THE JOURNAL OF COMPARATIVE NEUROLOGY 321:404420 (1.992)

Diencephalic Projections From the Superficial and Deep Laminae of the Medullary Dorsal Horn in the Rat KOICHI IWATA, DAN R. KENSHALO, JR., RONALD DUBNER, AND RICHARD L. NAHIN Neurobiology and Anesthesiology Branch, National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20892

ABSTRACT An important function of the medullary dorsal horn (MDH) is the relay of nociceptive information from the face and mouth to higher centers of the central nervous system. We studied the central projection pattern of axons arising from the MDH by examining the axonal transport of Phaseolus uulgaris-leucoagglutinin (PHA-L). Labeled axon and axon terminal distributions arising from the MDH were analyzed at the light microscopic level. After large injections of PHA-L into both superficial and deep laminae of the MDH in the rat, labeled axons were observed in the nucleus submedius of the thalamus (SUB), ventroposterior thalamic nucleus medialis (VPM),ventroposterior thalamic nucleus parvicellularis (VPPC), posterior thalamic nuclei (PO), zona incerta (ZI), lateral hypothalamic nucleus (LH), and posterior hypothalamic nucleus (PHI. Restriction of PHA-L into only the superficial laminae resulted in heavy axon and varicosity labeling in the SUB, VPM, PO, and VPPC and light labeling in LH. In contrast, after injections into deep laminae, labeled axons were mainly distributed in ZI and PH; some were also in VPM and LH, and fewer still in PO and SUB. Varicosities in VPM, SUB, and PO were significantly larger than those in VPPC, ZI, LH, and PH. Varicosity density was highest in SUB and lowest in the VPPC. We concluded that there are two distinct nociceptive pathways, one originating from the superficial MDH and terminating primarily in the dorsal diencephalon and the second originating from deep laminae of the MDH and terminating primarily in the ventral diencephalon. We propose that in the rat, input from the deeper laminae is primarily involved in the motivational-affective component of pain, whereas input from the superficial MDH is related to both the sensory-discriminative and motivational-affective component of pain. Published by Wiley-Liss, Inc., 1992 Key words: trigeminal, hypothalamus,thalamus, PHA-L, pain

The spinal trigeminal nucleus caudalis, also referred to as the medullary dorsal horn (MDH), is the medullary homologue of the spinal cord dorsal horn (Gobel et al., '77, '80). As such, the MDH shares many of the same functions as the spinal cord dorsal horn, including the processing of nociceptive input from the periphery. Thin primary afferent fibers thought to convey nociceptive input terminate in the superficial and deep laminae of the feline MDH (Gobel et al., '81; Hayashi, '85). These areas of the MDH contain many neurons responsive to noxious stimulation of the orofacial skin in cats (Hu et al., '81; Hu and Sessle, '84; Amano et al., '86; Sessle et al., '861, primates (Price et al., '76; Hoffman et al., '81; Maixner et al., '89) and rats (Renehan et al., '86).

PUBLISHED BY WILEY-LISS, INC., 1992

Anatomical studies in a number of species including the rat (Lund and Webster, '67; Fukushima and Kerr, '79; Peschanski, '84; Peschanski and Ralston, '85; Peschanski et al., '85; Kemplay and Webster, '891, cat (Burton et al., '79; Shigenaga et al., '83; Peschanski and Ralston, '851, monkey (Burton and Craig, '79; Rausell and Jones, '91) and hedgehog (Ring and Ganchrow, '831,as well as electrophysiological studies in the cat (Hu et al., '81) and monkey (Price et al., '76; Bushnell et al., '84) have suggested that the posterior thalamic nuclei receive nociceptive input from the trigeminal nuclei. Recent anatomical studies using retrograde tracing techniques have reported that the nucleus Accepted March 7, 1992

MDH PROJECTIONS TO DIENCEPHALON

405

submedius (SUB) in the rat (Craig and Burton, '81; Dado and Giesler, '90) and cat (Craig and Burton, '81; Craig, '87) and the lateral hypothalamic nucleus (LH) in the rat (Burstein et al., '87, '90) also receive direct projections from the MDH. However, with the exception of Peschanski ('841, the topography of MDH projections to the entire diencephd o n has not been systematically studied in the rat. Peschanski ('84) used large injections of wheat germ agglutininconjugated to horseradish peroxidase (WGA-HRP) to reexamine trigeminal afferent pathways to the diencephalon, but did not distinguish between pathways originating from the superficial or deep MDH. The present series of experiments used Phaseolus vulgaris-leucoagglutinin (PHA-L) as an anterograde tracer to study the diencephalic termination pattern of afferents originating from either the superficial or deep laminae of the MDH. PHA-L was chosen as the anterograde tracer, since recent reports indicate that PHA-L has several advantages over WGA-HRP (Gerfen and Sawchenko, '84; Cliffer and Giesler, '88): 1) PHA-L is mainly taken up by cell bodies and not significantly by passing fibers (however, see Rhoades et al., '89; Schofield, '90); 2) PHA-L injection sites are much more focally delineated than with HRP or other tracers; 3) axons and terminal varicosities anterogradely labeled with PHA-L are clearly identified. The present paper describes the dorsoventral and rostrocaudal distribution pattern of axons within diencephalic nuclei and reveals that, in general, cells in the superficial and deep laminae of MDH send their axons to different groups of nuclei. A preliminary report has been published (Iwata et al., '90).

MATERIALS AND METHODS Adult, male Sprague-Dawley rats (300-400 g) were anesthetized with sodium pentobarbital (40 mglkglml, i.p.1 and placed in a stereotaxic apparatus. The cisterna magna was opened to expose the MDH. During surgery, the animals were kept warm with a heating pad and the depth of anesthesia was maintained by continuous inhalation of methoxyflurane. Microinjections were made through a glass

micropipette (tip diameter of 15-25 pm) attached to a 1 p1 Hamilton syringe filled with 2.5% PHA-L (Vector Labs) in 10 mM Na-phosphate buffered (pH 8.0) saline. Penetrations were made between 0.2 and 0.8 mm caudal to the obex and 2.0 and 2.8 mm lateral to the midline. PHA-L (0.01 p1) was pressure injected (Casale et al., '88; Rhoades et al., '89) at two or three points along each penetration. Injection depths varied between 100 and 1,500 pm ventral to the obex. Penetrations and microinjections were separated by 200 pm. After each microinjection of PHA-L, the micropipette was kept in position for 15 minutes to prevent spread of PHA-L from the injection site. Two different sizes of injections were produced in the present study: 1) large injections extending from lamina I to lamina VI in which 10-15 PHA-L microinjections were made; 2) small injections restricted to either the superficial (1-111) or deep (IV-VI) laminae in which 6 to 9 microinjections were made. Penicillin G benzathine was injected (100,000 units/ml; 0.1 ml/kg, i.m.) after the surgery to prevent infection of traumatized tissues. After 14 to 21 days, animals were deeply anesthetized with sodium pentobarbital (80 mgikg, i.p.1 and perfused transcardially with 50 ml of phosphate-buffered saline (PBS; pH 7.4) followed by 4% paraformaldehyde in 0.1 M phosphate buffer. The brain was removed and placed in cold fixative for 24 hours and then transferred to cold phosphatebuffered 30% sucrose for 24-48 hours. Serial horizontal sections (50 pm thick) through the brain were cut from two animals with large injections. In addition, serial transverse sections (30 pm) were cut from animals with large (n = 3), superficial (n = 21, and deep (n = 2) injections. Sections were rinsed several times with PBS before immunocytochemical processing. Sections were processed for PHA-L immunocytochemistry as follows. They were first incubated for 30 minutes in PBS with 1.5% normal rabbit serum (NRS). Sections were then incubated for 48 hours in goat anti-PHA-L (1:2,000; Vector Labs), followed by successive 1 hour incubations in biotinylated rabbit-anti-goat (1:200; Vector Labs) and peroxidase-conjugated avidin-biotin complex (1:100; ABC: Vector Labs). We used a 6-10 minute incubation in 0.035%

