Osteocytic Osteolysis in PTH-treated Wild-type and Rankl^−/− Mice Examined by Transmission Electron Microscopy, Atomic Force Microscopy, and Isotope Microscopy

Abstract

To demonstrate the ultrastructure of osteocytic osteolysis and clarify whether osteocytic osteolysis occurs independently of osteoclastic activities, we examined osteocytes and their lacunae in the femora and tibiae of 11-week-old male wild-type and Rankl^−/− mice after injection of human parathyroid hormone (PTH) [1-34] (80 µg/kg/dose). Serum calcium concentration rose temporarily 1 hr after PTH administration in wild-type and Rankl^−/− mice, when renal arteries and veins were ligated. After 6 hr, enlargement of osteocytic lacunae was evident in the cortical bones of wild-type and Rankl^−/− mice, but not so in their metaphyses. Von Kossa staining and transmission electron microscopy showed broadly demineralized bone matrix peripheral to enlarged osteocytic lacunae, which contained fragmented collagen fibrils and islets of mineralized matrices. Nano-indentation by atomic force microscopy revealed the reduced elastic modulus of the PTH-treated osteocytic perilacunar matrix, despite the microscopic verification of mineralized matrix in that region. In addition, ⁴⁴Ca deposition was detected by isotope microscopy and calcein labeling in the eroded osteocytic lacunae of wild-type and Rankl^−/− mice. Taken together, our findings suggest that osteocytes can erode the bone matrix around them and deposit minerals on their lacunar walls independently of osteoclastic activity, at least in the murine cortical bone. (J Histochem Cytochem 68: –XXX, 2020)

Introduction

Parathyroid hormone (PTH) accelerates bone formation and resorption depending on dosing and administration intervals.^1–3 While its anabolic and catabolic effects on bones are noticeable in the span of few days, PTH also appears to elicit tissue responses by the minute. One such response may be “osteocytic osteolysis,” which is a concept originally proposed by Bélanger in the 1960s as a consequence not only of elevated PTH concentrations but also of calcium-deficient dieting.⁴ Throughout the 20th century, a few studies hinted that osteocytes were able to both erode^5–8 and deposit mineralized matrices.^9,10 Still, the concept of osteocytic osteolysis was not further explored and remained obscure; the concept was even discredited, since isolated avian osteocytes did not form resorption lacunae when cultured on whale dentin.¹¹ Nevertheless, several contemporary reports suggested that osteocytes and their canaliculi are key to mineral homeostasis of the bone matrix^12,13 and have attempted to shed light on the phenomenon of osteocytic osteolysis.^14–21 Currently, the concept has been brought back to attention, for example, osteocytic osteolysis after PTH administration or hyperparathyroidism,^14,22,23 during lactation and its heavy demand for calcium,^18,24 and when related with vitamin D²⁵ and sclerostin.^26,27 In these studies, the hypothesis was that osteocytes could erode the surrounding mineralized bone matrix, and would also deposit minerals on the once-eroded lacunar walls.

However, many questions continue to revolve around the idea of osteocytic osteolysis: one would have to consider, for instance, the role of osteoclasts and the degree of renal calcium excretion and reabsorption after PTH administration. Indeed, PTH treatment activates osteoclastic bone resorption and the consequent abundant secretion of acids and proteolytic enzymes that could pass through osteocytic canaliculi, reach the osteocytic lacunae, and eventually dissolve the perilacunar bone matrix.²⁸ On the other hand, it is still unclear whether osteocytic osteolysis is involved in the regulation of serum calcium; to verify it, one would have to exclude renal calcium excretion/reabsorption as a variable. Another important aspect of osteocytic osteolysis is determining whether it happens throughout the skeleton or if it is rather site-specific.

Osteocytes can build functional syncytia wherein osteocytic processes interconnect via gap junctions with neighboring osteoblasts and osteocytes, thus forming the osteocytic lacunar canalicular system (OLCS).^29–36 We have previously demonstrated that mature, well-mineralized bone develop a well-arranged OLCS, while immature bone exhibits an irregular, disorganized OLCS.^37,38 In fact, a well-arranged OLCS appears to be a consistent trait of a functional syncytium.^34,39 One evidence in favor of that notion was provided by Knothe Tate et al., who showed that histological samples of human osteomalacia featured highly connected but irregular OLCS, while late-stage osteoporosis specimens revealed a loss of connectivity in the OLCS.³⁴ The canaliculi network of OLCS may convey nutrition and signaling molecules to distant osteocytes, thus serving as a pipeline for the transmission of small molecules and minerals from the extracellular fluid.^40–44 Osteocytic osteolysis, therefore, may result of the work of not one, but of a collective of osteocytes; that is why it is important to determine whether osteocytes from regular or irregular OLCS exhibit osteolysis in response to exogenous PTH.

