Download Reprint () - Centre de recherche CERVO

THE JOURNAL OF COMPARATIVE NEUROLOGY 420:233–243 (2000)
Thalamic Projections From the WhiskerSensitive Regions of the Spinal
Trigeminal Complex in the Rat
PIERRE VEINANTE,1 MARK F. JACQUIN,2 AND MARTIN DESCHÊNES1*
Centre de Recherche Université Laval-Robert Giffard, Hôpital Robert Giffard, Québec
G1J 2G3, Canada
2
Department of Neurology, Washington University School of Medicine, St. Louis,
Missouri 63110
1
ABSTRACT
This study investigated the axonal projections of whisker-sensitive cells of the spinal
trigeminal subnuclei (SP5) in rat oral, interpolar, and caudal divisions (SP5o, SP5i, and
SP5c, respectively). The labeling of small groups of trigeminothalamic axons with biotinylated dextran amine disclosed the following classes of axons. 1) Few SP5o cells project to
the thalamus: They innervate the caudal part of the posterior group (Po) and the region
intercalated between the anterior pretectal and the medial geniculate nuclei. These fibers
also branch profusely in the tectum. 2) Two types of ascending fibers arise from SP5i:
Type I fibers are thick and distribute to the Po and to other regions of the midbrain, i.e.,
the prerubral field, the deep layers of the superior colliculus, the anterior pretectal
nucleus, and the ventral part of the zona incerta. Type II fibers are thin; branch sparsely
in the tectum; and form small-sized, bushy arbors in the ventral posterior medial nucleus
(VPM). Accordingly, a statistical analysis of the distribution of antidromic invasion
latencies of 96 SP5i cells to thalamic stimulation disclosed two populations of neurons:
fast-conducting cells, which invaded at a mean latency of 1.23 ⫾ 0.62 msec, and slowconducting cells, which invaded at a mean latency of 2.97 ⫾ 0.62 msec. 3) The rostral part
of SP5c contains cells with thalamic projections similar to that of type II SP5i neurons,
whereas the caudal part did not label thalamic fibers in this study. A comparison of SP5i
projections and PR5 projections in the VPM revealed that the former are restricted to
ventral-lateral tier of the nucleus, whereas the latter terminate principally in the upper
two tiers of the VPM. These results suggest a functional compartmentation of thalamic
barreloids that is defined by the topographic distribution of PR5 and type II SP5i
afferents. J. Comp. Neurol. 420:233–243, 2000. © 2000 Wiley-Liss, Inc.
Indexing terms: barrels; barreloids; vibrissa; ventral posterior medial nucleus; posterior group
nucleus; biotinylated dextran
Brainstem nuclei that receive vibrissal primary afferents
in rats include the principal trigeminal nucleus (PR5) and all
subdivisions of the spinal trigeminal complex (SP5). Each of
these (sub)nuclei contributes axons to the trigeminothalamic
tract, but the main stream of ascending fibers arises from
PR5 and from the interpolar division of SP5 (SP5i; Smith,
1975). Bulk labeling with anterograde tracers has shown
that both PR5 and SP5i axons innervate the ventral posterior medial (VPM) and the posterior group (Po) nuclei (Peschanski, 1984; Chiaia et al., 1991a; Williams et al., 1994).
For the moment, the axonal distribution of whisker-sensitive
cells of the oral and caudal divisions of SP5 (SP5o and SP5c,
respectively) remains ill defined.
© 2000 WILEY-LISS, INC.
At each level of the neuroaxis, the peripheral arrangement of the vibrissae is maintained in arrays of cellular
aggregates referred to as barrelettes (brainstem), barre-
Grant sponsor: Medical Research Council of Canada; Grant number:
MT-5877; Grant sponsor: National Institutes of Health; Grant numbers:
DE07662 and DE07734; Grant sponsor: UMDNJ Foundation; Grant sponsor: American Osteopathic Association.
*Correspondence to: Martin Deschênes, Centre de Recherche Université
Laval-Robert Giffard, Hôpital Robert Giffard, 2601 de la Canardiére, Québec G1J 2G3, Canada. E-mail: [email protected]
Received 11 October 1999; Revised 27 December 1999; Accepted 30
December 1999
234
P. VEINANTE ET AL.
loids (thalamus), and barrels (cortex). Whereas the
whisker-like patterning of the terminal fields of PR5 axons in rat VPM has been well documented both at the
ensemble level and the single-cell level (Williams et al.,
1994; Veinante and Deschênes, 1999), the topographic
distribution of SP5i projections remains unclear. Previous
studies reported either a complete overlap with PR5 terminal fields (Peschanski, 1984) or a patchy, scattered distribution lacking a clear, barreloid-like arrangement
(Chiaia et al., 1991a; Williams et al., 1994). Whatever the
case, SP5i terminals should be relatively abundant across
the field of barreloids, because electron microscopic studies consistently reported SP5i profiles presynaptic to VPM
cells dendrites (Chiaia et al., 1991a; Wang and Ohara,
1993; Williams et al., 1994). In addition, large numbers of
VPM neurons exhibit an overt multiwhisker responsiveness in PR5 lesioned rats (Rhoades et al., 1987) that
disappears after subsequent ablation of the SP5i (see also
Freidberg et al., 1999). For the moment, these ultrastructural and physiologic results are difficult to reconcile with
those provided by tract-tracing studies.
