Cellular mechanisms in axonal growth

EXPERIMENTAL

NEUROLOGY

64,649-698

(1979)

Cellular Mechanisms

in Axonal Growth’

RICHARD P. VERAA,BERNICE GRAFSTEIN, AND ROBERT A.Ross Research Division. National Paraplegia Lauderhill.

Foundation 4440 NW 19th Street, Suite L-l 1 I. Florida, 33313; and the Departments of Physiology and Neurology, Cornell University Medical College, New York, New York 10021 Received

January

3. I978

CONTENTS AXONAL GROWTH AFTER LESIONS OF THE CENTRAL NERVOUS SYSTEM IN DEVELOPING ANIMALS Growth of Abnormal Neural Connections after Focal Brain Lesions: Constraining Factors and Functional Effects Conservation of Axon Terminal Arborization Terminal Competition Behavioral Consequences Plasticity and Rigidity in the Developing Visual System of Rats Interaxonal Influences Age Dependence Special Affinity Organization and Reorganization of Synaptic Inputs in the Developing Hippocampus Mechanisms of Afferent Pattern Development Reorganization of Synaptic Inputs Mechanism of Reorganization Abbreviations: CNS-central nervous system, LGv, LGd-ventral, dorsal lateral geniculate body, PT-pretectal cell groups, SC-superior colliculus, Con A-concanavalin A, ACh-acetylcholine, EPSPs-excitatory postsynaptic potentials, NGF-nerve growth factor. 1 This report is a summary of the proceedings of the 4th Biennial Conference on Regeneration in the Central Nervous System sponsored by the National Paraplegia Foundation, May 31-June 2, 1976. The conference was chaired by Dr. Bemice Grafstein and supported by the U.S. Veterans Administration Contract V101(134)P-442 and by the donorofthe Wakeman Award for Research in the Neurosciences. Publication of this report was funded by a grant from the Eastern Paralyzed Veterans Association. Administration and clerical support was provided in part by the Community Service Council of Broward County, Inc., a United Way Agency. Requests for reprints should be addressed to Mr. Veraa. 649 0014-4886/79/060649-50$02.00/O Copyright 0 1979 by Academic Press, Inc. All rights of reproduction in any form reserved.

650

VERAA,

GRAFSTEIN,

AND ROSS

PROPERTIES OF GROWING AXONS Observations on the Properties of Growing Axons Leading Edge of Growing Axons Uptake of Exogenous Protein Mitogenic Effect of Schwann Cells by Growing Neurites Basal Lamina Formation by Schwann Cells in Contact with Neurites Optic Nerve Axons of Xenopus Tadpoles and Jimpy Mice Axon Development in Jimpy Mice Glial Response during Wallerian Degeneration and Regeneration inxenopus Optic Nerve Role of Ependyma in the Patterning of Neuronal Pathways during Spinal Cord Regeneration in Amphibians and Reptiles Tail Regeneration in Amphibians and Reptiles Axonal Preservation in the Crayfish Axonal Changes Occurring after Contact and Interaction with Target Tissue Synaptic Junction Development: Presynaptic Changes Synaptic Junction Development: Postsynaptic Changes Synaptic Metamorphosis FORMATION OF CELL MEMBRANES Overview Organization and Dynamics of Cell Surface Receptors and the Effect of Cytoskeletal Drugs on Transmembrane Receptor Control Formation and Renewal of Membranes in Visual Cells Somatic Supply of Proteins and Lipids Membrane Replacement Molecular Replacement Turnover of Acetylcholine Receptors and Control of Chemosensitivity in Skeletal Muscles Degradation of Receptor Kinetics of Turnover REACTION OF THE NERVE CELL BODY TO AXONAL INJURY Cell Body Responses of Vertebrate Neurons to Axonal Injury Cytological and Cytochemical Studies of Axon Reaction in Rubral Neurons of the Cat RNA Content of Normal and Axotomized Rubral Neurons Incorporation of Tritiated Leucine by Axotomized Rubral Neurons Conclusion Reduced Accumulation of Neurotransmitter Synthesizing Enzymes: A Biochemical Marker for the Retrograde Reaction in Intrinsic Neurons of the Central Nervous System Retrograde Reaction in Noradrenergic Neurons Retrograde Reaction of Nigrostriatal Dopaminergic Neurons Changes in Synthetic Enzymes during Collateral Sprouting in Mesolimbic Dopaminergic Neurons Discussion Changes in Mammalian Sympathetic Neurons after Interruption of Their Axons Effects of Postganglionic Axotomy Effects of Colchicine Application Trophic Effect of Nerve Growth Factor Axon Growth and Functional Restoration after Optic Nerve Transection in the Goldfish ESTABLISHMENT OF NEURONAL CONNECTIONS Nerve-Muscle Junction Formation in Vitro Redistribution of Acetylcholine Receptors during Formation of Neuromuscular Connections in Cell Culture

AXONAL

GROWTH

MECHANISMS

Synapse Formation and Neuron Cell Death during Embryogenesis Synapse Formation Cell Death after Nerve Terminal Formation Ultrastructural Studies In Viva REGENERATION OF INTRINSIC SPINAL CORD NEURITES Enzyme Therapy for Spinal Cord Injury Enzyme Preparations Morphological Recovery Electrophysiological Recovery Functional Recovery CONCLUDING STATEMENT

LIST OF PARTICIPANTS Dr. Stanley H. Appel Professor of Medicine Division of Neurology Duke University Medical Center Durham, North Carolina 27710 Dr. Kevin D. Barron Neurology and Research Services Veterans Administration Hospital Albany, New York 12208 Dr. Gary Bennett Department of Anatomy McGill University Montreal, P.Q., H3C 3G1, Canada Dr. Amico Bignami Spinal Cord Injury Service (128) Veterans Administration Hospital West Roxbury, Massachusetts 02132 Dr. Mary A. B. Brazier Brain Research Institute University of California at Los Angeles Los Angeles, California 90024 Dr. Mary Bunge Department of Anatomy Washington University School of Medicine St. Louis, Missouri 63110 Dr. Richard P. Bunge Department of Anatomy Washington University School of Medicine St. Louis, Missouri 63110 Dr. Carmine D. Clemente Director, Brain Research Institute

The Center for the Health Sciences University of California Los Angeles, California 90024 Dr. Monroe Cohen Department of Physiology McGill University Montreal, P.Q., H3C 3G1, Canada Dr. C. D. Cone Veterans Administration Hospital Hampton, Virginia 23667 Mrs. Charlotte M. Cone Veterans Administration Hospital Hampton, Virginia 23667 Dr. Carl W. Cotman Department of Psychobiology University of California Irvine, California 92664 Dr. Doris Dahl Spinal Cord Injury Service (128) Veterans Administration Hospital West Roxbury, Massachusetts 02132 Dr. Douglas M. Fambrough Department of Embryology Carnegie Institute of Washington 115 West University Parkway Baltimore, Maryland 21210 Dr. Gerald D. Fischbach Department of Pharmacology Harvard Medical School 25 Shattuck Street Boston, Massachusetts 021 IS

651

652

VERAA,

GRAFSTEIN,

Dr. Murray Goldstein Director, Stroke & Trauma Programs National Institute of Neurological and Communicative Disorders and Stroke National Institutes of Health Bethesda, Maryland 20014 Dr. Bemice Grafstein Department of Physiology Cornell University Medical College 1300 York Avenue New York, New York 10021 Dr. Barth A. Green Vice President for Research National Paraplegia Foundation Deputy Chief of Spinal Cord Injury Veterans Administration Hospital of Miami Miami, Florida 33 125 Dr. Lloyd Guth Department of Anatomy University of Maryland School of Medicine Baltimore, Maryland 21201 Mr. Mike Hammond Interpreter for Dr. Levon A Matinian Fogarty International Center Department of Health, Education, and Welfare Bethesda, Maryland 20014 Dr. Edward A. Kravitz Department of Neurobiology Harvard Medical School 25 Shattuck Street Boston, Massachusetts 02115

AND ROSS

Dr. Levon A. Matinian Chief, Laboratory of Neuroendocrinology Orbeli Institute of Physiology Academy of Science Yerevan, Armenian S.S.R. Dr. Irvine G. McQuarrie Department of Physiology Cornell University Medical College 1300 York Avenue New York, New York 10021 Dr. Marion Murray Department of Anatomy Medical College of Pennsylvania 3300 Henry Avenue Philadelphia, Pennsylvania 19179 Dr. Thomas F. Newcomb Assistant Chief Medical Director for Research Development Veterans Administration Central Office 810 Vermont Avenue, N.W. Washington, D.C. 20420 Dr. Garth L. Nicolson Salk Institute of Biological Studies San Diego, California 92112 Dr. Don A. Olson President, National Paraplegia Foundation Director, Education and Training Rehabilitation Institute of Chicago 345 E. Superior Street Chicago, Illinois 6061 I

Dr. Ray D. Lund Department of Biological Structure University of Washington Medical School Seattle, Washington 98195

Dr. Raleigh K. Pettegrew Department of Biology Denison University Granville, Ohio 43023 Dr. Guillermo R. Pilar Biological Science Group University of Connecticut Storm, Connecticut 06268 Dr. .I. W. Prineas Neurology Service (127) Veterans Administration Hospital East Orange, New Jersey 07019

Dr. Emanuele Mannarino Spinal Cord Injury Service (128) Veterans Administration Central Office 810 Vermont Avenue, N.W. Washington, D.C. 20420

Dr. Dale Purves Department of Physiology and Biophysics Washington University Medical School 660 S. Euclid Avenue St. Louis, Missouri 63110

Dr. A. Robert Lieberman Department of Anatomy University College London Gower Street London, WCI, England

AXONAL

GROWTH

MECHANISMS

653

Dr. Donald Reis Department of Neurology Cornell University Medical College 1300 York Avenue New York, New York 10021

Dr. Donald B. Tower Director, National Institute of Neurological and Communicative Disorders and Stroke National Institutes of Health Bethesda, Maryland 20014

Dr. Robert A. Ross Department of Neurology Cornell University Medical College 1300 York Avenue New York, New York 10021

Dr. Silvio S. Varon Department of Biology University of California, San Diego La Jolla, California 92037

Dr. Gerald E. Schneider Department of Psychology Massachusetts Institute of Technology El044 Cambridge, Massachusetts 02139 Dr. Richard L. Sidman Department of Neuroscience Childrens’ Hospital Medical Center 300 Longwood Avenue Boston, Massachusetts 02115 Dr. Marcus Singer Director, Department of Anatomy Case Western Reserve University Cleveland, Ohio 44106 Dr. Arnold J. Smolen Department of Anatomy Medical College of Pennsylvania 3300 Henry Avenue Philadelphia, Pennsylvania 19129

Mr. Richard P. Veraa Director, Research Division National Paraplegia Foundation 4440 N.W. 19th Street Suite L- 111 Lauderhill, Florida 333 13 Dr. Henry deF. Webster Laboratory of Neuropathology and Neuroanatomical Sciences National Institute of Neurological and Communicative Disorders and Stroke National Institutes of Health Bethesda, Maryland 20014 Dr. William F. Windle Professor Emeritus Denison University Granville, Ohio 43023 Dr. Richard W. Young Jules Stein Eye Institute University of California School of Medicine Los Angeles, California 90024

AXONAL GROWTH AFTER LESIONS OF THE CENTRAL NERVOUS SYSTEM IN DEVELOPING ANIMALS In young animals lesions of the central nervous system (CNS) may lead to axonal growth in the remaining neural tissue and to establishment of new synaptic connections. Although these compensatory changes may result in the establishment of functionally inappropriate connections, the rules governing this kind of neuroplasticity are important in revealing the nature of the stimuli that may induce neurons to form new connections and also the factors that may limit axonal growth and the establishment of synaptic terminals.

654

VERAA,

GRAFSTEIN,

AND ROSS

Growth of Abnormal Neural Connections after Focal Brain Lesions: Constraining Factors and Functional Effect? After lesions of the superior colliculus (121, 122, 124), retina (37,38), or lateral olfactory tract (26,27) of the Syrian hamster, abnormal growth patterns occur only if the damage is inflicted before specific ages. In the Syrian hamster the axons of the main optic tract terminate in the contralateral ventral and dorsal portions of the lateral geniculate body (LGv and LGd), the pretectal cell groups (PT), and the superficial gray layer of the superior colliculus (SC). In many animals, there is also a very small patch of retinofugal terminals in the lateral posterior nucleus (LP). However, the LP normally gets its visual input mainly from axons of superficial gray layer neurons of the SC; the PT and part of LGv also receive input from this part of the SC. If the upper layers of the SC are unilaterally ablated at birth-when optic tract axons are just reaching the tectumoptic axon terminals increase in the LGv, LP, and the pretectal region, in the same regions deprived of their input from the SC (Fig. 1). The abnormal projection to the LP is particularly striking because most cells in the LP normally receive no direct retinal projection. Furthermore, optic tract axons cross the tectal midline to occupy sites in the medial contralateral SC. The latter phenomenon is enhanced if the ipsilateral eye is removed at the time of SC ablation (Fig. 1). In that event, however,fewer terminals form in the LP. Two factors, in addition to undetermined chemical specificities, appear to be of special importance in determining when and where anomalous connections will form: conservation of axon terminal arborization, and competition for available terminal sites. Conservation of Axon Terminal Arborization. This hypothesis states that during development in the CNS, certain axonal populations tend to conserve the total amount of their terminal arborization such that if growth in one part of the axon is limited or exaggerated, compensatory changes result in another part of the axonal tree (28, 122). The above observations support this principle, although quantitative data on precise numbers of synapses are not available [however, see (97) for related data on the superior cervical ganglion]. This compensatory mechanism is manifested both in increased aberrant arborization and in decreases in numbers of normally expected terminals under different conditions. Thus, optic tract axons which fail to find adequate terminal space in the superior colliculus show “compensatory sprouting” in the thalamic nucleus lateralis posterior (122), and axons in the lateral olfactory tract which develop anomalous col* This section is based on the presentation by Dr. Gerald E. Schneider.

AXONAL

GROWTH MECHANISMS

655

FIG. 1. Upper dorsal view of the rostral brain stem of a normal adult hamster. The courses of the left and right optic tracts are indicated, with terminal areas depicted by open and filled circles. respectively. Lower left-similar view of brain stem of an adult hamster which has undergone ablation of the superficial layers of the right superior colliculus on the day of birth. The right optic tract has developed abnormal connections. Lower right-similar view in a case with a right colliculus lesion as in previous case, but the right eye was also removed at birth. For simplicity, the small ipsilateral projection of the retina is omitted. Abbreviations: DTN-dorsal terminal nucleus of the accessory optic system, LGd-dorsal nucleus of the lateral geniculate body, LGv-ventral nucleus of the lateral geniculate body, LP-nucleus lateralis posterior, PT-pretectal area, SC-superior colliculus, IC-inferior colliculus.

lateral branches and proximally show a “compensatory stunting” of their distal arbors (27). Terminal Competition. Axons tend to invade vacated terminal space and compete with other axons for occupancy. An example of this was presented by Schneider in which, after early unilateral lesions of the SC, when retinofugal axons from the two eyes invaded the remaining SC, the

656

VERAA,

GRAFSTEIN,

AND ROSS

distribution of axons from the left eye was constrained by the distribution of axons from the right eye. This phenomenon may depend on axonaxonal repelling interactions. (The principle of competition will be discussed further in the next section.) The two factors indicated above can result in the redirection of growing axons into systems from which they are normally completely separated. Thus, the optic tract can be caused to grow into the medial geniculate body of the auditory pathway (55, 122), and the lateral olfactory tract can grow into the neocortex (27). Functional correlates of these two most bizarre anomalies are unknown. However, Schneider went on to describe evidence of the functional impact of abnormal retinotectal connections. Behavioral Consequences. The development of anomalous axonal connections can have striking functional consequences, which may be adaptive or maladaptive (121, 122, 124). Thus, after partial or complete ablation of the midbrain tissue of the SC which normally receives optic tract input, all or part of the corresponding retina develops connections with the residual tectum (35,36,53, 125). Visually elicited orienting movements dependent on the SC are found in hamsters with such early lesions, although the accuracy and form of the movements may be abnormal. After unilateral ablation of superficial tectal layers at birth, together with removal of the ipsilateral eye, axons from the remaining eye cross over the region of damage, with some termination in surviving deep layers of the SC, and they form a large abnormal decussation at the tectal midline to terminate in the undamaged SC-on the wrong side of the brain (Fig. 1). Hamsters with such a connection display turning in the wrong direction in response to stimuli in part of the visual field. This abnormal behavior can be abolished in the mature animal by surgical transection of the anomalous recrossing bundle (123, 124). Plasticity

and Rigidity

in the Developing

Visual System of Rats3

The visual system of the rat shows a capacity for reordered growth patterns after perinatal lesions. Systematic alterations resulting from such lesions were studied to understand some of the factors influencing axonal growth in vivo (77). The main principles derived from this work and the results supporting them are outlined below. Znteraxonal Znfruences. The patterns of distribution of an axon population are influenced by other axon populations with which it interacts. For example, if one eye is removed before 5 to 10 days postnatal, optic axons of the other eye project bilaterally, innervating the visual centers on both sides of the brain (76). A local lesion of one eye at birth results in aberrant 3 This section is based on the presentation by Dr. Ray D. Lund.