In this study, to clarify whether osteocytic osteolysis occurs in association with or independently of osteoclastic activities after PTH administration, we decided to employ wild-type mice and receptor activator of nuclear factor-kappa B ligand-deficient mice (Rankl^−/− mice), which lack functional osteoclasts.⁴⁵ Renal arteries and veins were ligated to exclude the effect of renal calcium secretion/reabsorption on the measurement of serum calcium after PTH administration and, then, we examined osteocytes and their lacunae in PTH-treated wild-type mice and Rankl^−/− mice using transmission electron microscopy (TEM), isotope microscopy and nano-indentation with atomic force microscopy (AFM). Thus, our purposes was to determine whether osteocytic osteolysis takes place independently of osteoclastic function after PTH administration.

Materials and Methods

Animals

Eleven-week-old male ICR wild-type mice (n=92, CLEA Japan, Inc., Tokyo, Japan) and Rankl^−/− mice (n=30) kindly provided by Dr. Penninger from the Institute of Molecular Biotechnology of the Austrian Academy of Sciences, Vienna, Austria, were used in this study. The study followed the principles for animal care and research use set by Hokkaido University (Approved No. 15-0032 and 20-0019).

Blood Correction and Preparation of Histological Specimens of Wild-type Mice

For the measurement of serum concentration of calcium under anesthesia with an intraperitoneal injection of pentobarbital for body weight determination, the renal arteries and veins were ligated, blood was collected from wild-type mice at 0 min (just before PTH injection), 30 min, 1 hr, and 6 hr after a single injection of 80 µg/kg/dose of human PTH [1-34] (hPTH [1-34]) or phosphate buffered saline (PBS: control mice) into the external jugular vein (n=6 for each) of mice.

For histological examination, other wild-type mice were anesthetized as described above and were then injected with 80 µg/kg/dose of hPTH [1-34] directly into the external jugular vein, whereas control mice were injected with the same volume of PBS. After 1, 3, 6, and 9 h, control mice and PTH-treated mice (n=6 each) were perfused with 4% paraformaldehyde diluted in 0.1 M cacodylate buffer (pH 7.4) through the cardiac left ventricle. The tibiae and femora were immediately removed and immersed in the same fixative for 18 hr at 4C. Some samples were decalcified with solutions of 10% ethylenediamine tetraacetic disodium salt (EDTA-2Na) for light microscopic observation and 5% EDTA for TEM observation. The specimens decalcified with 10% EDTA were dehydrated in ascending ethanol solutions before paraffin embedding. For TEM observation, the femora and tibiae with and without decalcification were post-fixed with 1% osmium tetraoxide with a 0.1 M cacodylate buffer for 4 hr at 4C; then, they were dehydrated in ascending acetone solutions and embedded in epoxy resin (Epon 812, Taab, Berkshire, UK) as recently reported.⁴⁶ Ultrathin sections were prepared with an ultramicrotome (Sorvall MT-5000; Ivan Sorvall, Inc., Norwalk, CT) and were then stained with uranyl acetate and lead citrate for demineralized samples or left unstained for undecalcified specimens before TEM observation (Hitachi H-7100 Hitachi Co. Ltd, Tokyo, Japan) at 80 kV. For nano-indentation by AFM, the undecalcified specimens fixed with a paraformaldehyde solution were kept in cacodylate buffer until topography and elastic modulus were measured.

Blood Correction and Preparation of Histological Specimens of Rankl^−/− Mice

Rankl^−/−mice with the ligation of renal arteries and veins were injected with hPTH [1-34] (PTH-treated Rankl^−/−mice) or PBS (control Rankl^−/−mice) in the same regimen as implemented for wild-type mice, and 1 hr later (when there was a significant difference in the serum concentration of calcium in wild-type mice; Fig. 1), blood was collected for estimating the serum concentration of calcium (n=6 for each), and the femora and tibiae were extracted and fixed with 4% paraformaldehyde in 0.1 M cacodylate buffer for histological examination. Other Rankl^−/− mice injected with hPTH [1-34] were fixed with 4% paraformaldehyde solution at 6 hr after PTH injection, while control Rankl^−/− mice were injected with PBS instead of hPTH [1-34] and fixed as described above (n=6 for each). Some specimens were decalcified with 10% EDTA-2Na solution and dehydrated in ascending ethanol solutions before paraffin embedding. For TEM observation, the specimens with or without 5% EDTA-decalcification were post-fixed with 1% osmium tetraoxide and embedded in epoxy resin. Ultrathin sections were either stained with uranyl acetate and lead citrate or left unstained before microscopic observation. For the detection of calcein labeling, PTH-treated Rankl^−/− mice were intraperitoneally injected with 20 mg/kg of 0.1 % calcein solution three times at 2 weeks, 1 week, and 3 days before fixation (n=6). They were fixed with 4% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.4) and embedded in methyl methacrylate (MMA) before fluorescence microscopic observation.

