An Evolutionarily Conserved Role of Presenilin in Neuronal Protection in the Aging Drosophila Brain

Abstract

Mutations in the Presenilin genes are the major genetic cause of Alzheimer's disease. Presenilin and Nicastrin are essential components of γ-secretase, a multi-subunit protease that cleaves Type I transmembrane proteins. Genetic studies in mice previously demonstrated that conditional inactivation of Presenilin or Nicastrin in excitatory neurons of the postnatal forebrain results in memory deficits, synaptic impairment, and age-dependent neurodegeneration. The roles of Drosophila Presenilin (Psn) and Nicastrin (Nct) in the adult fly brain, however, are unknown. To knockdown (KD) Psn or Nct selectively in neurons of the adult brain, we generated multiple shRNA lines. Using a ubiquitous driver, these shRNA lines resulted in 80-90% reduction of mRNA and pupal lethality-a phenotype that is shared with Psn and Nct mutants carrying nonsense mutations. Furthermore, expression of these shRNAs in the wing disc caused notching wing phenotypes, which are also shared with Psn and Nct mutants. Similar to Nct, neuron-specific Psn KD using two independent shRNA lines led to early mortality and rough eye phenotypes, which were rescued by a fly Psn transgene. Interestingly, conditional KD (cKD) of Psn or Nct in adult neurons using the elav-Gal4 and tubulin-Gal80^ts system caused shortened lifespan, climbing defects, increases in apoptosis, and age-dependent neurodegeneration. Together, these findings demonstrate that, similar to their mammalian counterparts, Drosophila Psn and Nct are required for neuronal survival during aging and normal lifespan, highlighting an evolutionarily conserved role of Presenilin in neuronal protection in the aging brain.

Keywords: Alzheimer’s disease; brain; conditional knockdown; shRNA; γ-secretase.

Figure 1

Generation and validation of UAS-shPsn and UAS-shNct transgenic lines using a ubiquitous Gal4 driver. (A) Schematic gene structures of Psn and Nct and the target regions of the Psn and Nct shRNAs. Gray boxes indicate the 5′- or 3′-UTR. Colored boxes indicate exons. Hairpins indicate the shRNAs and the corresponding target regions. (B, C) Ubiquitous expression of all of the Psn and Nct shRNAs using the Act5c-Gal4 driver result in pupal lethality and ∼80–90% reduction of mRNA levels in Act5c > shPsn and Act5c > shNct whole third-instar larvae, compared to control (Act5c-Gal4/+). qRT-PCR analysis of Psn (B) and Nct (C) mRNA levels was performed using total RNA extracted from five whole third-instar larvae per genotype. Psn and Nct mRNA levels were normalized to rp49 mRNA levels as internal control. n = 4 independent experiments. All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post hoc comparisons, ****P < 0.0001.

Figure 2

Notching wing phenotypes in wing-specific Psn and Nct KD flies. (A) Expression of shPsn or shNct under the control of the c96-Gal4 driver results in thickened veins and/or notching wing margins. Representative adult wing images of control (c96-Gal4/+) and wing-specific Psn (c96 > shPsn) or Nct (c96 > shNct) KD flies are shown. The wing margin is also shown in higher magnification views. Red dots mark thickened veins or blistered phenotypes in the wing margins. Bar, 0.5 mm. (B) Representative adult wing images used to establish the scoring system for quantifying the severity of the wing phenotypes. Score 0: wild-type wing morphology. Score 1: mildly thickened L3 and L4 veins near the margin. Score 2: thickened L2, L3, L4, and L5 veins near the margin. Score 3: more enhanced vein phenotypes. Score 4: severe vein and blistered phenotypes. Score 5: severe margin loss. (C) Quantification of the severity of wing phenotypes using the scoring system shown in (B). At least 20 wings were scored per genotype. All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post hoc comparisons, ****P < 0.0001.

