Endogenous Nuclear RNAi Mediates
Behavioral Adaptation to Odor
Bi-Tzen Juang,1Chen Gu,1,2Linda Starnes,1,3Francesca Palladino,4Andrei Goga,1Scott Kennedy,5
and Noelle D. L’Etoile1,*
1Departments of Cell & Tissue Biology and Medicine, University of California, San Francisco, 513 Parnassus Avenue, San Francisco,
CA 94143-0512, USA
2Amunix, Inc., 500 Ellis Street, Mountain View, CA 94043, USA
3Chromatin Structure and Function Group, NNF Center for Protein Research, Faculty of Health Sciences, University of Copenhagen,
Blegdamsvej 3B, Room 4.3.07, 2200 Copenhagen N, Denmark
4E´cole Normale Supe ´rieure de Lyon, CNRS, Molecular Biology of the Cell Laboratory/ UMR5239, Universite ´ Claude Bernard Lyon,
69007 Lyon, France
5Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI 53706, USA
Most eukaryotic cells express small regulatory
RNAs. The purpose of one class, the somatic endog-
enous siRNAs (endo-siRNAs), remains unclear. Here,
we show that the endo-siRNA pathway promotes
odor adaptation in C. elegans AWC olfactory neu-
rons. In adaptation, the nuclear Argonaute NRDE-3,
which acts in AWC, is loaded with siRNAs targeting
odr-1, a gene whose downregulation is required for
adaptation. Concomitantwithincreased odr-1siRNA
in AWC, we observe increased binding of the HP1
homolog HPL-2 at the odr-1 locus in AWC and
reduced odr-1 mRNA in adapted animals. Phosphor-
ylation of HPL-2, an in vitro substrate of the EGL-4
kinase that promotes adaption, is necessary and
sufficient for behavioral adaptation. Thus, environ-
mental stimulation amplifies an endo-siRNA negative
feedback loop to dynamically repress cognate gene
expression and shape behavior. This class of siRNA
may act broadly as a rheostat allowing prolonged
stimulationto dampen gene expression and promote
cellular memory formation.
RNA interference (RNAi) has beenexploited asapowerful exper-
imental tool in both somatic and germ cells for over a decade
(Fire et al., 1998), and organisms ranging in complexity from
yeast to humans produce a range of endogenous small RNAs
of 20–30 nucleotides in length. Although it is apparent that
almost all cells of an organism are actively engaged in some
form of endogenous RNAi, its role, particularly in somatic
cells, remains unclear (reviewed in Ketting, 2011; Ghildiyal and
Endogenous small RNAs are grouped into three classes
according to their biosynthetic origin and the proteins they
bind: piwi-RNAs (piRNAs), micro RNAs (miRNAs), and endoge-
are encoded by genes, whereas in C. elegans, endo- siRNAs are
produced by RNA-dependent RNA polymerases that use
thousands of cellular messenger RNAs (mRNAs) as templates
to produce antisense small RNAs (Ghildiyal and Zamore, 2009;
Ketting, 2011; Gent et al., 2010; Gu et al., 2009). Small RNAs
have been linked to synaptic function and memory formation in
mammals (McNeill and Van Vactor, 2012). For instance, the
microRNA miR134 was shown to repress context-dependent
fear learning and long-term potentiation in mice (Gao et al.,
2010), and a piRNA has been shown to promote long-term
synaptic facilitation of cultured Aplysia sensory neurons (Rajase-
thupathy et al., 2012). However, the extent to which small RNAs
couple environmental stimuli to synaptic plasticity and the
mechanism by which small RNAs regulate experience-induced
behavioral changes remain a mystery.
Prolonged odor exposure induces a form of behavioral plas-
ticity termed adaptation. C. elegans is innately attracted to
food-related odors, but the attraction is diminished if starvation
accompanies the odor. The resulting odor-adapted state lasts
until the animal is fed (Colbert and Bargmann, 1997; Lee et al.,
2010). Odor sensation (Bargmann et al., 1993) and adaptation
(L’Etoile et al., 2002) occur within the olfactory sensory neuron
that is referred to as AWC. Whereas odor sensation requires
the guanylyl cyclase (GC) ODR-1, odor adaptation requires
downregulation of ODR-1 (L’Etoile and Bargmann, 2000).
Decreased intracellular cGMP, in part, drives the cGMP-depen-
dent protein kinase EGL-4 into the AWC nucleus (O’Halloran
et al., 2012). Once in the nucleus, EGL-4 is both necessary and
sufficient to induce long-lasting odor adaptation (Lee et al.,
2010). The mechanism by which nuclear EGL-4 triggers long-
lasting odor adaptation is not known.
Small RNAs can regulate gene expression in both the cyto-
plasm and nucleus. For instance, miRNAs and siRNAs act as
guides to target mRNAs for repression in the cytoplasm (re-
viewed in Ketting, 2011; Ghildiyal and Zamore, 2009). piRNAs
and siRNAs can enter nuclei to trigger cotranscriptional gene
silencing (nuclear RNAi) (Guang et al., 2008; Le Thomas et al.,
1010 Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc.
2013). During nuclear RNAi in C. elegans, the Argonaute (Ago)
transcripts that exhibit sequence complementarity to NRDE-3-
associated siRNAs (Guang et al., 2008; Guang et al., 2010).
NRDE-3 recruits the conserved nuclear protein NRDE-2 and
NRDE-4, to RNAi-targeted nascent transcripts to inhibit RNA
polymerase II (RNAP II) elongation (Guang et al., 2010; Burkhart
et al., 2011). In addition, genes targeted for silencing by the
nuclear RNAi pathway accumulate the repressive chromatin
mark, H3K9me3 (Guang et al., 2010; Burton et al., 2011). In the
C. elegans germline, piRNAs and siRNAs trigger nuclear RNAi
at thousands of genomic loci (Claycomb et al., 2009; Gu et al.,
2009; Ashe et al., 2012; Lee et al., 2012; Shirayama et al.,
2012), and the silencing effects can endure for more than five
generations (Vastenhouw et al., 2006; Buckley et al., 2012).
When nuclear RNAi is disabled, C. elegans germlines lose their
immortal character (Buckley et al., 2012).
of evidence indicate that, in the AWC olfactory sensory neurons
of adult-behaving C. elegans, endogenous RNAi promotes
odor adaptation by repressing the odr-1 gene. First, we show
cipitates (coIPs) odr-1-directed endo-siRNAs, and in adapted
Third, odor exposure diminishes the levels of odr-1 mRNA.
Fourth, in odor adaptation, HPL-2, a heterochromatin-binding
protein, is loaded onto the odr-1 locus. Additionally, we find
that phosphorylation of HPL-2 at sites that are in vitro targets
of the odor-responsive kinase EGL-4 is both necessary and
sufficient to promote odor adaptation in the AWC neurons of an
mentally relevant experiences may regulate gene expression,
thereby shaping behavior in a specific and dynamic fashion.
The Nuclear RNAi Argonaute NRDE-3 Is Required in the
AWC Sensory Neuron for Odor Adaptation
C. elegans is innately attracted to the odor, butanone. Attraction
is assessed by the chemotaxis assay shown in Figure 1A, which
allows quantification of the behavior in the form of a chemotaxis
index (CI) (Bargmann et al., 1993). Naive wild-type animals
exhibit a high CI to butanone, which decreases after 80 min of
butanone exposure in the absence of food (Colbert and
Bargmann, 1995). This experience-dependent decrease in CI is
termed long-term olfactory adaptation. If the adapted CI is
greater than one half of the naive CI, a strain is considered
To investigate the role of small RNAs in long-term olfactory
adaptation, we examined butanone adaptation in strains defec-
tive for major pathways producing RNAi in the soma, including
the microRNA, exogenous RNA (exo-RNAi), and endogenous
RNAi pathways. Animals lacking Dicer (DCR-1) were defective
for adaptation (Figure 1B). Dicer, an RNAase III, processes dou-
ble-stranded (dsRNA) into small noncoding RNAs (Grishok et al.,
the microRNA, exo-, and endo-siRNA interference pathways
(Grishok et al., 2001; Knight and Bass, 2001; Grishok et al.,
2005). These data suggest that Dicer-mediated processing of
dsRNA is required for adaptation. By contrast, the adapted CI
of strains bearing mutations in the pri-miRNA-processing RNase
III enzyme Drosha, DRSH-1 (Denli et al., 2004), the miRNA-
binding Ago, ALG-2 (Vasquez-Rifo et al., 2012), or the exo-
RNAi pathway Ago, RDE-1 (Tabara et al., 1999), were not
significantly different from the CI of wild-type controls (Figures
1B and S2 available online). These data suggest that, if
Dicer-mediated dsRNA processing is required for butanone
adaptation, microRNAs or the exoRNAi pathway are unlikely to
mediate this process.
