EXPERIMENTAL

NEUROLOGY

117,287-298

(19%)

Forelimb Motor Performance Following Cervical Spinal Cord Contusion Injury in the Rat GREGORY W. SCHRIMSHERANDPAUL Departments

of Neuroscience

and Neurological

Surgery,

University

The purpose of this study was to examine the degree, persistence, and nature of forelimb behavioral deficits following cervical spinal cord contusion injury in the rat. Forelimb reaching and pellet retrieval, forehead adhesive sticker removal, and vibrissae-induced forelimb placing were examined for 16 weeks following a weight-drop injury (10.0 g-2.6 cm) at the C,-C, spinal level. Nine of 13 rats studied were unable to perform the pellet retrieval task due to pronounced forelimb extension hypometria. However, these animals did carry out the forehead sticker removal and vibrissae-induced placing tasks. Therefore, the loss of reaching ability related to pellet retrieval was not due to generalized paralysis. This interpretation was further supported by evaluation of the rostrocaudal extent of relative motoneuron loss from l-mm divisions through the lesion zone. The extent of motoneuron pathology ranged from 2 to 6 mm but was largely confined to the C,-C, spinal segments. Morphometric assessments of axonal sparing revealed that pellet retrieval performance during the last month of observation was significantly correlated with fiber sparing in the dorsal columns and ventral white matter, whereas no significant correlation could be demonstrated with regard to dorsolateral white matter. While there were no conspicuous differences in qualitative assessments of damage to interneuron pools (i.e., laminae V to VII) between the nonreaching and retrieval-recovered rats, the possibility of combined white and gray matter pathology contributing to this deficit still exists. These initial findings thus demonstrate that the weight-drop contusion injury model can be adopted to studies of cervical spinal cord trauma in the rat. Such lesions yield permanent deficits in forelimb function lending to future studies of possible therapeutic interventions. Furthermore, performance deficits observed at 1 week postinjury in the placing and forehead sticker removal tasks can be predictive of any potential for long-range spontaneous recovery in pellet retrieval ability. o 1992 Academic PUSS, IN.

INTRODUCTION The first significant attempt to produce a clinically relevant model of spinal cord injury was that of Allen

of Florida,

J. REIER College

of Medicine,

Gainesville,

Florida

32610

(1). In this paradigm, a defined weight is dropped from a known height onto the dorsal surface of the exposed spinal cord. This contusion injury technique precipitates hemorrhage, ischemia, inflammation, edema, demyelination, and posttraumatic central cavitation similar to that seen in many cases of human spinal cord injury (7, 15, 16, 25, 42). Accordingly, contusion/compression injury models have been used extensively in rats and cats to investigate many aspects of spinal cord injury. Included among these are studies of locomotor and hindlimb reflex impairments (10,33), the effects on functional outcome of pharmacological intervention during the acute postinjury phase (4,22,31), remyelination and the recovery of axonal conduction (9), and the feasibility of neural tissue transplantation into contusion/compression injury sites (5, 24, 35, 36). Most studies involving this trauma model have been limited to the examination of hindlimb deficits following midthoracic spinal cord injuries. The degree to which contusion/compression lesion approaches can be extended to cervical spinal levels has only been briefly explored (17, 181, although neck injuries account for a large proportion of the human spinal cord injury population. Accordingly, the cervical spinal cord injury model is of interest as it can provide an important opportunity to utilize established behavioral tests for evaluatingforelimb deficits. Cervical spinal contusion lesions, however, can also introduce several degrees of complexity beyond the loss of long-tract projections given the possibility of concomitant interneuronal damage ordirect spinal motoneuron involvement, especially with regard to the phrenic and brachial motoneuron pools. Therefore, the challenge in developing an experimental model of cervical spinal cord contusion injury is to create a lesion that is readily survivable and which can be sufficiently controlled so as not to produce flaccid forelimb paralysis secondary to widespread motoneuron death. The present study has established the feasibility of a C,--C, spinal cord contusion injury model in the rat. This injury produced a complete, predictable, and longterm deficit in forelimb reaching ability related to a pellet retrieval task, although these animals could perform other forelimb motor tasks. We have also examined the

287 All

Copyright 0 1992 rights of reproduction

0014-46S6/92 $5.00 by Academic Press, Inc. in any form reserved.

288

SCHRIMSHER

relationships between the extent of white matter sparing, spinal motoneuron loss, and interneuronal pathology and the severity of the resulting impairment on the forelimb reaching and pellet retrieval task. Initial findings of this study have been reported previously (40,41). MATERIALS

AND

METHODS

Animal Groups and Surgery

Thirteen adult (250-350 g) female Sprague-Dawley rats (Zivic-Miller, Allison Park, PA) received a weightdrop contusion injury at spinal cord level C,-C, using a paradigm analogous to that described by Wrathall and co-workers (44) for midthoracic contusion lesions. Rats were anesthetized with an intraperitoneal injection of ketamine (Aveco, Ketaset; 40 mg/kg) and xylazine (Haver, Rompun; 6.7 mg/kg) and their eyes were protected with Lacri-Lube (Allergan) to prevent drying. The C, and C, vertebrae were exposed and dorsal laminectomies performed. The vertebral column was stabilized by attaching anchored Allis clamps to the spinous processes of the C, and C, vertebrae. These clamps also served to suspend the thorax above the operating surface. Contusion injury involved a 10.0 g weight that was dropped for a distance of 2.5 cm before striking a 2.5mm tip diameter Teflon impounder that rested freely on the exposed dural surface. The impounder was placed on the dural surface just prior to the weight-drop, and was removed immediately thereafter. After injury, the overlying musculature was sutured in layers and the skin was closed with surgical wound clips. The rats were given a subcutaneous injection of penicillin (TechAmerica, Dual-Pen; 100,000 U) and an intraperitoneal injection of isotonic saline (4.0 cc). The animals were housed individually during the acute postoperative recovery period (- 1 week) that was characterized by total forelimb impairment and lack of body weight support. During this period, bowel and bladder function, respiratory performance, and hydration status (based on the condition of the mucous membranes) were monitored three to four times daily. A diet of crushed rat chow (Ralston Purina), fresh apple, and shelled sunflower seeds was made freely available on the cage floor along with a floor-accessible water bottle. The food and bedding were changed three to four times daily. The body weight of each animal was monitored daily. Once the animals became self-reliant, they were housed in pairs. Behavioral Tests

The animals were maintained on a 12:12 h 1ight:dark cycle. Behavioral testing and training sessions were conducted during the light portion of the cycle. The rats were tested the day before the contusion injury and thereafter at weekly intervals for 16 weeks.

