Journal of Orthopaedic Research 986-75 Raven Press, Ltd., New York 0 1991 Orthopaedic Research Society
Effects of Propranolol on Bone Metabolism in the Rat Barbara Minkowitz, *Adele L. Boskey, *Joseph M. Lane, Hubert S. Pearlman, and ?Vincent J. Vigorita Department of Orthopaedics, Maimonides Medical Center, Brooklyn, *Departments of Ultrastructural Biochemistry and Collagen Tissue Research, Hospital for Special Surgery, New York; and fDepartment of Pathology, Lutheran Medical Center, Brooklyn, New York, U.S.A.
Summary: Propranolol, a nonspecific P-blocker, has many physiologic effects. Its effects on bone in vivo are unknown, although P receptor sites have been found on osteoblasts. In this study, the hypothesis tested was that low doses of propranolol could alter bone properties and enhance orthotopic endochondral bone formation. In a group of nonsurgical rats, propranolol treatment increased femoral torsional strength on biomechanical testing. In the rat surgical model used, right femora were fixed to a polyethylene plate and then defects were created mid-diaphysis and subsequently filled with demineralized bone matrix. These rats (defect rats) were randomly divided into groups that were given propranolol or a saline carrier for 19 consecutive days. In the defect rats, increased trabecular femoral metaphyseal mineral apposition rates were observed in propranolol-treated groups. Densitometry and roentgenographic scoring of callus formation after 12 weeks in propranolol-treated rats revealed increased callus and bone union. The results of this study indicate that propranolol treatment can significantly affect bone properties. Key Words: Propranolol-6-Adrenergic drug-Bone metabolism-Fracture healing-RatBone density.
The clinical effects of propranolol, a nonspecific P-blocker, include decreased hypertension, increased serum growth hormone levels, decreased thyroid hormone levels, decreased serum calcium levels, and increased vascular collagen crosslinking (8,13,20,21,29,33). In bone organ culture systems, propranolol has been shown to decrease resorptive effects of parathormone as well as other calciotropic hormones/agents (7,lO). In vivo propranolol affects the type I collagen in vascular tissues via interactions with lysyl-oxidase. This results in increased collagen cross-linking and increased tensile strength (2-4,24,28,30). In vitro, increased collagen synthesis in fetal lung fibroblast
culture was observed in the presence of propranolol (1). In vivo effects of P-blockers on bone are not well understood, although p receptor sites have been identified on osteoblasts in vitro (15,27). The significance of these receptor sites has never been studied in vivo or in vitro. In the present study, the hypothesis tested was that low doses of propranolol could alter bone properties and enhance orthotopic endochondral bone formation.
METHODS Experimental Groups One hundred thirty-nine male Sprague-Dawley rats (350 g) were purchased from Taconic Farms (Germantown, NY). They were divided into six treatment groups given propranolol or saline, with 40 rats serving as donors for the demineralized bone
Received August 30, 1989; accepted April 29, 1991. Address correspondence and reprint requests to Dr. B . Minkowitz care of A. L. Boskey at Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, U.S.A.
869
B. MINKOWITZ ET AL.
870
matrix needed for the rat femoral defect model. Three of the six groups, consisting of 12 animals each, were not manipulated (nonsurgical rats). The remaining three groups, consisting of 21 animals each, were surgically manipulated (defect rats). Treatments Intraperitoneal (i.p.) injections of saline or dilutions of injectable propranolol (Inderal, WyethAyerst, Philadelphia, PA) in saline were administered daily for 19 consecutive days. This treatment was started on postoperative day 1 for the defect animals. One group of nonsurgical and defect rats received saline, one nonsurgical and defect group received 50 pg propranolol(50-PRO), and the third nonsurgical and defect groups received 100 pg propranolol(lO0-PRO). Rats were fed Purina rat chow and were given water ad libitum. All experimental animals received humane care and treatment in compliance with the “Guide for the Care and Use of Lab Animals,” National Institutes of Health publication no. 86-23, revised 1985. Nonsurgical Rats Nonsurgical rats were all killed 12 weeks after the start of the treatments. Serum samples from nonsurgical rats were individually analyzed for calcium on a Beckman Astra autoanalyzer (Beckman Instruments, Fullerton, CA, U.S.A.) (14). The amount of T, present was analyzed using a T, radioimmunoassay (RIA) (9).
Defect Rats Demineralized Bone Matrix Preparation
Allogeneic demineralized bone matrix was prepared from combined femoral and tibia1 diaphyses cleaned of soft tissue and bone marrow, and subsequently lyophilized and then ground to a particle size of 45-300 pm in a Spex Mill (Spex Industries, Metuchen, NJ). This powder was extracted sequentially with 0.5 N hydrochloric acid, absolute ethanol, and ethyl ether to yield demineralized bone matrix (26). Defect Model
Three groups of 21 rats each underwent femoral defect formation (23). They were given chloral hydrate and procaine penicillin (60,000 U subcutaneous) before surgical manipulation. An incision was made laterally along the length of the right femur, which was then cleaned of soft tissue. A predrilled, high-density polyethylene plate (23 X 4 x 4 mm) was secured to the anterior surface of the femur by four threaded kirschner wires (1.2-mm diameter no. 26212 Zimmer). A 5-mm defect was created midfemur (twice the cross-sectional diameter) using a dental burr. Bone marrow on either side of the defect was flushed from the bone using a syringe loaded with isotonic saline. The defect was packed with the demineralized bone matrix powder (1 1). Rats were radiographed at 3-week intervals, to verify proper placement of the polyethylene plates. Histochemistry and Histopathology
Biomechanical Testing Both femora of nonsurgical groups were stabilized proximally and distally in a potting compound (cerubend alloy, Cenumetal Products, Bellefonte, PA, U.S.A.) for mechanical testing (5). Before testing, anterior-posterior and medial-lateral diameters of all bones were measured using a caliper. All bones were fractured in external rotation in a rapid load torsion tester (26). The stresshtrain curves generated were analyzed on an LCS Microplan I1 (Anaheim, CA) digitizing table connected to a minicomputer. Osseous biomechanical properties determined included the torsional strength and the stiffness (26). Right and left leg values were averaged to obtain a single value for each nonsurgical rat.
