The Ankyrin Repeat Domain of the TRPA Protein Painless Is Important for Thermal Nociception but Not Mechanical Nociception

Abstract

The Drosophila TRPA channel Painless is required for the function of polymodal nociceptors which detect noxious heat and noxious mechanical stimuli. These functions of Painless are reminiscent of mammalian TRPA channels that have also been implicated in thermal and mechanical nociception. A popular hypothesis to explain the mechanosensory functions of certain TRP channels proposes that a string of ankyrin repeats at the amino termini of these channels acts as an intracellular spring that senses force. Here, we describe the identification of two previously unknown Painless protein isoforms which have fewer ankyrin repeats than the canonical Painless protein. We show that one of these Painless isoforms, that essentially lacks ankyrin repeats, is sufficient to rescue mechanical nociception phenotypes of painless mutant animals but does not rescue thermal nociception phenotypes. In contrast, canonical Painless, which contains Ankyrin repeats, is sufficient to largely rescue thermal nociception but is not capable of rescuing mechanical nociception. Thus, we propose that in the case of Painless, ankryin repeats are important for thermal nociception but not for mechanical nociception.

Full text

The Ankyrin Repeat Domain of the TRPA Protein Painless
Is Important for Thermal Nociception but Not Mechanical
Nociception
Richard Y. Hwang
3
, Nancy A. Stearns
1
, W. Daniel Tracey
1,2,3
*
1Department of Anesthesiology, Duke University Medical Center, Durham, North Carolina, United States of America, 2Department of Cell Biology, Duke University
Medical Center, Durham, North Carolina, United States of America, 3Department of Neurobiology, Duke University Medical Center, Durham, North Carolina, United States
of America
Abstract
The Drosophila TRPA channel Painless is required for the function of polymodal nociceptors which detect noxious heat and
noxious mechanical stimuli. These functions of Painless are reminiscent of mammalian TRPA channels that have also been
implicated in thermal and mechanical nociception. A popular hypothesis to explain the mechanosensory functions of
certain TRP channels proposes that a string of ankyrin repeats at the amino termini of these channels acts as an intracellular
spring that senses force. Here, we describe the identification of two previously unknown Painless protein isoforms which
have fewer ankyrin repeats than the canonical Painless protein. We show that one of these Painless isoforms, that essentially
lacks ankyrin repeats, is sufficient to rescue mechanical nociception phenotypes of painless mutant animals but does not
rescue thermal nociception phenotypes. In contrast, canonical Painless, which contains Ankyrin repeats, is sufficient to
largely rescue thermal nociception but is not capable of rescuing mechanical nociception. Thus, we propose that in the case
of Painless, ankryin repeats are important for thermal nociception but not for mechanical nociception.
Citation: Hwang RY, Stearns NA, Tracey WD (2012) The Ankyrin Repeat Domain of the TRPA Protein Painless Is Important for Thermal Nociception but Not
Mechanical Nociception. PLoS ONE 7(1): e30090. doi:10.1371/journal.pone.0030090
Editor: Michael N. Nitabach, Yale School of Medicine, United States of America
Received September 2, 2011; Accepted December 13, 2011; Published January 25, 2012
Copyright: ß2012 Hwang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by grants to WDT from the Whitehall Foundation, the Alfred P. Sloan Foundation, and the National Institute of Neurological
Disorders and Stroke (5R01NS054899-05). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the
manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: dan.tracey@duke.edu
Introduction
Transient receptor potential (TRP) channels are important in
the senses of vision, taste, touch, hearing, nociception (mechanical,
thermal, and chemical), and thermosensation (reviewed in [1]).
painless encodes a Drosophila TRPA channel that is required for
both thermal and mechanical nociception in Drosophila [2–5]. In
addition, gustatory avoidance of isothiocyanate compounds, and
gravity perception in adult flies require painless. The molecular
mechanisms that allow painless to have polymodal sensory roles
that include thermosensory, chemosensory, and mechanical
signaling are not yet understood.
The nociception function of painless was initially found through
the investigation of nocifensive escape locomotion (NEL) behavior
that is seen in larvae exposed to noxious heat [4]. When
performing NEL, larvae rotate around the anterior posterior axis
in a corkscrew-like manner. painless mutant larvae show increased
sensory thresholds for nocifensive responses to noxious heat as well
as responses to noxious mechanical stimuli. Evidence suggests that
the Painless channel is a direct sensor of noxious heat. For
example, electrophysiological recordings from painless mutant
larval abdominal thermosensory neurons showed decreased firing
in response to noxious temperatures [4] and studies in heterolo-
gous expression systems have shown that Painless is a heat
activated thermoTRP with a threshold of approximately 39–42uC
[3]. This in vitro heat activation threshold for Painless is similar to
the 39–41uC behavioral threshold for triggering larval NEL. The
nociceptive function for Painless is likely mediated by nociceptive
Class IV multidendritic (mdIV) neurons which express the gene
and also show strong increases in firing at temperatures .39–41uC
[6]. Evidence suggests that mdIV neurons are nociceptors since
optogenetic activation of the mdIV neurons is sufficient to trigger
NEL and blocking the synaptic output of the mdIV neurons shows
that these neurons are necessary for responses to heat and
mechanical stimulation [7]. In addition, the pickpocket gene is
specifically expressed in the mdIV neurons and it is required for
mechanical nociception [8]. Although Painless is expressed in all
multidendritic neurons, only the class IV neurons have been found
to be activated at high temperatures [6]. This latter finding
suggests that tissue specific factors are likely to influence Painless
activity.