MUT-7, a putative 30to 50exonuclease, is required for accu-
mulation of endogenous 22 nucleotide siRNAs that bind the
WAGO clade of Agos (Yigit et al., 2006; Lee et al., 2006; Gu
et al., 2009) and accumulation of 26 nucleotide siRNAs (Zhang
et al., 2011), as well as transposon and transgene silencing,
exogenous RNAi, and proper chromosome segregation (Ketting
etal.,1999;Tabaraetal.,1999;Dernburg etal.,2000;Tops etal.,
2005). MUT-7 is also required for nuclear accumulation of
NRDE-3 (Guang et al., 2008). HPL-2 is one of two C. elegans
homologs of Heterochromatin Protein 1 (HP1) (Couteau et al.,
2002). HPL-2 is involved in multiple cellular events, including
gene regulation and DNA replication and repair (Couteau et al.,
2002; Coustham et al., 2006; Black and Whetstine, 2011), as
well as transgene silencing and piRNA-mediated gene silencing
in the gonad (Grishok et al., 2005; Burkhart et al., 2011; Ashe
et al., 2012; Buckley et al., 2012; Shirayama et al., 2012). Strains
that lacked MUT-7 or HPL-2 were defective for butanone adap-
tation (Figure 1B). These results suggest that heterochromatin
and possibly small RNAs promote odor adaptation downstream
Using mut-7 and hpl-2 promoter fusions to drive expression of
GFP-tagged MUT-7 or HPL-2, respectively, we observed GFP
expression in many cells, including both AWCs (Figure 1C). To
determine whether MUT-7 and HPL-2 act in the AWC neurons,
the site of odor sensation and adaptation, we asked whether
cell-specific expression of MUT-7 and HPL-2 could rescue the
odor adaptation defect of each corresponding mutant strain.
Expressing MUT-7 or HPL-2 solely within the AWC neurons
(from the AWC-specific ceh-36prom3promoter [Etchberger
et al., 2007]) of the respective mutant strain rescued its adapta-
tion defects (Figure 1D). These data indicate that MUT-7 and
HPL-2 act within AWC neurons to promote odor adaptation.
These factors could be required at the time of odor exposure
or developmentally. To distinguish between these possibilities,
we used the heat shock promoter phsp-16.2 (Stringham et al.,
1992) to express each factor in the adult immediately prior to
odor exposure. Heat-shock-driven expression restored adapta-
tion to the mut-7 and hpl-2 strains (Figure 1E). Consistent with a
requirement in the adult, neither morphology nor cell fate of the
AWC was altered by loss of HPL-2 or MUT-7 (Figure S1B and
Table S1). Together, these results indicate that the adaptation
defects of mut-7- and hpl-2-deficient animals are not due to
To address whether MUT-7 and HPL-2 act in the same
molecular pathway, we created mut-7;hpl-2 and control
Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc. 1011
Figure 1. HPL-2 and MUT-7 Act at the Time of Odor Exposure in the AWC Neurons to Promote Adaptation to Butanone
(A) Olfactory adaptationparadigm. Animalsexposed tobufferalone (naive)or butanone(adapted) for80minareplacedatthe‘‘origin’’ofanagar-lined10cmPetri
dish. Butanone is placed at the red and ethanol at the black ‘‘X.’’ Sodium azide (to paralyze the worms) was also placed at each ‘‘X.’’ Animals roam plates 2 hr
before counting. The CI is calculated by subtracting the number of animals at the diluent from the number at the odor and dividing this by the number of animals
that left the origin.
(B) Initial screen of mutant strains defective for siRNA pathways. Barsrepresentmean CIs of strains of theindicated genotype that had either been incubated with
buffer (?) or buffer-diluted butanone (+) for 80 min. Bars for wild-type represent the mean CI of pooled controls for all the strains. All error bars are SEM. The side-
by-side comparisons of each strain with wild-type controls are shown in Figure S1A. ** = p < 0.005, * = p < 0.05, and ‘‘n.s.’’ = p > 0.05. Unless otherwise noted, all
tests weretwo-tailedStudent’s ttest,andallassayswereperformed onseparatedayswith>100animals perassay.drsh-1,alg-2,dcr-1:n=4;rde-1mut-7:n=6;
hpl-2: n = 5.
(legend continued on next page)
1012 Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc.
fbf-1;hpl-2 double-mutant animals. We found that the ability of
hpl-2 or mut-7 single-mutant animals to adapt to odors was
similar to the ability of mut-7;hpl-2 double-mutant animals
(Figure 1F), but the adaptation defects of hpl-2 were enhanced
in the fbf-1;hpl-2 double-mutant strain. These data indicate
that MUT-7 and HPL-2 likely act in the same pathway within
AWC to promote odor adaptation at the time of odor
To probe the involvement of nuclear RNAi in adaptation, we
examined the nuclear Ago, NRDE-3. NRDE-3 interacts with a
subset of endo-22GRNAs and shuttles them into the nucleus,
where they direct cotranscriptional gene silencing (Guang
et al., 2008). NRDE-3 is expressed in the AWC neurons (B.-T.J.
and N.D.L., unpublished data), and the NRDE-3 null (nrde-
3(gg66)) was unable to adapt to butanone (Figure 1G). These
adaptation defects were rescued by expressing NRDE-3 solely
in the AWC neuron (Figures 1G and S1C), demonstrating that
the nuclear RNAi Argonaute NRDE-3 acts in AWC to promote
To better characterizethe nuclearRNAi pathway,wesurveyed
adaptation in siRNA-defective strains that were deemed chemo-
taxis proficient (Table 1 and Figure S1A). In C. elegans, RNAi can
be broken down into three steps: trigger processing, amplifica-
tion, and silencing (reviewed in Pak et al., 2012). We found that
trigger processing factors, Dicer and its partner RDE-4 (Tabara
et al., 2002; Duchaine et al., 2006), are required for adaptation.
(RdRP), RRF-3 (Simmer et al., 2002), was partially required as
rrf-3(pk1426) animals failed to adapt in five out of eight trials.
The silencing factor NRDE-3, along with its nuclear complex of
NRDE-2 (Guang et al., 2010) and NRDE-1 (Burkhart et al.,
2011), were each required. These results suggest that adapta-
tion requires trigger processing, possibly RdRP amplification,
and nuclear Ago-mediated silencing.
Biochemical and genetic analyses have implicated additional
factors in RNAi. Of the many factors shown to associate with
Dicer, DRH-2 (a Dicer-related DExH-box helicase [Tabara
et al., 2002]) and RDE-3 (a b-nucleotidyl transferase) (Duchaine
et al., 2006) were required for adaptation. Taken as a whole,
our genetic analysis indicates that the nuclear RNAi pathway is
likely to act in the AWC neuron to promote odor adaptation
downstream of DCR-1/RDE-4-mediated small RNA production.
odr-1 mRNA Decreases in Odor-Adapted Animals
To identify a target for siRNA in adaptation, we used quantitative
real-time PCR to probe endo-22GRNAs that map to AWC-
expressed genes (see Supplemental Information). We found
that the odr-1-derived 22GRNAs, odr-1.6 and odr-1.7, as well
as the unc-40-derived 22GRNA, unc-40.2, gave the most robust
signals. odr-1 encodes a GC whose downregulation is required
for odor adaptation (L’Etoile and Bargmann, 2000), and unc-40
(C) HPL-2 and MUT-7 are expressed in AWCs. Fluorescent confocal images of wild-type animals expressing the putative hpl-2 (top) or mut-7 (bottom) promoters
driving GFP-tagged versions of each protein. AWC is marked with ceh-36prom3promoter driving mCherry. Anterior is at the left for both images. Figure S1B is
associated with this panel.
third pair of bars) was expressed in AWC from pceh-36prom3in hpl-2(tm1489) or mut-7(pk204), respectively. **p = 0.002, *p = 0.0035, and n > 5 for each.
(E) Expression of HPL-2 or MUT-7 at the time of odor exposure rescued adaptation defects. hpl-2(tm1489) (left) or mut-7(pk204) (right) transgenic for the
respective cDNA under the control of the heat shock promoter (phsp16-2) were heated (+) 1 hr before odor exposure. Heat-treated animals’ exposed CI’s were
mut-7), n > 5 for each.
(F) HPL-2 and MUT-7 act in the same genetic pathway for adaptation. Mean naive (?) and exposed (+) CIs of animals of the indicted genotype. The adaptation
defects of the fbf-1(ok91) strain are due to loss of the translational control pathway (Kaye et al., 2009) that acts in parallel with hpl-2.
