Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

The membrane of peroxisomes in Saccharomyces cerevisiae is impermeable to NAD(H) and acetyl‐CoA under in vivo conditions.

The membrane of peroxisomes in Saccharomyces cerevisiae is impermeable to NAD(H) and acetyl‐CoA... EMBO Journal vol.14 no. 14 pp.3480-3486, 1995 The of in Saccharomyces The membrane peroxisomes is impermeable to NAD(H) and acetyl-CoA cerevisiae under in vivo conditions in steps of the biosynthesis of ether-linked phospholipids Carlo W.T.van Roermund, Ype Elgersma1, in while mammalian cells take place peroxisomes synthesis Ronald J.A.Wanders and Neena Singh2, reticulum. This involves is completed in the endoplasmic Henk F.Tabak"'3 export of the intermediate alkyl-dihydroxyacetone of and Academic membrane the Departments Clinical Biochemistry Pediatrics, phosphate across the single bounding per- University of Amsterdam, Meibergdreef 9, Medical Centre, Bosch et A related micro- oxisome (Van den al., 1992). of E.C.Slater 1105 AZ Amsterdam, 'Department Biochemistry, the of body-like organelle, glycosome trypanosomes, of 15, Institute, University Amsterdam, Meibergdreef the contains the major part of glycolytic pathway, implying 1105 AZ Amsterdam, The Netherlands and 2Institute of Pathology, Reserve Cleveland, OH 44106, USA that dihydroxyacetone phosphate, glycerol-3-phosphate, Case Western University, should be able to 3-phosphoglycerate and cofactors pass 3Corresponding author and the glycosomal membrane (Opperdoes Borst, 1977). should be considered C.W.T.van Roermund, Y.Elgersma and N.Singh of such metabolites across the However, how transfer as first authors equal is still a matter of peroxisomal membrane takes place One school of debate (reviewed by Borst, 1989). thought NADH We investigated how generated during peroxiso- to low molecular is that peroxisomes are freely permeable is reoxidized to and how the mal ,8-oxidation NAD+ the behavi- weight compounds. This was concluded from is end product of ,B-oxidation, acetyl-CoA, transported our of equilibrium density gradient peroxisomes upon to mitochondria in Saccharomyces from peroxisomes that several in sucrose and the finding centrifugation of the malate cerevisiae. Disruption peroxisomal dehy- such as D-amino acid oxidase, glycolate oxidase enzymes 3 resulted in (- drogenase gene (MDH3) impaired and urate oxidase failed to exhibit structure-linked latency oxidation as measured in intact cells, whereas capacity Duve and 1966). Direct permeability meas- (de Baudhuin, in In (-oxidation was normal cell lysates. perfectly urements clamp analysis provided evidence using patch cells were unable to grow addition, mdh3-disrupted in favour of this (Van Veldhoven et al., 1987). concept non-fermentable on oleate whereas on other growth The other opinion holds that this permeability observed that MDH3 is carbon sources was normal, suggesting in vitro is a result of their isolation and that peroxisomes of NADH involved in the reoxidation generated during in vivo are closed compartments. This concept finds rather than as fatty acid 3-oxidation functioning part the observation that peroxisomes in Hansenula support by the To the of of glyoxylate cycle. study transport acetyl have an acidic interior which implies restricted polymorpha units from we the peroxisomal peroxisomes, disrupted of the membrane toward protons permeability peroxisomal citrate gene (CIT2). The lack of phenotype of synthase (Nicolay et al., 1987; Waterham et al., 1990). In addition, the cit2 mutant indicated the presence of an alternative the observed latency of glycosomal enzymes suggests a formed the pathway for transport of acetyl units, by barrier for substrates and permeability phosphorylated carnitine acetyltransferase protein (YCAT). Disruption cofactors involved in trypanosomal glycolysis (Opperdoes blocked the (3- of both the CIT2 and YCAT gene and Borst, 1977). in in data oxidation intact cells, but not lysates. Our A solution to overcome a membrane barrier is to use suggest that the peroxisomal membrane is strongly as in mitochondria (reviewed by Walker to NAD(H) and acetyl-CoA in vivo, and shuttle-systems, impermeable and Runswick, 1993). Here, transport of reducing equiva- predict the existence of metabolite carriers in the lents from the cytosol to mitochondria is mediated by the peroxisomal membrane to shuttle metabolites from vice glycerol-3-phosphate/dihydroxyacetone phosphate shuttle peroxisomes to cytoplasm and versa. words: (-oxidation/carnitine (Zebe et al., 1959) or the malate/aspartate shuttle (Borst, Key acetyltransferase/citrate cycle/malate dehydrogenase whereas the transport of acetyl-CoA is mediated synthase/glyoxylate 1963), by an acetylcarnitine shuttle (Bieber, 1988). If similar shuttles are operative in peroxisomes it predicts the existence of a set of specific enzymes that participate in Introduction these exchange processes. Peroxisomes are essential subcellular organelles involved Here we have re-investigated the issue of peroxisome in a variety of metabolic processes. Their importance is permeability using a genetic approach to study how underlined by the recognition of an increasing number of the end of the (-oxidation of fatty acids in products inherited diseases in man in which one or more peroxiso- Saccharomyces cerevisiae, acetyl-CoA and reducing mal functions is impaired (Wanders et al., 1988; Moser, for further equivalents (NADH), leave the peroxisome 1991; Van den Bosch et al., 1992). metabolism in the cytosol and mitochondria. Our results For most of the enzymatic pathways delineated so far, indicate that peroxisomes are impermeable to these com- peroxisomes are dependent on efficient communication pounds and that specific shuttles are required to facilitate with the remainder of the cell. For instance, the first two across the peroxisomal membrane. transport 38) Oxford University Press Peroxisome impermeability to NAD(H) and acetyl-CoA Wild-type mdh3 mutant 1- [ .~~~ r. . . I AZ ', - Fig. 1. Subcellular location of malate dehydrogenase in S.cerevisiae. An organellar pellet was obtained by subcellular fractionation of oleate-grown cells and used for density gradient centrifugation on Nycodenz. Fraction 1 presents the bottom fraction, fraction 20 the top fraction. Succinate dehydrogenase and 3-hydroxyacyl-CoA dehydrogenase were measured as mitochondrial and peroxisomal markers, respectively. (A) Gradient of wild- type cells; (B) gradient of Amdh3 cells. Results 1 2 3 4 dehydrogenase 3 is present in Malate peroxisomes Transport of reducing equivalents from cytosol to mito- chondria in higher eukaryotes, has long been known to be mediated by the glycerol-3-phosphate/dihydroxyacetone phosphate shuttle and the malate/aspartate shuttle. In /MVHD analogy to the latter shuttle, a candidate enzyme for the reoxidation of NADH in peroxisomes of S.cerevisiae is the malate dehydrogenase enzyme (MDH). Earlier studies have revealed the existence of three MDH isozymes in S.cerevisiae (McAllister and Thompson, 1987; Minard and McAllister-Henn, 1991; Steffan and McAllister-Henn, 1992). The C-terminus of MDH3 ends in SKL (Steffan McAllister-Henn, 1992) which is an established and targeting signal (PTS) (Gould et al., 1989). peroxisomal Actin the presence of MDH3 in peroxisomes has not, However, as been demonstrated. yet, We disrupted the MDH3 gene and tested whether this resulted in the absence of malate dehydrogenase in the fractions. Therefore, cells were grown in peroxisomal 2. Northern-blot analysis of MDH3 expression. Cells were grown Fig. a medium containing oleate, a well-known inducer of on medium containing glucose (1), glycerol (2), oleate (3) or acetate followed by subcellular fractionation and peroxisomes, as sole carbon source. Total RNA (10 was used for each lane (4) gg) onto nitrocellulose, the filters were of the agarose gel. After blotting centrifugation of the organellar pellet. density gradient MDH3 gene or with the actin gene as a with the radiolabelled probed The results presented in Figure 1 show good resolution control. between peroxisomes and mitochondria as exemplified by of activity of succinate dehydrogenase the distinct profiles induction of MDH3 mRNA by growth and 3-hydroxyacyl-CoA dehy- showed profound mitochondrial marker) (a whereas expression of the actin gene (used as marker). Importantly, malate on oleate, drogenase (a peroxisomal is almost constant under the various growth showed a bimodal distribution a control) dehydrogenase activity The observed induction of MDH3 conditions 2). with the peroxisomal and mitochondrial (Figure profile coinciding to that found for the ,-oxidation enzymes. is similar in wild-type cells (Figure IA). No peroxisomal very fractions was observed in the Amdh3 cells (Figure MDH activity is essential for growth on oleate MDH3 encodes the peroxiso- MDH3 that the gene B), indicating for malate dehydro- a role peroxisomal To investigate mal malate dehydrogenase. of NADH, the in reoxidation