Measurement of instant rates of protein degradation in the livers of intact mice by the accumulation of bestatin-induced peptides.

Bestatin induces the accumulation of di- and tripeptide intermediates in cellular protein breakdown. In liver, a single set of bestatin-sensitive cytosolic peptidases are involved in the degradation to amino acids of the major classes of cellular proteins. Accumulation of bestatin-induced peptides, in isolated hepatocytes, is proportional to the rate of protein degradation (Botbol, V., and Scornik, O. A. (1989) J. Biol. Chem. 264, 13504-13509). Injection of 1 mg of bestatin into mice results in detectable amounts of hepatic intermediates in 15 min. We propose to use the accumulation of these peptides as a relative measurement of liver protein degradation. There is at present no other way to determine transient changes in protein breakdown in the tissues of intact animals. As an example of the applications of this procedure, we present the effects of a single meal on hepatic protein metabolism. Protein synthesis was estimated by the incorporation into liver protein of a massive dose of radioactive leucine (Scornik, O. A. (1974) J. Biol. Chem. 249, 3876-3883) and degradation of long-lived or short-lived proteins by the accumulation of bestatin-induced peptides, labeled in carboxy-C of their Leu or Arg moieties, 1 day or 1 h beforehand. A single meal resulted in an 18% increase in liver protein in 8 h, a 45% increase in the rate of hepatic protein synthesis, and a 3-fold decrease in the rate of breakdown of long-lived proteins. Short-lived proteins were not affected. To establish the efficiency with which bestatin-induced peptides accumulate in the livers of fasting mice, we compared them with the disappearance, in 1 day, of protein-bound 14C-guanidino-Arg residues, labeled by previous injection of 14C-bicarbonate (Swick, R. W., and Ip, M. M. (1974) J. Biol. Chem. 249, 6836-6841). From this comparison, we estimated that bestatin-induced Leu-labeled intermediates, accumulating in 15 min, represented 39% of the hepatic proteins degraded in that interval. For Arg-labeled intermediates the value was 55%. Correcting for these efficiencies, we estimate that in 4 h a meal decreased the rate of degradation of long-lived Arg-labeled proteins from 2.02 to 0.73%/h. For Leu-labeled proteins the estimated rates were 1.76 and 0.66%/h, respectively. Although a transient slowdown of liver protein degradation after a single meal had been suggested before, this is the first time that acute changes such as this can be determined directly in intact animals.(ABSTRACT TRUNCATED AT 400 WORDS)


Measurement of Instant Rates of Protein Degradation in the Livers of Intact Mice by the Accumulation of Bestatin-induced
Bestatin induces the accumulation of di-and tripeptide intermediates in cellular protein breakdown. In liver, a single set of bestatin-sensitive cytosolic peptidases are involved in the degradation to amino acids of the major classes of cellular proteins. Accumulation of bestatin-induced peptides, in isolated hepatocytes, is proportional to the rate of protein degradation (Botbol, V., and Scornik, 0. A. (1989) J. Biol. Chem. 264, 13504-13509). Injection of 1 mg of bestatin into mice results in detectable amounts of hepatic intermediates in 15 min. We propose to use the accumulation of these peptides as a relative measurement of liver protein degradation. There is at present no other way to determine transient changes in protein breakdown in the tissues of intact animals.
As an example of the applications of this procedure, we present the effects of a single meal on hepatic protein metabolism. Protein synthesis was estimated by the incorporation into liver protein of a massive dose of radioactive leucine (Scornik,  Biol. Chem. 249,3876-3883) and degradation of longlived or short-lived proteins by the accumulation of bestatin-induced peptides, labeled in carboxy-C of their Leu or Arg moieties, 1 day or 1 h beforehand. A single meal resulted in an 18% increase in liver protein in 8 h, a 45% increase in the rate of hepatic protein synthesis, and a 3-fold decrease in the rate of breakdown of long-lived proteins. Short-lived proteins were not affected.
To establish the efficiency with which bestatin-induced peptides accumulate in the livers of fasting mice, we compared them with the disappearance, in 1 day, of protein-bound 14C-guanidino-Arg residues, labeled by previous injection of 14C-bicarbonate ( From this comparison, we estimated that bestatin-induced Leu-labeled intermediates, accumulating in 15 min, represented 39% of the hepatic proteins degraded in that interval. For Arg-labeled intermediates the value was 55%. Correcting for these efficiencies, we estimate that in 4 h a meal decreased the rate of degradation of long-lived Arg-labeled proteins from 2.02 to 0.73%/h. For Leu-labeled proteins the estimated rates were 1.76 and 0.66%/h, respectively.
