0306-4522/90 $3.00 t 0.00 Fkrgamon Press plc 0 1990 IBRO
Neuroscience Vol. 39, No. 1, pp. 231-243, 1990 Printed in Great Britain
REINNERVATION AND RECOVERY OF MOUSE SOLEUS MUSCLE AFTER LONG-TERM DENERVATION A. IRINTCHEV, A. DRAGUHNand A. WERNIG* Physiol&&es Institut der Universitiit Bonn-Neurophysiologie,
WilhelmstraBe 31,530O Bonn 1, F.R.G.
Ak&aet--Reinnervation and recovery of the mouse soleus muscle were studied 2-10 months after denervation periods of about 7 months. To maintain denervation the right sciatic nerve was frozen 14 times at 2-week intervals. Though initially intermittent muscle reinnervation occurred, contractile force of denervated muscles was reduced to less than 10% of the contralateral muscles by the fifth nerve freezing and further declined thereafter. Following reinnervation, recovery of soleus muscle force proceeded slowly to reach plateau values after 5-6 months. Tetanic muscle force reached on average 72% (range X3-86%, rr = 12) of contralateral muscles after S-10 months, (P < 0.01, t-test for absolute values) and 87% of unoperated animals after 10 months (P < 0.05, n = 5). Muscle fibre diameters were significantly reduced in reinnervated muscles, but frequency distributions were normal and similarly shaped in reinnervated and control muscles, suggesting complete muscle reinnervation and the absence of denervated fibres even at 2 months of reinnervation. Total numbers of muscle fibres were similar in reinnervated (842 173 S.D., n = 15), contralateral (854 f 104 S.D., n = 15) and control soleus muscles (853 + 77 S.D., n = 5). The number of myelinated axons in regenerating soleus nerves reached control values by 3 months after the last freezing, continued to increase till 6 months (150% of control), and declined thereafter (125% at 9-10 months). In the contralateral soleus nerves the number of myelinated axons remained constant during this period. Nerve fibre diameters remained abnormally small; even after 10 months of reinnervation fibre diameters were unimodally distributed with a mean diameter of 3.3 pm in contrast to the bimodal distribution in intact nerves (mean values 3.9 and 9.0pm, respectively). Total fibre cross-section area per nerve increased with time but reached only 54% + 6 S.D., (n = 3) of contralateral nerves by 10 months. The relative thickness of the myelin sheath (g-ratio) returned to normal after 9-10 months. Anatomically, muscle reinnervation appeared to be complete by 7-8 weeks since unusually small muscle fibre profiles were absent. Functionally, however, innervation was rendered immature for the following reasons: acetylcholine-induced contractures (50 mg/l acetylcholine perchlorate) were still larger than normal (17% f 4 S.D., n = 6, of maximum tetanic force as compared to 3% + 1 S.D. in normal muscles, n = 12), nerve-evoked tetani (50 Hz for 2s) showed fatigue and peak values reached only 86% + 6 SD. (n = 5) of the muscle force obtained from direct muscle stimulation. Even by 3 months acetylcholine contractures were slightly elevated (8% + 2 SD., n = 6, P < 0.05) and block resistance of the nerveevoked responses in high magnesium/low calcium Tyrode (safety margin of transmission) was reduced. No deficits in nerve-related parameters were found after 5-t0 months of reinnervation. At 7-8 weeks of reinnervation fibre type distribution was abnormal since most muscle fibres stained for acid-stable myosin ATPase (Type I fibres, pH 4.5) only a few for alkaline-stable myosin ATPase (Type IIB, pH 10.3) and some for both (Type IIC). Consequently, the frequency of Type I fibres declined and by 3 months fibre type distribution approached values in normal control animals. Surprisingly, in contralateral muscles a similar SU~&IS followed by a decline in Type I fibres was observed. Running in wheels during the denervation and reinnervation periods had moderately positive effects on muscle force of both reinnervated (plus 17%) and contralateral control muscles (plus 14%) while training during the reinnervation period only (5-6 months) surprisingly was not effective at all. In conclusion it is suggested that muscle recovery after prolonged periods of denervation, though incomplete, is remarkably successful, when proper motoneurons reinnervate in suihcient numbers. The strikingly slow progress in recovery of muscle force despite early full reinnervation could be due to deficits in the regenerated nerves which suffered a heavy loss in axonal volume; alternatively, muscle recovery itself might be impaired, e.g. due to exhaustion of the satellite cell pool.
Despite the enormous capacity for regeneration of both peripheral nerve and muscle, clinical results of muscle reinnervation and recovery are often disappointing.1g,77 One potential problem that determines the outcome is the correct re-establishment of
*To whom correspondence
connections between neurons (motor and sensory) and the appropriate muscles.‘,*~” In the present inves-
tigation, therefore, we used repeated freezings rather than section of the nerve to denervate muscles, expecting less axonal misdirection.31,‘6.48,49One complication, however, added by this protocol is the intermittent reinnervation after the first few freezings (see Experimental Procedures). Factors limiting muscle recovery could also be located in the muscle itself. It is known that long-term
should be addressed.
