AMERICAN JOURNAL OF PHYSIOLOGY Vol. 230, No. 5, May 1976. Printed in U.S.A.
De novo synthesis of alanine perfused rat hindlimb BARBARA Department
GRUBB of Physiology,
Northwestern
University
GRUBB, BARBARA. De novo synthesis of alanine by the perfked rat hindlimb. Am. J. Physiol. 230(5): 1379-1384. 1976. Due to the disproportionately large production of alanine by muscle, it has been suggested that part of the alanine released by muscle is synthesized de novo by the transamination of glucose-derived pyruvate. This glucose-alanine conversion was quantitated in the isolated rat hindlimb perfused with a solution of bicarbonate buffer containing 2% albumin, 2.4% dextran, 2.5-15.9 MM glucose, 32-34% dog erythrocytes, and 0.05 PCilml [‘4C]glucose. Measurement of labeled alanine production allowed quantitation of de novo alanine synthesis. De novo derived alanine accounted for an average of 33% of the total alanine released by the perfused tissue (perfusate glucose concentration 8.3 mM>, concurrently 2.7% of the glucose taken up by the limb was converted to alanine. By increasing the glucose concentration perfusing the muscle, both the rate of glucose uptake and de novo alanine release were increased. Addition of insulin to the perfusate (700 pU/ml) resulted in a significant increase in the rate of glucose uptake and de novo alanine production, but the rate of total alanine release was significantly decreased by the hormone. It was concluded that de novo alanine production accounts for a sizeable portion of the total alanine released by muscle, nevertheless a comparatively small fraction of the glucose carbons are actually transformed to alanine.
dl
case; 1’4C]glucose;
insulin;
quantitation
skeletal muscle, alanine is released larger quantities than most other amino acids ‘(5, 6, substrate for hepatic > and serves as an important gluconeogenesis (17). While alanine comprises less than 10% of muscle protein (16), ZO-30% of the total amino acid output by the perfused rat hindlimb (27) and the human forearm in vivo (6, 22) is accounted for by alanine. It has been suggested that the catabolism of a specific polyalanyl protein could account for this disproportionately large production of alanine, yet such a protein has not been identified (4). Also, during starvation when muscle proteins are being broken down, alanine continues to be released in disproportionately large quantities (6). De novo synthesis of alanine by the transamination of glucose-derived pyruvate could best account for the large production of alanine (6, 17). Indirect estimates have been made as to the quantity of glucose taken up by muscle that may be converted to alanine. Felig and Wahren (7) estimated that 13 and 18% of the glucose taken up by the human leg and foreIN RAT AND HUMAN .
by the
Medical
School, Chicago, Illinois
60611
arm, respectively, may be converted to alanine. Pozefsky and Tancredi (23) found arterial pyruvate infusions resulted in up to a 6.7% increase in the release of alanine by the human forearm. However, it was not possible to determine how much of the alanine released by the arm prior to the pyruvate infusion was actually derived de novo. The purpose of this investigation was to determine the quantity of glucose taken up by muscle that can be converted to alanine and what fraction of the total alanine released was synthesized de novo. Changes in blood glucose concentration and the effect,s of insulin on the glucose-alanine conversion were examined by supplying labeled glucose to the perfused rat hindlimb and quantitating the glucose-alanine conversion. MATERIALS
AND METHODS
Surgical preparation. Male rats, 275-300 g, were allowed free access to food and water until surgery. The rats were anesthetized with sodium pentobarbital (6 mg/lOO g). A transverse abdominal incision was made through the skin, superficial and inferior epigastric vessels were ligated, an incision was made through the abdominal muscle layer, and all pelvic viscera were removed. The caudal artery as well as three pairs of lumbar vessels arising dorsal to the aorta were ligated. The skin was detached from the limb, but left loosely wrapped around the limb to prevent heat loss and desiccation. Only one limb was to be perfused; therefore, the blood supply to the contralateral limb was ligated. A tight ligature also was placed around the ankle joint so that the foot was not perfused. After all vessels not supplying the limb had been ligated, the rat was heparinized (100 U/l00 g) and a polyethylene cannula (PE190) filled with perfusate was placed in the aorta. Then, a larger polyethylene cannula (PE-240) was secured in the vena cava. On completion of cannulation, the rat was sacrificed with an intracardiac injection of sodium pentobarbital and placed in the environmental chamber P’erfusion medium. The perfusion medium consisted of Krebs-Ringer solution, bicarbonate buffer, bovine albumin (fraction V, Pentex-untreated to remove free fatty acids), dextran (mol wt 60,000-80,000), glucose, [14C]glucose, and washed dog erythrocytes. This artificial blood had a hematocrit of 32-34%, 2% albumin, 2.4% dextran, 2.5-15.9 mM glucose, and 0.05 p Ci/ml [14C]glucose. Porcine insulin, 700 pU/ml, was included
1379 Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
1380
B.
in the perfusate of some preparations. Details of perfusate preparation can be found elsewhere (32). Perfusion system. The perfusate was pumped from the arterial reservoir to the oxygenator by a variablespeed, two-channel, microinfusion roller pump (Holter RL 175) (Fig. 1). The limbs were perfused at a constant flow rate (1.40 ml/min per leg), which is in the in vivo range for blood flow through the femoral artery of the rat (24). After leaving the oxygenator, the perfusate entered the environmental chamber and passed through a glass filter packed with Dacron. A few centimeters beyond the filter, a bubble trap was inserted into the line, the side arm of which connected to a Statham pressure transducer, this, in turn, was connected to a Beckman RB Dynograph recorder. After passing through the bubble trap, the perfusate was conducted through the arterial cannula into the rat’s hindlimb. Perfusion effluent flowed from the vena cava into the venous cannula, and thence into the venous reservoir. The latter consisted of a large plastic test tube connected to a soda lime trap to prevent escape of labeled CO, (Fig. 1). The venous reservoir was placed in a beaker of ice to minimize glycolysis by the erythrocytes. The perfusate was not recirculated. AnaZysis of perfusate. The entire venous perfusate was collected for six successive 30-min periods. After collection, the venous samples were stored on ice, each of these 30-min samples was centrifuged and analyzed at the end of each experiment. Several arterial samples taken at varying times during the experiment were treated in a manner identical to that of the venous samples. Plasma alanine concentration was determined
a magnetrc
Stirrer
FIG. 1. Diagram of perfusion system. A plastic beaker in ice constituted arterial perfusate reservoir. By keeping perfusate on ice, glycolysis and bacterial growth were minimized. Perfusate was stirred continuously by a magnetic stirrer. Gentle mixing by Tefloncoated stirring rod caused less hemolysis than occurred in another type of reservoir system previously employed (32). Oxygenator consisted of 12 ft of coiled Silastic tubing (0.058 inch I’D, 0.077 inch OD, Dow Corning Corp.) enclosed in a l-gallon jar (1) into which a gas mixture of 95% O,-5% CO, was pumped. A thermoregulator placed under skin of perfused limb was set to maintain a temperature of 37°C. Since environmental chamber also was maintained at this temperature, perfusate entered hindlimb at 37°C. A CO, trap containing soda lime was placed in the venous reservoir to prevent escape of labeled CO,.