Although a transient slowdown of liver protein degradation after a single meal had been suggested before, this is the first time that acute changes such as this can be determined directly in intact animals. This novel procedure should be of use to establish the relative * This work was supported by Grant AM13336 from the National Institutes of Health and Grant DMB84-15377 from the National Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
importance of nutrients and hormones in the regulation of hepatic protein turnover.
Mammalian liver, because of its high rate of protein turnover and its preferential access to absorbed dietary amino acids in the portal circulation, plays an important role in whole body protein metabolism. Transient changes in protein synthesis and degradation result in the storage of excess amino acids as additional liver protein after a meal, and in the utilization of the stored protein as an endogenous source of amino acids between meals (1,2). Studies in perfused livers (3) and isolated hepatocytes (4) have established that liver protein turnover is subject to regulation by hormones, notably insulin and glucagon, and the plasma concentration of amino acids. To find out the relative physiological importance of these and other regulatory factors, however, isolated systems are no substitute for experiments in intact animals.
Instant rates of liver protein synthesis can be measured in live animals by a variety of procedures (2), but there has been up to now no reliable way to measure instant rates of protein breakdown. In perfused tissues or in isolated or cultured cells, average rates of protein degradation are usually estimated by the release to the medium of radioactive amino acids from labeled cellular proteins, in the presence of chasing concentrations of unlabeled amino acids. In intact animals this principle cannot be applied because the amino acids produced by protein breakdown are either reincorporated into proteins or catabolized. Average rates of protein degradation in tissues of intact animals have been estimated either as the balance between rates of protein synthesis and net protein changes, or by disappearance of radioactivity from protein labeled with metabolically unstable amino acids (2). These procedures require repeated determination in different groups of animals, over periods long enough to permit measurable changes in protein content and protein radioactivity, usually a day or longer. Transient changes over shorter intervals are usually too small for their accurate estimation with a reasonable number of animals. In this paper, we propose a novel way in which instant rates of liver protein degradation can be estimated in mice.
Work from this laboratory has shown that bestatin, an inhibitor of cytosolic peptidases, induces the accumulation of di-and tripeptide intermediates in the degradation of cellular proteins in liver (5, 6) and other mammalian cells (7-9). In liver, we have presented evidence that a single set of bestatinsensitive peptidases are involved in the degradation of the major classes of cellular proteins and have suggested that accumulation of bestatin-induced peptides can serve as a relative measure of protein breakdown in intact mice. The conclusion is validated by the simultaneous determination of protein degradation and accumulation of bestatin-induced 15 min (5) and provide a unique opportunity to estimate instant rates of protein breakdown. AS a n example, we present in this paper diurnal variations in the accumulation of bestatin-induced peptides in the livers of meal-fed mice. We discuss the advantages and limitations of this novel procedure.
Animals-Adult male CD1 mice (30-35 g) were purchased from Charles River Breeding Laboratories. The meal-fed animals were kept, 5 to a cage, with water ad libitum and food was provided only during the first 4 h of darkness (6-10 p.m.). Three days later their weight was determined before and after the meal; only animals that gained 2 or more grams after the 4-h period were used on the fourth day.
Food Intake and Urinary Nitrogen-For the measurement of food intake and urinary nitrogen the animals were kept for an additional day in individual metabolic cages. Urine was collected with 50 pl of glacial acetic acid (as a preservative), diluted, and frozen until the assay. Urinary nitrogen (urea + ammonia) was measured by conversion of the urea to ammonia by the action of urease and colorimetric determination of ammonia (Sigma kit #640).
Liver Protein-Transient changes in liver protein content after a meal are relatively small. We were concerned that a larger portal blood flow after a meal could result in increased residual blood in the organ and give misleading results. For this reason, total protein was determined in bloodless livers prepared as follows. Mice were killed at the indicated times by the intravenous injection of 0.2 ml of saline containing 12 mg of sodium pentobarbital and 10 USP units of heparin. The abdomen and thorax were cut open, the abdominal cava was ligated, the porta was sectioned, and 50 ml of saline were perfused through a 23-gauge needle inserted in the thoracic cava. The bloodless livers were homogeneized in 0.3 M sucrose. Protein was precipitated from aliquots of the homogenate by hot 10% trichloroacetic acid (90 "C, 15 min), washed with organic solvents (ethano1:ethyl ether:chloroform, 2:2:1), redissolved in 1% sodium dodecyl sulfate, and measured by the Lowry colorimetric reaction using bovine serum albumin as standard.
