Two Forms of Acetoacetyl Coenzyme A Thiolase in Yeast II. INTRACELLULAR LOCATION AND RELATIOSSHII’ TO GROWTH*

SUMMARY Two forms of acetoacetyl-CoA thiolase with different properties have been found in Saccharomyces cerevisiae. The enzyme with isoelectric point pH 5.3 (thiolase 5.3) is found in the cytosol while that with isoelectric point pH 7.8 (thio-lase 7.8) occurs in the mitochondria. When yeast are grown on glucose under aerobic conditions, thiolase activity shows two increases in activity during the growth period. Early in log phase, there is an increase in thiolase activity which sub-sequently reaches a plateau level. A second increase in activity at mid-log phase attains a peak value before decaying in late log to stationary phase. The biphasic curve is also observed during anaerobic growth on glucose or aerobic growth on galactose, indicating that removal of catabolite repression by glucose is unrelated to the rise in activity during midlog phase growth. A method for distinguishing the activities of both forms was developed. The early log phase activity contains predominantly thiolase while the midlog phase increase in activity is due to thiolase on galactose, which provides higher levels of mitochondria in early log phase, indicated increased thiolase

The enzyme with isoelectric point pH 5.3 (thiolase 5.3) is found in the cytosol while that with isoelectric point pH 7.8 (thiolase 7.8) occurs in the mitochondria.
When yeast are grown on glucose under aerobic conditions, thiolase activity shows two increases in activity during the growth period.
Early in log phase, there is an increase in thiolase activity which subsequently reaches a plateau level.
A second increase in activity at mid-log phase attains a peak value before decaying in late log to stationary phase.
The biphasic curve is also observed during anaerobic growth on glucose or aerobic growth on galactose, indicating that removal of catabolite repression by glucose is unrelated to the rise in activity during midlog phase growth.
A method for distinguishing the activities of both forms was developed. The early log phase activity contains predominantly thiolase 7.8, while the midlog phase increase in activity is due to thiolase 5.3. Growth on galactose, which provides higher levels of mitochondria in early log phase, indicated increased thiolase 7.8 activity.
Since the same enzyme appears in anaerobic growth, it indicates that thiolase 7.8 is also located in the promitochondria.
Yeast grown aerobically on glucose in the presence of d-threochloramphenicol, an inhibitor of mitochondrial development, exhibited no thiolase 7.8 activity, indicating that the expression of this enzyme is related to mitochondrial development. Thiolase 5.3 was not affected.
The possibility that thiolase controls the level of acetoacetyl-CoA and acetyl-CoA, negative and positive effecters of pyruvic carboxylase, is discussed.
III the preceding paper in this series we described some of the rnzymntic characteristics of the two forms of acetoacetyl-CoA  (1). The two forms of the enzyme are designated thiolase 5.3 and thiolase 7.8, the postscripts referring to their respective isoelectric points.
Prior to the work reported here, subcellular fractions of yeast which had been well characterized were assayed for thiolase activity by Xddleton and Apps (2). These workers showed that there is acetoacetyl-Cob thiolase activity associated with mitochondrial, cytosol, and nuclear fractions of Saccharomyces cerevisiae but not with the microsomal fraction.
In addition t,hcy demonstrated that the subcellular distribution of P-hydroxy-fl-methylglutaryl-Coh synthase resembles that of acetoacetyl-(loA thiolase.
We have investigated the occurrence of both thiolases in yeast cytosol and mitochondria.
Rather early in our studies we fouud that the level of acetoacet.yl-CoA thiolase activity in yeast was related to the growth status of the organism.
In this respect the enzymic act&it? appeared to resemble some yeast mitochondrial functions which are sensitive to glucose or catabolite repression (3). Since acetyl-CoA and acetoacetyl-Cob both regulate the gluconeogenie enzyme, pyruyate carboxylase (4), acetoacetyl-CoA thiolase provides a connecting link between carbohydrate, lipid, and energy metabolism.
