The Regulation of Isoleucine-Valine Biosynthesis in Saccharomyces cerevisiae

European J. Biochem. 3 (1968) 502-506 The Regulation of Isoleucine-Valine Biosynthesis in Saccharomyces cerevisiae 2. Identification and Characterization of Mutants Lacking the Acetohydroxyacid Synthetase P. T. MAGEE and H. DE ROBICHON-SZULMAJSTER Laboratoire d'enzymologie du C.N.R.S., Gif-sur-Yvette (Received September 26, 1967) Acetohydroxyacid synthetase activity has been characterized, independently of the acetoinforming system in Saccharomyces cerevisiae by a differential colorimetric assay. Maximum ph and K, values have been established for both enzymes with pyruvate as the substrate. Acetolactate formation by acetohydroxyacid synthetase has been shown directly by the separation of dinitrophenylosazones by thin-layer chromatography. Evidence for the biosynthetic role of the activity measured resides in its presence in wild-type strains, in its absence in the is, class of mutants (isoleucine-valine auxotrophs) and in its reappearance in prototrophic revertants of these mutants. The experiments of Strassman, Weinhouse and coworkers [ 1,2] have indicated that the pathway for biosynthesis of valine and isoleucine in yeast proceeds by means of the same reactions as in bacteria and Neurospora crash. Their experiments, however, were done by using labelled precursors and did not involve any studies of the enzymes implicated. Kakar and Wagner [3], using isoleucine-valine auxotrophs of 8. cerevisiae, classified each of four complementation groups according to the respective enzymatic defect (they were unable to find any class of mutants lacking either the acetohydroxyacid synthetase or the transaminase). Furthermore, they were able to find no enzymatic defect corresponding to the fifth group, since called is, [4] ; they suggested that is was associated with a disruption in a subcellular particle in which were localized the enzymes necessary for the synthesis of the two amino acids. The structural integrity of the particle would be necessary for the function of the enzymes. This paper reports the results of further studies on mutants of the is, locus. These studies have shown that the is, mutants lack the first enzyme of the combined pathway, the acetohydroxyacid synthetase. The failure of Kakar and Wagner to find this defect is attributable to a confusion between acetohydroxyacid synthetase and the acetoin-forming system of yeast. The latter system has been studied in detail by Juni [5]. Enzymes. Threonine deaminase or L-threonine hydrolyase (deaminating) (EC 4.2.1.16); acetolactate decarboxylase or 2-acetolactate carboxy-lyase (EC 4.1.1.5) ; pyruvic decarboxylase or 2-0x0-acid carboxy-lyase (EC 4.1.1.1). MATERIALS AND METHODS Strains The strains used include the ones described in the preceding paper, with, in addition, the following haploid mutant strains: MI4 (a, hi, is3); M2 (a, try, is,); M15 (a, hi, is,). They were kindly provided by Dr. D. C. Hawthorne. Selection of various revertants has been previously described [6]. Media, Growth and Cell-Free Extracts A complete description has been given in the first paper of this series. Assay of Enzymes The assay was a modification of the method of Ramakrishnan and Snellr71. Each tube contained 50 pmoles of sodium pyruvate, 2 pmoles of MgSO,, and 20pg of thiamine pyrophosphate. The volume was made up to 1 ml with 0.2 M potassium dihydrogen-phosphate adjusted with 1 M Tris buffer to the appropriate ph. The tubes were incubated in a 30" water bath for 5 minutes, then enzyme was added. After 20minutes the reaction was stopped by the addition of 0.1 ml of 6 N H,SO, or of 2.5 N NaOH. The tubes which had been acidified were heated for 15 minutes at) 55". At the end of this period, 1 ml of 0.5 /0 creatine was added to each tube, followed by 1 ml of 4 /0 a-naphthol in 2.5 N KaOH. The tubes were then mixed vigorously on a "Vortex Jr." mixer and incubated at 55" for 5 minutes. At the end of this time, they were removed and mixed again, then