Protein Synthesis-Mice were injected with a massive dose of L-[l-"C]leucine (91, 183, or 286 pmo1/100 g body weight, 0.01 pCi/ Fmol, intravenously). Animals were killed 5 min after the injection and incorporation of radioactivity into liver protein was measured. Protein synthesis was calculated from extrapolation of these values to infinity and the specific radioactivity of the injected amino acid (2, 10).
Bestatin-induced Peptides-Liver proteins were labeled by intraperitoneal injection of radioactive precursors. The nature of the precursor, the amount of radioactivity injected, and the times of injections are indicated in Tables I1 and 111. For labeling of leucine moieties, ~-[l-'~C]leucine was used. Arginine moieties were labeled with ~-[l-"C]ornithine. Because of the activity of urea cycle enzymes, ornithine was readily converted into carboxy-labeled arginine in the liver and incorporated as such into protein (5). Bestatin (0.2 ml, 5 mg/ml of saline) was injected intravenously and the mice were killed 15 min later. Based on our previous experience (5), the hepatic radioactive peptides were recovered as follows. Livers were homogeneized in enough ice-cold 0.3 M sucrose to make a final volume of 18 ml. Duplicate aliquots (0.18 ml) we used to determine protein radioactivity as in the previous section. The homogenate was then mixed with 2 ml of cold 100% trichloroacetic acid, allowed to precipitate in the cold for at least 1 h, and centrifuged (2000 rpm, 10 rnin). To eliminate carboyx-I4C from free leucine, a 5-ml aliquot of the acid supernatant was mixed with 0.85 ml of a 9% solution of ninhydrin in ethanol, and heated for 2 h at 90°C. Ninhydrin was then destroyed by adding 5 drops of HzOz (30 vol %) and heating 2 additional min. The solution was then extracted 3 times with 1 vol of water-saturated ethyl ether, and percolated through a column (15 X 6-mm inner diameter) of the cation-exchange resin AG-50W-X2 (200-400 mesh).
The peptides were eluted from the column with 4 ml of 1 N NH40H, dried in uacuo in a Speed-vac concentrator (Savant Instruments, Hycksville, NY), redissolved in 0.5 ml of 10% trichloroacetic acid, and its radioactivity determined by scintillation counting. Further purification of the peptides is possible (5) but not necessary in this case. Each experiment included control animals in which bestatin was omitted. The recovered radioactivity of these controls was subtracted from the experimental values to calculate the bestatin-induced peptide radioactivity.
Disappearance of Protein-bound "C-Guanidino-Arg Residues-To determine the extent to which bestatin-induced peptides accumulate in uiuo, we compared our procedure with that of Swick and Ip (11). These authors proposed the selective measurement of the guanidino-C of protein-bound arginine after labeling in uiuo with I4C-bicarbonate (11). In our experiment, mice were fasted for 22 h before the intraperitoneal injection of "C-bicarbonate (8 pCi). The radioactivity in the guanidino-C of protein-bound arginine was measured in two groups of mice, 24 and 48 h after the injection of 14C-bicarbonate (12). Liver protein was purified from liver homogenates by successive extractions with hot trichloroacetic acid, ethanokethyl ether:chloroform (2:2:1), acetone, and ethyl ether. The dried powder was dissolved in 6 N HCl and hydrolyzed in uacun for 36 h at 106 "C. Total amino acids in the hydrolysis were determined colorimetrically by the method of Rosen (13). and total arginine was calculated, from the known amino acid composition of liver proteins, as 4.16 mo1/100 mol of amino acids (14). The hydrolysate was dried in uacuo, excess C1was removed by addition of small amounts of the anion-exchange resin AG1 (bicarbonate form), and most amino acids were removed by percolation through a small column (5 X 6-mm inner diameter) of AG1 (OHform) (11). Arginine, because of the high pK of its guanidino group, passed through and was recovered in the effluent. Arginine was converted to urea by treatment with purified liver arginase, and in half of the sample urea was degraded to CO, and NH3 by further incubation with jack bean urease (12). Urea and radioactivity were determined in both samples, and the specific radioactivity of the I4C-urea (originally the I4C-guanidino group of the arginine) was determined by difference, as described earlier (12). Total radioactivity in protein-bound "C guanidino-arginine was calculated from this specific radioactivity and the total amount of arginine in the hydrolysate.

