Inhibitory effects from various types of dorsal column and raphe magnus stimulations on nociceptive withdrawal flexion reflexes

Brain Research 846 Ž1999. 72–86 www.elsevier.comrlocaterbres

Research report

Inhibitory effects from various types of dorsal column and raphe magnus stimulations on nociceptive withdrawal flexion reflexes Nayef E. Saade´ a

a,d,)

, Samir F. Atweh b , Alain Privat c , Suhayl J. Jabbur

d

Department of Human Morphology, Faculty of Medicine, American UniÕersity of Beirut, Beirut, Lebanon Department of Internal Medicine, Faculty of Medicine, American UniÕersity of Beirut, Beirut, Lebanon c U.336 INSERM, Montpellier, France d Department of Physiology, Faculty of Medicine, American UniÕersity of Beirut, Beirut, Lebanon

b

Accepted 17 August 1999

Abstract Most of the clinical and research reports agree about the analgesic effects of dorsal column ŽDC. stimulation, but there is no unanimity about the neural mechanisms involved in this stimulation. The aim of the present study was to compare the effects of segmental and rostral activation of the DCs and to investigate whether these effects are mediated through a brainstem spinal loop. Decerebrate–decerebellate cats were subjected to selective DC lesions at C 1 and C 3 spinal cervical levels and their reflex reactions to natural or electrical nociceptive stimuli were monitored either as withdrawal flexion reflexes or as motorneuronal discharges. Conditioning stimulation was performed as train of shocks Ž100 Hz, for 1 to 10 min or 300 Hz for 30 ms. applied on the DCs either rostral ŽDCr. or caudal ŽDCc. to the spinal lesions or on the raphe magnus ŽRM.. Conditioning trains for 5–10 min applied on DCr inhibited the withdrawal flexion reflexes recorded as toe flexion Ž90% of the control.. Comparisons of the effects of DCr, DCc or RM of conditioning stimuli were made on the discharges of 110 motorneurons recorded in isolated ventral root fibers. Conditioning stimulation applied to DCc produced short lived inhibition Žin about 60%. or facilitation Žin about 30% of the neurons. while DCr or RM conditioning produced inhibition in 90% of neurons which outlasted the duration of the conditioning trains. It was also shown that repetitive application of conditioning train on either DCr or RM resulted in longer duration of inhibition than that observed following DCc conditioning. We conclude that the stronger inhibition of motorneuronal discharges, evoked by nociceptive stimuli, is obtained by rostral activation of the DCs and that long term effects of DCst are mediated through a DC–brainstem–spinal loop. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Dorsal column; Pain modulation; Electroanalgesia; Raphe; Brainstem; Flexion reflexes

1. Introduction About three decades after the first application of the dorsal column ŽDC. electroanalgesia w23,43,59,62x there is still controversy about the efficacy of this technique and the neural mechanisms activated by such stimulation. The rationale for DC stimulation ŽDCst. emanated directly from Melzack and Wall’s w40x ‘‘gate control theory’’ which predicted that stimulation of the large primary afferent fibers will interfere with pain and from the analgesia in man produced by stimulating large fibers in peripheral ) Corresponding author. Department of Human Morphology, Faculty of Medicine, American University of Beirut, P.O. Box: 11-0236r41, Beirut, Lebanon. Fax: q961-1-744-464; e-mail: [email protected]

nerves or dorsal roots w65x. Other reports, however, suggested that the effects of electrical stimulation of peripheral nerves could also be attributed to a ‘‘frequency related conduction block’’ and that electrical stimulation of the anterior columns of the spinal cord was more effective in relieving pain than DCst w1,9,22,31x. In the clinical situation, DCst or dorsal spinal cord stimulation has been applied as a continuous train for minutes or hours in order to produce pain inhibition Žfor review, see Refs. w18,30,41x.. Several studies have shown that pain relief is obtained in those dermatomal areas where the patient feels parasthesia during the stimulation w32,34,44x. The modified somatic sensations Žincluding pain threshold. return immediately to normal level at the end of stimulation, but chronic pain is relieved for much

0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 9 9 . 0 2 0 0 3 - X

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

longer periods w32,34,37x. Deep and visceral reflexes are also modified during DCst w32,38x and this modification constituted an important indicator for the use of this technique in the treatment of spasticity w17,61,64x, of certain motor syndromes observed in multiple sclerosis w25x and of visceral pains and autonomic syndromes w12,36,39,63x. Garcia-Larrea et al. w16x have reported a depression of nociceptive flexion reflexes ŽRIII responses. in 52% of patients during DCst which was rapidly reversible Žafter interruption of the stimulation. but ‘‘significantly associated with subjective pain relief’’ and recommended the use of the depression of these reflexes as a major indication for the success of chronic pain relief. Despite the better controlled experiments on animals, uncertainties continued in interpreting the physiological effects of and pathways involved in DCst. A review of the literature reveals a wide range of possible actions at spinal and supraspinal levels following DCst. At the supraspinal level, DCst showed inhibitory effects on events produced by activation of the other ascending tracts in the thalamic w2,32,45x and cortical somatosensory centers w27,29,44,49x. At the spinal level, segmental DCst showed inhibitory effects upon different types of dorsal horn neurons driven by either nociceptive or non-nociceptive stimuli w6,15,20, 35,47,53x. Studies reported failure of DCst applied rostral to the spinal lesion to affect the discharge of spinothalamic w15x, or bulboreticular w60x neurons while other studies w6,47,53,54x reported inhibition of dorsal horn neurons through a DC–brainstem–spinal loop. Also, growing anatomical and physiological evidence has shown a complex organization of and projections from the DC nuclear system Žfor review, see Refs. w3,29,52x.. This led us to a careful reexamination of the effects of DCst in controlled animal experiments. For this purpose, comparisons were made between the effects of DCst rostral and caudal to selective DC lesions on spinal nociceptive mechanisms in decerebrate–decerebellate cats, to rule out possible involvement of cerebral and cerebellar neural loops. These effects were then compared to those of stimulation of raphe magnus ŽRM., well known as a major component of pain modulating centers Žfor reviews see Ref. w5,14,69x.. Spinal nociceptive reflexes have been often used in the study of nociceptive and analgesic mechanisms w8,16,46,68x. In this study, we investigated the effects of either intermittent or continuous conditioning stimulation on the late reflex discharges evoked by nociceptive stimuli. Part of the results have appeared in a short communication w50x. Results of this study showed that continuous conditioning DCst, for a few minutes, applied rostral to selective DC lesions could depress flexion reflexes evoked by nociceptive heat stimulation for a period of 10 to 20 min. Using shorter conditioning stimuli, the effects of rostral and segmental DC activation were also compared to each other and to those of RM stimulation on the C fiber reflex discharge recorded in single alpha motorneuron axons.

