- Email: [email protected]
Spinal somatosensory evoked potentials in mice and their developmental changes
~,~
!:
ELSEVIER
Brain & Development 1994; 16:44 51
Original Article
Spinal somatosensory evoked potentials in mice and their developmental changes Amayur P. Chandran *, PhD, Kenichiro Oda, MD, Hiroshi Shibasaki **, MD, Madhavan Pisharodi * MD
Spinal somatosensory evoked potentials (SEP) were recorded in 58 normal mice (C3H strain) divided into 4 groups according to age (3-, 6-, 9- and 12 weeks). Monopolar recordings of spinal SEP were made by subdermal needle electrodes from 3 vertebral levels, 'low-lumbar', 'high-lumbar' and 'mid-thoracic', by stimulating the tibial nerve bilaterally at the ankle. Three negative peaks, NI, NIl and NIII, presumably due to conduction through muscle afferents, cutaneous afferents (in the dorsal root or dorsal white column) and spinocerebellar tract, respectively, were recorded at the high-lumbar level in the 12-week-old mouse. Besides the NI and NII peaks, a small ventral root potential was also occasionally recorded at the low-lumbar level. At the mid-thoracic level, only NI and NIII were recordable. At both the high-lumbar and mid-thoracic levels, the negative peaks were superimposed over long duration ' s u m m a t i o n potentials' of opposite polarities. Well-defined standing potentials were also recorded at these two levels. The standing potentials could be the 'entry point potential' due to the entry of S1 root into the spinal cord at the T13 vertebral level. The summation potential presumably is due to a fixed generator located between the T7 and T12 vertebral levels resulting from intense synaptic activity at this level. In 3- and 6-week-old mice, the entry point potential was recorded in the low-lumbar SEP also, possibly due to less axial growth of the vertebral column at this stage of development. The summation potential and N I I I peak were reduced in size in these young mice, possibly due to less developed collaterals and synaptic activity. The somatosensory conduction velocity, measured between the low-lumbar and mid-thoracic recording sites, showed a highly significant increase during the 3 - 6 and 6 - 9 week periods, suggesting that a significant amount of myelination occurs in proprioceptive fibres postnatally in mice. Key words: Spinal somatosensory evoked potential; Subdermal monopolar recording; Bilateral tibial nerve stimula-
tion; Somatosensory conduction velocity; Developmental change; Mouse
1. I N T R O D U C T I O N Extensive studies have been made on spinal somatosensory evoked potentials (SEP) in man and their
National Institute of Neuroscience, NCNP, Kodaira, Tokyo, Japan Received 3 August 1993; accepted 20 October 1993 Correspondence address': A.P. Chandran, Brownsville Pain Research
Center, 844 Central Blvd Suite 1200, Brownsville, TX 78520, USA. Fax: (1) (210) 541-2070. Present addresses: * Brownsville Pain Research Center, 844 Central
Blvd Suite 600, Brownsville, TX 78520, USA; ** Department of Brain Pathophysiology,Kyoto University School of Medicine, Kyoto 606, Japan. 0387-7604/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSD1 0387-7604(93)E0103-P
utility is adequately understood [1-3]. However, such studies on animals are more limited [4-6]: in fact, reports on spinal SEP in mice are totally lacking in the literature. Since studies on neurodegenerative diseases are mainly performed on mice, an analysis of the spinal SEP, the ontogenesis of its wave forms, and the changes in normal values of spinal somatosensory conduction velocity during postnatal development in mice, could be of immense value in assessing neural deficits. The present work analyses the patterns of spinal SEP recorded at three vertebral levels - 'low-lumbar', 'high-lumbar' and 'mid-thoracic' - following bilateral stimulation of the tibial nerve at the ankle in normal mice, in order to identify the origin of different wave components in the spinal SEP. Variations in the spinal SEP during various ages were also studied. Further,
A.P. Chandran et aL / Brain & Development 1994; 16:44-51
the somatosensory conduction velocity was estimated at various ages to understand the extent of myelination during postnatal life.
2. MATERIALS AND METHODS The study was conducted in 58 normal mice (C3H strain) of either sex divided into 4 groups according to their age. Group I comprised 16 mice of age 3 weeks and mean weight 13.6 + 0.2 g (SEM). Similarly, group II comprised 17 mice (6 weeks) of mean weight 19.5 + 0.3 g, group III, 11 mice (9 weeks; 26.5 + 0.1 g) and group IV, 14 mice (12 weeks; 26.9 + 0.3 g). The animal to be investigated was lightly anesthetized by ether inhalation followed by sodium pentobarbital (Nembutal) 0.04 m g / g body weight, intraperitoneally. The animal was laid on its belly, the legs slightly stretched and the toes gently fixed to a wooden board by applying surgical tapes. The stimulating electrodes consisted of two thin copper wires introduced subcutaneously with the help of a 26-gauge hypodermic
EP
NI NIl NIZI
[
_I l • R~n I
/
needle. The cathode was implanted on the medial side of the ankle between the Achilles tendon and the medial malleolus and the anode 1 cm proximally on the lateral side. The ground electrode was applied to the tail. A midline incision was made on the back and the skin reflected slightly. Stainless steel needle electrodes with tip diameter of 0.03 mm were introduced subdermally between spinous processes at three vertebral levels: (i) L4-L5 intervertebral space (low-lumbar level, i.e. at the level of upper border of iliac crest), (ii) T13-L1 intervertebral space (high-lumbar level, i.e. just below the level of the lateral borders of the last floating ribs), and (iii) approximately at the T 6 - T 7 intervertebral space (mid-thoracic level, i.e. 0.5-0.7 cm rostral to the high-lumbar recording electrodes). Another stainless-steel needle (tip 0.03 ram) introduced subcutaneously into the ear acted as the reference electrode. Monopolar recordings of spinal SEP were simultaneously made from the 3 sites on a 4-channel electromyograph (Dantec-Counterpoint) and printouts
EP
A
;
45
NI N n
B
1 mslD "
,
II1'I
,
'i
.
lilI I ~ , i ,! , i
.
4kl.t ~ / . i .'pt"
/I
rr'~
.
IA.: ;I \
IIIt~
12o O,V/D .
. .
.
"
.
.
.
MT
.
.
.
.
.
.
.
.
ms
. . . . .
'
'
I . . . . . .
.
.
I 1 1 / %
. . . . . . .
.!i, ,
- - - -
.
.
.
IAig\ I
.
\
I
.
.
.
.
.
.
.
.
.
.
i
.
. . . . .
.
.
.
'VR .
I' Stimulus
.
.
.
HL'
.
= :1
Ill*
.
P
/I
/,
.
'
•
.,,..
.
.
.
.
.
LL"
.
1.00 mA
Fig. 1. Spinal SEP in a 12-week-old mouse recorded in low-lumbar (LL), high-lumbar (HL) and mid-thoracic (MT) leads following low (A) and high (B) intensity stimulation. EP, entry point potential (note only in HL and MT); NI, Nil and NIII, negative peaks according to the order of increasing onset latencies (note there is no NII in MT); VR, ventral root potential. (Note the long duration summation potentials (unmarked) of opposite polarities in HL and MT.) Changes during increasing strength of stimulus (B) include disappearance of VR at LL, appearance of NII at LL and increase in the sizes of all waves, especially at the HL level.
