G3: Genes|Genomes|Genetics Early Online, published on September 13, 2017 as doi:10.1534/g3.117.300263
1
Title
2
Serotonin drives predatory feeding behavior via synchronous feeding rhythms in
3
the nematode Pristionchus pacificus
4
5
Authors
6
Misako Okumura*, 2, Martin Wilecki* and Ralf J. Sommer*,1
7
*
8
Biology, 72076, Tuebingen, Baden-Wuerttemburg, Germany, 2Present address: Cell
9
biology Laboratory, Department of Biological Science, Graduate School of Science,
10
Max-Planck Institute for Developmental Biology, Department for Evolutionary
Hiroshima University, Hiroshima, 739-8526, Japan.
11
12
Reference numbers: 35
13
14
Short running title
15
Serotonin regulates worm predation
16
1
© The Author(s) 2013. Published by the Genetics Society of America.
17
Keywords
18
Predatory feeding behavior, Serotonin, tph-1, bas-1, Pristionchus pacificus
19
20
Corresponding author
21
Ralf J. Sommer, 1Max-Planck Institute for Developmental Biology, Department for
22
Evolutionary
23
[email protected], +49-7071-601-371
Biology,
Spemannstrasse
24
2
37,
72076,
Tuebingen,
Germany,
25
Abstract
26
Feeding behaviors in a wide range of animals are regulated by the neurotransmitter
27
serotonin, although the exact neural circuits and associated mechanism are often
28
unknown. The nematode Pristionchus pacificus can kill other nematodes by opening
29
prey cuticles with movable teeth. Previous studies showed that exogenous serotonin
30
treatment induces a predatory-like tooth movement and slower pharyngeal pumping in
31
the absence of prey, however, physiological functions of serotonin during predation and
32
other behaviors in P. pacificus remained completely unknown. Here, we investigate the
33
roles of serotonin by generating mutations in Ppa-tph-1 and Ppa-bas-1, two key
34
serotonin biosynthesis enzymes and by genetic ablation of pharynx-associated
35
serotonergic neurons. Mutations in Ppa-tph-1 reduced the pharyngeal pumping rate
36
during bacterial feeding compared to wild type. Moreover, the loss of serotonin or a
37
subset of serotonergic neurons decreased the success of predation, but did not abolish
38
the predatory feeding behavior completely. Detailed analysis using high-speed camera
39
revealed that the elimination of serotonin or the serotonergic neurons disrupted the
40
timing and coordination of predatory tooth movement and pharyngeal pumping. This
3
41
loss of synchrony significantly reduced the efficiency of successful predation events.
42
These results suggest that serotonin has a conserved role in bacterial feeding and in
43
addition drives the feeding rhythm of predatory behavior in Pristionchus.
44
4
45
Introduction
46
Animals utilize diverse food sources and exhibit many different feeding behaviors
47
depending on their ecology and physiology. In nematodes for example, their extreme
48
habitat diversity is reflected by a range of different feeding adaptations and food sources
49
including bacteria, fungi, protists and plant roots (Yeates et al. 1993). While the neural
50
mechanisms
51
Caenorhabditis elegans (Avery 2012), the neural regulation of the feeding behaviors
52
utilized by other diverse nematodes are unknown. One of the most striking examples of
53
nematode behaviors is predatory feeding in which certain nematode species, such as
54
Pristionchus pacificus, can use other nematodes as prey (Bento et al., 2010). Studying
55
the different feeding strategies of these two species can therefore provide insights into
56
the evolution of the neural mechanisms associated with feeding type switches and
57
predation.
regulating
bacterial
feeding
have
been
studied
intensively
in
58
P. pacificus has been established as a satellite model organism for comparison
59
with C. elegans and for integrative studies in evolutionary biology (Sommer et al.,
60
1996; Sommer, 2015). Various genomic tools are available, such as an annotated
5
61
genome, genetic transformation and reverse genetics through CRISPR/Cas9 engineering
62
(Dieterich et al. 2008; Schlager et al. 2009; Witte et al. 2015). In its natural
63
environment, P. pacificus often occurs in a necromenic association with scarab beetles,
64
that is, arrested, long-lived dauer larvae stay on a beetle and after the beetle’s death, the
65
worms exit the dauer stage and flourish on the carcass (Herrmann et al. 2007; Meyer et
66
al. 2017).
67
Predatory feeding in P. pacificus is dependent on the presence of teeth-like
68
denticles, which are capable of rupturing the cuticle of the prey. Interestingly, P.