AXONAL

GROWTH

MECHANISMS

657

growth patterns of remaining axons of the same eye and of axons of the other eye, thus resulting in an abnormal retinotectal map (80,81). Finally, corticotectal axons innervating the ipsilateral SC cross the midline and, in addition, innervate the contralateral SC if the normal input to that colliculus is removed at early postnatal times (70). Age Dependence. The amount of aberrant growth is reduced with increased age and varies with the pathway involved. Aberrant ipsilateral retinotectal axons can no longer be shown if the initial lesion is made later than 10 days postnatal (78); crossed corticotectal axons can be elicited by lesions made to 15 to 20 days postnatal (85). Axons can grow through damaged regions, and the capacity to do this is also restricted by age. This was shown with transtectal slits in fetal and neonatal rats and with unilateral tectal lesions in fetal rats (83). Special A&&y. This principle states that there are selective affinities between certain groups of nerve cells during development and that this factor influences, in part, the pattern of connections formed within the growing CNS. This was shown in the developing visual system of the rat by transplanting pieces of tectum from 15 to 16-day-old fetal rats into the region of the SC of newborn rats (79). These transplants receive axons from host pathways (e.g., retina, visual, and somatosensory cortices), and electron microscopic studies showed that synapses from the retinal fibers are formed within the appropriate layer of the transplant. Moreover, these axons may follow a circuitous route within the transplant to innervate those regions. It is suggested that this process is selective, for when fetal cortical tissue is transplanted into the SC, no innervation is seen (52). From these results, it appears that during postnatal development axons lose the capacity for major reordered growth and that this is due more to the properties of the axons themselves than to the region in which they are situated (because different afferents to the SC behave differently in this respect). Furthermore, it seems that the course an axon takes depends heavily on the environment in which it is growing and as such it is open to considerable modification. On the other hand, the pattern of final connections is more related to special affinities existing between certain groups of nerve cells and as such is more rigidly ordered. Organization and Reorganization of Synaptic Inputs in the Developing Hippocamp The synaptic inputs in the dentate gyrus of the rat are organized in a precise topography. Commissural-associational fibers make contact exclusively on the inner third (27%) of granule cell dendrites and entorhinal 4 This section is based on the presentation by Dr. Carl W. Cotman.

658

VERAA,

GRAFSTEIN,

AND

ROSS

and septal fibers on the outer two-thirds (73%). The boundary between entorhinal and commissural-associational fibers shows less than a lo-pm overlap [see (23) for a complete description of hippocampal circuitry]. Mechanisms of Afferent Pattern Development. Afferent fibers may be laminated at the earliest stages, or they may be mixed initially and sorted during the development process. To determine the organization of afferents on immature granule cells, pathways were traced by autoradiography in animals 4 days old, in which the number of synapses is only 15% of its final total and the dendrites are 25% of their final height. Cotman found that at 4 days postnatal, the granule cells already had their afferent fibers organized in laminae. Commissural fibers occupied the inner half, and entorhinal fibers occupied the outer half. As the dendrites grew, the afferent fibers moved upward, preserving their lamination pattern. Thus the transition from immature to mature lamination pattern appears to result from differential growth of the dendritic tree (20, 75). Reorganization ofSynaptic Znputs. A unilateral lesion of the entorhinal cortex triggers a dramatic reorganization of the synaptic inputs. When entorhinal input is removed at 11 days of age, the commissural and associational inputs expand to fill almost the entire dendritic tree. Septal afferent fibers and those from the contralateral entorhinal cortex become constrained to the outer one-fourth to one-sixth of the dendritic field. This region is largely unoccupied by the commissural-associational fibers. Thus, as in normal animals, the afferents are laminated, but now the boundary is shifted. The nature of the reorganization of septal inputs depends on the developmental state, and when synaptogenesis is complete (25 days), the proliferation of septal afferent fibers takes place over a larger part of the entorhinal zone (2 1- 23). Mechanism of Reorganization. These data illustrate the interdependence of afferent fiber development on the final organization of synapses. The complementary organization of afferents suggests that a competition between them might account for the pattern. For example, the septal afferents may be forced to the outer edge of the dendrites by the commissural afferents. However, a combined commissural and entorhinal lesion does not affect the abnormal development of septal inputs. Thus, in this case, simple competition does not account for the reorganization. A clue to a possible mechanism can be inferred from the nature of the entorhinal lesion that can elicit a proliferation of septal afferents fibers. Whether or not proliferation is seen depends on the portion of the entorhinal cortex removed. A lesion of the lateral entorhinal cortex, which removes input from the outer part of the dendrites, induces septal proliferation, but deafferentation of other portions of the dendritic tree does not. It is inferred that since dendrites are growing most rapidly at their tips, this portion of surface

AXONAL

GROWTH

659

MECHANISMS

remains unspecified and receptive to abnormal innervation. Other older portions of the surface may be specified already (86, 87). These data illustrate the dynamics of the organization and reorganization of developing circuitry. The state of the postsynaptic surface as well as the growth of new afferent fibers determines how inputs are organized. With the cessation of development, synaptogenesis becomes more fixed but is still modifiable in the dentate gyrus. PROPERTIES

OF GROWING

AXONS

During axonal growth, whether in embryonic development or after axonal injury, in viva or in vitro, the rate of advance of the growing axon and the direction in which it grows are determined by intrinsic properties of the axon tip and properties of the environment in which it grows. The growing axon tip shows a characteristic content of organelles and a plasma membrane that is highly specialized in both structure and function. In particular, the plasma membrane of the axon tip is specialized for the uptake of extracellular materials from its surroundings. This presumably enables the axon to sample the environment of the growing tip, possibly leading to alterations in the growing neuron in response to conditions in this environment. The contribution of glial cells and ependymal cells to axonal outgrowth may therefore not be limited to mechanical support, but may involve active transfer of some growth-stimulating and differentiationpromoting factors. Conversely, the growing axon may have an important influence on the metabolic activity of the supporting cells and target cells that it encounters. Observations

on the Properties

of Growing Axons5

Leading Edge of Growing Axons. The tips of growing axons appear as enlargements from which ruffling membranes or moving filopodia emanate. These terminal expansions or growth cones, as studied in cultures of sympathetic neurons, contain a variety of organelles, such as agranular endoplasmic reticulum, vacuoles, vesicles (some of which are dense-cored or coated), lysosomal structures, and mitochondria as well as filaments and microtubules (12). The regions of the cone exhibiting movement are filled primarily with a microfilamentous network. Uptake of Exogenous Protein. The uptake of exogenous proteins occurs much more readily in the growth cone, particularly at the leading edge near filopodia, than along the axon shaft or at the cell body. This has been studied by adding 2 to 5 mg peroxidase or 90 mg ferritin per milliliter cul5 This section

is based on the presentation

by Dr.

Mary

B. Bunge.

660

VERAA,

GRAFSTEIN,

AND

ROSS

ture medium to cultured sympathetic neurons. Ultrastructural studies of early stages in this process reveal that uptake into the growth cone occurs predominantly via tubular and uncoated vesicular structures that invaginate from the plasmalemma (13). Evidence suggests that the internal ends of the tubular channels may enlarge to form cup-shaped bodies which become lysosomal bodies, in particular multivesicular bodies. Label is not found free within the cytosol or within the bulk of the branched membranous structures considered to be agranular reticulum. Some label is found within coated vesicles. This endocytotic process may be a mechanism by which the axon samples the environment at its leading edge. Alternatively it may allow the axon to maintain a unique type of plasmalemma at its growing or searching end by contributing to a mechanism by which surface membrane components are recycled. Mitogenic Effect on Schwann Cells by Growing Neurites. This and the following study were made possible by the development of a technique for isolating pure populations of normal neurons, fibroblasts, and Schwann cells from cultures of fetal rat dorsal root ganglia (142). In cultures of isolated Schwann cells devoid of other cellular populations, the Schwann cells were found to exist in a quiescent or nondividing state for more than a month. When nerve cells are added to these cultures of quiescent Schwann cells, the Schwann cells not only divide but also in time ensheath the axons and form myelin, indicating that their functioning has not been impaired. This mitogenic effect of growing neurites was demonstrated by autoradiography after the addition of a tritiated thymidine label. In a bed of quiescent Schwann cells, the percentage of label in the Schwann cell nuclei is less than 2%. When neurons are added and nerve fibers enter the Schwann cell bed, then at least 90% of those Schwann cell nuclei are labeled with tritiated thymidine (143). In cultures where fibroblasts also were present, this phenomenon was noted only in the Schwann cell population. Basal Lamina Formation by Schwann Cells in Contact with Neurites. An important aspect of neuron-Schwann cell interaction is the formation of basal lamina. Basal lamina has not been observed in cultures of neurons alone, fibroblasts alone, or combined neurons and fibroblasts. In cultures of neurons and Schwann cells, however, basal lamina is consistently found on the surface of the Schwann cell (14, 141). The basal lamina becomes thicker with increasing age of the culture. It should be noted that in these preparations the Schwann cell appears on morphological grounds to be active. Rough endoplasmic reticulum and Golgi apparatus components are prominent. The elements of granular reticulum contain flocculent material. If the neuronal somata are removed from neuron-Schwann cell cul-

AXONAL

GROWTH

661

MECHANISMS

tures, the neurites rapidly degenerate and in the ensuing weeks the basal lamina may disappear. Whether or not the basal lamina disappears is dependent in part on the age of the culture; the older the culture the more likely the basal lamina is to be retained, at least for a few weeks. The basal lamina often disappears if the neurons are removed from a 3-week-old culture, and in these cultures the Schwann cells appear much less active morphologically. There are fewer organelles in the cytoplasm, ribosomes appear pale and less numerous, and cisterns of rough endoplasmic reticulum appear to have collapsed (141). If nerve cells are returned to these cultures, basal lamina is reformed by the Schwann cells and the morphology of the Schwann cell is again one of an actively synthesizing cell. It seems appropriate to present these new observations on Schwann cell cultures, not only because of the opportunities to better understand the aspects of Schwann cell functions discussed, but also because of the known role of the Schwann cell in guiding axon growth in peripheral nerve regeneration. Referring to this aspect of Schwann cell function in peripheral nerve repair, Ramon y Cajal (113) termed bands of Schwann cells upon which axons grow a tutorial cordon and considered the presence of this permissive and trophic terrain to be a key element in the relative success of peripheral nerve regeneration. Tissue culture methods are now being used to determine if these cells might serve a similar function for CNS axons. Optic Nerve Axons of Xenopus Tadpoles and Jimpy Mice6 Axon Development in Jimpy Mice. In the CNS of mammals and amphibia, glial cells (in particular oligodendrocytes) have a significant effect on the development and growth of axons they surround and myelinate (101). An example of this is seen in studies of changes in the axon diameters and axon-glial relationships in the optic nerve of developing normal mice and Jimpy, a mutant strain of mice with abnormal oligodendrocytes (139). At 9 days after birth, the optic nerve axons of both groups of mice are essentially normal in terms of their axon diameters. However, at 18 days, the oligodendrocytes in the Jimpy mice are greatly reduced in number and many have an abnormal appearance. Moreover, the percentage of axons undergoing myelination at this time is greatly reduced in Jimpy mice, and these axons are smaller in diameter than those seen in normal mice. Preliminary studies of freeze-fracture preparations suggest that abnormalities in the surface membrane of the optic nerve axons may be associated with the retarded growth and abnormal glial relationship that are seen in Jimpy optic axons. Gliai Response during Wallerian Degeneration and Regeneration in 6 This section

is based on the presentation

by Dr. Henry

deF.

Webster.

662

VERAA,

GRAFSTEIN,

AND

ROSS

Xenopus Optic Nerve. The responses of glial cells have also been examined during Wallerian degeneration and regeneration of transected Xenopus tadpole optic nerves. Unlike the mammalian CNS, the rate of degeneration occurs much more quickly in the tadpole nervous system. This may be due in part to the absence or paucity of scar formation at the site of the lesion. Also, astrocytes become phagocytic and remove degenerating fibers as well as forming a frame.work of radial processes which aligns the regenerating optic nerve fibers. Oligodendrocytes do not become phagocytic; they remain quiescent with their processes retracted until the arrival of the regenerating axons, when they proliferate and begin to remyelinate the optic axons (115). Role of Ependyma in the Patterning of Neuronal Pathways during Spinal Cord Regeneration in Amphibians and Reptiles’ Tail Regeneration in Amphibians and Reptiles. The lizard, Anolis carolinensis, regrows its tail after amputation and, within the new struc-

ture, a spinal cord is formed consisting of an ependymal-glial tube and descending nerve fibers (32). There is no formation of new nerve cells or of oligodendrocytes, but tracts of unmyelinated central axons appear which grow from the amputation stump to the tail tip. However, the functional effectiveness of the regenerated central neurites is uncertain for there are no apparent terminal connections for these axons. The most impressive fact in regrowth of the cord is the activity of the ependyma in the formation of the new spinal cord and in the patterning of the descending axons. Ependyma1 cells at the cut surface multiply and extend into the regrowing tail as a tube, closed at its caudal end. The ependymal cells have radial processes, and the spaces between these processes are strictly aligned fore-and-aft to form continuous longitudinal channels. The regenerating descending fibers grow within these channels, which are made in advance of the growing axons and thus apparently provide a pathway to attract and direct the fiber growth. Among the Amphibia the newt shows even more remarkable powers to regrow the spinal cord, whether in a regenerating tail or following ablation of a cord sector without amputation of a body part (33,95). Here, too, the ependyma plays an essential role in reconstructing the cord. It bridges the gap in the damaged cord and extends into the regenerating tail. Not only does the ependyma of the newt provide channels to guide growing axons, but in addition it later gives rise to new neurons and glia. Therefore, the mechanism for successful spinal cord regeneration in these vertebrates appears to lie in the ependyma, which apparently retains in the adult the developmental capacities of the embryo. 7 This section

is based

on the presentation

by Dr.

Marcus

Singer.

