- Email: [email protected]
Muscle adaptation during distraction osteogenesis in skeletally immature and mature rabbits
ELSEVIER
Journal of Orthopaedic Research 19 (2001) 897-905
Journal of Orthopaedic Research www.elsevier.com/locate/orthres
Muscle adaptation during distraction osteogenesis in skeletally immature and mature rabbits Kazunori Hayatsu, Patrick G. De Deyne * Department of Orthopaedic Surgery, University of Maryland School of Medicine, Maryland Center for Limb Lengthening and Reconstruction, MSTF, Room 400, 10 South Pine Street, Baltimore, M D 21201, USA
Received 11 September 2000; accepted 12 December 2000
Abstract The lack of adaptation of muscle is thought to be a major source of complications during distraction osteogenesis (DO). Although adaptation to DO varies with the regimen (lengthening rate >1 m d d a y and increase in bone length >20%) muscle contractures associated with DO may be a function of age. We tested this idea by subjecting skeletally mature and skeletally immature rabbits to an aggressive regimen of DO (1.4 mdday with a 20% increase in tibia1 length). By using immunofluorescenceto assess the presence of neonatal myosin heavy chain in sections from the tibialis anterior, we observed that the generation of new muscle tissue in response to DO was vigorous in young animals (27% positive fibers), whereas it was more muted in adult animals (9.9y0positive fibers). This adaptive response was associated with a pronounced proliferation of myoblasts in the young but not in the mature animals. Adult tibialis anterior subjected to DO showed a 50% loss in tetanic and twitch tension whereas those in young animals did not. This correlated with partial denervation of adult but not young muscle, as judged by morphological criteria. These experiments indicate that adaptation to DO depends not only on mechanical variables but also on skeletal maturity. 0 2001 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved.
Introduction Contractures in muscles subjected to distraction osteogenesis (DO) are thought to be caused by the lack of adaptation of the muscle 113,291. In animal studies, the adaptation of muscle is compromised if lengthening rates are greater than 1 m d d a y or if the total gain in limb length is more than 20% [33,37], suggesting that mechanical variables influence adaptation to DO. Other variables may also be involved, however. For example, the age of the distracted muscle may influence its adaptive response. Myoblast proliferation has been shown to increase when muscle is stretched [7] and tissue remodeling, especially of the extracellular matrix, occurs during DO [9,37]. Both these responses would be expected to occur more readily in young than in older muscle [3,4]. DO is frequently used to address pathological abnormalities in patients of different ages, but it is not known if muscles in older and younger individuals adapt equally well.
*Corresponding author. Tel.: +1-410-706-2703; fax: +1-410-7060311. E-mail address: [email protected] (P.G. De Deyne).
Here we tested the hypothesis that the adaptive potential of muscle undergoing lengthening by DO is reduced in skeletally mature animals compared to skeletally immature animals. We subjected young and mature rabbits to DO and then measured the effects of distraction on the tibialis anterior (TA) muscle. We assessed physiological parameters such as muscle length and weight, electrophysiological parameters such as maximum tetanic force, and changes in the phenotype of muscle fibers using markers for cell proliferation and the formation of new muscle tissue, as well as parameters for denervation. The distraction rate was set to be greater than normal to maximize the strain to the muscle in both age groups.
Materials and methods Animals
Our experimental model and was adapted from Kojimoto et al. [18] and others [7,33], an aggresive lengthening regimen was choosen based on the work by Simpson et al. [33]. The protocol was approved by the university’s Animal Care and Use Committee. Twelve skeletally immature (young, 3.5 & 0.5 months, 2.3 i0.2 kg) and 12 skeletally mature (adult, 12.7 f 1.8 months, 3.6 f 0.6 kg) male New Zealand White rabbits were studied. Six rabbits from each group served as sham
0736-0266/01/$ - see front matter 0 2001 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 1 ) 0 0 0 0 2 - X
898
K. Hayatsu, P. G. De Deyne / Journal of’ Orthopaedic Research 19 (2001) 897-905
controls, and six from each group underwent bone lengthening (DO) of the right tibia. The left tibia of each rabbit served as a non-treated control. Sham controls consisted of the operated, fractured, but nonlengthened right hind limb. Two distracted tibiae from the young animals were excluded from the study because of surgical complications and an incorrect reduction of the tibia at the end of DO. The Orthofix M-100 unilateral external fixator was applied with four 2-mm selftapping pins to the anteromedial aspect of the right tibia in each rabbit. After circumferential subperiosteal dissection, a transverse middiaphyseal osteotomy was created in the right tibia with a small oscillating saw between the second and third pins. During osteotomy, the TA was protected by retractors and absolute care was taken not to damage the TA. All animals were allowed unrestricted weight bearing immediately after recovering from anesthesia and two days after surgery the animals had an activity pattern that was indistinguishable from non-operated animals. DO was at a lengthening rate of 1.4 mm/ day (in two increments). Electrophysiology
In the anesthetized animal, a monoarticular muscle, the TA, was dissected, and with the neurovascular bundle intact, the peroneal nerve was stimulated supramaximally to induce a tetanic, isometric contraction [8]. Before attaching the most proximal part of the tendon to the load cell, its total length (myotendinous junction to insertion) was recorded. The proximal part of the tendon was attached to a load cell, and the preparation was mounted onto a frame in which the total length of the muscle was controlled. The common peroneal nerve was stimulated with a pulse of 1 ms duration at 100 Hz with a pulse train of 200 ms. Maximum tetanic force was measured by gradually increasing the voltage. The force data were used to determine the maximal tetanic tension and resting length. Single twitch recordings were made at the resting length. Determinat ion of physiological parameters
Muscle length was measured during the experiments with calipers. Lengths measurements were from the tibia1 tuberosity to the myotendinous junction and the position of the recording frame was also noted. The recording frame position and muscle length were directly correlated (mean r = 0.99). The resting length of the TA was obtained from the length-tension curve and corresponded to the muscle length at which peak force was developed.
Tissue preparation The TA muscle was dissected, cut in half, quick-frozen in liquid nitrogen-cooled 2-methyl butane, and stored at -8O”C, as above. At the time of experimentation, one half was used for biochemical assays and the other half was used for morphological studies. Western blot analysis
We used Western blot analysis [12] to assess semi-quantitative changes in muscle proteins. The tissue was pulverized and homogenized in sample buffer [19]. The samples were boiled and centrifuged, and the protein concentration of the supernatant was determined with amido black [12]. Samples were subjected to sodium dodecyl sulfatepolyacrylamide gel electrophoresis [19] and transferred to nitrocellulose electrophoretically. After the filters were incubated with the monoclonal antibodies, they were washed in a buffer and incubated with donkey anti-mouse antibodies conjugated to alkaline phosphatase (Jackson Laboratories, Wesgrove, PA). The bands were visualized by a chemiluminescent assay method (Tropix, Bedford, MA). Iininunofiuorescence
Cryosections (10 pm) were prepared from the frozen TA. For immunolabeling, the unfixed sections were incubated for 30 min in 1 mg/ml bovine serum albumin/phosphate-bufferedsaline, pH 7.2 (BSN PBS). For double labeling with two monoclonal antibodies, the samples were incubated with the first mouse primary antibody (5 pg/ml) for 1 h. After three washes with BSNPBS, the sections were incubated for 1 h with a rabbit anti-mouse Fab fragment (Jackson Laboratories,
Wesgrove, PA) diluted at 1:100. After another three washes in B S N PBS, the sections were incubated with a rhodamine-conjugated goat anti-rabbit antibody (1:lOO) for 30 min. After three washes with BSA/ PBS, the second monoclonal antibody was applied to the section (1--10 pg/ml) for 1 h. After another three washes with BSNPBS, a fluorescein-conjugated goat anti-mouse antibody was applied (1:100, 30 min). Negative controls for this method included the use of a nonimmune monoclonal antibody (MOP-C21; Sigma, St. Louis, MO) and the application of the rhodamine-labeled goat anti-rabbit antibody before the rabbit anti-mouse Fab fragment. After three more washes with BSA/PBS, the sections were mounted in 90% glycerol, 10% 1 M Tris-HCI, pH 8.0, supplemented with 1 mg/ml p-phenylenediamine, and observed through a Zeiss IM 135 microscope or a Zeiss 410 confocal microscope. Images were acquired with a CCD camera. The monoclonal antibodies were against neonatal myosin heavy chain (Novocastra, Newcastle upon Tyne, England), Ki-67, a marker for cell proliferation (Amac, Westbrook, ME), neural cell adhesion molecule (N-CAM), sarcomeric a-actin, and desmin (all from Sigma). Monotetramethylamine-rhodamine-cc-bungarotoxin, a specific marker for acetylcholine receptors, was obtained from Dr. R. Bloch (Department of Physiology, University of Maryland, Baltimore, MD). A8cetylcholinesteraseassay
The activity of acetylcholinesterase was measured with a radiometric assay [15]. Pulverized muscle was extracted with PBS containing 19’0 Triton X-100, 5 mM EGTA, 5 mM EDTA, 10 mM HEPES, 1 M NaC1, pH 7.5. After centrifugation (13,000 g at 4°C for 15 min), the hydrolysis of 1.57 pM3[H]-acetylcholine (30 pCi) was measured (3 h, RT). Butyrylcholine esterase activity was inhibited using tetra-isopropylpyrophosphoramide. The samples were read on a scintillation counter. After nonspecific background was subtracted, the counts were normalized to a positive control (electric eel acetylcholinesterase) and represented as nanomoles of acetylcholine hydrolyzedmidmg of protein. Analysis and statistics
As the data from the sham control group were not different from the data from the non-lengthened contralateral sides of the age-matched controls, we show only the data from the sham control group. All samples that were labeled by indirect immunofluorescence were randomized and scored in a blind fashion for the presence of the molecule ol‘ interest. Data from a pilot study indicated that a sample size of five was sufficient to obtain statistically reliable differences ( P < 0.05). A minimum of 100 muscle fibers were counted per section, and the data were expressed as the number of positive labeled fibers as a percent of total. To count the cells that were double labeled with antibodies against desmin and Ki-67, digital pictures were taken in a randomized fashion and the number of labeled cells were counted per area (10.4 x lo3 pmZ, or 512 x 512 pixels). Image analysis of the Western blots was performed from the integrated signal from line scans of the blots using Kodak ID‘” gel imaging software. The data from three independent experiments were expressed as YOof controls. Analysis of variance and appropriate post hoc tests (Tukey, Scheffi) were used to analyze all the data. Data sets that differed with a probability of P < 0.05 were considered to be significantly different. The figures depicting the separation of proteins by electrophoresis represent a typical experiment. All biochemical and immunological experiments were reproduced at least three times, and the figures represent typical results.
