- Email: [email protected]
Sympathetic nervous system as transmitter of mechanical loading in bone
Joint Bone Spine 70 (2003) 515–519 www.elsevier.com/locate/bonsoi
Original article
Sympathetic nervous system as transmitter of mechanical loading in bone Régis Levasseur a,b,*, Jean-Pierre Sabatier b,c, Céline Potrel-Burgot b, Bertrand Lecoq a,b, Christian Creveuil d, Christian Marcelli a,b b
a Service de Rhumatologie, CHU Cote de Nacre, 14033 Caen cedex, France Centre d’Etude du Tissu Osseux et de sa Résistance (CETOR), Faculté de Médecine, 14032 Caen cedex, France c Service de Médecine Nucléaire, CHU, 14033 Caen cedex, France d Service d’Informatique médicale, CHU, 14033 Caen cedex, France
Received 5 June 2003; accepted 10 July 2003
Abstract Sympathetic innervation has been demonstrated in bone. Adrenergic stimulation is one of the transmitters of bone loss by uncoupling between decreased bone formation and increased bone resorption. Objective. – By using a non-specific antagonist of b-adrenergic pathway (propranolol per os), we hypothesized that we could rescue the uncoupling induced mechanical unloading bone loss in the rat model of tail-suspension. Materials and methods. – Twenty-two female Wistar rats, 12 week-old, have been divided into three groups: eight tail-suspended rats (SR), six tail-suspended rats treated by propranolol (SRP) and eight non-suspended rats (NSR) during 30 days. Bone mineral density (BMD, g/cm2) has been measured by DXA (Hologic QDR-4500A) at D0 and D30 of the study, in the distal femoral metaphysis (DFM), the femoral diaphysis (FD), the whole body (WB, g) and body composition. Results. – Between D0 and D30, in DFM a significant variation in BMD is observed between NSR and SR (% BMD change: NSR +15.6 ± 3.1% vs SR –1.0 ± 1.4%, P < 0.0001) and BMD rescue in SRP group (% BMD change SRP +5.3 ± 1.5% vs SR –1.0 ± 1.4%, P = 0.03). In FD, gain of BMD is significant in NSR compared to SR (+17.5 ± 1.5% vs +8.2 ± 2.8%, P = 0.007) and to SRP (+17.5 ± 1.5% vs +10.1 ± 2.4%, P = 0.046). Gain in SRP group is not significant compared to SR group (P = 0.6). In WB, SRP gain more BMD than NSR (+14.0 ± 1.8% vs +5.4 ± 0.7%, P = 0.0002) and than SR (+14.0 ± 1.8% vs +7.8 ± 1.4%, P = 0.0043). There is no difference between NSR and SR groups (P = 0.19). Conclusion. – We demonstrate that b-adrenergic pathway of sympathetic nervous system is a major transmitter pathway of mechanical loading in rat bone. A specific study is necessary to analyse a possible systemic effect of propranolol in rat bone. Propranolol could be used to prevent the induced mechanical unloading bone loss as weightlessness. © 2003 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. Keywords: Bone mineral density; Hind limb unloading; Sympathetic nervous system; Mechanical load; Propranolol
1. Introduction Bone tissue is in a constant state of remodelling. This process involves osteoclasts, responsible for the degradation of the bone matrix, and also osteoblasts, the building blocks of reconstitution successively [1]. This remodelling allows bone tissue to maintain mass and mineralization levels adapted to the mechanical loading placed on it. Hormonal factors (estrogens, parathormone, vitamin D), growth factors (IGF-1, TGFb, BMPs), cytokines (interleukines, TNFa, * Corresponding author. E-mail address: [email protected] (R. Levasseur). © 2003 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. doi:10.1016/j.jbspin.2003.07.006
OPG), membrane receptors (LRP5) and local mechanical loading are the most influential regulators of bone cell activity. In the tail-suspension rat model, there is a deficit of bone mass gain in the hind limbs which are no longer subject to any mechanical loading [2]. This loss is linked to the uncoupling of diminished osteoblastic activity and the initially increased osteoclastic activity. It has recently been shown that the sympathetic system innervates bone tissue [3]. b-adrenergic receptors are present on the osteoblast surface and treatment using a non-selective antagonist of the b-adrenergic pathway (propranolol) permits the increase of bone mass in the murine model. This bone mass increase also
516
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519
