Stable and labile products of mitochondrial protein synthesis in vitro

Mitochondrial Nascent Chain Quality Control Determines Organelle Form and Function

Proteotoxicity has long been considered a key factor in mitochondrial dysfunction and human disease. The origin of the endogenous offending toxic substrates and the regulatory pathways to deal with these insults, however, have remained unclear. Mitochondria maintain a compartmentalized gene expression system that in animals is only responsible for synthesis of 1% of the organelle proteome. Because of the relatively small contribution of the mitochondrial genome to the overall proteome, the synthesis and quality control of these nascent chains to maintain organelle proteostasis has long been overlooked. However, recent research has uncovered mechanisms by which defects to the quality control of mitochondrial gene expression are linked to a novel cellular stress response that impinges upon organelle form and function and cell fitness. In this review, we discuss the mechanisms for a key event in the response: activation of the metalloprotease OMA1. This severs the membrane tether of the dynamin-related GTPase OPA1, which is a critical determinant for mitochondrial morphology and function. We also highlight the evolutionary conservation from bacteria of these quality-control mechanisms to maintain membrane integrity, gene expression, and cell fitness.

The role of AAA+ proteases in mitochondrial protein biogenesis, homeostasis and activity control

Mitochondria are specialised organelles that are structurally and functionally integrated into cells in the vast majority of eukaryotes. They are the site of numerous enzymatic reactions, some of which are essential for life. The double lipid membrane of the mitochondrion, that spatially defines the organelle and is necessary for some functions, also creates a physical but semi-permeable barrier to the rest of the cell. Thus to ensure the biogenesis, regulation and maintenance of a functional population of proteins, an autonomous protein handling network within mitochondria is required. This includes resident mitochondrial protein translocation machinery, processing peptidases, molecular chaperones and proteases. This review highlights the contribution of proteases of the AAA+ superfamily to protein quality and activity control within the mitochondrion. Here they are responsible for the degradation of unfolded, unassembled and oxidatively damaged proteins as well as the activity control of some enzymes. Since most knowledge about these proteases has been gained from studies in the eukaryotic microorganism Saccharomyces cerevisiae, much of the discussion here centres on their role in this organism. However, reference is made to mitochondrial AAA+ proteases in other organisms, particularly in cases where they play a unique role such as the mitochondrial unfolded protein response. As these proteases influence mitochondrial function in both health and disease in humans, an understanding of their regulation and diverse activities is necessary.

Principles of control over formation of structures responsible for respiratory functions of mitochondria

Topogenesis of mitochondrial proteins includes their synthesis in cytosol and mitochondria, their translocation across the outer and inner membranes, sorting to various mitochondrial compartments, and assembly of different protein complexes. These complexes are involved in transport functions, electron transfer through the respiratory chain, generation of transmembrane electrochemical potential, oxidative phosphorylation of ADP into ATP, etc. To perform these functions, a special stringent control is required over formation of submitochondrial structures and the mitochondrion as a whole. Such control is expected to rigorously eliminate not only misfolded proteins but also incorrectly incorporated subunits and is realized in mitochondria by means of numerous proteases with different functions and localizations. In the case of more complicated protein formations, e.g. supercomplexes, the protein quality is assessed by their ability to realize the integral function of the respiratory chain and, thus, ensure the stability of the whole system. Considering supercomplexes of the mitochondrial respiratory chain, the present review clearly demonstrates that this control is realized by means of various (mainly vacuolar) proteases with different functions and localizations. The contemporary experimental data also confirm the author's original idea that the general mechanism of assembly of subcellular structures is based on the "selection by performance criterion" and "stabilization by functioning".

The formation of respiratory chain complexes in mitochondria is under the proteolytic control of the m-AAA protease

