Fas-independent mitochondrial damage triggers cardiomyocyte death after ischemia-reperfusion.

The Fas/Fas ligand and mitochondria pathways have been involved in cell death in several cell types. We combined the genetic inactivation of the Fas receptor (lpr mice), on the one hand, to the pharmacological inhibition of the mitochondrial permeability transition pore (mPTP), on the other hand, to investigate which of these pathways is predominantly activated during prolonged ischemia-reperfusion. Anesthetized C57BL/6JICO (control) and C57BL/6-lpr mice were pretreated with either saline or cyclosporin A (CsA; 40 mg/kg, 3 times a day), an inhibitor of the mPTP, and underwent 25 min of ischemia and 24 h of reperfusion. After 24 h of reperfusion, hearts were harvested: infarct size was assessed by 2,3,5-triphenyltetrazolium chloride staining, myocardial apoptosis by caspase 3 activity, and mitochondrial permeability transition by Ca2+-induced mPTP opening using a potentiometric approach. Infarct size was comparable in untreated control and lpr mice, ranging from 77 +/- 5% to 83 +/- 3% of the area at risk. CsA significantly reduced infarct size in control and lpr hearts. Control and lpr hearts exhibited comparable increase in caspase 3 activity that averaged 57 +/- 18 and 49 +/- 5 pmol x min(-1) x mg(-1), respectively. CsA treatment significantly reduced caspase 3 activity in control and lpr hearts. The Ca2+ overload required to open the mPTP was decreased to a similar extent in lpr and controls. CsA significantly attenuated Ca2+-induced mPTP opening in both groups. Our results suggest that the Fas pathway likely plays a minor role, whereas mitochondria are preferentially involved in mice cardiomyocyte death after a lethal ischemia-reperfusion injury.

Low-pressure reperfusion alters mitochondrial permeability transition.

We hypothesized that low-pressure reperfusion may limit myocardial necrosis and attenuate postischemic contractile dysfunction by inhibiting mitochondrial permeability transition pore (mPTP) opening. Male Wistar rat hearts (n = 36) were perfused according to the Langendorff technique, exposed to 40 min of ischemia, and assigned to one of the following groups: 1) reperfusion with normal pressure (NP = 100 cmH(2)O) or 2) reperfusion with low pressure (LP = 70 cmH(2)O). Creatine kinase release and tetraphenyltetrazolium chloride staining were used to evaluate infarct size. Modifications of cardiac function were assessed by changes in coronary flow, heart rate (HR), left ventricular developed pressure (LVDP), the first derivate of the pressure curve (dP/dt), and the rate-pressure product (RPP = LVDP x HR). Mitochondria were isolated from the reperfused myocardium, and the Ca(2+)-induced mPTP opening was measured using a potentiometric approach. Lipid peroxidation was assessed by measuring malondialdehyde production. Infarct size was significantly reduced in the LP group, averaging 17 +/- 3 vs. 33 +/- 3% of the left ventricular weight in NP hearts. At the end of reperfusion, functional recovery was significantly improved in LP hearts, with RPP averaging 10,392 +/- 876 vs. 3,969 +/- 534 mmHg/min in NP hearts (P < 0.001). The Ca(2+) load required to induce mPTP opening averaged 232 +/- 10 and 128 +/- 16 microM in LP and NP hearts, respectively (P < 0.001). Myocardial malondialdehyde was significantly lower in LP than in NP hearts (P < 0.05). These results suggest that the protection afforded by low-pressure reperfusion involves an inhibition of the opening of the mPTP, possibly via reduction of reactive oxygen species production.

Mitochondrial permeability transition in cardiomyocyte apoptosis during acute graft rejection.

