Introduction

Alzheimer’s disease (AD) is the most common neurodegenerative disease characterized by progressive and irreversible decline of memory, cognitive function and language skills, eventually culminating in dementia1,2. The neuro-pathological hallmarks of AD are neurofibrillary tangles (NFTs) containing hyperphosphorylated aggregates of the microtubule-associated protein Tau and extracellular senile plaques (SPs) containing β-amyloid (Aβ) peptides3,4. The various species of Aβ peptides are derived from the amyloid precursor protein (APP), a large single-pass transmembrane protein expressed at high level in the brain5. APP is proteolytically processed in a highly complex fashion by a series of sequential proteases to generate a series of fragments, including Aβ peptides, N-terminal fragment (N-APP) and C-terminal intracellular domain (AICD)5,6. Mutations in APP have been associated with familial susceptibility to AD7,8 and augmented APP expression levels also increase the risk of AD9. In humans, the APP gene is located on chromosome 21. Interestingly, patients with Trisomy 21 (Down’s Syndrome) invariably develop AD10. Duplications of APP locus have also been discovered in some younger patients developing AD11. Though APP has been clearly implicated in the pathogenesis of AD, the genes that modulate APP’s functions remain largely unknown.

In the past decades, increasing evidence suggests that cell cycle re-entry plays a fundamental role in the pathogenesis of AD12,13,14. Neurons in the adult central neuronal system have classically been recognized as terminally differentiated, that are arrested in the G0 phase of cell cycle in a non-dividing and non-replicating state. However, in AD brain, vulnerable neurons display an activated cell cycle phenotype - abnormal elevation of cell cycle markers and re-expression of cell cycle regulators, suggesting these neurons have re-entered the cell cycle, yet they lack the ability to complete the cell cycle, which results in aberrant neuronal death15,16,17,18.

Polo kinases belong to a highly conserved Ser/Thr kinase family originally identified in Drosophila melanogaster19. Polo kinases contain the canonical serine/threonine kinase domain and a unique Polo Box Domain (PBD) that docks the kinases to target proteins20. Budding yeast and Drosophila each encodes only one Polo kinase, named as Cdc5 and Polo, respectively21. Five Polo-like kinases, Plk1, Plk2/SNK, Plk3/CNK/FNK, Plk4/SAK and Plk5, have been identified in mammals22. Polo kinases have been previously reported as key regulators of multiple aspects of the cell cycle, including mitotic entry, centrosome organization, spindle formation, chromosome segregation, mitotic exit and cytokinesis23,24,25,26,27,28,29,30,31. Intriguingly, Plk1 level is elevated in susceptible hippocampal and cortical neurons of AD patients32,33, yet the role of Plk1 in AD pathogenesis remains elusive.

Drosophila has been used as a powerful model organism to study the pathology and treatment of AD34,35,36,37,38. In the present study, we expressed human APP in the same pattern as its Drosophila homolog APPL and observed morphological abnormalities in the adult wing and larval neuromuscular junction (NMJ), locomotion defect, shortened lifespan and impaired choice ability. These phenotypes mimic several symptoms of AD pathology in a dose- and age-dependent manner. We found in a genetic screen that loss of Polo ameliorated APP-induced AD-like symptoms, presumably by suppressing APP-induced aberrant cell cycle re-entry.

Results

Loss of polo suppresses APP-induced wing expansion defect

To investigate the in vivo functions of human APP in animal development, we reasoned the best solution is to express APP in the same pattern as its Drosophila homolog APPL, which is expressed primarily in the central nervous system (CNS) in adults and 3rd instar larvae (Figure S1). To this end, we expressed UAS-APP under the control of Appl-Gal4 (Appl > APP) and observed a severe wing expansion defect (96% unexpanded, Fig. 1b; 4% partially expanded, Fig. 1c), as compared with Appl-Gal4 controls (Fig. 1a). The wing expansion behavior is regulated by the central nervous system whose dysfunction results in wing expansion deficit39. Intriguingly, expression of two truncated APP proteins with deletions in the C-terminal intracellular domain (AICD), APPΔCT (a truncated form of APP lacking AICD) and APPΔNPTY (APP with a deletion of the NPTY motif in AICD), failed to recapitulate the wing defect (Fig. 1d and Figure S2b), indicating that AICD (or more specifically, the NPTY motif) is indispensable for APP to impede wing expansion. This result is in agreement with the recent findings that AICD is necessary for APP-induced neuronal cell death in Drosophila40,41.

