JOURNALOF NEUROPHYSlOLoGY
Vol. 54, No. 5, November
1985. Prinrd
Somatotopic Organization in Cat Spinal Cord
Segments With Fused Dorsal Horns:
Caudal and Thoracic Levels
LOUIS
A. RITZ,
JAMES
L. CULBERSON,
AND
PAUL
B. BROWN
SUMMARY
AND
CONCLUSIONS
horn cell somatotopy, assuming that the presynaptic terminals’ somatotopy is in register
with that of dorsal horn cells (the presynaptic
somatotopy hypothesis; seeRef. 12).
1. We have explored the somatotopic organization of the two cat spinal cord regions
where the dorsal horns are fused (i.e., continuous across the midline): the caudal and INTRODUCTION
thoracic segments.We have mapped the lowthreshold component of dorsal horn cell reThe studies reported here are concerned
ceptive fields (RFs) in these segmentsand have with the somatotopic organization of cat dorsal
charted the locations of dorsal root low- horn cells at levels where the dorsal horns are
threshold mechanoreceptive dermatomes. We fused (continuous) acrossthe midline, without
also have determined the projections of caudal interruption by the dorsal columns. This ocand thoracic dorsal roots to laminae III and curs at caudal and thoracic segments(50). Our
objectives were to determine 2) whether the
IV by using degeneration techniques.
2. The dorsal skin of the tail or thorax is somatotopy of these segmentsis similar to that
represented laterally, and ventral skin is rep- of brachial and lumbosacral dorsal horn, and
resented at the midline, in the fused dorsal 2) whether the somatotopy of low-threshold
horns. Many caudal and thoracic dorsal horn
cutaneous mechanoreceptor projections to
units had RFs that crossedthe dorsal or ventral
laminae III and IV is in register with the dorsal
midline of the skin; these units were encoun- horn cell somatotopy (the presynaptic somatered near the edges or the midline, respec- totopy hypothesis: seeRef, 12).
tively, of the fused dorsal horns.
Across cat lumbosacral dorsal horn, Wall
3, The tail is fully representedwithin dorsal (62,63) found a mediolateral gradient of cells
root dermatomes S3to CaS.Roots more caudal responding to inputs from hindlimb skin; disthan Ca5 represent progressively smaller skin tal hindlimb is represented medially, and
areas of the distal tail. Adjacent dermatomes proximal hindlimb is represented laterally in
overlapped 15-65%. Thoracic dermatomes the dorsal horn. Byran et al. (14) demonhad a nearly vertical orientation; adjacent strated, for hindlimb spinocervical tract (SCT)
dermatomes overlapped by 30-75%.
neurons, that embryologically ventral skin is
4. Dorsal roots in caudal and thoracic re- represented medially in the dorsal horn, and
gions have crossedprojections to the medial embryologically dorsal skin is represented latand lateral (but not middle) portions of the erally. Brown and Fuchs (12) extended these
contralateral dorsal horn. These crossed pro- observations by describing the somatotopic
jections are a possibleanatomical substrate for organization of cat lumbosacral dorsal horn
RFs that cross the ventral or dorsal midline.
laminae I-VI in detail. They observed that 1)
5. The dorsal root projection patterns are there is somatotopic organization in the horizontal plane; 2) segmental representation
consistent with those that would be predicted
from the dorsal root dermatomes and dorsal within the dorsal horn closely resemblesthe
0022-3077/U
$1.50 Copyright 0 1985 The American
Physiological Society
1167
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
DepartmentsqfPhysi&gy and Anatomy, West Virginia University Medical Center,
Morgantown, West Virginia 26506
1168
RITZ,
CULBERSON,
BROWN
have a portion of their RFs on the contralateral
side of the body.
Based on these considerations
and on observations by others (e.g., Refs. 43, 57, 61) we
expected that at caudal and thoracic levels 1)
dorsal skin is represented laterally and ventral
skin medially in the dorsal horn; 2) for RF
continuity of dorsal horn representation of the
skin across the ventral midline of the tail or
thorax, dorsal root projections should terminate in the medial portion of the contralateral
dorsal horn (dorsal gray commissure);
3) for
dorsal horn RF continuity across the dorsal
midline of the tail or thorax, the dorsal roots
should terminate in the lateral portion of the
contralateral dorsal horn; 4) the trajectory of
segmental representation
within the dorsal
horn should parallel the trajectory of corresponding dorsal root dermatomes. We have
found these expectations to be correct. Preliminary observations
have been reported
elsewhere (52).
METHODS
Adult cats (2-5 kg), either male or female, were
used for these experiments. For electrophysiological
recording experiments, animals were anesthetized
with halothane, the trachea was intubated, and artificial respiration was begun. A carotid artery and
a jugular vein were cannulated to monitor blood
pressure and to infuse solutions, respectively. Halothane was discontinued, and the animal was either
decerebrated at a precollicular
level with a blunt
transection or infused with a-chloralose (70 mg/kg,
supplemented as required to maintain areflexia to
noxious stimulation), Similar results were obtained
with both preparations (4 decerebrate, 6 anesthetized). Rectal temperature and expired CO2 were
monitored and maintained within normal limits.
The urethra was cannulated to allow the bladder to
drain. The animal’s head and hips were securely
fixed in a Kopf animal frame. A laminectomy extending approximately
from T4 to T12 or from L5
to Gas exposed the relevant cord segments. A pool
of mineral oil covered the cord; the dura was cut,
and in the caudal cord the arachnoid was stripped
away, A vertebral clamp, bilateral pneumothorax,
and spinal cord platform were used to increase mechanical stability.
Single-unit discharges were recorded with stainless steel (23) or carbon fiber (1) microelectrodes.
