tßhnM18 mutants respond
less to sucrose.
First, we explored the involvement of OA and TA in the modulation of
sugar responsiveness. We used tßhnM18 mutant flies that lack OA and accumulate TA (Monastirioti et al.,
1996), tethered them to a hook glued between head and thorax and
tested their proboscis extension response to a serial dilution of
sucrose after 20 h of starvation. The proportion of flies
extending the proboscis increased with increasing sucrose
concentration to reach almost 90% (Fig. 1A). tßhnM18 mutant flies responded almost 40% less than their control (Fig. 1A).
The sum of all positive responses over the 7 sucrose presentations
was significantly different (Wilcoxon rank sum test, p = 2*10-12,
Fig. 1B). These results suggest that tßh-expressing and
consequently that OA and TA play a crucial role in controlling sugar
responses under starvation conditions.
Physiological response to starvation is weaker in tßhnM18 compared to wild type flies.
In order to determine whether tßhnM18 mutants might be less affected by starvation, we compared
carbohydrate content (trehalose plus glucose) in the hemolymph of
starved and fed flies. To this end, the hemolymph was extracted and
all glucose and trehalose was enzymatically converted into
spectrometrically measurable glucose. Trehalose is the “blood-sugar”
in insects (Thompson, 2003) and is degraded under starvation
conditions (Isabel et al., 2005; Meunier et al., 2007). Starvation
treatment reduced the carbohydrate content in both, mutants and wild
type flies (ANOVA: starvation p = 4*10-6,
F = 27.449, genotype p = 0.14, F = 2.261,
starvation x genotype p = 0.189, F = 1.777,
Fig. 1A). The change in carbohydrate level after starvation was
significantly smaller in tßh mutants compared to wild type
controls (Welch Two Sample t-Test, p = 0.0497, Fig. 1B). As
a second measure of starvation resistance, we recorded survival rate
under starvation conditions with ad libitum access to water.
As expected from their increased sugar content, tßhnM18 mutants survived longer than wild type controls (Wilcoxon rank sum
test, p = 0.039, Fig. 1). The prolonged survival cannot be
due to remaining eggs in the ovaries of tßh mutants
(Partridge et al., 1987; Monastirioti, 2003) since we found the
same survival effect on male tßh mutants (Fig. 1) and tßh-expression in mutant background did rescue egg-laying but
not the survival phenotype (unpublished observations). Our
experiments show that tßhnM18 mutants are
less affected by starvation compared to wild type.
Acute tßh induction rescues the sugar
response phenotype.
In order to rescue the tßhnM18 mutant sugar
response phenotype, we induced ubiquitous tßh expression in
the mutant background at different time points by means of the
heat-induced construct hsp-tßh. With heat shock-induced tßh expression 3 h before testing, we found an increase in the
mutants’ sugar response (paired Wilcoxon rank sum test, ap = 6*10-5, bp = 0, cp = 4*10-12, Fig. 1A). With heat
shock-induced tßh expression exclusively during the
starvation period, we did not find a rescue of the tßh mutant
phenotype (paired Wilcoxon rank sum test with bonferroni correction, ap = 0.008, bp = 0, hsp-tßh x tßhp = 1, Fig. 1B).
Those data indicate an acute role of the OA/TA-system during sugar
response.
Sensitivity of taste receptor neurons is lower in tßhnM18 mutants.
OA is known to modulate different kinds of receptors in insects
(Kass et al., 1988; Ramirez and Orchard, 1990; Pophof, 2000). In
order to test a potential role of OA on gustatory receptor
sensitivity we recorded the response of labellar sensilla to 100 mM
sucrose of fed and starved flies by the tip-recording method
(Hodgson et al., 1955; Hiroi et al., 2002). We found a decreased
sensillar response to sucrose stimulation after starvation in tßhnM18 mutants, compared to wild type controls (Fig. 1A, Wilcoxon rank sum
test, p = 0.008). The increase of spiking rate after
starvation seems to underlie great genetic variation since different
wild types exhibit different phenotypes (Fig. 1B, Wilcoxon rank sum
test, w1118p = 0.039, CantonSp = 0.001;
(Meunier et al., 2007; Inagaki et al., 2012; Nishimura et al.,
2012).
