Title: Drosophila Ionotropic Receptor 25a mediates circadian clock resetting by temperature
Authors: Chenghao Chen
1a
, Edgar Buhl
2a
, Min Xu
1
, Vincent Croset
4,+
, Johanna Rees
3
, Kathryn S.
Lilley
3
, Richard Benton
4
, James J. L. Hodge
2
, and Ralf Stanewsky
1*
a
both authors contributed equally
Affiliations:
1
Department of Cell and Developmental Biology, University College London, 21 University
Street, London, WC1E 6DE, UK.
2
School of Physiology and Pharmacology, University of Bristol, University Walk, Bristol, BS8
1TD, UK.
3
Cambridge Centre for Proteomics, Department of Biochemistry and Cambridge Systems
Biology Centre, University of Cambridge, Tennis Court Road, Cambridge, CB2 1QW, UK.
4
Center for Integrative Genomics, Faculty of Biology and Medicine, University of Lausanne, CH-
1015 Lausanne, Switzerland.
+
Present address: Centre for Neural Circuits and Behaviour, University of Oxford, Mansfield
Road, Oxford, OX1 3SR, UK.
*Correspondence to:
Ralf Stanewsky
Tel: +44(0)20-7679-6610
Email: r.stanewsky@ucl.ac.uk
2
Summary:
Circadian clocks are endogenous timers adjusting behaviour and physiology with the solar day
1
.
Synchronized circadian clocks improve fitness
2
and are crucial for human physical as well as
mental wellbeing
3
. Visual and non-visual photoreceptors are responsible for synchronizing
circadian clocks to light
4,5
, but clock-resetting is also achieved by alternating warmer (‘day’) and
colder (‘night’) temperatures with 2°-4°C difference only
6-8
. This temperature sensitivity is even
more remarkable considering that the period of circadian clocks (~24 h) is largely independent
of the surrounding ambient temperature
1,8
. Here we show that the Drosophila Ionotropic
Receptor 25a (IR25a) is required for behavioural synchronization to low-amplitude temperature
cycles. We found that this channel is expressed within sensory neurons of internal stretch
receptors in the fly body, which have previously been implicated in temperature
synchronization of the circadian clock
9
. IR25a is required for temperature-synchronized clock
protein oscillations in specific subsets of central clock neurons, defining the neural substrates
for temperature sensitivity within the circadian clock circuit. Extracellular leg nerve recordings
reveal temperature-and IR25a-dependent sensory responses and misexpression of IR25a
confers temperature-dependent firing of action potentials in heterologous neurons. We
propose that IR25a is part of a temperature input pathway to the circadian clock that is
responsible detecting small temperature differences. This pathway is operating in the absence
of the known temperate-preference regulating ‘hot’ and ‘cold’ sensors in the fly antenna
10,11
,
and hence revealing the existence of novel periphery-to-brain temperature signalling routes
involving IR25a function in peripheral sensory organs.
3
Main Text:
In Drosophila, daily activity rhythms are controlled by a network of ~150 clock neurons
expressing the clock genes period (per) and timeless (tim), which encode repressor proteins that
negatively feedback on their own promoters resulting in 24 h oscillations of clock molecules.
Temperature cycles (TC) synchronize molecular clocks present in external body parts and the
PNS in a tissue autonomous manner
9,12
, while synchronization of clock neurons in the brain
largely depends on temperature input from peripheral temperature receptors located in the
chordotonal organs (ChO) and on the gene nocte, which is also expressed in ChO
9,12,13
.
To identify novel proteins involved in temperature entrainment we expressed tagged NOCTE
versions in flies followed by purification of interacting partners and their identification by mass-
spectrometric analysis (see Methods and
14
). nocte mutants show defects in ChO morphology,
pointing to a structural role of NOCTE in ChO cilia
9
. Consequently, the majority of the identified
proteins (10/16) likely regulate function and dynamics of the ChO neuron cilia (Extended Data
Tab. 1). Since we were mainly interested in identifying potential temperature receptors, we
focused on other NOCTE-interacting proteins, particularly on Ionotropic Receptor 25a (IR25a)
(Extended Data Tab. 1). We verified the interaction by co-immunoprecipitation after
overexpressing IR25a and NOCTE in all clock cells using tim-gal4 (Extended Data Fig. 1a). IR25a
is a member of a divergent subfamily of ionotropic glutamate receptors, which function in
chemosensory detection rather than synaptic transmission, consistent with IR25a expression in
many different populations of sensory neurons in the antenna and labellum
15-17
. The function of
IR25a has only been analysed in olfactory neurons of the third antennal segment, where it acts
as a co-receptor with different odour-sensing IRs
15
.