At each rostrocaudal level studied (segments T,,
T9, TIO, Sz , S3, Cal , Ca2, Ca& successive dorsoventral electrode penetrations were spaced in ~200pm steps, beginning at the left dorsal root entry
zone, across the midline, and continuing to the right
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
dorsal root dermatomes;
3) a mediolateral
trajectory across the dorsal horn maps both a
distoproximal and a ventrodorsal progression
of receptive fields (RFs) across the hindlimb;
4) in L6, L7, and SI the foot representation
in the lumbosacral dorsal horn is disproportionately large relative to the more proximal
leg, which is represented further laterally in
the dorsal horn. The somatotopic organization
of the cat brachial dorsal horn is similar to
that of the lumbosacral dorsal horn (29). Subsequent studies have confirmed these general
findings (2, 5, 35).
As a first approximation
for describing the
assembly of dorsal horn RFs, Brown and
Fuchs (12) suggested the presynaptic somatotopy hypothesis: If primary afferent terminals are somatotopically arranged, then dorsal
horn RFs are determined by that portion of
primary afferent neuropil monosynaptically
connected to the dorsal horn neurons’ dendrites. For the lumbosacral cord the presynaptic somatotopy hypothesis is supported by
observations of the projections of dorsal roots
( 1 I), whole cutaneous nerves (30, 3 l), and
single primary afferent fibers (A. G. Brown,
P. B. Brown, R, E. W. Fyffe, and L. M. Pubols,
unpublished observations). These three sets of
projections terminate in laminae III-IV in a
pattern consistent with the hindlimb map of
Brown and Fuchs (12), with the exception of
a caudal spread of projections from branches
of the posterior femoral cutaneous nerve (3 1).
Brown and Noble (7), in a study of the connections between hair follicle afferent fibers
and SCT cells, and Brown et al. (IO), in a study
of the overlap of dendritic trees of adjacent
SCT neurons, provide further indirect evidence to support the presynaptic somatotopy
hypothesis. Note that Wall and colleagues (e.g.,
Refs. 19,64) and Meyers and Snow (e.g., Refs.
40,4 1) have reported afferent projections that
they conclude are more extensive rostrocaudally than would be predicted from the presynaptic somatotopy hypothesis.
Contralateral dorsal root projections have
been observed at several levels of spinal cord,
including sacrocaudal ( 17,37-39,58,59),
high
cervical ( 17, 18), and thoracic (55, 56) levels.
It is known that cat dorsal root dermatomes
end sharply at dorsal and ventral midlines (e.g.,
Refs. 13, 32). If the presynaptic somatotopy
hypothesis is correct, these crossed projections
suggest that some dorsal horn neurons may
AND
SOMATOTOPY
OF FUSED DORSAL
HORNS
1169
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
dorsal root entry zone, adjusting mediolateral adjacentsectionsat 0.12 mm intervalswerestained,
placementsto avoid blood vessels.For each unit either with the Fink-Heimer technique (20) to
encountered,the low-thresholdexcitatory compo- demonstratedegeneratedprimary afferent axons
nent of the RF wasoutlined on the skin and repro- and terminalsor with cresylviolet. Computer-genduced on a figurine drawing. The innocuous,cu- erateddorsal-viewmapswerepreparedon an LM2
taneousmechanoreceptiveRFs were determined computer(28) andwereusedto illustratethe extent
with brushesand with nonserratedforceps(innoc- of terminaldegenerationwithin laminaeIII and IV.
uouspressure);thesestimuli wereconsideredadeOur interpretationsof the degenerationdata are
quate for Acu,P-fibers.It is recognizedthat smaller basedon two assumptions:I) As a first approxiafferent fibers(e.g., Ref. 15) may alsobe activated mation we assumethat primary afferent input to
by suchstimuli.
laminaeIII and IV (e.g., Refs.4, 6, 8, 9, 45) is the
Dermatomesof caudaldorsalroots (4 animals; primary determinant of dorsal horn neuron lowsegments&Ca,) were mappedby recordingwith thresholdRFs(e.g., Refs.3, 7, 10).It isknown that
hook electrodesplaced under the roots. Derma- somelow-threshold mechanoreceptorsproject to
tomesfor thoracicdorsalrootscould not beassessedlamina II or V (e.g.,Refs,8, 37); their contribution
in thismannerbecause
the dorsalrootsaretoo short. to dorsalhorn neuron RFs isunknown. 2) We also
Therefore, thoracic dermatomes(3 animals;seg- assumethat the majority of primary afferent proments T3, T5 through TlO) were determined by jections to laminaeIII and IV emanatefrom lowmakingmultiple penetrationsthroughoutthe dorsal thresholdmechanoreceptors
(3). It isrecognizedthat
root gangliawith low-impedancemetal microelec- there may be someinput to sacrocaudallaminae
trodesandmappingthe collectiveinnervationfields III and IV from muscleafferent fibers,asthere is
obtained with multiunit recordings.Dermatomes for the lumbar enlargement(3, 27).
weretracedon the skinand werechartedon figurine
Note that vertebraeof the tail in catsare referred
drawings. Low-threshold cutaneous mechanore- to ascaudal vertebrae(47); accordingly, segments
ceptive input wasmapped,usingmanually applied caudalto S3arereferredto ascaudal,not coccygeal,
low-thresholdtactile stimuli.
segments.
