Influence of Dietary Protein Intake on Whole-Body Protein Turnover in Humans
Methods for measuring rates of protein synthesis and degradation in the whole body of humans with isotopes of carbon and nitrogen are described and attention is drawn to their relative merits and drawbacks for studying the nutritional control of protein metabolism. A review of published work on dietary protein and protein metabolism leads to the conclusion that protein is the major dietary determinant of whole-body protein turnover rates, and that energy intake is comparatively unimportant. Dietary protein affects protein turnover at two levels: an immediate response to the intake of protein in meals and a longer-term adaptation after a change in protein intake. An increase in the level of dietary protein enhances the response to meals, which mainly consists of a decrease in the rate of protein degradation. The adaptation to higher protein intakes involves an increase in the basal (postabsorptive) rates of both synthesis and degradation. Suggestions for future investigation include more detailed studies of the acute and adaptive responses, to facilitate understanding of dietary protein requirements, and the effects of very-high-protein intakes with continued development of techniques for studying protein turnover in individual tissues in humans. Diabetes Care 14:1189-98,1991
Peter |. Gariick, PhD Margaret A. McNurlan, PhD Peter L. Ballmer, MD
others might actively encourage very high intakes (e.g., some athletes). Conventionally, requirements have been investigated with nitrogen-balance techniques but this gives no information about the optimum intake, which must be assessed by other techniques (1). Nitrogen balance is the net result of the numerous metabolic reactions involved in the synthesis and degradation of cellular proteins. This process, protein turnover, is finely controlled so that the gain or loss of body protein can be determined for any particular metabolic or nutritional state. How rates of body protein synthesis and degradation are regulated by dietary factors is therefore an important step in understanding the metabolic significance of differences in dietary intake. We have therefore reviewed information on the responses of protein turnover to differences in the dietary intake of protein. We have limited the discussion on measurements of rates of whole-body protein turnover and, to avoid complications of interpretation resulting from growth, we concentrated mainly on work in adults. Additional information can be obtained from previous articles (1-7).
METHODOLOGICAL CONSIDERATIONS
A
dults are able to maintain body protein balance over a wide range of dietary intakes of protein. This raises the question not only of what is the minimum requirement of dietary protein but also what is the optimum intake in any particular circumstance? For example, certain groups of patients (e.g., those with kidney disease) might limit their intake of protein on medical advice, whereas
DIABETES CARE, VOL. 14, NO. 12, DECEMBER 1991
Most studies of protein turnover in humans have been measurements of rates in the whole body. The isotopic techniques for assessing these rates can be performed From the Clinical Metabolism Group, Rowett Research Institute, Aberdeen, Scotland; and the University of Bern, Department of Internal Medicine, Inselspital, Bern, Switzerland. Address correspondence and reprint requests to Peter J. Garlick, PhD, Rowett Research Institute, Aberdeen AB2 9SB, Scotland, UK.
1189
PROTEIN INTAKE AND BODY PROTEIN TURNOVER
noninvasively, whereas equivalent measurements on individual tissues are much more invasive and until now appear not to have been used for studying the effects of dietary protein intake in humans. Although two different approaches to the measurement of whole-body turnover have been used, they rely on essentially the same assumptions, which are that the body amino acids can be regarded as consisting of just two homogeneous pools or compartments; one containing all of the amino acids in protein and the other the free amino acids in the body fluids (i.e., plasma and tissue water). Protein turnover rates (i.e., the rates of exchange of amino acids between these two compartments) are then assessed by introducing a labeled amino acid into the free amino acid pool orally or by intravenous injection and assessing its dilution by unlabeled amino acids from the breakdown of dietary and body proteins or its disappearance into protein synthesis and catabolic pathways. It is assumed that during measurement, any label incorporated into protein is not released by protein degradation and thus recycled into the free amino acid pool. The method of calculation (stochastic analysis) is illustrated in Fig. 1 and described in more detail in previous articles (4,7). The choice of isotope determines the actual approach used. The stable isotopic form of nitrogen (15N), generally in the form of [15N]glycine, can be given orally or intravenously and by continuous infusion or single dose. Assessment of rates of protein turnover then depends on measurements of the excretion of the isotope in urinary end products of N metabolism (urea and ammonia). The advantage of this approach is that measurements are convenient to make, without stress or restriction of mobility to the subjects. However, measurement periods have frequently been long (e.g., 2-3 days), because in the original form of the method, described by Picou and Taylor-Roberts (8), the large size of the body urea pool caused a substantial delay in the excretion of label in urea. Therefore, the technique is unsuitable for observing rapid changes, such as those occurring after meals. Shorter measurement periods (e.g., 9 h) can be achieved by measurements on urinary ammonia (9), by the use of priming doses (10), or by estimating the amount of label retained in the body urea pool (11). The disadvantage is that the two end products, urea and ammonia, frequently give rise to different values for rates of turnover. Although a rationale for combining the rates obtained from the two end products has been provided (12), there remains a lack of detailed knowledge about the precursors of labeled urea and ammonia. The alternative approach is to use carbon labeling to trace the metabolism of a single amino acid (13,14). Carbon-14 (radioactive) or carbon-13 (stable) have been used extensively. Leucine labeled in the carboxyl group is particularly suitable (14), being given by continuous intravenous infusion for periods of >10 h, or for only 4 h if a priming dose is given (15,16). Mea-
1190
Dietary amino acids Labelled amino acid
Precursor methods
BODY PROTEIN
m
End product I N Excretion methods [ (NH3 and urea) FIG. 1. Two-pool model for protein metabolism used for calculating rates of whole-body protein turnover from isotopic labeling experiments. It is assumed that free amino acid pool size remains constant during measurement so that total entry to pool (from dietary intake [I] and body protein breakdown [B]) equals total exit (to oxidative processes [E] and protein synthesis [S]). Total turnover of free pool, or flux rate (Q) is given by Q = I + B = E + S (eq. 1). All other pathways of metabolism are ignored. Flux rate can be determined by infusing (at rate i, an isotopically labeled amino acid (either intravenously or orally) and determining isotopic enrichment of free amino acid precursor of protein synthesis. Carbon labeling traces metabolism of 1 amino acid (e.g., [1-13C]leucine). The isotopic enrichment of free leucine reaches constant plateau value (A max ), which is evaluated by measurements on plasma leucine, giving rise to plasma methods. Flux rate is then given by Q = i/Amax (eq. 2). Breakdown (B) and synthesis (S) of protein are calculated from Q with equation 1, where the dietary intake of amino acid (I) is known and rate of oxidation (E) can be determined from appearance of isotope in respiratory CO 2 . With 15N (e.g., [15N]glycine) label mixes in total amino-nitrogen pool and measurements are generally made on an oxidation product of free amino acid pool in urine (e.g., urea), giving rise to the term end product methods. Rate of flux is calculated from eq. 2, as for [1-13C]leucine, except that rates are given in terms of amounts of nitrogen rather than leucine. Values of (B) and (S) are calculated from eq. 1, where (I) is intake of nitrogen and (E) is urinary excretion of total nitrogen. Additional information can be obtained from previous articles (2,4,6,7).
