Altered patterns of reflex excitability subsequent to contusion injury of the rat spinal cord.

JOURNAL OF NEUROPHYSIOLOGY Vol. 68, No. 5, November 1992. Printed

in U.S.A.

Altered Patterns of Reflex Excitability Subsequent to Contusion Injury of the Rat Spinal Cord F. J. THOMPSON, P. J. REIER, C. C. LUCAS, AND R. PARMER Departments of Neuroscience and Neurological Surgery, College of iMedicine, University of Florida Brain Institute, Gainesville, Florida 32610 SUMMARY

AND

CONCLUSIONS

I. The present study investigated regulation of reflex excitability after experimental contusion injury of the spinal cord. 2. Four measures of H-reflex excitability were evaluated in normal rats and at 6,28, and 60 days after contusion injury at the Ts level: 1) reflex thresholds, 2) slope of the reflex recruitment curves, 3) maximal plantar H-reflex/maximal plantar M-response (H,,/M,,) ratios, and 4) rate-sensitive depression (i.e., the decrease in reflex magnitude relative to repetition rate). 3. Tested as a function of the afferent volley magnitude, the thresholds for reflex initiation fell progressively subsequent to contusion injury. No change was observed at 6 days postinjury, and the decrease at 28 days was not significant. However, by 60 days postinjury, the threshold had decreased by 23% of the maximal afferent volley, and this decrease was significant, [analysis of variance (ANOVA, P 5 0.01)] . ratios elicited in postcontusionanimals at 0.3 4. kax/wn, Hz werenot significantly different from thoserecordedin normal animals. 5. The slopesof the recruitment curvesweremarkedly reduced subsequentto contusion injury. The decreasewas greatestat 6 dayspostinjury. Although somerecovery toward normal occurred at 28 and 60 dayspostinjury, the slopesof recruitment curves in postcontusionanimalsremainedsignificantly decreased. 6. H-reflexeselicited at l-5 Hz were lesssensitiveto rate depressionin postcontusionanimalsthan in normal animalsat the samerespectivefrequencies.The decreasewasprogressivein onset,becomingsignificantby 28 days postinjury, and of an enduring nature, i.e., still significantly different from normal in the reflexestested60 dayspostinjury. 7. Rate sensitivity of the tibia1 nerve monosynaptic reflex (MSR) wasalsocomparedin normal and postcontusionanimals. Rate sensitivity of the tibial MSRs wassignificantly reducedat 28 and 60 dayspost-contusion,comparedwith normal animals. 6. Thesedata indicate that significantchangesin lumbar reflex excitability resultfrom midthoracic contusioninjury of the spinal cord. Thesechangesinclude reflex threshold,slopeof recruitment, and rate-sensitivedepression.Although recruitment slope was mostalteredin the shortestpostinjury interval tested,followed by somerecovery, the other changeswere progressivein onset and enduringin duration. INTRODUCTION

Several models of spinal cord injury have been used to investigate mechanisms related to dysfunction, plasticity, and the potential for stimulating regeneration or anatomic sparing. Lesions produced by complete or partial transection of the spinal cord have been studied most extensively

(e.g., de la Torre 1984; Young 1989). Because these injuries can yield reproducible deficits (Guth et al. 1980), they are considered particularly appropriate for the elucidation of structure-function relationships (Goldberger 1989). A limitation of the transection-type lesion is its departure from the form of injury commonly encountered in the clinical setting (Young 1989). The majority of human spinal cord injuries are produced by acceleration-related fracture / dislocations of the spine causing a contusion or compression of the spinal cord (Hughes 1988; Young 1989). The first attempt to mimic this form of injury experimentally is credited to Allen ( 19 1 1 ), who developed a weight-drop trauma model. Although this approach reproduces many of the histopathological features of injuries seen clinically ( Kakulas 1985 ) , experimental contusion /compression models have been challenged for their animal-to-animal variability (see reviews by Das 1989; de la Torre 1984; Young 1989). Accordingly, contusion/compression lesions have been viewed critically as to their usefulness for elucidating mechanisms associated with spinal cord dysfunction, functional sparing, or the recovery of function. These issues have been addressed by new experimental protocols reported to yield compression /contusion injuries with greater reproducibility (Anderson et al. 1988; Beattie et al. 1988; Black et al. 1986; Bresnahan et al. 1987; Kerasidis et al. 1987; Wrathall et al. 1985). From another perspective, the possibility has been raised that this type of injury may be more responsive to certain therapeutic strategies aimed at regeneration or functional sparing/recovery (Anderson et al. 199 1; Bracken et al. 1992; Reier et al. 1992; Stokes and Reier 1992; Young 1989). The feasibility of using fetal neural tissue grafts to repair acute and chronic contusion/compression lesions of the spinal cord has also been demonstrated (Anderson et al. 199 1; Reier et al. 1988, 1989, 1992; Stokes and Reier 1992; Winialski et al. 1987). Taken together, these considerations provide incentive for exploring this lesion model in more depth. However, apart from analyses of evoked potentials (Black et al. 1986; Fehlings et al. 1988; Shiau et al. 1992; Simpson and Baskin 1987; Zileli and Schramm 199 1) and axonal conduction ( Blight 1983, 1985 ) studies, neurophysiological changes have not been extensively investigated subsequent to contusion spinal injury. In the present study, specific attention was given to neurophysiological measures that may be related to the development of enhanced excitability of hindlimb reflexes, because

0022.3077/92 $2.00Copyright0 1992The AmericanPhysiological Society

1473

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (132.174.250.220) on September 18, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

1474

THOMPSON,

REIER, LUCAS, AND PARMER

segmental hyperreflexia is a hallmark of transection-type lesions in animals (Hultbom and Malmsten 1983a,b) and of spinal trauma in humans (e.g., Burke 1988; Landau 1974). Accordingly, four aspects of reflex excitability were tested: I) reflex thresholds, 2) slope of the reflex recruitment curve (i.e., an estimate of reflex gain), 3) maximal plantar H-reflex / maximal plantar M-response ( H,, / M-) ratios, and 4) rate-sensitive depression (i.e., the attenuation of reflex magnitude as a function of stimulus repetition rate). Evaluation of rate-sensitive depression of the tibial monosynaptic reflex (MSR) was also compared in normal and postcontusion animals to provide an additional test wherein reflex excitability was limited to central synapses. Preliminary results of this study have been summarized elsewhere (Thompson et al. 1989, 1990). METHODS

Animals A total of 43 adult femaleSprague-Dawleyrats (220-250 g initial weight) obtained from Zivic Miller Laboratories (Allison Park, PA) wereusedin theseexperiments.The ratswerehousedin an American Association for Lab Animal Care accreditedfacility, maintained on a 12-12h light/dark cycle and given Purina Rat Chow and water ad libitum. All postoperative care of theseanimals was supervisedby the attending veterinarian directly involved with the Spinal Cord Injury Program at the University of Florida.

minimally depressthe spinalmonosynapticreflex, and it doesnot alter the time courseof presynapticinhibition in contrastto other anesthetics,such aspentobarbital sodium,which are reported to depressthese reflexes and to prolong presynaptic inhibition ( Lodgeand Anis 1984;Tang and Schroeder1973). Depth of anesthesiawasregulatedby usinga doserate sufficientto block comeal reflexes,whisker tremor, and the pinna reflex. The trachea was intubated. Glycopyrrolate ( Robinul-V Injectable, A. H. Robins, 0.4 mg/kg) was given before anesthesiaand at 6-h intervals to decreaserespiratory secretions.Core body temperaturewasmonitored with a rectal thermistor and maintained between35-37°C using radiant heat. To prevent dehydration, subcutaneousinjections of 0.5 ml of isotonic glucose-saline, pH balancedto 7.4, were administeredat 0.5-h intervals. The axial skeletonwasimmobilized with spinalclamps,andthe lumbar enlargementwasexposedby laminectomy. The right tibial nerve wasthen exposedin the popliteal fossaand mounted on a stimulating bipolar electrode.Exposedtissueswere covered with warm mineral oil. The cord dorsumpotential wasrecordedfrom the lateral intermediatesulcusof the fifth lumbar cord segmentusinga silver ball electrode.The referenceelectrodewas located in the skin of the adjacent wound margin. Electromyograms (EMGs) were recordedfrom the digital interosseous musclesinsertingon the plantar surfaceof the fourth and fifth metatarsals.Thesemuscleswill be referredto asplantar muscles.The EMG electrodewasinserted percutaneouslyto lie in the sulcusbetweenthe fourth and fifth metatarsals.The referenceelectrodewaslocatedin the skin of the fifth digit.