A6breviation.s: ACh, acetylcholine; acid (alkaline) mATPase,
acid (alkaline)-preincubated myosine adenosine triphosphatase; SDH, succinate dehydrogenase. 231
232
A. IRINTCHEVet al.
denervated muscle fibres undergo degenerative changes;3.‘0.1’.‘4,2.75t h ough proliferation of satellite cells also occurs,3”0,55muscle tissue cannot be maintained after extended denervation periods. Generally, denervation periods of months or years have only rarely been studied in controlled animal experiments.324J528.3I
In the present experiments muscles were kept denervated for 7 months and muscle recovery was followed up till 10 months of reinnervation. Possible effects of voluntary running in wheels on the process of recovery were also studied. One group of animals was trained during both the denervation period and during reinnervation, one group started training after the last nerve freezing (present paper) while a third group, which surprisingly showed positive effects on muscle force, was trained during the denervation period only.34a In another paper, l9 the degree of axonal misdirection was estimated from labelling motoneuron pools of reinnervated tibialis anterior muscles retrogradely with horseradish peroxidase, and determining numbers, size and rostrocaudal distribution of motoneurons in the spinal cord. After multiple nerve freezing, the motoneuron pool was found to be broader than normal but less abnormal than after nerve section and suturing or cross-union of tibia1 and peroneal nerves. Similarly, co-ordination of antagonistic muscles (tibialis anterior and medial gastrocnemius) monitored from electromyograph recordings in freely running animals was well recovered after nerve freezings. Short accounts of this work have been published previously.78,8’” EXPERIMENTAL
PROCEDURES
Animals and animal care
Female C57Bl/6J mice obtained from Charles River Wiga (F.R.G.) at 2 months of age were kept individually in standard cages supplied with water and standard food ad libitum. Constant temperature (22.5 k 2.5”(Z), illumination (12-h on-ff cycle) and relative humidity (approximately 80%) were maintained in the animal rooms. Reinnervation was studied in 34 mice; two groups of untreated normal animals (n = 5 for both) investigated at 9 and 19 months of age served as controls. Mean body weight for all groups was
similar (30 g) at the time of the acute experiments. Muscle denervation by repeated nerve freezings Muscle denervation was maintained for 7 months with 14 repeated freezings of the right sciatic nerve at t-week intervals. In control experiments (10 solei of male NIH mice) it was found from comparing direct and indirect muscle stimulation in vitro that in the beginning muscle reinnervation occurred during the time periods between two nerve freezings but was absent after about eight freezings. Muscle atrophy progressed rapidly and muscle force declined to some 5-10% of the contralateral muscles after the fifth freezing. Other contractile parameters (twitch-tetanus ratio, time-to-peak and half-relaxation time of the twitch) and muscle mass approached values measured in long-term denervated mouse soleus muscle.80 Contralateral intact solei increased in force with increasing number of freeings, probably due to work hypertrophy caused by the impairment of the function of the denervated leg.
Surgical treatments were performed under pentobarbital (Nembutal) anaesthesia (4&&l mg/kg body weight). LOCd anaesthetic (Xylocain) was applied subcutaneously prior to skin incision. The sciatic nerve was exposed at the level of trochanter femoris and a 2-3mm segment was frozen by attaching to it a metal cryode cooled with liquid nitrogen w’.~’ for about 30 s. Skin incision was closed by one or two wound clips. The total surgical procedure took about 5 min. Multiple nerve freezings were very well tolerated and not a single animal was lost due to inflammatory reactions or other consequences of the denervation procedure. Experiments were performed according to the German laws for protection of animals, including the required permit. Wheel running
To study the effect of increased motor activity (by voluntary wheel running) on muscle recovery some animal cages were supplied with running wheels (6Ocm circumference, 9.5 cm width, 275 g weight) during different periods.4,34Nine animals exercised during both the denervation and reinnervation periods. Mean daily running activity of these animals was 5.8 km + 1.5 S.D. per day (range 2.5-7.3 km) during the time of denervation and 5.0 km f 1.5 S.D. per day during the reinnervation period (range 2.0-6.9 km). Other mice (n = 7) started running after the last operation, i.e. in the reinnervation period; running activity in this group was significantly higher (8.1 km f 0.8 S.D. per day, range 6.9-9.0 km; P < 0.005 compared to the previous group, t-test).Finally, a group of five animals started exercise after a reinnervation period of 40 days. Exercised