occurs because of an uncoupling mechanism between increased osteoblastic activity and normal or diminished osteoclastic activity [3]. We have put forth the hypothesis that the sympathetic system is a mediator of mechanical loading in bone and the study we have performed confirms that a non-selective pharmacological block of the b-adrenergic pathway (propranolol per os) allows the prevention of bone loss due to the absence of mechanical loading in the tail-suspended rat model. In addition, the innervation of adipose tissue has been recognized for over 20 years [4], which is why we have also studied the effects of suspension alone and of propranolol treatment on the whole body composition of female rats. 2. Materials and methods 2.1. Protocol Twenty-two female Wistar rats, 12 week- old, from the IFFA CREDO breeders (l’Abresle, France) were divided into three groups: a first group of non-suspended rats (NSR, n = 8), a second group of rats suspended by the tail (SR, n = 8), and a third group of rats suspended by the tail treated with propranolol (Sigma) which was diluted in their drinking water at a concentration of 0.5 g/l for a period of 30 days (SRP, n = 6). The rats were housed in the animal nursery of the Medicine University of Caen in individual cages furnished to allow suspension by the tail with an angle of incline permitting normal support on the front legs. They were kept at a temperature of 22 °C and subject to 12 h cycles of light and darkness. The rats were fed a normal diet containing 0.4% calcium, 0.3% phosphorus and 3 UI of vitamin D/g (UAR, villemoisson/orge). They were allowed to drink demineralized water in unlimited amounts for the 30 days of the study. Each rat was weighed on day 0 and day 30. 2.2. Densitometric analysis Bone mineral density (BMD, g/cm2) was measured on day 0 and day 30 of the protocol using dual energy X-ray absorp-
tiometry (DXA) with a hologic QDR-4500A (Waltham, MA) using a software for measuring small animals with ultra-high resolution mode. A daily quality control of the apparatus was performed using a spine and a stairs step phantoms. The rats were anesthetized with an intra-peritoneal injection of ketamine (100 mg/kg) and midazolam (0.1 ml). They were laid in a ventral position on a thin Plexiglas plate. The rats were stretched out to avoid a scoliotic position and the back legs were held in external rotation with adhesive tape (Fig. 1A). We studied the distal femoral metaphysis (DFM) and the femoral diaphysis (FD) of the right femur in ultra-high resolution mode. The hip and knee were at an angle of 90° while this measurement was taken (Fig. 1B). There is no significant difference in femoral length between suspended and non-suspended rats [5,6], which enables us to analyse the same metaphysis area in both parts of the study. For the analysis, the DFM area of interest was defined as the area corresponding to the secondary spongiosa under the growth plate containing the maximum of trabecular bone. It corresponds to a rectangle whose base is located 20 pixels (6.14 mm) from the epiphysis articular extremity and whose height is 10 pixels (3.07 mm) (Fig. 1B). The height of 10 pixels represents the area losing the most BMD in the ovariectomized rat model. We confirmed the interest in this area with a histological analysis (Levasseur and Sabatier, personal data). The area of femoral diaphysis interest (FD: exclusively cortical bone) corresponds to one-third of a femur when the femoral length is divided into three equal parts (Fig. 1B). We were also interested in the measurements of the whole body (WB), lean mass (LM, g) and fat mass (FM, g). We have excluded the tail from the area to be analysed because of the adhesive tape used for the two suspended groups. To calculate the reproducibility of the method, we measured the entire body 10 times and four rats femurs after repositioning. We obtained average variation coefficients (CV = standard deviation × 100/average) given in Table 1.
Fig. 1. Densitometric region of interest whole body (A) and femoral (B). FD, femoral diaphysis; DFM, distal femoral metaphysis.
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519 Table 2 BMD variations of the femoral cortical between D0 and D30
Table 1 Reproducibility of densitometric measurements CV
BMD WB 0.59%
BMD DFM 1.13%
BMD FD 1.20%
517
LM 0.51%
FM 4.68%
WB, whole body; DFM, distal femoral metaphysis; FD, femoral diaphysis; LM, lean mass; FM, fat mass.
NSR SR SRP
Mean (%) +17.5 +8.2 +10.1
SEE 1.5 2.8 2.4
P 0.007 NS 0.0458
vs SR vs SRP vs NSR
NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol; SEE, standard error of estimate.
2.3. Statistical analysis All results are expressed as average ± standard error of estimate. The average gain in BMD, lean mass or fat mass for each group match the average individual gain. They are expressed as percentage of the initial value. The quantitative data were analysed using ANOVA and when the ANOVA value F was significant, the Fisher PLSD test was used in a two by two comparisons. Values of P ≤ 0.05 were considered statistically significant. All statistical tests were performed using the Statview software (version 5.0).
