabourdin-Combe, C., Jurdic, P., Servet-Delprat, C. (2004) Immature dendritic cell transdifferentiation into osteoclasts: a novel pathway sustained by the rheumatoid arthritis microenvironment Blood 104,4029-4037[Abstract/Free Full Text]
    8. 8
  8. Miyazaki, T., Tanaka, S., Sanjay, A., Baron, R. (2006) The role of c-Src kinase in the regulation of osteoclast function Mod. Rheumatol. 16,68-74[CrossRef][Medline]
    9. 9
  9. Yao, Z., Li, P., Zhang, Q., Schwarz, E. M., Keng, P., Arbini, A., Boyce, B. F., Xing, L. (2006) Tumor necrosis factor-{alpha} increases circulating osteoclast precursor numbers by promoting their proliferation and differentiation in the bone marrow through up-regulation of c-Fms expression J. Biol. Chem. 281,11846-11855[Abstract/Free Full Text]
    10. 10
  10. Jacquin, C., Gran, D. E., Lee, S. K., Lorenzo, J. A., Aguila, H. L. (2006) Identification of multiple osteoclast precursor populations in murine bone marrow J. Bone Miner. Res. 21,67-77[CrossRef][Medline]
    11. 11
  11. De Bruijn, M. F. T. R., Slieker, W. A. T., van der Loo, J. C. M., Voerman, J. S. A., Van Ewijk, W., Leenen, P. J. M. (1994) Distinct mouse bone marrow macrophage precursors identified by differential expression of ER-MP12 and ER-MP20 antigens Eur. J. Immunol. 24,2279-2284[Medline]
    12. 12
  12. Nikolic, T., Bruijn, M. F., Lutz, M. B., Leenen, P. J. (2003) Developmental stages of myeloid dendritic cells in mouse bone marrow Int. Immunol. 15,515-524[Abstract/Free Full Text]
    13. 13
  13. De Vries, T. J., Schoenmaker, T., Beertsen, W., Van der Neut, R., Everts, V. (2005) Effect of CD44 deficiency on in vitro and in vivo osteoclast formation J. Cell. Biochem. 94,954-966[CrossRef][Medline]
    14. 14
  14. De Vries, T. J., Schoenmaker, T., Wattanaroonwong, N., van den Hoonaard, M., Nieuwenhuijse, A., Beertsen, W., Everts, V. (2006) Gingival fibroblasts are better at inhibiting osteoclast formation than periodontal ligament fibroblasts J. Cell. Biochem. 98,370-382[CrossRef][Medline]
    15. 15
  15. De Vries, T. J., Fourkour, A., Punt, C. J. A., Van de Locht, L. T. F., Wobbes, T., Van den Bosch, S., De Rooij, M. J. M., Mensink, E. J. B. M., Ruiter, D. J., Van Muijen, G. N. P. (1999) Reproducibility of detection of tyrosinase and MART-1 transcripts in the peripheral blood of melanoma patients: a quality control study using real-time quantitative PCR Br. J. Cancer 80,883-891[CrossRef][Medline]
    16. 16
  16. Inoue, M., Namba, N., Chappel, J., Teitelbaum, S. L., Ross, F. P. (1998) Granulocyte macrophage-colony stimulating factor reciprocally regulates {alpha}v-associated integrins on murine osteoclast precursors Mol. Endocrinol. 12,1955-1962[Abstract/Free Full Text]
    17. 17
  17. Lane, N. E., Yao, W., Nakamura, M. C., Humphrey, M. B., Kimmel, D., Huang, X., Sheppard, D., Ross, F. P., Teitelbaum, S. L. (2005) Mice lacking the integrin β5 subunit have accelerated osteoclast maturation and increased activity in the estrogen-deficient state J. Bone Miner. Res. 20,58-66[CrossRef][Medline]
    18. 18
  18. Iotsova, V., Caamano, J., Loy, J., Yang, Y., Lewin, A., Bravo, R. (1997) Osteopetrosis in mice lacking NF-{kappa}B1 and NF-{kappa}B2 Nat. Med. 3,1285-1289[CrossRef][Medline]
    19. 19
  19. Atkins, G. J., Kostakis, P., Vincent, C., Farrugia, A. N., Houchins, J. P., Findlay, D. M., Evdokiou, A., Zannettino, A. C. (2006) RANK expression as a cell surface marker of human osteoclast precursors in peripheral blood, bone marrow, and giant cell tumors of bone J. Bone Miner. Res. 21,1339-1349[CrossRef][Medline]
    20. 20
  20. Clanchy, F. I., Holloway, A. C., Lari, R., Cameron, P. U., Hamilton, J. A. (2006) Detection and properties of the human proliferative monocyte subpopulation J. Leukoc. Biol. 79,757-766[Abstract/Free Full Text]
    21. 21
  21. Most, J., Spotl, L., Mayr, G., Gasser, A., Sarti, A., Dierich, M. P. (1997) Formation of multinucleated giant cells in vitro is dependent on the stage of monocyte to macrophage maturation Blood 89,662-671[Abstract/Free Full Text]
    22. 22