After eachrecordingexperimentthe animal was
perfusedthroughthe heartwith 0,9%NaCl solution
followed by 10%formalin. After identifying spinal RESULTS
segmentsfrom which single-unitrecordingswere Receptive-field organization
made,cord segments
wereremovedand embedded
Electrophysiological
recordings were obin paraffin. Serial 25-pm sectionswere stainedfor
tained
from
102
single
units in caudal and
Nisslsubstanceand examinedunder dark field illuminationfor electrodetracksand for microlesions thoracic dorsal horns. Twenty units were localized, with lesions, to lamina III, IV, or V
placedat someof the recordingsites,
Drawingsof the RFsweretraced with a graphics of the dorsal horn. The remaining unmarked
digitizer and wereanalyzedon a PDP- 11computer units were recorded at microdrive depths conto examineRF geometry.The Spearmanrank cor- sistent with dorsal horn locations, principally
relation coefficient (rs: Ref. 54) wasusedto deter- lamina III, IV, or V. Note that RF data were
mine the correlation betweenRF area or length/ collected from several laminae because prewidth ratio and the dorsoventral position on the vious results (e.g., Refs. 12,3 1,35) suggest the
skin of the RF center. Testsfor the significanceof
neutral axis in
rs usedthe Student’st distribution (54). All tests existence of a somatotopically
the radial, or roughly dorsoventral, dimension.
weretwo tailed.
Catsusedfor degenerationstudieswereanesthe- That is, laminae I-VI of lumbosacral dorsal
tized by intraperitonealinjection of pentobarbital horn are in register with locations of the lowsodium(40 mg/kg),Hair overlying sacrocaudal
and/ threshold RF centers of dorsal horn cells.
or thoracic vertebraewasremoved,and under anIn our recording experiments, microelectiseptic conditions sacrocaudaland/or thoracic trode penetrations were placed, in sequence,
laminectomieswere performed, In three animals, from the left dorsal root entry zone to the right
one thoracic and one contralateral caudal dorsal dorsal root entry zone; this constitutes a reroot wereseveredcentralto the dorsalroot ganglion. cording plane. The entire width of the fused
In three other animals,a singlecaudaldorsalroot
was cut. After 5-7 days survival, animals were dorsal horns was sampled. In all recording
anesthetized with pentobarbital and perfused planes ( 16 caudal and 7 thoracic) RF centers
transcardiallywith a 0.9%salinewashfollowedby were obtained that shifted from left dorsal skin
10%formalin. Appropriate cord segments
werere- down the left side of the tail ur thorax to venmoved and stored in formalin for severalweeks. tral skin and up the right side to right dorsal
Serial 40-pm-thick frozen sectionswere cut, and skin of the tail or thorax.
1170
RITZ,
CULBERSON,
C&C
Ca 3.C
HG. 1. Dorsal-view
map of receptive fields within 2
caudal segments, obtained from 4 recording
planes from
1 animal. For all figures see text for details.
Figures 1 (caudal) and 2 (thoracic) use dorSal-view maps of the fused dorsal horns to illustrate the RFs obtained from single-unit re-
BROWN
cordings within a recording plane. These maps
were generated from units whose mediolateral
location could be accurately determined. The
segments are labeled in normalized form (i.e.,
Cal 0 is the rostra1 border of Ca2; T9 7 is the
point at which 70% of the T9 segment is rostra1
to it), with normalized dorsal horn widths. The
lateral borders of the fused dorsal horns are
marked by solid borders, and the midline by
a dashed line. The figurines in Fig. 1 are dorsal
views of the tail skin drawn as though it had
been unwrapped by cutting it along the dorsal
midline and laying it flat, epidermis down.
Thus, the RFs are seen as if the tail skin were
placed on a light box, and the RF boundaries
were viewed through the skin from the dermal
side. The ventral midline of the skin is represented by the dashed line of the figurine, and
the dorsal midline is represented as solid lines
at the edges of the figurine. In Fig. 2 the dorsal
and ventral borders of the figurines represent
the midlines of the thorax. In both figures, figurines are placed so that RF centers fall at the
recording sites.
In Figs. 1 and 2 the centers of RFs shift
from the left dorsal midline, around the ventral
T9.7
FIG. 2. Dorsal-view
map of receptive fields obtained
from 1 animal and the 2 TIO planes from another.
from
3 thoracic
recording
planes from
2 animals,
the Tg plane
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
Ca4.0-
AND
SOMATOTOPY
OF
FUSED
HORNS
1171
the ventral midline, and 24 units had RFs that
crossed the ventral midline. In contrast, 33
units had RFs that neither reached nor crossed
either midline.
Spearman rank correlation analysis revealed
that there was no significant correlation between RF area (rs = -0.25) or RF length/width
ratio (us = 0.18) and RF ventrodorsal location
on the tail, nor any between RF area (Ye =
-0.03) or RF length/width
ratio (us = -0.21)
and ventrodorsal RF location on the thorax.
Dorsal rool dermatomes
Dorsal root dermatomes from caudal segments were obtained in four experiments (Fig.
3). The proximal part of the tail is represented
in S3, and the distal portion of the tail in Ca5.
Dorsal roots caudal to CaS represented successively less of the distal portion of the tail.
Adjacent caudal dermatomes overlapped by
- 15-65% depending on the segments and
animal. The dorsal and ventral borders of the
CO5
>
Ca5
Co4
FIG.
3.
Drawings
of sacrocaudal
dorsal root dermatomes
from
0
4 animals.
Co6
U
Co7
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
surface to the right dorsal midline, as electrode
placement is moved from the left edge of the
fused dorsal horns, across the midline, to the
right edge. Of our 102 units, 99 follow this
general pattern. The three units that deviated
from this pattern were found at midline positions in caudal segments and had dorsal RFs.
In the seven recording planes obtained from
thoracic segments, all units in a plan had RF
centers at the same rostrocaudal body level;
i.e., the ventral RF centers were at the same
rostrocaudal position as the dorsal RF centers,
The same organization of RF centers was observed for 14 of 16 recording planes obtained
from caudal segments. In the other two caudal
places, ventral RF centers were slightly caudal
to the dorsal RF centers (not illustrated).
