Neurobiologyof Aging, Vol. 17, No. 4, pp. 613~i17, 1996 Copyright © 1996 ElsevierScienceInc. Printed in the USA. All rights reserved 0197-4580/96 $15.00 + .00 ELSEVIER

S0197-4580(96)00004-8

Overloading a Muscle Does Not Alter the Rate of Motoneuronal Loss in Aged Rats KENRO KANDA, 1 KEN HASHIZUME, 2 TAKASHI MIWA AND YASUKO MIWA

Department of Central Nervous System, Tokyo Metropolitan Institute of Gerontology, 35-2, Sakaecho, Itabashi-Ku, Tokyo 173, Japan R e c e i v e d 17 April 1995; R e v i s e d 11 S e p t e m b e r 1995; A c c e p t e d 21 S e p t e m b e r 1995 KANDA, K., K. HASHIZUME, T. MIWA AND Y. MIWA. Overloadinga muscledoes not alter the rate of motoneuronal loss in aged rats. NEUROBIOL AGING 17(4) 613-617, 1996.--Effects of increased activity on neuronal cell death was investigated in the motor nuclei innervating normal and overloaded medial gaslrocnemius (MG) muscles of Fischer 344 rats. The MG muscle was overloaded by the unilateral surgical ablation of synergists at the age of 17 months (group A) or 24 months (group B). When the rats reached the age of 24 and 28 months (group A) or 30 months (group B), motoneurons innervating the MG muscle were labeled bilaterally with horseradish peroxidase injected into the MG nerve• The wet weight of the muscle on the operated side was consistently heavier than that of the contralateral, intact side. The number of labeled neurons decreased with advancing age, and there was no difference in the magnitude of decline found between motor nuclei innervating intact and hypertrophied muscles. Thus, overloading the MG muscle did not retard or accelerate the age-related loss of motoneurons innervating this muscle. These findings indicate that the causal relationship between motoneuronal activity and death with advancing age needs to be studied further. Neuronal cell death Rat

Activity level

Motoneuron

Leg muscle

N E U R O N A L cell loss is a major manifestation of brain aging. The progress of this phenomenon seems to depend on the type of neurons and/or regions c f the brain (1,5). We have previously demonstrated that there is a significant decrease in the number of motoneurons innervating the medial gastrocnemius muscle in the hindlimb of aged rats (27 months old and older), with most of the decrease occurring among the larger alpha-motoneurons (14), especially those innervatinlg fast twitch motor units (19)• On the contrary, no sign of decline in the number of ulnar motoneurons innervating forelimb muscles has been found up until at least 27 months of age (11,13). Factors that may cause these differential age effects are not clear. Gamma-motoneurons seems to be more active than alpha-motoneurons because of their low response threshold to both periphe, ral and central stimulation and of their functional role (8). It has been reported that motoneurons innervating slow twitch motor units are more active compared to those innervating fast twitch motor units (2,16). It is also likely that motoneurons innervating hand muscles are more active than those innervating hindlimb muscles because forelimbs are used when grooming and feeding. Accordingly, groups of surviving motoneurons tend to be relatively more active compared to those preferentially dropping out dunng aging. Furthermore, in our previous experiments, long-term ,;wimming exercise retarded age-related changes in motoneurons and peripheral nerves (12). It has also been reported that no or :minimal loss has been found in a certain group of neurons with higher activity in the vertebrate brain and in

Overloading

Compensatory hypertrophy

the invertebrate nervous system (29,30)• These findings suggest that one of the factors influencing neuronal cell survival (or death) in the aging nervous system is its activity level• In the present experiments, we examined whether or not increased activity per se actually postponed motoneuronal cell loss. To increase activity of a particular group of motoneurons with minimal effects on hormonal, cardiovascular, and respiratory systems, we removed synergistic muscles sparing a muscle innervated by the group of motoneurons concemed. A preliminary report of some of these results has appeared elsewhere (20). METHOD The initial surgery for ablation of muscles was performed in 44 male Fischer 344 rats raised in a specific pathogen-free colony at the age of 17 months (group A, 35 animals) or 24 months (group B, 9 animals). The rats were anesthetized with pentobarbital sodium (45 mg/kg, IP), and the lateral gastrocnemius, soleus, and plantaris muscles were removed unilaterally. The central cut ends of nerves supplying these muscle were ligated with a suture. The operating side was selected randomly, and the surgery performed aseptically. After recovery, the rats were raised under usual laboratory conditions (one to three rats per cage, lights on 0600-1800 h, 22 +_ 1°C, food and water ad lib, specific pathogen-free condition) until the final experiment was performed at the ages of 24 and 28 months old (group A) or 30 months old (group B). We lost 13

1 To whom requests for r~prints should be addressed. 2 Present address: Facult3~of Health and Sport Sciences, Osaka University. 1-17, Machikaneyama, Toyonaka, Osaka 560, Japan. 613

614

KANDA ET AL.