73

2. Materials and methods 2.1. Surgical procedures Twenty male adult cats Ž3–5 kg. were initially anesthetized with either ether Ž n s 10. or Nembutal Ž35 mgrkg; n s 6. or a-chloralose Ž70 mgrkg; n s 4. and then decerebrated at infracollicular Ž n s 15. or midcollicular Ž n s 5. levels. Anesthesia was interrupted after decerebration and the animals were immobilized with Flaxedil Žexcept for five cats. and artificially ventilated to maintain the end-tidal CO 2 between 3.5–4.5%. Arterial blood pressure was maintained above 80 mmHg by intermittent i.v. infusion of 5% dextrose. During the experiments, rectal temperature and the temperature of the exposed parts of nervous tissue were thermostatically maintained at 378C. Midcollicular decerebration was performed after partial craniectomy by sectioning parts of the parietal and occipital cortices and the rostral midbrain. Infracollicular decerebration was performed after a large suboccipital craniectomy and suction of the cerebellum exposing the brainstem. With a fine curved forceps, the area at the base of the inferior colliculi was squeezed and progressively pinched off until a clear separation was apparent between the upper pons and the midbrain. The upper cervical spinal cord was exposed between the foramen magnum and the C4 segment by a dorsal laminectomy; the dorsal and ventral roots C 1 to C 3 were cut, and using a surgical microscope, the gracile and cuneate tracts were cut at C1 and C3 levels by pinching off, progressively deeper, with fine forceps until the central canal was reached ŽFig. 1.. Completeness of the DC cuts was ascertained by the failure to record, at the level of the DC nuclei, any afferent volley produced by supramaximal electrical stimulation applied either to the DCs caudal to the cuts or to the skin or nerves in any of the legs. In three cats, the DCs were cut at the level of C 3 –C 4 and dissected free till C 1 and mounted on a pair of platinum electrodes like a peripheral nerve. This procedure was not used routinely because the shortened survival of the DC fibers limited its use to 2 or 3 h after DC dissection. The lumbosacral spinal cord was exposed by a dorsal laminectomy and a longitudinal incision of the dura, and the ventral roots L 6 to S 1 were sectioned bilaterally. This operation precluded any possible antidromic activation of motorneurons of the corresponding spinal segments and allowed dissection of small bundles of fibers from L 7 or S 1 ventral roots for recording of a reflex discharge. A posterior medial longitudinal incision of the skin of a hind limb allowed access to the nerves of the leg. The sural and superficial peroneal nerves were dissected free for a distance of 2–3 cm, without cutting the distal ends, mounted on two pairs of platinum stimulating electrodes and covered with paraffin oil. Other nerves of the leg Žgastrocnemius, tibial and external popliteal. were ligated

74

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

Fig. 1. Microphotographs of the cervical spinal cord segments ŽC 1 , C 3 . showing the extents of the dorsal column lesions in four different cats ŽCa 1 –Ca 4 . and the surgical isolation of the dorsal columns ŽDC strand. in another two cats ŽCa 5 and Ca 6 ..

and cut. All the exposed nerves and central nervous structures were covered with warmed paraffin oil which was maintained at a temperature of 378C. 2.2. Stimulation and recording procedures 2.2.1. Test-stimulation Supramaximal electrical testing ŽT. stimuli were applied alternately to the sural and superficial peroneal nerves to elicit a C fiber afferent volley. This consisted of a single shock Ž0.2–0.5 ms duration. or a train of such shocks Ž100 Hz, 30 ms.. Single shock supramaximal stimulation was usually not sufficient to elicit late discharge of a motorneuron associated with C fiber afferent volley. Therefore, train stimulation was frequently used to elicit the late discharge of motorneurons as previously described w10,26,46x. When a single axonal discharge was isolated, the rate of stimulation was slowed to 1 stimulus every 10–20 s and the sensory modality and appropriate receptive field were determined using forceps of different sizes Žfor pressure and pinch., joint movement or radiant heat stimulation. 2.2.2. Recording Single axonal discharge of an alpha motorneuron was recorded in a small bundle of fibers isolated from L 7 or S 1 ventral roots and placed on a bipolar platinum electrode. Isolation of a small filament of ventral root containing one single active axon was made by progressive splitting or teasing of a rootlet using an operating microscope and fine forceps. A well separated axonal discharge produced by

peripheral stimulation can maintain a stable amplitude and shape for tens of minutes up to more than 1 h. Motorneuron discharges were amplified and either displayed on the screen of an oscilloscope for photography of samples or fed to a window discriminator and then averaged and stored by an STA I Neurograph system. Criteria defined by Chung et al. w10x and Woolf and Sweet w70x were used to discriminate between discharges of alpha and gamma motorneurons, and this study reports on the discharges of alpha motorneurons Žcharacterized by absent or low level of spontaneous activation and a well differentiated action potential from the background activation due mainly to g motorneurons..

2.2.3. Conditioning stimulation Conditioning DCst was applied to a position rostral to the C 1 DC lesions ŽDCr. through a bipolar electrode ŽDKI 85NE-200, 1 mm tip separation. placed between the midline and the ipsilateral gracile nucleus Žin reference to the stimulated peripheral hindleg nerves.; similar bipolar electrodes were used to stimulate the DCs caudal to ŽDCc. or between the two spinal cuts ŽDCb., and a stimulating coaxial electrode ŽDKI SNE-100, contact diameter 0.1 mm inner, 0.25 mm outer. was stereotaxically placed ŽP, 7–8; L, 0; V, 8., in reference to Berman’s atlas w4x, in RM. Two experimental protocols were used for conditioning stimulation: intermittent and continuous trains. The intermittent conditioning consisted of a train of pulses Ž300 Hz, 30 ms–0.15 ms duration of each shock. delivered at conditioning–testing ŽC–T. intervals ranging between

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

y500 and q1500 ms. Intervals were assigned a positive sign when the C shock preceded the T shock and a negative sign for the reverse order. The continuous conditioning consisted of a long Ž1–10 min. train Ž100 Hz. of stimulations. In most experiments, comparisons were made between the effects of conditioning stimulations applied, separately, to DCr, DCb, DCc and RM. In these comparisons, the same characteristics of train shocks and the same amount of current Ž0.05 to 0.1 mA in the individual shocks. were used at all sites of conditioning stimulations. In all experiments, data recording started 3 to 4 h after decerebration. This time allowed unparalyzed cats to elicit motor reaction in response to nociceptive stimulations and to show spontaneous respiration. The state of the preparation was evaluated by the efficacy of the conditioning from RM; in preparations where RM conditioning did not elicit an inhibitory effect on spinal reflexes, DCr conditioning was also ineffective and the preparation was discarded. This lack of effect could be attributed to a damage to the remaining brainstem. 2.2.4. Special procedures In one set of experiments Ž n s 5., recording of the flexion reflex of the toes was used. This was performed on unparalyzed cats 3 to 4 h after decerebration and decerebellation and after recovery of spontaneous respiration. The head, hip, knee and ankle of the animal were rigidly fixed in a stereotaxic frame. Laminectomy was performed at the upper cervical level for the DC lesions at C 1 and C 3 and for the application of the DC conditioning electrodes rostral to the spinal lesions. The lumbosacral vertebral column was not opened. A radiant heat spot was oriented to the center of the footpad in order to raise its temperature to 558C in each stimulation cycle for 3–7 s. One or two toes were ligated with a thread and connected to a transducer and the flexion reflex was recorded with a Physiograph. At the end of each experiment, the area of each flexion reflex response was measured by computerized planimetry and the results were evaluated by comparing the areas obtained during and after the DCst to the stable controls recorded before conditioning. 2.2.5. Data-analysis High intensity electrical stimulation generating C fiber afferent volleys produced an early and a late train of reflex discharges. The late discharge related to the C fiber activation and consequently to pain w10x started about 100–150 ms and ended about 500–1500 ms after stimulation. Analysis of the effects of conditioning at various sites was based on the variation in the number of spikes in the late discharge and was done by generating post stimulus histograms and by calculating the number of spikes in the late discharge. Each histogram was made of at least five consecutive responses, recorded at a rate of 2 or 3 per minute.

