JBC Origene Your Gene Company

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M308436200 on September 12, 2003

J. Biol. Chem., Vol. 278, Issue 49, 48997-49005, December 5, 2003
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
278/49/48997    most recent
M308436200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kondo-Okamoto, N.
Right arrow Articles by Okamoto, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kondo-Okamoto, N.
Right arrow Articles by Okamoto, K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Mmm1p Spans Both the Outer and Inner Mitochondrial Membranes and Contains Distinct Domains for Targeting and Foci Formation*

Noriko Kondo-Okamoto, Janet M. Shaw{ddagger}, and Koji Okamoto§

From the Department of Biology, University of Utah, Salt Lake City, Utah 84112

Received for publication, August 1, 2003 , and in revised form, September 10, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
In the yeast Saccharomyces cerevisiae, the integral membrane protein Mmm1p is required for maintenance of mitochondrial morphology and retention of mitochondrial DNA (mtDNA). Mmm1p localizes to discrete foci on mitochondria that are adjacent to mtDNA nucleoids in the matrix, raising the possibility that this protein plays a direct role in organizing, replicating, or segregating mtDNA. Although Mmm1p has been shown to cross the outer membrane with its C terminus facing the cytoplasm, the location of the N terminus has not been resolved. Here we show that Mmm1p spans both the outer and inner mitochondrial membranes, exposing its N terminus to the matrix. Surprisingly, deletion of the N-terminal extension decreased steady-state levels of the Mmm1 protein but did not affect mitochondrial morphology or mtDNA maintenance. Moreover, expression of Neurospora crassa MMM1, which naturally lacks a long N-terminal extension, substituted for loss of Mmm1p in budding yeast. These results indicate that the matrix-exposed portion of Mmm1p is not essential for mtDNA nucleoid maintenance. Additional studies revealed that the transmembrane segment and C-terminal domain of Mmm1p are required for foci formation and mitochondrial targeting, respectively. Our data suggest that the double membrane-spanning topology of Mmm1p at the membrane contact site is critical for formation of tubular mitochondria.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Mitochondria are dynamic organelles that change size, shape, and distribution in response to changing cellular needs (1–4). Because mitochondria are essential and cannot be synthesized de novo, most cell types employ cytoskeletal elements and molecular motors to partition the organelles into newly formed daughter cells during division (5). Studies performed in a variety of organisms including budding yeast reveal that maintenance of proper organelle morphology is critical for mitochondrial function. In Saccharomyces cerevisiae, mitochondria form highly branched tubular networks that continually divide and fuse (6–10). Mutations in a variety of nuclear genes disrupt wild-type mitochondrial morphology, converting the tubular network into small fragments, aggregates, large spheres, collapsed tubules, nets, or other aberrant shapes (11). In some cases, these morphological changes lead to loss of mitochondrial DNA (mtDNA)1 and defective respiratory activity (12–14). For example, mitochondria fragment and lose mtDNA in cells lacking the large GTPase Fzo1p (15, 16) or Ugo1p, (17) two integral outer membrane (OM) proteins that mediate mitochondrial fusion. Similar mtDNA loss is observed in cells lacking the intermembrane space (IMS) GTPase Mgm1p (18–21) required for mitochondrial fusion and inner membrane (IM) remodeling and the IM Rhomboid-type protease Mdm37p/Pcp1p/Rbd1p (11, 22–24) required for the biogenesis of Mgm1p. In the examples cited above, mtDNA loss is thought to be a secondary consequence of changes in mitochondrial shape, because there is currently no evidence that these proteins interact directly or indirectly with mtDNA nucleoids (mtDNA-protein complexes). A second class of mutations that converts tubular mitochondria into large spheres also leads to mtDNA loss. These mutations affect three integral OM proteins, called Mdm10p (25), Mdm12p (26), and Mmm1p (27), which are thought to form a complex. Roles for this complex in cytoskeletal attachment and mtDNA nucleoid maintenance have been proposed.

The function of the Mmm1 protein has been studied in some detail. S. cerevisiae Mmm1p is predicted to have a single transmembrane (TM) segment of ~25 residues (amino acids 92–116) (27). Cells lacking functional Mmm1p display defects in mitochondrial motility and inheritance, and mitochondria isolated from these cells are defective in ATP-sensitive actin binding activity in vitro (28, 29). Based on these observations, it was proposed that Mmm1p is required for interactions of mitochondria with the actin cytoskeleton. When tagged at the C terminus with the green fluorescent protein (GFP), Mmm1p-GFP localizes as discrete foci on the surface of mitochondria (30). Strikingly, these Mmm1p-GFP foci appear adjacent to mtDNA nucleoids in the matrix (30). Additional biochemical data suggest that Mmm1p resides at contact sites between the inner and outer mitochondrial membranes (30). These findings raise the possibility that Mmm1p is connected either directly or indirectly to mtDNA nucleoids.

Studies of the filamentous fungus Neurospora crassa MMM1 (NcMMM1) suggest that the cellular role of this protein varies in different cell types. In N. crassa, mitochondria form tubular arrays that co-align with microtubules rather than actin filaments (31). Unlike S. cerevisiae, N. crassa cannot survive in the absence of mitochondrial respiration, and loss of mtDNA results in lethality. Although NcMMM1 is also required for maintenance of normal mitochondrial morphology, it is dispensable for interactions with microtubules (32, 33). Moreover, the Mmm1 protein is not essential for mtDNA stability in N. crassa, because loss-of-function mutations are not lethal (32). Thus, although the role of Mmm1p in mitochondrial morphology maintenance has been conserved during evolution, its roles in cytoskeletal association and/or mtDNA retention may have been altered or lost.

The phenotypes observed in mmm1 mutants suggest that Mmm1p contains distinct domains with different functions. A better understanding of Mmm1p structure, topology, and domain function may reveal how this protein carries out activities within the cell. S. cerevisiae Mmm1p contains one predicted TM segment that spans the outer mitochondrial membrane leaving the C-terminal domain in the cytoplasm (27). Although the N terminus faces the interior of the organelle, its exact location (IMS or matrix) has not been determined. In addition, the specific functions of the Mmm1p N-terminal extension, TM segment, and C-terminal domain have not been explored. In this study, we determined the submitochondrial location of the Mmm1p N terminus and identified domains important for mitochondrial targeting and foci formation. In vivo site-specific cleavage and in vitro protease protection assays revealed that the N terminus of Mmm1p is exposed to the matrix, a topology that requires the protein to span both the outer and inner mitochondrial membranes. Although deletion of the N-terminal extension resulted in reduction of steady-state Mmm1p protein levels, it did not affect protein targeting to the organelle, maintenance of mitochondrial morphology, or mtDNA retention. Moreover, the mtDNA loss phenotype of the S. cerevisiae mmm1 null (mmm1{Delta}) mutant could be rescued by expressing NcMMM1, which naturally lacks a long N-terminal extension. These findings indicate that the matrix-exposed, N-terminal extension of Mmm1p does not play a direct role in mtDNA maintenance. We further show that the TM segment and C-terminal domain play essential roles in Mmm1p foci formation and mitochondrial targeting, respectively. These combined results suggest that the double membrane-spanning topology of Mmm1p at the membrane contact site is crucial for establishment and maintenance of tubular mitochondrial morphology. Implications of these findings with respect to mtDNA maintenance and cytoskeletal attachment are discussed. Models for the double membrane-spanning topology of Mmm1p are also presented.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Yeast Strains and Growth Conditions—All strains are derivatives of the FY10 strain (34). Standard genetic methods were used to grow, transform, and manipulate yeast (35) and bacterial (36) strains. Strains used in this study include: JSY1825 (MAT{alpha}, his3{Delta}200, leu2{Delta}1, lys2{Delta}-202, trp1{Delta}63, ura3–52) and NOY84 (MAT{alpha}, his3{Delta}200, leu2{Delta}1, lys2{Delta}-202, trp1{Delta}63, ura3–52, mmm1{Delta}::HIS3, [pRS416-MMM1-His]). Plasmids containing MMM1 constructs were introduced by plasmid shuffle into NOY84. Yeast transformants were grown at 30 °C in SC media or on dropout plates containing one of the following carbon sources: 2% dextrose (SD), 2% glycerol (SGly), 2% raffinose (SR), or 2% galactose plus 2% raffinose (SGalR).

