bastian & nguyenkim (2001) dendritic modulation of burst-like firing in sensory neurons

8/11/2019 Bastian & Nguyenkim (2001) Dendritic Modulation of Burst-Like Firing in Sensory Neurons

1/13


2/13

discharge resulting from the presence of objects in the animalsenvironment that have an impedance different from that of thesurrounding water. Electrocommunication stimuli also consistof changes in EOD amplitude and timing, but these result fromthe interaction of the EOD of an individual with that ofconspecifics, and these are typically spatially extensive influ-encing the activity of electroreceptors over large regions of thebody surface (see Turner et al. 1999 for recent reviews ofelectroreception). Electroreceptor afferents terminate within amedullary nucleus, the ELL, and the P-type or amplitudeencoding afferents provide excitatory synaptic input to basilarpyramidal cells (E cells). A second category of pyramidal cells,nonbasilar or I cells, are driven by inhibitory interneurons thatreceive excitatory afferent input (Maler 1979; Maler et al.1981; Saunders and Bastian 1984).

In addition to receiving electroreceptor afferent inputs, ELLpyramidal cells receive, via elaborate apical dendrites, massivesynaptic input descending from higher centers (Maler et al.1981; Sas and Maler 1983, 1987). One subdivision of thedescending pathways is thought to be involved in gain control(Bastian 1986a,b; Bastian and Bratton 1990; Nelson 1994)while another may provide positive feedback to accentuate

important stimulus features (Berman and Maler 1999; Bermanet al. 1997; Bratton and Bastian 1990; Maler and Mugnaini1994). Recent studies have also demonstrated robust synapticplasticity at the pyramidal cells apical dendrites (Bastian1999; Bell et al. 1997).

Studies in which single ELL pyramidal cells were intracel-lularly labeled and reconstructed showed that several physio-logical characteristics of these neurons were strongly corre-lated with neuronal morphology (Bastian and Courtright 1991).Apical dendritic size is highly variable, and both spontaneousfiring frequency as well as rate of adaptation to changes instimulus amplitude are negatively correlated with the size of acells apical dendritic arbor. This report extends these obser-vations showing that pyramidal cells are also highly variable in

terms of their tendency to produce bursts of action potentials,that apical dendritic size is strongly correlated with a cellstendency to burst, and that pharmacological reduction of de-scending excitation to these dendrites greatly reduces burstyfiring.

M E T H O D S

The South American weakly electric fish A. leptorhynchus wasexclusively used in these studies. Animals were housed in 50-gallonpopulation tanks at 2628C and with water conductivity rangingfrom 200 to 400 S. Experiments were done in a 39 44 12-cm-deep experimental tank containing water from the animalshome tank. Surgical techniques were the same as previously described(Bastian 1996a,b) and all procedures were in accordance with theUniversity of Oklahomas animal care and use guidelines.

Recording and stimulation

Extracellular recordings were made with metal-filled micropipettesconstructed as described by Frank and Becker (1964). Intracellularrecordings were made with borosilicate or aluminosilacate sharp elec-trodes pulled with a Brown-Flaming P-87 pipette puller and filled with3 M K-acetate. Initial electrode impedances ranged from 150 to 200M and were beveled (K. T. Brown BV-10 bevelor) until resistancesfell to between 60 and 100 M. For extracellular studies, electrodeswere advanced with a Kopf 650 hydraulic micro-drive and signals

with amplified with a WPI DAM50 preamplifier. For intracellulastudies, electrodes were advanced with a Burleigh piezoelectric microdrive and preamplified with a WPI 767 electrometer. Spike timesand times of EOD zero-crossings, and during intracellular recordingsmembrane potential waveforms were acquired with Cambridge Electronic Design 1401plus hardware and SpikeII software. Spike andelectric organ discharge (EOD) timing was measured with a resolutionof 0.1 ms, and analog waveforms were A/D converted at a rate of 12.5kHz. All subsequent data processing was done using Matlab (TheMathworks, Natick, MA).

The electric organ ofApteronotus is composed of modified motoneurons rather than muscle cells so the normal electric organ discharge remains intact during the neuromuscular blockade used inthese experiments. The pyramidal cell spontaneous firing propertiedescribed therefore refer to activity in the presence of the normabaseline receptor afferent activity, which is very constant in thabsence of modulations of the EOD amplitude. Electronically produced stepwise increases or decreases of the EOD, typically 1 mV/cmin amplitude and 300 ms in duration, were applied between electrodestraddling the fish and used as search stimuli. Stereotyped responseto this stimulus enabled categorization of cells as either basilar pyramidal cells (E cells) or nonbasilar pyramidal cells (I cells). The formerespond to increased EOD amplitude with short-latency increases inspike frequency while the latter respond to increased EOD amplitudewith reductions in firing frequency (Saunders and Bastian 1984).

Data analysis

Pyramidal cells were divided into subgroups, bursty or nonburstydepending on whether or not autocorrelograms of spontaneous activity deviated significantly from that expected assuming Poisson spiketrains (Abeles 1982). Autocorrelograms, 1-ms binwidth and 200-mduration, were produced from records of spontaneous activity typically containing 1,5002,000 spike times. For a few very low frequency cells, only 500 spike times were used. Spike trains werinitially displayed as plots of instantaneous frequency versus time andonly cells showing stable firing frequencies were studied. Expectedcorrelogram bin contents (y) was determined as

y f N t

wheref mean firing rate, n the number of spikes in the list, andt correlogram binwidth (1 ms). Autocorrelograms from Poissonspike trains are expected to be flat with bin contents approximatelyequal to y, ignoring the depression at the origin due to the cellsrefractory period. Initial peaks in the correlogram, determined to beunlikely assuming a Poisson spike train, were taken as the indicationof bursty firing. For each cell, a 99.9% confidence limit was calculatedfor the autocorrelogram expected bin contents as described by Abele(1982).

Given an expected bin contents, y, the probability of finding mspikes within a bin is

Pm; y eyym

m!

and these probabilities were accumulated to determine the smallestmsuch that

i0

m

P i; y 0.999

Autocorrelograms of spontaneous activity having initial peaks withbin contents exceedingmwere categorized as bursty, and examples oautocorrelograms from nonbursty and bursty cells are shown in Fig. 4A2and B2, respectively. The maximum ISI separating spikes considered to be within a burst was also determined from these correlo-grams. The time at which the falling phase of the initial peak crossed

1DENDRITIC CONTROL OF BURSTING


3/13

the 99.9% confidence limit was taken as the maximum ISI charac-teristic of bursts for that spike train, maximum intraburst interval(IBImax), and in cases where multiple early peaks remained above the99.9% threshold, the time of the minimum between the first two peakswas taken as IBImax.

