Species differences in the synaptic membranes of the end bulb of held revealed with the freeze-fracture technique.

THE ANATOMICAL RECORD 20557-63 (1983)
Species Differences in the Synaptic Membranes of the
End Bulb of Held Revealed With the
Freeze-Fracture Technique
D. E. MATTOX, G. R. NEISES, AND R. L. GULLEY
Department of Surgery, Division of Otorhinolaryngology, The Uniuersity of Texas Health
Science Center at San Antonio, TX 78284 (D.E.M.), and Labomtory of Neurootolaryngology, NINCDS, National Institutes of Health, Bethesda, MD 20014 (GAN.,
RLGJ
ABSTRACT
Two morphological differences distinguish the membranes of the
end bulb-spherical cell synapse in rats and mice from those in guinea pigs and
chinchillas. First, in freeze-fracture replicas, the membranes of rat and mouse
spherical cells lack perisynaptic aggregates which are present in the other species.
Second, small gap junctions are present between the end bulb and spherical cell
soma of rats and mice. These interspecies differences are not reflected in thinsectioned material. This observation points out the difficulty in attempting to
generalize about the significance of intramembrane specializations in synaptic
membranes.
The freeze-fracture technique has been used
in a variety of species to study the organization
of synaptic membranes in different areas of the
nervous system such as the cerebellum (Landis
and Reese, 1974;Korte and Rosenbluth, 1980),
the statoacoustic organ (Bagger-Sjoback and
Flock, 1977; Gulley and Bagger-Sjoback, 1979;
Bagger-Sjoback and Gulley, 19791, and the
neuromuscular junction (Heuser et al., 1974;
Rash and Ellisman, 1974). The distribution of
intramembrane particles is remarkably similar at these different contacts. Those differences that have been identified correlate with
differences in the shape of the presynaptic active zone (Bagger-Sjoback and Flock, 1977;
Korte and Rosenbluth, 1980)and with whether
or not the synapse is excitatory or inhibitory
(Landis and Reese, 1974). No intra- or interspecies differences have been described in the
organization of synaptic membranes.
The anterior branch of auditory nerve fibers
terminate in the rostral anteroventral cochlear
nucleus (AVCN) as a large calyceal terminals
on the soma of sperical cells. The ultrastructure of this terminal has been described in cat
(Ibata and Pappas, 19761, guinea pig (Pirsig et
al., 1969; Gulley et al., 1978a), chinchilla (Lenn
and Reese, 1966; Gulley et al., 1978a), and rat
(Gentschev and Sotelo, 1973). The internal organization of its membranes has been described in chinchilla and guinea pig using the
freeze-fracture technique (Gulley et al., 1978a).
Since the general cytological features of the
terminals seen in thin sections are similar in
these species, we assumed the organization of
their membranes would also be similar. However, distinct species differences are present in
the membranes of this synapse. This is the first
description of species differences in the organization of synaptic membranes.
METHODS AND MATERIALS
The cochlear nuclei of ten adult albino rats
and two brown rats were prepared for either
thin section or freeze-fracture study. Freezefracture study of an albino guinea pig, two tricolor guinea pigs, two albino mice, and two
brown mice were also included. All animals
apparently were able to hear, at least as demonstrable by a startle response to auditory
stimuli (Preyer’s reflex). Additional material
was reviewed from cats, chinchillas, and guinea
pigs used in previous studies. All animals were
anesthetized with 60% urethane (0.25 mYlOO
gm), perfused with 0.1 M sodium cacodylate
with 20 mM CaC12 followed by 3% glutaraldehyde, 2% paraformaldehyde in 0.1 M sodium
cacodylate, and 20 mM CaC12. After perfusion,
the head was removed and the skull was opened
and placed in fixative.
Freeze -Fracture
The brain was dissected 2 hr after perfusion,
and 250-km thick slices of the brain stem were
cut on a vibratome. The rostral AVCN was
0003-276X/83/2051-0057$02.500 1983 ALAN R. LISS, INC
h i v e d April 2,1981;accepted Sept. 29, 1982.