A.P. Chandran et al. / B r a i n & Det,elopment 1994; 16." 44-51
46
were obtained. The tibial nerves were simultaneously stimulated bilaterally at the ankle with repetitive square wave pulses (0.3 Hz) of 0.1 ms duration and of adequate strength (0.7-1.2 mA) from the constant current stimulator of the EMG. The band pass of the filter was kept 50 H z - 2 kHz. One hundred responses were averaged with respect to stimulus onset. The analysis time of the averager was kept at 10 ms from the start of the stimulus. For any given trial, at least two series of records were taken; one just above the threshold strength of stimulus and the other at a higher strength, usually double the threshold strength. For the purpose of calculating the somatosensory conduction velocity, the onset latency of the major negative wave was noted in each SEP with the help of cursors. The distance between the recording electrodes was measured accurately with the help of calipers. The somatosensory conduction velocity between low-lumbar and mid-thoracic levels was calculated from the difference in latencies of SEP recorded from the two levels and the distance between the two recording points. (Although SEP from the high-lumbar level was also recorded, the distance between the high-lumbar and mid-thoracic recording points were too short to make any accurate measurement of somatosensory conduction velocity.) The peak latencies of the negative waves
EP
were also noted for the calculation of interpeak latencies.
3. RESULTS The spinal SEP in the 12-week-old mouse showed the following major features (Figs. 1-3). (i) A triphasic response (positive-negative-positive) with a predominant negative component (NI) at the low-lumbar level; (ii) a small positive wave followed by a large negative component with usually 3 peaks, NI, NII and NIII at the high-lumbar level; and (iii) two negative waves of relatively small amplitude (NI and NIII) but longer duration, usually separated by a positive wave of variable duration at the mid-thoracic level. At the low-lumbar level, the NI and NII had peak latencies of 0.88 _+ 0.03 (S.E.M.) ms and 1.5 _+ 0.21 ms, respectively, with an interpeak latency of 0.62 _+ 0.02 ms. At the high-lumbar level, the NI, NIl and NIII had peak latencies of 1.22 + 0.04 ms, 1.88 _+ 0.05 ms and 2.27 _+ 0.06 ms, respectively. The N I - N I I interpeak latency was 0.67 + 0.04 ms and the N I - N I I I interpeak latency was 1.05 _+ 0.03 ms. At the mid-thoracic level, the NI and NIII had peak latencies of 1.49 _+ 0.04 ms and 2.57_+0.05 ms, respectively. The NI-NIII in-
EP
A
NI
N I NII NI11
|
I
"
-|
o.o
,.s
•
.I I
I!~
I .vl I
I
.
.
.
.
.
.
• I I
l
_i
•
m
,
I mslD
I.
; I I
10.0 •
. . . . .
I ~d~ I
I
i '1
.
I
I .
I.L#'~-
llf\
•
m
,l
I.
ms
•
i
,.,'V
MT
I
.
ll!
I
I
I
.'!
11 0
B
•
l~un. I,
11.11"
1
m
.
.
•
.
.
.
.
r~T
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
HL I
20/uVlD t
I
Stimulus
9.90
mA
I Stimulus
1.60 mA
Fig. 2. Spinal SEP in a 12-week-old mouse. S a m e n o m e n c l a t u r e s as for Fig. 1. Note the well-defined N i l on high intensity s t i m u l a t i o n (B).
47
A.P. Chandran et al. /Brain & Development 1994; 16:44-51
EP
NI N n
A
EP
N I NII NIII
I l[ o!ov D/ .," i
I
B
I
[ 10'I0
. . . .
II .iv,,\ :'. i . . . . .
ill
I~~i i t~' :1. . ~ . .M i . .~ . H
. . . .
.
MT___
l' t
. . . .
•
m ~
HLLi
.~
. ML
LI',
, Stimulus
1.20 mA
Fig.3. SpinalSEPin a 12-week-oldmouserecordedat 4 levels,includingmid-lumbar(ML)level.Samenomenclaturesas for Fig.1. Notethe appearanceof Nil in LLand MLanddisappearanceofVRin LLon highintensitystimulation(B). terpeak latency was 1.08 + 0.02 ms which was not significantly different (P > 0.05) from that at the highlumbar level. In fact, at both the high-lumbar and mid-thoracic levels, the N I - N I I I interpeak latencies varied within a very narrow range (0.8-1.2 ms). The NI latencies progressively increased from the low-lumbar to the mid-thoracic levels (Figs. 1-3). The Nil latencies also increased progressively from the low-lumbar to the high-lumbar levels (Figs. 2 and 3). The N I - N I I interpeak latency at the high-lumbar level was significantly higher (P < 0.05) as compared to that at the low-lumbar level. Although the distance between the low-lumbar and the high-lumbar sites was, on average, twice that between the high-lumbar and midthoracic sites, the difference in NI latency between the former pair was only 1.3 times that between the latter pair. A detailed analysis of the spinal SEP showed the following additional features. (i) At both the highlumbar and mid-thoracic levels there was a small but definite negative wave just after the stimulus artefact (EP; Figs. 1 and 2). This wave had a constant latency at both of these levels, a 'standing potential'. (ii) At the high-lumbar level, the three negative peaks were superimposed over a large negative 'summation potential'
with a long duration (Figs. 1 and 2). This summation potential was present also at the mid-thoracic level, but with opposite polarity and relatively smaller amplitude. (iii) A comparative study of the onset latency of negative peaks in the high-lumbar and mid-thoracic SEP revealed that the two negative peaks of mid-thoracic SEP corresponded to the NI and NIII peaks in the high-lumbar SEP (Fig. 1). (iv) Two more small negative potentials were occasionally seen at the low-lumbar level (Fig. 1). One of them, Nil (Fig. 2), occurred immediately after the first major negative potential, NI. The other, a relatively smaller one VR (Fig. 1), occurred after a short interval. This late negative potential was abolished on high intensity stimulation (Figs. 1 and 3). 3.1. Effect of an increase
in stimulus
strength
The SEP was recorded always at two stimulus strengths, one just adequate to evoke discernible wave forms at all recording sites (0.7-1.2 mA), and the other almost double the strength of the first. All the wave forms showed a significant increase in amplitude on the high intensity stimulation (Figs. 1 and 2) except for the small amplitude wave forms in the mid-thoracic
A.P, Chandran et al. /Brain & Decelopment 1994; 16." 44-51
48
EP
NI Nn
2./Run. .
/
,I.
.
,I
I
I
20~ uVlID , • I I I
I. I I
•I
I . I
!
!
,
I. I I
I
./.;
°
N nl
,
1 ms/D 1 0 . 0 ms
I I
.I
I
I
,
MT
only in 5 (out of 11) mice in the 9-week-old group. (ii) The size of the NIII peak in the high-lumbar SEP was relatively smaller as compared to that in 12-week-old mice. The decrease in size was greater in the midthoracic SEP (Fig. 4). (iii) There was a significant reduction in the size of summation potential especially in the mid-thoracic SEP. 3.3. Somatosensory conduction velocity
::[
,
# u,lD,
#
HL
~V/D t. I
•L L
I Stimulus
1.80 mA
Fig. 4. Spinal SEP in a 9-week-old mouse. Same nomenclatures as for Fig. 1. Note the well-marked entry point potential (EP) even at the LL level.