Figure 3

Severe early mortality, climbing defects and rough eyes in elav > shPsn flies. (A) Significant reduction of Psn mRNA level in elav > shPsn2. qRT-PCR analysis of Psn mRNA levels in third-instar larvae (whole larvae or dissected brains) or 3-day-old adults (whole flies or heads only). Psn mRNA levels were normalized to rp49 mRNA levels as internal control. Total RNA was extracted from whole larvae (five larvae per genotype), larval brains (15 brains per genotype), whole adult females (five adults per genotype) or adult female heads only (20 heads per genotype). n = 4 independent experiments. (B) Neuron-specific KD of Psn reduces pupa-to-adult viability. Viability was calculated by dividing the total number of flies by the total number of pupae, and shown as the percentage of pupae surviving to adulthood. No male flies eclosed from elav > shPsn2 pupae (0/219), and the number of adult females (83/219) were lower than anticipated. n = 11 independent experiments; ≥210 flies per genotype (∼20 flies per experiment) were used in the study. (C) Neuron-specific Psn KD results in defects on wing expansion. Percentage of elav > shPsn2 flies with expanded or wrinkled wing phenotypes. 63.9 ± 5.6% of elav > shPsn2 flies had normal expanded wings (EW; red) and 36.1 ± 5.6% of flies had wrinkled wings (WW; orange). n = 4 independent experiments. (D) Neuron-specific Psn KD causes defects in climbing ability. Only elav > shPsn2 females with normal expanded wings were used for the climbing assay. Bar indicates percentage of failed climbers. Age = 3 days, n = 8 independent experiments; ≥150 flies per genotype (∼20 flies per experiment) were used in the study. (E) Neuron-specific Psn KD causes severe mortality. Survival of Gal4 control (elav-Gal4/+, black), UAS control (UAS-shPsn2/+, pink), and neuron-specific Psn KD flies (elav-Gal4/+;; UAS-shPsn2/+) with expanded wings (EW, red) and wrinkled wings (WW, orange). Lifespans were plotted by the Kaplan-Meier method. (F) Neuron-specific Psn KD causes rough eye phenotypes. Representative images of the control and ealv > shPsn2 eyes are shown. Bar, 0.1 mm. All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post hoc comparisons. **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4

Similar but less severe phenotypes associated with a different Psn shRNA line. (A) Neuron-specific Psn KD using a different shRNA line results in significant reduction of Psn mRNA levels in 3-day-old adult heads of elav > shPsn3 males (83 ± 3.2%) and females (54 ± 2.8%), compared to controls. qRT-PCR analysis of Psn mRNA level in adult fly heads expressing shPsn3. Psn mRNA levels are normalized to rp49 mRNA levels as internal control. Total RNA was extracted from 20 adult heads per genotype. n = 4 independent experiments. (B) Neuron-specific Psn KD in elav > shPsn3 reduces pupa-to-adult viability (92/117). Viability was calculated by dividing the total number of flies by the total number of pupae, and shown as the percentage of pupae surviving to adulthood. n = 6 independent experiments, ≥110 flies per genotype (∼20 flies per experiment). (C) elav > shPsn3 causes defects in climbing ability in both males and females. Bar indicates percentage of failed climbers. Age = 3 days, n = 6 independent experiments, ≥110 flies per genotype (∼20 flies per experiment). (D, E) elav > shPsn3 causes severe mortality in male (D) and female (E) flies. Survival of Gal4 controls (elav-Gal4/+, black), UAS controls (UAS-shPsn3/+, pink), and elav > shPsn3 flies (elav-Gal4/+;; UAS-shPsn3/+, red). Lifespans were plotted using the Kaplan-Meir method. (F) elav > shPsn3 male flies exhibit severe rough eye phenotypes. (G) In contrast to females with a stronger shPsn2 line (elav > shPsn2), females of elav > shPsn3 show normal compound eyes. Representative images showing adult eye of controls and elav > shPsn3 flies. Bar, 0.1 mm. All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post hoc comparisons. ***P < 0.001, ****P < 0.0001.

