Research Reports - Science Direct

17 downloads 0 Views 3MB Size Report
respiration and were not driven by pressure or movement of the thorax. (5) Other ...... receptive fields were either cutaneous or were associated with respiratory ...
Brain Research, 77 (1974) 1-23 © Elsevier ScientificPublishing Company, Amsterdam - Printed in The Netherlands

1

Research Reports

S O M A T O T O P I C O R G A N I Z A T I O N OF T H E E X T E R N A L C U N E A T E N U C L E U S IN ALBINO RATS

SUZANN K. CAMPBELL*, T. D. PARKER** AND WALLY WELKER Department of Neurophysiology, University of Wisconsin, Madison, Wisc. 53706 (U.S.A.}

(Accepted March 19th, 1974)

SUMMARY

The pattern of representation of muscle receptive fields was delineated in the external cuneate nucleus (EC) of anesthetized, cerebellectomized albino rats. This nucleus was explored systematically using ball-tipped tungsten microelectrodes and closely-spaced electrode punctures. Single neurons were recorded from and the peripheral source and modality of their activating receptive fields were identified using natural stimulation o f exposed muscles of the rat's forequarter. Each muscle was stimulated by pulling the tendon, pressure to its belly with fine probes and/or by stretch induced by joint rotation. Of 585 single units recorded from 411 punctures in 31 animals, 243 were localized to EC and were activated by stretch at low threshold or punctate pressure o f muscles o f the ipsilateral forequarter. All cells definitely localized within EC by histological verification were activated only from receptive fields located in muscle. The musculotopic pattern of organization o f peripheral projections within the three-dimensional confines o f EC was rather detailed, with neck muscles represented in its rostrolateral pole, arm and shoulder muscles more caudomedially, and forearm and hand muscles progressively more caudally. There was no overlap o f representations from muscles located in different anatomical segments. EC units for which receptive field identification was certain were always activated from a single muscle. However, those units whose fields could not be identified precisely may have received convergent activation (or inhibition) from different muscles which our procedures did not allow * Present address: Division of Physical Therapy, School of Medicine, University of North Carolina, Chapel Hill, N.C. 27514, U.S.A. * * Present address: Department of Psychology, Loyola University of Chicago, Chicago, Ilk 60626, U.S.A.

2

S . K . C A M P B E L L C[ a].

us to detect. No units in EC were activated from cutaneous receptive fields, nor did we find convergence of activation from cutaneous and muscle receptive fields. EC cells were never activated from receptive fields in the hindlimb or contralateral forelimb.

INTRODUCTION

The search for understanding of the neural mechanisms responsible for control of posture and movement has revealed an incredibly complex assemblage of interconnected neural components, circuits, and systems in mammals. Sensory circuits from skin, muscles and joints reach motor control circuits in spinal cord, brain stem, cerebellum, thalamus, basal ganglia and cerebral neocortex8,9,12,~9,3°. The nuclear components at each of these levels are interconnected in ways that are only beginning to be understood. In order to determine how sensory feedback contributes to control and regulation o f movement and postures, it is necessary to know exactly how information from individual muscles, joints, and tendons is distributed to the several central circuits. Although the cerebellum is an important component of one of the motor control systems, little is known about the specific patterns of projections of afferent circuits to this structure. We have chosen to study the external cuneate nucleus (EC) of the medulla, which receives information from muscle receptors of the forequarter and transmits it to the cerebellum 6,7,1°,14,24,29,48. Research on EC has produced conflicting reports on (1) modality (cutaneous, tendon, primary and secondary muscle spindle) of receptive fields projecting afferents to EC 7,21,29,33,39, (2) convergence of input to single EC cells from different receptor modalities 7,29,33,4°, (3) bodily source of receptive fields 16,'~9, and (4) somatotopic organization of peripheral projections to this nucleus 6,21,27,3~,35. Conflicts in the literature may be related to the use of different experimental animals, and of different methods that vary in reliability and validity, as well as the failure to adequately localize recording sites (see Discussion). Our study was especially designed to delineate somatotopic (or musculotopic) organization of peripheral projections within EC, but some information regarding these other questions was also obtained. METHODS

Subjects Thirty-one adult male Holtzman albino rats were used in electrophysiological recording experiments, 13 in normal histological studies and 3 for normal anatomy of the musculoskeletal system.

Surgical preparation Subjects were anesthetized with sodium pentobarbital (50 mg/kg body weight) injected intraperitoneally and supplemented as needed to suppress reflex responses to pinching. The animal was respired mechanically via a tracheal cannula. Rectal temperature was maintained at 35 °C. The animal's head was held in a stere0taxic

SOMATOTOPICORGANIZATIONIN RAT EC

3

Fig. 1. Surface morphology of the rat medulla showing major sensory protuberances. A: dorsal and B: left dorsolateral views of a perfused medulla (rat No. 71-50). Dotted ellipse in 'A' indicates

general location of external cuneate (EC). C: photograph of dorsal view of portion of medulla in experimental rat (No. 74-330) with cerebellum removed. White dotted region indicates location of EC (also outlined in A). Typical pattern of electrode penetrations is indicated by black dots, three of which are numbered (1, 2, 3). Millimeter scales indicate enlargement factors of A, B (same as A), and C. Abbreviations: C1, level of entry of first cervical dorsal root; CN, cochlear nucleus; Cu, cuneate nucleus; EC, external cuneate nucleus; Gr, gracile nucleus; IC, inferior colliculus; arrow from O, obex; Tr, spinal trigeminal nucleus; V IV, fourth ventricle.

holder. The imedulla was exposed by removing occipital bone and aspirating the cerebellum. The atlas and axis were immobilized and fixed to a dam of acrylic constructed around the exposed edges of the cranium. The dura mater and leptomeninges overlying the caudal medulla were removed and the medullary surface was covered with warm mineral oil and photographed (Fig. 1C). The head and body were so supported that the entire left side was freely accessible to manipulation and stimulation. The skin and superficial fascia were partially dissected from the left forelimb leaving a majority o f the nerves to the skin intact so that cutaneous receptive fields could be identified in each experiment.

Electrodes and recording Tungsten ball-tipped microelectrodes were used and were prepared by the method of Parker et al. al. Ball diameters of these electrodes were between 5-15 #m. These electrodes permitted stable recording from single units despite vascular pulsation and medullary movements associated with respiration. Unit activity was amplified and recorded by conventional techniques.

4

S . K . CAMPBELL e l a t t.

Mapping The entire left dorsal medulla was explored by systematic micromapping techniques 21 to ensure thorough and complete sampling of the external cuneate nucleus. Records of depth below the surface and of medio-lateral and rostro-caudal coordinates were taken for each single neuron studied. Electrode tracks were marked by a mixture of india ink and Photoflo (Kodak) placed on the tip of the microelectrode. Boundaries of the responsive nuclear region were identified as lying between those adjacent electrode locations where the modality of the receptive fields changed from muscle to either cutaneous tissues, non-muscular deep tissues, silent regions, or areas containing steadily responding non-driveable units.

