Estimated intracortical bone turnover in the femur of growing macaquesImplications for their use as models in skeletal pathology.код для вставкиСкачать
THE ANATOMICAL RECORD 232:180-189 (1992) Estimated Intracortical Bone Turnover in the Femur of Growing Macaques: Implications for Their Use as Models in Skeletal Pathology DAVID B. BURR Departments of Anatomy and Orthopaedic Surgery, Indiana University School of Medicine, Indianapolis, Indiana 46202 ABSTRACT Although macaques have been used widely to study the dynamics of bone remodeling, there have been few comparisons to skeletal tissue turnover in humans. This study analyzes variation in bone microstructure with respect to gender, age, and body weight in growing macaques to determine whether the pattern of osteonal remodeling in macaques is like that in analogous bones of humans. Histomorphometric measurements were made on femoral midshaft crosssections of 54 macaques from the Cay0 Santiago skeletal collection. These data show that variation in bone microstructure occurs independent of gender, age, and body weight in macaques. Moreover, intracortical bone turnover in the macaque femur is much slower than in humans, making them poor models to study the effects of weightlessness, but good models to study low turnover skeletal dysplasias. The use of animal models to study human disease permits study of the disease while controlling all relevant variables that could affect it. Nonhuman primates are particularly useful because of their taxonomic affinities to humans. Macaques in particular have been used widely for biomedical studies of skeletal pathology (Bowden, 1979; Cann et al., 1980; Young and Schneider, 1981; Williams and Bowden, 1984; DeRousseau, 1985; Turnquist, 1983, 1985; Jerome et al., 1986; Young et al., 1986; Simmons et al., 1984, 1986; Miller et al., 1986). Remarkably, in view of their widespread use in biomedical research, we know relatively little about the processes of bone remodeling at the tissue level in these animals. Only a few studies have examined remodeling kinetics in rhesus macaques (Schock et al., 1972,1975; Przybeck, 1985). These studies have not examined the gender or body weight variation in relation to skeletal dynamics. Furthermore, variation has been incompletely characterized using only 2-3 samples at discrete age intervals. Preliminary data that suggested that the rate of intracortical bone turnover in macaques is quite low (Schaffler and Burr, 1984; Burr et al., 1989) support the contention that macaques may provide relatively good models for some human skeletal disease, e.g., senile osteoporosis. The broad objective of this work was to discover whether the processes of skeletal remodeling during maturation in rhesus macaques are like those in growing humans, allowing Macaca rnulatta to be used as an animal model for diseases that involve abnormal remodeling physiology. Several specific questions were posed: 1. Does variation in bone microstructure occur independent of gender, age, andlor body weight? 2. Are growing macaques different than adult macaques or adult humans in the extent of intracortical bone remodeling? 0 1992 WILEY-LISS, INC. 3. Are there maturational effects on bone remodeling in growing macaques that create different age-adjusted values for variables that describe remodeling in males and females? MATERIALS AND METHODS It is desirable to know something about the mechanical environment to which bone is exposed. Peak principal strain magnitudes and distribution have been collected in vivo in nonhuman primates only for the macaque mandible (Hylander, 1981, 1984) and tibia (Young et al., 1977) and for gibbon upper limb bones (Swartz et al., 1989). The midshaft of the femur in M . mulatta provides a site that 1) can be accurately defined in different animals; 2) is mechanically relevant and “comparable”across species; and 3) can be mechanically analyzed using beam theory even prior to collection of relevant in vivo strain data. All these reasons make this site suitable for analysis. One millimeter thick cross-sections were removed from the femoral midshafts of Cay0 Santiago derived M . rnulatta (n = 31 females and 23 males) ranging in age from 1.5-8.0 years of age. All femoral epiphyses are fused by 5.25 years in female and 6.0 years in male M . rnulatta (Cheverud, 1981). As these are among the last long bone epiphyses to fuse, this can be considered the age of skeletal maturity in this species. The mean age of males (4.57 2 1.79) was not significantly different than that in females (4.28 1.91). Only animals with known body weights were used. Body weights were taken from the medical records kept a t the Caribbean Primate Research Center (CPRC) in Puerto Rico. All animals had been born free-ranging on Cay0 Santiago, an island off the southeast coast of Puerto Rico, * Received February 8, 1991; accepted June 19, 1991. BONE TURNOVER IN MACAQUES but transferred to enclosures a t Sabana Seca in the CPRC at varying periods prior to death. Medical records were evaluated for all animals. Most animals had died as the result of wounds received from other animals, or of parasitic infections. We used no animals that died of metabolic diseases. The cross-sections were glued with cyanoacrylate to a plexiglass slide and hand ground to 150 microns thick using 600 grit silicon carbide paper, acid etched with HC1, and stained with toluidine blue. This makes identification of the cement line easier and facilitates identification of secondary osteons as distinct from primary osteons and vascular spaces. Secondary osteons represent remodeled bone and are defined by the presence of a cement line. Primary osteons represent the primary infilling of vascular spaces, are not replacement bone, and as such, have no cement line. Sections were viewed a t x212 