primary merent depolarization of muscle merents elicited ... · ( 1.2- 1.4 xt) produced a small...

JOURNALOF NEUROPHYSIOLOGY Vol. 70, No. 5, November 1993. Printed in C’.S..I

Primary Merent Depolarization of Muscle Merents Elicited by Stimulation of Joint Merents in Cats With Intact Neuraxis and During Reversible Spinalization

J. QUEVEDO, J. R. EGUIBAR, I. JIMENEZ, R. F. SCHMIDT, AND P. RUDOMIN Department of Physiology, Biophysics, and Neurosciences, Centro de Investigaci6n y de Estudios Avanzados de1 IPN, Mexico City 07000; Department of Physiological Sciences, Institute de Ciencias, Universidad Autdnoma de Puebla, Puebla 725 70, Mexico; and Institute of Physiology, University of Wiirzburg, D-8700 Wurzburg, Germany

SUMMARY AND CONCLUSIONS

1. In the anesthetized and artificially ventilated cat, stimulation of the posterior articular nerve (PAN) with low strengths ( 1.2- 1.4 XT) produced a small negative response (N, ) in the cord dorsum of the lumbosacral spinal cord with a mean onset latency of 5.2 ms. Stronger stimuli (> 1.4 XT) produced two additional components (N, and N3) with longer latencies (mean latencies 7.5 and 15.7 ms, respectively), usually followed by a slow positivity lasting lOO- 150 ms. With stimulus strengths above 10 XT there was in some experiments a delayed response (N,; mean latency 32 ms).

2. Activation of posterior knee joint nerve with single pulses and intensities producing N, responses only, usually produced no dorsal root potentials ( DRPs), or these were rather small. Stimula- tion with strengths producing N2 and N3 responses produced dis- tinct DRPs. Trains of pulses were clearly more effective than single pulses in producing DRPs, even in the low-intensity range.

3. Cooling the thoracic spinal cord to block impulse conduction, increased the DRPs and the N, responses produced by PAN stimulation without significantly affecting the N, responses. Re- versible spinalization also increased the DRPs produced by stimulation of cutaneous nerves. In contrast, the DRPs produced by stimulation of group I afferents from flexors were reduced.

4. Conditioning electrical stimulation of intermediate and high-threshold myelinated fibers in the PAN depressed the DRPs produced by stimulation of group I muscle and of cutaneous nerves.

5. Analysis of the intraspinal threshold changes of single Ia and Ib fibers has provided evidence that stimulation of intermediate and high threshold myelinated fibers in the posterior knee joint nerve inhibits the primary afferent depolarization (PAD) of Ia fibers, and may either produce PAD or inhibit the PAD in Ib fibers, in the same manner as stimulation of cutaneous nerves. In 7 / 16 group I fibers the inhibition of the PAD was increased during reversible spinalization.

6. The results obtained suggest that intermediate and high- threshold myelinated fibers in the PAN have the same actions on Ia and Ib fibers as intermediate and high-threshold cutaneous afferents and may therefore be considered as belonging to the same functional system. They further indicate that in anesthetized preparations the pathways mediating the PAD of group I fibers, as well as the pathways mediating the inhibition of the PAD, may be subjected to a descending control that is removed by spinalization.

INTRODUCTION

The possible role that joint afferents play in the control of movement has been a matter of discussion during the last few years. The presently available evidence suggests that

joint receptors do not participate in the sense of position, but have other important functions such as signaling joint movement, acting as joint limit detectors, and as nociceptors (for review see Proske et al. 1988; however, see Ferrel 1980).

Joint afferents produce flexion reflexes and have been considered as a part of the flexion reflex afferent (FRA) system (Eccles and Lundberg 1959; Holmqvist and Lund- berg 196 1; Lundberg et al. 1978). In 1963 Carpenter et al. showed that stimulation of high-threshold myelinated fibers of the posterior articular nerve (PAN) produced dorsal root potentials (DRPs) in spinal but not in decerebrate preparations. These investigators assumed that the DRPs evoked by stimulation of the joint nerve were due to primary afferent depolarization (PAD) of cutaneous and Ib fibers and that the PAD of Ia fibers was not significantly affected by articular inputs. In 1986 Schaible et al. reported that in the spinal cat stimulation of high-threshold fibers in the PAN produced a series of negative potentials in the cord dorsum that were not followed by positivity (P-waves). Be- cause the P-waves have been associated with the generation of PAD (Eccles et al. 1962), they suggested that joint afferents produce little or no PAD and presynaptic inhibition in large afferent fibers.

We now present a detailed analysis of the action of joint nerve afferents on DRPs as well as on the intraspinal threshold of single Ia and Ib afferents. Because there is evidence that the action of cutaneous and joint afferents is subject to a descending inhibitory control (Carpenter et al. 1963; Holmqvist and Lundberg 196 1 ), we also examined the effects of reversible spinal conduction block on the PAD produced by articular, cutaneous, and muscle afferents.

Our results indicate that PAN myelinated fibers with intermediate and high thresholds inhibit the PAD elicited in Ia fibers and may either produce PAD in some group Ib fibers or inhibit the PAD in other group Ib fibers, in the same manner as cutaneous afferents. The DRPs produced by cutaneous and knee joint myelinated afferents are increased significantly during reversible block of conduction in the thoracic spinal cord. Measuring intraspinal threshold changes of individual muscle afferents has provided additional evidence suggesting that both the excitatory and inhibitory actions exerted by cutaneous and knee joint inputs on the PAD of muscle afferents are modified during reversible spinalization.

0022-3077/93 $2.00 Copyright 0 1993 The American Physiological Society 1899

1900 QUEVEDO, EGUIBAR, JIMCNEZ, SCHMIDT, AND RUDOMIN

It is concluded that knee joint myelinated afferents with intermediate and high thresholds to electrical stimulation belong to the same functional category as cutaneous afferents, as initially proposed by Carpenter et al. ( 1963) and that the synaptic effectiveness of these afferents is subjected to a descending control, even in anesthetized preparations. Some of these findings have been published in abstract form (Quevedo et al. 199 1). After this work was submitted for publication Jankowska et al. ( 1993 ) published their observations on PAD of posterior knee joint nerve afferents with conduction velocities ~75 m/s. They found that PAD was induced from group I and group II muscle and cutaneous afferents, but not by stimulation of the knee joint nerves, that were however able to produce PAD in interosseus nerve afferents. Our studies complement theirs by showing that posterior knee joint nerve afferents have clear actions on the PAD of Ia and Ib fibers.

METHODS

General procedures The experiments were performed in 15 cats of either sex (2.5-

3.5 kg). During the dissection the cats were initially anesthetized with halothane, the trachea was cannulated, and polyethylene tubings were inserted into the central end of the right carotid artery to measure blood pressure and in the radial vein for the injection of fluids. At the end of the dissection, halothane was discontinued and ac-chloralose (60-70 mg/kg) or pentobarbital sodium (35 mg/kg) were given intravenously. Additional doses of anesthetic were administered throughout the experiment to maintain a deep level of anesthesia. The lumbosacral spinal cord was exposed and the ventral roots Si to L, sectioned. A small dorsal root filament ( L6) was prepared to record dorsal root potentials (DRPs, see below). In a couple of experiments small L, and L, dorsal rootlets were prepared in addition for simultaneous recordings of DRPs. A second laminectomy was performed at the midthoracic level to place the spinal cord cooling thermode.

The following nerves were dissected free, sectioned, and prepared for stimulation and / or recording : posterior biceps and semitendinosus ( PBSt ), gastrocnemius soleus (GS), sural (SU), superficial peroneus (SP), tibia1 (Tib), and the posterior articular nerve (PAN), also called posterior knee joint nerve. In those experiments aimed at measuring intraspinal excitability changes of afferent fibers, fine filaments were dissected from either the PBSt or GS nerves to allow recording of antidromic responses of single afferent fibers.

