Parathyroid hormone (PTH) has a significant role as an anabolic hormone in bone when administered by intermittent injection. Previous microarray studies in our laboratory have shown that the most highly regulated gene, monocyte chemoattractant protein-1 (MCP-1), is rapidly and transiently induced when hPTH(1-34) is injected intermittently in rats. Through further in vivo studies, we found that rats treated with hPTH(1-34) showed a significant increase in serum MCP-1 levels 2 hours after PTH injection compared with basal levels. Using immunohistochemistry, increased MCP-1 expression in osteoblasts and osteocytes is evident after PTH treatment. PTH also increased the number of marrow macrophages. MCP-1 knockout mice injected daily with hPTH(1-34) showed less trabecular bone mineral density and bone volume compared with wild-type mice as measured by peripheral quantitative computed tomography (pQCT) and micro-computed tomography (µCT). Histomorphometric analysis revealed that the increase in osteoclast surface and osteoclast number observed with intermittent PTH treatment in the wild-type mice was completely eliminated in the MCP-1 null mice, as well as much lower numbers of macrophages. Consequently, the lack of osteoclast and macrophage activity in the MCP-1 null mice was paralleled by a reduction in bone formation. We conclude that osteoblast and osteocyte MCP-1 expression is an important mediator for the anabolic effects of PTH on bone.
Parathyroid hormone (PTH) is the major hormone regulating calcium homeostasis, but it also has a significant role in bone remodeling. Bone remodeling is a coordinated and highly regulated process of bone resorption and bone formation necessary for maintaining bone health. PTH has a dual effect on bone turnover depending on the mode of administration. When given as intermittent injections, PTH stimulates bone remodeling with a net increase in bone formation, but continuous infusion causes bone loss.[1-4] The anabolic effects of parathyroid hormone are well documented in both animal and human studies.[5-8] Recombinant human parathyroid hormone is currently used as a treatment for osteoporosis.[9, 10] The mechanisms of this contradictory effect are not completely understood, although it has been established that osteoclastic resorption is necessary for the bone anabolic effect of PTH.[11-13] Mature osteoclasts are formed by the fusion of precursor cells of the monocyte/macrophage lineage. The PTH receptor (PTH1R) is located on the osteoblast, as well as stromal cells, osteoblast progenitors, and osteocytes. PTH-stimulated osteoclastic resorption is mediated through factors expressed by the osteoblast lineage cells, and osteocytes have been shown to be a major target of PTH-stimulated bone remodeling.[15-17]
There are many known genes that are regulated by both intermittent and continuous administration of PTH.[18, 19] In our laboratory,[20, 21] microarray studies and mRNA from 3-month-old female rats treated with intermittent human PTH(1-34) showed that the chemokine monocyte chemoattractant protein-1 (MCP-1, CCL2) was the most highly induced gene (100- to 250-fold). RT-PCR of RNA isolated from femurs of rats treated with hPTH(1-34) for 14 days showed that MCP-1 mRNA induction was very rapid, but transient, with maximal expression 1 hour after PTH injection. In UMR 106-01 cells, MCP-1 mRNA levels increased 20-fold around 90 minutes after hPTH(1-34) treatment, and MCP-1–secreted protein levels peaked 2 hours after treatment. MCP-1 increased 1 hour after hPTH(1-34) treatment in primary calvarial osteoblastic cells in the differentiation, proliferation, and mineralization stages. Osteoclastogenesis is induced by receptor activator of NF-κB ligand (RANKL) and macrophage colony-stimulating factor (M-CSF). The addition of MCP-1 to primary cultures containing M-CSF and RANKL significantly increased tartrate-resistant acid phosphatase (TRAP) mRNA expression, preosteoclast differentiation, and fusion.
Chemokines are small proteins that function primarily to direct the movement of circulating leukocytes and monocytes to sites of infection, inflammation, trauma, and ischemia. Recruitment of leukocytes in response to inflammatory stimuli is crucial for the cellular and adaptive immune responses. Chemokines are divided into four main subfamilies (C, CC, CXC, and CX3C) based on the location of the N-terminal cysteine residues. MCP-1, also known as CCL2, is a member of the CC subfamily of chemokines. Chemokines selectively activate leukocyte subpopulations. The principal function of MCP-1 is to stimulate monocyte and macrophage recruitment. MCP-1 predominantly initiates signal transduction through binding a G-protein coupled receptor CCR2. Monocytes express CCR1, CCR2, and CCR5, but MCP-1 binds only to CCR2 with high affinity. A close association exists between the immune system and the skeletal system. Immune cells and hematopoietic stem cells are found in the bone marrow, where they interact with bone cells, and the two systems share a number of regulatory molecules including cytokines, receptors, signaling molecules, and transcription factors. MCP-1 is detected at the site of tooth eruption, and it is implicated in the pathogenesis of inflammatory conditions characterized by monocyte cell infiltration such as bacterial infection, atherosclerosis, rheumatoid arthritis, and bone metastasis. Osteoblasts are the principal cells expressing MCP-1 in inflamed bone, and there is a significant correlation between MCP-1–positive cells and recruitment of monocytes and macrophages.
