Tetracycline administration restores osteoblast structure and function during experimental diabetes.код для вставкиСкачать
THE ANATOMICAL RECORD 231:25-34 (1991) Tetracycline Administration Restores Osteoblast Structure and Function During Experimental Diabetes TAKAHISA SASAKI, HARUKI KANEKO, NUNGAVARAM S. RAMAMURTHY, AND LORNE M. GOLUB Second Department of Oral Anatomy, School of Dentistry, Showa University, Shinagawa-ku, Tokyo, Japan (T.S., H.K.), and Department of Oral Biology and Pathology, School of Dental Medicine, Health Sciences Center, State University of New York, Stony Brook, New York (N.S.R., L.M.G.) ABSTRACT Osteopenia is a recognized complication of diabetes mellitus in humans and experimental animals. We recently found that tetracyclines prevent osteopenia in the streptozotocin-induced diabetic rat and that this effect was associated with a restoration of defective osteoblast morphology (Golub et al., 1990). The present study extends these initial ultrastructural observations by assessing osteoblast function in the untreated and tetracycline-treated diabetic rats. After a 3-week protocol, non-diabetic control and diabetic rats, including those orally administered a tetracycline, minocycline (MC), or a non-antimicrobial tetracycline analog (CMT), were perfusion-fixed with a n aldehyde mixture; the humeri were dissected and processed for ultracytochemical localization of alkaline phosphatase (ALPase) and Ca-ATPase activities. Some rats from each experimental group received a n intravenous injection of 3H-proline as a radioprecursor of procollagen, and the humeri were processed for light microscopic autoradiography. In addition, the osteoid volume in each experimental group was quantitatively examined by morphometric analysis of electron micrographs. During the diabetic state, active cuboidal osteoblasts in the endosteum of control rats were replaced by flattened bone-lining cells that contained few cytoplasmic organelles for protein synthesis (Golgi-RER system), and active transport (mitochondria). Treating diabetic rats with MC, and even more so with CMT, appeared to “restore” osteoblast structure. During diabetes, bone-lining cells incorporated little 3H-proline or secreted little labeled protein and produced only a very thin osteoid layer. Tetracycline administration to the diabetics increased both the incorporation of 3H-proline by osteoblasts and their secretion of labeled protein toward the osteoid matrix, in a pattern similar to that seen in the non-diabetic controls. Intense alkaline phosphatase (ALPase) activity was cytochemically demonstrated along the plasma membranes of osteoblasts in the non-diabetic control rats, but was completely absent from the bone-lining cells in the diabetics. Similar to that described above, CMT therapy restored the ALPase activity in the diabetic osteoblasts and the effect of MC was less dramatic. The distribution and intensity of Ca-ATPase in the osteoblast-plasma membranes of the different groups of rats were similar to that of ALPase, except for the absence of detectable Ca-ATPase in the MC-treated diabetics. These results suggest that diabetes-induced osteopenia reflects, a t least in part, impaired osteoblast structure and function and that tetracyclines, by a non-antimicrobial mechanism, may prevent this bone deficiency by normalizing these bone-lining cells. Osteopenia (Osteoporosis) is a complication of insulin-deficiency diabetes in humans and experimental animals (Kaneko et al., 1990 for review). Because tetracyclines have been shown to inhibit bone resorption induced in vitro by a variety of factors (Golub et al., lgS4; Rifkin et lgS4, Comes et lgS8), we recently determined the effect Of these antibiotics (and their non-antimicrobial analogs) on the development of osteopenia in experimental diabetes. Of extreme inter0 1991 WILEY-LISS, INC est, preliminary studies indicated that in vivo administration of tetracycline prevented 1) the loss of bone density, 2) the loss of the osteoid layer, and 3) the mor- Received June 17, 1990; accepted February 22, 1991, Address reprint requests to Dr. T. Sasaki, Second Dept. of Oral Anatomy, School of Dentistry, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142, Japan. 