69
pacificus displays a mouth-form dimorphism, which is influenced by environmental
70
factors (Bento et al. 2010; Ragsdale et al. 2013). The eurystomatous (Eu) mouth form is
71
characterized by i) a wide mouth opening, ii) a large, claw-like dorsal tooth, and iii) a
72
hook-shaped subventral tooth, which together enable predatory feeding (Figure 1A and
73
1B). In contrast, the stenostomatous (St) form has a narrow mouth opening and a single
74
flint-shaped dorsal tooth, which only allows bacterial feeding and scavenging, but not
75
predation (Figure 1C and 1D). In this study, we focus on the predatory feeding behavior
76
of the Eu form to gain insight into the neural regulation of predation.
6
77
Previous work identified a feeding mode switch between bacterial feeding and
78
predation in P. pacificus. During predation, the rate of pharyngeal muscle contractions
79
termed pumping was reduced by half when compared with bacterial feeding, and
80
additionally tooth movement was stimulated (Wilecki et al. 2015). Predatory-like
81
feeding rhythms were also triggered by exogenous serotonin treatment but not by other
82
monoamine neurotransmitters (Wilecki et al. 2015). However, the cellular and
83
mechanistic function of serotonin during feeding behaviors in P. pacificus remains
84
unknown. Among many animal taxa, pharmacological and genetic studies have shown
85
that serotonin regulates feeding behavior (Gillette 2006). In C. elegans, serotonin is
86
synthesized from tryptophan by two conserved enzymes called tryptophan hydroxylase
87
(TPH-1) and 5HTP/L-dopa decarboxylase (BAS-1) (Figure 1E). C. elegans tph-1
88
mutants exhibited a reduced pumping rate during bacterial feeding (Sze et al. 2000) and
89
serotonin activates pharyngeal pumping and isthmus peristalsis mimicking the effect of
90
bacterial food (Horvitz et al. 1982; Avery and Horvitz 1990; Song and Avery 2012).
91
Here, we investigated the role of serotonin in predatory feeding in P. pacificus.
92
We identified the P. pacificus tryptophan hydroxylase (Ppa-tph-1) and 5HTP/L-dopa
7
93
decarboxylase (Ppa-bas-1) and generated null mutants by utilizing the CRISPR/Cas9
94
system. Ppa-tph-1 and Ppa-bas-1 mutants exhibited a strong reduction in predation
95
efficiency in comparison to wild type animals. Using high-speed video microscopy, we
96
found that those mutants disrupt the temporal coordination of pharyngeal pumping and
97
tooth movement. Moreover, genetic ablation of the two serotonergic neurons NSM and
98
ADF mimic the phenotype of Ppa-tph-1 mutants. Our results provide a novel function
99
of serotonin for efficient killing during predation in P. pacificus by regulating the
100
coordination of pharyngeal pumping and tooth movement.
101
8
102
Materials and Methods
103
Experimental Model
104
All strains of P. pacificus were maintained at 20°C on NGM agar plates with E. coli.
105
OP50. All experiments were performed using young adult hermaphrodites by picking J4
106
larvae on the day before the assay. The following mutants and transgenic lines were
107
used:
108
Ppa-tph-1(tu1016) (RS3074), Ppa-bas-1(tu629) (RS2955), Ppa-bas-1(tu630) (RS2956),
109
Ex[Ppa-tph-1p::RFP,
110
Ex[Ppa-tph-1p::caspase-3(p12)::nz,
111
Ppa-egl-20p::RFP]
112
Ppa-tph-1p::RFP, Ppa-egl-20p::RFP] (tuEx256) (RS3065).
C.
elegans
(N2),
P.
pacificus
(PS312).
Ppa-egl-20p::RFP]
(tuEx259)
Ppa-tph-1(tu628)
(tuEx257)
(RS2946),
(RS3066),
Ppa-tph-1p::cz::caspase-3(p17), (RS3068),
Ex[tph-1p::caspase-3(p12)::nz,
113
114
Molecular cloning of Ppa-tph-1 and Ppa-bas-1
115
To identify Ppa-tph-1 and Ppa-bas-1, we utilized phylogenetic analysis between C.
116
elegans and P. pacificus as previously described (Baskaran et al. 2015). Total RNA of
117
the wild type strain (PS312) was purified using ZR RNA MicroPrep (Zymo Research,
9
118
R1060). cDNA was synthesized by reverse transcription and used for 5’ and 3’RACE
119
experiments using the SMARTer RACE 5’/3’ Kit (Clontech, 634858). PCR products
120
were sequenced using Sanger sequencing.