Plasmid Construction—For pRS416-MMM1-His and pRS315-MMM1-His, a BamHI-EcoRV fragment encoding the 0.52-kb MMM1 5'-UTR and Mmm1p (amino acids 1–426) plus His8, and an EcoRV-XhoI fragment encoding the 0.45-kb MMM1 3'-UTR were generated by PCR and cloned into the BamHI-XhoI sites of the low copy yeast vectors pRS416 and pRS315, respectively (American Type Culture Collection). For pRS315-mmm1{Delta} (2–75)-His, pRS315-mmm1{Delta} (2–90)-His, and pRS315-mmm1{Delta} (2–120)-His, the 1.1-kb BamHI-SalI fragment of pRS315-MMM1-His was replaced by a PCR-generated BamHI-SpeI fragment encoding the 0.52-kb MMM1 5'-UTR plus ATG and a PCR-generated SpeI-SalI fragment encoding Mmm1p (amino acids 76–190), (amino acids 91–190), and (amino acids 121–190), respectively. For pRS315-MMM1-GFP, a BamHI-XmaI fragment encoding the 0.52-kb MMM1 5'-UTR and Mmm1p (amino acids 1–426), XmaI-EcoRV fragment encoding HA plus GFP, and an EcoRV-XhoI fragment encoding the 0.45-kb MMM1 3'-UTR were generated by PCR and cloned into the BamHI-XhoI sites of pRS315. For pRS315-mmm1{Delta} (2–75)-GFP and pRS315-mmm1{Delta} (2–90)-GFP, the 1.1-kb BamHI-SalI fragment of pRS315-MMM1-GFP was replaced by a 0.87-kb BamHI-SalI fragment from pRS315-mmm1{Delta} (2–75)-His and a 0.82-kb BamHI-SalI fragment from pRS315-mmm1{Delta} (2–90)-His, respectively. For pRS315-T20(TM)-mmm1(C)-His, a BamHI-SpeI fragment encoding the 0.52-kb MMM1 5'-UTR plus ATG, a PCR-generated NheI-SpeI fragment encoding Tom20p (amino acids 2–50), and a SpeI-XhoI fragment encoding Mmm1p (amino acids 121–426) plus His8 and the 0.45-kb MMM1 3'-UTR were cloned into the BamHI-XhoI sites of the pRS315 variant that lacks an SpeI site in the MCS. For pRS315-T20(TM)-mmm1(C)-GFP, the 1.1-kb BamHI-SalI fragment of pRS315-MMM1-GFP was replaced by a 0.88-kb BamHI-SalI fragment from pRS315-T20(TM)-mmm1(C)-His. For pRS315-T20(TM)-GFP, the 1.4-kb SpeI-XhoI fragment of pRS315-T20(TM)-mmm1(C)-His was replaced by a PCR-generated SpeI-NcoI fragment encoding GFP (amino acids 2–56) and a 1.0-kb NcoI-XhoI fragment from pRS315-MMM1-GFP. For pRS315-NcM(TM)-mmm1(C)-His, the 0.67-kb BamHI-SpeI fragment of pRS315-T20(TM)-mmm1(C)-His was replaced by a PCR-generated BamHI-NheI fragment encoding the 0.52-kb MMM1 5'-UTR plus ATG and N. crassa MMM1 (amino acids 2–39). For pRS315-NcM(TM)-mmm1(C)-GFP, the 1.1-kb BamHI-SalI fragment of pRS315-MMM1-GFP was replaced by an 0.85-kb BamHI-SalI fragment from pRS315-NcM(TM)-mmm1(C)-His. For pRS425-MMM1-GFP, a 3.0-kb BamHI-XhoI fragment from pRS315-MMM1-GFP was cloned into the BamHI-XhoI sites of pRS425 (American Type Culture Collection). For pRS425-mmm1 (1–135)-GFP, pRS425-mmm1 (1–180)-GFP, pRS425-mmm1 (1–275)-GFP, and pRS425-mmm1 (1–365)-GFP, the 1.8-kb BamHI-XmaI fragment of pRS425-MMM1-GFP was replaced by PCR-generated BamHI-XmaI fragments encoding the 0.52-kb MMM1 5'-UTR plus Mmm1p (amino acids 1–135), (amino acids 1–180), (amino acids 1–275), and (amino acids 1–365), respectively. For pRS425-mmm1 (1–390)-GFP and pRS425-mmm1 (1–415)-GFP, the 0.71-kb SalI-XmaI fragment of pRS425-MMM1-GFP was replaced by PCR-generated SalI-XmaI fragments encoding Mmm1p (amino acids 191–390) and (amino acids 191–415), respectively. For pRS425-mmm1{Delta} (2–120)-GFP, the 1.1-kb BamHI-SalI fragment of pRS425-MMM1-GFP was replaced by a 0.74-kb BamHI-SalI fragment from pRS315-mmm1{Delta} (2–120)-His. For pRS315-TCS-MMM1-His and pRS315-TCS-MMM1-GFP, the 0.58-kb SpeI-SalI fragment of pRS315-MMM1-His and pRS315-MMM1-GFP, respectively, was replaced by a PCR-generated SpeI-NcoI fragment encoding 3xMyc, an NcoI-BglII fragment encoding tobacco etch virus (TEV) cleavage site (TCS) (ENLYFQG) plus 3xHA, and a BglII-SalI fragment encoding Mmm1p (amino acids 2–190). For pRS315-TCSmut-MMM1-His, the 0.12-kb NcoI-BglII fragment of pRS315-TCS-MMM1-His was replaced by a PCR-generated NcoI-BglII fragment encoding the mutated TCS (ANLDFRP) plus 3xHA. For pYX142-TCS-NcMMM1, an EcoRI-SpeI fragment encoding 3xMyc-TCS-3xHA and a SpeI-EcoRV fragment encoding N. crassa MMM1 (amino acids 2–415) were generated by PCR and cloned into the EcoRI-EcoRV sites of pYX142, a low copy yeast expression vector with the strong constitutive TPI1 promoter (Novagen). For pYX142-TCS-MMM1-His, the 1.4-kb NcoI-EcoRV fragment of pYX142-TCS-NcMMM1 was replaced by a 0.69-kb NcoI-SalI fragment from pRS315-TCS-MMM1-His and a 0.73-kb SalI-EcoRV fragment from pRS416-MMM1-His. For pYX142-Su9-TCS-GFP, a SpeI-BglII fragment encoding 3xHA plus TCS was generated by PCR, and the SpeI site was converted to blunt ends using mung bean nuclease. The fragment was cloned into the BamHI-BglII sites of pYX142-mtGFP (the BamHI site was converted to blunt ends using DNA polymerase I, Large Fragment) (37). For pYX142-b2-TCS-GFP, the 0.21-kb EcoRI-BglII fragment of pYX142-mtGFP was replaced by a PCR-generated EcoRI-NheI fragment encoding Cyb2p (amino acids 1–220) and a SpeI-BglII fragment encoding 3xHA plus TCS. For pRS416GAL1-b2-TEV, a PCR-generated NheI-BglII fragment encoding Cyb2p (amino acids 1–220) and a BamHI-XhoI fragment from pWP1039 (provided by Will Prinz) encoding the TEV protease were cloned into the XbaI-XhoI sites of pRS416GAL1 (American Type Culture Collection). All clones used in this study were verified by DNA sequencing.