AND REIER

Forelimb placing. The animals were held by hand from behind such that the forelimbs were free to move. The vibrissae on either the left or right side were then brought into contact with a tongue depressor clamped to a table top. The animal’s vibrissae were brushed against the tongue depressor a maximum of five times per trial during which placement of the ipsilateral forelimb was monitored. Forelimb placement was defined as positioning the underside of the forepaw on the upper surface of the tongue depressor in response to the vibrissae stimulation. Five sets of two alternating trials per side were performed, and these were separated by approximately 15-set intervals during which the rats were placed in a holding cage. A total of ten trials per forelimb were conducted during each testing session. Forelimb reaching. The testing apparatus (Fig. 1) consisted of a 30 X 25 X 30 cm (L X W X H) box with Plexiglas walls and a wire-mesh floor that was supported above a removable litter tray. Two 2.5 X 5.0-cm openings (W X H) were positioned at floor level in the front of the testing apparatus. Access through these openings was controlled by sliding Plexiglas doors. Plexiglas walls (10 X 10 cm) were positioned medially inside the apparatus on the left side of one access opening and the right side of the other access opening to bias use of either the left or right forelimb, respectively. A recessed tray was placed outside the testing apparatus 2.0 cm away from the front of the access openings. The tray was elevated 1.0 cm above the apparatus floor level. The rats were trained to reach with their forelimbs across the 2.0-cm gap between the access opening and the recessed tray and to retrieve a precision milled 45mg food pellet (BioServ) placed 1.0 cm from the edge of the tray (i.e., a total of 3.0 cm from the access opening). The rats were placed on a food deprivation schedule to facilitate training. This schedule consisted of an initial 48-h fast, after which the animals were given a daily rat chow ration (20-30 g) following each training session. The animals were trained for 2 weeks on the task. Under no circumstance would an animal have been allowed to drop below 80% of its initial body weight for the purpose of training. After the training period, normal task performance was established for each animal. Each rat was tested on 30 trials for both the left and right access openings. During a testing session, the rats were given 15 min to complete the 30 trials for each access opening. A trial consisted of 5.0 set of access to a pellet following an attempt to retrieve the pellet. A reaching attempt was operationally defined as a reaching movement which resulted in the rat contacting the recessed tray with its forelimb or extending its forelimb over the upper surface of the tray. The pellets were placed at the same position on the tray for each trial. Successful pellet retrieval was operationally defined as grasping the pellet

CERVICAL

SPINAL

CORD

CONTUSION

289

INJURY

FIG. 1. Forelimb reaching and pellet retrieval testing apparatus. The recessed tray and pellet are positioned for a trial of the left forelimb bias side of the testing apparatus. The access opening is indicated by “A,” the recessed pellet tray is indicated by “T,” the forelimb bias wall is indicated by “BW,” and the sliding door for the access opening is indicated by “D.” A pellet is located above the “T” in the photograph. See Materials and Methods for the description of the testing procedure and the dimensions of the testing apparatus.

and withdrawing it through the access opening. The first 5 trials for each access side were not scored and served to allow the rat to orient to the task. The subsequent 25 trials for each bias side (i.e., a total of 50 trials per testing session) were scored for both retrieval success and the forelimb used. If a rat used a given forelimb in more than 75% of the scored trials, it was classified as having a forelimb preference. Rats that used a particular forelimb less than 75% of the time were classified as ambidextrous. Data collection involved analysis of videotape recordings of each testing session. After normal testing, the rats received the contusion injury and were then tested on the reaching task once a week for 16 weeks using the same testing protocol. The rats also received a nonscored training period weekly between the scored testing sessions. The feeding schedule used during initial training was maintained during the postinjury period, except for the acute postsurgery recovery period when food was freely available. Forehead sticker removal. Rats were individually placed in a clean, empty cage identical to their home cages and left undisturbed for 5 min. A 1.9-cm diameter removable adhesive label (Avery) was then placed on the rat’s forehead, centered between the eyes and ears. The rats were observed for 3 min, and sticker removal performance for each forelimb was recorded. Sticker removal performance was graded on the following rating scale:

0, No removal attempted. 1, Forelimbs did not reach up as 2, Forelimbs reached as high as not contact the head medial to the 3, Forelimbs reached as high as head was contacted medial to the contact occurred. 4, Sticker was contacted by the attempt. 5, Sticker was removed from the

high as the sticker. the sticker, but did eye. the sticker and the eye, but no sticker forelimb

during

head.

Each testing session consisted of five trials separated approximately 30-set intervals. Light

the

by

Microscopy

After the 16-week postcontusion testing period, the rats were anesthetized with an overdose of sodium pentobarbital and perfused transcardially with isotonic saline followed by 5.0% glutaraldehyde and 4.0% paraformaldehyde in 0.1 M phosphate buffer. The level of the cervical spinal cord injury was confirmed, and the lesion site (i.e., the length of tissue containing cystic cavitation) was removed and cut into serial blocks of approximately l-mm thickness. The blocks were postfixed in osmium tetroxide, dehydrated, embedded in Epon, and sectioned at 2-pm thickness. The sections were stained with toluidine blue.