(G) Expression of NRDE-3 in AWC rescued the adaptation defects of the nrde-3 mutant strain. Mean CI of naive (?) and exposed (+) wild-type, ndre-3(gg66), and
NRDE-3 expressed in AWC (pceh-36prom3) of the nrde-3(gg66) mutant strain. Figure S1D is associated with this figure.
Error bars for each panel are SEM.
Table 1. Olfactory Adaptation Requires a Nuclear RNAi Pathway
Gene (Allele)Gene Function Butanone Adaptationa
RNase III nuclease defective
drsh-1(ok369) RNase III nucleasepartially chemotaxis
rrf-3(pk1426) RNA-dependent RNA
drh-1(tm1329) RNA helicase (RIG-I)wild-type
drh-2(ok951)RNA helicase (RIG-I)defective
drh-3(ne4253)RNA helicasechemotaxis defective
nrde-3(gg66) nuclear RNAi Argonautedefective
nrde-2(gg91)NRDE-3 binding nuclear
hpl-2(tm1489)histone H3 lysine 9
trimethyl binding (HP1)
aBehavioral assays are shown in Figure S1A.
bHeterozygous animals are marked with hT2::GFP(I,III).
Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc. 1013
Figure 2. Prolonged Odor Stimulation Dynamically Regulates odr-1-Derived 22G RNAs, Association of HPL-2 with the odr-1 Locus, and
Levels of odr-1 mRNA
(A) Diagram of the odr-1 and unc-40 genes. The odr-1 and unc-40 22GRNAs examined are indicated with arrows below the gene. The PCR amplicons for ChIP-
qPCR are in green. PCR amplicons for mRNA analysis are in red.
(B) Prolonged odor exposure decreases odr-1 mRNA levels. Bars represent the mean fold change in unc-40 (gray) or odr-1 (black) mRNA level as a function of
odor exposure (adapted/naive). RNA from animals of the indicated genotype was normalized to act-3 mRNA. The red line indicates ‘‘no change,’’ and the
‘‘no change’’ are different; p < 0.005. # indicates a difference of p = 0.034 (nonparametric, pair-wise comparison) in medians between the naive and adapted
values of the mRNA. p values displayed are from two-tailed Mann-Whitney test of medians. Chemotaxis behavior for each population and the individual data
points for each pair are shown in Figure S2A. Error bars represent SEM, and n > 3.
(C) Chemotaxis behavior correlates with the level of odr-1 mRNA in butanone-adapted animals. The butanone CI of odor-exposed animals was compared with
their odr-1 mRNA level (mRNA levels normalized to act-3 mRNA). Red circles indicate wild-types, and blue triangles indicate mut-7(pk204) animals expressing
(D) Prolonged odor exposure increases NRDE-3-bound odr-1 22GRNA levels. The first five bars represent mean fold change in total 22GRNAs normalized to
odor-insensitive sn2343 RNA in adapted versus naive animals of the indicated genotype. Error bars represent SEM. Red line indicates no change. * = p < 0.03,
(legend continued on next page)
1014 Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc.
(Hedgecock et al., 1990; Colon-Ramos et al., 2007). The gene
structure, along with the amplicons derived from mRNA,
22GRNA, and genomic DNA, are indicated in Figure 2A.
of target mRNA. To determine whether odr-1 message levels are
decreased in odor-adapted populations, we performed quanti-
tative real-time PCR on RNA collected from the same samples
that showed behavioral adaptation to butanone (Figure S2A).
We found that odr-1 mRNA decreased by approximately one
half in odor-adapted as compared to naive populations (Fig-
ure 2B, second bar). By comparison, unc-40 mRNA levels
were unchanged (Figure 2B, first bar). In mut-7(pk204) animals,
odr-1 mRNA levels were not odor responsive (Figure 2B, third
bar, Figure S2A for individual assays and behavior), but expres-
sion of MUT-7 solely within AWC partially restored odor res-
ponsiveness (Figure 2B, fourth bar). Thus, in odor-adapted
populations, the odr-1 mRNA decreases, and these changes
depend on odor exposure as well as functional MUT-7.
To understand whether the modest decrease in odr-1 mRNA
(Figure 2B) has a behavioral consequence, we asked whether
the level of odr-1 mRNA correlates with the CI of odor-adapted
populations. We found that the levels of odr-1 mRNA correlated
strongly with odor attractiveness (Figure 2C). The correlation be-
tween CI and odr-1 mRNA was even stronger in the mut-
7(pk204) strains that expressed MUT-7 solely in the AWC neuron
(Figure 2C). This indicates that the decreases we observe in odr-
1 mRNA in AWC could be responsible for the stably diminished
odor attractiveness that is the hallmark of long-term adaptation.
In the analysis described above, we examined mRNA from
whole worms, but two lines of evidence indicate that this drop
in mRNA occurs within the AWC neurons: loss of odr-1 leads
to the adapted phenotype, and this is rescued by expression
of ODR-1 in the AWC neurons (L’Etoile and Bargmann, 2000),
and overexpression of ODR-1 in AWC alone blocks adaptation
(L’Etoile and Bargmann, 2000). Taken together, the data impli-
cate downregulation of the odr-1 gene in AWC in butanone
odr-1-Directed 22GRNA Increases in the AWC Sensory
Neuron of Adapted Animals
To determine whether there is evidence for the endo-RNAi
pathway acting in adaptation, we used quantitative real-time
in naive and butanone-adapted populations. We found that
expression of the odr-1 22GRNA odr-1.7 increased by more
than 2-fold in adapted animals compared to naive controls (Fig-
ure 2D, second bar, and Figure S2B). The levels of a less abun-
dant 22GRNA, odr-1.6, and unc-40.2, however, did not change
significantly (red line indicates a ratio of 1:1 for adapted to naive
levels) (Figure 2D, first and third bars, and Figure S2B). Thus, a
22GRNA (odr-1.7) complementary to the odr-1 gene increases
in animals adapted to odor.
These measurements of 22GRNAs reflect levels throughout
the animal, including the germline (Gu et al., 2009). To determine
whether odr-1.7 22GRNA is regulated by odor specifically in
AWC, we analyzed 22GRNA from animals that expressed
MUT-7 only in AWC (Figure 1D). Though total odr-1.7 22GRNA
levels were insensitive to odor exposure in mut-7-defective ani-
mals, expression of MUT-7 in AWC restored odor responsive-
ness to this species of 22GRNA (Figure 2D, fourth and fifth
bars, and Figure S2B). Thus, the levels of odr-1.7 22GRNA are
increased by odor exposure when a factor required for
22GRNA accumulation (Gu et al., 2009) is expressed solely
within the AWC neuron.
odr-1 siRNAs Are Loaded onto NRDE-3 in Adaptation
To better understand how the nuclear RNAi pathway might func-
tion in odor adaptation, we asked whether odr-1.7 or unc-40.2
22GRNAs associate with NRDE-3. We probed this association
by IPing 3XFLAG-tagged NRDE-3 (see Figure S2C for behavior).
We found odr-1.6, 1.7 and unc-40.2 coimmunoprecipitated with
NRDE-3.Thelevelofodr-1.7 22GRNAinassociation withNRDE-
3 was increased significantly in adapted animals (Figure 2D,
seventh bar). By contrast, levels of coimmunoprecipitated odr-
1.6 or unc-40.2 22GRNA were not changed in the same animals,
indicating that NRDE-3 specifically binds more odr-1.7 22GRNA
inadapted animals. Thisfinding supports amodel inwhich ODR-
1 mRNA is reduced by NRDE-3/odr-1.7 22GRNA, mediating
downregulation of the odr-1 gene.
HPL-2 Associates with the odr-1 Locus in Odor-Adapted
One biochemical readout of siRNA/NRDE-3-directed silencing is
increased heterochromatin deposition at the targeted locus
(Burkhart et al., 2011; Guang et al., 2010; Gu et al., 2012). To un-
derstand whether odr-1.7/NRDE-3 might target the odr-1 locus
in the odor-adapted AWC neuron, we expressed 3XFLAG-
tagged heterochromatin associated factor, HPL-2, from the
odr-3 promoter (which drives expression in AWCs and four other
neurons; see Figure S2D for behavior). When we performed
chromatin immunoprecipitation (ChIP) of HPL-2 followed by
qPCR on naive and behaviorally adapted populations, we found
that ChIP of the odr-1 locus was increased in adapted AWC neu-
rons (Figure 2E). The greatest increase in HPL-2-associated
Wilcoxon signed-rank test for median values versus no change. p values displayed are the comparison of medians using an unpaired two sample Mann Whitney
nonparametric t test, n > 3. Figure S2B is associated with this panel. The last three bars represent mean fold change in pnrde-3::NRDE-3 coIPed 22GRNA (n = 6)
normalized to the odor-insensitive X-cluster. Error bars represent SEM. * = p < 0.04, Wilcoxon signed rank test for median values versus no change. Displayed p
values are from a pairwise, one-tailed t test, p = 0.0469 of medians. Figure S2C shows the behavior.