MCP-1 and CCR2 knockout mice both exhibit a deficiency in monocyte recruitment in response to inflammatory conditions.[23, 26] Analysis of the bone phenotype of CCR2–/– mice showed increased cortical BMD but less trabecular bone in lumbar spine and distal femur. They also showed less increase in bone volume after intermittent PTH treatment, indicating that signaling through CCR2 mediates the anabolic effect of PTH. In contrast, another group reported CCR2–/– mice have increased bone mineral density (BMD) and trabecular bone content attributed to a decrease in osteoclast development. Furthermore, ovariectomized CCR2–/– mice were also protected from osteoporosis. Even though the CCR2 receptor is the primary receptor for MCP-1, it can also bind other CC chemokines.[29, 30] Thus, to determine which chemokine is important in PTH's anabolic actions, we investigated the bone phenotype and effects of PTH treatment in MCP-1–/– mice. We found that MCP-1 expression is necessary for osteoclast and macrophage recruitment, osteoclast formation, and bone resorption associated with PTH anabolic effects.
Materials and Methods
All animals were maintained under conditions of a 12-hour light/dark cycle with standard rodent chow and water ad libitum. The Institutional Animal Care and Use Committee at UMDNJ-Robert Wood Johnson Medical School (Piscataway, NJ, USA) approved all treatments and procedures.
Serum MCP-1 in rats
Male Sprague-Dawley rats 4 months of age were obtained from Harlan Laboratories (Indianapolis, IN, USA). Rats were lightly anesthetized with isoflurane (Baxter, Deerfield, IL, USA), and blood was collected from the lateral tail vein. Serum MCP-1 levels were measured by ELISA using the rat MCP-1 ELISA kit from Invitrogen (Carlsbad, CA, USA). MCP-1 concentrations from rat samples were determined from a standard curve constructed using curve-fitting software (four-parameter algorithm). Minimal detectable concentration of MCP-1 is <8.0 pg/mL, assay precision: coefficient of variation (CV) = 5.7% intra-assay; 8.5% interassay. Rats were treated with hPTH(1-34) sc (Millipore, Billerica, MA, USA) or vehicle (saline) injections.
Four-month-old male rats (n = 10 per group) were treated with daily injections of vehicle (saline) or hPTH(1-34) at 80 µg/kg for 14 days. Blood samples were collected 0 to 6 hours on day 14 of treatment.
Dose effect and duration of treatment
Four-month-old male rats (n = 5 per group) were treated with daily injections of vehicle (saline) or hPTH(1-34) at 40 or 80 µg/kg for up to 14 days. Blood samples were collected 2 hours post-dose after 1, 7, and 14 days of treatment.
Four-month-old male and female rats (n = 5 per group) were treated with daily injections of vehicle (saline) or hPTH(1-34) at 80 µg/kg for 28 days. Blood samples were collected before beginning treatment and again 2 hours after injection on day 28 of treatment.
Tibias were collected 2 hours post-dose on days 1 to 3, 7, and 14 of treatment. Tibias were fixed in 10% neutral-buffered formalin, decalcified in 10% EDTA, then processed through graded ethanols, cleared in xylene, and embedded in paraffin. Sections (5 µm) were cut on a Shandon Finesse microtome (Thermo-Fisher Scientific, Waltham, MA, USA) and collected on slides. For MCP-1 IHC, the sections were deparaffinized, rehydrated, and immunostained with a 1:50 dilution of anti-rat MCP-1 antibody or isotype control (Millipore) with the HistoMouse Max DAB detection kit (Invitrogen). Sections were first blocked in blocking reagent. Primary antibodies were incubated at room temperature for 60 minutes followed by incubation with secondary antibody for 10 minutes. DAB chromogen for MCP-1 detection was applied for 3 minutes, then counterstained with hematoxylin for 1 minute.
For CD68[ED-1] IHC, additional sections were deparaffinized and rehydrated followed by antigen retrieval in 0.37% trypsin for 10 minutes at 37°C. Sections were immunostained with a 1:50 dilution of anti-rat CD68 [ED-1] (Millipore) or isotype control with HistoMouse Max AEC detection kit (Invitrogen). Sections were first blocked in blocking reagent. Primary antibodies were incubated at room temperature for 60 minutes followed by a 10-minute incubation with secondary antibody. AEC chromogen for CD68 detection was applied for 10 minutes, and the sections were counterstained with hematoxylin for 1 minute. Analysis of monocyte/macrophage number was done using a BioQuant image analysis system (Nashville, TN, USA). CD68[ED-1]–positive cells were counted in the secondary spongiosa between trabeculae. Osteoclasts and multinucleated cells were not included in the cell count. A total of 36 fields at 100× magnification were measured.
