Covalent cross-linking of the active sites of vesicle-bound cytochrome b5 and NADH-cytochrome b5 reductase

A water-soluble carbodiimide has been used to promote the formation of amide bonds between carboxyl residues on cytochrome b5 and lysyl residues on cytochrome b5 reductase. The visible and UV absorption spectrum of the purified cross-linked complex was identical with the sum of the spectra of the individual enzymes, and the average apparent molecular weight of the complex, determined by sodium dodecyl sulfate-gel electrophoresis, was within 12% of the sum of the apparent molecular weights of the two monomeric enzymes, indicating that the cross-linked derivative was a dimer containing one molecule each of cytochrome b5 and cytochrome b5 reductase. When reconstituted into phospholipid vesicles, the amphipathic derivative showed substantially reduced Vmax values with the soluble electron acceptors potassium ferricyanide, cytochrome b5 heme peptide and cytochrome c, and with the membrane-bound acceptors amphipathic cytochrome b5 and stearyl-CoA desaturase. The soluble catalytic fragment of the derivative, produced by limited digestion with subtilisin Carlsberg, showed similar decreases in Vmax values with the above soluble acceptors. In contrast, intradimer electron transfer in the soluble fragment, measured by stopped flow spectrophotometry at 2 degrees C was very efficient. Ninety per cent of the cytochrome b5 in the derivative was reduced with a first order rate constant of 51 s-1 upon the addition of NADH; the transfer of electrons from NADH to the reductase FAD prosthetic group, which is known to be the rate-limiting step in the reductase reaction mechanism, proceeded with an apparent rate constant of 57 s-1 under these conditions. These kinetic data show that the enzymes in the complex are cross-linked together at the surfaces involved in protein-protein contacts during electron transfer in an orientation similar to that assumed during electron transfer between the free proteins.     

Cytochrome b5 induced flip-flop of phospholipids in sonicated vesicles

Cytochrome b5 induced flip-flop of phosphatidylethanolamine (PE) in sonicated vesicles prepared from a 9:1 mixture of phosphatidylcholine (PC) to phosphatidylethanolamine was determined as follows. First, vesicles having a nonequilibrium distribution of PE across the bilayer were prepared by amidinating the external amino groups with isethionyl acetimidate. Amidinated cytochrome b5 was then added, and after the protein was completely bound, the rate of appearance of fresh PE on the outer surface was determined by removing aliquots at timed intervals and titrating the external amino groups with trinitrobenzenesulfonic acid. The results show an initial rapid phase of flip-flop (especially in the presence of salt) followed by a very slow phase, at 25 degrees C. Similar results were obtained when cytochrome b5 was introduced into the amidinated vesicles by spontaneous transfer from PC donor vesicles. These results indicate that the accumulation of the transferable ("loose") form of cytochrome b5 on the outer surface of a vesicle causes a transient, global destabilization of the bilayer that is relieved by lipid flip-flop. We speculate that this mechanism may be a significant driving force for the transfer of amphipathic molecules across membranes.     

Minimum requirements for immunogenic and antigenic activities of homologs of a synthetic peptide of influenza virus hemagglutinin.

Synthetic peptides of increasing length and corresponding in sequence to the C-terminal end of the HA1 molecule of influenza virus were constructed and examined for their immunogenic and antigenic properties. Peptides containing at least the four C-terminal amino acids, when coupled to keyhole limpet hemocyanin, were capable of eliciting antibody in BALB/c mice that bound to the 24-residue parent peptide H3 HA1 (305 to 328). In the absence of a carrier, the C-terminal decapeptide was the shortest peptide capable of eliciting antibody. The specificity of this antibody was indistinguishable from that of a monoclonal antibody to the parent peptide which recognizes an epitope encompassed by the C-terminal seven residues. All peptides containing at least the C-terminal four residues were able to inhibit completely the binding of this monoclonal antibody to the parent peptide. Taken together, these results indicate that (i) the tetrapeptide is capable of eliciting specific antibody when coupled to a carrier, (ii) this tetrapeptide possesses all of the antigenic information necessary to occupy the paratope of a monoclonal antibody elicited by the longer parent peptide, and (iii) the decapeptide contains all of the information necessary to elicit a specific immune response and therefore carries an epitope recognized by T cells as well as one recognized by B cells.     