Figure 1.

Serum concentration of calcium after PTH administration in wild-type mice. After ligation of renal arteries and veins, blood was corrected from mice at 0 min, 30 min, 1 hr, and 6 hr after PTH administration. Calcium level at 0 min is obtained immediately after the ligation without PTH administration. With ligation of renal arteries and veins, serum calcium level gradually increases in both control mice and PTH-treated mice. At 1 hr after PTH administration, however, the index shows slight but significant increase in the PTH-administered group when compared with the control group. Abbreviation: PTH, parathyroid hormone.

Topographic Images and Elastic Modulus Measured by Nano-indentation of AFM

The femoral cortical bone of wild-type mice with or without PTH administration was ground to be 0.5 mm of thickness and then sonicated in PBS. The surfaces of femoral cortical bone were polished using grinding wheels with grain sizes of 0.3 µm with irrigated PBS. The middle region of femora of the control specimens and specimens at 1 hr and 9 hr after PTH administration were examined (n=5, for each) by AFM (BioScope Catalyst™ AFM, Bruker Japan K.K. Nano Surfaces & Metrology Division Tokyo, Japan) thus allowing for the integration of AFM imaging with optical imaging, according to the methods used for mineralized tissues.^47–49 Topography and elastic modulus of osteocytic lacunae and the surrounding bone matrix were measured using the new PeakForce QNM^® mode.⁵⁰

Localization of ⁴⁴Ca Assessed by Isotope Microscopy

Wild-type mice administered with hPTH [1-34] with the regimen of 80 µg/kg/dose were kept for 5 days and were fed with regular diet and drinking water, and then, they were fed with 4.8 g/day of a diet containing 0.6% calcium, which is composed of 0.48% of calcium 44 (⁴⁴Ca; 2% of calcium is in the form of ⁴⁴Ca in nature, which is a stable isotope but rare in nature) and 0.12% of calcium 40 (⁴⁰Ca; 97% of calcium is in the form of ⁴⁰Ca in nature) for additional 2 days (in total, 7 days after PTH injection; n=6). Control mice were kept with regular diets for 5 days and then fed with the same content of ⁴⁰Ca for 2 days (n=6). Thereafter, the mice were anesthetized with an intraperitoneal injection of pentobarbital and fixed with a paraformaldehyde solution for being embedded in epoxy resin for TEM observation. Serial semi-thin sections of epoxy resin-embedded, undecalcified specimens of the middle region of tibial cortical bones were alternately mounted on silicon wafers and glass slides. Semi-thin sections mounted on silicon wafers were subjected to sputter coating with Au before observation under an isotope microscope^51–54 at Isotope Imaging Laboratory, Creative Research Institute, Hokkaido University, Sapporo, Japan.

We also counted the numbers of ⁴⁴Ca-labeled osteocytic lacunae and the total numbers of the lacunae in the oval-shaped observation field of isotope microscope images, which were selected in the middle region of the tibiae at random. The percentage of the ⁴⁴Ca-labeld lacunae was calculated by dividing the number of ⁴⁴Ca-labeld lacunae by the total number of all lacunae in the observation field of the control and PTH-administered cortical bone (n=6, each).

Immunohistochemistry for d1 and d2 Subunits of Vacuolar-type Proton Pump ATPase

After inhibition of endogenous peroxidase activity with 0.3% hydrogen peroxidase in methanol for 30 min, dewaxed paraffin sections were pretreated with 1% bovine serum albumin (BSA; Serologicals Proteins Inc. Kankakee, IL) in PBS (1% BSA–PBS) for 30 min. The sections were then incubated for 2–3 hr at room temperature with rabbit polyclonal antisera against d1 and d2 subunits of vacuolar-type proton pump ATPase (Immuno-Biological Laboratories. Co., Ltd, Fujioka, Japan; kindly provided by Daiichi-Sankyo Co. Ltd., Tokyo Japan) diluted at 1:100 with 1% BSA–PBS. This was followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit IgG (DakoCytomation, Glostrup, Denmark). For visualization of all immunoreactions, immune complexes were visualized using 3, 3’-diaminobenzidine tetrahydrochloride (Dojindo Laboratories, Kumamoto, Japan). As for negative control experiments, normal rabbit serum instead of antisera against d1 and d2 subunits was employed as a primary antibody.

