Citrate Synthase 1 Interacts with the Citrate Transporter of Yeast Mitochondria Elena V. Grigorenko The University of Texas Southwestern Medical Center, Howard Hughes Medical Institute, 4500 S. Lancaster Road, Dallas, Texas 75216, USA
W. Curtis Small, Lars-Olof Persson and Paul A. Srere* Pre-Clinical Science Unit of the Department of Veterans Affairs Medical Center and Biochemistry Department of The University of Texas Southwestern Medical Center, 4500 S. Lancaster Road, Dallas, Texas 75216, USA We have previously shown that citrate synthase binds to an intrinsic protein of the mitochondrial inner membrane (D’Souza and Srere, 1983). In this paper we present evidence that this citrate synthase binding protein is the citrate transporter. We have used citrate synthase 1 mutants of Saecharornyccs ccrevcsiae and transformants containing citrate synthase inactivated by sitedirected mutagenesis to study the effect of the CSl protein upon mitochondrial function (Kispal and Srere). In the present study citrate uptake and oxidation were measured during state 3 conditions (presence of 200 p~ ADP) in the mitochondria of several strains of Saccharomyces ccrcvesiuc: a parental strain containing wild-type mitochondrial citrate synthase (CSl) and strains derived from a CSI deficient strain in which the CSl gene was disrupted by insertion of the LEU2 gene. These strains were generated from the CS1- cells by transformation with vectors encoding site-specific mutants of CSl possessing very low levels of enzymatic activity. One such strain in this study was subsequently found to have undergone reversion to produce a strain which had activity very similar to wild type. Positive correlation between citrate uptake and the rate of citrate oxidation was found, suggesting coupling of the two processes. Both mitochondrial citrate uptake and oxidation were decreased in the mutant lacking any form of CSI protein. Reintroduction of mutagenized CSI into yeast causes an enhancement in the rate of state 3 oxygen consumption and of citrate uptake. The observed respiratory increase produces a state 3 oxygen consumption rate which is 1.5- to 17-fold greater than the rate of reintroduced mutagenized C S l activity; conversely, wild-type C S l enzyme levels are about 30-fold greater than the respiratory rate in wild-type cells. Further evidence for uncoupling of the process of citrate utilization from CSI activity is also seen in the revertant strain which has lower respiratory rate than the wild-type strain. This difference is not due to decreased concentration of the citrate carrier of those cells with lower respiratory rates, a change in equilibrium for citrate across the mitochondrial membrane, or to decreased citrate saturability of such mitochondria. These results, supported by the interaction observed during column chromatography between matrix-immobilized pig citrate synthase and the protein from yeast mitochondrial membrane preparations which exhibits citrate transport activity, strongly suggest physical interaction between CSl and the mitochondrial citrate transport protein.
INTRODUCTION It has been demonstrated that in many metabolic pathways there are specific interactions between enzymes which are responsible for catalyzing sequential reactions (see Srere, 1987, for review). In addition, interactions exist between enzymes involved in metabolic function and structural elements of cells. The term metabolon has been coined to categorize such complexes (Srere, 1985). For the enzymes of the Krebs TCA cycle, six of the eight possible sequential interactions have been demonstrated, and all of the enzymes of the TCA cycle have been shown to be bound to the inner surface of the inner mitochondria membrane (D’Souza and Srere, 1983, Robinson and Srere, 1985; Brent and Srere, 1987). There is evidence which suggests that malate dehydrogenase, isocitrate dehydrogenase, and a-ketoglutarate dehydrogenase are bound both physically and functionally to complex I of the inner mitochondrial membrane (Sumegi and Srere, 1984). Although the inner membrane binding *Author to whom correspondence should be addessed. (t> 1990 by John Wiley & Sons, Ltd
protein(s) for citrate synthase (D’Souza and Srere, 1983) have not been purified or identified, it is both reasonable and consistent with the original hypothesis to postulate that this binding protein may be the well-known citrate carrier of the inner mitochondrial membrane (Chappel and Haaroff, 1967). We have shown recently that yeast cells which lack mitochondrial citrate synthase (CSl) lose the ability to grow on acetate. We also have shown that ability to grow on acetate in these CSl - cells can be restored by transformation of these cells with CSl which has been inactivated ( < 1% wild-type activity) by site-directed alteration of a residue in the active site believed to be essential for catalysis (Kispal et al., 1989). The mutant cells contain cytosolic citrate synthase (CS2), and one would expect that with the aid of the mitochondrial citrate carrier that the missing mitochondrial citrate synthase step could be by-passed by its cytosolic counterpart. Since this did not occur we have investigated the citrate transport step in the mutant yeast strains. One possible reason for the inability of citrate synthesized in the cytosol by CS2 to replace CSI activity
0952-3499/90/02 154219 $05.00 JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5/6,1990 215
ELENA V . G R I G O R E N K O ET A L
for growth on acetate might be due to an insufficiency of the mitochondrial citrate transport mechanism in CSI cells. I t has been shown (Kispal et al., 1988) that the rate of oxidation of certain Krebs cycle intermediates is reduced when CS1 is absent, but preliminary results at that time did not identify a transport problem. In contrast. earlier work in hepatoma cells indicated that citrate production in mitochondria may be coupled to citrate transport (Moreadith and Lehninger, 1984). Presented here are the results of our current investigation of the relationship between citrate transport, CS 1 activity, and cellular respiration in mitochondria isolated from a wildtype yeast strain and those strains containing mutations of the CSI gene.