Figure 1
figure 1

Loss of polo suppresses APP-induced wing expansion defect.

(a–j) Light images of adult female flies are shown. Compared with Appl-Gal4 controls (a), expression of APP (b and c), but not APPΔCT (d), produced wing expansion defect in adults, almost half of which were suppressed partially or fully by RNAi-mediated down-regulation of polo (g–j) or in heterozygous polo1 mutant (k,l), but remained unaffected by expressing UAS-Dcr2 (e,f). (m) The knocking-down efficiencies of two independent polo-RNAi lines are measured by qRT-PCR assay. It was performed on cDNA isolated from brains of 3rd instar larvae for the indicated genotype. (n) Statistical analysis of the wing phenotype shown in b,c and e–l. The percentage of adult wing with morphological defects is shown. n > 300 for each genotype. The crosses were performed at 25 °C.

Full size image

To identify additional factors that regulate APP functions in vivo, we performed a genetic screen in Drosophila for UAS-RNAi lines that could dominantly suppress APP-induced wing phenotypes. We found that expression of two independent polo RNAi lines could partially suppress APP-induced wing defect (polo-IR-1, 42% partially expanded and 18% fully expanded; polo-IR-2, 37% partially expanded and 20% fully expanded, Fig. 1g–j,n). The knockdown efficiencies of the two polo-RNAi lines were verified by quantitative real-time PCR (qRT-PCR, Fig. 1m). To rule out the possibility that the effect of polo-RNAi on APP-induced phenotype is a result of Gal4 titration by another UAS sequence, we co-expressed UAS-Dicer2 (UAS-Dcr2) along with UAS-APP and found that UAS-Dcr2 could not alter APP-induced wing defect (94% unexpanded and 6% partially expanded, Fig. 1e,f,n). To further confirm the role of polo in APP-induced wing defect, we introduced polo1, a hypomorphic polo allele42, into Appl > APP background. We found that APP-induced wing defect was partially suppressed in heterozygous polo1 mutants (35% partially expanded and 8% fully expanded, Fig. 1k,l,n). Together, these data suggest that Polo is necessary for APP to impede wing expansion.

As Plk1 level is elevated in brain tissues of AD patients32,33, we wondered whether Polo expression was altered by expressing APP in Drosophila. To this end, we first examined polo mRNA in 3rd instar larval brain by in situ hybridization and qRT-PCR assay. We found polo mRNA level was not significantly affected by expressing APP (Figures S3a, S3b and S4). We next checked Polo protein level monitored by a GFP-Polo reporter43 and found GFP-Polo fusion protein was not considerably upregulated in the eye disc posterior to the morphogenetic furrow upon APP expression driven by GMR-Gal4 (Figures S3e and S3f). On the other hand, both polo mRNA expression in the brain and GFP-Polo expression in the eye disc were significantly downregulated by expressing two polo RNAi (Figures S3c, S3d, S3g and S3h), which validate the knockdown efficiencies of RNAi lines. Since Polo is a key component of the spindle assembly checkpoint (SAC) pathway29,44, we also examined whether overexpression of APP could affect the expression of other SAC pathway components. We performed qRT-PCR assay and found that aurA, aurB and msp1 expression in the brain were not significantly changed upon APP expression (Figure S4). Together, these results suggest that expression of APP has no effect on the expression of Polo or other components of the SAC pathway.

