_ iiaila 5:...cyfsaeiahvnv .. . .IO' 'Vhlz'; F173;).tvftru... ‘v EQ’VI? lit-:13.) 1 . t7iiyfas 03...!!!“ 2;)...— 3.‘ 7i .1 ”0!. 2991“}..9 In... «Avila-o ....... Ara-Ian... Jami... l r 3.7.. Er. .).. 7.i§xt§§§r . ~ til! gig-ittgtnl‘il ztt‘jgzgt‘voialzkhktfill . 1’1: I :1)! )5): 3!... it . sgtfif; .Jlliyf’llt §.I§§§ ‘ 32f... abtiisN4§fug , W15»: gingin‘llsi 15$“; V ax 71 Ml, . 3!... huh 2.. ’3‘, ’éfns‘u‘z. {be}; ~ th§1\X) 3.1:; I .3 till): .5 E Aling‘t‘l ‘ehlééfi izzz$£z .3954. .‘Eilr-EI-ZE ‘Aer p . ~ .I C’V‘Z. {Hkvvlh‘fli )7). .l. 7.4.41}? 12. ’3’?! x)- ...igigfilifik‘tiu\ fl... uni 1.3.5.. m 4r 55‘, z: $5: up.“ 1 . .>v...£..i.’i:€ {nil 2.1TKEARI4O91»>J. £213; a»: {tilt Iii)”; an?) D£L!’E’ big I L . ‘ (tr?! :9. . Tic 1,5,1 3 lg»; .zmrv-"Iqifnl‘rafixn {IDA‘P II.‘ wring». ., 32.111393... .....n$37...l.1§.. {5a. m 3.3351“. 1.3;} -45.? . x. . \ ind-3F!) etc? u...;:4_..2.1..:;:. Lu 5.....Ehm5; ,umrfiyiv .1 7.3. , .. . \— - " 7m,» ,. _ av - r '-'."‘l th.: This is to certify that the dissertation entitled Medullary Sources of Projections to the Kinesthetic Thalamus in the Raccoon presented by E . -Michael Os tapoff has been accepted towards fulfillment of the requirements for Ph ° D ° degree in Psychology/ Neuroscience { I 4 .J-‘Z/ ///{f //' V " I of /., , . Y 13‘; L/ Major p‘lfi‘cssor Datej71/7C’6Z/Héz /7/2 012771 MSU is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES “ V RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. THE MEDULLARY SOURCES OF PROJECTIONS TO THE KINESTHETIC THALAMUS IN THE RACCOON By Ernst—Michael Ostapoff A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Psychology and Neuroscience Program 1982 6/30é32~ ABSTRACT THE MEDULLARY SOURCES OF PROJECTIONS TO THE KINESTHETIC THALAMUS IN RACCOONS By Ernst-Michael Ostapoff Kinesthetic receiving regions in the medulla are known to project to cerebellum, but the medullary sources of projections to the kinesthetic thalamus are not known. The purposes of these studies were to 1) establish the distribution of cells projecting to the thalamus in comparison with that of cells projecting to the cerebellum; and 2) establish the cells of origin of the kinesthetic projections to the thalamus. We used the retrograde transport of horseradish peroxidase, either injected via a microliter syringe (for large injections into either the thalamus or cerebellum) or expressed from a simultaneous recording/injecting electrode (for small injections into the kinesthetic thalamus) in 25 raccoons. Seven nuclear subdivisions in the dorsal medulla were recogniZed. In 1) the central cluster region of the cuneate-gracile complex, in 2) cell group z and 3) the reticular portion of cell group x, 85-95% of the cells project to the thalamus as do 30% of the cells in 4) the basal subdivision of the cuneate. In the compact portion of cell group E.-M. Ostapoff x, 30% of the cells project to the ipsilateral and 30% to the contralateral cerebellum. Both the contralateral thalamus and ipsilateral cerebellum receive projections from: equal numbers of cells (20%) in 5) the rostral cuneate subdivision; 6) the external cuneate and 7) its medial tongue project predominantly to the cerebellum (72 and 62% of the cells respectively) with t0%-20% projecting to the thalamus. In the kinesthetic thalamus, projections from cell group 2 and the reticular portion of x are found lateral to those from the external cuneate-medial tongue and basal cuneate. Electrophysiological mapping of the dorsal medulla confirmed that all these nuclear subdivisions projecting to the kinesthetic thalamus receive projections from the deep tissues of the body, primarily from the forelimb. ACKNOWLEDGMENTS I wish to express my appreciation to the members of my committee Drs. Raymond Frankmann, Antonio Nunez, Denis Steindler, Charles Tweedle and special thanks to John I. Johnson, advisor and chairman, whose quiet dedication and persistance were invaluable to the completion of this work. I wish to also thank Dr. Glenn Hatton, whose critical review of my proposal helped focus this research effort. Cindy Smith assisted in the surgical and histological preparations. Animals were secured through the c00peration of the Division of Wildlife, State of Michigan, Department of Natural Resources. This research was supported by NSF Grant BNS-81-OBO731. ii TABLE OF CONTENTS LIST or FIGURES. .......... .' ................... . ........ . ..... . ...... vfi LIST OF ABBREVIATIONS......... .................................... ix CHAPTER I. .LITERATURE REVIEW ............. . ....... ................. 1 INTRODUCTION.. ...... .... ................................... ........ 1 CORTICAL CYTOARCHITECTONIC ORGANIZATION ............................ 1 FUNCTIONAL ORGANIZATION OF THE SOMATOSENSORY AND MOTOR CORTICIES... 1 THALAMO-CORTICAL RELATIONSHIPS ......................... ............ 5 Thalamic Nomenclature. ............ . ..... . ..... ... ............... 5 Carnivores ......... .. ........... ......... ......... .... ........ 6 Primates ................................................... ... 7 Thalamocortical Connections.. .................... ...... ......... 8 Thalamic Submodality Segregation ............ . ........ ........... 9 MEDULLO-THALAMIC RELATIONSHIPS .............................. . ...... 10 Medullothalamic Connections......... ....... ...... ........ ....... 10 Dorsal Sensory Nuclei of the Medulla............................ 1O Cytoarchitectonic Organization and Connections ................ 1O Physiology.................. ....... .... .......... ....... ..... . 1O Medullary Submodality Segregation............................... 14 Medullary Muscle Afferent Segregation ........................... 14 MUSCLE PROJECTIONS AS THE TESTABLE CASE OF SUBMODALITY SEGREGATION............ ..... ... ..... ............................... 17 iii A Segregated Muscle Afferent Pathway Within the Functional Organization of the Raccoon Somatosensory System ................ 19 CorteXOO0.0’OOOOOOOOOOOOOOOOOOOOOOO......OOOOOOOOOOOOOO....0... 19 ThalamuSOO00.0.00.........OOOOOOOOOOO.......OOOOOOOOOOOOOOOOO. 20 Medulla. O O OOOOOOOOOOOOOOOOOOO O ..... O O O 0 I O O O O O O O 000000000000000 20 OBJECTIVES OF THESE STUDIES..... ....... . ...... . .............. ...... 21 Possible Clinical Significance of the Proposed Studies .......... 22 CHAPTER II. COMPARISON OF CELLS IN THE MEDULLA PROJECTING TO THE THALAMUS VERSUS THOSE PROJECTING TO THE CEREBELLAR CORTEX. O O ...... O O O O O O O O O O C O O O O O O O O O O O O O I O O O O ..... O O O O O O O ...... O O O O O 24 INTRODUCTION 0000000000000 O ....... O I O 0 O O ..... 0 O O 0 O OOOOOOOOOOOOOOOOOO 24 METHODS ............................................................ 25 RESULTSOOOO0.0...O0.00......OOOOOOOOQOOOOOOOOO0....OOOOOOOOOOOOO... 27 Subdivisions of the Dorsal Mechanosensory Medulla............... 27 Medullothalamic Projections..... .......... . ..... ................ 35 Labeling Frequency as a Result of Large Thalamic Injections Centered in the Ventrobasal Complex ............................. 35 Bilateral Medullothalamic Projections ......................... 45 Other Sources of Medullothalamic Projections ......... ......... 45 Medullocerebellar Pro'jectionSOO00..............OOOOOOOOOOOOOOOOO 48 Bilateral Medullocerebellar Projections....................... 48 Other Medullocerebellar Projections..... ...................... 57 DISCUSSION..... ................................................ .... 57 Cuneate Nuclear Projections to Thalamus ......................... 57 External Cuneothalamic Projections......... ...... ....... ........ 58 Cell Groups Z and X Thalamic Projections........................ 59 Medullary Projections to Cerebellar Cortex....... ............... 59 iv Cell Groups x and z as a Complex... ..... ............. ......... 60 Nuclear Subregions Projecting to Both the Thalamus and Cerebellar Cortex............................................... 61 Rostral Cuneate............................................... 61 External Cuneate and Medial Tongue............................ 61 CHAPTER III. DORSAL MEDULLARY SOURCES OF PROJECTIONS TO THE KINESTHETIC THALAMUS... ...... . ....... ............ ........ .......... 65 INTRODUCTION.. ......... ............... .............. .... ...... ..... 65 METHODS.................. ..... ..................................... 67 Physiological Mapping of the Medulla ......................... ... 69 RESULTS... ................................................... ...... 70 Use of Pre-injection Thalamic Mapping to Localize Injections.... 7O Rostrolateral Injections ........................................ 79 Medial Injections ......... . ........... .. ...... . .............. ... 79 Injection into the Rostral Pole of the KVB ...................... 80 Other Medullary Projections to the KVB........... ..... .......... 9O Medullary Mapping Data ....... .... ....... . ...... ................. 97 DISCUSSION........ ..... . ........ ........................ ......... .. 109 The Kinesthetic Shell of the Ventrobasal Complex ................ 109 Nuclear Subdivisions in the Dorsal Medulla of the Raccoon ....... 111 Mapping Data..... ...... . ..... . ....................... . .......... 111 XZ Complex in Raccoons ............ . .......................... ... 113 Contributing Factors to These Results .................... ....... 114 Other Medullary Projections to the KVB. ..................... .... 115 Parallel Organization of Projections between Levels in the Ascending Somatosensory System.................. ......... ....... 116 Parallel Processing in the Ascending Somatosensory System....... 117 REFERENCES . ...... . .................... . . vi LIST OF FIGURES Figure 1.1 Atlas of the nuclear subdivisions of the dorsal mechanosensory medulla....... ....... ..... ..... ............. Figure 1.2 High magnification photomicrographs to show cytological details of cg x-c and x—r ...... .................. Figure 1.3 Reconstructions of large thalamic injection sites..... Figure 1.4 Labeling in the dorsal mechanosensory medulla following large thalamic injections.. ..... ...... ..... ..... Figure 1.5 Medullary cell labeling following large thalamic HRP injections shown in the horizontal plane. ....... ..... Figure 1.6 Photomicrographs showing the distinction between the cg x-c and x-r in the horizontal plane.. ..... ......... Figure 1.7 Photomicrograph of horizontal sections showing cell labeling in the ECu after thalamic injection of HRP............... Figure 1.8 Reconstructions of large cerebellar injections........ Figure 1.9 Cell labeling in the dorsal mechanosensory medulla following large cerebellar injections..................... Figure 1.10 Distribution of labeled cells in horizontal sections through the dorsal medulla following cerebellar injections........ Figure 1.11 Horizontal section illustrating the cellular bridge connecting the ECu and Mt.................................. Figure 1.12 Summary of cerebellar projecting versus thalamic vii 28 33 36 39 41 43 46 49 51 53 55 projecting cells in the dorsal mechanosensory medulla............. 62 Figure 2.1 Injection into the lateral KVB........................ 71 Figure 2.2 Examples of evoked potentials recorded through a recording/injecting micropipette........................ 77 Figure 2.3 Injection into the caudomedial KVB.................... 81 Figure 2.4 Injection in the medial KVB............................ 86 Figure 2.5 Injection near the rostral pole of the KVB............ 91 Figure 2.6 Summary of dorsal mechanosensory medullary projections to the KVB............................................ 95 Figure 2.7 Mapping data through the bCu ....... .... ....... ........ 98 Figure 2.8 Mapping data through the bCu in the vicinity of labeled cells resulting from kinesthetic specific injections of HRP into the KVB.................................................. 102 Figure 2.9 Mapping data from cg x—r in proximity to labeled cells from a kinesthetic specific injection into the lateral KVB.. 105 Figure 2.10 Examples of evoked potentials in the dorsal medulla... 