Electrical Silencing of PDF Neurons Advances

the Phase of non-PDF Clock Neurons in Drosophila

Ying Wu, Guan Cao, and Michael N. Nitabach1

Cellular & Molecular Physiology, Yale School of Medicine, New Haven, CT

Abstract Drosophila clock neurons exhibit self-sustaining cellular oscillations

that rely in part on rhythmic transcriptional feedback loops. We have previously
determined that electrical silencing of the pigment dispersing factor (PDF)-
expressing lateral-ventral (LNV) pacemaker subset of fly clock neurons via
expression of an inward-rectifier K+ channel (Kir2.1) severely disrupts free-
running rhythms of locomotor activity—most flies are arrhythmic and those that
are not exhibit weak short-period rhythms—and abolishes LNV molecular oscil-
lation in constant darkness. PDF is known to be an important LNV output signal.
Here we examine the effects of electrical silencing of the LNV pacemakers on mol-
ecular rhythms in other, nonsilenced, subsets of clock neurons. In contrast to pre-
viously described cell-autonomous abolition of free-running molecular rhythms,
we find that electrical silencing of the LNV pacemakers via Kir2.1 expression does
not impair molecular rhythms in LND, DN1, and DN2 subsets of clock neurons.
However, free-running molecular rhythms in these non-LNV clock neurons occur
with advanced phase. Electrical silencing of LNVs phenocopies PDF null muta-
tion (pdf 01) at both behavioral and molecular levels except for the complete abo-
lition of free-running cellular oscillation in the LNVs themselves. LNV electrically
silenced or pdf 01 flies exhibit weak free-running behavioral rhythms with short
period, and the molecular oscillation in non-LNV neurons phase advances in con-
stant darkness. That LNV electrical silencing leads to the same behavioral and
non-LNV molecular phenotypes as pdf 01 suggests that persistence of LNV molec-
ular oscillation in pdf 01 flies has no functional effect, either on behavioral rhythms
or on non-LNV molecular rhythms. We thus conclude that functionally relevant
signals from LNVs to non-LNV clock neurons and other downstream targets rely
both on PDF signaling and LNV electrical activity, and that LNVs do not ordinar-
ily send functionally relevant signals via PDF-independent mechanisms.

Key words pacemaker neuron, cellular oscillation, clock protein, phase advance, K+
channels, dose-dependent

Nearly all organisms that have been examined pos- activity in Drosophila melanogaster are controlled by
sess an intrinsic circadian timekeeping mechanism autonomous cellular oscillation of “clock neurons”
used to synchronize physiological functions with the whose activity patterns drive behavioral rhythms
rotation of the earth. Circadian rhythms of locomotor (Reppert and Weaver, 2000; Williams and Sehgal, 2001;

1. To whom all correspondence should be addressed: Michael N. Nitabach, Cellular & Molecular Physiology, Yale School of
Medicine, 333 Cedar Street, New Haven, CT 06520; e-mail: michael.nitabach@yale.edu.