Cytochrome b5, cytochrome c, and cytochrome P-450 interactions with NADPH-cytochrome P-450 reductase in phospholipid vesicles

Upon incubation of detergent-solubilized NADPH-cytochrome P-450 reductase and either cytochrome b5 or cytochrome c in the presence of a water-soluble carbodiimide, a 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide (EDC), covalently cross-linked complex was formed. The cross-linked derivative was a heterodimer consisting of one molecule each of flavoprotein and cytochrome, and it was purified to 90% or more homogeneity. The binary covalent complex between the flavoprotein and cytochrome b5 was exclusively observed following incubation of all three proteins including NADPH-cytochrome P-450 reductase, cytochrome b5, and cytochrome c in L-alpha-dimyristoylphosphatidylcholine vesicles, and no heterotrimer could be identified. The isolated reductase-cytochrome b5 complex was incapable of covalent binding with cytochrome c in the presence of EDC. No clear band for covalent complex formation between PB-1 and reductase was seen with the present EDC cross-linking technique. More than 90% of the cross-linked cytochrome c in the purified derivative was rapidly reduced upon addition of an NADPH-generating system, whereas approximately 80% of the cross-linked cytochrome b5 was rapidly reduced. These results showed that in the greater part of the complexes, the flavin-mediated pathway for reduction of cytochrome c or cytochrome b5 by pyridine nucleotide was intact. When reconstituted into phospholipid vesicles, the purified amphipathic derivative could hardly reduce exogenously added cytochrome c, cytochrome b5, or PB-1, indicating that the cross-linked cytochrome shields the single-electron-transferring interface of the flavoprotein. These results suggest that the covalent cross-linked derivative is a valid model of the noncovalent functional electron-transfer complex.     

The membrane-embedded segment of cytochrome b5 as studied by cross-linking with photoactivatable phospholipids

Vesicles were prepared from a 9:1 (mole/mol) mixture of dipalmitoyl phosphatidylcholine and the radioactively labeled phospholipids, 1-palmitoyl-2-omega-(m-diazirinophenoxy)undecanoyl-sn-glycero-3-phosphocholine (PC-I) or 1-palmitoyl-2-omega-(2-diazo-3,3,3-trifluropropionyloxy)lauroyl-sn- glycero-3-phosphocholine (PC-II). Rabbit liver cytochrome b5 was inserted into these vesicles spontaneously and the resulting vesicles containing the cytochrome b5 in the transferable form were photolyzed. Cytochrome b5 containing covalently cross-linked phospholipids was isolated by Sephadex LH-60 column chromatography using ethanol/formic acid as the solvent. Of the total radioactivity, 4.6% (PC-I) or 11.3% (PC-II) was linked to the protein; of the former, up to 51% was base-labile, while in the latter, 22% was base-labile. The sites of cross-linking of PC-I to the protein were investigated by fragmentation with trypsin, Staphylococcus aureas V8 protease, CNBr, and o-iodosobenzoic acid followed by Sephadex LH-60 chromatography and Edman sequencing (solid phase) of the appropriate fragments. The distribution of cross-linking was broad (Ser-104 to Met-130), showing a bell-shaped pattern with a significant peak at Ser-118. The labeling pattern is consistent with the previously proposed loop-back model for the membranous segment in the transferable form of cytochrome b5.     

Mass spectrometry to characterize the binding of a peptide to a lipid surface.