Figure 5

Partial phenotypic rescue of neuron-specific Psn KD using a Drosophila Psn transgene. (A) The UAS-hPSEN1 transgenic lines (BL33811 and BL33812) are not able to rescue the male pupal lethality or rough eye phenotype of elav > shPsn2. Representative images show the adult eye of each genotype. Bar, 0.1 mm (B–D) Using a Drosophila Psn transgenic line, UAS-Psn+14 (BL63243), the male pupal lethality, rough eye phenotype, and defective wing expansion in elav > shPsn2 are rescued. (B) Both male and female of elav > Psn+14 show rough eye phenotypes. elav > shPsn2, Psn+14 female flies show full rescue of the rough eye phenotype exhibited in elav > shPsn2. Representative images of adult eye of each genotype. Bar, 0.1 mm (C) The pupa-to-adult viability is significantly rescued in elav > shPsn2, Psn+14 (20/32; P < 0.01) compared to elav > shPsn2 (26/56). However, elav > shPsn2, Psn+14 flies showed reduced viability compared to control (P < 0.001). Viability was calculated by dividing the total number of flies by the total number of pupae, and shown as the percentage of pupae surviving to adulthood. n = 3 independent experiments. (D) The wrinkled wing phenotype of elav > shPsn2 is significantly rescued in elav > shPsn2, Psn+14 (P < 0.05). Specifically, female elav > shPsn2, Psn+14 flies showed normal wing phenotypes, while male elav > shPsn2, Psn+14 flies showed wrinkled wing phenotypes. Percentage of wrinkled wing phenotypes was calculated by dividing the number of flies with wrinkled wings by the total number of flies. All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Tukey’s post hoc comparisons. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 6

Severe early mortality, climbing defects, and rough eyes in elav > shNct flies. (A, D) Neuron-specific Nct KD using two independent lines, UAS-shNct2 (A) and UAS-shNct3 (D), results in significant reduction of Nct mRNA levels in third-instar larvae compared to controls. qRT-PCR analysis of Nct mRNA level in third-instar larvae (whole larvae or dissected brains) of elav > shNct2 (A) and elav > shNct3 (D). Nct mRNA levels are normalized to rp49 mRNA levels as internal control. Total RNA was extracted from whole larvae (five larvae per genotype) or larval brains (15 brains per genotype). n = 4–5 independent experiments. (B, E) Neuron-specific Nct KD results in decreased pupa-to-adult viability. Viability was calculated by dividing the total number of flies by the total number of pupae, and shown as the percentage of pupae surviving to adulthood. (B) Only 11 females elav > shNct2 eclosed from 225 pupae. n = 9 independent experiments. Total numbers of flies tested: elav-Gal4/+ = 218, UAS-shNct2/+ = 216, and elav > shNct2 = 225 (20 flies per experiment). (E) Only 13 females elav > shNct3 eclosed from 95 pupae; n = 5 independent experiments. Total numbers of flies tested: elav-Gal4/+ = 100, UAS-shNct3/+ = 97, and elav > shNct3 = 95 (20 flies per experiment). (C, F) Neuron-specific Nct KD causes defects in climbing ability. Bar indicates percentage of failed climbers. Age = 3 days. Gal4 and UAS controls: n = 6–8 independent experiments, ≥110 flies per genotype (20 flies per experiment). elav > shNct2: n = 5 flies (C) and elav > shNct3: n = 14 flies (F). (G, H) Females of both elav > shNct2 (G) and elav > shNct3 (H) exhibit rough eye phenotypes. Representative images showing the adult eye of each genotype. Bar, 0.1 mm All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post hoc comparisons. *P < 0.05, **P < 0.01, ****P < 0.0001.