surements are made of the isotopic enrichment (or specific radioactivity) of leucine in the plasma and of the production of labeled CO2 in the breath. The advantage of this method is that measurements can be made rapidly, and hence acute changes in metabolism detected. Moreover, an amino acid like leucine has very limited metabolism other than by protein turnover, unlike the complex transfers of 15 N from glycine within the total free amino-nitrogen pool. However, there remains uncertainty about the rates of protein
DIABETES CARE, VOL. 14, NO. 12, DECEMBER 1991
P.). GARLICK, M A McNURlAN, AND P.E. BALLMER
turnover obtained, because of differences in labeling of the plasma and tissue-free leucine pools. It has been suggested that the enrichment of the transamination product of leucine, a-ketoisocaproate, might provide a more realistic estimate of protein turnover (17,18), but transamination takes place mainly in the muscle (18), and it remains to be seen whether it will prove representative of all the other tissues in all metabolic conditions. Compared with methods with 1 5 N, this technique is more restrictive for the subject and requires more specialized equipment (i.e., 2 different mass spectrometers for measuring plasma leucine and breath CO2 enrichments, and a respiration chamber for measuring CO2 production). Studies of the effects of dietary protein have been made with both 15N and 13/14 C. From the above discussion, it will be apparent that neither method of measuring whole-body protein turnover rates can be regarded as totally definitive. Therefore, confirmation of observed changes by both methods can be taken as evidence that the changes occur. Note that differences in the time span of the measurements by the two approaches can mean that the results are not strictly comparable; [13C]leucine responding to acute changes and [15N]glycine giving an average sometimes over much longer periods.
RESPONSES TO DIETARY PROTEIN IN ADULT HUMANS The first attempt to assess the effect of varying dietary protein on human protein turnover was by Tschudy et al. (19) who used [15N]aspartate in a subject given four different levels of protein and energy. However, the first systematic study appears to be that of Steffee et al. (20) who measured protein turnover by oral administration of [15N]glycine for 60 h in volunteers given a diet containing a relatively normal amount of protein (1.5 g • kg" 1 • day"1) for 5 days followed by one containing a marginally adequate amount (0.38 g • kg" 1 • day" 1 ) for 12 days. Both synthesis and degradation of protein appeared to rise when the lower protein diet was given. It is not clear to what extent this result, showing a negative relationship between protein synthesis and dietary protein level, might have been influenced by the long periods of isotope infusion, because subsequent investigations, mostly with shorter protocols, have not provided confirmation of this result. Evidence that dietary protein positively affects whole-body protein turnover rates was first obtained by constant infusion of [1-14C]leucine into obese subjects on weight-reducing diets (Table 1). Measurements were made initially in subjects on a normal diet containing 8 MJ and 70 g protein/day, after which weightreducing diets of 2.1 MJ/day were given for 3 wk, followed by a second measurement of protein turnover. During each infusion, the subjects were given
DIABETES CARE, VOL. 14, NO. 12, DECEMBER 1991
TABLE 1 Effects of changes in dietary protein and energy intake on rates of whole-body protein turnover in obese subjects Rates of protein turnover (mmol leucine/h]1
Intake Protein (g/day)
Energy (kcal/day)
Oxidation
Synthesis
Degradation
70 50 70 0
2000 500 2000 500
2.3 1.9 2.4 0.4
7.9 6.9 7.2 4.4
6.6 6.0 6.3 4.8
Based on ref. 21. Measurements were made by constant intravenous infusion of |1-14C]leucine for 12 h in the fed state when subjects received the appropriate diet at rate of one-twelfth their daily ration each hour. Each subject received normal diet 3 days before measurement followed by 3-wk period on one of the low-energy diets before measurement was repeated.