Abbreviations

(from Paxinos and Watson, '82) 4

12

3v APT AV CL CM cu DM f F fr Gr ic LD LH LP LV MD Me MG ml

trochlear nucleus hypoglossal nucleus third ventricle anterior pretectal area anteroventral thalamic nucleus centrolateral thalamic nucleus centromedial thalamic nucleus cuneate nucleus dorsomedial hypothalamic nucleus fornix nucleus fields of Fore1 fasciculus retroflexus gracile nucleus internal capsule laterodorsal thalamic nucleus lateral hypothalamic nucleus lateroposterior thalamic nucleus lateral ventricle mediodorsal thalamic nucleus mesencephalon medial geniculate nucleus medial lemniscus

mlf mt opt PH PO PY Sol SP5 Sp5C Sp51 Sp50 SNR STh SUB VL

VM VMH VPL VPM VPPC ZI

medial longitudinal fasciculus mammillothalamic tract optic tract posterior hypothalamic nucleus posterior thalamic nuclei pyramidal tract nucleus of the solitary tract spinal trigeminal tract spinal trigeminal nucleus, caudal spinal trigeminal nucleus, interpolar spinal trigeminal nucleus, oral substantia nigra reticularis subthalamic nucleus nucleus submedius of the thalamus ventrolateral thalamic nucleus ventromedial thalamic nucleus ventromedial hypothalamic nucleus ventroposterior thalamic nucleus lateralis ventroposterior thalamic nucleus medialis ventroposterior thalamic nucleus parvicellularis zona incerta

K. IWATA ET AL.

406

3,3'-diaminobenzidine-tetra HCl (DAB; Sigma), 0.08% nickel ammonium sulfate, and 0.001% hydrogen peroxide in Trisbuffered saline (0.05M, pH 7.4) to develop the ABC reaction product, thus producing a distinctive black deposit. Between each incubation, tissue was washed for 15 minutes with PBS. Every other section was counterstained with Neutral Red. Labeled axons and varicosities were drawn with the aid of a camera lucida drawing tube. Axons were considered to represent projection pathways. While varicosities were taken to indicate functional synapses, such synapses can only be confirmed with electron microscopy. Each diencephalic nucleus was cytoarchitectonically identified according to the stereotaxic brain atlas of Paxinos and Watson ('82) and the descriptions as set forth by Jones ('85).The density and size of PHA-L-labeled varicosities were determined by placing a 100 x 100 pm grid over the diencephalic nuclei. The numbers of varicosities within this grid in ten transverse sections from three different rats were drawn, counted, and measured. Statistical differences in varicosity size or varicosity density at different locations were examined by an analysis of variance or chi-square. For chi-square analysis, the total numbers of varicosities from the three animals were pooled. Labeling in the ipsilateral diencephalon was not quantified because of the relatively small number of ipsilateral trigeminodiencephalic projections seen in the present study. We did not see any retrogradely labeled cell bodies in the diencephalon or brainstem after injection of PHA-L in the MDH; therefore, it is not likely that we were observing axon collaterals of retrogradely labeled neurons.

RESULTS Injection of PHA-L into the MDH produced anterograde labeling in a number of diencephalic nuclei. Figure 1 illustrates the rostrocaudal extent of a large injection including both the superficial and deep MDH. PHA-L is thought to be taken up preferentially by the cells within the injection core (filled area) of the illustrated injection site (Fig. 11, but not by cells in the injection halo (hatched area) (Gerfen and Sawchenko, '84). Within the core, individual neurons could not be identified. However, numerous heavily labeled cells could be seen within the injection halo. In addition, many fibers emanated from both the halo and core into both the dorsolateral and ventrolateral white matter. We assume that these are not fibers of passage, but rather axons of MDH neurons that we labeled by our injections. This assumption is supported by the observation that our PHA-L injections never labeled axons or varicosities within the VPL; this would be expected if PHA-L were picked up by fibers ascending from the spinal cord. Large PHA-L injections were generally confined to the MDH, although a small amount often spread medially to include the nucleus caudalis ventralis. Figure 2 illustrates several other examples of PHA-L injection sites. Figure 2A is a reconstruction in the near horizontal plane illustrating the full rostrocaudal extent of a large PHA-L injection. The dorsoventral extent of two large injections is seen in Figure 2B and C. Figure 2D and E gives photomicrographs of small injections restricted to either the superficial (Fig. 2D) or deep laminae (Fig. 2E). Figure 3 illustrates two lateral views of the diencephalon showing the approximate levels of the drawings in Figures 4-7.

-13.8

-14.6

Fig. 1. Camera lucida drawings illustrating the rostrocaudal spread of a large PHA-L injection that included both the superficial and deep MDH. The solid areas of the drawings illustrate the primary region of tracer uptake. Numbers indicate the anterior-posterior coordinates of the drawings based on the sterotaxic atlas of Paxinos and Watson ('82).

Projections after PHA-L injections including both the superficial and deep MDH After a large injection of PHA-L encompassing both the superficial and deep MDH (Fig. 11, a dense projection of labeled axons was found in SUB, VPM, VPPC, PO, LH, ZI, and, PH (Figs. 4 , 5 ) . Few labeled axons were observed in CM and CL (Fig. 4c). The vast majority of this labeling was contralateral to the PHA-L injections, but isolated axons

MDH PROJECTIONS TO DIENCEPHALON

mIf

12

SP

0

Fig. 2. Drawings and photomicrographs of PHA-L injection sites in the MDH. A, B, C: Examples of large injections in the MDH cut in either the horizontal (A) or transverse (B and C) plane. D: Photomicrograph illustrating a superficial injection. E: Photomicrograph of a deep injection. Total magnification in photomicrographs (D and E) x25.

407

408

K. IWATA ET AL.

A

B

Fig. 3. Lateral view of the diencephalon illustrating the approximate level of the drawings in Figures 4-7. A: Rostrocaudal levels of the coronal sections a-e in Figure 4 and a 4 in Figures 6 and 7. B: Dorsoventral levels of nearly horizontal sections a+ in Figure 5.

were occasionally observed in ipsilateral SUB, PH, and ZI (not shown). In SUB, axons were located in the dorsal portion and distributed along the full mediolateral extent of this nucleus. It appears that some axons in VPM or VPPC ascend to the middle part of SUB (Fig. 5b), and then course mediolaterally in the dorsal part of SUB (Figs. 4a, 5b). Many axons were labeled in the middle to caudal portion of VPM (Figs. 4b-d, 5a). The axons labeled in VPM were found to be larger as compared to those in the other thalamic nuclei (Fig. 8). In the middle portion of VPM, many axow were located in the dorsolaterd quadrant (Fig. 4a-c), whereas in the caudal portion of VPM, axons were distributed throughout the nucleus (Fig. 4d). The axons reaching the VPM appear to course through the lateral portion of PO and the medial portion of ZI (Fig. 5a,b). Labeled axons were also identified in the ventral portion of the VPPC. The axons in the VPPC were distributed within mediolaterally extending axon bundles (Figs. 4c, 5b). The axons reaching this nucleus appeared to ascend through the ventrocaudal part of VPM (Fig. 5b). Many labeled axons could be seen in PO. Axons were densely distributed just medial to the MG (Figs. 4e, 5a). Fibers reaching the PO appear to course through the lateral part of the mesencephalon (Fig. 5a) and magnocellular part of the medial geniculate complex (Figs. 4e, 5a). Some of the labeled fibers in PO appear to be distributed ultimately to the anterior pretectal area (Fig. 4e). In Figure 4c, many labeled axons can be seen in the ventral portion of ZI. Axons within ZI were distributed as a mediolaterally extending bundle of axons in both coronal and horizontal sections. The axons reaching this nucleus appear to course through the lateral part of the mesencephalon (Fig. 5b) and the medial region of the SNR (not shown). Within the hypothalamus two groups of axons were identified, those terminating in the lateral hypothalamus (LH) and those terminating in the posterior hypothalamus

(PH). Labeled axons were distributed in the dorsal portion of both the PH and the LH (Fig. 4b,d). Some axons in the LH appeared to course rostrally and dorsally to terminate either medially in the PH or laterally in the ZI (Fig. 5d).