Von Kossa Staining

Epoxy resin sections from undecalcified specimens were incubated with an aqueous solution of silver nitrate until dark brown/black staining of the bone tissue was discernible under light microscopy.^13,55

Quantification of Enlarged Osteocytic Lacunae After PTH Administration Into the Wild-type Mice

The diaphyseal cortex of long bones shows typical lamellar bone featuring geometrically regular distribution of osteocytes, while the metaphyseal cortical bone does not show the lamellar structures with irregular distribution of osteocytes.³⁸ Based on our previous report,³⁸ therefore, the assumed boxes (0.7 mm in height, parallel to the longitudinal axis, and 0.2 mm in width from the endosteal surface) were located in the tibial cortical bone of wild-type mice 0.2 mm below the cartilage plate (a box in the cortical bone in the metaphysis) and the middle of the tibial length (a box in the cortical bone in the diaphysis). The putative boxed area in the metaphyses is 0.7 mm in height and 0.9 mm in width, located 0.1 mm apart from the bottom of the center of the growth plate cartilage (see Fig. 3). We regarded the osteocytic lacunae as enlarged when they were distinctly larger with toluidine blue-stained areas (Compare Fig. 3D and F). The percentage of enlarged lacunae was calculated by dividing the number of enlarged lacunae by the total number of all lacunae in the assumed boxes.

Figure 3.

The areas for quantification and the percentage of enlarged osteocytic lacunae in control and PTH-administered tibiae. The assumed boxes (α, β, γ) in panels A, B, C, E are the region of quantification of enlarged osteocytic lacunae in metaphyseal trabeculae (α), as well as in the cortical bone of the metaphyses (β) and diaphyses (γ) of the control (A, C) and PTH-treated tibiae (B, E) (See Materials and Methods). Panels G, H, I show the percentage of enlarged lacunae in the metaphyseal trabeculae (G), metaphyseal region of the cortical bones (H), and diaphyseal region of the cortical bones (I) between the control and PTH-administered tibiae. The value is shown as mean ± standard deviation. Values of p<0.05 were considered significant. Bars: A, B, C, E: 0.4 mm, D, F: 10 µm. Abbreviation: PTH, parathyroid hormone.

Statistical Analysis

Statistical analysis was assessed by Student’s t-test for the percentages of the enlarged lacunae and the serum concentration of calcium in wild-type mice and Rankl^−/− mice, and all values are presented as mean ± standard deviation. Values of p<0.05 were considered significant.

Results

Serum Concentration of Calcium After PTH Administration in Wild-type Mice

Serum calcium was measured after PTH administration followed by ligation of renal arteries and veins. Consequently, calcium concentration rose in a time-dependent fashion both with and without PTH treatment (Fig. 1). However, at 1 h, PTH-treated mice showed a slight but significant increase in serum calcium.

Histological Examination of Osteocytic Lacunae After PTH Administration

Next, we examined the histology of PTH-administered metaphyseal trabeculae and cortical bone, which reportedly show irregular and regular OLCS distributions, respectively.^37,38,44 Without PTH injection, the histological features of osteocytes and their lacunae were similar in the metaphyseal trabeculae and in the cortical bone (Fig. 2A and B). In contrast, after PTH injection, osteocytic lacunae were enlarged only in the cortical bone, with wider pericellular spaces (Fig. 2F and G). Under phase-contrast microscopy, enlarged osteocytic lacunae were consistently seen after PTH injection (compare Fig. 2C and H). There are two isoforms of d subunits in the vacuolar-type ATPase: the d1 subunit is reported to be ubiquitously expressed, whereas the expression of d2 is restricted to osteoclasts.⁵⁶ Here, we examined the immunoreactivity for subunits d1 and d2 in mice after PTH administration. Subunit d1 immunoreactivity was widely observed, including in mouse osteoclasts, regardless of PTH treatment (data not shown). Conversely, only osteoclasts showed subunit d2 immunoreactivity in control mice. However, after PTH injection, some osteocytes in the cortical bones revealed immunoreactivity for subunit d2 (compare Fig. 2D, E, I, and J). The control experiments did not show any immunoreactivity (data not shown).

Figure 2.

Histological examination of osteocytic lacunae after PTH administration. Panels A, B, F, and G show toluidine blue staining of metaphyseal trabeculae (A, F) and the middle region of cortical bones (B, G) in control (A, B) and PTH-injected femora (F, G). Note that the configuration of osteocytes was matched with their lacunae in both metaphyses and cortical bones (See insets of A and B); however, some osteocytic lacunae in the PTH-treated cortical bone appear enlarged and show while opening areas (an arrowhead in an inset of G). Images of phase-contrast microscopy show expanded osteocytic lacunae in the middle region of PTH-administered femoral cortical bone, but not in the matched region of the control counterparts (compare arrows in panels C and H). Unlike control specimens, d2 subunit-immunopositive osteocytes (ocy) are seen in the middle region of femoral cortical bone after PTH administration (I, J) but not in control specimens (D, E). Bars: A, B, E, F, G, J: 10 µm, C, D, H, I: 20 µm. Abbreviation: PTH, parathyroid hormone.

