INFORMATION TO USERS The most advanced technology has been used to photo graph and reproduce this manuscript from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are re produced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. These are also available as one exposure on a standard 35mm slide or as a 17" x 23" black and white photographic print for an additional charge. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600
Order Number 8913623
Analyses of the response of dorsal root afferents following dorsal funiculus lesions and an examination of the sacral parasympathetic nucleus in amphibians
Campbell, H. Lee, Ph.D.
The Ohio State University, 1989
UMI 300 N. Zceb Rd. Ann Arbor, MI 48106
ANALYSES OF THE RESPONSE OF DORSAL ROOT
AFFERENTS FOLLOWING DORSAL FUNICULUS LESIONS AND
AN EXAMINATION OF THE SACRAL PARASYMPATHETIC
NUCLEUS IN AMPHIBIANS
DISSERTATION
Presented in Partial Fulfillment of the
Requirements for the Degree Doctor of Philosophy in
the Graduate School of the Ohio State University
By
H. LEE CAMPBELL, B.S.
* * * * *
The Ohio State University
1989
Dissertation Committee: Approved By
Jacqueline C. Bresnahan, Ph.I
Michael S. Beattie, Ph.D. Adviser James S. King, Ph.D. Department of Anatomy
George F. Martin, Ph.D. For the God who exists and is not silent ACKNOWLEDGEMENTS
I would like to thank Dr. Jacqueline C. Bresnahan and
Dr. Michael S. Beattie for sharing their time, knowledge, creativity and good humor with me.
I am entirely in the debt of Tina Van Meter, David
Norris, Dr. Gail Leedy and John Komon each for their own exceptional technical assistance and each for their friendship.
The faculty of the Department of Anatomy has proved to be of great help and encouragement, especially Dr.
James S. King and Dr. George F. Martin.
The loyal companionship of Tim Lipovsky has provided much in the way of acceptance, solace and motivation.
My parents have proved themselves without equal in their commitment to my well being and unshakable confidence. I am delighted to be called their son.
Finally, I would like to thank Debby for her patience, encouragement, inclination toward laughter and unselfish love.
1 1 1 VITA
Date of Birth...... July 8, 1954
Education...... B.S. in Biology, Oakland University, 1981 Ph.D. in Anatomy, The Ohio State University, 1988
Committees...... Graduate Committee, Dept, of Anatomy, 1987-1988 Council of Graduate Students, The Ohio State University, 1983-1984
Teaching Experience.. Gross Anatomy for Physical Therapists and Occupational Therapists, 1982-1983 Neuroanatomy for Medical Students, 1983, 1988 Histology for Medical Students, 1987-1988
PUBLICATIONS
Campbell, H.L., F.J. Liuzzi, M.S. Beattie, and J.C. Bresnahan (1982) Large circumferential cells of the developing Rana catesbeiana spinal cord are labelled after HRP application to lateral funiculus. Neuroscience Abstracts, 23 5.5, 821.
Campbell, H.L., M.S. Beattie, J.C. Bresnahan (1984) The response of dorsal root afferent fibers to dorsal funiculus lesions in developing rana catesbeiana tadpoles. Neuroscience Abstracts, 298.7, 1024.
Beattie, M.S., J.C. Bresnahan, H.L. Campbell, K.M. Conway, F.J. Liuzzi, and G. Lopate (1985) Studies of
iv growth and regeneration in developing anuran spinal cord. Paper presented: Neural Development, Plasticity & Regeneration Conference. Columbus, Ohio April 19-21.
Campbell, H.L., M.S. Beattie, J.C. Bresnahan (1987) Circumferential cells of the developing Rana catesbeiana lumbar spinal cord. Anatomy and Embryology 176:155-163.
Campbell, H.L., M.S. Beattie, and J.C. Bresnahan (1988) The response of dorsal root afferents to dorsal funicular lesions in developing and juvenile Xenopus laevis. American Association of Anatomists Abstracts, 23ID, 19a.
Campbell, H.L., M.S. Beattie, and J.C. Bresnahan (1988) The response of spared dorsal root afferent collaterals following dorsal funiculus lesions in juvenile xenopus laevis. Soc. Neurosci. Abstracts, 467.7, 1170.
FIELDS OF STUDY
Major Field: Neuroanatomy
Minor Fields: Anatomy, Histology
v TABLE OF CONTENTS
DEDICATION...... ii
ACKNOWLEDGEMENTS...... iii
VITA...... iv
LIST OF FIGURES...... viii
ABBREVIATIONS...... xi
INTRODUCTION...... 1
CHAPTER PAGE
I. A REVIEW OF ANURAN SPINAL CORD NEUROANATOMY...... 2
1.1 Introduction...... 2 1.2 Spinal Gray...... 4 1.3 Distribution of primary afferents...... 24 1.4 Segmental interconnections... 28 1.5 Ascending spinal systems..... 29 1.6 Descending supraspinal systems...... 35 1.7 Developmental considerations. 42 1.8 Conclus ion...... 44
II. REGENERATION AND PLASTICITY IN THE DEVELOPING AND ADULT AMPHIBIAN SPINAL CORD...... 46
2.1 Introduction...... 46 2.2 Rana catesbeiana...... 57 2.2.1 Methods...... 59
vi CHAPTER PAGE
II.
2.2.2 Results...... 63 2.2.3 Discussion...... 68 2.2.4 Conclusion...... 73
2.3 Xenopus laevis...... 75 2.3.1 Methods...... 76 2.3.2 Results...... 81 2.3.3 Discussion...... 89 2.3.4 Conclusion...... 99
III. OBSERVATIONS ON THE XENOPUS LAEVIS SACRAL SPINAL CORD...... 100
3.1 Introduction...... 100 3.2 Sacral parasympathetic nucleus...... 101 3.2.1 Methods...... 103 3.2.2 Results...... 104 3.2.3 Discussion...... 107
ILLUSTRATIONS...... 110
BIBLIOGRAPHY...... 165 LIST OF FIGURES
PAGE
Ranid experimental design...... 111
Illustration of the normal distribution of the ninth DR ...... 113
Blastema cells labeled with tritiated thymidine ...... 115
HRP labeled DRG axons in a blastema implant...... 117
Illustrations of aberrant axonal traj ectories...... 119
Micrographs of aberrant axonal trajectories...... 121
Illustrations of growth cones and retraction bulbs...... 123
Micrographs of aberrant fiber morphologies...... 125
Illustration of Xenopus experimental design...... 127
An illustration of the process used to digitize an image for analysis... 129
Normal distribution of tenth root primary afferents...... 131
Normal rostrocaudal distribution of tenth root primary afferents...... 133
viii PAGE
Micrographs showing distributions of DR afferents in the DTF and VTF.... 135
Illustration of the range of injury sites sizes...... 137
Micrographs of a root entry zone near an injury site...... 139
A graph showing the detected area for the DTF and VTF across groups... 141
Graphs showing the optical density of the DTF and VTF across groups.... 143
Graphs showing varicosity number and size distribution across groups.... 145
Figures showing examples of crossing fibers in normal, 3 day, 5 week, and 13 week cases...... 147
A graph showing the lengths of contralateral axons across groups... 149
Graphs showing contralateral axonal number and number of branch points.. 151
Illustrations showing the normal distribution of dorsal roots 10, 9 8, 7, 6 and also of the rubrospinal and raphespinal systems...... 153
Micrographs showing the distribution of elements at the tenth segment after thoracic HRP application..... 155
Photographs showing the surgical procedure used for labeling sacral spinal cord neurons...... 157
Neuronal elements labeled after HRP application to the tenth spinal nerve...... 159
ix FIGURE PAGE
62 Illustration showing neurons labeled after HRP application to the tenth nerve...... 161
63 A stereo-pair showing the distribution of cell bodies after HRP application to the tenth nerve.. 163
X ABBREVIATIONS
CF central field
CNS central nervous system
DC dorsal columns
DF dorsal field
DRG dorsal root ganglion or 'DR' for dorsal root
DTF dorsal terminal field [a.k.a. dorsal neuropil]
EM electron microscopy
GFAP glial fibrillary acidic protein
LCC large circumferential cells
LF lateral field
LMC lateral motor column
LT Lissauer's tract
MMC medial motor column
REZ root entry zone
VL ventrolateral field
VM ventromedial field
VTF ventral terminal field [a.k.a. ventral neuropil or lateral terminal field]
xi INTRODUCTION
The first chapter of this dissertation is a review of amphibian spinal cord neuroanatomy. A synopsis of available information about the spinal cord seemed necessary in order to provide a context for chapters II and III. Significant contributions to the understanding of frog spinal neuroanatomy have been made since
Ebbesson's ['76] thorough review. Chapter II contains the results of two experiments concerned with plasticity of primary afferents after dorsal column lesions in Rana catesbeiana and Xenopus laevis. Chapter III presents observations on the sacral spinal cord of juvenile
Xenopus laevis after HRP application to the 10th spinal nerve and pelvic viscera.
1 CHAPTER I
1.1 INTRODUCTION
The central nervous system [CNS] of amphibians is particularly useful for investigations of development and plasticity. Since frogs develop entirely in an external environment, their CNS is available for study and manipulation throughout embryonic and larval stages.
Questions pertaining to regeneration and plasticity can be profitably addressed because, within phylum chordata, the class amphibia has one of the widest ranges of regenerative responses to CNS injury.
Over a decade has passed since a neuroanatomic review of the amphibian spinal cord was written
[Ebbesson, '76]. Therefore, an attempt has been made here to assemble what is currently known about the subject. In addition, an effort has been made to compile a limited synopsis of spinal cord development where appropriate.
The adult anuran spinal cord occupies about half the length of the vertebral canal. It has eleven segments,
2 3
although the first roots are lost during development in
Xenopus laevis [Hughes and Tschumi, '58 and Thors et al.f
'80], Rana esculenta [Gaupp, 1896] and presumably in other anurans too. For this reason, most investigators number the first apparent adult spinal nerve as spinal nerve 2 [ten Donkelaar et al., '81]. However, some authors refer to ten segments in adult frogs with the first spinal nerve numbered 1, following the anatomic studies in Rana catesbeiana by Whitehouse and Grove ['47]
[for a discussion of this problem see Deucher, '75].
The cord has two obvious enlargements that develop concomitantly with the upper and lower limbs. According to ten Donkelaar ['81], spinal nerves 3 and 4 compose the brachial plexus in Xenopus laevis frogs. Silver ['42] asserted that a small branch of spinal nerve 2, plus spinal nerves 3 and 4 make up the brachial plexus in Rana pipiens. The Rana catesbeiana second spinal nerve is the sole innervation to the forelimb [Frank and Westerfield,
'82a; Jhaveri and Frank, '83 see also Joseph and
Whitlock, '68].
The lumbar enlargement consists of segments eight, nine and ten. Silver ['42] concluded that segment eleven contributes to the lumbar plexus, as well. Spinal nerves
8, 9, 10 and 11 combine in a variety of ways to make up 4 the lumbar plexus in ranids [Cruce, '74 and Lee & Farel,
' 88]. The various spinal nerve contributions may exclude
8 (rarely) or 11 (more often). The eleventh dorsal and ventral roots are not always present; their axons presumably exit the tenth roots in those cases.
1.2 SPINAL GRAY
The spinal gray of the adult amphibian lumbar enlargement can be parceled, according to cytoarchitectonic and hodological features, into the following fields: 1) the dorsal field [DF]; 2) the lateral field [LF]; 3) the lateral motor column [LMC] in the cervical [brachial] and lumbar enlargements; 4) the medial motor column; 5) the ventrolateral field [VL]; 6) the ventromedial field [VM] and 7) the central field [CF]
[Ebbesson, '76] [Fig. 2].
DORSAL FIELD
The dorsal field [Fig. 2], as characterized by
Ebbesson [Ebbesson, '76] corresponds to Rexed's laminae
I-IV [Rexed, '64]. Comparative neuroanatomic considerations of the distribution of cutaneous primary afferents in laminae I-V of the cat [Light and Perl, '79] and the frog dorsal terminal field [DTF] [Szekely et al.,
'82; Jhaveri and Frank, '83] suggest that they are comparable. Lamina V might also be included in the dorsal field. The substantia gelatinosa [SG] has been
difficult to distinguish in anurans. Until recent HRP
analysis of the distribution of cutaneous afferents , it
was held that the SG was either a thin strip of neuropil
along the lateral aspect of the dorsal horn [Silver, '42
called it the dorsal neuropil; Ebbesson, '76] or nearly
the entire dorsal field itself [Athias, 1897; Szekely,
'76a; Antal et al., '80; and Nikundiwe et al., '82]. Now
it is believed that Silver's dorsal neuropil corresponds to the SG [Szekely et al., '82; Jhaveri and Frank, '83].
This region contains small dendritic profiles 1-2 urn in diameter with closely approximated axonal elements and many astrocytic and ependymal processes. A specialized
synaptic arrangement has also been described that consists of a central, vesicle filled axon terminal in contact with small dendritic profiles and other axonal elements [Sotelo and Grofova, '76]. This glomerular organization is like the mammalian substantia gelatinosa where the central axonal profile is a primary afferent
[e.g. Beattie et al, '78; Ralston and Daly-Ralston, '81].
Two types of cells can be seen in the adult dorsal field:
1) circumferential cells and 2) cells within the gray matter that have a dorsal dendritic arbor.
Circumferentially oriented cells are evident at the outer perimeter of the dorsal field. In frontal sections their 6
two largest processes can be observed circumscribing the gray matter. This neuron is evident in embryonic and
larval anurans [Van Gehuchten, 1898; Athias, 1897;
Forehand and Farel, '82a & b; Roberts, '82; Campbell et al., '82 & '87]. These observations, along with the tendency for neurons in the periphery of the gray matter to be pigment bearing [Silver, '42 see also Hughes, '63 and other studies of peripherally placed somata Thors et al., '82 and van Meir, '86] indicate that this cell type has its origins early in development. At least three types of circumferential cell are evident in the dorsal field. One, the dorsolatoral commissural cell has been described in embryonic Xenopus [Roberts and Clarke,
'82a]. Its other dendrites extend, radially and longitudinally, into the lateral and dorsal funiculi.
Its axon crosses ventrally and ascends in the ventrolateral part of the white matter, at least to the hindbrain. It is believed to participate in embryonic swimming by phase coupling the motoneuron bursts from either side [Kahn and Roberts, '82]. Similar elements have been described in larval Rana [the cellules commissurale and/or the cellule marginale commissurale of
Athias, 1897; Campbell et al., '87] and ambystoma [the commissural cell of Herrick, #15]. No evidence exists 7
that this neuronal element persists after metamorphosis.
A second dorsal field circumferential cell, has been
termed the dorsal marginal cell [Forehand and Farel,
'82a]. Its axon ascends, ipsilaterally, to the caudal
hindbrain. Comparable elements have been described in
other amphibians [the cellule du cordon lateral of
Athias, 1897; Sala y Pons, 1892; Campbell et al., '87].
The dorsal marginal cell persists into adulthood.
Finally, a multipolar dorsomedian cell, has been termed
the dorsal commissural cell [Forehand and Farel, '82a].
This element can be labeled with HRP when the application
site is local, (i.e. within 2-3 segments). This class of
cell may need to be subdivided, for two dorsomedian
neurons, heterologous in other ways, were described in
salamanders [Van Gehuchten, 1897]. Van Gehuchten traced
the axon of one into the ipsilateral lateral funiculus
and the axon of another, across the cord, into the ventral funiculus. There is no evidence that this
segmental interneuron persists into adulthood. Of the three marginal cells described, only the dorsal marginal
cell is known to persist into adulthood. It sends an
ipsilateral ascending axon, through the lateral
funiculus, to the rhombencephalon [Forehand and Farel,
'82a].
Neurons with predominantly dorsal dendritic arbors 8
are the second general group of neurons observed in the
dorsal field [Ebbesson, '76]. These cells extend their
dendrites into the substantia gelatinosa and into the
dorsal funiculus, one subgroup has short dendrites, mainly limited to the dorsal field. Another subgroup has medial dendrites that cross the midline to arborize in
the contralateral dorsal field. From Golgi
impregnations, Athias [1897] described dorsomedial
neurons with dendrites extending into the dorsal
funiculus and their axons entering the dorsal and lateral
funiculi. The medial dorsomedial cells projected
dendrites into the dorsal funiculus and the lateral dorsomedial cells into the lateral funiculus. Larger cells in the same region sent axonal processes into the
lateral field [Athias, 1897, called them the cellules de
la substance gelatineuse de Rolando]. The relationship between Athias' dorsomedial and SG cells and those described by Ebbesson is not clear. A propriospinal cell, the dorsolateral neuron, has been described in embryonic Xenopus [Roberts and Clarke, '82a]. This cell is multipolar with longitudinal dendrites that extend rostrocaudally in the dorsal funiculus and the dorsal part of the lateral funiculus. Its axon ascends ipsilaterally in the dorsal aspect of the lateral 9 funiculus to the level of the rostral spinal cord.
There are few synaptic contacts on amphibian dorsal field neuronal cell bodies [Sotelo and Grofovd, '76]. It seems, therefore, that in anura most neural integration takes place on the extensive dendritic arbors of a limited number of cell types rather than through synaptic contact upon the proximal dendrites and somata of highly specialized neurons.
Substance P-, Met-enkephalin- and somatostatin-like immunoreactivity was found in the dorsal field of Rana esculenta [Lorez and Kemali, '81]. Of particular interest was the finding that all three peptides could be localized in neurons of the dorsal field, which, taken with the neuroanatomic evidence, argues for the presence of a peptidergic ascending propriospinal/supraspinal system [ibidem].
Cutaneous afferent collaterals terminate in the dorsal horn in an oval shaped dorsal terminal field [DTF]
[Szekely et al., '82; Jhaveri and Frank, '83] [Figs. 2,
31 & 35]. Cutaneous afferents from the dorsal skin surface of frogs preferentially terminate in the lateral part of DTF and those innervating ventral skin tend to terminate in the medial part of DTF. However, significant numbers of collaterals from both regions extend across the entire DTF [Szekely et al., '82]. In 10
the first two segments, trigeminal afferents also terminate in the dorsal field [ibidem]. Substance P-like
immunoreactivity is present in primary afferents [Adli et al, '88]. Somatostatin-like immunoreactivity is probably present in primary afferents also [Lorez and Kemali, 781;
Adli et al, 788].
LATERAL FIELD
The lateral field, as characterized by Ebbesson, probably corresponds to laminae V-VI of mammals
[Ebbesson, 776] [Figs. 2, 31 & 35]. However, the distribution of cutaneous afferents in the dorsal field and muscle afferents in the lateral field might support the notion that the lateral field corresponds to lamina
VI and VII [Liuzzi et al., 784],
There is no cytoarchitecturally obvious intermediolateral cell column in anurans [Nieuwenhuys,
764; Kusuma and ten Donkelaar, 779], however, neuronal elements constituting this structure were identified in xenopus after HRP application to thoracic ventral roots
[Nikundiwe et al., 782]. These preganglionic cells were more completely examined by application of HRP to the first sympathetic ganglion in Rana pipiens [Robertson,
787]. The neurons form two distinct columns of cells, a medial and a lateral group. The lateral group 11
corresponds to the mammalian principle intermediolateral
cell column and the medial group to the intercalated nucleus. The lateral group of cells was distributed homogeneously in a rostrocaudal direction and was seen to extend dendritic arborizations into the lateral
funiculus. The medial group of cells was divided into two groups: l) A medial nucleus intercalatus, whose dendrites are rostrocaudal in orientation, contains cells which are morphologically similar to those in n.
intercalatus pars paraependymalis in turtles [Leong et al., 783] and mammals [Petras and Cummings, '72?
Dalsgaard and Elfvin, '81]. 2) A lateral nucleus intercalatus was observed with cellular architecture reminiscent of that seen in mammals, with the exception that Rana pipiens did not have transverse cellular bands
[Robertson, '87], 65% of the preganglionic axons exit the third spinal nerve [Robertson, '87]. Since earlier studies by Pick ['57] concluded that sympathetic preganglionic axons exit spinal nerves two through ten,
HRP labeling of other sympathetic ganglia would be expected to reveal sympathetic preganglionic neurons throughout the amphibian spinal cord.
Ebbesson ['76] described small medial neurons and large lateral neurons in the lateral field. Small medial cells were entirely limited to the lateral field. Some 12
of the medial cells in Ebbesson's description may be constituents of n. intercalatus. Large lateral neurons were described as having dendrites that extended into all
fields and into the contralateral ventral field as well.
Embryonic Xenopus laevis has a laterally placed, unipolar commissural neuron, with dendrites that radiate into the lateral funiculus. The axon of this cell type crosses through the ventral commissure and then ascends and descends in the contralateral ventral funiculus [Roberts and Clarke, '82a]. The commissural neuron is similar to the cellule des cordons heteromeres of larval salamander
[Van Gehuchten, 1898], the correlation neuron of larval
Ambystoraa [Herrick, #15], or the cellule marginale commissurale of larval Rana temporaria [Athias, 1897].
This neuron, along with the dorsolateral commissural neuron from the dorsal field, are thought to participate in coupling the swimming response in the embryo [Kahn and
Roberts, '82]. There is no evidence that it persists into adulthood.
Another cell, evident in embryonic Xenopus, is a laterally located, multipolar interneuron with its dendrites extending radially into the lateral funiculus.
This propriospinal cell projects its dendrite caudally in the ipsilateral lateral funiculus [Roberts and Clarke, 13
'82a], An ascending ipsilateral axonal projection from
unipolar neurons is also observed in embryonic xenopus.
The axon ascends to the rhombencephalon [ibidem]. Athias
[1897] described two neuronal types in Golgi material
that he called cellules marginale and cellules du cordon
anterieur. Both of these cells are like the
anterolateral cell described by Forehand and Farel ['82
a&b] except for their more dorsal placement. Campbell et
al. ['82 and '87] have shown several examples of these
lateral field circumferential cells as well.
In the cervical and lumbar cord muscle primary
afferents terminate in a triangular shaped region, termed variously, the "ventral neuropil" [Jhaveri and Frank,
'83] "lateral" [Szekely, '76a] or "ventral terminal
field" [VTF] [Liuzzi et al., '84]. A lateral bundle of
la muscle spindle afferents continues from the VTF, along
the margin of the gray matter, into the lateral motor
column in lumbar cord [Szdkely, '76a] but it is not as prominent at brachial levels [Jhaveri and Frank, '83],
Ultrastructural observations of the lateral field have
shown the presence of large terminal and en passant
swellings which contain round vesicles. These elements
are generally apposed to dendrites and display both
synapses and gap junctions [Sotelo and Taxi, '70; Sotelo
and Grofova, '76]. Muscle primary afferents contact, 14
among other things, the dendrites of motoneurons
[Lichtman from Jhaveri and Frank, '83; Liuzzi et al '84 &
'85]. The lateral field receives rubrospinal, reticulospinal and raphespinal projections.
CENTRAL FIELD
The central field corresponds to lamina X [Ebbesson,
'76]. It is an area rich in astrocytic soraatae [Sasaki and Mannen, '81] and at least three types of neurons.
Roberts and Clarke ['82a] noted a ciliated ependymal cell in embryonic xenopus with an axon that projected, ipsilaterally in the ventral funiculus, to the hindbrain.
A comparable element has been described in adult Rana pipiens, with extensive dendritic arborizations in the ventral funiculus [Silver, '42]. Golgi preparations have revealed the other two neuronal types in adult frog
[Ebbesson, '76]. A fusiform cell with long thin dendrites invading adjacent regions, and an isodendritic cell with ventral dendrites located in the central field and dorsal dendrites in the dorsal field. The dorsal commissural nucleus evident in turtle [Kusuma and ten
Donkelaar, '79] and mammals [Petras and Cummings, '72;
Dalsgaard and Elfvin, '81] has not been identified in anurae.
Some primary afferents end in the central field 15
[Liuzzi et al, '84 see also Figs. 2, 31 and 35] particularly at thoracic segments from thoracic dorsal roots [Nikundiwe et al., #82]. There is moderate substance P- and somatostatin-like immunoreactivity in the central field [Lorez and Kemali, '81].
VENTROLATERAL FIELD
In light of the earlier discussions of the lateral and dorsal fields, it is difficult to assign a laminar designation to this region. Ebbesson described the ventrolateral field as homologous to lamina VII
[Ebbesson, '76]. Unlike lamina VII in mammals, the ventrolateral field receives few primary afferents. The neurons here are small to medium in size and can be placed into three categories on the basis of their dendritic arborizations [Silver, 42 and Ebbesson, '76].