magnification using a Zeiss Universal Research microscope. The sections were examined quantitatively using a n IBM-PC XT microcomputer and Bioquant image analysis software. Only secondary osteons were measured and the following calculations used for analysis: 181 tant determinant of variation in osteon development. To investigate the nature of this possible size-related source of variation, allometric equations (y = bxk) were calculated on log-transformed variables. The exponent of each relationship was estimated by both model I (least squares) and model I1 (major axis) regressions. Major axis (principal axis) fits to the logtransformed data acknowledge the fact that both x and y variables possess associated error terms, and estimates of the exponent k tend to be slightly higher than least squares values. Both regression models yield virtually identical results and inferences when the correlation between x and y variables is reasonably high. Because intraspecific allometric analyses with weight tend to have low correlation coefficients, reduced major axis (RMA) regression analysis may be preferable to either major axis or least squares methods (Rayner, 1981; Seim and Saether, 1983; Ricker, 1984). RMA fits to the data were also explored. Analysis of covariance was used to determine interactive effects of gender on this relationship. Values for all individuals for each of the variables were examined in multivariate space using the interactive statistical graphics package J M P (SAS Institute, 1. Percent osteonal area = (total summed osteon Inc., Cary, NC). Three individuals were outliers on several variables. These individuals “leverage” the rest of arealcortical area) x 100. 2. Osteons/mm2 (#/mm2): the number of osteons the data, artificially increasing the r2 value for a given with intact haversian canals per unit area of bone. regression. This can be misleading by suggesting a 3. Osteon fragmentdrnm’ (#/mm2): the number of strong relationship between variables when none expartial osteons without haversian canals per unit area ists. To avoid this, we adopted the conservative approach, and eliminated these three individuals (2 of bone. 4. Osteon population density/mm2 (OPD, #/mm2): males, 1 female) from the analysis. Two of these individuals were among the youngest (less than 2 years the sum of 2 and 3. 5. Mean osteon area (Ao, pm2): areas within the 0s- old) and we felt that the small amount of bone in the teonal cement line, measured for each osteon in a cross- cross-section could have biased the OPD counts. Analysis of covariance was used to assess differences section and averaged. 6. Mean wall thickness (MWT, pm) = ([osteon area in microstructure with respect to gender, age, and body weight. Linear regression analysis was used to deter- canal area]*2)/(osteon perimeter + canal perimeter); the average distance from cement line to haver- mine whether microstructure changed in linear fashsian canal wall. We use this method to avoid bias in- ion with age. Polynomial fits to the data were also introduced by relating to what location in a n osteon to vestigated. Finally logistic regression analysis was measure. According to Quarles and Lobaugh (1989) employed to further examine changes in bone microthickness can be measured either directly or indirectly structure with age. with the same error. Direct linear measurements rely RESULTS on the observer’s ability to randomly choose sites for Variation of Microstructure With Body Weight measurement in osteons that are not completely circuAs expected, body weight and age in growing lar. The method employed here eliminates bias by avmacaques are strongly correlated (Fig. 1; r2 = .79 to eraging all possible MWTs for a given osteon. 7. Activation frequency (mu, #/mm2/year) = OPDI predict age from weight, and .82 to predict weight from mean tissue age; this is the number of osteon creations age; P < .0001). The relationship between age and per unit time. Because the mean tissue age is not weight is different in male and female macaques, as known for these animals, and because the OPD is so demonstrated by significant interactions of each with small, we assumed t h a t mean tissue age = chronolog- gender (P < .01 and P < .03, respectively). Table 1 shows that the only microstructural variical age. 8. Bone formation rate (BFR, mm2/mm2/year) = ables that can predict body weight a t any acceptable confidence level are variables that are calculated using (mu * A,)/106 age (mu and BFR). Other than these, mean osteon area Wu et al. (1970) described the methods used to cal- is the best predictor, but explains only 34% of the variculate activation frequency and BFR from cross-sec- ation in weight. Except for percent osteonal bone, no tions without intravital labeling. Frost (1987 a,b) ex- interactive effects with sex were found when micropanded on the mathematical justification for these structural variables were regressed on body weight, measurements. The basis for the other measurements suggesting that gender is not a n important factor in is more fully described by Stout and Teitelbaum (19791, relationships between microstructure and weight. Activation frequency and BFR vary inversely with and Schaffler and Burr (1984). Within a taxonomic group, weight may be an impor- body weight (b = -1.33 and -1.09, respectively for 182 D.B. BURR macaques, both in laboratory and free-ranging settings. Logistic regressions showed that each microstructural variable predicted less than 6% of the variation in age. Because osteons do accumulate with age (at least in humans), logistic regressions using percent osteonal bone and OPD in combination were computed. These two variables together accounted for only 8% of the variation in age. Variation of Microstructure With Gender WEIGHT Fig. 1. Regression of age (years) vs. weight (kilograms) for growing macaques showing the strong correlation between the two (rz = .79). Age = .67-.22 (weight). The solid line represents the best least squares fit to the data, and the broken line represents one standard deviation about this line. Using either chi square or analysis of variance, no significant differences in bone microstructure were found between males and females (Table 3a,b). Males had larger mean osteon sizes and smaller OPD, but these were not significantly different than those in females. These data indicate that gender cannot be determined based on changes in bone microstructure. Covariance analysis highlighted no interactions between age and gender for any microstructural variable. At least as far as bone remodeling is concerned, females do not mature at a rate different than that in males. Differences in Microstructure in Growing and Adult Macaques The proportion of remodeled cortex is significantly greater in growing macaques than in adults, suggesting that apposition of primary lamellar bone may continue with age in macaques although remodeling of RMA fits), probably because of the strong relationship this new bone slows (Table 3a,b, Fig. 2). The percent of between weight and age in these growing animals. The remodeled cortex decreases after maturity because the relationship decreases exponentially with age, indicat- osteonal bone near the marrow cavity, having the ing that osteons do not accumulate very rapidly with greatest mean tissue age, is preferentially removed. maturation. This is reflected in the association be- The OPD in adult macaques, however, is greater than tween weight and the percentage of cortex occupied by in growing animals (Fig. 31, although no significant osteons (Table 1). Least squares fits show that the linear relationship exists between OPD and age within amount of remodeled cortex increases at only 5%of the the age range we studied (Fig. 4). We do not have data rate a t which the animals gain weight. Even RMA fits for osteon size in adult macaques, but the inverse reshow that the remodeled cortex increases at only 75% lation of OPD with percent of remodeled cortex can of the rate of weight gain. The OPD, however, de- only be possible if osteons become smaller with age. creases with weight gain (b = -0.26 for least squares Figure 5 shows bone microstructure a t the femoral fits and -0.83 for RMA fits; Table 1). midshaft in immature (1.8 years), young adult (6.1 The larger amount of remodeled cortex, but fewer years), and old (25 years) male macaques. These photoosteons, means that osteons become smaller as ma- micrographs also demonstrate that osteons do not accaques mature. This is indeed born out by inspection of cumulate with age in M . mulatta. the data (Table 1). Mean osteonal area increases at Differences in Microstructure in Macaques and Humans about one-third the rate that weight increases (r2 = .34; b = 0.43, RMA fit). Table 3a,b and Figures 2 and 3 clearly show that the potential for intracortical bone remodeling in maVariation of Microstructure With Age caques is less than that in humans. The proportion of Age accounts for little of the variation in the micro- femoral cortex that is remodeled t o secondary osteons structural variables examined (Table 2). Except for is much lower in both growing and mature adult those variables that are calculated using age (mu and macaques than it is in the human femur. Adult BFR), age predicts less than 20% of the variation in macaques have less than half the OPD found in adult each of the microstructural variables (r2 < .15). Age human femurs; growing macaques have even less. predicts mean osteonal size better than it predicts the DISCUSSION other variables (r2 = .15, P < .005), but the relationship is still far from linear. The need t o develop primate models to study the tisUsing polynomial fits to the data, age predicts 21- sue level effects of human skeletal physiology dictates 34% of the variation in the variables (Table 2 ) , again that normal bone remodeling during skeletal maturaignoring variables that use age in their calculation. tion be studied in some primate species. The extent to These fits are all statistically significant, a t P < .02 or which genetically determined developmental effects better. predominate over metabolic or mechanical influences, In humans, bone microstructure has been used to or how these factors might interact to produce the final predict age. This could be potentially useful in adult morphology, is not known. 183 BONE TURNOVER IN MACAQUES TABLE 1. Summary of regression fits for body weight vs. microstructure’ Variable % Osteonal bone Osteons/mm2 Fragments/mm2 MWT Mean area/osteon ( x OPD Activation frequency BFR r2 0.01 0.06 0.15 0.18 0.34 0.10 0.63 0.36 Least squares fit Intercept Slope 2.16 0.06 1.80 -0.19 0.69 -0.76* 3.81 0.17** 9.56 0.25** 2.09 -0.26* 1.94 -1.05** -2.59 -0.66** Major axis fit Intercept Slope 2.06 0.12 2.20 -0.45 6.01 -4.13 3.77 0.19 9.51 0.28 2.57 -0.57 2.53 -1.43 -1.80 1.16 RMA fit Intercept 1.09 2.78 2.63 3.45 9.28 2.98 2.38 -1.91 Slope 0.73 -0.81 -1.99 0.40 0.43 -0.83 -1.33 -1.09 ‘All data were log transformed. Least squares regressions were computed using the standard form of the allometric equation, log y = log a + b log (wei ht) The slope for least squares is defined as S , ISxx;the slope for major axis regressions is defined by 1/2S [S - S,, + ([S - S,,I2 + 4SX,2)’?.’, the’ slope for RMA regression is defined by C~yJS,,)”2 when S,, > 0, or the negative of this when S,, 2 0 . m e intercept E always defined as y -slope x. The r-value is the same for all regression fits. *P < .05. **P < .01. TABLE 2. Summary of regression fits for age vs. microstructure’ Variable % Osteonal bone Osteons/mm2 Fragments/mm2 MWT Mean area/osteon(x OPD Activation