Native mitochondrial creatine kinase forms octameric structures. II. Characterization of dimers and octamers by ultracentrifugation, direct mass measurements by scanning transmission electron microscopy, and image analysis of single mitochondrial creatine kinase octamers

Electron micrographs of negatively stained and metal-shadowed mitochondrial creatine kinase (Mi-CK) molecules purified as described by Schlegel et al. (Schlegel, J., Zurbriggen, B., Wegmann, E., Wyss, M., Eppenberger, H. M., and Wallimann, T. (1988) J. Biol Chem. 263, 16942-16953) revealed a homogeneous population (greater than or equal to 95%) of distinctly sized square-shaped, octameric particles with a side length of 10 nm that frequently exhibited a pronounced 4-fold axis of symmetry. The cube-like molecules consist of four dimers that are arranged around a stain-accumulating central cavity of 2.5-3 nm in diameter. This interpretation is supported by single particle averaging including correlation analysis by computer. Upon prolonged storage or high dilution, the cube-like octamers tended to dissociate into "banana-shaped" dimers. Sedimentation velocity and sedimentation equilibrium experiments yielded an s value of 12.8-13.5 S and an Mr of 328,000 +/- 25,000 for the octameric cubes. An s value of 5.0 S and a Mr of 83,000 +/- 8,000 was found under conditions which revealed banana-shaped dimers. These dimers proved to be very stable, as their dissociation into monomers of 45 kDa (s value = 2.0 S) required 6 M guanidine HCl. Thus, the oligomeric structures observed in the electron microscope are identified as Mi-CK dimers (banana-shaped structures) and cubical Mi-CK octamers assembled from four Mi-CK dimers. The octameric nature of native Mi-CK and the formation of Mi-CK dimers were confirmed by direct mass measurements of individual molecules by scanning transmission electron microscopy yielding a molecular mass of 340 +/- 55 kDa for the octamer and 89 +/- 27 kDa for the dimer. A structural model of Mi-CK octamers and the possible interaction with ATP/ADP-translocator molecules as well as with the outer mitochondrial membrane is proposed. The implications with respect to the physiological function of Mi-CK as an energy-channeling molecule at the producing side of the phosphoryl creatine shuttle are discussed.     

Mitochondrial creatine kinase from cardiac muscle and brain are two distinct isoenzymes but both form octameric molecules

Mitochondrial creatine kinase (Mi-CK) from chicken cardiac muscle and brain, recently shown to differ in their N-terminal amino acid sequences and to be encoded by multiple mRNAs (Hossle, H.P., Schlegel, J., Wegmann, G., Wyss, M., B?hlen, P., Eppenberger, H. M., Wallimann, T., and Perriard, J.C. (1988) Biochim. Biophys. Res. Commun. 151, 408-416) were separated on two-dimensional nonequilibrium pH-gradient electrophoresis gels and visualized as two distinct protein spots by immunoblotting. Analysis of the two proteins purified by specific elution from Blue-Sepharose with ADP (Wallimann, T., Zurbriggen, B., and Eppenberger, H. M. (1985) Enzyme 33, 226-231) followed by fast protein liquid chromatography cation exchange chromatography showed obvious differences in peptide maps, in immunological cross-reactivity with monoclonal antibodies, and in kinetic parameters. However, even though the two proteins were different, tissue-specific mitochondrial isoforms, both formed regularly-sized, perforated cube-like octameric structures with Mr of 364,000 +/- 25,000 and 352,000 +/- 20,000 for the cardiac and brain isoform, respectively. Electron microscopy of cardiac and brain Mi-CK octamers revealed cube-like molecules with a central cavity or transverse channel filled by negative stain. The octameric molecular structure of Mi-CK isoforms differs from the generally accepted dimeric arrangement of "cytosolic" muscle MM- and brain BB-CK.     

Mitochondrial creatine kinase from chicken brain. Purification, biophysical characterization, and generation of heterodimeric and heterooctameric molecules with subunits of other creatine kinase isoenzymes

