Periapical innervation of the ferret canine and the local retrograde neural changes after pulpectomy.

THE ANATOMICAL RECORD 220:318-327 (1988)
Periapical Innervation of the Ferret Canine and the
Local Retrograde Neural Changes
After Pulpectomy
G.R. HOLLAND
Division of Endodontics, Faculty of Dentistry, University ofAlberta,
Edmonton, Alberta, Canada T6G 2N8
ABSTRACT
The amputation of the dental pulp severs a population of axons that
are predominantly in the A6 and C fiber size range and are principally involved in
nociception. Local periapical neuromas, if they are formed after pulpectomy, may be
the sites of spontaneous nervous activity that may, in some circumstances, be
involved in the genesis of chronic pain. The periapical tissues from the mandibular
canines of four ferrets were examined 3 months after pulpectomies. Silver-stained
paraffin sections were examined in three dimensions a t the light microscope level.
Ultrathin sections were examined a t the electron microscope level. Compared with
contralateral and independent controls, the principal changes were the loss of the
periodontal plexus around the root apex, the extension of damage well below the
apical foramen, and the persistance of inflammation 12 weeks postoperatively.
While a somewhat disorderly mass of nerve fibers develops subapically, the arrangement has only some of the features usually associated with neuromas.
The tooth pulp has a rather unusual innervation. It
contains dense populations of A6 and C fibers (Johnsen
and Karlsson, 1974; Beasley and Holland, 1978), and the
sensation experienced from various stimuli applied to it
is pain (Anderson et al., 1970). While recent studies have
shown that the pulp does in fact contain A6 fibers (Cadden et al., 1982; Holland and Robinson, 1983) and may
be the source, under some circumstances, of sensations
other than pain (Grusser et al., 1982), it is probably true
that the predominant role of its innervation is nociceptive. Thus, when the dental pulp is destroyed, a preponderance of pain fibers are damaged. Tooth extraction or
pulpectomy is effectively amputation of a specialized
nociceptive population-a modality-specific form of deafferentation. The axons are severed 1-2 cm from their
termination and their target organ is removed.
It is possible that local changes after peripheral nerve
damage may be involved in the genesis of chronic pain.
Spontaneous activity originates in neuromas formed at
the site of damage or from the sensory ganglion cells
(Wall and Gutnik, 1974; Burchiel and Russell, 1985;
Wall and Devor, 1983). Little is known about the local
changes after damage to the nerves of the dental pulp.
Hansen (1980) found the existence of small traumatic
neuromas at the base of extraction sockets. Ratner et al.
(1979) described the existence of small intrabony defects
at the site of previous extractions in patients with trigeminal neuralgia. Histological examination of the contents of these defects shows neuroma-like tissue. After
curettage of these areas the symptoms of trigeminal
neuralgia are lost. No one has thus far looked at the
possibility of neuroma formation after damage
- limited
to the dental pulp.
The study reported here is a n initial step in examining
the local neural changes after pulpectomy.
0 1988 ALAN R. LISS, INC.
MATERIALS AND METHODS
Five young adult ferrets 4-6 months old a t initial
preparation were used for these observations. In four of
the animals, under both general (ketamine hydrochloride, 40 mgkg) and local (0.5 ml 2% lignocaine with
1:100,000 epinephrine) anesthesia the dental pulp was
amputated from the lower canine teeth on one side by
removing 2-3 mm of the crown tip and debriding the
root canal to the apical constriction with endodontic files
and 2% sodium hypochlorite solution. The canal space
was filled with gutta percha and zinc oxide and eugenol
cement. The fifth animal acted as a n unoperated control,
in which the teeth were left intact.
Three months after pulpectomy the animals were
killed by intravasuclar perfusion of ice-cold 2% glutaraldehyde and 0.1%osmium tetroxide in 0.01 M cacodylate buffer after initial clearing of the circulation with
0.01 M cacodylate buffer to which heparin and procaine
hydrochloride had been added. Perfusion was maintained for 15 min. Mandibles from three animals were
decalcified in formic acid. The canine teeth and supporting bone from these animals were embedded in paraffin
for light microscope examination. Two mandibles were
decalcified in 4% EDTA containing 1%glutaraldehyde
a t 4°C. The decalcification was monitored radiographically and when it was complete, the region of the mandible containing the root of the canine tooth was slit into
1-mm-thick slices, the radiographs being used to control
orientation and ensure that the periapical tissues were
included. These slices were then processed for electron
microscopy and embedded in LX112 resin.