The smallest of the three classes is contained entirely within the ventrolateral field. The second class is medium sized, with dendrites that extend into the lateral or ventral funiculi. The third category sends one of its dendritic processes, through the ventral commissure, into the contralateral ventral funiculus.
A group of neurons, dorsal and lateral to the lateral motor column [LMC], is known to project, ipsilaterally, to the reticular formation. These neurons, called anterolateral marginal cells, persist 16
into adulthood [Forehand and Farel, '82 a&b]. Large cells, structurally similar to motoneurons but situated dorsal to the LMC, project to the cerebellum [Grover and
Grusser-Cornehls, '84; Gonzalez et al., '84]. An extremely large cell, the large circumferential cell
[LCC], exists in the ventrolateral field of the lumbar enlargement in larval Rana catesbeiana [Campbell et al.,
'82 and '87]. The nucleus of this neuron is laterally eccentric, perhaps owing to the bundle of neurofilaments that pass through the somata from the dorsal to the ventral process. This distinctive somatic arrangement is reminiscent of unipolar cells. The dorsal circumferential dendrite of this neuron reaches the medial aspect of the ipsilateral dorsal funiculus and travels rostrocaudally within it. The ventral dendrite travels rostrocaudally within the contralateral ventral funiculus. Small radial dendrites are also evident about the LCC somata and are observed to extend into the ventral part of the lateral funiculus. The LCC receives numerous synaptic contacts from axons traversing the lateral funiculus. There is no evidence that this cell persists in adult ranid [Campbell et al., '82 & '87].
Another group of small circumferentially placed cells is also evident in that region of the ventrolateral field 17 lying between the lateral and medial motor columns
[Campbell et al., '87].
The ventrolateral field is a region that receives terminal arborizations of reticulospinal fibers including a serotonergic projection from the raphe [Soller, '77].
VENTROMEDIAL FIELD
Ebbesson ['76] compares this region with laminae
VIII in mammals. The neurons here are intrinsic to the field except for dendritic excursions into the contralateral ventromedial field. This area is a receiving area for tectospinal, vestibulospinal and some reticulospinal axonal endings [ibidem].
MEDIAL MOTOR COLUMN
The medial motor column [MMC] consists of large motoneurons that innervate axial musculature [Ariens
Kappers, '36; Silver, '42; Cruce, '74 and Ebbesson, /76].
The cells have extensive dendritic arborizations that enter all gray fields (except the central field) and the lateral, ventral and contralateral ventral funiculi. The number of medial motor neurons decreases during development from the onset of foot paddle stages (i.e. stage 54 in Xenopus laevis), to metamorphic climax [van
Meir, 785]. Medial motor neurons are lost at lumbar segments during larval stages. The reasons for this loss are not proved but the loss of tail musculature and the 18
turn-over of axial musculature at metamorphic climax
[Alley, in press] likely participate in this selective
cell death process. In the adult, the MMC begins caudal to the obex, near the ventromedial gray matter, where the hypoglossal nucleus ends. The large medial neurons continue to the seventh segment, re-emerge at the caudal part of the tenth segment and persist into the rostral filum terminale.
LATERAL MOTOR COLUMN
The lateral motor column [LMC] appears in the cervical and lumbar cord and innervates the musculature of the upper and lower limbs [Silver, '42? Cruce, '74 and
Ebbesson, '76] [Fig. 31 & 35]. The LMC develops concomitantly with the limbs during anuran larval development [Beaudoin, '55; Forehand and Farel '82a;
Liuzzi et al., '85], It may be subdivided into dorsal and ventral portions according to several criteria. In
Rana, the dorsal part of the LMC contains multipolar, more angular motoneurons, whereas the ventral part of the
LMC holds fusiform shaped cells [Cruce, '74 and Liuzzi et al., '84]. The dorsal LMC receives primary afferents, while the ventral LMC does not [Szdkely, '76a; Nikundiwe et al., '82; Liuzzi et al., '84? see also Blight and
Precht, '81], Primary afferent terminations in the 19 brachial dorsal LMC are not apparent [Jhaveri and Frank,
783]. The muscle groups innervated by the two neuron pools is also distinct. The ventral group, in lumbar cord, innervates hip flexors and knee extensors. The dorsal group, in the lumbar enlargement, innervates hip, ankle and toe extensors and knee, ankle and toe flexors
[Silver, '42 and Cruce, '74]. Finally, studies of the development of the two subdivisions revealed distinctions in the origins, migratory pathway and migratory interactions of the dorsal and ventral part of the LMC
[Forehand and Farel, '82a; Liuzzi et al., '83 & '85].
Cruce subdivides the dorsal part of the LMC into two longitudinal subdivisions in Rana catesbeiana: 1) the rostro-dorsal subdivision runs through segments 8 and 9.
The rostralmost part of this subdivision innervates knee flexors and caudally it innervates the hip and ankle extensors; 2 ) the caudo-dorsal group is largely in segment 10. Rostrally, it innervates ankle and toe flexors. Caudally, it innervates ankle and toe extensors.
LMC motoneurons have extensive dendritic arborizations. This quality was thought by Cajal ['09] to be typical of the primitive CNS and that it was related to the absence of elaborate axon collaterals. It may be that the motoneuron itself integrates the full 20
range of segmental and supraspinal input with little interneuronal intervention [Liuzzi et al., '84]. LMC motoneurons have between 5 and 6 primary dendrites that develop particular orientations during larval stages
[Liuzzi et al., '85] which are maintained into adulthood
[Bregman and Cruce, '80]. According to a Golgi study of
Rana pipiens lumbar LMC by Bregman and Cruce ['80], dendrites D-^ and D2 emerge from the dorsal aspect of the
LMC soma and extend into the lateral field, although branches do reach the lateral funiculus. As might be expected, input to dendrites D-^ and D2 is mainly segmental from dorsal root afferents [see section 1.3] and to a lesser extent, supraspinal via the rubrospinal and reticulospinal tracts [see section 1.6]. D3, D4 and
D5 extend into the lateral funiculus. These dendrites are not always present. Two out of three motoneurons possess a D4 dendrite, while about 6% of the motoneurons surveyed had D3 or D5 primary dendrites. Like D-^ and D2 ,
D6 is almost always present. It emerges from a ventrolateral position on the soma and extends into the ventral funiculus with little branching. The D7 dendrite is also present in nearly every case and it projects into the ventral funiculus. D7 also extends rostrocaudally over a considerable distance (i.e. over 500 um) and 21
branches into the gray matter. It is the likely
recipient of vestibulospinal and reticulospinal input as
well as primary afferent input in the lateral field
[Liuzzi et al., '84]. DQ present in about one out of
five cases, emerges from the dorsomedial region of the motoneuron and extends toward the central canal [Bregman
and Cruce, '80].
Motoneuron cell bodies and dendritic processes were
predicted to be in close apposition with one another and
to possess gap junctions based on the
electrophysiological studies of Grinnell ['66 & '70], who
recorded brief depolarizations from unstimulated motoneurons which were adjacent to antidromically activated motoneurons. These adjacent depolarizations were shown to be electrically mediated. One possible anatomic substrate for the electrotonic coupling is the gap junction. Gap junctions have been observed between motoneurons that are segmental and ipsilateral to each other [Sotelo and Taxi, '70; Sotelo and Grofova, '76?
Sonnhof et al., '77; Motorina, '78; Westerfield and
Frank, '82 and Collins, '83] intersegmental and ipsilateral to each other and segmental but contralateral to each other [Erulkar and Soller, '80]. However, these same studies show that gap junctions are relatively rare.
Early reports of their scarcity fueled speculation that 22
electrically mediated depolarizations are due to ephaptic effects mediated by the close apposition of large regions of motoneuron membrane. One ultrastructural analysis revealed that only a small percentage of the motoneuron membrane was in apposition with other motoneuron membrane in Bufo arenarum and Rana catesbeiana and that no gap junctions were apparent [Stensaas and Stensaas, '71]. An ultrastructural study by Szekely and Kosaras ['76b] examined cobalt labeled lumbar motoneurons of Rana esculenta and questioned the finality of the Stensaas and
Stensaas conclusions. To date, no clear conclusions have been reached, though some believe rare gap junctions to be a sufficient condition for physiological observations of electronic coupling [ibidem; see also Hatton et al,
787 and Cobbet et al, '87].
Considerable attention has been directed toward finding an anatomical substrate for the short latency monosynaptic reflex demonstrated physiologically
[Brookhart and Fadiga, '60? Cruce, '74; Simpson, '76;
Frank and Westerfield, '82]. These observations suggested the presence of axodendritic electrotonic contacts, axosomatic chemical synapses, or both. Some degeneration studies of primary afferents did not reveal axosomatic contacts upon LMC motoneurons [Chambers et 23 al., '60; Joseph and Whitlock, '6 8 ; Nikundiwe et al., '82
(note: this is an HRP study but probably showed labeled degenerating primary afferent elements owing to the survival times)]. However, axosomatic contacts between primary afferents and LMC motoneurons have been demonstrated in the dorsal aspect of the lumbar LMC using a variety of other techniques [Corvajci and Pellegrini,
'75; Szekely, '76a; Szekely and Czeh, '76c; Antal, '80;
Liuzzi et al., '84; Rosenthal and Cruce '84 & '85;
Shiriaev and Shupliakov, '8 6 ]. Physiological evidence suggests that contacts from primary afferents upon LMC motoneurons can be monosynaptic chemical synapses, electrotonic contacts or both [Sotelo and Grofova, '76;
Motorina, '78; Shapovalov, '80; Shapovalov and Shiriaev,
'80; Lichtman and Prank, '81; Blight and Precht, '81;
Frank and Westerfield, '82a & b; Adanina and Shapovalov,
'83; Shiriaev and Shupliakov, '8 6 ]. Anatomic studies lend support to the physiological observations of electrical contacts upon cell bodies and proximal dendrites [Sotelo and Taxi, '70; Sotelo and Grofova, '76;
Motorina, '78 and Adanina and Shapovalov, '83].
In the brachial cord few dorsal root axons project far enough to reach the LMC motoneurons that innervate the triceps brachii muscle [Jhaveri and Frank, '83], consequently most of these contacts with motoneurons are 24
axodendritic. This observation does not rule out a pattern similar to that seen in the lumbar LMC. Triceps motoneurons are in the ventral part of the brachial LMC, the comparable region in the lumbar LMC does not receive direct primary afferent contacts either [see Liuzzi et al., '84 for a discussion of this issue]. Each triceps motoneuron receives many Ca++-dependent monosynaptic contacts [Lichtman et al., '82; Frank and Westerfield,
'82a].
1.3 DISTRIBUTION OF PRIMARY AFFERENT AXONS
Dorsal root axons, upon entering the spinal cord, segregate into medial and lateral divisions [Cajal, '09] in a manner akin to that seen in mammals [e.g. Light and
Perl, '79]. The medial division, containing large diameter, myelinated axons, passes medially to fill the dorsal funiculus. The lateral division, consisting of thin, lightly myelinated or unmyelinated axons interspersed with some large diameter fibers, forms the dorsolateral fasciculus or Lissauer's tract [LT]. Both divisions contribute collateral axonal terminations to their segment of entry, segments rostral and caudal to their entry zone and to supraspinal regions [Ebbesson,
'76; Antal et al., '80? Szekely et al., '82; Nikundiwe et al., '82; Liuzzi et al., '84; Rosenthal and Cruce, '85]. 25
At least some of the elements in the lateral division are small diameter C fibers, excited by noxious stimuli
[toads, Maruhashi et al., '52].
Two fields receive primary afferent distribution, one in the dorsal field (i.e. the dorsal terminal field,
DTF) and the other in the lateral field (i.e. the lateral or ventral terminal field, VTF). The VTF develops earlier than the DTF in Rana catesbeiana [Liuzzi et al.,
'85] and in Xenopus laevis [van Mier, '86 ]. Axons from thoracic dorsal roots do not contribute to a VTF in the thoracic segments [Antal et al., '80; Nikundiwe et al.,
'82; Rosenthal and Cruce, *85]. A VTF is present at thoracic levels, but it is composed of primary afferents from brachial and lumbar DRG's [Antal et al., '80]. The
DTF and VTF have been discussed above [section 1.2).
Additionally, some small diameter axons enter a nucleus that is ventral to the tract of Lissauer [LT] at cervical and lumbar segments. Forehand and Farel ['82b] observed cells in this region, retrogradely labeled after HRP application to the lateral reticular formation in larval
Rana catesbeiana, that they termed an anuran lateral cervical nucleus. The same population of cells was not labeled after HRP application to the diencephalon.
Rosenthal and Cruce ['85] noted that since the rat lateral cervical nucleus projects to the thalamus in 26
addition to the reticular formation and that it is limited to the second cervical segment, the lateral spinal nucleus might be a better homolog for the nucleus in question. The lateral spinal nucleus extends along the entire rat cord and projects to the reticular formation not the thalamus. However, neither the lateral cervical nucleus nor the lateral spinal nucleus receive primary afferents in the rat [Bresnahan et al., '84].
For these reasons Rosenthal and Cruce were reluctant to name this nuclear arrangement. It is possible that these cells are displaced from laminae I and II, as is the case with some reptiles and birds [Cruce, '79 and Martin and
Brinkman, '70]. Substance P- and enkephalin-like immunoreactivity is high in this region [Lorez and
Kemali, '81; Adli et al., '8 8 ] as is somatostatin [Adli et a l ., '88]
Rosenthal and Cruce ['83] noted extensions of primary afferent axons into the contralateral dorsal gray at the tenth segment in Rana pipiens but not at other levels. However, in our experience crossing fibers of
9th root origin are present at their segment of entry
[Fig. 2] in Rana catesbeiana and 10***1 segment primary afferents were observed to cross at 8^ and 9th segments in Xenopus laevis. 27
Segmental arrangements of ascending and descending
DRG collaterals are evident in the dorsal funiculus
[Antal et al., '80; Nikundiwe et al., '82; Rosenthal and
Cruce, '85] but not in the tract of Lissauer [Rosenthal and Cruce, '85]. The rostrally ascending collaterals
from lumbar segments are located in the dorsomedial part of the dorsal funiculus. Fibers from the tenth segment are more medial than those from nine and eight [Rosenthal and Cruce, '85], Thoracic contributions to the dorsal funiculus are ventral to those from the lumbar region and brachial afferents are most lateral [Rosenthal and Cruce,
'85]. Small diameter ascending collaterals from lumbar levels reach the brachial spinal cord and those from brachial segments can reach as far as the glossopharyngeal nucleus. Large diameter ascending collaterals from the lumbar segments, extend terminal arborizations into the VTF's of each segment only as far rostral as the thoracic cord, however, a few thin fibers to the DTF are given off bilaterally throughout the course of the 'gracile fasciculus' [Antal et al., '80].
Large diameter ascending collaterals can extend 5-6 segments below their entry level. Muscle afferents entering the spinal cord as far rostral as segment 4 can reach the tenth segment.
The longest fibers in the tract of Lissauer [LT] 28
extend 2-3 segments below their entry segment and 5-6 segments above it [Antal et al., '80; Rosenthal and
Cruce, '85]. The termination of these elements in the cord occurs in a narrow semicircular zone around the tract of lissauer [Figs. 2, 31 & 35], This lateral terminal field [LTF] is apparently undescribed in other studies.
1.4 SEGMENTAL INTERCONNECTIONS
Quadrupedal locomotion [Grillner, '75; Szekely and
Czeh, '76c; Kusuma and ten Donkelaar, '80] and intersegmental coordination of sensorimotor activity
[Sperry, '51; Mendell and Hollyday, '76] require the presence of long and short propriospinal interconnections. Some propriospinal interconnections arise early in development [Roberts and Clarke, '82] and participate in the coordination of swimming movements of the axial musculature [Roberts and Kahn, '82; Kahn and
Roberts, '82]. Since quadrupedal locomotion requires coordination of limb and with axial movements [see
Szekely and Czdh, '76c for a review], these early propriospinal pathways may be retained or expanded upon during larval limb development. A thorough analysis of the propriospinal connections in anurans is not available. However, reptilian studies [Kusuma and ten Donkelaar, '80] indicate most propriospinal connections
are short [i.e. a few segments], ipsilateral projections
from neurons in the lateral parts of laminae VII and VIII
[i.e. comparable with the ventrolateral and ventromedial
fields]. They course rostrally and caudally along the
gray/white interface and terminate in the ventral gray.
In quadrupedal reptiles, long ascending and descending
axons, connecting the cervical and lumbar enlargements,
arise from neurons in lamina VIII [the ventromedial
field] and medial parts of VII. These axons travel in
the ventral and lateral funiculi and terminate
bilaterally [mainly contralateral], in lamina VIII and
the medial parts of VII. Propriospinal axons are thought
to make S and F type synaptic contacts on the somatae and
proximal dendrites of neurons in the medial ventral horn
in Bufo bufo [Corvaja and Grofova, '77]. In embryonic
Xenopus laevis, at least some propriospinal connections
are well established prior to the advent of descending
supraspinal projections [Nordlander et al., '85].
1.5 ASCENDING SPINAL SYSTEMS
Dorsal root axons or collaterals thereof, terminate
in the following supraspinal regions: 1 ) the dorsal
column nuclei? 2) the trigeminal sensory nuclei; 3) the vestibular nuclei? 4) the reticular formation; 5) the 30
cerebellum and 6 ) the midline commissural nucleus of
Cajal [Nikundiwe et al., '82] [see the DCN discussion
below].
DORSAL COLUMN SYSTEM
Whether anura possess dorsal column nuclei [DCN] was
an open question for some time [Ariens Kappers, '60;
Joseph and Whitlock, '68]. Investigators, using various
techniques sought to demonstrate the DCN [see Nieuwenhuys
and Opham, '76]. From Nissl stained material, Silvey et
al.['74] described a loose nuclear arrangement of neurons
in Rana pipiens that extended into the second spinal
segment. Nieuwenhuys and Opham ['76] were not able to
distinguish a homologous structure in Rana catesbeiana or
Rana esculenta from Nissl stained material. More
recently cobalt [Antal et al., '80] and HRP [Nikundiwe et
al., '82 ] techniques have been applied to the question.
Based on the patterns of primary afferent termination
contemporary investigators have concluded that anura do have gracile and cuneate nuclei.
The anuran DCN consists of small to medium sized cells? at the obex it is bounded by the medial vestibular
nucleus and the spinal tract of V. Segregation of the
loosely arranged neurons into n. gracilis and n. cuneatus
is clearer toward the upper obex but it is never as distinct as in mammals [Herrick, '30 from Ebbesson, '76]. 31
Antal describes the unusual organization of the axonal elements in the dorsal columns:
"At the level of the glossopharyngeal nucleus the brachial DR (dorsal root) fibres fan out and form a trough which, along a laterodorsal spiral course, close into a tube around the thoracic and lumbar fibres. The thoracic and lumbar fibres follow a similar spiral course, and concentric rings are formed with the most caudal fibres constituting the core. This organization becomes deranged at the rostral level of the facial motor nucleus, where the thoracic fibres come to an end and the brachial and lumbar fibres continue their course to the cerebellum." [Antal, et al. '80; p. 1313]
Many dorsal column fibers end within the DCN [Joseph and Whitlock, '68] but some have been observed to give off collaterals which continue to the cerebellum. In addition, some axonal elements from the gracile fasciculus decussate at the obex and terminate in the contralateral reticular formation [Antal et al., '80].
The caudal extent of the DCN has been disputed by some investigators. Nikundiwe et al. ['82] hold that the
DCN extend to the level described by Silvey et al. ['74]
[i.e. to the level of the first spinal roots, C-2].
Antal et al. ['80] assert that the caudal limit of the
DCN is the obex. Antal et al. described a nuclear arrangement, contiguous with the DCN and extending from the DCN to segment 2, that they suggest is homologous with the n. Bischoff found in tailed vertebrates. This cell column receives dorsal root terminations. Since n. 32
Bischoff receives proprioceptive primary afferent collaterals from the tail, it is unclear how this region is homologous to a nuclear group in adult tailess amphibians. The caudal nuclear column is indistinguishable in location and boundaries from the rostral DCN. Like the rostral DCN, it receives primary afferent terminations and it apparently sends an axonal projection to the thalamus [Ebbesson, '76], In the absence of better evidence to the contrary, it seems to be a caudal extension of the DCN.
Nikundiwe et al. ['82] describe primary afferent distribution to a nucleus they term the midline commissural nucleus of Cajal. It is conceivable that this nuclear arrangement is a ventral extension of the gracile nucleus or the commissural portion of the solitary nucleus.
DRG-TRIGEMINAL TRACT
In Rana esculenta, thin fibers from the lateral division of the dorsal roots course along the ventral aspect of the caudal spinal nucleus of V where they terminate. These elements are actually interposed between the lateral and medial division of the spinal tract of V. Around the glossopharyngeal nucleus the lateral division elements dwindle and axonal processes 33 from the medial division begin to enter the spinal n. of
V? this continues up to and into the principal nucleus of
V [Antal et al., '80],
DRG-VESTIBULAR TRACT
As suspected from the physiological studies of
Precht et al. ['74] who demonstrated evoked EPSP's, action potentials and dendritic spikes in vestibular neurons subsequent to stimulation of the frog spinal cord, a fascicle of dorsal root afferents from the cervical and lumbar enlargements continues to the vestibular nuclei [Ebbesson, '76]. These afferents terminate in three vestibular nuclei as defined by cytoarchitectural analysis [Matesz, '79] and cobalt labeling of DRG axons [Antal et al., '80]: 1) the medial vestibular nucleus; 2 ) the lateral vestibular nucleus and
3) the superior vestibular nucleus, the rostral pole of which is termed the n. cerebelli.
DRG-RETICULAR PATHWAYS
Most ascending spinal pathways, including those of dorsal root afferent origin, terminate in the rhombencephalic reticular formation. The medial part of this fasciculus, composed of primary afferents from brachial and lumbar segments, projects to the dorsal gray matter [Ebbesson, '76; Antal et al., '80]. These reticular regions, ventral and lateral to the solitary 34
tract, also receive input from the mesencephalic tract of
V [Matesz, '78 from Antal et al., '80]. The lateral part of the fasciculus, which conveys DRG axons from the thoracic cord, terminates in the lateral reticular field
[Ebbesson, '76],
DORSAL SPINOCEREBELLAR TRACT
Anurans have no demonstrable homologue for Clarke's column or the lateral cervical nucleus [Joseph and
Whitlock, '6 8 ; Ebbesson, '76; Nikundiwe et al., '82;
Antal et al., '80; Grover and Griisser-Cornehls, '84].
Primary afferent axons from lumbar and thoracic segments ascend ipsilaterally to the cerebellar granular layer
[Antal et al., '80; Grover and Griisser-Cornehls, '84;
Gonzalez et al., '84].
VENTRAL SPINOCEREBELLAR TRACT
This bilateral projection arises from neurons in the ventromedial and ventrolateral fields of the lumbar and brachial enlargements [Grover and Griisser-Cornehls, '84;
Gonzalez et al., '84]. One cluster of these cells is situated just dorsal to the LMC, with somatic sizes equal to those of the adjacent motoneurons. The ventral spinocerebellar cells send their axonal process through the contralateral ventral horn. The axons fasciculate laterally in the ventral funiculus and ascend to the 35 rhombencephalon where some of them re-decussate, at the cerebellar commissure. They terminate ipsilaterally as mossy fibers in the granular layer [Grover and Griisser-
Cornehls, '84].
SPINORETICULAR PATHWAYS
A terminal region in the cervical and lumbar enlargements, described as the nucleus of the dorsolateral funiculus [see Rosenthal and Cruce, '85 for discussion], projects bilaterally through the lateral funiculus to the superior reticular field [Ebbesson,
'76]. Marginal cells, called by various names [see section 1 .2 ], project bilaterally through the lateral funiculus to the inferior and middle reticular fields
[Ebbesson, '76; Forehand and Farel, '82b].
SPINOMESENCEPHALIC PATHWAYS
Two regions within the mesencephalon receive a minor contribution from intrinsic spinal neurons: 1 ) the magnocellular nucleus of the torus semicircularis, which is located in the path of what remains of ascending spinal afferents [Ebbesson, '76] and 2) the laminar nucleus of the torus, in the periaqueductal gray [Potter,
'65 from Ebbesson, '76]. These regions are regarded as homologous with the inferior colliculus of mammals
[Ebbesson, '76]. 36
1.6 DESCENDING SUPRASPINAL SYSTEMS
Until the advent of various neuroanatomical tracing techniques, little was known about the source of descending supraspinal systems. Ebbesson ['76], in his review, noted the presence of tectospinal, vestibulospinal, reticulospinal and trigeminospinal projections and suggested that anurans possessed a rubrospinal projection as well. A few recent efforts have contributed significantly toward an understanding of descending systems in adult anurans [Grover and Grusser-
Cornehls, '80; D'Ascanio and Corvaja, '81? ten Donkelaar et al., '81 & '82 and Toth et al., '85]. Telencephalic, diencephalic, mesencephalic and rhombencephalic regions have neuronal populations that project to various levels of the spinal cord.