The binding of an amphipathic alpha-helical peptide to small unilamellar lipid vesicles has been examined using chemical derivitization and mass spectrometry. The peptide is derived from the sequence of human apolipoprotein C-II (apoC-II), the protein activator of lipoprotein lipase (LpL). ApoC-II(19-39) forms approximately 60% alpha-helix upon binding to model egg yolk phosphatidylcholine small unilamellar vesicles. Measurement of the affinity of the peptide for lipid by spectrophotometric methods is complicated by the contribution of scattered light to optical signals. Instead, we characterize the binding event using the differential labeling of lysine residues by the lipid- and aqueous-phase cross-linkers, disuccinimidyl suberate (DSS) and bis(sulfosuccinimidyl) suberate (BS(3)), respectively. In aqueous solution, the three lysine residues of the peptide are accessible to both cross-linkers. In the presence of lipid, the C-terminal lysine residue becomes inaccessible to the lipid-phase cross-linker DSS, but remains accessible to the aqueous-phase cross-linker, BS(3). We use mass spectrometry to characterize this binding event and to derive a dissociation constant for the interaction (K(d) = 5 microM). We also provide evidence for the formation of dimeric cross-linked peptide when high densities of peptide are bound to the lipid surface.     

Purification and properties of a cross-linked complex between cytochrome c and cytochrome c peroxidase.

Cytochrome c (horse heart) was covalently linked to yeast cytochrome c peroxidase by using the cleavable bifunctional reagent dithiobis-succinimidyl propionate in 5 mM-sodium phosphate buffer, pH 7.0. A cross-linked complex of molecular weight 48 000 was purified in approx. 10% yield from the reaction mixture, which contained 1 mol of cytochrome c and 1 mol of cytochrome c peroxidase/mol. Of the total 40 lysine residues, four to six were blocked by the cross-linking agent. Dithiobis-succinimidylpropionate can also cross-link cytochrome c to ovalbumin, but cytochrome c peroxidase is the preferred partner for cytochrome c in a mixture of the three proteins. The cytochrome c cross-linked to the peroxidase can be rapidly reduced by free cytochrome c-557 from Crithidia oncopelti, and the equilibrium obtained can be used to calculate a mid-point oxidation-reduction potential for the cross-linked cytochrome of 243 mV. Mitochondrial NADH-cytochrome c reductase will reduce the bound cytochrome only very slowly, but the rate of reduction by ascorbate at high ionic strength approaches that for free cytochrome c. Bound cytochrome c reduced by ascorbate can be re-oxidized within 10s by the associated peroxidase in the presence of equimolar H2O2. In the standard peroxidase assay the cross-linked complex shows 40% of the activity of the free peroxidase. Thus the intrinsic ability of each partner in the complex to take part in electron transfer is retained, but the stable association of the two proteins affects access of reductants.     

Affinity labelling of the allosteric site of the L-lactate dehydrogenase of Lactobacillus casei

We have recently described the addition of 2-keto-3-butynoic acid to flavin-free flavocytochrome b2, a reaction which leads to the loss of flavin-binding capacity ('inactivation') [D. Pompon and F. Lederer (1982) Eur. J. Biochem. 129, 143-137]. For total inactivation, the extrapolated incorporation value was 0.9 mol reagent/mol subunit. In this work we report the results of sequence studies which elucidate the nature of the modification. The modified protein was cleaved with cyanogen bromide and the peptides separated on Sephadex G-100 and SP-Sephadex C-25. 14C-labeled peptides were digested with trypsin and chymotrypsin and smaller labeled fragments purified by chromatography on Sephadex G-50 and thin-layer fingerprinting. It is shown that three cysteine residues are fractionally labeled with nearly complete mutual exclusion. Furthermore, a fraction of the modified peptides is found under the form of cross-linked fragments, where two cysteines have added to the same ketobutynoate molecule. Only two of the possible cross-links were found. These results show that the three cysteines are close to one another in space in the flavin-free enzyme and hence probably also in the holoenzyme. These results, combined with those obtained in the affinity labeling reaction of holoenzyme with bromopyruvate [Alliel et al. (1982) Eur. J. Biochem. 122, 553-558], show that the three residues are located in or close to the active site. Their possible role is discussed.     