intraperoxisomal MDH is involved in reoxidation of genase If peroxisomal and Amdh3 strains were com- rates of wild-type one would expect induction of growth intraperoxisomal NADH, either oleate, acetate, ethanol on containing oleate since oleate is known to induce the pared plates activity by as the sole carbon source. Growth of Amdh3 or n-oxidation capacity and thus the production glycerol peroxisomal or was unaffected (not ethanol cells on acetate, glycerol Northern blot analysis indeed of NADH many-fold. 3481 Roermund et C.W.T.van aL the fatty acid. This gives a much better estimate of fatty VWV ird -type acid oxidation than the amount of [14C]CO2 alone, since only part of the acetyl-CoA produced during n-oxidation mdh3 is converted into C02 (Veerkamp et al., 1986). The results of Figure 4 suggest that the impairment in fatty acid 13- psf oxidation in Amdh3 intact cells is caused solely by the absence of malate peroxisomal dehydrogenase, and not by reduced induction or of the activity enzymes involved directly in ,B-oxidation, which include acyl-CoA oxidase, enoyl-CoA hydratase, 3-hydroxyacyl-CoA dehydrogenase Cis/E and 3-ketoacyl-CoA thiolase. Accumulation of 3-hydroxyacyl-CoA intermediates in the Amdh3 mutant ,Vcst If in vivo the block in 1-oxidation is indeed due to the inability to reoxidize peroxisomal NADH in the absence of MDH3, this should be reflected in the accumulation of the 3-hydroxyacyl-CoA ester in the A,ndh3 cells but not 800 - this notion in the in control cells. We tested experiment depicted in Figure 5. Oleate-induced wild-type and Amdh3 cells were incubated for 30 or 60 min with 0, radiolabelled acid. The various labelled fatty acyl-CoA esters, including the ester itself and the acyl-CoA 3- ax,p-unsaturated, hydroxy- and 3-ketoacyl-CoA were extracted and esters, separated on thin The results show that layer plates. significant levels of the ester were 3-hydroxyacyl-CoA found only in Amdh3 cells incubated with radiolabelled for or The fatty acid 30 60 min (Figure 5). observed accumulation of the 3-hydroxyacyl-CoA intermediate in the Amdh3 mutant suggests again that MDH3 participates in a redox shuttle rather than in the glyoxylate cycle. of from Removal acetyl-CoA peroxisomes is the end of the 1-oxidation of Acetyl-CoA product straight-chain fatty acids in An function yeast. important of the glyoxylate is the condensation of two C2 cycle units (acetyl-CoA) to the cell succinate, thereby enabling to form C4 carbon skeletons from C2 units. In 2 30 addition, 4|' 10o te 6zo >81 100n 1 20 14C if peroxisomes are impermeable for (acetyl-) CoA, this Time (hr) cycle is also important for releasing the CoA for continuing 1-oxidation cycles of and for facilitating the transport of of and mutant cells on Fig. 3. Growth wild-type cells oleate medium. carbon units across the Growth on a minimal oleate medium. The peroxisomal membrane. (A) plate containing pas21 mutant in the of is used as (disturbed assembly peroxisomes) control To test whether are peroxisomes indeed impermeable for no Growth curves of and mutant strains on growth. (B) wild-type for we tested the effect on acetyl-CoA, 1-oxidation by rich oleate medium. The strains shown are: cells wild-type (+), disrupting the peroxisomal citrate synthase (CIT2) gene Amdh3 cells (A), Acit2 cells (0), Aycat cells and Acit2/Aycat cells (El) (Lewin et al., 1990). Growth on oleate and As a oxidation of (O). control, wild-type cells were grown on the same medium without Tween/oleate (dashed line). [1-_4C]octanoic acid were investigated as described above for the Amdh3 mutant. The results depicted in Figure 6 shown), but on oleate was growth strongly impaired (second bar) show that oxidation of octanoate was normal (Figure 3), suggesting that malate in cells deficient in peroxisomal dehydro- peroxisomal citrate synthase. Moreover, genase is not taking part in the glyoxylate cycle, but is growth on oleate was indistinguishable from wild-type involved in the 1-oxidation of fatty acids. cells (Figure 3). However, since growth of Acit2 cells is To ascertain whether the inability to grow on oleate is almost normal on ethanol and acetate as well (not shown), caused directly by a block in peroxisomal 1-oxidation, we these results strongly suggest the existence of an efficient studied the oxidation of a 1-_4C-labelled fatty acid bypass that can compensate for the loss of peroxisomal in control (octanoic acid) and Amdh3 mutant cells. As citrate synthase. shown in Figure 4, oxidation of [1-_4C]octanoic acid was One possible bypass route would be the conversion of strongly impaired in the intact Amdh3 cells. Importantly, acetyl-CoA into acetylcarnitine via camitine acetyltrans- fatty acid oxidation was normal in Amdh3 cell-free lysates ferase (CAT) which is known to be present in mitochondria in which the membrane barriers of the different intra- and peroxisomes in higher eukaryotes and yeast (Markwell cellular organelles were absent and NAD+ was present in et al., 1973, 1976; Markwell and Bieber, 1976; Kawamoto excess. The rates of fatty acid oxidation represent the sum et al., 1978; accompanying paper, Elgersma et al., 1995). of and [14C]CO2 water-soluble material after extraction of The acetylcarnitine formed in peroxisomes might sub- 3482 Peroxisome impermeability to NAD(H) and acetyl-CoA *r)f ! _T 0 400 80;i 3 0O 1 0o WT mdh3 LI mdh3 = El WI Fig. 4. Octanoic acid p-oxidation in oleate-induced wild-type cells and Amdh3 cells. (A) p-oxidation in intact cells; (B) J-oxidation in cell lysates. acid oxidation is expressed as the sum of ['4C]C02 and water-soluble n-oxidation products produced. [1-14C]octanoic Time 0 0 30 60 30 60 (Figure 3). Furthermore, growth on (min) not to be affected acetate was normal (not shown). However, ethanol and Cells wt mdh3 wt wt mdh3 mdh3 disrupted both the CIT2 gene and the YCAT when we cells could no longer grow on oleate (Figure 3). gene, the the Acit2/ycat cells were also unable to grow on Since ethanol and acetate (not shown) we conclude that the YCAT protein is indeed indispensable for the net synthesis of C4 carbon units in case the CIT2 gene is deleted. Since the YCAT gene disruption results in the absence of both the peroxisomal and mitochondrial carnitine acetyl- transferase protein (Elgersma et al., 1995) the observed inability of Acit2/ycat cells to grow on oleate may be caused exclusively by the impermeability of mitochondria To investigate directly the permeab- towards acetyl-CoA. of the peroxisomal membrane towards ility properties we investigated the ability of the Acit2/ycat acetyl-CoA mutant to n-oxidize fatty acids as described above double for the Amdh3 mutant. Indeed, similar to what we found 5. TLC analysis of the 14C-labelled products derived from fatty Fig. the Amdh3 mutant, the p-oxidation was practically for acid oxidation in wild-type cells and Amdh3 cells. The products blocked in intact Acit2/ycat mutant cells, whereas the P- formed were analysed after incubating cells 0, 30 and 60 min with oxidation capacity was hardly affected in cell lysates in The right-hand panel shows a marker lane of [1-_4C]palmitate. which the membrane barriers are lost (compare Figure 6A CoA. [1_-4C]3-hydroxypalmitoyl This strongly suggests that there are with B, fourth bar). for the export of acetyl-CoA from two pathways only be transported to mitochondria where it could sequently either via conversion into glyoxylate cycle peroxisomes; be further oxidized to CO2 in the Krebs cycle. However, or via conversion into acetylcarnitine. Con- intermediates in the absence of the CIT2 protein, the citrate (or isocitrate) both pathways leads to the accumula- sequently, blocking mitochondria may also be retrieved for net formed in acetyl-CoA, which depletes the tion of peroxisomal of carbon skeletons in the glyoxylate cycle, as synthesis from free CoA for continued ,-oxidation or peroxisomes in 7. The possibility that peroxisomal and depicted Figure the f-oxidation enzymes by product inhibition inhibits citrate synthase may be able to take over mitochondrial and Osmundsen, 1989). (Hovik other's function has earlier been suggested for each based on deletion studies (Kispal et al., S.cerevisiae gene for the ethylamine-grown yeast Trichosporon 1988), and Discussion based on the enzymatic contents of peroxisomes cutaneum a novel to obtain information on We have used approach et al., 1986). (Veenhuis of peroxisomes in the yeast the permeability properties et al. (1995) demonstrated in the accompany- Elgersma is based on of genes This disruption specific is indeed present S.cerevisiae. that camitine acetyltransferase ing paper involved in the steps for preparatory and mitochondria of oleate-grown S.cerevi- encoding proteins in peroxisomes of metabolites from fatty acid ,B- the generated Both are encoded by the same gene, YCAT. transport siae. enzymes across the membrane. We first oxidation peroxisomal shows that deletion of the YCAT gene results in Figure 6 malate on dehydrogenase concentrated peroxisomal decrease in [1-14C]octanoic acid oxidation in intact a slight cloned and by the which was recently sequenced third bar) and in cell lysates (Figure 6B, (MDH3) cells (Figure 6A, and McAllister-Henn, of McAllister-Henn (Steffan group third However, the growth rate on oleate appeared bar). 