вход по аккаунту


Ultrastructure of the secretory epithelium nerve fibers and capillaries in the mouse sweat gland.

код для вставкиСкачать
THE ANATOMICAL RECORD 208:491-499 (1984)
Ultrastructure of the Secretory Epithelium, Nerve Fibers,
and Capillaries in the Mouse Sweat Gland
Department of Neurology, University of Minnesota, Minneapolis,
MN 55455
The ultrastructure of the mouse sweat gland was examined,
in support of neurological studies of sweat glands and their relationships to
the autonomic nervous system. It was found that the mouse sweat gland is
similar to that of the rat and has only one type of secretory cell. Many nerve
fibers are entwined with the secretory tubule and contain accumulations of
round, clear vesicles, some microtubules, but apparently no neurofilaments.
Cholinesterase is found in the clefts between nerve fibers and their ensheathing Schwann cells. The nerve fibers tend to run parallel with capillaries,
but have no close association with either the capillaries or the secretory
epithelium. Capillaries provide a n abundant blood supply to the sweat gland
and are fenestrated. The relationships between cellular elements of the sweat
gland provide no direct evidence of the mechanisms involved in neurogenic
sweating, although it seems likely that effector substances are diffusely
The mouse provides a convenient model for
the study of sweat glands and their innervation by the autonomic nervous system. Recent experimental neurological studies have
focused on the sweat glands of mouse volar
skin to determine their innervation by the
peripheral nerves (Kennedy et al., 1983a),
their responses to pharmacological and neurophysiological stimulation (Kennedy and
Sakuta, 1983),their response properties after
complete or partial denervation of the extremity (Kennedy and Sakuta, 1983; Kennedy et al., 1983a), and the course of
functional reinnervation (Kennedy and Sakuta, 1983). Some ultrastructural and histochemical features of the mouse sweat gland
have been reported by others (Kurosumi and
Kurosumi, 1970; Suzuki, 1973), but the
neural and vascular components have been
neglected. The present study provides new
ultrastructural information that can serve as
a basis for future studies of the mouse sweat
gland, its innervation, and its responses to
pathological conditions. It is expected that
work with this model system will lead to a
better understanding of how human sweat
glands are affected by diabetic neuropathy
(see Kennedy et al., 1983b,c) and other conditions involving the autonomic nervous
0 1984 ALAN R. LISS, INC.
Albino female Swiss-Webster mice weighing 30-35 gm were anesthetized with pentobarbital (“Nembutal,” Abbott, 120 m g k g
body weight). After excising the volar skin,
each animal was immediately sacrificed by
pneumothorax. The skin specimens were immersed in fixative, and small blocks of tissue
containing sweat glands were dissected €or
further processing.
For routine electron microscopy, the fixative was 5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.3). After 3 hours in this
solution at room temperature the specimens
were thoroughly washed in buffer, postfixed
for 1 hour in 2% osmium tetroxide in cacodylate buffer, dehydrated through a n ethanol
series, and embedded in epoxy resin. Silver
ultrathin sections were mounted on Formvar
films on grids and stained with uranyl acetate and lead citrate. Transmission electron
microscopy was done using Philips 300 and
JEOL lOOCX instruments.
To visualize cholinesterase by electron microscopy, the specimens were fixed and reacted according to the method of Karnovsky
Received May 2, 1983; accepted October 19, 1983.
(1964),incubating for 60 minutes a t 0°C. The
same procedure was used for light microscopy, except that the incubation was for 60
minutes a t room temperature and there was
no osmium fixation. The Koelle method for
cholinesterase (Hurley et al., 1953) was also
tested, but gave unsatisfactory results (see
Attempts were made to stain intrinsic
nerve fibers of the sweat gland by silver impregnation techniques (Barker and Ip, 1963;
Weddell and Pallie, 1954; Winkelmann,
1960), but these were unsuccessful (see Re-
sults). Zinc iodide-osmium tetroxide staining
of the nerve fibers was done by the method
of Betz et al. (1980), except that the concentration of the staining solution was reduced
by one-half.
