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.     

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.     

Association of protein kinase C with phospholipid monolayers: two-stage irreversible binding

The association of protein kinase C (PKC) with phospholipid (PL) monolayers spread at the air-water interface was examined. PKC-PL binding induced surface pressure changes that were dependent on the amount of PKC, the phospholipid composition of the monolayers, the presence of Ca2+, and the initial surface pressure of the monolayer (pi 0). Examination of surface pressure increases induced by PKC as a function of phospholipid surface pressure, pi 0, revealed that PKC-phosphatidylserine (PS) association had a critical pressure of 43 dyn/cm. Above this surface pressure, PKC cannot cause further surface pressure changes. This high critical pressure indicated that PKC should be able to penetrate many biological membranes which appear to have surface pressures of about 30 dyn/cm. PKC-induced surface pressure changes were Ca2+ dependent only for PL monolayers spread at a pi 0 greater than 26 dyn/cm. PKC alone (in the absence of PL) formed a film at the air-water interface with a surface pressure of about 26 dyn/cm. Calcium-dependent binding was studied at the higher surface pressures which effectively excluded PKC from the air-water interface. Subphase depletion measurements suggested that association of PKC with PS monolayers consisted of two stages: a rapid Ca2+-dependent interaction followed by a slower process that resulted in irreversible binding of PKC to the monolayer. The second stage appeared to involve penetration of PKC into the hydrocarbon region of the phospholipid. The commonly used in vitro substrates for PKC, histone and protamine sulfate, also associated with and penetrated PS monolayers with critical pressures of 50 and 60 dyn/cm, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)     

Interaction of lipid vesicles with monomolecular layers containing lung surfactant proteins SP-B or SP-C

Pulmonary surfactant contains two families of hydrophobic proteins, SP-B and SP-C. Both proteins are thought to promote the formation of the phospholipid monolayer at the air-fluid interface of the lung. The Wilhelmy plate method was used to study the involvement of SP-B and SP-C in the formation of phospholipid monolayers. The proteins were either present in the phospholipid vesicles which were injected into the subphase or included in a preformed phospholipid monolayer. In agreement with earlier investigators, we found that SP-B and SP-C, present in phospholipid vesicles, were able to induce the formation of a monolayer, as became apparent by an increase in surface pressure. However, when the proteins were present in a preformed phospholipid monolayer (20 mN/m) at similar lipid to protein ratios, the rate of surface pressure increase after injection of pure phospholipid vesicles into the subphase at similar vesicle concentrations was 10 times higher. The process of phospholipid insertion from phospholipid vesicles into the protein-containing monolayers was dependent on (1) the presence of (divalent) cations, (2) the phospholipid concentration in the subphase, (3) the size of the phospholipid vesicles, (4) the protein concentration in the preformed monolayer, and (5) the initial surface pressure at which the monolayers were formed. Both in vesicles and in preformed monolayers, SP-C was less active than SP-B in promoting the formation of a phospholipid monolayer. The use of preformed monolayers containing controlled protein concentrations may allow more detailed studies on the mechanism by which the proteins enhance phospholipid monolayer formation from vesicles.     

The interaction of 2,4-dinitrophenol with phospholipids at phospholipid-water interfaces

The effect of 2, 4-dinitrophenol, DNP, on monolayers of egg lecithin, hydrogenated egg lecithin, dipalmitoyl lecithin and mitochondrial lipids has been examined. Both the undissociated and dissociated forms of DNP bind to the phospholipid polar groups. Binding of the acid form leads to a decrease in monolayer surface potential and an expansion of the monolayer. The amount of penetration of the acid form into lecithin monolayers appears to depend on the London-Van der Waals attractions between the lecithin hydrocarbon chains. Binding of the 2,4-dinitro-phenolate anion is reflected in a decrease in surface potential for lecithin monolayers, and an increase in surface potential for mitochondrial lipid monolayers. The adsorption of dinitro-phenolate to egg lecithin has been further investigated by micro-electrophoresis of lecithin liposomes. It is suggested that binding of DNP to phospholipid-water interfaces is important in determining its action as an uncoupler of oxidative phosphorylation, and as a compound that increases the electrical conductance of artificial lipid membranes.     