H-reflex measurements

Plantar H-reflexes were elicited by stimulation of the tibial nel-veusingsingle200.pspulsesat 3-sinterstimulusinterval. The We used a modification of the Allen weight-drop method cathode was oriented proximal to the anode. Recordingsof the (Wrathall et al. 1985) to induce spinal cord injuries at the T, H-reflex typically consistof two EMG responses-an initial Mspinallevel in 25 animals.For consistencywith other descriptions wave and a later H-wave (Fig. 1). The M-responseis the result of of contusion injuries in the rat (Wrathall et al. 1985), the lesion the direct activation of the motor axons and doesnot involve a operations were performed under general anesthesiawith 4% spinalcircuit. The later H-reflex is a compound EMG responsein chloral hydrate ( 10ml/ kg ip). Eachanimal alsoreceiveda preop erative doseof penicillin ( Longicil, 100,000U im). After laminectomy, a 10-gweight wasdropped from a height of 2.5 cm onto a Teflon impounderthat restedfreely on the dura. After injury, the muscle,fascia, and skin were closedin layers. All animals were examineddaily for signsof distress,weight loss,dehydration, and boweland bladderdysfunction during the interval betweeninjury and the terminal physiologicalrecordingsessions at either 6,28, or 60 days.Manual expressionof bladderswasperformed 2-3 times daily aslong asrequired, and the animalsweremonitored for the possibilityof urinary tract infection, in which event the antibiotic Gentocin (Shering, 5 mg/kg im) wasadministered.The animals were paraplegicduring the first week and required assistancein bladder expressionduring this time. However, ambulation and bladderexpressionrecoveredprogressivelyduring the secondpostoperative week.At 1 and 2 mo, the animalsutilized quadrupedal locomotion, although the hind limbs appearedto be weak and 0 5 10 15 20 hypotonic. SeeKerasidiset al. ( 1987) and Noble and Wrathall Msec ( 1989) regarding details on locomotor recovery subsequentto FIG. 1. Top:cord dorsum potential recorded from the medial aspect of midthoracic contusioninjuries.

Contusion lesions

Acute electrophysiology recording experiments The animalswere anesthetizedby intramuscular injection of ketamine (Ketaset, Parke-Davis, 90 mg/kg) and maintained by constant intraperitoneal infusion of this drug (80420 mg kn-‘. h-l ). Ketamine waschosenbecauseit hasbeenshownto l

the dorsal root entry zone of L5, elicited by stimulation of the posterior tibial nerve at 1.4 X threshold for the appearance of the plantar H-reflex. Bottom:plantar H-reflex electromyogram (EMG) elicited by stimulation of the posterior tibial nerve at 1.4 x threshold for the appearance of the plantar H-reflex. Delivery of the stimulus for both traces occurred at the arrowat bottom.Amplitude calibration is 0.1 mV for the cord dorsum potential and 2.0 mV for the EMG. TPS, triphasic spike. Dotted lines: boundaries for tierent vollev waveform analvsis.


SPINAL REFLEX

EXCITABILITY

the plantar muscleelicited by the synapticactivation of motoneurons by muscleafferents. The M and H componentswere recordedin responseto a wide rangeof stimulusintensities.The total motor unit responsewas determinedby supramaximalstimulation of the axonsof the tibia1 nerve to produceM,,. The integratedareaof activity from baseline to baselinewas usedto compare the magnitude of M-wave and H-reflex waveforms. The prestimulus baselinewas usedto determinebaselinevaluesfor both the M and H waves.The H,,/ M,, ratio provided an index of the proportion of motoneurons recruited via a monosynapticreflex relative to the total motoneuron pool (Magladery et al. 1951; Taborikova 1966). Accordingly, the H-reflex magnitude,expressedasa fraction of the maximal M-wave, provided a standardthat could be referencedacrossanimals.To determinethe within-animal variability of the M,,, 5 consecutiveseriesof M,, waveformswererecordedin three normal animals,three28-day postcontusionanimals,and two 60-clay postcontusionanimals.The largeststandarddeviation calculated in the normal animalswas 1.07%of the mean, whereasit was 1.09%and 1.17% of the mean for the 28- and 60-day groups, respectively. H-reflex recruitment curves were produced by plotting reflex magnitudeas a function of the percent maximum magnitude of the dorsalroot afferent volley. The proportion of activated afferentswasestimatedusingRall’s method: a lead waslocated on the (intact) dorsalrootsnearthe dorsalroot entry zone to measurethe incoming afferent volley recordedasthe initial (i.e., Sl ) triphasic spike that was initiated peripherally in the tibial nerve (Rall 1955). The S1 triphasicspikewasintegratedfrom onsetto return to prestimulusbaseline.At higher stimulusintensities,the falling edgeof the triphasicspikewasinterrupted by the onsetof the Nl intramedullary field potential. At this point, a perpendicularwas droppedto the prestimulusbaselineto excludethe slowpotentials from the waveform areameasurement.(dotted lines in Fig. 1 define the limits for triphasicspike waveform analysis). The signal averageof 32 consecutiveSl and H-reflex waveforms were recordedduring a stimulusseriesthat beganat reflex threshold and then proceededin 10%stimulus increments from threshold to maximal Sl magnitude.Recruitment curveswereconstructed by plotting H-reflex magnitudeagainstthe percentmaximum magnitude of the correspondingS1 volley. The slopeof the recruitment curve wascalculated from a least-squareslinear regressionline fitted to the recruitment curve. The initial four points of the recruitment curve were most linear, and were thus usedfor this analysis.Theseinitial four points usedfor slopeanalysis( 1.O-1.3 T) are plotted in Fig. 2. The coefficientsof linearity calculated from the linear regressions for individual groupsrangedfrom 0.90 to 0.97. A changein the slopeof the recruitment curve wasinterpreted asa correspondingchangein reflex gain. Reflex magnituderesponses asa function of repetition rate were obtainedusingthe minimum stimulusintensity required to elicit maximal reflexes(H,,). A control repetition rate of 0.3 Hz was used for both H-reflexes and monosynaptic reflexes, consistent with the original studiesof Ecclesand Rall ( 1951) and Jefferson and Schlapp( 1953), which describedrate-sensitivedepressionas a diminishedresponseproduced by successivestimuli falling at intervals of w

H-reflex components In normal animals, H-reflexes were recorded in the plantar muscles and were composed of both M and H components (Fig. 1). The initial M-wave appeared after a latency of3.11 +O.l2(SE)ms(n=8)followedbytheH-wavewith a latency of 9.0 t 0.24 ms (n = 8) (Fig. 1, bottom). The M-wave was initiated at 1.O- 1.2 x threshold for the H-reflex and reached maximal amplitude between 2 and 3 X reflex threshold. The H-reflex was obliterated either by transection of the tibia1 nerve proximal to the stimulating electrode or by transection of the lumbar dorsal or ventral roots. The M-wave was affected only when the tibia1 nerve was transected distal to the stimulating electrode. In the initial experiments, these transections were performed to confirm 1) that the initial component designated as the M-wave was associated with the direct excitation of the motor nerves in the tibia1 nerve and 2) that the H-reflex component was elicited by centripetal conduction of the afferent activity and central excitation of the tibia1 motoneurons. Recruitment

Threshold Inspection of the starting point of these curves revealed that in the normal animals, the H-reflexes were initiated at 66.5 -t 2.16% (mean t SE, n = 10) of the maximal afferent volley. However, the thresholds fell progressively when tested at 6, 28, or 60 days subsequent to contusion injury. Whereas no change in reflex threshold was observed at 6 days postcontusion, by 28 days reflex threshold was reached at 56.4 t 4.4% (n = 10) of the maximal afferent volley. By 60 days postcontusion, reflex threshold had dropped to 43.1 t 6.52% (n = 8) of the maximal afferent volley. This latter drop of 23.5% from the normal reflex threshold represented a significant decrease in threshold relative to the normal animals [P 5 0.0 1 ( Fig. 3A )] .

70 60

Z

g

50

W

t= 40 4

B

8.0 6.0 W m 4.0 0 4 2.0 0.0

0.944

C x 0.8

curves

H-reflex recruitment patterns were examined by plotting reflex magnitudes against the magnitude of the corresponding Sl afferent volley produced by a series of stimulus intensities beginning at reflex threshold and proceeding in 10% increments. These recruitment curves exhibited a relatively linear early component (Fig. 2., stimulus intensities 1.O- 1.3 T) and an asymptotic top portion (i.e., saturation began at - 1.5 T). Further increase in stimulus intensity produced reflexes of smaller magnitude. This decrease was presumed by Magladery and McDougal ( 1950) to be due to collision of antidromic and orthodromic efferent impulses.

80

-II 0 > I-

RESULTS

:06 2

l

\

0.4 ::E 0.2

= 0.0 RSN

C-6

C-28

C-60

3. Summary of threshold, recruitment slope, and reflex magnitude for normal (RSN), contused 6day (C-6), contused 28day (C-28), and contused 60day (C-60) animals. (Significantly different from normal. *P I 0.05; **P I; 0.025.) analysis of variance (ANOVA), post hoc t test.) A : threshold. Comparison of group means for threshold initiation in normal and postcontusion animals. The threshold is defined as the minimal level of afferent volley required to elicit the test reflexes and is expressed in percent of the afferent magnitude. B: recruitment slope. Comparison of group means of slopes of the recruitment curves for the 4 groups. The slope of the recruitment curve was determined by a leastsquares linear regression of magnitude vs. percent of the maximal afferent volley (see text). The coefficients of linearity are included for each group. C: Maximal reflex ratio. Comparison of group means of maximal plantar H-reflex magnitude ( H,,), expressed as H,,/ maximal plantar M-reFIG.

span=

W,ax)*

postcontusion animals. By 28 and 60 days postcontusion, however, the slopes of the recruitment curves were 3.35 t 0.83( n = 10) and 2.66 t 0.20 (n = 8), respectively. Therefore although the slopes of the recruitment curves demonstrated some recovery toward normal, they remained significantly (P -C 0.05) decreased at the longest interval tested (Fig. 3B).