General Features
The secretory part of the gland is composed
of a convoluted secretory epithelium enveloped by a basal lamina and a sheath of fibroblasts (Figs. 1, 2). Capillaries and fascicles of
Fig. 1. Low-magnification electron micrograph of a mouse sweat gland. Principal elements
are the tubular, coiled secretory epithelium (El, capillaries (C), nerve fibers (N), and fibrocytic
sheaths (F).A, fat cells. x 1,280.
Fig. 2. A cross-section of the secretory epithelium of
a mouse sweat gland. The nucleus of a secretory cell is
seen at center. Its apical cytoplasm contains ribosomes,
endoplasmic reticulum, and small vesicles. At the upper
right, part of another secretory cell can be seen, which
has slightly darker cytoplasm. M, myoepithelial cell; g,
Golgi apparatus; p, basal plications of secretory cells; m,
lumenal microvilli; b, basal lamina; F, fibrocytic sheath.
x 15,500.
unmyelinated nerve fibers course among the
secretory coils (Fig. 1).
plasmic reticulum, a Golgi apparatus, and
modest quantities of glycogen (Fig. 2); 4) a
myofibrillar matrix; and 5 ) desmosomes atSecretory Cells
taching the cells to adjacent secretory cells
Only one cell type could be distinguished, (Fig. 3). The myofilament system is unusual
in contrast to eccrine glands of some other in that thick filaments (7-8 nm) are embedspecies that have two distinct types (Sato, ded in a matrix of granular cytoplasm (Figs.
1977). Some cells appear to have a cyto- 2, 3); in other myoepithelial cells, including
plasmic ground substance that is slightly those of the human eccrine sweat gland (Eldenser than that of neighboring cells (Figs. lis, 1965), thin filaments (5 nm) are obvious
1,21, and the mitochondria of these cells are and thick filaments are not seen.
usually more condensed than in other cells
Fibrocytic Sheath
(see Munger and Brusilow, 19711, but these
differences are not striking and there apThis is a loosely organized sheet of flat cells
pears to be a continuous range of these char- surrounding the secretory coil (Fig. 2). Adacteristics among a single population of herent and occluding junctions are common
secretory cells.
between cells of the sheath, but there are
Prominent features of the secretory cells also occasional openings between cells (Fig.
are 1)microvilli at the lumenal surface (Fig. 11, allowing for free diffusion between the
2); 2) a lumenal complex of intercellular junc- interstitial spaces and secretory epithelium.
tions including tight junctions, intermediate Collagen fibers are found between the sheath
junctions, and desmosomes (Fig. 2); 3) a n ir- cells and the secretory epithelium (Fig. 2).
regularly folded lateral surface with occaNerve Fibers
sional desmosomes but no intercellular
Unmyelinated fibers are located among
canaliculi (Fig. 2); 4)a deeply plicated basal
surface with hemidesmosomes a t the tip of coils of the secretory epithelium, outside the
each fold adjacent to the basal lamina (Figs. fibrocyte sheath (Fig. 1).They appear to have
2, 3); 5) an extensive complement of endo- no regular relation to the secretory coils, nor
plasmic reticulum, mostly of the smooth to the capillaries except that nerve fibers run
variety, but occasionally studded with ri- parallel to capillaries in many cases (Figs. 1,
bosomes (Fig. 2); 6) free ribosomes, more 9). In rare instances, the fibers penetrate beabundant in some cells than others (Fig. 2); tween the fibroblast sheath and the secretory
7) abundant clear vesicles, 50-200 nm in di- epithelium, but no ultrastructural specialameter (Fig. 2); 8) a small Golgi apparatus izations of the nerve fibers or nearby cells
(Fig. 2); 9) small quantities of glycogen; 10) have been seen in these cases.