58
D.E. MATFOX, G.R. NEISES, AND R.L. GULLEY
and chinchilla, each junctional aggregate is
encircled by up to six perisynaptic aggregates
(Fig. 6) on the spherical cell E-face opposite
channels of enlarged extracellular space. These
aggregates are not present in the rat on either
the E-face or P-face. The number of nonaggregate particles on the E-face of rat spherical
cells (Fig. 4,5) is similar to that in guinea pig
and chinchilla;' thus it is unlikely that the
particles constituting the aggregates in these
species are merely dispersed in the membranes
of the rat. Circular clusters of uniformly sized
large particles are present on the P-face of the
spherical cell associated with subadjacent subThin Sections
surface cisternae. They do not resemble the
The skull remained in fixative for 12 to 24 perisynaptic aggregate in size, shape, or parhr before dissection. After dissection, 250-pm ticle density.
thick slices were cut through the brain stem,
To ensure that the absence of the perisyand levels of the AVCN were selected which naptic aggregate is not an artifact of tissue
were identical to those used for freeze-fracture. preparation or fracturing, a rat and guinea pig
These were postfixed a t 4°C in 1.5%potassium were each perfused with one liter of fixative
ferrocyanide and 1%OsOlin 0.05 M sodium taken from the same batch. The tissue was
cacodylate buffer. All slices were dehydrated processed together through the same solutions
in a graded series of methanol and embedded and was fractured and replicated simultanein Epon. Sections, 1 to 2-pm thick, from each ously on a multiple-specimen stage. Perisyblock were examined for orientation before thin naptic aggregates were present in the guinea
sections were cut and stained with uranyl ace- pig, but were absent in the rat. Since some
tate and lead citrate. Thin sections and repli- albino strains have hearing disorders which
cas were examined in a Philips 201C electron
microscope.
Fig. 1. Thin section through an apposition of an end bulb
Frozen sections, 5-10 pm thick, of rat brain (E) and spherical cell (S) from an albino rat. Active zones
stem were cut and alternate sections stained (asterisks) appear as invaginations of the end bulb. Enwith cresyl-violet-luxolfast blue and the Bod- larged channels of extracellular space containing a glial
ian silver stain. Rapid Golgi and Golgi-Kopsch process (arrows) separate some of the active zones. 17,000
material, from neonate, young adult, and adult X .
rat used in an earlier study, was studied along
Fig. 2. Thin section through a portion of an end bulb (El
with the frozen sections, to obtain a general opposed to a spherical cell (S) in the albino rat. At least two
understanding of the organization of the ros- somatic appendages (stars) are present and synapse with
the end bulb (asterisks). Smaller, more irregular processes
tral AVCN in the rat.
around the appendages are probably from glial cells. 30,000
dissected from both sides and placed in 5%
glycerol in 0.1 M cacodylate buffer. The tissue
was transferred over a 1-hr period through 10%
and 15%glycerol-buffer to 20% glycerol-buffer
in which it remained for an additional hour.
The pieces were then frozen in Freon 22 and
fractured in a Balzers 301 freeze-fracture unit
at -119°C. Platinum-carbon replicas were made
with electron beam guns with the replica
thickness standardized with a quartz crystal
monitor. The replicas were cleaned in cold
methanol, Chlorox bleach, and water before
picking them up on uncoated grids.
RESULTS
In the rat, a single terminal or end bulb from
a primary auditory fiber envelops the dentritic
pole of each spherical cell in the dorsal portion
of the rostra1 AVCN. From this end bulb, numerous synapticprocesses are distributed. Other
non-primary terminals contact the remainder
of the soma. In thin sections (Fig. 11, the end
bulb and spherical cell resemble that seen in
the guinea pig and chinchilla (GuIley et al.,
1978a). In the rat, unlike the guinea pig or
chinchilla, the spherical cell soma has occasional finger-like appendages located on the
pole of the cell that is enveloped by the end
bulb (Figs. 2,4). In freeze-fracture replicas, the
E-face of the postsynaptic active zone has a
junctional aggregate (Fig. 3) identical to that
in guinea pig or chinchilla. In the guinea pig
X.
Fig. 3. Replica in which the fracture exposes a portion
of a junctional aggregate on the E-face of a spherical cell
membrane in an albino rat. The invagination of the presynaptic active zone is opposite the aggregate of large, tightly
packed, irregular particles which appears to be coextensive
with it. 62,000 x .
Fig. 4. A replica of the P-face of an end bulb (El opposed
to fragments of spherical cell E-face in a brown rat. The
perspective is from the inside of the spherical cell. Several
active zones (asterisks) and channels of enlarged extracellular space, one containing a glial process (arrow), are present. Cmfractured cytoplasm of aomatic appendages (stars)
protrude around and into the end bulb .No perisynaptic aggregates are present on the fragments of the E-face of the
spherical cell. 28,000 x .
lln each species, particles were caunted on 20 wm2 of membrane Eface which had a aimilar shadow angle.
SYNAPTIC MEMBRANES OF THE END BULB OF HELD
59
Fig. 5. A freeze-fracture replica of a spherical cell and
end bulb in an albino rat. The fracture crosses the spherical
cell cytoplasm (S) before exposing its E-face. Through windows in this E-face, the P-face of the end bulb (E)is seen.
Three active mnes (asterisks) are present. An arrow points
to an enlarged channel of extracellular space between two
active zones. No perisynaptic aggregates are present on the
spherical cell E-face (compare to Fig. 6). 40,000 X .
