A Trophic Role For Serotonin in The Development of A Simple Feeding Circuit

Original Paper
Dev Neurosci 2010;32:217237

DOI: 10.1159/000304888
Received: October 14, 2009

Accepted after revision: March 17, 2010
Published online: August 11, 2010
A Trophic Role for Serotonin in the

Development of a Simple Feeding Circuit
WendiS.Neckameyer
Department of Pharmacological and Physiological Science, Saint Louis University School of Medicine,
St. Louis, Mo., USA
Key Words
Branching ! Drosophila ! RNAi transgenic lines ! Tryptophan
hydroxylase ! Varicosities
Abstract
Correct differentiation and positioning of individual synapses during development is fundamental to the normal function of neuronal circuits. While classical transmitters such as
serotonin (5-HT) play a critical trophic role in neurogenesis
in addition to their functions as transmitters in the mature
nervous system, this process is not well understood. We used
a simple model to assess both development and function of
a specific behavioral circuit in the larval stage of the fruit fly
(Drosophila melanogaster). We show that, as in all other species examined, the neurotransmitter actions of 5-HT depress
feeding, and decreased neuronal 5-HT levels increase appetite. However, using transgenic tools, we show that constitutive knockdown of neuronal 5-HT synthesis to reduce 5-HT
levels during central nervous system (CNS) development results in increased branching of the serotonergic fibers projecting to the gut, as well as increased size and number of
varicosities along the neurite length. As larvae, these animals
display decreased feeding rates relative to controls, and,
when given exogenous 5-hydroxytryptophan, feeding is
significantly enhanced. Late-stage wild-type embryos ex-
Fax +41 61 306 12 34

E-Mail karger@karger.ch
www.karger.com
2010 S. Karger AG, Basel

03785866/10/03230217$26.00/0
Accessible online at:
www.karger.com/dne
posed to 5-HT to augment 5-HT levels during CNS development display, as mature larvae, a significant decrease in gut
fiber branching and total varicosity number, as well as increased feeding and a hyposensitivity to the effects of 5-HT.
Exposure of embryos unable to synthesize neuronal serotonin to 5-HT during late embryogenesis results in rescue of
the feeding behavior and abnormalities in the 5-HT gut fiber
architecture. These results demonstrate an inverse relationship between developmental 5-HT levels and complexity of
the fiber architecture projecting to gut tissue, which results
in a perturbed feeding pattern. We conclude that 5-HT is
tightly regulated during CNS development to direct the normal architecture and mature function of this neural circuit.
Copyright 2010 S. Karger AG, Basel
Introduction
Neuroactive substances function in the establishment

of neural networks before adopting their roles as transmitters in the mature central nervous system (CNS) [for
review, see Herlenius and Lagercrantz, 2001]. Seminal
earlier studies by Lauder and Kater and their colleagues
demonstrated that altered levels of specific transmitters
during embryogenesis affected the synaptic development
of mature neurons [Lauder and Krebs, 1978; Lauder et al.,
Wendi S. Neckameyer
Department of Pharmacological and Physiological Science
Saint Louis University School of Medicine
1402 South Grand Boulevard, St. Louis, MO 63104 (USA)
Tel. +1 314 977 6346, Fax +1 314 977 6411, E-Mail neckamws@slu.edu
1982; Haydon et al., 1984, 1987; Goldberg and Kater,

1989]. However, behavioral consequences of the developmental impairments could not be addressed in these
studies. Since perturbations in neurotrophic signaling
pathways are believed to underlie the etiology of several
depressive disorders [reviewed in Hasler et al., 2004;
Sodhi and Sanders-Bush, 2004], as well as autism spectrum disorders [Pardo and Eberhart, 2007] and the dysfunctional brain development observed in Down syndrome [Whittle et al., 2007], understanding the behavioral consequences of impairments in the development
of neural circuitry is imperative to delineating etiology
for these pathologies.
The indolamine serotonin (5-HT) has emerged as one
such trophic factor critical for development of synaptic
connectivity [see Gaspar et al., 2003], also serving as a
neurotransmitter and a neuromodulator in all animal
phyla studied [reviewed in Buznikov et al., 2001]. In
mammals, 5-HT affects the developing cerebral and prefrontal cortex [Luo et al., 2003; Janusonis et al., 2004;
Beique et al., 2004], and administration of 5-MT, a methylated derivative of 5-HT, affects sprouting of serotonergic fibers [Shemer et al., 1991]. Consistent use of selective
serotonin reuptake inhibitors as therapeutic targets for
depression stimulates hippocampal neurogenesis and
dendritic branching [reviewed in Cowan, 2007]. Developmental abnormalities in 5-HT pathways have been implicated in autism [Anderson et al., 1990; Chugani, 2002]
and schizophrenia [Dean, 2003]. Mice lacking serotonin
receptor 5-HT1A function during embryonic and postnatal stages display enhanced anxiety as adults [Gross et
al., 2002]; activation of the 5HT1A receptor promotes
neurite extension and survival of neuronal cells in culture [Fricker et al., 2005], suggesting that the trophic actions of 5-HT occur, at least in part, via signaling through
this receptor. This trophic role for 5-HT appears conserved across evolution. 5-HT acts as a regulatory signal
in the development of the antennal lobes during metamorphosis in the moth [Oland et al., 1995]. Total neurite
length is enhanced by 5-HT application to moth neuronal
cell populations in culture; in vitro and in vivo, 5-HT increases cell excitability and input resistance in antennal
neurons after depolarization [Mercer et al., 1995, 1996,
1999]. In Drosophila mutants deficient in dopa decarboxylase, the second enzyme in the biosynthetic pathway for
both dopamine and 5-HT, peripheral 5-HT fibers innervating the midgut display altered branching [Budnik et
al., 1989], consistent with later studies which suggest
5-HT is an autoregulator of serotonergic varicosity density [Sykes and Condron, 2005]. Again, these studies,
218
while demonstrating a critical role for 5-HT in the development of neural circuitry, did not address the behavioral consequences of these changes.
The rate-limiting step in 5-HT synthesis is the hydroxylation of tryptophan to 5-hydroxytryptophan (5-HTP),
which occurs via the action of tryptophan hydroxylase
(Tph); Tph is thus fundamental to supporting serotonergic transmission. In Drosophila, as in mammals, there are
two discrete Tph enzymes: Drosophila tryptophan-phenylalanine hydroxylase (DTPH) is largely nonneuronal,
expressed in the periphery, and corresponds to mammalian Tph1; Drosophila tryptophan hydroxylase (DTRH)
is exclusively neuronal and corresponds to mammalian
Tph2 [Walther et al., 2003; Zhang et al., 2004; Coleman
and Neckameyer, 2005; Neckameyer et al., 2007]. Biochemical regulation of these enzymes is strikingly similar to that of mammals [Coleman and Neckameyer, 2004,
2005].
In the larval stage of Drosophila, there are 84 bilaterally symmetric interneurons [Valles and White, 1988];
the recurrens nerve, which displays a high degree of
5-HT immunoreactivity, is part of the stomatogastric
nervous system and innervates the pharyngeal muscles,
the proventriculus (the larval foregut), and the midgut
[Budnik et al., 1989]. These fibers originate in the brain
and enter the proventriculus as small groups of 24 fascicles or bundles dotted with variscosities along the
lengths. They display relatively little branching until
reaching the anterior midgut (fig. 1). The cephalopharyngeal plates also display intense immunoreactivity
against 5-HT, as does the frontal nerve, which directly
connects the cephalopharyngeal plates to the brain
(fig.1). The larval mouth hooks (forming the most anterior part of the cephalopharyngeal plates) shovel food
into the gut at a relatively constant rate, since the animals
eat continuously during this stage of development. 5-HT
projections from the brain innervate the mouthparts and
the digestive tract, strong evidence for a role for 5-HT in
the modulation of feeding. This is consistent with the
observed role of 5-HT in appetite modulation in numerous species, including mammals [Lucki, 1992] and insects [e.g. Dacks et al., 2003; Novak and Rowley, 1994;
Novak et al., 1995]. While glutamate- and FMRFamidecontaining fibers are also present within the proventriculus, they are distinct from those that are serotonergic
[Budnik et al., 1989]. However, a role for 5-HT in modulating appetitive behavior in Drosophila has not yet been
formally established.
We previously noted that Drosophila mutants lacking
DTRH function, which contain very little neuronal seroNeckameyer
a
b
c
d
Cephalopharyngeal
plates
Fig. 1. Serotonergic fibers projecting to the
larval gut. Gut tissue was dissected from a

third instar larva and immunostained
with an antibody raised against 5-HT.
a Proventriculus and midgut. a = Proventriculus (foregut); b = fascicles of 5-HTimmunoreactive fibers; c = branching of
these fibers as they reach the midgut; d =
midgut. b, d Feeding circuit. c Close-up of
frontal nerve (e) and cephalopharyngeal
plates (f). Scale bars = 100 !m.
Brain
e
Recurrens
nerve
Frontal
nerve
Proventriculus
tonin (although there is no neuronal 5-HT synthesis, a

limited amount is taken up from the periphery by the
serotonin transporter), display decreased feeding as larvae [Neckameyer et al., 2007]. However, in all species examined, acute reduction in 5-HT levels in adult animals
increases feeding behavior, and increased 5-HT levels
decrease feeding [e.g. Dacks et al., 2003]. Serotonergic
cell bodies and some projections can be visualized in the
CNS by 1618 h after fertilized eggs have been laid (approximately 6 h before hatching), before there is a functional nervous system. No peripheral fibers appear to be
present until 24 h before hatching, and serotonergic fibers projecting to the gut are not observed until !2 h
before hatching [Budnik et al., 1989]. We propose that
the difference in feeding behavior in DTRH-null Drosophila reflects the critical need for 5-HT during neuronal development, which can be tested by comparing constitutive with temporal knockdown of neuronal DTRH.
Our data show that reduction in DTRH levels after the
CNS has developed results in increased feeding, with no
change in circuit architecture, as expected; but when
DTRH synthesis is disrupted during CNS development,
larval feeding is depressed, and serotonergic projections

to the gut display increased branching and an increase
in the number and size of varicosities along the neurite
length. When mature, feeding 5-HT results in an increased, rather than decreased, feeding rate. Conversely,
when 5-HT synthesis is induced after CNS development,
larval feeding is depressed, as expected. Drosophila embryos exposed to 5-HT during the last 6 h of embryonic
development display increased feeding as mature larvae,
with a significant decrease in branching and overall varicosity number relative to controls, and a hyposensitivity
to exogenous 5-HT. When Drosophila embryos carrying
a null mutation in DTRH are exposed to 5-HT during
the last 6 h of embryogenesis, their feeding behavior is
rescued and branching number, area, and size of varicosities are significantly decreased. These data demonstrate that neuronal 5-HT levels during development inversely correlate with the complexity of gut fiber architecture and impairment of the behavior modulated by
that neuronal circuitry, and suggests that levels of 5-HT
must be tightly regulated for normal development of the
behavioral circuit.
Trophic Role for Serotonin in the