SEP, which showed relatively smaller increase in amplitude. The extent of the increase in the amplitude of the negative peak NI in the low-lumbar and highlumbar SEP was the same. The Nil peak also showed a significant increase on the high intensity stimulation. In some mice in which SEP was recorded additionally from the L2-L3 intervertebral space ('mid-lumbar'), the Nil peak appeared just after NI on the high intensity stimulation (Fig. 3B). The increase in amplitude of Nil in the high-lumbar SEP was greater as compared to that of NII both at the low-lumbar and mid-lumbar levels (Fig. 3B). The summation potential also showed a significant increase in amplitude, more so at the high-lumbar levels (Figs. 1 and 2). The long latency response at the low-lumbar level, VR, disappeared on the high intensity stimulation (Figs. 1 and 3). 3.2. Spinal SEP in 3-, 6- and 9-week-old mice The SEP in the younger mice (3, 6 and 9 weeks; Fig. 4) showed the following significant differences from that in 12-week-old mice. (i) The standing potential was observed even in the low-lumbar SEP invariably in all mice belonging to the 3- and 6-week-old groups, but
The somatosensory conduction velocity at 12 weeks was 36.0_+ 1.0 (S.E.M.) m / s . It was only 17.1 _+ 0.8 m / s at 3 weeks, showing a highly significant increase ( P < 0.001) in the somatosensory conduction velocity during the 3-12 week period. At 6 weeks and 9 weeks, the somatosensory conduction velocity was 28.2 ± 0.8 m / s and 35.6 ± 1.3 m / s , respectively, showing a highly significant increase ( P < 0.001) during the 3-6 and 6-9 week periods. There was no significant increase in the somatosensory conduction velocity after 9 weeks.
4. DISCUSSION Recording of spinal SEP following stimulation of the tibial nerve at the ankle is considered to be a difficult procedure due to the presence of relatively less numbers of muscle afferents in the nerve, that are shown to contribute the most to the genesis of spinal SEP [7-9]. In the present work, however, such a procedure had to be resorted to because stimulation at a more proximal site (popliteal fossa), which is usually preferred in man and larger animals, could have drastically reduced the available distance between the stimulating and recording electrodes, especially in young (3and 6-week-old) mice. Furthermore, in order to compensate for the possible reduction in the size of SEP, bilateral simultaneous stimulation of the tibial nerves at the ankle was employed in the present study. Such a procedure has been shown to improve the quality and amplitude of SEP recordings [2,10]. An attempt has also been made to improve the signal-to-noise ratio using needle electrodes introduced subdermally between the spinous processes. The relative merits and demerits of using monopolar and bipolar recording methods have already been recognized [10]. In the present work, wherein the main stress was placed on the study of wave forms, their sources and spinal conduction, the monopolar method of recording was preferred. Bilateral stimulation of the tibial nerves at the ankle produced well-defined triphasic SEP with a predominant negative wave (NI) at the low-lumbar level, a complex potential with 3 negative peaks (NI, Nil, NIII) superimposed over a large summation potential at the high-lumbar level, and 2 negative waves (NI, NIII)
A.P. Chandran et al. /Brain & Development 1994; 16:44-51
superimposed over a positive summation potential at the mid-thoracic level (Figs. 1 and 2). There was a progressive increase in the onset latency of NI wave from the low-lumbar to the mid-thoracic levels, indicating that it is a travelling wave. On high intensity stimulation, the degree of increase in amplitude of the NI waves in both the low-lumbar and high-lumbar SEPs was similar. The NI peak being the first component of SEP, is most probably due to conduction in large diameter muscle afferents of the dorsal root [2,8]. The progressive increase in the latency of NII in the low-lumbar, mid-lumbar (when recorded) and highlumbar SEPs suggests that it is also a travelling wave conducted along the dorsal root. However, since the N I - N I I interpeak latency at the high-lumbar level was more than that at the low-lumbar level ( P < 0.05), it might be presumed that NII peak is due to conduction along fibres with lesser conduction velocity as compared to the large diameter muscle afferents that produced the NI peak. The NII could be elicited only at a higher stimulus strength, especially at the low-lumbar level (Fig. 2). Further, there was a marked increase in the amplitude of NII at the high-lumbar level (Fig. 3) due to a possible summation effect following bilateral stimulation [11]. All these observations make it highly probable that the NII peak is due to conduction along large diameter cutaneous afferents. There was no wave in the mid-thoracic SEP corresponding in latency to the NII of the high-lumbar recording (Figs. 1-3). It can be assumed that the impulses that cause NII in the caudal recordings are either not conducted rostrally or that they do not contribute sufficiently to SEP peaks at the mid-thoracic level, where records are basically of lower amplitude [2,8]. Cracco et al. [2] observed in man that sural nerve stimulation failed to produce any recordable potential in the thoracic leads. Assuming that the NII peak in the high-lumbar SEP is formed by large diameter cutaneous afferents, as suggeted above, it is quite possible that the NII peak is hardly recordable at the midthoracic level. Alternatively, it is also possible that most of the fibres contributing to the genesis of NII at the high-lumbar level are involved in spinal reflexes [11] and that only a few fibres ascend in the dorsal column to generate a recordable NII peak at the mid-thoracic level. It is quite possible that the NIII peak has its origin close to the root entry zone, since there was no peak corresponding to it in the caudal recordings and it appeared first in the high-lumbar recording (Figs. 1 and 3). Since the interpeak latency between NI and NIII in the high-lumbar recording was almost 1 ms and remained within narrow limits in all the recordings, it is quite possible that the NI wave might have dissociated into NI and NIII peaks at the high-lumbar level, the NI peak being presynaptic and NIII post-synaptic.
49
The anatomical correlates of this contention is that the muscle afferents in the dorsal root travel in both dorsal white column (NI, presynaptic) and Clarke's c o l u m n / spinocerebellar tract (NIII, post-synaptic), after they enter the spinal cord. The second negative peak in the mid-thoracic SEP corresponded to the NIII peak of the high-lumbar SEP (Figs. 1-3). As the interpeak latency between NI and NIII in the mid-thoracic SEP is almost the same as that at the high-lumbar recording, it is possible that the fibres responsible for NI and NIII waves have almost the same conduction velocity (although the distance between the high-lumbar and mid-thoracic recording sites was too short to make any definite conclusion). The possible origins of NI and NIII peaks at the high-lumbar level being dorsal white column and Clarke's column, respectively, both comprising proprioceptive fibres, as discussed above, it is likely that the NI and NIII peaks at the mid-thoracic level originate in dorsal white column and spinocerebellar tracts, respectively. A small negative wave with a relatively long latency was also recorded in the low-lumbar SEP in some 12-week-old mice (Figs. 1 and 3). The observation that this potential disappeared on the high intensity stimulation (Fig. 1, 3) indicates that it could be the ventral root potential as had been reported earlier [6,9]. The fact that it could be recorded only in a few animals may be due to the relatively small number of proprioceptive afferents in the tibial nerve at ankle. Alternatively, it may well be due to simultaneous stimulation of the adequate number of motor fibres that may antidromically occlude the ventral root potential at the low-lumbar level [9]. The high-lumbar SEP was of a more complex configuration and larger as compared to caudal or rostral recordings (Figs. 1-3). Such an SEP with maximal amplitude and complex configuration had been reported at T12 also in man [2,10]. The large negative wave with long duration (summation potential) could possibly be a summated activity of the 3 negative peaks, NI, Nil and NIII. However, this summation potential had an opposite polarity in the mid-thoracic SEP, although NI and NIII had negative polarities like those in the high-lumbar SEP, suggesting that the summation potential possibly had a separate generator. It appears that the afferent impulses cause the development of certain current sources, most probably due to intense synaptic activity. The reversal of polarity indicates that it is a fixed generator located between T7 and T12 vertebral levels similar to the source that produces N13 in the human SEP following median nerve stimulation [12,13]. It is generally accepted that the standing potential represents the junctional potential due to the entry of dorsal root potential into the spinal c o r d - the 'entry
