- Email: [email protected]
Effects of corticosteroid on the transport and metabolism of glutamine in rat skeletal muscle
376
Biochimica et Biophysica Acta, 1092 (1991) 376-383 © 1991 Elsevier Science Publishers B.V. 0167-4889/91/$03.50 ADONIS 016748899100166F
BBAMCR 12929
Effects of corticosteroid on the transport and metabolism of glutamine in rat skeletal muscle H a r i n d e r S. H u n d a l *, Philip Babij **, P e t e r M. T a y l o r , P e t e r W . Watt and M i c h a e l J. R e n n i e Deparlmen! of Anatomy & Ph.vsiolol{v. The University. Dut~dee(U.K.) (Received 7 September 1990)
Key words: Dexamethasone; Amino acid; Cell membrane; Trauma
Intramuscular glutamine falls with injury and disease in circumstances associated with increases in blood corticosteroids. We have investigated the effects of corticosteroid administration (0.44 m g / k g dexamethasone daily for 8 days, 200 g female rats) on intramuscular giutamine and Na +, muscle glutamine metabolism and sarcolemmal glutamine transport in the perfused hindlimb. After dexamethasone treatment intramuscular giutamine fell by 45% and Na + rose by Z~% (the respective muscle/plasma distribution ratios changed from 8.6 to 4.5 and 0.12 to 0.15); glutamine synthetase and giutaminase activities were unchanged at 475 + 75 and 60 + 19 nmol/g muscle per min. Glutamine output by the hindlimb of anaesthetized rats was increased from 31 to 85 nmol/g per min. Sarcolemmal glutamine transport was studied by paired-tracer dilution in the perfused hindlimb: the maximal capacity (Vmax) for glutamine transport into muscle (by Na+-glutamine symport) fell from 1058 4-310 to 395 4- 110 nmol/g muscle per min after dexamethasone treatment, accompanied by a decrease in the Km (from 8.1 + !.9 to 2.1 4. 0.4 mM glutamine). At physiological plasma glutamine concentration (0.75 mM) dexamethasone appeared to cause a proportional increase in sarcolemmal glutamine efflux over influx. Addition of dexamethasone (200 nM) to the perfusate of control rat hindlimbs caused acute changes in Vm, and K m Of glutamine transport similar to those resulting from &day dexamethasone treatment. The reduction in muscle glutamine concentration after dexamethasone treatment may be primarily due to a reduction in the driving force for intramuscular giutamine accumulation, i.e., in the Na + electrochemical gradient. The prolonged increase in muscle glutamine output after dexamethasone treatment (which occurs despite a reduction in the size of the intramuscular glutamine pool) appears to be due to a combination of (a) eceelerated sarcolemmal giutamine efflux and (b) increased intramuscular synthesis of glutamine.
Introduction It is well recognised that the metabolic response to injury and disease includes changes in muscle composition, irrespective of whether or not muscle was the initial site of trauma [1,2]. The longer term responses usually include loss of muscle mass, but even in the early stages of the acute response, e.g., to injury and trauma, when it is impossible to detect a loss of muscle
* Present Address: Department of Cell Biology, Hospital For Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G IX8. ** Present Address: Department of Physiology, University College London, London WC IE, U.K. Correspondence: M.J. Rennie, Department of Anatomy & Physiology, The University, Dundee DD1 4HN, U.K.
protein, there are marked losses of glutamine from the muscle free amino acid pool [3,4]. The ionic composition of sarcoplasm also changes with losses of K + (associated with depolarisation of the sarcolemma) and increases in Na + and Ca 2+ [5-8]. Furthermore, the acute metabolic responses to bodily injury are invariably associated with a rise in blood corticosteroid concentrations [9,10]. Although many of the changes in muscle have been related to changes in protein synthesis and breakdown, (e.g., see Ref. 2 for review) it seemed to us that the observed alterations in these were insufficiently rapid to account for the derangements in muscle amino acid balance observed in sick and injured patients. We therefore set out to see whether or not the kinetic properties of the Na+-dependent muscle glutamine transporter, which we have previously characterised in the perfused rat hindlimb [11], are modifiable, acutely and chroni-
377 cally, by corticosteroid (dexamethasone) administration. We have also investigated possible effects of dexamethasone on the intramuscular concentrations of glutamine and Na ÷ and on the maximal capacities of muscle glutamine synthetase and glutaminase. Materials and Methods
All chemicals were obtained from Sigma (Poole, U.K.) except bovine serum albumin (BSA) fraction V (Miles-Pentex, Kankakee, U.S.A) and sodium pentobarbitone (May and Baker, Dagenham, U.K). Radioactive tracers were purchased from Amersham International (Amersham, U.K) ([3H]glutamine at 6 T B q / m mol, [14C]mannitol at 20 GBq/mmol) or New England Nuclear (Boston, U.S.A) (46Sc-labelled microspheres). Female Wistar rats (Bantin & Kingman, Hull, U.K.) of approx. 200 g body weight were used. Rats had access to food and water until the night before study when food only was withdrawn. Food intake was monitored by providing animals with a fixed quantity of food every 12 h (20 g/rat) and weighing the food not eaten. 40 rats were divided into two groups: one group (15 rats) was treated with dexamethasone, the other group (25 rats) served as control. Dexamethasone (Sigma) was administered by suprascapular subcutaneous injection (0.44 m g / k g body weight as a fine suspension in 0.9% NaCl) every day for 8 days: control rats received saline. It was impractical to carry out all experimental measurements on each rat, therefore the experimental and control groups were sub-divided into groups of five rats. These groups were assigned for measurement of (a) muscle glutamine synthetase and glutaminase, muscle composition (e.g., amino acids, muscle N a + / K +, RNA, DNA and protein), blood flow and arteriovenous differences of amino acids in the hindlimb, (b) measurement of blood flow to individual muscles by the microsphere method, (c) glutamine transport kinetics in the perfused hindlimb, (d, e) control rats only, glutamine transport kinetics in the perfused hindlimb _+ 200 nM perfusate dexamethasone (5 rats each). Rats were anaesthetized with sodium pentobarbitone (60 mg/kg i.p.) prior to muscle sampling, determination of blood flow in vivo or preparation for hindlimb perfusion. For measurements of amino acids and Na + in muscle, soleus and gastrocnemius muscle of one leg were frozen rapidly using Wollenberger clamps cooled in liquid N 2. Approx. 0.37 MBq of [14C]mannitol, was injected 1 h before sampling of plasma and muscle for estimation of extracellular space. The gastrocnemius muscle of the opposite leg was removed rapidly, and used immediately for analysis of glutamine synthetase (rate of glutamine synthesis from 20 mM glutamate: pH 7.4) [12] and total glutaminase activity (rate of glutamate production from 20 mM glutamine: pH 8.0 _+ 10 mM
phosphate) [13]. To measure intramuscular Na ÷ the extraceilular space of single hindlimbs from five dexamethasone treated and five control rats were 'flushed' (via a femoral artery cannula) with 20 ml of ice-cold, isotonic sucrose solution. We found that this procedure washed out approx. 95% of calculated extracellular Na +, and a further flushing with 20 ml of sucrose solution yielded only 1.5% of Na + washed-out by the initial treatment. Muscle was then clamp-frozen. For estimation of muscle and plasma Na ÷, tissue was dried to constant weight at 90 ° C, powdered and dissolved in 1 M H N O 3 for flame photometry by standard methods [14]. Protein determination was by the Lowry technique [151. Whole hindquarter blood flow was measured using an ultrasonic flowmeter (Transonic Systems) which uses a wide-beam transit-time technique [16]. The ultrasound probe was placed around the aorta in the lower abdomen just above the iliac b,.'furcation so that flow to other tissues such as the kidneys was excluded. Flow became stable after 5 min and the mean of several readings over 15 min was taken as the hindquarter flow. Single hindlimb flow was obtained by d;viding the value by two. For measurement of blood flow to individual muscles, a microsphere method [171 was used utilising 15 #m 46Sc-labelled microspheres. Reference blood samples were collected over 15 s at 0.8 ml/min from the femoral artery. Femoral venous blood for arterio-venous difference measurements (0.1 ml, in duplicate) was taken from the leg opposite to that used for the blood-flow. Arterial blood (0.1 ml in duplicate) was sampled from the carotid artery or the abdominal aorta. In some experiments plasma was isolated by centrifugation (2000 X g for 20 min at 4°C). A neutralised perchloric acid 10% (w/v) extract of powdered muscle, blood or plasma was analyzed for amino acids by either an enzymatic method [lg] or by means of a Biotronik LCS000 amino acid analyzer. Muscle RNA and DNA were determined by a dual wavelength spectrophotometric method [19]. Glutamine transport kinetics were investigated using a paired-tracer isotope dilution technique [20] applied to the skinned perfused hindlimb [11.21]. Briefly, anaesthetized rats were prepared for single hindlimb perfusion by cannulation of the femoral artery and femoral vein. Hindlimbs were perfused with a cell free perfusate (Krebs-Henseleit bicarbonate buffer containing 6% BSA (w/v) and 5 mM glucose) which was pumped through the animals hindlimb, without recirculation, at a flow rate of 2 ml rain-i. Glutamine transport kinetics in hindlimbs of control and dexamethasone treated rats were determined under steady state conditions by the paired-tracer method (using [3H]glutamine as the transportable molecule and [14C]mannitol as the extracellular space marker). Perfusate glutamine was varied between 0.5 mM and 30 mM
378 (with at least 10 min allowed for equilibration at each glutamine concentration). In studies in which we investigated the acute effect of 200 nM dexamethasone, hindlimbs were pre-equilibrated with the hormone (for at least 30 min) and it was retained in the perfusate for the remainder of the perfusion period (up to 1.5 h). Calculation of fractional tracer efflux was determined as described by Yudilevich and Mann [20]. Values for VmaX and K m were calculated from the Hanes plot [22] (i.e., [S]/v t, vs. [S]; where [S] is perfusate glutamine concentration and ot is the unidirectional influx) obtained from least squares regression analysis using software written for the Apple lie microcomputer. Other experimental details were exactly as previously described [11]. Data are presented as the mean + 1 standard error (S.E.). Differences between means were tested for significance using Student's t-test and were considered significant if P < 0.05. Results
Body weight, food intake and muscle composition The initial body weight (mean + S.E.) of experimental animals was 200 _+4 g (n -- 40). After 8 days dexamethasone treatment body weight was reduced to 189 + 1 g (n = 15), although control rats gained weight to 212 + 2 g (n = 25)) over the same period. This represents an overall weight loss equivalent to 10.8~ of final control weight. Food intake in dexamethasone treated rats (14.1 :l= 0.28 g / d a y per rat) was significantly ( P < 0.05) lower than in controls (15.9 + 0.35 g / d a y per rat). In order to ensure that the treated rats were not lasing weight due to malnutrition we fed a group of age-andweight matched control rats 14 g food/day per rat for 8 days without any loss of body weight. This finding indicated that the weight loss experienced by the dexamethason~treated rats was unlikely to be due to under-
TABLE II
Intramuscular amino acids in gastrocnemius muscle of control and 8 day dexamethasone treated rats Amino acid
Control ( p m o l / g wet weight)
Dexamethasone (p.mol/g wet weight)
Glutamine Glutamate Alanine Aspartate Valine lsoleucine Leucine
4.55 +0.29 0.80 -+0.10 1.40 -+0.10 0.48 +0.14 0.13 +0.01 0.016 =l:0.002 0.046 + 0.006
2.51 +0.41 * * * !.43 +0.05 * * 1.37 _+0.04 0.44 -+0.04 0.15 -+0.01 0.027 _+0.002 * 0.066 -I-0.002 *
(4)
Values are mean + S.E. Asterisks denote values significantly different from control ( * P < 0.05, * * P < 0.01, * * * P < 0.001).