Loss of polo suppresses APP-induced larval NMJ and locomotion defects

The Drosophila neuromuscular junction (NMJ) is an excellent model for studying synapse formation and function45,46,47. Previous studies reported that expression of APP or APPL in the motor neurons disrupted axonal transport and increased the number of boutons48,49,50,51,52. We expressed APP specifically under the control of Appl-Gal4 and examined the synapse of larval segment A3, muscle 6/7. We observed a significant increase in the total number of boutons and branches at the NMJ in APP-expressing larvae (Fig. 2a,f,k and Figure S5a). To determine whether loss of polo could suppress the synaptic defects at the NMJ, we introduced polo RNAi or polo1 into Appl > APP background. Depletion of polo, although exhibited no discernable effect on NMJ synapse morphology (Fig. 2c–e,k and Figure S5a), significantly suppressed APP-induced increased number of boutons and branches at NMJ (Fig. 2h–k and Figure S5a). The type I boutons formed on larval muscles 6/7 could further be subdivided into type I big (1b) and type I small (1s) boutons53, which differ not only in their structural properties, but also in their functional properties. We found that the number of type 1b bouton was significantly reduced, whereas that of type 1s increased in Appl > APP larvae (Figures S5b and S5c). Both phenotypes were partially suppressed by depletion of polo, which by itself had no effect (Figure S5b and S5c).

Figure 2
figure 2

Loss of polo impedes APP-induced larval NMJ and locomotion defects.

(a–j) Confocal images of the synapse at segment A3, muscle 6/7 stained with HRP (red) and Dlg (green). Compared with controls (a), expression of Dcr2 (b) or two independent polo-RNAi lines (c,d), or heterozygous polo1 mutant (e) did not significantly influence the NMJ phenotype. Expression of APP (Appl > APP) resulted in an increase of branch and total bouton number (f), which was suppressed by RNAi-mediated knocking-down of polo (h,i) or in heterozygous polo1 mutant (j), but remained unaffected by expressing Dcr2 (g). Scale bar represents 25 μm. (k) Statistical analysis of total bouton number per muscle area in a–j is shown. Error bars mean + S.E.M. ***P < 0.001, n.s, not significant. One-way ANOVA Turkey’s multiple comparison test was used to determine significance between multiple different genotypes. n > 25. (l) APP expression induced a crawling deficit in 3rd instar larvae as compared with Appl-Gal4 controls. Knocking down polo significantly rescued the crawling deficits in APP-expressing larvae, but did not affect the crawling ability in control larvae. One-way ANOVA Turkey’s multiple comparison test was used to determine significance between multiple different genotypes. Error bars mean + S.E.M. ***P < 0.001, n.s, not significant. The crosses were performed at 25 °C.

Full size image

NMJs are functionally associated with larval locomotion ability. Consistent with the NMJ phenotype, expression of APP driven by Appl-Gal4 dramatically reduced the crawling speed of larvae, while loss of polo ameliorated this APP-induced larval locomotion defects (Fig. 2l). As a negative control, expression of Dcr2 did not suppress APP-induced larval NMJ and locomotion abnormalities (Fig. 2g,k,l and Figure S5). Together, these results indicate that Polo is required for APP-induced larval NMJ and locomotion defects.

Loss of polo suppresses APP-induced adult locomotion defects

Negative geotaxis (climbing) assay has been extensively applied in Drosophila to assess nervous system dysfunction in studying neurodegenerative diseases38. We found that expression of APP driven by Appl-Gal4 (Appl > APP) resulted in a climbing deficit in adult flies (Fig. 3a). While Appl-Gal4 control flies were able to climb at a rate of 1.7 cm/second, Appl > APP flies showed a dramatic reduction to less than 0.2 cm/second (Fig. 3a). We found that expression of polo RNAi but not Dcr2 and heterozygosity of polo1, could slightly but significantly suppress APP-induced climbing deficit (Fig. 3a). As a control, depletion of polo by RNAi or mutation had no apparent effect on flies’ climbing ability (Fig. 3a). Thus, loss of polo improves the locomotor dysfunction caused by APP expression.

Figure 3
figure 3

Loss of polo ameliorates APP-induced locomotion defect and shortened lifespan in adults.

(a) Compared with Appl-Gal4 controls, Appl > APP-induced adult climbing deficit was significantly rescued by loss of polo (Appl > APP + polo-IR-1, Appl > APP + polo-IR-2 and Appl > APP + polo1), but not the expression of Dcr2. One-way ANOVA Turkey’s multiple comparison test was used to determine significance between multiple different genotypes. Error bars mean + S.E.M. ***P < 0.001, n.s, not significant. The number of flies tested per genotype was n > 180. The crosses were performed at 25 °C. (b) Compared with Appl-Gal4 controls, expression of APP resulted in a drastically shortened lifespan, which was suppressed by loss of polo. The percentage survivorship was plotted against age. Reported p values comparing median lifespans in the right table are from Mantel-Cox log-rank statistical analysis. The number of flies tested per genotype was n > 210. The crosses were performed at 25 °C.