RESULTS AND DISCUSSION
Liver proteins were labeled in their leucine moieties by injecting the mouse with ~-[l-'~C]leucine and in their arginine moieties by injection of ~-[l-'~C]ornithine, which is readily converted to arginine in the liver because of the activity of the urea cycle enzymes. The use of carboxy-labeled precursors is required for the elimination of 14C label from free amino acids by acid-ninhydrin treatment (7). To study the degradation of short-lived proteins, bestatin (1 mg, intravenously) was injected 1 h after the radioactive precursor. For long-lived proteins, 24 h were allowed to elapse before the injection of bestatin. Measurable peptide radioactivity accumulates in 15 min ( 5 ) . In our previous work we have presented evidence that: ( a ) bestatin-induced peptides are intermediates in the degradation of cellular proteins. ( b ) They comprise di-and tripeptides ( c ) They result, at least in liver, from the inhibition of a single set of cytosolic peptidases affecting the degradation of cellular, but not extracellular proteins (6). ( d ) T h e effects of bestatin are not restricted to the degradation of a particular class of proteins. It induces accumulation of intermediates in the degradation of most hepatic proteins, including shortlived, long-lived, and abnormal proteins (6).
One corollary of these observations is that accumulation of bestatin-induced peptides could serve as a relative measure of liver protein breakdown in intact mice. T h e conclusion was validated directly in isolated hepatocytes, in which protein degradation and accumulation of peptides could be measured simultaneously. A substantial portion of the peptides (50% with leucine and 90% with arginine) were retained by the cells after a 30-min incubation (6). The reason for suggesting this experimental approach is that there is no other direct way to measure transient changes of protein degradation in tissues of intact animals (see Introduction). As an example of its application, we now present a study of diurnal variations of hepatic protein turnover in the livers of meal-fed mice.
Diurnal Variatwns in Hepatic Protein Content in Meal-fed Mice-Because feeding is discontinuous in most adult mammals, dietary amino acids must be stored as excess body proteins. The liver plays a major role in this storage. The hepatic storage of proteins is illustrated in Fig. 1. Mice are nocturnal animals; they eat primarily during the first few hours of darkness. If access to food is restricted to 4 h after light is turned off, they learn very quickly (2-3 days) to satisfy their daily food requirements in that period. Restriction of feeding to a well-defined daily meal is not drastically different from their spontaneous feeding pattern, but permits a sharp and controlled delineation of the diurnal cycle (15). In the 4h period, a 30-g mouse eats approximately 3 g of dry food containing 0.7 g of digestible protein. A large amount of these dietary amino acids is catabolized very rapidly. The digestion, absorption, and catabolism of excess amino acids is completed in a few hours, as shown by the pattern of urea excretion ( Four mice were placed in individual metabolic cages and trained to eat a single meal between 6-10 p.m. for 3 days. On the fourth day, the amount of food eaten was measured, and the urine was collected in the periods indicated; 2 h before the end of each period the mice received 2 ml of 0.3 M mannitol, to produce osmotic diuresis and thus favor the accurate timing of the urine samples. Average food intake in 4 h in this experiment was 3.10 z? 0.25 g, containing 680 mg of protein or 109 mg of protein N. Total urinary N excretion in the 10 h following the start of the meal was 60 mg, or 55% of the ingested protein N. Middle and bottom panels, liver weight and protein. These measurements were from separate experiments in which groups of 5 (weight) or 8 mice each (protein) were killed at the indicated times. For further details, see "Experimental Procedures." T o better visualize the diurnal nature of these changes, although the actual measurements are those represented by the vertical bars, the plot is repeated over a period of 2 days. imately 1.5 h (not shown); i.e. urea production peaks sooner than its excretion. The total amount of nitrogen eliminated in the urine during the 10-h period following the start of the meal in this experiment (60 mg) represents 0.38 g of catabolized amino acids, 55% of the ingested protein. Much of the other 0.3 g is probably stored as excess tissue proteins. Liver protein content increases in 8 h from 1.06 f 0.03 to 1.25 f 0.05 g/lOO g body weight ( p < 0.01; Fig. l), or about 0.06 g for a 30-g mouse. This excess protein is lost between meals when it serves as an endogenous source of amino acids.
