Content uploaded by John M Hawdon
Author content
All content in this area was uploaded by John M Hawdon
Content may be subject to copyright.
A common muscarinic pathway for diapause recovery
in the distantly related nematode species
Caenorhabditis elegans
and
Ancylostoma caninum
Heidi A. Tissenbaum*
†
, John Hawdon
‡
, Melissa Perregaux
‡
, Peter Hotez
‡
, Leonard Guarente*, and Gary Ruvkun
†§
*Department of Biology, Massachusetts Institute of Technology, Building 68-289, 77 Massachusetts Avenue, Cambridge, MA 02139; ‡Medical Helminthology
Laboratory, Yale University, 501 LEPH, 60 College Street, New Haven, CT 06520; and †Department of Molecular Biology, Massachusetts General Hospital and
Department of Genetics, Harvard Medical School, 50 Blossom Street, Boston, MA, 02144
Edited by William B. Wood, University of Colorado, Boulder, CO, and approved November 9, 1999 (received for review October 29, 1997)
Converging TGF-
b
and insulin-like neuroendocrine signaling path-
ways regulate whether Caenorhabditis elegans develops repro-
ductively or arrests at the dauer larval stage. We examined
whether neurotransmitters act in the dauer entry or recovery
pathways. Muscarinic agonists promote recovery from dauer arrest
induced by pheromone as well as by mutations in the TGF-
b
pathway. Dauer recovery in these animals is inhibited by the
muscarinic antagonist atropine. Muscarinic agonists do not induce
dauer recovery of either daf-2 or age-1 mutant animals, which
have defects in the insulin-like signaling pathway. These data
suggest that a metabotropic acetylcholine signaling pathway ac-
tivates an insulin-like signal during C. elegans dauer recovery.
Analogous and perhaps homologous cholinergic regulation of
mammalian insulin release by the autonomic nervous system has
been noted. In the parasitic nematode Ancylostoma caninum, the
dauer larval stage is the infective stage, and recovery to the
reproductive stage normally is induced by host factors. Muscarinic
agonists also induce and atropine potently inhibits in vitro recov-
ery of A. caninum dauer arrest. We suggest that host or parasite
insulin-like signals may regulate recovery of A. caninum and could
be potential targets for antihelminthic drugs.
Animals have developed many adaptations that allow them to
survive sporadic or seasonal declines in growth conditions.
Diapause, an alternative developmental state of low metabolic
activity, is an example of such an adaptation that is found in a
wide range of species (1). In the nematode, Caenorhabditis
elegans, the dauer diapause stage is induced by unfavorable
environmental conditions (2). Under normal growth conditions,
C. elegans develops through four larval stages (L1–L4) to the
reproductive adult stage. However, in response to a secreted
pheromone, low food levels, and high temperature (i.e., unfa-
vorable growth conditions) during the L1 lar val stage, animals
enter an alternative developmental pathway and arrest at the
dauer larval L3 stage. The C. elegans dauer larva has morpho-
logical, physiological, and anatomical changes that differ from a
reproductive L3 stage animal including: altered neuroanatomy,
radial shrinkage of a thick cuticle, sealing of mouth and anal
openings, suspension of feeding, and changes in metabolism (2).
Animals can remain arrested as dauer larvae for many months.
When pheromone levels and temperature decline, and food
levels increase, the animals molt and reenter the life cycle as an
L4 larva that is indistinguishable from animals that have not
arrested at the dauer stage (2). In parasitic nematodes, such as
the hookworm, Ancylostoma caninum, diapause arrest is an
obligatory part of the life history. The dauer stage in parasitic
nematodes is the infective stage that is morphologically, behav-
iorally, and functionally analogous to the C. elegans dauer (3–5).
Recovery to the reproductive stages in the parasitic nematodes
is induced upon host infection by unknown host factors.
Mutants that affect C. elegans dauer formation (daf) fall into
two categories: dauer constitutive and dauer defective. Dauer
constitutive mutations cause animals to enter the dauer stage
even under favorable growth conditions (low pheromone, high
food, and low temperature). Conversely, mutants that are dauer
defective do not arrest at the dauer stage even under conditions
of high pheromone (2). The mutations affecting dauer arrest
have been ordered into a genetic epistasis pathway (2). The
pathway represents the steps in a neuroendocrine signaling
system from sensory detection of food, pheromone, and tem-
perature to overall remodeling of the animal to form a full dauer.
The initiation and recovery of dauer arrest is regulated by
exposed sensory neurons. First, several mutants that do not
arrest as dauer larvae in response to the dauer pheromone have
structural abnormalities in the sensory amphid sensillum (2).
Second, killing particular sensory neurons with a laser mi-
crobeam induces dauer arrest, suggesting these neurons nor-
mally signal to induce reproductive development (2).
Molecular genetic analysis has identified a TGF-
b
pathway (2,
6) and an insulin-like signaling pathway (7) thought to couple the
sensory neural inputs to other secretory cell or to the target
tissues that are remodeled and metabolically shifted in the dauer
larvae (2, 7). DAF-7, the TGF-
b
ligand, is expressed in a pair of
exposed sensory neurons in reproductively growing animals but
not expressed in animals under conditions of high pheromone
(2). DAF-4, one of the TGF-
b
receptors, and DAF-3, a Smad
protein implicated in coupling TGF-
b
signals to the nucleus, are
expressed broadly in secretory as well as target tissues (6).