50
A.P. Chandran etal./Brain & Development 1994; 16." 44-51
point potential' [2,3,8]. The appearance of a well-defined entry point potential at the high-lumbar level in the present study (Figs. 1-3) suggests that, in the mouse, the S1 nerve root enters the spinal cord at T13 vertebral level, like as reported earlier in the rat [14]. In the present study, 3 major changes have been observed in the SEP pattern during postnatal development. (i) The entry point potential was present even at the low-lumbar level invariably in all 3- and 6-week-old mice and in half the number of 9-week-old mice (Fig. 4). It is possible that in younger mice ( < 9 weeks old) the S1 root enters at a lower level as compared to the adult mice. This corroborates well with the finding that the spinal cord ends at a lower level in younger rats so that the spinal segments correspond to lower vertebral levels as compared to adult rats [15]. The axial growth of the vertebral column during postnatal life may be the most probable reason for this developmental change. (ii) Another important difference oberved in the present study was the reduced amplitude of NIII in the high-lumbar and mid-thoracic SEP. Since, as discussed earlier, the NIII peak depends upon the synaptic activity in Clarke's column and conduction in the spinocerebellar tract, it is quite likely that the decreased amplitude of NIII is a consequence of decreased synaptic activity during early postnatal life. In fact, it had been reported that many collaterals in the spinal cord develop only postnatally and the synaptic activity shows a progressive development during postnatal growth in cats [16] and mice [17]. (iii) A reduction in the amplitude of summation potential especially in the mid-thoracic SEP, in the present study, also suggests a decreased synaptic activity during early postnatal life. Alternatively, the degree of myelination being negligible in newborn mice [18], the reduced conduction in the spinocerebellar tract may also contribute to the reduced size of the NIII peak. In the present study, the somatosensory conduction velocity was calculated from the latency differences between the SEP recorded at the low-lumbar and mid-thoracic levels. (A comparison between the caudal and rostral spinal conduction velocities was planned, but the distance between the high-lumbar and midthoracic recording sites was too short to attempt any meaningful comparison.) Since the entry point of S1 root was at the T13 vertebral level [14], the major part of the conduction velocity is along the dorsal root. However, in the younger mice, wherein the spinal levels correspond to lower vertebral levels, the conduction measured between the low-lumbar and midthoracic levels may involve more spinal component as compared to the adult mice. A comparison of the onset latencies of NI peaks in the low-lumbar, high-lumbar and mid-thoracic SEPs definitely suggested a significant decrease in the somatosensory conduction velocity across the high-
lumbar-mid-thoracic recording sites (although the decrease could not be quantified accurately due to the nearness of the two recording sites). This is expected as the fibres across the low-lumbar-high-lumbar recording sites are thicker due to their proximity to the dorsal root ganglion. The somatosensory conduction velocity showed a highly significant increase during the 3-6 and 6-9 week periods indicating a significant degree of myelination of the proprioceptive fibres in the dorsal root and dorsal white column. It is possible that, like in rats [18,19], myelination in mice is also almost a postnatal phenomenon. This may make mice more vulnerable to demyelinating diseases of juvenile onset and of the progressively degenerating type as compared to man. The present study is the first report on spinal SEP in mice. From the findings it may be concluded that conduction along the muscle afferents, cutaneous afferents and spinocerebellar tract produced separate peaks in spinal SEP. A fixed generator was identified, located between T7 and T13. The root entry of $1 is at T13 in adult mice, but in younger ones it is lower down, possibly due to less somatic growth. The reduced size of certain waves observed in the present study in young mice, especially in rostral recordings, indicated less synaptic activity. During the early postnatal life (3-9 weeks), there was a significant increase in the somatosensory conduction velocity indicating a substantial amount of myelination during this period.
5. REFERENCES 1. Chiappa KH, Yiannikas, C. Evoked potentials in clinical medicine. New York: Raven Press, 1983. 2. Cracco RQ, Cracco JB, Sranowski R, Vogel HB. Spinal evoked potentials. In: Desmedt JE, ed. Progress in clinical neurophysiology. VoL 7. Basel: Karger, 1980: 87-104. 3. Kimura J. Electrodiagnosis in d&eases of nerve and muscle: principles and practice. 2nd ed. Philadelphia: Davies, 1989. 4. Happel LT, Leblanc HJ, Kline DG. Spinal cord potentials evoked by peripheral nerve stimulation. Electroencephalogr Clin Neurophysiol 1975; 38: 349-54. 5. Rossini PM, Greco F, De Palma C, Pisano k Electrodiagnosis of the rabbit. Monitoring of spinal conduction in acute cord lesions versus clinical observation. Eur Neurol 1980; 19: 409-13. 6. Weitholter H, Hulser PJ. Lumbar spinal somatosensory evoked potentials in the rat after stimulation of the tibial nerve. Exp Neurol 1985; 89: 24-31. 7. Gandevia SC, Burke D, McKeon B. The projection of muscle afferents from the hand to cerebral cortex in man. Brain 1984; 107: 1-13. 8. Halonen JP, Stephen J J, Edgar MA, Ransford AD. Conduction properties of epidermally recorded spinal cord potentials following lower limb stimulation in man. Electroencephalogr Clin Neurophysiol 1989; 74: 161-74. 9. Yiannikas C, Shahani BT. The origins of lumbosacral evoked potentials in h u m a n s using a surface electrode technique. J Neurol Neurosurg Psychiatry 1988; 51: 499-508. 10. Kakigi R, Shibasaki H, Hashizume A, Kuriyowa Y. Short latency
A.P. Chandran et al. / Brain & Development 1994; 16." 44-51
11.
12.
13.
14.
somatosensory evoked spinal cord and scalp recorded potentials follwing posterior tibial nerve stimulation in man. Electroencephalogr Clin Neurophysiol 1982; 53: 602-11. Meinck HM, Benecke R, Kuster S, Conrad B. Cutaneomuscular organization in normal man and in patients with motor disorders. In: Desmedt JE, ed. Motor Control mechanisms in health and disease. New York: Raven press, 1983: 787-96. Allison T, Wood CC, McCarthy G, Hume AL, Goff WR. Short latency somatosensory evoked potentials in man, monkey, cat and rat: comparative latency analysis. In: Mauguiere F, Revol M, eds. Clinical application o f evoked potentials in neurology. New York: Raven Press, 1982: 303-10. Desmedt JE, Cheron G. Prevertebral (oesophageal) recording of subcortical somatosensory evoked potentials in man: the spinal P13 component and the dual nature of the spinal generators. Electroencephalogr Clin Neurophysiol 1981; 52: 257-75. Pamphlett RS. Spinal irritation does not inhibit distal axonal sprouting. Muscle Nerve 1988; 11: 493-501.
51
15. Brunquell PJ, Taylor GW, Holmes GL, Feldman DS. Ontogenesis of lumbar somatosensory evoked potentials after posterior nerve stimulation in the rat. Electroencephalogr Clin Neurophysiol 1990; 77: 112-8. 16. Skoglund S. The activity of muscle receptors in kitten. Acta Physiol Scand 1960; 50: 203-21. 17. Chandran AP, Oda K, Shibasaki H. Changes in motoneuron excitability during postnatal life in the mouse. Brain Dev (Tokyo) 1991; 13: 180-3. 18. Jacobson S. The sequence of myelination in the brain of albino rat. J Comp Neurol 1963; 121: 5-29. 19. Schonbach J, Hu KH, Friede RL. Cellular and clinical changes during myelination: autoradiographic, histological and biochemical data on myelination in the pyramidal tract and corpus callosum. J Comp Neurol 1968; 134: 21-38.