Yta10p (Afg3p) and Yta12p (Rcal1p), members of the conserved AAA family of ATPases, are subunits of the mitochondrial m-AAA protease, an inner membrane ATP-dependent metallopeptidase. Deletion of YTA10 or YTA12 impairs degradation of non-assembled inner membrane proteins and assembly of respiratory chain complexes. Mutations of the proteolytic sites in either YTA10 or YTA12 have been shown to inhibit proteolysis of membrane-integrated polypeptides but not the respiratory competence of the cells, suggesting additional activities of Yta10p and Yta12p. Here we demonstrate essential proteolytic functions of the m-AAA protease in the biogenesis of the respiratory chain. Cells harbouring proteolytically inactive forms of both Yta10p and Yta12p are respiratory deficient and exhibit a pleiotropic phenotype similar to Deltayta10 and Deltayta12 cells. They show deficiencies in expression of the intron-containing mitochondrial genes COX1 and COB. Splicing of COX1 and COB transcripts is impaired in mitochondria lacking m-AAA protease, whilst transcription and translation can proceed in the absence of Yta10p or Yta12p. The function of the m-AAA protease appears to be confined to introns encoding mRNA maturases. Our results reveal an overlapping substrate specificity of the subunits of the m-AAA protease and explain the impaired assembly of respiratory chain complexes by defects in expression of intron-containing genes in mitochondria lacking m-AAA protease.

Regulated protein degradation in mitochondria

Various adenosine triphosphate (ATP)-dependent proteases were identified within mitochondria which mediate selective mitochondrial protein degradation and fulfill crucial functions in mitochondrial biogenesis. The matrix-localized PIM1 protease, a homologue of the Escherichia coli Lon protease, is required for respiration and maintenance of mitochondrial genome integrity. Degradation of non-native polypeptides by PIM1 protease depends on the chaperone activity of the mitochondrial Hsp70 system, posing intriguing questions about the relation between the proteolytic system and the folding machinery in mitochondria. The mitochondrial inner membrane harbors two ATP-dependent metallopeptidases, the m- and the i-AAA protease, which expose their catalytic sites to opposite membrane surfaces and cooperate in the degradation of inner membrane proteins. In addition to its proteolytic activity, the m-AAA protease has chaperone-like activity during the assembly of respiratory and ATP-synthase complexes. It constitutes a quality control system in the inner membrane for membrane-embedded protein complexes.

Nucleo-mitochondrial interactions in mitochondrial gene expression

All proteins encoded by mitochondrial DNA (mtDNA) are dependent on proteins encoded by nuclear genes for their synthesis and function. Recent developments in the identification of these genes and the elucidation of the roles their products play at various stages of mitochondrial gene expression are covered in this review, which focuses mainly on work with the yeast Saccharomyces cerevisiae. The high degree of evolutionary conservation of many cellular processes between this yeast and higher eukaryotes, the ease with which mitochondrial biogenesis can be manipulated both genetically and physiologically, and the fact that it will be the first organism for which a complete genomic sequence will be available within the next 2 to 3 years makes it the organism of choice for drawing up an inventory of all nuclear genes involved in mitochondrial biogenesis and for the identification of their counterparts in other organisms.

Yta10p is required for the ATP-dependent degradation of polypeptides in the inner membrane of mitochondria

Incompletely synthesized polypeptides in the mitochondrial inner membrane are subject to rapid proteolysis. We demonstrate that Yta10p, a mitochondrial homologue of a conserved family of putative ATPases in Saccharomyces cerevisiae, is essential for this proteolytic process. Yta10p-dependent degradation requires divalent metal ions and the hydrolysis of ATP. Yta10p is an integral protein of the inner mitochondrial membrane exposing the carboxy terminus to the mitochondrial matrix space. Based on the presence of consensus binding sites for ATP, and for divalent metal ions found in a number of metal dependent endopeptidases, a direct role of Yta10p in the proteolytic breakdown of membrane-associated polypeptides in mitochondria is suggested.

Mitochondrial gene expression in Saccharomyces cerevisiae. Proteolysis of nascent chains in isolated yeast mitochondria optimized for protein synthesis

We demonstrate here that mitochondrial translation products synthesized by isolated yeast mitochondria are subject to rapid proteolysis. The loss of label from mitochondrial peptides synthesized in vitro comes from two distinct pools of peptides: one that is rapidly degraded (t1/2 of minutes) and one that is much more resistant to proteolysis (t1/2 of hours). As the length of the incubation period increases, the percentage of labelled peptides in the rapidly-turning-over pool decreases and cannot be detected after 60 min of incubation. This proteolysis is inhibited by chloramphenicol and is dependent on the presence of ATP. The loss of label during the chase occurs from fully completed translation products. The proteolysis observed here markedly affects measurements of rates of mitochondrial protein synthesis in isolated yeast mitochondria. In earlier work, in which proteolysis was not considered, mitochondrial translation was thought to stop after 20-30 min of incubation. In the present study, by taking proteolysis into account, we demonstrate that the rate of translation in isolated mitochondria is actually constant for nearly 60 min and then decreases to near zero by 80 min of incorporation. These findings have allowed us to devise a procedure for measuring the 'true' rate of translation in isolated mitochondria. In addition, they suggest that mitochondrial translation products which normally assemble with nuclear-encoded gene products into multimeric enzyme complexes are unstable without their nuclear-encoded counterparts.

