Guy J.-M. Lauquin

An appraisal of the functional significance of the inhibitory effect of long chain acyl-CoAs on mitochondrial transports

During the past few years, much attention has been paid to the inhibition by palmityl-CoA of some cellular enzymes and of mitochondrial transport systems. The particular interest of this inhibition is the possible role of acyl-CoA-sensitive enzymes in the mechanisms regulating lipogenesis and glucogenesis (for review see refs. [ l-3] . The inhibition by palmityl-CoA of the citrate carrier in mitochondria [4] has also been discussed in terms of a possible regulation of fatty acid synthesis. It is conceivable that the rate of citrate exit from mitochondria may control the rate of acetyl-CoA formation through the activity of the citrate-cleavage enzyme located in the cytosol, but it must be noted that other transport systems, such as the ADP carrier [3,5,6] and the dicarboxylate carrier [4] are also sensitive to palmityl-CoA (or oleyl-CoA). The phosphate transport appears to escape this inhibition [6] . Although a number of papers have been published on this subject, the majority lack the quantitation necessary for an adequate evaluation of the physiological significance of the inhibitory properties of long chain acyl-CoAs. In this paper, we compare the values of the inhibitor constant (Ki) for long chain acyl-CoAs of mitochon-

A constitutive role for GPI anchors in Saccharomyces cerevisiae : cell wall targeting

GPI anchors are widely represented among organisms and have several cellular functions. It has been proposed that in yeast there are two groups of GPI proteins: plasma membrane‐resident proteins, such as Gas1p or Yap3p, and cell wall‐targeted proteins, such as Tir1p or α‐agglutinin. A model has been proposed for the plasma membrane retention of proteins from the first group because of a dibasic motif located just upstream of the GPI‐anchoring signal. The results we report here are not in agreement with such a model as we show that constructs containing the C‐terminal parts of Gas1p and Yap3p are also targeted to the cell wall. We also detect the genuine Gas1p after cell wall treatment with Quantazyme or Glucanex glycanases. In addition, we show that the GPI‐anchoring signal from the human placental alkaline phosphatase (PLAP) is not compatible with the yeast machinery unless the human transamidase hGpi8p is co‐expressed. In this condition, this human signal is able to target a protein to the cell wall. Moreover, TIR1 proved to be a multicopy suppressor of Δgas1 mutation. The present findings suggest a constitutive role for GPI anchors in yeast: the cell wall targeting of proteins.

Regulation by low temperatures and anaerobiosis of a yeast gene specifying a putative GPI-anchored plasma membrane

Expression of the yeast Saccharomyces cerevisiae SRP1 (serine‐rich protein) gene is shown here to be induced both‐ by low temperature and anaerobic growth conditions. We show that anaerobic SRP1 expression is haem‐dependent; however, haem influence does not operate through the action of the hypoxic‐gene ROX1 repressor. The SRP1 promoter region displaying the stress‐responsive elements is restricted to its first 551 bp, upstream of the initiation codon, although an upstream activation site contained in upstream sequences is required for full promoter activity. In addition, we demonstrate that the TIP1 gene, sharing similar nucleotide and polypeptide structure with SRP1, and previously reported to be a cold‐shock‐inducible gene, is also a hypoxic gene. Srp1 protein production is similarly induced by low temperature and anaerobic growth conditions. This protein, detected in the plasma membrane fraction, is shown to be exposed on the cell surface via a glycosyl‐phosphatidylinositol membrane anchoring.

Interaction of azidonitrophenylaminobutyryl--ADP, a photoaffinity ADP analog, with mitochondrial adenosine triphosphatase. Identification of the labeled subunits.

Two photoactive derivatives of ATP, 8-azido-ATP [l] and azidonitrophenylaminopropionyl-ATP (NAP-ATP) [2] have been recently used to label F, -ATPase by photoaffinity. In the first compound, the azido group is located on the adenine ring of ATP. In the second one, the photosensitive adduct is attached to the ribose portion of ATP. Wagenvoord et al. [l] have reported that upon photoactivation, 8-azido-ATP binds covalently to the P-subunit of F,-ATPase. Russell et al. [l] have shown that NAPaATP also binds to F, -ATPase, but there were no data on the nature of the subunit(s) which carry the binding site(s) and no documentation on possible side effects due to the presence of an aryl group in the molecule. In this paper, we report on the interactions of N-Cazido-2-nitrophenylaminobutyryl-ADP (NAPa-ADP) (fig.1) with isolated F1 -ATPase and on the identification of the subunits which bind NAP, -ADP. For this purpose NAP4 -ADP was synthesized in radioactive form. When irradiated under visible light in the presence of F, -ATPase, NAP4 -ADP was found to bind covalently to the (Yand P-subunits of F1 -ATPase; this binding was paralleled by a lowering of the hydrolytic activity of F, -ATPase. Evidence is given, which shows that the ADP moiety of NAP,--ADP is responsible for the binding specificity of the whole molecule to FlATPase. pGiy OH OH