With regard to mechanical nociception, the role of Painless is
more poorly understood. The NEL responses to noxious
mechanical stimulation have an increased threshold in painless
mutant animals relative to wild type larvae which indicates an in
vivo requirement for Painless in mechanical nociception responses
[4]. However, in vitro studies on Painless expressed in a human
embryonic kidney cell line failed to detect Painless dependent
currents following hypo/hypertonic stimulation or direct touch
with a glass pipette [3]. This failure to detect mechanical currents
in Painless expressing cells may indicate that Painless is not directly
mechanosensitive. An indirect role for Painless in mechanotrans-
PLoS ONE | www.plosone.org 1 January 2012 | Volume 7 | Issue 1 | e30090

duction could occur if Painless functions downstream of another
mechanotranduction signaling molecule in the mdIV neurons.
Consistent with this possibility, the Degenerin/Epithelial Sodium
Channel (DEG/ENaC) protein Pickpocket is required for
mechanical nociception in Drosophila larvae but it is not required
for thermal nociception [8]. In addition, Painless channel activity
is strongly affected by intracellular Ca
2+
which may provide a
potential molecular mechanism for activation downstream of
neuronal activity [3]. Thus, Painless may have a direct role in
thermosensation but it may function downstream of Pickpocket in
mechanical signaling pathways.
On the other hand, heterologous expression studies are difficult
to interpret since mechanosensative channels may require
specialized signaling components that may not be present in
heterologous cells [9]. This may be particularly true with
expression of an insect channel in mammalian cells since the
required components in a mammalian cell, even if present, may
not be capable of interacting with the channel. In the best-
understood mechanotransduction system of C. elegans, the
mechanotransduction complex relies on the pore forming DEG/
ENaCs MEC-4 and MEC-10, extracellular proteins (MEC-1,
MEC-5, and MEC-9) and additional intracellular components
[10]. Similarly, in Drosophila, the extracellular protein NompA, is
required for the connection between the mechanosensory
neuronal sensory endings and cuticle in bristle mechanotransduc-
tion [11]. The identification of roles for extracellular proteins has
led to a model of mechanosensation involving tethers that anchor
the mechanosensitive channel or complex to intracellular and/or
extracellular components that efficiently transmit force to the
mechanosensitive channel.
Given the polymodal role of Painless in thermal nociception and
in mechanical nociception we were interested in the possibility that
particular domains of the Painless protein might play an important
role in various aspects of signaling. Here, we report three naturally
occurring RNA variants of painless that are predicted to encode
Painless protein isoforms which vary in the length of the ankyrin
repeat containing N-terminal domain. We utilize these naturally
occurring variants as tools to investigate the functional properties
of the ankyrin repeat domain of Painless through isoform specific
rescue experiments in vivo. Our results indicate that the longest
isoform, which contains the entire N-terminal ankyrin repeat
domain, is sufficient to rescue thermal nociception of a painless
mutant but does not rescue mechanical nociception. In addition,
this long protein isoform shows efficient localization to dendrites
and axons of the multidendritic neurons. In contrast, the shortest
isoform, which essentially lacks an ankyrin repeat domain, is
capable of rescuing mechanical nociception without rescuing
thermal nociception. This short isoform does not efficiently
localize to dendrites and is expressed at much lower levels in
comparison to the long isoform. Combined, these results suggest
that the polymodal functions of Painless may be explained by
distinct protein variants that have modality specific properties.
Furthermore, our results suggest that the N-terminal domain of
Painless is required for thermal nociception, but we do not find
evidence that it is required for mechanical nociception.
Results
Distinct isoforms of Painless vary in the length of their
N-terminal domain
Three transcripts of painless are detectable by northern blot
analysis [12]. To identify these different transcripts of painless,we
performed 59Rapid Amplification of Complementary cDNA Ends
(59RACE). Three 59RACE products were cloned and sequenced
and determined to represent three distinct 59ends for painless
transcripts. One of the RACE products was identical to the
previously described 59end of painless while the remaining two
RACE products encoded novel 59ends. A search of the expressed
sequence tag (EST) database identified 59EST sequences that
exactly matched the 59ends of the painless RACE products. Only a
single 39EST sequence for painless transcripts is found in the
database and this 39end is identical to that of the canonical painless
transcript. Combined, these data are consistent with the existence
of three painless transcripts that differ in the 59sequences but which
share a common 39end. Conceptual translation of the novel
transcripts predicted proteins that differed in the translational start
site of their N-terminal domains. Based on the predicted molecular
weights of the isoforms (103 kD, 72 kD and 60 kD), we refer to the
predicted protein isoforms as Painless
p103
, Painless
p72
, and
Painless
p60
.