The Increase in Protein Synthesis Is Insufficient to Account for the Accumulation of Liver Protein-After a protein-rich meal, hepatic protein synthesis is stimulated (16-21). This faster protein synthesis is largely or wholly due to a large proportion of ribosomes in polyribosomes (16). The increase in protein synthesis, however, is insufficient to account for the accumulation of liver protein, from which it has been inferred that protein breakdown must slow down (18,20). This is confirmed in Table I, where liver protein synthesis was measured by the injection of massive amounts of radioactive leucine. This procedure, previously developed in our laboratory, obviates the need to estimate the specific radioactivity of leucine in the precursor pool. The advantages of this procedure and its validation by comparison with the results of others was discussed in previous publications (2, 10). Liver protein synthesis increased in one experiment from 24.4 to 34.4 mg/h/100 g of body weight and in the other from 27.2 to 41.2 mg/h/100 g of body weight. The increase translates into the synthesis in 8 h of an excess of 80-112 mg protein/100 g of body weight. This is only about one-half of the actual gain in liver protein in this period, 192 mg/100 g of body weight (Fig. 1). The discrepancy between the protein gain and the increase in protein synthesis is even greater when we consider that only half of the newly synthesized proteins represent stable liver components. The other half comprises rapidly turning over or exported proteins (12, 21). In short, the stimulation of protein synthesis is not nearly enough to account for the liver protein gain after the meal; we must focus next on hepatic protein degradation.
Decreased Accumulation of Bestatin-induced Intermediates after a Meal-From the data just presented, it follows that the increase in protein synthesis is insufficient to produce by itself the rapid protein gain. A slowdown of protein breakdown

TABLE I Rates of protein synthesis in the liuers of meal-fed mice 4 h after the administration of food
Meal-fed mice were killed just before or 4 h after administration of the meal. Liver protein radioactivity (second column) was measured 5 min after the intravenous injection of a massive dose of ~-[ l -"Clleucine (286 pmo1/100 g body weight). Protein synthesis (third column) was estimated from this value, the specific radioactivity of the injected amino acid (0.01 pCi/pmol), the molecular weight of leucine (131), the molar proportion of leucine in liver protein (10.5% (13)), the extent to which it saturates the precursor pool at this dose (86%, determined previously by extrapolation to infinity of varying doses of leucine (2, 10) and confirmed in meal-fed mice in a separate experiment, not shown  (2,12), the loss of radioactive proteins in 8 h would be in the order of lo%, not large enough for accurate estimations with a reasonable number of animals. It is in transient situations, such as the effect of a single meal on hepatic protein metabolism, that the accumulation of bestatin-induced peptides proves most useful. After the meal, the accumulation of either arginineor leucine-labeled peptides decreases by a factor of three (Table 11, Experiments 1 and 2). Note that the radioactivity in bestatin-induced peptides, as a fraction of the corresponding protein radioactivity, is larger with arginine than with leucine. This is consistent with our previous observation, in isolated hepatocytes, that arginine-labeled peptides are retained more effectively within the cells (6).
One source of concern is that after the meal, bestatin may be less efficient in promoting the accumulation of peptides. This could occur because of changes in the extent of mitochondrial trapping of peptides (6), their leakage from the hepatocytes (6), the concentration of peptides in the cytosol, the activity of the peptidases, or their sensitivity to the inhibitor. To test possible differences in the efficiency of bestatin, we measured the accumulation of intermediates in the degradation of short-lived proteins, which should not be affected by the meal. In cultured cells and isolated hepatocytes, conditions that result in faster degradation of longlived proteins, such as serum or amino acid starvation, do not affect the rate of breakdown of short-lived proteins (22-24). Both classes of proteins are probably degraded by different mechanisms. Short-lived proteins may be predominantly broken down by the cytosolic ubiquitin-dependent system (25)(26)(27)(28). Stimulated degradation of long-lived proteins appears to depend on autophagy. This role of autophagy is particularly well documented in liver in the case of fasting (20, 29-33), glucagon, or amino acid deficiency (34)(35)(36)(37)(38)(39). There is also evidence of autophagy during stimulated protein degradation in cultured cells (40-42). Because of this background information, we expected that the meal would not affect the breakdown of short-lived hepatic proteins. In animals in which liver proteins were labeled by injection of carboxylabeled ornithine l h before the experiment, the meal had indeed little effect on the accumulation of bestatin-induced peptides (Table 11, Experiment 4). The experiment indicates that the livers are equally sensitive to bestatin in both conditions. Therefore, the effect of the meal on the accumulation of bestatin-induced peptides with long-lived proteins must reflect a real decrease in the rate of breakdown of these proteins.