(E)Prolonged odorexposurespecificallyincreasesHPL-2bindingtotheodr-1locusinaMUT-7-dependent fashion.Themeanratioof3XFLAG-HPL-2 expressed
in AWC (podr-3) coimmunoprecipitated odr-1 (dark bars) or unc-40 (light bars) in adapted versus naive animals is shown above the genotype of each population.
Error bars represent SEM. Also indicated is the PCR-amplified, coIPed region of each locus corresponding to ‘‘A,’’ ‘‘B,’’ and ‘‘C’’ in (A). Coimmunoprecipitated
Wilcoxon signed-rank test comparing median values to no change (the red line). The median value of odr-1 B was compared to unc-40 B; p = 0.0079 using a
two-tailed Mann Whitney test. n = 5. The final set of bars represents background from nontransgenic animals. Figure S2D shows the behavior.
Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc. 1015
ChIP (8-fold higher in adapted than in naive) was located just
downstream of the region encoding odr-1.7. Further, the odor-
dependent increase was not seen at the unc-40 locus. As a
specificity control for the 22GRNA pathway, we performed
ChIP from mut-7 loss-of-function animals, which show no in-
crease in odr-1.7 22GRNA levels in response to odor and like-
wise show no increase in odr-1 ChIP (Figure 2E). These results
show that odr-1 is a target for increased HPL-2 association in
the odor-adapted AWC. Though this is not the only interpreta-
tion, these results are most consistent with nuclear RNAi target-
ing this locus.
HPL-2 Is a Direct Phosphorylation Target of the Odor-
Responsive Kinase, EGL-4
How might an environmental signal such as odor intersect with
the endogenous nuclear RNAi pathway to mediate adaptation?
Prolonged odor stimulation causes nuclear accumulation of the
cGMP-dependent protein kinase EGL-4 (Figure 3A) (O’Halloran
et al., 2009; Lee et al., 2010), and nuclear EGL-4 is both neces-
sary and sufficient to induce long-term odor adaptation. Indeed,
expression of constitutively nuclear EGL-4 (NLS-EGL-4) in AWC
naive animals (Figure 3B) (Lee et al., 2010; O’Halloran et al.,
2009). MUT-7 or HPL-2 could thus act by promoting nuclear
accumulation of EGL-4. However, we found that nuclear accu-
mulation of EGL-4 was not altered in mut-7 or hpl-2 mutant
strains (Figure 3C). Three lines of evidence led us to hypothesize
activating MUT-7 and HPL-2. First, we found that constitutively
nuclear EGL-4 required both HPL-2 and MUT-7 to induce adap-
tation in naive animals (Figure 3B). Second, predicted EGL-4
phosphorylation sites within MUT-7 and HPL-2 (Figure 4A) are
required for adaptation (Figures 4B, 4C, and S3). Third, expres-
sion of phospho-defective MUT-7 in wild-type animals caused
adaptation defects, suggesting that MUT-7 phosphorylation is
required for this behavioral change (Figure S3D).
MUT-7 andHPL-2 mightbe direct targets of the EGL-4 kinase;
thus, we asked whether NLS-EGL-4 phosphorylates these fac-
tors in vitro. We were unable to purify full-length MUT-7, so we
focused on HPL-2. We found that C. elegans expressed immu-
nopurified NLS-EGL-4 phosphorylated recombinant HPL-2 and
that the level of32P incorporation diminished when the predicted
PKG phosphorylation sites within HPL-2 were mutated (Figures
4D and S3G). We therefore conclude that these sites are direct
targets of EGL-4 in vitro. Thus, it is likely that HPL-2, a nuclear
protein, is directly phosphorylated by EGL-4 once it enters the
Phosphorylation of HPL-2 at EGL-4 Target Sites Is Both
Necessary and Sufficient to Promote Odor Adaptation
If nuclear EGL-4 promotes odor adaptation by phosphorylating
HPL-2 or MUT-7, then mimicking phosphorylation at consensus
sites is predicted to promote adaptation in naive animals. To
test this, we replaced the serines and threonines at each pre-
dicted EGL-4 phosphorylation site in MUT-7 and each in vitro
verified site in HPL-2 with the phosphomimetic, glutamic acid
(Mansour et al., 1994). Expression of the phosphomimetic
form of MUT-7 in wild-type worms had no effect on chemotaxis.
Because only ?50% of known functions of phosphorylated res-
idues can be mimicked by glutamic acid substitutions (Macie-
jewski et al., 1995), we can make no conclusions about
MUT-7 phosphorylation. However, expressing the phosphomi-
metic form of HPL-2 in wild-type animals substantially reduced
naive attraction to butanone, whereas expression of the wild-
type HPL-2 had no effect (Figure 4E). Thus, mimicking phos-
phorylation of HPL-2 at EGL-4 target residues is sufficient to
Figure 3. HPL-2 and MUT-7 Act Downstream of Nuclear EGL-4
(A) Current model for long-term olfactoryadaptation of the AWC neuron. Acute
stimulation of AWC localized G-protein-coupled receptors (GPCR) by odor
(left) causes animals to chemotax toward the odor. After prolonged odor
exposure (right), the cGMP-dependent protein kinase (PKG) EGL-4 trans-
locates to the nucleus to cause animals to ignore the odor for prolonged pe-
riods of time.
(B) Once in the nucleus, EGL-4 requires HPL-2 and MUT-7 to promote
adaptation. The chemotaxis index of the indicated strains that express
constitutively nuclear EGL-4 from a transgene (+) were compared to their
siblings that did not carry this transgene (?). rde-2 is a control, adaptation-
proficient strain (Figure S1A). Importantly, all animals were naive to butanone.
n > 3 with >100 animals analyzed per assay. **p < 0.0001, two-tailed Student’s
t test. Bars represent the mean CIs, and the error bars represent SEM.
(C) HPL-2 and MUT-7 are not required for odor-induced nuclear entry of EGL-
4. GFP-tagged EGL-4 was expressed in either wild-type, hpl-2(tm1489), or
mut-7(pk204) strains. Animals were exposed to buffer alone (naive) or buta-
none for 80 min before imaging. The percentage of the population that showed
nuclear EGL-4 in one AWC neuron was determined.
1016 Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc.
promote behavior that resembles the adapted state. When each
site was analyzed individually, we found that HPL-2(S155E),
which lies in the chromo shadow domain (CSD), had the great-
est effect (Figure S3E).
engages the adaptation machinery in the absence of odor, or it
could nonspecifically diminish AWC function. To distinguish
between these possibilities, we expressed HPL-2 (all S/T to E) in
mutants that lack the downstream adaptation-promoting factor,
OSM-9 (Colbert and Bargmann, 1995). These animals were able
to chemotax significantly better to butanone than the parental
stream of OSM-9. We conclude that phosphorylation of HPL-2 at
EGL-4 target sites is sufficient to promote adaptation even in the
absence of odor exposure. Importantly, EGL-4 is the only PKG in
C. elegans that is required for odor adaptation (Figure S3H).
Thus, it is likely that odor acts via EGL-4 to activate HPL-2.
To understand whether the siRNA pathway was required for
HPL-2(all S/T to E) to induce adaptation, we asked whether
mut-7 was required for this gain-of-function phenotype. Loss-
of-function MUT-7 (mut-7(pk204)) suppressed the ectopic
adaptation seen in naive animals expressing HPL-2(all S/T to
E) (Figure 4F). Thus, phosphorylation of HPL-2 is both necessary
and sufficient for adaptation, but it requires fully functional
MUT-7. This is consistent with the ChIP studies in Figure 2E
that show that accumulation of HPL-2 at the odr-1 locus of
adapted worms requires functional MUT-7. The observation
that HPL-2(allS/T to E) promotes adaptation in the naive ani-
mal—and yet loss of MUT-7 blocks this adaptation—indicates
that, in the naive animal, there is sufficient MUT-7-dependent
RNAi to engage the adaptation process.
An emerging paradigm is that small noncoding RNAs provide
memory of nonself gene expression (Shirayama et al., 2012);
this work extends the role of siRNAs to encoding memory of
the environment and experience. We provided evidence that,
in the olfactory sensory neurons (AWCs) of adult-behaving
C. elegans, endogenous RNAi promotes odor adaptation by
repressing the odr-1 gene (Figure 5). Our data show that, in
response to prolonged odor exposure, odr-1-directed 22GRNAs
increase, and this increase is most likely to occur in the AWC
neuron (Figure 2D). We demonstrated that these 22GRNAs are
loaded on to the nuclear Ago, NRDE-3 (Figure 2D), that acts in
AWC (Figure 1G). NRDE-3 may shuttle the odr-1 22GRNA into
the AWC nucleus, and we have direct evidence that the HP1
homolog, HPL-2, is loaded on to the odr-1 gene in response to
odor (Figure 2E). We provide in vitro evidence that HPL-2 can
bephosphorylated bynuclear EGL-4(Figure 4).Mimickingphos-
phorylation of HPL-2 is sufficient to evoke adaptation behavior.