After the dissection the animal was fixed in a metal frame using spinal and pelvic clamps. The head was fixed in a stereotaxic appa- ratus. Pools were made with the skin flaps and filled with paraffin oil maintained between 36 and 37OC by means of radiant heat. The animals were paralyzed with pancuronium bromide (Pavu- lon, Organon) and artificially ventilated. Tidal volume was adjusted to have -4.5% of CO2 concentration in the expired air.

Stimulation of the PAN

Activation of low-threshold myelinated afferents (< 1.4- 1.5 XT) in the PAN was achieved with pulses 0.5 ms duration, 0.0% 0.2 V strength. Stimulation of intermediate and high-threshold myelinated afferents required higher voltages (2-4 V; see Schaible et al. 1986). In the initial series of experiments it was verified that with these stimulus intensities there was no concurrent activation of a neighboring nerve (Tib), as might be expected if there were current spread activating other nerves besides the PAN.

Central stimulation To stimulate the reticular formation (RF), the brain stem was

exposed by partial removal of the cerebellum and a tungsten monopolar electrode was introduced at 1.0-l .5 mm from the midline, 3-4 mm rostra1 to the obex, 3-3.5 mm from the surface, in the region corresponding to the gigantocellular reticular formation (see Rudomin et al. 1983, 1986). In some experiments the contralateral motor cortex was exposed and prepared for stimulation. Monopolar electrodes were placed on the posterior sigmoid gyrus, in the hindlimb region ( Landgren and Silfvenius 1969). The motor cortex was usually stimulated with trains of eight ano- da1 pulses at 700 Hz, twice the strength necessary to produce dis- tinct descending volleys in the cord dorsum (usually 20-50 V).

Recordings A silver ball electrode was placed on the cord dorsum at the

L,-L, level to record the responses produced by segmental and supraspinal stimulation [cord dorsum potential (CDP)] . The in- different electrode was placed on the paravertebral muscles. DRPs were recorded with a pair of Ag-AgCl hook electrodes. One electrode was placed on the distal end of the rootlet and the other as close as possible to the spinal cord (Barron and Matthews 1938; Eccles et al. 1962). Low-noise, high-gain differential amplifiers (band-pass filters 0.3 Hz to 10 kHz) were used to amplify CDPs and DRPs.

Intraspinal threshold changes of single aferents PAD was inferred from changes in the intraspinal threshold of

single group I afferents (Rudomin et al. 198 1, 1983; Schmidt 197 1; Wall 1958). The method employed is summarized in Fig. 1. A recording glass micropipette filled with 2 M NaCl was introduced into the spinal cord to search for the region in the intermediate nucleus with the largest extracellular field potentials produced by stimulation of the PBSt and GS nerves. Once in the optimal position, the micropipette was used to pass computer- controlled current pulses (of - 10 PA ), and recordings of antidromic responses of single afferent fibers were made from fine PBSt or GS nerve filaments. The stimulating pulses were given once per second and their intensity was automatically changed to produce antidromic responses with a constant firing index (set to 0.5). Each current pulse was integrated and the corresponding value was maintained until the next stimulating pulse. The output of the integrator was recorded by a penwriter to have a continuous estimate of the intraspinal threshold of the afferent fiber (Madrid et al. 1979). With this technique PAD of group I afferents is reflected as a reduction of the intraspinal threshold of the primary afferents, and inhibition of the PAD as a relative threshold increase (see Rudomin et al. 1981, 1983, 1986).

In several experiments we examined the threshold changes simultaneously in two group I fibers. To this end we used indepen- dent micromanipulators to place two different micropipettes within the intermediate nucleus. Once in optimal position, constant current pulses were passed through the micropipettes, and antidromic responses of two single afferent fibers were recorded from two fine nerve strands dissected either from the PBSt or GS nerves. The intraspinal threshold of the fibers was determined by alternating through each of the micropipettes computer-controlled current pulses, whose strength was automatically adjusted to keep a constant antidromic firing of the corresponding afferent fiber.

The changes in the intraspinal threshold of a single fiber produced by different stimulation paradigms were calculated as follows. It is assumed that the fiber’s resting threshold was 5 PA ( &) and that PBSt conditioning stimulation reduced this threshold to 3 PA ( Tpsst). The percentage threshold attained during PBSt conditioning stimulation would then be ( r,,,,/ TR) X 100 = ( 3 PA/ 5

PAD AND ARTICULAR AFFERENTS 1901

Penwriter: AAP + (to se8 changes in

firing index)

Penwriter: Change in stimulus p current due to conditioning I

stimuli *

COMPUTER: AAP YES: stim. current down AAP NO: stim current UD to

-1 maintain preset firing index 1 + A AAP yes/no

WINDOW DISCRIMINATOR r --.-- 3

Antidromic Action Potential

AAP

Adjustment signal

1 min

current , measurement,

I , \ 4

l current Generator

- - -

\ Conditioning Inputs eg. electrical stimulation of l cortex, cx Preparation isolated from

l Reticular Formation, RF ground (“floating” ) _ _

PA) X 100 = 60% relative to the resting threshold. Conditioning stimulation of the SU nerve had no effect on the resting threshold of Ia fibers but was able to inhibit the PAD produced by PBSt stimulation (as in Fig. 7 A). Assume that the threshold of the fiber, already lowered by PBSt stimulation, increased from 3 to 4 @A by the conditioning stimulation of SU ( T,, + PBSt). The threshold of the fiber during the concurrent stimulation of SU and PBSt, relative to the resting threshold would then be Tsu/ & = (4 PA/ 5 PA) X 100 = 80%. In other words, PBSt stimulation became less effective when tested during the background stimulation of the SU nerve. The effect of SU stimulation on the PBSt action would be calculated by [ ( 7&J + PBst - &Bst)/ TiJ X 100, which in this case was [(4 PA - 3 pA)j5 PA] X 100 = 20%.

Blockade of impulse conduction in the thoracic spinal cord A silver-plated thermode was placed over the surface of the ex-

posed spinal cord at the low thoracic level. The thermode was connected by thermally isolated tubings to reservoirs with warm (37°C) and cold (- 16°C) circulating fluid. The thermode had an

FIG. 1. Method used to estimate the intraspinal threshold changes of single group I muscle aRerents. See text for explanations.

attached thermocouple that allowed measurement of the temperature at the surface of the cord, below the cooling chamber. The cooling thermode was isolated with grease from the pool of oil. Control measurements were made with the probe connected to the warm liquid, which kept the spinal cord at 37°C. Switching to the cold mixture allowed a gradual cooling of the spinal cord surface (between 0 and -2.5 “C). The effectiveness of this procedure in blocking impulse conduction was assessed every time by recording in the lumbosacral spinal cord the descending volleys in the cord dorsum as well as the DRPs produced by constant stimulation of the reticular formation (Fig. 2). Switching to the warm liquid allowed restoration of impulse conduction in the spinal cord within a couple of minutes. This procedure proved to be quite satisfactory and was employed several times in the same experiment (see also Cervero et al. 199 1; Laird and Cervero 1990; Schaible et al. 199 1).

Data processing During the experiment cord dorsum potentials and dorsal root

potentials were averaged by means of a specially designed program

1902 QUEVEDO, EGUIBAR, JIMfiNEZ, SCHMIDT, AND RUDOMIN

A Cooling D Rewarming

F

- -,,,I -

z 2 :

1 37

OC -2.5

1 150 PV

50 ms

FIG. 2. A method to assess the effectiveness of the block of impulse conduction by cooling the thoracic spinal cord. A and D: changes in the area of the dorsal root potentials ( DRPs) produced in an L, dorsal rootlet by stimulation of the ipsilateral bulbar reticular formation (top traces) and temperature of the cooling probe (bottom traces). B, C, E, and F: cord dorsum potential (CDP; top traces) and the DRP (bottom traces) produced by stimulation of the reticular formation (8 pulses at 700 Hz, 60 PA). The arrows in A and D indicate the time at which records B-F were taken. Between C and E there is a gap of 5 min during which the cooling of the spinal cord was maintained. Note in E that blocking of the reticulospinal descending volley and of the DRP are complete. Note also that the depression of the DRPs by cooling the spinal cord is slower than the recovery during rewarming. Arrows in B show the time interval used for DRP integration in A and B.

and the results stored in the computer memory. Raw data were routinely recorded on digital tape and played back for further processing after the experiment. All traces in figures are averages of 32 responses taken at 1 Hz, unless otherwise stated.