nutrition. Differences between final hindlimb-muscle weights of control and dexamethasone-treated rats were in proportion to the difference in body weight (gastrocnemius 1.11 vs. 0.98 g, soleus 78 vs. 70 mg, anterior tibialis 312 vs. 275 mg wet weight, respectively; Table I). There were no significant differences between either tissue hydration or extracellular space volume of control and treated rats (Table l); in both groups intracellular volume was calculated to be approx. 0.61 m l / g wet weight. We have therefore expressed all data with wet weight as denominator. The concentration of protein in gastrocnemius muscle was lower in dexamethasonetreated rats than in controls (by 18%) and sarcoplasmic Na + concentration was 257o higher in dexamethasonetreated rats than in controls; K + concentration remained unaltered (Table !). Gastrocnemius glutamine fell in dexamethasone treated rats (by 45~) whilst intramuscular glutamate increased by 3-fold (Table II). We observed no significant changes in the muscle concentrations of alanine or aspartate but the concentrations of branched chain amino acids (valine, isoleucine and leucine) were all elevated although only signifi-
TABLE I
Effects o[ &day dexamethasone treatment on the composition of rat gastrocnemius muscle Data are presented as the mean -+ S.E. (per g muscle wet weight). (Number of determinations in parentheses).
Muscle weight (g) Tissue hydration (g H 2 O / g ) Extracellularvolume(ml/g) Protein concentration (rag/g) RNA (m8/8) D N A (mB/g) RNA/DNA RNA/alkali soluble protein Sodium concentration ( p m o l / g ) Potassium concentration ( p m o l / g )
(10) (4) (4) (4) (4) (4)
Control
Dexamet hasone-treated
!,11 + 0.015 0.75 +0.01 0.16 +0.03 189 +7 1,3 -+0.09 0.383 + 0.025 3.4 +0.01 0.007+0.0002 10.0 _+0.7 92.4 _+6.2
0.98 -+ 0.02 * * * 0.75 +0.01 0.13 :l:O.0l 155 _+5 * * 1.71 +0.29 0.424 + 0.020 4.0 +0.5 0.01 +0.001 * * 12.5 +0.4 * 92.2 + 5.8
Asterisks denote values significantly different from control ( * P < 0.05, * * P < 0.01, * * * P < 0.001).
(I0) 0o) (5) (5) (4) (4) (4) (4) (5) (5)
379 TABLE Ill Sarcoplasmic concentrations of glutamine and sodium (raM) and sarcoplasm /plasma distribution ratio (DR), in gastrocnemius muscle of control and dexamethasone-treated rats
Values are means+_S.E. from at least five rats in each case. Glutamine Control Dexamethasone
Na
muscle
plasma
DR
muscle
plasma
DR
7.3 + 0.50 3.9+0.60 * *
0.85 + 0.04 0.86 +0.06
8.6 + 0.16 4.5 +0.39 * *
16 + 1 20 + 0.6 *
138+ 2 130+ 6
0.116 + 0.005 0.154-1-0.002* *
Asterisks denote values significantly different from control ( * P < 0.01, * * P < 0.001). Plasma [glulamine] calculated as ([AI+ [V])/2.
cantly for isoleucine and leucine. Sarcoplasmic glutamine concentration ( # m o l / m l ) , calculated as [total muscle glutamine concentration ( # t o o l / g ) - (ECSV (extracellular space volume m l / g ) × plasma [glutamine] ( # m o l / m l ) ) ] / i n t r a c e l l u l a r volume (ml/g), is shown in Table Ill. The muscle distribution ratios for glutamine and N a + were lower and higher respectively in dexamethasone-treated rats than in controls (Table lit). M u s c l e activities o f glutamine metabolising e n z y m e s Glutaminase was detected in gastrocnemius muscle at low levels of activity, which were not significantly different belween control and dexamethasone-treated ra*.s (68 + :7 vs. 51 + 19 nmol min -~ g-~): approx. 55% of this activity was phosphate-dependent. Glutamine synthetase had much higher activity than glutaminase in this muscle, but there was no difference between synthetase activity of control and dexametbasone-treated rats when expressed per gram wet weight of tissue (475 + 80 vs. 460 + 50 nmol m i n - I g - m) although it was slightly elevated (by 20%) when expressed per total muscle protein (2.5 + 0.3 vs. 3 + 0.2 nmol rain- ~ m g protein in muscles of control and dexamethasone treated rats, respectively). Blood f l o w a n d amino acid balance across hindlimb muscle in vivo There was no significant difference between blood flow either to hindlimbs of control and dexametha-
sone-treated rats as measured using the Transonic blood flowmeter (Table IV) or to individual hindlimb muscles as determined by the microsphere method (e.g., gastrocnemius muscle, 0.072 + 0.015 vs. 0.075 + 0.021 m l / g per min). Arterial glutamine concentrations of control and dexamethasone-treated rats were not significantly different from one another (789 + 58 and 705 + 63 # m o l / l , respectively), but arteriovenous (A-V) concentration difference of glutamine across the hindlimb in vivo was significantly greater ( P < 0.05) in dexamethasone-treated rats ( - 3 0 3 _+ 48 # m o l / l ) than in controls ( - 1 1 1 + 51 # m o l / l ) . Hindlimbs of dexamethasone-treated rats exhibited significantly greater (2.7-fold higher; Table IV) net effluxes of glutamine than controls: there were no significant changes in net fluxes of glutamate and alanine (Table IV). Giutamine transport kinetics in perfused hindlimb The kinetic profile of glutamine i n f u x to hindlimb muscle of dexamethasone-treated rats was markedly different from that observed in control rats (Fig. 1). Kinetic analysis using the graphical method of Hanes (the advantages and limitations of which are discussed in Ref. 22) revealed that the maximal capacity of the transporter was reduced to approx. 37% of control value in steroid-treated rats (control 1058 + 310 nmol m i n - t g - i , dexamethasone-treated 395 + 110 nmol min - t g-~: P < 0.01). There was also a substantial difference in the concentration of perfusate glutamine at which influx
TABLE IV Hindlimb blood flow and net amino acid (AA) flux across hindlimb muscle of control and dexamethasone-treated rats
Blood flow (ml/hindlimb per rain) Control Dexamethasone
1.67 + 0.18 1.73 + 0 . 1 4
Net hindlimb AA flux = (nmol min - i g- i ) glutamine glutamate
alanine
- 31 +_14.3 85 + 13.5 *
- 20 _+12 - 32 + 21
-
+ 6.76 + 2.62 + 1.42+ 4.8
* P < 0.05; significantlydifferent from control. Flux = blood flow (ml/min per g muscle)xA-V concentration difference of AA. A single hindlimb of a 200 g rat contains approx. 6 g of muscle. Values are mean_+S.E. from five preparations in each case.
380 1000
3~J:ar~,~e ~otc~e
800
Control
E -
600
"7 ~= 4 0 0
-t
5
~
4-D. . . . . t . . . . . .