3483 C.W.T.van Roermund et aL 1 1.-_I .... I s: ] I .:: .. ..... ....... .. ... ::: :: ::: ........ ........ ...... i i ...... :: :: ... ....... ....... ...... ..... :::. .. .. :: :. .. 1 . 1. * :. 3 4 3 4 1 2 Fig. 6. Octanoic acid ,B-oxidation in oleate-induced wild-type and mutant cells. (A) n-oxidation in intact cells; (B) ,-oxidation in cell lysates. acid oxidation was measured as the sum of [14C]C02 and water-soluble material produced. The bars represent wild-type cells (1), [1-14C]octanoic Acit2 cells (2), cells (3) and Acit2/Avcat cells (4). Aycat acids Fatty Fatty acylCoA Malate Co OTCA NAD + Cycle C°2 MDH3 NADH Oxaloacetate m la + Carnitine Acetyl-CoA + 94 I-- Carnitine --- Acetyl-CoA Carnitine 1% 4' CAT CAT 4, Acetylcarnitine - Acetylcarnitine Acetylcarnitine O- + Oxaloacetate Acetyl-CoA CIT2 4, Oxaloacetate Citrate Citrate e Citrate ._u 4, I ACO ACO Malate Isocitrate socitrate Succinate ->Fumarate Succinate - I Succinate ICLI -?* > ICL1 Acetyl-CoA + Glyoxylate Glyoxylate _.. MLS 4, Malate - Cytosol Mitochondrion Peroxisome for the reoxidation of intraperoxisomal NADH and the pathways for the transport of acetyl-CoA. Our results do not rule out that other Fig. 7. Model malate and oxaloacetate are shuttled between peroxisome and cytosol. Since the presence of aconitase and isocitrate lyase in metabolites than has not been demonstrated unambiguously, these reactions might also take place in the cytosol (bold-dashed arrows). The thin arrows peroxisomes citrate or isocitrate for use in the glyoxylate cycle in case the CIT2 gene has been deleted. retrieval pathway of mitochondrial indicate the supposed CIT2, peroxisomal citrate synthase; ACO, aconitase; ICL, isocitrate lyase; MLS, malate dehydrogenase 3; CAT, camitine acetyltransferase; MDH3, malate synthase. and Thompson, 1987; and to complete impairment of growth on 1992). Earlier studies (McAllister peroxisomes, Minard and McAllister-Henn, 1991) had identified a medium containing oleate as the sole carbon source due In cell lysates mitochondrial (MDH I) and cytosolic malate dehydro- to a peroxisomal 5-oxidation deficiency. reoxidation of NAD+ is not required since this genase (MDH2). Based on the presence of a serine- where tripeptide at the C-terminus, the is in the reaction medium, fatty acid leucine-lysine (SKL) cofactor present to be located in peroxi- oxidation was completely normal. These data strongly MDH3 isoenzyme was speculated somes and McAllister-Henn, 1992). Our results suggest that the peroxisomal membrane is impermeable (Steffan indicate that this is indeed the case. Disruption of the to NAD(H) in vivo, and that malate dehydrogenase is MDH3 gene led to the total absence of MDH activity in involved in regeneration of intraperoxisomal NAD+. In 3484 Peroxisome impermeability to NAD(H) and acetyl-CoA such a shuttle mechanism malate dehydrogenase catalyses variety of different transport-proteins such as carnitine/ the reduction of oxaloacetate to malate with concomitant acetylcarnitine carriers and carriers. dicarboxylate formation of NAD+ from NADH, followed by the shuttling of malate versus oxaloacetate across the peroxisomal Materials and methods membrane. We cannot exclude other shuttles, for example, a malate/aspartate shuttle system. Such a shuttle would Yeast strains and culture conditions require the presence of aspartate aminotransferase protein All the gene disruptions used for this were made in the S.cerevisiae study strain BJ1991 Yeast (MATa, leu2, trpl, ura3-251, prbl-1122, pep4-3). in peroxisomes, which converts aspartate into oxaloacetate. transformants were selected and on minimal medium grown containing We are currently investigating whether this protein is 0.67% yeast nitrogen base without amino acids (YNB-WO)(DIFCO), indeed present in peroxisomes of S.cerevisiae. 2% glucose and amino acids (20 as needed. The media liquid gig/ml) The finding that Amdh3 cells are not impaired in growth used for growing cells for RNA curves or subcellular isolations, growth fractionation contained 0.5% potassium phosphate buffer, pH 0.3% on C2 carbon sources suggests that this enzyme does not 6.0, 2% 2% or 2% K- yeast extract, 0.5% peptone, and or glucose, glycerol, participate in the glyoxylate cycle. Moreover, the enzyme acetate, or 0.12% oleic acid/0.2% Tween-40 as carbon source. Before would then have to operate in two directions in the same one of minimal 0.3% shifting to these media, cells were grown on compartment, which is obviously impossible. Indications medium for at least 24 h. For RNA the cultures were glucose isolations, inoculated at such a that reached = 0.7-1.0 after that the kinetic parameters of glyoxysomal malate dehydro- density they OD6W0 -15 h. Oleic acid contained 0.1% oleic plates acid/0.4% Tween-40, genase are unfavourable to its participation in the glyoxyl- 0.67% base without amino acids yeast nitrogen (YNB-WO)(DIFCO), ate cycle of plant glyoxysomes were earlier reported by 0.1% yeast extract and amino acids as needed. (DIFCO) (20 ,ug/ml) Mettler and Beevers (1980). The consequence of these findings is that malate produced by the glyoxylate cycle Cloning procedures Standard DNA were carried out as described is transported out of the peroxisome followed techniques (Sambrook by retro- et al., 1989). The yeast MDH3 gene was two amplified using oligonucleo- conversion to oxaloacetate in the cytosol (via MDH2) or of and tide primers corresponding to with MDHI regions non-homology the mitochondria (via MDH 1) (Figure 7). MDH2. the Oligonucleotide sequence from 5' end of the was gene (5'- The results described in this paper also provide new The TTTGAATTCAAGCATAAAACAATCAAGG-3'). oligonucleotide of was sequence from the 3' end the gene (5'-GGATCCGATATGAGT- information on the way in which acetyl-CoA is transported A DNA was CAAGATACAAAGG-3'). S.cerevisiae genomic library from the interior of the peroxisome to mitochondria for in reaction the above used as a template a polymerase chain (PCR) using further metabolism in the Krebs cycle. It was a surprise two The PCR reaction was carried out 0.5 of primers. using gg template to observe that disrupting the CIT2 gene for peroxisomal 167 10 mM 0.5 units of DNA, oligonucleotides, dNTPs, Taq gg/ml 1.5 mM 10 mM citrate synthase did not lead to a polymerase (Boehringer Mannheim), Tris-HCl, deficiency to grow on MgC92, 50 mM KCI and 0.01I% BSA in a total volume of 50 The pH 8.4, ,ul. oleate or C2 compounds such as acetate or ethanol. A 1.2 kb was was EcoRI-BamHI annealing temperature 55°C. fragment Since assimilation of C2 compounds requires a functional into site of The obtained and subcloned the multiple cloning pUC19. glyoxylate cycle, this suggests that citrate or isocitrate 600 of the MDH3 EcoRV-NcoI fragment containing bp open reading blunt ended with a 2.2 kb from mitochondria can reach frame was replaced using the cytoplasm (or even the cloning, fragment Linearized DNA with the MDH3 of the LEU2 gene. plasmid disrupted peroxisomes) as has been proposed earlier (Veenhuis et al., was used to transform a strain BJ1991. gene haploid wild-type Leu+ 1986; Kispal et al., 1988). Our results indicate that the were confirmed for of the chromosomal MDH3 transformants disruption carnitine acetyltransferase protein is essential for this the PCR obtained on chromosomal DNA locus by analysing product with the same as used earlier to the bypass (Figure 7). primers amplify gene. The CIT2 deletion was made DNA of gene by isolating genomic The finding that intact cells with either the CIT2 or the cells PSY42-cit2 (Leu2-2, leu2-112, lvs2-801, CI72::URA3) (kindly YCAT gene disruption have almost normal capacity to for PCR and This was used as provided by A.Shyan R.Butow). template oxidize fatty acids whereas 3-oxidation is blocked when and with the CI72 primers (5'-GGATCCATGACAGTTCCTTATCTA-3') PCR The both genes are disrupted, indicates that there are only two (5'-CTATAGTTTGCTTTCAATGTT-3'). resulting fragment and ura+ transformants were was used to transform BJ 1991 cells, ways in which acetyl-CoA can leave the peroxisome: via in the PCR selected for integration CI72 gene by analysis. conversion into glyoxylate cycle intermediates or via was from DNA the 5' CAT- The YCAT gene amplified genomic using conversion into acetylcarnitine by carnitine acetyltrans- and the 3' A primer (5'-TTTGAATTCGAGAACTCTCTCAAAC-3') ferase. Consequently, since 5-oxidation is normal CAT-B primer TCTCCC-3') virtually (5'-TTTCTGCAGCGTAAGCCCTTl-l used for the PCR reaction The oligonucleotides. annealing temperature in cell-free lysates of Acit2/ycat cells (where the membrane 2.1 kb EcoRI-PstI was subcloned was The 55°C. resulting fragment barriers are absent), we conclude that CoA