RESULTS

Cord dorsum potentials produced by graded stimulation of the PAN

Stimulation of the posterior knee joint nerve with single pulses produced a series of negative potentials in the cord dorsum at the Ls-L, segmental level. With the lowest strengths of stimulation ( 1.2- 1.4 XT in Figs. 3A and 4A) there was only a small negative potential [N, response ac- cording to the nomenclature of Bernhard ( 1953)]. With further increase in stimulus strength additional components were recruited (N, and N, responses, see Fig. 3 A). These negative responses attained their maximal amplitude at ~4-10 XT but could grow further when stimulus strength was increased above lo-20 XT (Figs. 3A and 5 A and 0). In some experiments, when stimulus intensity was increased above 10 XT strength, a fourth negative wave ( N4) was recorded (not illustrated).

The mean onset latencies of the N,-N, responses recorded in five different experiments were of 5.2 t 0.5,7.5 t

1, 15.7 t 1.2, and 3 1.8 t 5 ms, respectively (mean t SD). The latencies of the N, and N, responses presently recorded agree well with the values reported by Schaible et al. ( 1986). However, the mean latency of the N, response recorded by these investigators was somewhat shorter (8.7 ms; range 7.5-13 ms).

In the experiment of Fig. 3 B stimulation of the PAN with a relatively low stimulus strength ( 1.7 XT), produced dis- tinctive N, and N, responses followed by a rather small P-wave. With stimuli of 2 1.8 XT strength, the N2 and N, responses were increased in size but were no longer followed by a P-wave, in confirmation of the observations of Schaible et al. ( 1986). However, in five of seven experiments, stimulation of the PAN with relatively low stimulus strengths produced N, and N, responses followed by P- waves that grew further with increasing strengths of stimulation, up to a maximum that was attained with stimuli 40- 60 XT (Fig. 4A).

Dorsal root potentials produced by stimulation of the PAN

The positive waves that follow the negative cord-dorsum potentials produced by stimulation of sensory nerves have been associated with the intraspinal depolarization of afferent fibers ( Eccles et al. 1962). However, because the magni- tude of these responses may depend on the recording site in the cord dorsum, relative to the level of entry of the stimulated afferents, it seemed desirable to record DRPs, in addition to the CDPs.

Figure 3 B shows the CDPs and the DRPs produced by PAN stimulation with several strengths. Stimulation of the PAN with single pulses 1.7 XT produced a small DRP. This response increased in size with increasing strengths up to a maximum that was attained with stimuli 27 XT (see also Fig. 4, A and B). In the experiment of Fig. 4 we recorded the DRPs produced by PAN stimulation simultaneously from small dorsal rootlets of three different segments (L,, L,, and L7), both with single shocks and with trains of stimuli. In general, stimulation of the PAN with trains of pulses produced larger DRPs than stimulation with single pulses. In this experiment the largest DRPs were obtained from the L, segmental level, both with single pulses or with trains of pulses, but DRPs of appreciable size were also recorded from the adjacent segments (L, and L7). This result fits with the known distribution of the PAN afferent fibers, which comprises mostly the L,-L, segments (Craig et al. 1988).

Efects of spinal block on DRPs

In seven experiments we found that reversible spinalization consistently increased the N, responses in the cord dorsum and the DRPs produced by PAN stimulation at wide range of intensities (Figs. 5 and 7). Typical results are illustrated in Fig. 5. In this experiment a train of pulses 5 XT produced small negative responses in the cord dorsum and a small DRP (Fig. 5 A). During spinalization, the N3 but not the N, and N, responses were increased and there was also a small increase in the DRP amplitude (Fig. 5 B). After rewarming the thoracic spinal cord the N, responses and the DRPs were again decreased (Fig. 5C).


t 4

10 ms

CDP l.7rT e DRP

FIG. 3. CDP and DRP produced by graded stimulation of the posterior articular nerve (PAN). A: CDPs evoked by stimulating the PAN with different strengths, as indicated. Stimulating pulses were of 0.5 ms duration. B: simultaneous recordings of CDPs and DRPs obtained in the same experiment but at a slower sweep.

3ma PAN stimulation with higher strengths (50 XT) pro-

duced larger N, and N, responses. Nevertheless, the N, responses remained practically unchanged following reversible spinalization whereas the N, responses were facilitated. The DRPs produced with PAN stimulation with this high stimulus strength were also increased during spinalization (Fig. 5, D-F).

In the experiment of Fig. 5, PAN stimulation with strengths up to 5 XT produced relatively small DRPs. In other experiments (n = 3) PAN stimulation with strengths as low as 2 XT produced DRPs that were also increased during spinal cold block (Fig. 7, A-C, top traces).

In agreement with what has been reported by other investigators (Burke et al. 197 1; Carpenter et al. 1963)) we found that the DRPs produced by stimulation of cutaneous afferents (with intensities in the range between 1.5 and 5 XT) were increased during spinal cord conduction block. In the

A SINGLE SHOCKS

PAN 1.2xT PAN 12xT PAN 60xT

DRP Lg.------en

L7 - +------+J-------

C TRAINS PAN 1.2xT PAN 2xT PAN 6xT

CDPk--- Lb

DRP Lg-m m

B 100

60

g

=: 40 E 2 k e 2o

experiment of Fig. 6A the DRPs were produced by stimulation of the SU nerve with pulses 1.5 XT strength. During spinalization these DRPs increased - 2.5 times ( Fig. 6 B) and were reduced again after the relief of the spinal block (Fig. 6C). The DRPs produced with stronger SU stimulation (5 XT) were also facilitated after spinalization (Fig. 6, D-F). The mean increase of the DRPs observed in three experiments during reversible spinalization was of 182 t 56% relative to control amplitude.

Unlike the DRPs produced by PAN and SU stimulation, the DRPs produced by stimulation of group I PBSt fibers were not facilitated during reversible spinalization, but depressed (to 80% in the example of Fig. 7, A and B; see also Carpenter et al. 1963). The mean change in the amplitude of the DRPs produced in three experiments by PBSt stimulation (with strengths 1.5 to 2 XT) during reversible spinalization was of 78.4 t 3.1% relative to control amplitude.

Jk-v2 IP

No \

/ -0, 0 ,a- d ‘.’ ODRP L5

/, , , ,~~ FIG. 4. Cord dorsum potentials and DRPs produced

l DRP Lg V DRP L7 by single pulse and repetitive stimulation of the PAN / nerve. CDPs were recorded at the L, level and DRPs

0 0.25 0.5 0.75 1.0 4.0 (v) from L,-L, dorsal rootlets, as indicated. A: CDPs and , I I 1 0 5 10 15 20 60 (XT) DRPs produced by single pulses (0.5 ms duration) ap-

STIMULUS INTENSITY plied to the PAN. Note that DRPs appear in L, with

- strengths as low as 1.2 XT. C: CDPs and DRPs produced by stimulation of the PAN with a train of 8 pulses at 700

100 1 . . . . . . ..f................................................ v Hz. Stimulus intensities are indicated. B and D: plot of

4

60 - amplitude vs. strength of responses partly illustrated in A and C, respectively. Note that trains of stimuli were more

g 60- effective than single pulses.

