Aprt deficiency shortens lifespan and induces metabolic and neurobehavioral defects.

(A) Aprt5 mutant flies have a reduced lifespan compared to wild-type flies (median lifespan: 38 and 50 days, respectively). Three independent experiments were performed on 150 males per genotype with similar results and a representative experiment is shown. Log-rank test (***p < 0.001). (B) HPLC profiles on head extracts revealed an increase in uric acid levels in Aprt5 mutant flies. Administration of 100 μg/ml allopurinol for 5 days before the test rescued uric acid levels. Mean of 3 independent experiments performed on 40 flies per genotype. One-way ANOVA with Tukey’s post-hoc test for multiple comparisons (*p < 0.05; **p < 0.01; ns: not significant). (C-E) Effect on climbing ability. (C) Aprt5 mutants show early defects in the SING paradigm that monitors locomotor reactivity and climbing performance. This deficit was already obvious at 1 da.E. (days after eclosion) and further decreased up to 8 d a.E., after which it did not change significantly up to 30 d a.E. Mean of 3 independent experiments performed on 50 flies per genotype. Unpaired Student’s t test (**p < 0.01; ***p < 0.001). (D, E) Administration of allopurinol does not rescue the motricity defects of Aprt-deficient mutants. Feeding the Aprt5 mutants with allopurinol (100 μg/ml) either in adults 5 days before the test (D) or throughout all developmental stages (E) did not alter the defects observed in SING behavior. Results of one experiment performed on 50 flies per genotype at 10 d a. E. Unpaired Student’s t test (***p < 0.001). (F) Downregulating Aprt by RNAi in all cells (da>AprtRNAi) also led to an early impairment in climbing responses in the SING assay at 10 d a.E., as compared to the driver (da/+) and effector (AprtRNAi/+) only controls. Mean of 3 independent experiments performed on 50 flies per genotype. One-way ANOVA with Tukey’s post-hoc test for multiple comparisons; (***p < 0.001). (G) Adult-specific inactivation of Aprt (tub-Gal80ts; da-Gal4>AprtRNAi) decreased startle-induced climbing abilities in the SING paradigm, suggesting that the locomotor impairment induced by Aprt deficiency is not caused by a developmental effect. Flies were raised at permissive temperature (18°C) in which Gal80ts suppressed Gal4-controlled AprtRNAi expression, and were shifted from 18 to 30°C for 3 days before the test (between 7 and 10 d a.E.) to activate transgene expression. Mean of 3 independent experiments performed on 50 flies per genotype. Two-way ANOVA with Sidak’s post-hoc test for multiple comparisons (***p < 0.001; ns: not significant).

Aprt knockdown in neurons or glial cells disrupts startle-induced locomotion in Drosophila.

(A) AprtRNAi expression in all neurons with elav-Gal4 decreased SING performance in 10-day-old flies. (B) Pan-neuronal expression of Drosophila Aprt with the UAS-Gal4 system partially rescued the locomotor response of Aprt5 mutants. (C-D) Downregulation of Aprt expression in all glia with repo-Gal4 (C) or in glial cell that express the glutamate transporter Eaat1 with Eaat1-Gal4 (D) also altered SING performances. (E) Aprt re-expression in glial cells with Eaat1-Gal4 driver did not rescue the climbing response of Aprt5 mutants. Results of 3 or 4 independent experiments performed on 50 flies per genotype at 10 d a. E. One-way ANOVA with Tukey’s post-hoc test for multiple comparisons (*p < 0.05; **p < 0.01; ns: not significant).

Aprt downregulation in DA neurons of the PAM cluster and in mushroom body neurons impairs startle-induced locomotion.

(A) RNAi-mediated Aprt inactivation in brain DA neurons except a large part of the PAM cluster with the TH-Gal4 driver did not lead to locomotor defects in the SING assay. (B) In contrast, Aprt knockdown in all dopaminergic neurons including the PAM cluster with the TH-Gal4, R58E02-Gal4 double-driver led to a decrease in SING performance. (C) Aprt downregulation in serotoninergic neurons with TRH-Gal4 did not alter startle-induced climbing response of the flies. (D-F) Aprt knockdown selectively in DA neurons of the PAM cluster using either R58E02-Gal4 (D), NP6510-Gal4 (E) or R76F05-Gal4 (F) significantly decreased climbing performance. (G-H) Aprt knockdown in all the mushroom body intrinsic neurons with 238Y-Gal4 (G) or VT30559-Gal4 (H) also led to a decrease in SING performance. Results of 3 or 4 independent experiments performed on 50 flies per genotype at 10 d a. E. One-way ANOVA with Tukey’s post-hoc test for multiple comparisons (*p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant).