frequency BFR r2 0.07 0.04 0.11 0.07 0.15 0.07 0.30 0.30 Least squares fit a b 7.23 0.73* 7.41 -0.45 1.74 -0.19* 50.34 2.15 16.36 1.26** 9.05 -0.61 5.82 -0.85** 0.07 -0.01** r2 0.29 0.31 0.21 0.24 0.26 0.34 0.52 0.44 Polynomial fit a bl 18.53 -5.04** 18.14 -5.93** - 1.03** 3.39 20.18 17.41** 6.37 6.36** 20.51 -6.46** 12.46 -4.23** 0.11 -0.03** b2 0.62** 0.59** 0.09” - 1.63** -0.55** 0.63** 0.36** 0.002** ‘Least squares regressions fit the equation y = a + bx, where y is the microstructural variable, x is the age, a is the intercept, and b is the slope. Polynomial regressions fit the equation y = a + b,x + b,x2. *P < .05. **P < .01 Some skeletal diseases depress intracortical bone formation, while a t the same time increasing formation along the periosteal surface (Wirth et al., 1982). These include some skeletal dysplasias found primarily in growing children, and also more common afflictions of age such as senile (type 11) osteoporosis. Intracortical activation frequency appears to be normally depressed in rhesus macaques. M. mulatta may prove to be a reasonable model for skeletal pathologies in which osteonal activation rates are slow, but a poor model to study the effects of rapid turnover. Probfems in Calculating Activation Frequency From Nonlabeled Bone The calculation of activation frequency from nonlabeled bone assumes that the mean age of the tissue is known. This is a function both of the amount of time osteons have been accumulating and the rate a t which osteons are lost from view. Underlying this is the assumption that OPD increases with age, at a rate determined by activation frequency. In other words, the activation frequency is the slope of the regression line between OPD and age. If OPD decreases with age, this assumption is violated, and the calculation of a n accurate activation frequency may be invalid. The slope of the regression line of OPD on age is negative (Table 2), suggesting that OPD decreases with age in macaques, and calling into question the validity of the activation frequency calculation. However, this decline is statistically insignificant, as is the r2 value, and Figure 4 shows that no true inverse relationship exists. Moreover, OPD is greater in mature macaques than in growing ones (Table 3a,b, Fig. 3) suggesting a n overall age-related increase in OPD. In view of these facts, the calculation of activation frequency in this sample of growing macaques is probably valid and is a n accurate reflection of the actual activation frequency. Is Variation in Bone Microstructure Independent of Gender, Age, and Body Weight? In a previous experiment, we observed a convergence between bone microstructure in the femur of a series of primates and the movement patterns of these primates defined kinesiologically (Schaff ler and Burr, 1984). These observations led us to hypothesize that the nature of the mechanical stimulus is an important determinant of bone morphology a t the microstructural level. This seemed reasonable because it is well known that mechanical stimuli control bone remodeling (Frost, 1983, 1987c,d; Rubin and Lanyon, 1984, 1985; Pead et al., 1988; Lanyon et al., 19821, and osteon formation is the result of bone remodeling events within the cortex. We concluded that the amount of osteonal bone is a n indicator of how a n animal uses its limbs. There are many influences on bone besides mechan- 184 D.B. BURR TABLE 3a. Histomorphometric data for macaque and human femur and rib' % FragOsteonal Osteonsi mentsi area mm' mm' (%) (#imm') (#/mm') Femur Immature macaque (female) Immature macaque (male) Adult macaque Adult human Rib Immature macaque (4-8 years) Immature human Adult macaque Adult human 10.262 (4.18) 9.18' (3.61) 6.803 (-1 45.304 (-) - 5.03' (2.40) 4.02' (1.71) 3.103 (-1 12.475 (-1 10.416 (4.60) 7.25' (-1 - - 17.02' (4.25) 9.38' (0.68) 3.535 (km) - (-1 6.7Z6 (3.01) 2.558 (-1 17.32' (6.85) 8.96' (2.52) Activation Mean frearea/ quencx OPD (#/mm / osteon (pm2) (#/mm') year) MWT 0.87' 60.08' (0.75) (15.89) 0.97' 62.62' (1.48) (11.13) - TABLE 3b. Histomorphometric data for macaque and human femur and rib' 63.0B7 (1.98) 69.007 (4.00) 58.0010 (13.00) 68.0011 (-1 'Numbers in parentheses are standard deviations. 'Data from current study; n = 31 female and 23 males; Mean ages = 4.28 and 4.57 years, respectively. 3Schaffler and Burr, 1984; n = 1; M . muluttu; age unknown. 4Evans, 1982; n z 54 specimens from 16 individuals (12 males and 4 females) ranging in age from 33-98 years, mean age = 64.81 years. 'Evans, 1976; n = 35 (mean age = 41.5 years). 'Przybeck, 1985; n = 6; age 4-8 years, mean age = 6.0 years. 7Schock et al., 1972; based on tetracycline labeling data in ribs from 13 macaques ages 3.5-4.6 years (mean age not given but we estimate it at approximately 4.22 years). 'Wu et al., 1970; data from the ribs of growing humans combine age groups 0-9 years (mean age = 6.0 years, n = 24) and 10-19 years (mean age = 15.80 years, n = 25); bone formation rates are calculated from OPD and tetracycline labeling data, respectively. 'Przybeck, 1985; n = 9; age 13-31 years, mean age = 22.67 years. 