Synthesis and Degradation of Fructose Diphosphate Aldolase Isoenzymes in Avian Brain*

The intracellular proteins of animal cells are continuously turning over; therefore, the concentration of a given protein is regulated both at the level of protein synthesis and at the level of protein degradation. Studies on the relative rates of turnover of isoenzymes, such as those of aldolase and lactate dehydrogenase, may help to clarify the mechanisms involved in protein turnover. The isoenzymes and subunit types are very similar proteins, and are located within the same intracellular compartments; yet, the concentrations of these proteins are independently regulated. The present paper describes the roles of synthesis and degradation in regulating aldolase isoenzyme concentrations in avian brain. as precursors of protein synthesis followed by isolation of the labeled isoenzymes by affinity chromatography or immunoprecipitation were performed to measure the relative rates of turnover of aldolases tetramers and subunit types. The results show that aldolases C, and AC, turn over at the same rates with apparent half-lives of about 4 Furthermore, a double isotope experiment showed that the of AC, turn same these are overestimates due to experiments, two experiments show that extensive isotope reutilization was not responsible for of essentially identical for and differences demonstrated by the employed


Synthesis and Degradation
of Fructose Diphosphate Aldolase Isoenzymes in Avian Brain* (Received for publication, December 16, 1974) HERBERT G. LEBHERZ~ WITH THE TECHNICAL ASSISTANCE OF ERIKA AB~CHERLI From the Institute for Cell Biology, Swiss Federal Institute of Technology, Hiinggerberg, CH-8049, Ziirich, Switzerland The intracellular proteins of animal cells are continuously turning over; therefore, the concentration of a given protein is regulated both at the level of protein synthesis and at the level of protein degradation. Studies on the relative rates of turnover of isoenzymes, such as those of aldolase and lactate dehydrogenase, may help to clarify the mechanisms involved in protein turnover. The isoenzymes and subunit types are very similar proteins, and are located within the same intracellular compartments; yet, the concentrations of these proteins are independently regulated. The present paper describes the roles of synthesis and degradation in regulating aldolase isoenzyme concentrations in avian brain. (a) Steady state concentrations (approximately 700 lg/g) of aldolase are present in chicken brain virtually during the entire life of the organism, and the relative levels of the isoenzymes do not change with age. (b) The aldolase protein of this tissue, as determined by ion exchange chromatography, is made up of 5% A and 95% C subunits. The relative levels of tetramers present in whole brain extracts, C, (81%), AC, (17%), A&, (24) o are quite similar to those predicted for the random association of 5% A and 95% C subunits. (c) Single and double isotope-labeling experiments using [3H]and ["Clleucine as precursors of protein synthesis followed by isolation of the labeled isoenzymes by affinity chromatography or immunoprecipitation were performed to measure the relative rates of turnover of aldolases tetramers and subunit types. The results show that aldolases C, and AC, turn over at the same rates with apparent half-lives of about 4 days. Furthermore, a double isotope experiment showed that the A and C subunits of aldolase AC, turn over at the same rate. (d) Although these half-lives are overestimates due to reutilization of the labeled precursors during the experiments, two control experiments show that extensive isotope reutilization was not responsible for the calculation of essentially identical half-lives for the isoenzymes and subunit types. (The specific radioactivity of the free amino acid pool rapidly decreased after administration of [3H]leucine; and large differences in the rates of turnover of other brain proteins were demonstrated by the methods employed here.) The present work suggests that the aldolases and subunit types have very similar rate constants for degradation and, therefore, that the degradative mechanisms of brain cells do not distinguish between isoenzymes or subunit types when selecting aldolase protein for degradation.
It follows that, although regulation at the level of degradation is important in determining total amounts of aldolase protein, regulation at the level of subunit synthesis (transcription, translation) is responsible for determining the relative levels of A and C subunits, which generate the tissue-specific isoenzyme pattern of avian brain. These conclusions are discussed in relation to the model proposed by Fritz and associates on the regulation of lactate dehydrogengse isoenzyme concentrations in animal cells.
