Exp Brain Res (1991) 84:635~42

BrainResearch 9 Springer-Verlag1991

Effects of electrical stimulation of the thoracic spinal cord on bladder and external urethral sphincter activity in the decerebrate cat B. Fedirchuk and S.J. Shefchyk Departments of Medicine and Physiology,Universityof Manitoba, 770 BannatyneAve., WinnipegCanada, R3E 0W3 Received August 17, 1990 / Accepted October 19, 1990 Summary. Electrical stimulation of the spinal cord above the sacral segments was used to produce coordinated micturition in the paralysed decerebrate cat. Stimulation of the superficial aspect of the dorsolateral funiculus (DLF) within the lower thoracic (T9-T13) segments produced a bladder contraction coordinated with decreased activity in the external urethral sphincter (EUS) branch of the pudendal nerve during which time fluid was expelled. In addition, a similar response was observed with DLF stimulation at the boundary of the L5/L6 segments. At the second cervical spinal segment, however, stimulation of a more lateral and ventral portion of the superficial spinal white matter was the only effective site for producing micturition. The spinal cord-evoked response was comparable to the micturition evoked by electrical stimulation the pontine micturition centre (PMC) within the brainstem. A bilateral lesion of the dorsal columns (DC) and the dorsolateral funiculi (DLF) at the lower thoracic levels abolished reflex micturition evoked by bladder distension. However stimulation rostral to the lesion, within the PMC or thoracic DLF, continued to produce coordinated bladder and sphincter response during voiding. Stimulation caudal to the lesion produced a decrease in pudendal nerve activity but did not produce a void or bladder pressure change. This reduction in pudendal nerve activity could be abolished with a second lesion of the superficial DLF caudal to the stimulation site. It was concluded that stimulation of the thoracic dorsolateral funiculus activates both ascending and descending fibres which can influence the bladder and/or sphincter muscles. The spinal cordevoked voiding was hypothesized to be due to activation of some portion of the ascending limb of the spinobulbospinal micturition reflex loop. The decreased activity produced by stimulation of the thoracic DLF caudal to a bilateral DC/DLF subtotal cord lesion may be mediated by fibres descending in the dorsolateral funiculus. The possibility that the spinal cord stimulation antidromically activated axons of neurons having segmental collaterals Offprint requests to: S.J. ShefchykDept. Physiology(address see

above)

capable of influencing pudendal neural activity cannot be exclused at this time.

Key words: Micturition - Spinal cord lesions - Thoracic spinal cord Pudendal

Introduction Since Barrington's description of a pontine micturition facilitatory region (Barrington 1921, 1925), the influence of the upper pons on bladder contractility and micturition has been examined both anatomically and electrophysiologically (see Kuru 1965; Loewy et al. 1979; McMahon and Spillane 1982; Sugaya et al. 1987; Shefchyk 1989). While this region, often referred to as the pontine micturition centre (PMC) is though to play an important role in facilitating micturition, neither the sensory pathways relaying to the PMC, nor the descending pathways mediatingthe PMC influences onto sacral spinal neurons have been adequately described. Although a variety of investigative approaches have examined the spinal cord pathways mediating ascending and/or descending systems in different preparations and species (Stewart 1899; Yamamoto et al. 1956; Nathan and Smith 1958; McMahon and Morrison 1982a-c), the trajectories of functionally identified pathways producing micturition by a spinobulbospinal loop remain uncertain. Using the measures of bladder pressure and EUS pudendal nerve activity, electrical stimulation of sites within the spinal white matter rostral to the sacral spinal segments was undertaken in an attempt to identify micturition related pathways. Subtotal lesions of the spinal cord were used to determine whether evoked responses were mediated via ascending or descending fibre tracts. Preliminary results have been reported (Fedirchuk and Shefchyk 1989).

Methods The data for this study was obtained from 29 male cats weighing from2.2 to 4.3 kg. Initialsurgicalprocedureswerecarried out usinga