the corresponding diet as small hourly meals. When the low-energy diet contained 50 g protein/day, the rates of synthesis and degradation of body protein were only slightly reduced compared with rates on the normal diet. However, when the low-energy diet contained no protein, both synthesis and breakdown of protein were substantially reduced (Table 1), suggesting that the energy deficit had little effect on turnover rates, whereas the absence of protein was critical. Similar data in obese subjects given low-energy diets containing different levels of protein have been reported from studies with [15N]glycine with measurements of the excretion of label in urea over 48 (22) or 120(23) h. In the experiment shown in Table 1, measurements were also made with [15N]glycine, given either as a single oral dose or as repeated oral dosages (equivalent to infusion). Measurement on urinary ammonia rather than urea permitted the measurement to be run in parallel with the [13C]leucine infusion over 12 h. The results were the same by the different techniques (21). The single-dose [15N]glycine technique, which is particularly suitable for frequent follow-up studies because measurements are only made on urine, was then used to follow the time course of protein turnover after the change in diet (Fig. 2). The fall in protein synthesis with the protein-free diet occurred mostly on the 1st day of the diet (Fig. 2). Similarly, after reintroduction of protein, there was a rapid increase in synthesis, suggesting that the higher rate of synthesis on proteincontaining diets is a direct response to the absorption of protein. The rapidity of changes in protein turnover with dietary changes was further illustrated by follow-up studies on obese patients on normal and low-energy diets. By giving 24-h infusions of [14C]leucine spanning a complete day, including 12 h of feeding and 12 h of fasting, diurnal variations were detected when the
1191
PROTEIN INTAKE AND BODY PROTEIN TURNOVER
8.0 MJ 70gProt.
2.1 MJ Zero Protein
2.1 MJ 50g Protein
FIG. 2. Rates of whole-body protein synthesis measured sequentially by a single dose of [15N]glycine in 1 obese subject given 3 different diets. From Garlick et al. (21).
normal diet was given. These variations were attributed to the varying rate of food absorption during the day (24). The original data suggested that feeding was associated with a rise in protein synthesis and amino acid oxidation and a fall in protein degradation (24,25), but more recent work has shown this to be partially in error (26). Because of the recycling of isotope from labeled leucine from the breakdown of labeled protein back into the free leucine pool during the latter part of the 24-h infusions, the synthesis rate in the fasting state was underestimated. Experiments comparing long and short infusion periods have shown that the response of synthesis to feeding had been exaggerated, and that breakdown is the major process regulating body protein mass with acute changes in dietary intake (26; Table 2). The responses of protein metabolism to feeding have been reviewed in more detail elsewhere (27). When the effects of the protein-free low-energy diet were examined separately in both the fed and fasted
states, no differences between the fed and fasted states were detected for protein synthesis, breakdown, or oxidation (28). Moreover, even in the fasted state, synthesis and degradation were depressed compared with the values observed during fasting after the normal diet. In particular, leucine oxidation was reduced to only 30% of its original value. These subjects were then measured again while consuming a low-energy protein-containing diet. The response to this was different from that observed in subjects who had been given this diet for the complete 3-wk period. The interpretation of these data is complicated by the problem with isotope recycling with the long infusions described above. However, recycling does not affect the measurement of leucine oxidation, and this showed that the response to meals containing protein was reduced by 3 wk of protein deprivation (28). These data show that although there is an immediate effect of dietary intake on protein turnover rates, there is also a metabolic adaptation, which took place during the 3 wk on the low-energy protein-free diet. Systematic studies of the effects of different levels of dietary protein, ranging from deficient to surfeit, have been conducted in healthy volunteers in the laboratory of Young. Separate measurements were made in the fed and fasted (postabsorptive) states with primed continuous infusion of [1-13C] leucine for 4-h periods so that isotope recycling was not a problem. Motil et al. (29) studied volunteers given diets containing normal energy, including 0.1, 0.6, and 1.5 g protein • kg" 1 • day" 1 for 7 days, whereas Yang et al. (3) gave levels of 0, 0.3, 0.6, and 1.5 g protein • kg" 1 • day" 1 . Their results are largely in accord with each other and are summarized in Table 3. Motil et al. also gave infusions of [15N]lysine in parallel with the [13C]leucine measurements and the data were consistent. In the fed state, leucine oxidation rose substantially with dietary protein intake, as might be expected from the higher rates of nitrogen excretion when protein in excess of the requirement for balance is given. In Yang et al., the increase was much more pronounced above a
TABLE 2 Apparent response of whole-body protein turnover to feeding measured with continuous infusion of [1-13C]leucine by 3 different protocols Rate of protein turnover (|xmol leucine kg ' • h' Fed state
Postabsorptive state Infusion period (h)
Oxidation
Synthesis
Degradation
Oxidation
Synthesis
Degradation
24 8 2X4
19 20 20
59 72 72
76 90 90
36(189) 31 (155) 46 (230)
79(134) 66 (92) 71 (99)
48 (63) 31 (34) 51 (57)
Based on ref. 26. Values in parentheses are percentages of postabsorptive values. The 24-h protocol was unprimed infusion, with subjects given hourly meals for the first 12 h and fasted for the second 12 h. The 8-h protocol was a primed infusion with fasting for first 4 h and feeding for second 4 h. In 2 x 4-h protocol, subjects were measured by primed infusion for last 4 h of 12-h feeding period. Fasting values are taken from line above (i.e., first 4 h of primed infusion in postabsorptive state). With 24-h protocol, there is an increase in protein synthesis with feeding, which is not apparent with either of the 2 shorter periods of infusion.