The longest (canonical) isoform, Painless
p103
, is predicted to
have 8 N-terminal ankyrin repeats, 6 transmembrane domains,
and an intracellular carboxy-terminal domain (CTD) [4]. Note
that four of the ankyrin repeats of Painless are somewhat
degenerate in amino acid sequence, and are thus not detected
by ankyrin repeat algorithms. The intracellular amino terminal
domain (NTD) which contains the ankyrin repeats consists of the
first 468 amino acids of Painless
p103
. The transcript encoding the
intermediate length isoform, Painless
p72
, is transcribed from the
same promoter as the transcript encoding Painless
p103
, but it
includes an alternately spliced second intron that results in the
utilization of an alternate ATG start codon (Figure 1). The result
of this splicing causes the NTD of Painless
p72
to lack the first 285
amino acids of Painless
p103
. The shortest isoform, Painless
p60
, uses
an alternate transcriptional start site that is downstream of the
transcriptional start site of the other two transcripts (Figure 1). This
structure causes the Painless
p60
protein to lack the first 385 amino
acids of Painless
p103
. The resulting intracellular NTD of
Painless
p60
is a relatively short 83 amino acids. Interestingly, the
alternate protein isoforms for Painless are similar in structure to N-
terminal protein variants of other TRP channels. For example,
there are two known protein isoforms of the Drosophila TRPA
channel Pyrexia: Pyx-A and Pyx-B. As in the case of Painless
p60
,
the Pyx-B protein is lacking the N-terminal ankyrin repeat
containing domain while the Pyx-B protein resembles Painless
p103
[13]. Similarly, the mammalian TRPV1 channel has a reported
isoform that lacks ankyrin repeats at its N-terminus [14].
Generation of transgenic flies expressing fluorescently
tagged Painless
p103
and Painless
p60
In order to test the functional roles of the Painless protein
isoforms in vivo, we generated transgenic files to specifically express
either the Painless
p103
isoform or the Painless
p60
isoform under
control of the GAL4/UAS system. In order to estimate expression
levels of the transgenes, an in-frame fusion of the Venus
Fluorescent Protein (VFP) was added at the C-terminus of both
constructs. We refer to these transgenes as UAS-painless
p103
::VFP
and UAS-painless
p60
::VFP. To examine the localization of the VFP
tagged Painless proteins, we crossed flies harboring UAS-
painless
p103
::VFP or UAS-painless
p60
::VFP to the painless
GAL4
driver
strain. In the case of UAS- painless
p103
::VFP the progeny of this cross
showed robust VFP fluorescence throughout the dendrites, cell
bodies, and axons of the painless
GAL4
expressing multidendritic
sensory neurons (Figure 2A–C). This fluorescence was easily
detectable in living animals using confocal microscopy (data not
shown). The expression levels produced from this transgene
exceeded that of the endogenous Painless protein as detected by
increased immunostaining with anti-Painless antibodies relative to
Function of Painless Ankyrin Domain
PLoS ONE | www.plosone.org 2 January 2012 | Volume 7 | Issue 1 | e30090

the wild type (data not shown). Indeed, wild type Painless protein
is localized to discrete punctae [4], while the over-expressed VFP-
Painless was present throughout the dendritic arbor. This wider
distribution is likely a consequence of relatively high expression
levels generated by the GAL4/UAS system.
In contrast, the VFP of UAS- painless
p60
::VFP was significantly
less intense. The VFP fluorescence from UAS- painless
p60
::VFP was
difficult to visualize in living larvae but was detectable by anti-GFP
immunostaining of fixed preparations (Figure 2 C–D). Interest-
ingly, the subcellular localization of Painless
p60
::VFP was distinct
from that of Painless
p103
::VFP. Although Painless
p60
::VFP was
easily detectable in the cell body the expression in dendrites and
axons was limited to the most proximal regions (Figure 2 D–F).
The subcellular localization and expression levels of the Pain-
less
p60
::VFP was observed in multiple independent UAS insertion
lines so it appears to be a property of the protein isoform as
opposed to position effects that might limit expression levels of a
particular UAS transgene insertion. These results suggest that the
N-terminal domain of Painless results in enhanced stability of the
Painless
p103
::VFP protein and more efficient localization to the
dendrites and axons of multidendritic neurons relative to the
Painless
p60
::VFP protein. Alternatively, the painless
p60
transcript
may be unstable, or poorly translated, relative to the painless
p103
transcript.
Pronounced thermal and mechanical nociception defects
in the pain
Gal4
/pain
NP7022
allelic combination
In order to study the potentially distinct roles of the
Painless
p103
::VFP and Painless
p60
::VFP isoforms, we performed
genetic rescue experiments in painless expressing tissues. To
achieve this, we took advantage of mutant alleles of painless that
express the yeast transcription factor GAL4 in painless expressing
cells. pain
Gal4
is one such mutant for painless which contains a
GAL4 enhancer trap P-element insertion in the 59untranslated
region (UTR) of the transcript encoding for Painless
p103
and the
shared 59UTR of the transcript encoding Painless
p72
. The pain
Gal4
allele shows GAL4 expression in multidendritic neurons, chordo-
tonal neurons, and a subset of neurons in the CNS [4].
