Download Cultured Olfactory Interneurons From Limax maximus: Optical and

JOURNALOFNEUROPHYSIOLOGY
Vol. 69. No. 6. June 1993. Prirllccl
it1 l’..S..d.
Cultured Olfactory Interneurons From Limax maximus: Optical and
Electrophysiological Studies of Transmitter-Evoked Responses
L. D. RHINES,
Biological
SIJMMARY
P. G. SOKOLOVE,
Computation
AND
J. FLORES,
Research Department,
D. W. TANK,
CONCLUSIONS
1. The olfactory processing network in the procerebral (PC)
lobe of the terrestrial mollusk Limax maximusexhibits a coherent
oscillation of local field potential that is modulated by odor input.
To understand the cellular basis of this oscillation, we developed a
cell culture preparation of isolated PC neurons and studied the
responses of isolated cells to stimulation
with neurotransmitters
known to be present in the PC lobe.
2. The distribution
of PC soma diameters suggests at least two
diferent populations of neurons. Approximately
95% of isolated
cells had soma diameters of 7-8 pm, with the remaining cells
having larger diameters ( 1O- 15 ,um).
3. Extracellular measurements of action potentials and optical
measurements of intracellular
calcium concentrations
in fura-2loaded cells were made. Serotonin and dopamine excited PC neurons and promoted transitions from steady to bursty activity. Both
amines elicited increases in intracellular calcium, presumably concomitant with the increase in action-potential
frequency.
4. Glutamate suppressed action-potential
firing and reduced
intracellular calcium. This effect was seen most clearly when glutamate was applied to cells excited by high potassium medium.
Quisqualate
is an effective glutamate agonist in this system,
whereas kainate is not.
5. Combined with anatomic and biochemical data and with
studies of the efiects of these neurotransmitters
on the oscillating
local field potential of the intact PC network, the data from isolated PC neurons are consistent with the hypothesis that dopamine and serotonin modulate network dynamics, whereas glutamate is involved in generating the basic oscillation of local field
potential in the PC.
6. The optical studies of fura-2-loaded
cells showed that several treatments that increase the rate of action-potential
production lead to elevations in intracellular
calcium. Optical studies of
intracellular calcium may be useful for multisite measurements of
activity in the intact, oscillating PC lobe network.
INTRODUCTION
AND
A T& T Bell Laboratories,
A. GELPERIN
A4urrave Hill, New Jersev* 07974
tion on the cellular physiology and intrinsic circuitry of
these oscillating networks will greatly aid attempts to determine the computational role of the collective network oscillations.
We are investigating cellular mechanisms and computational usesof coherent network oscillations with the useof
the procerebrum (PC) of the terrestrial gastropod Limax
maximus. The PC circuit contains of order lo4 local interneurons that receive direct input from the superior and inferior noses at the ends of the tentacles. The PC displays a
prominent 0.7-Hz oscillation in its LFP that is produced by
circuitry intrinsic to the PC (Gelperin and Tank 1990).
Structural studies of the synaptic zone of the PC clearly
indicate the importance of local circuit interactions based
on the density and reciprocal nature of the synaptic contacts (Chase and Tolloczko 1989; Zs.-Nagy and Sakharov
1970; S. Curtis, personal communication). Anatomically,
the PC is a very prominent part of the cerebral ganglion,
reflecting the dominant role of olfaction in the behavior of
Limax and Lima-x’s highly developed odor-learning ability
(Gelperin 1989; Sahley 1990). To characterize the basic
electrophysiology and transmitter sensitivity of PC neurons, we developed a cell culture preparation of isolated PC
neurons and measured their spontaneous electrical activity
and responsesto neurotransmitters known to be present in
the PC lobe.
Optical recordings provide an effective means to study
patterns of electrical activity in distributed networks (Grinvald 1985; Regehr et al. 1989). To implement optical recordings from the PC lobe circuit during odor processing,
an optical probe that reports on cell activity but that produces negligible alteration of cell activity is required. The
experiments reported here explore the utility of intracellular calcium measurements with the use of fura- for monitoring action-potential activity in individual PC lobe neurons and for eventual multisite, noninvasive monitoring of
PC lobe circuit function.
Recognition and classification of odors is thought to depend on distributed processing in collective neural networks in both mammalian and molluscan nervous systems.