acetyl-CoA (or The of the frame was into pUCl9 (pEL72). major part open reading itself) cannot freely membrane. pass the peroxisomal the 1281 of deleted, by replacing Accl-Bgl2 fragment (containing bp the LEU2 This It establish whether all the CAT has been notoriously difficult to open reading frame) by gene (pEL78). plasmid transform cells and Acit2 was used to (PSY142) wild-type (BJ1991) in of glyoxylate cycle enzymes are located peroxisomes in the YCAT were selected for cells. Leu+ transformants integration gene S.cerevisiae. The presence of isocitrate lyase and aconitase PCR by analysis. in peroxisomes is still a matter of debate, as indicated in Figure 7. However, these uncertainties do not fractionation and compromise Subcellular Nycodenz gradients as described der fractionations were (Van Leij Subcellular the implication we propose here with to the performed regard were used for continuous 16-35% et al., 1992). Organellar pellets impermeability of the peroxisomal membrane to NAD(H) of ml 42% with a cushion Nycodenz gradients (12 ml), Nycodenz and acetyl-CoA. The of the enzymes ,B-oxidation pathway mM mM KCI and 8.5% mM 1 dissolved in 5 6.0, EDTA, MES, pH for 2.5 h in a vertical rotor that produce these and the that sealed tubes were compounds enzymes sucrose. The centrifuged 000 at at 19 000 (MSE 8X35) (29 take care of the in removal of these r.p.m. g) 4°C. preparatory steps MLS I and compounds from peroxisomes (MDH3, CIT2, measurements fl-oxidation in doubt. YCAT) are all localized peroxisomes beyond any with water and in cells were washed phos- resuspended Oleate-grown of our results is in with = 2.5. of 20 of cell One of the predictions that, to analogy saline gl (PBS), Aliquots OD60W phate-buffered acid measurements in 200 were used for membrane must contain a mitochondria, the peroxisomal fatty suspension >-oxidation ,l 3485 C.W.T.van Roermund et al. Gould,S.J., medium containing PBS plus 10 IM [1-14C]palmitate or Keller,G.A., Hosken,N., Wilkinson,J. and Subramani,S. [1-14C]- (1989) J. Cell Biol., 1657-1664. octanoate. Reactions were allowed to proceed for 6 or 12 min at 30°C, 108, 100 of 1.3 M Heikoop,J.C., Van Roermund,C.W.T., followed by termination of reactions by adding ,l perchloric Just,W.W., Ofman,R., acid. Radiolabelled was Schutgens,R.B.H., Heymans,H.S.A., Wanders,R.J.A. and CO2 trapped overnight in 500 of 2 M NaOH. Tager,J.M. gl The were collected after (1990) J. Clin. Invest., 86, 126-130. 14C-labelled 5-oxidation products subsequently extracting the acidified material with as Hovik,R. and Osmundsen,H. (1989) Biochem. J., 263, 297-299. chloroformlmethanol/heptane described before et and in a Kawamoto,S., Ueda,M., Nozaki,C., Yamamura,M., Tanaka,A. and (Heikoop al., 1990) quantified liquid scintillation counter. Fatty acid I-oxidation activities were also measured Fukui,S. (1978) FEBS Lett., 96, 37-40. in cell-free lysates prepared by lysing protoplasts in a medium containing Kispal,G., Rosenkrantz,M., Guarente,L. and Srere,P. (1988) J. Biol. 0.1% Triton X-100, 5 mM MOPS, 1 mM EDTA and I mM Chem., 263, 11145-11149. pH 7.4, PMSF. The cell-free extract was subsequently incubated in reaction Lewin,A.S., Hines,V. and Small,G.M. (1990) Mol. Cell. Biol., 10, mM medium containing the following components: 150 Tris-HCI, pH 1399-1405. Mannaerts,G.P. 8.5, 5 mM ATP, 5 mM MgCl2, 2 mM NaCN, 100,M FAD, I mM and Van Veldhoven,P.P. (1993) Biochimie, 75, 147-158. 1 and Arch. Biochem. NAD, mM CoASH, 0.005% (w/v) Triton X-100, 1 mU/ml acyl-CoA Markwell,M.A.K. Bieber,L.L. (1976) Biophys., 172, 502-509. synthetase (Boehringer and 10 or Mannheim) gM [1-14C]palmitate Markwell,M.A.K., and [1-14C]octanoate. Reactions were allowed to proceed for 6 or 12 min, McGroarty,E.J., Bieber,L.L. Tolbert,N. (1973) J. Biol. Chem., 248, 3426-3442. followed by quantification of [14C]C02 and 14C-labelled :-oxidation as described above. Markwell,M.A.K., Tolbert,N.E. and Bieber,L.L. Arch. Biochem. products (1976) Biophys., 176, 479-488. Identification of intermediates McAllister,L. and Thompson,L.M. (1987) J. Bacteriol., 169, 5157-5166. acyl-CoA In order to the nature of the esters in Mettler,I.J. and Beevers,H. (1980) Plant Physiol., 66, 555-560. identify acyl-CoA accumulating mutant oleate-induced intact cells were incubated with 10 jM Minard,K.I. and McAllister-Henn,L. (1991) Mol. Cell. Biol., 11,370-380. cells, as described for 30 or 60 Reactions were Moser,H.W. (1991) Clin. Biochem., 24, 343-351. [1-_4C]palmitate above, min. terminated 100 tl 1.3 M acid. In order to all Munujos,P., Coll-Canti,J., Gonzalez-Sastre,F. and Gella,F.J. (1993) Anal. by perchloric hydrolyse CoA-esters 100 of 2 M NaOH was added and the Biochem., 212, 506-509. subsequently gl mixture was incubated at 50°C for 30 min. This was followed Nicolay,K., Veenhuis,M., Douma,A.C. and Harder,W. (1987) Arch. by addition of 10 of 0.5 M H2SO4 and 75 of sodium acetate Microbiol., 147, 37-41. ,ul buffer, gl If The pH 6.0. required, pH was adjusted to 4.0. fatty acids were then Opperdoes,F.R. and Borst,P. (1977) FEBS Lett., 80, 360-364. The and Molecular extracted with methanol/chloroformnheptane as described above. Sambrook,J., Fritsch,E.F. Maniatis,T. (1989) Cloning: A taken to under The Laboratory Manual, 2nd Cold lower layer was collected, washed and dryness N2. Ed., Spring Harbor Laboratory Press, in on TLC Cold Spring Harbor, NY. residue was taken up acetone, followed by chromatography conditions as described Bremer and with the et al. Anal. 76-85. using by Wojtczak (1972), Smith,P.K. (1985) Biochem., 150, that benzene was toluene. After the Steffan,J.S. and exception replaced by running, plate McAllister-Henn,L. (1992) J. Biol. Chem., 267, was dried and subjected to The of 24708-24715. autoradiography. recovery fatty acid derivatives during extraction and solvent acetone Van den Bosch,H., Schutgens,R.B.H., Wanders,R.J.A. and Tager,J.M. evaporation, solubilization and TLC chromatography were checked by determining (1992) Annu. Rev. Biochem., 61, 157-198. the recovery of and Van der Leij,I., Van den Berg,M., Boot,R., Franse,M., Distel,B. and [1-'4C]palmitoyl-CoA enzymatically synthesized enoyl-CoA esters and 3-hydroxyacyl-CoA esters from Tabak,H.F. (1992) J. Cell Biol., 119, 153-162. prepared Recoveries were >95%. Van Veldhoven,P.P., Just,W.W. and Mannaerts,G.P. (1987) J. Biol. Chem., [1-14C]palmitoyl-CoA. Radioactively labelled [1-14C]3-hydroxypalmitoyl CoA was 262, 4310-4318. synthe- Veenhuis,M., Van der sized 10 jM for 10 Klei,l.J. and Harder,W. (1986) Arch. Microbiol., enzymatically by incubating [1-14C]palmitate min with 50 U/ml crotonase (Sigma Co., St Louis, USA) and 10 U/ml acyl- 145, 39-50. CoA oxidase in a reaction as Veerkamp,J.H., Van (Sigma) medium described above. Further Moerkerk,H.T.B. and Glatz,J.C.G. (1986) Biochem. of the was Med. Metab. handling samples as described above for the identification of Biol., 35, 248-259. intermediates. Walker,J.E. and Runswick,M.J. (1993) acyl-CoA J. Bioenerg. Biomembr., 25, 435-446. Enzyme assays Wanders,R.J.A., Heymans,H.S.A., Schutgens,R.B.H., Barth,P.G., Van Malate dehydrogenase activity was measured as the den Bosch,H. and Tager,J.M. (1988) J. Neurol. Sci., 88, 1-39. oxaloacetate-depend- ent rate of NADH oxidation (A340 nm) in assay mixtures containing Wanders,R.J.A., IJlst,L., van Gennip,A.H., Jacobs,C. and Jager,J.D. 45 mM K3PO4, pH 7.4, 0.12 mM NADH, and 0.33 mM oxaloacetate (1990) J. Inherited Metab. Dis., 13, 311-314. and (Steffan McAllister-Henn, 1992). 3-hydroxyacyl-CoA dehydro- Waterham,H.R., Keizer-Gunnink,I., Goodman,J.M., Harder,W. and activities were genase measured on a Cobas-Fara centrifugal analyser Veenhuis,M. (1990) FEBS Lett., 267, 17-19. by following the A. and Bucher,T. 3-keto-octanoyl-CoA-dependent rate of NADH con- Zebe,E., Delbruck, (1959) Biochem. Z., 113, 114-124. sumption at 340 nm (Wanders et Succinate al., 1990). dehydrogenase was measured according to a recently described method (Munujos et al., Received on October 20, 1994; revised on March 17, 1995 1993). Protein concentrations were determined by the bicinchoninic acid method (Smith et al., 1985). Acknowledgements We are grateful to P.Borst, B.Distel and I.Braakman for helpful sugges- tions and stimulating discussions, to and R.Butow A.Shyan for providing the Acit2 mutant, and to W.H.Kunau and co-workers for providing the Nycodenz gradient protocol. References Bieber,L.L. (1988) Annu. Rev. Biochem., 57, 261-283. Borst,P. (1963) in Proceedings 5th International Congress Biochemistry, Pergamon Press, London, Vol. 2, pp. 233-247. Borst,P. (1989) Biochim. Biophys. Acta, 1008, 1-13. Bremer,J. and Wojtczak,A.B. (1972) Biochim. Biophys. Acta, 280, 515-530. de Duve,C. and Baudhuin,P. (1966) Physiol. Rev., 46, 323-357. Elgersma,Y., Van Roermund,C.W.T., Wanders,R.J.A. and Tabak,H.F. (1995) EMBO J., 14, 3472-3479. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

The membrane of peroxisomes in Saccharomyces cerevisiae is impermeable to NAD(H) and acetyl‐CoA under in vivo conditions.

