OLFACTORY RECEPTORS AND NERVE TRANSMISSION
IN ROCKFISH OF THE GENUS SEBASTES
Brendan Flannery
June 9, 1989
Hopkins Marine Station Spring Course
ABSTRACT
Work on olfaction in fish is significant in two major ways.
First, characterization of effective attractants obviously would impact
on commercial fishing. Second, examining basic olfactory mechanisms in
fish could further our general understanding of sensory mechanisms in
vertebrates. We examined the olfactory system in three abundant
rockfish of the Sebastes genus in order to initiate such a program of
research. We discovered extremely slow conduction velocities of in the
olfactory nerve of 15 to 25 cm/sec, and extremely small axons, even
compared to other fish. Our studies raise questions about the
organization of these axons in the olfactory nerve. Finally, tissue
dissociation proved effective in isolating single receptor cells in
high yield. Future research on olfaction in Sebastes rockfish looks
very promising.
INTRODUCTION
Herman Kleerekoper (1969), in his extensive review of olfaction in
fish, remarked, that "the olfactory organ is the most primitive among
the sensory receptors." Yet, we do not understand the mechanisms
involved in olfaction. Such events as receptor-stimulant interaction.
neural processing and recognition of odors remain unsolved puzzles and
great possibilities exist for significant future research.
Olfaction in fish has historically been a useful system with
which to address these questions because the sense of smell is highly
developed in these animals. For example, certain species can detect
alanine and other substances at extremely low concentrations (Hara,
1975). Migratory salmon find identify home stream by the "scent" of
small concentrations of chemicals (Ueda, cited in Bardach and Villars,
1974). We can infer the importance of olfaction for fish simply from
observing the extremely large size of the olfactory organ in relation
to that of the brain, and the tremendous number of receptor cells (up
to 10 million; Doving, cited in Kleerekoper, 1969).
Despite the historical interest in fish olfaction from a
behavioral point of view and the large number of readily accessible
receptor cells, the organ in fish has received comparatively little
attention in studies focusing on transduction mechanisms in single
receptor cells. Current research techniques of particular importance
are being applied in the study of olfactory receptor cells of other
species. Patch clamp recording enables visualization of membrane
currents mediated by receptor and channel proteins (Nakamura and Gold,
1987). Biochemical and molecular research analyzes properties of the
membrane proteins themselves (Lancet, 1986). Both approaches might
benefit from a large pool of sensory cells, such as occurs in fish.
The shortage of work on fish with these newer techniques is even more
surprising in light of the great commercial importance of these
animals.
We have initiated studies designed to develop a model experimental
system for examination of transduction mechanisms in olfactory receptor
cell of a fish. The present paper describes several aspects of the
anatomy and physiology of the olfactory organ and nerve in Sebastes,
genus of Scorpionidae rockfish common in the North Pacific. The
particular advantages of study on this animal include the ready access
to the organ and nerve, the easy dissociation of single receptor cells,
and the large number of species available for comparative purposes.
METHODS AND MATERIALS
Isolation of Olfactory Organ and Nerve
We examined three species of Sebastes rockfish: chrysomelus
(black and yellow), atrovirens (kelp), and caurinus (copper). All
fish were caught with barbless hooks baited with squid in the
intertidal zone of the Monterey Bay. They ranged in length from 20 to
25 cm. Fish were kept for up to two weeks in a tank with circulating
natural sea water at ambient sea water temperature (13-15 C), and
sustained with regular feeding.
Individual fish were killed by rapid decapitation, and one eye was
removed to expose the olfactory nerve, which runs along the inside of
the orbit. The olfactory nerve was then severed where it entered the
skull and separated from surrounding connective tissue. We enlarged
the naris, freed the olfactory organ from the pigmented tissue which
held it in place in the nasal cavity, and then teased out the entire
organ and intra-orbital section of nerve through the enlarged naris.
The nerve is easily damaged during this procedure at the site where it
passes through the ocular bone which separates the nasal cavity from
the eye socket. If the extraction of organ and nerve is carried out