EXPERIMENTAL Chemicals Triton X- 100. phenylsuccinic acid, and n-butylmalonic acid were from Aldrich Chemical Corp. (Milwaukee, W1, USA), dibutyl phthalate, a-cyano-3-hydroxycinnamic acid, D,L-isocitrate, oxaloacetate, coenzyme A, rotenone, antiniycin A, oligomycin, amino acids and adenine for yeast culture, and 1,2,3-benzene tricdrboxylate were from Sigma (St. Louis, MO, USA). Acetyl coenzyme A was prepared by the treatment of coenzyme A with acetic anhydride as described (Simon and Shemin, 1953). Dinonyl phthalate was from Eastman Kodak (Rochester, NY, USA). [I4C]citrate was from New England Nuclear (Boston, MA, USA). Yeast nitrogen broth was from Difco (Detroit, MI, USA). All other chemicals were reagent grade. Cell types Five strains of S. wrcwsiae were studied, a parental strain PSY 142 ( M a t a. leu2-2, lm2-112, lj*s2-801,ura3-52) and four mutant strains derived from this parent. These four strains were made CSI- by disruption of the CSI gene with the LEU2 gene as previously described (Kim et al., 1986). In three strains a second transformation introduced plasmids containing genes which coded for site-directed mutant CS 1 proteins containing substitutions of amino acids in the active site which are believed to participate in catalysis. In one plasmid the active site His313 was replaced with Gly (PSY 142/H313G);in another the active site His313 was replaced with Arg (PSY 142/H313R); and in the third the active site Asp414 was replaced with Gly (PSY 142/D414G). The plasmids also contained the LEU2 gene, and cells were selected by their ability to grow on synthetic medium which contained no leucine. The construction of these vectors has been described (Kispal c't al., 1989). Cells were routinely grown on complete synthetic medium (Kispal et ul., 1989) without leucine (wild type and CSI -) or without uracil and leucine (site-directed mutants) with 2% glucose as carbon source. Each of these mutants was characterized by a rate of mitochondrial citrate synthase activity which differs from the parental strain. The mutants lacking CSI had a level of citrate synthase activity which was not consistently significantly above background. The site-directed mutants PSY142/H313G, PSY142/H313R, and PSY142/D414G
were characterized by low mitochondrial citrate synthase activity, approximately 0.1-1.5% of the wild-type value. Even though these cells were maintained on selective medium, reversion rates were significant. Reversion was a particular problem in the case of the D414G strain which regained CSI activity approaching wild-type values in some cases. These revertants were studied as a control for the presence of CSI activity. Routinely, mitochondria were isolated from these cell lines by the method of Daum ct NI. (1972) from 500 mL cultures grown to stationary phase. Measurement of mitochondrial parameters Citrate synthase assays were conducted by the spectrophotometric method of Srere et al. (1963). Mitochondria for these determinations were suspended in a buffer containing 50 mM Tris HCI, pH 8.1, in 1 mM Na2EDTA, and 0.2% Triton X-100 at least 20 min before initiation of the assay. CS2 is known to be unstable at this pH, and > 90% of any residual CS2 activity in the mitochondrial preparation should be inactivated by this procedure (unpublished data). Mitochondria1 respiration was measured as previously described (Ohnishi et al.. 1966) using a Clark oxygen electrode. Protein concentrations were estimated by the method of Bradford (1976).
Measurement of citrate transport The uptake of I4C-labeled citrate was carried out in the medium used for respiratory studies with the addition of 8 p~ rotenone, 5 p~ antimycin A. and 8 p~ oligomycin. I4C-labeled citrate was used at a concentration of 2 mM and a specific activity of 0.8 pCi/pmol. The incubations were performed at 30 "C in a final volume of 300 pL. Incubations were initiated by the addition of a mitochondrial suspension containing I .21.5 mg protein. After a specific time of incubation (0 s, 10 s, 30 s, 10 min) a mixture of phenylsuccinic acid, n-butylmalonic acid, 1,2,3-benzene tricarboxylic acid, and a-cyano-3-hydroxycinnamic acid was added to yield final concentrations of 20, 20, 20, and 0.15 mM, respectively. The samples were then added to Eppendorf tubes containing 300 pL of saline medium layered atop a dibutyl phtha1ate:dinonyl phthalate mixture (5: I , v/v). The saline medium contained unlabeled citrate at the same concentration. The Eppendorf tubes were then centrifuged for 1 min at 7000 x g. The supernatant fluids were removed and the pellets dissolved in 300 pL of 88% formic acid. Finally, 200 pL of each dissolved pellet solution was counted in Aquasol in a liquid scintillation spectrometer.
Determination of mitochondrial citrate content Immediately after preparation, mitochondria were added to a solution of 0.6 M HCIO, and centrifuged at 3000 x g for 5 min. The supernatant was then neutralized to pH 7.4 with 20% (w/v) KOH containing 0.3 M MOPS, and citrate concentration was determined (Dugley, 1974). The protein concentration of these samples was determined by the method of Lowry et al. (1951).
216 JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5/6,1990
0 1990 by John Wiley & Sons, Ltd.
CITRATE TRANSPORT AND OXIDATION I N YEAST MITOCHONDRIA
RESULTS The rate of oxidation of citrate, malate, and succinate was measured in mitochondria in which coupled respiration was occurring. These measurements were made under state 3 conditions: i.e., the presence of 200 p ADP (Fig. I ) . The rate of oxidation of citrate and malate was decreased significantly in the mutant strains PSY 142/ CS 1 and PSY I42/H3 13G. The citrate oxidation rate was 24% of the wild-type value in the PSY142/CSI- mutant and 42% in PSY 142/H313 mitochondria. In the revertant PSY 142,D414G strain. the citrate- o r malate-dependent respiration of mitochondria isolated from these cells was
also lower than that of the parental strain. Succinate oxidation was not appreciably different in any of the strains studied. Table 1 compares the relative rates of state 3 respiration and CSI activities. Since the ratio of citrate synthase activity to respiratory rate is less than one for all mutant strains which did not undergo reversion it is readily apparent that their CSI activity is not sufficient to maintain their respiratory rate. The mutant lacking CSI had a significantly lower respiratory rate than all other strains tested. However, respiration still occurred in the absence of CSI, and this respiratory rate was greatly enhanced by the presence of mutant CS 1 possessing very low enzymatic activity. Further. in the cases of the parental strain and the PSY 142/D414G revertant, the presence of a large excess of CSI activity ( 30-fold that of the respiratory rate in both cases) did not greatly augment the rate of state 3 respiration over the values seen for those strains containing the site-directed mutants of CSI. Since the enzymatic conversions comprising the citric acid cycle proceed near the inner surface of the inner mitochondria1 membrane, we proposed that citrate synthase might interact with the citrate transporter system and may influence the rate of citrate and possibly malate (since this is also transported by the citrate transporter) flux in mitochondria. Therefore, we used radiolabeled citrate to study the citrate uptake in the mitochondria of the wild-type strain and three mutants. Fig. 2 summar-
-
0 25
0Citrats Malate Succinate
7
c
0.20
3
t
F 7
-P
0.15
$
's
.o
0.10
E a
14 I
v0 ON
0.05
PSY142
PSY14ZICSI-
PSY142IH313G
PSY1421D414G (Rev.)