Predicted insulin-like ligands of the DAF-2 insulin receptor have
been identified in the genome sequence but not yet shown to
engage DAF-2 or regulate dauer arrest (8).
Temperature also modulates dauer arrest because even null
mutations in the TGF-
b
pathway genes are temperature-
sensitive and most daf-2 alleles are temperature-sensitive (2, 9,
10). The temperature input to this neuroendocrine pathway is
mediated at least partially by the thermoregulatory AIY and
AIZ interneurons (11, 12), but how the activity of these neurons
is coupled to neuroendocrine outputs is unknown.
We reasoned that the neural pathways from the dauer regu-
latory sensory neurons and interneurons to neurosecretory cells
that signal target tissues are likely to utilize known neurotrans-
mitters. Glutamate, acetylcholine,
g
-aminobutyric acid, seroto-
nin, FMRFamide, and dopamine have been implicated in par-
ticular C. elegans behaviors (13, 14), and drugs that interact with
the mammalian receptors of these neurotransmitters affect
particular C. elegans behaviors (13, 14). Moreover, the C. elegans
genome sequence has identified members of the receptor su-
perfamilies for most of these neurotransmitters (15).
This paper was submitted directly (Track II) to the PNAS office.
Abbreviation: GSM, S-methyl-glutathione.
§To whom reprint requests should be addressed. E-mail: Ruvkun@molbiol.mgh.harvard.
edu.
The publication costs of this article were defrayed in part by page charge payment. This
article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C.
§1734 solely to indicate this fact.
460–465
u
PNAS
u
January 4, 2000
u
vol. 97
u
no. 1
To identify neurotransmitter inputs to the dauer neuroendo-
crine pathways, we tested a variety of neurotransmitter agonists
and antagonists for induction of dauer arrest or recover y. We
find that muscarinic agonists specifically promote dauer recov-
ery in pheromone-induced dauer larvae as well as particular
classes of dauer constitutive mutants. The muscarinic agonists do
not induce recovery of either daf-2-orage-1-induced dauer
larvae, which have defective insulin-like signaling. We show that
this muscarinic pathway also regulates A. caninum recovery from
dauer arrest. In mammals, muscarinic agonists promote insulin
release both in vivo and in vitro (17, 18). We suggest that
insulin-like secretory cells in the nematodes are regulated by
cholinergic inputs in a metabolic control pathway homologous to
mammalian autonomic input to pancreatic beta cell activity.
Materials and Methods
Strains and Growth Conditions. Animals were grown on standard
NG agar plates. In this study, the mutations in C. elegans used
were: LGI, daf-8(e1393); LGII, daf-22(m130),unc-4(e120),sqt-
1(sc13),age-1(m333 and mg44); LGIII, daf-7(e1372),daf-
2(e1370, e1391), daf-4(m63); LGIV, daf-1(m40),daf-10(e1387);
LGX, daf-12(m20).age-1 animals were maintained as marked
heterozygous strains [sqt-1(sc13) age-1(mg44)ymnC1 or unc-
4(e120) age-1(m333)ymnC1]. A. caninum were maintained as
described previously (18).
Dauer Recovery Assay. Minimal media plates were used for the
drug assays: 3.0 g NaCl, 20 g agarose (Sigma Type II A6877),
0.025 M KPO
4
(pH 6.0), 1 mM CaCl
2
, 1 mM MgSO
4
, and 5
m
gyml
cholesterol were added. In some assays, Escherichia coli (DH5})
bacteria arrested with streptomycin was added to each plate. All
of the drugs used were from Sigma. Dauer-stage animals were
transferred to drug plates at 25° without the addition of food. For
experiments presented in Figs. 1 and 2 and Table 1, about 10,000
L1s were placed in 10 ml of S medium (19) containing 1–2 ml of
a 0.4% (wtyvol) solution of E. coli DH5}bacteria in M9 solution
(19) arrested with streptomycin in a 25-ml flask on a rotating,
heated water bath at 25°C. For wild-type dauer lar vae, 600
m
lof
the 0.4% (wtyvol) bacterial solution and 20–50
m
l of pheromone
also were added to the flask as described in ref. 20. After 72 hr,
the cultures were centrifuged and the supernatant was removed.
Animals were resuspended in a preheated 25° solution of 1%
SDS and placed on a rocker at 25° for 30 min, and the SDS was
removed. Animals were washed with either M9 or S medium (19)
four to six times. After a final spin, 100–200 dauer larvae were
placed onto the drug plates without food and scored 24 and 48
hr later for dauer larvae and non-dauer adults. Because muta-
tions in age-1 are maternally rescued, age-1 dauer larvae were
isolated either by plating synchronized L1s or picking maternally
rescued unc-4 age-1 or sqt-1 age-1 non-dauer animals to plates at
25°C and, after 3 days, picking dauer larvae to the drug and
control plates. Consequently, the plates with age-1 had some
bacteria on them.