Many units had RFs that reached or crossed
the dorsal or ventral midline. Out of 102 units,
14 had RFs that reached the dorsal midline,
and 18 units had RFs that crossed the dorsal
midline. Thirteen units had RFs that reached
DORSAL
1172
RITZ,
CULBERSON,
T5
aT8,
FIG. 4, Drawings
of midthoracic
dorsal root dermatomes from 3 animals.
In the upper illustration
the T6
dermatome
was not determined.
BROWN
cause they fell within the surgical exposure,
ventral borders always reached the ventral
midline, but did not cross it.
Degeneration sludies
Individual thoracic and caudal dorsal roots
were sectioned, and the lesioned segment and
one to two segments rostra1 and caudal to the
lesioned root were examined for degenerating
axons and terminals. Useful degeneration data
were obtained in three caudal and three thoracic cases.
Dorsal-view
maps of dorsal root degeneration, as visualized using the Fink-Heimer
technique, are shown in Figs. 5 and 6. These
figures have the following format: The top of
each plot indicates which dorsal root was severed, and the labels at the left of the figure
indicate the rostra1 borders of spinal cord segments (e.g., Cazmois the rostra1 border of Caz).
The boxed area for each case identifies the
portion of the spinal cord that was examined
in Fink-Heimer-impregnated
material. The
dashed line, in the center of a box, indicates
the midline of the fused dorsal horns; the left
and right edges of the rectangle are the left and
right edges of the fused dorsal horns. The sections where degeneration was observed in
laminae III and IV are plotted at their fractional rostrocaudal levels within the segments,
and the left-right locations of the line segments
indicate the relative mediolateral locations of
the degeneration. The mediolateral widths and
rostrocaudal lengths of the segments are normalized. The absence of lines in the middle
of a degeneration pattern may mean either an
absence of degeneration or a gap in the sample.
Note that density of degeneration is not depicted on the dorsal-view maps, All figures are
plotted as though the left dorsal root had been
cut (i.e., if the right root was cut, the figure
has been left/right reversed).
In a few areas where abundant debris was
present, when it was not possible to discriminate between degenerating axons and zones
of likely afferent terminals, both are shown in
some figures. It was clear in the histological
material that most or all of the debris crossing
through the central one-third to one-half of
the contralateral dorsal horn was relatively
coarse, corresponding
to fibers of passage;
these are axons in small bundles passing toward their lateral target zones (17). No unequivocal terminal degeneration was ever seen
in this central area.
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
dorsal root dermatomes w‘ere very sharp, coinciding with dorsal and ventral midlines.
Rostra1 and caudal borders were less distinct,
but in most cases appeared to be tapered rather
than vertical, so that they extended further
rostrocaudally on the sides of the tail than on
the dorsal or ventral border. Note from Fig. 3
that there are a few areas of skin not represented by dorsal root dermatomes; our technique apparently led to some underrepresentation of the dermatomes,
Mid thoracic dorsal root dermatomes, obtained from three ani mals, had a nearly vertical orientation and overlapped adjacent dermatomes by ~30-75%
(Fig. 4). The dashed
lines of the dermatomes in the bottom figurine
represent the presumed dorsal extent of the
dermatomes; this could not be accurately assessed in this animal because of surgical damage to the dorsal rami. Although dorsal borders
sometimes could not be fully determined be-
AND
SOMATOTOPY
OF
FUSED
Cd
Ca2
Ca4
Ca3.0
Ca4.0
c 05.0
C a6.0
Cu.0
FIG, 5. Dorsal-view
maps illustrating
generation in laminae III and XV following
roots Ca, , Ca2, and Cad.
location
of delesions of dorsal
HORNS
1173
T4.0
T5.0
T6.0
T-Z0
T8.0
T9.0
TlO.O
Ttl.0
T12.0
T13.0
LI.0
L2.0
L3.0
FIG. 6.
Dorsal-view
maps displaying
generation
in laminae III and IV following
roots T6, TI1, and T1+
location of delesions of dorsal
DISCUSSION
Somatotupic organization of
dursul horn cells
Our unit recordings from caudal and thoracic dorsal horn demonstrate that dorsal skin
projects laterally and ventral skin projects medially in the cat sacrocaudal and thoracic dorsal horns. These results are consistent with data
obtained from the hindlimb (12, 62, 63), brachial (29), and thoracic (43) levels of the cat
spinal cord. Our results also agree with singleunit studies from thoracic cord of the rat (57).
Sacrocaudul and thoracic dermatomes
Our mapping of caudal and thoracic dermatomes agrees in principle with the few prior
studies at these cord levels. At caudal levels
Reid (46) found that the proximal tail is supplied by S3 and that the remaining caudal
dorsal roots distribute in overlapping bands to
sequentially more distal tail skin, until the tip
is innervated by the fifth caudal dorsal root.
A similar arrangement is seen in the spider
monkey (44), except that only the caudal-most
dorsal root innervates the tip of the tail, and
in the dog, except that S3 is shifted slightly
more toward the perineum and supplies a
smaller part of the proximal tail (2 1). In the
cat, dorsal roots S2 and S1 include successively
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
The following generalizations can be made,
based on the degeneration data used to generate Figs. 5 and 6: 1) Degenerated fibers are
found in laminae III and IV rostra1 to the lesioned dorsal root for at least two segments,
whereas the degeneration begins to disappear
caudally within the second segment (not fully
sampled in some of the caudal segments). 2)
The full width of the ipsilateral dorsal horn is
filled in the entry segment; this tapers in segments rostra1 and caudal to the entry segment.
