Endocrinology and Metabolism

Quantifying rates of protein synthesis in humans by use of 2H2O: application to patients with end-stage renal disease

Stephen F. Previs, Richard Fatica, Visvanathan Chandramouli, James C. Alexander, Henri Brunengraber, Bernard R. Landau

Abstract

A method is introduced for quantitating protein synthetic rates in humans by use of 2H2O. Its validity was tested in subjects with end-stage renal disease. Six clinically stable subjects, hemodialyzed three times weekly, ingested 2H2O to a body water 2H enrichment of ∼0.4%. On dialysis, body water enrichment declined to ∼0.1%. Enrichment of the α-hydrogen of plasma free alanine was also ∼0.4% before and ∼0.1% after dialysis. β-Hydrogen enrichment was ∼80-100% of α-hydrogen enrichment. 2H2O was ingested to replace 2H2O removed after each dialysis for 15-51 days, returning enrichment to ∼0.4%. Enrichment of alanine from plasma albumin gradually increased, with again ∼80-100% as much 2H in β- as in α-hydrogens. With continued dialyses, without 2H2O replacement, alanine from albumin enrichment gradually declined, whereas free alanine and water enrichments were negligible. The fractional albumin synthesis rate, calculated from the increase in enrichment in alanine from albumin, was 4.0 ± 0.5%/day, and from the decrease, 4.6 ± 0.2%/day. Thus body water enrichment in a subject given 2H2O can be maintained constant long term. A rapid exchange, essentially complete, occurs between the hydrogens of alanine and body water. An integrated measure over a long period of albumin's synthetic rate can be estimated from both the rise in enrichment of alanine from the protein during 2H2O ingestion and fall on 2H2O withdrawal, while the subject's living routine is uninterrupted. Estimates are in subjects with renal disease, but the method should be applicable to estimates of protein synthetic rates in normal subjects and in other pathological states.

  • protein
  • albumin
  • alanine
  • deuterium oxide
  • hemodialysis

several approaches are used to quantitate protein dynamics in humans (7, 38). Most depend on the labeling of protein following administration of a labeled amino acid. Sampling of the labeled amino acid in the systemic circulation is assumed to accurately reflect the label in the precursor pool from which synthesis of the protein proceeds. Assumptions made of a single pool of the amino acid and a uniform distribution of the labeled amino acid in that pool remain questionable. Primed continuous infusions and flooding doses of the labeled amino acid have been employed in attempts to achieve equilibration of label in the precursor pool. Rates of synthesis of proteins have also been estimated from the incorporation of 14C from [14C]bicarbonate into proteins, measurements of urea kinetics being required (18, 32), and rates of synthesis of albumin interpreted from the catabolism of radioactive iodine-labeled albumin (22, 32).

In 1941, Ussing (36) introduced the use of 2H2O in animals to measure rates of “protein renewal,” i.e., hydrolysis and resynthesis of proteins. He postulated that, if a concentration of 2H2O was maintained in the water phase of an animal, 2H in its proteins would increase with time, approaching a constant value determined by the renewal rates, the labeling occurring by synthesis of the proteins from free amino acids that had exchanged their H for 2H via transamination or similar processes or during their synthesis from substances of nonprotein origin. He reasoned that quantitation would be simplest if exchanges in the free amino acids were so rapid that their 2H contents were the same at every moment during synthesis, the transamination rate being a determinant. 2H2O was injected into rodents to a body water enrichment of ∼2.5%. Renewal rates were estimated from 2H contents in protein mixtures from various organs and in myosin and hemoglobin. Methods used to isolate the proteins remained a concern. Lack of heavy water in Denmark in 1941 prevented further study (36).

Rates of synthesis of fatty acids (30), cholesterol (33), and glucose (23) in humans have been estimated from 2H enrichments following 2H2O ingestion. We now report results in developing a method for quantitating rates of protein synthesis in humans following 2H2O ingestion. The method stems from the study of Ussing (36) and several more recent reports. Oshima and Tamiya (25) showed that alanine transaminase action results in the exchange of the hydrogen of 2H2O with both the α-hydrogen and β-hydrogens of alanine. Humphrey and Davies (13, 14) incubated cells from duckweed plant in a medium containing 3H2O and then transferred the cells to a 3H2O-free medium. The increase and then decline in the 3H in amino acids from protein in the cells provided a measure of protein turnover. Changes in 3H bound to the α-carbons of the amino acids were used in the quantitations. Commerford et al. (3) determined the distribution of 3H among the amino acids from serum proteins from mice exposed to 3H2O for 3 mo. The 3H specific activity of alanine was the same as that of tissue water.