Receptive field identification The animal's body was gently and continuously manipulated as the microelectrode was touched to the surface and as it was advanced into the medulla. When unit discharges were evoked, the electrode was moved slightly up or down to yield the largest discharge amplitude, and the body was then carefully explored to identify the unit's receptive field. When the general activating body region had been determined, the precise location, boundaries and modality of each receptive field were identified by the methods described below. The receptive fields were drawn on figurine drawings of forelimb muscles or body surface of the rat (Fig. 8). The following receptive field modalities were identified. (I) A receptive field was categorized as cutaneous if it was activated by gentle contact with skin or hair. The receptive field of a cutaneous unit remained the same even if the skin was moved with respect to deeper underlying tissue. Most such units were rapidly adapting. (2) Medium- or fast-adapting receptors responding to tap, vibration or pressure on periosteum, tendons, or other deep tissues were classified as pressure-activated. They were not activated by muscle stretch. (3) Joint units may have been encountered occasionally, but we did not obtain unambiguous proof that any deep-lying, tonicallyactivated receptor originated in a joint. Special dissection and isolation of joints is required for such identification 26. It is possible that some of the 95 receptive fields that we were unable to localize to a specific muscle had their origin in joints (see Results). (4) Respiratory units were classified as such if they fired cyclically with respiration and were not driven by pressure or movement of the thorax. (5) Other miscellaneous receptive fields, such as dental units, steadily firing non-driveable units, and tonically active units which responded to tapping the skull or head holder (presumably vestibular receptors) were found occasionally. (6) Stretch-activated units, which were of primary interest in this study, were tonically firing, partially adapting units activated by stretch of a muscle by pulling its tendon or by punctate pressure of its muscle belly. Some of these units produced spike trains cyclically with respiration, but they also could be activated independently of respiration by mechanical movements of the chest. When a unit was tentatively identified as stretch-activated, the following general procedure was used to locate the receptive field precisely. (1) Since moving a joint stretches the muscles attached across it, joint rotation was the first procedure used to

SOMATOTOPIC ORGANIZATION IN RAT

EC

5

locate muscles containing a stretch receptor. Each joint of the forelimb was passively moved through its normal ranges of motion and the unit's general response characteristics were ascertained. Differential response to movement of joints in different directions helped to identify which muscle of a group contained the receptor. As each joint was manipulated, all other joints were immobilized manually. Whether the unit responded throughout the full range of stretch of its muscle was noted since this helped to differentiate muscle receptors from joint receptors, the latter being maximally activated through only a small part of the total range of motion, and not necessarily at the extremes of range. (2) The muscle bellies probably involved were then carefully explored by gentle pressure with a small glass probe or fine wire until the threshold receptive field was found which, when stimulated, produced steady and reliable firing of the unit. (3) The relevant muscle was then verified by pulling its tendon. (4) For those muscles that were deep and not readily accessible, the limb and shoulder joints were manipulated in as many different ways as possible in order to produce either reliable activation or cessation of the unit's response. Such manipulations, along with deep probing with a blunt stylus, made it possible to identify, if not a specific muscle, then the general region (i.e., lateral neck, ventral thorax) of the receptive field.

Localization of recording sites and reconstruction of musculotopic pattern The brains of all animals were perfused intracardially with formolsaline. To assist in histological identification of microelectrode tracks, tungsten wires (40/tm diameter) were inserted into the medulla through each of two or three selected electrode punctures. They were removed just prior to histological processing. The medulla was either frozen or embedded in paraffin and sectioned at 20 #m. Frozen sections were stained with cresyl violet, and paraffin sections were stained alternately with thionin or hematoxylin. Sections no thicker than 20 #m were important for identifying microelectrode tracks. Ten normal rat brains, embedded in celloidin and sectioned in one of the 3 major planes, were used to reconstruct the 3-dimensional relationships of the EC and adjacent nuclei of the medulla. Alternate sections were stained either with thionin or hematoxylin. Musculotopic patterns of sensory projections to EC were reconstructed by collating the electrophysiological and electrode location data with the neuroanatomical structures defined microscopically on stained sections (Fig. 7). Two-dimensional plots of somatotopic patterns were made by first superimposing a 200/zm grid rectangle on photographs (enlarged to the same magnification) of dorsal views of all medullas contributing to a particular plot. Then every electrode track location could be referred to the two axes of the rectangle and plotted as depicted in Fig. 5. Identification of receptive fields of muscles in the rat was aided by study of a freezedried specimen of the rat's forelimb, dissected and prepared according to the method of Hildebrand 17. Locations of skeletal attachments of muscles were studied on a skeleton of the forequarter. Although other researchers have identified EC neurons by antidromic activation of cells by electrical stimulation of the inferior cerebellar peduncle, we did not feel that

6

s.K. CAMPBELL ('f ~1/, A

,I,Mex

~'ext

4'flex

'~ext

Wrist Flexion

B t ,

l " ~-~

. ~ JiillI,L,t, LUld.~il,htl,.t

tldl~L~litttJl[ttltild~,ll~ll lhl_llllllJlllltllallliLl~!.l, tlita,ttllllm hl~lla[l u.~.,tllkh[

' ~IT~l~lqTIFl~lTIMlrll['lIT'll~rl~lrITlmtrfll111~11'l~lUl-ltIlll ITlllll~ITllllP11~[ll~rllrlBlr11TITl'll~?'lT~ll Radial Deviation

C h , t iiJi ,j~ till, l~ll.lhiltllaltuLl iill llL!Iilili

liliall.~lti I lik itll I1 tillLll

l lll lihIl~iI i,, J,, i 1111

l' ' ' Till 11 Illl"TIT~TTTFIIIllllfrtlTillIIlIIT[IT[ Ill]llll~llI/lIIII IIIIlFII I litlTlll~ 11 l IF III Tendon Pull _

_

~

5

4

I

*

,

,

,

I

1 See.

150bLV I

Extensor Digitorum

2 Fig. 2. Unit discharge patterns recorded from a stretch-activated receptive field in extensor digitorum muscle (rat No. 71-266, Unit 9). A: alternate ventroflexion (flexion, downward arrow) and dorsiflexion (extension, upward arrow) of wrist. B : radial deviation combined with ventroflexion of wrist (arrow indicates onset). Stretch on the muscle was maintained throughout the record. It produced maximal activation of the unit. C: gentle pulling of tendon of this muscle. Amplitude and time scales shown at lower right. Figurine shows location of muscle belly (in black). Arrow near figurine indicates location of tendon grasped by forceps.

this was a relevant procedure for our purposes since the identification methods we use have been repeatedly demonstrated as highly accurate and reliable. Moreover, there is the possibility that some neurons in the cuneate nucleus, as well as those in EC, project to the cerebellum zl. RESULTS

Modality classification In 31 experiments, 585 single units were recorded from 411 punctures traversing EC and other nuclei adjacent or subjacent to it in the dorsal medulla. The modality and receptive field were identified for 550 units. Of these, 258 were activated from cutaneous receptive fields, 244 by muscle stretch, 26 by pressure to other deep-lying receptive fields, and 2 from tooth stimulation. One unit was activated by sl~imulation of vestibular receptors, 5 were steadily firing, non-driveable, 14 were associated with inspiration or expiration, and 35 units did not have receptive fields that could be identified. Joint units were not identified with certainty. The cutaneous, tooth and deep-lying fields projected to the gracile, cuneate or trigeminal nuclei. Clear convergence o f input to single units from receptive fields representing different modalities was not demonstrated. That is, there were no multimodal units. With one exception (a receptive field on the head), all units activated by muscle stretch were located within the boundaries o f what is reasonably defined as EC by anatomical criteria. The remainder were localized in adjacent medullary nuclei. Spike discharge patterns from

7

SOMATOTOPIC ORGANIZATION IN RAT E C

B Serratus anterior

R

a

d

~

malor

~

External intercostal

InternalIntercostal

3 ~"

D

C

SupraspnatusTeres malor

rsi ~ ~ / / ~ / / / ~ / y . r n a s t o i d e/u/~/s~'7~ ~;~/~.~,~ ~= Levator

/ ~

~.