10 15 20 Perfusote glutamme (rnlVl)
25
~*
~0
Fig. 1. Relationship between rate of unidirectional transport of 81utamine and perfusate glutamine concentration in perfused skinned
rat hindlimbfrom control animals and rats treated with desamethasone for 8 days, Resultsare mean values:1:S,E, for 5-8 rats. Analysis by the Hanesmethodyieldsthe followingkineticparameters. Control (O), I/re.~ 1058+310 nmol min-I g muscle-I, gm 8.1-+1.9 mM; dexamethasone (11), 352-+98 nmol min-t g muscle i, Km 2.10 + 0.35 raM.
kinetics of sarcolemmal glutamine exchange were reproduced acutely in perfused hindlimb when dexamethasone was added to perfusate at 200 nM (a concentration chosen to mimic those of corticosteroids found in stress and injury): in these experiments unidirectional glutamine uptake was found to be lower in control limbs (by 329~) at perfusate glutamine concentrations of 0.55 mM but only significantly lower (by 65%) at high extracellular glutamine concentrations (Table V). A kinetic analysis (by the Hanes plot) using data from six different perfusate glutamine concentrations revealed that Vm.,~ and K m of glutamine influx were 370 + 6 nmol m i n - t g-~ and 3.05 + 0.25 mM, respectively, and fractional tracer recovery was 0.82 _+ 0.05. The results indicate that glutamine exchange across the sarcolemmal membrane of rat hindlimb accelerates after dexamethasonetreatment.
Discussion was apparently half-maximal (control 8.1 + 1.9 raM, dexamethasone-treatcd 2.1 + 0.35 mM, P < 0.05). At physiological arterial glutamine concentration (approx. 0.75 mM in the present experiments) the mean glutamine influx to hindlimb muscle was therefore 90 nmol rain-t g-t in control rats and 104 nmol rain-t g - t in dexamethasone-treated rats. We have estimated sarcolemmal glutamine efflux as the sum of influx and net flux across the hindlimb: this gives mean efflux values of 121 nmol rain -I g - i and 189 nmol rain -t g - i for hindlimbs of control and dexamethasone-treated rats, respectively. A semi-quantitative estimate of sarcolemmal glutamine efflux in the perfused hindlimb was also obtained from the recovery of [~H]glutamine tracer in venous perfusate. The fractional recovery of [3H]glutamine extracted by the hindlimb (over the 4 rain post-injection period) was 0.50 :i: 0.07 in controls and 0.80+0.07 in dexamethasone-treated rats, i.e., glutamine exchange was higher after dexamethasonetreatment. The effects of 8-day dexamethasone-treatment on the
TABLE V
Acute effect of dexamethasone (200 nM) on unidirectional glutamine uptake in perfused rat skeletal muscle Pertusate [glutaminel
- Dexamethasone (nmol m i n - i 8 - i )
+ Dexamethasone (nmol rain- i g - t )
73+ 8 370+ 40 955 + 120
57+3 263_+5 338 +_6 *
froM)
0.55 5.5 ?-0
Values are from 3 - 4 preparations. Asterisk denotes value significantly different from control * P < 0.01.
The present results confirm previously reported effects of corticosteroid administration on the rat, i.e., body mass, muscle mass and intramuscular free amino acid concentrations decrease and there is an increased output of glutamine from muscle to blood. The most complete study of the effects of corticosteroid on muscle glutamine balance was that of Muhlbacher et al. [23], who followed the time-course of the effects of steroid administration in dogs. Our results on muscle mass and composition agree with theirs, except that we found no statistically significant decrease in glutaminase as they reported in their dogs. The major new findings of the present work are (a) the alterations in the apparent kinetics of glutamine transport across the muscle membrane as a result of both chronic and acute treatment with dexamethasone, and (b) the observations made in living rats that chronic exposure to dexamethasone results in a reduced sarcolemmal Na + concentration gradient. The amino acid transporter which mediates glutamine influx to muscle appears to be related to the syslem-N transporter of the liver sinusoidal-membrane [24] inasmuch as it is Na+dependent and has a similar substrate specificity (i.e., glutamine, Asn and His). The muscle glutamine-transporter (called by us N m [11]) differs from that in the liver in being insulin-sensitive and pH insensitive; the present results extend our knowledge of its hormone sensitivity. We do not know at present whether the effect of dexamethasone is specific to system-N m in skeletal muscle, but we have preliminary evidence of an effect of dexamethasone on the expression of amino acid transporters in rat liver. Sinusoidal membrane vesicles prepared from livers of rats treated with dexamethasone under identical conditions to those reported in the present study show a 2-fold induction of system-N transport and a 6-fold
381 induction of Na+-glutamate transport via a different transporter (system X-, Ref. 25 and unpublished). It is a well documented fact, confirmed in the present results, that muscle output of glutamine increases after corticosteroid treatment [23] or injury [4] despite a decreased sarcolemmal glutamine gradient. What are the pathophysiological mechanisms underlying the alterations in the net efflux of muscle glutamine, the decreased muscle glutamine pool size and the increases in muscle Na+? in our experiments moderate dexamethasone treatment over 8 days, designed to mimic the pathophysiological blood corticosteroid concentrations observed in injury, increased hindlimb glutamine output by approx. 54 nmol min- 1 g- i (i.e., 3.24 # m o l / g per h, equivalent to 7070 of the intramuscular glutamine pool/h). The measured decrease in muscle glutamine concentration over 8 days of dexamethasone treatment (2.04 #tool/g) could support a mean increase in glutamine output of only 0.18 nmol min -1 g-i over this period indicating that sources other than the intracellular glutamine pool were also making large contributions towards the observed increase in glutamine output. In muscle, glutamine can be synthesized from the branched chain amino acids, from glutamate and aspartate all of which are present in muscle protein (the BCAA contributing 2070 of the amino acid residues) and whose muscle free concentrations increase during dexamethason¢ treatment (indicating no shortage of precursor substrates for glutamine synthesis during glucocorticoid treatment). However, calculating on the basis that glutamine accounts for between 40 to 6070 of the observed nitrogen efflux from muscle, net protein breakdown (which was equivalent to 34 m g / g muscle over 8 days) could account for only 2270 (8-12 nmol glutamine/min per g muscle) of the total observed output (54 nmol min-i g-i). The strong implication of these calculations is that dexamethasone has increased net glutamine synthesis in muscle but this alone could not explain the increase in glutamine output since the simultaneous marked fall in the sarcolemmal distribution ratio of glutamine would tend to retard net glutamine ¢fflux from muscle. The accelerated sarcolemmal glutamine exchange reported here would therefore appear to be a prerequisite for increased release of muscle glutamine under certain pathophysiological circumstances. It has recently been reported [26,27] that dexamethasone increases glutamine synthetase activity in rat skeletal muscle when it is administered at doses above 0.5 mg kg-i day-1 [26]. Induction of glutamine synthetase activity appears to be at the transcriptional level since there is an overall increase in the level of mRNA encoding the enzyme [26] with a resultant increase in the number of enzyme molecules (i.e., a Vma~ effect). We therefore investigated whether glutamine synthetase activity can be stimulated in muscle extracts of dexa-
methasone treated rats at saturating glutamate concentrations (given that the K m for glutamate is - 2 mM the reaction rates measured are likely to reflect the maximal enzyme capacity in vitro). Under these conditions we observed no significant changes in the total activities of glutamine synthetase and glutaminase in muscle after dexamethasone-treatment; although when expressed on the basis of total muscle protein there was an indication of a slight rise in glutamine synthetase activity (of approx. 2070). Therefore increased net flux through glutamine synthetase in our studies appears to be due largely to local activation of enzyme in vivo, rather than de novo enzyme synthesis. Induction of glutamine synthetase mediated via changes in transcriptional events may, however, be an additional means by which skeletal muscle may increase its glutamine production in response to dexamethasone treatment. It is central to our argument that changes in the rate of muscle glutamine synthesis (by whatever mechanism) cannot account for the fact that the intramuscular glutamine pool falls after dexamethasone treatment (i.e., the lowered sarcolemmal glutamine gradient occurs despite an apparent increase in muscle synthesis of glutamine). It appears, therefore, that the sarcoplasmic glutamine pool is not simply a store of glutamine. We have suggested [11] that the high intramuscular glutamine concentration is maintained by concentrative uptake of glutamine across the sarcolemma via N a + / glutamine cotransport. The distribution ratio of glutamine across the sarcolemma may therefore be dependent to a large extent on the electrochemical Na + gradient across this membrane. Our results extend the available information on the effects of corticosteroid on muscle by demonstrating a rise of intramuscular Na + concentration, although we were unable to detect any associated change in the muscle membrane-potential, which w a s ' - - 8 0 mV in leg extensor muscle (EDL) isolated from hindlimbs of dexamathasone-treated and control rats (unpublished data of Ward, Hundal and Taylor). A rise in intracellular Na + concentration has also been observed in other tissues exposed to corticosteroid [28,29] and may possibly be related to increased 'leakiness' of the sarcolemma to Na + and also to modulation of the activity of the N a + / K + pump. We have previously shown that a reduction in the inward Na +electrochemical gradient which is associated with a fall in the maximal capacity of glutamine transport (Vmax) into muscle but with no significant change in transport g m [4]. Our recent studies in sarcolemmal vesicles [30] and in the perfused rat hindlimb [4] demonstrate that sarcolemmal glutamine transport falls when the Na gradient is dissipated by either (a) replacement of the inwardly directed concentration gradient of NaC! with KCI or choline chloride, (b) depolarisation of the membrane with the cation ionophore valinomycin or (c) denervation of hindlimb muscle. In denervated muscle