Site-specific TEV Cleavage Assay—Yeast cells harboring two plasmids encoding a TEV protease and a substrate protein were pregrown to an A600 of 1.5–3 in SD media. To induce TEV protease expression, cells were washed three times with 1 ml of SGalR media, resuspended to an A600 of 0.4–1, and grown to an A600 of 1–2 in SGalR media. 2 A600 units of cells were collected 3 h (for Su9-TCS-GFP), 6 h (for b2-TCS-GFP), or 16 h (for TCS-Mmm1p-His, TCSmut-Mmm1p-His, and TCS-Mmm1p-GFP) after induction, resuspended in 1 ml of H2O, and lysed by adding 160 µl of 1.85 M NaOH, 7.4% 2-mercaptoethanol on ice for 10 min. The protein extracts were precipitated with 6% trichloroacetic acid on ice for 10 min. The precipitates were rinsed with 1 ml of 1 M Tris base, resuspended in 60 µl of SDS gel-loading buffer, and boiled for 3 min. 15 µl of each sample was analyzed by SDS-PAGE and Western blotting.

Protease Protection Assay—Yeast cells expressing TCS-Mmm1p-His or TCS-Mmm1-GFP were grown in SR media, and the mitochondria were isolated as described previously (38), except that oxalyticase (Enzogenetics) was used for spheroplast formation (6 x 104 units in 30-ml suspension containing 2 x 103 A600 units of cells). Mitochondria (200 µg of protein) were resuspended in 1 ml of HS buffer (20 mM HEPES-KOH, pH 7.4, 0.6 M sorbitol), H buffer (20 mM HEPES-KOH, pH 7.4), or HT buffer (20 mM HEPES-KOH, pH 7.4, 1% Triton X-100). If indicated, the samples were treated with proteinase K at a final concentration of 200 µg/ml, and kept on ice for 30 min. The reaction was stopped by the addition of 2 mM phenylmethylsulfonyl fluoride on ice for 5 min. For the samples in HS and H buffers, centrifugation was performed at 16,000 x g for 10 min at 4 °C. The pellet was washed with 1 ml of HS and H buffers containing 1 mM phenylmethylsulfonyl fluoride, collected by centrifugation at 16,000 x g for 5 min at 4 °C, and resuspended in SDS gel-loading buffer. The sample in HT buffer was directly added to SDS gel-loading buffer. The samples corresponding to 30 µg of mitochondria were boiled for 3 min and analyzed by SDS-PAGE and Western blotting.

Immunoblot Analysis—Western blots were decorated with mouse monoclonal antibodies generated against HA (1:1,000; University of Utah Core Facility), GFP (1:1000; Covance), or 3-phosphoglycerate kinase (1:2,000; Molecular Probes) and rabbit antisera raised against the N-terminal domain of Fzo1p (1:2,000; Shaw laboratory), Cyb2p (1:2,000; Carla Koehler), or Ssc1p (1:1,000; Elizabeth Craig) in TBS buffer containing 5% non-fat dry milk for 1 h and washed three times with TBS buffer for 10 min. The blots were decorated with the goat anti-mouse or -rabbit horseradish peroxidase conjugate (1:8,000; Sigma) in TBS buffer containing 5% non-fat dry milk for 1 h and washed three times with TBS buffer for 10 min. Proteins were detected using ECL Plus and Hyperfilm ECL (Amersham Biosciences). All steps were performed at room temperature.

Microscopy and Imaging—Cells were visualized by a Zeiss Axioplan 2 Imaging microscope (Carl Zeiss) equipped with differential interference contrast (DIC) optics, epifluorescence capabilities, and a Zeiss Plan-Apochromat x100 (numerical aperture 1.4) objective. Filter sets (excitation/beamsplitter/emission) for GFP and red fluorescent protein fluorescence were BP 470 ± 20/FT 495/LP 525 ± 25, and BP 546 ± 6/FT 560/BP 575–640, respectively. Images were captured using a Zeiss AxioCam Mm monochrome digital camera and Zeiss AxioVision 3.1 software and assembled using Adobe Photoshop (Adobe Systems). Plasmids encoding organelle marker proteins and a staining probe used in this study were pVT100U-mtGFP (37), pRS424ADH-mtRFP (Shaw laboratory), and MitoFluor Red 589 (Molecular Probes) for mitochondria and pJK59 (Sec63p-GFP) (39) for the endoplasmic reticulum.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
The N Terminus of Mmm1p Is Exposed to the Mitochondrial Matrix—To investigate whether the N terminus of Mmm1p resides in the mitochondrial matrix or the IMS, we performed an in vivo site-specific cleavage assay using TEV protease (40) targeted to either the matrix or the IMS. This protease recognizes the seven-amino acid consensus sequence, EXXYXQ(S/G). Cleavage occurs between the conserved glutamine and serine or glycine residues (41). In vivo processing by TEV protease in budding yeast has been applied successfully to examine localization and topology of peroxisomal and mitochondrial proteins tagged with TCS (42, 43).

We constructed plasmids encoding TEV proteases fused to the C termini of different mitochondrial presequences including Su9 (amino acids 1–69) from ATPase subunit 9 of N. crassa and b2 (amino acids 1–220) from cytochrome b2 of S. cerevisiae (Su9-TEV and b2-TEV, respectively). Su9 and b2 are well characterized presequences that specifically target various passenger polypeptides, including TEV protease, to the matrix and the IMS, respectively (37, 43, 44).2 To determine the location of the N terminus, Mmm1p-His was fused to the C terminus of a wild-type TCS (TCS-Mmm1p-His) or a mutated TCS (TCSmut-Mmm1p-His) sequence (Fig. 1A). In addition, GFP was fused to the C terminus of TCS-Mmm1p (TCS-Mmm1p-GFP) (Fig. 1A). Each construct also contained a 1x–3x HA epitope tag, allowing recognition by immunoblot analysis. The TCS-Mmm1p-His, TCSmut-Mmm1p-His, and TCS-Mmm1p-GFP constructs were fully functional and restored normal mitochondrial networks and glycerol growth when expressed in mmm1{Delta} cells (data not shown). In addition, TCS-Mmm1p-GFP formed mitochondrial foci (data not shown).



View larger version (47K):
[in this window]
[in a new window]
  		FIG. 1.
The N terminus of Mmm1p is exposed to the mitochondrial matrix. A, schematic representation of proteins used for in vivo site-specific cleavage and in vitro protease protection assays. Mmm1p is 426 amino acids in length. The TM segment designated in Burgess et al. (27) is shown (amino acids 92–116). B, in vivo site-specific cleavage in cells harboring plasmids that encode a TEV protease (top) and a substrate protein (left). The expression of TEV proteases was repressed by dextrose (lanes 1 and 3) or induced by galactose (lanes 2 and 4). Protein extracts from 0.5 A600 units of cells were analyzed by immunostaining with antibodies specific for HA (for TCS-Mmm1p-His, TCSmut-Mmm1p-His, and TCS-Mmm1p-GFP expressed in mmm1{Delta} cells) and GFP (for Su9-TCS-GFP and b2-TCS-GFP expressed in wild-type cells). The asterisk indicates a form of Su9-TCS-GFP produced by the mitochondrial processing peptidase. TEV protease cleavage products for each substrate are marked to the left of each panel by an arrow, open arrowhead, closed arrowhead, or closed circle. C, in vitro PK treatment of mitochondria isolated from mmm1{Delta} cells expressing TCS-Mmm1p-His or TCS-Mmm1p-GFP. Mitochondria (30 µg of protein) were mock-treated (lanes 1 and 5), treated with 200 µg/ml PK under osmotic conditions (lanes 2 and 6), or subjected to osmotic shock (OS) in the absence (lanes 3 and 7) or presence (lanes 4 and 8) of Triton X-100 (TX-100). Samples were analyzed by immunostaining with antibodies specific for HA (for TCS-Mmm1p-His and TCS-Mmm1p-GFP), Fzo1p (for the OM protein), Cyb2p (for the IMS protein), and Ssc1p (for the matrix protein). Asterisks indicate forms produced by nonspecific degradation.