Figure 1. Oxygen consumption in state 3 respiration in isolated yeast mitochondria of PSY142 derivatives. Cells were grown overnight in complete medium with glucose minus leucine. Mitochondria were isolated as described by Daum ef a/. (1972). 0 2 consumption was measured using a Clark electrode with 2 mM final concentrations of either citrate, malate or succinate as a substrate. ADP was added in a concentration of 200 PM. Values are the average of duplicate determinations of five yeast cultures SD.
*
Table 1. Comparison of citrate synthase 1 activity with state 3 oxygen consumption in mitochondria of BY142 cells and derivative tines
Strain
PSY142 PSY142JCSlPSY142JH313G PSY142JH313R PSYl42JD414G PSYl42JD414G (reversion)
State 3 Column l/column 2 CS1 activity oxygen consumption CS1 activity wnol/min/mg protein pnol/min/mg protein State 3 O2cons
15.5
< 0.004 0.23 0.02 0.03 12.8
0.52 0.1 4 0.38 0.30 0.1 8 0.37
30.2
0.70 0.06 0.18 34.4
Time (sec)
Cells were grown and harvested and mitochondria isolated as described in Materials and Methods. Values reported are the mean of four experiments.
(01990 by John Wiley & Sons, Ltd.
Figure 2. Dynamics of 14C-labeledcitrate uptake in mitochondria of PSY142 yeast derivatives. Experimentswere performed as described in Experimental. The data points indicated at time zero were obtained by the initial incubation of yeast mitochondria with 14Clabeled citrate in the presence of mixture of transport inhibitors. Values reported are corrected for extra-mitochondria space citrate and are the mean of triplicate determinations from four yeast cultures. 0 PSY142, PSY142/CS1-, PSY142/H313G, 0 PSYl42JD414G (reversion).
+
JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5J6.1990 217
ELENA V. GRIGORENKO E T A L .
Table 2. Determination of citrate content of isolated mitochondria preincubated with 2 mM citrate Citrate concentration prnollrng protein
Strain
*
4.52 1.83 4.94 i0.72 4.34;t 1.51 4.66 =k 0.68
PSY1421WT PSY142/CSI PSY1421H313G PSY142lD414G (reversion)
Intact isolated mitochondria were incubated and their citrate content measured as described in Materials and Methods. Values shown represent the mean f SD of four determinations.
izes the results of such experiments. Citrate influx into the mitochondria of PSY 142/CSI was not detectable, and the rate in PSY 142/H313G was approximately onequarter that of the parental strain. The PSY 142/D414G revertant and the parental strain have very similar dynamics of citrate uptake in the first 30 s of incubation, but after 20 min the amount of citrate uptake in the revertant was less than the wild type. In every case most of the uptake of citrate occurred during the first 30 s, and further incubation to 10 min did not greatly change these values. In some experiments, before measuring the uptake of I4C citrate into these mitochondria, we incubated the mitochondria in the presence of 2 mM unlabeled citrate to make sure that the citrate transporter was saturated with citrate. Mitochondria1 citrate concentration is essentially the same for all strains studied (Table 2). These results are similar to our previous observation in these strains (Kispal ef al., 1988). In these mitochondria, the differences in citrate uptake are maintained (data not shown). Therefore, these differences in citrate influx cannot be explained by the difference in intramitochondrial citrate available for exchange. Fig. 3 shows that citrate influx has a high degree of positive correlation to citrate oxidation in mitochondria. These data indicate that the mitochondria of those mutants which have a low rate of citrate oxidation are also characterized by a low rate of citrate uptake. Here ~
.$
250
* 0
L
I
t
"
I
0
0
0
.
0 " 150
+!
G
o
............................... 0
-5
5
10
15
20
25
Citrate uptake, nmollminlmg protein
Figure 3. Correlation between uptake of 14C-labeled citrate and ADP-dependent citrate oxidation in yeast mitochondria of PSY142 derivatives. The uptake of 14C-labeled citrate was carried out in the medium used for respiratory studies as described in Experimental. 0 PSY142, PSY142/CS1-, 0 PSY142/G313G, H PSY142/ D414G (reversion).
+
the amount of citrate uptake is more closely linked to the amount of citrate oxidation in the mitochondria than to their level of CSl activity. To ascertain that the absence of CSl was not causing a reduction in the amount of citrate transport protein present in the mitochondrion, separate experiments were conducted in which the carrier was isolated from the mitochondrial membrane of both wild type cells and those deficient in the CSl activity. This lipoprotein fraction was then reconstituted into liposomes (Claeys and Azzi, 1989) and citrate transport was measured (Luthy and Azzi, 1989) as previously described. These experiments showed that the membranes of CSI- mitochondria contain at least as much citrate carrier protein activity (73 nmol/min/mg) as the mitochondria from wild-type yeast (75 nmol/min/mg). In our recent studies on the purification of the yeast citrate transporter from yeast mitochondria we have observed that transporter in the extract of membrances is bound to a Sepharose column containing immobilized pig citrate synthase and can be eluted from the column by an increase in the ionic strength of the eluant. An increase in specific activity of the transporter was obtained by this procedure, but pure transport protein is not available as yet for further experiments.
