Research ArticleCELL BIOLOGY

Large-scale RNAi screen identified Dhpr as a regulator of mitochondrial morphology and tissue homeostasis

See allHide authors and affiliations

Science Advances  18 Sep 2019:
Vol. 5, no. 9, eaax0365
DOI: 10.1126/sciadv.aax0365
  • Fig. 1 Screen strategy and the classification of typical phenotypes.

    (A) Diagram to show the screen strategy. (B) Diagram to show how different phenotypes were categorized. L, length; D, diameter. (C to K) The typical mitochondrial morphologies in fat body tissues were shown for each phenotype category. (L to V) Genes that were reported to regulate mitochondrial morphology were identified in this screen. The mitochondrial morphology in the fat body cells with the indicated gene knockdown is shown. Scale bars, 5 μm. Green signals were mitoGFP signals to indicate the mitochondrial morphology. Blue signals were nuclear 4′,6-diamidino-2-phenylindole (DAPI) staining. The boxed region in each image was enlarged in the inset to show the typical mitochondrial morphology. The phenotype category for each genotype was listed on the top of the image.

  • Fig. 2 The reduction of the expression of genes encoding ETC components leads to mitochondrial morphology defects.

    (A) Control (CTRL). (B to J) The decrease of the ETC components led to swollen mitochondria. (K and L) The knockdown of CTPsyn by two independent RNAi lines led to a similar mitochondrial phenotype. Scale bars, 5 μm. Green signals were mitoGFP signals to indicate the mitochondrial morphology. Blue signals were nuclear DAPI staining. The boxed region in each image was enlarged and shown in the inset to present the typical mitochondrial morphology. The phenotype category for each genotype is listed on the top of the image. (M to Q) The TEM images of the fat body tissues in the animals with indicated genotypes. The swollen mitochondria (red arrows) in the RNAi tissues were identified by the remaining cristae (yellow arrows). The green arrows indicate the lipid droplets in the fat body tissues. (R) Quantification of the ratio of abnormal mitochondria (mitochondria with less cristae) in (M) to (Q). n = 4; ***P < 0.001, one-way analysis of variance (ANOVA)/Bonferroni’s multiple comparisons test. (S) Quantification of the average mitochondrial diameter in (M) to (Q). n = 6 images for each genotype; **P < 0.01; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (T) ATP production in the tissues with indicated genotypes. CTPsyn or ND-15 RNAi tissues had reduced ATP production. n = 3 replicates; 20 larvae per replicate; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (U to W) ROS production was increased in the CTPsyn or ND-15 RNAi cells. The green signals of DCFH-DA staining indicated the production of ROS. DAPI staining (blue) indicated the nuclei. Scale bars, 5 μm. (X) Quantification of the relative level of DCFH-DA in (U) to (W). n = 4 images for each genotype; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test.

  • Fig. 3 The loss of Dhpr leads to mitochondrial defects similar to those in Pink1 mutant animals.