290

SCHRIMSHER

Morphometry

A section from the epicenter (i.e., the region of maximum pathology) of each lesion was examined using a Zeiss Axiophot microscope and morphometrically assessed with a Vidas Image Analysis System (Zeiss, KONTRON). The perimeter of each section was measured and the area defined by a circle having an equivalent circumference was calculated. This area was regarded as the maximal area (MA) that theoretically could have been occupied by the spinal cord at that level. This calculation was necessitated by the fact that cystic cavities in the lesion can collapse during histological processing (see Figs. 4h, 4j, and 4k). Consequently, measuring the actual areas of sections would have produced randomly variable data that could not be compared across animals. The areas of sparing within five regions of defined white matter were then assessed for each section. These five areas were the dorsal columns, the left and right dorsolateral white matter (defined as the white matter lateral to the dorsal horns and superior to a line that would bisect the dorsal and ventral halves of the spinal cord), and the left and right ventral white matter (defined as the white matter between the dorsal/ventral spinal cord bisector and the ventromedial fissure). Using 2-pm plastic sections, our assessments of spared white matter were based upon a discrimination between regions in which preserved axons and myelin predominated versus areas in which these elements were represented only modestly (Fig. 2). The area measurements for each region of spared white matter area were divided by the calculated maximal area of the respective section to allow comparisons between animals. In considering spared white matter, our measurements obviously included some foci of axonal pathology. Likewise, areas deemed as being nonspared sometimes contained scattered axons with normal morphology. Because these factors preclude exact measurements of sparing, nonparametric analyses of these data were performed. In 9 of 13 animals, the most caudal l-mm block of embedded tissue from the lesion site (i.e. the length of tissue containing cystic cavitation) yielded sections that were free of gray matter pathology. In these animals, an estimate of motoneuron number was obtained from one random cross-section from each of the l-mm blocks taken through the rostrocaudal extent of the lesion site and compared to the number of motoneurons in the section free of gray matter pathology from the most caudal block. The total number of presumptive motoneurons in a given cross-section was based on counts of large perikaryal profiles in the ventral horns, including cell body profiles at the level of the nucleus, as well as through nonnuclear regions of the soma where Nissl body aggregates were still present. A relative motoneuron index was thus derived as follows: sections having greater

AND

REIER

than 70% of the motoneuron total determined in the most caudal, nonpathological section was defined as having minimal or no damage; 40 to 70% of the caudalmost section motoneuron number was scored as moderately damaged, 10 to 40% of the caudalmost section motoneuron number was rated as being severely damaged; and less than 10% of the caudalmost section motoneuron number was defined as being maximally damaged. Interneuron pool pathology in laminae V to VII was qualitatively assessed as none (no obvious gray matter disruption), moderate (numerous neuronal cell bodies present along with some gray matter pathology), severe (few neuronal cell bodies discernible within the gray matter pathology), and total (no neuronal cell bodies present). Statistics

Correlations between morphological measurements and behavioral outcomes were based on Spearman’s rank order correlations. Comparisons of morphological measurements between groups of rats were based on the Mann-Whitney u test. Means are reported -1-1standard deviation. Significance values for all tests were chosen at P < 0.05. RESULTS General Clinical Features

A primary concern was that a-spinal C,-Cs contusion posed the potential for direct or evolving damage to the phrenic motoneuron pool which extends from spinal levels C, to C, (20). None of the rats, however, showed respiratory distress following the injury. The cervical contusions did demand strict attention to food and water availability and intake, since the first few days following injury typically involved total forelimb impairment. However, the capacity to maintain a quadrupedal stance rapidly recovered within 7 days following injury, and all rats thereby became self-reliant and no longer required any extraordinary care. General Histopathology

and Morphometric

Assessments

The lesion zones extended approximately 4 to 8 mm along the neuraxis. The pathology found at the epicenters was characterized by partitioned cystic cavities resulting from gray matter necrosis with some occasional sparing of the superficial dorsal horns (Figs. 3 and 4). The cysts were of maximum diameter at the lesion epicenter and tapered off in the dorsal columns both rostra1 and caudal to this level. White matter pathology was most prevalent in the dorsal columns, ventromedial funiculi, and the medial portions of the lateral funiculi. Axonal sparing was greatest at the subpial

CERVICAL

SPINAL

CORD

CONTUSION

INJURY

291

FIG. 2. Epicenter pathology gradient in the ventrolateral cervical spinal cord following contusion injury. The dashed line in the photomicrograph illustrates the demarcation between tissue that was deemed as spared (left of dashed line) and nonspared (right of dashed line) following the contusion injuries. Note that the degree of axonal sparing is greatest at the pial border (arrow) and approaches being nonexistent at the cystic cavity border. 20x.

margin and decreased to relatively little or no fiber preservation near the cysts (Fig. 2). Based on the comparative estimates of motoneuron counts, the length of moderate to maximal motoneuron pathology was seen along the neuraxis was over a distance of approximately 2 to 6 mm (Table 1). However, the rostra1 extent of motoneuron pathology may be somewhat exaggerated since these assessments were based on comparisons of motoneuron cell counts from normal-appearing gray matter in the most caudal part of the lesion (i.e., closer to the brachial enlargement). Damage to the interneuron pools of laminae V-VII (32) was similarly evaluated and found to have extended approximately 2 to 5 mm along the neuraxis (Table 1). Damage to the lamina VIII interneuron pool essentially mirrored the extent of pathology identified in the lamina IX motoneuron pools (32). Behavior Reaching and pellet retrieval. The normal pellet retrieval movement consisted of: (1) an extension of the limb through the access opening, (2) an extension of the

digits once the paw was near the target, (3) pellet contact with flexion of the digits, and (4) withdrawal of the limb with the grasped pellet through the opening. Following contusion injury, a profound reaching hypometria was the main performance-limiting factor in 9 of the 13 lesioned rats, and these animals never recovered the ability to make an operationally defined reaching attempt during the entire 16-week postinjury period (Figs. 4e-4m). This lack of reaching ability was not due to task inattentiveness. These animals remained at the access opening during each weekly postinjury testing session and would either try to use their forelimbs or tongues to retrieve pellets from the platform. These postinjury limb extension efforts would result in forward motion of the forelimb through the access opening, but the maximal forward progress would proceed only to wrist level. During these ineffective reaching efforts, the forepaw and forelimb typically would not be elevated above the level of the grid floor of the testing apparatus. This reaching impairment resulted more from a lack of limb elevation and shoulder flexion rather than elbow extension, although the latter was also limited.