Evidence indicates that acute cardiac graft rejection is associated with cardiomyocyte apoptosis. Mitochondrial permeability transition (MPT) induces apoptotic cell death. We sought to determine whether MPT might play a role in cardiomyocyte apoptosis in the rat model of heterotopic cardiac transplantation. Syngenic and allogenic transplantations were performed, and both native and grafted hearts were harvested 3 or 5 d after transplantation for detection of acute rejection, assessment of Ca(2+)-induced MPT, and myocardial apoptosis by TUNEL staining and caspase 3 activity. Allogenic grafts developed severe acute rejection at day 5 with concomitant cardiomyocyte apoptosis (apoptotic index: 7.1 +/- 1.0% vs. 1.0 +/- 0.2% in syngenic hearts, and caspase 3 activity: 38 +/- 25 vs. 5 +/- 9 nmol/mg, in allogenic vs. syngenic grafts, respectively). At day 5, Ca(2+)-induced MPT was dramatically altered in allogenic when compared with syngenic grafts (mean Ca(2+) overload averaged 0 +/- 20 vs. 280 +/- 30 microM in allogenic and syngenic grafts, respectively). NIM811, a nonimmunosuppressive derivative of cyclosporin A (CsA), that specifically inhibits the MPT pore, did not alter acute rejection, but significantly delayed Ca(2+)-induced MPT pore opening, attenuated caspase 3 activity and cardiomyocyte apoptosis in allogenic grafts. This suggests that mitochondrial permeability transition pore opening may play an important role in cardiomyocyte apoptosis associated with acute cardiac graft rejection.

Palmitic and stearic acids bind Ca2+ with high affinity and form nonspecific channels in black-lipid membranes. Possible relation to Ca2+-activated mitochondrial pores.

A mitochondrial hydrophobic component that forms Ca2+-induced nonspecific ion channels in black-lipid membranes (Mironova et al., 1997) has been purified and its nature elucidated. It consists of long-chain saturated fatty acids--mainly palmitic and stearic. These fatty acids, similar to the mitochondrial hydrophobic component, bind Ca2+ with high affinity in comparison with unsaturated fatty acids, saturated fatty acids with shorter aliphatic chains, phospholipids, and other lipids. Ca2+-binding is inhibited by Mg2+ but not by K+. For palmitic acid, the Kd for Ca2+ was 5 microM at pH 8.5 and 15 microM at pH 7.5, with the Bmax of 0.48 +/- 0.08 mmol/g. This corresponds to one Ca2+ ion for eight palmitic acid molecules. The data of IR spectroscopy confirm that Ca2+ does not form ionic bonds with palmitic and stearic acids under hydrophobic conditions. It has been found that in the presence of Ca2+, palmitic and stearic acids, but not unsaturated FFA induce a nonspecific permeability in black-lipid membranes. Addition of Ca2+ in order to induce the permeability transition, increases the extractable amount of palmitic and stearic acids, the effect being prevented by a phospholipase A2 inhibitor. The possible involvement of palmitic and stearic acids in the mitochondrial nonspecific permeability is discussed.

Calcium-binding properties of the mitochondrial channel-forming hydrophobic component.

A hydrophobic, low-molecular weight component extracted from mitochondria forms a Ca2+-activated ion channel in black-lipid membranes (Mironova et al., 1997). At pH 8.3-8.5, the component has a high-affinity binding site for Ca2+ with a Kd of 8 x 10(-6) M, while at pH 7.5 this Kd was decreased to 9 x 10(-5) M. Bmax for the Ca2+-binding site did not change significantly with pH. In the range studied, 0.2 +/- 0.06 mmol Ca2+/g component were bound or one calcium ion to eight molecules of the component. The Ca2+ binding was strongly decreased by 50-100 mM Na+, but not by K+. Treatment of mitochondria with CaCl2 prior to ethanolic extraction resulted in a high level of Ca2+-binding capacity of the partially purified component. Cyclosporin A, a specific inhibitor of the mitochondrial permeability transition, when added to the mitochondrial suspension, decreased the Ca2+-binding activity of the purified extract severalfold. The calcium-binding capability of the partially purified component correlates with its calcium-channel activity. This indicates that the channel-forming component might be involved in the permeability transition that stimulates its formation.

Sub-mitochondrial localization of the catalytic subunit of pyruvate dehydrogenase phosphatase.

Using a specific antibody against the PDP catalytic subunit, PDPc, precise localization of this subunit in mitochondria was performed. Sub-fractionation of purified mitochondria by controlled swelling processes led to the isolation of outer membranes, matrix space and inner membrane vesicles which were purified on a sucrose density gradient. In this study, we demonstrated that PDPc was not recovered as a soluble protein in the matrix space but was associated with the inner membrane. Moreover, Triton X-114 phase partitioning performed on inner membranes showed that PDPc behaved both as a hydrophilic and as a hydrophobic protein, thus suggesting two different forms of this enzyme.