AXONAL

GROWTH

MECHANISMS

663

Axonal Preservation in the Crayfish. Singer and Salpeter (129) presented

evidence in 1966 that the Schwann myelin sheath transports substances into the axon. Further evidence for a pathway through the SchmidtLantermann clefts was reported in a series of publications using tracer substances (59, 128), and extensive transfer of materials from glia to axons has now been demonstrated by other workers (40, 67). A notable example of glial influence on axonal metabolism is the failure of the transected crayfish giant motor axon to undergo Wallerian degeneration. Hoy et al. (51) observed that motor function of the crayfish claw is reestablished sooner than could be expected by regeneration of the giant motor axon. They postulated that, shortly after transection, the cut ends of the axons must fuse by “primary intention” and in this way reestablish the continuity and integrity of the axon. Ultrastructural analysis of this phenomenon (93, 94) showed that such healing by primary intention does not occur, but that the severed segment of the axon and its muscle endings persist indefinitely, and that the proximal stump of the cut axon regenerates many fine processes which grow along the severed segment and make contact with it. The severed segment itself does not sprout. The persistence of the severed segment is apparently related to the activity of surrounding glia which becomes intimately applied to the severed segment, thickens greatly, and apparently provides sustenance for the maintenance of the severed segment. Persistence of the separated axon may also be related to trophic activity of the proximal motor sprouts on the distal segment of the cut axon. When the glial mechanism of axonal maintenance is understood, it is conceivable that the principle may be applied to prevent Wallerian degeneration in the nervous system of vertebrates. Axonal Changes Occurring after Contact and Interaction with Target Tissueg

The addition of explants of rat embryonic thoracic spinal cord to dissociated sympathetic neurons has permitted the study of morphological changes in neurite growth cones and target cells during synapse formation (114). It was established that in this culture system neurites from the spinal cord establish cholinergic synapses with the cocultured autonomic neurons (57). Upon initial contact between the outgrowing neurites from the spinal cord and the isolated autonomic neurons, morphological changes are seen both presynaptically and postsynaptically. Synaptic Junction Development: Presynaptic Changes. Changes in the growth cone begin upon arrival at the target neuron (114). The growth cone filopodia become extensively applied to the plasmalemma of the sympaw This section is based on the presentation by Dr. Richard P. Bunge

664

VERAA,

GRAFSTEIN,

AND

ROSS

thetic neuron and manifest numerous pun&ate regions in which the opposing plasma membranes are separated by only 70 to 80 A. In this transmission electron microscopic study, however, no gap junctions or tight junctions were observed at this early phase of contact. Within 2 to 3 days, the typical appearance of the growth cone is altered with the loss of filopodia and the appearance of synaptic vesicles, which become clustered opposite the postsynaptic density. As the synapse matures, synaptic vesicles increase in number, cleft width increases, and the lysosomal structures and branched membranous reticulum, which are abundant within the active growth cone, become greatly diminished. Synaptic Junction Development: Postsynaptic Changes. Alterations occur within the target cells subsequent to contact by the presynaptic neurite (114). The Golgi apparatus hypertrophies and an increased number of coated vesicles is observed in its vicinity. Similar vesicles are seen in continuity with the plasma membrane near the site of presynaptic contact, resulting in alocalized “thickening” at the region of the postsynaptic membrane. By this method of study, the development of the postsynaptic density appears to be the primary element of synapse formation, with the presynaptic changes coming later. Tracer studies with peroxidase and ferritin confirm that the coated vesicles arise from the Golgi region and travel to the somal surface. These vesicles are associated with the postsynaptic density at all stages of maturation of the synapse. Further studies are needed to determine if these Golgi-derived vesicles are incorporated specifically into selected regions of the cell membrane, and if this is the mechanism for confining certain characteristics of the neuronal surface to the synaptic region. Synaptic Metamorphosis. Recent studies of sympathetic neurons in tissue culture demonstrate that the synapses formed between the principal neurons become transformed, in time, from having noradrenergic properties to synapses characteristic of cholinergic neurons [for reviews see (15, 99)]. During the first 2 weeks in culture, an extensive neuritic network is formed between the principal neurons of the rat superior cervical ganglion, and within this network synapses are formed. These neurons exhibit properties of adrenergic neurons in that they (a) synthesize and store catecholamines, (b) selectively take up and release norepinephrine, (c) are destroyed by guanethidine, a specific adrenergic toxin, and(d) show the appearance of dense-cored vesicles in preparations fixed with potassium permanganate. At 1 week in culture, the neurons contain a very low activity of choline acetyltransferase, the enzyme necessary for biosynthesis of acetylcholine. The synapses are apparently nonfunctional at this stage, for high concentrations of catecholamines fail to elicit membrane potential changes in these neurons (96). At 7 to 8 weeks in culture, dramatic changes are seen (54). At that time,

AXONAL

GROWTH

MECHANISMS

665

choline acetyltransferase activity and the ability of the culture to synthesize acetylcholine has greatly increased. Although quantitative cytochemical studies indicate that synapses seen after 1 week in culture contain predominantly (80%) dense-core vesicles, after 8 weeks the majority of synapses contain clear vesicles. In addition, an increasing number of functional, excitatory, nicotinic cholinergic synapses can be identified electrophysiologically. At intermediate times during this metamorphosis, synaptic terminals have been shown to contain half dense-core and half clear vesicles-with an eventual sorting out with time so that there are some synapses that have primarily dense-core vesicles and others (the majority) that have primarily clear vesicles. Patterson and his colleagues [for review see (loo)] examined the different sets of in vitro conditions which influence this transformation and showed that it is promoted by the addition of nonneuronal cells to the culture, or by the addition of conditioned media, i.e., media in which various cell types (particularly heart muscle cells) have been grown. For example, sympathetic neurons grown under certain conditions will remain adrenergic. However, addition of suitable conditioned medium will cause these cells to become cholinergic within several weeks, a time course similar to that seen with the addition of nonneuronal cells to the culture. If, as these observations suggest, the adrenergic neuron is gradually becoming cholinergic in certain tissue culture environments, then these autonomic neurons exhibit a type of plasticity generally not recognized. It would be critical to know if other types of neurons are capable of a type of plasticity which allows a shift in neurotransmitter production in response to altered environmental conditions. FORMATION

OF CELL

MEMBRANES

One essential feature in axonal regeneration is the reformation of the cell membrane. This involves special problems for the nerve cell because the area of membrane is so large in proportion to the volume of the cell body (where much of the membrane synthesis occurs) and also because the membrane shows an unusual degree of regional specialization at the nerve terminal. Nevertheless, the basic process is undoubtedly similar to that in other cells, and more is known about how the process operates in cells other than neurons, e.g., visual receptor cells and muscle cells. Overviews

Tremendous advances have been made during the last few years in terms of the biochemical elucidation of membrane structure and function. As our understanding of membranes has grown we have looked less and less at the y This section is based on the presentation by Dr. Gary Bennett.

666

VERAA,

GRAFSTEIN,

AND

ROSS

structural level-microscopically-and looked increasingly at the biochemical level. As observed in the electron microscope a few years ago, the plasma membrane was seen as a simple three-layer structure thought to consist of the lipid bilayers sandwiched between two layers of protein. Many histochemical studies at that time showed the presence of carbohydrate material on the external surface of virtually all plasma membranes, and it was thought that this was probably a separate layer which constituted a cell coat (112). Now it is seen that this layer represents the carbohydrate-rich hydrophilic ends of membrane glycoproteins and glycolipids; many of the glycoproteins extend through the membrane by means of a hydrophobic portion attached to another hydrophilic segment which extends into the cytoplasm (88). These proteins, or combinations of them, appear as the intramembranous particles seen in freeze-fracture studies (88). A relationship was shown between these structures that further emphasizes the concept that the membrane is essentially a two-dimensional viscous solution in which lipids can flow fairly rapidly and in which proteins can also move laterally through the fluid layer. Thus we see that the membrane is not as static as was believed, but is really a very dynamic structure (88). Recent studies described below showed that the plasma membrane is also intimately associated with the cytoplasmic microfilaments and microtubules (Nicolson, this conference), so that we can see now that the plasma membrane is simply part of a more complex organelle. These studies revealed mechanisms whereby information can be transferred from the outside to the inside of the cell and perhaps explain various phenomena such as the promotion of locomotion by various external factors. They also shed light on the mechanisms described earlier of the induction of Schwann cell proliferation (M. Bunge, this conference). Another dynamic feature of the plasma membrane that has become clear is the continual and fairly rapid turnover of membrane components-not only proteins, but also glycoproteins (6) and phospholipids. Organization and Dynamics of Cefl Surface Receptors and the Effect of Cytoskeletal Drugs on Transmembrane Receptor ControP”

The cytoskeletal structure of the cell is a key element in maintaining its morphologic shape and is probably extremely important in such phenomena as cell mobility and neural sprouting. If the cytoskeletal system is disrupted, modified, or destroyed, cells tend to assume an equilibrium configuration-a spherical shape. They do not show highly differentiated or specialized surface regions (41). I0 This section

is based

on the presentation

by Dr.

Garth

L. Nicolson.

AXONAL

GROWTH

MECHANISMS

667

Recent observations demonstrate that the cytoskeletal system-microtubules and microfilaments in particular-are also important in the exertion of transmembrane control of the display and mobility of cell surface receptor molecules. The surfaces of mammalian cells are bounded by fluid, dynamic plasma membranes that function in a variety of important cellular processes (130). The receptors on cell surfaces are free to diffuse laterally in the plane of the membrane unless their mobility is restrained by planar associations, lipid domain formation, peripheral protein restraints, or membrane-associated cytoskeletal restraints (8890). This last mechanism of cellular control of surface receptor mobility was examined by the use of colchicine, which interferes with microfilaments, cytochalasin B, which interferes with microtubules, and by the use of local anesthetic agents. Tertiary amine local anesthetic agents can facilitate cell surface receptor redistribution and cell agglutination induced by antibodies and lectins (89, 107). Concanavalin A (Con A)-induced agglutination of mouse BALB/3T3 cells and redistribution of Con A receptors into clusters and patches is enhanced by low concentrations of dibucaine (2 x low3 M), tetracaine (5 X lob4 M), lidocaine (5 x 10V3 M), or procaine (10 x 10V3 M) (98). However, those drugs when used at similar concentrations, inhibit ligandinduced capping-but not clusteringof immunoglobulin receptors on mouse B lymphocytes (107). It is proposed that these effects on receptor mobility in fibroblasts and lymphocytes result from the action of anesthetic agents on membrane-associated cytoskeletal assemblies involved in the transmembrane control of cell surface receptor mobility (89). The effects of local anesthetic agents on ligand-induced redistribution can be duplicated by treating cells with colchicine together with cytochalasin B. In addition, electron microscopic examination of anesthetic agent-treated BALB/3T3 cells reveals that these drugs rapidly and reversibly induce cell rounding concomitant with losses in membrane-associated cytoskeletal organization and/or plasma membrane attachment. Both microtubule and micofilament organizations are disrupted in cells treated with an anesthetic agent (91). Tertiary amine local anesthetic agents are known to interact with membrane anionic lipids resulting in displacement of membrane-bound calcium (98, 126). It was proposed that calcium displacement from membranes affects disengagement of membrane-associated microfilament systems, and the resulting increase in cytoplasmic calcium causes depolymerization of cellular microtubules (89). The possible involvement of membrane-bound and cellular calcium in cytoskeletal integrity was investigated by utilizing calcium-specific ionophores (A23 187 and X537A) to manipulate cellular calcium concentration. In media containing calcium and cytochalasin B, calcium-specific ionophores caused effects similar to tertiary amine local anesthetics, whereas potassium-specific ionophores had no effect (106),

668

VERAA,

GRAFSTEIN,

AND

ROSS

suggesting that at least some of the properties of local anesthetic agents can be explained by their indirect action on cytoskeletal assemblies. The implications of these observations in neurological systems and particularly in receptor aggregation during synapse formation recommend themselves to similar studies in neural development. Formation

and Renewal

of Membranes

in Visual Cells”

An important issue in the growth-and regrowth-of nerve fibers is the availability of appropriate materials for their construction. It was shown (140) that growing as well as mature axons are constructed of materials synthesized in the neuronal somata and transported along the axon. A system in which a similar phenomenon occurs as a continuing process is that of the visual cells of the retina, which has been intensively studied (147). The outer segment of rod visual cells in the retina of vertebrate animals consists of a tall stack of closely packed disc-shape membranes enveloped by the outer membranes of the cell. The predominant protein in the membranes is opsin, the light-absorbing visual pigment protein. The membranes of the outer segment-which is the cell’s light-detecting deviceare continually renewed, providing a model system for studying the formation of cell membranes. The outer segment membranes lack the metabolic machinery to generate the molecules required for their own renovation. These have to be supplied from elsewhere in the cell. The only connection between the outer segment membrane system and the remainder of the cell is a hollow cilium which serves as a supply channel. Somatic Supply of Proteins and Lipids. Visual cells in vertebrate animals are remarkably compartmentalized. Almost all the proteins, including opsin, are synthesized in the myoid region of the inner segment, which is the cell’s manufacturing center for all major classes of proteins but not for nucleic acids. Those proteins destined for the outer segment are somehow “sorted out” from those intended for other sites of utilization. At the level of the ribosomes, a side chain of carbohydrate begins to be added to opsin. First glucosamine, then mannose, is added. Next, the new proteins begin to migrate toward the outer segment, most or all of them passing through the Golgi apparatus, where the last of the carbohydrate-additional glucosamine -is added to the opsin. The proteins then bypass the mitochondria clustered in the outer portion of the inner segment (the ellipsoid region) and traverse the connecting cilium to reach the outer segment. Phospholipids are also produced in the myoid region of the inner segment and are delivered to the outer segment along the same ciliary supply line, probably protected from the aqueous environment by carrier proteins. ” This section

is based

on the presentation

by Dr.