Drug Assay in
A. caninum
.Hookworm-infective L3 animals were
collected from 1- to 4-week-old coproculture by the Baermann
technique and decontaminated with 1% HCl in BU buffer (50
mM Na
2
PO
4
y22 mM KH
2
PO
4
y70 mM NaCl, pH 6.8) (5, 21) for
30 min at 22°C. Approximately 250 L3 animals were incubated
in each individual wells of a 96-well plate containing 0.1 ml of
RPMI 1640 tissue culture medium, supplemented with 0.25 mM
Hepes, pH 7.2y100 units/ml penicilliny100
m
g/ml streptomyciny
100
m
g/ml gentamyciny2.5
m
g/ml amphotericin B. The L3 ani-
mals were activated to resume development and feeding by
including 10% (volyvol) canine serum and 25 mM S-methyl-
glutathione (GSM; ref. 22) in the medium. Nonactivated L3
animals were incubated in RPMI 1640 alone. The agonists were
tested for activation by incubation with the L3 animals in the
absence of the normal stimulus (i.e., serum 1GSM), whereas
atropine was tested in the presence of either the normal stimulus
or the agonists. Animals were incubated at 37°C in 5% CO
2
for
24 hr. The percentage of feeding was determined by incubating
the animals with 2.5 mgyml FITC-BSA for 2–3 hr and counting
the number of L3 animals that ingested the labeled BSA by
microscopic examination under epif luorescent illumination (23).
Each treatment was done in triplicate, and each experiment was
repeated at least once.
Dauer Arrest Assay. Animals were grown at 15° and placed in a
bleach solution to isolate eggs. One hundred to 200 eggs or
synchronized L1 larvae then were added to plates containing a
Fig. 1. The effect of muscarinic agonists and an antagonist on dauer recovery
in C. elegans and A. caninum.(A) Oxotremorine, a synthetic muscarinic
agonist, promotes dauer recovery in both C. elegans and A. caninum. daf-
2(e1370) fails to recover at all drug concentrations. The scale for the drug
concentration for A. caninum is 103. All points are the average of two
experiments, where at least two plates were scored with the exception of 100
mM daf-7,N2,daf-2, where only one experiment was done. (B) Atropine can
specifically inhibit the muscarinic agonist-induced response. In C. elegans,at
1 mM oxotremorine, as the concentration of atropine, a muscarinic antago-
nist, is increased, dauer recovery is completely inhibited. In A. caninum L3
incubated with serum and GSM plus atropine (5 mM), dauer recovery was
inhibited by 99.5% (from 86.7% to 0.5%). Similarly, in A. caninum larvae, 0.5
mM arecoline and increasing amounts of atropine cause dauer recovery to be
completely inhibited. Concentrations of 1–5 mM of a drug are used routinely
in drug assays in C. elegans (13, 14, 24). The unusually high doses may be due
to a cuticle permeability barrier. Drugs in the following classes were tested on
daf-7(e1372) and daf-2(e1370) mutant strains for dauer recovery and on wild
type and the daf-22(m130) mutant for dauer induction and had no reproduc-
ible effect on dauer recovery or formation. (a) Noradrenalineyadrenaline:
cocaine*, imipramine*, reserpine*, tetrabenazine*, clonidine, isoproteronol,
epinephrine, propanolol, and metopolol; (b) serotonin: 5-hydroxytryptamine,
p-chlorophenylalanine, and spiperone; (c) opioids: hydromorphone, meperi-
dine, nalorphine, PCP*, and methaqualone; (d) dopamine: dopamine and
metoclopramide; (e)
g
-aminobutyric acid: muscimol, diazepam, pentobarbi-
tol, bicuculline, and picrotoxin; (f) glutamate: NMDA, kainate,
a
-amino-
adipate, glutamate diethyl ester, and quisqualate. *, Affects multiple signal-
ing pathways.
Tissenbaum et al. PNAS
u
January 4, 2000
u
vol. 97
u
no. 1
u
461
NEUROBIOLOGY
drug and food at 20°C. When the non-dauer larvae had reached
the gravid adult hermaphrodite stage and were beginning to lay
eggs, each plate was examined visually for the presence of dauer
larvae and non-dauer larvae. After this, animals were rinsed off
the plate into a plastic dish containing 1% SDS (dauer larvae are
the only larval stage resistant to this treatment). After 30 min,
dishes were examined for the presence of dauer larvae and
non-dauer larvae.
Results
Cholinergic Input to Dauer Recovery. We tested drugs that affect the
following mammalian neuronal pathways for effects on C.
elegans dauer induction and dauer recovery: adrenergicy
noradrenergic, serotonergic, cholinergic, glutaminergic, dopa-
minergic, GABAergic, and opioid. Most drugs tested did not
affect induction or recover y from dauer arrest (Fig. 1). However,
multiple, unrelated muscarinic agonists promote dauer recovery.