A confusing issue with most studies that used massive
tracer injections to map SP5 projections to the thalamus
relates to their lack of specificity with regard to the different populations of orofacial afferents. This makes it
hard to reach strong conclusions about the projections of
whisker-sensitive trigeminothalamic cells. A second problem concerns the heterogeneity of this cellular population:
Intracellular staining of cells antidromically invaded from
the thalamus revealed various morphologic types of
vibrissa-responsive neurons across the SP5 subnuclei
(Jacquin et al., 1986a, 1988, 1989; Renehan et al., 1986;
Jacquin and Rhoades, 1990). A similar diversity was highlighted by parvalbumin and calbindin immunostaining of
retrogradely labeled trigeminothalamic cells (BennettClarke et al., 1992). It appears likely that these different
populations of neurons give rise to different patterns of
axonal projections. When studied at a single-cell level, for
instance, the axonal projections from the PR5 demonstrate far more complexity than what is apparent immediately in conventional tract-tracing studies (Veinante
Abbreviations
3v
An
Ang
APT
AV
fr
LG
MG
Po
PR5
PRF
RT
SC
sm
SP5
SP5c
SP5i
SP5o
TR
VL
VPL
VPm
ZI
ZId
third ventricle
anterior group nuclei
angular nucleus
anterior pretectal nucleus
anterior ventral nucleus
fasciculus retroflexus
lateral geniculate nucleus
medial geniculate nucleus
posterior group nucleus
principal trigeminal nucleus
prerubral field
reticular thalamic nucleus
superior colliculus
stria medullaris
spinal trigeminal nucleus
caudal division of the spinal trigeminal nucleus
interpolar division of the spinal trigeminal nucleus
oral division of the spinal trigeminal nucleus
thalamic reticular nucleus
ventral lateral nucleus
ventral posterior lateral nucleus
ventral posterior medial nucleus
zona incerta
dorsal division of the zona incerta
and Deschênes, 1999). In the current study, we sought to
determine the single-cell make-up of SP5 projections to
the thalamus by labeling small groups of axons that arise
from the whisker-responsive regions of the three SP5 subnuclei.
MATERIALS AND METHODS
Anterograde labeling experiments
Experiments were carried out in 52 adult rats (SpragueDawley) in accordance with federally prescribed and university animal care and use guidelines (Olfert et al., 1993).
Rats were anesthetized with ketamine (75 mg/kg) and
xylazine (5 mg/kg), and injections of biotinylated dextran
amine (BDA; molecular weight, 10,000; Molecular Probes,
Eugene, OR) were made with small-sized micropipettes
(5– 8 ␮m) into the three SP5 subnuclei. The stereotaxic
coordinates of the atlas of Paxinos and Watson (1986)
were used to target the injections in SP5o and SP5i. The
subnucleus caudalis and the caudal part of SP5i were
reached through the opening of the cisterna magna by
using the obex as a landmark. After recording vibrissaeevoked responses to ascertain the correct placement of the
injections, BDA (2% in 0.5 M potassium acetate) was
ejected by iontophoresis (positive current pulses of 300 –
1,000 nA, 500 msec duration, half-duty cycle for 30 minutes). After a survival period of 5–7 days, animals were
perfused with saline followed by a fixative containing 4%
paraformaldehyde and 0.5% glutaraldehyde in 0.1 M
phosphate buffer (PB), pH 7.4. Brains were postfixed in
the same fixative for 2 hours, cryoprotected overnight in
30% sucrose, and cut at 75 ␮m on a freezing microtome
along the horizontal plane. Sections were processed for
cytochrome oxidase and BDA histochemistry according to
previously described protocols (Wong-Riley, 1979;
Horikawa and Armstrong, 1988). Finally, sections were
mounted on gelatin-coated slides, dehydrated in alcohols,
cleared in toluene, and coverslipped without counterstaining. Labeled trigeminothalamic fibers were drawn with a
camera lucida by using ⫻25 or ⫻40 objectives. Photomicrographs were taken with a digital camera (Agfa-Gevaert
NV, Montsel, Belgium) and were processed with Photoshop software (version 3.0; Adobe Systems, Mountain
View, CA). Apart from brightness and contrast adjustments, they were printed without modifications.
Control experiments
Control experiments were made to address anatomic
issues raised by the anterograde labeling experiments.
These included calbindin immunostaining of sections containing labeled fibers to determine whether some of these
fibers terminate in the anterior intralaminar nuclei. After
bilateral injections of BDA into SP5i, one brain was cut at
50 ␮m along the frontal plane, and alternate sections were
processed either for cytochrome oxidase and BDA reactions or for revealing BDA and calbindin immunoreactivity. Once the BDA reaction was terminated, sections were
rinsed three times in 0.01 M phosphate-buffered saline
(PBS, pH 7.4), and incubated overnight in a solution containing 5% normal horse serum, 0.2% Triton X-100, and
anticalbindin D-28K antiserum (Sigma; St. Louis, MO;
dilution, 1:2,500). After three rinses in PBS, sections were
incubated for 1 hour in the secondary antibody (biotinylated horse immunoglobulin G; Vector Laboratories, Bur-
SP5 PROJECTIONS IN THE RAT
235
Fig. 1. Photomicrographs of biotinylated dextran amine (BDA) injection sites in the oral (A), interpolar (B), and caudal (C) divisions of the spinal trigeminal nucleus (SP5o, SP5I, and SP5c, respectively).
All photomicrographs were taken from horizontal sections. 7n, Seventh nerve. Scale bar ⫽ 500 ␮m.
lingame, CA), rinsed again three times in PBS, and reacted with the avidin-biotin-peroxidase complex (ABC;
Vector Laboratories). Bound peroxidase was revealed by
using 3,3⬘-diaminobenzidine tetrahydrochloride (DAB;
Sigma) as a substrate.