METHODS
The mammalian olfactory bulb displays collective behavior
in the form of a coherent 40. to 80-Hz oscillation in local Cell cult we
field potential (LFP). Odors elicit unique patterns of spaPC neurons were harvested by anesthetizing slugs (4-6 gm) with
tial variation in the amplitude of the LFP oscillation decold
(4°C) for 30 min and then removing the brain and buccal
pending on the particular odor stimulus and the animal’s
mass
to sterile saline containing 0.1% penicillin /streptomycin
and
behavioral state (Baird 1986; Freeman and Skarda 1985;
1% amphotericin
B (S-A). Saline composition
was as follows (in
Skinner et al. 1990). The procerebral (PC) lobe of gastro- mM): 55.4 Na’, 4.2 K+, 7.0 Ca2+, 4.6 Mg2+, 80.1 Cl-, 0.2
pod mollusks, like the olfactory bulb of mammals, receives H,PO,, 2.5 HCO,, 5.0 g 1ucose, 10 IV-2-hydroxyethylpiperazinedirect olfactory input and also shows a coherent oscillation
IV’-2-ethanesulfonic
acid ( HEPES), pH 7.6. Sterile conditions
of its LFP that is modulated in frequency and waveform by were used in all subsequent steps. Transfer pipettes were presoaked in saline containing 45 mg/ml bovine serum albumin (Sodor input (Gelperin and Tank 1990). Detailed informa1940
0022-3077/93 $2.00 Copyright 0 1993 The American Physiological Society
OPTICAL
AND ELECTRICAL
STUDIES
OF OLFACTORY
INTERNEURONS
1941
11PA
4 set
FIG.
3. Loose-patch whole-cell recording from an isolated cultured PC
neuronto illustratethe levelof spontaneous
activity typicallyobserved
in
thesecellsin culture.
times with S-BSA at room temperatureand transferredto a 3-ml
bottle with 1.5 ml of salinecontaining 10mg/ml proteasetype IX
(Sigma; S-P). PCs were incubated in S-P at 35°C for 1 h, then
washedfour times with S-BSA and transferredto a 1.5-m] tube
containinga Teflon-coatedstirring bar and 0.25 ml of S-BSA.PCs
were stirred at 4°C until tissuedispersionwascomplete,generally
I 1h. At that point, 0.8 ml salinewasaddedto the cell suspension,
and two drops of cell suspensionwere placed in the center of a
polylysine-coatedcoverslip. After standing undisturbedfor 2 h,
3.5 ml of S-BSA was added, and the chamberscontaining the
coverslipswere allowed to stand undisturbedovernight at room
temperature. For someexperimentscell density wasincreasedby
placingthe cell suspensioninsidea 6-mm (ID) glassring resting
on the polylysine substrate.The glassring wasremovedafter cell
attachment. Cellswere incubated at room temperaturein S-BSA
for up to 20 days.
Electrophysiological
recording
The coverslipbearingthe PC neuronswassealedto the bottom
of a Plexiglaschamber that was then filled with normal saline.
Cellswere visualizedwith the useof a ZeissIM35 inverted microscope.Patch electrodeswerebrought to the surfaceof the cell, and
FIG. 1. Photomicrograph of a typxal 7-day culture of neurons isolated
gentle suction wasapplied to obtain a whole-cell,loose-patchrefrom the procerebrum of Limus maximus
The cells have regrown numerous processes in a culture medium containing only salts, buffer, glucose, cording. Membrane current recordingswere obtainedwith a List
EPC7 amplifier and displayedwith the useof a Nicollet model
and bovine serum albumin. Scale bar, 20 Nrn.
2090oscilloscopeanda Gould 220chart recorder.Neurotransmitterswereappliedto cellsby applying brief( OS-5 s) pressurepulses
BSA). Cerebralgangliawith PC lobesattachedweretransferredto (Picospritzer) to an electrode filled with neurotransmitter and
S-BSAat 4°C and PC lobeswereseparatedfrom cerebralganglia 0.1% Fast Green. The cell chamber was continuously perfused
and desheathed.Isolated PC lobesin S-BSA were washedthree during the experiment, and the puffer electrodewaslocated upstreamof the neuron under study. Visualization of the cloud of
Fast Green moving over the cell allowed determination of the
approximate duration of neurotransmitter exposure.Control experimentswereconductedto assess
responses
to FastGreenalone.
A
I IIII II
1111 lltl.
Id III
III II
CELL DIAMETER,
FIG. 2. Histogram of cell body diameters of procerebral (PC) neurons
grown in cultures like that shown in Fig. I. The major size class is 6-9 pm
with a small number of cells in the range of IO- 13 pm soma diameter. n =
233.