All four muscarinic agonists tested, carbachol, oxotremorine,
pilocarpine, and arecoline, promote recovery of dauer larvae
induced by mutation as well as by pheromone although ox-
otremorine was the most potent (Fig. 1 and Table 1). Muscarinic
agonists induce recovery of dauer larvae induced by defective
TGF-
b
signaling in the daf-7(e1372) mutant, with a defect in the
TGF-
b
ligand, as well as daf-8 and daf-14 mutants, which encode
members of the Smad family of signal transduction transcription
factors that act downstream of the daf-7 ligand (A. Estevez, M.
Sundermeyer, M. Estevez, K. V. King, and D. L. Riddle, personal
communication; T. Inoue, and J. H. Thomas, personal commu-
nication) (Table 1). The muscarinic agonists do not induce
recovery of daf-2 mutants, which have defects in the C. elegans
homologue of the mammalian insulin receptor gene, or in age-1
mutants, which have a defect in a PI-3-kinase that acts down-
stream of the daf-2 receptor (refs. 2 and 9; Fig. 1 and Table 1).
Thus, the muscarinic recovery pathway depends on insulin-like
signaling.
The infective ‘‘dauer’’ L3 of the hookworm A. caninum can be
stimulated to resume feeding and development in vitro by
incubation with canine serum and GSM, but not by tissue culture
medium alone (18). When A. caninum L3 were incubated with
either oxotremorine or arecoline without canine serum or GSM
in the tissue culture medium, 60–80% of the animals recovered,
as indicated by the resumption of feeding (Fig. 1A). Therefore,
muscarinic agonists mimicked the recovery induced by serum
and GSM.
Fig. 1Ashows the oxotremorine dose-response curve for
wild-type pheromone-induced dauer lar vae, daf-7(e1372), daf-
2(e1370), and A. caninum dauer larvae. Oxotremorine and
pilocarpine induce maximum recover y of daf-7(e1372) dauer
larvae at 5 mM concentration, whereas pheromone-induced
dauer larvae reach maximum recovery at 1 mM (Fig. 1Aand
data not shown). A. caninum L3 dauer larvae also reach maxi-
mum recovery at 5 mM oxotremorine (Fig. 1A), but fail to
recover when incubated with pilocarpine (data not shown).
Pilocarpine had the least effect on all the strains tested (Table
1). Additionally, it had no effect on A. caninum (data not shown).
The maximal response for arecoline is 10-fold lower than for the
other agonists in both C. elegans and A. caninum (Table 1 and
data not shown).
Atropine Specifically Inhibits Dauer Recovery. To determine the
specificity of the muscarinic response, we added both oxotremo-
rine and atropine, a muscarinic antagonist, varying the concen-
tration of antagonist (Fig. 1B). In 1 mM oxotremorine, 40% of
the daf-7(e1372) dauer larvae recover at 25°C, the nonpermissive
temperature for daf-7(e1372ts). However, in combination with 1
mM atropine, 1 mM oxotremorine induced only 5% recovery; at
5 mM atropine, the 1 mM oxotremorine response is completely
abolished. For wild-type pheromone-induced dauer lar vae, the
results are almost identical (Fig. 1B). This suggests that the
drug-induced recovery is a specific muscarinic response, because
in mammals, atropine affects muscarinic and not nicotinic
receptors (25, 26).
Atropine can partially inhibit C. elegans dauer recovery in-
duced by food signals. When bacteria are added to pheromone-
induced dauer larvae at 25°C, 91% of the animals recover,
compared with 9% recovery without food (Fig. 2A). Atropine
partially inhibits the food-induced recovery of dauer larvae,
from 91% without drug to 75% with drug. However, pheromone
reduced this response much more severely (from 91% to 20%).
Thus, food induces more than the muscarinic pathway to trigger
dauer recovery.
In A. caninum, recovery induced by serum and GSM was
inhibited nearly completely by atropine (5 mM) (Fig. 2 A; from
86.7% to 0.5%). Moreover, A. caninum L3 incubated with 0.5
mM arecoline and 1.0 mM atropine failed to recover. These data
indicate that the muscarinic pathway is a major recovery signal
from arrest in hookworm.
Temperature downshifts in the presence of food induce dauer
recovery in animals bearing mutations in the TGF-
b
- or insulin-
like-signaling pathways (2, 6, 9). Temperature downshift in the
absence of food partially induces dauer recovery in daf-2 mutants
and does not induce dauer recovery in daf-7 mutants (Fig. 2B).
Similarly, bacterial food at 25°C does not allow reproductive
development of either daf-2 or daf-7 mutants (ref. 2; data not
shown). However, temperature downshift in the presence of food
induces more than 40% recovery of both mutants (Fig. 2B). This
temperature shift plus food recovery in both daf-7 and daf-2
mutants is potently inhibited by atropine (Fig. 2B). Similar
inhibition of dauer recovery was observed with another daf-2
allele, daf-2(e1391) (Fig. 2B).