382 [4] there is a fall in the Na gradient (muscle Na increases by 26~: similar to that in the present studies), resting sarcolemmal potential is depolarised (by 15 mV) and unidirectional glutamine transport falls by 40%. This depression in membrane transport can be reversed by partly restoring the Na gradient by exposure of d e n e ~ a t e d muscle membranes to the Na channel blocker ~ X (this facilitates reduction in intramuscular Na through a reduction in membrane Na permeability and extrusion of muscle Na by the Na pump). These observations are consistent with the notion that changes in muscle glutamine distribution during catabolic episodes may be closely associated to changes in muscle Na distribution, The present results also show an association between reductions in transport Vm.x and the Na + gradient, although with an additional reduction in transport Km. It is therefore possible that the K m effect is unrelated to the change in Na + gradient, and may be the result either of changes in membrane fluidity (which may affect the operation/kinetic characteristics of membrane proteins) or of direct dexamethasone action on specific membrane proteins (see below). Dexamethasone treatment results in an increased fractional turnover of the muscle glutamine pool from 0.026/min to 0.074/min (fractional turnover is calculated as sarcolemmal glutamine efflux/muscle glutamine concentration). The accelerated sarcolemmal fluxes of glutamine could result from a change in the exchange properties of the System N m transporter as other kinetic characteristics of this System are altered by dexamethasone treatment. We have recently demonstrated that System N m in sarcolemmal vesicles is capable of mediating Na + dependent glutamine transport into and out of muscle [30], but we cannot exclude a number of other possibilities such as (a) the existence of two cartiers with similar substrate specificities that may be differentially modulated by corticosteroids and (b) alterations in the V m a J K m ratio for glutamine efflux (e.g,, by trans, stimulation of the efflux mechanism, alIosteric modification of transporter molecules or alterations in membrane lipid properties). We have discussed our results, in vivo, in terms of specific effects of dexamethasone itself rather than of possible concurrent side-effects such as alterations in the concentrations of other hormones (e.g., insulin, glueagon). The fact that we have demonstrated acute effects of corticosteroid on the Vm,~ and K m of glutamine transport (i.e., when dexamethasone was added to the pcrfusate of hindlimbs of control rats) indicates that we ought to be entitled to interpret the data, at least partly, in terms of direct glucocorticoid action. Other acute effects (within 0.5-1 h) of dexamethasone have been observed in incubated preparations of rabbit digit extensor muscles (on muscle protein synthesis) [31]. The short term effects may reflect rapid cell penetration by the steroid (and subsequent effects on intracellular
metabolism) a n d / o r modification of membrane phospholipids through alterations in the activities of phospi~olipases (e.g., phospholipase A 2)The dexamethasone effects on glutamine transport kinetics appear to be very rapid and therefore they may play an important role in the rapid alterations of amino acid balance in muscle tissue observed after injury and the onset of disease. However the changes in the kinetic characteristics of the N m-transporter persist (for at least 8 days in this study) in their effect and will predispose any damaged muscle to an increase in the loss of amino acid nitrogen. Furthermore, the kinetic characteristics of splanchnic glutamine uptake [24,32,25] are such as to make the rate of hepatic ureagenesis proportional to the amino acid supply [33]. Thus, interactions between the muscle and splanchnic mechanisms may help explain why the increased provision of amino acids to severely injured and septic patients does not result in the immediate resumption of positive nitrogen balance in muscle or the whole-body [2]. The increase in intracellular Na + of rat skeletal muscle after dexamethasone treatment may increase intramuscular Ca 2+ by inhibition of cation-exchange mechanisms, which in turn may affect numerous metabolic processes including protein turnover. It is noteworthy that in many circumstances of muscle wasting associated with trauma, in which there is likely to be increased corticosteroid secretion, intramuscular Na + (and occasionally Ca 2+) concentration is elevated [5,8]. In addition, an alteration in the size of the intramuscular glutamine pool may itself affect muscle metabolism. Since glutamine appears to be an essential amino acid for the growth of cells in tissue culture [34,35] it occurred to us that it may have an influence on skeletal muscle protein synthesis. We have discovered that there is indeed a good correlation between skeletal muscle protein synthetic rate and intramuscular glutamine concentration both in vivo [36] and in the perfused rat hindlimb [37]. There is also evidence that glutamine reduces protein breakdown in cultured skeletal muscle [38] and we have also observed glutamine inhibition of protein breakdown by an insulin-independent mechanism in the perfused rat hindlimb [39]. More recently we have been able to demonstrate that the administration of glutamine (in the form of the dipeptide alanyiglutamine) to normal healthy subjects results in a stimulation of muscle protein synthesis [40]. The modulation of sarcolemmal transport properties (e.g., by hormones or other effectors as a result of injury or disease) could therefore influence the metabolic behaviour of muscle and predispose it towards increased loss of amino acids, particularly glutamine, to the rest of the body. Since glutamine appears to be important for the normal functioning of white cells, and dividing cells in general [35], this apparently pathophysiological behaviour may have some benefits.
383
Acknowledgements We are grateful to Dr. M.R. Ward for making the membrane potential measurements. This work was supported by Action Research for the Crippled Child, Ajinomoto GmbH, British Diabetic Association, The Medical Research Council, Pfrimmer-Kabi GmbH, The Rank Prize Funds, The Wellcome Trust and The University of Dundee. References 1 Cuthbertson, D.P. (1971) Q. J. Med. 25, 233-245. 2 Rennie, M.J. (1985) Br. Med. Bull. 41,257-264. 3 Ftirst, P., Elwyn, D.H., Askanazi, J., Kinney, J.M. (1982) in Clinical Nutrition: (Wesdorp, R.I.C. and Soeters, P.B., eds.), pp. 10-17, Churchill Livingstone, Edinburgh. 4 Hundal, H.S., Babij, P., Watt, P.W., Ward, M.R. and Rennie, M.J. (1990) Am. J. Physiol. 259, E148-E154. 5 Bergstrom, J., FUrst, P., Holmstrom, B.O., Vinnars, E., Askanazi, J,, Elwyn, D.H., Michelson, C.B. and Kinney, J.M. (1981) Ann. Sur8. 193, 810-816. 6 Cunningham, J.N., Jr., Carter, N.W., Rector, F.C., Jr. and Seldin, D.W. (1971) J. Clin. Invest. 50, 49-59. 7 Trunkey, D.D., lllner, H., Wagner, 1.Y. and Shires, F.T. (1973) Surgery 74, 241-250. 8 Turinsky, J. and Gonnerman, W.A. (1982) J. Surg. Res. 33, 337344. 9 Moore, F.D. (1957) Recent Prog. Horm. Res. 13, 511-576. 10 Wilmore, D.W. (1986) Clin. Nutr. 5, 9-19. 11 Hundal, H.S., Rennie, M.J. and Watt, P.W. (1987) J. Physiol. 393, 283-305. 12 Lund, P. (1970) Biochem. J. 118, 35-39. 13 Huang, Y-Z. and Knox, W.E. (1976) Enzyme 21,408-426. 14 Moore, R.D., Munford, J.W. and Pillsworth, T.J. (1983) J. Physiol. 338, 277-294. 15 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 16 Burton, R.C. and Gorewitt, R.C. (1984) Med. Electronics 15, 68-73. 17 Rothwell, N.J. and Stock, MJ. (1981) Pflugers Arch. 389, 239-242. 18 Bergmeyer H.U. (1965) Methods of Enzymatic Analysis, Verlag Chemie-Academic Press, New YorK.
19 Tsanev, R. and Markov, G.G. (1960) Biochim. Biophys. Acta. 42, 442-452. 20 Yudilevich, D.L. and Mann, G.E. (1982) Fed. Proc. 41, 3045-3053. 21 Hundal, H.S., Rennie, M.J. and Watt, P,W. (1989) J. Physiol. 408, 93-114. 22 Price, N.C., Stevens, L. (1982) Fundamentals of Enzymology, Oxford University Press, Oxford. 23 Muhlbacher, F., Kapadia, C.R., Colpoys, M.F., Smith, R.J. and Wilmore, D.W. (1984) Am. J. Physiol. 247, E75-E83. 24 Kilberg, M.S., Handlogten, M.E. and Christensen, H.N. (1980) J. Biol. Chem. 255, 4011-4019. 25 Taylor, P.M. and Rennie, M.J. (1987) FEBS Lett. 221,370-374. 26 Max, S.R., Mill, J., Mearow, M., Konagaya, M., Konagaya, Y., Thomas, J.W., Banner, C. and Vitkovic, L. (1988) Am. J. Physiol. 255, E397-E403. 27 Tischler, M.E., Henriksen, EJ. and Cook, P.H. (1988) Muscle Nerve 11,752-756. 28 Rayson, B. and Gupla, g. (1985) J. Biol. Chem. 260, 12740-12743. 29 Klein, L.E., Barlolomei, M. and Lo, C.S. (1987) Am. J. Physiol. 249, H570-575. 30 Ahmed, A., Taylor, P.M and Rennie, M.J. (1990) Am. J. Physiol. 259, E284-E291. 31 Reeds, P.J. and Palmer, R.M. (1984) Biochem. J. 219, 953-957. 32 Joseph, S.K., Bradford, N.M. and McGivan, J.D. (1978) Biochem. J. 176, 827-836. 33 Meijer, A.J., Lof, C., Ramos, I.C. and Verhoeven, A.J. (1985) Eur. J. Biochem. 148, 189-196. 34 Kovacevic, I.C.S. and McGivan, J.D. (1983) Physiol. Rev. 63, 547-605. 35 Newsholme, E.A., Crabtree, B. and Ardawi, M.S.M. (1985) Q. J. Exp. Physiol. 70, 473-489. 36 Rennie, M.J., Hundal, H.S., Babij, P., MacLennan, P., Taylor, P.M., Watt, P.W., Jepson, M.M. and Millward, D.J. (1986) Lancet ii, 1008-1012. 37 MacLennan, P.A., Brown, R.A. and Rennie, M.J. (1987) FEBS Left. 215, 187-191. 38 Smith, R.J. (1985) in Intracellular protein catabolism (Khairallah, E.A., Bond, J., Bird, J.D., eds.) pp. 633-635, Alan R. Liss, New York. 39 MaeLennan, P.A., Smith, K., Weryk, B., Watt, P.W. and Rennie, M.J. (1988) FEBS Left. 237, 133-136. 40 Barua, J.M., Smith, K., Wilson, E., Scrimgeour, C. and Rennie, M.J. (1991) Diabetologica Lalina, in press.