In the DFM, between D0 and D30, the ANOVA analysis of variations in the three groups is significant (F = 17.1, P < 0.0001). The BMD variation between D0 and D30 is significantly different between the SR group and the NSR group (NSR: +15.6 ± 3.1% vs SR: –1.0 ± 1.4%, P < 0.0001) (Fig. 2). The SRP group shows a significant BMD gain compared to SR group (SRP: +5.3 ± 1.5% vs SR: –1.0 ±1.4%, P = 0.03). However, the BMD gain in SRP group is significantly lower than that of the NSR group (P = 0.006) (Fig. 2). In the FD, between D0 and D30, the ANOVA analysis is significant (F = 5.0, P = 0.018). There is a signifi-
cantly greater BMD increase in the NSR group than in the SR group (+17.5 ± 1.5% vs +8.2 ± 2.8%, P = 0.007) and in the RSP (+17.5 ± 1.5% vs +10.1 ± 2.4%, P = 0.046) (Table 2). There is no significant difference in BMD when comparing the RSP and the SR groups (P = 0.6). With regards to the WB, between D0 and D30, the ANOVA analysis is significant (F = 10.4, P = 0.0009); there is a significantly higher BMD increase in the RSP group than in the NSR (+14.0 ± 1.8% vs +5.4 ± 0.7%, P = 0.0002) and in the SR group (+14.0 ± 1.8% vs +7.8 ± 1.4%, P = 0.0043) (Fig. 3). There is no significant difference between the NSR and the SR groups (P = 0.19). We also measured weight gain in the three groups between D0 and D30. The propranolol taken by the suspended rats (SRP) caused a slowing in weight gain as compared to the NSR group (NSR: +7.8 ± 0.6% vs SRP: +3.9 ± 1.1%) (Table 3). However, the ANOVA analysis of the three groups is not significant (F = 2, P = 0.15). We observed a very significant loss of fat mass (F = 14.6, P = 0.0001) in the SRP group compared to the NSR (–35.6 ± 5.5% vs +1.6 ± 4.3%, P < 0.0001) and to the SR groups (–35.6 ± 5.5% vs –13.8 ± 4.5%, P = 0.005) (Fig. 4). We did not observe a significant difference among the three groups concerning the lean mass increase (Table 3).
Fig. 2. BMD variations of distal femoral metaphysis between D0 and D30. NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol.
Fig. 3. BMD variations of whole body between D0 and D30. NSR, nonsuspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol.
3. Results
Table 3 Weight, fat mass and lean mass variations between D0 and D30 NSR SR SRP
Weight gain (%) +7.8 ± 0.6 +5.7 ± 1.8 +3.9 ± 1.1
vs SR vs SRP vs NSR
P NS NS NS
FM gain (%) +1.6 ± 4.3 –13.8 ± 4.5 –35.6 ± 5.5
P 0.0256 0.0051 <0.0001
LM gain (%) +8.1 ± 2.1 +6.7 ± 1.6 +9.1 ± 1.6
NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol; NS, non-significant; LM, lean mass; FM, fat mass.
P NS NS NS
518
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519
Fig. 4. Fat mass variations of whole body between D0 and D30. NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol.