Labeling after PHA-L injections restricted to either superficial or deep laminae Discrete injections of PHA-L into either the superficial or deep MDH consistently produce more restricted patterns of anterograde labeling in the diencephalon. After PHA-L injections restricted to the superficial laminae of MDH, we observed a number of labeled axons in the dorsal portion of SUB (Fig. 6a). The relatively small number of axons labeled probably reflects both the small amount of PHA-L injected and the fact that the PHA-L injections spared the most ventral and most dorsal parts of the MDH (Fig. 1D). However, the distribution pattern of labeled axons in SUB was similar to that seen after large injections. Superficial injections also produced dense labeling in the dorsolateral part of VPM (Fig. 6b,c), in VPPC (Fig. 6b) and PO (Fig. 6d) and a small amount in LH (Fig. 612). The distribution pattern of axons in these nuclei was similar to that after large injections encompassing both superficial and deep laminae, When PHA-L was injected into the deep laminae of MDH, many axons were labeled in PH and ZI, and a small number of axons in LH (Fig. 7b-d) and the lateral portion of VPM (Fig. 7a-c). An occasional isolated axon was observed within SUB and PO (not shown), but no labeling was seen in VPPC (Fig. 7). In summary, the diencephalic nuclei receiving inputs from superficial or deep laminae were clearly distinguishable in this experiment. After PHA-L injections in the superficial laminae, the vast majority of labeled axons were restricted to nuclei located in the dorsal portion of the diencephalon (SUB, VPM, PO, and VPPC); on the other hand, deep injections produced labeling primarily within

a

C

d b

e Fig. 4. a-e: Camera lucida drawings of transverse sections through the diencephalon showing the distribution of labeled axons and varicosities following the large PHA-L injection illustrated in Figure 1. The most rostra1 section is indicated by a in this figure. Each number indicates the distance (mm) from bregma.

L

J

a

C

\

Fig. 5. a-e: Camera lucida drawings of horizontal sections through the diencephalon showing the distribution of labeled axons and varicosities. The labeling is based on the large PHA-L injection illustrated in Figure 2A. Rostral (R) and lateral (L) are indicated in a. The most dorsal section is indicated by a in this figure.

MDH PROJECTIONS TO DIENCEPHALON

2.9

411

K. IWATA ET AL.

412

3.8

b Fig. 7. a-d Camera lucida drawings of transverse sections through the diencephalon showing the distribution of labeled axons and varicosities. The labeling is based on the small deep PHA-L injection illustrated in Figure 1E. The most rostra1 section is indicated by a in this figure. Each number indicates the distance (mm) from bregma.

ventrally situated nuclei (ZI, LH, and PHI. Injections of PHA-L that included both superficial and deep laminae produced dense labeling in all areas seen with restricted injections (SUB, W M , PO, VPPC, ZI, LH, and PHI, plus sparse labeling in CM and CL.

Morphological characteristics and density of trigeminal-diencephalicvaricosities Although many different sizes of axonal varicosities were noted within each nucleus receiving MDH input, clear

MDH PROJECTIONS TO DIENCEPHALON

A

413

w

.4

6‘

,

4

5/

Fig. 8. Camera lucida drawings of labeled varicosities and axons in the contralateral diencephalon. and PH (F).Scale bar = Labeled axons and varicosities in VPM (A), SUB (B),VPPC (C), 21 (D), LH (E), 10 pm.

differences were seen in the mean varicosity size between nuclei (Fig. 8-10). For instance, within VPM varicosities were larger on average than those in other nuclei, being located at the terminal end of relatively large-diameter axons (Figs. 8A, 9A). These exceptionally large varicosities were restricted to the middle portion of W M ; nonvaricose axons were not seen in this area. Intermediate-sizedvaricosities were found in SUB (Figs. 8B, 9B), PO (Fig. 9 0 , and VPPC (Fig. 8C). Varicosities at the terminal ends of axons, thin fibers with en passant structures, and isolated varicosities were all intermingled within SUB (Figs. 8B, 9B), VPPC (Fig. 8C), and PO (Fig. 9C). The varicosities within ZI (Fig. 8D), LH (Figs. 8E, 9E), and PH (Figs. 8F, 9F) showed similar sizes and arrangements. Many of them had very clear en passant structures. Labeled axons and isolated varicosities were intermingled as in SUB, PO, and VPPC. A histogram illustrating the mean area per labeled varicosity in various diencephalic nuclei is presented in Figure 10. A total of 3,295 varicosities was analyzed in this study. The mean varicosity size in VPM was 3.2 0.2 pm2 (mean f SEM, n = 361). These varicosities were significantly larger than those in other nuclei (SUB, PO, VPPC, PH, LH, ZI, and LH) (P < 0.01). The mean size of varicosities in SUB was 2.2 ? 0.1 pm2 (n = 760). These were

*

significantly larger ( P < 0.01) than those in PO, VPPC, PH, ZI, and LH. The mean size of varicosities in PO was 1.1 2 0.1 pm2 (n = 373). Varicosities in PO were significantly larger than those in VPPC, PH, ZI, and LH (P < 0.01). Varicosities in VPPC, PH, ZI, and LH were similar in sizes and arrangement within the nucleus [VPPC, 0.7 f 0.03 pm2 (n = 260); PH, 0.6 f 0.02 pm2 (n = 612); ZI, 0.5 & 0.01 pm2 (n = 626); LH, 0.5 f 0.02 pm2 (n = 303)l. After large injections of PHA-L, we observed a significantly higher terminal density within SUB than all other nuclei (chi-square, P < 0.01; Fig. 11).PH and ZI, in turn, contained higher terminal densities than VPM, PO VPPC, or LH (chi-square, P < 0.01: Fig. 11).However, no overall difference was seen between terminal density in the thalamus and hypothalamus, each region containing subdivisions of high and low density. No correlation was seen between terminal density and terminal size (df = 7, r2 = 0.01).

DISCUSSION In the present report, we have used the anterograde tracer PHA-L to delineate the terminal arborization and density of axons originating from the superficial and deep portions of the MDH. In particular, we present the first definitive evidence in the rat that the predominate dienceph-

Fig. 9. Photomicrographs of labeled varicosities and axons in the contralateral diencephalon. Labeled axons and varicosities are in VPM (A),SUB (B),ZI (C), LH (D),and PH (El. Scale bar in E = 60 +m in A and B and 40 W r n in C-E.

connections arise primarily from the superficial MDH. The findings for VPM are in close agreement with Shigenaga et al. ('83), who found that tracer injections into the cat VPM produced the greatest degree of retrograde labeling in the superficial laminae of the MDH. In the present study, we did not see a large projection from the deep MDH to the PO as suggested by the retrograde labeling study of Dado and Giesler ('90; Fig. 6). These discrepancies may be a result of methological differences. For instance, the retrograde tracer injections of Dado and Giesler ('90) may have extended into ZI or been picked up by fibers coursing toward ZI through the MLF; either of these occurrences would result in a number of deep MDH neurons being labeled. Alternatively our series of small, restricted injections of PHA-L could have a bias toward revealing only the most prominent projection pathways originating from the MDH, such that we would underesti-

4N i

E

t

T

v

3

3-

v) +I

m

2m 2 .-E In

.-8 k>

'-

5

m

r" Fig. 10. Mean area per labeled varicosity after PHA-Linjections in the MDH.The varicosity areas in VPM,SUB, and PO were significantly different (ANOVA; P < 0.01) from each other. The varicosity sizes in each of these nuclei were also different from those in W P C ,PH,ZI, and LH (ANOVA, P < 0.01).