!:
ELSEVIER
Brain & Development 1994; 16:44 51
Original Article
Spinal somatosensory evoked potentials in mice and their developmental changes Amayur P. Chandran *, PhD, Kenichiro Oda, MD, Hiroshi Shibasaki **, MD, Madhavan Pisharodi * MD
Spinal somatosensory evoked potentials (SEP) were recorded in 58 normal mice (C3H strain) divided into 4 groups according to age (3-, 6-, 9- and 12 weeks). Monopolar recordings of spinal SEP were made by subdermal needle electrodes from 3 vertebral levels, 'low-lumbar', 'high-lumbar' and 'mid-thoracic', by stimulating the tibial nerve bilaterally at the ankle. Three negative peaks, NI, NIl and NIII, presumably due to conduction through muscle afferents, cutaneous afferents (in the dorsal root or dorsal white column) and spinocerebellar tract, respectively, were recorded at the high-lumbar level in the 12-week-old mouse. Besides the NI and NII peaks, a small ventral root potential was also occasionally recorded at the low-lumbar level. At the mid-thoracic level, only NI and NIII were recordable. At both the high-lumbar and mid-thoracic levels, the negative peaks were superimposed over long duration ' s u m m a t i o n potentials' of opposite polarities. Well-defined standing potentials were also recorded at these two levels. The standing potentials could be the 'entry point potential' due to the entry of S1 root into the spinal cord at the T13 vertebral level. The summation potential presumably is due to a fixed generator located between the T7 and T12 vertebral levels resulting from intense synaptic activity at this level. In 3- and 6-week-old mice, the entry point potential was recorded in the low-lumbar SEP also, possibly due to less axial growth of the vertebral column at this stage of development. The summation potential and N I I I peak were reduced in size in these young mice, possibly due to less developed collaterals and synaptic activity. The somatosensory conduction velocity, measured between the low-lumbar and mid-thoracic recording sites, showed a highly significant increase during the 3 - 6 and 6 - 9 week periods, suggesting that a significant amount of myelination occurs in proprioceptive fibres postnatally in mice. Key words: Spinal somatosensory evoked potential; Subdermal monopolar recording; Bilateral tibial nerve stimula-
tion; Somatosensory conduction velocity; Developmental change; Mouse
1. I N T R O D U C T I O N Extensive studies have been made on spinal somatosensory evoked potentials (SEP) in man and their
National Institute of Neuroscience, NCNP, Kodaira, Tokyo, Japan Received 3 August 1993; accepted 20 October 1993 Correspondence address': A.P. Chandran, Brownsville Pain Research
Center, 844 Central Blvd Suite 1200, Brownsville, TX 78520, USA. Fax: (1) (210) 541-2070. Present addresses: * Brownsville Pain Research Center, 844 Central
Blvd Suite 600, Brownsville, TX 78520, USA; ** Department of Brain Pathophysiology,Kyoto University School of Medicine, Kyoto 606, Japan. 0387-7604/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSD1 0387-7604(93)E0103-P
utility is adequately understood [1-3]. However, such studies on animals are more limited [4-6]: in fact, reports on spinal SEP in mice are totally lacking in the literature. Since studies on neurodegenerative diseases are mainly performed on mice, an analysis of the spinal SEP, the ontogenesis of its wave forms, and the changes in normal values of spinal somatosensory conduction velocity during postnatal development in mice, could be of immense value in assessing neural deficits. The present work analyses the patterns of spinal SEP recorded at three vertebral levels - 'low-lumbar', 'high-lumbar' and 'mid-thoracic' - following bilateral stimulation of the tibial nerve at the ankle in normal mice, in order to identify the origin of different wave components in the spinal SEP. Variations in the spinal SEP during various ages were also studied. Further,
A.P. Chandran et aL / Brain & Development 1994; 16:44-51
the somatosensory conduction velocity was estimated at various ages to understand the extent of myelination during postnatal life.
2. MATERIALS AND METHODS The study was conducted in 58 normal mice (C3H strain) of either sex divided into 4 groups according to their age. Group I comprised 16 mice of age 3 weeks and mean weight 13.6 + 0.2 g (SEM). Similarly, group II comprised 17 mice (6 weeks) of mean weight 19.5 + 0.3 g, group III, 11 mice (9 weeks; 26.5 + 0.1 g) and group IV, 14 mice (12 weeks; 26.9 + 0.3 g). The animal to be investigated was lightly anesthetized by ether inhalation followed by sodium pentobarbital (Nembutal) 0.04 m g / g body weight, intraperitoneally. The animal was laid on its belly, the legs slightly stretched and the toes gently fixed to a wooden board by applying surgical tapes. The stimulating electrodes consisted of two thin copper wires introduced subcutaneously with the help of a 26-gauge hypodermic
EP
NI NIl NIZI
[
_I l • R~n I
/
needle. The cathode was implanted on the medial side of the ankle between the Achilles tendon and the medial malleolus and the anode 1 cm proximally on the lateral side. The ground electrode was applied to the tail. A midline incision was made on the back and the skin reflected slightly. Stainless steel needle electrodes with tip diameter of 0.03 mm were introduced subdermally between spinous processes at three vertebral levels: (i) L4-L5 intervertebral space (low-lumbar level, i.e. at the level of upper border of iliac crest), (ii) T13-L1 intervertebral space (high-lumbar level, i.e. just below the level of the lateral borders of the last floating ribs), and (iii) approximately at the T 6 - T 7 intervertebral space (mid-thoracic level, i.e. 0.5-0.7 cm rostral to the high-lumbar recording electrodes). Another stainless-steel needle (tip 0.03 ram) introduced subcutaneously into the ear acted as the reference electrode. Monopolar recordings of spinal SEP were simultaneously made from the 3 sites on a 4-channel electromyograph (Dantec-Counterpoint) and printouts
EP
A
;
45
NI N n
B
1 mslD "
,
II1'I
,
'i
.
lilI I ~ , i ,! , i
.
4kl.t ~ / . i .'pt"
/I
rr'~
.
IA.: ;I \
IIIt~
12o O,V/D .
. .
.
"
.
.
.
MT
.
.
.
.
.
.
.
.
ms
. . . . .
'
'
I . . . . . .
.
.
I 1 1 / %
. . . . . . .
.!i, ,
- - - -
.
.
.
IAig\ I
.
\
I
.
.
.
.
.
.
.
.
.
.
i
.
. . . . .
.
.
.
'VR .
I' Stimulus
.
.
.
HL'
.
= :1
Ill*
.
P
/I
/,
.
'
•
.,,..
.
.
.
.
.
LL"
.
1.00 mA
Fig. 1. Spinal SEP in a 12-week-old mouse recorded in low-lumbar (LL), high-lumbar (HL) and mid-thoracic (MT) leads following low (A) and high (B) intensity stimulation. EP, entry point potential (note only in HL and MT); NI, Nil and NIII, negative peaks according to the order of increasing onset latencies (note there is no NII in MT); VR, ventral root potential. (Note the long duration summation potentials (unmarked) of opposite polarities in HL and MT.) Changes during increasing strength of stimulus (B) include disappearance of VR at LL, appearance of NII at LL and increase in the sizes of all waves, especially at the HL level.