J Orthop Res, Vol. 9, No. 6 , 1991
Six rats in each defect group were labeled i.p. with tetracycline hydrochloride (86 pg/kg) on postoperative day 1 and with doxycycline (43 pg/kg) on postoperative day 10. These rats were killed on postoperative day 14 for evaluation of femoral metaphyseal trabecular mineral apposition rates in both operated and nonoperated femora. For histomorphometric studies, we chose to analyze mineral apposition rates, which are directly derived and do not require data manipulation, to evaluate changes in bone metabolism (25). Trabecular bone rather than cortical bone labels were analyzed because of artifactual distortion of cortical bone during processing. Undecalcified femoral metaphyses were coronally sectioned at a thickness of 5 pm using a heavy-duty microtome (Polycut S, Reichert-Jung,
PROPRANOLOL EFFECTS ON RAT BONE METABOLISM Heidelberg, Germany). Unstained sections were analyzed (identifications were masked by numbercoding the slides) using fluorescent microscopy (Nikon Fluophot, Tokyo, Japan). The distance between parallel double tetracycline labels present were measured using an eyepiece micrometer. Values were then divided by 9, the number of days between labels, to yield the mineral apposition rate in microns per day as previously described (12,25). Femoral calluses harvested from five rats in each defect group after 12 weeks were processed for routine histopathologic examination, in which the amount and type of new bone formation was evaluated. Bone Formation After the rats were killed, femora of defect and nonsurgical rats were harvested with the soft tissue intact. High-quality fine-focus roentgenograms (Faxitrons) were obtained using a Keleket machine. Assessment of mineral content of bone as a function of optical density (OD) was accomplished on a TBX Transmission Densitometer (Cones Instruments, Solon, OH) (23). Determination was made of the O D of the femora mid-diaphysis reflected as the percentage of light transmitted through the radiographs, and absorbed by the tissues present. In the nonsurgical groups, right versus left density values were compared. In the defect groups, values for the ratio of right femoral callus OD, over left midfemoral bone OD were compared (23). Faxitron radiographs of the defect rats, with identities masked, were scored for the amount of callus bridging the defect. Points were awarded as follows: 1 for no callus, 2 for callus but no union, and 3 for union (17). Single photon absorptiometry was performed on right femora of defect rats with a Norland Bone Densitometer (Data Products, Fort Atkinson, WI) to obtain callus values as a ratio of bone mineral content to bone width in gm/cm/cm (6). Specimens with polyethylene plate, screws, muscle, and other soft tissue intact were positioned on their medial side, perpendicular to the photon beam of the machine in a Styrofoam box, with air used as the surrounding medium. The distance between the screws was just adequate to avoid the photon beam, allowing similar alignment of all specimens. The contribution to the overall specimen density of the polyethylene plate and of the Styrofoam box was
871
determined and subtracted from the single photon absorptiometry values generated. Calluses from six specimens in each of the defect groups were manually harvested from the defect area of the femora, with sample identities concealed by number coding. The samples were preweighed and then dry weights were determined after drying the samples in an oven at 110°C for 24 h. The samples were subsequently heated at 600°C for 24 h, cooled, and weighed to allow determination of callus ash weights (mineral content). Statistics Data were statistically analyzed using x2 test, a two-tailed, unpaired t test, and/or analysis of variance, p values were considered significant if 60.05. All data are reported as means 2 SD (31).
RESULTS Nonsurgical Rats Serum analysis from nonsurgical rats showed no differences in T, levels between groups (saline 3.04 ? 1.03 pg/dl, 50-PRO 2.78 2 1.12 kg/dl, and 100PRO 3.65 2 1.10 pg/dl). There was a slight increase in calcium levels in the propranolol-treated rats (saline 8.47 2 0.73 mg/dl, 50-PRO 8.64 2 0.76 mg/dl, and 100-PRO 9.23 t 0.75 mg/dl). Biomechanical Testing Femora of nonsurgical rats biomechanically tested were of approximately the same dimensions with regard to length, anterior-posterior, and lateral middiaphyseal diameters (data not shown). Seven specimens were lost at the time of testing due to slippage of the specimens in the potting compound. There was a 24% increase in torsional strength found in 50-PRO rats and a 33% increase in strength found in 100-PRO rats when compared with the saline rats. No significant differences were found with regard to stiffness (Table 1). Defect Rats Defect rats evaluated in this study consisted of 20 saline, 17 50-PRO, and 21 100-PRO-treated animals. All of these rats had adequate polyethylene plate fixation to their right femora and received a full
J Orthop Res, Vol. 9, No. 6 , 1991
B. M I N K 0 WITZ ET AL.
872
TABLE 1. Biomechanical evaluation of nonsurgical rat femora" Group
50 pg PRO
Saline Torsional strength' (N-M) t SD Stiffness (NMP) 2 SD
0.37
2
0.14 (N
0.052
-t
0.011
=
10)
0.45
2
0.07 (N
100 pg PRO =
11)
0.49 t 0.07 (N
0.052 2 0.001
0.060
2
=
11)
0.012
NM, newton meter; N, number of rats per group; PRO, propranolol. Averaged values for right and left femora from biomechanical analysis using a rapid load torsion tester. Analysis of variance p = 0.09 for nonsurgical groups; t test p < 0.05 for 100 pg PRO versus saline.
'
course of treatment. No animals were lost to infection in this study. Behavioral changes in the rats were not observed in this study either as a result of surgery or of treatment. Treatment in the form of i.p. administration of propranolol, although unable to assure identical absorption to oral or intravenous routes of administration, was able to assure similar dosing of the animals. All of the rats had similar weights (450 k 20 g) at time of killing. Histochemistry
Examination of femoral metaphyseal trabecular tetracycline double labeling of defect rats revealed statistically significant increases in mineral apposition rates in both the 50-PRO and the 100-PRO groups of greater or equal to twice that of the saline group (p < 0.05). Differences between right and left femora within each group and between 50-PRO and 100-PRO groups were not significant (Table 2). Bone Formation
Optical densitometry and single photon absorptiometry showed significantly increased callus forma-
tion in propranolol groups compared with the saline defect group. Callus formation of the 50-PRO group was found to be increased 33% compared with the saline group by both methods. Callus formation in the 100-PRO group was increased 49% using optical densitometry and 28% using single photon absorptiometry. Optical density values were determined by two different observers and results were reproduced within 2%. Significant correlation of optical densitometry to bone ash weight (Y = 0.88) can be found in the literature (23). Using optical densitometry, no differences were found between nonsurgical femora as a function of right versus left or of group versus group (not shown). The precision for the single photon studies was found to be 3% when repeat values were observed by the' same person for femora at different times. This method has been found to significantly (Y = 0.99) correlate with mineral content (ash weight) with an accuracy of 2 4 % (6). The single photon absorptiometry values for 100-PRO and the 50-PRO groups are not significantly different (Table 3 ) . Right defect femora showed varying degrees of
TABLE 3. Optical densities (OD) and single photon absorptiometry of defect rats" Group
TABLE 2. Trabecular femoral metaphyseal mineral apposition ratesa Group
No. of rats/ group Mean ( p d d a y ) i SD
Saline
50 pg PRO
100 pg PRO
6
6
6
0.43 2 0.09
0.87
-t
0.17'
0.88 2 0.14'
Defect rats treated with propranolol (PRO) or saline were labeled with tetracycline hydrochloride on postoperative day 1 and doxycycline on postoperative day 10 and were sacrificed postoperative day 14 for analysis. p < 0.05 (analysis of variance) compared with control.