~ Jr

"

1~/C~la~all

s

"

Triceps/ /,

Cutaneous External maximus Serratus anteriorabdomi nals

Fig. 3. Musculature of the left forequarter of the albino rat. A: skeleton showing normal relationships of bones of the forequarter and head. B: deep thoracic muscles (pectoral muscles reflected). C: scapulohumeral and scapulothoracic muscles. D: superficial forequarter musculature. Cut edges of muscles are indicated by open circles.

one stretch-activated unit are illustrated in Fig. 2. These patterns demonstrate the activating effects on unit activity o f joint manipulation and of muscle stretch by pulling a tendon. The firing pattern appears qualitatively similar to that reported for first-order Ia fibers innervating muscle spindles12, 2°.

Muscle anatomy Our experimental procedures were designed to localize receptive fields in muscles and other deep tissues as accurately as possible given their anatomical complexity. Muscles are oriented in spiral and diagonal patterns with respect to skeletal structures so that the normal direction o f pull or stretch o f a muscle seldom occurs in a single plane. Muscles also interweave and overlie one another in complex arrangements. This structural complexity provides rotatory components to movements. Functionally, muscles are not simple flexors or extensors. Instead, because of their various spatial orientations and attachments, they may exert pull along several vectors. Movement of a joint in a single plane was seldom sufficient to permit receptive field localization.

S.K. (AMPBELL ('1 (/].

8

A

B Supra,spinatus S u b s c a P ~ T r

iceps brachii

l' /

//

\ ?

~.~,,~/,./,,/~(Biceps

L\/

//

~ [ Humerus

'rY/Y/J/,/,ld~.#~~,'IkJ~ ~\,,\ E"itroch-

brachii

Tere m Latissimus dorsi 4

p Extensorcarpi Extensordigiti " ,.~..~aris quarti a,ndq'6inti Extensor

pollicis

carpi radialis digitorum Iongusand brevis

D

\

\

Pronator Flexorcarpi ~ teiesradialis Flexordigitorum

umans

Iongus

sum~mus

Fig. 4. Musculature of the left forelimb. A: medial view of the bones of the forelimb. B: arm muscles and muscles deep (medial) to the scapula; from same view as A. C: dorsal forearm muscles (wrist and long digital extensors). D: ventral forearm muscles (wrist and long digital flexors). Tendon of palmaris longus has been cut to show underlying digital flexor muscles.

A receptive field was located following selective manipulation of each and all joints o f the body segment through all their possible planes of motion. The spiral and diagonal arrangement of the forequarter muscles and their spatial relationships to skeletal structures in the albino rat are illustrated in Figs. 3 and 4.

Localization of stretch-activated receptive fields Of the 244 neurons activated by gentle muscle stretch or pressure to a muscle belly, one was located in the head, 27 in the neck, 33 in the thorax and muscles of respiration, 81 in shoulder muscles, 16 in the upper arm, 63 in the forearm, and 23 in the hand. We did not perform quantitative tests to assess whether the receptive fields derived from primary or secondary spindle endings or from tendon organs. However, our stimulating procedures were sufficiently delicate and precisely localized to the activating tissues so that tendon organs were probably not being stimulated. Table I lists the muscles o f the rat forequarter, their innervation according to GreenO 5, and the number o f units identified which were activated from receptive fields in each muscle. The receptive field location of 95 units could be identified only with respect to general body region, and the number o f these units for each region is also included in Table I.

SOMATOTOPIC ORGANIZATION IN RAT

EC

9

All units whose receptive fields were clearly identified were activated by stretch of a single muscle. The region of EC from which the most complete data were obtained was that containing cells activated from receptive fields in arm and forearm muscles which were easily accessible to stretch and pressure stimulation. Of the 79 stretch receptive fields identified in arm and forearm muscles, only 6 could not be localized to one specific muscle, in most cases because the unit was lost before its receptive field could be carefully identified. We seldom attempted to identify individual intrinsic hand muscles because of their small size in the rat. In regions of EC activated by stimulation of proximal musculature (shoulder, neck, thorax), relatively more units were found whose receptive fields were not easily identified within a single muscle. We assume that many units, whose receptive fields could be localized only to a general region, were activated from inaccessible small or deep muscles which could not be stimulated in isolation. However, the possibility exists that they may also have had convergent fields from different muscles or were activated from joint receptors (see Discussion). Nevertheless, occasionally we were able to identify receptive fields in single deep muscles which could be stimulated in relative isolation because of their simpler physical relationships with other muscles and with skeletal structures. Sequentially alternating stretch of muscles having antagonistic functions did not produce convergent activation of any of the stretch-activated units. However, since antagonists were never stretched simultaneously, our data do not argue against some form of interaction of input from antagonists during muscular cocontraction (cf. Ros6n and SjSlund40). No units were found to be activated by muscle stretch in the hindquarter 16, and all receptive fields were ipsilateral. No units were identified that were activated from the following muscles: cutaneous maximus, serratus posterior superior, subclavius, the scapular elevators, deltoideus, teres major, the anconeus and epitrochleoanconeus, and the pronator quadratus (Table I). Some of these muscles are small and deeply situated, and possibly they are among those which activated units whose receptive fields could be identified by general region only. The cutaneous maximus, however, is a large superficial muscle and might be expected to be amply represented in EC. It may be that this muscle, associated with skin movement, has small numbers of stretch receptors and therefore a very small representation in EC, or none at all. It is also possible that afferents from muscle receptors in cutaneous maximus project to Clarke's column rather than to EC. Indeed, dorsal roots C5-T7 also project to that nucleus in the monkey 41. Failure to find units activated from receptive fields in the deltoideus and teres major is puzzling since they are well-defined, relatively superficial muscles. It is possible that they are represented under a regularly situated blood vessel (see Fig. 1) or in some other small portion of EC that our electrode might have consistently missed.

Somatotopic organization The results regarding somatotopic organization were summarized and schematically represented in a 2-dimensional plot of EC as viewed from the dorsal surface of the medulla (Fig. 5). Although combining data from different animals gives an

4B 3C 3C, 4B 3C, 4B

Rhomboids Latissimus dorsi Occipitoscapularis, levator scapulae and claviculae Pectoralis major and minor Subclavius Serratus anterior Acromiodeltoidius; spinodeltoidius

Subscapularis Infraspinatus and teres minor Supraspinatus Teres m a j o r

Arm

-3C, D, 4B 3D 3B -3B, C, D 3C, D

Trapezius

Shoulder

4B --

Brachialis

3C, D --3B

Biceps brachii and coracobrachialis

Flexors

Regional identification only

3D

Cutaneous maximus Serratus posterior superior Diaphragm External and internal intercostals Regional identification only

Thorax

-3D -----

Platysma Sternocleidomastoideus lnfrahyoid muscles Anterior vertebral muscles Lateral vertebral muscles (Scalenes) Posterior vertebral muscles Regional identification only

Neck

Figure

Muscle

Location

MUSCLES OF NECK, THORAX AND FORELIMB OF ALBINO RAT

TABLE I

Shoulder and elbow flexion, supination Elbow flexion

Scapular elevation, depression, adduction Scapular d o w n w a r d rotation Extension, medial rotation Elevation Flexion, adduction Depression Protraction Flexion, adduction, medial rotation; extension, abduction, lateral rotation Medial rotation Lateral rotation Abduction Extension, adduction, medial rotation