A.P. Chandran et al. / B r a i n & Det,elopment 1994; 16." 44-51
46
were obtained. The tibial nerves were simultaneously stimulated bilaterally at the ankle with repetitive square wave pulses (0.3 Hz) of 0.1 ms duration and of adequate strength (0.7-1.2 mA) from the constant current stimulator of the EMG. The band pass of the filter was kept 50 H z - 2 kHz. One hundred responses were averaged with respect to stimulus onset. The analysis time of the averager was kept at 10 ms from the start of the stimulus. For any given trial, at least two series of records were taken; one just above the threshold strength of stimulus and the other at a higher strength, usually double the threshold strength. For the purpose of calculating the somatosensory conduction velocity, the onset latency of the major negative wave was noted in each SEP with the help of cursors. The distance between the recording electrodes was measured accurately with the help of calipers. The somatosensory conduction velocity between low-lumbar and mid-thoracic levels was calculated from the difference in latencies of SEP recorded from the two levels and the distance between the two recording points. (Although SEP from the high-lumbar level was also recorded, the distance between the high-lumbar and mid-thoracic recording points were too short to make any accurate measurement of somatosensory conduction velocity.) The peak latencies of the negative waves
EP
were also noted for the calculation of interpeak latencies.
3. RESULTS The spinal SEP in the 12-week-old mouse showed the following major features (Figs. 1-3). (i) A triphasic response (positive-negative-positive) with a predominant negative component (NI) at the low-lumbar level; (ii) a small positive wave followed by a large negative component with usually 3 peaks, NI, NII and NIII at the high-lumbar level; and (iii) two negative waves of relatively small amplitude (NI and NIII) but longer duration, usually separated by a positive wave of variable duration at the mid-thoracic level. At the low-lumbar level, the NI and NII had peak latencies of 0.88 _+ 0.03 (S.E.M.) ms and 1.5 _+ 0.21 ms, respectively, with an interpeak latency of 0.62 _+ 0.02 ms. At the high-lumbar level, the NI, NIl and NIII had peak latencies of 1.22 + 0.04 ms, 1.88 _+ 0.05 ms and 2.27 _+ 0.06 ms, respectively. The N I - N I I interpeak latency was 0.67 + 0.04 ms and the N I - N I I I interpeak latency was 1.05 _+ 0.03 ms. At the mid-thoracic level, the NI and NIII had peak latencies of 1.49 _+ 0.04 ms and 2.57_+0.05 ms, respectively. The NI-NIII in-
EP
A
NI
N I NII NI11
|
I
"
-|
o.o
,.s
•
.I I
I!~
I .vl I
I
.
.
.
.
.
.
• I I
l
_i
•
m
,
I mslD
I.
; I I
10.0 •
. . . . .
I ~d~ I
I
i '1
.
I
I .
I.L#'~-
llf\
•
m
,l
I.
ms
•
i
,.,'V
MT
I
.
ll!
I
I
I
.'!
11 0
B
•
l~un. I,
11.11"
1
m
.
.
•
.
.
.
.
r~T
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
HL I
20/uVlD t
I
Stimulus
9.90
mA
I Stimulus
1.60 mA
Fig. 2. Spinal SEP in a 12-week-old mouse. S a m e n o m e n c l a t u r e s as for Fig. 1. Note the well-defined N i l on high intensity s t i m u l a t i o n (B).
47
A.P. Chandran et al. /Brain & Development 1994; 16:44-51
EP
NI N n
A
EP
N I NII NIII
I l[ o!ov D/ .," i
I
B
I
[ 10'I0
. . . .
II .iv,,\ :'. i . . . . .
ill
I~~i i t~' :1. . ~ . .M i . .~ . H
. . . .
.
MT___
l' t
. . . .
•
m ~
HLLi
.~
. ML
LI',
, Stimulus
1.20 mA
Fig.3. SpinalSEPin a 12-week-oldmouserecordedat 4 levels,includingmid-lumbar(ML)level.Samenomenclaturesas for Fig.1. Notethe appearanceof Nil in LLand MLanddisappearanceofVRin LLon highintensitystimulation(B). terpeak latency was 1.08 + 0.02 ms which was not significantly different (P > 0.05) from that at the highlumbar level. In fact, at both the high-lumbar and mid-thoracic levels, the N I - N I I I interpeak latencies varied within a very narrow range (0.8-1.2 ms). The NI latencies progressively increased from the low-lumbar to the mid-thoracic levels (Figs. 1-3). The Nil latencies also increased progressively from the low-lumbar to the high-lumbar levels (Figs. 2 and 3). The N I - N I I interpeak latency at the high-lumbar level was significantly higher (P < 0.05) as compared to that at the low-lumbar level. Although the distance between the low-lumbar and the high-lumbar sites was, on average, twice that between the high-lumbar and midthoracic sites, the difference in NI latency between the former pair was only 1.3 times that between the latter pair. A detailed analysis of the spinal SEP showed the following additional features. (i) At both the highlumbar and mid-thoracic levels there was a small but definite negative wave just after the stimulus artefact (EP; Figs. 1 and 2). This wave had a constant latency at both of these levels, a 'standing potential'. (ii) At the high-lumbar level, the three negative peaks were superimposed over a large negative 'summation potential'
with a long duration (Figs. 1 and 2). This summation potential was present also at the mid-thoracic level, but with opposite polarity and relatively smaller amplitude. (iii) A comparative study of the onset latency of negative peaks in the high-lumbar and mid-thoracic SEP revealed that the two negative peaks of mid-thoracic SEP corresponded to the NI and NIII peaks in the high-lumbar SEP (Fig. 1). (iv) Two more small negative potentials were occasionally seen at the low-lumbar level (Fig. 1). One of them, Nil (Fig. 2), occurred immediately after the first major negative potential, NI. The other, a relatively smaller one VR (Fig. 1), occurred after a short interval. This late negative potential was abolished on high intensity stimulation (Figs. 1 and 3). 3.1. Effect of an increase
in stimulus
strength
The SEP was recorded always at two stimulus strengths, one just adequate to evoke discernible wave forms at all recording sites (0.7-1.2 mA), and the other almost double the strength of the first. All the wave forms showed a significant increase in amplitude on the high intensity stimulation (Figs. 1 and 2) except for the small amplitude wave forms in the mid-thoracic
A.P, Chandran et al. /Brain & Decelopment 1994; 16." 44-51
48
EP
NI Nn
2./Run. .
/
,I.
.
,I
I
I
20~ uVlID , • I I I
I. I I
•I
I . I
!
!
,
I. I I
I
./.;
°
N nl
,
1 ms/D 1 0 . 0 ms
I I
.I
I
I
,
MT
only in 5 (out of 11) mice in the 9-week-old group. (ii) The size of the NIII peak in the high-lumbar SEP was relatively smaller as compared to that in 12-week-old mice. The decrease in size was greater in the midthoracic SEP (Fig. 4). (iii) There was a significant reduction in the size of summation potential especially in the mid-thoracic SEP. 3.3. Somatosensory conduction velocity
::[
,
# u,lD,
#
HL
~V/D t. I
•L L
I Stimulus
1.80 mA
Fig. 4. Spinal SEP in a 9-week-old mouse. Same nomenclatures as for Fig. 1. Note the well-marked entry point potential (EP) even at the LL level.