3) The tapering rostra1 and caudal portions of
the degeneration display no consistent medial
or lateral shift in location. 4) Degeneration is
observed in the medial and lateral parts of the
contralateral dorsal horn, at the level of the
lesioned dorsal root and sometimes in adjacent
segments; unequivocal terminal degeneration
is not seen in the middle region of the contralateral dorsal horn. 5) Contralateral degeneration is much less dense than ipsilateral degeneration (not discernible in the dorsal-view
maps) and has much less rostrocaudal extent.
6) The absolute longitudinal extent of the degeneration (not discernible in the normalized
figures) within the segments sampled, ranged
from ~40 to 75 mm for thoracic regions and
from 14 to 19 mm for caudal segments. In
some cases the degeneration appeared to extend beyond the segments sampled.
DORSAL
1174
RITZ,
CULBERSON,
AND
BROWN
ganglionic transport of horseradish peroxidase,
Smith (56) found that dorsal rami project to
the lateral portion of the contralateral dorsal
horn and that ventral rami project to the medial portion of the contralateral dorsal horn.
In the rat (24), dorsolateral tail afferent fibers
project laterally in the dorsal horn, and ventrolateral tail afferent fibers project medially
in the dorsal horn. These data from the rat
certainly concur with our observations
on
crossed and uncrossed dorsal root projections.
Dorsul rsot project ions
With respect to the anatomy of dorsal root
afferent projections, there is little need to address ipsilateral fiber distributions;
it is well
known that dorsal horn laminae III and IV
receive dense input to their entire mediolateral
extent in the entry segment for dorsal root fibers (e.g., Refs. 1 1, 16, 26, 57). Crossed projections, on the other hand, offer some clues
to the possible somatotopy of cells responding
to stimulation of the tail or thorax. Crossing
afferent fibers at sacrocaudal levels are extensive and have been studied at several levels of
resolution. Dorsal roots generate crossed input
as shown by our previous work (17) and by
others (36,49). The crossed projection patterns
of individually stained primary afferent fibers
have been demonstrated (37, 39, 5 1). Many
contralaterally coursing fibers are of visceral
origin and pass principally to the dorsal gray
commissure (42). Our studies (present work,
and Ref. 17) and especially those of Matsushita
and Tanami (38) establish the predominant
targets of the majority of crossed afferent fibers
as being medial and lateral dorsal horn areas,
especially in laminae III and IV, Earlier results
( 17, 38) and those reported here (Figs. 5 and
6) establish a possible structural substrate for
primary afferent innervation of cells in medial
or lateral dorsal horn that have RFs extending
across the ventral or dorsal midlines, respectively, of the skin.
Lacking prior data on cat thoracic levels,
we note that in rat thoracic dorsal horn it has
been found, by use of degeneration of thoracic
spinal nerves (22) and transganglionic
transport of horseradish peroxidase within thoracic
spinal nerves (65), that dorsal rami of spinal
nerves (which innervate skin up to the dorsal
midline) project laterally, and ventral rami of
spinal nerves (which innervate down to the
ventral midline) project medially in the ipsilateral dorsal horn. Also in rat, by use of trans-
Sumatot0pic appropriateness of
dorsal root proJections
The dorsal root projections of the sacrocauda1 and thoracic cord are somatotopically
appropriate, as dorsal horn RFs overlapping a
root’s dermatome were found in all projection
regions for that root, and a dorsal root projects
to all regions where RFs of dorsal horn cells
overlap the dermatome. This is not true for
hindlimb projections of branches of the posterior femoral cutaneous nerve (3 l), where
projections in lamina V and ventral lamina
IV reach more caudally than would be expected from the low-threshold
component of
dorsal horn RFs. However, these seemingly
inappropriate projections could be to sacrocaudal lamina V multireceptive
neurons,
which can have very large RFs (53); these
neurons have dendrites that penetrate laminae
III and IV, and in order to assemble such large
RFs, caudal extensions of posterior femoral
cutaneous nerve projection zones would be
somatotopically
appropriate.
Crossed projections are rare in segments
whose dermatomes
rarely reach the dorsal
midline and never reach the ventral midline,
e.g., Lb and L7 (17, 33). The absence of crossed
projections is somatotopically
appropriate in
that such crossed projections would result in
discontinuous
bilateral RFs, which have not
been observed. The lack of terminal degeneration in the middle part of the contralateral
dorsal horn is also somatotopically
appropriate, given the lack of dorsal horn cells with
contralateral RF components on the side of
the tail or trunk.
It is interesting to note that the levels at
which the ipsilateral dorsal horn is filled from
the edge to the midline correspond closely to
the levels at which there are crossed projections
just across the midline and to the contralateral
edge. In contrast, the tapered rostra1 and cauda1 extensions of the ipsilateral projections oc-
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
smaller skin areas of the proximal tail and
progressively larger skin areas of perineum and
hindlimb (13). The larger dermatomes illustrated by Kuhn (32) for cat also suggest that
SI may contribute to innervation of proximal
tail, as do the results of Brown and Koerber
(I 3). At thoracic cord levels the only difference
in our observations
and those of Hekmatpanah (25), who also described cat thoracic
dermatomes, is that he reported a slight rostrodorsal-caudoventral
tilt to the dermatomes.