To further dissect the role of distinct molecular features Painless
protein isoforms in either mechanical or thermal nociception, we
used expression of GAL4 in the pain
Gal4
mutant allele in
combination with the pain
NP7022
allele. Mutant animals with the
pain
Gal4
/pain
NP7022
genotype showed pronounced thermal nocicep-
tion and mechanical nociception defective phenotypes. In wild
type larvae gently touched with a probe heated to a noxious
temperature (46uC), nocifensive escape behavior was rapidly
triggered (Figure 3A). In contrast, pain
Gal4
/pain
NP7022
larvae
showed a response that is typical for other mutant alleles of
painless (Figure 3B). In mechanical nociception tests, wild type
larvae stimulated with a 50 mN Von-Frey fiber showed nocifen-
sive responses in 70% of trials whereas pain
Gal4
/pain
NP7022
trans-
heterozygous larvae only responded approximately 40% of the
time (Figure 4). Thus, pain
Gal4
/pain
NP7022
trans-heterozygotes
exhibited robust thermal and mechanical nociception defective
phenotypes.
Modality specific isoforms of Painless
We next tested whether expression of the Painless
p103
::VFP or
the Painless
p60
::VFP proteins would be sufficient to rescue the
polymodal aspects of the painless nociceptive phenotype. We
hypothesized that the Painless
p103
isoform would functionally
rescue mechanical nociception because of the predicted role the
N-terminal domain ankyrin repeats in the gating spring model for
mechanosensation. In contrast, the Painless
p60
::VFP protein might
rescue thermal nociception but fail to rescue mechanical
nociception.
The results were contrary to these expectations as expression of
the Painless
p103
::VFP in the pain
Gal4
/pain
NP7022
mutant background
(pain
Gal4
/pain
NP7022
; UAS-Painless
p103
::VFP/+) showed a nearly
complete rescue of nociception responses to a 46uC thermal
stimulus (Figure 3A-C). In contrast to the rescue experiments using
UAS-Painless
p103
::VFP trangenes, the pain
Gal4
/pain
NP7022
; UAS
Painless
p60
::VFP/+animals did not show rescue of the mutant
responses to thermal nociception stimuli (Figure 3D).
The failure of Painless
p60
::VFP to rescue thermal nociception
might have been due its lower expression levels in the sensory
neurons or it might indicate that Painless
p60
::VFP encodes a non-
functional channel. Thus to further test the functional properties of
Painless
p60
::VFP we tested if it could rescue defective mechanical
nociception behaviors of pain
Gal4
/pain
NP7022
mutants. Expression of
this transgene in the mutant background improved the mechanical
nociception responses of the mutant animals (Figure 4) such that
they were no longer different from wild type levels. This result
Figure 1. Schematic diagram of transcription units and
proteins for the newly identified
painless
transcripts. (A.) The
painless
p103
transcription unit. (B.) The painless
p78
transcript shares the
first non-coding exon of painless
p103
but utilizes an alternative
downstream splice acceptor. (C.) The painless
p60
transcript initiates
from an alternate promoter that is downstream of the promoter for
painless
p103
and painless
p78
. (D.) The predicted proteins for the three
isoforms differ in the length of the n-terminal domain. Ankryin repeats
are depicted as red.
doi:10.1371/journal.pone.0030090.g001
Function of Painless Ankyrin Domain
PLoS ONE | www.plosone.org 3 January 2012 | Volume 7 | Issue 1 | e30090

suggested that the Painless
p60
::VFP transgene was functional, and
that the expression level of this transgene was sufficient to rescue
mechanical nociception, but not thermal nociception painless
mutant phenotypes.
Most surprisingly, the Painless
p103
::VFP protein, which was
expressed at significantly higher levels than Painless
p60
::VFP, was
unable to rescue mechanical nociception (Figure 4). If anything,
expression of Painless
p103
::VFP enhanced mechanical nociception
defects of the pain
Gal4
/pain
NP7022
mutant larvae. These results do
not conform to predictions of the gating spring model for the
ankyrin repeat domain. Our results suggest that the ankyrin repeat
domain of Painless
p103
is important for thermal signaling and not
for mechanical signaling. The Painless
p60
isoform, lacking ankyrin
repeats shows complimentary functions, rescuing mechanical
signaling without rescue of thermal signaling.
Discussion
We have found three isoforms of Painless that vary in the length
of the ankyrin repeat containing N-terminal domain with the
longest isoform containing the full N-terminal domain and the
shortest isoform, containing only a small portion of the N-terminal
domain. The Painless
p103
isoform is capable of rescuing the
thermal nociception phenotype of mutant animals in the absence
of functional rescue for mechanical nociception phenotypes. The
Painless
p60
short isoform is capable rescuing mechanical nocicep-
tion in the absence of functional rescue for thermal nociception.
These findings do not support a previously proposed hypothesis in
which ankyrin repeats serve as a compliant gating spring in
mechanosensing TRP channels.
Our results suggest the possibility that distinct isoforms of
Painless are dedicated to specific sensory modalities which require
Painless function. The Painless
p103
isoform may primarily be
required for thermal nociception whereas the Painless
p60
isoform
may be specifically involved in mechanical nociception. Interest-
ingly, this finding suggests that the N-terminal domain of Painless
may have an important function in temperature sensing rather
than in mechanotransduction. Note that our data do not exclude
the possibility that more than one isoform of Painless may be
present in functional channels in vivo. This caveat is necessary to
consider because the pain
Gal4
/pain
NP7022
mutant larvae are not null
for the painless locus. Residual expression of the different isoforms
may allow for the formation of heteromeric channels in our rescue
experiments despite the fact that the rescue transgenes are
designed to express a single isoform. To determine definitively
whether the distinct isoforms are genuinely sufficient for modality
specific rescue, the experiments described here must be repeated
in a DNA null mutant for painless. Efforts to generate a null allele
for painless in our laboratory will make this ideal experiment
possible in the near future. In addition, a painless null allele would
allow for tissue specific rescue experiments in the nociceptor
neurons. Interpretation of the results of this study must be
tempered by the caveat that pain-GAL4 is expressed in cells that
are not nociceptive in addition to the nociceptors themselves.