Pharmacological techniques and current injection

Micropressure ejection techniques were used to apply the non-

NMDA glutamate antagonist 6-cyano-7-nitroquinoxaline-2,3-dione(CNQX) to local regions of the ELL molecular layer containing theapical dendrites of a recorded cell. Multibarrel pipettes were pulled toa fine tip and broken back to a total tip diameter of about 10 m.Typically two barrels were filled with a 1 mM solution of disodiumCNQX, two barrels were filled with 1 mM glutamate, and a fifthcontrol barrel contained distilled water. After a well- isolated single-unit extracellular recording or a stable intracellular recording wasestablished, the pressure pipette was slowly advanced into an appro-priate region of the ELL molecular layer while periodically ejectingpuffs of glutamate. Typically ejection duration ranged from 50 to100 ms, and ejection pressure was usually 40 psi. As described earlier(Bastian 1993), proximity to the apical dendrite of the recorded cellwas indicated by short-latency increases in firing rate following glu-tamate ejection. Following correct placement, CNQX was delivered

as a series of pulses (e.g., 100-ms puffs at 0.5 Hz for 20 s), and thistreatment typically resulted in tonic alterations in pyramidal activitylasting approximately 5 min. In experiments where the effects of tonichyperpolarization on pyramidal cell activity were studied, negativecurrent injection, typically 0.5 nA for 5 min, was applied via thebridge circuit of the electrometer. Unless indicated otherwise, samplemeans are given 1 SE.

R E S U L T S

Spontaneous activity

Spontaneous firing frequencies of 126 pyramidal cells wererecorded from the lateral and centrolateral segments of the ELLof 30 fish in which the animals neurogenic electric organ

discharge was intact. Recording site within the lateral andcentrolateral segments were estimated from surface landmarksand recording depth, and no obvious differences among cellsfrom these regions were seen. Cells were categorized as eitherE or I type, excited or inhibited by increasing electric organdischarge amplitude, respectively, and the distribution of spon-taneous firing frequencies is shown in Fig. 1A. Spontaneousfiring rates for this sample of E and I cells were not different,averaging 18.32 1.34, n 62 and 17.72 1.03, n 64,respectively, and these values are similar to those seen inearlier studies of this and the related fish Eigenmannia vire-scens (Bastian 1986a; Metzner et al. 1998).

The coefficient of variation (CV) of each pyramidal cellsrecord of ISIs was calculated as an initial estimate of spiketrain variability and, as described for Eigenmannia(Metzner etal. 1998),CVs were highly variable ranging from 0.45 to 2.24.The CVs were also negatively correlated with these cellsspontaneous firing frequency as shown in Fig. 1B. The CV isexpected to be reduced with increasing spike frequency sincefiring must become more regular as the ISI approaches thecells refractory period; however, effects of refractory periodalone cannot account for the range ofCVs seen in this sampleof pyramidal cells. Given a cell with an absolute refractoryperiod of 5 ms, a spontaneous firing rate of 5 spikes/s, and a CVof 1.5, increasing firing rate to 50 spikes/s is expected to

decreaseCVby about 25% (Gabbiani and Koch 1998, equation9.12). The reduction inCVs seen across pyramidal cells havinga similar range of firing frequencies was approximately 50% ortwice as large as predicted given refractory effects alone.

Unlike most weakly electric fish, the neurogenic electricorgan discharge in Apteronotus remains intact under neuromuscular blockade, hence the spontaneous activity referredto herein is activity in the presence of the normal dischargeElectroreceptor afferents time of firing is strongly phase

coupled to the EOD waveform (Bastian 1981a; Hagiwara et al1965; Hopkins 1976) while the ELL pyramidal cells generallyshow a weaker phase relationship to the EOD cycle (Bastian1981b). Phase histograms of firing times within the period othe EOD cycle were produced for a subset of the cells studiedPhase coupling was measured by computing the mean vectolength for each histogram (Batschelete 1981); this statisticranges from zero for histograms of activity unrelated to theEOD period to 1.0 for perfectly synchronized firing. Thstrength of the phase coupling to the EOD varied significantlyamong pyramidal cells, and phase histograms of spontaneous

FIG. 1. Coefficient of variation of pyramidal cell interspike intervals inegatively correlated with spontaneous firing rate.A: histogram of spontaneoufiring rates of 126 pyramidal cells. Black bars indicate basilar pyramidal celland white bars indicate nonbasilar cells. B: interspike interval coefficient ovariation versus spontaneous firing frequency. and E , basilar and nonbasilapyramidal cells, respectively.

12 J. BASTIAN AND J. NGUYENKIM


4/13


5/13

cells consisted of two phases: an initial peak at short ISIsfollowed by a longer tail (Fig. 4, B1 and C1) as described forpyramidal cells of the related fish Eigenmannia virescens(Metzner et al. 1998). The latter ISIH pattern is typical for cellshaving a tendency to produce short high-frequency bursts or

clusters of action potentials.A variety of techniques have been used to determine whetheror not to classify a cell as bursty and to define the ISI sizeconsidered to be characteristic of spikes within a burst. Theserange from identifying features such as the first trough orinflection point following the peak of ISI distributions orautocorrelation functions (Metzner et al. 1998; Turner et al.1996) to statistical methods based on the occurrence probabil-ity of short ISI sequences (Legendy and Salcman 1985). Amethod described by Abeles (1982) was adopted for identify-ing bursty pyramidal cells and for determining the ISIs char-acteristic of bursts. Autocorrelograms were constructed, andthe expected bin contents based on the cells average firing rateas well as the 99.9% confidence limit for the expected bin

contents ( and - - -, respectively, Fig. 4, A2C2) were deter-mined. Cells showing initial peaks in the autocorrelogramexceeding the confidence interval were categorized as bursty.

Autocorrelograms of approximately 20% of the cells studied(12 E cells and 13 I cells) were either flat, as in the case of Fig.4A2,or had small initial peaks that did not exceed the thresholdconfidence interval. These cells typically had simple exponen-tial ISI distributions, had relatively high rates of spontaneousactivity, and their average CV was 0.84. These cells werecategorized as nonbursty. The majority of cells (80%) werecategorized as bursty and the maximum intraburst interval(IBImax) of spike doublets, or longer bursts, was taken as thetime at which the falling phase of the autocorrelograms initialpeak crossed the confidence limit (arrow of Fig. 4B2). In caseswhere the correlogram contained multiple peaks that exceededthe threshold, as in Fig. 4C2, the time of the minimum follow-ing the first peak was taken as IBImax. The value of IBImaxforeach cell was then used to identify bursts of from 2 through 10spikes in the records of spontaneous activity.

The values of IBImaxfor 50 E cells and 51 I cells averaged15.98 0.58 and 15.29 0.71 ms, respectively, and thevalues of IBImax were negatively correlated with the cellsspontaneous firing frequencies (Fig. 5A). After identifying theIBImaxfor a given cell, burst probability was determined as theratio of the total number of bursts, each burst considered as a

single event, to the total number of events (burst events plussingle spikes) in the sample. Burst probabilities for E and cells averaged 0.25 0.014 and 0.22 0.012, respectivelyand were not significantly correlated with pyramidal cell spontaneous firing frequency (Fig. 5B).

Pyramidal cells were categorized as bursty based on significant deviations from the properties of a Poisson spike train

FIG. 4. Pyramidal cell interval histograms are highlyvariable ranging from simple exponential interval distributions (A1) to biphasic distributions showing an initial sharppeak characteristic of burst-like firing (B1and C1).A2C2autocorrelograms of the spike trains of A1C1. , th

expected bin contents assuming random ISIs; - - -, 99.9%confidence intervals of the expected bin contents.2, timeselected as the maximum interspike intervals characteristiof bursts (IBImax).

FIG. 5. Summaries of maximum intraburst intervals (IBImax) and bursprobabilities vs. pyramidal cell spontaneous firing frequency. Maximum intraburst intervals are negatively correlated with spontaneous firing rate (A, r0.63,P 0.001), but burst probabilities are not (B, r 0.03,P 0.79).C, single exponential fit to the tail of the ISI histogram (ISIH) shown andextrapolated to the intersection with the rising phase of the initial peak. Cinset: expanded y axis view showing region of the ISIH ( ) replaced withextrapolated values from the exponential fit to the tail of the ISIH. The meanof this distribution, following removal of the initial peak, was used to estimatethe non-Poisson bursts or the fraction of all bursts that could not beaccounted for assuming an underlying Poisson process (see text). D: relation onon-Poisson bursts to the cells spontaneous firing frequencies (r 0.73P 0.001).



6/13

that is, bursty cells are, by definition, those that produce anexcess of spike clusters over those expected from a Poissonspike train. The probabilities shown in Fig. 5B are overesti-mates of the excess bursts since some number ISIs shorterthan IBImaxwould be expected to occur even in a Poisson spiketrain. Therefore an alternative measure of burst probabilitydesigned to show the proportion of a cells total bursts aboveor in excess of those expected was also calculated. This wasdone by removing the fraction of bursts expected given aPoisson spike train of an average firing frequency predictedfrom the tail region of the ISIH: first, data from the tail of theISIH were fit with a single exponential, which was extrapolatedto its intersection with the start of the initial peak of the ISIH.An example is shown by the smooth line of Fig. 5C.The peakof the histogram above this extrapolation, including binsthrough 2 times IBImax, was then replaced with values pre-dicted by the exponential fit to the tail. The solid line shown inthe inset of Fig. 5C shows the altered initial phase of theresulting histogram, and the dotted line indicates values thatwere replaced. The mean spike frequencies (f) determined fromhistograms with the initial peaks removed in this manner wereused as the basis for predictions of the number of bursts

expected due to an underlying Poisson process. The probabilityofn events p(n) in a given time period, T, is given as:

pn fTnefT/n!

The probabilities of 2 spikes occurring within a time T IBImax,of 3 spikes occurring within T 2 IBImax, . . . , through10 spikes occurring within T 9 IBImaxwere determined andsubtracted from the actually observed probabilities for burstscontaining 2 through 10 spikes. Each probability determinedfrom the Poisson distribution was itself corrected for the smalleffects of refractory period. No spikes occur during a cellsrefractory period; however, given the Poisson distribution,there is a small probability of 2 or more spikes occurring withinthe time equal to the refractory period. So, for example, the

probability of two spikes occurring during T IBImax wascorrected by subtracting the probability of two spikes occurringduring the time equal to the refractory period.

The resulting corrected probabilities represent a cells ten-dency to fire bursts in excess of those accounted for by anunderlying Poisson process. These non-Poisson bursts, ex-pressed as a percentage of the total burst probability, areplotted in Fig. 5D,and this measure is strongly correlated withthe cells spontaneous firing rates (r 0.73,P 0.001). Themajority of bursts seen in low-frequency spike trains are non-Poisson, but for higher-frequency cells, approximately 50% ofthe bursts seen are expected given a Poisson spike train. Giventhe strong negative correlation between spontaneous firing

frequency and apical dendritic size, this analysis indicates thatcells with larger apical dendrites produce more bursts in excessof those expected.

Intraburst spike interval distributions

To determine if characteristic patterns of ISIs occur withinbursts, successive intervals within bursts of six spikes wereanalyzed for 64 pyramidal cells. Ten such bursts were pro-duced by the cell of Fig. 6A within a 3-min record, and thesuccessive ISIs within each burst are shown by E. The meansof these intervals are shown by F. No systematic pattern of

ISIs was seen for this cell except for the relatively long intervalat the end of the burst. Mean ISIs within bursts containing sixspikes from a sample of 10 cells are shown by E of Fig. 6B

The second through fifth intervals are expressed as a percent-age of the first, and no systematic pattern of successive inter-vals within bursts was seen for these cells. The F shows thegrand mean of within-burst intervals normalized in this mannefor the 64 cells analyzed, and no statistically significant trendin interval length was found.

A second analysis was performed on all bursts containingthree or more spikes. The differences between successive intervals within each burst were determined as described byTurner et al. (1996), and histograms of these interval differences were produced. Figure 6C shows the distribution ointraburst interval (IBI) differences for the cell of Fig. 6A. Thathe interval differences are approximately normally distributedwith a mean (0.035 ms) not significantly different from zeroindicates that there are no simple trends in interval lengthwithin bursts. Successive intervals neither increase nor decrease systematically; rather, for this cell, intervals vary randomly about a mean length of approximately 16 ms. Themeans of IBI difference distributions, like that of Fig. 6 C, weredetermined for 99 cells and are summarized in the histogram ofFig. 6D. The mean of this distribution (0.016 ms) is also nodifferent from zero. Although, in in vitro preparations, significant numbers of pyramidal cells showed trends or nonrandompatterns of IBI differences (Turner et al. 1996), in vivo, noconsistent patterning of IBIs is seen.

FIG. 6. No systematic variation is seen among successive intervals withinbursts.A: plots of successive interspike intervals (E) from 10 bursts of 6 spikerecorded from a basilar pyramidal cell with a spontaneous firing frequency o11.9 spikes/s and a CV of 1.46. , means of these successive intervals. Bmeans of successive ISIs from burst of 6 spikes produced by 10 differenpyramidal cells (E). Each mean interval is plotted as the percentage of the meanof the first interval. , grand mean of intraburst intervals within bursts of 6spikes from the 64 pyramidal cells studied. C: histogram of intraburst intervadifferences for bursts of from 3 to 10 spikes for the cell ofA. D: histogram omeans of IBI histograms like that ofCfor the 99 cells studied.



7/13

Blockade of descending excitation reduces burst-likebehavior

Previous in vivo studies showed that localized blockade ofexcitatory inputs to pyramidal cell apical dendrites can be

achieved via micropressure ejection of glutamate antagonistswithin the ELL dorsal and ventral molecular layers (Bastian1993). This technique was used to determine if alterations ofsynaptic inputs to pyramidal cell apical dendrites influencedbursting behavior. As in previous studies, a multibarrel pres-sure pipette was positioned in close proximity to the apicaldendrite of a pyramidal cell being recorded from. As thepressure pipette was advanced through the molecular layer,brief puffs of glutamate were periodically ejected, and short-latency responses to the glutamate were taken to indicateproximity of the pipette to the recorded cell. Previous pressureejection experiments coupled with extracellular labeling ofejection site and intracellular labeling of the recorded cellverified that short-latency excitatory responses occurred when

the pipette was very close to or within the cells dendritic arbor(Bastian 1993). Slightly longer-latency inhibitory responseswere typically seen when the pressure pipette was furtheroutside the dendritic arbor where glutamate preferentially ac-tivated inhibitory interneurons. Following recording of 200 s ofspontaneous firing, the non-NMDA antagonist CNQX wasejected. Typically 50-ms puffs were delivered at a rate of oneper 2 s, resulting in a reduction in the cells spontaneous firingrate and a decrease in the cells tendency to produce spikebursts. The reduction of spontaneous activity is not due toCNQX blockade of glutamatergic receptor afferent input topyramidal cells. As shown earlier (Bastian 1993) and repeatedin this study (data not shown), responses to changes in EODamplitude are increased rather than decreased by CNQX ap-plication to the molecular layers.

Figure 7, A, 1 and 2, and B, 1 and 2, shows the intervalhistograms and autocorrelograms of a bursty basilar pyramidalcell prior to and during the application of CNQX, respectively.Spontaneous firing rate was reduced from 10.0 to 5.1 spikes/s,and this reduction in firing frequency was mainly associatedwith a loss of spikes separated by short intervals. This resultsin a large reduction in the early peak of the ISIH and reductionin the height of the initial peak in the autocorrelogram. Theinset of Fig. 7B1 shows the initial phases of the intervalhistograms acquired before and after recovery from CNQX

(thin lines) superimposed on that taken during application ofthe antagonist (thick line). Burst probability for this cell wainitially 0.35 given an IBImax of 23 ms. In the presence oCNQX, this was reduced to 0.14 and IBImaxwas increased to26 ms. Cells typically recovered within 3 min of the cessation

of CNQX application, and this cells spontaneous rate returnedto 10.2 spikes/s while burst probability recovered to 0.39 withan IBImaxof 24 ms (Fig. 7C, 1 and 2).

The effects of CNQX application on the spontaneous firingpatterns of 24 pyramidal cells (11 basilar and 13 nonbasilarare summarized in Fig. 8. There was no significant differencein the behavior of these cell types so the data were pooled. Onaverage, both total burst probability and non-Poisson bursprobability were significantly reduced during CNQX application (Fig. 8, black and dark gray bars, P 0.005,t-tests), andthese measures returned toward their initial values followingtermination of CNQX application (Fig. 8, light gray bars)However, only 13 of the 24 cells were recorded throughcomplete recovery. The coefficients of variation were alsoreduced by CNQX treatment in most cells (16 of 24); however

FIG. 7. The non-N-methyl-D-aspartate (NMDA) glutamate antagonist [6-cyano-7-nitroquinoxalene-2,3-dion(CNQX)] reduces pyramidal cell bursting. A, 1 and2: ISIHand autocorrelogram of a basilar pyramidal cells spontaneous activity (10.0 sp/s,CV, 1.27, burst probability, 0.35prior to ejection of CNQX within the electrosensory lateraline lobe (ELL) dorsal molecular layer. B, 1and2: activityof this same cell during CNQX application (spontaneourate, 5.1 sp/s, CV, 1.15, burst probability, 0.14). Insetsuperposition of ISIH before and after CNQX application(thin lines) and ISIH during CNQX (thick line). C, 1 and2: ISIH and autocorrelogram following recovery fromCNQX (spontaneous rate, 10.2 sp/s, CV, 1.26, burst probability, 0.39).

FIG. 8. CNQX reduces pyramidal cell bursting and rates of spontaneouactivity. Overall burst probability as well as non-Poisson burst probability wasignificantly reduced by CNQX treatment (black and dark gray bars, P 0.00and 0.004, respectively, n 24,t-tests). Probabilities after recovery were nosignificantly different from pretreatment probabilities. Average spontaneoufiring frequency before, during and after CNQX are shown by the black, darkgray, and light gray hatched bars, respectively, and the white hatched bashows average frequency of the spike trains recorded prior to CNQX application following replacement of all bursts with single spikes.



8/13

the averageCvof all cells during CNQX (1.04 0.05) was notsignificantly less than that prior to treatment (1.09 0.04).Although the reduction in burst probability during CNQXinjection averaged approximately 45% for all cells studied,seven of these were only minimally affected by this treatment(burst probabilities changed by less than 5%). It is not knownif the lack of significant CNQX effects in these cells is a resultof failure to deliver effective CNQX doses or if the firingcharacteristics of these cells are truly insensitive to this treat-ment.

The reductions in spontaneous firing frequency seen withCNQX application (compare Fig. 8, black and dark grayhatched bars) were not due to simply shifting the ISI distribu-tions to longer values. Instead, as shown by Fig. 7B1, inset, thisantagonist preferentially reduced the probability of short ISIs,and this loss of the higher-frequency spike bursts must con-tribute to the observed reduction in spontaneous rate. Forcomparison with the CNQX effects on firing rate, the burstswere artificially removed from the pre-CNQX spike trains andreplaced with single spikes. The average frequency of thesemodified spike trains, which contained no bursts, is comparableto the frequency seen in the presence of CNQX (Fig. 8, hatched

white bar). Although this manipulation of pre-CNQX spiketrains indicates that loss of burst spikes alone could account forthe reductions in firing rate seen with CNQX treatment, theapplication of CNQX also resulted in increased numbers oflong (more than 300 ms) ISIs; this also contributes to the loweraverage firing rates seen.

Hyperpolarizing current injection reduces burst-likebehavior

Possible mechanisms by which CNQX blockade of pyrami-dal cell apical dendritic inputs could reduce spike bursts in-clude elimination of patterned synaptic inputs that directly

evoke pyramidal cell bursting as well as reduction of tonicexcitatory inputs resulting in hyperpolarization and reducedprobability that a burst-threshold is exceeded. To determineif moderate hyperpolarization is sufficient to reduce burstsintracellular recordings of spontaneous activity before, duringand after hyperpolarizing current injection were comparedFigure 9A1 shows two typical 1-s segments of intracellularlyrecorded basilar pyramidal cell spontaneous activity prior to

hyperpolarizing current injection. Spontaneous firing rate wainitially 15.3 spikes/s and the CV was 1.24. Constant curreninjection,0.5 nA, initially silenced pyramidal cells but typically within 30 s spontaneous firing resumed at a lower bustable rate. Figure 9A2 shows 1-s epochs of spontaneous activity after the cell adopted a new steady-state firing frequencyThis cell was hyperpolarized by 1.5 mV, which reduced spontaneous firing frequency to 8.6 spikes/s andCV to 1.07, and thismall hyperpolarization reduced burst probability from an initial valued of 0.24 to 0.04. Interval histograms of this cellspontaneous firing before and during current injection areshown in Fig. 9A3by the thin and thick lines, respectively, andas shown for CNQX blockade of descending excitation, hyperpolarization preferentially reduced the initial peak corresponding to the shortest ISIs.

A pressure ejection pipette was also in place while recordingfrom this cell so intracellular activity was monitored before andduring CNQX blockade of inputs to the cells apical dendriteComparison of the records (Fig. 9B, 1and2) shows that CNQXblockade resulted in similar changes in pyramidal cell spontaneous activity and bursting as did hyperpolarizing curreninjection. The brackets above the spike records show burstidentified based on IBImaxof 12 and 13 ms for the data of Fig9, Aand B, respectively. Spontaneous firing rate was reducedfrom 15.5 to 8.4 spikes/s,CVwas reduced from 1.3 to 0.74, andburst probability was reduced from 0.27 to 0.03. The interva

FIG. 9. Hyperpolarization of pyramidal cells mimics the effects of CNQX application to the vicinity othe apical dendrites.A1: 2 randomly selected segmentof spontaneous activity recorded from a basilar pyramidal cell. A2: randomly selected 1-s segments ospontaneous activity during continuous 0.5-nA current injection.A3: ISIHs of spontaneous activity befor(thin line) and during (thick line) 0.5-nA I injection

B, 1 and 2: randomly selected pairs of 1-s records o

spontaneous activity before and during CNQX application, respectively. B3: ISIHs of normal spontaneouactivity (thin line), of activity during CNQX (heavyline) and of activity with CNQX at a dosage reducedby 50% (dotted line).



9/13

histograms of Fig. 9B3summarize this cells firing before andduring the CNQX application (thin and thick lines, respectively),and the dashed line shows that the effects of CNQX were dosedependent; 50% of the original dose was applied in this case.

Figure 10 summarizes the effects of hyperpolarization on thespontaneous activity of 14 pyramidal cells (7 basilar, 7 non-basilar). Current magnitude was typically 0.5 nA, and theaverage hyperpolarization that resulted was 3.25 mV. As with

CNQX treatment, overall burst probability as well as non-Poisson burst probabilities were significantly decreased byhyperpolarization as was spontaneous firing rate. The effects ofremoving bursts of from 2 through 10 spikes from spike trainsrecorded at normal membrane potential and replacing themwith single events was also examined. As seen with CNQXtreatment, the resulting reduction in firing frequency due toartificial burst removal closely matched that due to hyperpo-larization (Fig. 10, dark gray and clear hatched bars). Follow-ing hyperpolarization burst probabilities recovered but, unlikethe recovery from CNQX where probabilities and spike ratesremained somewhat depressed, following hyperpolarizationboth burst probabilities and firing rates were larger than ini-

tially, although these increases were not statistically signifi-cant. The result that hyperpolarization mimicked the effects ofCNQX treatment is compatible with the idea that the CNQXblocked tonic excitation and altered a threshold for burst gen-eration. However, this result does not rule out the possibilitythat CNQX treatment also eliminates patterned dendritic in-puts. The intracellular recordings from cells that were alsotreated with CNQX did not reveal any consistent effects of thisdrug on resting potential; hyperpolarization was not systemat-ically observed in response to this treatment. Although it ispossible that the slow onset of CNQX effects along with smallchanges in impalement quality and other artifactual sourcesmembrane potential drift precluded clear-cut indications ofCNQX-caused hyperpolarization, the lack of changes in resting

potential may indicate that patterned inputs are also blocked bythis treatment.

D I S C U S S I O N

The results of this and of previous studies (Bastian 1993show that there is considerable variability among ELL pyramidal cells in terms of their physiology and morphology awell as in the distribution of their neurotransmitter receptortypes (Bottai et al. 1997, 1998; Dunn et al. 1999) and in thepresence of various intracellular signaling molecules (Bermanand Maler 1999; Berman et al. 1995; Zupanc et al. 1992). Thevariability of pyramidal cell apical dendrite morphology iparticularly striking (Fig. 3, AC), and the strong correlationbetween spontaneous firing rate and dendritic structure (Fig3D) enables us to predict the morphological characteristics ofcells studied with extracellular recording techniques. Cellwith the smallest apical dendrites (deep basilar pyramidal cells)also lie deeper within the ELL laminae (Bastian and Courtrigh1991) and exhibited the highest firing rates and the lowesprobabilities of bursts in excess of those expected based on aPoisson spike train, and their spontaneous activity showed the

strongest phase coupling to the electric organ discharge. Thisindicates that these cells are principally driven by receptoafferent inputs whose activity is also strongly phase-coupled tothe EOD. Conversely, the lowest-frequency cells have thlargest apical dendrites, they are found most superficiallywithin the pyramidal cell lamina, and their activity shows littleto no phase relation to the EOD waveform. These cells activ-ity is likely to be more strongly influenced by apical dendriticinputs; their spontaneous activity is more irregular (high CV)and these cells show the highest probability of producingunexpected spike bursts. The correlation between bursting andextensive apical dendrites has been previously demonstrated inmammalian cortical pyramidal cells in vitro (Mason and Larkman 1990) and modeling studies also showed that apical den-

dritic size alone, without differential distributions of ion chan-nels across morphological categories, was correlated with acells ability to produce spike bursts (Mainen and Sejnowsk1996).

The tendency of various ELL pyramidal cell types to pro-duce bursts may not be constant; instead, as sensory processingrequirements change, bursty behavior may be modulated optimizing the performance of these cells for the task at hand. Thelarge reductions in burst probability seen following blockade ofdescending excitation to the pyramidal cell apical dendrites andELL inhibitory interneurons supports the idea that bursty firingmay be under descending control.

Sources of apical dendritic inputs

The pyramidal cell apical dendrites receive input from twodistinct sources. Distal regions of the dendrites receive glutamatergic inputs, via AMPA as well as NMDA receptors, fromtypical cerebellar parallel fibers (Berman et al. 1997) andenormous numbers of these granule cell axons comprise theELL dorsal molecular layer (DML) (Maler 1979; Maler et al1981). The granule cell bodies are found in a region superficiato the DML termed the posterior eminentia granularis (EGp)and the granule cells receive descending electrosensory inputas well as proprioceptive information and, possibly corollary

FIG. 10. Hyperpolarizing current injection causes changes in pyramidal cellfiring characteristics similar to the effects of CNQX. Overall burst probabilityas well as non-Poisson burst probability was significantly reduced by I injec-tion (black and dark gray bars, P 0.03 and 0.04, respectively, n 14,t-tests). Ten cells were recorded through recovery, and burst probabilities afterrecovery were not significantly different from prehyperpolarization probabil-ities. Average spontaneous firing frequency before, during and after CNQX areshown by the black, dark gray, and light gray hatched bars, respectively, andthe white hatched bar shows average frequency of the spike trains recordedprior to I injection after replacement of all bursts with single spikes.



10/13


11/13

was positively correlated with the size of a neurons dendriticarbor.

In addition to showing that burst probability is related tocellular morphology, our results suggest mechanisms by whichdendritic inputs may modulate a cells firing pattern. In vitrostudies showed that the pattern of ELL pyramidal cell activityevoked by depolarizing current injection could shift from atonic to a bursty pattern contingent on the degree to which thecell was depolarized (Lemon and Turner 2000). In the presentstudy, blockade of direct excitatory and disynaptic inhibitoryinputs to the apical dendrites reduced bursting and hyperpo-larizing current injection, resulting in membrane potentialchanges as small as 1 to 2 mV, mimicked this result. Hencemodulation of descending inputs could bias pyramidal cells ina way that determines whether responses to electrosensorystimuli consist of bursts, enabling the cells to act as featuredetectors (Gabbiani and Metzner 1999; Gabbiani et al. 1996;Metzner et al. 1998) or consist of more tonic changes in firingrate that are better suited to encoding detailed informationabout a stimulus.

Bursts seen in vivo and in vitro also showed some importantdifferences. In vitro burst durations were much longer and

bursts contained many more spikes than seen in vivo. In theformer case, the average numbers of spikes/burst ranged from8 to 61 depending on the ELL subdivision or map recordedfrom (Turner et al. 1996), while in vivo the average wasbetween 2 and 3 spikes/burst (Gabbiani et al. 1996; and thisstudy). Spike intervals within bursts recorded from cells invitro were typically shorter, ranging from 3 to 13 ms (Lemonand Turner 2000) compared with 7 to 25 ms in vivo, and thismay indicate differences in the temporal characteristics of theDAPs under in vitro and in vivo conditions. In vitro burst ISIsalso often show serial patterning; alternating long and shortintervals as well as serially decreasing intervals are seen and, invitro, the last ISI of a burst is typically the shortest sincerefractory effects terminate in vitro bursts (Lemon and Turner

2000; Turner and Maler 1999; Turner et al. 1996). No consis-tent intraburst spike interval patterning was seen in vivo, andthe last ISIs within bursts were not systematically differentfrom any other intraburst ISI. The significantly shorter burstdurations seen in vivo plus the absence of indications that shortISIs lead to burst termination suggests that processes other thanrefractory effects terminate bursts in vivo.

One possibility is that inhibitory interactions within the ELLterminate bursts and detailed analyses of inhibition in the ELLhave recently appeared (Berman and Maler 1998ac, 1999).Dorsal molecular layer parallel fibers excite pyramidal cells aswell as inhibitory interneurons, DML stellate cells, and ventralmolecular layer neurons, and both of these provide GABAAinhibition to pyramidal cells (Berman and Maler 1998c). Stel-late inhibition primarily reduces the amplitude of individualparallel fiber excitatory postsynaptic potential (EPSPs), whileVML neuron inhibition is of longer duration and may modulatepersistent Na channel currents (Berman and Maler 1998c,1999). Given the importance of the persistent Na current inburst generation, VML cell inhibition is especially attractive asa candidate mechanism for burst termination in vivo. Appro-priately timed excitatory parallel fiber inputs to pyramidal cellscould initially augment the somatic spike-evoked depolarizingafterpotentials initiating bursts, and the disynaptically evokedinhibition could repolarize the cell terminating the burst. In

addition, to terminating individual bursts, the results of thehyperpolarizing current injection experiments suggest that inhibitory inputs to the proximal apical dendrites and somaticregions could also modulate overall burst probability by participating in the control of a cells membrane potential.

In addition to the possibility that DML parallel fibers modulate pyramidal cell burst probability in a tonic fashion byaltering the cells membrane potential and therefore the probability that a burst threshold will be exceeded, it is also possible that excitatory ventral molecular layer inputs triggeindividual bursts. Nucleus praeminentialis stellate cells projecbilaterally to the ELL ventral molecular layers and providepowerful excitatory inputs to the initial segments of the py-ramidal cells apical dendrites (Sas and Maler 1983, 1987)These cells have very low rates of spontaneous activity, buthey respond with high-frequency bursts of activity to certainpatterns of electrosensory stimuli (Bratton and Bastian 1990)and stellate cell evoked EPSPs studied in vivo and in vitroshow extreme frequency-dependent posttetanic potentiation(Bastian 1996b; Berman and Maler 1998b, 1999; Wang andMaler 1997, 1998). The reciprocally topographic relationshipsbetween ELL pyramidal cells and the nP stellate cells and the

nP stellate cells burst-like responses to appropriate electrosensory stimuli and the highly facilitating pyramidal cell EPSPsevoked by the descending stellate activity all support the ideathat this circuit behaves as a positive feedback loop that greatlyamplifies responses to certain patterns of electrosensory inpuperhaps by evoking pyramidal cell bursts.

Electrosensory information serves at least two separate categories of behavior. Electrocommunication behaviors involvean animals detection of conspecifics discharges followed bythe generation responses such as the jamming avoidance response (Heiligenberg 1991) or chirps (Bullock 1969; Larimerand Macdonald 1968). The stimuli received during electrocommunication consist of the sum of an animals own dischargeplus that of a conspecific. This summation produces a bea

waveform consisting of cyclic patterns of amplitude and relative phase modulations (Heiligenberg 1991). These EOD modulations are spatially extensive and influence pyramidal celreceptive field centers and antagonistic surrounds simultaneously. Previous studies demonstrated that pyramidal celspike bursts encode the occurrence of EOD amplitude increases and decreases more reliably than either electroreceptoafferent spikes or pyramidal cell single spikes (Gabbiani andMetzner 1999; Metzner et al. 1998), and accurate representation of the timing of these stimulus features is critical for thecontrol of these electrocommunication behaviors (Heiligenberg 1991).

Electrolocation stimuli also generate EOD amplitude andphase modulations, but unlike electrocommunication signalsthese AMs are spatially localized and typically have a movement component. Hence electrolocation stimuli may affecsubdivisions of pyramidal cell receptive fields sequentially. Ain the case of electrocommunication stimuli, increases or de-creases in EOD amplitude comprise important electrolocationstimulus features and reliably encoding these as bursts ospikes produced by a subset of the somatotopically mappedpyramidal cell population could, for example, indicate thcurrent position of a moving object. However, it also seemslikely that detailed information about the time course of theAM waveform may be important for electrolocation and cell



12/13

with high firing rates are better suited for encoding this infor-mation (Metzner et al. 1998). Although the highest spontane-ous firing rates seen for pyramidal cells, 4050 spikes/s, arestill well below that typical for receptor afferents (150400spikes/s), additional studies focusing on these nonbursty py-ramidal cells are needed to determine their potential for en-coding detailed information about the spatially localized EODmodulations that occur during electrolocation.

We thank Drs. L. Maler and R. Turner for helpful discussions.This work was supported by National Institute of Neurological Disorders

and Stroke Grant NS-12337 to J. Bastian.Present address of J. Nguyenkim: Division of Biology and Biomedical

Sciences, Washington University, 660 S. Euclid Ave., St. Louis, MO 63110.

REFERENCES

ABELESM. Quantification, smoothing, and confidence limits for single-unitshistograms. J Neurosci Methods 5: 317325, 1982.

BASTIANJ. Electrolocation. I. How the electroreceptors of Apteronotus albi-fronscode for moving objects and other electrical stimuli. J Comp Physiol144: 465479, 1981a.

BASTIANJ. Electrolocation. II. The effects of moving objects and other elec-trical stimuli on the activities of two categories of posterior lateral line lobecells in Apteronotus albifrons. J Comp Physiol 144: 481494, 1981b.

BASTIANJ. Gain control in the electrosensory system mediated by descending

inputs to the electrosensory lateral line lobe.J Neurosci6: 553562, 1986a.BASTIANJ. Gain control in the electrosensory system: a role for the descending

projections to the electrosensory lateral line lobe. J Comp Physiol [A]158:505515, 1986b.

BASTIANJ. The role of amino acid neurotransmitters in the descending controlof electroreception. J Comp Physiol [A] 172: 409423, 1993.

BASTIAN J. Pyramidal-cell plasticity in weakly electric fish: a mechanism forattenuating responses to reafferent electrosensory inputs. J Comp Physiol[A] 176: 6373, 1995.

BASTIAN J. Plasticity in an electrosensory system. I. General features of adynamic sensory filter. J Neurophysiol 76: 24832496, 1996a.

BASTIAN J. Plasticity in an electrosensory system. II. Postsynaptic eventsassociated with a dynamic sensory filter. J Neurophysiol 76: 24972507,1996b.

BASTIANJ. Plasticity in an electrosensory system. III. Contrasting properties ofspatially segregated dendritic inputs.J Neurophysiol79: 18391857, 1998a.

BASTIANJ. Modulation of calcium-dependent postsynaptic depression contrib-utes to an adaptive sensory filter. J Neurophysiol 80: 33523355, 1998b.

BASTIAN J. Plasticity of feedback inputs in the apteronotid electrosensorysystem. J Exp Biol 202: 13271337, 1999.

BASTIANJ ANDBRATTON B. Descending control of electroreception. I. Prop-erties of nucleus praeeminentialis neurons projecting indirectly to the elec-trosensory lateral line lobe. J Neurosci 10: 12261240, 1990.

BASTIAN J AND COURTRIGHT J. Morphological correlates of pyramidal celladaptation rate in the electrosensory lateral line lobe of weakly electric fish.

J Comp Physiol [A]168: 393407, 1991.BATSCHELETE. Circular Statistics in Biology. New York: Academic, 1981, p.

721.BELLCC, BODZNICKD, MONTGOMERYJ, AND BASTIANJ. The generation and

subtraction of sensory expectations within cerebellum-like structures. BrainBehav Evol50: 1731, 1997.

BERMANNJ, HINCKEMT, AND MALER L. Inositol 1,4,5-trisphosphate receptorlocalization in the brain of a weakly electric fish (Apteronotus leptorhyn-

chus) with emphasis on the electrosensory system. J Comp Neurol 361:512524, 1995.BERMAN NJ AND MALER L. Inhibition evoked from primary afferents in the

electrosensory lateral line lobe of the weakly electric fish (Apteronotusleptorhynchus). J Neurophysiol 80: 31733196, 1998a.

BERMANNJ ANDMALER L. Interaction of GABAB -mediated inhibition withvoltage-gated currents of pyramidal cells: computational mechanism of asensory searchlight. J Neurophysiol 80: 31973213, 1998b.

BERMAN NJ AND MALER L. Distal versus proximal inhibitory shaping offeedback excitation in the electrosensory lateral line lobe: implications forsensory filtering. J Neurophysiol80: 32143232, 1998c.

BERMANNJANDMALERL. Neural architecture of the electrosensory lateral linelobe: adaptations for coincidence detection, a sensory searchlight and fre-quency-dependent adaptive filtering. J Exp Biol 202: 12431253, 1999.

BERMAN NJ, PLANT J, TURNER R, AND MALER L. Excitatory amino aci

transmission at a feedback pathway in the electrosensory system. J Neurophysiol78: 18691881, 1997.

BOTTAI D, DUNNRJ, ELLISW, AND MALERL. N-methyl-D-aspartate receptor mRNA distribution in the central nervous system of the weakly electric fish

Apteronotus leptorhynchus. J Comp Neurol 389: 6580, 1997.BOTTAI D, MALER L, AND DUNNRJ. Alternative RNA splicing of the NMDA

receptor NR1 mRNA in the neurons of the teleost electrosensory systemJ Neurosci18: 51915202, 1998.

BRATTON B AND BASTIAN J. Descending control of electroreception. II. Prop

erties of nucleus praeeminentialis neurons projecting directly to the electrosensory lateral line lobe. J Neurosci 10: 12411253, 1990.

