Faculty Bibliography

A general feature of the developing nervous system is the activity-dependent rearrangement of genetically defined, synaptic connections. A parallel process occurs at the developing neuromuscular junction as activity-dependent synapse withdrawal reduces the initial polyneuronal innervation of individual muscle fibers to a mononeuronal innervation within the first few weeks after birth. Because members of the neurotrophin gene family influence motor neuron differentiation and survival, we examined whether or not they also influence synaptic rearrangements in neonatal muscles. We found that treatment with brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3), or neurotrophin-4/5 (NT-4/5) causes the transient retention of multiple synaptic contacts on neonatal myofibers. However, the combination of both electrophysiological and histological assays revealed that the majority of such supernumerary synaptic contacts are functionally inactive or 'silent.' There also occurs an increase in the number of retracting axons. Because BDNF mRNA is expressed in developing muscle and the trkB tyrosine kinase receptor for BDNF is expressed by neonatal motor neurons, our results suggest that BDNF may play an endogenous role in the refinement of synaptic connectivity that occurs in skeletal muscles after birth

We show that leukemia inhibitory factor (LIF) plays a physiological role in the programmed withdrawal of synapses from neonatal muscles. First, LIF mRNA is present in embryonic skeletal muscle and is developmentally regulated. We detect high levels of LIF mRNA at embryonic day 17 (E17) in mouse hind leg muscles. The content of LIF mRNA falls 10-fold between E17 and birth and then remains low in the neonate and adult. The decrease in LIF mRNA in skeletal muscle coincides with the end of secondary myogenesis and the completion of the adult number of myofibers. Second, treatment of the mouse tensor fascia latae (TFL), a superficial muscle of the hind leg, with LIF from birth (100 ng/day), transiently delays the withdrawal of excess inputs from polyneuronally innervated myofibers by approximately 3 days. The midpoint of the process is shifted from 7.5 +/- 10.2 +/- 0.6 days of age. LIF treatment delays synapse withdrawal by altering its timing without an appreciable effect on its rate. Third, in mice homozygous for a disruption of the LIF gene, the midpoint in the reduction of multiply innervated TFL myofibers occurs 1 day earlier, at 6.5 +/- 0.5 days of age. Muscle fiber number is unchanged in LIF null mice. Treatment with LIF does not alter the rate of neonatal growth, the number of muscle fibers in the TFL, or the reappearance of inputs that have been eliminated. Instead, LIF appears to delay maturation of the motor unit by transiently delaying the onset of synapse withdrawal. We hypothesize that this is a necessary component of a selective process that will operate simultaneously and equally on multiple, competing motor units

The mutation gly93-->ala of Cu,Zn superoxide dismutase (SOD) is found in patients with familial amyotrophic lateral sclerosis and causes motor neuron disease when expressed in transgenic mice. The progression of clinical and pathological disease was studied in a line of mice designated G1H. Clinical disease started at 91 +/- 14 days of age with fine shaking of the limbs, followed by paralysis and death by 136 +/- 7 days of age. Pathological changes begin by 37 days of age with vacuoles derived from swollen mitochondria accumulating in motor neurons. At the onset of clinical disease (90 days), significant death of somatic motor neurons innervating limb muscles has occurred; mice at end-stage disease (136 days) show up to 50% loss of cervical and lumbar motor neurons. However, neither thoracic nor cranial motor neurons show appreciable loss despite vacuolar changes. Autonomic motor neurons also are not affected. Mice that express wild-type human Cu,Zn SOD remain free of disease, indicating that mutations cause neuron loss by a gain-of-function. Thus, the age-dependent penetrance of motor neuron disease in this transgenic model is due to the gradual accumulation of pathological damage in select populations of cholinergic neurons

Mutations of human Cu,Zn superoxide dismutase (SOD) are found in about 20 percent of patients with familial amyotrophic lateral sclerosis (ALS). Expression of high levels of human SOD containing a substitution of glycine to alanine at position 93--a change that has little effect on enzyme activity--caused motor neuron disease in transgenic mice. The mice became paralyzed in one or more limbs as a result of motor neuron loss from the spinal cord and died by 5 to 6 months of age. The results show that dominant, gain-of-function mutations in SOD contribute to the pathogenesis of familial ALS

The ability of ciliary neurotrophic factor (CNTF) to induce sprouting by undamaged adult motor neurons was studied in gluteal muscles of adult ICR mice. Low doses of CNTF (0.02 mg kg-1 day-1) only induced sprouting in gluteus muscles that were beneath the site of injection, whereas high doses of CNTF (0.4-1.2 mg kg-1 day-1) acted systemically to induce motor neuron sprouting. We found little difference between the type or quality of sprouting induced by CNTF compared with muscle paralysis. Both stimuli induced sprouts of the same length, although muscle paralysis tended to induce more sprouts per end-plate. Paralysis also induced more nodal sprouting than did CNTF, but both were weaker stimuli for nodal sprouting than was partial denervation. In addition to its effects on motor neuron sprouting, high doses of CNTF induced loss of up to 36% of the body weight of treated mice. The substantial wasting caused by CNTF indicates that the factor has potent cachectic activity

Ciliary neurotrophic factor (CNTF) and basic fibroblast growth factor (bFGF) were tested for effects on sprouting by motor neurons innervating the adult mouse gluteus muscle. Factors were delivered by subcutaneous injection directly over the surface of the superior gluteus muscle once daily for 7 d and then end plates and axons were visualized by combined silver and cholinesterase staining. CNTF (500 ng daily) induced sprouting both from end plates and from the subset of nodes of Ranvier that are closest to the end plate. The effect of CNTF was potentiated twofold by coadministration of bFGF at doses of 2-20 ng daily, whereas treatment with bFGF alone failed to induce sprouting from either end plates or nodes of Ranvier. The sprouting stimulus delivered by the factors showed limited penetrance into the muscle and restricted lateral spread from the injection site