Molecular associations and surface-active properties of short- and long-N-acyl chain ceramides

The behaviour of N-hexadecanoylsphingosine (Cer16), N-hexanoylsphingosine (Cer6) and N-acetylsphingosine (Cer2) in aqueous media and in lipid-water systems, monolayers and bilayers has been comparatively examined using Langmuir balance and fluorescence techniques. Cer16 behaves as an insoluble non-swelling amphiphile, not partitioning into the air-water interface, thus not modifying the surface pressure of the aqueous solutions into which it is included. By contrast both Cer6 and Cer2 behave as soluble amphiphiles, up to approx. 100 microM. At low concentrations, they become oriented at the air-water interface, increasing surface pressure in a dose-dependent way up to ca. 5 microM bulk concentration. At higher concentrations, the excess ceramide forms micelles, critical micellar concentrations of both Cer6 and Cer2 being in the 5-6 microM range. When the air-water interface is occupied by a phospholipid, 6Cer2 and Cer6 become inserted in the phospholipid monolayer, causing a further increase in surface pressure. This increase is dose dependent, and reaches a plateau at ca. 2 microM ceramide bulk concentration. Both Cer2 and Cer6 become inserted in phospholipid monolayers with initial surface pressures of up to 43 and 46 mN m(-1), respectively, which ensures their capacity to become inserted into cell membranes whose monolayers are estimated to support a surface pressure of about 30 mN m(-1). Both Cer2 and Cer6, but not Cer16, had detergent-like properties, such as giving rise to phospholipid-ceramide mixed micelles, when added to phospholipid monolayers or bilayers. The short-chain ceramides form large aggregates and precipitate at concentrations above approx. 100 microM. These results are relevant in cell physiology studies in which short- and long-chain ceramides are sometimes used as equivalent molecules, in spite of their different biophysical behaviour.     

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.     

Ubiquinones have surface-active properties suited to transport electrons and protons across membranes

Surface-active properties of ubiquinones and ubiquinols have been investigated by monomolecular-film techniques. Stable monolayers are formed at an air/water interface by the fully oxidized and reduced forms of the coenzyme; collapse pressures and hence stability of the films tend to increase with decreasing length of the isoprenoid side chain and films of the reduced coenzymes are more stable than those of their oxidized counterparts. Ubiquinone with a side chain of two isoprenoid units does not form stable monolayers at the air/water interface. Mixed monolayers of ubiquinol-10 or ubiquinone-10 with 1,2-dimyristoyl phosphatidylcholine, soya phosphatidylcholine and diphosphatidylglycerol do not exhibit ideal mixing characteristics. At surface pressures less than the collapse pressure of pure ubiquinone-10 monolayers (approx. 12mN.m(-1)) the isoprenoid chain is located substantially within the region occupied by the fatty acyl residues of the phospholipids. With increasing surface pressure the ubiquinones and their fully reduced equivalents are progressively squeezed out from between the phospholipid molecules until, at a pressure of about 35mN.m(-1), the film has surface properties consistent with that of the pure phospholipid monolayer. This suggests that the ubiquinone(ol) forms a separate phase overlying the phospholipid monolayer. The implications of this energetically poised situation, where the quinone(ol) is just able to penetrate the phospholipid film, are considered in terms of the function of ubiquinone(ol) as electron and proton carriers of energy-transducing membranes.     

Adult rat cardiomyocytes cultured in creatine-deficient medium display large mitochondria with paracrystalline inclusions, enriched for creatine kinase

In adult regenerating cardiomyocytes in culture, in contrast to fetal cells, mitochondrial creatine kinase (Mi-CK) was expressed. In the same cell, two populations of mitochondria, differing in shape, in distribution within the cell and in content of Mi-CK, could be distinguished. Immunofluorescence studies using antibodies against Mi-CK revealed a characteristic staining pattern for the two types of mitochondria: giant, mostly cylindrically shaped, and, as shown by confocal laser light microscopy, randomly distributed mitochondria exhibited a strong signal for Mi-CK, whereas small, "normal" mitochondria, localized in rows between myofibrils, gave a much weaker signal. Transmission EM of the giant mitochondria demonstrated paracrystalline inclusions located between cristae membranes. Immunogold labeling with anti-Mi-CK antibodies revealed a specific decoration of these inclusions for Mi-CK. Addition of 20 mM creatine, the substrate of Mi-CK, to the essentially creatine-free culture medium caused the disappearance of the giant cylindrically shaped mitochondria as well as of the paracrystalline inclusions, accompanied by an increase of the intracellular level of total creatine. Replacement of creatine in the medium by the creatine analogue and competitor beta-guanidinopropionic acid caused the reappearance of the enlarged mitochondria. It is believed that the accumulation of Mi-CK within the paracrystalline inclusions, similar to those observed in certain myopathies, represents a compensatory effect of the cardiomyocytes to cope with a metabolic stress situation caused by low intracellular total creatine levels.     

