Brain Research, 108 (1976) 25-36 © Elsevier ScientificPublishing Company, Amsterdam - Printed in The Netherlands
25
STRIATAL A F F E R E N T CONNECTIONS IN T H E T U R T L E ( C H R Y S E M Y S P I C T A ) AS REVEALED BY R E T R O G R A D E A X O N A L T R A N S P O R T OF H O R S E R A D I S H PEROXIDASE*
ANDRE PARENT* * Laboratoire de Neurobiologie, Facultd de Mddecine, Universitd Laval, Qudbec GIK 7P4 (Canada) (Accepted October 13th, 1975)
SUMMARY Horseradish peroxidase (HRP, 3 0 ~ solution, 0.1-0.3 /zl, 72 h) was injected unilaterally into the basal striatum (STR) and the dorsal ventricular ridge (DVR) of adult turtles (Chrysemys picta) in order to demonstrate the cells of origin of some afferents to these telencephalic structures. After selective STR injection, HRP-labeled cells were visualized in the dorsal thalamus and midbrain tegmentum, ipsilaterally. At thalamic level, HRP-positive neurons were located around nucleus rotundus, i.e., mainly within nuclei dorsomedialis anterior, dorsolateralis anterior and less abundantly in nuclei ventralis and reuniens. At midbrain level, a large population of labeled neurons was disclosed within the ventrolateral portion of rostral tegmentum. Other HRP-positive neuronal somata were found scattered throughout the lateral portion of the caudal midbrain tegmentum. In addition, labeled axons were visualized in both peduncles of the lateral forebrain bundle (LFB) after STR injection. The HRP-positive fibers of the dorsal peduncle of the LFB were followed up to the ipsilateral labeled thalamic cells where they appear to arise, whereas the HRP-containing axons of the ventral peduncle were traced down to the lateral midbrain tegmentum where they appear to arborize. Most of the H R P injections into the DVR were confined to the mediodorsal quadrant of the rostral half of the DVR. In such a case, a very large number of HRPpositive cells were disclosed within all thalamic nuclei surrounding nucleus rotundus, ipsilaterally. In addition, numerous labeled neurons were also found in nucleus * Presented in part at the 5th Annual Meeting of the Societyfor Neurosciences,New York, November 2-6, 1975. * * Address for reprints: Laboratoires de Neurobiologie, Pavilion Notre-Dame, 2075, rue de Vitr6, Qudbec, Canada GIJ 5B3.
26 rotundus itself and within nucleus reuniens. No HRP-positive cells were disclosed caudally to the meso-diencephalic junction after DVR injection.
INTRODUCTION
Recently, the use of various histochemical methods has been introduced in the field of comparative neurobiology. The highly specific Falck-Hillarp fluorescence histochemical method for biogenic amines 2, for instance, has been used in a series of comparative studies of the morphological organization of central monoamine neurons in the brain of various vertebrate species including the turtle (Chrysemys picta) 17, 19-21. Such an approach has been of valuable help in the attempt to identify some homologies in the telencephalon of various vertebrateslL These histochemical investigations have shown, among other things, that there exists a striking parallelism between the organization of central monoamine neurons in both the turtle and the rat 21. The basal striatum of the turtle, for instance, has been shown to receive a direct and massive catecholaminergic input from a large population of neurons located within the rostral midbrain tegmentum. This meso-striatal catecholaminergic pathway of the turtle brain appears closely related to the mammalian nigro-striatal dopamine pathway 17. The existence of the nigro-striatal pathway in mammals, first mapped out by means of the Falck-Hillarp histochemical method, has been recently confirmed with the help of a new fiber tracing method using horseradish peroxidase (HRP) as a protein marked °,14. Since our knowledge of the basal ganglia connections in 'lower' vertebrates is still very limited, it appeared to us of interest to use such a new and promising neuroanatomical tool to study the afferent systems to the basal striatum (STR) and the dorsal ventricular ridge (DVR) of Chrysemys picta is. MATERIAL AND METHODS
Twelve adult turtles (Chrysemys picta) have been used in the present study. The HRP histochemical method was adapted from LaVail et al. 11. Intracerebral injections of a 30 ~o solution of HRP (type VI, Sigma Chemical Co., St. Louis) were stereotaxically made within the basal striatum (STR) (first group of six animals) and in the so-called 'dorsal ventricular ridge' (DVR) of Johnston7 (second group of six animals). The enzyme was dissolved in normal saline and administered through a 5 #i Hamilton syringe in quantities ranging from 0.1 to 0.3/zl over a period of 15-20 min. The animals were anesthetized with ketamine hydrochloride (Ketalar, ParkeDavis, 10 mg/kg, i.m.) and their head was maintained in a rat Horsley-Clarke stereotaxic apparatus. Successful unilateral injections in the STR were obtained using the following coordinates: 5.5 mm rostral to the interaural line, 2.0 mm lateral to the midline and 4.5 mm deep from the cortical surface. Unilateral injections involving mainly the dorsomedial quadrant of the DV R (see Results) were obtained
27 using the following coordinates: 3.0 mm rostral to the interaural line, 2.0 mm lateral to the midline and 2.5 mm deep from the cortical surface. The turtles which had a carapace length ranging from 17 to 20 cm were kept at room temperature for 48-72 h after the injections. The animals were then killed by transcardiac perfusion at room temperature with a paraformaldehyde (2 ~o) and gluteraldehyde (1 9/0) mixture in 0.1 M phosphate buffer (pH 7.3). In most cases, however, the perfusion with the fixative was preceded by perfusion with normal saline in order to remove as many erythrocytes as possible. After the perfusion, the brains were removed from the skull, placed in a fresh fixative for 2-4 h and then rinsed overnight at 4 °C in 0.1 M phosphate buffer containing 5 ~ sucrose. The brains were then sectioned on a freezing microtome. Sections 50-80/,m thick were collected serially in a 16-compartment plastic box containing 5 ~o sucrose in 0.1 M phosphate buffer (pH 7.3, 4 °C). The sections were then transferred to a solution containing 0.05~ 3,3'-diaminobenzidine tetrahydrochloride (Sigma) in Tris.HC1 buffer (pH 7.6). For demonstrating the presence of peroxidase the sections were incubated for a further 30 min at 40 °C in a medium containing the same concentration of diaminobenzidine Tris.HC1 but with the addition of 0.03 ml of 30 hydrogen peroxide per 100 ml, according to Graham and Karnovsky 5. Following this, the sections were washed in distilled water and mounted on slides. The microscopy was performed with a Leitz microscope equipped with a dark-field condenser. RESULTS
In the present study the striatal HRP injections are all confined to the lateral (or somatic) portion without involving the medial (or olfactive) subdivision o f the basal striatum (STR) 4,15. The injection sites varying in size with the amount injected, have their largest extent in the middle region of the STR and spare the rostral tip of that structure (Figs. 1 and 2A). In the vicinity of the injection sites a few striatal and lateral (piriform) cortex neurons can be visualized. These neurons and their proximal processes are stained in their entirety and therefore appear to have been labeled by direct uptake of the enzyme (see ref. 8). In addition, numerous labeled fibers occur at the fringes of the injection loci. Some of these fibers can be followed for a short distance up to the dorsal cortex (via the pallial thickening) and within the core of the dorsal ventricular ridge (Fig. 1). Other HRP-containing fibers can be traced for a much longer distance within the lateral forebrain bundle (see below). After selective unilateral injection of HRP within the STR of Chrysernyspicta numerous HRP-positive neuronal somata are present in the dorsal thalamus and midbrain tegmentum, ipsilaterally. In contrast to the HRP-containing neurons occurring in the vicinity of the injection site, the thalamic and midbrain labeled neurons show a typical granular appearance considered to be a distinguishing feature of cells labeled by retrograde transport of HRP from a distant injection site (see ref. 8). At the thalamic level, cells containing granular reaction product are found in all