Fig. 6. A replica of the P-face of a n end bulb (El viewed
through windows in the E-face of a spherical cell (S) of tricolor guinea pig. Two active zones (asterisks) are present
on the end bulb. The spherical cell E-face surrounding these
active zones has numerous perisynaptic aggregates on the
E-face (arrowheads). 40,000 x .
SYNAPTIC MEMBRANES OF THE END BULB OF HELD
61
Fig. 7. A replica illustrating a gap junction between an
end bulb (E) and spherical cell (S)in an albino rat. The
fracture exposes the E-face of the end bulb (E) and a portion
of the spherical cell P-face. A portion of crossfracture cytoplasm containing synaptic vesicles, some of which are clustered around a dome-shaped active zone (arrow), identifies
the neuronal nature of the E-face membranes. The large
size of the terminal and the multiple circular active zones
identify it as an end bulb. A gap junction (arrowhead) is
present on the spherical cell P-face adjacent to the fractured
end bulb membrane and is illustrated at higher magnification in the inset. 21,000 x . Inset 73,000 x .
conceivably could affect the perisynaptic aggregate, spherical cells in the AVCN of brown
rats were also examined. They too lacked perisynaptic aggregates (Fig. 41, as did the spherical cells of both pigmented and unpigmented
strains of mice. The spherical cells of the albino
guinea pig have the aggregatesa2
Small gap junctions, about 0.1 pm or less in
diameter, are present between the end bulbs
and the spherical neurons of both rats and mice
(Fig. 7). These junctions were usually found
near active zones, and were only identified in
freeze-fracture replicas. Their frequency is difficult to quantitate; however, they do not appear to be more common than one for every 25
active zones identified.
DISCUSSION
% an unpublished study by Drs. E. Kane and R. Gulley, perisynaptic aggregates were identified in the bushy cells of the cat.
Perisynaptic aggregates were originally described at the spherical cell-end bulb contact
in the AVCN of chinchillas and guinea pigs
(Gulley et al., 1978a). They are also present in
the medial nucleus of the trapezoid body of the
guinea pig opposite the calyces of Held (Mattox, unpublished observations).At both of these
synapses, these aggregates are on the E-face
of the membrane opposed to channels of enlarged extracellular space surrounding the
postsynaptic active zone. The aggregates are
irregularly shaped and have a heterogeneous
population of small and medium-sized particles. Both of the synapses where perisynaptic
aggregates have been described are in auditory
brain stem nuclei, and they both have a large
area of apposition with many small active zones
surrounded by channels of enlarged extracel-
62
D.E. MATTOX, G.R. NEISES, AND R.L. GULLEY
lular space. The function of the perisynaptic
aggregates is unknown; however, based upon
their distribution in the membrane, Gulley et
al. (1978a)suggested four possibilities: 1)Ionic
pumps to maintain ionic concentrations in the
face of sustained high rates of activity, 2) extrajunctional receptors, 3) enzymes for neurotransmitter degradation, or 4) sites of transient attachment for cisterns of endoplasmic
reticulum.
Subsequent studies eliminated some of these
possibilities. After dederentation, perisynaptic aggregates are removed from the membrane (Gulley et al., 1977) but reappear when
sprouting occurs in the nonprimary boutons
(Collins and Gulley, 1979). The fact that these
aggregates disappear after deafferentation argues that they are not extrajunctional receptors for neurotransmitters since, in other systems, this type of extrajunctional receptor
increases in response to deafferentation. Either
aspartate or glutamate is the neurotransmitter for primary auditory terminals (Wenthold
and Gulley, 1978;Wenthold, 1980).Since postsynaptic enzymes are not believed to play a
major role in degrading these neurotransmitters, it seems unlikely that these aggregates
are enzymes involved in transmitter degradation. Cisternae of endoplasmic reticulum are
common in the rat where perisynaptic aggregates are not present; thus, these aggregates
are probably not exclusively sites for transient
attachment for these structures. There is evidence that the function of these aggregates may
be related to the activity of the synapse, which
might be the case if they were ionic pumps. In
the waltzing guinea pig, a genetically induced,
postnatal loss of hair cells is accompanied by
the cessation of VIII nerve firing. In these animals, the number of perisynaptic aggregates
decrease in the spherical cell membrane after
the period ofhair cell loss (Gulley et al., 1978b).
Yet, if these particles do represent structures
involved in synaptic activity, their function
must not be related strictly to VIII nerve activity, since they are not present in rats and
mice with apparently normal hearing. Thus we
are left with no compelling informationon which
to base a hypothesis on the function of the perisynaptic aggregate.