75

For each neuron, the first control was made of three to five consecutive histograms sampled during 15 min before starting the conditioning, and showing relatively stable number of reflex discharges. During conditioning at different C–T intervals, conditioning stimuli at DCr, DCb, DCc and RM were applied in random order, and the effects at a given interval were compared to the controls obtained before conditioning at this interval. 2.2.6. Histology At the end of each experiment, an electrolytic lesion was made in the RM, and the spinal cord and the brainstem were cut and stained ŽNissl. to verify the position of the electrode in RM and the extent of the DC lesions ŽFig. 1..

3. Results 3.1. Effects of rostral DC nuclear actiÕation on the flexion reflex After a recovery period of 3–4 h from the initial dose of anesthesia, a decerebrate–decerebellate and unparalyzed cat was quite unstable because it was difficult to establish an appropriate threshold for nociceptive heat stimulation eliciting a stable reflex reaction. Often, the same stimulation Žwith identical parameters of duration and temperature. could produce either a massive reaction involving both hindlimbs or no reaction at all. For this reason a subanesthetic dose of Nembutal Ž10–15 mgrkg, i.p.. was administered in order to obtain a stable reaction for about 1 h. Furthermore, stimulation rate was maintained at less than 1r90 s in order to avoid sensitization and increase of the motor reaction. Control responses were obtained after averaging 5 to 10 trials showing less than 15% variations during 10 to 20 min. Fig. 2A shows samples of polygraph records obtained from one experiment and the inhibition of the motor reaction by continuous DCst applied rostral to the spinal lesions at 100 Hz for 5 min. DCst produced a complete depression of the response during conditioning; this inhibition was evidenced by delaying and decreasing the amplitude of the response and was maintained after the end of conditioning. Complete recovery was obtained 5 to 10 min after the end of conditioning. The curve in Fig. 2B, obtained from averaging five trials made on five different preparations, shows the time course of inhibition produced by DCst at 100 Hz for 10 min. In general, the inhibition started about 1 min after the beginning of conditioning and reached its maximum within 5 min. Recovery of the response began at the end of conditioning and was complete after 20–30 min. In one cat, conditioning stimulation in RM was examined, using the same parameters as for DC conditioning,

76

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

Fig. 2. Long term inhibitory effects of continuous conditioning stimuli applied on the dorsal column rostral to the spinal lesions. In ŽA. samples of polygraph recording of electromecanogram of toe flexion reflex induced by radiant heat applied to the footpad Žlower trace in each panel., before Žcontrol. and after DCr conditioning stimulation Ž100 Hz, 5 min.. Note that the inhibitory effects are expressed by decreasing the amplitude and delaying the onset of the mechanical contraction as shown at 5 and 12 min following conditioning. Complete recovery of the amplitude and latency of the reflex are observed 25 min after the beginning of the conditioning train. The curve in ŽB., shows the time course of the inhibition of the flexion reflexes by DCr conditioning Ž10 min train. observed in five different cats. Each point represents the average" S.E.M. of five measurements at the indicated time interval.

and this produced similar, but stronger, inhibitory effects to those described after DCst. 3.2. Effects of conditioning stimuli on the C fiber reflex discharge of single alpha motorneurons Results of the first part of this study showed that rostral conditioning activation of the DCs can modulate global

reflex reaction to nociceptive stimuli for a period of time outlasting that of conditioning. But this type of preparation did not allow comparisons between the effects of DC conditioning rostral and caudal to the spinal lesions and to those of raphe stimulation for the following reasons: first, the instability of unanesthesized and unparalysed preparation; second, the slow rate of stimulation Ž1 to 2rmin. limiting the possibilities of sampling and of use of more

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

77

Table 1 Sensory modalities and sites of electrical stimulation eliciting discharges of motorneurons Sensory input

Number of cells

Total

Flexion of toes

Extension of toes

Pressure

Pinch

Polymodal

Unknown

4

3

8

18

36

41

than one type of conditioning at the same time; third, the necessity to use a high intensity of heat stimulation repeated over minutes or hours which can produce changes at the site of stimulation Žsee Ref. w7x.. For these reasons the comparative part of this study was made on C fiber evoked discharges recorded from single axons of alpha motorneurons. A total of 110 alpha motorneuron axons were isolated and the effects of various conditioning stimuli were studied on their late reflex discharges produced by C fiber afferent volleys. Sixty nine axons Ž62% of total. were also discharged by natural peripheral activations and of these 62 responded to high threshold cutaneous nociceptive stimulation such as pressure, pinch or radiant heat ŽTable 1.. Natural activation of the remaining 41 motorneurons could not be determined. These motorneurons were driven by electrical stimulation applied to the sural Ž n s 27., superficial peroneal Ž n s 63. or both nerves Ž n s 20.. Fig. 3 shows a sample record of a late motorneuronal discharge and the effects of various conditioning stimuli. This neuron was also driven by continuous pinch to the skin of the leg

110

Electrical stimulation Sural

Superficial peroneal

Both

27

63

20

and inhibited by rostral DC activations with a continuous train at 100 Hz ŽFig. 3, A3 and B3.. 3.3. Effects of intermittent stimulation Table 2 gives a summary of the effects of conditioning stimulations at the various DC sites and RM on the late discharges of motorneurons evoked by high intensity electrical stimulations. Using similar parameters for conditioning stimulations Žsame C–T intervals, shock train frequency, duration and intensity of the single shocks., stimulation at DCr inhibited a higher percent Ž90%. of cells than at DCc Ž61%., and the reverse was true for facilitation, while one neuron only was influenced by conditioning stimulation between the two dorsal cuts. The strongest inhibitory effects were obtained with RM conditioning stimulation and even with lower current intensity than that used for DC conditioning. In 14 out of 46 motorneurons, antagonistic effects were observed between conditioning stimulation to DCr and DCc, in a way that when the first produced inhibition the

Fig. 3. The effects of continuous train of conditioning stimuli Ž100 Hz, 60 s. from different origins on the motorneuronal reflex discharge in S1 ventral rootlet, induced either by electrical stimulation of the sural nerve or by pinching the footpad. The recording of reflex discharges was made at the time of application of the indicated conditioning stimulus, except for control and pinch-recovery which were recorded either before or after the end of the conditioning train. Late discharges were evoked by a single electrical test shock indicated by a dot in the lower case of each trace in rows 1 and 2.

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

78

Table 2 Summary of the effects of conditioning stimulation, at various sites, on the C-fiber-related late reflex discharge Conditioning from

Number of tested cells

Inhibition Ž%.

No effect Ž%.

Facilitation Ž%.

DCr, rostral to cuts DCb, between cuts DCc, caudal to cuts Raphe

70

63 Ž90.

3 Ž4.2.

4 Ž5.7.

30

1 Ž3.33.

29 Ž96.66.

0

46

28 Ž60.8.

6 Ž13.04.

12 Ž26.08.

41

40 Ž97.5.

1 Ž2.5.