We postulated that, after 2H2O ingestion by humans, the hydrogens in free alanine in the systemic circulation by exchange would rapidly approach the 2H enrichment in body water and that enrichment would then be the enrichment of the alanine used in the protein's synthesis. The rate of protein synthesis could then be estimated from the rate of increase in the 2H enrichment of the alanyl units of that protein, since the hydrogens of those units could no longer be exchanged, i.e., be subject to the action of transaminase. These postulates were tested in patients with end-stage renal disease undergoing chronic hemodialysis. Changes in the 2H enrichments in free plasma alanine could then be compared with the changes in 2H2O content in body water as a result of dialysis, providing a measure of the rapidity of the exchange. The rate of increase in the 2H enrichments in alanine from plasma albumin, measured during 2H2O exposure, and the rate of decline, following 2H2O removal by dialysis, could both then be used to estimate the rate of synthesis of the albumin. Because the rate estimated from the decline depends solely on measurements of enrichments in albumin, whereas the rate estimated from the increase also depends on the measurements of the enrichment in free alanine, comparison of the estimates provides a test of the assumption that the enrichment of the free alanine was that of the alanine used in the synthesis of the albumin. A portion of this work has been reported in abstract form (27).

MATERIALS AND METHODS

Subjects. Six patients, three men and three women, were recruited from the Ohio Renal Care East Dialysis Center. They were clinically stable while on hemodialysis treatment for a minimum of 12 mo. They are characterized in Table 1. Three were dialyzed at the Center on Mondays, Wednesdays, and Fridays and three on Tuesdays, Thursdays, and Saturdays. Dialyses were begun in the morning and were of 3.5-4.5 h in duration. The Institutional Review Board of the Cleveland Clinic Foundation approved the protocol and informed-consent document. Written informed consent was obtained from each subject.

View this table:
Table 1.

Characteristics of the subjects

Study design. The three subjects dialyzed on Fridays began the study by drinking, over a weekend, a loading dose of 2H2O (99.9 atom % 2H) purchased from Isotec (Miamisburg, OH). The other three subjects drank the dose over a Sunday and Monday. Thus the subjects drank the loading dose over a 2-day period when not dialyzed. The dose was calculated to give a body water enrichment of 0.4-0.5%, i.e., 4-5 ml of 2H2O/kg body water (23). The dose was ingested in four equal portions with ≥1 h separating the ingestion of each portion to avoid side effects of vertigo and nausea (23, 33).

After ingestion of the loading dose, venous blood (10 ml) was collected from the first two subjects via their indwelling catheters before each dialysis for 26 days, 5 ml were collected for 40 days from the second two subjects, for 54 days from the fifth subject, and for 70 days from the sixth. Blood was also collected from the first subject after his first two dialyses following the ingestion of the loading dose and from the other subjects only at the end of their first dialyses.

On returning home after each dialysis, the first two subjects for the first 15-17 days, the second two subjects for the first 26 days, the fifth for the first 31 days, and the sixth for the first 51 days drank a dose of 2H2O equal to 75% of their loading doses, an amount estimated to replace the amount removed by the dialysis. The 75% estimate was based on the expectation that 2H2O would be removed by dialysis to about the same extent as urea (Table 1) and on previous measurements that we had made in two other patients who had ingested 2H2O and in whom we had also then measured body water enrichment before and at the end of dialysis. The replacement dose was ingested in three equal portions spaced ∼1-2 h apart.

Analyses. Blood samples were collected in EDTA Vacutainer tubes (Becton Dickinson, Franklin Lakes, NJ) to which soybean trypsin inhibitor was added (9). Plasma was separated by low-speed centrifugation and stored at -80°C until analyzed.