Development of a Feeding Circuit
219
Methods
Fly Culture
Flies were maintained in pint bottles containing standard
agar-cornmeal-yeast food at 25 C on a 12-hour light-dark cycle;
wandering third instar larvae for immunohistochemical analyses
were collected from these bottles. Staged larvae for behavioral
studies were collected from a population cage maintained at 25 C
on a 12-hour light-dark cycle. Females were allowed to lay eggs
overnight on apple juice-agar plates, and newly hatched larvae
were collected by maintaining plates with newly deposited eggs at
25 C for 24 h, and collecting first instars by migration of the animals onto yeast paste in the center of the agar plate. Second and
early third instar larvae were obtained by allowing first instars to
age for 24 or 48 h, respectively. For the 5-HTP studies, second instar larvae were fed yeast paste or yeast paste plus 10 mg/ml 5-HTP
(Sigma, St. Louis, Mo., USA) for 24 h before behavioral analyses
(details for this procedure can be found in Neckameyer [1996]).
Fly Strains
All strains were maintained as described above.
w1118 parental strain for the insertional mutation for DTRH
and the RNAi transgenic lines. These flies have white eyes, and
are isogenic for the second and third chromosomes. Obtained
from the Bloomington stock center.
pBac[PB]CG9122c01440 insertional mutation in the DTRH
gene generated by Exelixis, Inc., and obtained from the Bloomington stock center, first described in Thibault et al. [2004]. We
refer to this stock as pBacTRH [Neckameyer et al., 2007].
pP[w+mW.hs = GawB]elav C155 pan-neuronal Gal4 driver, obtained from the Bloomington stock center, which constitutively
drives expression in neuronal tissue. This driver initiates expression within neural tissues beginning after embryonic stage 9,
peaking within the next few hours, and decreasing afterwards
[Lin and Goodman, 1994].
pP[ELAV-GeneSwitch] inducible pan-neuronal Gal4 driver
obtained from Haig Keshishian. Referred to as GSelav. After exposure to the induction agent RU486 (mifeprestone), the driver is
activated and expression of the upstream activator sequence
(UAS) construct is observed within 3 h; peak expression occurs
by 22 h, and levels of expression are maintained for the next 24 h
[Osterwalder et al., 2001; data not shown]. Exposure of GSelav
second instar larvae to RU486 thus results in expression of the
transgene only in animals exposed to the ligand, so that uninduced animals serve as controls.
CSwu a wild-type Canton S strain established in the laboratory of Martin Heisenberg in 1978.
Generation of Transgenics
Complementary DNA encoding DTRH (CG9122, FBgn0035187)
was subcloned into both the SympUAS vector [Giordano et al.,
2002] downstream of the yeast Gal4 UAS (to generate DTRH
dsRNA) and a standard pUAS vector (stock No. 1000, Drosophila
Genomics Resource Center). These constructs were injected into
w1118 embryos using standard transformation techniques [Robertson et al., 1988]. The RNAi lines were generated as described in
Neckameyer et al. [2007] for DTPH RNAi transgenic flies; the
pUAS transgenics were generated using the services of Genetic Services, Inc. (Cambridge, Mass., USA). Two independent RNAi transgenic lines were used (TRHE, on chromosome 2, and TRHA, on
chromosome 3) to titrate DTRH expression levels, and thus, neuronal 5-HT synthesis. Similarly, two independent UAS lines for in220
duction of TRH levels were used (TRH32, on chromosome 2, and

TRH27 on chromosome 3). All lines were crossed with either the
pan-neuronal Gal4 promoter elavC155 or to the inducible pan-neuronal Gene Switch Gal4 transgenic promoter (GSelav) to drive constitutive or inducible neuronal expression, respectively, of these
transgenes [elav is expressed in all central and peripheral neurons;
Robinow and White, 1988; Yao and White, 1994]. The GSelav construct contains a human progesterone receptor-ligand-binding domain, which binds to UAS in the presence of the antiprogestin
RU486 to induce expression of UAS fused to the transgene of interest. Controlled feeding of RU486 permits discrete manipulation of
UAS activation; induction begins approximately 3 h after exposure
to RU486, and reaches a peak of expression at 22 h [Roman et al.,
2001]. The DTRH sequences used are unique and will not affect
expression of any other gene.
Analysis of DTRH Expression in the Transgenic Lines
Induction and knockdown of DTRH expression and 5-HT levels were assessed by quantitation of changes in fluorescent intensity of specific neurons after visualization with an antibody raised
against DTRH [Coleman and Neckameyer, 2005] or against 5-HT
(Spring Biosciences, Fremont, Calif., USA). The GSelav driver
[Osterwalder et al., 2001] was used to induce temporal expression
of the TRH RNAi transgene or the TRH UAS transgene using
RU486 (Mifepristone, Sigma). Second instar larvae (carrying one
copy each of the GSelav driver and the DTRH transgene) were
submerged in either RU486 (6 mg/ml in 80% ethanol, induced) or
80% ethanol (uninduced control) for 2 min, and aged for 24 h
(quantitation of neuronal DTRH levels) or for 4044 h (quantitation of neuronal 5-HT levels). Peak expression of the GSelav driver occurs approximately 21 h after induction [Osterwalder et al.,
2001], and we anticipated that changes in 5-HT levels would occur
about 24 h later. The optimum time point for reduced 5-HT levels
was estimated based on 24-hour pharmacological inhibition of
tyrosine hydroxylase to reduce dopamine levels after exposure to
a tyrosine hydroxylase inhibitor [Neckameyer, 1996]. The 5-HT
cell pattern in each brain was visualized under fluorescence, and
the single pair of neurons in the 8th abdominal neuromere [Valles
and White, 1988; Chen and Condron, 2008] was photographed at
the same magnification and exposure. These neurons were chosen since they were easily identified and quantified. The average
density of pixel intensity was sampled from seven regions within
each neuron, covering the entire cytoplasmic region (Northern
Eclipse, Empix Imaging, North Tonawanda, N.Y., USA). Density
was averaged for each singlet pair in larval brains from induced
and uninduced tissues treated and assessed in parallel, and statistical analyses were performed using Students t test.
Levels of constitutive knockdown were assessed by dissecting
larval brains from each genotype (elav C155/w1118; elav C155/DTRHE,
and elav C155/DTRHA). Each fly thus carried either one copy of the
elav Gal4 driver and one copy of the RNAi transgene, or the elav
Gal4 driver alone, which served as a control. Quantitation of
knockdown was accomplished as described above comparing
DTRH immunoreactivity in the neurons in the 8th abdominal
neuromere, and all genotypes were assessed in parallel. Statistical
analyses were performed using one-way ANOVAs followed by
Dunnetts post-test. Constitutive induction of the DTRH UAS
transgenes was assessed by observation of ectopic expression of
DTRH in numerous 5-HT and non-5-HT cells within the larval
brain (for DTRH levels) and by pixel intensity of the singlet neurons as described above (for 5-HT levels).
Neckameyer
The observed phenotypes are not due to ectopic DTRH expression in non-serotonergic neurons. DTRH is expressed solely
within 5-HT neurons; the only homologous genes are DTPH and
Drosophila tyrosine hydroxylase, both expressed in DA neurons.
However, there is considerable third base wobble, greatly limiting
homology at the DNA level, so it is not likely that the dsRNAi
DTRH transgene will affect the expression of these genes [Coleman and Neckameyer, 2005]. Although both the GSelav and
elav C155 promoters drive ectopic DTRH expression, by immunohistochemical analyses, 5-HT expression is limited to 5-HT neurons (data not shown). This is likely a consequence of substrate
and cofactor requirements.
Immunohistochemistry of Proventricular Tissue
The proventriculus and midgut from wandering third instar
larva were dissected in phosphate-buffered saline (PBS), fixed for
1 h (4% EM grade formaldehyde in 1! PBS) and washed thoroughly in PBT (1! PBS, 0.1% protease-free bovine serum albumin, 0.1% Triton X-100). Tissues were incubated in primary antiserotonin antibody and then washed thoroughly in PBT, followed
by incubation in secondary antibody (Alexa Fluor 568 goat antimouse or anti-rabbit IgG; 1:400 dilution; Invitrogen Molecular
Probes, Carlsbad, Calif., USA). After several washes in PBT, the
tissues were incubated in 4 mM sodium carbonate, mounted in 4%
n-propyl gallate in 20 mM sodium carbonate, and viewed under
fluorescence. To enhance 5-HT immunoreactivity, dissected gut
tissues were preincubated in 10 6 M 5-HT for 1 h at room temperature before extensive washing and incubation with the primary antibody. Previous studies [Budnik et al., 1989; Sykes and
Condron, 2005] have demonstrated that this concentration of exogenous 5-HT does not affect neuronal architecture or varicosity
density in immunohistochemical analyses.
Analysis of Neuronal Circuitry
Serotonergic fibers projecting to the proventriculus from each
genotype were assessed in at least six independent immunohistochemical experiments and were photographed at the same resolution. Quantification of varicosities and branching of fibers were
analyzed using Neuroleucida, version 5 and Neuroexplorer (MBF
Bioscience, Chicago, Ill., USA). Individual projections were traced
within the body of the proventriculus, and varicosity number,
branches, and number of large varicosities per 100-!m length
were quantified. Area (!m2) per large varicosity was also determined, as was the number of varicosities 5 !m2 or greater in diameter along the neurite length. A varicosity was defined as a
brighter, discrete unit sufficiently enlarged beyond the size of the
fiber, and large varicosities were defined as those larger than
1 !m2 (i.e. larger than the width of the neurite fiber).
Behavioral Paradigms
Retraction of the cephalopharyngeal sclerites is a standard
way to assess larval feeding rate, having been used for decades in
population evolution studies as a correlate of rapid development
[e.g. Sewall et al., 1975; Joshi and Mueller, 1988].
Feeding
A single second instar larvae was placed in the center of a 2%
agar-filled Petri dish overlaid with a 2% yeast solution, and the
number of mouth hook contractions was counted for 1 min after
a 30-second acclimation period [Neckameyer, 1996].