'OSchock et al., 1975; based on tetracycline labeling data from 23 late adolescent male rhesus macaques (specific ages not provided). "Pirok et al., 1966; n = 226 males (ages 0-71-t years; mean age = 37.17 years) and 147 females (ages 0-71 years, mean age = 35.26 years). ical ones. It is possible that other influences contributed to the correlation of movement pattern and bone microstructure. Before we could conclude that microstructure and limb usage were related, it was necessary to address the normal variation related to differences in age, gender, and body weight that could significantly influence the microstructure of bone. Data from the present study suggest strongly that variation in bone microstructure occurs independent of variation in body weight, age, and gender in growing macaques. Moreover, there are no maturaticnal effects. Age-matched male and female macaques have comparable bone microstructures. Intraspecifically, body weight is not a n important determinant of OPD, although weight may be related to osteon size (Table 1).The implication of this is that increased body size has no effect on the initiation of new remodeling sites (i.e., the activation frequency and the rate of bone turnover), but does have some effect on the individual vigor of osteoblasts and osteoclasts. As in growing humans, i t suggests a decreasing individual cell level activity for osteoclasts with the onset of ma- Femur Immature macaque (female) Immature macaque (male) Adult macaque Adult human Rib Immature macaque (4-8 years) Immature human Adult macaque Adult human 21,963' (5,572) 22,965' (4,954) BFR (mm2/ mm2/ year) 1.88' 0.034' (1.87) (0.029) 1.26' 0.026' (0.87) (0.012) 38,5714 (5,925) 6.0g2 (3.04) 4.77' (1.86) 9.353 (4.29) 22.875 (5.23) 25,2006 (4,020) 17.116 (6.92) - 9.80' (1.75) 17.316 0.4356 (10.82) (0.26) 27.907 0.3607 (13.00) (0.18) 13.705 0.45g5 - 21,333' (2,357) 38,333l' (4,150) 34.41' (6.76) 18.34' (3.14) - (-1 - (-) 10.00~ 0.210~ (5.00) (0.151 11.81' (2.25) 8.78' 0.186' (2.32) (0.07) 2.67' 0.08810 (-1 (-1 0.0948 (0.014) 0.037' (0.005) 0.03011 'Numbers in parentheses are standard deviations. 'Data from current study; n = 31 females and 23 males; mean ages = 4.28 and 4.57 years, respectively. 3Larson, unpublished manuscript, 1986; n = 12; age not specified. 4The mean of data from Currey, 1964 (ages 23-89 years, mean age = 57.47 years), Jowsey, 1966 (ages 20-90 years, mean age not specified); and Martin et al., 1980 (ages 37-96 years, mean age = 69.26 years); n = 65. %tout and Gavan, unpublished manuscript; n = 104. 'Przybeck, 1985; n = 6; age 4-8 years, mean age = 6.0 years. 7Schock et al., 1972; based on tetracycline labeling data in ribs from 13 macaques ages 3.5-4.6 years (mean age not given but we estimate it at approximately 4.22 years). 'Wu et al., 1970; data from the ribs of growing humans combine age groups 0-9 years (mean age = 6.0 years, n = 24) and 10-19 years (mean age = 15.80 years, n = 25); bone formation rates are calculated from OPD and tetracycline labeling data, respectively. 'Przybeck, 1985; n = 9; ages 13-31 years, mean age = 22.67 years. "Mean of Stout and Gavan, unpubished manuscript; and Wu et al., 1970. "Parfitt, 1983, age not given. ''Pirok et al., 1966; n = 226 males (ages 0-71 + years; mean age = 37.17 years) and 147 females (ages 0-71 years, mean age = 35.26 years). turity. Because MWT did not also decline (Tables 1,a), osteoblastic refilling of resorption cavities is not affected. Decreased osteon area without a n accompanying change in MWT implies that osteoclasts are less active or fewer in adult macaques, and that the intracortical resorption-formation balance shifts in greater favor of formation as maturity is reached. The decline of OPD during development supports the idea that fewer BMUs are activated as adolescent monkeys ap- 185 BONE TURNOVER IN MACAQUES 41 50 35 40 30 30 T 25 20 20 15 10 ‘ o c t FEMUR Fig. 2. The amount of femoral cortex remodeled in adult macaques is less than 25% of that in humans, indicating the conservative nature of intracortical bone turnover in macaques. The proportion of remodeled cortex actually declines in adult macaques, suggesting that remodeling of the cortex does not keep pace with apposition of new periosteal lamellar bone. Data for adult macaques are from Schaffler and Burr (1984), while those for adult humans are from Evans (1982). Standard errors are not reported by Evans. Data for adult M . mulutta are based on a single sample because these are the only data available, although several other macaque species were examined by Schaffler and Burr (1984) with similar results. “ 9 FEMUR RIB Fig. 3. OPD in the femoral midshaft is greater in adult macaques than in growing macaques. However, there are still fewer than half as many osteons in adult macaque femurs as in adult human femurs. Intracortical remodeling in the macaque femur is more conservative than in the rib, where OPD is greater than in humans. Data for adult humans and adult macaques were taken from Larson (unpublished material) (n = 12 macaques) and Stout and Gavan, (unpublished material) (n = 104 humans). proach skeletal maturity. This also means that, in a global sense, fewer osteoclasts are recruited. Significant cortical-endosteal resorption tends to remove the larger osteons found close to the marrow cavity and skews the osteon population toward smaller sizes (Takahashi et al., 1965). A process like this occurs in adult humans as well (Currey, 1964; Landeros and Frost, 1964; Jowsey, 1966; Hattner et al., 1964). As ent in growing and adult primates from the genera cortical-endosteal resorption and periosteal apposition Ateles and Pan. Data presented here suggest that the occur with age, osteonal bone is destroyed near the rate of accumulation of osteons in growing macaques marrow cavity and replaced with nonosteonal bone does not change significantly with age. The trend periosteally. This process cannot fully account for the throughout the growth years is for no change in the changes observed in growing monkeys. Although it amount of remodeled cortex (Table 2). The OPD is would reduce the number of osteons per unit area of greater in adult macaques, as expected and as observed bone, it would also reduce the percentage of bone cortex in humans and other species (Currey, 1964; Atkinson, occupied by osteons. The amount of remodeled cortex 1965; Kerley, 1965; Evans and Bang, 1967; Singh and did not decrease in growing macaques. Moreover, the Gunberg, 1970; Ortner, 1975; Martin et al., 1980), alincrease in