The primary structure of cytochrome c1 from Neurospora crassa

Genome-wide CRISPRi screening identifies OCIAD1 as a prohibitin client and regulatory determinant of mitochondrial Complex III assembly in human cells

Dysfunction of the mitochondrial electron transport chain (mETC) is a major cause of human mitochondrial diseases. To identify determinants of mETC function, we screened a genome-wide human CRISPRi library under oxidative metabolic conditions with selective inhibition of mitochondrial Complex III and identified ovarian carcinoma immunoreactive antigen (OCIA) domain-containing protein 1 (OCIAD1) as a Complex III assembly factor. We find that OCIAD1 is an inner mitochondrial membrane protein that forms a complex with supramolecular prohibitin assemblies. Our data indicate that OCIAD1 is required for maintenance of normal steady-state levels of Complex III and the proteolytic processing of the catalytic subunit cytochrome c₁ (CYC1). In OCIAD1 depleted mitochondria, unprocessed CYC1 is hemylated and incorporated into Complex III. We propose that OCIAD1 acts as an adaptor within prohibitin assemblies to stabilize and/or chaperone CYC1 and to facilitate its proteolytic processing by the IMMP2L protease.

Linear Discriminant Analysis Identifies Mitochondrially Localized Proteins in Neurospora crassa

Besides their role as powerhouses, mitochondria play a pivotal role in the spatial organization of numerous enzymatic functions. They are connected to the ER, and many pathways are organized through the mitochondrial membranes. Thus, the precise definition of mitochondrial proteomes remains a challenging task. Here, we have established a proteomic strategy to accurately determine the mitochondrial localization of proteins from the fungal model organism Neurospora crassa. This strategy relies on both highly pure mitochondria as well as the quantitative monitoring of mitochondrial components along their consecutive enrichment. Pure intact mitochondria were obtained by a multistep approach combining differential and density Percoll (ultra) centrifugations. When compared with three other intermediate enrichment stages, peptide sequencing and quantitative profiling of pure mitochondrial fractions revealed prototypic regulatory profiles of per se mitochondrial components. These regulatory profiles constitute a distinct cluster defining the mitochondrial compartment and support linear discriminant analyses, which rationalized the annotation process. In total, this approach experimentally validated the mitochondrial localization of 512 proteins including 57 proteins that had not been reported for N. crassa before.

Structure, organization and expression of the genes encoding mitochondrial cytochrome c(1) and the Rieske iron-sulfur protein in Chlamydomonas reinhardtii

The sequence and organization of the Chlamydomonas reinhardtii genes encoding cytochrome c(1) ( Cyc1) and the Rieske-type iron-sulfur protein ( Isp), two key nucleus-encoded subunits of the mitochondrial cytochrome bc(1) complex, are presented. Southern hybridization analysis indicates that both Cyc1 and Isp are present as single-copy genes in C. reinhardtii. The Cyc1 gene spans 6404 bp and contains six introns, ranging from 178 to 1134 bp in size. The Isp gene spans 1238 bp and contains four smaller introns, ranging in length from 83 to 167 bp. In both genes, the intron/exon junctions follow the GT/AG rule. Internal conserved sequences were identified in only some of the introns in the Cyc1 gene. The levels of expression of Isp and Cyc1 genes are comparable in wild-type C. reinhardtii cells and in a mutant strain carrying a deletion in the mitochondrial gene for cytochrome b (dum-1). Nevertheless, no accumulation of the nucleus-encoded cytochrome c(1) or of core proteins I and II was observed in the membranes of the respiratory mutant. These data show that, in the green alga C. reinhardtii, the subunits of the cytochrome bc(1) complex fail to assemble properly in the absence of cytochrome b.