'
J Orthop Res, Vol. 9, No. 6 , 1991
Saline
OD callus-defect femora N = 13 Mean i SD 0.55 i 0.15 Callus-defect femora Single photon absorption N = 8 Mean i SD 0.18 2 0.03
50 pg PRO
100 pg PRO
N = 11 0.73 i 0.20'
N = 15 0.82 2 0.26b
N = 6 0.24 2 0.05'
N = 9 0.23 t 0.05'
~~
N, number of animals per group. Callus formation analysis of propranolol (PRO) and salinetreated defect rats as a percentage of light absorbed by the specimen (OD) and by bone mineral content divided by bone weight (g/cm/cm) using a single photon absorptiometer. p < 0.05 (analysis of variance) compared with controls.
'
PROPRANOLOL EFFECTS ON RAT BONE METABOLISM bridging at the defect gap (Fig. 1). Based on the scoring system 0% of the saline defect rats, 25% of the 50-PRO rats and 29% of the 100-PRO rats had unions by the 12th week (x2 p < 0.05) (Table 4). Calluses manually harvested from saline and propranolol defect rats at 12 weeks had similar percentages of mineral content (63% 4%).
*
Histopathology Histopathology of the callus from defects taken from rats that were killed showed new bone formation in the periosteal, cortical, and endosteal zones in all the samples. A unique observation seen in propranolol rats not found in saline-treated animals was that 50% of the samples examined from the 100-PRO and 50-PRO defect groups revealed marked increased chondroid periosteal new bone formation (Fig. 2A and B). DISCUSSION
Propranolol appears to have significant effects on normal bone and on endochondral bone formation
873
TABLE 4. Visual scoring from fine-focus radiographs of calluses ( I , no callus; 2 , callus but nonunion; 3, union)"
Score
Saline (%) (N = 14)
50 pg PRO (%)b (N = 12)
100 Pg PRO (%)b (N = 14)
1 2 3
36 64 0
42 33 25
35 36 29
N, number of rats per group. Scoring of callus formation in femoral gaps of defect rats treated with saline or propranolol (PRO) is reported as the percent of rats per group with each score. p < 0.05 (xz test) compared with controls. a
in the defect model. Biomechanical studies on nonsurgical animals showed increased strength in propranolol compared with saline-treated rats. In this experiment, minimally increased serum calcium levels in the 100 pg PRO group rats at the time of the kill revealed that 9 weeks after ending therapy changes can still be seen in serum calcium levels. No effects were found on T, levels, implying that there is no effect on T, levels based on the
FIG. 1. Representative fine-focus radiographs from the 100 pg propranolol-treated defect animals at 12 weeks show different extents of right femoral defect callus formation. Illustrated from left to right: no callus, nonunion, union.
J Orthop Res, Vol. 9, No. 6, 1991
874
B . MINKOWITZ ET AL. B
A
FIG. 2. Representative light microscopic photographs from callus of rat defects depicting new bone formation in periosteal, cortical, and endosteal zones with increased chondroid periosteal new bone formation in propranolol groups with saline groups (A).
equilibrium found between T, and T,. This agrees with findings by others using a similar propranolol treatment regimen in rats but differs from the clinical literature (13,16). Perhaps the human thyroid gland is more responsive to propranolol than is the rat’s. In propranolol-treated defect rats, increased femoral metaphyseal mineral apposition rates were noted early in the course of treatment. In defect rats, increased callus formation was observed at the time of the kill using optical densitometry as well as with single photon absorptiometry studies of the calluses. Visual findings on fine-focus radiographs of these calluses showed increased callus formation and union rates in femora of propranolol-treated rats. It would appear that in these defect rats a hypermetabolic state was activated by propranolol treatment, leading to increased bone formation and fracture healing. Densitometric studies of changes in healing callus were noted in the defect model used. This model accentuated differences between the groups studied by prolonging fracture healing time in the control animals from weeks to months. Density measurements in this study obtained by single photon absorptiometry appear to be more sensitive and reliable than those obtained by OD, which is dependent on many parameters not easily controlled and a lack of an optimal method to standardize OD measurements. Internal standards in single photon absorptiometer equipment increase accuracy and reproducibility of density value determinations. Limitations in sensitivity of single photon absorptiometry is seen with regard to the lack of differences between nonsurgical group femora, which were shown to exhibit increased strength in torsional J Orthop Res, Vol. 9, N o , 6 , 1991
(B)as compared
testing. This may mean that densitometry is not a good reflection of the biomechanical status of bones or that torsion testing bones is a more sensitive method of exhibiting changes in bone properties. Effects of propranolol in this study seem to be directed toward endochondral bone formation. This is well demonstrated in the metaphyseal region of bones studied for mineral apposition rates (12,18). Other investigators have reported that fracture healing affects bone mineral apposition rates in the rat femur contralateral to the fracture, possibly associated with systemic effects of fractures, which was not seen in this study (18). Effects on endochondral bone formation are also seen when callus production in defect rats was evaluated. Callus formation was hastened, and callus ash weights reveal that the same type of callus mineralization in propranolol- and saline-treated rats was present. Histologically, our observations that chondroid tissue was present in the propranolol group is consistent with recent literature demonstrating known osteoinductive endochondral cellular potential of the periosteum (22). In conclusion, this study demonstrates the existence of P-adrenergic control of bone formation. Propranolol may systemically stimulate bone metabolism or directly affect osteoblasts through p a d renergic or membrane-stabilizing mechanisms. A hypermetabolic state is induced in bone, resulting in increased rates of endochondral bone formation and increased bone strength (10,19,32). This may result from a disinhibition of an osseous metabolic pathway. Acknowledgment: We thank the Departments of Biomechanics and Nuclear Medicine at the Hospital for Special
PROPRANOLOL EFFECTS ON RAT BONE METABOLISM Surgery, New York, NY; Clinical Laboratories at Maimonides Medical Center, Brooklyn, NY; and Department of Orthopedics, Mt. Sinai Medical Center, New York, NY. Special thanks to C. Rimnac, Ph.D., Hospital for Special Surgery; J. Evangelista, M.D., Lutheran Medical Center; and D. M. Kahn, S. Varano, and M. Ross, M.D., Maimonides Medical Center. This project received the AOA-Zimmer Annual Travel Award and the Brooklyn Orthopedic Society Mercer Rang-Freundlich Award. Support for this study was supplied by grants from Maimonides Medical Center Research and Development Foundation and the Orthopedic Research Education Foundation.