carefully, with a minimum of lateral pulling and without squeezing the
nerve in the forceps, the preparation obtained shows no difference in
characteristics of nerve transmission compared to those observed with
in situ preparations.
All experiments were carried out in a saline solution designed for
Pacific sanddab (Citharichthys sordidus) by Gilly (1987): 200 NaCl, 4
Cacl , 4MgCl, 3KCl, 10 Hepes (mmol/1), buffered to 7.2 with NaOH.
All chemicals were obtained from Sigma Chemical (St. Louis, MO).
Transmission Electron Microscopy
We fixed pieces of nerve and rosette in 28 glutaraldehyde in fish
saline for two hours, then washed in clean saline and postfixed in 2%
OsO in artificial sea water for one hour. Osmolarity of the
glutaraldehyde was adjusted to approximate that of the fish saline by
dilution with deionized water. After dehydration in graded ethanol.
tissue was embedded in LR White. Sections approximately 0.15 microns
thick (silver-gold) were stained with uranyl acetate and lead citrate
and viewed at an accelerating voltage of 60kV in a Phillips 201
electron microscope.
Recording Action Currents in Nerve
The method of extracellular recording is described in Gilly, Yee
and Burkhardt (1986). Glass suction electrodes ranged in diameter
from O.5mm to O.1mm, and we obtained best recordings when the nerve
precisely filled the tip of the suction electrode, thereby creating a
fairly tight seal. The stimulating electrode consisted of two Pt-Ir
wires (70 um diameter) separated by 150 um and insulated by a thin film
of silicone glue.
We calculated conduction velocities by dividing inter-electrode
distance by the time lapse between the onset of the shock artifact and
an estimate of median height of the depolarizing action current on a
storage oscilloscope. Uncertainty in determining the median height of
the first spike leads to an estimated maximum error of 10% in the
conduction velocities.
Isolation of Single Cells
Pieces of a rosette consisting of 4-5 lamellae were dissected
manually and placed in normal saline containing 5 mg/ml Protease XIV
(Sigma) for 50 minutes. The pieces of tissue were then washed in clean
saline and immediately examined under an inverted compound microscope
(4Ox objective). We also found that "natural" cell dissociation took
place when slices of the rosette were simply left in saline at room
temperature for two hours. In both cases, we dispersed cells from the
lamellae by gently shaking the tissue manually.
RESULTS
Examination of Olfactory Organ and Nerve
The rosettes removed from adult rockfish of 20-25 cm length filled
the nasal cavity underneath the incurrent and excurrent nostrils. The
rosette is slightly elliptical in shape, with 20-24 lamellae. Very
small lamellae lie at the most anterior end, and significantly larger
ones at the posterior end of the nasal cavity (Figure ja). This
anatomy agrees with that described for similar fishes by Kleerekoper
(1969).
Observation of a living rosette under a compound microscope (40x
water immersion objective) revealed motile cilia on the "top" edge of
each lamella (i.e., the edge facing the inhalant naris). Beating of
these cilia propelled fluid down each surface of the lamella towards
the center of the rosette and outwards between the lamellae. What
appeared to be receptor cells were visible on the lower half of the
flat sides of the lamellae. Whether those cells were restricted to
these areas could not be determined.
The olfactory nerve extracted with the rosette measured up to 3 cm
in length. The whole nerve consists of about 8 major bundles, each
with numerous subdivisions (Figure 1b). The tissue sheath on each
individual bundle does not bind to that surrounding adjacent bundles.
Thus, for when we cut the sheath wrapping the whole nerve, the separate
bundles would fall apart. This made it feasible to obtain a single
bundle of axons (several hundred thousand axons) several om in length.
thereby permitting stimulation and recording of action currents in a
specific group of axons. Although the dissection becomes more
difficult, each major bundle can be further "de-sheathed" to expose a
number of minor bundles or filaments, referred to collectively as "fila
olfactoria" by Kleerekoper (1969). Recording can also be done from
these smaller filaments.
Each major bundle seems to innervate one specific region of the
rosette, most likely a wedge of several lamellae. Where the major
bundles approach the bottom of the rosette, distinct branching occurs,
as depicted in Figure 1b. In some instances, one of these smaller
filaments or fila could be traced into a single lamella.
Transmission Electron Microsopy