In the case of mechanosensitive TRP channels, it has been
proposed that the NTD may play an important role in
transmitting cytoskeletal force to the channels [15]. This theory
was developed largely based upon the unusually large number
(twenty-nine) of ankyrin repeats found at the N-terminus of Nomp-
C [15]. The Nomp-C channel is a TRP channel that was first
identified in a screen for uncoordinated mutants that had defects
in mechanically induced currents in Drosophila bristle neurons [16].
Interestingly, Nomp-C is also required for hearing in both
Drosophila and Zebrafish [17,18]. In addition, a conserved role
for Nomp-C in mechanotransduction has been found in C. elegans
[19]. The ankyrin repeats of Nomp-C have been hypothesized to
function as a flexible spring that can transmit force to the
mechanosensitive channel possibly while being anchored to a
cellular component such as the cytoskeleton [15]. Consistent with
Figure 2. Distinct subcellular localization of Painless
p103
::VFP and Painless
p60
::VFP. A–F show the dorsal cluster of multidendritic neurons
of larval abdominal segments immunostained with anti-GFP antibody (green) and anti-HRP (magenta). (A) Anti-GFP staining of pain
Gal4
UAS-
Painless
p103
::VFP larva. (B) Anti-HRP staining of pain
Gal4
UAS-Painless
p103
::VFP. (C) Merge of A and B. (D) Anti-GFP staining of pain
Gal4
UAS-
Painless
p60
::VFP line 1. (E) Anti-HRP staining of pain
Gal4
UAS-Painless
p60
::VFP line 1. (F) Merge of D and E. Note that the exposure times for acquisition of
Painless
p60
::VFP signal was significantly longer than for Painless
p103
::VFP.
doi:10.1371/journal.pone.0030090.g002
Function of Painless Ankyrin Domain
PLoS ONE | www.plosone.org 4 January 2012 | Volume 7 | Issue 1 | e30090

this gating spring hypothesis, evidence has also shown that the
ankyrin repeats of ankyrin-R indeed behave as a Hookean
molecular spring [20]. Widespread interest in this theory stems
from the fact that evidence suggests that a flexible spring element
may be involved in gating of the elusive mechanosensitive ion
channel in hair cells of the inner ear. It was originally proposed
that the extracellular tip links near the tips of stereocilia might be
the gating spring. However, the identification of cadherin-23
[21,22], and protocadherin-15 as components of tip links [23]
argued against tips-links serving as the gating spring because
molecular simulations predict that cadherin molecules are stiffer
than the ‘‘gating spring’’ in hair cells [24]. These simulations
provided further support to the idea that the compliant spring
element in hair cells might exist intracellularly [25].
A potential role for the NTD in thermoTRP heat sensing has
also been previously suggested. In the case of TRPV4, deletion of
ankyrin repeats abolishes heat activated currents [26] but not its
response to hypotonic stimulation [27]. Even so, a role for the
NTD in temperature sensing may not be generally applicable to all
thermoTRP channels. The two isoforms of Pyrexia, a thermo-
sensitive Drosophila TRPA channel differ in the presence of the
N-terminal domain and yet both are thermosensitive [13]. This
suggests that the N-terminal domain of Pyrexia may not be
required for heat activation. Still other evidence implicates the
CTD of TRPV1 and TRPM8 in thermosensing. Deletion of the
CTD alters TRPV1’s temperature sensitivity and domain
swapping of the TRPM8 CTD with the TRPV1 CTD causes a
switch of function in thermosensitivity [28,29]. In addition,
residues surrounding the pore have been implicated in allosteric
modulation of temperature sensing of TRPV3 and TRPV1.
Interestingly, recent results from our laboratory [30] and others
[31] support the idea that the amino terminal domains of TRPA1
channels play a role in heat sensing. Alternative splicing of
Drosophila TrpA1 (dTrpA1) generates transcripts that encode either
heat sensitive (dTRPA1-A, dTRPA1-D) or heat insensitive
isoforms of dTRPA1 (dTRPA1-B, dTRPA1-C) [30]. TRP
Ankryin Caps (TACs) vary between the various dTRPA1 isoforms
and the TACs determine the heat sensing properties. The
dTRPA1-C heat insensitive isoform was found to be required
for thermal nociception but not mechanical nociception. It is
possible that the mechanical nociception function of dTrpA1 could
Figure 3. Isoform specific rescue of thermal nociception in
painless
mutant animals. (A) Nocifensive response latency of control
(Canton-S) larvae, n = 235. (B) pain
NP7022
/pain
Gal4
larvae are defective in
thermal nociception, n = 122 (compared to Canton-S(p,0.001). (C)
Expression of painless
p103
::VFP in NP7022/pain-Gal4 partially rescues the
thermal nociception phenotype (statistically different from both
Canton-S (p,0.001 and NP7022/pain-Gal4 (p,0.001)) (D) Expression
of painless
p60
::VFP does not rescue the thermal nociception phenotype
of pain
NP7022
/pain
Gal4
n = 141 (statistically different from Canton-
S(p,0.001) but not different from NP7022/pain-Gal4 (p.0.05)). The
Wicoxon Rank Sum test was used for all statistical analysis.