Phosphorylation of HPL-2 would repress the odr-1 gene and
ultimately lead to the reduced levels of odr-1 mRNA seen in
adapted animals (Figure 2B). This reduction in odr-1 mRNA cor-
relates strongly with behavior (Figure 2C). One gap in this model
is that we do not know whether NRDE-3 or odr-1 22GRNA bind
the odr-1 locus. An alternate explanation is that odr-1 is
repressed by a factor that is itself negatively regulated by a sec-
this scheme, the repressive factor that binds to the odr-1 regula-
at the same part of the odr-1 gene that encodes the odr-1
22GRNA bound by NRDE-3. However, the proposed model is
more parsimonious and consistent with the data than the alter-
native model and leads to the exciting hypothesis that RNAi
may act broadly as a biological rheostat to allow stimulation to
dampen gene expression and may promote cells to alter their re-
sponses as a function of previous stimulation.
Specificity of Odor Adaptation within AWC Neurons
Butanone adaptation does not affect attraction to benzaldehyde
or isoamyl alcohol (Colbert and Bargmann, 1995), so how would
downregulation of ODR-1, a GC required for all AWC responses
(L’Etoile and Bargmann, 2000), specifically adapt the butanone
response? The other odors are sensed by both left and right
AWCs, and butanone is sensed by only one AWC (Wes and
Bargmann, 2001). Indeed, prolonged butanone exposure results
in nuclear EGL-4 in only one AWC (Lee et al., 2010). Thus,
reducing the levels of ODR-1 in the butanone responsive neuron
should not affect chemotaxis mediated by the other AWC.
Furthermore, each odor requires different factors for adaptation
(Colbert and Bargmann, 1995), and thus, each response may
have unique sensitivity to the level of ODR-1.
The Nuclear RNAi Pathway Acts with HP1 in Odor
We foundthat the nuclear Argonaute NRDE-3is required in AWC
for odor adaptation, and it binds odr-1 siRNA in an odor-
dependent fashion. Prior work showed that NRDE-3 acts in the
nucleus along with NRDE-2, NRDE-1, and NRDE-4 to establish
H3K9me3 marks on the target locus, thereby silencing transcrip-
tion (Burkhart et al., 2011; Burton et al., 2011; Gu et al., 2012;
Guang et al., 2010). This connection between endo-siRNA,
H3K9me3 marks, and gene silencing was originally reported in
S. pombe, in which silencing involves deposition of H3K9me3
marks directed by siRNAs produced from pericentromeric
repeat regions and the mating type locus (Aygu ¨n and Grewal,
2010). In pombe, these siRNAs induce a transcriptional silencing
complex (RITS) that localizes chromatin to specific nascent
transcripts. A feed-forward silencing loop is established as chro-
modomain proteins, including the HP1 homolog, and methyl-
more methyltransferases. Concurrently, RNA-dependent RNA
polymerase complexes (RDRCs) are recruited, thus increasing
siRNA production (Hayashi et al., 2012; Rougemaille et al.,
2012; Yamanaka et al., 2013). A direct link between chromatin,
RNAi, and RITS was demonstrated when the CSD of pombe
HP1 was shown to interact with several members of the RNAi
and RITS machinery via the HP1-binding protein, Ers1 (Rouge-
maille et al., 2012). Because Ers1 interacted specifically with
the CSD of yeast HP1, and we show that in C. elegans, phos-
phorylation of this domain is sufficient to induce adaptation,
we speculate that the C. elegans HPL-2 CSD likewise nucleates
RNAi factors on genes such as odr-1 whose silencing promotes
adaptation. Indeed, because loss of mut-7 suppressed the gain-
of-function HPL-2(S155E), MUT-7 may either act along with or
Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc. 1017
Figure 4. Phosphorylation of HPL-2 and MUT-7 at Predicted PKG Sites Is Required for Adaptation
(A) Schematic of HPL-2 and MUT-7. (Top) HPL-2 contains an N-terminal chromodomain, a C-terminal chromo shadow domain, and four predicted PKG
phosphorylation sites. (Bottom) MUT-7 contains two predicted functional domains—cytosolic fatty acid binding domain and 30to 50exonuclease—and seven
predicted PKG phosphorylation sites.
(B) Phosphorylation of predicted PKG target sites in MUT-7 is required for adaptation. Mean CIs of wild-type or mut-7(pk204) strains expressing the indicated
form of MUT-7 in AWC. Figure S3D shows individual lines. n = 3 and p value is from a two-tailed Student’s t test. The lines rescued the sterility defects of
mut-7(pk204) (Figure S3B).
(legend continued on next page)
1018 Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc.
downstream of activated HPL-2. Thus, our data are consistent
withHPL-2 being recruited
H3K9me3 marks and perhaps also nucleating an RNAi-based
to siRNA-targetedloci by
Chromatin Marks in Behavior
HPL-2 loads onto the odr-1 locus in odor-adapted AWCs. This
may reflect deposition of a heterochromatic mark. Such marks
have been implicated in both neuronal development, as well as
in stimulus-induced changes in behavior. H3K9me3-mediated
silencing of all but the active olfactory receptor allows for mono-
allelic expression of odor receptors in the mammalian nasal
epithelium (Magklara et al., 2011). In rodents, behavioral addic-
tion to cocaine has been shown to increase H3K9me2 marks
within a key brain reward region (Renthal et al., 2009), and regu-
lation of H3K9 methylation is important for addiction-induced
neuroplasticity (Maze et al., 2010). These studies highlight the
importance of histone methylation marks in regulating long-
marks to specific locations could be a key regulated process. It
via the action of endo-siRNAs.
Evidence that mammals have a dicer-dependent class of
22GRNAs is currently lacking (Babiarz et al., 2011). In
S. pombe, however, siRNA species derived from protein-coding
genes were not identified until nuclear exosome deficient cells
were used (Yamanaka et al., 2013). Such degradation processes
might also conceal endo-siRNAs in higher eukaryotes. Though
no RNA-dependent RNA polymerase has yet been identified in
mammals, it is possible that other classes of small RNAs such
as mitrons (miRNAs processed from introns) play an analogous
function in the mammalian brain or that RNA polymerase I, II,
or III might be recruited to produce small antisense RNAs
(Filipovska and Konarska, 2000; Lehmann et al., 2007; Greco-
Stewart et al., 2009). These RNAs could similarly direct deposi-
tion of chromatin marks and affect behavior.
Odor Regulates Chromatin Changes
Our workindicates thatanenvironmental signal islikely to actvia
a kinase to amplify the small RNA-directed process. Kinases
have been widely appreciated to effect behavioral responses:
mitogen-activated protein kinases, calcium calmodulin-depen-
dent protein kinase II, protein kinase C, and protein kinase A
can each contribute to the formation of long-term memory sub-
sequent to repeated training (Dash et al., 2007; Gerstner et al.,
2009). Indeed, EGL-4 acts via a histone deacetylase (HDA class
I) in the nucleus of uterine epithelium cells to promote egg laying
(Hao et al., 2011). Here, we demonstrate that the HP1 homolog,
HPL-2, is a direct target of this odor-dependent kinase.
In both yeast and mammals, HP1 phosphorylation has been
shown to regulate HP1’s repressive activity in response to inter-
and intracellular signals. Although many studies highlighted the
important role played by modifications of the CD (Shimada and
Murakami, 2010), our observations suggest that modifications
of the CSD may be equally important. The CSD serves as a
2 cDNA in AWC are shown. n > 3 with >100 individuals per assay. p value is from an unpaired Student’s t test.
(D) Nuclear EGL-4 phosphorylates HPL-2 in vitro. (Left) 3XFLAG-nuclear EGL-4 kinase was immunopurified from worms (behavior in Figure S3F) and incubated
with purified HPL-2 and32P ATP. The reactions were resolved on a gel and stained with Coomassie blue as loading control (lower) followed by autoradiography
(upper). (Right) Quantification of five independent kinase assays.32P phosphorylated HPL-2 was normalized to Coomassie stained band. Values shown for
mutant HPL-2 substrate are shown as fold reduction of phosphorylation relative to HPL2-wild-type, which was set to 1. Error bars represent mean ± SEM (p <
0.0001; two-tailed Student’s t test, n = 5).