Table 1. Induction of dauer recovery by muscarinic agonists
Drug daf-7(e1372) daf-8(e1377) daf-14(m77) daf-2(e1370) age-1(m333) age-1(mg44)
No drugyno food 1.3 61.3 4.5 60.1 0 0 0 0
Pilocarpine 8.8 66.0 17.7 60.4 2.5 60.1 0 ND ND
Arecoline 34.3 624.4 34.0 60.7 0.2 6,0.1 0 0 0
Oxotremorine 41.2 624.5 36.2 60.3 14.3 6,0.1 0 0 0
Oxotremorine atropine 5.2 62.7 4.8 60.01 ,0.1 6,0.1 0 ND ND
Experiments were performed as in Materials and Methods. Animals were scored 24 –48 h after being placed as dauer larvae on plates containing drugs. Values
are mean 6SE. Numbers shown indicate the percent dauer recovery from two trials, where each plate was at least in duplicate for daf-7(e1372), daf-2(e1370),
and age-1(m333). The remaining strains show values from one experiment in which each drug was tested on at least two plates, although similar trends have
been observed previously. For each drug tested the number of animals are in the following order: no drug no food; 1 mM pilocarpine; 1 mM arecoline; 1 mM
oxotremorine; and 1 mM oxotremorine with 1 mM atropine. Number of animals tested are: daf-7(e1372), 2345, 692, 756, 1363, 1227; daf-8(e1377), 1582, 354,
652, 753, 925; daf-14(m77), 1300, 676, 698, 902, 1101; daf-2(e1370), 872, 222, 258, 207, 460; age-1(m333), 113, not done (ND), 74, 74, ND; age-1(mg44), 71, ND,
57, 52, ND.
462
u
www.pnas.org Tissenbaum et al.
Neurotransmitter Regulation of Dauer Arrest. We examined
whether exogenous application of neurotransmitters could
mimic the dauer pheromone to induce dauer arrest. Drugs that
affect mammalian neuronal pathways including adrenergicy
noradrenergic, serotonergic, cholinergic, glutaminergic, dopa-
minergic, GABAergic, and opioid were examined for effects on
C. elegans dauer induction (Table 1). We tested these drugs for
induction of dauer arrest in wild-type and daf-22 mutants, a
mutant that does not secrete pheromone, but arrests at the dauer
stage when exposed to exogenous pheromone (2). None of the
drugs tested induced dauer arrest under favorable growth con-
ditions. The drugs were active because several of the drugs
caused either paralysis, death, or egg-laying defects.
Discussion
Arrest at the dauer stage is a nematode survival strategy that is
a specific example of the phyletically general diapause arrest
(1–3). In C. elegans, dauer arrest occurs under harsh environ-
mental conditions whereas in the hookworm, A. caninum,a
parasitic nematode, diapause is a nonconditional stage in the life
cycle (3, 4). Dauer recovery is regulated by levels of pheromone,
food, and temperature in C. elegans, whereas in A caninum,
unknown host factors induce dauer recovery upon infection
(3–5).
We have shown that muscarinic agonists cause dauer recovery
in both C. elegans and A. caninum and that this recovery is
inhibited specifically by the muscarinic antagonist atropine (Fig.
1). The endogenous neurotransmitter at muscarinic receptors is
acetylcholine, which, in vertebrates, functions at cholinergic
synapses in both the peripheral and central nervous system (26).
Acetylcholine has a wide variety of functions in vertebrate
signaling including sympathetic and parasympathetic ganglion
cells as well as the adrenal medulla, synapses within the central
nervous system, and motor end plates on skeletal muscle inner-
vated by somatic motoneurons (26). Muscarinic receptors are
found in muscle, the autonomic ganglia, the central nervous
system, and secretory glands. These receptors couple to G
proteins and signal on longer time scales than nicotinic recep-
tors. Both muscarinic and nicotinic receptors also have been
found in invertebrates such as Drosophila and C. elegans (13, 14,
26–28).
The nicotinic receptor has been the primary focus of the
studies on cholinergic signaling in the worm. The drug levami-
sole, a nicotinic agonist, is toxic to C. elegans, causing muscle
hypercontraction (13). Mutants that are resistant to this drug
have revealed components of a nicotinic-signaling cascade (13).
Levamisole has no effect on dauer recovery (data not shown),
suggesting that the nicotinic receptor pathway does not regulate
dauer arrest.
Fewer studies have been done on muscarinic signaling in C.
elegans. Binding studies on crude homogenates of C. elegans have
shown that they contain muscarinic receptors that have the
potential to bind to the muscarinic ligands [
3
H]QNB (29) and
[
3
H]N-methylscopalamine (30) with high affinity (13). These
receptors were found in both C. elegans adults and L1 and L2
larvae (29). Several muscarinic receptor homologues have been
identified in the C. elegans genome sequence database (ref. 28;
H.A.T., G. Sandoval, and G.R., unpublished observation).
Both arecoline and pilocarpine are naturally occurring drugs
from the betel nut seed and the Pilocarpus leaf, respectively,
whereas oxotremorine and carbachol are synthetic drugs (26).
Arecoline, pilocarpine, and oxotremorine have the same sites of
action on mammalian metabotropic cholinergic receptors, but
arecoline also acts on nicotinic receptors (26). Atropine, a
natural product, specifically inhibits mammalian muscarinic
responses (26). Because all of the drug-induced dauer recovery
was inhibited by atropine, and this drug does not inhibit nicotinic
signaling in C. elegans, we conclude that dauer recovery is
mediated by muscarinic signaling.