1192
DIABETES CARE, VOL. 14, N O . 12, DECEMBER 1991
P.). GARLICK, M A McNURlAN, AND P.E. BALLMER
TABLE 3 Effects on varying dietary protein and energy intake on whole-body protein turnover measured by primed intravenous infusion of [1-13C]leucine Intake (kg
Rates of protein metabolism (\imo\ leucine • kg" 1 • h-')
Vh)
Postabsorptive state Study 1
2
3 4
Fed state
Protein (g)
Energy (kcal)
Oxidation
Synthesis
Degradation
Oxidation
Synthesis
Degradation
0.1 0.6 1.5 0 0.3 0.6 1.5 0.6 0.6 1 1.3
44 44 44 48 48 48 48 43 54 27 35
13 22 18 8 10 9 10 16 16 17 13
76 89 113 66 62 71 88 81
85 107 125 75 71 80 98 92 92 86 79
12 22 46 9 7 8 22 18
64 102 113 57 58 85 107 88 92 53 55
66 82 67 66 49 60 46 66 65 43 30
81
69 64
12
29 28
Data are from: study 1, Motil et al. (29); study 2, Yang et al. (3,30); study 3, Motil et al. (31); and study 4, Bruce et al. (32). Studies 1, 2, and 3 involved healthy young men adapted to diets for 7 days before measurement. Study 4 involved healthy women who underwent no prior adaptation. During the fed phase of the measurement, the diets were given at a rate of one-twelfth the intake values given in the table each hour.
dietary intake of 0.6 g, consistant with the view that this intake is close to the requirement value. In the fasted state, the variations in oxidation with intake were much less pronounced, and in the study of Yang et al., they were absent. Rates of whole-body protein synthesis also changed in response to dietary protein, but in the fed state, the changes were more pronounced. During both feeding and fasting, there were low rates of synthesis at low intakes of protein, then rising at intakes of 0.6 and 1.5 g protein • kg" 1 • day" 1 . Protein breakdown did not respond to intake in the fed state, appearing to be depressed to the same low level by feeding at any level of protein. However, while fasting, breakdown rates were also lower with the lower intakes of protein. These changes in rates during the fasting state are additional evidence that protein metabolism not only responds to the immediate intake of food but also undergoes adaptation to varying intakes of protein over a longer period. These data can also be viewed from a different perspective. The differences in protein synthesis and degradation between the fasted and fed states represent the responses to feeding, and these vary with the dietary protein content. At very low intakes of protein there was very little effect of feeding on any of the parameters measured, consistent with the data described above with low-energy diets (28). At the higher levels of intake, the effect of feeding was mainly to inhibit protein degradation and enhance leucine oxidation. The effect of feeding on protein synthesis, however, even at the high protein intakes, is not definitive. The data of Yang et al. (3) suggest that synthesis rises on feeding at high intakes, whereas the results of Motil
DIABETES CARE, VOL. 14, N O . 12, DECEMBER 1991
et al. (31) show no changes at any protein level. Unfortunately, studies of the effects of feeding on whole-body rates of protein turnover reveal one of the difficulties of interpreting the data with a simple twopool model of protein metabolism (Fig. 1). Although it might be reasonable to regard the splanchnic tissues as part of a single homogeneous body pool in the fasted state, during feeding, the unlabeled amino acid from the diet first enters the gut and liver. These tissues extract a part and modify its isotopic enrichment before it can enter the peripheral circulation, which is the site of blood sampling. Methods that use multiple labels are being developed in an attempt to analyze such data more rigorously (33). Supporting evidence for the pattern of changes of protein turnover with dietary protein intake described above comes from a study by Meredith et al. (34), who gave diets containing 0.6, 0.9, and 1.2 g proteinkg" 1 • day" 1 to endurance-trained men for 10-day periods. Protein turnover was determined by repeated oral dosages of [15N]glycine over 60-h periods with measurements on urinary urea. Nitrogen excretion (protein oxidation) increased with increasing dietary protein, as did the rate of whole-body protein synthesis, but protein breakdown remained unaltered. The estimated protein requirement to maintain nitrogen balance was —0.9 g protein • kg" 1 • day" 1 , but no differences in the responses to changes in dietary protein were apparent when diets above and below the maintenance level were given. This study showed no differences in protein kinetics or protein requirement between young and middle-aged men. A somewhat different picture is suggested by Stuart et al. (35), who measured protein turnover by primed
1193
PROTEIN INTAKE AND BODY PROTEIN TURNOVER constant infusion of [1-13C]leucine in subjects given 0.6 or 1.0 g protein • kg" 1 • day" 1 for 5 days. Measurements were made only in the fasting state and showed that leucine oxidation was higher and protein degradation lower on the higher-protein diet. However, protein synthesis was not significantly altered. Although the authors discuss several differences in the experimental procedures that might have caused the responses to be different from those reported previously (e.g., dietary energy level and their use of ketoisocaproate to calculate rates of turnover compared with plasma leucine), the explanation for the discrepancies is not clear. Stuart et al. (35) also examined the response of protein kinetics to insulin by the insulin clamp procedure. However, despite the lower protein degradation on the higher-protein diet, the maximal and half-maximal suppressions of degradation by insulin were not different. In the studies discussed so far, none have investigated the metabolic effects of very-high-dietary protein levels. There appears to be only one report of protein turnover rates in adults given excess protein (30). Primed infusions of [1-13C]leucine were given to volunteers in the postabsorptive state after receiving 1.5 or 3.9 g protein • kg" 1 • day" 1 for 8 days. Only rates of protein degradation were reported, because these were part of a more extensive study of alanine kinetics and were unaltered by the change in dietary protein. Although increases in dietary protein
Methods for measuring rates of protein synthesis and degradation in the whole body of humans with isotopes of carbon and nitrogen are described and attention is drawn to their relative merits and drawbacks for studying the nutritional control of protein metabolism. A review of published work on dietary protein and protein metabolism leads to the conclusion that protein is the major dietary determinant of whole-body protein turnover rates, and that energy intake is comparatively unimportant. Dietary protein affects protein turnover at two levels: an immediate response to the intake of protein in meals and a longer-term adaptation after a change in protein intake. An increase in the level of dietary protein enhances the response to meals, which mainly consists of a decrease in the rate of protein degradation. The adaptation to higher protein intakes involves an increase in the basal (postabsorptive) rates of both synthesis and degradation. Suggestions for future investigation include more detailed studies of the acute and adaptive responses, to facilitate understanding of dietary protein requirements, and the effects of very-high-protein intakes with continued development of techniques for studying protein turnover in individual tissues in humans. Diabetes Care 14:1189-98,1991