20 -

O-

ODRP L5 Ll¶ L7

0 0.25 0.5 0.75 1.0 4.0 (VI

0 5 10 15 20 60 (XT) STIMULUS INTENSITY

1904 QUEVEDO, EGUIBAR, JIMENEZ, SCHMIDT, AND RUDOMIN

A B C INTACT NEURAXIS COLD BLOCK AFTER REWARMING

CDP

DHI' -- **

0-e-a . . . . . , ..,. .r\ . . INTACT

D Eli F

PAN 50xT -iiL

20 ms

FIG. 5. Effects of spinal conduction block on the CDPs and DRPs produced by stimulation of the PAN. A-C: responses produced by stimulation of the PAN with a train of 8 pulses at 700 Hz, 5 XT. D-F: same, but with PAN stimulation 50 XT. A and D: CDPs and DRPs recorded before the spinal block (intact neuraxis). B and E: responses obtained during spinalization (cold block). C and F: responses obtained 5 min after rewarming the spinal cord. In B and E the dots show the responses obtained before the spinal block. The trace labeled Diff was obtained by subtracting the CDP recorded before spinalization from the CDP obtained during spinalization. Vertical lines indicate the N2 and NJ components of the cord dorsum response. Note that spinal block facilitates the N, response as well the DRPs. Further explanations in text.

Interaction between DRPs

INTERACTION WITH DRPS OF CUTANEOUS ORIGIN. we have shown above that stimulation of the PAN with single shocks or with trains of pulses produces appreciable DRPs in the L,-L, segments. Although the recording of DRPs

discloses the existence of PAD, it does not allow assessment of the type of fibers (Ia, Ib, or cutaneous) that are being depolarized. However, interactions between DRPs of various origins have provided useful information pertaining to the neuronal sets that mediate the PAD (Brink et al. 1984).

In five experiments we have examined the effects of reversible spinalization on the DRPs produced by separate and concurrent PAN and SU nerve stimulation. In Fig. 6A the PAN was stimulated with a train of seven pulses 700 Hz, 10 XT strength, and the SU nerve was stimulated with a single pulse 1.5 XT. With this intensity, SU stimulation produced a relatively small DRP that was only slightly affected by PAN conditioning stimulation (compare trace b with trace c-a in the lower set of records). This suggests that with these stimulus intensities there was little overlap in the pathways activated by these two inputs (see Brink et al. 1984). During cold block of the thoracic spinal cord the DRPs produced by stimulation of either the PAN or SU were almost doubled in size (Fig. 6B). The PAN-conditioned SU-DRP (trace labeled as c-a) was now smaller than the unconditioned SU-DRP (trace labeled b).

Stronger stimulation of the SU nerve (5 XT) produced larger DRPs, and these were also depressed after conditioning stimulation of the PAN (Fig. 6 D). During reversible spinalization the depression of the test SU-DRPs due to PAN stimulation was also increased (Fig. 6E). The DRPs produced by SU stimulation (with stimulus strengths rang- ing between 1.5 and 5 XT) were reduced by PAN conditioning stimulation ( lo-20 XT) to 89.2 t 5.8% relative of control amplitude (rz = 5). During spinal cord block the SU-evoked DRPs were reduced by PAN conditioning to 52.2% (n = 2). Because PAN stimulation evoked rather large DRPs, it is difficult to know how much of the DRP depression observed during the interaction represents occlusion or inhibition. INTERACTION WITH DRPS OFGROUP I ORIGIN. Conditioning stimulation of the PAN also depressed the DRPs produced

A INTACT NEURAXIS

aa

B COLD BLOCK C AFTER REWARMING

FIG. 6. Interaction between DRPs produced by stimulation of the PAN and sural (SU) nerves. A-C: PAN stimulation was a train of 8 pulses, 700 Hz, and 10 XT. SU stimulation was a single pulse 1.5 XT. D-F: the same, but stimulus strength for PAN and SU was 10 and 5 XT, respectively. A and D: responses obtained in a preparation with intact neuraxis. B and E: responses obtained during cold conduction block of the thoracic spinal cord. C and F: recovery. Each panel shows, from top

“s--

to bottom, the DRPs produced by PAN stimulation (trace a), by SU stimulation (trace b), and by concurrent stimulation of PAN and SU (trace c). The lowest pair of traces shows the unconditioned DRP produced by SU stimulation (trace b) and the PAN-conditioned SU-DRP (trace c-a). See text for further explanations.

35 ms


A INTACT NEURAXIS

PAN 2xT la/

A

b

D

a

b

PAN 1OxT /--

B COLD BLOCK C AFTER REWARMING

E

c-a

by activation of group I PBSt fibers. This is illustrated in Fig. 7, taken from the same experiment as that of Fig. 6. As shown in Fig. 7A, stimulation of the PAN with a train of pulses 2 XT produced a DRP that was about five times smaller than the DRP produced by stimulation of group I PBSt fibers ( traces a and b in Fig. 7A). Concurrent stimulation of both nerves (PAN stimulation preceding PBSt stimulation by 60 ms) markedly reduced the DRP evoked by the PBSt stimulus (trace c-a in the lowest set of records). The depression of the PBSt-DRP was significantly increased when the conditioning PAN stimulation was raised to 10 XT (Fig. 7, D-F). Reversible spinalization increased the depression of the PBSt-DRPs produced by conditioning stimulation of the PAN, both with low and high stimulus strengths (Fig. 7, B and E).

In preparations with intact neuraxis, PAN conditioning stimulation ( lo-20 XT) depressed the group I PBSt-DRPs to 76.3 t 6.4% (~1 = 6) of control amplitude. During spinalization these DRPs were reduced to 67.7 t 14.5%. It is to be noted in the experiment of Fig. 7 that during reversible spinalization the amount of depression of the test DRP appears to be larger than the increase of the conditioning PAN-DRP. This suggests that increased inhibition rather than occlusion may be responsible for the increased depression of the test PBSt-DRPs produced by PAN during reversible spinalization (see below).

Intraspinal threshold changes of single group I aferents

PREPARATIONS WITH INTACTNEURAXIS. Analysis ofthein- traspinal threshold changes of single afferents allows identi- fication of the types of afferent fibers that are depolarized by segmental and descending inputs (Rudomin et al. 198 1, 1983, 1986; for a review of earlier literature see Schmidt 197 1 and Burke and Rudomin 1977 ) . In anesthetized preparations with intact neuraxis, group I muscle afferents have been shown to have three different PAD patterns (Rudo- min et al. 1986). Most group Ia fibers from flexors and

FIG. 7. Interaction between DRPs produced by PAN and posterior biceps and semitendinosus ( PBSt ) stimulation. Same format and experiment as that of Fig. 6. A-C: DRPs produced by PAN stimulation (train of 8 pulses at 700 Hz, 2 XT strength) and PBSt simulation (train of 4 pulses at 300 Hz, 2 XT strength). D-F: the same but PAN stimulus strength was of 10 XT.

35

extensors have a type A PAD pattern. They are more effec- tively depolarized by stimulation of group I fibers from flexors and also by stimulation of vestibulospinal fibers, but not by stimulation of cutaneous, reticulospinal, rubrospinal and corticospinal fibers that instead inhibit the PAD produced in these fibers by PBSt or by vestibulospinal stimulation. Group Ib fibers have type B and type C PAD patterns. They are strongly depolarized by other group Ib fibers and by supraspinal stimulation. Quite interestingly, cutaneous fibers may either produce PAD (type B PAD pattern) or inhibit the PAD of Ib fibers (type C PAD pattern; see Rudo- min et al. 198 1, 1983, 1986). It has been suggested that the action of cutaneous fibers on the PAD of Ib fibers depends on the balance of excitatory and inhibitory descending inputs that act on the segmental interneurons mediating the PAD (Rudomin et al. 1986; Rudomin 1990, 199 1).

Figure 8A shows the intraspinal threshold changes typical of a group I fiber with a type A PAD pattern. It may be seen that the intraspinal threshold of this fiber was practically unaffected by stimulation of the motor cortex with a train of eight pulses 700 Hz of 20 and 30 V, or by stimulation of the SU and SP nerves with single pulses of 11 and 13 XT, respectively. Stimulation of the PAN with pulses 14 XT (2 V) also failed to significantly change the intraspinal threshold of the fiber. On the other hand, stimulation of the PBSt nerve with a train of four pulses at 300 Hz and 3.5 XT strength produced a sustained reduction of the intraspinal threshold of the fiber (from ~4 to 3 PA). When tested against the background PAD produced by PBSt stimulation, all of the above stimuli produced a relative increase in the fibers threshold. That is, they inhibited the PAD induced by PBSt stimulation (Rudomin et al. 1983, 1986).