Enrichments in 2H2O in plasma, equated to that in body water, was determined as described by Yang et al. (39). Plasma (50 μl) was spiked with acetone and NaOH. After 24 h to allow exchange between the hydrogens of water and the acetone, the acetone was extracted into chloroform and assayed for 2H enrichment by gas chromatographymass spectrometry. A correction, a 2.3% increase, was made for the small water volume used to dissolve the EDTA and trypsin inhibitor [0.1 ml in the 5-ml Vacutainer and 0.2 ml in the 10-ml Vacutainer, assuming blood is 85% water (5)]. Enrichment in body water is expressed in atom percent excess, i.e., the percentage of the hydrogen atoms in body water, above natural abundance, that are deuterium atoms.

For assaying the 2H enrichment in free alanine in plasma, 2.0 ml of methanol were added to 0.2 ml of plasma. The supernatant was evaporated, and the residue, dissolved in 0.25 N HCl, was applied to a cation exchange column (AG 50W-X8, hydrogen form; Bio-Rad Laboratories, Hercules, CA). The column was washed with water and then the amino acids eluted with 4 N NH4OH. The eluate was evaporated.

For assaying the 2H enrichment in alanine from plasma albumin, 1 ml of trichloroacetic acid was added to 200 μl of plasma. The precipitate was washed with 5% trichloracetic acid, and albumin was then extracted from it into absolute ethanol (1, 2, 16). The preparation gave a band with the mobility of human albumin on acrylamide gel electrophoresis. There were two other bands containing ∼2% of the protein applied. The ethanol was evaporated, 3 ml of 6 N HCl were added, and the mixture was heated at 100°C for 24 h. The hydrolysate, diluted to 0.5 N HCl, was also applied to a AG 50W-X8 column, the column was washed with water, the amino acids were eluted with 4 N NH4OH, and the eluate was evaporated.

The eluates, containing the free alanine from the plasma and the alanine from the albumin, were treated with N,N-dimethylformamide dimethyl acetal (Pierce, Rockford, IL) as described by Thenot and Horning (35). The 2H labeling of alanine was determined using an Agilent 5973N-MSD equipped with an Agilent 6890 GC system. A DB17-MS capillary column (30 m × 0.25 mm × 0.25 μm) was used in all analyses. The temperature program was 90°C initial, hold 5 min, increase by 5°C/min to 130°C, increase by 40°C/min to 240°C, and hold for 5 min. The split ratio was 15:1 with a helium flow of 1 ml/min. Alanine elutes at ∼12 min. The mass spectrometer was operated in the electron impact mode. Selective ion monitoring of mass-to-charge ratios (m/z) 158 and 159 (total 2H enrichment of alanine) and 143 and 144 (2H enrichment of the α-hydrogen of alanine) was performed using a dwell time of 10 ms/ion.

The molecular weight of the unlabeled derivative is 158 and that of the derivative containing a single deuterium atom is 159. At a body water enrichment of 0.4%, labeling of alanine molecules with more than one deuterium atom is insignificant. Thus the m/z 159/(159 + 158) is the fraction of the alanine molecules enriched in deuterium. That fraction, expressed as a percentage, is the mole percent excess, enrichment in the alanine, i.e., above natural abundance. On electron impact ionization the methyl group, carbon-3 of the alanine with its three β-hydrogens is cleaved, yielding mass ion 143, i.e., a fragmentation 15 less in molecular weight. Thus the m/z 144/(144 + 143) is the fraction of the alanine molecules having a deuterium atom bound to their α-carbons; hence, expressed as a percentage, it is the atom percent excess of 2H in the α-hydrogen. The fraction of alanine molecules enriched in deuterium, less the fraction enriched in deuterium bound to the α-carbon, is the fraction of molecules having deuterium bound to their three β-carbons. That fraction divided by 3, again expressed as a percentage, is then the atom percent excess of 2H in each β-hydrogen.

To amplify the detection of low levels of 2H, the method developed by Katanik et al. (21) was used. Briefly, 2H-labeled molecules typically fractionate on gas chromatography columns (e.g., 2H-labeled alanine elutes slightly before unlabeled alanine). Integrating a limited region of the chromatographic peaks improves the detection of 2H enrichment (21). The integration routine was set to quantify the initial 25% of the chromatographic peaks.

