Brain Research, 586 (1992) 36-43 © Igg2 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00
36
BRES 17838
Identification of nociceptive neurons in the medial thalamus: morphological studies of nociceptive neurons with intracellular injection of horseradish peroxidase Kenji Sugiyama, Hiroshi Ryu and Kenichi U e m u r a Department of Neurosurgery, tlamamatsu Unirersity School of Medicbw, Handa-cho, itamamatsu (Japan) (Accepted 29 January 1902)
Key words: Wide dynamic range neuron; Tap neuron; Medial thalamus; Parafascicular nucleus; Subparafascicular nucleus; Horseradish l~¢roxidase study; lntracellular injection of horseradish peroxidase
Somatosensory neurons including nociceptive ones in the medial thalamus have been studied extracellularly, and are classified into three types: nociceptive-specific, wide dynamic range, and tap neurons. However, the morphological characteristics of these three neurons have not yet been clarified We studied the morphological characteristics of the neurons by iontophoretic injection of HRP into single neurons in 32 cats. Nine wide dynamic range neurons and two tap neurons were electrophysiologically identified and successfully stained with HRP in and around the parafascicular and subparafascicular nuclei, and in the mcdiodorsal nucleus. All nine wide dynamic range neurons had fewer dendrites which formed scanter tufts, whereas the two tap neurons had many more dendrites which formed denser tufts: i,e, wide dynamic range neurons were of the isodc:ndritic type, and the tap neurons were of the allodendritic type, according to Ram6n-Moliner's classification. The axons of the two tap neuron~ run antero-laterally whereL,s those of wide dyr,.amic range neurons ran in many directions, These morphological differences suggest that lap net,ron,~ may have more specialized and fixed ft, nctions than wide dynamic range net,rons.
INTRODUCTION Somatosensory neurons including nociceptive ones in the medial thalamus have been studied extrac¢llu. hlrly in cats '~'',~7,~~.'~,rats ".~'s,rabbits r' and monkeys '~J~,~'~. They have been located mainly in and around the parafascicular, subparafascicular, central lateral, and mediodorsal nuclei T M ~,~'~.These neurons are classified into three types: nociceptive, wide dynamic range and tap neurons a,'~J~.2:. Different from the neurons in the ventral posterior nuclei, they have broad, bilateral receptive fields, Nociceptive and wide dynamic range neurons have another unique characteristic, the generation of long-lasting discharges after noxious stimuli a0plicd to the periphery .~.4.~','~,~1.14,2:,Clinically, some have reported contralateral pain relief by making lesions in these nucleiC'), and others have reported bihtteral pain relief by stimulating them 27, although the mechanisms of pain relief have not yet been clarified, Many have studied these medial thalamic, non-
specific nuclei histologically by use of degeneration techniques ~'~, intra-tissue injections of tritiated amino acids ~'~'~ or of HRP ?''"'''~''-''j''~=, and other methods 2'2.'=". However, the morphological characteristics of these three different neuronal types have not yet been clarified. We studied the morphological characteristics of neurons in the medial thalamus by iontophoretic injections of HRP into single neurons, with the hope of clarifying a part of the mechanism for the perception of noxious information, MATERIALS AND METHODS Thirty-two cats weighing between 2.5 and 4.0 kg were used in this experiment. They were anesthetized initially by intra-peritoneal injection of sodium pentobarbital (50 mg/kg), followed by inhalation of halothane (0..~%), immobilized with suxamethonium chloride (I mg/kg/h), and artificially ventilated. Blood pressure, end-tidal CO 2 and rectal temperature were monitored and maintained within the physiological range. A craniectomy was performed and a part of the parietal lobe was aspirated to the ventricle to permit stereotactic placement of recording micropipettes into the thalamus. The sural
(~rreslxmd(,nce: K. Sugi)ama, Department of Neurosurgery, Hamamatsu University School of Medicine, 3600 Handa-cho, Hamamatsu 431-31, Japan.