636 mixture of nitrous oxide and halothane anaesthesia. A tracheostomy was performed and arterial blood pressure was measured from one carotid artery. The external jugular vein was cannulated and a bicarbonate/glucose buffer solution (0.84 g NaHCO 3 and 5 g glucose in 100 ml distilled water) was administered intravenously at a rate of 2 4 ml/h. The animal's temperature was monitored and maintained between 37 ~ and 39 ~ using heating lamps. The urinary bladder was exposed via a ventral midline incision and an infant feeding tube (size Fr. 5) was inserted through a small puncture in the bladder wall and sutured into place. This cannula was later connected to an infusion-pumpand pressure transducer to monitor bladder pressure. In two animals bipolar electromyographic electrodes were placed in the left side of the external urethral sphincter muscle to monitor EUS activity prior to paralysis. A lateral incision at the base of the tail exposed the ischiorectal cavity and was extended to expose the following peripheral nerves on the right side: pudendal nerve branches to the external urethral sphincter (EUS), external anal sphincter and sensory pudendal branch including the dorsal penile nerve (see Martin et al. 1974), and the nerves innervating the posterior biceps/semitendinosus and semimembranosus/anterior biceps hindlimb muscles. The nerves were freed from connective tissue, cut and ligated in preparation for mounting on bipolar recording/stimulatingelectrodes when the animal was later placed in the spinal cord frame. The spinal cord was exposed by laminectomies at several levels in different animals. In all cats the ninth to thirteenth thoracic (T9-T13) laminae were removed; in six animals an additional laminectomy was made either at the upper cervical level (C1-C3) or at the fourth to seventh lumbar (L4-L7) levels. Clamps were placed at the lumbar and/or thoracic levels to support the vertebral column and the cat was placed in the spinal cord recording frame. The skin surrounding the peripheral nerves and laminectomies was fashioned into a pool and filled with mineral oil heated to body temperature. The head was placed in a stereotaxic headholder and the cranium exposed. The bone over the occipital and parietal lobes was removed and a precollicular postamammillary decerebration was performed. The anaesthesia was discontinued at this point and the animal was paralysed using Flaxedil (2-3 mg/kg initial dose, supplementary doses 1 mg/ kg/ h) and artificially ventilated (expired COz maintained at 2.5-5%). All animals were capable of reflex voiding evoked by bladder distension (2 ml/min) following the decerebration. The volume required for evoking reflex micturition varied from animal to animal and ranged from 10 to 40 ml. In addition, micturition was evoked using electrical stimulation within the central nervous system. Prior to each stimulation trial the bladder was infused with a volume of warm saline less than half of that required to produce a distension reflex void in that animal. Electrical stimulation of the PMC site was done using a steel monopolar cathode electrode (exposed tip 200 #m; 0.2 ms square wave pulses at 100 Hz, currents 20 250 #A; also see Shefchyk 1989). The spinal cord was stimulated using monopolar tungsten cathode electrodes (exposed tip 60 100 #m; DC resistance 0.7 1.5 M r ; 0.2 ms square wave pulses at 100 Hz, 20 500 ~A). In 3 animals where bipolar electrodes were used to stimulate a portion of the spinal white matter, a pair of silver ball electrodes were placed on the surface of the exposed spinal cord with a maximum interelectrode distance of 3 mm. Stimulus strengths were kept at threshold (usually < 100 #A) for evoking the effect and were raised only after cord lesions. Higher stimulus strengths were used to distinguish between an increased threshold for a response and a complete loss of the response. No attempt was made to standardize the duration of the stimulation because it was evident that once having initiated the void, the presence or absence of the stimulation did not determine the duration or pattern of the void; rather the void proceeded as necessary to empty the bladder contents. Extensive stimulus current spread did not appear to be a major factor in the experiments because movement of the electrode 0.5 mm or 0.1 mm from the optimal site within the brainstem or spinal white matter respectively, abolished the evoked coordinated bladder and sphincter activity. Since it was the coordination between the bladder and sphincter

activity during voiding and not the urodynamic details that was the focus of these experiments, the selected stimulation procedures were deemed sufficient for the purpose of the present study. At the end of the experiment stimulation sites within the brainstem and spinal cord were electrolytically lesioned and 21/29 brainstem sites as well as 7/29 cord sites were confirmed histologically. Partial lesions of the spinal cord were made under a dissecting microscope using fine scissors and jeweUer's forceps. Prior to these lesions the spinal cord was cooled with frozen saline and following the lesion the tissue was immediately rewarmed with mineral oil (37 ~ The lesions were visually verified at the end of each experiment and complete histological confirmation was done in 3 animals. Electroneurograms (ENGs) or electromyograms (EMGs) were amplified, filtered and digitized (1 or 2 KHz). The bladder pressure and stimulus marker records were also digitized (330 or 500 Hz and 1 or 2 KHz, respectively). Data acquisition and subsequent display and analysis was done using a Masscomp 5400 computer and custom software (InfoWest).