Surface-active properties of the antitumour ether lipid 1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine (edelfosine)

The surface activity and interaction with lipid monolayers and bilayers of the antitumour ether lipid 1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine (edelfosine) have been studied. Edelfosine is a surface-active soluble amphiphile, with critical micellar concentrations at 3.5 microM and 19 microM in water. When the air-water interface is occupied by a phospholipid, edelfosine becomes inserted in the phospholipid monolayer, increasing surface pressure. This increase is dose-dependent, and reaches a plateau at ca. 2 microM edelfosine bulk concentration. The ether lipid can become inserted in phospholipid monolayers with initial surface pressures of up to 33 mN/m, which ensures its capacity to become inserted into cell membranes. Upon interaction with phospholipid vesicles, edelfosine exhibits a weak detergent activity, causing release of vesicle contents to a low extent (<5%), and a small proportion of lipid solubilization. The weak detergent properties of edelfosine can be related to its very low critical micellar concentrations. Its high affinity for lipid monolayers combined with low lytic properties support the use of edelfosine as a clinical drug. The surface-active properties of edelfosine are similar to those of other "single-chain" lipids, e.g. lysophosphatidylcholine, palmitoylcarnitine, or N-acetylsphingosine.     

The effect of ferriprotoporphyrin IX and chloroquine on phospholipid monolayers and the possible implications to antimalarial activity

Ferriprotoporphyrin IX intercalates into phospholipid membranes, as evidenced from its effect on the surface pressure of monolayers composed of different phospholipids. Ferriprotoporphyrin intercalation is enhanced by membrane hydrophobicity and decreased by negative surface potential. Chloroquine enhances the effect of ferriprotoporphyrin in relatively hydrophobic membranes but reduces it in monolayers composed of highly unsaturated phospholipids. These results are consistent with the differential effect of chloroquine on ferriprotoporphyrin-induced lysis of erythrocytes and of malarial parasites, thus supporting the membrane-lesion hypothesis of antimalarial action.     

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.     

Molecular associations and surface-active properties of short- and long-N-acyl chain ceramides

The behaviour of N-hexadecanoylsphingosine (Cer16), N-hexanoylsphingosine (Cer6) and N-acetylsphingosine (Cer2) in aqueous media and in lipid-water systems, monolayers and bilayers has been comparatively examined using Langmuir balance and fluorescence techniques. Cer16 behaves as an insoluble non-swelling amphiphile, not partitioning into the air-water interface, thus not modifying the surface pressure of the aqueous solutions into which it is included. By contrast both Cer6 and Cer2 behave as soluble amphiphiles, up to approx. 100 μM. At low concentrations, they become oriented at the air-water interface, increasing surface pressure in a dose-dependent way up to ca. 5 μM bulk concentration. At higher concentrations, the excess ceramide forms micelles, critical micellar concentrations of both Cer6 and Cer2 being in the 5-6 μM range. When the air-water interface is occupied by a phospholipid, 6Cer2 and Cer6 become inserted in the phospholipid monolayer, causing a further increase in surface pressure. This increase is dose dependent, and reaches a plateau at ca. 2 μM ceramide bulk concentration. Both Cer2 and Cer6 become inserted in phospholipid monolayers with initial surface pressures of up to 43 and 46 mN m⁻¹, respectively, which ensures their capacity to become inserted into cell membranes whose monolayers are estimated to support a surface pressure of about 30 mN m⁻¹. Both Cer2 and Cer6, but not Cer16, had detergent-like properties, such as giving rise to phospholipid-ceramide mixed micelles, when added to phospholipid monolayers or bilayers. The short-chain ceramides form large aggregates and precipitate at concentrations above approx. 100 μM. These results are relevant in cell physiology studies in which short- and long-chain ceramides are sometimes used as equivalent molecules, in spite of their different biophysical behaviour.     

Protein-liposome interactions and their relevance to the structure and function of cell membranes

Summary Recent studies on the interactions of soluble proteins, membrane proteins and enzymes with phospholipid model membranes are reviewed. Similarities between the properties of such systems and the behavior of biomembranes, such as alterations in the redox potential of cytochromec after binding to membranes and effects of phospholipid fluidity on (Na + K) ATPase activity, are emphasized. The degree of correspondence between the behavior of model systems and natural membranes encourages the continuing use of model membranes in studies on protein-lipid interactions. However, some of the data on the increase of surface pressure of phospholipid monolayers by proteins and increases in the permeability of liposomes indicate that many soluble proteins also have a capability to interact hydrophobically with phospholipids. Thus a sharp distinction between both peripheral and integral membrane proteins and non-membrane proteins are not seen by these techniques. Cautious use of such studies, however, should lead to greater understanding of the molecular basis of cell membrane structure and function in normal and pathological states. Studies implicating protein-lipid interactions and (Na + K) ATPase activity in membrane alterations in disease states are also briefly discussed.An invited article.     