(E) Phosphorylation of HPL-2 at a predicted PKG phosphorylation site in the CSD is sufficient to decrease butanone chemotaxis in naive animals. CIs of wild-type
animals expressing the indicated form of HPL-2 in AWC, n > 3. Figure S3E shows CIs of individual lines. All strains expressed similar levels of the indicated
transgenes as assessed by GFP intensity. p values are from a two-tailed Student’s t test.
(F) Phosphorylation of HPL-2 at the EGL-4 phosphorylated sites is sufficient to promote adaptation in naive animals. CIs of naive wild-type, osm-9, or mut-7
animals either expressing the phosphomemetic HPL-2(S/Tto E) (+) or not (?). In all panels of this figure, the bars represent the mean values, and the error bars
represent SEM; n > 5. p value is from an unpaired two-tailed Student’s t test.
Figure 5. Prolonged Stimulation Induces Long-Term Olfactory
Adaptation in the AWC Neurons via an siRNA and Chromatin-
Model for how prolonged butanone stimulation may lead to long-lasting
olfactory adaptation in the AWC neuron. Asterisks indicate processes and
factors shown to act in AWC. Odor exposure stimulates a seven trans-
7 (dashed arrow). Phosphorylated HPL-2 promotes adaptation in a 22GRNA
dependent process by binding to H3K9me3 (yellow flags). Phosphorylated
NRDE-3 Ago complex to direct H3K9me3 to odr-1 gene. Phosphorylated
lower levels of odr-1 mRNA decreases the animal’s attraction to butanone.
Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc. 1019
platform for the assembly of other chromatin (Couteau et al.,
2002) and RNAi (Rougemaille et al., 2012) -associated proteins
and may therefore represent an attractive target for dynamic
regulation of transcriptional states. The CSD is required for
HP1 homodimerization and formation of an interaction platform
with proteins containing the PxVxL interaction motif (Cowieson
et al., 2000; Thiru et al., 2004). Though basal silencing requires
phosphorylation of the CSD (Zhao et al., 2001), our data indicate
that CSD phosphorylation may also be used for signal respon-
sive silencing in neurons.
Other kinases may act in a similar fashion to EGL-4 in other
cells and organisms to allow developmental or environmental
signals to enhance small-RNA-dependent gene silencing. By
regulating RITS, all siRNA-producing loci could be coordinately
silenced at a point in time, and the ensuing chromatin changes
would ensure stable silencing. Such widespread silencing by
siRNAs may allow experiences to alter expression of whole
cohorts of genes in the context of both development and
For a complete list of strains used, please see the Supplemental Information.
Bristol N2 was the wild-type strain.
Plasmid Construction and Transgenic Strains
Details of plasmid construction can be found in the Supplemental Information.
Behavioral assays were conducted on day one adults as described (Colbert
and Bargmann, 1995). More details are presented in the Supplemental Infor-
mation. For heat shock experiments, worms on their original growth plates
were exposed to 30?C for 1 hr and then recovered at 20?C for 2 hr prior to
Kinase Assay with Nuclear EGL-4
To evaluate nuclear NLS FLAG-EGL-4 kinase activity, 100 mg of worm lysate
was immunoprecipitated using anti-FLAG M2 magnetic beads (Sigma-
Aldrich). Bead-bound immunoprecipitates were washed extensively with
kinase buffer. Then kinase assays were performed directly on the beads by
adding 1.5 mg of substrate (HPL-2 WT, HPL-2[all S/T-A], or Histone H1),
2 mCi32P ATP (PerkinElmer), and 25 mM cGMP (Sigma-Aldrich).
Details are in the Supplemental Information.
Isolation of NRDE-3-Associated Small RNA
50–60 plates of adult animals expressing 3XFLAG::GFP::NRDE-3 were
collected, and half the population was incubated with SBasal alone, and the
other half was incubated with SBasal plus butanone for 80 min. Behavior of
?100 animals from each was assessed. Extracts were made from the remain-
ing animals as described (Guang et al., 2008 and Supplemental Information).
Isolation of HPL-2-Associated DNA
podr-3::3XFLAG::GFP::HPL-2 was integrated into the genome and out-
crossed five times. 100 plates of adult animals were harvested, and half
were exposed to buffer and half to butanone and buffer. ?100 animals from
each were assayed. The remaining animals were processed for ChIP (Gerstein
et al., 2010). Only populations that showed an adapted CI of 0.05–0.3 were
used. Details of the ChIP are in the Supplemental Information. Quantitation
of coimmunoprecipitated DNA is described in the Supplemental Information.
Quantitative Real-Time PCR
For RNA analysis, 5 plates of day one adult animals were collected and treated
to the adaptation protocol, and their behavior was assessed. Total RNA was
isolated as described in the Supplemental Information. Total RNA from entire
worms was used in 22GRNA and mRNA quantitation as described in Supple-
To quantify HPL-2-associated DNA, ChIP results were analyzed by
qPCR using Brilliant III Ultra-Fast SYBR Green qPCR Master Mix (Agilent
housekeeping gene act-3 did not change with odor. Please see Extended
Experimental Procedures and Table S2 for details and primers.
Supplemental Information includes Extended Experimental Procedures, three
figures,andtwotablesand canbefoundwiththisarticleonline athttp://dx.doi.
We thank Maria Gallegos, Mehrdad Matloubian, Bassem Al-Sady, Jonathan
Isaiah Gent, Georgia Woods, and Robert Blelloch; members of the L’Etoile
lab: Damien O’Halloran, Scott Hamilton, Sarah Gerhart, Mary Bethke, and
Chantal Brueggemann; and members of the Kennedy lab: Bethany Buckley
and Kirk Burkhart, for critical reading and helpful discussions. We thank Shi-
Yu Chen for purified MUT-7 and HPL-2. We thank Ahmed Elewa and Craig
Mello for good discussions, worm strains, and odr-1 22G RNA sequences;
Ebenezer Yamoah for use of his qPCR system; and Kohta Ikegami and Susan
Strome for ChIP protocols and advice. We also thank Nadia Gronachon, Kim
Collins, Christopher Morales, Anu Gupta, and Aarati Asundi for help with
worm culture; Bethany Buckley, the Caenorhabditis Genetics Center and the
National Bioresource Project for numerous strains, and Yuji Kohara for yk
cDNA clones. We thank Leah Frater (Anderson lab), Derek Pavelec (Kennedy
lab), Jay Maniar, and Jonathan Gent (Fire lab) for producing and sharing
unpublished sequencing reads. N.D.L. acknowledges support from NSF
0954258 and NIH R01DC005991. B.-T.J. conceived of and carried out
experiments and cowrote the manuscript. L.S. developed, optimized, and
performed kinase assays. CG performed RNA and DNA analysis. L.S. and
C.G. contributed equally to the manuscript. A.G., F.P., and S.K. provided
unpublished protocols, reagents, support and insight. N.L. analyzed data
and wrote the manuscript.
Received: April 7, 2013
Revised: July 16, 2013
Accepted: August 1, 2013
Published: August 29, 2013
Aoki, K., Moriguchi, H., Yoshioka, T., Okawa, K., and Tabara, H. (2007). In vitro
analyses of the production and activity of secondary small interfering RNAs in
C. elegans. EMBO J. 26, 5007–5019.
Ashe, A., Sapetschnig, A., Weick, E.M., Mitchell, J., Bagijn, M.P., Cording,
A.C., Doebley, A.L., Goldstein, L.D., Lehrbach, N.J., Le Pen, J., et al. (2012).
piRNAs can trigger a multigenerational epigenetic memory in the germline of
C. elegans. Cell 150, 88–99.
Aygu ¨n,O.,andGrewal, S.I.(2010).Assemblyandfunctionsofheterochromatin
Babiarz, J.E., Hsu, R., Melton, C., Thomas, M., Ullian, E.M., and Blelloch, R.
(2011). A role for noncanonical microRNAs in the mammalian brain revealed
by phenotypic differences in Dgcr8 versus Dicer1 knockouts and small RNA
sequencing. RNA 17, 1489–1501.
Bargmann, C.I., Hartwieg, E., and Horvitz, H.R. (1993). Odorant-selective
genes and neurons mediate olfaction in C. elegans. Cell 74, 515–527.
1020 Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc.
Black, J.C., and Whetstine, J.R. (2011). Chromatin landscape: methylation
beyond transcription. Epigenetics 6, 9–15.
Buckley, B.A., Burkhart, K.B., Gu, S.G., Spracklin, G., Kershner, A., Fritz, H.,
Kimble, J., Fire, A., and Kennedy, S. (2012). A nuclear Argonaute promotes
multigenerational epigenetic inheritance and germline immortality. Nature
Burkhart, K.B., Guang, S., Buckley, B.A., Wong, L., Bochner, A.F., and
Kennedy, S. (2011). A pre-mRNA-associating factor links endogenous
siRNAs to chromatin regulation. PLoS Genet. 7, e1002249.