105 viii bCu Ca cCu cg CGr Cu CuGr da DCN df ECu FA Gr Hc IC IFT LCN ABBREVIATIONS basal region of the cuneate n. caudate n. cluster region of the cuneate n. cell group central grey of the midbrain cuneate n. cuneate-gracile nuclear complex area receiving projections from dorsal column nuclei area receiving projections from external cuneate n. fast adapting response gracile n. area receiving projections from heterogenous area of SI cortex horseradish peroxidase inferior colliculus infratrigeminal n. the deep tissues of the arm the deep tissues of the leg cutaneous receptors in the head kinesthetic subdivision of the ventrobasal complex area receivng projections from cutaneous receptors in the leg lateral cervical n. ix LGN MAc MB No Mt rCu RN SA SI VA Vc VL VLc VPI VPLc VPLo VPM lateral geniculate n. muscle afferent region of the SI cortex mammlllory body primary motor area of cortex medial tongue extension of the ECu rostral region of the cuneate n. red n. slowly adapting response primary Somatosensory area of the cerebral cortex ventroanterior n. of the thalamus ventrobasal complex of the thalamus glabrous skin representation area of SI cortex ventrolateral n. of the thalamus ventrolateral n., caudal part ventroposteroinferior n. ventroposterolateral n., caudal part ventroposterolateral n., oral part ventroposteromedial n. cell group x, compact part cell group x, reticular part area receiving projections from cutaneous receptors in the indicated digits of the hand CHAPTER I. LITERATURE REVIEW INTRODUCTION Elucidating the principles of structural organization and how this organization may relate to the function of the central nervous system is one of the primary goals of neuroscience. The hypothesis that a segregation of projections into functional subunits, within which the elements respond to single stimulus submodalities, which exist at each level (or synapse) of the ascending somatosensory system is garnering increasing experimental support. Data suporting this organizational principle has been most extensively studied at the most easily accessible level of this three level system, the cerebral cortex. At the preceding levels, from which the cortex receives its information (and perhaps derives its organization) much less is known of the organization of the sensory thalamus .and relatively little of this organization at the medullary level. CORTICAL CYTOARCHITECTONIC ORGANIZATION To preface a discussion of cortical functional organization, a brief description of the architectonic organization of sensory and motor cortex is necessary. Traditionally, these two cortical regions are considered to be separated by the central sulcus. with the motor cortex rostral and the sensory cortex caudal (Johnson '80). Precentral cortex or area 4 (that region lying anterior to the central sulcus) can be characterized by the presence of giant pyramidal cells in layer V. Postcentral cortex or area 3b can be characterized by an expanded granular layer IV as well as the presence of the outer layer of Baillarger (the cell free stripe between layers IV and V). These designations were described by Brodmann ('03) in man and primates and by Hassler & Muhs-Clement ('64) in cats. In the depths of the central sulcus lies an area (designated 3a) which is cytoarchitecturally characterized as the transition zone between the giant pyramidal cells in layer V of area 4 and the attenuation of the granular layer IV of the sensory cortex (Jones & Porter '80). There appears to be a variable degree of species specific overlap of these cytoarchitectural features (see Jones & Porter '80). In human cortex (Brodman '03), new world monkeys (Bonin '38, Jones '75, Sanides ’68) and cats (Hassler & Muhs-Clement '64) there is some overlap; in old world monkeys (Jones, Coulter & Hendry '78, Jones, Wise & Coulter '79) and raccoons (Johnson, Ostapoff & Warach '82) there is no overlap of the giant pyramidal cells of motor cortex and the granular layer of sensory cortex. FUNCTIONAL ORGANIZATION OF THE SOMATOSENSORY AND MOTOR CORTICES The classical conceptions of the functional organization of the primary somatosensory cortical areas (SI) have recently come under increased scrutiny. Early workers, using rather large surface electrodes or large "microelectrodes" (approximately 500 um in diameter) with large distances between recording sites, found that in the cerebral cortex there were several areas in which stimulation of the peripheral body resulted in evoked activity (review, Johnson '80). The largest such area has come to be known as the primary somatosensory area, or SI. Further, these evoked responses are organized so that stimulation of adjacent area of the body evokes responses in adjacent loci in the cortex. This functional organization has been termed somatotopy and can be demonstrated in virtually any mammal. Rather than a single representation of the body surface (Woolsey '58) several laboratories, including. our own, using finely detailed micromapping techniques have recently proposed that multiple representations of all or parts of the body in fact exist in SI in a wide variety of species (raccoon; Johnson et al. '82, cat; Dykes, Rasmussan,& Hoeltzell '80; primates, Kaas, Nelson, Sur, Lin, & Merzenich '79, monkey; Zimmerman '68, galago; Sur, Nelson & Kaas '80; tree squirrel, Sur, Nelson, & Kaas '78; opossum, Pubols, Pubols, DiPette & Sheely '76). The most compelling evidence derives from maps of large numbers of electrode* penetrations recording evoked activity in single animals in which the pattern of responses recorded in closely spaced electrode penetrations describe not a single somatopic representation but a split representation. For example, evoked responses from a proximal digit may lie on both sides of the representation of the distal tip of that digit forming a mirror image within which each of these representations appear to maintain somatotopy (e.g. grey squirrels, Sur et a1. '78) or the representation of the proximal digit may have on either side representations of the distal digit (owl monkey, Merzenich, Kaas, Sur & Lin '78 or the representations may be found in serial order, i.e. proximal digit, distal digit, proximal digit, distal digit as in Rhesus monkey (Paul, Merzenich and Goodman '72). These multiple body representations may be segregated on the basis of response characteristics (presumably related to stimulus submodality, i.e. cutaneous slow and fast adapting, deep, pacinian, etc.). The pattern of physiological responses one obtains when recording from these cortical areas adjoining the central sulcus remains fairly constant. Most rostral is the classical primary motor cortex (area 4) in which stimulation of cutaneous and muscle« afferents may evoke physiological responses (Murphy, Wong & Kwan '75, Lucier, Ruegg & Wiesendanger '75, Hore, Preston, Durkovic & Cheney '76). Its primary physiological feature is the low threshold this cortex displays for electrical stimulation of muscle‘ movements (Woolsey '58, Hardin, Arumugasamy & Jameson '68). In the rostral part of SI cortex one can record responses from either light tactile peripheral stimulation or from electrical stimulation of cutaneous nerves (this is generally regarded as related to the architectonic cortical area 3b). In between these two regions, often in the depths of the central sulcus, are responses to peripheral stimulation of the deeper lying tissues or to electrical stimulation of the muscle afferent nerves, (usually considered area 3a (monkey: Phillips, Powell & Wiesendanger '71, Merzenich et al. '78; cat: Kaas et al. '79). Recently interest has focussed on this cortical region (3a) lying mainly in the depths of the central sulcus because it responds short latency to muscle la afferents (Phillips et al. 1971) and lies between the classical sensory cortex (area 3b) and motor cortex (area 4). Area 3a is not the only cortical cytoarchitectonic area receiving muscle afferent input in monkeys. Area 4, in conscious monkeys, also responds to peripheral stimulation (Lemon '79, Tanji & Wise'8l , Wise & Tanji '81). Area 2 in monkeys, also considered a part of 81, receives muscle afferent information (Merzenich et al. '78) as does the SII cortex (Andersson, Landgren and Wolsk '66). We have not, however, recorded unit responses to peripheral stimulation in the motor cortex (Area 4) in anesthetized raccoons (Johnson et al. '82) nor have we investigated the caudal SI area for muscle afferent input (analogous to area 2 in the monkey). THALAMO-CORTICAL RELATIONSHIPS It has, of course, long been known that the primary sources of input to the cerebral cortex are via the dorsal thalamus (Poliak '32). In general it has been found that (using the terminology developed in the carnivore thalamus) the ventrobasal complex innervates the SI sensory cortex and the ventrolateral complex innervates the motor cortex. Thalamic Nomenclature A brief discourse on the various thalamic .nomeclatures in necessary both to accurately define the regions of interest to this study as well as to provide a basis for comparisons between published reports on different species. Recently an attempt has been made in this regard (Jones '81) but a more detailed description of the particular thalamic region included in this study appears necessary. Thalamic nomenclature has historically proceeded from the parcellation and naming of nuclei on the basis of the distribution pattern of cell bodies (Nissl cytoarchitectonics) or of myelinated fibers (myelo-architectonics), often with little data concerning the connectivity or functional characteristics of the regions being studied. Only later was experimental demonstration that these architectonic divisions represent functional subunits provided as a basis for defining nuclear regions (often leading to another classification scheme). This has of course led to confusion in nomenclature especially as regards analagous nuclei in different species. Carnivores. The region of interest to this study, defined by connectivity and physiological function, is the rostral pole of the thalamic region receiving afferent innervation from the somatosensory (via the medial lemniscus) system. In carnivores, an early description of the dog and cat thalamus by Rioch ('29) subdivided the ventral nucleus of the dorsal thalamus into several subdivisions. The subdivisions of interest here are, in order of their appearance in transverse sections from rostral to caudal: n. ventralis anterior, n. ventralis pars medialis, n. ventralis pars externa, n. ventralis pars arcuata. Briefly the defining cytoarchitectural characteristics of each of these are: n. ventralis anterior- this forms the rostral pole of the ventral nucleus. Caudally there are large polygonal cells which are distributed between horizontally running fibers. n. ventralis pars medialis- this forms the medial portion of the ventral nucleus and extends in its caudal portions dorsolaterally to cap the n. ventralis p. arcuata. The cells are medium sized and polygonal with no large cell bodies. n. ventralis externa- the cellular characteristics in this division include a mixture of very large cells similar to those in n. vent. anterior and small elongated ones in the anterior regions of this nucleus. The nucleus lies in the ventrolateral part of the middle of the ventral nucleus more or less surrounded by the n. ventralis pars arcuata. n. ventralis pars arcuata- this comprises the main portion of n. ventralis and is roughly semilunar in shape, hence the name. It is said to contain three types of cells. 1) those similar to n. ventralis anterior. 2) small and medium cells similar to l) but lighter staining and smaller. 3)cells similar to the largest in n. ventralis externa. By 1952 a simplified scheme, based on available experimental data was being used by Rose and Mountcastle in which the ventral nucleus of the thalamus was subdivided into essentially three "complexes"; the ventrolateral (VL, comprisng the n. ventralis anterior (VA) of Rioch. This was thought to receive a major input from the cerebellum (via ~the superior cerebellar peduncle) and project to motor cortex; the ventrobasal complex (VB, including the n. ventralis pars externa and pars arcuata) as these nuclei received a major input from the medial lemniscus and spinothalamic pathways and that the region of thalamus responding to tactile stimulation was coextensive with these nuclei; and the ventromedial complex (including the n. ventralis medialis) which was distinguished by a lack of connectional data as well as unresponsiveness to tactile stimulation. Primates. The nomenclature used in primates is different from that for carnivores. However analogies based on connectivity and physiological function may be drawn. The region of interest to this study in carnivores lies at the border between the thalamic areas receiving cerebellar input and projecting to motor cortex (VL complex) and the area receiving medial lemniscus input and projecting to somatosensory cortex (VB complex). In monkeys the corresponding nuclei, using the nomenclature of Olszewski ('52), are n. ventroposterolateralis pars oralis (VPLo) and n. ventrolateralis pars caudalis (VLc) for the VL complex (Tracey, Asanuma, Jones & Porter '80, Thatch & Jones '79) and n. ventroposterolateralis pars caudalis (VPLc) and n. ventroposteromedalis (VPM) for the VB complex (Mountcastle & Henneman '52, Berkley '80, Boivie '78, and Kalil '81). Remarkably, the cytoarchitectonic descriptions of these nuclei (Olszewski '52) in primates are in close accord to those corresponding areas in carnivores (Jones et al. '79, Sakai '82). Thalamocortical Connections Area 4 (motor cortex) is reciprocally connected only to the VL complex (cats, Jones & Burton '74; raccoons, Sakai '8l) or the equivalent region in monkey, VPLo and VLc (Jones et al. '79, Tracey et al. '80), portions of which nuclei can also be characterized by receiving deep cerebellar nuclear projections (Thatch & Jones '79. Hendry, Jones & Graham '79, Kalil '81). The central portions of SI (area 3b) are reciprocally connected to the central core of the VB complex (monkey: Lin, Merzenich, Sur & Kaas '79; cat: Jones & Powell '68, Jones & Leavitt '73). The thalamic cells projecting to cortical area 3a are to some extent known in monkeys and cats. Jones et al. ('79) and Friedman & Jones ('81) have shown that area 3a, responding to nerve stimulation of muscle Ia afferents, projects to a "shell" region of VPLc extending from the anterior pole throughout the anteroposterior extent of the dorsal aspect of this nucleus. In contrast, area 2 (which also receives muscle afferent information) may project only to the caudal part of this shell (Friedman & Jones '81). In monkeys, deep pressure stimuli activated single unit discharges in both VPLo and VPLc (Horne & Porter '80, Horne & Tracey '79; Loe, Whitsel, Dreyer & Metz '77; Poggio & Mountcastle '63). The cervico- and spinothalamic projections appear to project to both the VB (VPLc) and VL (VPLo-VLc) in monkeys (Jones & Powell '68) and cats (Berkley '80). Thalamic Submodality Segregation Evidence for the segregation of stimulus submodalies in the thalamus is quite scanty. Early studies did not achieve a sufficiently high resolution (i.e. closely