REFERENCES 1 . Berg RA, Moss J, Baum BJ, Crystal RG: Regulation of collagen production by the beta-adrenergic system. J Clin Invest 67:1457-1462, 1981 2. Boucek J, Gunja-Smith 2 , Noble NL, Simpson CF: Modulation by propranolol of the lysyl cross-links in aortic elastin and collagen of the aneurysm-prone turkey. Biochem Pharmacol32:275-280, 1983 3. Brophy CM, Tilson JE, Tilson MD: Propranolol delays the formation of aneurysms in the male Blotchy mouse. J Surg Res 44:687689, 1988 4 . Brophy CM, Tilson JE, Tilson MD: Propranolol stimulates the crosslinking of matrix components in skin from the aneurysm-prone Blotchy mouse. J Surg Res 46:33&332, 1989 5. Burstein AH, Frankel VH: Technical note-a standard test for laboratory animal bone. J Biomech 4:155-158, 1971 6. Cameron JR, Mazess RB, Sorenson JA: Precision and accuracy of bone mineral determination by direct photon absorptiometry. Invest Radio1 3: 11-20, 1989 7. Chambers TJ, McSheehy PMJ, Thomson BM, Fuller K: The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit bones. Endocrinology 116:234-239, 1985 8 . Chihara K, Kodama H, Kaji H, Tetsuya K, Kashio Y, Okimura Y, Abe H, Fujita T: Augmentation by propranolol of growth hormone releasing hormone-( 1-44)-NH2 induced growth hormone release in normal short and normal children. J Clin Endocrinol Metab 61:229-233, 1985 9. Chopra IJ: A radioimmunoassay for measurement of thyroxine in unextracted human serum. J Clin Endocrinol Metab 37: 177-182, 1973 10. Dietrich JW, Mundy GR, Raisz LG: Inhibition of bone resorption in tissue culture by membrane-stabilizing drugs. Endocrinology 104:1644-1648, 1979 1 1 . Einhorn TA, Lane JM, Burstein AH, Kopman CR, Vigorita VJ: The healing of segmental bone defects induced by demineralized bone matrix. J Bone Joint Surg [Am] 66:274279, 1984 12. Einhorn TA, Vigorita VJ, Simon G, Devlin VJ, Warman J, Evangelista J: The mineral response of intact bones to remote skeletal injury. Trans Orthop Res Soc 13:540, 1988 13. Franklyn JA, Wilkens MR, Wilkinson R, Ramsden DB, Sheppard TC: The effect of propranolol on circulating thyroid hormone measurements in thyrotoxic and euthyroid subjects. Acta Endocrinol (Copenh) 108:351-355, 1985
875
14. Henry RJ, Cannon DC, Winkelman JW: Clinical Chemistry Principles and Techniques, 2nd ed, New York, Harper and Row, 1974 15. Imai Y, Rodan SB, Rodan GA: Effects of retinoic acid on alkaline phosphatase messenger ribonucleic acid, catecholamine receptors and G protein in ROS 17/23 cells. Endocrinology 122:456-463, 1988 16. Klein I: Thyroxine induced cardiac hypertrophy: time course of development and inhibition by propranolol. Endocrinology 123:203-210, 1988 17. Lane JM, Betts F, Posner AS, Yue DW: Mineral parameters in early fracture repair. J Bone Joint Surg [Am] 66:12891293, 1984 18. Lane JM, Sandhu HS: Current approaches to experimental bone grafting. Orthop Clin North A m 18:213-225, 1985 19. Levitzki A: From epinephrine to CAMP. Science 241:8W 805, 1988 20. Lund B, Hindberg I, Sorenson OH: Hypocalcaemic effect of propranolol in rats. Horm Metab Res 7:176-180, 1974 21. Mauras N, Blizzard RM, Thorner MO, Rogol AD: Selective beta,-adrenergic receptor blockade with atenolol enhances growth hormone and mediated growth hormone release in man. Metabolism 36:369-372, 1987 22. Nakahara H, Bruder SB, Goldberg VM, Caplan AI: In vivo osteochondrogenic potential of cultured cells derived from the periosteum. Clin Orthop 259:223-232, 1990 23. Nottebaert M, Lane JM, Juhn A , Burstein AH, Schneider R, Klein C, Sinn RS, Dowling C, Cornell C, Catsimpoolas N: Omental angiogenic lipid fraction and bone repair. An experimental study in the rat. J Orthop Res 7:157-169, 1989 24. Opsahl W, Zeronian H, Ellison M, Lewis D, Rucker RB, Riggins RS: Role of copper in collagen cross-linking and its influence on selected mechanical properties of chick bone and tendon. J Nutr 112:708-716, 1982 25. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR: Bone histomorphometry: standardization of nomenclature, symbols and units. J Bone Min Res 2:595610, 1987 26. Reddi AH: Cell biology and biochemistry of endochondral bone development. Coll Relat Res 1:209-226, 1981 27. Rodan SB, Rodan GA: Dexamethasone effects on betaadrenergic receptors and adenylate cyclase regulatory proteins Gs and Gi in ROS 17/23 cells. Endocrinology 118: 2510-2518, 1986 28. Rosenquist JB, Baylink DJ, Spengler DM: The effect of BAPN on bone mineralization. Proc Soc Exp Biol Med 154:31&3 13, 1977 29. Serlin MM, Orme ML’E, Baber NA, Sibeon RG, Laws E, Breckenridge A: Propranolol in the control of blood pressure: a dose response study. Clin Pharmacol Ther 2 7 5 8 6 588, 1980 30. Spengler DM, Baylink DJ, Rosenquist JB: Effect of betaaminopropionitrile on bone mechanical properties. J Bone Joint Surg [Am] 59:670-672, 1977 31. Stantona-Glanz: Primer of Biostatistics, New York, McGraw Hill, 1974 32. Stiles GL, Caron MG, Lefkowitz RL: Beta-adrenergic receptors: biochemical mechanisms of physiological regulation. Physiol Rev 64:661-743, 1984 33. Wilkins MR, Franklyn JA, Woods KL, Kendall MJ: Effect of propranolol on thyroid homeostasis of healthy volunteers. Postgrad Med J 61:391-394, 1985
J Orthop Res, Vol. 9,No. 6 , 1991
Effects of Propranolol on Bone Metabolism in the Rat Barbara Minkowitz, *Adele L. Boskey, *Joseph M. Lane, Hubert S. Pearlman, and ?Vincent J. Vigorita Department of Orthopaedics, Maimonides Medical Center, Brooklyn, *Departments of Ultrastructural Biochemistry and Collagen Tissue Research, Hospital for Special Surgery, New York; and fDepartment of Pathology, Lutheran Medical Center, Brooklyn, New York, U.S.A.