We have attempted to estimate the percentage of enlarged lacunae that was examined in the three different regions—metaphyseal trabeculae, metaphyseal region of the cortical bone, and diaphyseal region of the cortical region as shown in Fig. 3. The percentage of enlarged lacunae was 7.47 ± 0.38 (control specimens) and 8.90 ± 0.79 (PTH-treated specimens) in the metaphyseal trabecular bones with no significance (Fig. 3G), 8.21 ± 1.82 (control specimens) and 8.89 ± 0.79 (PTH-treated specimens) in the metaphyseal region of the cortical bone with no significance (Fig. 3H), and 4.85 ± 0.28 (control specimens) and 24.69 ± 2.25 (PTH-treated specimens) in the diaphyseal region of the cortical bone with significance (p=0.0000054) (Fig. 3I).

Ultrastructure of Demineralized Bone Matrix Surrounding Osteocytes After PTH Administration

As shown in Fig. 4, at 3 hr after PTH injection, demineralization of the bone matrix surrounding osteocytes was noticeable in the central portion of cortical bones (Fig. 4C and D), different from control specimens (Fig. 4A and B). Nine hours later, demineralization was evident in the vicinity of some osteocytes (Fig. 4E and F). Under TEM observation, PTH-injected specimens showed enlarged osteocytic lacunae containing amorphous organic materials, fragments of mineral matrix and irregular walls (Fig. 4G). In addition, some osteocytes featured small vesicles close to their cell membranes (Fig. 5A and B). Collagen fibrils near these vesicles were thinned, and some vesicles appear fused with the cell membrane, suggesting the secretion of components produced intracellularly. Other osteocytes were nested in enlarged lacunae, which contained amorphous materials, thin collagen fibrils, and small islets of mineral crystals (Fig. 5C to E).

Figure 4.

Demineralized bone matrix surrounding osteocytes after PTH administration. Panels A to F demonstrate von Kossa staining of the middle region of tibial cortical bone, and panels B, D, and F are highly magnified images of A, C, and E, respectively. In control specimens, the configuration of osteocytes (ocy) is matched with their lacunar walls, which are surrounded by a well-mineralized matrix (dark brown color) (B). Conversely, at 3 hr after 80-µg/kg PTH treatment, blue-colored organic components (a white arrow in the inset of panel D) are exposed in the periphery of osteocytic lacunae (C and D). At 9 hr after PTH administration, broad, blue-colored organic components are shown to be expanded (E and F). Note that the blue area indicates organic components, while dark brown color indicates mineralized matrix (a white arrow in F). Under TEM observation, as shown in panel G, the walls of osteocytic lacunae are irregular, featuring a broad demineralized area (asterisks) in the middle region of tibial cortical bone. Please note the many fragmented mineralized islets (white arrows in an inset). Bars: A, C, E: 10 µm, B, D, F: 5 µm, G: 2 µm. Abbreviation: PTH, parathyroid hormone.

Figure 5.

Ultrastructure of osteocytes and surrounding materials after PTH administration. When observing the middle region of the PTH-treated tibial cortical bone under TEM, some osteocytes (ocy) are located slightly enlarged lacunae (A), featuring small vesicles close to the cell membrane (red arrows, B). At higher magnification, markedly thinned collagen fibrils (an asterisk in B) are associated with the cell membrane of the osteocyte, and some vesicles seem to be fused with the cell membrane (black arrowheads, B). Other PTH-treated osteocytes were shown to be located in the enlarged lacuna (C), including amorphous materials (an asterisk, C). When observing at higher magnification, the small diameter of collagen fibrils (an asterisk, D) and small islets of mineral crystals (See arrows and inset, E) are discernible. Bars: A: 2 µm, B: 1 µm, C: 1.6 µm, D, E: 0.5 µm. Abbreviation: PTH, parathyroid hormone.

Nano-indentation of the Bone Matrix Surrounding Osteocytic Lacunae Assessed by AFM

Using AFM, the most commonly applied method of testing a tissue’s mechanical properties,⁵⁷ we assessed the topography and the elastic modulus of the bone matrix peripheral to osteocytic lacunae, although that bone matrix was not evidently eroded (Fig. 6). In control specimens, AFM showed images matching the topography with the elastic modulus of the osteocytic lacunae and their surrounding matrix (Fig. 6A). In contrast, 1 hr after PTH treatment, the elastic modulus surrounding a few of lacunae was slightly smaller despite the presence of perilacunar bone matrix (Fig. 6B). Nine hours after PTH administration, a markedly reduced elastic modulus was detected in the periphery of some osteocytic lacunae, even though the lacunae did not seem evidently enlarged when their topography was considered (Fig. 6C).