Interactions between the cytochrome b, cytochrome c1, and Fe-S protein subunits at the ubihydroquinone oxidation site of the bc1 complex of Rhodobacter capsulatus

Ubihydroquinone:cytochrome (cyt) c oxidoreductase (bc1 complex and its plant counterpart b6f complex) is a vital component of energy-transducing systems in most organisms from bacteria to eukaryotes. In the facultative phototrophic (Ps) bacterium Rhodobacter capsulatus, it is constituted by the cyt b, cyt c1, and Rieske Fe-S protein subunits and is essential for Ps growth. Of these subunits, cyt b has two nontransmembrane helices, cd1 and cd2, which are critical for its structure and function. In particular, substitution of threonine (T) at position 163 on cd1 with phenylalanine (F) or proline (P) leads to the absence of the bc1 complex. Here, Ps+ revertants of B:T163F were obtained, and their detailed characterizations indicated that position 163 is important for the assembly of the bc1 complex by mediating subunit interactions at the Qo site. The loss of the hydroxyl group at position 163 of cyt b was compensated for by the gain of either a hydroxyl group at position 182 of cyt b or 46 of the Fe-S protein or a sulfhydryl group at position 46 of cyt c1. Examination of the mitochondrial bc1 complex crystal structure [Zhang, Z., Huang, L., Shulmeister, V. M., Chi, Y.-I., Kim, K. K., Hung, L.-W., Crofts, A. R., Berry, E. A., and Kim, S.-H. (1998) Nature 392, 677-684] revealed that the counterparts of B:G182 (i.e., G167) and F:A46 (i.e. , A70) are located close to B:T163 (i.e., T148), whereas the C:R46 (i.e., R28) is remarkably far from it. The revertants contained substoichiometric amounts of the Fe-S protein subunit and exhibited steady-state and single-turnover, electron transfer activities lower than that of a wild-type bc1 complex. Interestingly, their membrane supernatants contained a smaller form of this subunit with physicochemical properties identical to those of its membrane-bound form. Determination of the amino-terminal amino acid sequence of this soluble Fe-S protein revealed that it was derived from the wild-type protein by proteolytic cleavage at V44. This work revealed for the first time that position 163 of cyt b is important both for proper subunit interactions at the Qo site and for inactivation of the bc1 complex by proteolytic cleavage of its Fe-S protein subunit at a region apparently responsible for its mobility during Qo site catalysis.

Two distinct and independent mitochondrial targeting signals function in the sorting of an inner membrane protein, cytochrome c1

Proteins of the mitochondrial inner membrane display a wide variety of orientations, many spanning the membrane more than once. Some of these proteins are synthesized with NH2-terminal cleavable targeting sequences (presequences) whereas others are targeted to mitochondria via internal signals. Here we report that two distinct mitochondrial targeting signals can be present in precursors of inner membrane proteins, an NH2-terminal one and a second, internal one. Using cytochrome c1 as a model protein, we demonstrate that these two mitochondrial targeting signals operate independently of each other. The internal targeting signal, consisting of a transmembrane segment and a stretch of positively charged amino acid residues directly following it, initially directs the translocation of the preprotein into the intermembrane space. It then inserts into the inner membrane from the intermembrane space side in a delta psi-dependent manner and thereby determines the orientation the protein attains in the inner membrane. Analysis of a number of other presequence-containing protein of the inner membrane suggest that they too contain such internal targeting signals.

Sorting of cytochrome b2 to the intermembrane space of mitochondria. Kinetic analysis of intermediates demonstrates passage through the matrix

Precytochrome b2 is targeted to the mitochondrial intermembrane space by a dual targeting sequence comprising 80 amino acids. A kinetic analysis of intramitochondrial sorting was performed. The intermediate-size form accumulated transiently in the matrix. When import was performed in the presence of metal chelators to prevent the first processing by the matrix processing peptidase, > 40% of the imported precursor was localized in the matrix. A deletion of 13 amino acids in the intermembrane space sorting sequence caused partial inhibition of the first processing, and a transient accumulation of the precursor form in the matrix was also observed. The decrease in this matrix-localized precursor form paralleled an increase in the mature-size form in the intermembrane space. A point mutation in the mitochondrial targeting sequence (N-terminal to the sorting sequence) resulted in missorting to the matrix space. Furthermore, a chimeric protein consisting of the initial 85 residues of cytochrome b2 fused to dihydrofolate reductase was partially targeted to the matrix at 15 degrees C, but not at 25 degrees C. Together, the results presented here indicate that cytochrome b2 passes through the matrix on its sorting pathway to the intermembrane space.

Cytochromes c1 of kinetoplastid protozoa lack mitochondrial targeting presequences

We have used the polymerase chain reaction to amplify cDNA fragments that encode the amino-terminal sequences of cytochrome c1 from two distantly related kinetoplastid species, Crithidia fasciculata and Bodo caudatus. Cloning and sequencing of these fragments have revealed that these proteins lack conventional mitochondrial targeting presequences.