The freeze-fracture technique is a valuable
morphological tool for learning about synaptic
specializations. Nonetheless, as we learn more
about synaptic organization with this tool, we
also learn more about the limitations of the
technique. Through this process we are being
forced to rethink some of the earlier general-
izations derived from this technique (cf. Landis
and Reese, 1974 to Gulley and Reese, 1981).
The discussion of the perisynaptic aggregate
in this manuscript is an example of the devolution of ideas about function of intramembrane specialization which were based solely
on morphological observations. The fact that
these aggregates appear in the membranes of
spherical neurons of some species but not others may ultimately be an important clue as to
their function, but it also should serve as a
caution against speculating too freely about
the significance of intramembrane specializations in synaptic membranes.
ACKNOWLEDGMENTS
This work was supported by the Laboratory
of Neuro-otolaryngology,National Institute of
Neurological and Communicative Disorders and
Stroke, the National Institute of General Medical Sciences, and by grants from the Deafness
Research Foundation and NIH (NSRO115058).
We wish to acknowledge Marianne Parakkal
for her assistance in the preparation of the
manuscript.
LITERATURE CITED
Bagger-Sjoback, D. and A. Flock (1977) Freeze-fracturing
of the auditory basillar papilla in the lizard Calotes uersicolor. Cell Tiss. Res. 177:431443.
Bagger-Sjoback, D. and R.L. Gulley (1979) Synaptic structures in the type I1 hair cell in the vestibular system of
the guinea pig. Freeze-fracture and TEM study. Acta Otolaryngol. 88:401-411.
Colline, S.D. and R.L. Gulley (1979) Quantitative study of
synaptic remodeling in the cochlear nucleus following
deafferentation. Neurosci. Abstr. 5:623.
Gentshev, T. and C. Sotelo (1973) Degenerative patterns in
the ventral cochlear nucleus of the rat after primary deafferentation: an ultrastructural study. Brain Rea. 62:37-60.
Gulley, R.L., R.J. Wenthold, and G. Neises (1977) Remodeling of neuronal membranes as a n early response to deafferentation. A freeze-fracture study. J. Cell Biol.
75:837-850.
Gulley, R.L., D.M.D. Landis, and T.S. Reese (1978a) Internal
organization of membranes a t end bulbs of Held in the
rostral anteroventral cochlear nucleus. J. Comp. Neurol.
180:707-742.
Gulley, R.L., R.J. Wenthold, and G. Neises (1978b) Changes
in the synapses of spiral ganglion cells in the rostral anteroventral cochlear nucleus of the waltzing guinea pig
following hair cell loss. Brain Res. 158:279-294.
Gulley, R.L. and D. Bagger-Sjoback (1979) Freeze-fracture
studies on the synapse between the type I hair cell and
the calyceal terminal in the guinea pig vestibular system.
J. Neurocytol. 8r591-603.
Gulley, R.L. and T.S. Reese (1981) Cytoskeletal organization
a t the postsynaptic complex. J. Cell Biol. 91:298-302.
Heuser, J.E., T.S. Reese, and D.M.D. Landis (1974) Functional changes in frog neuromuscular junctions studied
with freeze-fracture.J. Neurocytol. 3:109-131.
Ibata, Y.and G.D. Pappas (1976) The fine structure of synapses in relation to the large spherical neurons in the
anteroventral cochlear nucleus of the cat. J. Neurocytol.
5:395-406.
SYNAPTIC MEMBRANES OF THE END BULB OF HELD
Korte, G.E. and J. Rosenbluth (1980)Freeze-fracture study
ofthe postsynaptic membrane ofthe cerebellar mossy fiber
synapse in the frog. J. Comp. Neurol. 193:689-700.
Landis, D.M.D. and T.S.Reese (1974)Differences in membrane structure between excitatory and inhibitory synapses in the cerebellar cortex. J. Comp. Neurol. 155:93-126.
Lenn, N.J. and T.S.Reese (1966)The fine structure of nerve
endings in the nucleus of the trapezoid body and the ventral cochlear nucleus. Am. J. Anat. 118:375-390.
Pirsig, W., M. Reinecke, and I. Lehmann (1969)Die U1trastruktur der Synapsen im ventralen Nucleus Coch-
63
leans des Meerschweinchens. Arch. klin. exp. Ohr-Nas.
Lu Kehlk Heilk 195:156-168.
Rash, J.E. and M.H. Ellisman (1974)Studies of excitable
membranes. I. Macromolecular specializations of the neuromuscular junction and non-junctional sarcolemma. J.
Cell Biol. 63567686.
Wenthold, R.J. (1980)Neurochemistry of the auditory system. Ann. Otol. Rhinol. Laryngol. 89:121-131.
Wenthold, R.J., and R.L. Gulley (1977)Aspartic and glutamic acid levels in the cochlear nucleus following auditory nerve lesion. Brain Res. 138:111-123.