Slope The slope of the recruitment curve was observed to be 6.79 t 1.34 (n = 10) in normal animals, but had dropped significantly (P x 0.025) to 0.65 t 0.2 1 (n = 7) in the 6-day

Maximal

reflex ratios

In normal animals, maximal H-reflexes were produced using a stimulus repetition rate of 0.3 Hz. When expressed


SPINAL REFLEX TABLE

1.

EXCITABILITY

Frequency, Hz

SE

P 0.00 1 0.001 0.001 0.001 0.001

46.5

4.1

3.00

30.3 20.9

6.0 4.6

4.00 5.00

17.6 16.9

5.0 5.6

Rate sensitivity of the tibial H-reflex in normal animals. Values shown at each frequency are expressed as percentage of the magnitude obtained at the control repetition rate of 0.3 Hz. The P values shown were derived from a within group ANOVA, post hoc t test comparisons of respective means, relative to the reference obtained at 0.5 Hz (Marks 1982). N = 10. ANOVA, analysis of variance.

as a ratio of the M,,, (i.e., Hmax/Mmax), the mean value obtained for maximal H-reflexes in normal animals was 0.48 t 0.05 (n = 10). Although the mean plantar H,,,/ M max ratios obtained in postcontusion animals at 6 days (0.31 t 0.07, n = 7)) appeared decreased, and appeared increased at 28 days (0.53 t 0.08, n = 10) and 60 days (0.6 1 t 0.07, n = 8), respectively, none of these changes in reflex ratio after contusion injury were significantly different from normal. However, the mean of the 60 day postcontusion H,,/ M,,, reflex ratios were significantly (P 5 0.05) larger than the mean reflex ratios obtained in 6-day

I

ratios

Rate sensitivity of H-reflex magnitude NORMAL ANIMALS. In normal animals, the magnitude of the H-reflex was subject to attenuation by repetitive activation at frequencies greater than 0.3 Hz (Table 1; Fig. 4A ; see also Meinck 1976). Waveforms elicited at 0.3 Hz, and subsequently at 1 and 5 Hz, are superimposed in Fig. 4A to demonstrate this point. In this example, the reflex elicited at l-Hz repetition rate was attenuated to 56.3% of the magnitude seen at 0.3 Hz. As a group (n = lo), the magnitude recorded in normal animals at 1 Hz was 40.9 t 5.9 of the magnitude recorded at 0.3 Hz. The depression in reflex amplitude that occurred in response to increase in reflex repetition rate was even more accentuated at frequencies of 2-5 Hz ( Table 1; Fig. 5A ) . Accordingly, the reflex magnitudes recorded at l-5 Hz were significantly attenuated compared with the 0.5.Hz reference (P s 0.001, within-group repeated measures ANOVA, post hoc t test). For this withingroup analysis, 0.3 Hz was used as the control frequency, and 0.5 Hz was used as the reference frequency (Marks 1982). Rate depression has been reported to be influenced by the magnitude of the reflexes tested (Lloyd and Wilson 1957 ). We examined whether the patterns of rate depres-

c, , * , , , , , ,

A, , , , II

1477

INJURY

Hoc t Test

% Control

1.00 2.00

CONTUSION

postcontusion animals. The mean reflex magnitude of each group are compared in Fig. 3C.

Rate sensitivity: tibia1 H-reflex ANOVA/Post

AFTER

,

11

I

I

Mmax

II

-1.0

Hmox -

I

0.3

I

I

Hz

c

Hz

L

t 1.0

Hz

I1

1111111

II

I

II I

I

II

I

II

1

4. Rate sensitivity of tibia1 H-reflexes and tibia1 monosynaptic reflexes ( MSRs). A : plantar H-reflex waveforms from a normal animal elicited by stimulation of posterior tibia1 nerve ( 1.5 T) at 0.3, 1, and 5 Hz, respectively. B: 28day postcontusion, H-reflexes elicited as in A except from an animal 28 days subsequent to midthoracic contusion injury. C: a pair of tibia1 MSR waveforms from a normal animal elicited by stimulation of posterior tibia1 nerve ( 1.3 T) at 0.3, 1, and 5 Hz, respectively. D: 60 day postcontusion, tibia1 MSRs elicited as in C except from an animal 60 days subsequent to midthoracic contusion injury. Each waveform is the signal average of 16 consecutive traces. FIG.


1478

THOMPSON,


sion of maximal reflexes are representative of reflexes of submaximal amplitude by comparison of rate depression in maximal versus half H,, in 3 normal animals (Fig. 5A). We observed no significant difference in rate depression as a function of H-reflex amplitude. POSTCONTUSION. At 6 days after contusion injury, H-reflex magnitude, as a function of reflex repetition rate, was not significantly different from that recorded in normal animals. However, a significant change was seen by 28 days postcontusion in the rate sensitivity of H-reflex magnitude. Sample waveforms to illustrate this change are shown in Fig. 4 B. Note that the 1-Hz waveform shows minimal attenuation compared with the 0.3-Hz waveform. The mean reflex magnitudes elicited at 1 Hz in the 28day postcontusion animals were attenuated to only 9 1.O. t 9.8% (n = 10) of that seen at 0.3 Hz. These waveforms were in sharp contrast with the substantial attenuation produced in the l-Hz waveform in Fig. 4A recorded from a normal animal. A scatter plot of these data (Fig. 6A) from individual animals demonstrates that the range of responses included rate-sensitive potentiation, which replaced the normal pattern of rate-sensitive depression in 3 of 10 28day postcontusion animals. Comparison of group mean reflex magnitudes at l-5 Hz (Fig. 6B) revealed that reflex magnitudes in 28day postlesion animals were significantly larger (i.e., less atten-

A .

1.0

l A + V

B

NORMAL C-6 C-28 C-60

N=lO N=7 N=lO N=8

FREQUENCY (Hz)

4 mn ILU

NORMAL C-6 C-28 C-60

ho0 aL g

80

o

60

N=lO N=7 N=lO N=6

A ,100 0 z 80 z 0 60 0

Kv m

N=lO NORMAL (Hmax) NORMAL N=3 (l/2 Hmax)

1.0

2.0

3.0

4.0

5.0

FREQUENCY (Hz)

t

FIG. 6. Rate sensitivity of H-reflex. A : scatter plot comparison of individual experimental values of rate-sensitive H-reflex depression at l-5 Hz compared with the magnitude at the 0.3.Hz control in normal and 3 postcontusion groups. B: comparison ofgroup means for rate-sensitive attenuation at l-5 Hz compared with the magnitude at 0.3 Hz. The magnitude is expressed as percent of the magnitude at 0.3 Hz [*P s 0.05;**P 5 0.005, significantly larger than normal, analysis of variance (ANOVA), post hoc t test). Variability is expressed as mean t, SE.

v I 1:o

210 310 FREQUENCY

410

510

(Hz)

B Kq m

FREQUENCY

NORMAL (MSRmax) NORMAL (~32 l/2

N=15 N=3 max)

(Hz)

FIG. 5. Rate sensitivity of maximal vs. half maximal test reflexes. Comparison of rate sensitive of the tibial H-reflex (A ), and tibial monosynaptic reflex [ MSR (B)], at maximal and half maximal reflex magnitudes, respectively.

uated) at l-3 Hz than those recorded at comparable frequencies in normal animals (P 5 0.05). This loss of rate-sensitive depression of reflex magnitude continued to be marked when tested 60 days postinjury (Fig. 6 B). The mean reflex magnitudes of the 60.day postcontusion animals were significantly less attenuated at 1-5 Hz compared with those recorded at comparable frequencies in the normal animals (P 5 0.05 ). In the absence of the normal pattern of rate-sensitive depression, the relative reflex magnitude in both the 28. and the 60day postcontusion animals was increased compared with normal animals when elicited at test frequencies of 21 Hz.

Rate-sensitive depression of tibia1 MSR Because the H-reflex is recorded as a function of central and peripheral (neuromuscular junction) synapses, the tib-


SPINAL REFLEX

EXCITABILITY

ial MSR magnitude was tested to provide an additional measure of rate-sensitive depression of reflex excitability that would not be complicated by the involvement of peripheral synapses. In normal animals, the tibial MSR was significantly reduced in magnitude at test frequencies of l-5 Hz compared with the 0.5.Hz reference (P 5 0.001, within-group ANOVA; post hoc t test, Table 2). For this within-group analysis, 0.3 Hz was used as the control frequency, and 0.5 Hz was used as the reference frequency (Marks 1982). At l-Hz repetition rate, the MSR magnitude was decreased to 60.8 t 3.5% (n = 15) of the magnitude at 0.3.Hz repetition rate. Example waveforms are shown in Fig. 4C. The pattern of rate-sensitive depression was similar to that seen with H-reflexes (Fig. 4A), except that the relative magnitude of the depression was 15.20% greater in the H-reflex than at comparable frequencies in the MSR. In three normal animals, rate depression was compared in maximal versus half maximal MSRs (Fig. 5 B). There were no significant differences in rate depression as a function of reflex amplitude. POSTCONTUSION. When tested at 28 and 60 days postcontusion, the tibial MSR magnitudes were significantly less rate sensitive than were seen in normal animals at corresponding frequencies (P < 0.05 ). For example, MSR magnitude elicited at 1 Hz in the 60day postcontusion animals was reduced to only 84.8 t 2.5% (n = 6) of the respective magnitude at 0.3.Hz repetition rate compared to 60.8 t 3.5% in normal animals. Sample waveforms recorded from a 60.day postcontusion animal are shown in Fig. 40. Group means of the tibial MSR rate sensitivity for normal and postcontusion animals are compared in Fig. 7. Note that at l-5 Hz, the MSR reflex amplitudes are significantly larger (i.e. less attenuated) compared to the MSRs in the normal animals. These observations indicate that the tibial MSR elicited in the 28. and 60.day postcontusion animals revealed significantly less rate-sensitive depression compared to normal animals at test frequencies, l-5 Hz.