Aprt-deficient flies sleep less and walk slower than wild-type flies.

(A) Quantification of total spontaneous locomotor activity during day and night over 5 LD cycles. Aprt5 mutants show no difference in spontaneous locomotion with wild-type flies during the day but a higher activity at night. 3 independent experiments were performed on 32 flies per genotype and mean ± SEM was plotted. Unpaired Student’s t test (***p < 0.001; ns: not significant). (B) Sleep pattern during a typical 24h LD cycle showing that the total amount of sleep is smaller during day and night in Aprt5 mutants compared to wild-type flies. (C) Quantification of day (ZT1-12), night (ZT13-24) and total sleep in Aprt5 mutants. ZT is for zeitgeber. (D) Locomotion speed during waking is reduced in Aprt5 mutants. (E) The average sleep bout duration is also decreased, indicating that Aprt5 mutants have a difficulty to maintain sleep. (F) Sleep pattern of elav>AprtRNAi flies, showing that knockdown of Aprt in all neurons led to sleep reduction during the night, similarly to the mutant condition, and an even more profound sleep defect during the day. (G) Quantification of total amount of sleep when Aprt was downregulated in all neurons (elav-Gal4), all glial cells (repo-Gal4) and both neurons and glial cells (elav-Gal4; repo-Gal4). Except for glia only, the resulting effect was a significant sleep reduction. For sleep and locomotion speed, means ± SEM were plotted. Unpaired Student’s t test (C-E) and one-way ANOVA with Tukeys’s post-hoc test for multiple comparisons (G) (*p < 0.05; **p < 0.01; ***p < 0.001; ns: not significant).

Aprt deficiency increases DA synthesis and content in the Drosophila brain.

(A) Representative confocal projections of TH-immunostained whole-mount adult brains from wild-type flies and Aprt5 mutants. MB: mushroom body. Scale bars: 100 μm. (B) Quantification of TH immunofluorescence intensity normalized to the controls in the entire brain. 4 to 6 brains were dissected per experiment and genotype, and 6 independent experiments were performed (**p < 0.01). (C) Representative confocal projections of DA immunostaining in whole-mount adult brains of wild type and Aprt5 mutants. Scale bars: 100 μm. (D) Quantification of DA immunofluorescence intensity over the entire brain showed a slight increase in DA content in Aprt5 mutants compared to wild-type controls. 6 brains were dissected per experiment and genotype, and 3 independent experiments were performed. Unpaired Student’s t test (**p < 0.01). (E) mRNA levels of tyrosine hydroxylase neuronal form DTH1 are markedly increased in Aprt5 mutant heads compared to wild-type flies. Results of 6 independent RT-qPCR experiments carried out on 3 to 4 different RNA extractions from 20-30 male heads per genotype. Unpaired Student’s t test (**p < 0.01). (F) Conversely, overexpressing Aprt in all cells with the da-Gal4 driver (da>Aprt) reduced mRNA level of DTH1 in heads, as compared to the driver (da/+) and effector (Aprt/+) controls. Mean of 3 independent experiments performed on 3 different RNA extractions from 20-30 male fly heads. One-way ANOVA with Tukey’s post-hoc test for multiple comparisons (*p < 0.05, **p < 0.01). (G) Representative western blot of wild-type and Aprt5 mutant head extracts probed with anti-TH and anti-actin beta antibodies. (H) Quantification of DTH1 protein levels in adult wild-type and Aprt5 mutant heads by western blots showed an increase in DTH1 protein level (60 kDA) in Aprt5 mutants. Actin (Act5C, 42kDa) was used as a loading control. Results are the mean of four determinations in two independent experiments. Unpaired Student’s t test (**p < 0.01).