cross-sectional area levels off a t about 6 though OPD never approaches that in humans, even in years in both male and female macaques. Therefore, old animals (Fig. 5). Data from growing macaques are between 6 and 8 years there is not much periosteal consistent with data from adult macaques in suggestapposition of primary lamellar bone that could skew ing some inhibition on intracortical bone remodeling the results. In combination, this suggests a real decline processes relative to humans. in activation frequency, although perhaps a smaller Are Macaques Different Than Humans in the Extent of decline than indicated by the data. Osteonal Remodeling? Are Growing Macaques Different Than Adult Macaques in the Extent of Osteonal Remodeling? In previous work (Schaffler and Burr, 1984) we observed that the accumulation of osteons was not differ- There are limited data about activation frequency and BFR from the femur of any nonhuman primate, and most of the data for humans come from the rib or iliac crest. Comparative data are sketchy. The data we D.B. BURR 186 12 10 8 0 P 5 0 4 2 0 I 2 3 4 5 6 7 8 AGE Fig. 4. Regression of OPD (#/rnrn2) vs. age (years) for growing macaques showing no relationship between the two variables (r2 = .07). OPD = 9.05 - 0.61 (age). The solid line represents the best least squares fit to the data, and the broken lines represent one standard deviation about this line. have collected for the femoral midshaft are unique for that bone. Rhesus macaques have been used to study the effects of reduced loading on bone (Schock et al., 1975; Young and Schneider, 1981; Young et al., 1983,1986; Wronski and Morey, 1983) partly as a means to evaluate the effects of hypogravic environments on skeletal physiology. One reason for using macaques is that they represent an animal taxonomically closer to humans than dogs or rats. A good model for human skeletal physiology requires that the rate at which bone remodeling can be activated, and the resulting bone morphology, are also like those in humans. The data presented here suggest that intracortical remodeling in the long bones of macaques is distinct from those same processes in older humans. Although good information does not exist on the BFR in adult macaque or adult human femora, the low OPD suggests that the intracortical BFR in macaques is significantly lower than that in humans. It is possible that the reason that humans appear to have a greater OPD and higher BFR is that osteons have had a longer period over which to accumulate. Macaques in this study have had less than 8 years to accumulate osteons, whereas the humans we surveyed had 40-90 years. However, our data suggest that osteon accumulation is unlikely to be solely a function of age. There is little correlation between age and proportion of remodeled cortex (Fig. 2). OPD in the femur is only slightly greater in adult macaques than in immature macaques (Fig. 3). Moreover, Figures 5e,f clearly show the absence of an osteonized femoral cortex even in a 25 year old macaque. This is an age at which there is no question that human bone has a high OPD. OPD in the rib is greater than that in adult humans, suggesting that the low OPD in macaque femora may be less a function of an inherent inability to form osteons, than a locally conservative BFR in the macaque femur. Finally, data from Jowsey (1963) shows that human bone is heavily osteonized even by the age of 2 years. Clearly, better data is needed for younger humans and for older macaques before the issue can be resolved. Stout and Gavan (unpublished manuscript) and Larson (unpublished manuscript) observed depressed intracortical bone remodeling in the macaque femur. Likewise, Schock et al. (1975) found that immobilization in rhesus monkeys did not evoke an increased remodeling space. There was no increased activation frequency following 2 months of immobilization. This is contrary to the observation in humans that immobilization for even short periods of time significantly elevates the activation frequency for intracortical bone remodeling and alters global bone balance (Heaney, 1962; Minaire et al., 1974; Stout, 1982; Chantraine et al., 1986). The significance of these observations is that macaques may make poor models to study normal human skeletal tissue kinetic processes, and may not accurately reflect the extent to which bone can be lost in the weight bearing bones in hypogravic environments. The inhibition of remodeling activation would slow the loss of bone and lead to an underestimate of the severity of bone loss. However, macaques may make ideal models to study certain human skeletal pathologies, including age-related osteoporosis in which cortical bone turnover is depressed. It is important to understand that these observations apply only to cortical bone, and only to the intracortical envelope. As such, they do not speak to differences between macaques and humans in rates of formation or resorption drifts that may occur on cortical-endosteal and periosteal surfaces as a function of modeling processes, but apply only to intracortical remodeling. In disuse, and in many metabolic bone diseases, significant bone loss occurs through modeling processes on trabecular and cortical-endosteal surfaces as well. Bone turnover rates on these envelopes are greater than intracortically, and it is possible that the dynamics of these surfaces in the macaque more closely approximate those in humans. Direct comparisons between trabecular bone dynamics in humans and macaques have never been made, but are needed. These data paint quite a different picture of bone remodeling in macaques than data collected from rib biopsies. Przybeck (1985) found OPD in the macaque rib was 60% greater than OPD in the human rib, and that remodeling rates were comparably high (Fig. 3). Using