spaced electrode penetrations) nor were the response characteristics observed in such a way as to detect submodality differences (e.g. Rose & Mountcastle '52, Mountcastle & Henneman '52, Welker & Johnson '65). Recently a study satisfying these criteria has been reported using squirrel monkeys (Dykes, Sur, Merzenich, Kaas & Nelson '81). These authors were able to show discrete segregation of two of four response characteristics (fast and slow adapting cutaneous, pacinian and deep) in the VPLc and the ventroposteroinferior n. of the squirrel monkey. Deep responses were segregated dorsally and pacinian responses ventrally (VPI) while the two cutaneous stimulation response categories were mixed in the central core of VPLc in a complex fashion. Also, in the macaque monkey, Maendly, Ruegg, Wiesendanger, Wiesendanger, Lagowski & Hess ('82) have shown that stimulation of forelimb nerves and muscle pulls results in activation of la muscle afferents in a specific rostrodorsal region of VPLc. Thus there is evidence for a segregated region of VB thalamus 10 receiving projections from deep tissue including muscles and having connections with cortical area 3a. MEDULLO-THALAMIC RELATIONSHIPS Medullothalamic Connections The projections from the dorsal column nuclei (DCN, cuneate-gracile nuclei) to the thalamus have been studied in the cat and monkey (for reviews see Berkley '80s, Kalil '81, Boivie '81). It is known that the major projection zone of the DCN corresponds to VB (VPLc —VPM in monkey) and not to VL (VPLo). These projection systems have been largely treated as single functional units and therefore only a few studies anatomically tracing specific submodality pathways have been done. It is known that the region in thalamus which projects to cortical area 3a receives input from the spinothalamic neurons in cats and monkeys (Applebaum, Leonard, Kenshalo, Martin & Willis '79, Berkley, '80, Boivie '78), although the spinothalamic nuclei receive little if any group 1 muscle afferents (Foreman, Kenshalo, Schmitt & Willis '79). Lesions and tritiated amino acid injections into the DCN (cuneate-gracile) result in the dorsorostral portion of VB being lightly if at all labeled (Boivie & Bowman '81, Kalil '81). Dorsal Sensory Nuclei of the Medulla Cytoarchitectonic Organization and Connections. The dorsal somatosensory nuclei associated with the post-cranial body in the medulla are the gracile (Gr), medially and the cuneate (Cu) laterally 11 near the obex. Lying more rostrally is the external cuneate (ECu) nucleus and rostral to the Cr, Cu, and ECu and caudal to the vestibular nuclei are Z and X, two cell groups also associated with the somatosensory system. Within the Gr and Cu are subunits recognized by some authors on the basis of cytoarchitecture, connectivity and physiological responses. These include: ' l) the central "cluster" region. In most mammals, to a varing degree, the cells in the central core of both the Cu and Cr exhibit a specific dendritic and cellular arrangement which gives the appearance of local aggregates of cells separated by thin axonal fascicles, called variously lobules, clusters, bricks, or nests (cat, Kuypers & Tuerk ’64, Hand '66; monkey, Albright '78, Albright & Haines '78; raccoon, Johnson, Welker & Pubols '68). These clusters have long been considered the primary thalamic projection region of the dorsal medulla (Lund & Webster '67, Cheek, Rustioni & Trevino '75, Hand & van Winkle '77) and receive the densest innervation from the dorsal root fibers (cats, Rustioni & Macchi '68, Hand '66, Kuypers & Tuerk '64, Keller & Hand '70 ; monkeys, Chang & Ruch '47, Ferraro & Barrera '35, Albright '78, Albright & Haines '78). 2) A caudoventral subunit containing large fusiform cells (Kuypers & Tuerk '64), the so called basal cells. This region receives, in addition to a sparse dorsal column input, a cortical input (Kuypers & Tuerk '64, Weisberg & Rustioni '79), a dorsolateral column (Spinal) input (Rustioni & Kaufmann '77, Rustioni '74, '77, Rustioni & Molenaar '75, Miller & Basbaum '76). 3) A rostral subunit containing many different sizes and shapes of 12 cells including some very large ones. These cells are not organized into clusters (Kuypers & Tuerk '64). This region also (like the caudoventral) receives sparse dorsal column innervation, dense dorsolateral column and cortical input (cat: Kuypers & Tuerk '64, Rustioni & Kaufmann '77, Weisberg & Rustioni '79; monkey, Rustioni, Hayes & O'Neill '79). In addition it receives input from other subcortical sources (red nucleus, Edwards '72; reticular formation, Kuypers '60, Sotgui & Marini '77). Some of these cells project to the thalamus but many project to extra-thalamic targets (e.g. cerebellum, Warren, Rowinski, Maliniak, Haring & Pubols '80, Cheek et al. '75, Rinvik & Walberg '75; other brainstem nuclei, Berkley '75, Hand & van Winkle '77; tectum, Berkley & Hand '78, spinal cord, Burton & Loewy '77). The external cuneate (ECu) nucleus is characterized by the presence of very large multipolar cells located in the dorsolateral aspect of the medulla, rostral to the obex. The ECu receives its major input via the dorsal columns and this input originates in the cervical and upper thoracic dorsal root ganglia (Rustioni & Macchi '68). Cell groups z and x, first described by Brodal & Pompeiano ('57a) are characterized by their small cell size and varying cell shape (in contrast to the neighboring nuclei except Gr) as well as their connectivity. Cell group x, bordered laterally by the restiform body, rostrally by the descending vestibular nucleus and caudally by the large cells of the ECU does not receive primary vestibular nerve input (as contrasted with the DVN) nor dorsal column input (as contrasted with the ECu) but does receive heavy input from the dorsolateral column (cats: Brodal & Pompeiano '57, Rustioni & Molenaar '75; monkey, 13 Albright & Haines '78). It projects to the cerebellar cortex and to higher levels of the brainstem (Brodal & Pompeiano '57a, b, Brodal '81). Cell group z has similar cytoarchitecture to the rostral of the Gr, however is is separated from Gr by a narrow cell free pole strip and receives a major input from the dorsolateral column (cat: Rustioni '74, Rustioni & Molenaar '75; monkey: Albright '78). It projects, not to the cerebellar cortex as does cg x but to the rostral pole of the VB complex (Grant, Boivie & Silfvenius '73). Physiology. The cluster region of the CuGr is the most heavily studied physiologically. This subunit responds primarily to very light tactile stimulation of the glabrous surfaces of the fore- (Cu) and hindpaws (Gr) (cats, Dykes, Rasmusson, Stretavan & Rehman '82; raccoon, Johnson et a1. '68). The basal region of the CuGr is very difficult to study physiologically. Miller & Basbaum ('76) and Dykes et al. ('82) in cats reported a high number of responses to stimulation of the deep tissues but this was not reported in raccoons (Johnson et al. '68). The rostral region of these nuclei exhibit larger receptive fields than the cluster region (Kruger, Siminoff & Witkovsky '61, Kruger '61, Perl, Whitlock & Gentry '62, Winter '65, Johnson et al. '68, Dykes et al. '82) and response modality segregation is difficult to detect except in the region where there is a grouping of large cells, similar in appearance to those of the ECu, and these respond to stimulation of- the deep tissues in cat (Dykes et al. '82) and raccoon (Johnson_et al. '68). 14 Medullary Submodality Segregation Recently the hypothesis has been advanced that the segregation of submodalities seen at the cortical and thalamic levels is maintained throughout the ascending somatosensory pathway (Dykes, '82, Johnson et al. '82, Berkley '80). To substantiate this hypothesis, it will be necessary to demonstrate l) segregation of submodalities in the medulla, and 2) the connections of these segregated regions with the corresponding subunits at the thalamic and cortical levels. To my knowledge the degree to which submodalities are segregated in the medulla of monkeys has not been investigated using modern techniques although a crude separation of muscle versus cutaneous may been made between the ECu and Cu nuclei. The most recent mapping study of the dorsal medulla in cats (Dykes et al. '82) indicates that there are discrete areas in the dorsal medulla responding in separable ways (i.e. fast adapting versus slow adapting) to cutaneous and deep stimulation. Within the cluster region of the dorsal column nuclei, these authors found that the cutaneous slow adapting (SA) and fast adapting (FA) responses were intermixed (as reported previously by Kruger et al. '61, Gordon & Jukes '64s, b). The deep responses however, were somewhat more segregated. The deep SA responses were limited to the ECu and adjoining rostal subunit of the Cu. The deep FA responses were more diffusely organized and were found in the rostro-medial ECu and the ventral portions of the Cu. _Medullary Muscle Afferent Segregation Cells in ECu physiologically respond mainly with the fastest 15 propogated evoked potentials in sensory nerves supplying the muscles of the forelimb and neck (Ia afferents) and project to the ipsilateral cerebellar cortex (Cooke, Larson, Oscarsson & Sjolund '7la, b). In addition to the ECu, cg z and x have also been reported to respond to electrical stimulation of muscle afferent nerves innervating the periphery of the extremities (both hind and fore limbs. A hindlimb muscle afferent pathway from the medulla to the cerebral cortex of the cat has been described physiologically (Langren & Silfvenius '69, '70, '71). This pathway was said to project from the dorsolateral funiculus via the cg z in the medulla to an ill-defined (stereotaxic coordinates only were given) region of the thalamus to an equally ill-defined region of "sensorimotor" cortex. Jones & Porter ('80) interpreted this cortical area as corresponding to area 3a. Grant, Boivie & Silfvenius ('73), using the Fink Heimer technique following lesions in the dorsal medulla (only one of which was confined to cg z) described the projections of a medullary relay nucleus for muscle afferents to the thalamus. These projections terminated in the caudo-lateral part of VL, immediately adjacent to the rostral dorso-lateral part of VB. Obviously when anyone draws a line on a section describing "the border" between two nuclei (in this case VL and VB) there is some degree of uncertainty. In a more recent article, Hendry et al. ('79) showed that the caudolateral portion of VL does not receive deep cerebellar nuclear input nor does it project to area 4 of the cortex. This region was also identified as analagous to that region receiving spino and cervicothalamic input by another author (Boivie '71a) and not to include regions of VB to which the gracile nucleus (Boivie '71b) and the cervicothalamic fibers (from the lateral cervical nucleus, Boivie 16 '78) project. A forelimb muscle afferent pathway to the cortex has also been described physiologically in the cat (Rosen '69a, b, Rosen & Sjolund '73s, b, Rosen & Asanuma '73). Again with scanty histological evidence, it appears that the rostro-ventral portion of the Cu (Oscarsson & Rosen '63) receives muscle afferent projections via the dorsal columns and projects via the thalamus (site undeterminable from these data) to the cerebral cortex (apparently to both preand postcruciate cortex). Two major problems with these studies (both the hind and forelimb muscle afferent projections) needing resolution are: l) the majority of these studies are strictly physiological, with recording sites determined by electrical nerve stimulation at the periphery and antidromic electrical stimulation at projection sites. Dykes et al. ('82) state two problems with electrical stimulation of peripheral nerves with regard to the study of stimulus submodality segregation. These were that the axon diameters and conduction velocities overlap so extensively that stimulation of single submodalities in mixed nerves is essentially impossible and that unusual inputs (i.e. those not demonstrable by natural receptor stimulation) can be evoked by electrical stimulation (c.f. Dykes & Gabor '81, Dostrovsky, Jabbur & Millar '78). And 2) histological verification of these projections, preferable with some physiological confirmation of the neuronal responses of the regions experimentally under study is necessary. The meager histology presented in these studies of projection sites consists, at best, of the location of the stimulating electrode, which of course could be activating fibers of passage. 