miR-277 targets the proapoptotic gene-hid to ameliorate Aβ42-mediated neurodegeneration in Alzheimer's model

Abstract

Alzheimer's disease (AD), an age-related progressive neurodegenerative disorder, exhibits reduced cognitive function with no cure to date. One of the reasons for AD is the accumulation of Amyloid-beta 42 (Aβ42) plaque(s) that trigger aberrant gene expression and signaling, which results in neuronal cell death by an unknown mechanism(s). Misexpression of human Aβ42 in the developing retina of Drosophila exhibits AD-like neuropathology. Small non-coding RNAs, microRNAs (miRNAs), post-transcriptionally regulate the expression of their target genes and thereby regulate different signaling pathways. In a forward genetic screen, we identified miR-277 (human ortholog is hsa-miR-3660) as a genetic modifier of Aβ42-mediated neurodegeneration. Loss-of-function of miR-277 enhances the Aβ42-mediated neurodegeneration. Whereas gain-of-function of miR-277 in the GMR > Aβ42 background downregulates cell death to maintain the number of neurons and thereby restores the retinal axonal targeting defects indicating the functional rescue. In addition, gain-of-function of miR-277 rescues the eclosion- and climbing assays defects observed in GMR > Aβ42 background. Thus, gain-of-function of miR-277 rescues both structurally as well as functionally the Aβ42-mediated neurodegeneration. Furthermore, we identified head involution defective (hid), an evolutionarily conserved proapoptotic gene, as one of the targets of miR-277 and validated these results using luciferase- and qPCR -assays. In the GMR > Aβ42 background, the gain-of-function of miR-277 results in the reduction of hid transcript levels to one-third of its levels as compared to GMR > Aβ42 background alone. Here, we provide a novel molecular mechanism where miR-277 targets and downregulates proapoptotic gene, hid, transcript levels to rescue Aβ42-mediated neurodegeneration by blocking cell death. These studies shed light on molecular mechanism(s) that mediate cell death response following Aβ42 accumulation seen in neurodegenerative disorders in humans and provide new therapeutic targets for neurodegeneration.

Fig. 1. Gain-of-function of miR-277 modulates the Aβ42-mediated neurodegeneration.

A Wild-type larval eye imaginal disc develops in to B an adult eye comprises of ~800 organized ommatidia. Note that the eye imaginal disc is stained with a membrane-specific marker, discs large (Dlg; green), and a pan neural marker ELAV (red) that marks the photoreceptors (C, D) GMR-Gal4, drives the expression of target transgenes in the developing eye, serves as a control. E, F Misexpression of human Aβ42 in the developing (E) eye imaginal disc (GMR > Aβ42) leads to highly reduced (F) adult eye phenotype with fused ommatidia. G, H GMR>miR-277, I, J GMR>miR-277 mutant, and K, L GMR > Df(3 R)miR-277KO eye imaginal discs and adult eye also served as controls. M, N Gain-of-function of miR-277 in the background of GMR > Aβ42 (GMR > Aβ42+ miR-277) results in significant rescue in eye disc and adult eye as compared to the GMR > Aβ42. O–R However, reducing miR-277 levels in GMR > Aβ42 flies using (O, P) miR-277 mutant (GMR > Aβ42+ miR-277 mutant), (Q, R) Df(3 R)miR-277KO [GMR > Aβ42 + Df(3 R)miR-277KO] enhances GMR > Aβ42 neurodegenerative phenotype. S Bar graph shows frequency as average between 3 repetitions. Two hundred flies were counted per repetition (200 × 3 = 600) to calculate the frequency for each genotype (1. GMR-Gal4, 2. GMR > Aβ42, 3. GMR > Aβ42+miR-277, 4. GMR > Aβ42+miR-277 mutant, 5. GMR > Aβ42 + Df(3 R)miR-277KO). Statistical analysis was performed using the student’s t test for independent samples. T Quantitative analyses of severity score of neurodegenerative phenotypes(s) in eye. Flies from each genotype were selected for scoring according to the criteria described in methods section. Comparisons were made using non-Parametric: Mann–Whitney t test. U Quantitative analyses of area of the eye. The surface area of the eye (within white dotted line) was calculated using Image J. Statistical analysis was performed using the student’s t test for independent samples. The surface area is significantly rescued in GMR > Aβ42+miR-277 (n = 5; p = 0.0000000146) whereas significantly reduced in GMR > Aβ42+miR-277 mutant (n = 5; p = 0.017) and GMR > Aβ42 + Df(3 R)miR-277KO (n = 5; p = 0.02) as compared to GMR > Aβ42. V Relative expression of miR-277 at the transcriptional level using quantitative PCR (qPCR) in genotypes (1. GMR-Gal4 2. GMR > Aβ42, 3. GMR > Aβ42+ miR-277). Triplicate was used for the calculation. Statistical analysis was performed using student’s t test for independent samples. The data plotted shows mean ± SEM (Standard Error of Mean), and symbols above the error bar signify as ***p value < 0.001, **p value < 0.01, *p value < 0.05, and not significant (ns), p value > 0.05 respectively. The orientation of all imaginal discs is identical with posterior to the left and dorsal up. Scale bar = 100 μm.