General description of lesion histology LESIONS. The spinal cords of animals killed 6 days after weight-drop injury were characterized by a devel-

AFTER CONTUSION

1479

INJURY

A l

inn IVV

$ e + Z

+ V

NORMAL C-28 C-60

N=15 N=S N=6

80

0

t)

60

I-

z 40 0 QI w 20 e

0

5:o

110 FREQUENCY

w

(Hz) -

2100 z IZ 0

o I$ 0

+-

NORMAL t**

80 --

N=15 N=5 N=6

60 -40--

E 20 -a. 0 FREQUENCY

(Hz)

FIG. 7. Rate sensitivity of tibial monosynaptic reflex ( MSR). A : scatter plot comparison of individual experimental values of rate-sensitive tibial MSR depression at l-5 Hz compared with the magnitude at the 0.39Hz control in normal and 2 postcontusion groups. B: comparison of normal, 28-&y, and 60day postcontusion group means for rate-sensitive depression at l-5 Hz. The magnitude at each frequency is expressed as percent of the 0.3-Hz amplitude. [*P r 0.05;***P r 0.00 1, significantly larger (i.e., less attenuated) than normal, analysis of variance (ANOVA), post hoc t test]. Variability is expressed as mean k SE.

CONTUSION

TABLE 2.

Rate sensitivity: tibia1 MSR ANOVAfPost

Hoc t Test

Frequency, Hz

% Control

SE

P

1.00 2.00 3.00 4.00

60.8 43.8 36.1 30.4 26.4

3.5 4.1 3.9 3.7 3.4

0.001

5.00

Q.001 0.001 0.001

0.001

Rate sensitivity of the tibial MSR in normal animals. Values shown at each frequency are expressed as percentage of the magnitude obtained at the control repetition rate of 0.3 Hz. The P valuesshown were derived from a within group ANOVA, post hoc t test comparisons of respective means, relative to the reference obtained at 0.5 Hz (Marks 1982). N = 15. MSR, monosynaptic reflex; ANOVA, analysis of variance.

oping central core of necrosis involving gray and white matter (Figs. 8A and 9, A and B). The lesion epicenters contained dense accumulations of macrophages and other inflammatory cells that were either seen in well-developed cysts or that delineated areas of cavitation. There was some variability in the degree and pattern of tissue erosion. Representative examples are shown in Fig. 9, A and B. In some cases, the spinal cords displayed a large symmetric zone of central necrosis. Some sparing of lateral and ventrolateral white matter was seen, but each of these sectors contained many profiles of degenerating axons as well (Blight and DeCriscito 1986; Bresnahan 1978). Other specimens revealed a more asymmetric lesion, with some apparently persisting axons in the dorsal columns and partial preservation of the superficial dorsal horn. Tissue specimens obtained at 28 or 60 days postcontusion injury revealed large cysts extending 3.75-4;6 mm in the rostrakaudal plane. Representative histological speci-


1480

THOMPSON,


FIG. 8. Transverse 2-pm sections are shown of plastic-embedded spinal cord specimens that were obtained at the level of the lesion epicenter. A: at 6 days post-injury, an evolving core of central necrosis is evidenced by small cysts (cy) and numerous macrophages, infiltrating inflammatory cells, and cellular debris. This region of tissue necrosis is surrounded by a rim of more intact-appearing white matter. Even at this low magnification, however, much of this rim of tissue exhibits vacuolation characteristic of spongiform degeneration of myelinated fibers. B: by 28 days, prominent cysts are seen that now contain fewer cells than seen earlier. Although a well-defined rim of tissue still remains, some vacuolation is still seen within the white matter. C: spinal cord at 60 days postinjury appeared similar to that seen in the previous micrograph. Again, some degeneration is still indicated in the remaining subpial tissue. D and E: these 2 micrographs are higher-magnification views of regions of spared tissue shown in A and B. In both examples, small intact-appearing axons are seen immediately subjacent to the pia. Those fibers located more internally show considerable degenerative changes as exhibited by swollen axons (E, b ) and distorted myelin sheaths (D, b ). A-C: X30, D-E: X 120.

mens for 28-day and 60-day postcontusion injuries are shown in Figs. 8, B and C, respectively. Cavitation was either centrally located or restricted to the dorsal aspect of the spinal cord. These zones of necrosis were almost com-

pletely in the myelin tissue.

surrounded by persisting white matter, although, as 6-day material, many focal regions of axonal and breakdown were seen in most areas of persisting Representative camera lucida silhouettes of the cys-


SPINAL

REFLEX

EXCITABILITY

@MB G

H

FIG. 9. Computer-assisted drawings of representative lesion epicenters seen in 2-pm sections of plastic-embedded tissue obtained at 8 (A and B), 28 (C-E), and 60 days (F-H) postinjury. As noted in the text, the areas of tissue necrosis and sparing take into account the fact that regions of apparent tissue preservation exhibited substantial, long-term axonal degeneration. Accordingly, the extent of tissue degeneration shown in black extends beyond the limits of the cystic cavities seen in these specimens. For example, Figs. 8, B and C, are tissue sections on which drawings in Fig. 9, E and F, are based, respectively.

tic necrosis for 28- and 60.day postcontusion shown in Fig. 9, C-E and F-H, respectively.

injuries are

DISCUSSION

The results of the present study indicate that significant changes in lumbar reflex excitability result from midthoracic contusion injury of the rat spinal cord. These changes include reduced reflex threshold, decreased slope of recruitment, and diminished rate-sensitive depression. Although recruitment slope was most altered at the shortest postinjury interval tested and exhibited some recovery, the other changes were progressive in onset and enduring. H-reflex In previous studies of motoneuron excitability after hemisection injury (Malmsten 1983), the reflex magnitudes caudal to the lesion were compared with those of the uninjured side. Using the intact side as a standard, this method permitted comparison across animals of left /right ratios instead of absolute values. Because contusion injury often involves bilateral damage, this approach was not feasible in the present study. Therefore a recording strategy had to be adopted that would provide sensitive measures of reflex ac-

AFTER

CONTUSION

INJURY

1481

tivity that could be compared across animals. The H-reflex (Magladery and McDougaI 1950) offered a useful method with these considerations in mind. It has been shown, for example, that the magnitude of the H-reflex is proportional to the number of motoneurons activated (Magladery and McDougal 1950) and that this reflex provides a quantitative index of neurotransmission from Ia afferents to target motoneurons (Burke et al. 1984; MagIadery et al 195 1; Wolpaw 1987). In addition, as noted earlier (see METHratio facilitates comparison of data ODS) 9 the wn,,/ wn, across animals. Although the H-reflex is extremely useful for measuring reflex amplitude, there is a potential caveat of frequency response data. The H-reflex EMG recordings may not exclusively reflect the frequency-following characteristics of central monosynaptic synapses, because possible changes in the frequency response of the neuromuscular junction could also influence the data. Therefore experiments using normal and spinally contused animals were incorporated in this study in which the rate-sensitive depression of tibiaI MSRs was also evaluated. As discussed below, agreement was observed between both of these recording paradigms. Reflex thresholds and recruitment curves The plantar H-reflexes were used in this study because they were observed to be more stable than those recorded from the triceps surae (also see Meinck 1976). This is consistent with the proportionally large fraction of tibial motoneurons that innervate the plantar muscles (Crockett et al. 1987). The plantar H-reflex appeared at 66.6% of the maximal afferent volley elicited by stimulation of the tibiaI nerve. This seems proportionately large compared with thresholds for cat triceps surae monosynaptic reflexes (Rall 1955; Hunt 1955). However, the adult rat monosynaptic reflexes have been compared with the less secure activation patterns in kittens ( Kaizawa and Takahashi 1970). These patterns in kittens have been attributed to an immature state of the intraspinal collaterals of primary afferents and correspondingly less secure functional connections between Ia afferents and motoneurons (Skoglund 1960). However, the relatively steep slope of the recruitment curve indicated that once the minimum reflex excitability level had been reached, additional units were readily recruited with minimal increases in the relative magnitude of the afferent volley. At 6 days postcontusion, although the threshold was unchanged, a substantial decrease in slope of the recruitment curve was obtained. This change in slope is interpreted as a decrease in the gain of reflex excitability (i.e., a diminished rate of growth of reflex magnitude per unit increase in afferent input). Collectively, these observations indicated that at 1 wk subsequent to midthoracic contusion injury, lumbar reflex excitability was significantly depressed. Although remaining significantly diminished compared with normal, these recruitment patterns demonstrated some recovery by 28 days and 60 days postinjury. Recordings in the 28.day postcontusion animals also revealed that the thresholds for reflex initiation were decreased and reached significant levels by 60 days postinjury. These