Summary: Propranolol, a nonspecific P-blocker, has many physiologic effects. Its effects on bone in vivo are unknown, although P receptor sites have been found on osteoblasts. In this study, the hypothesis tested was that low doses of propranolol could alter bone properties and enhance orthotopic endochondral bone formation. In a group of nonsurgical rats, propranolol treatment increased femoral torsional strength on biomechanical testing. In the rat surgical model used, right femora were fixed to a polyethylene plate and then defects were created mid-diaphysis and subsequently filled with demineralized bone matrix. These rats (defect rats) were randomly divided into groups that were given propranolol or a saline carrier for 19 consecutive days. In the defect rats, increased trabecular femoral metaphyseal mineral apposition rates were observed in propranolol-treated groups. Densitometry and roentgenographic scoring of callus formation after 12 weeks in propranolol-treated rats revealed increased callus and bone union. The results of this study indicate that propranolol treatment can significantly affect bone properties. Key Words: Propranolol-6-Adrenergic drug-Bone metabolism-Fracture healing-RatBone density.
The clinical effects of propranolol, a nonspecific P-blocker, include decreased hypertension, increased serum growth hormone levels, decreased thyroid hormone levels, decreased serum calcium levels, and increased vascular collagen crosslinking (8,13,20,21,29,33). In bone organ culture systems, propranolol has been shown to decrease resorptive effects of parathormone as well as other calciotropic hormones/agents (7,lO). In vivo propranolol affects the type I collagen in vascular tissues via interactions with lysyl-oxidase. This results in increased collagen cross-linking and increased tensile strength (2-4,24,28,30). In vitro, increased collagen synthesis in fetal lung fibroblast
culture was observed in the presence of propranolol (1). In vivo effects of P-blockers on bone are not well understood, although p receptor sites have been identified on osteoblasts in vitro (15,27). The significance of these receptor sites has never been studied in vivo or in vitro. In the present study, the hypothesis tested was that low doses of propranolol could alter bone properties and enhance orthotopic endochondral bone formation.
METHODS Experimental Groups One hundred thirty-nine male Sprague-Dawley rats (350 g) were purchased from Taconic Farms (Germantown, NY). They were divided into six treatment groups given propranolol or saline, with 40 rats serving as donors for the demineralized bone
Received August 30, 1989; accepted April 29, 1991. Address correspondence and reprint requests to Dr. B . Minkowitz care of A. L. Boskey at Hospital for Special Surgery, 535 East 70th Street, New York, NY 10021, U.S.A.
869
B. MINKOWITZ ET AL.
870
matrix needed for the rat femoral defect model. Three of the six groups, consisting of 12 animals each, were not manipulated (nonsurgical rats). The remaining three groups, consisting of 21 animals each, were surgically manipulated (defect rats). Treatments Intraperitoneal (i.p.) injections of saline or dilutions of injectable propranolol (Inderal, WyethAyerst, Philadelphia, PA) in saline were administered daily for 19 consecutive days. This treatment was started on postoperative day 1 for the defect animals. One group of nonsurgical and defect rats received saline, one nonsurgical and defect group received 50 pg propranolol(50-PRO), and the third nonsurgical and defect groups received 100 pg propranolol(lO0-PRO). Rats were fed Purina rat chow and were given water ad libitum. All experimental animals received humane care and treatment in compliance with the “Guide for the Care and Use of Lab Animals,” National Institutes of Health publication no. 86-23, revised 1985. Nonsurgical Rats Nonsurgical rats were all killed 12 weeks after the start of the treatments. Serum samples from nonsurgical rats were individually analyzed for calcium on a Beckman Astra autoanalyzer (Beckman Instruments, Fullerton, CA, U.S.A.) (14). The amount of T, present was analyzed using a T, radioimmunoassay (RIA) (9).
Defect Rats Demineralized Bone Matrix Preparation
Allogeneic demineralized bone matrix was prepared from combined femoral and tibia1 diaphyses cleaned of soft tissue and bone marrow, and subsequently lyophilized and then ground to a particle size of 45-300 pm in a Spex Mill (Spex Industries, Metuchen, NJ). This powder was extracted sequentially with 0.5 N hydrochloric acid, absolute ethanol, and ethyl ether to yield demineralized bone matrix (26). Defect Model
Three groups of 21 rats each underwent femoral defect formation (23). They were given chloral hydrate and procaine penicillin (60,000 U subcutaneous) before surgical manipulation. An incision was made laterally along the length of the right femur, which was then cleaned of soft tissue. A predrilled, high-density polyethylene plate (23 X 4 x 4 mm) was secured to the anterior surface of the femur by four threaded kirschner wires (1.2-mm diameter no. 26212 Zimmer). A 5-mm defect was created midfemur (twice the cross-sectional diameter) using a dental burr. Bone marrow on either side of the defect was flushed from the bone using a syringe loaded with isotonic saline. The defect was packed with the demineralized bone matrix powder (1 1). Rats were radiographed at 3-week intervals, to verify proper placement of the polyethylene plates. Histochemistry and Histopathology
Biomechanical Testing Both femora of nonsurgical groups were stabilized proximally and distally in a potting compound (cerubend alloy, Cenumetal Products, Bellefonte, PA, U.S.A.) for mechanical testing (5). Before testing, anterior-posterior and medial-lateral diameters of all bones were measured using a caliper. All bones were fractured in external rotation in a rapid load torsion tester (26). The stresshtrain curves generated were analyzed on an LCS Microplan I1 (Anaheim, CA) digitizing table connected to a minicomputer. Osseous biomechanical properties determined included the torsional strength and the stiffness (26). Right and left leg values were averaged to obtain a single value for each nonsurgical rat.