26 T. SASAKI ET AL. phologic degeneration of osteoblasts, all associated with the diabetic state in rats (Golub et al., 1990).However, the mechanisms involved in ameliorating these diabetes-induced skeletal abnormalities are as yet unknown. Skeletal tissue metabolism is highly regulated by bone cells such as osteoblasts and osteoclasts. Multifunctional osteoblasts synthesize and secrete organic bone matrix consisting mainly of type I collagen (Frank and Frank, 1969;Wright and Leblond, 1981; Takagi et al., 1983) and regulate bone mineralization (Fleisch and Neiman, 1961;Yoshiki et al., 1972;Fukushima and Goshi, 1983;Akisaka et al., 1988).In addition, alkaline phosphatase (ALPase) and calciumtransporting ATPase (Ca-ATPase) of osteoblasts are thought to be involved in mineralization of bone matrix (Fleisch and Neiman, 1961;Yoshiki et al., 1972;Jaffe, 1976; Nijweide et al., 1981; Fukushima and Goshi, 1983; Fauran-Clavel and Oustrin, 1986; Register et al., 1986; Akisaka et al., 1988; Shen et al., 1988). ALPase is widely recognized as a marker enzyme for osteoblast and/or osteogenic cells (Yoshiki e t al., 1972; Wlodarski and Reddi, 1986). Ca-ATPase participates in the active transport of calcium ions to create local high concentrations of this mineral at the extracellular sites of calcification (Melancon and DeLuca, 1970; Fukushima and Goshi, 1983;Akisaka et al., 1988).To extend our initial studies (Golub et al., 1990) on the prevention of osteopenia, this report describes diabetes-induced alterations in procollagen production (assessed autoradiographically) and in the cytochemical markers of osteoblast activity (assessed ultrastructural cytochemistry) and the response to tetracycline therapy. MATERIALS AND METHODS Adult (4-month old) male Sprague-Dawley (Taconic Farms, Germantown, NY) rats (350-400 g body weight) were rendered diabetic by intravenous injection of streptozotocin (70 mg/kg body weight) as previously described (Golub et al., 1978).Some of these rats (4 rats per each group) were administered, by oral gavage, minocycline (Lederle Labs, Pearl River, NY) or 4-de-dimethylaminotetracycline (a chemically modified, non-antimicrobial analog in which the dimethylamino group on carbon-4 of the tetracycline molecule is removed), which was synthesized and characterized by us using techniques described previously (Golub et al., 1987;McCormick et al., 1957;Boothe et al., 1958). Minocycline and the chemically modified tetracycline analog (CMT) were suspended in 2% carboxymethylcellulose and 20 mg per day was administered to the appropriate rats throughout the 21-day protocol. The remaining diabetic and non-diabetic (control) rats were treated with vehicle (carboxymethylcellulose) alone. On day 21, the rats were fixed by intracardiac perfusion with a mixture of 1% formaldehyde and 1% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.3) for 10 min a t room temperature. After perfusion, the humeri were dissected and immersed in the same fixative for 1 h r a t 4°C. They were then demineralized in 10% disodium ethylenediamine tetracetic acid (pH 7.3) for 2 months at 4"C, and cross sections of 40-50 pm thickness were prepared from mid-diaphysis of the humeri using a microslicer (Dohan EM, Osaka). For alkaline phosphatase (ALPase) cytochemistry, sections were prepared as above and preincubated in 0.1 M buffer containing 10 mM magnesium sulfate for 2 hr. These slices were then incubated in the following media for 30 rnin a t room temperature (about 22°C). The medium for ALPase, according to Mayahara e t al. (1967),consisted of 28 mM tris-HC1 buffer (pH 8.5), 20 mM P-glycerophosphate, 3.9 mM magnesium sulfate, 2.0 mM lead citrate, and 8% sucrose. Final pH was adjusted a t 9.0-9.2. For cytochemical controls, some slices were incubated in a medium lacking substrate (P-glycerophosphate) or in a complete medium containing 10 mM levamisole, a potent inhibitor of ALPase. For calcium-transporting ATPase (Ca-ATPase) cytochemistry, the sections described above were reactivated in 0.1M sodium cacodylate buffer containing 10 mM CaC1, for 2 hr, and then incubated in the following media for 15 min at room temperature. The medium for Ca-ATPase, according to Ando e t al. (1981),consisted of 3.0 mM ATP (disodium salts, Sigma Chemical), 250 mM glycine-KOH buffer (pH 9.01,10 mM CaCl,, 4.0 mM lead citrate (dissolved in 50 mM KOH), and 5 mM levamisole. Final pH was adjusted to 9.0-9.2 using 10 M KOH. To ensure the substrate and calcium dependency of the enzymatic activity, some sections were incubated in a medium lacking either the substrate (ATP) or the enzyme activator CaCl,. After incubation, these sections were postfixed with 1.5% potassium ferrocyanide-reduced 1% osmium tetroxide for 15 min at 4°C and block-stained with ethanolated 1% uranyl acetate for 15 min. For conventional ultrathin sectioning, the bone tissues were fixed and demineralized as described above without preparation of microslicer sections for cytochemistry. All these specimens were dehydrated through a graded ethanol series and embedded in Quetol 812 (Nisshin EM, Tokyo). Thin sections were cut using a diamond knife on a Reichert-Jung Ultracut OmU-4, and stained with 1% tannic acid (pH 7.01,uranyl acetate, and lead citrate for morphological observation or lightly stained with uranyl acetate for cytochemistry. They were examined with a Hitachi H-800electron microscope at 75 kV. For light microscopic autoradiography, the rats were anesthetized with sodium nembutal and injected through the right or left jugular vein with phosphateproline (New Enbuffered saline