Results Electrical s t i m u l a t i o n within the p o n t i n e m i c t u r i t i o n centre (PMC) produces a b l a d d e r c o n t r a c t i o n c o o r d i n a t e d with decreased external urethral sphincter activity as depicted in Fig. 1A (also see Shefehyk 1989). The brainstem site at which such reproducible voiding was evoked using the lowest possible s t i m u l a t i o n currents was defined as the P M C for each animal. Figure 1B shows the 21 histologically confirmed P M C sites, all of which were f o u n d between P1 to P3, L1 to L4 a n d H2 to H-3 (Berman 1968). I n each case, the v o l u m e within the b l a d d e r prior to s t i m u l a t i o n was less t h a n 50% of the volume threshold for distension-evoked voiding reflexes. D u r i n g P M C - e v o k e d m i c t u r i t i o n the bladder peak pressure reached 2 0 - 4 0 m m H g (varying s o m e w h a t between animals) a n d the void typically lasted 1 0 - 4 0 s with n o m e a s u r a b l e residual volumes observed. The E U S E M G activity was e x a m i n e d prior to paralysis in two cats a n d a n obvious similarity between the E U S E M G activity a n d the E N G recording from the cut contralateral E U S b r a n c h of the p u d e n d a l nerve was observed (see Fig. 1A). I n the majority of the a n i m a l s the decrease in E U S activity appeared linked to the b l a d d e r pressure increase a n d the voiding cycle, while in the m i n o r i t y of a n i m a l s (35 %) the decreased E U S activity d u r i n g some of the voiding cycles d u r i n g the experiment appeared to be coupled with the electrical s t i m u l a t i o n a n d was m a i n t a i n e d b e y o n d the period that the bladder contracted a n d fluid was expelled. There was n o way to predict when this latter p a t t e r n would occur n o r did a n y particular m a n i p u l a t i o n (CNS stimulus site or cord lesion) a p p e a r to predispose the a n i m a l to this pattern. I n addition, in the m a j o r i t y of animals in which the two separate cut branches of the p u d e n d a l nerve i n n e r v a t i n g the E U S a n d anal sphincter muscles c o n t a i n e d multiple efferent units a n d p r o d u c e d viable recordings t h r o u g h o u t the experiment, the efferent o u t p u t to these two muscles d u r i n g evoked voiding could differ as illustrated in Fig. 2. U n i l a t e r a l s t i m u l a t i o n of the superficial aspect of the dorsolateral funiculus ( D L F ) j u s t lateral to the dorsal root entry zone at the thoracic segments T9 to T12 ( n = 19) p r o d u c e d a n increase in b l a d d e r pressure a n d decreased

637

A

/t~N ~

i 2OmmHg

EUSEMG EUSENG

I

0

I

I

20

I

l

I

40 s

Fig. 1A, B. PMC-evoked micturition and summary of PMC stimulation sites. Panel A shows the bladder pressure and EUS activity pattern produced by electrical stimulation of the PMC (200 #A). The period of stimulation is indicated with the solid bar below the last trace. The second trace is the EMG recording obtained from the left portion of the external sphincter muscle while the ENG recording was obtained from the cut right branch of the pudendal nerve

innervating the external urethral sphincter. Panel B is a schematic of the brainstem (P2) showing the 21 PMC stimulation sites which were histologically verified. Note, the section used for illustration is P2 while the dots represent sites found between P1 and P3. (IC, inferior colliculus; CNF, cuneiform nucleus; BC, brachium conjuctivum; LC, locus coeruleus)

activity in the EUS pudendal E N G during which time a void occurred (Fig. 2A). The effective stimulation current ranged from 20/~A to 200 #A and was usually less than 100 #A. The onset of the bladder pressure change was detected within 2 s following the start of the stimulus and the latency for the decreased activity in the EUS E N G was usually less than 1 s. Only in a few trials was the delay longer. However, in these cases the decreased EUS activity coincided with, or followed, the start of the bladder contraction. The peak bladder pressure and duration of the voiding cycles were similar to those reported earlier for the PMC-evoked voids. Within any animal these voiding parameters tended to remain constant throughout the experiment, and only with signs of hemorrhaging within the bladder or deterioration of the overall state of the animal did these parameters change. Figure 2 illustrates the similarity of effects produced by electrical stimulation of the superficial surface of the D L F at T12 (Fig. 2A) and at the border between the L5 and L6 (n--2) segments (Fig. 2B) in the same animal. In contrast, stimulation of the D L F at the C2 level (n = 4) failed to produce micturition. However, when the electrode was positioned in a more ventrolateral region of the cervical white matter, voiding with coordinated bladder and sphincter activity was observed (Fig. 2C, different animal than 2A and B; also see M c M a h o n and Morrison 1982c). To the right of each panel in Fig. 2 are schematic representations of the regions effective for evoking coordinated micturition at the various spinal cord levels examined as well as the regions tested which did not produce voiding. It should be noted that dorsal column stimulation at any level of the spinal