28 2mm
IF
IL~~~p, ~ E
'
I
Fig. 1. Camera lucida drawings of transverse half sections through the brain of Chrysemys picta (case No. 210/2) showing the effect of typical striatal injections (hatched areas) of HRP. The drawings are set out in a rostro-caudal order from A to I. The full circles indicate neuronal somata that have been labeled following retrograde axonal transport of HRP. Labeled fibers in the vicinity of the injection site and within the lateral forebrain bundle are indicated as thin filaments. Abbreviations used in this and subsequent figures are dc: dorsal cortex; dla: nucleus dorsolateralis anterior; dma: nucleus dorsomedialis anterior; dp: lateral forebrain bundle, dorsal peduncle; dvr: dorsal ventricular ridge; h: hypothalamus; ha: habenular nuclei; ip: interpeduncular nucleus; lc: lateral (piriform) cortex; lfb: lateral forebrain bundle; mc: medial (hippocampal) cortex; oc: optic chiasma; ot: optic teetum; pt: pallial thickening; re: nucleus reuniens; rt: nucleus rotundus; s: septum; str: basal striatum; vp: lateral forebrain bundle, ventral peduncle; III: oculomotor nerve root fibers.
nuclei surrounding nucleus rotundus, i.e., mainly within nuclei dorsomedialis anterior, dorsolateralis anterior and much less abundantly in nuclei ventralis and reuniens (according to Papez's nomenclaturel6). N o t all the neurons in each o f these nuclei are labeled and most o f the labeled cells are f o u n d in the central portion o f the nuclei (Figs. 1 2B, C). In one case in which the S T R injection site was much larger, also involving the ventral part of the pallial thickening and the rostral tip o f the D V R , a few labeled cells were disclosed within nucleus geniculatus lateralis, p a r s dorsalis. In all 6 turtles, however, no HRP-positive cells o c c u r in nucleus rotundus itself after S T R injection (Figs. 1 and 2C). Contralaterally to the injection site the whole dorsal thalamus is devoid o f labeled cells. At the midbrain level a large population o f labeled neuronal somata is found in the ventrolateral portion of the rostral midbrain tegmentum on the injection side (Figs. 1 and 3A, B). M o s t o f these HRP-positive neurons are lying immediately caudal to the o c u l o m o t o r nerve r o o t fiber level (Fig. 1). The H R P granules are evenly distributed within the cytoplasm and invade the proximal dendrites o f these midbrain neurons (Fig. 3A). M o r e caudally, at the level o f the interpeduncular nucleus, a small
Fig. 2. A: microphotograph of a transverse section through the telencephalon showing the largest extent of a typical striatal HRP injection site (case No. 210/2). B and C: microphotographs (darkfield illumination) of transverse sections through the dorsal thalamus showing HRP-positive cells within the core of nucleus dorsomedialis anterior (B) and in nucleus dorsolateralis anterior (C), ipsilaterally to the striatal injection. D and E: microphotographs (dark-field illumination) of transverse sections through the rostra1 diencephalon (level D in Fig. 1) illustrating the numerous labeled fibers that occur in the lateral forebrain bundle after striatal HRP injection. The lateral forebrain bundle is shown in its more medial extension in the hypothalamus (D) and in its lateral extension in the striatum @). Magnifications indicated.
30 number of HRP-positive neurons are scattered throughout the lateral midbrain tegmentum (Fig. 1). No HRP-containing neurons are seen caudally to the isthmus of the hindbrain after striatal injection. As mentioned above, numerous labeled fibers can be observed near the striatal injection site, some of them invading caudally the lateral forebrain bundle (LFB). HRP-containing fibers are present in both the dorsal and ventral peduncles of the LFB (Figs. 1, 2D, E and 3C). The HRP-positive fibers of the dorsal peduncle can be traced as far as the level of the dorsal thalamic labeled neurons where they appear to arise. They reach the perirotundus nuclei by spreading out both around the surface and through the rostral half of nucleus rotundus. The ventral peduncle fibers can be followed down to the lateral border of the midbrain tegmentum where they appear to arborize. As they penetrate the midbrain tegmentum the labeled fibers of the
Fig. 3. A: microphotograph (dark-field illumination) showing medium-sized labeled neurons located in the rostral midbrain tegmentum, ipsilaterally to the HRP striatal injection. B: microphotograph (dark-field illumination) of a transverse section through the large population of labeled cells (level H in Fig. 1) that occurs in the ventrolateral portion of the rostral midbrain tegmentum, ipsilaterally to the striatal injection. In the left half of the figure, transversally cut labeled fibers can be observed (asterisk) in the terminal field of the 'tractus strio-tegmentalis ventralis'. C: microphotographs (dark-field illumination) of a transverse section through the diencephalon showing the ventral and dorsal peduncles of the lateral forebrain bundle, both containing numerous HRP-positive fibers after striatal injection. Magnifications indicated.
31 ventral peduncle progressively break out along the lateral border. In frontal sections, the very small H R P granules encountered within this portion of the labeled fibers are diffusely distributed in the lateral tegmentum (Figs. 1 and 3B). The H R P granules gradually disappear as we reach the caudal half portion of the midbrain tegmentum. There appears to be no direct continuity between the HRP-containing cell bodies of the midbrain and the labeled fibers of the ventral peduncle, in contrast to the labeled fibers of the dorsal peduncle which merge in thalamic cells.
Dorsal ventricular ridge injections The H R P injection sites in the DVR are all mainly confined to the dorsomedial quadrant of the rostral half of this structure (Figs. 4 and 5A). In most cases, some H R P material appears to have also diffused in the lateral ventricle since a light diffusion zone can be noticed more especially around the basis of the lateral ventricle (Fig. 4). In all the animals of this group, however, no H R P material can be found further than 100-150/am from the ependymal surface. Within this diffusion zone a few weakly labeled neuronal somata can be visualized along the ependymal border of both the STR and the septum. After DVR injections a very large number of HRP-positive neurons occurs in the dorsal thalamus, ipsilaterally. Labeled cells are found in most dorsal thalamic nuclei including nucleus rotundus. They are more abundant in the rostral half of the nucleus rotundus. The number of labeled neurons present in all nuclei surrounding nucleus rotundus, i.e., nuclei dorsomedialis anterior, dorsolateralis anterior, ventralis and reuniens, was much greater than after striatal injection. At the level of nucleus 2mm
dma
~.~./
\V(I'..'