Another source of concern is that in the intact mouse, part of the peptides could derive their radioactivity from long-lived proteins indirectly, through reincorporation of the label into short-lived proteins. In experiments with isolated cells, the distinction between long-and short-lived proteins is unambiguous, as long as it is properly measured in the presence of unlabeled amino acids in the medium, which prevents the TABLE I1 Effect of food in the hepatic accumulation of bestatin-induced peptides Long-lived liver proteins (Experiments 1, 2, and 3) were labeled in their leucine or arginine moieties by intraperitoneal injection of 10 pCi of either carboxy-labeled leucine or ornithine, 30 h beforehand. Short-lived proteins were labeled by injection or 2 pCi of ornithine 75 min before killing. Where indicated, bestatin (1 mg) was injected intravenously 15 min before killing. Peptides were purified from liver homogenates and their radioactivity determined after acid hydrolysis and ninhydrin treatment as described before and expressed as a ratio between peptide and protein radioactivity (5). Average total liver protein radioactivity in Experiments 1-3 was 1115 f 30, 887 f 32, and 909 f 32 X lo3 cpm; in Experiment 4, it was 219 f 9 X lo3 cpm before the meal, and 164 f 10 X lo3 cpm after the meal. The number of mice are shown in parentheses; values are average f range for duplicate mice or f S.E. for three or more animals. In Experiments 1, 2, and 4, the effect of food is studied in meal-fed animals at the indicated times after the start of the meal. In Experiment 3, food was withdrawn from untrained mice from 7 a.m. to 7 p.m., after which the animals were forced to ingest by intragastric tubing 25 mg of leucine dissolved in 1 ml of water at 40 "C, 4,3,2, and 1 h before the injection of bestatin. In a separate experiment it was determined that when ~-[l-'~C]leucine was added to a single forced fed solution, the excretion of 14C02 (an indication of the rate of absorption and distribution of the ingested leucine) peaked between 30 and 45 min afterwards. The total amount of leucine ingested in Experiment 3 in 4 h, 0.1 g, is equivalent to the amount of leucine contained in the 4-h meal in the other exneriments.  reincorporation into short-lived proteins of the radioactive amino acids produced by degradation of long-lived proteins. It could be argued that in the intact mouse significant reincorporation occurs, and that enough label accumulates in short-lived proteins to be a source of radioactive intermediates. If that were the case, the effect of a meal could be due, not to a decrease in protein breakdown, but to isotopic dilution by the large amounts of unlabeled leucine or arginine reaching the liver during absorption of digested food. If such unlabeled amino acids chase the labeled ones in the liver, and prevent their reincorporation, and if a large portion of the radioactive peptides derive directly from newly synthesized short-lived proteins, a decrease of peptide radioactivity could be conceivable even in the absence of slower protein breakdown. Although this possibility seemed remote, we judged it was important to eliminate it by the following experiment (Table 11, Experiment 3). Long-lived liver proteins were labeled with radioactive leucine and the animals were deprived of food from 7 a.m. to 7 p.m. Instead of a meal we gave these mice, by intragastric tubing, 100 mg of leucine over a period of 4 h, an amount equivalent to that contained in 5 g of food. We then injected bestatin to these and other control mice which did not receive leucine. The radioactivity of bestatin-induced peptides was the same in both groups. We therefore conclude that the effect of the meal in the accumulation of leucinelabeled peptides is due to a decrease in protein breakdown and not to isotopic dilution by the unlabeled leucine in the food.