0

second produced facilitation Ž n s 12. and vice versa for two other neurons. This antagonism was never observed between conditioning stimulation to DCr and RM Ž n s 40. and only one neuron was not affected by both types of conditioning. 3.4. Time courses of the conditioning testing (C–T) interÕals In classical C–T experiments using peripheral or central stimulations, that recruit mainly myelinated fibers, the

conduction times of the C or T volleys are rarely unequal or could have differences measured in a few milliseconds when working on small animals. For these reasons, C–T intervals are given positive values in the sense that the C shock must precede T shock by a time ranging from tens to hundreds of milliseconds. In the present study, the spectrum of velocities of the fiber systems involved in C–T intervals ranged from 0.5 ms for the smallest unmeylinated fibers Žrecruited by the T volley from the periphery or C volley descending from the brainstem. to 30–90 mrs for the myelinated fibers in the DCs. As a simple illustration, a DC conditioning volley applied caudal to the spinal lesions can reach L 7 or S 1 segments in 4 to 5 ms, while the Test C fiber afferent volley can reach the same spinal segments after 125–150 ms; consequently, a T shock applied 100 ms before the C shock Žcoming from DCc. would reach the spinal cord 25 ms after the C shock. Figs. 4 and 5 illustrate the effects of various conditioning stimuli at different C–T intervals. In Fig. 5, DC conditioning rostral and caudal to the spinal lesions showed opposite effects, while in Fig. 6, both had inhibitory effects. These two figures also show that conditioning

Fig. 4. Sample records of early and late discharges of a motorneuron illustrating the effects of the various types of conditioning using positive and negative values for conditioning-testing intervals Ži.e., negative, T preceding C shocks; positive, C preceding T shocks. as indicated on the right side of each row of panels. T and C shocks are indicated by dots and arrows, respectively. The time scale is 100 ms for all panels except for those in row 5. Note the similar inhibitory effects elicited by DCr and RM conditioning, while DC conditioning caudal to the lesions ŽDCc. produced facilitatory effects.

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

79

Fig. 5. Sample records of early and late discharges of a motorneuron showing inhibitory effects from the various sites of conditioning stimuli, using positive and negative C–T intervals. T and C shocks are indicated by dots and dashes, respectively. The records were obtained from a preparation with dissected and isolated dorsal columns ŽDC strand.. The conditioning from the dorsolateral funiculus ŽDLF. was applied at C 2 level, ipsilateral to the recording site, to assess the integrity of fibers traveling in that area following DC lesions. C–T intervals were y100 ms, 0 ms and q100 ms in panels 2–5 of columns A, B and C, respectively.

stimuli from different origins could affect the C fiber reflexes at a long range of C–T intervals where the C shocks can either precede or follow the T shock. For this reason, a correct determination of the time course of inhibition of the late discharge must satisfy the following conditions: first, the search for motorneurons which could elicit a stable late discharge in response to a single T shock at C fiber strength and the use of a short train for the C shocks Ž30 ms., in order to reduce the problem of central recruitment and the overlap between the times of the C and T shocks; second, the use of negative Žwhen T shock precedes C shock. and positive Žwhen the C shock precedes the T shock. C–T intervals ranging between y500 and q1000 ms. These conditions were fulfilled in seven motorneurons and four of them were inhibited by DCr, DCc and RM conditioning stimuli ŽFig. 6B. and three of them were facilitated by DCc conditioning stimuli ŽFig. 6A.. The

inhibition of these neurons by all types of conditioning started at y300 ms ŽT–C interval 300 ms. and ended at C–T of 1000 to 1500 ms. The time course of this inhibition showed three peaks: the first at y150 to y50 ms; the second at q50 to q150 ms; and the third late and less important at q300 to q500 ms. Facilitation of three motorneurons by DCc conditioning stimuli showed a shorter duration and a simpler shape with a first peak between y200 and y100 ms and a second peak between q150 and q300 ms. 3.5. Effects of continuous conditioning stimulation The effects of various conditioning stimuli are summarized in Table 3, which shows that the results obtained by continuous conditioning are comparable to those of intermittent conditioning stimulation.

80

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

Fig. 6. Time courses of the effects of the various conditioning stimuli on the late discharges of seven motorneurons. Three out of the seven motorneurons were facilitated by DC conditioning caudal to the lesions ŽA. and all of them were inhibited by RM and DC conditioning rostral to the lesions ŽB.. DC conditioning stimuli applied between the two DC lesions did not alter the discharges of three motorneurons ŽA.. The negative segments correspond to the time intervals where T shocks preceded C shocks, while the positive segments correspond to the classical paradigms for conditioning ŽC precedes T shocks..

The time course of either inhibition or facilitation elicited in seven motorneurons by conditioning trains at 100 Hz applied to the three sites are illustrated in Fig. 7. The maximum effects are seen during 5 to 10 s after

starting conditioning. All neurons were inhibited by DCr and RM conditioning stimuli ŽFig. 7A.. This inhibition peaked after 5 to 10 s Ž95% for RM, and 80% for DCr. and was maintained at about 50%–60% of control level at

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86 Table 3 Summary of the effects of conditioning stimulation Ž100 Hz, 60 s., at various sites, on the C-fiber-related late reflex discharge Conditioning from

Number of tested cells

Inhibition Ž%.

No effect Ž%.

Facilitation Ž%.

DCr, rostral to Cuts DCc, caudal to Cuts DCb, between Cuts Raphe

40

36 Ž90.

3 Ž2.5.

1 Ž2.5.

40

23 Ž57.5.

4 Ž10.

13 Ž32.5.

7

0

7 Ž100.

0

26

26 Ž100.

0

0

the end of the conditioning train. Recovery of the responses to normal levels was slightly more delayed for DCr than for RM conditioning and ceased within 1 min after the end of conditioning. The effects of conditioning stimulation of DCc on the same population of neurons were different and are shown in Fig. 7B. Four neurons were inhibited and three were facilitated. Inhibitory effects were less pronounced Ž50%. than those obtained from DCr and ended before cessation of the conditioning. On the other hand, facilitatory effects were stronger Ž200%., peaked at the beginning of conditioning train and showed a progressive decrease reaching the control level at the end of conditioning. Thus, the inhibitory effects of DCst rostral to the spinal lesion and of RM stimulation are cumulative and outlast the time of conditioning, while the effects of DCst caudal to the spinal lesions are extinguished either before Žfor inhibition. or at the end Žfacilitation. of the conditioning train ŽFig. 7A and B..