Interaction of cholesterol hydroxylating cytochrome P-450 with cytochrome b5

Some new relations between cytochrome P-450-dependent monooxygenases were discovered. Cytochrome b5, a representative of "microsomal" monooxygenases, was shown to form a highly specific complex with cytochrome P-450scc, a member of the "ferredoxin" monooxygenase family. This interaction is characterized by a dissociation constant, Kd, of 0.28 microM. The cytochrome P-450scc-cytochrome b5 complex may be cross-linked with water-soluble carbodiimide. Using proteolytic modification of cytochrome b5, it was shown that both hydrophilic and hydrophobic fragments of cytochrome b5 are involved in the interaction with cytochrome P-450scc. Cytochrome b5 immobilized via amino groups is an effective affinity matrix for cytochrome P-450scc purification. The role of some amino acid residues in cytochrome P-450scc interaction with cytochrome b5 was studied. The role and the nature of complexes in cytochrome P-450-dependent monooxygenases as well as interrelationships between "microsomal" and "ferredoxin" monooxygenases are discussed.     

Tyrosine-371 contributes to the positive cooperativity between the two cAMP binding sites in the regulatory subunit of cAMP-dependent protein kinase I

The regulatory (R) subunit of cAMP-dependent protein kinase I has been expressed in Escherichia coli, and oligonucleotide-directed mutagenesis was initiated in order to better understand structural changes that are induced as a consequence of cAMP-binding. Photoaffinity labeling of the type I holoenzyme with 8-azidoadenosine 3',5'-monophosphate (8-N3cAMP) leads to the covalent modification of two residues, Trp-260 and Tyr-371 [Bubis, J., & Taylor, S.S. (1987) Biochemistry 26, 3478-3486]. The site that was targeted for mutagenesis was Tyr-371. The intention was to establish whether the interactions between the tyrosine ring and the adenine ring of cAMP are primarily hydrophobic in nature or whether the hydroxyl group is critical for cAMP binding and/or for inducing conformational changes. A single base change converted Tyr-371 to Phe. This yielded an R subunit that reassociated with the catalytic subunit to form holoenzyme and bound 2 mol of cAMP/mol of R monomer. The cAMP binding properties of the holoenzyme that was formed with this mutant R subunit, however, were altered: (a) the apparent Kd(cAMP) was shifted from 16 to 60 nM; (b) Scatchard plots showed no cooperativity between the cAMP binding sites in the mutant in contrast to the positive cooperativity that is observed for the wild-type holoenzyme; (c) the Hill coefficient of 1.6 for the wild-type holoenzyme was reduced to 0.99. The Ka's for activation by cAMP were altered in the mutant holoenzyme in a manner that was proportional to the shift in Kd(cAMP).(ABSTRACT TRUNCATED AT 250 WORDS)     

Relation of the Escherichia coli dnaX gene to its two products--the tau and gamma subunits of DNA polymerase III holoenzyme.

The Escherichia coli DNA polymerase III holoenzyme 71.1 kDa tau subunit is a 643 amino acid protein encoded by the dnaX gene. This gene also encodes the holoenzyme 56.5 kDa gamma subunit. The tau factor (as a tau'-LacZ' fusion protein) has been isolated and shown to be cleaved in vitro to form gamma and a 135 kda C-terminal cleavage product. The tau'-LacZ' fusion protein, gamma, and the C-terminal cleavage product have been isolated. N-terminal sequencing has demonstrated that tau and gamma share the same N-terminal sequences and that tau is proteolytically cleaved in vitro between residues 498 and 499 to form gamma. In addition, residues 420-440 were shown to be present in both tau and gamma by use of antibody specific for a synthetic peptide corresponding to that sequence. Some mechanism functions in vivo to ensure that tau and gamma are synthesized in a ratio of about one-to-one, as shown by radioimmune precipitation of tau and gamma from cellular extracts.     