Skin m o v e m e n t Raises ribs 2-5 Inspiration Respiration

Skin m o v e m e n t Flexion, rotation Flexion Flexion Flexion Extension

Action

C5-7

C5-7

C5-7 C5-7 C5-6 C6-7

C5 C6--7 C3-5 C5-6; C8, TI C6 C6-8 C6-7

XI, C2-3

C8, T1 * TI~ C4-5 ! Tl-12 $

VII*, C* roots XI*, C 3 ~ C1-3 C1-4 C2-7 C1-8

Innervation

I

9

23

9 6 2 --

8 3 -27 -2 --

1

18

--15

1 1 ----25

.No. o f units

7z

4C 4C 4C 4C 4C

Wrist extensors Extensor carpi radialis longus and brevis Extensor carpi ulnaris

Wrist and long digital extensors Extensor digitorum (communis) Extensor indicis Extensor digiti quarti and quinti

Small digit 1 and digit 5 muscles Regional identification only

lntrinsics (short digital) Dorsal interossei Volar interossei Lumbricales m

4C

4D 4D

Wrist and long digital flexors Flexor digitorum sublimus Flexor digitorum profundus

Digit I extensors Extensor pollicis (longus and brevis) Regional identification only

4D 4D 4D

4D

Wrist flexors Flexor carpi radialis Palmaris longus Flexor carpi ulnaris

Radio-ulnar Supinator Pronator teres Pronator quadratus

3C, D, 4B 4B

Abduction digits 2-4 Adduction digits 2-5 Metacarpophalangeal flexion, phalangeal extension Flexion, abduction, opposition

Extension digit 1

Extension digits 2-5 Extension digits 2-3 Extension digits 4--5

Radial wrist extension Ulnar wrist extension

Flexion proximal joints digits 2-5 Flexion distal joints digits 2-5

Radial wrist flexion Wrist flexion Ulnar wrist flexion

Supination Pronation Pronation

Shoulder and elbow extension Elbow extension Elbow extension

* VII, cranial nerve VlI; XI, cranial nerve XI; C, cervical dorsal roots; T, thoracic dorsal roots.

Total

Hand

Forearm

Extensors Triceps brachii Epitrochleoanconeus Anconeus Regional identification only

C7-T1

C7-TI C7-T1 C7-TI

C6-8

C6-8 C6-8 C6-8

C5-7 C6-8

C7-T1 C7-T1

C6-7 C7-T1 C7-T1

C5-7 C6-7 C7-TI

C6-8 C7-T1 C7-8

243

23

m

m

23

3

8

6

1

7

1 1

2

1

1

5

7--.

>

>

o o

O

s.K. CAMPBELL dl ~1/.

12

2.8

28

:~:' "

O

~A

~ "

~L[2 T

28

{:

2,4

2.4

2.4

^tep,

-k

Io

'

~

2.0

^.QO 'k

16 | ~

~ o°

1.6

~@

1.2 8 .8

"410

En

o

o 2.0 i

'

O

A

,.>!,

#.

o

o o _°E~e

B

A

~

I

.8

I

.4

I

m m.

FOREQUARTERCOMPOSITE

I

i 0be×

.4

1.0

'

~

I m m.

SINGLEEXPERIMENT

.4]0 Ob+ex

I

Z~

I

m m,

[ Obex

ARMANDFOREARM

5 Fig. 5. Muscular somatotopy in EC. Data from different dorsoventral levels are projected upon a 2-dimensional horizontal plane of the medulla. Each symbol represents receptive field of a single neural unit. Key to symbols indicated in enclosed rectangles. A: combined data from 141 units in 15 experiments illustrate caudomedial-rostrolateral somatotopic organization of those body segments of the forequarter indicated in key. Forearm and intrinsic hand muscles are represented in the most caudomedial part of EC while neck muscle receptive fields project to the rostrolateral pole. B: data from 22 units in a single experiment (rat No. 72-352) illustrate medial location of forearm and hand muscle projections and more rostrolateral location of projections from shoulder and arm muscles. Apparent overlap of arm and shoulder projections is due to collapse of three dimensions to two (see Fig. 6). The neck muscle representations were not extensively sampled in this experiment. C: combined data on arm and forearm muscle projections from 69 units in 23 experiments. Antagonistic muscle groups acting at the same joint are represented by the same geometric symbol, but open symbols refer to flexor muscles while filled symbols refer to extensors (see Table I). Note that there is almost no overlap of projections from arm and forearm muscles even though data from many experiments were combined in this plot. Projections from muscles acting only at the wrist (represented by squares) are located in the rostral part of the forearm region, whereas the representations from muscles which act on both the wrist and the digits (represented by circles) are located throughout the rostrocaudal extent of the forearm projection region. The general recording location of all data shown and the spatial orientation of all three plots on the medulla are shown in the brain diagram at top, center. Composite plots were made as described in Methods.

appearance o f overlap of receptive field representations from different body parts, overlap was not found in a single experiment, even in closely spaced punctures where cells recorded from were only 100-200ffm apart (Fig. 5B). Moreover, two units recorded from a single electrode position were never activated from a single muscle. Fig. 5 shows that the more proximal forequarter segments are represented rostrolaterally and the more distal segments project caudomedially. The apparent overlap of arm and shoulder muscle projections in the 2-dimensional plots is not real. The

SOMATOTOPICORGANIZATIONIN RAT EC

13

arm representations lie superficial to the shoulder representations as shown in Fig. 6, a schematic 3-dimensional representation o f muscular somatotopy within EC. Our detailed findings regarding somatotopy are described in the following paragraphs*. N e c k muscle stretch receptors project their afferents to neurons in the rostrolateral pole of EC. Shoulder muscle projections were found near the middle of the nucleus throughout much o f the rostrocaudal extent o f EC partially surrounding a central core of representations of extensor muscles o f the wrist and digits. Thus, an electrode passing through this central region frequently encountered a unit activated from a shoulder muscle dorsally, followed by a forearm muscle, and then another shoulder muscle unit more ventrally. Fore,arm muscle projections continue into the caudal pole of EC and will be discussed in detail below. The afferents from thoracic and upper arm muscles projected to the most lateral portions of EC at its middle third rostrocaudally (Figs. 5 and 6). The deep muscles of the thorax and muscles of respiration were represented in a crescentshaped lamina lateral to the shoulder muscle projections. Receptors in intercostals and diaphragm were inferred from their synchronous activation with either thoracic or diaphragmatic components o f respiration. More specific identification was made by deep pressure into these regions with a probe. Laterality o f receptive field in the diaphragm could not be judged. Representations of upper arm muscles formed a curved lamina medial to the thoracic representations and dorsal to the shoulder muscle projections. The elbow flexor muscles appeared to project more rostrally and the elbow extensor muscles more caudally in this lamina. A composite of the data from 23 rats regarding somatotopy o f individual arm and forearm muscle receptive fields is illustrated in Fig. 5C. Stretch receptors in those forearm muscles which flex the wrist and digits activated cells in the most medial portion of EC, but only in its middle third rostrocaudally. The lateral part o f the forearm flexor muscle projections partially overlie those of the deep core o f wrist extensor muscle representations while medially the forearm flexor representations lie dorsal to the intrinsic hand region, which was the most medioventral part of EC.