Biochimica et Biophysica Acta, 1092 (1991) 376-383 © 1991 Elsevier Science Publishers B.V. 0167-4889/91/$03.50 ADONIS 016748899100166F
BBAMCR 12929
Effects of corticosteroid on the transport and metabolism of glutamine in rat skeletal muscle H a r i n d e r S. H u n d a l *, Philip Babij **, P e t e r M. T a y l o r , P e t e r W . Watt and M i c h a e l J. R e n n i e Deparlmen! of Anatomy & Ph.vsiolol{v. The University. Dut~dee(U.K.) (Received 7 September 1990)
Key words: Dexamethasone; Amino acid; Cell membrane; Trauma
Intramuscular glutamine falls with injury and disease in circumstances associated with increases in blood corticosteroids. We have investigated the effects of corticosteroid administration (0.44 m g / k g dexamethasone daily for 8 days, 200 g female rats) on intramuscular giutamine and Na +, muscle glutamine metabolism and sarcolemmal glutamine transport in the perfused hindlimb. After dexamethasone treatment intramuscular giutamine fell by 45% and Na + rose by Z~% (the respective muscle/plasma distribution ratios changed from 8.6 to 4.5 and 0.12 to 0.15); glutamine synthetase and giutaminase activities were unchanged at 475 + 75 and 60 + 19 nmol/g muscle per min. Glutamine output by the hindlimb of anaesthetized rats was increased from 31 to 85 nmol/g per min. Sarcolemmal glutamine transport was studied by paired-tracer dilution in the perfused hindlimb: the maximal capacity (Vmax) for glutamine transport into muscle (by Na+-glutamine symport) fell from 1058 4-310 to 395 4- 110 nmol/g muscle per min after dexamethasone treatment, accompanied by a decrease in the Km (from 8.1 + !.9 to 2.1 4. 0.4 mM glutamine). At physiological plasma glutamine concentration (0.75 mM) dexamethasone appeared to cause a proportional increase in sarcolemmal glutamine efflux over influx. Addition of dexamethasone (200 nM) to the perfusate of control rat hindlimbs caused acute changes in Vm, and K m Of glutamine transport similar to those resulting from &day dexamethasone treatment. The reduction in muscle glutamine concentration after dexamethasone treatment may be primarily due to a reduction in the driving force for intramuscular giutamine accumulation, i.e., in the Na + electrochemical gradient. The prolonged increase in muscle glutamine output after dexamethasone treatment (which occurs despite a reduction in the size of the intramuscular glutamine pool) appears to be due to a combination of (a) eceelerated sarcolemmal giutamine efflux and (b) increased intramuscular synthesis of glutamine.
Introduction It is well recognised that the metabolic response to injury and disease includes changes in muscle composition, irrespective of whether or not muscle was the initial site of trauma [1,2]. The longer term responses usually include loss of muscle mass, but even in the early stages of the acute response, e.g., to injury and trauma, when it is impossible to detect a loss of muscle
* Present Address: Department of Cell Biology, Hospital For Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G IX8. ** Present Address: Department of Physiology, University College London, London WC IE, U.K. Correspondence: M.J. Rennie, Department of Anatomy & Physiology, The University, Dundee DD1 4HN, U.K.
protein, there are marked losses of glutamine from the muscle free amino acid pool [3,4]. The ionic composition of sarcoplasm also changes with losses of K + (associated with depolarisation of the sarcolemma) and increases in Na + and Ca 2+ [5-8]. Furthermore, the acute metabolic responses to bodily injury are invariably associated with a rise in blood corticosteroid concentrations [9,10]. Although many of the changes in muscle have been related to changes in protein synthesis and breakdown, (e.g., see Ref. 2 for review) it seemed to us that the observed alterations in these were insufficiently rapid to account for the derangements in muscle amino acid balance observed in sick and injured patients. We therefore set out to see whether or not the kinetic properties of the Na+-dependent muscle glutamine transporter, which we have previously characterised in the perfused rat hindlimb [11], are modifiable, acutely and chroni-
377 cally, by corticosteroid (dexamethasone) administration. We have also investigated possible effects of dexamethasone on the intramuscular concentrations of glutamine and Na ÷ and on the maximal capacities of muscle glutamine synthetase and glutaminase. Materials and Methods
All chemicals were obtained from Sigma (Poole, U.K.) except bovine serum albumin (BSA) fraction V (Miles-Pentex, Kankakee, U.S.A) and sodium pentobarbitone (May and Baker, Dagenham, U.K). Radioactive tracers were purchased from Amersham International (Amersham, U.K) ([3H]glutamine at 6 T B q / m mol, [14C]mannitol at 20 GBq/mmol) or New England Nuclear (Boston, U.S.A) (46Sc-labelled microspheres). Female Wistar rats (Bantin & Kingman, Hull, U.K.) of approx. 200 g body weight were used. Rats had access to food and water until the night before study when food only was withdrawn. Food intake was monitored by providing animals with a fixed quantity of food every 12 h (20 g/rat) and weighing the food not eaten. 40 rats were divided into two groups: one group (15 rats) was treated with dexamethasone, the other group (25 rats) served as control. Dexamethasone (Sigma) was administered by suprascapular subcutaneous injection (0.44 m g / k g body weight as a fine suspension in 0.9% NaCl) every day for 8 days: control rats received saline. It was impractical to carry out all experimental measurements on each rat, therefore the experimental and control groups were sub-divided into groups of five rats. These groups were assigned for measurement of (a) muscle glutamine synthetase and glutaminase, muscle composition (e.g., amino acids, muscle N a + / K +, RNA, DNA and protein), blood flow and arteriovenous differences of amino acids in the hindlimb, (b) measurement of blood flow to individual muscles by the microsphere method, (c) glutamine transport kinetics in the perfused hindlimb, (d, e) control rats only, glutamine transport kinetics in the perfused hindlimb _+ 200 nM perfusate dexamethasone (5 rats each). Rats were anaesthetized with sodium pentobarbitone (60 mg/kg i.p.) prior to muscle sampling, determination of blood flow in vivo or preparation for hindlimb perfusion. For measurements of amino acids and Na + in muscle, soleus and gastrocnemius muscle of one leg were frozen rapidly using Wollenberger clamps cooled in liquid N 2. Approx. 0.37 MBq of [14C]mannitol, was injected 1 h before sampling of plasma and muscle for estimation of extracellular space. The gastrocnemius muscle of the opposite leg was removed rapidly, and used immediately for analysis of glutamine synthetase (rate of glutamine synthesis from 20 mM glutamate: pH 7.4) [12] and total glutaminase activity (rate of glutamate production from 20 mM glutamine: pH 8.0 _+ 10 mM
phosphate) [13]. To measure intramuscular Na ÷ the extraceilular space of single hindlimbs from five dexamethasone treated and five control rats were 'flushed' (via a femoral artery cannula) with 20 ml of ice-cold, isotonic sucrose solution. We found that this procedure washed out approx. 95% of calculated extracellular Na +, and a further flushing with 20 ml of sucrose solution yielded only 1.5% of Na + washed-out by the initial treatment. Muscle was then clamp-frozen. For estimation of muscle and plasma Na ÷, tissue was dried to constant weight at 90 ° C, powdered and dissolved in 1 M H N O 3 for flame photometry by standard methods [14]. Protein determination was by the Lowry technique [151. Whole hindquarter blood flow was measured using an ultrasonic flowmeter (Transonic Systems) which uses a wide-beam transit-time technique [16]. The ultrasound probe was placed around the aorta in the lower abdomen just above the iliac b,.'furcation so that flow to other tissues such as the kidneys was excluded. Flow became stable after 5 min and the mean of several readings over 15 min was taken as the hindquarter flow. Single hindlimb flow was obtained by d;viding the value by two. For measurement of blood flow to individual muscles, a microsphere method [171 was used utilising 15 #m 46Sc-labelled microspheres. Reference blood samples were collected over 15 s at 0.8 ml/min from the femoral artery. Femoral venous blood for arterio-venous difference measurements (0.1 ml, in duplicate) was taken from the leg opposite to that used for the blood-flow. Arterial blood (0.1 ml in duplicate) was sampled from the carotid artery or the abdominal aorta. In some experiments plasma was isolated by centrifugation (2000 X g for 20 min at 4°C). A neutralised perchloric acid 10% (w/v) extract of powdered muscle, blood or plasma was analyzed for amino acids by either an enzymatic method [lg] or by means of a Biotronik LCS000 amino acid analyzer. Muscle RNA and DNA were determined by a dual wavelength spectrophotometric method [19]. Glutamine transport kinetics were investigated using a paired-tracer isotope dilution technique [20] applied to the skinned perfused hindlimb [11.21]. Briefly, anaesthetized rats were prepared for single hindlimb perfusion by cannulation of the femoral artery and femoral vein. Hindlimbs were perfused with a cell free perfusate (Krebs-Henseleit bicarbonate buffer containing 6% BSA (w/v) and 5 mM glucose) which was pumped through the animals hindlimb, without recirculation, at a flow rate of 2 ml rain-i. Glutamine transport kinetics in hindlimbs of control and dexamethasone treated rats were determined under steady state conditions by the paired-tracer method (using [3H]glutamine as the transportable molecule and [14C]mannitol as the extracellular space marker). Perfusate glutamine was varied between 0.5 mM and 30 mM
378 (with at least 10 min allowed for equilibration at each glutamine concentration). In studies in which we investigated the acute effect of 200 nM dexamethasone, hindlimbs were pre-equilibrated with the hormone (for at least 30 min) and it was retained in the perfusate for the remainder of the perfusion period (up to 1.5 h). Calculation of fractional tracer efflux was determined as described by Yudilevich and Mann [20]. Values for VmaX and K m were calculated from the Hanes plot [22] (i.e., [S]/v t, vs. [S]; where [S] is perfusate glutamine concentration and ot is the unidirectional influx) obtained from least squares regression analysis using software written for the Apple lie microcomputer. Other experimental details were exactly as previously described [11]. Data are presented as the mean + 1 standard error (S.E.). Differences between means were tested for significance using Student's t-test and were considered significant if P < 0.05. Results
Body weight, food intake and muscle composition The initial body weight (mean + S.E.) of experimental animals was 200 _+4 g (n -- 40). After 8 days dexamethasone treatment body weight was reduced to 189 + 1 g (n = 15), although control rats gained weight to 212 + 2 g (n = 25)) over the same period. This represents an overall weight loss equivalent to 10.8~ of final control weight. Food intake in dexamethasone treated rats (14.1 :l= 0.28 g / d a y per rat) was significantly ( P < 0.05) lower than in controls (15.9 + 0.35 g / d a y per rat). In order to ensure that the treated rats were not lasing weight due to malnutrition we fed a group of age-andweight matched control rats 14 g food/day per rat for 8 days without any loss of body weight. This finding indicated that the weight loss experienced by the dexamethason~treated rats was unlikely to be due to under-
TABLE II
Intramuscular amino acids in gastrocnemius muscle of control and 8 day dexamethasone treated rats Amino acid
Control ( p m o l / g wet weight)
Dexamethasone (p.mol/g wet weight)
Glutamine Glutamate Alanine Aspartate Valine lsoleucine Leucine
4.55 +0.29 0.80 -+0.10 1.40 -+0.10 0.48 +0.14 0.13 +0.01 0.016 =l:0.002 0.046 + 0.006
2.51 +0.41 * * * !.43 +0.05 * * 1.37 _+0.04 0.44 -+0.04 0.15 -+0.01 0.027 _+0.002 * 0.066 -I-0.002 *
(4)
Values are mean + S.E. Asterisks denote values significantly different from control ( * P < 0.05, * * P < 0.01, * * * P < 0.001).