Download PDF
7 pages

Loading next page...
 
/lp/springer-journals/the-membrane-of-peroxisomes-in-saccharomyces-cerevisiae-is-impermeable-RFM6Apc39V

References (1)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1995
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/j.1460-2075.1995.tb07354.x
Publisher site
See Article on Publisher Site

Abstract

EMBO Journal vol.14 no. 14 pp.3480-3486, 1995 The of in Saccharomyces The membrane peroxisomes is impermeable to NAD(H) and acetyl-CoA cerevisiae under in vivo conditions in steps of the biosynthesis of ether-linked phospholipids Carlo W.T.van Roermund, Ype Elgersma1, in while mammalian cells take place peroxisomes synthesis Ronald J.A.Wanders and Neena Singh2, reticulum. This involves is completed in the endoplasmic Henk F.Tabak"'3 export of the intermediate alkyl-dihydroxyacetone of and Academic membrane the Departments Clinical Biochemistry Pediatrics, phosphate across the single bounding per- University of Amsterdam, Meibergdreef 9, Medical Centre, Bosch et A related micro- oxisome (Van den al., 1992). of E.C.Slater 1105 AZ Amsterdam, 'Department Biochemistry, the of body-like organelle, glycosome trypanosomes, of 15, Institute, University Amsterdam, Meibergdreef the contains the major part of glycolytic pathway, implying 1105 AZ Amsterdam, The Netherlands and 2Institute of Pathology, Reserve Cleveland, OH 44106, USA that dihydroxyacetone phosphate, glycerol-3-phosphate, Case Western University, should be able to 3-phosphoglycerate and cofactors pass 3Corresponding author and the glycosomal membrane (Opperdoes Borst, 1977). should be considered C.W.T.van Roermund, Y.Elgersma and N.Singh of such metabolites across the However, how transfer as first authors equal is still a matter of peroxisomal membrane takes place One school of debate (reviewed by Borst, 1989). thought NADH We investigated how generated during peroxiso- to low molecular is that peroxisomes are freely permeable is reoxidized to and how the mal ,8-oxidation NAD+ the behavi- weight compounds. This was concluded from is end product of ,B-oxidation, acetyl-CoA, transported our of equilibrium density gradient peroxisomes upon to mitochondria in Saccharomyces from peroxisomes that several in sucrose and the finding centrifugation of the malate cerevisiae. Disruption peroxisomal dehy- such as D-amino acid oxidase, glycolate oxidase enzymes 3 resulted in (- drogenase gene (MDH3) impaired and urate oxidase failed to exhibit structure-linked latency oxidation as measured in intact cells, whereas capacity Duve and 1966). Direct permeability meas- (de Baudhuin, in In (-oxidation was normal cell lysates. perfectly urements clamp analysis provided evidence using patch cells were unable to grow addition, mdh3-disrupted in favour of this (Van Veldhoven et al., 1987). concept non-fermentable on oleate whereas on other growth The other opinion holds that this permeability observed that MDH3 is carbon sources was normal, suggesting in vitro is a result of their isolation and that peroxisomes of NADH involved in the reoxidation generated during in vivo are closed compartments. This concept finds rather than as fatty acid 3-oxidation functioning part the observation that peroxisomes in Hansenula support by the To the of of glyoxylate cycle. study transport acetyl have an acidic interior which implies restricted polymorpha units from we the peroxisomal peroxisomes, disrupted of the membrane toward protons permeability peroxisomal citrate gene (CIT2). The lack of phenotype of synthase (Nicolay et al., 1987; Waterham et al., 1990). In addition, the cit2 mutant indicated the presence of an alternative the observed latency of glycosomal enzymes suggests a formed the pathway for transport of acetyl units, by barrier for substrates and permeability phosphorylated carnitine acetyltransferase protein (YCAT). Disruption cofactors involved in trypanosomal glycolysis (Opperdoes blocked the (3- of both the CIT2 and YCAT gene and Borst, 1977). in in data oxidation intact cells, but not lysates. Our A solution to overcome a membrane barrier is to use suggest that the peroxisomal membrane is strongly as in mitochondria (reviewed by Walker to NAD(H) and acetyl-CoA in vivo, and shuttle-systems, impermeable and Runswick, 1993). Here, transport of reducing equiva- predict the existence of metabolite carriers in the lents from the cytosol to mitochondria is mediated by the peroxisomal membrane to shuttle metabolites from vice glycerol-3-phosphate/dihydroxyacetone phosphate shuttle peroxisomes to cytoplasm and versa. words: (-oxidation/carnitine (Zebe et al., 1959) or the malate/aspartate shuttle (Borst, Key acetyltransferase/citrate cycle/malate dehydrogenase whereas the transport of acetyl-CoA is mediated synthase/glyoxylate 1963), by an acetylcarnitine shuttle (Bieber, 1988). If similar shuttles are operative in peroxisomes it predicts the existence of a set of specific enzymes that participate in Introduction these exchange processes. Peroxisomes are essential subcellular organelles involved Here we have re-investigated the issue of peroxisome in a variety of metabolic processes. Their importance is permeability using a genetic approach to study how underlined by the recognition of an increasing number of the end of the (-oxidation of fatty acids in products inherited diseases in man in which one or more peroxiso- Saccharomyces cerevisiae, acetyl-CoA and reducing mal functions is impaired (Wanders et al., 1988; Moser, for further equivalents (NADH), leave the peroxisome 1991; Van den Bosch et al., 1992). metabolism in the cytosol and mitochondria. Our results For most of the enzymatic pathways delineated so far, indicate that peroxisomes are impermeable to these com- peroxisomes are dependent on efficient communication pounds and that specific shuttles are required to facilitate with the remainder of the cell. For instance, the first two across the peroxisomal membrane. transport 38) Oxford University Press Peroxisome impermeability to NAD(H) and acetyl-CoA Wild-type mdh3 mutant 1- [ .~~~ r. . . I AZ ', - Fig. 1. Subcellular location of malate dehydrogenase in S.cerevisiae. An organellar pellet was obtained by subcellular fractionation of oleate-grown cells and used for density gradient centrifugation on Nycodenz. Fraction 1 presents the bottom fraction, fraction 20 the top fraction. Succinate dehydrogenase and 3-hydroxyacyl-CoA dehydrogenase were measured as mitochondrial and peroxisomal markers, respectively. (A) Gradient of wild- type cells; (B) gradient of Amdh3 cells. Results 1 2 3 4 dehydrogenase 3 is present in Malate peroxisomes Transport of reducing equivalents from cytosol to mito- chondria in higher eukaryotes, has long been known to be mediated by the glycerol-3-phosphate/dihydroxyacetone phosphate shuttle and the malate/aspartate shuttle. In /MVHD analogy to the latter shuttle, a candidate enzyme for the reoxidation of NADH in peroxisomes of S.cerevisiae is the malate dehydrogenase enzyme (MDH). Earlier studies have revealed the existence of three MDH isozymes in S.cerevisiae (McAllister and Thompson, 1987; Minard and McAllister-Henn, 1991; Steffan and McAllister-Henn, 1992). The C-terminus of MDH3 ends in SKL (Steffan McAllister-Henn, 1992) which is an established and targeting signal (PTS) (Gould et al., 1989). peroxisomal Actin the presence of MDH3 in peroxisomes has not, However, as been demonstrated. yet, We disrupted the MDH3 gene and tested whether this resulted in the absence of malate dehydrogenase in the fractions. Therefore, cells were grown in peroxisomal 2. Northern-blot analysis of MDH3 expression. Cells were grown Fig. a medium containing oleate, a well-known inducer of on medium containing glucose (1), glycerol (2), oleate (3) or acetate followed by subcellular fractionation and peroxisomes, as sole carbon source. Total RNA (10 was used for each lane (4) gg) onto nitrocellulose, the filters were of the agarose gel. After blotting centrifugation of the organellar pellet. density gradient MDH3 gene or with the actin gene as a with the radiolabelled probed The results presented in Figure 1 show good resolution control. between peroxisomes and mitochondria as exemplified by of activity of succinate dehydrogenase the distinct profiles induction of MDH3 mRNA by growth and 3-hydroxyacyl-CoA dehy- showed profound mitochondrial marker) (a whereas expression of the actin gene (used as marker). Importantly, malate on oleate, drogenase (a peroxisomal is almost constant under the various growth showed a bimodal distribution a control) dehydrogenase activity The observed induction of MDH3 conditions 2). with the peroxisomal and mitochondrial (Figure profile coinciding to that found for the ,-oxidation enzymes. is similar in wild-type cells (Figure IA). No peroxisomal very fractions was observed in the Amdh3 cells (Figure MDH activity is essential for growth on oleate MDH3 encodes the peroxiso- MDH3 that the gene B), indicating for malate dehydro- a role peroxisomal