containing (2,3-3H) gland Nuclear) in a dose of 20 mCiikg body weight. At 20 min and 4 h r after proline injection, the rats were perfusion fixed through the left ventricle with a mixture of 2% glutaraldehyde and 2% formaldehyde in a 0.1 M sodium cacodylate buffer (pH 7.3)for 10 min at room temperature. Dissected humeri were demineralized and embedded a s described above. Then 0.5-p.mthick sections were cut with a diamond knife, mounted on glass microscopic slides, and dipped in Kodak NTB2 emulsion diluted 1:2 with distilled water. After exposure for 7 days a t 4"C, the autoradiographs were developed in Dektol (Kodak Ltd.) for 2 min at 17"C, stained with 1% toluidine blue for 5 min a t 4WC, and examined with a n Olympus BHS light microscope. To evaluate the amount of osteoid produced by the osteoblasts in each experimental group, the osteoid Figs. 1-4. Ultrathin sections of osteoblasts and/or bone-lining cells in the non-diabetic control (l), untreated diabetic (21, MC-treated diabetic (3), and CMT-treated diabetic rats (4). Tannic acid-uranyl acetate-lead citrate staining. x 8,000. 28 T. SASAKI ET AL. Figs. 5-8 TETRACYCLINES AND DIABETES-INDUCED OSTEOPENIA 29 Golgi apparatus, RER, and mitochondria. Thus these bone-lining cells were almost completely devoid of cytoplasmic organelles necessary for protein synthesis Number of Osteoid volume osteoblasts and active transport (Fig. 2). The bone-lining cells in the diabetic rats exhibited examined Groups (pm2)/10pm * S.D. dramatic morphological changes in response to minocy36.8 * 6.0 28 Control cline (MC) administration (Fig. 3 ) and, even more so, in Untreated diabetes 3.4 k 0.7 22 response to the chemically modified non-antimicrobial 16.0 2 11.0 26 MC-treated diabetes tetracycline (CMT) (Fig. 4).In both the MC- and CMTCMT-treated diabetes 35.1 2 4.2 26 treated diabetic rats, the osteoid layer (very thin in the untreated diabetic rats) was again present on the mineralized bone surfaces. The bone-lining cells in the MCvolume (pm2) per unit osseous surface (10 pm) of os- treated diabetic rats exhibited a mix of flattened cells, teoblast was examined on the photographic prints pre- characteristic of the untreated diabetic bones, plus oval pared at a final magnification of 4,000-10,000 x normal-looking cells (Fig. 3 ) . The osteoblasts in the (original magnification of 2,000-5,000 x using a Pias CMT-treated diabetic group could not be distinguished LA-525R image analyzer attached to a NEC personal morphologically from those in the control group, both computer PC-9801 IVX. In each experimental group, groups exhibited active osteoblasts with similar cuboi22-28 osteoblasts were examined. Total examined ar- dal outline and a well-developed Golgi-RER system eas of osteoid were 858.8 pm2 (control), 67.6 pm2 (un- (Figs. 1, 4). treated diabetes), 462.7 pm2 (MC-treated diabetes) and Histomorphometric Analysis of Osteoid Volume 921.6 pm2 (CMT-treated diabetes), respectively. Quantitative analysis of osteoid volume in each exRESULTS perimental group revealed that experimentally inOsteoblast Ultrastructure duced diabetes resulted in a 91% reduction of osteoid Because some of the ultrastructural features of the volume. MC treatment partially corrected the reduced osteoblasts in the different groups were described pre- osteoid in the diabetics, whereas CMT therapy apviously (Golub et al., 19901, these are only reviewed peared to result in a complete restoration of this defect briefly herein. In the non-diabetic control rats, the (Table 1). mineralized bone surfaces of the periosteum of the hu3H-ProlineAutoradiography meri were covered with a n unmineralized osteoid layer In the non-diabetic controls, at 20 min after 3H-pro(Fig. 1).The osteoblasts showed a cuboidal and/or oval outline with numerous cellular processes extending to- line injection, the silver grains appeared over the cell wards the osteoid and into the bone canaliculi. The cell bodies of osteoblasts and fibroblasts in the periosteal bodies contained a well-developed Golgi apparatus, mi- surfaces of the humeri (Fig. 5a). The greatest uptake of tochondria, numerous cisternae of rough endoplasmic radioprecursor was seen in the osteoblasts. Young osreticulum (RER), and free polyribosomes (Fig. 1). Thus teocytes also incorporated radioprecursor but to a much osteoblasts in the control rats showed a cellular mor- lesser extent. A few grains were detected over the osphology consistent with actively synthesizing and se- teoid matrix but were rarely observed over the mineralized bone matrix. At 4 h r after 3H-proline injection, creting cells. In contrast, the active-looking osteoblasts were miss- the silver grains were concentrated over the osteoid ing and the adjacent osteoid volume was dramatically matrix facing the osteoblast layer (Fig. 5b). Some decreased on the bone surfaces of the untreated dia- grains were also observed over the osteoblast-cell