DISCUSSION The functional coupling of transport proteins to metabolic enzymes has been reported for a number of systems (see Pande and Murthy, 1988 for review). In the present study we have shown that citrate transport is absent when citrate synthase protein is missing and that there is a positive correlation between citrate uptake and its oxidation. Since citrate transporter can be extracted and measured in proteoliposomes this suggests that the citrate transporter is regulated by the presence of citrate synthase and by tight coupling to the rate of utilization of citrate. It would appear that transporter is active in liposomes and not in situ. While this is an unusual observation it is clear that the environment of the transporter in situ is quite different from its environment in the liposomes. Citrate oxidation, in turn, is mediated by the Krebs cycle which appears to become more efficient when CSl protein is physically present, even when the primary carbon source is citrate. The physical presence of CSl with negligible activity causes an approximate doubling of the mitochondrial respiratory rate on citrate over mitochondria in those strains lacking the enzyme. This augmentation of respiratory rate may be responsible for the ability of these cells to grow on acetate. An excess ( - 130-fold)of CSl activity in the wild type causes only another 2-fold increase in respiration rate while a revertant strain (PSY 142/D414G) with essentially the same CSI activity as wild-type oxidizes both citrate and malate at a slower rate. We have shown previously that citrate synthase binds to a protein of the mitochondrial inner membrane (D'Souza and Srere, 1983; Brent and Srere, 1987). One hypothesis consistent with these data and with earlier studies of the Krebs cycle and related events is that the presence of the CSl protein is important to interact physically with other component(s) of the TCA cycle. The CSI protein in this case need not be catalytically active to be efficacious as long as it retains a
218 JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 516.1990
0 1990 by John Wiley & Sons, Ltd.
CITRATE TRANSPORT A N D O X I D A T I O N IN YEAST M I T O C H O N D R I A
steric conformation close to that of the native enzyme. Using this model, it may be inferred from the data discussed above that the revertant protein does not interact as well with other components of the TCA cycle as the native protein even though it has a very similar level of enzymatic activity and that the observations are consistent with the binding of citrate synthase to citrate transporter in order to obtain transport activity in situ. There are other systems in which the activity of a transport protein may depend on coupling to an enzyme. Yeast has three hexokinases and a high K, and low K, uptake systems for glucose and fructose. Bisson and Fraenkel (Bisson and Fraenkel, 1983a,b) have shown that in mutant yeast lacking the two hexokinases (but not the glucokinase), fructose showed only the high K, uptake while glucose exhibited both uptake systems. In a mutant lacking all three kinases, glucose was subjected only to the high K, system. Reinsertion of the missing kinases on a plasmid restored low K, hexose uptake. Fraenkel and Bisson speculate that the three kinases have a role other than their enzymatic one, such as a direct role of hexose uptake by binding to the carriers. Another example of such a system is ornithine uptake by mitochondria. Raijman and her associates (Cohen et al., 1987) loaded mitochondria with unlabeled ornithine.
When these mitochondria were incubated with labeled ornithine, the specific activity of the biosynthesized citrulline was the same as that for the exogenous labeled ornithine. It is possible, therefore, that the transport of ornithine is coupled directly to the matrix enzymes for its utilization. In addition, carnitine transacylase has been postulated to be functionally linked to mitochondrial matrix enzymes (Pande and Murthy, 1988). Perhaps the best-known mitochondrial system which links transport with metabolism is the proton-driven ATP synthase (ATPase) of the inner mitochondrial membrane. It is likely that in all these systems interaction between transporters and sequential enzymes are essential for in situ metabolic activity. With such examples in mind, experiments are now in progress to clarify further and characterize the role of CSI in the function of the citrate transporter.
Acknowledgements This work was supported by grants from the Department of Veterans Affairs, National Science Foundation, and United States Public Health Service. E.V.G. was a Chilton Foundation Fellow. We wish to thank Dr J. Johnson for his assistance and Mrs Penny Kerby for secretarial assistance.
REFERENCES Bisson, L F., and Fraenkel, D. G. (1983a).Proc. Narl. Acad. S o . USA 80,1730-1 734. Bisson. L F., and Fraenkel, D G. (1983b).J. Bacteriol. 155,
995-1000 Bradford. M M (1976).Anal. Biochem. 72, 248-254. Brent, L G , and Srere, P. A. (1987).J. B i d . Chem. 262,319-325. Chappell, B., and Haaroff, K. N. (1967).In Biochemistry of Mitochondria. ed by E. C Slater, 2. Kaniuga, and L. Wojtczak. pp. 75-91 Academic Press, London and PWN-Polish Scientific Publishers, Warsaw. Claeys, D , and Azzi, A. (1 989) J. B i d . Chem. 264,14627-1 4630. Cohen, N S.. Cheung, C - W , and Raijman, L. (1987).J B i d Chem 262,203-208. Daum, G., Bohni, P C., and Schatz, G. (1972). J. B i d . Chem. 257.
13028-1 3033. D'Souza, S . F , and Srere, P A. (1983).J. Biol Chem. 258.
4706-4709 Dugley. S. (1974).In Methods of Enzymatic Analysis, ed. by H . V. Bergmeyer, pp. 1562-1 565. Harcourt Brace Jovanovich, New York Kim. K - S , Rosenkrantz. M. S , and Guarente, L. P. (1986) Mol. Ce// Biol 6,1936-1 942. Kispal, G , Rosenkrantz. M., Guarente, L , and Srere, P. A (1988).J. B i d Chem 263. 1 1 145-1 1 149.
xj 1990 by John Wiley & Sons, Ltd.
G.,Evans, C. T.. Malloy, C , and Srere. P. A ( 1 989) J. Biol Chem. 264,11204-1 121 0 . Lowry, 0 . H . , Rosebrough. N. J., Farr. A L.. and Randall, R. T. (1951 ). J. Blol Chem. 193, 265-275 Luthy. R., and Azzi. A. (1989).Anal. Biochem. 177,323-326. Moreadith, R W., and Lehninger, A L. (1984). J fhol Chem 259, Kispal.
621 5-6221. Ohnishi. T , Kawaguchi. K , and Hagihara, B (1966) J Biol. Chem
241, 1797-1 806 Pande. S . V.. and Murthy. M. S R. (1988).In Microcompartmenfation, ed. by D. P Jones, pp. 93-114 CRC Press, Boca Raton. Florida. Robinson. J. B. Jr., and Srere, P. A. (1985) J. B i d Chem. 262,
10800-10805. Simon. E. J.. and Shemin, D. (1953).J. Am Chem. SOC.2520. Srere. P A (1987) Ann. Rev. Biochem. 56,89-1 24 Srere. P. A., Brazil, M., and Gonen, L (1963) Acra Chem Scand.