Locomotion
Each animal was placed on a 2% agar substrate and allowed to
acclimate for 30 s. The larva was then observed as it crawled over
the substrate for a period of 1 min, and each posterior to anterior
contractile wave was counted [Neckameyer, 1996]. In general, the
same animal was first assessed for locomotor behavior, and then
for feeding.
Statistics
Statistical analyses were accomplished by one-way ANOVA
using Tukeys or Dunnetts post-hoc tests (for constitutive knockdown with elav C155 Gal4) or by Students t test (for temporal
knockdown with GSelav or for embryonic studies). There were 40
animals from 46 independent experiments.
Embryonic Exposure to 5-HT
Staged embryos were aged until 16 h after egg laying, dechorionated in 50% chlorox (Walmart), washed in embryo wash (0.7%
NaCl, 0.05% Triton X-100), rinsed with H2O, air dried and then
devitellinized by exposure to octane (Sigma Aldrich, electronic
grade) for 5 min. Embryos were incubated in 10 6 M 5-HT in serum-free medium (Sf-900 II SFM 1!, Gibco) for 46 h until
hatching, and then placed in yeast paste on an agar plate and kept
at room temperature. Animals were analyzed for feeding behavior
as second instar larvae; analysis of architecture of the serotonergic
fibers projecting to the gut was accomplished using animals aged
to third instar.
Results
Neuronal 5-HT Acts as a Transmitter in the Brain to

Modulate Feeding in Drosophila
While the stomatogastric system is heavily innervated
by 5-HT (fig.1), and 5-HT is known to modulate feeding
in several other species, including insects, formal demonstration that 5-HT modulates feeding in Drosophila has
not been established. We have previously reported that
DTRH is solely responsible for neuronal 5-HT synthesis;
the peripheral Tph, DTPH, which also encodes a phenylalanine hydroxylase activity, is expressed only in dopaminergic, and not serotonergic, neurons [Neckameyer et
al., 2007]. We therefore generated transgenic RNAi lines
to knockdown DTRH levels, as well as transgenic UAS
lines to induce DTRH expression. We used a GeneSwitch
Gal4 pan-neuronal transgenic line to induce expression
of both transgenes in second instar larvae; the GeneSwitch
construct contains a human progesterone receptor-ligand-binding domain that binds Gal4 to UAS sequences
in the presence of the antiprogestin RU486. Controlled
feeding of RU486 permits discrete manipulation of UAS
activation [Roman et al., 2001]; since Drosophila do not
contain a receptor that binds RU486, there is no expression in its absence. Induction of expression occurs within
3 h and peaks at 22 h, with lowered expression over at least
221
GSelav TRH RNAi uninduced

GSelav TRH RNAi RU486
60
40
20
175
150
125
**
100
75
50
25
0
50
40
30
20
100
75
50
25
0
40
20
0
GSelav RNAi uninduced control

GSelav RNAi uninduced 5-HTP
GSelav RNAi RU486 control
GSelav RNAi RU486 5-HTP
***
***
***
70
60
50
40
30
20
10
0
GSelav UAS uninduced control

GSelav UAS uninduced 5-HTP
GSelav UAS RU486 control
GSelav UAS RU486 5-HTP
Tukey's multiple comparison test
GSelav32 uninduced control vs. GSelav32 uninduced 5-HTP
GSelav32 uninduced control vs. GSelav32 RU486 control
GSelav32 uninduced control vs. GSelav32 RU486 5-HTP
GSelav32 uninduced 5-HTP vs. GSelav32 RU486 control
GSelav32 uninduced 5-HTP vs. GSelav32 RU486 5-HTP
GSelav32 RU486 control vs. GSelav32 RU486 5-HTP
p value
p < 0.001
p < 0.001
p < 0.001
p > 0.05
p < 0.001
p < 0.001
Fig. 2. Neuronal 5-HT acts as a neurotransmitter to modulate
feeding behavior in Drosophila larvae. Larval brains were dissected from control (uninduced) or induced (6 mg/ml RU486 for
2 min) second instar larvae 40 h after induction using the
GeneSwitch elav inducible pan-neuronal driver to express DTRH
transgenic constructs in the CNS. The 5-HT cell pattern in each
brain was visualized with a monoclonal antibody raised against
5-HT and viewed under fluorescence. The two unpaired neurons
in the 8th abdominal neuromere (boxed, a) were photographed at
the same magnification and exposure, and the pixel intensity of
the neuronal fluorescence was determined for both knockdown
of 5-HT synthesis using a TRH RNAi transgenic construct (b) and
induction of 5-HT synthesis using a TRH UAS transgenic construct (c). Uninduced and induced animals were assessed in parallel. GSelav TRH RNAi uninduced, n = 37 neurons; GSelav TRH
RNAi RU486, n = 46 neurons; GSelav TRH UAS uninduced, n =
43 neurons; GSelav TRH UAS RU486, n = 36 neurons. *p ! 0.05,
222
60
Body wall contractions/min
125
**
80
10
150
GSelav TRH UAS uninduced

GSelav TRH UAS RU486
60
Body wall contractions/min
Mouth hook contractions/min

Mouth hook contractions/min
175
80
100
Pixel intensity
Pixel intensity
100
**p ! 0.01, Students t tests. d Treatment with 10 mg/ml 5-HTP,

the product of the DTRH reaction, decreases feeding, and decreasing 5-HT levels by reducing DTRH expression increases
feeding rate. Animals with reduced neuronal DTRH synthesis
when given exogenous 5-HTP display normal feeding. *p ! 0.05,
**p ! 0.01, one-way ANOVA followed by Dunnetts multiple comparisons post-test. e Locomotion is unaffected, demonstrating
that pharyngeal contractions in the animals are normal. f Exogenous 5-HTP and induction of neuronal 5-HT synthesis results in
decreased feeding. The effect is additive if induced animals are
also given 5-HTP. *** p ! 0.001, one-way ANOVA followed by
Tukeys multiple comparisons post-test. g Locomotion is unaffected, demonstrating that pharyngeal contractions in the animals are normal. n = 40 for behavioral analyses, from 46 independent experiments. Lines above the graph depict standard error
of the mean.
Neckameyer
the next 24 h [Osterwalder et al., 2001; data not shown].

Exposure of GSelav second instar larvae to RU486 thus
results in expression of the transgene only in animals exposed to the ligand, so that the uninduced animals serve
as age- and environment-matched controls.
Although elav directs expression within all neural tissue, DTRH expression is limited to 5-HT neurons. When
analyzing knockdown using a DTRH RNAi, the homology between DTRH and its sister enzymes (Drosophila
tyrosine hydroxylase and DTPH) at the DNA level is sufficiently limited that the DTRH RNAi and UAS constructs will not affect expression of these genes [Neckameyer and White, 1992; Neckameyer et al., 2007]. Thus,
no other cells were directly affected, permitting controlled and specific manipulation of neuronal 5-HT synthesis in staged larvae. When analyzing induction of
DTRH using a UAS transgene, while DTRH protein was
ectopically observed in the CNS, the 5-HT pattern was
unchanged except for intensity, implying that all the necessary factors for 5-HT synthesis (substrate, cofactor)
were only available in 5-HT neurons.
Changes in DTRH expression were assessed by quantitation of changes in fluorescent intensity in specific
5-HT neurons after visualization with the DTRH antibody (data not shown) as well as with an antibody raised
against 5-HT (fig.2). DTRH protein is undetectable in
brain by Western immunoblot analysis, since it is expressed in a limited subset of neurons within the CNS
[Neckameyer et al., 2007], and a decrease in neuronal Tph
activity levels would not be observed with these transgenes, since DTPH Tph activity would not be affected and
would obscure results from enzymatic assays of brain tissues. The intense 5-HT expression within the ring gland,
which is physically associated with the brain lobes, could
obscure HPLC or ELISA analyses of 5-HT levels in dissected larval brains. We thus chose to quantitate DTRH
and 5-HT levels within easily identifiable 5-HT larval
neurons. The two unpaired, bilaterally symmetric neurons in the 8th abdomere (fig.2a) were assessed in third
instar larval brains from uninduced and induced GSelav/
DTRH RNAi (fig. 2b) and Gselav/DTRH UAS (fig. 2c)
animals (each carrying a single copy of the inducible
GSelav Gal4 driver and a single copy of the TRH transgene) treated in parallel (quantitation occurred 24 h after
induction for DRTH and 44 h after induction for 5-HT
levels). Fluorescence from these neurons was not affected
by the fluorescence from the dense serotonergic neuropil
found within the brain lobes and ventral ganglia. The
other neurons within the ventral ganglion are bilaterally
symmetrical doublets, and often partially overlap, which
affects the fluorescence intensity. The relative fluorescence of each neuron, photographed at the same exposure
and magnification, was determined by quantitating the
relative intensity of a circular sampling region placed at
seven different locations and covering the cytoplasm of
each cell. This number was averaged to give a score for
each neuron. Both DTRH (uninduced, n = 14 neurons;
induced, n = 22 neurons; p = 0.0002, Students t test, data
not shown) and 5-HT levels (fig.2b; p ! 0.05) were reduced when expression of the DTRH RNAi transgene
was induced in larvae during the second instar, demonstrating that reduction in DTRH levels resulted in diminished 5-HT synthesis. Similarly, induction of DTRH expression resulted in increased 5-HT fluorescence (fig.2c;
p ! 0.01); relative levels of DTRH could not be assessed
since the protein is ectopically expressed in the brain (see
fig.7a).
We then examined these animals for changes in feeding rate by feeding 5-HTP (the product of the DTRH reaction) to increase 5-HT levels (this approach has been
used successfully to increase neuronal 5-HT levels [e.g.
Kaplan et al., 2008]), and by inducing knockdown of the
RNAi construct to decrease DTRH levels and thus 5-HT
synthesis (fig. 2d). In mammals, increased brain serotonin levels decrease appetite, and lesioning of 5-HT neurons increases feeding behavior [see Blundell, 1986, for
review]. Since Drosophila larvae at this developmental
stage (second to third instar) engage full-time in feeding
activity in the presence of food, the continuous contraction of the mouth hooks is equivalent to the feeding rate
[Sewall et al., 1975], in anticipation of the enormous energy demands for metamorphosis. Also, in contrast to
adults, there is no gender difference in feeding rate [de
Miranda and Eggleston, 1988].
As expected, when the animal was placed in a yeast
solution, feeding 5-HTP decreased its feeding rate (measured by the number of mouth hook contractions), and a
reduction in neuronal 5-HT synthesis increased feeding
rate. Animals with reduced neuronal DTRH levels fed
5-HTP displayed a feeding rate indistinguishable from
normal. To establish that the effect of 5-HTP was not due
to a generalized physiological impairment, we also assayed locomotor behavior, which is stable in second and
third instar larvae [Sewall et al., 1975]. Foraging larvae
use their mouth hooks to dig into the substrate, which
may facilitate pulling the larvae forward by muscular
body wall contractions. These same muscles are used by
the animal to shovel food into the larval pharynx. Changes in 5-HT levels either by feeding 5-HTP or by reducing
DTRH levels had no effect on locomotion (fig.2e).