Full size image

Loss of polo rescues APP-induced adult lifespan shortening

Lifespan assay provides a statistically robust test of the neurological integrity of a fly. Previous studies reported that expression of human neurodegenerative disease genes would reduce lifespan of Drosophila34,36,54. Consistently, expression of APP driven by Appl-Gal4 resulted in a drastically shortened lifespan, which was partially rescued by depletion of polo, but remained unaffected by expressing Dcr2 (Fig. 3b). Together, these results suggest that loss of polo rescues lifespan of APP-expressing flies.

Loss of polo attenuates adult specific APP-induced locomotion and lifespan defects

Given that AD is an age-related disease, to better study the pathological functions of APP in adult neurodegeneration, we sought to express APP specifically in the adult stage, thus avoiding its earlier developmental effects. Since the Gal4/UAS system is sensitive to temperature, to inhibit APP expression in earlier development, we raised the embryos at a lower temperature (17 °C) throughout the larva and pupa stages and shifted the freshly eclosed flies to 25 °C to allow the expression of APP in adulthood. Although Appl > APP flies reared at 25 °C displayed abnormal wing phenotype (Fig. 1b,c), those raised at 17 °C did not produce any obvious morphological defects (Figures S6a and S6b). In addition, no apparent climbing defect was observed in 2-day old Appl > APP flies, as compared with the controls (Fig. 4a). These data suggest that the developmental effects of APP were effectively avoided.

Figure 4
figure 4

Loss of polo attenuates adult-specific APP-induced locomotion decline.

(a) Adult-specific expression of APP (Appl > APP) had no discernable effect on climbing ability in 2-day old flies, but displayed a climbing deficit in 17- and 27-day old flies, as compared with Appl-Gal4 controls. Knocking down polo significantly rescued APP-induced climbing deficits. (b) Knocking down polo did not affect the climbing ability in control flies at different time-points. (c) Compared with controls (Appl-Gal4), Appl > APP flies displayed a much severer age-dependent decline of climbing ability, which was ameliorated by knocking-down polo. (d) Knocking down polo did not affect the age-dependent decline of climbing ability in control flies. Flies were raised at 17 °C and shifted to 25 °C after eclosion. One-way ANOVA Turkey’s multiple comparison test was used to determine significance between multiple different genotypes. Error bars mean + S.E.M. ***p < 0.001, n.s, not significant. The number of flies tested per genotype was n > 150.

Full size image

Previous studies suggested that flies’ locomotion capability diminishes with aging37,54,55,56. Consistently, we observed a decreased climbing ability in control flies at 17 days and 27 days, respectively (Fig. 4a). Interestingly, the climbing ability of APP-expressing flies declined more dramatically with aging (Fig. 4a), indicative of an age-related neurodegeneration resulted from adult-specific expression of APP. Expressing polo RNAi, but not Dcr2, considerably rescued the locomotion defect in APP-expressing flies at the corresponding age points (Fig. 4a), though loss of polo by itself did not affect flies’ climbing ability as compared with age-matched controls (Fig. 4b).

To accurately quantify the development of age-dependent locomotion disability, we defined a performance index (PI) - the relative decline of the climbing ability between aged and young (2-day old) flies in a climbing assay: PI = velocityaged/velocityyoung. With aging, the Appl-Gal4 control group displayed a gentle decline in the climbing ability (Fig. 4c), whereas APP-expressing flies began to display a miniscule but statistically significant locomotor dysfunction at 7-day old, which became more obvious and drastic at 17- and 27-day old (Fig. 4c). Knocking-down polo significantly ameliorated the decline of locomotor function in APP-expressing flies (Fig. 4c), but did not affect the climbing ability in control flies (Fig. 4d).

We also checked the effect of adult-specific expression of APP on lifespan at 29 °C. Consistent with the locomotion disability, APP expression resulted in a significantly shortened lifespan, which was partially rescued by knocking-down polo, but remained unchanged by expressing Dcr2 (Figure S7). These results indicate that loss of polo could ameliorate adult-specific APP-induced locomotor dysfunction and lifespan shortening.