800 (u

E

s'" F

x 0

r m In

.-.In 0 .g

Q) . I -

z

200

n "

SUB

ZI

PH

PO VPM LH VPPC

Fig. 11. Number of varicosities per nuclei after PHA-Linjections in the MDH. The varicosities from three different animals (area sampled equals 10 x 100 pm2/rat) were pooled. The density of varicosities within SUB was significantly higher than all other nuclei (chi-square; P < 0.01). PH and ZI had higher varicosity densities than VPM,PO, LH, or VPPC (chi-square;P < 0.01).

alic projection of the deep MDH is to PH and ZI, while that of the superficial MDH is to SUB. Furthermore, we have demonstrated that different regions of the diencephalon receive different sized varicosities originating from overlapping locations in the MDH.

VPM, PO, and intralaminar projections It has been known for some time that the VPM and PO are densely innervated by neurons in the rat MDH (Lund and Webster, '67; Fukushima and Kerr, '79; Giesler et al., '79; Peschanski, '84; Granum, '86; LeDoux et al., '87; Kemplay and Webster, '89). We have confirmed these innervations and have further demonstrated that these

illustrated by Dado and Giesler ('90; Fig. 6). Interestingly, while the present report and Dado and Giesler ('90) agree that the superficial MDH has a more substantial projection to the PO than does the deep MDH in the rat, Shigenaga et al. ('83) have identified the opposite in the cat. This may be another example of interspecies variations. An important but not exclusive function of the spinothalamic and trigeminothalamic tracts is to relay nociceptive information to the diencephalon (Willis and Coggeshall, '78; Dubner and Bennett, '83). The VPM and PO have been thought to be the main recipients of this nociceptive information. Electrophysiological studies in different species have shown that many MDH neurons responding to noxious mechanical and thermal stimulation of the facial skin can be antidromically activated from thalamic nuclei in the monkey (Price et al., '761, cat (Sessle and Greenwood, '76; Hu et al., 'Bl), and rat (Renehan et al., '86). Many neurons within VPM respond to noxious stimulation of the periphery in both the rat (Shigenaga et al., '73) and cat (Woda et al., '75; Albe-Fessard et al., '77; Yokota et al., '85, '86; Kni& and Vahle-Hinz, '87). Some of these neurons project to feline SI cortex (Yokota et al., '86). The lateral thalamus, including VPM, is thought to mediate the discriminative aspect of pain (e.g., Mehler et al., '60). The significant contribution the superficial MDH makes to the VPM suggests that some neurons within the superficial MDH are concerned with the quality, intensity, and temporal aspects of pain in the rat. Previous work identified abundant spinal cord and trigeminal input to CL and CM in the rat (Fukushima and Kerr, '79; Giesler et al., '79; Granum, '86) and cat (Shigenaga et al., '83). However, early data in the cat clearly demonstrate that only neurons in laminae VII and VIII of the spinal cord and their trigeminal homologue, nucleus caudalis ventralis (NCV; Gobel et al., '77; Shigenaga et al., '831, project to CL and CM. In the present study, our PHA-L injections barely impinge on the NCV, thereby accounting for the relative lack of intralaminar labeling. Unlike Peschanski ('84), we never identified fierents in or around the fasciculus retroflexus (FR) after tracer injections into the MDH. However, as is the case with CL and CM, it is likely our injections were too dorsally situated within the MDH to label neurons projecting to the FR. Shigenaga et al. ('83) found that retrograde tracer injections encompassing FR in the cat only labeled neurons within the NCV. Similarly, Giesler et al. ('79) found that

416

K. IWATA ET AL.

tracer injections into the medial thalamus that included the FR and surrounding gray matter labeled deep neurons but not dorsal horn neurons within the C1 spinal cord of the rat. Finally, spinal cord studies suggest that species variability may especially influence MDH projections to the FR region; anterograde tracer studies using WGA-HRP have failed to identify spinal cord projections to FR in the squirrel monkey (Mantyh, '83) and cat (Craig and Burton, '851, but reveal the projections in the raccoon (Craig and Burton, '85) and macaque monkey (Apkarian and Hodge, '89), complicating interpretation of the present results.

These characteristics make it unlikely that the MDH projection to SUB would transmit information related to the sensory-discriminative aspects of pain. SUB projections to the orbital cortex have also been described in the rat (Herkenham, '79; Price and Slotnick, '83; Ma et al., '88) and cat (Craig et al., '82). The orbital cortex is generally associated with aversive behavior, autonomic arousal, and emotional control (Warren and Akert, '64; Rosenkilde, '79; Tsubokawa et al., '811, suggesting that the trigeminalcortical pathway via SUB mediates the emotional and aversive components of the pain experience.

SUB projections

ZI projections

Recent anterograde studies have described axons arising from the MDH that project to the dorsal portion of SUB in the cat and rat (Craig and Burton, '81; Peschanski, '84; Craig, '87; Ma et al., '88). Retrograde tracer studies in the rat (Dado and Giesler, '90) and cat (Craig and Burton, '81) indicated that most neurons sending axons to the SUB are distributed in the superficial laminae of the MDH. We have confirmed these findings by demonstrating that substantial numbers of labeled axons could be observed in the dorsal portion of SUB only after injection of PHA-L into superficial laminae. This absence of a large projection from the deep MDH to SUB is consistent with Dado and Giesler ('go), who found that 4-5 times as many neurons are labeled in the superficial MDH, as the deep MDH, after Fluoro-Gold injections in to SUB. It has been stated that PHA-L injections restricted to lamina I of the cat MDH will exclusively label fibers in SUB (Craig, '87). Unfortunately, injection sites were not included in this report. Craig ('87) also did not examine the labeling produced by PHA-L injections into deep laminae, making it difficult to compare these results from the cat to the present work in the rat. There is some controversy as to the topography of inputs to SUB. Craig and Burton ('81) report that, in the cat, trigeminal inputs terminate in the caudal half of SUB, while spinal inputs end in the rostral half. However, this topography appears to break down in the rat. Despite injection sites that covered the rostral two-thirds of SUB, Dado and Giesler ('90) could not identify significant numbers of spinal cord neurons projecting to SUB. Similarly, after tracer injections into the MDH, our anterograde labeling, like that of Craig and Burton ('81), is heaviest in the rostral one-half to three-quarters of SUB. Therefore the evidence does not support a clear topographic distribution to SUB inputs in the rat. Most lamina I neurons in the MDH respond either exclusively or preferentially to noxious stimulation in the monkey (Willis et al., '74; Price et al., '76; Hoffman et al., '81) and rat (Salt et al., '83).In the rat, some of these nociceptive neurons can be antidromically activated from SUB (Dostrovsky et al., '87). It is therefore not surprising that many neurons within SUB respond either exclusively or preferentially to noxious peripheral stimulation as demonstrated in the cat (Craig, '87) and rat (Dostrovsky et al., '87; Dostrovsky and Guilbaud, '88, '90;Miletic and Coffield, '89). However, the response characteristics of nociceptive neurons in SUB are quite different from those in the superficial MDH. Nociceptive neurons in SUB usually have large bilateral receptive fields and often respond with a delayed onset and delayed offset to peripheral noxious stimulation (cat: Craig, '87; rat: Dostrovsky et al., '87; Dostrovsky and Guilbaud, '88; Miletic and Coffield, '89).