Efficiency of Accumulation of Bestatin-induced Peptides: Comparison with the Disappearance of Protein-bound 14C-
Cuanidino-Arg-To determine the extent to which bestatininduced peptides accumulate in vivo, we compared our procedure with that of Swick and Ip (11). These authors proposed the selective measurement, in liver, of the guanidino-C of protein-bound arginine after labeling in vivo with 14C-bicarbonate. Because of the high activity of the urea cycle in this tissue: ( a ) 14C02 is preferentially incorporated in the guanidino-C of hepatic arginine (via carbamyl phosphate in the hepatic mitochondria), and little of it ends up in other tissues; ( 6 ) most of the guanidino-14C of the arginine returning to the amino acid pools turns over in the urea cycle before it can be reincorporated. This ingenious procedure remains the only reliable means of estimating liver protein degradation by the disappearance of protein radioactivity. As with other measurements of its kind, however, it requires an interval between samples long enough to detect the change in radioactivity, which for long-lived proteins is no less than 1 day. It is thus of no use in following acute changes in the breakdown of longlived proteins, such as those studied in this paper. Nevertheless, we can compare both procedures in fasted animals, where the diurnal variations caused by food intake are avoided. Since the results with short-lived proteins indicate that food intake makes no difference in the sensitivity of liver to bestatin (Table 11), a comparison of both procedures in the absence of food should suffice for our purpose.
Mice were fasted for 22 h before the intraperitoneal injection of 14C-bicarbonate, ~-[l-'~C]leucine, or ~-[l-'~C]ornithine (a precursor of hepatic protein-bound arginine (5)). The leucine and ornithine-labeled animals were used to determine the accumulation of bestatin-induced peptides, as in the previous experiments, except that 36 h were allowed to elapse between the injection of the label and that of bestatin, to coincide with the midpoint of the following determination. The radioactivity in the guanidino-C of protein-bound arginine was measured in two groups of mice, 24 and 48 h after the injection of 14C-bicarbonate. As with all measurements based on the degradation of labeled cellular proteins, the measured rate depends on the time elapsed between the injection of the precursor and the determination. In liver, we have previously shown that one-half of the newly synthesized protein is either exported (as plasma proteins) or degraded within 3 h (12). The labeled proteins remaining after 3 h are metabolically heterogeneous. The protein radioactivity disappears thereafter at decreasing rates; as the more rapidly turning over components are progressively depleted, the average rate at which the remaining proteins disappear becomes slower. In livers of mice labeled with 14C-bicarbonate we have shown that during the first day, the apparent Kd is 0.66 day", during the second day, it is 0.33 day-', and afterwards 0.23 day" (12).
In the Experiment A of Table 111, the apparent K d for guanidino-labeled proteins disappearing between 24 and 48 h after injection of 14C-bicarbonate was 0.327 day". If we take this to represent the true rate of degradation of the labeled proteins during that interval, we can now compare to this rate, under the same conditions, the accumulation of bestatininduced peptides 36 h after injection of the corresponding precursors. As shown in Experiment B in Table III, were used to determine the accumulation of bestatin-induced peptides, as in Table 11, except that 36 h were allowed to elapse between the injection of the label and that of bestatin, to coincide with the midpoint of the following determination. The radioactivity in the guanidino-C of protein bound arginine (part A) was measured in triplicate in a pool of livers from two groups of mice, killed 24 and 48 h after the injection of "C-bicarbonate (see "Experimental Procedures"). All values are averages f S.E. (number of animals in the group). The apparent K d for the degradation of guanidino-"C-protein, between 24 and 48 h after injection of the label, was calculated assuming an exponential disappearance during that interval (12). In parts B and C, peptides were purified from liver homogenates and their radioactivity determined after acid hydrolysis and ninhydrin treatment as described before and expressed as a ratio between peptide and protein radioactivity (5). Average total liver protein radioactivity in B was 489 f 19 X lo3 cpm; and in C, 267 t 6 X lo3 cpm. For comparison, the apparent Kd for bestatin-induced peptides the text. accumulating in 15 min was recalculated for 1 day as explained in A. Disappearance of protein-bound "C-euanidino areinine  (7) 1.86 0.179 (55%) as Measure of Liver Protein Degradation h, or 12.9%/day. This value represents 39% of the protein degradation calculated by the procedure of Swick and Ip (11). The experiment was also repeated under the same conditions with livers labeled in the carboxy-C of Arg moieties by the injection of ~-[l-'~C]ornithine 36 h beforehand (Table 111, Experiment C). The arginine-labeled peptides represented 0.186% of the hepatic protein radioactivity, or 55% of the true protein degradation. As with the experiments in Table 11 discussed before, the more efficient accumulation of argininelabeled intermediates is consistent with our previous observation in isolated hepatocytes, that these peptides are better retained within these cells (6).