Tibias were collected 2 hours post-dose after 6 weeks of treatment, which was administered 5 days per week. The samples were processed and immunostained for MCP-1 as described above for rats. For F4/80 and MAC-3, the sections were treated similarly to the MCP-1 sections with a 1:300 dilution of anti-mouse F4/80 or MAC-3 antibody or isotype control (Serotec, Oxford, UK). Primary antibodies were incubated at room temperature for 60 minutes followed by incubation with secondary biotinylated antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) at 1:200 dilution and streptavidin (Dako, Carpinteria, CA, USA) at 1:300 dilution for 30 minutes each. AEC chromogen for F4/80 and MAC-3 detection was applied for 5 minutes, then counterstained with hematoxylin for 1 minute.
Bone effects of PTH in MCP-1 null mice
Breeding pairs of MCP-1–/– mice and wild-type controls (C57/BL6) were purchased from The Jackson Laboratory (Bar Harbor, ME, USA), and colonies were established and crossed for 10 generations. The genotypes of preweanling mice were confirmed by PCR following the Jackson Laboratory genotyping protocol. Male and female wild-type and MCP-1–/– mice at 4 months of age (n = 12 per group) were treated with daily injections of vehicle (saline) or hPTH(1-34) at 80 µg/kg sc 5 days per week for 6 weeks. Mice at 6 months of age (n = 8 per group) were also treated in the same manner. All animals were injected with 15 mg/kg calcein (Sigma, St. Louis, MO, USA) 9 and 2 days before necropsy. Tibias were excised; one was processed for histology, and the other was used for ex vivo peripheral quantitative computed tomography (pQCT) and micro-computed tomography (μCT) analysis of trabecular bone architecture. One femur was also collected for pQCT analysis of cortical bone.
Peripheral quantitative computed tomography
The total and trabecular BMD of the proximal tibia was evaluated using an XCT Research SA (Norland/Stratec Medizintechnik, Pforzheim, Germany). After running a scout scan for a length of 10 mm, the pQCT scan was initiated 1.4 mm distal from the proximal epiphysis. The scan is 1 mm thick with a voxel size of 90 μm. Using an iterative algorithm, soft tissue (density below 223 mg/cm3) was automatically removed and the density of the remaining bone reported as total density (mg/cm3). A concentric peel of the outer 55% of the bone was used to determine trabecular density (mg/cm3). Cortical BMD was measured on the femur. The femur length was determined after a scout scan was run for the full length of the bone. The pQCT scan was initiated at the mid-diaphysis of the femur. The scan is 1 mm thick with a voxel size of 90 μm. Using an iterative algorithm, tissue with density below 500 mg/cm3 was automatically removed. The density of the remaining bone is reported as cortical density (mg/cm3). The analysis software was also used to determine cortical thickness (Ct.Th), periosteal perimeter (Ps.Pm), and endocortical perimeter (Ec.Pm).
Measurements of trabecular architecture were done on the proximal tibia metaphysis using a μCT 20 (Scanco Medical Ag, Bassersdorf, Switzerland). After an initial scout scan, a total of 100 slices were obtained on each bone sample starting 1.4 mm distal to the proximal epiphysis. The area for analysis was outlined within the trabecular compartment, excluding the cortical and subcortical bone. Every 10 sections were outlined, and the intermediate sections were interpolated with the contouring algorithm to create a volume of interest. A three-dimensional analysis was done to determine bone volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), and trabecular separation (Tb.Sp). The μCT measurements follow the guidelines reported by Bouxsein and colleagues.
Tibias were fixed in 70% (v/v) ethanol and embedded undecalcified in methyl methacrylate (Polysciences, Warrington, PA, USA). Longitudinal tissue sections (5 and 10 μm) were cut on a Polycut S microtome (Leica, Bannockburn, IL, USA). The 5-μm sections were stained with Goldner's Trichrome stain and the 10-μm sections were left unstained for dynamic histomorphometric measurements. Quantitative histomorphometry was done using a BioQuant image analysis system (Nashville, TN, USA). Fifty fields in the secondary spongiosa at 40× were quantified. Static histomorphometry was measured on 5-μm Goldner's stained sections. Fluorochrome (calcein)-based parameters of bone formation were measured on unstained 10-μm sections under fluorescent light. Bone area (B.Ar), bone surface (BS), osteoid surface (Os.S), osteoclast surface (Oc.S), osteoclast number (N.Oc), single-labeled surface (sLS), double-labeled surface (dLS), and label thickness (L.Th) were measured. Derived indices of bone resorption and formation including percent bone area (B.Ar/T.Ar), percent osteoid surface (OS/BS), percent osteoclast surface (Oc.S/BS), osteoclast number per millimeter of bone surface (N.Oc/BS), percent mineralizing surface (MS/BS) calculated as dLS + ½ sLS, mineral apposition rate (MAR), and bone formation rate/bone surface referent (BFR/BS) were calculated. These measurements follow the standard nomenclature approved by the American Society for Bone and Mineral Research.