An alkaline thiol proteinase in the liver mitochondria of bullfrog, Rana catesbeiana

The mitoplasts were prepared from bullfrog (Rana catesbeiana) liver mitochondria by treatment with digitonin and were then separated into the matrix and inner membrane fractions. The matrix fraction thus obtained was free of lysosomal contaminations and exhibited a distinct proteinase activity. pH dependency of the matrix proteinase activity measured in the presence and absence of iodoacetamide revealed that the matrix contained at least two kinds of proteinase, a major alkaline thiol proteinase having an optimal pH at 8.5 and a minor neutral proteinase having an optimal pH at 7.5. The major matrix proteinase activity was strongly inhibited by leupeptin, chymostatin, antipain and E64-C, an inhibitor of Ca2+-dependent thiol proteinase, while it was scarcely affected by diethylpyrocarbonate. The activity was also inhibited by DTNB and p-chloromercuribenzoate. Addition of hydrocarbon compounds such as ethylene glycol, glycerol, Triton X-100 and poly (ethylene glycol) to the reaction mixture was found to decrease the matrix proteinase activity. Neither cytochrome c nor glutamate dehydrogenase was hydrolyzed when subjected to the matrix proteinase activity in vitro. On the other hand, cytochrome c oxidase was effectively hydrolyzed, and the enzyme associated with the mitochondrial innermembrane fragments was partially hydrolyzed by the major matrix proteinase activity.

Proteolysis in mitochondrial preparations and in lysosomal preparations derived from rat liver

A method of preparing rat liver mitochondria with low residual contamination by lysosomal proteases is described. Preparations of mitochondria are divided into two equal portions, one of which is supplemented with a lysosomal fraction. The addition of the lysosomal fraction causes an increase in proteolysis of between 26- and 56-fold at pH 5.0 in four similar experiments. This increase matches the increase in the lysosomal marker beta-glucuronidase and indicates that all proteolysis at pH 5.0 is due to enzymes of the lysosomal fraction. Above pH 7.0, the addition of a lysosomal supplement increases proteolysis by 1.5- to 5-fold only, suggesting that in the absence of a lysosomal supplement very little of the observed proteolysis is due to enzymes of lysosomal origin. A method of calculating the contribution to total proteolysis of enzymes of the lysosomal fraction or of the mitochondrial fraction is described. The calculations show that at pH 7.0 and above, more than 93% of the observed proteolysis is due to enzymes originating in the mitochondrial fraction. The results support the view of other workers that rat liver mitochondria contain an endogenous neutral proteolytic system capable of degrading mitochondrial proteins to acid-soluble products.

Identification and partial purification of a heart mitochondrial membrane proteinase

Membrane-bound proteinase activity was demonstrated by a solid-phase assay system in both beef heart and rat liver mitochondria. The activity was sensitive to SH reagents and assorted proteinase inhibitors. Although stimulated by nonionic detergents, it became labile when solubilized by detergents. The proteinase activity from heart mitochondria copurified with the ADP:ATP translocator protein. Gel electrophoresis of this preparation revealed the translocator polypeptide as well as a number of minor components. In solubilized mitochondria the ADP:ATP translocator polypeptide slowly disappeared upon standing at 0 degrees C as revealed by polyacrylamide gel electrophoresis under denaturing conditions. The loss of this polypeptide was prevented by addition of proteinase inhibitors as well as the translocator affinity ligand, carboxyatractylate. These observations confirm the presence of an integral membrane proteinase in mitochondria and suggest a structural and enzymatic interaction between the proteinase and the ADP:ATP translocator.