4
To investigate if IR25a is co-expressed with nocte in ChO we first analysed IR25a expression in
femur and antennal ChO using an IR25a-gal4 line, previously shown to at least partially reflect
the IR25a expression pattern in the
third antennal segment
15
(Extended Data Fig. 2a). IR25a-
gal4 driven mCD8-GFP prominently labelled subsets of ChO neurons in the femur, which
showed substantial overlap with nompC-QF driven QUAS-Tomato signals (Fig. 1 a-c). nompC-QF
is expressed in larval ChO
18
as well as in the adult femur ChO (Fig. 1d, e) Comparison of IR25a-
driven mCD8-GFP and nuclear Ds-Red signals with those of other ChO neuron drivers (F-gal4
and nocte-gal4
9
), suggests that IR25a is transcribed in a small subset of femur ChO neurons
and Johnston’s Organ (JO) neurons (Fig. 1c, Extended Data Fig. 1b-g). We also detect
endogenous IR25a mRNA in the femur and leg (Extended Data Fig. 2b, e). To determine if
IR25a-gal4 ChO signals reflect endogenous IR25a expression, we performed IR25a antibody
staining on femur ChO (Fig. 1f, g). Double labelling with a neuronal marker revealed IR25a
signals within ChO neuron cell bodies and ciliated dendrites, similar as in coeloconic neurons of
the antenna
16
. This subcellular distribution of IR25a was confirmed after co-expression of an
mCherry-IR25a fusion protein with the dendritic cap marker NOMPA-GFP (Fig. 1h), which
showed expression along the ChO cilia, clearly distinct from the dentritic cap. Together, these
results show that IR25a is expressed in subsets of antennal and femur ChO neurons and the
IR25a-gal4 driver reflects this pattern.
Since nocte
1
mutants do not synchronize to 16°C : 25°C TC in constant light (LL)
9,12
(Extended
Data Fig. 3a) we analysed IR25a
-/-
mutants
16
under these conditions. Unlike nocte
1
, the IR25a
-/-
flies synchronized well to this regime and we obtained similar results at warmer TC (Extended
Data Fig. 3a). To test the possibility that IR25a is specifically required for synchronization to
5
small temperature intervals
7,13
, we subjected IR25
-/-
flies to a series of TC with an amplitude of
2°C only. Surprisingly, and in contrast to wild type, IR25a
-/-
mutants did not synchronize to any
of the shallow TC in LL or constant darkness (DD) (Fig. 2a-e, Extended Data Fig. 3b, 4c). While in
LL wild type and IR25a rescue flies showed a clear activity peak in the 2
nd
part of the warm
period before and after the 6 h shift of the TC, IR25a
-/-
mutants were constantly active
throughout the TC, apart from a short period of reduced activity at the beginning of the warm
phase of TC1 (Fig. 2a, Extended Data Fig. 3b). In DD, control flies slowly advanced (or delayed)
their evening activity peak during phase-advanced (or delayed) TC (Fig. 2b, Extended Data Fig.
4c). The phase of this activity peak was maintained in the subsequent free running conditions
(DD, const. 25°C) indicating stable re-entrainment of the circadian clock (Fig. 2b, Extended Data
Fig. 4). In contrast, IR25a mutants did not shift their evening peak during the TC; instead it
remained at its original phase throughout the experiment (Fig. 2b; see also Extended Data Fig.
4c and Fig. 2d for phase quantification).
To quantify entrainment in LL, we determined the ‘Entrainment Index’ (EI) for each genotype
and condition, whereas for most DD experiments we calculated the phase difference of the
main activity peak upon release into constant conditions between IR25a mutants and controls
(see Methods). In all 2°C-amplitude TC tested the EI of IR25a
-/-
flies was significantly lower (LL)
and phase calculation indicated no, or a significantly reduced phase shift compared to controls
(Fig. 2c-e). The same non-synchronization phenotype was observed in IR25a
-
/Df(IR25a) flies,
and temperature synchronization was fully restored in IR25a
-/-
flies expressing a genomic
rescue construct (rescue) (Fig. 2a-d, Extended Data Fig. 3b). IR25a
-/-
mutants synchronize to
light and have normal free-running and temperature compensated periods (Fig. 2b, Extended