17,S129-Sl34 Srere, P. A. (1985).Trends Biochem. S o . 10,1 09-1 1 0 Sumegi. B and Srere, P. A (1984).J. B i d Chem 259, 15040-
15045 Received 1 1 September 1990; accepted (revised) 13 November 1990.
JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5/6,1990 219
W. Curtis Small, Lars-Olof Persson and Paul A. Srere* Pre-Clinical Science Unit of the Department of Veterans Affairs Medical Center and Biochemistry Department of The University of Texas Southwestern Medical Center, 4500 S. Lancaster Road, Dallas, Texas 75216, USA We have previously shown that citrate synthase binds to an intrinsic protein of the mitochondrial inner membrane (D’Souza and Srere, 1983). In this paper we present evidence that this citrate synthase binding protein is the citrate transporter. We have used citrate synthase 1 mutants of Saecharornyccs ccrevcsiae and transformants containing citrate synthase inactivated by sitedirected mutagenesis to study the effect of the CSl protein upon mitochondrial function (Kispal and Srere). In the present study citrate uptake and oxidation were measured during state 3 conditions (presence of 200 p~ ADP) in the mitochondria of several strains of Saccharomyces ccrcvesiuc: a parental strain containing wild-type mitochondrial citrate synthase (CSl) and strains derived from a CSI deficient strain in which the CSl gene was disrupted by insertion of the LEU2 gene. These strains were generated from the CS1- cells by transformation with vectors encoding site-specific mutants of CSl possessing very low levels of enzymatic activity. One such strain in this study was subsequently found to have undergone reversion to produce a strain which had activity very similar to wild type. Positive correlation between citrate uptake and the rate of citrate oxidation was found, suggesting coupling of the two processes. Both mitochondrial citrate uptake and oxidation were decreased in the mutant lacking any form of CSI protein. Reintroduction of mutagenized CSI into yeast causes an enhancement in the rate of state 3 oxygen consumption and of citrate uptake. The observed respiratory increase produces a state 3 oxygen consumption rate which is 1.5- to 17-fold greater than the rate of reintroduced mutagenized C S l activity; conversely, wild-type C S l enzyme levels are about 30-fold greater than the respiratory rate in wild-type cells. Further evidence for uncoupling of the process of citrate utilization from CSI activity is also seen in the revertant strain which has lower respiratory rate than the wild-type strain. This difference is not due to decreased concentration of the citrate carrier of those cells with lower respiratory rates, a change in equilibrium for citrate across the mitochondrial membrane, or to decreased citrate saturability of such mitochondria. These results, supported by the interaction observed during column chromatography between matrix-immobilized pig citrate synthase and the protein from yeast mitochondrial membrane preparations which exhibits citrate transport activity, strongly suggest physical interaction between CSl and the mitochondrial citrate transport protein.
INTRODUCTION It has been demonstrated that in many metabolic pathways there are specific interactions between enzymes which are responsible for catalyzing sequential reactions (see Srere, 1987, for review). In addition, interactions exist between enzymes involved in metabolic function and structural elements of cells. The term metabolon has been coined to categorize such complexes (Srere, 1985). For the enzymes of the Krebs TCA cycle, six of the eight possible sequential interactions have been demonstrated, and all of the enzymes of the TCA cycle have been shown to be bound to the inner surface of the inner mitochondria membrane (D’Souza and Srere, 1983, Robinson and Srere, 1985; Brent and Srere, 1987). There is evidence which suggests that malate dehydrogenase, isocitrate dehydrogenase, and a-ketoglutarate dehydrogenase are bound both physically and functionally to complex I of the inner mitochondrial membrane (Sumegi and Srere, 1984). Although the inner membrane binding *Author to whom correspondence should be addessed. (t> 1990 by John Wiley & Sons, Ltd
protein(s) for citrate synthase (D’Souza and Srere, 1983) have not been purified or identified, it is both reasonable and consistent with the original hypothesis to postulate that this binding protein may be the well-known citrate carrier of the inner mitochondrial membrane (Chappel and Haaroff, 1967). We have shown recently that yeast cells which lack mitochondrial citrate synthase (CSl) lose the ability to grow on acetate. We also have shown that ability to grow on acetate in these CSl - cells can be restored by transformation of these cells with CSl which has been inactivated ( < 1% wild-type activity) by site-directed alteration of a residue in the active site believed to be essential for catalysis (Kispal et al., 1989). The mutant cells contain cytosolic citrate synthase (CS2), and one would expect that with the aid of the mitochondrial citrate carrier that the missing mitochondrial citrate synthase step could be by-passed by its cytosolic counterpart. Since this did not occur we have investigated the citrate transport step in the mutant yeast strains. One possible reason for the inability of citrate synthesized in the cytosol by CS2 to replace CSI activity
0952-3499/90/02 154219 $05.00 JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5/6,1990 215
ELENA V . G R I G O R E N K O ET A L
for growth on acetate might be due to an insufficiency of the mitochondrial citrate transport mechanism in CSI cells. I t has been shown (Kispal et al., 1988) that the rate of oxidation of certain Krebs cycle intermediates is reduced when CS1 is absent, but preliminary results at that time did not identify a transport problem. In contrast. earlier work in hepatoma cells indicated that citrate production in mitochondria may be coupled to citrate transport (Moreadith and Lehninger, 1984). Presented here are the results of our current investigation of the relationship between citrate transport, CS 1 activity, and cellular respiration in mitochondria isolated from a wildtype yeast strain and those strains containing mutations of the CSI gene.