 
Processing of the TCS-tagged Mmm1p variants by TEV protease is predicted to generate proteins reduced in size by ~5 kDa that can be detected with anti-HA antibodies. When TCS-Mmm1p-His was coexpressed with Su9-TEV, low levels of the ~5-kDa shorter form were detected (Fig. 1B, lane 2, arrow; for each substrate, Su9-TEV and b2-TEV were expressed under identical conditions and for identical times). In contrast, very little of the ~5-kDa shorter form of TCS-Mmm1p-His was produced by the coexpression of b2-TEV (Fig. 1B, lane 4). In time course experiments, the amount of cleavage product produced by expression of Su9-TEV protease was consistently larger than that produced by b2-TEV expression (data not shown). The cleavage of TCS-Mmm1p-His by Su9-TEV occurred within the TCS, because Su9-TEV did not generate the ~5-kDa shorter form of the TCSmut-Mmm1p-His substrate (Fig. 1B, lane 2). Moreover, specific Su9-TEV-dependent processing was observed in cells expressing TCS-Mmm1p-GFP (Fig. 1B, lane 2, open arrowhead). Although the processing of TCS-Mmm1p-His and TCS-Mmm1p-GFP was not complete, additional control experiments indicated that it was specific to Su9-TEV. In particular, the substrates Su9-TCS-GFP and b2-TCS-GFP were processed only by Su9-TEV and b2-TEV, respectively (Fig. 1B, lane 2, closed arrowhead and lane 4, closed circle). The latter results demonstrate that TEV proteases are active in the matrix and IMS and that the processing occurs only when both a TEV protease and a substrate are localized in the same mitochondrial subcompartment. Together, these results suggest that the N terminus of Mmm1p is exposed in the matrix.

To further examine the topology of Mmm1p, we carried out a conventional protease protection assay in vitro. Mitochondria were isolated from mmm1{Delta} cells expressing TCS-Mmm1p-His or TCS-Mmm1p-GFP and incubated in the presence or absence of proteinase K (PK) with no treatment (to keep the organelle intact), osmotic shock treatment (to expose the IMS), or Triton X-100 treatment (to solubilize the organelle). As shown in Fig. 1C, lanes 1 and 5, anti-HA antibodies decorated the full-length ~80-kDa TCS-Mmm1p-His and ~107-kDa TCS-Mmm1p-GFP polypeptides in mitochondrial samples that were not treated with protease. In contrast, when intact mitochondria were exposed to PK (200 µg/ml), an ~40-kDa proteolytic fragment was detected (Fig. 1C, lanes 2 and 6). The apparent molecular weight of the protected polypeptide was larger than predicted for this substrate (a cleavage product consisting of 3xMyc-TCS-3xHA plus the first 120-amino acid residues of Mmm1p should migrate at ~22 kDa). This size increase was not because of incomplete proteolysis, as treatment with 1 mg/ml PK also generated an ~40-kDa form (data not shown). During the course of this study, we noticed that all variants of Mmm1p migrated slower (appeared larger) on SDS gels than predicted. The reason for this altered migration is not known. Nevertheless, when osmotically shocked mitochondria containing TCS-Mmm1p-His or TCS-Mmm1p-GFP were treated with PK, the ~40-kDa clipped form was not degraded further, even though the OM was clearly ruptured, allowing the IMS protein Cyb2 to be digested (Fig. 1C, lanes 3 and 7). In contrast, the protected ~40-kDa form was completely degraded and could not be detected in PK-treated samples containing Triton X-100 (Fig. 1C, lanes 4 and 8). Under this condition, both mitochondrial membranes were lysed as assessed by degradation of the matrix protein Ssc1. Because the HA tag recognized in these experiments resides near the N termini of the substrates, these results indicate that the N termini of TCS-Mmm1p-His and TCS-Mmm1p-GFP can only be digested when PK has access to proteins in the matrix. Similar results were obtained when experiments were performed using a different protease, trypsin (data not shown). Based on the combined results, we conclude that Mmm1p has the topology of Nmatrix-Ccytoplasm, spanning the outer and inner mitochondrial membranes.

The N-terminal Extension of Mmm1p Is Not Required for Maintenance of Mitochondrial Morphology and MtDNA—To test whether the N-terminal extension of Mmm1p is critical for mitochondrial morphology and/or mtDNA maintenance, we constructed plasmids encoding Mmm1 mutant proteins that lack amino acid residues 2–90 and 2–120 (Mmm1{Delta}(2–90)p-His and Mmm1{Delta}(2–120)p-His, respectively) (Fig. 2A). Mmm1{Delta}(2–90)p-His lacks nearly the entire N-terminal extension in front of the predicted TM segment (amino acids 92–116) (27). Both the N-terminal extension and the TM segment are deleted in Mmm1{Delta}(2–120)p-His. When expressed in mmm1{Delta} cells, Mmm1{Delta}(2–90)p-His restored normal mitochondrial tubular networks that were indistinguishable from those in mmm1{Delta} cells expressing full-length Mmm1p-His (see Fig. 2B and Table I). In contrast, mmm1{Delta} cells expressing Mmm1{Delta}(2–120)p-His exhibited the large spherical mitochondria typically seen in mmm1{Delta} cells containing an empty plasmid (see Fig. 2B and Table I).



View larger version (41K):
[in this window]
[in a new window]
  		FIG. 2.
The N-terminal extension of Mmm1p is not essential for mitochondrial morphology or mtDNA maintenance. A, schematic representation of Mmm1p deletion mutant proteins. B, mmm1{Delta} cells carrying plasmid pRS315 alone (no protein expressed) or encoding Mmm1p-His, Mmm1{Delta}(2–90)p-His, or Mmm1{Delta}(2–120)p-His were transformed with pVT100U-mtGFP, grown to log phase in SD media, and observed by fluorescence microscopy to visualize mitochondria. Bar, 5 µm. C, cells described in B (without pVT100U-mtGFP) were grown overnight in SD media, spotted as 2-fold serial dilutions, and incubated for three (SD plates) or six (SGly plates) days.

 


View this table:
[in this window]
[in a new window]
  		TABLE I
Mitochondrial membrane morphology in mmm1 {Delta} cells

Numbers are percent of cells grown to log phase in SD media at 30°C (n = 200). Mitochondrial morphology was visualized with matrix-targeted GFP and scored in the three different categories shown.