JOURNAL OF BIOLOGICAL RHYTHMS, Vol. 23 No. 2, April 2008 117-128

DOI: 10.1177/0748730407312984
© 2008 Sage Publications

117
118 JOURNAL OF BIOLOGICAL RHYTHMS / April 2008
Young and Kay, 2001). Cellular circadian oscillation
depends in part on multiple interlocking transcrip-
tional feedback loops involving the genes period (per),
timeless (tim), vrille (vri), and Par domain protein 1
(Pdp1), which encode transcription factors (Sehgal
et al., 1994; Gekakis et al., 1995; Myers et al., 1995;
Sehgal et al., 1995; Blau and Young, 1999; Cyran et al.,
2003). PERIOD (PER) and TIMELESS (TIM) proteins
serve to inhibit transcription of the genes that encode
them, and thereby constitute the core of a negative
transcriptional feedback loop essential for cellular
oscillation (Sehgal et al., 1999; Young and Kay, 2001).
Our previous work has demonstrated that, in addition
to these transcriptional feedback loops, free-running
(but not environmentally entrained) cellular rhythms
also rely on depolarization-dependent events occur-
ring at the clock neuron plasma membrane (Nitabach
et al., 2002, 2005).
In addition to cellular rhythmicity per se, the coor-
dination of multiple autonomous cellular oscillations
occurring in multiple anatomically and neurochemi-
cally distinguishable subsets of clock neurons is
required for normal circadian behavior (Helfrich-
Forster, 1998; Renn et al., 1999; Helfrich-Forster et al.,
2000; Kaneko et al., 2000; Blanchardon et al., 2001;
Peng et al., 2003; Lin et al., 2004; Nitabach et al., 2006).
A number of recent studies indicate that distinct
groups of clock neurons function autonomously to
control the morning and evening peaks of anticipatory
locomotor activity (Grima et al., 2004; Stoleru et al.,
2004). Furthermore, coordinated cellular oscillations
among these groups of neurons are required for nor-
mal free-running circadian behavior (Lin et al., 2004;
Stoleru et al., 2005; Nitabach et al., 2006). Pdf 01-null
mutant flies, LNV-ablated flies, or LNV-electrically
silenced flies exhibit severe defects in free-running
rhythmicity (Helfrich-Forster, 1998; Renn et al., 1999;
Nitabach et al., 2002), suggesting a key role for the pig-
ment dispersing factor (PDF)-expressing lateral-
ventral subset of clock neurons (LNV) in circadian
behavior. Cyclic release of PDF from the sLNV dorso-
medial terminals is thought to provide an important
circadian signal to PDF sensitive downstream targets
(Helfrich-Forster et al., 2000; Park et al., 2000; Lin et al.,
2004; Nitabach et al., 2006). The identification of the
PDF receptor (PDFR) further confirms the key role of
PDF neuropeptide in circadian control of locomotor
activity, because PDFR mutant flies exhibit severe
deficits in free-running locomotor rhythms similar to
those seen in pdf 01 flies (Hyun et al., 2005; Lear et al.,
2005; Mertens et al., 2005).
Interestingly, the study of pdf 01-null mutant flies indi-
cates that PDF release from LNV pacemaker neurons is
not necessary for the persistence of free-running clock
protein oscillation in clock neurons (Lin et al., 2004). In
the absence of PDF, the phase relationships of indepen-
dent cellular oscillators disperse and lose synchrony,
thus explaining, at least in part, the severe deficits in
free-running locomotor rhythms exhibited by pdf 01
flies. Rendering LNVs hyperexcitable via expression of a
slowly inactivating bacterial Na+ channel interferes
with cyclic release of PDF, and results in desynchrony
among subsets of clock neurons and complex behav-
ioral rhythms (Nitabach et al., 2006). This suggests that
properly regulated neuronal excitability is necessary for
cyclic PDF release from LNV pacemaker neurons and
consequent proper coordination of cellular rhythmicity
among the various subsets of clock neurons.
We have previously shown that electrical silencing
of LNV pacemaker neurons by targeted expression of
an open rectifier or inward rectifier K+ channel stops
free-running PERIOD and TIMELESS oscillations in
LNVs of flies maintained in constant darkness (DD),
but does not interfere with clock protein oscillation in
diurnal 12h:12h light/dark (LD) conditions (Nitabach
et al., 2002). This indicates that neuronal electrical
activity in pacemaker neurons is essential to the nor-
mal function of the free-running cellular clock. LNV-
electrically silenced flies exhibit similar behavioral
phenotypes in DD and LD as pdf 01 mutant or LNV-
ablated flies (Renn et al., 1999). In the current study, we
have examined the effects of graded electrical silencing
of LNV pacemaker neurons on (1) membrane proper-
ties of PDF neurons, (2) detailed patterns of locomotor
activity in DD and LD, and (3) cellular oscillation in
non-LNV subsets of clock neurons. We find a graded
effect of varying degrees of electrical silencing on circa-
dian behavior, with the most extreme silencing pheno-
copying pdf 01 in both LD and DD conditions, although
an intermediate level of silencing phenocopies pdf 01 in
DD but that behaves as wild-type in LD. Although LNV
electrical silencing abolishes free-running cellular
rhythms in the LNVs, cellular rhythms continue to free-
run in non-LNV clock neurons, albeit with advanced
phase. Based on these results, we propose that the mol-
ecular oscillation in non-LNV clock neurons can persist
and drive weak rhythms with short period in the
absence of electrical activity or molecular rhythms in
the LNVs themselves.
MATERIALS AND METHODS
Fly Strains
pdf-GAL4, UAS-Kir2.1, UAS-dORKΔ-NC, and UAS-
dORKΔ-C fly lines are as described previously (Brand
 Wu et al. / AUTONOMOUS NON-LNV OSCILLATION
and Perrimon, 1993; Renn et al., 1999; Baines et al.,
2001; Nitabach et al., 2002). dORKΔ-C was con-
structed by truncating the C-terminal cytoplasmic
domain of native Drosophila open rectifier potassium
channel (dORK), which results in a constitutively
open channel (Nitabach et al., 2002). dORKΔ-NC is a
nonconducting version of dORKΔ with pore region
mutation. The fly strain If/Cyo;UAS-DsRedII/TM is
from Miesenböck laboratory (Yale University). We gen-
erated the fly strain pdf-gal4/Cyo;UAS-DsRedII/TM.
DsRedII is a variant of the original DsRed, a red fluo-
rescent protein derived from the Discosoma corals.
Pdf-GAL4 virgins were crossed to UAS males to gener-
ate male progeny for behavioral analysis and
immunostaining. Pdf-GAL4;UAS-DsRedII virgins were
crossed with UAS males to generate progeny for elec-
trophysiological recordings from dORKΔ-C- or Kir2.1-
expressing flies.
Brain Dissection and Electrophysiology
Flies 3-7 days posteclosion are anesthetized with
CO2 and then immersed in 70% ethanol for ~30 s
before being transferred to a dissection chamber
filled with external recording solution. External solu-
tion consists of (in mM): 101 NaCl, 3 KCl, 1 CaCl2, 4
MgCl2, 1.25 NaH2PO4, 5 glucose, 20.7 NaHCO3, pH
7.2 with osmolarity of 250 mmol/kg. After head cuti-
cle, eyes, proboscis, and trachea are removed with
fine forceps, the brain is dissected free and placed on
the floor of the recording chamber, frontal side up. A
mammalian brain slice “harp” holder with a 0.5-mm
interval between each nylon fiber is placed on top of
the fly brain to secure it during recording. Fly brains
are continuously perfused with external solution
bubbled with 95% O2/5% CO2 at room temperature.
LNVs are visualized using an Olympus fixed-stage
upright microscope (BX50WI) by their DsRed fluo-
rescence; lLNVs are distinguished by their sizes and
anatomical locations. The immediate area surround-
ing the lLNVs is enzymatically digested with focal
application of protease XIV (2 mg/mL; Sigma).
Whole-cell recordings are performed using borosili-
cate standard wall capillary glass pipettes (Sutter
Instrument Company, Novato, CA), and data are
acquired with Axopatch 200B amplifier, Digidata