SEP, which showed relatively smaller increase in amplitude. The extent of the increase in the amplitude of the negative peak NI in the low-lumbar and highlumbar SEP was the same. The Nil peak also showed a significant increase on the high intensity stimulation. In some mice in which SEP was recorded additionally from the L2-L3 intervertebral space ('mid-lumbar'), the Nil peak appeared just after NI on the high intensity stimulation (Fig. 3B). The increase in amplitude of Nil in the high-lumbar SEP was greater as compared to that of NII both at the low-lumbar and mid-lumbar levels (Fig. 3B). The summation potential also showed a significant increase in amplitude, more so at the high-lumbar levels (Figs. 1 and 2). The long latency response at the low-lumbar level, VR, disappeared on the high intensity stimulation (Figs. 1 and 3). 3.2. Spinal SEP in 3-, 6- and 9-week-old mice The SEP in the younger mice (3, 6 and 9 weeks; Fig. 4) showed the following significant differences from that in 12-week-old mice. (i) The standing potential was observed even in the low-lumbar SEP invariably in all mice belonging to the 3- and 6-week-old groups, but
The somatosensory conduction velocity at 12 weeks was 36.0_+ 1.0 (S.E.M.) m / s . It was only 17.1 _+ 0.8 m / s at 3 weeks, showing a highly significant increase ( P < 0.001) in the somatosensory conduction velocity during the 3-12 week period. At 6 weeks and 9 weeks, the somatosensory conduction velocity was 28.2 ± 0.8 m / s and 35.6 ± 1.3 m / s , respectively, showing a highly significant increase ( P < 0.001) during the 3-6 and 6-9 week periods. There was no significant increase in the somatosensory conduction velocity after 9 weeks.
4. DISCUSSION Recording of spinal SEP following stimulation of the tibial nerve at the ankle is considered to be a difficult procedure due to the presence of relatively less numbers of muscle afferents in the nerve, that are shown to contribute the most to the genesis of spinal SEP [7-9]. In the present work, however, such a procedure had to be resorted to because stimulation at a more proximal site (popliteal fossa), which is usually preferred in man and larger animals, could have drastically reduced the available distance between the stimulating and recording electrodes, especially in young (3and 6-week-old) mice. Furthermore, in order to compensate for the possible reduction in the size of SEP, bilateral simultaneous stimulation of the tibial nerves at the ankle was employed in the present study. Such a procedure has been shown to improve the quality and amplitude of SEP recordings [2,10]. An attempt has also been made to improve the signal-to-noise ratio using needle electrodes introduced subdermally between the spinous processes. The relative merits and demerits of using monopolar and bipolar recording methods have already been recognized [10]. In the present work, wherein the main stress was placed on the study of wave forms, their sources and spinal conduction, the monopolar method of recording was preferred. Bilateral stimulation of the tibial nerves at the ankle produced well-defined triphasic SEP with a predominant negative wave (NI) at the low-lumbar level, a complex potential with 3 negative peaks (NI, Nil, NIII) superimposed over a large summation potential at the high-lumbar level, and 2 negative waves (NI, NIII)
A.P. Chandran et al. /Brain & Development 1994; 16:44-51
superimposed over a positive summation potential at the mid-thoracic level (Figs. 1 and 2). There was a progressive increase in the onset latency of NI wave from the low-lumbar to the mid-thoracic levels, indicating that it is a travelling wave. On high intensity stimulation, the degree of increase in amplitude of the NI waves in both the low-lumbar and high-lumbar SEPs was similar. The NI peak being the first component of SEP, is most probably due to conduction in large diameter muscle afferents of the dorsal root [2,8]. The progressive increase in the latency of NII in the low-lumbar, mid-lumbar (when recorded) and highlumbar SEPs suggests that it is also a travelling wave conducted along the dorsal root. However, since the N I - N I I interpeak latency at the high-lumbar level was more than that at the low-lumbar level ( P < 0.05), it might be presumed that NII peak is due to conduction along fibres with lesser conduction velocity as compared to the large diameter muscle afferents that produced the NI peak. The NII could be elicited only at a higher stimulus strength, especially at the low-lumbar level (Fig. 2). Further, there was a marked increase in the amplitude of NII at the high-lumbar level (Fig. 3) due to a possible summation effect following bilateral stimulation [11]. All these observations make it highly probable that the NII peak is due to conduction along large diameter cutaneous afferents. There was no wave in the mid-thoracic SEP corresponding in latency to the NII of the high-lumbar recording (Figs. 1-3). It can be assumed that the impulses that cause NII in the caudal recordings are either not conducted rostrally or that they do not contribute sufficiently to SEP peaks at the mid-thoracic level, where records are basically of lower amplitude [2,8]. Cracco et al. [2] observed in man that sural nerve stimulation failed to produce any recordable potential in the thoracic leads. Assuming that the NII peak in the high-lumbar SEP is formed by large diameter cutaneous afferents, as suggeted above, it is quite possible that the NII peak is hardly recordable at the midthoracic level. Alternatively, it is also possible that most of the fibres contributing to the genesis of NII at the high-lumbar level are involved in spinal reflexes [11] and that only a few fibres ascend in the dorsal column to generate a recordable NII peak at the mid-thoracic level. It is quite possible that the NIII peak has its origin close to the root entry zone, since there was no peak corresponding to it in the caudal recordings and it appeared first in the high-lumbar recording (Figs. 1 and 3). Since the interpeak latency between NI and NIII in the high-lumbar recording was almost 1 ms and remained within narrow limits in all the recordings, it is quite possible that the NI wave might have dissociated into NI and NIII peaks at the high-lumbar level, the NI peak being presynaptic and NIII post-synaptic.
49
The anatomical correlates of this contention is that the muscle afferents in the dorsal root travel in both dorsal white column (NI, presynaptic) and Clarke's c o l u m n / spinocerebellar tract (NIII, post-synaptic), after they enter the spinal cord. The second negative peak in the mid-thoracic SEP corresponded to the NIII peak of the high-lumbar SEP (Figs. 1-3). As the interpeak latency between NI and NIII in the mid-thoracic SEP is almost the same as that at the high-lumbar recording, it is possible that the fibres responsible for NI and NIII waves have almost the same conduction velocity (although the distance between the high-lumbar and mid-thoracic recording sites was too short to make any definite conclusion). The possible origins of NI and NIII peaks at the high-lumbar level being dorsal white column and Clarke's column, respectively, both comprising proprioceptive fibres, as discussed above, it is likely that the NI and NIII peaks at the mid-thoracic level originate in dorsal white column and spinocerebellar tracts, respectively. A small negative wave with a relatively long latency was also recorded in the low-lumbar SEP in some 12-week-old mice (Figs. 1 and 3). The observation that this potential disappeared on the high intensity stimulation (Fig. 1, 3) indicates that it could be the ventral root potential as had been reported earlier [6,9]. The fact that it could be recorded only in a few animals may be due to the relatively small number of proprioceptive afferents in the tibial nerve at ankle. Alternatively, it may well be due to simultaneous stimulation of the adequate number of motor fibres that may antidromically occlude the ventral root potential at the low-lumbar level [9]. The high-lumbar SEP was of a more complex configuration and larger as compared to caudal or rostral recordings (Figs. 1-3). Such an SEP with maximal amplitude and complex configuration had been reported at T12 also in man [2,10]. The large negative wave with long duration (summation potential) could possibly be a summated activity of the 3 negative peaks, NI, Nil and NIII. However, this summation potential had an opposite polarity in the mid-thoracic SEP, although NI and NIII had negative polarities like those in the high-lumbar SEP, suggesting that the summation potential possibly had a separate generator. It appears that the afferent impulses cause the development of certain current sources, most probably due to intense synaptic activity. The reversal of polarity indicates that it is a fixed generator located between T7 and T12 vertebral levels similar to the source that produces N13 in the human SEP following median nerve stimulation [12,13]. It is generally accepted that the standing potential represents the junctional potential due to the entry of dorsal root potential into the spinal c o r d - the 'entry