and sedentary mice were investigated at different periods of reinnervation. In vitro isometric tension measurements Contraction measurements of reinnervated and contralateral muscles were performed after reinnervation times (time after last nerve freezing) of 7 weeks to 10 months as previously described.4 Muscles from intact animals were studied at 9 and 19 months of age. Soleus nerve-muscle preparations were dissected from anaesthetized animals, transferred to aerated (95% O,, 5% CO,) Tyrode’s solution in lucite chambers and connected to a force transducer. Composition of Tyrode’s solution was (in mM): 125 NaCl, 24 NaHCO,, 5.37 KCl, 1.0 MgCl,, 1.8 CaCl, and 5% glucose. Two muscles from one animal (experimental and contralateral) were studied simultaneously in two chambers with a common perfusion system. Temperature was kept at 25.O”C (range 24.5-25.5”C) throughout the measurements to reduce metabolism and prevent 0, and glucose deficits of inner fibres in vitro.@ Muscles were stimulated electrically through a pair of silver electrodes in the bath (direct muscle stimulation). Nerves were stimulated through suction electrodes with 0.1 ms, 5-6 V pulses (indirect muscle stimulation). The optimal pretension for maximal force generation in each muscle was determined while applying single electrical pulses (0.5 ms, 20 V) to the muscle. Subsequently, voltage levels were gradually increased until a plateau in isometric muscle force was reached. Amplitudes twice the lowest voltage sufficient to cause a plateau value (typically 20-25 V) were used throughout the experiment. Stimulus intensities up to 35-40 V and pulse durations up to 0.7 ms did not produce larger than plateau values in both reinnervated (n = 3) and intact (n = 3) muscles. Thus, we assume that with our usual stimulation procedure the whole muscle was completely activated.‘3Js”5 Further increase in charge application (mainly stimulus duration) led to notable enlargement (up to three-fold) of the twitch amplitude both in reinnervated and intact muscles, but no plateau was reached.” It can be assumed that with longer pulses more than one action potential fired per pulse and/or additional electromechanical coupling is triggered by the depolarizing pulses directly (electrical contracture). Muscle performance rapidly declined after such vigorous stimulation, probably
233
Reinnervation of long-term denervated muscle because of muscle fibre damage,” but was well maintained for many hours after the usual stimulation protocol? Muscles and nerves were stimulated with single pulses and tetani (50 and IO0 Hz for 2 s). The sequence of measurements on any muscle was kept constant in all experiments with adequate time between single measurements (3min between tetanic/twitch stimulations). The variables measured were: twitch (P,) and tetanic tension (PO), timeto-peak of the twitch. Muscle sensitivity to acetylcholine [ACh) was tested with 50 mgjl ACh per&orate in Tyrode applied directly to the chambers by quickly exchanging the solutions. ACh sensitivity was defined as the ratio of the ACh contracture peak amplitude to the amplitude of the preceding tetanus in per cent (%Po’). After ACh contractures muscles were allowed to recover for sufficient time (minimum of 12 mitt, 12-15 mm typically). One physiolo~~l measure of muscle inne~ation is the ratio indirect- to direct-evoked tension (functionat innervation ratio}.4+i2For normal muscles this ratio is 100% and is reduced in partially denervated muscles. Following complete ~inne~ation, however, the ratio remains reduced for a period of time, possibly due to immaturity of synaptic transmission, and is thus a poor measure for reinnervation.r’+3rJ4 We also found good correlations between indirect/direct force ratio and ACh sensitivity at ACh perchlorate concentrations of 5 mg/l (r = 0.87, P < 0.01, r-test for regression) and SOmg/l (r = 0.91, P < 0.01, n = 7 for each concentration; not shown); in reinnervated muscles obviously extrasynaptic receptors for ACh persist after functionally incomplete or immature reinnervation. Thus, normal sensitivity to ACh, but a poor functional innervation ratio, can be taken as evidence for acute nerve damage during preparation. Safety margin of transmission in nerve-muscle preparations was determined from the block resistance in high magnesium/low calcium (2.0 mM/O.S mM) Tyrode’s solution After 30 min of calibration in the blocking solution, the degree of ~ansmis~on block was estimated from the relative decline of the indirect twitch and tetanus amplitudes as compared to normal bathing solution. After the tension measurements muscles were blotted dry and weighed with parts of the distal and proximal tendons present. Morphologicalprocedures After weighing, the muscles were placed on pieces of calf-liver, pinned to approximately resting length and rapidly frozen in isopentane precooled in liquid nitrogen. The liver pieces had been fixed in 