Unlike Peschanski ('84),we identified afferents within ZI, PH, and LH after anterograde tracer injections into the MDH. These differences are most likely due to the superior anterograde labeling produced by PHA-L. However, it is possible that the presence of a number of retrogradely labeled neurons and associated dendritic processes within ZI, as reported by Peschanski ('841, could have masked anterograde labeling after WGA-HRP injections into the MDH (Peschanski, '84). It should be noted that Peschanski ('84) does find anterograde labeling in ZI after WGA-HRP injections into the more rostrally situated trigeminal nuclei, principalis and interpolaris, suggesting spread of our tracer injections into these nuclei. We feel this is not the case, since our findings are in accord with observations in the cat and monkey. Craig and Burton ('85; also Burton et al., '79) found intense anterograde labeling in ZI after tracer injections into the cat MDH. Similarly, Apkarian and Hodge ('89) observed moderate labeling in ZI after WGAHRP injections in to the monkey lumbar spinal cord. There are also supportive observations in the rat. Dado and Giesler ('90) illustrated a retrograde tracer injection site that partially encompassed ZI and overlying structures; this injection produced moderate labeling within the MDH. In addition, Menetrey et al. ('84) were able to identify antidromically a dorsal horn neuron in rat lumbar spinal cord that projected to ZI. There is increasing evidence that the ZI is involved in processing somatosensory information. An early electrophysiological study in the monkey demonstrated that the neurons in the ventral portion of this nucleus received somatosensory inputs (Kruger and Albe-Fessard, '60). Behavioral experiments in the cat indicate that electrical stimulation of the ZI area produces escape behavior (Kaelber, '77; Kaelber and Smith, '79). Anatomically, ZI has connections with a number of brainstem nuclei in the rat that are known to process somatosensory information, including the dorsal column nuclei and the periaqueductal gray matter (Roger and Cadusseau, '85; ShammahLagnado et al., '85; Spencer et al., '88). Our findings of dense axonal and varicosity labeling in the ventral portion of ZI after injection of PHA-L in the deep laminae of MDH support a role of the ZI in somatosensory processing. ZI also receives generous limbic input (Shammah-Lagnado et al., '85). It may be that integrated sensory and limbic input in the ZI is regulated by the somatosensory cortex (Roger and Cadusseau, '85).

VPPC projections In both the rat (Norgren and Leonard, '71, '75) and cat (Nomura et al., '79; Nishikawa et al., '881, the VPPC receives abundant sensory input from the tongue. Along with this lingual input, neurons in the VPPC receive inputs

417

MDH PROJECTIONS TO DIENCEPHALON from the solitary tract and the pontine parabrachial nucleus (rat: Norgren and Leonard, '71, '75; Cechetto and Saper, '87; cat: Nomura et al., '79). Electrophysiological studies in the cat have shown that neurons in VPPC respond to innocuous thermal, mechanical, and chemical stimuli of the tongue (Ganchrow and Erickson, '72; Nishikawa et al., '88). In addition, the rat VPPC has been shown to receive a dense projection from the parabrachial nucleus (Cechetto and Saper, '87) that overlaps with VPPC inputs from the superficial MDH, as demonstrated in the present study. This pathway from the parabrachial nucleus is thought to carry visceral information to the thalamus and relay it to the insular cortex. Since neurons in the superficial MDH respond almost exclusively to noxious peripheral stimulation, our findings suggest that neurons in the ventral portion of VPPC may be involved in processing nociceptive information from the tongue, as well as gustatory sensation.

LH and PH projections We found axonal and varicosity labeling in LH after injection of PHA-L in either the superficial or deep laminae of MDH. This is consistent with Burstein et al. ('901, who reported that neurons in the rat MDH were labeled after large Fluoro-Gold injections into the lateral hypothalamus. Since only sparse retrograde labeling was seen in the MDH after medial hypothalamic injections, Burstein et al. ('90) went on to propose that the lateral but not medial hypothalamus processes trigeminal nociceptive information. The present findings have identified an additional hypothalamic region receiving trigeminal input; terminal labeling was identified in the posterior hypothalamus after PHA-L injections into deep laminae of the MDH. Furthermore, the posterior hypothalamus appears to receive a higher degree of MDH input than does the LH (see Figs. 3-6, 10). It is likely that the injections sites of Burstein et al. ('90) were rostra1 to the major locus of anterograde terminal labeling in the present study, thereby missing trigeminal input to the posterior hypothalamus. In the present study, anterogradely labeled hypothalamic fibers or terminals were rarely encountered ipsilateral to PHA-L injections into the MDH. It is possible that our small tracer injections could have a bias toward labeling major pathways with a resultant underestimation of the ipsilateral projections. However, the present data is consistent with Burstein et al. ('go), who reported that the contralateral trigeminohypothalamic projection was 3-4 times as large as the ipsilateral projection. Indicative of the multifunctionary role of the hypothalamus is the wide variety of afferent inputs it receives (Shigenaga et al., '73; Scott, '77; Kanouse et al., '85; Ono et al., '86; Kai et al., '88; Ter Horst et al., '89; Yamamoto et al., '89). Some neurons in the hypothalamus respond to noxious stimulation of the body (Kai et al., '88), while other neuron types respond to or precede changes in heart rate, blood pressure, respiration, and feeding behavior (Hamamura et al., '84; Smith and Devito, '84; Ono et al., '86; Yamamoto et al., '89). Trigeminal nociceptive input to the hypothalamus would have a potential influence on all of these somatic and visceral functions and resulting behaviors.

Varicosities and terminal labeling Electron microscopic studies describe at least three different types of varicosities in VPM and SUB of the rat: large

round (LR), small round (SR), and flat (F) types (Peschanski and Ralston, '85; Peschanski et al., '85; Ma et al., '87a,b, '88). It was proposed that nonnoxious and noxious information originating from the spinal cord or MDH were transmitted to thalamic neurons by morphologically similar axonal terminals of the LR type (Ma et al., '87a,b, '88). The size of labeled varicosities identified in VPM and SUB in the present study fall within the range of LR terminal size seen in these nuclei using electron microscopy, with LR terminals in the SUB being smaller than those in the VPM. Since most MDH input to VPM and all input to the SUB originates from the superficial laminae of MDH, it follows that LR terminals originate primarily from the superficial laminae. Large terminals may be indicative of secure input onto target neurons (Ralston, '83). This secure source of sensory input could account for the robust, long-lasting excitation of neurons in the SUB of both the rat (Dostrovsky et al., '87; Dostrovsky and Guilbaud, '88, '90; Miletic and Coffield, '89) and cat (Craig, '87) after application of noxious peripheral stimulation. On the other hand, the smaller varicosities identified in the PO, W P C , ZI, PH, and LH appear to originate from either superficial or deep laminae of the MDH. We are not aware of any electron microscopic studies that have specifically examined the morpholoa of afferents in the PO, VPPC, ZI, LH, or PH. However, the small varicosities identified in PO, VPPC, and 21 resemble the numerous SR terminals identified throughout the thalamus (for review, see Ralston, '83).Although a large proportion of SR terminals originate from cortical neurons, the origin of the remainder remains a question (Jones, '85). Some of these terminals may originate from the MDH and spinal cord. We can make little comment about the pattern of terminal labeling in the hypothalamus, except to note that spinohypothalamic tract neurons are known to have slower conduction velocities than neurons projecting to the ventrobasal thalamus (compare Giesler et al., '76, '81 to Burstein et al., '87). It is likely these slow conduction velocities are a function of the thin afferents reaching the hypothalamus as seen in the present study. Hecently, retrograde tracing techniques have been used to compute the total number of spinal cord neurons projecting to either the thalamus or hypothalamus in the rat (Burstein et al., '90a,b). It was found that comparable numbers of cells project to the thalamus and hypothalamus, at least 9,500 and 9,000, respectively. From these values, equivalent terminal densities within the thalamus and hypothalamus might be predicted after PHA-L injections into the MDH. This was indeed the case, with thalamic nuclei having a mean density of 476 93 terminals/30 x 100 pm2 and hypothalamic nuclei a mean density 154 terminals/30 x 100 pm2. Terminal density of 457 cannot be compared directly to cell numbers, since a given cell may have extensive terminal arborization; however, the ratio of our values (hypothalamic terminalsithalamic terminals = 0.96) is close to that of Burstein et al. ('90a,b; spinohypothalamic neuronsispinothalamic neurons = 0,951, suggesting similar divisions of ascending trigeminal and spinal pathways within the diencephalon. Differences in the terminal density seen within different diencephalic nuclei after PHA-L injections reflect the degree of input originating from the MDH, but not necessarily the efficacy of these inputs. For instance, large, round terminals within VPM (Ralston, '83) can simultaneously interact with several different targets (Peschanski et a]., '85; Ma et al., '87a). Multiple synapses of a single terminal