Richard

W. Young.

AXONAL

GROWTH

MECHANISMS

669

This picture of the synthesis and subsequent fate of opsin was made by autoradiographic tracing of labeled amino acid precursors, mannose, and glucosamine. Phospholipids were traced by radioactive choline, glycerol, and fatty acids (7, 8, 145, 147). Membrane Replacement. At the base of the outer segment, new discs are repeatedly formed by infolding of the outer cell membrane. New molecules arriving from the inner segment are inserted into the membrane in this zone of growth. The carbohydrate side chain is thought to assure the proper orientation of opsin in the membrane. The bilayer of phospholipids in the outer segment membranes has the fluid characteristics of olive oil. Opsin floats freely in the lipid, rotating incessantly, and diffusing laterally. Some of the opsin inserted into the growing membrane infoldings diffuses into the outer membrane. However, the new discs soon lose their attachment to the outer cell membrane, preventing further exchange of their proteins by diffusion out of the discs. Repeated formation of new discs at the base of the stack gradually displaces older discs along the outer segment. To balance this continuous process, groups of discs are intermittently shed from the apical end of the outer segment by a process involving infolding of the outer membrane and separation of membrane-enclosed packets of old discs as the infolding cell membranes join, thus pinching off the packet of old discs. Shedding of discs from the outer ends of the rods occurs according to a daily rhythm, synchronized by the light cycle (68, 148). The detached packets of membranes are phagocytosed and digested by the adjacent layer of cells (the pigment epithelium). This represents a process of renewal by membrane replacement. In the retina of the rhesus monkey, and probably in the human retina as well, each rod assembles 80 to 90 new membranous discs daily, and completely replaces its stack of membranes every 9 to 12 days, that is, roughly 35 times each year. In removing the debris of shed membranes, the pigment epithelial cells are probably the most active phagocytes in the body (144). Molecular Replacement. The membranes are also renewed by a continuous molecular replacement, a process which differs with the type of molecule. Opsin is not replaced in the membrane, but some of the proteins associated with the membranes are renewed as the discs move along the outer segment. New phospholipid molecules delivered from the myoid region repeatedly take the place of older molecules of the same kind in the membranes. Furthermore, some of the constituents of the membrane phospholipids, such as fatty acids and choline (but not the glycerol “backbone”), are continually renewed by exchange within the membrane (146). The purpose of this process of balanced formation and degradation is probably preventive maintenance. Parts are replaced before they wear out. The gradual accumulation of defects provoked by a variety of environ-

670

VERAA,

GRAFSTEIN,

AND ROSS

mental hazards as well as by functional activity is prevented by the continual reconstruction of the membranes, which is only one aspect of the general renewal of the cell’s constituents. No matter what its age, the visual cell is largely an assemblage of new molecules. Turnover of Acetylcholine Receptors and Control of Chemosensitivity in Skeletal Muscles’2

Studies of the turnover of acetylcholine (ACh) receptors have provided useful strategies for investigating metabolic aspects of plasma membrane proteins. The ACh receptor is a glycoprotein of which 3 to 6% of molecular weight is carbohydrate, the remainder being polypeptide chains. The molecule is composed of 5 to 10 subunits held together by strong noncovalent bonds. Portions of its surface have a great affinity for lipids, permitting firm attachment of the molecule in the plasma membrane with its binding sites extending outward and its inner portion extending into the cytoplasm. a-Bungarotoxin is a small protein which attaches avidly to ACh binding sites, forming an extremely stable complex with the receptor. Because a-bungarotoxin is such a strong ligand for the receptors, radioactive derivatives of it are extremely useful in studying the metabolism of receptors. If chick muscle cells in culture are allowed to interact with unlabeled cY-bungarotoxin, all receptor sites exposed on the membrane are bound and unavailable for further toxin binding. As new receptors are subsequently metabolized and enter the plasma membrane, the rate of their production and incorporation can be determined by challenging the cells again at various time intervals with radioactively labeled a-bungarotoxin, which will thus label new receptors only. Receptor appearance was found to be a linear process occurring at a rate dependent on the metabolism of the cell (29). Uncoupling oxidative phosphorylation with dinitrophenol rapidly stops the incorporation of new receptor into plasma membrane, indicating that the process has an energy requirement. Blocking protein biosynthesis has no immediate effect on the incorporation mechanism, but incorporation decreases as the previously synthesized internal receptor pool is depleted. Reversal of such a block after 1 h results in a hiatus of receptor incorporation 2 to 3 h later. It is proposed that all precursor receptor molecules within the cell are membrane-bound. This is inferred from the identification of receptors by labeled a-bungarotoxin in the membrane fraction of cell homogenates. A more direct strategy for the study of receptor metabolism involves feeding the cells with heavy amino acids prepared with the heavy isotopes ‘* This section is based on the presentation by Dr. Douglas M. Fambrough.

AXONAL

GROWTH

MECHANISMS

671

carbon 13, nitrogen 15, and deuterium. These amino acids are exactly the same size and are metabolized in exactly the same way as normal amino acids, but are heavier per unit volume. In these experiments, existing surface receptors are labeled with lz5iodine-labeled cr-bungarotoxin, and the cells are fed with heavy amino acids and, at various times thereafter, are dissolved and challenged as described above but, in this case with L31iodine-labeled a-bungarotoxin. By this means, all receptors are radioactively labeled, but the receptor-toxin complexes formed since application of the heavy amino acids are slightly denser and can be separated by centrifugation. This enables one to study the kinetics of these processes (3 1). Degradation of Receptor. The stability of the toxin-receptor complex also makes a-bungarotoxin an important tool in the study of the ultimate destruction of receptor molecules. After radioactively iodinated toxin is bound to ACh receptors on muscle cells, there is a process by which the toxin-receptor complexes are catabolized in the lysosomal system. When the a-bungarotoxin is degraded, iodotyrosine is created. This is a compound to which cells are extremely permeable and which cannot be metabolized by the cells. Radioactive iodotyrosine thus produced provides a measure of the degradation rate ofreceptor by the cell, which these findings indicate is relatively constant at approximately 3% per hour. The degradation rate measured in pulse-chase experiments with heavy amino acids is very slightly faster than the rate estimated by the iodinated toxin method. Kinetics of Turnover. From these and other observations (29-31, 34), one can derive a mathematical expression for the overall kinetics of receptor turnover in this sytem as follows: Letting k = rate constant for receptor degradation, r = number of new receptors appearing in the surface membrane per hour, RO = total number of surface receptors present at time t = 0, then the number of receptors present at t = 0 and still present at a later time t is equal to R,,edkt and the number of new receptors formed in time t is equal to [r/k](l - eekl). Thus the total number of receptors at time t, Rt, is equal to Roe-“’

As t approaches R, = r/k.

infinity,

+ [r/k](l

- e-kf).

the first term goes to zero as does emkt, leaving

The metabolism of skeletal muscle ACh receptor is drastically affected by innervation. Electrical stimulation of cultured or denervated adult skeletal muscle mimics innervation in causing a drastic decline in the number of ACh receptors. The effect of electrical stimulation is mediated through a block in receptor biosynthesis or incorporation into plasma membrane; such stimulation causes little change in the degradation rate.

672

VERAA,

GRAFSTEIN,

AND ROSS

The above description is for extrajunctional ACh receptors. ACh receptors at neuromuscular junctions are organized into very dense clusters in the postsynaptic membrane. These receptor-rich regions of membrane seem to be metabolically stable and will persist for months after denervation. The turnover rate for junctional ACh receptors is not known but is slow compared to that for extrajunctional receptors. This difference in turnover rate may be of central importance in the accumulation of receptors at neuromuscular junctions during development and in the maintenance of localized chemosensitivity at the adult neuromuscular junction. REACTION

OF THE NERVE

CELL

BODY TO AXONAL

INJURY

Materials required for axonal outgrowth are mainly produced by synthesis in the nerve cell body. In many neurons, axonal injury results in marked changes in the morphologic and metabolic properties of the cell body, and some of those changes may be critical in determining whether or not regeneration of the severed axon can occur. The changes in the injured neuron may extend to the intact axon collaterals and may be accompanied by changes in the presynaptic boutons, so that axotomy may result in altered function of neurons not directly affected by the original lesion. Cell Body Responses of Vertebrate Neurons to Axonal Znjury’3

Most vertebrate neurons, both in the central and peripheral nervous system, display a series of changes (termed “axon reaction”) after interruption of their axons (43,73,74). The retrograde changes in the cell bodies of peripheral neurons are associated with the regenerative growth of the injured axons, and the significant features of the response are therefore of obvious interest in the context of the relatively much poorer regenerative growth of the axons of intrinsic neurons of the CNS. The metabolic changes that occur in injured neurons were much studied in recent years. Changes in RNA and protein metabolism were examined chiefly in relation to motor neurons (138). In rodent hypoglossal motor neurons, axotomy is followed by a burst of nucleolar RNA synthesis, an increase in the amount of nucleolar RNA, an increase in the rate of transfer of newly synthesized RNA from nucleus to cytoplasm and in cytoplasmic RNA content, cytoplasmic protein synthesis, and cytoplasmic protein content (134, 135, 138). After proximal lesions (close to the cell body) the response to axotomy occurs more rapidly than after distal injury (74), but the duration of the period of increased nuclear RNA synthesis is much shorter. This results in only a slight increase in cytoplasmic RNA content I3 This section is based on the presentation by Dr. A. R. Lieberman.

AXONAL

GROWTH

MECHANISMS

673

compared, for example, to a doubling of RNA content with more distal lesions (135, 138). Thus when axons are interrupted close to their cell bodies, chromatolysis appears to be more severe, and cell death may be more extensive (74). However, it is still not clear why the response should be so short when the greatest length of axon must be replaced. After its initiation, the metabolic response proceeds to its completion and cytoplasmic RNA content decreases to normal values irrespective of whether or not the regenerative efforts of the axon are successful. If, however, the axon fails to make an appropriate neuromuscular contact, RNA content will decrease to less than normal (135, 138), and cells will atrophy and may die (74). Concurrent with the anabolic response, the synthesis and cellulifugal passage of transmitters and of transmitter-related molecules is inhibited (73). The classical chromatolytic reaction, in which the granular endoplasmic reticulum becomes fragmented and dispersed and some membrane components are lost, may be closely related to this selective inhibition (73). Mobilization of the lysosome system, marked by an early and large increase in the number, size, and internal complexity of hydrolase-containing cytoplasmic bodies, is probably related in part to the consequent selective degradation of the protein-synthesizing machinery (73). More widespread autodigestive processes almost certainly occur as a result of other aspects of perikaryal remodelling such as dendritic retraction (see below) and changes in the proximal axon and its collaterals (58, 117). The morphologically detectable perikaryal changes may be more an expression of selective inhibition of protein synthesis and cellular remodelling than of enhanced protein synthesis and may not, therefore, serve as a particularly useful guide to the metabolic potency of the retrograde reaction in a given neuron. Dramatic changes also occur at the cell surface: axotomized craniospinal motor neurons, sympathetic ganglion cells, and other neurons, suffer a partial and selective loss of axon terminals synapsing on their somata and dendrites (9, 17, 18, 110, 132, 138). The loss of synapses involves the active participation of microglial cells (9), which proliferate (134, 138), and of astrocytes, which hypertrophy (136, 138). The partial deafferentation results in retraction of the dendritic tree (133) and is associated with loss of receptors from the soma and dendrites (11,39, 119) and with a depressed responsiveness of the cell to its normal inputs (10,60, 110). Axon terminals and receptors are restored and the full dendritic tree is reestablished only if the neuron reestablishes functional neuromuscular contact (110, 138). The loss of synaptic contacts and receptors from axotomized motor neurons could be viewed as a phenomenon parallel to the inhibition of transmitter synthesis; these changes indicate a shift in the functional status

674

VERAA,

GRAFSTEIN,

AND

ROSS

of the cell from its normal state to one in which receptive and transmitting functions take second place to structural renewal. A crucial question is do the responses of the hypoglossal motor neuron constitute an adequate model for other axotomized neurons? There are clear indications that the metabolic responses of sensory ganglion cells, and of goldfish retinal ganglional cells (whose axons regenerate after interruption) are similar to those of motor neurons (16,45,56,84,137). But what of intrinsic neurons of the CNS, whose responses to axotomy are apparently extremely variable (73), and whose metabolic responses have been scarcely investigated? Some intrinsic CNS neurons apparently lack the ability to respond to axotomy by making appropriate enzymatic responses, or by increasing RNA and protein synthesis, findings of obvious relevance to the debate on CNS regeneration (3-5). Other injured intrinsic neurons, however, notably the noradrenergic neurons of the locus ceruleus and the dopaminergic neurons of the substantia nigra, display biochemical changes similar to those of peripheral adrenergic neurons (116, 118), and it is probably premature to conclude that all intrinsic CNS neurons are inherently incapable of generating the necessary metabolic response. Cytological

and Cytochemical Studies of Axon Reaction in Rubral Neurons of the Carl4

RNA and protein metabolism was studied in axotomized mammalian intrinsic neurons. Axon reaction was produced in one red nucleus of the cat by left-side lateral funiculotomy at C2, which interrupted the completely crossed rubrospinal tract. RNA Content of Normal and Axotomized Rubral Neurons. Microspectrophotometric measurements of RNA were made in histological sections of the caudal portion of the red nucleus using the azure B method of Shea (127) after DNase digestion (5). Chromatolysis was apparent in large neurons of the reacting (rightside) red nucleus 6 days postoperatively. At 9 to 14 days survival, a majority of nerve cells of all sizes were affected, and swollen chromophobic profiles of large neurons appeared. Nucleolar vacuoles and basophilic bodies were frequent 2 to 14 days after axotomy. The red nucleus ipsilateral to surgery (left-side, nonreacting) was unaffected except that there appeared appreciable numbers of nucleolar bodies 6 days postoperatively. At 28 to 60 days after funiculotomy there was partial restoration of Nissl substance, but virtually all neurons of the right red nucleus were shrunken and pale. I4 This

section

is based on the presentation

by Dr.

Kevin

D. Barron.

AXONALGROWTHMECHANISMS

675

In neurons of all sizes, statistically significant nuclear atrophy was present in the reacting red nucleus 2 through 60 days postoperatively. Nucleolar atrophy occurred 14 to 60 days after funiculotomy. Neither nuclear nor nucleolar enlargement was found at any survival time. Cytoplasmic atrophy first was documented 14 days after operation, preceded by a small (18.6%) cytoplasmic enlargement at 9 days. Cells that underwent enlargement were all intensely chromatolytic. Very few neurons disappeared completely, because even with 60 days survival, when neuronal atrophy and pallor were conspicuous, there was only a 4% difference between the right and left red nuclei. Nine days after axotomy, large neurons showed a statistically significant 24% depletion of cytoplasmic RNA concentration, which signified a decrease in RNA content, even in neurons that had increased in volume. A few of the swollen large rubral neurons were so chromophobic as to preclude microspectrophotometric measurement of their RNA concentration, and often lacked a definable nucleus or nucleolus. Fourteen days after funiculotomy there was further reduction (32%) in cytoplasmic RNA concentration of the large neurons. Cytoplasmic atrophy was now apparent. At 28 and 60 days survival, even further depletion of cytoplasmic RNA was seen in the large neurons and small and medium-size neurons also exhibited loss of cytoplasmic RNA. Nucleolar RNA depletion first appeared in large neurons at 28 days, but in neurons of the middle size range, nucleolar RNA values were unchanged at 28 and 60 days postoperatively. The smallest nerve cells of the reacting red nucleus showed definite depletion of nucleolar RNA 2,28, and 60 days postoperatively. In summary, during axon reaction rubral neurons became depleted of cytoplasmic and nucleolar RNA. Cytoplasmic depletion appeared to preclude nucleolar change in large neurons. Nuclear and nucleolar atrophy also accompanied rubral axon reaction. Cytoplasmic enlargement appeared transiently 9 days postoperatively, but profound cytoplasmic atrophy then supervened. The cytological and cytochemical observations indicated a neuronal response to axotomy which was regressive in character. In contrast, in companion cytological and cytochemical studies of cervical motor neurons responding to brachial plexectomy, a significant increase in nucleolar RNA content was found 6 days postoperatively and was associated with nucleolar enlargement (Barron, K. D., et ul., unpublished data). Incorporation

of Tritiated

Leucine

by Axotomized

Rubral

Neurons.

Fourteen kittens, newly weaned and 7 to 10 weeks of age, were injected with [H3]leucine 0.5 to 24 h before killing, 1 to 30 days after unilateral high

676

VERAA,

GRAFSTEIN,

AND

ROSS

cervical rubrospinal tractotomy (3), and histoautoradiographs of the red nuclei were prepared. Axon reaction wh evident histologically 24 h after surgery but had already begun to disappear 10 to 30 days postoperatively. A transient cytoplasmic enlargement was observed at early survival times, most marked at 3 days postoperatively. However, nucleolar, nuclear, and cytoplasmic shrinkage were otherwise the hallmarks of rubral axon reaction in these young animals. Incorporation of labeled amino acid was reduced, suggesting that the axon reaction in the red nucleus of newlyweaned kittens is associated with diminished protein synthesis (3). In contrast, axotomized cervical motoneurons of kittens of like age showed normal or increased incorporation of amino acid (24). Conclusion. Definition of the nature of axon reaction in intrinsic neurons is crucial to elucidation of the reasons for the known failure of most interrupted mammalian CNS fiber tracts to regenerate. It may be suggested, from the results reported for retrograde reaction in the feline red nucleus, that the frequent failure of axonal repair in the mammalian CNS lies in the innately regressive nature of the axon reaction of some intrinsic mammalian neurons. Reduced Accumulation of Neurotransmitter Synthesizing Enzymes: Biochemical Marker for the Retrograde Reaction in Intrinsic Neurons of the Central Nervous SystemI

A

Techniques have been developed for the immunocytochemical localization of neurotransmitter synthesizing enzymes within the cell. Studies have been focused on the dynamic changes induced in cells undergoing reversible somatic axon reaction. The central catecholaminergic and mesolimbic dopaminergic neurons were selected for study because they provide a useful preparation in which to examine some of the biochemical and morphological processes associated with the regeneration of intrinsic neurons of the CNS. Neurotransmitter enzymes were studied as representative proteins produced by chemically identifiable cells whose changes might be expected to yield insights into changes in patterns of protein tumover in reactive cells. Retrograde Reaction in Noradrenergic Neurons. The response to lesions of the ascending axons of noradrenergic neurons was examined after electrolytic lesions of the lateral hypothalamus. Activity and amounts of tyrosine hydroxylase and dopamine-p-hydroxylase in noradrenergic cell bodies of the nucleus locus ceruleus were measured. The response was triphasic: within the first 24 to 48 h, the activities of tyrosine hydroxylase and dopamine-/3-hydroxylase increased to nearly 150% of normal values; I5 This

section

is based

on the presentation

by Dr.