Oscillating Ca2+-induced channel activity obtained in BLM with a mitochondrial membrane component.

Oscillations in ion fluxes and membrane potential may be observed in cells and in mitochondria as well. We obtained Ca2+-induced oscillations in channel activity in black-lipid membranes reconstituted with hydrophobic components extracted from mitochondria. Mitoplasts prepared from purified rat liver mitochondria were extracted with ethanol followed by Folch extraction and further partial purification by silicic acid chromatography. Channel activity was measured in lipid bilayers formed from bovine brain lipids and 10% cardiolipin with addition of the purified fractions. The conductance with 10 mM Ca2+ was 100 pS or its multiples. Ca2+ gradients of 4: 1 induced oscillating channel activity for several hours, with initial open states of 40 s and closed states of 56 s; the open times gradually decreasing to 8.6 s. No channel activity was seen without added fractions. The channel activity was associated with a Ca2+-binding lipid, nonpolar, low-molecular-weight fraction that in gel electrophoresis was not stained with Coomassie Blue and did not contain carbohydrate-staining material. 1H-Nuclear magnetic resonance spectra of the substance showed the presence of aliphatic chains and carbonyls, but the detailed structure remains to be elucidated.

The 8 kD product of the putative oncogene MTCP-1 is a mitochondrial protein.

An unusually small (8 kD) protein (p8MTCP-1) is coded by the putative oncogene MTCP-1 (also called c6.1B), involved in the translocation t(X;14)(q28;q11) associated with some mature T-cell proliferations. Here, we show by subcellular fractionation and by confocal microscopy that this protein is located in the mitochondria. This localization orientates toward a role of p8MTCP-1 in the mitochondrial metabolism which may be relevant for the oncogenic process.

Further characterization of mitochondrial outer membrane: evidence for the presence of two endogenous sialylated glycoproteins.

The distribution of sialic acid-containing glycoproteins was investigated in highly purified mitochondrial membranes using labeled Sambucus nigra agglutinin as a detection system. Two sialylated glycoproteins were shown to be true components of the mitochondrial outer membrane. Relative to monoamine oxidase activity, these glycoproteins were found to be preferentially located in the "free" outer membrane fraction. As sialic acid is thought to be involved in molecular recognition, a role for these glycoproteins in mediating the interactions between mitochondria and other sub-cellular organelles is considered.

Involvement of mitochondrial contact sites in the subcellular compartmentalization of phospholipid biosynthetic enzymes.

The concerted synthesis of phospholipids derived from serine involving two microsomal enzymes (phosphatidylserine synthase and phosphatidylethanolamine N-methyltransferase) and a mitochondrial one (phosphatidylserine decarboxylase) occurs in reconstituted cell-free systems. Subfractionation of crude mitochondria after swelling and separating on a sucrose density gradient resulted in the isolation of two contact site-enriched fractions from total outer membranes and inner membranes, respectively. Estimation of marker enzyme activities shows a high recovery of glucose-6-phosphate phosphatase (a marker for the endoplasmic reticulum) associated with contact site-enriched fractions. Accordingly, the linked synthesis of phosphatidylserine, phosphatidylethanolamine, and at a lesser extent phosphatidylcholine can occur. This biosynthetic pathway was absent from purified contact site-enriched fractions correlative with the absence of glucose-6-phosphate phosphatase activity. Reconstitution experiments, including contact site-enriched fractions incubated with endoplasmic reticulum-rich fraction, led to the restoration of the linked synthesis of phospholipids, thereby demonstrating that a reversible association between these two fractions can occur. These functional interactions between the endoplasmic reticulum and mitochondria are confirmed at the ultrastructural level using either chemical or physical fixation before resin embedding. These results show that the interorganelle trafficking of lipids may involve only highly specialized microdomains of both membranes, thereby allowing the maintenance of a specific lipid composition and distribution within membranes.