1482

THOMPSON,

REIER,

changes in reflex excitability could be attributed to many factors, including progressive pathology at the injury site as well as possible plasticity-related changes in the segmental circuits responsible for reflex initiation. The amount of excitatory signal required to initiate reflex discharge and the subsequent efficiency of recruitment gradation depend on the distribution of excitability among the motoneurons of the test pool and the amount and distribution of background synaptic bias (Hunt 1955; Kernel1 and Hultborn 1990; Rall 1955). A decrease in reflex threshold with a corresponding displacement of the recruitment curve along the abscissa was Qroduced in the recruitment of monosynaptic reflexes in the cat spinal cord (Hunt 1955) and in a numerical model of motoneuron pool recruitment (Kernel1 and Hultbom 1990) by increasing the background excitation using inputs that overlap the intrapool distribution of the excitatory drive; the shape of recruitment curve was dependent on the intrapool distribution of test excitatory and inhibitory inputs (Hunt 1955; Kernel1 and Hultborn 1990). We interpret the changes in excitability in the present study to be consistent with the view that initially spinal injury resulted in a marked alteration in the distribution of background excitation and inhibition, producing a substantial depression of reflex excitability. Then, during the first few postinjury weeks, progressive changes occurred that resulted in an increase in the reflex excitability of motoneurons. Noteworthy among the progressive changes that have been suggested to lead to an increase in reflex excitability are I ) an increase in the intrinsic excitability of the motoneurons and 2) enhanced synaptic input (which could also secondarily produce a net increase in background excitation ) . Within hours of thoracic spiMOTONEURON EXCITABILITY. nal transection in cats, intracellular recordings revealed 200% increases in Ia excitatory postsynaptic potentials (EPSPs) in medial gastrocnemius (fast) motoneurons (Nelson et al. 1979) or 5 100% increases in soleus (slow) motoneurons (Cope et al. 1988). A decrease in threshold and alteration of minimum firing rate of lumbar motoneurons was reported after acute dorsal hemisection of the thoracic spinal cord (Carp et al. 199 1; Powers and Rymer 1988). However, these changes do not appear to be enduring because intracellular recordings in animals with chronic spinal lesions do not demonstrate significantly enlarged Ia EPSPs (Munson et al. 1986; Nelson et al. 1979) or significant alterations in membrane electrical properties relative to motor unit type (Munson et al. 1986). Therefore an enduring alteration in the intrinsic excitability of motoneurons does not appear to be the most compelling explanation for persisting increases in the excitability of spinal reflexes subsequent to spinal cord injury. Experiments designed to correENHANCED SYNAPTIC INPUT. late temporal patterns in the recovery of hind limb motor function in cats with alterations in the distribution of labeled dorsal root axons and terminals in spinal gray matter have provided evidence for enhancement of segmental synaptic inputs via collateral sprouting (e.g., Murray and Goldberger 1974). CaudaI and ipsilateral to lower thoracic

LUCAS,

AND

PARMER

hemisections, all hind limb locomotor and reflex activity was considerably depressed during the acute postinjury period. Over several weeks, a progressive recovery of limb use was accompanied by the appearance of increased hind limb reflexes ( Murray and Goldberger 1974). Histological analysis revealed that dorsal root distributions and terminals in the spinal gray matter were increased on the side caudal to hemisection compared to the intact side. These authors proposed that recovery of locomotor function was facilitated by an adaptive increase in hind limb reflex activity produced in part by enhanced segmental synaptic input secondary to collateral sprouting of the primary afferent terminals. Later, progressive recovery of locomotor function was also suggested to be complimented by collateral sprouting of descending fibers (Goldberger and Murray 1988). Lumbar monosynaptic reflexes were also increased in amplitude 34 days subsequent to partial axotomy of the ascending axon branches by dorsal column transection at the L, or L, segmental level (Decima.and Morales 1983). This enhancement was correlated with an increased rate of rise of Ia EPSPs in lumbar motoneurons (Decima et al. 1986). In both lesion paradigms, (i.e., hemisection or partial axotomy ) the delayed appearance of the behavioral or physiological changes was proposed to be consistent with primary afferent plasticity. It is conceivable that the same principle applies to the delayed increase in reflex excitability subsequent to contusion lesion. Spinal cord injury: reflex magnitude

and reflex gain

Several investigators have reported increases in the magnitude of maximal monosynaptic reflexes subsequent to thoracic hemisection in rats (Malmsten 1983) and cats (Hultborn and Malmsten 1983a,b; McCouch et al. 1958; Teasdall et al. 1958). In the present study, however, despite the extensive tissue destruction precipitated by contusion injury, no clear indication of increased reflex magnitude ratios per se was exhibited in reflexes elicited at a control frequency of 0.3 Hz. The apparent contrast in motoneuron excitability after hemisection (Malmsten 1983) and contusion damage to the rat spinal cord could be attributed to several differences between hemisection and contusion lesions. However, the previous studies (e.g., Hultborn and Malmsten 1983a,b; Malmsten 1983) utilized a single repetition frequency of 1 Hz. As noted in RESULTS and discussed below, we found significant differences between normal and postcontusion animals in 1-Hz rate-sensitive depression of reflex magnitude. Therefore until rate-sensitive depression is specifically tested in hemisection lesioned animals, it will be uncertain if the MSRs observed caudal to the hemisected side were actually increased in magnitude or appeared increased because they were not as rate-sensitive as the intact control and therefore less attenuated. However, the present findings suggest that rate-sensitive depression is a critical consideration in spinal reflex studies, particularly after experimental injury. Rate sensitivity of H-reflex and MSRs Our observation that reflex magnitudes were largest at the slowest repetition rate (0.3 Hz) is consistent with the


SPINAL

REFLEX

EXCITABILITY

reduction in reflex amplitude related to long-acting inhibitory processes initiated by a preceding reflex event (e.g., rate-sensitive depression, Lloyd and Wilson 1957). The “control” 0.3.Hz stimulus repetition rate was selected for the present studies to permit an interval of sufficient duration between events for the inhibitory processes produced by the preceding reflex to decay to a relatively low level of influence (Kaizawa and Takahashi 1970; Lloyd and Wilson 1957). In normal rats, we observed that tibia1 H-reflex magnitudes were 56% smaller at 1.O Hz than at 0.3-Hz repetition rates, whereas in the 28.day contused animals they were only 10% smaller. We speculated that the greater reflex magnitudes seen in the postcontused animals at 1 Hz was due to a significant failure in the expression of rate-sensitive depression. To explore this possibility more fully, we compared the levels of reflex attenuation in normal and spinally contused animals using repetition frequencies of 1-5 Hz relative to the 0.3.Hz baseline in each group. The patterns of rate-sensitive depression of the plantar H-reflexes in our normal animals were consistent with those previously reported in other studies of the rat (Meinck 1976). However, the decline in reflex amplitude as a function of frequency was significantly less at 28 and 60 days postcontusion than in the normal animals. These observations were also paralleled by our tibia1 MSR data. In normal animals, the pattern of rate-sensitive depression was similar to the previous MSR data obtained in the rodent (Kaizawa and Takahashi 1970). Again, however, the degree of reflex attenuation was significantly reduced at frequencies of l-5 Hz when tested at 28 and 60 days postcontusion. The correspondence of the H-reflex and MSR data thus indicate that the altered patterns of rate-sensitive depression in spinally contused animals are indicative of alterations occurring in central mechanisms associated with rate-sensitive depression ( Lloyd 1957; Lloyd and Wilson 1957) rather than being a reflection of peripheral synaptic dysfunction. Rate-sensitive depression has been reported to be considerably greater in submaximal monosynaptic reflexes than at maximal amplitudes in normal cats (Lloyd and Wilson 1957)) but no difference in rate depression was found as a function of reflex amplitude of L, monosynaptic reflexes in normal rats (Kaizawa and Takahashi 1970). We tested this question by comparing rate depression of H,,, versus half H maxand tibia1 MSRs in three normal animals (Fig. 5, A and B). Our data are in agreement with the prior observations of Kaizawa on rat monosynaptic reflexes. We observed no significant differences in rate depression as a function of either H-reflex or tibia1 MSR amplitude. These observations suggest that rate depression can be produced by activity in the lowest-threshold afferent fibers.