Donald

Reis.

AXONAL

GROWTH

MECHANISMS

677

by days 5 through 7 activity decreased to 50 to 60% of control values, remaining there until 3 to 4 weeks, when complete recovery occurred. The reduction in enzyme activity represents reduced amounts of enzyme protein, the magnitude of which depends on the proximity of the lesions to the locus ceruleus. Retrograde Reaction of Nigrostriatal Dopaminergic Neurons. The retrograde reaction in neurons of the substantia nigra consisted of an elevation of tyrosine hydroxylase activity during the first 24 to 48 h followed by a rapid and permanent decrease to 40 to 50% of control values. The permanent decrease of tyrosine hydroxylase in the substantia nigra can be demonstrated histologically to be the result of death of dopaminergic neurons. However, lesions placed distally within the terminal fields of the striatum resulted in a reversible reduction in the activity and amount of tyrosine hydroxylase lasting from days 7 to 14, roughly parallel to the reversible reduction in transmitter enzymes in central noradrenergic neurons. Changes in Synthetic Enzymes during Collateral Sprouting in Mesolimbic Dopaminergic Neurons. Dopaminergic neurons of the mesolimbic

system projecting from the mesencephalic cell bodies of the A10 group into the olfactory tubercle were studied to determine whether or not such neurons could undergo collateral sprouting in response to the removal of a nondopaminergic input into the olfactory tubercle, and if such growth might be associated with changes in the activity of tyrosine hydroxylase within the parent cell bodies of the A10 group. A major nondopaminergic input into one olfactory tubercle was removed by ablating the ipsilateral olfactory bulb. Within the olfactory tubercle, tyrosine hydroxylase increased 10 to 14 days after ablation of the olfactory bulb, gradually reaching a maximum of 125% of control values at 21 days and remaining elevated for at least 60 days. This increase in tyrosine hydroxylase could not be attributed to changes in its activity in noradrenergic terminals, as no change in dopamine-@-hydroxylase activity was observed. By 30 days after bulbectomy, specific uptake of tritiated dopamine into synaptosomes of the olfactory tubercle was increased to 12 1% of control values. At that time, immunocytochemical staining of tyrosine hydroxylase by the peroxidase-antiperoxidase method demonstrated by light microscopy a striking increase in the intensity and number of stained fibers within the olfactory tubercle. Within dopaminergic cell bodies of the A10 group, tyrosine hydroxylase activity was transiently elevated to 121% of control values by the 4th day, returning to the control by 10 days, Thus, the reactive changes oftyrosine hydroxylase during collateral sprouting differ from those initiated by axonal lesions and those associated with regenerative sprouting. They

678

VERAA,

GRAFSTEIN,

AND

ROSS

differ in that during collateral sprouting enzyme activity is not decreased, but rather shows a brief increase in the 1st week after the ablation of nondopaminergic input. Discussion. These studies have several general implications. First, they indicate that a reversible reduction in the accumulation of enzymes subserving transmitter biosynthesis appears to be a general concomitant of the retrograde reaction in at least two types of CNS neurons, the noradrenergic and dopaminergic neuronal systems, provided that the lesions are not so close to the cell body as to produce retrograde cell death. A second implication relates to the reduced accumulation of enzyme protein which occurs in response to axonal injury. This reduced accumulation may reflect a reduced synthesis of enzyme protein at a time when total protein biosynthesis is increasing. This suggests that during the retrograde reaction in response to axonal injury there may be a reordering of priorities of protein biosynthesis favoring proteins required for reconstitution of axonal surface at the expense of those required for neurotransmission. It would seem that the neuron is being informed of the extent of its cell surface -which requires specific proteins-and the cell body is somehow aware of the extent of its synaptic territory. In the case of axotomy, a great deal of territory is lost, and the nerve cell, having to reconstitute its membranes, may essentially shut down its production of the enzyme to repair its surfaces. In collateral sprouting, on the other hand, the presumed amount of the terminal field which is altered is relatively small with regard to the size of the nerve cell and the need for producing just enough nutrients and transmitter enzymes to fill these terminals is subserved by a small burst of activity. In any case, it seems clear that some of the enzymes involved in neurotransmission are regulated by the integrity of the axonal structure and the extent of its surface. Changes in Mammalian Interruption

Sympathetic Neurons of Their Axons I6

after

By analogy with events at the neuromuscular junction of vertebrates, one might expect the synaptic input of a neuron to exert some measure of control over postsynaptic properties (110). For neurons, however, a quite different mechanism of control is possible: because nerve cells send their axons to contact other nerve, muscle, or gland cells, the axonal extension to, and functional connection with, a target might also exert some influence on neuronal properties. Such a control mechanism, which is suggested by the dramatic changes that nerve cells undergo after interruption of their axons (111) has been explicitly investigated in a series of largely bioI6 This section is based on the presentation by Dr. Dale Purves.

AXONAL

GROWTH

MECHANISMS

679

chemical experiments (138). In the work summarized here the influence of the postganglionic axon extension to the periphery was examined with intracellular recordings and electron microscopy in the superior cervical ganglion of adult guinea pigs. Effects of Postganglionic Axotomy. Within 4 to 7 days after crushing the major postganglionic branches of the superior cervical ganglion, the amplitude of excitatory postsynaptic potentials (EPSPs) recorded in ganglion cells in response to maximal preganglionic nerve stimulation declines markedly (108). Counts of synapses per unit area of electron microscopic sections showed a commensurate reduction in synaptic profiles, suggesting that the major reason for synaptic depression after axotomy is a loss of synapses from the dendrites of ganglion cells. Within 1 month, postganglionic axons begin to reinnervate their peripheral targets as evidenced by sympathetic effects such as pupillary dilation in response to stimulation. Coincident with this, surviving ganglion cells begin to recover their synapses, and EPSP amplitudes return to normal within about 3 months. If, however, the postganglionic nerves are ligated, thus preventing regeneration of the majority of ganglion cell axons, synaptic recovery does not occur, and within 3 months most of the affected neurons die and are removed by phagocytic action (108). Effects of Colchicine Application. These findings suggest that the integrity of a ganglion cell’s axon is necessary for the maintenance of preganglionic synaptic contacts and ultimately for neuronal survival. To investigate the likelihood that the sequence of events after axotomy represents interference with a physiological control mechanism, those experiments were repeated using local colchicine application instead of mechanical interruption (103, 109). A single drop of 0.1 M colchicine solution was applied to one of the major postganglionic branches for 30 min. Four to seven days later the properties of neurons antidromically driven from a point several millimeters distal to the point of drug application were compared to neurons in the same region but projecting to the periphery via an untreated postganglionic nerve. Although there was no electrophysiologic evidence of axon damage, treated neurons showed marked depression of synaptic transmission. Electron microscopical counts of synapses in different regions of the ganglion confirmed that the drug treatment caused loss of synapses from the dendrites of treated, but not untreated, neurons. A number of other axotomy effects were also mimicked by colchicine. These findings suggest that loss of synapses after axotomy is not primarily a response to neuronal injury per se, but is the result of interference with a normal control mechanism which depends on axoplasmic transport. Trophic Effect of Nerve Growth Factor. A natural question in the sympathetic nervous system is whether or not the protein nerve growth

680

VERAA,

GRAFSTEIN,

AND ROSS

factor (NGF) might be involved in such a mechanism of control. A number of investigators showed that this agent is specifically taken up by sympathetic nerve terminals, is retrogradely transported to sympathetic ganglion cells, and can prevent some of the biochemical effects of axotomy in young animals (71, 72). To examine this possibility, silastic pellets impregnated with NGF were implanted near ganglia in which a population of neurons had had their axons crushed (92). After 4 to 7 days-the interval at which postaxotomy synaptic depression is maximal-the properties of axotomized and normal neurons were compared by means of intracellular recording in ganglia treated with NGF. Exogenous NGF released from the pellets during a period of days could largely prevent the decrease in EPSP amplitude after axotomy, as well as some of the other changes which usually occur, such as the development of regenerative responses in dendrites (108). This result fulfills one of the criteria that would establish NGF as atrophic agent serving to maintain synaptic function in the peripheral sympathetic system. Consistent with this is the additional finding that systemic treatment of adult guinea pigs with antiserum to NGF caused a loss of ganglionic synapses and other changes which are, in many ways, similar to the effects of postganglionic axotomy (92). The implication of these results is that mature synapses in sympathetic ganglia are maintained by a retrograde influence which requires the extension of ganglion cell axons to the periphery. The evidence available so far suggests that this retrograde influence may be mediated in part by the protein nerve growth factor. Axon Growth and Functional Restoration after Optic Axon Transection in the Goldfisht7 The optic axons of the goldfish show extremely successful regeneration. Regenerative axonal growth begins a few days after transection of the optic tract, and visual function recovers in approximately 3 to 4 weeks (42, 44, 84). Approximately 7 to 10 days after transection of the optic tract, regenerating fibers are observed in the optic tectum. Those fibers are small, slender, and very densely packed. Indeed, there is at this time an excessive number of small-caliber sprouts growing from the cut end of the optic tract to deploy into the tectum. By 3 to 4 weeks after transection, regenerating fibers are still small, unmyelinated, and densely packed, although they have enlarged considerably in diameter. It is characteristic of the regenerating system that, rather than having a single fiber going through the striatum opticum and terminatI7 This section is based on the presentation by Dr. Marion Murray.

AXONAL

GROWTH

681

MECHANISMS

ing in the neuropil, large fascicles are observed consisting of 20 to 30and sometimes more-unmyelinated axons. Toward the end of this period, restoration of function can be demonstrated, although functional return by no means represents the end point of the regenerative process. The diameter of regenerated fibers continues to increase for at least 3 months after the lesion, and there are indications that at later periods it may still be increasing slowly approaching, but never quite achieving, the normal diameter. Another event that occurs late in the regenerative process is the myelination of the regenerating fibers. At 6 weeks after transection virtually all regenerating fibers remain unmyelinated. At 4 months after transection, although myelin sheaths are more substantial, 70% of the new fibers remain unmyelinated. In addition to retinotectal fibers described above, another system of retinal fibers also regenerates to terminate in the dorsal part of the diencephalon. As these terminations are closer to the optic tract, one might expect their regeneration prior to termination in the tectum, but that is not the case. Autoradiographic tracing of regenerating fibers has shown that at 3 to 4 weeks post-transection there is intense labeling in the optic tract and also in the optic tectum. At that time, however, there is no sign of any projection into the diencephalon. Approximately 2 weeks later (6 weeks postlesion), label appears in the diencephalic nuclei. In these studies it was also noted that labeling is considerably more intense on the regenerating side than on the control side in unilateral preparations, which is consistent with electron microscopic observations that at this and later periods regeneration processes are still active as the axons undergo considerable increase in diameter. The late development of diencephalic projections suggests two possibilities. First, there may be two populations of retinal ganglion cells, a rapidly regenerating group that gives rise to the projection to the tectum, and another population, more slowly regenerating, giving rise to the diencephalic projection. The second interpretation is that the diencephalic projection may derive from collaterals of retinotectal fibers that develop only when primary projections to the tectum are completed. The latter possibility is suggested as most consistent with other observations. ESTABLISHMENT

OF NEURONAL

CONNECTIONS

No matter how vigorous axonal regeneration might be after nervous system injury, functional recovery could not occur unless the growing axons were able to reestablish effective synaptic endings with their target cells. One particularly useful system for studying the mechanism involved in establishing such connection has been the neuromuscular junction, where

682

VERAA,

GRAFSTEIN,

AND

ROSS

there are clear criteria for identification of synaptic structures, e.g., the accumulation of ACh receptor molecules. One indication of the importance for the growing neuron of establishing effective synaptic connections is that during development nerve cells die if they fail to establish appropriate connections with the periphery. Nerve-Muscle

Junction

Formation

in Vitro18

Neuromuscularjunction formation was studied in chick embyro cell and tissue cultures in terms of the relationship of the ACh receptors and sites of innervation on the muscle membrane. The principal question addressed was whether motor axons induce new clusters of ACh receptors (at sites of transmitter release) or seek out preexisting clusters. In adult innervated muscle, receptors are restricted to the immediate subsynaptic membrane. The great majority of the muscle membrane is insensitive to ACh, whereas when the fibers are denervated, extrajunctional receptors appear as reported above (see section by Fambrough). The present study undertook to determine the precise distribution of the receptors in the vicinity of nerve terminals and whether or not this distribution changes during the process of synapse formation in vitro. In earlier studies of muscle fibers in the absence of neurons it was determined that the myotubes are sensitive to ACh over their entire length. However, relative peaks or “hot spots” were noted, with sensitivities an order of magnitude greater than the background sensitivity of the nearby membrane. Are these hot spots preferred sites of synapse formation? Innervation was studied after the addition of 4-day spinal cord explants to the above described muscle cultures. With electrophysiologic recording the presence of synapses was demonstrable quite rapidly-within I5 to 20 h after the explants settled on the muscle culture surface. It was noted that in many cases large varicosities along neurites overlying myotubes were not functional, whereas in some cases functional contact was achieved by an otherwise nondescript nerve fiber in contiguity with a muscle fiber, Acetylcholine sensitivity was mapped, and it was found that in every case a relative peak of ACh sensitivity was located within 2 Frn of the site of transmitter release. In subsequent studies, innervation was achieved with spinal cord explants cut from 16-day embryos. The initial outgrowth from these older explants is restricted to the ventral horn, and very large distinctive neurites that are not seen in younger explants emerge. Sites of transmitter release were located by extracellular recording or focal extracellular stimulation (in the presence of 10e7 M tetrodotoxin). A technique was devised for I8 This section

is based

on the presentation

by Dr. Gerald

D. Fischbach.

AXONAL

GROWTH

683

MECHANISMS

rapid and precise mapping of ACh sensitivity. An ACh ionophoretic electrode was moved over the surface and photographed on 16-mm movie film at each test site. Responses to ACh were analyzed simultaneously with a small on-line computer system. The camera and computer permitted the mapping of as many as 100 points in 30 min. This generated a very accurate two-dimensional picture of the distribution of the surface ACh receptors in the muscle fibers before contact with a nerve and immediately after. It was found that relative peaks or hot spots of ACh sensitivity in uninnervated muscle membranes are extremely stable. Relative peaks can be detected at the same position for 3 or 4 days. After blockage of all exposed receptors with cr-bungarotoxin, new receptors appear rapidly (as discussed by Fambrough above), and hot spots reappear in exactly the same place. Thus stable clusters are composed of receptors with rapid turnover times. As neurites contact nerve fibers, however, it was seen that they are indeed capable of inducing new areas of high ACh sensitivity. This was demonstrated by watching a growth cone migrate to the muscle cell membrane, observing a new synapse form, and seeing a new area of high sensitivity appear where one previously did not exist. This process is extremely rapid and increased areas of sensitivity have been mapped within 60 min after nerve fiber contact. Thus neurites can create new hot spots. They need not seek out and synapse on preexisting regions of high receptor density. More studies will be required to determine the origin of receptors that accumulate at newly formed synapses. Redistribution of Acetylcholine Receptors during Formation Neuromuscular Connections in Cell Culture’s

of

Myotomal muscle cells from l-day-old embryos ofxenopus laevis were cultured as a monolayer with and without neural tube cells. They developed striations within 1 day and remained mononucleated and stationary. When neural tube cells were included in the cultures many of the cells became innervated: spontaneous twitching and contractions evoked by electrical stimulation of neuronal perykarya were observed only in nerve-contacted cells, and this activity was abolished by curare and by a-bungarotoxin (2). By 2 days in culture all noninnervated muscle cells had developed one or more characteristic patches of ACh receptors, as revealed by staining with fluorescent conjugates of a-bungarotoxin. Such receptor patches were uncommon on innervated cells; instead the fluorescent staining was usually confined to the path of nerve-muscle contact. This latter form of staining often extended for greater distances than the patches seen on nonle This

section

is based on the presentation

by Dr.