223
Mouth hook
contractions/min
Pixel intensity
150
**
100
50
**
200
150
***
100
50
***
40
20
200
*** ***
***
***
100
50
elavC155/w1118-controlelav
elavC155/w1118-10 mg/ml 5-HTP
elavC155/TRHA-control
elavC155/TRHA-10 mg/ml 5-HTP
DTRHA/w1118-control
DTRHA/w1118-10 mg/ml 5-HTP
Mouth hook
contractions/min
Mouth hook
contractions/min
150
60
250
200
w1118/TRHA
elavC155/w1118
elavC155/TRHE
elavC155/TRHA
80
Body wall
contractions/min
elavC155/w1118
elavC155/TRHE
elavC155/TRHA
200
w1118-control
w1118-5-HTP
pBacTRH-control
pBacTRH-5-HTP
150
**
**
**
100
50
0
Fig. 3. Constitutive reduction in neuronal 5-HT results in de-
were assessed in parallel. *p ! 0.05; **p ! 0.01, one-way ANOVAs

followed by Dunnetts multiple comparisons post-test. The animals were then assayed for feeding (b) and locomotor behavior (c).
An additional control, DTRH/w1118, was included in these studies.
Feeding was decreased in DTRH knockdown animals, and locomotion was unaffected. d Feeding rate was dramatically increased
in elav C155/DTRHA larvae given 10 mg/ml 5-HTP for 24 h. e Consistent with the results obtained with the DTRH knockdown
lines, the DTRH-null pBacTRH mutation was similarly assessed
for feeding rate and compared with the parental w1118 controls.
n = 40 for each genotype and condition. Lines above the graph
depict standard error of the mean. **p ! 0.01; ***p ! 0.001, oneway ANOVAs followed by Tukeys post-test.
Similarly, feeding 5-HTP to uninduced GSelav/TRH

UAS animals reduced appetite; induction of neuronal
5-HT synthesis also reduced feeding, and induced animals given 5-HTP displayed an additive effect (fig.2f).
Again, motor behavior was normal (fig.2g), demonstrating that changes in neuronal 5-HT specifically affected
appetitive behavior. Thus, 5-HT acts as a neurotransmitter to modulate feeding in Drosophila, as it does in all
other species tested.
DTRH expression during neuronal development. It was

not feasible to reduce 5-HT levels during the last 6 h of
embryonic development via pharmacological inhibition
of DTRH or via neuronal induction of the DTRH transgene, since neither approach would have resulted in reduced 5-HT levels during the critical time frame. We
thus chose to use the pan-neuronal elav driver to reduce
DTRH levels, and thus 5-HT synthesis, beginning from
stage 9 of embryogenesis. Since the elavC155 Gal4 driver
is strongest during late embryogenesis, and displays reduced expression in the larval CNS, use of this driver allowed us to knockdown DTRH levels during embryogenesis, with a lesser reduction in expression during the
larval stages. We identified two independent insertions
of the same dsRNAi transgene (DTRHE, on chromosome 2, and DTRHA, on chromosome 3), and quantified
pressed feeding behavior and abnormal responses to exogenous

5-HT. Two independent RNAi transgenic lines (DTRHE, inserted
on chromosome 2, and DTRHA, on chromosome 3) were used to
titrate constitutive knockdown of DTRH levels throughout CNS
development. Larval brains were dissected from each genotype
(elavC155/w1118, n = 9; elav C155/DTRHE, n = 10, and elavC155/
DTRHA, n = 14). The 5-HT cell pattern in each brain was visualized with a polyclonal antibody raised against DTRH [Coleman
and Neckameyer, 2005] and viewed under fluorescence. a The
pixel intensity of the DTRH immunofluorescence of the two unpaired neurons in the 8th abdominal neuromere was determined
to ascertain the extent of knockdown of DTRH. All genotypes
Constitutively Reduced Neuronal Serotonin Levels

throughout Development Result in Perturbed Larval
Feeding Behavior
To address the trophic rather than transmitter role for
5-HT in the modulation of feeding rate, we used a constitutive pan-neuronal driver (elavC155 Gal4) to perturb
224
Neckameyer
the relative intensity of DTRH expression in these animals using the approach described above to demonstrate
that a single copy of the driver and the RNAi construct
significantly decreased DTRH levels, and DTRHA resulted in a greater knockdown of DTRH levels than did
DTRHE, thus permitting titration of the decrease in
DTRH expression and 5-HT synthesis (fig.3a). Progeny
from parental animals (w1118, the strain used for generation of these transgenics, and the elav Gal4 driver) were
used as controls. Decreased levels of DTRH knockdown
directly correlated with reduced feeding relative to controls (w1118/DTRHA and elavC155/w1118; fig.3b) and general motor activity was unaffected (fig.3c), demonstrating the effect was specific to feeding behavior and not
the result of a generalized impaired physiology. While
consistent with our published observations for the null
DTRH mutation, pBacTRH [Neckameyer et al., 2007],
these findings are contrary to what was observed for
knockdown of the DTRHA transgene using the inducible pan-neuronal promoter, suggesting a critical role for
5-HT during CNS development in the maturation of this
neural circuitry.
We predicted, since feeding exogenous 5-HTP decreases appetite, and since appetite is also depressed in
animals with developmentally reduced 5-HT levels, that
exposing these animals to 5-HTP would significantly
further impair feeding rate. Surprisingly, we observed
that feeding 5-HTP to elavC155/DTRHA (fig. 3d) and
pBacTRH animals (fig.3e) resulted in a highly increased
appetite; locomotor control was unaffected (data not
shown). These results clearly demonstrate that although
5-HT still acts to modulate feeding at this circuit, its actions are impaired as a consequence of reduced neuronal
5-HT levels during development.
Titration of Neuronal 5-HT Synthesis during CNS
Development Affects the Architecture of 5-HT Fibers
Projecting to the Proventriculus
Using Drosophila mutants deficient in dopa decarboxylase, and thus unable to synthesize either dopamine or
serotonin, Budnik et al. [1989] noted that these animals
displayed increased branching of serotonergic fibers projecting to the midgut, as well as increased numbers of
varicosities; however, these studies could not distinguish
whether this phenotype was a consequence of altered serotonin or altered dopamine levels, or both, and whether
the perturbed fibers were sensitive to these factors acting
neuronally or peripherally. We thus ascertained whether
manipulating neuronal DTRH levels to titrate 5-HT levels throughout CNS development would correlate with
this phenotype, as it did with feeding rate. This was accomplished by quantitating the numbers of branch
points, as well as the size and number of varicosities,
along the lengths of the serotonergic fibers projecting to
the foregut. We limited our analyses to the fibers within
the center of the proventriculus, since they are easily visualized and, in control animals, display little branching;
the earlier studies with dopa decarboxylase mutants assessed the branching of these fibers in the midgut, which
normally displays greater branching. It is not possible to
quantitate changes in branching and varicosity number
within the frontal nerve and recurrens nerve, since they
are tightly bundled, and single fibers cannot be adequately resolved, nor can individual varicosities be distinguished (see fig.1).
We analyzed gut tissues from the two transgenic RNAi
lines driven by the pan-neuronal elav promoter, as well as
from the pBacTRH mutation. The pBacTRH mutation results in barely detectable levels of 5-HT within the CNS;
although the mutation does not express DTRH protein,
there is a small amount of 5-HT taken up via the serotonin transporter into the nervous system [Neckameyer
et al., 2007]. The w1118 allele was used in the generation of
the elav Gal4 driver [Lin and Goodman, 1994] as well as
the pBacTRH mutation, but they still differ in their genetic backgrounds. In addition, w1118 animals are blind,
which impacted their behavioral analyses (w1118 animals
move and feed at a reduced rate relative to elav Gal4/w1118,
data not shown; see also Mackenzie et al. [1999]). Therefore, the mutation and the transgenic knockdown lines
were not directly compared.
We observed a correlative increase in arborization
with decreasing levels of neuronal 5-HT in the DTRH
transgenic knockdown animals relative to controls
(fig.4). While there was some overlap in branching phenotype, animals with constitutively decreased levels of
neuronal 5-HT showed greater complexity of arborization.
We also assessed the total number of varicosities per
unit length (0.1 mm), defining a varicosity as a brightly
staining, discrete vesicle containing 5-HT dotting the
neurite length, with an area about the width of the neurite
fiber, which is less than 1 !m2. Large varicosities were
defined as those enlarged beyond the width of the neurite
(area greater than 1 !m2). Ultrastructural analysis of
similar varicosities in the CNS neuropil suggests they are
presynaptic terminals that release 5-HT [Sykes and Condron, 2005]. In general, the number of total varicosities
increased in animals with perturbed DTRH levels (fig.5).
While large varicosities could be observed in the gut proDev Neurosci 2010;32:217237
225
elav/w1118
elav/TRHE
elav/TRHA
2.0
Branches/0.1 mm length
1.5
elavC155/w1118
elavC155/TRHE
elavC155/TRHA
w1118
pBacTRH
1.0
**
elavC155/w1118
elavC155/TRHE
elavC155/TRHA
w1118
pBacTRH
**
*
0.5
Fig. 4. Decreasing amounts of neuronal serotonin during critical
jecting fibers from 45 independent gut dissections (controls);

elav C155/TRHE, 45 projecting fibers from 34 independent gut dissections; elav C155/TRHA, 54 projecting fibers from 39 independent gut dissections; w1118 and pBacTRH, 18 and 19 projecting
fibers from 17 and 20 independent gut dissections, respectively.
Lines above the graph depict standard error of the mean. Significance was assessed by one-way ANOVAs followed by Dunnetts
post-tests (elav C155/w1118, DTRHA/w1118, elav C155/TRHE, elav C155/
TRHA) or Students t tests (w1118 and pBacTRH). * p ! 0.01;
** p ! 0.001. These groups were not analyzed together because
they did not share the same genetic background.
jections of control animals, animals with constitutively

decreased neuronal 5-HT levels displayed a far greater
number of these vesicles; some even in the range of 25
50 !m2 (fig.5a). While the pBacTRH mutant did not display the greatest increase in varicosity number (fig.5b),
gut projections from the pBacTRH mutant appeared to
have a significantly greater number of larger varicosities
(15 !m2; fig.5c). Closer observation of the larger vari-
cosities revealed that they were composed of numerous

smaller ones (fig. 5d), suggesting that, with decreasing
amounts of neuronal 5-HT during development, they
sufficiently increase in number and then fuse. Thus,
while pBacTRH gut fibers may appear to have fewer discrete varicosities relative to the neuronal TRH knockdowns, the significantly larger vesicles along the lengths
of the gut fibers may contain a far greater number of
developmental windows inversely correlate with the degree of

branching of serotonergic fibers projecting to the proventriculus.
The mouthparts, brain and foregut were dissected from third instar larvae and immunostained with an anti-serotonin antibody.
The samples were viewed under fluorescence and photographed
at the same magnification. a Representative examples of increased
arborization in animals with constitutively decreased neuronal
5-HT levels. Scale bar = 20 !m. b Quantitation of arborization.
Branch points for projecting fibers were quantitated using the
Neuroleucida analysis program. elav C155/w1118, 50 projecting fibers from 40 independent gut dissections; DTRHA/w1118, 53 pro-
226
Neckameyer
Varicosities/0.1 mm length
50
elav/TRHE
***
***
**
40
30
20
10
0
elav/TRHA
3
Varicosities >5 m2/0.1 mm length
elav/w1118
elavC155/w1118
DTRHA/w1118
elavC155/TRHE
elavC155/TRHA
w1118
pBacTRH
***
***
Fig. 5. Decreasing amounts of neuronal serotonin during development inversely correlates with varicosity number and varicosity
size of serotonergic fibers projecting to the proventriculus. The
mouthparts, brain and foregut were dissected from third instar
larvae and immunostained with anti-serotonin. The samples
were viewed under fluorescence and photographed at the same
magnification. a Representative examples of the increase in large
varicosities in animals with constitutively decreased neuronal
5-HT levels. Scale bar = 20 !m. Number of total varicosities (b)
and large varicosities with an area measuring greater than 5 mm2
(c) for projecting fibers were quantitated using Neuroleucida.
elav C155/w1118, 52 projecting fibers from 40 independent gut dissections; DTRHA/w1118, 53 projecting fibers from 45 independent
gut dissections; elav C155/TRHE, 46 projecting fibers from 34 independent gut dissections; elavC155/TRHA, 58 projecting fibers
from 39 independent gut dissections; w1118 and pBacTRH, 23 and
24 projecting fibers from 17 and 20 independent gut dissections,
respectively. Lines above the graph depict standard error of the
mean. Significance was assessed by one-way ANOVAs followed
by Tukeys post-tests (elav C155/w1118, DTRHA/w1118, elav C155/
TRHE, elav C155/TRHA) or Students t tests (w1118 and pBacTRH).
** p ! 0.01; *** p ! 0.001. These groups were not analyzed together because they did not share the same genetic background.
d Examples of less dense varicosities (from pBacTRH gut fibers),
which appear to contain multiple smaller varicosities. Scale bar =
10 !m.
smaller, discrete varicosities whose membranes have

fused to form the large ones, consistent with a correlative
increase in varicosity number with decreasing developmental neuronal 5-HT levels.
To demonstrate that these abnormalities were the result of perturbations in developmental neuronal 5-HT
levels, and not a consequence of altered neuronal 5-HT
transmission, we analyzed the architecture of serotoner-
gic fibers projecting to the proventriculus in GSelav/

TRHA RNAi animals, inducing neuronal expression of
the RNAi transgene during second instar, and analyzing
the animals during early third instar (4044 h after induction, when 5-HT levels have been significantly reduced; fig.6). No differences were observed in arborization of fibers (fig.6a), varicosity number (fig.6b), number
of large varicosities (fig.6c), or in varicosity area (fig.6d).