4. Discussion The recent discovery of a sympathetic innervation of bone tissue and its role in the regulation of bone remodelling [3] is of major interest in known situations where uncoupling between osteoclasts and osteoblasts occurs (immobilization, algodystrophy, stress fracture). The tail-suspended rat model, recognized for both its absence of mechanical loading and its weightlessness [2,7,8], is a model known to cause uncoupling between osteoblasts and osteoclasts. We have put forth the hypothesis that the sympathetic nervous system could be a mediator in bone loss due to the absence of mechanical loading in this model. We used a pharmacological approach treating tail-suspended rats with a non-specific b-adrenergic inhibitor, propranolol per os, to show that there is significant bone mass recovery in the treated tail-suspended group as compared to the merely suspended group. We have also demonstrated for the first time that the sympathetic nervous system is a mediator of mechanical loading in bone. This strongly suggests that the reduction of the gain in bone mass, observed in the tail-suspended group whose hindlimbs were not subject to mechanical loading, is partially due to an activation of the sympathetic system. This effect occurs in both the trabecular part and the cortical part; however, the bone mass recovery induced by the taking of propranolol in the tail-suspended group is only significant in the trabecular bone rich metaphysis area. There is a tendency towards recovery in the cortical area but it is not significant in our study. Although the effect is present in both regions, it is stronger in the trabecular area and, considering the small number of beta-blocked animals (n = 6), is not significant in the cortical area. We did not observe any effect on longitudinal bone growth as has previously been shown [5,6]; this might explain the absence of innervation in growth plate and the absence of b-adrenergic receptors in the growth plate chondrocytes. Under the effect of propranolol, we have observed a BMD increase over the whole body. This may be the result of two effects: a general systemic effect and a local effect due to changes in mechanical loading. In the literature, a systemic effect was observed in mice [3] and in rats, showing an increase in bone mineral apposition rate [9]. On
the other hand, in a post-menopausal female population, we did not observe any effect of beta-blockers on bone mass [11]. The model we used did not allow us to distinguish a systemic effect from a local effect. It would be useful to perform another study whose aim is to specifically analyse the systemic effect. In the tail-suspended rats, taking propranolol, we observed a tendency toward weight loss. This effect of tail-suspension on weight is known [8]. We have not seen that tail-suspension has any effect on lean mass. Stein et al. [9] observed a decrease in weight of the soleus muscle most likely because of this muscle decreased activity due to the suspension. In our study, the lean mass measurement, concerning the entire body, resulted in the effects of variations with mechanical loading on the weight-bearing frontlimbs and the suspended hindlimbs. We observed a significant decrease in fat mass in the tail-suspended group an in the tail-suspended group treated with propranolol. The sympathetic system regulates the metabolism of fats by changing the anti-lipolytic a-adrenergic and lipolytic b-adrenergic balance [10]. It is possible that suspension acts either on the decrease in fat storage or their heightened consumption or both mechanisms at the same time. However, we can only report this effect without being able to explain it, as our model does not permit us to regulate the a-adrenergic pathway. In conclusion, our study shows that the sympathetic nervous system is implicated in the physiopathology of osteoporosis due to the absence of mechanical loading (weightlessness, immobilization) and allows us to envision this responsibility for other pathologies such as algodystrophy and stress fracture. Hence, beta-lockers could be used as a therapeutic solution in these different situations.
Acknowledgements We would like to thank Mr. Bruno Philoxène for his help in the animal nursery and Mrs. Muriel Henry for her help in preparing the manuscript.
References [1]
Frost HM. Tetracycline-based histological analysis of bone remodeling. Calcif Tissue Res 1969;3:211–37.
[2]
Morey-Holton ER, Globus RK. Hindlimb unloading of growing rats: a model for predicting skeletal changes during space flight. Bone 1998; 22:83S–8S.
[3]
Takeda S, Elefteriou F, Levasseur R, Liu X, Zhao L, Parker KL, et al. Leptin regulates bone formation via the sympathetic nervous system. Cell 2002;111:305–17.
[4]
Fain JN, Garcija-Sainz JA. Adrenergic regulation of adipocyte metabolism. J Lipid Res 1983;24:945–66.
[5]
Machwate M, Zerath E, Holy X, Pastoureau P, Marie PJ. Insulin-like growth factor-I increases trabecular bone formation and osteoblastic cell proliferation in unloaded rats. Endocrinology 1994;134:1031–8.
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519 [6]
Spengler DM, Morey ER, Carter DR, Turner RT, Baylink DJ. Effects of spaceflight on structural and material strength of growing bone. Proc Soc Exp Biol Med 1983;174:224–8.
[7]
Vico L, Novikov VE, Very JM, Chappard D, Alexandre C. Effects of a 40 day tail-suspension on rat weight-bearing bones. Physiologist 1990;33:S96–7.
[8]
Vico L, Bourrin S, Very JM, Radziszowska M, Collet P, Alexandre C. Bone changes in 6-mo-old rats after head-down suspension and a reambulation period. J Appl Physiol 1995;79:1426–33.
[9]
519
Stein T, Schluter M, Galante A, Soteropoulos P, Tolias P, Grindeland R, et al. Energy metabolism pathways in rat muscle under conditions of simulated microgravity. J Nutr Biochem 2002;13:471. [10] Christin L, Ravussin E, Bogardus C, Howard BV. The effect of propranolol on free fatty acid mobilization and resting metabolic rate. Metabolism 1989;38:439–44. [11] Levasseur R, Dargent P, Sabatier JP, Marcelli C, Bréart G. Current use of beta-blockers is not associated with bone mineral density and fracture risk in postmenopausal elderly women: results from the EPIDOS study, submitted for publication.