:
Fig. 4. Camera lucida drawings of transverse half sections through the forebrain of Chrysemyspieta (case No. 221/1) illustrating the effect of typical dorsal ventricular ridge injections (hatched areas) of HRP. The drawings are set out in a rostro-eandal order from A to F. The full circles indicate neuronal somata that have been labeled following retrograde axonal transport of HRP.
32
Fig. 5. A: microphotograph of a transverse section through the rostral half of the dorsal ventricular ridge showing the largest extent of a typical H R P injection site in that structure (case No. 221/I). B: microphotograph (dark-field illumination) of a transverse section through the dorsal thalamus that illustrates the numerous labeled cells present within and around the rostral half of nucleus rotundus, ipsilaterally (right side) to a dorsal ventricular injection. C: microphotograph (dark-field illumination) of a transverse section through nucleus reuniens showing the large number of HRPpositive cells that occurs in this nucleus ipsilaterally to a dorsal ventricular ridge injection. D : microphotograph (dark-field illumination) that illustrates some of the large neurons of nucleus rotundus whose cell body and proximal processes are completely filled with H R P material. These neurons appear to have been labeled by a direct uptake of the enzyme by the damaged portions of their axon (see text). Magnifications indicated.
33 dorsomedialis anterior, for instance, the HRP-positive cells are extremely abundant and closely packed (Figs. 4 and 5B). The nucleus reuniens also contains a large population of labeled cells that prolongs itself caudally within the pretectal area (Fig. 5C). In contrast to the results obtained after the STR injection, no HRP-containing neurons are present caudally to the meso-diencephalic junction after HRP injection into the DVR. In one case, however, in which the DVR injection site did also involve the most dorsal portion of the STR, weakly labeled neuronal somata were outlined in the midbrain tegmentum. These labeled neurons are identical in type and size to those observed in the same midbrain areas after HRP injection confined to the STR. The amount of HRP granules in these midbrain neurons is very small, however, in comparison to the strong labeling resulting from a massive STR injection. In this peculiar case, a few labeled fibers are found in the vicinity of that portion of the DVR injection site reaching the STR. Some of the reactive fibers can be followed back to large neurons of nucleus rotundus that appear completely filled with HRP material, thus contrasting with other labeled cells of that area displaying the typical granular appearance (Fig. 5D). Therefore, it may be presumed that these large nucleus rotundus neurons have been labeled by a direct uptake of the enzyme by the injured segment of their axon coursing through the injection area. All the HRP-positive neurons described in the present study after HRP injection either in the STR or DVR have been found ipsilaterally to the injection. It is worth mentioning, however, that a few labeled cells have been observed within the contralateral perirotundus area in one case of DVR injection. These neurons were lying along the ependymal surface at the level of nucleus dorsomedialis anterior. They were very weakly labeled in comparison to the ipsilateral dorsal thalamic neurons. The relevance of this finding remains to be evaluated since the injection site in this peculiar case was apparently identical to other DVR injection sites that did not show such a contralateral labeling. DISCUSSION
The present results confirm the existence in the turtle (Chrysemyspicta) of a midbrain-striatal connection that has been previously mapped out by means of the Falck-Hillarp histochemical method for monoamines. The large population of HRPpositive neurons found within the rostral half of the midbrain tegmentum after STR injection is strikingly similar to the large group of catecholamine-containing cells disclosed in the same midbrain area of the turtle20, ~1. In both cases the neuronal somata which are of the same type and size occupy the ventrolateral portion of the rostral midbrain tegmentum. The catecholamine-containing neurons of that area have been shown to project via the medial forebrain bundle upon the STR. This midbrain-striatal catecholaminergic pathway of the turtle brain has been tentatively related to the mammalian nigro-striatal dopaminergic pathway 17. On the other hand, a small number of elongated HRP-positive cells are also scattered throughout the lateral portion of the caudal half of the midbrain tegmentum
34 after STR injection. These labeled cells are strikingly similar in size and shape to the 5-hydroxytryptamine (serotonin)-containing neurons disclosed in the same midbrain area by means of the Falck-Hillarp method 20. Numerous other serotonin-containing neurons occur along the raphe region of the turtle lower brain stem but none of these cells have been labeled after H R P injection in the STR. Therefore, it may be postulated that most serotonin-type axon terminals present in the STR of Chrysemys picta arise from the serotonin-containing neuronal somata lying at the level of the caudal midbrain tegmentum. This condition would resemble closely the situation in mammalian brain in which most telencephalic serotonin axon terminals appear to arise from nuclei raphe dorsalis and raphe medialis (or centralis superior) both lying at caudal midbrain or isthmus level (see ref. 12). The present study also provides important data with regard to the exact origin of the dorsal thalamic projection to the turtle STR recently disclosed by means of the silver anterograde degeneration technique 6. The thalamo-striatal afferent in Chrysemys picta seems to originate from neurons belonging to various dorsal thalamic nuclei surrounding nucleus rotundus, i.e., nuclei dorsomedialis anterior, dorsolateralis anterior and, to a lesser extent, nucleus ventralis, which all contain HRP-labeled cells after STR injection. On such a basis the perirotundus nuclei may be compared to the mammalian intralaminar nuclei which are the major thalamic source of the thalamo-striatal afferent. Owing to the fact that no HRP-positive cells were detected within nucleus rotundus itself after STR injection, it may be presumed that this nucleus does not contribute significantly to the turtle thalamo-striatal connection. In contrast the nucleus rotundus, which is the major thalamic target of the optic tectum, appears to project to the DVR of the turtle. Numerous HRP granules are indeed found within the neurons of this nucleus after DVR injection as shown in this study. These results are in apparent contradiction with the fact that ablations of the DVR do not produce retrograde cell changes in nucleus rotundus whereas lesions within the STR induced marked retrograde changesg,2L Such a discrepancy may be explained by the fact that STR lesions inevitably damage the more proximal portion of nucleus rotundus axons which course through the STR en route to the DVR 6. Destruction involving only the terminal arborizations of nucleus rotundus axons in the DVR could be insufficient in order to induce retrograde cell alterations. On the other hand, a large amount of neurons of the perirotundus area (including nucleus reuniens) also show a positive H R P reaction after DVR injection. Therefore, the perirotundus area seems to project to both the STR and the DVR. Whether both of these projections arise from the same or different neuronal populations in the perirotundus nuclei remains to be evaluated. It must also be mentioned that the nucleus reuniens, that is heavily labeled after DVR injection, has been recently reported to receive auditory fibers and to send its axons to the DVR in the lizard Iguana 3. The lateral forebrain bundle (LFB) in Chrysemys picta as well as in other reptiles is usually divided into dorsal and ventral peduncles. In the past, the dorsal peduncle has been associated with various 'thalamo-striatal' pathways whereas the ventral peduncle was thought to be composed mainly of 'strio-tegmental' pathways 1.