    (A) Control (CTRL). (B to E) The RNAi of indicated genes in the fat body tissues led to similar mitochondrial phenotypes. (F) Dhpr overexpression (Dhpr OE) rescued the mitochondrial defects in Dhpr RNAi fat body tissues. (G and H) TEM analysis of the fat body tissues in the animals with indicated genotypes. Dhpr RNAi led to swollen of the mitochondria and the reduction of cristae in fat body tissues. (I) Quantification of the ratio of cells with large mitoGFP puncta in (A) to (F). n = 5; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (J) Quantification of the ratio of mitochondria with less cristae in (G) to (H). n = 3 images of each genotype; ***P < 0.001, two-tailed unpaired Student’s t test. (J′) Quantification of the average mitochondrial diameter in (G) to (H). n = 47; ***P < 0.001, two-tailed unpaired Student’s t test. (K to N) Toluidine blue staining of the fly thorax thick sections with indicated genotypes. The boxed region was enlarged in the inset to show the detailed defects. In the 35-day-old flies, flight muscle fragments were disorganized (red arrows) in Dhpr17-4 and Pink1[B9] flies. (N) Dhpr genomic fragment (Dhpr G) could rescue the muscle defects in Dhpr17-4 mutant. (K′ to N′) TEM analysis of the thin sections of the fly thorax. The mitochondria in Dhpr17-4 mutant muscles are swollen and have less cristae. (O) Quantification of the ratio of degenerated muscle in (K) to (N). n = 4; **P < 0.01; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (P) Quantification of the ratio of mitochondria in (K′) to (N′). The diagram on top of this panel shows the typical normal mitochondria (gray) and abnormal mitochondria. The abnormal mitochondria were categorized into three types based on cristae morphology. Type I (blue): Mitochondria lost some cristae, and there are obvious empty spaces between tightly packed cristae. Type II (orange): The cristae of the mitochondria are disorganized and loose. Type III (red): Mitochondria lost most of its cristae, and more than half of the mitochondria area was empty. n = 3 images for each genotype; ***P < 0.001, χ2 (and Fisher’s exact) test. (Q to S) Anti-TH staining (green) of the adult brains from the 35-day-old animals with indicated genotypes. The TH neurons in different brain regions were labeled in (Q). The number of TH neurons in Dhpr and Pink1 mutant animals is significantly reduced in DM, DL1, and PM regions. (T) Quantification of the number of TH neurons from the animals with indicated genotypes. n = 45; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. n.s., not significant. (U to W) The mitochondria become swollen in the TH neurons from the 5-day-old animals when Dhpr were knocked down in TH neurons. Five-day-old Pink1 mutants also had swollen mitochondria in the TH neurons. Mitochondria were labeled by mitoGFP (green). The TH neurons were labeled by anti-TH staining (red). (X) Quantification to show the ratios of cells with enlarged mitochondria per TH neuron cluster. n = 40; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test.

  • Fig. 4 Pink1 and park overexpression could partially rescue mitochondrial defects in Dhpr mutants.

    (A to H) Mitochondrial morphology in the fat body tissues of the animals with indicated genotypes was accessed through confocal microscopy by examining the mitoGFP signals (green). The nuclei were labeled with DAPI (blue). The boxed regions were enlarged in the insets to show the detailed mitochondrial morphology. Dhpr RNAi led to large mitoGFP puncta formed inside the fat body cells. Pink1/park overexpression could reduce the large mitoGFP puncta formation in Dhpr RNAi tissues. However, Dhpr overexpression could not rescue the mitochondrial defects in Pink1 RNAi tissues. (I) Quantification of the ratio of cells with large mitoGFP puncta in (A) to (H). Ri, RNAi; OE, overexpression; n = 5; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (J to O) Toluidine blue staining of the fly thorax thick sections from 3-day-old flies with indicated genotypes. The boxed regions were enlarged in the insets to show the detailed defects. Muscle-specific knockdown of Dhpr led to disorganized muscle fragments indicated by the holes (red arrows) on the muscle fragments. Pink1 and park overexpression largely rescued the muscle defects. (P) Quantification of the ratio of degenerated muscles in (J) to (O). n = 4; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (J′ to O″) The thin sections of the fly thorax were examined by TEM. Dhpr RNAi led to the enlargement of mitochondria and the reduction of cristae. Pink1 overexpression in wild-type animals reduced the mitochondrial size and led to mild cristae abnormality. The overexpression of Pink1 also reduced the mitochondrial size of the Dhpr RNAi animals and largely rescued the mitochondrial cristae defects caused by Dhpr RNAi. park overexpression could rescue the mitochondrial morphology defects in some muscle fragments (O′) but could not do so in the other muscle fragments (O″). (Q) Quantification of the ratio of mitochondria in (J′) to (O″). The diagram on top of this panel showed the typical normal mitochondria (gray) and abnormal mitochondria. The abnormal mitochondria were categorized into three types based on cristae morphology. Type I (blue): Mitochondria lost some cristae, and there are obvious empty spaces between tightly packed cristae. Type II (orange): The cristae of the mitochondria are disorganized and loose. Type III (red): Mitochondria lost most of its cristae, and more than half of the mitochondria area was empty. n = 3 images of each genotype; ***P < 0.001, χ2 (and Fisher’s exact) test. (Q′) Quantification of the mitochondrial diameters in (J′) to (O″). n = 27; **P < 0.01; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (R and S) Toluidine blue staining of the fly thorax thick sections from 3-day-old flies with indicated genotypes. The boxed regions were enlarged in the insets to show the detailed defects. Pink1[B9] had severe muscle degeneration, and Dhpr knockdown further enhanced the muscle defects of Pink1[B9]. (R′ and S′) Thin sections of the fly thorax were examined by TEM. Three-day-old Pink1[B9] mutants had both intact mitochondria (R′) and severely damaged mitochondria (R″). Almost all of the mitochondria were abnormal when knocking down Dhpr in Pink[B9] mutant background. (T) Quantification of the ratio of degenerated muscles in (J), (K), (R), and (S). n = 7 to 10; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (U) Quantification of the ratio of mitochondria in (J′), (K′), (R′), and (S′). n = 3 images for each genotype; ***P < 0.001, χ2 (and Fisher’s exact) test. The categories of mitochondria are the same as those in (Q). The scale bars in the immunofluorescence images are 5 μm. The scale bars in the TEM images are as indicated.