bodbetic rats (Fig. 2). Instead, the mineralized bone was ies and the superficial but not the deeper layers of the covered with flattened bone-lining cells characterized mineralized bone matrix (Fig. 5b). In the untreated by flattened nuclei and deficient cytoplasm poor in diabetics, few if any silver grains were observed over the flattened bone-lining cells and bone matrix at either time period (Figs. 6a,b). At 20 min after isotope injection, in both MC- and CMT-treated diabetic rats, the silver grains appeared Fig. 5a,b. Autoradiographs of osteoblasts in humeri of non-diabetic over the more normal-looking osteoblast-like cell bodcontrol rats a t 20 min (Fig. a) and 4 h r (Fig. b) after 3H-proline injection. Silver grains are seen over the osteoblasts a t 20 min (Fig. a) ies that replaced the flattened bone-lining cells (Figs. 7a, 8a). At 4 h r after injection, a moderate amount of and over newly formed osteoidibone matrices a t 4 hr (b). x 600. silver grains were observed over the osteoblasts and Fig. 6a,b. Autoradiographs of bone-lining cells (arrowheads) of un- the osteoid matrix in the MC- and CMT-treated rats, treated diabetic rats at 20 min (Fig. a) and 4 hr (Fig. b) after isotope but the number of grains was fewer than that in the injection. Note the absence of silver grains over the tissues. X 600. non-diabetic controls (Figs. 7b, 8b). TABLE 1. Quantitative analysis of osteoid volume per unit osseous surface (10 am) of osteoblasts Fig. 7a,b. Autoradiographs of osteoblasts (arrowheads) in minocycline (MCbtreated diabetic rats at 20 min (Fig. a) and 4 hr (Fig. b) after isotope injection. Fewer silver grains are visible over the osteoid of these rats (b) compared to the pattern seen in the non-diabetic controls. x 600. Fig. 8a,b. Autoradiographs of osteoblasts (arrowheads) in CMTtreated diabetic rats a t 20 min (Fig. a) and 4 hr (Fig. b) after isotope injection. Silver grains can be seen over the osteoblasts (a) and over the osteoid (b). x 600. Alkaline Phosphatase (ALPase) In the control rats, a n intense enzymatic reaction for ALPase was demonstrated as electron-dense precipitates of lead phosphates along the extracellular aspects of the plasma membranes of the osteoblasts (Fig. 9). The cell processes penetrating the osteoid also exhibited a strong enzymatic reaction. The reaction products Figs. 9-1 2. Ultracytochemical localization of ALPase activity along the plasma membranes of osteoblasts and/or bone-lining cells in the non-diabetic control (9), untreated diabetic (lo), and MC-treated (111, and CMT-treated diabetic rats (12). Enzymatic reaction can be seen as electron-dense precipitates of lead phosphates a t the extracellular site of the plasma membranes. Unstained sections. x 10,000. TETRACYCLINES AND DIABETES-INDUCED OSTEOPENIA were also diffusely distributed in the osteoid layer close to the osteoblast-plasma membranes at the osseous cell surfaces (Fig. 9). Cytochemical controls demonstrated that the enzymatic activity was dependent on the presence of the substrate and sensitive to the enzyme inhibitor levamisole in the incubation medium (data not shown). The ALPase activity was completely absent from the flattened bone-lining cells in untreated diabetic rats (Fig. 10). However, when the diabetic rats were treated with MC, relatively weak ALPase activity was seen to be localized along the osseous and lateral cell surfaces of the flattened and/or oval osteoblasts. Tiny precipitates of enzymatic reaction were also noted in the osteoid layer (Fig. 11). In contrast, the osteoblasts in CMT-treated diabetic rats exhibited intense ALPase activity comparable to the pattern seen in osteoblasts of the control rats (compare Figs. 9 and 12). Ca-ATPase The enzymatic reaction of Ca-ATPase activity was demonstrated along the extracellular aspects of the plasma membranes of the osseous and lateral cell surfaces of osteoblasts in the control rats (Fig. 13).In cytochemical controls, omission of either substrate or the enzyme activator CaC1, from the incubation medium eliminated all evidence of this enzymatic reaction (data not shown). No evidence of reaction products could be seen in the atrophic bone-lining cells of the untreated diabetic rats (Fig. 14). Although the osteoblasts in MC-treated diabetic rats showed a more normal oval outline, they did not exhibit Ca-ATPase activity (Fig. 15). However, the osteoblasts in the CMT-treated diabetic rats showed intense Ca-ATPase reaction a s seen in the control rats (Fig. 16). DISCUSSION Structural and Cytochemical Changes of Osteoblasts During Experimental Diabetes Osteoporosis (osteopenia) has been described a s a complication of diabetes mellitus in humans and animals (Deleew and Abs, 1975, 1977; Levin et al., 1976; McNair e t al., 1979; Ramamurthy e t al., 1973; Santiago et al., 1977). Although the mechanisms responsible for this disorder of the skeletal tissues are not yet known, i