Data were analyzed by Dunnett's one-way analysis of variance (ANOVA) or two-way ANOVA followed by Tukey's multiple comparison test using SigmaStat software (SPSS Sciences, Chicago, IL, USA). Data are expressed as the mean ± standard error with p < 0.05 considered significant.
Effect of PTH on serum MCP-1 levels in rats
Because the induction of MCP-1 mRNA in rats injected with PTH was so profound (100- 250-fold 1 hour after injection), we were interested in assessing if this was reflected in changes in serum levels of MCP-1. The time course and dosing duration used to examine serum MCP-1 were modeled after the treatment regimen used in our in vivo mRNA work, which followed the anabolic protocols for rats developed previously by others.[2, 19] To determine the time point where maximal serum levels of MCP-1 are achieved after injection with PTH, rats were injected daily with 80 μg/kg hPTH(1-34) sc for 14 days, and serial blood samples were collected after the last injection. Results show that peak serum MCP-1 levels occur 2 hours after PTH injection with levels approaching baseline by 4 hours (Fig. 1A). There was a 1.5- to 2-fold increase in serum MCP-1 levels 2 hours after PTH injection compared with basal levels.
The dose response of PTH treatment and the effect of the duration of PTH treatment on serum MCP-1 levels were further investigated. Rats were injected daily with vehicle (saline), 40 μg/kg or 80 μg/kg hPTH(1-34) sc for 14 days. Blood was collected 2 hours after injection on days 1, 7, and 14. Fig. 1B shows that with the 40-μg/kg dose of hPTH[1-34], serum MCP-1 levels increase in comparison to vehicle-treated animals with the magnitude of the increase dependent on the duration of PTH treatment. MCP-1 levels increased by 24%, 53%, and 79% compared with vehicle on days 1, 7, and 14, respectively. In contrast, 80 μg/kg of hPTH(1-34) produced a maximal increase in serum MCP-1 from the first injection, and no further increase was observed with duration of treatment. A previous study with hPTH(1-34) at doses of 40, 80, and 120 μg/kg showed that maximum serum levels of MCP-1 were attained at the 80-μg/kg dose, and no further increase was observed with a higher dose of PTH (data not shown).
Possible sex differences in serum MCP-1 levels were examined after a 1-month treatment period with PTH. In addition, we were also interested to see if the increase in serum MCP-1 was maintained with an extended duration of PTH treatment. Male and female rats (n = 6/group) were injected daily with vehicle (saline) or 80 μg/kg hPTH(1-34) sc for 28 days. Serum samples were analyzed for MCP-1 at baseline before treatment began and again on day 28 2 hours after the last hPTH(1-34) injection. Serum MCP-1 levels in the vehicle groups treated for 28 days were not different from the baseline levels before treatment began. Two hours after PTH treatment on day 28, serum levels increased twofold in both males and females. Serum levels in the female rats were 40% to 50% less than in male rats at baseline and after treatment (Fig. 1C). Overall, these data provide quantitative proof that MCP-1 protein is increased by PTH in vivo.
Immunohistochemistry of tibias in PTH-treated rats
In an effort to determine if osteoblasts express MCP-1 in response to PTH in vivo, bone samples were processed for immunohistochemistry to detect the chemokine. Rats were injected daily with vehicle (saline) or hPTH(1-34) at 80 mg/kg sc for up to 14 days. The tibias were collected 2 hours after the injections on days 1, 2, 3, 7, and 14. Two hours after 1 PTH injection, very little difference in staining was detected in the active cuboidal osteoblasts lining the bone surface in comparison to saline treatment. Increased MCP-1 expression in osteoblasts was evident after 2 days of PTH injections, and by 3 days a pronounced increased staining was observed (Fig. 2A). Maximal staining was sustained through 7 and 14 days of treatment. Staining was also present in other bone marrow cells, including monocytes, macrophages, osteoclasts, and also in some osteocytes. The intensity of staining in these cells also increased after PTH injections. As shown in Fig. 2B, monocytes and macrophages in the bone marrow, as well as osteoclasts attached to the bone surface, were stained with CD68[ED-1], a lysosomal membrane marker for monocytes and macrophages. After 14 days of treatment with 80 mg/kg hPTH[1-34], there was an obvious increase in the number of CD68[ED-1]–positive cells. Cell counts revealed that monocytes/macrophages in the marrow increased by 56% in the PTH-treated rats (average total cells = 245 ± 33) compared with vehicle treatment (average total cells = 156 ± 29). The increase was statistically significant at p < 0.05 (n = 5).