Loss of polo ameliorates APP-induced male courtship choice defect

Cognitive decline is another critical hallmark of AD1. Male courtship preference behavior, a paradigm for decision making in animals, to a certain degree, reflects the discriminative and cognitive ability. Previous study reported that Drosophila males, when provided with both younger and older virgin females, tended to be fastidious and preferred to court younger ones57. Interestingly, this courtship preference was eliminated in aged males or males expressing APP in the courtship specific neurons driven by fru-Gal457, suggesting that expression of APP in Drosophila could mimic aging-induced choice disorder and cognitive impairment. Consistent with the previous report, we found 3-day old fru-Gal4 control males preferred younger virgin females to older ones, while fru > APP males were unable to distinguish younger females from older ones (Fig. 5a). Knocking-down polo did not influence males’ courtship preference behavior, but ameliorated APP-induced male choice dysfunction and partially restored their courtship preference to younger mates (Fig. 5a).

Figure 5
figure 5

Loss of polo restores APP-induced male courtship choice disorder.

(a) Courtship index of 3-day-old naive males in choice assays toward younger virgin females (white) and older ones (grey) are shown. The courtship index of control males (fru-Gal4) to younger virgin females is much higher than their courtship index to older ones. Expression of APP in male courtship circuit (fru > APP) eliminates male’s courtship preference to younger mates, while knocking down polo restores the choice preference. (b) Preference index of 3-day-old naive males in courtship choice assays. Mean ± standard error of the mean (SEM), n > 28 for each genotype; Kruskal-Wallis test, Dunn's post-hoc was used to determine significance for each comparison. ***p < 0.001, *p < 0.05, n.s, not significant. The crosses were performed at 25 °C. Box-and-whisker plots for CIs show 1–99 percentiles and means (+), n > 28 for each genotype; Wilcoxon matched pairs test was used to determine significance between intragroup courtships toward younger females and older ones.

Full size image

To accurately quantify the extent of a male’s preference for younger or older females, we also checked the preference index (PI) that indicate the relative difference between male’s courtship percentage toward younger females and that toward older ones in a choice assay: PI = (CIy-CIo)/(CIy + CIo)57. Consistently, the PI of APP-expressing flies drastically dropped compared with that of controls and was significantly suppressed by knocking down polo (Fig. 5b). Thus, these results indicate that Polo is also required for APP-induced cognitive disability.

Loss of polo impedes APP-induced retinal degeneration and cell cycle re-entry

Previous work suggested that expression of human neurodegenerative disease genes induced neurotoxicity in fly eye58,59,60. To investigate the mechanism by which loss of polo ameliorates APP-induced neurotoxicity, we drove APP expression in developing eyes by the GMR-Gal4 driver61. We found that pan-retinal expression of APP resulted in eye deterioration characterized by small and rough eyes with reduced pigmentation (Fig. 6a,e,i). Although knocking-down polo alone had no distinguishable effect on eye development (Fig. 6c,d,i), it partially suppressed APP-induced retina degeneration (Fig. 6g–i), suggesting Polo is indispensable for APP-induced retina toxicity in adult eyes.

Figure 6
figure 6

Loss of polo suppresses APP-induced retina degeneration and cell cycle re-entry.

(a–h) Light images of Drosophila adult eyes are shown. The white dashed lines outline the eye contour. Compared with the control (a,b), expression of APP (GMR > APP) induced small and rough eyes with reduced pigmentation (e). Loss of polo did not affect eye development (c,d), but suppressed APP-induced retina degeneration (g,h). (i) Statistical analysis of eye size shown in (a–j). The fold change of eye size compared with control is shown. The size of adult eye from control group (GMR-Gal4) was normalized to a value of 1. "Fold change" represents the size of adult eye from indicated genotype normalized with the control. One-way ANOVA Turkey’s multiple comparison test was used to determine significance between multiple different genotypes. Error bars mean + S.E.M. ***P < 0.001, n.s, not significant, n > 15 for each genotype. The crosses were performed at 25 °C. (j–q) Fluorescent images of 3rd instar larval eye discs stained with phospho-histone 3 (PH3) showing the number of mitotic cells in M phase. White arrowheads indicate the morphogenetic furrow (MF). Compared with the control (j,k), reduction of polo did not affect the PH3 level (l,m). Expression of APP significantly upregulated the level of PH3 posterior to MF (n), which was suppressed by loss of polo (p,q). (r) Statistical analysis of mitotic cell number posterior to the MF in (j–q). The number of PH3-positive cells posterior to the MF is shown. One-way ANOVA Turkey’s multiple comparison test was used to determine significance between multiple different genotypes. Error bars mean + S.E.M. ***P < 0.001, n.s, not significant, n > 15 eye discs for each genotype. The crosses were performed at 25 °C.