Observations of the olfactory nerve in cross section (Figure 2)
reveal the fine structure of the axonal bundles. A Schwann cell
nucleus is visible and varying numbers of tiny axons are ensheathed by
the processes sent out by such Schwann cells. The number of Schwann
cells and their exact three-dimensional disposition could not be
determined in our observations. Each Schwann cell must ensheath
thousands of small axons, however, and the average axonal diameter is
approximately 0.05 microns (histogram, Figure 3). The extensive
shrinkage seen in the preparation makes accurate diameter estimation
impossible.
Use of the electron microscope to view epithelium will be
discussed in conjunction with the results on isolation of single cells.
Electrophysiolgy
Extracellular recordings of olfactory nerve activity were made
with a suction electrode, applied either en passant or to the cut end
of the nerve. Following a brief electric shock applied to the nerve,
we observed a bi-phasic, graded spike typical of compound activity from
a large number of axons with similar properties. (Figure 4).
Conduction velocity was very slow, as would be expected for non¬
myellinated axons of such small diameter (see below). Increasing shock
strengths showed that the greater recruitment of axons at higher
voltage did not appreciably change the time course of the response.
However, in many instances, shocks near threshold did reveal a dramatic
change in time course. This change in time course could assume either
of two representations: two distinct peaks would appear at different
thresholds, as depicted in Figure 5, or a "shoulder" would arise in the
depolarizing phase of the spike, as can be seen in Figure 6. These
phenomena suggest that more than one class of axon may exist.
Conduction velocities measured ranged from 15 to 25 cm/sec. The
variation occurred from nerve to nerve, and velocities did not depend
on distance the stimulating and recording electrode. Figure 6a shows
two records from the same nerve for inter-electrode distances of O.4
and 0.8 cm. Figure 6b shows three records from the same nerye at
distances of 0.4, 0.6 and 0.8 cm. The conduction velocity is measured
in each response as approximately 15 cm/sec. Increasing inter¬
electrode distance did cause a widening of the recorded signal, and a
slower rise to peak voltage (Figure 6).
Isolation of single sensory cells
Several cell types are visible in the material dissociated from
the sensory epithelium, as documented in Figure 7. The yield of
receptor cells from this procedure was excellent, although the protease
seemed to remove the long sensory cilia which are expected to be
associated with sensory receptors (Kleerekoper, 1969). In an attempt
to visualize ciliated sensory cells, we also observed cells shaken from
"naturally" dissociated tissue. The yield from this method was much
lower, and most of the sensory cells seen had broken off from the
tissue in the company of support cells and were not easily visualized.
Between the two methods, several specific cell types could be found
which appear to correspond well to those depicted in the schematic
drawing at the top of Figure 7 (taken from Bardach and Villars, 1974).
Basal cells did not appear in the dissociated material in any number.
Figures 8 and 9 show electron micrographs of a section through the
olfactory epithelium in a plane normal to the lamellar face. The same
type of cells seen in the dissociated material were also seen in fixed
tissue. A receptor equipped with cilia appears in Figure 8a. Basal
cells appear at the bottom of Figure 9. Figure 8b shows two receptor¬
like cells without cilia, most likely a simple result of the cutting
plane. Although only a small sample size was analyzed, we observed no
receptor cells in slices from the top of the lamellae. All of the
receptor cells were found on the side of one lamellae, facing another
lamellae. This leads to the idea that ciliated cells on top of the
lamellae may serve to move water down between the folds towards the
receptors.
DISCUSSION
Before choosing the rockfish, we compared the size of olfactory
organs in several intertidal fish. Compared to specked sanddabs
(Citharichthys stigmaeus), black surfperch (Embiotica lateralis) and
monkey-faced eel (Cebidichthys viacens), the olfactory rosette in
rockfish was the largest relative to body size. It was discovered that
in sand dabs and eel, the nasal cavity would fill with mucus when the
animal was killed. This made the dissection of these animals
difficult. Because this effect was not observed in rockfish, the
removal of the organ from rockfish was much simpler.
An interesting organization of the olfactory nerve exists which