EXPERIMENTAL Chemicals Triton X- 100. phenylsuccinic acid, and n-butylmalonic acid were from Aldrich Chemical Corp. (Milwaukee, W1, USA), dibutyl phthalate, a-cyano-3-hydroxycinnamic acid, D,L-isocitrate, oxaloacetate, coenzyme A, rotenone, antiniycin A, oligomycin, amino acids and adenine for yeast culture, and 1,2,3-benzene tricdrboxylate were from Sigma (St. Louis, MO, USA). Acetyl coenzyme A was prepared by the treatment of coenzyme A with acetic anhydride as described (Simon and Shemin, 1953). Dinonyl phthalate was from Eastman Kodak (Rochester, NY, USA). [I4C]citrate was from New England Nuclear (Boston, MA, USA). Yeast nitrogen broth was from Difco (Detroit, MI, USA). All other chemicals were reagent grade. Cell types Five strains of S. wrcwsiae were studied, a parental strain PSY 142 ( M a t a. leu2-2, lm2-112, lj*s2-801,ura3-52) and four mutant strains derived from this parent. These four strains were made CSI- by disruption of the CSI gene with the LEU2 gene as previously described (Kim et al., 1986). In three strains a second transformation introduced plasmids containing genes which coded for site-directed mutant CS 1 proteins containing substitutions of amino acids in the active site which are believed to participate in catalysis. In one plasmid the active site His313 was replaced with Gly (PSY 142/H313G);in another the active site His313 was replaced with Arg (PSY 142/H313R); and in the third the active site Asp414 was replaced with Gly (PSY 142/D414G). The plasmids also contained the LEU2 gene, and cells were selected by their ability to grow on synthetic medium which contained no leucine. The construction of these vectors has been described (Kispal c't al., 1989). Cells were routinely grown on complete synthetic medium (Kispal et ul., 1989) without leucine (wild type and CSI -) or without uracil and leucine (site-directed mutants) with 2% glucose as carbon source. Each of these mutants was characterized by a rate of mitochondrial citrate synthase activity which differs from the parental strain. The mutants lacking CSI had a level of citrate synthase activity which was not consistently significantly above background. The site-directed mutants PSY142/H313G, PSY142/H313R, and PSY142/D414G
were characterized by low mitochondrial citrate synthase activity, approximately 0.1-1.5% of the wild-type value. Even though these cells were maintained on selective medium, reversion rates were significant. Reversion was a particular problem in the case of the D414G strain which regained CSI activity approaching wild-type values in some cases. These revertants were studied as a control for the presence of CSI activity. Routinely, mitochondria were isolated from these cell lines by the method of Daum ct NI. (1972) from 500 mL cultures grown to stationary phase. Measurement of mitochondrial parameters Citrate synthase assays were conducted by the spectrophotometric method of Srere et al. (1963). Mitochondria for these determinations were suspended in a buffer containing 50 mM Tris HCI, pH 8.1, in 1 mM Na2EDTA, and 0.2% Triton X-100 at least 20 min before initiation of the assay. CS2 is known to be unstable at this pH, and > 90% of any residual CS2 activity in the mitochondrial preparation should be inactivated by this procedure (unpublished data). Mitochondria1 respiration was measured as previously described (Ohnishi et al.. 1966) using a Clark oxygen electrode. Protein concentrations were estimated by the method of Bradford (1976).
Measurement of citrate transport The uptake of I4C-labeled citrate was carried out in the medium used for respiratory studies with the addition of 8 p~ rotenone, 5 p~ antimycin A. and 8 p~ oligomycin. I4C-labeled citrate was used at a concentration of 2 mM and a specific activity of 0.8 pCi/pmol. The incubations were performed at 30 "C in a final volume of 300 pL. Incubations were initiated by the addition of a mitochondrial suspension containing I .21.5 mg protein. After a specific time of incubation (0 s, 10 s, 30 s, 10 min) a mixture of phenylsuccinic acid, n-butylmalonic acid, 1,2,3-benzene tricarboxylic acid, and a-cyano-3-hydroxycinnamic acid was added to yield final concentrations of 20, 20, 20, and 0.15 mM, respectively. The samples were then added to Eppendorf tubes containing 300 pL of saline medium layered atop a dibutyl phtha1ate:dinonyl phthalate mixture (5: I , v/v). The saline medium contained unlabeled citrate at the same concentration. The Eppendorf tubes were then centrifuged for 1 min at 7000 x g. The supernatant fluids were removed and the pellets dissolved in 300 pL of 88% formic acid. Finally, 200 pL of each dissolved pellet solution was counted in Aquasol in a liquid scintillation spectrometer.
Determination of mitochondrial citrate content Immediately after preparation, mitochondria were added to a solution of 0.6 M HCIO, and centrifuged at 3000 x g for 5 min. The supernatant was then neutralized to pH 7.4 with 20% (w/v) KOH containing 0.3 M MOPS, and citrate concentration was determined (Dugley, 1974). The protein concentration of these samples was determined by the method of Lowry et al. (1951).
216 JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5/6,1990
0 1990 by John Wiley & Sons, Ltd.
CITRATE TRANSPORT AND OXIDATION I N YEAST MITOCHONDRIA
RESULTS The rate of oxidation of citrate, malate, and succinate was measured in mitochondria in which coupled respiration was occurring. These measurements were made under state 3 conditions: i.e., the presence of 200 p ADP (Fig. I ) . The rate of oxidation of citrate and malate was decreased significantly in the mutant strains PSY 142/ CS 1 and PSY I42/H3 13G. The citrate oxidation rate was 24% of the wild-type value in the PSY142/CSI- mutant and 42% in PSY 142/H313 mitochondria. In the revertant PSY 142,D414G strain. the citrate- o r malate-dependent respiration of mitochondria isolated from these cells was
also lower than that of the parental strain. Succinate oxidation was not appreciably different in any of the strains studied. Table 1 compares the relative rates of state 3 respiration and CSI activities. Since the ratio of citrate synthase activity to respiratory rate is less than one for all mutant strains which did not undergo reversion it is readily apparent that their CSI activity is not sufficient to maintain their respiratory rate. The mutant lacking CSI had a significantly lower respiratory rate than all other strains tested. However, respiration still occurred in the absence of CSI, and this respiratory rate was greatly enhanced by the presence of mutant CS 1 possessing very low enzymatic activity. Further. in the cases of the parental strain and the PSY 142/D414G revertant, the presence of a large excess of CSI activity ( 30-fold that of the respiratory rate in both cases) did not greatly augment the rate of state 3 respiration over the values seen for those strains containing the site-directed mutants of CSI. Since the enzymatic conversions comprising the citric acid cycle proceed near the inner surface of the inner mitochondria1 membrane, we proposed that citrate synthase might interact with the citrate transporter system and may influence the rate of citrate and possibly malate (since this is also transported by the citrate transporter) flux in mitochondria. Therefore, we used radiolabeled citrate to study the citrate uptake in the mitochondria of the wild-type strain and three mutants. Fig. 2 summar-
-
0 25
0Citrats Malate Succinate
7
c
0.20
3
t
F 7
-P
0.15
$
's
.o
0.10
E a
14 I
v0 ON
0.05
PSY142
PSY14ZICSI-
PSY142IH313G
PSY1421D414G (Rev.)