1200 A/D hardware, and pClamp 8.0 software (Axon
Instruments, Union City, CA). Recording pipettes are
filled with internal solution consisting of (in mM):
102 potassium gluconate, 17 NaCl, 0.085 CaCl2, 4 Mg-
ATP, 0.5 Na-GTP, 0.94 EGTA, and 8.5 HEPES, pH 7.2,
and osmolarity of 235 mmol/kg. Gigaohm seals are
119
achieved before recording in cell-attached configura-
tion in voltage-clamp mode, followed by break-in to
whole-cell configuration while in voltage-clamp
mode. To confirm maintenance of a good seal and
absence of damage to the cell, a 60-mV hyperpolariz-
ing pulse is imposed on each cell while in whole-cell
voltage-clamp from a holding potential of –60 mV.
Only if the resulting inward leak current is less than
–50 pA is that cell used for subsequent current-clamp
measurements of resting potential and action poten-
tial firing. Resting membrane potentials are deter-
mined after stabilization of the membrane potential
following the transition from voltage-clamp to
current-clamp mode and for cells with oscillating
membrane potential is defined at the trough of the
oscillation. Action potential firing rates are computed
over the 5-min period following the transition from
voltage-clamp to current-clamp mode. The ampli-
tude of sinusoidal oscillation about resting potential
is defined as the difference between peak and trough.
The significance tests are performed using ANOVA
with Tukey-Kramer multiple comparisons on the dif-
ference of resting membrane potentials among differ-
ent genotypes.
Circadian Behavioral Analysis
Locomotor activity of individual adult male flies (1-
4 days post eclosion) was measured at 25 °C using the
TriKinetics infrared beam-crossing system. The flies
were first monitored for 5~6 days in 12 h:12 h LD con-
ditions, and free-running locomotor activity was then
monitored in constant darkness (DD) for 2 further
weeks. Total infrared beam crosses in 10-min bins
were recorded and circadian rhythms were analyzed
in 20-min blocks by Lomb-Scargle periodogram using
Actimetrics (Wilmette, IL) ClockLab software. Signifi-
cant circadian rhythmicity was defined as presence of
a peak in periodogram power that exceeds the p = 0.05
significance line, and the percentages of arrhythmic
flies were compared using χ2 analyses. The strength of
rhythmicity was defined as height of the Lomb-Scargle
periodogram peak in arbitrary units, and compared
among different genotypes by ANOVA with Tukey-
Kramer multiple comparisons. Total activities were
calculated for 5 days of LD or 12 days of DD, and
compared among different genotypes by ANOVA
with Tukey-Kramer multiple comparisons. Average
numbers of activity events per half-hour bin per fly
were also calculated and histograms for relative activi-
ties were generated for LD and DD behavior of the
flies. In addition, the times of morning and evening
120 JOURNAL OF BIOLOGICAL RHYTHMS / April 2008
activity peaks in LD were computed, and these peak
phases were analyzed statistically by ANOVA with
Tukey-Kramer multiple comparisons.
Immunocytochemistry
Adult fly brains were dissected and processed
for anti-PDP1 (1:10,000) immunocytochemistry as
described previously (Nitabach et al., 2006). PDP1 were
visualized using a Texas Red-conjugated secondary
antibody (1:300). Because the DN2 subgroup of clock
neurons is frequently indistinguishable anatomically
from the nearby DN1 clock neurons, DN2 staining was
scored for each brain hemisphere only as the presence
or absence of a distinct pair of neurons ventral to the
DN1s, and analyzed by comparing the percentage of
brain hemispheres exhibiting such a distinct pair. Anti-
PDP1 immunofluorescent images were collected using
a charge-coupled device (CCD) camera mounted on a
Zeiss (Oberkochen, Germany) Axioskop or Axiophot
epifluorescent photomicroscope. In brief, the average
background pixel value for each image was subtracted
from the threshold-selected pixels of that image to
yield the final threshold-selected background-
subtracted images, which were pseudocolored (with
warmer colors representing greater pixel values) and
used for quantitative analysis. The integrated pixel val-
ues of the threshold-selected background-subtracted
images were normalized within each cell group and
genotype in light/dark cycle or constant darkness to
the time point with the greatest average integrated
pixel value. The statistical analysis of anti-PDP1 stain-
ing was performed using ANOVA and Tukey-Kramer
multiple comparisons.
RESULTS
Δ-C on LNV
Graded Effect of Kir2.1 and dORKΔ
Membrane Electrical Properties
To assess the effects of K+ channel subunit expres-
sion on LNV membrane properties, we performed
whole-cell electrophysiological recordings in whole
brain explants of flies expressing either Kir2.1
(inward rectifier K+ channel) or conducting dORKΔ-
C (Drosophila open rectifier K+ channel). Both Kir2.1
inward rectifier and dORKΔ-C open rectifier K+ chan-
nels have no voltage dependence or time dependence
of channel gating, have a substantial open probabil-
ity at rest, and electrically silence neurons expressing
them by decreasing the input resistance at rest and
shifting the resting potential in the hyperpolarizing
direction toward the K+ equilibrium potential.
Inward rectifier and open rectifier channels differ
biophysically in that inward rectifiers have a higher
conductance for inward flux of K+ ions than for out-
ward, whereas open rectifiers have equal inward and
outward conductance. This biophysical difference is
not important in vivo, where the neuronal membrane
potential almost never becomes more hyperpolar-
ized than the K+ reversal potential, and thus K+ ions
almost never flow inward. Nonconducting point-
mutant dORKΔ-NC has no effect on input resistance
or resting potential and is used as a negative control.
To simplify the nomenclature, pdf>dORKΔ-C is used
to represent the fly pdf-gal4;UAS-dORKΔ-C, and anal-
ogous nomenclature is applied to other genotypes.
The role of large LNV (lLNV) clock neurons in free-
running circadian behavior is not well understood.
However, our membrane property experiment
focused on lLNVs because it is much easier to obtain
reliable recordings from lLNV than small LNVs
(sLNVs) and it is reasonable to provisionally assume
that electrical silencing has similar effects on pdf-
gal4-driven K+ channel expressing lLNV or sLNVs.
Flies expressing DsRed fluorescent protein and K+
channel subunits in the LNVs were maintained in LD
conditions. Whole-cell current-clamp recordings
were made on lLNV clock neurons, and the resting
membrane potential and spontaneous action poten-
tial firing were monitored from ZT2 to ZT7 of the first
half of the day. In wild-type flies, the resting mem-
brane potential of lLNVs (n = 33) averages –49.2 mV
and 25 of 33 lLNVs (75.8%) fire spontaneous action
potentials. In contrast, lLNVs expressing Kir2.1 (II) or
dORKΔ-C are totally silent with no spontaneous
action potential firing (data not shown). Consistent
with the lack of spontaneous action potential firing,
K+ channel-expressing lLNVs also exhibit signifi-
cantly hyperpolarized resting membrane potential in
comparison to control lLNVs (p < 0.001; Fig. 1). The
average resting potential of lLNVs expressing Kir2.1
or dORKΔ-C from the independent C1 or C2 UAS
transgene insertions are –77.6 mV (n = 9), –66.8 mV
(n = 9), and –64.1 mV (n = 12), respectively. Kir2.1-
expressing lLNVs exhibit significantly more hyperpo-
larized resting membrane potential in comparison to
dORKΔ-C (p < 0.05). Although the average resting
potential of flies expressing dORKΔ-C from the C1 or
C2 insertions are not significantly different, the trend
is for greater hyperpolarization with C1. These
results are consistent with the observations of Park
and Griffith on dORKΔ-C-expressing lLNVs (Park
 Wu et al. / AUTONOMOUS NON-LNV OSCILLATION 121