  • Fig. 5 The catalytic activity of Dhpr is required for its mitochondrial function.

    (A to D and I to L) Confocal images showing mitochondrial morphology in the fat body tissues of the third-instar larvae with indicated genotypes. The green signals from mitoGFP labeled the mitochondria, and the blue signals were nuclei labeled with DAPI. Scale bars, 5 μm. Wild-type Dhpr overexpression rescued the mitochondrial defects in Dhpr RNAi fat bodies. DhprG16D overexpression did not rescue the mitochondrial defects but worsened it in Dhpr RNAi tissues. Nos RNAi did not cause obvious mitochondrial defects in wild-type fat body tissues but greatly enhanced the mitochondrial defects in Dhpr RNAi fat body tissues. (E to H and M to P) Toluidine blue staining of the fly thorax thick sections from 3-day-old flies with indicated genotypes. The red arrows indicated the disorganized muscle fragments. (E′ to H′ and M′ to P′) TEM images of thorax thin sections showing the detailed morphology of mitochondria and muscle fibers. The scale bars are as indicated. (Q) Quantification of the ratio of cells with large mitoGFP puncta in (A) to (D) and (I) to (L). n = 5; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (Q′) Quantification of the average number of large mitoGFP puncta per affected cell in (A) to (D) and (I) to (L). n = 10; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (R) Quantification of the ratio of degenerated muscle in (E) to (H) and (M) to (P). n = 4 to 10; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (S) Quantification of the ratio of mitochondria in (E) to (H) and (M) to (P). The diagram on top of this panel shows the typical normal mitochondria (gray) and abnormal mitochondria. The abnormal mitochondria were categorized into three types based on cristae morphology. Type I (blue): Mitochondria lost some cristae, and there are obvious empty spaces between tightly packed cristae. Type II (orange): The cristae of the mitochondria are disorganized and loose. Type III (red): Mitochondria lost most of its cristae, and more than half of the mitochondria area was empty. n = 3 images for each genotype; ***P < 0.001, χ2 (and Fisher’s exact) test.

  • Fig. 6 Dhpr regulates mitochondrial morphology by modulating S-nitrosylation of Drp1.

    (A to H and E′ to H′) Drp1 overexpression could rescue Dhpr RNAi-induced mitochondrial defects in fat body tissues (A to D) and muscles (E to H and E′ to H′). (I) Quantification of the ratio of cells with large mitoGFP puncta in (A) to (D), (P), and (S). n = 5; **P < 0.01; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (J) Quantification of the ratio of degenerated muscle in (E) to (H). n = 4 to 10; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (K) Quantification of the ratio of mitochondria in (E′) to (H′). The diagram on top of this panel shows the typical normal mitochondria (gray) and abnormal mitochondria. The abnormal mitochondria were categorized into three types based on cristae morphology. Type I (blue): Mitochondria lost some cristae, and there are obvious empty spaces between tightly packed cristae. Type II (orange): The cristae of the mitochondria are disorganized and loose. Type III (red): Mitochondria lost most of its cristae, and more than half of the mitochondria area was empty. n = 3 images for each genotype; ***P < 0.001, χ2 (and Fisher’s exact) test. (K′) Quantification of the average mitochondrial diameter in (E′) to (H′). n = 27; **P < 0.01; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test. (L and L′) Dhpr RNAi reduced S-nitrosylation of Drp1. (L′) Quantification of (L). n = 3 replicates; 15 individual flies per replicate; ***P < 0.001, two-tailed unpaired Student’s t test. (M and M′) C643 site was a major S-nitrosylation site of Drp1 in vivo. C643A mutation in Drp1 greatly reduced S-nitrosylation of Drp1. (M′) Quantification of (M). n = 3 replicates; 15 individual flies per replicate; ***P < 0.001, two-tailed unpaired Student’s t test. (N to Q) Overexpression of Drp1, but not Drp1C643A, could rescue mitochondrial defects caused by Drp1 RNAi. (S) Drp1C643A overexpression cannot rescue Dhpr RNAi-induced mitochondrial defects in fat body tissues. (A to D, N to Q, and S) Images of fat body cells with indicated genotypes. Mitochondria were marked by mitoGFP signals (green), and nuclei were marked by DAPI staining (blue). Drp1C643A-3HA expression was indicated by anti-HA staining (red). Scale bars, 5 μm. (E to H) Toluidine blue staining of the 3-day-old fly thorax thick sections with indicated genotypes. The red arrows indicated the disorganized muscle fragments. (E′ to H′) TEM images of thorax thin sections showing the detailed morphology of mitochondria and muscle fibers. The scale bars are as indicated. (R) Quantification of the ratio of cells with large mitoGFP puncta in (A) and (N) to (S). n = 7; ***P < 0.001, one-way ANOVA/Bonferroni’s multiple comparisons test.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/9/eaax0365/DC1