Nuclear-magnetic-resonance studies on the conformation of membrane-bound alpha-mating factor. Transferred nuclear Overhauser effect analysis

The C-H proton resonances of alpha-mating factor, yeast pheromone, in 2H2O solution were assigned. The phase transition temperature of perdeuterated dipalmitoylglycerophosphocholine (suspension) was found to be 35.5 degrees C. In the presence of vesicles of this phospholipid, the exchange broadening and transferred nuclear Overhauser effect (TRNOE) of peptide proton resonances (at 50 degrees C) were analyzed. The mode of binding of this peptide with the phospholipid bilayer was elucidated. The N-terminal nine residues (Trp1-Gly9) are tightly bound to the bilayer, while the C-terminal four residues (Gln10-Tyr13) are left free in aqueous phase. This is consistent with the previous observation that the C-terminal three residues (Pro11-Tyr13) are not essential for the activity of this pheromone [Masui, Y. et al. (1977) Biochem. Biophys. Res. Commun. 78, 534-538]. Furthermore, from the TRNOE analyses, the conformation of the membrane-bound N-terminal part of alpha-mating factor was elucidated; the residues Trp1-Gln5 form a compact helical structure while the residues Lys7-Gly9 form an extended structure. A similar TRNOE was also observed for an active decapeptide analog Trp1-Gln10. This confirms the previous conclusion that the physiological activities of this pheromone and analog peptides are correlated with the conformations of membrane-bound peptide molecules [Higashijima, T. et al. (1983) FEBS Lett. 159, 229-232].     

Lipid-protein interactions in membrane models. Fluorescence polarization study of cytochrome b5-phospholipids complexes.

According to previous authors, cytochrome b5, when extracted from bovine liver by a detergent method, is called cytochrome d-b5. On the other hand, the protein obtained after trypsin action, which eliminates an hydrophobic peptide of about 54 residues, is called cytochrome t-b5. Fluorescence polarization of the dansyl phosphatidylethanolamine probe inserted into phospholipid vesicles is very sensitive to the binding of proteins, and so is a useful method to study lipid-protein interactions. The chromophore mobility, R, decreases markedly when dipalmitoyl phosphatidylcholine vesicles are incubated with cytochrome d-b5, whereas R does not change for cytochrome c and cytochrome t-b5. This can be interpreted as a strengthening of bilayer, only due to the interaction of the hydrophobic peptide tail. Interaction of dipalmitoly phosphatidylcholine vesicles with cytochrome d-b5 occurs either below or above the melting temperature of the aliphatic chains (41 degrees C). Even for a high protein to lipid molar ratio (1 molecule of protein for 40 phospholipid molecules), the melting temperature is apparently unaffected. Phosphatidylserine and phosphatidylinositol do not interact at pH 7.7 with cytochrome d-b5, because electrostatic forces prevent formation of complexes. At low pH, the interaction with the protein occurs, but the binding is mainly of electrostatic nature.     

Primary structure of the wall peptidoglycan of leprosy-derived corynebacteria.

The cell walls isolated from axenically grown leprosy-derived corynebacteria were submitted to various chemical and enzymatic degradations. The glycan strands of the wall peptidoglycan are essentially composed of N-acetylglycosaminyl-N-acetylmuramic acid disaccharide units. Small amounts of N-acetylglycosaminyl-N-glycolylmuramic acid (less than 10%) were also detected. The muramic acid residues of adjacent glycan strands are substituted by amidated tetrapeptide units which, in turn, are cross-linked through direct linkages extending between the C-terminal D-alanine residue of one tetrapeptide and the mesodiaminopimelic acid residue of another tetrapeptide. Such a structure is very similar to that of the wall peptidoglycan found in the taxonomically related microorganisms of the Corynebacterium, Mycobacterium, and Nocardia groups.     

Peptides that mimic the pseudosubstrate region of protein kinase C bind to acidic lipids in membranes.