In addition, the remaining rhythmic pdf>dORKΔ-C1

−20 flies or pdf>Kir2.1 flies display weak rhythms with
increasing degree of silencing amplitude significantly smaller than that of the C1
Resting Membrane Potential (mV)

insertion and each of the dORKΔ-NC-expressing

−40 genotypes (Fig. 2C; p < 0.0001). The differences in
proportion of rhythmicity between flies expressing
Kir2.1 or dORKΔ-C from the C1 insertion and each of
−60 the dORKΔ-NC-expressing genotypes are highly sta-
tistically significant (χ2 > 44; p < 0.0001). Pdf> dORKΔ-
C2 flies exhibit no significant deficits in free-running
−80
locomotor rhythms, presumably due to less severe
electrical silencing in comparison to pdf> dORKΔ-C1
and pdf> Kir2.1 flies (Fig. 1). The reason expression
−100
from the C2 insertion has no effect on behavioral
2

1
-C

C
l

.1
ro

rhythymicity is uncertain, given that it appears that

-

r2
KΔ

KΔ
nt

Ki
co

the degree of electrical silencing for C1 and C2 are

dO

dO

similar. One possibility is that the difference in silenc-

Figure 1. Large LNV clock neurons are electrically silenced by
dORKΔ Δ or Kir2.1 K+ channel expression. Control flies or flies
expressing dORKΔ Δ or Kir2.1 were entrained in LD conditions.
Resting membrane potentials of lLNVs were recorded from
whole-brain explants under the whole-cell current-clamp con-
figuration during ZT2 to ZT7. Each data point represents the
resting membrane potential (in mV) of single lLNV. The average
resting membrane potentials of lLNVs are –49.2 ± 6.9 mV (n = 33),
–64.1 ± 6.2 mV (n = 12), –66.8 ± 3.9 mV (n = 9), and –77.6 ± 12.2 mV
(n = 9; mean ± SD) in control, pdf>dORKΔ-C2, pdf>dORKΔ-C1,
and pdf>Kir2.1 flies, respectively. Large LNV clock neurons
expressing Kir2.1 or dORKΔ Δ-C are significantly hyperpolarized
compared to control (p < 0.001). Kir2.1 expression in LNVs
induced more hyperpolarized resting membrane potential com-
pared to dORKΔ Δ-C (p < 0.05).
and Griffith, 2006), and demonstrate graded degree
of electrical silencing of LNVs by Kir2.1 and dORKΔ-
C expressed from 2 independent chromosomal inser-
tion lines.
Dose-Dependent Disruption of Circadian
Rhythms of Locomotor Activity by Electrical
Silencing of LNVs
Flies expressing Kir2.1, conducting dORKΔ-C, or
nonconducting dORKΔ-NC specifically in the LNV
pacemaker neurons were entrained for 5-6 days in
LD and then released into DD. More than 87% of con-
trol flies expressing dORKΔ-NC in the LNVs from
either of 2 independent UAS-dORKΔ-NC transgenes
exhibit free-running circadian rhythms of locomotor
activity with period averaging 24.6 h as assayed by
Lomb-Scargle periodogram analysis (Fig. 2). In con-
trast, more than 60% of pdf>dORKΔ-C1 flies (60%) or
pdf>Kir2.1 flies from 2 independent Kir2.1 insertions
(65% and 67%) are behaviorally arrhythmic (Fig. 2).
ing between these 2 lines is greater at some location
on the neuronal arbor electrically distant from the
recording electrode at the soma. In addition, weakly
rhythmic pdf>Kir2.1 flies show statistically significant
shorter period than pdf> dORKΔ-NC flies (Fig. 2; p <
0.001). The averaged actograms for different geno-
type of flies also indicate a dose-dependent effect of
electrical silencing on free-running period and
pdf>Kir2.1 flies from 2 independent insertions show
weak rhythmicity with shorter period relative to
dORKΔ-NC- or dORKΔ-C-expressing flies (Fig. 2B).
The above periodogram analysis was based on the 2
weeks of behavior immediately after release into DD
from the entraining LD environment. We performed a
more refined analysis of the average relative activity
for the first and second days after release into DD (DD-
D1,2) and the third and fourth days (DD-D3,4) of flies
expressing dORKΔ-NC, dORKΔ-C, or Kir2.1, including
arrhythmic and weakly rhythmic flies. Kir2.1 expres-
sion in LNVs induces a severe phenotype with progres-
sive loss of rhythms of locomotor activity from
DD-D1,2 to DD-D3,4 compared to flies expressing
dORKΔ-NC or dORKΔ-C (Fig. 3, bottom 2 rows of pan-
els). Additionally, Kir2.1-expressing flies have a uni-
modal distribution of activity on DD-D1,2 with a peak
at the subjective light-to-dark transition. In contrast,
pdf>dORKΔ-NC and pdf>dORKΔ-C2 flies have a
bimodal activity distribution on DD-D1,2 and DD-D3,4
with 1 peak near subjective dawn and the other near
subjective dusk. The gradual shift from bimodal to uni-
modal distribution of free-running locomotion
observed in pdf>dORKΔ-C1 flies from DD-D1,2 to DD-
D3,4 is consistent with less severe electrical silencing
by dORKΔ-C expressed from the UAS-dORKΔ-C1
insertion than by Kir2.1 (Figs. 1, 2).
122 JOURNAL OF BIOLOGICAL RHYTHMS / April 2008

We also analyzed the loco-

motor behavior in LD of
flies expressing dORKΔ-NC,
dORKΔ-C, or Kir2.1 in the LNVs.
Flies expressing Kir2.1 exhibit a
~1.5-h phase advance of
evening anticipation and a
damping of morning anticipa-
tion in comparison to other
genotypes. Although dORKΔ-
NC and dORKΔ-C flies begin
to increase their locomotor
activity about 2 h before
lights-off, Kir2.1 flies begin to
increase their activity about 3.5
h before lights-off (Fig. 3, top
panels). This alteration in antic-
ipatory behavior is similar to
that observed in pdf 01-null
mutant flies and flies lacking
PDF-expressing LNVs (Renn
et al., 1999). No phase shift of
evening anticipation occurs in
dORKΔ-C flies, consistent with
the less severe electrical silenc-
ing and free-running behav-
ioral phenotypes induced by
dORKΔ-C than by Kir2.1.
Control pdf>dORKΔ-NC flies
exhibit a small increase of rela-
tive locomotor activity in the
Figure 2. Behavioral phenotypes induced by electrical silencing. (A) Double-plotted locomo- middle of the night and before
tor actograms are shown, spanning 12 days in DD of representative male progeny of the indi- the morning anticipation. This
cated genotypes after release from diurnal 12h:12h LD entraining conditions. The gray and increase of relative locomotor
black bars above each actogram indicate subjective day and subjective night, respectively.
Pdf>dORKΔ-NC1 and Pdf>dORKΔ-NC4 control flies expressing nonconducting dORKΔ Δ K+
activity is more pronounced in
channels in the LNVs exhibit statistically significant free-running rhythms of locomotor activ- pdf>Kir2.1 flies, and is consis-
ity. In contrast, more than 60% of pdf>Kir2.1 or pdf>dORKΔ-C1 flies expressing Kir2.1 or con- tent with the phenotype of
ducting dORKΔ Δ K+ channels in the LNVs are behaviorally arrhythmic. The remaining pdf>Kir2.1
pdf 01-null flies (Renn et al.,
flies exhibit weak rhythmicity with period of 22.8 h, significantly shorter than 24.6 h in
pdf>dORKΔ-NC flies. Pdf>dORKΔ-C2 flies are statistically indistinguishable from nonconduct- 1999).
ing controls. (B) The averaged actograms of all flies (including rhythmic and arrhythmic flies) We calculated total aver-
of the indicated genotypes are shown. Pdf>Kir2.1 flies from 2 different insertions exhibit weak age activity of dORKΔ-NC-,
rhythmicity with period of 23.0 h or 22.8 h, shorter than 24.7 h in pdf>dORKΔ-NC1, 24.6 h in
pdf>dORKΔ-NC4 (amplitude 116), 24.5 h in pdf>dORKΔ-C2 (amplitude 141), and 24.6 h in dORKΔ-C-, or Kir2.1-expressing
pdf>dORKΔ-C1 flies (amplitude 39), indicating that electrical silencing of PDF neurons causes flies maintained in LD for 5
weak free-running circadian rhythms with short period. (C) A summary of the percentages of days and DD for 2 weeks (Fig.
flies exhibiting arrhythmic (gray bar) and rhythmic (white bar) locomotor activity assayed over
the first 12 days in DD. n, number of flies tested. The difference in proportion of behavioral
2C). Flies of all genotypes are
phenotypes between each of the electrically silenced genotypes except for pdf>dORKΔ-C2 and substantially more active in LD
the nonconducting controls are each statistically significant; for each of the marked χ2 values, p than in DD. Flies expressing
< 0.0001. LD or DD activity designates the averaged total activity among 5 days in LD or 12 days
Kir2.1 from either of 2 indepen-
in DD conditions (counts per minute, mean ± SD). Period designates the period value (h, mean
± SD) and amplitude the strength of rhythmicity (mean ± SD) for rhythmic flies from different dent insertions exhibit signifi-
genotypes; for marked period or amplitude values with *, p < 0.001. cantly higher activity levels
 Wu et al. / AUTONOMOUS NON-LNV OSCILLATION 123