35 The present study has shown that numerous HRP-contaning fibers occur at the level of both peduncles of the LFB after STR injection. On the one hand, HRP material appears to migrate retrogradely from the STR injection site to the cells of the perirotundus nuclei via the dorsal peduncle fibers, thus confirming the existence of a thalamo-striatal system among the fibers of the dorsal peduncle. On the other hand, HRP material appears to migrate anterogradely within the ventral peduncle fibers down to the ventrolateral border of the midbrain tegmentum. These labeled fibers have a course that corresponds closely to the so-called 'tractus strio-tegmentalis ventralis' terminating in large part along the ventrolateral border of the midbrain tegmentum1. The labeled fibers present in the ventral peduncle of the LFB after STR injection could represent a reptilian equivalent of the well-documented strio-nigral pathway of mammals. In mammalian brain, the strio-nigral fibers also penetrate the midbrain tegmentum through a latero-ventral course 13. The presence of numerous labeled fibers in the LFB of Chrysemys picta after STR injection could be the result, at least in part, of a direct uptake of HRP material by axons that could have been damaged by the tip of the canula (see Fig. 2A). It is also possible that in the brain of reptiles, as in other poikilotherms, the slower metabolic rate may facilitate the mapping of the nervous fibers with the HRP technique. The HRP granules could migrate along the axons of the turtle brain at a much lower rate than the migration rate estimated for mammalian central nervous fibers. In fact, 72 h was found to be the ideal survival period in order to obtain an optimal retrograde labeling of Chrysemys picta neurons in comparison to the 2a, "!8 h period disclosed for mammalian brains TM. Very few labeled neurons were obtained even 48 h after successful HRP injection in the turtle brain. The effect of survival periods longer than 72 h on the intensity of the HRP retrograde labeling remains, however, to be investigated. In any event, the results of this study suggest that the HRP histochemical method may be a new and promising approach to the study of the brains of lower vertebrates. ACKNOWKEDGEMENTS The author would like to express his deep gratitude to Dr. F. Dord, Mr. R. Boucher and Mr. R. Marchand for their valuable technical assistance at various stages of this study. The histological work was performed by Mrs. D. Picard. This study was supported by Grant MT-732 of the Medical Research Council of Canada. The author is holder of a Scholarship of the Medical Research Council.
REFERENCES 1 ARIi~NS-KAPPERS,C. U., HUBER, G. C., AND CROSBY,E. C., The Comparative Anatomy of the Nervous System of Vertebrates including Man, Hafner, New York, 1960. 2 FALCK, B., HILLARP, N.-A., THmME,G., AND TORP, A., Fluorescence of catecholamines and related compounds condensed with formaldehyde,J. Histochem. Cytochem., 10 (1962) 348-354.
36 3 FOSTER, R. E., AND PEELE, T. L., Thalamotelencephalic auditory pathways in the lizard (Iguana iguana), Anat. Rec., 181 0975) 530 (abstract). 4 GOEDBY, F., AND GAMBLE, H. J., The reptilian cerebral hemispheres, Biol. Rev., 32 (1957) 383420. 5 GRAHAM, R. C., JR., AND KARNOVSKY, M.J., The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique, J. Histochem. Cytochem., 14 0966) 291-302. 6 HALL, W. C., AND EBNER, F. F., Thalamotelencephalic projections in the turtle (Pseudemys scripta), J. comp. Neurol., 140 (1970) 101-122. 7 JOHNSTON,J. B., The cell masses in the forebrain of the turtle, Cistudo carolina, J. comp. Neurol., 25 (1915) 393-468. 8 JONES, E. G., AND LEAV1TT,R. Y., Retrograde axonal transport and the demonstration of nonspecific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei in the rat, cat and monkey, J. comp. Neurol., 154 (1974) 349-378. 9 KRUGER,L., AND BERKOW1TZ,E. C., The main afferent connections of the reptilian telencephalon as determined by degeneration and electrophysiological methods, J. comp. Neurol., 115 0960) 125-141. 10 KUYPERS, H. G. J. M., K1EVIT, J., AND GROEN-KLEVANT,A. C., Retrograde axonal transport of horseradish peroxidase in the rat's forebrain, Brain Research, 58 (1973) 470-477. 11 LAVAIL, J. H., WINSTON, K. R., AND TISH, A., A method based on retrograde intraaxonal transport of protein for identification of cell bodies of origin of axons terminating within the CNS, Brain Research, 58 (1973) 470-477. 12 LORENS, S. A., AND GULDBERG, H. C., Regional 5-hydroxytryptamine distribution following selective midbrain raphae lesions in the rat, Brain Research, 78 (1974) 45-56. 13 NAUTA, W. J. H., AND MEHLER, W. R., Projections of the lentiform nucleus in the monkey, Brain Research, I (1966)3-42. 14 NAUTA, H. J. W., PRITZ, M. B., AND LASEK, R. J., Afferents to the rat caudoputamen studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method, Brain Research, 67 0974) 219-238. 15 NORTHCUTT, R. G., The Telencephalon of the Western Painted Turtle (Chrysemys picta belli), Illinois Biological Monographs, University of Illinois Press, Urbana, l[1., 1970. 16 PAPEZ, J. W., Thalamus of turtles and thalamic evolution, J. comp. Neurok, 61 (1935) 433-475. 17 PARENT, A., Demonstration of a catecholaminergic pathway from the midbrain to the strioamygdaloid complex in the turtle (Chrysemys picta), J. Anat. (Lond.), 114 (1973) 379-387. 18 PARENT,A., AND BOUCHER,R., Horseradish peroxidase study of the striatal afferent connections in the turtle (Chrysemyspicta), Proc. Soc. Neurosci., 5 (1975) 195 (abstract). 19 PARENT, A., AND OLIVIER, A., Comparative histochemical study of the corpus striatum, J. Hirnforseh., 12 (1970) 73-81. 20 PARENT,A., AND POIRIER,L. J., Occurrence and distribution of monoamine-containing neurons in the brain of the painted turtle (Chrysemyspicta), J. Anat. (Lond.), 110 (1971) 81-89. 21 PARENT, A., AND POITRAS, D., The origin and distribution of catecholaminergic axon terminals in the cerebral cortex of the turtle (Chrysemys picta), Brain Research, 78 (1974) 345-358. 22 POWELL, Z. P. S., AND KRUGER, L., The thalamic projection upon the telencephalon in Lacerta viridis, J. Anat. (Lond.), 94 (1960) 528-542.