121
122
Generation of Ppa-tph-1 and Ppa-bas-1 mutants by CRISPR/Cas9 system
123
CRISPR mutants were generated as previously described for P. pacificus (Witte et al.
124
2015). To generate Ppa-tph-1(tu628), Cas9 protein (ToolGen Inc.) and sgRNAs
125
(ToolGen Inc.) against Ppa-tph-1 and Ppa-pdl-1 were co-injected into PS312 young
126
adults and screened for the dumpy co-CRISPR marker. Dumpy animals were isolated
127
and the F2 were analyzed further for mutations in Ppa-tph-1. To obtain an additional
128
Ppa-tph-1 mutant allele (tu1016), we synthesized a new sgRNA using the EnGen
129
sgRNA Synthesis Kit, S. pyogenes Cas9 (NEB, #E3322S) and injected PS312 young
130
adult animals. Mutant candidates were obtained by high resolution melting curve
131
analysis (LightCycler 480 High Resolution Melting Master, Roche, 04909631001) of F1
132
progeny of injected animals, confirmed by Sanger sequencing. To generate Ppa-bas-1
133
mutants, Cas9 protein (ToolGen Inc.) and sgRNA (ToolGen Inc.) against Ppa-bas-1
10
134
were injected into PS312 young adults following the same protocol as for Ppa-tph-1.
135
To generate gRNAs, the following gene specific sequences were used;
136
Ppa-pdl-1: TCTCGTGAACGCCAACCGCGTGG
137
Ppa-tph-1(tu628): AGATTCCAAGCAATCGTCATCGG
138
Ppa-tph-1(tu1016): ATGCAATAACGGGAGCTCCTTGG
139
Ppa-bas-1(tu629), Ppa-bas-1(tu630): CTCCTCAGTACCCAGAGGAATGG
140
All mutants were backcrossed with PS312 three times before performing further
141
analysis.
142
143
Generation of transgenic lines
144
A 4.7 kb upstream region of Ppa-tph-1 was amplified by PCR and fused with
145
TurboRFP and rpl-23 3’UTR to generate Ppa-tph-1p::RFP. Caspase-3 (p12)::nz and
146
cz::caspase-3 (p17) were amplified from Pmec-18 caspase-3 (p12)::nz [TU#813]
147
(Addgene #16082) and Pmec-18 cz::caspase-3 (p17) [TU#814] (Addgene #16083),
148
respectively (Chelur and Chalfie 2007). The Ppa-tph-1 promoter, each reconstituted
149
caspase
subunit
and
the
rpl-23
3’UTR
11
were
fused
to
construct
150
Ppa-tph-1p::caspase-3(p12)::nz and Ppa-tph-1p::cz::caspase-3(p17), respectively.
151
These transgenes (10 ng/µl), Ppa-egl-20p::TurboRFP (10 ng/µl) as co-injection maker
152
and PS312 genomic DNA (60 ng/µl) were digested by the appropriate restriction
153
enzymes and were injected into young adults of PS312 as previously described
154
(Schlager et al. 2009).
155
156
Immunostaining
157
Immunostaining was performed as previously described (Desai et al. 1988). Briefly,
158
several plates with adult worms were washed in M9 buffer three times and fixed with
159
4% paraformaldehyde/ PBS at 4°C overnight. Samples were washed with 0.5% Triton
160
X-100/PBS (PBST) three times and treated with 5% -mercaptoethanol/ 1% TX-100/
161
0.1 M Tris (pH 7.5) overnight at 37°C. After being washed in 1% TX-100/ 0.1 M Tris
162
(pH 7.5) three times and in collagenase buffer (1 mM CaCl2/ 1% TX-100/ 0.1 M Tris
163
(pH 7.5)), samples were incubated 1-3 hours at 37°C in 2000 Units/ml Collagenase type
164
IV (SIGMA, C5138) in collagenase buffer. The samples were rinsed with PBST three
165
times and blocked at room temperature for one hour in 1% BSA/PBST. They were
12
166
incubated at 4°C overnight in the first antibody solution and rinsed with PBST three
167
times. After a second blocking step for one hour in 1% BSA/ PBST, samples were
168
incubated for 2-4 hours at 37°C with the secondary antibody solution. Samples were
169
washed with PBST three times and mounted on agarose pads with Vectashield (Vector
170
Laboratories, Inc, H-1000) for imaging. Images were obtained using a TCS SP8
171
confocal microscope (Leica) and were processed using ImageJ. The following
172
antibodies were used in this study; anti-serotonin (1:100, rabbit, SIGMA, S5545),
173
anti-RFP (1:500, mouse, Cell Biolabs, AKR-021), goat anti-rabbit IgG secondary
174
antibody Alexa Fluor 488 (1:100, Thermo Fisher, A-11008), and goat anti-mouse IgG
175
Cy3 (1:100, Jackson ImmunoResearch Laboratories Inc. 115-165-146).