t is generally agreed that decreased bone formation rather than accelerated osteoclastic bone resorption is a key factor (Shires et al., 1981). As a result of inducing diabetes, the unmineralized osteoid layer separating the mineralized bone from the osteoblast layer was dramatically reduced in volume, and cuboidal active osteoblasts appeared to be replaced by inactive flattened bone-lining cells, which were mostly devoid of cytoplasmic organelles necessary for protein synthesis (Golgi apparatus and RER) and active transport (mitochondria). In fact, these bone-lining cells incorporated little 3H-proline and secreted almost no labeled proteins. Mitochondria are involved in part in regulation of Ca-ATPase activity and cytosolic calcium concentration (Borle, 1973). Such bone-lining cells are deficient in cytoplasmic organelles and seem to be functionally inactive, since they produce little procollagen and lack both Ca-ATPase and ALPase activities, all 31 being necessary for matrix formation and mineralization. These osteoblastic functions are regulated by la,25dihydroxyvitamin D, (1,25(OH),D,), and acute and chronic streptozotocin (STZI-induced diabetes has been shown to decrease plasma 1,25(OH),D3 levels (Schneider e t al., 1977; Schedl et al., 1978; Hough et al., 1981;Shires e t al., 1981; Ishida et al., 1985). Schedl et al. (1978) suggested that a primary result of insulin deficiency in diabetes is a n abnormality in vitamin D, metabolism leading to impaired 1,25(OH),D, production. We recently found that, in the STZ-induced diabetic rat, osteoclasts lost their ruff led border-clear zone complex (Kaneko et al., 1990). This osteoclast abnormality in diabetes may result from a prior inactivation of the osteoblasts and decreased release of soluble cytokine(s) by these cells required for osteoclast differentiation (Heath et al., 1985; McSheehy and Chambers, 1986a,b; Thomson e t al., 1986). These results are consistent with earlier reports that diabetes-induced bone loss (osteopenia) results from impaired bone formation rather than increased osteoclastic bone resorption (Hough et al., 1981; Shires et al., 1981; Golub et al., 1990; Kaneko et al., 1990). Several possibilities may account for the above findings. Because insulin can increase the osteoblastic production of proteoglycans and collagen, the latter by increasing procollagen mRNA possibly by increasing its stability and half-life (Craig et al., 1989; Kream et al., 1985; Weiss et al., 1981), insulin-deficiency diabetes may inhibit the production of new matrix constituents by the atrophic osteoblasts (Klein et al., 1985; Schneir e t al., 1981). This inhibition of matrix formation was clearly confirmed in this morphometric study. A deficiency of insulin-like growth factor-1 (IGF-1) may also contribute to bone deficiency due to suppressed bone formation (Scheiwiller et al., 1986), although it has recently been shown that the administration of IGF-1 to diabetic rats does not restore bone growth and density to the same extent a s insulin therapy (Binz et al., 1990). Another possibility is that the osteoid deficiency might result from pathologically excessive collagenase activity, which would be expected to accelarate the degradation of newly synthesized bone collagen (Cowen et al., 1985; Delaisse et al., 1985). Such a n effect would also be consistent with the elevated collagenase activity seen in rat skin and gingiva during diabetes (Mohanam and Bose, 1983; Ramamurthy and Golub, 1983). Although collagenase in bone is now known to originate from osteoblasts, not osteoclasts (Sakamoto and Sakamoto, 1984), and although osteoblasts also produce plasminogen activator that can contribute to the activation of latent collagenase (Hamilton e t al., 19841, it is questionable whether the atrophic osteoblasts in the diabetic animals secrete increased levels of these proteins (enzymes) to excessively resorb the osteoid. At the present time, the relative contribution of decreased protein synthesis and increased extracellular collagen degradation to the loss of osteoid in the diabetic state cannot be assessed. The Effects of Antibiotic and Non-Antibiotic Tetracyclines on Osteoblasts in Diabetes-Induced Bone Disease In preliminary studies, we have reported that tetracycline administration can prevent the development of Figs. 13-1 6. Ultracytochemical localization of Ca-ATPase along the plasma membranes of osteoblasts and/or bone-lining cells in the non-diabetic control (13),untreated diabetic (14), and MC-treated (15),and CMT-treated diabetic rats (16).Unstained sections. X 10,000. TETRACYCLINES AND DIABETES-INDUCED OSTEOPENIA osteopenia in long bones of streptozotocin-induced, insulin-deficient, diabetic rats (Golub et al., 1990). The oral administration of doxycycline, a semi-synthetic antimicrobial tetracycline, to these diabetic rats appeared to increase the abnormally deficient bone mass, bone