Received May 13, 1987; accepted July 23, 1987.
PERIAPICAL INNERVATION AFTER PULPECTOMY
Fig. 1. Periapical region immediately adjacent to the apical canal (A) entering the tooth pulp
from a n unoperated tooth. C, cementum; B, bone; NF, nerve fibers. Hematoxylin and eosin.
Fig. 2. Periapical region showing separate nerve fiber bundles (NF)passing to the periodontal
ligament (upper) and to the apical canal of the tooth (lower). Unoperated tooth, siIver-stained.
B, bone; C, cementum of root apex.
3 19
320
G.R. HOLLAND
Fig. 3. Nerve fiber bundles (NF) in the periodontal ligament lateral to the apical foramen. B,
bone; C , cementum. Unoperated tooth, silver-stained.
Light Microscopy
The paraffh blocks containing the canine roots and
surrounding tissues were serially sectioned a t 8 pm.
Every 10th section was stained with hemotoxylin and
eosin. The region where the blood vessels and nerves
entered the tooth pulp at the root apex was identified
from those sections and 50 serial sections to each side of
this central plane were stained with silver to demonstrate the innervation by the method described by Linder (1978).
The hemotoxylin- and eosin-stained sections of the
periapical region were examined for evidence of gross
changes in the structure of the tissues. The silver-stained
sections were used to construct three-dimensional reconstructions of the periapical innervation. Each section of
the periapical series was examined in the photomicroscope and traced by means of a camera lucida attachment at a magnification of 200 X. Only the profile of the
root apex and the silver-stained nerves were traced onto
d e a r acrylic sheets. These were assembled into a reconstruction in two ways. Initially the sheets were merely
manually superimposed, the profile of the root apex
being used as a fixed point. A final composite reconstruction was made by adding a final sheet to the pile and
tracing all the underlying nerve fibers onto it, effectively compressing the information from a 800-pm slice
of tissue onto one sheet. In addition the nerve and apical
profiles were digitized with software (Bioquant, R&M
Biometrics, Nashville, TN) that would, as well as store
the data, build the reconstruction and rotate it in various planes to determine whether the relationships we
saw in single sections or in the superimposed series
occurred in three dimensions.
Electron Microscopy
The plastic-embedded blocks were all initially sectioned at 1 pm and the first sections of each block were
examined after toluidine blue staining. Regions of the
block containing the subapical nerve fibers were located
and the blocks were trimmed to include them only. These
were then thin-sectioned, stained with uranyl acetate
and lead citrate, and examined in the electron microscope. The innervation was located and examined and
low-power montages were prepared of some components.
Axon sizes were measured with a digitizing pad and
appropriate software (Bioquant, R&M Biometrics, Nashville, TN).
RESULTS
Nerve Plexus in the Periapical Region
In individual sections stained with toluidine blue or
hemotoxylin and eosin the dental branches of the inferior alveolar nerve can be distinguished only with difficulty and little can be discerned of their distribution or
arrangement except that they pass through clear channels in the bone and are almost always adjacent to blood
vessels (Fig. 1).
The use of a selective silver stain shows the neural
arrangement more clearly. In single sections small seg-
PERIAPICAL INNERVATION AFTER PULPECTOMY
321
appear widely separated compared with the inferior alveolar nerve and the endoneurium is sparse and poorly
developed. Some smaller bundles containing only 2-5
myelinated axons also occur, as do occasional bundles
containing only nonmyelinated axons. Occasional nonmyelinated fibers containing only one mitochondria-rich
axon are seen close to the root apex. The bundles sometimes pass close to the bone, being separated from it
only by a lining of osteoblasts. Immediately adjacent to
the root tip the bundles are much smaller and appear
equidistantly spaced from the cemental surface.
A sampling of the nerve bundles shows that 45% of
the axons in this area are nonmyelinated. The myelinated axons range in size from 4 to 38 pm in circumference. The size distribution is askewed unimodal with
the mode at 15 pm (mean 17.13; SD 5.93); 35% of the
myelinated axons are in the A/3 size range.
Occasional lamellated axons were seen close to the
tooth apex.