227
50
the CNS has matured does not affect the

architecture of serotonergic fibers projecting to the proventriculus. Tissues were dissected from control (uninduced) or induced (6 mg/ml RU486 for 2 min) GSelav/
TRHA larvae 4044 h after induction.
a Branch number. b Varicosity number.
c Number of large varicosities per 0.1 mm.
d Varicosity area. Control, 30 projecting
fibers from 24 independent gut dissections; RU486, 28 projecting fibers from 21
independent gut dissections. Lines above
the graph depict standard error of the
mean. Significance was assessed by Students t tests.
0.50
0.25
Large varicosities/
0.1 mm length
Varicosities/
0.1 mm length
Fig. 6. Reduction in neuronal 5-HT after
0.75
20
10
2
1
3
2
1
analyses of third larval instar CNS tissues from animals carrying

a UAS TRH transgenic construct expressed in the CNS and visualized with an antibody raised against DTRH. a GSelavTRH
UAS32, uninduced. As expected, the pattern is restricted to 5-HT
neurons. b GSelavTRH UAS32, induced 6 mg/ml RU486 for 2 min
as third instar larvae. DTRH is expressed ectopically in all neurons, demonstrating the protein is induced. Brains were dissected
24 h after induction. c The same transgene driven by elavC155 Gal4.
d An independent UAS TRH (TRH UAS27) on a separate chro-
Pixel intensity
Fig. 7. Induction of DTRH levels during CNS development does

not result in increased 5-HT levels. ad Immunohistochemical
228
30
30
40
GSelav TRHA control

GSelav TRHA RU486
Varicosity
area (mm2)
Branches/
0.1 mm length
1.00
10
elavC155/w1118
elavC155/TRH UAS27
elavC155/TRH UAS32
20
mosome, also driven by elav C155 Gal4. Both transgenes are expressed ectopically, as expected. Scale bar = 50 !m. e In contrast
to expressing the DTRH UAS after the larval CNS has matured
(fig.2), there is no increase in 5-HT levels, measured by the pixel
intensity of the two unpaired neurons in the 8th abdominal neuromere in brains from newly hatched first larval instars visualized
with an antibody raised against 5-HT. elav C155/w1118, n = 16;
elav C155/UAS27, n = 20, and elav C155/UAS32, n = 8. Significance
was assessed by one-way ANOVAs followed by Tukeys post-tests.
Lines above the graph depict standard error of the mean.
Neckameyer
Fig. 8. Embryos exposed to 5-HT during
CSwu-control
CSwu-5-HT
150
100
50
0
200
150
100
50
0
***
70
Body wall
contractions/min
Mouth hook
contractions/min
200
Mouth hook
contractions/min
the last 6 h of development display increased feeding as larvae. Wild-type (CSwu)

embryos were collected for 2 h, aged for
14 h, dechorionated, devitellinized, and exposed to 10 6 M 5-HT in serum-free media
until hatching. Control embryos were processed in parallel but not exposed to 5-HT.
Treated embryos displayed a significant increase in feeding rate (a) with no change in
locomotion (b). *** p ! 0.001, Students t
tests. c The effect of exogenous 5-HT was
determined by feeding 10 mg/ml 5-HTP
for 24 h before assaying for feeding. Although exogenous 5-HTP decreases feeding rate, the response is not as robust as it
is for controls. Statistical analyses accomplished by one-way ANOVAs followed by
Tukeys multiple comparisons post-test
(n = 40). Lines above the graph depict standard error of the mean.
60
50
40
30
20
10
0
CSwu-control-control
CSwu-control-5-HTP
CSwu-5-HT-control
CSwu-5-HT-5-HTP
Tukeys multiple comparison test
CSwu-control-control vs. CSwu-control-5-HTP
CSwu-control-control vs. CSwu-5-HT-control
CSwu-control-control vs. CSwu-5-HT-5-HTP
CSwu-control-5-HTP vs. CSwu-5-HT-control
CSwu-control-5-HTP vs. CSwu-5-HT-5-HTP
CSwu-5-HT-control vs. CSwu-5-HT-5-HTP
p value
p < 0.001
p < 0.001
p < 0.05
p < 0.001
p < 0.001
p < 0.05
5-HT Synthesis Is Tightly Regulated during CNS

Development
To determine whether increased 5-HT levels above a
normal threshold also affected development of the serotonergic feeding circuit, we attempted to use the elav C155
driver to express a TRH UAS transgene in the developing CNS. When DTRH expression was induced in second larval instars using the GSelav driver, DTRH was
expressed in cells throughout the larval CNS (fig.7a, b),
which results in increased neuronal 5-HT levels (see
fig.2c). 5-HT was observed only in 5-HT neurons (data
not shown), suggesting that although DTRH may be expressed in dopamine neurons, regulatory mechanisms
must be in place to ensure that only the correct transmitter is synthesized (tyrosine hydroxylase, the ratelimiting enzyme in dopamine synthesis, is a sister enzyme that shares many biochemical features, including
an absolute requirement for tetrahydrobiopterin as cofactor). When the constitutive pan neuronal elav Gal4
driver was used to induce the DTRH UAS transgenes,
DTRH was also expressed ectopically (fig.7c, d). However, unlike the inducible promoter, this induction did
not result in increased neuronal 5-HT levels in late-stage
(second and third instar) larvae (data not shown). Since
it was possible that DTRH synthesis and 5-HT levels
might be normalized over time, the CNSs of newlyhatched larvae were examined for increased 5-HT levels.
However, even at this early stage, no increase in neuronal 5-HT was observed (fig.7e). Since the enzyme activ-
ity is limited by cofactor and substrate availability, it is

likely that these factors preclude increased synthesis
during this developmental period, even though DTRH
levels are induced. These data suggest that regulatory
mechanisms are in place to not only restrict 5-HT expression only to 5-HT neurons, but that the developing
CNS is highly sensitive to perturbations above normal
5-HT threshold levels.

Increased Levels of 5-HT during Development

Result in Increased Feeding and Decreased Gut
Fiber Complexity
Since constitutive induction of neuronal 5-HT during
CNS development was not feasible, staged wild-type
(CSwu) embryos were aged until 16 h after egg laying
(when projections from serotonergic cell bodies are first
observed), and directly exposed to 106 M 5-HT until
hatching (fig. 8). In contrast to what was observed for
constitutive neuronal induction of transgenic animals
carrying a DTRH RNAi transgene, the feeding rate of
these animals, when assessed as second instar larvae, was
significantly increased (fig.8a); locomotor behavior was
normal (fig.8b). Larvae exposed to 5-HT during the last
6 h of embryogenesis were also capable of reducing their
feeding rate when given exogenous 5-HTP as second instar larvae, although not to wild-type levels, suggesting a
hyposensitivity to the neurotransmitter actions of serotonin (fig.8c). In addition, branching of gut fibers was
significantly decreased (fig.9a), the opposite of that ob229
0.5
Control
+5-HT
50
40
Large varicosities/
0.1 mm length
1.0
Varicosities/0.1 mm length
Branches/0.1 mm length
1.5
30
20
10
0
Control
+5-HT
***
2
1
0
Control
+5-HT
Fig. 9. Embryos exposed to 5-HT during the last 6 h of develop-
number of individual varicosities was similar (b), the number of

large varicosities was also significantly less (c). There were 29 projections from 25 guts from 6 independent experiments for control
and 36 projections from 30 guts from 5 independent experiments.
*p ! 0.05, ***p ! 0.001, Students t tests. Lines above the graph
depict standard error of the mean. Representative 5-HT immunostaining of control (d) and treated (e) embryos. Scale bar =
50 !m.
served when neuronal 5-HT levels were decreased during

development. While the total number of varicosities
along the fiber lengths was unchanged (fig.9b), the number of large (11 !m2) varicosities was significantly decreased relative to controls (fig.9c). Since the larger varicosities appear to be composed of multiple fused smaller
varicosities, this would suggest an overall decrease in varicosity number, the converse of that observed with reduced neuronal developmental 5-HT levels. Thus, developmental levels of neuronal 5-HT inversely correlate with
the amount of branching of serotonergic fibers projecting
to the gut, as well as varicosity number along the fiber
lengths. In addition, too little neuronal 5-HT during development results in animals with decreased feeding,
while excess 5-HT during development results in in-
creased feeding relative to controls, demonstrating that

the effects of developmental 5-HT on fiber architecture
and behavioral output of the mature circuit can be titrated, and suggesting that there is a concentration threshold
above and below which the integrity of the circuit is compromised.
ment display decreased branching of 5-HT fibers projecting to the

gut, as well as fewer overall numbers of serotonergic varicosities.
Wild-type (CSwu) embryos were collected for 2 h, aged for 14 h,
dechorionated, devitellinized, and exposed to 10 6 M 5-HT in serum-free media until hatching. Control embryos were processed
in parallel but not exposed to 5-HT. a Treated embryos displayed
decreased branching of gut fibers relative to controls. While the
230
Feeding and Gut Fiber Architecture Abnormalities

in DTRH Null Mutants Are Rescued by Exposure to
5-HT during the Last 6 Hours of Embryogenesis
pBacTRH embryos were aged until 16 h after egg laying and directly exposed to 106 M 5-HT until hatching,
when they were placed in yeast paste without additional
exogenous 5-HT. When assessed as second instar larvae,
their feeding behavior was indistinguishable from that of
Neckameyer
w1118
pBacTRH
pBacTRH 5-HT
pBacTRH +5-HT
150
***
200
***
Mouth hook
contractions/min
Mouth hook
contractions/min
200
pBacTRH 5-HT 5-HTP

pBacTRH 5-HT +5-HTP
pBacTRH +5-HT 5-HTP
pBacTRH +5-HT +5-HTP
100
50
150
100
50
0
Tukeys multiple comparison test
pBacTRH 5-HT 5-HTP vs. pBacTRH 5-HT +5-HTP
pBacTRH 5-HT 5-HTP vs. pBacTRH +5-HT 5-HTP
pBacTRH 5-HT 5-HTP vs. pBacTRH +5-HT +5-HTP
pBacTRH 5-HT +5-HTP vs. pBacTRH +5-HT 5-HTP
pBacTRH 5-HT +5-HTP vs. pBacTRH +5-HT +5-HTP
pBacTRH +5-HT 5-HTP vs. pBacTRH +5-HT +5-HTP
to pBacTRH animals that were not exposed to any treatment as

well as to pBacTRH animals exposed only to serum-free medium;
treated animals were indistinguishable from the w1118 parental
controls. b Treatment with 5-HT during the last 6 h of embryogenesis also rescues the abnormal feeding response to the actions
of exogenous 5-HTP. ***p ! 0.001, one-way ANOVAs followed by
Tukeys multiple comparisons post-test (n = 40). Lines above the
graph depict standard error of the mean.
Varicosity
area (m2)
1.0
***
0.5
2.0
1.5
1.0
40
2
0
***
***
0
50
pBacTRH control
pBacTRH +5-HT
0.5
8
Varicosities/
0.1 mm length
Branches/
0.1 mm length
1.5
Varicosities >5 m2/