density, and mineral and organic constituents even though the severity of hyperglycemia was unaffected by the drug treatment (Golub e t al., 1990). The current study suggests that minocycline (MC) and its chemically modified non-antibiotic analogue (CMT) may ameliorate abnormalities in both osteoblast structure and function in diabetes-induced disease in mature rats. In diabetic rats treated with either MC or CMT: (1) the osteoid volume was much greater than that seen in untreated diabetic rats, (2) osteoblasts showed cuboidal shape with cytoplasmic organization similar to that in control rats, (3)the osteoblasts incorporated 3H-proline similar to those in control rats, and (4) the osteoblasts in CMT-treated rats also exhibited near-normal enzymatic activities of ALPase and CaATPase, which were severely deficient in diabetic osteoblasts (and less severely deficient in MC-treated diabetics). Only Ca-ATPase activity could not be detected in the osteoblasts in MC-treated diabetics. Although the MC-treated and CMT-treated diabetic rats were administered the same oral dose (20 mg per day) of these drugs, it is possible that the blood levels of the drugs resulting from these two treatments were different. In this regard, Yu e t al. (1990) recently demonstrated that the oral administration of CMT produced much higher blood levels than tetracycline-HC1. Thus it is suggested that CMT is absorbed more efficiently than MC as well. From the present study, it cannot be determined whether these similarities in cytological features of osteoblasts in normal and MC- or CMT-treated diabetic bones reflect a prevention of the development of diabetic changes or a return of diabetes-induced abnormalities to a more normal state. Another question currently being addressed is whether these antibiotics (which have long been known to bind metal ions) restore bone cell function by modulating cytoplasmic calcium levels; preliminary studies suggest t h a t this is possible a t least for osteoclasts (Rifkin et al., 1991). Consistent with the promotion of osteoid production in the mature animal by MC or CMT administration, Polson e t al. (1989) described a significantly increased deposition of bone matrix in the marrow spaces of the mandible of adult squirrel monkeys administered tetracycline. Schneir et al. (1990) also reported that MC administration increased the production of collagen in the atrophic skin of streptozotocin-diabetic rats. However, the drug’s ability to increase osteoid deposition in the diabetic’s skeletal tissue could also reflect the inhibition of osteoblast collagenase and a resultant decreased rate of degradation of newly secreted bone collagen. In fact, the in vivo tetracycline administration to STZ-induced diabetic rats reduced the pathologically-excessive collagenase activity in skin and gingiva and prevented the loss of skin collagen mass (Golub et al., 1983, 1984, 1987). Ramamurthy e t al. (1990) further demonstrated that tetracyclines can inhibit the collagenase secreted by osteoblasts in culture. In summary, it is suggested that (1) diabetes-induced bone loss (osteopenia) results from decreased protein syn- 33 thetic and phosphatase activities in osteoblasts and (2) tetracycline administration is effective in preventing (or reversing ?) this bone cell abnormality. ACKNOWLEDGMENTS This study was supported in part by a grant #DE03987 from the National Institute of Dental Research (NIH).The authors thank Ms. T. Shiraishi for her technical assistance. LITERATURE CITED Akisaka, T., T. Yamamoto, and C.V. Gay 1988 Ultracytochemical investigation of calcium-activated adenosine triphosphatase (CaATPase) in chick tibia. J. Bone Mineral Res., 3t19-25. Ando, T., K. Fujimoto, H. Mayahara, H. Miyayama, and K Ogawa 1981 A new one-step method for the histochemistry and cytochemistry of Ca-ATPase activity. Acta Histochem. Cytochem., 14:705-726. Binz, K., E.B. Hunziker, C. Schmid, and E.R. Frosch 1990 Osteoporosis in adult streptozotocin diabetic rats is cured by insulin but not by insulin-like growth factor I (IGF-I). Transact. Orthopaed. Res. SOC.,36th Ann. Meeting, New Orleans, p. 566. Boothe, J.H., G.E. Bonvicino, C.W. Waller, J.P. Petisi, R.W. Wilkinson, and R.B. Broschard 1958 Chemistry of the tetracycline antibiotics. I. Quaternary derivatives. J. Am. Chem. SOC.,80:16541657. Borle, A.B. 1973 Calcium metabolism a t the cellular level. Fed. Proc., 32t1944-1950. Cowen, K.S., M. Sakamoto, and S. Sakamoto 1985 Monensin inhibits collagenase production to osteoblastic cell cultures, and also inhibits both collagenase release and bone resorption in mouse calvaria cultures. Biochem. Int., llr273-280. Craig, R.G., D.W. Rowe, D.N. Petersen, and B.E. Kream 1989 Insulin increases the steady state level of a-l(I)procollagen mRNA in the osteoblast-rich segment of fetal rat calvaria. Endocrinology, 125: 1430-1437. Delaisse, J.M., Y. Eeckhout, C. Sear, A. Galloway, K. McCullah, and