Bone mineral density of MCP-1–/– mice treated with PTH
To further investigate the role of MCP-1 in the anabolic action of PTH in vivo, 4-month-old male wild-type (C57/BL6) mice and MCP-1–/– mice were injected with either vehicle (saline) or hPTH(1-34) at 80 mg/kg sc 5 days per week for 6 weeks. After 6 weeks, pQCT of the proximal tibiae (Fig. 3A) showed that PTH treatment significantly increased both total and trabecular BMD in wild-type mice, whereas MCP-1–/– mice showed little response. Trabecular bone significantly increased in wild-type mice treated with PTH in comparison to vehicle-treated mice (vehicle: 452.18 ± 6.61 mg/cm3; PTH: 521.48 ± 13.50 mg/cm3). In contrast, the MCP-1 null mice showed a slight increase in trabecular BMD that is not statistically different from the corresponding vehicle-treated MCP-1 mice (vehicle: 453.93 ± 23.57 mg/cm3; PTH: 485.90 ± 13.98 mg/cm3). PTH-treated MCP-1 null mice have approximately 7.0% less trabecular BMD than the corresponding PTH-treated wild-type mice, although the difference is not statistically significant. There are no differences between the wild-type mice and MCP-1–/– mice treated with vehicle. Overall, the effect of PTH on BMD was blunted in the MCP-1 null mice compared with wild-type mice, suggesting that MCP-1 is necessary for the PTH anabolic effect.
The cortical bone compartment was relatively unaffected by the absence of MCP-1. The BMD and cortical thickness significantly increased in PTH-treated wild-type and MCP-1–/– mice compared with the corresponding vehicle-treated mice. There were no differences in femur length or endocortical perimeter. Periosteal perimeter was significantly increased in the PTH-treated wild-type mice but not in the PTH-treated MCP-1–/– mice (Table 1).
|Femur length (mm)||15.70 ± 0.10||15.63 ± 0.10||15.59 ± 0.12||15.23 ± 0.08|
|Bone mineral density (mg/cm3)||996.17 ± 7.51||1046.37 ± 11.26a||988.33 ± 8.08||1023.56 ± 7.98a|
|Cortical thickness (mm)||0.327 ± 0.007||0.363 ± 0.007a||0.336 ± 0.006||0.358 ± 0.007a|
|Periosteal perimeter (mm)||4.89 ± 0.08||5.18 ± 0.07a||4.95 ± 0.06||5.03 ± 0.05|
|Endocortical perimeter (mm)||2.84 ± 0.09||2.90 ± 0.07||2.84 ± 0.05||2.78 ± 0.09|
The C57/B6 mouse strain is skeletally mature at 4 months of age; however, peak bone mass may not be achieved until 6 months of age. To assess the effect of the anabolic response of PTH on older animals, 6-month-old male wild-type and MCP-1–/– mice were also treated with vehicle or hPTH[1-34], 80 mg/kg sc 5 days per week for 6 weeks. The same trends observed in the 4-month-old mice were also seen in the 6-month-old mice. In fact, the 6-month-old PTH-treated MCP-1–/– mice showed significantly lower total, trabecular, and cortical BMD compared with the PTH-treated wild-type mice. However, the increases in cortical thickness did not reach statistical significance (Table 2).
|Total BMD (mg/cm3)||556.27 ± 17.37||630.07 ± 19.58a||555.57 ± 20.81||567.88 ± 14.66b|
|Trabecular BMD (mg/cm3)||478.13 ± 21.67||557.71 ± 24.88a||479.90 ± 20.14||490.65 ± 14.54b|
|Femur length (mm)||15.69 ± 0.12||15.75 ± 0.12||15.77 ± 0.15||15.58 ± 0.16|
|Cortical BMD (mg/cm3)||984.90 ± 15.07||1071.85 ± 8.69a||987.59 ± 18.28||1023.10 ± 13.04b|
|Cortical thickness (mm)||0.345 ± 0.010||0.366 ± 0.017||0.330 ± 0.009||0.351 ± 0.008|
|Periosteal perimeter (mm)||4.99 ± 0.11||5.07 ± 0.12||5.19 ± 0.11||5.17 ± 0.10|
|Endocortical perimeter (mm)||2.82 ± 0.10||2.77 ± 0.12||3.12 ± 0.12||2.96 ± 0.13|
Differences between female PTH-treated wild-type and MCP-1–/– mice are not as evident as in the male mice. Female mice treated from 4 months of age (Supplemental Fig. S1; Supplemental Table S1) and 6 months of age (Supplemental Table S2) show some of the same trends observed with the male animals, eg, a lack of an increase in total and trabecular BMD in the female MCP-1–/– mice. Lower bone mass (7% to 10% bone volume for 4-month-old females compared with 13% to 16% bone volume for males at the same age) and the influence of estrogen may be contributing factors in the smaller changes observed in female mice.