facilitates extracellular recording. Individual bundles of nerve
fibers appear to directly innervate primarily, and perhaps solely, one
region of the rosette. This arrangement facilitates mapping of the
rosette for responsiveness when recording from one of these bundles.
and will be useful in further testing of Mozell's (1970) hypothesis
that there exist regions of different specificities on the sensory
epithelium. Further, if such spatial patterning exists, the
correspondence between specific filaments and lamellae may play an
important role in signal processing.
The observed effects of increasing inter-electrode distance on the
time course of the recorded action currents fits well with a hypothesis
that axons in most bundles are of very close, but not identical
diameter. As the distance traveled by the response increases, the
small variation in conduction velocities causes the spike to broaden.
Often, a difference in conduction time between groups of slightly
smaller or larger axons appears as a "shoulder" which is smoothed out
in the broadening peak as distance traveled increases (see Figure 6b).
Several of our results raise the question of whether there are
separable or definable fiber types in olfactory nerve. Certainly, the
appearance of two peaks or of a "shoulder" on some responses suggest
that two classes of axons exist. A variety of waveforms was recorded:
two clear spikes (Figure 5), "shoulders" (Figure 6) and very smooth
shape (Figure 4). This variation suggests that the distribution of
different diameter axons may not be uniform throughout the nerye. In
the cross-section of nerve in Figure 2 and accompanying histogram in
Figure 3, neither distinct spatial segregation nor proof for two
classes of axons can be demonstrated. Additional anatomical
measurement and a larger sample size would be necessary to solve this
point.
Packaging of fibers of larger or smaller diameters in different
Schwann cells could explain the phenomena observed in the two traces of
Figure 5. The faster, larger axons making the first spike should be
the first to depolarize, exhibiting a lower threshold, as seen in the
top trace. However, in the bottom trace, the smaller axons involved in
creating the slower spike show a lower threshold. However, we propose
this phenomena to be the result of placing the stimulating electrode
closer to a region enriched in smaller axons. The more distant
regions containing the larger axons would thus be excited only at
higher voltages. Whether the differentiation of bundles of fibers has
a specific purpose is a question which will require much future
investigation.
Unfortunately, attempts to electrically stimulate the sensory face
of the epithelium yielded no interpretable results. More careful
mapping and recording from smaller nerve bundles might yield positive
results. Recording from nerve while applying chemical stimulants to
the sensory epithelium is technically feasible with this organ, but our
limited attempts at chemical stimulation do not warrant serious
discussion.
Dissociation of single putative sensory cells from the olfactory
epithelium was very successful. The simple methods used and the
resultant high yield make it seem likely that additional technical
refinements would produce even better results. It definitely appears
that single cell recording techniques such as patch clamp should work
well for studying olfaction in fish. Furthermore, the ready access to
the large olfactory organ in rockfish and the success we had with
dissociation methods prompts us to suggest use of these rockfish for
future biochemical and molecular research on the olfactory system.
Using fish as models for olfactory research shows great promise,
not only in relation to fisheries biology and commercial fishing, but
12
E
also for elucidating basic mechanisms governing olfaction in
vertebrates.
ACKNOW LEDGMENTS
1 thank Bruce Hopkins for operating the Transmission Electron
Microscope, and Dr. Stuart Thompson for helpful advice. I appreciate
my parents' patient waiting for the conclusion of this project.
BIBLIOGRAPHY
Bardach, J.E. and Villars, Trudy. 1974. "The Chemical Senses of
Fishes."in Grant, P.T. and Mackie, A.M., eds. Chemoreception in Marine
Organisms. London: Academic Press. pp. 49-97.
Bardach, J. 1975. "Chemoreception in Aquatic Animals." in Denton, D.
and Coghlan, J., eds. Olfaction and Taste V New York: Academic Press.
pp. 121-132.
Bernard, Rudy. 1975.
"Can Taste Neuron Specificity Coexist with
Multiple Sensitivity?" in Denton, D. and Coghlan, J., eds. Olfaction
and Taste V New York: Academic Press. pp. 11-14.