81

train, the recovery time becomes longer and shows a second phase of inhibition occurring at 2 or 3 min after the end of conditioning stimulation and lasting for 8 to 10 min. This late phase of inhibition becomes more evident with the second paradigm of DCr conditioning, using a longer conditioning train of 3–5 min on the C fiber late reflex discharges of five motorneurons. The curve in Fig. 8B shows that the inhibition produced by a long train DCr conditioning presents two main components. The first one is a rapid peak occurring during the conditioning train followed immediately by an incomplete recovery. The second long-lasting component starts few minutes after the end of conditioning and lasts for 20 to 25 min. A second long train of DCr conditioning applied to the same cell

3.6. CumulatiÕe effects of rostral DC actiÕation Short train Ž40–60 s, 100 Hz. DCr conditioning produced an immediate inhibition of the C fiber reflexes, reaching its maximum within 5–15 s and declining gradually before the end of the conditioning train; total recovery occurred within 1 or 2 min after the cessation of the conditioning. This rapid decline of inhibition, even during conditioning, cannot account for the long lasting inhibition of pain seen in patients or observed on cats in the first parts of this study. For this reason, two additional DCr conditioning paradigms were used. In the first paradigm, three or more conditioning trains Ž40–60 s, 100 Hz. were applied successively on DCr in a manner that application of the next train was made when the C fiber late discharge of a neuron had recovered from the previous conditioning stimulation. Fig. 8A illustrates the results of repetitive DCr conditioning obtained from four different motorneurons. These results reveal two main observations. First, in each neuron the amount of inhibition was constant and did not show any significant variations with repetitive conditioning when using the same parameters for the conditioning train Že.g., intensity of shock, duration, frequency.. Second, after the third conditioning

Fig. 7. Time courses of the inhibitory and facilitatory effects of conditioning stimulation Žshort trains at 100 Hz, 60 s., at the various sites, on the late motorneuronal late reflex discharges. Inhibitory effects produced by raphe ŽRM. and DC conditioning rostral to spinal lesions ŽDCr. are illustrated in ŽA.. Each point in the curves represents the average"S.E.M. of the discharge of seven motorneurons as compared to the control before or after conditioning. ŽB. illustrates the facilitation of discharges of three motorneurons and the inhibition of the discharges of the other four motorneurons by the same conditioning train applied caudal to the spinal lesions ŽDCc..

82

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

nociceptive mechanisms through a brainstem spinal loop and to compare these results to the more documented effects of either segmental DC activation or raphe stimulations on the same spinal mechanisms. DC conditioning stimuli applied rostral to the spinal lesions produced significant inhibition of the flexion withdrawal reflex and inhibited the C fiber reflex discharge of about 90% of the motorneurons studied. DC lesions at C1 and C3 level prevented any possible antidromic activation of the spinal cord. The DC strand preparation Žthree cats. precluded the possibility of spread of current to neighboring fiber systems or other structures in the spinal cord or the brainstem. Furthermore, decerebellation and decerebration implied that any DC nuclear influence at the level of the spinal cord must involve ponto-bulbar mechanisms. The positive answer to this first question led to several other questions. First, is the C fiber reflex discharge a valid test for pain-related behavior? Second, what are the possible mechanisms underlying the different effects obtained from DC conditioning rostral and caudal to the spinal lesions? Third, is there any link between the effect of conditioning from DCr and from RM? Fourth, what are the neural mechanisms underlying the time course of the observed inhibition? 4.1. Validity of the test for pain

Fig. 8. Time courses of the long term inhibitory effects of DC repetitive conditioning trains applied rostral to the spinal lesions. A1 represents the effects of conditioning train Ž100 Hz, 60 s. repeatedly applied on the same population of four neurons. Note the increased time of inhibition becoming more evident with the third train Žcond. III.. ŽB. illustrates the effects of repeated conditioning trains of longer duration Ž100 Hz, 3 min.. The values were obtained from five neurons subjected to three trains of conditioning, and correspond to their average response to the third application of the conditioning stimulation.

produced the same effects but with longer lasting late inhibition. The effects of the same paradigms of conditioning Žeither with repetitive short train or long train. stimulations were tested with either RM or DCc. The former produced similar effects to those of DCr conditioning, while the latter showed only the first peak of either inhibition or facilitation lasting for the time of conditioning or for few minutes after its cessation.

4. Discussion The principal aim of this study was to investigate whether rostral activation of the DCs can modulate spinal

Withdrawal flexion reflexes in humans are associated with second Žburning. pain sensation and activation of C fiber afferents w8,46,68x. In animal experimental models, the late reflex discharges in alpha motorneurons associated with cutaneous C fiber afferent volleys ŽC fiber reflexes. and various nociceptive stimuli, have been often used as a test for pain-related behavior w10,26,46,57,58x. All motorneurons used in this study were discharged by electrical stimulations at C fiber strength applied to cutaneous or mixed nerves, and most of them were also activated by cutaneous nociceptive stimuli. These motorneurons fulfilled the criteria for the identification of alpha motorneurons defined by Chung et al. w10x and most of them appear to belong, at least, to the flexion reflex arc as defined by Hagbarth w19x and Holmqvist and Lundberg w21x. The effects of conditioning stimuli at the different sites were tested on the late reflex discharges, produced by the A delta and C fiber afferent volleys. A possible supraspinal contribution to this late discharge could not be totally excluded, but our finding that progressive lesions of the lateral tracts of the spinal cord produced an increase in this discharge and a decrease in inhibition by DCr or RM conditioning w55x could be correlated with previous evidence concerning a tonic descending inhibition of flexor reflexes in decerebrate preparations w11,21x. Furthermore, the inhibition of nociceptive spinal reflexes by rostral DC activation is in line with our previous description of a DC input into the brainstem reticular formation and its in-

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

hibitory interactions with activity coming through the ascending fibers of the anterolateral columns w51,52,56x. 4.2. Possible neural mechanisms inÕolÕed in conditioning stimulations at DCr and DCc Although the late discharge in the majority of the motorneurons was inhibited by DC conditioning rostral to Ž90%. or caudal to Ž61%. the spinal lesions, the neuronal mechanisms are most likely to be quite different. This assumption is further supported by the antagonistic effects shown on 25 motorneurons inhibited by DCr and facilitated by DCc. Several studies have shown that DC stimulation caudal to spinal lesions Žsingle or short train of shocks. inhibited up to the two thirds of the various types of dorsal horn neurons Žrelay or non-relay, nociceptive or non-nociceptive.; this inhibition followed a short period of excitation and lasted about 300 ms w15,20,35,53x. Most of these studies attributed the inhibitory effects to the activation of gating mechanisms, in the dorsal horn, between the inputs of large myelinated and thinly myelinated or unmyelinated fibers as has been initially theorized by Melzack and Wall w40x. Conditioning at DCc stimulation produced mixed inhibitory and excitatory effects on spinal motorneurons. Similar effects have been recently reported by Rees and Roberts w47x on the response of dorsal horn neurons to intense cutaneous stimuli in decerebrate rats. This substantiates earlier studies by Brown and Martin w6x and Saade´ et al. w53x showing inhibitory effects of rostral DC activation upon various types of dorsal horn neurons Žrelay and non-relay, low threshold or high threshold mechanoreceptive or wide dynamic range.. The fact that these neurons, and mainly the nociceptive driven, may intervene in the polysynaptic circuits involved in flexion withdrawal reflex w58x could explain the inhibition of these reflexes by rostral DC activation. However, the neural mechanisms and fiber systems involved in DCr conditioning are different from those activated by segmental or DCc stimulation. 4.3. Possible link between conditioning stimulations at DCr and RM Inhibitory effects from RM conditioning were predicted from the well documented role of this structure as a pain modulating center Žfor review, see Refs. w5,14,69x.. Our study, however, provides a clear demonstration of inhibitory effects from RM stimulation on the C fiber late reflex discharge. This finding is closely related to two other important observations: the parallelism between the effects of RM and DCr conditioning stimulations and the time courses of the inhibition. The first observation could be related to our previous finding of a DC input into the inferior raphe complex and periaqueductal gray w28,51x. For the second observation, we have introduced a condi-