The intracellular proteins of animal cells are continuously being catabolized and resynthesized, a process known as * This investigation was supported by the Swiss National Science Foundation Grant 3.8640.72. Preliminary accounts of part of this work have been presented at the Federation of American Societies of Experimental Biology meeting in Atlantic City, New Jersey (April, 1973)

and the Third International
Conference on Isozymes at Yale University, New Haven, Connecticut, April, 1974 (1, 2). $ Present address; Department of Molecular Biology, Roswell Park Memorial Institute, Buffalo, New York 14203. turnover (3,4). Although the detailed mechanisms involved in the turnover process are not well understood, several parameters have been shown to influence the rates of protein degradation. Among them are: the structural characteristics of the protein (3, 4), including subunit size (5,6); the properties of the cells in which the protein is localized (3,4,7); and the composition of the intracellular dehydrogenase and aldolase, may help to clarify the mechanisms involved in protein degradation.
Using these systems, possible variations in the rates of turnover of different subunit types based on differences in gross conformation and primary structure, on differences in subunit size, or on differences in intracellular localization are eliminated, since the subunit types are structurally homologous, are of the same size, and are found in association with each other within the cell (2,(9)(10)(11)(12)(13)(14). Thus, any preferential selection of the isoenzymes or subunits for degradation would be determined by relatively minor structural and conformational differences, and/or by differences in abilities or inabilities to interact with the components of the cell.
Fritz and associates (15,16) recently proposed that preferential degradation of different subunit types of lactate dehydrogenase is of major importance in regulating the relative levels of the isoenzymes in mammalian cells. Their model is of considerable interest, since it had previously been assumed (9,17) that regulation at the level of transcription is primarily, if not solely, responsible for maintaining the well known tissuespecific isoenzyme patterns of animal cells. The model of Fritz et al. was a major stimulus to initiate the present, studies on the in uivo turnover of aldolase isoenzymes and their subunits. Chicken brain was selected for study since it is a major vertebrate tissue which contains the easily separable isoenzymes belonging to the A-C hybrid set.

Methods
Analytical Methods-Aldolase activity was determined at 25O as previously described (19). Activities were expressed as micromoles of fructose-P, cleaved per min per ml of enzyme solution, and specific activities as units of activity per mg of protein.
Protein concentrations of crude extracts, resolubilized trichloroacetic acid precipitates, and immunoprecipitates were determined by the method of Lowry et al. (20). Concentrations of pure aldolase solutions were measured by absorbance at 280 nm, using the extinction coefficients (0.9) of rabbit aldolases A, and C, (13,21). Concentrations of dilute aldolase solutions were estimated from activity measurements using specific catalytic activities of 16 and 8 for A and C subunit protein, respectively (18).
Cellulose-polyacetate electrophoresis and staining for aldolase activity were performed as previously described (22). Polyacrylamide gel electrophoresis was done at pH 9.5 in 5% gels (23); electrophoresis was carried out with a current of 1.5 ma/gel and afterward, the gels were stained for proteins with 1% Amido schwarz or 0.25% Coomassie blue, followed by destaining in acetic acid/methanol/water, l/1/8.
Tritium and "C were counted in the Beckman LS-150 liauid  (2,(25)(26)(27)(28). In contrast to 10 to 13 ml of kquasol. Trichloroacetic acid precipitates were dissolved in 1 N NaOH, and 0.4-ml aliquots were counted in 10 ml of Aquasol; muscle or other tissues, which contain aldolases with large prior to counting, the NaOH was neutralized by the addition of 0.6 ml numbers of A subunits (a), all extractable aldolase of chicken of 1 N HCl. In single label experiments, [3H]toluene standards were brain (approximately 700 pg/g wet weight) could be solubilized used to convert counts per minute to disintegrations per minute.
with buffers of low ionic strength.