GRUBB
by a microfluorometric enzymatic assay (14). Plasma glucose concentration was determined on a Technicon AutoAnalyzer by the standard ferricyanide reduction method (10). There was no drop in the glucose concentration of the perfusate during its passage through the perfusion system, indicating no significant glucose utilization by the erythrocytes. Also, there was no measurable glucose uptake by the erythrocytes in the venous samples collected and stored on ice. The diaphragm, when supplied with labeled glucose, is able to synthesize labeled glutamine, glutamate, and aspartate in addition to alanine (18). Other types of skeletal muscles may also produce these labeled amino acids. In the present experiments, labeled alanine was separated from the other labeled products and labeled glucose by use of ion exchange resins. To remove glutamate, aspartate, lactate, and pyruvate, the deprotein ized sample, pH 7.0, was applied to an anion exchange column (1 g Dowex-1, acetate form). Because glutamine will pass through the anionic column at pH 7.0, glutamine was converted to glutamate by acidifying and boiling samples for 1 h before applying the samples to the resin. The samples were allowed to cool and brought to pH 7 before applying to the column. Neutral amino acids as well as glucose pass through this column. Deionized distilled water was used to wash the anionic column. This water effiuent containing labeled glucose and alanine was adjusted to pH 2.0 and applied to a cationic exchange resin (1 g Dowex 50, sodium form). The positively charged alanine adhered to the resin, the glucose was washed through the resin. Alanine was eluted from the column with 4 N HCl and subsequently counted on a scintillation counter. In order to provide evidence that only labeled alanine was present in the eluate from the cationic exchange column, three experiments were performed in which venous samples were collected from the perfused limbs. Tritiated alanine was added to the samples before deproteinization, thus they presumably contained only [“Hlalanine and [14C]alanine after having run through both columns. A small volume of the eluate from the cationic column was applied to a thin-layer plate (Eastman) for two-dimensional resolution by electrophoresis (300 V, 10 mA for 30 min, buffer pH 1.85, acetic acid: formic acid:H,O, 78:25:877, vol/vol) followed by conventional ascending thin-layer chromatography (n-butyl alcohol:acetic acid:H,O, 3:l:l) (28). The plate was dried and scanned for radioactivity by an Actigraph III Nuclear Chicago Radiochromatography system. In each experiment, the plate exhibited only one peak which contained the C3H]alanine added to the sample as well as a 14C compound. Because the [3H]alanine peak coincided with the 14C peak it was concluded that the 14C compound was alanine. If other labeled compounds were present, they occurred in such small quantities that they represented background counts. Since recovery of [14C]alanine from the columns was never lOO%, but varied between 90-95% under optimal conditions, the variation in alanine recovery was monitored by adding a known quantity of [3H]alanine to each sample prior to deproteinization. The percent recovery
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
DE
NOVO
ALANINE
SYNTHESIS
BY
RAT
1381
HINDLIMB
of [3H]alanine was then used to correct the [14C]alanine counts in all samples. Presentation of results. Rates of glucose uptake and alanine release were determined by multiplying the arteriovenous concentration difference of these two substances by the perfusion flow rate. The percent glucose that was converted to alanine was determined as follows: 1) Counts per minute (cpm) [14C]glucose uptake = cpm arterial [14C]glucose/ml x % cold glucose uptake. The percent cold glucose taken up was determined by dividing the arteriovenous glucose concentration difference by the arterial glucose concentration and multiplying by 100. 2) Percent glucose uptake converted to alanine = cpm [14C]alanine/ml cpm [14C]glucose uptake/ml
x 100
3) Percent of total alanine release derived de novo = glucose uptake rate X % glucose converted to alanine Total alanine release (per min)
x2
(The numerator is multiplied by 2 since 1 mol of glucose yields 2 mol of alanine.) All results are expressed per 4.5 g dry leg, the average dry weight of the perfused tissue (73.3% H20, 17 g wet wt), because the results per unit dry weight were slightly less variable than when expressed as a function of wet weight.
function of perfusion time (Fig. 2), but approached a steady state after perfusion period 4. The glucose uptake rates measured in this study are within the range of those reported for other perfused rat hindlimb preparations. Ruderman et al. (26) reported that the glucose uptake rate of their perfused hindlimb was 0.5-1.7 pmollmin per 30 g (perfusate glucose concentration 5.5 mM). If expressed in relation to dry tissue weight, as used in this paper, these values would be approximately 0.28-0.96 pmol/min per 4.5 g. In another study the perfused rat hindlimb preparation was found to have a higher glucose uptake rate, 3.8 pmol/min per 30 g (2.15 pmol/min per 4.5 g), but the glucose concentration in the perfusate was 10 mM (29). Figure 3 shows the continual release of alanine throughout the entire 3-h perfusion period. The rate of total alanine release (de novo plus endogenously derived alanine from muscle proteolysis) did not appear to be systematically related to perfusion time. The rate of de novo alanine release (labeled alanine) increased as a function of perfusion time, but the rate of release approached a steady state by perfusion periods 5-6. De novo derived alanine accounted for an average of 33% of the total alanine released by the perfused tissue (periods 5-6); concurrently, 2.7% of the glucose taken up by the limb was converted to alanine. Data are given for periods 5 and 6 only, since a steady-state rate of glucose uptake had not been attained until this time. Effect of glucose-uptake rate on glucose-alanine conversion. Glucose concentrations were varied in the perfusate and the effect of glucose uptake on the rate of alanine release was determined. Within the range of
RESULTS
Viability of preparation. The preparation employed in this investigation remained viable throughout the perfusion as judged by the rate of oxygen consumption and lactate production. The rate of oxygen consumption by this preparation was 4.90 t .36 pmol/min per 4.5 g dry leg, and a steady state of oxygen consumption was maintained for the 3-h perfusion period (unpublished observations). This rate of oxygen consumption agrees well with estimates of oxygen consumption by the rat leg in vivo, 0.05 ml/min per leg (12) (2.38 pmol/min per 4.5 g); the perfused rat hindlimb, 10.8 pmol/min per 30 g (26) (6.12 pmol/min per 4.5 g); rat gracilis muscle in vivo, 0.95 ml/min per 100 g (11) (6.80 pmol/min per 4.5 g)* As part of another study, the rate of lactate production by the limb was measured and found to remain stable after the first 30 min of perfusion (unpublished observations). In a deteriorating preparation, a progressive decrease in oxygen consumption and an increase in lactate production would be expected. Thus, according to these criteria this preparation remained viable throughout the experimental period. Glucose uptake rate. Since preliminary experiments showed the rate of glucose uptake was quite variable during the first 30 rnin of perfusion, no chemical analyses were performed on this venous sample. All of the limbs took up a significant quantity of glucose. The rate of glucose uptake increased slowly as a
.
’
’
1
’
2 30 Minute
2. Effect of perfusion time hindlimb. “Plasma” glucose
on glucose concentration
FIG.
rat
’
’
3 4 Perfuston
J
5 6 Period
uptake rate by perfused was 8.3 mM. Data
shown are meanS * SE (N = ‘)* .30
UI-25: 3
9-G Tot al Alanine
i!.-E
5
;
De Novo
123456 30 Minute FIG.
alanine
3. Effect release.
Alanine
.lO
Perfusion
of perfusion time on rate Plasma glucose concentration
shown are means k SE (N = 9).
Release
Release t
T
Period
of total and de novo was 8.3 mM. Data
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
1382
B. GRUBB
glucose concentrations studied, the rate of glucose uptake during periods 5 and 6 was found to be linearly related to plasma glucose concentration (Fig. 4). The glucose-uptake rates shown in Fig. 4 are plotted against their respective rates of alanine release (Figs. 5 and 6). The rate of de novo alanine release- (Fig. 5) was correlated to the rate of glucose uptake. As illustrated in Fig. 6, no correlation existed between the glucose-uptake rate and the rate of total alanine release; nor was there a correlation between the glucose-uptake rate and the percent glucose that was converted to alanine. On the average, 2.7% of the glucose taken up by the limb was converted to alanine. Effect of insulin on glucose-alanine conversion. In some experiments, porcine insulin was included in the perfusate to determine the effect of this hormone on alanine release. An analysis of variance indicated that the rate of glucose uptake by the limbs was significantly enhanced by insulin (F = 21.97, P c 0.01) (Fig. 7). In this preparation, inclusion of 0.7 mu/ml porcine insulin enhanced the glucose uptake rate 22% (average of all periods). The level of insulin used was a pharmacological dose, however the stimulated uptake rate was in the physiological range. Other studies have reported a much more pronounced stimulation of glucose uptake with insulin. Ruderman et al. (26) found the rate of glucose uptake by the perfused hindlimb preparation 0 0
2.0 r 2.4 -
L
1 2
1 4
1 1 6 8 ( Glucose
1 1 10 12 1 mM
’ 14
1 16
FIG. 4. Effect of perfusate glucose concentration on rate of glucose uptake by perfused rat hindlimb. Data shown are from perfusion periods S-and 6 only since a steady-state rate of glucose uptake had not been attained until this time (Fig. 2). .14 1
0
(r=.475 1
’
.4
FIG. 5. Correlation novo derived alanine 5 and 6.