83

tioning-testing paradigm in the study of inhibition and facilitation whereby the testing could precede the conditioning shock. The reason for this was explained in the Results section. Most of the earlier studies on the effect of RM conditioning on pain employed conditioning stimulation with a continuous train of shocks. In the few studies using intermittent small train or single conditioning shocks, the C fiber related discharges in the test response were not systematically studied w6,15,33,47x. Furthermore, the reported inhibition started early Ž10–25 ms. and ended after 500–700 ms. In our study, using minimum current intensity to elicit inhibitions and just suprathreshold test stimuli for the C fiber reflex discharge, the curve of inhibition showed three main peaks: the first, between y300 and y50 ms ŽT shock preceding C shock.; the second, between q50 and q200 ms; and the third starting at q500 ms Žsee Fig. 6.. Inhibitory effects, from conditioning stimulations at DCr and RM in our study, started at about y400 ms and ended at about q2000 ms. Duration and strength of inhibition could be made longer and stronger, respectively, if we used stronger currents for testing and conditioning stimuli, especially for RM stimulation. In view of the different sizes of fibers and, therefore, conduction velocity spectrum Žfrom 1 to 70 mrs. of the raphe–spinal system w66,67x, the first peak of inhibition can be attributed to the descending volley via the large fiber system reaching the spinal cord within 3 to 5 ms Žconduction distance about 25 cm.; this volley can inhibit motorneuronal firing if it preceded or coincided with the late discharge. The second peak could be attributed to the descending volley Žfrom the brainstem. in the small fiber system reaching the segmental level after a latency ranging between 50 and 150 ms. The third peak could be attributed to the various synaptic mechanisms at the spinal level depending on the collateralization of the raphe spinal fibers and on the nature of liberated neurotransmitters and neuromodulators w5,13,69x. However, the observed parallelism between the effects of DCr and RM conditionings does not imply that DCr effects are mediated, exclusively, through RM. This is supported by the following observations: first, DCst has been shown to activate various brainstem structures in addition to RM w28,51,52,56x; second, RM electrical stimulation could involve neighboring areas and passing fibers; third, unpublished data from our laboratory have shown that anesthetic block of the RM area produced a partial reversal of the inhibitory effects of DCr conditioning. 4.4. The possible neural mechanisms underlying the obserÕed long term inhibition Conditioning stimulation with repetitive short trains Ž1 min. or longer trains Ž3–5 min. applied to either DCr or RM produced long term inhibition of the C fiber reflexes comparable to that obtained in the model of toes flexion reflex elicited by nociceptive radiant heat, while the same

84

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

conditioning stimuli applied to the DC caudal to spinal lesion produced only short lived inhibition or facilitation. Comparison of the effects of long train conditioning applied to DCr and DCc, shows that both exhibited an early phase of inhibition recovering by the end of the conditioning train, while DCr conditioning exhibited a late long lasting inhibition persisting after the end of the train. The early phase of inhibition could be correlated with the parasthesia and the raised thresholds of somatic sensations Žincluding pain. and of the flexion reflexes observed in patients during DC stimulation and with the reversal of these changes at the end of stimulation w16,24,32,34,44x. Altered sensations and reduction of flexion reflexes during DCst are in line with observations from previous studies showing that DCst exerts a short term inhibition Žms to sec. of nociceptive and non-nociceptive neurons, recorded in the dorsal horn w6,15,35,53x or at supraspinal levels w29,44,45x. These neurons could be involved, at spinal level, in the polysynaptic circuits of the nociceptive flexion reflexes and at any or all central levels in the processing of perceived somatic sensations. The major finding of this study, however, is the second phase of the long lasting inhibition produced by DCr and RM conditioning which presented the following important features: Ž1. It appeared after either repetitive short conditioning trains or after a long conditioning train and showed accumulation or summation of effects of previous conditionings. Ž2. It started after the end of conditioning and its duration and intensity were related to the number of repetitions andror the durations of the conditioning trains. Ž3. It appeared to be mediated by ponto-bulbo-spinal mechanisms because it was elicited by conditioning stimulation to either DCr or RM, but not by DCc conditioning. This ponto-bulbo-spinal loop could provide a neural substrate for the long term inhibition of nociceptive flexion reflexes by DCr conditioning and could also be related to our earlier report on long term inhibition of acute and tonic pain by DC nuclear stimulations in behaving rats with chronic DC lesions w54x. Furthermore, this loop could provide an experimental substrate for the clinical observations of long term relief of chronic pain by repetitive DC stimulation and after the recovery of normal sensations and flexion reflexes w16,18,30,41x. Two research groups have recently addressed the question of spinal vs. supraspinal contribution to the long-term inhibitory effects of DC stimulation on spinal activities induced by nociceptive stimuli w42,47,48x. Using mononeuropathic rats, Meyerson et al. w42x and Ren et al. w48x reported long lasting inhibitory effects of spinal cord stimulation on the early reflex discharges related to the activation of Ab fibers in the nerve ligated leg, and a short lived inhibition of its equivalent in the intact leg. However, the C fiber mediated reflex discharge was not affected by spinal cord stimulation. Both reports also demonstrated that this inhibitory effect was not affected by acute transection of the spinal cord. Our present results on the short

lived effects of the DC conditioning, applied caudal to the DC cuts, correlate well with those reported on the intact leg by Meyerson et al. w42x and Ren et al. w48x. The discrepancies concerning the effects on the C fiber evoked late discharges could be attributed to two major differences between our experimental protocols. First, the mononeuropathic rats used in both studies, might have had the well known syndromes produced by injuries to the peripheral nerves. Such syndromes include allodynia, due to abnormal function of the afferent A fibers, and reflex sympathetic dystrophy attributed, at least partially, to an abnormal function of spinal efferents. Neither study provided a simultaneous comparison between the effects of DC stimulation rostral and caudal to selective DC lesions. Further support to our hypothesis about the long lasting inhibition of nociceptive behavior by rostral DC activation w53,54x is provided by the work of Rees and Roberts w47x performed on intact animals. This study described short and long lasting inhibitions of the activities of dorsal horn neurons by DC conditioning stimuli applied caudal and rostral to selective DC lesions, respectively. A minor discrepancy between our results and those of Rees and Roberts w47x is that the inhibitory effects of rostral DC activation were present in cats subjected either to transcollicular or subcollicular decerebration. In conclusion, the results of the present study provide an experimental substrate for the long-term relief from pain produced by DC stimulation. This effect appears to be mediated, to a large extent, by the activation of brainstem centers known to produce analgesia. Acknowledgements This work was supported by grants from the Lebanese National Council for Scientific Research and by the Franco–Lebanese C.E.D.R.E. project. References w1x H. Bantli, J.R. Bloedel, D.M. Long, P. Thienprasit, Distribution of activity in spinal pathways evoked by experimental dorsal column stimulation, J. Neurosurg. 42 Ž1975. 290–295. w2x H. Bantli, J.R. Bloedel, P. Thienprasit, Supraspinal interactions resulting from experimental dorsal column stimulation, J. Neurosurg. 42 Ž1975. 296–300. w3x K.J. Berkley, R.J. Budell, A. Blomquist, M. Bull, Output systems of the dorsal column nuclei in the cat, Brain Res. Rev. 11 Ž1986. 199–225. w4x A.L. Berman, The Brain Stem of the Cat, University of Wisconsin Press, Madison, 1968. w5x J.-M.R. Besson, Supraspinal modulation of the segmental transmission of pain, in: H.W. Kosterlitz, L.Y. Terenius ŽEds.., Pain and Society, Dahlem Konferenzen 1980, Verlag Chemie, Weinheim, pp. 161–182. w6x A.G. Brown, H.F. Martin III, Activation of descending control of the spinocervical tract by impulses ascending the dorsal columns and relaying through the dorsal column nuclei, J. Physiol. 235 Ž1973. 535–550.