292

SCHRIMSHER

AND

REIER

metric during reaching attempts, and their retrieval effectiveness was limited because their digits or paw would often catch on the edge of the pellet platform. This impairment resulted from inadequate elevation of the limb during the reaching motion and impaired shoulder flexion. The fourth rat recovered pellet retrieval ability at 1 week postinjury (Fig. 5). This animal did not show hypometric impairment, and its spinal cord exhibited the least pathology (Fig. 4a). Overall, the normal mean pellet retrieval success rate for all rats (n = 13) was 33.8 +- 5.1 retrievals per 50 attempts, while the mean success rate for the retrieval-recovered rats (n = 4) was only 10.8 + 7.5 retrievals per 50 attempts at 16 weeks postinjury.

FIG. 3. Transverse the cervical contusion pellet retrieval ability, ability. 2.5~.

sections through the epicenters of three of injuries. The animal depicted in A recovered while those in B and C failed to regain reaching

Three of the remaining four animals recovered pellet retrieval ability at 3 to 6 weeks postcontusion (Figs. 4b, 4c, 4d, and Fig. 5). Relative to preoperative baseline scores, their performance at 16 weeks postcontusion was significantly reduced (P -C 0.0001, Fisher’s exact probability test). These rats were also markedly hypo-

Forehead sticker removal. Normal rats can readily remove adhesive stickers placed on the forehead. Nine rats were tested on the full five-trial sticker removal protocol. Three rats received their injuries prior to the development of this protocol and received a maximum of only two trials per session. One rat was averse to being held while the sticker was affixed and would not perform the task normally. That animal was excluded from this portion of the study. Following cervical contusion injury, a full range of impairments in sticker removal was observed (see Materials and Methods for rating scale). The temporal progression of postinjury recovery in sticker removal ability occurred in the sequence defined by the rating criteria (Table 2). By 16 weeks postcontusion, all tested rats had attained either a rating of 3 or 4. The basis of the severe initial deficits was a lack of shoulder flexion and immobility at the elbow. But with time, the rats recovered shoulder and elbow flexion and could contact their foreheads or the sticker with their forelimbs (ratings 3 and 4). Actual sticker removal (rating 5) following the injury was very uncommon and occurred only three times. The nature of this deficit appeared to be an inability and/or lack of strength in the digits to grasp the edge of the sticker and pull it away. Typically, the stickers could not be dislodged by a simple swiping movement of the paw. Based on comparisons of normal to postlesion sticker removal performance in the nine rats that received the full five-trial protocol, postinjury sticker removal performance was significantly impaired (P < 0.05, Wilcoxon matched pairs test) at each week postinjury. Interestingly, those rats that were able to contact the forehead sticker with a forepaw (rating 4) at 1 week postinjury also recovered pellet retrieval ability (n = 4). Rats that did not recover forelimb reaching ability showed more severe sticker removal impairments (ratings 0 through 2) at Week 1 (n = 8). The probability of this predictive relationship between l-week sticker removal performance and eventual recovery of pellet retrieval ability was 0.002 (Fisher’s exact probability).

CERVICAL

SPINAL

CORD CONTUSION

INJURY

293

FIG. 4. Tracings depicting the epicenter pathology for each of the cervical contusion lesions. Black areas represent the regions of cystic cavitation, dark gray regions represent areas of tissue deemed as nonspared, light gray regions represent intact gray matter, and white regions represent areas of white matter deemed as spared. The rats depicted in A-D recovered pellet retrieval ability, while those in E-M never regained the ability to make a reaching attempt.

Vibrissae-induced placing. Normal rats readily and reliably place their forelimbs on a surface in response to vibrissae contact. Following cervical spinal cord contusion injury the rats demonstrated a wide range of impairment on this task. These deficits primarily resulted from a lack of shoulder flexion. Rats that eventually recovered pellet retrieval ability (n = 4) demonstrated 50% or greater placing accuracy at 1 week postinjury, whereas those that did not recover the ability to reach (n = 9) demonstrated a gross initial placing impairment. As a group, these nine rats showed 5.0 f 7.5% placing accuracy at 1 week postinjury and significantly reduced placing accuracy at all weeks postinjury (P < 0.001, post hoc Scheffe’s test) (Fig. 6). The probability of a chance occurrence of the predictive relationship between 1 week placing accuracy and long-term pellet retrieval recovery was 0.0014 (Fisher’s exact probability). Anatomical-Behavioral