Mitochondrial creatine kinase interaction with heterogeneous monolayers: Effect on lipid lateral organization

Our study highlights the tight relationship between protein binding to monolayers and the phase-state of the phospholipids. Interaction of mitochondrial creatine kinase with phospholipidic membranes was analysed using a two-phase monolayer system containing anionic phospholipids under chain mismatch conditions. Monolayers were made up of mixtures of DMPC/DPPG or DPPC/DMPG containing 40% negatively charged phospholipids which is approximately the negative charge content of the mitochondrial inner membrane. Langmuir isotherms of these monolayers showed that they underwent a phase transition from a liquid expanded state to a liquid-condensed phase at about 2 mN/m and 5 mN/m respectively. Interface morphology modifications caused by injection of mtCK under these monolayers at low or high surface pressure were monitored by Brewster angle microscopy. This work provides evidence that the presence at the air/water interface of discrete domains with increased charge density, may lead to difference in partition of soluble proteins such as mtCK, interacting with the lipid monolayer. Conversely these proteins may help to organize charged phospholipid domains in a membrane.     

Divalent cation-induced interaction of phospholipid vesicle and monolayer membranes

The effects of phospholipid vesicles and divalent cations in the subphase solution on the surface tension of phospholipid monolayer membranes were studied in order to elucidate the nature of the divalent cation-induced vesicle-membrane interaction. The monolayers were formed at the air/water interface. Various concentrations of unilamellar phospholipid (phosphatidylserine, phosphatidylcholine and their mixtures) vesicles and divalent cations (Mg²⁺, Ca²⁺, Mn²⁺, etc.) were introduced into the subphase solution of the monolayers. The changes of surface tension of monolayers were measured by the Wilhelmy plate (Teflon) method with respect to divalent ion concentrations and time.When a monolayer of phosphatidylserine and vesicles of phosphatidylserine/phosphatidylcholine (1 : 1) were used, there were critical concentrations of divalent cations to produce a large reduction in surface tension of the monolayer. These concentrations were 16 mM for Mg²⁺, 7 mM for Sr²⁺, 6 mM for Ca²⁺, 3.5 mM for Ba²⁺ and 1.8 mM for Mn²⁺. On the other hand, for a phosphatidylcholine monolayer and phosphatidylcholine vesicles, there was no change in surface tension of the monolayer up to 25 mM of any divalent ion used. When a phosphatidylserine monolayer and phosphatidylcholine vesicles were used, the order of divalent ions to effect the large reduction of surface tension was Mn²⁺ > Ca²⁺ > Mg²⁺ and their critical concentrations were in between the former two cases. The threshold concentrations also depended upon vesicle concentrations as well as the area/molecule of monolayers. For phosphatidylserine monolayers and phosphatidylserine/phosphatidylcholine (1 : 1) vesicles, above the critical concentrations of Mn²⁺ and Ca²⁺, the surface tension decreased to a value close to the equilibrium pressure of the monolayers within 0.5 h.This decrease in surface tension of the monolayers is interpreted partly as the consequence of fusion of the vesicles with the monolayer membranes. The     

Interaction of the antimicrobial peptide pheromone Plantaricin A with model membranes: implications for a novel mechanism of action

Plantaricin A (plA) is a 26-residue bacteria-produced peptide pheromone with membrane-permeabilizing antimicrobial activity. In this study the interaction of plA with membranes is shown to be highly dependent on the membrane lipid composition. PlA bound readily to zwitterionic 1-stearoyl-2-oleoyl-sn-glycero-3-phosphocholine (SOPC) monolayers and liposomes, yet without significantly penetrating into these membranes. The presence of cholesterol attenuated the intercalation of plA into SOPC monolayers. The association of plA to phosphatidylcholine was, however, sufficient to induce membrane permeabilization, with nanomolar concentrations of the peptide triggering dye leakage from SOPC liposomes. The addition of the negatively charged phospholipid, 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-rac-glycerol POPG (SOPC/POPG; molar ratio 8:2) enhanced the membrane penetration of the peptide, as revealed by (i) peptide-induced increment in the surface pressure of lipid monolayers, (ii) increase in diphenylhexatriene (DPH) emission anisotropy measured for bilayers, and (iii) fluorescence characteristics of the two Trps of plA in the presence of liposomes, measured as such as well as in the presence of different quenchers. Despite deeper intercalation of plA into the SOPC/POPG lipid bilayer, much less peptide-induced dye leakage was observed for these liposomes than for the SOPC liposomes. Further changes in the mode of interaction of plA with lipids were evident when also the zwitterionic phospholipid, 1-palmitoyl-2-oleoyl-sn-glycerol-3-phosphoethanolaminne (POPE) was present (SOPC/POPG/POPE, molar ratio 3:2:5), thus suggesting increase in membrane spontaneous negative curvature to affect the mode of association of this peptide with lipid bilayer. PlA induced more efficient aggregation of the SOPC/POPG and SOPC/POPG/POPE liposomes than of the SOPC liposomes, which could explain the attenuated peptide-induced dye leakage from the former liposomes. At micromolar concentrations, plA killed human leukemic T-cells by both necrosis and apoptosis. Interestingly, plA formed supramolecular protein-lipid amyloid-like fibers upon binding to negatively charged phospholipid-containing membranes, suggesting a possible mechanistic connection between fibril formation and the cytotoxicity of plA.     

