INFORMATION TO USERS

This reproduction was made from a copy of a document sent to us for microfilming. While the most advanced technology has been used to photograph and reproduce this document, the quality of the reproduction is heavily dependent upon the quality of the material submitted.

The following explanation of techniques is provided to help clarify markings or notations which may appear on this reproduction.

1.The sign or “ target” for pages apparently lacking from the document photographed is “ Missing Page(s)” . I f it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure complete continuity.

2. When an image on the film is obliterated with a round black mark, it is an indication of either blurred copy because of movement during exposure, duplicate copy, or copyrighted materials that should not have been filmed. For blurred pages, a good image of the page can be found in the adjacent frame. If copyrighted materials were deleted, a target note will appear listing the pages in the adjacent frame.

3. When a map, drawing or chart, etc., is part of the material being photographed, a definite method of “sectioning” the material has been followed. It is customary to begin film ing at the upper left hand corner o f a large sheet and to continue from left to right in equal sections with small overlaps. I f necessary, sectioning is continued again—beginning below the first row and continuing on until complete.

4. For illustrations that cannot be satisfactorily reproduced by xerographic means, photographic prints can be purchased at additional cost and inserted into your xerographic copy. These prints are available upon request from the Dissertations Customer Services Department.

5. Some pages in any document may have indistinct print. In all cases the best available copy has been filmed.

University M icrofilm s International 300 N. Zeeb Road Ann Arbor, Ml 48106 8305355

Liuzzi, Francis Joseph

THE DEVELOPMENT OF DORSAL ROOT AFFERENTS AND THE LATERAL MOTOR COLUMN IN THE BULLFROG LUMBAR ENLARGEMENT AS SHOWN BY HRP INJURY FILLING OF DORSAL AND VENTRAL ROOTS

The Ohio State University PH.D. 1982

University Microfilms

International 300 N. Zeeb Road, Ann Arbor, M I 48106 PLEASE NOTE:

In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a checkV mark .

1. Glossy photographs or pages

2. Colored illustrations, paper or print____

3. Photographs with dark background ^

4. Illustrations are poor copy______

5. Pages with black marks, not original copy__

6. Print shows through as there is text on both sides of page______

7. Indistinct, broken or small print on several pages___

8. Print exceeds margin requirements_____

9. Tightly bound copy with print lost in spine______

10. Computer printout pages with indistinct print______

11. Page(s)______lacking when material received, and not available from school or author.

12. Page(s)______seem to be missing in numbering only as text follows.

13. Two pages numbered______. Text follows.

14. Curling and wrinkled pages______

15. Other______

University Microfilms International THE DEVELOPMENT OF DORSAL ROOT AFFERENTS AND THE LATERAL

MOTOR COLUMN IN THE BULLFROG LUMBAR ENLARGEMENT AS SHOWN

BY HRP INJURY FILLING OF DORSAL AND VENTRAL ROOTS.

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of the Ohio State University

By

Francis J. Liuzzi, B.S., M.S.

*****

The Ohio State University

1982

Reading Committee: Approved by

Michael S. Beattie, Ph.D. Co-Adviser

Jacqueline C. Bresnahan, Ph.D. Co-Adviser

George F. Martin, Ph.D.

James S. King, Ph.D. Adviser

Adviser

Department of Anato For Rosemary, Cisco and Guy

ii ACKNOWLEDGMENTS

I wish to thank Dr. Michael S. Beattie and Dr. Jacqueline C.

Bresnahan for allowing me to be a part of their laboratory and for their encouragement throughout my graduate career.

I am indebted to Tina Hatten Van Meter for her technical expertise and assistance.

Karl Rubin not only provided technical advice but close friendship that made my tenure at Ohio State more pleasant and at times fun.

I would also like to thank all of the faculty of the Department of Anatomy, especially Dr. James S. King and Dr. George F. Martin, for their knowledge, advice and encouragement.

Finally, I want to thank my wife Rosemary and my sons Cisco and

Guy for their love and patience. This degree is as much theirs as it is my own. VITA

NAME: Francis Joseph Liuzzi

DATE OF BIRTH: October 4, 1949

PLACE OF BIRTH: Albany, New York

EDUCATION: Grade School and High School P.S. 25, Albany, N.Y. 1954-1958 Saddlewood Elementary School, Colonie, N.Y. 1959-1961 Lisha Kill Junior High School, Colonie, N.Y. 1962-1963 Colonie Cenfral High School, Colonie, N.Y. 1964-1967

College and Degrees Union College, Schenectady, N.Y. B.S. Psychology 1971 Rensselaer Polytechnic Institute, Troy, N.Y. M.S. Psychology 1975 Ohio State University, Columbus, Ohio PhD. Anatomy 1982

TEACHING EXPERIENCE: Introductory Psychology, 1972, 1973 Physiological Psychology Labora­ tory, 1973 Electron Microscopy for Histo- technologists, 1975-1979 Gross Anatomy for Dental Students, 1980 Neuroanatomy for Dental Students, 1981 Neuroanatomy for Medical Students, 1981

iv RESEARCH TRAINING: Synaptogenesis of axosomatic synapse of primary afferents on motor neurons in the ventral horn of ranid tadpoles as shown by HRP injury filling. Comparative organization of primary afferents as shown by HRP root labelling.

SCIENTIFIC SOCIETIES: Electron Microscopic Society of America Society for Neuroscience

HONORS: Presidential Fellowship - The Ohio State University - 1981-1982

ABSTRACTS AND PUBLICATIONS:

Beattie, M.S., Bresnahan, J.C. , Ho, R.H. and Liuzzi, F. , 1979 "Distribution of met-enkephalin, substance-P and soma­ tostatin in the spinal cord dorsal horn after hemisection in the rat." Neurosci. Absts., 5, 719.

Beattie, M.S., Bresnahan, J.C. and Liuzzi, F. . 1980, "Ultra­ structure of dorsal root afferents to the ventral horn of the cat." Neurosci. Absts., 6 , 435.

Beattie, M.S., Bresnahan, J.C. and Liuzzi, F. . 1981, "Aberrant dorsal root axons in the dorsal horn after chronic spinal hemisection in the cat." Pain, Abstracts of the Third World Congress on Pain, Suppl. 11-319, S239.

Beattie, M.S., Bresnahan, J.C. and Liuzzi, F., 1981, "Aberrant axons from the dorsal roots after chronic spinal hemi­ section in the cat." Neurosci. Absts., in press.

Liuzzi, F.J., Beattie, M.S. and Bresnahan, J.C. , "Development of the relationship between dorsal root afferents and the lateral motor column of the bullfrog spinal cord," Neurosci. Absts., in press.

Beattie, M.S., Liuzzi, F.J., and Bresnahan, J.C., "Migrating motoneurons and their relationship to developing dorsal root afferents in the bullfrog tadpole spinal cord," Neurosci. Absts. in press.

Campbell, H.L., Liuzzi, F.J., Beattie, M.S., and Bresnahan, J.C., "Large circumferential cells of the developing Rana catesbeiana spinal cord are labelled after HRP application to lateral funiculus, " Neurosci. Absts., in press.

v Liuzzi, Francis J., Beattie, Michael S. and Bresnahan, Jacqueline C., 1982, "Development of the monosynaptic reflex arc in the lumbar spinal cord of Rana catesbeiana as shown by simultaneous labelling of dorsal and ventral roots with HRP," Anat. Rec.. 202, 113A.

Liuzzi, F.J., Beattie, M.S. and Bresnahan, J.C., "Dorsal root afferents contact migrating motoneurons," submitted 1982. TABLE OF CONTENTS

Page

DEDICATION ...... ii

ACKNOWLEDGMENTS ...... iii

VITA ...... iv

LIST OF FIGURES ...... ix

INTRODUCTION ...... 1

MATERIAL AND METHODS ...... 5

Subjects ...... 5 Surgery ...... 5 Tissue Preparation ...... 6 Analysis ...... 7

RESULTS: ADULT FROG ...... 9

Primary Afferent Distribution...... 9 LMC Motoneuron Morphology ...... 10 Primary Afferent-LMC Interactions ...... 11 Electron Microscopic Observations ...... 12

RESULTS: LARVAL FROG ...... 14

FOOT PADDLE STAGES ...... 14 Primary Afferents ...... 14 LMC Motoneurons ...... 15 Primary Afferent-LMC Motoneuron Interactions ...... 16

FOOT OR PREMETAMORPHIC STAGES ...... 17 Primary Afferents ...... 17 LMC Motoneuron Morphology ...... 19 Primary Afferent-LMC Motoneuron Interactions ...... 20 Electron Microscopic Observations ...... 22

vii Page

MIGRATING LMC MOTONEURONS ...... 22

DISCUSSION ...... 26

Primary Afferent Distribution in Adult Frog ...... 26 The LMC in the Adult Frog ...... 31 Development of Primary Afferents ...... 33 Development of the LMC Motoneurons ...... 37 Migrating Motoneurons ...... 39 Summary ...... 41

ILLUSTRATIONS ...... 44

BIBLIOGRAPHY ...... 100

viii LIST OF FIGURES

FIGURE PAGE

1 Drawing of a transverse section through the adult lumbar cord ...... 46

2 Transverse section through the entry zone of an HRP injury-filled ninth dorsal root in an adult bullfrog ...... 48

3 High magnification micrograph of the moto­ neuron seen in the box in Figure 2 48

4 Transverse section through the lumbar enlargement of an adult bullfrog in which the ninth dorsal and ventral roots were injury-filled with HRP...... 50

5 Section similar to Figure 4 through the lumbar enlargement of an adult bullfrog ...... 50

6 Drawing of a fusiform motoneuron located in the ventral LMC of an adult bullfrog ...... 52

7 Drawing of dorsally situated LMC motoneuron in an adult bullfrog ...... 54

8 High magnification micrograph of an HRP filled motoneuron dendrite within the lateral field in an adult bullfrog ...... 56

9 Type C terminal in contact with an LMC motoneuron soma in the lumbar enlargement of an adult bullfrog ...... 58

10 HRP labelled primary afferent terminal in synaptic contact with an LMC motoneuron soma in the lumbar enlargement of an adult bullfrog ...... 58

ix FIGURE PAGE

11 Labelled primary afferent terminal in con­ tact with an LMC motoneuron dendrite follow- HRP injury-filling of the ninth dorsal root in an adult bullfrog ...... 60

12 HRP labelled primary afferent terminal in contact with an LMC motoneuron dendrite in an adult bullfrog ...... 60

13 Lightly labelled primary afferent terminal in contact with a small dendrite within the dorsal LMC ...... 60

14 Four stages of hindlimb development in the ranid tadpole ...... 62

15 Drawing of a transverse section through the lumbar enlargement of a stage VI tadpole ...... 64

16 Transverse section through lumbar spinal cord of a stage VI tadpole ...... 66

17 High magnification micrograph of rectangular area in Figure 16 66

18 Drawing of a transverse section through the lumbar enlargement of a stage X tadpole ...... 68

19 Transverse section through the lumbar enlargement of a stage X tadpole following HRP injury filling of dorsal and ventral roots ...... 71

20 High magnification of the dorsal LMC and the lateral field dorsal to it ...... 71

21 High magnification micrograph of unlabelled LMC in stage X tadpole ...... 71

22 Labelled LMC motoneuron ...... 71

23 Drawing of a dorsally positioned LMC motoneuron of a stage X tadpole ...... 73

24 Drawing of the LMC motoneuron shown in Figure 22 75

x FIGURE PAGE

25 Electron micrograph of type F synapse on a primary motoneuron of a stage VI tadpole ...... 77

26 Electron micrograph of a type C synapse on a primary motoneuron of a stage VI tadpole ...... • • • • 77

27 Electron micrograph of an early type S synapse on an HRP filled LMC motoneuron in a stage X tadpole ...... 77

28 Drawing of a transverse section through the lumbar enlargement of a stage XIV tadpole .... 79

29 Maraglas embedded transverse section through the root entry zone of a stage XIII tadpole ... 82

30 Higher magnification micrograph of the dorsal quadrant of the section shown in Figure 29 «... 82

31 Higher magnification micrograph of the lower quadrant of the section shown in Figure 29 •• • • 82

32 Drawing of a dorsally positioned LMC moto­ neuron in a stage XIV tadpole ...... 84

33 Transverse section through the lumbar enlargement of a stage XVII tadpole with HRP injury filling of dorsal and ventral roots • • • • 87