Applying the “fractionation method” required the use of standards of 2H-labeled alanine. Typical standard curves yield a similar response in the measured-to-expected ratio regardless of whether the total labeling of alanine or that of the α-hydrogen is quanitified (i.e., the slopes of the linear regression are the same). However, the y-intercept is ∼1.1% greater when the total labeling of alanine vs. that of the α-hydrogen is measured. That is expected because the ion of mass 158 (and 159) contains one additional carbon atom compared with the ion of m/z 143 (and 144). Coefficients of variation on five separate determinations of the enrichments in alanines with enrichments of 0.10, 0.25, 0.60, 1.0, 1.5, and 2.0%, prepared from l-[2- 2H]alanine (98 atom % 2H, purchased from Isotec), were, respectively, 8.8, 8.6, 5.7, 5.1, 2.9, and 2.0%.

Calculations. Enrichments in body water and free alanine in plasma during 2H2O administration were calculated as the means of those enrichments just before the beginning of dialysis on days 1-17 and 19 in the first two subjects, on days 1-29 in the second two subjects, days 1-33 in the fifth subject, and days 1-54 in the sixth. The enrichment in each of the β-hydrogens of alanine in plasma and from albumin was calculated as a percentage of the enrichment in the α-hydrogen.

The fractional synthetic rate (FSR) of albumin was estimated using the mean enrichment in free alanine in plasma from day 1 to day 17, calculated from the enrichments before the dialyses during those days, i.e., days 1, 3, 5, 8, 10, 12, and 15. That mean was corrected for the lower enrichment in plasma alanine occurring during each dialysis, 3.5-4.5 h, the time, ∼4 h, until 2H2O replacement was begun, and the period until replacement was complete, ∼4 h. Because ∼75% of the 2H2O in body water was removed by dialysis, and free alanine had the enrichment of body water, the enrichment in free alanine during the ∼4 h following a dialysis was 25% of that before dialysis. The mean enrichment of the alanine during the ∼4 h of dialysis was then 62.5% of that before dialysis, i.e., [25 + (100-25)/2], assuming a linear decline in enrichment during dialysis. During the ∼4 h of 2H2O replacement, assuming a linear increase in enrichment, the mean was also 62.5% of that before dialysis. Thus the mean enrichment during those ∼12 h was [(4 + 4)(0.625) + 4(0.25)]/12 = 0.5, i.e., 50% of that before dialysis. Because there are 384 h in 16 days and there were 7 dialyses from days 1 to 17, the corrected mean was the uncorrected mean times [300 + (12)(7)(0.5)]/384, i.e., times 0.89.

Enrichment in alanine from albumin is not changed by proteolysis, since the enrichment in the alanyl units of albumin removed by proteolysis is the same as in the alanyl units remaining in the albumin. Thus the rate of increase in the enrichment in alanine from albumin is determined by the rate of synthesis of the alanyl units of albumin from the plasma free alanine. The free alanine has a higher enrichment than the alanyl units of the albumin until all the units have been synthesized from the enriched free alanine, i.e., until the enrichment of the alanine from albumin is that of the free alanine. Subtracting from the enrichment in free alanine the enrichment in the albumin as it increases from zero then effectively allows the natural logarithmic plot of the daily decline in the difference to reflect the decline in the unenriched albumin. That decline in the difference must be at the same fractional rate as the increase in the enrichment in albumin. The rate of decrease in the enrichment of the alanine from albumin, when 2H2O is withdrawn, is determined by the rate of synthesis from then-unenriched free alanine. The enrichment in the alanine from the albumin will continue to decrease until it is unenriched.

Therefore, the natural logarithms of the differences between the corrected mean enrichment in free alanine in plasma and the enrichment in alanine from albumin before each dialysis on those 16 days were plotted as a function of the day of dialysis. The values were fitted to a linear regression line, the negative of the slope then equaling the FSR, i.e., the fraction of the albumin pool synthesized per day, expressed in %/day by multiplying by 100. The relationship between the enrichment in free alanine (AF) and the FSR is then expressed by AF[1-e-(FSR)t], where t is time in days, the enrichment rising from 0 to AF. The FSR of albumin for each subject was also calculated from the slope of the linear regression line obtained on plotting the natural logarithms of the enrichment in alanine from albumin as a function of the days after 2H2O withdrawal. The relationship between the enrichment in alanine from albumin and the FSR is then Math, where t is time in days after withdrawal of 2H2O and Math is the enrichment in alanine from albumin at the time of withdrawal, the enrichment falling from Math to 0. If the rate of synthesis is more or less than the rate of proteolysis, the size of the pool of albumin will change. The FSR is assumed to remain constant; i.e., if the pool size changes, the same percentage of the pool is still assumed to be synthesized per unit time.