’
’
.8 1.2 Glucose yMole/min
0
(r=.078 2
FIG.
of total 6.
6. Lack alanine
,
1 .4
of correlation release. Data
. 1 1 1 J 1.2 1.6 2.0 24 28 Glucose Uptake Rate gMole/min/4.5g Dry Leg
between glucose-uptake shown are from perfusion
&&;:
$2
n.s.1
1 .8
rate and rate periods 5 and
f-@
2
.0 I
1 1
*
2 30 Minute
1
1
3 4 Perfusion
1
‘I
5 6 Period
FIG. 7. Effect of insulin on rate of glucose uptake by perfl xsed rat hindlimb. Plasma glucose concentrations was 8.3 mM. I No insulin in perfusate, means * SE (N = 9), 3 700 PUlml porcine insulin included in perfusate, means t SE (N = 5).
increased 3.3 to 14-fold with the inclusion of 12.5 mu/ml insulin (a substantially higher concentration than used in the present investigation). Strohfeldt et al. (29) reported that insulin (1.0 mu/ml) increased the glucoseuptake rate of their hindlimb preparation 3 times the basal level. The insulin-stimulated glucose-uptake rate obtained in the present investigation seems quite low compared to the stimulated uptake rates reported for other perfused limb preparations. The reason for this is unknown, however; the level of insulin in the perfusate (700 pU/ml) was determined by radioimmunoassay and the biological activitv may have been significantly less than the immunoreactivity because the insulin was stored in a frozen solution rather than being prepared immediately prior to use. The rate of total alanine release (Fig. 8) was depressed by insulin (F = 6.38, P < 0.05), whereas the rate of de novo alanine release was significantly enhanced (F = 33.64, P < 0.01). A significantly greater percentage of the total alanine release was derived from glucose in the presence of insulin (51.5%, period 5-6) in contrast with 33% derived de novo in the absence of insulin (F = 54.7, P < 0.01). While only a small fraction of the glucose-uptake rate was converted to alanine in the presence of insulin (3.7%, periods 5-G) this was significantly greater than the 2.7% glucose-alanine conversion in the absence of the hormone (F = 93.00, P < 0.01).
PC.011 ’
’
’
1.6 2.0 2.4 Uptake Rate /4.5g Dry Leg
1
DISCUSSION
2.8
between glucose-uptake rate and rate of de release. Data shown are from perfusion periods
This study provides evidence that a large part of the alanine released from muscle is synthesized de novo by the transamination of glucose-derived pyruvate rather than produced by muscle proteolysis. Of the total ala-
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
DE
NOVO
ALANINE
SYNTHESIS .30r
Total
11
30
RAT
HINDLIMB
Alanine
1
1
BY
2 Minute
1.1
3 4 Perfusion
1
1
5 6 Period
8. Effect of insulin on rate of total and de nova derived alanine released by perfused rat hindlimb. No insulin; * total alanine release, + de no-vo alanine release, means 2 SE (N = 9). Insulin 700 pU/ml; $ total alanine release, ’ # de novo alanine release, means + SE (N = 5). FIG.
nine released by the perfused rat limb, 33% was found to be derived de novo from glucose during perfusion periods 5 and 6. By including [14C]glucose in the medium of an incubated diaphragm preparation (cold glucose concentration 5 mM), Odessey et al. (19) found that 29% of the alanine produced by the diaphragm was derived from glucose. This figure agrees very well with the quantity of alanine derived de novo (33%) by the perfused rat hindlimb. Insufficient information was reported in the study by Odessey et al. (19) to allow determination of the quantity of glucose taken up by the diaphragm that was actually converted to alanine. Based on indirect estimates, Felig and Wahren (7) estimated that as much as 13 and 18% of the glucose taken up by the human leg and forearm, respectively, may be converted to alanine, but the present study reveals that only 2.7% of the glucose taken up by the perfused rat hindlimb, after a steady-state rate of glucose uptake has been reached (periods 5 and 6)) is converted to alanine. Rat muscle protein contains about 6.4% alanine (16), yet Ruderman and Lund (27) found that alanine accounted for 22.5% of the total amino acids released by the perfused rat hindquarter. The present investigation revealed that 33% of the alanine released by the rat hindlimb was of de novo origin. Of the 22.5% amino acid release accounted for by alanine, approximately 33% would be alanine derived from glucose, suggesting that 15% of the total amino acid released by the rat hindquarter was alanine of endogenous or non-glucosederived origin. If the rate of alanine release reflects the muscle amino acid concentration, a question still exists as to the source of the non-glucose-derived alanine responsible for the disproportionately large production of endogenously derived alanine (15%) in relation to the quantity of alanine comprising muscle protein (6.4%). Since muscle possesses the enzymes necessary to produce pyruvate from amino acids such as glutamine and aspartate (20), muscle may synthesize alanine from non-glucose-derived pyruvate. Furthermore, pyruvate may be supplied by the breakdown of muscle glycogen (present in muscle before perfusion ); consequently, the
1383 alanine derived therefrom would not be labeled. To what extent these pathways provide carbon for alanine synthesis is not known. The present investigation demonstrates that as the glucose uptake rate by muscle increases, the rate of de novo alanine release likewise increases. As the rate of glucose uptake by muscle increases, the additional glucose is phosphorylated and can then either be converted to pyruvate through glycolysis or converted to glycogen. Since pyruvate supplies the carbon precursor for the de novo alanine, it would be expected that the rate of alanine synthesis and release would be correlated to the glucose-uptake rate. If production of glycogen is enhanced under these conditions, this glycogen would be labeled and thus the alanine produced from pyruvate derived from glycogenolysis would be recovered as labeled (de nova) alanine. While the rate of de novo derived alanine was increased as a result of the enhanced glucose-uptake rate, no change in total alanine release was observed. The constant rate of total alanine release implies that the rate of alanine output derived from endogenous sources was decreasing while de novo derived alanine was increasing. In this respect, glucose has been shown to decrease the rate of proteolysis in rat skeletal muscle (3, 8, 9) and, therefore, a reduction in the release of alanine derived from protein degradation would be expected: Other investigators have reported that alanine release was not influenced by the addition of glucose or pyruvate to the incubation medium (21), and it was concluded that the alanine released by the muscle was derived from sources other than pyruvate. The present investigation provides data indicating that the output of de novo derived alanine can increase while total alanine release does not change. Insulin administration reduces the release of many amino acids by muscle, but alanine release is not affected significantly (22). Moreover, a rise in the alanine concentration has been noted following insulin administration (2). It has been shown that insulin enhances amino acid transport (15) and protein synthesis (31), while independently decreasing protein degradation (8). The net effect of these processes would be reflected in a decrease in amino acid release by the tissue. The anomalous behavior of alanine in response to insulin administration is thought to be due to the insulin-stimulated glucose transport and utilization, resulting in an increased availability of pyruvate for de novo alanine synthesis (4). In contrast to other reports, the present study has shown that insulin inhibits the rate of total alanine release (Fig. 8). Despite this small, but statistically significant, reduction in total alanine release, the rate of de novo derived alanine release was significantly enhanced by insulin; slightly more than 50% of the alanine released by muscle was derived from glucose carbons. The diminished tota! alanine release may reflect the difference between insulin-stimulated de novo alanine synthesis (enhancing alanine release) and insulin-stimulated amino acid transport (recapture), protein synthesis, and decreased proteolysis, the latter three
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
1384 actions of the hormone serving to diminish alanine release by muscle. A significantly greater percentage of the glucose taken up by the limb was converted to alanine in the presence of insulin. This increased glucose-alanine transformation cannot be due simply to an increased pyruvate availability secondary to the insulin-stimulated glucose transport, because increasing the glucoseuptake rate by increasing the glucose concentration did not enhance the percentage of the glucose that was converted to alanine despite the fact that de novo alanine production was increased. Direct hormonal stimulation of glutamate-pyruvate transaminase, as occurs with a hepatic transaminase (30), might account for the increased de novo synthesis of alanine, but since the transaminase in muscle is in such excess, it is unlikely that an increase in its activity could increase de novo synthesis of alanine. In summary, the alanine produced from glucose is a
B. GRUBB
sizeable portion of the total alanine released by muscle; nevertheless, a compa ratively smal 1 fraction of the glucose carbons are actually transformed to alanine. The rate of de novo alanine release may be accelerated bY enhancing the glucose-uptake rate, either by increasi -ng the glucose concentration perfusing the muscle or by adding insulin to the perfusate. The release of endogenously derived alanine appears to be inhibited by either increasing the glucose-uptake rate or by adding insulin to the perfusate. The author is grateful to Drs. B. A. Schottelius, W. T. Stauber, and J. F. Snarr for the helpful suggestions made in preparing this manuscript. The author thanks Dr. Hosley, Eli Lilly Co., for kindly supplying the’porcine insulin used in this investigation. Present address of author: Dept. of Zoology, Duke University, Durham, N. C. 27706. Received
for publication
18 August
1975.