Relations between Aprt and molecular components of adenosinergic signaling. (A-B) Impacts of the lack of Aprt activity on the adenosinergic system.

(A) Adenosine level was measured in whole flies or heads of Aprt5 flies by UPLC. Compared to wild-type flies, adenosine level was significantly reduced in the mutants. Results of 3 independent experiments performed with 5 males per genotype in triplicates, and 2 independent experiments with 30 heads per genotype in triplicates. Two-way ANOVA with Sidak’s post-hoc test for multiple comparisons (*p < 0.05; **p < 0.01). (B) Aprt5 mutation did not affect AdoR expression but induced a marked increase in adenosine transporter Ent2 mRNA abundance. 3-6 different RNA extractions were performed on 20-30 male heads. Results of 3 to 6 independent experiments. Two-way ANOVA with Sidak’s post-hoc test for multiple comparisons (***p < 0.001; ns: not significant). (C) Null AdoRKGex mutants showed decreased Aprt expression (left panel) and a stronger decrease in Aprt activity (right panel). 4 independent RNA extractions were carried out on 20-30 male heads and 4 independent real-time PCR determinations were done per RNA sample. For Aprt activity, 3 independent determinations were performed on 20 whole flies per genotype. Unpaired Student’s t test (***p < 0.001).

Aprt deficiency triggers a seizure-like phenotype.

(A) At 30 d a.E., Aprt5 mutants need a much longer time than wild-type flies to recover from a strong mechanical shock, showing a bang-sensitive (BS) paralysis comparable to tonic-clonic seizure. Results of 3 independent experiments performed on 50 flies per genotype. Unpaired Student’s t test; **p < 0.01. (B) At 30 d a. E., hemizygous Aprt5 mutants also showed a marked BS phenotype. Results of 2-4 independent experiments performed on 50 flies per genotype. One-way ANOVA with Dunnett’s post-hoc test for multiple comparisons (***p < 0.001). (C) RNAi-mediated downregulation of Aprt in all cells (da>AprtRNAi) also led to BS phenotype in adults at 30 d a.E., but not with the driver and effector controls. Results of 3 independent experiments performed on 50 flies per genotype. One-way ANOVA with Tukeys’s post-hoc test for multiple comparisons (*p < 0.05). (D) Aprt knockdown by RNAi during the adult stage for 3 days before the test was not sufficient to induce bang-sensitivity, suggesting that this phenotype could be caused by a developmental defect or a longer downregulation of Aprt. Results of 2 independent experiments performed on 50 flies per genotype. Two-way ANOVA with Sidak’s post-hoc test for multiple comparisons; ns: not significant. (E, F) The BS phenotype of 30-day-old Aprt5 mutants was rescued by feeding either 500 µM adenosine (ado) (E) or 500 µM N6-methyladenosine (m6A) (F) during all developmental stages plus 5 days before the test. Results of 4 or 6 independent experiments performed on 50 flies per genotype. One-way ANOVA with Tukeys’s post-hoc test for multiple comparisons (***p < 0.001, ns: not significant).

Expression of a pathogenic mutant isoform of human HGPRT induces neurobehavioral defects in flies.

(A-B) Ubiquitous expression of human HPRT1 with da-Gal4. (A) Amplification of human HPRT1 transcripts detected by RT-PCR in head extracts of da>HPRT1-WT and da>HPRT1-I42T flies. A band with lower intensity was also detected in the effector controls (UAS-HPRT1-WT/+ and UAS-HPRT1-I42T/+), and not in the driver control (da/+), which indicates a small amount of driver-independent transgene expression. (B) Quantification of the previous experiment. (C, D) Expression of the LND-associated I42T isoform in all neurons (elav>HPRT1-I42T), but not of the wild-type form (elav>HPRT1-WT), induced an early locomotor SING defect at 15 d a.E. (C) and a prominent bang-sensitivity phenotype at 30 d a.E. (D), as compared to the driver (elav/+) and effector (UAS-HPRT1 I42T/+) only controls. Results of 3 independent experiments performed on 50 flies per genotype. One-way ANOVA with Tukeys’s post-hoc test for multiple comparisons (**p < 0.01; ***p < 0.001; ns: not significant).