TELENCEPHALIC PROJECTIONS
Ebbesson ['76] was unable to discern a telencephalic projection to the spinal cord in anurans, although he was able to trace some fibers into the dorsal column nuclei.
Degeneration studies in Bufo marinus and Ambystoma tigrinum purported to show an ipsilateral projection into the cervical spinal cord from the telencephalon [Kokoros from ten Donkelaar, '81? Kokoros and Northcutt, '77].
HRP injections into the rostral cervical spinal cord 37
labeled cells ipsilaterally in the ventral striatum of
Xenopus laevis [ten Donkelaar et al., '81], Cobaltic-
lysine injections into the medullary ventral funiculus of
Rana esculenta labeled neurons in the middle third of the ventral striatum, whose axons projected through the
lateral forebrain bundle into the ventral funiculus [Toth
et al., '85]. The cytoarchitecture of these neurons is
like that seen in Golgi preparations from the same area
[Kicliter and Ebbesson, '76],
DIENCEPHALIC PROJECTIONS
Descending supraspinal axons arise from four
identified nuclear groups in the Xenopus laevis diencephalon, two from the preoptic area, one hypothalamic and one thalamic [ten Donkelaar et al.,
'81], These neurons are predominantly ipsilateral, with their axonal projections passing through the ventral and
lateral funiculi and their cell bodies located within:
1) the periventricular preoptic nucleus which projects as
far as the thoracic cord; 2 ) the magnocellular preoptic nucleus, which also projects to the thoracic cord; 3) the serotonergic periventricular hypothalamic nucleus, the
axons of which reach lumbar levels and 4) the ventrolateral thalamic nucleus, whose axonal processes also attain lumbar levels. Tdth et al., ['85], were not able to label the same cells with cobaltic-lysine 38
injections into the cervical cord of Rana esculenta but
confirmed ten Donkelaar's observations with cobalt
injections into the medulla. Interestingly, Toth et al.,
did label three groups of neurons in the diencephalon
after cervical cord injections of cobalt which were
largely ipsilateral, with their axons in the ventral
funiculus and somatae located within: 1 ) the posterior
thalamic nucleus; 2 ) the lateral thalamic nucleus, pars
posteroventralis and 3) the tuberculum posterius.
MESENCEPHALIC PROJECTIONS
Supraspinal projections to the cord arise from
mesencephalic neurons within: 1 ) the periventricular
gray, a possible homologue of the Edinger-Westphal
nucleus, which sends axonal processes bilaterally
[Ebbesson, '76] (ipsilaterally according to Forehand and
Farel ['85b]) through the ventral and lateral funiculi to
the level of the thoracic cord; 2 ) the nucleus of the medial longitudinal fasciculus, that project bilaterally,
with an ipsilateral predominance, through the medial
longitudinal fasciculus [MLF] as far as the thoracic cord
[D'Ascanio et al., #79; ten Donkelaar et al., '81; Toth
et al., '85]; 3) the presumptive red nucleus, the axons
of which decussate at the level of their emergence and
travel down the dorsal aspect of the lateral funiculus as 39
far as the lumbar enlargement [Ebbesson, '76; ten
Donkelaar et al., '81; T6th, '85]; 4) the anterodorsal, anteroventral and posteroventral tegmental nuclei, (which
likely represent the mesencephalic reticular formation in anurans [T6th et al., '85]), the axons of which project bilaterally through the ventral funiculus and, in the case of the anterodorsal and posteroventral nuclei, reach lumbar levels [Potter, '65? ten Donkelaar et al., '81;
Toth et al., '85]; 5) a presumptive anuran homologue of the inferior colliculus [Ariens Kappers, '60? Larsell,
'67; Potter, '65; Ebbesson, '76; Rubinson and Skiles,
'73], the caudal ventral part of torus semicircularis,
[Toth et al., '85] which projects ipsilaterally as far as the thoracic cord [ten Donkelaar et al., '81], and neurons in the laminar nucleus of the torus semicircularis, which project ipsilaterally to the cervical cord [ten Donkelaar et al., '81]? 6) the mesencephalic nucleus of V, which were is labeled in xenopus [ten Donkelaar et al., '81] but is labeled in
Rana esculenta [T6th, et al., '85? see Ebbesson, '76] and in toad [D'Ascanio et al., '79] after injections of tracers into the cervical cord; 7) the optic tectum, contralaterally projecting to the cervical cord from layer 6 and ipsilaterally projecting to the cervical cord from layer 7. This projection is slight in Xenopus 40
laevis [ten Donkelaar et al., '81], but larger in ranids
[Rubinson, '6 8 ? Grover and Griisser-Cornehls, '80;
Forehand and Farel, '85b? L&zar et al., '83] and in toads
[D'Ascanio and Corvaja, '79]. It is interesting that anurans which are obligate lurker hunters have a larger tectospinal projection than Xenopus laevis which can dine
adequately upon immobile detritus.
It is curious, also, that torus semicircularis, the presumptive anuran homologue of the inferior colliculus, has a descending spinal projection. Frogs and toads are very specialized auditory animals but unlike most vertebrates, they must turn their bodies with their limbs
in order to orient their head toward a novel auditory stimulus. Body turning rather than neck turning is necessary because anurans have one of the shortest post- otic head sequences amongst the vertebrates [Deucher,
'75]. This need for rapid orientation toward auditory stimuli and for limb movements to accomplish it, may explain the survival value of torospinal projections.
RHOMBENCEPHALIC PROJECTIONS
The largest volume of descending projections to the anuran spinal cord arise from distinct groups of neurons in the rhombencephalon. These include cells from: 1) the reticular formation, which project bilaterally 41 through the ventral and lateral funiculi at least to the lumbar enlargement and which arise, specifically, from the isthmal reticular nucleus, the superior reticular nucleus, the medial reticular nucleus, the inferior reticular nucleus and the raphe [Mensah, '74 from
Ebbesson, '76; ten Donkelaar et al., '81; ten Donkelaar and de Boer-van Huizen, '82; Toth et al., '85]; 2) an analog of the locus coeruleus, medial to nucleus isthmi
[see ten Donkelaar et al., '81 for a discussion of this assertion], which projects ipsilaterally as far as the lumbar enlargement; 3) the central gray, which contributes a small bilateral projection throughout the cord; 4) the cerebellar nucleus, which is thought to be homologous with the mammalian fastigial nucleus, and which projects contralaterally throughout the cord [ten
Donkelaar et al., '81]; 5) the rostral and caudal lateral line nuclei project as far as the thoracic segments. The rostral lateral line nucleus projects ipsilaterally and the caudal lateral line nucleus (a.k.a. n. caudalis of VIII [Nieuwenhuys, '76] or the medial vestibular n. [Matesz, '79]) projects contralaterally;
6 ) the descending nucleus of V, the axons of which pass through the ventral funiculus into the first few cervical segments; 7) the nucleus of the solitary tract, which projects bilaterally to cervical and contralaterally to 42
thoracic and lumbar segments? 8 ) the large cells of the ventral nucleus of VIII [Nikundiwe and Niewenhuys, '83], which have predominantly ipsilateral projections throughout the cord. A caudal, contralaterally projecting group of cells may be homologous to mammalian inferior and/or medial vestibular nuclei, while the rostral, large celled group may be homologous with the lateral vestibular nucleus (i.e. Dieter's) [van Mier,
'8 6 ] .
Reticulospinal projections from the medial, superior and isthmal reticular nuclei reach the cord via the ventral funiculus, whereas those from the inferior reticular nucleus travel along in the ventral part of the lateral funiculus. Raphespinal projections are diffusely interspersed throughout the ventral and lateral funiculi.
1.7 DEVELOPMENTAL CONSIDERATIONS
Amphibians move successively through an embryonic and a larval phase before entering adulthood. The development of the amphibian spinal cord thus entails the sequential establishment of neuronal circuitry necessary for the different behaviors of larval and adult life
[Kollros,'81; Fox, '84). For example, the embryonically born Rohon-Beard cell (Hughes, '57 and Lamborghini, '80), 43
Mauthner cell (Szepsenwol, from Ebbesson, '76; Willis,
'48), certain kinds of interneurons [Athias, 1897?
Forehand and Farel, '82a; Lamborghini, '80; Roberts and
Clarke, '82a; Campbell et al., '87], and primary motor
neurons are essential for the behavior of the swimming
embryo and larvae [Herrick, '15? Roberts and Kahn, '82b;
Kahn and Roberts, '82] but not for the adult tetrapod.
It is toward this developmental endpoint that, as the
larvae becomes adult in form, the Rohon-Beard system is
replaced with the dorsal root ganglion system, other
descending supraspinal innervation supersedes the
Mauthner cell system and a secondary motor neuron pool
arises to innervate the developing limbs [Forehand and
Farel, '82 a & b; Liuzzi et al., '85]. Reticulospinal
axons reach the spinal cord quite early in development
[ten Donkelaar and de Boer-van Huizen, '82]. Those from
the inferior and medial reticular nuclei reach the
rostral cord in Xenopus by stage 28 [van Mier, '8 6 ] and the caudal cord by stage 37 [Nordlander, '84 & '85].
Reticulospinal projections from the interstitial nucleus
of the MLF and the Mauthner cell reach the rostral cord by stage 30 [van Mier, '8 6]. Between stages 33-42 axons
from the superior reticular formation and from the ventral nucleus of VIII reach the rostral cord [van Mier,
'86] and the tail cord by stage 39 [Nordlander, '84 & 44
'85]. Serotonergic [Parent, '81; Ueda et al., '84]
raphespinal projections from the inferior raphe reach the
rostral cord by stage 32 and the caudal cord by stages
35/36 [van Mier, '86; Beattie et al., '87]. A
coeruleospinal projection arrives in the lumbar
enlargement around stage 55 and by stage 58, when the hindlimbs are used for locomotion, a rubrospinal projection exists [ten Donkelaar and de Boer-van Huizen,
'82]. Serotonergic axonal elements from the periventricular hypothalamic nucleus and a projection
from the ventrolateral thalamic nucleus arrive around stages 57-59 [van Mier, '86],
X.8 CONCLUSION
The advent of recent methodology has expanded what is known about normal amphibian spinal neuroanatomy. Since
Ebbesson's ['76] excellent review much has been learned about the development and organization of the LMC, the distribution of DRG primary afferents, the identity of many ascending and descending spinal systems and about the development of the spinal cord. A review of the literature seemed an effective way to frame the results of the experiments in chapters II and III. For, these experiments have to do with the disparate topics of plasticity and normal frog sacral neuroanatomy, unrelated to one another, but related to the broader concerns outlined here. CHAPTER II
2.1 INTRODUCTION
The studies contained in this chapter examine lesion induced changes in the normal distribution of
lumbar DRG axons after DC injury.
Injury to the nervous system can cause immediate motor, sensory or cognitive deficit. Following traumatic
injury many events occur, at the lesion site and at places removed from the lesion site. Those events occurring at the site of a lesion include, hemorrhage,
loss of vascular perfusion, loss of the blood brain barrier, edema, degeneration of processes and cellular elements, retraction of neuronal processes, apparent regenerative attempts by spared fibers, the hypertrophy of astroglia and proliferation of microglia [see e.g.
Bresnahan, '78; Becker and Povlishock, '85; Beattie et al., '88]. Observations that show evidence for plasticity in regions removed from the actual lesion site include, denervation supersensitivity, the use of alternative pathways, the occupation of vacated synaptic
46 47
sites through collateral sprouting of spared fibers, and pruning responses of spared collaterals [see e.g.
Goldberger, '85? Cotman, '85? Beattie and Bresnahan,
'82].
REGENERATION OF EXTRALOCAL AXONAL PROJECTIONS
CNS plasticity includes lesion induced activities that are apparently directed toward the circumvention of a disrupting lesion by the axons transected at the time of injury. The peripheral nervous system [PNS], in all species, has the capacity to regenerate through a lesion.
The extent of such regeneration varies according to the type of lesion [Brushart, '80; Wall et al., '83?
McQuarrie, '86], The specificity of re-established connections is also variable [Sperry, '63; Guth and
Bernstein, '61; Moyer et al., '53; Devor and Wall, '81;
Westerfield and Frank, '83? Farel, '86; Lee and Farel,
'88]. However, the ability of all species to achieve some measure of regenerative success in the PNS is unequivocal.
All species do not exhibit regeneration after CNS trauma. Species from most metazoan phyla can regenerate nervous elements. These include; plateyhelminthes; rhynchocoeles; annelids [Hickman, '70], arthropods
[Edwards and Palka, '76; Carbonetto and Muller, '77;
Nordlander and Singer, '82], and echinoderms [Hickman, 48
'70]. However, with the advent of the dorsal tubular arrangement of the metazoan CNS [i.e. phylum chordata], the capacity for CNS regeneration begins to decline.
Furthermore, this regenerative impotence increases with the complexity of classes within phylum chordata such that avians and mammals exhibit very little successful
CNS regeneration.
Cyclostomes [class: Agnatha], one of the classes of simple forms within phylum chordata, can regenerate their giant unmyelinated axons in a manner adequate to restore the rostral control of behavior [Hibbard, '63; Rovainen,
'74 & '76; Wood and Cohen, '81; Borgens et al., '80; Yin and Selzer, '84; Cohen et al., '88]. The regeneration is not complete, as only one out of three giant axons regenerate past the lesion site [Cohen and Hall, '86] and only 50% of all cut axons grow more than 2.5 mm past the injury site [Croop et al, '88]. Axons of Mauthner,
Muller anteriorly projecting dorsal cells, and giant interneuron axons are among those that regenerate
[ibidem; Yin and Selzer, '83 & #84], Regenerating axons extend through the lesion site in association with the subpial zone [Cohen, '88 and Croop et al., '88].
Chondrichthyes and osteichthyes are also capable of regeneration [Bernstein and Gelderd, '70; Bernstein and 49
Bernstein, '73; Coggeshall and Youngblood, '83; Levine,
'83; Berry, '82]. However, regeneration is selective for particular subsets of neurons such as: retinal ganglion cells [Sperry, '44]; reticulospinal neurons [Bunt and
Fill-Moebs, '84] and fibers from the thalamus after tectal ablation [Levine, '83]. Some fibers, in teleost, can successfully pass through a spinal cord lesion but fail to extend very far, or establish appropriate connections, like Rohon-Beard cell axons and the ascending limb of dorsal root afferents. Vestibulospinal and tectospinal fibers appear not to successfully regenerate through the same lesion site. Fibers that do regenerate through spinal cord lesions are often located in the dorsolateral funiculus and absent from the dorsal columns [Bernstein and Gelderd, '70; Bunt and Fill-Moebs,
'84] .
Urodeles have the capacity to regenerate axons through lesions in the spinal cord [Spallanzani, 1768;
Hooker, '15; Butler and Ward, '66; Egar and Singer, '72;
Stensaas, '83] in association with organized arrangements of ependymal cell processes [Egar et al., '70; Nordlander and Singer, '78; Michel and Reier, '79; Singer et al.,
'79; Simpson, '70 & '83]. The origin of the fibers which successfully pass through the lesion site and the completeness of the regeneration which does occur has not 50
been documented/ except for the Mauthner cell axon
[Holtzer, '52 and Simpson, '83]. Axons do not regenerate through the dorsal columns in urodeles [Stensaas, '83].
Like urodeles, reptiles can regenerate some of their descending tracts into reconstituted tail cord
[Simpson, '70; Egar et al., '72], but they apparently lack the capacity for restoration of fiber tracts when the lesion occurs at some midpoint of the cord [Egar et al., '72], A possible exception is the turtle [Satini and Browner, '88]. Most of the axons that regenerate into the lizard tail cord are from intrinsic spinal neurons, however, axons from reticular, raphe, and vestibular nuclei can also regenerate to some degree
[Duffy, et al, '88]. Axons from the red nucleus do not regenerate [ibidem].
Anurans can regenerate retinal ganglion cell axons
[Sperry, '44; Constantine-Paton and Law, '78; Gaze and
Jacobson, '63; Gaze and Keating, '70; Sullivan et al.,
'84], though the regenerated axons tend to be unmyelinated ones [Stelzner, '85]. Anurans can also regenerate cranial nerve VIII [Zakon and Capranica, '81],
Mauthner cell axons during larval stages [Lee, '82], ascending DRG axon collaterals prior to stage VIII
[Clarke et al., '86] and some descending axons prior to 51
and during metamorphic climax [Forehand and Farel, '82b;
Michel and Reier, '79; Lopate et al., '85]. It has been suggested that much of the larval regeneration consists
of new axon growth rather than true regeneration [Davis
and Farel, 785].
The speculation that many developmental examples of regeneration are due to extension of developing axon projections through a lesion site rather than regrowth of cut fibers, confounds efforts to interpret experimental observations. In anurans axons which pass through a spinal cord lesion site tend to appear in the dorsolateral funiculus and rarely occupy the dorsal columns [Lopate et al., '85; Forehand and Farel, 782c] unless the lesion occurs before stage VII [Clarke et al.,
786 and Holder, et al., 787]. Adult or non-metamorphic larval spinal cord does not regenerate [Piatt and Piatt,
758; Forehand and Farel, 782b]. However, some recent studies show limited regeneration of dorsal root ganglion cell axons into the spinal gray and Lissauer7s tract of adult anurans if the lesion is made outside the spinal cord [Katzenstein and Bohn, 784? Sah and Frank, 784;
Liuzzi, 786a and 786b; Frank and Sah, 786].
As mentioned in the preceeding synopsis, the tail and tail spinal cord of larval anurans and some species of adult reptile and urodele can regenerate following 52
amputation [Spallanzani, 1768; Hooker, '15; Holtzer, H.,
'51 & '56; Piatt, '55; Butler, E. G. and Ward, '66;
Simpson, '70; Egar and Singer, '72; Egar et al., '72;
Stensaas, '83]. Descending supraspinal tracts and intrinsic spinal neurons extend into the reconstituted tail cord [ibidem and Duffy et al, '88]. In these cases ependymal cells proliferate and then organize so as to leave contiguous channels, through which regenerating axons pass [Egar et al., '70; Nordlander and Singer, '78;
Michel and Reier, '79; Singer et al., '79; Simpson, '70 &
'83]. A similar process occurs after mid-thoracic lesions in urodeles [Butler and Ward, '65; Clemente, '64;
Michel and Reier, '79; Simpson, '83; Stensaas, '83].
Undifferentiated mesenchymal cells, blastema cells, have been hypothesized to be organizing centers for tail regeneration [Simpson, '83]. Fibroblastic infiltration rather than blastema formation occurs after mid-thoracic lesions in adult frogs and reptiles [ibidem].
Avians and mammals have some capacity for regeneration though it is often developmentally dependent and restricted in adults to very small subsets of neurons. The developmental dependence of some kinds of regeneration has led to speculation that the observed lesion circumvention may be due to the ingrowth of 53 undamaged, developing axons [Bregman and Goldberger, '82
& '83]. Some regeneration has been observed by: retinal axons [McConnell and Berry, '82 and Berry, '76]; neurohypophyseal fibers [Adams et al., '68; Beck et al.,
'69; Kiernan, '71; Raisman, '73 and Berry, '76]; olfactory axons [Barber, '81 & '82]; adrenergic axons
[Sievers and Klemm, '82; Bjorklund and Stenevi, '79;
Bjorklund, '82; Jonsson and Hallman, '82; Crutcher and
Collin, '82; see Crutcher, '87 for a review]; cholinergic fibers [Barnes and Worall, '68; Svengaard et al., '76;
Risling et al., '84; Bjorklund and Stenevi, '79]; a variety of different fibers in the adult rat brain
[Foerster, '82] and many fiber types in fetal and neonatal mammals [Kalil and Reh, '79; Bernstein and
Stelzner, '83; Reh and Kalil, '82; Schreyer and Jones,
'83; Bregman and Goldberger, '82; Martin and Xu, '88].
Some mammalian CNS neurons show the capacity for regeneration into PNS grafts [Kao, '74; Richardson et al., '83].
When cataloging the regenerative response of axonal projections within phylum chordata, what becomes evident is that the transition from species that are capable of regeneration to those that are largely unsuccessful is gradual. It is also evident that certain features are common to many types of regeneration in the CNS. For 54
instance, effective adult mammalian regenerates tend to be small diameter, unmyelinated or lightly myelinated, noradrenergic or cholinergic fibers, that grow for short distances. Regenerating, small and large diameter axons in nonmammalian species often occupy regions that are lightly myelinated and rarely occupy regions that are heavily myelinated. Particular developmental stages are conducive to regeneration of large fibers through or around a lesion site in anurans and in mammals, though whether this is new growth rather than the regeneration of transected fibers is not always clear. Also, in adult anurans, when the central processes of dorsal roots are crushed and allowed to regenerate into the spinal cord, the long projecting tracts in the dorsal funiculus are not re-established.
REORGANIZATION OF NEURONAL ELEMENTS
In addition to the response of damaged axons at the site of injury [i.e. regeneration], other localized plasticity may occur consequent to injury. These types of reorganization include collateral sprouting, pruning responses of spared collaterals and ingrowth of foreign axonal populations into denervated regions [Davis, '85].
When terminal fields are lost, they may be replaced by the growth of the collaterals from undamaged axons in 55 the affected region, [Liu and Chambers, '58; Tsukahara,
'81; Cotman & Nieto-Sampedro, '81] even in the adult CNS
[Murray and Goldberger, '74; Cotman, '76 and '79?
Steward, '82]. Whatever the etiology of this phenomenon, the result may produce limited behavioral recovery
[Goldberger and Murray, '80] or may contribute to behavioral deficit [McCouch, '58]. Sprouting can arise from local neuron groups [Tessler et al., '84] or from neurons removed from the site of denervation that normally project to [Liu and Chambers, '58] or near the denervated region [Goldberger and Murray, '80 & '82]. It has been suggested that elements similar to those lost through denervation have a competitive advantage over sprouts from other sources [Goldberger and Murray, '82;
Murray and Goldberger, '86].
A "pruning effect" is the growth response of the undamaged portion of an axonal arborization after partial axotomy. It is like regeneration in that a growth response occurs consequent to injury, but is unlike regeneration in that the elements which expand are not at the lesion site [Schneider and Jhaveri, '74]. This phenomenon has been extensively studied in the cerebellar cortex [Pickel et al., '73 & '74] and the hippocampus
[Gage et al., '83; Davis and Haring, '83; see also Cotman and Lynch, '76 and Cotman, '85 for a discussion of these 56
issues].
The current study attempts to contribute to an understanding of some CNS responses to trauma through examination of the reaction of DRG central processes to a lesion. The question of whether rostrally ascending DRG axons can pass through or around a thoracic DC lesion has been studied in Rana catesbeiana larvae and Xenopus laevis juveniles. In addition, the reactions of the collateral plexus of lumbar DRG cells caudal to DC lesions were examined. 57
2.2 PLASTICITY IN LARVAL AND ADULT RANIDS
The purpose of this experiment is to examine the response of DRG central processes to lesions of the thoracic DC in larval and juvenile frogs and to identify examples of plasticity for further study. The method used to demonstrate DRG axons in the present experiment was injury filling of spinal roots. Application of HRP to the central cut end of the spinal dorsal root provides a Golgi-like picture of the labeled elements and fills axons of all calibers, thereby permitting careful observation of those elements at both the light [Light and Perl, '77, '79; Proshansky and Egger, '77] and electron microscopic levels [Beattie et al., '78 & 79;
Mawe et al., '83]. Injury filling of afferents has been used to discern the normal pattern of primary afferent input in anurans [Fig. 2] [Nikundiwe et al., '82; Liuzzi et al., '84; Rosenthal and Cruce, '85], the initial growth of this afferent system into the spinal cord
[Forehand and Farel, '82b; Liuzzi et al., 85; van Meir,
'86] and regeneration of DRG axons into the cord
[Forehand and Farel, '82c; Liuzzi and Lasek, '84 & '86a].