Results
To assess the skeletal maturity of the animals, radiographs were obtained of the left and right tibiae. The radiographs allowed us to visually assess reduction, pin placement, and, at the end of distraction, degree of bone remodeling. Fig. 1 shows typical radiographs of a young skeletally immature and an adult skeletally mature rabbit. The arrows (Fig. l(A)) point to the growth plates
K. Hayatsu, P.G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
899
Fig. 1. Radiographs of distracted tibiae. The radiographs show the amount of distraction in the young (A) and adult (B) rabbits. Arrows in A point to the growth plates.
that are still open in the tibia of the young rabbit. The radiographs were also used to measure the amount of lengthening and the length of the tibia. No additional bone growth or inhibition was observed in the tibiae of the sham operated or control young rabbits (data not shown). In skeletally immature rabbits subjected to DO, the tibiae increased in length from 96.3 f 5.2 mm to 119.5 f 5.9 mm, a 23 f 1% increase. In skeletally mature rabbits subjected to the identical DO regimen (rate and increase in length), the response was similar, the tibiae increased from 110 f 1.6 mm to 133 f 2.9 mm, a 21 f 3% increase, as determined by radiography. Thus, DO caused identical changes in relative lengths of the limbs in skeletally mature and skeletally immature rabbits. We used indirect immunofluorescence to detect changes in the phenotype of muscle fibers, as defined by the expression of neonatal myosin heavy chain [8]. We expected to see some level of expression of neonatal myosin in the muscle from control young rabbits (Fig. 2(A)), as reported earlier [I 13. Antibodies against neonatal myosin heavy chain detected positive labeling in the lengthened TA of both young and adult rabbits (Fig. 2(C) and (D)), consistent with the idea that the generation of new sarcomeres contribute to the adaptive response in lengthened muscle. A vigorous expression was noted in the muscle from lengthened tibiae of young rabbits: many fibers in these muscles showed strong (Fig. 2(C), avvow) or weaker labeling (Fig. 2(C), arrowhead). A less pronounced but still significant response was observed in the TA from adult rabbits (Fig. 2(D)). We quantitated the number of labeled fibers, compared to all fibers in the muscle. In young control rabbits, 4% f 3 (mean k S.D.) of the fibers contained neonatal myosin heavy chain, and in adult control rabbits, 1 f 0.2% of the fibers contained neonatal myosin heavy chain. During DO, the number of fibers expressing the
Muscle Fibers Containing Neonatal MHC as a % of Total
E
T. 20
10
0 YSham
YDO
Adsham
Ad Do
Distraction Osteogenesis Treatment (Mean+S.E., n.6)
Fig. 2. Immunofluorescent labeling of the TA. Antibodies against neonatal myosin heavy chain were used to visualize the regenerative response. Occasional myofibers are labled in the TA from young shamoperated rabbits (A). Such labeling was not seen in muscle fibers from adult rabbits (B). DO induces the expression of neonatal myosin heavy chain in muscle fibers in both young and mature rabbits. The labeled myofibers are the same size as the surrounding muscle fibers, but a more robust response is seen in young (C) than in adult (D) muscle. The percent of total muscle fibers containing neonatal myosin heavy chain is shown in E.
neonatal form of myosin heavy chain increased to 2 7 . 6 f 7.6% in TA from young rabbits and to 9.9 f 0.7% in TA of adult rabbits ( P < 0.01, n = 4). The data (Fig. 2(E)) show that the adaptive response, as measured by the amount of muscle fibers expressing neonatal myosin heavy chain [I 6,3 11, is considerably more robust in skeletally immature rabbits. This difference is not associated with changes in the distraction variables, because the distraction rate and total amount of lengthening were the same and our results suggest that skeletal maturity influences the adaptive response.
K. Hayatsu. P. G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
900
To compare the extent of cell proliferation in the muscle of the two age groups, we used a marker for cell proliferation (Ki-67), a nuclear antigen that is present in all phases of the mitotic cycle but that is absent in quiescent cells (Go). Desmin, a muscle-specific intermediate filament protein, was used as a marker for muscle fibers, myotubes, and myoblasts. Western blot analysis (Fig. 3) allowed a semiquantitative analysis of Ki-67. Relative to the amount of all fast twitch myosin heavy chain (internal control; Fig. 3, Fast-MHC, bottom panel), there is a significant (P< 0.01, n = 3) increase in the amount of desmin and Ki-67 in the TA subjected to DO compared to sham controls (densitometric analysis is represented on the right). These data suggest that DO leads to an important increase in the amount of desmin in both age groups that is indicative of myogenesis and that is paralleled by an increase in a marker for cell proliferation.
To localize this proliferative response, we performed double labeling immunofluorescence with antibodies to desmin and Ki-67 (Fig. 4). The data showed numerous proliferating cells (Fig. 4(F), arrows) in the muscle from distracted tibiae of young rabbits and not in the sham controls (Fig. 4(E) and (G)). Proliferating cells were observed less frequently in the muscle from distracted tibiae from adult rabbits, and if seen they were sparsely distributed along the muscle fiber (Fig. 4(H)). Double labeling with antibodies against desmin allowed us to distinguish between proliferating myoblasts and non-muscle cells. We noted that in the young muscle subjected to DO most proliferating cells were desminpositive (Fig. 4(J)), suggesting that they are myoblasts or satellite cells. In the adult muscle subjected to DO, the Ki-67-positive cells were rarely labeled with antibodies against desmin (Fig. 4(L)). By counting the cells that
0
A
kD
10-67
Ki67
203 Desmin
i
r *
I
48.7
-
Fast- MHC
203 -
Fig. 3. Western blot analysis of muscle homogenates. Equal amounts (50 pg of protein) of the muscle homogenates (see Materials and Methods) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose. Antibodies against markers for proliferating cells (Ki-67), muscle (desmin), and an internal control (fast myosin heavy chain) were used to visualize and quantitate the retative amounts of the molecules. The data are representative of three independent experiments and suggest a relative increase of desmin and Ki-67 during DO. Densitometric analysis is shown on the right.