Original article
Sympathetic nervous system as transmitter of mechanical loading in bone Régis Levasseur a,b,*, Jean-Pierre Sabatier b,c, Céline Potrel-Burgot b, Bertrand Lecoq a,b, Christian Creveuil d, Christian Marcelli a,b b
a Service de Rhumatologie, CHU Cote de Nacre, 14033 Caen cedex, France Centre d’Etude du Tissu Osseux et de sa Résistance (CETOR), Faculté de Médecine, 14032 Caen cedex, France c Service de Médecine Nucléaire, CHU, 14033 Caen cedex, France d Service d’Informatique médicale, CHU, 14033 Caen cedex, France
Received 5 June 2003; accepted 10 July 2003
Abstract Sympathetic innervation has been demonstrated in bone. Adrenergic stimulation is one of the transmitters of bone loss by uncoupling between decreased bone formation and increased bone resorption. Objective. – By using a non-specific antagonist of b-adrenergic pathway (propranolol per os), we hypothesized that we could rescue the uncoupling induced mechanical unloading bone loss in the rat model of tail-suspension. Materials and methods. – Twenty-two female Wistar rats, 12 week-old, have been divided into three groups: eight tail-suspended rats (SR), six tail-suspended rats treated by propranolol (SRP) and eight non-suspended rats (NSR) during 30 days. Bone mineral density (BMD, g/cm2) has been measured by DXA (Hologic QDR-4500A) at D0 and D30 of the study, in the distal femoral metaphysis (DFM), the femoral diaphysis (FD), the whole body (WB, g) and body composition. Results. – Between D0 and D30, in DFM a significant variation in BMD is observed between NSR and SR (% BMD change: NSR +15.6 ± 3.1% vs SR –1.0 ± 1.4%, P < 0.0001) and BMD rescue in SRP group (% BMD change SRP +5.3 ± 1.5% vs SR –1.0 ± 1.4%, P = 0.03). In FD, gain of BMD is significant in NSR compared to SR (+17.5 ± 1.5% vs +8.2 ± 2.8%, P = 0.007) and to SRP (+17.5 ± 1.5% vs +10.1 ± 2.4%, P = 0.046). Gain in SRP group is not significant compared to SR group (P = 0.6). In WB, SRP gain more BMD than NSR (+14.0 ± 1.8% vs +5.4 ± 0.7%, P = 0.0002) and than SR (+14.0 ± 1.8% vs +7.8 ± 1.4%, P = 0.0043). There is no difference between NSR and SR groups (P = 0.19). Conclusion. – We demonstrate that b-adrenergic pathway of sympathetic nervous system is a major transmitter pathway of mechanical loading in rat bone. A specific study is necessary to analyse a possible systemic effect of propranolol in rat bone. Propranolol could be used to prevent the induced mechanical unloading bone loss as weightlessness. © 2003 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. Keywords: Bone mineral density; Hind limb unloading; Sympathetic nervous system; Mechanical load; Propranolol
1. Introduction Bone tissue is in a constant state of remodelling. This process involves osteoclasts, responsible for the degradation of the bone matrix, and also osteoblasts, the building blocks of reconstitution successively [1]. This remodelling allows bone tissue to maintain mass and mineralization levels adapted to the mechanical loading placed on it. Hormonal factors (estrogens, parathormone, vitamin D), growth factors (IGF-1, TGFb, BMPs), cytokines (interleukines, TNFa, * Corresponding author. E-mail address: [email protected] (R. Levasseur). © 2003 Éditions scientifiques et médicales Elsevier SAS. All rights reserved. doi:10.1016/j.jbspin.2003.07.006
OPG), membrane receptors (LRP5) and local mechanical loading are the most influential regulators of bone cell activity. In the tail-suspension rat model, there is a deficit of bone mass gain in the hind limbs which are no longer subject to any mechanical loading [2]. This loss is linked to the uncoupling of diminished osteoblastic activity and the initially increased osteoclastic activity. It has recently been shown that the sympathetic system innervates bone tissue [3]. b-adrenergic receptors are present on the osteoblast surface and treatment using a non-selective antagonist of the b-adrenergic pathway (propranolol) permits the increase of bone mass in the murine model. This bone mass increase also
516
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519