The relative amounts of the isoenzymes in brain extracts were determined by ion exchange chromatography on DEAE-Sephadex (Fig. 1). The levels of isoenzymes were found to be: C, (81%), AC, (17%), and A&, (2%). Only trace amounts of aldolases A,C or A, were ever detected. Since aldolase subunit associations are very stable irz uitro (18), it may be assumed that the relative levels of isoenzymes observed were the same as those within the intact brain. From this tetramer distribution, it was calculated that A and C subunits comprise 5% and 95%, respectively, of the aldolase protein.
Brain tissue is quite heterogeneous (29) and, therefore, the isoenzyme content of whole brain extracts represents the sum contributions of all brain regions and cell types present in the intact tissue. Some differences in relative proportions of isoenzymes and in specific aldolase activity were observed when comparing different brain regions. For example, the ventral cerebrum contained relatively higher levels of A&, and lower levels of C, as compared with the other brain regions tested (medulla, optic lobe, cerebellum, and dorsal cerebrum).
Since aldolase subunits appear to associate in a random fashion (30), comparisons between the relative levels of isoenzymes observed in tissue extracts and those expected assuming random combination of subunits from a single subunit pool may give indications of the degree of cellular heterogeneity, with respect to isoenzyme content, of animal tissues (31). If observed and predicted values are the same, then tissue heterogeneity may not be a problem in interpreting turnover data. In the present studies, good agreement between observed and predicted values was obtained. However, since the isoenzyme distribution of chicken brain is highly skewed toward aldolase C, (Fig. l), the possibility that significant differences in the isoenzyme patterns of the two predominant cell types of brain (neurons and glial cells) may exist could not be eliminated using such calculations.
In view of these observations, cellular heterogeneity with respect to isoenzyme content must be considered in interpreting the turnover data described below, since the purpose here was to compare the rates of turnover of aldolase isoenzymes within the same cell(s).
There are advantages to performing the present turnover experiments on young chicks. Because of the smaller body weight and higher brain weight to body weight ratio, approxi- mately loo-fold lower levels of radioactive amino acids are required to obtain the same extent of labeling of brain proteins of 2-to a-week-old chicks as compared with those required for adult chickens. In using still developing animals, however, it must be demonstrated that steady state concentrations of the proteins under study are present at an early age. The relationship between aldolase isoenzyme content of brain and age (body weight) was therefore determined.
As shown in Fig. 2, aldolase activity/g of brain has reached a constant maximal level prior to 1 week of age (body weight, GO to 70 g). Specific catalytic activity has also reached the steady state level (0.145 to 0.155 unit/mg of soluble protein) by this time, and no changes in relative concentrations of the isoenzymes were indicated by electrophoresis.
Thus, steady state concentrations of the isoenzymes have been accumulated in brain prior to 2 weeks of age, and young chicks were used for the turnover experiments described below. Since the brain did show considerable growth during the first 2 to 3 weeks after hatching, the brain weight versus body weight relationship in Fig. 2 was used to correct for increases in total aldolase content (brain mass) during the experiments.
Isolation of Brain Aldolase Zsoenzymes-In the turnover experiments described below, aldolases AC, and C, were isolated by one of two purification procedures. In the affinity chromatography method, the aldolases were specifically eluted from phosphocellulose by fructose-P, (see "Methods" for details). Elution profiles of the Whatman (containing aldolase AC J and Schleicher-Schuell (containing aldolase C ,) phosphocellulose columns are shown in Fig. 3. The two aldolase preparations thus obtained were tested for purity by electrophoresis. Fig. 4 (upper) compares the activities in the two preparations with those in a crude brain extract. Polyacrylamide gel electrophoresis showed that the C, preparation was homogeneous, while the AC, preparation contained only small amounts of contaminating aldolases (A,C, and C,). Densitometry tracings of the gels showed that contamination of aldolase AC, by the other isoenzymes was usually about 10% (Fig. 4, lower).