34 Higher magnification micrograph of the LMC and the more ventral terminal field ...... 87

35 Higher magnification micrograph of the dorsal LMC ...... 87

36 High magnification micrograph of the lateral funiculus ...... 87

37 Transverse section through the LMC of a stage XIV tadpole with HRP injury filling of the ninth dorsal root ...... 89

38 Transverse section through the LMC of a stage XV tadpole with HRP injury filling of the ninth dorsal root ...... 89

xi FIGURE PAGE

39 Parasagittal section through LMC in stage XVIII tadpole ...... 89

40 Labelled primary afferent terminal in synaptic contact with labelled LMC motoneuron dendrite within the lateral field of a stage XIV tadpole ...... 91

41 Labelled primary afferent terminal making synaptic contact with a labelled LMC moto­ neuron dendrite just dorsal to the LMC ...... 91

42 Labelled primary afferent terminal synap- sing upon a labelled motoneuron dendrite within the lateral field ...... 91

43 Primary afferent terminal within the dorsal LMC synapsing upon a large labelled LMC motoneuron dendrite in a stage XVII tadpole ... 91

44 Drawings of the dorsal and ventral migra­ tory pathways as observed in transverse sections through the rostral lumbar enlarge­ ment following HRP injury-filling of the ninth dorsal and ventral roots ...... 93

45 Transverse section of the dorsal migratory pathway shown in the drawing in Figure 44A .... 95

46 Transverse section of the ventral migratory pathway shown in the drawing in Figure 44B .... 95

47 Parasagittal section through the rostral lumbar enlargement of a stage XIV tadpole following HRP injury filling of the ninth dorsal and ventral roots...... 97

48 High magnification micrograph of area within the rectangle in Figure 47 97

49 Electron micrograph of a synaptic contact on an HRP filled dorsally migrating moto­ neuron ...... 99

50 Vesicle filled profile in apposition to a labelled migrating motoneuron within the dorsal pathway ...... 99

xii FIGURE PAGE

51 Electron micrograph of a labelled primary afferent profile, similar to the one at the arrowhead in Figure 48 99

xiii INTRODUCTION

Amphibians hold a unique position in phylogeny between aquatic and terrestrial vertebrates. Moreover, their larval development is distinct in that their central nervous system undergoes a metamor­ phosis from one organized strictly for swimming to one organized for quadripedal locomotion. During metamorphosis, the frog's lumbosacral spinal cord undergoes extensive changes to accommodate the innervation of the developing hindlimb. Some of the main changes are the pro­ liferation, migration and differentiation of a distinct lateral motor column (LMC) containing motoneurons which innervate the hindlimbs.

In addition to the development of motoneuron efferents to the muscles of the maturing hindlimb, there is a concurrent development of dorsal root afferents which carry information centrally irrom cutaneous and deep receptors. The purpose of the present investigation is to study the development of the dorsal root afferents and the motoneurons in the lumbar spinal cord of larval bullfrogs as a function of hindlimb development using injury filling of dorsal and ventral roots with horseradish peroxidase (HRP).

Early studies of the distribution of dorsal root afferents in spinal cord of the adult frog using either the Nauta method for demon­ strating degenerating axons (Joseph and Whitlock, 1968) or auto­ radiographic procedures (Lasek, Joseph and Whitlock, 1968) failed to show a projection to the ventral horn. These experiments were in 2 contrast to similar studies in mammals where dorsal root afferents were shown to enter lamina IX of the ventral horn (see Ebbesson,

1976). Consequently, it was suggested that a shift of dorsal root afferent terminals occurs as one goes up the phylogenetic scale, from the distal dendrites in amphibia to the proximal dendrites and somata of motoneurons in mammals (Corvaja and Pellegrini, 1975).

The purported absence of primary afferent axosomatic contacts on frog LMC motoneurons was also suggested to be the anatomical basis for the results of physiological studies of the monosynaptic reflex arc in frogs. These studies, conducted by a number of investigators

(see Simpson, 1976, for a review) reported that in the frog primary afferent input seldom evokes motoneuron discharge. It was, there­ fore, held that the inefficacy of the monosynaptic activation of LMC motoneurons was due to the restriction of primary afferent input to the more distal portion of the LMC motoneuron dendrites (Simpson,

1976).

More sensitive anatomical techniques, however, indicate that there is primary afferent input to the proximal dendrites and somata for anuran motoneurons. In contrast to the degeneration and auto­ radiographic studies, Szekely (1976) using cobaltous chloride ionto- phoretic filling of lumbar dorsal roots in Rana esculenta showed that dorsal root afferent axons in the adult frog do enter the ventral horn and come in close apposition to the somata of the motoneurons.

In the brachial spinal cord of Rana catesbeiana, Frank and Westerfield

(1982) using HRP injury filling of triceps brachii peripheral nerve reported similar findings. They reported, however, that afferent axons from the triceps brachii terminated no closer than 20-30 ^im

to the somata of motoneurons innervating the same muscle.

In light of more recent physiological evidence for a "typical" monosynaptic reflex arc (Szekely and Czeh, 1976) coupled with the

recent anatomical evidence (Szekely, 1976; Frank and Westerfield,

1982) for a substantial input to LMC motoneurons in the adult frog,

it is of interest to reassess the relationship between primary afferents and motoneurons, and to determine when and how this relation­ ship is established during hindlimb development. To address this question, application of HRP to the cut dorsal and/or ventral roots

to demonstrate primary afferent axons and motoneuron somata and dendrites seems to be particularly appropriate. HRP applied to cut dorsal roots in cats (Proshansky and Egger, 1977) has been shown to diffusely fill primary afferent axons rendering them visible for light (Proshansky and Egger, 1977) and electron microscopic examin­ ation (Beattie, Bresnahan and King, 1978). Similarly, HRP applied to cut ventral roots has been shown to diffusely fill motoneuron somata and dendrites (Light and Metz, 1978; Chu-Wang, Oppenheim and

Farel, 1981).

To date, many developmental studies within the central nervous system have relied upon correlation with the adult structure to identify various elements. However, in the absence of an easily characterized organization within the ventral horn during develop­ ment, unequivocal identification of presynaptic elements as well as postsynaptic targets has been difficult (Oppenheim, Chu-Wang and

Foelix, 1975; Bodian, 1975). In contrast, the use of HRP anterograde filling of dorsal roots

and retrograde filling of ventral roots in the present study has the

advantage of labelling a specific developing afferent input to a specific, circumscribed column of developing motoneurons. Moreover, developing secondary and tertiary dendrites of LMC motoneurons,

though removed from the LMC proper, can clearly be identified in an immature neuropil at the light and electron microscopic levels. An additional advantage, at least by hindlimb stage X (Taylor and

Kollros, 1946) is that migrating motoneurons, presumably destined for the LMC send axons into the ventral root and are, therefore, retrogradely filled with HRP when it is applied to the cut ventral root (Chu-Wang, Oppenheim and Farel, 1981).

The present study provides an opportunity to observe at the light and electron microscopic levels, the development of a labelled specific input, dorsal root afferents, in relation to labelled motoneurons already extant in the LMC as well as others which are in the process of migrating to the LMC. MATERIAL AND METHODS

SUBJECTS:

Thirty-two bullfrog (Rana catesbeiana) tadpoles obtained from

Riverside Biological were grouped according to the staging criteria

set forth by Taylor and Kollros (1946) . The stages chosen for study

were the foot paddle stages (VI-X) and the foot stages (XI-XVIII).

It is during these two major stages or phases that maximal hindlimb

development occurs as well as the concurrent development of the

spinal cord lumbar enlargement. In addition to tadpoles, 18 adult

Rana catesbeiana and Rana pipiens were used to establish the mature

pattern of primary afferent-motoneuron connectivity.

SURGERY:

Individual tadpoles were anesthetized in a 1.0 percent solution

of urethane or by hypothermia and placed in a petri dish filled with

distilled water. Laminectomies to expose the lumbar enlargement

(spinal segments 8 , 9, 10) were performed with a pair of 45 degree

iridectomy scissors under a Zeiss dissecting microscope. Care was

taken to expose the lateral aspect of the cord to allow access to ventral as well as dorsal roots. The ninth dorsal and/or ventral roots were cut and chips of dried, concentrated HRP (Type VI, Sigma

Chemical Co.) were applied with fine microforceps to the proximal cut ends. One hour was allowed to elapse during which the site was

5 kept as dry as possible and during which chips of HRP were continually

applied. Excess HRP was then flushed from the site with saline and the

muscle and skin sutured. The tadpoles were then returned to their

home aquarium. Surgery in the adult frogs was essentially the same

as that for the tadpoles except that Fluothane was used as the

anesthetic.

TISSUE PREPARATION:

Following a survival time of 24 hours, the animals were re­

anesthetized and intracardially perfused with a gravity perfusion

apparatus, with phosphate buffered saline. A pulse of lidocaine

(O.lcc) was injected directly into the heart at the initiation of

the saline perfusion in order to facilitate vasodilation.

Following a five to ten minute flushing of the vascular system,

the animals were perfused for 30 minutes with a fixative containing

1.0 percent glutaraldehyde and 3.0 percent paraformaldehyde in 0.1M phosphate buffer (pH 7.4). The spinal cord was then re-exposed and

the entire animal was immersed in fresh fixative for six hours after which it was placed in a 30.0 percent phosphate buffered sucrose solution overnight.

On the following day, the spinal cord was dissected out and the lumbar enlargement was blocked by making a transverse cut at the eighth root and another at the eleventh spinal root. The blocked lumbar enlargement was then sectioned parasagittally or transversely at 60 jum on an Oxford vibratome. Sections were rinsed in phosphate buffer, intensified with 5.0 percent cobaltous chloride (Adams, 1977) incubated in diaminobenzidine

(DAB) (Baker Chemical) and reacted with H2O2.

For light microscopy, sections were mounted on chrome-alum coated slides and counter stained with cresy1-violet. Those sections selected for electron microscopy (EM) were post-fixed in 1.0 percent osmium tetroxide, stained en-bloc with uranyl acetate, dehydrated through a graded series of ethanols and flat embedded in Maraglas between two sheets of aclar plastic (Allied Chemical Co.). This flat embedding yielded smooth, clear sheets of Maraglas only a few micrometers thicker than the tissue sections they contained. The Maraglas sheets were cut with a scissors and individual sections were taped to micro­ scope slides for light microscopic analysis and photography.

ANALYSIS:

Light microscopic analysis was conducted on both cresyl-violet and plastic sections from adult frogs and from tadpoles at various stages of hindlimb development to determine three major pieces of information. First, cases in which the dorsal roots were labelled were examined to determine the pattern of primary afferent distri­ bution particularly in relation to the LMC. Secondly cases in which the ventral root was labelled with HRP were examined to determine the typical patterns and extent of dendritic arborization of the moto­ neurons in the LMC as well as the areas of overlap with primary afferents. Labelled motoneuron somata and dendrites were drawn at

100X using an Olympus microscope with drawing tube. Third, cases which had simultaneous labelling of dorsal and ventral roots were examined to determine when and where labelled primary afferent axon en passant and terminal swellings made apparent contact with LMC motoneuron dendrites and somata.

Electron microscopy was performed on the same Maraglas sections used in the light microscopic analysis. After light microscopic observations, the LMC and the gray matter containing primary afferents and/or LMC motoneuron dendrites were trimmed from the Maraglas sections with the aid of a Zeiss dissection microscope. The optical clarity of the Maraglas embedded sections allowed very precise excision of the area of interest. Selected pieces of Maraglas embedded sections were then glued with cyanoacrylic glue to a blank Maraglas capsule. After the glue dried, the block was trimmed further for ultramicrotomy.

One-micron thick sections were cut on a Reichert Om U2 ultra­ microtome. They were stained with toluidine blue and examined to determine the precise location of labelled elements. Drawings were made of one-micron thick sections to aid in locating specific areas and structures on the electron microscope.

Thin sections were picked up on formvar coated slotted grids and stained with uranyl acetate and lead citrate. Grids were examined on a Phillips 301 electron microscope. Labelled elements and montages of LMC motoneuron somata and dendrites were photographed at 9,100X.

Suspected synaptic contacts were examined and photographed at 25,000X. RESULTS: ADULT FROG

Primary afferent distribution. Injury filling of lumbar dorsal roots with HRP in the adult ranid spinal cord reveals the pattern of primary afferent distribution illustrated in Figure 1. An extensive medial division (MD) can be distinguished which contains primarily large diameter fibers which fill the dorsal funiculus (Figure 1). More laterally, smaller fibers of the lateral division (LD) are inter­ spersed with some large diameter axons within the dorsolateral funiculus (Figure 1).