Further evidence for both functional and structural microcompartmentation within the membranes of two associated organelles, mitochondrion and endoplasmic reticulum.

We previously demonstrated that the translocation of phospholipids between the mitochondrion and the endoplasmic reticulum occurs via highly specialized membrane microdomains of both organelles that are in situ closely associated. As understanding of the interactions between both organelles requires characterization of the translocation sites organization, we first analysed the amino acid compositions of these sites. Using principal component analysis, we have shown that the translocation sites exhibit characteristic patterns when compared with the membranes from which they are derived. The results are discussed in terms of both functional and structural microcompartmentation within the membranes of mitochondria and endoplasmic reticulum.

Use of Percoll gradients for isolation of human placenta mitochondria suitable for investigating outer membrane proteins.

Human mitochondria were isolated from placenta by a combination of differential and Percoll gradient centrifugation, resulting in a highly pure and intact preparation as assessed by marker enzyme analysis and electron microscopy. The advantages over previous methods are the rapidity of the procedure and the excellent resolution of mitochondria and lysosomes. Moreover, the high extent of intactness of the mitochondria so obtained made them particularly well suited for investigating outer membrane proteins. Taking advantage of this method, we have purified human mitochondrial porin. The purified protein consists of a single unglycosylated polypeptide of molecular mass 33 kDa.

Mitochondrial dolichyl-phosphate mannose synthase. Purification and immunogold localization by electron microscopy.

Mitochondrial dolichyl-phosphate mannose synthase has been purified to homogeneity using an original procedure, reconstitution into specific phospholipid vesicles and sedimentation on a sucrose gradient as final step. The enzyme has an apparent molecular mass of 30 kDa on an SDS/polyacrylamide gel. Increased enzyme activity could be correlated with this polypeptide band. A specific antibody was raised in rabbits against this transferase. Specific IgG obtained from the immune serum removed enzymatic activity from a detergent extract of mitochondrial outer membrane and reacted specifically with the 30-kDa band on immunoblots. Furthermore, an immunocytochemical experiment proved the localization of dolichyl-phosphate mannose synthase on the cytosolic face of the outer membrane of mitochondria.

Sialylation processes in mitochondria: evidence for two distinct sialyltransferases located in the outer membrane.

Previous studies have shown that purified mitochondrial outer membrane is able to catalyze the transfer of sialic acid from CMP-Neu5Ac to an exogenous asialoglycoprotein acceptor, asialofetuin. Considering the heterogeneity of the glycan chains borne by this glycoprotein, an investigation of mitochondrial sialyltransferase activities was undertaken. Our data provide evidence for the existence of two distinct sialyltransferases in purified mitochondrial outer membranes. The use of different acceptor substrates, the temperature dependence of these enzymes, and their different sensitivity towards a sulfhydryl reagent, p-CMB, allowed us to discriminate between a galactoside alpha(2-3) sialyltransferase and a galactoside alpha(2-6) sialyltransferase presumably involved in the sialylation of O- and N-glycan chains of glycoprotein, respectively. These results are discussed in terms of mitochondrial autonomy for post-translational events.

Investigation of glycosylation processes in mitochondria and microsomal membranes from human skeletal muscle.

Glycoconjugates are directly involved in major skeletal muscle functions. As little is known about glycosylation processes in muscle, we investigated glycoconjugate synthesis in subcellular fractions from human skeletal muscle tissue. Mitochondria and microsomal membranes were prepared from muscle biopsies by thorough mechanical disruption and differential centrifugations. This procedure resulted in the isolation of intact mitochondria (1 mg protein/g muscle) and of a microsomal fraction (1.5 mg protein/g muscle). Glycosyltransferases were studied in both subcellular fractions using either dolichylmonophosphate as a polyprenic acceptor or chemically modified fetuin as a glycoprotein substrate. Our results provide evidence for high rates of glycosylation in muscle. The highest activities were obtained with GDP-mannose: dilichylmonophosphate mannosyltransferase, a key enzyme in glycosylation process (220 pmol/mg per h in mitochondria and 1,550 pmol/mg per h in microsomal membranes). Substantial individual variations were observed for dolichol pathway glycosyltransferases but low individual variations were found for glycosyltransferases involved in maturation of glycoproteins. The role which glycosylation defects may play in muscle dysfunction has yet to be defined.