G. Vaes 1985 A new synthetic inhibitor of mammalian tissue collagenase inhibits bone resorption in culture. Biochem. Biophys. Res. Commun., 133r483-490. Deleew, I.H., and R. Abs 1975 Bone mass in diabetic. Diabetologia, 2t337-342. Deleew, I.H., and R. Abs 1977 Bone mass and bone density in maturity-type diabetics measured by the lz5I photon-absorption technique. Diabetes, 26:1130-1135. Fleisch, H., and W.F. Neiman 1961 Mechanism of calcification: role of collagen, polyphosphate and phosphatase. Am. J . Physiol., 200: 1926-1300. Frank, R.M., and P. Frank 1969 Autoradiographie quantitative de l’osteogenese en microscopie electronique a l’aide de proline tritiee. Z. Zellforsch 99,121-133. Frauran-Clavel, M.J., and J Oustrin 1986 Alkaline phosphatase and bone calcium parameters. Bone, 7t95-99. Fukushima, O., and N. Goshi 1983 Neutral ATP-hydrolyzing enzyme activity on the plasma membrane of osteoblasts. Acta Histochem. Cytochem., 16r216-222. Golub, L.M., R.A. Greenwald, E.J. Zebrowski, and N.S. Ramamurthy 1978 The effect of experimental diabetes on the molecular characteristics of soluble rat-tail tendon collagen. Biochim. Biophys. Acta, 534:78-81. Golub, L.M., T.F. McNamara, G. DAngelo, R.A. Greenwald, and N.S. Ramamurthy 1987 A non-antibacterial chemically-modified tetracycline inhibits mammalian collagenase activity. J. Dent. Res., 66t1310-1314. Golub, L.M., N.S. Ramamurthy, H. Kaneko, T. Sasaki, B. Rifkin, and T.F. McNamara 1990 Tetracycline administration prevents diabetes-induced osteopenia in the rats: Initial observations. Res. Commun. Chem. Pathol. Pharmacol., 68:27-40. Golub, L.M., H.M. Lee, G. Lehrer, A. Nemiroff, T.F. McNamara, R. Kaplan, and N.S. Ramamurthy 1983 Minocycline reduces gingival collagendlytic activity during diabetes: preliminary observations and a proposed new mechanism of action. J. Periodont. Res., 18r516-526. Golub, L.M., N. Ramamurthy, T.F. McNamara, B. Gomes, M. Wolff, A. Casino, A. Kapoor, J. Zambon, S. Ciancio, M. Schneir, and H. Perry 1984 Tetracyclines inhibit tissue collagenase activity. A new mechanism in the treatment of periodontal disease. J. Periodont. Res., 19:651-655. 34 T. SASAKI ET AL. Games, B.C., L.M. Goluh, and N.S. Ramamurthy 1984 Tetracyclines inhibit parathyroid hormone-induced bone resorption in organ culture. Experientia, 40t1273-1275. Hamilton, J.A., S.R. Lingelhach, N.C. Partridge, and T.J. Martin 1984 Stimulation of plasminogen activator in osteohlast-like cells by hone-resorbing hormones. Biochem. Biophys. Res. Commun., 122; 230-236, 1984. Heath, J.K., J. Saklatvala, M.C. Meikle, S.J. Atkinson, and J.J. Reynolds 1985 Pig interleukin l(cataho1in)is a potent stimulator of bone resorption in vitro. Calcif. Tissue Int., 37:95-97. Hough, S., L.V. Avioli, M.A. Bergfeld, M.D. Fallon, E. Slatopolsky, and S.L. Teitelhaum 1981 Correction of abnormal bone and mineral metabolism in chronic streptozotocin-induced diabetes mellitus in the rat by insulin therapy. Endocrinology, 108t22282234. Ishida, H., Y. Seino, S. Nishi, N.Kitano, M. Seno, T. Taminato, and S. Matsukura 1985 Effects of insulin on altered mineral and vitamin D metabolism in streptozotocin-induced diabetes. Acta Endocrinologica, 108t231-236. Jaffe, N.R. 1976 Alkaline phosphatase activity, characterization, and suhcellular distribution during initial skeletogenesis in the prenatal rat limb. Calcif. Tissue Res., 21:153-162. Kaneko, H., T. Sasaki, N. Ramamurthy, and L.M. Golub 1990 Tetracycline administration normalizes the structure and acid phosphatase activity of osteoclasts in streptozotocin-induced diabetic rats. Anat. Rec., 227:427-436. Klein. L.. X.I. Ying-Yan, and K.M. Wong 1985 Effect of short-term streptozotocin-Ynduced diabetes on b&e resorption, mineralization and collagen formation in the rat. Calcif. Tissue Int., 37:ahst. 107. Kream, B.E., M.D. Smith, E. Canalis, and L.G. Raisz 1985 Characterization of the effect of insulin on collagen synthesis in fetal rat bone. Endocrinology, 116t296-302. Levin, M.E., V.S. Boisseau, and L.V. Avioli 1976 Effects of diabetes militus on hone mass in juvenile and adult-onset diabetes. N. Engl. J . Med., 294t241-245. McCormick, J.R.D., S.M. Fox, L.L. Smith, B.A. Bitler, J. Reichenthal, V.E. Origoni, W.H. Muller, R. Winterbottom, and A.P. Doerschuk 1957 Studies of the reversible epimerization occurring in the tetracycline family. The preparation, properties and proof of structure of some 4-epi-tetracyclines. J. Am. Chem. SOC.,79t28482859. McNair, P., S. Madshad, C. Christiansen, M.S. Christiansen, O.K. Faher, C. Binder, and I. Transbo 1979 Bone loss in diabetes: effects of metabolic state. Diahetologia, 17t283-286. McSheehy, P.M.J., and T.J. Chambers 1986a Osteohlastic cells mediate osteoclastic responsiveness to parathyroid hormone. Endocrinology, 118t824-828. McSheehy, P.M.J., and T.J. Chambers 198613 Osteoblast-like cells in the presence of parathyroid hormone release