0.1 mm length
Fig. 10. Embryos lacking DTRH function exposed to 5-HT for the
last 6 h of embryonic development display normal feeding behavior. pBacTRH embryos, which are null for DTRH function and
contain barely detectable levels of neuronal 5-HT, were collected
for 2 h, aged for 14 h, dechorionated, devitellinized, and exposed
to 10 6 M 5-HT in serum-free media until hatching. Control embryos were processed in parallel but not exposed to 5-HT. a Treated embryos displayed a significant increase in feeding rate relative
Summary
***
***
n.s.
n.s.
***
***
30
20
10
0
Fig. 11. Exposure of embryos lacking DTRH function to 5-HT for

the last 6 h of embryonic development decreases the complexity
of the serotonergic gut fibers projecting to the proventriculus.
pBacTRH embryos were collected for 2 h, aged for 14 h, dechorionated, devitellinized, and exposed to 10 6 M 5-HT in serum-free
media until hatching. Control embryos were processed in parallel
but not exposed to 5-HT. a Treated embryos displayed decreased
branching of gut fibers relative to controls. The number of large

varicosities (b) and the average varicosity area (c) were also significantly less. d The total number of individual varicosities was
similar. There were 38 projections from 27 guts from 8 independent experiments for control, and 35 projections from 23 guts
from 9 independent experiments. ***p ! 0.001, Students t tests.
Lines above the graph depict standard error of the mean.

231
the parental w1118 (fig.10a), and the hypersensitivity to

exogenous 5-HTP was also rescued (fig.10b). Consistent
with the changes in feeding behavior, branching of the
proventricular gut fibers was decreased (fig.11a), as was
the varicosity area (fig. 11b) and number of extremely
large varicosities (fig. 11c). The varicosity number was
unchanged (fig.11d), but since the larger varicosities appear to be composed of numerous small varicosities, a
decrease in the number of large varicosities, and in varicosity area, would suggest that the overall number of
varicosities has been reduced, as expected. These results
provide confirmation that it is the developmental actions
of serotonin that affect the architecture and function of
the 5-HT feeding circuit in the DTRH knockdown and
mutant animals.
Discussion
Neurotransmitters can function as growth factors in

developing neuronal and nonneuronal tissues before
adopting their roles as signaling molecules in the mature
nervous system, since stimulation of their receptors, also
present, can activate downstream intracellular signaling
cascades. Thus, the machinery is in place for neurotransmitters to exert pleiotropic effects in both development
and function of nervous tissue [reviewed in Weiss et al.,
1998]. 5-HT is present not only in the most primitive animal, the sponge [Salmoun et al., 2002], but it acts as a
trophic factor to stimulate the development of specialized
feeding cells in juvenile hydra [McCauley, 1997]. Here, we
present data to show that 5-HT acts as a trophic factor in
the development of a specific neural circuit in Drosophila,
as well as acting as a neurotransmitter to modulate the
behavior arising from the mature circuit.
Serotonergic Feeding Circuit in Drosophila
The cephalopharyngeal plates are strongly 5-HT-immunoreactive; a continuous serotonergic nerve fiber connects the plates to the brain and then the brain to the
proventriculus (the larval foregut) [Budnik et al., 1989].
The latter part of this nerve, the recurrens nerve, fasciculates in the proventriculus into individual fibers that are
dotted with serotonergic varicosities along the lengths;
additional fibers branch off when the fibers reach the
midgut. This defines the larval serotonergic feeding
circuit, which is an integral part of the stomatogastric
pattern generator. Although both glutaminergic and
FRMFamide fibers innervate the foregut, they are distinct from the 5-HT circuitry, and do not change their
232
architecture as a consequence of impaired 5-HT levels

[Budnik et al., 1989], although presumably all contribute
to the generation of proventricular muscular contractions.
The control condition used when exposing larvae to
exogenous 5-HTP differed from that used when assessing
basal feeding rate, although the same yeast source was
used. The minor differences in control conditions slightly affected responses when comparing the two different
conditions, demonstrating that even small changes in the
normal environment may contribute to the variability
observed in the behavioral outcomes. We have previously observed this for adult behavioral paradigms [Neckameyer and Matsuo, 2008], as have others, not only in Drosophila [e.g. Martin et al., 1999], but in mice as well
[Wahlsten et al., 2006]. Therefore, although we observed
the same changes in behaviors (i.e. decrease or increase
in feeding rate) in different experiments, the actual number of mouth hook contractions between control conditions may differ, depending upon the environment.
Serotonin modulates feeding behavior in all species
tested; rodent studies have revealed the involvement of
5-HT1 and 5-HT2 receptor agonists in promoting the hypophagic response to 5-HT [De Vry and Schreiber, 2000].
Consistent with these studies, we have shown that increased neuronal 5-HT levels in the mature larva decrease feeding rate, and reduced 5-HT levels within the
CNS increase feeding. However, since 5-HT is also known
to act as a trophic factor in the development of neural circuitry, and, given the extensive innervation of the mouthparts and gut by 5-HT, we asked whether the trophic effects of 5-HT could impair the development, and thus the
mature function, of this feeding circuit.
Serotonin Is Required during Development of the
Feeding Circuit for the Normal Architecture of 5-HT
Projections from the Brain to the Gut
Using transgenic, genetic and pharmacological tools
to constitutively titrate DTRH and 5-HT levels during
CNS development, we have shown that developmental
5-HT levels inversely correlate with the degree of arborization and of varicosity number and size along the length
of neurites projecting to the foregut from the CNS. Our
results are consistent with the seminal work of Kater and
colleagues, who demonstrated that application of 5-HT
to cultured Helisoma B19 buccal ganglion neurons repressed neurite outgrowth, with secondary effects on
synaptogenesis [Haydon et al., 1984; Goldberg and Kater,
1989; reviewed in Goldberg, 1998]. Similarly, treatment of
Helisoma embryos with a Tph inhibitor affected the deNeckameyer
velopment of early differentiating serotonergic neurons

(ENC1s) by increasing the arborization of a neurite projecting from these neurons to the foot primordium, which
occurred through the actions of 5-HT autoreceptors
[Diefenbach et al., 1995]. Later studies demonstrated that
5-HT induced a calcium influx-dependent inward sodium current via the actions of cyclic-nucleotide-gated
channels [reviewed in Goldberg, 1998]. 5-HT has been
shown to induce growth cone collapse and inhibit neurite
outgrowth in Helisoma by depolymerizing F-actin, a critical cytoskeletal component of growth cones [Torreano et
al., 2005]. While these studies demonstrated localized actions of 5-HT directly on growth cone activity in isolated
neurons, further studies have shown that 5-HT can have
more widespread effects on developing neuronal populations [Goldberg, 1998]. In addition, the Helisoma C1 cerebral ganglion, which extends through the buccal ganglia to target tissues required for the feeding response,
may release 5-HT nonsynaptically to bind to 5-HT receptors on the B19 neuron and elsewhere [Goldberg, 1998].
Thus, the actions of serotonin in the developing nervous
system may be specific for a given cell type and target tissue, and at a given developmental window. Development
of the serotonergic innervation of target tissues for feeding in Drosophila is similar to that of the snail in that innervation of the buccal mass occurs during late embryogenesis, with increasing numbers of 5-HT-immunoreactive fibers in muscle during the early juvenile stages
[Goldberg and Kater, 1989; Budnik et al., 1989; Balog and
Elekes, 2008]. In both systems, depletion of 5-HT during
this critical developmental time resulted in increased arborization of 5-HT fibers projecting to the target tissues.
However, these analyses were limited to components of
the feeding circuit; while gross development was unaffected, and simple behavior (locomotion) was normal in
our studies, other circuitry could have been differentially
affected.
Earlier studies suggested that 5-HT acts to inhibit neurite outgrowth in actively elongating fibers, but it can also
reinitiate growth in fibers whose outgrowth has been arrested; the former occurs in both embryonic and mature
neurons, but the latter only occurs in embryonic neurons
[Goldberg et al., 1991]. In addition, the local milieu must
be important, since individual neurites projecting from
the same neuron can respond differently to 5-HT exposure [Goldberg et al., 1991]. Depletion of serotonin in
the developing neocortex results in depressed dendritic
growth of cortical interneurons [Durig and Hornung,
2000], and 5-HT applications to moth antennal lobe neurons cultured in vitro increases branching of these cells
[Kim et al., 2005], the converse of what we have observed

for serotonergic fibers projecting from the Drosophila
brain to the foregut. Thus, the actions of 5-HT during the
developing CNS can be both inhibitory and stimulatory.
While the actions of 5-HT induce different signaling
pathways in these developing circuits, so that the effect of
5-HT application or reduction may have opposing consequences for the neuronal architecture, 5-HT still acts as
a trophic factor to initiate these changes.
5-HT also acts to spatially regulate varicosity formation along the neurite length. The diameter of normal
serotonergic varicosities in the neuropil of the Drosophila
distal abdominal ganglia are approximately 0.51.0 !m
in size [Sykes and Condron, 2005], comparable to the majority of varicosities observed in the serotonergic projections to the gut in our control animals. Application of
exogenous 5-HT to intact Drosophila larval ventral nerve
cords in culture resulted in a reversible loss of serotonergic varicosities in second and third instar larvae [Sykes
and Condron, 2005], consistent with our observations
that reduced 5-HT levels increase varicosity size and
number and increased 5-HT levels reduce them. Since we
did not observe any change in varicosity number or size
when comparing control and induced GSelav TRH RNAi
animals (fig.6), the effect of reduced 5-HT levels on varicosity density in animals with constitutively reduced
neuronal 5-HT levels must have occurred prior to second
instar (2 days after hatching). Assessment of reduced
5-HT synthesis during more narrowly defined developmental windows will also clarify whether this effect represents an ongoing process or whether there is a more
limited developmental window.
Peripheral fibers begin to project from serotonergic
neurons 24 h before hatching, and the fibers projecting
to the gut are not observed for another 2 h [Budnik et al.,
1989]. In addition, these fibers must grow with the enlarging proventriculus, which increases in size throughout the larval molts. Since knockdown (using RNAi) and
knockout (via the pBacTRH mutation) of DTRH occurred across this developmental window, and since embryonic exposure to exogenous 5-HT also occurred over
this same time span, it is also possible the changes in arborization reflected temporal sensitivity to 5-HT levels.
It is conceivable that perturbations in regional 5-HT concentrations at specific times direct growth of the fasciculating fibers to limit arborization until they enter the
midgut, which would occur shortly before hatching.
Expression levels of the serotonin transporter might
be upregulated to counteract limited neuronal 5-HT synthesis by having increased reuptake from the periphery.