~ ~~ ~ ~ Val. 3, No. 4, 1068 P. T. MAGEE arid H. DE ROBICHON-SZULMAJSTER 503 centrifuged to remove precipitated protein. (The heating after addition of the reagents causes the color to develop completely in 5 minutes, rather than in over 40 minutes at room temperature as described previously [7].) The tubes were read at 530 mp. in a Zeiss spectrophotometer. The entire assay was run in duplicate and the average of the two tubes taken as the final value. The amount of acetoin present in each tube was calculatjed from the standard curve made by adding known amounts of acetoin to a reaction mixture minus enzyme and then adding either acid or base. It is important to use two standard curves, as the intensity of color in a reaction mixture stopped by the addition of base is only 75O/, as great as in an acidified reaction mixture. In this assay, the amount of acetoin present after acidification measures acetolactate plus acetoin, since acetolactate undergoes spontaneous decarboxylation in acid to yield acetoin. The assay of a reaction mixture to which base has been added measures only acetoin. Thus the difference between the two values is a measure of acetolactate in a system in which both compounds are present. Measurement of Protein Protein content was determined by the method of Lowry et al. [S]. Formation of Dinitroplhenylosazones and Separation in Thin-Layer Chromatography Standard reaction mixtures (5 ml) containing 1 ml of the appropriate cell-free extract (about 5 mg of protein) were incubated for 2 hours at 30". The reaction was stopped by the addition of 50ml of 0.3O/, dinitrophenylhydrazine in 2 N HC1. After further incubation at 30" for 60 minutes, the precipitate was washed with absolute ethanol at 60" until the effluent was clear. (The dinitrophenylhydrazone of pyruvate is soluble in hot ethanol, the osazone of acetolactate is not [I]. Thus this treatment eliminates most of the great excess of pyruvate which is present in the reaction mixture.) The residual precipitate was dissolved in 3 ml of pyridineether (1 : 2, v/v) and 5 pl of the resulting solution was spotted on silica gel plates. The chromatogram was developed in benzene-acetic acid (9 : I, vlv). RESULTS Juni [5] has demonstrated the existence, in S. cerevisiae, of an enzyme system which produces one mole of acetoin from two moles of pyruvate or from one mole of pyruvate and one mole of acetaldehyde. (Since yeast possesses no acetolactate decarboxylase [5], acetolactate is not an intermediate in acetoin biosynthesis.) The ph optimum of this system was reported to be near 5.5. The assay used by Kakar and Wagner 131 for the acetohydroxyacid synthetase was conducted at ph 6.5 and it seemed possible that they were measuring acetoin rather than acetolactate formation. We therefore examined the relative production of acetolactate and acetoin from pyruvate in extracts of the wild type and of M15, an is, mutant, at ph 6.5. The table shows that the mutant produced very little acetolactate but a large amount of acetoin in this assay, whereas the wild type (M14R1, a revertant of M14) produced approximately 50 /, acetolactate and 50 /, acetoin. For comparison, a revertant of M15, M15R2, had regained the ability to produce acetolactate and indeed produced it at a higher level than did the wild-type. We then undertook to investigatc more fully the ph Table. Production of acetolactate and acetoin at pli 6.5 by extracts of S. cerevisiae The extracts were prepared and assayed as described in methods. They contained the following amounts of protein: M14R,, 5 mg/ml; ML5, 7.5 mg/rnl; M15R2, 10 mg/ml Aeetoin Extract Acetoin Acetolactate - - Acetolactatc nmolcs/20 rniniing protein M14R, 79.6 81.5 0.977 Mi5 141.1 8.9" 15.8 M15R2 63.7 112.5 0.561 8. This represents a difference, rnensurcd in the assay, of 2.7 nmoles, well within the experimental error. dependence of acetolactate formation. Fig. 1 shows the variation of the activity of the two enzyme systems with ph in wild-type extracts, in is, extracts, and revertant extracts. This figure shows clearly that the ph optimum of the acetolactate synthetase lies at 7.2, that MI5 lacks this activity, and that, in contrast to Escherichia coli and N. crassa, S. cerevisiae contains only one enzyme active in acetolactate synthesis between ph 6.0 and 9.0. The latter fact is shown by the monotonic ph curve and by the loss and recovery of the activity in single genetic steps. The acetoin synthetase differs from the acetolactate synthetase in properties other than its ph maximum. In order to differentiate more clearly between the two systems, we investigated their respective Michaelis constants. Fig. 2 shows the variation of enzyme activity with pyruvate concentration in M14R, at ph 6.0, where only acetoin is produced and at ph 7.2 with acetolactate measured as the product. The figure shows that the acetoinforming system does not follow classical Michaelis- Menten kinetics, but that the activity is proportional to the square of the pyruvate concentration. The acetolactate synthetase, on the other hand, shows classical kinetics. The Michaelis constant for pyruvat,e in the latter system is 8.6~10-~ M; for acetoin