nutrition. Differences between final hindlimb-muscle weights of control and dexamethasone-treated rats were in proportion to the difference in body weight (gastrocnemius 1.11 vs. 0.98 g, soleus 78 vs. 70 mg, anterior tibialis 312 vs. 275 mg wet weight, respectively; Table I). There were no significant differences between either tissue hydration or extracellular space volume of control and treated rats (Table l); in both groups intracellular volume was calculated to be approx. 0.61 m l / g wet weight. We have therefore expressed all data with wet weight as denominator. The concentration of protein in gastrocnemius muscle was lower in dexamethasonetreated rats than in controls (by 18%) and sarcoplasmic Na + concentration was 257o higher in dexamethasonetreated rats than in controls; K + concentration remained unaltered (Table !). Gastrocnemius glutamine fell in dexamethasone treated rats (by 45~) whilst intramuscular glutamate increased by 3-fold (Table II). We observed no significant changes in the muscle concentrations of alanine or aspartate but the concentrations of branched chain amino acids (valine, isoleucine and leucine) were all elevated although only signifi-
TABLE I
Effects o[ &day dexamethasone treatment on the composition of rat gastrocnemius muscle Data are presented as the mean -+ S.E. (per g muscle wet weight). (Number of determinations in parentheses).
Muscle weight (g) Tissue hydration (g H 2 O / g ) Extracellularvolume(ml/g) Protein concentration (rag/g) RNA (m8/8) D N A (mB/g) RNA/DNA RNA/alkali soluble protein Sodium concentration ( p m o l / g ) Potassium concentration ( p m o l / g )
(10) (4) (4) (4) (4) (4)
Control
Dexamet hasone-treated
!,11 + 0.015 0.75 +0.01 0.16 +0.03 189 +7 1,3 -+0.09 0.383 + 0.025 3.4 +0.01 0.007+0.0002 10.0 _+0.7 92.4 _+6.2
0.98 -+ 0.02 * * * 0.75 +0.01 0.13 :l:O.0l 155 _+5 * * 1.71 +0.29 0.424 + 0.020 4.0 +0.5 0.01 +0.001 * * 12.5 +0.4 * 92.2 + 5.8
Asterisks denote values significantly different from control ( * P < 0.05, * * P < 0.01, * * * P < 0.001).
(I0) 0o) (5) (5) (4) (4) (4) (4) (5) (5)
379 TABLE Ill Sarcoplasmic concentrations of glutamine and sodium (raM) and sarcoplasm /plasma distribution ratio (DR), in gastrocnemius muscle of control and dexamethasone-treated rats
Values are means+_S.E. from at least five rats in each case. Glutamine Control Dexamethasone
Na
muscle
plasma
DR
muscle
plasma
DR
7.3 + 0.50 3.9+0.60 * *
0.85 + 0.04 0.86 +0.06
8.6 + 0.16 4.5 +0.39 * *
16 + 1 20 + 0.6 *
138+ 2 130+ 6
0.116 + 0.005 0.154-1-0.002* *
Asterisks denote values significantly different from control ( * P < 0.01, * * P < 0.001). Plasma [glulamine] calculated as ([AI+ [V])/2.
cantly for isoleucine and leucine. Sarcoplasmic glutamine concentration ( # m o l / m l ) , calculated as [total muscle glutamine concentration ( # t o o l / g ) - (ECSV (extracellular space volume m l / g ) × plasma [glutamine] ( # m o l / m l ) ) ] / i n t r a c e l l u l a r volume (ml/g), is shown in Table Ill. The muscle distribution ratios for glutamine and N a + were lower and higher respectively in dexamethasone-treated rats than in controls (Table lit). M u s c l e activities o f glutamine metabolising e n z y m e s Glutaminase was detected in gastrocnemius muscle at low levels of activity, which were not significantly different belween control and dexamethasone-treated ra*.s (68 + :7 vs. 51 + 19 nmol min -~ g-~): approx. 55% of this activity was phosphate-dependent. Glutamine synthetase had much higher activity than glutaminase in this muscle, but there was no difference between synthetase activity of control and dexametbasone-treated rats when expressed per gram wet weight of tissue (475 + 80 vs. 460 + 50 nmol m i n - I g - m) although it was slightly elevated (by 20%) when expressed per total muscle protein (2.5 + 0.3 vs. 3 + 0.2 nmol rain- ~ m g protein in muscles of control and dexamethasone treated rats, respectively). Blood f l o w a n d amino acid balance across hindlimb muscle in vivo There was no significant difference between blood flow either to hindlimbs of control and dexametha-
sone-treated rats as measured using the Transonic blood flowmeter (Table IV) or to individual hindlimb muscles as determined by the microsphere method (e.g., gastrocnemius muscle, 0.072 + 0.015 vs. 0.075 + 0.021 m l / g per min). Arterial glutamine concentrations of control and dexamethasone-treated rats were not significantly different from one another (789 + 58 and 705 + 63 # m o l / l , respectively), but arteriovenous (A-V) concentration difference of glutamine across the hindlimb in vivo was significantly greater ( P < 0.05) in dexamethasone-treated rats ( - 3 0 3 _+ 48 # m o l / l ) than in controls ( - 1 1 1 + 51 # m o l / l ) . Hindlimbs of dexamethasone-treated rats exhibited significantly greater (2.7-fold higher; Table IV) net effluxes of glutamine than controls: there were no significant changes in net fluxes of glutamate and alanine (Table IV). Giutamine transport kinetics in perfused hindlimb The kinetic profile of glutamine i n f u x to hindlimb muscle of dexamethasone-treated rats was markedly different from that observed in control rats (Fig. 1). Kinetic analysis using the graphical method of Hanes (the advantages and limitations of which are discussed in Ref. 22) revealed that the maximal capacity of the transporter was reduced to approx. 37% of control value in steroid-treated rats (control 1058 + 310 nmol m i n - t g - i , dexamethasone-treated 395 + 110 nmol min - t g-~: P < 0.01). There was also a substantial difference in the concentration of perfusate glutamine at which influx
TABLE IV Hindlimb blood flow and net amino acid (AA) flux across hindlimb muscle of control and dexamethasone-treated rats
Blood flow (ml/hindlimb per rain) Control Dexamethasone
1.67 + 0.18 1.73 + 0 . 1 4
Net hindlimb AA flux = (nmol min - i g- i ) glutamine glutamate
alanine
- 31 +_14.3 85 + 13.5 *
- 20 _+12 - 32 + 21
-
+ 6.76 + 2.62 + 1.42+ 4.8
* P < 0.05; significantlydifferent from control. Flux = blood flow (ml/min per g muscle)xA-V concentration difference of AA. A single hindlimb of a 200 g rat contains approx. 6 g of muscle. Values are mean_+S.E. from five preparations in each case.
380 1000
3~J:ar~,~e ~otc~e
800
Control
E -
600
"7 ~= 4 0 0
-t
5
~
4-D. . . . . t . . . . . .
10 15 20 Perfusote glutamme (rnlVl)
25
~*
~0
Fig. 1. Relationship between rate of unidirectional transport of 81utamine and perfusate glutamine concentration in perfused skinned
rat hindlimbfrom control animals and rats treated with desamethasone for 8 days, Resultsare mean values:1:S,E, for 5-8 rats. Analysis by the Hanesmethodyieldsthe followingkineticparameters. Control (O), I/re.~ 1058+310 nmol min-I g muscle-I, gm 8.1-+1.9 mM; dexamethasone (11), 352-+98 nmol min-t g muscle i, Km 2.10 + 0.35 raM.
kinetics of sarcolemmal glutamine exchange were reproduced acutely in perfused hindlimb when dexamethasone was added to perfusate at 200 nM (a concentration chosen to mimic those of corticosteroids found in stress and injury): in these experiments unidirectional glutamine uptake was found to be lower in control limbs (by 329~) at perfusate glutamine concentrations of 0.55 mM but only significantly lower (by 65%) at high extracellular glutamine concentrations (Table V). A kinetic analysis (by the Hanes plot) using data from six different perfusate glutamine concentrations revealed that Vm.,~ and K m of glutamine influx were 370 + 6 nmol m i n - t g-~ and 3.05 + 0.25 mM, respectively, and fractional tracer recovery was 0.82 _+ 0.05. The results indicate that glutamine exchange across the sarcolemmal membrane of rat hindlimb accelerates after dexamethasonetreatment.