233
5-HT transporter mRNAs are first observed in stage 15

embryos, before 5-HT can be detected in serotonergic
neurons [Demchyshyn et al., 1994]. However, in the
DTRH null mutant pBacTRH, 5-HT levels in the brain
are barely detectable [Neckameyer et al., 2007]; since all
neuronal 5-HT must occur via reuptake from circulating
5-HT, it is therefore unlikely 5-HT transporter expression is significantly enhanced. While DTRH levels can be
upregulated during early CNS development, this does not
result in increased levels of 5-HT. While it is also possible
that expression of the 5-HT transporter is downregulated, it is not clear that such downregulation would have
any significant effect on neuronal 5-HT levels, since
neuronal and peripheral 5-HT synthesis are distinct
pathways [Neckameyer et al., 2007]. In addition, when
exposure to enhanced 5-HT is limited to late embryogenesis (fig. 9), the larvae still display aberrant feeding
and neurite architecture of fibers projecting to the foregut, suggesting that the last 6 h of embryogenesis provide
a critical developmental window; similar treatment of
pBacTRH embryos rescues both the behavioral and gut
architecture abnormalities (fig.10, 11).
It is possible that constitutive knockdown of DTRH
synthesis could affect expression of the any of the four
serotonin receptors to compensate for limited 5-HT levels, since all the receptors are expressed in the CNS during embryonic stage 1617 [Saudou et al., 1992; Colas et
al., 1995]. However, it is not yet known which of these receptors are involved in the development of the serotonergic innervation of the foregut. Since 5-HT signaling must
be propagated via the actions of 5-HT receptors, it is conceivable that, depending upon the temporal expression of
a given receptor, the actions of 5-HT may act to both induce and repress neurite extension and arborization.
Changes in Gut Fiber Architecture Correlate with
Impairments in Feeding Behavior, the Functional
Output of the Mature Circuit
The architectural changes in this 5-HT circuit as a consequence of changes in 5-HT levels during development
also correlate with impaired feeding response in the mature larva. The precise causative mechanisms for the behavioral aberrations remain to be elucidated. While reduced developmental 5-HT results in an increase in the
number of varicosities, and the formation of large vesicles
that likely arise from the fusion of numerous smaller vesicles, their contribution to the behavioral phenotype is
unknown. The small vesicles along the neurite length are
likely to be synaptic varicosities that contain release sites
for serotonin. However, increased varicosity number and
234
size correlates with decreased DTRH levels, and thus

5-HT synthesis. It is possible that a stimulus resulting in
release of 5-HT from these vesicles can surpass the normal
concentration threshold and functionally affect downstream signaling pathways. The increased arborization
and abnormal neurite structures may also result in local
release of 5-HT that surpasses a threshold. Manipulation
of Helisoma embryos to transiently depress extracellular
5-HT levels increased synaptic connectivity; enhanced
dye coupling between neurons was observed, which correlated with the increase in neurite outgrowth [Goldberg
and Kater, 1989]. Both large and small 5-HT varicosities
have been identified along axonal processes in rat raphe
5-HT neurons; the larger varicosities comprise less than
10% of the total vesicle population, and may play a different functional role [Benzekhroufa et al., 2009]. These may
be analogous to the fine 5-HT fibers, with small varicosities, and beaded 5-HT fibers, with larger varicosities,
found in the retrohippocampal areas in the rabbit; these
fibers are spatial distinct and are thus also likely to serve
different functions [Bjarkam et al., 2005].
While feeding in animals with constitutively reduced
neuronal 5-HT is depressed, their response to exogenous
5-HTP as mature larvae suggests a hypersensitivity to its
neurotransmitter actions. Conversely, embryos exposed
to 5-HTP during development of the circuitry are hyposensitive to 5-HTP when mature. Therefore, sensitivity of
the feeding circuit to the neurotransmitter actions of
5-HT also appears to correlate with the degree of arborization and varicosity size and number.
It is also possible that perturbations in developing
5-HT levels may affect either the frontal or recurrens
nerves; however, these nerve fibers are highly 5-HT immunoreactive and tightly aligned, making quantitation
of changes in the fibers difficult to assess. Similarly, the
cephalopharnygeal plates are intensely 5-HT immunoreactive (see fig.1). When placed on an agar substrate, larvae will insert their mouth hooks into the agar with each
forward muscular contraction, and as the animal moves,
the tracks left by the mouth hook indentations are clearly
visible. Thus, locomotor behavior on an agar substrate
not only acts as a control to demonstrate that the animals
are not generally physiologically impaired, but damage to
the frontal nerve and cephalopharyngeal plates as a consequence of perturbed developmental 5-HT levels should
have affected this aspect of locomotor behavior, which it
did not. Constitutive reduction in 5-HT synthesis had no
effect on locomotion, neither did temporal up- or downregulation of 5-HT levels, or exposure of embryos to exogenous 5-HT during the last 6 h of embryogenesis. Since
Neckameyer
the recurrens nerve is merely an extension of the frontal

nerve, it is reasonable to assume that any alterations in
the frontal nerve would also occur in the recurrens nerve.
Thus, while we cannot rule out that the frontal nerve, recurrens nerve, or the cephalopharyngeal plates do not
have an effect on the changes in feeding, since serotonin
does not appear to alter mouth hook contractions except
when the animal is placed in a food source, it is reasonable to assume that any changes as a consequence of perturbations in developmental neuronal 5-HT levels are
likely limited to the proventricular and midgut tissues.
In summary, neuronal developmental 5-HT levels inversely correlate with the amount of branching, and the
number and size of varicosities of 5-HT fibers projecting
from the brain to the gut. Developmental neuronal 5-HT
levels directly correlate with changes in feeding behavior
in mature larvae and an inappropriate feeding response to
exogenous 5-HT. These data demonstrate that 5-HT acts
as a trophic factor in CNS development, with direct effects
on neuronal architecture, and impairment of the behavior
modulated by that neuronal circuitry. Many converging
lines of evidence now suggest that abnormal levels of 5-HT
during CNS development result in permanent alterations

in neuronal circuitry and may explain the antidepressant
actions of selective serotonin reuptake inhibitors in neurogenically active regions of the adult brain. Inhibition of
5-HT synthesis increases the rate of proliferation of serotonergic neurons in the dentate gyrus [reviewed in Djavadian, 2004]; chronic fluoxetine treatment enhances dendritic arborization of newborn hippocampal granule cells
[Wang et al., 2008] and increases hippocampal pyramidal
neuron dendritic spine density [Hajszan et al., 2005].
Thus, understanding the trophic actions of serotonin on
developing neurons in the immature and mature CNS will
provide critical insight into not only a fundamental question in developmental neuroscience, but into the therapeutic actions of antidepressant treatments.
Acknowledgements
This work has been supported by NSF Grant No. 0616062 to
W.S.N. We would like to acknowledge Dr. W. Michael Panneton
for the generous use of microscopes and analysis programs, and
Parag Bhatt for technical assistance.
References
Anderson G, Horne, W, Chatterjee D, Cohen D
(1990): The hyperserotonemia of autism.
Ann N Y Acad Sci 600:331340.
Balog G, Elekes K (2008): Functional neuroanatomy of the 5-HTergic system in the developing and adult buccal complex of the pond
snail, Lymnea stagnalis. Acta Biol Hung
59(suppl):5559.
Beique J, Chapin-Penick E, Mladenovic L, Andrade R (2004): Serotonergic facilitation of
synaptic activity in the developing rat prefrontal cortex. J Physiol 556:739754.
Benzekhroufa K, Liu B, Tang F, Teschemacher A,
Kasparov S (2009): Adenoviral vectors for
highly selective gene expression in central
serotonergic neurons reveal quantal characteristics of serotonin release in the rat brain.
BMC Biotechnol 9: 23.
Bjarkam C, Sorensen J, Geneser F (2005): Distribution and morphology of serotonin-immunoreactive axons in the retrohippocampal
areas of the New Zealand white rabbit. Anat
Embryol 210: 199207.
Blundell J (1986): Serotonin manipulations and
the structure of feeding behaviour. Appetite
7(suppl):3956.
Budnik V, Wu C, White K (1989): Altered
branching of serotonin-containing neurons
in Drosophila mutants unable to synthesize
serotonin and dopamine. J Neurosci 9:2866
2877.