 
Although Mmm1p-His and Mmm1{Delta}(2–90)p-His restored glycerol growth when shuffled into mmm1{Delta} cells, mmm1{Delta} cells expressing Mmm1{Delta}(2–120)p-His or containing an empty plasmid did not grow on glycerol medium (Fig. 2C). No glycerol growth defect was observed in mmm1{Delta} cells expressing Mmm1{Delta}(2–90)p-His at 37 °C (data not shown). Consistent with this growth pattern, the morphology and number of 4',6-diamidino-2-phenylindole-stained mtDNA nucleoids in mmm1{Delta} cells expressing Mmm1p-His or Mmm1{Delta}(2–90)p-His were similar, whereas mmm1{Delta} cells expressing Mmm1{Delta}(2–120)p-His or containing an empty plasmid lacked such structures (data not shown). The failure of Mmm1{Delta}(2–120)p-His to complement mutant phenotypes in mmm1{Delta} cells is likely because of the absence of the TM segment, which is required for mitochondrial membrane insertion (see Fig. 6). Our data suggest that the N-terminal extension of Mmm1p is dispensable for the maintenance of normal mitochondrial tubular networks and mtDNA.



View larger version (41K):
[in this window]
[in a new window]
  		FIG. 6.
The C-terminal domain of Mmm1p plays an essential role in mitochondrial targeting. A, schematic representation of Mmm1p deletion variants fused to the N terminus of GFP. Wild-type cells containing pRS425 encoding each GFP fusion protein or pJK59 encoding the ER marker Sec63p-GFP were transformed with pRS424ADH-mtRFP, grown to log phase in SD media, and observed by fluorescence microscopy. Localization of these GFP fusion proteins is summarized at the right. B, representative images of wild-type cells expressing Mmm1p-GFP, Mmm1 (1–135)p-GFP, Mmm1 (1–415)p-GFP, Mmm1{Delta}(2–120)p-GFP, or Sec63p-GFP. Bar, 5 µm. DIC, differential interference contrast.

 
Yeast mmm1{Delta} Cells Expressing NcMMM1 Can Maintain Normal Mitochondrial Tubular Networks and MtDNA—Although S. cerevisiae Mmm1p is fairly similar to the N. crassa homologue NcMMM1 (30% identity and 54% similarity), there is a significant difference between the two proteins at the N termini. In particular, Mmm1p has ~90 amino acid residues N-terminal to the predicted TM segment, whereas NcMMM1 has ~10 amino acid residues (32). The fact that Mmm1{Delta}(2–90)p-His is as functional as Mmm1p-His prompted us to test whether NcMMM1 can substitute for the yeast homologue. We constructed a yeast expression vector encoding NcMMM1 fused to the C terminus of a TCS (TCS-NcMMM1). The same vector was used for the expression of TCS-Mmm1p-His as a control. When expressed in mmm1{Delta} cells, TCS-NcMMM1 restored normal mitochondrial tubular networks that were virtually identical to those in mmm1{Delta} cells expressing TCS-Mmm1p-His (see Fig. 3A and Table I). In addition, the glycerol growth defect of mmm1{Delta} cells was suppressed by the expression of TCS-NcMMM1 or TCS-Mmm1p-His at 30 °C (Fig. 3B) and 37 °C (data not shown). 4',6-diamidino-2-phenylindole staining confirmed that normal mtDNA nucleoids were present in mmm1{Delta} cells expressing TCS-NcMMM1 or TCS-Mmm1p-His (data not shown). In vivo TEV cleavage and in vitro protease protection assays suggest that the N terminus of NcMMM1 is also exposed to the matrix of yeast mitochondria.3 Together, these results demonstrate that, despite the lack of an N-terminal extension, NcMMM1 can perform the function of Mmm1p in yeast.



View larger version (57K):
[in this window]
[in a new window]
  		FIG. 3.
N. crassa MMM1 can substitute for Mmm1p in yeast. A, mmm1{Delta} cells carrying the plasmid pYX142 alone (no protein expressed) or encoding TCS-Mmm1p-His or TCS-NcMMM1 were transformed with pVT100U-mtGFP, grown to log phase in SD media, and observed by fluorescence microscopy to visualize mitochondria. Bar, 5 µm. B, cells described in A (without pVT100U-mtGFP) were grown overnight in SD media, spotted as 2-fold serial dilutions, and incubated for three (SD plates) or six (SGly plates) days.

 

Deletion of the Mmm1p N-terminal Extension Does Not Affect Foci Formation but Reduces the Steady-state Level of Mmm1p— To examine whether the N-terminal extension of Mmm1p is required for foci formation, we constructed plasmids encoding Mmm1p-GFP mutant proteins lacking the amino acid residues 2–75 and 2–90 (Mmm1{Delta}(2–75)p-GFP and Mmm1{Delta}(2–90)p-GFP, respectively) (Fig. 4A). When expressed in mmm1{Delta} cells, both Mmm1{Delta}(2–75)p-GFP and Mmm1{Delta}(2–90)p-GFP formed discrete foci on mitochondrial tubules identical to those observed in mmm1{Delta} cells expressing Mmm1p-GFP (Fig. 4B) (30). The number of cells exhibiting these Mmm1p foci, the number of foci per cell, and the fluorescence intensity of the foci were similar for all three constructs, Mmm1{Delta}(2–75)p-GFP, Mmm1{Delta}(2–90)p-GFP, and Mmm1p-GFP (data not shown). Moreover, the morphology of restored mitochondrial networks was identical in mmm1{Delta} cells expressing Mmm1{Delta}(2–75)p-GFP, Mmm1{Delta}(2–90)p-GFP, and Mmm1p-GFP (Fig. 4B) (30). These observations suggest that the N-terminal extension of Mmm1p is dispensable for foci formation.



View larger version (47K):
[in this window]
[in a new window]
  		FIG. 4.
Deletion of the Mmm1p N-terminal extension does not affect foci formation but reduces steady-state protein levels. A, schematic representation of Mmm1p deletion variants fused to the N terminus of GFP. B, mmm1{Delta} cells carrying the plasmid pRS315 encoding Mmm1p-GFP, Mmm1{Delta}(2–75)p-GFP, or Mmm1{Delta}(2–90)p-GFP were grown to log phase in SD media, stained with MitoFluor Red 589, and observed by fluorescence microscopy. Bar, 5 µm. C, protein extracts from 0.5 A600 units of cells described in B were analyzed by immunostaining with antibodies specific for GFP (for Mmm1p-GFP, Mmm1{Delta}(2–75)p-GFP, and Mmm1{Delta}(2–90)p-GFP) and 3-phosphoglycerate kinase (as a cytoplasmic protein loading control). WT, wild-type.

 
Surprisingly, we found that the steady-state level of Mmm1{Delta}(2–90)p-GFP was 4- to 6-fold lower than that of Mmm1p-GFP and Mmm1{Delta}(2–75)p-GFP (Fig. 4C). This observation indicates that amino acid residues 75–90 somehow affect expression or turnover of the Mmm1 mRNA or protein. Because mutant fusion proteins lacking the N-terminal extension function as well as the wild-type fusion protein produce similar numbers of Mmm1p foci, these data suggest that the steadystate level of Mmm1p expressed in wild-type cells is not limiting.