17 Also the standard mapping procedures using natural peripheral stimulation allow a much larger region to be explored as well as allowing the testing of many submodalities simultaneously. In this fashion, one does not decide a priori which submodality or receptor sites will be studied. Using these methods, it can be said that not only cg z, the rostral Cu and Ben receive muscle afferent projections but also the ventral and caudal portions of the CuGr (cat: Dykes et a1. '82; tree squirrels: Ostapoff, Johnson & Albright '83). Where these latter cells project is now unknown. Recently Boivie and coworkers described an external cuneothalamic pathway in the monkey. (Boivie, Grant, Albe-Fessard, and Levant '75, Boivie and Bowman '81). The ECN has long been known to receive a massive group 1 afferent input (Cooke et a1. '7lb). Therefore this pathway is a potential source for the pathway to the muscle afferent rostral cap of the VB thalamus. From the available evidence it may be hypothesized that the muscle afferent projections, at least some of them, may be segregated within the three levels of the somatosensory system and that this case of submodality segregation may be the most readily demonstrable using current physiological and anatomical techniques. Significantly, little if anything is known about the brainstem connections to the muscle afferent region in the thalamus of any species. MUSCLE PROJECTIONS AS THE TESTABLE CASE OF SUBMODALITY SEGREGATION The muscle afferent projection system has been chosen for study as it appears to be the best candidate for demonstrating submodality 18 segregation. As mentioned above there is suggestive evidence that evoked responses to stimulation of the deep tissues of the postcranial body occur preferentially clustered together as do responses to stimulation of the skin at each level of the somatosensory system. There are also anatomical distinctions in cortex as well as in the medulla which can be correlated with this response submodality segregation. In addition, it is easier to discriminate between pathways subserving receptors located in the deep tissues and those conveying the cutaneous submodalities than to discriminate between the cutaneous submodalities themselves. The muscle sensory receptors (e.g. golgi tendon organ, muscle spindles, etc) will respond to a variety of stimuli (e.g. deep pressure, joint rotation, muscle stretching, etc.) but as a class they are separable from the majority of the cutaneous receptors on the basis of stimulus intensity. Many muscle sensory receptors respond to deep pressure (substantial indentation of the skin) while most cutaneous receptors respond to very light tactile stimulation (little or no indentation of the skin) in intact preparations. Muscle dissection (e.g. Maendly et a1. '81) allows one to selectively stimulate individual muscles and tendons (the connective tissue does however allow for some transfer of the stimulus to neighboring structures). Obviously large scale dissections are inappropriate in experiments in which the animal is expected to recover from the anesthetics for the duration of the survival time and were not used here. The animal cannot remain under the anesthetics during the entire survival time as this inhibits transport (personal observation). There are three reasons why choosing the muscle afferent stimulation may discriminate submodality pathways better than one of 19 the cutaneous submodalities. 1) Not all cutaneous receptors have been identified with respect to their response characteristics and adequate stimuli (i.e. the nature of the stimulus which best elicits responses). 2) Those cutaneous receptors which have been adequately described in terms of their response characteristics to specific stimuli show large overlap in both the stimulus which will elicite responses as well as the properties of that response. For instance, pacinian corpuscles follow high frequency, low amplitude sine waves applied to the skin (tuning points in excess of 100-200 Hz) and have broad response fields without distinct boundaries but have an obvious focus of maximal sensitivity (Burgess & Perl '73). Simple dermal corpuscles have a rapidly adapting response to small displacements of the skin but do not follow high frequency stimulation and have small response fields with distinct boundaries (Munger & Pubols '72). There is considerable overlap in stimulus parameters which will activate, albeit not optimally, several classes of receptors (Pubols '80). 3) The projections of the cutaneous submodalities appear to be more complexly organized (i.e. each response subunit appears smaller and the subunits are more intermixed) than do the muscle projections (Dykes et al. '82, Douglas, Ferrington & Rowe '78). The arrangement of the cortical cutaneous subunits also appears to vary considerably from species to species, e.g. the location of the hairy digit representation versus the volar digit representation (c.f. Carlson & Welt '80). A Segregated Muscle Afferent Pathway Within the Functional Organization of the Raccoon Somatosensory System Cortex. In the raccoon sensory cortex we have already determined 20 (Johnson et al. '82) that the functional organization of the rostral reaches of SI in the hand area is organized into three subdivisions. These were- designated, in rostrocaudal sequence, MAc (for the area of cortex responding to deep stimulation, including muscle afferents) , Hc (for the area of cortex responding to stimulation of digit claws (one or more) or more than one digit, either hairy or volar stimulation, and Vc (for the area of cortex responding to cutaneous stimulation of the volar forepaw glabrous skin pads and exhibiting a precise detailed somatotopic organization) on the basis of the submodality of the peripheral stimulus necessary to evoke responses. Rostral to the MAc region is the classical motor cortex, Mc (Hardin et al. '68). These designated regions can be roughly correlated with the cortical architectionic areas defined in other species. Thus, area 4 is analagous to Mc; 3a to MAc; and area 3b with Hc and Vc. Thalamus. More recent work in progress in our laboratory indicates the thalamus of the raccoon, as in the squirrel monkey (Dykes et al. '81) and macaque monkey (Maendly et al. '81), there exists at least a rostral cap to the ventrobasal complex which responds to peripheral stimulation of the deep tissues of the body (including the muscles, Wiener, Johnson & Ostapoff '82). This appears to represent a muscle afferent region segregated from the largely cutaneous representation found ventro-caudally in most of the rest of the VB complex. Medulla. At the medullary level in raccoons there is also physiological evidence for the segregation of stimulus modality (Johnson et al. '68). Although that study was not designed to 21 specifically segregate responses on the basis of stimulaus submodality, nonetheless it was noted that the majority of responses to stimulation of the deep tissues of the post-cranial body were segregated spatially to the external cuneate nucleus (ECN) and to'a "medial tongue" (MT) of large multipolar cells extending from the ECN medially along the ventral border of the main cuneate nucleus. These nuclear regions may correspond to all or part of the hypothetical medullary muscle afferent area. Separate from these muscle representations, within the cuneate-gracile nuclei (CuGr), responses from stimulation of the volar skin of the hands and feet were found mainly within the central "lobule" regions of the CuGr, while the responses from the claws and hairy skin stimulation were found more externally in the CuGr. OBJECTIVES OF THESE STUDIES To begin to understand the functional significance of any system in the brain, detailed information is necessary on the anatomical relationships within that system as well as the physiological properties of each level of organization. It is now known that muscle afferent information reaches various cortical areas, presumably subserving different functions in each area. Further, these cortical areas appear to be reciprocally connected with specific subunits of the ventral nucleus of the thalamus. At the level of the thalamus, the anatomical and physiological relationships between these subunits are only beginnning to be worked out. The anatomical projections to these thalamic areas are virtually unknown. The body of data describing the physiology of medullary nuclei which may be involved in these pathways is therefore difficult 22 to interpret, especially in light of the several subunits or areas already known in the thalamus and cortex (best described in the monkey). It is important therefore that specific nuclear groups projecting to discrete, physiologically identified thalamic and cortical subunits be identified. Once the projections from the medulla to cortex via thalamus are known then the physiological characteristics of each of the levels within these pathways may be fully studied: the interactions between levels can be analyzed, and related to the functional significance of each of the separated pathways. Intra-system submodality segregation may be better demonstrable in raccoons than in other commonly used non-primate species because of the highly elaborated ascending somatic sensory pathway in this animal (Welker & Seidenstein '59, Welker & Johnson '65, Welker, Johnson & Pubols '64, Johnson et al. '68) which is correlated with the raccoon's relatively high manual dexterity (Welker '69). These experiments were designed to determine the medullary input to one of the segregated zones in the VB thalamus; the one most accessible to study at this time, the segregated zone of projections from deep tissues of the forelimb which includes afferents from the muscles. Possible Clinical Significance of the Proposed Studies A potential application of the study of the muscle afferent projection pathway to the cerebral cortex is in the study of Parkinson's Disease induced tremors. Ohye and his co-workers in Japan have been sucessful in controlling drug resistant muscle tremors in humans by placing stereotaxic lesions in a restricted portion of the 23 thalamus in man, called Vim. This region is characterized by group 1 muscle afferent input and lies near the boundary between VL and VPLc (Cooper, Samra & Bergmann '69, Ohye, Fukamachi, Miyazaki, Isobe, Kakajima & Shibazaki '77). Presently the only animal model available is the monkey (Ohye, Imai, Nakajima, Shibazaki & Hirai '79). It would be of some benefit if an animal such as the raccoon could be developed which had sufficient similarity to the human condition but without the expense of primates in elucidating the anatomical and functional characteristics of the pathway controlling the drug resistant muscular tremors resulting from this disease. CHAPTER II. COMPARISON OF CELLS IN THE MEDULLA PROJECTING TO THE THALAMUS VERSUS THOSE PROJECTING TO THE CEREBELLAR CORTEX CHAPTER II. COMPARISON OF CELLS IN THE MEDULLA PROJECTING TO THE THALAMUS VERSUS THOSE PROJECTING TO THE CEREBELLAR CORTEX 24 INTRODUCTION The dorsal column nuclei (DCN), at the spino-medullary junction, wherein fibers of the spinal dorsal 'columns terminate, have been segregated into two major subdivisions according to the destination of their output: the cuneate-gracile nuclear complex (CuGr) projecting to thalamus and the external cuneate nucleus (ECu) projecting to cerebellum. Precise boundaries between CuGr and ECu have never been clear. Some nuclear regions in the vicinity of this vague boundary have been reported to project to both thalamus and cerebellum (in raccoons, Johnson, Welker & Pubols '68, Haring '81, Haring & Rowinski '82). This raises questions about the simple division into two nuclear regions with distinct output pathways. Should we recognize additional subdivisions with diverse outputs? Or is the segregation by output less than total for any of the nuclear subdivisions? Several arrays of further subdivision of the dorsal column nuclei have been proposed, based upon cytoarchitecture, corticobulbar input and connections to and from spinal gray matter. The purpose of the present study is to establish the distribution of cells projecting to thalamus, in comparison with that of cells projecting to cerebellum, and to relate these distributions to cytoarchitecturally recognizable subdivisions in the dorsal column nuclei of raccoons. Questions concerning projections to other targets. (spinal cord, inferior olive, tectum, pretectal nuclei) are net addressed here, and remain for future study. 25 METHODS Fourteen raccoons were used in this series of experiments. All animals were imobilized with an intramuscular injection of 1.0 ml (100 mg) Ketamine (Vetalar) to facilitate subsequent anesthetization by intraperitoneal injection of dial (45 mg/kg)-urethane (180 mg/kg). For the thalamic injection group (n=8), the muscles overlying the dorsal skull were retracted and a 75 mm hole bored in the skull to expose the cerebral cortex overlying the thalamus. The head was then positioned and cemented in place in the stereotaxic planes using a special headholder designed to align the center of the external ear canal and the inferior margin of the ocular orbit in the horizontal plane. In each case sufficient exploratory microelectrode mapping penetrations were made in order to determine the position of the Ventrobasal complex (VB). The recording electrode was then replaced by a 5 ‘ul syringe whose tip was lowered to the apprOpriate coordinates and 1.0-4.0.ul of 20% horseradish peroxidase (HRP, containing approximately equal amounts of Sigma VI, Bohringer-Mannheim type I and Miles brands dissolved in tris buffer pH 8.3 with .025 M KCl and 3% lysophasphotidyl choline) was delivered in 0.1-0.2‘ul increments with 5 minutes between increments. For the cerebellar injection group (n=6) the muscles attached to the occipital pole of the skull were retracted and the cerebellar cortex was exposed from the midline to approximately 1 cm lateral. The skull was then positioned in the stereotaxic planes as before. The needle of the 5‘ul syringe was aligned horizontally and introduced into 'the cerebellum at the vermal-paravermal junction and advanced rostrally to a point lying in the intermediate portion of the anterior lobe 26 (lobules IV and V). Injection of 1-3.0 ul were made as in the thalamic injections. In addition a series of 3 to 5 smaller surface injections (0.1-0.2 pl each) were made up to 1 mm deep into the dorsal aspect of the paramedian lobule under visual guidance ipsilateral to the anterior lobe injection. Following injection, all the animals were allowed to survive for 2-4 days (4 days proved optimal for retrograde transport) and then perfused with 500 ml of 