With the efficiency of accumulation of bestatin-induced peptides thus known, it is possible to estimate absolute rates of breakdown. For instance, using the value of 55% for Argpeptides (Table 111), if we now return to Experiment 1 in Table 11, the rate of degradation of long-lived proteins (defined in this experiment as the radioactive proteins remaining in the liver 1 day after injection of the precursor) was, before the meal, 0.278 X (60 min/l5 min) X (100/55) = 2.02%/h. Because we know, from the study of short-lived proteins (Experiment 4, Table 11) that food intake has little or no effect on the sensitivity of the liver to bestatin, we can calculate in the same way that this rate dropped to 0.73, 0.69, and 1.03%/h, 4, 8, and 16 h after the beginning of the meal. With the value of 39% for Leu-labeled intermediates (Table   111), in Experiment 2 of Table 11, degradation dropped 4 h after the meal from 1.76 to 0.66%/h.
The comparison with the procedure of Swick and Ip (11) can be in principle applied to any other condition in which unchanging degradation of liver proteins can be studied for at least 1 day. To ascertain that bestatin efficiency is not affected by acute changes, such as food intake, comparison of bestatin-induced intermediates in the degradation of shortlived proteins can be used, as already discussed in the previous section. As with all experiments with pulse-labeled proteins, including the procedure of Swick and Ip (11) and measurements in isolated or cultured cells, the apparent Kd will necessarily be affected by the length of the chase. A complete description of the breakdown of radioactive cellular proteins can only be attained either by continuous administration of the precursor, until all classes of proteins are uniformly labeled; or by extensive analysis of the apparent rates obtained after progressively longer chase intervals (1). We should stress, however, that our major interest here is not to account for the increase in liver protein content after a meal by a detailed balance of protein synthesis and degradation. This can be done more accurately during periods of sustained liver growth, including the refeeding of protein-depleted animals (21). Where this novel procedure should be irreplaceable is in establishing, in intact animals, the relative importance of nutrients and hormones in the transient changes in protein metabolism after a meal. For this, the slower degradation of long-lived proteins is most important; these are the proteins that are most clearly affected in perfused livers (3) and isolated hepatocytes (4).
Present Limitations of the Procedure-As a way to estimate transient changes in protein breakdown, the accumulation of bestatin-induced peptides has at present some limitations. ( a ) Compared with cultured cells, hepatocytes are very sensitive to bestatin (see Fig. 3 in Ref 6). Other cell types may require larger concentrations, which may be difficult to achieve in intact animals. The maximum solubility of bestatin in aqueous media is 5 mg/ml. Because it is undesirable to inject excessive volumes of fluid to animals, solubility may become a limitation in studies with other tissues. ( b ) Bestatin is at present expensive and this limits its use to small animals. Its chemical structure is, however, relatively simple and it can be synthesized (43); we thus expect that cost represents only a transient problem. None of these limitations constitute an impediment for the application of the procedure to studies in the livers of small animals. We expect it will also be useful for studies in other tissues.
Conclusions-From the experiments presented above we conclude that 4 h after a meal, storage of excess liver protein results from a 45% increase in protein synthesis and a 65% decrease in the degradation of long-lived proteins. A decrease in liver protein breakdown after a meal was suggested before because the rate of hepatic protein synthesis did not increase enough (18,20), and based on electron microscopic evidence of decreased hepatic autophagy (20,31). Slower protein degradation was found in perfused livers derived from mice refed after a short period of starvation (44). A sustained decrease in protein degradation has been documented in livers after recovery from protein depletion, both by balance between protein synthesis and accumulation and by the disappearance of radioactive liver proteins labeled by injection of 14C-bicarbonate (21). This is, however, the first time that transient changes in hepatic protein breakdown can be demonstrated directly in intact animals. From studies with perfused livers (3) and isolated hepatocytes (4), we know some of the parameters that could be responsible for the effects. After a meal, higher levels of circulating amino acids and insulin, and lower concentrations of glucagon could all contribute to stimulation of net protein accumulation in body tissues (3, 45). Measurement of bestatin-induced peptides will now permit the establishment of the relative importance of these or other variables in liver of intact animals. The procedure may also prove useful for studies in other tissues.