37 nerve on one side was exposed for stimulation and recording of the compound action potentials, These experimental procedures were carried out under the ethical guidelines made by the Physiological Society of Japan 30 and by the Committee for Research and Ethical Issues of IASP (The International Association for the Study of Pain)3-~. Glass micropipetles for recording and HRP injection were filled with 7% HRP (Toyobo Co., Osaka, Japan) in 0.05 M Tris-HCI buffer (pH 8.6) with 0,2 M KCI. The electrode resistance varied from 15 to 40 MD. The sural nerve was stimulated with stainless steel electrodes (rate !/s, duration 300 p.s, interval 5 ms, train 3, intensity about 800/zA) and the compound action potentials were recorded with the method of Dong et alfl. A stimulus intensity of 800/zA was used to produce noxious stimulation which was confirmed by monitoring A~ and C fiber compound action potentials. Mechanical stimuli such as tap or pinch were also applied to the periphery to identify the characteristics of the neurons. The responses of the thalamic neurons were stored on magnetic tapes and analyzed later with a signal processor (7T08 NEC San-el Co,, Tokyo, Japan). HRP was injected iontophoretically through the recording electrode. A depolarizing constant current with a pulse duration of 200 ms and an intensity of 10-20 nA was passed through the electrode at 2.5 pulses/s for 7-15 min. The identity of the recorded spikes was checked every 2 rain during HRP injection. The cats were sacrificed 3-10 h after the HRP injection, The animal was perfused transcardially with 0.I M phosphate buffer solution and then with 2'% paraformaldehyde in 0.I M phosphate buffer (pH 7,4). The brain was removed and stored for 48 h in a 0.I M phosphate buffer solution containing 30% sucrose at 4°C. Serial transverse sections (75./zm thickness) were treated for HRP with the
Ni-Co enhanced diaminobenzidine method (Adams' method)I. The sections were mounted on gelatin-coated slides, air-dried for over 5 days, infiltrated with acetone, and microscopically examined. Drawings were prepared at a magnification of × 400 or x 1000, using a microscope. The sections were then counterstained with Cresyl violet and the neurons were located.
RESULTS
A total of 125 units were isolated and were classified into 4 groups. Twenty-one out of 125 units were 'tap units' which were driven only by innocuous brisk taps. Eighteen units were 'nociceptive units' which were driven only by noxious stimuli. Twenty-seven units were 'wide dynamic range units' (or 'noxious-tap units') which were driven by both noxious and tap stimuli. The remaining 59 units were termed 'spontaneous units', since they showed only spontaneous discharges and no time-locked responses to any stimuli. Eleven neurons out of 125 units were succe~s~tully stained with HRP. Nine out of the ll neurons were considered to be 'wide dynamic neurons', and the rest were thought to be innocuous 'tap neurons'. We could
Electric stimulation A z16----~,. . . . . . . . . . . . . .
100
m80¢,
C
;,
i.i,
Pinch
32~
Iltlm,
•
•
I
•
m
........ ,
\
\, m
ioo )Jm
\
1 8ec
Tap 32w sum.
•
II
•
II
K807265f
|
IJ
~
/
Fig. 1. A wide dynamic range neuron found in the parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram shows initial discharges and long after discharges with electric stimulation ( zx). The three tall vertical lines are shock artifacts of electric stimuli. This neuron also responded to both noxious (pinching) and innocuous (tapping) stimuli. The small bars indicate the period of each stimulation. B: the location of the neuron ( • ) . See Abbreviations section. C: camera lucida drawing of the neuron stained with HRP in the coronal plane. The neuron has fewer dendrites and scanter tufts than the tap neurons. These dendrites show the bipolar alinement. Its axon (arrow) runs in the antero-ventro-lateral direction.
38
A
Electric stimulation ~t
bl
t
LP
Q.
0
~,m.
hi I
o O-
100 msec
I -
Y'I 1
Fr, 6.5
Pinch 32'
0 1 sec
Tap
321
stl~,
I
I
K903165b
•
O
ioo~m
1 II0¢
Fig. 2. A wide dynamic range neuron found on the border of the parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram shows initial discharges, long after.discharl~es and inhibition with electric stimulation ( .. ). The three tall vertical lines are shock artifacts of electric stimuli. This neuron also responded to both noxious (pinching) and innocuous (tapping) stimuli, although the responses to the noxious stimuli ,re I,rger than those to the innocuous ones. The small bars indicate the period of ca~:lz .~timulation. B: the Iocutkm of the neuron ( # ). See Abbreviations section. C': camera lucida drawing of the neuron stained with HRP in the coronal plane. The neuron has fewer dendrites lind scanter tufts than the tap neurons. These dendrites show the phme alinement. Its axon (~lrrow) runs in the dorso-l,teral direction.
TABLE !
Summary of the morphologk'al differences between wide dynamic range and tap ne~rmes The density of dendritic spines showed no statistical difference between the two groups,
Wide dynamic Number of neurons Locations of neurons Ram6n-Moliner's classification Number of dendrites Tuft formation Number of dendritic spines(/10/~m) Density of dendritic spines(/l0 ~m) Axonal direction
Pf 5 ( I ) * Spf 2 ( I ) * * isodendritic type fewer scanter 180-250 417±0,6 antero-lateral antero-dorso-lateral dorso-lateral postero-ventro-medial ventro-medial
* A wide dynamic range neuron found on the lateral border of the Pf, A wide dynamic range neuron found in the ventral area adjacent to the Spf,
* *
Tap nellron ranRe neuron
! 2 I I I
2 MD i Spf ! allodendritic type richer denser 900-1100 5.3:1:0.7 antero-lateral 2
39
A
Electric stimulation _z 1 6 - - ¥ m
/,
O.