Figure 8 B shows one example of a group I fiber with a type B PAD pattern. In this fiber, stimulation of the SP, SU, and PAN, as well as stimulation of the bulbar reticular formation (RF) reduced the intraspinal threshold of the fiber. Stimulation of the PBSt nerve, in this case with a train of pulses 3 XT strength, also reduced the intraspinal threshold


A TYPE A

PBSt 3.5xT cx cx SP SU PAN

3oV G = E = CX Cx PAN PAN SU 3

-- 21

3ov 2OV 7xT 14xT 1lxT 4

3

B TYPE B

PBSt 3xT SP SU PAN RF ---

11

9

7

C TYPE C

PBSt 3.5xT cx cx SP su 18

15 ac

12

PA I. 1 min

FIG. 8. Effects of PAN stimulation on the intraspinal threshold of single group I muscle afferents with different PAD patterns. A-C: effects of segmental and supraspinal stimulation on the intraspinal threshold of two single PBSt afferent fibers (A and B) and on a single gastrocnemius soleus (GS) afferent fiber (C). PBSt stimulus was in all cases a train of 4 pulses at 300 Hz applied 25 ms before the threshold testing pulse. SU and superficial peroneus (SP), one pulse applied 45 ms before the threshold testing pulse; PAN, Cortex (Cx), and reticular formation (RF) were stimulated with a train of 8 pulses at 700 Hz preceding the threshold testing pulse by 75 ms. See text for explanation of the primary afferent depolarization (PAD) patterns. Note that the effects produced by PAN stimulation are in the same direction as those produced by cutaneous afferents (SU and SP).

of the fiber, and this effect was not reverted by any of the above segmental and descending inputs. In fact, these inputs produced no additional threshold reduction when tested against the background PAD produced by PBSt stimulation, as expected if there were complete occlusion between the actions produced by these inputs. When the PBSt was stimulated with lower strengths and there was a smaller PAD, the effects produced by stimulation of the SU, SP, PAN, RF, and cortex occluded with those produced by PBSt stimulation.

Finally, Fig. 8C illustrates the intraspinal threshold changes produced by PAN-conditioning volleys in another group I fiber. In this case stimulation of the cortex reduced the intraspinal threshold of the fiber and this effect occluded with the effect produced by PBSt stimulation. On the other hand, stimulation of the SU produced practically no threshold changes but stimulation of the SP and PAN produced a relatively small threshold reduction. However, both SP and PAN reverted the effect produced by PBSt stimulation, suggesting that this fiber had a type C PAD pattern (see Rudomin et al. 1986).

Figure 9 summarizes data from the experiments performed in cats with intact neuraxis. Because PBSt and GS fibers behaved similarly, all the data have been pooled together. Figure 9, A, C, and E, shows the mean threshold as percentage of control obtained during each of the condi-

tioning procedures (indicated at the bottom of the figure). Figure 9, B, D, and F, shows the changes in the PAD due to the additional conditioning with cutaneous, PAN, or descending stimulation. These changes are expressed as the percentage increase or decrease in the threshold of the fiber, already reduced by PBSt conditioning stimulation (see METHODS). Thus, a positive value means reduction of the PAD, whereas a negative value implies increased PAD (Ru- domin et al. 1983, 1986). Figure 9, A, C, and E, illustrates the basic features of the various types of PAD patterns. Be- cause the number of fibers with type B and type C PAD patterns is rather small, the statistical significance of the differences between the effects produced by the various inputs cannot be assessed. However, it must be pointed out that the observed mean intraspinal threshold changes are in good agreement with previous data (Rudomin et al. 1983, 1986), namely, that SP and SU stimulation produce a larger threshold reduction in fibers with type B than in fibers with type A or C PAD pattern and that descending stimulation reduces the intraspinal threshold of group I fibers with type B and C PAD patterns but has very little effect on fibers with type A PAD pattern.

The more striking differences in the PAD patterns appear when considering the effects of SP and SU conditioning

A B

C D 6

100 z 1

.,............................................................. T

I 6 4

-8 100

- .s 90 - 1

- 6 3

- 2 01 x-

-d”

- -4 5

-3 :

- v QO- . Y

0 80. 02

Jv) -2x-

g$ 4 4 70- -0:w

to= Wa

a* 60. --2 g;

x0 c 50 - - -4 5

W - -6 0

40 -

E F

PBSt SU SP PAN Cx RF 3 a m v) G

FIG. 9. Summary histograms of intraspinal threshold changes of group I afferents produced by segmental and descending inputs. A, C, and E: percentage changes (*SD) of the intraspinal threshold of single GS and PBSt Ia and Ib fibers produced by segmental and supraspinal stimulation. B, D, and E: effects of conditioning stimulation on the threshold changes produced by the indicated inputs. Inhibition of PAD is expressed as a positive change and facilitation of PAD as a negative change. Number of fibers tested is indicated in each column. NT, not tested. Stimulation paradigms same as in Fig. 8. Further explanations in text.


stimulation on the PAD elicited by PBSt stimulation. These conditioning inputs inhibit the PAD of group I fibers with type A and C PAD patterns (Fig. 9, B and F) but have a minor effect on the PAD of fibers with a type B PAD pattern (Fig. 9 D). It is to be noted that the effects produced by stimulation of the PAN were in the same direction as those produced by SU and SP stimulation. It thus seems fair to conclude that medium and high threshold myelinated afferents in the posterior knee joint nerve have similar actions on the PAD of Ia and Ib fibers as intermediate and high- threshold cutaneous afferents. EFFECTS OF REVERSIBLE SPINALIZATION. Analysis of the changes produced in the DRPs during reversible spinalization allows disclosure of descending influences acting onto the segmental pathways that mediate PAD. However, it does not provide insights on how these descending influences act on the pathways mediating the PAD of specific types of afferents. This was examined by studying the effects produced by reversible spinalization on intraspinal threshold changes of single Ia and Ib fibers produced by several inputs. Altogether, we examined 16 group I fibers in

A CONTROL

B COLD BLOCK

RFPANSU SP

6.0

4.5

3.2

7.0

5.2

3.1

6.0

4.5

3.2

v . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . * . . . . . . . o . . . . . . . . . . . . . . . . . . . . . . . . . . . ‘u\ ww I:-:

C REWARMING 6.0

4.5

' 3.2

FIG. 10. Effects of reversible spinalization on the PAD of pairs of afferent fibers. Intraspinal threshold changes produced by segmental and descending stimulation were determined simultaneously in 2 afferent fibers. A: control. B: during conduction block of thoracic spinal cord. C: after rewarming. Top and bottom traces: intraspinal threshold changes. Same format as that of Fig. 8. Stimulation procedures are indicated. PBSt stimulation was a train of 4 pulses 300 Hz, 2 XT applied 25 ms before the threshold testing pulse. SU and SP, one pulse applied 50 ms before the threshold testing pulse; PAN, Cortex, and RF were stimulated with a train of 8 pulses at 700 Hz preceding the threshold testing pulse by 75 ms. Further explanations in text.

TABLE 1. Changes in the PAD of single group Ia and IbJibers during reversible spinalization

PAD pattern

PAD Produced by PBSt Inhibition of PAD Produced by Stimulation SU, SP, or PAN conditioning

No No Decreased effect Increased Increased effect Decreased

Type A 318 4/8 l/8 418 218 218 Type B 10 l/2 O/2 O/2 212 o/2 Type C 416 l/6 l/6 316 l/6 216 Total 8/16 6/16 2/16 7116 5/16 4/16

PAD was inferred from the changes in the intraspinal threshold of the fiber. Experiments were performed as in Fig. 10. Stimulus intensity ranges were as follows: PBSt, 1.2% 12.5 XT; PAN, 1.15-66 XT; SU, 6-50 XT; and SP, 12-25 XT. Each box gives the number of fibers showing a particular behavior relative to the total number of fibers tested. PAD, primary afferent depolarization; PBS& posterior biceps and semitendinosus; SU, sural; SP, superficial peroneus; PAN, posterior articular nerve.

detail: 8 fibers had a type A PAD pattern, 2 fibers a type B PAD pattern, and 6 fibers a type C PAD pattern.