K. Hayatsu, P.G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
90 1
Fig. 4.Double immunofluorescent labeling of the TA. Frozen sections of muscle were double labeled with antibodies against desmin (A-D) and Ki67 (E-H). Image overlays (I-L) allow the visualization of cells containing both molecules (J, e.g. arrows). DO leads to numerous cells that were positively labeled with antibodies against desmin and Ki-67 (B, F, and J, arrows). In the TA from adult rabbits, DO leads to a marginal increase in proliferating cells, that in most cases were not labeled with antibodies to desmin (L).
were double labeled with antibodies against desmin and Ki-67, we found in the experimental TA from young animals 0.58 f0.24/103 pm2 Ki-67 positive cells (n = 5 ) , of which half were also labeled with antibodies against desmin and were localized appropriately. We also observed several proliferating cells in the distal region of the distracted muscle from young animals that were not labeled with antibodies against desmin (data not shown), consistent with remodeling of the connective tissue. In the experimental TA from adult animals we only counted 0.08 f 0.05/103 pm2 Ki-67 positive cells (n = 4) and these proliferating cells were rarely positive for desmin. These results suggest that DO-induced proliferation of myogenic cells only occurs to a significant extent in young muscle. We performed electrophysiological experiments to investigate the functional changes that occur in muscle subjected to DO (Table 1). Our major finding is that the TA in young rabbits adapted to DO without functional deficits and that muscle strength was not affected. By
contrast, the TA in adult rabbits adapted to DO poorly, and muscle weakness was significant. The maximum tetanic tension in single twitch tension was less than 50% that of the controls (Table 1). In the adult rabbits the wet weight of the muscle was not statistically different between the sham controls and the experimental group but in the young rabbits a significant increase in wet weight was noted ( P = 0.035, P = 0.528) in the lengthened TA. As expected, the resting length increased during DO, in both young and mature muscles, in agreement with an earlier report [8]. To examine possible denervation associated with DO, we studied the neuromuscular junction of control and DO muscle using morphological techniques (Fig. 5). The acetylcholine receptors at the synapse are associated with N-CAM, which is lost following denervation [5,6,26]. In control samples, the morphology of the neuromuscular junction was normal (Fig. 5(I) and (K)), as it was in sections from distracted muscle from young rabbits (Fig. 5(J), arrow). In samples from the distracted
902
K. Hayatsu, P.G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
Table 1 Physiological data ~
Electro-physiological
Young
Adult
parameters
Sham (n = 5, mean fSEM)
DO (n = 4 , mean fSEM)
Sham (n = 6, mean fSEM)
DO (n = 6, mean fSEM)
Pmax (N) Twitch tension (N) Wet weight (g) Resting length (cm)
12.59 f 1.90 2.18 10.71 2.06 f0.07 7.34 f0.1
12.17 f0.57 2.73 f0.59 2.21 10.12' 8.4 f0 2 '
19.30f 1.69 4.62 f 0.91 3.77f0.19 8.07 & 0.22
7.36i2.31' 1.18 f0.37' 3.82 +0.15 9.52f0.14'
~
< 0.05. P = 0.075.
*P
**
Fig. 5. N-CAM at the neuromuscular junction in the TA. Frozen sections of muscle were double labeled with rhodamine-conjugated bungarotoxin (A-D) and antibodies against N-CAM (E-H). Image overlay (I-L) allows visualization of both molecules at the neuromuscular junction (J, arrow). In the TA from young rabbits, the DO does not lead to a loss of N-CAM associated with the acetylcholine receptor (B, F, and J, arrows). In the TA from adult rabbits, N-CAM is lost from the neuromuscular junction (D, H, and L, arrow), suggestive of denervation. An increase in intracellular labeling is also observed (H and L, asterisks).
TA of adult rabbits, however, DO led to the loss of NCAM (Fig. 5(H) and (L), urrows) from the synaptic regions (Fig. 5(D) and (L), arrows). In addition, we noted that a majority of the muscle fibers had intracellular labeling for N-CAM (Fig. 5(L), asterisks). Thus our results are consistent with the idea that DO leads to denervation of at least some of the muscle fibers of the
TA. We tried to confirm these findings by quantifying the activity of acetylcholinesterase, which, in rabbits, increases upon of denervation [20]. However, no significant changes in enzyme activity were observed in DO muscle from either young or adult rabbits ( P = 0.3) despite the observed trend that is consistent with denervation (Table 2). The combined results from the
Table 2 Acetylcholine esterase assay Treatment
nmoles Ach hydrolyzed min-’ nig protein-’ (mean SEM)
Young control Young Sliani Young DO Adult control Adult Sham Adult DO
3.5.56 7.73 4 1 . 2 1 10.48 48.96i12.9.5 41.932~17.2 34.05 2C 13.6 53.38 10.9
*
*
*
functional, morphological, and biochemical experiments suggest that a partial denervation occurs in muscle in skeletally mature rabbits subjected to DO.
Discussion
A con~plicationof DO, the development of contractures, has been interpreted as resulting from an insufficient adaptive response of skeletal muscle to the DO regimen. A number of investigators [21,32,33,37] have observed increased complications with a lengthening rate of more than 1 mm/day or with an increase in limb length of more than 20‘1/0. As skeletal maturity may also influence the successful outcome of DO and because DO is performed both in children and adults, we wanted to know whether the adaptation of muscle to DO could be a function of skeletal maturity. Adaptation of muscle subjected to D O might be expected to be more complete in a skeletally immature individual because their higher circulating levels of growth hormone in may support more cell proliferation and tissue growth [lo], and because younger tissues are in general considered t o be more “plastic” [24]. Our results confirm that muscle subjected to DO in skeletally immature rabbits experiences an increased proliferation of myoblasts or satellite cells while maintaining contractile function. Despite the use of a DO regimen that was likely to lead to the development of fibrosis and other complications [33,37], the muscle from the skeletally immature rabbits adapted without developing measurable weakness 121,321. The proliferation of myogenic cells during D O has been reported before [7] and it is quite possible that during D O the adaptive process of muscle, i.e. the proliferation of myogenic cells [7] and even the addition of sarcomeres [33], may depend on bioactive peptides. In growing animals the circulating levels of growth hormone may create an environment in which the adaptation of muscle to DO is facilitated. This process may be mediated through insulin-like growth factor-1 [lo], an important trophic factor for muscle 121, the synthesis of which is in part controlled by growth hormone [lo]. In contrast, the TA from the skeletally mature rabbits showed an adaptive response that was significantly di-
minished and that resulted in severe muscle weakness. It is well known that regeneration in muscle decreases with age [3,4,24]. In models of muscle injury, regeneration is preceded by degenerative events such as inflammatory reactions and the remodeling of the extracellular matrix [22,35,36]. We do not know to what extent these events occur in muscle during D O or whether they hamper the adaptive response of muscle in skeletally mature animals. Nevertheless skeletally mature animals do show an adaptive process, as indicated by the re-expression of neonatal myosin heavy chain in the skeletally mature animals. Howcver our western blots (Fig. 3) and our indirect immunofluorescence experiments (Fig. 4) suggest that this upregulation of neonatal myosin in skeletally mature animals occurs in the absence of significant proliferation of myogenic cells. Proliferation of nonmuscle cells may contribute to additional complications associated with DO such fibrosis [33,37]. Our morphological and biochcniical assays further suggested that DO causes at least partial denervation of the TA in skeletally mature rabbits. This was most clearly indicated by the loss of N-CAM from the neuroniuscular junction of older animals and was consistent with a trend towards higher activity of acetylcholine esterase in muscles subjected to DO. This conclusion is consistent with clinical observations of nerve damage that occurred with DO in achondroplastic dwarfs [30]. In children, a transient denervation-induced muscle weakness of the quadriceps has also been noted, even with small amounts of femoral lengthening (2 cm) (281 and long-term weakness of the quadriceps muscle in children has been observed in femoral lengthening averaging 7 cm [23]. As our results are only suggestive, however, further studies of denervation associated with DO of young and older muscle will be needed. That the adaptation of muscle in skeletally immature rabbits is more robust than in adults is not surprising. In young animals, growth hormone influences the growth and maturation of a number of skeletal tissues [14,34] and some investigators have proposed that increased systemic levels of growth hormone can stimulate a local growth response in muscle [ I ,25,38]. In addition, growth hormone regulates the synthesis of insulin-like growth factor I (IGF-1) [lo], which has been implicated as promotor of muscle proliferation and maturation [2,17,27]. Consistent with the idea that it plays a role in the adaptive response of young muscle to DO, published experiments indicate that IGF-1 acts by a local, paracrine mechanism, especially when the muscle is exposed to stretch [39]. In summary, our report shows that young muscle adapts to DO better than adult muscle when similar distraction variables are applied. Our findings show a vigorous adaptive response in muscle from skeletally immature rabbits without any functional deficits. In contrast, a more modest adaptive response, muscle
K. Huyntsu, P. G. Dr Drj.ne I Journul of' Orthoprirdic Resrurch 15, (2001) 897-905
904
weakness, and a partial denervation are seen in muscle from skeletally mature rabbits. One goal of our future research is to identify therapeutic interventions to prevent these complications in adult animals.