Those data indicate a role of the
OA/TA-system in the starvation-dependent modulation of receptor
potentials.
Differential outcome of OA/TA-receptor manipulations on survival
and sugar responsiveness
Because the tßh mutation leads to increased TA and decreased
OA levels (Monastirioti et al., 1996), we performed additional
experiments to disentangle the relative importance of each amine in
the regulation of survival and sugar response. We tested published
and novel mutants for several OA- and TA-receptors in our PER and
survival under starvation conditions assays (Fig. 1, Table 1).
TyrRf05682 was generated by piggyback
transposition (Thibault et al., 2004), real-time PCR revealed a
100-fold decrease of mRNA transcript (Zhang and Blumenthal,
submitted). TyrRII∆29 and the double mutant TyrRII-TyrR∆124 are deletion mutants of 13567316-13576610 and the entire region
13567316-13579400, respectively. TyrRII has low mRNA transcript
levels such that quantification of mutant levels is not possible
(Zhang and Blumenthal, submitted). Octß2R∆3.22 and Octß2R∆4.3 are deletion mutants of 17,539 bp and 33,489
bp, respectively. Deletions were confirmed by genomic PCR. oamb and honoka were described elsewhere (Han et al.,
1998; Kutsukake et al., 2000).
The two TA-receptor mutants TyrRf05682 and honoka showed a decreased sugar response (Wilcoxon rank sum
test with correction for multiple measurements, TyrRf05682p = 0.003, honokap = 0.012) and an increased
survival (TyrRf05682p = 0.002, honokap = 0.009) comparable to tßhnM18 mutants. Interestingly, the two phenotypes could be decorrelated
indicated by the differential behavioral outcome of the tested
receptors: Octß2R and the double mutant TyrII-TyrR∆124 show an increase in survival (Octβ2R∆3.22p = 0.001, Octß2R∆4.3p = 0.021, TyrRII-CG7431∆124p = 0.015)
and a normal sugar response (Octß2R∆3.22p = 0.825, Octß2R∆4.3p = 0.06),
while TyrRII∆29 shows normal survival (p = 0.354) and a decrease in
sugar response (p = 0.007). Finally, the oamb mutants showed no phenotype at all (survival: oamb286p = 0.397, oamb584p = 0.867; sugar response: oamb286p = 0.506, oamb584p = 0.388),
in contrast to a previously published report (Erion et al., 2012).
The receptor mutant data suggest that flies can exhibit a wild type
survival simultaneously with a lower sugar response (TyrRII∆24),
or a higher survival simultaneously with a wild type sugar response
(Octß2R∆3.22 and Octß2R∆4.3).
That indicates that starvation affects sugar responsiveness and
survival via different pathways.
Neuronal and non-neuronal tßh expression rescues PER
OA and TA act both inside and outside of the nervous system,
functioning as either a neurotransmitter or -hormone in insects
(Cole et al., 2005). Thus, we explored whether the sugar response
phenotype of tßh mutants was a result of alterations in
neurons inside or outside of the brain or in non-neuronal cells. For
this experiment, we expressed UAS-tßh in tßhnM18 mutant males driven by different GAL4-lines. We found a significant
increase in sugar response compared to the respective mutant control
when we used the ubiquitous Actin-promoter (Wilcoxon rank sum test,
p = 0.016, Fig. 1), the pan-neuronal nSyb-promoter
(p = 0.013, Fig. 1), or the non-neuronal Tdc1-GAL4 driver
(Cole et al., 2005) (p = 0.028, Fig. 1). In contrast, tßh expression in subsets of OA/TA-neurons in thoracic nerve
cord and the brain by using either Tdc2- or NP7088-GAL4 did not
significantly affect the mutants’ response (Tdc2p = 0.098, NP7088p = 0.58, Fig. 1) in contrast to a
previous report (NP7088-Gal4, Scheiner et al., 2014). These results
indicate that tßh expression induced in neurons in the
central nervous system as well as in non-neuronal cells is sufficient
to enhance the sugar responsiveness of tßhnM18 mutant flies.