cord tested failed to produce micturition or a measurable response in either the bladder pressure or sphincter E N G recordings. The stimulating electrode at the thoracic level was always positioned so as to avoid placement on thoracic dorsal rootlets, and in 3 cats bilateral section or ligation of the mixed spinal roots within one segment rostral and caudal to the thoracic stimulation site was done. This procedure had no effect on the thoracic DLF-evoked voiding, bladder pressure changes or EUS response. In addition, stimulation of the nearest thoracic dorsal root using comparable stimulus parameters did not produce a change in bladder pressure or pudendal nerve activity. In 2 cats in which the lumbosacral (L7-$3) dorsal roots were cut bilaterally, thoracic stimulation produced coordinated voiding with no measurable residual volumes indicating that sacral segmental inputs were not determining the evoked response. As reported previously (Shefchyk 1989), the deafferentation did not alter the peak pressure attained, the duration of the bladder contraction, or the timing of the decreased EUS activity. No measurement of other urodynamic parameters (i.e. flow rate, urethral pressure) were made and therefore no comment can be made about the presence of more subtle changes. To assess whether ascending or descending pathways were producing the DLF-evoked voiding, subtotal lesions of the thoracic spinal cord were made. Typically, two monopolar stimulating electrodes were placed in the superficial aspect of the thoracic spinal cord at two different segments 2 to 3 cm apart and the effects of stimulation at each of the D L F sites were evaluated before

638 A T12

A TIO

Bladderwessure

Bladder pressure

I 20 mmHg

~LJS I=N~

o 0

10

20

30

40

$

1'o

~o

6o

1o

~o

6o

~o s

B TI= Bladder pressure

B LS/LS

EUS ENG

o

;o

~o

.o

,o

.o s

Bladder pressure

0

10

20

30

40

s

C C2 I 40rnmHg 9

|

__i

EUS ENG

6 Anal ENG ~,i ~lrT~l~

rl

r

q

rv~

Fig. 2A-C. Spinal cord-evoked micturition in the decerebrate cat. Panels A - C illustrate the bladder pressure increase and coordinated decrease in external urethral sphincter ENG evoked by electrical stimulation of the spinal white matter at T12, L5/L6, and C2 respectively. To the right of each panel is a schematic of the corresponding spinal cord section and the sites which were tested and which evoked micturition (hatched areas) or did not (-) and following a subtotal lesion of the spinal cord between the two electrodes. As illustrated in Fig. 3, lesions removing a 5 - 1 0 m m length of the dorsal columns between the electrodes (n = 9) did not disrupt the voiding evoked by stimulation of the D L F rostral or caudal to the lesion. This type of lesion verified that although the stimulation sites were superficially located, they remained effective following major trauma of the dorsal aspect of the cord. A lesion of the dorsal columns and dorsal portion of the lateral funiculus was done in 7 animals. This type of lesion, depicted in the upper right corner of Fig. 4, did not disrupt P M C - e v o k e d micturition (Fig. 4C) but did abolish distension-evoked reflex micturition in all the cats (not shown). These observations provide evidence that the D C / D L F lesions disrupt the ascending arm of the spinobulbospinal loop while leaving a significiant portion of the bulbospinal component relatively unaffected. Electrical stimulation of the D L F rostral to the bilateral D C / D L F lesion (Fig. 4D) produced a bladder

1'0

40

30

10

~0

~0

s

Fig. 3A-C. Spinal cord evoked micturition following dorsal column lesions. Panels A, B illustrate the response evoked by stimulation of the thoracic DLF at T10 (50/IA) and T12 (100 #A) respectively, prior to the lesion of the dorsal columns. Panel t2 shows the evoked responses from stimulation of the DLF caudal and rostral to the bilateral dorsal column lesion (same strengths as prelesion trials). Pressure calibration in A applies to B, C also

contraction coordinated with decreased EUS pudendal activity similar to prelesion responses (Fig. 4B) in all 7 animals tested. The stimulus strengths used to evoke a response did not change following such lesions. The delay for the onset of decreased EUS activity seen in Fig. 4D following the cord lesion was likely related to the presence of a small bladder contraction preceding the stronger contraction which produced the void. The decreased EUS activity coincided with the larger bladder contraction. This pattern of bladder pressure change was not observed in other animals, normally one uniform wave of increased pressure was observed. Figure 5 shows further observations from the same animal as illustrated in Fig. 4. Electrical stimulation of the T11 D L F caudal to the D C / D L F lesion was ineffective for producing a bladder contraction, although a decrease in EUS pudendal nerve activity was observed no fluid was expelled (Fig. 5B). The decreased E N G activity was observed within 500 ms of the stimulus onset, ranging from 100 ms to 500 ms. Determining the latency of the onset of

639

Control

Post lesion TIO

A PMCstim. ~

'

/

~i::i:~

/

C Bladderpressure

~

EUS E N n

0

10

20

30

\

40

B T9 stim. ~

s

10

0

20

30

D Bladderpressure

s

20 mmHg

EUS Elk' ~ ~k..LLL.La~L,I~L&

.........