Acknowledgements We would like to acknowledge the support of the National Center for Medical Rehabilitation Research at the NIH (KOIHDO1165), Maryland Center for Limb Lengthening and Reconstruction, and the Association for the Study and the Advancement of the Methods by Ilizarov. We would also like to thank Rick Meyer for technical assistance, Dori Kelly for editorial help, Drs. Dror Paley and John Herzenberg for the helpful discussions and especially Dr. Robert Bloch for a critical reading of the manuscript.
References Aroniadou-Anderjaska V. Lemon PW. Gilloteaux J . Effects of exogenous growth hormone on skeletal muscle of young female rats. Tissue Cell 1996;28:719 24. Barton-Davis ER. Shoturina DI. Musaro A, Rosenthal N. Swecney HL. Viral mediated expression of insulin-like growth factor 1 blocks the aging-related loss of skeletal muscle function. Proc Natl Acad Sci USA 1998;95:15603-7. Cannon JG. Intrinsic and extrinsic factors in muscle aging. Ann NY Acad Sci 1998:854:72-7. Carlson BM. Muscle regeneration and aging. Monographs Develop Biol 1992:23:189 95. Cashman NR, Covault J. Wollman RL, Sanes JR. Neural cell adhesion molecule in normal, denervated. and myopathic h u m a n muscle. Ann Neurol 1987:21:481-9. Covault J, Sanes JR. Neural cell adhesion molecule (N-CAM) accumulates in denervated and paralyzed skeletal muscles. Proc Natl Acad Sci USA 1985:82:4544-8. Day CS. Moreland MS, Floyd SS, Huard J. Limb lengthening promotes muscle growth. J Orthop Res 1997;15:227-34. De Deyne PG, Hayatsu K. Meyer R, Paley D, Herzenberg JE. Muscle regeneration and fiber type transformation during distraction osteogenesis. J Orthop Res 1999;17:560-70. De Deyne PG, Meyer R, Paley D, Herzenberg JE. The adaptation of perimuscular connective tissue during distraction osteogenesis. Clin Orthop 2000;379:259--69. Florini JR, Ewton DZ, Coolican SA. Growth hormone and the insulin-like growth factor system in myogenesis. Endocr Rev 1996: I7:48 1-5 I7. Gondret F. Lefaucheur JP, D'Albis A, Bonneau M . Myosin isoform transitions in four rabbit muscle during postnatal growth. J Muscle Res Cell Mot 1996;17:657-67. Harlow E. Lane D. Immiinoblotting. In: Harlow E. Lane D. editors. Antibodies: a laboratory manual. Cold Spring Harbor Laboratory: 1988. p. 471-504. Herzenberg JE, Scheufele LL, Paley D, Bechtel R. Tepper S. Knee range of motion in isolated femoral lengthening. Clin Orthop I994:301:49--54. Isaksson OG, Ohlsson C, Nilsson A, Isgaard J. Lindahl A. Regulation of cartilage growth by growth hormone and insulinlike growth factor I. Pediatr Nephrol 1991;5:451-3.
Johnson CD, Russell RL. A rapid, simple radiometric gassay for cholinesterase. suitable for multiple determinations. Anal Biochem 1975:64:229 38. Karsch~Mizrdchi I, Travis M. Blau H , Leinwand LA. Expression and DNA sequence analysis o f a human embryonic skeletal muscle myosin heavy chain gene. Nuc Ac Res 1989;17:6167-79. Keller HL, St. Pierre S, Eppihimer LA, Cannon JG. Association of IGF-I and IGF-I1 with myofiber regeneration in vivo. Muscle Nerve 1999:22:347-54. Kojimoto H, Yasui N. Goto T, Matsuda S, Shimomura Y. Bone lengthening in rabbits by callus distraction. The role of periosteiini and endosteum. J Bone J Surg [Br] 1988:70:543-9. Laemmli U K . Cleavage Of StruCturd proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680-5. Lai J, Jedrzejczyk J, PizLey JA, Green D , Barnard EA. Neural control of the forms of acetylcholinestera~ein slow mammalian muscles. Nature 1986:321:724. Lee DY, Choi IH. Chung CY, Chung PH, Chi JG, Suh YL. Effect of tibia1 lengthening on the gastrocnemius muscle. A histopathologic and inorphometric study in rabbits. Acta Orthop Scand 1993;64:688 92. Lef'aucheur JP, Gjata B. Lafont H , Sebille A. Angiogenic and inflammatory responses following skeletal muscle injury are altered by immune neutralization of endogenous basic fibroblast growth factor, insulin-like growth factor- 1 and transforming growth factor-beta 1. J NeuroiinmLinol 1996:70:3744. Maffulli N , Fixsen JA. Muscular strength after callotasis limb lengthening. J Ped Orthop 1995;15:212-6. Marsh DR, Criswell DS, Carson JA, Booth FW. Myogenic regulatory factors during regeneration of skeletal muscle in young. adult, and old rats. J Appl Physiol 1997:83:1270-5. McCall GE, Goulet C, Roy RR, Grindeland RE, Boorman GI, Bigbee AJ. Hodgson JA, Greenisen MC, Edgerton VR. Spaceflight suppresses exercise-induced release of bioassayable growth hormone. J Appi Physiol 1999:87: 1207-12. Moscoso LM. Merlie JP, Sanes J R . N-CAM, 43K-rapsyn, and Slaminin mRNAs are concentrated at synaptic sites in muscle fibers. Mol Cell Neurosci 1995;6:80-9. Musaro A. Roseiithal N. Maturation of the inyogenic program is induced by postmitotic expression of insulin-like growth factor I. Mol Cell Biol 1999; 19:31 15-24. Oey PL, Engelbert RH. van Roermond PM, Weineke G H . Temporary muscle weakness in the early phase of distraction during femoral lengthening. Clinical and electr.omyogl-apliical observations. Electromyogr Clin Neurophysiol 1999;39:2 17-20. Paley D. Problems. obstacles, and complications of limb lengthening by the Ilizarov technique. Clin Orthop 1990:81-104. Polo A, Zambito A, Aldegheri R, Tinazzi M, Rizzuto N. Nerve conduction changes during lower limb lengthening. Somatosensory evoked potentials (SEPs) and F-wave results. Electromyogr Clin Neurophysiol 1999;39:13944. Sartore S, Gorza L, Schiafino S. Fetal myosin heavy chains in regenerating muscle. Nature 1982;298:294-6. Schumacher B, Keller J, Hvid I. Distraction effects on muscle. Leg lengthening studied in rabbits. Acta Orthop Scand 1994;65:647-
so. Simpson AH, Williams PE, Kyberd P, Goldspink G, Kenwright J. The response of muscle to leg lengthening. J Bone J Surg [Br] 1995:77:630-6. Spagnoli A, Rosenfeld R G . The mechanisins by which growth hormone brings about growth. The relative contributions of growth hormone and insulin-like growth factors. Endocrinol Metab Clin North Am 1996;25:615-31. Stauber WT, Fritz VK, Dahlmann B. Extracellular matrix changes following blunt trauma to rat skeletal muscles. Exp Mol Pathol 1990;52:69-86.
K. Huyutsu, P.G. DP Deqne I Joitnlul of Ortkop~~edic Research 19 (2001) 897-905
[36] Tidball JG, St. Pierre BA. Apoptosis of macrophages during the resulution of muscle inflammation. J Leukoc Biol 199639: 380-8. [37] Williams PE, Kyberd P, Simpson AH, Kenwrighl J, Goldspink G. The morphological basis of increased stiKness of rabbit tibialis anterior muscles during surgical limb-lengthening. J Anat l998;l 93:131-8.
905
[38] Wilson VJ, Rattray M, Thomas CR. Moreland BH. Schulster D. Growth hormone increases IGF-I, collagen I and collagen I11 gene expression in dwarf rat skeletal muscle. Mol Cell Endocrinol 1995;I 15:187-97. [39] Yang H, Alnaqeeb M, Simpson AH, Goldspink G. Changes in muscle fibre type, muscle mass and IGF-I gene expression in rabbit skeletal muscle subjected to stretch. J Aiiat 1997:190:613 22.