0
o 0l l l , i
L
&,,_k_~.
'al
I
',
i
100 m s e c
Pinch 32
/~i,~e!
'... j,d
Fr. 7.0
G
o J UklLii. h.-.a
1 SeC
Tap 321 [tim. u
_ _ 100 pm
K902147c
I tee
Fig. 3. A wide dynamic range neuron found in the parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram show initial discharges and after-discharges with electric stimulation ( A ). This neuron showed poor after-discharges, although it also responded to both noxious (pinching) and innocuous (tapping) stimuli, which proved this neuron to be a wide dynamic range neuron. The small bars indicate the period of each stimulation B: the location of the neuron ( • ). See Abbreviations section. C: camera lucida drawing of the neuron stained with HRP in the coronal plane. The neuron has fewer dendrites and scanter tufts than the tap neurons. These dendrites show the bipolar alinement. Its axon (arrow) runs in the postero-ventro-medial direction.
not inject HRP successfully into purely nociceptive neurons in this experiment. Figs. 1-5 show drawings of five wide dynamic neurons, their unit activities, and their locations in the thalamus. Figs. 6-7 show drawings of two tap neurons, their unit activities, and their locations. Among the nine wide dynamic neurons, six were found in and around the parafascicular nucleus and three were in and around the subparafascicular nucleus. Two tap neurons were found in the mediodorsal and the subparafascicular nuclei. The unit activities of wide dynamic neurons showed initial discharges of 30-50 ms after the electric stimulation followed by long after-discharges lasting for 200-500 ms. These neurons also responded to mechanical stimuli including both pinching and tapping (Figs. IA-5A). The unit activities of tap neurons, on the other hand, showed only initial discharges ~,~ter the electric stimulation and no response to pinching (Figs. 6A and 7A). The HRP study revealed morphological differences between wide dynamic and tap neurons (Figs. 1C-7C, Table I). Wide dynamic neurons had fewer dendrites which formed scanter tufts, whereas tap neurons had many more
dendrites which formed denser tufts. The number of dendritic spines were about 180-250 on one dendrite in wide dynamic neurons, and about 900-1,100 in tap neurons. However, there was no statistically significant difference in the density of dendritic spines between wide dynamic (4.7 + 0.6/10 ~m) and tap neurons (5.3 + 0.7/10 pm). The dendritic spines were omitted in the drawings. The axons (arrow in the figures) were identified morphologically by the absence of dendritic spines. The axons of the two tap neurons were seen to run antero.laterally, whereas those of wide dynamic neurons run in antero-lateral, antero-dorso-lateral, dorso-lateral, postero-ventro-medial and ventro-medial directions, respectively.
DISCUSSION Histological studies of these medial thalamic, nonspecific nuclei have used degeneration techniques 13, intra-tissue injections of tritiated amino acids 12'e'~or of HRP 7.1°,!6.e1.29,32, and other methods e'2°'es. Anatomical
40
A
Electric stimulation ,
z16- ? L_~|~ IL~
"0N,,L
o
,-
'
L,L
Puk
.
LP
\
'
p
i ~ .~ ,,,., =,.bh~h kBi=~ 100 msec
Fr. 7.5
.. / s
Pinch
_f
32 I Ibm i
I
sec
Tap 32~-
W
tOO~ m
OA._,I,,__L_L_I, .I.L_JL, LI, K903164d
t lee
\
Fig. 4. A wide dynamic range neuron found in 1he parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli, Pulse density histogram shows initial discharges and long after.discharges with electric stimulation ( ,x ), The three tall vertical lines are shock artifilcts of electric stimuli, This neuron also responded to both noxious (pi.ching) and innocuous (tapping) stimuli, .ltht~ugh the re,,,ponses to the irlnocuotis stimuli are sm;dler than those to noxiotis ones, The small h.rs indic.re the p~riod of each stimulatio.. B: the location of the neuron ( $ ), Scc Abbrcviati;)lls section, (': c.mcra lucid, drawing of the: netffon stained with IIRP in the coronal plane, The .euron has fewer dendrites =rod s~:.nter tufts tha. the t.p neurons, These dcndrltes show the htpohtr ttllncntettt, Its axon (arrow) ru.s in the vcntro-medial direction.