....

0

'-;j

10

i. I. . . . . . . . . J I = , ~ L ~ a

.~I.L~JI.,,,,,L ,L. I,J&,l.-..il. ""u " I L~,J,~UL,t.

........... " " , . . . . . r- ' ' 4 ' ' ' ' '

20

30

. . . . . . . ,' . . . . . . . . r , , , ,

40

s

0

10

20

30

40

s

Fig. 4A-D. PMC and thoracic DLF evoked micturition following bilateral dorsal column and dorsolateral funiculus lesions. Panels A, B show the responses evoked from the PMC (100 #A) and T9 DLF (75/~A) respectively prior to the spinal cord lesion. Following the

DC/DLF lesion at the T10 level (upper right in panel C), micturition was evoked by stimulation rostral to the lesion, either from the PMC (panel C) or the T9 DLF (panel D). Pressure calibration in D applies to all panels

the decreased EUS activity from the onset of stimulation was often a problem because the loss of small single units within these E N G signals was difficult to detect and therefore, it is likely the reported latencies are overestimated. Stimulation of the D L F caudal to a bilateral D C / D L F lesion could also produce a decrease in ongoing hindlimb muscle nerve activity at similar latencies. The stimulus-evoked decrease in pudendal nerve activity was abolished with a second lesion of the D L F caudal to the stimulating electrode (see Fig. 5C) in each of the 4 animals subjected to the second lesion. In several cats the D L F lesions were extended to include more of the lateral funiculi. These lesions, although not always disrupting the coordination between the bladder contraction and decreased EUS E N G activity, did result in much weaker bladder pressure responses, with peak pressures rarely exceeding 10 mmHg and no voiding occurring. Since repeated lesions during the terminal phase of the experiments were likely to produce a further compromise in spinal cord blood flow, interpretation of the effects of the lateral lesions versus non-specific deterioration of the experimental preparation (often 8 h postdecerebration) were perceived as a serious limitation and such manipulations were not pursued in this series of experiments.

matter which when electrically stimulated can; 1) produce micturition characterized by bladder contraction coordinated with decreased activity in the external urethral sphincter pudendal efferents, and 2) modulate EUS pudendal nerve activity without producing micturition in the cord lesioned animal. Although D L F stimulation at the lumbar level as well as ventrolateral funiculus stimulation at the cervical level could produce micturition, the more detailed investigation of the fibres mediating the evoked responses in the present study was restricted to stimulation and lesions within the thoracic spinal segments. Stimulation using a monopolar electrode placed superficially within the thoracic spinal white matter just lateral to the dorsal root entry zone could be extremely effective for producing micturition. It is likely that micturition produced by electrical stimulation of the D L F reflects activation of ascending fibres which can influence brainstem micturition regions in a manner which results in coordinated voiding. We conclude that ascending fibres are involved in the evoked micturition because of the observations that bilateral lesions of the D L F rostral to the spinal cord stimulation site eliminated voiding responses while similar lesions caudal to the stimulation site did not. It is known that a functionally important portion of the ascending information originates in the sacral spinal cord (i.e. pelvic afferent information from bladder; for reviews see Kuru 1965; de Groat 1975) and it is likely that at least a portion Of the ascending fibres conveying this sensory information were activated by the spinal cord stimulation in the DLF. Kuru (1965) described two main ascending bladder specific pathways, one in the dorsal columns (Yamamoto

Discussion

Based on the observations reported in this study we propose that there are multiple pathways within the dorsolateral funiculi of the thoracic spinal cord white

640

A

T9

T13

U

U

1

B

2

Post lesion 1

Bladder pressure

20 mmHg I

EUS ENG

C

Post lesion 1 + 2

Bladder pressure FI IR PN~

0

10

20

30

s

40

Fig. 5A-C. Alterations of DLF-evoked responses following subtotal spinal cord lesions. Panel A is a schematic illustrating the levels of the lesions and the stimulation level in the same animal as illustrated in Fig. 4. Panel B shows the response following a bilateral dorsal

column and DLF lesion at T10 (lesion 1); electrical stimulation of the T11 DLF caudal to the lesion failed to evoke a bladder response even with a stimulation current of 200/~A but a decrease in the EUS pudendal ENG (75 #A) was produced. Panel C shows the effectsof a second lesion of the DLF at T12 caudal to the T11 stimulus site (lesion 2); stimulation currents of 200 and 500 #A no longer produced the decreased efferent firing in the EUS pudendal ENG