I
I
urn
I
I
I I
I/ I
I
I
I II
Ill
II
I I
t
4. Briefpulsesof dopamineactivatePCneuronsin culture.A: a
quiescent
cell is transformedinto a burstycell by a 3-spuff of IO-“ M
dopamine.
B: cellshowing
slowsteadyactivity istransformed
intoabursty
cellby a 3-spuff of lo-” M dopamine.
FIG.
1942
RHINES,
SOKOLOVE,
FLORES,
TANK,
AND
GELPERIN
FIG. 5.
Prolonged dopamine application ( 15 s, 10m4M) produces a brief period of inhibition followed by a period of
generating spike clusters and bursts. Dopamine excitation is followed by a period of decreased activity.
For each neurotransmitter,
15-20 cells were tested, each cell being
used for tests of a single neurotransmitter.
Desensitization
effects
were minimized by giving brief pulses of transmitter to cells being
continually perfused with normal saline and moving up the saline
perfusion stream as successive cells were sampled. Neurotransmitters were used at the following concentrations
in the puffer electrode: dopamine ( 100 PM with 1 mg/ml ascorbate), serotonin
( 100 PM), FMRFamide
( 10 PM), small cardioactive peptide B
(SCP,; 10 PM), and glutamate ( 100 PM). Cells were also activated by puffs of five times normal (20 mM) potassium saline. In
some experiments cells were exposed to neurotransmitters
added
to the saline used to perfuse the experimental chamber.
Opt ical recording
Measurements of free calcium levels in individual PC neurons
were made on cells loaded with fura- by the use of the membrane
permeant acetoxymethylester
form of the probe. Cells were labeled by transferring the coverslip bearing the PC neurons to a
solution of 5 PM fura-2AM for 12 min and then returning the cells
to normal saline. Images were collected with a cooled charge-coupled device (CCD) camera (model 220, Photometrics)
with computer (Mac II, Apple Computer) controlled image acquisition and
display, as described previously (Regehr and Tank 1992). Intracellular calcium was determined from the ratio of images at 340
and 380 nm excitation (Grynkiewicz
et al. 1985).
Each neurotransmitter
was tested on at least three different cultures 1- 10 days after isolation. Typically, two image sequences
were obtained from each culture. The series of digital images taken
before, during, and after agonist application was processed to obtain the 340/380 ratio for groups of 15-20 cells during a pretreatment baseline period and throughout the response to neurotrans-
FIG. 6. Serotonin activates isolated PC neurons. A: 500-ms puff of
10M4 M serotonin directly onto the PC neuron elicits a brief period of
activation. B: 800-ms puff of 10e4 M serotonin on a cell connected by a
process to the recorded cell produces activation with a delav.
mitter, which was applied from a puffer electrode in a continuous
flow of saline. The second series of optical records from a given
culture was obtained from cells upstream of the first transmitter
application. The responses of 1-day and 5- to 1O-day cultures were
similar, so the data have been grouped.
RESULTS
PC neurons isolated by the method described above attached to the polylysine substrate readily, and many cells
initiated growth of new processes within 12 h. The cells
grew well for several weeks in a culture medium consisting
only of S-BSA. A typical group of PC neurons after 7 days
in culture is shown in Fig. 1. A histogram ofcell body diameters measured from projected images of micrographs is
shown in Fig. 2. Two size classes are present. Cell bodies
6-9 pm diam (7.1 t 0.5, mean t SD, n = 220) form the
most prevalent size class (94.4%). The other class is composed of a smaller group of cells of larger size with soma
diameters 9- 15 pm ( 10.4 t 1.2, y1= 13). Measurements on
a second population of cultured cells gave similar results,
with the size class from 6 to 10.5 pm (mean of 7.97 t 0.78,
y2= 554) forming 95.7% of the population and the size class
from 10.5 to 14.5 pm (mean of 12.1 t 1.4, n = 25) comprising 4.3% of the population. A small group of cells significantly larger ( lo- 15 pm) than the major size class ( 7-9 pm)
FIG. 7. Glutamate application inhibits isolated PC neurons. Two l-s
puffs of 1O-4 M glutamate were delivered to a cell showing an elevated rate
of spontaneous firing.
OPTICAL
AND
ELECTRICAL
STUDIES
OF
OLFACTORY
INTERNEURONS
1943
90% (39 of 43 cells in 1 set of 5- to lo-day-old cultures).