Liver mitochondria contain an ATP-dependent, vanadate-sensitive pathway for the degradation of proteins

A large fraction (30-50%) of the various proteins synthesized within isolated rat liver mitochondria were degraded to amino acids within 60 min after synthesis. Incomplete mitochondrial polypeptides resulting from the incorporation of puromycin were degraded even more extensively (80% per hr). Protein breakdown was measured by the appearance of acid-soluble radioactivity and by the disappearance of labeled polypeptides detected on NaDodSO4/polyacrylamide gel electrophoresis. The amino acids generated by proteolysis were transported rapidly out of the mitochondria and no peptide intermediates accumulated in the organelle. This degradative process did not involve lysosomes or lysosomal enzymes and was markedly stimulated by ATP either generated within the mitochondria or supplied exogenously. An inhibitor of respiration (cyanide) or uncouplers of oxidative phosphorylation (oligomycin, dinitrophenol) reduced proteolysis when mitochondria were provided substrates for ATP generation. When exogenous ATP was provided, these agents did not affect proteolysis, but degradation was then sensitive to atractyloside, an inhibitor of adenine nucleotide transport. Vanadate, an inhibitor of various ATPases, blocked proteolysis even in the presence of ATP and caused a marked stabilization of nearly all polypeptide bands. Thus, mitochondria--like bacteria or the cytosol of animal cells--contain a pathway for complete degradation of proteins which seems to selectively remove polypeptides with abnormal structures. Within this organelle, ATP hydrolysis appears necessary for an initial step in this degradative process.

Regulation of mitochondrial biogenesis. Occurrence of non-functioning components of the mitochondrial respiratory chain in Saccharomyces cerevisiae grown in the presence of proteinase inhibitors: evidence for proteolytic control over assembly of the respiratory chain

Yeast was grown in glucose- or galactose-containing media without or with proteinase inhibitors, phenylmethanesulphonyl fluoride and pepstatin. Culture growth was practically not affected by these compounds. Yeast growth on glucose in the presence of either phenylmethanesulphonyl fluoride or pepstatin entails accumulation of cytochromes c, c1, b and aa3 to a 25--30% excess above the control by the stationary phase, while cell respiration is unaffected. During growth on galactose the maximal cytochrome content (per unit weight of biomass) is reached in the mid-exponential phase and then decreases by 30--40% towards the stationary phase, while cell respiration remains constant. Addition of phenylmethanesulphonyl fluoride or pepstatin in the mid-exponential phase blocks the decrease in cytochrome levels and has no effect on cell respiration. Mitochondrial populations isolated from stationary-phase control and phenylmethanesulphonyl fluoride-grown cells glucose cultures display identical succinate oxidase and partial-respiratory-chain activities, despite the differences in cytochrome contents. However, the activities of individual respiratory complexes measured after maximal activation are nearly proportional to the amounts of corresponding components. The same situation holds true for mitochondrial populations from mid-exponential-phase, stationary-phase control and stationary-phase inhibitor-grown cells of galactose cultures. The findings suggest that the 'surplus' respiratory-chain components do not participate in electron flow because of the lack of interaction with adjacent carriers.

The localization of an intracellular membrane-bound proteinase from rat liver

To localize the membrane-bound, histone-degrading proteinase, which was previously isolated from the mitochondrial fraction, nuclei, mitochondria, lysosomes, peroxisomes, smooth and rough endoplasmic reticulum, ribosomes and plasma membranes were prepared from a rat liver homogenate according to established methods and characterized by marker enzyme activities. The isolated subcellular fractions were treated with digitonin, and subjected to a discontinuous sucrose gradient centrifugation. The material, which sedimented through 1.74 M surcrose was analyzed in respect to the various marker enzymes and for proteolytic activity. Proteinase activity was found in the material obtained after digitonin treatment and step gradient centrifugation of mitochondria. This finding shows the occurrence of a proteinase in mitochondria; After fractionation of mitochondria into outer and inner membrane, intermembrane fraction and matrix, the proteinase could be localized exclusively in the inner mitochondrial membrane. A possible physiological function of the enzyme during the biosynthesis of inner membrane constituents is discussed.

The dicyclohexylcarbodiimide-binding protein of rat liver mitochondria as a product of the mitochondrial protein synthesis

A product of mitochondrial protein synthesis in rat liver mitochondria, characterized by a low molecular weight (Mr is less than 10000) and an unusually high hydrophobicity, has been identified as the dicyclohexylcarbodiimide-binding protein and as a peptide of the hydrophobic sector of the mitochondrial ATPase complex. The purified protein still possesses the ability of bind dicyclohexylcarbodiimide.