504 Isoleucine-Valine Biosynthesis in S. cerevisiae. 2 European J. Biocheiii. M14 R1 A '"1 M15R2 C loot B 60 70 8.0 90 Fig.1. Variation of production of acetolactate and acetoin with ph in extracts of S. cerevisiae. The extracts contained the following amounts of protein: (A) M14R,, 12.0 mg/ml; (B) M15, 5.4 mg/ml; (C) M15R,, 5.5 mg/ml. All were grown on minimal medium plus histidine, isoleucine and dine 0 500 1000 VllSI 0 6.0 70 80 9.0 8 h M14 R1 CELL-FREI formation the concentration of pyruvate required for half-maximal activity is 1.4 x M. Acetohydroxybutyrate and acetolactate are apparently not taken up by S. cerevisiae, since we failed in all attempts to substitute these compounds for isoleucine and valine in fulfilling the nutritional requirements of auxotrophs. This was true whether the block was at the is, locus or at the is, locus. Thc latter class of mutants lacks threonine deaminase, requires only isoleucine, and thus should grow on acetohydroxybutyrate. We therefore undertook to demonstrate chemically that M15 fails to produce acetolactate at ph 7.2 and that M14R, and M15R2 10-3.V/ [ s] Fig. 2. Uependence on pyruvate concentration of acetoin and acetolactate formation. (A) Plot of u us. V/[SIz for acetoin formation at ph 6.0. (B) Plot of V vs. V/[S] for acetolactate formation at ph 7.2. Both experiments were done with cell-free extracts of M14R,. K, in such a plot is equal to minus the slope of the line

Vol. 3, No. 4, 1968 P. T. MAGEE and H. DE ROBICHON-SZULMAJSTER 605 do. Fig. 3 shows the results of a thin-layer chromatogram of the dinitrophenylosazones of standard assay mixtures, each containing one of the three types of extracts. It can be seen that only M14R, and M15R, produced acetolactate at ph 7.2 and that only the dinitrophenylhydrazone of pyruvate is present in the mixture corresponding to M15. The dinitrophenylosazone of acetoin seems to form very slowly under the reaction conditions employed as shown by the fact that a mock reaction mixture containing 12pmoles of acetoin produces no osazone upon reaction at 30" for 40minutes, in the presence of 0.30/,, dinitrophenylhydrazone. ORIGIN ;OLVENT ---- FRONT 0 0 0 0 0 0 0 0 0-0 0 o..... AL A MI5 MI5 M14 P R4 Fig. 3. Thin-layer chromutogrupjq of dinitrophenylosazones of reaction mixtures containing wild-type mutant and revertant extracts. M14R,, MI5 and M15R, refer to reaction mixtures containing these extracts. P is a pyruvate standard; AL, acetolactate; A, acetoin. The differential assay showed that M14R, produced 3.4 pmoles of acetolactate and 7.5 pmoles of acetoin; Mi5 produced 3 pmoles of acetolactate and 6 pmoles of acetoin; and M15R, produced 1.3 pmoles of acetolactate and 7.8 pmoles of acetoin DISCUSSION All the foregoing data lead to the conclusion that the is, mutants of S. cerevisiae lack the enzyme acetohydroxyacid synthetase. The differential assay, while admittedly indirect, shows a clear difference between extracts of auxotrophs and protoprophs, and this difference can be correlated with the disappearance and reappearance of the ability to form acetolactate as shown by chromatography of the products of the reaction. The inability of S. cerevisiae to grow on the acetohydroxyacids, which would have been an additional proof of the nature of the mutants, is probably due to the fact that the compounds cannot enter the cells in the ionized form, whereas in the nonionized form they spontaneously decarboxylate [9]. The fact that even mutants lacking the R1 threonine deaminase will not accept acetohydroxybutyrate in lieu of isoleucine provides avidence for this interpretation. In the table the mutant's production of acetoin is equivalent to the production by the wild-type of acetoin-acetolactate combined. This might lead to the conclusion that the altered enzyme is functioning to make acetoin, rather than