Monroe

Cohen.

684

VERAA,

GRAFSTEIN,

AND ROSS

innervated cells, thereby indicating that the nerve had induced the accumulation of at least some of the receptors at sites of contact. Similar patterns of fluorescent staining were also observed within 1 day of adding neural tube cells to 2- to 3-day-old muscle cultures. It therefore follows that innervation not only prevented the development of receptor patches but also caused their disappearance. All these nerve-induced changes in receptor distribution also occurred in the presence of agents which block neuromuscular transmission. Additional experiments were carried out to determine whether or not the accumulation of ACh receptors at sites of nerve-muscle contact involves a process of receptor redistribution. Three-day-old muscle cultures were stained with fluorescent toxin and maintained thereafter in native toxin to ensure that newly synthesized receptors would not be stained. Neural tube cells were then added. When the cultures were examined 1 to 3 days later, characteristic patterns of stain were observed at sites of nerve-muscle contact. Of particular significance were examples where the stain along the path of contact extended for greater distances than the patches of stain seen on noninnervated cells, for such examples indicate that some of the receptors must have originally existed elsewhere on the muscle cell. Successive observations on individual muscle cells confirmed that prestained receptors accumulate at sites of nerve-muscle contact and further revealed the formation of new receptor patches on noncontacted muscle cells. On the basis of these findings it is concluded that: (a) ACh receptors change their position in the sarcolemma and accumulate at sites of nervemuscle contact, (b) this accumulation is due to the release of a neural substance or to some property of the axolemma, and(c) receptor redistribution also occurs spontaneously (1, 2). Synapse Formation and Neuron Cell Death during EmbryogenesiszO

Studies were undertaken to characterize the principles governing the formation of neuronal connections in a simple avian effector system composed of central preganglionic elements which impinge on the cells of the ciliary ganglion, the axons of which form junctions with effector organs in the eye. Synapse Formation. The system chosen for study includes central neurons located in the bird midbrain which synapse with the connecting neurons of the avian ciliary ganglion. This ganglion consists of two types of cells-the choroid cells and the ciliary cells. Choroid cells synapse with the smooth muscle of the choroidal coat of the eye and ciliary cells innervate the striated muscle of the iris and ciliary body. This system thus provides two stages of neuron-target interaction, for the ganglion cells *OThis section is based on the presentation by Dr. Guillermo R. Pilar.

AXONAL

GROWTH

MECHANISMS

685

serve as the target of the central neurons, and the iris and choroidal muscle are the target of the ganglion cells. Similarly, both neuron-neuron and neuromuscular interactions occur and are available for study (66). The time course of synapse formation was studied (62). The ganglion cells are innervated by the central neurons at approximately stage 30 (day 7 of embryonic life). The ganglion cells themselves form synapses with the peripheral target at stage 34 (day 8). Neuromuscular junction formation is complete by stage 39 (13 days of incubation). The ganglion cells migrate from the cranial end of the neural crest very early in development, and migration is completed at about stage 25. At that time, the ganglion cells appear to be electrically as well as biochemically differentiated, because they conduct action potentials and have acetylcholinesterase and choline acetyltransferase activities (19). This suggests that the ganglion cells, which are pluripotential during migration (69), immediately express cholinergic characteristics at the termination of their migration. During the period of synapse formation a luxuriant outgrowth of somal processes was noted. It is hypothesized that the presence of these pseudodendrites increases the likelihood of the central neurons finding appropriate synaptic targets. After the preganglionic synapses are formed, these processes are retracted. Another feature observed during this phase of synaptogenesis is that the presynaptic cells selectively innervate the ciliary and choroid ganglion cells from the very first establishment of those connections. As is also true in the mature animal (61) only the faster conducting, myelinated preganglionic fibers synapse with the ciliary cells (62, 65). Thus, “specific affinities ,’ ’ as originally proposed by Sperry ( 13 l), seem to be involved in the formation of synapses in this system. Similarly, the ciliary and choroida1 cells send their axons only in the direction of their proper target organs (65). The ciliary cells have faster conducting axons and have larger diameters than the axons that form the choroidal nerves, and it appears that the decision to innervate the appropriate peripheral target is made very close to the cell soma. Concomitant with synapse formation at the neuromuscular junctions is a 200-fold increase in the activity of choline acetyltransferase (19). The increased activity of the enzyme seems to be triggered by synapse formation. During synapse formation, an elaborate reorganization of rough endoplasmic reticulum appears in the ganglionic cells, and this coincides with the time of increased synthesis of the enzyme, which is destined for export (104). In experiments in which the target organs were removed early in development (63), the ganglion cells possessed very little rough endoplasmic reticulum, and most of the ribosomes were observed to be unbound, a situation comparable to that in the normal cell prior to the establishment of terminals (104).

686

VERAA,

GRAFSTEIN,

AND

ROSS

Cell Death after Nerve Terminal Formation. In the normal ganglion it was observed that half the cells degenerate and die during peripheral synapse formation (64). This normal cell death can be ameliorated if the number of neurons competing for interaction with the target is decreased (105). Both sets of observations suggest that because these ganglion cells can be rescued from death, they are not necessarily destined to die and are not defective. Furthermore, the evidence suggests that the neurons interacting with the target compete for a limited amount of some trophic substance, a limited number of synaptic sites, or a substance that can be taken up only after the formation of synapses. Thus cells that fail to compete successfully for the target organ become superfluous and, as a consequence, degenerate. The dependence of neurons on the target for survival was also demonstrated when the optic vesicle was removed at earlier stages of embryonic development (65). The ganglia in those operated birds develop normally until stages 34 to 39, the time during which normal ganglion cells would be expected to form peripheral synapses. At that time most of the peripherally deprived ganglion cells die, confirming previous observations (49, 50). Thus, it was concluded that this neuron cell death plays an important role in the shaping of neuronal circuitry (25). Ultrastructural Studies. Electron microscopic studies of neuronal cell death in normal ganglia and in peripherally deprived ganglia (65, 104) further support the postulated interaction between target organ and neuron. If normal cell death results simply because neurons have failed to form peripheral connections, one would expect that the appearance of cell death at the ultrastructural level would be similar in the two cases. This concept, however, is not supported by the findings, for the mechanism of cell death appears to be different in each case. Cell death in the peripherally deprived neurons was signaled by nuclear changes followed by freeing of ribosomal polysomes from the rough endoplasmic reticulum, and presumably cessation of protein synthesis. In contrast, “normal” cell death was characterized by dilation of the rough endoplasmic reticulum, with eventual cytoplasmic disruption and nuclear changes appearing only secondarily. This type of ceil death in the normal ganglion occurs only in cells that have already interacted with the periphery, and is possibly indicative of competition for synaptic sites. The failure to form and maintain synapses could result in the accumulation of transmission-related proteins with consequent cisternal dilation and eventual cell death. In Vivo REGENERATION

OF INTRINSIC SPINAL CORD NEURITES One approach to the problem of CNS regeneration is an empirical search for means of promoting neural regeneration, for example, by the injection

AXONAL

GROWTH

MECHANISMS

687

of chemicals that might be expected, for one reason or another, to be effective in stimulating axonal growth. Some interesting studies of this kind have been conducted in the USSR, involving the use of proteolytic enzymes to enhance recovery from spinal cord injury in rats. Enzyme Therapy for Spinal Cord Injury21

The effect of proteolytic and mucolytic enzyme preparations in creating favorable conditions for spinal cord regeneration was tested in female rats 6 to 7 weeks of age, after complete spinal cord transection at the level of TS. At the time of operation, and daily for varying times thereafter, enzyme preparations, either alone or in combination, were administered and the degree of morphological, electrophysiological, and functional recovery was examined for as long as 340 days (82). Enzyme Preparations. The enzyme preparations administered were lidase, hyaluronidase, trypsin, and elastase. The lidase preparation, extracted from cattle, was enzymatically similar to hyaluronidase. Each enzyme preparation used was purified to its crystalline form. After spinal cord transection, the incision was flushed, and the enzyme preparation was infused topically. Daily thereafter, for the first 15 days, enzyme preparations were administered topically at the lesion site as well as systemically (intramuscularly). Subsequently, daily enzyme preparations were administered systemically for as long as 3 months. Control animals were rats with similar complete spinal cord transections followed by daily administration of buffered physiologic solution. The best results, with respect to the percentage of animals (a) with partial or complete recovery of symatic function, (b) with a corresponding morphological improvement at the lesion site, or (c) with a longer span of life, were obtained with a combined administration of the various enzyme preparations, specifically hyaluronidase followed by trypsin, or trypsin followed by elastase. The joint administration of proteolytic as well as mucolytic and proteolytic enzymes increased considerably the effectiveness of treatment, compared with administration of each enzyme alone. It was proposed that the effect is due to the fact that the different enzymes are active on different peptide bonds of protein molecules, thereby permitting a more complete proteolysis to occur when more than one type of enzyme is used. Another important property of such proteolytic enzymes as trypsin and elastase is the selective enzyme hydrolysis by them of necrotic tissue -particularly elastin and fibrin-without affecting healthy tissue. Moreover, such hydrolysis may provide the necessary substrates for increased protein synthesis during regeneration. *I This section is based on the presentation by Dr. Levon A. Matinian. Another synopsis of Dr. Matinian’s experiments has appeared elsewhere (102).

688

VERAA,

GRAFSTEIN,

AND

ROSS

Morphological Recovery. The joint administration of these enzymes brought about good and stable recovery of the morphologic elements of the nervous tissue. In cases of good functional recovery, the scar formation at the lesion was less, the recovery of vascularization (as evidenced by an increase in the number and size of blood vessels) was better, the content of tissue hyaluronic acid-a measure of the amount of connective tissue present- was lower, and the growth of nerve fibers through the loose scar tissue was greater than that seen in cases of poor functional recovery and controls. New fibers from both proximal and distal parts of the spinal cord grew through the lesion area to make contact with neurons of both stumps. The nerve cell bodies close to the lesion were found to be larger, hyperchromic, and showing many more processes than the control. Moreover, there was an increase in size and number of synapses in this area of the spinal cord. Electrophysiologicaf Recovery. Enzyme therapy brought about a gradual reestablishment of afferent and efferent conduction through the lesion site. During the initial postoperative period (24 to 27 days), the appearance of slight regeneration was accompanied by the return of afferent nerve impulse conduction through the lesion site. However, the rate of afferent conduction was greatly reduced, and the latent period of the evoked cortical response was markedly prolonged. From 7 to 9 weeks, the rate of impulse conduction through the lesion improved, increasing 1.5 to 2 times. The latent period was also reduced. By 9 to 14 months postlesion, the rate of conduction of afferent impulses was still 2.5 to 3 times slower than normal, and the latent period of the evoked cortical response was prolonged. That conduction is indeed occurring through regenerating fibers across the lesion is confirmed by the observation that both afferent and efferent conductivity are lost after a second transection of the spinal cord. The recovery of conduction through the lesion is thought to be due largely to intraspinal fibers which grow across the lesion. Also, regenerating nerve root fibers play a role in establishing afferent pathways from the periphery. The observation that the rate of afferent impulse conduction is still slow after 9 to 14 months may result from impulse conduction, not along the normal pathways, but via slower conducting fibers, such as the spinothalamic pathway. Functional Recovery. The first sign of recovery of motor function was seen between 1.5 and 2 months after the operation and was immediatefy preceded by recovery of visceral organ function. In the evaluation of results, “partial” recovery denoted recovery of visceral and sensory functions, while “complete” recovery signified the additional restoration of coordinated motor function in the experimental animals. Of those animals surviving longer than 3 months, greater than 85% showed partial or complete functional recovery after the administration of (a) lidase, (b) hyal-

AXONAL

GROWTH

MECHANISMS

689

uronidase and trypsin, and (c) trypsin and elastase. Moreover, such recovery was long-lasting, being observed for as long as 340 days following the lesion. CONCLUDING

STATEMENT

This Conference, like the three preceding ones in this series (46-48, 120) has been developed on the strength of the idea that regeneration in the CNS can occur, and that such regeneration can be the basis for developing a cure for paraplegia and other conditions resulting from injury to the CNS. At present we are still far from understanding how to induce and regulate the regenerative process, and to reach our goal we must still look for the answers to some fundamental questions such as: How can the axons of central neurons be induced to sprout? What circumstances are conducive to maintenance of axonal growth? What mechanisms must be mobilized for the production of materials required for this growth? What factors determine when synaptic terminals will be formed? What changes occur in the presynaptic and postsynaptic elements as a consequence of terminal formation? In seeking answers to these questions we must rely not only on experiments on the CNS itself, but on studies of regeneration in the peripheral nervous system, studies of nervous system development, and even studies of renewal processes in nonneuronal cells, in order to reach an understanding of the basic cellular mechanisms involved in growth and repair. One important area, for example, lies in the analysis of events in the injured nerve cell. In the course of this Conference we have considered the nature of some of the metabolic changes that may be elicited, including changes in RNA and protein metabolism (Lieberman, Barr-on), and changes in synaptic transmitter function in both the injured neuron itself (Reis) and its presynaptic connection (Purves). How those events may contribute to the success or failure of regeneration is a critical question that remains to be explored, and in this connection we have much to lean from analyzing the phenomenally successful regeneration that may be observed in the CNS of lower vertebrates (Murray). Regeneration involves the replacement of the neuronal structures amputated by injury, including the plasma membrane, with its characteristic differences in structure among its various regions. General principles of plasma membrane formation (Bennett, Nicolson) are likely to be the same in nerve cells as in other types of cells, and the mechanisms leading to regional specializations of structure in visual receptor cells (Young) and muscle cells (Fambrough) may provide some models for corresponding neuronal processes. Investigation of axonal outgrowth after injury of the CNS in young

690

VERAA,

GRAFSTEIN,

AND ROSS

animals reveals some of the conditions under which such growth occurs and the constraints that may be operating. One important principle that emerges is that the size of the terminal field may be relatively constant (Schneider). Another is that in the establishment of connections competition for terminal space occurs among different inputs. However, there is a limitation on the kinds of input that will be accepted, presumably reflecting special affinities among nerve cells, which become increasingly restricted in the course of development (Lund, Cotman). A feedback mechanism operating to enhance the survival of neurons with appropriate synaptic connections by eliminating neurons that have failed to connect has also been demonstrated (Pilar). Some postsynaptic factors that might influence the establishment of synaptic connections have been explored in detail at neuromuscular junctions in tissue culture preparations combining muscle and nerve cells. Analysis of the disposition of ACh receptors during the establishment of neuromuscular contact, for example, showed that although regions of receptor aggregation build up spontaneously before innervation occurs, the incoming motor axons do not invariably make their connection at these sites. On the other hand, when an effective connection is established, the receptor molecules do accumulate there (Fischbach, Cohen). Morphological changes in both the pre- and postsynaptic elements during synapse formation indicate that a close mutual interaction between the two elements is a requisite for synaptogenesis (R. Bunge). The role of the “supporting” cells, i.e., glial or Schwann cells, must not be neglected. Glial cells clearly influence the growth and differentiation of nerve cells (Webster, R. Bunge), and successful axonal regeneration in the CNS of lower animals may depend on appropriate responses from the glial cells and ependymal cells (Singer). Conversely, the activity of the supporting cells may be influenced by the approach of regenerating axons (Webster, M. Bunge). Eventually, the principles gleaned from studies like those described above must be translated into techniques that can be used to enhance regeneration in the CNS. In this respect, the pioneering attempts by Matinian to apply enzymes in the treatment of the injured spinal cord represent a bold and direct approach to this problem and encourage us toward further efforts in this direction. REFERENCES 1.