Trabecular architecture of MCP-1–/– mice treated with PTH
Structural analysis of the trabecular bone of the 4-month-old male mice by µCT confirms the results seen by pQCT. Compared with vehicle treatment, trabecular bone volume increased by 27% in wild-type mice treated with PTH, whereas no increase was observed in PTH-treated MCP-1 null mice (Fig. 3B). PTH also significantly increased trabecular thickness in PTH-treated wild-type mice, but very little change was seen in MCP-1–/– mice treated with PTH compared with the vehicle-treated animals. No significant differences were noted in trabecular number or trabecular separation (Table 3). Similar trends were also seen in 4-month-old female mice; however, no changes in trabecular thickness were observed (Supplemental Table S3). It appears that the increased bone volume after PTH treatment in male wild-type mice is achieved by increasing trabecular thickness, a process that is absent in the male MCP-1 null mice.
|Trabecular Number (1/mm)||3.95 ± 0.10||3.78 ± 0.17||4.48 ± 0.91||3.90 ± 0.71|
|Trabecular Thickness (mm)||0.065 ± 0.001||0.083 ± 0.003a||0.064 ± 0.002||0.069 ± 0.002b|
|Trabecular Separation (mm)||0.255 ± 0.007||0.273 ± 0.007||0.221 ± 0.005||0.257 ± 0.005|
Histomorphometry of MCP-1–/– mice treated with PTH
Histomorphometric analysis of the 4-month-old male mice revealed that bone area (B.Ar/Tt.Ar) (Fig. 4A) was significantly increased in the PTH-treated wild-type mice compared with vehicle-treated mice (vehicle: 7.49 ± 1.15%; PTH: 12.16 ± 1.65%). In comparison, the PTH-treated MCP-1–/– mice showed only a slight nonsignificant increase in B.Ar/Tt.Ar with PTH treatment compared with vehicle (vehicle: 7.29 ± 0.91%; PTH: 8.15 ± 0.60%). The twofold increase in osteoclast surface (Oc.S) and osteoclast number (Oc.N) observed with intermittent PTH treatment in the wild-type mice was completely eliminated in the MCP-1 null mice (Fig. 4B, C). As shown in Fig. 4D, the lack of osteoclast activity in the PTH-treated MCP-1 null mice was paralleled by a reduction in osteoid formation (OS/BS); however, this decrease was not statistically different from the PTH-treated wild-type mice. Dynamic parameters demonstrate that mineral apposition rate (MAR) does not differ between PTH-treated wild-type and MCP-1–/– mice (Fig. 5A). There was a significant increase over vehicle with PTH treatment in wild-type as well as MCP-1–/– mice. Furthermore, mineralizing surface (MS/BS) and bone formation rate (BFR/BS) significantly increased in PTH-treated MCP-1–/– mice; however, the effect was attenuated compared with wild-type mice (Fig. 5B, C). Morphometric parameters in the vehicle-treated wild-type mice did not differ from vehicle-treated MCP-1 null mice. Similar results were observed in female mice; however, MS/BS and BFR/BS in the female MCP-1–/– mice were not significantly increased with PTH treatment (Supplemental Table S4).
Immunohistochemistry of tibias in PTH-treated MCP-1 null mice
Tibial sections of wild-type and MCP-1–/– mice treated from 4 months of age were immunostained for MCP-1 and the macrophage markers MAC-3 and F4/80. The mice were injected daily with 80 mg/kg hPTH(1-34) or vehicle (saline) for 6 weeks. The bones were collected 2 hours after the last injection. PTH-treated wild-type mice showed a considerable increase in MCP-1 expression with prominent staining in the bone-lining cells compared with vehicle treatment (Fig. 6A). Demonstration of monocytes and macrophages in the secondary spongiosa using MAC-3 or F4/80 antibodies revealed that there is very low expression of MAC-3 and residual expression of F4/80 in the MCP-1–/– mice, both with and without PTH treatment, in comparison to the wild-type mice (Fig. 6B). The wild-type mice show an increase in MAC-3 and F4/80 staining after PTH treatment.