TABLE1 MUSCLE WET WEIGHT,MOTONEURONNUMBER,AND SOMASEE Motoneuron* Group (Body Weight, g)

Group A 17-month-old (468 _+3) 24-month-old (428 ± 29) 28-month-old (403 _+39) Group B 30-month-old (390 _+22)

Side (n)

Muscle Weight (mg)

Number

Soma Size (W2)

Intact (4)

725 _+31

132.8 _+ 7.8

869 _+ 57

Intact (4) Operated (4) Intact (13) Operated (13)

664 ± 25 862 _+19 564 _+26 749 _+54

127.8 + 7.0 126.5 _+10.5 115.5 _+ 8.0 117.2 + 6.5

734 _+122 731 _+124 850_+ 119 865 _+ 89

Intact (7) Operated (7)

505 _+39 643 _+35

114.3 + 9.2 112.0 _+11.3

Values are expressed as group means -+ SD. * Including both alpha- and gamma-motoneurons.

rats by natural death during this waiting period. Four rats, which were in a very poor general condition at the time of the final experiment, were not used. The remaining, apparently healthy animals were anesthetized with a mixture of halothane and nitrous oxide, and the nerve innervating the medial gastrocnemius (MG) muscle was freed from surrounding tissues that might include regenerating axons from the proximal ends of the cut nerves. A fresh solution of 40% horseradish peroxidase (HRP) was injected into the MG nerve bilaterally. Two days later, the rats were perfused transcardially with warmed physiological saline (300 ml) followed by a cooled fixative (700 ml of a mixture of 1.25% gultaraldehyde and 1% paraforrnaldehyde in phosphate buffer) under deep anesthesia with pentobarbital (50 mg/kg, IP). The lumbar spinal cord was removed and immersed in 30% sucrose solution overnight. The spinal segments were identified by the insertion of the dorsal roots and the cord was trimmed to include the L4 to S 1 segment. Serial sections (40 txm in thickness) were cut horizontally, and processed by the TMB method (23). In three cases, motoneurons innervating muscles other than the MG muscle were apparently labeled, and these were not used for further analysis. Consequently, data obtained from 4 24-month-old and 13 28-month-old rats in group A, 7 30-month-old rats in group B, and 2 additional rats in group A, which were sacrificed immediately at age of 17 months without muscle removal, were subjected to the final analysis. With the aid of montages obtained from each section, each labeled motoneuron was identified under a microscope, and counted [see (14) for more detailed methods]. At this stage, the experimenter (K.H.) was not aware of the operated and intact sides• The soma cross-sectional area of the identified motoneurons was measured using a tablet digitizer (HDG-1111B, Hitachi Seiko, Ltd., resolution: 0.05 ram) connected to a personal computer (Macintosh Ilfx), The outline of each cell body magnified by 533 times was traced on the tablet with the aid of a drawing tube attached to the microscope. The soma cross-sectional areas of labeled motoneurons were distributed bimodally, with an apparent transition at about 500 p~m2 between small and large groups. We assumed that large cells were alpha-motoneurons and that small cells were gamma-motoneurons (14). The MG muscles were also removed and weighed.

Statistical Analysis The mean values among the different groups were tested with a oneway ANOVA, or a repeated measures ANOVA; post hoc

comparisons were made with the Scheff6's procedure. The criterion for significant statements was p < 0.05. RESULTS

Within a week of synergistic muscle removal, the rats recovered and behaved normally with apparently unaffected use of the operated leg. It was impossible to identify the operated side when the rat walked in the cage just before sacrifice. No contracture and ankylosis was found in the operated leg, and the range of movement of knee and foot joints in the operated side was similar to that on the control side. There was a significant age effect on body weight, F(3, 20) = 3.900, p < 0.05. Body weight declined gradually with advancing age (Table 1). The 28- and 30-month-old rats were significantly lighter than 17-month-old rats. A significant between (age) effect, F(2, 19) = 43.06,p < 0.0001, and within (overloading) effect, F(1, 19) = 412.63, p < 0.0001, were found for wet weight of the MG muscle (Table 1). The muscles in 28- and 30-month-old rats were significantly lighter than those in 24- and 28-month-old rats, respectively, although the difference between 17- and 24-month-old

28 M

24 M (GroupA)

30 M (GroupB)

( G r o u p A)

900 •

J 1

OS

4

700 "

ii i!