We felt, t,herefore, that it was important to investigate the variation in thiolase activity during yeast growth as part of a study of the physiological roles of the two enzymes.
S. cerevisiae ATCC 9896 were grown in a microfermentor in standard medium (I). For studies on the intracellular locations of the two forms of acctoacctyl-Cod thiolase, yeast mere grown on medium containing 0.8% glucose to early stationary phase.
Yeast mitochondrial and cytosol fractions were prepared by a modification (2) of the method of Duell, Inoue, and I-tter (5). The modified method uses mannitol instead of sorbitol to stabilize the subcellular fractions, and breaks the sphcroplasts with a Potter-Elvehjem homogenizer instead of a vortex mixer. For the preparation of spheroplasts, 15 g of yeast, harvested in early stationary phase, were suspended in 50 ml of 0.2 M 2.mercaptoethanol, 25 mM EDT& pH 8.0. The suspension was incubated for 30 min at 30" and centrifuged.
The superuatant solution was tlicn centrifuged at 9,000 X g for 10 min. The pellet was resusl~entlcd in 0.6 AI mamlitol, 1 m&l I~XY~A, 20 mitI Tris, l)II 7.2, cclltrifugcd again at 1,000 X g for 5 miri followed by ccntrifugatioii at 9,000 x g for 10 min. The pellet sediment& between 1,000 x g and 9,000 x g is referred to as the mitochondrial fraction. The supernntant solution from the original 9,000 x g ccntrifugntion n-as centrifuged at 48,000 x g for 30 min. The 48,000 x g supernatant solution is referred to as the cytosol.
Small quantities of yeast were available for studies involving the time course of thiolase appearance during the early portion of the growth curve. Samples of yeast harvested from the microfermentor weighed less than 2.5 g. These samples were diluted to 5 ml nith water, 0.5 ml sf toluenc was added, and the samples were then incubated at 37" with shaking.
Autolysis was allowed to take place for 3.5 hours and was stopl)cd 1)~ centrifugstion of the remaining whole cells and cell debris. All other methods used were described in the first paper in this series (1)

Intrucellulur
Localion--Middleton and Apps (2) sur\eyctl subcellular fractions of yeast for acetoacetyl-Cob thiolasc :iud found activity in both the cytosol and mitochondrial frartiolls. They did not determine whether the enzymatic activities I\-erc caused by the same enzyme.
t'ytoplasmic and mitochondrial protein were subjecicd to inoclrctric focusing as shown in Fig. 1. Thiolase n-as released from the mitochondrial fraction only after sonication. The mitochondrial fraction (about 25 mg of protein), suspentlcd in 6 ml of 50 InM Tris, pH 8.2, was sonicated for 30 set with a Bronson Sonifier at full power. After that time thiolase activity was no longer pelleted by centrifugation for 10 min at 9,000 x g; in addition there was a 50% increase in the total thiolase activity. Thiolase 5.3 is the dominant species in yeast cytosol (Fig.  1A). The ratio of thiolase 5.3 to thiolase 7.8 in the c*>?osol preparation is 5.4, wherea,s the same yeast, when lyscd with toluene followed by electrofocusing, resulted in a thiolasc 5.3 to thiolase 7.8 ratio of 1.7 (1). Fig. 1B shows quite clearly that the only thiolase activity associated with the yeast mitocholldrial fraction is thiolase 7.8. Thiolase 5.3, except for :I small amount of contamination, is absent from the mitochondrial fractiou.
Cross-contamination of the subcellular fractions was invcstigated using cytochrome oxidase as a mitochondrial marker anal glucose 6-phosphate dehydrogenase as a cytosol marker. 11'~ were not able to detect any contamination of washed mitocholldria with cytosol in our best preparation.
Similarly, that mitochondrial fraction (not shown) contained no detrctablc thiolase 5.3. The cytosol, after centrifugation at 48,000 x g for 30 min, contained no cytochrome oxidase activity.
