A protein complex containing Mdm10p, Mdm12p, and Mmm1p links mitochondrial membranes and DNA to the cytoskeleton-based segregation machinery

Abstract

Previous studies indicate that two proteins, Mmm1p and Mdm10p, are required to link mitochondria to the actin cytoskeleton of yeast and for actin-based control of mitochondrial movement, inheritance and morphology. Both proteins are integral mitochondrial outer membrane proteins. Mmm1p localizes to punctate structures in close proximity to mitochondrial DNA (mtDNA) nucleoids. We found that Mmm1p and Mdm10p exist in a complex with Mdm12p, another integral mitochondrial outer membrane protein required for mitochondrial morphology and inheritance. This interpretation is based on observations that 1) Mdm10p and Mdm12p showed the same localization as Mmm1p; 2) Mdm12p, like Mdm10p and Mmm1p, was required for mitochondrial motility; and 3) all three proteins coimmunoprecipitated with each other. Moreover, Mdm10p localized to mitochondria in the absence of the other subunits. In contrast, deletion of MMM1 resulted in mislocalization of Mdm12p, and deletion of MDM12 caused mislocalization of Mmm1p. Finally, we observed a reciprocal relationship between the Mdm10p/Mdm12p/Mmm1p complex and mtDNA. Deletion of any one of the subunits resulted in loss of mtDNA or defects in mtDNA nucleoid maintenance. Conversely, deletion of mtDNA affected mitochondrial motility: mitochondria in cells without mtDNA move 2-3 times faster than mitochondria in cells with mtDNA. These observations support a model in which the Mdm10p/Mdm12p/Mmm1p complex links the minimum heritable unit of mitochondria (mtDNA and mitochondrial outer and inner membranes) to the cytoskeletal system that drives transfer of that unit from mother to daughter cells.

Figure 1.

Mdm12p-Myc– and Mdm10p-Myc–containing structures are adjacent to mtDNA nucleoids. (A) Growth on glucose-(YPD) and glycerol (YPG)-containing media of wild-type parent (YPH252), Mdm12p-Myc–(IBY118), and Mdm10p-Myc (IBY113)–expressing cells. Equal amounts of cells were spotted onto plates and grown for 3 d. (B) Mdm12p-Myc– and Mdm10p-Myc–expressing cells were grown at 30°C to early mid-log phase. Fixed cells were stained for Mdm12p-Myc (a) or Mdm10p-Myc (b) by indirect immunofluorescence with anti-Myc antibody (green) and for mitochondria by an antibody raised against mitochondrial OM proteins (red). Z-sections through the cell were collected and projected into a single image. (C) Mdm12p-Myc– (a, c, and e) and Mdm10p-Myc (b, d, and f)–expressing cells were grown and fixed as for B. Tagged proteins were visualized using anti-Myc antibody (green). mtDNA was visualized using DAPI (red). Z-sections through the cell were collected, deconvolved, and projected with the Volocity 3-D rendering tool. DAPI signal is projected at maximum opacity, whereas Myc fluorescence is projected at 25–30% of maximum opacity, resulting in more solid fluorescent signal and removal of background. Three views of the 3-D models rotated 0° (a and b), 70° (c and d), and 150° (e and f) are shown. Nuclear DNA was resolved as large objects (arrows) and mtDNA resolved as small dots. Bar, 1 μm.

Figure 2.

Defects in mtDNA nucleoid stability in mdm12Δ mutants. MDM12 (MYY291) (a and b) and mdm12Δ cells (MYY624) (c and d) were grown at 24°C to mid-log phase in synthetic, glucose-based liquid media (SC). Fixed cells were stained for mitochondrial OM proteins by indirect immunofluorescence by using anti-mitochondrial OM antibody (a and c). mtDNA was visualized using DAPI (b and d). Z-sections throughout the cells were collected, deconvolved, and projected. Arrowheads mark nuclear DNA. Arrows point to mtDNA. Bar, 2 μm.

Figure 3.

Mitochondria in mdm12Δ cells do not exhibit long-distance movement. Wild-type (MYY291) (A and C) and mdm12Δ mutant cells (MYY624) (B and D) expressing a fusion protein consisting of the mitochondrial signal sequence of CS1-GFP were grown at 23°C in YP-raffinose media. Cells were viewed by differential interference contrast (A and B), and GFP-labeled mitochondria were visualized by fluorescence microscopy (C and D). Mitochondrial movements in CS1-GFP cells were monitored by time-lapse imaging. Images were acquired at 20-s intervals over a period of 10 min. The points denote the position of the tips of organelles at 20-s intervals in wild-type cells (E) and in mdm12Δ mutant cells (F). Bar, 1 μm.