The pattern of activity was irregular with no cells showing
regular bursting. About 10% of the sampled cells were in the
large ( lo- 15 pm) size class, however, these large cells
showed the same pattern of spontaneous activity and transmitter responses as the smaller, much more numerous size
class. Figure 3 shows a typical record to indicate the signalto-noise ratio obtained in these recordings and the variability in interspike interval observed during spontaneous activity of PC neurons in culture. Occasional cells were quiescent unless activated, for example by a brief application of
high-K+ saline. Some recordings showed more than one
size class of action potential, probably reflecting activity in
a neuronal soma and one or more processes from other cells
monitored
simultaneously,
as reported by Forda et al.
800
600
(1982).
HI-K
T
800
700
z
r
g
0
;
e
ml
3
J
ti
0
POST
600
500
400
300
d
100
0
PRE
DA
POST
11: nonspecific
excitation
by high potassium medium (2.5 times
normal)
greatly elevates internal calcium
concentration
in cultured
PC
neurons
loaded with fura-2AM.
Means and standard
deviations
are
shown. B: application
of 1W4 M dopamine
cxcitcs PC neurons and greatly
elevates internal calcium.
Measurement
of the dopamine
response was
made at the peak of the calcium elevation.
FIG. 8.
Dopamine application with puff duration of 3 s activated
quiescent cells (Fig. 4A) or changed the pattern of firing in
spontaneously active cells from slow, single spikes at irregular intervals to a distinctly bursty pattern (Fig. 4B). In
some cases the dopamine application produced a brief period of inactivity before emergence of the bursty mode (Fig.
5 ). The interspike intervals shown in Fig. 5 indicate that PC
neurons are capable of producing action potentials at rates
as high as 20 spikes/s, at least for brief periods. This is also
evident in the spike clusters of Fig. 4B, which show apparent spike amplitude reduction at the highest firing rates.
Serotonin application directly to the recorded neuron
elicited excitation with clustered action potentials (Fig.
6A ). In some situations the serotonin puff was delivered to
cells located downstream from the recorded cell but connected to the recorded cell by a clearly visible process. In the
example shown (Fig. 6 B) a 0.8-s puff on the downstream
cells produced a long-lasting train in the recorded cell with
no sign of spike clustering. This may represent indirect activation of the recorded cell by synaptic activation from the
downstream presynaptic cells, but we cannot exclude the
possibility that a remote process from the recorded cell was
activated directly.
Results with direct application
of the tetrapeptide
FMRFamide
( PheMetArgPhe-amide)
and SCP, were weak
and inconsistent at the puffer concentration tested ( 10
PM). Some cells showed very weak excitation, whereas
others showed weak inhibition.
Control puffs with only
Fast Green in the electrode showed small effects of the same
tY Pee
was also seen in paraffin-embedded serial sections of the PC
lobe stained with hematoxylin and eosin. Some members of
the larger size class may be derived from the scattered small
clusters of 15-pm FMRFamide-like
immunoreactive
cells
known to be present in the PC lobe (Cooke and Gelperin
1988).
Whole-cell loose-patch recordings from PC neurons in 5to lo-day cultures showed that many of the cells generate
action potentials spontaneously at an average rate of 0.8 t
0.07/s (mean t SE, 18 active cells). The proportion of
sampled cells showing spontaneous action potentials before
stimulation with neurotransmitter
was typically close to
Glutamate application consistently inhibited action-potential production by PC cells. Inhibition was clear in cells
with an elevated level of spontaneous firing ( Fig. 7) or in
cells activated by exposure to high-K+ saline (data not
shown). Figure 7 shows two trials of glutamate application
to the same neuron to indicate the reproducibility
of the
effects reported here. The 0.5-s puffs of glutamate resulted
in cessation of activity for 5-8 s, with no evidence for postinhibitory rebound.
Culcium imuging
Cultured PC neurons take up fura-2AM readily. The fura
loading protocol used in these experiments results in internal concentrations of fura- that are high enough so that
OPTICAL
AND
ELECTRICAL
STUDIES
digital images can be obtained with modest exposure times
but low enough that free [ Ca]i levels are not altered by the
fura-2, as judged by the ability of the cells to maintain norma1 patterns of spontaneous activity and resting calcium
PRE
w
20001800"
z
c
1600-
8
L
1400-
9
1200-
s
IOOO-
ii
800
2
u
600-
5
400:
200
K+GLU
HI-K
POST
-
.
.
4
.
-
.
-
PRE
C
HI-K
HI-K
KNATE
HI-K
POST
1400
1200
z
c
*
1000
lo.
2
800
d
i
600
z
E
400
200
PRE
HI-K
QUIS
HI-K
POST
1
OF
OLFACTORY
INTERNEURONS
1945
levels in the 50- to lOO-nM range for several hours. If the
cells are activated by high potassium medium ( 10 mM, 2.5
times normal), the great majority of the cells respond with
increases in [ Ca]i into the 700- to 900-nM range with recovery of normal calcium levels after return to normal saline.