Two major regions of primary afferent arborization are apparent in the dorsal horn and intermediate gray. One major terminal zone

(I, Figures 1 and 2) corresponds to the dorsal neuropil of Silver

(1942) or the dorsal field described by Ebbeson (1976) henceforth referred to as the dorsal field. Located in the dorsal portion of the spinal gray, this terminal field receives fibers from the lateral division as well as collaterals from some of the larger medial di­ vision axons. The second major terminal zone (II, Figures 1 and 2) corresponds to the lateral field described by Ebbesson (1976) which includes the deep dorsal horn and the intermediate gray. Collaterals from the longitudinally oriented fibers of the dorsal columns form five or six distinct fascicles which enter the dorsal horn (Figure 1, small arrows). Some of these large fibers contribute branches to the

9 10 dorsal terminal field. Others traverse the dorsal gray unbranched and form a dense plexus of axons and varicosities within the deeper lateral field. Additionally, some large diameter fibers which con­ tribute to this deep terminal field continue into the ventral horn

(Figure 2, arrows), where labelled axons may be seen entering the dorsal portion of the LMC.

LMC motoneuron morphology. In transverse Nissl stained sections the somata of LMC motoneurons lie within a well demarcated nucleus of 8-12 large neurons which range in shape from multipolar, angular cells most dorsally to distinctly fusiform types in the ventralmost region (Figure 2).

LMC motoneurons were labelled by application of HRP to the cut ninth ventral root in three cases. This procedure provides a Golgi- like picture of the motoneuron soma and its dendritic arborization

(Figures 4 and 5). HRP labelled LMC motoneurons have extensive dendritic arborizations (Figures 4-7). There are typically five or six primary dendrites which arise from characteristic positions on the motoneuron soma, and which arborize in particular regions of the spinal gray and white matter. Of the five or six primary dendrites encountered, three have branches that ramify within the region of the deeper primary afferent terminal field (Figures 1, 6 and 7). These are the dorsolateral and dorsomedial primary dendrites. The dorso­ lateral primary dendrites (designated and D2 by Bregman and Cruce,

1980) arborize primarily within the lateral intermediate gray and the dorsal portion of the lateral funiculus (Figures 4-7). The more medial of the dorsolateral dendrites (D^, Figures 6 and 7) has branches which 11 ramify predominantly within the lateral field, although some branches extend into the lateral funiculus. The more lateral of the dorso­ lateral dendrites (D2 , Figures 6 and 7) has branches in the lateral field as well as extensive arborization within the lateral funiculus, where terminal dendritic branches can be traced as far as the pial surface where they contribute to the subpial (Szekely, 1976) or extra- medullary (Stensaas and Stensaas, 1971) plexus (Figures 1 and 5, arrowheads).

A dorsomedial dendrite is also usually present and branches in the region of the lateral field (Figures 6 and 7). This dendrite can arise from positions on the soma which suggest that it corresponds to either a branch of D 7 or to Dg of Bregman and Cruce (1980). Other primary dendrites arising from HRP labelled LMC motoneurons arborize either in the medial ventral horn and/or in the ventral white matter as described by Bregman and Cruce (1980) and are not in position to receive dorsal root input.

Primary afferent-LMC interactions. In Nissl stained transverse sections from cases where the dorsal root alone has been injury filled with HRP, labelled primary afferent axons can be observed leaving the deeper terminal field to enter the ventral horn and the dorsal portion of the LMC (Figure 2, arrows). These large diameter fibers exhibit varicosities along their length. In experiments where both dorsal and ventral roots are injury filled with HRP, the primary afferent vari­ cosities are often seen in close apposition to labelled motoneuron dendrites, and dorsal root axons can be seen wrapping around the dorsally extending processes of labelled LMC neurons (Figure 8). 12

Within the LMC proper, HRP labelled axons exhibit varicosities in close apposition to motoneuron somata and primary dendrites in the more dorsal portion of the LMC (Figure 3, block arrow). They do not enter more ventral regions.

Electron microscopic observations of primary afferents. HRP labelled profiles are easily distinguished from unlabelled elements with the electron microscope. HRP reaction product is diffusely distributed and aggregates on membrane surfaces (Figures 10-13) as reported in other species (Beattie, Bresnahan and King, 1978 and

Mason, 1982). In the region just dorsal to the LMC, numerous HRP labelled primary afferent terminals are observed in contact with

intermediate size and small dendrites (Figures 11-13). Further ventrally in the LMC, labelled profiles are in apposition to the primary dendrites and somata of motoneurons (Figure 10) , although labelled axosomatic profiles with unequivocal synaptic speciali­ zations are encountered only infrequently. More ventrally in the

LMC progressively fewer labelled profiles are seen until most ventrally, in the region of the fusiform motoneurons, no labelled profiles are observed.

Several types of synaptic terminal have been described previously in the LMC and intermediate gray of the ranid spinal cord (Sotelo and Grafova, 1976). In the present study, three types have been distinguished: type C, type F and type S. Type C

terminals are characterized by small pleomorphic clear vesicles and

typically contain one or two larger dense-cored vesicles (Figure 10,

large block arrow). Type C terminals are seen in contact with LMC 13 motoneuron primary dendrites and somata, and are associated with a postsynaptic subsurface cistern (Figures 9 and 10, block arrows, see also Sotelo and Grafova, 1976). Type C terminals were not observed to be labelled with HRP after dorsal root applications in the present study.

Type F terminals contain pleomorphic clear vesicles with no dense-cored vesicles and have little or no postsynaptic density associated with them (Figure 12). These terminals are found in contact with motoneuron somata as well as with large, intermediate and small diameter dendritic profiles. Dorsal root injury filling with HRP did not label type F terminals in contact with LMC moto­ neuron somata or dendrites.

Type S terminals are characterized by spherical agranular vesicles, occasional dense-cored vesicles and they are typically associated with some postsynaptic dense material (Figure 12). Type

S terminals also contact LMC motoneuron somata and dendrites. Some type S terminals contacting motoneuron somata and dendrites, were labelled after HRP application to dorsal roots (Figures 10-13).

These HRP labelled terminals contain small spherical, agranular vesicles (30-35 nm. diam.) and an occasional dense-cored vesicle

(Figure 11, block arrow). These terminals are associated with a postsynaptic density, but it is sparse compared with primary afferent terminations on motoneurons in mammals (e.g. Ralston and

Ralston, 1979). Occasionally, labelled terminals are quite dark and may be undergoing degeneration (Figure 12). RESULTS: LARVAL FROG

The development of dorsal root primary afferents, the LMC and their interaction in the lumbar spinal cord of the bullfrog occurs during the foot paddle stages (VI-X) and the foot or premetamorphic stages (XI-XVIII) of hindlimb development as illustrated in Figure 14.

FOOT PADDLE STAGES (VI-X) :

Primary afferents. During the foot paddle stages, the primary afferent axons give rise to collaterals that penetrate a poorly differentiated spinal gray matter.

In the first foot paddle stage (VI), the hindlimb ends as a flattened paddle (Figure 14). At stage VI, HRP applied to the cut dorsal roots fills dorsal funiculus axons (Figures 15 and 16). A few larger diameter axons penetrate the dorsal gray matter along the margin between the gray and white matter (Figures 16 and 17). Vari­ cosities, between 1.0 and 2.0 pm in diameter, are present along the length and at the tips of these larger fibers (Figure 17, block arrows). In addition to the large diameter fibers, smaller, fine

HRP labelled fibers are observed within the developing dorsal horn between tightly packed, small (5.0-6.0 p m in diameter) spherical cells which populate the region (Figures 15 and 16, arrows). Along these fine fibers, small varicosities, 0.5 pm or less in diameter are observed (Figure 17, arrow). It appears that HRP labelled axons

14 15

travel no further ventrally than the dorsal portion of the developing

intermediate gray (Figures 15 and 16), the region which in the adult

corresponds to the lateral field (Ebbesson, 1976).

The final foot paddle stage, stage X, is marked-by the appear­

ance of five distinct, though rudimentary digits (Figure 14). At this

stage, dorsal root injury filling with HRP reveals a marked change in

the distribution of labelled primary afferent axons. By hindlimb

stage X, HRP labelled afferents fill the lateral two-thirds of the dorsal horn and extend into the intermediate gray (Figures 18 and 19,

PAs). In addition, a few primary afferent axons can be seen growing

into the ventral horn (Figure 20) and an occasional labelled axon is

observed entering the LMC (Figure 21, block arrow).

Many HRP labelled primary afferent axons have varicosities along

their length and at their tips (Figure 20, block arrow). These vari­

cosities are 0 .5-1.0 urn in diameter.

LMC motoneurons. Nissl stained transverse sections through the lumbar enlargement of stage VI tadpoles revealed that the LMC was a

tightly packed clump of bipolar, fusiform motoneurons, each measuring an average of 7.0 jam wide (Figure 16).

By stage X, HRP applied to cut ventral roots (Figure 19) demon­ strated the early differentiation of the primary dendrites. Figures

23 and 24 are drawings of two LMC motoneurons, both of which are located in the dorsal LMC in the rostral end of spinal segment nine.

Each motoneuron is multipolar with somata measuring approximately

15.0 urn at its greatest dimension. The motoneuron in Figure 23 has two major dendrites and D2- Dendrite has secondary and 16

tertiary branches which extend both dorsally into the lateral field

and laterally into the LF. Dendrite D2 has secondary branches ex­

tending laterally. The more dorsal branch appears to taper into a

long fine unbranched process which extends almost 150 /urn into the

LF, nearly to the pial surface. The soma also gives rise to a

number of fine, unbranched processes. One, designated Dg because

of its relative position along the cell soma, extends without

branching nearly 60 /am into the lateral field. The motoneuron

shown in Figure 24 also has dendrites designated D]_, D2 and Dg.

Dendrite Dj^ has secondary branches extending dorsally into the

lateral field. Dendrite D2 , similarly to D 2 in Figure 24, has secondary branches the more dorsal of which tapers into a fine process extending into the LF.

Primary afferent-LMC motoneuron interaction. The first apparent

contact by lumbar dorsal root afferent axons with LMC motoneuron

somata and dendrites is made during the foot paddle stages.

At the beginning of the foot paddle stages, stage VI, no labelled primary afferent axons are observed entering the ventral horn (Figures 15 and 16). By stage X the end of the foot stages,

HRP labelled primary afferent axons enter the LMC (Figure 21) and

their varicosities are observed in close apposition to LMC moto­ neuron dendrites (Figure 22, white arrows, and Figure 24, small arrows).

At the electron microscopic level, in the one stage VI

tadpole examined, no identifiable synaptic terminals of any type were observed in contact with LMC motoneuron somata. In contrast, 17 mature appearing synaptic profiles are present on the large primary motoneurons in the ventromedial gray (Figures 25 and 26). Primary motoneurons (PMNs), large motoneurons located in the ventromedial gray, were observed throughout the foot paddle and foot stages

(Figures 18, 19, 33 and 44). Those in the lumbar enlargement with axons in the ninth ventral root innervate axial muscles involved in swimming (Forehand and Farel, in press).

By hindlimb stage X, a few vesicle filled profiles are observed in contact with HRP labelled LMC motoneuron somata (Figure 27, asterisk). These are in marked contrast to the mature appearing synapses on labelled LMC motoneuron dendrites within the LF. How­ ever, in the stage X tissue examined, no labelled primary afferent terminals have been observed in contact with labelled LMC motoneuron somata or dendrites at the electron microscopic level.

FOOT OR PREMETAMORPHIC STAGES (XI-XVII):

The foot stages or premetamorphic stages are those during which differentiation of the foot takes place (Taylor and Kollros, 1946).

During these stages (XI-XVII), the hindlimb evolves from one with five rudimentary digits to one which morphologically looks like the adult, except for its size (Figure 14).

Primary afferents. By hindlimb stages XIII and XIV, dorsal root injury-filling with HRP reveals discemable separation of primary afferent fibers into lateral and medial divisions (Figure 30).

Coincident with the differentiation of lateral and medial divisions of dorsal root axons is the early appearance of two primary afferent terminal fields (Figures 28 and 30, I and II). The first (I) lies 18 just ventral to the dorsal funiculus in the dorsal gray (Figures

28-30). This area corresponds to the dorsal field (Ebbesson, 1976) in the adult. Examination of this region reveals fine HRP labelled axons with varicosities along their lengths (Figure 30, arrowheads).