I1) 600" '-I

E

"4 Q 0 0

EIr~ 500"

.1 400

I

I

I

I

i

i

i

i

i

i

i

Rats

i

i

I

i

[] I

I

Cl

= J

I

I

I

Int~t =lde

FIG. 1. Wet weight of the medial gastrocnemius muscle in the intact (open squares) and operated (filled circles) legs. Each pair of data connected by a vertical bar was obtained from the same rat. Increase in weight of the operated side was observed consistently in all animals. Even hypertrophied muscles showed a decrease in weight with advancing age.

MOTONEURONAL LOSS IN AGED RATS

¢/) ¢.,-

2--i ¢-

615

E

12o

-0 E

110

140 -

2

130

[]

¢,.. o

O

"6

i

Ul i-

o

::3

Z 100 0

O

Previous e:
•

Intact side of Group A

•

Intact side of Group B

130

I

I

I

I

I

I

5

10

15

20

25

30

I

•

•

120

E o

!i!T

I

[] []

[] • i "" •

110

e~ E

\

30 M (Group B)

28M (GroupA)

24 M (Group A)

140

[]

•

,o

[]

[]

lOO

Z

I I I I I I I I I I I I I

I I I I I I I

35

•

Rats

Operated side

[] ,n==,~

Age (months) FIG. 2. Mean number of labeled MG motoneurons innervating the intact leg in rats of various ages. MG motor nucleus in 28- and 30-month-old rats contained a significantly smaller number of motoneurons compared to that in 17-month-old rats. Open circle: data obtained in the previous experiments (14); filled circle: group A in the present experiment; filled square: group B in the present experhnent. Note that decrease in number of motoneurons is consistent over different series of experiments. The vertical bar indicates the standard dev!iation of the mean.

rats was not significant. The operated side was consistently heavier by 27-32% than that for the contralateral, intact side (Fig. 1). A repeated-measures ANOVA for the number of labeled motoneurons showed a significant between effect, F(2, 21) = 6.47, p < 0.01, power = 0.858 (T~ble 1 and Fig. 3). Compared to the data obtained in the previous experiments [134.5 _+3.2 for 5-month-old rats; 137.0 +_ 4.0 for 10 to 13-month-old rats; 121.4 _+ 3.9 for 26-month-old rats, and 107.2 _+6.9 for 31-month-old rats; (14)], a decrease in the number of MG motoneurons innervating the intact side seems to be a quite consistent phenomenon (Fig. 2). On the other hand, there was no within (overloading) effect, F(1, 21) = 0.06, p > 0.8, power = 0.05. The mean numbers of labeled motoneurons innervating hypertrophied muscles of the operated side were not different from the values of the intact side. Figure 3 shows the individual values. There was no significant within effect for soma sizes of both presumed alpha-motoneurons, F(1, 10) = 0.23, p > 0.6; power = 0.07, and presumed gamnm-motoneurons, F(1, 10) = 0.14, p > 0.7; power = 0.05, either (Fig. 4).

FIG. 3. Comparison of the number of labeled MG motoneurons between the intact (open squares) and operated (filled circles) sides. No consistent trend was found between the intact and operated sides.

muscles. We did not find, however, any difference in the soma cross-sectional area of MG motoneurons between the intact and hypertrophied sides. These contradictory results might be due to the different period of hypertrophy. Finkelstein et al. examined animals 6 to 8 weeks after ablation of synergistic muscles, while in the present experiment the final examination was performed 7 or 11 months after the operation. EMG amplitude of hypertrophied plantaris muscle decreased slightly between 30 and 35 days following ablation of its synergists (7). This late change might be due to an increase in tension production by motor units in hypertrophied muscle because individual muscle fibers enlarged (31). Thirty percent of increase in the MG muscle weight does not seem to be sufficient to compensate for loss of its three synergists. Load born by the MG muscle appears to remain higher without its synergists compared to the counterpart of the intact limb throughout the period of the present experiment. Contracture or ankylosis was not found in the operated leg. Furthermore, we did not notice any discernible difference in limb movement during locomotor activity between the intact and operated sides. It is very likely, therefore, that more motoneurons were activated on the operated side than on the intact side in addition to an increase in activity of low threshold motoneurons throughout the period of the present experiment.

o~-Motoneurons

DISCUSSION

A

~Motoneurons 500

1200 '

Activity Level of Motoneurons Innervating Overloaded Muscle 1000.