Features

Cumulative pellet retrieval performance during the last 4 weeks of observation for all 13 rats was significantly correlated with axonal sparing in the dorsal columns (r8 = 0.6790, P < 0.05) and the ventral white matter (ipsilateral to the preferred forelimb in retrieval-recovered rats and the unilateral side with the greatest ventral sparing in nonreaching rats) (r* = 0.8041, P < 0.001). However, dorsolateral white matter (ipsilateral

to the preferred forelimb in retrieval-recovered rats and the unilateral side with the greatest dorsolateral sparing in nonreaching rats) was not significantly correlated with pellet retrieval performance (rS = 0.3517). As noted earlier, motoneuron and interneuron pathology could be uniformly assessedin 9 rats (Table 1). No significant correlations were observed regarding pellet retrieval performance relative to the rostrocaudal extent of either motoneuron (rS = -0.2184) or interneuron (rS = -0.6023) pool damage ipsilateral to the preferred forelimb in retrieval-recovered rats or the side with the greatest extent of pathology in nonreaching rats. The total sparing of white matter was significantly (2 = 2.623, P < 0.01) greater in retrieval-recovered rats (42.5 + 13.0% MA) than in nonreaching rats (22.6 * 3.8% MA). Axonal sparing in the dorsal columns was significantly (2 = 2.392, P < 0.05) greater in retrievalrecovered rats (6.8 f 2.6% MA) as compared to nonreaching rats (2.1 f 1.3% MA). Interestingly, the dorsal column corticospinal tract was severely disrupted in all animals. As illustrated in Fig. 7, tissue sections taken slightly rostra1 to the lesion epicenters of reaching-recovered rats showed extensive degenerative pathology ranging from highly vacuolated white matter to deeply stained gliotic tissue. Comparable pathology also was seen in the dorsal column corticospinal tract caudal to the level of lesion. Significantly (2 = 2.777, P < 0.01) greater sparing of the ventral white matter ipsi-

294

SCHRIMSHER

AND

TABLE Gray Rat ID

_I_

K

1

Pathology

< Rostra1

Ml-

I

Matter

REIER

Assessments

Epicenter

Caudal

MM/l

MMM/III

MMM/III

M/III

-/I

MMMIIII

MMMIIII

MMMIIII

>

-/MMMIIII

-I-

-l-

Note. Each box represents a l-mm segment of tissue through a lesion site. The epicenter is defined as the block with the greatest degree of gray matter pathology. “M” stands for motoneuron pools and “I” for interneuronal pools of laminae V to VII. The degree of neuronal pool pathology is represented as follows: “NNN” represents maximal or total destruction, “NN” represents severe damage, “N” represents moderate damage, and “-” represents minimal to no damage to the neuronal pool. The “Rat ID” designation in the far left column refers to the cord sections depicted in Fig. 3. The three animals depicted above the dark line recovered pellet retrieval ability, while the six animals depicted below the line did not recover the ability to make a reaching attempt. See Materials and Methods for a details of the pathology criteria and assessment procedures.

lateral to the preferred forelimb was also seen in retrieval-recovered rats (13.0 f 3.2% MA) as compared to the spinal cord side with the greatest ventral sparing in nonreaching rats (6.7 * 1.2% MA). On the other hand, no difference (2 = 1.157) was found in dorsolateral sparing ipsilateral to the preferred forelimb in retrievalrecovered rats (6.8 + 2.6% MA) as compared to the spinal cord side with the greatest dorsolateral sparing in nonreachers (4.9 + 2.0% MA).

DISCUSSION

The consequences of spinal cord injury can vary considerably based on the severity and nature of the lesion. Obviously, the spinal level at which trauma occurs is a major factor in this regard and can present alternative problems with regard to the development of future therapeutic approaches. Cervical spinal cord injuries are particularly interesting because the resulting deficits can be a product of a combination of white matter long tract interruption and/or spinal motoneuron and interneuron damage. Therefore, in the present investigation, we have examined the feasibility of producing experimental lesions of the cervical spinal cord that closely mimic those encountered in the clinical setting. Despite the potentially devastating nature of these lesions, our findings show that relatively extensive contusion injuries can be made at the C,-C, spinal level after which

the animals progressively regain self-sufficiency without exhibiting severe respiratory distress or flaccid paralysis. The lesions, however, were of sufficient magnitude to produce a complete and long-term (i.e., 16 weeks) impairment in forelimb motor performance in these animals. Forelimb Reaching Deficits

Nine of the 13 rats studied showed a permanent failure to make an operationally defined reaching attempt with regard to the pellet retrieval task. These rats, however, could use their forelimbs in attempting forehead sticker removal for which recovery was seen in all cases. Forelimbs also were utilized in placing tasks, locomotion, and grooming, and the animals could execute flexion of the shoulder when in an upright rearing or supported posture during the sticker removal or placing tasks. In contrast, shoulder flexion and forelimb elevation were markedly impaired when the animals were in a quadrupedal stance while attempting the reaching movement required in the pellet retrieval task. The rats, however, were not conspicuously ataxic and did not show any obvious lack of balance during locomotion or quadrupedal stance. Therefore, the differences between the postural components of the tasks (upright versus quadrupedal) do not appear to be the sole basis for the lack of reaching ability.

CERVICAL

!i?

‘-d

25

5

20

SPINAL

CORD

E 2

15 10

0 N

1

2

3

4

5

6

7

8

Weeks

9 10111213141516

Postinjury

FIG. 5. Pellet retrieval performance for the four rats that recovered retrieval ability. “N” indicates normal task performance while the numerals indicate the number of weeks postcontusion. The symbols indicate the total number of retrievals for each bias side on which the requisite 25 retrieval attempts were made (see Materials and Methods). Open triangles represent the performance of the rat depicted in Fig. 4a, closed triangles represent the performance of the rat depicted in Fig. 4b, the open squares represent the performance of the rat depicted in Fig. 4c, and the closed circles represent the performance of the rat depicted in Fig. 4d.