The TM Segment of Mmm1p Is Essential for Foci Formation—To investigate the role of the Mmm1p TM segment in foci formation, we constructed plasmids encoding chimeric proteins containing the TM segment of Tom20p or NcMMM1 and the large C-terminal domain of Mmm1p-His (T20(TM)-Mmm1(C)p-His and NcM(TM)-Mmm1(C)p-His, respectively) (Fig. 5A). For localization studies, GFP was fused at the C terminus of T20(TM)-Mmm1(C)p and NcM(TM)-Mmm1(C)p (T20(TM)-Mmm1(C)p-GFP and NcM(TM)-Mmm1(C)p-GFP, respectively) (Fig. 5A). Tom20p contains a single TM segment at the N terminus embedded in the outer mitochondrial membrane and has the topology of Nin-Ccytoplasm (45). A previous study (46) showed that the human Tom20 TM segment plus some flanking residues fused at the N terminus of GFP can be targeted and inserted into the OM in humans.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 5.
The double membrane-spanning topology of Mmm1p is crucial for foci formation. A, schematic representation of chimeric fusion proteins. B, wild-type cells carrying the plasmid pRS315 encoding T20(TM)-GFP, T20(TM)-Mmm1(C)p-GFP, or NcM(TM)-Mmm1(C)p-GFP were grown to log phase in SD media, stained with MitoFluor Red 589, and observed by fluorescence microscopy. Bar, 5 µm. C, mmm1{Delta} cells carrying the plasmid pRS315 encoding T20(TM)-Mmm1(C)p-His or NcM(TM)-Mmm1(C)p-His were transformed with pVT100U-mtGFP, grown to log phase in SD media, and observed by fluorescence microscopy to visualize mitochondria. Bar, 5 µm. D, mmm1{Delta} cells carrying the plasmid pRS315 alone (no protein expressed) or encoding Mmm1p-His, T20(TM)-Mmm1(C)p-His, or NcM(TM)-Mmm1(C)p-His were grown overnight in SD media, spotted as 2-fold serial dilutions, and incubated for 3 days (SD plates) or 6 days (SGly plates). DIC, differential interference contrast.

 
When expressed in wild-type cells, GFP fused to the C terminus of the Tom20p TM segment (T20(TM)-GFP; see Fig. 5A) colocalized with mitochondria and was evenly distributed throughout the tubular network (Fig. 5B). Similarly, T20(TM)-Mmm1(C)p-GFP localized throughout the mitochondrial network without forming visible foci (Fig. 5B). In contrast, NcM(TM)-Mmm1(C)p-GFP formed foci that colocalized with mitochondrial tubules (Fig. 5B). These results indicate that the TM segment of Mmm1p is essential for foci formation and can be replaced with that of NcMMM1. Interestingly, T20(TM)-Mmm1(C)p-His did not restore normal mitochondrial tubular networks and glycerol growth in mmm1{Delta} cells, whereas NcM(TM)-Mmm1(C)p-His suppressed these defects in the mutant (see Fig. 5, C and D, and Table I). These data, together with the observation that the N terminus of NcMMM1 is exposed to the matrix of yeast mitochondria,3 suggest that the double membrane-spanning topology is crucial for Mmm1p to localize as discrete foci where the protein performs its function.

The C-terminal Domain of Mmm1p Is Indispensable for Targeting to Mitochondria—To define the region required for targeting Mmm1p to mitochondria, we expressed plasmids encoding GFP fused to the C terminus of various Mmm1p deletion mutants in wild-type cells (Fig. 6A). Surprisingly, Mmm1(1–135)p-GFP was targeted to perinuclear structures similar to those labeled with the endoplasmic reticulum (ER) marker Sec63p-GFP (39) (Fig. 6B). Because we already showed that removal of the N terminus does not interfere with Mmm1p targeting or foci formation, this finding indicates that the TM segment of Mmm1p does not contain a complete signal for targeting to mitochondria. Additional studies established that Mmm1(1–180)p-GFP and Mmm1(1–275)p-GFP were also targeted to the ER, whereas Mmm1(1–365)p-GFP and Mmm1(1–390)p-GFP were targeted to both the ER and cytoplasm (Fig. 6A). In contrast, Mmm1(1–415)p-GFP formed discrete foci that were colocalized with the mitochondrial network (Fig. 6B) and restored normal mitochondrial morphology and glycerol growth in mmm1{Delta} cells (data not shown). These results indicate that the last 11 amino acids of Mmm1p are not required for mitochondrial targeting and foci formation. Mmm1{Delta}(2–120)p-GFP, a mutant lacking the TM segment, was targeted to the cytoplasm (Fig. 6B). Collectively, these observations demonstrate that, together with the TM segment, the C-terminal domain plays an important role in targeting Mmm1p to mitochondria.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Double Membrane-spanning Topology of Mmm1p—Previous studies suggested the following: 1) that Mmm1p spans the OM with its C terminus exposed to the cytoplasm (27), and 2) that Mmm1p forms discrete foci at membrane contact sites, adjacent to mtDNA nucleoids in the matrix (30). Although the latter finding raised the possibility that the N terminus of Mmm1p spans the IM, the submitochondrial location of the N terminus of Mmm1p has not been reported (30). In this study, we used in vivo site-specific cleavage and in vitro protease protection assays to demonstrate that the N terminus of Mmm1p is exposed to the mitochondrial matrix. Together, these findings establish that Mmm1p spans both mitochondrial membranes with the topology of Nmatrix-Ccytoplasm.

Tim23p is the only other protein reported to span both mitochondrial membranes in yeast (47). This protein contains an N-terminal domain anchored in the OM, a middle (hydrophilic) domain exposed to the IMS, and four C-terminal TM helices embedded in the IM (47). Unlike Tim23p, Mmm1p contains only one predicted TM segment and is unlikely to contain a stretch exposed to the IMS. This interpretation is supported by our observation that the N-terminal extension of Mmm1p was not degraded further by externally added protease even when the OM was ruptured by osmotic shock (Fig. 1C).

In the absence of a second TM segment, at least three different models could account for the double membrane-spanning topology of Mmm1p. First, Mmm1p might span both mitochondrial membranes once via its single TM segment, as depicted in Fig. 7A. This scenario is problematic, if the TM helix of Mmm1p is only ~25 amino acid residues in length (amino acids 92–116), as predicted previously (Kyte-Doolittle hydrophobicity prediction) (27). Unless this region adopts an unconventional structure (e.g. a dramatically extended helix), additional amino acids flanking the predicted TM helix would be required to span two lipid bilayers. In fact, other prediction programs including SOSUI (amino acids 102–123), HMMTOP (amino acids 101–120), and TMHMM (amino acids 100–122) predict slightly different TM helices within the Mmm1 protein (ExPASy Molecular Biology Server). When the results of all four prediction programs are combined, an Mmm1p TM segment of 32 residues (amino acids 92–123) results. If a TM segment of this length exists in Mmm1p, it might be sufficient to span both mitochondrial membranes. Second, the TM segment of Mmm1p could be embedded in the OM, and the N-terminal extension could span the IM through an aqueous pore formed by other members of a protein complex (Fig. 7B). Third, the C-terminal flanking region of the Mmm1p TM segment could span the OM through an aqueous pore formed by proteins in the OM (Fig. 7C). Further studies are required to define the length, position, structure, and environment of the double membrane-spanning domain of Mmm1p.



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 7.
Schematic models for the double membrane-spanning topology of Mmm1p. Asterisks mark hypothetical pore-forming proteins in the outer or inner membrane. See text for details.

 
A recent study (48) indicates that two additional OM proteins, Mdm10p and Mdm12p, form mitochondrial foci adjacent to mtDNA nucleoids. Like the mmm1{Delta} strain, cells lacking Mdm10p and Mdm12p contain spherical mitochondria and lose mtDNA nucleoids (25, 26). In addition, all three proteins behave as if they form a complex in immune precipitation experiments (48), and localization of Mmm1p to mitochondrial foci does not occur in the absence of Mdm10p and Mdm12p (48).4 Based on these observations, it is likely that Mdm10p and Mdm12p interact with Mmm1p, forming a complex at the membrane contact site. The topology of Mdm10p and Mdm12p with respect to both mitochondrial membranes and the domains of those proteins required for interaction with Mmm1p are important issues to be addressed in the future.