We showed in the present experiments that long-term overloading of the medial gastrocnemius muscle by removal of its synergistic muscles did not alter the magnitude of decline in the number of retrogradely labeled motoneurons innervating this muscle with advancing age, although this procedure induced a clear compensatory hypertrophy in the muscle. Gardiner et al. (7) have reported increased amplitude and integral of EMG bursts of rat plantaris during treadmill walking following surgical removal of its synergists. Morphological study of rat plantaris motoneurons following compensatory hypertrophy of the muscle has shown that those motoneurons become smaller (4). Conduction velocity of motor axons decreases after muscle hypertrophy (4). These findings suggest increased recruitment: and rate of discharge of motor units in the overloaded muscle. In the present experiment, compensatory hypertrophy was similarly induced in the MG muscle by removal of its synergists: the lateral gastrocnemius, soleus and plantaris

I

4OO

800 300 600 200 400 o u}

100

200 0 Side Age

0

--\o"~¢" 17 M

24 M

28 M

17M

24 M

28 M

FIG. 4. Soma cross-sectional area of labeled motoneurons. There was no difference between the intact (open column) and operated (shaded column) sides for both alpha- and gamma-motoneurons. Large alpha- and small gamma-motoneurons were distinguished from the frequency distribution and cumulative graph of soma size in individual rats.

616

KANDA ET AL.

Decrease in Number of Labeled Motoneurons Decrease in number of labeled motoneurons may reflect an actual decrease in number by cell death and/or deficit in axonal transport. Neurons whose axon has withdrawn beyond the tracer injection site (e.g., dying back phenomenon) are not labeled, either, if indeed those motoneurons actually exist. The rate of fast anterograde axonal transport becomes slow with age (6,22). Although little is known about fast retrograde axonal transport in aged rats, it is likely that it declines with age similarly to that of anterograde transport because similar mechanisms are proposed for both axonal transport (18). Frolkis et al. (6) reported that fast anterograde axonal transport in the sciatic nerve of 26- to 28month-old rats was 200 _+ 40 mm/24 h. The distance between the MG motor nucleus and the HRP injection site was about 100 mm in the present experiments. In our previous experiment in which HRP was injected into the MG muscle, labeled motoneurons were first observed after 13 h, and the number of labeled neurons reached the plateau level after 20 h (27). Thus, a 48-h waiting period from HRP injection and sacrifice seems to be sufficient to label motoneurons while avoiding the problem of degradation of HRP in the soma (3,25). Furthermore, analysis of labeled and unlabeled neurons in and near the medial gastrocnemius motor nucleus showed that the ratio of these two groups of neurons did not significantly alter with advancing age (Kanda, unpublished data). Therefore, a decrease in the number of labeled motoneurons seems to be due mainly to neuronal cell death. Neuronal Cell Loss In the present experiments we did not see any effect of increased neuronal activity on cell death. The followings are possible explanations and comments. First, beneficial effects of increased activity might be canceled by the concomitant hazardous effect. Increased neuronal activity may affect the aging process of the nervous system in completely different ways according to currently proposed aging theories. It has been reported that application of nerve growth factor (NGF) relieves cholinergic neurons in the basal forebrain of adult rats from their death following axonal lesions (15,21,32). This indicates that neuronal survival at least partly depends on the supply of trophic substance from the target organ even after development. It is, therefore, possible that neuronal cell death in the aging brain is caused by a deficit of trophic substance(s). Uptake and transport of trophic substance may depend on the neuronal activity (17). Increased activity may augment the supply of trophic substances that are transported retrogradely from the target organs and/or orthogradely through presynaptic neurons. Thus, increased activity may lengthen the life span of neurons. On the other hand, enhanced metabolism may increase the formation of free-radicals, which may damage various cell functions, and eventually cause cell death (10,28). That is, in-