*

*

418

K. IWATA ET AL.

could compensate for the relatively low density of terminals within VPM. Another consideration is the location of the synapse on the target neuron. Trigeminothalamic neurons tend to synapse onto proximal dendrites within VPM and SUB (Peschanski et al., '85). Such input is more likely to influence cell activity than distal input. Despite such factors, the combination of low terminal density and small terminal size within VPPC and LH suggests that these nuclei do not play a significant role in processing input from the MDH.

CONCLUSIONS In the present study we have characterized the projection patterns of anterogradely labeled axons and varicosities terminating primarily in either the dorsal (VPM, SUB, PO, and VPPC) or ventral (ZI and PH) diencephalon after injection of PHA-L into either the superficial or deep laminae of the MDH, respectively. These data suggest the existence of two distinct pathways originating primarily from either superficial or deep laminae of the MDH. It is likely these pathways are carrying nociceptive information (Willis and Coggeshall, '78; Dubner and Bennett, '83) as well as thermal and tactile sensation (Dostrovsky and Hellon, '78; Willis and Coggeshall, '78). Input originating from the MDH may be subject to parallel processing via these multiple pathways. Whereas input from the deeper laminae appears to be involved in the motivational-affective component of pain (ZI, PHI, input originating from the superficial MDH becomes involved in both the sensorydiscriminative (VPM) and motivational-affective (SUB) aspects of pain (Melzack and Casey, '68; Price and Dubner, '77).

NOTE ADDED IN PROOF' Since this paper was first submitted, Cliffer et al. (J. Neurosci. 11:852-8681, and Craig (J. Comp. Neurol. 313: 377-393) have used PHA-L to identify ascending spinall trigeminal pathways to the diencephalon in the rat and cat, respectively.

LITERATURE CITED Albe-Fessard, D., B. Nashold, B. Pollin, and A. Woda (1977) Thalamic and mid-brain responses to dental pulp afferent messages in awake cats. J. Physiol. (Paris) 73:347-357. Amano, N., J.W. Hu, and B.J. Sessle (1986) Responses of neurons in feline trigeminal subnucleus caudalis (medullary dorsal horn) to cutaneous, intraoral, and muscle afferent stimuli. J. Neurophysiol. 55227-243. Apkarian, A.V., and C.J. Hodge (1989) Primate spinothalamic pathways: 111. Thalamic terminations of the dorsolateral and ventral spinothalamic pathways. J. Comp. Neurol. 288:493-511. Burstein, R., K.D. Cliffer, and G.J. Giesler, Jr. (1987) Direct somatosensory projections from the spinal cord to the hypothalamus and telencephalon. J. Neurosci. 7~4159-4164. Burstein, R., K.D. Chiffer, and G.J. Giesler, Jr (1990) The cells of origin of the spinothalamic tract of the rat: a quantitative reexamination. Brain Res. 511:329-337. Burstein R., K.D. Cliffer, and G.J. Giesler, Jr. (1990) Cells of origin of the spinohypothalamic tract in the rat. J. Comp. Neurol. 291:329-344. Burton, H., and A.D. Craig, Jr. (1979) Distribution of trigeminothalamic projection cells in cat and monkey. Brain Res. 161:515-521. Burton, H., A.D. Craig, Jr., D.A. Poulos, and J.T. Molt (1979) Efferent projections from temperature sensitive recording loci within the marginal zone of the nucleus caudalis of the spinal trigeminal complex in the cat. J. Comp. Neurol. 183753-777.

Bushnell, M.C., G.H. Duncan, D. Dubner, and L.F. He (1984) Activity of trigeminothalamic neurons in medullary dorsal horn of awake monkeys trained in a thermal discrimination task. J. Neurophysiol. 52:170-187. Casale, E.J., A.R. Light, and A. Rustioni (1988) Direct projection of the corticospinal tract to the superficial laminae of the spinal cord in the rat. J. Comp. Neurol. 278.275-286. Cechetto, D.F., and C.B. Saper (1987) Evidence for a viscerotopic sensory representation in the cortex and thalamus in the rat. J. Comp. Neurol. 262:27-45. Cliffer, K.D., and G.J. Giesler, J r . (1988) PHA-L can he transported anterogradely through fibers of passage. Brain Res. 458:185-191. Craig, A.D., Jr. (1987) Medial thalamus and nociception: The nucleus submedius. In J.M. Besson, G. Guilbaud, and M. Peschanski (eds): Thalamus and Pain. Amsterdam: Elsevier, pp. 227-243. Craig, A.D., J r . , and H. Burton (1981) Spinal and medullary lamina I projection to nucleus submedius in medial thalamus: A possible pain center. J. Neurophysiol. 45443-466. Craig, A.D., Jr., and H. Burton (1985) The distribution and topographical organization in the thalamus of anterogradely transported horseradish peroxidase after spinal injections in cat and raccoon. Exp. Brain Res. 58:227-254. Craig, A.D., Jr., S.J. Wiegand, and J.L. Price (1982) The thalamo-cortical projection of the nucleus submedius in the cat. J. Comp. Neurol. 206:28-48. Dado, R.J., and G.J. Giesler, Jr. (1990) Afferent input to nucleus submedius in rats: Retrograde labeling of neurons in the spinal cord and caudal medulla. J. Neurosci. 10:2672-2686. Dostrovsky, J.O., and R.F. Hellon (1978) The representation of facial temperature in the caudal trigeminal nucleus of the cat. J. Physiol. (Lond.) 277:29-47. Dostrovsky, J.O., and G. Guilhaud (1988)Noxious stimuli excite neurons in nucleus submedius of the normal and arthritic rat. Brain Res. 460:269280. Dostrovsky, J.O., and G. Guilbaud (1990) Nociceptive responses in medial thalamus of the normal and arthritic rat. Pain 40;93-104. Dostrovsky, J.O., J.G. Broton, and N.K. Warma (1987) Functional properties of subnucleus caudalis lamina I neurons projecting to nucleus submedius. In R.F. Schmidt, H.G. Schaible, and C. Vahle-Hinz (eds): Fine M e r e n t Nerve Fibers and Pain. Weinheim: VCH, pp. 355-366. Dubner, R., and G.J. Bennett (1983) Spinal and trigeminal mechanisms of nociception. Annu. Rev. Neurosci. 6:381418. Fukushima, T., and F.W.L. Kerr (1979) Organization of trigeminothalamic tracts and other thalamic afferent systems of the brainstem in the rat: Presence of gelatinosa neurons with thalamic connections. J. Comp. Neurol. 183:169-184. Ganchrow, D., and R.P. Erickson (1972) Thalamocortical relations in gustation. Brain Res. 36:289-305. Gerfen, C.R., and P . Sawchenko (1984) An anterograde neuroanatomical tracing method that shows the detail morphology of neurons, their axons and terminals: Immunohistochemical localization of an axonally transported plant lectin, Phaseolus uulguris leucoagglutinin (PHA-L). Brain Res. 290:219-238. Giesler, G.J., Jr., D. M e n h e y , G. Guilhaud, and J.M. Besson (1976) Lumbar cord neurons a t the origin of the spinothalamic tract in the rat. Brain Fies. 118:320-324. Giesler, G.J., Jr., D. MenBtrey, and A.I. Bashaum (1979) Differential origins of spinothalamic tract projections to medial and lateral thalamus in the rat. J. Comp. Neurol. 184:107-126. Giesler, G.J., Jr., R.P. Yezierski, K.D. Gerhart, and W.D. Willis (1981) Spinothalamic tract neurons that project to medial and/or lateral thalamic nuclei: Evidence for a physiologically novel population of spinal cord neurons. J. Neurophysiol. 46:1285-1308. Gobel, S., W.M. Falls, and S. Hockfield (1977) The division of the dorsal and ventral horns of the mammalian caudal medulla into eight layers using anatomical criteria. In D.J. Anderson and B. Matthews (eds): Pain in the Trigeminal Region. Amsterdam: ElsevieriNorth-Holland, pp. 443453. Gobel, S., S. Hockfield, and M.A. Ruda (1980) Anatomical similarities between medullary and spinal dorsal horns. In Y. Kawamura and R. Dubner (eds): Oral-Facial Sensory and Motor Function. Tokyo: Quintessence, pp. 211-223. Gobel, S., W.M. Falls, and E. Humphrey (1981) Morphology and synaptic connections of ultrafine primary axons in lamina I of the spinal dorsal horn: Candidates for the terminal axonal arbors of primary neurons with unmyelinated (C) axons. J. Neurosci. I: 1163-1179. Granum, S.L. (1986) The spinothalamic system of the rat. I. Locations of cells of origin. J. Comp. Neurol. 247:159-180.