This effect was quantified by measuring the 340/380-nm
ratio and converting this to [ Ca]i for a number of individual PC cells before, during peak response to, and after recovery from exposure to high potassium medium. The results
of these measurements are shown in Fig. 8A. Activation of
action-potential
production by the nonspecific stimulus of
high [K], leads to a dramatic rise in [ Cali that is fully reversible.
Dopamine application also leads to a dramatic increase
in [ Cali, as shown in Fig. 8 B. The measurements were
taken before, at the peak of, and after recovery from the
application of 100 PM dopamine. The digital images analyzed in Fig. 8 B were not obtained with the temporal resolution necessary to study changes in the pattern of spike production evident in the records of Figs. 4 and 5, however, the
increase in average rate of spike production caused by dopamine is more than sufficient to account for the increase
in [ Ca]i evident in Fig. 8B.
Recent improvements
in the imaging system software
have allowed rapid sequences of images to be obtained.
These digital “movies” were used to study cyclic changes in
[ Ca]i in dopamine-stimulated
cells that presumably reflect
activity level changes like those shown in Figs. 4 and 5.
Figure 9 shows such a movie. Several neurons show fluctuating levels of [Cal, rather than a simple sustained increase. This culture was 7 days old and therefore contained
processes between cells and perhaps functional synapses.
For this reason the responses shown cannot rigorously be
partitioned between network effects and effects endogenous
to the cell showing the changes in [Cal+
The effects of serotonin on [ Ca]i are more variable from
cell to cell than the effects of dopamine. Some cells respond
immediately to a puff of serotonin, whereas others respond
with a delay or show no response at all (data not shown).
For those cells that respond to serotonin with an increase in
[Cal;, the response is long lasting, often persisting long after
complete washout of the serotonin. This difference between
dopamine and serotonin in time course of calcium response
parallels the difference in time course of electrophysiological responses to these two amines (Gelperin et al. 1992).
Glutamate suppresses action-potential
production in PC
neurons (cf. Fig. 7) and leads to a decrease in [ Cali. Although glutamate lowers [ Ca]i in resting unstimulated PC
neurons, the effects of glutamate are seen most clearly
against a background of enhanced activity elicited by elevated [K], . The experiment involves taking five pairs of
FIG. 10. A : glutamate application ( I Oe4 M ) suppresses the increase in
intracellular calcium elicited by high potassium medium. B: kainate ( lop3
M ) is not an effective glutamate agonist because kainate was not effective
in reversing the elevation in intracellular calcium caused by high potassium medium. As shown by the postkainate high potassium response, the
decrease in internal calcium during kainate application is due to accommodation to the continued presence of the high potassium stimulation. C:
quisqualate ( 1Oe3 M ) is an effective glutamate agonist as shown by its
ability to suppress the internal calcium increase elicited by high potassium
medium.
1946
RHINES,
SOKOLOVE,
FLORES,
images successively, namely during I ) prestimulation
baseline, 2) high [ K],, 3) high [ K], + agonist, 4) high [K],
recovery, and 5) posttreatment baseline. Figure 1OA shows
the results of such an experiment using 100 PM glutamate.
Glutamate inhibition of high [ K],-stimulated
cells reduces
[Cali to the level found in resting unstimulated neurons.
The same experimental protocol was used to test the effects
of 1 mM kainate (Fig. 1OB) and 1 mM quisqualate (Fig.
1OC). Kainate is not an effective glutamate agonist in this
system, whereas quisqualate is a very effective glutamate
agonist. The quisqualate-type glutamate receptor in this
system is not affected by 6-cyano-7-nitroquinoxaline-2,3dione (CNQX; data not shown).
DISCUSSION
The network of local interneurons in the PC lobe of Limax displays complex dynamics that are responsive to input elicited by natural odors applied to the olfactory neuroepithelium at the tip of the superior tentacle (Gelperin and
Tank 1990). To understand the genesis of the network dynamics, it is necessary to understand the cellular and synaptic properties of the individual procerebral interneurons.
These olfactory interneurons are now shown to generate
action potentials, and the nature of their responsiveness to
several of the neurotransmitters
endogenous to the Limax
PC lobe is demonstrated.