176
177
Egg laying assay
178
Worms were synchronized by bleaching and single J4 larvae were picked to each plate.
179
Worms were transferred to new plates every 12 hours and the number of eggs was
180
counted. To count the number of egg laying in short time intervals, one-day old adults
181
were transferred to 96 well plates with 50 µl of M9 buffer or 10 mM serotonin solution.
13
182
After two-hour incubation at 20°C, the number of eggs was counted. The number of
183
eggs in the uterus was counted by observing three-day old adults with an Axioplan 2
184
microscope (Zeiss).
185
186
Corpse assay
187
The corpse assay was performed as previously described (Wilecki et al. 2015; Lightfoot
188
et al. 2016). To collect young larvae as prey, freshly starved plates of C. elegans (N2)
189
were washed with M9 buffer and larvae were filtered twice with 20 µm Nylon Net
190
Filters (Millipore, NY2004700). Young larvae were centrifuged and 3 µl of worm pellet
191
were dropped on unseeded 60 mm NGM assay plate. Five predators were added to the
192
assay plates. After two hours, the number of corpses was counted. Five plates of each
193
strain were tested per experiment and the experiments were repeated three times. The
194
data is the sum of all the experiments.
195
196
Biting assay
197
The bite assay was performed as previously described (Wilecki et al. 2015; Lightfoot et
14
198
al. 2016). In short, young larvae were collected as mentioned in previous section.
199
Predators were transferred to the assay plate with young larvae of C. elegans and they
200
were allowed to recover for 15 minutes. After recovery, the predator was observed using
201
a light stereomicroscope for 10 minutes and the numbers of biting, killing and feeding
202
events were counted. 4–7 animals of each strain were tested per experiment and the
203
experiments were repeated two or three times. The data are the sum of all the
204
experiments.
205
206
Pumping and tooth movement
207
Pumping and tooth movement were observed as previously described (Wilecki et al.
208
2015; Lightfoot et al. 2016). Worms were transferred to 35 mm NGM assay plates with
209
E. coli OP50 or plates full with young larvae of C. elegans (N2). After 15-minute
210
recovery time, pharyngeal pumping and tooth movement were video-recorded using a
211
high-speed camera (Axio Imager A1, Zeiss) with 50 Hz for 15 seconds. For treatment
212
with serotonin or dopamine, worms were incubated in 5 mM solutions of the
213
neurotransmitters for four hours and pharyngeal pumping and tooth movement were
15
214
observed within 20 minutes after transferring worms onto assay plates. The timing of
215
tooth closure and muscle relaxation were observed and plotted as tooth movement and
216
pumping, respectively.
217
218
Statistical Analysis
219
Data were processed and plotted using Excel and R 3. 2. 1 and statistical analysis was
220
performed using R 3. 2. 1. All data are represented as mean± SEM unless otherwise
221
noted. Statistical details of experiments, including statistical tests, statistical
222
significance, the exact n numbers, can be found in the figure legends. “n” is denoting as
223
the number of animals used except for corpse assay (Figure 2E). When p values are
224
lower than 0.05, the results were considered to be statistically significant. No statistical
225
methods were used to predetermine sample size. The experiments were not randomized,
226
and the investigators were not blinded to allocation during experiments and outcome
227
assessment.
228
229
Data availability
16
230
All materials and data used in this study, including mutant, transgenic lines, and
231
plasmids are available upon request from the corresponding authors.
232
233
17
234
Results
235
CRISPR/Cas9 induced Ppa-tph-1 mutants show conserved functions in pharyngeal
236
pumping and egg-laying
237
To study the involvement of serotonin in predatory feeding behavior, we
238
generated mutants in the tryptophan hydroxylase (Ppa-tph-1) gene of P. pacificus using
239
the CRISPR/Cas9 system because TPH-1 is an essential enzyme for serotonin
240
biosynthesis (Figure 1E and 1F). We obtained two independent alleles of Ppa-tph-1,
241
tu628 with a 10-bp insertion and tu1016 with a 2-bp deletion and a 22-bp insertion that
242
resulted in a premature stop codon (Figure 1F). To confirm that the mutant alleles are
243
indeed
244
anti-serotonin antibody (Wilecki et al. 2015). In wild type animals, serotonin is
245
expressed in three pairs of neurons (NSM, ADF, I1) in the head region and in the
246
hermaphrodite-specific neurons (HSN) in the vulva region (Figure 1G). Serotonin
247
expression was completely absent in Ppa-tph-1(tu628) and Ppa-tph-1(tu1016) mutants
248
(Figure 1H and 1I). Thus, both Ppa-tph-1 alleles are unable to produce serotonin and
249
are likely strong reduction-of-function or null alleles.
non-functional,
we
performed
immunostaining
18
experiments
with
an
250
In C. elegans, serotonin is involved in feeding behavior and Cel-tph-1
251
mutants show a reduction in the rate of pharyngeal pumping on bacterial food (Sze et al.