Fig. 2. Gain-of-function of miR-277 can restore axonal targeting impaired by Aβ42-mediated neurodegeneration.

Chaoptin (24B10) marks the photoreceptors and their neurons. The photoreceptor neurons bundle up in the optic stalk and innervate the medulla and lamina of the brain. A In the wild-type eye imaginal disc, the retinal axons marked by 24B10 innervates the two layers of the brain marked with yellow arrows. B GMR-Gal4 serves as the control. C Misexpression of Aβ42 (GMR > Aβ42) results in impaired axonal targeting from retina to the brain. The retinal axons fail to innervate the two layers of the brain and end abruptly. Controls (D) GMR>miR-277 (E) GMR>miR-277 mutant, and (F) GMR > Df(3 R)miR-277KO do not impact axonal targeting. G Gain-of-function of miR-277 in the background of GMR > Aβ42 (GMR > Aβ42+ miR-277) significantly restores the axonal targeting to near wild-type. Loss-of-function of miR-277 in GMR > Aβ42 flies using (H) miR-277 mutant (GMR > Aβ42+ miR-277 mutant), and (I) Df(3 R)miR-277KO (GMR > Aβ42 + Df(3 R)miR-277KO) disrupt the axonal targeting. The orientation of all imaginal discs is identical with posterior to the left and dorsal up. The magnification of all eye imaginal discs is ×20. J Bar graph shows the frequency of restoration of axonal targeting phenotype. Sample size was 5 for each genotype 1. GMR-Gal4, 2. GMR > Aβ42, 3. GMR > Aβ42+miR-277, 4. GMR > Aβ42+miR-277 mutant, 5. GMR > Aβ42 + Df(3 R)miR-277KO. GMR > Aβ42+miR-277 significantly restores the axonal targeting (n = 5; p = 0.000029) whereas GMR > Aβ42+ miR-277 mutant (n = 5; p = 0.37) and GMR > Aβ42 + Df(3 R)miR-277KO (n = 5; p = 0.38) disrupt the axonal targeting as compared to GMR > Aβ42. The data plotted shows mean ± SEM (Standard Error of Mean), and symbols above the error bar signify as ***p value < 0.001, **p value < 0.01, *p value < 0.05, and not significant (ns), p value > 0.05 respectively. Scale bar = 100 μm.

Fig. 3. Gain-of-function of miR-277 in the GMR > Aβ42 background can prevent neuronal cell death.

Terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay is employed to mark the cells undergoing cell death. A Wild-type (Canton-S) and B GMR-Gal4 eye imaginal discs show a few TUNEL-positive nuclei. C Misexpression of Aβ42 using GMR-Gal4 driver (GMR > Aβ42) shows increased levels of TUNEL positive nuclei as compared to (A) wild-type eye imaginal disc. D GMR>miR-277, E GMR>miR-277 mutant, and F GMR > Df(3 R)miR-277KO serve as controls. G Gain-of-function of miR-277 in the background of GMR > Aβ42 (GMR > Aβ42+ miR-277) significantly rescue the cell death as compared to GMR > Aβ42. Loss-of-function of miR-277 in GMR > Aβ42 flies using H miR-277 mutant (GMR > Aβ42+ miR-277 mutant), and (I) Df(3 R)miR-277KO (GMR > Aβ42 + Df(3 R)miR-277KO) show elevated levels of TUNEL positive nuclei. A–I TUNEL-positive nuclei were counted within yellow dotted line, the region of interest, for the statistical analysis. A, C, G–J TUNEL-positive nuclei in photoreceptor cells of five eye imaginal discs per genotype were counted (1. Canton-S, 2. GMR-Gal4, 3. GMR > Aβ42, 4. GMR>miR-277, 5. GMR > Aβ42+ miR-277, 6.GMR> miR-277 mutant, 7. GMR > Aβ42+ miR-277 mutant, 8. GMR > Df(3 R)miR-277KO, 9. GMR > Aβ42 + Df(3 R)miR-277KO). Statistical analysis was performed using student’s t test for independent samples. GMR > Aβ42+miR-277 exhibits significant reduction in TUNEL-positive nuclei as compared to GMR > Aβ42 (n = 5; p = 0.000000014) whereas GMR > Aβ42+ miR-277 mutant (n = 5; p = 0.083) and GMR > Aβ42 + Df(3 R)miR-277KO (n = 5; p = 0.056) show slight increase in TUNEL positive nuclei as compared to GMR > Aβ42. The data plotted shows mean ± SEM (Standard Error of Mean), and symbols above the error bar signify as ***p value < 0.001, **p value < 0.01, *p value < 0.05, and not significant (ns), p value > 0.05 respectively. The orientation of all imaginal discs is identical with posterior to the left and dorsal up. Scale bar = 100 μm.

Fig. 4. Gain-of-function of miR-277 downregulates cell death and plays a role in growth.

Pupal retinae were stained with Discs large (Dlg) which marks the membrane (green), and a pan neural marker ELAV (red) which marks photoreceptors. A Pupal retina of wild-type shows hexagonal ommatidia and single layer of interommatidial cells. B GMR > Aβ42 shows disrupted pupal retina as compared to A wild-type pupal retina. C pupal retina of GMR>miR-277 shows excess interommatidial cells as compared to the control (A) wild-type pupal retina. D GMR > Aβ42+ miR-277 restores the number and shape of ommatidial and interommatidial cells. E Bar graph represents the number of photoreceptors within the area of 500px × 500px. We examined five pupal retinae per genotype (n = 5) (1. GMR-Gal4, 2. GMR > Aβ42, 3. GMR>miR-277, 4. GMR > Aβ42+ miR-277). Statistical analysis was performed using student’s t test for independent samples. The number of photoreceptors is significantly restored in GMR > Aβ42+ miR-277 (n = 5; p = 0.0000018) as compared to GMR > Aβ42. F Bar graph represents the number of pigment cells (secondary, tertiary, and bristle cells) within the area of 500px × 500px. We examined five pupal retinae per genotype (n = 5) (1. GMR-Gal4, 2. GMR > Aβ42, 3. GMR>miR-277, 4. GMR > Aβ42+ miR-277). Statistical analysis was performed using student’s t test for independent samples. The number of pigment cells is significantly restored in GMR > Aβ42+ miR-277 (n = 5; p = 0.000064) as compared to GMR > Aβ42. Error bars show standard error of mean (mean ± SEM), and symbols above the error bar signify as ***p value < 0.001, **p value < 0.01, * p value < 0.05, and not significant (ns), p value > 0.05 respectively. Scale bar = 50 μm.

Fig. 5. Modulation of miR-277 in GMR > Aβ42 flies reduces ROS production.