et al. 1956) and a second in the lateral funiculi (Kuru 1956). The micturition observed in the present experiments was not due to current spread and subsequent activation of the dorsal column fibres since lesions of the dorsal columns did not disrupt P M C or cord-evoked micturition. In fact direct stimulation of the dorsal columns alone did not produce voiding. Fibres which ascend in the dorsolateral funiculus and which display changes in activity related to the state of the bladder have been reported (McMahon and Morrison 1982a; Kamikawa et al. 1962) and likely overlap with the sacrobulbar pathways Kuru (1965) discussed. Such pathways are candidates for the ascending fibres mediating the spinal cord-evoked micturition in the present study. Other investigators have described ascending sensory tracts in the dorsolateral funiculus of the spinal cord (Menetrey et al. 1982; Apkarian et al. 1985; Jones et al. 1985; Hylden et

al. 1986) or in the cervical ventrolateral funiculus (Yezierski 1988) but the role of these tracts in micturition reflexes or bladder sensation has not been directly examined. In the present study an effort was made to determine if thoracolumbar segmental afferents or efferents might be mediating the evoked responses. One possibility considered was that the electrical stimulation could have activated lumbar/thoracic segmental afferents and resulted in the activation of sympathetic efferents. Since bilateral ligation or section of the mixed roots immediately rostral and caudal to the thoracic stimulation site did not alter the results, it was felt that activation of segmental thoracic efferents was not mediating the micturition. In addition, stimulation of the thoracic dorsal rootlets which may activate thoracolumbar sympathetic outflow, failed to produce micturition although D L F stimulation at the same thoracic segment did. These observations are consistent with previous reports (Elliot 1907; de Groat and Saum 1971; McMahon and Morrison 1982a) in which activation of the sympathetic efferents functioned in a manner consistent with continence reflexes and not the expulsion phase of voiding. A second possibility considered was that the electrical stimulation activated nonspecific ascending sensory fibres which somehow facilitated the responses observed. McMahon and Morrison (1982a, b) reported that afferents from either sacral parasympathetic or sympathetic systems could activate nonspecific ascending pathways which in turn accessed brainstern regions that then sent a signal to the sacral cord. In their synopsis, either micturition or defecation occurred depending on a "gating mechanism" determined by the pattern of sacral segmental pelvic afferent input (i.e. rectal versus bladder distension; Floyd et al. 1982; McMahon and Morrison 1982c). Although the electrical stimulation of the D L F may have activated ascending units such as they described, we do not feel that the D L F stimulation activated an ascending system which had nonspecific sacral spinal cord actions. The electrical stimulus-evoked micturition was observed with a fluid volume in the bladder less than 50% of the volume required to produce a reflex void. It was assumed that this amount of distension, being insufficient to initiate reflex micturition, would not activate any sacral gating mechanisms. In addition, deafferentation of the sacral segments by transecting the dorsal roots was performed and had no effect on the cord-evoked micturition. This observation strengthens our contention that segmental afferent input played a minor role in the present experiments. As described in the results, lesions of the D L F rostral to a thoracic spinal cord stimulation site abolished the evoked micturition but revealed a stimulus evoked decrease in tonic pudendal nerve activity. Bilateral lesions of the dorsolateral portion of the spinal white matter result in a release of spinal reflex pathways (Holmqvist and Lundberg 1959; Engberg et al. 1968; also Downman and Hussain 1958) due to the removal of descending pathways which influence spinal neurons. Stimulation of D L F inhibits polysynaptic reflexes influencing motoneurons innervating hindlimb muscles (Holmqvist and Lundberg 1959) and these actions are thought to be mediated by the