J Orthop Res, Vol. 9, No. 6 , 1991
Six rats in each defect group were labeled i.p. with tetracycline hydrochloride (86 pg/kg) on postoperative day 1 and with doxycycline (43 pg/kg) on postoperative day 10. These rats were killed on postoperative day 14 for evaluation of femoral metaphyseal trabecular mineral apposition rates in both operated and nonoperated femora. For histomorphometric studies, we chose to analyze mineral apposition rates, which are directly derived and do not require data manipulation, to evaluate changes in bone metabolism (25). Trabecular bone rather than cortical bone labels were analyzed because of artifactual distortion of cortical bone during processing. Undecalcified femoral metaphyses were coronally sectioned at a thickness of 5 pm using a heavy-duty microtome (Polycut S, Reichert-Jung,
PROPRANOLOL EFFECTS ON RAT BONE METABOLISM Heidelberg, Germany). Unstained sections were analyzed (identifications were masked by numbercoding the slides) using fluorescent microscopy (Nikon Fluophot, Tokyo, Japan). The distance between parallel double tetracycline labels present were measured using an eyepiece micrometer. Values were then divided by 9, the number of days between labels, to yield the mineral apposition rate in microns per day as previously described (12,25). Femoral calluses harvested from five rats in each defect group after 12 weeks were processed for routine histopathologic examination, in which the amount and type of new bone formation was evaluated. Bone Formation After the rats were killed, femora of defect and nonsurgical rats were harvested with the soft tissue intact. High-quality fine-focus roentgenograms (Faxitrons) were obtained using a Keleket machine. Assessment of mineral content of bone as a function of optical density (OD) was accomplished on a TBX Transmission Densitometer (Cones Instruments, Solon, OH) (23). Determination was made of the O D of the femora mid-diaphysis reflected as the percentage of light transmitted through the radiographs, and absorbed by the tissues present. In the nonsurgical groups, right versus left density values were compared. In the defect groups, values for the ratio of right femoral callus OD, over left midfemoral bone OD were compared (23). Faxitron radiographs of the defect rats, with identities masked, were scored for the amount of callus bridging the defect. Points were awarded as follows: 1 for no callus, 2 for callus but no union, and 3 for union (17). Single photon absorptiometry was performed on right femora of defect rats with a Norland Bone Densitometer (Data Products, Fort Atkinson, WI) to obtain callus values as a ratio of bone mineral content to bone width in gm/cm/cm (6). Specimens with polyethylene plate, screws, muscle, and other soft tissue intact were positioned on their medial side, perpendicular to the photon beam of the machine in a Styrofoam box, with air used as the surrounding medium. The distance between the screws was just adequate to avoid the photon beam, allowing similar alignment of all specimens. The contribution to the overall specimen density of the polyethylene plate and of the Styrofoam box was
871
determined and subtracted from the single photon absorptiometry values generated. Calluses from six specimens in each of the defect groups were manually harvested from the defect area of the femora, with sample identities concealed by number coding. The samples were preweighed and then dry weights were determined after drying the samples in an oven at 110°C for 24 h. The samples were subsequently heated at 600°C for 24 h, cooled, and weighed to allow determination of callus ash weights (mineral content). Statistics Data were statistically analyzed using x2 test, a two-tailed, unpaired t test, and/or analysis of variance, p values were considered significant if 60.05. All data are reported as means 2 SD (31).
RESULTS Nonsurgical Rats Serum analysis from nonsurgical rats showed no differences in T, levels between groups (saline 3.04 ? 1.03 pg/dl, 50-PRO 2.78 2 1.12 kg/dl, and 100PRO 3.65 2 1.10 pg/dl). There was a slight increase in calcium levels in the propranolol-treated rats (saline 8.47 2 0.73 mg/dl, 50-PRO 8.64 2 0.76 mg/dl, and 100-PRO 9.23 t 0.75 mg/dl). Biomechanical Testing Femora of nonsurgical rats biomechanically tested were of approximately the same dimensions with regard to length, anterior-posterior, and lateral middiaphyseal diameters (data not shown). Seven specimens were lost at the time of testing due to slippage of the specimens in the potting compound. There was a 24% increase in torsional strength found in 50-PRO rats and a 33% increase in strength found in 100-PRO rats when compared with the saline rats. No significant differences were found with regard to stiffness (Table 1). Defect Rats Defect rats evaluated in this study consisted of 20 saline, 17 50-PRO, and 21 100-PRO-treated animals. All of these rats had adequate polyethylene plate fixation to their right femora and received a full
J Orthop Res, Vol. 9, No. 6 , 1991
B. M I N K 0 WITZ ET AL.
872
TABLE 1. Biomechanical evaluation of nonsurgical rat femora" Group
50 pg PRO
Saline Torsional strength' (N-M) t SD Stiffness (NMP) 2 SD
0.37
2
0.14 (N
0.052
-t
0.011
=
10)
0.45
2
0.07 (N
100 pg PRO =
11)
0.49 t 0.07 (N
0.052 2 0.001
0.060
2
=
11)
0.012
NM, newton meter; N, number of rats per group; PRO, propranolol. Averaged values for right and left femora from biomechanical analysis using a rapid load torsion tester. Analysis of variance p = 0.09 for nonsurgical groups; t test p < 0.05 for 100 pg PRO versus saline.
'
course of treatment. No animals were lost to infection in this study. Behavioral changes in the rats were not observed in this study either as a result of surgery or of treatment. Treatment in the form of i.p. administration of propranolol, although unable to assure identical absorption to oral or intravenous routes of administration, was able to assure similar dosing of the animals. All of the rats had similar weights (450 k 20 g) at time of killing. Histochemistry
Examination of femoral metaphyseal trabecular tetracycline double labeling of defect rats revealed statistically significant increases in mineral apposition rates in both the 50-PRO and the 100-PRO groups of greater or equal to twice that of the saline group (p < 0.05). Differences between right and left femora within each group and between 50-PRO and 100-PRO groups were not significant (Table 2). Bone Formation
Optical densitometry and single photon absorptiometry showed significantly increased callus forma-
tion in propranolol groups compared with the saline defect group. Callus formation of the 50-PRO group was found to be increased 33% compared with the saline group by both methods. Callus formation in the 100-PRO group was increased 49% using optical densitometry and 28% using single photon absorptiometry. Optical density values were determined by two different observers and results were reproduced within 2%. Significant correlation of optical densitometry to bone ash weight (Y = 0.88) can be found in the literature (23). Using optical densitometry, no differences were found between nonsurgical femora as a function of right versus left or of group versus group (not shown). The precision for the single photon studies was found to be 3% when repeat values were observed by the' same person for femora at different times. This method has been found to significantly (Y = 0.99) correlate with mineral content (ash weight) with an accuracy of 2 4 % (6). The single photon absorptiometry values for 100-PRO and the 50-PRO groups are not significantly different (Table 3 ) . Right defect femora showed varying degrees of
TABLE 3. Optical densities (OD) and single photon absorptiometry of defect rats" Group
TABLE 2. Trabecular femoral metaphyseal mineral apposition ratesa Group
No. of rats/ group Mean ( p d d a y ) i SD
Saline
50 pg PRO
100 pg PRO
6
6
6
0.43 2 0.09
0.87
-t
0.17'
0.88 2 0.14'
Defect rats treated with propranolol (PRO) or saline were labeled with tetracycline hydrochloride on postoperative day 1 and doxycycline on postoperative day 10 and were sacrificed postoperative day 14 for analysis. p < 0.05 (analysis of variance) compared with control.