The same methods have been employed in the current experiment to address not only regeneration of the central cut ends of the dorsal root axons but also their undamaged local collateral plexus. Several questions are 58
addressed by these studies: l) Do dorsal root afferent axons regenerate and traverse or bypass dorsal funicular lesions in larval or adult anura?; 2) Is the ability to do so, assuming it is present, dependent upon developmental stage? 3) Can cellular elements (e.g. blastema cells) grafted into the spinal cord, induce regeneration of ascending DRG collaterals? and 4) Do DRG axonal elements, caudal to the thoracic lesion and spared after the injury, reorganize in some discernible fashion?
Preliminary accounts of these experiments have been reported [Campbell et al., '84, '88a & '88b]. 59
2.2.1 METHODS: Rana catesbeiana
TISSUE PREPARATION
Twenty-three Rana catesbeiana tadpoles [stages I-XX of Taylor and Kollros, '46? Fig. 1] and four juveniles were used in this study. The normal HRP labeled anurans used in this study were cases available in the lab which were prepared by Liuzzi et al. ['83, '84 & '85].
Tadpoles and juveniles were anesthetized by immersion in an aqueous solution of 6% tricaine methanesulfonate (MS-
222) . Larger frogs were anesthetized by hypothermia followed with an injection of MS-222, at 0.1-0.4mg/g body weight, into the dorsal lymph sac. The thoracic spinal cord was exposed by the removal of the laminae of the fourth and fifth vertebrae in juvenile anurae and the sixth and seventh vertebrae in larval subjects. A Beaver microblade was used to make a unilateral, shallow cut through the dorsal funiculus. A fine glass needle was used to investigate the lesion and cut any spared regions. Muscle and skin were consecutively approximated with 7-0 silk.
The ranid preparations were allowed to survive the initial dorsal funiculus lesions for periods ranging from
48 hours to 14 weeks. After recovery, axons of the eighth or ninth dorsal roots, ipsilateral to the lesion, were cut and chips of dried, concentrated HRP (type IV, 60
Sigma Chemical Co.) were applied to the proximal cut ends
[Fig. l]. The time of HRP application varied from 20-40 minutes. At the end of this procedure excess HRP was flushed away and the wound edges were approximated with
7-0 silk suture. After 12-24 hours the animals were perfused intracardially with 0.9% saline solution followed by fixative containing 1% gluteraldehyde and 3% paraformaldehyde (pH 7.3). The spinal cord and brain were extirpated and immersed in 30% sucrose overnight.
Transverse sections through the sacral, lumbar, and thoracic cord were cut at 60um on a freezing microtome and mounted on glass slides. Tissue from animals intended for sequential light and electron microscopic analysis in studies subsequent to this one were sectioned frontally on a vibratome. Sections were placed in a solution of 5% cobalt chloride for intensification of the reaction product [Adams, '77]. After rinsing the sections in 0.1 M phosphate buffer, they were incubated at 37°C in 0.05% diaminobenzidine (DAB) for 20 minutes.
4ml of 0.06% H2O2/100 ml of DAB solution were then added for 5-10 minutes while the sections were gently agitated on a rotator. The slides were then rinsed, stained for
Nissl substance with cresylecht violet, dehydrated in a series of alcohols to xylene and coverslipped. After HRP 61 processing, sections taken on the vibratome were rinsed in sodium acetate, placed in 1% osmium tetroxide for 1 hour, stained en block with uranyl acetate, dehydrated in graded alcohols to propylene, infiltrated with Maraglas and flat embedded between two sheets of Aclar [Allied
Chemical Co.],
Blastema cells used in spinal cord implantation were generated within tail amputation stumps, ten days after amputation of the distal third of the tadpole tail. The tip of the stump was excised, denuded of apparent epidermis and implanted within a dorsal column lesion rostral to the lumbar enlargement in three animals. The implant was secured with a closely applied bone flap and the wound edges approximated with 7-0 silk suture. The implants were autologous in order to minimize host rejection of the implant and facilitate survival of the host. After varying periods of survival, the ipsilateral eighth dorsal root was exposed, transected and chips of
HRP applied in the manner previously described. In five additional cases the nuclei of the rapidly dividing blastema cells were labeled by two I.P. injections
[lOuCi] of tritiated thymidine given at a twelve hour interval. The donor blastema was harvested 3 hours after the second injection and implanted into the dorsal columns of unlabeled hosts (n=3). The other two animals 62
were sacrificed after 24 hours and used as controls. The implantation subjects were perfused at intervals ranging from 48 hours to 14 days. 60um sections of the region of implantation were cut on a freezing microtome and mounted on clean glass slides. Incorporated thymidine was demonstrated using standard autoradiographic procedures.
The tissue from one of the three long term blastema implants was cut on a vibratome and every third section was prepared for electron microscopic analysis in the manner previously described.
DATA ANALYSIS
Light microscopic analysis was performed, using a
Leitz Orthoplan, microscope with bright field, dark field and differential interference contrast optics. Areas of interest were photographed and drawn at several magnifications. The analysis particularly focused on the root entry zones of the labeled roots and the lesion sites in order to determine if any localized or long trajectory types of reorganization or regeneration occurred. 63
2.2.2 RESULTS: Rana catesbeiana
REGENERATION OF DRG AXONS THROUGH OR AROUND A DC LESION
Useable results were obtained from 11 DC lesioned tadpoles [Fig. 1], The interoperative time periods for those cases with DC lesions are listed followed by the stage at lesion and then the stage at assay: 1 week-1 case (XIX-XIX); 2 weeks-2 cases (XX-XX; XIV-XIV); 3 weeks-2 cases (XX-XX; XIX-XIX); 4 weeks-2 cases (XII-XII;
XIX-juvenile); 8 weeks-1 case (XI-XII); 10 weeks-3 cases
(VII-XV; XII-XV; XII-XV); 13 weeks-1 case (XV-juvenile).
In the normal cases, the distribution of DR afferents entering their respective segments was as
Liuzzi et al. ['84 & '85] described [Fig.2]. Dorsal root axons enter the spinal cord and segregate into a medial and lateral division. The lateral division axons are smaller in diameter and distribute through the LT.
Medial division axons are larger in diameter and distribute through the DC. Three terminal arborizations are seen throughout the course of their distribution: 1) a terminal field around the tract of Lissauer; 2) a DTF and 3) a VTF. Occasional axons project contralaterally and these projections increase in the caudal spinal cord.
A small fascicle of axons passes into the dorsal aspect of the LMC from the VTF. During stages VI-X DR axons have scarcely entered the dorsal field. Numerous 64
examples of growth-cone like structures, 3-10um in diameter, can be seen. By stages XIII-XIV the medial and lateral subdivisions, the DTF and VTF and the terminal field around LT are seen. By stage XVII the DR distribution is adult in appearence.
The cases receiving blastema implants fell into one of two categories. Either the subjects received heterologous implants labeled with tritiated thymidine
(n=3) or homologous implants and were used for HRP studies (n=3). In addition, amputated tadpoles (n=2) were labeled with tritiated thymidine and used as controls. Useable heterologous implants were: 2 days-1 case (XX); 4 days-1 case (XIV); 14 days-1 case (XVIII).
There was one useable homologous implant that survived 12 weeks (i.e. a stage XII tadpole that was assayed at stage
XVII). Mesenchymal cells, presumed to be blastema cells were labeled with tritiated thymidine in the regenerating amputation stump 24 hours after injection [Fig. 3].
H-thymidine labeled transplants, placed into thoracic lesions remained intact for the duration of the control study [Fig. 4). The homologous implant appeared, grossly and microscopically [Fig. 5, outline], to be where it was placed at the time of implantation. A portion of the eighth dorsal root entry zone entered the caudal part of 65 the implant [Fig. 5]. Axons were seen in the implant
[Fig. 6] through several sections. Interestingly, relatively few varicosities could be observed upon axons in the transplanted tissue.
Unilateral lesions of the thoracic dorsal columns
[DC] were effective in cutting rostrally ascending DRG axon collaterals in at least 6 out of 8 cases as evidenced by the complete absence of HRP labeled elements rostral to the lesion site. The lesion size ranged from transections of the DC to injuries which extended as far as the dorsal part of the ventrolateral field, including part of the contralateral dorsal horn.
HRP application to dorsal roots 8 or 9 at various times after thoracic dorsal funiculus injury reveals the disposition and response of rostrally projecting primary afferent collaterals to injury. Rostrally projecting DRG collaterals typically extended as far as the caudal part of the lesions and did not appear to grow through or around lesions in larval or juvenile Rana catesbeiana.
At and near the lesion site collaterals arborized in the dorsal gray matter but did not do so in a predictable pattern. The arborizations were disorganized. In two cases HRP labeled axons were observed rostral to the putative lesion sites. The fibers were observed in the intermediate part of the DC along the pial surface. 66
Whether these fibers represent successful regeneration or
not is unclear. In one of the cases the lesion site was
difficult to identify and in the other the lesion site
was at the entry zone of the labeled root.
The morphology and disposition of HRP filled
elements at or caudal to a lesion site was quite abnormal
in both larval and juvenile anurans. Aberrant axonal
trajectories were frequently observed and included
altered orientation of fibers as they entered the cord
and further extensions of fibers into regions minimally
innervated in normal cases. For example, axons from the
tract of Lissauer extended into the lateral funiculus
[Figs. 7 & 12], more than in normal subjects [e.g.
compare Fig. 2 with Fig. 7]. This was true especially if
the lesion site was close to the root entry zone and if the survival time exceeded 4 weeks [n=3]. The smallest
caliber axons observed with this abnormal extension were associated with the subpial border along the lateral
funiculus. Fibers in the contralateral dorsal gray were more numerous and their arborizations more extensive in the lesion cases than in normal cases [e.g. compare Fig.
2 with Figs. 8 & 9? also see Fig. 11 a high power micrograph of a crossing fiber at the midline]. Subjects that survived the longest had the greatest propensity for 67 exhibiting these fibers [n=5]. Occasionally, axons from the lateral division showed a pronounced medial to lateral orientation at the root entry zone [n=3] [Fig.
10]. This pattern was seen at various stages and survival times [e.g. 1 week XIV-XIV? 10 weeks VII-XV? 13 weeks XV-juvenile]. Axons were often enlarged, with club shaped endings reminiscent of retraction bulbs [Figs. 15,
16, 18 & 19 arrowheads]. Some of the large swellings
[e.g. the structures in Figs. 14 & 16, block arrowheads, are approximately 25um in diameter] had fine protoplasmic extensions which made them appear like growth cones [e.g.
Figs 14, 16, 17, 19 and 20 small arrows]. Large varicosities were observed along axonal elements [e.g.
Figs. 16, 18 & 20 open arrowheads]. Altered axonal morphology of some sort was seen in all DC lesion cases.
There was some evidence for consecutive retraction and expansion of terminal arborizations when the lesions were near the root entry zones [e.g. compare Fig. 13, 2 week survival (stage XX) with Fig. 15, 4 week survival
(stage XII) and with Fig. 8, 13 week survival
(juvenile)]. 68
2.2.3 DISCUSSION
The appearance of the larval and adult rana catesbeiana lumbar spinal cord at the eighth and ninth segment in HRP labeled cases did not differ, in general, from other descriptions of the anuran lumbar cord. These studies examined the afferent spinal system with a variety of techniques [Sala y Pons, 1892; Gehuchten,
1887; Athias, 1897; Ramon y Cajal, '09; Ariens-Kappers,
'36; Silver, '42; Nieuwenhuys, '64; Stensaas and
Stensaas, '71; Cruce, '74; Ebbesson, '76] including the use of HRP histochemistry [Forehand and Farel, '82a & b;
Jhaveri and Frank, '83; Liuzzi et al., '83, '84 & '85;
Rosenthal, '85], and cobaltic-lysine [Szekely et al.,
'82).
As predicted by similar experiments [Piatt and
Piatt, '58; Stensaas, '83; Katzenstein and Bohn, '84; Sah and Frank, '84; Liuzzi and Lasek, '85, '86a & b; Clarke et al., '86; Frank and Sah, '86], rostrally projecting collaterals of lumbar DRG's were not observed to regenerate through or around lesions of the dorsal funiculus in larval or juvenile Rana catesbeiana.
Similar studies in larval Rana temporaria have shown a propensity for lumbar DRG axons to traverse thoracic DC lesions in larvae younger than stage VIII but not afterward [Clarke et al., '86). This study did not 69
include tadpoles from stages before VIII [Fig. 1], which
lends support to their observation of the stage-dependent
regenerative responsiveness of DR axons. All of the
studies concur that after lesions of DR axons longitudinally oriented axonal processes are particularly scarce in the dorsal columns [DC] but are found in the LT and the gray matter [Katzenstein and Bohn, '84? Sah and
Frank, '84; Liuzzi, '83 and '86a; Frank and Sah, '86; present study]. One question which arises is whether the mature DC is non-permissive for axonal extension while the immature DC or other regions, like the LT or gray matter, are permissive.
Whether DRG axons are transected by dorsal root crush [Liuzzi and Lasek, '85, '86a & b], dorsal root cut
[Katzenstein and Bohn, '84], dorsal root freezing [Frank and Sah, '86], or DC injury [Clarke et al., '86], the consequences include accumulation of many small, glial fibrillary acidic protein [GFAP] negative [Liuzzi and
Miller, '85 in dorsal root crushes] glia that appear to be microglia [Liuzzi and Lasek, '86]. Also, astrocytic processes hypertrophy [e.g. see Reier, '83], leading some investigators to identify "glial scars" as a barrier to regeneration [ibidem]. Radial glial processes are widely separated in the mature DC [Miller and Liuzzi, '86]. 70
This coupled with the observation that HRP filled, regenerating axonal elements are generally associated with glial elements [i.e. subpial glial endfeet or radial glial processes] led Liuzzi and Lasek ['86a] to suggest that the DC was non-permissive for axonal extension because of the paucity of glial processes available for axonal attachment [i.e. the DC is heavily myelinated whilst the LT is lightly myelinated].
Long after DC injury and/or DRG axonal transection, myelin debris remains in the anuran DC [Stensaas, '83;
Liuzzi and Lasek, '84]. This led Stensaas ['83] to suggest that myelin breakdown products may be deleterious to axonal extension, a factor reduced or absent from the lightly myelinated DLF [see also Berry, '82].
A second issue which arises in analyses of regeneration in developing organisms is whether the apparent reconstitution of axonal projections is attributable to the regrowth of cut axons or the ingrowth of developing axonal populations. The stage dependent regenerative potency of DC axons in larval Rana temporaria is critical around stage VIII [Clarke et al.,
'86] and the affinity for developing DRG axons for the DC is acute [Holder et al., '87]. Also, at this stage, DRG axons have scarcely begun to enter the spinal gray
[Liuzzi et al., '83], although a few ascending DRG axons 71
have reached the dorsal column nuclei [DCN] as early as stage I [Forehand and Farel, '82a & b]. Many other examples of developmentally dependent regeneration are seen in anura [Michel and Reier, '79; Forehand and Farel,
'82c; Davis and Farel, '85; Lopate et al., '85] and mammals [Kalil and Reh, '79; Bregman and Goldberger, '82
& '83; Bernstein and Stelzner, '83; Schreyer and Jones,
'83; Tolbert and ti Der, '87; Martin and Xu, '88].
BLASTEMA IMPLANTATION
The tail and tail spinal cord of larval anurans and some species of adult reptile and urodele can regenerate following amputation [Spallanzani, 1768; Hooker, '15;
Holtzer, H., '51 & '56; Piatt, '55; Butler, E. G. and
Ward, '66; Simpson, '70; Egar and Singer, '72; Egar et al., '72; Stensaas, '83]. Since blastema cells have been hypothesized to participate in this regeneration
[Simpson, '83], an experiment was devised to introduce tissue, rich in blastema cells, into the thoracic spinal cord. Tissue containing these mesenchymal cells can survive implantation into the thoracic cord [Fig. 4] and can support the ingrowth of DRG axons [Figs. 5 & 6]. In the only surviving long term graft the implant supported axonal ingrowth, however, rostrally projecting regeneration of DRG axons in the DC was not observed. 72
REORGANIZATION OF DRG AFFERENTS AFTER ROSTRAL LESIONS
Examination of the HRP filled elements caudal to the lesion site both after short and longer survival periods showed that these elements are capable of plasticity.
Examples of what were taken to be growth cones [Figs. 14,
16, 19 and 20 block arrowheads] and elements similar to retraction bulbs [Fig. 16; large arrow and 18] could be seen. Also, axonal expansion into areas not normally as densely occupied by DRG elements were noted [Figs. 7-12].
The structures thought to be growth cones were considerably larger [i.e. averaging 25um in diameter] than comparable elements in developing anurans which do not exceed 10 um [Liuzzi et al., '85 and Nordlander, '84] or in adult regeneration of motoneuron axons into the spinal cord [i.e. about 10 um in diameter] [Liuzzi and
Lasek, '86]. Liuzzi and Lasek ['86] do show large bulbous appendages along the shaft of regenerating motoneuron axons which are about the same size as the elements seen in this study, but which lack filopodial processes. Xenopus laevis embryos display larger and more complex growth cones than larvae and axons growing along preformed paths have small growth cones [Nordlander
'84]. Furthermore, pioneer growth cones become quite large in various insect embryos at pathway choice points
[Berlot and Goodman, '84; Bently and Caudy, '83; Bently 73
and Keshishian, '82]. Local guidance cues are present in
the spinal cord which direct DR axons and motoneuron
axons along similar courses of arborization [Liuzzi and
Lasek, '86 a & b]. This suggests that the size of the elements observed in this study are related to exploratory movements of the growing tips of axons.
These types of events indicated that DRG cells exhibited some reorganization including possible collateral sprouting, ingrowth into denervated areas and/or a pruning response. While there were striking changes in the morphologies of axonal elements at the root entry zone, the observation of increases in terminal fields due to sprouting or pruning was judged to be more subtle. A systematic, quantitative analysis of these potential reactions to DC lesions was therefore deemed necessary [see section 2.3].
2.2.4 CONCLUSION
Injury-filling of dorsal roots with HRP allows detailed examination of the morphology and distribution of identified dorsal column axons and their collateral arborizations in the larval and juvenile frog. This allows analysis of the effects of both direct axotomy and the indirect effects of axotomy on intact collaterals to the spinal gray caudal to a lesion. 74
It appears that the capacity for DRG axons to
traverse a DC injury is developmentally dependent.
Whether this is due to the maturity of the axonal
population or the milieu the axons must traverse is not
known. It is also not known whether ingrowing rather
than transected axons are the constituents of early
successful regenerates.
Tissue from regenerating tadpole tail (i.e. tail
blastema) can survive transplantation into the dorsal
columns, and HRP labeled axons enter the implant. The
implant does not seem to support the rostral extension of
DR axons.
Transection of the dorsal columns at various stages
of development produces an array of axonal responses which include the formation of large, swollen tips, growth cone-like endings, aberrant axonal orientation and trajectories, and in some cases, an apparent early retraction of segmental collaterals caudal to the lesion.
This cluster of observations is implicative of collateral sprouting in response to denervation or of a "pruning effect" as a result of the damage to the ascending DRG axonal elements. The following study is an attempt to address these observations in a more quantitative fashion. 75
2.3 PLASTICITY IN JUVENILE XENOPUS
Evidence from experiment 2.2 suggested that DRG
axonal elements are capable of reorganization after DC
lesions. The intent of this experiment was to
quantitatively describe some of those differences,
focusing on the projections of the tenth dorsal root.
Juvenile and larval Xenopus laevis frogs received
unilateral lesions of their thoracic dorsal funiculi.
Afterwards, HRP was applied to the cut 10th spinal nerve peripherally [Fig. 21 & 54]. Several questions were
addressed: 1) Do the dorsal and ventral terminal fields
[DTF and VTF, respectively], expand or retract after thoracic DC lesion? 2) If so, are the changes due to an
increase in the density of elements or an increase in the
area occupied by the DTF and VTF, or both? 3) Do the
sizes of axonal varicosities increase consequent to a thoracic lesion? 4) Does sprouting of DRG axonal
elements into the contralateral cord after thoracic DC
injury occur? A preliminary account of this experiment has been reported [Campbell et al., '88a & b]. 76
2.3.1 METHODS: Xenopus laevis
TISSUE PREPARATION
Fifty Xenopus laevis tadpoles [stages 57-61 of
Nieuwkoop, '56] and seventy-two juvenile frogs were used.
Tadpoles and juveniles were anesthetized by immersion in an aqueous solution of 6% tricaine methanesulfonate (MS-
222) . Larger juvenile frogs were anesthetized by injection of MS-222, at 0.1-0.4mg/g body weight, into the dorsal lymph sac. The spinal cord exposures and DC lesions were made as previously described [Section
2 .2 .1 ].
The juvenile xenopus preparations were allowed to survive for 3 days [n=5], l week [n=7], 2 weeks [n=8], 3 weeks [n=5], 5 weeks [n=5], 7 weeks [n=5], 10 weeks [n=5] or 13 weeks [n=5] prior to sacrifice. After these post lesion periods, the axons of the 10tl1 spinal nerve, ipsilateral to the dorsal funicular lesion, were cut peripherally and chips of dried, concentrated HRP (type
IV, Sigma Chemical Co.) were applied to the proximal cut ends [Fig.34]. The period of HRP application ranged from
15-30 minutes. At the end of this procedure excess HRP was flushed away and the wound edges were approximated with 7-0 silk suture. After 56 hours the animals were perfused intracardially with fixative and their spinal 77
cords prepared after the manner described in experiment I
[2.2.1], At least one subject from each timed survival
group was sectioned on a vibratome with every third
section being prepared for sequential light and electron
microscopy (EM) as described in the first experiment
[2.2.1] (EM will not be reported here).
In four cases, DC lesions of the thoracic cord were
performed, followed by the application of HRP chips to
the same cuts. The intent of this procedure was to
backfill tenth segment DRG collaterals that reach the
lesion site. The animals were perfused after 24 hours.
In seventeen cases without lesions, the 10th spinal nerve
was injury filled with HRP. These animals were perfused
after 56 hours. The tissue was sectioned and processed
as previously described.
DATA ANALYSIS
This analysis focused on the 10fc^ root entry zones
[REZ] and the thoracic lesion sites in order to determine
if any localized reorganization of axons at the REZ had
taken place. In order to measure the local collateral plexus, optical density readings of HRP labeled elements
in five sections through the tenth REZ in normal and
lesioned animals were obtained with an interactive
computer-assisted image analysis system. The light microscopic image was picked up by a television camera 78
(Dage-MTI series 68, Newvicon, attached directly to a
Leitz orthoplan microscope), the video signal was
directed through a real time video processor (Nippon
Avionics, model Image E), then into a computer (Magiscan
2A, Joyce-Loebl) for digitization and analysis [Fig. 22],
The video processor served to change the video signal
from a logarithmic intensity function into a linear density function. It also effectively removed density variances picked up from the light source, the microscope optics and the camera by subtracting a reference image of those variances from the section being examined [Fig. 23
& 24]. Finally, the video processor served to adjust the
signal toward a range appropriate for digitization by the computer. The video image amplification was adjusted so as to assign a value of zero (0.00) to the unstained tissue and a value of one (1.00) to the most densely labeled HRP filled neural elements [Fig. 25]. The video image was then digitized [Fig. 26] and the computer analysis begun. Once digitized, the entire hemicord, gray matter, the HRP-filled dorsal and ventral terminal fields and the lateral motor columns were outlined with a light pen. The computer was then instructed to calculate the detected area and optical density for each region.
The computer analysis was conducted on low power images 79
(6.3X). One-way analyses of variance were performed
comparing the detected area and density of HRP labeled
elements from the normal and lesioned animals. When a
significant overall effect was noted, post-hoc analyses
were made using the Scheffd test.
Additional analyses were conducted in order to
determine whether: 1) the number and size of HRP labeled
varicosities had increased in the ventral terminal field
as a result of the DC lesion and 2) if more fibers
crossed the midline and if the extent of their
contralateral arborizations increased as a result of DC
injury. In order to address the first concern, HRP
filled varicosities within the ventral terminal field were drawn at 100X using a drawing tube attached to the microscope. The method of sampling involved drawing all
of the varicosities within a 120 x 120 mm square (i.e.
representing 660 X 660 um in the tissue). The square was placed so that it framed a sample of varicosities from the medial-most aspect of the ventral terminal field in the normal group [n=5], 3-day [n=4], 5 week [n=5] and 13 week [n=4] post-lesion groups. The varicosity drawings were analyzed using a Zeiss Videoplan morphometries
computer by tracing the contours onto a digitization pad.