intravital tetracycline labels, Schock et al. (1972) measured formation rates in juvenile monkey ribs that were 1.7 times greater than in juvenile humans. Based on these observations Przybeck (1985) suggested that macaques might represent excellent models for human bone remodeling. The reasons for the differences between rib and femur are unclear. They may be related to specific local factors (relationships of periosteum or marrow to bone cortex, for example), relative relationships of cortical and trabecular bone, or reflect different mechanical environments. BONE TURNOVER I N MACAQUES 187 Fig. 5. Photomicrographs showing that osteons do not accumulate with age in macaques. In each case, the higher power photomicrograph was taken from the region of the femur that had the most secondary osteons. The few secondary osteons that exist are confined to the bone with the oldest mean tissue age, i.e., near the cortical- endosteal surface. The remainder of the cortex is composed of primary lamellar bone and primary vascular spaces. The box on the low power photomicrograph highlights the area from which the higher power micrograph was taken. (a,b) 1.8 years, x 9.5, x 50; (c,d) 6.1 yrs., x 6.3, x 32; (e,D 25 years, x 6, x 32 (on overleaf). CONCLUSIONS ACKNOWLEDGMENTS 1. Variation in bone microstructure occurs independent of gender, age, and body weight in growing macaques. 2. Microstructure of the femoral midshaft cannot be used to estimate age or body weight in macaques in which these life data are unknown. 3. Bone turnover in the femur in macaques is much slower than in humans. Because of this, macaques probably make poor models to study the effects of weightlessness, but may make good models to study low turnover skeletal dysplasias. 4. Remodeling rates in macaque rib and femur are significantly different, suggesting that local factors control bone formation rate to a significant degree. This research was supported by NIH grant P40RRO3640-04 to the Caribbean Primate Research Center (Project I1 (El 4) and by NIH grant AR 39708. I wish to thank Satoshi Mori for preparation of the photomicrographs, Cary Johnson for help with tissue preparations, and Terri Stafford for help with microscopic analysis, and to express my gratitude to Jean Turnquist, Matt Kessler and the Caribbean Primate Research Center for access to the Cay0 Santiago-derived skeletal collections and for laboratory facilities. My appreciation is also extended to Bill Jungers for explaining the basis for various regression fits to the data, and to Bruce Martin and Harold Frost for reading and criticizing the manuscript prior to submission. D.B. BURR 188 f Fig. 5e-f LITERATURE CITED Atkinson, P.J. 1965 Changes in resorption spaces in femoral cortical bone with age. J . Pathol. Bacteriol., 89:173-178. Bowden, D.M. 1979 Aging in Nonhuman Primates. Van Nostrand Reinhold, New York. Burr, D.B., R.Y. Nishikawa, and D.P. Van Gerven 1989 Bone growth and remodeling in Cay0 Santiago-derived Macaca mulatta. Puerto Rico Health Sci. J., 8:191-196. Cann, C.E., H.K. Genant, and D.R. Young 1980 Comparison of vertebral and peripheral mineral losses in disuse osteoporosis in monkeys. Radiology, 134:525-529. Chantraine, A,, B. Nusgens, and Ch. M. Lapiere 1986 Bone remodeling during development of osteoporosis in paraplegia. Calcif. Tissue Int., 38r323-327. Cheverud, J.M. 1981 Epiphyseal union and dental eruption in Macaca mulatta. Am. J. Phys. Anthropol., 56r157-167. Currey, J.D. 1964 Some effects of ageing in human Haversian systems. J . Anat., 98:69-75. DeRousseau, C.J. 1985 Aging in the musculoskeletal system of rhesus monkeys: 111. Bone loss. Am. J. Phys. Anthropol., 68:157-168. Evans, F.G. 1976 Mechanical properties and histology of cortical bone from younger and older men. Anat. Rec., 185:l-12. Evans, F.G. 1982 Bone and bones. J. Biomech. Eng., 104:l-5. Evans, F.G., and S. Bang 1967 Differences and relationships between the physical properties and the microscopic structure of human femoral tibia1 and fibular cortical bone. Am. J. Anat., 120:79-88. Frost, H.M. 1983 A determinant of bone architecture. The minimum effective strain. Clin. Orthop., 175:286-292. Frost, H.M. 1987a Secondary osteon populations: an algorithm for determining mean bone tissue age. Yrbk. Phys. Anthropol., 30: 221-238. Frost, H.M. 1987b Secondary osteon population densities: an algorithm for estimating the missing osteons. Yrbk. Phys. Anthropol., 30:239-254. Frost, H.M. 1987c Bone “mass” and the “mechanostat”: a proposal. Anat. Rec., 219:l-9. Frost, H.M. 1987d The mechanostat: a proposed pathogenic mechanism of osteoporoses and the bone mass effects of mechanical and nonmechanical agents. Bone Miner., 2:73-85. Hattner, R., 0. Landeros, and H.M. Frost 1964 Are there cell behavioral modes that are unaffected by aging? Gerontology, 4(part II):34. Heaney, R.P. 1962 Radiocalcium metabolism in disuse osteoporosis in man. Am. J . Med., 33:188-200. Hylander, W.L. 1981 Patterns of stress and strain in the macaque mandible. In: Craniofacial Biology. D.S. Carlson, ed. Monograph No. 10, Craniofacial Growth Series, Center for Human Growth and Development, Ann Arbor, pp. 1-35. Hylander, W.L. 1984 Stress and strain in the mandibular symphysis of primates: a test of competing hypotheses. Am. J . Phys. Anthropol., 64:l-46. Jerome, C.P., D. Kimmel, J.A. McAlister, and D.S. Weaver 1986 Effects of ovariectomy on iliac trabecular bone in baboons (Papio anubis). Calcif. Tissue Int., 39:206-208. Jowsey, J . 