Affect of Pulpectorny on the Periapical Plexus
Fig. 4. Reconstruction of the periapical innervation made by superimposing camera lucida tracings from silver-stained sections. The
hatched area represents the profile of the tooth apex. The larger bundle
to the left is part of the inferior alveolar nerve trunk. Two large but
separate branches innervate the tooth pulp and the periodontal ligament around the tooth apex. The plexus around the apex is contained
within the periodontal ligament. Unoperated tooth.
The principal difference in operated specimens is the
presence of areas of inflammation in the subapical tissues (Fig. 7). In some sections it appears that the periodontal space innervation is sparse (Fig. 8), and in some
sections it appears that branching is much more extensive in the bone between the inferior alveolar nerve
trunk and the tooth apex.
The manually superimposed compound images confirm these superficial impressions (Fig. 9). Only occasional bundles are found in the periodontal space of the
apex. The branching pattern in the bone is more extensive, resulting in a larger number of more randomly
arranged small branches.
The computer reconstruction (Fig. 10) shows the
changes even more clearly. As the image is rotated
through O", 40°,80°,and 120" the presence of a proliferated mass of nerves in the bone beneath the apex is
confirmed as is the almost total loss of the periodontal
plexus, those fibers that remain being largely a greater
distance from the cemental surface than in the control
apices.
The tissue is infiltrated with chronic inflammatory
cells including lymphocytes and macrophages. The nerve
bundles appear structurally normal even within the inflamed zone. Inflammatory cells do not appear within
the epineurium and there is no overt evidence of neural
degeneration. The structural characteristics noted for
the control side were maintained on the operated side.
A higher proportion of the axons were nonmyelinated
(59%)and these axons showed a somewhat higher incidence of axoaxonal apposition and exposure of the axolemma to the extracellular space. The size range and
distribution of the myelinated axons were similar to
those of the control side. The mean size of myelinated
axons was 18.17 pm (SD 7.711, not statistically significantly different from that on the control side (MannWhitney, P < 0.01).
The occasional lamellated receptor terminal was also
present in the operated as in the control pericemental
zone.
ments of the bundles traveling towards the apex may be
seen (Fig. 2), as well as individual bundles in the periodontal space (Fig. 3).
The images produced by manually superimposing serial tracings give a clearer presentation of the branching
of these nerve bundles (Fig. 4).A large trunk leaves the
inferior alveolar nerve and divides dichotomously several times before reaching the root apex. The daughter
branches continue in the same general direction, ending
in different segments of the apical periodontal space.
Once approximated to the root apex some branches enter the apical foramen, but the majority end in the
periodontal space by branching extensively and forming
a definite neural plexus around the root apex midway
between the cemental and bone surfaces. The examination of individual sections (Fig. 5) shows that in this
periodontal plexus most fibers are oriented parallel to
the cemental surface and most accompany arterioles or
capillaries.
Examination of the computer-generated reconstructions (Fig. 6) allows extension of these observations into
three dimensions. At the 0" rotation (around the Y axis)
(Fig. 6A) the density of the periodontal plexus can be
seen, its relationship parallel to the root surface and the
larger bundles that supply its branches. Rotating the
image through 40" (Fig. 6B), 80" (Fig. 6 0 , and 120"
(Fig. 6D) shows that the relationships that were seen in
two dimensions persist in the third dimension and that
the periodontal plexus is indeed a three-dimensional
network completely enveloping the root apex on all sides.
The nerve bundles derived from the inferior alveolar
DISCUSSION
nerve are relatively small when found in the immediThe normal arrangement of the innervation of the
ately subapical zone. Each bundle contains 10-30 myelinated axons accompanied by a similar number of periapical tissues of the ferret consists of mixed myelinnonmyelinated axons. The axons within the bundles ated and nonmyelinated fibers in small bundles derived
-
322
G.R. HOLLAND
Fig. 5. A segment of the periodontal ligament (PDL) between cementum (C) and bone (B).