Peter |. Gariick, PhD Margaret A. McNurlan, PhD Peter L. Ballmer, MD
others might actively encourage very high intakes (e.g., some athletes). Conventionally, requirements have been investigated with nitrogen-balance techniques but this gives no information about the optimum intake, which must be assessed by other techniques (1). Nitrogen balance is the net result of the numerous metabolic reactions involved in the synthesis and degradation of cellular proteins. This process, protein turnover, is finely controlled so that the gain or loss of body protein can be determined for any particular metabolic or nutritional state. How rates of body protein synthesis and degradation are regulated by dietary factors is therefore an important step in understanding the metabolic significance of differences in dietary intake. We have therefore reviewed information on the responses of protein turnover to differences in the dietary intake of protein. We have limited the discussion on measurements of rates of whole-body protein turnover and, to avoid complications of interpretation resulting from growth, we concentrated mainly on work in adults. Additional information can be obtained from previous articles (1-7).
METHODOLOGICAL CONSIDERATIONS
A
dults are able to maintain body protein balance over a wide range of dietary intakes of protein. This raises the question not only of what is the minimum requirement of dietary protein but also what is the optimum intake in any particular circumstance? For example, certain groups of patients (e.g., those with kidney disease) might limit their intake of protein on medical advice, whereas
DIABETES CARE, VOL. 14, NO. 12, DECEMBER 1991
Most studies of protein turnover in humans have been measurements of rates in the whole body. The isotopic techniques for assessing these rates can be performed From the Clinical Metabolism Group, Rowett Research Institute, Aberdeen, Scotland; and the University of Bern, Department of Internal Medicine, Inselspital, Bern, Switzerland. Address correspondence and reprint requests to Peter J. Garlick, PhD, Rowett Research Institute, Aberdeen AB2 9SB, Scotland, UK.
1189
PROTEIN INTAKE AND BODY PROTEIN TURNOVER
noninvasively, whereas equivalent measurements on individual tissues are much more invasive and until now appear not to have been used for studying the effects of dietary protein intake in humans. Although two different approaches to the measurement of whole-body turnover have been used, they rely on essentially the same assumptions, which are that the body amino acids can be regarded as consisting of just two homogeneous pools or compartments; one containing all of the amino acids in protein and the other the free amino acids in the body fluids (i.e., plasma and tissue water). Protein turnover rates (i.e., the rates of exchange of amino acids between these two compartments) are then assessed by introducing a labeled amino acid into the free amino acid pool orally or by intravenous injection and assessing its dilution by unlabeled amino acids from the breakdown of dietary and body proteins or its disappearance into protein synthesis and catabolic pathways. It is assumed that during measurement, any label incorporated into protein is not released by protein degradation and thus recycled into the free amino acid pool. The method of calculation (stochastic analysis) is illustrated in Fig. 1 and described in more detail in previous articles (4,7). The choice of isotope determines the actual approach used. The stable isotopic form of nitrogen (15N), generally in the form of [15N]glycine, can be given orally or intravenously and by continuous infusion or single dose. Assessment of rates of protein turnover then depends on measurements of the excretion of the isotope in urinary end products of N metabolism (urea and ammonia). The advantage of this approach is that measurements are convenient to make, without stress or restriction of mobility to the subjects. However, measurement periods have frequently been long (e.g., 2-3 days), because in the original form of the method, described by Picou and Taylor-Roberts (8), the large size of the body urea pool caused a substantial delay in the excretion of label in urea. Therefore, the technique is unsuitable for observing rapid changes, such as those occurring after meals. Shorter measurement periods (e.g., 9 h) can be achieved by measurements on urinary ammonia (9), by the use of priming doses (10), or by estimating the amount of label retained in the body urea pool (11). The disadvantage is that the two end products, urea and ammonia, frequently give rise to different values for rates of turnover. Although a rationale for combining the rates obtained from the two end products has been provided (12), there remains a lack of detailed knowledge about the precursors of labeled urea and ammonia. The alternative approach is to use carbon labeling to trace the metabolism of a single amino acid (13,14). Carbon-14 (radioactive) or carbon-13 (stable) have been used extensively. Leucine labeled in the carboxyl group is particularly suitable (14), being given by continuous intravenous infusion for periods of >10 h, or for only 4 h if a priming dose is given (15,16). Mea-
1190
Dietary amino acids Labelled amino acid
Precursor methods
BODY PROTEIN
m
End product I N Excretion methods [ (NH3 and urea) FIG. 1. Two-pool model for protein metabolism used for calculating rates of whole-body protein turnover from isotopic labeling experiments. It is assumed that free amino acid pool size remains constant during measurement so that total entry to pool (from dietary intake [I] and body protein breakdown [B]) equals total exit (to oxidative processes [E] and protein synthesis [S]). Total turnover of free pool, or flux rate (Q) is given by Q = I + B = E + S (eq. 1). All other pathways of metabolism are ignored. Flux rate can be determined by infusing (at rate i, an isotopically labeled amino acid (either intravenously or orally) and determining isotopic enrichment of free amino acid precursor of protein synthesis. Carbon labeling traces metabolism of 1 amino acid (e.g., [1-13C]leucine). The isotopic enrichment of free leucine reaches constant plateau value (A max ), which is evaluated by measurements on plasma leucine, giving rise to plasma methods. Flux rate is then given by Q = i/Amax (eq. 2). Breakdown (B) and synthesis (S) of protein are calculated from Q with equation 1, where the dietary intake of amino acid (I) is known and rate of oxidation (E) can be determined from appearance of isotope in respiratory CO 2 . With 15N (e.g., [15N]glycine) label mixes in total amino-nitrogen pool and measurements are generally made on an oxidation product of free amino acid pool in urine (e.g., urea), giving rise to the term end product methods. Rate of flux is calculated from eq. 2, as for [1-13C]leucine, except that rates are given in terms of amounts of nitrogen rather than leucine. Values of (B) and (S) are calculated from eq. 1, where (I) is intake of nitrogen and (E) is urinary excretion of total nitrogen. Additional information can be obtained from previous articles (2,4,6,7).