Adenine nucleotide transport in sonic submitochondrial particles. Kinetic properties and binding of specific inhibitors

1. A procedure for preparation of sonic submitochondrial particles competent for adenine nucleotide transport is described. ADP or ATP transport was assayed, in the presence of oligomycin, in a saline medium made of 0.125 M KCl, 1 mM EDTA, 10 mM 4-morpholinopropane sulfonic acid buffer, pH 6.5. 2. Sonic particles transport ADP and ATP by an exchange diffusion process. Externally added ADP (or ATP) is exchanged with internal ADP and ATP with a stoichiometry of one to one. The V value for ADP transport 5 degrees C was between 2 and 3 nmol/min per mg protein. 3. The transport system in sonic particles is specific for ADP and ATP. It is strongly dependent on temperature. The activation energy between 0 and 9 degrees C is approx. 35 kcal/mol. The optimum pH is 6.5, 4, Like in intact mitochondria, externally added ADP is transported into sonic particles faster at a given concentration than externally added ATP. The V value for ADP transport is 1.5-2 times higher than the V value for ATP transport. 5. The transition from the energized to the deenergized state in sonic particles results in a decrease of the pH gradient across the membrane (internal pH less than external pH) and in a 2-4 fold increase in the Km value for ATP. This latter effect is opposite that found for transport of added ATP in intact mitochondria (Souverijn, J.H.M., Huisman, L.A., Rosing J. and Kemp, Jr., A. (1973) Biochim. Biophys. Acta 305, 185-198). Energization has no effect on the V value of ATP transport in sonic particles. 6. In contrast to intact mitochondria, inhibition of ADP transport in sonic particles by bongkrekic acid does not have any lag-time and does not depend on pH. The inhibition caused by bongkrekic acid is a mixed type inhibition with a Ki value of 1.2 micronM. Atractyloside and carboxyatractyloside do not inhibit ADP transport in sonic particles, unless the particles have been preloaded with these inhibitors during the sonication. 7. Palmityl-CoA added to sonic particles inhibits efficiently ADP transport. The mixed type inhibition found with palmityl-CoA has a Ki value of 1.6 micronM. 8. [3H]Bongkrekic acid binds to sonic particles readily and with high affinity. Bongkrekic acic binding to sonic particles does not depend on pH and it has a saturation plateau, corresponding approximately to 1.3 mol of site per mol of cytochrome a. The number of [3H]atracytloside binding sites is much lower (one-fifth of the bongkrekic acid). External carboxyatractyloside does not compete with [3H]bongkrekic acid for binding to sonic particles. However, when carboxyatractyloside is present inside the particles, it inhibits the binding of [3H]bongkrekic acid.

Comparison of mitochondrial cationic channels in wild‐type and porin‐deficient mutant yeast

Bilayers were formed at the tip of microelectrodes from a suspension of proteoliposomes derived from wild‐type and porin‐deficient mutant yeast mitochondria. In both preparations, identical cationic channels of large conductance were recorded. This result rules out any relationship between this channel and the outer membrane voltage‐dependent anion channel, the activity of which is carried by porin. The ionic selectivity and the voltage‐dependence of the yeast cationic channel suggest that it is related to that recently described in mammalian mitochondria. This hypothesis is further supported by the fact that both channels are blocked by a mitochondrial addressing peptide.

The respiration of cells and mitochondria of porin deficient yeast mutants is coupled

Several mutants of yeast lacking the porin gene have been found stable and viable on glucose or glycerol media. Ethanol-supported respiration of porin-free mutant and wild cells appeared equally coupled in vivo being similarly depressed by inhibitors of ADP/ATP translocase or of ATP synthase and stimulated by the uncoupler FCCP. The absence of porin in isolated mutant mitochondria hardly impaired the electron flux but increased the requirement for Mg2+ (or Ca2+) and for ADP and carboxyatractylate concentrations necessary to drive effectively state 3 - state 4 and state 4 - state 3 transitions, respectively. The existence of another porin species, possibly controlled by bivalent cations, is postulated.