Nerve fibers (arrow) are restricted largely to the central region of the ligament adjacent to
blood vessels. Unoperated tooth, silver-stained.
from the inferior alveolar nerve. The bundles branch
dichotomously within the bone between the nerve trunk
and the tooth apex. Some bundles travel directly to the
root apex and, while giving some branches to the periodontal space, enter the apical foraminae to supply the
dental pulp. Other bundles divert from the most direct
root to the apex and pass through bone to reach the
periodontal space more coronally. There is little branching within the bone but once the bundles reach the
periodontal space extensive division occurs to form a
network of smaller bundles and individual fibers, a distinct periodontal plexus. Three-dimensional reconstruction emphasizes the density of this plexus within a
relatively confined but well-defined area. It appears to
surround the apex on all sides. The plexus is almost
entirely limited to the middle third of the ligament
space with relatively nerve-free zones between it and
both the cementum and bone. The nerve bundles almost
always accompany blood vessels. An initial impression
is that this plexus is much denser in the area immediately around the apex than a t more coronal levels though
this point requires further investigation. This description corresponds with those by Byers (1985) of the rat
molar and by Kubota and Osanai (1977) of the Japanese
shrew mole.
The role of the periodontal innervation has been subject to some debate (Hannam, 1985).While the response
properties of individual fibers have been described in
some detail (Hannam, 19691, the activity of this input in
reflex and volitional jaw movements is unclear. This
study can add little to the debate on function but such
a n extensive plexus, precisely positioned in the periodontal space, must provide during mastication a substantial input on both tension and compression. The
proximity of nerves and blood vessels is interesting.
Presumably the blood vessels are placed such that their
patency is maintained during tooth movement. This
would, one would suppose, be the position of minimum
stress and not the optimum position for receptors. It may
be that the terminals as opposed to the fibers are positioned in more appropriate sites. This requires further
study.
The composition of the nerve bundles as they approach
the apex differs from the composition of the inferior
alveolar nerve (IAN) (Holland, 1978; Fried and Hildebrand, 1982) in that, unlike the inferior alveolar nerve
trunk, the myelinated spectrum is unimodal and from
the intrapulpal innervation (Holland and Robinson,
1983; Johnsen and Karlsson, 1974; Fried and Hildebrand, 1981) in the higher proportion of AP fibers. The
proportion of non-myelinated axons is similar to that of
the IAN and much smaller than that of the pulp. The
high proportion of axoaxonal exposure reported in pulpal nerves (Holland, 1981) is absent.
PERIAPICAL INNERVATION AFTER PULPECTOMY
\
,
v::.,.>:.:.:.>>::
:.>:.>:..
........
.........
...............
........
I
6A
v.......
::::::.
.........
:.:.:.:.:.>:.
323
......> ...:.>:1
..........
.............
2mm
4
6C
u
2mm
6D
Fig. 6. Computer reconstruction of the periapical innervation of a control tooth rotated around
the Y axis. The profile of the tooth apex has been shaded for clarity. Rotation are as follows. A
0". B: 40". C: 80". D: 120". The relationship of the nerve fibers in the periodontal ligament to
the root surface remains relatively constant, suggesting it is a three-dimensional plexus present
on all sides of the root surface.
The nerve bundles supplying the periapical tissues
contain both pulpal and periodontal fibers. The difference in composition between them and the pulpal nerves
would suggest that the bulk of the AP fibers go to the
periodontal tissues, though within the ligament the fiber composition is as yet unknown. This would correlate
well with the known mechanoreceptive role of the periodontal ligament and the nociceptive function of the
dental pulp. The even higher proportion of A@fibers in
the inferior alveolar nerve trunk would presumably be
an expression of the cutaneous and mucosal fibers it
contains, which are absent from the periapical bundles.
It seems likely too that a larger proportion of the nonmyelinated axons enter the dental pulp than the periodontal ligament. This study did not include a detailed
examination of the receptors or nerve terminals though
we did observe both mitochondria-rich axonal segments
and lamellated Paccinian corpuscle-like endings. If
structure reflects function as elsewhere (Sakada, 1971),
it would suggest that these lamellated bodies are highfrequency, fast-adapting transducers and would probably signal the initial onset of tooth movement.
324
G.R. HOLLAND
Fig. 7. Nerve fibers passing through an area of subapical inflammation beneath a pulpectomized tooth. The infiltrate of leukocytes suggests a mild degree of inflammation. No damage to
the nerve bundle is apparent. Hematoxylin and eosin. B, bone; C , cementum; NF, nerve fiber
bundles.
Fig. 8. Segment of periodontal ligament (PDL) from a pulpectomized specimen. No nerve
fibers are present in this area. Silver stain (no counterstain). B, Bone.