* In the discussion which follows, the muscles of the proximal forequarter segments will be referred to as neck, shoulder, and thoracic muscles (see Table I for muscles included in each group). A functional classification of the muscle groups of the arm and forearm according to the major joint action produced by their contraction will be used in order to discuss the projectional specificity of these groups. Muscles of the arm which move the elbowjoint will be referred to as arm flexors and extensors. Forearm muscles include the radio-ulnar muscles which act to produce pronation and supination, as well as those muscles whose bellies are located in the forearm segment but have primary actions on either the wrist or the digits. They include the wrist flexors and extensors which are primary movers of the wrist joint, and the long digital flexors and extensors, which are primary movers of the digits but also have corresponding effects on the wrist joints. As an example of this double action, the extensor digitorum acts to produce extension of both the wrist and the digits. The long extensors of the first digit, however, have no significant effect on extension of the wrist joint and are considered primary digital extensors. Therefore, 4 general groups of forearm muscles will be considered: (1) radio-ulnar muscles; (2) wrist muscles; (3) long digital muscles (excluding long extensors of the first digit); and (4) long extensors of digit 1. Finally, hand muscles refers to the intrinsic muscles of the hand which do not cross the wrist joint and therefore act only at the metacarpophalangeal and interphalangeal joints.

s.K. CAMPBELLel u/.

14

6 A

r" --1

1

--2 --3 --4 -5

-6 -7

/

/

/

/,f

/

EC'~

/

\ //

/'

I ,

-'~

I \\ ~ i// / / / q I

\

le'/ t I

-~-

,@

N= T = S= A= F= H=

/

/

/ /

_ ~

/

~iI

Neck Thorax Shoulder Arm Forearm Hand

Fig. 6. Muscular somatotopy. Composite schematic reconstruction of the 3-dimensional somatotopic organization of EC. A : reconstructions of transverse sections of EC at 7 rostrocaudal levels showing projections of muscles from 6 major subdivisions of the forequarter. B: 3-dimensional reconstruction of the major forequarter representations in EC; dorso-caudo-medial aspect. Seven rostrocaudal levels of sections shown in A indicated at right. C: outline diagram of rat medulla from dorsocaudo-medial view showing locations and orientation of EC representations. Key to body segments at lower right.

The organization o f the region o f EC receiving projections f r o m forearm extensor muscles demonstrated a segregation o f representations based on functional a n a t o m y o f the muscles. Muscles with primary actions on either the wrist (extensor carpi radialis) or the digits (extensor pollicis) are represented separately at rostral and caudal levels respectively in the forearm extensor region, while muscles which have actions at both joints (extensor digitorum) appear to activate cells at all rostrocaudal levels in the central core o f EC. There are three muscles whose projections to E C appear to be somatotopically aberrant. Three f o r e a r m muscles appear to project within the arm muscle region (Fig. 5C and Table I). (1) This finding m a y be related to the fact that the p r o n a t o r teres

SOMATOTOPIC ORGANIZATION IN RAT E C

15

muscle crosses the elbow joint and is innervated by the musculocutaneous nerve (from dorsal root segments C6 and 7) which also provides the innervation to the biceps muscle. (2) The supinator muscle is a synergist of the biceps muscle in producing forearm supination and is innervated by cervical dorsal roots 5-7. Based on innervation, function, and projection pattern to EC, it may be that the radio-ulnar muscles should then be classified as 'arm' muscles. (3) The flexor carpi radialis, a wrist flexor, was also represented within the arm region of EC. The projection of this muscle's afferents to the arm region is perhaps related to the fact that, in the rat, this muscle is innervated by the musculocutaneous nerve which also innervates the biceps and pronator teres 1~. The other forearm wrist and digital flexors are innervated by the median and ulnar nerves from C7-T1. These suggestions are based on little data, however, and require further study. We did not obtain a detailed picture of somatotopic organization among adjacent EC cells since we did not make a special effort to record from more than one unit of EC at a single recording locus. However, on 4 occasions we recorded from two adjacent cells at a single electrode location. The spike amplitude and wave form of such units were clearly different, as were their activating muscles. Two of the pairs (extensor digitorum and extensor pollicis; extensor carpi radialis and extensor pollicis) included anatomically adjacent muscles, and the third pair (supinator and pronator) was composed of antagonist muscles. The fourth pair (shoulder and wrist muscles) consisted of two muscles from different anatomical regions. The latter two muscles, however, were not precisely identified. While these fragmentary data are insufficient to allow precise description of the relationships between adjacent cells within EC, they do suggest that there may be specific kinds of spatial associations between central representations of synergist and antagonist muscles.

Anatomical relationships of the external cuneate nucleus In order to reconstruct the 3-dimensional pattern of muscular somatotopy within EC, it was necessary to study the normal anatomy of this nucleus as well as that of those nuclei that lie adjacent to it. The dorsolateral aspect of the caudal medulla contains several populations of neurons receiving somatic sensory input (Fig. 7). The major structures are the trigeminal (Tr), cuneate (Cu), gracile (Gr) and external cuneate (EC) nuclei, but several additional smaller nuclear groups may also be components of somatic sensory circuits. These include nucleus 'z '2, the ventral portions of the cuneate nucleus 34-36,as-4°, and 3 additional nuclear groups overlying the spinal nucleus of the trigeminal complex, which we have called the ectotrigeminal nucleus (ET), the trigeminal plate nucleus (TP), and the paratrigeminal nucleus (PT). Subnucleus 't', between Cu and Tr, contains the representation of cutaneous receptive fields from the ear and adjacent head and neck skin in raccoons 21 and sheep 49 as well as in rats (see Fig. 8). The relative location, shape, and orientation of these cell groups in the caudolateral medulla of the rat can be understood from Figs. 7 and 8. We have receptive field data in the rat only for EC, Tr, Cu and Gr. Anatomical features of only EC will be discussed here.

16

S. K, C A M P B E L L ('1 ~li.

Fig. 7. Nuclear groups of dorsal lateral caudal medulla. A : transverse hematuxylin-stained section of left medulla of rat No. 72-349 (section 149) at a plane just rostral to the obex. B: outline diagram of section in A showing location of 7 nuclear groups. Cu, cuneate nucleus; EC, external cuneale nucleus; ET, ectotrigeminal nucleus; Gr, gracile nucleus; PT, paratrigeminal nucleus: RB, restiforrn body; t, triangular subnucleus of the cuneate nucleus; TP, trigeminal plate nucleus; Tr, trigeminal nucleus; V, trigeminal tract. C: cytoarchitecture of main body of EC (transverse section 252, thionin stain) at level slightly rostral to that shown in "A'. Scales in each figure indicate magnification.

The main body of the rat's EC is situated rostral, dorsal and lateral to the cuneate-gracile complex (Cu-Gr). Although the dorsal and lateral boundaries of EC are sharply delineated by fiber tracts and by the dorsal surface of the medulla, the rostral and caudal boundaries are less definitive, as are the medial and ventral boundaries at certain rostrocaudal levels. The rostral pole of EC nests within the caudal pole of the descending (spinal, inferior) vestibular nucleus, but in serial sections in all 3 major planes the separation of these two nuclei can be easily recognized (see ref. 1). The caudal pole of EC splinters and appears as interrupted strands or islands of cells within the cuneate fascicle. Throughout the caudal half of EC synectant strands of cell bodies and neuropil interconnect EC ventrally with Cu, Gr, and Tr. These strands have been named 'nucleus interstitialis' by von Monakow 2s, but were believed to contain EC neurons by Ferraro and Barrera 11. We do not have evidence that this is so. At more rostral levels of EC, Cu and Gr become progressively smaller, are less definitive in their outlines, lie more closely apposed to EC, and the cells between EC and C u - G r appear transitional in size, shape and staining characteristics. The prominence of large cells in EC is a characteristic feature of all mammals that have been examined 28. However, medium-sized and smaller cells are also clearly