Collaterals from the large diameter axons in the dorsal funi­ culus form a few fascicles (Figure 30, arrows) which, similar to those in the adult, enter the dorsal horn. Branches of some of these collateral axons ramify in the dorsal terminal field. Others pass through this region unbranched and pass ventrally to contribute to a plexus of large diameter primary afferents in the deeper dorsal horn and the intermediate gray (Figures 28-30, II). This plexus lies in the region described by Ebbesson (1976) as the lateral field in the adult. Varicosities measuring up to 2.0 *im in diameter are found along the length of fibers in this plexus, some at apparent branch points (Figure 31, block arrow). As in stage X, HRP labelled primary afferents at stages XIII and XIV are observed entering the

LMC (Figure 37, arrow) with approximately the same frequency.

The end of the foot stages is attained when the tadpole hind­ limb reaches an adult-like appearance at stage XVII (Figure 14).

HRP injury-filling of dorsal roots at hindlimb stage XVII reveals a primary afferent distribution which is qualitatively indistinguishable from that observed in the adult spinal cord

(Compare Figure 33 with Figure 2). The medial and lateral divisions of primary afferent fibers, composed of large and small caliber axons respectively, are now clearly defined (Figure 33). Moreover, the two distinct terminal fields of primary afferents which lie in the dorsal (I) and lateral (II) fields are well differentiated (Figure 33). 19

LMC motoneuron morphology. During the foot stages, the LMC evolves

from the tightly packed clump of small fusiform motoneurons which characterize the LMC during the foot paddle stages (Figures 15, 16 and 19) to a less tightly packed accumulation of large motoneurons ranging in shape from multipolar dorsally to fusiform more ventrally

(Figures 31 and 33).

Figure 32 is a drawing of a stage XIV motoneuron from the dorsal

LMC. It has five primary dendrites emanating from a large (35 /um in length) soma. Three dendrites have been assigned designations accord­ ing to the scheme of Bregman and Cruce (1980) in the adult. The D2 dendrite, similar to the D2 dendrite in Figure 23, has secondary branches which enter the lateral funiculus. At stage XIV, however, these secondary branches of the D2 dendrite give rise to third and fourth order branches.

Two primary dendrites (D^ and Dg) have branches which arborize in the region dorsal to the LMC which corresponds to the lateral field in the adult. The dorsolateral dendrite (D-^) has a few branches which extend into the LF (Figure 32). Its predominant arborization, however, is in the lateral field. Here, the D^ dendrite has third and fourth order branches. The dorsomedial dendrite (Dg) is a very fine dendrite. It extends more than 180 jjm from the cell soma into the lateral field where it has two fine secondary branches (Figure 32).

The Dg dendrite of the stage XIV motoneuron in Figure 32 should be compared to the Dg dendrite of the stage X motoneuron shown in

Figure 23. 20

HRP injury-filling of ventral roots at hindlimb stage XVII revealed an essentially adult appearing LMC (Figure 33). In two stage XVII tadpoles with extensive ventral root injury filling with

HRP, labelled elements were so dense as to render it impossible to draw a single motoneuron with its dendritic arbors. However, a com­ parison of Figure 33 with Figure 4 and 5 shows that the pattern and extent of dendritic arborizations are the same for stage XVII tad­ poles and adult frogs. The D 2 dendrites of the LMC motoneurons have arborizations predominantly in the lateral funiculus with their terminal dendrites reaching the pial surface. Here they turn and run parallel to the pial surface contributing to the subpial

(Szekely, 1976) or extramedullary plexus (Stensaas and Stensaas,

1971) (Figure 36, block arrow). Within the LF, near the pial surface, numerous hollow appearing varicosities (Figure 36, arrows) are observed along the length of some HRP labelled dendrites. In addition to arborization within the LF, some branches of the D2 dendrites pass into the lateral field along the gray-white margin

(Figure 35, closed block arrow).

The dendrite, just as in the adult, arborizes predominantly within the lateral field (Figure 34). It also contributes some branches to the LF (Figure 34). In addition, branches of the D7 dendrite or the Dg dendrite arborize in the more medial portion of the lateral field (Figure 34).

Primary afferent-LMC motoneuron interactions. In some tadpoles of the foot stages, only the dorsal roots were injury filled with HRP

(Figures 37 and 38). In Nissl stained transverse sections of the 21

lumbar enlargements of these cases only one or two primary afferent

axons are observed entering the LMC proper (Figures 37 and 38). In

these cases and in others with both dorsal and ventral root injury

filling with HRP (Figure 35) a few primary afferent varicosities are

observed in apposition to LMC motoneuron somata. Parasagittal

sections through the LMC from a stage XVIII tadpole with both dorsal

and ventral roots labelled demonstrate that only a small number of

primary afferent axons enter the LMC proper (Figure 39). Here,

varicosities are observed in close apposition to motoneuron somata

and primary dendrites (Figure 39, arrow).

During the foot stages, increasing primary afferent-LMC moto­

neuron interaction is evident in the region of the lateral field.

At hindlimb stages XIII and XIV relatively fewer labelled dendrites

are observed in comparison to the number in the LF (Figure 31). How­

ever, in the more ventral aspect of the lateral field, immediately

dorsal to the LMC, primary afferent varicosities are observed in

close apposition to labelled LMC motoneuron dendrites (Figure 31,

arrows). In contrast, at hindlimb stage XVII, large numbers of

labelled dendrites are observed in the lateral field (Figures 34 and

35). These dendrites, branches of D]_, D2, Dg and D7 are seen in

Figure 34 to form a dendritic field which is overlapped by the more ventral terminal field of primary afferents. Moreover, closer exam­

ination reveals labelled primary afferent axons wrapping labelled

dendrites (Figure 35, open block arrow). Varicosities of these

afferent axons are seen in close apposition to LMC dendrites. 22

Electron microscopic observations. At the electron microscopic level,

HRP labelled primary afferent terminals are first positively identified

in synaptic contact with HRP labelled dendrites at hindlimb stages XIII

and XIV (Figures 40 and 41). Although numerous appositions between

labelled elements are observed, those which can be unquestionably

characterized as axodendritic synapses are infrequent. These axo­

dendritic synapses have been observed within the lateral field along

the gray-white margin immediately dorsal to the LMC. Within the LMC

proper, there were no labelled primary afferent terminals observed

in contact with labelled motoneuron somata or primary dendrites. This

observation at the electron microscopic level reflects the apparent

paucity of labelled primary afferent axons observed entering the LMC

at the light level at these stages (Figures 37 and 38).

Examination of the lateral field and LMC at hindlimb stage XVII

shows more frequent axodendritic synapses with both elements labelled

(Figures 42 and 43). These axodendritic synapses are found throughout

the lateral field and are often just dorsal to the LMC on larger den­

drites (Figure 42). No axosomatic synapses with both elements labelled have been observed.

MIGRATING LMC MOTONEURONS:

Frequently, in hindlimb stages X to XIV, HRP applied to cut lumbar ventral roots labels round and fusiform cells measuring 8.0-12.0 *im in diameter between the ventricular zone (VZ) and the LMC. These migrat­

ing motoneurons have been observed by Grobstein (1979) using HRP injury filling of ventral roots in fetal rat spinal cord and by

Bemelsman and Farel (1980) using this same procedure in bullfrog tad­ poles . 23

In the present study, the largest numbers of migrating moto­

neurons have been observed in three of seven stage XIV tadpoles. In

these animals, two distinct pathways of migration are encountered

(Figure 44). One pathway, the dorsal migratory pathway, forms an arc

through the developing lateral field between the upper portion of the

ventral quadrant of the VZ and the dorsal LMC (Figures 44A and 45).

The second pathway, the ventral migratory pathway, forms an arc along

the ventral margin of the developing gray matter between the lower portion of the ventral quadrant of the VZ and the ventral LMC

(Figures 44B and 46).

Labelled neurons in both the dorsal and ventral migratory path­ ways have been observed rostral to a column of postmigratory LMC motoneurons which were also labelled and which had begun to establish dendritic arborizations. Figure 47 is a parasagittal section through

the lumbar enlargement of a stage XIV tadpole with both ninth dorsal and ventral roots labelled with HRP. Here a band of small, spherical cells are observed to be labelled rostrodorsal to the LMC. These are cells in the dorsal migratory pathway which are shown in higher mag­ nification in Figure 48. Additionally, a few cellular profiles

(Figure 48, arrows) are observed to be HRP positive just dorsal to the primary motoneurons (PMNs). These are cells within the ventral migratory pathway. The migrating motoneurons closer to the VZ are typically fusiform in transverse section (Figures 44-46) and round in sagittal section (Figures 47 and 48). In both the dorsal and the ventral pathways, the migrating motoneurons typically exhibit two processes which are medio-laterally oriented (Figure 44). The 24

medially extending process can often be traced into the VZ while the

laterally extending process can frequently be followed into the develop­

ing LMC (Figure 44). In addition to these medial and lateral processes,

an axon can often be traced toward or into the ventral root, particu­

larly in the ventral migration (Figure 44B, block arrow).

Other than the positions of the migratory pathways, two character­

istics distinguish the migrating motoneurons of the dorsal and the

ventral pathways. One, the cells within the dorsal pathway are smaller

(range: 8 .0-8 .5 um in diameter) than those within the ventral pathway

(range: 11.0-12.0 um in diameter). This size difference is evident in

the drawings in Figure 44. More significant, perhaps, is the differ­

ence in the environments through which the cells migrate. The dorsally migrating motoneurons follow a path through the lateral field. By hindlimb stage XIV when this migration appears more evident, the

lateral field is heavily populated with primary afferent axons and

their varicosities (Figure 44A, arrows). At the light microscopic

level, labelled primary afferent varicosities are frequently observed

in close apposition to labelled somata and processes of dorsally mi­ grating motoneurons (Figure 44A, block arrow, and Figure 48). In contrast, as seen in Figure 44B, the migrating motoneurons within the ventral pathway do not pass through a field of primary afferent axons and varicosities.

Electron microscopic observations indicate that the migrating motoneurons are contacted by vesicle filled elements confirming the finding of Chu-Wang, Oppenheim and Farel (1981). Some examples of this are shown in Figures 49 and 50. Figure 49 shows a profile filled 25 with a number of pleomorphic vesicles and large dilated cisterna of smooth endoplasmic reticulum (SER). Additionally, there appears to be a synaptic.junctional complex with some presynaptic dense material as well as some synaptic cleft material between the white block arrows.

Figure 50 shows a rather large profile in close apposition to a labelled migrating motoneuron soma. This profile is filled with small, agran­ ular vesicles 30-35 nm. in diameter. No identifiable synaptic complex can be observed.

In addition to unlabelled profiles, electron microscopy of the area shown in Figure 48 revealed that primary afferent profiles are in contact with labelled migrating motoneuron somata. Figure 51 shows a labelled primary afferent profile, similar to the one at the block arrow in Figure 48, in contact with a labelled migrating motoneuron soma. The labelled profile has a large number of vesicles, ranging in size from 30 to 35 nm. in diameter. No identifiable synaptic complex is evident, although there is some evidence for some type of junction at the white block arrow where the intermembrane gap widens and some cleft material is apparent. DISCUSSION

PRIMARY AFFERENT DISTRIBUTION IN THE ADULT FROG:

The results of the experiments on adult Rana catesbeiana in the present study generally substantiate those of Szekely (1976) on adult

Rana esculenta. Injury filling of dorsal roots in the lumbar spinal cord of adult frogs reveals a separation of primary afferent fibers into a medial division of large diameter fibers and a lateral division of mostly smaller diameter axons. The medial division enters the dor­ sal columns while the lateral division forms a compact bundle in the dorsolateral aspect of the white matter which may be homologous to the tract of Lissauer (Snyder, 1977) in mammals. The primary afferents terminate in two distinct regions of the spinal gray. A dorsal terminal field lies within the area termed the dorsal field by Ebbesson (1976).

It probably corresponds to laminae I through IV of Rexed in the cat rather than the substantia gelatinosa alone as proposed by Szekely

(1976), since both large and small fibers from the dorsal roots inner­ vate this region (see Ebbesson, 1976). A second area of primary afferent termination lies in the deep dorsal horn and the intermediate gray. This region lies within Ebbesson!s lateral field and receives afferents exclusively from collaterals of dorsal root fibers located in the posterior funiculus. These collaterals pass without branching through the dorsal terminal field to ramify within the lateral field.

26 27

Szekely (1976) tentatively identified this region as nucleus proprius and assumed its homology to laminae IV-VI in the mammalian cord. A more detailed comparison of the primary afferent input to the dorsal horn in the anuran frog and mammalian spinal cord will be the subject of a subsequent publication.