Figure 10 illustrates the effects produced by reversible spinalization simultaneously on the intraspinal threshold changes of a pair of Ia fibers from the same PBSt nerve filament. The data of Fig. 1OA were taken before the spinal conduction block and show that stimulation of the SU, SP, and PAN as well as RF stimulation had practically no effects on the resting intraspinal threshold of these two fibers. However, stimulation of the PBSt nerve was able to reduce the threshold of both fibers. Conditioning stimulation of the PAN inhibited the PAD produced by PBSt stimulation. The inhibition was barely noticeable with pulses 1.1 XT strength but was very clear with pulses 24 XT strength. On the other hand, stimulation of the SU ( 8 XT), SP (4 XT), and RF (with pulses 240 PA strength) produced a rather small inhibition, suggesting that both fibers had a type A PAD pattern and were most likely from muscle spindles. Figure 10 B shows the effects produced by these same stimuli while impulse conduction in the thoracic spinal cord was blocked by cooling. The resting threshold of both fibers was virtually unchanged and was not affected by SU, SP, and PAN stimulation. Stimulation of the PBSt nerve with the same strength used before spinalization reduced the intraspinal threshold of the fiber to about the same extent, but there was a significant increase in the inhibition produced by PAN, SU, and SP stimulation. This was particularly clear when testing the effects of PAN with low stimulus strengths ( 1.1 XT), but was also seen with SU and SP stimulation. After rewarming the spinal cord, the inhibition of the PBSt-PAD produced by SU and SP stimulation was again rather small or absent in both fibers. The inhibition produced by PAN was also decreased after removal of the conduction block (Fig. 1OC).

The results obtained in the 16 afferent fibers during reversible spinalization are summarized in Table 1. During reversible spinalization the PAD produced by PBSt stimulation was reduced in 8 / 16 fibers, increased in 2/ 16 fibers, and remained the same in 6 / 16 fibers. The net effect would be a small reduction of the PBSt-induced PAD, which seems to be in agreement with the diminution of group I DRPs produced during reversible spinalization (Fig. 7).


The inhibition that SU, SP, and PAN stimulation produces on the PBSt-induced PAD was increased after reversible spinalization in 4/ 8 fibers with a type A PAD pattern, decreased in 2 / 8 fibers and remained the same in 2 / 8 fibers. In fibers with a type C PAD pattern the SU, SP, and PAN induced inhibition was increased in 3 / 6 fibers, decreased in 2/6, and remained the same in l/6 fibers.

DISCUSSION

Stimulation of articular aflerents produces PAD

The N1 response produced by electrical stimulation of the PAN appears to be generated by the activation of the largest myelinated afferent fibers with conduction velocities between 43 and 87 m/s (Schaible et al. 1986). Fibers within this range of conduction velocities appear to have diameters between 8 and 17 pm and to arise from Vater Paccinian and Golgi encapsulated endings, which subserve slow adapta- tion receptors discharging at most joint angles with an out- ward tibia1 twist (Burgess and Clark 1969; Freeman and Wyke 1967). The N2 response is generated by medium- sized afferents with conduction velocities between 33 and 53 m/s ( Schaible et al. 1986) and diameters between 5 and 8 pm, which may originate from Ruffini endings that discharge only at or near the extremes of flexion and extension (Burgess and Clark 1969). Finally, the N3 response is generated by the smallest myelinated fibers with conduction velocities (27 m/s (Schaible et al. 1986), probably with axon diameters between 2 and 5 pm. These afferents appear to originate from free endings that are probably nociceptors responding only to bending and twisting procedures considered noxious (Boyd and Kalu 1979; Burgess and Clark 1969; Skoglund 1956).

In confirmation of Schaible et al. ( 1986), we have seen that in some experiments stimulation of the PAN nerve, even with high stimulus strengths, produced negative potentials in the cord dorsum that were not followed by a positive potential. However, in many experiments, particularly when using trains instead of single pulses, stimulation of the PAN nerve produced these positive potentials even with the weakest stimuli. It is possible that large negative cord dorsum responses obscure the detection of the positive potential. Also, it may be a matter of the segmental level of recording of these potentials relative to the entry level of the joint afferents (E. Jankowska, personal communication).

Simultaneous recordings of CDPs and DRPs have given a clear indication that regardless of whether or not PAN stimulation produced negative cord dorsum responses that were followed by positive potentials, DRPs were indeed generated, particularly with trains of stimuli above 1.5 XT, but sometimes even with lower strengths. Nevertheless, the largest contribution to the DRPs results from activation of medium and small myelinated fibers in the intermediate and high-threshold range that also produce the N, and the N, responses. Therefore, it may be concluded that these DRPs are due to activation of afferents activated by ex- treme knee joint movements that are not necessarily noxious as well as by afferents that are indeed activated by noxious stimulation.

Articular aferents have similar actions as cutaneous aferents on pathways producing PAD of group Ijibers

We have found that stimulation of intermediate and high-threshold PAN myelinated fibers may depress the DRPs produced by group I muscle afferents as well as the DRPs produced by cutaneous afferents with thresholds > 1.5 XT. If it is considered that stimulation of the PAN also produced DRPs, the depression of the test DRPs can be ascribed to occlusion in the pathways producing PAD (Brink et al. 1984), to presynaptic inhibition of the afferent fibers producing the test DRP ( Eccles et al. 1962), and/or to inhibition exerted on the interneurons mediating the PAD (Lund et al. 1965; Rudomin et al. 1983, 1986).

Studies presented here on the changes in the intraspinal threshold of single group I afferents from muscles have indicated that stimulation of the PAN nerve inhibits the PAD of Ia fibers and may either produce PAD in a fraction of Ib fibers or inhibit the PAD in other Ib fibers. These inhibitory actions could contribute significantly to the PAN-induced depression of the DRPs produced by stimulation of group I afferents. Thus it seems that PAN afferents behave in the same manner as cutaneous afferents, particularly those fibers activated with stimulus strengths > 1.5 XT, i.e., with strengths producing the N, and N3 responses.

Is there a descending inhibitory control on the segmental actions of joint receptors?

During spinal cord block the N, responses are practically unaffected and there is a large increase of the N, responses (Fig. 5 ). These negative cord dorsum responses appear to be generated by synaptic activation of spinal cord interneurons (see Willis and Coggeshall 199 1). Therefore, it is suggested that in the anesthetized preparation with intact neuraxis there is a tonic descending inhibition that affects the interneurons responding mostly, if not exclusively, to noxious activation of joint afferents. Spinalization would remove this descending inhibition and lead to the increased N3 responses. This finding fits well with the recent observations of Cervero et al. ( 199 1) and of Schaible et al. ( 199 1) where it is shown that reversible spinalization leads to a significant increase in the resting discharges as well as in the responses of spinal interneurons after noxious compression of the knee and the adjacent thigh and lower leg.

The increase in the PAD produced by cutaneous afferents during spinalization (Fig. 6) confirms previous findings indicating that the pathways mediating the PAD elicited by flexor reflex afferents (FRA) are under a tonic descending inhibitory control even in anesthetized preparations (Holmqvist and Lundberg 196 1). Because the DRPs produced by medium and high-threshold knee joint afferents was also increased during spinalization, inclusion of knee joint afferents with thresholds > 1.5 XT within the FRA category seems fully justified (see Lundberg et al. 1978). In this regard it is very interesting to note that spinalization also facilitated the inhibitory actions of cutaneous and knee joint inputs on the PAD of a significant fraction of Ia and Ib fibers (Fig. 10 and Table 1). This could be indica- tive of the removal of a descending inhibition acting on the pathway from cutaneous and PAN fibers to group I afferents with tvpe A and C PAD patterns.