connections to and from tile medial thalamus have been described in detail. For example, the parafascicular nucleus receives inputs from bulbothalamic and spinothalamic pathways, from the substantia nigra pars rcticulata, parabrachial pontinc nucleus, periaqueductal gray, and deep layers of the superior colliculus, and provides efferent projections to the cerebral cortex, striatum and subthalamic nucleus. Based on these findings and on the electrophysiological characteristics, it has been speculated that nociceptire neurons in the medial thalamus are related to either the affective aversive aspect of the nociceptive stimuli aas''-a, the scnsorimotor responses to noxious stimuli:", or the arousal system :,~s. in this experiment, we tbund many morphological differences between wide dynamic range neurons and tap neurons in the medial thalamus. The main differences are the spatial distribution of their dendrites and the axonal directions. The difference in the number of
dendritic spines seems to depend on the number of branches of the dendrites, because the density of dendritic spines showed no statistical difference between these two neurons. Ram6n-Molincr classified vertebrate neurons into three categories according to the morphology of their dendrites ''~'2~'. The neurons include isodendritic, aliodendritic, and idiodendritic types. The isodendritic type is characterized by few relatively straight dendrites and scanter tufts, which tend to orient themselves along a given plane or to diverge to two opposite directions. The ailodendritic type is defined as more specialized neurons than the isodendritie type, but not so highly specialized as the idiodendritic type. Neurons of this latter type are richer in dendrites and have denser tufts than the isodendritic ones, and are also are divided into several subgroups depending on their locations. The idiodendritic type is defined as the most specialized neurons. Each neuron of this type depends so
41
A
Electric stimulation '~
z 16 nW
8 o,1-" _.IJI J L ~ , ~ . ~ . ~ , ~ 100 msec
Fr. 8.0
\
'
Pinch 32~ slim. o
.... J M , . j , , k l
,.
l',..iJb
....
O 1 SOC
Tap
32~
s~m. =
•
I
•
I
I
o '~ ,u.ul~, II, I,, lu.,u.A K812226d i1 8 e c
i
100 um
Fig, 5. A wide dynamic range neuron fcand in the area adjacent to the subparafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram shows initial discharges and long after-discharges with electric stimulation ( A ). The three tall vertical lines are shock artifllcts of electric stimuli. Although this neuron shows poor after-discharges, it also responded to both noxious (pinching) and innocuous (tapping) stimuli, which proved it to be a wide dynamic range neuron. The small bars indicate the period of each stimulation. B: the location of the neuron ( • ) . See Abbreviatiozls section. C: camera lueida drawing of the zleuron stained with HRP in the coronal phme. The neuron has l'~wer dendrites ,nd scanter tufts thai| th~ tap neurons. Its axon (arrow) runs in tile antero-dorso.lateral direction.
specifically on its location that one can easily identify the location even by the shape of any single neuron of this type. In our experiment, all of nine wide dynamic range neurons in the medial thalamus were of the isodendritic type: Four showed bipolar orientation of the dendrites and two others tended to orient themselves along a plane. The two tap neurons were of the allodendritic type. Their axons tended to run anterolaterally, suggesting that they project their axons to the same target and play the same role, though further studies are necessary to draw a definite conclusion. The axons of wide dynamic range neurons, on the other hand, tended not to run in a fixed single direction but rather in many directions. This seems to
indicate that wide dynamic range neurons have connections not only with the ascending reticular system but with many other systems, including the hypothalamus or the limbic system, it is very interesting that, as Ram6n-Moliner pointed out, the neurons of the hypothalamus or of the reticular formation belong to the isodendritic type and the neurons of specific thalamic relay nuclei, such as ventral posterior nuclei, to the allodendritic type 26. These findings may suggest that tap neurons are more specialized neurons and have more distinct and fixed functions than wide dynamic range neurons in the somatosensory system, and that tap and wide dynamic range neurons play quite different roles in the medial thalamus.
42 Electric stimulation A| ~e---~
-~-! ..... 1~
~ -~--~.
,.p "t, " " / } ,
..o,~
i
I+ ! CL i
~
k ~.~'i
i.A .,~., o
0''
'
~-T
|
I
i 100 mse¢
Zl
_
x°"'~
'~
Hyde//
--'--
Fr. 8;0
Pinch
321
a
0i
I
h
|o 1 II0¢
l
Tap
/
32r ,%~..
.
.
.
.
.
.
.