25
STRIATAL A F F E R E N T CONNECTIONS IN T H E T U R T L E ( C H R Y S E M Y S P I C T A ) AS REVEALED BY R E T R O G R A D E A X O N A L T R A N S P O R T OF H O R S E R A D I S H PEROXIDASE*
ANDRE PARENT* * Laboratoire de Neurobiologie, Facultd de Mddecine, Universitd Laval, Qudbec GIK 7P4 (Canada) (Accepted October 13th, 1975)
SUMMARY Horseradish peroxidase (HRP, 3 0 ~ solution, 0.1-0.3 /zl, 72 h) was injected unilaterally into the basal striatum (STR) and the dorsal ventricular ridge (DVR) of adult turtles (Chrysemys picta) in order to demonstrate the cells of origin of some afferents to these telencephalic structures. After selective STR injection, HRP-labeled cells were visualized in the dorsal thalamus and midbrain tegmentum, ipsilaterally. At thalamic level, HRP-positive neurons were located around nucleus rotundus, i.e., mainly within nuclei dorsomedialis anterior, dorsolateralis anterior and less abundantly in nuclei ventralis and reuniens. At midbrain level, a large population of labeled neurons was disclosed within the ventrolateral portion of rostral tegmentum. Other HRP-positive neuronal somata were found scattered throughout the lateral portion of the caudal midbrain tegmentum. In addition, labeled axons were visualized in both peduncles of the lateral forebrain bundle (LFB) after STR injection. The HRP-positive fibers of the dorsal peduncle of the LFB were followed up to the ipsilateral labeled thalamic cells where they appear to arise, whereas the HRP-containing axons of the ventral peduncle were traced down to the lateral midbrain tegmentum where they appear to arborize. Most of the H R P injections into the DVR were confined to the mediodorsal quadrant of the rostral half of the DVR. In such a case, a very large number of HRPpositive cells were disclosed within all thalamic nuclei surrounding nucleus rotundus, ipsilaterally. In addition, numerous labeled neurons were also found in nucleus * Presented in part at the 5th Annual Meeting of the Societyfor Neurosciences,New York, November 2-6, 1975. * * Address for reprints: Laboratoires de Neurobiologie, Pavilion Notre-Dame, 2075, rue de Vitr6, Qudbec, Canada GIJ 5B3.
26 rotundus itself and within nucleus reuniens. No HRP-positive cells were disclosed caudally to the meso-diencephalic junction after DVR injection.
INTRODUCTION
Recently, the use of various histochemical methods has been introduced in the field of comparative neurobiology. The highly specific Falck-Hillarp fluorescence histochemical method for biogenic amines 2, for instance, has been used in a series of comparative studies of the morphological organization of central monoamine neurons in the brain of various vertebrate species including the turtle (Chrysemys picta) 17, 19-21. Such an approach has been of valuable help in the attempt to identify some homologies in the telencephalon of various vertebrateslL These histochemical investigations have shown, among other things, that there exists a striking parallelism between the organization of central monoamine neurons in both the turtle and the rat 21. The basal striatum of the turtle, for instance, has been shown to receive a direct and massive catecholaminergic input from a large population of neurons located within the rostral midbrain tegmentum. This meso-striatal catecholaminergic pathway of the turtle brain appears closely related to the mammalian nigro-striatal dopamine pathway 17. The existence of the nigro-striatal pathway in mammals, first mapped out by means of the Falck-Hillarp histochemical method, has been recently confirmed with the help of a new fiber tracing method using horseradish peroxidase (HRP) as a protein marked °,14. Since our knowledge of the basal ganglia connections in 'lower' vertebrates is still very limited, it appeared to us of interest to use such a new and promising neuroanatomical tool to study the afferent systems to the basal striatum (STR) and the dorsal ventricular ridge (DVR) of Chrysemys picta is. MATERIAL AND METHODS
Twelve adult turtles (Chrysemys picta) have been used in the present study. The HRP histochemical method was adapted from LaVail et al. 11. Intracerebral injections of a 30 ~o solution of HRP (type VI, Sigma Chemical Co., St. Louis) were stereotaxically made within the basal striatum (STR) (first group of six animals) and in the so-called 'dorsal ventricular ridge' (DVR) of Johnston7 (second group of six animals). The enzyme was dissolved in normal saline and administered through a 5 #i Hamilton syringe in quantities ranging from 0.1 to 0.3/zl over a period of 15-20 min. The animals were anesthetized with ketamine hydrochloride (Ketalar, ParkeDavis, 10 mg/kg, i.m.) and their head was maintained in a rat Horsley-Clarke stereotaxic apparatus. Successful unilateral injections in the STR were obtained using the following coordinates: 5.5 mm rostral to the interaural line, 2.0 mm lateral to the midline and 4.5 mm deep from the cortical surface. Unilateral injections involving mainly the dorsomedial quadrant of the DV R (see Results) were obtained
27 using the following coordinates: 3.0 mm rostral to the interaural line, 2.0 mm lateral to the midline and 2.5 mm deep from the cortical surface. The turtles which had a carapace length ranging from 17 to 20 cm were kept at room temperature for 48-72 h after the injections. The animals were then killed by transcardiac perfusion at room temperature with a paraformaldehyde (2 ~o) and gluteraldehyde (1 9/0) mixture in 0.1 M phosphate buffer (pH 7.3). In most cases, however, the perfusion with the fixative was preceded by perfusion with normal saline in order to remove as many erythrocytes as possible. After the perfusion, the brains were removed from the skull, placed in a fresh fixative for 2-4 h and then rinsed overnight at 4 °C in 0.1 M phosphate buffer containing 5 ~ sucrose. The brains were then sectioned on a freezing microtome. Sections 50-80/,m thick were collected serially in a 16-compartment plastic box containing 5 ~o sucrose in 0.1 M phosphate buffer (pH 7.3, 4 °C). The sections were then transferred to a solution containing 0.05~ 3,3'-diaminobenzidine tetrahydrochloride (Sigma) in Tris.HC1 buffer (pH 7.6). For demonstrating the presence of peroxidase the sections were incubated for a further 30 min at 40 °C in a medium containing the same concentration of diaminobenzidine Tris.HC1 but with the addition of 0.03 ml of 30 hydrogen peroxide per 100 ml, according to Graham and Karnovsky 5. Following this, the sections were washed in distilled water and mounted on slides. The microscopy was performed with a Leitz microscope equipped with a dark-field condenser. RESULTS
In the present study the striatal HRP injections are all confined to the lateral (or somatic) portion without involving the medial (or olfactive) subdivision o f the basal striatum (STR) 4,15. The injection sites varying in size with the amount injected, have their largest extent in the middle region of the STR and spare the rostral tip of that structure (Figs. 1 and 2A). In the vicinity of the injection sites a few striatal and lateral (piriform) cortex neurons can be visualized. These neurons and their proximal processes are stained in their entirety and therefore appear to have been labeled by direct uptake of the enzyme (see ref. 8). In addition, numerous labeled fibers occur at the fringes of the injection loci. Some of these fibers can be followed for a short distance up to the dorsal cortex (via the pallial thickening) and within the core of the dorsal ventricular ridge (Fig. 1). Other HRP-containing fibers can be traced for a much longer distance within the lateral forebrain bundle (see below). After selective unilateral injection of HRP within the STR of Chrysernyspicta numerous HRP-positive neuronal somata are present in the dorsal thalamus and midbrain tegmentum, ipsilaterally. In contrast to the HRP-containing neurons occurring in the vicinity of the injection site, the thalamic and midbrain labeled neurons show a typical granular appearance considered to be a distinguishing feature of cells labeled by retrograde transport of HRP from a distant injection site (see ref. 8). At the thalamic level, cells containing granular reaction product are found in all