Anti-PDP1 immunofluo-
pdf>dORKΔ-NC1 pdf>dORKΔ-NC4 pdf>dORKΔ-C2 pdf>dORKΔ-C1 pdf>Kir2.1 (II) rescence was detected in the
pdf>Kir2.1 (III)
0.10
0.08
nuclei of clock neurons and
LD
0.06
0.04
quantified by integrating back-
0.02 ground-subtracted pixel inten-
0.00
0.05 sities in anatomical subgroups
0.04
DD-D1,2 0.03
of clock neurons from each
0.02
0.01
brain hemisphere. Control
0.00
pdf>dORKΔ-NC flies express-
0.05
0.04 ing nonconducting dORKΔ-
DD-D3,4 0.03
0.02 NC in the LNVs exhibit a
0.01
0.00 similar temporal pattern of
PDP1 accumulation in the
Figure 3. Behavioral phenotypes induced by electrical silencing. Average activity histograms sLN , LN , and DN1 cell
V D
indicate relative levels of locomotor activity versus time for flies of the indicated genotypes.
groups, with peak levels late at
Each bar represents 30 min cumulative activity, with white and black bars indicating the day
and night phases in LD, respectively. In the DD plots, gray and black bars designate subjective night or subjective night (p <
day and night, respectively. LD behavior (5 days) and free-running behavior during DD-D1,2 or 0.001) and trough levels during
DD-D3,4 are shown. Flies expressing Kir2.1 exhibit damped lights-on anticipation and phase day or subjective day (Figs.
advance of lights-off anticipation in LD conditions. These flies also display unimodal rhyth-
micity on DD-D1,2 and behavioral arrhythmicity on DD-D3,4 of constant darkness. In contrast, 4-6). In contrast, anti-PDP1
both pdf>dORKΔ-C1 and pdf>dORKΔ-C2 flies display similar LD behavior to pdf>dORKΔ-NC immunofluorescence in the
flies. However, pdf>dORKΔ-C1 flies exhibit a gradual shift from biomodal to unimodal rhyth- sLN s of pdf>Kir2.1 flies is
V
micity in DD.
barely detectable in 6 of 16
brain hemispheres on DD-D2
than dORKΔ-C- or dORKΔ-NC-expressing flies in LD and only 1 of 18 hemispheres on DD-D4. These results
(p < 0.05). In DD, Kir2.1- and dORKΔ-C-expressing flies establish that Kir2.1 expression abolishes PDP1 oscilla-
do not differ significantly from control dORKΔ-NC- tion in the sLNVs in DD (p < 0.001), consistent with our
expressing flies, except that the UAS-dORKΔ-C2 line previous observations of the effects of Kir2.1 expres-
exhibits a significantly higher level of activity (p < sion on TIM and PER oscillations (Nitabach et al., 2002;
0.001). In summary, although the most extreme degree Nitabach et al., 2005). However, when maintained in
of LNV electrical silencing by Kir2.1 increases overall LD, Kir2.1 expression does not interfere with PDP1
activity in LD, there is no consistent effect of electrical oscillation, again consistent with our previous observa-
silencing on overall locomotor activity in DD. tions of the effects of Kir2.1 expression on TIM and PER
oscillations. In fact, pdf>Kir2.1 flies exhibit a signifi-
cantly higher level of PDP1 accumulation at ZT14,
Electrical Silencing of LNVs Stops Free-Running
PDP1 Oscillation in sLNVs and Speeds PDP1 ZT18, and ZT22 than control flies (p < 0.001), and Kir2.1
Oscillations in LNDs, DN1s, and DN2s expression causes phase advance of the onset of PDP1
accumulation in sLNVs in LD (Fig. 4). These results
To assess the effect of electrical silencing on cellular indicate that LNV electrical silencing stops free-
rhythmicity, we performed anti-PDP1 immunos- running, but not entrained, PDP1 clock protein oscilla-
taining in control pdf> dORKΔ-NC4 and experimental tion in sLNV neurons, consistent with our previous
pdf>Kir2.1 (II) adult fly brains on DD-D2 and DD-D4. observations of the effects of electrical silencing on PER
Flies were entrained in LD and then released into DD and TIM oscillation (Nitabach et al., 2002; Nitabach
for 2 or 4 days or kept in LD before the fly brains were et al., 2005).
fixed at different times of day, followed by anti-PDP1 In contrast to the effects of LNV Kir2.1 expression
immunocytochemistry. The large LNV neurons were on LNV cellular oscillation, free-running PDP1 oscil-
excluded from analysis because we never observe anti- lations in LND, DN1, and DN2 clock neurons are nei-
PDP1 staining in the lLNVs in DD, consistent with prior ther abolished nor damped by LNV Kir2.1 expression.
reports of absence of functional cellular oscillation in Kir2.1 expression in LNVs induces a phase advance of
lLNVs under free-running conditions (Yang et al., 1998; PDP1 oscillation in LND, DN1, and DN2 clock neu-
Veleri et al., 2003; Grima et al., 2004; Nitabach et al., rons but without affecting amplitude. In pdf>Kir2.1
2006). flies, the LNDs exhibit a peak of PDP1 accumulation
124 JOURNAL OF BIOLOGICAL RHYTHMS / April 2008