0.9% saline followed by 2-4 liters of 1.25% glutaraldehyde, 1% paraformaldehyde in a .1 M phosphate buffer (pH 7.3) followed by 2 liters of 3% sucrose in the phosphate buffer. Brains were removed, blocked and infiltrated at 4 degrees C with 30% sucrose in the buffer for 2-3 days prior to frozen sectioning at 40 or 60 um in one of the stereotaxic planes. Series of alternate sections were processed with either tetramethylbenzidine (Mesulum & Mufson '80) counterstained with neutral red, or cobalt-intensified diaminobenzidine (Adams '77) counterstained with thionine. All the sections were systematically searched for HRP positive cells at a magnification of 125x and representative sections at five levels through the dorsal sensory medulla were chosen for illustration. The distribution of labeled cells in the dorsal medulla was plotted using a Zeiss drawing tube at a magnification of 50X. These drawings were then transferred to a standardized series of section drawings traken from one animal to facilitate comparisons. Nuclear subregions were based on descriptions of the medulla in the literature of both the raccoon (Johnson et al. '68) and the cat (Kuypers & Tuerk '64, Brodal & Pompieano '57). In one animal from each group (cerebellar injected and thalamic injected) counts were made of labeled and total numbers of cells within \ 26 (lobules IV and V). Injection of 1-3.0 pl were made as in the thalamic injections. In addition a series of 3 to 5 smaller surface injections (0.1-0.2 pl each) were made up to 1 mm deep into the dorsal aspect of the paramedian lobule under visual guidance ipsilateral to the anterior lobe injection. Following injection, all the animals were allowed to survive for 2-4 days (4 days proved optimal for retrograde transport) and then perfused with 500 ml of 0.9% saline followed by 2—4 liters of 1.25% glutaraldehyde, 1% paraformaldehyde in a .1 M phosphate buffer (pH 7.3) followed by 2 liters of 3% sucrose in the phosphate buffer. Brains were removed, blocked and infiltrated at 4 degrees C with 30% sucrose in the buffer for 2-3 days prior to frozen sectioning at 40 or 60 um in one of the stereotaxic planes. Series of alternate sections were processed with either tetramethylbenzidine (Mesulum & Mufson '80) counterstained with neutral red, or cobalt-intensified diaminobenzidine (Adams '77) counterstained with thionine. All the sections were systematically searched for HRP positive cells at a magnification of 125x and representative sections at five levels through the dorsal sensory medulla were chosen for illustration. The distribution of labeled cells in the dorsal medulla was plotted using a Zeiss drawing tube at a magnification of 50X. These drawings were then transferred to a standardized series of section drawings traken from one animal to facilitate comparisons. Nuclear subregions were based on descriptions of the medulla in the literature of both the raccoon (Johnson et al. '68) and the cat (Kuypers & Tuerk '64, Brodal & Pompieano '57). In one animal from each group (cerebellar injected and thalamic injected) counts were made of labeled and total numbers of cells within \ 27 the nuclear subdivisions recognized in this study in several adjacent sections (3-6). The subregions counted included the cluster, rostral and basal cuneate, the external cuneate and its medial tongue, and the cell group x, compact and reticular portions. No fewer than 115 and up to 300 cells were counted for each sample. The percent of the cells labeled is simply the number of labeled cells divided by the total number of cells. These numbers are used to indicate relative density of projections, not as an absolute number of cells projecting to either the cerebellar cortex or the thalamus. RESULTS Subdivisions of the Dorsal Mechanosensory Medulla (Figure 1.1) Nuclei in the dorsal mudulla reported to receive mechanosensory projections include the Cuneate and Gracile nuclei (CuGr), the External Cuneate nucleus (ECu) and cell groups z and x (cg z, cg x). The CuGr have been further subdivided on the basis of cytoarchitecture (Cajal '09, Kuypers & Tuerk '64, Hand '66, Johnson et al. '68), corticofugal projections (Kuypers & Tuerk '64) and non-primary spinal afferents (Rustioni '74, Rustioni & Molenaar '75). These subregions include a central cell cluster region (cCu) which in the raccoon shows cytoarchitecture and connections to those in the cat (Cajal '09, Kuypers & Tuerk '64, Ellis & Rustioni '81) except that the clusters are very large and well separated by fiber fascicles (Johnson et al. '68). Other subregions of the CuGr described in the cat apply to the raccoon as well. The rostral pole of the cuneate nucleus (rCu) in 28 Figure 1.1 Atlas of the nuclear subdivisions of the dorsal mechanosensory medulla. The nuclear subregions of interest to this study are shown here in these photomicrographs of transverse sections through the dorsal medulla. These are taken from one of the animals receiving a large injection of HRP into the thalamus (animal no. 51OL). These frozen sections were reacted with DAB and counterstained with thionine. The levels shown here are approximately those drawn in the summary diagram (Figure 1.12) and their rostrocaudal locations are shown in the inset, top, right. Bar equals 1 mm. A,A'. At a level aproximately 3.5 mm rostral to the obex the tight clustering of cells in cg x-c is apparent lateral and dorsal to the descending vestibular nucleus. Ventral and lateral to the cg x-c is the rostral pole of the ECu. A higher magnification of the area indicated by the box in A' is shown in Figure 1.2 A. B,B'. Approximately one half mm caudal to the previous section the loose arrangement of cells in the cg x-r is seen in approximately the same location. A higher magnification of the area outlined by the box in B' is shown in Figure 1.2 B. C,C'. This section, approximately 1 mm rostral to the obex, shows the rCu, ECu and Mt subdivisions of the DCN. D,D'. The division between the cCu and bCu is quite apparent in this figure taken at a level 1.5 mm caudal to the obex. Note the relative lack of unlabeled cells in the cCu while only some of the cells in the bCu appear labeled. The polymorphous ring external to the cCu is indicated in D' by the dotted line. E,E'. By 2.9 mm caudal to the obex the clustering of cells in the CuGr is much less obvious. This figure shows labeled cells in the caudal portions of the Cu, Cr, bCu and central cervical nuclei. Figure 29 AI Abbreviations: central cervical n. cuneate n. basal region of Cu cluster region of Cu rostral region of Cu cluster region of Cu . external cuneate n. cell roup f rest: arm body cell group x-compact . cell group x- reticular descending vestibular n. Bl hypoglossal n. DI I ’ ‘ \ I I \ cCu \ cGr 1 bCu 44 El 120 ‘9’ o 28 O 3 1.1 Figure V I: u‘ ‘w-l‘d ...!- 31 raccoon as well. The rostral pole of the cuneate nucleus (rCu) in raccoon is somewhat smaller than that in the cat but as in the cat it contains a polymorphic collection of cell types including large and small triangular, multipolar and fusiform cell bodies (Keller & Hand ’70). Another subregion of the cuneate nucleus of interest here is the basal Cuneate (bCu), the polymorphic region lying beneath the cCu as described in the cat (Cajal '09, Kuypers & Tuerk '64). This region, in contrast to the more dorsal cCu receives corticofugal projections (Kuypers & Tuerk '64) in the cat. Physiological mapping studies in cat (Dykes, Rassmusson, Stretavan & Rehman '82) and raccoon (Ostapoff & Johnson '83c) have shown that this region receives projections from the deep tissues of the forelimb). Unlike the cat but similar to the monkey (Ferraro & Barrera '35s, b), along the ventral border of the rCu in raccoon is a group of large multipolar cells, similar in morphology and physiological properties to the ECu (Johnson et al. '68, Haring '81, Haring & Rowinski '82) to which this group of cells is connected by a cellular bridge. This cell group was called the medial tongue (Mt) extension of the ECu by Johnson et al. ('68). Cell groups x and z in the raccoon form a continuous cell zone lying between the vestibular nuclei medial and rostral and the rCu and ECu lateral and caudal. We have subdivided this zone into three parts: cell group z lying caudomedial (cg z), cell group x-reticular (cg x-r) lying intermediate between cg z and cell group x-compact (cg x-c) lying rostrolateral. These subdivisions may also be seen in the cat though cg x—r is much less obvious (personal observation). Cg z in the 32 raccoon appears much smaller in mediolateral extent than reported in cat (Brodal & Pompeiano '57s, b) and galago (Albright '78, Albright & Haines '78) thOugh it Occupies a similar position just rostral to the Gracile nucleus and caudal to the caudal poles of the medial and descending vestibular nuclei. Cg x is located, as in the cat, dorsal to the rostral pole of the ECu and in raccoon extends medial and caudal, dorsal to the descending vestibular nucleus. In the raccoon two subregions of differing cytoarchitecture are discernable. The mediocaudal two-thirds of the cell group is composed of scattered strings of cells lying within the fibers of the restiform body. This subgroup is here called cg x-r (Figures 1.1B, 1.2B). The rostrolateral one-third of cg x is composed of medium sized cells in a rather more compact arrangement. This subregion more closely resembles the description of cg x given by Brodal & Pompeiano ('57a). It is separated from the rostral ECu (ventrally) by a thin fiber fascicle and in horizontal sections is also separated from the descending vestibular nucleus in a similar manner. We will refer to this subregion as cg x-c (Figures 1.1A, 1.2A). Stedmann's medical dictionary defines the obex as "the point of the dorsal surface of the medulla oblongata that marks the caudal angle of the rhomboid fossa or fourth ventricle. It corresponds to a small transverse medullary fold overhanging the narrow lower end of the fourth ventricle between the two tuberculum nuclei gracilis" (PP. 973). In the raccoon this bridge of grey matter, bordered caudally by the separation of the two gracile nuclei and rostrally by the merging of the central canal with the fourth ventricle is aproximately 800 ‘um in the anteroposterior dimension, as measured from 40 um frozen sections. 33 Figure 1.2 High magnification photomicrographs to show cytological details of cg x-c and x—r. A. This is a high magnification photomicrograph (466 X) of the cg x-c taken from the same section as shown in Figure 1.1 A, at the location of the box in Figure 1.1 A'. Dorsal is up and medial to the left. Note that these medium sized cells are closely arranged and separated from the descending vestibular nucleus (bottom, left) by a relatively cell free stripe. B. This is a photomicrograph taken at the same magnification as Figure 1.2 A but from the same section as Figure 1.1 B, at the location of the box in Figure 1.1 B', to show the cells in the cg x-r. These cells appear somewhat larger and are obviously scattered within a fiber bundle. Most of the cells in this photomicrograph show HRP labeling and appear black. One cell (arrow) apears to have a single process labeled (presumably the axon hillock). Bar equals 50 pm. 34 Figure 1.2,, v“ ' .__ e ,_, -_ _i_---_.-s..._._._....-..- --t . _ , 35 For the purposes of describing the transverse levels illustrated in this paper, the obex, designated 0.0, will refer to any section in this 800 um. Levels rostral to the obex will be designated by a "+" sign followed by the distance from the midpoint of this grey matter bridge. Levels caudal to the obex will be designated by a "-" sign followed by the distance caudal to the midpoint of the bridge. Medullothalamic Projections Figure 1.3 presents examples of the thalamic sites of HRP injection. All the injections sites reported for this group were confined to one side of the thalamus and the resultant labeled cells observed in the medulla contralateral to the injection site unless specifically stated. Labeling Frequency as a Result of Large Thalamic Injections Centered in the Ventrobasal Complex. A dual labeling pattern of cells in the dorsal sensory medulla was observed following large injections of HRP in the thalamus (Figures 1.4, 1.5). In cg z and x—r (but not cg x-c, Figures 1.6 A, B) and the cluster region of the cuneate and gracile at least 85% of the total number of cells contained HRP reaction product. In addition, the labeling in these nuclei can be characterized as extremely dense, often obscuring the nucleus of the labeled cells. In contrast, the basal and rostral regions of the CuGr showed many fewer cells labeled (less than 30%). The large multipolar cells in the medial two-thirds of the ECu as well as similar cells in the entire Mt also showed relatively light labeling both in terms of frequency of labeled cells and density of 36 Figure 1.3 Reconstructions of large thalamic injection sites. Series of sections through the thalamus of two of the animals receiving large thalamic injections of HRP are shown here. The drawing at the top, left depicts a transverse section through the diencephalon with the appropriate locations of the sections drawn in the sagittal (A, 28-141) and horizontal (B, I-IV) planes. On the right (A) is the injection site for animal 510L (see Figure 1.4 for the distribution of labeled cells in the medulla). At the lower left (B) are drawings of horizontal sections showing the injection site in animal 597 (see Figure 1.5 for the distribution of labeled cells in the dorsal medulla in this case). In both animals the dense core of the injection site (as visualized with DAB) includes the entire lateral two-thirds of the thalamus, sparing the midline nuclei as well as the colliculi of the midbrain. DAB-thionine counterstain. The bar equals 1 mm. C. This is a photomicrograph of a sagittal section (28) through the center of the large thalamic injection in animal 510L (see Figure 1.3 A). This section was reacted with DAB and thionine counterstained. Obviously, the entire dorsoventral extent and the rostrocaudal extent of the thalamus to the level of the lateral and medial geniculate nuclei is involved in the injection. Bar equals 1 mm. 37 Figure Mal Honzo 38 Figure 1.3 Figure 1.4 Labeling in the dorsal mechanosensory medulla following large thalamic injections. Shown here is a series of transverse sections through the medulla of animal 510L (Figure 1.3 shows the thalamic injection site). Nuclear subdivisions are indicated by abbreviations. These drawings are taken from the sections shown in Figure 1.1. Nuclear subdivisions which had greater than 85% of the cells labeled (the cg x-r and the cCuGr) are shaded black. The left diagonal shading indicates nuclear areas in which 10-25% of the cells were labeled. These included the bCu, rCu, Mt and Ecu. The vestibular nuclei contained small numbers of labeled cells and these are shown by the dots. Numbered arrows on the right indicate the level of the horizontal sections shown in Figure 1.5. Bar equals 1 mm. 4o ' QB .