28 2mm
IF
IL~~~p, ~ E
'
I
Fig. 1. Camera lucida drawings of transverse half sections through the brain of Chrysemys picta (case No. 210/2) showing the effect of typical striatal injections (hatched areas) of HRP. The drawings are set out in a rostro-caudal order from A to I. The full circles indicate neuronal somata that have been labeled following retrograde axonal transport of HRP. Labeled fibers in the vicinity of the injection site and within the lateral forebrain bundle are indicated as thin filaments. Abbreviations used in this and subsequent figures are dc: dorsal cortex; dla: nucleus dorsolateralis anterior; dma: nucleus dorsomedialis anterior; dp: lateral forebrain bundle, dorsal peduncle; dvr: dorsal ventricular ridge; h: hypothalamus; ha: habenular nuclei; ip: interpeduncular nucleus; lc: lateral (piriform) cortex; lfb: lateral forebrain bundle; mc: medial (hippocampal) cortex; oc: optic chiasma; ot: optic teetum; pt: pallial thickening; re: nucleus reuniens; rt: nucleus rotundus; s: septum; str: basal striatum; vp: lateral forebrain bundle, ventral peduncle; III: oculomotor nerve root fibers.
nuclei surrounding nucleus rotundus, i.e., mainly within nuclei dorsomedialis anterior, dorsolateralis anterior and much less abundantly in nuclei ventralis and reuniens (according to Papez's nomenclaturel6). N o t all the neurons in each o f these nuclei are labeled and most o f the labeled cells are f o u n d in the central portion o f the nuclei (Figs. 1 2B, C). In one case in which the S T R injection site was much larger, also involving the ventral part of the pallial thickening and the rostral tip o f the D V R , a few labeled cells were disclosed within nucleus geniculatus lateralis, p a r s dorsalis. In all 6 turtles, however, no HRP-positive cells o c c u r in nucleus rotundus itself after S T R injection (Figs. 1 and 2C). Contralaterally to the injection site the whole dorsal thalamus is devoid o f labeled cells. At the midbrain level a large population o f labeled neuronal somata is found in the ventrolateral portion of the rostral midbrain tegmentum on the injection side (Figs. 1 and 3A, B). M o s t o f these HRP-positive neurons are lying immediately caudal to the o c u l o m o t o r nerve r o o t fiber level (Fig. 1). The H R P granules are evenly distributed within the cytoplasm and invade the proximal dendrites o f these midbrain neurons (Fig. 3A). M o r e caudally, at the level o f the interpeduncular nucleus, a small
Fig. 2. A: microphotograph of a transverse section through the telencephalon showing the largest extent of a typical striatal HRP injection site (case No. 210/2). B and C: microphotographs (darkfield illumination) of transverse sections through the dorsal thalamus showing HRP-positive cells within the core of nucleus dorsomedialis anterior (B) and in nucleus dorsolateralis anterior (C), ipsilaterally to the striatal injection. D and E: microphotographs (dark-field illumination) of transverse sections through the rostra1 diencephalon (level D in Fig. 1) illustrating the numerous labeled fibers that occur in the lateral forebrain bundle after striatal HRP injection. The lateral forebrain bundle is shown in its more medial extension in the hypothalamus (D) and in its lateral extension in the striatum @). Magnifications indicated.
30 number of HRP-positive neurons are scattered throughout the lateral midbrain tegmentum (Fig. 1). No HRP-containing neurons are seen caudally to the isthmus of the hindbrain after striatal injection. As mentioned above, numerous labeled fibers can be observed near the striatal injection site, some of them invading caudally the lateral forebrain bundle (LFB). HRP-containing fibers are present in both the dorsal and ventral peduncles of the LFB (Figs. 1, 2D, E and 3C). The HRP-positive fibers of the dorsal peduncle can be traced as far as the level of the dorsal thalamic labeled neurons where they appear to arise. They reach the perirotundus nuclei by spreading out both around the surface and through the rostral half of nucleus rotundus. The ventral peduncle fibers can be followed down to the lateral border of the midbrain tegmentum where they appear to arborize. As they penetrate the midbrain tegmentum the labeled fibers of the
Fig. 3. A: microphotograph (dark-field illumination) showing medium-sized labeled neurons located in the rostral midbrain tegmentum, ipsilaterally to the HRP striatal injection. B: microphotograph (dark-field illumination) of a transverse section through the large population of labeled cells (level H in Fig. 1) that occurs in the ventrolateral portion of the rostral midbrain tegmentum, ipsilaterally to the striatal injection. In the left half of the figure, transversally cut labeled fibers can be observed (asterisk) in the terminal field of the 'tractus strio-tegmentalis ventralis'. C: microphotographs (dark-field illumination) of a transverse section through the diencephalon showing the ventral and dorsal peduncles of the lateral forebrain bundle, both containing numerous HRP-positive fibers after striatal injection. Magnifications indicated.
31 ventral peduncle progressively break out along the lateral border. In frontal sections, the very small H R P granules encountered within this portion of the labeled fibers are diffusely distributed in the lateral tegmentum (Figs. 1 and 3B). The H R P granules gradually disappear as we reach the caudal half portion of the midbrain tegmentum. There appears to be no direct continuity between the HRP-containing cell bodies of the midbrain and the labeled fibers of the ventral peduncle, in contrast to the labeled fibers of the dorsal peduncle which merge in thalamic cells.
Dorsal ventricular ridge injections The H R P injection sites in the DVR are all mainly confined to the dorsomedial quadrant of the rostral half of this structure (Figs. 4 and 5A). In most cases, some H R P material appears to have also diffused in the lateral ventricle since a light diffusion zone can be noticed more especially around the basis of the lateral ventricle (Fig. 4). In all the animals of this group, however, no H R P material can be found further than 100-150/am from the ependymal surface. Within this diffusion zone a few weakly labeled neuronal somata can be visualized along the ependymal border of both the STR and the septum. After DVR injections a very large number of HRP-positive neurons occurs in the dorsal thalamus, ipsilaterally. Labeled cells are found in most dorsal thalamic nuclei including nucleus rotundus. They are more abundant in the rostral half of the nucleus rotundus. The number of labeled neurons present in all nuclei surrounding nucleus rotundus, i.e., nuclei dorsomedialis anterior, dorsolateralis anterior, ventralis and reuniens, was much greater than after striatal injection. At the level of nucleus 2mm
dma
~.~./
\V(I'..'