Usually several nerve fibers are enclosed
occasional lipofuscin granules, especially in
older animals; and 11)mitochondri- in
. a con- by a single Schwann cell sheath (Figs. 4-6).
densed or expanded condition. Gap junctions Accumulations of round, clear vesicles (30between secretory cells were not observed in 40 nm diameter) are seen very frequently
the mouse and have not been reported to be (Figs. 4-6,9). Where vesicles are present, the
present in the sweat glands of any other spe- nerve fibers usually have diameters in the
cies. Although gap junctions may indeed be range of 0.5-1.5 pm. In those parts of the
present, it appears that they are not common nerve fibers where vesicles are not seen, the
enough to provide for a great deal of commu- fibers are 0.4-0.6 pm in diameter. Microtunication of intracellular signals among cells bules are present in the nerve fibers (Fig. 4),
of the secretory epithelium.
but neurofilaments appear to be absent. Occasionally, a large, dense-cored vesicle (75Myoepithelial Cells
100 nm) may be seen among the clear vesiThese are found adjacent to the basal lam- cles, but there are no collections of denseina, between the bases of secretory cells (Fig. cored vesicles. Faint, cytoplasmic, submem2). Cytological features are very similar to braneous densities reminiscent of presynthose reported by Ellis (1965) for human ec- aptic structures are sometimes seen among
crine sweat glands, including 1)a scalloped collections of vesicles (Figs. 5, 6). However,
basal surface with hemidesmosomes adja- no synapse-like close contacts have ever been
cent to the basal lamina (Figs. 2, 3); 2) sub- observed between nerve fibers and any other
surface vesicles and associated pits (Fig, 3); cell type.
3) a peripheral nucleus surrounded by nonfiElectron microscopy of specimens stained
brillar cytoplasm with mitochondria, endo- for cholinesterase (Karnovsky method, 1964)
show reaction product primarily in the extracellular spaces between axons and their ensheathing Schwann cells (Fig. 6). Lesser,
background amounts of reaction product can
be seen adherent to the surfaces of other cells
and to collagen. In light microscopy, using
the same cholinesterase technique, reaction
product is seen to have a branching fibrillar
or ribbonlike distribution (Fig. 7). By correlating the ultrastructural and light microscopic data, it can be concluded that the light
microscopic technique is selective for fascicles of unmyelinated nerve fibers intermingled with the secretory tubules of the sweat
gland. In contrast, light microscopy of sweat
glands stained by the Koelle method (Hurley
et al., 1953) shows reaction product in a diffuse, nonfibrillar distribution. Nerve fibers
approaching the gland through the dermis
are usually not stained by either cholinesterase reaction.
Silver stains for nerve fibers (e.g., Barker
and Ip, 1963; Weddell and Pallie, 1954; Winkelmann, 1960) do not impregnate the nerve
fibers lying directly adjacent to the secretory
epithelium, although they do work well for
fibers leading up to the sweat glands and for
sensory fibers in the nearby dermis. The lack
of reactivity to silver stain may be related to
the apparent absence of neurofilaments in
these parts of the fibers (see above). The zinc
iodide-osmium tetroxide technique does impregnate fibers intermingled with the gland
tubules (Fig. 8), a s might be expected of fibers containing large numbers of cholinergic
vesicles (Akert and Sandri, 19751, but the
reaction has been difficult to control in our
The sweat glands have a good vascular supply, with a network of capillaries interwoven
with the secretory coils (Fig. 1).In the electron microscope, it can be seen that the capillaries are of the fenestrated type (Fig. 91,
and they are often accompanied by nerve
fibers (Fig. 9).
Our finding that the mouse sweat gland
does not have more than one type of secretory cell is in agreement with Kurosumi and
Kurosumi (19701, who also studied mouse
sweat glands, and with the findings of Munger and Brusilow (1971) and Wechsler and
Fisher (1968) for rat sweat glands. Munger
and Brusilow (1971) contend that this uniformity of cell types is distinctive enough
that the rat gland should be considered to be
neither eccrine nor apocrine, although they
propose no alternative designation. Kurosumi and Kurosumi (1970)refer to the mouse
plantar glands as “eccrine,” and we tend to
agree with this nomenclature because of the
apparent homology between mouse volar
sweat glands and the eccrine glands found
on palmar and volar skin of other species
(Sato, 1977).
Our observations on the ultrastructure of
the secretory epithelium are in agreement
with the report by Kurosumi and Kurosumi
(1970), except for their implication that cytoplasmic organelles of the secretory cells generally have a polarized apical or basal
distribution. We observed segregation of mitochondria toward the basal pole in some
cells; otherwise, the vesicles, smooth and
rough endoplasmic reticulum, free ribosomes, and other inclusions (e.g., lipofuscin)
were distributed rather uniformly.