Full size image

Given that Polo is a key regulator of cell cycle and aberrant cell cycle-triggered cell death has been associated with neuronal degeneration in AD brain18, we wondered whether abnormal cell cycle re-entry is involved in GMR > APP-induced eye degeneration. To address this issue, we examined cell cycle state of third instar larval eye discs. We incubated eye imaginal disc with Bromo-2′-deoxyuridine (BrdU) to ascertain whether APP could induce cells re-entering S phase or re-synthesizing DNA. Compared with control eye discs, expression of APP by GMR-Gal4 significantly increased the number of cells incorporated BrdU (Figure S8). Since previous study also reported that the level of phospho-histone H3 (PH3), a marker of M phase, increased in hippocampal neurons in AD patients62, we wondered whether expression of APP could further promote cells to go through the G2/M transition into M phase. In control eye discs, only a few scattered cells were labeled with PH3 posterior to the second mitotic wave (SMW), suggesting that most cells posterior to SMW had exited the cell cycle (Fig. 6j)63. Expression of APP posterior to the morphogenetic furrow (MF) driven by GMR-Gal4 increased the number of PH3 + cell in the posterior part of eye discs (Fig. 6n,r), suggesting that expression of APP promoted cells re-entering cell cycle to M phase. Knocking-down polo significantly suppressed GMR > APP-induced increase of PH3 + cells (Fig. 6p–r), suggesting that loss of polo is able to abrogate APP-induced neurotoxicity by preventing aberrant cell cycle re-entry. These results likely suggest that knocking-down polo partially suppresses APP-induced retinal degeneration by preventing APP-induced cell cycle re-entry.

Discussion

Drosophila has emerged as an excellent animal model to study human neurodegenerative diseases36,41,64,65. Ectopic expression of AD-related proteins (various isoforms of Tau proteins, Aβ peptides and APP) in flies produced a number of phenotypes, which greatly contributed to the understanding of their in vivo functions and the pathogenesis of AD18,34,38,41,50,54. APP belongs to an evolutionary conserved protein family that also includes the mammalian APLP1 and APLP2, the Drosophila APPL and the C. elegans APL-166. APPL was reported to be exclusively expressed in the nervous system in embryonic stages67, yet its expression in later stage remains elusive. Here, using UAS-GFP to track the expression pattern of Appl-Gal4, we found that Appl is also predominantly expressed in the larval and adult CNS (Figure S1). Thus, compared with previous studies expressing APP pan-neuronally by elav-Gal4, here we generated a fly AD model by expressing APP in the same pattern as Appl. Based on this model, we performed a genetic loss-of-function screen to search dominant modifiers of ectopic APP-induced phenotypes. Genes identified from the screen presumably function in Appl-expressing neurons and are more likely to interact with Appl in development.

Interestingly, we found that expression of APP driven by Appl-Gal4 caused wing expansion defect. As an insect that would undergo complete metamorphosis, in Drosophila, the hardening of the expanded wings marks the end of morphological development68,69,70. Previous studies reported that the wing-expansion program is governed by neuroendocrine networks39,70,71, for instance, the neurohormone bursicon functions within the Drosophila CNS to modulate wing expansion behavior39. In addition, the network underlying wing expansion has been shown to include a dozen of neurons implicated in both larval and pupal ecdysis in Drosophila72,39,73,74,75. For example, suppression of excitability of bursicon-expressing neurons or ablation of CCAP-expressing neurons by cell death gene during development would lead to a severe wing expansion deficit in adult flies39,73,76. Accordingly, expression of APP by Appl-Gal4 likely caused malfunction or cell death in the CNS, resulting in the wing expansion deficits. Interestingly, AICD is necessary for APP to generate this phenotype (Fig. 1d). This result is in agreement with recent findings that AICD is required for APP-induced cell death in Drosophila40,41.