Spinal

Cord Pathology

and Forelimb

Motoneurons. General shown that spinal nerves

Deficits

anatomical studies have C&-C, in the rat innervate

CONTUSION

295

INJURY

muscles involved in lateral and flexive movements of the head, forward movement of the scapulae, extension of the shoulder joint, and elevation of the forelimb. Spinal nerves C, and C, contribute to the innervation of muscles involved in flexion of the shoulder and inward movement of the forelimb (21, 23). While direct or evolving damage to the motoneuron pools of these spinal segments following the contusion injury at spinal level C!,-C, could contribute to the pattern of forelimb impairment noted in this study, no overt differences were observed in the general rostrocaudal extent of motoneuron loss between the retrieval-recovered and nonreaching rats (Table 1). Nevertheless, the possibility of differential damage to motoneuron pools for particular muscles still cannot be eliminated, but this would imply a consistently similar degree of specific motoneuron involvement in all nine nonreaching animals. White matter. Previous studies involving similar pellet retrieval paradigms have demonstrated that lesions of the sensorimotor cortex and basal ganglia can produce a comparable reaching deficit characterized by extension hypometria, lack of coordinated limb movement, and impaired grasp ability (12,34,38,39,43). The reaching impairments produced by the cervical contusion injury appeared qualitatively similar to the deficits associated with these cortical/basal ganglia lesions. Focal corticospinal tract disruption produced by pyramidotomy results in a grasp deficit, but limb extension is not compromised although rats with pyramidal lesions involving concomitant damage to the medial lemniscus showed decreased forelimb extension capacity (13). While there was a significantly greater sparing of the dorsal columns in the rats that had recovered pellet

TABLE

2

Forehead Sticker Removal Performance Injury

N

1

2

3

4

5

8

Note. An injury status of “N” indicates normal task performance numerals in the far left column represent the behavioral criteria shaded boxes represent the modal performance of rats (n = 9) that the table indicate the number of these animals that performed at

7

Status 8

9

10

11

12

13

14

15

16

while the numerals indicate the number of weeks postcontusion. The large performance ratings that are detailed under Materials and Methods. The received the full five trial per session testing protocol, while the numerals in a given criteria level for each testing period.

296

SCHRIMSHER

N

1

2

3

4

5

6

Weeks

7

8

9

10111213141516

Postinjury

FIG. 6. Vibrissae-induced forelimb placing performance. “N” indicates normal task performance while the numerals indicate the number of weeks postcontusion. Open circles represent the mean placing accuracy of reaching-recovered rats (n = 4) and filled circles represent the placing accuracy of nonreaching rats (n = 9). The bars represent one standard deviation.

retrieval ability in the present investigation, the corticospinal portion of the dorsal columns (11) was disrupted severely in all animals whether pellet retrieval ability was ultimately recovered or not. This suggests that the ascending sensory fiber components of the dorsal col-

AND

REIER

umns (6,8, 14,191 could play an important role in mediating the performance of skilled forelimb reaching movements in the rat. It is important to bear in mind that cortical/basal ganglia lesions not only disrupt direct corticospinal projections, but also interrupt cortical input to brainstem nuclei giving rise to other descending spinal systems (e.g., the reticular and vestibular nuclei). In this regard, we observed a significant correlation between sparing of ventral white matter and retained pellet retrieval performance. This was coupled with a significantly greater preservation of ventral white matter ipsilateral to the preferred limb in retrieval-recovered rats as compared to the greatest unilateral ventral sparing in nonreaching rats. This is consistent with the fact that both the reticulospinal and vestibulospinal projections have been suggested to exert a major influence over axial and proximal limb musculature (29). Propriospinal interneurons. No correlation was found between interneuronal pathology and pellet retrieval performance during the last 4 weeks of behavioral testing. This result does not eliminate the possibility that damage to propriospinal neurons and/or their projections may still contribute to the lack of reaching ability. In the cat, for example, C&C!, propriospinal neurons project to forelimb motoneurons and receive monosynaptic corticospinal, rubrospinal, and reticulospinal inputs (26, 27). This circuitry is hypothesized to integrate input from subcortical centers that update the execution of a cortical command (3). Disruption of the C&-C!, propriospinal projections at spinal level C, pro-

FIG. 7. Dorsal columns of rats that recovered pellet retrieval ability. These sections were taken 2 to 3 mm rostra1 of the Note the prominent pathology in the region of the dorsal column corticospinal tract of each animal (arrows). “FG” denotes fasciculus gracilis and “C” denotes the location of the central canal. The dorsal columns illustrated in A and B are from whose lesion epicenters are depicted in Figs. 4a and 4b, respectively. These animals had the greatest degree of dorsal column the injuries. 10X.

lesion epicenters. the location of the the same animals sparing following

CERVICAL

SPINAL

CORD

duces an ataxic and aiming deficit in targeted reaching, but does not abolish the ability to execute the reaching movement. In contrast, a dorsal lesion at C!, that disrupts corticospinal and rubrospinal inputs to the C,C, propriospinal interneurons and forelimb motoneuron pools impairs the ability to make a reaching attempt when examined at early testing intervals. However, the ability to make a reaching attempt recovered in 3 to 12 weeks (2). If comparable circuits are present in the rat, then the reaching deficits seen in the current study could result from a combined loss of propriospinal interneuron and descending supraspinal inputs to the forelimb motoneuron and interneuron pools. The latter consideration would be congruent with the fact that there was significantly less dorsal column and ventral white matter sparing in the nonreaching rats although differences in interneuronal survival were not obvious. Cervical Spinal

Cord Contusion

as an Injury

CONTUSION

spinal injury is near nuclei giving rise to descending spinal projections, the regeneration potential of these neurons may be greatly enhanced relative to injury at more caudal spinal levels (28,30, 37). The location of the injury site immediately rostra1 to the major forelimb motoneuron pool of the brachial enlargement also raises interesting possibilities related to intraspinal fetal cell transplants. In principle, such grafts essentially involve the introduction of embryonic gray matter which could serve as a novel functional relay between separated regions of the spinal cord (28,3.5,36). Given the possible involvement of interneuronal populations, such as the C,-C, propriospinal neurons, the cervical spinal cord contusion model could therefore provide a unique opportunity for testing this proposed role of transplants under a clinically relevant condition. ACKNOWLEDGMENTS