surements are made of the isotopic enrichment (or specific radioactivity) of leucine in the plasma and of the production of labeled CO2 in the breath. The advantage of this method is that measurements can be made rapidly, and hence acute changes in metabolism detected. Moreover, an amino acid like leucine has very limited metabolism other than by protein turnover, unlike the complex transfers of 15 N from glycine within the total free amino-nitrogen pool. However, there remains uncertainty about the rates of protein
DIABETES CARE, VOL. 14, NO. 12, DECEMBER 1991
P.). GARLICK, M A McNURlAN, AND P.E. BALLMER
turnover obtained, because of differences in labeling of the plasma and tissue-free leucine pools. It has been suggested that the enrichment of the transamination product of leucine, a-ketoisocaproate, might provide a more realistic estimate of protein turnover (17,18), but transamination takes place mainly in the muscle (18), and it remains to be seen whether it will prove representative of all the other tissues in all metabolic conditions. Compared with methods with 1 5 N, this technique is more restrictive for the subject and requires more specialized equipment (i.e., 2 different mass spectrometers for measuring plasma leucine and breath CO2 enrichments, and a respiration chamber for measuring CO2 production). Studies of the effects of dietary protein have been made with both 15N and 13/14 C. From the above discussion, it will be apparent that neither method of measuring whole-body protein turnover rates can be regarded as totally definitive. Therefore, confirmation of observed changes by both methods can be taken as evidence that the changes occur. Note that differences in the time span of the measurements by the two approaches can mean that the results are not strictly comparable; [13C]leucine responding to acute changes and [15N]glycine giving an average sometimes over much longer periods.
RESPONSES TO DIETARY PROTEIN IN ADULT HUMANS The first attempt to assess the effect of varying dietary protein on human protein turnover was by Tschudy et al. (19) who used [15N]aspartate in a subject given four different levels of protein and energy. However, the first systematic study appears to be that of Steffee et al. (20) who measured protein turnover by oral administration of [15N]glycine for 60 h in volunteers given a diet containing a relatively normal amount of protein (1.5 g • kg" 1 • day"1) for 5 days followed by one containing a marginally adequate amount (0.38 g • kg" 1 • day" 1 ) for 12 days. Both synthesis and degradation of protein appeared to rise when the lower protein diet was given. It is not clear to what extent this result, showing a negative relationship between protein synthesis and dietary protein level, might have been influenced by the long periods of isotope infusion, because subsequent investigations, mostly with shorter protocols, have not provided confirmation of this result. Evidence that dietary protein positively affects whole-body protein turnover rates was first obtained by constant infusion of [1-14C]leucine into obese subjects on weight-reducing diets (Table 1). Measurements were made initially in subjects on a normal diet containing 8 MJ and 70 g protein/day, after which weightreducing diets of 2.1 MJ/day were given for 3 wk, followed by a second measurement of protein turnover. During each infusion, the subjects were given
DIABETES CARE, VOL. 14, NO. 12, DECEMBER 1991
TABLE 1 Effects of changes in dietary protein and energy intake on rates of whole-body protein turnover in obese subjects Rates of protein turnover (mmol leucine/h]1
Intake Protein (g/day)
Energy (kcal/day)
Oxidation
Synthesis
Degradation
70 50 70 0
2000 500 2000 500
2.3 1.9 2.4 0.4
7.9 6.9 7.2 4.4
6.6 6.0 6.3 4.8
Based on ref. 21. Measurements were made by constant intravenous infusion of |1-14C]leucine for 12 h in the fed state when subjects received the appropriate diet at rate of one-twelfth their daily ration each hour. Each subject received normal diet 3 days before measurement followed by 3-wk period on one of the low-energy diets before measurement was repeated.