Muscarinic agonists potently induced recovery of dauer larvae
induced by pheromone or mutations in the daf-7 TGF-
b
group
of genes, but did not induce recovery of the daf-2 and age-1
Fig. 2. Atropine specifically inhibits dauer recovery in C. elegans and A.
caninum.(A) Wild-type pheromone-induced dauer larvae placed on plates
containing bacterial food, no bacteria and no pheromone, bacteria and 1 mM
atropine, and pheromone at 25°C and scored 24–42 hr later for dauer larvae
and reproductive L4yadults. Experiments were performed at least twice. With
no food or pheromone, 91% of the animals remain arrested (n51,141). Dauer
larvae placed on plates with food recovered efficiently, with less than 10%
remaining arrested at the dauer stage (n52,596). The addition of 1 mM
atropine in the presence of food partially inhibited dauer recovery: 25%
remained arrested at the dauer stage (n51,311). Eighty-nine percent of the
animals maintained on plates with pheromone but no food (n51,027)
remained arrested at the dauer stage. The pheromone preparation contained
bacterial contaminants that may have been used as a food source. In A.
caninum incubated with 10% serum and 25 mM GSM, 9% of the infective
larvae remained as dauer larvae and did not resume feeding. The addition of
atropine (0.5 mM) to the serum and GSM completely inhibited recovery of A.
caninum L3, and no worms resumed feeding (data not shown). (B)daf-
7(e1372) (horizontally striped bars) and daf-2(e1370) (diagonally striped bars)
dauer larvae from 25°C liquid cultures were placed onto plates at 15°C.
Animals were scored for the presence of dauer larvae and reproductive adults
after 2 days. Temperature downshift induced dauer recovery only very slightly
in daf-2(1370) animals (4%, n5320) and not in either daf-7(1372) (0.2%, n5
330) or daf-2(1391) animals, where 100% of the animals remained as dauers
(n5164). At 15°C with food, 65% of the daf-2(e1370) (n5384) and 43% of
the daf-7(e1372) dauer larvae (n5587) recovered and 76% of daf-2(e1391)
animals recovered (n5458). Atropine at 1 mM potently inhibits dauer
recovery of daf-2(e1370) to 18% (n5228), daf-7(e1372) to 6% (n5405), and
daf-2(e1391) to 23% (n5363) dauer larvae on plates at 15°C with food. For
daf-2(1370) and daf-7(e1372), each experiment was performed two or three
times, whereas for daf-2(e1391), the numbers are from only one trial. The
difference between the effects of atropine on pheromone-induced dauer
larvae and either daf-2-ordaf-7-induced dauer larvae may be because the
pheromone-induced dauer larvae had been arrested longer than the daf-2
and daf-7 dauer larvae. Older dauer larvae will recover when exogenous
pheromone is removed, even without the addition of food, but younger dauer
larvae do not (2). Alternatively, pheromone may inhibit both the TGF-
b
- and
insulin-like-signaling pathways (and perhaps other signals) whereas daf-2 or
daf-7 mutants may only decrease one endocrine signal.
Tissenbaum et al. PNAS
u
January 4, 2000
u
vol. 97
u
no. 1
u
463
NEUROBIOLOGY
insulin-signaling mutants. Thus, the cholinergic input to dauer
recovery depends on insulin-like signaling. We suggest that
muscarinic agonists induce recovery of the TGF-
b
pathway
mutant dauer larvae or pheromone-induced dauer larvae by
stimulating signaling in the daf-2 insulin-like pathway. In this
way, cholinergic stimulation can induce recovery in animals with
defective TGF-
b
pathway genes but not in animals with defective
insulin-like pathway genes.
In vertebrates, studies link the muscarinic and insulin signaling
pathways. Both adrenergic and cholinergic fibers innervate
secretory cells in the vertebrate islet of Langerhans (16, 31).
Consistent with the suggestion that muscarinic inputs increase C.
elegans insulin-like signaling, mammalian autonomic cholinergic
fibers enhance insulin secretion (29, 31). Pharmacological stim-
ulation with acetylcholine or carbachol induces insulin release
both in vivo and in vitro. This induction is completely abolished
by atropine, showing that it is mediated by activation of musca-
rinic receptors on the beta cells (16, 32, 33). In mammalian
systems, binding of acetylcholine to the beta cell muscarinic
receptor causes activation of sodium channels, which, in turn,
leads to a change in membrane potential to induce insulin release
(ref. 33; Fig. 3).
These data suggest the model shown in Fig. 3 for dauer
recovery in C. elegans. When pheromone levels decrease and
food levels increase, acetylcholine is secreted from an as yet
unidentified neuron and binds to the muscarinic receptor on an
insulin-like secreting neuron or other cell. This induces secretion
of an insulin-like signal, in turn, to induce dauer recovery.
Insulin-secreting pancreatic beta cells have many neuronal fea-
tures and are thought to be specialized ‘‘ganglia’’ related to the
enteric nervous system of lower vertebrates (34). In addition,
proteins related to insulin are produced by neurons that regulate
metabolism in Limulus (35). Distant relatives of insulin are
found in the C. elegans genome database (ref. 8; S. Pierce and
G.R., unpublished observations). We suggest that the secretory
cells expressing such an insulin-like gene will also express
muscarinic receptors and have connections to food, pheromone,
and temperature sensory neurons (Fig. 3).