Finally, four experiments were conducted to map in
frontal sections the terminal fields of PR5 and SP5i
afferents. Rats were anesthetized with ketamine/
xylazine, and BDA was injected iontophoretically into
the whisker-sensitive division of both subnuclei by
means of coarse micropipettes (tip diameter ⬇ 15 ␮m)
and large-current pulses (3– 4 ␮A, 7 seconds on/7 seconds off). After a survival period of 6 days, animals were
perfused, and the tissue was processed as described
above.
Electrophysiological experiments
Data pertaining to the conduction velocity of thalamic
projecting SP5i neurons were obtained in the course of a
separate series of experiments carried out previously by
one of the authors (for a description of the methods used,
see Jacquin et al., 1989). In these experiments, the latencies of antidromic invasion to VPM stimulation were measured for 96 SP5i neurons. Most of these cells responded to
whisker stimulation, and some were injected intracellularly with horseradish peroxidase. These data were used
in conjunction with the anatomic data base to assess the
hypothesis that SP5i projections to the thalamus arise
from two populations of neurons that have different conduction velocities.
RESULTS
Data base
In horizontal sections of the brainstem, cytochrome oxidase staining clearly delineated SP5 subnuclei. Six sites
were located in SP5o, 16 were located in SP5i, and 17 were
located in SP5c (see Fig. 1). The thalamic projections of 51
darkly stained trigeminothalamic fibers were reconstructed totally; then, proceeding backward, the main
trunks and collateral branches were mapped to the other
terminal sites in the mesencephalon. An additional group
of 65 axons was reconstructed in part to verify whether
their projections conformed to those of the fully reconstructed fibers.
Projections from SP5o
Previous anatomic studies reported few labeled cells in
SP5o after the injection of retrograde tracers into the
thalamus (Smith, 1975; Silverman and Kruger, 1985;
Bruce et al., 1987; Mantle-St.-John and Tracey, 1987;
Bennett-Clarke et al., 1992). Accordingly, large injections
of BDA (1 ␮A) made with relatively coarse micropipettes
(⬇15 ␮m) labeled very few SP5o trigeminothalamic fibers
(typically, 0 –3 axons per injection site). The reconstruction of these axons (n ⫽ 7) revealed a common pattern of
innervation in the caudal-ventral part of the Po and in the
crescent-shaped region adjacent to the anterior pretectal
nucleus (Fig. 2). In matching horizontal sections that were
stained for calbindin, this later region stained positively
and displayed a sieve-like appearance. It comprised the
ethmoid nucleus, the medial part of the medial geniculate
236
P. VEINANTE ET AL.
Fig. 2. Terminal fields of three SP5o axons in the thalamus and tectum. All reconstructions were
made from horizontal sections. For abbreviations, see list. Scale bar ⫽ 500 ␮m.
nucleus (MGm), and the caudalmost extension of Po intercalated between the pretectal and the medial geniculate nuclei. Axons from the SP5o formed loose clusters of
terminations in caudal Po and more dispersed clumps in
the rest of their terminal fields. As a rule, SP5o axons also
gave off a number of branches in the superior colliculus
(SC), and some innervated the dorsal division of the zona
incerta (ZId).
Projections from SP5i
In agreement with previous electrophysiological studies
(Woolston et al., 1982; Jacquin et al., 1986a, 1989), recordings made in SP5i prior to the injections disclosed almost
exclusively units that responded to the movement of multiple whiskers. On the basis of the distribution of terminations in the upper mesencephalon and thalamus and of
the shape of terminal arbors, two types of axons make up
the ascending projections from this subnucleus.
Type I axons (n ⫽ 23) have diameters of 2– 4 ␮m and
consistently showed up after all injections made into SP5i.
Collectively, these axons project to a number of sites in the
thalamus and upper brainstem. The most robust projections are to the superior colliculus, the pretectum, the
ventral division of the zona incerta (ZIv), and the prerubral field, including the parvocellular division of the red
nucleus. Terminal fields in these regions often are so
dense that even the labeling of only two or three fibers
makes their complete reconstruction hazardous. Tracing
the main branches of single axons clearly shows, however,
that most distribute terminals to each of these targets. In
the thalamus, terminal fields are less dense and are concentrated principally in Po (Figs. 3, 5A). After giving off a
series of collaterals in the mesencephalon and ZIv, type I
axons enter the thalamus and divide in a number of secondary branches that form one or multiple clusters of
medium-sized terminations (2– 4 ␮m) concentrated principally in a shell-like region over the dorsal aspect of VPM
and/or the caudal part of VPM lacking barreloids. Among
these fibers, some also innervate the dorsal part of the
ventral-lateral nucleus (VL) and/or give off branches that
make a distinct cluster of terminations in a dorsomedial
region of Po situated behind the anterior ventral nucleus.
This later region is not an actual part of the intralaminar
thalamus, as evidenced by its lack of calbindin immunoreactivity in double-stained sections. Rather, it seems to
correspond to the angular nucleus in the atlas of Paxinos
and Watson (1986). Finally, most axons distribute scattered terminals throughout caudal Po between the
parafascicular and the lateral geniculate nuclei. Although
many type I axons give off branches that ascend through
VPM, none gives off terminations in the field of barreloids.
Type II axons (n ⫽ 15) are observed more frequently
after injections made into the caudal, barrelette-patterned
region of SP5i or the rostral part of SP5c. Injections made
into the rostral SP5i label few of these axons, and, when it
is present, the staining is often faint. Most type II axons
are of small diameter (1–2 ␮m), they branch very sparsely
in the tectum and MGm, and none terminates in ZI. In
VPM, they make bushy terminal fields containing one to
three clumps of medium-sized boutons (2– 4 ␮m; see Figs.