the corresponding diet as small hourly meals. When the low-energy diet contained 50 g protein/day, the rates of synthesis and degradation of body protein were only slightly reduced compared with rates on the normal diet. However, when the low-energy diet contained no protein, both synthesis and breakdown of protein were substantially reduced (Table 1), suggesting that the energy deficit had little effect on turnover rates, whereas the absence of protein was critical. Similar data in obese subjects given low-energy diets containing different levels of protein have been reported from studies with [15N]glycine with measurements of the excretion of label in urea over 48 (22) or 120(23) h. In the experiment shown in Table 1, measurements were also made with [15N]glycine, given either as a single oral dose or as repeated oral dosages (equivalent to infusion). Measurement on urinary ammonia rather than urea permitted the measurement to be run in parallel with the [13C]leucine infusion over 12 h. The results were the same by the different techniques (21). The single-dose [15N]glycine technique, which is particularly suitable for frequent follow-up studies because measurements are only made on urine, was then used to follow the time course of protein turnover after the change in diet (Fig. 2). The fall in protein synthesis with the protein-free diet occurred mostly on the 1st day of the diet (Fig. 2). Similarly, after reintroduction of protein, there was a rapid increase in synthesis, suggesting that the higher rate of synthesis on proteincontaining diets is a direct response to the absorption of protein. The rapidity of changes in protein turnover with dietary changes was further illustrated by follow-up studies on obese patients on normal and low-energy diets. By giving 24-h infusions of [14C]leucine spanning a complete day, including 12 h of feeding and 12 h of fasting, diurnal variations were detected when the
1191
PROTEIN INTAKE AND BODY PROTEIN TURNOVER
8.0 MJ 70gProt.
2.1 MJ Zero Protein
2.1 MJ 50g Protein
FIG. 2. Rates of whole-body protein synthesis measured sequentially by a single dose of [15N]glycine in 1 obese subject given 3 different diets. From Garlick et al. (21).
normal diet was given. These variations were attributed to the varying rate of food absorption during the day (24). The original data suggested that feeding was associated with a rise in protein synthesis and amino acid oxidation and a fall in protein degradation (24,25), but more recent work has shown this to be partially in error (26). Because of the recycling of isotope from labeled leucine from the breakdown of labeled protein back into the free leucine pool during the latter part of the 24-h infusions, the synthesis rate in the fasting state was underestimated. Experiments comparing long and short infusion periods have shown that the response of synthesis to feeding had been exaggerated, and that breakdown is the major process regulating body protein mass with acute changes in dietary intake (26; Table 2). The responses of protein metabolism to feeding have been reviewed in more detail elsewhere (27). When the effects of the protein-free low-energy diet were examined separately in both the fed and fasted
states, no differences between the fed and fasted states were detected for protein synthesis, breakdown, or oxidation (28). Moreover, even in the fasted state, synthesis and degradation were depressed compared with the values observed during fasting after the normal diet. In particular, leucine oxidation was reduced to only 30% of its original value. These subjects were then measured again while consuming a low-energy protein-containing diet. The response to this was different from that observed in subjects who had been given this diet for the complete 3-wk period. The interpretation of these data is complicated by the problem with isotope recycling with the long infusions described above. However, recycling does not affect the measurement of leucine oxidation, and this showed that the response to meals containing protein was reduced by 3 wk of protein deprivation (28). These data show that although there is an immediate effect of dietary intake on protein turnover rates, there is also a metabolic adaptation, which took place during the 3 wk on the low-energy protein-free diet. Systematic studies of the effects of different levels of dietary protein, ranging from deficient to surfeit, have been conducted in healthy volunteers in the laboratory of Young. Separate measurements were made in the fed and fasted (postabsorptive) states with primed continuous infusion of [1-13C] leucine for 4-h periods so that isotope recycling was not a problem. Motil et al. (29) studied volunteers given diets containing normal energy, including 0.1, 0.6, and 1.5 g protein • kg" 1 • day" 1 for 7 days, whereas Yang et al. (3) gave levels of 0, 0.3, 0.6, and 1.5 g protein • kg" 1 • day" 1 . Their results are largely in accord with each other and are summarized in Table 3. Motil et al. also gave infusions of [15N]lysine in parallel with the [13C]leucine measurements and the data were consistent. In the fed state, leucine oxidation rose substantially with dietary protein intake, as might be expected from the higher rates of nitrogen excretion when protein in excess of the requirement for balance is given. In Yang et al., the increase was much more pronounced above a
TABLE 2 Apparent response of whole-body protein turnover to feeding measured with continuous infusion of [1-13C]leucine by 3 different protocols Rate of protein turnover (|xmol leucine kg ' • h' Fed state
Postabsorptive state Infusion period (h)
Oxidation
Synthesis
Degradation
Oxidation
Synthesis
Degradation
24 8 2X4
19 20 20
59 72 72
76 90 90
36(189) 31 (155) 46 (230)
79(134) 66 (92) 71 (99)
48 (63) 31 (34) 51 (57)
Based on ref. 26. Values in parentheses are percentages of postabsorptive values. The 24-h protocol was unprimed infusion, with subjects given hourly meals for the first 12 h and fasted for the second 12 h. The 8-h protocol was a primed infusion with fasting for first 4 h and feeding for second 4 h. In 2 x 4-h protocol, subjects were measured by primed infusion for last 4 h of 12-h feeding period. Fasting values are taken from line above (i.e., first 4 h of primed infusion in postabsorptive state). With 24-h protocol, there is an increase in protein synthesis with feeding, which is not apparent with either of the 2 shorter periods of infusion.