RESULTS

Figure 1 depicts in the six subjects the 2H enrichments in body water and in the four hydrogens bound to the α- and β-carbons and only the hydrogen bound to the α-carbon of alanine, free in plasma and released on hydrolysis of the plasma albumin. Body water enrichment before each dialysis (Fig. 1, •) remained constant, ranging from 0.36 to 0.44% (Table 2), from before the first dialysis, after the ingestion of the 2H2O loading dose, until discontinuation of the replacement dose. At the end of the first two dialyses in the first subject and after the first dialysis in the other five subjects and before the beginning of ingestion of the replacement dose, enrichment had declined by ∼80% (Table 2). After the two dialyses, following discontinuing replacement, the enrichment in body water was negligible [days 19 and 21-26 in the first two subjects, days 31-40 in the second two subjects, days 33-54 in the fifth subject, and days 54-70 in the sixth subject (Fig. 1)].

Fig. 1.

Plots of enrichments for the 6 subjects in body water and alanine free in plasma and from albumin. Enrichments are in the hydrogen of body water (•), the α-hydrogen in free plasma alanine (▪) and in alanine from albumin (▴), and in all 4 hydrogens (α + β) of free plasma alanine (□) and alanine from albumin (▵). Enrichments in the hydrogen of body water and the α-hydrogen of alanine are in atom % excess; enrichment in alanine, α + β, is in mole % excess, i.e., %alanine molecules having a deuterium atom in one of their hydrogens, whether α or β.

View this table:
Table 2.

Enrichment in body water and enrichment in α-hydrogen of plasma free alanine, their ratio, enrichment in the 4 hydrogens (α + β) and enrichment in α-hydrogen relative to enrichments in the 4 hydrogens of plasma free alanine and of alanine from albumin, enrichment in β-hydrogen as percentage of that in α-hydrogen, and %decline in enrichment in body water and hydrogens of plasma free alanine on dialysis

The enrichment of the α-hydrogen of the free alanine in plasma before each dialysis (Fig. 1, ▪), when the replacement doses were given, was about that of body water (ratios 0.99-1.14; Table 2). The enrichment of a β-hydrogen was 80-100% (Table 2) of the enrichment of the α-hydrogen. At the end of the first two dialyses of subject 1 and after the first dialysis of each of the other five subjects, the enrichment of the alanine had declined by ∼75%, about the same percent decline as of body water (Table 2). When the replacement doses of 2H2O were stopped, the enrichment of alanine declined, just as did body water, to negligible levels (Fig. 1). In subject 2 (Fig. 1), in whom on days 10 and 12 blood drawing was delayed until 20 min after dialysis was begun, there was a small decline in both enrichments in body water and alanine. In subject 6, for an unknown reason, body water was lower on days 29 and 31 than on the other days he ingested 2H2O, with a corresponding reduction in the enrichment in alanine on those days (Fig. 1).

The enrichment in alanine from albumin gradually increased (Fig. 1, ▵ for α- plus β-hydrogens and ▴ for α-hydrogen), whereas the initial enrichment in body water was maintained and then declined when 2H2O ingestion was stopped and dialyses continued. Enrichment of alanine from albumin reached 37 and 47% of that in plasma in the first two subjects, 61 and 80% in the second two subjects, 78% in the fifth subject, and 68% in the sixth. These percentages are estimated from the mean of the enrichment in four hydrogens of plasma alanine relative to the peak enrichment of the four hydrogens in alanine from albumin (Fig. 1), corrected using the factor of 0.89, for the lower enrichment in plasma alanine during dialysis and until replacement was complete (see Calculations).