Easton, Dexter. 1965. "Impulses at the Artifactual Nerve End," in Cold
Spr. Harb. Symp. quan. Biol. 30, pp. 15-27.
Gasser, Herbert. 1950. "Unmedullated Fibers Originating in Dorsal Root
Ganglia, in Journal of General Physiology. 33, pp. 651-690.
Gasser, Herbert. 1956. "Olfactory Nerve Fibers," in Journal of General
Physiology. 39, pp. 473-496.
Gesteland, Robert. 1971. "Neural Coding in Olfactory Receptor Cells,
in Beidler, Lloyd, ed. Handbook of Sensory Physiology, Volume IV:
Chemical Senses, Part I: Olfaction. Berlin: Springer-Verlag. pp. 132-
150.
Gilly, W., Yee, A. and Burkhardt, J. 1986. J. Exp. Bio. 128, 287.
Gilly, W. and Aladjem, E. 1987. Journal of Muscle Research and Cell
Motility. 8. pp.407-417.
1975. "Molecular Structure and Stimulatory
Hara, Toshiaki.
Effectiveness of Amino Acids in Fish Olfaction." in Denton, D. and
Coghlan, J., eds. Olfaction and Taste V New York: Academic Press. pp.
223-227.
Kleerekoper, Herman. 1969. Olfaction in Fishes. Bloomington: Indiana
University Press. pp. 1-98.
Lancet, Doron. 1986. "Vertebrate Olfactory Reception," in Annual Review
of Neuroscience. 9. pp. 329-55
Laverack, M. S. 1974. "The Structure and Function of Chemoreceptor
Cells." in Grant, P.T. and Mackie, A.M., eds. Chemoreception in Marine
Organisms. London: Academic Press. pp. 1-49.
Mozell, Maxwell. 1971. "Spatial and Temporal Patterning," in Beidler,
Lloyd, ed. Handbook of Sensory Physiology, Volume IV: Chemical Senses,
Part I: Olfacti n. Berlin: Springer-Verlag. pp. 205-216.
Nakamura, T. and Gold, G. 1987. "A cyclic nucleotide-gated conductance
in olfactory receptor cilia. Nature. 325, 442-444.
FIGURES
Figure 1. Drawing of olfactory organ as viewed from front and back.
From back of rosette, branching of major bundles into filaments is
apparent. Major bundles, fila olfactoria,
pigmented tissue
and orientation is labeled.
Figure 2. Cross section of Sebastes chrysomelus olfactory nerve.
10,000x. Scale: 1 micron. Schwann cell nucleus (Nu.), Schwann cell
ensheathing processes (Pr), and axons (ax.) are labelled. Shrinkage of
tissue in preparation is apparent by amount of plastic (white areas) in
picture.
Figure 3. Histogram representing distribution of axon sizes as
measured from Figure 2.
Figure 4. Typical bi-phasic, graded action current in olfactory nerve
of Sebastes atrovirens, as observed with extracellular recording.
Shocks: 20, 40, 60, 80 & 100 V with 0.4 msec duration. Interelectrode
distance: O.6 cm.
Figure 5. Responses showing two peaks. Top: Larger, faster axons
(first peak) have lower threshold while smaller axons are recruited at
greater shock strength. Shocks: 10-50V with 1Omsec (long) duration.
Inter-electrode distance: 1.2 cm. Bottom: Recruitment of fibers now
reversed. Smaller axons (second peak) show lower threshold. Shocks:
10-40 V with 0.4 msec duration. Inter-electrode distance: 0.6 cm.
Figure 6. Effect of changing inter-electrode distance on response.
Both pictures are from the same preparation. Notice appearance of
"shoulder" consistently on all spikes. Top: Two spikes represent
inter-electrode distances of 0.4 and 0.8 cm. Shocks: 45V at 0.4 cm,
45-60 V at 0.8 cm with duration of 0.4 msec. Bottom: Three spikes
represent distances of 0.4, 0.6 and 0.8 cm. Shock: 65V at all three
positions with O.4 msec duration.
Figure 7. Identification of cells dissociated from olfactory
epithelium based on drawings by Holl (cited in Bardach-Villars). a)
Original drawing by Holl. b-g) All pictures of cells dissociated with
Protease, except for e) which came from "natural" dissociation. All
cells were seen under inverted compound microscope at 600x. b)two
apparent types of receptor cells; c)elongate receptor cell; d)"cigar-
shaped" receptor cell; e)ciliated cell; f) ciliated cell with cilia
removed; g)goblet cell.
Figure 8. Detail of sensory epithelium from olfactory organ of
Sebastes chrysomelus. a) 4,500x. b) 4,500x. Resolution of receptor
cells (rc), supporting cells (sc), and ciliated cells (cc).
Figure 9. Transmission micrograph of sensory epithelium from
olfactory organ of Sebastes chrysomelus. 1,500x. Goblet cells (gc),
receptor cells (rc) and basal cells (bc) are visible.
.


Single
lamella

119

L


pigwented
fissue linig
ose
Onerser and
asal
c

Silamant
Sila
olsactore
V
maje
Burde
Figure
ft
C
25
15-
10
Fi
d d
.12
500 0,02-0.04 0.04-0.06 0.06-0.08 0.08-0.10 0.10-0.12
Axon Diameters (in microns)
ire 3
20 msec
Figure 4
O.4
40 msec
20 msec
0.2
O.
Figure
20 msec
20 msec
Figure 6
e



a



L

Ner
D
Detail of an olfactory lamella of the eel (from Holl. 1965).
(1) receptors (2) supporting cells (3) ciliated cells (4) basal cells (5) goblet cell
c)
d)

a)
Figure
25 mcros
24
2
Figure 8
104
Figure 9