Figure 7

Age-dependent climbing defects and shortened lifespan in adult neuron-specific Psn and Nct KD flies. (A) Strategy for imposing temporal control of shRNA expression using Gal80^ts. (B) Adult neuron-specific conditional KD of Psn (AN-Psn cKD; red) and Nct (AN-Nct cKD; blue) results in significant reduction of Psn (red) or Nct (blue) mRNA level in 10-day-old male fly heads, but no significant difference in 1-day-old male fly heads compared to control. qRT-PCR analysis of Psn and Nct mRNA level in the male adult heads of 1 and 10 day-old flies. Total RNA was extracted from 20 male adult heads. Psn and Nct mRNA levels were normalized to rp49 mRNA levels as internal control. n = 4 independent experiments. All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post hoc comparisons. **P < 0.01, ***P < 0.001. (C) Adult neuron-specific Psn or Nct KD causes defects in climbing ability at the age of 50 days. Bar indicates percentage of failed climbers. n ≥ 5 independent experiments, ≥100 flies per genotype (20 flies per experiment). All data are expressed as mean ± SEM. Statistical analysis was performed using two-way ANOVA with Dunnett’s post hoc comparisons. Two-way ANOVA showed main effects of genotype [F(2, 45) = 6.471; P = 0.0034] and age [F(2, 45) = 556.1; P < 0.0001] with an interaction between these factors [F(4, 45) = 9.905; P < 0.0001]. ns, nonsignificant; ****P < 0.0001. (D–F) Adult neuron-specific Psn and Nct cKD reduce lifespan. AN-Psn/Nct cKD flies showed a greater reduced lifespan relative to AN-Psn cKD and AN-Nct cKD flies. Survival of control (black), AN-Psn cKD (red, D), AN-Nct cKD (blue, E), and AN-Psn/Nct cKD (green, F). Lifespans were plotted by the Kaplan-Meier method. (G–I) Impairment of γ-secretase activity in AN-Psn cDK flies. (G) Strategy for assessing γ-secretase dependent cleavage of the APP-LV reporter system. (H) Western analysis shows dramatic increases of γ-secretase substrate hAPP-CTF-LV in the adult head of male AN-Psn cKD; APP^LV flies, indicating reduction of γ-secretase activity. β-actin was used as loading control. Asterisk marks nonspecific band. (I) Quantification of hAPP-CTF-LV levels in AN-Psn cKD; APP^LV and control fly heads. hAPP-CTF-LV levels were normalized to β-actin and full length APP protein; n = 3 independent experiments. All data are expressed as mean ± SEM. Statistical analysis was performed using unpaired Student’s t-test. ****P < 0.0001.

Figure 8

Age-dependent neurodegeneration in adult neuron-specific Psn and Nct cKD flies. (A) H&E staining in frontal brain sections revealed age-dependent neurodegeneration. Increased number of vacuoles (marked by arrowheads) indicates neurodegeneration in the brain. Bar, 100 μm. (B) Quantification of vacuoles in the brain of control (elav-Gal4/y;; tub-Gal80^ts/+, black), AN-Psn cKD (elav-Gal4/y;; tub-Gal80^ts/UAS-shPsn2, red), AN-Nct cKD (elav-Gal4/y;; tub-Gal80^ts/UAS-shNct2, blue), and AN-Psn/Nct cKD flies (elav-Gal4/y;; tub-Gal80^ts, UAS-shPsn2/UAS-shNct2, green). Neurodegeneration is minimal in the brain of control flies. The brains of AN-Psn cKD and -Nct cKD flies show significant age-dependent neurodegeneration compared to brains of control flies at the age of 30, 40, and 50 days. The brains of AN-Psn/Nct cKD flies show significant age-dependent neurodegeneration compared to the brains of control flies at the age of 20, 30, and 40 days. All AN-Psn/Nct cKD flies died before the age of 50 days. Total number of vacuoles >5 μm in diameter was counted throughout all sections of entire brains. n = 10–17 brains per genotype. All data are expressed as mean ± SEM. Statistical analysis was performed using two-way ANOVA with Dunnett’s post hoc comparisons. Two-way ANOVA showed main effects of genotype [F(3, 197) = 22.77; P < 0.0001] and age [F(4, 197) = 57.92; P < 0.0001], with an interaction between these factors [F(12, 197) = 19.94; P < 0.0001]; ns, nonsignificant; *P < 0.05, **P < 0.01, ****P < 0.001, ****P < 0.0001.

Figure 9

Increased apoptosis in the brain of adult neuron-specific Psn and Nct cKD flies. (A) Apoptotic cells (marked by arrowheads) were identified by TUNEL staining. Bar, 10 μm. (B) Quantification of TUNEL+ cells. The number of TUNEL+ cells is increased in AN-Psn cKD, AN-Nct cKD, and AN-Psn/Nct cKD brains compared to control brains; n = 11–14 brains per genotype at 30 days of age. All data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post hoc comparisons. ****P < 0.0001.