SOMATOTOPIC ORGANIZATIONIN RAT EC

17

present scattered among the larger cells (Fig. 7C) 44. A quantitative study of cell sizes in EC of raccoons 21 revealed a bell-shaped distribution of cell sizes. We have not measured EC neurons in rats. Many cells at the borders o f EC have a fusiform shapel~ but whether or not this feature is associated with functional differences is not known. Fiber bundles that enter or leave EC do so in groups that tend to split the nucleus into cellular clusters. Although these clusters suggest that EC may be composed o f several distinct subnuclei, we did not obtain sufficient data from adjacent EC neurons to indicate whether or not such cellular aggregates have somatotopic significance as they do in Cu and Gr 21. In many mammals EC is fractionated into numerous individual subnuclei. In a particular species, these subnuclei appear similar in different specimens and on both sides o f the same specimen with respect to their number, shape, size, and relative location. Such subnuclei are consistently present on several adjacent serial sections. Multiple subnucleation is especially prominent in gibbons, chimpanzees and humans. Axons from at least the larger neurons in EC project into the cerebellum. The destinations of axons from medium-sized and small cells are not known.

Anatomical-physiological correlations (Fig. 8) Although we were primarily interested in units activated from muscle stretch receptors, we recorded from a large area o f the medulla in order to obtain data on boundaries between the various medullary nuclei surrounding EC. All cells that we have recorded from that were localized without question in EC were activated by gentle stretch of, and pressure into, muscles of the ipsilateral forequarter. The one unit activated by stretch of a head muscle was not identified histologically so its precise location is unknown. All EC units responded by steadily discharging as long as the stretch stimulus was applied (Fig. 2). Units with cutaneous receptive fields were never found in EC. Whenever the recording electrode passed from EC into surrounding nuclei, the activating receptive fields shifted abruptly from muscle receptors to either non-muscular deep pressure or cutaneous receptors. In addition, the unit discharges shifted from slowly adapting to relatively rapidly adapting patterns activated by low threshold stimulation o f skin, hairs, deep periosteal, fascial or tendinous tissue. The shift o f receptive fields from muscle to either cutaneous or deep nonmuscular tissues is illustrated in Fig. 8. Data from 4 different animals recorded at 5 different rostrocaudal levels of EC are shown. All electrode tracks illustrated were identified histologically and the anatomical locations of the recording loci were estimated from depth readings recorded in the experimental protocol. Whenever the electrode was moved medial to EC, the units recorded were either not activated by sensory stimulation (NIL punctures, not shown in Fig. 8), or their receptive fields were either cutaneous or were associated with respiratory movements (e.g. Fig. 8C and D). At rostral levels of EC, Cu and Gr are not present, but at midrostrocaudal levels where EC overlies Cu and is just lateral to Gr, cutaneous receptive fields that activated units medial to EC were located about the midriff (Fig. 8C and D). More caudally, where Cu extends more medially relative to EC, punctures medial to EC encountered chest receptive fields (Fig. 8E).

l~

S.K.

B

1718

|

/

/

:~-~ a

c

,

C'I d].

CAMPBliLL

'~ >

"

I

I

I ....

~ ....

~

I

/~'~'~

I mm

54

1

1ll 13

" ~..,

,

1

i

I

e-j

--

[I

pC

~' il

"9

'

i/

"

,,"

ntT

I

~__i

SOMATOTOPIC ORGANIZATION1N RAT E C

19

Units located ventromedial to E C were activated by hair m o v e m e n t or light t o u c h o f skin o f the forelimb, neck or ear (Fig. 8A and C-E). These units were located in the cuneate nucleus. Those units recorded in subnucleus triangularis (t, see Fig. 7) were activated f r o m receptive fields on the ear and periaural regions (Fig. 8C-E). The few receptive fields f o u n d f r o m deep periosteal, fascial, or tendinous tissues in the lower arm or wrist activated neurons in rostral regions o f the cuneate nucleus in a border zone between E C and Cu (Fig. 8B-D). Muscle receptor input to rostroventral portions o f Cu has been reported in cat a4-~7 and r a c c o o n 21, but we did not obtain records f r o m this region in the rat. Units activated f r o m pressure and cutaneous receptive fields in the head were f o u n d in punctures lateral to E C or ventral to the lateral part o f E C (Fig. 8A, D and E). These units were all located in the spinal trigeminal nucleus (Tr). DISCUSSION The external cuneate nucleus (EC) was o f special interest to us because o f its probable involvement in sensory control and regulation o f the m a n y specialized behaviors utilizing musculature o f the forequarter exhibited by most mammals. We were particularly interested in how specific the patterns o f projections were f r o m stretch receptors o f individual muscles o f the forelimb to EC. Previous electrophysiological studies have shown some degree o f somatotopic organization6,21,27,aa, ~s,39. But we believe that methods used were not precise enough to answer this question. They neither attempted to stimulate individual muscles n o r systematically explored all parts o f EC21,27,38,3L Or, they used electrical stimulation o f peripheral nerves 6,7, instead o f natural mechanical stimulation21,ag, 40. N o study has yet utilized all these i m p o r t a n t ingredients. In the present study, using m i c r o m a p p i n g methods and threshold levels o f natural stimulation o f single muscles (wherever possible), we f o u n d that there was a highly organized musculotopic pattern o f projections f r o m forequarter muscles to EC. There was no overlap o f projections f r o m different forequarter segments (e.g.

Fig. 8. Anatomical-physiological correlations. Diagrams A-E: transverse sections through left medulla from 5 different experiments. Each section taken from a different rostrocaudal level of EC (dotted region) whose location is indicated by a dotted ellipse in key diagram of medulla at lower right. Numbered arrows on each section indicate locations of electrode tracks. Each black dot on arrows identifies estimated locus of single unit from which recordings were obtained. Receptive field activating each unit is depicted on a figurine of left forequarter of rat. Each figurine or pair of figurines is numbered to correspond to a numbered electrode track. Receptive fields of successively dorsoventral units are lettered sequentially from 'a'. Lettered receptive fields shown are of either muscle stretch (black patches), cutaneous (dotted patches with solid outlines) or deep pressure (dotted patches) origin. Nil punctures medial and lateral to EC are not shown. Abbreviations of muscle names: ADM, abductor digiti minimi (intrinsic hand muscle); ECR, extensor carpi radialis; ED, extensor digitorum; EDQ, extensor digiti quinti; FCU, flexor carpi ulnaris; FDP, flexor digitorum profundus; FCR, flexor carpi radialis; PI, palmar interosseus (intrinsic hand muscle); P, pectoralis major;?, unidentified muscle; resp. ?, respiratory-related unit, no receptive field identified. Animal numbers and section numbers indicated on bottom of each drawing. 1 mm scale shown near top center.