Role of the Mmm1p TM Segment in Foci Formation and Generation of Contact Sites—Our finding that the double membrane-spanning topology of Mmm1p is critical for foci formation (Fig. 5) supports the notion that Mmm1p functions at contact sites between the inner and outer mitochondrial membranes. Although we cannot rule out the possibility that a subgroup of Mmm1p molecules is distributed throughout mitochondria in a single membrane-spanning fashion, the results presented here suggest that the membrane contact site is the place where Mmm1p forms a functional complex. Despite three decades of research on mitochondrial structure and protein import, proteins that direct the formation of these morphologically defined membrane contact sites have proven difficult to identify (49). Here we demonstrate that the double membrane-spanning protein Mmm1 serves as a marker of this intramitochondrial structure. Is Mmm1p a structural determinant of the mitochondrial membrane contact site, attaching the OM to the IM? Based on the observation that the IM is strikingly disorganized in mmm1{Delta} cells, a role for Mmm1p in forming this structure has been proposed (30). However, the disorganized IM observed in mmm1{Delta} mutants could also be a secondary consequence of converting mitochondrial membranes from tubular networks to spherical compartments. Moreover, it has not been firmly established whether the membrane contact sites are lost and/or destabilized in mitochondria lacking Mmm1p. High resolution electron microscopy and three-dimensional tomography of mmm1{Delta} mitochondria will help to address this issue.

Role of the Mmm1p N-terminal Extension in MtDNA Nucleoid Maintenance—The unique Nmatrix-Ccytoplasm topology of Mmm1p raised the possibility that this is a bifunctional protein with two topologically distinct domains, an N-terminal extension for mtDNA nucleoid maintenance and a C-terminal domain for mitochondrial morphology maintenance. This hypothesis predicts that deletion of the N-terminal extension will lead to mtDNA loss but will not affect mitochondrial morphology. However, data presented here indicate that the N-terminal extension of Mmm1p is not required for maintenance of normal tubular networks or mtDNA nucleoids (Fig. 2). Thus, it is unlikely that Mmm1p directly anchors mtDNA nucleoids to the matrix side of the IM. This interpretation is supported by our observation that, despite the lack of an N-terminal extension, the N. crassa MMM1 homologue is fully functional in budding yeast (Fig. 3).

How is Mmm1p involved in maintenance of mtDNA? We cannot exclude the possibility that Mmm1p functions indirectly to attach mtDNA nucleoids to the IM. For example, Mmm1p may interact with one or more proteins whose matrix-exposed domains directly bind mtDNA nucleoids. In the absence of functional Mmm1p, such protein(s) might not target, assemble, or function properly to anchor mtDNA nucleoids at the membrane contact site, ultimately resulting in mtDNA loss. Alternatively, unknown factor(s) might anchor mtDNA nucleoids to the membrane contact site independently of Mmm1p. In this scenario, mtDNA nucleoid aggregation and the subsequent mtDNA loss phenotypes of mmm1 mutants (30) may be a secondary consequence of mitochondrial morphology defects that grossly alter IM structure.

Role of Mmm1p in Cytoskeletal Attachment—Although a role for Mmm1p in cytoskeletal attachment has been proposed, two lines of evidence suggest that this protein does not directly mediate interactions between mitochondria and the cytoskeleton. First, in N. crassa, microtubules are the cytoskeletal elements utilized for mitochondrial morphology, distribution, and inheritance (31). However, in vitro studies showed that NcMMM1-depleted mitochondria still bind to microtubules (33), indicating that NcMMM1 is not essential for microtubule-dependent mitochondrial behavior. Second, we show here that NcMMM1 can substitute for Mmm1p in S. cerevisiae (Fig. 3). In contrast to N. crassa, S. cerevisiae utilizes the actin cytoskeleton for mitochondrial attachment and movement (5, 50). Thus, the primary function of Mmm1p has apparently been conserved between S. cerevisiae and N. crassa, independent of the cytoskeletal system used by the cell to regulate mitochondrial dynamics. Based on these results, it seems unlikely that the Mmm1 protein acts as a direct molecular bridge between mitochondria and cytoskeletal elements to maintain tubular networks.

How, then, can we reconcile the observation that S. cerevisiae mitochondria bind actin filaments in an Mmm1p-dependent fashion in vitro (28)? One possibility is that S. cerevisiae and N. crassa Mmm1 proteins bind molecular adaptors that bridge interactions between mitochondria and actin filaments in both cell types. Loss of actin interactions would have a severe affect in budding yeast, because actin filaments and cables are the major cytoskeletal system found in the cytoplasm. In contrast, both actin and microtubules are found as prominent cytoplasmic structures in N. crassa. Because N. crassa relies most heavily on microtubules for mitochondrial dynamics and presumably has additional proteins that mediate microtubule attachment, loss of Mmm1p actin binding activity produces only subtle (if any) defects. Identifying the complete set of cytoskeleton-binding and motor proteins that interact with mitochondria in both organisms will help to clarify this discrepancy.

The Mmm1p Cytoplasmic Domain Is Required for Mitochondrial Targeting—To our surprise, we found that Mmm1p mutant proteins lacking the cytoplasmic C-terminal domain were targeted to the ER rather than mitochondria (Fig. 6). This result indicates that all or a part of the C-terminal domain serves as a subsignal for mitochondrial targeting of Mmm1p. To date, only one other protein is known to require cytoplasmic signals for mitochondrial targeting. An {alpha}-helix at the N terminus of the single TM protein Bax was shown recently (51) to contribute significantly to mitochondrial targeting. These cytoplasmic subsignals may interact with receptor proteins or specific recruiting factors that facilitate faithful targeting.

Why are Mmm1p mutants lacking the C-terminal domain targeted to the ER? This phenomenon may be because of the hydrophobicity and/or length of the membrane-spanning domain. For example, human Tom20 variants with increased hydrophobicity and length in the TM segment are targeted to ER/Golgi compartments (46). Similarly, increasing the TM domain hydrophobicity of another outer mitochondrial membrane protein, Fis1, also results in ER targeting (52). In the absence of cytoplasmic C-terminal sequences, the excess hydrophobicity and/or length of the Mmm1p TM segment may define a default signal for ER targeting.

What are the proteins that interact with the C-terminal domain of Mmm1p and recruit Mmm1p to mitochondria? As described above, recent studies suggest that at least two other proteins, Mdm10p and Mdm12p, are required to target Mmm1p to mitochondria (48).4 In addition to components of the OM protein import machinery, other as yet unidentified proteins may play a role in targeting Mmm1p to mitochondria. Studying how Mmm1p is targeted to mitochondria, inserted into the outer and inner membranes, and assembled into a complex at the membrane contact site will help us to understand the molecular function of Mmm1p and the mechanisms and pathways underlying protein transport to a distinct intramitochondrial structure, the membrane contact site.


   FOOTNOTES
 
* This work was supported in part by National Institutes of Health Grant GM-53466 (to J. M. S.) and by a grant from the United Mitochondrial Disease Foundation (to K. O.). The University of Utah DNA and Peptide Facility was supported in part by NCI, National Institutes of Health Grant CA42014. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} To whom correspondence may be addressed: Dept. of Biology, University of Utah, 257 S. 1400 E., Salt Lake City, UT 84112. Tel.: 801-585-6205; Fax: 801-581-2174; E-mail: shaw{at}bioscience.utah.edu. § To whom correspondence may be addressed: Dept. of Biology, University of Utah, 257 S. 1400 E., Salt Lake City, UT 84112. Tel.: 801-587-9209; Fax: 801-581-2174; E-mail: kokamoto{at}biology.utah.edu.