Discussion was apparently half-maximal (control 8.1 + 1.9 raM, dexamethasone-treatcd 2.1 + 0.35 mM, P < 0.05). At physiological arterial glutamine concentration (approx. 0.75 mM in the present experiments) the mean glutamine influx to hindlimb muscle was therefore 90 nmol rain-t g-t in control rats and 104 nmol rain-t g - t in dexamethasone-treated rats. We have estimated sarcolemmal glutamine efflux as the sum of influx and net flux across the hindlimb: this gives mean efflux values of 121 nmol rain -I g - i and 189 nmol rain -t g - i for hindlimbs of control and dexamethasone-treated rats, respectively. A semi-quantitative estimate of sarcolemmal glutamine efflux in the perfused hindlimb was also obtained from the recovery of [~H]glutamine tracer in venous perfusate. The fractional recovery of [3H]glutamine extracted by the hindlimb (over the 4 rain post-injection period) was 0.50 :i: 0.07 in controls and 0.80+0.07 in dexamethasone-treated rats, i.e., glutamine exchange was higher after dexamethasonetreatment. The effects of 8-day dexamethasone-treatment on the
TABLE V
Acute effect of dexamethasone (200 nM) on unidirectional glutamine uptake in perfused rat skeletal muscle Pertusate [glutaminel
- Dexamethasone (nmol m i n - i 8 - i )
+ Dexamethasone (nmol rain- i g - t )
73+ 8 370+ 40 955 + 120
57+3 263_+5 338 +_6 *
froM)
0.55 5.5 ?-0
Values are from 3 - 4 preparations. Asterisk denotes value significantly different from control * P < 0.01.
The present results confirm previously reported effects of corticosteroid administration on the rat, i.e., body mass, muscle mass and intramuscular free amino acid concentrations decrease and there is an increased output of glutamine from muscle to blood. The most complete study of the effects of corticosteroid on muscle glutamine balance was that of Muhlbacher et al. [23], who followed the time-course of the effects of steroid administration in dogs. Our results on muscle mass and composition agree with theirs, except that we found no statistically significant decrease in glutaminase as they reported in their dogs. The major new findings of the present work are (a) the alterations in the apparent kinetics of glutamine transport across the muscle membrane as a result of both chronic and acute treatment with dexamethasone, and (b) the observations made in living rats that chronic exposure to dexamethasone results in a reduced sarcolemmal Na + concentration gradient. The amino acid transporter which mediates glutamine influx to muscle appears to be related to the syslem-N transporter of the liver sinusoidal-membrane [24] inasmuch as it is Na+dependent and has a similar substrate specificity (i.e., glutamine, Asn and His). The muscle glutamine-transporter (called by us N m [11]) differs from that in the liver in being insulin-sensitive and pH insensitive; the present results extend our knowledge of its hormone sensitivity. We do not know at present whether the effect of dexamethasone is specific to system-N m in skeletal muscle, but we have preliminary evidence of an effect of dexamethasone on the expression of amino acid transporters in rat liver. Sinusoidal membrane vesicles prepared from livers of rats treated with dexamethasone under identical conditions to those reported in the present study show a 2-fold induction of system-N transport and a 6-fold
381 induction of Na+-glutamate transport via a different transporter (system X-, Ref. 25 and unpublished). It is a well documented fact, confirmed in the present results, that muscle output of glutamine increases after corticosteroid treatment [23] or injury [4] despite a decreased sarcolemmal glutamine gradient. What are the pathophysiological mechanisms underlying the alterations in the net efflux of muscle glutamine, the decreased muscle glutamine pool size and the increases in muscle Na+? in our experiments moderate dexamethasone treatment over 8 days, designed to mimic the pathophysiological blood corticosteroid concentrations observed in injury, increased hindlimb glutamine output by approx. 54 nmol min- 1 g- i (i.e., 3.24 # m o l / g per h, equivalent to 7070 of the intramuscular glutamine pool/h). The measured decrease in muscle glutamine concentration over 8 days of dexamethasone treatment (2.04 #tool/g) could support a mean increase in glutamine output of only 0.18 nmol min -1 g-i over this period indicating that sources other than the intracellular glutamine pool were also making large contributions towards the observed increase in glutamine output. In muscle, glutamine can be synthesized from the branched chain amino acids, from glutamate and aspartate all of which are present in muscle protein (the BCAA contributing 2070 of the amino acid residues) and whose muscle free concentrations increase during dexamethason¢ treatment (indicating no shortage of precursor substrates for glutamine synthesis during glucocorticoid treatment). However, calculating on the basis that glutamine accounts for between 40 to 6070 of the observed nitrogen efflux from muscle, net protein breakdown (which was equivalent to 34 m g / g muscle over 8 days) could account for only 2270 (8-12 nmol glutamine/min per g muscle) of the total observed output (54 nmol min-i g-i). The strong implication of these calculations is that dexamethasone has increased net glutamine synthesis in muscle but this alone could not explain the increase in glutamine output since the simultaneous marked fall in the sarcolemmal distribution ratio of glutamine would tend to retard net glutamine ¢fflux from muscle. The accelerated sarcolemmal glutamine exchange reported here would therefore appear to be a prerequisite for increased release of muscle glutamine under certain pathophysiological circumstances. It has recently been reported [26,27] that dexamethasone increases glutamine synthetase activity in rat skeletal muscle when it is administered at doses above 0.5 mg kg-i day-1 [26]. Induction of glutamine synthetase activity appears to be at the transcriptional level since there is an overall increase in the level of mRNA encoding the enzyme [26] with a resultant increase in the number of enzyme molecules (i.e., a Vma~ effect). We therefore investigated whether glutamine synthetase activity can be stimulated in muscle extracts of dexa-
methasone treated rats at saturating glutamate concentrations (given that the K m for glutamate is - 2 mM the reaction rates measured are likely to reflect the maximal enzyme capacity in vitro). Under these conditions we observed no significant changes in the total activities of glutamine synthetase and glutaminase in muscle after dexamethasone-treatment; although when expressed on the basis of total muscle protein there was an indication of a slight rise in glutamine synthetase activity (of approx. 2070). Therefore increased net flux through glutamine synthetase in our studies appears to be due largely to local activation of enzyme in vivo, rather than de novo enzyme synthesis. Induction of glutamine synthetase mediated via changes in transcriptional events may, however, be an additional means by which skeletal muscle may increase its glutamine production in response to dexamethasone treatment. It is central to our argument that changes in the rate of muscle glutamine synthesis (by whatever mechanism) cannot account for the fact that the intramuscular glutamine pool falls after dexamethasone treatment (i.e., the lowered sarcolemmal glutamine gradient occurs despite an apparent increase in muscle synthesis of glutamine). It appears, therefore, that the sarcoplasmic glutamine pool is not simply a store of glutamine. We have suggested [11] that the high intramuscular glutamine concentration is maintained by concentrative uptake of glutamine across the sarcolemma via N a + / glutamine cotransport. The distribution ratio of glutamine across the sarcolemma may therefore be dependent to a large extent on the electrochemical Na + gradient across this membrane. Our results extend the available information on the effects of corticosteroid on muscle by demonstrating a rise of intramuscular Na + concentration, although we were unable to detect any associated change in the muscle membrane-potential, which w a s ' - - 8 0 mV in leg extensor muscle (EDL) isolated from hindlimbs of dexamathasone-treated and control rats (unpublished data of Ward, Hundal and Taylor). A rise in intracellular Na + concentration has also been observed in other tissues exposed to corticosteroid [28,29] and may possibly be related to increased 'leakiness' of the sarcolemma to Na + and also to modulation of the activity of the N a + / K + pump. We have previously shown that a reduction in the inward Na +electrochemical gradient which is associated with a fall in the maximal capacity of glutamine transport (Vmax) into muscle but with no significant change in transport g m [4]. Our recent studies in sarcolemmal vesicles [30] and in the perfused rat hindlimb [4] demonstrate that sarcolemmal glutamine transport falls when the Na gradient is dissipated by either (a) replacement of the inwardly directed concentration gradient of NaC! with KCI or choline chloride, (b) depolarisation of the membrane with the cation ionophore valinomycin or (c) denervation of hindlimb muscle. In denervated muscle