Figure 1. Oxygen consumption in state 3 respiration in isolated yeast mitochondria of PSY142 derivatives. Cells were grown overnight in complete medium with glucose minus leucine. Mitochondria were isolated as described by Daum ef a/. (1972). 0 2 consumption was measured using a Clark electrode with 2 mM final concentrations of either citrate, malate or succinate as a substrate. ADP was added in a concentration of 200 PM. Values are the average of duplicate determinations of five yeast cultures SD.
*
Table 1. Comparison of citrate synthase 1 activity with state 3 oxygen consumption in mitochondria of BY142 cells and derivative tines
Strain
PSY142 PSY142JCSlPSY142JH313G PSY142JH313R PSYl42JD414G PSYl42JD414G (reversion)
State 3 Column l/column 2 CS1 activity oxygen consumption CS1 activity wnol/min/mg protein pnol/min/mg protein State 3 O2cons
15.5
< 0.004 0.23 0.02 0.03 12.8
0.52 0.1 4 0.38 0.30 0.1 8 0.37
30.2
0.70 0.06 0.18 34.4
Time (sec)
Cells were grown and harvested and mitochondria isolated as described in Materials and Methods. Values reported are the mean of four experiments.
(01990 by John Wiley & Sons, Ltd.
Figure 2. Dynamics of 14C-labeledcitrate uptake in mitochondria of PSY142 yeast derivatives. Experimentswere performed as described in Experimental. The data points indicated at time zero were obtained by the initial incubation of yeast mitochondria with 14Clabeled citrate in the presence of mixture of transport inhibitors. Values reported are corrected for extra-mitochondria space citrate and are the mean of triplicate determinations from four yeast cultures. 0 PSY142, PSY142/CS1-, PSY142/H313G, 0 PSYl42JD414G (reversion).
+
JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5J6.1990 217
ELENA V. GRIGORENKO E T A L .
Table 2. Determination of citrate content of isolated mitochondria preincubated with 2 mM citrate Citrate concentration prnollrng protein
Strain
*
4.52 1.83 4.94 i0.72 4.34;t 1.51 4.66 =k 0.68
PSY1421WT PSY142/CSI PSY1421H313G PSY142lD414G (reversion)
Intact isolated mitochondria were incubated and their citrate content measured as described in Materials and Methods. Values shown represent the mean f SD of four determinations.
izes the results of such experiments. Citrate influx into the mitochondria of PSY 142/CSI was not detectable, and the rate in PSY 142/H313G was approximately onequarter that of the parental strain. The PSY 142/D414G revertant and the parental strain have very similar dynamics of citrate uptake in the first 30 s of incubation, but after 20 min the amount of citrate uptake in the revertant was less than the wild type. In every case most of the uptake of citrate occurred during the first 30 s, and further incubation to 10 min did not greatly change these values. In some experiments, before measuring the uptake of I4C citrate into these mitochondria, we incubated the mitochondria in the presence of 2 mM unlabeled citrate to make sure that the citrate transporter was saturated with citrate. Mitochondria1 citrate concentration is essentially the same for all strains studied (Table 2). These results are similar to our previous observation in these strains (Kispal ef al., 1988). In these mitochondria, the differences in citrate uptake are maintained (data not shown). Therefore, these differences in citrate influx cannot be explained by the difference in intramitochondrial citrate available for exchange. Fig. 3 shows that citrate influx has a high degree of positive correlation to citrate oxidation in mitochondria. These data indicate that the mitochondria of those mutants which have a low rate of citrate oxidation are also characterized by a low rate of citrate uptake. Here ~
.$
250
* 0
L
I
t
"
I
0
0
0
.
0 " 150
+!
G
o
............................... 0
-5
5
10
15
20
25
Citrate uptake, nmollminlmg protein
Figure 3. Correlation between uptake of 14C-labeled citrate and ADP-dependent citrate oxidation in yeast mitochondria of PSY142 derivatives. The uptake of 14C-labeled citrate was carried out in the medium used for respiratory studies as described in Experimental. 0 PSY142, PSY142/CS1-, 0 PSY142/G313G, H PSY142/ D414G (reversion).
+
the amount of citrate uptake is more closely linked to the amount of citrate oxidation in the mitochondria than to their level of CSl activity. To ascertain that the absence of CSl was not causing a reduction in the amount of citrate transport protein present in the mitochondrion, separate experiments were conducted in which the carrier was isolated from the mitochondrial membrane of both wild type cells and those deficient in the CSl activity. This lipoprotein fraction was then reconstituted into liposomes (Claeys and Azzi, 1989) and citrate transport was measured (Luthy and Azzi, 1989) as previously described. These experiments showed that the membranes of CSI- mitochondria contain at least as much citrate carrier protein activity (73 nmol/min/mg) as the mitochondria from wild-type yeast (75 nmol/min/mg). In our recent studies on the purification of the yeast citrate transporter from yeast mitochondria we have observed that transporter in the extract of membrances is bound to a Sepharose column containing immobilized pig citrate synthase and can be eluted from the column by an increase in the ionic strength of the eluant. An increase in specific activity of the transporter was obtained by this procedure, but pure transport protein is not available as yet for further experiments.
DISCUSSION The functional coupling of transport proteins to metabolic enzymes has been reported for a number of systems (see Pande and Murthy, 1988 for review). In the present study we have shown that citrate transport is absent when citrate synthase protein is missing and that there is a positive correlation between citrate uptake and its oxidation. Since citrate transporter can be extracted and measured in proteoliposomes this suggests that the citrate transporter is regulated by the presence of citrate synthase and by tight coupling to the rate of utilization of citrate. It would appear that transporter is active in liposomes and not in situ. While this is an unusual observation it is clear that the environment of the transporter in situ is quite different from its environment in the liposomes. Citrate oxidation, in turn, is mediated by the Krebs cycle which appears to become more efficient when CSl protein is physically present, even when the primary carbon source is citrate. The physical presence of CSl with negligible activity causes an approximate doubling of the mitochondrial respiratory rate on citrate over mitochondria in those strains lacking the enzyme. This augmentation of respiratory rate may be responsible for the ability of these cells to grow on acetate. An excess ( - 130-fold)of CSl activity in the wild type causes only another 2-fold increase in respiration rate while a revertant strain (PSY 142/D414G) with essentially the same CSI activity as wild-type oxidizes both citrate and malate at a slower rate. We have shown previously that citrate synthase binds to a protein of the mitochondrial inner membrane (D'Souza and Srere, 1983; Brent and Srere, 1987). One hypothesis consistent with these data and with earlier studies of the Krebs cycle and related events is that the presence of the CSl protein is important to interact physically with other component(s) of the TCA cycle. The CSI protein in this case need not be catalytically active to be efficacious as long as it retains a
218 JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 516.1990
0 1990 by John Wiley & Sons, Ltd.