Previous studies in our laboratory have shown that MCP-1 is the most dramatically upregulated gene after intermittent hPTH(1-34) injection in vivo.[20, 21] These results served as the basis for determining if such high-fold induction of MCP-1 mRNA is reflected in an increase in circulating levels of MCP-1 in vivo after PTH injection. We found that in rats treated with 80 μg/kg hPTH(1-34) for 14 days, the serum levels of MCP-1 significantly increase with peak levels almost double that of baseline values attained 2 hours after the injection. This is consistent with the time course of changes in MCP-1 levels observed with osteoblastic cells in culture. The magnitude of the increase in serum MCP-1 is less than the fold increase reported for mRNA in bone. This is not surprising because MCP-1 concentration likely increases locally in the bone microenvironment to act in a paracrine fashion to attract monocytes and preosteoclasts to the vicinity of the bone surface.
MCP-1 serum levels show a dose-dependent effect of PTH treatment. Low-dose PTH (40 μg/kg) showed a ramping up of serum levels with increasing duration of treatment, whereas higher doses (80 μg/kg) produced maximal serum levels of MCP-1 after one injection that did not change with increased duration of treatment. There is a positive association between serum MCP-1 levels and serum PTH levels as demonstrated in our study, where serum MCP-1 levels tended to be greater in women with higher than normal serum PTH levels. In the present study, we have not determined if the major source of the serum MCP-1 levels is from bone because other tissues produce and secrete MCP-1, most notably adipose tissue. However, the rapid response and clear correlation with the time course of mRNA levels observed in the bones of rats injected with PTH supports the notion that the increased MCP-1 is derived from PTH action on cells of the osteoblast lineage. We have also shown that sex differences are apparent in that female rats have lower basal levels of serum MCP-1 than male rats, as well as after treatment with PTH, suggesting that MCP-1 may be modulated by estrogen. Studies in both humans and animals have shown that estradiol suppresses MCP-1 expression.[37-39] In all, these studies provide quantitative evidence that MCP-1 is increased by PTH.
We performed immunohistochemistry to identify the cells in bone that produce MCP-1 in response to PTH treatment. Increased MCP-1 protein expression was evident in osteoblasts after the second PTH injection with maximal expression by the third injection in rats. We also observed production of MCP-1 in some osteocytes. Others have shown that MCP-1 is produced by osteoblasts in vivo.[25, 40, 41]
MCP-1 is produced by a variety of cell types such as fibroblasts, vascular endothelium, monocytic and microglial cells, and smooth muscle cells. Monocytes, macrophages, and T lymphocytes are a major source of MCP-1, possibly regulated through the action of other chemokines, cytokines, and growth factors.[25, 29] Also, osteoclasts and some mononuclear cells in the bone marrow were shown to express increased MCP-1 after PTH injections. Some of the stained cells may be the result of secreted MCP-1 bound to a receptor on a responsive cell rather than a source of the protein. Mature osteoclasts are known to express MCP-1. This is likely because of an increase in RANKL induction, and it would occur after the first injection of PTH in our treatment protocol. MCP-1 expression in osteoclasts is highly stimulated by RANKL, and MCP-1 expression by the osteoclast may further enhance osteoclastogenesis in an autocrine manner.
The skeletal effects of PTH are mediated by a G-protein–coupled receptor (PTH1R) located on cells of the osteoblast lineage, so PTH-stimulated bone resorption is orchestrated by factors expressed by these cells. Recent evidence suggests that osteocytes may also have a primary role in PTH-stimulated action on bone formation.[15-17] PTH affects osteoclast maturation and function indirectly by stimulating the expression of RANKL and M-CSF.[44, 45] RANKL stimulation by intermittent PTH is a transient event,[21, 44] and MCP-1 expression parallels that of RANKL, therefore osteoblastic and osteocytic expression of MCP-1 may assist RANKL in osteoclast recruitment, differentiation, and fusion of osteoclast precursors. The transient response is important in order to turn off the osteoclastogenic signal quickly so that bone resorption is of short duration. This resorptive activity is followed by increased numbers and activity of osteoblasts that build bone, resulting in a net gain in bone mass. Direct contact between preosteoclasts and osteoblasts is necessary for osteoclastogenesis; therefore, attracting cells of the monocyte/macrophage lineage close to the osteoblast is essential to promote osteoclast maturation. When MCP-1 is applied close to bone in vivo, monocyte infiltration is enhanced. Our recent experiments identifying monocytes and macrophages using CD68[ED-1], F4/80, and MAC-3 immunohistochemistry show that monocyte and macrophage numbers are increased in bone marrow in response to PTH, and this is abolished in the absence of MCP-1.