REFERENCES 1. BARTLET, R., B. NOYES, AND P. DRINKER. A simple reliable membrane oxygenator for organ perfusion. J. Appl. Physiol. 29: 758-759, 1970. 2. CARSTEN, A., B. HALLGRAM, R. JOGENBURG, A. SVANBORG, AND L. WERKO. Amino acid and free fatty acid in plasma in diabetes. Acta Med. Stand. 179: 361-370, 1966. 3. CASSENS, R. G. MuscZe BioZogy. New York: Marcel Dekker, 1971, vol. 1, p. 111-118. 4. FELIG, P. The glucose-alanine cycle. MetaboZism 22: 179-204, 1973. 5. FELIG, P., E. MARLISS, 0. OWEN, AND G. CAHILL. Role of substrate in the regulation of hepatic gluconeogenesis in fasting man. Advan. EnzymoZ. Regulation 7: 41-46, 1968. 6. FELIG, P., AND T. POZEFSKY. Alanine: key role in gluconeogenesis. Science 167: 1003-1004, 1969. 7. FELIG, P., AND J. WAHREN. Amino acid metabolism in exercising man. J. CZin. Invest. 50: 2703-2714, 1971. 8. FULKS, R. M., J. LI, AND A. GOLDBERG. Effects of insulin, glucose, and amino acids on protein turnover in rat diaphragm. J. BioZ. Chem. 250: 290-298, 1975. 9. GOLDBERG, A., E. HOWELL, J. LI, S. MARTEL, AND W. PROUTY. Physiological significance of protein degradation in animal and bacterial cells. Federation Proc. 33: 1112-1120, 1974. 10. HOFFMAN, W. A rapid photoelectric method for the determination of glucose in blood and urine. J. BioZ. Chem. 120: 51-55, 1937. 11. HONIG, C. R., J. L. FRIERSON, AND C. N. NELSON. 0, transport and Va, in resting muscle: significance for tissue-capillary exchange. Am. J. Physiol. 220: 357-363, 1971. 12. Jansky, L., AND J. S. HART. Participation of skeletal muscle and kidney during nonshivering thermogenesis in cold acclimated rats. Can. J. Biochem. PhysioZ. 41: 953-964, 1963. 13. JEFFERSON, L., D. RANNELS, B. MUNGER, AND H. MORGAN. Insulin in the regulation of protein turnover in heart and skeletal muscle. Federation Proc. 33: 1098-1104, 1974. 14. KARL, I., A. PAGLIARA, AND D. KIPNIS. A rapid microfluorometric enzymatic assay for the determination of alanine and pyruvate in plasma and tissues. J. Lab. CZin. Med. 80: 434-441, 1972. 15. KIPNIS, D. M., AND M. W. NOALL. Stimulation of amino acid transport by insulin in the isolated rat diaphragm. Biochem. Biophys. Acta 28: 226-227, 1958. 16. KOMINZ, D., A. HOUGH, P. SYMONDS, AND K. LAKI. The amino acid composition of action, myosin, tropomyosin and meromyosin. Arch. Biochem. Biophys. 50: 148-159, 1954. 17. MALLETTE, L., J. EXTON, AND C. PARK. Control of gluconeogene-
18.
19.
20.
21.
22.
23. I 24. 25. 26.
27. 28. 29.
30.
31.
32.
sis from amino acids in perfused rat liver. J. BioZ. Chem. 244: 5713-5723, 1969. MANCHESTER, K., AND F. YOUNG. Location of 14C protein from isolated rat diaphragm incubated in vitro with 14C amino acids and with 14C0,. Biochem. J. 72: 136-141, 1959. ODESSEY, R., E. A. KHAIRALLAH, AND A. L. GOLDBERG. Origin and possible significance of alanine production by skeletal muscle. J. BioZ. Chem. 249: 7623-7629, 1974. OPIE, L., AND E. NEWSHOLME. Activity of fructose 1,6-diphosphatase-phosphofructokinase and phosphoenolpyruvate carboxykinase in red and white muscle. Biochem. J. 103: 391-399, 1967. OZAND, R., J. TILDON, R. WAPNIR, AND M. CORNBLATH. Alanine formation by rat muscle homogenate. Biochem. Biophys. Res. Commun. 53: 251-257, 1973. POZEFSKY, T., P. FELIG, J. TOBIN, J. SOELDNER, AND G. CAHILL. Amino acid balance across tissue of the forearm in postabsorptive man. Effects of insulin at two dose levels. J. CZin. Invest. 48: 2273-2282, 1969. POZEFSKY, T., AND R. G. TANCREDI. Effects of intrabrachial arterial infusion of pyruvate on forearm tissue metabolism. J. CZin. Invest. 51: 2359-2369, 1972. ROSS, E. Regional blood flow in the rat. J. AppZ. Physiol. 21: 1273-1275, 1966. RUDERMAN, N., AND M. BERGER. The formation of glutamine and alanine in skeletal muscle. J. BioZ. Chem. 249: 5500-5506, 1974. RUDERMAN, N. B., C. R. S, HOUGHTON, AND R. HEMS. Evaluation of the isolated perfused rat hindquarter for the study of muscle metabolism. Biochem. J. 124: 639-651, 1971. RUDERMAN, N., AND P. LUND. Amino acid metabolism in skeletal muscle. IsrueZ J. Med. Sci. 8: 295-302, 1971. SCOTT, R. CZinicaZ AnaZysis by Thin Layer Chromatography Techniques. Ann Arbor, Mich.: Humphrey, 1969, p. 84-87. STROHFELDT, P., H. KETTL, AND K. F. WEINGES. Perfusion of the isolated rat hindlimb with a synthetic medium. Hormone Metab. Res. 6: 167-168, 1974. WICKS, W. D., C. A. BARNETT, AND J. B. MCKIBBIN. Interaction between hormones and cyclic AMP in regulating specific hepatic enzyme synthesis. Federation Proc. 33: 1105-1111, 1974. WOOL, I. G., AND P. CAVINCHI. Protein synthesis by skeletal muscle ribosomes: effect of diabetes and insulin. Biochemistry 6: 1231-1242,’ 1967. ZIVIN, J., AND J. SNARR. A stable preparation for rat brain perfusion: effect of flow rate on glucose uptake. J. AppZ. PhysioZ. 32: 658-663, 1972.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
De novo synthesis of alanine perfused rat hindlimb BARBARA Department
GRUBB of Physiology,
Northwestern
University
GRUBB, BARBARA. De novo synthesis of alanine by the perfked rat hindlimb. Am. J. Physiol. 230(5): 1379-1384. 1976. Due to the disproportionately large production of alanine by muscle, it has been suggested that part of the alanine released by muscle is synthesized de novo by the transamination of glucose-derived pyruvate. This glucose-alanine conversion was quantitated in the isolated rat hindlimb perfused with a solution of bicarbonate buffer containing 2% albumin, 2.4% dextran, 2.5-15.9 MM glucose, 32-34% dog erythrocytes, and 0.05 PCilml [‘4C]glucose. Measurement of labeled alanine production allowed quantitation of de novo alanine synthesis. De novo derived alanine accounted for an average of 33% of the total alanine released by the perfused tissue (perfusate glucose concentration 8.3 mM>, concurrently 2.7% of the glucose taken up by the limb was converted to alanine. By increasing the glucose concentration perfusing the muscle, both the rate of glucose uptake and de novo alanine release were increased. Addition of insulin to the perfusate (700 pU/ml) resulted in a significant increase in the rate of glucose uptake and de novo alanine production, but the rate of total alanine release was significantly decreased by the hormone. It was concluded that de novo alanine production accounts for a sizeable portion of the total alanine released by muscle, nevertheless a comparatively small fraction of the glucose carbons are actually transformed to alanine.