Buznikov G, Lambert H, Lauder J (2001): Serotonin and serotonin-like substances as regulators of early embryogenesis and morphogenesis. Cell Tissue Res 305:177186.
Chen J, Condron B (2008): Branch architecture
of the fly larval abdominal serotonergic neurons. Dev Biol 320:3038.
Chugani D (2002): Role of altered brain serotonin mechanisms in autism. Mol Psychiatry
7:S16S17.
Colas J, Launay J, Kellermann O, Rosay P, Maroteaux L (1995): Drosophila 5-HT2 serotonin receptor: coexpression with fushitarazu during segmentation. Proc Natl Acad
Sci USA 92:54415445.
Coleman C, Neckameyer W (2004): Substrate
regulation of serotonin and dopamine synthesis in Drosophila. Invert Neurosci 5: 85
96.
Coleman C, Neckameyer W (2005): Serotonin
synthesis by two distinct enzymes in Drosophila melanogaster. Arch Insect Biochem
Physiol 59:1231.
Cowan D (2007): Serotonin and neuronal growth
factors a convergence of signaling pathways. J Neurochem 101:11611171.
Dacks A, Nickel T, Mitchell B (2003): An examination of serotonin and feeding in the flesh
fly Neobellieria bullata (Sarcophagidae: Diptera). J Insect Behavior 16:121.
Dean B (2003): The cortical serotonin2A receptor and the pathology of schizophrenia: a
likely accomplice. J Neurochem 85:113.
Demchyshyn L, Pristupa Z, Sugamori K, Barker
E, Blakely R, Wolfgang W, Forte M, Niznik
H (1994): Cloning, expression, and localization of a chloride-facilitated, cocaine-sensitive serotonin transporter from Drosophila
melanogaster. Proc Natl Acad Sci USA 91:
51585162.
De Miranda J, Eggleston P (1988): Larva competition in Drosophila melanogaster. I. Estimation of larval growth parameters. Heredity
60:205212.
De Vry J, Schreiber, R (2000): Effects of selected
serotonin 5-HT1 and 5-HT2 receptor agonists on feeding behavior: possible mechanisms of action. Neurosci Biobehav Rev 24:
341353.
Diefenbach T, Sloley B, Goldberg J (1995): Neurite branch development of an identified serotonergic neuron from embryonic Helisoma: evidence for autoregulation by serotonin.
Dev Biol 167:282293.
Djavadian R (2004): Serotonin and neurogenesis
in the hippocampal dentate gyrus of adult
mammals. Acta Neurobiol Exp 64:189100.
Durig J, Hornung J (2000): Neonatal serotonin
depletion affects developing and mature
mouse cortical neurons. Dev Neurosci 11:
833837.
235
Fricker A, Rios C, Devi L, Gomes I (2005): Serotonin receptor activation leads to neurite
outgrowth and neuronal survival. Mol Brain
Res 138:228235.
Gaspar P, Cases O, Maroteaux L (2003): The developmental role of serotonin: news from
mouse molecular genetics. Nat Rev Neurosci
4:10021012.
Giordano E, Rendina R, Pelus I, Furia M (2002):
RNAi triggered by symmetrically transcribed transgenes in Drosophila melanogaster. Genetics 160:637648.
Goldberg J (1998): Serotonin regulation of neurite outgrowth in identified neurons from
mature and embryonic Helisoma trivolvis.
Perspect Dev Neurobiol 5: 373387.
Goldberg J, Kater S (1989): Expression and function of the neurotransmitter serotonin during development of the Helisoma nervous
system. Dev Biol 131:483495.
Goldberg J, Mills L, Kater S (1991): Novel effects
of serotonin on neurite outgrowth in neurons cultured from embryos of Helisoma
trivolvis. J Neurobiol 22:182194.
Gross C, Zhuang X, Stark K, Ramboz S, Ossting
R, Kirby L, Santarelli L, Beck S, Hen R (2002):
Serotonin 1A receptor acts during development to establish normal anxiety-like behavior in the adult. Nature 416:396400.
Hajszan T, MacLusky N, Leranth C (2005):
Short-term treatment with the antidepressant fluoxetine triggers pyramidal dendritic
spine synapse formation in rat hippocampus. Eur J Neurosci 21:12991303.
Hasler G, Drevets W, Manji H, Charney D
(2004): Discovering endophenotypes for major depression. Neuropsychopharmacol 29:
17651781.
Haydon P, McCobb P, Kater S (1984): Serotonin
selectively inhibits growth cone motility and
synaptogenesis of specific identified neurons. Science 226:561564.
Haydon P, McCobb P, Kater S (1987): The regulation of neurite outgrowth, growth cone motility, and electrical synaptogenesis by serotonin. J Neurobiol 18:197215.
Herlenius E, Lagercrantz H (2001): Neurotransmitters and neuromodulators during early
human development. Early Human Dev 65:
2137.
Janusonis S, Gluncic V, Rakic P (2004): Early serotonergic projections to Cajal-Retzius cells:
relevance for cortical development. J Neurosci 24:16521659.
Joshi A, Mueller L (1988): Evolution of higher
feeding rate in Drosophila due to densitydependent natural selection. Evolution 42:
10901093.
Kaplan D, Zimmerman G, Suyama K, Meyer
T, Scott M (2008): A nucleostemin family
GTPase, NS3, acts in serotonergic neurons to
regulate insulin signaling and control body
size. Genes Dev 22:18771893.
236
Kim J, Sung D, Park C, Park H, Park C, Jeon S,

Kang P, Kwon O, Lee B (2005): Brain-derived
neurotrophic factor promotes neurite
growth and survival of antennal lobe neurons in brain from the silk moth, Bombyx
mori in vitro. Zool Sci 22:333342.
Lauder J, Krebs H (1978): Serotonin as a differentiation signal in early neurogenesis. Dev
Neurosci 1:1530.
Lauder J, Wallace J, Krebs H, Petrusz P, McCarthy K (1982): In vivo and in vitro development of serotonergic neurons. Brain Res Bull
9:605625.
Lin D, Goodman C (1994): Ectopic and increased
expression of fasiculin II alters motoneuron
growth cone guidance. Neuron 13:507523.
Lucki I (1992): 5-HT1 receptors and behavior.
Neurosci Behav Res 16:8393.
Luo X, Persico A, Lauder J (2003): Serotonergic
regulation of somatosensory cortical development: lessons from genetic mouse models.
Dev Neurosci 25:173183.
Mackenzie S, Brooker M, Gill T, Cox G, Howells
A, Ewart G (1999): Mutations in the white gene
of Drosophila melanogaster affecting ABC
transporters that determine eye colouration.
Biochim Biophys Acta 1419:173185.
Martin J, Ernst R, Heisenberg M (1999): Temporal pattern of locomotor activity in Drosophila melanogaster. J Comp Physiol A 184: 73
84.
McCauley D (1997): Serotonin plays an early
role in the metamorphosis of the hydrozoan
Phialidium gregarium. Dev Biol 190: 229
240.
Mercer A, Hayashi J, Hildebrand J (1995): Modulatory effects of 5-HT on voltage-activated
currents in cultured antennal lobe neurons
of the sphinx moth Manduca sexta. J Exp
Biol 198:613627.
Mercer A, Kirchoff B, Hildebrand J (1996): Enhancement of serotonin of the growth in vitro of antennal lobe neurons of the sphinx
moth Manduca sexta. J Neurobiol 29:4964.
Mercer A, Kloppenburg P, Hildebrand J (1999):
Serotonin-induced changes in the excitability of cultured antennal-lobe neurons of the
sphinx moth Manduca sexta. J Comp Physiol
A 178:2131.
Neckameyer W (1996): Multiple roles for dopamine in Drosophila development. Dev Biol
176:209219.
Neckameyer W, Coleman C, Eadie S, Goodwin S
(2007): Compartmentalization of neuronal
and peripheral serotonin synthesis in Drosophila melanogaster. Genes Brain Behav 6:
256269.
Neckameyer W, Matsuo H (2008): Distinct neural circuits reflect sex, sexual maturity, and
reproductive status in response to stress in
Drosophila melanogaster. Neuroscience 156:
841856.
Neckameyer W, White K (1992): A single locus
encodes both phenylalanine hydroxylase
and tryptophan hydroxylase activities in
Drosophila. J Biol Chem 267: 199206.
Novak M, Ribeiro J, Hildebrand J (1995): 5-HT

in the salivary glands of adult female Aedes
aegypti and its role in the regulation of salivation. J Exp Biol 198: 167174.
Novak M, Rowley W (1994): Serotonin depletion
affects blood-feeding but not host-seeking
ability in Aedes triseriatus (Diptera: Culicidae). J Med Entomol 31:600606.
Oland L, Kirschenbaum S, Pott W, Mercer A,
Tolbert L (1995): Development of an identified serotonergic neuron in the antennal lobe
of the moth and effects of reduction in serotonin during construction of olfactory glomeruli. J Neurobiol 28:248267.
Osterwalder T, Yoon K, White B, Keshishian H
(2001): A conditional tissue-specific transgene expression system using inducible
GAL4. Proc Natl Acad Sci USA 98: 12596
12601.
Pardo C, Eberhart C (2007): The neurobiology of
autism. Brain Pathol 17:434447.
Robertson H, Preston C, Phillis R, JohnsonSchlitz D, Benz W, Engels W (1988): A stable
genomic source of P element transposase in
Drosophila melanogaster. Genetics 118: 461
470.
Robinow S, White K (1988): The locus elav of
Drosophila melanogaster is expressed in neurons at all developmental stages. Dev Biol
126:294303.
Roman G, Endo K, Zong L, Davis R (2001):
P[Switch], a system for spatial and temporal
control of gene expression in Drosophila
melanogaster. Proc Natl Acad Sci USA 98:
1260212607.
Salmoun M, Devijver C, Daloze D, Braekman J,
van Soest R (2002): 5-Hydroxytryptaminederived alkaloids from two marine sponges
of the genus Hyrtios. J Nat Prod 65: 1173
1176.
Saudou F, Boschert U, Amlaiky N, Plassat J, Hen
R (1992): A family of Drosophila serotonin
receptors with distinct intracellular signaling properties and expression patterns.
EMBO J 11:717.
Sewall D, Burnet B, Connolly K (1975): Genetic
analysis of larval feeding behavior in Drosophila melanogaster. Genet Res (Cam) 24:
163173.
Shemer A, Azmitia E, Whitaker-Azmitia P
(1991): Dose-related effects of prenatal 5-methoxytryptamine (5-MT) on development of
serotonin terminal density and behavior.
Dev Brain Res 59:5963.
Sodhi M, Sanders-Bush E (2004): Serotonin and
brain development. Int Rev Neurobiol 59:
111174.
Sykes P, Condron B (2005): Development and
sensitivity to serotonin of Drosophila varicosities in the central nervous system. Dev
Biol 286:207216.
Neckameyer
Thibault S, Singer M, Miyazaki W, Milash B,

Dompe N, Singh C, Buchholz R, Demsky M,
Fawcett R, Francis-Lang H, Ryner L, Cheung
L, Chong A, Erickson C, Fisher W, Greer K,
Hartouni S, Howie E, Jakkula L, Joo D, Killpack K, Laufer A, Mazzotta J, Smith R, Stevens L, Stuber C, Tan L, Ventura R, Woo A,
Zakrajsek I, Zhao L, Chen F, Swimmer C,
Kopczynski C, Duyk G, Winberg M, Margolis J (2004): A complementary transposon
tool kit for Drosophila melanogaster using P
and piggyBac. Nat Genet 36:283287.
Torreano P, Waterman-Storer C, Cohan C
(2005): The effects of collapsing factors on
F-actin content and microtubule distribution of Helisoma growth cones. Cell Motil
Cytoskeleton 60:166179.

Valles A, White W (1988): Serotonin-containing

neurons in Drosophila melanogaster : development and distribution. J Comp Neurol
268:414428.
Wahlsten D, Bachmanov A, Finn D, Crabbe J
(2006): Stability of inbred mouse strain differences in behavior and brain size between
laboratories and across decades. Proc Natl
Acad Sci USA 103:1636416369.
Walther D, Peter J, Bashammakh S, Hortnagl H,
Voits M, Fink H, Bader M (2003): Synthesis
of serotonin by a second tryptophan hydroxylase isoform. Science 299:76.
Wang J, David D, Monckton J, Battaglia F, Hen
R (2008): Chronic fluoxetine stimulates maturation and synaptic plasticity of adult-born
hippocampal granule cells. J Neurosci 28:
13741384.
Weiss E, Maness P, Lauder J (1998): Why do neurotransmitters act like growth factors? Perspect Dev Neurobiol 5: 323335.
Whittle N, Sartori S, Dierssen M, Lubee G,
Singewald N (2007): Fetal Down syndrome
brains exhibit aberrant levels of neurotransmitters critical for normal brain development. Pediatrics 120:14651471.
Yao K, White W (1994): Neural specificity of elav
expression: defining a Drosophila promoter
for directing expression to the nervous system. J Neurochem 63:4151.
Zhang X, Beaulieu J, Sotnikova T, Gainetdinov
R, Caron M (2004): Tryptophan hydroxylase-2 controls brain serotonin synthesis.
Science 305: 217.
237

A Trophic Role For Serotonin in The Development of A Simple Feeding Circuit

Uploaded by

Copyright:

Available Formats

You might also like

A Trophic Role For Serotonin in The Development of A Simple Feeding Circuit

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Trophic Role For Serotonin in The Development of A Simple Feeding Circuit

Uploaded by

Copyright:

Available Formats

Original Paper

Dev Neurosci 2010;32:217237

Received: October 14, 2009