4, 5B). In horizontal sections, terminal fields form thin
(⬇80 ␮m), narrow sagittal bands (⬇100 ⫻ 250 ␮m) that
most often are present in sections cut deep through the
thalamus.
Projections from SP5c
Whereas recordings made in SP5o and SP5i prior to the
injections disclosed almost exclusively units that responded to the movement of multiple whiskers, the recordings made in SP5c yielded a number of single-whisker
cells; however, very few of these cells seem to project to the
thalamus. Of 17 injections made into this subnucleus, only
those located rostrally (n ⫽ 6 injections), within ⬇0.8 mm
from the caudal border of SP5i, produced anterograde
labeling in the thalamus. Other injection sites resulted in
heavy staining of intersubnuclear axons that formed a
SP5 PROJECTIONS IN THE RAT
237
Fig. 3. A–D: Terminal fields of type I SP5i axons in the posterior
group nucleus (Po) and the upper brainstem. Thalamic arborizations
are reconstructed fully, and arrows indicate the other terminal sites
in the midbrain. Note the absence of terminations in the ventral
posterior medial nucleus (VPM) and the presence of distinct clusters
in the angular and ventral lateral nuclei. For other abbreviations, see
list. Scale bar ⫽ 500 ␮m.
thin, sagittal band across the full length of the trigeminal
complex; however, the thalamus and upper brainstem
were completely free of labeling in these cases. Sites that
give rise to thalamic projections are characterized by the
presence of multiwhisker cells. Trigeminothalamic axons
arising from these sites (six fibers were reconstructed) are
238
P. VEINANTE ET AL.
Fig. 4. A–D: Terminal fields of type II SP5i axons in the VPM. Note the small sizes and elongated
aspect of terminal fields, which are ⬇80 ␮m thick. Drawings were made from horizontal sections. For
abbreviations, see list. Scale bar ⫽ 200 ␮m.
of small diameter (1–2 ␮m) and form bushy terminal fields
similar to those made by type II SP5i fibers. Likewise,
terminal fields are distributed principally in the ventral
part of VPM (Fig. 4).
Small numbers of thin-diameter axons labeled from the
rostral part of SP5c projected to both VPM and Po. They
arborized in the caudal part of VPM that lacks barreloids
and in the neighboring region of Po. Although none of
these fibers was drawn completely, partial reconstruction
revealed that they also give off terminal branches in MGm
and tectum.
Comparative distribution of PR5 and SP5i
projections in VPM
A remarkable feature of SP5i axons that terminated in
VPM was their preferential distribution in horizontal sec-
SP5 PROJECTIONS IN THE RAT
239
Fig. 5. Photomicrographs of SP5 terminal fields in the thalamus. A: Terminal field of type I axons in
the rostral Po (shell region over the VPM). B: A single type II axon in the VPM. Note the clumped
terminations of type II axons. Scale bars ⫽ 100 ␮m in A; 50 ␮m in B.
tions cut deep through the thalamus. Therefore, we cut
two brains along the frontal plane to visualize better the
location of terminal fields. After a large BDA injection into
the whisker-responsive division of SP5i (two cases), the
anterograde labeling in VPM was found exclusively in the
ventral-lateral tier of the nucleus that is contiguous to the
ventral posterior lateral nucleus (VPL). Figure 6A shows
the distribution of darkly labeled terminal fields in a representative frontal section passing through VPM. After a
similar injection made into PR5 (two cases), labeled terminal fields were found in a more restricted zone of VPM
corresponding to the size of one or two barreloids. This
zone contains dense clusters of terminations, principally
in the upper two tiers of the nucleus, with a sparser
distribution of boutons in the lower tier adjacent to VPL
(Fig. 6B,C). Thus, these results show that the terminal
fields of PR5 and SP5i axons form complementary projection patterns in VPM, except in the lower tier of the
nucleus, where both projections seem to overlap.
Conduction velocities of thalamic projecting
SP5i cells
The anatomic results described above show that SP5i
projections to the thalamus arise from two classes of fibers
that have different diameters. One would expect this feature to be related to cellular populations that have differ-
ent axonal conduction velocities. The histogram of Figure
7 shows the distribution of the antidromic invasion latencies of 96 SP5i units that were backfired from the contralateral VPM. At first glance, this distribution looks
neither clearly unimodal nor bimodal. Thus, a test of bimodality was carried out to determine whether the distribution contains one population or a mixture of two populations of neurons (McLachlan and Basford, 1988). This
test compared the likelihood of two hypothesis: The observed distribution of latencies arises either from a single
population of normally distributed latencies (hypothesis
P1) or from a mixture of two populations of normally
distributed latencies (hypothesis P2). In the latter case, it
was assumed that the variance within the two populations
was the same, but the mean latency was allowed to differ.
The likelihood ratio of the two hypothesis (L2/L1) indicates
which of hypothesis is more likely and twice its logarithmic value follows a ␹2 distribution that can be used to test
the null hypothesis of a single population (hypothesis P1)
against the alternative hypothesis (P2). This test clearly
favors the P2 hypothesis, with a likelihood ratio ⫽ 518
(␹2 ⫽ 12.5; 2 degrees of freedom; P ⫽ 0.002). Accordingly,
the histogram would contain two populations of thalamic
projecting cells consisting of 74% of fast-conducting cells
invaded at a mean latency of 1.23 ⫾ 0.62 msec (P1) and
240
P. VEINANTE ET AL.
Fig. 6. A–C: Anterograde labeling of principal trigeminal nucleus
(PR5) and SP5i afferents in the VPM. A shows the distribution of SP5i
type II terminal fields in the ventral lateral tier of the VPM. Scattered
black dots in the VPM and Po are fragments of labeled axons. The
distribution of PR5 afferents is shown in two consecutive sections in B
and C. Note that terminal fields are concentrated mainly in the upper
two tiers of the VPM. All photomicrographs were taken from frontal
sections, and asterisks delineate the VPM. Scale bars ⫽ 500 ␮m in A;
250 ␮m in B,C.