Integration and regulation of mitochondrial assembly in yeast

The interactions between the mitochondrial and nucleocytoplasmic systems required for mitochondriogenesis have been investigated at several different levels. Those involved in the formation of functional enzyme complexes have been studied using cytochrome oxidase: this multimeric (2 X 7 and 2 X 6 subunits for enzymes from yeast and beef heart respectively) has been resolved, and the mitochondrial contribution has been shown to be dispensible for catalytic function proper. Using novel mutants, with a mitochondrial mode of inheritance, a mitochondrial gene product localized in the oligomycin-sensitive ATPase has been implicated in the assembly not only of this complex, but of cytochrome oxidase as well. Interactions required for the genetic competence of the mitochondrial system have become apparent as a result of studies in the mechanism of action of the highly effective mitochondrial mutagen ethidium bromide. This agent first becomes covalently inserted into mitochondrial DNA and, after its excision, eventually results in extensive degradation of the macromolecule. The excision reaction has now been shown to be performed by a complex between the oligomycin-sensitive ATPase and a DNA-binding protein presumably involved in recognizing the damage. On the level of replication and expression of the mitochondrial genome studies using thermolabile mutants have demonstrated that these processes appear independent of the replication of nuclear DNA but not of its expression.

Analysis of proteins synthesized in mitochondria of cultured mammalian cells. An assessment of current approaches and problems in interpretation

1. The conditions which enable highly efficient utilization of [35S]methionine by cultured mammalian cells and the resolution of selectively labelled mitochondrial products are described. 2. Analysis of mitochondria purified from cells labelled in the presence or absence of inhibitors of cytoplasmic (or mitochondrial) protein synthesis indicated that about 5% of the [35S]methionine incorporated into mitochondrial proteins results from synthesis on mitoribosomes. 3. The electrophoretic profile of the detergent-solubilized proteins of mitochondrial isolated from cells which were labelled in the presence of 50 mug/ml emetine was similar to those obtained with extracts prepared by direct solbuilization of the intact cells after incorporation of label. 4. Pulse-labelling studies suggested that the components resolved by electrophoresis and autoradiography under the conditions described, apparently represent discrete and stable end products radiography under the conditions described, apparently represent discrete and stable end products of mitochondrial protein synthesis. No post-synthetic modification or degradation of these products was detected. 5. Erythromycin was found to suppress the synthesis of additional labelled products which were detected in extracts of one cell line, when analysed by procedures which normally detected only mitochondrially synthesized proteins. These additional bands were attributed to the synthetic activity of Mycoplasma.

Formation, size, and solubility in chloroform/methanol of products of protein synthesis in isolated mitochondria of rat liver and Zajdela hepatoma

The number, size, solubility in chloroform/methanol and some aspects of the formation of the components labeled by radioactive amino acids in isolated mitochondria of rat liver and Zajdela hepatoma were studied. Isolated mitochondria were labeled with radioactive amino acids under various conditions, and the distribution of radioactivity in sodium dodecylsulfate-polyacrylamide gels after electrophoresis of mitochondrial membrane fraction was analysed. 1. Isolated mitochondria of rat liver and Zajdela hepatoma incroporated radioactive amino acids almost exclusively into the membrane fraction. Electrophoretic analysis of this fraction revealed the presence of 15 distinct peaks of radioactivity with corresponding apparent molecular weights of 10 000 to 58 000. The electrophoretic mobility of the labeled components was identical and the general pattern of the radioactivity distribution in the gel for the rat liver and the tumour mitochondria was very similar. 2. Components of the membrane fraction of rat liver mitochondria labeled in vitro displayed an unequal solubility in acidic (2 mM HC1) chloroform/methanol (2/1) mixture; as detected by sodium dodecylsulfate-polyacrylamide gel electrophoresis a single labeled component with apparent molecular weight of 10 000 was soluble in neutral chloroform/methanol. 3. Inverse relation was observed between amino acid incorporation activity of isolated mitochondria and the portion of the label incorporated into the component with apparent molecular weight 10 000. The identity of this component with that soluble in neutral chloroform/methanol mixture has been indicated. 4. The rate of incorporation of [3H]leucine by isolated Zajdela hepatoma mitochondria into the components with lower (10 000-25 000) apparent molecular weights decreased with time, whereas that into components with higher (above 25 000) apparent molecular weight remained approximately constant within the time interval tested (30 min). 5. From the total radioactivity incorporated into the membrane fraction during 5-min pulse labeling of isolated Zajdela hepatoma mitochondria by [3H]leucine up to 25% was recovered in the region of the gel corresponding to a component with apparent molecular weight 10 000. After 25 min chase the radioactivity in this region decreased about 3.5 times while the specific radioactivity of the total membrane fraction did not change significantly. The pattern of radioactivity distribution observed after the pulse was preserved by chloramphenicol. 6. Unlabeled sonicated mitochondria or postribosomal supernatant from rat liver regenerating in the presence of chloramphenicol were incubated with neutral chloroform/methanol extract of in vitro with [14C]leucine labeled rat liver mitochondria. After this incubation several labeled components with apparent molecular weights above 10 000 were recovered in the electrophoreograms of the originally unlabeled fractions.