In a recent study it has been shown that mitochondrial creatine kinase from chicken brain (Mia-CK) and heart (Mib-CK) are two distinct isoenzymes differing in ten out of the thirty N-terminal amino acids (Hossle, J.P., Schlegel, J., Wegmann, G., Wyss, M., B?hlen, P., Eppenberger, H.M., Wallimann, T., and Perriard J.C. (1988) Biochem. Biophys. Res. Commun. 151, 408-416). The present article describes the purification and biophysical characterization of the mitochondrial creatine kinase isoenzyme from chicken brain (Mia-CK). Gel permeation chromatography, direct mass measurements of individual molecules by scanning transmission electron microscopy, and analytical ultracentrifugation confirmed the existence of two different oligomeric forms, dimeric and octameric Mia-CK, with molecular masses of 85 kDa and 306-352 kDa and with sedimentation constants of 4.9-5.3 and 11.6-12.0 S, respectively. In addition, it was tested if Mia- and Mib-CK can form heterodimeric and heterooctameric molecules with subunits of other CK isoenzymes. By denaturation in urea or guanidine hydrochloride and subsequent renaturation, MiaMib-CK and surprisingly also MiaM-CK heterodimers could be generated. In contrast, no heterodimers were obtained between Mib- and M- or B-CK. Furthermore, reoctamerization of a mixture of Mia- and Mib-CK homodimers led to the formation of MiaMib-CK heterooctamers. In these heterooctamers, the Mia- and Mib-CK homodimers remained the fundamental building blocks. No subunit exchange between adjacent dimers within the heterooctamer could be observed even after storage for 3 months at 4 degrees C. The relevance of these data on the structural organization of the Mi-CK octamer and on the physiological aspects of tissue-specific isoenzyme expression are discussed.     

Native mitochondrial creatine kinase forms octameric structures. I. Isolation of two interconvertible mitochondrial creatine kinase forms, dimeric and octameric mitochondrial creatine kinase: characterization, localization, and structure-function relationships

The mitochondrial isoform of creatine kinase (Mi-CK, EC 2.7.3.2) purified to homogeneity from chicken cardiac muscle by the mild and efficient technique described in this article was greater than or equal to 99.5% pure and consisted of greater than or equal to 95% of a distinct, octameric Mi-CK protein species, with a Mr of 364,000 +/- 30,000 and an apparent subunit Mr of 42,000. The remaining 5% were dimeric Mi-CK with an apparent Mr of 86,000 +/- 8,000. Octamerization was not due to covalent linkages or intermolecular disulfide bonding. Upon dilution into buffers of low ionic strength and alkaline pH, octameric Mi-CK slowly dissociated in a time-dependent manner (weeks-months) into dimeric Mi-CK. However, the time scale of dimerization was reduced to minutes by the addition to diluted Mi-CK octamers of a mixture of Mg2+, ADP, creatine and nitrate known to induce a transition-state analogue complex (Milner-White, E.J., and Watts, D. C. (1971) Biochem. J. 122, 727-740). The conversion was fully reversible, and octamers were reformed by simple concentrations of Mi-CK dimer solutions to greater than or equal to 1 mg/ml at near neutral pH and physiological salt concentrations in the absence of adenine nucleotide. After separation of the two Mi-CK species by gel filtration, electron microscopic analysis revealed uniform square-shaped particles with a central negative-stain-filled cavity in the octamer fractions and "banana-shaped" structures in the dimer fractions. Mi-CK was localized inside the mitochondria by immunogold labeling with polyclonal antibodies. A dynamic model of the octamer-dimer equilibrium of Mi-CK and the preferential association of the octameric Mi-CK form with the inner mitochondrial membrane is discussed in the context of regulation of Mi-CK activity, mitochondrial respiration, and the CP shuttle.     

The structure of mitochondrial creatine kinase and its membrane binding properties

The biochemical and biophysical characterization of the mitochondrial creatine kinase (Mi-CK) from chicken cardiac muscle is reviewed with emphasis on the structure of the octameric oligomer by electron microscopy and on its membrane binding properties. Information about shape, molecular symmetry and dimensions of the Mi-CK octamer, as obtained by different sample preparation techniques in combination with image processing methods, are compared. The organization of the four dimeric subunits into the Mi-CK complex as apparent in the end-on projections is discussed and the consistently observed high binding affinity of the four-fold symmetric end-on faces towards many support films and towards each other is outlined. A study on the oligomeric state of the enzyme in solution and in intact mitochondria, using chemical crosslinking reagents, is presented together with the results of a search for a possible linkage of Mi-CK with the adenine nucleotide translocator (ANT). The nature of Mi-CK binding to model membranes, demonstrating that rather the octameric than the dimeric subspecies is involved in lipid interaction and membrane contact formation, is resumed and put into relation to our structural observations. The findings are discussed in light of a possiblein vivo function of the Mi-CK octamer bridging the gap between outer and inner mitochondrial membranes at the contact sites.     