26% of slow-conducting cells invaded at a mean latency of
2.97 ⫾ 0.62 msec (P2).
cells, we demonstrated the existence of two main types of
ascending fibers to the thalamus: thick fibers that distribute to the Po and to other regions of the midbrain and thin
fibers that arborize principally in the VPM. In addition,
the SP5 projection to VPM was shown to be restricted to
the lower tier of the nucleus, where PR5 projections are
sparse but present.
DISCUSSION
Two principal findings resulted from this study. By using BDA to label small groups of whisker-sensitive SP5
SP5 PROJECTIONS IN THE RAT
241
collicular and incertal cells to whisker displacements
(Tiao and Blakemore, 1976; Chalupa and Rhoades, 1977;
Yamasaki and Krauthamer, 1990; Chiaia et al., 1991b;
Diamond et al., 1992; Nicolelis et al., 1992) also matches
with the projection patterns observed in the current study.
Thus, it seems unlikely that our anatomic data base was
contaminated much by nonwhisker-related axons.
Comparison with previous studies
Fig. 7. Distribution of antidromic invasion latencies of 96 SP5i
cells backfired from the thalamus. For explanation of statistics, see text.
Technical comments
BDA is a very efficient and stable anterograde tracer
that stains cell processes in a Golgi-like manner. This
tracer seemingly enters into the cells through the smallsized axonal and dendritic branches that are severed by
the micropipette (Pinault, 1996; Chen and Aston-Jones,
1998). This is probably the reason why labeling appears
solid and uniform instead of granular, as with tracers that
can be taken up actively (e.g., lectins or cholera toxin).
Thus, after extracellular deposits, the labeling of fibers of
passage is unavoidable. This drawback, which is common
to most tracers, can be minimized by the use of small-sized
micropipettes. The fact that injections made into the different trigeminal subnuclei labeled axons with different
projection patterns is a good indication that BDA uptake
through axon collaterals is not a critical issue in the current study. However, when large pipettes are used to
pressure inject tracers, as in most previous studies, the
probability of breaking fine collaterals increases as the
square of the pipette diameter. Large injections into PR5,
for instance, in combination with the use of sensitive histochemical methods, are likely to yield false-positive results, because many trigeminothalamic axons from SP5i
send collaterals to this nucleus (Jacquin et al., 1986a,
1990).Consequently, labeling in the VPM may appear
more extensive than what we report in this study and may
be more abundant in the lower tier of the nucleus, where
we found only a sparse projection. This contention is supported by the spatial distribution of the terminal fields of
single PR5 axons that were labeled in a previous series of
experiments (Veinante and Deschênes, 1999). Reexamination of these data confirms the preferential arborization of
PR5 fibers in the two dorsal-medial tiers of VPM.
Tracer injections were made into the whiskerresponsive regions of SP5. It remains possible, however,
that some labeled fibers may have receptive fields on
neighboring regions of the mystacial pad (i.e., guard hairs,
intervibrissa fur). This possibility actually is impossible to
dismiss; however, insofar as the projections to the barreloids are concerned, there is no doubt that they arise from
the whisker-sensitive populations of the trigeminal complex. The well-documented responsiveness of Po and of
Thalamic projections from the SP5 have been described
previously in studies using anterograde degeneration and
a variety of neuronal tracers (Lund and Webster, 1967;
Smith, 1975; Erzurumlu and Killackey, 1980; Peschanski,
1984; Chiaia et al., 1991a; Williams et al., 1994). Because
most studies used massive lesions or injections that likely
involved many types of second-order orofacial afferents,
there exist some minor mismatches between prior results
and the current set of whisker-related data. For instance,
we did not observe any SP5 projection to the submedius or
to the anterior intralaminar nuclei, likely because these
regions receive input from afferents of other submodalities
(Ma et al., 1988). Labeled terminal fields were not observed in the ventral-medial wing of VPM, which contains
the upper lip and lower mandibular representations. However, the other projection sites to the thalamus and the
midbrain were confirmed again by the current study.
Although it can never be said that the current sample of
SP5 axons is representative of all fiber types that make up
the ascending projections from this division of the trigeminal
complex, it came rather as a surprise that none of the labeled
fibers actually innervated both the field of barreloids and the
Po. Such axons were expected from prior double-retrograde
labeling studies, which estimated that they constitute ⬇8%
of the ascending contingent from SP5i (Chiaia et al., 1991a).
Our negative result hardly can be ascribed to the small size
of the injections and, thus, to a sampling bias, because injections of similar size made into the PR5 showed that 4% of
ascending fibers from that nucleus terminate in both VPM
and Po (Veinante and Deschênes, 1999). Interpolaris cells
with a dual thalamic projection, thus, may be absent or very
few in number. The double-retrograde labeling observed after injections of tracers into VPM and Po likely resulted from
tracer uptake by damaged branches of SP5i axons that ascend through VPM to reach Po.