Polo kinase is one of the best characterized Ser/Thr protein kinase. As a key mitotic regulator, Polo kinase governs multiple aspects of mitosis including mitotic entry, centrosome maturation, spindle formation, chromosome segregation and cytokinesis23,24,25,26,27,28,29,30,31. Recent studies reported that Polo was involved in activation of the anaphase-promoting complex/cyclosome (APC/C) and allowed mitotic exit29,44. Genetic and biochemical studies have shown that Polo plays crucial role in controlling Spindle Assembly Checkpoint (SAC) pathway to keep the maintenance of genomic stability during mitosis and is required for correct localization of AurB and promotes recruitment of Mps129,44.

Intriguingly, Plk1, the homolog of Polo, is highly present in susceptible hippocampal and cortical neurons of AD patients compared with age-matched controls32. However, the exact role of Plk1 in the pathology of AD has remained unclear. In this study, we examined both mRNA and protein expression levels of polo and found that overexpression of APP did not significantly affect the expression of polo (Figures S3 and S4). In addition, we did not observe any obvious alteration on the expression of aurB and msp1 (Figures S4), which encode important components of SAC pathway. Thus, up-regulation of Plk1 in neurons of AD patients may not be a direct outcome of elevated APP activities.

In the present study, we expressed human APP, an AD causative gene, in the same temporal-spatial pattern as its Drosophila homolog Appl. By taking advantage of the temperature-dependent activity of the Gal4/UAS system, we raised flies at different temperatures to investigate the pathological outcomes of APP expression in development or adult-onset. We demonstrated that expression of APP resulted in morphological defect in adult wings, synaptic abnormalities in larval NMJ, locomotion decline, male choice dysfunction and shortened lifespan. Some of the phenotypes resemble AD-like symptoms. Genetic depletion of polo ameliorated APP-induced morphological and behavioral defects. Finally, loss of polo suppressed APP-induced eye degeneration and cell cycle re-entry. These findings not only help us gain further understandings of the physiological and pathological functions of APP, but also shed light on discovering a potential cure or prevention of AD.

Materials and Methods

Fly Strains

Unless otherwise indicated, flies were kept on a cornmeal and agar medium at 25 °C according to standard protocols and the data are depicted from female flies.

Drosophila strains used include: UAS-APP, UAS-APP695ΔCT and Appl-Gal440, GMR-Gal461, Frutless-Gal457 and GFP-Polo43 were previously described. UAS-APPΔNPTY (29869), UAS-polo-IR-1 (33042), UAS-polo-IR-2 (35146), UAS-Dcr2 (24651), UAS-GFP (32198) and polo1 (546) were obtained from the Bloomington Drosophila Stock Center.

Behavioral testing

For behavioral experiments, the Appl-GAL4 line was outcrossed with w1118 line for eight generations.

For larvae crawling assay, the procedures were performed as described40. 3rd instar larvae were collected and rinsed in 1% PBS and subsequently transferred onto 3% agarose plate. Larvae were allowed to rest for 3 min before videotaping. The locomotion behavior was analyzed by the Nikon software (NIS-Elements D, Nikon). Each genotype was tested with five larvae and each larva was tested for five times. For the track of the larvae, each genotype was recorded for 10 min and repeated for five times.

For Rapid Iterative Negative Geotaxis (climbing) assay, a modified version of Nichols was used77. Briefly, Flies were collected within 24 hours after eclosion. 25 ~ 30 flies were placed in a vertical vial (20 cm height, 2.5 cm diameter) and tapped to the bottom of the vial. After 5 seconds, a picture was taken and the average height climbed was then recorded. Velocity defined as (height/5s) was calculated. Each analysis was repeated 4 times with 60 seconds resting interval. The number of flies tested per genotype was n > 100. For Fig. 3, the analysis was performed 2 days post-eclosion (d.p.e.). For Fig. 4, the flies were raised at 17 °C and shifted to 25 °C after eclosion. The analysis was tested serially for 2 ~ 27 d.p.e. The performance index (PI) was defined as Velocityn d.p.e./Velocity2d.p.e.