Figure 6.

Nano-indentation by AFM on bone matrix surrounding osteocytic lacunae. Panels A, B, and C represent the topography (left panels) and the elastic modulus (right panels) in the middle region of femoral cortical bone of the control specimen (A) and the specimens at 1 hr (B) and 9 hr (C) after PTH administration. In control specimens, the images of topography (the upper and lower panels show two- and three-dimensional images, respectively) are matched with the elastic modulus of osteocytic lacunae and their surrounding matrix (A). Please note that in the graphs, both arrows in the topography and elastic modulus indicate the ridge at the same height. Red lines indicate the margins of the lacunae. However, at 1 hr after PTH administration, the elastic modulus surrounding the lacuna was relatively reduced (B). Note the slightly expanded diameters of two- and three-dimensional images of the elastic modulus and also notice that the index of the elastic modulus is lower than the topography in the graph (see black arrows). At 9 hr after PTH injection, the markedly reduced elastic modulus is observed in the graph, which is consistent with the reduced size of osteocytic lacunae seen in two- and three-dimensional images (C). Abbreviations: AFM, atomic force microscopy; PTH, parathyroid hormone.

Enlarged Osteocytic Lacunae in Rankl^−/− Mice After PTH Administration

One hour after PTH injection and after ligation of renal arteries and veins, serum calcium was significantly higher in PTH-treated Rankl^−/− mice than in control Rankl^−/− mice (Fig. 7A). Consistent with the histological findings in wild-type mice, it was possible to see a larger area of demineralized bone matrices surrounding the osteocytic lacunae in PTH-treated Rankl^−/− mice (Fig. 7B to E). Under TEM observation, osteocytic lacunae in the cortical bone of PTH-treated Rankl^−/− mice were somewhat enlarged and featured rather irregular walls compared to control Rankl^−/− mice (Fig. 7F and G). Amorphous materials, which appeared to be degraded collagen fibrils, were present in the enlarged osteocytic lacunae seen after PTH administration.

Figure 7.

Enlarged osteocytic lacunae of Rankl^−/− mice after PTH administration. At 1 hr after PTH injection, the serum calcium level was significantly higher in PTH-treated Rankl^−/− mice (with ligation of renal arteries and veins) than in control Rankl^−/− mice (which received the ligation of renal arteries and veins but not PTH treatment) (A). At 6 hr after PTH administration, demineralized bone matrices surrounding the osteocytic lacunae in the middle region of Rankl^−/− tibial cortical bone appear expanded (D, E), when compared with the control counterparts without PTH (B, C). Note that a blue-colored area (a white arrowhead, E) can be seen in the vicinity of the osteocyte (ocy). Under TEM observation, Rankl^−/− osteocytic lacunae are enlarged after PTH administration (F) when compared with the control specimen without PTH treatment (G). The amorphous organic materials are seen in the PTH-treated lacuna (an asterisk, G), and bundles of collagen fibrils are seen in the lacuna wall (white arrowheads, G). Bars: B, D: 10 µm, C, E: 5 µm, F: 1.2 µm, G: 1.4 µm, H: 0.8 µm. Abbreviations: PTH, parathyroid hormone; TEM, transmission electron microscopy.

Mineral Deposition on the Walls of Osteocytic Lacunae

Seven days after PTH administration in wild-type mice, TEM observation showed the presence of electron-dense lines sprouting from the irregular walls of the osteocytic lacunae (Fig. 8A). Isotope microscopy observation revealed ⁴⁴Ca deposition in the vicinity of some enlarged osteocytic lacunae in the PTH-treated wild-type cortical bone (Fig. 8B and C). ⁴⁴Ca deposition was seen in the region corresponding to the toluidine blue-stained peripheral bone matrix of osteocytic lacunae (Compare Fig. 8B and C), implying that ⁴⁴Ca had been deposited on the once-demineralized bone matrix surrounding the osteocytic lacunae. From the statistical analysis, as shown in Fig. 8D, the percentage of ⁴⁴Ca-labeled osteocytic lacunae was 3.94 ± 2.50 versus 17.90 ± 4.36 in the control and the PTH-treated specimens (p=0.0194), respectively. Consistent with the wild-type mice, mineral deposition at the osteocytic lacunae in cortical bones of Rankl^−/− mice was examined after 7 days of PTH administration. Under TEM observation, amorphous electron-dense lines were visible outside the osteocytic lacunae in the cortical bone (Fig. 9A). Under fluorescence microscopic observation, calcein labeling was detected on the walls of the osteocytic lacunae in Rankl^−/− mice, suggesting that mineral deposition was taking place in the lacunae (Fig. 9C and D).