M. J., AND M. W. COHEN. 1977. Nerve-induced and spontaneous redistribution of acetylcholine receptors on cultured muscle cells. J. Physiol. (London)

ANDERSON,

268: 757-773. 2. ANDERSON, M.

the distribution (London)

J., M. W. COHEN, of acetylcholine

268: 73 I- 756.

E. ZORYCHTA. 1977. Effects of innervation on receptors on cultured muscle cells. .I. Physiol.

AND

AXONAL

GROWTH

MECHANISMS

691

3. BARRON, K. D., M. P. DENTINGER, L. R. NELSON, AND M. E. SCHEIBLY. 1976. Incorporation of tritiated leucine by axoromized rubral neurons. Bruin Res. 116: 251266. 4. BARRON, K. D., J. B. OLDERSHAW, AND J. BERNSOHN. 1966. Hydrolase cytochemistry of retrograde neuronal degeneration in feline lateral geniculate body. J. Neurupathol. Exp. Neural. 25: 443-478. 5. BARRON, K. D., S. SCHREIBER, J. L. COVA, AND M. E. SCHEIBLY. 1977. Quantitative cytochemistry of RNA in axotomized feline rubral neurons. Brain Res. 130: 469481. 6. BENNETT, G., AND C. P. LEBLOND. 1977. Biosynthesis of the glycoproteins present in plasma membrane, lysosomes and secretory materials as visualized by radioautography. Histochem. J. 9: 393-417. 7. BIBB, C., AND R. W. YOUNG. 1974. Renewal of fatty acids in the membranes of visual cell outer segments. J. CeN Biol. 61: 327-343. 8. BIBB, C., AND R. W. YOUNG. 1974. Renewal of glycerol in the visual cells and pigment epithelium of the frog retina. J. Cell Biol. 62: 378-389. 9. BLINZINGER, K., AND G. W. KRELJTZBERG. 1968. Displacement of synaptic terminals from regenerating motoneurons by microglial cells. 2. Zellforsch. 85: 145 157. 10. BRENNER, H. R., AND E. W. JOHNSON. 1976. Physiological and morphological effects of postganglionic axotomy on presynaptic nerve terminals. J. Physiol. (London) 260: l43- 158. 11. BRENNER, H. R., AND A. R. MARTIN. 1976. Reduction in acetylcholine sensitivity of axotomized ciliary ganglion cells. J. Physiol. (London) 260: 159- 175. 12. BUNGE, M. B. 1973. Fine structure of nerve fibers and growth cones of isolated sympathetic neurons in culture. J. Cell Biol. 56: 713-735. 13. BUNGE. M. B. 1977. Initial endocytosis of peroxidase or ferritin by growth cones of cultured nerve cells. J. Neurocytol. 6: 407-439. 14. BUNGE, M., J. JEFFREY, AND P. WOOD. 1977. Different contributions of Schwann cells and libroblasts to the collagenous components of peripheral nerve. J. CeN Biol. 75: 16la (abstract). 15. BUNGE, R. P., M. JOHNSON, AND C. D. Ross. 1978. Nature and nurture in the development of the autonomic neuron. Science 199: 1409- 1416. 16. BURRELL, H. R., L. A. DOKAS, AND B. W. AGRANOFF. 1978. RNA metabolism in the goldfish retina during optic nerve regeneration. J. Neurochem. 31: 289-298. 17. CHEN. D. H. 1978. Qualitative and quantitative study of synaptic displacement in chromatolyzed spinal motor neurons of the cat. J. Comp. Neural. 177: 635-664. 18. CHEN, D. H.. W. W. CHAMBERS, AND C. N. Lru. 1977. Synaptic displacement in i&acentral neurons of Clarke’s nucleus following axotomy in the cat. Exp. Neural. 57: 1026- 1041. 19. CHIAPPINELLI, V., E. GIACOBINI, G. PILAR, AND H. UCHIMURA. 1976. Induction of cholinergic enzymes in chick ciliary ganglion and iris muscle cells during synaptic formation. J. Physiol. (London) 257: 749-766. 20. COTMAN, C. W. 1978. Synapse formation and plasticity in the developing dentate gyrus. Pages 47-65 in K. LIVINGSTONE AND 0. HORNYKEIWICZ, Eds., Limbic Mechanism. The Continuing Evolution of the Limbic System Concept. Plenum Press, New York. 21. COTMAN, C. W., AND G. S. LYNCH. 1976. Reactive synaptogenesis in the adult nervous system: the effects of partial deafferentation on new synapse formation. Pages 69- 108 in S. BARONDES, Ed.. Neuronrrl Recognition. Plenum Press, New York. 22. COTMAN, C. W., D. A. MATTHEWS, D. TAYLOR, AND G. LYNCH. 1973. Synaptic rearrangement in the dentate gyrus: Histochemical evidence ofadjustments after lesions in immature and adult rats. Proc. Nnti. Arad. Sci. U.S.A. 70: 3473-3477.

692

VERAA,

GRAFSTEIN,

AND

ROSS

23. COTMAN, C. W., AND J. V. NADLER. 1978. Reactive synaptogenesis in the hippocampus. Pages 227-271. in C. W. COTMAN, Ed., Neuronal Plasticity. Raven Press, New York. 24. COVA, J. L., AND K. D. BARRON. 1977. Uptake of leucine-3H by axotomized cervical motoneurons. Abstract. Anat. Rec. 187: 558. 25. COWAN, W. M. 1973. Neuronal death as a regulative mechanism in the control of cell number in the nervous system. Pages 19-41 in Development and Aging in the Nervous System. Academic Press, New York. 26. DEVOR, M. 1975. Neuroplasticity in the sparing or deterioration of function after early olfactory tract lesions. Science 190: 998- 1000. 27. DEVOR, M. 1976. Neuroplasticity in the rearrangement of olfactory tract fibers after neonatal transection in hamsters. .I. Comp. Neurol. 166: 31-58. 28. DEVOR, M., AND G. E. SCHNEIDER. 1975. Neuroanatomical plasticity: the principle of conservation of total axonal arborization. Pages 191-200 in F. VITAL-DURAND AND M. JEANNEROD, Eds., Aspects of Neural Plasticity. I.N.S.E.R.M., Paris, France. 29. DEVREOTES, P. N., AND D. M. FAMBROUGH. 1975. Acetylcholine receptor turnover in membranes of developing muscle fibers. J. Cell Biol. 65: 335-358. 30. DEVREOTES, P. N., AND D. M. FAMBROUGH. 1976. Turnover of acetylcholine receptors in skeletal muscle. Cold Spring Harbor Symp. Quant. Biol. 40: 237-251. 31. DEVREOTES, P. N., J. M. GARDNER, AND D. M. FAMBROUGH. 1977. Kinetics of biosynthesis of acetylcholine receptor and subsequent incorporation into plasma membrane of cultured chick skeletal muscle. Cell 10: 365-373. 32. EGAR, M., S. B. SIMPSON, AND M. SINGER. 1970. The growth and differentiation of the regenerating spinal cord of the lizard, Anolis carolinensis. J. Morphol. 131: 131- 152. 33. EGAR, M., AND M. SINGER. 1972. The role of ependyma in spinal cord regeneration in the urodele, Triturus. Exp. Neural. 37: 422-430. 34. FAMBROUGH, D. M., P. N. DEVREOTES, AND D. 3. CARD. 1977. The synthesis and degradation of acetylcholine receptors. Pages 202-236 in G. A. COTTRELL AND P. N. R. USHERWOOD, Eds., The Synapse. Blackie, Glasglow. 35. FINLAY, B. L. 1976. Neuronal Specificity and Plasticity in the Hamster Superior Colliculus: Electrophysiological Studies. Ph.D. thesis, M.I.T., Cambridge, Mass. 36. FINLAY, B., K. WILSON, AND G. E. SCHNEIDER. 1979. Anomalous ipsilateral retinotectal projections in Syrian hamsters with early lesions: topography and functional capacity. J. Comp. Neural. 183: 721-740. 37. FROST, D. O., AND G. E. SCHNEIDER. 1975. Plasticitiy of retinotectal projections after partial lesions of retina in newborn Syrian hamsters. Sot. Neurosci. Abstr. 1: 495. 38. FROST. D. O., AND G. E. SCHNEIDER. 1979. Plasticity of retinofugal projections after partial lesions of the retina in newborn Syrian hamsters. J. Comp. Neurol. 39. FUMAGALLI, L., G. DERENZIS, AND N. MIANI. 1978. cY-Bungarotoxin-acetylcholine receptors in the chick ciliary ganglion: effects of deafferentation and axotomy. Brain Res. 153: 87-98. 40. GAINER, H., I. TASAKI, AND R. J. LASEK. 1973. Evidence for the glia-neuron protein transfer hypothesis from intracellular perfusion studies of solid giant axons. J. Ce// Biol. 74: 524-530. 41. GOLDMAN, G. C., A. F. MIRANDA, A. D. DEITCH, AND S. W. TANNENBAUM. 1975. Action of cytochalasin D on cells of established lines. III. Zeiosis and movements at the cell surface. J. Cell Biol. 64: 644-667. 42. GRAFSTEIN, B. 1971. Role of slow axonal transport in nerve regeneration. Acta Neuropathol. Suppl. 5: 144-152. 43. GRAFSTEIN, B. 1975. The nerve cell body response to axotomy. Exp. Neurol. 48: 32-51.

AXONAL

GROWTH

MECHANISMS

693

44. GRAFSTEIN, B., AND M. MURRAY. 1%9. Transport of protein in goldfish optic nerve during regeneration. Exp. Neural. 25: 494-508. 45. GUNNING, P. W., P. L. KAYE, AND L. AUSTIN. 1977. In vivo synthesis of rapidlylabelled RNA within the rat nodose ganglia following vagotomy. J. Neurochem. 2%: 1245- 1248. 46. GIJTH, L. 1974. Axonal regeneration and functional plasticity in the central nervous system. Exp. Neurol. 45: 606-654. 47. GUTH, L., AND W. F. WINDLE. 1970. The enigma of central nervous regeneration. Exp. Neural. 28(Suppl. 5): l-43. 48. GUTH, L., AND W. F. WINDLE. 1973. Physiological, molecular, and genetic aspects of central nervous system regeneration. Exp. Neural. 39: iii-xvi. 49. HAMBURGER, V. 1975. Cell death in the development of the lateral motor column of the chick embryo. J. Comp. Neural. 160: 535-546. 50. HAMBURGER, V., AND R. LEVI-M• NTALCINI. 1949. Proliferation, differentiation and degeneration in the spinal ganglion of the chick under normal and experimental conditions. J. Exp. 2001. 111: 457-501. 51. HOY, R. R., D. BITTNER, AND D. KENNEDY. 1967. Regeneration in crustacean motoneurones; evidence for axonal fusion. Science 156: 251-252. 52. JAEGER, C. B., AND R. D. LUND. 1978. Development of cerebral cortex transplanted to cerebral cortex or tectum of newborn rat hosts. Sot. Neurosci. Absfr. 4: 116. 53. JHAVERI, S. R., AND G. E. SCHNEIDER. 1974. Retinal projections in Syrian hamsters: normal topography, and alterations after partial tectum lesions at birth. Ana?. Rec. 178: 383.

JOHNSON, M., D. Ross, M. MEYERS, R. REES, R. BUNGE, E. WAKSHULL, AND H. BURTON. 1976. Synaptic vesicle cytochemistry changes when cultured sympathetic neurones develop cholinergic interactions. Nature (London) 262: 308-3 10. 55. KALIL, R., AND G. E. SCHNEIDER. 1975. Abnormal synaptic connections of the optic tract in the thalamus after midbrain lesions in newborn hamsters. Bruin Res. 100: 54.

690-698. 56. 57.

58.

59. 60. 61. 62.

63. 64.

65.

KAYE, P. L., P. W. GUNNING, AND L. AUSTIN. 1977. In vivo synthesis of stable RNA within the rat nodose ganglia following vagotomy. J. Neurochem. 28: 1241- 1243. Ko, C.-P., H. BURTON, AND R. P. BUNGE. 1976. Synaptic transmission between rat spinal cord explants and dissociated superior cervical ganglion neurons in tissue culture. Bruin Res. 117: 437-460. KREUTZBERG, G. W., AND P. SCHUBERT. 1971. Volume changes in the axon during regeneration. Acta Neuropathol. 17: 220-226. KRISHNAN, N., AND M. SINGER. 1973. Penetration of peroxidase into peripheral nerve fibers. Am. J. Anat. 136: l-15. KUNO, M. 1976. Responses of spinal motor neurons to section and restoration of peripheral motor connections. Cold Spring Harbor Symp. Quant. Biol. 40: 457-463. LANDMESSER, L., AND G. PILAR. 1970. Selective reinnervation of two cell population4 in the adult pigeon ciliary ganglion. J. Physiol. (London) 211: 203-2 16. LANDMESSER, L., AND G. PILAR. 1972. The onset and development of transmission in the chick ciliary ganglion. J. Physiol. (London) 222: 691-713. LANDMESSER, L., AND G. PILAR. 1974. Synapse formation during embryogenesis on ganglion cells lacking periphery. J. Physiol. (London) 241: 715-736. LANDMESSER, L., AND G. PILAR. 1974. Synaptic transmission and cell death during normal ganglionic development. J. Physiol. (London) 241: 738-750. LANDMESSER, L.. AND G. PILAR. 1976. Fate of ganglionic synapses and ganglion cell axons during normal and induced cell death. J. Cell Biol. 68: 357-374.

694

VERAA,

GRAFSTEIN,

AND

ROSS

L., AND G. PILAR. 1978. Interactions between neurons and their targets synaptogenesis. Fed. Proc. 37: 2016-2022. LASEK, R. J., H. GAINER, AND R. J. PRZYBYLSKI. 1974. Transfer of newly synthesized proteins from Schwann cells to the squid giant axon: glia, neurons, secretion. Proc. Nat/. Acad. Sci. U.S.A. 71: 1188-1192. LAVAIL, M. M. 1976. Rod outer segment disk shedding in rat retina: relationship to cyclic lighting. Science 194: 1071-1074. LEDOUARIN, N. M., AND N. M. TEILLER. 1974. Experimental analysis of the migration and differentiation of neuroblasts of the autonomic nervous system and of neuroectodermal mesenchymal derivatives, using a biological cell marking technique. Dev. Biol. 41: 162-184. LEONG, S. K., AND R. D. LUND. 1973. Anamolous bilateral corticofugal pathways in albino rats after neonatal lesions. Brain Res. 62: 218-221. LEVI-M• NTALCINI, R. 1975. Regulatory processes in the sympathetic adrenergic neuron. Pages 437-448 in M. SANTINI, Ed., Go& Centennial Symposium Proceedings. Raven, New York. LEVI-M• NTALCINI, R. 1976. The nerve growth factor: its widening role and place in neurobiology. Adv. Biochem. Psychopharmacoi. 15: 237-250. LIEBERMAN, A. R. 1971. The axon reaction: a review of the principal features of perikaryal responses to axon injury. Int. Rev. Neurobiol. 14: 49- 124. LIEBERMAN, A. R. 1974. Some factors affecting retrograde neuronal responses to axonal lesions. Pages 71- 105 in R. BELLAIRS AND E. G. GRAY, Eds., Essays on the Nervous System. Clarendon Press, Oxford. Lou, R., C. W. COTMAN, AND G. LYNCH. 1977. The development of afferent lamination in the fascia dentata of the rat. Brain Res. 121: 229-243. LUND, R. D. 1972. Anatomic studies on the superior colliculus. Invest. Ophthalmo/.