Jhaveri and Frank (1981) have recently compared HRP injury fill­ ing of muscle and cutaneous nerves in the brachial spinal cord of adult Rana catesbeiana. They report that the central projections of cutaneous and muscle afferents do not overlap. The dorsal primary afferent field receives input only from cutaneous nerves, while the ventral field (i.e. the lateral field of Ebbesson) receives only muscle afferents. This is clearly different from the pattern observed in the cat and other mammals (Brown, 1982) where muscle and cutaneous afferents overlap considerably in their distribution within the dorsal horn.

In the present study, large caliber primary afferent axons were found to leave the deeper plexus of dorsal root termination and to travel in the lateral intermediate zone to enter the ventral horn, particularly the LMC. Within the LMC, afferent fibers were observed in close apposition to the somata and primary dendrites of LMC moto­ neurons in the dorsal portion of the LMC. No dorsal root fibers were seen to invade the more ventral portion of the LMC. Szekely (1976) has reported similar findings at the light microscopic level. In the present study, the synaptic nature of these appositions was confirmed with electron microscopy. 28

Frank and Westerfield (1982) used HRP to label nerves of the

triceps brachii muscle in the bullfrog. They reported that primary

afferent varicosities, which they presumed to be synaptic, were observed no closer to labelled triceps brachii motoneurons than 20 um, and were typically more than 100 um away from the triceps brachii motoneuron pool.

In the present study, using whole root labelling with HRP, definite monosynaptic axosomatic synapses were observed. However, Frank and

Westerfield's (1982) results are consonant with the present study if one assumes that the organization of the LMC in the brachial cord is similar to that described by Cruce (1974) for the lumbar enlargement in the bullfrog. According to this organization, the triceps brachii muscle would be innervated by motoneurons located in the ventral LMC, since its origin is in the dorsal muscle mass. Thus, on the basis of the observations of Szekely (1976) and the present study, it would not be expected that primary afferents would extend into the ventral LMC where the triceps brachii motoneuron pool is located.

Electron microscopic observation of the adult lumbar cord after dorsal root injury filling with HRP revealed labelled primary afferent terminals in the lateral field, frequently in contact with dendrites, many of which were presumably LMC motoneuron dendrites. More ventrally, fewer dorsal root terminals were encountered. Those profiles that were seen, were found in contact with large primary dendrites and identified

LMC motoneuron somata. All of the labelled synaptic profiles observed contained small, agranular, spherical vesicles with an occasional dense cored vesicle. The synaptic contacts were characterized by a small amount of postsynaptic dense material. 29

These observations substantiate and extend the degeneration ex­ periments of Corvaja and Pellegrini (1975) in the toad. They reported that two days after dorsal root transection degenerating boutons are observed in the intermediate gray and less frequently in the ventral horn. No boutons, however, were observed in contact with motoneuron somata. Sotelo and Grafova (1976) felt that the boutons which

Corvaja and Pellegrini observed degenerating after dorsal root tran­ section are type NFS boutons characterized by spherical vesicles dis­ persed about a central core of neurofilaments and mitochondria.

According to these authors, type NFS boutons synapse upon dendritic profiles and less frequently upon motoneuron somata. Because they have been observed to undergo degeneration after dorsal root tran­ section, Sotelo and Grafova felt these boutons may be homologous to

Conradi's (1969) type M bouton in the cat. However, the type M bouton of Conradi (1969) was always associated with postsynaptic

Taxi bodies and was frequently the postsynaptic target of small type

P boutons which contain pleomorphic vesicles and are never observed to contact motoneuron somata.

Perhaps, the primary afferent terminals in the present study are the same as Sotelo and Grafova's type NFS boutons. This is difficult to say with certainty because neurofilaments are not easily seen in densely labelled terminals. As to whether the labelled terminals are homologous to Conradi's (1969) type M boutons in the cat would be doubtful in that they were never observed to be associated with post­ synaptic Taxi bodies nor were they observed to be postsynaptic to type P boutons. It is possible, however, that the labelled terminals 30

in the present study are homologous to Conradi’s type S boutons which

contain spherical vesicles with an occasional dense-cored vesicle,

contact motoneuron sbmata, and which also disappeared after dorsal

root transection in the cat.

The presence of primary afferent contacts on the somata of

spinal motoneurons has been observed in the cat using degeneration methods (Conradi, 1969; McLaughlin, 1972) and more recently at the

light (Burke, Walmsley, and Hodgson, 1979) and electron microscopic

(Beattie, Bresnahan and Liuzzi, 1980) levels using HRP. Although

quantitative studies are lacking, it is clear that in the cat, as in

frog (present study: Szekely, 1976) most of the primary afferent

contacts on motoneurons are on dendrites rather than somata. Never­

theless, the clear presence of significant numbers of dorsal root

synapses on motoneuron cell bodies in the cat and frog distinguish

these two species from the monkey, where recent studies failed to

find labelled 'axosomatic synapses on motoneurons after tritiated

amino acid injections into the dorsal root ganglia (Ralston and

Daly-Ralston, 1979). It has been suggested (Corvaja and Pellegrini,

1975) that a shift of primary afferent terminals from the distal dendrites of motoneurons to proximal dendrites and somata occurs with progression up the phylogenetic scale. The above considerations

suggest this concept may not be valid.

Type C boutons, similar to those described by Conradi (1969) in

cat, the type L bouton described by Bodian (1975) in monkey and the

type C bouton described by Sotelo and Grafova (1976) in the frog, were

regularly observed in contact with motoneuron somata. These large

terminals characterized by small pleomorphic vesicles, occasional 31 dense-cored vesicles, long synaptic contacts and a postsynaptic sub­ surface cisterna were never labelled by dorsal root injury filling with HRP. This strongly suggests that the type C terminal in the frog is not of dorsal root origin. Rather, the present study supports the conclusion from studies in cat (Conradi, 1969; McLaughlin, 1972) and monkey (Bodian, 1975) that C terminals originate from a source other than the dorsal roots, possibly propriospinal neurons as these remain in the toad ventral horn after removal of suprasegmental input

(Corvaja and Grafova, 1977).

THE LMC IN THE ADULT FROG:

Retrograde injury-filling of LMC motoneurons with HRP in the adult yielded a "Golgi-like" appearance of somata and dendrites much like that attained by Szekely (1976), using cobaltous chloride and like that obtained in cat motoneurons with HRP (Light and Metz, 1978).

Motoneuron morphologies were also comparable to those detailed by

Bregman and Cruce (1980) who analyzed Golgi impregnated LMC motoneurons in the adult bullfrog. It was useful in the present study to use

Bregman and Cruce's designations (D^, D2, D7 and Dg) for specific dendrites and their arborizations in the adult in order to describe the corresponding dendrite in the tadpoles. The Di and D2 dendrites probably receive a convergence of segmental input and suprasegmental input. By virtue of their arborizations, thus, these dendrites may provide a compartment for the integration and processing of input before it reaches the motoneuron soma. A similar organization might be proposed for the D7 dendrite which often has branches which enter the lateral field to receive primary afferent input. Since this 32 dendrite may also receive vestibulospinal input (Bregman and Cruce,

1980), it could provide a compartment wherein vestibulospinal input may be integrated with segmental input. Perhaps the rigid, stereo­

type motor behavior of the frog is consistent not with a lack of integration of input to the motoneuron as suggested by Bregman and

Cruce (1980) but rather with an integration of input on motoneuron dendrites.

It is interesting that anatomically, on the basis of the dis­ tribution of labelled primary afferent axons within the adult LMC, it can be roughly divided into a dorsal and ventral portion. In general, the dorsal portion which receives primary afferent input is composed of more angular, multipolar somata. In contrast, the more ventral LMC contains more typically fusiform somata.

Moreover, the division of the LMC into dorsal and ventral portions is also supported by functional considerations of the muscle groups innervated by neurons in the two portions. For example, within the ninth spinal segment, the knee flexors and hip extensors are innervated, in general, by motoneurons which lie within the more dorsal LMC (Cruce, 1974). In contrast, knee ex­ tensors and hip flexors are innervated by motoneurons which lie within the more ventral LMC (Cruce, 1974).

The adult LMC may also be divided into dorsal and ventral portions on the basis of developmental observations. Of interest in the present study was the presence of two distinct migratory pathways for LMC motoneurons, one dorsal and one ventral. This observation suggests that motoneurons of the dorsal LMC may have a 33 different origin and path of migration than those of the ventral LMC.

In addition, the dorsal pathway passes through the most ventral ex­ tent of the primary afferent distribution and cells in this pathway interact with this afferent input. The ventral pathway, in contrast, does not pass through the primary afferent distribution and thus the cells within this pathway have no opportunity to interact with this afferent input during early development. In the adult, this dis­ tinction is maintained as mentioned above, i.e., the primary afferents distribute to the proximal dendrites and somata of the dorsally po­ sitioned motoneurons in the LMC while not to those of the ventrally positioned ones.

DEVELOPMENT OF PRIMARY AFFERENTS:

The primary afferent input to the lumbar enlargement in the frog develops during the foot paddle (VI-X) and the foot or premetamorphic stages (XI-XVII).

At the beginning of the foot paddle stages, stage VI, primary afferent axons have invaded the primitive dorsal horn. Large diameter fibers along the gray white margin have varicosities along their lengths and at their tips. These varicosities resemble the axonal growth cones described by Mason (1982) following HRP injury-filling of developing retinogeniculate axons in kittens. Additionally, close examination of the dorsal horn reveals numerous smaller diameter fibers among the spherical cells which populate the region. The large and small diameter primary afferent fibers reach nearly to the intermediate gray matter. These results are inconsistent with the recent work of

Forehand and Farel (1982, in press) in the bullfrog, in which they 34 report that labelled dorsal root fibers scarcely impinged upon the dorsal horn until stages VIII and IX. One possible explanation for this inconsistency may be that their longer survival time, two to three days versus 24 hours in the present study, results in the de­ generation of afferent fibers.

By the end of the foot paddle stages, stage X, labelled primary afferent axons have entered the intermediate gray. Moreover, a few labelled fibers are seen in the LMC proper. In addition, primary afferent axon varicosities are observed in apposition to LMC moto­ neuron dendrites. These observations are supported by the electro- physiological results reported by Forehand and Farel (1982, in press).

They showed that the latency of LMC motoneuron response to dorsal root stimulation dropped rapidly between stages V and X after which it gradually reaches adult levels. Thus between stages V and X the anatomical substrate for the myotatic reflex is established. Of further interest, in support of establishment of the myotatic reflex arc closer to the end of the foot paddle stages, Forehand and Farel noted that antidromic stimulation of hindlimb motoneurons could elicit dorsal root potentials only in tadpoles older than stages VIII and IX.

The establishment of the adult pattern of primary afferent dis­ tribution to the lumbar spinal cord occurs during the foot stages.

It is during this time that the hindlimb attains its adult form

(Taylor and Kollros, 1946). By the middle of the hindlimb stages, stages XIII and XIV, the lateral and medial divisions of the dorsal root are clearly identifiable. These divisions may become evident 35 as a consequence of increased myelinization of the large diameter fibers of the medial division rather than the result of a sudden appearance of a certain class of fibers. Coincident with the defin­ ition of lateral and medial divisions of dorsal root fibers is the distribution of two primary afferent terminal fields. This dis­ tinction occurs earlier in the brachial cord at stage X (Jhaveri and Frank, 1980) possibly reflecting a rostrocaudal gradient of maturation within the tadpole spinal cord. By hindlimb stages XIII and XIV, the deep terminal field of primary afferent axons appears significantly more dense than at stage X. Moreover, maturation within this field is probably still taking place at this time as evidenced by the presence of growth cone-like varicosities.

Hindlimb stage XVII marks the end of the foot stages. At this time the hindlimb has attained an adult appearance. In the spinal cord, it appears that the adult distribution of primary afferents has been achieved. Similar observations have been made by Jhaveri and Frank (1980) in the brachial cord of bullfrog tadpoles and recently by Forehand and Farel (1982, in press) in the lumbar cord.

One impression that has not been quantified is the observation of apparently fewer primary afferent varicosities in the LMCs of stage XVII and XVIII tadpoles compared to the LMCs of adult frogs.

Indeed, Forehand and Farel (1982, in press) have shown that the response of hindlimb motoneurons to dorsal root stimulation in stage

XVIII tadpoles is very similar to that recorded in the adult frog.

Further investigation is needed to determine whether the adult number of primary afferent axosomatic contacts on LMC motoneurons is attained 36 during the metamorphic stages. It is during this period that the tadpole changes from a strictly aquatic animal to one capable of hopping and when necessary walking.