One way analyses of variance were performed, comparing the number and average size of varicosities in normal 80
cases with those of the lesioned groups. Shifts in varicosity size were evaluated using analyses of variance.
Measurements of the number and extent of contralateral axonal arborizations at the 10th REZ were determined from 4Ox camera lucida drawings of 5 consecutive sections through the 10th root entry zone in normal [n=5], 3-day [n=4], 5 week [n=5] and 13 week [n=4] cases. A Zeiss Videoplan morphometries computer was used to determine the total length of HRP filled elements occupying the contralateral dorsal gray. The number of branch points observed in each group was tabulated.
Also, counts were made of HRP filled elements in the medial 1/3, central 1/3 and lateral 1/3 of the contralateral cord. Comparisons of the proportion of contralateral elements in each of the three divisions, across groups were made with the chi square test. 81
2.3.2 RESULTS: Xenopus laevis
Application of HRP to the tenth spinal nerve labels primary afferents and motoneurons [Figs. 27-31].
The pattern of primary afferent distribution at the tenth segment is similar to that seen at other levels of the lumbar or brachial cord [Cajal, '09; Ebbesson, '76;
Antal et al., '80? Szdkely et al., '82; Nikundidwe et al., '82; Liuzzi et al., '84; Rosenthal and cruce, '85].
Tenth root afferents enter the spinal cord and separate into medial and lateral divisions [see Fig. 27 & 31c].
The medial division, contains large diameter myelinated axons, some of which travel directly to the VTF [Figs. 27
& 31c] and the majority of which extend into the dorsal funiculus. Some of the fibers that pass directly to the
VTF pass through the dorsal part of the DTF in transit
[see Fig. 31a-e]. The lateral division is composed of small diameter unmyelinated or lightly myelinated fibers.
Three distinct regions of tenth root afferent termination are apparent from the REZ through the thoracic cord. One is located around the LT which exclusively contributes to its composition [Figs. 27 and
31a-e, medial to HLTM]. The other two fields, the DTF
[Figs. 27, 28 and 31] and VTF [Fig. 27, 29 and 31] receive their DRG axonal termination from medial division fibers in the dorsal funiculus and from the dorsal root. 82
In some instances, particularly at segments rostral to
the lO^1 REZ, fibers with en passant boutons in the DTF
continue into the VTF and arborize [Fig. 32 & 33).
Crossing fibers from the DTF and VTF occasionally extend
contralaterally at the tenth segment [Fig. 30] and at the
rostral lumbar segments. Caudal to the tenth segment
many of these elements cross to the contralateral gray.
Collaterals from the medial division of the tenth
segment extend rostrally, in the dorsal funiculus into
the brainstem, assuming a midline location in their
ascent. Axonal processes leave this medial bundle and
arborize within dorsal and ventral terminal fields well
into the thoracic spinal cord [Fig. 31a & b). In the thoracic cord, fibers entering the VTF leave from the
lateral part of the 'gracile fasciculus' whilst those terminating in the DTF emerge from the medial aspect
[Fig. 31a].
Motoneurons in the lateral motor column, medial motor column and previously undescribed neurons in the lateral field were labeled following HRP application to the tenth spinal nerve. The labeled lateral field cells were located from the caudal 10th through what was presumed to be the 11th segment [see chapter 3],
Unilateral lesions of the dorsal column of the 83
thoracic cord cut the rostral collaterals of a contingent
of 10th root afferents. The number of axons transected
can be assessed by an examination of section A in Fig.
31. This section is at the approximate level at which
the dorsal column lesions were made. The extent of the
lesions is shown in Fig. 35 and always included the
dorsal column fibers and the LT fibers. Figure 54 is a
section through the tenth root entry zone and shows the
result of HRP application to a dorsal quadrant lesion in
the thoracic cord. Many HRP filled elements are seen in
the DC and in the lO*'*1 root. The HRP filled elements in
the lateral funiculus of figure 54 are likely propriospinal, rubrospinal and scattered raphespinal axons.
The rostrally projecting collaterals of lumbar DRG's were not observed to regenerate through or around lesions of the dorsal funiculus in juvenile cases. Fibers were observed just caudal to the lesions and some axons were seen in the gliotic area [Fig. 34, 36 and 38]. However, no labeled elements were traced rostral to the lesion in any case. In one of the cases, the lesion was misplaced just rostral to the 10th root entry zone. The entering fibers [see Fig. 38] were quite similar to those observed in the Rana catesbeiana cases described in section 2.2.2.
Large club like endings in the dorsal root were observed 84
[Fig. 36 & 37], Abnormal axonal trajectories were seen
[Figs. 36 and 38]. Small diameter fibers coalesced along
the DC subpia [Fig. 38; small arrows]. Figs. 36 [i.e.
ventral to the root] and 38 [i.e. large arrow] show axon
collaterals bent into hairpin turns at the root entry
zone. The high cell density evident in the DC [Fig. 38]
suggests that these abnormal fibers are in a gliotic
area. These results suggest that lesions close to the
entry zone have somewhat different effects than distant
lesions on the reaction to injury of DRG central axons.
Of the fifty xenopus tadpoles lesioned forty-eight
died at metamorphic climax. Therefore, analysis was not
possible.
A comparison of the primary afferent distribution
in normal and DC lesioned xenopus juveniles, at the tenth
segment, showed that significant changes in the size and
density of the dorsal and ventral terminal fields develop
after thoracic DC injury.
Figure 39 shows that the relative areas of the DTF
and VTF increased subsequent to rostral/ipsilateral DC
injury. An analysis of variance indicated that the detected areas of both fields were significantly different across groups (VTF: F=16.31, df=3, 30, p < 0 .0001; DTF: F=6.5, df=3, 30, p<0.005). Post-hoc 85 analysis of pooled data (early*= 1-3 wks.; middle= 5-7 wks.; late= 10-13 wks.) using the Scheff6 test showed that: 1) the dorsal terminal field is larger than normal in the 5-7 week groups (normal vs. middle: F= 5.24, df=
3, 30, and p<0.01) and 2) the ventral terminal field is larger than normal in the 5-7 week groups (normal vs. middle: f= 5.99, df=3, 30, p<0.005) and then declines in size by 10-13 weeks, although it does not return to the size it was at 1-3 weeks (early vs. late: F= 6.77, df= 3,
30, p<0.01).
Figures 40 and 41 show that the summed optical density of the HRP filled elements within the DTF and VTF increased subsequent to rostral/ipsilateral DC injury.
An analysis of variance indicated that the optical density of each field changed over time after the lesion
(VTF: F=14.1, df=3, 30, p<0.0001; DTF: F= 7.17, df= 3,
30, p<0.001). Post-hoc analysis of the pooled data
(early= 1-3 wks.; middle= 5-7 wks.; late- 10-13 wks.) using the Scheffd test showed that: 1) the dorsal terminal field is more dense in the late [10-13 weeks] group than in the early [1-3 weeks] (early vs. late: F=
6.66, df= 3, 30, and p<0.01) and 2) the ventral terminal field is more dense in the late group (10-13 weeks) than in the normal or other post-lesion groups (late vs. normal F= 5.06, df= 3, 30 and p<0.01; late vs. early 86
[1-3 weeks] F= 13.71, df= 3, 30 and p<0.01; late vs. middle [5-7 weeks] F= 3.96, df= 3,30, and p< 0.05).
The average size of varicosities in the medial aspect of the ventral terminal field was not significantly different across groups [F=0.39, df= 3, 14 and p> 0.10]. Within the range of varicosity sizes, there were no significant shifts in proportion across groups
[Fig. 43] nor was there a significant increase in number of varicosities within the sampling area [F= 62.59, df=
3, 14 and p< 0.38] [Fig. 42].
The number of axons that occupy the contralateral hemicord at the tenth segment and the extensiveness of their terminal arborizations was greater in the lesion groups [e.g. compare normal (Fig. 44) with examples from
3 day (Fig. 45), 5 week (Fig. 46) and 13 week (Fig. 47)].
The sum of the lengths of HRP filled elements in the contralateral cord at the tenth segment was markedly higher in the thirteen week group than in any other group
[Fig. 48]. An analysis of variance showed a significant difference across groups [F= 32.06, df= 3, 14 and p<0.0468]. Post-hoc analysis using the Bonferroni test demonstrated that the 13 week post-lesion group was indeed significantly larger than either the control group
[F= 7.01, df= 1, 14 and p<0.05] or the early groups [F= 87
8.55, df= 1, 14 and p<0.05].
The gray matter directly contralateral to the 10th root entry zone was parceled into medial, intermediate and lateral thirds and the HRP labeled fibers were counted for each region. The purpose of this analysis was to ascertain if the probability of finding HRP labeled elements in the more lateral aspects of the contralateral spinal cord increased after DC lesion [Fig.
49], An analysis of variance showed no significant difference in the numbers of fibers across groups [F=
34.55, df= 3, 14 and p<0.0.75]. However, because the inter-animal variability was high [see the error bars in figure 49] a chi-square test was performed on the same data. This analysis showed that the proportion of crossing fibers in each zone was significant across groups [x2= 29.86, p< 0.01]. Subsequent testing of pooled data showed that the proportion of fibers appearing in the lateral aspect of the contralateral cord was significantly greater in the thirteen week group than in any of the other groups [x2= 26, p< 0.01]. The data were pooled for the previous test in such a way that four groups were compared with one another: 1) normal, 3 day and 5 week medial; 2) normal, 3 day and 5 week intermediate and lateral; 3) 13 week medial and 4) 13 week intermediate and lateral. 88
Another measure of the extensiveness of the contralateral arborizations was to count the number of points at which axonal elements branched [Fig. 50]. An analysis of variance showed a significant increases in branching across groups [F= 31.18, df= 3, 14 and p<
0.0001]. Subsequent analysis, using the Bonferroni test, showed that the number of branch points was significantly greater in the 13 week group than in the control group
[F= 19.1, df= 3, 14 and p< 0.01]. 89
2.3.3 DISCUSSION
The results of this experiment suggest that regeneration of tenth DR axons through or around a thoracic DC lesion does not occur in juvenile Xenopus frogs. However, plasticity of tenth DR collaterals at the tenth segment was evident. The central processes of these DRG cells crossed into the contralateral dorsal field consequent to an ipsilateral thoracic DC lesion.
Also, the density and detected area of the ipsilateral dorsal and ventral terminal fields increased, as measured by a computer aided image analysis system.
As anticipated by a similar experiment in Rana catesbeiana [Chapter 2.2] and analogous studies [Piatt and Piatt, '58; Stensaas, '83; Katzenstein and Bohn, #84;
Sah and Frank, '84; Liuzzi and Lasek, '85a & b; Clarke et al., '86; Frank and Sah, '86], cut DRG axons were not observed to regenerate through the DC lesion.
Regeneration through the dorsal column region was not observed even when the lesion was placed very close to the root entry zone. So it appears that this region is not permissive for regrowth [see 2.2.3 for a complete discussion of this issue].
The results of this experiment show that after a thoracic DC lesion both the number of axons crossing into the contralateral dorsal and lateral fields and the 90
length of these elements increases [Figs. 44-50], This suggests that the caudal collateral plexus can increase in response to a rostral lesion. Such increases in the collateral arborization extents in the contralateral spinal gray could potentially be induced by denervation of that region. Alternatively, pruning of the rostral collaterals of the DRG neurons might induce growth of collaterals into the contralateral spinal gray. An additional possibility is that both stimuli could act together. The results of the control procedure in the present experiment in which HRP was introduced into the lesion site in the thoracic cord, show that there are labelled descending fibers projecting contralaterally into the spinal gray. The source of these fibers is not known, but based on extrapolation from other experiments that label descending systems from the brainstem [ten
Donkelaar et al., '81; Toth, '85] without contralateral extension, it is probable that they are propriospinal in origin. Also, since these axons can be traced back to the lateral funiculus they are probably not backfilled primary afferents.
Another interesting observation was that increases in contralateral sprouting did not occur in segments that normally have very few crossing fibers. This implies 91
that denervation or pruning are necessary but not
sufficient conditions for the observed growth effects.
Perhaps the extracellular milieu is permissive for
contralateral axonal sprouting in the caudal segments,
for segments 10 and 11 normally contain significant
numbers of axons from the contralateral tenth dorsal
roots. In R. pipiens [Rosenthal and Cruce, '84 and '85],
no fibers were reported to cross except at the 10th
segment. Tenth segment afferents extend very heavily
into the contralateral spinal cord at the caudal level.
Caudal spinal cord may have a greater capacity for
stimulating axonal growth. Certainly, studies in tail
cord of reptiles [Simpson, '70; Egar et al., '72],
urodeles [Spallanzani, 1768; Hooker, '15; Holtzer, '51 &
'56; Butler and Ward, '66; Egar and Singer, '72;
Stensaas, '83] and larval amphibians [Piatt, '55], have
shown extraordinary examples of regenerative abilities by
neurons in this region of the spinal cord.
COMPUTER IMAGING RESULTS
Experiments involving large lesions of the anuran
rostral spinal cord [Piatt and Piatt, '58] or dorsal
roots [Frank and Westerfield, '82b; Liuzzi and Lasek, '83
& '85; Katzenstein and Bohn, '84; Sah and Frank, '84;
Frank and Sah, '86] have produced many interesting
examples of central DRG axonal plasticity. Small lesions 92
involving dorsal root afferents, as in the present
experiment, could result in subtler types of plasticity.
Therefore, quantitative analysis was used to insure that
changes which occur in response to small lesions were
detected. Computer assisted image analysis provided
rapid, measured answers to the questions of this study,
particularly in the context of confounding problems such
as variability in the density of HRP reaction product,
differences in cord size and fibers of passage through
the DTF.
This experiment shows that DC lesions which
axotomize some central processes of the 10th DRG are
followed by plasticity. The detected area and density of
DRG elements in the DTF and VTF increases. The density
of these terminal fields is largest at 10-13 weeks while
their detected area is largest at 5-7 weeks.
INCREASES IN THE AREA OF THE DTF AND VTF
Why would the detected area of the DTF and VTF
increase 5-7 weeks after a thoracic DC lesion? Four possible explanations are: 1) the white matter decreases
in size increasing the ratio of terminal field area to hemisection area and thereby falsely inflating the
relative size of the terminal fields; 2) HRP is transported to the tips of axons not normally filled, for 93 a limited time; 3) the gray matter expands 5-7 weeks after a thoracic DC lesion or 4) sprouting responses are expressed by DRG axons subsequent to rostral DC injury.
If tenth segment white matter were to shrink in size after 5-7 weeks post-lesion and if the gray matter were to remain unchanged during that time, so as not to reduce the terminal field area, then there would necessarily be an apparent, albeit false, increase in the relative size of the DTF and VTF. However, there is no obvious decrease in cross-sectional area of the ipsilateral tenth segment white matter after thoracic DC injury. In a similar experiment where unilateral hemisections were performed in the thoracic cord and where both tenth spinal nerves were labeled, there was no measurable shrinkage at the tenth segment [Norris et al., '88].
Therefore, it seems unlikely that the significant expansion of the DTF and VTF is due to changes in the size of the white matter at segment 10.
If, for a limited period of time [i.e. 5-7 weeks after a thoracic DC lesion], HRP is carried to parts of the DRG axonal tree not normally filled with HRP, then there would be an apparent increase in DTF and VTF size.
HRP is believed to completely fill the central processes of DRG axons through the mechanism of diffusion [Beattie et al., 778] and if labeled distal to the ganglion 94
through active transport as well [Mawe et al., '83],
Large and small axonal elements, and fine terminal arborizations are filled with HRP reaction product in each group, whether normal or lesioned. It is improbable that an observed increase in DTF and VTF size is due to short term changes in the ability of HRP to diffuse into diffusion restricted domains within the axonal tree. It is possible that thoracic DC lesions would enhance active transport [McQuarrie, '86] but there is no example of lesion primed neurons transporting material into normally occult elements.
If the tenth segment gray matter, ipsilateral to a thoracic lesion, were to briefly expand [i.e. within 5-7 weeks] then the DTF and VTF would appear to increase in size. It is difficult to simply measure the gray matter in this experiment because the white matter is normally quite cellular and the morphometric analysis required that the tissue be unstained and that it be examined at low power [6.3X], However, in cat the gray matter does not shrink in response to a variety of lesions
[Goldberger and Murray, '82; Murray and Goldberger, '86].
Based upon these observations and deductions, it seems that the significant increase in DTF and VTF size at 5 & 7 weeks is due to either dennervation induced 95 collateral sprouting, a pruning response or both. In
order to address these questions it is important to consider what kinds of denervation could conceivably be caused by a thoracic DC lesion. For these purposes a lesion larger than typically produced will be considered.
In the case of a 7 ^ segment dorsal quadrant lesion, besides cutting rostral ascending branches of the 10th dorsal root, the following axonal systems would be interrupted: 1) some rostral axonal extensions from dorsal roots 7-9 [Fig. 53] [note: descending projections from DRG segments rostral to the lesion do not normally reach the tenth segment (Antal et al, '80; Rosenthal and
Cruce '85)]; 2) the rubrospinal tract [Fig. 52]; 3) a few scattered raphespinal projections [Fig. 52]; 4) some of the descending propriospinal projections from the brachial cord.
It is possible that the effect of cutting the rostral processes of collaterals from the seventh, eighth or ninth segments would cause their caudal extensions at the tenth segment to chronically retract. Expansion, not retraction of these terminations within the tenth segment, would be expected based on the results of this experiment but lesions closer to the cell bodies might produce a different result. If rostral primary afferents expand their tenth segment dorsal and ventral terminal 96
fields at times later than the tenth segment primary afferents [i.e. after 5-7 weeks], then the observation that the terminal fields shrink at 10-13 weeks might be explained as competition induced retraction [Goldberger and Murray, '85].
It is possible that consistently extensive lesions in the thoracic cord could consistently denervate the ipsilateral tenth segment by interrupting the rubrospinal tract, some raphespinal and some propriospinal projections. However, these axonal projections do not distribute equally to the dorsal and ventral terminal fields. Rubrospinal and propriospinal axons from the dorsal part of the lateral funiculus end in the LMC, lateral field, and ventrolateral field [Ebbesson, '76; ten Donkelaar, #82]. Raphespinal projections from the lateral funiculus end within the LMC, lateral and medial parts of the dorsal field but very slightly within the lateral field [van Meir, '86].
The changes seen in the DTF and the VTF are quite parallel, suggesting that the mechanism driving their expansion is the same for both. If denervation were the primary stimulus for expansion, the lesser degree of denervation in the DTF might be expected to result in a smaller effect. The extent of the lesions outside the DC 97 varies between animals and the normal distributions of the rubrospinal, raphespinal and propriospinal are not equal to both terminal fields. The parallel expansion of the DTF and VTF is, therefore, consistent with a pruning effect. Denervation induced collateral sprouting is also possible, however.
INCREASES IN THE DENSITY OF THE DTF AND VTF
The density of the DTF and VTF could increase at 10-
13 weeks for a number of reasons: 1) tenth segment spinal gray, ipsilateral to a thoracic lesion could decrease in size; 2) HRP diffusion could be enhanced at this time, revealing a previously hidden expanse of DRG terminal arborization; 3) varicosities within the terminal fields could increase in size or number; 4) sprouting or pruning effects, restricted to the terminal fields could be manifest; 5) the terminal fields could shrink without an equivalent absorption of retracted processes.
Gray matter does not shrink appreciably after dorsal rhizotomy in cats [Goldberger and Murray, '82; Murray and
Goldberger, '86]. The ipsilateral tenth segment hemicord does not shrink after thoracic DC lesions [Norris et at,
'88]. These lines of evidence lead to the supposition that the increase in density seen in this experiment is 98
not due to shrinkage of the gray matter.
It is improbable that the ability of HRP to diffuse
into occult axonal domains is briefly enhanced after 10-
13 weeks, for reasons previously discussed.
If the number or size of the varicosities within the terminal fields increased then the density of those
regions would concomitantly increase. Varicosities from the medial aspect of the VTF were measured. There was no significant change in the size, number or proportional distribution of these elements within the sample [Figs.
42 and 43]. Therefore, the increase in density is probably not due to changes in the number or size of varicosities, at least in the region sampled.
Increases in VTF and DTF density could be due to an increase in the number of axonal elements from sprouting or pruning effects. However, this would require that the expressed plasticity be restricted to the bounds of the terminal fields, because the size of these regions is not significantly larger than normal at 10-13 weeks post lesion [Fig. 39],
It is possible that the terminal fields expand through some sprouting mechanism after thoracic DC injury and then, because of some competitive interaction between tenth dorsal root axons and other sprouting elements, withdraw their newly generated processes into their 99
normal terminal fields by 10-13 weeks. Competition and
specificity has been widely investigated [e.g. see
Cotman, '85; Murray, '86; Frank and Sah, '86; Mackler and
Selzer, '87; Liu and Chambers, '58; Pubols and Sessle,
'87] .
2.3.4 CONCLUSION
The adult vertebrate CNS, with some exceptions, is a difficult milieu for regenerating axons to traverse.
The two broad concerns that confront the investigator of spinal cord injury have to do with events related to the re-formation of long axonal projections and the events related to the re-establishment of functionally appropriate synaptic connections. In this study a portion of the axonal arborization was compromised resulting in sprouting of the spared collaterals. Some evidence was shown for the existence of local guidance cues and zones of restriction. The results of this experiment also support the use of computer aided image analysis in investigations of more subtle types of plasticity. CHAPTER III
3.1 INTRODUCTION
AMPHIBIAN CLOACA AND ASSOCIATED STRUCTURES
The gross anatomy of the anuran pelvic viscera is
shown in figure 58. The anuran cloaca is a tubular
structure that receives the lower bowel anteriorly
(labeled "C") and the bladder ventrally (labeled uBn).
Wolffian ducts and oviducts end on the dorsal cloaca. At
its distal end a striated sphincter encircles the cloaca
[Fig. 58, arrow].
The Xenopus laevis urinary bladder is a diaphanous
blind sac that is situated on the ventral aspect of the
cloaca. The anuran analogues for mammalian ureters, the
Wolffian ducts, do not terminate in the bladder, they end
in the cloaca. Urine can flow into the bladder from the
cloaca but does not necessarily do so [Deucher, '75]. A
sphincter is present around the bladder neck at its
interface with the cloaca [ibidem]. The cited descriptions of this sphincter do not state whether it is
composed of striated or smooth muscle. The anuran
100 101
bladder functions in water regulation [Chester-Jones et
al., '72]. Urine excretion from the bladder is a
consequence of its use in systemic water balance.
Aldosterone and cortisol stimulate sodium transport by
the urinary bladder [Chester-Jones et al., '72].
In anurans, branches from spinal nerves 9, 10 and in
some cases 11 compose the ischiococcygeal plexus. Nerves
pass from this plexus to the bladder, cloaca, oviducts
and dorsal lymph hearts [Duellman and Trueb, '86],
The present experiment sought to determine the
organization of the spinal cord components innervating
the Xenopus pelvic viscera. In addition, an attempt was made to characterize the organization of the sacral
neural components in the region which might be comparable
to those described for mammalian species. In mammals the
pelvic nerve innervates the bowel and bladder, whereas
the pudendal nerve innervates the striated urethral and
anal sphincters [DeGroat, '81],
3.2 SACRAL AUTONOMIC INNERVATION IN XENOPUS LAEVIS
The previous Xenopus laevis study provided evidence
for anuran sacral preganglionic neurons and motor neurons
associated with pelvic musculature as a result of
applying HRP to the proximal cut end of the tenth spinal
nerve. In this experiment these observations were 102
extended in order to provide additional information on the innervation of pelvic structures. 103
3.2.1 METHODS
TISSUE PREPARATION
The 10th spinal nerve was labeled in six Xenopus laevis juveniles [Fig. 57 labeled "10 SpN"]. In Xenopus laevis a nerve leaves the ishiococcygeal plexus, adjacent to the urostyle, to innervate the bowel, proximal cloaca and bladder. Further caudal another nerve emerges from the plexus to innervate the cloacal sphincter [Figs. 57 &
58], The cloacal sphincter [Fig. 58, arrow] or the, presumptive pelvic or pudendal nerves [Figs. 57, "PE" or
"PU"] were labeled with HRP in four xenopus juveniles.
The HRP application and tissue processing proceeded as previously described.
DATA ANALYSIS
Sections prepared for light and EM analysis were studied, using a Leitz Orthoplan microscope with bright field and differential interference contrast optics.
Areas of interest were photographed and drawn at several magnifications.