Burton, N.O., Burkhart, K.B., and Kennedy, S. (2011). Nuclear RNAi maintains
heritable gene silencing in Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA
108, 19683–19688. http://dx.doi.org/10.1073/pnas.1113310108.
Claycomb, J.M., Batista, P.J., Pang, K.M., Gu, W., Vasale, J.J., van Wolfswin-
kel, J.C., Chaves, D.A., Shirayama, M., Mitani, S., Ketting, R.F., et al. (2009).
The Argonaute CSR-1 and its 22G-RNA cofactors are required for holocentric
chromosome segregation. Cell 139, 123–134.
Colbert, H.A., and Bargmann, C.I. (1995). Odorant-specific adaptation path-
ways generate olfactory plasticity in C. elegans. Neuron 14, 803–812.
Colbert, H.A., and Bargmann, C.I. (1997). Environmental signals modulate
olfactory acuity, discrimination, and memory in Caenorhabditis elegans.
Learn. Mem. 4, 179–191.
Colon-Ramos, D.A., Margeta, M.A., and Shen, K. (2007). Glia promote local
synaptogenesis through UNC-6 (netrin) signaling in C. elegans. Science 318,
Coustham, V., Bedet, C., Monier, K., Schott, S., Karali, M., and Palladino, F.
(2006). The C. elegans HP1 homologue HPL-2 and the LIN-13 zinc finger pro-
teinformacompleximplicatedinvulval development. Dev.Biol.297,308–322.
Couteau, F., Guerry, F., Muller, F., and Palladino, F. (2002). A heterochromatin
protein 1 homologue in Caenorhabditis elegans acts in germline and vulval
development. EMBO Rep. 3, 235–241.
Cowieson, N.P., Partridge, J.F., Allshire, R.C., and McLaughlin, P.J. (2000).
Dimerisation of a chromo shadow domain and distinctions from the chromo-
domain as revealed by structural analysis. Curr. Biol. 10, 517–525.
Dash, P.K., Moore, A.N., Kobori, N., and Runyan, J.D. (2007). Molecular activ-
ity underlying working memory. Learn. Mem. 14, 554–563.
Denli, A.M., Tops, B.B., Plasterk, R.H., Ketting, R.F., and Hannon, G.J. (2004).
Processing of primary microRNAs by the Microprocessor complex. Nature
Dernburg, A.F., Zalevsky, J., Colaia ´covo, M.P., and Villeneuve, A.M. (2000).
Transgene-mediated cosuppression in the C. elegans germ line. Genes Dev.
Duchaine, T.F., Wohlschlegel, J.A., Kennedy, S., Bei, Y., Conte, D., Jr., Pang,
K., Brownell, D.R., Harding, S., Mitani, S., Ruvkun, G., et al. (2006). Functional
proteomics reveals the biochemical niche of C. elegans DCR-1 in multiple
small-RNA-mediated pathways. Cell 124, 343–354.
Etchberger, J.F., Lorch, A., Sleumer, M.C., Zapf, R., Jones, S.J., Marra, M.A.,
Holt, R.A.,Moerman, D.G., and Hobert, O.(2007). Themolecular signatureand
cis-regulatory architecture of a C. elegans gustatory neuron. Genes Dev. 21,
Filipovska, J., and Konarska, M.M. (2000). Specific HDV RNA-templated tran-
scription by pol II in vitro. RNA 6, 41–54.
Fire, A., Xu, S., Montgomery, M.K., Kostas, S.A., Driver, S.E., and Mello, C.C.
(1998). Potent and specific genetic interference by double-stranded RNA in
Caenorhabditis elegans. Nature 391, 806–811.
Su, S.C., and Tsai, L.H. (2010). A novel pathway regulates memory and plas-
ticity via SIRT1 and miR-134. Nature 466, 1105–1109.
Gent, J.I., Lamm, A., Maniar, J.M., Parameswaran, P., Tao, L., and Fire, A.Z.
(2010). Distinct stages of siRNA synthesis in an endogenous RNAi pathway
in C. elegans soma. Mol Cell 37, 679–689.
Gerstein, M.B., Lu, Z.J., Van Nostrand, E.L., Cheng, C., Arshinoff, B.I., Liu, T.,
Yip, K.Y., Robilotto, R., Rechtsteiner, A., Ikegami, K., et al.; modENCODE
Consortium. (2010). Integrative analysis of the Caenorhabditis elegans
genome by the modENCODE project. Science 330, 1775–1787.
Gerstner, J.R., Lyons, L.C., Wright, K.P., Jr., Loh, D.H., Rawashdeh, O.,
Eckel-Mahan, K.L., and Roman, G.W. (2009). Cycling behavior and memory
formation. J. Neurosci. 29, 12824–12830.
Ghildiyal, M., and Zamore, P.D. (2009). Small silencing RNAs: an expanding
universe. Nat. Rev. Genet. 10, 94–108.
Greco-Stewart, V.S., Schissel, E., and Pelchat, M. (2009). The hepatitis delta
virus RNA genome interacts with the human RNA polymerases I and III.
Virology 386, 12–15.
Grishok, A., Pasquinelli, A.E., Conte, D., Li, N., Parrish, S., Ha, I., Baillie, D.L.,
Fire, A., Ruvkun, G., and Mello, C.C. (2001). Genes and mechanisms related to
RNA interference regulate expression of the small temporal RNAs that control
C. elegans developmental timing. Cell 106, 23–34.
Grishok, A.,Sinskey,J.L., and Sharp, P.A. (2005). Transcriptional silencing of a
transgene by RNAi in the soma of C. elegans. Genes Dev. 19, 683–696.
Moresco, J.J., Youngman, E.M., Keys, J., Stoltz, M.J., et al. (2009). Distinct
argonaute-mediated 22G-RNA pathways direct genome surveillance in the
C. elegans germline. Mol. Cell 36, 231–244.
Gu, S.G., Pak, J., Guang, S., Maniar, J.M., Kennedy, S., and Fire, A. (2012).
Amplification of siRNA in Caenorhabditis elegans generates a transgenera-
Guang, S., Bochner, A.F., Pavelec, D.M., Burkhart, K.B., Harding, S., Lacho-
wiec, J., and Kennedy, S. (2008). An Argonaute transports siRNAs from the
cytoplasm to the nucleus. Science 321, 537–541.
Guang, S., Bochner, A.F., Burkhart, K.B., Burton, N., Pavelec, D.M., and
Kennedy, S. (2010). Small regulatory RNAs inhibit RNA polymerase II during
the elongation phase of transcription. Nature 465, 1097–1101.
Hao, Y., Xu, N., Box, A.C., Schaefer, L., Kannan, K., Zhang, Y., Florens, L.,
Seidel, C., Washburn, M.P., Wiegraebe, W., and Mak, H.Y. (2011). Nuclear
cGMP-dependent kinase regulates gene expression via activity-dependent
recruitment of a conserved histone deacetylase complex. PLoS Genet. 7,
Hayashi, A., Ishida, M., Kawaguchi, R., Urano, T., Murakami, Y., and
Nakayama, J. (2012). Heterochromatin protein 1 homologue Swi6 acts in con-
cert with Ers1 to regulate RNAi-directed heterochromatin assembly. Proc.
Natl. Acad. Sci. USA 109, 6159–6164.
Hedgecock, E.M., Culotti, J.G., and Hall, D.H. (1990). The unc-5, unc-6, and
unc-40 genes guide circumferential migrations of pioneer axons and
mesodermal cells on the epidermis in C. elegans. Neuron 4, 61–85.
Kaye, J.A., Rose, N.C., Goldsworthy, B., Goga, A., and L’Etoile, N.D. (2009). A
3’UTR pumilio-binding element directs translational activation in olfactory
sensory neurons. Neuron 61, 57–70.
Kennedy, S., Wang, D., and Ruvkun, G. (2004). A conserved siRNA-degrading
RNase negatively regulates RNA interference in C. elegans. Nature 427,
Ketting, R.F. (2011). The many faces of RNAi. Dev. Cell 20, 148–161.
Ketting, R.F., Haverkamp, T.H., van Luenen, H.G., and Plasterk, R.H. (1999).
Mut-7 of C. elegans, required for transposon silencing and RNA interference,
is a homolog of Werner syndrome helicase and RNaseD. Cell 99, 133–141.
Knight, S.W., and Bass, B.L. (2001). A role for the RNase III enzyme DCR-1 in
RNA interference and germ line development in Caenorhabditis elegans.
Science 293, 2269–2271.
L’Etoile, N.D., and Bargmann, C.I. (2000). Olfaction and odor discrimination
are mediated by the C. elegans guanylyl cyclase ODR-1. Neuron 25, 575–586.