Structure of the mitochondrial creatine kinase octamer: high-resolution shadowing and image averaging of single molecules and formation of linear filaments under specific staining conditions.

The combination of high-resolution tantalum/tungsten (Ta/W) shadowing at very low specimen temperature (-250 degrees C) under ultrahigh vacuum (less than 2 x 10(-9) mbar) with circular harmonic image averaging revealed details on the surface structure of mitochondrial creatine kinase (Mi-CK) molecules with a resolution less than 2.5 nm. Mi-CK octamers exhibit a cross-like surface depression dividing the square shaped projection of 10 x 10 nm into four equally sized subdomains, which correspond to the four dimers forming the octameric Mi-CK molecule. By a combination of positive staining (with uranyl acetate) and heavy metal shadowing, internal structures as well as the surface relief of Mi-CK were visualized at the same time at high resolution. Computational image analysis revealed only a single projection class of molecules, but the ability of Mi-CK to form linear filaments, as well as geometrical considerations concerning the formation of octamers by four equal, asymmetric dimers, suggest the existence of at least two distinct faces on the molecule. By image processing of Mi-CK filaments a side view of the octamer differing from the top-bottom projections of single molecules became evident showing a funnel-like access each form the top and bottom of the octamer connected by a central channel. The general structure of the Mi-CK octamer described here is relevant to the localization of the molecule at the inner-outer mitochondrial contact sites and to the function of Mi-CK as an "energy channeling" molecule.     

Mitochondrial creatine kinase mediates contact formation between mitochondrial membranes

Purified mitochondrial creatine kinase (Mi-CK) (EC 2.7.3.2) from chicken heart was shown to interact simultaneously with purified inner and outer mitochondrial membranes, thereby creating an intermembrane chondrial membranes, thereby creating an intermembrane were purified from rat liver and thus were fully devoid of Mi-CK. Intermembrane contact formation was demonstrated by measuring the binding of inner membrane vesicles to outer membranes spread at the air-water interface. Mi-CK also mediated intermembrane adhesion when membranes formed with total lipid extracts of both membranes were used, pointing to the role of lipids as potential membrane anchors of Mi-CK in the mitochondrial intermembrane space. Other enzymes of the intermembrane space that (like Mi-CK) are also cationic, as well as cytosolic isoenzymes of creatine kinase, failed to induce contact formation. Thus, of the proteins tested, membrane contact formation was specific for Mi-CK. The two oligomeric forms of Mi-CK (octamer and dimer) differed in their ability to mediate intermembrane adhesion, the octamer being more potent. Highly basic peptides, i.e. poly-L-lysines, were shown to strongly interact with membranes formed with lipid extracts of mitochondrial membranes: they both induced intermembrane binding and fusion. Interestingly, the extent of contact formation mediated by poly-L-lysines was lower than that of octameric Mi-CK. The implications of these findings on the function and localization of Mi-CK and on the structure of the mitochondrial intermembrane compartment are discussed.     

Interfacial behavior of cytoplasmic and mitochondrial creatine kinase oligomeric states

Adsorption to the air/water interface of isoenzymes of creatine kinase was investigated using surface pressure-area isotherms and Brewster angle microscopy (BAM) observations. Octameric mitochondrial creatine kinase (mtCK) exhibits a significant affinity for the air/water interface. Whatever the mode of formation of the interfacial film, i.e., injection of the protein in the subphase or spreading onto the buffer surface, the final arrangement and conformation adopted by mtCK molecules lead to a similar result. In contrast, the dimeric isoenzymes mtCK and cytosolic MMCK do not induce any surface pressure variation. However, when the subphase contains 0.3M NaCl, both isoenzymes adsorb to the interface. When treated with 0.8 or 3M GdnHCl, muscle creatine kinase (MMCK) becomes surface active and occupies a greater surface than mtCK. This result contrasts with previous observations, often derived from monomeric proteins, that their surface activity is increased upon unfolding. It underlines the possible influence exerted by the protein oligomeric state on its interfacial activity. At a subphase pH of 8.8, which corresponds to the pI of octameric mtCK, the profiles of the isotherms obtained with dimeric and octameric states and the resistance to compression of the protein monolayers are significantly affected when compared to those recorded at pH 7.4. These data suggest that the octamer is more hydrophobic than the dimer and may contribute to explaining why octamers bind to the inner mitochondrial membrane while dimers do not.     