325
PERIAPICAL INNERVATION AFTER PULPECTOMY
1m m
Fig. 9. Reconstruction of the periapical innervation made by superimposing camera lucida tracings from silver-stained sections. The
hatched area represents the profile of the tooth apex. The large bundle
to the right is part of the inferior alveolar nerve. Most of the periodontal plexus is absent. The accumulation of small plexus nerve fibers in
the subapical region is more extensive than in unoperated specimens.
Pulpectomy has a dramatic effect on the innervation
of the periapical region when observed 3 months later.
It was somewhat surprising to find residual inflammation at this time. The degree of inflammation did not
cause smicient bone loss to be detectable radiographically and consists of a leukocytic infiltration with little
overt evidence of destruction. Within inflammed areas
in the pulp, the nerves are the last elements to be affected (Torneck, 1974).There were no inflammatory cells
within the perineurium and the structure of most of the
myelinated axons appeared normal. The nonmyelinated axons, however, commonly. showed degenerative
changes. Presumably the protection afforded the unmyelinated axons from extracellular toxins is not as effective as the multiple layered sheath of the myelinated
axons.
What is seen a t this 3-month stage is probably partial
recovery from the acute trauma. It may be that observations at earlier stages (currently under way) and later
(planned) will show a different pattern, perhaps more
extensive damage initially and maybe a more normal
pattern finally. If this is so, the repair process would
appear to be extended compared to that observed after
simple transverse section of a nerve trunk in this region
(Holland and Robinson, 19851, where after similar time
periods the peripheral innervation had almost entirely
recovered.
The composition of subapical bundles differed little
from controls in terms of myelinated axons. A somewhat
higher proportion of nonmyelinated axons after pulpec-
J
c
10B
./
Fig. 10. Computer reconstruction of the periapical innervation of a pulpectomized tooth
rotated around the Y axis. The profile of the tooth apex has been shaded. The Rotation is as
follows. A 0". B: 40".C: 80'; D: 120". By comparison with the control specimens (Fig. 6 ) it
appears that the periodontal plexus has virtually disappeared and a network of small nerve
fibers between the tooth apex and the nerve trunk is present.
i
G.R. HOLLAND
326
2mm
1oc
.*
Fig. 10, continued.
tomy may indicate that the reparative process still is
under way. The major change discernible, particularly
in the three-dimentional reconstructions is the loss of
the periodontal plexus in the ligament and the increased
branching and accumulation of nerves below the apex
but outside the ligament. If this change were to persist
long, it may explain the differences in sensory discrimination between vital and nonvital teeth as being due to
a loss of periodontal rather than pulpal input. While it
would be attractive to consider the subapical neural
reaction to be a neuroma, examination of this ultrastructure denies this. It appears merely to be a proliferation
and increased branching with less confusion than demonstrated in neuromas (Blumberg and Janig, 1982). It
may well later organize and from it branches may return t o the periodontal ligament and reform the plexus.
At 3 months more extensive recovery may have been
expected. After similar periods of time we have seen
pulp and dentin reinnervated from nerve section lesions
some distance away (Holland, Matthews and Robinson,
1987). It may be that severed pulpal nerve attempts to
recover and reinnervate the target organ that has gone
and that this reorganization has a longer time course
than that occurring after simple transection of a nerve
trunk.
ACKNOWLEDGMENTS
This study was supported by the Medical Research
Council of Canada and the Alberta Heritage Foundation
for Medical Research. I would like to thank Mrs. E.
Pehowich and Mrs. V. Burgess for their expert technical
assistance and Mrs. H. Ridyard for typing the
manuscript.
LITERATURE CITED
Anderson, D.J., A.G. Hannam, and B. Matthews (1970) Sensory mechanisms in mammalian teeth and their supporting structures. Physiol. Rev., 50:171-195.
Beasley, W.L., and G.R. Holland (1978)A quantitative analysis of the
innervation of the pulp
.~ of the cat’s canine tooth. J. Comp. Neurol.,
178:487-494.
Blumberg, H., and H. Janig (1982) Changes in unmyelinated fibers
including postganglionic fibers of a shin nerve after peripheral
neuroma formation. J. Auton. Nerv. Syst., 6:173-183.
Burchiel, K.J., and L.C. Russell (1985)Spontaneous activities of ventral root axons following peripheral nerve injury. J. Neurosurg.,
62:408-413.