anatomically close DN2s in

A sLNV Anti-PDP1 Immunocytochemistry B Staining Intensity some brains to the DN1 neuron
pdf>dORKΔ-NC group. Taken together, these
ZT2 ZT6 ZT10 ZT14 ZT18 ZT22 pdf>Kir2.1
LD 1.00
results suggest that electrical
pdf>dORKΔ-NC silencing of LNVs via Kir2.1
0.75
expression leads to acceleration
0.50
of free-running cellular oscilla-
0.25
pdf>Kir2.1 tion of the LND, DN1, and DN2
0.00 clock neurons.
CT2 CT6 CT10 CT14 CT18 CT22 2 6 10 14 18 22
DD-D2 1.00
pdf>dORKΔ-NC 0.75
DISCUSSION
0.50
0.25
pdf>Kir2.1 We have previously shown
0.00 that LNV electrical silencing
2 6 10 14 18 22
CT2 CT6 CT10 CT14 CT18 CT22 induces severe deficits in free-
DD-D4 1.00
running behavioral rhythms—
pdf>dORKΔ-NC 0.75
with substantial arrhythmic-
0.50 ity and weak short-period
0.25 rhythms—and stops sLNV
pdf>Kir2.1
0.00 molecular oscillations (Nitabach
2 6 10 14 18 22
et al., 2002, 2005). In this study
we establish that free-running
Figure 4. Kir2.1 expression in LNV pacemaker neurons stops free-running PDP1 clock protein molecular oscillation in non-
oscillation in sLNVs. Pdf>dORKΔ-NC and pdf>Kir2.1 flies were kept in LD conditions or LN clock neurons continues in
V
released into DD for 2 or 4 days after entrainment, and fly brains were then processed for anti-
flies with electrically silenced
PDP1 immunofluorescence at the indicated time points. Bar graphs show normalized integrated
anti-PDP1-staining intensities (mean ± SEM). Representative pseudocolored photographs of LNVs, albeit with advanced
clock neurons of the indicated genotypes at the indicated time points are shown. Control phase (Figs. 5-7). On the basis
pdf>dORKΔ-NC flies exhibit a peak level of PDP1 accumulation in the sLNVs at ZT22 or CT22 of these results we propose that
(p < 0.001), late in night or subjective night. In contrast, PDP1 oscillation is nearly abolished in
pdf>Kir2.1 flies in DD (p < 0.001). In LD conditions, pdf>Kir2.1 flies exhibit earlier onset of PDP1 (1) the weak short-period free-
oscillation and statistically significant higher PDP1 accumulation relative to control running behavioral rhythms of
pdf>dORKΔ-NC flies (p < 0.001). n > 10 brain hemispheres for each experimental group. flies with electrically silenced
LNVs are determined by cellu-
on DD-D2 at CT22, the same as in control lar oscillation in non-LNV clock neurons, and (2) the
pdf>dORKΔ-NC flies, but by DD-D4, this peak has persistence of molecular oscillation in non-LNV clock
shifted to CT18, 4 h earlier than in control flies (p < neurons is independent of both molecular oscillation
0.001; Fig. 5). Kir2.1 expression in LNVs also phase and electrical activity of LNVs, which abolishes action
advances PDP1 accumulation in DN1s, with peak potential firing and hyperpolarizes the plasma
accumulation of PDP1 at CT14 on DD-D2, 4 h earlier membrane.
than in control flies, and at CT10 on DD-D4 (p < 0.001; When LNVs are electrically silenced by expression
Fig. 6). DN2s of control flies exhibit peak PDP1 accu- of Kir2.1 or dORKΔ-C K+ channels (Fig. 1), most flies
mulation at CT14 on DD-D2 and at CT10 on DD-D4 exhibit arrhythmicity and the remainder exhibit
(Fig. 7). Although the DN2s of pdf>Kir2.1 flies exhibit weak rhythmicity with short periods (Fig. 2). The
a peak of PDP1 accumulation in phase with control severity of behavioral phenotypes in LD and DD
flies on DD-D2, by DD-D4, PDP1 accumulation correlates with the severity of electrical silencing
peaks at CT6, 4 h earlier than in controls (Fig. 7). DN1 induced by expression of Kir2.1 or dORKΔ-C (Figs. 1-
neurons exhibit small phase advance in control flies 3). The most severe behavioral phenotype induced
on DD-D2. Because the DN2s are close to DN1s and by Kir2.1 phenocopies that of pdf 01 both in LD and
drift out of phase with DN1s progressively in DD DD, whereas that of dORKΔ-C (expressed from the
(Veleri et al., 2003), the small phase advance apparent C1 transgene insertion) phenocopies pdf 01 only in DD,
in control flies likely reflects the attribution of but appears wild-type in LD (Fig. 3). Morning and
 Wu et al. / AUTONOMOUS NON-LNV OSCILLATION 125

Stoleru et al., 2004; Stoleru

A LND Anti-PDP1 Immunocytochemistry BStaining Intensity et al., 2005).
pdf>dORKΔ-NC Consistent with previous
ZT2 ZT6 ZT10 ZT14 ZT18 ZT22 pdf>Kir2.1
LD analysis of the effect of LNV
1.00
pdf>dORKΔ-NC
0.75
electrical silencing on TIM and
PER oscillation in LNVs, electri-
0.50
cal silencing of LNVs stops free-
0.25
pdf>Kir2.1 running, but not entrained,
0.00
2 6 10 14 18 22 PDP1 oscillation in the sLNVs
CT2 CT6 CT10 CT14 CT18 CT22
DD-D2 (Fig. 4). In contrast to these cell-
1.00
pdf>dORKΔ-NC
autonomous effects, however,
0.75
LNV electrical silencing does
0.50 not diminish the molecular
pdf>Kir2.1 0.25 oscillations in non-LNV clock
0.00 neurons, and rather induces
2 6 10 14 18 22
CT2 CT6 CT10 CT14 CT18 CT22 phase advance of free-running
DD-D4 1.00 PDP1 oscillation in the LNDs,
pdf>dORKΔ-NC 0.75 DN1s, and DN2s (Figs. 5-7).
0.50 This is consistent with the effects
0.25 of pdf 01-null mutation on non-
pdf>Kir2.1
0.00 LNV cellular oscillation (Lin
2 6 10 14 18 22 et al., 2004). Phase advances of
PDP1 oscillation in non-LNV
Figure 5. Kir2.1 expression in LNV pacemaker neurons causes phase advanced PDP1 oscilla-
clock neurons in flies with elec-
tion in LNDs in DD. Control pdf>dORKΔ-NC flies exhibit a similar temporal pattern of PDP1 trically silenced LNVs suggests
accumulation in the LNDs as in the sLNVs, with a peak at ZT22 or CT22 (p < 0.001). In contrast, that in the absence of LNV elec-
pdf>Kir2.1 flies exhibit a peak of PDP1 accumulation at CT18 on DD-D4, 4 h earlier than on DD-
trical activity, non-LNV clock
D2 (p < 0.001). Kir2.1 expression in LNVs does not affect PDP1 oscillation in LND neurons in LD.
n > 12 brain hemispheres for each experimental group and error bars indicate SEM. neurons free-run with their own
intrinsic short-period rhythms.
This interpretation is consistent
evening anticipatory locomotor activity is thought to with the role of LNVs in sending a daily phase-resetting
be driven by “morning” and “evening” (“M” and signal to non-LNV clock neurons (Stoleru et al., 2005).
“E”) cells, which have been localized to the sLNVs Since the non-LNV clock neurons, LNDs and DNs,
and LNDs/DN1s, respectively (Grima et al., 2004; exhibit phase-advanced cellular oscillations in
Stoleru et al., 2004). In constant darkness, pdf>Kir2.1 pdf>Kir2.1 flies, we propose that the weak short-period
flies exhibit unimodal rather than bimodal free-running behavioral rhythms of the flies with elec-
distribution of free-running behavior with a peak trically silenced LNVs are induced by accelerated mol-
near the subjective light-to-dark transition on DD- ecular oscillation in non-LNV clock neurons. In the
D1,2 (Fig. 3), indicative of persistent function of E, absence of resetting signal from LNVs to LND/DN neu-
but not M, cells. The evening peak of free-running rons, the molecular oscillation in those non-LNV clock
behavior in Kir2.1-expressing flies disappears in neurons are not sufficient to ensure normal circadian
most flies by DD-D3,4, despite sustained molecular rhythmicity in constant darkness, yet the molecular
oscillation in E cells (Figs. 5, 6; Veleri et al., 2003). In oscillations in the LND/DN neurons still sustain with
LD, flies with electrically silenced LNVs exhibit a no discernible change of amplitude for at least 4 days
decrease of morning peak activity and phase advance in constant darkness (Figs. 4-7). These results further
of evening peak anticipation (Fig. 3). These results emphasize the role of sLNV in maintaining free-
are consistent with previous observations of the running behavioral rhythms in constant darkness.
effect of pdf 01-null mutation on circadian behavior In relation to the underlying mechanisms for syn-
(Renn et al., 1999), and further support the key role of chronizing multiple independent oscillators in circa-
sLNV in determining free-running behavioral dian circuits, it is thought that neuropeptide PDF is
rhythms in DD and the morning anticipatory peak an intrinsic coupling signal within the circadian clock
(Renn et al., 1999; Grima et al., 2004; Lin et al., 2004; circuit that synchronizes multiple oscillators that
126 JOURNAL OF BIOLOGICAL RHYTHMS / April 2008