Oligomeric state and membrane binding behaviour of creatine kinase isoenzymes: Implications for cellular function and mitochondrial structure

The membrane binding properties of cytosolic and mitochondrial creatine kinase isoenzymes are reviewed in this article. Differences between both dimeric and octameric mitochondrial creatine kinase (Mi-CK) attached to membranes and the unbound form are elaborated with respect to possible biological function. The formation of crystalline mitochondrial inclusions under pathological conditions and its possible origin in the membrane attachment capabilities of Mi-CK are discussed. Finally, the implications of these results on mitochondrial energy transduction and structure are presented.     

Novel staining pattern of skeletal muscle M-lines upon incubation with antibodies against MM-creatine kinase

Incubation of chicken skeletal muscle fibers with an excess of anti-M- creatine kinase (CK) immunoglobulin G and an excess of anti-M-CK Fab fragments leads to heavy decoration of the M-line (Wallimann, T., D.C. Turner, and H.M. Eppenberger, 1977, J. Cell Biol. 75:297-317) and to removal of the electron-dense M-line structure (Walliman, T., G. W. Pelloni, D.C. Turner, and H.M. Eppenberger, 1978, Proc. Natl. Acad. Sci. USA., 75:4296-4300), respectively. On the other hand, incubation with low concentrations of monovalent anti-M-CK Fab did not extract but rather decorated the M-line, giving rise to a distinct two-line staining pattern. A similar double-line staining pattern, although less pronounced, was also observed within the M-line of paraformaldehyde- prefixed myogenic cells, which after permeabilization were incubated with low concentrations of divalent anti-M-CK antibody. In both cases, the two decorated lines appearing in the middle of the A-band were spaced axially 42-44 nm apart and correspond most likely to the two M4 and M4'' m-bridge rows described by Sjostrom and Squire (1977, J. Mol. Biol., 109:49-68; 1977, J. Microscopy., 111:239-278). It is concluded that the muscle-specific form of creatine kinase, MM-CK, contributes mainly to the electron density of these M4 and M4'' m-bridges within the M-line structure. This specific labeling pattern is a further demonstration that CK is an integral part of the M-line.     

The structural features of beef heart mitochondrial creatine kinase.

Two forms of mitochondrial creatine kinase (Mi-CK) having Mr 320 kDa and 240 kDa as determined by gel-filtration on Sephacryl S-300 in 0.1 M Tris-HCl pH 7.4 were investigated. The sedimentation coefficient values for these two forms were found to be identical and equal to 12.3 S. When studied by electron microscopy the main type of images for the 320 kDa and 240 kDa Mi-CK appeared as annular particles, 12-14 nm in diameter, with a well-detected subunit structure and a central hollow, 3-4nm in diameter filled with the dye. The results of the averaging of the main type of individual Mi-CK images and particles of the two-dimensional crystal layer point to the overall geometry of the Mi-CK molecule structure as containing eight subunits arranged by a 4-fold symmetry around the central hollow. It may be that the eight identical subunits of crystalline Mi-CK are arranged with a P422 symmetry. However in both cases the averaged main images do not show a mirror symmetry. The multiplicity of the observed projections close to annular one provides additional evidence in favour of the great lability and structural mobility of the Mi-CK subunits. It allows to assume that two forms (320 kDa and 240 kDa) are not the different oligomers but they are two functionally distinct conformational states of octameric molecule of Mi-CK.     

Over-expression, purification and characterization of the oligomerization dynamics of an invertebrate mitochondrial creatine kinase

Mitochondrial creatine kinase (MtCK) plays a central role in energy homeostasis within cells that display high and variable rates of ATP turnover. Vertebrate MtCKs exist primarily as octamers but readily dissociate into constituent dimers under a variety of circumstances. MtCK is an ancient protein that is also found in invertebrates including sponges, the most primitive of all multi-cellular animals. We have cloned, expressed, and purified one of these invertebrate MtCKs from a marine polychaete worm, Chaetopterus variopedatus (CVMtCK). Size exclusion chromatography and dynamic light scattering (DLS) were used to characterize oligomeric state in comparison with that of octameric chicken sarcomeric isoform (SarMtCK). At protein concentrations >1 mg/ml, CVMtCK was predominantly octameric (>90%). When diluted to 0.1 mg/ml, CVMtCK dissociated into dimers much more rapidly than SarMtCK when observed under identical conditions. The rate of dissociation for both MtCKs increased as temperature rose from 2 to 28 degrees C, and in CVMtCK, fell at higher incubation temperatures. The fraction of octameric CVMtCK at equilibrium increased with temperature and then fell. Temperature transition studies showed that octamers and dimers were rapidly interconvertible on a similar time scale. Importantly, when CVMtCK was converted to the transition state analog complex (TSAC), both size exclusion chromatography and DLS showed that there was minimal dissociation of octamers into dimers while SarMtCK octamers were highly unstable as the TSAC. These results clearly show distinct differences in octamer stability between CVMtCK and SarMtCK, which could impact function under physiological circumstances. Furthermore, the large yield of recombinant protein and high stability of CVMtCK in the TSAC suggest that this protein might be a good target for crystallization efforts.     