Rate-sensitive depression: neuroanatomical/functional considerations Eccles and Rall ( 195 1) described a long period of reflex depression induced by its activation. They correlated the decreased reflex amplitude to an inhibitory depression of the presynaptic spike. Because the total time course affecting reflex ampli tude subsequen t to reflex repe tition lasts for

AFTER

CONTUSION

INJURY

1483

seconds, Lloyd and Wilson ( 1957 ) termed this attenuation “low-frequency depression.” Because reflex depression occurs without concomitant alterations in motoneuron excitability, rate-sensitive depression has also been ascribed to a long-acting presynaptic inhibition by several investigators (Curtis and Eccles 1960; Evanson 1956; Frank and Fuortes 1957; Kaizawa and Takahashi 1970; Lloyd 1957; Lloyd and Wilson 1957). Evidence continues to mount that supports the idea that presynaptic inhibition includes y-aminobutyric acid (GABA)-mediated changes in Ia presynaptic transmitter release. Recently, the GABA-b agonist L-baclofen was proposed to modify rate-sensitive changes in Ia synaptic transmission by reducing presynaptic Ca+* influx and the concomitant level of transmitter release from Ia-tierent terminals ( Lev-Tov et al. 1988). Reduction in rate-sensitive depression was found in experimental hind limb rigidity preparations in cats (Murayama and Smith 1965) and dogs (Gelfan 1966) produced by ischemia of the lumbosacral spinal cord. Cellular counts in the gray matter of the seventh lumbar segment in rigid preparations revealed survival of only 25% of the normal interneuron population (Gelfan 1966). Consequently, such deficits in the normal expression of rate-sensitive depression were attributed to intemeuronal cell loss (see also Eccles et al. 196 1; Jankowska et al. 198 1). Because thoracic spinal cord injury was involved in our present studies, a large-scale loss of lumbar interneurons is an unlikely basis for the loss of rate-sensitive depression. It is now recognized, however, that several descending pathways influence spinal presynaptic inhibitory interneurons and therefore play a role in modulating rate depression associated with presynaptic inhibition (Andersen et al. 1962; Carpenter et al. 1966; Fung and Barnes 1987; Jimenez et al. 1987; Lundberg and Vyklicky 1966; Martin et al. 1979; Proudfit et al. 1980; Rudomin et al. 1980, 198 1, 1983; Sastry and Sinclair 1977). Therefore the possibility exists that contusion injury resulted in the loss of descending axons that were involved in the control of reflex inhibition. Although the importance of presynaptic mechanisms has been emphasized, postsynaptic factors that regulate frequency transmission at the Iamotoneuron synapse (Koerber and Mendell 199 1) also warrant continued scrutiny. Future studies will focus on selective lesions of specific supraspinal pathways that are believed to modulate rate-sensitive depression. Such information may then be used to gain a greater understanding of the contusion injury model in rodents. Such further investigation may be germane to the loss of rate-sensitive depression in humans with spinal cord injuries (e.g., Delwaide 1973; Ishikawa et al. 1966). It is interesting that, compared with normal animals, rate-sensitive depression was only slightly reduced at 6 days postcontusion but was significantly reduced at 28 and 60 days postcontusion. Although a progressive loss of specific fiber populations was not apparent across the three postinjury intervals studied, the slow decline in rate-sensitive depression may reflect subtle secondary pathology leading to the degeneration of fibers or some other circuitry changes that were not expressed shortly after injury. It is also possible that this deficit in rate-sensitive depression may be related to the delayed expression of pharmacological changes in the lumbar cord subsequent to injury (e.g., Faden 1986).


1484

THOMPSON,


An alternative, though not mutually exclusive, consideration is that these physiological changes may be part of an adaptive recovery process, as mentioned earlier, secondary to injury of descending tracts (Burke 1988; Goldberger and Murray 1988; Murray and Goldberger 1974). A final consideration is the possibility that injury of the primary afferent ascending collaterals could have resulted in an alteration in the excitability of the segmental terminals ( Decima et al. 1986). In conclusion, these studies have demonstrated an intriguing change in the pattern of motoneuron excitability after contusion injury of the adult rat spinal cord. An expression of hyperreflexia was observed that wa .s not correlated sim ply to an increase in maximal H-reflex magnitudes per se but was correlated to the progressive decrease in reflex th .reshold and loss of rate-sensitive depression. These deficits progressed slowly to significant magnitudes and, being sustained at 2 mo, are interpreted to be enduring changes. Several earlier reports emphasized that repetition rate must be carefully considered in evaluating reflex excitability in normal animals (Eccles and Rall 195 1; Jefferson and Schlapp 1953; Lloyd and Wilson 1957). The present findings underscore these observations and emphasize the particular importance of rate-sensitive depression in the evaluation of reflex excitability subsequent to spinal cord injury. Experiments are now in progress to further investigate underlying neuroanatomic substratum and related neurophysiological and neuropharmacological mechanisms essential to the expression of rate-sensitive depression that could be relevant to future evaluation of various transplantation and other strategies directed at improving or sparing function in the damaged spinal cord (Thompson et al. 1990). The authors sincerely appreciate comments on the manuscript by Drs. Charles D. Barnes, Richard Johnson, William G. Luttge, John B. Munson, Louis A. Ritz, and Charles J. Vierck, the superb technical assistance of E. Grygotis and B. Zeller O’Steen, and the assistance of D. Winialski in the replication of the impact injury device. This work was support by National Institute of Neurological and Communicative Disorders and Stroke Grants NO 1-NS-7-2300 and 5-PO 1-NS275 11 and by the State of Florida Impaired Drivers and Speeders Trust Fund. Address for reprint requests: F. J. Thompson, Box J-244 JHMHC, Dept. of Neuroscience, College of Medicine, University of Florida, Gainesville, FL 32610. Received 6 May 199 1; 16 June 1992. REFERENCES

ALLEN,A. R. Surgery of experimental lesion of spinal cord equivalent to crush injury of fracture dislocation of spinal column. A preliminary report. JAMA 57: 878-880, 19 11. ANDERSEN, P., ECCLES, J. C., AND SEARS, T. A. Presynaptic inhibitory action of cerebral cortex on the spinal cord. Nature Lond. 194: 740-74 1, 1962. ANDERSON, D. K., BRAUGHLER, J. M., HALL, E. D., WATERS, T. R., MCCALL, J. M., AND MEANS, E. D. Effects of treatment with U-74006F on neurological outcome following experimental spinal cord injury. J. Neurosurg. ANDERSON,

69: 562-567,

1988.

D. K., REIER, P. J., WIRTH, E. D., THEELE, D. P., MARECI, T., BROWN, S. A., AND ZELLER, B. E. Delayed grafting of fetal CNS tissue into chronic compression lesions of the adult cat spinal cord. Restor. Neural. Neurosci. 2: 309-325, 199 1. BEA’ITIE, M. S., STOKES, B. T., AND BRESNAHAN, J. C. Pharmacological

approaches to treatment of brain and spinal cord injury. In: Experimental Spinal Cord Injury: Strategies Based on Anatomic and Behavioral

-for Acute and Chronic Intervention Studies. New York: Plenum, 1988,

p. 43-74. P., MARKOWITZ, R. S., COOPER, V., MECHANIC, A., KUSHNER, H., DAMAJANOV, I., FINKELSTEIN, S. D., AND WACHS, K. C. Models of spinal cord injury. 1. Static load technique. Neurosurgery 19: 752-762, 1986. BLIGHT, A. R. Axonal physiology of chronic spinal cord injury in the cat: intracellular recording in vitro. Neuroscience 10: 147 1- 1486, 1983. BLIGHT, A. R. Delayed demyelination and macrophage invasion: a candidate for “secondary” cell damage in spinal cord injury. Cent. Nerv. Syst. Trauma 2: 299-3 15, 1985. BLIGHT, A. R. AND DECRISCITO, V. Morphometric analysis of experimental spinal cord injury in the cat: the relation of injury intensity to survival of myelinated axons. Neuroscience 19: 32 l-34 1, 1986. BRACKEN, M. B., SHEPARD, M. J., COLLINS, W. F., JR.,HOLFORD, T. R., BASKIN, D. S., EISENBERG, H. M., FLAMM, E., LEO-SUMMERS, L., MAROON, J. C., MARSHALL, L. F., PEROT, P. L., JR.,PIEPMEIER, J., SONNTAG, V. K. H., WAGNER, F. C., JR.,WILBERGER, J. L., WIN-N,H. R., AND YOUNG, W. Methylprednisolone or naloxone treatment after acute spinal cord injury: one-year follow-up data. Results of the second National Acute Spinal Cord Injury Study. J. Neurosurg. 76: 23-3 1, 1992. BRESNAHAN, J. C. An electron-microscopic analysis of axonal alterations following blunt contusion of the spinal cord of the rhesus monkey (macaca mulatta). J. Neural. Sci. 37: 59-82, 1978. BRESNAHAN, J. C., BEATTIE, F. D., TODD, F. D., AND NOYES, D. H. A behavioral and anatomical analysis of spinal cord injury produced by a feedback controlled impaction device. Exp. Neural. 95: 548-570, 1987. BURKE, D. Spasticity as an adaptation to pyramidal tract injury. In: AdBLACK,