:
Fig. 4. Camera lucida drawings of transverse half sections through the forebrain of Chrysemyspieta (case No. 221/1) illustrating the effect of typical dorsal ventricular ridge injections (hatched areas) of HRP. The drawings are set out in a rostro-eandal order from A to F. The full circles indicate neuronal somata that have been labeled following retrograde axonal transport of HRP.
32
Fig. 5. A: microphotograph of a transverse section through the rostral half of the dorsal ventricular ridge showing the largest extent of a typical H R P injection site in that structure (case No. 221/I). B: microphotograph (dark-field illumination) of a transverse section through the dorsal thalamus that illustrates the numerous labeled cells present within and around the rostral half of nucleus rotundus, ipsilaterally (right side) to a dorsal ventricular injection. C: microphotograph (dark-field illumination) of a transverse section through nucleus reuniens showing the large number of HRPpositive cells that occurs in this nucleus ipsilaterally to a dorsal ventricular ridge injection. D : microphotograph (dark-field illumination) that illustrates some of the large neurons of nucleus rotundus whose cell body and proximal processes are completely filled with H R P material. These neurons appear to have been labeled by a direct uptake of the enzyme by the damaged portions of their axon (see text). Magnifications indicated.
33 dorsomedialis anterior, for instance, the HRP-positive cells are extremely abundant and closely packed (Figs. 4 and 5B). The nucleus reuniens also contains a large population of labeled cells that prolongs itself caudally within the pretectal area (Fig. 5C). In contrast to the results obtained after the STR injection, no HRP-containing neurons are present caudally to the meso-diencephalic junction after HRP injection into the DVR. In one case, however, in which the DVR injection site did also involve the most dorsal portion of the STR, weakly labeled neuronal somata were outlined in the midbrain tegmentum. These labeled neurons are identical in type and size to those observed in the same midbrain areas after HRP injection confined to the STR. The amount of HRP granules in these midbrain neurons is very small, however, in comparison to the strong labeling resulting from a massive STR injection. In this peculiar case, a few labeled fibers are found in the vicinity of that portion of the DVR injection site reaching the STR. Some of the reactive fibers can be followed back to large neurons of nucleus rotundus that appear completely filled with HRP material, thus contrasting with other labeled cells of that area displaying the typical granular appearance (Fig. 5D). Therefore, it may be presumed that these large nucleus rotundus neurons have been labeled by a direct uptake of the enzyme by the injured segment of their axon coursing through the injection area. All the HRP-positive neurons described in the present study after HRP injection either in the STR or DVR have been found ipsilaterally to the injection. It is worth mentioning, however, that a few labeled cells have been observed within the contralateral perirotundus area in one case of DVR injection. These neurons were lying along the ependymal surface at the level of nucleus dorsomedialis anterior. They were very weakly labeled in comparison to the ipsilateral dorsal thalamic neurons. The relevance of this finding remains to be evaluated since the injection site in this peculiar case was apparently identical to other DVR injection sites that did not show such a contralateral labeling. DISCUSSION
The present results confirm the existence in the turtle (Chrysemyspicta) of a midbrain-striatal connection that has been previously mapped out by means of the Falck-Hillarp histochemical method for monoamines. The large population of HRPpositive neurons found within the rostral half of the midbrain tegmentum after STR injection is strikingly similar to the large group of catecholamine-containing cells disclosed in the same midbrain area of the turtle20, ~1. In both cases the neuronal somata which are of the same type and size occupy the ventrolateral portion of the rostral midbrain tegmentum. The catecholamine-containing neurons of that area have been shown to project via the medial forebrain bundle upon the STR. This midbrain-striatal catecholaminergic pathway of the turtle brain has been tentatively related to the mammalian nigro-striatal dopaminergic pathway 17. On the other hand, a small number of elongated HRP-positive cells are also scattered throughout the lateral portion of the caudal half of the midbrain tegmentum
34 after STR injection. These labeled cells are strikingly similar in size and shape to the 5-hydroxytryptamine (serotonin)-containing neurons disclosed in the same midbrain area by means of the Falck-Hillarp method 20. Numerous other serotonin-containing neurons occur along the raphe region of the turtle lower brain stem but none of these cells have been labeled after H R P injection in the STR. Therefore, it may be postulated that most serotonin-type axon terminals present in the STR of Chrysemys picta arise from the serotonin-containing neuronal somata lying at the level of the caudal midbrain tegmentum. This condition would resemble closely the situation in mammalian brain in which most telencephalic serotonin axon terminals appear to arise from nuclei raphe dorsalis and raphe medialis (or centralis superior) both lying at caudal midbrain or isthmus level (see ref. 12). The present study also provides important data with regard to the exact origin of the dorsal thalamic projection to the turtle STR recently disclosed by means of the silver anterograde degeneration technique 6. The thalamo-striatal afferent in Chrysemys picta seems to originate from neurons belonging to various dorsal thalamic nuclei surrounding nucleus rotundus, i.e., nuclei dorsomedialis anterior, dorsolateralis anterior and, to a lesser extent, nucleus ventralis, which all contain HRP-labeled cells after STR injection. On such a basis the perirotundus nuclei may be compared to the mammalian intralaminar nuclei which are the major thalamic source of the thalamo-striatal afferent. Owing to the fact that no HRP-positive cells were detected within nucleus rotundus itself after STR injection, it may be presumed that this nucleus does not contribute significantly to the turtle thalamo-striatal connection. In contrast the nucleus rotundus, which is the major thalamic target of the optic tectum, appears to project to the DVR of the turtle. Numerous HRP granules are indeed found within the neurons of this nucleus after DVR injection as shown in this study. These results are in apparent contradiction with the fact that ablations of the DVR do not produce retrograde cell changes in nucleus rotundus whereas lesions within the STR induced marked retrograde changesg,2L Such a discrepancy may be explained by the fact that STR lesions inevitably damage the more proximal portion of nucleus rotundus axons which course through the STR en route to the DVR 6. Destruction involving only the terminal arborizations of nucleus rotundus axons in the DVR could be insufficient in order to induce retrograde cell alterations. On the other hand, a large amount of neurons of the perirotundus area (including nucleus reuniens) also show a positive H R P reaction after DVR injection. Therefore, the perirotundus area seems to project to both the STR and the DVR. Whether both of these projections arise from the same or different neuronal populations in the perirotundus nuclei remains to be evaluated. It must also be mentioned that the nucleus reuniens, that is heavily labeled after DVR injection, has been recently reported to receive auditory fibers and to send its axons to the DVR in the lizard Iguana 3. The lateral forebrain bundle (LFB) in Chrysemys picta as well as in other reptiles is usually divided into dorsal and ventral peduncles. In the past, the dorsal peduncle has been associated with various 'thalamo-striatal' pathways whereas the ventral peduncle was thought to be composed mainly of 'strio-tegmental' pathways 1.