i
K807054b
,•.=
Fig, 6. A tap neuron found in the subparafascicular nucleus. A: responses of the tap neuron to electric and mechanical stimuli. Pulse density histogram shows only initial discharges with electric stimulation (~). This neuron responded only to innocuous (tapping) stimuli, and did not respond to noxkius (pinching) stimuli. These findings proved this neuron to be a tap neuron, The small bars indicate the period of each stimulation, B: the location of the neuron (o), See Abbreviations section. C: c:lmera lucida drawing of the neuron stained with HRP in the coronal plane, The neuron has richer dendrites and denser tufts than the wide dynamic range neurons. Its axon (arrow) runs in the antero-latend aspect,
A
Electric stimulation
BI
).......... ,.; ...... ~>i ~
~,
-r"°io
36
BRES 17838
Identification of nociceptive neurons in the medial thalamus: morphological studies of nociceptive neurons with intracellular injection of horseradish peroxidase Kenji Sugiyama, Hiroshi Ryu and Kenichi U e m u r a Department of Neurosurgery, tlamamatsu Unirersity School of Medicbw, Handa-cho, itamamatsu (Japan) (Accepted 29 January 1902)
Key words: Wide dynamic range neuron; Tap neuron; Medial thalamus; Parafascicular nucleus; Subparafascicular nucleus; Horseradish l~¢roxidase study; lntracellular injection of horseradish peroxidase
Somatosensory neurons including nociceptive ones in the medial thalamus have been studied extracellularly, and are classified into three types: nociceptive-specific, wide dynamic range, and tap neurons. However, the morphological characteristics of these three neurons have not yet been clarified We studied the morphological characteristics of the neurons by iontophoretic injection of HRP into single neurons in 32 cats. Nine wide dynamic range neurons and two tap neurons were electrophysiologically identified and successfully stained with HRP in and around the parafascicular and subparafascicular nuclei, and in the mcdiodorsal nucleus. All nine wide dynamic range neurons had fewer dendrites which formed scanter tufts, whereas the two tap neurons had many more dendrites which formed denser tufts: i,e, wide dynamic range neurons were of the isodc:ndritic type, and the tap neurons were of the allodendritic type, according to Ram6n-Moliner's classification. The axons of the two tap neuron~ run antero-laterally whereL,s those of wide dyr,.amic range neurons ran in many directions, These morphological differences suggest that lap net,ron,~ may have more specialized and fixed ft, nctions than wide dynamic range net,rons.
INTRODUCTION Somatosensory neurons including nociceptive ones in the medial thalamus have been studied extrac¢llu. hlrly in cats '~'',~7,~~.'~,rats ".~'s,rabbits r' and monkeys '~J~,~'~. They have been located mainly in and around the parafascicular, subparafascicular, central lateral, and mediodorsal nuclei T M ~,~'~.These neurons are classified into three types: nociceptive, wide dynamic range and tap neurons a,'~J~.2:. Different from the neurons in the ventral posterior nuclei, they have broad, bilateral receptive fields, Nociceptive and wide dynamic range neurons have another unique characteristic, the generation of long-lasting discharges after noxious stimuli a0plicd to the periphery .~.4.~','~,~1.14,2:,Clinically, some have reported contralateral pain relief by making lesions in these nucleiC'), and others have reported bihtteral pain relief by stimulating them 27, although the mechanisms of pain relief have not yet been clarified, Many have studied these medial thalamic, non-
specific nuclei histologically by use of degeneration techniques ~'~, intra-tissue injections of tritiated amino acids ~'~'~ or of HRP ?''"'''~''-''j''~=, and other methods 2'2.'=". However, the morphological characteristics of these three different neuronal types have not yet been clarified. We studied the morphological characteristics of neurons in the medial thalamus by iontophoretic injections of HRP into single neurons, with the hope of clarifying a part of the mechanism for the perception of noxious information, MATERIALS AND METHODS Thirty-two cats weighing between 2.5 and 4.0 kg were used in this experiment. They were anesthetized initially by intra-peritoneal injection of sodium pentobarbital (50 mg/kg), followed by inhalation of halothane (0..~%), immobilized with suxamethonium chloride (I mg/kg/h), and artificially ventilated. Blood pressure, end-tidal CO 2 and rectal temperature were monitored and maintained within the physiological range. A craniectomy was performed and a part of the parietal lobe was aspirated to the ventricle to permit stereotactic placement of recording micropipettes into the thalamus. The sural
(~rreslxmd(,nce: K. Sugi)ama, Department of Neurosurgery, Hamamatsu University School of Medicine, 3600 Handa-cho, Hamamatsu 431-31, Japan.