Assembly of the Rieske iron-sulfur subunit of the cytochrome bc1 complex in the Escherichia coli and Rhodobacter sphaeroides membranes independent of the cytochrome b and c1 subunits

The Rieske iron-sulfur subunit of the cytochrome bc1 complex from Rhodobacter sphaeroides has been expressed in Escherichia coli and also in a strain of Rb. sphaeroides lacking the other subunits of the bc1 complex. PCR products encoding the full-length subunit were introduced into expression vectors to produce the subunit alone or the subunit fused behind the mature portion of the E. coli maltose binding protein (MBP), but lacking the MBP signal sequence. These proteins are both located in the cytoplasmic membrane. The unfused Rieske subunit assembles a Rieske-like iron-sulfur cluster, but with EPR characteristics which differ from the normal rhombic signal observed in the cytochrome bc1 complex. The overproduced MBP fusion protein, on the other hand, does not contain an EPR-detectable iron-sulfur cluster. Subfragments of the Rieske subunit lacking the amino-terminal hydrophobic anchor also lack the iron-sulfur cluster were expressed in E. coli. When expressed in Rb. sphaeroides in the absence of the cytochrome b and c1 subunits, the fully metalated Rieske subunit with the diagnostic gy = 1.90 EPR signal is observed in the cytoplasmic membrane. The fact that the Rieske subunit has an assembled iron-sulfur cluster and is bound to either the E. coli or the Rb. sphaeroides membrane in the absence of the other subunits of the bc1 complex demonstrates a mode of membrane attachment independent of the other components of the complex. These data are consistent with models in which the Rieske subunit is bound to the membrane via a single membrane-spanning helix located near the amino terminus.(ABSTRACT TRUNCATED AT 250 WORDS)

The distribution of charged amino acids in mitochondrial inner-membrane proteins suggests different modes of membrane integration for nuclearly and mitochondrially encoded proteins

We have analyzed the amino acid distribution in seven nuclearly encoded and five mitochondrially encoded inner membrane proteins with experimentally well characterized topologies. The mitochondrially encoded proteins conform to the 'positive inside' rule, i.e. they have many more positively charged residues in their non-translocated as compared to translocated domains. However, most of the nuclearly encoded proteins do not show such a bias but instead have a surprisingly skewed distribution of Glu residues with an almost ten times higher frequency in the intermembrane space than in the matrix domains. These findings suggest that some, but possibly not all, nuclearly encoded inner membrane proteins may insert into the membrane by a mechanism that does not depend on the distribution of positively charged amino acids.

Cytochrome c1 from potato: a protein with a presequence for targeting to the mitochondrial intermembrane space

Here we report the primary structure of potato cytochrome c1, a nuclear-encoded subunit of complex III. Using heterologous antibodies directed against cytochrome c1 from yeast two types of clones were isolated from an expression library, suggesting that at least two different genes are present and expressed in the genome. Northern blot analysis reveals that slightly varying levels of cytochrome c1 transcripts are present in all potato tissues analysed. A 1304 bp insert of one of the cDNA clones (pC13II) encodes the entire 320 amino acids of the precursor protein corresponding to a molecular weight of 35.2 kDa. As revealed by direct amino acid sequence determination of the cytochrome c1 protein another cDNA clone (pC18I) encodes the major form of cytochrome c1 present in potato tuber mitochondria. Western blots of subfractionated potato mitochondria show that the mature protein present in the membrane fraction is smaller than the pC13II encoded protein synthesized in Escherichia coli. The transient presequence of the protein is 77 amino acids long and has a bipartite polarity profile characteristic of presequences involved in targeting to the intermembrane space of fungal mitochondria. It consists of a positively charged NH2-terminal part which resembles "matrix targeting domains" and an adjacent hydrophobic region showing sequence similarities to "intramitochondrial sorting domains". The amino-terminal region of potato cytochrome c1 is the first presequence of a plant protein of the mitochondrial intermembrane space to be determined and may be useful in the study of intramitochondrial sorting in plants.