Locations of neurons labeled after HRP application to the pelvic and pudendal nerves, presumed to be preganglionics and motor neurons associated with the cloacal sphincter were plotted in two (2) cases. The plotted sections were reconstructed with the aid of an
IBM AT computer, digitizing pad and Auto-CAD software. 104
3.2.2 RESULTS
HRP application to the tenth spinal nerve labelled two groups of neurons in the caudal portion of the 10th segment (Fig. 59), one group in the lateral field and one in the ventral lateral field in the ventral horn, positioned medial to the lateral motor column (Fig. 59, open block arrow).
The dorsal group can be divided on cytoarchitectonic grounds into three groups of neurons: 1) a dorsal subdivision containing small cells with short, mediolaterally oriented dendrites (Fig. 61; Fig. 62, open block arrows); 2) a lateral subdivision containing small neurons with dorsoventrally oriented dendrites (Fig. 62, closed block arrows) and occasional large neurons which have a dorsal dendrite extending into the lateral funiculus (Fig. 60; Fig. 62 long arrow); and 3) an intermediate subdivision which contains neurons not clearly classed with the other two subdivisions. In some cases, axons from these labelled neurons could be followed as far as the proximal part of the 10th ventral root (Fig. 62, small arrows). The neurons in this dorsal group are in close apposition to synaptic swellings on labelled primary afferents (Figs. 60 & 61, straight arrows). 105
The ventral group consists of a tight cluster of small neurons. Their dendritic arbors are generally mediolaterally oriented.
Labeling of either the two nerves issuing from the ischiococcygeal plexus (Fig. 57, PE and PU) or the cloacal sphincter revealed the same populations of cells in the lateral field and the ventral field. HRP applications to the nerves and cloacal sphincter in different cases labeled these populations of neurons differentially, but unfortunately, not in a predictable manner. It was concluded that there was spread of the enzyme to the other pelvic structures precluding clear identification of the terminal sites of the labeled neurons. Nevertheless, these applications did not label the neurons in the lateral motor column projecting to the hindlimb which are a major component of the 10th nerve.
This allowed at least some differentiation of the target sites, (i.e. hindlimb vs. pelvic structures). In addition it should be noted that the neurons in the 10th nerve cases were diffusely plus granularly filled with HRP whereas in the cases with the more peripheral applications, the label was most frequently only granular.
In the cases with the pelvic structures labeled, the ventral group of cells in the lateral field could be 106
easily distinguished. This column of cells extended well
into the tenth spinal segment; in the 10th root labelled cases, the dense labeling of the LMC obscures the presence of this group of small cells. The intermediate and lateral subdivisions of the dorsal group extended
into the caudal aspect of the 10th segment whereas the medial subdivision neurons could be observed throughout the lumbar and into the thoracic levels. A stereo pair showing the position of labelled neurons in the dorsal and ventral groups is shown in Fig. 63. 107
3.2.3 DISCUSSION
HRP application to the tenth spinal nerve or the pelvic viscera reveals groups of neurons that appear to be amphibian sacral preganglionics and motoneurons distributing to pelvic structures. These elements have not been previously described.
The dorsal group of neurons described in the present experiment, from a comparative anatomic viewpoint, appear to be preganglionic neurons (PGN's).
Those located caudally appear to compose a sacral parasympathetic nucleus (SPN). In an area corresponding to laminae V and VI in mammals, a medial to lateral band of neurons were labeled that are positioned similarly to the dorsal band (DB) cells innervating the distal colon in mammalian species [deGroat et al., '81; Nadelhaft et al., '80; Mawe et al., '86], The primary axis of these neurons extends mediolaterally. Along the edge of the gray matter in the region corresponding to lamina VII in mammals, a string of neurons was also labeled. These elements seemed to be similar to lateral band (LB) neurons that innervate the urinary bladder in mammalian species [e.g. Morgan et al, '79]. The typical orientation of these laterally placed neurons is dorsoventral. Both the DB and LB-1ike neurons in the present experiment received contacts from HRP filled 108
primary afferents entering the spinal cord from the tenth root. Interestingly, similar primary afferent input to sacral preganglionic neurons was observed in the cat
[Mawe et al., '84 & '86]. This input was shown to be synaptic by ultrastructural criteria. The presence of direct dorsal root afferent contacts on preganglionic neurons in such widely divergent species suggests conservation of function for this pathway through higher vertebrates.
Also labeled when the presumptive pelvic nerve was cut and HRP applied, was a medial rostrally extensive but sparse cell group that presumably corresponds to the sympathetic intercalated nucleus described in Xenopus thoracic cord [Robertson, '87]. It was observed by Pick
['57] on the basis of physiological studies, that sympathetic preganglionic axons exited spinal nerves 2-
10. In the current experiment, the label in these cells was almost always light and particulate, which made it difficult to resolve in tenth nerve fills. These cells may have been labeled as a result of diffuse HRP labeling paravertebral sympathetics.
Finally, a ventral group of neurons, the rostral extent which was hidden by the IMC in tenth nerve fills, was also observed. This ventral group of cells is 109 positioned in the ventral horn, similar to the somatic motoneuron group which innervates the sphincter muscles in mammals. This nucleus, termed Onuf's nucleus in man
[Onuf, '08] and its homologue in other species, innervates striated muscle of the external urethral and anal sphincters [de Groat et al., '81; Petras and
Cummings, '78; Schroder, '81]. The conclusion of analogy between the ventral group and Onuf's nucleus is based upon the relatively small size of the neurons, their location in the ventral horn and the fact that they are labelled following HRP application to the pelvic region.
Given the rather primitive nature of anuran sexual and eliminative functions, the similarities to mammals in the organization of the anuran sacral cord is unexpected.
However, it serves to underscore the ubiquity of at least some basic principles of spinal cord organization.
Further study of these circuits in anurans seems, therefore, to be warranted. FIGURE LEGENDS Ill
FIGURE 1
Figure 1 shows the experimental design used for the
ranid experiments. The column labeled "LESION" in the
first table shows developmental stages of the tadpoles and juveniles when they received DC lesions [0P.1 in
illustration]. The column labeled "HRP APPLICATION"
ennumerates the developmental stages at which HRP was applied to the ninth dorsal root in previously lesioned and control animals [OP.2 in illustration]. The last column shows the number of cases from which useable histological preparations were obtained. The asterisk is to indicate that the normal material, prepared by Frank
Liuzzi, was available in the lab. The illustration on the bottom of the figure shows the sequence of surgical operations. 0P.1 indicates the dorsal column lesion and
OP.2 indicates the labeling procedure. EXPERIMENTAL DESIGN
DORSAL COLUMN LESIONS AS A FUNCTION OF DEVELOPMENT LESION HRP APPLICATION + HISTOLOGY I-XI [n=] 5 0 0 XII-XIV [n=J 9 8 3 XV-XXII [n=] 9 6 6 JUVENILE [n=] 4 7 2 TOTAL [n=] 27 21 11 CONTROL HRP APPLICATION AS A FUNCTION OF DEVELOPMENT* I-XI [n=] 6 2 XII-XIV [n=] 10 4 XV-XXII [n=] 8 5 JUVENILE [n=] - 5 3 TOTAL [n=] ~ 29 14
LESION'
______TIME 113
FIGURE 2
Figure 2 illustrates the normal distribution of HRP
filled ninth dorsal root afferents at the ninth spinal
segment. The central projections from DRG cells
segregate into a medial and lateral division. The medial
division fibers are larger in diameter and pass into the
dorsal column [DC]. Lateral division elements are
smaller in diameter and enter the tract of Lissauer [LT].
Cutaneous afferents terminate largely in the dorsal field
[DF] in a region termed the dorsal terminal field [DTF].
Muscle afferents terminate chiefly in the lateral field
[LF] in a region called the ventral terminal field [VTF].
These elements also extend into the central and ventrolateral fields [labeled CF and LF respectively] as well as the dorsal aspect of the lateral motor column
[LMC]. [LF= lateral funiculus; VF= ventral funiculus;
VM= ventromedial field and cc= central canal]. Bar=
200um.
115
FIGURES 3 and 4
Figures 3 and 4 show presumptive mesenchymal cells labeled with tritiated thymidine. Figure 3 is an example of tritiated thymidine label over the nuclei of cells in the amputation stump 24 hours after injection [arrows].
Bar= Sum. Figure 4 shows tritiated thymidine label over nuclei of cells implanted into the spinal cord of Rana tadpoles after 14 days [arrows] [bv= blood vessel]. Bar=
5um. .{fcWP^ rrrtmamqr-
•ivV 117
FIGURES 5 AND 6
Figures 5 and 6 show HRP filled elements within a
blastema implant. In this case an autologous implant, 10
days after amputation, was placed in the DC of a stage
XII tadpole. After 12 weeks the animal had metamorphosed
into a juvenile. In figure 5, dots demarcate the presumptive interface between the spinal cord and blastema implant [cc= central canal]. Bar= 200um.
Figure 6 is a higher power micrograph of HRP labeled
fibers [arrows] within the implant. Bar= 30um.
119
FIGURES 7-9
Figures 7-9 show two different types of aberrant axonal trajectories observed after rostral dorsal column
[DC] lesions. Figure 7 illustrates an extension of elements from the tract of Lissauer into the lateral funiculus [LF] in an animal lesioned at stage XII and labeled at stage XIX, 9 weeks later [see figure 12 and compare figure 7 with 2], Bar- 300um. Figure 8 shows an enhancement of the contralateral axonal arborization. In this case the animal was lesioned at stage XVIII and labeled as a juvenile [i.e. after 14 weeks]. Bar= 200um.
Figure 9 [inset] shows a higher magnification illustration of the area shown in 8. Bar= 75um. uiriOOc
© 121
FIGURES 10-12
Figures 10-12 show micrographs of some of the aberrant axonal trajectories described in the text. Figure 10 shows a medial-to-lateral axonal bundle [arrows] arising from the lateral division and arching under the dorsal column. In this case the lesion and label occurred at the same stage with 1 week between operations. Bar=
75um. Figure 11 is a high power micrograph of a crossing fiber at the dorsal midline [small arrows] [lesion-
XVIII; label- juvenile; 14 weeks]. The crossing fiber seems to be associated with a blood vessel along its course across the midline. One collateral [i.e. at the top of the micrograph, ipsilaterally] turns 90 degrees with respect to its original orientation at the midline.
Bar= 30um. Figure 12 is a high power micrograph of axons from the tract of Lissauer passing into the lateral funiculus [axonal elements within the bounds of the arrows] [lesion- XII; label- XIX; 9 weeks], Bar= 50 um.
123
FIGURES 13-16
Figures 13-16 show elements reminiscent of growth
cones, retraction bulbs and hypertrophic axons. Figure
13 is an illustration of a case shown photographically in
figure 17. In this case large blebs were observed upon
axons in the labeled dorsal root 2 weeks after a DC
lesion [lesion- XX; label- XX], Bar= 200um. At higher
power [Figure 14], these elements [e.g. block arrowhead]
were seen in some cases to have thin extensions
reminiscent of growth cone filopodia [small arrows].
Bar= 100 um. Figure 15 is drawn from a tadpole lesioned
at stage XII and labeled 4 weeks later at stage XVII. In
this case the normal pattern of dorsal root input has
been significantly altered by the lesion. Bar= 200um.
Figure 16 shows two large bulbous swellings [block arrow heads] with many thin filopodia-like extensions [small
arrows]. Shown in the upper left is a large fiber [large
arrow] [shown also in figure 18]. Bar= lOOum.
125
FIGURES 17-20
Figures 17-20 show examples of aberrant fiber morphologies. Figure 17 is a micrograph of the elements
drawn in figure 13 & 14. The small arrows indicate elements like those seen in growth cones including
filopodia and lamellapodia. Bar= 30um. Figure 18 shows a hypertrophic fiber with associated terminal enlargements [block arrowhead]. Enlarged varicosities can also be seen [open arrowheads] [lesion- XII; label-
XVII; 4 weeks], Bar= 20um. Figure 19 is a micrograph of a hypertrophic fiber with a terminal swelling [block arrowhead]. The small arrow points to a small extension
from the edge of the swelling [lesion XV; label- juvenile; 4 weeks]. Bar= 15um. Figure 20 shows a micrograph of the swelling drawn in figure 16; small arrows point to its filopodial extensions. A hypertrophic fiber with a terminal swelling quite similar to that shown in figure 19 is also present in this field
[block arrowhead]. Several examples of large varicosities are also shown [open arrowheads]. Bar=
7 Sum.
127
FIGURE 21
Figure 21 illustrates the Xenopus laevis experimental design. The numbers of normal cases and cases with DC lesions are shown. 0P.1 and OP.2 are separated by intervals from 1-13 weeks. The illustration at the bottom shows the sequence of the surgical preparations. EXPERIMENTAL DESIGN
NORMAL DORSAL COLUMN LESIONS 1 WK. 2 WK. 5 WK. 7 WK. 10 WK. 13 WK. n= 6 5554 5 4
JUVENILE XENOPUS LAEVIS PERIPHERAL NERVE LABELLED WITH HRP...SACRIFICED AFTER 56 HOURS
LESION HRP 10mm
O P.l O P. 2 129
FIGURES 22-26
Figures 22-26 illustrate the process used to digitize a microscopic image for quantitative analysis. The image in figure 22 is picked up by a television camera attached to a microscope, generated through a real time video processor and a computer. Figure 23 is generated by adjusting the real time video processor brightness and contrast. This expands the density range of the logarithmic intensity function. Figure 24 shows the same section after subtracting density variances contributed by the light source optics of the microscope, camera and glass slide. Figure 25 was generated by readjusting the real time video processor so that the background has a value equal to 0.00 and the densest areas have a value of
1.00. Figure 26 is a digitization of the previous image.
These are black and white photographs of random [Figs.
22-25] and density-coded [Fig. 26] pseudo-color images.
Bar= 300um. >* «. vJ V'V Jr
sb££&« 131
FIGURE 27-30
Figures 27-30 show some aspects of the normal distribution of 10th root primary afferents. Figure 27 shows the elements labeled after HRP application to the
4 * V i 10 spinal nerve. The motoneurons with axons in the
10th nerve innervating the distal hindlimb musculature, are labeled in the lateral motor column [LMC]. Distal motoneuron dendrites coalesce along the pia to form the subpial plexus [small arrows]. The afferents distribute to the dorsal terminal field [DTF] and the ventral terminal field [VTF]. The areas shown in Figs 28, 29 and
30 are indicated. [cc= central canal] Bar= lOOum.
Figure 28 is a higher power micrograph of a portion of the DTF. The arrow indicates a large fiber passing through the DTF en route to the VTF. Bar= 25um. Figure
29 is a high power micrograph of part of the VTF showing a dense network of primary afferent terminations [bv= blood vessel]. Bar= 2Sum. Figure 30 shows the midline with some axonal elements passing into the contralateral dorsal horn [arrows]. Bar= Sum. ' " ‘ i 133
FIGURE 31
This figure shows the rostro-caudal distribution of 4*1^ * 10 1 root primary afferents at thoracic [A], upper lumbar
[B], 10th segment [C], sacral [D], and caudal [E] spinal
levels. The sections shown are not arranged to indicate
the relative distance of the segments from one another.
A higher power micrograph of section A can be seen in
figure 32. Bar= lOOum. E 135
FIGURES 32-34
Figures 32 & 33 demonstrate that axons passing through
the DTF to the VTF elaborate en passant swellings within
the DTF [arrows] at thoracic [figure 32; Bar= 50um] and
lumbar [figure 33; Bar= 50um] levels. Figure 34 is a micrograph taken of a section 120um caudal to a DC lesion
site 5 weeks after the injury. At the lesion site no
fibers can be seen. This shows primary afferent fibers within regions of gliosis. Bar= 50um.
137
FIGURE 35
Figure 35 illustrates the range of gliosis at lesion sites consequent to DC injury. The smallest lesion [dark stippling], an intermediate lesion and the largest lesion
[light stippling] are indicated.
139
FIGURE 36-38
Figures 36-38 are taken from one case with a lesion near the 10tl1 root e n t r y zone and surviving for 13 weeks.
Figure 36 is a micrograph of a section through the root entry zone [cc= central canal]. Bar= 50um. Figure 37 is a higher power micrograph showing an enlargement at the end of an axonal element in the root. Bar= 25um. Figure
38 is from a section adjacent to that in figure 36 [the top of the micrograph is dorsal and the right edge is medial]. The small arrows point to a subpial plexus.
The open block arrowhead points to an enlarged bouton.
Hypertrophic fibers and elements oriented in hairpin turns [arrow] are in evidence and elements [circled] reminiscent of growth cones in the gray matter can also be seen. Bar= 50um.
141
FIGURE 39
Figure 39 shows the mean values and standard errors of the mean for the detected area of the ventral terminal field [VTF] and the dorsal terminal field [DTF] expressed as a % of the hemicord area at 1 week, 2 weeks, 5 weeks,
7 weeks, 10 weeks and 13 weeks after thoracic DC injury.
The open triangle and square are the values for the normal group. DETECTED AREA
Vtf
d tf
0 5 10 weeks post-lesion 143
FIGURES 40-41
Figures 40-41 show the mean optical density [+/- the standard error of the mean] of the dorsal and ventral terminal fields at the 10th segment in normal frogs [open square and triangle] and at 1, 2, 5, 7, 10 and 13 weeks after DC injury. The density units represent the average of the summed pixel densities per section for each terminal field, x 104. density units d > density uni tn o j 5 10 5 0 ------5 1 ------OPTICAL DENSITY DORSAL TERMINAL FIELD ek post-lesion weeks OPTICAL DENSITY VENTRAL TERMINAL lrIFLO ek post-lesion weeks 10 - --
145
FIGURES 42-43
Figures 42 and 43 show that, within the sample areas examined, varicosity number and size distribution are not significantly altered by DC lesions. Figure 42 shows the results from varicosity counts at 3 days, 5 weeks and 13 weeks as compared to normal counts [open squares].