PMT1 MANNOSYL TRANSFERASE IS INVOLVED IN CELL WALL INCORPORATION OF SEVERAL PROTEINS IN SACCHAROMYCES CEREVISIAE

We constructed hybrid proteins containing a plant α‐galactosidase fused to various C‐terminal moieties of the hypoxic Srp1p; this allowed us to identify a cell wall‐bound form of Srp1p. We showed that the last 30 amino acids of Srp1p, but not the last 16, contain sufficient information to signal glycosyl‐phosphatidylinositol anchor attachment and subsequent cell wall anchorage. The cell wall‐bound form was shown to be linked by means of a β1,6‐glucose‐containing side‐chain. Pmt1p enzyme is known as a protein‐O‐mannosyltransferase that initiates the O‐glycosidic chains on proteins. We found that a pmt1 deletion mutant was highly sensitive to zymolyase and that in this strain the α‐galactosidase–Srp1 fusion proteins, an α‐galactosidase–Sed1 hybrid protein and an α‐galactosidase–α‐agglutinin hybrid protein were absent from both the membrane and the cell wall fractions. However, the plasma membrane protein Gas1p still receives its glycosyl‐phosphatidylinositol anchor in pmt1 cells, and in this mutant strain an α‐galactosidase–Cwp2 fusion protein was found linked to the cell wall but devoid of β1,6‐glucan side‐chain, indicating an alternative mechanism of cell wall anchorage.

A covalent tandem dimer of the mitochondrial ADP/ATP carrier is functional in vivo.

The adenine nucleotide carrier, or Ancp, is an integral protein of the inner mitochondrial membrane. It is established that the inactive Ancp bound to one of its inhibitors (CATR or BA) is a dimer, but different contradictory models were proposed over the past years to describe the organization of the active Ancp. In order to decide in favor of a single model, it is necessary to establish the orientations of the N- and C-termini and thus the parity of the Ancp transmembrane segments (TMS). According to this, we have constructed a gene encoding a covalent tandem dimer of the Saccharomyces cerevisiae Anc2p and we demonstrate that it is stable and active in vivo as well as in vitro. The properties of the isolated dimer are strongly similar to those of the native Anc2p, as seen from nucleotide exchange and inhibitor binding experiments. We can therefore conclude that the native Anc2p has an even number of TMS and that the N- and C-terminal regions are exposed to the same cellular compartment. Furthermore, our results support the idea of a minimal dimeric functional organization of the Ancp in the mitochondrial membrane and we can suggest that TMS 1 of one monomer and TMS 6 of the other monomer in the native dimer are very close to each other.

Photoaffinity labeling of the adenine nucleotide carrier in heart and yeast mitochondria by an arylazido ADP analog

Abstract 1. 1. Arylazido analogs of ADP and ATP ( N -4-azido-2-nitrophenyl-aminobutyryl-ADP and N -4-azido-2-nitrophenylaminobutyryl-ATP have been prepared in radioactive form and used in photolabeling experiments to identify the adenine nucleotide carrier in mitochondria and sonic submitochondrial particles. 2. 2. When added in the dark to beef heart mitochondria, azidonitrophenyl-aminobutyryl-ADP binds to the adenine-nucleotide carrier. It is not transported across the membrane to the matrix space, but it inhibits ADP transport in mito-chondria. The inhibition is of a mixed type with a K i value of about 10 μM. 3. 3. The nitrene derivative formed upon photoirradiation of tritiated azidonitrophenylaminobutyryl-ADP or -ATP binds to a polypeptide of apparent molecular weight 30 000 in beef heart mitochondria and 37 000 in Saccharomyces cerevisiae mitochondria. The photolabeling is prevented by preincubation of the mitochondria with atractyloside or carboxyatractyloside. 4. 4. Photoirradiation of sonic submitochondrial particles from beef heart (inside-out particles) with tritiated azidonitrophenylaminobutyryl-ADP or -ATP results in the labeling of the 30 000-dalton polypeptide and also in the labeling of higher molecular weight peptides (50 000–55 000) probably belonging to F 1 -ATPase. Addition of bongkrekic acid specifically decreases the photolabeling of the 30 000-dalton polypeptide. 5. 5. An arylazido derivative of atractyloside ( N -4-azido-2-nitrophenylaminobutyryl atractyloside) binds upon photoirradiation to the 30 000-dalton polypeptide in beef heart mitochondria and to the 37 000-dalton polypeptide in S. cerevisiae mitochondria. 6. 6. Since the adenine nucleotide carrier is readily damaged by ultraviolet light, nitro-arylazido analogs of ADP and ATP or of atractyloside, which are photoactivated in visible light, were used in preference to other azido analogs, which require ultraviolet light for photoactivation. 7. 7. Data presented in this paper support the view that the same mitochondrial protein belonging to the adenine nucleotide transport system is able to bind ADP (or ATP) and atractyloside.