Biogenesis of mitochondrial membranes in Neurospora crassa. Mitochondrial protein synthesis during conidial germination

The conidia of Neurospora crassa entered logarithmic growth after a 1-h lag period at 30 degrees C. Although [14C]leucine is incorporated quickly early in growth, cellular protein data indicated that no net protein synthesis occurred until after 2 h of growth. Neurospora is known to produce ethanol during germination even though respiratory enzymes are present. Also, Neurospora mitochondria isolated from cells less than 3-h old are uncoupled. Since oxygen uptake increased during germination, was largely cyanide-sensitive, and reached a maximum at 3 h, it is hypothesized that during early germination the uncoupled electron transport chain merely functions to dispose of reducing equivalents generated by substrate level ATP production. The rate of protein synthesis in vitro by mitochondria isolated from 0-8-h-old cells increased as did cell age. Mitochondrial protein synthesis in vivo, assayed in the presence of 100 mug cycloheximide/ml, increased from low levels in the cinidia to peak levels at 3-4 h of age and then slowly decreased. The rate of mitochondrial protein synthesis in vivo was linear for at least 90 min in 0-4-h-old cells, but declined after 15 min of incorporation in 6 and 8-h-old cells. The products of mitochondrial protein synthesis in vivo were analyzed with dodecylsulfate gel electrophoresis and autoradiography. Early in germination 80% of the synthesis was of two small proteins (molecular weights 7200 and 9000). At 8 h 85% of the radioactivity was in 10 larger proteins (12 200 to 80 000). Within the high-molecular-weight class, proteins of between 12 000 and 21 500 molecular weight were preferentially lavelled early in germination, whereas after 8 h of growth proteins of 27 500 to 80 000 molecular weight were preferentially labelled. It is hypothesized that the 7200 and 9000-molecular-weight products of mitochondrial protein synthesis combine with other proteins to form the larger proteins found later in growth. The availability of these other proteins in cells of different ages could affect the rate of mitochondrial protein synthesis in vivo.

Lipophilic proteins encoded by mitochondrial and nuclear genes in Neurospora crassa

Mitochondrial proteins soluble in neutral chloroform-methanol (2:1) were separated from lipids by ether precipitation and resolved by Sephadex G-200 filtration in the presence of dodecylsulfate into two major fractions eluting in the excluded region (peak I) and in a region of an apparent molecular weight 8000 (peak II). Residual phospholipids are found only in peak II. Peak I consists of several aggregated small polypeptides of molecular weights around 8000, which can be disaggregated by mild oxidation with performic acid. Cycloheximide stimulates almost two-fold incorporation of radioactive phenylalanine into peak I proteins but inhibits labelling of peak II proteins by 95%. Chloramphenicol and ethidium bromide inhibit the synthesis of peak I proteins by 70% and 95% respectively, but do not affect labelling of peak II proteins. At least 30% of the translation products of mitochondrial DNA in vitro behave like peak I proteins: they are soluble in organic solvents, they aggregate in dodecylsulfate buffer after removal of phospholipids and they contain species of molecular weights around 8000 that disaggregate upon oxidation. The data strongly suggest that the proteins of peak I are encoded by mitochondrial genes and synthesized on mitochondrial ribosomes, whereas the proteins of peak II are encoded by nuclear genes and synthesized on cytoplasmic ribosomes. Both groups of lipophilic proteins are very similar in their molecular weights, but the mitochondrially coded peak I proteins have the unique property of forming large heat-stable aggregates in the presence of dodecylsulfate.