MDH PROJECTIONS TO DIENCEPHALON Hamamura, M., K. Shihuki, and K. Yagi (1984) Noxious input to supraoptic neurosecretory cells in the rat. Neurosci. Res. 2 4 9 4 1 . Hayashi, H. (1985) Morphology of terminations of small and large myelinated trigeminal primary afferent fihers in the cat. J. Comp. Neurol. 240: 7 1-89. Herkenham, M. (1979) The afferent and efferent connections of the ventromedial thalarnic nucleus in the rat. J. Comp. Neurol. 183:487-518. Hoffman, D.S., R. Dubner, R.L. Hayes, and T.P. Medlin (1981) Neuronal activity in medullary dorsal horn of awake monkeys trained in a thermal discrimination task. I. responses to innocuous and noxious thermal stimuli. J. Neurophysiol. 46:409427. Hu, J.W., and B.J. Sessle (1984) Comparison of responses of cutaneous nociceptive and nonnociceptive brain stem neurons in trigeminal suhnucleus caudalis (medullary dorsal horn) and suhnucleus oralis to natural and electrical stimulation of tooth pulp. J. Neurophysiol. 5239-52. Hu, J.W., J.O. Dostrovsky, and B.J. Sessle (1981) Functional properties of neurons in cat trigeminal subnucleus caudalis (medullary dorsal horn). I. Response to oral-facial noxious and nonnoxious stimuli and projections to thalamus and suhnucleus oralis. J. Neurophysiol. 45:173-192. Iwata, K., R.L. Nahin, and D.R. Kenshalo, Jr. (1990) Medullary dorsal horn projections to the diencephalic nuclei in the rat. SOC. Neurosci. Ahst. 16:705. Jones, E.G. (1985) The Thalamus. New York: Plenum Press, pp. 43-84 and 323424. Kaelber, W.W. (1977) Suhthalamic nociceptive stimulation in the cat: Effect of sensory lesions and rostra1 fiber projections. Exp. Neurol. 56.574-597, Kaelher, W.W., and T.B. Smith (1979) Projections of the zona incerta in the cat, with stimulation controls. Exp. Neurol. 63:177-200. Kai, Y., Y. Oomura, and N. Shimizu (1988) Responses of lateral hypothalamic neurons to periaqueductal gray stimulation and nociceptive stimuli. Brain Res. 461:107-117. Kanouse, K., T. Nakayama, Y. Ishikawa, T. Hosono, T. Kaminaga, and A. Shosaku (1985) Responses of thalamic and hypothalamic neurons to scrotal warming in rats: Nonspecific responses? Brain Res. 328:207-213. Kemplay, S.,and K.E. Wehster (1989) A quantitative study of the projections of the gracile, cuneate and trigeminal nuclei and the medullary reticular formation to the thalamus in the rat. Neuroscience 32153-167. KnifRi, K.-D., and C. Vahle-Hinz (1987) Theperipheryofthe cat’s ventroposteromedial nucleus (VPM,): Nociceptive neurones. In J.M. Besson, G. Guilbaud, and M. Peschanski (eds): Thalamus and Pain. Amsterdam: Elsevier, pp. 245-257. Kruger, L., and D. Albe-Fessard (1960) Distribution of responses to somatic afferent stimuli in the diencephalon of the cat under chloralose anesthesia. Exp. Neurol. 2 4 2 2 4 6 7 . LeDoux, J.E., D.A. Ruggiero, R. Forest, R. Stornetta, and D.J. Reis (1987) Topographic organization of convergent projections to the thalamus from the inferior colliculus and spinal cord in the rat. J. Comp. Neurol. 2641123-146. Lund, R.D., and K.E. Webster (1967) Thalamic afferents from the spinal cord and trigeminal nuclei a n experimental anatomical study in the rat. J.Comp. Neurol. 130.313-328. Ma, W., M. Peschanski, and H.J. Ralston (1987a) The differential synaptic organization of the spinal and lemniscal projections to the ventrohasal complex of the rat thalamus. Evidence for convergence of the two systems upon single thalamic projections. Neuroscience 22925-934. Ma, W., M. Peschanski, and H.J. Ralston (1987h) Fine structure of the spinothalamic projections to the central lateral nucleus of the rat thalamus. Brain Res. 414:187-191. Ma, W., M. Peschanski, and P.T. Ohara (1988)Fine structure of the dorsal part of the nucleus suhmedius of the rat thalamus: An anatomical study with reference to possible pain pathways. J. Neurosci. 26:147-159. Maixner, W., R. Dubner, D.R. Kenshalo, Jr., M.C. Bushnell, and J.-L. Oliveras (1989) Responses of monkey medullary dorsal horn neurons during the detection of noxious heat stimuli. J. Neurophysiol. 62437449. Mantyh, P.W. (1983) The spinothalamic tract in the primate: A reexamination using wheat germ agglutinin conjugated to horseradish peroxidase. Neuroscience 9347-862. Mehler, W.R., M.W. Feferman, and W.J.H. Nauta (1960) Ascending axon degeneration following anterolateral cordotomy. An experimental study in the monkey. Brain 83:718-750. Melzack, R., and R.L. Casey (1968) Sensory, motivational, and central control determinants of pain. In D.R. Kenshalo, Sr. (ed): The Skin Senses. Springfield: Charles C. Thomas, pp. 4 2 3 4 4 3 .