\ /\ |64 120 _ 1. “SI cGr Cells Projecting To The Thalamus ‘ ' \V I >85°/. k .W <25°/s 44. ““ Cu 2] l <—3l mm l “\\\‘ ... 28 (1 ~CeC 41 Figure 1.5 Medullary cell labeling following large thalamic HRP injections shown in the horizontal plane. Shown here are two horizontal sections through the dorsal medulla of animal .597 (see Figure 1.3 A for injection site). The distinction between cg x-c and x—r (see also Figure 1.6) as well as that between the rCu and cCu (see Figure 1.7) may be seen in horizontal sections. The relative numbers of labeled cells are indicated as in Figure 1.4. Numbered arrows across the top indicate the level of the transverse sections shown in Figure 1.4. Bar equals 1 mm. 42 Figure 1.5 IIII > 859 Cells Projecting To The Thalamus 0 Nil ( 2534 43 Figure 1.6 Photomicrographs showing the distinction between the cg x-c and x-r in the horizontal plane. A. This is a horizontal section through the cg x from an animal receiving a thalamic injection (597). Rostral is to the left, and lateral is down. In this section portions of the cCu, rCu, ECu, cg x-r and x-c may be seen. Cells in the cg x-r are labeled following thalamic injections while those in the cg x—c are not. Bar equals 1 mm. This is better seen at higher magnification. B. Higher magnification of the area outlined in the line drawing A' containing portions of both the cg x-c and x-r. Many cells in the x-r are labeled while those in the x-c do not show label. This section was reacted with TMB and neutral red counterstained. Bar equals 100 um. 44 Figure 1.6 I 'I 01; 20.: f “l” :5 .,- . , . A: ’17.” sf - $53 33' 454%; 3‘5 as: . A595» ' "2’24” as". 5...: ..wg'é .‘i'W'QVWJ-sa $1.... ."r‘N'r ._ '52? .4 _.",'a;:‘:::" 323;. :g:‘ "5:3 a §:%":P ,o‘l", ,1 "z " . ... .1 .'-“.. ”"4" f' "- -" . ' I. D. .’ 1 5 ' 4' ’ :. . .l'n"l I .“l’ . ‘.. l . I . I ~ I . "1'3," .M‘ I! n I p": - O '. ' s ' g ' ‘ ’.a ,l’ N ‘ '- Kira" :a ' . ' I. . ... . 70;. In no ' o I. .' ~' : My -. - '. " \ o.\. .. ~ . . \ \ \‘ .', r \m . ...l‘.o 45 label (Figure 1.7 A-C). Bilateral Medullothalamic Projections In three cases (506, 586 and 597) very large complete filling of one side of the thalamus with HRP resulted in a very small number of labeled cells in the DCN on the side ipsilateral to the injection. These cells were found in the cluster region most often inbetween, not within the clusters or in the region near the border between the cuneate and gracile nuclei. Rostrally, in the cg z ipsilateral to the injection a very few HRP positive cells in the same cases were observed. We do not believe that these labeled cells were the result of HRP spread across the midline of the thalamus as in these cases the injection sites were clearly confined to one side. Other Sources of Medullothalamic Projections Within the rostrocaudal extent of the medulla examined in these studies, labeled neurons were observed following large thalamic injections in other nuclei as well as those in the dorsal mechanosensory medulla. Caudally these included the lateral cervical nucleus. Very heavy filling of nearly all the cells in the upper two cervical levels (only those levels were included in our tissue blocks) was observed. The Central Cervical nucleus also showed some scattered cell filling at these levels. Rostrally, we observed substantial numbers of filled cells in a cell group associated with the lateral reticular nucleus called the infratrigeminal nucleus (IFT) by Berman ('68). We also observed a few scattered labeled cells in the descending and medial vestibular nuclei. 46 Figure 1.7 Photomicrograph of horizontal sections showing cell labeling in the ECu after thalamic injection of HRP. A, A'. This is a section through the center of the ECu from the same animal as in Figure 1.6, approximately 600 pm ventral to show the location of some of the labeled cells in the ECu following thalamic injections. Figures 1.7 B,C are higher magnifications of labeled cells in the ECu at the locations indicated in A',below. Bar equals 1 mm. B. Two heavily labeled cells (arrow) and one lightly (double arrow) cell near the caudal pole of the ECu are shown here. C. Two heavily labeled cells (arrows), located more medially but still within the ECu, following large thalamic injections. B, C. Bar equals 50 pm. 47 Figure 1.7 Fizfmang‘fwt m - ~ - _ - . ‘ V .. .‘ . ’ , ' .’ F ,‘ - - .‘ . .a’ \ .- ' .... ’ . ens-a raw “1 v.23»: 4-1‘.‘ ., ;; _ 22...... 2: ‘ “A“ '~..‘.'"";. “wind-11* ". “*2“. - ‘6‘"; r...3:5..‘fv ' 'Tz'; . , ”=- ' a. : -. a -«3'.'v .~' ""-a “4:91,? «a: - - :--;~ - -.-: .." .'. -- ~ . ' m §~ ow..o;..5.rv. C ..w’ a .. 0 1.... ¢‘- 4' .-. . \ .y-J. -‘ 3", . ‘ ‘4. ajnfi’a‘id‘; .' *‘t‘i‘lffli °- .vIfi ‘ ' ‘._.;J.- ' .' . ’4 "'1." ' .‘r ’f a 0'. ‘w‘..’- 'I- q. _ - . . . ~01 ’1‘ : _.. '1 .. 9. .‘c. \ '0 o. t :4?" .95 53‘? 5::41“ " " ' “a" ‘17 r - M. -".0’ "5?"? -, ";..“\k.q3 "‘t.“?'91 fi‘cgm. 0" v ..r -u _ f "- ‘ ‘-"‘- . . ' ’7‘ ‘ '1': ("5‘ .' 5'.0°n‘1.' ’57? ”:3 *’ '2‘."": unfit)"; '7'”ng l: "a. ' ' ""‘ “cf". ' ". A“’ "3:? - . u-o “ " ‘ ...‘V I .o 3"!" _I/=_Ofl’ o - {... a ,. _: -l . , , ‘ .. .‘N .. 71-. .x@ . ,-.¢.. 9"2' , .. ' ‘350‘..dfi . . o . 4 . I t .1. $.13; Us... ;\.‘\«9“;. _ '§“ .. -i ‘ . ..‘-'I . n . _ .‘ ,. ..... .‘ Q' . .. ...: 3V. ‘5 ’2'"... On 0“; w r. ‘ . ‘ .- ’l' T ‘ ' .... r "‘3 (‘::.": ..‘E' .9. 3“- ..- 'g; *u:£ - ' _ 1“. nif- . ;- t . I .' t." .' \{‘ “-.h ‘ - :,.".?.C{“- A, '. :é"' I’ 3","' ‘9". ‘Q'v‘ ,jf‘a ' ' ‘: -' ‘_ '0 O r. o. ' -- 33.": '1 . ‘3’ .s‘ 'Q :.~,:"“ , xv H135} 1 —. (.- -- er. . .‘-. .. ':‘- '~ ~:- are? rpm-rm..- 4- " '1‘: ' O A .n' I'. 1' n . . I‘ " a" . " " '97:. . \ a 0 'J' Ifu' T a I" ‘ ‘Il . ..‘r: i .O n.,‘ ‘. .‘ f. r ’ .c ._ . :_. ‘ \‘.Ak.. .- ‘ : .. .- “lu‘ . ‘. - “-.' ,{ .... ..’ 1 ‘3”‘-‘-.-‘ ' _,, . "‘ ”bi-4 -. . ' ‘ .‘~: - .~ I. I, . ""..".. ;..c .“ -~ . ' ‘ '0‘ . ° . I ' , , Ml ‘1 l' _ .I . \ r «w... ~‘;-,.-~*'t:- 2: .~:-‘- . . . . . .04 ’- ~ . ' ... "u.';"‘."’ ‘ . . ‘ ' ... .. t . 'i“-*. '0 Q" 3‘? It, .“ ' c". .'- ‘ .’- ‘ gyro. 1r ‘ . . a. W} ‘0 - --".-r O . I‘ .‘ ' .‘a‘-=-".‘:-r'm-'--- - «has. :..;- . '1' '34!" "’.";\-g ’ if... 'fio‘. ’0 f. “ . r ’3." ‘ ...: . ' r‘. 7;" . 5 co'k-i'.” Up - :~‘."b‘"‘ . in". " " ‘ ." ‘- ~--‘.¢=~-»r.«.-s-~ ..., ' ..~. .. ‘5'?— .... .. - . ... '- -. +3 . 5 a" . ‘. g.“". ‘ 7"‘ ‘ 'o. ‘ l.. ‘. \ ’ ~ , . V. ‘.. . . - l g z . .’ 1‘ ...‘ 09‘ gy. .‘\ O . ._. - - . . » " .. . u ‘- ' o A r -' -= mm: tame -- - . . . .- .4. 3;, ., - .- ; .- ' . a 7 ' 'ofifl'qu‘a) \- ' "i“ . ”Q ‘ 5 r.“ ' ‘0 E3 .1 ' 1 . . '0 r . ‘. . . ‘~ ' - a .' a f. ’ . .. : ' . .o!‘ . . a :- o ‘ ‘ n. I. . ,4 .. 48 Medullocerebellar Projections The regions of the cerebellar cortex injected with HRP in these studies (Figure 1.8) were those reported in the cat to receive somatosensory projections (Cooke, Larson, Oscarsson & Sjolund '71 a, b). Hence the majority of label observed was located in the mechanosensory medulla. Again two patterns in the pr0portion of labeled versus unlabeled cells were observed, in these areas ipsilateral to the injection site (Figures 1.9, 1.10). The cells in the ECu, long known as a primary source of mechanosensory projections to the cerebellar cortex were observed to be virtually all filled with dense accumulations of HRP reaction product. The Mt extension of the ECu lying beneath the rCu was also very heavily labeled as was (Figure 1.11) a cell group in the caudoventral descending vestibular nucleus resembling that called cg f by Brodal & Pompieano ('57a). Fewer labeled cells (approximately 50%), though these were no less densely labeled, were observed in cg x-c. A small group of labeled cells in the cCu, near the caudal pole of the ECu were also observed. These were situated at the extreme lateral edge of the Cu, outside of the central cell clusters. Some scattered labeled cells were also observed in the external polymorphous ring surrounding the cell clusters, and in the rCu. Bilateral Medullocerebellar Projections In all the above cell groups (ECu, Mt, cg x and f, and the Cu group) labeled cells were observed on the side contralateral to the injection site, though with differing proportions of labeled cells and 49 Figure 1.8 Reconstructions of large cerebellar injections Two of the large cerebellar injections used in this study are shown here in series of transverse (animal 513, left) and sagittal (591, right) sections. The approximate location of these sections are shown in the diagram of the dorsal cerebellar surface in the upper. left. See Figures 1.9 and 1.10 for the distribution of labeled cells in each case. The injection sites in this figure are indicated both for TMB (diagonal stripe) and DAB (solid black) from adjacent sections. Note that the anterior injection sites do not extend to the midline of the vermis even in the TMB reacted sections. This was true for all the animals used in this group. Bar equals 1 mm. Figure 1.8 :42 82 52 A L+R o L p ‘ AAkP '“ Sogx’ttol v M 駧§fzgfffzé22 7" D -ronsverse 52+:- §HE E 66 ' ' 82 TMB DAB 126 142 51 Figure 1.9 Cell labeling in the dorsal mechanosensory medulla following large cerebellar injections. This figure shows a series of transverse sections through the dorsal medulla of animal 513. Areas in which 60-75% of the cells in these sections were labeled are indicated in black. These include the ECu and the Mt. The right diagonal shading indicates areas in which 17-33% of the cells were labeled. These include the rCu and cg x-c. In the rCu most of the largest cells in the subdivision were labeled, some smaller cells were also labeled. In the cg x-c the majority of labeled cells were in the lateral one-third of the cell group (in this one-third approximately half of the cells were labeled). Areas with only a few labeled cells are indicated by the dots. These included the vestibular nuclei, the polymorphous ring surrounding the dorsal and lateral edge of the cCu and the base of the cg x-r. These latter cels were found exclusively at the ventral boundary of the cg x-r where it borders on the descending vestibular nucleus and not in the cells strands lying more dorsally. The approximate level of these section is shown by the numbered arrows in Figure 1.10. Bar equals 1 mm. .52 Figure 1.9 CeHs Prmécflng To The Cerebenuw - >6O°/. <33°/. 53 Figure 1.10 Distribution of labeled cells in horizontal sections through the dorsal medulla following cerebellar injections. This figure show the distribution of labeled cells following cerebellar injections in animal 591 (see Figure 1.8 B for the injection sites). The frequencies of labeled cells in the nuclear subdivisions are indicated as in Figure 1.9 except in this case almost 80% of the cells in the ECu were-labeled: black 60-80%; diagonal shading 17—33%; dots less than 17% of cells labeled. ECu—Mt nuclear areas showed the highest frequency of labeled cells while the cg x-c and rCu were intermediate and the vestibular nuclei and polymorphous ring around the cCu showed a few scattered labeled cells. The approximate level of these sections is shown by the numbered arrows in Figure 1.9. Bar equals 1 mm. 54 Figure 1.10 261 229 ISI [05 Al‘s—P 21 l l l l dorsal ventral II. ) 609% Cells Pro‘e'cfin To The Cerebellum I g <33°/. Figure 1.11 Horizontal section illustrating the cellular bridge connecting the ECu and Mt. This is a horizontal section intermediate between those shown in Figure 1.10 from the same animal illustrating both the cerebellar projections of the cells in the ECu (laterally) and Mt (medially) as well as the cellular bridge connecting the two nuclear regions. This section was reacted with TMB and neutral red counterstained. Bar equals 1 mm. «..., 56 Figure 1.11 3.. ' ' - V - ‘ ' ’; n ' ... .1 .. .- . . ..‘ AWW -— ll .. U . ,I o ' ‘ "'.. ’ ." ' ' v a .’. ... "“".- x 6 a; o 4 . . fl, ...,n I . » d . w. . . a _ u- .a 0 ~'N fi g v- . . v 8'- v . 'I‘ o .6 _ ! fl . ‘ ‘ ‘ ...: l ‘ ‘ ‘ ~33???“ 524$? ~". " . c: I J . ’ '. . ~ ’. I .1” 5" v ' '4‘", *.O I. I . : ..\ - , v" . . ‘4‘". . . ._ _ ‘. ‘ 6 0 OJ . . ‘ . l . .{0 F") r... a . ' . . .. elf. ‘ ‘ i " I" ‘ I . a ‘ a \ ....s, g — . .. .. ‘4. .. .3 . . —' . I ‘ ' . u - ‘- « « {"" ‘ t ' I ‘ z p- ,. ~ ~x-luaw ' n o o" ' . , ‘_ ’ . . r‘ . .1 v‘ -". . E“... I ‘ . \ t ' . I. . $.‘ I .’.‘ 'fi.‘ 1 m. a -. . . . _ __ .v o 3 ‘w 0“. ‘0. '.‘ - i “ "o c w ‘ . ' .." . ”4:" ‘ " o- ” . ‘9 . .. .4 . n ' p p. a \ - .. . 