In mammalian olfactory systems the necessity of modulatory input from outside the olfactory bulb for behavioral
modification of olfactory bulb electrophysiology is well documented (Gervais et al. 1988; Gray et al. 1986). We have
shown that the coherent network oscillation in LFP displayed by the PC is modulated in frequency and waveform
by dopamine and serotonin (Gelperin et al. 1992). The
data on cellular responses and network responses to these
two amines combine to support the hypothesis that these
amines change the pattern of firing of PC neurons and
hence the dynamics of the PC network. In Limax the aminergic modulatory input may be engaged by the unconditioned stimuli known to be effective in mediating odoraversion learning, such as quinidine sulfate, COZ gas, and
electric shock (Gelperin 1975; Sahley 1990). In Aplysia several lines of evidence indicate that the tail shock unconditioned stimulus used in associative conditioning of the gill
and siphon withdrawal response can activate a serotonergic
pathway (Glanzman et al. 1989; Hawkins 1989; Mackey et
al. 1989).
The PC neurons are remarkable for their ability to survive several weeks, extend extensive new processes, and
form new synapses in a medium containing only salts, glucose, and BSA. The PC cells are also notable for their very
high nucleus to cytoplasm ratio and for being uniformly
diploid, in contrast to the polyploidy of many other central
molluscan neurons (Chase and Tolloczko 1987). We did
not study in a systematic way the ability of cultured PC cells
to form chemical synapses in vitro, although several of our
measurements strongly suggest this possibility. Other systems of cultured molluscan neurons have shown these cells
capable of recreating aspects of their in vivo connectivity in
vitro (Kleinfeld et al. 1990; Syed et al. 1990). By plating
Limax PC neurons at even higher densities than used in
TANK,
AND
GELPERIN
this study and allowing more time for synaptic reconnection, PC cell aggregates displaying a coherent network oscillation may be produced. The axosomatic synapses made in
the apical cell body layer of the intact PC are adequate to
mediate the coherent network oscillation, as demonstrated
by the fact that the isolated apical cell layer devoid of its
adjacent neuropil still shows the network oscillation (Gelperin et al. 1992).
In addition to dopamine, serotonin, and glutamate, the
PC lobe contains neurons immunoreactive
for two peptide
neurotransmitters:
FMRFamide
(Cooke and Gelperin
1988) and SCP, (A. Gelperin and J. Flores, unpublished
observations). Measurements of peptide effects on PC lobe
LFP oscillation frequency or amplitude do not reveal a
clear effect of FMRFamide
or SCP, at the concentrations
used in this study (Gelperin et al. 1992), although concentrations lo- 100 times higher do produce effects on the PC
lobe oscillation (Flores and Gelperin, unpublished observations). Five members of the FMRFamide-related
peptide
family have been isolated from extracts of whole Limax
(Krajniak et al. 1989) thus all five are candidate neurotransmitters in the PC lobe. FMRFamide
and SCP, affect
rhythmic motor networks in the Limax feeding system
(Cooke et al. 1985; Prior and Watson 1988), increasing the
likelihood that some members of these two peptide families
have modulatory effects on the PC lobe network. Studies of
ion channel modulation in the tactile sensory cells of Aplysia show that FMRFamide
augments the S-type potassium
channel (Abrams et al. 1984; Sweatt and Kandel 1989) and
blocks a rapidly inactivating calcium channel (Edmonds et
al. 1990). SCP, has been shown to increase adenosine 3’,5’cyclic monophosphate levels in PC neurons (Yamane and
Gelperin 1987 ) .
The studies reported here show that fura- is a well-tolerated probe whose fluorescence can monitor average rates of
action-potential
production. Because a number of studies
in mammalian
olfactory bulb show the existence of spatial
patterns of oscillation amplitude variation correlated with
odor detection ( Bressler 1988; Skinner et al. 1990), we wish
to make such multisite measurements in the Limax PC
lobe. Improvements
in techniques for labeling substantial
numbers of neurons in the mammalian
slice preparation
with fura- (Regehr and Tank 199 1) may facilitate complete labeling of the intact PC lobe network to enable optical recording of the coherent PC oscillation and its modulation by odor input. Recent multisite optical recordings
from the Limax PC lobe stained with the voltage-sensitive
dyes RH 155 (Fee et al. 1992) and Di-4-ANEPPS (Gelperin
et al. 1992) support the view that odor input to the PC lobe
results in both spatial and temporal patterns of network
modulation.
Present address: L. D. Rhines, Harvard Medical School; P. G. Sokolove,
Dept. of Biological Sciences, University of Maryland Baltimore County.
Address for reprint requests: A. Gelperin, Rm lC464, AT&T Bell Labs.,
600 Mountain Ave., Murray Hill, NJ 07974.
Received 26 May 1992; accepted in final form 17 December 1992.