Figure 43 displays the cumulative frequency distribution of varicosity sizes [i.e. the percentage of varicosities of sizes ranging from 0.1-1.0um= 1; l.l~2.0um= 2;...9.1-
10.0um= 10] [square= normal; triangle= 3 days; diamond= 5 weeks; circle= 13 weeks]. % VARICOSITIES FREQUENCY [X 10] 20 25 15 10 5 0 0 0 ; 5 6 8 6 4 NUMBER OF VARICOSITIES OF NUMBER WEEKS POST-LESION WEEKS UUAIE FREQUENCY CUMULATIVE ITIUIN [V SIZE] DISTRIBUTION AREA BINS AREA
147
FIGURES 44-47
These figures show examples of crossing fibers in a normal animal at the 1 0 ^ segment [figure 44], 3 days after a lesion [figure 45], 5 weeks after a lesion
[figure 46] and 13 weeks after a lesion [figure 47],
Bar= 50um. \
/ ' / 1 J ' , / , t « f » ; “ ; W 1 t / 0 ' - ~ • r \ r ta - ’ — — " "V-'' \ t 05 i t 1 / «; '< \ © ------149 FIGURE 48 Figure 48 shows the sum of the lengths of contralateral elements at the lo^*1 segment in each group [normal, 3 days, 5 weeks and 13 weeks] divided by the number of subjects in each group [the lengths are 102 times larger than shown]. CROSSING FIBER LENGTH 10 TIME [WEEKS POST-LESION] 151 FIGURES 49-50 Figure 49 shows the number of fibers counted in each of three regions of the contralateral gray matter in normal [0 week post-lesion], 3 day, 5 and 13 week post lesion subjects. Standard error bars are shown. [M=square=medial; I=diamond=intermediate; L=triangle=lateral]. Figure 50 is a graph showing the number of branch points of contralateral fibers in an animals 10th segment contralateral gray matter, summed for all subjects in each group [normal= 0; 3 day; 5 week and 13 week] and divided by the number of subjects in each group. FIBER FREQUENCY M-L 60 >- 50 CJ 40 CD IXJ az LU 30 20 10 0 0 5 10 TIME [WEEKS POST-LESION] NUMBER OF BRANCH POINTS 80 70 >— C_) 60 CD CD 50 LU GC 40 30 20 10 0 0 5 10 TIME [WEEKS POST-LESION] 153 FIGURES 51-53 Figures 51-53 are to illustrate the normal distribution of the 1 0 ^ dorsal root [figure 51], relevant descending supraspinal systems [figure 52] and the dorsal roots 6-9 [figure 53]. [Ra= raphespinal projections? Rn= rubrospinal projections]. See the text for further information. 155 FIGURES 54-55 Figures 54 and 55 show HRP labeled elements at the i a U 10tn segment after HRP application to a large thoracic lesion. The arrows in figure 54 indicate neurons filled with HRP reaction product. Bar= lOOum. Figure 55 shows a higher power micrograph of the dorsal midline. Indicated [arrows] is one of the axonal elements arising from the lateral funiculus and passing into the contralateral dorsal gray matter. Bar= 30um. S i - A v V V CC 157 FIGURES 56-58 Figures 56-58 illustrate the surgical procedure used to label neuronal elements in the sacral spinal cord. Figure 56 shows a dorsal incision with an arrow pointing to the lo*'*1 spinal nerve. Figure 57 is a higher power photograph showing the 10th spinal nerve [10 Sp.n.] and the putative pelvic [PE] and pudendal [PU] nerves. Figure 58 shows extirpated pelvic viscerae including the terminal end of the colon [C], the bladder [B] and cloaca [the arrowhead indicates the location of the cloacal sphincter]. 159 FIGURES 59-61 Figures 59-61 show neuronal elements in the caudal (sacral) cord labeled after HRP application to the 10th spinal nerve. Figure 59 shows the relationship of these elements to one another, the open block arrow indicates a ventral cluster of neurons. The large lateral neuron surrounded by the rectangle labeled '607 has a lateral dendrite which extends into the lateral funiculus. Bar= 50um. Figure 60 is a higher power micrograph showing the large lateral neuron with an example of an HRP labeled primary afferent closely approximated to it [arrows]. Out of focus and dorsal to the large cell is a labeled cell whose ventral dendrite wraps incompletely around and is close to the proximal part of the large cell's dendrite. Bar= 25um. Figure 61 shows a labeled neuron with many closely approximated varicosities of dorsal root origin [arrows], Bar= 25um. 161 FIGURE 62 Figure 62 is an illustration showing sacral neurons labeled after HRP application to the 1 0 ^ spinal nerve. The rostral section shows a large lateral neurons with two lateral dendrites extending into the lateral funiculus to the subpial plexus and an apparent axon which can be traced exiting the ventral root [small arrows]. Shown in all three sections [asterisks] is a cluster of small neurons in the ventral horn, a group of small neurons in the lateral gray [block arrowhead] and a smattering of neurons in the intermediate lateral field [open block arrowheads]. Bar= lOOum. 163 FIGURE 63 Figure 63 is a stereo-pair showing the distribution of cell bodies labeled after HRP application to the 1 0 ^ spinal nerve, excluding those of the LMC. BIBLIOGRAPHY Adams, J. C. (1977) Technical considerations on the use of horseradish peroxidase as a neuronal marker. Neurosci. 2, 141-145. Adams, J. H., Daniel, P. M. and Prichard, M. M. L. (1968) Degeneration and regeneration of hypothalamic nerve fibres in the neurohypophysis after pituitary stalk section in the ferret. J. Comp. Neurol. 135, 121-144. Adanina, V. O. and Shapovalov, A. I. (1983) Mixed synapses of primary afferent fibers on the motoneurons of the frog, indentified with the help of neuroplasmatic transport of horseradish peroxidase and the electron microscope. Dokl. Akad. Nauk SSSR 269, 213-2 33. Adli, D. S. H., Rosenthal, B. M., Yuen, G. L., Ho, R. H. and Cruce, W. L. R. (1988) Immunohistochemical localization of substance-P, somatostatin, enkephalin and Serotonin in the spinal cord of the northern leopard frog, Rana pipiens. J. Comp. Neurol. 275, 106-116. Antal, M., Tornai, I. and Szekely, G. (1980) Longitudinal extent of dorsal root fibres in the spinal cord and brain stem of the frog. Neurosci. 5, 1311-1322. Ariens-Kappers, C. V., Huber, G. C. and Crosby, E. C. (1960) The comparative anatomy of the nervous system of vertebrates, including man, Vol. I-III. Hafner (New York). Athias, M. (1897) Structure histologique de la moelle epiniere du tetard de la grenouille (rana temporairia). Bibl.Anat. 5, 58-89. Banker, G. A. (1980) Trophic interactions between astroglial cells and hippocampal neurons in culture. Science, 209, 809. 165 166 Barber, P. C. (1981) Regeneration of vomeronasal nerves into the main olfactory bulb in the mouse. Brain Res. 216, 239-251. Barnes, C. D. and Worall, N. (19 68) Reinnervation of spinal cord by cholinergic neurons. J. Neurophys. 31, 680-694. Beattie, M. S., Bresnahan, J. and King, J. S. (1978) Ultrastructural identification of dorsal root primary afferent terminals after anterograde injury filling with horseradish peroxidase. Brain Res. 153, 127-134. Beattie, M. S., Bresnahan, J. C. and King, J. S. (1979) Light and electron microscopic observations of dorsal root terminations in the marginal and gelatinous layers of the dorsal horn of the cat after anterograde injury filling with horseradish peroxidase. In: Advances in pain research and therapy, Eds. Bonica, J., Liebeskind, J., and Albe-Fessard, D., New York, Raven Press, 3: 845- 851. Beattie, M. S., Bresnahan, J. (1982) Neuronal plasticity: Implications for spinal trauma, In: Head injury, basic and clinical aspects (eds. R. G. Grossman & P. L. Gildenberg) Raven Press, New York, 57-68. Beattie, M. S., Stokes, B. T. and Bresnahan, J. (1988) Experimental spinal cord injury. Strategies for acute and chronic intervention based on anatomic, physiological and behavioral studies. In: Pharmacological approaches to the treatment of brain and spinal cord injury (eds. D.G. Stein & B. A. Sabel) Plenum, 43-74. Beck, E., Daniel, P. M. and Prichard M. M. L. (1969) Regeneration of hypothalamic nerve fibers in the goat. Neuroendocrinology 5, 161-182. Becker, D. P. and Povlishock, J. T.(eds.) (1985) Central nervous system trauma status report. Berger, P. J. and Burnstock, G. (1979) Autonomic nervous system In: Biology of the Reptilia, Vol. 10, Neurology B, (Eds. Gans, C., Northcutt, R. G. and Ulinski, P.) Academic Press (London) 1-57. Bernstein, D. R. and Stelzner, D. I. (1983) Plasticity of the corticospinal tract following midthoracic spinal 167 injury in the postnatal rat. J. Comp. Neurol. 221, 382- 400. Bernstein, J. J. and Bernstein, M. E. (1969) Ultrastructure of normal regeneration and loss of regenerative capacity following teflon blockage in goldfish spinal cord. Exp. Neurol. 24, 537-557. Bernstein, J. J. and Gelderd, J. B. (1970) Regeneration of long spinal tracts in goldfish. Brain Res. 20, 33-38. Berry, M. (1982) Post-injury myelin-breakdown products inhibit axonal growth: An hypothesis to explain the failure of axonal regeneration in the mammalian central nervous system. Biblthca Anat. 23, 1-11. Bjorklund, A. and Stenevi, U. (1979) Regeneration of monoaminergic and cholinergic neurons in the mammalian central nervous system. Physiol. Rev. 59, 62-100. Bjorklund, A. (1982) Regeneration of central monaminergic and cholinergic connections as revealed in experiments with intracerebral neural transplants. In: Growth and Regeneration of Axons in the Nervous System 23, 93-94. Blight, A. R. (1978) Golgi staining of "primary" and "secondary" motoneurons in the developing spinal cord of an amphibian. J. Comp. Neurol. 1180, 679-690. Blight, A. R. and Precht, W. (1981) Electrical transmission between primary afferents and motoneurons related to function. In: Regulatory functions of the CNS. Principles of motion and organization, (eds. J. Szentagothai, M. Palkovits and V. Hamori) Pergamon Press, 117-120. Borgens, R. B., Cohan, M. J., and Jaffe (1980) Large and persistent electrical currents enter the transected lamprey spinal cord. Proc. Natl. Acad. Sci. USA 77 1209-1213. Borgens, R. B. Roederer, E. and Cohen, M. J. (1981) Enhanced spinal cord regeneration in lamprey by applied electric fields. Science 213 611-617. Borgens, R. B., Blight, A. R., Murphy, D. J. and Stewart, L. (1986) Transected dorsal column axons within the guinea pig spinal cord regenerate in the presence of an applied electric field. J. Comp. Neurol. 250, 168-180. 168 Bregman, B. S. and Cruce, W. L. R. (1980) Normal dendritic morphology of frog spinal motoneurons: A Golgi study. J. Comp. Neurol. 193, 1035-1045. Bregman, B. s. and Goldberger, M. E. (1982) Anatomical plasticity and sparing of function after spinal cord damage in neonatal cats. 217, 553-555. Bregman, B. S. and Golberger, M. E. (1983) Infant lesion effect. III. Anatomical correlates of sparing and recovery of function after spinal cord damage in newborn and adult cats. Dev. Brain Res. 9, 137-154. Brenowitz, G. L., Collins, w. F., III, and Erulkar, S. D. (1983) Dye and electrical coupling between frog motoneurons. Brain Res. 274, 371-375. Bresnahan, J. C. (1978) An electron microscopic analysis of axonal alterations following blunt contusion of the spinal cord of the rhesus monkey (Macaca mulatta), J. Neurol. Sci. 37, 59-82. Brookhart, J. M. and Fadiga, E. (1960) Potential fields initiated during monsynaptic activation of frog motoneurones. J. Physiol. (Lond.) 150, 633-655. Brushart, T. M. (198 0) Alteration in connections between muscle and anterior horn motoneurons after peripheral nerve repair. Science 208, 603-604. Bunge, R. P., Johnson, M. I. and Thuline, D. (1983) Spinal cord reconstruction using cultured embryonic spinal cord strips. In: Spinal Cord Reconstruction (Eds. Kao, C. C., Bunge, R. P. and Reier, P. J.) Raven Press (New York) 341-3 58. Bunt, S. M. and Fill-Moebs, P. (1984) Selection of pathways by regenerating spinal cord fiber tracts. Dev. Brain Res. 16, 307-311. Butler, E. G. and Ward, M. B. (1965) Reconstitution of the spinal cord following ablation in urodele larvae. J. Expl Zool. 160, 47-66. Butler, E. G. and Ward, M. B. (1966) Reconstruction of the spinal cord after ablation in adult Triturus. Dev. Biol. 15, 464-486. 169 Cajal, R. (1909) Histologie du systeme nerveux de 1'homme et des vertebres. 2 vols. Paris: Maloine Madrid: Instituto Ramon y Cajal. Campbell, H. L., Beattie, M. S. and Bresnahan, J. B. (1982) Large circumferential cells of the developing Rana cates-beiana spinal cord are labelled after HRP application to lateral funiculus. Soc. Neurosci. Abstr. 8 , 820. Campbell, H. L., Beattie, M. S. and Bresnahan, J. B. (1984) The response of dorsal root afferent fibers to dorsal funiculus lesions in developing Rana catesbeiana tadpoles. Soc. Neurosci. Abstr. Campbell, H. L., Beattie, M. S. and Bresnahan, J. B. (1987) Circumferential cells of the developing Rana catesbeiana lumbar spinal cord. Anatomy and Embryology, 176, 155-163. Campbell, H. L., Beattie, M. S. and Bresnahan, J. B. (1988a) The response of dorsal root afferent fibers to dorsal funiculus lesions in developing and juvenile Xenopus laevis. Am. Ass. Anat. Abstract Campbell, H. L., Beattie, M. S. and Bresnahan, J. B. (1988b) The response of spared dorsal root afferent collaterals following dorsal funiculus lesions in juvenile Xenopus laevis. Soc. Neurosci. Abstr. Carbonetto, S. and Muller> K. J. (1977) A regenerating neurone in the leech can form an electrical synapse in its severed axon segment. Nature 267, 450-451. Chambers, W. W., Sprague, J. M. and Liu, C. N. (1960) Anatomical organization of the frog and cat spinal cord, dorsal root and propriospinal pathways. Am. J. Med. Sci. 240, 122-125. Chester-Jones, I. D., Bellamy, K.K., Chan, B. K., Follett, I. W., Henderson, J. G., Phillips, J. G. and Snart, R. S. (1972) Biological actions of steroid hormones in nonmammalian vertebrates. In: Steroids in Nonmaramalian Vertebrates, Academic Press (New York) 414- 480. Clarke, J. D. W., Tonge, D. A. and Holder, N. H. K. (1986) Stage-dependent restoration of sensory dorsal columns following spinal cord transection in anuran 170 tadpoles. Proc. R. Soc. Lond. B 227f 67-82. Clemente, C. D. (1964) Regeneration in the vertebrate central nervous system. Int. Rev. Neurobiol. 6, 257-299. Cobbett, P., Yang, Q. Z. and Hatton, G. I. (1987) Incidence of dye coupling among magnocellular paraventricular nucleus neurons in male rats is testosterone dependent. Brain Res. Bull. 18, 365-370. Coggeshall, R. E. and Youngblood C. S. (1983) Recovery from spinal transection in fish: regrowth of axons past the transection. Neurosci. Let. 38, 227-231. Cohen, M. J. and Hall, G. F. (1986) Control of neuron shape during development and regeneration. Neurochem. Path. 5(3), 331-344. Cohen, A. H., Mackler, S. A. and Selzer, M. E. (1988) Behavioral recovery following spinal transection: functional regeneration in the lamprey CNS. Trends Neurosci. 11 (5) 227-231. Collins, W. F., III (1983) Organization of electrical coupling between frog lumbar motoneurons. J. Neurophysiol. 49, 730-744. Constantine-Paton, M. and Law, M. I. (1978) Eye-specific termination bands in tecta of three-eyed frogs. Science 202 (4365), 639-641. Corvaja, N. and Pellegrini, M. (1975) Ultrastructure of dorsal root projections in the toad spinal cord. Arch. Ital. Biol. 113, 122-149. Corvajd, N. and Grofova, I. (1977) The medial region of the toad ventral horn. An electron microscopic study of normal and degenerating terminals of propriospinal and supraspinal origin. Arch. Ital. Biol. 115, 263-293. Corvaja, N., Pellegrini, M. and Buisseret-Delmas, C. (1978) Ultrastructure of supraspinal dorsal root projections in the toads. I. The obex region. Brain Res. 142, 413-424. Cotman, C. W. and Lynch, G. (1976) Reactive synaptogenesis in the adult nervous system. In: neuronal recognition, (ed. S. Barondes) Plenum, New York, 69-108. 171 Cotman, C. W. (ed.) (1979) Neuronal Plasticity. Raven Press, New York. Cotman, C. W. (ed.) (1981) Synaptic plasticity. The Guilford Press, New York. Cotman, C.W., Nieto-Sampedro, M. and Harris, E.W. (1981) Synapse replacement in the nervous system of adult vertebrates. Physiol. Rev. 61, 684-784. Croop, R. S., Snedeker, J. A. and Selzer, M. E. (1988) Probabilities of regeneration among lamprey reticulospinal neurons. Proc. Soc. Neurosci. Abstract 267.10, 656. Cruce, W. L. R. (1974) The anatomical organization of hindlimb motoneurons in the lumbar spinal cord of the frog, Rana catesbiana. J. Comp. Neurol. 153, 59-76. Cruce, W. L. R. (1979) Spinal cord in lizards. In: Biology of the reptilia, Eds. Gans, Northcutt and Ulinski vol. 9 Academic Press (New York). Crutcher, K. A. and Collin, F. (1982) In vitro evidence for two distinct hippocampal growth factors: basis of neuronal plasticity? Science 217, 67-68. Crutcher, K. A. (1987) Sympathetic sprouting in the central nervous system: a model for studies of axonal growth in the mature mammalian brain. Brain Res. Rev. 12, 203-233. Dalsgaard, C. J. and Elfvin, L. G. (1981) The distribution of the sympathetic neurons in the cat spinal cord projecting to the stellate ganglion. J. Comp. Neurol. 185, 23-30. D'Ascanio, P. and Corvaja, N. (1981) Spinal projections from the rhombencephalon in the toad. Archs. Ital. Biol. 119, 139-150. Davis, J. N. and Haring J. H. (1983) Evidence for pruning of locus coeruleus neurons in the rat hippocampus. Ann. Neurol. 14, 263. Davis, J. N. (1985) Neuronal rearrangements after brain injury: A proposed classification. In: Central Nervous System Trauma. Status Report (Eds. Becker, D. P. and Povlishock, J. T.) National Institute of Neurological and 172 Communicative Disorders and Stroke, National Institues of Health. Ch. 35, 491-501 Davis, G. R. and Farel, P. B. (1985) Quantitative analysis of the development of descending projections in the bullfrog. Proc. Soc. Neurosci. Abstract 178.6, 604 Deuchar, E. M. (1975) Xenopus: The South African clawed frog. Wiley & Sons (London). Devor, M. and Wall, P. D. (1981) Plasticity in the spinal cord sensory map following peripheral nerve injury in rats. J. Neurosci. 1, 676. Duellman, W. E. and Trueb, L. (1986) Biology of Amphibians. McGraw-Hill Book Co. (New York) 1-670. Duffy, M. T., Davis, B. M. and Simpson, S. B. (1988) Axons in regenerated lizard tail are primarily of local origin. Proc. Soc. Neurosci. Abstract 267.11, 656. Ebbesson, S. O. E. (1969) Brain stem afferents from spinal cord in a sample of reptilian and amphibian species. Ann. N.Y. Acad. Sci. 167, 80-102. Ebbesson, S. 0. E. (1976) Morphology of the spinal cord, In: Frog Neurobiology Eds. R. Llinas and W. Precht, Springer-Verlag, 679-703. Edwards, J. S. and Palka, J. (1976) Neural generation and regeneration in insects. In: Simpler Networks and Behavior, (Ed. Fentress, J. c.) 167-185. Egar, M. and Singer, M. (1972) The role of ependyma in spinal regeneration in the urodele Triturus. Exp. Neurol. 37, 422-430. Egar, M., Simpson, S. B., and Singer, M. (1972) The growth and differentiation of the regenerating spinal cord of the lizard, Anolis carolinensis. J. Morph. 131, 1311-152. Erulkar, S. D. and Soller, R. W. (1980) Interactions among lumbar motoneurons on opposite sides of the frog spinal cord: morphological and electrophysiological studies. J. Comp. Neurol. 192, 473-488. Farel, P. (1986) Specificity of neuromuscular connections during early development and following 173 regeneration of motor axons in the bullfrog. Neurochem. Path. 5, 187-203. Finger, s. and Almli, C. R. (1985) Brain damage and neuroplasticity: mechnisms of recovery or development? Brain Res. Rev. 10, 177-186. Foerster, A. P. (1982) spontaneous regeneration of cut axons in adult rat brain. J. Comp. Neurol. 210, 335-356. Forehand, C. J. and Farel, P. B. (1982a) Spinal cord development in anuran larvae: I. primary and secondary neurons. J .Comp.Neurol. 209, 386-394. Forehand, C.J. and Farel, P. B. (1982b) Spinal cord development in anuran larvae: II. ascending and descending pathways. J .Comp.Neurol. 209, 395-408. Forehand, C. J. and Farel, P. B. (1982c) Anatomical and behavioral recovery from the effects of spinal cord transection: dependence on metamorphosis in anuran larvae. J. Neurosci. 2(5), 654-662. Fox, H. (1984) In: Amphibian Morphogenesis. 81-89. Humana Press, New Jersey. Frank, E. and Westerfield, M. (1982a) Synaptic organization of sensory and motor neurones innervating triceps brachii muscles in the bullfrog. J. Physiol. (Lond.) 324, 479-494. Frank, E. and Westerfield, M. (1982b) The formation of appropriate central and peripheral connections by foreign sensory neurones of the bullfrog. J ; Physiol. (Lond.) 324 495-505. Frank, E. and Sah, D.W. (1986) Reformation of specific synaptic connections by regenerating sensory axons in the spinal cord of the bullfrog. Neurochem. Path. 5, 165- 185. Gage, F. H., Bjorklund, A. and Stenevi, U. (1983) Reinnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and noradrenergic afferents. Brain Res. 268, 27-37. Gaze, R. M. and Jacobson, M. (1963) A study of the retinotectal projection during regeneration of the optic 174 nerve in the frog. Proc. R. Soc. Lond. (Biology), 157, 420-448. Gaze, R. M. and Keating, M. J. (1970) Further studies on the restoration of the contralateral retinotectal projection following regeneration of the optic nerve in the frog. Brain Res. 21, 183-195. Gaupp, E. (1896) A. Eckers und R. Wiedersheims Anatomie des frosches. F. Viewig & Sohn, Braunschweig. Golberger, M. E. and Murray M. (1980) Locomotor recovery after deafferentation of one side of the cat's trunk. Exp. Neurol. 67, 103-117. Goldberger, M. E. and Murray, M. (1982) Lack of sprouting and its presence after lesion of the cat spinal cord. Brain Res. 241, 227-239. Goldberger, M. E. and Murray, M. (1983) Recovery of function and anatomical plasticity after damage to the adult and neonatal spinal cord, In: Synaptic plasticity (ed. C. W. Cotman) Guilford Press, New York, 77-110. Grillner, S. (1975) Locomotion in vertebrates-Central mechanisms and reflex interaction. Physiol. Rev. 55, 247- 304. Grinnell, A. D. (1966) A study of the interaction between motoneurons in the frog spinal cord. J. Physiol. (Lond.) 182, 612-648. Grinnell, A. D. (1970) Electrical interaction between antidromically stimulated frog motoneurons and dorsal root afferents: enhancement by gallamine and TEA. J. Physiol. (Lond.) 210, 17-43. Gonzalaz, A. ten Donkelaar, H. J. and de-Boer-van Huizen, R. (1984) Cerebellar connections in Xenopus laevis. An HRP study. Anat. Embryol. 169, 167-176. Grover, B. G. and Grusser-Cornehls, U. (1984) Cerebellar afferents in the frogs Rana esculenta and Rana temporaria. Cell Tiss. Res. 237, 259-267. Guth, L. and Bernstein, J. J. (1961) Selectivity in the re-establishment of synapses in the superior cervical sympathetic ganglion of the cat. Exp. Neurol. 4, 59-69. 175 Hatton, G. I., Yang, Q. Z. and Cobbett, P. (1987) Dye coupling among immunocytochemically identified neurons in the supraoptic nucleus: increased incidence in lactating rats. Neurosci. 21 (3), 923-930. Hefti, F., Dravid, A. and Hartikka, J. (1984) Chronic intraventricular injections of nerve growth factor elevate hippocampal choline acetyltransferase activity in adult rats with partial septo-hippocampal lesions. Brain Res. 293, 305-318. Herrick, C. J. and Coghill, G. E. (1915) The development of reflex mechanisms in amblystoma. J. Comp. Neurol. 25 (1), 65-85. Hibbard, E. (1963) Regeneration in the severed spinal cord of chordate larvae of Petromyzon marinus. Exp. Neurol. 7, 175-185. Hickman, J. L. (1978) Principles of Zoology. Academic Press (New York). Holder, N., Clarke, J. D. W. and Tonge, D. (1987) Pathfinding by dorsal column axons in the spinal cord of the frog tadpole. Development 99, 577-587. Holzer, H. (1951) Reconstruction of the urodele spinal cord following unilateral ablation. J. Exp. Zool. 117, 523-558. Holzer, H. (1956) The inductive activity of the spinal cord in urodele tail regeneration. J. Morphol. 99, 1-39. Hooker, D. (1915) Studies on regeneration in the spinal cord. J. Comp. Neurol. 25, 467-495. Hughes, A. (1957) The development of the primary sensory system in Xenopus Laevis (Daudin). J. Anat. (London) 91, 324-338. Hughes, A. F. W. and Tschumi P. A. (1958) The factors controlling the development of the dorsal root ganglia and ventral horn in Xenopus laevis. J. Anat. 92, 498-527. Hughes, A. (1963) On the labelling of larval neurones by melanin of ovarian origin in certain anura. J.Anat. 7, 217-224. Xnagaki, S., Senba, E., Shiosaka, S., Takagi, H. Kawai, 176 Y., Takatsuki, K. Sakanaka, M.f Matsuzaki, T., and Tohyama, M. (1981) Regional distribution of substance P- like immunoreactivity in the frog brain and spinal cord: Immunohistochemical analysis. J. Comp. Neurol. 201, 243- 254. Jhaveri, S. and Frank, E. (1983) Central projections of the brachial nerve in bullfrogs: Muscle and cutaneous afferents project to different regions of the spinal cord. J. Comp. Neurol. 221 304-312. Jonsson, G. and Hallman, H. (1982) Response of central monoamine neurons following an early neurotoxic lesion. In: Regeneration of Axons in the CNS, Biblthca Anat. 23, 76-92. Joseph, B. S. and Whitlock, D. G. (1968) Central projections of selected spinal dorsal roots in anuran amphibians. Anat. Rec. 160, 279-288. Kahn, J. A. and Roberts, A. (1982) Experiments on the central pattern generator for swimming in amphibian embryos. Phil. Trans. R. Soc. London B. 298, 229-243. Kalil, K. and Reh, T. (1979) Regrowth of severed axons in the neonatal CNS. Science 2 05, 1158-1161. Kao, C. C. (1974) Comparison of healing process in transected spinal cords grafted with autogenous brain tissue, sciatic nerve and nodose ganglion. Exp. Neurol. 