Temperature acts as a modulator for dauer recovery (refs.
2 and 14; Fig. 2). The thermoregulatory circuit for temperature
sensation and output of that information to motor and endo-
crine pathways has been identified (11, 12, 36). This pathway
consists of the thermosensory neuron AFD coupled to the
interneurons AIY and AIZ (11, 12, 14, 36). ttx-3, a gene that
affects AIY function, is expressed exclusively in the AIY
interneurons (11, 12). Mutations in ttx-3 decouple this ther-
moregulatory pathway from the dauer pathway: daf-7; ttx-3
double mutant animals form dauer larvae that recover at high
temperature, unlike daf-7 single mutants (12). However, daf-2;
ttx-3 double mutant dauer larvae do not recover at high
temperature, like the daf-2 mutant alone (O. Hobert and G.R.,
unpublished observation). We suggest that thermosensor y
signals through the thermoregulatory AIY and AIZ interneu-
rons couple to insulin-like secretory neurons (Fig. 3). Given
that rates of growth and metabolism are intimately connected
to cultivation temperature in invertebrates, the coupling of
thermosensation to metabolic control is reasonable. Such a
coupling of thermosensory input to metabolic control by the
daf-2 insulin-like signaling pathway may be analogous (or even
homologous) to the hypothalamic modulation of autonomic
input to the pancreatic beta cells (31, 33, 34).
A muscarinic signaling pathway also induces recovery of the
hookworm infective L3 from their arrested ‘‘dauer’’ state. Re-
covery from dauer arrest in hookworm occurs in the host in
response to an undefined host-specific signal. We suggest that
up-regulation of an insulin-like molecule by a cholinergic path-
way also causes dauer recovery upon entr y into the host in A.
caninum. Known muscarinic signaling drugs may constitute
Fig. 3. A model for cholinergic input induction of dauer recovery. In dauer pheromone or in a daf-7 mutant, the DAF-7 TGF-
b
ligand is not produced by the
ASI sensoryysecretory neuron. Therefore, there is no activation of the DAF-1 and DAF-4 TGF-
b
receptors or downstream DAF-8 and DAF-14 Smad proteins, and
this results in high DAF-3 Smad activity in signaling cells or target tissues. In pheromone without muscarinic agonists, no insulin-like signal is released, causing
less insulin receptor signal transduction to the transcription factor DAF-16, which, in combination with unregulated DAF-3, induces dauer arrest. Under
pheromone-induced or daf-7 mutation-induced dauer induction conditions, muscarinic stimulation causes release of an insulin-like DAF-2 ligand, which
stimulates the DAF-2yAGE-1-signaling pathway to DAF-16 inactivation. Because daf-7 mutants can recover in muscarinic agonists, the TGF-
b
signaling pathway
is not required for dauer recovery. Under normal conditions of dauer recovery upon release from pheromone and addition of food and low temperature, we
suggest that acetylcholine released via temperature or food pathways binds to the muscarinic receptor on the insulin-like signaling cell to cause an increase in
insulin release. We suggest that temperature is coupled via the interneurons AIY and AIZ to the DAF-2 insulin-like signaling pathway rather than the TGF-
b
signaling pathway because mutations in the thermoregulatory gene ttx-3 can suppress mutations in daf-7 and not mutations in daf-2 (ref. 11; O. Hobert and
G.R., unpublished observations).
464
u
www.pnas.org Tissenbaum et al.
novel chemotherapeutic strategies to perturb the dauer mainte-
nance process in invertebrate hosts as well as the recovery
process in human hosts.
We thank Jim Thomas for providing some of the strains used in this study
and Ann Sluder for the minimal media protocol. Some of the strains were
obtained from the Caenorhabditis Genetics Center, which is supported
by the National Institutes of Health National Center for Research
Resources. We thank members of the Ruvkun and Kaplan labs and Allan
Dines for helpful discussions, suggestions, and critical reading of the
manuscript. This work was funded by a grant from Hoechst and National
Institutes of Health Grant R01AG14161 to G.R. Part of this work was
completed while H.A.T. was supported by the Helen Hay Whitney
Foundation.
1. Tauber, M. J., Tauber, C. A. & Masaki, S. (1986) in Seasonal Adaptation of
Insects (Oxford Univ. Press, New York).
2. Riddle, D. L. & Albert, P. S. (1997) in C. elegans II, eds. Riddle, D. L.,
Blumenthal, T., Meyer, B. J. & Priess, J. R. (Cold Spring Harbor Lab. Press,
Plainview, NY), pp. 739 –768.
3. Riddle, D. L. & Bird, A. F. (1985) J. Nematol. 17, 165–168.
4. Schmidt, G. D. & Roberts, L. S. (1985) in Foundations of Parasitology (Times
MirroryMosby, St. Louis).
5. Hawdon, J. M. & Schad, G. A. (1991) in Developmental Adaptations in
Nematodes., ed. Toft, C. A. (Oxford Univ. Press, Oxford), pp. 274 –298.