Electron microscopy in the present study substantiates the observations made at the light microscopic level. At early foot paddle stages, no labelled primary afferent axons are observed in the LMC. Moreover, the only identifiable synapses in the ventral horn were those on the primary motoneurons which innervate axial muscles used in swimming. This provides evidence suggesting the advanced maturity of the central motor program for swimming involving primary motoneurons (Stenhouwer and Farel, 1980) compared to central mechanisms of hindlimb locomoter activity involving LMC motoneurons.

By hindlimb stage X, electron microscopy reveals apparent immature synaptic profiles in contact with LMC motoneuron somata.

Failure to observe labelled primary afferent terminals in contact with labelled motoneuron somata or dendrites reflects their rare occurrence as evidenced by the light microscopic data.

The first axodendritic synapses of labelled primary afferent terminals on labelled LMC dendrites were observed at stages XIII and

XIV in the lateral field along the gray-white margin just dorsal to the LMC. As one progressed dorsomedially, or into the LMC proper, fewer such synapses were encountered. By the end of the foot stages, hindlimb stage XVII, axodendritic synapses with both elements labelled were more frequently encountered in the lateral field reflecting the mature state of the primary afferents and LMC motoneuron dendrites in the region. Although axodendritic synapses between primary afferent terminals and large dendrites of LMC motoneurons were observed, no

axosomatic synapses with both elements labelled were encountered.

This, again, reflects their scarcity at the light microscopic level.

DEVELOPMENT OF THE LMC MOTONEURONS:

The lateral motor column is first evident as a distinct group

of tightly packed cells during the limb bud stages (I-V) (Beaudoin,

1955). Forehand and Farel (1982, in press) have retrogradely filled

LMC motoneurons by application of HRP to the cut ventral root as early

as hindlimb stage III. They show a tightly packed clump of labelled motoneurons with processes extending into the LF. No processes

apparently extend into the developing lateral field of gray matter.

These results were consistent with electrophysiological data which

showed very early short-latency responses of LMC motoneurons to stim­ ulation of the LF.

In the present study, motoneurons within the LMC were consistently labelled after hindlimb stage X. Stage X LMC motoneurons have primary dendrites that can be assigned designations (D^ etc.) similar to the primary dendrites of adult LMC motoneurons (Bregman and Cruce, 1980) .

At hindlimb stage X the dendrite designated D£ has secondary branches which taper into long fine unbranched processes extending into the LF nearly to the pial surface. These laterally extending branches of the

D2 dendrite reach considerable distances from the somata, much further than branches of the other primary dendrites. Additionally, in the present study, mature appearing synapses on labelled LMC motoneuron dendrites have been observed within the LF at the ultrastructural level in stage X tadpoles. This suggests that LF input to the D2 38 dendrite, i.e., suprasegraental input, probably matures before inputs to the other dendrites. Such an interpretation is consistent with recent electrophysiological data of Forehand and Farel (in press) in bullfrog tadpoles demonstrating that LMC motoneuron response to LF stimulation matures before LMC motoneuron response to dorsal root stimulation. Moreover, it is consistent with the electron micro­ scopic study of Oppenheim, Chu-Wang and Foelix (1975) in the spinal cord of chick embryos showing that mature synapses first appear most laterally in the LMC along the gray-white margin.

By hindlimb stage XIV, the D2 dendrite has fourth and fifth order branches. This branching occurs along the entire length of the dendrite as well as at the tip. This suggests that, as Morest (1969) has observed in developing mammalian CNS, branching can occur along the shaft as well as at the distal end of the dendrite.

The development of the and Dg dendrites parallel the develop­ ment of the primary afferent input to the lateral field. At hindlimb stage X the dendrite has secondary branches extending into the lateral field. By stage XIV it has fifth order branches in the lateral field. The increasing density of primary afferent axons and varicosities coincides with increasing branching of the dorsally ex­ tending dendrites. Since primary afferent axons are present in the lateral field before dendritic branches extend into the region, the data in the present study provides circumstantial evidence for Morest's

(1969) suggestion that afferent axons may influence or induce dendritic branching. Experiments are underway to determine the effect of removal of the afferents by ganglionectomy on the branching of the D-^ dendrites of LMC motoneurons. 39

By hindlimb stage XVII the dendritic arborizations of the LMC motoneurons are very similar to those in the adult. Examination of the LF at this stage reveals hollow swellings or cavitations along the lengths of some labelled dendrites. Perhaps these cavitations are similar to the ones described by Falls and Gobel (1979) along dendrites of neurons within the substantia gelatinosa of kittens. If so, they might represent early stages of dendritic involution and disinte­ gration. This suggests that there may be a pruning of excessive dendrites within the LF. Another possibility which should be mentioned is that this is a response to the axotomy. Further investigation of this phenomenon is needed in this system.

MIGRATING MOTONEURONS:

Retrograde injury filling of larval motoneurons with HRP labelled not only motoneurons extant in the LMC and primary motoneurons, but also migrating motoneurons between the ventricular zone and the LMC.

Similar observations have been reported in embryonic chick trigeminal motor nucleus (Heaton, Moody and Koosier, 1978), in fetal rat spinal cord (Grobstein, 1979) and larval bullfrog spinal cord (Farel and

Bemelsman, 1980; Chu-Wang, Oppenheim and Farel, 1981).

In the present study, two quite distinct pathways of motoneuron migration were observed, particularly at hindlimb stage XIV. There is a dorsal migrating pathway of smaller, more spherical cells which passes through the lateral field and a ventral migrating pathway of larger, more fusiform cells which pass along the ventral gray-white margin just dorsal to the somata of the primary motoneurons. These dorsal and ventral pathways were reported by Farel and Bemelsman (1980) 40

in the lumbar enlargement of larval frogs while apparently only a dorsal migration was observed by Grobstein (1978) in fetal rat spinal

cord.

The differences in the sizes of the dorsal and ventral migrating motoneurons, coupled with the discrete locations of the two pathways strongly suggests that these migrating motoneurons are two separate populations. Moreover, this argument is further strengthened by evidence of two motoneuron types within the adult LMC. It appears that the smaller dorsally migrating motoneurons populate the dorsal

LMC which in general innervates muscles of ventral muscle mass origin

(Cruce, 1974). Similarly, the ventrally migrating motoneurons appear to populate the ventral LMC which in general innervates muscles of dorsal muscle mass origin (Cruce, 1974).

Chu-Wang, et al. (1981) have reported that migrating motoneurons receive mature appearing synapses. They, however, were not able to specify the source of these contacts. In the present study, synapses were also observed on the migrating motoneuron somata, substantiating their findings. In addition, dorsal root injury-filling with HRP revealed that one source of contacts on the dorsally migrating cells is primary afferent axons. Indeed, at the light microscopic level, the dorsal migrating pathway is seen to pass through the ventral most extent of the primary afferent distribution. Moreover, numerous pri­ mary afferent varicosities are observed in close apposition to the somata of these cells, a finding which is corroborated by electron microscopic observations. 41

Thus, a third important difference is evident between the dorsally and ventrally migrating motoneurons. The dorsally mi­ grating motoneurons receive primary afferent contacts during migra­ tion while those in the ventral pathway do not. Of particular interest is the observation that in the adult, it is motoneurons in the more dorsal LMC that receive axosomatic input from primary afferent axons.

Although quantitative analysis was not performed, there appeared to be more axosomatic contacts of primary afferent axons on migrating motoneurons than on motoneurons already within the LMC. Such an observation supports the suggestion of Chu-Wang et al. (1981) that these synapses are transitory in nature. This is the first observa­ tion of a known afferent, which has been shown to normally make synaptic contact on an adult population of neurons, make contact with those neurons during their migration and later break that contact only to be reestablished when the neurons reach their destination. Since the migrating motoneuron axons have not reached the hindlimb (Chu-Wang et al., 1981), one might speculate that the primary afferent axons may be carrying some sort of information to the migrating cells from the developing hindlimb and thus in some way program the connectivity of these cells. Experiments have begun in this laboratory to investigate the influence of the primary afferents and other afferents on moto­ neuron migration and development.

SUMMARY:

The LMC of the bullfrog lumbar enlargement is a discrete, well- demarcated group of motoneurons. Its organization, the location of specific motor pools, has been well documented (Cruce, 1974). 42

Moreover, it has recently been shown (Bregman and Cruce, 1980) that the origins and arborizations of the primary dendrites of LMC moto­ neurons are highly regular.

The present study examined the dorsal root primary afferent distribution, particularly in relation to the LMC in the adult bull­ frog as well as in the developing lumbar spinal cord of larval frogs during hindlimb development. It was found that primary afferent axons form two terminal fields the more ventral of which overlaps portions of the dendritic arborizations of LMC motoneuron primary dendrites. Moreover, labelled primary afferent axons make synaptic contact on these dendrites as well as on the somata of the more dorsal LMC motoneurons. The presynaptic elements of these synaptic contacts are S-type terminals with small, agranular spherical vesicles.

During development, collaterals of primary afferent axons invade the developing gray matter. By hindlimb stage X when five discrete digits can first be distinguished, primary afferent axons have con­ tacted LMC motoneurons. Development of LMC motoneuron dendrites, which in the adult receive primary afferent input, closely parallels the development of the ventral terminal field. This development appears essentially complete by hindlimb stage XVII when the two primary afferent terminal fields are distinct.

Migrating motoneurons on their way to the LMC are retrogradely labelled when HRP is applied to the ventral root between hindlimb stages X and XIV. These migrating motoneurons are observed in two pathways, a dorsal and a ventral migratory pathway. Migrating moto­ neurons receive synaptic contacts. In the dorsal migratory pathway, some of these contacts are primary afferent in origin. The results of the present and recent studies (Forehand and

Farel, 1982, in press) on the development of the spinal cord in larval frogs suggest that the LMC with its established organization provides a good model for in vivo studies of afferent influences on dendritic development. Moreover, the ease with which migrating moto­ neurons are labelled opens up numerous possibilities for in vivo studies of neuronal migration. ILLUSTRATIONS

44 45

FIGURE 1

Drawing of a transverse section through the adult lumbar cord.

HRP labelled primary afferent axons and LMC motoneurons are shown on the right. Dorsal root afferents form medial (MD) and lateral

(LD) divisions. Larger caliber axons of the medial division enter

the dorsal funiculus (DF). Smaller caliber fibers form a bundle in

the dorsolateral funiculus labelled LD. Two discrete terminal fields are observed. The more dorsal terminal field (I) lies within the dorsal field of gray matter. The more ventral terminal field (II) lies in the lateral field of gray matter. Some primary afferent axons leave the ventral terminal field to pass ventrally into the dorsal LMC. The four somata of LMC motoneurons depicted are actual cells which were retrogradely filled with HRP. The dendritic arbor­ izations of the ventral fusiform motoneuron shown are typical of these cells. Branches of dorsolateral dendrites extend into the LF where they reach the pial surface to contribute to the subpial plexus

(arrowheads). Small arrows indicate fascicles of collaterals enter­ ing the dorsal gray. Bar, 200 jum. DORSAL FIELD

LATBtAL FIELD PLATE I

FIGURE 2

Transverse section through the entry zone of an HRP injury-

filled ninth dorsal root in an adult bullfrog. The dorsal funiculus

(DF) contains labelled large diameter axons of the medial division.

Within the dorsolateral funiculus, the axons of the lateral division

contribute to a darkly labelled bundle. The two terminal fields of primary afferent fibers (I and II) are evident. Labelled axons leave

the more ventral terminal field (II) to enter the dorsal LMC (arrows).

Here labelled varicosities are observed in close apposition to moto­ neuron somata (within box). Nissl. Bar, 200 >im.

FIGURE 3

High magnification micrograph of the motoneuron seen in the box in Figure 2. Note the varicosities in close apposition to the primary dendrite and the soma (block arrows). Nissl. Bar, 10 jum.

49

PLATE II

FIGURE 4

Transverse section through the lumbar enlargement of an adult

bullfrog in which the ninth dorsal and ventral roots were injury-

filled with HRP. Motoneurons with more fusiform somata are located

in the more ventral LMC (arrow). Nissl. Bar, 200 pm.

FIGURE 5

Similar section through the lumbar enlargement of an adult

bullfrog. In this section, it can be seen that motoneurons located

in the dorsal LMC tend to be more angular (arrow). All LMC moto­

neurons have dendrites which arborize within the lateral funiculus

(LF). The terminal branches of these dendrites contribute to the

extramedullary neuropil of the subpial plexus (arrowheads).

Nissl. Bar, 200 pm.