All electrophysiological studies agree on the fact that
whisker-sensitive SP5i cells that project to the thalamus
have multiwhisker receptive fields (Woolston et al., 1982;
Jacquin et al., 1986a, 1989). In one of these studies (Jacquin et al., 1986a), SP5i neurons were backfired from the
thalamus and labeled intracellularly. Although the numbers of neurons studied were small, there was a clear
dichotomy among this cellular population (see Table 1 in
Jacquin et al., 1986a). Half of the sample consisted of cells
with thick axons (2–5 ␮m) that were invaded antidromically at latencies shorter than 1 msec, whereas the rest of
the sample consisted of cells with thinner axons (1–2 ␮m)
that were backfired at latencies longer than 1.5 msec.
These results are fully in line with the two types of axons
that were labeled by our BDA injections and with the
statistical analysis of the distribution of antidromic latencies. Thus, it seems that two populations of neurons relay
vibrissae information from SP5i to the thalamus: fastconducting cells that project to Po and slow-conducting
cells that project to VPM. In all likelihood, these two
classes of neurons correspond to the large and small tri-
242
P. VEINANTE ET AL.
geminothalamic cells that were identified previously in
the SP5i by Phelan and Falls (1991).
Ultrastructural and electrophysiological studies of the trigeminal projections in the rat VPM provided evidence that
both PR5 axons and SP5i axons make synapses with the
same relay cells (Chiaia et al., 1991b; Wang and Ohara,
1993; Freidberg et al., 1999). In light of the current data, it is
clear that both types of prethalamic afferents should contact
a fair proportion of cells in a thalamic barreloid. Only cells
situated dorsally in barreloids may have dendrites that are
out of reach of type II SP5i axons. However, if their dendrites
extend across the VPM/Po border, then these neurons may
be the target of PR5 axons and type I afferents from the SP5i
that terminate in the shell region of the VPM. For the moment, this remains an open issue, and additional labeling
studies will be required to settle the question. The segregation of SP5i axons in the ventral-lateral tier of the VPM
suggests a functional compartmentation within barreloids.
Such a notion has been introduced and discussed previously
by Land et al. (1995) on the basis of the different cytochrome
oxidase reactivity of the dorsal and ventral aspects of barreloids and of their differential retrograde labeling after horseradish peroxidase injections made at different depths of a
barrel column. Here again, a single-cell study may reveal an
as yet unsuspected specificity of connections between different regions of a thalamic barreloid and its corresponding
cortical barrel.
Projections from SP5i to barreloids
In sections cut in an oblique sagittal plane through the
VPM, cytochrome oxidase-rich barreloids form curved, tapering cylinders that extend through the thickness of the
nucleus (Land et al., 1995). In the dorsal-medial part of
the VPM, the long axis of barreloids lies normal to the
VPM/Po border but bends horizontally in the ventrallateral part adjacent to VPL. Most PR5 trigeminothalamic
fibers that project to VPM have single-whisker receptive
fields and form small-sized, bushy arbors that are restricted to the dimension of a single barreloid (Williams et
al., 1994; Veinante and Deschênes, 1999). Although the
shape of individual barreloids could not be visualized in
our material, the terminal field of single type II axons
from the SP5i appears isomorphic with the size and
curved structure of barreloids in the ventral lateral part of
VPM. Thus, it seems reasonable to propose that single
type II axons from the SP5i convey multiwhisker information to a single thalamic barreloid.
Receptive field of VPM cells
The PR5 and SP5i are innervated by the same first-order
vibrissal afferents that branch repeatedly throughout the
trigeminal column (Hayashi, 1980, 1985; Jacquin et al.,
1986b). The PR5 is situated next to the entry of the trigeminal nerve and at ⬇7 mm from the VPM, whereas the barrelette region of the SP5i lies ⬇2.5 mm more caudally. In
addition, single-whisker PR5 neurons are much smaller
than the multiwhisker units of the SP5i (Jacquin et al., 1988,
1989; Veinante and Deschênes, 1999), indicating that they
likely possess a higher input resistance that would reduce
the rise time of synaptic potentials. Assuming a factor of
proportionality of 6 m/second per micrometer of axon diameter (Hursh, 1939; Rushton, 1951), the conduction velocity of
PR5 axons would be about twice that of type II SP5i fibers
(PR5 axons diameter, 2–3 ␮m, as measured from a previous
data base; Veinante and Deschênes, 1999). Together, all of
these factors add up to favor a faster transmission through
the PR5 than the SP5i channel. Thus, these crude anatomic
considerations give full support to the conclusions reached
by Lee et al. (1994b), who provided clear physiological evidence that a late-arriving, multiwhisker input to the VPM
arises from the SP5i.
The above-described anatomic organization is consistent with and, indeed, clarifies previous functional studies
of vibrissa-sensitive cells in rat VPM. On striking a single
vibrissa, relay cells in the corresponding barreloid first
will be activated by the firing of fast-conducting PR5 afferents. Activation will be followed by inhibition arising
from reticular thalamic cells that will prevent the latearriving, multiwhisker input from the SP5i to reach
threshold. The critical point here is how effective is the
inhibition. Strong inhibition will confer on thalamic cells a
single-whisker responsiveness, whereas inhibition that is
not so strong will allow these cells to manifest a multiwhisker sensitivity. Because the inhibition induced by
reticular thalamic cells is strongly state-dependent (Steriade and Deschênes, 1984), it becomes obvious that the
weight of SP5i inputs will increase as the level of anesthesia diminishes (Freidberg et al., 1999). After lesion of
the reticular nucleus, most relay cells will respond to
multiple whiskers (Lee et al., 1994a). Lesion of the PR5,
however, will prevent the early induction of reticular inhibition and confer on relay cells the multiwhisker responsiveness that characterizes SP5i afferents. In contrast, SP5i
lesion will reduce the receptive field to one or two whiskers (e.g., see Rhoades et al., 1987; Freidberg et al., 1999).