doi:10.1371/journal.pone.0030090.g003
Figure 4. Isoform specific rescue of mechanical nociception
phenotypes in
painless
mutant animals. The graph shows the
percentage of animals that respond to 50 mN Von Frey stimulus with
Nocifensive Escape Locomotion behavior. Canton-S n = 249, pain
NP7022
/
pain
Gal4
n = 201, pain
NP7022
/pain
Gal4
;UAS-painless
p60
::VFP n = 343,
pain
NP7022
/pain
Gal4
n = 201, pain
NP7022
/pain
Gal4
; UAS-painless
p103
::VFP
n = 71. (One way ANOVA p,0.0001, pair-wise comparisons to Canton-
S were performed with the Dunnett post-hoc multivariate t-distribution,
**indicates p,0.01 and *** indicates p,0.001 relative to Canton-S).
doi:10.1371/journal.pone.0030090.g004
Function of Painless Ankyrin Domain
PLoS ONE | www.plosone.org 5 January 2012 | Volume 7 | Issue 1 | e30090

rely on an as yet unidentified isoform for this channel that lacks
ankyrin repeats as we have now found for painless.
In addition, chimeric channels between heat sensitive snake or
Drosophila TRPA1 channels made with the heat insensitive human
TRPA1 channel, support the idea that heat sensor domains reside
within the ankyrin repeat region [31]. Our in vivo analyses of
painless are in definite support of this idea.
Although the Painless NTD is apparently not required for
mechanical nociception it would be premature to extrapolate from
this finding to other TRP channels involved in mechanotransduction.
Additional experiments will be necessary to test whether or not
Painless or other TRPA channels encode pore-forming subunits of
mechanosensory channels. As mentioned above, Painless may
function downstream of another mechanosensory such as Pickpocket.
Interestingly, our results suggest that if Painless does function
downstream of Pickpocket, then the method of activation for Painless
is unlikely to be dependent on Ca
2+
influx, since the putative EF hand
of the Painless NTD in absent in the Painless
p60
isoform.
Our results also indicate that the Painless ankyrin repeats play an
important role in the sub-cellular localization and the expression level
of Painless isoforms. The Painless
p60
::VFP transgene was expressed at
relatively low levels in vivo and its subcellular localization was
restricted to proximal dendrites, the cell soma, and proximal axons.
In contrast, the Painless
p103
::VFP isoform was expressed robustly and
throughout the multidendritic neurons including the sensory
dendrites. These findings suggest the possibility that Painless
p103
::VFP
and Painless
p60
::VFP proteins function in distinct subcellular
compartments for thermal and mechanical nociception. Thermal
nociception signaling by Painless
p103
mayoccurinthemoredistal
sensory dendrites whereas the mechanical nociception function for
Painless may reside in proximal regions of dendrites, likely in
amplification of upstream mechanosensory transduction signals
initiated by the DEG/ENaC Pickpocket.
Materials and Methods
RNA Ligase Mediated Rapid Amplification of cDNA Ends
(RLM-RACE)
The 59ends of painless transcripts were isolated by 59RACE using the
FirstChoice (RLM-RACE kit (RNA Ligase Mediated Rapid Ampli-
fication of cDNA Ends, Ambion Inc.) according to the instructions of
the manufacturer. RNA was isolated from a mixed population of first
and second instar Drosophila w
1118
larvae using Trizol reagent
(Invitrogen/Life Technologies) and treated with DNAse-I (DNA-free
Kit (Ambion Inc.). The 59phosphate from any uncapped RNA in the
preparation was then removed by treatment with Calf Intestine
Alkaline Phosphatase (CIP). The CIP reaction was terminated by
phenol-chloroform extraction, and the RNA was precipitated with
isopropanol. The 59capwasremovedfromthemRNAbytreatment
with Tobacco Alkaline Pyrophosphatase (TAP) enzyme, leaving
monophosphated 59ends. A 45-base RNA adaptor from the
FirstChoice (RLM-RACE kit was then ligated to these newly exposed
monophosphated 59ends with T4 RNA Ligase.
First strand cDNA was generated using random decamer
primers and M-MLV Reverse Transcriptase. Nested PCR was
performed to amplify painless 59ends, using forty cycles and 5 min
extension times for both inner and outer reactions. The forward
primers were supplied by the RACE kit (59Race outer and inner
forward primers) and are complimentary to sequences present in
the ligated 59adaptor. The outer reaction painless specific reverse
primer 59-GGATGGTAAATACGGCTAAGAC-39and the inner
reaction painless-specific reverse primer 59-TTCGTGGAACTT-
GAGGAGGCGTG-39were used. PCR products were examined
by agarose-TAE gel electrophoresis. The products were gel-
purified, cloned into pCR-XL-TOPO cloning vector (Invitrogen/
Life Technologies), and introduced into electrocompetent E.coli of
the TOP10 strain (Invitrogen/Life Technologies). DNA from
Kanamycin-resistant colonies was digested with EcoRI restriction
endonuclease to release the cloned inserts, and the relative
molecular weight of the inserts was determined by gel electropho-
resis. Clones containing inserts with unique molecular weights
were selected for sequencing. The inserts of clones were sequenced
using M13 F and M13R against the TOPO XL vector.