Some recent advances relating to prokaryotic cytochrome c reductases and cytochrome c oxidases

Prokaryotic systems provide excellent experimental opportunities for exploring structure/function relationships for the complex, membrane-bound, multisubunit enzymes responsible for the reduction and subsequent oxidation of c-type cytochromes in respiratory or photosynthetic electron transport chains. Two points are made in this mini-review: (1) The eukaryotic and prokaryotic aa3-type cytochrome c oxidases are members of an apparently large superfamily of structurally related respiratory oxidases. This superfamily displays considerable variation in terms of the heme prosthetic groups (a or b) as well as the substrate oxidized (quinol or cytochrome c). The relationships among these enzymes help to facilitate explorations of how they work. (2) Molecular biology techniques can be used to generate intact, redox-active, water-soluble domains of membrane-bound subunits. These soluble domains can be used for detailed examination, including obtaining high resolution structure by NMR techniques or by X-ray crystallography. This approach is being used to study the soluble heme-binding domain of cytochrome c1 from the bc1 complex of Rhodobacter sphaeroides.

Interaction between cytochrome C and di- and tripeptides

By observing changes in the absorbance spectrum between 340 and 650 nm, we found that tyrosyltyrosylphenylalanine (TTP) interacts with cytochrome C (CC). TTP caused the characteristic changes of CC reduction, namely, an increased optical density at 524 and 550 nm and a hyperchromic shift at 416 nm. The apparent dissociation constant (Kd) was 2.9 x 10(-3) M. Most of the reducible CC at 20 uM concentration was reduced by 10 mM TTP. TTP was more potent than all other peptides tested, including the previously reported tyrosylphenylalanine. That the carboxyl terminal phenyl group was essential for reduction was shown by comparing variously substituted di- and tripeptides. Reduction by TTP increased with increasing pH and buffer concentration at constant pH. A combination of superoxide dismutase and catalase failed to inhibit the reduction. We found no effect of TTP on O2 consumption of isolated intact mitochondria. Our data demonstrate that small peptides can serve as probes of the topography and electron density of CC and that the TTP-CC interaction may provide insight as an analog of in-vivo processes.

Biogenesis of mitochondrial c-type cytochromes

Cytochromes c and c1 are essential components of the mitochondrial respiratory chain. In both cytochromes the heme group is covalently linked to the polypeptide chain via thioether bridges. The location of the two cytochromes is in the intermembrane space; cytochrome c is loosely attached to the surface of the inner mitochondrial membrane, whereas cytochrome c1 is firmly anchored to the inner membrane. Both cytochrome c and c1 are encoded by nuclear genes, translated on cytoplasmic ribosomes, and are transported into the mitochondria where they become covalently modified and assembled. Despite the many similarities, the import pathways of cytochrome c and c1 are drastically different. Cytochrome c1 is made as a precursor with a complex bipartite presequence. In a first step the precursor is directed across outer and inner membranes to the matrix compartment of the mitochondria where cleavage of the first part of the presequence takes place. In a following step the intermediate-size form is redirected across the inner membrane; heme addition then occurs on the surface of the inner membrane followed by the second processing reaction. The import pathway of cytochrome c is exceptional in practically all aspects, in comparison with the general import pathway into mitochondria. Cytochrome c is synthesized as apocytochrome c without any additional sequence. It is translocated selectively across the outer membrane. Addition of the heme group, catalyzed by cytochrome c heme lyase, is a requirement for transport. In summary, cytochrome c1 import appears to follow a "conservative pathway" reflecting features of cytochrome c1 sorting in prokaryotic cells. In contrast, cytochrome c has "invented" a rather unique pathway which is essentially "non-conservative."

The 49 K subunit of NADH: ubiquinone reductase (complex I) from Neurospora crassa mitochondria: primary structure of the gene and the protein

The primary structure of the 49 K subunit of the respiratory chain NADH:ubiquinone reductase (complex I) from Neurospora crassa was determined by sequencing cDNA, genomic DNA and the N-terminus of the mature protein. The sequence lengths correlate to a molecular mass of 54,002 daltons for the preprotein and 49,239 daltons for the mature protein. The presequence consists of 42 amino acids of typical composition for sequences which target nuclear-encoded proteins into mitochondria. The mature protein consists of 436 amino acids and shows 64% similarity to a 49 K subunit of bovine heart NADH:ubiquinone reductase and 33% to a predicted translation product of an open reading frame in the chloroplast DNAs of Marchantia polymorpha and Nicotiana tabacum. Evidence for an iron-sulfur cluster in the subunit is discussed.