In the immunoprecipitation method, aldolases AC, and C, were partially purified by ammonium sulfate fractionation (see "Methods") and separated from each other by chromatography on DEAE-Sephadex, as described in Fig. 1. The isoenzymes were then specifically precipitated with antiserum directed toward aldolase C,. Ouchterlony double diffusion tests with the antiserum are shown in Fig. 5. A single precipitin line was produced with crude brain extracts which completely fused with the lines produced with pure aldolases AC, and C,. No precipitin line formed with crude muscle extracts (containing aldolase A,). The antiserum, therefore, appears to be monospecific for aldolase C subunit protein, and recognizes FIG. 3. Isolation of aldolases AC, and C, by substrate elution from phosphocellulose. The brain aldolases were partially purified as described under "Methods." The aldolase preparation was applied to a Whatman (upper) phosphocellulose column (1.8 cm x 10 cm) and the large breakthrough peak (containing aldolase C,) was applied to a Schleicher-Schuell (lower) column (1.8 cm x 14 cm). The two columns were developed as described under "Methods." The aldolases were eluted with fructose-P, (B) and remaining proteins were eluted with 2 M NaCl (C) (see "Methods" for details). Purity of brain aldolase preparations. Aldolases AC, and C, were isolated as described in Fig. 3. Upper, cellulose polyacetate electrophoresis followed by staining for aldolase activity. a, Crude brain extract; b, aldolase AC,; c, aldolase C,. Lower, polyacrylamide gel electrophoresis. The gels were stained for protein with Amido schwarz and were then scanned at 610 nm. ---, AC, preparation, -, C, preparation.

UNITS OF ALDOLASE ADDED
FIG. 5. Characterization of rabbit antiserum directed toward chicken aldolase C,. Upper, Ouchterlony double diffusion tests. Outer wells contain crude brain (B), crude muscle (M), or purified aldolases, as indicated. Aldolase concentrations in all preparations were between 0.15 and 0.20 mg/ml. The precipitin lines were stained for protein or for activity as described under "Methods." Lower, immunotitration of chicken brain extracts. Tubes contained 0.5 ml of antiserum and increasing volumes of diluted brain extracts. All volumes were adjusted to 3 ml with 10 mM Tris-Cl, and all samples contained 4% NaCI. The tubes were incubated at room temperature for 1 hour and overnight at 4". After centrifugation at 3000 x g for 15 min, the supernatant fractions were assayed for aldolase activity. The precipitate fractions were washed in the above buffer, including 4% NaCl and assayed for protein as described under "Methods." AC, and C, with apparent immunological identity. The antigen-antibody complexes are enzymatically active, as shown by direct staining of the precipitin lines for aldolase activity (Fig. 5, upper right). A typical immunotitration curve using the antiserum and crude brain extract (Fig. 5) showed that each ml of antiserum precipitated approximately 0.54 unit (67 bug) of brain aldolase. Control experiments, in which unlabeled aldolase C, was added to [3H]leucine-labeled liver extracts (containing labeled aldolase B,), followed by immunoprecipitation with the antiserum and washing of the immunoprecipitate, showed that all radioactivity was removed from the antigen-antibody complexes. Therefore, the antiserum could be used to precipitate specifically aldolases containing C subunits from solutions containing mixtures of proteins.

Time Course of [SH]Leucine
Incorporation into Aldolases AC, and C,-The rates of turnover of aldolases AC, and C, in chicken brain were first investigated by measuring the time courses and extents of labeling of the isoenzymes after administration of [3H]leucine.
As shown in Fig. 6, incorporation of isotope into both aldolases followed similar time courses with incorporation into both enzymes tapering off after about 2 hours. Taken alone, this observation suggests that the two isoenzymes turn over at similar rates, since the time course of isotope incorporation is related to the rate of protein turnover (5,32). However, the fact that AC, became labeled to a considerably higher extent than C,, in itself, suggests that AC, turns over more rapidly than C,. Since A and C subunits contain approximately the same number of leucyl residues per polypeptide chain (33, 34), the greater extent of labeling of aldolase AC, cannot be explained as reflecting a higher specific activity of leucine in A as compared to C subunits. One possible explanation for the conflicting data of Fig. 6 may be that the isoenzymes turn over at similar rates, but that different portions of the different isoenzymes were derived from different cell types (see above). In this case, differences in the specific radioactivity of leucine at the sites of protein synthesis in the different cell types may be responsible for the different extent of labeling of the two aldolases.