37 nerve on one side was exposed for stimulation and recording of the compound action potentials, These experimental procedures were carried out under the ethical guidelines made by the Physiological Society of Japan 30 and by the Committee for Research and Ethical Issues of IASP (The International Association for the Study of Pain)3-~. Glass micropipetles for recording and HRP injection were filled with 7% HRP (Toyobo Co., Osaka, Japan) in 0.05 M Tris-HCI buffer (pH 8.6) with 0,2 M KCI. The electrode resistance varied from 15 to 40 MD. The sural nerve was stimulated with stainless steel electrodes (rate !/s, duration 300 p.s, interval 5 ms, train 3, intensity about 800/zA) and the compound action potentials were recorded with the method of Dong et alfl. A stimulus intensity of 800/zA was used to produce noxious stimulation which was confirmed by monitoring A~ and C fiber compound action potentials. Mechanical stimuli such as tap or pinch were also applied to the periphery to identify the characteristics of the neurons. The responses of the thalamic neurons were stored on magnetic tapes and analyzed later with a signal processor (7T08 NEC San-el Co,, Tokyo, Japan). HRP was injected iontophoretically through the recording electrode. A depolarizing constant current with a pulse duration of 200 ms and an intensity of 10-20 nA was passed through the electrode at 2.5 pulses/s for 7-15 min. The identity of the recorded spikes was checked every 2 rain during HRP injection. The cats were sacrificed 3-10 h after the HRP injection, The animal was perfused transcardially with 0.I M phosphate buffer solution and then with 2'% paraformaldehyde in 0.I M phosphate buffer (pH 7,4). The brain was removed and stored for 48 h in a 0.I M phosphate buffer solution containing 30% sucrose at 4°C. Serial transverse sections (75./zm thickness) were treated for HRP with the
Ni-Co enhanced diaminobenzidine method (Adams' method)I. The sections were mounted on gelatin-coated slides, air-dried for over 5 days, infiltrated with acetone, and microscopically examined. Drawings were prepared at a magnification of × 400 or x 1000, using a microscope. The sections were then counterstained with Cresyl violet and the neurons were located.
RESULTS
A total of 125 units were isolated and were classified into 4 groups. Twenty-one out of 125 units were 'tap units' which were driven only by innocuous brisk taps. Eighteen units were 'nociceptive units' which were driven only by noxious stimuli. Twenty-seven units were 'wide dynamic range units' (or 'noxious-tap units') which were driven by both noxious and tap stimuli. The remaining 59 units were termed 'spontaneous units', since they showed only spontaneous discharges and no time-locked responses to any stimuli. Eleven neurons out of 125 units were succe~s~tully stained with HRP. Nine out of the ll neurons were considered to be 'wide dynamic neurons', and the rest were thought to be innocuous 'tap neurons'. We could
Electric stimulation A z16----~,. . . . . . . . . . . . . .
100
m80¢,
C
;,
i.i,
Pinch
32~
Iltlm,
•
•
I
•
m
........ ,
\
\, m
ioo )Jm
\
1 8ec
Tap 32w sum.
•
II
•
II
K807265f
|
IJ
~
/
Fig. 1. A wide dynamic range neuron found in the parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram shows initial discharges and long after discharges with electric stimulation ( zx). The three tall vertical lines are shock artifacts of electric stimuli. This neuron also responded to both noxious (pinching) and innocuous (tapping) stimuli. The small bars indicate the period of each stimulation. B: the location of the neuron ( • ) . See Abbreviations section. C: camera lucida drawing of the neuron stained with HRP in the coronal plane. The neuron has fewer dendrites and scanter tufts than the tap neurons. These dendrites show the bipolar alinement. Its axon (arrow) runs in the antero-ventro-lateral direction.
38
A
Electric stimulation ~t
bl
t
LP
Q.
0
~,m.
hi I
o O-
100 msec
I -
Y'I 1
Fr, 6.5
Pinch 32'
0 1 sec
Tap
321
stl~,
I
I
K903165b
•
O
ioo~m
1 II0¢
Fig. 2. A wide dynamic range neuron found on the border of the parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram shows initial discharges, long after.discharl~es and inhibition with electric stimulation ( .. ). The three tall vertical lines are shock artifacts of electric stimuli. This neuron also responded to both noxious (pinching) and innocuous (tapping) stimuli, although the responses to the noxious stimuli ,re I,rger than those to the innocuous ones. The small bars indicate the period of ca~:lz .~timulation. B: the Iocutkm of the neuron ( # ). See Abbreviations section. C': camera lucida drawing of the neuron stained with HRP in the coronal plane. The neuron has fewer dendrites lind scanter tufts than the tap neurons. These dendrites show the phme alinement. Its axon (~lrrow) runs in the dorso-l,teral direction.
TABLE !
Summary of the morphologk'al differences between wide dynamic range and tap ne~rmes The density of dendritic spines showed no statistical difference between the two groups,
Wide dynamic Number of neurons Locations of neurons Ram6n-Moliner's classification Number of dendrites Tuft formation Number of dendritic spines(/10/~m) Density of dendritic spines(/l0 ~m) Axonal direction
Pf 5 ( I ) * Spf 2 ( I ) * * isodendritic type fewer scanter 180-250 417±0,6 antero-lateral antero-dorso-lateral dorso-lateral postero-ventro-medial ventro-medial
* A wide dynamic range neuron found on the lateral border of the Pf, A wide dynamic range neuron found in the ventral area adjacent to the Spf,
* *
Tap nellron ranRe neuron
! 2 I I I
2 MD i Spf ! allodendritic type richer denser 900-1100 5.3:1:0.7 antero-lateral 2
39
A
Electric stimulation _z 1 6 - - ¥ m
/,
O.