    Supplementary Materials and Methods

    Fig. S1. The complex analysis of the hits from the mitochondrial morphology screening.

    Fig. S2. The mitochondrial morphology in the fat body tissues with indicated gene RNAi is shown.

    Fig. S3. Sixteen genes encoding proteasome components were identified in this mitochondrial morphology screening.

    Fig. S4. Twenty-seven genes encoding spliceosome components were identified in this mitochondrial morphology screening.

    Fig. S5. The reduction of the enzymes involved in tyrosine and lysine metabolism led to abnormal mitochondrial morphology.

    Fig. S6. Loss of Dhpr leads to the reduction of life span and increase of the ROS production.

    Fig. S7. The overexpression of Pink1 or park partially rescues muscle defects caused by Dhpr RNAi.

    Fig. S8. The genetic interaction between Dhpr and genes whose products consume or produce BH4.

    Fig. S9. The genetic interaction between Dhpr and other core machinery of mitochondrial fusion and fission.

    Fig. S10. The model of how Dhpr regulates mitochondrial morphology.

    Table S1. The list of the RNAi lines used in this screening and their corresponding genes.

    Table S2. The annotation of the phenotypes and the quantification data.

    Table S3. The list of genes that had two or three independent RNAi lines and genes that had been reported to be involved in regulating mitochondria.

    Table S4. The list of protein complexes required for mitochondrial morphology maintenance.

    Table S5. The lists of genes encoding spliceosome, proteasome, and electron transfer chain components that have been identified in this screen.

  • Supplementary Materials

    The PDF file includes:

    • Supplementary Materials and Methods
    • Fig. S1. The complex analysis of the hits from the mitochondrial morphology screening.
    • Fig. S2. The mitochondrial morphology in the fat body tissues with indicated gene RNAi is shown.
    • Fig. S3. Sixteen genes encoding proteasome components were identified in this mitochondrial morphology screening.
    • Fig. S4. Twenty-seven genes encoding spliceosome components were identified in this mitochondrial morphology screening.
    • Fig. S5. The reduction of the enzymes involved in tyrosine and lysine metabolism led to abnormal mitochondrial morphology.
    • Fig. S6. Loss of Dhpr leads to the reduction of life span and increase of the ROS production.
    • Fig. S7. The overexpression of Pink1 or park partially rescues muscle defects caused by Dhpr RNAi.
    • Fig. S8. The genetic interaction between Dhpr and genes whose products consume or produce BH4.
    • Fig. S9. The genetic interaction between Dhpr and other core machinery of mitochondrial fusion and fission.
    • Fig. S10. The model of how Dhpr regulates mitochondrial morphology.

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Table S1 (Microsoft Excel format). The list of the RNAi lines used in this screening and their corresponding genes.
    • Table S2 (Microsoft Excel format). The annotation of the phenotypes and the quantification data.
    • Table S3 (Microsoft Excel format). The list of genes that had two or three independent RNAi lines and genes that had been reported to be involved in regulating mitochondria.
    • Table S4 (Microsoft Excel format). The list of protein complexes required for mitochondrial morphology maintenance.
    • Table S5 (Microsoft Excel format). The lists of genes encoding spliceosome, proteasome, and electron transfer chain components that have been identified in this screen.

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article