1 The abbreviations used are: mtDNA, mitochondrial DNA; OM, outer membrane; IMS, intermembrane space; IM, inner membrane; TM, transmembrane; GFP, green fluorescent protein; NcMMM1, N. crassa MMM1; UTR, untranslated region; TEV, tobacco etch virus; TCS, TEV cleavage site; HA, hemagglutinin; PK, proteinase K; ER, endoplasmic reticulum. Back

2 J. W. Thatcher and J. M. Shaw, unpublished data. Back

3 N. Kondo-Okamoto, unpublished data. Back

4 N. Kondo-Okamoto, unpublished data. Back


   ACKNOWLEDGMENTS
 
We thank Benedikt Westermann (University of Munich, Munich, Germany) for plasmids pVT100U-mtGFP and pYX142-mtGFP and the cDNA encoding NcMMM1, Will Prinz (National Institutes of Health, Bethesda, MD) for plasmids pJK59 and pWP1039, Carla Koehler (University of California, Los Angeles, Los Angeles, CA) for anti-Cyb2p antisera, Elizabeth Craig (University of Wisconsin, Madison, WI) for anti-Ssc1p antisera, and members of the Shaw laboratory for stimulating discussions and careful review of the manuscript. N. K.-O. and K. O. are deeply grateful to Walter Neupert (University of Munich, Munich, Germany) for generous support during the initial stage of this study.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Hermann, G. J., and Shaw, J. M. (1998) Ann. Rev. Cell Dev. Biol. 14, 265–303[CrossRef][Medline] [Order article via Infotrieve]
  2. Yaffe, M. P. (1999) Science 283, 1493–1497[Abstract/Free Full Text]
  3. Griparic, L., and van der Bliek, A. M. (2001) Traffic 2, 235–244[CrossRef][Medline] [Order article via Infotrieve]
  4. Jensen, R. E., Aiken Hobbs, A. E., Cerveny, K. L., and Sesaki, H. (2001) Microsc. Res. Tech. 51, 573–583
  5. Boldogh, I. R., Yang, H.-C., and Pon, L. A. (2001) Traffic 2, 368–374[CrossRef][Medline] [Order article via Infotrieve]
  6. Nunnari, J., Marshall, W. F., Straight, A., Murray, A., Sedat, J. W., and Walter, P. (1997) Mol. Biol. Cell 8, 1233–1242[Abstract]
  7. Yoon, Y., and McNiven, M. A. (2001) Curr. Biol. 11, R67-R70[CrossRef][Medline] [Order article via Infotrieve]
  8. Shaw, J. M., and Nunnari, J. (2002) Trends Cell Biol. 12, 178–184[CrossRef][Medline] [Order article via Infotrieve]
  9. Westermann, B. (2002) EMBO Rep. 3, 527–531[CrossRef][Medline] [Order article via Infotrieve]
  10. Mozdy, A. D., and Shaw, J. M. (2003) Nat. Rev. Mol. Cell. Biol. 4, 468–478[CrossRef][Medline] [Order article via Infotrieve]
  11. Dimmer, K. S., Fritz, S., Fuchs, F., Messerschmitt, M., Weinbach, N., Neupert, W., and Westermann, B. (2002) Mol. Biol. Cell 13, 847–853[Abstract/Free Full Text]
  12. Berger, K. H., and Yaffe, M. P. (2000) Trends Microbiol. 8, 508–513[CrossRef][Medline] [Order article via Infotrieve]
  13. Contamine, V., and Picard, M. (2000) Microbiol. Mol. Biol. Rev. 64, 281–315[Abstract/Free Full Text]
  14. Scott, S. V., Cassidy-Stone, A., Meeusen, S. L., and Nunnari, J. (2003) Curr. Opin. Cell Biol. 15, 482–488[CrossRef][Medline] [Order article via Infotrieve]
  15. Hermann, G. J., Thatcher, J. W., Mills, J. P., Hales, K. G., Fuller, M. T., Nunnari, J., and Shaw, J. M. (1998) J. Cell Biol. 143, 359–373[Abstract/Free Full Text]
  16. Rapaport, D., Brunner, M., Neupert, W., and Westermann, B. (1998) J. Biol. Chem. 273, 20150–20155[Abstract/Free Full Text]
  17. Sesaki, H., and Jensen, R. E. (2001) J. Cell Biol. 152, 1123–1134[Abstract/Free Full Text]
  18. Jones, B. A., and Fangman, W. L. (1992) Genes Dev. 6, 380–389[Abstract/Free Full Text]
  19. Guan, K., Farh, L., Marshall, T. K., and Deschenes, R. J. (1993) Curr. Genet. 24, 141–148[CrossRef][Medline] [Order article via Infotrieve]
  20. Shepard, K. A., and Yaffe, M. P. (1999) J. Cell Biol. 144, 711–720[Abstract/Free Full Text]
  21. Wong, E. D., Wagner, J. A., Gorsich, S. W., McCaffery, J. M., Shaw, J. M., and Nunnari, J. (2000) J. Cell Biol. 151, 341–352[Abstract/Free Full Text]
  22. Esser, K., Tursun, B., Ingenhoven, M., Michaelis, G., and Pratje, E. (2002) J. Mol. Biol. 323, 835–843[CrossRef][Medline] [Order article via Infotrieve]
  23. Herlan, M., Vogel, F., Bornhovd, C., Neupert, W., and Reichert, A. S. (2003) J. Biol. Chem. 278, 27781–27788[Abstract/Free Full Text]
  24. McQuibban, G. A., Saurya, S., and Freeman, M. (2003) Nature 423, 537–541[CrossRef][Medline] [Order article via Infotrieve]
  25. Sogo, L. F., and Yaffe, M. P. (1994) J. Cell Biol. 126, 1361–1373[Abstract/Free Full Text]
  26. Berger, K. H., Sogo, L. F., and Yaffe, M. P. (1997) J. Cell Biol. 136, 545–553[Abstract/Free Full Text]
  27. Burgess, S. M., Delannoy, M., and Jensen, R. E. (1994) J. Cell Biol. 126, 1375–1391[Abstract/Free Full Text]
  28. Boldogh, I., Vojtov, N., Karmon, S., and Pon, L. A. (1998) J. Cell Biol. 141, 1371–1381[Abstract/Free Full Text]
  29. Yang, H.-C., Palazzo, A., Swayne, T. C., and Pon, L. A. (1999) Curr. Biol. 9, 1111–1114[CrossRef][Medline] [Order article via Infotrieve]
  30. Aiken Hobbs, A. E., Srinivasan, M., McCaffery, J. M., and Jensen, R. E. (2001) J. Cell Biol. 152, 401–410[Abstract/Free Full Text]
  31. Steinberg, G., and Schliwa, M. (1993) J. Cell Sci. 106, 555–564[Abstract]
  32. Prokisch, H., Neupert, W., and Westermann, B. (2000) Mol. Biol. Cell 11, 2961–2971[Abstract/Free Full Text]
  33. Fuchs, F., Prokisch, H., Neupert, W., and Westermann, B. (2002) J. Cell Sci. 115, 1931–1937[Abstract/Free Full Text]
  34. Winston, F., Dollard, C., and Ricupero-Hovasse, S. L. (1995) Yeast 11, 53–55[CrossRef][Medline] [Order article via Infotrieve]
  35. Burke, D., Dawson, D., and Stearns, T. (2000) Methods in Yeast Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
  36. Sambrook, J., and Russell, D. W. (2001) Molecular Cloning, 3rd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
  37. Westermann, B., and Neupert, W. (2000) Yeast 16, 1421–1427[CrossRef][Medline] [Order article via Infotrieve]
  38. Diekert, K., de Kroon, A. I., Kispal, G., and Lill, R. (2001) Methods Cell Biol. 65, 37–51[Medline] [Order article via Infotrieve]