Turnover of Aldolases AC, and C,-The rates of turnover of aldolases AC, and C, were next investigated by measuring the decline in specific radioactivity of the isoenzymes after administration of [3H]leucine.
Since difficulties in the isolation of aldolase C, at some times were encountered, reliable data for the turnover of this isoenzyme were not obtained in this experiment.
The data for AC, are given in Fig. 7. As expected from what is known about protein degradation in animal cells (3-5), the decline in specific radioactivity followed first order kinetics. The open circles show the data prior to correction for increase in brain mass during the experiment, while the closed circles show the same data after correction for the increase in brain mass (i.e. total aldolase activity), using the brain weight versus body weight relationship of Fig. 2. From the slope of the corrected curve, an apparent half-life of 4.2 days was calculated for aldolase AC,.
The relative rates of turnover of aldolases AC 3 and C 4 were then determined using the double isotope method described by Arias et al. (35). Using this method, relative rates of turnover of proteins can be accurately determined in a single animal. The animal receives one isotopic form of an amino acid (in this case, [3H]leucine) followed some time later by the same amino acid labeled with another isotope (['%]leucine).
The proteins are then isolated and their 14C:3H ratios determined. Since 2 4 6 TIME (hrs) FIG. 6. Time course of [3H Jleucine incorporation into aldolases AC, and C, of chicken brain. Chicks received 174 PCi of [3H]leucine/100 g of body weight. At the times indicated, two chicks were taken, and the brain aldolases were isolated by affinity chromatography as described under "Methods." Concentrations of aldolase protein were estimated by activity measurements, as described under "Methods" in order to calculate disintegrations per min/mg. 5971 proteins which turn over rapidly are synthesized faster and degraded faster than proteins which turn over more slowly, the former will have higher 14C: 3H ratios than the latter (35). Also, any differences between the ratios of the proteins will be increased as the time interval between administration of the two isotopes is increased; no differences in the ratios would be observed if the isotopes were administered simultaneously. In the first series of experiments, chicks received [3H]leucine followed either 2, 5, or 9 days later by ["Clleucine.
One day after the last injection, the brain aldolases were isolated by affinity chromatography, and their 1rC:3H ratios determined. The data are given in Table I. Since different amounts of isotope were given to each group of chicks, the important values here are the normalized ratios for the two isoenzymes isolated from each group. These values are quite close to those expected for proteins that turn over at the same rate (1.00). For comparison, note the very different values that are predicted assuming that aldolase AC3, with an apparent half-life of 4.2 days, turns over twice as fast as aldolase C, (last column).
Purification of the isoenzymes by affinity chromatography results only in the isolation of functional aldolase tetramers (tetramers capable of binding substrate) and, therefore, the experiments described above were concerned with the turnover of active aldolase molecules. Another double isotope experiment was performed to determine the relative rates of turnover of the isoenzymes using immunological competence as the criterion for aldolase isolation.
The data at the bottom of Table I suggest that immunologically competent forms of aldolases AC 3 and C, also turn over at similar rates.