0
o 0l l l , i
L
&,,_k_~.
'al
I
',
i
100 m s e c
Pinch 32
/~i,~e!
'... j,d
Fr. 7.0
G
o J UklLii. h.-.a
1 SeC
Tap 321 [tim. u
_ _ 100 pm
K902147c
I tee
Fig. 3. A wide dynamic range neuron found in the parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram show initial discharges and after-discharges with electric stimulation ( A ). This neuron showed poor after-discharges, although it also responded to both noxious (pinching) and innocuous (tapping) stimuli, which proved this neuron to be a wide dynamic range neuron. The small bars indicate the period of each stimulation B: the location of the neuron ( • ). See Abbreviations section. C: camera lucida drawing of the neuron stained with HRP in the coronal plane. The neuron has fewer dendrites and scanter tufts than the tap neurons. These dendrites show the bipolar alinement. Its axon (arrow) runs in the postero-ventro-medial direction.
not inject HRP successfully into purely nociceptive neurons in this experiment. Figs. 1-5 show drawings of five wide dynamic neurons, their unit activities, and their locations in the thalamus. Figs. 6-7 show drawings of two tap neurons, their unit activities, and their locations. Among the nine wide dynamic neurons, six were found in and around the parafascicular nucleus and three were in and around the subparafascicular nucleus. Two tap neurons were found in the mediodorsal and the subparafascicular nuclei. The unit activities of wide dynamic neurons showed initial discharges of 30-50 ms after the electric stimulation followed by long after-discharges lasting for 200-500 ms. These neurons also responded to mechanical stimuli including both pinching and tapping (Figs. IA-5A). The unit activities of tap neurons, on the other hand, showed only initial discharges ~,~ter the electric stimulation and no response to pinching (Figs. 6A and 7A). The HRP study revealed morphological differences between wide dynamic and tap neurons (Figs. 1C-7C, Table I). Wide dynamic neurons had fewer dendrites which formed scanter tufts, whereas tap neurons had many more
dendrites which formed denser tufts. The number of dendritic spines were about 180-250 on one dendrite in wide dynamic neurons, and about 900-1,100 in tap neurons. However, there was no statistically significant difference in the density of dendritic spines between wide dynamic (4.7 + 0.6/10 ~m) and tap neurons (5.3 + 0.7/10 pm). The dendritic spines were omitted in the drawings. The axons (arrow in the figures) were identified morphologically by the absence of dendritic spines. The axons of the two tap neurons were seen to run antero.laterally, whereas those of wide dynamic neurons run in antero-lateral, antero-dorso-lateral, dorso-lateral, postero-ventro-medial and ventro-medial directions, respectively.
DISCUSSION Histological studies of these medial thalamic, nonspecific nuclei have used degeneration techniques 13, intra-tissue injections of tritiated amino acids 12'e'~or of HRP 7.1°,!6.e1.29,32, and other methods e'2°'es. Anatomical
40
A
Electric stimulation ,
z16- ? L_~|~ IL~
"0N,,L
o
,-
'
L,L
Puk
.
LP
\
'
p
i ~ .~ ,,,., =,.bh~h kBi=~ 100 msec
Fr. 7.5
.. / s
Pinch
_f
32 I Ibm i
I
sec
Tap 32~-
W
tOO~ m
OA._,I,,__L_L_I, .I.L_JL, LI, K903164d
t lee
\
Fig. 4. A wide dynamic range neuron found in 1he parafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli, Pulse density histogram shows initial discharges and long after.discharges with electric stimulation ( ,x ), The three tall vertical lines are shock artifilcts of electric stimuli, This neuron also responded to both noxious (pi.ching) and innocuous (tapping) stimuli, .ltht~ugh the re,,,ponses to the irlnocuotis stimuli are sm;dler than those to noxiotis ones, The small h.rs indic.re the p~riod of each stimulatio.. B: the location of the neuron ( $ ), Scc Abbrcviati;)lls section, (': c.mcra lucid, drawing of the: netffon stained with IIRP in the coronal plane, The .euron has fewer dendrites =rod s~:.nter tufts tha. the t.p neurons, These dcndrltes show the htpohtr ttllncntettt, Its axon (arrow) ru.s in the vcntro-medial direction.