6. Patterson, G. I., Koweek, A., Wong, A., Liu, Y. & Ruvkun, G. (1997) Genes
Dev. 11, 2679–2690.
7. Ogg, S., Paradis, S., Gottlieb, S., Patterson, G. I., Lee, L., Tissenbaum, H. A.
& Ruvkun, G. (1997) Nature (London) 389, 994–999.
8. Duret, L., Guex, N., Peitsch, M. C. & Bairoch, A. (1998) Genome Res. 8,
348–353.
9. Kimura, K., Tissenbaum, H. A., Liu, Y. & Ruvkun, G. (1997) Science 277,
942–946.
10. Malone, E. A. & Thomas, J. H. (1994) Genetics 136, 879 – 886.
11. Hobert, O., D’Alberti, T., Liu, Y. & Ruvkun, G. (1998) J. Neurosci. 18,
2084–2096.
12. Hobert, O., Mori, I., Yamashita, U., Honda, H., Ohshima, Y., Liu, Y. &
Ruvkun, G. (1997) Neuron 19, 345–357.
13. Rand, J. B. & Nonet, M. L. (1997) in C. elegans II, eds. Riddle, D. L.,
Blumenthal, T., Meyer, B. J. & Priess, J. R. (Cold Spring Harbor Lab. Press,
Plainview, NY), pp. 611– 643.
14. Bargmann, C. I. & Mori, I. (1997) in C. elegans II, eds. Riddle, D. L.,
Blumenthal, T., Meyer, B. J. & Priess, J. R. (Cold Spring Harbor Lab. Press,
Plainview, NY), pp. 717–737.
15. Bargmann, C. I. (1998) Science 282, 2028 –2033.
16. Ahren, B., Taborsky, G. J., Jr., & Porte, D., Jr. (1986) Diabetologia 29, 827–836.
17. Miller, R. E. (1981) Endocr. Rev. 2, 471– 494.
18. Hawdon, J. M. & Schad, G. A. (1993) Exp. Parasitol. 77, 489– 491.
19. Sulston, J. & Hodgkin, J. (1988) in The Nematode Caenorhabditis elegans, ed.
Wood, W. B. (Cold Spring Harbor Lab. Press, Plainview, NY), pp. 587– 606.
20. Gottlieb, S. & Ruvkun, G. (1994) Genetics 137, 107–120.
21. Hawdon, J. M. & Schad, G. A. (1991) J. Helm. Soc. Wash. 58, 140 –142.
22. Hawdon, J. M., Jones, B. F., Perregaux, M. A. & Hotez, P. J. (1995) Exp.
Parasitol. 80, 205–211.
23. Hawdon, J. M. & Schad, G. A. (1990) J. Parasitol. 76, 394–398.
24. Aver y, L., Bargmann, C. I. & Horvitz, H. R. (1993) Genetics 134, 455–464.
25. Lef kowitz, R. J., Hoffman, B. B. & Taylor, P. (1996) in The Phar macological
Basis of Therapeutics, eds. Hardman, J. G. & Limbird, L. E. (McGraw–Hill, New
York), pp. 105–139.
26. Brown, J. H. & Taylor, P. (1996) in The Pharmacological Basis of Therapeutics,
eds. Hardman, J. G. & Limbird, L. E. (McGraw–Hill, New York), pp. 141–160.
27. Haim, N., Nahum, S. & Dudai, Y. (1979) J. Neurochem. 32, 543–552.
28. Lee, Y. S., Park, Y. S., Chang, D. J., Hwang, J. M., Min, C. K., Kaang, B. K.
& Cho, N. J. (1999) J. Neurochem. 72, 58– 65.
29. Yamamura, H. I. & Snyder, S. H. (1974) Proc. Natl. Acad. Sci. USA 1725–1729.
30. Burgermeister, W., Klein, W. L., Nirenberg, M. & Witkop, B. (1978) Mol.
Pharmacol. 14, 240 –256.
31. Woods, S. C. & Porte, D., Jr. (1974) Physiol. Rev. 54, 596– 619.
32. Latifpour, J., Gousse, A., Yoshida, M. & Weiss, R. M. (1992) J. Urol. 147,
760–763.
33. Henquin, J.-C. (1994) in Joslin’s Diabetes Mellitus, eds. Kahn, C. R. & Weir,
G. C. (Lea & Febiger, Malvern, PA), pp. 56– 80.
34. Bonner Weir, S. & Smith, F. E. (1994) in Joslin’s Diabetes Mellitus, eds. Kahn,
C. R. & Weir, G. C. (Lea & Febiger, Malvern, PA), pp. 15–28.
35. Roovers, E., Vincent, M. E., van Kesteren, E., Geraets, P. M., Planta, R. J.,
Vreugdenhil, E. & van Heerikhuizen, H. (1995) Gene 162, 181–188.
36. Mori, I. & Ohshima, Y. (1995) Nature (London) 376, 344 –348.
Tissenbaum et al. PNAS
u
January 4, 2000
u
vol. 97
u
no. 1
u
465
NEUROBIOLOGY