The large variability of the effects of functional spinalization on the PAD produced by PBSt stimulation and on the inhibition of this PAD after conditioning stimulation of cutaneous and joint afferents is reminiscent of the varia- tions in the tonic suppression of reflex transmission in low spinal cats after additional lesions of the ipsilateral dorsolat- era1 funiculus at the L,-L, level (Cavallari and Petersson 1989), which may be related to the “state” of the preparation, namely the amount of tonic activation of specific sets of propriospinal neurons. It may also reflect the high degree of differentiation in the control of transmission from cutaneous and knee joint receptors to group I afferents (see Jankowska et al. 1993).

Schaible et al. ( 199 1) have found that during acute in- flammation of the knee there is increased effectiveness of the descending inhibition. They suggested that the inflam- mation-evoked hyperexcitability of spinal neurons to the afferent input from the inflamed knee, as well as to the input from regions remote to the inflamed joint, is counter- acted by a progressive enhancement of the descending inhibition. Our data complement their observations by showing that the segmental pathways, which are activated by electrical stimulation of both intermediate and high-threshold myelinated fibers in the PAN and lead to the generation of the N, responses in the cord dorsum as well as to PAD, are also subjected to a descending inhibitory input that is sup- pressed by spinalization. It is not clear, however, if the increase in these responses during spinal block is due to the interruption of a supraspinal inhibitory loop that requires afferent input to be active or to the suppression of a tonic descending inhibitory influence that may not require afferent inputs for its expression.

Recent findings of Schaible et al. ( 199 1) show that spinalization increases background activity and the evoked responses of spinal interneurons without reversing the sign of these responses. In agreement with this we found that spinalization does not appear to change the PAD patterns of single muscle afferents. Fibers with a type A PAD pattern re- main with the same pattern during spinalization (Fig. lo), and the same can be said pertaining to fibers with type B and C PAD patterns (not illustrated). It thus seems that, contrary to our initial expectations (Jimenez et al. 1988; Rudomin et al. 1986), the balance between excitatory and inhibitory descending influences acting on the pathway from cutaneous to Ib fibers may not be responsible for the expression of type B and type C PAD patterns shown by these fibers. However, this may require further investiga- tion in view of the observations of Holmqvist and Lund- berg ( 196 1) where it is shown that in the high pontine preparation, stimulation of high-threshold afferents in the PAN inhibits PBSt or GS monosynaptic reflexes, whereas in spinal cat the PBSt monosynaptic reflexes became facilitated, but the GS monosynaptic reflexes are still inhibited. It is to be noted that the action of group II afferents on contralat- era1 motoneurons also differs considerably in intact and spinal preparations, probably because of the elimination of descending noradrenergic and serotonergic pathways (Bras et al. 1989a,b, 1990; Noga et al. 1992). In this regard it should be mentioned that Harrison and Jankowska ( 1989) have shown that stimulation of high-threshold fibers in the PAN produces PAD in group II fibers. It will be very inter-

esting to see if such a PAD is also affected after spinalization and if it is modulated by activation of descending noradrenergic and serotonergic pathways.

Some functional implications

In the decerebrate cat small extension movements of the knee facilitate monosynaptic reflexes elicited in extensor motoneurons and inhibit monosynaptic reflexes of flexor motoneurons. This action appears to be mediated by large PAN afferent fibers and has been considered as a positive feedback response that tends to further increase knee joint extension (Grigg et al. 1978). On the other hand, electrical stimulation of PAN fibers with intermediate and high thresholds, which convey information from knee joint limit and nociceptive receptors, facilitates activity of flexor motoneurons and opposes further extension of the joint (He et al. 1988; Skoglund 1956). The presynaptic disinhibition of Ia input from flexor muscles demonstrated in this work, together with the increased gamma drive produced by stimulation of knee joint afferents (Baldissera et al. 1972; Bax- endale et al. 1992) will also increase the flexor reflex. How- ever, the presynaptic disinhibition of Ia input from extensors will have the opposite action. The latter appears somewhat paradoxical, but not surprising, if it is considered that Ia afferents do not only synapse with motoneurons but also have connections with spinal interneurons and provide information to supraspinal nuclei which may well be used for other purposes, including the programming of compen- satory reactions of descending origin.

To this general scheme we must add the presynaptic effects of knee joint afferents on Ib inputs. At present time the implications of this presynaptic control in terms of reflex activity are more difficult to predict because, as shown here, knee joint afferents produce PAD in one set of Ib fibers and inhibit the PAD in another set, and there is no information on whether these two types of Ib fibers have the same or different central actions. It is generally accepted that Ib fibers (from flexors as well as from extensors) excite flexor motoneurons and inhibit extensor motoneurons via poly- synaptic pathways (see Baldissera et al. 1972). The presynaptic disinhibition produced by knee joint afferents on Ib fibers with a type C PAD pattern would then reinforce the flexor reflex. However, the presynaptic inhibition of Ib fibers with type B PAD pattern would decrease their contribution to the flexor reflex. Again, there seems to be some paradox here in terms of reflex activity, which is only an expression of the complexity of the interactions between segmental afferents.

We thank Prof. E. Jankowska for useful comments to the manuscript, J. Gonzalez and C. Rodriguez for the implementation of the computer pro- grams required for data acquisition and processing, A. Rivera and C. Leon for technical support, and Dr. H. G. Schaible for lending us the thermode for spinal block.

This work was partly supported by a grant from the Volkswagen-Stif- tung, by the National Institute of Neurological Disorders and Stroke Grant NS-09 196, and by the Sistema National de Investigadores, Mexico.

Address for reprint requests: P. Rudomin, Dept. of Physiology, Biophys- ics, and Neurosciences, CINVESTAV-IPN, Apartado Postal 14-740, Mex- ico D.F. 07000, Mexico.

Received 27 October 1992; accepted in final form 29 April 1993.

1910 QUEVEDO, EGUIBAR, JIMENEZ, SCHMIDT, AND RUDOMIN

REFERENCES

BALDISSERA, F., HULTBORN, H., AND ILLERT, M. Integration in spinal neuronal systems. In: Handbook of Physiology. The Nervous System. Motor Control. Bethesda, MD: Am. Physiol. Sot., 198 1, sect. 1, vol. II, part 1, p. 509-595.

BARRON, D. A. AND MATTHEWS, B. H. The interpretation of potential changes in the spinal cord. J. Physiol. Lond. 92: 276-32 1, 1938.

BAXENDALE, R. H., DAVEY, N. I., ELLAWAY, P. H., AND FERREL, W. R. The interaction between joint and cutaneous afferent inputs in the regu- lation of fusimotor neuron discharge. In: Muscle Afferents and SpinaZ Control of Movement, edited by L. Jami, E. Pier-rot-Deseilligny, and D. Zytnicki. New York: Pergamon, 1992, p. 95- 104.

BERNHARD, C. G. Analysis of the spinal cord potentials in leads from the cord dorsum. In: The Spinal Cord, edited by J. L. Malcolm and J. A. B. Gray. Boston, MA: Little, Brown, and Co., 1953, p. 43-60.

BOYD, I. A. AND KALU, K. U. Scaling factor relating conduction velocity and diameter for myelinated afferent nerve fibres in the cat hindlimb. J. Physiol. Lond. 289: 277-297, 1979.

BRAS, H., CAVALLARI, P., JANKOWSKA, E., AND MCCREA, D. Comparison of effects of monoamines on transmission on spinal pathways from group I and II muscle afferents in the cat. Exp. Brain Res. 76: 27-37, 1989a.

BRAS, H., JANKOWSKA, E., AND NOGA, B. R. Depression of transmission from group II muscle afferents to spinal interneurones by descending monoaminergic pathways. Eur. J. Neurosci. 2, Suppl. : 187, 1989b.

BRAS, H., JANKOWSKA, E., NOGA, B. R., AND SKOOG, B. Comparison of effects of various types of NA and 5-HT agonists on transmission from group II muscle afferents in the cat. Eur. J. Neurosci. 2: 1029- 1039, 1990.

BRINK, E., JANKOWSKA, E., AND SKOOG, B. Convergence onto interneurons subserving primary afferent depolarization of group I afferents. J. Neurophysiol. 5 1: 432-449, 1984.