Journal of Orthopaedic Research 19 (2001) 897-905
Journal of Orthopaedic Research www.elsevier.com/locate/orthres
Muscle adaptation during distraction osteogenesis in skeletally immature and mature rabbits Kazunori Hayatsu, Patrick G. De Deyne * Department of Orthopaedic Surgery, University of Maryland School of Medicine, Maryland Center for Limb Lengthening and Reconstruction, MSTF, Room 400, 10 South Pine Street, Baltimore, M D 21201, USA
Received 11 September 2000; accepted 12 December 2000
Abstract The lack of adaptation of muscle is thought to be a major source of complications during distraction osteogenesis (DO). Although adaptation to DO varies with the regimen (lengthening rate >1 m d d a y and increase in bone length >20%) muscle contractures associated with DO may be a function of age. We tested this idea by subjecting skeletally mature and skeletally immature rabbits to an aggressive regimen of DO (1.4 mdday with a 20% increase in tibia1 length). By using immunofluorescenceto assess the presence of neonatal myosin heavy chain in sections from the tibialis anterior, we observed that the generation of new muscle tissue in response to DO was vigorous in young animals (27% positive fibers), whereas it was more muted in adult animals (9.9y0positive fibers). This adaptive response was associated with a pronounced proliferation of myoblasts in the young but not in the mature animals. Adult tibialis anterior subjected to DO showed a 50% loss in tetanic and twitch tension whereas those in young animals did not. This correlated with partial denervation of adult but not young muscle, as judged by morphological criteria. These experiments indicate that adaptation to DO depends not only on mechanical variables but also on skeletal maturity. 0 2001 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved.
Introduction Contractures in muscles subjected to distraction osteogenesis (DO) are thought to be caused by the lack of adaptation of the muscle 113,291. In animal studies, the adaptation of muscle is compromised if lengthening rates are greater than 1 m d d a y or if the total gain in limb length is more than 20% [33,37], suggesting that mechanical variables influence adaptation to DO. Other variables may also be involved, however. For example, the age of the distracted muscle may influence its adaptive response. Myoblast proliferation has been shown to increase when muscle is stretched [7] and tissue remodeling, especially of the extracellular matrix, occurs during DO [9,37]. Both these responses would be expected to occur more readily in young than in older muscle [3,4]. DO is frequently used to address pathological abnormalities in patients of different ages, but it is not known if muscles in older and younger individuals adapt equally well.
*Corresponding author. Tel.: +1-410-706-2703; fax: +1-410-7060311. E-mail address: [email protected] (P.G. De Deyne).
Here we tested the hypothesis that the adaptive potential of muscle undergoing lengthening by DO is reduced in skeletally mature animals compared to skeletally immature animals. We subjected young and mature rabbits to DO and then measured the effects of distraction on the tibialis anterior (TA) muscle. We assessed physiological parameters such as muscle length and weight, electrophysiological parameters such as maximum tetanic force, and changes in the phenotype of muscle fibers using markers for cell proliferation and the formation of new muscle tissue, as well as parameters for denervation. The distraction rate was set to be greater than normal to maximize the strain to the muscle in both age groups.
Materials and methods Animals
Our experimental model and was adapted from Kojimoto et al. [18] and others [7,33], an aggresive lengthening regimen was choosen based on the work by Simpson et al. [33]. The protocol was approved by the university’s Animal Care and Use Committee. Twelve skeletally immature (young, 3.5 & 0.5 months, 2.3 i0.2 kg) and 12 skeletally mature (adult, 12.7 f 1.8 months, 3.6 f 0.6 kg) male New Zealand White rabbits were studied. Six rabbits from each group served as sham
0736-0266/01/$ - see front matter 0 2001 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 0 2 6 6 ( 0 1 ) 0 0 0 0 2 - X
898
K. Hayatsu, P. G. De Deyne / Journal of’ Orthopaedic Research 19 (2001) 897-905
controls, and six from each group underwent bone lengthening (DO) of the right tibia. The left tibia of each rabbit served as a non-treated control. Sham controls consisted of the operated, fractured, but nonlengthened right hind limb. Two distracted tibiae from the young animals were excluded from the study because of surgical complications and an incorrect reduction of the tibia at the end of DO. The Orthofix M-100 unilateral external fixator was applied with four 2-mm selftapping pins to the anteromedial aspect of the right tibia in each rabbit. After circumferential subperiosteal dissection, a transverse middiaphyseal osteotomy was created in the right tibia with a small oscillating saw between the second and third pins. During osteotomy, the TA was protected by retractors and absolute care was taken not to damage the TA. All animals were allowed unrestricted weight bearing immediately after recovering from anesthesia and two days after surgery the animals had an activity pattern that was indistinguishable from non-operated animals. DO was at a lengthening rate of 1.4 mm/ day (in two increments). Electrophysiology
In the anesthetized animal, a monoarticular muscle, the TA, was dissected, and with the neurovascular bundle intact, the peroneal nerve was stimulated supramaximally to induce a tetanic, isometric contraction [8]. Before attaching the most proximal part of the tendon to the load cell, its total length (myotendinous junction to insertion) was recorded. The proximal part of the tendon was attached to a load cell, and the preparation was mounted onto a frame in which the total length of the muscle was controlled. The common peroneal nerve was stimulated with a pulse of 1 ms duration at 100 Hz with a pulse train of 200 ms. Maximum tetanic force was measured by gradually increasing the voltage. The force data were used to determine the maximal tetanic tension and resting length. Single twitch recordings were made at the resting length. Determinat ion of physiological parameters
Muscle length was measured during the experiments with calipers. Lengths measurements were from the tibia1 tuberosity to the myotendinous junction and the position of the recording frame was also noted. The recording frame position and muscle length were directly correlated (mean r = 0.99). The resting length of the TA was obtained from the length-tension curve and corresponded to the muscle length at which peak force was developed.
Tissue preparation The TA muscle was dissected, cut in half, quick-frozen in liquid nitrogen-cooled 2-methyl butane, and stored at -8O”C, as above. At the time of experimentation, one half was used for biochemical assays and the other half was used for morphological studies. Western blot analysis
We used Western blot analysis [12] to assess semi-quantitative changes in muscle proteins. The tissue was pulverized and homogenized in sample buffer [19]. The samples were boiled and centrifuged, and the protein concentration of the supernatant was determined with amido black [12]. Samples were subjected to sodium dodecyl sulfatepolyacrylamide gel electrophoresis [19] and transferred to nitrocellulose electrophoretically. After the filters were incubated with the monoclonal antibodies, they were washed in a buffer and incubated with donkey anti-mouse antibodies conjugated to alkaline phosphatase (Jackson Laboratories, Wesgrove, PA). The bands were visualized by a chemiluminescent assay method (Tropix, Bedford, MA). Iininunofiuorescence
Cryosections (10 pm) were prepared from the frozen TA. For immunolabeling, the unfixed sections were incubated for 30 min in 1 mg/ml bovine serum albumin/phosphate-bufferedsaline, pH 7.2 (BSN PBS). For double labeling with two monoclonal antibodies, the samples were incubated with the first mouse primary antibody (5 pg/ml) for 1 h. After three washes with BSNPBS, the sections were incubated for 1 h with a rabbit anti-mouse Fab fragment (Jackson Laboratories,
Wesgrove, PA) diluted at 1:100. After another three washes in B S N PBS, the sections were incubated with a rhodamine-conjugated goat anti-rabbit antibody (1:lOO) for 30 min. After three washes with BSA/ PBS, the second monoclonal antibody was applied to the section (1--10 pg/ml) for 1 h. After another three washes with BSNPBS, a fluorescein-conjugated goat anti-mouse antibody was applied (1:100, 30 min). Negative controls for this method included the use of a nonimmune monoclonal antibody (MOP-C21; Sigma, St. Louis, MO) and the application of the rhodamine-labeled goat anti-rabbit antibody before the rabbit anti-mouse Fab fragment. After three more washes with BSA/PBS, the sections were mounted in 90% glycerol, 10% 1 M Tris-HCI, pH 8.0, supplemented with 1 mg/ml p-phenylenediamine, and observed through a Zeiss IM 135 microscope or a Zeiss 410 confocal microscope. Images were acquired with a CCD camera. The monoclonal antibodies were against neonatal myosin heavy chain (Novocastra, Newcastle upon Tyne, England), Ki-67, a marker for cell proliferation (Amac, Westbrook, ME), neural cell adhesion molecule (N-CAM), sarcomeric a-actin, and desmin (all from Sigma). Monotetramethylamine-rhodamine-cc-bungarotoxin, a specific marker for acetylcholine receptors, was obtained from Dr. R. Bloch (Department of Physiology, University of Maryland, Baltimore, MD). A8cetylcholinesteraseassay
The activity of acetylcholinesterase was measured with a radiometric assay [15]. Pulverized muscle was extracted with PBS containing 19’0 Triton X-100, 5 mM EGTA, 5 mM EDTA, 10 mM HEPES, 1 M NaC1, pH 7.5. After centrifugation (13,000 g at 4°C for 15 min), the hydrolysis of 1.57 pM3[H]-acetylcholine (30 pCi) was measured (3 h, RT). Butyrylcholine esterase activity was inhibited using tetra-isopropylpyrophosphoramide. The samples were read on a scintillation counter. After nonspecific background was subtracted, the counts were normalized to a positive control (electric eel acetylcholinesterase) and represented as nanomoles of acetylcholine hydrolyzedmidmg of protein. Analysis and statistics
As the data from the sham control group were not different from the data from the non-lengthened contralateral sides of the age-matched controls, we show only the data from the sham control group. All samples that were labeled by indirect immunofluorescence were randomized and scored in a blind fashion for the presence of the molecule ol‘ interest. Data from a pilot study indicated that a sample size of five was sufficient to obtain statistically reliable differences ( P < 0.05). A minimum of 100 muscle fibers were counted per section, and the data were expressed as the number of positive labeled fibers as a percent of total. To count the cells that were double labeled with antibodies against desmin and Ki-67, digital pictures were taken in a randomized fashion and the number of labeled cells were counted per area (10.4 x lo3 pmZ, or 512 x 512 pixels). Image analysis of the Western blots was performed from the integrated signal from line scans of the blots using Kodak ID‘” gel imaging software. The data from three independent experiments were expressed as YOof controls. Analysis of variance and appropriate post hoc tests (Tukey, Scheffi) were used to analyze all the data. Data sets that differed with a probability of P < 0.05 were considered to be significantly different. The figures depicting the separation of proteins by electrophoresis represent a typical experiment. All biochemical and immunological experiments were reproduced at least three times, and the figures represent typical results.