Byers, M.R. (1985) Sensory innervation of periodontal ligament of rat
molars consists of unencapsulated Riffini-like-mechano-receptors
and free nerve endings. J. Comp. Neurol., 231500-518.
Cadden, S.W., S.W.J., Lisney, and B. Matthews (1982)A fibre innervation of tooth pulp in the cat, with a discussion of the functions of
nerves supplying tooth-pulp. In: Anatomical, Physiological and
Pharmacological Aspects of Trigeminal Pain. B. Matthews and
R.G. Hill, eds. Excerpta Medica: Amsterdam, pp. 41-50.
Fried, K., and C. Hildebrand (1981)Pulpal axons in developing, mature
and ageing feline permanents incisors. J. Comp. Neurol., 203:3751.
Fried, K., and C. Hildebrand (1982)Axon number and size distribution
in the developing feline inferior alveolar nerve. J. Neurol. Sci.,
53:169-180.
Grusser, O.J., W. Kollman, and E. Mijatovic (1982) Evidence for cold
perception from human teeth. In: Anatomical, Physiological and
Pharmacological Aspects of Trigeminal Pain. B. Matthews and
R.G. Hill, eds. Excerpta Medica: Amsterdam, pp. 77-95.
Hannam, A.G. (1969)The response of periodontal mechanoreceptors in
the dog to controlled loading of the teeth. Arch. Oral Biol., 14:781791.
Hannam, A.G. (1985)The innervation of the periodontal ligament. In:
The Periodontal Ligament in Health and Disease. B.K.B. Berkovitz and B.J. Moxham, eds. Pergamon: Oxford, pp. 173-196.
Hansen, H.J. (1980) Neuro-histological reactions following tooth extractions. Int. J. Oral Surg., 9:411-426.
Holland, G.R. (1978) Fibre number and sizes in the inferior alveolar
nerve of the cat. J. Anat., 127:343-352.
Holland, G.R. (1981) The effect of buffer molarity on axonal exposure
and axoaxonal apposition in the rat molar pulp. Anat. Rec.,
201:47 1-476.
Holland, G.R., and P.P. Robinson (1983) The number and size of axons
at the apex of the cat’s canine tooth. Anat. Rec., 205:215-222.
Holland, G.R., and P.P. Robinson (1985) Reinnervation of the canine
tooth pulp after section of the inferior alveolar nerve in the cat.
Brain Res., 329:300-303.
Holland, G.R., B. Matthews, and P.P. Robinson (1987)An electrophysiological and morphological study of cat dentine following denervation and reinnervation. J. Physiol., 384:31-44.
PERIAPICAL INNERVATION AFTER PULPECTOMY
327
Johnsen, D., and U.L. Karlsson (1974) Electron microscopic quantitaperiostea to mechanical stimuli. In: Oral-Facial Sensory and Motor
tions of cat primary and permanent incisor innervation. Arch Oral
Mechanisms. R. Dubner and Y. Kawamita, eds. Appleton-CenturyBiol., 19:671-678.
Crofts, New York, pp. 105-122.
Kubota, K., and K. Osanai (1977) Periodontal sensory innervation of Torneck, C. (1974) Changes in the fine structure of the dental pulp in
human caries pulpitis. Part 1. Nerves and blood vessels. J. Oral
the dentition of the Japanese shew-mole. J. Dent. Res., 56,531-537.
Pathol., 3.7-82.
Linder, J.E. (1978) A simple and reliable method for the silver impregnation of nerves in parafin sections of soft and mineralized tissues. Wall, P.D., and M. Devor (1983) Sensory afferent impulses originate
from dorsal root ganglia as well as from the periphery in normal
J. Anat., 127.543-551.
and nerve injured rats. Pain, 17;321-339.
Ratner, E.J., P. Person, D.J. Kleinman, G. Shklar, and S.S. Socransky
(1979) Jawbone cavities and trigeminal and atypical facial neural- Wall, P.D., and M. Gutnik (1974)Ongoing activity in peripheral nerves:
The physiology and pharmacology of impulses originating from a
gias. Oral Surg. Oral Med. Oral Pathol., 48:3-20.
neuroma. Exp. Neurol., 43:580-593.
Sakada, S. (1971) Response of Golgi-Mazzoni corpuscles in the cat