REFERENCES
ABRAMS, T. W., CASTELLUCCI,
V. F., CAMARDO,
J. S., KANDEL, E. R.,
AND LLOYD, P. E. Two endogenous neuropeptides modulate the gill and
OPTICAL
AND ELECTRICAL
STUDIES
siphon withdrawal reflex in Aplysia by presynaptic facilitation involving
CAMP-dependent closure of a serotonin-sensitive potassium channel.
Proc. Natl. Acad. Sci. USA 8 1: 7956-7960, 1984.
BAIRD,
B. Nonlinear dynamics of pattern formation and pattern recognition in the rabbit olfactory bulb. Physica 22D: 150-175, 1986.
BRESSLER,
S. L. Changes in electrical activity of rabbit olfactory bulb and
cortex to conditioned odor stimulation. Behav. Neurosci. 102: 740-747,
1988.
CHASE, R. AND TOLLOCZKO,
B. Evidence for differential DNA endoreplication during the development of a molluscan brain. J. Neurobiol. 18:
395-406, 1987.
CHASE,
R. AND TOLLOCZKO,
B. Interganghonic dendrites constitute an
output pathway from the procerebrum of the snail Achatinafulica. J.
Comp. Neural. 283: 143-152, 1989.
COOKE, I. AND GELPERIN,
A. Distribution of FMRFamide-like immunoreactivity in the nervous system of the slug Limax maximus. Cell Tissue
Res. 253: 69-76, 1988.
COOKE, I. R. C., DELANEY,
K., AND GELPERIN,
A. Complex computation
in a small neural network. In: Memory Systems of the Brain, edited by
N. M. Weinberger, J. L. McGaugh, and G. Lynch. New York: Guilford,
1985, p. 173-191.
EDMONDS,
B., KLEIN, M., DALE, N., AND KANDEL,
E. R. Contributions of
two types of calcium channels to synaptic transmission and plasticity.
Science Wash. DC 250: 1142- 1147, 1990.
FEE, M. S., DELANEY,
K. R., GELPERIN,
A., STEPNOSKI,
R. A., TANK,
D. W., AND KLEINFELD,
D. Spatiotemporal variations in the activity of
an oscillating neuronal network in the terrestrial mollusc Limax (Abstract ). Biophys. J. 6 1: A5 18, 1992.
FORDA,
S. R., JESSELL, T. M., KELLY, J. S., AND RAND, R. P. Use of the
patch electrode for sensitive high resolution extracellular recording.
Brain Res. 249: 37 l-378, 1982.
FREEMAN,
W. J. AND SKARDA, C. A. Spatial EEG patterns, nonlinear dynamics and perception: the neo-Sherringtonian view. Brain Res. Rev.
10: 147-175, 1985.
GELPERIN,
A. Rapid food-aversion learning by a terrestrial mollusk.
Science Wash. DC 189: 567-570, 1975.
GELPERIN,
A. Neurons and networks for learning about odors. In: Perspectives in Neural Systems and Behavior, edited by T. Carew and D. Kelley.
New York: Liss, 1989, p. 12 I- 136.
GELPERIN,
A., GERVAIS, R., DELANEY,
K., FEE, M. S., FLORES, J., TANK,
D. W., AND KLEINFELD,
D. Olfactory interneuron oscillations: optical
and electrical measurements of intrinsic and odor-modulated activity.
Sot. Neurosci. Ahstr. 18: 303, 1992.
GELPERIN,
A., RHINES, L., FLORES, J., AND TANK, D. Coherent network
oscillations by olfactory interneurons: modulation by endogenous
amines. J. Neurophysiol. 69: 1930- 1939, 1993.
GELPERIN,
A. AND TANK, D. W. Odor-modulated collective network oscillations of olfactory interneurons in a terrestrial mollusc. Nature Lond.
345: 437-440, 1990.
GERVAIS,
R., HOLLEY, A., AND KERVERNE,
B. The importance of central
noradrenergic influences on the olfactory bulb in the processing of
learned olfactory cues. Chem. Senses 13: 3- 12, 1988.
GLANZMAN,
D. L., MACKEY,
S. L., HAWKINS,
R. D., DYKE, A. M.,
LLOYD,
P. E., AND KANDEL,
E. R. Depletion of serotonin in the nervous
system of Aplysia reduces the behavioral enhancement of gill withdrawal as well as the heterosynaptic facilitation produced by tail shock.