10% formalin for long periods of time, washed thoroughly in running tap water for several hours, and in Tyrode’s solution for 24 h prior to use. This preparation allows performance of any ~st~he~cal procedure used here without interference from the liver tissue. The use of the liver pieces as support secures and facilitates cutting of cross-sections, especially through very atrophic muscles, and more importantly, assures that no part of the cross-sectional area is lost during cutting. Serial cross-sections (IOpm) cut at the level of the end-plate region were collected on coverslips. Due to the specific muscle architecture cross-sections at this level contain all muscle fibres.” The end-plate region was recognized in rapid Toluidine Blue staining of control sections from the disappearance of proximal tendon and the entry of the muscle nerve. Serial cross-sections from each muscle were stained with 1% Toluidine Blue/l% borax and for the following enzyme activities: glycogen phosphorylase,)6 suc&ate dehydrogenase (SDH),% myofibrillar ATPase after acid (PH 4.5) and alkali (PH 10.3) preincubation.q,*3 Overfapping videoprints (videocamera AVT-CD and videoprinter Sony UP-104) from selected stained sections were glued together to reconstruct the whole cross-se&on (final rna~~~cation x 250-580) and used as “bard” copies
for quantitative evaluations. Total numbers of muscle hbms were dete~in~ from sections stained for glycogen phosphorylase activity, muscle fibre diameters from Toluidine Blue stain&. Mean orthogonal diameters (mean of the longest axis and a short one passing through the middle of the longest at right angle) were determined according to the method of Song et al.” as modified by Schmitt@ and measured with a digitizing tablet.*’ All fibres in a crosssection were measured except mechanically damaged tibres or areas with any artifact. All muscles used for determination of fibre diameters were examined at three levels for “split” muscle fibres (Toluidine Blue, mATPase, spaced serial sections, 108 pm intervals, see Ref. 34). Total muscle cross-sectional area was calculated from the mean orthogonal diameter of a whole muscle cross-section assuming circular shape. Nerves to soleus muscles were sampled from nervemuscle preparations following tension measurement. The whole nerve trunk (part of n. ischiadicus, n. tibia& and the nerve to m. soleus) was fixed in Baker’s formol-calcium at 4°C overnight and the nerves dissected free from muscle and conn~tive tissue under high-power ma~ifi~tion. After pos~xation in osmium, the nerves were embedded gat in Epon. Semithin (1 pm) cross-sections were cut with glass knives from the most distal unbranched part of the muscle nerves73 and these were stained with Toluidine Blue. Videoprints (magnification x 1480 to x 1880) were used to count the number of myelinated nerve fibres, measure fibre diameters and estimate the g-ratio (axonifibre diameter, see Refs 20 and 68). Statistical tests were considered significant at the 5% level. Unless otherwise indicated, the r-test is used to compare means. RESULTS Muscle recovery
after long-term ~ne~atio~
Tetanic force of reinnervated muscles 7-8 weeks after the last operation was 25% of contralateral muscles (Table I), increased three-fold in the following 3 months and reached on the average 69% and 77% of contralateral muscle force at 5-6 and 910 months, respectively (P < 0.05, Fig. 1 and Table 1). Twitch force improved better than tetanic force. Regain of muscle weight, similar to tetanic force, was slow but approached control values by 910 months. Time-to-peak of the twitch continuously decreased, attaining values smaller than the contralateral ones at 3 months of reinne~ation. T~tch-tetanus ratio declined with time of ~inn~~ation without reaching control values (Table 1). Muscle weights were somewhat larger in contralateral muscles of experimental animals than in untreated controls of comparable age (Table 1, P < 0.05), indicating work hypertrophy in contralateral muscles (also see below). A similar trend (P -C0.1) was visible for tetanic muscle force. In contmlateral control muscles some reduction in force and muscle weight occurred between 1 l-l 5 and 18-19 months of age (Fig. 1, upper part and Table 1). This might be due to ageing,Sa since in 19-month-old control animals tetanic force and muscle weight (Table 1) were signi~cantiy lower than in animals at 9 months of age (tetanic force 224 mN -t_12S.D., muscle weight 16 mg & 1 SD., 10 muscles, five animals, P < 0.05; data not included in Table 1).
11+2
8&l*
616
No. of muscles/animals
45&6*
51 St:5*
13 & 1** 8188
4+2
40 t 7**
5.26 + 0.03**
44rt 11**
169 f 29**
14-U 5-6 months
14&2 4/4b
2.6
38 & 3**
0.26 Ifr0.03**
43&3
170 + 9**t
18-19 9-10 months
..--
818
17 & 3
351
50 + 7
0.23 & 0.01
56 + 6
245 + 30
1415 (56 months)
16-t 1 414”
4.2
46&3
0.17 f 0.03t
37 & 13t
223 -t_26
18-19 (9-10 months)
Contralateral muscies
815
14 z!z2it
441.