Figure 8.

Mineral deposition onto the walls of wild-type osteocytic lacunae after PTH administration. As shown in panel A, some osteocytic lacunae in the middle region of tibial cortical bone are encompassed by external amorphous electron-dense materials (white arrows) encompassing the lacunar walls (black arrows). There are lysosomes (Ly) in the osteocytes (ocy) and amorphous organic materials (an asterisk), and their lacunar walls are still irregular. Panels B and C are light microscopic images of an undecalcified, toluidine blue-stained section (B) and ⁴⁴Ca localization assessed by isotope microscopy (C) at 7 days after PTH injection in wild-type mice. White labeling representing ⁴⁴Ca deposition (C) is seen in the region of toluidine blue-stained area in the periphery of the osteocytes (ocy; white arrows in B and C). Panel D is statistical analysis on the percentage of ⁴⁴Ca-labeled osteocytic lacunae in the middle region of tibial cortical bone between the control and PTH-administered specimens. The value is shown as mean ± standard deviation. Values of p<0.05 were considered significant. Bars: A: 2 µm, B, C: 20 µm. Abbreviation: PTH, parathyroid hormone.

Figure 9.

Mineral deposition onto the walls of Rankl^−/− osteocytic lacunae after PTH administration. After 7 days of PTH administration, TEM observation revealed external, amorphous, electron-dense bands (black arrows) encompassing the lacunar walls (white arrowheads) in the Rankl^−/− tibial cortical bone. Panels B and C are images of an osteocyte (ocy) in the middle region of cortical bone under dark field (B) and bright field (C) of fluorescence microscopy. Please note the calcein labeling (green color) in the periphery of the osteocyte (See both arrows in panels B and C). Bars: A: 2 µm, B, C: 20 µm. Abbreviations: PTH, parathyroid hormone; TEM, transmission electron microscopy.

Discussion

In this study, we showed that after PTH administration (1) serum calcium rose temporarily in mice after ligation of renal arteries and veins, even with the halted osteoclastic bone resorption seen in Rankl^−/− mice, (2) some degree of bone demineralization is evident in the periphery of osteocytic lacunae in the cortical bones of wild-type and Rankl^−/− mice, but not in the metaphyseal trabeculae, (3) the elastic modulus of perilacunar bone matrices was lower, although bone matrix was present in a corresponding area of the topographic image, and (4) calcein labeling and ⁴⁴Ca isotope were detected on the walls and in periphery of osteocytic lacunae, which was suggestive of bone mineral deposition. There has been speculation that osteclastic bone resorption produces acids and proteolytic enzymes that might pass through the osteocytic canaliculi and erode the bone matrix peripheral to the lacunar and canalicular network.²⁸ While that was an elegant hypothesis, this study revealed that the widening of osteocytic lacunae and the calcium deposition seen after PTH administration would still occur in the absence of osteoclastic activity, as shown in the PTH-treated Rankl^−/− specimens. Thus, it is safe to assume that osteoclasts do not participate in PTH-driven osteocytic osteolysis, thus strengthening the concept of osteocytic osteolysis as described elsewhere.^{4–10,14–27}

Still, not all osteocytes generated osteolysis in their lacunae after PTH administration. Here, enlarged lacunae were seen in the cortical bone, where regularly arranged OLCS can be seen.^37,38,44 On the other hand, lacunar size seemed unchanged in the metaphyseal trabeculae, where a disorganized OLCS is the norm. When properly organized with highly connected cytoplasmic processes, the OLCS could develop a regularly arranged continuous passageway of osteocytic canaliculi/cytoplasmic processes, which may enable the OLCS to act as a functional syncytium, likely transporting small mineral particles and sensing external physicochemical stimuli.^42,58 Based on our findings, we conjectured that osteocytic osteolysis may occur when regularly arranged osteocytes, that is, in the cortical bone, respond to a single injection of PTH by eroding the inner walls of their lacunae, thus widening the space around them. However, the biological and physiological significance of osteocytic osteolysis is still unknown and the reason why only a small group of osteocytes in the cortical bone are triggered to enlarge their lacunae needs further exploration.

Taking into account our observations that the increase in serum calcium after PTH administration was time-dependent and that only osteocytes from a particular region eroded their surrounding matrix, one may conclude that the kidney—not the osteocytes—is the major regulator of serum calcium concentrations. Regarding the measurement of serum calcium concentration, we limited time to at most 6 hours to prevent systemic complications, such as uremia, in the mice after the ligation of renal arteries and veins. Still, it is interesting that serum calcium was lower in Rankl^−/− mice than it was in wild-type mice. Although osteoclastic bone resorption may be, in part, involved in the regulation of serum calcium, osteocytic osteolysis does not seem to be involved in the regulation of serum calcium concentration. Alternatively, it might be involved in a cascade that leads to focal remodeling of mature bone in response to hormonal and mechanical stimuli. Exciting possibilities and tentative explanations abound, but further studies are necessary to unveil the biological significance of osteocytic osteolysis.