51

FIGURE 6

Drawing of a fusiform motoneuron located in the ventral LMC

(insert) of an adult bullfrog. The positions of the primary dendrites and the regions in which their branches arborize are typical of these motoneurons. Three dendrites, those designated

Df, D2 and D 7 (see Bregman and Cruce, 1980) have branches which ramify within the lateral field of gray matter. Dendrite T>2 has its primary arborization in the LF with terminal branches which approach the pial surface (PS). Bar, 50 Insert bar, 200 /am. 6 53

FIGURE 7

Drawing of dorsally situated LMC motoneuron (insert) in an adult bullfrog. The soma is more angular than the ventrally positioned cell in Figure 6. Although complete dendritic arborizations could not be traced, the position of the primary dendrites and their initial branches allowed them to be assigned designations according to the work of

Bregman and Cruce (1980). Here, two dendrites, D-^ and Dg have arbor­ izations that enter the lateral field of gray matter. Bar, 50 /am.

Insert bar, 200 >jm. 7 55

FIGURE 8

High magnification micrograph of an HRP filled motoneuron dendrite within the lateral field in an adult bullfrog. The block arrows mark the positions of varicosities along an HRP filled primary afferent axon which apparently is wrapping around the dendrite. Osmium.

Maraglas embedded. Bar, 10 /urn.

57

PLATE III

FIGURE 9

Type C terminal in contact with an LMC motoneuron soma in the lumbar enlargement of an adult bullfrog. Block arrows indicate postsynaptic subsurface cistern. The ninth dorsal root was injury- filled with HRP. Bar, 0.5 jum.

FIGURE 10

HRP labelled primary afferent terminal in synaptic contact with an LMC motoneuron soma in the lumbar enlargement of an adult bullfrog.

The labelled terminal has small (30-34 nm) , spherical vesicles. A small amount of postsynaptic dense material is evident along the length of the synaptic apposition (arrows). It is important to note that on either side of the labelled terminal, there is an unlabelled type C terminal. Small block arrows indicate postsynaptic subsurface cistern. Block arrows indicate a dense-cored vesicle. Bar, 0.5 jum.

59

PLATE IV

FIGURE 11

Labelled primary afferent terminal in contact with an LMC moto­ neuron dendrite following HRP injury-filling of the ninth dorsal root in an adult bullfrog. The labelled terminal has small spherical vesicles as well as a dense-cored vesicle (block arrow). Bar, 0.5 /urn.

FIGURE 12

HRP labelled primary afferent terminal in contact with an LMC motoneuron dendrite in an adult bullfrog. The terminal is very dark and may be undergoing degeneration. A small amount of postsynaptic dense material is evident along the length of the synaptic contact

(arrows). Two unlabelled synaptic terminals are also in contact with the dendrite. One is a type S with small spherical vesicles and a dense-cored vesicle. The other is a type F terminal with flattened or pleomorphic vesicles. Bar, 0.5 jam.

FIGURE 13

Lightly labelled primary afferent terminal in contact with a small dendrite within the dorsal LMC. The HRP reaction product can clearly be seen to be accumulated on the outer membranes of the vesicles and mitochondrion. A small accumulation of dense material is evident along the postsynaptic membrane (arrows). Bar, 0.5 jum.

61

FIGURE 14

Four stages of hindlirab development in the ranid tadpole

(after Taylor and Kollros, 1946). Hindlimb stage VI marks the beginning of the foot paddle stages. At this time the hindlimb is

a flattened paddle. Stage X marks the end of the foot paddle stages

and is characterized by the appearance of five distinct but primitive digits. Stages XIII and XVII belong to the foot or premetamorphic stages. Stage XIII is a mid-foot stage marked by the differentiation of the web between digits 4 and 5. Stage XVII marks the end of the foot stages, a time when the hindlimb appears in its adult form.

This stage heralds the end of hindlimb development and the beginning of metamorphosis, the transition from an aquatic form to an amphibian form. Bar, 10 mm. n MAX

-*r-,

:~ y .‘

MIX FIGURE 15

Drawing of a transverse section through the lumbar enlargement of a stage VI tadpole. HRP labelled primary afferent axons of the ninth dorsal root are shown on the right. Primary afferent axons are seen in the dorsal funiculus (DF). A few large diameter axons are observed entering the developing dorsal horn along the gray- white margin (block arrows). These axons have varicosities along their lengths and at their tips. More medially, finer caliber axons are seen entering the dorsal horn (small arrows). These too have swellings along their lengths and at their tips. No labelled axons are observed in the presumptive lateral field. The LMC at stage VI is a well demarcated clump of tightly packed fusiform cells in the lateral ventral horn. Bar, 200 *im. LATERAL FIELD 65

PLATE V

FIGURE 16

Transverse section through lumbar spinal cord of a stage VI

tadpole. Ninth dorsal root injury-filling with HRP reveals labelled

axons in the dorsal funiculus (DF). A few large diameter axons

leave the dorsal column and pass ventrally along the gray-white margin (block arrow). These axons have varicosities along their

lengths and at their tips (small arrows). Additionally, numerous

finer caliber fibers are observed between the small spherical cells

of the developing dorsal horn. The LMC is a tightly packed clump

of larger fusiform cells (7.0 Mm in width). The LMC is easily

recognized against a background of small spherical cells which

populate the rest of the developing ventral horn. Bar, 50 M m -

FIGURE 17

High magnification micrograph of rectangular area in Figure 16.

Note the larger diameter axons along the gray-white margin. The

block arrows point to large varicosities (1.5 Mm) along the length

of one of the fibers. Smaller fibers can be seen more medially.

These too have varicosities along their lengths (arrow). Bar, 10 Mm.

67

FIGURE 18

Drawing of a transverse section through the lumbar enlargement

of a stage X tadpole. The result of a ninth dorsal and ventral root

injury fill with HRP is depicted on the right. Larger diameter fibers of dorsal root origin have entered the lateral field. Many of these larger fibers have varicosities along their lengths and at their

tips (arrows). Some primary afferent fibers (PAs) have varicosities

that are observed in close relationship with dendrites of motoneurons in the more dorsal LMC. Retrograde injury-filling of the LMC moto­ neurons shows that these cells have long, fine dendrites extending out into the LF, some approaching the pial surface. PMN, primary moto­ neurons. DF, dorsal funiculus. Arrowheads, subpial plexus.

Bar, 200 ^im.

69

PLATE VI

FIGURE 19

Transverse section through the lumbar enlargement of a stage X

tadpole following HRP injury filling of dorsal and ventral roots.

HRP labelled primary afferent axons fill the more lateral two-thirds

of the developing lateral field. Ventral root injury-filling with

HRP has retrogradely filled LMC motoneurons as well as a large pri­

mary motoneuron located in the ventromedial ventral horn. VZ,

ventricular zone. P A s , primary afferents. DF, dorsal funiculus.

Nissl. Bar, 100 jam.

FIGURE 20

High magnification of the dorsal LMC and the lateral field

dorsal to it. The block arrow indicates two varicosities at the end of a primary afferent axon in the vicinity of a dorsally situated,

labelled LMC motoneuron. These varicosities are smaller (0.5-1.0 jam

in diameter) than those observed along the larger diameter axons in

the dorsal field at stage VI. Nissl. Bar, 10 jum.

FIGURE 21

High magnification micrograph of unlabelled LMC in stage X

tadpole. Dorsal root injury filling with HRP reveals primary afferent axons that enter the LMC (block arrow). Osmium. Maraglas embedded. Bar, 10 um. FIGURE 22

Labelled LMC motoneuron. A labelled primary afferent axon comes in close relationship to dendrites of this cell (white arrows).

Figure 24 is a drawing of this motoneuron which more clearly shows the relationship of the primary afferent axon, to the motoneuron dendrites. Nissl. Bar, 10 ;um.

FIGURE 23

Drawing of a dorsally positioned LMC motoneuron of a stage X tadpole (see insert). The cell has three large dendrites with secondary branches. Additionally, numerous fine dendrites arise at various points from the soma. Three of the dendrites have tentatively been assigned designations of , D2 and Dg according to their point of origin from the cell and the regions where their arborization appears to be developing. The secondary branch of D2 tapers into a fine unbranched process which extends through the LF nearly to the pial surface. Bar, 50 jam. Insert bar, 200 wm.

FIGURE 24

Drawing of the LMC motoneuron shown in Figure 22. Dendrites were tentatively assigned adult designations, D^, D2 and Dg based

on point of origin and region of arborization. Secondary branches

of the D2 dendrite extend into the LF virtually unbranched toward

the pial surface. The smaller arrows indicate a primary afferent

axon which has a number of varicosities in close relationship with

the proximal portions of D-^ and D2. The larger arrows indicate

large swellings along primary afferent axons similar to those

shown in Figure 20. Bar, 50 /urn. Insert bar, 200 /um. D2

,08 76

PLATE VII

FIGURE 25

Electron micrograph of type F synapse on a primary motoneuron

(PMN) of a stage VI tadpole. Bar, 0.5 jum.

FIGURE 26

Electron micrograph of a type C synapse on a primary motoneuron

(PMN) of a stage VI tadpole. Bar, 0.5 jam.

FIGURE 27

Electron micrograph of an early type S synapse (*) on an HRP filled LMC motoneuron in a stage X tadpole. Bar, 0.5 jum.

78

FIGURE 28

Drawing of a transverse section through the lumbar enlargement of a stage XIV tadpole. On the right, HRP labelled primary afferent axons are depicted as is a retrogradely filled LMC motoneuron. Two divisions of primary afferent axons are discerned, a medial division

(MD) of larger diameter fibers in the dorsal funiculus (DF) and a lateral division (LD) of smaller fibers which contribute to a bundle of fibers in the dorsolateral funiculus. Larger diameter axons form fascicles (arrows) which drop through the dorsal horn.

Some contribute to the more ventral terminal field (II) of primary afferents. Additionally, a more dorsal terminal field is discernable

(I) . Bar, 200 jam. ,MD

DORSAL FIELD

LATERAL FIELD

VZ

LMC

28

j 80

PLATE VIII

FIGURE 29

Maraglas embedded transverse section through the root entry zone of a stage XIII tadpole. Both the dorsal and ventral roots have been injury filled with IIRP. Osmium. Bar, 100 jum. DF, dorsal funiculus.

LF, lateral funiculus. VZ, ventricular zone. I, more dorsal terminal field. II, more ventral terminal field. P A s , primary afferents.

FIGURE 30

Higher magnification micrograph of the dorsal quadrant of the section shown in Figure 29. Labelled dorsal root fibers can be seen to contribute to discernable medial and lateral divisions which enter the dorsal funiculus (DF) and a small bundle of fibers in the dorso­ lateral funiculus, respectively. Fascicles of large diameter fibers

(arrows) drop through the superficial dorsal horn to contribute to the deeper more ventral terminal field II of primary afferents in the lateral field of gray matter. This plexus of labelled afferents appears dense with more apparent branching than observed at stage X

(see Figure 19). Additionally, a dorsal terminal field (I) can be clearly discerned. Here, some smaller diameter fibers ramify in the region just ventral to the dorsal funiculus (arrowheads). Osmium.

Bar, 50 jam. 81

FIGURE 31

Higher magnification micrograph of the lower quadrant of the

section shown in Figure 29. The LMC motoneurons send numerous

dendrites into the LF (large arrows). Moreover, in the region just

dorsal to the LMC, dendrites are observed extending into the lateral

field where they have primary afferent axons and varicosities in

close association with them (small arrows). Moreover, varicosities

are still observed along labelled primary afferent axons, some at possible branch points (block arrowhead). Osmium. Bar, 50 pm.

FIGURE 32

Drawing of a dorsally positioned LMC motoneuron in a stage XIV tadpole. The three dendrites designated D^, T)2 and Dg can be compared to those of the stage X LMC motoneurons in Figures 23 and 24. Both the

Dj and D 2 dendrites display higher orders of branching as well as more extensive arborizations. The Dg dendrite extends dorsomedially into the lateral field. P S , pial surface. Bar, 50 um. Insert bar, 200 jum. ,D8

AXON— ► 85

PLATE IX

FIGURE 33

Transverse section through the lumbar enlargement of a stage

XVII tadpole with HRP injury filling of dorsal and ventral roots.

The cord looks essentially like that of the adult (compare to

Figures 2, 4 and 5). Clearly defined medial and lateral divisions of dorsal root afferents are evident. Moreover, discrete dorsal (I) and ventral (II) terminal fields of primary afferent axons are evident

in the dorsal and lateral fields respectively. The LMC motoneurons have laterally extending dendrites which fill the lateral funiculus nearly to the root entry zone dorsally. Additionally, a distinct subpial plexus (white arrows) is seen filled with HRP labelled

terminal dendrites of LMC motoneurons. Nissl. Bar, 50 pm.