vances

in Neuro&y

Functional

Recovery

in Neurological

Disease,

edited by S. G. Waxman. New York: Raven, 1988, p. 401-423. BURKE, D., GANDEVIA, C., AND MC~ON, B. Monosynaptic and oligosynaptic contributions to human ankle jerk and H-reflex. J. Neurophysiol. 52: 435-448, 1984. CARP, J. S., POWERS, R. K., AND RYMER, W. Z. Alterations in motoneuron properties induced by acute dorsal spinal hemisection in the decerebrate cat. Exp. Brain Res. 83: 539-548, 199 1. CARPENTER, D., ENGBERG, I., AND LUNDBERG, A. Primary afferent depolarization evoked from the brain stem and the cerebellum. Arch. Ital. Biol. 104: 73-85, 1966. COPE, T. C., HICKMAN, K. R., AND BOTTERMAN, B. R. Acute effects of spinal transection on EPSPs produced by single homonymous Ia-fibers in soleus alpha-motoneurons. J. Neurophysiol. 60: 1678- 1694, 1988. CROCKETT, D. P., HARRIS, S. L., AND EGGAR, M. D. Plantar motoneuron columns in the rat. J. Comp. Neurol. 265: 109-l 18, 1987. CURTIS, D. R. AND ECCLES, J. C. Synaptic action during and after repetitive stimulation. J. Physiol. Lond. 150: 374-398, 1960. DAS, G. D. Perspectives in anatomy and pathology of paraplegia in experimental animals. Brain Res. Bull. 22: 7-22, 1989. DE LA TORRE, J. C. Spinal cord injury models. Prog. Neurobiol. 22: 289344, 1984. DECIMA, E. E., LORENZO, D., AND MALLACH, L. Effects of “partial axotomy” upon synaptic function. Exp. Brain Res. 64: 464-468, 1986. DECIMA, E. E. AND MORALES, F. R. Long-lasting facilitation of a monosynaptic pathway as a result of “partial” axotomy of its presynaptic elements. Exp. Neurol. 79: 532-55 1, 1983. DELWAIDE, P. J. Human monosynaptic reflexes and presynaptic inhibition: an interpretation of spastic hyperreflexia. In: New Developments in Electromyography and Clinical Neurophysiology, edited by J. E. Desmedt. Basel: Karger, 1973, p. 508-522. ECCLES, J. C., ECCLES, R. M., AND MAGNI, F. Central inhibitory action attributable to presynaptic depolarization produced by muscle afferent volleys. J. Physiol. Land. 159: 147-166, 196 1. ECCLES, J. C. AND RALL, W. Effects induced in a monosynaptic reflex by its activation. J. Neurophysiol. 14: 353-376, 195 1. EVANSON, J. M. The presynaptic localization of the depression following monosynaptic reflex activity (Abstract). J. Physiol. Land. 132: 6 1P, 1956. FADEN, A. I. Changes in TRH immunoreactivity in spinal cord after experimental spinal injury. Neuropeptides 7: 1 l- 18, 1986. FEHLINGS, M. G., TATOR, C. H., LINDEN, R. D., AND PIPER, I. R. Motor and somatosensory evoked potentials recorded from the rat. Electroencephalogr. Clin. Neurophysiol. 69: 65-78, 1988.


SPINAL REFLEX

EXCITABILITY

FRANK, K. AND FUORTES, M. G. F. Presynaptic and postsynaptic inhibition of monosynaptic reflexes. Fed. Proc. 16: 39-40, 1957. FUNG, S. J. AND BARNES, C. D. Presynaptic facilitatory action of locus coeruleus stimulation upon hindlimb sensory impulse transmission in decerebrate cats. Arch. Ital. Biol. 125: 187-200, 1987. GELFAN, S. Altered spinal motoneurons in dogs with experimental hindlimb rigidity. J. Neurophysiol. 29: 583-6 1 1, 1966. GOLDBERGER, M. E. Methods for assessing recovery of hindlimb movements after SC1 in cats. In: Conference Proceedings: Criteria for Assessing Recovery ofFunction: Behavioral Methods, edited by M. Brown and M. E. Goldberger. Springfield, NJ: APA, 1989, p. 56-6 1. GOLDBERGER, M. E. AND MURRAY, M. Patterns of sprouting and implications for recovery of function. In: Advances in Neurology: Functional Recovery in Neurological Disease, edited by S. G. Waxman. New York: Raven, 1988, vol. 47, p. 36 l-385. GUTH, L., BREWER, C. R., COLLINS, W. F., GOLDBERGER, M. E., AND PERL, E. R. Criteria for evaluating spinal cord regeneration experiments. Exp. Neural. 69: 1-3, 1980. HUGHES, J. T. Pathologic Changes AJer Spinal Cord Injury. Oxford, UK: Oxford Univ. Press, 1988, p. 34. HULTBORN, H. AND MALMSTEN, J. Changes in segmental reflexes following chronic spinal cord hemisection in the cat. I. Increased monosynaptic and polysynaptic ventral root discharges. Acta Physiol. Stand. 1 19: 405-422, 1983a. HULTBORN, H. AND MALMSTEN, J. Changes in segmental reflexes following chronic spinal cord hemisection in the cat. II. Conditioned monosynaptic reflexes. Acta Physiol. Stand. 1 19: 423-433, 1983b. HUNT, C. C. Monosynaptic reflex response of spinal motoneurons to graded afferent stimulation. J. Gen. Physiol. 38: 8 13, 1955. ISHIKAWA, K., OTT, K., PORTER, R. W., AND STUART, D. Low frequency depression of the H wave in normal and spinal man (Abstract). Exp. Neurol. 15: 140- 156, 1966. JANKOWSKA, E., MCCREA, D., RUDOMIN, P., AND SYKOVA, E. Observations on neuronal pathways subserving primary afferent depolarization. J. Neurophysiol. 46: - 19203, 198 1. JEFFERSON, A. A. AND SCHLAPP, W. Some effects of repetitive stimulation of afferents on reflex conduction. In: The Spinal Cord. London: Churchill, 1953, p. 99-l 19. JIMENEZ, I., RUDOMIN, P., AND SOLODKIN, M. Mechanisms involved in the depolarization of cutaneous afferents produced by segmental and descending inputs in the cat spinal cord. Exp. Brain Res. 69: 195-207, 1987. KAIZAWA, J. AND TAKAHASHI, I. The low frequency depression of rat monosynaptic reflex. Tohoku J Exp. Med. 100: 383-394, 1970. KAKULAS, B. A. Pathomorphological evidence for residual spinal cord functions. In: Recent Achievements in Restorative Neurology: Upper Motor Neuron Functions and Dysfllnctions, edited by J. C. Eccles and M. R. Dimitrijevic. Basel: Karger, 1985, vol. 1. KERASIDIS, H., WRATHALL, J. R., AND GALE, K. Behavioral assessment of functional deficit in rats with contusive spinal cord injury. J. Neurosci. Methods 20: 167- 189, 1987. KERNELL, D. AND HULTBORN, H. Synaptic effects on recruitment gain: a mechanism of importance for the input-output of motoneurone pools? Brain Res. 507: 176-179, 1990. KOERBER, H. R. AND MENDELL, L. M. Modulation of synaptic transmission at Ia-afferent fiber connections on motoneurons during high frequency stimulation: role of postsynaptic target. J. Neurophysiol. 65: 590-597, 1991. LANDAU, W. M. Spasticity the fable of a neurological demon and the emperor’s new therapy. Arch. Neural. 3 1: 2 17-2 19, 1974. LEV-TOV, A., MEYERS, D. E., AND BURKE, R. E. Activation of type B gamma-aminobutyric acid receptors in the intact mammalian spinal cord mimics the effects of reduced presynaptic Ca2+ influx. Proc. Natl. Acad. Sci. USA 85: 5330-5334, 1988. LLOYD, D. P. C. Monosynaptic reflex response of individual motoneurons as a function of frequency. J. Gen. Physiol. 40: 435-450, 1957. LLOYD, D. P. C. AND WILSON, V. J. Reflex depression in rhythmically active monosynaptic reflex pathways. J. Gen. Physiol. 40: 409-426. 1957. LODGE, D. AND ANIS, N. A. Effects of ketamine and three other anesthetics on spinal reflexes and inhibitions in the cat. Br. J. Anaesth. 56: 1 1431151, 1984. LUNDBERG, A. AND VYKLICKY, L. Inhibition of transmission to primary afferents by electrical stimulation of the brain stem. Arch. Ital. Biol. 104: 86-97, 1966.