SOMATOTOPY
QF
FUSED
7
Medial and luferul shifts of
dorsal root projections
A number of investigators (26, 34, 48, 49,
59,60) have reported that there is a systematic
medial or lateral shift in the location of dorsal
root projections within the dorsal horn rostra1
or caudal to the root entry segment. We have
observed lateral shifts of dorsal root C111. cutaneous nerve (30, 3 1), and single ‘aff&ent
(A. G. Brown, P. B. Brown, R. E. W. Fyffe,
and L. M. Pubols, unpublished observations)
HORNS
1175
projections as they course rostra1 or caudal into
the area of enlarged representations of the distal hindlimb (12). Imai and Kusama (26) reported a similar pattern for the forelimb enlargement. Most, if not all, of these observations can be adequately accounted for by the
som .atotopy of the enlargement regions (12,
29) . That is, ascending or descending projections of roots with relatively proximal dermatomes can be expected to shift laterally
when entering the segments where the medial
representation of a foot is expanded; ascending
or descending projections of roots with relatively proximal dermatomes can be expected
to expand m edially when leaving the foot representation.
The lack of a medial or lateral shift of caudal
or thoracic dorsal root projections rostra1 or
caudal to their parent segments, although
contradi cting somk earlier reports, 1s consistent
with the absence, in most cases, of a significant
tilt of the caudal or thoracic dorsal root dermatomes relative to the dorsal horn segmental
representations. Such a relative tilt could yield
medial or lateral shifts rostrally or caudally,
as we suggested in an earlier publication (3 1).
ACKN0WLEDGMENTS
This research was supported
by United States Public
Health Service (USPHS) Grant NS- 1206 1. The LM2 computer was provided
under USPHS Grant RR-00374.
Present address of L. A. Ritz: Dept.
Box 5265, JHMHC,
Univ. of Florida,
32610.
of Neurosurgery,
Gainesville,
FL
Received
May 1985.
in final form
11 October
1984; accepted
29
REFERENCES
1. ARMSTRONG-JAMES,
M. ANDMILLAR,
J.Carbonfibre
microelectrodes.
J. Neurosci. Melh. 1: 279-287,
1979.
2. BRENOWITZ,
G. L. AND PUBOLS, L. M. Increased receptive field size of dorsal horn neurons following
chronic spinal cord hemisections
in cats. Brain Rex
216: 45-59, 1981.
3. BROWN, A. G. Organization
in the Spinal Cord. Berlin:
Springer-Verlag,
198 1.
4. BROWN,
A. G., FYFFE, R. E. W., AND NOBLE, R.
Projections
from Pacinian
corpuscles
and rapidly
adapting
mechanoreceptors
of glabrous skin to the
cat’s spinal cord. J. Ph ysiol. London
307 : 38 5-400,
1980.
5. BROWN, A. G., FYFFE, R. E. W., NOBLE, R., ROSE,
P. K., AND SNOW, P. J. The density, distribution
and
topographical
organization
of spinocervical
tract neurones in the cat. J. Physiol. London 300: 409-428,
1980.
6. BROWN,
A. G., FYFFE, R, E. W., ROSE, P. K., AND
7.
8.
9.
10.
SNOW, P. J. Spinal cord collaterals from axons of type
II slowly adapting units in the cat. J. Physiol. London
316: 469-480,
1981.
BROWN, A. G. AND NOBLE, R. Connexions
between
hair follicle afferent fibres and spinocervical
tract neurones in the cat: the synthesis of receptive
fields. J.
Physiol. London 323: 77-91, 1982.
BROWN, A. G., ROSE, P. K,, AND SNOW, P. J. The
morphology
of hair follicle afferent fibre collaterals in
the spinal cord of the cat. J. Physiol. Londm
272:
779-797,
1977.
BROWN, A. G., ROSE, P. K., AND SNOW, P. J. Morphology and organization
of axon collaterals from afferent fibres of slowly adapting
type I in cat spinal
cord. J. Physiol. London 277: 15-27, 1978.
BROWN, A. G., ROSE, P. K., AND SNOW, P. J. Dendritic trees and cutaneous receptive fields of adjacent
spinocervical
tract neurones
in the cat. J. Physiol.
London 300: 429-440,
1980.
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
cur at levels where there is no crossed projection. This is somatotopically
appropriate for
caudal segments, given that the dermatomes
are longer rostrocaudally
on the sides of the
tail than on the dorsal or ventral border. However, the thoracic dermatomes are bandlike,
not tapered, and the segmental representations
appear to be similarly shaped (Fig. 2). This
would suggest that projections
of thoracic
dorsal roots should not be tapered rostrally or
caudally; rather, they should end relatively
abruptly spanning most of the ipsilateral dorsal horn at either end. The observed tapered
projections suggest that 1) our dermatomes are
inadequately sampled by our microel ectrode
recordings from the ganglia, and the dermatomes are actually tapered like the caudal dermatomes; or 2) the segmental representations
of dorsal and ventral trunk skin are less extensive rostrocaudally than for skin on the side
of the trunk; or 3) there is less complete
monosynaptic
convergence for dorsal and
ventral skin than for skin on the side of the
trunk.
DORSAL
1176
RITZ,
CULBERSON,
BROWN
29. KOERBER,
H. R. Somatotopic
organization
of cat
brachial spinal cord. Exp. Neural. 69: 48 l-492, 1980.
30. KOERBER,
H. R. AND BROWN,
P. B. Projections
of
two hindlimb
cutaneous nerves to cat dorsal horn. J.
Neurophysiol.
44: 259-269,
1980.
31. KOERBER,
H. R. AND BROWN,
P. B. Somatotopic
organization
of hindlimb
cutaneous
nerve projections
to cat dorsal horn, J. Neurophysiol.
48: 48 l-489, 1982.
32. KUHN,
R. A. Organization
of tactile dermatomes
in
cat and monkey. J. Neurophysiol.
16: 169-182,
1953.
33. KUKULINSKY,
D. H. AND BROWN,
P. B. Cat L4-Sl
dermatomes
determined
using signal averaging. Neurosci. Lett. 13: 79-82, 1979.