66. LANDMESSER, during in vivo 67.

68. 69.

70. 71. 72. 73. 74. 75. 76.

11: 434-441.

77. LUND, R. D. 1978. Development and Plasticity of the Brain, p. 370. Oxford University Press, New York and London. 78. LUND, R. D., T. J. CUMMINGHAM, AND J. S. LUND. 1973. Modified optic projections after unilateral eye removal in young rats. Brain Behav. Evol. 8: 51-72. 79. LUND, R. D., AND S. D. HAUSCHKA. 1976. Transplanted neural tissue develops connections with host rat brain. Science 193: 582-584. 80. LUND, R. D., AND J. S. LUND. 1973. Reorganization of the retinotectal pathway in rats after neonatal retinal lesions. Exp. Neurol. 40: 377-390. 81. LUND, R. D., AND J. S. LUND. 1976. Plasticity in the developing visual system: the effects of retinal lesions made in young rats. J. Comp. Neural. 169: 133-154. 82. MATINIAN, L. A., AND A. S. ANDREASIAN. 1976. Enzyme Therapy in Orgunic Lesions of the Spinal Cord (Elena Tanasescu, Trans.), pp. v-156. UCLA Brain Information Service, Los Angeles, Calif. 83. MILLER, B. F., AND R. D. LUND. 1975. The pattern of retinotectal connections in albino rats can be modified by fetal surgery. Brain Res. 91: 119- 125. 84. MURRAY, M., AND B. GRAFSTEIN. 1969. Changes in the morphology and amino acid incorporation of regenerating goldfish optic neurons. Exp. Neural. 23: 544-560. 85. MUSTARI, M. J., AND R. D. LUND. 1976. An aberrant crossed visual corticotectal pathway in albino rats. Brain Res. 112: 37-42. 86. NADLER, J. V., C. W. COTMAN, AND G. LYNCH. 1977. Histochemical evidence of altered development of cholinergic fibers in the rat dentate gyrus following lesions. I. Time course after complete unilateral entorhinal lesion at various ages. J. Comp. Neural.

171: 561-588.

AXONAL

GROWTH

MECHANISMS

695

87. NADLER, J. V., C. PAOLETTI, C. W. COTMAN, AND G. LYNCH. 1977. Histochemical evidence of altered development of cholinergic fibers in the rat dentate gyrus following lesions. II. Effects of partial entorhinal and simultaneous multiple lesions. J. Camp. Neural. 171: 589-604. 88. NICOLSON, G. L. 1976. Transmembrane control of the receptors on normal and tumor cells. I. Cytoplasmic influence over cell surface components. Biochim. Biophys. Acfa 457: 57- 108. 89. NICOLSON, G. L., AND Cl. POSTE. 1976. Cell shape changes and transmembrane receptor uncoupling induced by tertiary amine local anesthetics. J. Supramol. Struct. 5: 65-72. 90. NICOLSON, G. L., G. POSTE, AND T. H. JI. 1977. The dynamics of cell membrane organization. Pages 1-73 in G. POSTE AND G. L. NICOLSON, Eds. Cell Surface Reviews, Vu/. 3: Dynamic Aspects of Cett Surface Organization. North-Holland, Amsterdam. 91. NICHOLSON, G. L., J. R. SMITH, AND B. POSTE. 1976. Effect oflocal anesthetics on cell morphology and membrane-associated cytoskeletal organization in BALB/3T3 cells. J. Cell Biol. 68: 395-402. 92. NJ.&, A., AND D. PURVES. 1978. The effects of nerve growth factor and its antiserum on synapses in the superior cervical ganglion of guinea-pig. J. Physiol. (Londoni 277: 53-75. 93. NORDLANDER, R. H., AND M. SINGER. 1972. Electron microscopy of severed motor fibers in the crayfish. Z. Zelljorsch. 126: 157- 181. 94. NORDLANDER, R. H., AND M. SINGER. 1973. Degeneration and regeneration of severed crayfish sensory fibers: an ultrastructural study. J. Camp. Neural. 152: 17% 192. 95. NORDLANDER, R. H., AND M. SINGER. 1978. The role of ependyma in regeneration of the spinal cord in the urodele amphibian tail. J. Comp. Neural. 180: 349-374. 96. OBATA, K. 1974. Transmitter sensitivities of some nerve and muscle cells in culture. Bruin Res. 73: 71-88. 97. GSTBERC, A.-J. C., G. RAISMAN, P. M. FIELD, L. L. IVERSEN, AND R. E. ZIGMOND. 1976. A quantitative comparison of the formation of synapses in the rat superior cervical sympathetic ganglion by its own and by foreign nerve fibers. Bruin Res. 107: 445-470. 98. PAPAHADJOPOULOS, D. 1972. Studies on the mechanism of action of local anesthetics with phosphohpid model membranes. Biochim. Biophys. Acra 265: 169- 186. 99. PATTERSON, P. 1978. Environmental determination of autonomic neurotransmitter function. Annu. Rev. Neurosci. 1: l- 17. 100. PATTERSON, P., D. POTTER, AND E. FURSHPAN. 1978. The chemical differentiation of nerve cells. Sci. Am. 239(No. 7): 50-59. 101. PETERS. A.. S. L. PALAY, AND H. DEF. WEBSTER. 1976. The Fine Structure of the Nervous System. The Neurons atzd Supporting Cells. W. B. Saunders & Co., Philadelphia. 102. PETTEGREW, R. K., AND W. F. WINDLE. 1976. Factors in recovery from spinal cord injury. Exp. Neural. 53: 815-829. 103. PILAR, G., AND L. LANDMESSER. 1972. Axotomy mimicked by localized colchicine application. Science 177: 1116- 1118. 104. PILAR, G., AND L. LANDMESSER. 1976. Ultrastructural differences during embryonic cell death in normal and peripherally deprived ciliary ganglia. J. Cell Biol. 68: 339-356. 105. PILAR, G., L. LANDMESSER, AND L. BURSTEIN. 1979. Competition for survival among developing ciliary ganglion cells. J. Neurophysiol., in press.

696

VERAA,

GRAFSTEIN,

AND

ROSS

106. POSTE, G., AND G. L. NICOLSON. 1976. Calcium ionophores A23187 and X537A affect cell agglutination by lectins and capping of lymphocyte surface immunoglobulins. Biochim. Biophys. Acta 426: 148-155. 107. POSTE, G., D. PAPAHADJOPOLJLOS, AND G. L. NICOLSON. 1975. The effect of local anesthetics on transmembrane cytoskeletal control of cell surface mobility and distribution. Proc. Natl. Acad. Sci. U.S.A. 12: 4430-4434. 108. PURVES, D. 1975. Functional and structural changes in mammalian sympathetic neurones following interruption of their axons. J. Physiol. (London) 252: 429-463. 109. PURVES, D. 1976. Functional and structural changes in mammalian sympathetic neurones following colchicine application to postganglionic nerves. J. Physiol. (London) 259: 159- 175. 110. PURVES, D. 1976. Long term regulation in the vertebrate peripheral nervous system. Pages 125-177 in R. PORTER, Ed., International Review of Physiology, Neurophysiology II, Vo/. 10. University Park Press, Baltimore/London/Tokyo. 111. PURVES, D., AND A. NJ&. 1978. Trophic maintenance of synaptic connections in autonomic ganglia. Pages 27-48 in C. W. COTMAN, Ed., Neuronal Plasticity. Raven, New York. 112. RAMBOURG, A., AND C. P. LEBLOND. 1967. Electron microscopic observations on the carbohydrate rich cell coat present at the surface of cells in the rat. J. Cell Biol. 32: 27-53. 113. RAM~N y CAJAL, S. 1929. Degeneration and Regeneration of the Nervous System (R. M. May, Trans. and Ed.), Vol. I, pp. xix-396; Vol. II, viii-373. Oxford University Press, London. 114. REES, R. P., M. B. BUNGE, AND R. P. BUNGE. 1976. Morphological changes in the neurite growth cone and target neuron during synaptic junction development in culture. .I. Cell Biol. 68: 240-263. 115. REIER, P. J., AND H. DEF. WEBSTER. 1974. Regeneration and remyelination ofxenopus tadpole optic nerve fibers following transection or crush. J. Neurocytol. 3: 591-618. 116. REIS, D. J., G. GILAD, V. M. PICKEL, AND T. H. JOH. 1978. Reversible changes in the activities and amounts of tyrosine hydroxylase in dopamine neurons of the substantia nigra in response to axonal injury as studied by immunochemical and immunocytochemical methods. Brain Res. 144: 325-342. 117. REIS, D. J., Ross, R. A., AND T. H. JOH. 1974. Some aspects of the reaction of central and peripheral noradrenergic neurons to injury. Pages 109-125 in K. FUXE, L. OLSON, Neurons.

AND

Y.

ZOTTERMAN,

Eds.,

Dynamics

of Degeneration

and

Growth

in

Pergamon Press, Oxford and New York. 118. Ross, R. A., T. H. JOH, AND D. J. REIS. 1975. Reversible changes in the accumulation and activities of tyrosine hydrolase and dopamine-P-hydroxylase in neurons of nucleus locus coeruleus during the retrograde reaction. Brain Res. 92: 57-72. 119. ROTTER, A., N. J. M. BIRDSALL, A. S. V. BURGEN, P. M. FIELD, AND G. RAISMAN. 1977. Axotomy causes loss of muscarinic receptors and loss of synaptic contacts in the hypoglossal nucleus. Nature (London) 266: 734-735. 120. SCHNEIDER, D. 1972. Regenerative phenomena in the central nervous system. J. Neurosurg. 37: 129- 136. 121. SCHNEIDER, G. E. 1970. Mechanisms of functional recovery following lesions of visual cortex or superior colliculus in neonate and adult hamsters. Brain Behav. Evol. 3: 295-323. 122. SCHNEIDER, G. E. 1973. Early lesions of superior colliculus: factors affecting the formation of abnormal retinal projections. Brain Behav. Evol. 8: 73- 109. 123. SCHNEIDER, G. E. 1975. Discussion. Pages 56-59, 223-226 in Outcome of Severe

AXONAL Damage

to the Central

Nervous

GROWTH MECHANISMS System,

697

Ciba Foundation Symposium 34. Elsevier,

Amsterdam. 124.

G. E. 1977. Growth of abnormal neural connections following focal brain lesions: constraining factors and functional effects. Pages 5-26 in W. H. SWEET, S. OBRADOR, AND J. G. MARTIN-RODRIGUEZ, Eds., Neurological Treatment in Psychiatry. Pain and Epilepsy. University Park Press, Baltimore. 125. SCHNEIDER, G. E., AND S. R. JHAVERI. 1974. Neuroanatomical correlates of spared or altered function after brain lesions in the newborn hamster. Pages 65- 109 in D. G. STEIN, J. J. ROSEN, AND N. BUTTERS, Eds., Plasticity and Recovery of Function in the Central Nervous System. Academic Press, New York. 126. SEEMAN, P. 1972. The membrane actions of anesthetics and tranquilizers. Pharmacol. SCHNEIDER,

Rev.

24: 583-655.

127. SHEA, J. R. 1970. A method forin situ cytophotometric estimation of absolute amount of ribonucleic acid using azure B. J. Histochem. Cytochem. 18: 143- 152. 128. SINGER, M., N. KRISHNAN, AND D. A. FYFE. 1972. Penetration of ruthenium red into peripheral nerve fibers. Anat. Rec. 173: 375-390. 129. SINGER, M., AND M. M. SALPETER. 1966. Transport of tritium-labeled I-histidine through the Schwann and myelin sheaths into the axon of peripheral nerves. Nature (London)

210:

1225-

1227.

130. SINGER, S. J., AND G. L. NICOLSON. 1972. The fluid mosaic model of the structure of cell membranes. Science 175: 720-731. 131. SPERRY,R. W. 1965. Embryogenesis of behavioral nerve nets. Pages 161-186 in R. L. DEHAAN AND H. URSPRUNG, Eds., Organogenesis. Holt, Rinehart and Winston, New York. 132. SUMNER, B. E. H., AND F. I. SUTHERLAND. 1973. Quantitative electron microscopy on the injured hypoglossal nucleus in the rat. J. Neurocytol. 2: 315-328. 133. SUMNER, B. E. H.,AND W. E. WATSON. 1971. Retractionandexpansionofthedendritic tree of motor neurones of adult rats induced in vivo. Nature (London) 233: 273-275. 134. WATSON, W. E. 1965. An autoradiographic study of the incorporation of nucleic acid precursors by neurones and glia during nerve regeneration. J. Physiol. (London) 180: 741-753. 135. WATSON, W. E. 1%8. Observations on the nucleolar and total cell body nucleic acid of injured nerve cells. J. Physiol. (London) 1%: 655-676. 136. WATSON, W. E. 1972. Some quantitative observations upon the responses of neuroglial cells which follow axotomy of adjacent neurones. J. Physiol. (London) 225: 415-435. 137. WATSON, W. E. 1973. Some responses of neurones of dorsal root ganglia to axotomy. J. Physiol.

(London)

231: 41-42.

138. WATSON, W. E. 1974. Cellular responses to axotomy and to related procedures. Br. Med. Bull. 30: 112- 115. 139. WEBSTER, H. DEF., P. J. REIER, J.-M. MATTHIEU, AND R. H. QUARLES. 1976. Axonal abnormalities in optic nerves of Jimpy mice. J. Neuropathol. Exp. Neural. 35: 309. 140. WEISS, P. 1961. The concept of perpetual neuronal growth and proximodistal substance convection. Pages 220-250 in S. S. KETY AND J. ELKES, Eds., Regional Neurochemistry. Pergamon Press, New York. 141. WILLIAMS, A. K., P. WOOD, AND M. B. BUNGE. 1976. Evidence that the presence of Schwann cell basal lamina depends upon interaction with neurons. J. Cell Biol. 70: 138a (abstract). 142. WOOD, P. 1976. Separation of functional Schwann cells and neurons from normal peripheral nerve tissue. Brain Res. 115: 361-375.

VERAA,

698

GRAFSTEIN,

AND ROSS

143. WOOD, P., AND R. BUNGE. 1975. Evidence that sensory axons are mitogenic for Schwann cells. Nature (London) 265: 662-664. 144. YOUNG, R. W. 1971. The renewal of rod and cone outer segments in the rhesus monkey. J. Cell Biol. 49: 303-318. 145. YOUNG, R. W. 1973. Renewal systems in rods and cones. Ann. Ophthalmol. 5: 843-854.

146. YOUNG, 147. 148.

R. W. 1974. Biogenesis and renewal of visual cell outer segment membranes. 18: 215-223. 1976. Visual cells and the concept of renewal. Invest. Ophthalmol. 15:

Exp. Eye Res. YOUNG, R. W. 700-725. YOUNG, R. W. 573-578.

1978. Visual cells, daily rhythms, and vision research. Vision Res. 18: