MCN Molecular and Cellular Neuroscience 19, 402– 416 (2002)

doi:10.1006/mcne.2001.1088, available online at http://www.idealibrary.com on

tipE Regulates Na ⴙ-dependent Repetitive Firing

in Drosophila Neurons

Dianne D. Hodges,* ,1 Daewoo Lee,* Charles F. Preston,*

Kevin Boswell,* Linda M. Hall, † and Diane K. O’Dowd* ,2
*Department of Anatomy and Neurobiology and Department of Developmental and
Cell Biology, University of California at Irvine, Irvine, California 92697-1280; and
†
Functional Insect Genomics, 1105 Kennedy Place, Suite 4, Davis, California 95616

The tipE gene, originally identified by a tempera- INTRODUCTION

ture-sensitive paralytic mutation in Drosophila, en-
codes a transmembrane protein that dramatically influ- Cell-specific changes in electrical excitability during
ences sodium channel expression in Xenopus oocytes. early development are critical in formation of mature
There is evidence that tipE also modulates sodium neural circuits (Spitzer et al., 1994). Modulation of neu-
channel expression in the fly; however, its role in regu-
ronal excitability has also been implicated in mediating
lating neuronal excitability remains unclear. Here we
plasticity in the nervous system (Desai et al., 1999;
report that the majority of neurons in both wild-type and
Aizenman and Linden, 2000; Armano et al., 2000). Elec-
tipE mutant (tipE ⴚ ) embryo cultures fire sodium-depen-
dent action potentials in response to depolarizing cur- trophysiological studies demonstrate that alterations in
rent injection. However, the percentage of tipE ⴚ neu- the number, type, localization, and/or posttranslational
rons capable of firing repetitively during a sustained modifications of voltage-gated ion channels can influ-
depolarization is significantly reduced. Expression of a ence neuronal excitability (Barish, 1986; Huguenard et
tipE ⴙ transgene, in tipE ⴚ neurons, restores repetitive al., 1988; O’Dowd et al., 1988; Spitzer, 1991; Turrigiano et
firing to wild-type levels. Analysis of underlying cur- al., 1995; Massengill et al., 1997; Catterall, 2000). How-
rents reveals a slower rate of repolarization-dependent ever, the molecular mechanisms underlying regulation
recovery of voltage-gated sodium currents during re- of excitability are less clear.
peated activation in tipE ⴚ neurons. This phenotype is Using a genetic approach in Drosophila, progress has
also rescued by expression of the tipE ⴙ transgene. been made in identifying genes involved in mediating
These data demonstrate that tipE regulates sodium- neuronal excitability. Shaker and ether-a-go-go (eag), mu-
dependent repetitive firing and recovery of sodium cur- tants with hyperexcitable phenotypes, exhibit anoma-
rents during repeated activation. Furthermore, the du- lous repetitive firing in larval motor axons and identify
ration of the interstimulus interval necessary to fire a potassium channel genes important in determining the
second full-sized action potential is significantly longer
excitability properties of these neurons (Wu and
in single- versus multiple-spiking transgenic neurons,
Ganetzky, 1992; Littleton and Ganetzky, 2000). Temper-
suggesting that a slow rate of recovery of sodium cur-
ature-sensitive paralytic mutants identify additional
rents contributes to the decrease in repetitive firing in
tipE ⴚ neurons.
genes, such as paralytic (para) and no action potential
(nap), that were recognized as playing a role in medi-
ating neuronal excitability based on temperature-de-
pendent blockade of action potential conduction in lar-
1
Present address: Purdue Pharma Ltd., 213 Technology Drive, val nerve fibers (Wu and Ganetzky, 1992). The para gene
Irvine, CA 92618.
2
encodes a voltage-gated sodium channel ␣ subunit
To whom correspondence and reprint requests should be ad-
dressed at the Department of Anatomy and Neurobiology, University
(Loughney et al., 1989), whereas nap encodes an RNA
of California, Irvine, CA 92697-1280. Fax: (949) 824-1105. E-mail: helicase involved in editing of para transcripts (Reenan
dkodowd@uci.edu. et al., 2000).

1044-7431/02 $35.00
© 2002 Elsevier Science (USA)
402 All rights reserved.
tipE Regulates Neuronal Excitability 403

tipE mutant flies, similar to nap and para mutants,

exhibit a rapid and reversible temperature-sensitive pa-
ralysis (Kulkarni and Padhye, 1982). A reduction in the
number of sodium channel binding sites in head mem-
branes from tipE ⫺ flies (Jackson et al., 1986) and a de-
crease in sodium current density in cultured tipE ⫺ neu-
rons (O’Dowd and Aldrich, 1988) suggest that tipE may
regulate sodium channel expression. Enhanced temper-
ature sensitivity for nerve conduction failure in para;
tipE double mutants, compared with para alone, sup-
ports the suggestion that tipE can modulate axonal
conduction properties (Ganetzky, 1986). The cloning of
tipE revealed that the gene product is a novel integral
membrane protein with two membrane spanning re-
gions (Feng et al., 1995a). The tipE protein does not form
a functional channel by itself when expressed in Xeno-
pus oocytes but coexpression of tipE with para cRNA
alters both the expression levels and the fast kinetic FIG. 1. Wild-type and tipE ⫺ Drosophila neurons grown in primary
properties of the para-encoded voltage-gated sodium dissociated cell culture. Neuronal clusters interconnected by overlap-
channels (Feng et al., 1995a; Warmke et al., 1997). Taken ping branched neuritic processes in a wild-type (A) and a tipE ⫺ (C)
together these data suggest that tipE may define a new culture. Isolated neurons in a wild-type (B) and a tipE ⫺ (D) culture.
class of proteins that regulates electrical excitability Cultures were grown for 2 days in vitro in DDM1, fixed in 4%
paraformaldehyde, and stained with fluorescein-conjugated anti-HRP
through an interaction with the para sodium channel. antibodies. Scale bars, 20 ␮m.
Analysis of excitability in tipE ⫺ mutants has been
limited to extracellular recordings in larval motor neu-
rons that, interestingly, demonstrated apparently nor-
mal action potential propagation even at behaviorally some with simple neurites and others with elaborately
nonpermissive temperatures (Ganetzky, 1986). Using branched processes. Neurons could be found in clus-
cell cultures that contain subpopulations of primary ters, where there was contact between neighboring cells
Drosophila neurons exhibiting distinct firing pheno- (Figs. 1A and 1C), and in isolation (Figs. 1B and 1D).
types, we explored the role of tipE in regulation of The whole-cell recording technique was used to exam-
neuronal excitability. A line of transgenic flies carrying ine the firing properties of neurons at 2 and 3 days in
the wild-type tipE ⫹ gene under the control of a heat- vitro. All recordings were performed blind with respect
shock promoter, in a tipE ⫺ background (Feng et al., to genotype. Initial studies were conducted to deter-
1995b), was crucial in determining the electrical pheno- mine if a correlation between morphological features
types associated with tipE. Our results demonstrate that and neuronal excitability could be established. Because
tipE plays a role in regulating sodium-dependent repet- there did not appear to be systematic differences in the
itive firing properties in cultured Drosophila neurons. electrical properties of isolated versus clustered neu-
rons all data were grouped for statistical analysis.
The majority of neurons in both the tipE ⫺ (76 ⫾ 5%,
n ⫽ 10 platings) and the wild-type (82 ⫾ 5%, n ⫽ 10
RESULTS platings) cultures were electrically excitable based on
the ability to elicit one or more action potentials in
tipE ⴚ Neurons Exhibit a Decrease in Repetitive
response to a 600-ms, suprathreshold depolarizing cur-
Firing, Spontaneous Firing, and
rent step. The electrically excitable neurons were
Action Potential Amplitude
grouped into three broad classes: single spiking, graded
To determine if the tipE gene plays a role in regulat- multiple spiking, or multiple spiking (Figs. 2A and 2B).
ing neuronal excitability we compared the firing prop- The single spiking neurons were characterized by a
erties of embryonic tipE ⫺ and wild-type neurons grown single action potential elicited at the beginning of the
in dissociated cell culture. Cultures from both geno- step depolarization (Fig. 2A). In this class of neurons,
types contained heterogeneous populations of neurons: neither changes in holding potential nor increases in the
404 Hodges et al.

scribed in “giant” Drosophila neurons grown in cell

culture (Zhao and Wu, 1997). The tonic firing pattern
was characterized by a relatively constant interspike
interval during a train of action potentials (Fig. 2B). The
adaptive subclass was characterized by a decrease in
frequency during the spike train. Neurons were in-
cluded in this class if the first interspike interval was
less than 70% of the last interspike interval in the train.
Some, but not all, adaptive cells contain a doublet at the
beginning of the spike train as seen in the adaptive cell
shown in Fig. 2B. In the delayed subclass, action poten-
tials were usually observed after a latency of ⬎100 ms
from the onset of the stimulus (Fig. 2B). However, in a
small number of cells in this group the delay was ⬍100
ms but the interspike interval became progressively
shorter throughout the action potential train. Repeated
depolarizing stimuli, separated by 2–5 s at rest, resulted
in reproducible firing patterns in neurons within each
of the three classes.
Spontaneous action potentials were also observed in
some of the cultured cells. Neurons were classified as
spontaneously firing if: (1) action potentials were ob-
FIG. 2. Embryonic Drosophila neurons grown in DDM1 exhibit het- served in extracellular recordings obtained in the cell-
erogeneous firing properties. (A) Representative whole-cell current attached configuration and/or (2) action potentials
clamp recordings obtained from two neurons illustrating the single- were observed in intracellular current clamp recordings
spiking and the graded multiple-spiking firing classes. (B) The mul- at the cell’s resting potential in the absence of step
tiple-spiking class is composed of three subclasses; examples of each
depolarizations (Fig. 2C).
type are illustrated. Voltage traces in both A and B were recorded in
response to a 600 suprathreshold depolarizing current step from a Neurons in all three firing classes were observed in
negative holding potential. (C) Spontaneous action potentials re- both wild-type and tipE ⫺ cultures. However, there was
corded from a single neuron, first in a cell-attached recording config- a significant difference in the distribution of neurons
uration (extracellular), followed by recordings made in the whole-cell within these three classes (Fig. 3). In wild-type cultures
recording configuration (intracellular), in the absence of depolarizing
the most prevalent class of neurons was multiple spik-
current injection. All electrophysiological recordings in these and the
subsequent figures were obtained at room temperature from neurons ing, comprising approximately 70% of all electrically
at 2–3 days in vitro. excitable cells, with the single-spiking class represent-
ing about 25% of the total. In contrast, single-spiking
neurons were the most abundant class (50% of total) in
the tipE ⫺ cultures, with only 25% of the neurons exhib-
amplitude of current injected were capable of inducing iting multiple-spiking properties (Figs. 3A and 3B).
additional action potentials during the 600-ms step. The Among tipE ⫺ neurons that were multiple spiking, the
graded multiple-spiking neurons fired two to six action distribution within the three subclasses was similar to
potentials at the beginning of a 600-ms depolarizing that of wild type (Table 1). Spontaneously firing neu-
current step with a steady decrement in amplitude of rons were observed in both tipE ⫺ and wild-type cul-
each successive action potential following the first one tures. However, there was a twofold decrease in the
or two. In these neurons no spikes were elicited after incidence of spontaneously firing neurons in the tipE ⫺
the first 300 ms (Fig. 2A). cultures compared with wild type (Fig. 3C). Because the
The third class, multiple-spiking neurons, fired trains vast majority of spontaneously firing neurons detected
of action potentials throughout the 600-ms depolarizing were in the multiple-spiking class, this decrease is likely
current step and each action potential was of approxi- to reflect the reduction in the percentage of multiple-
mately equal amplitude (Fig. 2B). The multiple-spiking spiking neurons in the tipE ⫺ cultures.
category could be further subdivided into three sub- Analysis of the first spike induced by suprathreshold
classes (tonic, adaptive, and delayed) as previously de- depolarization in each neuron revealed that the action
tipE Regulates Neuronal Excitability 405

Rescue of Reduced Repetitive Firing and

Spontaneous Firing by Expression of the tipE ⴙ
Transgene in Differentiated tipE ⴚ Neurons
The reduced repetitive firing, spontaneous firing, and
action potential amplitude in tipE ⫺ neurons are consis-
tent with the hypothesis that tipE regulates these prop-
erties in Drosophila neurons. However, one caveat to
this interpretation is that the comparison was made
between populations of neurons harvested from two fly
strains in which the contribution of differences in ge-
netic backgrounds is unknown. To determine if the
altered electrical properties in tipE ⫺ neurons are the
consequence of a mutation in the tipE gene we studied
a transgenic line (tipE ⫺:tipE ⫹) containing the wild-type
tipE transgene (tipE ⫹), under the control of a heat shock
promoter, in the tipE ⫺ background.
To monitor expression of the transgene and deter-
mine if it could be regulated in the neurons by heat
shock, cultures were prepared from wild-type, tipE ⫺,
and tipE ⫺:tipE ⫹ embryos. Half of the cultures in each
genotype were subjected to three 1-h heat shocks (see
Experimental Methods for details). Non-heat-shocked
cultures were maintained continuously at room temper-
ature. RNA was prepared from all the cultures at 42 h
FIG. 3. The frequencies of repetitive firing, spontaneous firing, and
after plating. Primers (M1 and M2) flanking a single
action potential amplitude are decreased in tipE ⫺ neurons. (A) Rep-
resentative recordings illustrating the most abundant firing class in point mutation in the tipE cDNA (removing an RsaI
wild-type and tipE ⫺ neurons. (B) There was a significant reduction in restriction enzyme site) generated PCR products of dis-
the percentage of neurons in the multiple-spiking (MS) class in tipE ⫺ tinct sizes in wild-type and tipE ⫺ neurons following
cultures compared with wild-type, with a corresponding increase in RsaI digestion (Fig. 4A). In this analysis we found that
the % of neurons in the single-spiking (SS) class. The mean percentage
wild-type and tipE ⫺ neurons, in both non-heat-shocked
of total excitable cells within each firing class was determined by
calculating the percentage observed in nine separate experiments in and heat-shocked cultures, expressed only wild-type or
which three or more neurons were examined. (C) The fraction of mutant tipE mRNA, respectively. In contrast, tipE ⫺:
neurons firing spontaneous action potentials, as a function of the total tipE ⫹ neurons express both mutant and wild-type tipE
number of neurons examined, was significantly lower in tipE ⫺ com- mRNA, even in the absence of heat shock (Fig. 4B). The
pared with wild-type. The means represent data obtained from three
relative abundance of the wild-type tipE product in the
or more neurons in four separate platings for both genotypes. (D) The
amplitude of the first action potential (AP) in each train (peak to tipE ⫺:tipE ⫹ cultures was dramatically increased (⬎10-
trough) was significantly reduced in tipE ⫺ neurons compared to wild fold) following heat shock. However, since the absolute
type. Bars indicate SEMs. * P ⬍ 0.05, **P ⬍ 0.01, ***P ⬍ 0.001, levels of tipE mRNA produced are not known it was
Student’s t test. possible that, even in the absence of heat shock, wild-
type tipE levels might be sufficient to rescue the mutant
firing phenotype. Therefore, in all the electrophysiology
potential amplitude in the tipE ⫺ neurons was reduced
compared to wild type (Fig. 3D). In contrast, the action
potential duration was not different between the two TABLE 1
genotypes (Table 2). No significant changes in the input Subclass Distribution of Multiple Spiking (MS) Neurons
resistance or the resting membrane potential were de- Tonic (%MS) Adaptive (%MS) Delayed (%MS)
tected (Table 2). There was a large range in the size of
the cultured neurons but the mean capacitance of the Wild type 53.9 ⫾ 10.8 25.3 ⫾ 10.3 21.3 ⫾ 6.1
tipE ⫺ 57.4 ⫾ 13.1 20.4 ⫾ 11.7 18.8 ⫾ 13.2
population examined in the two genotypes was similar
(Table 2). Note. 9 platings; mean ⫾ SEM.
406 Hodges et al.

TABLE 2
Comparison of Electrical Properties in Wild-Type and tipE ⫺ Neurons at 2–3 DIV

Action potential duration Input resistance Resting potential Capacitance

(ms) (GOhm) (mV) (pA)

Wild type 6.5 ⫾ 1.2 (35) 1.3 ⫾ 0.1 (43) ⫺38.0 ⫾ 2.0 (53) 22.3 ⫾ 1.3 (45)
tipE ⫺ 6.7 ⫾ 1.0 (33) 1.4 ⫾ 0.2 (46) ⫺37.0 ⫾ 2.1 (43) 19.0 ⫾ 1.2 (54)

Note. Mean ⫾ SEM (No. of neurons).

studies comparisons included analysis of neurons in firing classes were readily apparent between tipE ⫺ and
tipE ⫺:tipE ⫹ cultures examined after heat shock and in wild-type neurons. In four independent experiments,
sibling cultures that were maintained continuously at half of the tipE ⫺:tipE ⫹ cultures were exposed to two 1-h
ambient temperature. heat shocks at 42 and 49 h after plating while the
To address the role of tipE in neuronal excitability, remaining cultures were not heat shocked (Fig. 5, top).
independent of development, tipE ⫺:tipE ⫹ cultures were To control for the affects of heat shock alone, wild-type
grown in the absence of heat shock for the first 2 days cultures prepared in parallel were exposed to the same
in vitro, by which time differences in the three major heat-shock regime. Cultures were coded and examined
blind with respect to genotype and heat-shock condi-
tions. Examples of firing properties in three different
neurons recorded from a wild-type (⫹HS), a tipE ⫺:tipE ⫹
(⫺HS), and a tipE ⫺:tipE ⫹ (⫹HS) culture are illustrated
in Fig. 5A. In the absence of heat shock, there were very
few multiple-spiking neurons in the tipE ⫺:tipE ⫹ cul-
tures, with the majority of excitable cells split between
the graded multiple-spiking and single-spiking firing
classes (Fig. 5B), similar to the distribution seen previ-
ously in the tipE ⫺ cultures. These data demonstrate that
the level of wild-type tipE product in the absence of
heat shock does not rescue the mutant firing phenotype
in transgenic cultures. However, the altered firing type
distribution was fully rescued within 24 h after heat
shock (⫹HS), with over 90% of the neurons in the
FIG. 4. Heat shock drives expression of wild-type tipE mRNA in a multiple-spiking firing class and the remainder classi-
tipE ⫺ background. (A) Schematic representation of the tipE cDNA fied as single spiking, similar to the distribution seen in
with the orientation of primer pair M1 (radioactively labeled) and M2. wild-type cultures (⫹HS) examined in parallel (Fig. 5B).
Positions of two RsaI restriction enzyme sites, R1* (eliminated in the The percentage of spontaneously firing neurons, low in
tipE mutant) and R2 are shown. RsaI digestion of the PCR product
generated by amplification using M1/M2 yields a labeled fragment of
the tipE ⫺:tipE ⫹ cultures in the absence of heat shock
85 bp in tipE ⫺ and 66 bp in wild type. (B) Autoradiogram of RsaI (⫺HS), was also rescued following heat shock (⫹HS)
digests of PCR products from wild-type, tipE ⫺, and tipE ⫺:tipE ⫹ (⫾HS) (Fig. 5C). Unexpectedly, the reduced action potential
cultures. Only the 66-bp product is amplified in RNA prepared from amplitude in the transgenic neurons was not rescued
wild-type neurons. Amplification of only the 85-bp product is ob- following heat shock (Fig. 5D).
served in RNA prepared from tipE ⫺ neurons. Heat shock does not
alter expression of these products in either the wild-type or the tipE ⫺
neurons. RNA prepared from tipE ⫺:tipE ⫹ transgenic embryo cultures Repolarization-Dependent Recovery of Sodium
expresses both the wild-type and the mutant product even in the
absence of heat shock. However, following heat shock there is a Currents and Excitability in Wild-Type, Mutant,
dramatic increase in the relative abundance of the wild-type (66 bp) and Transgenic Neurons
versus the mutant (85 bp) message. Cultures were heat shocked by
incubation in a 37°C, 5% CO 2 incubator for 1 h at 16, 33, and 40 h after Previous studies had demonstrated that the wild-
plating (see Experimental Methods for details). Total RNA was ex- type tipE gene product upregulates the amplitude and
tracted 2 h after the last heat shock. alters the kinetic properties of para sodium currents in a
tipE Regulates Neuronal Excitability 407

heterologous expression system (Feng et al., 1995a;

Warmke et al., 1997). In addition, a decrease in sodium
current density in tipE ⫺ versus wild-type neurons was
observed at 1 day in culture (O’Dowd and Aldrich,
1988). Therefore, we asked if there were changes in the
sodium current properties that might contribute to the
mutant firing phenotypes seen in the present culture
condition. These studies were complicated by the fact
that the sodium currents in the majority of the electri-
cally excitable neurons could not be well voltage-
clamped, precluding a classical quantitative biophysical
analysis of the underlying sodium channel properties.
However, comparison of features of the whole-cell cur-
rents elicited by step depolarizations, using identical
recording conditions for wild-type, mutant, and trans-
genic neurons, allowed us to identify sodium current
properties linked to tipE expression.
The first comparison focused on the maximal sodium
current activated in mutant and wild-type neurons.
Individual neurons were stimulated with a series of
increasing depolarizing voltage steps in the presence of
cesium in the internal solution to block outward potas-
sium currents. The sodium current density in each neu-
ron was determined by normalizing the peak inward
sodium current to the whole-cell capacitance. There
was a 25% reduction in the peak sodium current density
in the tipE ⫺ versus wild-type neurons (Fig. 6A). The
sodium current density in the transgenic neurons in the
FIG. 5. Expression of wild-type tipE message in a tipE ⫺ background
absence of heat shock was similarly low. However, this
restores wild-type distribution of firing phenotypes and spontaneous reduced sodium current density was not rescued fol-
firing levels in differentiated neurons. The heat-shock regime used in lowing heat shock (Fig. 6A). The inability to rescue
these experiments is illustrated on the top. Half of the cultures in each sodium current density is consistent with the inability
genotype were subjected to two 1 h, 37°C heat shocks at 42 and 49 h to rescue the action potential amplitude, suggesting
after plating. Electrophysiological recordings were done at 66 –74 h
after plating (see Experimental Methods for details). (A) Representa-
that these two properties are linked. These data further
tive action potential trains recorded from neurons in the three differ- demonstrate that recovery of robust repetitive firing
ent culture conditions tested, wild-type (⫹HS), tipE ⫺:tipE ⫹ (⫺HS), does not require rescue of the sodium current density.
and tipE ⫺:tipE ⫹ (⫹HS). (B) Heat shock rescued the altered firing class In light of our finding that most tipE mutant neurons
distribution seen in the mutant neurons. The distribution of neurons, are capable of firing a single action potential but are
normalized to total number of excitable cells, among the three major
firing classes was determined in the different cultures. The % multi-
compromised in their ability to fire repetitively, we
ple-spiking (MS) and % single-spiking (SS) neurons are significantly investigated the recovery of sodium currents during
different in wild type (⫹HS) and tipE ⫺:tipE ⫹ (⫹HS) compared to repetitive activation. Neurons were subject to two iden-
tipE ⫺:tipE ⫹ (⫺HS) (ANOVA, *P ⬍ 0.05, **P ⬍ 0.01, Fisher’s pro- tical voltage steps from ⫺75 to ⫺5 mV, separated by a
tected least significant difference). Heat shock did not alter the dis- 10-ms interstimulus interval at ⫺75 mV (Fig. 6B). The
tribution of firing classes in wild-type cultures. Mean percentages
were determined by calculating the percentages in four independent
amplitude of the current elicited by the second step was
experiments in which 4 – 6 neurons/culture condition were examined
in each experiment. The total number of neurons in each group was
18 wild type (⫹HS), 22 tipE ⫺:tipE ⫹ (⫺HS), and 21 tipE ⫺:tipE ⫹ (⫹HS).
(C) The level of spontaneous firing is restored to wild-type levels in
transgenic neurons following heat shock. The fraction of neurons potential (AP) amplitude is not rescued by induction of the tipE ⫹
firing spontaneous action potentials, as a function of the total number transgene: there is no significant difference between the mean ampli-
of neurons examined, is significantly different between tipE ⫺:tipE ⫹ tude in neurons in ⫹HS (n ⫽ 13) and ⫺HS (n ⫽ 12) tipE ⫺:tipE ⫹
neurons ⫹HS and ⫺HS (**P ⬍ 0.01, Student’s t test). (D) The action cultures (P ⬎ 0.05, Student’s t test). Bars on all graphs indicate SEM.
408 Hodges et al.

in the mutant was also seen in the transgenic neurons in

the absence of heat shock. Following heat shock, so-
dium current recovery was restored to wild-type levels
(Fig. 6B). There was no decrement in sodium current
amplitude following a 2-s interstimulus interval at ⫺75
mV, the standard time between repeated trials, in either
mutant or rescued transgenic neurons. These findings
demonstrate that the tipE mutation results in a slower
rate of repolarization-dependent recovery of sodium
currents during repetitive activation.
A straightforward interpretation of these data would
be that the tipE mutation slows but does not block
recovery of the underlying sodium channels from inac-
tivation. In the absence of excess sodium channels, this
would decrease the probability of repetitive spiking
during a sustained depolarization. The poor space
clamp in the cultured neurons makes a more detailed
investigation of the kinetics of recovery of sodium
currents from inactivation problematic. However, if
slowed recovery from inactivation contributes to the
tipE ⫺ firing phenotype, then one would predict that
repolarization should be necessary, and 2 s interstimu-
lus interval sufficient, for firing a second spike in tipE ⫺
FIG. 6. Sodium currents in wild-type, tipE ⫺, and transgenic neurons (single-spiking) neurons. Therefore, five single-spiking
(tipE ⫺:tipE ⫹). (A) Sodium current density is reduced in tipE ⫺ neurons
but is not rescued by induction of the tipE ⫹ transgene. Sodium current
neurons, in transgenic cultures in the absence of heat
density was calculated by normalizing the maximal amplitude so- shock, were held between ⫺50 and ⫺60 mV and sub-
dium current to the whole-cell capacitance in each neuron in which a jected to prolonged depolarizing steps (5 s) separated
sodium current was observed. There is a significant decrease in mean by 2 s intervals. All neurons fired one action potential at
sodium current density in tipE ⫺ compared to wild-type neurons. *P ⬍ the onset of each depolarizing step. These data demon-
0.05, Student’s t test, unpaired. The mean sodium current densities in
transgenic neurons (tipE ⫺:tipE ⫹) under ⫹HS and ⫺HS conditions are
strate that an interstimulus interval is necessary and 2 s
not significantly different from each other (P ⬎ 0.05, Student’s t test). is sufficient for firing a second full-sized action poten-
(B) Recovery of sodium currents from repetitive activation is reduced tial in tipE ⫺ neurons that are classified as single spiking.
in the mutant and rescued by heat-shock induction of the tipE ⫹ The relationship between interstimulus duration and
transgene. This was examined using the two-step protocol illustrated; recovery of excitability was examined in cultures of
holding potential and recovery voltages were ⫺75 mV, test steps were
to ⫺5 mV for 10 ms, interstimulus interval was 10 ms. The sodium
transgenic neurons, half that were heat shocked and
current amplitude elicited by the second test step was normalized to half that served as controls. Neurons were held at volt-
that elicited by the first step. The normalized sodium current ampli- ages between ⫺50 and ⫺60 mV and given two identical,
tude in tipE ⫺ neurons (n ⫽ 35) is significantly lower than in wild type suprathreshold, depolarizing, current steps. The step
(n ⫽ 29) (***P ⬍ 0.001, Student’s t test, unpaired). The reduction depolarizations were separated by intervals of varying
observed in the transgenic neurons in the absence of heat shock
(tipE ⫺:tipE ⫹ ⫺HS; n ⫽ 12) is rescued in the heat-shocked transgenic
duration (Fig. 7A). An example of the typical behavior
neurons (tipE ⫺:tipE ⫹ ⫹HS; n ⫽ 13) (***P ⬍ 0.001, Student’s t test, of a single-spiking transgenic neuron, in the absence of
unpaired) (n ⫽ 7). Bars in all graphs indicate SEM. heat shock, is shown in the top traces in Fig. 7B. The
second action potential was reduced in amplitude when
the two depolarizing pulses were separated by short
interstimulus intervals of 2 and 10 ms (first and second
normalized to that elicited by the first step. Under these pair). The action potential amplitude was similar in the
conditions, the amplitude of the second sodium current two steps when the interstimulus interval was in-
was approximately 90% of the first sodium current, in creased to 100 ms (third pair, Fig. 7B). In contrast, in a
wild-type neurons. In contrast, in tipE ⫺ neurons the multiple-spiking transgenic neuron, following heat
second sodium current was only 80% of the control shock rescue, there was no significant decrement in the
amplitude (Fig. 6B). The reduced level of recovery seen amplitude of the action potential elicited by a second
tipE Regulates Neuronal Excitability 409

step depolarization, even when the interstimulus inter-

val was as short as 2 ms (Fig. 7B, bottom). To quantify
these data, the action potential amplitude elicited by the
second step was normalized to that elicited by the first
step and plotted as a function of interstimulus interval,
for a number of neurons in each firing class (Fig. 7C). As
illustrated in the graph, the interstimulus duration re-
quired for recovery of the ability to fire a full-sized
action potential was longer in the single-spiking than in
the multiple-spiking transgenic neurons. In graded
multiple-spiking neurons the required interstimulus
duration was intermediate to the other two classes. The
longer interstimulus duration necessary to fire a second
full-sized action potential in single- versus graded ver-
sus multiple-spiking classes is consistent with the sug-
gestion that a slowed rate of repolarization-dependent
recovery of sodium currents contributes to the decrease
in repetitive firing in tipE ⫺ neurons.
In addition to sodium currents, voltage-gated potas-
sium currents are critical in determining many aspects
of neuronal excitability, including repetitive firing
properties (Wu and Ganetzky, 1992). Potassium cur-
rents were induced by a series of 300-ms depolarizing
voltage steps between ⫺55 and ⫹55 mV. Individual
neurons in both genotypes had varying levels of tran-
sient and sustained currents. The peak and plateau
FIG. 7. Recovery of excitability in transgenic neurons. (A) Neu- current amplitudes in wild-type and tipE ⫺ neurons
rons in transgenic cultures were held at ⫺55 mV and stimulated
were determined at ⫹55 and normalized to the whole-
with the two-step protocols illustrated using three interstimulus
intervals (2, 10, 100 ms). (B) Whole-cell current clamp recordings cell capacitance. There was no significant difference in
from a single-spiking neuron in the absence of heat shock (tipE ⫺ : the peak (wt 63.9 ⫾ 8 pA/pF, n ⫽ 25; tipE ⫺ 62.6 ⫾ 4
tipE ⫹ ⫺HS) (top trace) and a multiple-spiking neuron in the pres- pA/pF, n ⫽ 48) nor in the plateau current density (wt
ence of heat shock (tipE ⫺ :tipE ⫹ ⫹HS) (bottom trace). In the single- 33.5 ⫾ 3 pA/pF, n ⫽ 22; tipE ⫺ 35.1 ⫾ 2.5 pA/pF, n ⫽
spiking neuron the action potential elicited by the second step in
47) between the two genotypes. Since there were no
each pair was reduced in amplitude for the interstimulus intervals
of 2 and 10 ms. When the interstimulus interval was extended to prior studies indicating that tipE regulates potassium
100 ms the second action potential was similar in amplitude to the channels, we did not compare the properties of voltage-
first. In a multiple-spiking neuron in a transgenic culture that had gated potassium currents in tipE ⫺ and wild-type neu-
been heat shocked, there was no difference in action potential rons further.
amplitude in the first and second steps even when the interstimu-
lus interval was 2 ms. (C) The amplitude of the first action potential
(measured from peak to trough) generated in the second step in Coexpression of tipE and para mRNA
each pair was normalized to the amplitude of the first action
in Wild-Type Neurons
potential in the first step and plotted as a function of interstimulus
interval. Multiple-spiking neurons showed no decrease in the nor- Previous studies from our laboratory have demon-
malized action potential amplitude at any of the interstimulus
strated that the para gene encodes functional sodium
intervals examined. In contrast, the normalized action potential
amplitude was significantly reduced in graded multiple spiking channels in neurons cultured from wild-type Drosophila
and even further reduced in single-spiking neurons at inter- embryos (O’Dowd et al., 1989). If tipE is regulating para
vals shorter than 20 ms. The values for single- and multiple- expression, thereby influencing sodium currents and
spiking classes, between 2 and 12 ms, when evaluated in a firing properties in Drosophila neurons as suggested
point-by-point comparison are significantly different (ANOVA,
from the oocyte studies, then tipE and para must be
P ⬍ 0.001 (2– 8 ms), P ⬍ 0.01 (10 –12 ms), Fisher’s protected least
significant difference). Each data point represents the mean value expressed at the same time and in the same cells. RT-
for the indicated number of neurons in each firing class. Bars PCR with two distinct primer sets (see Experimental
indicate SEM. Methods) was used to examine expression of para and
410 Hodges et al.

tipE in RNA harvested from wild-type cultures between

2 and 48 h in vitro. A third primer set was included to
amplify ribosomal protein 49 (O’Connell and Rosbash,
1984), serving as a control for normalization of input
RNA amounts. The autoradiogram in Fig. 8A illustrates
the PCR products generated from total RNA harvested
from wild-type neurons at the indicated times in cul-
ture. The percentage of the maximal levels of expres-
sion, after normalization to rp49 values, is shown in Fig.
8B. Expression of both genes was initiated about 12 h
after plating, peaked at 33 h, and subsequently declined
to approximately 50 –70% of maximum by 48 h (Fig. 8B).
The similar time courses of tipE and para expression are
consistent with a role for tipE in regulating para sodium
channels during normal neuronal development.
To determine if para and tipE are coexpressed in
individual neurons, we used a multiplex, single-cell
RT-PCR approach to examine their expression in wild-
type neurons at 2 days in culture. These analyses were
performed with two primer pairs in a single RT-PCR.
One primer pair (paraDP3/DP4) flanked the alterna-
tively spliced exons i and a in the para gene, amplifying
distinct PCR products representing the para splice vari-
ants containing one or both alternative exons i and a.
The second primer set (tipEComF/R) amplified a single
PCR product from the tipE transcripts. A representative
autoradiogram of PCR products amplified from RNA
harvested from individual cultured neurons is shown
FIG. 8. tipE and para sodium channel mRNA expressions exhibit
in Fig. 8C. As previously reported there was cell-to-cell similar temporal and spatial distributions in cultured neurons.
variability in expression of para splice variants. The (A) Autoradiogram of the RT/PCR products separated on an
majority of cells that express sodium currents and para 8% polyacrylamide gel reveals similar temporal patterns of expres-
mRNA also express tipE (Fig. 8C). In six experiments, in sion of tipE and para mRNA in the cultured neurons. rp49 is
observed at a constant level at all of the ages examined, indicating
which eight or more neurons were examined/experi-
that expression of this gene is not developmentally regulated and
ment, 92.2 ⫾ 3.1% of the para-expressing neurons also that starting RNA template concentrations were similar in each
expressed tipE. These data demonstrate that the para sample. Lane W represents a negative control in which water is
and tipE gene products could interact directly or indi- substituted for RNA. PCR products were generated from total
rectly, in single neurons, to influence repetitive firing. RNA prepared from cultured neurons using three individual
primer sets that amplify rp49, para, and tipE transcripts. (B) For
temporal comparison the product levels for tipE and para were
normalized to the maximum levels observed for each. Data were
obtained from five independent RNA preparations at each age.
DISCUSSION Bars indicate SEMs. (C) Autoradiogram from a typical experiment
showing strong amplification of one or more of three alternatively
spliced para PCR products from 7/10 cells, with 6/7 of these
Although little is known about the firing properties of
neurons also expressing the tipE product. In addition, low levels of
Drosophila neurons in vivo (Ikeda and Kaplan, 1970; both para and tipE are seen in one additional neuron (Cell 9). The
Tanouye et al., 1981) studies in cell culture reveal sub- experiment was accepted for analysis on the basis of clean inter-
populations of neurons with distinct firing patterns elic- mingled medium (M) and water (W) controls that were processed
ited by depolarizing current pulses (O’Dowd, 1995; in parallel with the cells. PCR products from total RNA harvested
from whole pupae (P) are shown for comparison. Amplification of
Zhao and Wu, 1997). Many of the multiple-spiking
the cell contents was performed using 32 P-labeled primers that
neurons in this study have spike discharge patterns that amplify the region surrounding alternative para exons i and a
closely resemble those of the well-characterized regu- (paraDP3/DP4) and tipE primers that detect all transcripts from
lar-spiking or fast-spiking firing classes of mammalian the tipE region.
tipE Regulates Neuronal Excitability 411

cortical neurons described in the animal and in disso- tipE Regulates Sodium-Dependent
ciated cell culture (Connors and Gutnick, 1990; Massen- Repetitive Firing
gill et al., 1997). The single-spiking neurons in the Dro-
Assessment of the firing properties in primary neu-
sophila cultures are similar to the recently described
rons from genetic mutants is a useful strategy for ex-
on-spiking neurons found in the rodent auditory cortex,
amining the role of specific genes in regulating neuro-
which fire only one or two spikes that occur within 10
nal excitability. Alterations in spontaneous activity of
ms of the onset of a maintained intracellular depolar-
neurons cultured from Hyperkinetic mutant embryos
ization (Metherate and Aramakis, 1999). The presence
(Yao and Wu, 1999) supported an early study indicating
of similar firing classes in the cultured Drosophila neu-
that this gene, encoding a K channel ␤ subunit, is in-
rons and rodent cortical neurons suggests strong con- volved in regulation of neuronal firing properties
servation of the functional elements contributing to (Ikeda and Kaplan, 1970). Our analysis of tipE ⫺ neurons
CNS circuitry between these distantly related species. revealed reductions in repetitive firing, spontaneous
It should also be noted that the resting potentials of firing, action potential amplitude, peak sodium current
the Drosophila neurons reported in this study are more density, and sodium current recovery during repeated
depolarized than is standard for many mature mamma- activation, suggesting that these are linked to each other
lian neurons. However, hyperpolarizing shifts in mem- and to tipE. Rescue experiments, involving expression
brane potential, from ⫺40 to ⫺65 mV, have been re- of the wild-type tipE transgene in tipE ⫺ neurons, con-
ported during early development in some populations firmed that tipE is important in regulation of repetitive
of mammalian cortical neurons (Agmon et al., 1996; firing, spontaneous firing, and the rate of recovery of
Zhou and Hablitz, 1996). This suggests that the depo- sodium currents during repeated activation. Our data
larized resting potentials could be related to the rela- also demonstrate that induction of tipE ⫹ expression in
tively young age at which most of the recordings were transgenic neurons beginning at 2 days, after neurons
obtained, 2–3 days of the neuronal birth date. More have already established their firing properties, is suf-
negative resting potentials of ⫺55 mV have been re- ficient to rescue the mutant firing phenotypes. This
ported for Drosophila “giant neurons” examined at suggests that regulation of tipE may play a role, not
slightly later stages, between 2 and 5 days in culture only in establishment of neuronal firing phenotype, but
(Yao and Wu, 1999). In addition, we have observed also in modulation of firing properties in differentiated
more hyperpolarized resting potentials (⫺55 mV) when neurons.
recordings are done at 3– 4 days (unpublished data). The slower rate of recovery of sodium currents dur-
Despite the depolarized resting potentials, intracellu- ing repetitive activation in tipE ⫺ neurons predicts that a
lar (whole cell) recordings revealed spontaneous action diminished sodium current will be available for gener-
potentials in the absence of current injection in some ation of the second spike in an action potential train in
neurons. This does not seem likely to be injury-induced the mutant neurons. This could thus contribute to the
spiking as spontaneously active neurons were observed decrease in probability of mutant neurons firing repet-
at a similar frequency in extracellular (cell attached) itively during sustained depolarization. Concomitant
recordings. Previous studies from our lab have also rescue of sodium current recovery and repetitive firing,
demonstrated the presence of action potential mediated following induction of the tipE ⫹ transgene in tipE ⫺ neu-
spontaneous excitatory postsynaptic currents in many rons, suggests linkage between these two phenotypes.
of these cultured neurons, in which activity in the pre- The difference in the level of recovery of sodium cur-
synaptic neuron is clearly independent of technical rents during repolarization seen between wild-type and
artifacts that could be potentially associated with mutant neurons, though significant, was not large (ap-
whole-cell recording electrodes (Lee and O’Dowd, proximately 10%), and therefore it was not clear how
1999). Finally, recordings from neurons in the Drosoph- this property might influence repetitive firing rates.
ila embryonic nerve cord have revealed large spontane- However, analysis of the recovery of excitability as a
ously active currents, thought to underlie action poten- function of interstimulus interval in the different firing
tials, in neurons held at ⫺40 mV, that were rarely seen classes is consistent with the suggestion that reduced
in those held at more hyperpolarized potentials (Baines rate of recovery of sodium currents contributes to the
and Bate, 1998). Together these findings suggest that decrease in repetitive firing in mutant neurons. In sin-
young embryonic Drosophila neurons, both in vivo and gle-spiking neurons, an interstimulus interval was re-
in vitro, are excitable at relatively depolarized voltages. quired for recovery of the ability to fire a second action
412 Hodges et al.

potential. In addition, the duration of the interstimulus logues may identify novel pathways involved in regu-
interval necessary to fire a second full-sized action po- lation of sodium currents that can influence action po-
tential was significantly longer in single- versus multi- tential propagation in mammalian neurons.
ple-spiking transgenic neurons. Most of the spontaneously firing neurons in wild-
In Drosophila, as in mammals, the sodium channels type cultures were in the multiple-spiking class. Alter-
that underlie the whole-cell sodium currents are tran- ations that decrease the probability of firing a second
siently activated by a sustained depolarizing voltage spike in the mutant neurons in response to depolariza-
step and recovery from inactivation requires return to tion could also decrease the probability of firing
hyperpolarized potentials (O’Dowd and Aldrich, 1988). spontaneously. However, additional changes in the un-
A decrease in the rate of recovery from inactivation of derlying currents may contribute to the reduced spon-
the underlying sodium channels is one mechanism that taneous activity in the mutant neurons. For example, in
could contribute to the reduced recovery of sodium oocytes, coexpression of the wild-type tipE product in-
currents seen in the tipE ⫺ neurons. Studies in other fluenced both the density and the fast decay kinetics of
systems have clearly demonstrated a relationship be- the para sodium currents (Warmke et al., 1997). The fast
tween rate of recovery of sodium channels from inacti- kinetic properties of sodium currents were not assessed
vation and repetitive firing. In hippocampal pyramidal in the present study due to inadequate voltage-clamp in
neurons spikes in the dendrites are attenuated by high- excitable cells. Therefore, tipE might also affect fast
frequency stimulation and this is correlated with a rel- gating properties of sodium channels that could con-
atively slow rate of recovery of sodium channels from tribute to the altered firing phenotypes observed.
inactivation (Colbert et al., 1997; Jung et al., 1997). A The oocyte studies further suggested that tipE might
computational model supports the hypothesis that de- be functioning like sodium channel ␤ subunits (␤1 and
layed recovery of sodium channels from inactivation ␤2) as these are known to influence both expression
can result in attenuation of action potentials (Migliore, levels and fast kinetic properties of mammalian sodium
1996). Additionally, hyperexcitability characterized by channels (Isom et al., 1994). A newly identified ␤ sub-
elevated firing frequencies in spinal sensory neurons unit (␤3), cloned from human and rat, has been shown
following injury has been associated with the emer- to influence the rate of sodium current recovery from
gence of sodium currents that recover rapidly from inactivation (Morgan et al., 2000), similar to the role
inactivation (Cummins and Waxman, 1997; Cummins et suggested for tipE by the present study. Our single-cell
al., 2000). However, in the present study the majority of RT-PCR analyses demonstrate that tipE is coexpressed
the data on sodium currents were obtained from neu- with para in most cells, and coimmunoprecipitation in
rons that could not be well voltage-clamped. Therefore, Xenopus oocytes suggests that the two proteins can
we cannot rule out the possibility that a use-dependent physically associate (L. M. Hall and C. Ericsson, unpub-
change in space constant, rather than a change in the lished results). Taken together these data suggest that,
sodium channel inactivation properties, could contrib- although tipE has little amino acid sequence identity
ute to the observed decrease in recovery of the currents. with sodium channel ␤ subunits, it could be functioning
For example, a failure to reach the same membrane as an auxiliary subunit important in regulating sodium
potential during the two sequential depolarizing steps channel function in wild-type Drosophila neurons. A
could cause a reduction in amplitude of the sodium prediction of this hypothesis is that wild-type neurons
current evoked by the second pulse. We do not believe that fire multiple spikes express more tipE than those
this was a factor since the latency and waveform of the that fire only single action potentials. A quantitative
currents, also influenced by space constant, did not analysis of gene and/or protein levels, not undertaken
vary significantly between the two steps (Fig. 6B). In in the present studies, would be necessary to address
addition, for this mechanism to account for the differ- this question.
ences seen between tipE ⫺ and wild-type neurons and The inability of wild-type tipE transgene expression
the rescue by tipE ⫹, it would necessitate invoking to rescue the reduced sodium current density and ac-
genotype-specific differences in the properties of use- tion potential amplitude in transgenic neurons was sur-
dependent alterations in space clamp. In either case, our prising. It is possible that these features, while related to
rescue studies clearly demonstrate that tipE is impor- each other, are not necessarily linked to tipE. However,
tant for regulating recovery of sodium currents from we cannot rule out the possibility that tipE plays a role
repeated activation and sodium-dependent repetitive in regulation of sodium current density and action po-
firing. Therefore, isolation of vertebrate tipE ortho- tential amplitude in primary neurons. For example,
tipE Regulates Neuronal Excitability 413

induction conditions or the timing of the assay could be stimulation due to the reduced sodium current density
suboptimal for detecting regulation mechanisms in- and depressed recovery of sodium currents during re-
volving coassembly of tipE products with para sodium peated stimulation. Furthermore, a rise in temperature
channels prior to membrane insertion. In either case, speeds up the gating kinetics of all channels and may
rescue of the repetitive firing phenotype in the absence result in potassium currents overwhelming the altered
of restoration of sodium current density and action sodium currents leading to an even more pronounced
potential amplitude demonstrates that these can be alteration at elevated temperatures. This could contrib-
functionally separated. ute to the temperature-induced paralysis. The reduced
sodium current density and altered repolarization-
dependent sodium current recovery in tipE ⫺ mutant
tipE ⴙ Is Not Necessary for Repetitive Spiking
neurons could also contribute to the enhanced sensitiv-
in All Drosophila Neurons
ity to temperature-induced action potential blockade
Repetitive firing, spontaneous activity, and fast re- previously reported in the para;tipE double mutants
covery of sodium currents from repeated activation in (Ganetzky, 1986).
some of the tipE ⫺ neurons demonstrate that tipE ⫹ ex-
pression is not necessary for manifestation of these
electrophysiological phenotypes in all cultured neu- EXPERIMENTAL METHODS
rons. It is possible that a tipE homologue, identified in a
recent analysis of the Drosophila genome (Littleton and Drosophila stocks and cell culture. Embryos were
Ganetzky, 2000), encodes a protein that substitutes for collected from Canton-S homozygous wild-type, tipE
the mutant tipE in cells that fire repetitively. Alterna- sepia (tipE ⫺), and w;tipE sepia flies transformed with a
tively, the tipE mutant used in this study, an EMS- wild-type tipE cDNA under control of the heat-shock
induced recessive mutation (Kulkarni and Padhye, promoter (tipE ⫺:tipE ⫹) (Feng et al., 1995b). Neurons
1982) resulting in a premature stop codon, may act as a were prepared from midgastrula-stage embryos and
hypomorph rather than a true null (Feng et al., 1995b). cultured in Drosophila defined medium 1 (DDM1) at
The tipE ⫺ neurons with apparently wild-type properties 22–25°C and 4 –5% CO 2, as previously described
could be due to residual function of the mutant protein. (O’Dowd, 1995). Cultures stained with anti-horseradish
This second possibility seems less likely because previ- peroxidase (HRP) antibodies were fixed in 4% parafor-
ous studies have shown that mutant tipE was not able to maldehyde for 30 min at room temperature followed by
rescue adult paralysis (Feng et al., 1995b). Additionally, a 1-h incubation with fluorescein-conjugated anti-HRP
mutant tipE cRNA expressed in Xenopus oocytes does antibodies (1:100; Organon Teknika). Coverslips were
not enhance para sodium current expression (M. Chopra mounted on glass slides. Images were acquired with a
and L. M. Hall, unpublished observations). Spot cooled CCD camera (Diagnostic Instruments)
mounted on a Nikon Optiphot microscope and pre-
pared for presentation in Adobe PhotoShop.
Functional Significance
Electrophysiological recordings. To minimize po-
In nap and para mutants, a temperature-dependent tential bias in selection of cells for analysis, whole-cell
blockade of action potential propagation in larval motor recordings from wild-type, tipE ⫺, and tipE ⫺:tipE ⫹ trans-
nerves has been associated with the temperature-sensi- gene neurons were performed blind with respect to
tive paralysis (Wu and Ganetzky, 1992). In contrast, it genotype and heat-shock treatment. Unpolished re-
was unclear why tipE ⫺ larvae exhibit normal extracel- cording pipettes had open pipette resistances of 2–5
lularly recorded action potential propagation in motor M⍀. For assessment of firing properties the internal
nerves both at the behaviorally permissive and at the pipette solution contained (in mM): potassium glu-
nonpermissive temperature (Ganetzky, 1986). Our data conate (120), NaCl (20), EGTA (1.1), CaCl 2 (0.1), MgCl 2
demonstrate that tipE ⫺ neurons are capable of generat- (2), Hepes (10), pH 7.2. Sodium currents were examined
ing a single action potential in response to a discrete using an internal solution containing (in mM): d-glu-
stimulus, consistent with the apparently normal com- conic acid (120), cesium hydroxide (120), NaCl (20),
pound action potential recorded in larval motor neu- CaCl 2 (0.1), MgCl 2 (2), EGTA (1.1), Hepes (10), pH 7.2.
rons. However, our findings suggest that action poten- The external solution contained (in mM): NaCl (140),
tial propagation in tipE ⫺ mutant nerves could be KCl (3), MgCl 2 (4), CaCl 2 (1), Hepes (5), pH 7.2. A 5-mV
compromised during high-frequency, repetitive nerve liquid junction potential has been subtracted from all
414 Hodges et al.

TABLE 3
Primer Pairs Used in RT-PCR Studies

Primer name Sequence Nucleotide positions GenBank accession No.

paraComF TGCCATGTCGTATGACGAATTG 1555–1576 M32078

paraComR TCTCGCCGCCAACAAATAGC 1796–1777 M32078
paraDP3 ATGTCCATTCGGAGCGTCGA 1826–1845 M32078
paraDP4 CTGGGCATCCTGATATGTTGACA 2083–2061 M32078
tipEComF AAACTTGAGCAAGACGATGACGAC 2341–2364 U27561
tipEComR TCTTTTTCGGGTTGGGTCTCC 2545–2525 U27561
tipEM1 TACTATGTGGGAGCCAGGCT 1694–1713 U27561
tipEM2 GAGTAGTAGCAGGGGAACTTCATGC 1815–1791 U27561
rp49F AAGATGACCATCCGCCCAGCATAC 417–440 X00848
rp49R CTCGTTCTTCTTGAGACGCAGG 876–855 X00848

membrane potentials noted in this report. Whole-cell 1995) using the primer pairs shown in Table 3. Ampli-
capacitance was determined by measuring the area un- fied products, visualized by inclusion of 2–5 ⫻ 10 5 dpm
der the capacitative transient current record obtained of 32P-end-labeled forward primers in the PCR, were
immediately after break into the cell. Data were col- separated by electrophoresis on 8 or 10% nondenatur-
lected and analyzed using a List EPC-7 patch-clamp ing polyacrylamide gels. The amount of product was
amplifier, a Dell computer, and pCLAMP software quantified by phosphorimager analysis (Molecular Dy-
(Axon Instruments). All recordings were performed at namics, Sunnyvale, CA).
room temperature. Identification of wild-type and mutant tipE PCR
Heat-shock induction of tipE ⴙ expression in trans- products was performed by RsaI restriction enzyme
genic neurons. Cultures were prepared from midgas- analysis of an aliquot of the PCR products using stan-
trula-stage embryos obtained from wild-type, tipE ⫺, dard procedures (Sambrook et al., 1989). In the devel-
and tipE ⫺:tipE ⫹ flies. For PCR analysis of gene expres- opmental study a single reverse transcription reaction
sion, half of the cultures from each genotype were heat was performed on each RNA sample for each time
shocked by transfer to a 37°C, 5% CO 2 incubator for 1 h point. This was divided into three equal aliquots in
at 16, 33, and 40 h after plating. The remainder of the which PCR products were amplified using primers spe-
time they were maintained at ambient temperature (22– cific for ribosomal protein 49 (rp49) (21 cycles) or para or
25°C). The sibling cultures were maintained continu- tipE (25 cycles). Cycle numbers were chosen to yield
ously in a 5% CO 2 incubator at ambient temperature for products within the linear range of amplification. To
42 h. Total RNA was extracted at 42 h (2 h after the last minimize differences in reaction conditions, primers of
heat shock) from both control and heat-shocked cul- similar size and specific activities were used. Phospho-
tures. For the electrophysiological studies, half of the imager optical density measurements for developmen-
cultures from each genotype were heat shocked by tally regulated PCR products were normalized to opti-
transfer to a 37°C, 5% CO 2 incubator for 1 h at 42 and cal density values obtained from PCR amplification of
49 h after plating. The remainder of the time they were rp49, a mRNA that is not developmentally regulated
maintained in a 5% CO 2 incubator at ambient temper- (O’Connell and Rosbash, 1984). Single-cell amplifica-
ature. The sibling cultures were maintained continu- tion of total RNA aspirated from neurons after electro-
ously at ambient temperature. All electrophysiological physiological recordings was performed as previously
recordings were done at 66 –74 h after plating. described (O’Dowd et al., 1995).
RT-PCR analysis of gene expression in cultured neu- Primer pairs. para: To amplify a single product
rons. Total RNA from cultured neurons was prepared common to all para transcripts the primer set para-
using Tri-Reagent (Molecular Research Center, Inc., ComF/R was used. To examine the distribution of para
Cincinnati, OH) according to a single-step method transcripts containing alternatively spliced exons a and
(Chomczynski and Sacchi, 1987). First-strand cDNA i, a primer pair (paraDP3/DP4) flanking these exons
was generated by random-primed reverse transcription was used. tipE: For developmental profiles and single-
of total RNA, and PCR amplification of the cDNA was cell experiments, PCR amplification of tipE mRNA was
performed as previously described (O’Dowd et al., performed using the primer pair tipEComF/R. To dif-
tipE Regulates Neuronal Excitability 415

ferentiate between wild-type and tipE mutant mRNA, Feng, G., Deak, P., Kasbekar, D. P., Gil, D. W., and Hall, L. M. (1995b).
primers tipEM1/M2 were used. rp49: rp49F/R primer Cytogenetic and molecular localization of tipE: A gene affecting
sodium channels in Drosophila melanogaster. Genetics 139: 1679 –
pair was used for PCR amplification of ribosomal pro- 1688.
tein transcripts. Sequences, nucleotide positions, and Ganetzky, B. (1986). Neurogenetic analysis of Drosophila mutations
GenBank accession numbers for all of the primer sets affecting sodium channels: Synergistic effects on viability and
used are detailed in Table 3. nerve conduction in double mutants involving tipE. J. Neurogenet. 3:
19 –31.
Huguenard, J. R., Hamill, O. P., and Prince, D. A. (1988). Develop-
mental changes in sodium conductances in rat neocortical neurons:
ACKNOWLEDGMENTS Appearance of a slowly inactivating component. J. Neurophysiol. 59:
778 –794.
This work was supported by NIH Grants NS27501 and NS01854 to Ikeda, K., and Kaplan, W. (1970). Patterned neural activity of a mutant
D.K.O’D., NIH Grant NS16204 and an American Cancer Society Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 66: 765–772.
Scholar Award to L.M.H., and American Heart Association Postdoc- Isom, L. L., DeJongh, K. S., and Catterall, W. A. (1994). Auxiliary
toral Grant 95-98 to D.D.H. subunits of voltage-gated ion channels. Neuron 12: 1183–1194.
Jackson, F. R., Wilson, S. D., and Hall, L. M. (1986). The tip-E muta-
tions of Drosophila decrease saxitoxin binding and interact with
other mutations affecting nerve membrane excitability. J. Neuro-
REFERENCES genet. 3: 19 –31.
Jung, H. Y., Mickus, T., and Spruston, N. (1997). Prolonged sodium
Agmon, A., Hollrigel, G., and O’Dowd, D. K. (1996). Functional channel inactivation contributes to dendritic action potential atten-
GABAergic synaptic connections in neonatal mouse barrel cortex. uation in hippocampal pyramidal neurons. J. Neurosci. 17: 6639 –
J. Neurosci. 16: 4684 – 4695. 6646.
Aizenman, C. D., and Linden, D. J. (2000). Rapid, synaptically driven Kulkarni, S. J., and Padhye, A. (1982). Temperature-sensitive paralytic
increases in the intrinsic excitability of cerebellar deep nuclear mutations on the second and third chromosomes of Drosophila
neurons. Nat. Neurosci. 3: 109 –111. melanogaster. Genet. Res. 40: 191–199.
Armano, S., Rossi, P., Taglietti, V., and D’Angelo, E. (2000). Long-term Lee, D., and O’Dowd, D. K. (1999). Fast excitatory synaptic transmis-
potentiation of intrinsic excitability at the mossy fiber-granule cell sion mediated by nicotinic acetylcholine receptors in Drosophila
synapse of rat cerebellum. J. Neurosci. 20: 5208 –5216. neurons. J. Neurosci. 19: 5311–5321.
Baines, R. A., and Bate, M. (1998). Electrophysiological development Littleton, J. T., and Ganetzky, B. (2000). Ion channels and synaptic
of central neurons in the Drosophila embryo. J. Neurosci. 18: 4673– organization: Analysis of the Drosophila genome. Neuron 26: 35– 43.
4683. Loughney, K., Kreber, R., and Ganetzky, B. (1989). Molecular analysis
Barish, M. E. (1986). Differentiation of voltage-gated potassium cur- of the para locus, a sodium channel gene in Drosophila. Cell 58:
rent and modulation of excitability in cultured amphibian spinal 1143–1154.
neurones. J. Physiol. 375: 225–229. Massengill, J. L., Smith, M. A., Son, D. I., and O’Dowd, D. K. (1997).
Catterall, W. A. (2000). From ionic currents to molecular mechanisms: Differential expression of K4-AP currents and Kv3.1 potassium
The structure and function of voltage-gated sodium channels. Neu- channel transcripts in cortical neurons that develop distinct firing
ron 26: 13–25. phenotypes. J. Neurosci. 17: 3136 –3147.
Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA Metherate, R., and Aramakis, V. B. (1999). Intrinsic electrophysiology
isolation by acid guanidinium thiocyanate–phenol– chloroform ex- of neurons in thalamorecipient layers of developing rat auditory
traction. Anal. Biochem. 162: 156 –159. cortex. Dev. Brain. Res. 115: 131–144.
Colbert, C. M., Magee, J. C., Hoffman, D. A., and Johnston, D. (1997). Migliore, M. (1996). Modeling the attenuation and failure of action
Slow recovery from inactivation of Na ⫹ channels underlies the potentials in the dendrites of hippocampal neurons. Biophys. J. 71:
activity-dependent attenuation of dendritic action potentials in hip- 2394 – 403.
pocampal CA1 pyramidal neurons. J. Neurosci. 17: 6512– 6521. Morgan, K., Stevens, E., Shah, B., Cox, P., Kixon, A., Lee, K., Pinnock,
Connors, B. W., and Gutnick, M. J. (1990). Intrinsic firing patterns of R., Hughes, J., Richardson, P., Mizuguchi, K., and Jackson, A.
diverse neocortical neurons. Trends Neurosci. 13: 99 –104. (2000). ␤3: An additional auxiliary subunit of the voltage-sensitive
Cummins, T., Dib-Hajj, S., Black, J., and Waxman, S. (2000). Sodium sodium channel that modulates channel gating with distinct kinet-
channels and the molecular pathophysiology of pain. In Progress in ics. Proc. Natl. Acad. Sci. USA 97: 2308 –2313.
Brain Research, pp. 3–19. Elsevier, Amsterdam. O’Connell, P. O., and Rosbash, M. (1984). Sequence, structure, and
Cummins, T., and Waxman, S. (1997). Downregulation of tetrodo- codon preference of the Drosophila ribosomal protein 49 gene.
toxin-resistant sodium currents and upregulation of rapidly Nucleic Acids Res. 12: 5495–5513.
repriming tetrodotoxin-sensitive sodium current in small spinal O’Dowd, D. K. (1995). Voltage-gated currents and firing properties of
sensory neurons after nerve injury. J. Neurosci. 17: 3503–3514. embryonic Drosophila neurons grown in a chemically defined me-
Desai, N. S., Rutherford, L. C., and Turrigiano, G. G. (1999). Plasticity dium. J. Neurobiol. 27: 113–126.
in the intrinsic excitability of cortical pyramidal neurons. Nat. Neu- O’Dowd, D. K., and Aldrich, R. W. (1988). Voltage-clamp analysis of
rosci. 2: 515–520. sodium channels in wildtype and mutant Drosophila neurons.
Feng, G., Deák, P., Chopra, M., and Hall, L. M. (1995a). Cloning and J. Neurosci. 8: 3633–3643.
functional analysis of tipE, a novel membrane protein that enhances O’Dowd, D. K., Gee, J. R., and Smith, M. A. (1995). Sodium current
Drosophila para sodium channel function. Cell 82: 1001–1011. density correlates with expression of specific alternatively spliced
416 Hodges et al.

sodium channel mRNAs in single neurons. J. Neurosci. 15: 4005– potentials associated with the Shaker complex locus of Drosophila.
4012. Proc. Natl. Acad. Sci. USA 78: 6548 – 6552.
O’Dowd, D. K., Germeraad, S., and Aldrich, R. W. (1989). Alterations Turrigiano, G., LeMasson, G., and Marder, E. (1995). Selective regu-
in the expression and gating of Drosophila sodium channels by lation of current densities underlies spontaneous changes in the
mutations in the para gene. Neuron 2: 1301–1311. activity of cultured neurons. J. Neurosci. 15: 3640 –3652.
O’Dowd, D. K., Ribera, A. B., and Spitzer, N. C. (1988). Development Warmke, J., Reenan, R., Wang, P., Qian, S., Arena, J., Wang, J.,
of voltage-dependent calcium, sodium and potassium currents in Wunderler, D., Liu, K., Kaczorowski, G., Wan der Ploeg, L.,
Xenopus spinal neurons. J. Neurosci. 8: 792– 805. Ganetzky, B., and Cohen, C. (1997). Functional expression of Dro-
Reenan, R. A., Hanrahan, C. J., and Ganetzky, B. (2000). The mle napts sophila para sodium channels: Modulation by the membrane pro-
RNA helicase mutation in Drosophila results in a splicing catastro-
tein tipE and toxin pharmacology. J. Gen. Physiol. 110: 119 –133.
phe of the para Na ⫹ channel transcript in a region of RNA editing.
Wu, C.-F., and Ganetzky, B. (1992). Neurogenetic studies of ion chan-
Neuron 25: 139 –149.
nels in Drosophila. In Ion Channels, pp. 261–314. Plenum, New York.
Sambrook, J. M., Fritsch, E. F., and Maniatis, T. (1989). Molecular
Yao, W. D., and Wu, C. F. (1999). Auxiliary hyperkinetic beta subunit
Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press,
Cold Spring Harbor, New York. of K ⫹ channels: Regulation of firing properties and K ⫹ currents in
Spitzer, N. C. (1991). A developmental handshake: Neuronal control Drosophila neurons. J. Neurophysiol. 81: 2472–2484.
of ionic currents and their control of neuronal differentiation. Zhao, M.-L., and Wu, C.-F. (1997). Alterations in frequency coding
J. Neurobiol. 22: 659 – 673. and activity dependence of excitability in cultured neurons of Dro-
Spitzer, N. C., Gu, X., and Olson, E. (1994). Action potentials, calcium sophila memory mutants. J. Neurosci. 17: 2187–2199.
transients and the control of differentiation of excitable cells. Curr. Zhou, F.-M., and Hablitz, J. J. (1996). Postnatal development of mem-
Opin. Neurobiol. 4: 70 –77. brane properties of layer I neurons in rat neocortex. J. Neurosci. 16:
Tanouye, M. A., Ferrus, A., and Fujita, S. C. (1981). Abnormal action 1131–1139.

Received August 21, 2001

Revised November 30, 2001
Accepted December 7, 2001