For lifespan analysis, flies were collected within 24 hours after eclosion and placed in a food vial at a density of 20 ~ 25 flies per vial, transferred to fresh vials every 2 to 3 days and dead flies were counted at that time. The number of flies tested per genotype was n > 200. Data are presented as survival curves and analysis was performed using log-rank tests to compare between groups.

For male courtship choice assay, the procedures were performed as described57. Courtship behavior assays were performed at approximately the same time each day (within 1 hour at the beginning of the 12-hour illumination half of the cycle) in round observation chambers (1.5 cm in diameter and 0.3 cm deep) by observing courtship behavior of males (3-day old). Choice-assays were performed by pairing a naive male with 2 younger (3 ~ 5-day old) wild type virgin females and 2 older (30 ~ 35-day old) ones. All tests were recorded for 10 minutes with an HDR-CX270 digital video camera (Sony) and analyzed with Noldus EthoVision XT software (Noldus Information Technology). The courtship index (CI) was calculated as the percentage of time that a male courted the females during a 10-minute period. The preference index (PI) defined as (CIy-CIo)/(CIy + CIo) was calculated. The number of male flies tested per genotype was n > 28.

Immunohistochemistry and BrdU labelling

Eye discs were dissected from 3rd instar larvae in cold PBS and fixed in 4% paraformaldehyde. After proper washes, the discs were blocked in 10% horse serum and stained with primary antibody (rabbit anti-PH3, Sigma). Subsequently, fluorescence conjugated secondary antibodies were used for signal detection.

For BrdU labelling78, eye discs were dissected from 3rd instar larvae in PBS, incubated with 200 μg/ml BrdU (Sigma) in Schneider’s media for 40 min then fixed in 5% formaldehyde, then washed in PBST and hydrolyzed in 2 N HCl. After proper washes, the discs were blocked in 10% horse serum and stained with primary antibody (mouse anti-BrdU, Becton Dickinson); subsequently, fluorescence conjugated secondary antibodies were used for signal detection.

For NMJ, 3rd instar larvae were dissected in PBS and fixed in 4% paraformaldehyde for 20 minutes and washed with PBS containing 0.3% Triton X-100 (PBT). Larval body walls were stained overnight with primary antibody (anti-Dlg and HRP-TRITC) in 1% normal donkey serum (NDS) in PBT at 4 °C, followed by 3 washes in PBT, 2 hours incubation with secondary antibody and Phalloidin-TRITC diluted in 1% NDS in PBT at room temperature, 4 washes with PBT. Larval body walls were mounted in Vectashield (Vector Labs, H-1000). Confocal images were collected from Leica confocal microscope SP5 equipped with an oil immersion objective (X63 HCX PL APO 1.25). Leica Application Suite Advanced Fluorescence software was used to capture, process and analyze images.

In Situ Hybridization

In situ hybridization was performed on larvae brains as previously described79. Digoxigenin (DIG)-labeled (Roche) polo RNA probes were generated by transcribing plasmid pGEM-T Easy, with SP6 polymerase (antisense probe) or with T7 polymerase (sense probe), respectively.

Quantitative Real Time RT-PCR

Total RNA was isolated from the brains of wandering third instar larvae from different genotypes using TRIzol reagent (Invitrogen). n > 60 brains of indicated larvae were collected. RT-PCR was performed as previously described80. It was performed in triplicate and the fold change was calculated by using the ΔΔCT method using RP49 as control. The results from 4 independent tests. Primers for polo are kindly provided by Dr. Lipsick at Stanford University81. Primers for Rp49 are kindly provided by Dr. Ketu Mishra at Yale University. The primers for the target genes analyzed are as follows: for aurB, forward primer 5′TGATGTTCAAAGAGGAGCTGC 3′ and reverse primer 5′-CTCGTCGTGGAACCAAGTGAG-3′; msp1, forward primer 5′ -AGGACGACTTTAACACGCCAT-3′ and reverse primer 5′ -CCCTCGTTTGGACTTGGAAAGA-3′; aurA, forward primer 5′ -AGCCCAACAGCGAGAATATGG-3′ and reverse primer 5′ -GGAAGCTATGGAATTGGAGCCT-3′

Additional Information

How to cite this article: Peng, F. et al. Loss of Polo ameliorates APP-induced Alzheimer's disease-like symptoms in Drosophila. Sci. Rep. 5, 16816; doi: 10.1038/srep16816 (2015).