  22. Ishida, N., Hayashi, K., Hoshijima, M., Ogawa, T., Koga, S., Miyatake, Y., Kumegawa, M., Kimura, T., Takeya, T. (2002) Large scale gene expression analysis of osteoclastogenesis in vitro and elucidation of NFAT2 as a key regulator J. Biol. Chem. 277,41147-41156[Abstract/Free Full Text]
    23. 23
  23. Mizoguchi, T., Muto, A., Udagawa, N., Arai, A., Yamashita, T., Hosoya, A., Ninomiya, T., Nakamura, H., Yamamoto, Y., Kinugawa, S., Nakamura, M., Nakamichi, Y., Kobayashi, Y., Nagasawa, S., Oda, K., Tanaka, H., Tagaya, M., Penninger, J. M., Ito, M., Takahashi, N. (2009) Identification of cell cycle-arrested quiescent osteoclast precursors in vivo J. Cell Biol. 184,541-554[Abstract/Free Full Text]
    24. 24
  24. Arai, F., Miyamoto, T., Ohneda, O., Inada, T., Sudo, T., Brasel, K., Miyata, T., Anderson, D. M., Suda, T. (1999) Commitment and differentiation of osteoclast precursor cells by the sequential expression of c-Fms and receptor activator of nuclear factor {kappa}B (RANK) receptors J. Exp. Med. 190,1741-1754[Abstract/Free Full Text]
    25. 25
  25. Geissmann, F., Jung, S., Littman, D. R. (2003) Blood monocytes consist of two principal subsets with distinct migratory properties Immunity 19,71-82[CrossRef][Medline]
    26. 26
  26. Pettit, A. R., Chang, M. K., Hume, D. A., Raggatt, L. J. (2008) Osteal macrophages: a new twist on coupling during bone dynamics Bone 43,976-982[CrossRef][Medline]
    27. 27
  27. Sunderkotter, C., Nikolic, T., Dillon, M. J., van Rooijen, N., Stehling, M., Drevets, D. A., Leenen, P. J. M. (2004) Subpopulations of mouse blood monocytes differ in maturation stage and inflammatory response J. Immunol. 172,4410-4417[Abstract/Free Full Text]
    28. 28
  28. Chang, M. K., Raggatt, L. J., Alexander, K. A., Kuliwaba, J. S., Fazzalari, N. L., Schroder, K., Maylin, E. R., Ripoll, V. M., Hume, D. A., Pettit, A. R. (2008) Osteal tissue macrophages are intercalated throughout human and mouse bone lining tissues and regulate osteoblast function in vitro and in vivo J. Immunol. 181,1232-1244[Abstract/Free Full Text]
    29. 29
  29. Leenen, P. J. M., Campbell, P. A. (1993) Heterogeneity of mononuclear phagocytes. An interpretive review Blood Cell Biochem 5,29
    30. 30
  30. Pixley, F. J., Stanley, E. R. (2004) CSF-1 regulation of the wandering macrophage: complexity in action Trends Cell Biol. 14,628-638[CrossRef][Medline]
    31. 31
  31. Everts, V., Korper, W., Jansen, D. C., Steinfort, J., Lammerse, I., Heera, S., Docherty, A. J. P., Beertsen, W. (1999) Functional heterogeneity of osteoclasts: matrix metalloproteinases participate in osteoclastic resorption of calvarial bone but not in resorption of long bone FASEB J. 13,1219-1230[Abstract/Free Full Text]
    32. 32
  32. Everts, V., Korper, W., Hoeben, K. A., Jansen, I. D., Bromme, D., Cleutjens, K. B., Heeneman, S., Peters, C., Reinheckel, T., Saftig, P., Beersten, W. (2006) Osteoclastic bone degradation and the role of different cysteine proteinases and matrix metalloproteinases: differences between calvaria and long bone J. Bone Miner. Res. 21,1399-1408[CrossRef][Medline]
    33. 33
  33. Andersen, T. L., Boissy, P., Sondergaard, T. E., Kupisiewicz, K., Plesner, T., Rasmussen, T., Haaber, J., Kolvraa, S., Delaisse, J. M. (2007) Osteoclast nuclei of myeloma patients show chromosome translocations specific for the myeloma cell clone: a new type of cancer-host partnership? J. Pathol. 211,10-17[CrossRef][Medline]
    34. 34
  34. Alnaeeli, M., Park, J., Mahamed, D., Penninger, J. M., Teng, Y. T. (2007) Dendritic cells at the osteo-immune interface: implications for inflammation-induced bone loss J. Bone Miner. Res. 22,775-780[CrossRef][Medline]
    35. 35
  35. Alnaeeli, M., Penninger, J. M., Teng, Y. T. (2006) Immune interactions with CD4+ T cells promote the development of functional osteoclasts from murine CD11c+ dendritic cells J. Immunol. 177,3314-3326[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Figure 1
Right arrow All Versions of this Article:
jlb.0708402v1
85/6/919    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Vries, T. J.
Right arrow Articles by Everts, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Vries, T. J.
Right arrow Articles by Everts, V.