-1s ran be seen in Fig. L4  Growth conditions were standardized in a New 13runswick microfermentor. Fig. 2 shows the growth curve for X. cerevisiae ATC'C 9896 grown aerobically on glucose. Plotted in the same figure is the thiolase activity of yeast lysates determined during the growth experiment.
The biphasic growt'h curve is characteristic of catabolite repression (6). The total specific activity of the thiolase is not constant throughout the growth experiment.
After an initial period of 3 hours during which the enzyme is barely detectable, the specific activity rises to a level of 40 to 80 units per mg of protein.
This value remains constant until about midlog phase of the fermentative portion of growth and will be referred to as early enzyme. At about midlog phase there is a second increase in specific activity -. -.
--w 4. Thiolase activity during aerobic growth on galactose. Yeast were grown in continuous culture in a New Brunswick microfermentor on growth medium containing galactose instead of glucose. The yeast were harvested by siphoning and lysed under standard conditions.

Thiolase
Specific Activity (units per mg of protein) refers to the specific activity of the lysates. which reaches a peak value of about 150 units per mg of protein.
This peak occurs at the end of the log phase (13 hours) and will be referred to as peak enzyme.
The specific activity then decreases with further growth.

Relationship of Growth
Sfatus to Appearance of Early and Peale Enzymes-Yeast, when grown aerobically on glucose, do not show much respiratory activity as long as the glucose is present (7). In addition they show few complete mitochondrial structures (8). When the glucose in the growth medium is exhausted there is rapid development of both mitochondrial structures and respiratory activity.
Prior to the exhaustion of glucose from the growth medium the yeast grow fermentatively even though oxygen is present.
The first exponential portion of the growth curve in Fig. 2 represents fermentative growth while the second exponential portion represents oxidative growth. Yeast, when grown anaerobically, do not synthesize complete mitochondria but do elaborate promitochondria (9). The latter structures contain some mitochondrial enzymes but not others (10). Promitochondria appear to be more fragile than mitochondria when visualized with the electron microscope (9).
By growing yeast anaerobically, it should be possible to determine whether or not the appearance of peak enzyme is due to the conversion of promitochondria into mitochondria or de nouo mitochondrial synthesis. This was done and the results shown in Fig. 3. The growth curve is typical of fermentatively grown yeast. Interestingly, the thiolase activity curve of Fig. 3 is substantially the same as the comparable curve in the aerobic experiment of Fig. 2. It must be concluded, therefore, that the appcurance of the peak enzyme is not due to mitochondrial synthesis.
The inhibition of the expression of mitochondrial enzymes by glucose is referred to as catabolite repression.
It is responsible for the lack of respiratory activity of yeast grown on glucose. If the expression of the peak enzyme is due to a release from catabolite repression, growth of the organism on galactose (which is much less effective than glucose in causing catabolite repression (11)) should result in substantially different thiolase acCvit,y curve during growth.
Yeast were grown on galactose and thiolase activity was determined.
The results are shown in Fig. 4. It is clear that the over-all shape of the thiolase specific activity curve is similar to the glucose curves shown in Figs. 2 and 3. It is important to note that the peak enzyme activity still occurs and therefore is not a function of catabolite repression.
An interesting feature of the thiolase activity curve of the galactose-grown yeast is the fact, that the early enzyme activity is higher than in the glucose grown yeast as shown in Figs. 2 and 3. Since yeast harvested in early log phase on galactose contained well developed mitochondria, the increased early enzyme activity seen in Fig. 4 was probably due to the thiolase present in these st'ructures.
Similarly the early enzyme activity of glucose grown cells seen in Figs. 2 and 3 was probably due to thiolase present in promitochondria.