1963 Microradiography of bone resorption. In: Mechanisms of Hard Tissue Destruction. R.F. Sognnaes, ed. American Association for the Advancement of Science, Washington, D.C., pp. 447469. Jowsey, J. 1966 Studies of Haversian systems in man and some animals. J. Anat., 100:857-864. Kerley, E.R. 1965 The microscopic determination of age in human bone. Am. J . Phys. Anthropol., 23:149-163. Landeros, O., and H.M. Frost 1964 The cross section size of the osteon. Henry Ford Hosp. Med. Bull., 12517-525. Lanyon, L.E., A.E. Goodship, C.J. Pye, and J.H. MacFie 1982 Mechanically adaptive bone remodelling. J. Biomech., 15:141-154. Martin, R.B., J.C. Pickett, and S. Zinaich 1980 Studies of skeletal remodeling in aging men. Clin. Orthop., 149:268-282. Miller, L.C., D.S. Weaver, J.A. McAlister, and D.R. Koritnik 1986 Effects of ovariectomy on vertebral trabecular bone in the cynomologus monkey (Macacafascicularzs). Calcif. Tissue Int., 38:6265. Minaire, P., P. Meunier, C. Edouard, J . Bernard, P. Courpron, and J. Bourret 1974 Quantitative histological data on disuse osteoporosis. Comparison with biological data. Calcif. Tissue Int., 17.5773. Ortner, D.J. 1975 Aging effects on osteon remodeling. Calcif. Tissue Int., 18t27-36. Parfitt, A.M. 1983 The physiologic and clinical significance of bone histomorphometric data. In: Bone Histomorphometry: Techniques and Interpretation. R.R. Recker, ed. CRC Press, Inc., Boca Raton, pp. 143-223. Pead, M.J., T.M. Skerry, and L.E. Lanyon 1988 Direct transformation from quiescence to bone formation in the adult periosteum following a single brief period of bone loading. J. Bone Miner. Res., 3:647-656. Pirok, D.J., J.R. Ramser, H. Takahashi, A.R. Villanueva, and H.M. Frost 1966 Normal histological, tetracycline and dynamic parameters in human, mineralized bone sections. Henry Ford Hosp. Med. Bull., 14:195-218. Przybeck, T.R. 1985 Histomorphology of the rib: bone mass and cortical remodeling. In: Behavior and Pathology of Aging in Rhesus Monkeys. R.T. Davis and C.W. Leathers, eds. Alan R. Liss, Inc., New York, pp. 303-326. Quarles, L.D., and B. Lobaugh 1989 Equivalency of various methods for estimating osteoid seam width. J. Bone Miner. Res., 4t671677. Rayner, J.M.V. 1981 Linear relations in biomechanics: the statistics of scaling functions. J . 2001. Lond., 206t415-439. Ricker, W.E. 1984 Computation and uses of central trend lines. Can. J . Zool., 62:1897-1905. Rubin, C.T., and L.E. Lanyon 1984 Regulation of bone formation by applied dynamic loads. J . Bone Joint Surg., 66A:397-402. Rubin, C.T., and L.E. Lanyon 1985 Regulation of bone mass by mechanical strain magnitude. Calcif. Tissue Int., 37:411-417. Schaffler, M.B., and D.B. Burr 1984 Primate cortical bone microstructure: relationship to locomotion. Am. J. Phys. Anthropol., 65: 191-197. Schock, C.C., F.R. Noyes, M.M. Crouch, and C.H.E. Mathews 1975 BONE TURNOVER IN MACAQUES The effects of immobility on long bone remodelling in the rhesus monkey. Henry Ford Hosp. Med. J . , 23:107-116. Schock, C.C., F.R. Noyes, and A.R. Villaneuva 1972 Measurement of Haversian bone remodelling by means of tetracycline labelling in rib of rhesus monkeys. Henry Ford Hosp, Med. J., 20:131-144. Seim, E., and B-E Saether 1983 On rethinking allometry: which regression model to use? J . Theor. Biol., 104:161-168. Simmons, D.J., C. Parvin, K.C. Smith, P. Frame, and L. Kazarian 1986 Effect of rotopositioning on the growth and maturation of mandibular bone in immobilized rhesus monkeys. Aviat. Space Environ. Med., 57:157-161. Simmons, D.J., J.E. Russell, W.V. Walker, B. Grazman, C. Oloff, and L. Kazarian 1984 Growth and maturation of mandibular bone in otherwise totally immobilized rhesus monkeys. Clin. Orthop., 182:220-230. Singh, I.J., and D.L. Gunberg 1970 Estimation of age a t death in human males from quantitative histology of bone fragments. Am. J. Phys. Anthropol., 33:373-381. Stout, S.D. 1982 The effects of long-term immobilization on the histomorphology of human cortical bone. Calcif. Tiss. Int. 34:337342. Stout, S.D., and S.L. Teitelbaum 1979 Use of histology in ancient bone research. Yrbk. Phys. Anthropol., 22t220-249. Swartz, S.M., J.E. Bertram, and A.A. Biewener 1989 Telemetered in vivo strain analysis of locomotor mechanics of brachiating gibbons. Nature, 342:270-272. Takahashi, H., B. Epker, and H.M. Frost 1965 Relation between age and size of osteons in man. Henry Ford Hosp. Med. Bull., 13: 25-31. Turnquist, J.E. 1983 Influence of age, sex, and caging on joint mobility in the patas monkey (Erythrocebus patas). Am. J . Phys. Anthropol., 61:211-221. 189 Turnquist, J.E. 1985 Passive joint mobility in patas monkeys (Erythrocebus patas): rehabilitation of caged animals after release into a free-ranging environment. Am. J . Phys. Anthropol., 67:l-6. Williams, D.D., and D.M. Bowden 1984 A nonhuman primate model for the osteopenia of aging. In: Comparative Pathobiology of Major Age-Related Diseases: Current Status and Research Frontiers. D.G. Scarpelli and G. Migaki, eds. Alan R. Liss, New York, pp. 207-219. Wirth, C.R., J. Kay, and R. Bourke 1982 Diaphyseal dysplasia (Englemann’s syndrome). A case report demonstrating a deficiency in cortical Haversian system formation. Clin. Orthop., 171:186195. Wronski, T.J., and E.R. Morey 1983 Inhibition of cortical and trabecular bone formation in the long bones of immobilized monkeys. Clin. Orthop., 181r269-276. Wu, K., K.E. Schubeck, H.M. Frost, and A.R. Villanueva 1970 Haversian bone formation rates determined by a new method in a mastodon, and in human diabetes mellitus and osteoporosis. Calcif. Tissue Int., 22r204-219. Young, D.R., W.H. Howard, and D. Orne 1977 In vivo bone strain telemetry in monkeys ( M . nernestrzna).J . Biomech. Eng., 1:104109. Young, D.R., and V.S. Schneider 1981 Radiographic evidence of disuse osteoporosis in the monkey ( M . nernestrina). Calcif. Tissue Int., 33:631-639. Young, D.R., W.J. Niklowitz, and C.R. Steele 1983 Tibia1 changes in experimental disease osteoporosis in the monkey. Calcif. Tiss. Int., 35304-308. Young, D.R., W.J. Niklowitz, R.J. Brown, and W.S.S. Jee 1986 Immobilization-associated osteoporosis in primates. Bone, 7t109117.