soluble factor that stimulates osteoclastic bone resorption. Endocrinology, 119: 1654-1659. Mayahara, H., H. Hirano, T. Saito, and K. Ogawa 1967 The new lead citrate method for the ultracytochemical demonstration of nonspecific alkaline phosphatase. Histochemie, 1lt88-96. Melancon, M.J., and H.F. DeLuca 1970 Vitamin D stimulation of calcium-dependent adenosine triphosphatase in chick intestinal brush borders. Biochemistry 9,1658-1664. Mohanan, S., and S.M. Bose 1983 Influence of streptozotocin and alloxan induced diabetes in the rat on collagenase and certain lysosomal enzymes in relation to the degradation of connective tissue proteins. Diahetologia, 25t66-70. Niiweide. P.J.. E.W.M. Kawilarane de Haas. and A.M. Wassenaar 198f Alkaline phosphatase a& calcification, correlated or not? Metah. Bone Dis. Rel. Res., 3:61-66. Polson, A.M., O.J. Bouwsma, T.F. McNamara, and L.M. Golub 1990 Enhancement of normal hone density after tetracycline administration. J. Periodontal., (in press). ~ _1 ~~~ ~ Ramamurthy, N.S., and L.M. Goluh 1983 Diabetes increases collagenase activity in extracts of rat gingiva and skin. J . Periodont. Res., 17t455-462. Ramamurthy, N.S., E.J. Zehrowski, and L.M. Golub 1973 Alloxan diabetes and reduced hone density in rat mandible. Res. Commun. Chem. Pathol. Pharmacol., 5514-619. Ramamurthy, N.S., A.T. Vernillo, H. Lee, L.M. Goluh, and B. Rifkin 1990 Tetracyclines inhibit collagenase from osteohlastic rat UMR cells. J. Dent. Res., 69(Sp 1ss):Ahstr. 1932, p. 350. Register, T.C., F.M. McLean, M.G. Low, and R.W. Wuthier 1986 Roles of alkaline phosphatase and labile internal mineral in matrix vesicle-mediated calcification. J. Biol. Chem., 261t9354-9360. Rifkin, B., H. Donahue, K. Iijima, M. Oligorsky, and C. Ruhin 1991 Tetracyclines attenuate extracellular calcium-stimulated cytosolic calcium mobilization in osteoclasts. J. Dent. Res., 70(Sp Iss): ahst. (in press). Rifiin, B., F. Sanavi, A. Kleckner, B. Games, N.S. Ramamurthy, and L.M. Goluh 1988 Nan-antibacterial tetracycline inhibits hone resorption in organ culture. J. Dent. Res., 67(Sp Iss): ahst. 2068. Sakamoto, S., and M. Sakamoto 1984 Osteohlast collagenase: Collagenase synthesis by clonally derived mouse osteogenic (MC3T3E l ) cells. Biochem. Internat., 951-58. Santiago, J.V., W.H. McAlister, S.K. Ratzan, Y. Bussman, M.W. Haymond, G. Shakelford, and V.V. Weldon 1977 Decreased cortical thickness and osteopenia in children with diabetes mellitus. Clin. Endocrinol. Metah., 45:845-848. Schedl, H.P., H. Heath, and J . Wenger 1978 Serum calcitonin and parathyroid hormone in experimental diabetes: Effects of insulin treatment. Endocrinology, 103t1368-1373. Scheiwiller, E., H-P. Guler, J. Merryweather, C. Scandella, W. Maerki, J. Zapf, and E.R. Froesch 1986 Growth restoration of insulindeficient diabetic rats by recombinant human insulin-like growth factor I. Nature 323t169-171. Schneider, L.E., H.P. Schedl, T. McCain, and M.R. Haussler 1977 Experimental diabetes reduces circulating 1,25-dihydroxyvitamin D in the rat. Science, 196:1452-1454. Schneir, M., N. Ramamurthy, and L.M. Goluh 1981 Diabetes reducesmandibular collagen synthesis rather than enhancing its cataholism. J. Dent. Res., 60 (Sp 1ss):AGO: 1332. Schneir, M., N.S. Ramamurthy, and L.M. Goluh 1990 Minocycline treatment of diabetic rats normalizes skin collagen production and mass. Matrix: Collagen Rel. Res., 10t112-123. Shen, V., K. Hruska, and L.V. Avioli 1988 Characterization of a (Ca, Mg)-ATPase system in the osteohlast plasma membrane. Bone, 9t325-329. Shires, R., S.L. Teitelbaum, M.A. Bergfeld, M.D. Fallon, E. Slatopolsky, and L.V. Avioli 1981 The effect of streptozotocin-induced chronic diabetes mellitus on hone and mineral homeostasis in the rat. J. Lab. Clin. Med., 97:231-240. Takagi, M., R.T. Parmley, Y. Toda, and F.R. Denys 1983 Ultrastructural cytochemistry of complex carbohydrates in osteohlasts, osteoid, and hone matrix. Calcif. Tissue Int., 35t309-319. Thomson, B.M., J. Saklatvala, and T.J. Chambers 1986 Osteohlasts mediate interleukin 1 stimulation of hone resorption by rat osteoclasts. J. Exp. Med., 164t104-112. Weiss, R.E., A.H. Gorn, and M.E. Nimni 1981 Abnormalities in the biosynthesis of cartilage and hone proteoglycans in experimental diabetes. Diabetes, 3Ot670-677. Wlodarski, K.H., and A.H. Reddi 1986 Alkaline phosphatase a s a marker of osteoinductive cells. Calcif. Tissue Int., 39t283-285. Wright, G.M., and C.P. Lehlond 1981 Immunohistochemical localization of procollagens. J . Histochem. Cytochem., 29:791-804. Yoshiki, S., T. Umeda, and Y. Kurahashi 1972 An effective reactivation of alkaline phosphatase in hard tissues completely decalcified for light and electron microscopy. Histochemie, 29t296-304. Yu, Z., M. Leung, N.S. Ramamurthy, T. McNamara, and L.M. Golub 1990 Serum levels of chemically-modified tetracycline: A comparison to tetracycline. J. Dent. Res., 69(Sp Iss): abstr. 1092, p. 245.