To investigate mal malate dehydrogenase. of NADH, the in reoxidation intraperoxisomal MDH is involved in reoxidation of genase If peroxisomal and Amdh3 strains were com- rates of wild-type one would expect induction of growth intraperoxisomal NADH, either oleate, acetate, ethanol on containing oleate since oleate is known to induce the pared plates activity by as the sole carbon source. Growth of Amdh3 or n-oxidation capacity and thus the production glycerol peroxisomal or was unaffected (not ethanol cells on acetate, glycerol Northern blot analysis indeed of NADH many-fold. 3481 Roermund et C.W.T.van aL the fatty acid. This gives a much better estimate of fatty VWV ird -type acid oxidation than the amount of [14C]CO2 alone, since only part of the acetyl-CoA produced during n-oxidation mdh3 is converted into C02 (Veerkamp et al., 1986). The results of Figure 4 suggest that the impairment in fatty acid 13- psf oxidation in Amdh3 intact cells is caused solely by the absence of malate peroxisomal dehydrogenase, and not by reduced induction or of the activity enzymes involved directly in ,B-oxidation, which include acyl-CoA oxidase, enoyl-CoA hydratase, 3-hydroxyacyl-CoA dehydrogenase Cis/E and 3-ketoacyl-CoA thiolase. Accumulation of 3-hydroxyacyl-CoA intermediates in the Amdh3 mutant ,Vcst If in vivo the block in 1-oxidation is indeed due to the inability to reoxidize peroxisomal NADH in the absence of MDH3, this should be reflected in the accumulation of the 3-hydroxyacyl-CoA ester in the A,ndh3 cells but not 800 - this notion in the in control cells. We tested experiment depicted in Figure 5. Oleate-induced wild-type and Amdh3 cells were incubated for 30 or 60 min with 0, radiolabelled acid. The various labelled fatty acyl-CoA esters, including the ester itself and the acyl-CoA 3- ax,p-unsaturated, hydroxy- and 3-ketoacyl-CoA were extracted and esters, separated on thin The results show that layer plates. significant levels of the ester were 3-hydroxyacyl-CoA found only in Amdh3 cells incubated with radiolabelled for or The fatty acid 30 60 min (Figure 5). observed accumulation of the 3-hydroxyacyl-CoA intermediate in the Amdh3 mutant suggests again that MDH3 participates in a redox shuttle rather than in the glyoxylate cycle. of from Removal acetyl-CoA peroxisomes is the end of the 1-oxidation of Acetyl-CoA product straight-chain fatty acids in An function yeast. important of the glyoxylate is the condensation of two C2 cycle units (acetyl-CoA) to the cell succinate, thereby enabling to form C4 carbon skeletons from C2 units. In 2 30 addition, 4|' 10o te 6zo >81 100n 1 20 14C if peroxisomes are impermeable for (acetyl-) CoA, this Time (hr) cycle is also important for releasing the CoA for continuing 1-oxidation cycles of and for facilitating the transport of of and mutant cells on Fig. 3. Growth wild-type cells oleate medium. carbon units across the Growth on a minimal oleate medium. The peroxisomal membrane. (A) plate containing pas21 mutant in the of is used as (disturbed assembly peroxisomes) control To test whether are peroxisomes indeed impermeable for no Growth curves of and mutant strains on growth. (B) wild-type for we tested the effect on acetyl-CoA, 1-oxidation by rich oleate medium. The strains shown are: cells wild-type (+), disrupting the peroxisomal citrate synthase (CIT2) gene Amdh3 cells (A), Acit2 cells (0), Aycat cells and Acit2/Aycat cells (El) (Lewin et al., 1990). Growth on oleate and As a oxidation of (O). control, wild-type cells were grown on the same medium without Tween/oleate (dashed line). [1-_4C]octanoic acid were investigated as described above for the Amdh3 mutant. The results depicted in Figure 6 shown), but on oleate was growth strongly impaired (second bar) show that oxidation of octanoate was normal (Figure 3), suggesting that malate in cells deficient in peroxisomal dehydro- peroxisomal citrate synthase. Moreover, genase is not taking part in the glyoxylate cycle, but is growth on oleate was indistinguishable from wild-type involved in the 1-oxidation of fatty acids. cells (Figure 3). However, since growth of Acit2 cells is To ascertain whether the inability to grow on oleate is almost normal on ethanol and acetate as well (not shown), caused directly by a block in peroxisomal 1-oxidation, we these results strongly suggest the existence of an efficient studied the oxidation of a 1-_4C-labelled fatty acid bypass that can compensate for the loss of peroxisomal in control (octanoic acid) and Amdh3 mutant cells. As citrate synthase. shown in Figure 4, oxidation of [1-_4C]octanoic acid was One possible bypass route would be the conversion of strongly impaired in the intact Amdh3 cells. Importantly, acetyl-CoA into acetylcarnitine via camitine acetyltrans- fatty acid oxidation was normal in Amdh3 cell-free lysates ferase (CAT) which is known to be present in mitochondria in which the membrane barriers of the different intra- and peroxisomes in higher eukaryotes and yeast (Markwell cellular organelles were absent and NAD+ was present in et al., 1973, 1976; Markwell and Bieber, 1976; Kawamoto excess. The rates of fatty acid oxidation represent the sum et al., 1978; accompanying paper, Elgersma et al., 1995). of and [14C]CO2 water-soluble material after extraction of The acetylcarnitine formed in peroxisomes might sub- 3482 Peroxisome impermeability to NAD(H) and acetyl-CoA *r)f ! _T 0 400 80;i 3 0O 1 0o WT mdh3 LI mdh3 = El WI Fig. 4. Octanoic acid p-oxidation in oleate-induced wild-type cells and Amdh3 cells. (A) p-oxidation in intact cells; (B) J-oxidation in cell lysates. acid oxidation is expressed as the sum of ['4C]C02 and water-soluble n-oxidation products produced. [1-14C]octanoic Time 0 0 30 60 30 60 (Figure 3). Furthermore, growth on (min) not to be affected acetate was normal (not shown). However, ethanol and Cells wt mdh3 wt wt mdh3 mdh3 disrupted both the CIT2 gene and the YCAT when we cells could no longer grow on oleate (Figure 3). gene, the the Acit2/ycat cells were also unable to grow on Since ethanol and acetate (not shown) we conclude that the YCAT protein is indeed indispensable for the net synthesis of C4 carbon units in case the CIT2 gene is deleted. Since the YCAT gene disruption results in the absence of both the peroxisomal and mitochondrial carnitine acetyl- transferase protein (Elgersma et al., 1995) the observed inability of Acit2/ycat cells to grow on oleate may be caused exclusively by the impermeability of mitochondria To investigate directly the permeab- towards acetyl-CoA. of the peroxisomal membrane towards ility properties we investigated the ability of the Acit2/ycat acetyl-CoA mutant to n-oxidize fatty acids as described above double for the Amdh3 mutant. Indeed, similar to what we found 5. TLC analysis of the 14C-labelled products derived from fatty Fig. the Amdh3 mutant, the p-oxidation was practically for acid oxidation in wild-type cells and Amdh3 cells. The products blocked in intact Acit2/ycat mutant cells, whereas the P- formed were analysed after incubating cells 0, 30 and 60 min with oxidation capacity was hardly affected in cell lysates in The right-hand panel shows a marker lane of [1-_4C]palmitate. which the membrane barriers are lost (compare Figure 6A CoA. [1_-4C]3-hydroxypalmitoyl This strongly suggests that there are with B, fourth bar). for the export of acetyl-CoA from two pathways only be transported to mitochondria where it could sequently either via conversion into glyoxylate cycle peroxisomes; be further oxidized to CO2 in the Krebs cycle. However, or via conversion into acetylcarnitine. Con- intermediates in the absence of the CIT2 protein, the citrate (or isocitrate) both pathways leads to the accumula- sequently, blocking mitochondria may also be retrieved for net formed in acetyl-CoA, which depletes the tion of peroxisomal of carbon skeletons in the glyoxylate cycle, as synthesis from free CoA for continued ,-oxidation or peroxisomes in 7. The possibility that peroxisomal and depicted Figure the f-oxidation enzymes by product inhibition inhibits citrate synthase may be able to take over mitochondrial and Osmundsen, 1989). (Hovik other's function has earlier been suggested for each based on deletion studies (Kispal et al., S.cerevisiae gene for the ethylamine-grown yeast Trichosporon 1988), and Discussion based on the enzymatic contents of peroxisomes cutaneum a novel to obtain information on We have used approach et al., 1986). (Veenhuis of peroxisomes in the yeast the permeability properties et al. (1995) demonstrated in the accompany- Elgersma is based on of genes This disruption specific is indeed present S.cerevisiae. that camitine acetyltransferase ing paper involved in the steps for preparatory and mitochondria of oleate-grown S.cerevi- encoding proteins in peroxisomes of metabolites from fatty acid ,B- the generated Both are encoded by the same gene, YCAT. transport siae. enzymes across the membrane. We first oxidation peroxisomal shows that deletion of the YCAT gene results in Figure 6 malate on dehydrogenase concentrated peroxisomal decrease in [1-14C]octanoic acid oxidation in intact a slight cloned and by the which was recently sequenced third bar) and in cell lysates (Figure 6B, (MDH3) cells (Figure 6A, and McAllister-Henn, of McAllister-Henn (Steffan group third However, the growth rate on oleate appeared bar). 