641 dorsal reticulospinal system (Engberg et al. 1968). The DLF-evoked decrease in pudendal nerve activity following subtotal cord lesions observed in the present study may be mediated by mechanisms similar to those influencing polysynaptic hindlimb reflexes. We have observed that following bilateral D C / D L F cord lesions, D L F stimulation caudal to the lesion decreased the polysynaptic volleys in the pudendal muscle nerves evoked by stimulation of the sensory branch of the pudendal nerve (Fedirchuk and Shefchyk unpublished observations). Such reflex depression would be consistent with the D L F evoked decrease in tonic pudendal nerve activity observed in the present study since the tonic activity is believed to be due to polysynaptic reflex activation of the pudendal motoneurons (Garry et al. 1959; Bradley and Teague 1977; Mackel 1979). One can hypothesize that within the D L F there are descending fibres which can modify segmental polysynaptic pudendal reflexes. Furthermore, the decrease in pudendal nerve activity produced by this pathway following dorsal cord lesions appears to be independent of the descending pathway mediating the decreased activity of the EUS pudendal nerve observed during micturition. Bilateral D L F lesions did not result in dyssynergic voiding patterns during evoked micturition; bladder contractions coordinated with decreased EUS pudendal activity were maintained after such lesions. In addition to the hypothesized descending pathways, one cannot rule out the possibility that the electrical stimulation antidromically activates ascending fibres of neurons with segmental collaterals which could influence the segmental circuitry controlling pudendal efferent activity. Ascending fibres which have segmental collaterals have been described (Rastad et al. 1977; Bras et al. 1988) and at least one population, the spinocervical axons, do travel within the D L F (Loewy 1974) and therefore could mediate changes in segmental circuitry and pudendal efferent output. The possibility that segmental afferents activate the tract neurons with collateral inhibitory actions directly onto pudendal motoneurons is unlikely since intracellular recordings from pudendal motoneurons have not revealed significant inhibitory synaptic input produced by peripheral afferent stimulation (Fedirchuk et al. 1990). However, one cannot exclude the possibility that activated axons collaterals of ascending tract cells could influence segmental interneurons interposed in a pathway to the pudendal motoneurons and exert their actions premotoneuronally. To date, such possibilities have not been examined in detail for cat lumbosacral reflex systems. During the present experimental series it was observed that bilateral lesions of the D L F abolished reflex distension-evoked voiding but P M C or rostral cord stimulation continued to evoke coordinated voiding. In cases where the D L F lesions were extended to include more of the lateral funiculi, the bladder and EUS coordination remained but the magnitude of the bladder contractions were decreased and often insufficient to expel fluid. A number of explanations can be proposed beyond the secondary effects previously discussed in the Results section. The lesions may have disrupted a portion of the descending micturition specific pathway and resulted in an incomplete activation of sacral neurons. It is also possible

that within the D L F there are facilitatory pathways, either tonic or phasic, akin to those proposed by M c M a h o n and Morrison (1982a, b, c) which set the level of activity for sacral neurons. With disruption of these fibres one might expect to see a less effective activation of the sacral micturition circuitry. In the case of the bladder parasympathetic neurons, this might result in a less vigorous bladder contraction. The evaluation of such facilitatory pathways was not possible in the present experimental paradigm. The brainstem regions and spinal trajectories of descending fibres examined by Holstege and coworkers (Holstege and Kuypers 1982; Holstege et al. 1986; Holstege and Tan 1987) do at least partially overlap with the regions stimulated in the present study. The brainstem connections of the P M C site which in turn may project to the spinal cord have not yet been described. Such projections may travel to areas not directly examined by Holstege and coworkers in their work which focused on direct connections to the spinal cord. The presence and relative importance of the direct and indirect pathways from the P M C to the sacral spinal cord and their relation to the results presented in the present study remain to be determined.

Acknowledoements. This work was funded by grants from the Medical Research Council of Canada to SJS. BF is a recipient of a Manitoba Health Research Council studentship. The authors gratefully acknowledge M. Setterbom for her unfailing assistance and Dr. D. McCrea for his helpful comments.

References Apkarian AV, Stevens RT, Hodge CJ (1985) Funicular location of ascending axons of lamina I cells in the cat spinal cord. Brain Res 334:160 164 Barrington FJF (1921) The relation of the hind-brain to micturition. Brain 4:23 53 Barrington FJF (1925) The effect of lesion of the hind- and mid-brain on micturition in the cat. Q J Exp Physiol 15:81-102 Berman AL (1968) The brain stern of the cat: a cytoarchitectonic atlas with stereotaxic coordinates. The University of Wisconsin Press, Madison Wisc Bradley WE, Teague CT (1977) Synaptic events in pudendal motoneurons of the cat. Exp Neurol 56:237-240 Bras H, Cavallari P, Jankowska E (1988) Demonstration of initial axon collaterals of cells of origin of the ventral spinocerebellar tract in the cat. J Comp Neurol 273:584 592 de Groat WC (1975) Nervous control of the urinary bladder of the cat. Brain Res 87:201 211 de Groat WC, Saum WR (1971) Adrenergic inhibition in mammalian parasympathetic ganglia. Nature New Biol 231 : 188-189 Dowman CBB, Hussain A (1958) Spinal tracts and supraspinal centres influencing visceromotor and allied reflexes in cats. j Physiol 141:489 499 Elliott TR (1907) The innervation of the bladder and urethra. J Physiol 35:367 445 Engberg I, Lundberg A, Ryall RW (1968) Reticulospinal inhibition of transmission in reflex pathways. J Physiol 194:201-203 Fedirchuk B, Shefchyk SJ (1989) Effects of suprasacral spinal cord stimulation on bladder and external urethral sphincter activity in the decerebrate cat. Soc Neurosci Abstr 251.7 Fedirchuk B, Hochman S, Shefchyk SJ (1990) Synaptic inputs onto pudendal motoneurons from sacral segmental and brainstem pathways in the cat. Soc Neurosci Abstr 48.10