dl
case; 1’4C]glucose;
insulin;
quantitation
skeletal muscle, alanine is released larger quantities than most other amino acids ‘(5, 6, substrate for hepatic > and serves as an important gluconeogenesis (17). While alanine comprises less than 10% of muscle protein (16), ZO-30% of the total amino acid output by the perfused rat hindlimb (27) and the human forearm in vivo (6, 22) is accounted for by alanine. It has been suggested that the catabolism of a specific polyalanyl protein could account for this disproportionately large production of alanine, yet such a protein has not been identified (4). Also, during starvation when muscle proteins are being broken down, alanine continues to be released in disproportionately large quantities (6). De novo synthesis of alanine by the transamination of glucose-derived pyruvate could best account for the large production of alanine (6, 17). Indirect estimates have been made as to the quantity of glucose taken up by muscle that may be converted to alanine. Felig and Wahren (7) estimated that 13 and 18% of the glucose taken up by the human leg and foreIN RAT AND HUMAN .
by the
Medical
School, Chicago, Illinois
60611
arm, respectively, may be converted to alanine. Pozefsky and Tancredi (23) found arterial pyruvate infusions resulted in up to a 6.7% increase in the release of alanine by the human forearm. However, it was not possible to determine how much of the alanine released by the arm prior to the pyruvate infusion was actually derived de novo. The purpose of this investigation was to determine the quantity of glucose taken up by muscle that can be converted to alanine and what fraction of the total alanine released was synthesized de novo. Changes in blood glucose concentration and the effect,s of insulin on the glucose-alanine conversion were examined by supplying labeled glucose to the perfused rat hindlimb and quantitating the glucose-alanine conversion. MATERIALS
AND METHODS
Surgical preparation. Male rats, 275-300 g, were allowed free access to food and water until surgery. The rats were anesthetized with sodium pentobarbital (6 mg/lOO g). A transverse abdominal incision was made through the skin, superficial and inferior epigastric vessels were ligated, an incision was made through the abdominal muscle layer, and all pelvic viscera were removed. The caudal artery as well as three pairs of lumbar vessels arising dorsal to the aorta were ligated. The skin was detached from the limb, but left loosely wrapped around the limb to prevent heat loss and desiccation. Only one limb was to be perfused; therefore, the blood supply to the contralateral limb was ligated. A tight ligature also was placed around the ankle joint so that the foot was not perfused. After all vessels not supplying the limb had been ligated, the rat was heparinized (100 U/l00 g) and a polyethylene cannula (PE190) filled with perfusate was placed in the aorta. Then, a larger polyethylene cannula (PE-240) was secured in the vena cava. On completion of cannulation, the rat was sacrificed with an intracardiac injection of sodium pentobarbital and placed in the environmental chamber P’erfusion medium. The perfusion medium consisted of Krebs-Ringer solution, bicarbonate buffer, bovine albumin (fraction V, Pentex-untreated to remove free fatty acids), dextran (mol wt 60,000-80,000), glucose, [14C]glucose, and washed dog erythrocytes. This artificial blood had a hematocrit of 32-34%, 2% albumin, 2.4% dextran, 2.5-15.9 mM glucose, and 0.05 p Ci/ml [14C]glucose. Porcine insulin, 700 pU/ml, was included
1379 Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
1380
B.
in the perfusate of some preparations. Details of perfusate preparation can be found elsewhere (32). Perfusion system. The perfusate was pumped from the arterial reservoir to the oxygenator by a variablespeed, two-channel, microinfusion roller pump (Holter RL 175) (Fig. 1). The limbs were perfused at a constant flow rate (1.40 ml/min per leg), which is in the in vivo range for blood flow through the femoral artery of the rat (24). After leaving the oxygenator, the perfusate entered the environmental chamber and passed through a glass filter packed with Dacron. A few centimeters beyond the filter, a bubble trap was inserted into the line, the side arm of which connected to a Statham pressure transducer, this, in turn, was connected to a Beckman RB Dynograph recorder. After passing through the bubble trap, the perfusate was conducted through the arterial cannula into the rat’s hindlimb. Perfusion effluent flowed from the vena cava into the venous cannula, and thence into the venous reservoir. The latter consisted of a large plastic test tube connected to a soda lime trap to prevent escape of labeled CO, (Fig. 1). The venous reservoir was placed in a beaker of ice to minimize glycolysis by the erythrocytes. The perfusate was not recirculated. AnaZysis of perfusate. The entire venous perfusate was collected for six successive 30-min periods. After collection, the venous samples were stored on ice, each of these 30-min samples was centrifuged and analyzed at the end of each experiment. Several arterial samples taken at varying times during the experiment were treated in a manner identical to that of the venous samples. Plasma alanine concentration was determined
a magnetrc
Stirrer
FIG. 1. Diagram of perfusion system. A plastic beaker in ice constituted arterial perfusate reservoir. By keeping perfusate on ice, glycolysis and bacterial growth were minimized. Perfusate was stirred continuously by a magnetic stirrer. Gentle mixing by Tefloncoated stirring rod caused less hemolysis than occurred in another type of reservoir system previously employed (32). Oxygenator consisted of 12 ft of coiled Silastic tubing (0.058 inch I’D, 0.077 inch OD, Dow Corning Corp.) enclosed in a l-gallon jar (1) into which a gas mixture of 95% O,-5% CO, was pumped. A thermoregulator placed under skin of perfused limb was set to maintain a temperature of 37°C. Since environmental chamber also was maintained at this temperature, perfusate entered hindlimb at 37°C. A CO, trap containing soda lime was placed in the venous reservoir to prevent escape of labeled CO,.