20

S . K . CAMPBELl. el a].

neck, shoulder, arm, forearm, hand), and 61 ~,,, of all neurons were activated strongly and tonically by stretching single muscles. For the remaining EC units, we could not identify a single activating muscle source. In many of these latter cases, we believe that the receptive fields lay either in the deeper less accessible muscles, or in muscles that were so intimately associated with one another mechanically that they could not be stretched alone without also moving the others. We expected that if there were convergence of activation, each of the two or more contributing muscle sources might have been clearly identified separately as well. But this was never the case. It is possible that convergent activation may have required stretch of all contributing muscles at the same time. However, we made no careful quantitative attempts to do this. It is important to note that our data regarding specificity of projections applies only to delicate, near-threshold natural stimulation. It is possible that stronger stimulation would be more likely to reveal convergent activation, but we do not believe this occurs. It was impossible to stimulate the tiny individual muscles of the rat's forelimb strongly since we did not immobilize the joints, or mechanically isolate single muscles from one another. We found no evidence of tendon ~9,4°, joint 33 or cutaneous23,"9, 33 activation o f neurons in EC, although stimulation of such receptive fields did activate neurons in other nuclei adjacent to EC. But again, we did not systematically test for convergence of activating input from these sources with that from muscle. It is of course possible that the use of an anesthetic or subnormal body temperatures prevented convergence. Our data do suggest, however, that extreme care must be taken to identify histologically the exact location of the recording electrode, since EC lies so close to the cuneate nucleus which receives cutaneous and joint input, and to the intermediate cellular region where we believe we may have recorded tendon-activated cells. Since our primary purpose was to search for sources of stretch-activation of EC neurons, we did not employ the special methods and tests required4,a3,39, 42,4a to identify exact modality of the contributing receptors (primary muscle spindle endings, secondary endings, Golgi tendon organs). Nor did we search for inhibitory effects on EC neurons reported by others v,18,29,3a,3s,4°. It is likely that in addition to specificity of musculotopic projections to EC, there may also be a highly ordered and specific pattern of inputs from other receptors that are activated when muscles contract or are stretched. Ros6n and Sj61und's 39,4° fine studies o f EC in cat have shown unequivocally that primary muscle spindle endings are the predominant source (77 ~ , n = 106) of activating input to EC cells. Eighty-seven percent of the ceils that were tested in their sample (n = 76) were activated by stretch of only a single muscle. Although we did not identify specific submodalities o f muscle receptive fields, we believe that our gentle stimulation methods tended to activate spindle primary endings preferentially. About 13 ~ of Ros6n and Sj61und's sample of 76 units were activated by pulling more than one tendon, and in all cases the convergent receptors were located in synergistic muscles. The receptors in these convergent cases were believed not to be tendon organs and probably were spindle primaries. Ros6n and Sj61und studied forearm muscles primarily. In our forearm sample 8 % of the EC cells (n = 63) might have had convergent fields since their receptive fields could not be localized to a single muscle. If it is true that the

SOMATOTOPIC ORGANIZATION IN RAT

EC

21

units for which we were unable to find single receptive fields were of the convergent type, then we suggest that certain muscle pairs or groups might exhibit greater convergence than others. Ros6n and Sj61und found inhibitory convergence in 4070 of their sample. We have no data bearing on this. How the EC--cerebellar circuit functions in normal behavior is not clear. It certainly will be active during all movements. Movements may be fast ballistic or slow, automatic or learned, intentional or signal-elicited. Different behaviors differ with respect to their composition of postural and skilled movement patterns. Every movement has an onset, acceleration, force, velocity, range, duration, deceleration, termination of contraction, and either returns to resting muscle tone, or begins the next movement sequence. Muscle spindle primary and secondary endings and tendon organs are so deployed that they can selectively signal different, but overlapping ranges of several of these stimulus parameters. Each muscle that is active in a given movement has its own particular spatiotemporal pattern of sensory feedback from these receptors. The great spatiotemporal precision exhibited by most complex skilled movements suggests not only that each muscle involved is under precise motor control, but that the sensory feedback from it is distinguishable in central circuits from the feedback from other adjacent muscles. It is this view that prompted the present micromapping experiments in search of musculotopic organization. Much more needs to be known about microdetails of circuit construction; about decoding and encoding features within both EC and cerebellum; about excitatory and inhibitory interactions between different types and locations of receptor modalities; and about the role of the several interconnecting ascendingS, 29 and descending7,22,3a circuits that are known to exist. In order to answer such questions, a battery of specialized test procedures must be utilized. Although electrical stimulation of nerves provides specific information about timing relationships6,V,1S,19,29, there are dangers to its use 19,39. Several receptor types may send their fibers through the nerve stimulated3,39. Conduction velocity can no longer be considered a sufficient single indicator of receptor modality39. The following strategies are suggested for use in the future. Natural stimulation should always be used to test any hypotheses derived from electrical stimulation studies. When natural stimulation is used, quantitative parametric control of the available stimulus variables is essential, as is determination of response thresholds to minimal values of each stimulus parameter. Quantitative analyses of cell discharge profiles and of temporal patterns of these to variations of input patterns from first-order neurons will be important if we are to understand the dynamic functions of this nucleus. Micromapping procedures are also essential in any study of EC. These include thorough sampling of the entire neuron population, closely-spaced microelectrode penetrations, single-unit recording and histological identification of recording sites. Care must be taken in identifying boundaries between adjacent neuron populations since adjacent cells in such regions may have different types of afferent sources. Recording from adjacent cells in EC and determining whether they have mutually interacting effects will also be important. Finally, only recording from appropriate samples of neurons in EC in awake animals that are performing specific movement

22

S.K. CAMPBELL ~t ~l[.

patterns is the ultimate source of i n f o r m a t i o n regarding how EC normally operates d u r i n g behavior 32,45-~7. A l t h o u g h n o t all these procedures can be used in any given experiment, the n o r m a l functions o f neural circuits controlling movements and postures c a n n o t be u n d e r s t o o d until such i n f o r m a t i o n is available. ACKNOWLEDGEMENTS This project was supported by U.S. Public Health Service G r a n t s GM-46,838 a n d NB 6625. We wish to t h a n k Dee U r b a n for p r o d u c i n g the drawings, T. P. Stewart a n d Shirley H u n s a k e r for p r e p a r i n g photographs, Jo A n n Ekleberry for preparing the histological materials, D e b b y Hager for t y p i n g the m a n u s c r i p t , Peter T h u r l o w for technical assistance, a n d R. M. Benjamin, A. L. Berman, H. J. Ralston III, C. N. W o o l s e y a n d G. M. Shambes for their helpful c o m m e n t s in reading versions o f the manuscript.