Turnover of A and C Subunits of Aldolase AC,-It is now known that subunits which comprise the same oligomeric structure may turn over independently of each other (36-38). Therefore, if aldolase A and C subunits turn over independently and at different rates, the measured rate of turnover of aldolase AC, would be largely a measure of the turnover of C subunits. This is apparent from the subunit composition of aldolase AC,. Consider the situation (Fig. 8) in which it is assumed that C subunits, whether present in homo-or heterotetramers, turn over half as fast as A subunits. The apparent half-life of AC, (4.2 days) would be quite close to that of C, (4.8 days), while the differences in half-lives of A (2.4 days) and C (4.8 days) subunits would be readily discernible (Fig. 8). A double isotope experiment was performed to clarify this point. Rate of turnover of chicken brain aldolase AC,. Chicks received 80 PCi of [3H]leucine/100 g body weight and at the times indicated, the brains from two chicks were taken for analysis. Aldolase AC, was isolated by affinity chromatography, and the disintegrations per min/mg of protein determined from catalytic activity and radioactivity measurements (see "Methods"). 0---0, data uncorrected for increase in brain mass during the experiment; O-0, data corrected for increase in brain mass using the brain weight versus body weight relationship of Fig. 2. On the 7th day, the aldolases were isolated by the immunoprecipitation method. Incubation mixtures contained approximately 120 pg and 130 pg of aldolases AC, and C,, respectively, and 2.0 ml of antiserum.
Final volumes were adjusted to 14 ml, and the samples contained 4% NaCl. Incubations were performed as described in Fig. 5. The washed immunoprecipitates were taken up in 0.5 ml of water and assayed for SH and "C content, as described under "Methods." As shown in Table II, the 14C:3H ratios of the A and C Five chicks each received [3H]leucine followed 6 days later by ['"Clleucine.
On the 7th day, the brain aldolases were subunits of aldolase AC, were essentially identical, demon-isolated by affinity chromatography. As shown in Fig. 9, considerable contamination of the AC, preparation by aldolase C, (33%) occurred.' The 14C:3H ratios of the two isoenzyme preparations (0.89) were found to be identical. The A and C subunits of the AC, preparation were then separated as follows. The labeled aldolase was added to a 39-fold molar excess of unlabeled aldolase A,, and the tetramers were dissociated to subunits by titration to pH 2.3 (18,30,39,40). Reassociation to tetramers was then effected by titration to pH 8 with complete recovery of activity (see legend of Fig. 10 for details). Because of the large excess of unlabeled A subunits present, the labeled A and labeled C subunits were sequestered into different tetramers (A,A* and A&*, respectively) during the reassociation process. The two tetramers containing the labeled subunits were then separat,ed by chromatography of DEAE-Sephadex (Fig. 10). The fact that only aldolases A, and A,C were detected after reassociation demonstrates the complete randomization of subunit associations as a result of the pH 2.3 treatment.
No cross-contamination of the A, and A,C preparations, and hence, of labeled A and C subunits, was indicated by electrophoresis (Fig. 10). The two subunit preparations were treated as described in Table II Table II. The brain aldolases were isolated by affinity chromatography and subjected to electrophoresis as described under "Methods." The gels were stained with Coomassie blue and scanned at 570 nm to determine the relative levels of the isoenzymes in the aldolase AC, preparation. FIG. 10 (right).
Separation of A and C subunits of aldolase AC,. Dissociation-Reassociation-The labeled aldolase AC, preparation (see Fig. 9) containing 0.65 mg of protein was added to a solution containing 25 mg of unlabeled aldolase A,. The sample was diluted to 40 ml with water, and was adjusted to 0. Five chicks each received 500 ICi of [3H]leucine followed 6 days later by 82 /*Ci of ["Clleucine. On the 7th day, the brain aldolases were isolated by affinity chromatography. A and C subunits of the aldolase AC, preparation were separated as described in Fig. 10. The two subunit preparations were diluted to equal volumes (19 ml) with 1 mM 2.mercaptoethanol, and adjusted to equal protein concentration (0.7 mg/ml) by addition of bovine serum albumin. The preparations were extensively dialyzed against 1 mM 2-mercaptoethanol, and then lyophilized to dryness. Residues were taken up in 1 ml of water and counted in 10 ml of Aquasol.
Subunit 3H "C "C:SH Relatively few studies on the turnover of brain proteins have been performed, and in most of these studies, the problems of isotope reutilization were not considered. Furthermore, the report by Stewart and Urban (45) that the cellular amino acid pool of mouse brain remained highly labeled for a long time after administration of [3H]leucine suggests that isotope reutilization in brain cells might be quite extensive. Consequently, control experiments were needed to show that t,he similar 14C:3H ratios of the aldolases and subunit types do reflect similar rates of turnover of these proteins.