Fly Strains
To generate the UAS-painless
p103
::YFP and UAS-painless
p60
::YFP
plasmids the open reading frames were amplified from BACR08I14
using the forward primer for painless
p103
59CAC CAT GGA CTT
TAA CAA CTG C 93 or the forward primer for painless
p60
59
CACCATG GAT ATC AAC TCG AGA CCA 39. The reverse
primer 59CCGGTCCTGGACCAGCT39was used for both
constructs. The PCR products for the respective gene products were
then cloned in the pENTR-D vector (Invitrogen) for use with the
Drosophila Gateway Cloning System. The resulting painless
p60
and
painless
p103
ENTRvectorswerethenusedassubstratesforclonase
mediated recombination with the pTWV destination vector. Fully
sequenced inserts from this reaction were found to contain wild type
painless sequences and were used to transform the w
1118
Drosophila
strain via P-element mediated transformation. Inserts from trans-
formed flies were mapped to a chromosome and the expression levels
of the VFP transgenes were determined by crossing to the painless
GAL4
driver strain. For behavioral experiments inserts on the third
chromosome were used. Flies with the genotype w;painless
GAL4
,w;
painless
GAL4
;UAS-painless
p103
::YFP/K87 (T(2:3) SM6a TM6b Cy Tb),
or painless
GAL4
;UAS-painless
p60
::YFP/K87 were crossed to pain
NP7022
/
K87 and Tb
+
progeny were selected and tested for nociception
behavioral responses as described previously.
Immunostaining and confocal microscopy
For visualization of the painless isoforms, the pain
GAL4
driver was
crossed to UAS-painless
P103
::VFP (venus fluorescent protein) or UAS-
painless
P60
::VFP. Trans-heterozygous larvae were dissected and
filleted in Ca
2+
free HL3 saline (70 mM NaCl, 5 mM KCl,
20 mM MgCl
2
, 10 mM NaHCO
3
, 5 mM trehalose, 115 mM
sucrose, and 5 mM HEPES [pH 7.2]) followed by fixation for
1 hour in 4% paraformaldehyde. Primary mouse anti-GFP
(1:1000) and anti-mouse Alexa Fluor 488 (Molecular probes,
1:1000) secondary were used to detect VFP. Neurons were
counterstained with rabbit anti-HRP (1:500, anti-horseradish
peroxidase) and the secondary anti-rabbit Alexa Fluor 568
(Molecular probes, 1:1000). Images (102461024) were taken with
a Zeiss LSM 5 Live Confocal system using a 4061.3 N/A oil
immersion lens. The two channels were collected separately in
multi-track mode (anti-GFP: excitation 488 nm, emission 500–
525 nm) (anti-HRP excitation 532 nm, emission 560–675). The
confocal micrographs are presented maximum intensity projec-
tions of confocal Z-stacks.
Acknowledgments
We thank the members of the Tracey lab who provided useful advice and
suggestions. We thank the Duke University Fly Core for the generation of
transgenic flies harboring the transgenes used in this study.
Author Contributions
Conceived and designed the experiments: RH NS WDT. Performed the
experiments: RH NS WDT. Analyzed the data: RH NS WDT. Wrote the
paper: RH NS WDT.
Function of Painless Ankyrin Domain
PLoS ONE | www.plosone.org 6 January 2012 | Volume 7 | Issue 1 | e30090

References
1. Damann N, Voets T, Nilius B (2008) TRPs in our senses. Curr Biol 18:
R880–889.
2. Al-Anzi B, Tracey WD, Benzer S (2006) Response of Drosophila to wasabi is
mediated by painless, the fly homolog of mammalian TRPA1/ANKTM1. Curr
Biol 16: 1034–1040.
3. Sokabe T, Tsujiuchi S, Kadowaki T, Tominaga M (2008) Drosophila painless is
aCa2+-requiring channel activated by noxious heat. J Neurosci 28: 9929–9938.
4. Tracey WD, Jr., Wilson RI, Laurent G, Benzer S (2003) painless, a Drosophila
gene essential for nociception. Cell 113: 261–273.
5. Xu SY, Cang CL, Liu XF, Peng YQ, Ye YZ, et al. (2006) Thermal nociception
in adult Drosophila: behavioral characterization and the role of the painless
gene. Genes, Brain and Behavior 0: 1–12.
6. Xiang Y, Yuan Q, Vogt N, Looger LL, Jan LY, et al. (2010) Light-avoidance-
mediating photoreceptors tile the Drosophila larval body wall. Nature.
7. Hwang RY, Zhong L, Xu Y, Johnson T, Zhang F, et al. (2007) Nociceptive
neurons protect Drosophila larvae from parasitoid wasps. Curr Biol 17:
2105–2116.
8. Zhong L, Hwang RY, Tracey WD (2010) Pickpocket is a DEG/ENaC protein
required for mechanical nociception in Drosophila larvae. Curr Biol 20:
429–434.