35 The present study has shown that numerous HRP-contaning fibers occur at the level of both peduncles of the LFB after STR injection. On the one hand, HRP material appears to migrate retrogradely from the STR injection site to the cells of the perirotundus nuclei via the dorsal peduncle fibers, thus confirming the existence of a thalamo-striatal system among the fibers of the dorsal peduncle. On the other hand, HRP material appears to migrate anterogradely within the ventral peduncle fibers down to the ventrolateral border of the midbrain tegmentum. These labeled fibers have a course that corresponds closely to the so-called 'tractus strio-tegmentalis ventralis' terminating in large part along the ventrolateral border of the midbrain tegmentum1. The labeled fibers present in the ventral peduncle of the LFB after STR injection could represent a reptilian equivalent of the well-documented strio-nigral pathway of mammals. In mammalian brain, the strio-nigral fibers also penetrate the midbrain tegmentum through a latero-ventral course 13. The presence of numerous labeled fibers in the LFB of Chrysemys picta after STR injection could be the result, at least in part, of a direct uptake of HRP material by axons that could have been damaged by the tip of the canula (see Fig. 2A). It is also possible that in the brain of reptiles, as in other poikilotherms, the slower metabolic rate may facilitate the mapping of the nervous fibers with the HRP technique. The HRP granules could migrate along the axons of the turtle brain at a much lower rate than the migration rate estimated for mammalian central nervous fibers. In fact, 72 h was found to be the ideal survival period in order to obtain an optimal retrograde labeling of Chrysemys picta neurons in comparison to the 2a, "!8 h period disclosed for mammalian brains TM. Very few labeled neurons were obtained even 48 h after successful HRP injection in the turtle brain. The effect of survival periods longer than 72 h on the intensity of the HRP retrograde labeling remains, however, to be investigated. In any event, the results of this study suggest that the HRP histochemical method may be a new and promising approach to the study of the brains of lower vertebrates. ACKNOWKEDGEMENTS The author would like to express his deep gratitude to Dr. F. Dord, Mr. R. Boucher and Mr. R. Marchand for their valuable technical assistance at various stages of this study. The histological work was performed by Mrs. D. Picard. This study was supported by Grant MT-732 of the Medical Research Council of Canada. The author is holder of a Scholarship of the Medical Research Council.
REFERENCES 1 ARIi~NS-KAPPERS,C. U., HUBER, G. C., AND CROSBY,E. C., The Comparative Anatomy of the Nervous System of Vertebrates including Man, Hafner, New York, 1960. 2 FALCK, B., HILLARP, N.-A., THmME,G., AND TORP, A., Fluorescence of catecholamines and related compounds condensed with formaldehyde,J. Histochem. Cytochem., 10 (1962) 348-354.
36 3 FOSTER, R. E., AND PEELE, T. L., Thalamotelencephalic auditory pathways in the lizard (Iguana iguana), Anat. Rec., 181 0975) 530 (abstract). 4 GOEDBY, F., AND GAMBLE, H. J., The reptilian cerebral hemispheres, Biol. Rev., 32 (1957) 383420. 5 GRAHAM, R. C., JR., AND KARNOVSKY, M.J., The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique, J. Histochem. Cytochem., 14 0966) 291-302. 6 HALL, W. C., AND EBNER, F. F., Thalamotelencephalic projections in the turtle (Pseudemys scripta), J. comp. Neurol., 140 (1970) 101-122. 7 JOHNSTON,J. B., The cell masses in the forebrain of the turtle, Cistudo carolina, J. comp. Neurol., 25 (1915) 393-468. 8 JONES, E. G., AND LEAV1TT,R. Y., Retrograde axonal transport and the demonstration of nonspecific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei in the rat, cat and monkey, J. comp. Neurol., 154 (1974) 349-378. 9 KRUGER,L., AND BERKOW1TZ,E. C., The main afferent connections of the reptilian telencephalon as determined by degeneration and electrophysiological methods, J. comp. Neurol., 115 0960) 125-141. 10 KUYPERS, H. G. J. M., K1EVIT, J., AND GROEN-KLEVANT,A. C., Retrograde axonal transport of horseradish peroxidase in the rat's forebrain, Brain Research, 58 (1973) 470-477. 11 LAVAIL, J. H., WINSTON, K. R., AND TISH, A., A method based on retrograde intraaxonal transport of protein for identification of cell bodies of origin of axons terminating within the CNS, Brain Research, 58 (1973) 470-477. 12 LORENS, S. A., AND GULDBERG, H. C., Regional 5-hydroxytryptamine distribution following selective midbrain raphae lesions in the rat, Brain Research, 78 (1974) 45-56. 13 NAUTA, W. J. H., AND MEHLER, W. R., Projections of the lentiform nucleus in the monkey, Brain Research, I (1966)3-42. 14 NAUTA, H. J. W., PRITZ, M. B., AND LASEK, R. J., Afferents to the rat caudoputamen studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method, Brain Research, 67 0974) 219-238. 15 NORTHCUTT, R. G., The Telencephalon of the Western Painted Turtle (Chrysemys picta belli), Illinois Biological Monographs, University of Illinois Press, Urbana, l[1., 1970. 16 PAPEZ, J. W., Thalamus of turtles and thalamic evolution, J. comp. Neurok, 61 (1935) 433-475. 17 PARENT, A., Demonstration of a catecholaminergic pathway from the midbrain to the strioamygdaloid complex in the turtle (Chrysemys picta), J. Anat. (Lond.), 114 (1973) 379-387. 18 PARENT,A., AND BOUCHER,R., Horseradish peroxidase study of the striatal afferent connections in the turtle (Chrysemyspicta), Proc. Soc. Neurosci., 5 (1975) 195 (abstract). 19 PARENT, A., AND OLIVIER, A., Comparative histochemical study of the corpus striatum, J. Hirnforseh., 12 (1970) 73-81. 20 PARENT,A., AND POIRIER,L. J., Occurrence and distribution of monoamine-containing neurons in the brain of the painted turtle (Chrysemyspicta), J. Anat. (Lond.), 110 (1971) 81-89. 21 PARENT, A., AND POITRAS, D., The origin and distribution of catecholaminergic axon terminals in the cerebral cortex of the turtle (Chrysemys picta), Brain Research, 78 (1974) 345-358. 22 POWELL, Z. P. S., AND KRUGER, L., The thalamic projection upon the telencephalon in Lacerta viridis, J. Anat. (Lond.), 94 (1960) 528-542.