acetolactate, or that the mutant has acquired an acetolactate decarboxylase. The striking difference between the ph curves of total acetoin production (acid curves, Fig. 1) in Mi5 and M14R, seems to rule out even an abortive role for biosynthetic enzyme in the mutant. Furthermore, the mutation is recessive (one would expect the auxotrophy caused by a decarboxylase to be dominant), and direct assay for decarboxylase reveaas 110 such activity. It is worthwhile mentioning that there is no well-documented function for the acetoin-forming enzyme, although Juni has provided evidence that it is associated with and may be a side reaction of pyruvic decarboxylase. In any casc, the different kinetics as well as the genetic data show that acetoin formation is completely separate from acetolactate biosynthesis in yeast, in contrast t,o N. crasm and K. aerogenes, where acetoin is formed by decarboxylation of acetolactate. The kinetic characteristics of the two reactions are interesting. It has been shown that the two molecules of pyruvate required for acetoin synthesis are not equivalent since acetaldehyde can replace one of these molecules but not both [5]. Since the K, plot for pyruvate indicates a bimolecular reaction, the rate-limiting step must be the condensation rather than either of the two preliminary reactions pyruvate may be supposed to undergo. For acetolactate, on the other hand, the K, plot indicates a monomolecular reaction. Therefore the rate limiting step is probably not the condensation but may very well be formation of hydroxyethylthiamine pyrophosphate. All the revertants studied have regained the activity measured in the differential assay, an indication that this activity is the physiological one. Experiments to be described in the following paper of this series concerning the inhibition of the enzyme by valine also evidence of its function in the biosynthetic pathway. As indicated in the results section, the evidence clearly indicates that there is oniy one acetolactate synthetase active between ph 5.5 and p1-i 9.0 in 8. cerevisiae. This is in contrast to the case in E. coli and N. crassa where there are two, one with an optimum near ph 6.0 and the other active at a more basic ph [7]. The functional reason for the two acetolactate-forming enzymes in these organisms is not yet clear, since only the one with a ph optimum near ph 8.0 seems to be active under normal conditions. This can best be shown in E. coli K12, where

506 NAGEE and ROBICHON-SZULMAJSTER: Isoleucine-Valine Biosynthesis in 8. cerewisiae. 2 European J. Biochem. inhibition of this enzyme by valine is the major cause of the well-known bacteriostatic effect of valine. Ramakrishnan and Adelberg [ 101, however, have been able to show that the enzyme with a ph optimum near 6.0 can function if one selects the mutants of the ph 8.0 enzyme by first resistance, and then sensitivity to a-aminobutyrate, a valine antagonist. The present paper, in conjunction with their results clearly indicates that the reason that mutants for acetolactate synthetase in the enterobacteria eluded selection for so long was the presence of two enzymes, since such mutants arise with apparently normal frequency in S. eerevisiae. This work has been supported by a grant from the DBIBgation GBnBrale & la Recherche Scientifique et Technique (61-FR-063). One of us (P. T. M.) was a postdoctoral fellow of the American Cancer Society, grant nr. PF 243. REFERENCES 1. Strassman, M., Thomas, A. J., and Weinhouse, S., J. Am. Chem. Soc. 77 (1955) 1261. 2. Lewis, K. F., and Weinhouse, S., J. Ana. Chem. Soc. 80 (1958) 4913. 3. Kakar, S. M., and Wagner, R. B., Genetics, 49 (1964) 213. 4. Hawthorne, D. L., Personal communication. 5. Juni, E., J. Bid. Chem. 236 (1961) 2303. 6. de Robichon-Szulmajster, H., and Magee, P. T., European J. Biochem. 3 (1968) 492. 7. Ramakrishnan, A. N., and Snell, E. E., J. Biol. Chem. 235 (1960) 2316. 8. Lowry, O., Rosebrougli, N. J., Farr, A. L., and Randall, R. J., J. BioZ. Chem. 193 (1951) 265. 9. Krampitz, L. O., Arch. Biochem. 17 (1958) 81. 10. Ramakrishnan, T., and Adelberg, E. A., J. Bacteriol. 89 (1965) 654. P. T. Magee s present address: Department of Microbiology Yale University School of Medicine New Haven, Conn., U.S.A. H. de Robichon-Szulmajster Laboratoire d Enzymologie du C.N.R.S. 91 Gif-sur-Yvette, France