The cytoplasmic form of protein kinase C (PKC) is inactive, probably because the pseudosubstrate region in its regulatory domain blocks the substrate-binding site in its kinase domain. Calcium ions cause a translocation to the membrane: maximum activation requires a negative lipid such as phosphatidylserine (PS) and the neutral lipid diacylglycerol (DAG) but the mechanism by which PS and DAG activate PKC is unknown. Pseudosubstrate region 19-36 of PKC-beta has six basic and one acidic amino acids and region 19-29 has five basic and no acidic amino acids. Since any binding of basic residues in the pseudosubstrate region to acidic lipids in the membrane should stabilize the active form of PKC, we studied how peptides with amino acids equivalent to residues 19-36 and 19-29 of PKC-beta bound to phospholipid vesicles. We made equilibrium dialysis, filtration, and electrophoretic mobility measurements. The fraction of bound peptide is a steep sigmoidal function of the mol fraction of negative lipid in the membrane, as predicted from a simple theoretical model that assumes the basic residues provide identical independent binding sites. The proportionality constant between the number of bound peptides/area and the concentration of peptide in the bulk aqueous phase is 1 micron for a membrane with 25% negative lipid formed in 0.1 M KCl. Equivalently, the association constant of the peptide with the membrane is 10(4) M-1, or the net binding energy is 6 kcal/mol. Thus the interaction of basic residues in the pseudosubstrate region with acidic lipids in the membrane could provide 6 kcal/mol free energy towards stabilizing the active form of PKC.     

Membrane interaction of Escherichia coli hemolysin: flotation and insertion-dependent labeling by phospholipid vesicles

The 1,024-amino-acid acylated hemolysin of Escherichia coli subverts host cell functions and causes cell lysis. Both activities require insertion of the toxin into target mammalian cell membranes. To identify directly the principal toxin sequences dictating membrane binding and insertion, we assayed the lipid bilayer interaction of native protoxin, stably active toxin, and recombinant peptides. Binding was assessed by flotation of protein-liposome mixtures through density gradients, and insertion was assessed by labeling with a photoactivatable probe incorporated into the target lipid bilayer. Both the active acylated hemolysin and the inactive unacylated protoxin were able to bind and also insert. Ca(2+) binding, which is required for toxin activity, did not influence the in vitro interaction with liposomes. Three overlapping large peptides were expressed separately. A C-terminal peptide including residues 601 to 1024 did not interact in either assay. An internal peptide spanning residues 496 to 831, including the two acylation sites, bound to phospholipid vesicles and showed a low level of insertion-dependent labeling. In vitro acylation had no effect on the bilayer interaction of either this peptide or the full-length protoxin. An N-terminal peptide comprising residues 1 to 520 also bound to phospholipid vesicles and showed strong insertion-dependent labeling, ca. 5- to 25-fold that of the internal peptide. Generation of five smaller peptides from the N-terminal region identified the principal determinant of lipid insertion as the hydrophobic sequence encompassing residues 177 to 411, which is conserved among hemolysin-related toxins.     

Characterization of the covalent cross-links of the active sites of amidinated cytochrome b5 and NADH:cytochrome b5 reductase

Preparations of amidinated cytochrome b5 and cytochrome b5 reductase, cross-linked by using a soluble carbodiimide to promote the formation of covalent bonds between carboxyl groups of the hemeprotein and nucleophilic residues of the flavoprotein at the surfaces involved in protein-protein contacts during electron transfer, have been used to characterize the charge pair interactions that occur during electron transfer between the free proteins. Sequence analyses of tryptic, V8 protease-, and Asp-N protease-generated peptides show that the heme propionyl carboxyl group at the surface of the cytochrome forms an ester bond with Ser162 of the reductase, thus implicating Lys163 as the normal participant in ionic bonding between the active sites of the two proteins. Moreover, Lys41 and Lys125 directly form amide bonds with carboxyl residues on the active-site surface of the cytochrome. In the case of Lys41, this involves Glu52 and/or Glu60, and Glu47 and/or Glu48 for Lys125, again implicating these residues as the groups that form charge pairs during normal interactions between the active sites of the two proteins.     