occurs because of an uncoupling mechanism between increased osteoblastic activity and normal or diminished osteoclastic activity [3]. We have put forth the hypothesis that the sympathetic system is a mediator of mechanical loading in bone and the study we have performed confirms that a non-selective pharmacological block of the b-adrenergic pathway (propranolol per os) allows the prevention of bone loss due to the absence of mechanical loading in the tail-suspended rat model. In addition, the innervation of adipose tissue has been recognized for over 20 years [4], which is why we have also studied the effects of suspension alone and of propranolol treatment on the whole body composition of female rats. 2. Materials and methods 2.1. Protocol Twenty-two female Wistar rats, 12 week- old, from the IFFA CREDO breeders (l’Abresle, France) were divided into three groups: a first group of non-suspended rats (NSR, n = 8), a second group of rats suspended by the tail (SR, n = 8), and a third group of rats suspended by the tail treated with propranolol (Sigma) which was diluted in their drinking water at a concentration of 0.5 g/l for a period of 30 days (SRP, n = 6). The rats were housed in the animal nursery of the Medicine University of Caen in individual cages furnished to allow suspension by the tail with an angle of incline permitting normal support on the front legs. They were kept at a temperature of 22 °C and subject to 12 h cycles of light and darkness. The rats were fed a normal diet containing 0.4% calcium, 0.3% phosphorus and 3 UI of vitamin D/g (UAR, villemoisson/orge). They were allowed to drink demineralized water in unlimited amounts for the 30 days of the study. Each rat was weighed on day 0 and day 30. 2.2. Densitometric analysis Bone mineral density (BMD, g/cm2) was measured on day 0 and day 30 of the protocol using dual energy X-ray absorp-
tiometry (DXA) with a hologic QDR-4500A (Waltham, MA) using a software for measuring small animals with ultra-high resolution mode. A daily quality control of the apparatus was performed using a spine and a stairs step phantoms. The rats were anesthetized with an intra-peritoneal injection of ketamine (100 mg/kg) and midazolam (0.1 ml). They were laid in a ventral position on a thin Plexiglas plate. The rats were stretched out to avoid a scoliotic position and the back legs were held in external rotation with adhesive tape (Fig. 1A). We studied the distal femoral metaphysis (DFM) and the femoral diaphysis (FD) of the right femur in ultra-high resolution mode. The hip and knee were at an angle of 90° while this measurement was taken (Fig. 1B). There is no significant difference in femoral length between suspended and non-suspended rats [5,6], which enables us to analyse the same metaphysis area in both parts of the study. For the analysis, the DFM area of interest was defined as the area corresponding to the secondary spongiosa under the growth plate containing the maximum of trabecular bone. It corresponds to a rectangle whose base is located 20 pixels (6.14 mm) from the epiphysis articular extremity and whose height is 10 pixels (3.07 mm) (Fig. 1B). The height of 10 pixels represents the area losing the most BMD in the ovariectomized rat model. We confirmed the interest in this area with a histological analysis (Levasseur and Sabatier, personal data). The area of femoral diaphysis interest (FD: exclusively cortical bone) corresponds to one-third of a femur when the femoral length is divided into three equal parts (Fig. 1B). We were also interested in the measurements of the whole body (WB), lean mass (LM, g) and fat mass (FM, g). We have excluded the tail from the area to be analysed because of the adhesive tape used for the two suspended groups. To calculate the reproducibility of the method, we measured the entire body 10 times and four rats femurs after repositioning. We obtained average variation coefficients (CV = standard deviation × 100/average) given in Table 1.
Fig. 1. Densitometric region of interest whole body (A) and femoral (B). FD, femoral diaphysis; DFM, distal femoral metaphysis.
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519 Table 2 BMD variations of the femoral cortical between D0 and D30
Table 1 Reproducibility of densitometric measurements CV
BMD WB 0.59%
BMD DFM 1.13%
BMD FD 1.20%
517
LM 0.51%
FM 4.68%
WB, whole body; DFM, distal femoral metaphysis; FD, femoral diaphysis; LM, lean mass; FM, fat mass.
NSR SR SRP
Mean (%) +17.5 +8.2 +10.1
SEE 1.5 2.8 2.4
P 0.007 NS 0.0458
vs SR vs SRP vs NSR
NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol; SEE, standard error of estimate.
2.3. Statistical analysis All results are expressed as average ± standard error of estimate. The average gain in BMD, lean mass or fat mass for each group match the average individual gain. They are expressed as percentage of the initial value. The quantitative data were analysed using ANOVA and when the ANOVA value F was significant, the Fisher PLSD test was used in a two by two comparisons. Values of P ≤ 0.05 were considered statistically significant. All statistical tests were performed using the Statview software (version 5.0).