J. Neurosci. 9: 4200-42 13, 1989.
GRAY,
C. M., FREEMAN,
W. J., AND SKINNER, J. E. Chemical dependen-
OF OLFACTORY
INTERNEURONS
1947
ties of learning in the rabbit olfactory bulb: acquisition of the transient
spatial pattern change depends on norepinephrine. Behav. Neurosci.
100: 585-596, 1986.
GRINVALD,
A. Real-time optical mapping of neuronal activity: from single
growth cones to the intact mammalian brain. Annu. Rev. Neurosci. 8:
263-305, 1985.
GRYNKIEWICZ,
G., POENIE, M., AND TSIEN, R. Y. A new generation of
Ca2+ indicators with greatly improved fluorescence properties. J. Biol.
Chem. 260: 3440-3450, 1985.
HAWKINS,
R. D. Localization of potential serotonergic facilitator neurons
in Aplysia by glyoxylic acid histofluorescence combined with retrograde
fluorescent labeling. J. Neurosci. 9: 42 14-4226, 1989.
KLEINFELD,
D., PARSONS,
T. D., RACCUIA-BEHLING,
F., SALZBERG,
B. M., AND OBAID, A. L. Foreign connections are formed in vitro by
Aplysia califirnica interneurone L 10 and its in vivo followers and nonfollowers. J. Exp. Biol. 154: 237-255, 1990.
KRAJNIAK,
K. G., GREENBERG,
M. J., PRICE, D. A., DOBLE, K. E., AND
LEE,
T. D. The identification, localization and pharmacology of
FMRFamide-related peptides and SCP, in the penis and crop of the
terrestrial slug, Limax maximus. Comp. Biochem. Physiol. C Comp.
Pharmacol. 94: 485-492, 1989.
MACKEY,
S. L., KANDEL,
E. R., AND HAWKINS,
R. D. Identified serotonergic neurons LCB 1 and RCB 1 in the cerebral ganglia of Aplysia produce
presynaptic facilitation of siphon sensory neurons. J. Neurosci. 9: 42274235, 1989.
PRIOR,
D. J. AND WATSON,
W. H. The molluscan neuropeptide SCP,
increases the responsiveness of the feeding motor program in Limax
maximus. J. Neurobiol. 19: 87- 105, 1988.
REGEHR,
W. G., CONNOR,
J. A., AND TANK, D. W. Optical imaging of
calcium accumulation in hippocampal pyramidal cells during synaptic
activation. Nature Land. 341: 533-536, 1989.
REGEHR,
W. G. AND TANK, D. W. Selective fura- loading of presynaptic
terminals and nerve cell processes by local perfusion in mammalian
brain slice. J. Neurosci. Methods 37: 1 1I- 1 19, 199 1.
REGEHR,
W. G. AND TANK, D. W. Calcium concentration dynamics produced by synaptic activation of CA1 hippocampal pyramidal cells. J.
Neurosci. 12: 4202-4223, 1992.
SAHLEY,
C. L. The behavioral analysis of associative learning in the terrestrial mollusc Limax maximus: the importance of interevent relationships. In: Connectionist Modeling and Brain Function: The Developing
Interjke, edited by S. Hanson and C. Olson. Cambridge, MA: MIT
Press, 1990, p. 36-73.
SKINNER,
J. E., MARTIN,
J. L., LANDISMAN,
C. E., MOMMER,
M. M.,
FULTON,
K., MITRA, M., BURTON, W. D., AND SALTZBERG,
B. Chaotic
attractors in a model of neocortex: dimensionalities of olfactory bulb
surface potentials are spatially uniform and event related. In: Chaos in
Brain Function, edited by E. Basar. Berlin: Springer-Verlag, 1990, p.
119-134.
SWEATT,
J. D. AND KANDEL,
E. R. Persistent and transcriptionally-dependent increase in protein phosphorylation in long-term facilitation of
Aplysia sensory neurons. Nature Land. 339: 5 l-54, 1989.
SYED, N. I., BULLOCH,
A. G. M., AND LUKOWIAK,
K. In vitro reconstruction of the respiratory central pattern generator of the mollusk
Lymnaea. Science Wash. DC 250: 282-285, 1990.
YAMANE,
T. AND GELPERIN,
A. Aminergic and peptidergic amplification
of intracellular cyclic AMP levels in a molluscan neural network. Cell.
Mol. Neurobiol. 7: 29 l-30 1, 1987.
ZS.-NAGY,
I. AND SAKHAROV,
D. A. The fine structure of the procerebrum
of pulmonate molluscs, IIelix and Limax. Tissue Cell 2: 399-4 11, 1970.