Von Kossa staining and TEM observation demonstrated that, after PTH administration, the quantity of demineralized perilacunar matrix increased gradually, as amorphous materials—degraded organic components such as collagen—accumulated in the widened lacunae. Such finding suggests that osteocytic demineralization is preceded by the breakdown of organic components of the perilacunar matrix, presumably driven by acids secreted by osteocytes. Indeed, in our immunoreactivity experiments showed that osteocytes in PTH-treated cortical bones stained positive for subunit d2 of the vacuolar-type ATPase. Since subunits d2 are expressed mainly in osteoclasts,⁵⁶ it is intriguing that osteocytes would express these specific subunit after PTH injection. This apparently minor finding has its meaning expanded if one takes into account earlier reports of osteocytes synthesizing tartrate-resistant acid phosphatase and cathepsin K, which are osteoclast-derived enzymes.^18,26 In our study, however, we could not detect tartrate-resistant acid phosphatase nor cathepsin K in osteocytes after PTH administration (data not shown). As shown in this study, immunolocalization of the d2 subunit of V-ATPase would presumably induce acidic micro-circumstance in the osteocytic lacunae, which may result in the dissolving of bone minerals. However, it seems difficult to define how osteocytes would digest surrounding extracellular matrices in a way that is different from osteoclastic bone resorption. Thus, investigations on which types of proteolytic enzymes are secreted from osteocytes seem worthwhile, as much as studies on the specifics of how osteocytes degrade their surrounding bone matrix.

Regardless of the type of proteolytic enzymes produced by them, it seems that osteocytes dissolve the components of bone matrix in a way that is challenging to detect under conventional light microscopy. Nano-indentation by AFM, on the other hand, is a powerful tool for assessing subtle changes in the mechanical properties of the bone matrix surrounding the osteocytic lacunae. The mechanical weakening represented by the lower elastic modulus suggests that a small amount of mineral is removed from the bone matrix, as previously reported.⁵⁹ In this study, our initial aim to show the discrepancy between topography and elastic modulus by AFM was to verify, in the absence of visible defects by microscopic observation, that the physical properties, such as the elastic modulus, might be altered in the surrounding matrix of some osteocytic lacunae. As a consequence, we could provide clues for the discrepancy between topography and elastic modulus, even though it is not visible under microscopic observation. However, it was difficult, using our technique, to examine the topography and elastic modulus of all the osteocytic lacunae in the middle region of the tibial cortical bone using AFM, due to technical noise from organic materials. Therefore, in this study, we did not quantify the percentage of enlarged lacunae. Nevertheless, more detail and convincing data can be obtained by the statistical analysis on the quantification of the topography and the elastic modulus by, for instance, comparison with parameters obtained from light-microscopic analyses. In addition, in order to fastidiously demonstrate the elastic modulus, it is better to adjust, for example, the direction of observation lines to be parallel to the longitudinal axis of the tibiae or the direction of regularly oriented osteocytes and their cytoplasmic processes for measurement of the depth of lacunae (topography) and the elastic modulus between the control and PTH-administered specimens. Taken together, we can see that it is necessary to carefully display the quantification of the topography and the elastic modulus of osteocytic lacunae and the surrounding matrix in comparison to those parameters estimated by light-microscopic observation in future.

But can osteocytes deposit new bone? From the 1970s to the 1980s, the specialized literature alluded to the fact that osteocytes do form bone, albeit not as quickly as osteoblasts.^9,10,60 Our experiments showed calcein labeling in the cortical bone of Rankl^−/− mice as well as ⁴⁴Ca localization in the perilacunar matrix of wild-type cortical bones. This is evidence that osteocytes could deposit bone minerals onto the once-eroded lacunar walls. It appears to be a mineralization mechanism that is not matrix vesicle-mediated,^61–63 since we did not find matrix vesicles in the osteocytic lacunae after PTH administration (data not shown). Still, it is noteworthy that toluidine blue staining allowed for ⁴⁴Ca localization in perilacunar areas (Fig. 8), which correspond to the region where the bone matrix had been demineralized. Therefore, we have shown compelling evidence that mature osteocytes can deposit bone minerals on their lacunar walls.

In conclusion, the present study has collected several microscopic findings that strengthened the concept of osteocytic osteolysis, which is independent of osteoclastic activity, discernible in mature cortical bones with organized OLCS, and is subjected to a compensatory mineral deposition. Some of the questions raised here, however, warrant the need of further studies, particularly those concerning the biological significance of osteocytic osteolysis.