FIGURE 34

Higher magnification micrograph of the LMC and the more ventral

terminal field (II) . Labelled LMC motoneuron dendrites can be observed coursing dorsally into the primary afferent plexus within the lateral

field (arrows). D p D2 , Dy and Dg indicate the approximate areas of

the lateral field where branches of these dendrites arborize. Nissl.

Bar, 50 pm.

FIGURE 35

Higher magnification micrograph of the dorsal LMC. Note LMC motoneurons coursing diagonally from lower right to upper left. Primary afferent axon varicosities are frequently found in close apposition to these dendrites (open block arrow). Closed block arrow, branch of D2 dendrite. White arrow, primary afferent in apparent contact with LMC motoneuron soma. Nissl. Bar, 10 pm.

FIGURE 36

High magnification micrograph of the lateral funiculus.

Numerous HRP labelled LMC dendrites are evident. Some dendrites have hollow varicosities (arrow) along their lengths. Terminal dendrites run parallel to the pial surface to contribute to the extramedullary neuropil of the subpial plexus (block arrow). Nissl.

Bar, 50 pm. >••'7 88

PLATE X

FIGURE 37

Transverse section through the LMC of a stage XIV tadpole with

HRP injury filling of the ninth dorsal root. Note the one primary

afferent axon entering the LMC (arrow). Compare with the adult

(Figure 2). Nissl. Bar, 50 pm.

FIGURE 38

Transverse section through the LMC of a stage XV tadpole with

HRP injury filling of the ninth dorsal root. A primary afferent varicosity is evident in the dorsal LMC (arrow). For comparison with adult, see Figure 2. Nissl. Bar, 50 pm.

FIGURE 39

Parasagittal section through LMC in stage XVIII tadpole. Dorsal root injury filling with HRP revealed only a few primary afferent axons in close apposition to LMC motoneuron somata (arrow). Osmium.

Maraglas embedded. Bar, 50 pm.

90

PLATE XI

FIGURE 40

Labelled primary afferent terminal (PA) in synaptic contact with

labelled LMC motoneuron dendrite within the lateral field of a stage

XIV tadpole. Bar, 0.5 pm.

FIGURE 41

Labelled primary afferent terminal (PA) making synaptic contact with a labelled LMC motoneuron dendrite just dorsal to the LMC.

Stage XIV. Bar, 0.5 pm.

FIGURE 42

Labelled primary afferent terminal (PA) synapsing upon a labelled motoneuron dendrite within the lateral field. Stage XVII. Bar, 0.5 pm.

FIGURE 43

Primary afferent terminal (PA) within the dorsal LMC synapsing upon a large labelled LMC motoneuron dendrite in a stage XVII tadpole.

Bar, 0.5 pm.

92

FIGURE 44

Drawings of the dorsal (A) and ventral (B) migratory pathways

as observed in transverse sections through the rostral lumbar enlarge­

ment following HRP injury-filling of the ninth dorsal and ventral roots.

The smaller (Range: 8.0-8.5 pm) dorsally migrating motoneurons (A) pass

through the developing ventral terminal field of primary afferents

(arrows). More laterally, many of the cells appear to be contacted

by primary afferent varicosities. The motoneurons of the ventral

migratory pathway (B) are larger (Range: 11.0-12.0 pm) and appear more

fusiform than those dorsally. Primary afferent axons are not observed

in the vicinity of the ventral migration (see insert). PMN, primary motoneurons. Bars, 50 pm. Insert bars, 200 pm. VZ

LF

PMN

VZ

PMN 94

PLATE XII

FIGURE 45

Transverse section of the dorsal migratory pathway shown in the drawing in Figure 44A. Osmium. Maraglas embedded. Bar, 50 pm.

FIGURE 46

Transverse section of the ventral migratory pathway shown in the drawing in Figure 44B. Osmium. Maraglas embedded. Bar, 50 pm.

96

PLATE XIII

FIGURE 47

Parasagittal section through the rostral lumbar enlargement of a stage XIV tadpole following HRP injury filling of the ninth dorsal and ventral roots. Caudally, a fairly mature LMC is evident while rostrally, small round labelled cells are observed midway between the dorsal and ventral funiculi. Additionally, small round cells are evident ventrally (arrows) just dorsal to the primary moto­ neurons (PMNs) . Osmium. Maraglas embedded. Bar, 50 jum'.

FIGURE 48

High magnification micrograph of area within the rectangle in

Figure 47. The labelled migrating cells are within a field of labelled primary afferent axons and varicosities. Some primary afferent varicosities are in close apposition to the migrating somata (block arrow). Osmium. Maraglas embedded. Bar, 10 pm. ROSTRAL

PMNS 98

PLATE XIV

FIGURE 49

Electron micrograph of a synaptic contact on an HRP filled dorsally migrating motoneuron (MMN). This terminal has pleomorphic vesicles and numerous profiles of smooth endoplasmic reticulum (SER).

Same magnification as Figure 51. Bar, 0.50 pm.

FIGURE 50

Vesicle filled profile in apposition to a labelled migrating motoneuron (MMN) within the dorsal pathway. Same magnification as

Figure 51. Bar, 0.50 pm.

FIGURE 51

Electron micrograph of a labelled primary afferent profile, similar to the one at the arrowhead in Figure 48. This profile is filled with vesicles (30-35 nm.in diameter) and is in close appo­ sition to a labelled migrating motoneuron soma. The white block arrow indicates a possible junction between the profile and the cell. Bar, 0.50 pm. IWT

MMN MMN BIBLIOGRAPHY

Adams, J.C., 1977, "Technical Considerations in the Use of Horseradish Peroxiaas as a Neuronal Marker," Neurosci., 2, 141-145.

Beattie, M., Bresnahan, J. and King, J., 1978, "Ultrastructural Identification of Dorsal Root Primary Afferent Terminals After Antero­ grade Filling with Horseradish Peroxidas," Brain Res., 153,127-134.

Beattie, M.S., Bresnahan, J.C. and Liuzzi, F.J., 1980, "Ultrastructure of Dorsal Root Afferents to the Ventral Horn of the Cat," Neurosci. Absts., 6, 435.

Beaudoin, A.R., 1955, "The Development of Lateral Motor Column Cells in the Lumbosacral Cord in Rana Pipiens," Anat. Rec., 121, 81-96.

Bodian, D., 1968, "Development of Fine structure of Spinal Cord in Monkey Fetuses, II. Pre-Reflex Period to Period of Long Inter- segmental Reflex," J. Comp. Neurol. , 135, 113-166.

Bodian, D . , 1975, "Origin of Specific Synaptic Types in the Motoneuron Neuropil of the Monkey," J. Comp. Neurol., 159, 225-244.

Bregman, B.S. and Cruce, W.L.P., 1980, "Normal Dendritic Morphology of Frog Spinal Motoneurons: A Golgi Study," J. Comp. Neurol. , 193, 1035-1045.

Brown, A.G., 1981, Organization in the Spinal Cord, Springer- Verlag, New York.

Burke, R.E., Walmsley, B. and Hodgson, J.A., 1979, "HRP Anatomy of Group la Afferent Contacts on Alpha Motoneurons," Brain Res., 160, 347-352.

Chu-Wang, I.W., Oppenheim, R.W. and Farel, P.B. , 1981, "Ultrastructure of Migrating Spinal Motoneurons in Anuron Larvae," Brain Res., 213, 307-318.

Conradi, S., 1969, "Ultrastructure of Dorsal Root Boutons on Lumbosacral Motoneurons of the Adult Cat, As Revealed by Dorsal Root Section." Acta Physiol. Scand., suppl., 332, 85-115.

100 101

Corvaja, J. and Grafova, I., 1978, "Vestibulospinal Projections in the Toad, An Experimental Electron Microscopic Study," Neurosci., 3, 619-628.

Corvaja, N. and Pellegrini, M. , 1975, "Ultrastructure of Dorsal Root Projections in the Toad Spinal Cord," Arch. Ital. Biol., 113, 122-148.

Cruce, W.L.R., 1974, "The Anatomical Organization of Hindlimb Motoneurons in Lumbar Spinal Cord of the Frog, Rana Catesbeiana," J. Comp. Neurol. , 153, 59-76.

Ebbesson, S.O.E., 1976, "Morphology of the Spinal Cord," in Neurobiology of the Frog, eds. R. Llinas and W. Precht, Springer- Verlag, New York, 679-706.

Falls, W. and Gobel, S., 1979, "Golgi and EM Studies of the Formation of Dendritic and Axonal Arbors: The Intemeurons of the Substantia Gelatinosa of Rolando in Newborn Kittens, J. Comp. Neurol., 187, 1-18.

Farel, P.B. and Beraelsman, S.E. , 1980, "Retrograde Labeling of Migrating Spinal Motoneurons in Bullfrog Larvae," Neurosci. Lett.. 18, 133-136.

Forehand, C.J. and Farel, P.B. , (In Press), "Spinal Cord Development in Anuran Larvae," J. Comp. Neurol.,

Frank, E. and Westerfield, M., 1982, "Synaptic Organization of Sensory and Motor Neurons Innervating Triceps Brachii Muscles in the Bullfrog," J. Physiol. , 324, 479-494.

Grobstein, C.S., 1979, "The Development of Afferents and Motoneurons in Rat Spinal Cord," Neurosci. Absts., 517, 161.

Heaton, M.B., Moody, S.A. and Koosier, M.E., 1978, "Peripheral Innervation by Migrating Neuroblasts in the Chick Embryo," Neurosci. Lett. , 10, 55-59.

Jhaveri, S. and Frank, E., 1981, "Non-Overlapping Projections of Muscle and Cutaneous Afferents in the Brachial Spinal Cord of Adult Bullfrogs," Soc. Neurosci. Absts., 7, 610.

Joseph, B.S. and Whitlock, D.G., 1968, "Central Projections of Selected Spinal Dorsal Roots in Anuron Amphibians," Anat. Rec., 160, 279-288.

Lasek, R., Joseph, B.S. and Whitlock, D.G., 1968, "Evaluation of Radioautographic Neuroanatomical Tracing Method," Brain Res., 8 , 319-336. 102

Light, A.R. and Metz, C . , 1978, "The Morphology of the Spinal Cord Efferent and Afferent Neurons Contributing to the Ventral Roots of the Cat," J. Comp. Neurol., 179, 819-833.

Mason, C.A., 1982, "Development of Terminal Arbors of Retino- Geniculate Axons in the Kitten, I. Light Microscopical Observations." Neurosci., 7, 541-560.

McLaughlin, B.J., 1972, "Dorsal Root Projections to the Motor Nuclei of the Cat Spinal Cord," J. Comp. Neurol., 144, 429-460.

Morest, D.K., 1969, "The Growth of Dendrites in the Mammalian Brain," Anat. Entwickl-Gesch., 130, 265-305.

Oppenheim, R.W., Chu-Wang, I.W. and Foelix, R.F., 1975, "Some Aspects of Synaptogenesis in the Spinal Cord of the Chick Embryo: A Quantitative Electron Microscopic Study," J. Comp. Neurol., 161, 383-418.

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.

Ralston, H.J. and Daly-Ralston, D., 1979, "Identification of Dorsal Root Synaptic Terminals on Monkey Ventral Horn Cells by Election Microscopy Autoradiography," J. Neurocytol., 8, 151-166.

Silver, M.L., 1942, "The Motoneurons of the Spinal Cord of the Frog," J. of Comp. Neuro. , 77, 1-33.

Simpson, J.I., 1976, "Functional Synaptology of the Spinal Cord," in Neurobiology of the Frog, eds. R. Llinas and W. Precht, Springer-Verlag, New York, 728-749.

Snyder, R., 1977, "The Organization of the Dorsal Root Entry Zone in Cats and Monkeys," J. Comp. Neurol., 174, 47-70.

Sotelo, C. and Grafova, I., 1976, "Ultrastructural Features of the Spinal Cord," in Neurobiology of the Frog, eds. R. Llinas and W. Precht, Springer-Verlag, New York, 707-727.

Stehouwer, D.J. and Farel, P.B., 1980, "Central and Peripheral Controls of Swimming in Anuron Larvae." Brain Res., 218, 131-140.

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. 103

Szekely, G., 1976, "Morphology of Motoneurons and Dorsal Root Fibers in the Frog's Spinal Cord," Brain Res.. 103, 275-290.

Szekely, G. and Czeh, G., 1976, "Organization and Locomotion," in Neurobiology of the Frog, eds. R. Llinas and W. Precht, Springer- Verlag, New York, 765-792.

Taylor, A.C. and Kollros, J.J., 1946, "Stages in the Normal Development of Rana Pipiens Larvae," Anat. Rec. . 94, 7-23.