34. LAMOTTE,
C. Distribution
of the tract of Lissauer and
the dorsal root fibers in the primate spinal cord. J.
Cump. Neural. 172: 529-562,
1977.
35. LIGHT, A. R. AND DURKOVIC,
R, G. Features of laminar and somatotopic
organization
of lumbar spinal
cord units receiving cutaneous inputs from hindlimb
receptive fields. J. Neurophysiol.
52: 449-458,
1984.
36. LIGHT, A. R. AND PERL, E, R. Reexamination
of the
dorsal root projection
to the spinal dorsal horn including observations
on the termination
of coarse and
fine fibers. J. Comp. Neural.
186: 117- 13 1, 1979.
37. LIGHT, A. R. AND PERL, E. R. Spinal termination
of
functionally
identified primary
afferent neurons with
slowly conducting
myelinated
fibers. J. Comp. Newt,/.
186: 133-150,
1979.
38. MATSUSHITA,
M. AND TANAMI,
T. Contralateral
termination
of primary
afferent axons in the sacral and
caudal segments of the cat, as studied by anterograde
transport
of horseradish
peroxidase. J. Camp. Neural.
220: 206-218,
1983.
39. MENSE,
S., LIGHT,
A. R., AND PERL, E. R. Spinal
terminations
of subcutaneous
high-threshold
mechanoreceptors.
In: Spinal Cord Sensation,
edited by
A. G. Brown and M. Rethelyi.
Edinburgh:
Scottish
Academic,
198 1, p. 79-86.
40 MEYERS, D. E. R. AND SNOW, P. J. Anatomical
demonstration
of somatotopicall
y inappropriate
projections of single hair follicle afferent fibres to the cat
spinal cord (Abstract).
Proc. Int. Congr. Physiol. Sci.
29 Sydney 1983, vol. 15, p. 135.
41. MEYERS,
D. E, R. AND SNOW, P. J. Somatotopically
inappropriate
projections of single hair follicle afferent
fibres to the cat spinal cord. J. Physiol. London 347:
59-73, 1984.
42. MORGAN,
C., NADELHAFT,
I,, AND DE GROAT,
W. C. The distribution
of visceral primary
afferents
from the pelvic nerve to Lissauer’s tract and the spinal
gray matter and its relationship
to the sacral parasympathetic
nucleus. J. Comp. New-o/. 201: 415-440,
1981,
43. POMERANZ,
B., WALL,
P. D., AND WEBER,
W. V.
Cord cells responding to fine m yelinated afferents from
viscera, muscle and skin. J. Physiol. London 199: 5 ll532, 1968.
44. PUBOLS,
B. H., JR. AND PUBOLS,
L. M. Forelimb,
hindlimb
and tail dermatomes
in the spider monkey
(Ate/es). Brain Behav. Evol. 2: 132-l 59, 1969.
45. RALSTON, H.J.,III,
LIGHT, A.R., RALSTON, D.D.,
AND PERL, E. R. Morphology
and synaptic relationships of physiologically
identified low-threshold
dorsal
root axons stained with intra-axonal
horseradish
peroxidase in the cat and monkey,
J. Neurophysiol.
5 1:
777-792,
1984.
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
1 1, BROWN,
P. B. AND CULBERSON,
J. L. Somatotopic
organization
of hindlimb
cutaneous dorsal root projections to cat dorsal horn. J. Neurophysiol.
45: 137143, 1981.
12. BROWN,
P. B. AND FUCHS, J. L. Somatotopic
representation of hindlimb
skin in cat dorsal horn. J. Neurophysio(.
38: l-9, 1975.
P. B, AND KOERBER,
H. R. Cat hindlimb
13. BROWN,
tactile dermatomes
determined
with single-unit
recordings. J. Neurophysiol.
4 1: 260-267,
1978.
14. BRYAN, R. N., TREVINO,
D. L., COULTER,
J. D., AND
WILLIS,
W. D. Location and somatotopic
organization
of the cells of origin of the spinocervical
tract. Exp.
Bruin Rex 17: 177- 189, 1973.
15. BURGESS,
P. R. AND PERL, E. R. Cutaneous
mechanoreceptors
and nociceptors.
In: Handbook
ofSensor,v PhysioIogy.
Somatosensory
System, edited by A,
Iggo. Berlin: Springer-Verlag,
1973, vol. 2, p. 29-78.
16. CULBERSON,
J, L. AND BROWN,
P. 8. Projections
of
hindlimb
dorsal roots to lumbosacral
spinal cord of
cat. J. Neurophysiol.
5 1: 5 16-528, 1984.
17. CULBERSON,
J. L., HAINES,
D. E,, KIMMEL,
D. L.,
AND BROWN,
P. B. Contralateral
projection of primary
afferent fibers to mammalian
spinal cord. Exp. Neural.
64: 83-97, 1979.
18. CULBERSON,
J. L. AND KIMMEL,
D. L. Primary
afferent fiber distribution
at brachial and lumbosacral
spinal cord levels in the opossum (Didelphis
marsupialis virginiana).
Brain Behav. Evol. 12: 229-246,
1975.
19. DEVOR, M., MERRILL,
E. G,, AND WALL,
P. D. Dorsal
horn cells that respond to stimulation
of distant dorsal
roots. J. Physiol. London 270: 5 19-53 1, 1977.
20. FINK, R. P. AND HEIMER,
L. Two methods for selective silver impregnation
of degenerating
axons and
their synaptic endings in the central nervous system.
Brain Rex 4: 369-374,
1967.
21 FLETCHER,
T. F. AND KITCHELL,
R. L. The lumbar,
sacral and coccygeal tactile dermatomes
of the dog.