Model

To date, most spinal cord injury research involving contusion/compression lesions has concentrated on thoracic level injuries and deficits primarily associated with white matter damage. Trauma sites near the interneuronal and motoneuronal populations of the limb enlargements, however, introduce other levels of anatomical and functional complexity by virtue of potential gray matter involvement. This issue needs to be addressed experimentally not only for its practical relevance but also because of the biologically intriguing possibilities for future studies of spinal cord functional sparing and recovery. The potential variabilReproducibility of the lesion. ity encountered with a contusion lesion is of concern since spontaneous recovery secondary to failure to produce an adequate lesion could lead to misinterpretation of experimental results. On the other hand, lesion variability in the chronically injured spinal cord is of less concern since a stable baseline deficit can be determined and thus serve as a point of reference for assessing any therapeutic strategy. In that regard, the lesions produced in this study are certainly compatible with chronic injury models. Furthermore, we have found that performance on the forehead sticker removal and vibrissae-induced placing tests at l-week postinjury could be used to screen effectively for potential spontaneous recovery of pellet retrieval ability at later postlesion times. This predictive information would be of considerable value in studies of therapeutic approaches involving a short delay of at least 1 week in the initiation of an experimental treatment protocol. Theoretical implications for functional repair. Despite the challenges involved, contusion injury of the cervical spinal cord provides a uniquely advantageous model in which to examine several issues regarding potential treatments. For example, since the cervical level

297

INJURY

The research was supported by NIH Grant MH15737 (G.W.S.), NIH Grant NS72300 (P.J.R.), NIH Grant NS27511 (P.J.R.), the Mark F. Overstreet Fund for Spinal Cord Injury Research, and the funding through the State of Florida Speeders and Impaired Drivers Trust. The authors wish to thank Ms. Minnie Smith and Ms. Barbara O’Steen for their excellent technical assistance. Sincere appreciation is also extended to Dr. Charles Vierck, Jr., Dr. Brad Stokes, and Mr. Edward Wirth, III for their advice and support.

REFERENCES 1.

ALLEN, A. R. 1911. Surgery of experimental lesion equivalent to crush injury of fracture dislocation umn. J. Am. Med. Assoc. 57: 878-880.

of spinal of spinal

cord col-

2. ALSTERMARK,

B., A. LUNDBERG, U. NORRSELL, AND E. SYBIRSKA. 1981. Integration in descending motor pathways controlling the forelimb in the cat. 9. Differential behavioural defects after spinal cord lesions interrupting defined pathways from higher centres to motoneurones. Exp. Brain Res. 42: 299-318.

3.

ALSTERMARK, B., T. GORSKA, A. LUNDBERG, AND L.-G. PETTERSSON. 1990. Integration in descending motor pathways controlling the forelimb in the cat. 16. Visually guided switching of target-reaching. Exp. Brain Res. 80: l-11.

4. ANDERSON,

D. K., T. R. WATERS, AND E. D. MEANS. treatment with alpha tocopherol enhances neurologic after experimental spinal cord compression injury. trauma 5: 61-67.

1988. Prerecovery J. Neuro-

5. ANDERSON,

6.

7.

8.

9.

D. K., P. J. REIER, E. D. WIRTH, III, D. P. THEELE, T. MARECI, AND S. A. BROWN. 1991. Delayed grafting of fetal CNS tissue into chronic compression lesions of the adult cat spinal cord. Rest. Neurol. Neurosci. 2: 309-325. ARVIDSSON, J., AND K. PFALLER. 1990. Central projections of C,-C, dorsal root ganglia in the rat studied by anterograde transport of WGA-HRP. J. Comp. Neural. 292: 349-362.

BALENTINE, J. D. 1978. Pathology of experimental spinal cord trauma. I. The necrotic lesion as a function of vascular injury. Lab. Invest. 39: 236-253. BASBAUM, A. I., AND P. J. HAND. 1973. Projections of cervicothoracic dorsal roots to the cuneate nucleus of the rat, with observations on cellular “bricks”. J. Comp. Neural. 148: 347-360. BLIGHT, A. R., AND W. YOUNG. 1989. Central axons in injured

298

10.

11. 12. 13. 14.

15. 16.

17. 18. 19.

20.

21. 22.

23. 24. 25.

26.

27.

28.