44, 197-209. Katzenstein, M. B. and Bohn, R. C. (1984) Regeneration of transected dorsal root ganglion cell axons into the spinal cord in adult frogs (Xenopus laevis). Brain Res. 300, 188-191. Kicliter, E. and Ebbesson, S. O. E. (1976) Organization of the "nonolfactory" telencephalon. In: Frog Neurobiology, (Eds. Llinas, R. and Precht, W.) Springer- Verlag (Berlin) 946-974. Kiernan, J. A. (1971) Pituicytes and the regenerative properties of neurosecretory and other axons in the rat. J. Ant. 109, 97-114. Kokoros, J. J. and Northcutt, R. G. (1977) Telencephalic efferents of the tiger salamander Ambystoma tigrinum tigrinum (Green). J. Comp. Neurol. 173, 613-628. 177 Kollros, J. J. (1981) Transitions in the nervous system during amphibian metamorphosis. In: Metamorphosis a problem in developmental biology. 2nd ed. (Eds. L. I. Gilbert and E. Frieden) 445-458. Plenum Press, New York. Kusuma, A. and ten Dontelaar H. J. (1979) Staining of dorsal root primary afferent fibers by anterograde movement of horseradish peroxidase and retrograde labeling of motoneurons and preganglionic autonomic cells in the turtle spinal cord. Neurosci. Lett. 14, 141-146. Kusuma, A. and ten Donkelaar, H. J. (1980) Propriospinal fibers interconnecting the spinal enlargements in some quadrupedal reptiles. J. Comp. Neurol. 193, 871-891. Lamborghini, J. E. (1980) Rohon-Beard cells and other large neurons in Xenopus embryos originate during gastrulation. J. Comp. Neurol. 189, 323-333. Larsell, o. (1967) The comparative anatomy and histology of the cerebellum from Myxinoids through Birds. University of Minnesota Press (Minneapolis). Lazdr, Gy., Toth, P., Csank, Gy. and Kicliter, E. (1983) Morphology and location of tectal projection neurons in frogs: a study with HRP and cobalt-filling. J. Comp. Neurol. 215, 108-120. Lee, M. T. (1982) Regeneration and functional reconnection of an idendified vertebrate central neuron. J. Neurosci. 2, 1793-1811. Lee, M. T. and Farel, P. B. (1988) Guidance of regenerating motor axons in larval and juvenile Bullfrogs. J. Neurosci. 8(7), 2430-2437. Leedy, M. G., Beattie, M. S. and Bresnahan, J.C. (1987) Testosterone-induced plasticity of synaptic inputs to adult mammalian motoneurons. Brain Res. 424, 386. Leong, S. K . , Tay, S. S. W. and Wong, W. C. (1983) Preganglionic neurons projecting to the first thoracic sympathetic ganglion in the terrapin (Trionyx sinensis), J. Auton. Nerv. Syst. 9, 585-593. Levine, R. L. (1983) Widespread regeneration of central axons through the central nervous system of the goldfish. Dev. Brain Res. 9, 416-419. 178 Lichtman, j. w. and Frank, E. (1981) Projections of individual muscle sensory fibers to homonymous and heteronymous motoneurons in the bullfrog. Soc. Neurosci. Abstr. 7, 362. Lichtman, J. W., Jhaveri, S. and Frank, E. (1984) Anatomical basis of specific connections between sensory axons and motor neurons in the bullfrog's brachial spinal cord. J. Neurosci. 4, 1754-1763. Light, A. R. and Metz, C. B. (1977) The morphology of the spinal cord efferent and afferent neurons contributing to the ventral roots of the cat. J. Comp. Neurol. 179, 501- 516. Light, A. R. and Perl, E. R. (1979) Spinal termination of functionally identified primary afferent neurons with slowly conducting myelinated fibers. J. Comp. Neurol. 186, 133-150. Liuzzi, F. J., Beattie, M. S. and Bresnahan, J. C. (1983) Dorsal root afferents contact migrating motoneurons in the developing frog spinal cord. 262, 299-302. Liuzzi, F. J., Beattie, M. S. and Bresnahan, J. C. (1984) The relationship of dorsal root afferents to motoneuron somata and dendrites in the adult bullfrog: A light and electron microscopic study using horseradish peroxidase. 11 No. 4, 951-961. Liuzzi, F. J. and Lasek, R. J. (1984) Electron microscopy of HRP injury-filled regenerated axons in the adult frog spinal cord. Proc. Soc. Neurosci. Abstract 9, 695. Liuzzi, F. J., Beattie, M. S. and Bresnahan, J. C. (1985) The development of the relationship between dorsal root afferents and motoneurons in the larval bullfrog spinal cord. 14, 377-392. Liuzzi, F. J. and Miller, R. H. (1985) Immunohistochemical characterization of normal and reactive glia in the adult frog spinal cord. Proc. Soc. Neurosci. Abstract 11, 396. Liuzzi, F.J. and Lasek, R. J. (1986a) Dorsal root axonal regeneration in the adult frog spinal cord, a model of vertebrate CNS regeneration. Neurochem. Path. 5, 237- 253. 179 Liuzzi, F. J. and Lasek, R. J. (1986b) Regeneration of motoneuron axons into the adult frog spinal cord after ventral-to-dorsal-root anastamosis. J. Comp. Neurol. 247, 111-122. Longo, F. M. (1984) Neuronotrophic activities in cerebral spinal fluid of head trauma patients. Exp. Neurol. 84, 207. Lopate, 6. L., Beattie, M. S. and Bresnahan, J. C. (1985) Restitution of descending spinal pathways after spinal transection in Xenopus occurs during, but not after metamorphosis. Proc. Soc. Neurosci. Abstract 175.10, 590. Lorente de No, R. (1921) La regeneracion de la medulla espinal en las larvas de batracio. Trab. Lab. Invest. Biol. 19, 147-183. Lorente de No, R. (1924) Etudes sur le cerveau posterieur. III. Sur les connexions extra-cerebelleuses des fascicules afferents au cerveau, et sur la fonction de cet organe. Trav. Lab. Rech. Biol. Univ. Madrid 22, 51-65. Lorez, H. P. and Kemali, M. (1981) Substance P-, Met- Enkephalin- and somatostatin-like immunoreactivity distribution in the frog spinal cord. Neurosci. Lett. 26, 119-124. Mackler, S. A. and Selzer, M. E. (1987) Specificity of synaptic regeneration in the spinal cord of the larval sea lamprey. J. Physiol. 388, 183-198. Manthorpe, M., Nieto-Sampedro, M., Skaper, S. D., Lewis, E. R., Barbin, E. R., Longo, F. M., Cotman, C. W. and Varon, S. (1983) Neuronotrophic activity in brain wounds of the developing rat. Correlation with implant survival in the wound cavity. Brain Res. 267, 47-56. Martin, A. H. and Brinkman, R. (1970) The dorsal horn of the avian spinal cord, a re-examination. Experientia 26, 887-889. Martin, G. F. and Xu, X. M. (1988) Evidence for developmental plasticity of the rubrospinal tract. Studies using the North American opossum. Dev. Brain Res. 39, 303-308. 180 Maruhashi, J., Mizuguchi, K. and Tasaki, I. (1952) Action currents in single afferent nerve fibres elicited by stimulation of the skin of the toad and the cat. J. Physiol. 117, 129-151. Matesz, C. (1979) Central projection of the VUIth cranial nerve in the frog. Neurosci. 4, 2061-2071. Have, G. M., Bresnahan, J. C. and Beattie, M. S. (1983) Ultrastructure of HRP-labeled neurons: A comparison of two sensitive techniques. Brain Res. Bull. 10, 551-558. McConnell, P. and Berry, M. (1982) Regeneration of axons in the mouse retina after injury. Biblthca Anat. 23, 26- 37. McCouch, G. P., Austin, G. M . , Liu, W. and Liu, C. Y. (1958) sprouting as a cause of spasticity. J. Neurophys. 21, 205-216. McQuarrie, I. G. (1986) Stuctural protein transport in elongating motor axons after sciatic nerve crush effects of a conditioning lesion. Neurochem. Pathol. 5 (3) 153- 164. Mendell, L. M. and Hollyday, M. (1976) Spinal reflexes in anurans with and altered periphery. In: Frog Neurobiology, Eds. Llinas, R. and Precht, W. Springer- Verlag (Berlin) 793-810. Mensah, P. L. (1974) The course and distribution of descending fibers of the lateral funiculus of the amphibian spinal cord. Doctoral Thesis Univ. Ca. Irvine. Michel, M. E. and Reier, P. J. (1979) Axonal-ependymal associations during early regeneration of the transected spinal cord in Xenopus laevis tadpoles. J. Neurocytol. 8, 529-548. Miller, R. H. and Liuzzi, F. J. (1986) Regional specialization of radial glial cells of the adult frog spinal cord. J. Neurocytol. 15, 187-196. Motorina, M. V. (1978) Ultrastructural features of synapses of gap juction type in frog spinal motor nuclei. Arkhiv. Anatomii, Gistologii i Embriologii, 75(4), 27-34. Moyer, E. K., Kimmel, D. L. and Winborne, L. W. (1953) Regeneration of sensory spinal nerve roots in young and 181 senile rats. J. Comp. Neurol. 98, 283-307. Murray, M. and Goldberger, M. E. (1974) Restitution of function and collateral sprouting in the cat spinal cord: The partially hemisected animal. J. Comp. Neurol. 158, 19-36. Murray, M. and Goldberger, M. E. (1986) Replacement of synaptic terminals in lamina II and Clarke's nucleus after unilateral dorsal rhizotomy in adult cats. Murray, M. (1986) Reactive Synaptogenesis in the CNS. A Comparison of Regenerating and Sprouting systems. Neurochem. Path. 5, 205-220. Nieto-Sampedro, M., Manthrope, M., Barbin, G., Varon, S. and Cotman, C. W. (1983) Injury-induce neuronotrophic activity in adult rat brain: Correlation with survival of delayed implants in the wound cavity. J. Neurosci. 3, 2219-2229. Nieuwenhuys, R. (1964) Comparative anatomy of the spinal cord. In: Progress in Brain Research, Elsevier 11, 1-57. Nieuwenhuys, R. and Opdam, P. (1976) Structure of the brain stem. In: Frog Neurobiology Eds. Llinas R. and Precht, W . ; Springer-Verlag (Berlin) 811-855. Nieuwkoop, P. D. and Faber, J. (1956) Normal table of xenopus laevis (daudin). A systematical and chronological survey of the development from the fertilized egg till the end of metamorphosis. North- Holland. Nikundiwe, A. M., de Boer-van Huizen, R. and ten Donkelaar, H. J. (1982) Dorsal root projections in the clawed toad (Xenopus laevis) as demonstrated by anterograde labeling with horseradish peroxidase. Neuroscience, 7 (9), 2089-2103. Nikundiwe, A. M. and Nieuwenhuys, R. (1983) The cell masses in the brain stem of the South African clawed frog, Xenopus laevis: A topological analysis. J. Comp. Neurol. 213, 199-219. Nobel, M., Fok-Seang, J. and Cohen, J. (1984) Glia are a Unique Substrate for the in vitro Growth of Central Nervous System Neurons. J.Neurosci. 4, 1892-1903. 182 Noble, G. K. (1931) The biology of the amphibia. McGraw- Hill (New York and London). Nordlander, R. N. and Singer, M. (1978) The role of ependyma in regeneration of the spinal cord in the urodele amphibian tail. J. Comp. Neurol. 180, 349-374. Nordlander, R. N. and Singer, M. (1982) Z. Zellforsch. Mikrosk. Anat. 126, 157. Nordlander, R. H. (1984) Developing descending neurons of the early Xenopus tail spinal cord in the caudal spinal cord of early Xenopus. J. Comp. Neurol. 228, 117-128. Nordlander, R. H., Baden, S. T. and Ryba, T. M. J. (1985) Development of early brainstem projections to the tail spinal cord of Xenopus. J. Comp Neurol. 231, 519-529. Norris, D., Beattie, M. S. and Bresnahan, J. B. (1988) Evidence for sprouting of dorsal root terminal fields after spinal cord hemisection in juvenile Xenopus frogs. Neurosci. Abst. 267.8 Parent, A. (1981) The anatomy of seotonin-containing neurons across phylogeny. In: Serotonin, Neurotransmission and Behaviour (Eds. Jacobs, B. L. and Gelperin, A.) MIT Press (Cambridge) 3-34. Petras, J. M. and Cummings, J. F. (1972) Autonomic neurons in the spinal cord of the rhesus monkey: a correlation of the findings of cytoarchitectonics and sympathectomy with fiber degeneration following dorsal rhizotomy. J. Comp. Neurol. 146,189-218. Piatt, J. (1955) Regeneration in the central nervous system of amphibia. In: Regeneration in the Central Nervous System, (Ed. Windle, W. F.) Charles C. Thomas (Springfield, 111.) 20-46. Piatt, J. and Piatt, M. (1958) Transection of the spinal cord in the adult frog. Anat. Rec. 131, 81-95. Pick, J. (1957) Sympathectomy in amphibians. J. Comp. Neurol. 107, 169-208. Pickel, V. M., Krebs, H. and Bloom, F. E. (1973) Proliferation of norepinephrine-containing axons in the rat cerebellar cortex after peduncle lesions. Brain Res. 59, 169-179. 183 Pickel V. M., Siegal, M. and Krebs, H. (1974) Axonal proliferation following lesions of cerebellar peduncles. A combined fluorescence microscopic and radioautographic study. J. Comp. Neurol. 155, 43-59. Potter, H. D. (1965) Mesencephalic auditory region of the bullfrog. J. Neurophysiol. 228, 1132-1154. Precht, W., Richter, A. Ozawa, S. and Shimazu, H. (1974) Intracellular study of frog's vestibular neurons in relation to the labyrinth and spinal cord. Exp. Brain Res. 19, 377-393. Proshansky, E. and Egger, M. D. (1977) Staining of the dorsal root projection to the cat's dorsal horn by anterograde movement of HRP. Neurosci. Lett. 5, 103-110. Pubolis, L. M. and Sessle, B. J. (1987) Effects of injury on trigeminal and spinal somatosensory systems. Neurology and neurobiology 30, 143-297. Raisman, G. (1973) An ultrastructural study of the effects of hypophysectomy on the supraoptic nucleus of the rat. J. Comp. Neurol. 147, 181-208. Ralston, III, H. J. (1968) Dorsal root projections to dorsal horn neurons in the cat spinal cord. J. Comp. Neurol. 132, 275-302. Reier, P. J., Stensaas, L. J. and Guth, L. (1983) The astrocytic scar as an impediment to regeneration in the central nervous system, In: Spinal Cord Reconstruction, (eds. C. C. Kao, R. P. Bunge and R. J. Reier), 163-195. Rexed, B. (1964) Some aspects of the cytoarchitectonics and synaptology of the spinal cord. In: Progress Brain Research 11, 58-92. Richardson, P. M., Aguayo, A. J. and McGuiness (1983) Role of sheath cells in axonal regeneration, In: Spinal Cord Reconstruction, (eds. C. C. Kao, R. P. Bunge and R. J. Reier) 293-304. Risling, M., Hildebrand, C. and Cullheim, S. (1984) Invasion of the L7 ventral root and spinal pia mater by new axons after sciatic nerve division in kittens. Exp. Neurol. 83, 84-97. 184 Roberts, A. and Clarke, J. D. W. (1982a) The neuroanatomy of an amphibian embryo spinal cord. Phil. Trans. R. Soc. London B 296, 195-212. Roberts, A. and Kahn, J. A. (1982b) Intracellular recordings from spinal neurons during 'swimming' in paralysed amphibian embryos. Phil. Trans. R. Soc. London B. 296, 195-212. Robertson, D. R., (1987) Sympathetic preganglionic neurons in frog spinal cord. J. Autonom. Nerv. Sys. 18, 1-1 1 . Rosenberg, M. B., Grossman, M. H. and Breakefield, X. O. (1984) Brain Res. 295, 35. Rosenthal, B. M. and Cruce, W. L. R. (1984) Ipsilateral and contralateral projections of primary afferents in frog caudal lumbar spinal cord: A light and electron microscopic study. Anat. Rec. 2 08 152A. Rosenthal, B. M. and Cruce, W. L. R. (1985) Distribution and ultrastructure of primary afferent axons in Lissauer's tract in the northern leopard frog (Rana pipiens). Rovainen, C. M. (1976) Regeneration of Muller and Mauthner axons after spinal transection in larval lampreys. J. Comp. Neurol. 168, 545-554. Rubinson, K. (1968) Projections of the tectum opticum of the frog. Brain Behav. Evol. 1, 529-561. Rubinson, K. and Skiles, M. (1973) Efferent projections from the superior olivary nucleus in Rana catesbeiana. Anat. Rec. 175, 431-432. Sah, D. and Frank, E. (1984) Regeneration of sensory- motor synapses in the spinal cord of the bullfrog. J. Neurosci. 4, 2784-2791. Sala y Pons, C. (1892) Estructura de la medulla espinal de los batracios. Trab.Lab.Invest.Histol. 3-22. Santini, D. and Browner, R. H. (1988) Regeneration of descending pathways after spinal cord trauma in the red eared turtle. Proc. Soc. Neurosci. Abstract 267.15, 657. Sasaki, H. and Mannen, H. (1981) Morphological analysis 185 of astrocytes in the bullfrog (Rana catesbeiana) spinal cord with special reference to the site of attachment of their processes. J. Comp. Neurol. 198, 13-35. Schneider, G. E. and Jhaveri, S. R. (1974) Neuroanatomical correlates of spared or altered function after brain lesion in the newborn hamster. In: Plasticity and Recovery of Function in the Central Nervous System. Academic Press (New York) 65-109. Schreyer, D. J. and Jones, E. G. (1983) Growing corticospinal axons bypass lesions of neonatal rat spinal cord. Neurosci. 9, 31-40. Shiriaev, B. I. and Shupliakov, 0. V. (1986) Synaptic organization of dorsal root projections to lumbar motoneurons in the clawed toad (Xenopus laevis) Exp. Brain Res. 63 135-142. Shapovalov, A. I. (1980) Interneuronal synapses with electrical, dual and chemical mode of transmission in vertebrates. Neurosci. 5, 113-1124. Shapovalov, A. I. and Shiriaev, B. I. (1980) Dual mode of junctional transmission at synapses between primary afferent fibres and raotoneurones in the amphibian J. Physiol. (Lond.) 306, 1-15. Sievers, J. and Klemm, H. P. (1982) Locus coeruleus- cerebellum: interaction during development. Biblthca Anat. 23, 56-75. Silver, M. (1942) The motoneurons of the spinal cord of the frog. J.Comp.Neurol. 77:1-39. Silvey, G. E., Gulley, R. L. and Davidoff, R. A. (1974) The frog dorsal column nucleus. Brain Res. 73, 421-437. Simpson, J. I. (1976) Functinal synaptology of the spinal cord. In: Frog Neurobiology, eds. Llinas R. & Precht, W. Springer-Verlag, 728-749. Simpson, S. B. (1970) Studies on regeneration of the lizard's tail. Am. Zool. 10, 157-165 Simpson, S. B. (1983) Fasciculation and guidance by the ependyma. In: Spinal Cord Reconstruction, (Eds. Kao, C. C . , Bunge, R. P. and Reier, P. J.) Raven Press (New York) 151-162. 186 Singer, M., Nordlander, R. H. and Egar, M. (1979) Axonal guidance during embryogenesis and regeneration in the spinal cord of the newt: the blueprint hypothesis of neuronal pathway patterning. J. Comp. Neurol. 185, 1-22. Smith, C. L. and Frank, E. (1988) Specificity of sensory projections to the spinal cord during development in bullfrogs. J. Comp. Neurol. 269, 96-108. Soller, R. W. (1977) Monoaminergic inputs in frog motoneurons: An anatomical study using fluorescence histochemical and silver degeneration studies. Brain Res. 122, 445-458. Sonnhof, U., Richter, D. W. and Taugner, R. (1977) Electrotonic coupling between frog spinal motoneurons. An electrophysiological and morphological study. Brain Res. 138, 197-215. Sotelo, C. and Taxi, M. (1970) Ultrastructural aspects of electotonic junctions in the spinal cord of the frog. Brain Res. 17, 137-141. Sotelo, C. and Grofova, I. (1976) Ultrastructural features of the spinal cord. In: Frog Neurobiology eds. R. Llinaas and W. Precht, Springer-Verlag Berlin, 707- 727. Spallanzani, L. (1768) Prodromo di un opera da impremersi sopra le riproduzioni animale dato urn luce. G, Montanari, Modena. Sperry, R. W. (1944) Optic nerve regeneration with return of vision in anurans. J. Neurophysiol. 7, 57-69. Sperry, R. W. (1951) Mechanisms of neural maturation. In: Handbook of experimental psychology, Ed. Stevens, S. s. Wiley (New York) 263-280. Stelzner, D. J. (1985) Axonal regeneration in the amphibian visual system: similarities and differences with normal development. In: Neural Development, Plasticity & Regeneration: A Discussion of Common Mechanisms. Stelzner, D. J., Weber, E. D., and BryzGornia, W. F. (1986) Sparing of function in developing spinal cord: anatomical substrate. In: Development and Plasticity of 187 the Mammalian Spinal Cord M. E. Goldberger, A. Gorio, and M. Murray (eds.) Liviana Press, Springer Verlag, Fidia Research Series 3f 81-99. Stensaas, L. J. and Stensaas, S. S. (1971) Light and electron microscopy of motoneurons and neuropile in the amphibian spinal cord. Brain Res. 31, 67-84. Stehouwer, D. J. (1986) Behavior of larval and juvenile Bullfrogs (Rana catesbeiana) following chronic spinal transection. Behavioral and Neural Biol. 45, 120-134. Stensaas, L. J. (1983) Regeneration in the spinal cord of the newt Notopthalmus (triturus) pyrrhogaster. In: Spinal Cord Reconstruction (Eds. Kao, C. C., Bunge, R. P. and Reier, P. J.) Raven Press (New York) 121-149. Steward, 0. (1982) Assessing the functional significance of lesion-induced neuronal plasticity. Int. Rev. Neurobiol. 23, 197-254. Sullivan, K., Conway, K. M. and Hunt, R. K. (1984) Demonstration of a polarizing signal that reverses future retinotectal patterns across nucleopore filter barriers in Xenopus embryonic eye. Cell Diff. 14, 33-45. Svendgaard, N. A., Bjorklund, A. and Stenevi, U. (1976) Regeneration of central cholinergic neurons in the adult brain. Brain Res. 102, 1-22. Szekely, G. (1976a) The morphology of motoneurons and dorsal root fibers in the frog's spinal cord. Brain Res. 130, 275-290. Szekely, G. and Kosaras, B. (1976b) Dendro-dendritic contacts between frog motoneurons shown with the cobalt labeling technique. Brain Res. 108, 194-198. Szekely, G. and Czeh, G. (1976c) Organization of locomotion. In: Frog Neurobiology, Eds. Llinas, R. and Precht, W. Springer-Verlag (Berlin) 765-792. Szekely, G., Matesz, K., Baker, R. E., and Antal, M. (1982) The termination of cutaneous nerves in the dorsal horn of the spinal cord in normal and in skin-rotated frogs. Exp. Brain Res. 45, 19-28. Tassava, R. A. and Olsen, C. L. (1982) Higher vertebrates do not regenerate digits and legs because the wound 188 epidermis is not functional. A hypothesis. Differentiation 22, 151-155. Taxi, J. (1976) Morphology of the autonomic nervous system. In: Frog Neurobiology, (Eds. Llinas, R. and Precht, W.) Springer-Verlag (Heidelberg) 93-150. Taylor, A. C. and Kollros, J. J. (1946) Stages in the normal development of Rana pipiens larvae. Anat.Rec., 94, 7-23. ten Donkelaar, H. J., de Boer-van Huizen, R., Schouten, F. T. M. and Eggen, S. J. (1981) Cells of origin of descending pathways to the spinal cord in the clawed toad (Xenopus laevis). Neurosci. 6(11), 2297-3112. ten Donkelaar, H. J. and de Boer-van Huizen, R. (1982) Observations on the development of descending pathways from the brain stem to the spinal cord in the clawed toad, Xenopus laevis. Anat. Embryol. 163, 461-473. ten Donkelaar, H. J. (1982) Organization of descending pathways to the spinal cord in Amphibians and Reptiles. In: Progress in Brain Research, (Eds. Kuypers, H. G. J. M. and Martin, G. F.) Elsevier (Amsterdam) 57, 25-67. Tessler, A., Himes, B. T., Soper, K., Murray, M. Goldberger, M. E. and Reichlin, S. (1984) Recovery of substance P but not somatostatin in the cat spinal cord after unilateral lumbosacral dorsal rhizotomy: a qualitative study. Brain Res. 305, 95-102. Thors, F., de Kort, E. J. M. and Nieuwenhuys, R. (1982) On the development of the spinal cord of the clawed frog, Xenopus laevis. II. Experimental analysis of differentiation and migration. Anat. Embryol. 164, 443- 454. Tolbert, D. L. and ti Der, T. (1987) Redirected growth of pyramidal tract axons following neonatal pyramidotomy in cats. J. Comp. Neurol. 260, 299-311. Toth, P., Csank, Gy. and Laz&r, Gy. (1985) Morphology of the cells of origin of descending pathways to the spinal cord in Rana esculenta. A tracing study using cobaltic- lysine complex. J. Hirnforsch 4, 365-383. Ueda, S., Nojyo, Y. and Sano, Y. (1984) Immunohistochemical demonstrations of the serotonin 189 neuron system in the central nervous system of the bullfrog, Rana catesbeiana. Anat. Embryol. 169, 219-229. Van Gehuchten, A. (1895) La moelle epiniere de la truite (Trutta fario). Cellule 11, 111-174. Van Gehuchten, A. (1898) La moelle epiniere des larvaes des Batraciens (Salamandra maculosa) Archs Biol., Paris 15, 599-619. Van Mier, P. L. L. (1986) The development of the motor system in the clawed toad, Xenopus laevis. Doctoral Dissertaion, University of Nijmegen. Varon, S. O., Manthorpe, M. and Williams, L. R. (1983) Neuronotrophic and neurite-promoting factors and their clinical potentials. Dev. Neurosci. 6, 73-100. Wall, J. T., Felleman, D. J. and Kaas, J. H. (1983) Recovery of normal topography in the somatosensory cortex of monkeys after nerve crush and regeneration. Science 221, 771. Westerfield, M. and Frank, E. (1982) Specificity of electrical coupling among neurons innervating forelimb muscles of the adult bullfrog. J. Neurophysiol. 48, 904- 913. Westerfield, M. and Powell, S.L. (1983) Selective Reinnervation of Limb Muscles by Regenerating Frog Motor Axons. Dev. Brain Res. 10, 301-304. Whitehouse, R. H. and Grove, A. J. (1947) Dissection of the frog. University Tutorial Press, Cambridge. Wilhelm, G. B. and Coggeshal, R. E. (1981) An electon microscopic analysis of the dorsal root in the frog. J. Comp. Neurol. 196, 421-429. Wood, M. R. and Cohen, M. J. (1981) Synaptic regeneration and glial reactions in the transected spinal cord of the lamprey. J.Neurocytol. 10, 57-79. Woodburne, R. T. (1939) Certain phylogenetic anatomical relations of localizing significance for the mammalian central nervous system. J. Comp. Neurol. 71, 215-257. Yin, H. S. and Selzer, M. E. (1983) Axonal regeneration in lamprey spinal cord. J. Neurosci. 3(6), 1135-1144. 190 Yin, H. S. and Selzer, M. E. (1984) Electrophysiological evidence of regeneration of lamprey spinal neurons. Exp. Neurol. 83, 618-628. Zakon, H. and Capranica, R. R. (1981) An anatomical and physiological study of regeneration of the eighth nerve in the leopard frog. Brain Res. 209, 325-338.