J. Comp. Neural. 128: 171-180,
1966.
22 GRANT, G. AND YGGE, J. Somatotopic
organization
of the thoracic spinal nerve in the dorsal horn demonstrated with transganglionic
degeneration.
J. Comp.
Neural. 202: 357-364,
198 1.
23. GREEN, J. D. A simple microelectrode
for recording
from the central nervous system. Nature London 182:
962, 1958.
24. GROSSMAN,
M. L., BASBAUM,
A. I., AND FIELDS,
H. L. Afferent and efferent connections
of the rat tail
flick reflex (a model used to analyze pain control
mechanisms).
J. Camp. Neural. 206: 9-l 6, 1982.
25. HEKMATPANAH,
J, Organization
of tactile dermatomes, C1 through L4, in cat. J. Neurophysial.
24: 129140, 1961.
26. IMAI, Y. AND KUSAMA,
T. Distribution
of the dorsal
root fibers in the cat. An experimental
study with the
Nauta method. Brain Rex 13: 338-359,
1969.
27. ISHIZUKA,
N., MANNEN,
H., HONGO, T,, AND SASAKI,
S. Trajectory
of group Ia afferent fibers stained with
horseradish
peroxidase in the lumbosacral
spinal cord
of the cat: three dimensional
reconstructions
from serial sections. J, Comp. NeuroI. 186: 189-2 i 2, 1979.
28. KEHL, T. H,, DUNKEL,
L,, AND Moss, C. LM2-a
logic machine minicomputer.
IEEE Computer
8: 1222, 1975.
AND
SOMATOTOPY
OF FUSED DORSAL
57.
58.
59
’
60.
6 1,
62.
63*
64.
65.
1177
nization of primary afferent projections to the rat thoracic spinal cord. J. Camp, Neural. 220: 29-43, 1983.
SMITH, C. L. Skin near the dorsal and ventral midlines
is represented bilaterally in the spinal cord of the rat.
Sot. Neurosci. Abstr. 9: 473, 1983.
SPRAGUE,J. M. The distribution of dorsal root fibres
on motor cells in the lumbosacral spinal cord of the
cat, and the site of excitatory and inhibitory terminals
in monosynaptic pathways. Proc. R. Sot. London Ser.
B 149: 534-556, 1958.
SPRAGUE, J. M. AND HA, H. The terminal fields of
dorsal root fibers in the lumbosacral spinal cord of
the cat, and the dendritic organization of the motor
nuclei. Pr0g. Brain Rex 11: 120- 154, 1964.
STERLING, P. ANDKUYPERS, H.G. J. M.Anatomical
organization of the brachial spinal cord of the cat. I.
The distribution of dorsal root fibers, Brain Res. 4:
1-15, 1967.
SZENTAGOTHAI, J. Neuronal and synaptic arrangement in the substantia gelatinosa Rolandi. J. Comp.
Neural. 122: 2 19-239, 1964.
WALL, P. D. Cord cells responding to touch, damage,
and temperature of skin. J. Neurophysiol. 23: 197210, 1960.
WALL, P. D. The laminar organization of dorsal horn
and effectsof descending impulses. J. Physiol. London
188: 403-423, 1967.
WALL, P. D. AND WERMAN, P. The physiology and
anatomy of long range afferent fibres within the spinal
cord. J. Physiol. London 255: 321-334, 1976.
YGGE, J. AND GRANT, G. The organization of the
thoracic spinal nerve projection in the rat dorsal horn
demonstrated with transganglionic transport of
horseradish peroxidase. J. Camp. Neural. 2 16: 1-9,
1983.
Downloaded from http://jn.physiology.org/ by 10.220.33.4 on June 18, 2017
46. REID, K. H. Dermatomes and skin innervation density
in the cat’s tail.Exp. Neural. 26: l-l 6, 1970,
47. REIGHARD, J. ANDJENNINGS, H. S. Anatomy of the
Cat. New York: Holt, Rinehart, & Winston, 1935.
48. RETHELYI, M. AND SZENTAGOTHAI, J. Distribution
and connections of afferent fibres in the spinal cord.
In: Handbook qf Sensory Physiology, Somatosensory
System, edited by A. Iggo, Berlin: Springer-Verlag,
1973, vol. 2, p. 207-252.
49, RETHELYI, M., TREVINO, D. L., AND PERL, E. R.
Distribution of primary afferent fibers within the sacrococcygeal dorsal horn: an autoradiographic study.
J. Camp. Neural. 185: 603-62 1, 1979.
50. REXED, B. A cytoarchitectonic atlas of the spinal cord
in the cat. J. Camp. Neural. 100: 297-379, 1954.
51, RITZ, L. A. AND BROWN, P. B. Projections of fibers
from slowly-adapting type I receptors to cat sacrocauda1 spinal cord. Sue. Neurosci. Abstr. 11: 220, 1985.
52. RITZ, L. A., CULBERSON,J. L., AND BROWN, P. B.
Sornatotopic organization of cat spinal cord segments
with fused dorsal horns. Sue. hreurosci. Abstr. 9: 473,
1983.
53 RITZ, L. A. AND GREENSPAN, J. D, Morphological
features of lamina V neurons receiving nociceptive
input in cat sacrocaudal spinal cord. J. Camp. Neurcll.
238: 440-452, 1985.
54. SIEGEL,S, Nclnparametric Statistics for the Behavioral
Sciences. New York: McGraw-Hill, 1956.
55. SMITH, C. L. Afferents which supply tactile domes
near the dorsal and ventral midlines project bilaterally
to the thoracic spinal cord in the rat. J. Physiol. London 334: 72-73P, 1983.
56, SMITH, C. L. The development and postnatal orEa-
HORNS