Figure 2A depicts the linear regression line, with its correlation coefficient (r2) calculated for each subject from the natural logarithms of the differences between the mean enrichment in the α- plus β-hydrogens of plasma free alanine and the enrichments in the α- plus β-hydrogens of alanine from albumin on dialysis days 1-17 and in Fig. 2B for just the enrichments in the α-hydrogen. The FSR, estimated from the slopes of the lines during that 16-day period, was from the enrichments in the α- plus β-hydrogens 4.0 ± 0.5%/day (mean ± SE) and from the α-hydrogen 4.4 ± 0.6%/day.

Fig. 2.

A: natural logarithms of differences between the mean enrichment in the α- + β-hydrogens of plasma alanine and the enrichments in the α- + β-hydrogens of alanine from albumin before each dialysis from days 1 through 17 plotted as a function of the day of dialysis. The measure of the fractional synthesis rate (FSR, in %/day) is then the negative of the slope of the linear regression line fitted to the data. B: same as in A except just for enrichment in the α-hydrogen. Correlation coefficients (r2) are recorded.

In Fig. 3A, the natural logarithms of the α- plus β-enrichments and in Fig. 3B of just the α-enrichments in alanine from albumin, for each subject, are plotted as a function of the days after 2H2O withdrawal. The data have also been fitted to linear regression lines and correlation coefficients given. The FSR estimated from α- plus β-enrichments for the six subjects (Fig. 3A) was 4.6 ± 0.2%/day (mean ± SE). The rate was much more variable, 7.1 ± 1.7%/day, estimated from the α-enrichments (Fig. 3B). Attempts to assay enrichment in the α-hydrogen of alanine from albumin from subject 2 were unsuccessful.

Fig. 3.

Natural logarithms of enrichments in the α- + β-hydrogens (A) and α-hydrogen of alanine from albumin (B) plotted as a function of days after withdrawal of 2H2O. Enrichment at day 0 is that before the 1st dialysis following discontinuing 2H2O. FSR (%/day) then equals the negative of the slope of the linear regression line fitted to the data. Correlation coefficients (r2) are also recorded.

DISCUSSION

Ingested 2H2O distributes uniformly throughout body water (31). The carbon-bound α- and β-hydrogens of alanine exchange with the hydrogens of body water. Origins of the alanine are synthesis, release from stored protein, and food ingestion. The exchange is an intracellular process catalyzed by alanine transaminase. The rate of exchange of the α- and β-hydrogens can differ (4, 25). Alanine transaminase activity is highest in liver (20), the site of synthesis of albumin. Exchange is assumed to be so rapid that the 2H enrichment of the alanine used in synthesis of the protein approaches that in body water. The enrichments of the alanine used in synthesis and of the free alanine in plasma are assumed to be the same.

Studying dialyzed patients allowed testing of those assumptions. Rapid exchange between the hydrogens of free alanine in plasma and the hydrogens of body water is evidenced by similar declines in their enrichments on dialysis. Exchange of the α-hydrogen with the hydrogen of water was complete, as the enrichments of the α-hydrogen and water were the same (• vs. ▪ in Fig. 1 and 2H2O/α column in Table 2). That finding also provides evidence for the enrichment of intracellular alanine being the same as that in plasma alanine. Thus the α-hydrogen of intracellular alanine could not have an enrichment higher than that of the 2H of body water, its precursor, and, hence, that of plasma alanine. Furthermore, to the extent that the precursor of enriched plasma alanine is intracellular alanine, the enrichment in that intracellular alanine could not have been less than the enrichment in plasma alanine. Because transaminase activity, released from cells, is found in plasma, along with alanine, pyruvate, glutamate, and α-ketoglutarate, exchange of the hydrogens of alanine with those in body water could have occurred to an extent in the systemic circulation.

In the first four subjects, enrichment in the β-hydrogen was ∼80% of that in the α-hydrogen, and thus exchange with the 2H of 2H2O was extensive but not complete. However, in the last two subjects, the enrichment of the β-hydrogen was ∼100% of that in the α-hydrogen. We have been unable to find an explanation for that difference. An exchange of ∼80% in the β-hydrogens is in accord with the finding that hydrogens bound to carbon-6 of blood glucose from fasting subjects given 2H2O have ∼80% of the 2H enrichment of the hydrogen bound to carbon-5 of the glucose (23). That is because the β-hydrogens of pyruvate, which become enriched during the transamination, are the precursors of the hydrogens bound to carbon-6 of glucose formed by gluconeogenesis, hepatic pyruvate and alanine being in isotopic equilibrium (17). The hydrogen at carbon-5 of the glucose takes on the enrichment of body water in the hydration of phosphoenolpyruvate. Further evidence for the enrichment of plasma alanine reflecting that of the labeled alanine used in the synthesis of the albumin is the similar percentages of 2H in the α- and β-hydrogens of plasma alanine and of alanine from albumin (Table 2).