Surface-active properties of the antitumour ether lipid 1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine (edelfosine)

The surface activity and interaction with lipid monolayers and bilayers of the antitumour ether lipid 1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine (edelfosine) have been studied. Edelfosine is a surface-active soluble amphiphile, with critical micellar concentrations at 3.5 μM and 19 μM in water. When the air-water interface is occupied by a phospholipid, edelfosine becomes inserted in the phospholipid monolayer, increasing surface pressure. This increase is dose-dependent, and reaches a plateau at ca. 2 μM edelfosine bulk concentration. The ether lipid can become inserted in phospholipid monolayers with initial surface pressures of up to 33 mN/m, which ensures its capacity to become inserted into cell membranes. Upon interaction with phospholipid vesicles, edelfosine exhibits a weak detergent activity, causing release of vesicle contents to a low extent (< 5%), and a small proportion of lipid solubilization. The weak detergent properties of edelfosine can be related to its very low critical micellar concentrations. Its high affinity for lipid monolayers combined with low lytic properties support the use of edelfosine as a clinical drug. The surface-active properties of edelfosine are similar to those of other “single-chain” lipids, e.g. lysophosphatidylcholine, palmitoylcarnitine, or N-acetylsphingosine.     

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.     

Interactions of cytochrome c and [14C]-carboxymethylated cytochrome c with monolayers of phosphatidylcholine, phosphatidic acid and cardiolipin

1. The interactions between cytochrome c (native and [(14)C]carboxymethylated) and monolayers of phosphatidylcholine, phosphatidic acid and cardiolipin at the air/water interface was investigated by measurements of surface radioactivity, pressure and potential. 2. On a subphase of 10mm-or m-sodium chloride, penetration of cytochrome c into egg phosphatidylcholine monolayers, as measured by an increase of surface pressure, and the number of molecules penetrating, as judged by surface radioactivity, were inversely proportional to the initial pressure of the monolayer and became zero at 20dynes/cm. The constant of proportionality was increased when the cytochrome c was carboxymethylated or decreased when the phospholipid was hydrogenated, but the cut-off point remained at 20dynes/cm. 3. Penetrated cytochrome c could be removed almost entirely by compression of the phosphatidylcholine monolayer above 20dynes/cm. 4. With phosphatidic acid and cardiolipin monolayers on 10mm-sodium chloride the binding of cytochrome c was much stronger and cytochrome c penetrated into films nearing the collapse pressure (>40dynes/cm.). The penetration was partly electrostatically facilitated, since it was decreased by carrying out the reaction on a subphase of m-sodium chloride, and the relationship between the surface pressure increment and the initial film pressure moved nearer to that observed with phosphatidylcholine. 5. Surface radioactivity determinations showed that [(14)C]carboxymethylated cytochrome c was still adsorbed on phosphatidic acid and cardiolipin monolayers after the cessation of penetration. This adsorption was primarily electrostatic in nature because it could be prevented and substantially reversed by adding m-sodium chloride to the subphase and there was no similar adsorption on phosphatidylcholine films. 6. The penetration into and adsorption on the three phospholipid monolayers was examined as a function of the pH of the subphase and compared with the state of ionization of both the phospholipid and the protein, and the area occupied by the latter at an air/water interface. 7. It is concluded that the binding of cytochrome c to phospholipids can only be partially understood by a consideration of the ionic interaction between the components and that subtle conformational changes in the protein must affect the magnitude and stability of the complex. 8. If cytochrome c is associated with a phospholipid in mitochondria then cardiolipin would fulfil the characteristics of the binding most adequately.