43 + 4
0.24 f 0.02tT
46 & 6
196 & 18
19
Intact control animals
Values are means f SD. Symbols indicate values significantly different (P < 0.05, t-test) from values for: *muscles reinnervated for 56 months; **respective group of contralateral muscles; tcontralateral muscles of younger animals; stage-matched contralateral muscles; fintact control animals. “
Neuroscience Vol. 39, No. 1, pp. 231-243, 1990 Printed in Great Britain
REINNERVATION AND RECOVERY OF MOUSE SOLEUS MUSCLE AFTER LONG-TERM DENERVATION A. IRINTCHEV, A. DRAGUHNand A. WERNIG* Physiol&&es Institut der Universitiit Bonn-Neurophysiologie,
WilhelmstraBe 31,530O Bonn 1, F.R.G.
Ak&aet--Reinnervation and recovery of the mouse soleus muscle were studied 2-10 months after denervation periods of about 7 months. To maintain denervation the right sciatic nerve was frozen 14 times at 2-week intervals. Though initially intermittent muscle reinnervation occurred, contractile force of denervated muscles was reduced to less than 10% of the contralateral muscles by the fifth nerve freezing and further declined thereafter. Following reinnervation, recovery of soleus muscle force proceeded slowly to reach plateau values after 5-6 months. Tetanic muscle force reached on average 72% (range X3-86%, rr = 12) of contralateral muscles after S-10 months, (P < 0.01, t-test for absolute values) and 87% of unoperated animals after 10 months (P < 0.05, n = 5). Muscle fibre diameters were significantly reduced in reinnervated muscles, but frequency distributions were normal and similarly shaped in reinnervated and control muscles, suggesting complete muscle reinnervation and the absence of denervated fibres even at 2 months of reinnervation. Total numbers of muscle fibres were similar in reinnervated (842 173 S.D., n = 15), contralateral (854 f 104 S.D., n = 15) and control soleus muscles (853 + 77 S.D., n = 5). The number of myelinated axons in regenerating soleus nerves reached control values by 3 months after the last freezing, continued to increase till 6 months (150% of control), and declined thereafter (125% at 9-10 months). In the contralateral soleus nerves the number of myelinated axons remained constant during this period. Nerve fibre diameters remained abnormally small; even after 10 months of reinnervation fibre diameters were unimodally distributed with a mean diameter of 3.3 pm in contrast to the bimodal distribution in intact nerves (mean values 3.9 and 9.0pm, respectively). Total fibre cross-section area per nerve increased with time but reached only 54% + 6 S.D., (n = 3) of contralateral nerves by 10 months. The relative thickness of the myelin sheath (g-ratio) returned to normal after 9-10 months. Anatomically, muscle reinnervation appeared to be complete by 7-8 weeks since unusually small muscle fibre profiles were absent. Functionally, however, innervation was rendered immature for the following reasons: acetylcholine-induced contractures (50 mg/l acetylcholine perchlorate) were still larger than normal (17% f 4 S.D., n = 6, of maximum tetanic force as compared to 3% + 1 S.D. in normal muscles, n = 12), nerve-evoked tetani (50 Hz for 2s) showed fatigue and peak values reached only 86% + 6 SD. (n = 5) of the muscle force obtained from direct muscle stimulation. Even by 3 months acetylcholine contractures were slightly elevated (8% + 2 SD., n = 6, P < 0.05) and block resistance of the nerveevoked responses in high magnesium/low calcium Tyrode (safety margin of transmission) was reduced. No deficits in nerve-related parameters were found after 5-t0 months of reinnervation. At 7-8 weeks of reinnervation fibre type distribution was abnormal since most muscle fibres stained for acid-stable myosin ATPase (Type I fibres, pH 4.5) only a few for alkaline-stable myosin ATPase (Type IIB, pH 10.3) and some for both (Type IIC). Consequently, the frequency of Type I fibres declined and by 3 months fibre type distribution approached values in normal control animals. Surprisingly, in contralateral muscles a similar SU~&IS followed by a decline in Type I fibres was observed. Running in wheels during the denervation and reinnervation periods had moderately positive effects on muscle force of both reinnervated (plus 17%) and contralateral control muscles (plus 14%) while training during the reinnervation period only (5-6 months) surprisingly was not effective at all. In conclusion it is suggested that muscle recovery after prolonged periods of denervation, though incomplete, is remarkably successful, when proper motoneurons reinnervate in suihcient numbers. The strikingly slow progress in recovery of muscle force despite early full reinnervation could be due to deficits in the regenerated nerves which suffered a heavy loss in axonal volume; alternatively, muscle recovery itself might be impaired, e.g. due to exhaustion of the satellite cell pool.
Despite the enormous capacity for regeneration of both peripheral nerve and muscle, clinical results of muscle reinnervation and recovery are often disappointing.1g,77 One potential problem that determines the outcome is the correct re-establishment of
*To whom correspondence
connections between neurons (motor and sensory) and the appropriate muscles.‘,*~” In the present inves-
tigation, therefore, we used repeated freezings rather than section of the nerve to denervate muscles, expecting less axonal misdirection.31,‘6.48,49One complication, however, added by this protocol is the intermittent reinnervation after the first few freezings (see Experimental Procedures). Factors limiting muscle recovery could also be located in the muscle itself. It is known that long-term
should be addressed.