Mitochondrial creatine kinase with atypical pI values detected in serum of a patient with ovarian hepatoid yolk sac tumor

Atypical mitochondrial creatine kinase (creatine N-phosphotransferase, CK, EC 2.7.3.2) was detected in the serum of a patient with carcinoma of germ cell origin, probably hepatoid yolk sac tumor. The pI of the oligomeric atypical mitochondrial CK (Mi-CK) was found at the acidic side compared to that of the typical ubiquitous Mi-CK (uMi-CK), while the molecular size of the atypical Mi-CK was similar to that of the typical uMi-CK. The pIs of the oligomeric and the dimeric atypical Mi-CKs became the same as those of the typical uMi-CK upon treatment with 2-mercaptoethanol. Therefore, the atypical Mi-CK was suggested to be an oxidized form of uMi-CK, and the oxidation might have occurred in the mitochondria because the oligomeric atypical Mi-CK had atypical pIs. The physicochemical characteristics of the oxidized uMi-CK were similar to those of the typical uMi-CK.     

Determination of isoelectric points of oligomeric forms of mitochondrial creatine kinase

Using isoelectrofocusing in three pH gradients differing in the initial pH value of the ampholyte gel mixture and in gradient pH range, the isoelectric points for the dimeric and octameric forms of mitochondrial creatine kinase from bovine heart and pigeon breast muscle were determined. The isoelectric points for the dimer and octamer are equal to 9.67 +/- 0.01 and 8.93 +/- 0.05 for the heart enzyme and to 9.56 +/- 0.08 and 8.91 +/- 0.23 for the skeletal muscle enzyme. The correctness of identification of the oligomeric forms of mitochondrial creatine kinase was confirmed by ultracentrifugation in a sucrose density linear gradient. Since creatine kinase is known to bind to mitochondrial membrane cardiolipin by electrostatic forces, it can be assumed that both oligomeric forms of the enzymes can bind to the membranes. However, the properties of the creatine kinase dimer suggest its greater ability to bind to mitochondrial membranes.     

The role of an absolutely conserved tryptophan residue in octamer formation and stability in mitochondrial creatine kinases

In most organisms, mitochondrial creatine kinase (MtCK) is present as dimers and octamers with the latter predominating under physiological conditions. An absolutely conserved tryptophan residue (Trp-264 in chicken sarcomeric MtCK) appears to play a key role in octamer stability. Recently, it has been shown that the sponge Tethya aurantia, a member of the most ancient group of living multi-cellular animals, expresses an obligate, dimeric MtCK that lacks this absolutely conserved tryptophan residue, instead possessing a tyrosine in this position. In the present study we confirm that the absolutely conserved tryptophan residue is lacking in other sponge MtCKs where it is instead substituted by histidine or asparagine. Site directed mutations of the Trp-264 in expression constructs of chicken sarcomeric MtCK and the octameric MtCK from the marine worm Chaetopterus destabilized the octameric quaternary structure producing only dimers. A Tyr-->Trp mutation in an expression construct of the Tethya MtCK construct failed to produce octamerization; Tyr-->His and Tyr-->Asn mutations also yielded dimers. These results, in conjunction with analysis of homology models of Chaetopterus and Tethya MtCKs, strongly support the view that while the absolutely conserved tryptophan residue is important in octamer stability, octamer formation involves a complex suite of interactions between a variety of residues.     

Influence of glycosylation on the oligomeric structure of yeast acid phosphatase

Secreted yeast acid phosphatase is found to be an octamer under physiological conditions rather than a dimer, as previously believed. The octameric form of the enzyme dissociates rapidly into dimers at pH below 3 and above 5, or by treatment with guanidine hydrochloride or urea, without further dissociation of dimers. Crosslinking experiments revealed that the dissociation of the octamer occurs through very unstable hexamers and tetramers, showing that the octamer is built of dimeric units. Dissociation to dimer was in all cases accompanied with a loss of most of the enzyme activity. The underglycosylated acid phosphatase, with less than eight carbohydrate chains per subunit, secreted from cells treated with moderate tunicamycin concentrations, contained besides octamers a high proportion of the dimers. With decreasing levels of enzyme glycosylation, the proportion of dimers increases and the amount of octamers correspondingly decreases. Furthermore, underglycosylated octamers were found to be significantly less stable than the fully glycosylated ones. This showed that carbohydrate chains play a significant role in the octamer formation in vivo, and in stabilization of the enzyme octameric form.     