Triggering of mannosyltransferase activity in inner mitochondrial membranes by dolichyl-monophosphate incorporation mediated through phospholipids or fatty acids.

The activity of GDPmannose:dolichyl monophosphate mannosyltransferase in inner mitochondrial membranes can be triggered by dolichyl-monophosphate incorporation mediated through phospholipids or fatty acids. The efficiency of this incorporation and the efficiency of the enzyme activity are not equivalent. Among a variety of amphiphiles which were tested, the highest mannosyltransferase activity was obtained with the mixture of lipids extracted from the outer mitochondrial membranes. The results presented here appear consistent only with a mechanism involving collisional contacts of the phospholipid vesicles and fusion with the membranes. ESR spectroscopy confirms that (a) the incorporation process is followed by solubilization of dolichyl monophosphate molecules in the lipid phase and (b) the general organization of the inner mitochondrial membranes is not perturbed by the addition of dolichyl monophosphate.

Organization of mitochondrial UDP-glucose:dolichylmonophosphate glucosyltransferase in the outer membrane.

Previous studies have shown the existence of an autonomous mitochondrial UDP-glucose: dolichylmonophosphate glucosyltransferase, located in mitochondrial outer membrane of liver cells. To improve our knowledge about the topographical aspects of glycosylation in mitochondria, we have investigated the organization of this enzyme in intact mitochondria, using controlled proteolysis with trypsin and sensitivity towards amino-acid specific reagents. Our data provides evidence: --for a mitochondrial glucosyltransferase facing the cytoplasmic side of the outer membrane --and for the involvement of histidine and tryptophan residues as well as sulfhydryl groups in the catalytic activity of the enzyme.

Glycosyltransferase activities in normal and leukaemic monocytic cells.

Glycosyltransferase activities were measured in normal monocytes and in leukaemic monoblasts. Biosynthesis of glycosylated derivatives of dolichyl-monophosphate, which act as intermediates in glycosylation, was measured. Transfer of mannose from GDP-mannose was greatly increased in leukaemic monoblasts. Galactosyltransferase activities, using endogenous protein acceptors, were increased in leukaemic cells of the monocytic lineage compared to normal cells. No significant difference was observed on specific exogenous glycoprotein acceptors. These selective increases of some glycosyltransferase activities in normal and leukaemic monocytic cells can be correlated either with different expression of specific carbohydrate structures or with changes in glycosylation regulation.

Topological orientation of mitochondrial GDPmannose: dolichyl-monophosphate mannosyltransferase in the outer membrane.

Previous studies have shown the existence of an autonomous mitochondrial GDPmannose:dolichylmonophosphate mannosyltransferase, located in mitochondrial outer membrane of liver cells. As nothing is known about glycosylation sites in mitochondria, we have investigated the topological orientation of this enzyme in intact mitochondria, using controlled proteolysis with trypsin. Mitochondria were purified sequentially by mild ultrasonic treatment and sucrose density gradient. Purity and homogeneity of mitochondrial fraction were assessed by electron microscopy and specific marker enzymes measures. Our data provide evidence for a mitochondrial GDPmannose:dolichylmonophosphate mannosyltransferase facing the cytoplasmic side of the outer membrane. However, the exposure of this enzyme to the water phase has been shown to be dependent on the ionic strength of the environment.

Glucosyltransferase activity in mitochondria. Biosynthesis of glucosyl-phosphoryl-dolichol in inner mitochondrial membranes.

1. Inner mitochondrial membranes are able to transfer [14C]glucose from UDP-[14C]glucose onto dolichylmonophosphate. 2. Synthesis of dolichyl-phosphoryl-glucose takes place only in the presence of exogenous dolichyl-monophosphate loaded into phospholipid vesicles. 3. Neutral phospholipids interact preferentially with the membrane-bound enzyme. The effect of phospholipids is not related to the length of fatty acid chains but a correlation between the activation and the degree of unsaturation of fatty acid chains has been found. 4. This enzyme required divalent cations for activity. Such a requirement might be related to lipid-protein interactions which favour a suitable conformation of glycosyltransferase.