BURGESS, P. R. AND CLARK, F. J. Characteristics of knee joint receptor in the cat. J. Physiol. Lond. 203: 3 17-335, 1969.

BURKE, R. E. AND RUDOMIN, P. Spinal neurons and synapses. In: Hand- book of Physiology, The Nervous System. Bethesda, MD: Am. Physiol. Sot., 1977, sect. 1, vol I, p. 877-944.

BURKE, R. E., RUDOMIN, P., VYKLICKY, L., AND ZAJAC, F. E. Primary afferent depolarization and flexion reflexes produced by radiant heat stimulation of the skin. J. Physiol. Lond. 2 13: 185-2 14, 197 1.

CARPENTER, D., ENGBERG, I., FUNKENSTEIN, H., AND LUNDBERG, A. De- cerebrate control of reflexes to primary afferents. Acta Physiol. Stand. 59:424-437,1963.

CAVALLARI, P. AND PETTERSSON, L.-G. Tonic suppression of reflex transmission in low spinal cats. Exp. Brain Res. 77: 20 l-2 12, 1989.

CERVERO, F., SCHAIBLE, H.-G., AND SCHMIDT, R. F. Tonic descending inhibition of spinal cord neurones driven by joint afferents in normal cats and in cats with an inflamed knee joint. Exp. Brain Res. 83: 675- 678,1991.

CRAIG, A. D., HEPPELMAN, N. B., AND SCHAIBLE, H.-G. The projection of the medial and posterior articular nerves of the cat’s knee to the spinal cord. J. Comp. Neural. 276: 279-288, 1988.

ECCLES, R. M. AND LUNDBERG, A. Synaptic actions in motoneurones by afferents which may evoke the flexion reflex. Archives Italiennes de Bio- Zogie97: 199-221, 1959.

ECCLES, J. C., MAGNI, F., AND WILLIS, W. D. Depolarization of central terminals of group I afferent fibres from muscle. J. Physiol. Land. 160: 62-93,1962.

FERRELL, W. R. The adequacy of stretch receptors in the cat knee joint for signalling joint angle throughout a full range of movement. J. Physiol. Lond. 299: 85-99, 1980.

FREEMAN, M. A. R. AND WYKE, B. The innervation of the knee joint. An anatomical and histological study in the cat. J. Anat. 10 1: 505-532, 1967.

GRIGG, P., HARRIGAN, E. P., AND FOGARTY, K. E. Segmental reflexes mediated by joint afferent neurons in cat knee. J. Neurophysiol. 4 1: 9-14, 1978.

HARRISON, P. J. AND JANKOWSKA, E. Primary afferent depolarization of

central terminals of group II muscle afferents in the cat spinal cord. J. Physiol. Land. 4 11: 7 l-83, 1989.

HE, X., PROSKE, U., SCHAIBLE, H.-G., AND SCHMIDT, R. F. Acute inflam- mation of the knee joint in the cat alters responses of flexor motoneurons to leg movements. J. Neurophysiol. 59: 326-340, 1988.

HOLMQVIST, B. AND LUNDBERG, A. Differential supraspinal control of synaptic actions evoked by volleys in the flexion reflex afferents in alpha motoneurones. Acta Physiol. Stand. 54, Suppl. 186: l-5 1, 196 1.

JANKOWSKA, E., RIDDELL, J. F., AND MCCREA, D. A. Primary afferent depolarization of myelinated fibres in the joint and interosseus nerves of the cat. J. Physiol. Lond. 466: 115-l 3 1, 1993.

JIM~NEZ, I., RUDOMIN, P., AND SOLODKIN, M. PAD patterns ofphysiolog- ically identified afferent fibers from the medial gastrocnemius muscle. Exp. Brain Res. 7 1: 643-657, 1988.

LANDGREN, S. AND SILFVENIUS, H. Projection to cerebral cortex of group I muscle afferents from the cat’s hindlimb. J. Physiol. Land. 200: 353- 373,1969.

LAIRD, J. M. A. AND CERVERO, F. Tonic descending influences on the receptive field properties of nociceptive neurons in the sacral spinal cord of the rat. J. Neurophysiol. 63: 1022- 1032, 1990.

LUND, S., LUNDBERG, A., AND VYIUICKY, L. Inhibitory action from the flexor reflex afferents on transmission to Ia afferents. Acta Physiol. Stand. 64: 345-355, 1965.

LUNDBERG, A., MALMGREN, K., AND SCHOMBURG, E. D. Role of joint afferents in motor control exemplified by effects on reflex pathways from Ib afferents. J. Physiol. Land. 284: 327-343, 1978.

MADRID, J., ALVARADO, J., DUTTON, H., AND RUDOMIN, P. A method for the dynamic continuous estimation of excitability changes of single fiber terminals in the central nervous system. Neurosci. Lett. 11: 253-258, 1979.

NOGA, B. R., BRAS, H., AND JANKOWSKA, E. Transmission from group II muscle afferents is depressed by stimulation of locus coeruleus/ subcoer- uleus, Kolliker-Fuse and Raphe nuclei in the cat. Exp. Brain Rex 88: 502-516, 1992.

PROSKE, U., SCHAIBLE, H.-G., AND SCHMIDT, R. F. Joint receptors and kinaesthesia. Exp. Brain Res. 72: 2 19-224, 1988.

QUEVEDO, J., EGUIBAR, J. R., JIM~NEZ, I., SCHMIDT, R. F., AND RUDO- MIN, P. Modulation of synaptic effectiveness of group I muscle fibers by articular afferents. Sot. Neurosci. Abstr. 17: 1024, 199 1.

RUDOMIN, P. Presynaptic control of synaptic effectiveness of muscle spindle and tendon organ afferents in the mammalian spinal cord. In: The Segmental Motor System, edited by M. D. Binder and L. M. Mendell. New York: Oxford University Press, 1990, p. 349-380.

RUDOMIN, P. Presynaptic inhibition of muscle spindle and tendon organ afferents in mammalian spinal cord. Trends Neurosci. 13: 499-505, 1991.

RUDOMIN, P., ENGBERG, I., AND JIM~NEZ, I. Mechanisms involved in presynaptic depolarization of group I and rubrospinal fibers in cat spinal cord. J. Neurophysiol. 46: 532-548, 198 1.

RUDOMIN, P., JIM~NEZ, I., SOLODIUN, M., AND DuE~;~As, S. Sites of action of segmental and descending control of transmission on pathways mediating PAD of Ia- and Ib-afferent fibers in cat spinal cord. J. Neuro- physiol. 50: 743-769, 1983.

RUDOMIN, P., SOLODKIN, M., AND JIM~NEZ, I. PAD and PAH response patterns of group Ia and Ib fibers to cutaneous and descending inputs in the cat spinal cord. J. Neurophysiol. 56: 987-1006, 1986.

SCHAIBLE, H.-G., NEUGEBAUER, V., CERVERO, F., AND SCHMIDT, R. F. Changes in tonic descending inhibition of spinal neurons with articular input during the development of acute arthritis in the cat. J. Neurophys- iol. 66: 1021-1032, 1991.

SCHAIBLE, H.-G., SCHMIDT, R. F., AND WILLIS, W. D. Responses of spinal cord neurones to stimulation of articular afferent fibres in the cat. J. Physiol. Lond. 372: 575-593, 1986.

SCHMIDT, R. F. Presynaptic inhibition in the vertebrate central nervous system. Ergeb. Physiol. 63: 20-101, 197 1.

SKOGLUND, S. Anatomical and physiological studies of the knee joint innervation in the cat. Acta Physiol. Stand. 36, Suppl. 124: 33- 10 1, 1956.

WALL, P. D. Excitability changes in afferent fibre terminations and their relation to slow potentials. J. Physiol. Land. 142: l-2 1, 1958.

WILLIS, W. D. AND COGGESHALL, R. E. Sensory Mechanisms of the Spinal Cord. New York: Plenum, 199 1.

primary merent depolarization of muscle merents elicited ... · ( 1.2- 1.4 xt) produced a small...

Documents