Protein sorting to mitochondria: evolutionary conservations of folding and assembly

According to the endosymbiont hypothesis, mitochondria have lost the autonomy of their prokaryotic ancestors. They have to import most of their proteins from the cytosol because the mitochondrial genome codes for only a small percentage of the polypeptides that reside in the organelle. Recent findings show that the sorting of proteins into the mitochondrial subcompartments and their folding and assembly follow principles already developed in prokaryotes. The components involved may have structural and functional equivalents in bacteria.

Apocytochrome c: an exceptional mitochondrial precursor protein using an exceptional import pathway

The cytochrome c import pathway differs markedly from the general route taken by the majority of other imported proteins, which is characterized by the import involvement of namely, surface receptors, the general insertion protein (GIP), contact sites and by the requirement of a membrane potential (delta psi). Unique features of both the cytochrome c precursor (apocytochrome c) and of the mechanism that transports it into mitochondria, have contributed to the evolution of a distinct import pathway that is not shared by any other mitochondrial protein analysed thus far. The cytochrome c pathway is particularly unique because i) apocytochrome c appears to have spontaneous membrane insertion-activity; ii) cytochrome c heme lyase seems to act as a specific binding site in lieu of a surface receptor and; iii) covalent heme addition and the associated refolding of the polypeptide appears to provide the free energy for the translocation of the cytochrome c polypeptide across the outer mitochondrial membrane.

Early steps in mitochondrial protein import: receptor functions can be substituted by the membrane insertion activity of apocytochrome c

The process of insertion of precursor proteins into mitochondrial membranes was investigated using a hybrid protein (pSc1-c) that contains dual targeting information and, at the same time, membrane insertion activity. pSc1-c is composed of the matrix-targeting domain of the cytochrome c1 presequence joined to the amino terminus of apocytochrome c. It can be selectively imported along either a cytochrome c1 route into the mitochondrial matrix or via the cytochrome c route into the intermembrane space. In contrast to cytochrome c1, pSc1-c does not require the receptor system/GIP for entry into the matrix. The apocytochrome c in the pSc1-c fusion protein appears to exert its membrane insertion activity in such a manner that the matrix-targeting sequence gains direct access to the membrane potential-dependent step. These results attribute an essential function to the receptor system in facilitating the initial insertion of precursors into the mitochondrial membranes.

Mitochondrial protein import

Most mitochondrial proteins are synthesized as precursor proteins on cytosolic polysomes and are subsequently imported into mitochondria. Many precursors carry amino-terminal presequences which contain information for their targeting to mitochondria. In several cases, targeting and sorting information is also contained in non-amino-terminal portions of the precursor protein. Nucleoside triphosphates are required to keep precursors in an import-competent (unfolded) conformation. The precursors bind to specific receptor proteins on the mitochondrial surface and interact with a general insertion protein (GIP) in the outer membrane. The initial interaction of the precursor with the inner membrane requires the mitochondrial membrane potential (delta psi) and occurs at contact sites between outer and inner membranes. Completion of translocation into the inner membrane or matrix is independent of delta psi. The presequences are cleaved off by the processing peptidase in the mitochondrial matrix. In several cases, a second proteolytic processing event is performed in either the matrix or in the intermembrane space. Other modifications can occur such as the addition of prosthetic groups (e.g., heme or Fe/S clusters). Some precursors of proteins of the intermembrane space or the outer surface of the inner membrane are retranslocated from the matrix space across the inner membrane to their functional destination ('conservative sorting'). Finally, many proteins are assembled in multi-subunit complexes. Exceptions to this general import pathway are known. Precursors of outer membrane proteins are transported directly into the outer membrane in a receptor-dependent manner. The precursor of cytochrome c is directly translocated across the outer membrane and thereby reaches the intermembrane space. In addition to the general sequence of events which occurs during mitochondrial protein import, current research focuses on the molecules themselves that are involved in these processes.