SCHRIMSHER cat spinal cord recover electrophysiological function following remyelination by Schwann cells. J. Neural. Sci. 91: 15-34. BRESNAHAN, J. C., M. S. BEA’ITIE, F. D. TODD, III, AND D. H. NOYES. 1987. A behavioral and anatomical analysis of spinal cord injury produced by a feedback-controlled impaction device. Exp. Neurol. 95: 548-570. BROWN, L. T. 1971. Projections and termination of the corticospinal tract in rodents. Exp. Brain Res. 13: 432-450. CASTRO, A. J. 1972. The effects of cortical ablations on digital use in the rat. Brain Res. 37: 173-185. CASTRO, A. J. 1972. Motor performance in rats: The effects of pyramidal tract section. Brain Res. 44: 313-323. CLIFFER, K. D., AND G. J. GIESLER. 1989. Postsynaptic dorsal column pathway of the rat. III. Distribution of ascending afferent fibers. J. Neurosci. 9: 3146-3168. DAS, G. P. 1989. Perspectives in anatomy and pathology of paraplegia in experimental animals. Brain Res. Bull. 22: 7-32. DE LA TORRE, J. C., M. T. RICHARD, AND L. P. IVAN. 1986. Experimental problems in spinal cord neural reconstruction. Erp. BrainRes. 13:317-325. FADEN, A. I., T. J. JACOBS, AND J. W. HOLADAY. 1982. Comparisons of early andlate naloxone treatment in experimental spinal cord injury. Neurology 32: 677-681. FADEN, A. I., AND T. J. JACOBS. 1985. Effect of TRH analogs on neurologic recovery after experimental spinal trauma. Neurology 35: 1331-1334. GIESLER, G. J., R. L. NAHIN, AND A. M. MADSEN. 1984. Postsynaptic dorsal column pathway of the rat. I. Anatomical studies. J. Neurophys. 51:260-275. GOSHGARIAN, H. G., AND J. A. RAFOLS. 1981. The phrenic nucleus of the albino rat: A correlative HRP and Golgi study. J. Comp.Neurol. 201:441-456. GREENE, E. C. 1963. Anatomy of the Rat. Heffner Publishing Co., New York. HALL, E. D., AND P. A. YONKERS. 1990. Preservation of motor nerve function during early degeneration by the 21-aminosteroid anti-oxidant U74006F. Brain Res. 513: 244-247. HEBEL, R., AND M. W. STROMBERG. 1986. Anatomy and Embryology of the Laboratory Rat. Biomed Verlag, Worthsee, Germany. HOO~I,ER, D. W., AM, J. R. WRATHALL. 1991. Implantation of neuronal suspensions into contusive injury sites in the adult rat spinal cord. Acta Neuropathol. 81: 303-311. HUGHES, J. T. 1988. Pathological changes after spinal cord injury. In Spinal Cord Dysfunction: Assessment (L. S. Illis, Ed.), pp. 34-40. Oxford Univ. Press, Oxford. ILLERT, M., A. LUNDBERG, AND R. TANAKA. 1977. Integration in descending motor pathways controlling the forelimb in the cat. 3. Convergence on propriospinal neurones transmitting disynaptic excitation from the corticospinal tract and other deacending tracts. Exp. Brain Res. 29: 323-346. ILLERT, M., A. LUNDBERG, Y. PADEL, AND R. TANAKA. 1978. Integration in descending motor pathways controlling the forelimb in the cat. 5. Properties of and monosynaptic excitatory convergence on C3-C4 propriospinal neurones. Exp. Brain Res. 33: 101-130. JAKEMAN, L. B., AND P. J. REIER. 1991. Axonal projections between fetal spinal cord transplants and the adult rat spinal cord:

AND REIER

29.

30.

31.

32.

33. 34. 35.

36.

37.

38.

39. 40. 41.

42. 43.

44.

A neuroanatomical tracing study of local interactions. J. Comp. Neurol. 307: 311-334. KUYPER~, H. G. J. M. 1981. Anatomy of descending pathways. In Handbook of Physiology: The Nervous System (J. M. Brookhart, V. B. Mountcastle, V. B. Brooks, and S. R. Geiger, Eds.), Vol. 2, pp. 597-666. American Physiology Society, Bethesda, MD. LAM-I-BARREAULT, N., P. REGA, AND P. GAUTHIER. 1991. Axonal regeneration from central respiratory neurons of the adult rat into peripheral nerve autografts: Effects of graft location within the medulla. Neurosci. L&t. 125: 121-124. MEANS, E. D., D. K. ANDERSON, T. R. WATERS, AND L. KALAF. 1981. Effect of methylpredniaolone in compression trauma to the feline spinal cord. J. Neurosurg. 55: 200-208. MOLANDER, C., Q. Xv, C. RIVERO-MELIAN, AND G. GRANT. 1989. Cytoarchitectonic organization of the spinal cord of the rat. II. The cervical and upper thoracic cord. J. Comp. Neurol. 289: 375-385. NOBLE, L. J., AND J. R. WRATHALL. 1989. Correlative analyses of lesion development and functional status after graded spinal cord contusive injuries in the rat. Exp. Neurol. 103: 34-40. PISA, M. 1988. Motor functions of the striatum in the rat: Critical role of the lateral region in tongue and forelimb reaching. Neuroscimce24:453-463. REIER, P. J., D. K. ANDERSON, F. J. THOMPSON, AND B. T. STOKES. 1992. Neural tissue transplantation and CNS trauma: Anatomical and functional repair of the injured spinal cord. J. Neurotrauma 9: S223-8248. REIER, P. J., B. T. STOKES, F. J. THOMPSON, AND D. K. ANDERSON. 1992. Fetal cell grafts into resection and contusion/compression injuries of the rat and cat spinal cord. Exp. Neurol. 115:177-E@. RICHARDSON, P. M., V. M. K. ISSA, AND A. J. AGUAYO. 1984. Regeneration of long spinal axons in the rat. J. Neurocytol. 13: 165-182. SABOL, K. E., D. B. NEILL, S. A. WAGES, W. H. CHURCH, AND J. B. JUSTICE. 1985. Dopamine depletion in a striatal subregion disrupts performance of a skilled motor task in the rat. Brain Res. 335: 33-43. SCHNEIDER, J. S., AND U. E. O-AL. 1984. Behaviorally specific limb use deficits following globus pallidus lesion in rats. Brain Res. 308: 341-346. SCHRIMSHER, G. W., F. J. THOMPSON, AND P. J. REIER. 1990. Behavioral correlates of cervical spinal cord contusion injury in the rat. Sot. Neurosci. Abstr. 16: 112. SCHRIMSHER, G. W., AND P. J. REIER. 1991. Forelimb motor performance following cervical spinal cord contusion injury, dorsal column lesion, or dorsolateral funiculus lesion in the rat. Sot. Neurosci. Abstr. 17: 1025. WAXMAN, S. G. 1989. Demyelination in spinal cord injury. J. Neurol. Sci. 91: 1-14. WHISHAW, I. Q., W. T. O’CONNOR, AND S. B. DUNNETT. 1986. The contributions of motor cortex, nigrostriatal dopamine and caudate-putamen to skilled forelimb use in the rat. Brain 109: 805-843. WRATHALL, J. R., R. K. PETTEGREW, AND F. HARVEY. 1985. Spinal cord contusion in the rat: Production of graded, reproducible, injury groups. Exp. Neurol. 88: 108-122.

Forelimb motor performance following cervical spinal cord contusion injury in the rat.

The purpose of this study was to examine the degree, persistence, and nature of forelimb behavioral deficits following cervical spinal cord contusion ...
4MB Sizes 0 Downloads 0 Views