'
J Orthop Res, Vol. 9, No. 6 , 1991
Saline
OD callus-defect femora N = 13 Mean i SD 0.55 i 0.15 Callus-defect femora Single photon absorption N = 8 Mean i SD 0.18 2 0.03
50 pg PRO
100 pg PRO
N = 11 0.73 i 0.20'
N = 15 0.82 2 0.26b
N = 6 0.24 2 0.05'
N = 9 0.23 t 0.05'
~~
N, number of animals per group. Callus formation analysis of propranolol (PRO) and salinetreated defect rats as a percentage of light absorbed by the specimen (OD) and by bone mineral content divided by bone weight (g/cm/cm) using a single photon absorptiometer. p < 0.05 (analysis of variance) compared with controls.
'
PROPRANOLOL EFFECTS ON RAT BONE METABOLISM bridging at the defect gap (Fig. 1). Based on the scoring system 0% of the saline defect rats, 25% of the 50-PRO rats and 29% of the 100-PRO rats had unions by the 12th week (x2 p < 0.05) (Table 4). Calluses manually harvested from saline and propranolol defect rats at 12 weeks had similar percentages of mineral content (63% 4%).
*
Histopathology Histopathology of the callus from defects taken from rats that were killed showed new bone formation in the periosteal, cortical, and endosteal zones in all the samples. A unique observation seen in propranolol rats not found in saline-treated animals was that 50% of the samples examined from the 100-PRO and 50-PRO defect groups revealed marked increased chondroid periosteal new bone formation (Fig. 2A and B). DISCUSSION
Propranolol appears to have significant effects on normal bone and on endochondral bone formation
873
TABLE 4. Visual scoring from fine-focus radiographs of calluses ( I , no callus; 2 , callus but nonunion; 3, union)"
Score
Saline (%) (N = 14)
50 pg PRO (%)b (N = 12)
100 Pg PRO (%)b (N = 14)
1 2 3
36 64 0
42 33 25
35 36 29
N, number of rats per group. Scoring of callus formation in femoral gaps of defect rats treated with saline or propranolol (PRO) is reported as the percent of rats per group with each score. p < 0.05 (xz test) compared with controls. a
in the defect model. Biomechanical studies on nonsurgical animals showed increased strength in propranolol compared with saline-treated rats. In this experiment, minimally increased serum calcium levels in the 100 pg PRO group rats at the time of the kill revealed that 9 weeks after ending therapy changes can still be seen in serum calcium levels. No effects were found on T, levels, implying that there is no effect on T, levels based on the
FIG. 1. Representative fine-focus radiographs from the 100 pg propranolol-treated defect animals at 12 weeks show different extents of right femoral defect callus formation. Illustrated from left to right: no callus, nonunion, union.
J Orthop Res, Vol. 9, No. 6, 1991
874
B . MINKOWITZ ET AL. B
A
FIG. 2. Representative light microscopic photographs from callus of rat defects depicting new bone formation in periosteal, cortical, and endosteal zones with increased chondroid periosteal new bone formation in propranolol groups with saline groups (A).
equilibrium found between T, and T,. This agrees with findings by others using a similar propranolol treatment regimen in rats but differs from the clinical literature (13,16). Perhaps the human thyroid gland is more responsive to propranolol than is the rat’s. In propranolol-treated defect rats, increased femoral metaphyseal mineral apposition rates were noted early in the course of treatment. In defect rats, increased callus formation was observed at the time of the kill using optical densitometry as well as with single photon absorptiometry studies of the calluses. Visual findings on fine-focus radiographs of these calluses showed increased callus formation and union rates in femora of propranolol-treated rats. It would appear that in these defect rats a hypermetabolic state was activated by propranolol treatment, leading to increased bone formation and fracture healing. Densitometric studies of changes in healing callus were noted in the defect model used. This model accentuated differences between the groups studied by prolonging fracture healing time in the control animals from weeks to months. Density measurements in this study obtained by single photon absorptiometry appear to be more sensitive and reliable than those obtained by OD, which is dependent on many parameters not easily controlled and a lack of an optimal method to standardize OD measurements. Internal standards in single photon absorptiometer equipment increase accuracy and reproducibility of density value determinations. Limitations in sensitivity of single photon absorptiometry is seen with regard to the lack of differences between nonsurgical group femora, which were shown to exhibit increased strength in torsional J Orthop Res, Vol. 9, N o , 6 , 1991
(B)as compared
testing. This may mean that densitometry is not a good reflection of the biomechanical status of bones or that torsion testing bones is a more sensitive method of exhibiting changes in bone properties. Effects of propranolol in this study seem to be directed toward endochondral bone formation. This is well demonstrated in the metaphyseal region of bones studied for mineral apposition rates (12,18). Other investigators have reported that fracture healing affects bone mineral apposition rates in the rat femur contralateral to the fracture, possibly associated with systemic effects of fractures, which was not seen in this study (18). Effects on endochondral bone formation are also seen when callus production in defect rats was evaluated. Callus formation was hastened, and callus ash weights reveal that the same type of callus mineralization in propranolol- and saline-treated rats was present. Histologically, our observations that chondroid tissue was present in the propranolol group is consistent with recent literature demonstrating known osteoinductive endochondral cellular potential of the periosteum (22). In conclusion, this study demonstrates the existence of P-adrenergic control of bone formation. Propranolol may systemically stimulate bone metabolism or directly affect osteoblasts through p a d renergic or membrane-stabilizing mechanisms. A hypermetabolic state is induced in bone, resulting in increased rates of endochondral bone formation and increased bone strength (10,19,32). This may result from a disinhibition of an osseous metabolic pathway. Acknowledgment: We thank the Departments of Biomechanics and Nuclear Medicine at the Hospital for Special
PROPRANOLOL EFFECTS ON RAT BONE METABOLISM Surgery, New York, NY; Clinical Laboratories at Maimonides Medical Center, Brooklyn, NY; and Department of Orthopedics, Mt. Sinai Medical Center, New York, NY. Special thanks to C. Rimnac, Ph.D., Hospital for Special Surgery; J. Evangelista, M.D., Lutheran Medical Center; and D. M. Kahn, S. Varano, and M. Ross, M.D., Maimonides Medical Center. This project received the AOA-Zimmer Annual Travel Award and the Brooklyn Orthopedic Society Mercer Rang-Freundlich Award. Support for this study was supplied by grants from Maimonides Medical Center Research and Development Foundation and the Orthopedic Research Education Foundation.