Because the enrichment in the α-hydrogen of plasma alanine is the same as that in water, in practice the rate of synthesis of a protein could be calculated only from the enrichment in body water and the α-hydrogen of the alanine of the protein. That would be analogous to quantitating gluconeogenesis by using 2H2O (23), where, because of rapid exchange, catalyzed by phosphohexoisomerase, the enrichment of the hydrogen bound to carbon-2 of glucose is the same as of the hydrogen of body water. The limitation of a low enrichment of the hydrogen at carbon-2 (as well as at carbon-5) in glucose was overcome in the quantitation of gluconeogenesis by recourse to a hexa-methylenetetramine derivative (23). Here, we used isotopic discrimination to increase the sensitivity of the assay of the enrichment of the α-hydrogen (21). Although the α-hydrogen of plasma alanine was then found to have the enrichment in water, at relatively low enrichments encountered in the α-hydrogen of alanine from albumin, the assay appeared less reliable for quantitating the rate of synthesis of the protein than assay of the sum of the enrichments in all four hydrogens, and that assay is then recommended for use.

When alanine is converted to alanyl units of albumin, the α-and β-hydrogens are no longer exchangeable, peptide formation removing their accessibility to transamination. When the albumin is hydrolyzed using HCl, the deuterium remains bound to the alanine carbons and therefore can be assayed. When the replacement dose of 2H2O is stopped and body water enrichment becomes negligible, the [2H]alanine released on proteolysis is exposed to transamination, hence losing its label by exchange with the unlabeled body water. Thus the rate of synthesis is quantitated from the decline in enrichment in the albumin without the need to correct for continued synthesis from [2H]alanine. The mean FSRs calculated from the increase and decline in the enrichments in alanine from albumin were ∼4.5%/day. Because the calculation of the FSR from the increase, but not the decline, depends on assuming that the enrichment in the plasma alanine was that in the alanine used in the synthesis, that the FSRs are similar supports the assumption that the enrichment in plasma alanine was that in alanine used in the synthesis.

Conceivably, the synthesis of albumin from [2H]alanine released by proteolysis of proteins labeled during 2H2O administration, including albumin, could have occurred after 2H2O withdrawal. However, that labeled alanine could not then have found expression in the enrichment of plasma alanine, since that enrichment was then negligible and would not have experienced transamination with the loss of its label before conversion into albumin.

Because synthesis and breakdown of albumin occur intracellularly, enrichment in albumin at a given time may be different at sites other than the intravenous site sampled (6, 8), but relative changes in the enrichment in the plasma albumin over time presumably still reflect changes in the enrichment at those sites. The FSR, calculated from the increase in enrichment in the albumin, could be an underestimate if alanine from dietary sources was absorbed into the portal vein, and, having a lower enrichment than in alanine peripheral vein blood, was then used in synthesis. The enrichment of the α-hydrogen in free alanine in plasma equaling that in the hydrogen of body water indicates that equilibration of the alanine hydrogens was complete at least by the time alanine entered the hepatic vein.

As indicated, the correction factor for the lower alanine enrichment during dialysis and until 2H2O replacement was complete is an approximation. Thus the synthesis of albumin is less, about one-half as much (6), in the fasted than in the fed state, and the ∼12 h on each of the 3 days/wk of dialysis, when plasma alanine enrichments were reduced, were during meal times. The fractional proteolytic rate should have equaled the FSR unless the size of the albumin pool changed. The subjects were clinically stable and their serum albumin concentrations did not change.

A rate of ∼4.5%/day is comparable to rates obtained by other methods (2, 10, 34). However, in those methods, measurements are made over a period of hours under a steady-state condition, e.g., the fasted or fed state. A labeled precursor is given in sufficient amount to label adequately the protein for measurement within that period, and the study is usually done in a clinical center setting. Thus recently, the effect of hemodialysis on the turnover of proteins in patients with end-stage renal disease undergoing chronic dialysis was measured using state-of-the-art primed constant infusions of stable isotope-labeled leucine and phenylalanine (15, 29). The patients were fasted and measurements were made over an 8.5-h period in a clinical center.