First, the approximate time courses of disappearance of tritium from the total, trichloroacetic acid-soluble, and trichloroacetic acid-insoluble fractions of brain extracts (17,000 x g supernatant fraction) were measured. As shown in Fig. 11, the tritium in the trichloroacetic acid-soluble and -insoluble fractions were nearly equal at all times after [3H]leucine administration. However, when the soluble fractions were lyophilized, the tritium counts in the residues rapidly decreased with time. Less than 3% and 0.3% of the trichloroacetic acid-soluble counts were found in the residues after 1 and 7 days, respectively (Fig. 11) 11. Kinetics of disappearance of tritium from the total, trichloroacetic acid-soluble, and trichloroacetic acid-insoluble fractions of brain extracts (17,000 x g supernatant). Chicks received 35 PC1 of [3H]leucine/100 g body weight. At the times indicated, the brain from one chick was homogenized in 10 volumes of water, and the homogenate centrifuged at 17,000 x g for 30 min. For protein precipitation, % volume of 5OY" trichloroacetic acid was added to brain extract and, after 4 hours at 0", the soluble and insoluble material was separated by centrifugation at 3000 x g for 10 min. The soluble fraction was extracted three times with 2 volumes of ether to remove the trichloroacetic acid. Aliquots of the aqueous phase were lyophilized to dryness and the residues taken up in water. The precipitates were washed b) suspension in 6 to 8 volumes of 10% trichloroacetic acid containing 5 mM L-leucine, and the precipitates were recovered by centrifugation as before. The washing procedures were repeated two more times, and the final pellet was disolved in N NaOH. All brain fractions were counted for tritium content as described under "Methods." cellular amino acid pool may not reflect the specific radioactivity and turnover of amino acids that serve as precursors of protein synthesis at the site(s) of protein synthesis; these precursors may actually bypass the large cellular amino acid pool altogether (48). Isotope reutilization should influence apparent half-lives of all soluble brain proteins, and, therefore, the detection of large differences in the relative rates of turnover of the soluble brain proteins would strengthen the argument that the brain aldolases do, in fact. turn over at similar rates. This was accomplished by an experiment which is analogous to those used to demonstrate a correlation between protein subunit size and rate of protein turnover in other systems (5,6,50,51).
One chicken received [ 3H ]-and ['*C lleucine simultaneously, and another received the isotopes 6 days apart. On the 7th day, the soluble brain proteins were dissociated to subunits with sodium dodecyl sulfate, and the subunits were separated according to size by gel filtration.
The results are shown in Fig.  12. As expected, no consistent differences in the 14C:3H ratios of the protein subunits were observed when the isotopes were administered simultaneously. However, large differences in the 14C:3H ratios were apparent when the isotopes were given 6 days apart and, as expected, larger subunits had higher ratios than smaller ones. Thus, differences in the rates of turnover of brain proteins were readily detected with the procedures employed here. protein of this tissue is composed of 5% A and 95% C subunits. Turnover experiments were performed to determine if regulation at the level of protein degradation was involved in maintaining the highly different steady state concentrations of the subunit types. Since brain is quite heterogeneous, both with respect to its regional and cellular composition, the measured relative rates of turnover of aldolase tetramers (AC, and C,) do not necessarily reflect the turnover of different isoenzymes within the same cells. However, the observation that A and C subunits of the AC, heterotetramer turn over at the same rate does reflect the turnover of the subunit types within the same cell(s); the subunits must have been associated with each other in the intact tissue, since rearrangement of aldolase subunit associations does not occur in tissue homogenates or during aldolase purification2 (18). Thus, these data suggest that the degradative mechanisms of brain cells do not distinguish between A and C subunits when selecting aldolase protein for degradation.