otherwise free-run indepen-

dently (Helfrich-Forster et al.,
A DN1 Anti-PDP1 Immunocytochemistry B Staining Intensity
pdf>dORKΔ-NC 2000; Park et al., 2000; Lin
ZT2 ZT6 ZT10 ZT14 ZT18 ZT22 pdf>Kir2.1 et al., 2004; Stoleru et al., 2005;
LD 1.00
pdf>dORKΔ-NC
Nitabach et al., 2006). Interest-
0.75 ingly, electrical silencing of
0.50 LNVs phenocopies the pdf 01-
pdf>Kir2.1 0.25 null mutant at both behav-
0.00 ioral and molecular levels
2 6 10 14 18 22
CT2 CT6 CT10 CT14 CT18 CT22 (Renn et al., 1999) except for
DD-D2
1.00
the complete abolition of
pdf>dORKΔ-NC 0.75
free-running cellular oscilla-
0.50 tion in the LNVs themselves
pdf>Kir2.1 0.25 (Nitabach et al., 2002). LNV-
0.00 electrically silenced or pdf 01-
2 6 10 14 18 22
CT2 CT6 CT10 CT14 CT18 CT22 null flies exhibit weak free-
DD-D4 1.00
running behavioral rhythms
pdf>dORKΔ-NC 0.75
with short period, and the mol-
0.50
ecular oscillation in non-LNV
0.25 neurons phase advances in con-
pdf>Kir2.1
0.00 stant darkness. The pdf 01-null
2 6 10 14 18 22
mutant has no PDF peptide,
yet molecular oscillation still
Figure 6. Kir2.1 expression in LNV pacemaker neurons causes phase-advanced PDP1 oscilla-
sustains in sLNVs, albeit with
tion in DN1 neurons in DD. Control pdf>dORKΔ-NC flies exhibit a similar temporal pattern of gradually dispersing phase in
PDP1 accumulation in the DN1s as in the sLNVs, with a peak late in night or subjective night constant darkness (Lin et al.,
(p < 0.001). Pdf>Kir2.1 flies exhibit a phase advance of peak PDP1 accumulation from CT14 on DD-
2004). In contrast, LNV electri-
D2 to CT10 on DD-D4 (p < 0.001). Kir2.1 expression does not affect PDP1 oscillation in DN1 neu-
rons in LD. n > 12 brain hemispheres for each experimental group and error bars indicate SEM. cal silencing abolishes the
molecular oscillation of LNVs
(Nitabach et al., 2002). That
A DN2 Anti-PDP1 Immunocytochemistry B % of Hemispheres LNV electrical silencing leads
pdf>dORKΔ-NC
CT2 CT6 CT10 CT14 CT18 CT22 pdf>Kir2.1 to the same behavioral and
DD-D2 100
non-LNV molecular pheno-
pdf>dORKΔ-NC 75
types as pdf 01 thus suggests
50
that persistence of LNV molec-
pdf>Kir2.1 25 ular oscillation in pdf 01 flies
0 has no functional effect, either
2 6 10 14 18 22
CT2 CT6 CT10 CT14 CT18 CT22
100 on behavioral rhythms or on
DD-D4
75
non-LNV molecular rhythms.
pdf>dORKΔ-NC
We thus conclude that func-
50
tionally relevant signals from
25
pdf>Kir2.1 LNVs to non-LNV clock neu-
0
2 6 10 14 18 22 rons and other downstream
targets rely both on PDF sig-
Figure 7. Kir2.1 expression in LNV pacemaker neurons causes phase-advanced free-running naling and LN V electrical
PDP1 clock protein oscillation in DN2 neurons. The bar graph shows the percentage of brain activity, and that LNVs do
hemispheres with a pair of anatomically distinguishable PDP1-positive DN2s. The DN2s in not ordinarily send function-
control pdf>dORKΔ-NC flies exhibit a peak of detectability at CT14 on DD-D2, as do pdf>Kir2.1
flies (p < 0.001; χ2). However, by DD-D4, pdf>Kir2.1 flies exhibit a different temporal profile of ally relevant signals via PDF-
PDP1 accumulation from control DN2 clock neurons, with peak accumulation 4 h earlier at CT6. independent mechanisms.
 Wu et al. / AUTONOMOUS NON-LNV OSCILLATION
ACKNOWLEDGMENTS
Work in the laboratory of M.N.N. is supported in
part by the National Institute of Neurological Disorders
and Stroke, National Institutes of Health (grants
R01NS055035 and R01NS056443), and by the Whitehall
Foundation. Y.W. is supported by an NINDS Ruth L.
Kirschstein NRSA Postdoctoral Fellowship. The
authors thank Justin Blau for providing anti-PDP1 anti-
serum and Michael Schwartz for the use of an epifluo-
rescence microscope.
REFERENCES
Baines RA, Uhler JP, Thompson A, Sweeney ST, and Bate M
(2001) Altered electrical properties in Drosophila neu-
rons developing without synaptic transmission. J
Neurosci 21:1523-1531.
Blanchardon E, Grima B, Klarsfeld A, Chelot E, Hardin PE,
Preat T, and Rouyer F (2001) Defining the role of
Drosophila lateral neurons in the control of circadian
rhythms in motor activity and eclosion by targeted
genetic ablation and PERIOD protein overexpression.
Eur J Neurosci 13:871-888.
Blau J and Young MW (1999) Cycling vrille expression is
required for a functional Drosophila clock. Cell 99:661-671.
Brand AH and Perrimon N (1993) Targeted gene expression
as a means of altering cell fates and generating domi-
nant phenotypes. Development 118:401-415.
Cyran SA, Buchsbaum AM, Reddy KL, Lin MC, Glossop
NR, Hardin PE, Young MW, Storti RV, and Blau J (2003)
vrille, Pdp1, and dClock form a second feedback loop in
the Drosophila circadian clock. Cell 112:329-341.
Gekakis N, Saez L, Delahaye-Brown AM, Myers MP, Sehgal
A, Young MW, and Weitz CJ (1995) Isolation of timeless
by PER protein interaction: Defective interaction
between timeless protein and long-period mutant