Localization of Reactive Cysteine Residues by Maleidoyl Undecagold in the Mitochondrial Creatine Kinase Octamer

Octamers of mitochondrial creatine kinase (Mi-CK) wore modified with the thiol-specific reagents N-ethylmaleimide or the gold-coupled derivative, maleidoyl undecagold. The kinetics of inhibition of the Mi-CK catalysis was shown to be comparable for both reagents, suggesting that the large gold cluster complex is accessible to the reactive cysteines. SDS-PAGE analysis revealed that two of eight cysteines per Mi-CK monomer were labeled with maleidoyl undecagold with a similar affinity for the functional maleimide group. Gel exclusion chromatography of labeled molecules showed that the octameric structure of Mi-CK was preserved after thiol modification. Freeze-dried gold-labeled octamers visualized by electron microscopy under cryoconditions were enhanced in contrast and showed a well-preserved fourfold symmetry of the end-on view, Image analysis of gold-labeled Mi-CK exhibited an averaged end-on view with four strong contrast signals located at the periphery of the notamer, whereas the center of the molecule remained electron translucent. We conclude that the two cysteine residues per monomer labeled with maleidoyl undecagold are located at the octamer's perimeter and we discuss the possible role of these reactive cysteines in enzyme catalysis.     

Interaction of mitochondrial creatine kinase with model membranes. A monolayer study

The interaction of mitochondrial creatine kinase (Mi-CK; EC 2.7.3.2) with phospholipid monolayers and spread mitochondrial membranes at the air/water interface has been investigated. It appeared that Mi-CK penetrated into these monolayers as evidenced by an increase in surface pressure upon incorporation of Mi-CK. The increase in surface pressure was dependent on (1) the amount and (2) the oligomeric form of Mi-CK in the subphase, as well as on (3) the initial surface pressure and (4) the phospholipid composition of the monolayer. In this experimental system Mi-CK was able to interact equally well with both inner and outer mitochondrial membranes.     

Processing of Mgm1 by the rhomboid-type protease Pcp1 is required for maintenance of mitochondrial morphology and of mitochondrial DNA

The structure of mitochondria is highly dynamic and depends on the balance of fusion and fission processes. Deletion of the mitochondrial dynamin-like protein Mgm1 in yeast leads to extensive fragmentation of mitochondria and loss of mitochondrial DNA. Mgm1 and its human ortholog OPA1, associated with optic atrophy type I in humans, were proposed to be involved in fission or fusion of mitochondria or, alternatively, in remodeling of the mitochondrial inner membrane and cristae formation (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; Wong, E. D., Wagner, J. A., Scott, S. V., Okreglak, V., Holewinske, T. J., Cassidy-Stone, A., and Nunnari, J. (2003) J. Cell Biol. 160, 303-311; Sesaki, H., Southard, S. M., Yaffe, M. P., and Jensen, R. E. (2003) Mol. Biol. Cell, in press). Mgm1 and its orthologs exist in two forms of different lengths. To obtain new insights into their biogenesis and function, we have characterized these isoforms. The large isoform (l-Mgm1) contains an N-terminal putative transmembrane segment that is absent in the short isoform (s-Mgm1). The large isoform is an integral inner membrane protein facing the intermembrane space. Furthermore, the conversion of l-Mgm1 into s-Mgm1 was found to be dependent on Pcp1 (Mdm37/YGR101w) a recently identified component essential for wild type mitochondrial morphology. Pcp1 is a homolog of Rhomboid, a serine protease known to be involved in intercellular signaling in Drosophila melanogaster, suggesting a function of Pcp1 in the proteolytic maturation process of Mgm1. Expression of s-Mgm1 can partially complement the Deltapcp1 phenotype. Expression of both isoforms but not of either isoform alone was able to partially complement the Deltamgm1 phenotype. Therefore, processing of l-Mgm1 by Pcp1 and the presence of both isoforms of Mgm1 appear crucial for wild type mitochondrial morphology and maintenance of mitochondrial DNA.