The multiple chemokine receptors found on monocytes and the range of receptor usage suggest that chemokines may have redundant functions. This redundancy is likely considering the high degree of structural similarity between the MCP family of chemokines and the responses of leukocyte subsets to these chemokines. Redundancy ensures a minimal sufficient level of monocyte attraction for survival. MCP-1–/– mice are viable, develop normally, and show no obvious differences in phenotype from wild-type mice. Lu and colleagues have shown that there are normal levels of circulating leukocytes and macrophages in MCP-1–/– mice. Although these mice exhibit no obvious problems in the unstressed state, exposure to antigenic and nonantigenic challenges in several inflammatory models revealed significant defects in monocyte/macrophage recruitment and immunological response. Similar differences have also been demonstrated in CCR2–/––deficient mice. These studies confirm that despite functional redundancy, MCP-1 and its receptor CCR2 are critical in recruiting macrophages to sites of inflammation and to respond to PTH.
In this study, we have examined the skeletal response of MCP-1 knockout mice to intermittent hPTH(1-34) treatment. We did not observe a basal bone phenotype in the MCP-1–/– mice compared with the wild-type mice. As mentioned previously, redundancy in the chemokine system may contribute to migration of preosteoclastic cells to permit normal bone development. However, like the immunological response, the skeletal system cannot respond effectively when challenged. The most compelling results of this study are that the anabolic effect of PTH is reduced in MCP-1–/– mice treated with hPTH(1-34) for 6 weeks, indicating that MCP-1 is an important mediator for this response. Trabecular BMD in PTH-treated MCP-1 null mice is less than PTH-treated wild-type mice. More importantly, analysis of trabecular architecture revealed that bone volume does not increase with PTH treatment in MCP-1–/– mice, whereas cortical bone was less affected by the absence of MCP-1. In 4-month-old male mice, cortical BMD and cortical thickness increase in PTH-treated MCP-1–/– mice similar to wild-type mice. These data are consistent with analysis of the bone phenotype of CCR2–/– mice that show an increase in cortical BMD in the lumbar vertebrae and distal femur but less of an increase in trabecular bone volume after intermittent PTH treatment. The results with CCR2–/– mice indicate that signaling through CCR2 mediates the anabolic effect of PTH on bone.
Remarkably, the significant increase in osteoclast surface (Oc.S/BS) and osteoclast number (N.Oc/BS) seen in the 4-month-old male wild-type mice treated with hPTH(1-34) was completely abolished in the MCP-1–/– mice, confirming that MCP-1 is a main factor in osteoclast formation. The lack of recruitment of monocytes in the absence of MCP-1 likely resulted in diminished osteoclast formation and, therefore, fewer trabecular sites of osteoclastic resorption. Staining of marrow monocytes and macrophages with F4/80 and MAC-3 show that there are fewer monocytes in this tissue in the MCP-1–/– mice compared with wild-type mice. Bone turnover is not increased in the MCP-1–/– mice as it is in wild-type mice. Several studies have established that osteoclastic resorption is necessary for the bone anabolic effect of PTH.[11, 13] Inhibiting osteoclast differentiation in an in vivo mouse model blunted the anabolic response of PTH. Clinical and preclinical studies with combined bisphosphonate and PTH treatment revealed that the anabolic action of PTH was diminished in comparison to PTH treatment alone.[12, 13] Osteoblast function does not appear to be compromised by the lack of MCP-1 in the null mice. Mineral apposition rate in the PTH-treated MCP-1–/– mice was not different from the wild-type mice, indicating that osteoblast activity was not diminished. Osteoid surface (OS/BS), mineralizing surface (MS/BS), and bone formation rate (BFR/BS) all increase in PTH-treated MCP-1–/– mice; however, the effect is attenuated compared with wild-type mice. We believe that this attenuated response is because of fewer resorbing sites, therefore initiating less recruitment of osteoblasts and subsequent bone formation.
In conclusion, our results show that MCP-1 is a key molecular mediator for the anabolic effects of hPTH(1-34) on bone. MCP-1, secreted by cells of the osteoblast lineage in response to PTH stimulation, is necessary for the recruitment of monocytes and preosteoclastic cells and assists in the formation of mature osteoclasts. The transient increase in osteoclastic activity leads to a subsequent increase in bone formation, resulting in a net increase in bone mass. The absence of MCP-1 eliminates PTH-induced osteoclast formation and bone resorption. Subsequently, fewer resorption lacunae are present for new bone formation, leading to decreased recruitment of osteoblasts. Thus, the anabolic effect of PTH is attenuated.
All authors state that they have no conflicts of interest.
This work was supported in part by NIH grants DK48109 and DK47420 to NCP.
Authors' roles: Study design: JT and NP. Data collection: JT, AV, ES, NB, JJ, and CB. Data analysis: JT, AV, and ES. Data interpretation: JT, NM, and NP. Draft manuscript: JT and NP. Manuscript revisions: JT, NP, NM, and CB. JT takes responsibility for the integrity of the data analysis.