In comparing the mouse sweat gland with
that of the rat (Landis and Keefe, 1983; Munger and Brusilow, 1971; Wechsler and Fisher,
1968),there appear to be no major differences
in ultrastructure of the secretory epithelium
or the nerve fibers. The ultrastructure of capillaries in the rat sweat gland has not been
It is known that stimulation of mouse peripheral nerves causes sweating, and denervation of skin precludes sweating in the
affected area (Kennedy et al., 1983a). Sweating can also be produced by local injection of
pilocarpine, suggesting that sudomotor nerve
fibers are cholinergic (Kennedy and Sakuta,
1983). Accordingly, one might expect that
there would be recognizable synapses between the nerve fibers and sudoeffector cells
in sweats glands, and that acetylcholinesterase would be localized in the vicinity of the
synapses. However, the only presynaptic
structures that are recognizable are the axon
swellings that contain clear vesicles and submembraneous densities (Fig. 5). These structures are not found in close association with
either the secretory epithelium or the capillaries, but rather, they appear to relate only
to the intercellular spaces and the Schwann
cell sheath. Cholinesterase is localized primarily in the intercellular spaces between
axons and Schwann cells. Similar results regarding human (Rechardt et al., 1976) and
rat (Landis and Keefe, 1983) sweat glands
have been reported previously, and in those
studies the reaction has been shown to be
due to specific acetylcholinesterase. In light
microscopy, cholinesterase appears to be uniformly distributed along nerve fascicles associated with the gland, with no areas of
concentration that might be suspected to be
nerve terminals. These results prevent us
from identifying the postsynaptic target cells,
and it is impossible to decide whether the
action of neurotransmitters is to directly
stimulate the secretory epithelium or to affect capillaries in such a way as to secondarily induce the production of sweat by the
secretory epithelium (see also Jenkinson,
1973; Lundberg et al., 1979).The innervation
of sweat glands has also been studied in other
species by electron microscopy (see, e.g., Hellmann, 1955; Jenkinson, 1970; Langer, et al.,
1981; Rechardt et al., 1976; Uno and Montagna, 1975),but these efforts, too, have failed
to reveal any close relationships between
nerve fibers and other components of the
The scarcity of gap junctions in the secretory epithelium should relate to the mode of
its innervation. If cells of the epithelium were
highly communicative through gap junctions, then a few synapses closely coupled to
a few single secretory cells could excite the
whole epithelium via gap junction coupling.
This would help to explain why conventional
synapses were not observed; i.e., they could
be rare. On the other hand, if gap junctions
were rare and intercellular communication
poor, then neural signals would have to influence the secretory cells by synapsing directly
on many or all of them, or by triggering a
diffuse activating system with access to many
Fig. 3. A myoepithelial cell of the mouse sweat gland.
Thick myofilaments are apparent, but thin filaments are
not. v, subsurface vesicles; a, desmosome linking the
myoepithelial cell to a n adjacent secretory cell; b, basal
lamina. ~26,100.
Fig. 4. Vesicles (v) and microtubules (t)can be seen in
nerve fibers of the sweat gland. Microfilaments (D can
be found in the ensheathing Schwann cells, but not in
nerve fibers. ~44,200.
Fig. 5. A number of small nerve fibers associated
with the sweat gland. Two dense-cored vesicles (d) can
be seen among the smaller lucent vesicles. A small subsurface density (s) is typical; such structures usually
have some closely associated vesicles and face a large
extracellular space. ~29,600.
or all of the secretory cells. Our findings, that
secretory cell gap junctions are rare or absent, and that direct synapses are also rare
or absent, tend to support the hypothetical
presence of a diffuse activation mechanism.