connections to and from tile medial thalamus have been described in detail. For example, the parafascicular nucleus receives inputs from bulbothalamic and spinothalamic pathways, from the substantia nigra pars rcticulata, parabrachial pontinc nucleus, periaqueductal gray, and deep layers of the superior colliculus, and provides efferent projections to the cerebral cortex, striatum and subthalamic nucleus. Based on these findings and on the electrophysiological characteristics, it has been speculated that nociceptire neurons in the medial thalamus are related to either the affective aversive aspect of the nociceptive stimuli aas''-a, the scnsorimotor responses to noxious stimuli:", or the arousal system :,~s. in this experiment, we tbund many morphological differences between wide dynamic range neurons and tap neurons in the medial thalamus. The main differences are the spatial distribution of their dendrites and the axonal directions. The difference in the number of
dendritic spines seems to depend on the number of branches of the dendrites, because the density of dendritic spines showed no statistical difference between these two neurons. Ram6n-Molincr classified vertebrate neurons into three categories according to the morphology of their dendrites ''~'2~'. The neurons include isodendritic, aliodendritic, and idiodendritic types. The isodendritic type is characterized by few relatively straight dendrites and scanter tufts, which tend to orient themselves along a given plane or to diverge to two opposite directions. The ailodendritic type is defined as more specialized neurons than the isodendritie type, but not so highly specialized as the idiodendritic type. Neurons of this latter type are richer in dendrites and have denser tufts than the isodendritic ones, and are also are divided into several subgroups depending on their locations. The idiodendritic type is defined as the most specialized neurons. Each neuron of this type depends so
41
A
Electric stimulation '~
z 16 nW
8 o,1-" _.IJI J L ~ , ~ . ~ . ~ , ~ 100 msec
Fr. 8.0
\
'
Pinch 32~ slim. o
.... J M , . j , , k l
,.
l',..iJb
....
O 1 SOC
Tap
32~
s~m. =
•
I
•
I
I
o '~ ,u.ul~, II, I,, lu.,u.A K812226d i1 8 e c
i
100 um
Fig, 5. A wide dynamic range neuron fcand in the area adjacent to the subparafascicular nucleus. A: responses of the wide dynamic range neuron to electric and mechanical stimuli. Pulse density histogram shows initial discharges and long after-discharges with electric stimulation ( A ). The three tall vertical lines are shock artifllcts of electric stimuli. Although this neuron shows poor after-discharges, it also responded to both noxious (pinching) and innocuous (tapping) stimuli, which proved it to be a wide dynamic range neuron. The small bars indicate the period of each stimulation. B: the location of the neuron ( • ) . See Abbreviatiozls section. C: camera lueida drawing of the zleuron stained with HRP in the coronal phme. The neuron has l'~wer dendrites ,nd scanter tufts thai| th~ tap neurons. Its axon (arrow) runs in tile antero-dorso.lateral direction.
specifically on its location that one can easily identify the location even by the shape of any single neuron of this type. In our experiment, all of nine wide dynamic range neurons in the medial thalamus were of the isodendritic type: Four showed bipolar orientation of the dendrites and two others tended to orient themselves along a plane. The two tap neurons were of the allodendritic type. Their axons tended to run anterolaterally, suggesting that they project their axons to the same target and play the same role, though further studies are necessary to draw a definite conclusion. The axons of wide dynamic range neurons, on the other hand, tended not to run in a fixed single direction but rather in many directions. This seems to
indicate that wide dynamic range neurons have connections not only with the ascending reticular system but with many other systems, including the hypothalamus or the limbic system, it is very interesting that, as Ram6n-Moliner pointed out, the neurons of the hypothalamus or of the reticular formation belong to the isodendritic type and the neurons of specific thalamic relay nuclei, such as ventral posterior nuclei, to the allodendritic type 26. These findings may suggest that tap neurons are more specialized neurons and have more distinct and fixed functions than wide dynamic range neurons in the somatosensory system, and that tap and wide dynamic range neurons play quite different roles in the medial thalamus.
42 Electric stimulation A| ~e---~
-~-! ..... 1~
~ -~--~.
,.p "t, " " / } ,
..o,~
i
I+ ! CL i
~
k ~.~'i
i.A .,~., o
0''
'
~-T
|
I
i 100 mse¢
Zl
_
x°"'~
'~
Hyde//
--'--
Fr. 8;0
Pinch
321
a
0i
I
h
|o 1 II0¢
l
Tap
/
32r ,%~..
.
.
.
.
.
.
.
i
K807054b
,•.=
Fig, 6. A tap neuron found in the subparafascicular nucleus. A: responses of the tap neuron to electric and mechanical stimuli. Pulse density histogram shows only initial discharges with electric stimulation (~). This neuron responded only to innocuous (tapping) stimuli, and did not respond to noxkius (pinching) stimuli. These findings proved this neuron to be a tap neuron, The small bars indicate the period of each stimulation, B: the location of the neuron (o), See Abbreviations section. C: c:lmera lucida drawing of the neuron stained with HRP in the coronal plane, The neuron has richer dendrites and denser tufts than the wide dynamic range neurons. Its axon (arrow) runs in the antero-latend aspect,
A
Electric stimulation
BI
).......... ,.; ...... ~>i ~
~,
-r"°io