It is possible to inhibit both promitochondrial and mitochondrial development by growing yeast in the presence of high concentrations of D( -)-threochloramphenicol (la), an inhibitor of mitochondrial protein synthesis. When the yeast were grown in the presence of the ant,ibiotic, there was a total inhibition of the expression of early enzyme (Fig. 5). The expression of peak enzyme was not affected. It must be concluded that it is early enzyme that is associated with the promitochondria and Issue of July 25, 1971 J. A. Komblatt and H. Rudney 4427 mitochondria and not peak enzyme. The cause for the late appearance of the peak enzyme is not yet known.
Type of Thiolase in Early and Peak Enzyme Activity---Jsoelectric focusing was performed on both early and peak enzyme activities.
No activity was detected after isoelectric focusing of the early enzyme; this point will be dealt with later. TSOClectric focusing of the peak enzyme showed that at least 90% of the activity was thiolase 5.3.
In the preceding paper, thiolase 7.8 was shown to use thiol donors other than CoA much more efficiently thsn thiolasc 5.3 (1). Thiolase 7.8 uses dithiothreitol 69% as effectively as CoA. Thiolase 5.3, on the other hand, uses dithiothreitol only 11% as effectively as CoA. This fact was used to probe the nature of the thiolase occurring at different points in the growth cycle. The ratio of thiolase activity using dithiothreitol as substrate to thiolase activity using CoA as substrate was determined for a riumber of synthetic mixtures of thiolase 5.3 and thiolase 7.8. The dithiothreitol to VoA ratio was shown to vary linearly with the percentage of each thiolasc.
The dithiothreitol to CoA ratios were then determined in the anaerobic growth experiment, as shown in Table I Yeast were grown aerobically in a New Brunswick microfermentor. They were harvested and lysed under standard conditions. 2'hioZase Specific Activity (units per mg of protein) refers to the specific activity of the lysntes.
were able to determine the 1)erccntage of each thiolase by cornparison with the known standards.
It can be seen in Table I  that the majority of the early c~~zgmc is thiolase 7.8. The peak enzyme tends to be almost exclusively thiolase 5.3. To further substantiate the view that, thiolase 7.8 is the early enzyme, samples of thiolase taken from the peak enzyme in the growth curves of glucose-aerobic and glucose-chloramlhenicol experiments were subjected to isoelectric focusing.
The results arc shown in Fig. 6. The chloraml)henicol-treated yeast contain no thiolnse 7.8, whereas the yeast grown without chlorampheni-co1 do contain it. Although the percentage of thiolase i.8 in the untreated yeast is small, the value, being 4-fold greater than the limit of detection, is significant.
It should be cmphasized that the chloramphenicol-treated yeast contain absolutely no detectable thiolase 7.8 in this experiment.
Except for the chloramphenicol-treated yeast, we have not \-et observed a peak enzyme preparation that did Irot' contain thiolase 7.8. <'hloramphcnicol had IIO rffect on act irity whtln added directly to either thiolase 5.3 or 7.8.
As mentioned earlier, attempts had bcrt~ made to isoelcctritally focus the early enzyme but no activity was recovered on eluting the column.
If thiolasc    was inactivated or degraded on the column by proteolytic P'IIzymes, then addition of authentic thiolase 7.8 to early enzyme followed by focusing of the mixture should result in total degradation or inactivation of the added thiolase. This experiment was carried out and the results are presented in Table  II.
Of the 400 units of early thiolase added to the column, all but 15y0 of the enzyme was inactivated, This 15% eluted at the position of thiolase 5.3. In the experiment in which authentic thiolase 7.8 was added to early enzyme extract, all of the added thiolase 7.8 was inactivated plus 90% of the early enzyme. The results of this experiment explain why no thiolase 7.8 is seen when the early enzyme is subjected to isoelectric focusing. DISCUSSION In S. cerevisiae acetoacetyl-CoA thiolase is found in both the cytosol and mitochondria (2). The mitochondrial form and the cytosol form are not the same (Fig. 1). Cytosol thiolsse has an isoelectric point of 5.3, mitochondrial thiolase an isoelectric point of 7.8.