Results
To assess the skeletal maturity of the animals, radiographs were obtained of the left and right tibiae. The radiographs allowed us to visually assess reduction, pin placement, and, at the end of distraction, degree of bone remodeling. Fig. 1 shows typical radiographs of a young skeletally immature and an adult skeletally mature rabbit. The arrows (Fig. l(A)) point to the growth plates
K. Hayatsu, P.G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
899
Fig. 1. Radiographs of distracted tibiae. The radiographs show the amount of distraction in the young (A) and adult (B) rabbits. Arrows in A point to the growth plates.
that are still open in the tibia of the young rabbit. The radiographs were also used to measure the amount of lengthening and the length of the tibia. No additional bone growth or inhibition was observed in the tibiae of the sham operated or control young rabbits (data not shown). In skeletally immature rabbits subjected to DO, the tibiae increased in length from 96.3 f 5.2 mm to 119.5 f 5.9 mm, a 23 f 1% increase. In skeletally mature rabbits subjected to the identical DO regimen (rate and increase in length), the response was similar, the tibiae increased from 110 f 1.6 mm to 133 f 2.9 mm, a 21 f 3% increase, as determined by radiography. Thus, DO caused identical changes in relative lengths of the limbs in skeletally mature and skeletally immature rabbits. We used indirect immunofluorescence to detect changes in the phenotype of muscle fibers, as defined by the expression of neonatal myosin heavy chain [8]. We expected to see some level of expression of neonatal myosin in the muscle from control young rabbits (Fig. 2(A)), as reported earlier [I 13. Antibodies against neonatal myosin heavy chain detected positive labeling in the lengthened TA of both young and adult rabbits (Fig. 2(C) and (D)), consistent with the idea that the generation of new sarcomeres contribute to the adaptive response in lengthened muscle. A vigorous expression was noted in the muscle from lengthened tibiae of young rabbits: many fibers in these muscles showed strong (Fig. 2(C), avvow) or weaker labeling (Fig. 2(C), arrowhead). A less pronounced but still significant response was observed in the TA from adult rabbits (Fig. 2(D)). We quantitated the number of labeled fibers, compared to all fibers in the muscle. In young control rabbits, 4% f 3 (mean k S.D.) of the fibers contained neonatal myosin heavy chain, and in adult control rabbits, 1 f 0.2% of the fibers contained neonatal myosin heavy chain. During DO, the number of fibers expressing the
Muscle Fibers Containing Neonatal MHC as a % of Total
E
T. 20
10
0 YSham
YDO
Adsham
Ad Do
Distraction Osteogenesis Treatment (Mean+S.E., n.6)
Fig. 2. Immunofluorescent labeling of the TA. Antibodies against neonatal myosin heavy chain were used to visualize the regenerative response. Occasional myofibers are labled in the TA from young shamoperated rabbits (A). Such labeling was not seen in muscle fibers from adult rabbits (B). DO induces the expression of neonatal myosin heavy chain in muscle fibers in both young and mature rabbits. The labeled myofibers are the same size as the surrounding muscle fibers, but a more robust response is seen in young (C) than in adult (D) muscle. The percent of total muscle fibers containing neonatal myosin heavy chain is shown in E.
neonatal form of myosin heavy chain increased to 2 7 . 6 f 7.6% in TA from young rabbits and to 9.9 f 0.7% in TA of adult rabbits ( P < 0.01, n = 4). The data (Fig. 2(E)) show that the adaptive response, as measured by the amount of muscle fibers expressing neonatal myosin heavy chain [I 6,3 11, is considerably more robust in skeletally immature rabbits. This difference is not associated with changes in the distraction variables, because the distraction rate and total amount of lengthening were the same and our results suggest that skeletal maturity influences the adaptive response.
K. Hayatsu. P. G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
900
To compare the extent of cell proliferation in the muscle of the two age groups, we used a marker for cell proliferation (Ki-67), a nuclear antigen that is present in all phases of the mitotic cycle but that is absent in quiescent cells (Go). Desmin, a muscle-specific intermediate filament protein, was used as a marker for muscle fibers, myotubes, and myoblasts. Western blot analysis (Fig. 3) allowed a semiquantitative analysis of Ki-67. Relative to the amount of all fast twitch myosin heavy chain (internal control; Fig. 3, Fast-MHC, bottom panel), there is a significant (P< 0.01, n = 3) increase in the amount of desmin and Ki-67 in the TA subjected to DO compared to sham controls (densitometric analysis is represented on the right). These data suggest that DO leads to an important increase in the amount of desmin in both age groups that is indicative of myogenesis and that is paralleled by an increase in a marker for cell proliferation.
To localize this proliferative response, we performed double labeling immunofluorescence with antibodies to desmin and Ki-67 (Fig. 4). The data showed numerous proliferating cells (Fig. 4(F), arrows) in the muscle from distracted tibiae of young rabbits and not in the sham controls (Fig. 4(E) and (G)). Proliferating cells were observed less frequently in the muscle from distracted tibiae from adult rabbits, and if seen they were sparsely distributed along the muscle fiber (Fig. 4(H)). Double labeling with antibodies against desmin allowed us to distinguish between proliferating myoblasts and non-muscle cells. We noted that in the young muscle subjected to DO most proliferating cells were desminpositive (Fig. 4(J)), suggesting that they are myoblasts or satellite cells. In the adult muscle subjected to DO, the Ki-67-positive cells were rarely labeled with antibodies against desmin (Fig. 4(L)). By counting the cells that
0
A
kD
10-67
Ki67
203 Desmin
i
r *
I
48.7
-
Fast- MHC
203 -
Fig. 3. Western blot analysis of muscle homogenates. Equal amounts (50 pg of protein) of the muscle homogenates (see Materials and Methods) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose. Antibodies against markers for proliferating cells (Ki-67), muscle (desmin), and an internal control (fast myosin heavy chain) were used to visualize and quantitate the retative amounts of the molecules. The data are representative of three independent experiments and suggest a relative increase of desmin and Ki-67 during DO. Densitometric analysis is shown on the right.