REFERENCES 1 BERMAN,A. L., The Brain Stem of the Cat, Univ. of Wisc. Press, Madison, Wisc., 1968. 2 BRODAL, A., AND POMPEIANO, O., The vestibular nuclei in the cat, J. Anat. (Lond.), 91 (1957) 438-454. 3 BROWN, A. G., AND HAYDEN, R. W., The distribution of cutaneous receptors in the rabbit's hind limb and differential electrical stimulation of their axons, J. Physiol. (Lond.), 213 (1971) 495-506. 4 BROWN, M. C., ENGBERG, 1., AND MATTHEWS, P. B. C., The relative sensitivity to vibration of muscle receptors of the cat, J. Physiol. (Lond.), 192 (1967) 773-780. 5 BURTON,J. E., BLOEDELL,J. R., AND GREGORY,R. S., Electrophysiological evidence for an input to lateral reticular nucleus from collaterals of dorsal spinocerebellar and cuneocerebellar fibers, J. Neurophysiol., 34 (1971) 885-897. 6 COOKE, J. D., LARSON, B., OSCARSSON,O., AND SJOLUND, B., Origin and termination of cuneocerebellar tract, Exp. Brain Res., 13 (1971) 339-358. 7 COOKE,J. D., LARSON,B., OSCARSSON,O., AND SJOLUND, B., Organization of afferent connections to cuneocerebellar tract, Exp. Brain Res., 13 (1971) 359-377. 8 ECCLES, J. C., The cerebellum as a computer: Patterns in space and time, J. Physiol. (Lond.), 229 (1973) 1-32. 9 ECCLES,J. C., ITO, M., AND SZENT.~GOTHAI,J,, The Cerebellum as a Neuronal Machine, Springer, New York, 1967. 10 EKEROT,C. F., AND LARSON,B., Differential termination of the exteroceptive and proprioceptive components of the cuneocerebellar tract, Brain Research, 36 (1972) 420-424. 11 FERRARO,A., AND BARRERA, S. E., The nuclei of the posterior funiculi in Macacus rhesus. An anatomic and experimental investigation, Arch. neurol. Psychiat., 33 (1935) 262-275. 12 GRANIT, R., The Basis of Motor Control, Academic Press, New York, 1970. 13 GRANIT, R., SKOGLUND, S., AND THESLEFE, S., Activation of muscle spindles by succinylcholine and decamethonium. The effects of curare, Acta physiol, scand., 28 (1953) 134-151. 14 GRANT,G., Projection of the external cuneate nucleus onto the cerebellum in the cat: An experimental study using silver methods, Exp. Neurol., 5 0962) 179-195. 15 GREENE, E. C., Anatomy of the Rat, Hafner, New York, 1955. 16 HAND, P. J., Lumbosacral dorsal root terminations in the nucleus gracilis of the cat, J. comp. Neurol., 126 (1966) 137-156. 17 HILDEBRAND,M., Anatomical Preparations, Univ. Calif. Press, Los Angeles, 1968. 18 HOLMQVIST,B., OSCARSSON,O., AND ROS!~N, I., Functional organization of the cuneocerebellar tract in the cat, Acta physiol, scand., 58 (1963) 216-235. 19 HUNT,C. C., Relation of function to diameter in afferent fibers of muscle nerves, J. gen. Physiol., 38 (1954) 117-131. 20 JANSEN, J. K. S., AND RUDJORD, T., Dorsal spinocerebellar tract: Response pattern of nerve fibers to muscle stretch, Science, 149 (1965) 1109-1111.

SOMATOTOPIC ORGANIZATION IN RAT E C

23

21 JOHNSON,J. I., JR., WELKER,W. I., AND PUBOLS,B. H., JR., Somatotopic organization of raccoon dorsal column nuclei, J. comp. Neurol., 132 (1968) 1-44. 22 KERR, F. W. L., The descending pathway to the lateral cuneate nucleus, the nucleus of Clarke and the ventral horn, Anat. Rec., 160 (1968) 375 (abstract). 23 KJERULE, T.D., AND LOESER, J.D., Neuronal hyperactivity following deafferentation of the lateral cuneate nucleus, Exp. Neurol., 39 (1973) 70-85. 24 Liu, C. N., Afferent nerves to Clarke's and the lateral cuneate nuclei in the cat, Arch. Neurol. Psychiat. (Chic.), 75 (1956) 67-77. 25 MONAKOW, C. VON, Neue experimentelle Beitrag zur Anatomie der Schleife, Neurol. Zbl., 4 (1885) 265-268. 26 MOUNTCASTLE,V. B., AND POWELL, T. P. S., Central neural mechanisms subserving position sense and kinesthesis, Johns Hopk. Hosp. Bull., 105 (1959) 173-200. 27 NORD, S. G., Somatotopic organization in the spinal trigeminal nucleus, the dorsal column nuclei and related structures in the rat, J. comp. NeuroL, 130 (1967) 343-356. 28 O'NEAL, J. T., AND WESTRUM, L. E., The fine structural synaptic organization of the cat lateral cuneate nucleus. A study of sequential alterations in degeneration, Brain Research, 51 (1973) 97-124. 29 OSCARSSON,O., Functional organization of the spino- and cuneocerebellar tracts, Physiol. Rev., 45 (1965) 495-522. 30 OSCARSSON,O., Functional significance of information channels from the spinal cord to the cerebellum. In M. D. YAHR AND D. P. PURPURA (Eds.). Neurophysiological Basis of Normal and Abnormal Motor Activities, Raven Press, Hewlett, New York, 1967, pp. 93-117. 31 PARKER,T. D., STRACHAN,D. D., AND WELKER, W. I., Tungsten ball microelectrode for extracellular single-unit recording, Electroenceph. clin. Neurophysiol., 35 (1973) 647-651. 32 PHILLIPS, M. I., (Ed.), Brain Unit Activity During Behavior, Thomas, Springfield, II1., 1973. 33 RAUSH, J., Electrophysiology of the External Cuneate Nucleus in the Cat, P h . D . Dissertation, Univ. of Washington, 1969. 34 Ros~N, I., Functional organization of Group I activated neurons in the cuneate nucleus of the cat, Brain Research, 6 (1967) 770-772. 35 ROS~N, I., Localization in caudal brain stem and cervical spinal cord of neurones activated from forelimb Group I afferents in the cat, Brain Research, 16 (1969) 55-71. 36 Ros~N, I., Afferent connections to Group I activated cells in the main cuneate nucleus of the cat, J. Physiol. (Lond.), 205 (1969) 209-236. 37 Ros~N, I., AND ASANVMA, H., Natural stimulation of Group I activated cells in the cerebral cortex of the awake cat, Exp. Brain Res., 16 (1973) 247-254. 38 ROS~N, I., AND SJ6LUNO, B., Natural stimulation of Group I activated cells in the cuneate nuclei of the cat, Acta physiol, scand., Suppl. 330 (1969) 118 (abstract). 39 ROSr~N, I., AND SJ6LUND, B., Organization of Group I activated cells in the main and external cuneate nuclei of the cat: Identification of muscle receptors, Exp. Brain Res., 16 (1973) 221-237. 40 ROS~N, I., AND SJ6LUND, B., Organization of Group I activated cells in the main and external cuneate nuclei of the cat: Convergence patterns demonstrated by natural stimulation, Exp. Brain Res., 16 (1973) 238-246. 41 SHRIVER,J. E., STEIN, B. M., AND CARPENTER,M. B., Central projections of spinal dorsal roots in the monkey. I. Cervical and upper thoracic dorsal roots, Amer. J. Anat., 123 (1968) 27-74. 42 STUART,D. G., GOSLOW, G. E., MOSHER, C. G., AND REINKIN6, R. M., Stretch responsiveness of Golgi tendon organs, Exp. Brain Res., 10 (1970) 463-476. 43 STUART, D. G., MOSHER, C. G., GERLACH, R. L., AND REINKING, R. M., Selective activation of Ia afferents by transient muscle stretch, Exp. Brain Res., 10 (1970) 477-489. 44 TABER, E., The cytoarchitecture of the brain stem of the cat. 1. Brain stem nuclei of the cat, J. comp. Neurol., 116 (1961) 27-69. 45 THACH, W.T., Discharge of cerebellar neurons related to two maintained postures and two prompt movements. I. Nuclear cell output, J. Neurophysiol., 33 (1970) 527-536. 46 THACH, W.T., Discharge of cerebellar neurons related to two maintained postures and two prompt movements. II. Purkinje cell output and input, J. Neurophysiol., 33 (1970) 537-547. 47 THACH, W. T., Cerebellar output: Properties, synthesis and uses, Brain Research, 40 (1972) 89-97. 48 WALKER,E. A., AND WEAVER,T. A., JR., The topical organization and termination of the fibers of the posterior columns in Macaca mulatta, J. comp. Neurol., 76 (1942) 145-158. 49 WOUDENBERG, R.A., Projections of mechanoreceptive fields to cuneate-gracile and spinal trigeminal nuclear regions in sheep, Brain Research, 17 (1970) 417-437.