Dihydroethidium (DHE) is employed to detect ROS production in cells. A Wild-type (Canton-S) eye imaginal discs show a few ROS puncta. B GMR > Aβ42 shows elevated ROS puncta as compared to A wild-type eye imaginal disc. C Gain-of-function of miR-277 in the background of GMR > Aβ42 (GMR > Aβ42+ miR-277) significantly reduced the ROS production as compared to GMR > Aβ42. Loss-of-function of miR-277 in GMR > Aβ42 background using D miR-277 mutant (GMR > Aβ42+ miR-277 mutant), and E Df(3 R) miR-277KO [GMR > Aβ42 + Df(3 R)miR-277KO] result in increased ROS production. A–F ROS puncta were counted within yellow dotted line, the region of interest, for the statistical analysis. ROS puncta were quantified in photoreceptor cells of five eye imaginal discs per genotype (1. Canton-S, 2. GMR > Aβ42, 3. GMR > Aβ42+ miR-277, 4. GMR > Aβ42+ miR-277 mutant, 5. GMR > Aβ42 + Df(3 R)miR-277KO). Statistical analysis was performed using student’s t test for independent samples. GMR > Aβ42+miR-277 exhibits significant reduction in ROS puncta as compared to GMR > Aβ42 (n = 5; p = 0.000024), whereas GMR > Aβ42+ miR-277 mutant (n = 5; p = 0.1) shows slight increase and GMR > Aβ42 + Df(3 R)miR-277KO (n = 5; p = 0.02) shows significant increase in ROS puncta as compared to GMR > Aβ42. The data plotted shows mean ± SEM (Standard Error of Mean), and symbols above the error bar signify as ***p value < 0.001, **p value < 0.01, *p value < 0.05, and not significant (ns), p value > 0.05, respectively. The orientation of all imaginal discs is identical with posterior to the left and dorsal up. Scale bar = 100 μm.

Fig. 6. Overexpression of miR-277 reduces the mortality rate and rescues the climbing defects in AD flies.

A The bar graph represents the number of flies eclosed. We compared the number of flies eclosed in 1. Elav Gal4, 2. Elav > Aβ42, 3. Elav > Aβ42+ miR-277 and 4. Elav > Aβ42+ miR-277 mutant 5. Elav > Aβ42 + Df⁽3 R)miR-277KO background and validated that overexpression of miR-277 in the Elav > Aβ42 background rescues the Elav > Aβ42 mortality rate. We counted 270 flies in three independent biological sets from each background and plotted it on a graph in GraphPad Prism. The data plotted shows mean ± SEM (Standard Error of Mean), and symbols above the error bar signify as ***p value < 0.001, **p value < 0.01, *p value < 0.05, and not significant (ns), p value > 0.05, respectively. B Improved climbing activity when miR-277 is overexpressed in AD flies. We compared the number of flies which crossed 10 cm from the bottom of the vial. Graph represents the climbing activity of the flies from day 1–day 30.

Fig. 7. Gain-of-function of miR-277 rescues the hid-mediated cell death.

A GMR>hid, B GMR>rpr, and C GMR>grim show reduced eye phenotype due to caspase-mediated cell death. D GMR>hid+ miR-277 results in significant rescue as compared to GMR>hid flies. E GMR>rpr+ miR-277 and F GMR>grim+ miR-277 do not show a significant change in eye phenotype as compared to their respective controls B GMR>rpr and C GMR>grim. G Bar graph shows frequency as average between three repetitions. Two hundred flies were counted per repetition (200 × 3 = 600) to calculate the frequency for each genotype (1. Canton-S, 2. GMR-Gal4, 3. GMR>hid, 4. GMR>hid+ miR-277, 5. GMR>rpr 6. GMR>rpr+ miR-277, 7. GMR>grim, 8. GMR>grim+ miR-277). Statistical analysis was performed using the student’s t test for independent samples. H Quantitative analyses of severity score of neurodegenerative phenotypes(s) in eye. Flies from each genotype were randomly selected for scoring according to criteria described in the methods section. Comparisons were made using non-parametric: Mann–Whitney t test. I Quantitative analyses of area of the eye. The surface area of the eye (within white dotted line) was calculated using Image J. Statistical analysis was performed using the student’s t-test for independent samples. The surface area of GMR>hid+ miR-277 is significantly rescued (n = 5; p = 0.0000000073) as compared to GMR>hid. Error bars show standard error of mean (mean ± SEM). The orientation of all imaginal discs is identical with posterior to the left and dorsal up. Scale bar = 100 μm.