creased activity may shorten the life span of neuronal cells. It is, therefore, possible that we did not see any change in survival rate of MG motoneurons in the present experiment because these two effects canceled each other. Secondly, the amount of increase in activity was not enough to produce detectable changes in cell survival. We did not directly monitor motoneuronal activity by recording EMG. Gardiner et al. (7) reported about a 100% increase in EMG activity of plantaris muscle whose synergists had been removed. We have previously shown that swimming exercise during a similar period (from 17 to 27 months of age) retarded motoneuronal loss with advancing age (12). The MG muscle are driven to their maximum contraction rate during the power stroke phase of each swim cycle (9,26). The mean number of strokes was about 2.1/s for all animals examined. Supposing that this continued throughout a swimming period of 30 min, the MG muscle would be driven more than 3600 times in each training day. The total number of discharges for motor units in the extensor digitorum longus and the soleus muscles of freely moving rats was reported to range from 2,660 to 495,800 per day (16). Thus, we would like to think that beneficial effects of physical exercise like swimming are due mainly to hormonal and other environmental changes induced by exercise rather than to increased motoneuronal activity per se. In conclusion, the present experiment could not provide direct evidence for the beneficial (or harmful) effect of increased activity per se on motoneuronal cell death. This might also mean that increased activity cannot be a potent antiaging factor for motoneurons in the rat spinal cord. Our previous findings of preferential loss of alpha-motoneurons, especially those belonging to fasttwitch motor units in the hindlimb (11,13,14,19), could not be explained solely by the difference in activity level. Mettling et al. (24) have shown in the chick embryo that brachial and lumbar motoneurons are intrinsically different in terms of sensitivity to trophic factors. The differential progress of cell loss among different groups of motoneurons might be due mainly to their intrinsic nature. Finally, we should pay attention to the type II error of a statistical test. Generally the power of a statistical test increases with an increase in sample size. The causal relationship between neuronal activity per se and death with advancing age needs to be studied further. ACKNOWLEDGEMENTS K.H. was at the Department of Kinesiology, TMIG when the experiments were performed. This study was supported in part by a Grant-in-Aid for Science Research from the Ministry of Education, Science and Culture, Japan (06670094), and a grant from the Life Science Promotion Foundation to K.K. We express our gratitude to Drs. S. Mizuno, K. Nakazato, and I. Tatsumi for their valuable advice on statistical methods. We also thank Ms. S. Asaki for her excellent histology technique.

REFERENCES

1. Brody, H. Organization of cerebral cortex. III. A study of aging in the human cerebral cortex. J. Comp. Neurol. 102:511-556; 1955. 2. Burke, R. E. Motor units: Anatomy, physiology, and functional organization. In: Handbook of physiology. The nervous system. II, part 1. Bethesda: Am. Physiol. Soc.; 1981:345--422. 3. Chu-Wang, I.W.; Oppenheim, R.W. Uptake, intra-axonal transport and fate of horseradish peroxidase in embryonic spinal neurons of the chick. J. Comp. Neurol. 193:753-756; 1980. 4. Finkelstein, D. I.; Lang, J. G.; Luff, A.R. Functional and structural changes of rat plantaris motoneurons following compensatory hypertrophy of the muscle. Anat. Rec. 229:129-137; 1991. 5. Hood, D.G.; Coleman, P.D. Neuron numbers and sizes in aging

6.

7. 8. 9.

brain: Comparisons of human, monkey, and rodent data. Neurobiol. Aging 9:453-463; 1988. Frolkis, V.V.; Tanin, S.A.; Marcinko, V.I.; Kulchitsky, O.K.; Yasechko, A. V. Axoplasmic transport of substances in motoneuronal axons of the spinal cord in old age. Mech. Ageing Dev. 29:19-28; 1985. Gardiner, P.; Michel, R.; Browman, C.; Noble, E. Increased EMG of rat plantaris during locomotion following surgical removal of its synergists. Brain Res. 380:114-12 ] ; 1986. Granit, R. The basis of motor control. London: Academic Press; 1970. Gruner, J. A.; Altman, J. Swimming in the rat. Exp. Brain Res. 40: 374-382; 1980.