A6breviation.s: ACh, acetylcholine; acid (alkaline) mATPase,
acid (alkaline)-preincubated myosine adenosine triphosphatase; SDH, succinate dehydrogenase. 231
232
A. IRINTCHEVet al.
denervated muscle fibres undergo degenerative changes;3.‘0.1’.‘4,2.75t h ough proliferation of satellite cells also occurs,3”0,55muscle tissue cannot be maintained after extended denervation periods. Generally, denervation periods of months or years have only rarely been studied in controlled animal experiments.324J528.3I
In the present experiments muscles were kept denervated for 7 months and muscle recovery was followed up till 10 months of reinnervation. Possible effects of voluntary running in wheels on the process of recovery were also studied. One group of animals was trained during both the denervation period and during reinnervation, one group started training after the last nerve freezing (present paper) while a third group, which surprisingly showed positive effects on muscle force, was trained during the denervation period only.34a In another paper, l9 the degree of axonal misdirection was estimated from labelling motoneuron pools of reinnervated tibialis anterior muscles retrogradely with horseradish peroxidase, and determining numbers, size and rostrocaudal distribution of motoneurons in the spinal cord. After multiple nerve freezing, the motoneuron pool was found to be broader than normal but less abnormal than after nerve section and suturing or cross-union of tibia1 and peroneal nerves. Similarly, co-ordination of antagonistic muscles (tibialis anterior and medial gastrocnemius) monitored from electromyograph recordings in freely running animals was well recovered after nerve freezings. Short accounts of this work have been published previously.78,8’” EXPERIMENTAL
PROCEDURES
Animals and animal care
Female C57Bl/6J mice obtained from Charles River Wiga (F.R.G.) at 2 months of age were kept individually in standard cages supplied with water and standard food ad libitum. Constant temperature (22.5 k 2.5”(Z), illumination (12-h on-ff cycle) and relative humidity (approximately 80%) were maintained in the animal rooms. Reinnervation was studied in 34 mice; two groups of untreated normal animals (n = 5 for both) investigated at 9 and 19 months of age served as controls. Mean body weight for all groups was
similar (30 g) at the time of the acute experiments. Muscle denervation by repeated nerve freezings Muscle denervation was maintained for 7 months with 14 repeated freezings of the right sciatic nerve at t-week intervals. In control experiments (10 solei of male NIH mice) it was found from comparing direct and indirect muscle stimulation in vitro that in the beginning muscle reinnervation occurred during the time periods between two nerve freezings but was absent after about eight freezings. Muscle atrophy progressed rapidly and muscle force declined to some 5-10% of the contralateral muscles after the fifth freezing. Other contractile parameters (twitch-tetanus ratio, time-to-peak and half-relaxation time of the twitch) and muscle mass approached values measured in long-term denervated mouse soleus muscle.80 Contralateral intact solei increased in force with increasing number of freeings, probably due to work hypertrophy caused by the impairment of the function of the denervated leg.
Surgical treatments were performed under pentobarbital (Nembutal) anaesthesia (4&&l mg/kg body weight). LOCd anaesthetic (Xylocain) was applied subcutaneously prior to skin incision. The sciatic nerve was exposed at the level of trochanter femoris and a 2-3mm segment was frozen by attaching to it a metal cryode cooled with liquid nitrogen w’.~’ for about 30 s. Skin incision was closed by one or two wound clips. The total surgical procedure took about 5 min. Multiple nerve freezings were very well tolerated and not a single animal was lost due to inflammatory reactions or other consequences of the denervation procedure. Experiments were performed according to the German laws for protection of animals, including the required permit. Wheel running
To study the effect of increased motor activity (by voluntary wheel running) on muscle recovery some animal cages were supplied with running wheels (6Ocm circumference, 9.5 cm width, 275 g weight) during different periods.4,34Nine animals exercised during both the denervation and reinnervation periods. Mean daily running activity of these animals was 5.8 km + 1.5 S.D. per day (range 2.5-7.3 km) during the time of denervation and 5.0 km f 1.5 S.D. per day during the reinnervation period (range 2.0-6.9 km). Other mice (n = 7) started running after the last operation, i.e. in the reinnervation period; running activity in this group was significantly higher (8.1 km f 0.8 S.D. per day, range 6.9-9.0 km; P < 0.005 compared to the previous group, t-test).Finally, a group of five animals started exercise after a reinnervation period of 40 days. Exercised and sedentary mice were investigated at different periods of reinnervation. In vitro isometric tension measurements