GRUBB
by a microfluorometric enzymatic assay (14). Plasma glucose concentration was determined on a Technicon AutoAnalyzer by the standard ferricyanide reduction method (10). There was no drop in the glucose concentration of the perfusate during its passage through the perfusion system, indicating no significant glucose utilization by the erythrocytes. Also, there was no measurable glucose uptake by the erythrocytes in the venous samples collected and stored on ice. The diaphragm, when supplied with labeled glucose, is able to synthesize labeled glutamine, glutamate, and aspartate in addition to alanine (18). Other types of skeletal muscles may also produce these labeled amino acids. In the present experiments, labeled alanine was separated from the other labeled products and labeled glucose by use of ion exchange resins. To remove glutamate, aspartate, lactate, and pyruvate, the deprotein ized sample, pH 7.0, was applied to an anion exchange column (1 g Dowex-1, acetate form). Because glutamine will pass through the anionic column at pH 7.0, glutamine was converted to glutamate by acidifying and boiling samples for 1 h before applying the samples to the resin. The samples were allowed to cool and brought to pH 7 before applying to the column. Neutral amino acids as well as glucose pass through this column. Deionized distilled water was used to wash the anionic column. This water effiuent containing labeled glucose and alanine was adjusted to pH 2.0 and applied to a cationic exchange resin (1 g Dowex 50, sodium form). The positively charged alanine adhered to the resin, the glucose was washed through the resin. Alanine was eluted from the column with 4 N HCl and subsequently counted on a scintillation counter. In order to provide evidence that only labeled alanine was present in the eluate from the cationic exchange column, three experiments were performed in which venous samples were collected from the perfused limbs. Tritiated alanine was added to the samples before deproteinization, thus they presumably contained only [“Hlalanine and [14C]alanine after having run through both columns. A small volume of the eluate from the cationic column was applied to a thin-layer plate (Eastman) for two-dimensional resolution by electrophoresis (300 V, 10 mA for 30 min, buffer pH 1.85, acetic acid: formic acid:H,O, 78:25:877, vol/vol) followed by conventional ascending thin-layer chromatography (n-butyl alcohol:acetic acid:H,O, 3:l:l) (28). The plate was dried and scanned for radioactivity by an Actigraph III Nuclear Chicago Radiochromatography system. In each experiment, the plate exhibited only one peak which contained the C3H]alanine added to the sample as well as a 14C compound. Because the [3H]alanine peak coincided with the 14C peak it was concluded that the 14C compound was alanine. If other labeled compounds were present, they occurred in such small quantities that they represented background counts. Since recovery of [14C]alanine from the columns was never lOO%, but varied between 90-95% under optimal conditions, the variation in alanine recovery was monitored by adding a known quantity of [3H]alanine to each sample prior to deproteinization. The percent recovery
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
DE
NOVO
ALANINE
SYNTHESIS
BY
RAT
1381
HINDLIMB
of [3H]alanine was then used to correct the [14C]alanine counts in all samples. Presentation of results. Rates of glucose uptake and alanine release were determined by multiplying the arteriovenous concentration difference of these two substances by the perfusion flow rate. The percent glucose that was converted to alanine was determined as follows: 1) Counts per minute (cpm) [14C]glucose uptake = cpm arterial [14C]glucose/ml x % cold glucose uptake. The percent cold glucose taken up was determined by dividing the arteriovenous glucose concentration difference by the arterial glucose concentration and multiplying by 100. 2) Percent glucose uptake converted to alanine = cpm [14C]alanine/ml cpm [14C]glucose uptake/ml
x 100
3) Percent of total alanine release derived de novo = glucose uptake rate X % glucose converted to alanine Total alanine release (per min)
x2
(The numerator is multiplied by 2 since 1 mol of glucose yields 2 mol of alanine.) All results are expressed per 4.5 g dry leg, the average dry weight of the perfused tissue (73.3% H20, 17 g wet wt), because the results per unit dry weight were slightly less variable than when expressed as a function of wet weight.
function of perfusion time (Fig. 2), but approached a steady state after perfusion period 4. The glucose uptake rates measured in this study are within the range of those reported for other perfused rat hindlimb preparations. Ruderman et al. (26) reported that the glucose uptake rate of their perfused hindlimb was 0.5-1.7 pmollmin per 30 g (perfusate glucose concentration 5.5 mM). If expressed in relation to dry tissue weight, as used in this paper, these values would be approximately 0.28-0.96 pmol/min per 4.5 g. In another study the perfused rat hindlimb preparation was found to have a higher glucose uptake rate, 3.8 pmol/min per 30 g (2.15 pmol/min per 4.5 g), but the glucose concentration in the perfusate was 10 mM (29). Figure 3 shows the continual release of alanine throughout the entire 3-h perfusion period. The rate of total alanine release (de novo plus endogenously derived alanine from muscle proteolysis) did not appear to be systematically related to perfusion time. The rate of de novo alanine release (labeled alanine) increased as a function of perfusion time, but the rate of release approached a steady state by perfusion periods 5-6. De novo derived alanine accounted for an average of 33% of the total alanine released by the perfused tissue (periods 5-6); concurrently, 2.7% of the glucose taken up by the limb was converted to alanine. Data are given for periods 5 and 6 only, since a steady-state rate of glucose uptake had not been attained until this time. Effect of glucose-uptake rate on glucose-alanine conversion. Glucose concentrations were varied in the perfusate and the effect of glucose uptake on the rate of alanine release was determined. Within the range of
RESULTS
Viability of preparation. The preparation employed in this investigation remained viable throughout the perfusion as judged by the rate of oxygen consumption and lactate production. The rate of oxygen consumption by this preparation was 4.90 t .36 pmol/min per 4.5 g dry leg, and a steady state of oxygen consumption was maintained for the 3-h perfusion period (unpublished observations). This rate of oxygen consumption agrees well with estimates of oxygen consumption by the rat leg in vivo, 0.05 ml/min per leg (12) (2.38 pmol/min per 4.5 g); the perfused rat hindlimb, 10.8 pmol/min per 30 g (26) (6.12 pmol/min per 4.5 g); rat gracilis muscle in vivo, 0.95 ml/min per 100 g (11) (6.80 pmol/min per 4.5 g)* As part of another study, the rate of lactate production by the limb was measured and found to remain stable after the first 30 min of perfusion (unpublished observations). In a deteriorating preparation, a progressive decrease in oxygen consumption and an increase in lactate production would be expected. Thus, according to these criteria this preparation remained viable throughout the experimental period. Glucose uptake rate. Since preliminary experiments showed the rate of glucose uptake was quite variable during the first 30 rnin of perfusion, no chemical analyses were performed on this venous sample. All of the limbs took up a significant quantity of glucose. The rate of glucose uptake increased slowly as a
.
’
’
1
’
2 30 Minute
2. Effect of perfusion time hindlimb. “Plasma” glucose
on glucose concentration
FIG.
rat
’
’
3 4 Perfuston
J
5 6 Period
uptake rate by perfused was 8.3 mM. Data
shown are meanS * SE (N = ‘)* .30
UI-25: 3
9-G Tot al Alanine
i!.-E
5
;
De Novo
123456 30 Minute FIG.
alanine
3. Effect release.
Alanine
.lO
Perfusion
of perfusion time on rate Plasma glucose concentration
shown are means k SE (N = 9).
Release
Release t
T
Period
of total and de novo was 8.3 mM. Data
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
1382
B. GRUBB
glucose concentrations studied, the rate of glucose uptake during periods 5 and 6 was found to be linearly related to plasma glucose concentration (Fig. 4). The glucose-uptake rates shown in Fig. 4 are plotted against their respective rates of alanine release (Figs. 5 and 6). The rate of de novo alanine release- (Fig. 5) was correlated to the rate of glucose uptake. As illustrated in Fig. 6, no correlation existed between the glucose-uptake rate and the rate of total alanine release; nor was there a correlation between the glucose-uptake rate and the percent glucose that was converted to alanine. On the average, 2.7% of the glucose taken up by the limb was converted to alanine. Effect of insulin on glucose-alanine conversion. In some experiments, porcine insulin was included in the perfusate to determine the effect of this hormone on alanine release. An analysis of variance indicated that the rate of glucose uptake by the limbs was significantly enhanced by insulin (F = 21.97, P c 0.01) (Fig. 7). In this preparation, inclusion of 0.7 mu/ml porcine insulin enhanced the glucose uptake rate 22% (average of all periods). The level of insulin used was a pharmacological dose, however the stimulated uptake rate was in the physiological range. Other studies have reported a much more pronounced stimulation of glucose uptake with insulin. Ruderman et al. (26) found the rate of glucose uptake by the perfused hindlimb preparation 0 0
2.0 r 2.4 -
L
1 2
1 4
1 1 6 8 ( Glucose
1 1 10 12 1 mM
’ 14
1 16
FIG. 4. Effect of perfusate glucose concentration on rate of glucose uptake by perfused rat hindlimb. Data shown are from perfusion periods S-and 6 only since a steady-state rate of glucose uptake had not been attained until this time (Fig. 2). .14 1
0
(r=.475 1
’
.4
FIG. 5. Correlation novo derived alanine 5 and 6.