L’Etoile, N.D., Coburn, C.M., Eastham, J., Kistler, A., Gallegos, G., and
Bargmann, C.I. (2002). The cyclic GMP-dependent protein kinase EGL-4
regulates olfactory adaptation in C. elegans. Neuron 36, 1079–1089.
Le Thomas, A., Rogers, A.K., Webster, A., Marinov, G.K., Liao, S.E., Perkins,
E.M., Hur, J.K., Aravin, A.A., and To ´th, K.F. (2013). Piwi induces piRNA-guided
Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc. 1021
transcriptional silencing and establishment of a repressive chromatin state.
Genes Dev. 27, 390–399.
Lee, R.C., Hammell, C.M., and Ambros, V. (2006). Interacting endogenous and
exogenous RNAi pathways in Caenorhabditis elegans. RNA 12, 589–597.
Lee, J.I., O’Halloran, D.M., Eastham-Anderson, J., Juang, B.T., Kaye, J.A.,
Scott Hamilton, O., Lesch, B., Goga, A., and L’Etoile, N.D. (2010). Nuclear en-
try of a cGMP-dependent kinase converts transient into long-lasting olfactory
adaptation. Proc. Natl. Acad. Sci. USA 107, 6016–6021.
Lee, H.C., Gu, W., Shirayama, M., Youngman, E., Conte, D., Jr., and Mello,
C.C. (2012). C. elegans piRNAs mediate the genome-wide surveillance of
germline transcripts. Cell 150, 78–87.
Lehmann, E., Brueckner, F., and Cramer, P. (2007). Molecular basis of RNA-
dependent RNA polymerase II activity. Nature 450, 445–449.
Maciejewski, P.M., Peterson, F.C., Anderson, P.J., and Brooks, C.L. (1995).
Mutation of serine 90 to glutamic acid mimics phosphorylation of bovine pro-
lactin. J. Biol. Chem. 270, 27661–27665.
Magklara, A., Yen, A., Colquitt, B.M., Clowney, E.J., Allen, W., Markenscoff-
Papadimitriou, E., Evans, Z.A., Kheradpour, P., Mountoufaris, G., Carey, C.,
et al. (2011).An epigenetic signature for monoallelic olfactory receptor expres-
sion. Cell 145, 555–570.
Mansour, S.J., Matten, W.T., Hermann, A.S., Candia, J.M., Rong, S., Fuka-
sawa, K., Vande Woude, G.F., and Ahn, N.G. (1994). Transformation of
mammalian cells by constitutively active MAP kinase kinase. Science 265,
Maze, I., Covington, H.E., 3rd, Dietz, D.M., LaPlant, Q., Renthal, W., Russo,
S.J., Mechanic, M., Mouzon, E., Neve, R.L., Haggarty, S.J., et al. (2010).
Essential role of the histone methyltransferase G9a in cocaine-induced plas-
ticity. Science 327, 213–216.
McNeill, E., and Van Vactor, D. (2012). MicroRNAs shape the neuronal
landscape. Neuron 75, 363–379.
O’Halloran, D.M., Altshuler-Keylin, S., Lee, J.I., and L’Etoile, N.D. (2009).
Regulators of AWC-mediated olfactory plasticity in Caenorhabditis elegans.
PLoS Genet. 5, e1000761.
O’Halloran, D.M., Hamilton, O.S., Lee, J.I., Gallegos, M., and L’Etoile, N.D.
(2012). Changes in cGMP levels affect the localization of EGL-4 in AWC in
Caenorhabditis elegans. PLoS ONE 7, e31614.
Pak, J., Maniar, J.M., Mello, C.C., and Fire, A. (2012). Protection from feed-
forward amplification in an amplified RNAi mechanism. Cell 151, 885–899.
Rajasethupathy, P., Antonov, I., Sheridan, R., Frey, S., Sander, C., Tuschl, T.,
and Kandel, E.R. (2012). A role for neuronal piRNAs in the epigenetic control of
memory-related synaptic plasticity. Cell 149, 693–707.
Renthal, W., Kumar, A., Xiao, G., Wilkinson, M., Covington, H.E., 3rd, Maze, I.,
Sikder, D., Robison, A.J., LaPlant, Q., Dietz, D.M., et al. (2009). Genome-wide
analysis of chromatin regulation by cocaine reveals a role for sirtuins. Neuron
Rougemaille, M., Braun, S., Coyle, S., Dumesic, P.A., Garcia, J.F., Isaac, R.S.,
Natl. Acad. Sci. USA 109, 11258–11263.
Shimada, A., and Murakami, Y. (2010). Dynamic regulation of heterochromatin
function via phosphorylation of HP1-family proteins. Epigenetics 5, 30–33.
Shirayama, M., Seth, M., Lee, H.C., Gu, W., Ishidate, T., Conte, D., Jr., and
Mello, C.C. (2012). piRNAs initiate an epigenetic memory of nonself RNA in
the C. elegans germline. Cell 150, 65–77.
Simmer, F., Tijsterman, M., Parrish, S., Koushika, S.P., Nonet, M.L., Fire, A.,
Ahringer, J., and Plasterk, R.H. (2002). Loss of the putative RNA-directed
RNA polymerase RRF-3 makes C. elegans hypersensitive to RNAi. Curr.
Biol. 12, 1317–1319.
Stringham, E.G., Dixon, D.K., Jones, D., and Candido, E.P. (1992). Temporal
and spatial expression patterns of the small heat shock (hsp16) genes in trans-
genic Caenorhabditis elegans. Mol. Biol. Cell 3, 221–233.
Tabara, H., Sarkissian, M., Kelly, W.G., Fleenor, J., Grishok, A., Timmons, L.,
Fire, A., and Mello, C.C. (1999). The rde-1 gene, RNA interference, and trans-
poson silencing in C. elegans. Cell 99, 123–132.
Tabara, H., Yigit, E., Siomi, H., and Mello, C.C. (2002). The dsRNA binding
protein RDE-4 interacts with RDE-1, DCR-1, and a DExH-box helicase to
direct RNAi in C. elegans. Cell 109, 861–871.
Thiru, A., Nietlispach, D., Mott, H.R., Okuwaki, M., Lyon, D., Nielsen, P.R.,
Hirshberg, M., Verreault, A., Murzina, N.V., and Laue, E.D. (2004). Structural
basis of HP1/PXVXL motif peptide interactions and HP1 localisation to hetero-
chromatin. EMBO J. 23, 489–499.
Tops, B.B., Tabara, H., Sijen, T., Simmer, F., Mello, C.C., Plasterk, R.H., and
Ketting, R.F. (2005). RDE-2 interacts with MUT-7 to mediate RNA interference
in Caenorhabditis elegans. Nucleic Acids Res. 33, 347–355.
Vasquez-Rifo, A., Jannot, G., Armisen, J., Labouesse, M., Bukhari, S.I.A.,
Rondeau, E.L., Miska, E.A., and Simard, M.J. (2012). Developmental charac-
terization of the microRNA-specific C. elegans Argonautes alg-1 and alg-2.
PLoS ONE 7, e33750. http://dx.doi.org/10.1371/journal.pone.0033750.
Vastenhouw, N.L., Brunschwig, K., Okihara, K.L., Mu ¨ller, F., Tijsterman, M.,
and Plasterk, R.H. (2006). Gene expression: long-term gene silencing by
RNAi. Nature 442, 882.
Wes, P.D., and Bargmann, C.I. (2001). C. elegans odour discrimination
requires asymmetric diversity in olfactory neurons. Nature 410, 698–701.
Yamanaka, S., Mehta, S., Reyes-Turcu, F.E., Zhuang, F., Fuchs, R.T., Rong,
Y., Robb, G.B., and Grewal, S.I. (2013). RNAi triggered by specialized
machinery silences developmental genes and retrotransposons. Nature 493,
Yigit, E., Batista, P.J., Bei, Y., Pang, K.M., Chen, C.C., Tolia, N.H., Joshua-Tor,
L., Mitani, S., Simard, M.J., and Mello, C.C. (2006). Analysis of the C. elegans
Cell 127, 747–757.
Zhang, C., Montgomery, T.A., Gabel, H.W., Fischer, S.E.,Phillips, C.M., Fahlg-
ren, N., Sullivan, C.M., Carrington, J.C., and Ruvkun, G. (2011). mut-16 and
other mutator class genes modulate 22G and 26G siRNA pathways in Caeno-
rhabditis elegans. Proc. Natl. Acad. Sci. USA 108, 1201–1208. http://dx.doi.
Zhao, T., Heyduk, T., and Eissenberg, J.C. (2001). Phosphorylation site muta-
tions in heterochromatin protein 1 (HP1) reduce or eliminate silencing activity.
J. Biol. Chem. 276, 9512–9518.
1022 Cell 154, 1010–1022, August 29, 2013 ª2013 Elsevier Inc.