9. Goodman MB, Ernstrom GG, Chelur DS, O’Hagan R, Yao CA, et al. (2002)
MEC-2 regulates C. elegans DEG/ENaC channels needed for mechanosensa-
tion. Nature 415: 1039–1042.
10. O’Hagan R, Chalfie M (2006) Mechanosensation in Caenorhabditis elegans. Int
Rev Neurobiol 69: 169–203.
11. Chung YD, Zhu J, Han Y, Kernan MJ (2001) nompA encodes a PNS-specific,
ZP domain protein required to connect mechanosensory dendrites to sensory
structures. Neuron 29: 415–428.
12. Tracey WD, Wilson RL, Laurent G, Benzer S (2003) painless, a Drosophila
Gene Essential for Nociception. Cell 113: 261–273.
13. Lee Y, Lee Y, Lee J, Bang S, Hyun S, et al. (2005) Pyrexia is a new thermal
transient receptor potential channel endowing tolerance to high temperatures in
Drosophila melanogaster. Nat Genet 37: 305–310.
14. Schumacher MA, Moff I, Sudanagunta SP, Levine JD (2000) Molecular cloning
of an N-terminal splice variant of the capsaicin receptor. Loss of N-terminal
domain suggests functional divergence among capsaicin receptor subtypes. J Biol
Chem 275: 2756–2762.
15. Howard J, Bechstedt S (2004) Hypothesis: a helix of ankyrin repeats of the
NOMPC-TRP ion channel is the gating spring of mechanoreceptors. Curr Biol
14: R224–226.
16. Walker RG, Willingham AT, Zuker CS (2000) A Drosophila mechanosensory
transduction channel. Science 287: 2229–2234.
17. Gopfert MC, Albert JT, Nadrowski B, Kamikouchi A (2006) Specification of
auditory sensitivity by Drosophila TRP channels. Nat Neurosci 9: 999–1000.
18. Sidi S, Friedrich RW, Nicolson T (2003) NompC TRP channel required for
vertebrate sensory hair cell mechanotransduction. Science 301: 96–99.
19. Li W, Feng Z, Sternberg PW, Xu XZ (2006) A C. elegans stretch receptor
neuron revealed by a mechanosensitive TRP channel homologue. Nature 440:
684–687.
20. Lee G, Abdi K, Jiang Y, Michaely P, Bennett V, et al. (2006) Nanospring
behaviour of ankyrin repeats. Nature 440: 246–249.
21. Siemens J, Lillo C, Dumont RA, Reynolds A, Williams DS, et al. (2004)
Cadherin 23 is a component of the tip link in hair-cell stereocilia. Nature 428:
950–955.
22. Sollner C, Rauch GJ, Siemens J, Geisler R, Schuster SC, et al. (2004) Mutations
in cadherin 23 affect tip links in zebrafish sensory hair cells. Nature 428:
955–959.
23. Ahmed ZM, Goodyear R, Riazuddin S, Lagziel A, Legan PK, et al. (2006) The
tip-link antigen, a protein associated with the transduction complex of sensory
hair cells, is protocadherin-15. J Neurosci 26: 7022–7034.
24. Sotomayor M, Corey DP, Schulten K (2005) In search of the hair-cell gating
spring elastic properties of ankyrin and cadherin repeats. Structure 13: 669–682.
25. Corey DP, Sotomayor M (2004) Hearing: tightrope act. Nature 428: 901–903.
26. Watanabe H, Vriens J, Suh SH, Benham CD, Droogmans G, et al. (2002) Heat-
evoked activation of TRPV4 channels in a HEK293 cell expression system and
in native mouse aorta endothelial cells. J Biol Chem 277: 47044–47051.
27. Liedtke W, Choe Y, Marti-Renom MA, Bell AM, Denis CS, et al. (2000)
Vanilloid receptor-related osmotically activated channel (VR-OAC), a candidate
vertebrate osmoreceptor. Cell 103: 525–535.
28. Brauchi S, Orta G, Salazar M, Rosenmann E, Latorre R (2006) A hot-sensing
cold receptor: C-terminal domain determines thermosensation in transient
receptor potential channels. J Neurosci 26: 4835–4840.
29. Vlachova V, Teisinger J, Susankova K, Lyfenko A, Ettrich R, et al. (2003)
Functional role of C-terminal cytoplasmic tail of rat vanilloid receptor 1.
J Neurosci 23: 1340–1350.
30. Zhong L, Bellemer A, Yan H, Honjo K, Robertson J, et al. (2011)
Thermosensory and Nonthermosensory Isoforms of Drosophila melanogaster
TRPA1 Reveal Heat -Sensor Domains of a ThermoTRP Channel. Ce ll
Reports;10.1016/j.celrep.2011.11.002.
31. Cordero-Morales JF, Gracheva EO, Julius D (2011) Cytoplasmic ankyrin repeats
of transient receptor potential A1 (TRPA1) dictate sensitivity to thermal and
chemical stimuli. Proc Natl Acad Sci U S A;10.1073/pnas.1114124108.
Function of Painless Ankyrin Domain
PLoS ONE | www.plosone.org 7 January 2012 | Volume 7 | Issue 1 | e30090