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86 w7x R.E. Burke, P. Rudomin, L. Vyklicky, F.E. Zajac III, Primary afferent depolarization and flexion reflexes produced by radiant heat stimulation of the skin, J. Physiol. 213 Ž1971. 185–214. w8x I.G. Campbell, E. Carstens, L.R. Watkins, Comparison of human pain sensation and flexion withdrawal evoked by noxious radiant heat, Pain 45 Ž1991. 259–268. w9x J.N. Campbell, Examination of possible mechanisms by which stimulation of the spinal cord in man relieves pain, Appl. Neurophysiol. 44 Ž1981. 181–186. w10x J.M. Chung, Z.R. Fang, C.L. Cargill, W.D. Willis, Prolonged, naloxone-reversible inhibition of the flexion reflex in the cat, Pain 15 Ž1983. 35–53. w11x M. Dougherty, S. Shea, C.N. Liu, W.W. Chambers, Effects of spinal cord lesions on cutaneously elicited reflexes in the decerebrate cat. Tonic bulbospinal and spinobulbar inhibitory systems, Exp. Neurol. 26 Ž1970. 551–570. w12x T. Eliasson, L.E. Augutinsson, C. Mannheimer, Spinal cord stimulation in severe angina pectoris — presentation of current studies, indications and clinical experience, Pain 65 Ž1996. 169–179. w13x H.L. Fields, S.D. Anderson, Evidence that raphe–spinal neurons mediate opiate and midbrain stimulation-produced analgesia, Pain 5 Ž1978. 333–349. w14x H.L. Fields, A.I. Basbaum, Brainstem controls of spinal pain-transmission neurons, Annu. Rev. Physiol. 40 Ž1978. 217–248. w15x R.D. Foreman, J.E. Beall, A.E. Applebaum, J.D. Coulter, W.D. Willis, Effects of dorsal column stimulation on primate spinothalamic tract neurons, J. Neurophysiol. 39 Ž1976. 534–546. w16x L. Garcia-Larrea, M. Sindou, F. Mauguiere, Nociceptive flexion reflexes during analgesic neurostimulation in man, Pain 39 Ž1989. 145–156. w17x P.L. Gildenberg, Treatment of spasmodic torticollis by dorsal column stimulation, Appl. Neurol. 41 Ž1978. 113–121. w18x J.M. Gybels, R. Kupers, Central and peripheral electrical stimulation of the nervous system in the treatment of chronic pain, Acta Neurochir. Suppl. 38 Ž1987. 64–75. w19x K.E. Hagbarth, Excitatory and inhibitory skin areas for flexor and extensor motorneurones, Acta Physiol. Scand. 26 Ž1952. 1–58, Suppl. 94. w20x H.O. Handwerker, A. Iggo, M. Zimmermann, Segmental and supraspinal actions on dorsal horn neurons responding to noxious and non-noxious skin stimuli, Pain 1 Ž1975. 147–165. w21x B. Holmqvist, A. Lundberg, Differential supraspinal control of synaptic actions evoked by volleys in the flexion reflex afferents in alpha motorneurones, Acta Physiol. Scand. 54 Ž1961. 1–51, Suppl. 186. w22x R. Hoppenstein, Electrical stimulation of the ventral and dorsal columns of the spinal cord for relief of chronic intractable pain: Preliminary report, Surg. Neurol. 4 Ž1975. 187–194. w23x Y. Hosobuchi, J.E. Adams, P.R. Weinstein, Preliminary percutaneous dorsal column stimulation prior to permanent implantation, J. Neurosurg. 37 Ž1972. 242–245. w24x J.P. Hunter, P. Ashby, Segmental effects of epidural spinal cord stimulation in humans, J. Physiol. 474 Ž1994. 407–449. w25x L.S. Illis, E.H. Sedgwick, A.E. Oygar, M.A. Awadalla, Dorsal column stimulation in the rehabilitation of patients with multiple sclerosis, Lancet 26 Ž1970. 1383–1385. w26x G.A. Iwamoto, H. Ryu, I.H. Wagman, Contribution of unmyelinated afferents from the sural nerve to spinal reflex activity, Exp. Neurol. 62 Ž1978. 260–267. w27x S.J. Jabbur, N.E. Saade, ´ N.R. Banna, Excitatory and inhibitory interactions between dorsal and ventral column inputs into the cat postcruciate cortex, in: E. Grastyan, P. Molnar ŽEds.., Adv. Physiol. Sci., Vol. 16, Sensory Functions, Pergamon, Budapest, 1981, pp. 171–180. w28x S.J. Jabbur, N.E. Saade, ´ S.F. Atweh, Dorsal column ŽDC. input into periaqueductal gray ŽPAG. in decerebrate decerebellate cats, Society for Neuroscience ŽUSA. 10 Ž1984. 101, Abstracts.