50
A.P. Chandran etal./Brain & Development 1994; 16." 44-51
point potential' [2,3,8]. The appearance of a well-defined entry point potential at the high-lumbar level in the present study (Figs. 1-3) suggests that, in the mouse, the S1 nerve root enters the spinal cord at T13 vertebral level, like as reported earlier in the rat [14]. In the present study, 3 major changes have been observed in the SEP pattern during postnatal development. (i) The entry point potential was present even at the low-lumbar level invariably in all 3- and 6-week-old mice and in half the number of 9-week-old mice (Fig. 4). It is possible that in younger mice ( < 9 weeks old) the S1 root enters at a lower level as compared to the adult mice. This corroborates well with the finding that the spinal cord ends at a lower level in younger rats so that the spinal segments correspond to lower vertebral levels as compared to adult rats [15]. The axial growth of the vertebral column during postnatal life may be the most probable reason for this developmental change. (ii) Another important difference oberved in the present study was the reduced amplitude of NIII in the high-lumbar and mid-thoracic SEP. Since, as discussed earlier, the NIII peak depends upon the synaptic activity in Clarke's column and conduction in the spinocerebellar tract, it is quite likely that the decreased amplitude of NIII is a consequence of decreased synaptic activity during early postnatal life. In fact, it had been reported that many collaterals in the spinal cord develop only postnatally and the synaptic activity shows a progressive development during postnatal growth in cats [16] and mice [17]. (iii) A reduction in the amplitude of summation potential especially in the mid-thoracic SEP, in the present study, also suggests a decreased synaptic activity during early postnatal life. Alternatively, the degree of myelination being negligible in newborn mice [18], the reduced conduction in the spinocerebellar tract may also contribute to the reduced size of the NIII peak. In the present study, the somatosensory conduction velocity was calculated from the latency differences between the SEP recorded at the low-lumbar and mid-thoracic levels. (A comparison between the caudal and rostral spinal conduction velocities was planned, but the distance between the high-lumbar and midthoracic recording sites was too short to attempt any meaningful comparison.) Since the entry point of S1 root was at the T13 vertebral level [14], the major part of the conduction velocity is along the dorsal root. However, in the younger mice, wherein the spinal levels correspond to lower vertebral levels, the conduction measured between the low-lumbar and midthoracic levels may involve more spinal component as compared to the adult mice. A comparison of the onset latencies of NI peaks in the low-lumbar, high-lumbar and mid-thoracic SEPs definitely suggested a significant decrease in the somatosensory conduction velocity across the high-
lumbar-mid-thoracic recording sites (although the decrease could not be quantified accurately due to the nearness of the two recording sites). This is expected as the fibres across the low-lumbar-high-lumbar recording sites are thicker due to their proximity to the dorsal root ganglion. The somatosensory conduction velocity showed a highly significant increase during the 3-6 and 6-9 week periods indicating a significant degree of myelination of the proprioceptive fibres in the dorsal root and dorsal white column. It is possible that, like in rats [18,19], myelination in mice is also almost a postnatal phenomenon. This may make mice more vulnerable to demyelinating diseases of juvenile onset and of the progressively degenerating type as compared to man. The present study is the first report on spinal SEP in mice. From the findings it may be concluded that conduction along the muscle afferents, cutaneous afferents and spinocerebellar tract produced separate peaks in spinal SEP. A fixed generator was identified, located between T7 and T13. The root entry of $1 is at T13 in adult mice, but in younger ones it is lower down, possibly due to less somatic growth. The reduced size of certain waves observed in the present study in young mice, especially in rostral recordings, indicated less synaptic activity. During the early postnatal life (3-9 weeks), there was a significant increase in the somatosensory conduction velocity indicating a substantial amount of myelination during this period.
5. REFERENCES 1. Chiappa KH, Yiannikas, C. Evoked potentials in clinical medicine. New York: Raven Press, 1983. 2. Cracco RQ, Cracco JB, Sranowski R, Vogel HB. Spinal evoked potentials. In: Desmedt JE, ed. Progress in clinical neurophysiology. VoL 7. Basel: Karger, 1980: 87-104. 3. Kimura J. Electrodiagnosis in d&eases of nerve and muscle: principles and practice. 2nd ed. Philadelphia: Davies, 1989. 4. Happel LT, Leblanc HJ, Kline DG. Spinal cord potentials evoked by peripheral nerve stimulation. Electroencephalogr Clin Neurophysiol 1975; 38: 349-54. 5. Rossini PM, Greco F, De Palma C, Pisano k Electrodiagnosis of the rabbit. Monitoring of spinal conduction in acute cord lesions versus clinical observation. Eur Neurol 1980; 19: 409-13. 6. Weitholter H, Hulser PJ. Lumbar spinal somatosensory evoked potentials in the rat after stimulation of the tibial nerve. Exp Neurol 1985; 89: 24-31. 7. Gandevia SC, Burke D, McKeon B. The projection of muscle afferents from the hand to cerebral cortex in man. Brain 1984; 107: 1-13. 8. Halonen JP, Stephen J J, Edgar MA, Ransford AD. Conduction properties of epidermally recorded spinal cord potentials following lower limb stimulation in man. Electroencephalogr Clin Neurophysiol 1989; 74: 161-74. 9. Yiannikas C, Shahani BT. The origins of lumbosacral evoked potentials in h u m a n s using a surface electrode technique. J Neurol Neurosurg Psychiatry 1988; 51: 499-508. 10. Kakigi R, Shibasaki H, Hashizume A, Kuriyowa Y. Short latency
A.P. Chandran et al. / Brain & Development 1994; 16." 44-51
11.
12.
13.
14.
somatosensory evoked spinal cord and scalp recorded potentials follwing posterior tibial nerve stimulation in man. Electroencephalogr Clin Neurophysiol 1982; 53: 602-11. Meinck HM, Benecke R, Kuster S, Conrad B. Cutaneomuscular organization in normal man and in patients with motor disorders. In: Desmedt JE, ed. Motor Control mechanisms in health and disease. New York: Raven press, 1983: 787-96. Allison T, Wood CC, McCarthy G, Hume AL, Goff WR. Short latency somatosensory evoked potentials in man, monkey, cat and rat: comparative latency analysis. In: Mauguiere F, Revol M, eds. Clinical application o f evoked potentials in neurology. New York: Raven Press, 1982: 303-10. Desmedt JE, Cheron G. Prevertebral (oesophageal) recording of subcortical somatosensory evoked potentials in man: the spinal P13 component and the dual nature of the spinal generators. Electroencephalogr Clin Neurophysiol 1981; 52: 257-75. Pamphlett RS. Spinal irritation does not inhibit distal axonal sprouting. Muscle Nerve 1988; 11: 493-501.
51
15. Brunquell PJ, Taylor GW, Holmes GL, Feldman DS. Ontogenesis of lumbar somatosensory evoked potentials after posterior nerve stimulation in the rat. Electroencephalogr Clin Neurophysiol 1990; 77: 112-8. 16. Skoglund S. The activity of muscle receptors in kitten. Acta Physiol Scand 1960; 50: 203-21. 17. Chandran AP, Oda K, Shibasaki H. Changes in motoneuron excitability during postnatal life in the mouse. Brain Dev (Tokyo) 1991; 13: 180-3. 18. Jacobson S. The sequence of myelination in the brain of albino rat. J Comp Neurol 1963; 121: 5-29. 19. Schonbach J, Hu KH, Friede RL. Cellular and clinical changes during myelination: autoradiographic, histological and biochemical data on myelination in the pyramidal tract and corpus callosum. J Comp Neurol 1968; 134: 21-38.