Chemokine and chemokine receptor expression during colony stimulating factor-1-induced osteoclast differentiation in the toothless osteopetrotic rat: a key role for CCL9 (MIP-1gamma) in osteoclastogenesis in vivo and in vitro

Abstract

Osteoclasts differentiate from hematopoietic precursors under systemic and local controls. Chemokines and receptors direct leukocyte traffic throughout the body and may help regulate site-specific bone resorption. We investigated bone gene expression in vivo during rapid osteoclast differentiation induced by colony-stimulating factor 1 (CSF-1) in Csf1-null toothless (tl/tl) rats. Long-bone RNA from CSF-1-treated tl/tl rats was analyzed by high-density microarray over a time course. TRAP (tartrate-resistant acid phosphatase)-positive osteoclasts appeared on day 2, peaked on day 4, and decreased slightly on day 6, as marrow space was expanding. TRAP and cathepsin K mRNA paralleled the cell counts. We examined all chemokine and receptor mRNAs on the arrays. CCL9 was strongly induced and peaked on day 2, as did its receptor, CCR1, and regulatory receptors c-Fms (CSF-1 receptor) and RANK (receptor activator of nuclear factor kappaB). Other chemokines and receptors showed little or no significant changes. In situ hybridization and immunohistochemistry revealed CCL9 in small, immature osteoclasts on day 2 and in mature cells at later times. Anti-CCL9 antibody inhibited osteoclast differentiation in culture and significantly suppressed the osteoclast response in CSF-1-treated tl/tl rats. While various chemokines have been implicated in osteoclastogenesis in vitro, this first systematic analysis of chemokines and receptors during osteoclast differentiation in vivo highlights the key role of CCL9 in this process.

Figure 1.

TRAP-positive osteoclasts in tl rats in response to CSF-1 injections. (A) TRAP staining of proximal tibial metaphysis reveals few osteoclasts by day 2 (red, arrows), and well-established populations at days 4 and 6 (images were obtained with a Zeiss Axioskop 2 Plus microscope with a 20 ×/0.5 NA brightfield objective). None were seen at day 0 (not shown). (B) Cell counts were performed in a window of defined size in 2 areas, area A near the chondro-osseous junction, and area B, in the lower metaphysis, in replicate slides of multiple animals (n = 24-36; bars show mean + 1 SD).

Figure 2.

CCL-9 and CCR1 mRNA expression in long bones determined by microarray and real-time PCR. tl rats were treated for the days indicated and total bone RNA was analyzed. CCL9 and CCR1 mRNA were highest after 2 days of treatment, after which the levels subsided. Means + 1 SD are shown. Compared with untreated animals (day 0), all time points were significantly elevated. (A-B) *Difference from day 0, P < .01. (C) *Difference from day 0, P < .005; **P < .02.

Figure 3.

In situ hybridization for CCL-9 mRNA during CSF-1 time course, proximal tibia of tl rat. No CCL-9-expressing cells were visible at day 0 (A). CCL-9-positive cells are indicated in panels B-D with arrows. Many small cells were detected at day 2 (B,E), and larger osteoclasts were seen at days 4 (C,F) and 6 (D,G). Higher magnification revealed cytological features of the CCL-9-expressing cells. Arrows in panels F and G indicate some individual nuclei. The cells seen at day 2 were typically smaller, mononuclear, and less differentiated, while at days 4 and 6, osteoclasts generally contained multiple nuclei and were better differentiated. The CCL-9 mRNA is diffusely distributed throughout the cytoplasm in the differentiating mononuclear pre-osteoclasts in panel E, but appears highly concentrated in the ruffled border area in days 4 (F) and 6 (G). *Bone trabeculae. Original magnifications: × 100 (A-D); × 630 (E-G). No counterstain was used.

Figure 4.

Immunohistochemistry for CCL-9 in the proximal tibia of tl rats. Immunohistochemistry after 2 (A), 4 (B), or 6 (C) days of CSF-1 injections is shown. Orange-brown color indicates CCL-9 protein. At day 2, the staining was most evident in small, newly formed osteoclasts, while at days 4 and 6, large, mature osteoclasts were present with much weaker, more diffuse labeling. *Bone trabeculae; v indicates a prominent vacuole in the large osteoclast in panel B. Scale bar in panel C equals 21 μm. Shown are 10-μm paraffin sections without counterstain. Objective magnification: 63 ×/1.25 NA.

Figure 5.

Inhibition of CCL9 blocks osteoclast differentiation in vitro. Rat BMCs were induced to differentiate into osteoclasts using CSF-1 and RANKL. Anti-CCL9 antibody blocked the formation of TRAP-positive, multinucleated cells. Means + 1 SD are shown. For both antibody concentrations tested, the decrease was highly significant. *P < .001.

Figure 6.

Injections of anti-CCL9 suppress osteoclast differentiation in vivo. tl/tl rats were injected either with CSF-1 alone or with anti-CCL9 antibody and CSF-1, and TRAP⁺ osteoclasts were counted on multiple sections of the proximal femur from 3 animals for each condition after 4 days of injections. Anti-CCL9 significantly reduced the osteoclastogenic response to CSF-1 injections. Means + 1 SD are shown. In both areas, differences were highly significant (*P < .001).