’
’
.8 1.2 Glucose yMole/min
0
(r=.078 2
FIG.
of total 6.
6. Lack alanine
,
1 .4
of correlation release. Data
. 1 1 1 J 1.2 1.6 2.0 24 28 Glucose Uptake Rate gMole/min/4.5g Dry Leg
between glucose-uptake shown are from perfusion
&&;:
$2
n.s.1
1 .8
rate and rate periods 5 and
f-@
2
.0 I
1 1
*
2 30 Minute
1
1
3 4 Perfusion
1
‘I
5 6 Period
FIG. 7. Effect of insulin on rate of glucose uptake by perfl xsed rat hindlimb. Plasma glucose concentrations was 8.3 mM. I No insulin in perfusate, means * SE (N = 9), 3 700 PUlml porcine insulin included in perfusate, means t SE (N = 5).
increased 3.3 to 14-fold with the inclusion of 12.5 mu/ml insulin (a substantially higher concentration than used in the present investigation). Strohfeldt et al. (29) reported that insulin (1.0 mu/ml) increased the glucoseuptake rate of their hindlimb preparation 3 times the basal level. The insulin-stimulated glucose-uptake rate obtained in the present investigation seems quite low compared to the stimulated uptake rates reported for other perfused limb preparations. The reason for this is unknown, however; the level of insulin in the perfusate (700 pU/ml) was determined by radioimmunoassay and the biological activitv may have been significantly less than the immunoreactivity because the insulin was stored in a frozen solution rather than being prepared immediately prior to use. The rate of total alanine release (Fig. 8) was depressed by insulin (F = 6.38, P < 0.05), whereas the rate of de novo alanine release was significantly enhanced (F = 33.64, P < 0.01). A significantly greater percentage of the total alanine release was derived from glucose in the presence of insulin (51.5%, period 5-6) in contrast with 33% derived de novo in the absence of insulin (F = 54.7, P < 0.01). While only a small fraction of the glucose-uptake rate was converted to alanine in the presence of insulin (3.7%, periods 5-G) this was significantly greater than the 2.7% glucose-alanine conversion in the absence of the hormone (F = 93.00, P < 0.01).
PC.011 ’
’
’
1.6 2.0 2.4 Uptake Rate /4.5g Dry Leg
1
DISCUSSION
2.8
between glucose-uptake rate and rate of de release. Data shown are from perfusion periods
This study provides evidence that a large part of the alanine released from muscle is synthesized de novo by the transamination of glucose-derived pyruvate rather than produced by muscle proteolysis. Of the total ala-
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
DE
NOVO
ALANINE
SYNTHESIS .30r
Total
11
30
RAT
HINDLIMB
Alanine
1
1
BY
2 Minute
1.1
3 4 Perfusion
1
1
5 6 Period
8. Effect of insulin on rate of total and de nova derived alanine released by perfused rat hindlimb. No insulin; * total alanine release, + de no-vo alanine release, means 2 SE (N = 9). Insulin 700 pU/ml; $ total alanine release, ’ # de novo alanine release, means + SE (N = 5). FIG.
nine released by the perfused rat limb, 33% was found to be derived de novo from glucose during perfusion periods 5 and 6. By including [14C]glucose in the medium of an incubated diaphragm preparation (cold glucose concentration 5 mM), Odessey et al. (19) found that 29% of the alanine produced by the diaphragm was derived from glucose. This figure agrees very well with the quantity of alanine derived de novo (33%) by the perfused rat hindlimb. Insufficient information was reported in the study by Odessey et al. (19) to allow determination of the quantity of glucose taken up by the diaphragm that was actually converted to alanine. Based on indirect estimates, Felig and Wahren (7) estimated that as much as 13 and 18% of the glucose taken up by the human leg and forearm, respectively, may be converted to alanine, but the present study reveals that only 2.7% of the glucose taken up by the perfused rat hindlimb, after a steady-state rate of glucose uptake has been reached (periods 5 and 6)) is converted to alanine. Rat muscle protein contains about 6.4% alanine (16), yet Ruderman and Lund (27) found that alanine accounted for 22.5% of the total amino acids released by the perfused rat hindquarter. The present investigation revealed that 33% of the alanine released by the rat hindlimb was of de novo origin. Of the 22.5% amino acid release accounted for by alanine, approximately 33% would be alanine derived from glucose, suggesting that 15% of the total amino acid released by the rat hindquarter was alanine of endogenous or non-glucosederived origin. If the rate of alanine release reflects the muscle amino acid concentration, a question still exists as to the source of the non-glucose-derived alanine responsible for the disproportionately large production of endogenously derived alanine (15%) in relation to the quantity of alanine comprising muscle protein (6.4%). Since muscle possesses the enzymes necessary to produce pyruvate from amino acids such as glutamine and aspartate (20), muscle may synthesize alanine from non-glucose-derived pyruvate. Furthermore, pyruvate may be supplied by the breakdown of muscle glycogen (present in muscle before perfusion ); consequently, the
1383 alanine derived therefrom would not be labeled. To what extent these pathways provide carbon for alanine synthesis is not known. The present investigation demonstrates that as the glucose uptake rate by muscle increases, the rate of de novo alanine release likewise increases. As the rate of glucose uptake by muscle increases, the additional glucose is phosphorylated and can then either be converted to pyruvate through glycolysis or converted to glycogen. Since pyruvate supplies the carbon precursor for the de novo alanine, it would be expected that the rate of alanine synthesis and release would be correlated to the glucose-uptake rate. If production of glycogen is enhanced under these conditions, this glycogen would be labeled and thus the alanine produced from pyruvate derived from glycogenolysis would be recovered as labeled (de nova) alanine. While the rate of de novo derived alanine was increased as a result of the enhanced glucose-uptake rate, no change in total alanine release was observed. The constant rate of total alanine release implies that the rate of alanine output derived from endogenous sources was decreasing while de novo derived alanine was increasing. In this respect, glucose has been shown to decrease the rate of proteolysis in rat skeletal muscle (3, 8, 9) and, therefore, a reduction in the release of alanine derived from protein degradation would be expected: Other investigators have reported that alanine release was not influenced by the addition of glucose or pyruvate to the incubation medium (21), and it was concluded that the alanine released by the muscle was derived from sources other than pyruvate. The present investigation provides data indicating that the output of de novo derived alanine can increase while total alanine release does not change. Insulin administration reduces the release of many amino acids by muscle, but alanine release is not affected significantly (22). Moreover, a rise in the alanine concentration has been noted following insulin administration (2). It has been shown that insulin enhances amino acid transport (15) and protein synthesis (31), while independently decreasing protein degradation (8). The net effect of these processes would be reflected in a decrease in amino acid release by the tissue. The anomalous behavior of alanine in response to insulin administration is thought to be due to the insulin-stimulated glucose transport and utilization, resulting in an increased availability of pyruvate for de novo alanine synthesis (4). In contrast to other reports, the present study has shown that insulin inhibits the rate of total alanine release (Fig. 8). Despite this small, but statistically significant, reduction in total alanine release, the rate of de novo derived alanine release was significantly enhanced by insulin; slightly more than 50% of the alanine released by muscle was derived from glucose carbons. The diminished tota! alanine release may reflect the difference between insulin-stimulated de novo alanine synthesis (enhancing alanine release) and insulin-stimulated amino acid transport (recapture), protein synthesis, and decreased proteolysis, the latter three
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
1384 actions of the hormone serving to diminish alanine release by muscle. A significantly greater percentage of the glucose taken up by the limb was converted to alanine in the presence of insulin. This increased glucose-alanine transformation cannot be due simply to an increased pyruvate availability secondary to the insulin-stimulated glucose transport, because increasing the glucoseuptake rate by increasing the glucose concentration did not enhance the percentage of the glucose that was converted to alanine despite the fact that de novo alanine production was increased. Direct hormonal stimulation of glutamate-pyruvate transaminase, as occurs with a hepatic transaminase (30), might account for the increased de novo synthesis of alanine, but since the transaminase in muscle is in such excess, it is unlikely that an increase in its activity could increase de novo synthesis of alanine. In summary, the alanine produced from glucose is a
B. GRUBB
sizeable portion of the total alanine released by muscle; nevertheless, a compa ratively smal 1 fraction of the glucose carbons are actually transformed to alanine. The rate of de novo alanine release may be accelerated bY enhancing the glucose-uptake rate, either by increasi -ng the glucose concentration perfusing the muscle or by adding insulin to the perfusate. The release of endogenously derived alanine appears to be inhibited by either increasing the glucose-uptake rate or by adding insulin to the perfusate. The author is grateful to Drs. B. A. Schottelius, W. T. Stauber, and J. F. Snarr for the helpful suggestions made in preparing this manuscript. The author thanks Dr. Hosley, Eli Lilly Co., for kindly supplying the’porcine insulin used in this investigation. Present address of author: Dept. of Zoology, Duke University, Durham, N. C. 27706. Received
for publication
18 August
1975.