Fig. 8. hid is the target of miR-277.

A In vitro dual-luciferase reporter assays of miR-mRNA interactions in S2 cell. The mean ± SEM of the relative luciferase expression ratio (firefly luciferase/Renilla luciferase, Luc/R-luc) was calculated for four biological replicates and compared with the control miRNA, miR-1. Statistical analysis was performed using student’s t test for independent samples. The results of dual-luciferase reporter assays of target 3′UTR of drice, hid, and Dark have showed that miR-277 can efficiently target hid and Dark except for drice. B Relative expression of hid at the transcriptional level using quantitative PCR (qPCR) in genotypes (1. GMR-Gal4 2. GMR > Aβ42, 3. GMR > Aβ42+ miR-277, 4. GMR > Aβ42+ miR-277 mutant, 5. GMR > Aβ42 + Df(3 R)miR-277KO). Triplicate was used for the calculation. Statistical analysis was performed using student’s t test for independent samples. hid transcript levels were significantly downregulated in GMR > Aβ42+ miR-277 (n = 3; p = 0.001) whereas slightly increased in GMR > Aβ42+ miR-277 mutant (n = 3; p = 0.14) and GMR > Aβ42 + Df(3 R)miR-277KO) (n = 3; p = 0.1) as compared to GMR > Aβ42. C Bar graph represents average intensity of hid GFP levels within yellow dotted line, region of interest, of five eye imaginal discs per genotype (n = 5) (1. hid5’-GFP, 2. GMR > Aβ42+ hid5’-GFP, 3. GMR > Aβ42+ miR-277+ hid5’-GFP, 4. GMR > Aβ42+ miR-277 mutant+ hid5’-GFP, 5. GMR > Aβ42 + Df(3 R) miR-277KO+ hid5’-GFP). Statistical analysis was performed using student’s t test for independent samples. hid GFP reporter is significantly downregulated in GMR > Aβ42+ miR-277+ hid5’-GFP (n = 5; p = 0.002) whereas slightly upregulated in GMR > Aβ42+ miR-277 mutant+ hid5’-GFP (n = 5; p = 0.75) and GMR > Aβ42 + Df(3 R)miR-277KO+ hid5’-GFP (n = 5; p = 0.29) as compared to GMR > Aβ42+ hid5’-GFP. Error bars show standard error of mean (mean ± SEM), and symbols above the error bar signify as ***p value < 0.001, **p value < 0.01, *p value < 0.05, and not significant (ns), p value > 0.05 respectively. E GMR > Aβ42+ hid5’-GFP results in elevated levels of hid-GFP levels as compared to the wild-type (D) hid5’-GFP. F GMR > Aβ42+ miR-277+ hid5’-GFP results in the significant downregulation of hid-GFP levels, whereas G GMR > Aβ42+ miR-277 mutant+ hid5’-GFP (H) GMR > Aβ42 + Df(3 R)miR-277KO+ hid5’-GFP result in the upregulation of hid-GFP levels. The orientation of all imaginal discs is identical with posterior to the left and dorsal up. Scale bar = 100 μm.

Fig. 9. miR-277 ameliorates the neurodegenerative phenotype of GMR > Aβ42 fly eye.

Misexpression of Aβ42 (GMR > Aβ42 flies) results in the accumulation of amyloid plaques and causes aberrant activation of caspase Hid resulting in neurodegenerative adult eye phenotype. Misexpression of miR-277 in the background of GMR > Aβ42 (GMR > Aβ42+ miR-277), targets 3′UTR of hid mRNA, silences its expression, and hence rescues the reduced eye phenotype observed in GMR > Aβ42 flies.