Contraction measurements of reinnervated and contralateral muscles were performed after reinnervation times (time after last nerve freezing) of 7 weeks to 10 months as previously described.4 Muscles from intact animals were studied at 9 and 19 months of age. Soleus nerve-muscle preparations were dissected from anaesthetized animals, transferred to aerated (95% O,, 5% CO,) Tyrode’s solution in lucite chambers and connected to a force transducer. Composition of Tyrode’s solution was (in mM): 125 NaCl, 24 NaHCO,, 5.37 KCl, 1.0 MgCl,, 1.8 CaCl, and 5% glucose. Two muscles from one animal (experimental and contralateral) were studied simultaneously in two chambers with a common perfusion system. Temperature was kept at 25.O”C (range 24.5-25.5”C) throughout the measurements to reduce metabolism and prevent 0, and glucose deficits of inner fibres in vitro.@ Muscles were stimulated electrically through a pair of silver electrodes in the bath (direct muscle stimulation). Nerves were stimulated through suction electrodes with 0.1 ms, 5-6 V pulses (indirect muscle stimulation). The optimal pretension for maximal force generation in each muscle was determined while applying single electrical pulses (0.5 ms, 20 V) to the muscle. Subsequently, voltage levels were gradually increased until a plateau in isometric muscle force was reached. Amplitudes twice the lowest voltage sufficient to cause a plateau value (typically 20-25 V) were used throughout the experiment. Stimulus intensities up to 35-40 V and pulse durations up to 0.7 ms did not produce larger than plateau values in both reinnervated (n = 3) and intact (n = 3) muscles. Thus, we assume that with our usual stimulation procedure the whole muscle was completely activated.‘3Js”5 Further increase in charge application (mainly stimulus duration) led to notable enlargement (up to three-fold) of the twitch amplitude both in reinnervated and intact muscles, but no plateau was reached.” It can be assumed that with longer pulses more than one action potential fired per pulse and/or additional electromechanical coupling is triggered by the depolarizing pulses directly (electrical contracture). Muscle performance rapidly declined after such vigorous stimulation, probably
233
Reinnervation of long-term denervated muscle because of muscle fibre damage,” but was well maintained for many hours after the usual stimulation protocol? Muscles and nerves were stimulated with single pulses and tetani (50 and IO0 Hz for 2 s). The sequence of measurements on any muscle was kept constant in all experiments with adequate time between single measurements (3min between tetanic/twitch stimulations). The variables measured were: twitch (P,) and tetanic tension (PO), timeto-peak of the twitch. Muscle sensitivity to acetylcholine [ACh) was tested with 50 mgjl ACh per&orate in Tyrode applied directly to the chambers by quickly exchanging the solutions. ACh sensitivity was defined as the ratio of the ACh contracture peak amplitude to the amplitude of the preceding tetanus in per cent (%Po’). After ACh contractures muscles were allowed to recover for sufficient time (minimum of 12 mitt, 12-15 mm typically). One physiolo~~l measure of muscle inne~ation is the ratio indirect- to direct-evoked tension (functionat innervation ratio}.4+i2For normal muscles this ratio is 100% and is reduced in partially denervated muscles. Following complete ~inne~ation, however, the ratio remains reduced for a period of time, possibly due to immaturity of synaptic transmission, and is thus a poor measure for reinnervation.r’+3rJ4 We also found good correlations between indirect/direct force ratio and ACh sensitivity at ACh perchlorate concentrations of 5 mg/l (r = 0.87, P < 0.01, r-test for regression) and SOmg/l (r = 0.91, P < 0.01, n = 7 for each concentration; not shown); in reinnervated muscles obviously extrasynaptic receptors for ACh persist after functionally incomplete or immature reinnervation. Thus, normal sensitivity to ACh, but a poor functional innervation ratio, can be taken as evidence for acute nerve damage during preparation. Safety margin of transmission in nerve-muscle preparations was determined from the block resistance in high magnesium/low calcium (2.0 mM/O.S mM) Tyrode’s solution After 30 min of calibration in the blocking solution, the degree of ~ansmis~on block was estimated from the relative decline of the indirect twitch and tetanus amplitudes as compared to normal bathing solution. After the tension measurements muscles were blotted dry and weighed with parts of the distal and proximal tendons present. Morphologicalprocedures After weighing, the muscles were placed on pieces of calf-liver, pinned to approximately resting length and rapidly frozen in isopentane precooled in liquid nitrogen. The liver pieces had been fixed in 10% formalin for long periods of time, washed thoroughly in running tap water for several hours, and in Tyrode’s solution for 24 h prior to use. This preparation allows performance of any ~st~he~cal procedure used here without interference