0. .' b ‘ ‘ ‘ .c . . ...“ . ,9 . o- .a’ v... .\ : a... . ‘ . . a .' .3 . " , , - . u.‘ .... . “’2fiY'“” r. "~'.1.\-. ‘ ..-..px.. av ‘ >Y’r cg! ‘. v . ' . .9 .\..>\ " . . . '.\ 9- ~ we» *1 ~.. .gnmsowugz ..4 4 v ”p.. .-J. 7, ¢ ..p. o, _. ‘ -" :7 ...l-: -. . _,)' '_ (a “7 ..av.; ,_ . . . 07‘, - n." I“... ' M ‘ ... . 023. f l. ' v ‘ w‘ ('2‘: -"’".¢:"'l-. ' 9 '.~ ‘ t.’ (03". 1""... "o W v . o , on .F . \ .- P4 _.~ \ ~ 7 , . “ .- I u ".._'\‘ .‘ ' ' 1.)“ - $._;‘. . .. ...a, .‘ . .. '. ' ..V - ' ‘ I" ‘ . 9:; .' \. wake '- . . . . ‘ ' 57 density of cell labeling. A more detailed consideration of the bilateral distribution of these projections has been presented elsewhere (Ostapoff '82, Ostapoff & Johnson '83a). Other Medullocerebellar Projections Following these injections into the cerebellar ‘cortex labeled cells were observed in the lateral reticular nucleus of both sides, with the ipsilateral side showing more labeled cells. The inferior olive contralateral to the injection site was also heavily labeled. The descending and medial vestibular nuclei and the central cervical nucleus (the latter bilaterally) had a few scattered labeled cells in our material. DISCUSSION Cuneate Nuclear Projections to Thalamus The dual nature of the labeling observed in these experiments further justifies the parcellation of the major nuclei in the dorsal mechanosensory medulla into several subregions. The heavy labeling in the cluster region of the cuneate and gracile nuclei (cCu) (including the external polymorphous ring), long considered the primary thalamic projecting areas (Lund & Webster '67, Cheek, Rustioni & Trevino '75, Hand & van Winkle '77) is starkly contrasted by the relatively sparse labeling in the two other subregions of the CuGr considered here (i.e. the rCu and bCu). The rCu in the cat has been reported to project to a wide variety of non-thalamic targets (other brainstem nuclei, Berkley '75. Hand & van Winkle '77; tectum, Berkley & Hand .782 Robards '79; 58 spinal cord, Burton & Loewy '77). Considering the relatively few cells in this region labeled by very large thalamic injections of HRP, it would not be surprising if a similar heterogenous set of projections to non-thalamic targets from the rCu was present in the raccoon. The bCu has been reported to receive, in addition to dorsal column 'projections, a cortical input (Kuypers & Tuerk '64, Weisberg & Rustioni '79), and a dorsolateral funiculus input (Rustioni & Kaufmann '77, Rustioni & Molenaar '75, Miller & Basbaum '76). Projections from the bCu to the spinal cord and lateral cervical nucleus have been described (Burton & Loewy '77, Craig '78). It is clear from the data presented here that only a few cells in these two regions of the Cu project to the thalamus in the raccoon. A more precise localization of the terminal field of the bCu projections to the thalamus in the raccoon is presented elsewhere (Ostapoff & Johnson '83a). External Cuneothalamic Projections A projection to thalamus from the ECu has not been previously reported in carnivores. However such a projection has been recently reported in monkeys by Boivie and coworkers (Bovie, Grant, Albe-Fessard & Levante '75, Boivie & Bowman '81). The projection reported here for the raccoon appears to be somewhat more sparse (i.e. fewer cells were labeled) than that reported for monkey (Boivie et al '75). Whether or not, as Boivie and Bowman ('81) suggest,this pathway represents a phylogenetically recent development, its presence in the monkey and raccoon and. reportedly not in the cat would suggest a relationship of this pathway to the forepaw manipulative behaviors exhibited by the two former species. 59 Cell Group 2 and x Thalamic Projections These present studies confirm that in the raccoon most of the cells in cg z project to the thalamus as reported in the cat (Grant, Boivie & Silfvenius '73). The present experiments demonstrate that most of the cells in the cg x-r also project to the thalamus agrees in part with a report that some (2 of 8 reported) cells in an apparently similar location receiving hindlimb muscle afferent projections could be antidromically activated from the thalamus (Johansson & Silfvenius '77a, b). Medullary Projections to Cerebellar Cortex The association of the Mt region with the ECu in the raccoon is supported by the obvious heavy labeling of the majority of cells in the ECu, Mt and cells connecting the two, following injection of HRP into the somatosensory areas of the cerebellar cortex (Figures 1.9, 1.10). relative sparcity of labeled cells in the rest of the rCu in raccoon This contrasts somewhat with reports in the cat (Cooke et al. '71a). These authors found that cerebellar relay cells activated by cutaneous afferents were intermingled with lemniscal neurons. This lack of mingling of cerebellar projecting and thalamic projecting neurons in the raccoon rCu has been previously reported (Haring '81). In the latter report the majority of cells rostral to the obex, not in the ECu, projecting to the cerebellum were located in the Mt. A few cells in the polymorphous ring surrounding the cCu projecting to the cerebellar cortex in raccoon as described here have also been reported (Warren, Rowinski, Maliniak, Haring & Pubols '80). 60 Approximately one—third of the cells in the cg x-c in this study projected to the areas’ of the cerebellar cortex injected with HRP. None of the cells in the dorsal strings of cells of cg x were labeled following these injections, though some few of the cells along the border between the cg x-c and the descending vestibular nucleus were labeled. In the eat some cg x cells were reported to project to the somatosensory areas of the cerebellar cortex (Johansson & Silfvenius '77a, b) while Brodal & Pompeiano ('57a) describe a projection from cg x to the vestibular portions of the cerebellar cortex. We did not investigate this latter projection but the projection from cg x-c to our cerebellar injection sites was approXimately equal bilaterally. Cell Groups x and 2 as a Complex. Despite the "clear distinction" of cg x from cg z in the cat (Brodal & Pompeiano '57a, p. 446), in the raccoon we view these cell groups as forming a continuous band, caudomedial to rostrolateral, separating the nuclei of the dorsal columns (CuGr, ECu) from those of the vestibular complex (DV, MV). This region might therefore better be designated as a complex (the xz complex) much as we do for the cuneate and gracile nuclei. Segregation of the cg z from the cg x has also been made on the basis efferent projections. In cats, cg 2 has been shown to project to the thalamus (Landgren & Silfvenius '70,'71) while cg x projects to various areas of the cerebellar cortex (Brodal '81). The present study also shows that in raccoon, cg x—r, in addition to cg z projects heavily to the thalamus while cg x-c projects, in part, to the somatosensory areas of the cerebellar cortex (see also Ostapoff & Johnson '83a). The subdivision here called cg x-r projects, parallel to cg z, to the 61 thalamus but clearly lies within the regions shown as cg x in the cat (Brodal & Pompeiano '57a, text-figure 3; Johansson & Silfvenius 77c, text-figure 2, c.f. our Figures 1.4-1.6). A better designation for the cg x-r might be the intermediate reticular portion of the xz complex (cg xz-r). A more detailed consideration of this region is currently in preparation (Ostapoff & Johnson '83d). Nuclear Subregions Projecting to Both the Thalamus and Cerebellar Cortex Figure 1.12 shows a summary of the distribution and relative density of labeled cells following either cerebellar of thalamic injections. Three of the nuclear subregions identified in this study appear to contain subpopulations of cells which label following both large thalamic and large cerebellar injections. Rostral Cuneate. The polymorphic rCu is one of these subregions. In our cell counts approximately equal numbers of cells were labeled following both cerebellar injection (17%) and thalamic injection (20% . External Cuneate and Medial Tongue. The ECu and Mt also contain labeled cells following both cerebellar and thalamic injections. In this case however, the majority of cells in the ECu (greater than 70%) and Mt (greater than 60%) project to the cerebellar cortex while smaller numbers of cells in these subregions (9 and 21% respectively) project to the thalamus (Ostapoff & Johnson '83c). The few cells in the polymorphic ring surrounding the cCu and projecting to the cerebellar cortex are similarly scattered among the majority of cells projecting to the thalamus. 62 Figure 1.12 Summary of cerebellar projecting versus thalamic projecting cells in the dorsal mechanosensory medulla. This figure shows a series of drawings of transverse sections through the raccoon medulla. Left diagonal shading indicates areas in which cells project to the thalamus. Closely spaced diagonal lines indicate that the majority of the cells were labeled. These areas include the cg x—r and the cluster region of the CuGr. Widely spaced lines indicate that a minority of the cells in those areas project to the thalamus. These areas include the ECu (9%), Mt (21%), rCu (20%) and bCu (26%). Right diagonal shading indicates areas in which cells were labeled after our cerebellar injections. Closely spaced diagonal lines indicate areas in which at least 62% of the cells were labeled. The areas included the ECu (70%) and the Mt (62%). Widely spaced diagonal lines indicate areas in which fewer cells were labeled from our cerebellar injections. These areas include the cg x-C (33%) and the rCu (17%). Three nuclear subregions stand out as containing cells which label following both cerebellar and thalamic injections. These are the rCu, in which aproximately equal numbers of cells project to each target and the ECu-Mt in which the majority of cells project to the cerebellar cortex but a sizeable minority project to the thalamus (9% of the cells in the ECu and 21% of the cells in the Mt). Bar equals 1 mm. 63 Cells Projecting To The Thalamu versus Cells Projecting 'I To The Cerebellum/ 64 It is logically possible that some of these cells project to both the thalamus and the cerebellar cortex. Two studies designed to address this question have failed to demonstrate either anatomically (cat, Rustioni, Hayes & O'Neill '79) or physiologically (raccoon, Haring '81) that there are cells sending axon collaterals to both the cerebellar cortex‘ and thalamus. What may rather be the case is that some cells in regions projecting primarily to the cerebellar cortex may be sending a sample of this information to the thalamus (and presumably from there to the cerebral cortex; and vice versa, a sample of thalamically directed information similarly arrives in cerebellar cortex. The cg x is thought to be innervated by axon collaterals from the dorsal spinocerebellar tract in cats (Johansson & Silfvenius '77a). If this is case in raccoons, then functionally the cells in the cg x—r may serve an analagous sampling function for the information being conveyed by the tract. An analagous situation may exist in other subregions of the DCN known to project to non-thalamic targets. That is, those cells in the bCu and rCu projecting to the thalamus may represent the means by which cerebral cortex is informed about the integration of activity taking place at the medullary level and projected to non-thalamic targets. CHAPTER III. DORSAL MEDULLARY SOURCES OF PROJECTIONS TO THE KINESTHETIC THALAMUS 65 INTRODUCTION Evoked responses from stimulation of the deep tissues of the body can be recorded in the transitional cortex (area 3a) between the classical motor (area 4) and somatosensory (area 3b) cerebral cortices in monkeys (Phillips, Powell & Wiesendanger '71; Tanji & Wise '81; Wise & Tanji '81; Merzenich, Kaas, Sur & Lin '78), cat (Kaas, Nelson, Sur, Lin & Merzenich '79) and raccoon (Johnson, Ostapoff & Warach '82). A similar kinesthetic zone or shell has recently been described physiologically in the monkey thalamus (Dykes, Sur, Merzenich, Kaas & Nelson '81; Maendley, Ruegg, Wiesendanger, Wiesendanger, Lagowsky & Hess '81), and raccoon (Wiener, Johnson & Ostapoff '82). Reciprocal thalamocortical projections of this kinesthetic shell in the thalamus with area 3a cortex in the monkey have also been shown (Jones & Friedman '82; Jones, Friedman & Hendry '82). Little is known of the sources of these kinesthetic projections to the thalamus. Physiological studies using both natural and electrical stimulation of nerves presumed to carry muscle afferent fibers have reported several nuclear regions in the dorsal medulla that respond to kinesthetic stimulation. These include the caudoventral (or basal) Cuneate and Gracile nuclei (cat: Dykes, Rasmusson, Stretavan & Rehman '82; tree squirrel: Ostapoff, Johnson & Albright, in press); the ventral portion of the rostral Cuneate nucleus (cat: Dykes et al. '82; Rosen '69a,b, Rosen & Sjolund '73a,b; Rosen & Asanuma '75; raccoon: Johnson, Welker & Pubols '68); the External Cuneate nucleus (cats: 66 Cooke, Oscarsson & Sjolund 71a,b; Dykes et al. '82; raccoons: Johnson et al. '68; tree squirrel: Ostapoff et al., in press) and cell groups 2 and x (Landgren & Silfvenius '69, '70, '71). Anatomical demonstration that any of these nuclear regions responding to kinesthetic stimulation project to the kinesthetic- thalamus is somewhat limited. Grant, Boivie & Silfvenius ('73) using the Fink-Heimer _technique, following lesions in the dorsal medulla (only one of which was confined to cell group z) reported that projections from this area of the dorsal medulla terminated in the caudolateral part of the ventrolateral nucleus (VL), immediately adjacent to the rostral dorsolateral part of the ventrobasal complex (VB). In a more recent article, Hendry, Jones & Graham ('79) showed that the caudal part of the VL does not receive deep cerebellar input nor does it project to area 4 of the cortex. This region was also identified as analagous to that region receiving spino- and cervicothalamic input by another author (Boivie '70, '71a) and not to include regions of VB to which the gracile nucleus (Boivie '71b) and the cervicothalamic fibers (Boivie '78) project. A pathway from the ventral portion of the rostral Cuneate nucleus via the thalamus by which forelimb kinesthetic information may reach the cerebral cortex of cat was described by Oscarsson and Rosen ('63). Recently Boivie and coworkers described an External Cuneothalamic pathway in the monkey (Boivie, Grant, Albe-Fessard & Levant '75, Bovie & Bowman '81). Recent work in our laboratory (Wiener et al. '82) has shown that the raccoon VB has a large kinesthetic region as compared with the monkey (Dykes et al. '81, Maendley et al. '81, Jones & Friedman '82, 66 Cooke, Oscarsson & Sjolund 71a,b; Dykes et al. '82; raccoons: Johnson et al. '68; tree squirrel: Ostapoff et al., in press) and cell groups 2 and x (Landgren & Silfvenius '69, '70, '71). Anatomical demonstration that any of these nuclear regions responding to kinesthetic stimulation project to the kinesthetic. thalamus is somewhat limited. Grant, Boivie & Silfvenius ('73) using the Fink-Heimer (technique, following lesions in the dorsal medulla (only one of which was confined to cell group 2) reported that projections from this area of the dorsal medulla terminated in the caudolateral part of the ventrolateral nucleus (VL), immediately adjacent to the rostral dorsolateral part of the ventrobasal complex (VB). In a more recent article, Hendry, Jones & Graham ('79) showed that the caudal part of the VL does not receive deep cerebellar input nor does it project to area 4 of the cortex. This region was also identified as analagous to that region receiving spino- and cervicothalamic input by another author (Boivie '70, '71a) and not to include regions of VB to which the gracile nucleus (Boivie '71b) and the cervicothalamic fibers (Boivie '78) project. A pathway from the ventral portion of the rostral Cuneate nucleus via the thalamus by which forelimb kinesthetic information may reach the cerebral cortex of cat was described by Oscarsson and Rosen ('63). Recently Boivie and coworkers described an External Cuneothalamic pathway in the monkey (Boivie, Grant, Albe-Fessard & Levant '75, Bovie & Bowman '81). Recent work in our laboratory (Wiener et al. '82) has shown that the raccoon VB has a large kinesthetic region as compared with the monkey (Dykes et al. '81, Maendley et al. '81, Jones & Friedman '82, 67 Jones et al. '82). In this region we made small injections of horseradish peroxidase (HRP), closely guided by simultaneous physiological recording of evoked responses to peripheral stimulation, to determine the medullary sources of the kinesthetic projections. METHODS Nine raccoons were used for this series of experiments. Surgical and histological procedures were as previously described (Ostapoff & Johnson '83b). In each animal a grid of standard tungsten microelectrode penetrations (Johnson, Ostapoff & Wwarach '82) was made in the rostral third of the ventrobasal complex, specifically locating the kinesthetic region (KVB) with respect to the underlying cutanous projections. All injections reported in this study were made while simultaneously recording evoked potentials to peripheral stimulation of the forelimb. Target sites in the KVB fulfilled the following criterion with respect to modality of effective stimulation. The responses at the injection site could be elicited only by stimulation of the deep tissues of the forelimb. Deep tissue was defined in two ways: 1) either joint movement or visibly large indentation of the tissues was necessary to evoke maximal responses and 2) in areas of the body covered by loose skin (e.g. upper arm) the receptive field must have remained fixed with respect to the deep tissues as the superficial skin was displaced. The recording microelectrode was then replaced by an HRP filled glass pipette with a tip inner diameter of 50 to 80 um into which a moveable tungsten 'filament (diameter 30-70 um) was inserted so that 20-40 um of tungsten extended beyond the glass tip. This was then lowered into the KVB until evoked responses to 68 kinesthetic stimulation of the forelimb were recorded through the exposed tungsten tip. The tungsten was then retracted, allowing the HRP to be drawn by capillary action to the glass tip. This HRP was then expressed by re-extending the tungsten. At this time the receptive field could be reéexamined to ensure against inadvertant electrode movement. The size of the injection site could to some extent be controlled by: varying the the amount of time allowed for HRP filling of the tip and the number and frequency of tungsten withdrawal/extension cycles made. The injections were targeted to lie in either the medial or lateral halves of the KVB as determined by the underlying cutaneous representation (lateral half overlying the representation of digits 4 and 5, medial half overlying projections from digits 1, 2 or 3). In most cases the relationship between these kinesthetic responses to the underlying cutaneous projections was known either by advancing the injecting electrode into the cutaneous projections before injecting the HRP at more dorsal levels or by subsequent histological reconstruction of the pre-injection mapping penetrations. This allowed us to place the kinesthetic projections within the organizational and stereotaxic framework developed by us in a separate fine grain mapping study of the organization of projections in the KVB (Wiener et a1 '82) Following a survival time of 3—4 days, all animals except those selected for post-injection mapping of their medullas (see below), were perfused and their brains treated for routine HRP visualization as in previous studies (Ostapoff & Johnson '83b). All the sections through the dorsal medulla were systematically examined for labelled cells at a magnification of 125K and 69 reconstructions of the location of all labeled cells in every fourth section through the dorsal medulla were made with the aid of a drawing tube attached to a Zeiss microscope at a magnification of 50X. These were then fitted onto a standardized series of drawings of the raccoon dorsal medulla at six representative levels to facilitate comparisons. Previous studies in the raccoon have identified six sub—regions of the nuclei in the dorsal medulla other than the cluster region of the CuGr projecting to the thalamus (Ostapoff & Johnson '82b). These include the base of the cuneate nucleus (bCu), the heterogenous portion of the rostral Cuneate (rCu), the External Cuneate nucleus (ECu) and its medial tongue (Mt), and cell groups 2 (cg z) and x (cg x). In addition cell group x was further subdivided into a reticular portion (cg x-r) projecting to the thalamus and a compact portion (cg x-c) projecting in part to the cerebellar cortex (Ostapoff & Johnson '83b). Physiological Mapping of the Medulla The mapping data for the primary afferent projections to nuclei in the medulla of the raccoon (Johnson et al. '68) extensively describe projections to the Cuneate, Gracile and External Cuneate nuclei but do not include data from the base of the Cuneate nucleus (bCu) nor cell groups 2 (cg z) and x (cg x). In order to firmly establish that these nuclear areas do indeed convey kinesthetic information, on the third or fourth survival day following HRP injection into the KVB, four animals were anesthetized as as before and surgically prepared for mapping of their medullas in the regions of the bCu and cg z and x. The medulla and caudal cerebellar vermis were exposed and rows of electrode penetrations made in the region of the bCu rostral to the spinal 70 cord-medullary junction, and in the region of cg z and x (approximately 1 to 4 mm rostral to the obex and 3 to 5 mm lateral to the midline) with spacing of approximately .75 mm (between penetrations. These latter penetrations were made through the overlying cerebellar cortex so as not to damage the cg z and x which lie very close to the dorsal surface of the medulla. Small microlesions were made in the vicinity of kinesthetic responses so that the actual electrode tip location could be determined. rAt the conclusion of the post-injection medullary mapping experiments the animals were perfused and their brains treated similarly to the rest of the subjects in this study. RESULTS Use of Pre-injection Thalamic Mapping to Localize Injections The results of our fine grain mapping study of the projections to the KVB (Wiener et al. '82) were used to locate the injection sites in this study with respect to the overall patterns of projections to the VB. The physiological data as well as as the histological reconstructions of the pre-injection electrode penetrations when compared to the same data from the injecting pipette were considered in characterizing the location of the injection sites. This allows us to describe the injection sites reported here both in terms of their anatomical location and more importantly, in terms of their relationship to the observed organization of somatosensory projections to the VB complex. An example of this can be seen in Figure 2.1. Though the kinesthetic projections to the VB are somatotopically organized, the receptor fields are quite large (presumably due to our 71 Figure 2.1 Injection into the lateral KVB. This figure shows an example of the data used to accurately localize the injection sites in this study. At the upper right is a drawing of a horizontal section showing the approximate locations of the transverse sections drawn in A-D. This drawing is adapted from Welker and Johnson ('65, Figure 5, pp771) to show the rostral kinesthetic shell of the VB described by Wiener et al. '82). Indicated in this drawing are the representations of the cutaneous leg (L), digits of the hand (5—1), and head (H) and the (KVB) representations of the deep tissues of the leg and trunk (df) and of the arm (da) as determined by recording experiments . A—D. Line drawings of transverse sections through the VBC showing reconstructions of the pre-injection electrode penetrations (1-6) and the track of the recording/injecting pipette (Il) as well as the maximal spread of the injection site (indicated by heavy circle) for animal 533 LT. Levels A-C are separated by approximately 0.75 mm, D is 1.5 mm caudal to C. A. Responses to stimulation of the hairs covering the ankle and dorsal midline above the shoulder were recorded in penetrations 1 and 2 respectively. Penetration 3 had a locus responding to stimulation of the deep tissues of the lower forearm above a response to pad C of the hand. These responses indicate that this row of penetrations is approximately 1—1.5 mm caudal to the rostral pole of the KVB. B. The receptive fields encountered in penetration 4 and I1 are shown in the figurines to the right, center. Two loci responding to stimulation of the deep tissues of the lower arm (A and B) lay above a response to stimulation of the claw of digit 3 of the hand (C) in penetration 4. In penetration I1, the injection was made at a locus whose receptive 72 field was in the deep tissues of the elbow. These responses indicate that the injection was made into the lateral portion of the forelimb KVB in a zone responding best to joint movements lateral to the rostral pole of the digit 3 cutaneous representation. C. The responses recorded in penetration 5 (deep tissues of the upper arm above claw of digit 2 of the hand above the hairy dorsum of digit 2) indicate that this level passes through the rostral part of the digit 2 representation. D. Penetration 6 passed through the representations of the distal glabrous surfaces of digits 2 and 1 (upper three loci) and entered the cutaneous representation of the head in the ventroposteromedial nucleus. Regardless of which thalamus was actually injected in this and following figures, injections will be depicted as in left thalami. Likewise labeled cells in the medulla will be depicted as in right medullas. The pre—injection mapping penetrations both rostral and caudal to the injection sites in all cases were reconstructed from histological sections and used to place the injection site but only those penetrations at the same transverse level as the recording/ injecting pipette track are shown in all subsequent figures. On figurines in this and all subsequent figures, shaded areas indicate subcutaneous and black areas indicates cutaneous receptive fields. Bar equals 1 mm. E. Distribution of labeled cells following injection into the lateral KVB. Shown here is a standard series of drawings from transverse sections through the right dorsal medulla in the raccoon. Approximate levels of the transverse sections are shown in the inset at the upper left, numbers show distance from the obex in mm. Dots in this figure and all similar ones indicate the approximate location of all cells observed to contain HRP granules in every fourth 73 section through the contralateral dorsal medulla. In all cases this is described as the right side of the medulla. In this case the injection resulted in substantial labeling in cg x-r and z and the medial most portion of the Mt. Note the paucity of labeled cells in the ECu, lateral Mt or any of the nuclei projecting cutaneous information to the thalamus (cCuGr). 74 Figure 2.1 75 Figure 2.1 E a D MJ+