REFERENCES 1. BARTLET, R., B. NOYES, AND P. DRINKER. A simple reliable membrane oxygenator for organ perfusion. J. Appl. Physiol. 29: 758-759, 1970. 2. CARSTEN, A., B. HALLGRAM, R. JOGENBURG, A. SVANBORG, AND L. WERKO. Amino acid and free fatty acid in plasma in diabetes. Acta Med. Stand. 179: 361-370, 1966. 3. CASSENS, R. G. MuscZe BioZogy. New York: Marcel Dekker, 1971, vol. 1, p. 111-118. 4. FELIG, P. The glucose-alanine cycle. MetaboZism 22: 179-204, 1973. 5. FELIG, P., E. MARLISS, 0. OWEN, AND G. CAHILL. Role of substrate in the regulation of hepatic gluconeogenesis in fasting man. Advan. EnzymoZ. Regulation 7: 41-46, 1968. 6. FELIG, P., AND T. POZEFSKY. Alanine: key role in gluconeogenesis. Science 167: 1003-1004, 1969. 7. FELIG, P., AND J. WAHREN. Amino acid metabolism in exercising man. J. CZin. Invest. 50: 2703-2714, 1971. 8. FULKS, R. M., J. LI, AND A. GOLDBERG. Effects of insulin, glucose, and amino acids on protein turnover in rat diaphragm. J. BioZ. Chem. 250: 290-298, 1975. 9. GOLDBERG, A., E. HOWELL, J. LI, S. MARTEL, AND W. PROUTY. Physiological significance of protein degradation in animal and bacterial cells. Federation Proc. 33: 1112-1120, 1974. 10. HOFFMAN, W. A rapid photoelectric method for the determination of glucose in blood and urine. J. BioZ. Chem. 120: 51-55, 1937. 11. HONIG, C. R., J. L. FRIERSON, AND C. N. NELSON. 0, transport and Va, in resting muscle: significance for tissue-capillary exchange. Am. J. Physiol. 220: 357-363, 1971. 12. Jansky, L., AND J. S. HART. Participation of skeletal muscle and kidney during nonshivering thermogenesis in cold acclimated rats. Can. J. Biochem. PhysioZ. 41: 953-964, 1963. 13. JEFFERSON, L., D. RANNELS, B. MUNGER, AND H. MORGAN. Insulin in the regulation of protein turnover in heart and skeletal muscle. Federation Proc. 33: 1098-1104, 1974. 14. KARL, I., A. PAGLIARA, AND D. KIPNIS. A rapid microfluorometric enzymatic assay for the determination of alanine and pyruvate in plasma and tissues. J. Lab. CZin. Med. 80: 434-441, 1972. 15. KIPNIS, D. M., AND M. W. NOALL. Stimulation of amino acid transport by insulin in the isolated rat diaphragm. Biochem. Biophys. Acta 28: 226-227, 1958. 16. KOMINZ, D., A. HOUGH, P. SYMONDS, AND K. LAKI. The amino acid composition of action, myosin, tropomyosin and meromyosin. Arch. Biochem. Biophys. 50: 148-159, 1954. 17. MALLETTE, L., J. EXTON, AND C. PARK. Control of gluconeogene-
18.
19.
20.
21.
22.
23. I 24. 25. 26.
27. 28. 29.
30.
31.
32.
sis from amino acids in perfused rat liver. J. BioZ. Chem. 244: 5713-5723, 1969. MANCHESTER, K., AND F. YOUNG. Location of 14C protein from isolated rat diaphragm incubated in vitro with 14C amino acids and with 14C0,. Biochem. J. 72: 136-141, 1959. ODESSEY, R., E. A. KHAIRALLAH, AND A. L. GOLDBERG. Origin and possible significance of alanine production by skeletal muscle. J. BioZ. Chem. 249: 7623-7629, 1974. OPIE, L., AND E. NEWSHOLME. Activity of fructose 1,6-diphosphatase-phosphofructokinase and phosphoenolpyruvate carboxykinase in red and white muscle. Biochem. J. 103: 391-399, 1967. OZAND, R., J. TILDON, R. WAPNIR, AND M. CORNBLATH. Alanine formation by rat muscle homogenate. Biochem. Biophys. Res. Commun. 53: 251-257, 1973. POZEFSKY, T., P. FELIG, J. TOBIN, J. SOELDNER, AND G. CAHILL. Amino acid balance across tissue of the forearm in postabsorptive man. Effects of insulin at two dose levels. J. CZin. Invest. 48: 2273-2282, 1969. POZEFSKY, T., AND R. G. TANCREDI. Effects of intrabrachial arterial infusion of pyruvate on forearm tissue metabolism. J. CZin. Invest. 51: 2359-2369, 1972. ROSS, E. Regional blood flow in the rat. J. AppZ. Physiol. 21: 1273-1275, 1966. RUDERMAN, N., AND M. BERGER. The formation of glutamine and alanine in skeletal muscle. J. BioZ. Chem. 249: 5500-5506, 1974. RUDERMAN, N. B., C. R. S, HOUGHTON, AND R. HEMS. Evaluation of the isolated perfused rat hindquarter for the study of muscle metabolism. Biochem. J. 124: 639-651, 1971. RUDERMAN, N., AND P. LUND. Amino acid metabolism in skeletal muscle. IsrueZ J. Med. Sci. 8: 295-302, 1971. SCOTT, R. CZinicaZ AnaZysis by Thin Layer Chromatography Techniques. Ann Arbor, Mich.: Humphrey, 1969, p. 84-87. STROHFELDT, P., H. KETTL, AND K. F. WEINGES. Perfusion of the isolated rat hindlimb with a synthetic medium. Hormone Metab. Res. 6: 167-168, 1974. WICKS, W. D., C. A. BARNETT, AND J. B. MCKIBBIN. Interaction between hormones and cyclic AMP in regulating specific hepatic enzyme synthesis. Federation Proc. 33: 1105-1111, 1974. WOOL, I. G., AND P. CAVINCHI. Protein synthesis by skeletal muscle ribosomes: effect of diabetes and insulin. Biochemistry 6: 1231-1242,’ 1967. ZIVIN, J., AND J. SNARR. A stable preparation for rat brain perfusion: effect of flow rate on glucose uptake. J. AppZ. PhysioZ. 32: 658-663, 1972.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 15, 2019.