AFTER CONTUSION

INJURY

1485

J. W. AND MCDOUGAL, D. B. ElectrophysiologicaI studies of nerve and reflex activity in man. I. Identification of certain reflexes in the electromyogram and the conduction velocity of peripheral nerve fibers. Bull. Johns Hopkins Hosp. 86: 265-290, 1950. MAGLADERY, J. W., PORTER, W. E., PARK, A. M., AND TEASDALL, R. D. Electrophysiological studies of nerve and reflex activity in normal man, two-neurone reflex and identification of certain action potentials from spinal roots and cord. Bull. Johns Hopkins Hosp. 88: 499-5 19, 195 1. MALMSTEN, J. Time course of segmental reflex changes after chronic spinal cord hemisection in the rat. Acta Physiol. Stand. 119: 435-443, 1983. MARKS, R. G. Analyzing Research Data, the Basis ofBiomedical Research Methodology. New York: Van Nostrand Reinhold, 1982, p. 90. MARTIN, R. F., HABER, L. H., AND WILLIS, W. D. Primary afferent depolarization of identified cutaneous fibers following stimulation in medial brain stem. J. Neurophysiol. 42: 779-790, 1979. MCCOUCH, G. P., AUSTIN, G. M., AND LIU, C. Y. Sprouting as a cause of spasticity. J. Neurophysiol. 2 1: 205-2 16, 1958. MEINCK, H. M. Occurrence of the H-Reflex and the F-wave in the rat. Electroencephalogr. and Clin. Neurophysiol. 4 1: 530-533, 1976. MUNSON, J. B., FOEHRING, R. C., LOFTON, S. A., ZENGEL, J. E., AND SYPERT, G. W. Plasticity of medial gastrocnemius motor units following cordotomy in the cat. J. Neurophysiol. 55: 6 19-634, 1986. MURAYAMA, S. AND SMITH, C. M. Rigidity of hind limbs of cats produced by occlusion of spinal cord blood supply. Neurology 15: 565-577, 1965. MURRAY, M. AND GOLDBERGER, M. E. Restitution of function and collateral sprouting in the cat spinal cord: the partially hemisected animal. J. Comp. Neural. 158: 19-36, 1974. NELSON, S. G., COLLATOS, T. C., NIECHAJ, A., AND MENDELL, L. M. Immediate increase in Ia-motoneuron synaptic transmission caudal to spinal cord transection. J. Neurophysiol. 42: 655-664, 1979. NOBLE, L. T. AND WRATHALL, J. R. Correlative analysis of lesion develop ment and functional status after graded spinal cord contusion injuries in the rat. Exp. Neural. 103: 34-40, 1989. POWERS, R. K. AND RYMER, W. Z. Effects of acute dorsal spinal hemisection on motoneuron discharge in the medial gastrocnemius of the decerebrate cat. J. Neurophysiol. 59: 1540-l 556, 1988. PROUDFIT, H. K., LARSON, A. A., AND ANDERSON, E. G. The role of GABA and serotonin in the mediation of raphe-evoked spinal cord dorsal root potentials. Brain Res. 195: 149-165, 1980. RALL, W. Experimental monosynaptic input-output relations in the mammalian spinal cord. J. Cell. Comp. Physiol. 46: 413-438, 1955. REIER, P. J., ENG, L. F., AND JAKEMAN, L. B. Reactive astrocyte and axonal outgrowth in the injured CNS: is gliosis really an impediment to regeneration? In: Neural Regeneration and Transplantation, edited by F. J. Seil. New York: Liss, 1989, p. 183-209. REIER, P. J., HOULE, J. D., JAKEMAN, L. B., WINIAUKI, D., AND TESSLER, A. Transplantation of fetal spinal cord tissue into acute and chronic hemisection and contusion lesions of the adult rat spinal cord. Prog. Brain Res., 78: 173-l 79, 1988. REIER, P. J., STOKES, B. T., THOMPSON, F. J., AND ANDERSON, D. K. Fetal cell grafts into resection and contusion/compression injuries of the adult rat and cat spinal cord. Exp. Neural. 115: 177-l 88, 1992. RUDOMIN, P., ENGBERG, E., JANKOWSKA, E., AND JIMENEZ, I. Evidence of two different mechanisms involved in the generation of presynaptic depolarization of the afferent and rubrospinal fibers in the cat spinal cord. Brain Res. 189: 256-26 1, 1980. RUDOMIN, P., ENGBERG, I., AND JIMENEZ, I. Mechanisms involved in presynaptic depolarization of Group I and rubrospinal fibers in cat spinal cord. J. Neurophysiol. 46: 532-548, 198 1. RUDOMIN, P., JIMENEZ, I., SOLODKIN, M., AND DUENAS, S. Sites of action of segmental and descending control of transmission on pathways mediating control of transmission on pathways mediating PAD of Ia and b afferent fibers in cat spinal cord. J. Neurophysiol. 50: 743-769, 1983. SASTRY, B. S. R. AND SINCLAIR, J. G. Tonic inhibitory influences of a supraspinal monoaminergic system on presynaptic inhibition of an extensor monosynaptic reflex. Brain Res. 124: 109-120, 1977. SHIAU, J. S., ZAPPULLA, R. A., AND NIEVES, J. The effect ofgraded spinal cord injury on the extrapyramidal and pyramidal motor evoked potentials of the rat. Neurosurgery 30: 76-84, 1992. SIMPSON, R. K. AND BASKIN, D. S. Corticomotor evoked potentials in acute and chronic blunt spinal cord injury in the rat: correlation with neurological outcome and histological damage. Neurosurgery 20: 13 l137, 1987. MAGLADERY,


1486

THOMPSON,


S. The reactions to tetanic stimulation of the two-neuron arc in the kitten. Acta Physiol. Stand. 50: 238-253, 1960. STOKES, B. T. AND REIER, P. J. Fetal grafts alter chronic behavioral outcome after contusion damage to the adult rat spinal cord. Exp. Neural. 116: 1-12, 1992. TABORIKOVA, H. Fraction of the motoneuron pool activated in the monosynaptic H-reflexes in man. Nature Lond. 209: 206-207, 1966. TANG, A. H. AND SCHROEDER, L. A. Spinal-cord depressant effects of k&mine and etoxadrol in the cat and the rat. Anesthesiology 39: 37-43, 1973. TEASDALL, R. D., MAGLADERY, J. W., AND RAMEY, E. H. Changes in reflex patterns following spinal cord hemisection in cats. Bull. Johns Hopkins Hosp. 103: 223-235, 1958. THOMPSON, F. J., REIER, P. J., SCHFUMSHER, G. W., JAKEMAN, L. B. W., WINIALSIU, D., LUCAS, C. C., AND RAY, L. R. Early changes in the control of primary afferent excitability following contusion injury of the rodent spinal cord. Sot. Neurosci. Abstr. 15: 124 1, 1989. SKOGLUND,

F. J., REIER, P. J., SCHRIMSHER, G. W., LUCAS, C. C., AND R. Excitability studies in the rodent spinal cord following contusion injury and neural tissue transplantation. Sot. Neurosci. Abstr. 16: 38, 1990. WINIALSKI, D., HOULE, J. D., JAKEMAN, L. B., AND REIER, P. J. Transplantation of fetal rat spinal cord into longstanding contusion injuries of adult rat spinal cord. Sot. Neurosci. Abstr. 13: 75 1, 1987. WOLPAW, J. R. Operant conditioning of primate spinal reflexes: the H-reflex. J. Neurophysiol. 57: 443-459, 1987. WRATHALL, J. R., PE~EGREW, R. K., AND HARVEY, F. Spinal cord contusion in the rat: production of graded, reproducible injury groups. Exp. Neural. 88: 108-122, 1985. YOUNG, W. Recovery mechanisms in spinal cord injury implication for regenerative therapy. In: Neural Regeneration and Transplantation. New York: Liss, 1989, p. 157-l 69. ZILELI, M. AND SCHRAMM, J. Motor versus somatosensory evoked potentials after acute experimental spinal cord injury in rats. Acta Neurochir. 108: 140-147, 1991. THOMPSON, RAY, L.


The experimental contusion injury of the spinal cord in sheep.

Dopamine is produced in the rat spinal cord and regulates micturition reflex after spinal cord injury.

(18)F-FDG-PET imaging of rat spinal cord demonstrates altered glucose uptake acutely after contusion injury.

A monitored contusion model of spinal cord injury in the rat.

Trigemino-cervical reflex in spinal cord injury.

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

Thoracic rat spinal cord contusion injury induces remote spinal gliogenesis but not neurogenesis or gliogenesis in the brain.

Central necrosis following contusion to the sheep's spinal cord.

Fetal grafts alter chronic behavioral outcome after contusion damage to the adult rat spinal cord.

Urodynamic patterns after traumatic spinal cord injury.

Three-dimensional computer-assisted analysis of graded contusion lesions in the spinal cord of the rat.

Regulatory effects of intermittent noxious stimulation on spinal cord injury-sensitive microRNAs and their presumptive targets following spinal cord contusion.

Glutamate antagonists applied to midbody spinal cord segments reduce the excitability of the fictive rostral scratch reflex in the turtle.

Priming Neural Circuits to Modulate Spinal Reflex Excitability.

Elevation and clearance of extracellular K+ following contusion of the rat spinal cord.

Influence of spinal cord lesion level and severity on H-reflex excitability and recovery curve.

Interleukin-33 treatment reduces secondary injury and improves functional recovery after contusion spinal cord injury.

compression injuries of the rat and cat spinal cord.

Limited Functional Effects of Subacute Syngeneic Bone Marrow Stromal Cell Transplantation After Rat Spinal Cord Contusion Injury.

Immediate Effects of Therapeutic Ultrasound on Quadriceps Spinal Reflex Excitability in Patients With Knee Injury.

Bilateral contusion-compression model of incomplete traumatic cervical spinal cord injury.

Spatiotemporal patterns and essential role of MSK1 expression after rat spinal cord injury.

Differential Histopathological and Behavioral Outcomes Eight Weeks after Rat Spinal Cord Injury by Contusion, Dislocation, and Distraction Mechanisms.

Posture-dependent changes in corticomotor excitability of the biceps after spinal cord injury and tendon transfer.