Projections from SP5i to Po
The SP5 projection to Po is extensive, with some clusters,
and it is particularly dense in the shell region over the VPM
and in the caudal ventral part of the ventrobasal complex,
where it is difficult to distinguish Po from the nonbarreloid
region of VPM. This latter region demonstrates heavy retrograde labeling after tracer injections into the second somatosensory cortical area (Spreafico et al., 1987). Vibrissal information processed in this cortical area likely transits through
this zone of the posterior thalamus, which receives input
from all subnuclei of the trigeminal complex.
The distinct clusters of terminations in the angular and
ventral lateral nuclei also clearly were present in the
photomicrographs published by Williams et al. (1994; see
Fig. 2). It remains unclear whether these thalamic regions
project to the barrel field and/or to other whisker-related
areas of the neocortex. Because these two regions receive
a robust collateral input from layer 5 cells of the barrel
field, whose axons also project to the deep layers of the
superior colliculus (unpublished observations), they may
be involved in the motor control of whiskers. The same
comment also applies to the shell region over the VPM,
which receives a layer 5 corticothalamic input and contains cells that project to both the somatic sensory and the
motor cortices (Deschênes et al., 1998). Thus, the hodology
of prethalamic and corticothalamic connections in Po suggests that this nucleus may be more involved in the processing of information related to active whisking than to
passive displacements of the vibrissae.
ACKNOWLEDGMENTS
This work was supported by grant MT-5877 from the
Medical Research Council of Canada to M.D., by National
SP5 PROJECTIONS IN THE RAT
Institutes of Health grants DE07662 and DE07734, and
by grants from the UMDNJ Foundation and the American
Osteopathic Association to M.F.J. The authors are grateful to Dr. Chantal Mérette from the Group of Biostatistics,
who carried out the bimodality test on the data. They also
thank Caroline Varga for her helpful technical assistance.
LITERATURE CITED
Bennett-Clarke CA, Chiaia NL, Jacquin MF, Rhoades RW. 1992. Parvalbumin and calbindin immunocytochemistry reveals functionally distinct cell groups and vibrissa-related patterns in the trigeminal brainstem complex of the adult rat. J Comp Neurol 320:323–338.
Bruce LL, McHaffie JG, Stein BE. 1987. The organization of trigeminotectal and trigeminothalamic neurons in rodents: a double-labeling study
with fluorescent dyes. J Comp Neurol 262:315–330.
Chalupa LM, Rhoades RW. 1977. Response of visual, somatosensory and
auditory neurons in the golden hamster’s superior colliculus. J Physiol
(London) 270:595– 626.
Chen S, Aston-Jones G. 1998. Axonal collateral-collateral transport of tract
tracers in brain neurons: false anterograde labelling and useful tool.
Neuroscience 82:1151–1163.
Chiaia NL, Rhoades RW, Bennett-Clarke CA, Fish SE, Killackey HP.
1991a. Thalamic processing of vibrissal information in the rat: I. Afferent input to the medial ventral posterior and posterior nuclei.
J Comp Neurol 314:201–216.
Chiaia NL, Rhoades RW, Fish SE, Killackey HP. 1991b. Thalamic processing of vibrissal information in the rat: II. Morphological and functional
properties of medial ventral posterior nucleus and posterior nucleus
neurons. J Comp Neurol 314:217–236.
Deschênes M, Veinante P, Zhang Z-W. 1998. The organization of corticothalamic pathways: reciprocity versus parity. Brain Res Rev 28:286 –308.
Diamond ME, Armstrong-James M, Ebner FF. 1992. Somatic sensory
responses in the rostral sector of the posterior group (Pom) and in the
ventral posterior medial nucleus (VPM) of the rat thalamus. J Comp
Neurol 318:462– 476.
Erzurumlu RS, Killackey HP. 1980. Diencephalic projections of the subnucleus interpolaris of the brainstem trigeminal complex in the rat.
Neuroscience 5:1891–1901.
Freidberg MH, Lee SM, Ebner FF. 1999. Modulation of receptive field
properties of thalamic somatosensory neurons by the depth of anesthesia. J Neurophysiol 81:2243–2252.
Hayashi H.1980. Distribution of vibrissae afferent fiber collaterals in the
trigeminal nuclei as revealed by intra-axonal injection of horseradish
peroxidase. Brain Res 183:442– 446.
Hayashi H. 1985. Morphology of central terminations of intra-axonally
stained large myelinated primary afferent fibers from facial skin in the
rat. J Comp Neurol 237:195–215.
Horikawa K, Armstrong WE. 1988. A versatile means of intracellular
labeling: injection of biocytin and its detection with avidin conjugates.
J Neurosci Methods 25:1–11.
Hursh JB. 1939. Conduction velocity and diameter of nerve fibers. Am J
Physiol 127:131–139.
Jacquin MF, Rhoades RW. 1990. Structure and response properties of cells
in the trigeminal subnucleus oralis. Somatosens Motor Res 7:265–288
Jacquin MF, Mooney RD, Rhoades RW. 1986a. Morphology, response properties, and collateral projections of trigeminothalamic neurons in
brainstem subnucleus interpolaris of rat. Exp Brain Res 61:457– 468.
Jacquin MF, Woerner D, Szczepanik AM, Riecker V, Mooney RD, Rhoades
RW. 1986b. Structure-function relationships in rat brainstem subnucleus interpolaris. I. Vibrissa primary afferents. J Comp Neurol 243:
266 –279.