419 Menetrey, D., J. De Pommery, and F. Roudier (1984) Properties of deep spinothalamic tract cells in the rat, with special reference to ventromedial zone of lumbar dorsal horn. J. Neurophysiol. 52:612-624. Miletic, V., and J.A. Coffield (1989) Responses of neurons in the rat nucleus submedius to noxious and innocuous mechanical cutaneous stimulation. Somatosens. Motor Res. 6:567-587. Nishikawa, Y., N. Koyama, and T. Yokota (1988) Ipsilateral somatosensory representation within the lateral subdivision of the nucleus ventralis posteromedialis parvocellularis of the cat thalamus. Brain Res. 458.394396. Nomura, S.,N. Mizuno, K. Itoh, K. Matsuda, T. Sugimoto, and Y. Nakamura (1979) Localization of parahrachial nucleus neurons projecting to the thalamus or the amygdala in the cat using horseradish peroxidase. Exp. Neurol. 64~375-385. Norgren, R., and C.M. Leonard (1971) Taste pathways in rat brainstem. Science 1973: 1136-1 139. Norgren, R., and C.M. Leonard (1975) Ascending central gustatory pathways. J. Comp. Neurol. 150:217-238. Ono, T., K. Sasaki, H. Nishino, M. Fukuda, and R. Shihata (1986) Feeding and diurnal related activity of lateral hypothalamic neurons in freely behaving rats. Brain Res. 373.92-102. Paxinos, G., and C. Watson (1982) The Rat Brain in Stereotaxic Coordinates. New York: Academic Press. Peschanski, M. (1984) Trigeminal afferents to the diencephalon in the rat. Neuroscience 12:465487. Peschanski, M., and H.J. Ralston, 111. (1985) Light and electron microscopic evidence of transneuronal labeling with WGA-HRP to trace somatosensorypathways to the thalamus. J. Comp. Neurol. 236.29-41. Peschanski, M., F. Roudier, H.J. Ralston, 111, and J.M. Besson (1985) Ultrastructural analysis of the terminals of various somatosensory pathways in the ventrobasal complex of the rat thalamus: An electronmicroscopic study using wheat germ agglutinin conjugated to horseradish peroxidase as an axonal tracer. Somatosens. Res. 3:75-87. Price, D.D., and R. Dubner (1977) Neurons that suhserve the sensorydiscriminative aspects of pain. Pain 3.307438. Price, D.D., R. Duhner, and J.W. Hu (1976) Trigeminothalamic neurons in nucleus caudalis responsive to tactile, thermal, and nociceptive stimulation of the monkey’s face. J.Neurophysiol. 39:936-953. Price, J.L., and B.M. Slotnick (1983) Dual olfactory representation in the rat thalamus: An anatomical and electrophysiological study. J. Comp. Neurol. 21563-77. Ralston, H.J., I11 (1983) The synaptic organization of the ventrohasal thalamus in the rat, cat and monkey. In G. Macchi, A. Rustioni, and R. Spreafico (eds): Somatosensory Integration in the Thalamus Amsterdam: Elsevier, pp. 241-250. Rausell, E., and E.G. Jones (1991) Chemically distinct compartments of the thalamic VPM nucleus in monkeys relay principal and spinal trigeminal pathways to different layers of the somatosensory cortex. J. Neurosci. 11.226-237. Renehan, W.E., M.F. Jacquin, R.D. Mooney, and R.W. Rhoades (1986) Structure-function relationships in rat medullary and cervical dorsal horns. 11. Medullary dorsal horn cells. J. Neurophysiol. 55;1187-1201. Rhoades, R.W., S.E. Fish, N.L. Chiaia, C. Bennett-Clarke, and R.D. Mooney (1989) Organization of the projections from the trigeminal brainstem complex to the superior colliculus in the rat and hamster: Anterograde tracing with Phaseolus uulgaris leucoagglutinin and intra-axonal injection. J. Comp. Neurol. 289:641-656. Ring, G., and D. Ganchrow (1983) Projections of nucleus caudalis and spinal cord to brain stem and diencephalon in the hedgehog (Erinaceus europaeus and Paruechinus aethiopicus): A degeneration study. J. Comp. Neurol. 216,132-151. Roger, M., and J. Cadusseau (1985) M e r e n t s to the zona incertain the rat: A combined retrograde and anterograde study. J. Comp. Neurol. 241:480492. Rosenkilde, C.E. (1979) Functional heterogeneity of the prefrontal cortex in the monkey: A review. Behav. Neurol. Biol. 25:301-345. Salt, T.E., R. Morris, and R.G. Hill (1983) Distribution of substance-Presponsive and nociceptive neurones in relation to substance P-immunoreactivity within the caudal trigeminal nucleus of the rat. Brain Res. 273.217-228. Schofield, B.R. (1990) Uptake of P h a s e o h vulgaris leucoagglutinin (PHA-L) hy axons of passage. J. Neurosci. Methods 35:47-56. Scott, J.W. (1977) A measure of extracellular unit responses to repeated stimulation applied to observations of the time course of olfactory responses. Brain Res. 132247-258.

420 Sessle, B.J., and L.F. Greenwood (1976) Inputs to trigeminal brain stem neurones from facial, oral pulp and pharyngolaryngeal tissues. I. Responses to innocuous and noxious stimuli. Brain Res. 1171211-226. Sessle, B.J., J.W. Hu, N. Amano, and G. Zhong (1986) Convergence of cutaneous, tooth pulp, visceral, neck and muscle d e r e n t s onto nociceptive and non-nociceptive neurones in trigeminal suhnucleus caudalis (medullary dorsal horn) and its implications for referred pain. Pain 27:219-235. Shammah-Lagnado, S.J., N. Negrao, and J.A. Ricardo (1985) Afferent connections of the zona incerta: A horseradish peroxidase study in the rat. Neuroscience 15:109-134. Shigenaga, Y., S. Matano, K. Okada, and A. Sakai (1973) The effect of tooth pulp stimulation in the thalamus and hypothalamus of the rat. Brain Res. 63t402-407. Shigenaga, Y., Z. Nakatani, T. Nishimori, S. Suemune, R. Kuroda, and S. Matamo (1983) The cells of origin of cat trigeminothalamic projections: Especially in the caudal medulla. Brain Res. 277201-222. Smith, O.A., and J.L. Devito (1984) Central neural integration for the control of autonomic responses associated with emotion. Annu. Rev. Neurosci. 7:43-65. Spencer, S.E., W.B. Sawyer, and A.D. Loewy (1988) L-glutamate stimulation of the zona incerta in the rat decreases heart rate and blood pressure. Brain Res. 458:72-81. Ter Horst, G.J., P. De Boer, P.G.M. Luiten, and J.D. van Willigen (1989) Ascending projections from the solitary tract nucleus to the hypothala-

K. IWATA ET AL. mus. A Phaseolus uulgaris lectin tracing study in the rat. Neuroscience 31:785-797. Tsubokawa, T., Y. Katayama, Y. Ueno, and N. Moriyasu (1981) Evidence for the involvement of frontal cortex in pain-related cerebral events in cats: Increase in local blood flow by noxious stimuli. Brain Res. 21 7r179-185. Warren, J.M., and K. Akert (1964) The Frontal Granular Cortex and Behavior. New York MacGraw-Hill. Willis, W.D., and R.E. Coggeshall(1978) Sensory Mechanisms of the Spinal Cord. New York: Plenum Press. Willis, W.D., D.L. Trevino, J.D. Coulter, and R.A. Maunz (1974) Responses of primate spinothalamic tract neurons to natural stimulation of hindpaw. J. Neurophysiol. 37t358-372. Woda, A,, J. Azerad, G. Guilbaud, and J.M. Besson (1975) Etude microphysiologique des projections thalamiques de la pulpe dentaire chez le chat. Brain Res. 891193-213. Yamamoto, T., R. Matsuo, Y. Kiyomitsu, and R. Kitamura (1989) Response properties of lateral hypothalamic neurons during ingestive behavior with special reference to licking of various taste solutions. Brain Res. 4811286-297. Yokota, T., N. Koyama, and N. Matsumoto (1985) Somatotopic distribution of trigeminal nociceptive neurons in ventrobasal complex of cat thalamus. J. Neurophysiol. 53:1387-1400. Yokota, T., Y.Nishikawa, and N. Koyama (1986) Tooth pulp input to the shell region of nucleus ventralis posteromedialis of the cat thalamus. J. Neurophysiol. 56r80-98.