Measurements using 2H2O are made over many days under daily living conditions, i.e., fasting, eating, sleeping, etc. The subjects should be in steady state as regards their daily routine during the period of measurement, e.g., diet, exercise, and medications. Except for drinking 2H2O, the subject's usual daily routine is unaltered; e.g., there is no need for admission to a clinical research center and infusions. Cost is only for 2H2O and blood sampling. A subject with a body water weight of 40 kg, would, to achieve and maintain a body water enrichment of 0.5%, ingest a loading dose of 200 ml and then 450 ml/wk, i.e., 3 × 150 ml/dialysis. The method then has the potential to measure the effect of treatment, e.g., diet, frequency of dialysis, and/or medication, on protein turnover, a possible measure of a patient's well-being, while the patient maintains his or her normal routine.

To apply the method to subjects not undergoing dialysis, after giving a loading dose to maintain a constant body water enrichment, 2H2O need be replaced only to the extent of the daily turnover of body water. The daily replacement dose in normal circumstances would then be expected to be ∼10% of the loading dose (33), i.e., in a subject having a body water weight of 40 kg the loading dose again would be 200 ml and the daily dose 20 ml to maintain an enrichment of 0.5%. Because the half-life of 2H2O in body water is many days, when replacement is stopped, a correction to the estimated rate of synthesis would be required for the continued synthesis of labeled protein. An exception would be when the protein measured has such a long half-life that it is still adequately labeled when body water enrichment becomes negligible.

For the method to be applied to a protein other than albumin, the turnover time of the protein would have to be sufficiently long to allow intake and mixing of the 2H2O with body water and then measurements of the rate of increase in enrichment in the protein. Rates of synthesis of tissue proteins have been measured in rodents by means of this approach (28) and mass isotopomer distribution analysis (12), but at body water enrichments of 2.5-3.0%. 2H2O enrichments of 1.5-2.0% in body water have also recently been maintained in humans for many weeks, initiated by administering a loading dose of 2H2O over 18-24 h (24). We aimed for a body water enrichment of ∼0.5% because of extensive evidence for safety at that enrichment (19, 23, 26). In our subjects, the need to replace the 2H2O removed by dialysis within several hours further limited the size of the dose that could be given. No side effects occurred as a result of the 2H2O administration. By use of similar dose schedules none have occurred in normal and diabetic subjects.

Estimates made by measuring 2H enrichments in an amino acid other than alanine may prove as satisfactory as using alanine and perhaps more so. Thus Commerford et al. (3) reported in mice given 3H2O for 3 mo that the 3H specific activities of the hydrogens of glycine from serum proteins were also the same as in body water. Specific activities of the hydrogens of serine, glutamate, and aspartate were somewhat less those of glycine (3). Glycine units represent only 1% of the amino acids units of albumin, but large percentages are in collagen and its derivatives (11).

In conclusion, 2H2O on ingestion equilibrates with body water. 2H2O removed by dialysis can be replaced to maintain body water enrichment near constant for long periods. The enrichment in the α-hydrogen of free plasma alanine rapidly equilibrates with that in body water, as evidenced by their same enrichments at the beginning and end of dialysis. The enrichment of the β-hydrogen is 80-100% of that in water. The rate of increase in the 2H enrichment in alanine from plasma albumin on 2H2O ingestion provides an estimate of the rate of synthesis of the albumin. The decrease in that enrichment on discontinuing 2H2O ingestion provides another estimate of that rate. The estimates are obtained without interrupting the daily routine of subjects.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-14507, DK-07319, and DK-058126, a grant from the Diabetes Association of Greater Cleveland, the Case Western Reserve University Diabetes Research Fund with contributions in memory of Leo Demsey, and the Daniel S. Stein Foundation of the Jewish Federation of Cleveland.

Acknowledgments

We thank Dr. K. Sreekumaran Nair for advice and guidance. We gratefully acknowledge the help of nurse Leslie LaMastra in recruiting and caring for the patients.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

References

View Abstract