The present work defines the roles of synthesis and degradation in regulating aldolase isoenzyme concentrations in avian brain. Although the isoenzymes and subunit types are continuously being degraded and resynthesized, no net change in amounts of these proteins occurs with time. This observation necessitates that the rate of synthesis (VJ of each protein is equal to its rate of degradation (V,). Synthesis follows pseudozero order kinetics, so V, can be expressed in terms of a zero order rate constant for synthesis (K,). Degradation follows pseudo-first order kinetics, and V, can be expressed as the product of a first order rate constant for degradation (K,) and the steady state level (P) of the isoenzyme or subunit type (3,4). Thus, the steady state level (P), = K,/Kd, where K, = In 2/t '/L (3, 4). Since the half-lives of the isoenzymes and subunit types calculated here (4.2  to define numerical values for K, and K,. The important question concerns the relative rates of turnover of the subunits and isoenzymes. Since aldolase A and C subunits apparently have the same rate constants for degradation, and since the concentration of C subunits is 19-fold higher than that of A subunit, it follows that, per unit of time, 19 C subunits are degraded and resynthesized for every A subunit which turns over. It can be calculated from the present work (Table I and Fig. 8) that at least 1 x 1Ol4 A and 19 x 10" C subunits are synthesized and destroyed per gram of brain each day. Although regulation at the level of degradation is involved in determining total aldolase protein concentration, it is suggested that regulation at the level of subunit synthesis (transcription, translation) is predominantly, if not solely, responsible for determining the relative levels of aldolase isoenzymes and subunit types in the avian brain. Whether or not this conclusion can be generalized to explain the regulation of aldolase isoenzyme concentrations in other tissues, or in tissues of other organisms must await the outcome of additional experiments.
In contrast to the above emphasis on regulation at the level of subunit synthesis for the control of aldolase isoenzyme concentrations, Fritz and associates have proposed that regulation at the level of degradation is of major importance in controlling the relative concentrations of lactate dehydrogenase isoenzymes in mammalian cells (15,16). Their reports that lactate dehydrogenase B subunits turn over very much faster than A subunits in skeletal muscle, while A subunits turn over faster than B subunits in cardiac muscle, pointed to a high degree of specificity in the selection of subunit types for degradation, and suggested that this specificity was determined by the characteristics of the cells in which the isoenzymes were localized. However, an entirely different interpretation of their data has recently been published (31), which suggests that the nonrandom distribution of lactate dehydrogenase subunits and differential turnover of the isoenzymes reported by these workers (15,16,52,53) may simply reflect cellular heterogeneity, with respect to isoenzyme content, of the tissues studied; different rates of turnover of lactate dehydrogenase in different cell types of the same tissue would not be unexpected, since rates of turnover of proteins are partially determined by the properties of the cells in which the proteins are localized (3,4,7). Consequently, it is not yet known whether or not mechanisms other than those that control rates of subunit synthesis are involved in the regulation of relative concentrations of the isoenzymes within the same cell. Correlations between rates of protein turnover in vivo and susceptibility to proteolytic inactivation and digestion in vitro has been observed (6,54,55). Also, it has been suggested that rates of protein turnover may be determined by the rates at which proteins spontaneously denature within the cell (56). If these observations and suggestions truly bear on the turnover process, then it would not be totally unexpected to find different isoenzymes of aldolase or lactate dehydrogenase turning over at different rates within the same cell. For example, different isoenzymes of lactate dehydrogenase show different susceptibilities to proteolytic inactivation in vitro, and with some isoenzymes, this susceptibility may be modified by the coenzyme NAD (57). Also, different isoenzymes of aldolase3 and lactate dehydrogenase (58) show different sus-ceptibilities to heat denaturation and, in the case of lactate dehydrogenase, certain metabolites protect some isoenzymes z2, from denaturation, but not others (58). Meaningful correla-27, tions between rates of degradation of isoenzymes in viva and the behavior of isoenzymes in vitro may yet be found. If so, 28. such correlations may help to clarify the mechanisms involved in intracellular protein degradation. 29.