Figure 1.
Western blot analysis showing specificity of antisera that recognize Drosophila dMfn, Opa1, and Drp1.
(A) Protein extracts from wt flies and flies expressing either the UAS-dmfn-RNAiVienna or UAS-dmfn-RNAiGuo transgenes targeting the dmfn gene were subjected to western blot analysis with an anti-dMfn antiserum. The hsp70-GAL4 driver was used to induce expression of the dmfn-RNAi transgenes, and flies were collected for analysis 24 to 48 hrs after a 3-hr heat shock protocol. The arrow indicates the location of the dMfn band at 94 kDa. The asterisks (*) indicate two nonspecific bands detected at 80 kDa and 110 kDa. (B) Protein extracts from wt flies, flies bearing the UAS-Opa1-FLAG transgene and the mesoderm-specific 24B-GAL4 driver, and flies bearing the UAS-opa1RNAi transgene and the muscle-specific dmef2-GAL4 driver were subjected to western blot analysis with an anti-Opa1 antiserum. (C) Protein extracts from wt flies, flies bearing the UAS-Drp1-HA transgene and the muscle-specific dmef2-GAL4 driver, and flies heterozygous for the Df(2L)Excel6008 deletion chromosome (which removes the drp1 gene) were subjected to western blot analysis with an affinity-purified anti-Drp1 antiserum.
Figure 2.
Perturbations of PINK1 and parkin specifically influence dMfn abundance.
Protein extracts from wt males (m), PINK1B9 hemizygous males, wt males and females (m and f) and park25 males and females were subjected to western blot analysis with an affinity-purified anti-dMfn antiserum (A), an anti-Opa1 antiserum (C), an affinity-purified anti-Drp1 antiserum (D), an anti-complex V β antiserum (A, C, D), an anti-VDAC antiserum (A, C, D), and an anti-actin antiserum (A, C, D). Arrow in (A) indicates location of the dMfn band at 94 kDa and the asterisks (*) indicate nonspecific bands. (B) Quantification of dMfn abundance in PINKB9 and park25 null mutants relative to wt controls. Because PINK1B9 males are sterile and the PINK1 gene resides on the X chromosome, we are only able to generate male PINK1B9 mutants and thus used wt males as the control population for these mutants. The ratio of dMfn to actin abundance was obtained from three independent blots for each sample analyzed; the ratios for wt controls were set at a value of 1 and mutant ratios were normalized to the wt ratios. *p<0.05 by Student’s t-test. (E) Protein extracts from mitochondrial and cytosolic fractions (see Materials and Methods section for details on fractionation) were subjected to western blot analysis with an affinity-purified anti-Drp1 antiserum, an anti-complex V β (CompV) antiserum, an anti-VDAC antiserum, and an anti-actin antiserum. (F) Protein extracts from wt flies, flies overexpressing Parkin (hsp70-GAL4/UAS-Parkin) and flies overexpressing PINK1 (hsp70-GAL4/UAS-PINK1) were subjected to western blot analysis with an affinity-purified anti-dMfn antiserum, an anti-complex V β antiserum, an anti-VDAC antiserum, and an anti-actin antiserum. Flies were subjected to a 1-hr heat shock and collected for analysis 24 hrs following the heat shock. The arrow indicates the location of the dMfn band at 94 kDa and the asterisks (*) indicate the two nonspecific bands.
Figure 3.
dMfn is ubiquitinated in a PINK1- and Parkin-dependent fashion.
Affinity purified anti-dMfn antiserum was used to immunoprecipitate dMfn from wild-type flies, PINK1B9 mutants, and park25 mutants. As a control for specificity, anti-dMfn antiserum was also used to immunoprecipitate dMfn from flies with hsp70-GAL4 driven expression of UAS-dmfn-RNAiVienna. In the left panel, 3% of the lysate input used in the immunoprecipitations was subjected to western blot analysis using an anti-ubiquitin antiserum to show that general ubiquitination levels were similar in all genotypes. In the middle and right panels, dMfn immunoprecipitates from all four genotypes were subjected to western blot analysis with either anti-ubiquitin antiserum (middle panel) or anti-dMfn antiserum (right panel). Arrow indicates the unmodified dMfn species detected in wt flies, PINK1B9 mutants, and park25 mutants, with reduced levels in flies expressing UAS-dmfn-RNAiVienna. Arrowhead indicates location of ubiquitinated dMfn species. These analyses were replicated at least three times with similar results.
Figure 4.
Affinity purified anti-Parkin antiserum was used to immunoprecipitate Parkin from wt flies and park25 null mutants. As park25 flies do not produce any Parkin protein, this immunoprecipitation serves as a control for any nonspecific binding of other proteins to the anti-Parkin antiserum. Three percent of the total amount of lysate used for the immunoprecipitations as well as the Parkin immunoprecipitates were subjected to western blot analysis with the anti-dMfn antiserum. Only in the wt flies did a band corresponding to the dMfn band appear indicating an interaction between Parkin and dMfn. The asterisks (*) indicate the nonspecific bands.
Figure 5.
A potential model by which the PINK1/Parkin pathway promotes mitochondrial fragmentation and turnover.
Previous work has shown that PINK1 localizes to the outer mitochondrial membrane with its kinase domain facing the cytoplasm [28], [39], where it is required to recruit Parkin selectively to damaged mitochondria to promote the autophagic clearance of these mitochondria [27], [28], [29], [30], [31]. Our model postulates that upon Parkin recruitment to damaged portions of the mitochondrial reticulum, it ubiquitinates (Ub) the mitochondrial fusion-promoting factor Mitofusin, thus tagging it for degradation, or otherwise inactivating its fusion-promoting activity. Subsequent Drp1-dependent mitochondrial fission would yield a damaged mitochondrial product that is deficient in Mitofusin, and thus unable to re-enter the mitochondrial network. These small damaged mitochondrial fission products would instead be targeted for autophagic degradation. Although our figure depicts a model in which Parkin binds prior to mitochondrial fission, Parkin could also be recruited to damaged mitochondria following fission. An alternative model is that the ubiquitination of Mitofusin serves as a signal for the turnover of damaged mitochondria. These possible models are not mutually exclusive.