1192
DIABETES CARE, VOL. 14, N O . 12, DECEMBER 1991
P.). GARLICK, M A McNURlAN, AND P.E. BALLMER
TABLE 3 Effects on varying dietary protein and energy intake on whole-body protein turnover measured by primed intravenous infusion of [1-13C]leucine Intake (kg
Rates of protein metabolism (\imo\ leucine • kg" 1 • h-')
Vh)
Postabsorptive state Study 1
2
3 4
Fed state
Protein (g)
Energy (kcal)
Oxidation
Synthesis
Degradation
Oxidation
Synthesis
Degradation
0.1 0.6 1.5 0 0.3 0.6 1.5 0.6 0.6 1 1.3
44 44 44 48 48 48 48 43 54 27 35
13 22 18 8 10 9 10 16 16 17 13
76 89 113 66 62 71 88 81
85 107 125 75 71 80 98 92 92 86 79
12 22 46 9 7 8 22 18
64 102 113 57 58 85 107 88 92 53 55
66 82 67 66 49 60 46 66 65 43 30
81
69 64
12
29 28
Data are from: study 1, Motil et al. (29); study 2, Yang et al. (3,30); study 3, Motil et al. (31); and study 4, Bruce et al. (32). Studies 1, 2, and 3 involved healthy young men adapted to diets for 7 days before measurement. Study 4 involved healthy women who underwent no prior adaptation. During the fed phase of the measurement, the diets were given at a rate of one-twelfth the intake values given in the table each hour.
dietary intake of 0.6 g, consistant with the view that this intake is close to the requirement value. In the fasted state, the variations in oxidation with intake were much less pronounced, and in the study of Yang et al., they were absent. Rates of whole-body protein synthesis also changed in response to dietary protein, but in the fed state, the changes were more pronounced. During both feeding and fasting, there were low rates of synthesis at low intakes of protein, then rising at intakes of 0.6 and 1.5 g protein • kg" 1 • day" 1 . Protein breakdown did not respond to intake in the fed state, appearing to be depressed to the same low level by feeding at any level of protein. However, while fasting, breakdown rates were also lower with the lower intakes of protein. These changes in rates during the fasting state are additional evidence that protein metabolism not only responds to the immediate intake of food but also undergoes adaptation to varying intakes of protein over a longer period. These data can also be viewed from a different perspective. The differences in protein synthesis and degradation between the fasted and fed states represent the responses to feeding, and these vary with the dietary protein content. At very low intakes of protein there was very little effect of feeding on any of the parameters measured, consistent with the data described above with low-energy diets (28). At the higher levels of intake, the effect of feeding was mainly to inhibit protein degradation and enhance leucine oxidation. The effect of feeding on protein synthesis, however, even at the high protein intakes, is not definitive. The data of Yang et al. (3) suggest that synthesis rises on feeding at high intakes, whereas the results of Motil
DIABETES CARE, VOL. 14, N O . 12, DECEMBER 1991
et al. (31) show no changes at any protein level. Unfortunately, studies of the effects of feeding on whole-body rates of protein turnover reveal one of the difficulties of interpreting the data with a simple twopool model of protein metabolism (Fig. 1). Although it might be reasonable to regard the splanchnic tissues as part of a single homogeneous body pool in the fasted state, during feeding, the unlabeled amino acid from the diet first enters the gut and liver. These tissues extract a part and modify its isotopic enrichment before it can enter the peripheral circulation, which is the site of blood sampling. Methods that use multiple labels are being developed in an attempt to analyze such data more rigorously (33). Supporting evidence for the pattern of changes of protein turnover with dietary protein intake described above comes from a study by Meredith et al. (34), who gave diets containing 0.6, 0.9, and 1.2 g proteinkg" 1 • day" 1 to endurance-trained men for 10-day periods. Protein turnover was determined by repeated oral dosages of [15N]glycine over 60-h periods with measurements on urinary urea. Nitrogen excretion (protein oxidation) increased with increasing dietary protein, as did the rate of whole-body protein synthesis, but protein breakdown remained unaltered. The estimated protein requirement to maintain nitrogen balance was —0.9 g protein • kg" 1 • day" 1 , but no differences in the responses to changes in dietary protein were apparent when diets above and below the maintenance level were given. This study showed no differences in protein kinetics or protein requirement between young and middle-aged men. A somewhat different picture is suggested by Stuart et al. (35), who measured protein turnover by primed
1193
PROTEIN INTAKE AND BODY PROTEIN TURNOVER constant infusion of [1-13C]leucine in subjects given 0.6 or 1.0 g protein • kg" 1 • day" 1 for 5 days. Measurements were made only in the fasting state and showed that leucine oxidation was higher and protein degradation lower on the higher-protein diet. However, protein synthesis was not significantly altered. Although the authors discuss several differences in the experimental procedures that might have caused the responses to be different from those reported previously (e.g., dietary energy level and their use of ketoisocaproate to calculate rates of turnover compared with plasma leucine), the explanation for the discrepancies is not clear. Stuart et al. (35) also examined the response of protein kinetics to insulin by the insulin clamp procedure. However, despite the lower protein degradation on the higher-protein diet, the maximal and half-maximal suppressions of degradation by insulin were not different. In the studies discussed so far, none have investigated the metabolic effects of very-high-dietary protein levels. There appears to be only one report of protein turnover rates in adults given excess protein (30). Primed infusions of [1-13C]leucine were given to volunteers in the postabsorptive state after receiving 1.5 or 3.9 g protein • kg" 1 • day" 1 for 8 days. Only rates of protein degradation were reported, because these were part of a more extensive study of alanine kinetics and were unaltered by the change in dietary protein. Although increases in dietary protein