382 [4] there is a fall in the Na gradient (muscle Na increases by 26~: similar to that in the present studies), resting sarcolemmal potential is depolarised (by 15 mV) and unidirectional glutamine transport falls by 40%. This depression in membrane transport can be reversed by partly restoring the Na gradient by exposure of d e n e ~ a t e d muscle membranes to the Na channel blocker ~ X (this facilitates reduction in intramuscular Na through a reduction in membrane Na permeability and extrusion of muscle Na by the Na pump). These observations are consistent with the notion that changes in muscle glutamine distribution during catabolic episodes may be closely associated to changes in muscle Na distribution, The present results also show an association between reductions in transport Vm.x and the Na + gradient, although with an additional reduction in transport Km. It is therefore possible that the K m effect is unrelated to the change in Na + gradient, and may be the result either of changes in membrane fluidity (which may affect the operation/kinetic characteristics of membrane proteins) or of direct dexamethasone action on specific membrane proteins (see below). Dexamethasone treatment results in an increased fractional turnover of the muscle glutamine pool from 0.026/min to 0.074/min (fractional turnover is calculated as sarcolemmal glutamine efflux/muscle glutamine concentration). The accelerated sarcolemmal fluxes of glutamine could result from a change in the exchange properties of the System N m transporter as other kinetic characteristics of this System are altered by dexamethasone treatment. We have recently demonstrated that System N m in sarcolemmal vesicles is capable of mediating Na + dependent glutamine transport into and out of muscle [30], but we cannot exclude a number of other possibilities such as (a) the existence of two cartiers with similar substrate specificities that may be differentially modulated by corticosteroids and (b) alterations in the V m a J K m ratio for glutamine efflux (e.g,, by trans, stimulation of the efflux mechanism, alIosteric modification of transporter molecules or alterations in membrane lipid properties). We have discussed our results, in vivo, in terms of specific effects of dexamethasone itself rather than of possible concurrent side-effects such as alterations in the concentrations of other hormones (e.g., insulin, glueagon). The fact that we have demonstrated acute effects of corticosteroid on the Vm,~ and K m of glutamine transport (i.e., when dexamethasone was added to the pcrfusate of hindlimbs of control rats) indicates that we ought to be entitled to interpret the data, at least partly, in terms of direct glucocorticoid action. Other acute effects (within 0.5-1 h) of dexamethasone have been observed in incubated preparations of rabbit digit extensor muscles (on muscle protein synthesis) [31]. The short term effects may reflect rapid cell penetration by the steroid (and subsequent effects on intracellular
metabolism) a n d / o r modification of membrane phospholipids through alterations in the activities of phospi~olipases (e.g., phospholipase A 2)The dexamethasone effects on glutamine transport kinetics appear to be very rapid and therefore they may play an important role in the rapid alterations of amino acid balance in muscle tissue observed after injury and the onset of disease. However the changes in the kinetic characteristics of the N m-transporter persist (for at least 8 days in this study) in their effect and will predispose any damaged muscle to an increase in the loss of amino acid nitrogen. Furthermore, the kinetic characteristics of splanchnic glutamine uptake [24,32,25] are such as to make the rate of hepatic ureagenesis proportional to the amino acid supply [33]. Thus, interactions between the muscle and splanchnic mechanisms may help explain why the increased provision of amino acids to severely injured and septic patients does not result in the immediate resumption of positive nitrogen balance in muscle or the whole-body [2]. The increase in intracellular Na + of rat skeletal muscle after dexamethasone treatment may increase intramuscular Ca 2+ by inhibition of cation-exchange mechanisms, which in turn may affect numerous metabolic processes including protein turnover. It is noteworthy that in many circumstances of muscle wasting associated with trauma, in which there is likely to be increased corticosteroid secretion, intramuscular Na + (and occasionally Ca 2+) concentration is elevated [5,8]. In addition, an alteration in the size of the intramuscular glutamine pool may itself affect muscle metabolism. Since glutamine appears to be an essential amino acid for the growth of cells in tissue culture [34,35] it occurred to us that it may have an influence on skeletal muscle protein synthesis. We have discovered that there is indeed a good correlation between skeletal muscle protein synthetic rate and intramuscular glutamine concentration both in vivo [36] and in the perfused rat hindlimb [37]. There is also evidence that glutamine reduces protein breakdown in cultured skeletal muscle [38] and we have also observed glutamine inhibition of protein breakdown by an insulin-independent mechanism in the perfused rat hindlimb [39]. More recently we have been able to demonstrate that the administration of glutamine (in the form of the dipeptide alanyiglutamine) to normal healthy subjects results in a stimulation of muscle protein synthesis [40]. The modulation of sarcolemmal transport properties (e.g., by hormones or other effectors as a result of injury or disease) could therefore influence the metabolic behaviour of muscle and predispose it towards increased loss of amino acids, particularly glutamine, to the rest of the body. Since glutamine appears to be important for the normal functioning of white cells, and dividing cells in general [35], this apparently pathophysiological behaviour may have some benefits.
383
Acknowledgements We are grateful to Dr. M.R. Ward for making the membrane potential measurements. This work was supported by Action Research for the Crippled Child, Ajinomoto GmbH, British Diabetic Association, The Medical Research Council, Pfrimmer-Kabi GmbH, The Rank Prize Funds, The Wellcome Trust and The University of Dundee. References 1 Cuthbertson, D.P. (1971) Q. J. Med. 25, 233-245. 2 Rennie, M.J. (1985) Br. Med. Bull. 41,257-264. 3 Ftirst, P., Elwyn, D.H., Askanazi, J., Kinney, J.M. (1982) in Clinical Nutrition: (Wesdorp, R.I.C. and Soeters, P.B., eds.), pp. 10-17, Churchill Livingstone, Edinburgh. 4 Hundal, H.S., Babij, P., Watt, P.W., Ward, M.R. and Rennie, M.J. (1990) Am. J. Physiol. 259, E148-E154. 5 Bergstrom, J., FUrst, P., Holmstrom, B.O., Vinnars, E., Askanazi, J,, Elwyn, D.H., Michelson, C.B. and Kinney, J.M. (1981) Ann. Sur8. 193, 810-816. 6 Cunningham, J.N., Jr., Carter, N.W., Rector, F.C., Jr. and Seldin, D.W. (1971) J. Clin. Invest. 50, 49-59. 7 Trunkey, D.D., lllner, H., Wagner, 1.Y. and Shires, F.T. (1973) Surgery 74, 241-250. 8 Turinsky, J. and Gonnerman, W.A. (1982) J. Surg. Res. 33, 337344. 9 Moore, F.D. (1957) Recent Prog. Horm. Res. 13, 511-576. 10 Wilmore, D.W. (1986) Clin. Nutr. 5, 9-19. 11 Hundal, H.S., Rennie, M.J. and Watt, P.W. (1987) J. Physiol. 393, 283-305. 12 Lund, P. (1970) Biochem. J. 118, 35-39. 13 Huang, Y-Z. and Knox, W.E. (1976) Enzyme 21,408-426. 14 Moore, R.D., Munford, J.W. and Pillsworth, T.J. (1983) J. Physiol. 338, 277-294. 15 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 16 Burton, R.C. and Gorewitt, R.C. (1984) Med. Electronics 15, 68-73. 17 Rothwell, N.J. and Stock, MJ. (1981) Pflugers Arch. 389, 239-242. 18 Bergmeyer H.U. (1965) Methods of Enzymatic Analysis, Verlag Chemie-Academic Press, New YorK.
19 Tsanev, R. and Markov, G.G. (1960) Biochim. Biophys. Acta. 42, 442-452. 20 Yudilevich, D.L. and Mann, G.E. (1982) Fed. Proc. 41, 3045-3053. 21 Hundal, H.S., Rennie, M.J. and Watt, P,W. (1989) J. Physiol. 408, 93-114. 22 Price, N.C., Stevens, L. (1982) Fundamentals of Enzymology, Oxford University Press, Oxford. 23 Muhlbacher, F., Kapadia, C.R., Colpoys, M.F., Smith, R.J. and Wilmore, D.W. (1984) Am. J. Physiol. 247, E75-E83. 24 Kilberg, M.S., Handlogten, M.E. and Christensen, H.N. (1980) J. Biol. Chem. 255, 4011-4019. 25 Taylor, P.M. and Rennie, M.J. (1987) FEBS Lett. 221,370-374. 26 Max, S.R., Mill, J., Mearow, M., Konagaya, M., Konagaya, Y., Thomas, J.W., Banner, C. and Vitkovic, L. (1988) Am. J. Physiol. 255, E397-E403. 27 Tischler, M.E., Henriksen, EJ. and Cook, P.H. (1988) Muscle Nerve 11,752-756. 28 Rayson, B. and Gupla, g. (1985) J. Biol. Chem. 260, 12740-12743. 29 Klein, L.E., Barlolomei, M. and Lo, C.S. (1987) Am. J. Physiol. 249, H570-575. 30 Ahmed, A., Taylor, P.M and Rennie, M.J. (1990) Am. J. Physiol. 259, E284-E291. 31 Reeds, P.J. and Palmer, R.M. (1984) Biochem. J. 219, 953-957. 32 Joseph, S.K., Bradford, N.M. and McGivan, J.D. (1978) Biochem. J. 176, 827-836. 33 Meijer, A.J., Lof, C., Ramos, I.C. and Verhoeven, A.J. (1985) Eur. J. Biochem. 148, 189-196. 34 Kovacevic, I.C.S. and McGivan, J.D. (1983) Physiol. Rev. 63, 547-605. 35 Newsholme, E.A., Crabtree, B. and Ardawi, M.S.M. (1985) Q. J. Exp. Physiol. 70, 473-489. 36 Rennie, M.J., Hundal, H.S., Babij, P., MacLennan, P., Taylor, P.M., Watt, P.W., Jepson, M.M. and Millward, D.J. (1986) Lancet ii, 1008-1012. 37 MacLennan, P.A., Brown, R.A. and Rennie, M.J. (1987) FEBS Left. 215, 187-191. 38 Smith, R.J. (1985) in Intracellular protein catabolism (Khairallah, E.A., Bond, J., Bird, J.D., eds.) pp. 633-635, Alan R. Liss, New York. 39 MaeLennan, P.A., Smith, K., Weryk, B., Watt, P.W. and Rennie, M.J. (1988) FEBS Left. 237, 133-136. 40 Barua, J.M., Smith, K., Wilson, E., Scrimgeour, C. and Rennie, M.J. (1991) Diabetologica Lalina, in press.