When yeast are grown on glucose, thiolase activity is not constant throughout the growth period (Fig. 2). Prior to midlog phase, the specific activity of the thiolase is 40 to 80 units per mg (early enzyme).
After midlog phase the specific activity increases to about 150 units per mg (peak enzyme).
We attempted to determine what causes this rapid increase in thiolase activity.
We reasoned that the increase in thiolase activity might be associated with either mitochondrial development or release from catabolite repression.
Therefore, yeast were grown under conditions that would prevent the conversion of promitochondria into mitochondria (Fig. 3) or prevent catabolite repression from occurring (Fig. 4). The results of these two experiments demonstrated unequivocally that the appearance of peak en-zyme was not related to mitochondrial development or catabolite repression.
When the orga.nisms are grown on galactose, the specific activity of the early enzyme was higher than when growth took place on glucose. This indicated that the rally enzyme might be associated with both promitochondria and mitochontlria.
Since it appeared that the early enzyme was associated with promitochondria, the yeast were grown in the presence of D (-)-threochloramphenicol.
This antibiotic prevents mitochoadrial and promitochondrial protein synthesis both in vitro (12) and in vivo (13). If the early enzyme were really present in the promitochondria, growth of the yeast on media containing chloramphenicol might inhibit t,he appearance of the early cnzyme. Fig. 5 shows this to be the case. When all early enzyme activity was abolished (Fig. 5), chloramphenicol had no effect on peak enzyme. We conclude that it is highly likely that the early enzyme is associated with both promitochondria and mitochondria, and that the peak enzyme activity is probably present in the cytosol.
It is not necessary to postulat,e that early enzyme is synthesized by the mitochondrial genetic apparatus, although this is a 1)ossibility.
It is just as likely that a component of the mitochondrial protein is necessary for inclusion of early enzyme in the mitochondrion.
In this case, thiolase which was synthesized on the cytoplnsmic ribosomes could not enter or attach to the mitochondrion and might be particularly susceptible to degradation.
A third possibility is that chloramphenicol inhibits the mitochondrial synthesis of a protein which is required to initiate some cytoplasmic protein synthesis.
Since thiolase 7.8 is found in the mitochondria (Fig. l), and since it is not found in chloramphenicol-treated yeast (Fig. 5), it is quite likely that it corresponds to early enzyme.
It was not possible to show by isoelectric focusing that early enzyme and thiolase 7.8 were the same. All of the evidence directly indicating that the early enzyme is thiolase 7.8 is provided by enzymatic assays based on the differential response of the true enzymes to thiol donors such as CoA and dithiothreitol. It is clearly shown in Table I that the activity in the early enzyme is primarily due to thiolase 7.8 while that in the peak is thiolase 5.3. .4tt,empts to isolate the early enzyme activity by electrofocusing were unsuccessful. This is probably due to proteolytic enzymes expressed durin g early growth since addition of purified thiolase 7.8 to early growth extracts also resulted in destruction of activity (Table II).
It is likely that yeast contain several proteolytic enzymes, and that the particular proteases present are determined by the growth phase, among other things. Manney (14) has shown yeast tryptophan synthetase is susceptible to proteolysis and that the sensitivity of this enzyme to degradation is a function of growth.
He has also shown that different proteases are present during each growth phase. A similar system may be present in the yeast strain used in this study.
Here, as mentioned earlier, there would be an additional complication in that mitochondrial protein may protect thiolase 7.8 from degradation late in growth, but not early when there are fewer mitochondria.
All of the preceding evidence supports the conclusion that the early enzyme is predominantly thiolase 7.8. On the basis of studies reported here and the properties of the two forms of acetoacetyl-CoA thiolase reported in the preceding paper (1) it seems reasonable to propose physiological roles for the thiolase as outlined in Fig. 7. In most instances definitive 4430 Location and Relation to Growth of Thiolases in Yeast Vol. 246,No. 14