CITRATE TRANSPORT A N D O X I D A T I O N IN YEAST M I T O C H O N D R I A
steric conformation close to that of the native enzyme. Using this model, it may be inferred from the data discussed above that the revertant protein does not interact as well with other components of the TCA cycle as the native protein even though it has a very similar level of enzymatic activity and that the observations are consistent with the binding of citrate synthase to citrate transporter in order to obtain transport activity in situ. There are other systems in which the activity of a transport protein may depend on coupling to an enzyme. Yeast has three hexokinases and a high K, and low K, uptake systems for glucose and fructose. Bisson and Fraenkel (Bisson and Fraenkel, 1983a,b) have shown that in mutant yeast lacking the two hexokinases (but not the glucokinase), fructose showed only the high K, uptake while glucose exhibited both uptake systems. In a mutant lacking all three kinases, glucose was subjected only to the high K, system. Reinsertion of the missing kinases on a plasmid restored low K, hexose uptake. Fraenkel and Bisson speculate that the three kinases have a role other than their enzymatic one, such as a direct role of hexose uptake by binding to the carriers. Another example of such a system is ornithine uptake by mitochondria. Raijman and her associates (Cohen et al., 1987) loaded mitochondria with unlabeled ornithine.
When these mitochondria were incubated with labeled ornithine, the specific activity of the biosynthesized citrulline was the same as that for the exogenous labeled ornithine. It is possible, therefore, that the transport of ornithine is coupled directly to the matrix enzymes for its utilization. In addition, carnitine transacylase has been postulated to be functionally linked to mitochondrial matrix enzymes (Pande and Murthy, 1988). Perhaps the best-known mitochondrial system which links transport with metabolism is the proton-driven ATP synthase (ATPase) of the inner mitochondrial membrane. It is likely that in all these systems interaction between transporters and sequential enzymes are essential for in situ metabolic activity. With such examples in mind, experiments are now in progress to clarify further and characterize the role of CSI in the function of the citrate transporter.
Acknowledgements This work was supported by grants from the Department of Veterans Affairs, National Science Foundation, and United States Public Health Service. E.V.G. was a Chilton Foundation Fellow. We wish to thank Dr J. Johnson for his assistance and Mrs Penny Kerby for secretarial assistance.
REFERENCES Bisson, L F., and Fraenkel, D. G. (1983a).Proc. Narl. Acad. S o . USA 80,1730-1 734. Bisson. L F., and Fraenkel, D G. (1983b).J. Bacteriol. 155,
995-1000 Bradford. M M (1976).Anal. Biochem. 72, 248-254. Brent, L G , and Srere, P. A. (1987).J. B i d . Chem. 262,319-325. Chappell, B., and Haaroff, K. N. (1967).In Biochemistry of Mitochondria. ed by E. C Slater, 2. Kaniuga, and L. Wojtczak. pp. 75-91 Academic Press, London and PWN-Polish Scientific Publishers, Warsaw. Claeys, D , and Azzi, A. (1 989) J. B i d . Chem. 264,14627-1 4630. Cohen, N S.. Cheung, C - W , and Raijman, L. (1987).J B i d Chem 262,203-208. Daum, G., Bohni, P C., and Schatz, G. (1972). J. B i d . Chem. 257.
13028-1 3033. D'Souza, S . F , and Srere, P A. (1983).J. Biol Chem. 258.
4706-4709 Dugley. S. (1974).In Methods of Enzymatic Analysis, ed. by H . V. Bergmeyer, pp. 1562-1 565. Harcourt Brace Jovanovich, New York Kim. K - S , Rosenkrantz. M. S , and Guarente, L. P. (1986) Mol. Ce// Biol 6,1936-1 942. Kispal, G , Rosenkrantz. M., Guarente, L , and Srere, P. A (1988).J. B i d Chem 263. 1 1 145-1 1 149.
xj 1990 by John Wiley & Sons, Ltd.
G.,Evans, C. T.. Malloy, C , and Srere. P. A ( 1 989) J. Biol Chem. 264,11204-1 121 0 . Lowry, 0 . H . , Rosebrough. N. J., Farr. A L.. and Randall, R. T. (1951 ). J. Blol Chem. 193, 265-275 Luthy. R., and Azzi. A. (1989).Anal. Biochem. 177,323-326. Moreadith, R W., and Lehninger, A L. (1984). J fhol Chem 259, Kispal.
621 5-6221. Ohnishi. T , Kawaguchi. K , and Hagihara, B (1966) J Biol. Chem
241, 1797-1 806 Pande. S . V.. and Murthy. M. S R. (1988).In Microcompartmenfation, ed. by D. P Jones, pp. 93-114 CRC Press, Boca Raton. Florida. Robinson. J. B. Jr., and Srere, P. A. (1985) J. B i d Chem. 262,
10800-10805. Simon. E. J.. and Shemin, D. (1953).J. Am Chem. SOC.2520. Srere. P A (1987) Ann. Rev. Biochem. 56,89-1 24 Srere. P. A., Brazil, M., and Gonen, L (1963) Acra Chem Scand.
17,S129-Sl34 Srere, P. A. (1985).Trends Biochem. S o . 10,1 09-1 1 0 Sumegi. B and Srere, P. A (1984).J. B i d Chem 259, 15040-
15045 Received 1 1 September 1990; accepted (revised) 13 November 1990.
JOURNAL OF MOLECULAR RECOGNITION, VOL. 3, No. 5/6,1990 219