An atypical heme-binding structure of cytochrome c1 of Euglena gracilis mitochondrial complex III

Complex III was purified from submitochondrial particles prepared from Euglena gracilis. The purified complex consisted of 10 subunits and lost antimycin sensitivity. The Euglena complex III showed an atypical difference absorption spectrum for cytochrome c1 with its alpha-band maximum at 561 nm. The pyridine ferrohemochrome prepared from covalently bound heme in the Euglena complex III had an alpha-peak at 553 nm. This wavelength is the same as that of pyridine ferrohemochrome prepared from Euglena mitochondrial cytochrome c (c-558), the heme of which is linked to only a single cysteine residue through a thioether bond. Cytochrome c1 which was a heme-stained subunit with a molecular mass of 32.5 kDa was isolated from the purified complex III and its N-terminal sequence of 46 amino acids was determined. On the basis of apparent homologies to cytochromes c1 from other sources, this sequence included the heme-binding region. However, the amino acid at position 36, corresponding to the first cysteine involved in heme linkage in other cytochromes c1, was phenylalanine. Position 39, corresponding to the second cysteine, was not identified despite the treatment for removal of the heme and carboxymethylation of the expected cysteine. The unidentified amino acid is assumed to be a derivative of cysteine to which the heme is linked through a single thioether bond. The histidine-40 corresponding to the probable fifth ligand for heme iron was conserved in Euglena cytochrome c1.

Resonance Raman spectroscopy indicates a lysine as the sixth iron ligand in cytochrome f

The resonance Raman spectrum of turnip cytochrome f is similar to that of other c-type cytochromes with the exception of a single band at 1532 cm-1 which is shifted to lower frequency relative to its position (1542-1545 cm-1) in other c-type cytochromes. Comparison of the frequency of this band with that in alkylated cytochrome c at high pH suggests that the sixth heme iron ligand in cytochrome f is a deprotonated lysine amino group rather than a methionine sulfur. Comparison of the amino-acid sequences of cytochromes f and c1 suggests lysine-145 as a likely candidate for the sixth heme iron ligand in cytochrome f.

Import of proteins into mitochondria: a multi-step process

Translocation of precursor proteins from the cytosol into mitochondria is a multi-step process. The generation of translocation intermediates, i.e. the reversible accumulation of precursors at distinct stages of their import pathway into mitochondria ('translocation arrest'), has allowed the experimental characterization of distinct functional steps of protein import. These steps include: ATP-dependent unfolding of precursors; specific recognition of precursors by distinct receptors on the mitochondrial surface; interaction of precursors; specific recognition of precursors by distinct receptors on the mitochondrial surface; interaction of precursors with a general insertion protein ('GIP') in the outer mitochondrial membrane; membrane-potential-dependent translocation into the inner membrane at contact sites between both membranes; proteolytic processing of precursors; and intramitochondrial sorting of precursors via the matrix space ('conservative sorting'). The functional characteristics unveiled by studying mitochondrial protein import appear to be of general interest for investigations on intracellular protein sorting.

Successive translocation into and out of the mitochondrial matrix: targeting of proteins to the intermembrane space by a bipartite signal peptide

We investigated the import and sorting pathways of cytochrome b2 and cytochrome c1, which are functionally located in the intermembrane space of mitochondria. Both proteins are synthesized on cytoplasmic ribosomes as larger precursors and are processed in mitochondria in two steps upon import. The precursors are first translocated across both mitochondrial membranes via contact sites into the matrix. Processing by the matrix peptidase leads to intermediate-sized forms, which are subsequently redirected across the inner membrane. The second proteolytic processing occurs in the intermembrane space. We conclude that the hydrophobic stretches in the presequences of the intermediate-sized forms do not stop transfer across the inner membrane, but rather act as transport signals to direct export from the matrix into the intermembrane space.

Mitochondrial porin of Neurospora crassa: cDNA cloning, in vitro expression and import into mitochondria

cDNA encoding porin of Neurospora crassa, the major protein component of the outer mitochondrial membrane, was isolated and the nucleotide sequence was determined. The deduced protein sequence consists of 283 amino acids (29,979 daltons) and shows sequence homology of around 43% to yeast porin; however, no significant homology to bacterial porins was apparent. According to secondary structure predictions, mitochondrial porin consists mainly of membrane-spanning sided beta-sheets. Porin was efficiently synthesized in vitro from the cDNA; this allowed us to study in detail its import into mitochondria. Thereby, three characteristics of import were defined: (i) import depended on the presence of nucleoside triphosphates; (ii) involvement of a proteinaceous receptor-like component on the surface of the mitochondria was demonstrated; (iii) insertion into the outer membrane was resolved into at least two distinct steps: specific binding to high-affinity sites and subsequent assembly to the mature form.