In the DFM, between D0 and D30, the ANOVA analysis of variations in the three groups is significant (F = 17.1, P < 0.0001). The BMD variation between D0 and D30 is significantly different between the SR group and the NSR group (NSR: +15.6 ± 3.1% vs SR: –1.0 ± 1.4%, P < 0.0001) (Fig. 2). The SRP group shows a significant BMD gain compared to SR group (SRP: +5.3 ± 1.5% vs SR: –1.0 ±1.4%, P = 0.03). However, the BMD gain in SRP group is significantly lower than that of the NSR group (P = 0.006) (Fig. 2). In the FD, between D0 and D30, the ANOVA analysis is significant (F = 5.0, P = 0.018). There is a signifi-
cantly greater BMD increase in the NSR group than in the SR group (+17.5 ± 1.5% vs +8.2 ± 2.8%, P = 0.007) and in the RSP (+17.5 ± 1.5% vs +10.1 ± 2.4%, P = 0.046) (Table 2). There is no significant difference in BMD when comparing the RSP and the SR groups (P = 0.6). With regards to the WB, between D0 and D30, the ANOVA analysis is significant (F = 10.4, P = 0.0009); there is a significantly higher BMD increase in the RSP group than in the NSR (+14.0 ± 1.8% vs +5.4 ± 0.7%, P = 0.0002) and in the SR group (+14.0 ± 1.8% vs +7.8 ± 1.4%, P = 0.0043) (Fig. 3). There is no significant difference between the NSR and the SR groups (P = 0.19). We also measured weight gain in the three groups between D0 and D30. The propranolol taken by the suspended rats (SRP) caused a slowing in weight gain as compared to the NSR group (NSR: +7.8 ± 0.6% vs SRP: +3.9 ± 1.1%) (Table 3). However, the ANOVA analysis of the three groups is not significant (F = 2, P = 0.15). We observed a very significant loss of fat mass (F = 14.6, P = 0.0001) in the SRP group compared to the NSR (–35.6 ± 5.5% vs +1.6 ± 4.3%, P < 0.0001) and to the SR groups (–35.6 ± 5.5% vs –13.8 ± 4.5%, P = 0.005) (Fig. 4). We did not observe a significant difference among the three groups concerning the lean mass increase (Table 3).
Fig. 2. BMD variations of distal femoral metaphysis between D0 and D30. NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol.
Fig. 3. BMD variations of whole body between D0 and D30. NSR, nonsuspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol.
3. Results
Table 3 Weight, fat mass and lean mass variations between D0 and D30 NSR SR SRP
Weight gain (%) +7.8 ± 0.6 +5.7 ± 1.8 +3.9 ± 1.1
vs SR vs SRP vs NSR
P NS NS NS
FM gain (%) +1.6 ± 4.3 –13.8 ± 4.5 –35.6 ± 5.5
P 0.0256 0.0051 <0.0001
LM gain (%) +8.1 ± 2.1 +6.7 ± 1.6 +9.1 ± 1.6
NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol; NS, non-significant; LM, lean mass; FM, fat mass.
P NS NS NS
518
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519
Fig. 4. Fat mass variations of whole body between D0 and D30. NSR, non-suspended rat; SR, suspended rat; SRP, suspended rat treated by propranolol.