The sweat glands of a number of species,
including the horse (Hellmann, 1955; but see
also Bell and Montagna, 1972) cat (Hellman,
19.551, and dog (Morishima, 1970),are known
to have a n adrenergic innervation. In man
and the macaque, which have both cholinergic and adrenergic innervation of sweat
glands (Sato, 1977; Sat0 and Sato, 19811,
there are two types of terminal-like specializations of the nerve fibers-one with predominantly clear vesicles and the other with
dense-cored, presumably adrenergic vesicles
Wno, 1977; Uno and Montagna, 1975).In the
mouse, no concentrations of dense-cored vesicles were observed, although a few isolated,
large, dense-cored vesicles were often seen
among the concentrations of clear vesicles
(Fig. 5). It therefore appears that adrenergic
nerve fibers are not present. This would agree
with a report (Jenkinson, 1970) that monoamine oxidase, a marker enzyme for adrenergic fibers, is not found in the nerve fibers of
mouse sweat glands, and with pharmacological findings (Hayashi, 1968) that mouse
sweat glands do not respond specifically to
adrenergic agents.
The functional linkage between activity in
sudomotor nerve fibers and the production of
sweat remains a puzzle. In studies of macaque eccrine sweat glands isolated in vitro,
Sat0 and Sat0 (1981) found that exposure to
cholinergic and adreneregic agonists would
cause the glands to produce sweat, even
though the sweat gland vascular supply was
disrupted. This is evidence against the possibility that sweating is primarily dependent
on local changes in blood flow or capillary
permeability. Alternative possibilities are
that the secretory epithelium itself may be
sensitive to neurotransmitters, or that another cell type, for example, the Schwann cell
or capillary endothelial cell, may respond to
neurotransmitters and affect sweating by release of a secondary sudoeffector substance.
The finding that the Schwann cell sheath in
rat (Landis and Keefe, 19831, human (Rechardt et al., 19761, and mouse sweats glands
is associated with relatively high levels of
cholinesterase activity leads us to suggest
the possibility that the Schwann cells may
be the primary target cells in cholinergic
sweating, or may play a regulatory role in
that process.
Thanks to Richard Landis for his technical
assistance. This work was supported by grant
#BNS 8300313 from the National Science
Foundation (U.S.A.).
Akert, K., and C. Sandri (1975)Significance of the Maillet method for cytochemical studies of synapses. In:
Golgi Centennial Symposium: Perspectives in Neurobiology. M. Santini, ed. Raven, New York, pp. 387-399.
Barker, D., and M. C. Ip (1963) A silver method for
demonstrating the innervation of mammalian muscle
in teased preparations. J. Physiol. (Lond.), 169t73P74P (Abstract).
Bell, M., and W. Montagna (1972) Innervation of sweat
in horses and dogs.
- Br. J. Dermatol., 86:160163.
Betz, W.J., J.H. Caldwell, and R.R. Ribchester (1980)
Sprouting of active nerve terminals in partially inactive muscles in the rat. J. Physiol. (Lond.), 303:281297.
Ellis, R.A. (1965)Fine structure of the myoepithelium of
the eccrine sweat glands of man. J. Cell Biol., 27:
Hayashi, H. (1968) Functional activity of the sweat
glands of the mouse. Tohoku J. Exp. Med., 95:289-295.
Hellmann, K. (1955) Cholinesterase and amine oxidase
in the skin: A histochemical investigation. J. Physiol.
(Lond.), 129t454-463.
Hurley, H.J., Jr., W.B. Shelley, and G.B. Koelle (1953)
The distribution of cholinesterases in human skin, with
special reference to eccrine and apocrine sweat glands.
J. Invest. Dermatol., 21t139-147.
Jenkinson, D. McE. (1970) The distribution of nerves,
monoamine oxidase and cholinesterase in the skin of
the guinea-pig, hamster, mouse, rabbit and rat. Res.
Vet. Sci., 11t60-70.
Jenkinson, D. McE. (1973) Comparative physiology of
sweating. Br. J. Dermatol., 88~397-406.
Fig. 6. Electron micrograph of cholinesterase reaction
product, in relation to nerve fibers (N) and their ensheathing Schwann cells (S). e, extracellular space.
x 17,800.
Fig. 7. Light micrograph of three dissected, wholemounted sweat glands stained for cholinesterase (Karnovsky method, 1964). Numerous darkly stained fibers
course among the lucent coils of secretory epithelium.
x 135.
Fig. 8. Light micrograph of a dissected, wholemounted sweat gland stained by the zinc iodide-osmium
tetroxide technique for nerve fibers. Numerous fibers
are seen amongst secretory tubules of the gland, with
irregular elaborations. The duct (D) has no associated
nerve fibers. x240.