BULLOCK TH. Species differences in effect of electroreceptor input on electriorgan pacemakers and other aspects of behavior in electric fish. Brain Beha

Evol2: 85118, 1969.CONNORS W AND GUTNICKMJ. Intrinsic firing patterns of diverse neocortica

neurons. Trends Neurosci 13: 99104, 1990.DUNN RJ, BOTTAI D, AND MALER L. Molecular biology of the Apteronotu

NMDA receptor NR1 subunit. J Exp Biol 202: 13191326, 1999.FRANKK AND BECKERMC. Microelectrodes for recording and stimulation. In

Physical Techniques in Biological Research. New York: Academic, 1964vol. 5, part A, p. 2384.

GABBIANI F AND KOCH C. Principles of spike train analysis. In: Methods inNeuronal Modeling(2nd ed.), edited by Segev I and Koch C. CambridgeMA: MIT Press, 1998, p. 313360.

GABBIANIF ANDMETZNERW. Encoding and processing of sensory informationin neuronal spike trains. J Exp Biol 202: 12671279, 1999.

GABBIANIF, METZNER W, W ESSELR, ANDKOCHC. From stimulus encoding tofeature extraction in weakly electric fish. Nature 384: 564567, 1996.

HAGIWARA S, SZABO T, AND ENGER PS. Electroreceptor mechanisms in high-frequency weakly electric fish, Sternarchus albifrons. J Neurophysio28: 784799, 1965.

HEILIGENBERGW. Neural Nets in Electric Fish. Cambridge, MA: MIT Press1991.

HOPKINSCD. Stimulus filtering and electroreception: tuberous electroreceptorin three species of gymnotid fish. J Comp Physiol 111: 171207, 1976.

LARIMERJL AND MACDONALD JA. Sensory feedback from electroreceptors toelectromotor pacemaker centers in gymnotids. Am J Physiol 214: 12531261, 1968.

LEGENDYCR ANDSALCMAN. Bursts and recurrences of bursts in the spike trainof spontaneously active striate cortex neurons.J Neurophysiol53: 9269391985.

LEMONN ANDTURNERRW. Conditional spike backpropagation generates bursdischarge in a sensory neuron.J Neurophysiol 84: 15191530, 2000.

LISMANJE. Bursts as a unit of neural information: making unreliable synapsesreliable.Trends Neurosci 20: 3843, 1997.

MAINEN ZF AND SEJNOWSKI TJ. Influence of dendritic structure on firingpattern in model neocortical neurons. Nature 382: 363366, 1996.

MALERL. The posterior lateral line lobe of certain gymnotoid fish: quantitativelight microscopy. J Comp Neurol 183: 323363, 1979.

MALER L AND MUGNAINI E. Organization and function of feedback to thelectrosensory lateral line lobe of gymnotiform fish, with emphasis on searchlight mechanism. J Comp Physiol [A] 173: 667670, 1993.

MALER L AND MUGNAINI E. Correlating gamma-aminobutyric acidergic circuits and sensory function in the electrosensory lateral line lobe of gymnotiform fish. J Comp Neurol 345: 224252, 1994.

MALER L, SAS EK, ANDROGERSJ. The cytology of the posterior lateral linlobe of high-frequency electric fish (Gymnotidae): dendritic differentiationand synaptic specificity in a simple cortex. J Comp Neurol 195: 871391981.

MASON A AND LARKMAN A. Correlations between morphology and electrophysiology of pyramidal neurons in slices of rat visual cortex. II. Electrophysiology.J Neurosci 10: 14151428, 1990.

METZNER W, KOCH C, WESSEL R, AND GABBIANI F. Feature extraction byburst-like spike patterns in multiple sensory maps. J Neurosci 18: 22832300, 1998.

NELSONME. A mechanism for neuronal gain control by descending pathwaysNeural Comp6: 242254, 1994.

NELSONME, XU Z, AND PAYNE JR. Characterization and modeling of P-typeelectrosensory afferent responses to amplitude modulations in a wave-typeelectric fish. J Comp Physiol [A] 181: 532544, 1997.

SAS E AND MALER L. The nucleus Praeeminentialis: a Golgi study of feedback center in the electrosensory system of a gymnotid fish. J Comp

Neurol221: 127144, 1983.



13/13

SASE AND MALERL. The organization of afferent input to the caudal lobe ofthe cerebellum of the gymnotid fish Apteronotus leptorhynchus. Anat Em-bryol177: 5579, 1987.

SAUNDERSJ AND BASTIANJ. The physiology and morphology of two types ofelectrosensory neurons in the weakly electric fish Apteronotus leptorhyn-chus. J Comp Physiol [A] 154: 199209, 1984.

SHUMWAY CA AND MALER L. GABAergic inhibition shapes temporaland spatial response properties of pyramidal cells in the electrosensorylateral line lobe of gymnotiform fish. J Comp Physiol [A] 164: 391407,1989.

TURNERRW AND MALERL. Oscillatory and burst discharge in the Apteronotidelectrosensory lateral line lobe. J Exp Biol 202: 12551265, 1999.

TURNERRW, MALER L, ANDBURROWSM. Electroreception and Electrocom-munication.Cambridge, UK: Company of Biologists, 1999, p. 11671458.

TURNER RW, MALER L, DEERINCK T, LEVINSON SR, AND ELLISMAN MH.TTX-sensitive dendritic sodium channels underlie oscillatory discharge in avertebrate sensory neuron. J Neurosci 14: 64536471, 1994.

TURNERRW, PLANTJR, AND MALER L. Oscillatory and burst discharge acroselectrosensory topographic maps. J Neurophysiol 76: 23642382, 1996.

WANGD ANDMALERL. The immunocytochemical localization of glutamate inthe electrosensory system of the gymnotiform fish, Apteronotus leptorhynchus. Brain Res 653: 215222, 1994.

WANG D AND MALER L. Plasticity of the direct feedback pathway in thelectrosensory system of Apteronotus leptorhynchus. J Neurophysiol 7818821889, 1997.

WANG D AND MALER L. Differential role of Ca/calmodulin-dependent kinases in posttetanic potentiation at input selective glutamatergic pathwaysProc Natl Acad Sci USA 95: 71337138, 1998.

WESSEL R, KOCH C, AND GABBIANI F. Coding of time-varying electric fieldamplitude modulations in a wave-type electric fish. J Neurophysiol 7522802293, 1996.

ZUPANCG, AIREYJA, MALERL, SUTKO JL, AND ELLISMANMH. Immunohistochemical localization of ryanodine binding protein in the central nervousystem of gymnotiform fish. J Comp Neurol 325: 135151, 1992.


bastian & nguyenkim (2001) dendritic modulation of burst-like firing in sensory neurons

Documents