3483 C.W.T.van Roermund et aL 1 1.-_I .... I s: ] I .:: .. ..... ....... .. ... ::: :: ::: ........ ........ ...... i i ...... :: :: ... ....... ....... ...... ..... :::. .. .. :: :. .. 1 . 1. * :. 3 4 3 4 1 2 Fig. 6. Octanoic acid ,B-oxidation in oleate-induced wild-type and mutant cells. (A) n-oxidation in intact cells; (B) ,-oxidation in cell lysates. acid oxidation was measured as the sum of [14C]C02 and water-soluble material produced. The bars represent wild-type cells (1), [1-14C]octanoic Acit2 cells (2), cells (3) and Acit2/Avcat cells (4). Aycat acids Fatty Fatty acylCoA Malate Co OTCA NAD + Cycle C°2 MDH3 NADH Oxaloacetate m la + Carnitine Acetyl-CoA + 94 I-- Carnitine --- Acetyl-CoA Carnitine 1% 4' CAT CAT 4, Acetylcarnitine - Acetylcarnitine Acetylcarnitine O- + Oxaloacetate Acetyl-CoA CIT2 4, Oxaloacetate Citrate Citrate e Citrate ._u 4, I ACO ACO Malate Isocitrate socitrate Succinate ->Fumarate Succinate - I Succinate ICLI -?* > ICL1 Acetyl-CoA + Glyoxylate Glyoxylate _.. MLS 4, Malate - Cytosol Mitochondrion Peroxisome for the reoxidation of intraperoxisomal NADH and the pathways for the transport of acetyl-CoA. Our results do not rule out that other Fig. 7. Model malate and oxaloacetate are shuttled between peroxisome and cytosol. Since the presence of aconitase and isocitrate lyase in metabolites than has not been demonstrated unambiguously, these reactions might also take place in the cytosol (bold-dashed arrows). The thin arrows peroxisomes citrate or isocitrate for use in the glyoxylate cycle in case the CIT2 gene has been deleted. retrieval pathway of mitochondrial indicate the supposed CIT2, peroxisomal citrate synthase; ACO, aconitase; ICL, isocitrate lyase; MLS, malate dehydrogenase 3; CAT, camitine acetyltransferase; MDH3, malate synthase. and Thompson, 1987; and to complete impairment of growth on 1992). Earlier studies (McAllister peroxisomes, Minard and McAllister-Henn, 1991) had identified a medium containing oleate as the sole carbon source due In cell lysates mitochondrial (MDH I) and cytosolic malate dehydro- to a peroxisomal 5-oxidation deficiency. reoxidation of NAD+ is not required since this genase (MDH2). Based on the presence of a serine- where tripeptide at the C-terminus, the is in the reaction medium, fatty acid leucine-lysine (SKL) cofactor present to be located in peroxi- oxidation was completely normal. These data strongly MDH3 isoenzyme was speculated somes and McAllister-Henn, 1992). Our results suggest that the peroxisomal membrane is impermeable (Steffan indicate that this is indeed the case. Disruption of the to NAD(H) in vivo, and that malate dehydrogenase is MDH3 gene led to the total absence of MDH activity in involved in regeneration of intraperoxisomal NAD+. In 3484 Peroxisome impermeability to NAD(H) and acetyl-CoA such a shuttle mechanism malate dehydrogenase catalyses variety of different transport-proteins such as carnitine/ the reduction of oxaloacetate to malate with concomitant acetylcarnitine carriers and carriers. dicarboxylate formation of NAD+ from NADH, followed by the shuttling of malate versus oxaloacetate across the peroxisomal Materials and methods membrane. We cannot exclude other shuttles, for example, a malate/aspartate shuttle system. Such a shuttle would Yeast strains and culture conditions require the presence of aspartate aminotransferase protein All the gene disruptions used for this were made in the S.cerevisiae study strain BJ1991 Yeast (MATa, leu2, trpl, ura3-251, prbl-1122, pep4-3). in peroxisomes, which converts aspartate into oxaloacetate. transformants were selected and on minimal medium grown containing We are currently investigating whether this protein is 0.67% yeast nitrogen base without amino acids (YNB-WO)(DIFCO), indeed present in peroxisomes of S.cerevisiae. 2% glucose and amino acids (20 as needed. The media liquid gig/ml) The finding that Amdh3 cells are not impaired in growth used for growing cells for RNA curves or subcellular isolations, growth fractionation contained 0.5% potassium phosphate buffer, pH 0.3% on C2 carbon sources suggests that this enzyme does not 6.0, 2% 2% or 2% K- yeast extract, 0.5% peptone, and or glucose, glycerol, participate in the glyoxylate cycle. Moreover, the enzyme acetate, or 0.12% oleic acid/0.2% Tween-40 as carbon source. Before would then have to operate in two directions in the same one of minimal 0.3% shifting to these media, cells were grown on compartment, which is obviously impossible. Indications medium for at least 24 h. For RNA the cultures were glucose isolations, inoculated at such a that reached = 0.7-1.0 after that the kinetic parameters of glyoxysomal malate dehydro- density they OD6W0 -15 h. Oleic acid contained 0.1% oleic plates acid/0.4% Tween-40, genase are unfavourable to its participation in the glyoxyl- 0.67% base without amino acids yeast nitrogen (YNB-WO)(DIFCO), ate cycle of plant glyoxysomes were earlier reported by 0.1% yeast extract and amino acids as needed. (DIFCO) (20 ,ug/ml) Mettler and Beevers (1980). The consequence of these findings is that malate produced by the glyoxylate cycle Cloning procedures Standard DNA were carried out as described is transported out of the peroxisome followed techniques (Sambrook by retro- et al., 1989). The yeast MDH3 gene was two amplified using oligonucleo- conversion to oxaloacetate in the cytosol (via MDH2) or of and tide primers corresponding to with MDHI regions non-homology the mitochondria (via MDH 1) (Figure 7). MDH2. the Oligonucleotide sequence from 5' end of the was gene (5'- The results described in this paper also provide new The TTTGAATTCAAGCATAAAACAATCAAGG-3'). oligonucleotide of was sequence from the 3' end the gene (5'-GGATCCGATATGAGT- information on the way in which acetyl-CoA is transported A DNA was CAAGATACAAAGG-3'). S.cerevisiae genomic library from the interior of the peroxisome to mitochondria for in reaction the above used as a template a polymerase chain (PCR) using further metabolism in the Krebs cycle. It was a surprise two The PCR reaction was carried out 0.5 of primers. using gg template to observe that disrupting the CIT2 gene for peroxisomal 167 10 mM 0.5 units of DNA, oligonucleotides, dNTPs, Taq gg/ml 1.5 mM 10 mM citrate synthase did not lead to a polymerase (Boehringer Mannheim), Tris-HCl, deficiency to grow on MgC92, 50 mM KCI and 0.01I% BSA in a total volume of 50 The pH 8.4, ,ul. oleate or C2 compounds such as acetate or ethanol. A 1.2 kb was was EcoRI-BamHI annealing temperature 55°C. fragment Since assimilation of C2 compounds requires a functional into site of The obtained and subcloned the multiple cloning pUC19. glyoxylate cycle, this suggests that citrate or isocitrate 600 of the MDH3 EcoRV-NcoI fragment containing bp open reading blunt ended with a 2.2 kb from mitochondria can reach frame was replaced using the cytoplasm (or even the cloning, fragment Linearized DNA with the MDH3 of the LEU2 gene. plasmid disrupted peroxisomes) as has been proposed earlier (Veenhuis et al., was used to transform a strain BJ1991. gene haploid wild-type Leu+ 1986; Kispal et al., 1988). Our results indicate that the were confirmed for of the chromosomal MDH3 transformants disruption carnitine acetyltransferase protein is essential for this the PCR obtained on chromosomal DNA locus by analysing product with the same as used earlier to the bypass (Figure 7). primers amplify gene. The CIT2 deletion was made DNA of gene by isolating genomic The finding that intact cells with either the CIT2 or the cells PSY42-cit2 (Leu2-2, leu2-112, lvs2-801, CI72::URA3) (kindly YCAT gene disruption have almost normal capacity to for PCR and This was used as provided by A.Shyan R.Butow). template oxidize fatty acids whereas 3-oxidation is blocked when and with the CI72 primers (5'-GGATCCATGACAGTTCCTTATCTA-3') PCR The both genes are disrupted, indicates that there are only two (5'-CTATAGTTTGCTTTCAATGTT-3'). resulting fragment and ura+ transformants were was used to transform BJ 1991 cells, ways in which acetyl-CoA can leave the peroxisome: via in the PCR selected for integration CI72 gene by analysis. conversion into glyoxylate cycle intermediates or via was from DNA the 5' CAT- The YCAT gene amplified genomic using conversion into acetylcarnitine by carnitine acetyltrans- and the 3' A primer (5'-TTTGAATTCGAGAACTCTCTCAAAC-3') ferase. Consequently, since 5-oxidation is normal CAT-B primer TCTCCC-3') virtually (5'-TTTCTGCAGCGTAAGCCCTTl-l used for the PCR reaction The oligonucleotides. annealing temperature in cell-free lysates of Acit2/ycat cells (where the membrane 2.1 kb EcoRI-PstI was subcloned was The 