85

w29x S.J. Jabbur, N.E. Saade, ´ Interactions between dorsal column and ventral spinal inputs in somatosensory thalamocortical areas and in various brainstem nuclei in the cat, in: M.M. Rowe, W.D. Willis, Jr. ŽEds.., Development, Organization, and Processing in Somatosensory Pathways, Neurology and Neurobiology, Vol. 14, Alan R. Liss, New York, 1985, pp. 223–229. w30x F.U. Krainick, U. Thoden, Spinal cord stimulation, in: P.D. Wall, R. Melzack ŽEds.., Textbook of Pain, Churchill Livingston, London, 1989, pp. 920–924. w31x S.J. Larson, A. Sances, J.F. Cusick, G.A. Meyer, T. Swiontek, A comparison between anterior and posterior spinal implant systems, Surg. Neurol. 4 Ž1975. 180–186. w32x S.J. Larson, A. Sances Jr., D.H. Riegel, G.A. Meyer, D.E. Dallmann, T. Swiontek, Neurophysiological effects of dorsal column stimulation in man and monkey, J. Neurosurg. 41 Ž1974. 217–223. w33x A.R. Light, E.J. Casale, D.M. Menetrey, The effects of focal stimu´ lation in nucleus raphe magnus and periaqueductal gray on intracellularly recorded neurons in spinal laminae I and II, J. Neurophysiol. 56 Ž1986. 555–571. w34x U. Lindblom, B.A. Meyerson, Influence on touch, vibration and cutaneous pain of dorsal column stimulation in man, Pain 1 Ž1975. 257–270. w35x U. Lindblom, D.N. Tapper, Z. Wiesenfeld, The effect of dorsal column stimulation on the nociceptive response of dorsal horn cells and its relevance for pain suppression, Pain 4 Ž1977. 133–144. w36x B. Linderoth, B.A. Meyerson, Dorsal column stimulation: modulation of somatosensory and autonomic function, The Neurosciences 7 Ž1995. 263–277. w37x S. Marchand, M.C. Bushnell, P. Molina-Negro, S.N. Martinez, G.H. Duncan, The effects of dorsal column stimulation on measures of clinical and experimental pain in man, Pain 45 Ž1991. 249–257. w38x M. Meglio, B. Cioni, E. D’amico, G. Ronzoni, G.F. Rossi, Epidural spinal cord stimulation for the treatment of neurogenic bladder, Acta Neurochir. 54 Ž1980. 191–199. w39x M. Meglio, B. Cioni, G.F. Rossi, S. Sandric, P. Santarelli, Spinal cord stimulation affects the central mechanisms of regulation of heart rate, Appl. Neurophysiol. 49 Ž1986. 139–146. w40x R. Melzack, P.D. Wall, Pain mechanisms: a new theory, Science 150 Ž1965. 971–979. w41x B.A. Meyerson, Electric stimulation of the spinal cord and brain, in: J.J. Bonica ŽEd.., The Management of Pain, Lea and Febiger, Philadelphia, 1990, pp. 1862–1877. w42x B.A. Meyerson, B. Ren, P. Herregodts, B. Linderoth, Spinal cord stimulation in animal models of mononeuropathy: effects on the withdrawal response and the flexor reflex, Pain 61 Ž1995. 229–243. w43x B.S. Nashold Jr., H. Friedman, Dorsal column stimulation for control of pain. Preliminary report on 30 patients, J. Neurosurg. 36 Ž1972. 590–597. w44x B. Nashold Jr., G. Somjen, H. Friedman, Paresthesias and EEG potentials evoked by stimulation of the dorsal funiculi in man, Exp. Neurol. 36 Ž1972. 273–287. w45x J.K. Nyquist, J.H. Greenhoot, Responses evoked from the thalamic centrum medianum by painful input: suppression by dorsal funiculus conditioning, Exp. Neurol. 39 Ž1973. 215–222. w46x D.D. Price, Characteristics of second pain and flexion reflexes indicative of prolonged central summation, Exp. Neurol. 37 Ž1972. 371–387. w47x H. Rees, M.H.T. Roberts, Antinociceptive effects of dorsal column stimulation in the rat: involvement of the anterior pretectal nucleus, J. Physiol. 417 Ž1989. 375–388. w48x B. Ren, B. Linderoth, B.A. Meyerson, Effects of spinal cord stimulation on the flexor reflex and involvement of supraspinal mechanisms: An experimental study in mononeuropathic rats, J. Neurosurg. 84 Ž1996. 244–249. w49x N.E. Saade, ´ S.J. Jabbur, N.R. Banna, Interactions of isolated dorsal column and ventral spinal ascending volleys on single neurons in the post-cruciate cortex ŽSmI. of the cat, Brain Res. 201 Ž1980. 446–451.

86

N.E. Saade´ et al.r Brain Research 846 (1999) 72–86

w50x N.E. Saade, ´ S.F. Atweh, M.S. Tabet, S.J. Jabbur, Inhibition of nociceptive withdrawal flexion reflexes through a dorsal column– brainstem–spinal loop, Brain Res. 335 Ž1985. 306–308. w51x N.E. Saade, ´ A.S. Jundi, S.J. Jabbur, N.R. Banna, Dorsal column input to inferior raphe centralis neurons, Brain Res. 250 Ž1982. 345–348. w52x N.E. Saade, ´ N.A. Salibi, N.R. Banna, A.L. Towe, S.J. Jabbur, Spinal input pathways affecting the medullary gigantocellular reticular nucleus, Exp. Neurol. 80 Ž1983. 582–600. w53x N.E. Saade, ´ M.S. Tabet, N.R. Banna, S.F. Atweh, S.J. Jabbur, Inhibition of nociceptive evoked activity in spinal neurons through a dorsal column–brainstem–spinal loop, Brain Res. 339 Ž1985. 115– 118. w54x N.E. Saade, ´ M.S. Tabet, S.A. Souiedan, M. Bitar, S.F. Atweh, S.J. Jabbur, Supraspinal modulation of nociception in awake rats by stimulation of the dorsal column nuclei, Brain Res. 369 Ž1986. 307–310. w55x N.E. Saade, ´ S.F. Atweh, S.J. Jabbur, Descending inhibitory pathways from the brainstem modulating the late reflex discharge ŽLRD. evoked by C-fiber afferent volleys, Pain Suppl. 5 Ž1993. S269. w56x N.A. Salibi, N.E. Saade, ´ N.R. Banna, S.J. Jabbur, Dorsal column input into the reticular formation, Nature 288 Ž1980. 481–483. w57x E.D. Schomburg, H. Steffens, Synaptic responses of lumbar amotorneurones to selective stimulation of cutaneous nociceptors and low threshold mechanoreceptors in the spinal cat, Exp. Brain Res. 62 Ž1986. 335–342. w58x J. Schouenborg, B.H. Sjolund, Activity evoked by A- and C-afferent fibers in rat dorsal horn neurons and its relation to a flexion reflex, J. Neurophysiol. 50 Ž1983. 1108–1121. w59x C.N. Shealy, J.T. Mortimer, J.B. Reswick, Electrical inhibition of pain by stimulation of the dorsal columns, Anesth. Analg. 46 Ž1967. 489–491.

w60x A.G. Shetter, J.R. Atkinson, Dorsal column stimulation: its effect on medial bulboreticular unit activity evoked by noxious stimuli, Exp. Neurol. 54 Ž1977. 185–198. w61x J. Siegfried, J.U. Krainick, H. Haas, C. Adorjani, M. Meyer, U. Thoden, Electrical spinal cord stimulation for spastic movement disorders, Appl. Neurophysiol. 41 Ž1978. 134–141. w62x W.H. Sweet, J.G. Wepsic, Stimulation of the posterior columns of the spinal cord for pain control: indications, technique, and results, Clin. Neurol. Neurosurg. 21 Ž1974. 278–310. w63x R.C. Tallis, L.S. Illis, E.M. Sedgwick, C. Hardwidge, J.S. Garfield, Spinal cord stimulation in peripheral vascular disease, J. Neurol. Neurosurg. Psychiat. 43 Ž1983. 478–484. w64x U. Thoden, J.U. Krainick, H.M. Straburg, D. Wenzel, Modulation of monosynaptic reflexes by dorsal column stimulation ŽDCS. in spasticity, Pflugers Arch. Gesamte Physiol. Menschen Tiere 365 Ž1976. R41. w65x P.D. Wall, W.H. Sweet, Temporary abolition of pain in man, Science 155 Ž1967. 108–109. w66x D.C. West, J.H. Wolstencroft, Strength-duration characteristics of myelinated and non-myelinated bulbospinal axons in the cat spinal cord, J. Physiol. 377 Ž1983. 37–50. w67x D.C. West, J.H. Wolstencroft, Location and conduction velocity of raphe spinal neurones in nucleus raphe magnus and raphe pallidus in the cat, Neurosci. Lett. 5 Ž1977. 147–151. w68x J.C. Willer, Comparative study of perceived pain and nociceptive flexion reflex in man, Pain 3 Ž1977. 69–80. w69x W.D. Willis, The raphe–spinal system. Brainstem control of spinal cord function, Res. Top. Physiol. 6 Ž1984. 141–214. w70x C.J. Woolf, J.E. Swett, The cutaneous contribution to the hamstring flexor reflex in the rat: an electrophysiological and anatomical study, Brain Res. 303 Ž1984. 299–312.