MOTONEURONAL LOSS IN AGED RATS

10. Harman, D. The aging process. Proc. Natl. Acad. Sci. USA 78:71247128; 1981. 11. Hashizume, K,; Kanda, K Neuronal dropout is greater in hindlimb motor nuclei than in forelimb motor nuclei in aged rats. Neurosci. Lett. 113:267-269; 1990. 12. Hashizume, K.; Kanda, K. Effects of swimming exercise on agerelated changes in motoneurons and peripheral nerves in the rat. Soc. Neurosci. Abstr. 19:1742; 1993. 13. Hashizume, K.; Kanda, K. Differential effects of aging on motoneurons and peripheral nerves innervating the hindlimb and forelimb muscles of rats. Neurosci. Res. 22:189-196; 1995. 14. Hashizume, K.; Kanda, K.; Burke, R. E. Medial gastrocnemius motor nucleus in the rat: Age-related changes in the number and size of motoneurons. J. Comp. Neurol. 269:425-430; 1988. 15. Hefti, F. Nerve growth factor (NGF) promotes survival of septal cholinergic neurons after fimbrial transections. J. Neurosci. 6:2155-2162; 1986. 16. Hennig, R.; Lcmo, T. Fir.ng patterns of motor units in normal rats. Nature 314:164-166; 198.';. 17. Heuser, J. E.; Reese, T. S. Evidence for recycling of synaptic vesicle membrane during transmilter release at the frog neuromuscular junction. J. Cell Biol. 57:315-344; 1973. 18. Hirokawa, N. Axonal transport and the cytoskeleton. Curr. Opin. Neurobiol. 3:724-731; 1993. 19. Kanda, K.; Hashizume, K. Changes in properties of the medial gastrocnemius motor units in aging rat. J. Neurophysiol. 91:737-746; 1989. 20. Kanda, K.; Hashizume, K.; Iwata, K. Loss of motoneurons innervating normal and hypertrophied medial gastrocnemius muscle in aged rats. Soc. Neurosci. Abstr. 20::L711; 1994. 21. Kromer, L. F. Nerve growth factor treatment after brain injury prevents neuronal death. Science 235:214-216; 1987. 22. McMartin, D. N.; O'Conner, J. A. Effect of age on axoplasmic trans-

617

23.

24.

25.

26.

27.

28.

29. 30.

31.

32.

port of cholinesterase in rat sciatic nerves. Mech. Ageing Dev. 10: 241-248; 1979. Mesulam, M.-M. Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: A noncarcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents. J. Histochem. Cytochem. 26:106-117; 1986. Mettling, C.; Camu, W.; Henderson, C. E. Embryonic wing and leg motoneurons have intrinsically different survival properties. Development 118:1149-1156; 1993. Neuhuber, W.; Niedede, B. Differential labeling by horseradish peroxidase of small and large spinal neurons of rats. Neurosci. Lett. 20:131-134; 1980. Roy, R. R.; Hirota, W. K.; Kuehl, M.; Edgerton, V.R. Recruitment patterns in the rat hindlimb muscle during swimming. Brain Res. 337: 175-178; 1985. Sato, H.; Kanda, K.; Hashizume, K.; Yamada, J. Nerve stimulation does not necessarily enhance retrograde HRP labeling of rat motoneuron. Neurosci. Lett. 107:81-84; 1989. Strong, R.; Mattamal, M. B.; Andorn, A. C. Free radicals, the aging brain, and age-related neurodegenerative disorders. In: Yu, B. P., ed. Free radicals in aging. Boca Raton, FL: CRC Press; 1993:223-246. Swaab, D. F. Brain aging and Alzheimer's disease, "wear and tear" vs. "use it or lose it." Neurobiol. Aging 12:317-324; 1991. Van Der Woude, P. F.; Goudsmit, E.; Wierda, M.; et al. No vasopressin cell loss in the human hypothalamus in aging and Alzheimer's disease. Neurobiol. Aging 16:11-18; 1995. Walsh, J. V.; Burke, R. E.; Rymer, W. Z.; Tsairis, P. Effect of compensatory hypertrophy studied in individual motor units in medial gastrocnemius muscle of the cat. J. Neurophysiol. 41:496-508; 1978. Williams, L. R.; Varon, S.; Perterson, G. M. Continuous infusion of nerve growth factor prevents basal forebrain neuronal death after timbria foruix transection. Proc. Natl. Acad. Sci. USA 83:9231-9235; 1986.