PERL. Science 270:811-815.
Grima B, Chelot E, Xia R, and Rouyer F (2004) Morning and
evening peaks of activity rely on different clock neurons
of the Drosophila brain. Nature 431:869-873.
Helfrich-Forster C (1998) Robust circadian rhythmicity of
Drosophila melanogaster requires the presence of lateral
neurons: A brain-behavioral study of disconnected
mutants. J Comp Physiol [A] 182:435-453.
Helfrich-Forster C, Tauber M, Park JH, Muhlig-Versen M,
Schneuwly S, and Hofbauer A (2000) Ectopic expression
of the neuropeptide pigment-dispersing factor alters
behavioral rhythms in Drosophila melanogaster. J
Neurosci 20:3339-3353.
Hyun S, Lee Y, Hong S-T, Bang S, Paik D, Kang J, Shin J, Lee
J, Jeon K, Hwang S, Bae E, and Kim J (2005) Drosophila
GPCR Han is a receptor for the circadian clock neu-
ropeptide PDF. Neuron 48:267-278.
Kaneko M, Park JH, Cheng Y, Hardin PE, and Hall JC
(2000) Disruption of synaptic transmission or clock-
gene-product oscillations in circadian pacemaker cells
127
of Drosophila cause abnormal behavioral rhythms. J
Neurobiol 43:207-233.
Lear BC, Merrill CE, Lin J-M, Schroeder A, Zhang L, and
Allada R (2005) A G protein-coupled receptor, Groom-
of-PDF, is required for PDF neuron action in circadian
behavior. Neuron 48:221-227.
Lin Y, Stormo GD, and Taghert PH (2004) The neuropeptide
pigment-dispersing factor coordinates pacemaker inter-
actions in the Drosophila circadian system. J Neurosci
24:7951-7957.
Mertens I, Vandingenen A, Johnson EC, Shafer OT, Li W,
Trigg JS, De Loof A, Schoofs L, and Taghert PH (2005) PDF
receptor signaling in Drosophila contributes to both circa-
dian and geotactic behaviors. Neuron 48:213-219.
Myers MP, Wager-Smith K, Wesley CS, Young MW, and
Sehgal A (1995) Positional cloning and sequence analy-
sis of the Drosophila clock gene, timeless. Science
270:805-808.
Nitabach MN, Blau J, and Holmes TC (2002) Electrical
silencing of Drosophila pacemaker neurons stops the
free-running circadian clock. Cell 109:485-495.
Nitabach MN, Sheeba V, Vera DA, Blau J, and Holmes TC
(2005) Membrane electrical excitability is necessary for
the free-running larval Drosophila circadian clock. J
Neurobiol 62:1-13.
Nitabach MN, Wu Y, Sheeba V, Lemon WC, Strumbos J,
Zelensky PK, White BH, and Holmes TC (2006)
Electrical hyperexcitation of lateral ventral pacemaker
neurons desynchronizes downstream circadian oscilla-
tors in the fly circadian circuit and induces multiple
behavioral periods. J Neurosci 26:479-489.
Park D and Griffith LC (2006) Electrophysiological and
anatomical characterization of PDF-positive clock neu-
rons in the intact adult Drosophila brain. J Neurophysiol
95:3955-3960.
Park JH, Helfrich-Forster C, Lee G, Liu L, Rosbash M, and
Hall JC (2000) Differential regulation of circadian pace-
maker output by separate clock genes in Drosophila.
Proc Natl Acad Sci U S A 97:3608-3613.
Peng Y, Stoleru D, Levine JD, Hall JC, and Rosbash M
(2003) Drosophila free-running rhythms require inter-
cellular communication. PLoS Biol 1:E13.
Renn SC, Park JH, Rosbash M, Hall JC, and Taghert PH
(1999) A pdf neuropeptide gene mutation and ablation
of PDF neurons each cause severe abnormalities of
behavioral circadian rhythms in Drosophila. Cell 99:
791-802.
Reppert SM and Weaver DR (2000) Comparing clockworks:
Mouse versus fly. J Biol Rhythms 15:357-364.
Sehgal A, Ousley A, Yang Z, Chen Y, and Schotland P (1999)
What makes the circadian clock tick: Genes that keep
time? Recent Prog Horm Res 54:61-84.
Sehgal A, Price JL, Man B, and Young MW (1994) Loss
of circadian behavioral rhythms and per RNA oscilla-
tions in the Drosophila mutant timeless. Science 263:
1603-1606.
Sehgal A, Rothenfluh-Hilfiker A, Hunter-Ensor M, Chen Y,
Myers MP, and Young MW (1995) Rhythmic expression
of timeless: A basis for promoting circadian cycles in
period gene autoregulation. Science 270:808-810.
128 JOURNAL OF BIOLOGICAL RHYTHMS / April 2008
Stoleru D, Peng Y, Agosto J, and Rosbash M (2004) Coupled
oscillators control morning and evening locomotor
behaviour of Drosophila. Nature 431:862-868.
Stoleru D, Peng Y, Nawathean P, and Rosbash M (2005) A
resetting signal between Drosophila pacemakers synchro-
nizes morning and evening activity. Nature 438:238-242.
Veleri S, Brandes C, Helfrich-Forster C, Hall JC, and
Stanewsky R (2003) A self-sustaining, light-entrainable
circadian oscillator in the Drosophila brain. Curr Biol
13:1758-1767.
Williams JA and Sehgal A (2001) Molecular components of
the circadian system in Drosophila. Annu Rev Physiol
63:729-755.
Yang Z, Emerson M, Su HS, and Sehgal A (1998) Response
of the timeless protein to light correlates with behav-
ioral entrainment and suggests a nonvisual pathway for
circadian photoreception. Neuron 21:215-223.
Young MW and Kay SA (2001) Time zones: A compara-
tive genetics of circadian clocks. Nat Rev Genet 2:
702-715.