252
2000). We examined the pumping rate during bacterial feeding in P. pacificus to identify
253
if this function of serotonin is conserved between both species. We found a significantly
254
lower pumping rate of Ppa-tph-1 mutants in comparison to wild type (p < 2.9*10-7),
255
indicating that serotonin controls the pumping rate during bacterial feeding in P.
256
pacificus (Figure 2A). As C. elegans serotonin expression in the HSN neurons regulates
257
egg-laying behavior (Sze et al. 2000), we also investigated its potential role in P.
258
pacificus. The lifetime fecundity of the P. pacificus wild type reference strain (PS312) is
259
approximately 150 progeny, whereas Ppa-tph-1 mutants showed a significant reduction
260
with an average of 50 progeny (Figure 2B). We also counted egg laying after a two-hour
261
treatment with serotonin or M9 buffer control treatments. Although the serotonin
262
treatment suppressed egg laying in wild type confirming previous results (Gutierrez and
263
Sommer 2007), it partially rescued the egg-laying defect in Ppa-tph-1 mutant animals
264
(Figure 2C, p < 1.0*10-7). The number of eggs in utero also increased in
265
Ppa-tph-1(tu628) resulting in an egg-laying defective (Egl) phenotype in 80% of mutant
19
266
adults (Figure 2D). In contrast, wild type P. pacificus eggs are usually released in the
267
four or eight cell stages and an Egl phenotype is observed extremely rarely. Together,
268
these data suggest that the function of serotonin during pharyngeal pumping is
269
conserved between P. pacificus and C. elegans, and a proper amount of serotonin is
270
important for egg laying in P. pacificus.
271
272
Ppa-tph-1 is required for efficient killing during predation
273
Next, we set out to test the involvement of serotonin in predatory feeding. We
274
performed two killing assays termed “corpse assays” and “biting assays” that were
275
established previously (Wilecki et al. 2015; Lightfoot et al. 2016). In short, in corpse
276
assays five P. pacificus adults are incubated as predators with thousands of C. elegans
277
L2 larvae as prey on unseeded NGM plates and the number of corpses are counted after
278
a two-hour time period. Comparing the Ppa-tph-1 mutants with wild type animals we
279
found a strong reduction in the number of corpses, however, corpses were still
280
detectable and not completely eliminated (Figure 2E). Specifically, five wild type
281
predators kill between 50 and 60 C. elegans larval prey, whereas Ppa-tph-1 mutants kill
20
282
around 30 animals. In order to examine the predatory behavior in greater detail, we
283
conducted bite assays. For this, one predator is observed over a 10-minute time interval
284
and the number of biting, killing and feeding events on prey larvae are quantified. We
285
observed that Ppa-tph-1 mutants bite and killed less prey than wild type animals (Figure
286
2F). In addition, the number of feeding events was also decreased in Ppa-tph-1 mutants.
287
Together these results suggest that serotonin is required for efficient killing during
288
predation and that the absence of serotonin reduces, but does not completely eliminate,
289
predatory killing and feeding behavior.
290
291
Serotonin is necessary for the coordination of pharyngeal pumping and tooth
292
movement
293
As serotonin appeared to be involved in the efficiency of predatory feeding,
294
we investigated its mechanistic function in more detail. For that, we quantified the
295
feeding rhythm during predation using a high-speed camera system. We found that in P.
296
pacificus wild type animals pharyngeal pumping decreased during predation
297
maintaining an approximately 1:1 ratio of pumping and tooth movement (Fig. 3A and
21
298
3B), confirming previous observations (Wilecki et al. 2015). In Ppa-tph-1 mutants,
299
pharyngeal pumping was often similar, but tooth movement was significantly lower
300
than in wild type animals (Figure 3A, p < 1.1*10-4). Similarly, the ratio of pharyngeal
301
pumping to tooth movement was also decreased in Ppa-tph-1 mutants (Figure 3B, p