55°C. resulting fragment barriers are absent), we conclude that CoA acetyl-CoA (or The of the frame was into pUCl9 (pEL72). major part open reading itself) cannot freely membrane. pass the peroxisomal the 1281 of deleted, by replacing Accl-Bgl2 fragment (containing bp the LEU2 This It establish whether all the CAT has been notoriously difficult to open reading frame) by gene (pEL78). plasmid transform cells and Acit2 was used to (PSY142) wild-type (BJ1991) in of glyoxylate cycle enzymes are located peroxisomes in the YCAT were selected for cells. Leu+ transformants integration gene S.cerevisiae. The presence of isocitrate lyase and aconitase PCR by analysis. in peroxisomes is still a matter of debate, as indicated in Figure 7. However, these uncertainties do not fractionation and compromise Subcellular Nycodenz gradients as described der fractionations were (Van Leij Subcellular the implication we propose here with to the performed regard were used for continuous 16-35% et al., 1992). Organellar pellets impermeability of the peroxisomal membrane to NAD(H) of ml 42% with a cushion Nycodenz gradients (12 ml), Nycodenz and acetyl-CoA. The of the enzymes ,B-oxidation pathway mM mM KCI and 8.5% mM 1 dissolved in 5 6.0, EDTA, MES, pH for 2.5 h in a vertical rotor that produce these and the that sealed tubes were compounds enzymes sucrose. The centrifuged 000 at at 19 000 (MSE 8X35) (29 take care of the in removal of these r.p.m. g) 4°C. preparatory steps MLS I and compounds from peroxisomes (MDH3, CIT2, measurements fl-oxidation in doubt. YCAT) are all localized peroxisomes beyond any with water and in cells were washed phos- resuspended Oleate-grown of our results is in with = 2.5. of 20 of cell One of the predictions that, to analogy saline gl (PBS), Aliquots OD60W phate-buffered acid measurements in 200 were used for membrane must contain a mitochondria, the peroxisomal fatty suspension >-oxidation ,l 3485 C.W.T.van Roermund et al. Gould,S.J., medium containing PBS plus 10 IM [1-14C]palmitate or Keller,G.A., Hosken,N., Wilkinson,J. and Subramani,S. [1-14C]- (1989) J. Cell Biol., 1657-1664. octanoate. Reactions were allowed to proceed for 6 or 12 min at 30°C, 108, 100 of 1.3 M Heikoop,J.C., Van Roermund,C.W.T., followed by termination of reactions by adding ,l perchloric Just,W.W., Ofman,R., acid. Radiolabelled was Schutgens,R.B.H., Heymans,H.S.A., Wanders,R.J.A. and CO2 trapped overnight in 500 of 2 M NaOH. Tager,J.M. gl The were collected after (1990) J. Clin. Invest., 86, 126-130. 14C-labelled 5-oxidation products subsequently extracting the acidified material with as Hovik,R. and Osmundsen,H. (1989) Biochem. J., 263, 297-299. chloroformlmethanol/heptane described before et and in a Kawamoto,S., Ueda,M., Nozaki,C., Yamamura,M., Tanaka,A. and (Heikoop al., 1990) quantified liquid scintillation counter. Fatty acid I-oxidation activities were also measured Fukui,S. (1978) FEBS Lett., 96, 37-40. in cell-free lysates prepared by lysing protoplasts in a medium containing Kispal,G., Rosenkrantz,M., Guarente,L. and Srere,P. (1988) J. Biol. 0.1% Triton X-100, 5 mM MOPS, 1 mM EDTA and I mM Chem., 263, 11145-11149. pH 7.4, PMSF. The cell-free extract was subsequently incubated in reaction Lewin,A.S., Hines,V. and Small,G.M. (1990) Mol. Cell. Biol., 10, mM medium containing the following components: 150 Tris-HCI, pH 1399-1405. Mannaerts,G.P. 8.5, 5 mM ATP, 5 mM MgCl2, 2 mM NaCN, 100,M FAD, I mM and Van Veldhoven,P.P. (1993) Biochimie, 75, 147-158. 1 and Arch. Biochem. NAD, mM CoASH, 0.005% (w/v) Triton X-100, 1 mU/ml acyl-CoA Markwell,M.A.K. Bieber,L.L. (1976) Biophys., 172, 502-509. synthetase (Boehringer and 10 or Mannheim) gM [1-14C]palmitate Markwell,M.A.K., and [1-14C]octanoate. Reactions were allowed to proceed for 6 or 12 min, McGroarty,E.J., Bieber,L.L. Tolbert,N. (1973) J. Biol. Chem., 248, 3426-3442. followed by quantification of [14C]C02 and 14C-labelled :-oxidation as described above. Markwell,M.A.K., Tolbert,N.E. and Bieber,L.L. Arch. Biochem. products (1976) Biophys., 176, 479-488. Identification of intermediates McAllister,L. and Thompson,L.M. (1987) J. Bacteriol., 169, 5157-5166. acyl-CoA In order to the nature of the esters in Mettler,I.J. and Beevers,H. (1980) Plant Physiol., 66, 555-560. identify acyl-CoA accumulating mutant oleate-induced intact cells were incubated with 10 jM Minard,K.I. and McAllister-Henn,L. (1991) Mol. Cell. Biol., 11,370-380. cells, as described for 30 or 60 Reactions were Moser,H.W. (1991) Clin. Biochem., 24, 343-351. [1-_4C]palmitate above, min. terminated 100 tl 1.3 M acid. In order to all Munujos,P., Coll-Canti,J., Gonzalez-Sastre,F. and Gella,F.J. (1993) Anal. by perchloric hydrolyse CoA-esters 100 of 2 M NaOH was added and the Biochem., 212, 506-509. subsequently gl mixture was incubated at 50°C for 30 min. This was followed Nicolay,K., Veenhuis,M., Douma,A.C. and Harder,W. (1987) Arch. by addition of 10 of 0.5 M H2SO4 and 75 of sodium acetate Microbiol., 147, 37-41. ,ul buffer, gl If The pH 6.0. required, pH was adjusted to 4.0. fatty acids were then Opperdoes,F.R. and Borst,P. (1977) FEBS Lett., 80, 360-364. The and Molecular extracted with methanol/chloroformnheptane as described above. Sambrook,J., Fritsch,E.F. Maniatis,T. (1989) Cloning: A taken to under The Laboratory Manual, 2nd Cold lower layer was collected, washed and dryness N2. Ed., Spring Harbor Laboratory Press, in on TLC Cold Spring Harbor, NY. residue was taken up acetone, followed by chromatography conditions as described Bremer and with the et al. Anal. 76-85. using by Wojtczak (1972), Smith,P.K. (1985) Biochem., 150, that benzene was toluene. After the Steffan,J.S. and exception replaced by running, plate McAllister-Henn,L. (1992) J. Biol. Chem., 267, was dried and subjected to The of 24708-24715. autoradiography. recovery fatty acid derivatives during extraction and solvent acetone Van den Bosch,H., Schutgens,R.B.H., Wanders,R.J.A. and Tager,J.M. evaporation, solubilization and TLC chromatography were checked by determining (1992) Annu. Rev. Biochem., 61, 157-198. the recovery of and Van der Leij,I., Van den Berg,M., Boot,R., Franse,M., Distel,B. and [1-'4C]palmitoyl-CoA enzymatically synthesized enoyl-CoA esters and 3-hydroxyacyl-CoA esters from Tabak,H.F. (1992) J. Cell Biol., 119, 153-162. prepared Recoveries were >95%. Van Veldhoven,P.P., Just,W.W. and Mannaerts,G.P. (1987) J. Biol. Chem., [1-14C]palmitoyl-CoA. Radioactively labelled [1-14C]3-hydroxypalmitoyl CoA was 262, 4310-4318. synthe- Veenhuis,M., Van der sized 10 jM for 10 Klei,l.J. and Harder,W. (1986) Arch. Microbiol., enzymatically by incubating [1-14C]palmitate min with 50 U/ml crotonase (Sigma Co., St Louis, USA) and 10 U/ml acyl- 145, 39-50. CoA oxidase in a reaction as Veerkamp,J.H., Van (Sigma) medium described above. Further Moerkerk,H.T.B. and Glatz,J.C.G. (1986) Biochem. of the was Med. Metab. handling samples as described above for the identification of Biol., 35, 248-259. intermediates. Walker,J.E. and Runswick,M.J. (1993) acyl-CoA J. Bioenerg. Biomembr., 25, 435-446. Enzyme assays Wanders,R.J.A., Heymans,H.S.A., Schutgens,R.B.H., Barth,P.G., Van Malate dehydrogenase activity was measured as the den Bosch,H. and Tager,J.M. (1988) J. Neurol. Sci., 88, 1-39. oxaloacetate-depend- ent rate of NADH oxidation (A340 nm) in assay mixtures containing Wanders,R.J.A., IJlst,L., van Gennip,A.H., Jacobs,C. and Jager,J.D. 45 mM K3PO4, pH 7.4, 0.12 mM NADH, and 0.33 mM oxaloacetate (1990) J. Inherited Metab. Dis., 13, 311-314. and (Steffan McAllister-Henn, 1992). 3-hydroxyacyl-CoA dehydro- Waterham,H.R., Keizer-Gunnink,I., Goodman,J.M., Harder,W. and activities were genase measured on a Cobas-Fara centrifugal analyser Veenhuis,M. (1990) FEBS Lett., 267, 17-19. by following the A. and Bucher,T. 3-keto-octanoyl-CoA-dependent rate of NADH con- Zebe,E., Delbruck, (1959) Biochem. Z., 113, 114-124. sumption at 340 nm (Wanders et Succinate al., 1990). dehydrogenase was measured according to a recently described method (Munujos et al., Received on October 20, 1994; revised on March 17, 1995 1993). Protein concentrations were determined by the bicinchoninic acid method (Smith et al., 1985). Acknowledgements We are grateful to P.Borst, B.Distel and I.Braakman for helpful sugges- tions and stimulating discussions, to and R.Butow A.Shyan for providing the Acit2 mutant, and to W.H.Kunau and co-workers for providing the Nycodenz gradient protocol. References Bieber,L.L. (1988) Annu. Rev. Biochem., 57, 261-283. Borst,P. (1963) in Proceedings 5th International Congress Biochemistry, Pergamon Press, London, Vol. 2, pp. 233-247. Borst,P. (1989) Biochim. Biophys. Acta, 1008, 1-13. Bremer,J. and Wojtczak,A.B. (1972) Biochim. Biophys. Acta, 280, 515-530. de Duve,C. and Baudhuin,P. (1966) Physiol. Rev., 46, 323-357. Elgersma,Y., Van Roermund,C.W.T., Wanders,R.J.A. and Tabak,H.F. (1995) EMBO J., 14, 3472-3479.

Journal

The EMBO Journal – Springer Journals

Published: Jul 1, 1995

There are no references for this article.