Jacquin MF, Golden JP. Panneton WM. 1988. Structure and function of
barrel “precursor” cells in trigeminal nucleus principalis. Dev Brain
Res 43:309 –314.
Jacquin MF, Barcia M, Rhoades RW. 1989. Structure-function relationships in rat brainstem subnucleus interpolaris. IV. Projection neurons.
J Comp Neurol 282:45– 62.
Jacquin MF, Chiaia NL, Haring JH, Rhoades RW. 1990. Intersubnuclear
connections within the rat trigeminal brainstem complex. Somatosens
Motor Res 7:399 – 420.
Land PW, Buffer SA, Yaskoski JD. 1995. Barreloids in adult rat thalamus:
three-dimensional architecture and relationship to somatosensory cortical barrels. J Comp Neurol 355:573–588.
243
Lee SM, Freidberg MH, Ebner FF. 1994a. The role of GABA-mediated
inhibition in the rat ventral posterior medial thalamus. I. Assessment
of receptive field changes following thalamic reticular nucleus lesion.
J Neurophysiol 71:1702–1715.
Lee SM, Freidberg MH, Ebner FF. 1994b. The role of GABA-mediated
inhibition in the rat ventral posterior medial thalamus. II. Differential
effects of GABAa and GABAb receptor antagonists on responses of
VPm neurons. J Neurophysiol 71:1716 –1726.
Lund RD, Webster KE. 1967. Thalamic afferents from the spinal cord and
trigeminal nuclei: an experimental anatomical study in the rat. J Comp
Neurol 130:301–312.
Ma W, Peschanski M, Ohara PT. 1988. Fine structure of the dorsal part of
the nucleus submedius in the rat thalamus: an anatomical study with
reference to possible pain pathways. Neuroscience 26:147–159.
Mantle-St-John LA, Tracey DJ. 1987. Somatosensory nuclei in the brainstem of the rat: independent projections to the thalamus and cerebellum. J Comp Neurol 255:259 –271.
McLachlan G, Basford K. 1988. Mixture models. New York: Dekker.
Nicolelis MAL, Chapin JK, Lin RCS. 1992. Somatotopic maps within the
zona incerta relay parallel GABAergic somatosensory pathways to the
neocortex, superior colliculus and brainstem. Brain Res 577:134 –141.
Olfert ED, Cross BM, McWilliams AA. 1993. Guide to the care and use of
experimental animals. Bradda: Canadian Council on Animal Care.
Paxinos G, Watson C. 1986. The rat brain in stereotaxic coordinates, 2nd
ed. Sydney: Academic Press.
Peschanski M. 1984. Trigeminal afferents to the diencephalon in the rat.
Neurosci 12:465– 487.
Phelan KD, Falls WM. 1991. A comparison of the distribution and morphology of thalamic, cerebellar and spinal projection neurons in rat
trigeminal nucleus interpolaris. Neuroscience 40:497–511.
Pinault D. 1996. A novel single-cell staining procedure performed in vivo
under electrophysiological control: morpho-functional features of juxtacellularly labeled thalamic cells and other central neurons with biocytin or neurobiotin. J Neurosci Methods 65:113–136.
Renehan WE, Jacquin MF, Mooney RD, Rhoades RW. 1986. Structurefunction relationships in rat medullary and cervical dorsal horns. II.
Medullary dorsal horn cells. J Neurophysiol 55:1187–1201.
Rhoades RW, Belford GR, Killackey HP. 1987. Receptive field properties of
rat ventral posterior medial neurons before and after selective kainic
acid lesions of the brainstem trigeminal complex. J Neurophysiol 57:
1577–1600.
Rushton WAH. 1951. A theory of the effect of fibre size in medullated
nerve. J Physiol (London) 115:101–122.
Silverman JD, Kruger L. 1985. Projections of the rat trigeminal sensory
nuclear complex demonstrated by multiple fluorescent dye retrograde
transport. Brain Res 361:383–388.
Smith RL. 1975. The ascending fiber projections from the principal sensory
trigeminal nucleus in the rat. J Comp Neurol 148:423– 446.
Spreafico R, Barbaresi P, Weinberg RJ, Rustioni A. 1987. SII-projecting
neurons in the rat thalamus: single- and double-retrograde tracing
study. Somatosens Res 4:359 –375.
Steriade M, Deschênes M. 1984. The thalamus as a neuronal oscillator.
Brain Res Rev 8:1– 63.
Tiao Y-C, Blakemore C. 1976. Functional organization in the superior
colliculus of the golden hamster. J Comp Neurol 168:483–504.
Veinante P, Deschênes M. 1999. Single- and multi-whisker channels in the
ascending projections from the principal trigeminal nucleus in the rat.
J Neurosci 19:5085–5095.
Wang B-R, Ohara PT. 1993. Convergent projections of trigeminal afferents
from the principal nucleus and subnucleus interpolaris upon rat ventral posteromedial thalamic neurons. Brain Res 629:253–259.
Williams MN, Zahm DS, Jacquin MF. 1994. Differential foci and synaptic
organization of the principal and spinal projections to the thalamus in
rats. Eur J Neurosci 6:429 – 453.
Wong-Riley MTT. 1979. Changes in the visual system of monocularly
sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry. Brain Res 171:11–28.
Woolston DC, LaLonde JR, Gilson JM. 1982. Comparison of response
properties of cerebellar- and thalamic-projecting interpolaris neurons.
J Neurophysiol 48:160 –173.
Yamasaki DSG, Krauthamer GM. 1990. Somatosensory neurons projecting
from the superior colliculus to the intralaminar thalamus in the rat.
Brain Res 523:188 –194.