K. Hayatsu, P.G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
90 1
Fig. 4.Double immunofluorescent labeling of the TA. Frozen sections of muscle were double labeled with antibodies against desmin (A-D) and Ki67 (E-H). Image overlays (I-L) allow the visualization of cells containing both molecules (J, e.g. arrows). DO leads to numerous cells that were positively labeled with antibodies against desmin and Ki-67 (B, F, and J, arrows). In the TA from adult rabbits, DO leads to a marginal increase in proliferating cells, that in most cases were not labeled with antibodies to desmin (L).
were double labeled with antibodies against desmin and Ki-67, we found in the experimental TA from young animals 0.58 f0.24/103 pm2 Ki-67 positive cells (n = 5 ) , of which half were also labeled with antibodies against desmin and were localized appropriately. We also observed several proliferating cells in the distal region of the distracted muscle from young animals that were not labeled with antibodies against desmin (data not shown), consistent with remodeling of the connective tissue. In the experimental TA from adult animals we only counted 0.08 f 0.05/103 pm2 Ki-67 positive cells (n = 4) and these proliferating cells were rarely positive for desmin. These results suggest that DO-induced proliferation of myogenic cells only occurs to a significant extent in young muscle. We performed electrophysiological experiments to investigate the functional changes that occur in muscle subjected to DO (Table 1). Our major finding is that the TA in young rabbits adapted to DO without functional deficits and that muscle strength was not affected. By
contrast, the TA in adult rabbits adapted to DO poorly, and muscle weakness was significant. The maximum tetanic tension in single twitch tension was less than 50% that of the controls (Table 1). In the adult rabbits the wet weight of the muscle was not statistically different between the sham controls and the experimental group but in the young rabbits a significant increase in wet weight was noted ( P = 0.035, P = 0.528) in the lengthened TA. As expected, the resting length increased during DO, in both young and mature muscles, in agreement with an earlier report [8]. To examine possible denervation associated with DO, we studied the neuromuscular junction of control and DO muscle using morphological techniques (Fig. 5). The acetylcholine receptors at the synapse are associated with N-CAM, which is lost following denervation [5,6,26]. In control samples, the morphology of the neuromuscular junction was normal (Fig. 5(I) and (K)), as it was in sections from distracted muscle from young rabbits (Fig. 5(J), arrow). In samples from the distracted
902
K. Hayatsu, P.G. De Deyne I Journal of Orthopaedic Research 19 (2001) 897-905
Table 1 Physiological data ~
Electro-physiological
Young
Adult
parameters
Sham (n = 5, mean fSEM)
DO (n = 4 , mean fSEM)
Sham (n = 6, mean fSEM)
DO (n = 6, mean fSEM)
Pmax (N) Twitch tension (N) Wet weight (g) Resting length (cm)
12.59 f 1.90 2.18 10.71 2.06 f0.07 7.34 f0.1
12.17 f0.57 2.73 f0.59 2.21 10.12' 8.4 f0 2 '
19.30f 1.69 4.62 f 0.91 3.77f0.19 8.07 & 0.22
7.36i2.31' 1.18 f0.37' 3.82 +0.15 9.52f0.14'
~
< 0.05. P = 0.075.
*P
**
Fig. 5. N-CAM at the neuromuscular junction in the TA. Frozen sections of muscle were double labeled with rhodamine-conjugated bungarotoxin (A-D) and antibodies against N-CAM (E-H). Image overlay (I-L) allows visualization of both molecules at the neuromuscular junction (J, arrow). In the TA from young rabbits, the DO does not lead to a loss of N-CAM associated with the acetylcholine receptor (B, F, and J, arrows). In the TA from adult rabbits, N-CAM is lost from the neuromuscular junction (D, H, and L, arrow), suggestive of denervation. An increase in intracellular labeling is also observed (H and L, asterisks).
TA of adult rabbits, however, DO led to the loss of NCAM (Fig. 5(H) and (L), urrows) from the synaptic regions (Fig. 5(D) and (L), arrows). In addition, we noted that a majority of the muscle fibers had intracellular labeling for N-CAM (Fig. 5(L), asterisks). Thus our results are consistent with the idea that DO leads to denervation of at least some of the muscle fibers of the
TA. We tried to confirm these findings by quantifying the activity of acetylcholinesterase, which, in rabbits, increases upon of denervation [20]. However, no significant changes in enzyme activity were observed in DO muscle from either young or adult rabbits ( P = 0.3) despite the observed trend that is consistent with denervation (Table 2). The combined results from the
Table 2 Acetylcholine esterase assay Treatment
nmoles Ach hydrolyzed min-’ nig protein-’ (mean SEM)
Young control Young Sliani Young DO Adult control Adult Sham Adult DO
3.5.56 7.73 4 1 . 2 1 10.48 48.96i12.9.5 41.932~17.2 34.05 2C 13.6 53.38 10.9
*
*
*
functional, morphological, and biochemical experiments suggest that a partial denervation occurs in muscle in skeletally mature rabbits subjected to DO.
Discussion
A con~plicationof DO, the development of contractures, has been interpreted as resulting from an insufficient adaptive response of skeletal muscle to the DO regimen. A number of investigators [21,32,33,37] have observed increased complications with a lengthening rate of more than 1 mm/day or with an increase in limb length of more than 20‘1/0. As skeletal maturity may also influence the successful outcome of DO and because DO is performed both in children and adults, we wanted to know whether the adaptation of muscle to DO could be a function of skeletal maturity. Adaptation of muscle subjected to D O might be expected to be more complete in a skeletally immature individual because their higher circulating levels of growth hormone in may support more cell proliferation and tissue growth [lo], and because younger tissues are in general considered t o be more “plastic” [24]. Our results confirm that muscle subjected to DO in skeletally immature rabbits experiences an increased proliferation of myoblasts or satellite cells while maintaining contractile function. Despite the use of a DO regimen that was likely to lead to the development of fibrosis and other complications [33,37], the muscle from the skeletally immature rabbits adapted without developing measurable weakness 121,321. The proliferation of myogenic cells during D O has been reported before [7] and it is quite possible that during D O the adaptive process of muscle, i.e. the proliferation of myogenic cells [7] and even the addition of sarcomeres [33], may depend on bioactive peptides. In growing animals the circulating levels of growth hormone may create an environment in which the adaptation of muscle to DO is facilitated. This process may be mediated through insulin-like growth factor-1 [lo], an important trophic factor for muscle 121, the synthesis of which is in part controlled by growth hormone [lo]. In contrast, the TA from the skeletally mature rabbits showed an adaptive response that was significantly di-
minished and that resulted in severe muscle weakness. It is well known that regeneration in muscle decreases with age [3,4,24]. In models of muscle injury, regeneration is preceded by degenerative events such as inflammatory reactions and the remodeling of the extracellular matrix [22,35,36]. We do not know to what extent these events occur in muscle during D O or whether they hamper the adaptive response of muscle in skeletally mature animals. Nevertheless skeletally mature animals do show an adaptive process, as indicated by the re-expression of neonatal myosin heavy chain in the skeletally mature animals. Howcver our western blots (Fig. 3) and our indirect immunofluorescence experiments (Fig. 4) suggest that this upregulation of neonatal myosin in skeletally mature animals occurs in the absence of significant proliferation of myogenic cells. Proliferation of nonmuscle cells may contribute to additional complications associated with DO such fibrosis [33,37]. Our morphological and biochcniical assays further suggested that DO causes at least partial denervation of the TA in skeletally mature rabbits. This was most clearly indicated by the loss of N-CAM from the neuroniuscular junction of older animals and was consistent with a trend towards higher activity of acetylcholine esterase in muscles subjected to DO. This conclusion is consistent with clinical observations of nerve damage that occurred with DO in achondroplastic dwarfs [30]. In children, a transient denervation-induced muscle weakness of the quadriceps has also been noted, even with small amounts of femoral lengthening (2 cm) (281 and long-term weakness of the quadriceps muscle in children has been observed in femoral lengthening averaging 7 cm [23]. As our results are only suggestive, however, further studies of denervation associated with DO of young and older muscle will be needed. That the adaptation of muscle in skeletally immature rabbits is more robust than in adults is not surprising. In young animals, growth hormone influences the growth and maturation of a number of skeletal tissues [14,34] and some investigators have proposed that increased systemic levels of growth hormone can stimulate a local growth response in muscle [ I ,25,38]. In addition, growth hormone regulates the synthesis of insulin-like growth factor I (IGF-1) [lo], which has been implicated as promotor of muscle proliferation and maturation [2,17,27]. Consistent with the idea that it plays a role in the adaptive response of young muscle to DO, published experiments indicate that IGF-1 acts by a local, paracrine mechanism, especially when the muscle is exposed to stretch [39]. In summary, our report shows that young muscle adapts to DO better than adult muscle when similar distraction variables are applied. Our findings show a vigorous adaptive response in muscle from skeletally immature rabbits without any functional deficits. In contrast, a more modest adaptive response, muscle
K. Huyntsu, P. G. Dr Drj.ne I Journul of' Orthoprirdic Resrurch 15, (2001) 897-905
904
weakness, and a partial denervation are seen in muscle from skeletally mature rabbits. One goal of our future research is to identify therapeutic interventions to prevent these complications in adult animals.