4. Discussion The recent discovery of a sympathetic innervation of bone tissue and its role in the regulation of bone remodelling [3] is of major interest in known situations where uncoupling between osteoclasts and osteoblasts occurs (immobilization, algodystrophy, stress fracture). The tail-suspended rat model, recognized for both its absence of mechanical loading and its weightlessness [2,7,8], is a model known to cause uncoupling between osteoblasts and osteoclasts. We have put forth the hypothesis that the sympathetic nervous system could be a mediator in bone loss due to the absence of mechanical loading in this model. We used a pharmacological approach treating tail-suspended rats with a non-specific b-adrenergic inhibitor, propranolol per os, to show that there is significant bone mass recovery in the treated tail-suspended group as compared to the merely suspended group. We have also demonstrated for the first time that the sympathetic nervous system is a mediator of mechanical loading in bone. This strongly suggests that the reduction of the gain in bone mass, observed in the tail-suspended group whose hindlimbs were not subject to mechanical loading, is partially due to an activation of the sympathetic system. This effect occurs in both the trabecular part and the cortical part; however, the bone mass recovery induced by the taking of propranolol in the tail-suspended group is only significant in the trabecular bone rich metaphysis area. There is a tendency towards recovery in the cortical area but it is not significant in our study. Although the effect is present in both regions, it is stronger in the trabecular area and, considering the small number of beta-blocked animals (n = 6), is not significant in the cortical area. We did not observe any effect on longitudinal bone growth as has previously been shown [5,6]; this might explain the absence of innervation in growth plate and the absence of b-adrenergic receptors in the growth plate chondrocytes. Under the effect of propranolol, we have observed a BMD increase over the whole body. This may be the result of two effects: a general systemic effect and a local effect due to changes in mechanical loading. In the literature, a systemic effect was observed in mice [3] and in rats, showing an increase in bone mineral apposition rate [9]. On
the other hand, in a post-menopausal female population, we did not observe any effect of beta-blockers on bone mass [11]. The model we used did not allow us to distinguish a systemic effect from a local effect. It would be useful to perform another study whose aim is to specifically analyse the systemic effect. In the tail-suspended rats, taking propranolol, we observed a tendency toward weight loss. This effect of tail-suspension on weight is known [8]. We have not seen that tail-suspension has any effect on lean mass. Stein et al. [9] observed a decrease in weight of the soleus muscle most likely because of this muscle decreased activity due to the suspension. In our study, the lean mass measurement, concerning the entire body, resulted in the effects of variations with mechanical loading on the weight-bearing frontlimbs and the suspended hindlimbs. We observed a significant decrease in fat mass in the tail-suspended group an in the tail-suspended group treated with propranolol. The sympathetic system regulates the metabolism of fats by changing the anti-lipolytic a-adrenergic and lipolytic b-adrenergic balance [10]. It is possible that suspension acts either on the decrease in fat storage or their heightened consumption or both mechanisms at the same time. However, we can only report this effect without being able to explain it, as our model does not permit us to regulate the a-adrenergic pathway. In conclusion, our study shows that the sympathetic nervous system is implicated in the physiopathology of osteoporosis due to the absence of mechanical loading (weightlessness, immobilization) and allows us to envision this responsibility for other pathologies such as algodystrophy and stress fracture. Hence, beta-lockers could be used as a therapeutic solution in these different situations.
Acknowledgements We would like to thank Mr. Bruno Philoxène for his help in the animal nursery and Mrs. Muriel Henry for her help in preparing the manuscript.
References [1]
Frost HM. Tetracycline-based histological analysis of bone remodeling. Calcif Tissue Res 1969;3:211–37.
[2]
Morey-Holton ER, Globus RK. Hindlimb unloading of growing rats: a model for predicting skeletal changes during space flight. Bone 1998; 22:83S–8S.
[3]
Takeda S, Elefteriou F, Levasseur R, Liu X, Zhao L, Parker KL, et al. Leptin regulates bone formation via the sympathetic nervous system. Cell 2002;111:305–17.
[4]
Fain JN, Garcija-Sainz JA. Adrenergic regulation of adipocyte metabolism. J Lipid Res 1983;24:945–66.
[5]
Machwate M, Zerath E, Holy X, Pastoureau P, Marie PJ. Insulin-like growth factor-I increases trabecular bone formation and osteoblastic cell proliferation in unloaded rats. Endocrinology 1994;134:1031–8.
R. Levasseur et al. / Joint Bone Spine 70 (2003) 515–519 [6]
Spengler DM, Morey ER, Carter DR, Turner RT, Baylink DJ. Effects of spaceflight on structural and material strength of growing bone. Proc Soc Exp Biol Med 1983;174:224–8.
[7]
Vico L, Novikov VE, Very JM, Chappard D, Alexandre C. Effects of a 40 day tail-suspension on rat weight-bearing bones. Physiologist 1990;33:S96–7.
[8]
Vico L, Bourrin S, Very JM, Radziszowska M, Collet P, Alexandre C. Bone changes in 6-mo-old rats after head-down suspension and a reambulation period. J Appl Physiol 1995;79:1426–33.
[9]
519
Stein T, Schluter M, Galante A, Soteropoulos P, Tolias P, Grindeland R, et al. Energy metabolism pathways in rat muscle under conditions of simulated microgravity. J Nutr Biochem 2002;13:471. [10] Christin L, Ravussin E, Bogardus C, Howard BV. The effect of propranolol on free fatty acid mobilization and resting metabolic rate. Metabolism 1989;38:439–44. [11] Levasseur R, Dargent P, Sabatier JP, Marcelli C, Bréart G. Current use of beta-blockers is not associated with bone mineral density and fracture risk in postmenopausal elderly women: results from the EPIDOS study, submitted for publication.