642 Floyd K, McMahon SB, Morrison JFB (1982) Inhibitory interactions between colonic and vesical afferents in the micturition reflex of the cat. J Physiol 322:45-52 Garry RC, Roberts TDM, Todd JK (1959) Reflexes involving the external urethral sphincter in the cat. J Physiol 149:653-665 Holmqvist B, Lundberg A (1959) On the organization of the supraspinal inhibitory control of interneurones of various spinal reflex arcs. Arch Ital Biol 97:340-356 Holstege G, Griffiths D, deWall H, Dalm E (1986) Anatomical and physiological observations of supraspinal control of bladder and urethral sphincter muscles in the cat. J Comp Neurol 250: 449-461 Holstege G, Kuypers HGJM (1982) The anatomy of brain stem pathways to the spinal cord in cat: a labelled amino acid tracing study. Progr Brain Res 57:145-175 Holstege G, Tan J (1987) Supraspinal control of motoneurons innervating the striated muscles of the pelvic floor including urethral and anal sphincters in the cat. Brain 110:1323-1344 Hylden JLK, Hayashi H, Dubner R, Bennett GJ (1986) Physiology and morphology of the lamina I spinomesencephalic projection. J Comp Neurol 247:505-515 Jones MW, Hodge CJ Jr., Apkarian AV, Stevens RT (1985) A dorsolateral spinothalamic pathway in cat. Brain Res 335: 188-193 Kamikawa K, Matsuo S, Koshino K, Kuru M (1962) Analysis of lateral column units related to vesical reflexes. Exp Neurol 6: 271-284 Kuru M (1956) The spino-bulbar tracts and the pelvic sensory vagus: further contributions to the theory of the sensory dual innervation of the viscera. J Comp Neurol 104:207-231 Kuru M (1965) Nervous control of micturition. Physiol Reviews 45: 425 -494 Loewy AD (1974) Measurement of nerve fiber diameters in the region of the dorsal spinocerebellar and spinocervical tracts in the cat. J Comp Neurol 155: 231-250. Loewy AD, Saper CB, Baker RP (1979) Descending projections from the pontine micturition center. Brain Res 172:533-538 Mackel R (1979) Segmental and descending control of the external urethral and anal sphincters in the cat. J Physiol 294:105-122

Martin WD, Fletcher, TF, Bradley WE (1974) Innervation of feline perineal musculature. Anat Rec 180:15-30 McMahon SB, Morrison JFB (1982a) Spinal neurones with long projections activated from the abdominal viscera of the cat. J Physiol 322:1-20 McMahon SB, Morrison JFB (1982b) Two groups of spinal interneurones that respond to stimulation of the abdominal viscera of the cat. J Physiol 322:21-34 McMahon SB, Morrison JFB (1982c) Factors that determine the excitability of parasympathetic reflexes to the cat bladder. J Physiol 322:35-43 McMahon SB, Spillane K (1982) Brain stem influences on the parasympathetic supply to the urinary bladder of the cat. Brain Res 234:237-249 Menetrey D, Chaouch A, Binder D, Besson, JM (1982) The origin of the spinomesencephalic tract in the rat: an anatomical study using the retrograde transport of horseradish peroxidase. J Comp Neurol 206:193-207 Nathan PW, Smith MC (1958) The centrifugal pathway for micturition within the spinal cord. J Neurol Neurosurg Psychiat 21: 177-189 Rastad J, Jankowska E, Westman J (1977) Arborization of initial axon collaterals of spinocervical tract cells stained intracellularly with horseradish peroxidase. Brain Res 135:1-10 Shefchyk SJ (1989) The effects of lumbosacral deafferentation on pontine micturition centre-evoked voiding in the decerebrate cat. Neurosci Lett 99:175-180 Stewart CC (1899) On the course of impulses to and from the cat's bladder. Am J Physiol 2:182-202 Sugaya K, Matsuyama K, Takakuskai K, Mori S (1987) Electrical and chemical stimulations of the pontine micturition centre. Neurosci Lett 80:197-201 Yamamoto S, Sugihara S, Kuru M (1956) Microelectrode studies on sensory afferents in the posterior funiculus of cat. Jap J Physiol 6:68-85 Yezierski RP (1988) Spinomesencephalic tract: projections from the lumbosacral spinal cord of the rat, cat, and monkey. J. Comp Neurol 267:131-146