REFERENCES 1 . Berg RA, Moss J, Baum BJ, Crystal RG: Regulation of collagen production by the beta-adrenergic system. J Clin Invest 67:1457-1462, 1981 2. Boucek J, Gunja-Smith 2 , Noble NL, Simpson CF: Modulation by propranolol of the lysyl cross-links in aortic elastin and collagen of the aneurysm-prone turkey. Biochem Pharmacol32:275-280, 1983 3. Brophy CM, Tilson JE, Tilson MD: Propranolol delays the formation of aneurysms in the male Blotchy mouse. J Surg Res 44:687689, 1988 4 . Brophy CM, Tilson JE, Tilson MD: Propranolol stimulates the crosslinking of matrix components in skin from the aneurysm-prone Blotchy mouse. J Surg Res 46:33&332, 1989 5. Burstein AH, Frankel VH: Technical note-a standard test for laboratory animal bone. J Biomech 4:155-158, 1971 6. Cameron JR, Mazess RB, Sorenson JA: Precision and accuracy of bone mineral determination by direct photon absorptiometry. Invest Radio1 3: 11-20, 1989 7. Chambers TJ, McSheehy PMJ, Thomson BM, Fuller K: The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit bones. Endocrinology 116:234-239, 1985 8 . Chihara K, Kodama H, Kaji H, Tetsuya K, Kashio Y, Okimura Y, Abe H, Fujita T: Augmentation by propranolol of growth hormone releasing hormone-( 1-44)-NH2 induced growth hormone release in normal short and normal children. J Clin Endocrinol Metab 61:229-233, 1985 9. Chopra IJ: A radioimmunoassay for measurement of thyroxine in unextracted human serum. J Clin Endocrinol Metab 37: 177-182, 1973 10. Dietrich JW, Mundy GR, Raisz LG: Inhibition of bone resorption in tissue culture by membrane-stabilizing drugs. Endocrinology 104:1644-1648, 1979 1 1 . Einhorn TA, Lane JM, Burstein AH, Kopman CR, Vigorita VJ: The healing of segmental bone defects induced by demineralized bone matrix. J Bone Joint Surg [Am] 66:274279, 1984 12. Einhorn TA, Vigorita VJ, Simon G, Devlin VJ, Warman J, Evangelista J: The mineral response of intact bones to remote skeletal injury. Trans Orthop Res Soc 13:540, 1988 13. Franklyn JA, Wilkens MR, Wilkinson R, Ramsden DB, Sheppard TC: The effect of propranolol on circulating thyroid hormone measurements in thyrotoxic and euthyroid subjects. Acta Endocrinol (Copenh) 108:351-355, 1985
875
14. Henry RJ, Cannon DC, Winkelman JW: Clinical Chemistry Principles and Techniques, 2nd ed, New York, Harper and Row, 1974 15. Imai Y, Rodan SB, Rodan GA: Effects of retinoic acid on alkaline phosphatase messenger ribonucleic acid, catecholamine receptors and G protein in ROS 17/23 cells. Endocrinology 122:456-463, 1988 16. Klein I: Thyroxine induced cardiac hypertrophy: time course of development and inhibition by propranolol. Endocrinology 123:203-210, 1988 17. Lane JM, Betts F, Posner AS, Yue DW: Mineral parameters in early fracture repair. J Bone Joint Surg [Am] 66:12891293, 1984 18. Lane JM, Sandhu HS: Current approaches to experimental bone grafting. Orthop Clin North A m 18:213-225, 1985 19. Levitzki A: From epinephrine to CAMP. Science 241:8W 805, 1988 20. Lund B, Hindberg I, Sorenson OH: Hypocalcaemic effect of propranolol in rats. Horm Metab Res 7:176-180, 1974 21. Mauras N, Blizzard RM, Thorner MO, Rogol AD: Selective beta,-adrenergic receptor blockade with atenolol enhances growth hormone and mediated growth hormone release in man. Metabolism 36:369-372, 1987 22. Nakahara H, Bruder SB, Goldberg VM, Caplan AI: In vivo osteochondrogenic potential of cultured cells derived from the periosteum. Clin Orthop 259:223-232, 1990 23. Nottebaert M, Lane JM, Juhn A , Burstein AH, Schneider R, Klein C, Sinn RS, Dowling C, Cornell C, Catsimpoolas N: Omental angiogenic lipid fraction and bone repair. An experimental study in the rat. J Orthop Res 7:157-169, 1989 24. Opsahl W, Zeronian H, Ellison M, Lewis D, Rucker RB, Riggins RS: Role of copper in collagen cross-linking and its influence on selected mechanical properties of chick bone and tendon. J Nutr 112:708-716, 1982 25. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR: Bone histomorphometry: standardization of nomenclature, symbols and units. J Bone Min Res 2:595610, 1987 26. Reddi AH: Cell biology and biochemistry of endochondral bone development. Coll Relat Res 1:209-226, 1981 27. Rodan SB, Rodan GA: Dexamethasone effects on betaadrenergic receptors and adenylate cyclase regulatory proteins Gs and Gi in ROS 17/23 cells. Endocrinology 118: 2510-2518, 1986 28. Rosenquist JB, Baylink DJ, Spengler DM: The effect of BAPN on bone mineralization. Proc Soc Exp Biol Med 154:31&3 13, 1977 29. Serlin MM, Orme ML’E, Baber NA, Sibeon RG, Laws E, Breckenridge A: Propranolol in the control of blood pressure: a dose response study. Clin Pharmacol Ther 2 7 5 8 6 588, 1980 30. Spengler DM, Baylink DJ, Rosenquist JB: Effect of betaaminopropionitrile on bone mechanical properties. J Bone Joint Surg [Am] 59:670-672, 1977 31. Stantona-Glanz: Primer of Biostatistics, New York, McGraw Hill, 1974 32. Stiles GL, Caron MG, Lefkowitz RL: Beta-adrenergic receptors: biochemical mechanisms of physiological regulation. Physiol Rev 64:661-743, 1984 33. Wilkins MR, Franklyn JA, Woods KL, Kendall MJ: Effect of propranolol on thyroid homeostasis of healthy volunteers. Postgrad Med J 61:391-394, 1985
J Orthop Res, Vol. 9,No. 6 , 1991