Fig. 9. A cross-section of a capillary to the mouse
sweat gland, showing obvious fenestrations (h). Nearby
nerve fibers (N) can be recognized by their content of
synaptic vesicles. x 16,200.
Karnovsky, M.J. (1964) The localization of cholinesterase activity in rat cardiac muscle by electron microscopy. J. Cell Biol., 23t217-232.
Kennedy, W.R., and M. Sakuta (1983)Collateral reinnervation of sweat glands. Ann. Neurol. (in press).
Kennedy, W.R., M. Sakuta, and D.C. Quick (1983a) Rodent eccrine sweat glands: A case of multiple efferent
innervation. Neuroscience (in press).
Kennedy, W.R., M. Sakuta, D. Sutherland, andF.C. Goetz
(1983b) The sweating deficiency in diabetes mellitus:
Methods of quantitation and clinical correlation. Neurology (in press).
Kennedy, W.R.. M. Sakuta, D. Sutherland, and F.C. Goetz
Quantitation of the sweating deficiency in diabetes mellitus. Ann. Neurol. (in press).
Kurosumi. K.. and U. Kurosumi (1970) Electron microscopy of the mouse plantar eccrine sweat glands. Arch.
Histol. Jpn. 31~455-475.
Landis, S.C., and D. Keefe (1983) Evidence for neurotransmitter plasticity in vivo: Developmental changes
in properties of cholinergic sympathetic neurons. Dev.
Biol., 98:349-372.
Langer, J., H.H. Goebel, and S. Veit (1981) Eccrine sweat
glands are not innervated in hereditary sensory neuropathy type IV.Acta Neuropathol. 54t199-202.
Lundberg, J.M., T. Hokfelt, M. Schultzberg, K. UvnasWallensten, C. Kohler, and S.I. Said (1979)Occurrence
of vasoactive intestinal polypeptide (VIP)-like immunoreactivity in certain cholinergic neurons of the cat:
Evidence from combined immunohistochemistry and
acetylcholinesterase staining. Neuroscience, 4:15391559.
Morishima, T. (1970) Histochemical studies on the munoaminergic nerve in the skin chiefly by means of the
fluorescence method of Falck and Hillarp. Acta Dermato-Venereol., 50t329-337.
Munger, B.L., and S.W. Brusilow (1971) The histophysiology of rat plantar sweat glands. Anat. Rec.,
Rechardt, L., T. Waris, and A. Rintala (1976)Innervation
of human axillary sweat glands. Scand. J. Plast. Reconstr. Surg., 10t107-112.
Sato, K. (1977) The physiology, pharmacology, and biochemistry of the eccrine sweat gland. Rev. Physiol.
Biochem. Parniacol., 79t51-131.
Sato, K., and F. Sat0 (1981) Pharmacologic responsiveness of isolated single eccrine sweat glands. Am. J.
Physiol., 240rR44-R51.
Suzuki, H. (1973) An electron microscopic study of keratinization in the mouse intraepidermal sweat duct.
Arch. Histol. Jpn. 35265-282.
Uno, H. (1977) Sympathetic innervation of the sweat
glands and piloarrector muscles of macaques and human beings. J. Invest. Dermatol, 69:112-120.
Uno, H., and W. Montagna (1975) Catecholaminecontaining nerve terminals of the eccrine sweat glands of
macaques. Cell Tissue Res., 158:l-13.
Wechsler, H.L., and E.R. Fisher (1968) Eccrine glands of
the rat. Response to induced sweating, hypertension,
uremia, and alteration of sodium state. Arch. Dermatol., 97,189-201.
Weddell, G., and W. Pallie (1954) The value of ‘spreading
factors’ in the demonstration of tissue nebral elements. Q.J. Microsc. Sci., 95:389-397.
Winkelmann, R.K. (1960)Nerve Endings in Normal and
Pathologic Skin. Thomas, Springfield.
Без категории
Размер файла
1 232 Кб
ultrastructure, capillaries, fiber, epithelium, secretory, nerve, gland, mouse, sweat
Пожаловаться на содержимое документа