from the liver tissue. The use of the liver pieces as support secures and facilitates cutting of cross-sections, especially through very atrophic muscles, and more importantly, assures that no part of the cross-sectional area is lost during cutting. Serial cross-sections (IOpm) cut at the level of the end-plate region were collected on coverslips. Due to the specific muscle architecture cross-sections at this level contain all muscle fibres.” The end-plate region was recognized in rapid Toluidine Blue staining of control sections from the disappearance of proximal tendon and the entry of the muscle nerve. Serial cross-sections from each muscle were stained with 1% Toluidine Blue/l% borax and for the following enzyme activities: glycogen phosphorylase,)6 suc&ate dehydrogenase (SDH),% myofibrillar ATPase after acid (PH 4.5) and alkali (PH 10.3) preincubation.q,*3 Overfapping videoprints (videocamera AVT-CD and videoprinter Sony UP-104) from selected stained sections were glued together to reconstruct the whole cross-se&on (final rna~~~cation x 250-580) and used as “bard” copies
for quantitative evaluations. Total numbers of muscle hbms were dete~in~ from sections stained for glycogen phosphorylase activity, muscle fibre diameters from Toluidine Blue stain&. Mean orthogonal diameters (mean of the longest axis and a short one passing through the middle of the longest at right angle) were determined according to the method of Song et al.” as modified by Schmitt@ and measured with a digitizing tablet.*’ All fibres in a crosssection were measured except mechanically damaged tibres or areas with any artifact. All muscles used for determination of fibre diameters were examined at three levels for “split” muscle fibres (Toluidine Blue, mATPase, spaced serial sections, 108 pm intervals, see Ref. 34). Total muscle cross-sectional area was calculated from the mean orthogonal diameter of a whole muscle cross-section assuming circular shape. Nerves to soleus muscles were sampled from nervemuscle preparations following tension measurement. The whole nerve trunk (part of n. ischiadicus, n. tibia& and the nerve to m. soleus) was fixed in Baker’s formol-calcium at 4°C overnight and the nerves dissected free from muscle and conn~tive tissue under high-power ma~ifi~tion. After pos~xation in osmium, the nerves were embedded gat in Epon. Semithin (1 pm) cross-sections were cut with glass knives from the most distal unbranched part of the muscle nerves73 and these were stained with Toluidine Blue. Videoprints (magnification x 1480 to x 1880) were used to count the number of myelinated nerve fibres, measure fibre diameters and estimate the g-ratio (axonifibre diameter, see Refs 20 and 68). Statistical tests were considered significant at the 5% level. Unless otherwise indicated, the r-test is used to compare means. RESULTS Muscle recovery
after long-term ~ne~atio~
Tetanic force of reinnervated muscles 7-8 weeks after the last operation was 25% of contralateral muscles (Table I), increased three-fold in the following 3 months and reached on the average 69% and 77% of contralateral muscle force at 5-6 and 910 months, respectively (P < 0.05, Fig. 1 and Table 1). Twitch force improved better than tetanic force. Regain of muscle weight, similar to tetanic force, was slow but approached control values by 910 months. Time-to-peak of the twitch continuously decreased, attaining values smaller than the contralateral ones at 3 months of reinne~ation. T~tch-tetanus ratio declined with time of ~inn~~ation without reaching control values (Table 1). Muscle weights were somewhat larger in contralateral muscles of experimental animals than in untreated controls of comparable age (Table 1, P < 0.05), indicating work hypertrophy in contralateral muscles (also see below). A similar trend (P -C0.1) was visible for tetanic muscle force. In contmlateral control muscles some reduction in force and muscle weight occurred between 1 l-l 5 and 18-19 months of age (Fig. 1, upper part and Table 1). This might be due to ageing,Sa since in 19-month-old control animals tetanic force and muscle weight (Table 1) were signi~cantiy lower than in animals at 9 months of age (tetanic force 224 mN -t_12S.D., muscle weight 16 mg & 1 SD., 10 muscles, five animals, P < 0.05; data not included in Table 1).
11+2
8&l*
616
No. of muscles/animals
45&6*
51 St:5*
13 & 1** 8188
4+2
40 t 7**
5.26 + 0.03**
44rt 11**
169 f 29**
14-U 5-6 months
14&2 4/4b
2.6
38 & 3**
0.26 Ifr0.03**
43&3
170 + 9**t
18-19 9-10 months
..--
818
17 & 3
351
50 + 7
0.23 & 0.01
56 + 6
245 + 30
1415 (56 months)
16-t 1 414”
4.2
46&3
0.17 f 0.03t
37 & 13t
223 -t_26
18-19 (9-10 months)
Contralateral muscies
815
14 z!z2it
441.
43 + 4
0.24 f 0.02tT
46 & 6
196 & 18
19
Intact control animals
Values are means f SD. Symbols indicate values significantly different (P < 0.05, t-test) from values for: *muscles reinnervated for 56 months; **respective group of contralateral muscles; tcontralateral muscles of younger animals; stage-matched contralateral muscles; fintact control animals. “