Figure 4.

Mdm10p-Myc and Mdm12p-Myc localize to mtDNA in the absence of actin structures. Yeast cells expressing Mdm10p-Myc– (IBY113) (A–D) or Mdm12-Myc (IBY118) (E–H)–expressing cells were grown at 30°C to mid-log phase. Then 0.4 mM Lat-A dissolved in dimethyl sulfoxide (C, D, G, and H) or an equal volume of dimethyl sulfoxide alone (A, B, E, and F) was added to each culture. After 10-min incubation under growth conditions, cells were fixed and stained for the Myc tag by using anti-Myc antibody (green) and for mtDNA by using DAPI (red) (A, C, E, and G). F-actin–containing structures were visualized with rhodamine phalloidin (B, D, F, and H). Bar, 1 μm.

Figure 5.

Mmm1p, Mdm10p, and Mdm12p coimmunoprecipitate with each other. Mitochondria were purified from an untagged strain (D273-10b) (lanes 1 and 4), strains expressing only HA-tagged construct (DNY108 in A and B; DNY366 in C) (lanes 3 and 6) or strains coexpressing either Mmm1p-Myc with Mdm10p-HA (DNY365 in A), Mdm12p-Myc with Mdm10p-HA (DNY364 in B), or Mdm12p-Myc with Mmm1p-HA (DNY422 in C) (lanes 2 and 5). Mitochondria were solubilized in a 0.5% digitonin buffer (lanes 1–3) and from this lysate, Myc-tagged proteins were immunoprecipitated (see MATERIALS AND METHODS) using a monoclonal anti-Myc antibody (lanes 4–6). Immunoprecipitated proteins were probed using anti-Myc, anti-HA, and control, anti-OM45 antibodies. OM45 is an abundant integral OM protein that does not immunoprecipitate with Mmm1p, Mdm10p, or Mdm12p.

Figure 6.

Localization of proteins in deletion mutants. mmm1Δ mutant cells expressing Mdm12p-Myc (HCY340; a–d) or Mdm10p-Myc (HCY330; e–h), and mdm12Δ mutant cells expressing Mmm1p-Myc (HCY351; i–l) or Mdm10p-Myc (HCY361; m–p) were grown at 24°C to early mid-log phase in synthetic, glucose-based liquid media (SC-his, -ura). Fixed cells were stained for the Myc epitope and mitochondrial OM proteins by indirect immunofluorescence by using anti-Myc antibody (a, e, i, and m) and anti-OM antibody (b, f, j, and n), respectively. mtDNA was visualized using DAPI (c, g, k, and o). d, h, l, and p show the overlays of Myc epitope (green), mitochondrial OM (red), and DNA (blue) staining. Bar, 1 μm.

Figure 7.

Mdm10p-Myc and Mdm12p-Myc show punctate localization to mitochondria in the absence of mtDNA. Mdm12p-Myc–(HCY372) (a–c) and Mdm10p-Myc (HCY371) (d–f)–expressing cells were grown at 30°C to early mid-log phase. Fixed cells were stained for the Myc epitope and mitochondrial OM proteins by indirect immunofluorescence using anti-Myc antibody (b and e) and anti-OM antibody (c and f), respectively. mtDNA was visualized using DAPI (a and b). Bar, 2 μm.

Figure 8.

Mitochondrial velocities in rho⁰ cells are higher than wild-type cells. (A) Mitochondrial morphology in rho⁰ and respiratory deficient mutants is normal: wild-type (KAY40; a–c), rho⁰ cells derived from KAY40 (d–f), QCR9 parent (BY4743; g–i), and qcr9Δ mutant (RG24813; j–l) cells expressing CS1-GFP were grown to mid-log phase at 23°C in synthetic, glucose-based liquid media (SC-ura). After fixation, nuclear structures (a, d, g, and j) and actin cytoskeleton (c, f, i, and l) were visualized by staining with DAPI and rhodamine-phalloidin, respectively. CS1-GFP was used to detect mitochondria (b, e, h, and k). Arrows point to examples of colocalization of mitochondria with actin cables in rho⁰ cells. Bar, 5 μm. (B) To measure velocities of mitochondrial movement, cells grown as described above were harvested and immediately visualized by fluorescence microscopy. Time-lapse images of mitochondria in living cells were acquired at 20-s intervals over a period of 10 min. All velocities are averages of 50 measurements. Error bars represent SD. Mitochondria in KAY40 rho⁰ cells moved significantly faster than in KAY40 rho⁺ cells (p <0.001, t test).