Acknowledgements We would like to acknowledge the support of the National Center for Medical Rehabilitation Research at the NIH (KOIHDO1165), Maryland Center for Limb Lengthening and Reconstruction, and the Association for the Study and the Advancement of the Methods by Ilizarov. We would also like to thank Rick Meyer for technical assistance, Dori Kelly for editorial help, Drs. Dror Paley and John Herzenberg for the helpful discussions and especially Dr. Robert Bloch for a critical reading of the manuscript.
References Aroniadou-Anderjaska V. Lemon PW. Gilloteaux J . Effects of exogenous growth hormone on skeletal muscle of young female rats. Tissue Cell 1996;28:719 24. Barton-Davis ER. Shoturina DI. Musaro A, Rosenthal N. Swecney HL. Viral mediated expression of insulin-like growth factor 1 blocks the aging-related loss of skeletal muscle function. Proc Natl Acad Sci USA 1998;95:15603-7. Cannon JG. Intrinsic and extrinsic factors in muscle aging. Ann NY Acad Sci 1998:854:72-7. Carlson BM. Muscle regeneration and aging. Monographs Develop Biol 1992:23:189 95. Cashman NR, Covault J. Wollman RL, Sanes JR. Neural cell adhesion molecule in normal, denervated. and myopathic h u m a n muscle. Ann Neurol 1987:21:481-9. Covault J, Sanes JR. Neural cell adhesion molecule (N-CAM) accumulates in denervated and paralyzed skeletal muscles. Proc Natl Acad Sci USA 1985:82:4544-8. Day CS. Moreland MS, Floyd SS, Huard J. Limb lengthening promotes muscle growth. J Orthop Res 1997;15:227-34. De Deyne PG, Hayatsu K. Meyer R, Paley D, Herzenberg JE. Muscle regeneration and fiber type transformation during distraction osteogenesis. J Orthop Res 1999;17:560-70. De Deyne PG, Meyer R, Paley D, Herzenberg JE. The adaptation of perimuscular connective tissue during distraction osteogenesis. Clin Orthop 2000;379:259--69. Florini JR, Ewton DZ, Coolican SA. Growth hormone and the insulin-like growth factor system in myogenesis. Endocr Rev 1996: I7:48 1-5 I7. Gondret F. Lefaucheur JP, D'Albis A, Bonneau M . Myosin isoform transitions in four rabbit muscle during postnatal growth. J Muscle Res Cell Mot 1996;17:657-67. Harlow E. Lane D. Immiinoblotting. In: Harlow E. Lane D. editors. Antibodies: a laboratory manual. Cold Spring Harbor Laboratory: 1988. p. 471-504. Herzenberg JE, Scheufele LL, Paley D, Bechtel R. Tepper S. Knee range of motion in isolated femoral lengthening. Clin Orthop I994:301:49--54. Isaksson OG, Ohlsson C, Nilsson A, Isgaard J. Lindahl A. Regulation of cartilage growth by growth hormone and insulinlike growth factor I. Pediatr Nephrol 1991;5:451-3.
Johnson CD, Russell RL. A rapid, simple radiometric gassay for cholinesterase. suitable for multiple determinations. Anal Biochem 1975:64:229 38. Karsch~Mizrdchi I, Travis M. Blau H , Leinwand LA. Expression and DNA sequence analysis o f a human embryonic skeletal muscle myosin heavy chain gene. Nuc Ac Res 1989;17:6167-79. Keller HL, St. Pierre S, Eppihimer LA, Cannon JG. Association of IGF-I and IGF-I1 with myofiber regeneration in vivo. Muscle Nerve 1999:22:347-54. Kojimoto H, Yasui N. Goto T, Matsuda S, Shimomura Y. Bone lengthening in rabbits by callus distraction. The role of periosteiini and endosteum. J Bone J Surg [Br] 1988:70:543-9. Laemmli U K . Cleavage Of StruCturd proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680-5. Lai J, Jedrzejczyk J, PizLey JA, Green D , Barnard EA. Neural control of the forms of acetylcholinestera~ein slow mammalian muscles. Nature 1986:321:724. Lee DY, Choi IH. Chung CY, Chung PH, Chi JG, Suh YL. Effect of tibia1 lengthening on the gastrocnemius muscle. A histopathologic and inorphometric study in rabbits. Acta Orthop Scand 1993;64:688 92. Lef'aucheur JP, Gjata B. Lafont H , Sebille A. Angiogenic and inflammatory responses following skeletal muscle injury are altered by immune neutralization of endogenous basic fibroblast growth factor, insulin-like growth factor- 1 and transforming growth factor-beta 1. J NeuroiinmLinol 1996:70:3744. Maffulli N , Fixsen JA. Muscular strength after callotasis limb lengthening. J Ped Orthop 1995;15:212-6. Marsh DR, Criswell DS, Carson JA, Booth FW. Myogenic regulatory factors during regeneration of skeletal muscle in young. adult, and old rats. J Appl Physiol 1997:83:1270-5. McCall GE, Goulet C, Roy RR, Grindeland RE, Boorman GI, Bigbee AJ. Hodgson JA, Greenisen MC, Edgerton VR. Spaceflight suppresses exercise-induced release of bioassayable growth hormone. J Appi Physiol 1999:87: 1207-12. Moscoso LM. Merlie JP, Sanes J R . N-CAM, 43K-rapsyn, and Slaminin mRNAs are concentrated at synaptic sites in muscle fibers. Mol Cell Neurosci 1995;6:80-9. Musaro A. Roseiithal N. Maturation of the inyogenic program is induced by postmitotic expression of insulin-like growth factor I. Mol Cell Biol 1999; 19:31 15-24. Oey PL, Engelbert RH. van Roermond PM, Weineke G H . Temporary muscle weakness in the early phase of distraction during femoral lengthening. Clinical and electr.omyogl-apliical observations. Electromyogr Clin Neurophysiol 1999;39:2 17-20. Paley D. Problems. obstacles, and complications of limb lengthening by the Ilizarov technique. Clin Orthop 1990:81-104. Polo A, Zambito A, Aldegheri R, Tinazzi M, Rizzuto N. Nerve conduction changes during lower limb lengthening. Somatosensory evoked potentials (SEPs) and F-wave results. Electromyogr Clin Neurophysiol 1999;39:13944. Sartore S, Gorza L, Schiafino S. Fetal myosin heavy chains in regenerating muscle. Nature 1982;298:294-6. Schumacher B, Keller J, Hvid I. Distraction effects on muscle. Leg lengthening studied in rabbits. Acta Orthop Scand 1994;65:647-
so. Simpson AH, Williams PE, Kyberd P, Goldspink G, Kenwright J. The response of muscle to leg lengthening. J Bone J Surg [Br] 1995:77:630-6. Spagnoli A, Rosenfeld R G . The mechanisins by which growth hormone brings about growth. The relative contributions of growth hormone and insulin-like growth factors. Endocrinol Metab Clin North Am 1996;25:615-31. Stauber WT, Fritz VK, Dahlmann B. Extracellular matrix changes following blunt trauma to rat skeletal muscles. Exp Mol Pathol 1990;52:69-86.
K. Huyutsu, P.G. DP Deqne I Joitnlul of Ortkop~~edic Research 19 (2001) 897-905
[36] Tidball JG, St. Pierre BA. Apoptosis of macrophages during the resulution of muscle inflammation. J Leukoc Biol 199639: 380-8. [37] Williams PE, Kyberd P, Simpson AH, Kenwrighl J, Goldspink G. The morphological basis of increased stiKness of rabbit tibialis anterior muscles during surgical limb-lengthening. J Anat l998;l 93:131-8.
905
[38] Wilson VJ, Rattray M, Thomas CR. Moreland BH. Schulster D. Growth hormone increases IGF-I, collagen I and collagen I11 gene expression in dwarf rat skeletal muscle. Mol Cell Endocrinol 1995;I 15:187-97. [39] Yang H, Alnaqeeb M, Simpson AH, Goldspink G. Changes in muscle fibre type, muscle mass and IGF-I gene expression in rabbit skeletal muscle subjected to stretch. J Aiiat 1997:190:613 22.