Reversely-oriented cytochrome b561 in reconstituted vesicles catalyzes transmembrane electron transfer and supports extravesicular dopamine beta-hydroxylase activity

Cytochrome b561 from bovine adrenal chromaffin vesicles contains two heme B prosthetic groups. We verified that purified cytochrome b561 can donate electron equivalents directly to cytochrome c. The purified cytochrome b561 was successfully reconstituted into cholesterol-phosphatidylcholine-phosphatidylglycerol vesicles by a detergent-dialysis and extrusion method. When ascorbate-loaded vesicles with cytochrome b561 were mixed with ferricytochrome c, the intravesicular ascorbate was able to reduce external thiazole blue or cytochrome c. The reduction of thiazole blue or cytochrome c was dependent on the presence of cytochrome b561 in the vesicle membranes. Pre-treatment of cytochrome b561 with diethylpyrocarbonate suppressed the reduction of extravesicular cytochrome c significantly, confirming that the reduction was not due to leakage of ascorbate from the vesicles. The topology of the reconstituted cytochrome b561 in the vesicle membranes was examined by treatment with trypsin followed by SDS-PAGE and MALDI-TOF-MS analyses. Only one major cleavage site at Lys191 was identified, indicating that cytochrome b561 was reconstituted into the membranes in an inside-out orientation irrespective of the modification with diethylpyrocarbonate. The addition of a soluble form of dopamine beta-hydroxylase to the external medium resulted in the successful reconstitution of the hydroxylation activity towards tyramine, an analogue of dopamine, suggesting that a direct electron transfer via complex formation occurred. This activity was enhanced significantly upon the addition of ferricyanide as a mediator between cytochrome b561 and dopamine beta-hydroxylase.     

Structure of cytochrome b5 and its topology in the microsomal membrane

The complete amino acid sequence of human and chicken liver microsomal cytochrome b5 was determined. The amino termini of cytochrome b5 from four other mammalian species were examined in order to determine their complete covalent structure. As in the rat species, cytochrome b5 preparations from man, rabbit, calf and horse had an acetylated alanine as the first residue. In contrast, the pig cytochrome had alanine at the amino terminus. The amino terminus of the chicken cytochrome b5 was also unmodified, and extended three residues absent in the mammalian species. In order to investigate whether the carboxy-terminal segment of cytochrome b5 is located on the cytosolic or the luminal side of the microsomal membrane, rabbit liver microsomes were treated with trypsin and subjected to gel filtration and high-pressure liquid chromatography. The nonpolar peptide isolated from these microsomes lacked the terminal hexapeptide, indicating that when cytochrome b5 is bound to intact microsomes, the carboxy terminus is located on the cytosolic side of the membrane and does not extend in the lumen of the endoplasmic reticulum.     

Analysis of the quaternary structure of the MutL C-terminal domain

The dimeric DNA mismatch repair protein MutL has a key function in communicating mismatch recognition by MutS to downstream repair processes. Dimerization of MutL is mediated by the C-terminal domain, while activity of the protein is modulated by the ATP-dependent dimerization of the highly conserved N-terminal domain. Recently, a crystal structure analysis of the Escherichia coli MutL C-terminal dimerization domain has been reported and a model for the biological dimer was proposed. In this model, dimerization is mediated by the internal (In) subdomain comprising residues 475-569. Here, we report a computational analysis of all protein interfaces observed in the crystal structure and suggest that the biological dimer interface is formed by a hydrophobic surface patch of the external (Ex) subdomain (residues 432-474 and 570-615). Moreover, sequence analysis revealed that this surface patch is conserved among the MutL proteins. To test this hypothesis, single and double-cysteine variants of MutL were generated and tested for their ability to be cross-linked with chemical cross-linkers of various size. Finally, deletion of the C-terminal residues 605-615 abolished homodimerization. The biochemical data are fully compatible with a revised model for the biological dimer, which has important implications for understanding the heterodimerization of eukaryotic MutL homologues, modeling the MutL holoenzyme and predicting protein-protein interaction sites.