- Email: [email protected]
Regulation of flux through glutaminase and glutamine synthetase in isolated perfused rat liver
Btochtmwa et Btop/~vs.a ,4
272
Elsevier Biomedical Pre,~ BBA 21355
REGULATION O F FLUX T H R O U G H GLUTAMINASE AND G L U T A M I N E S Y N T H E T A S E IN ISOLATED PERFUSED RAT LIVER DIETER H A U S S I N G E R a. W O L F G A N G G E R O K a and H E L M U T SIES b
" Mediztmsche Khnik, Hugstetterstrasse 55, Unlversttat Fretburg, D- 7800 Fretburg, and h Instttut fur Ph~ stologt~che Chemle 1, UmverTttat Diisseldorf, Dasseldorf (F R. G.) (Received August 25th, 1982)
Key words. Glutammase," Glutamlne synthetase, Glutamme fluff, (Rat hver)
1. Glutaminase and glutamine synthetase are simultaneously active in the intact liver, resulting in an energy consuming cycling of glutamine at a rate up to 0.2 pmol per g per min. 2. An increase in portal glutamine concentration was followed by an increased flux through glutaminase, but flux through glutamine synthetase remained unchanged. Glutaminase flux was also increased by ammonium ions or glucagon; these effects were additive. 3. Glutamine synthetase flux was increased by ammonium ions, but this activation was partly overcome by increasing portal glutamine concentrations. Glutamine synthetase flux was slightly increased by glucagon at portal glutamine concentrations of about 0.2-0.3 mM, but was strongly inhibited above 0.6 mM. 4. During experimental metabolic acidosis there was an increased net release of glutamine by the liver, being due to opposing changes of flux through glutaminase and glutamine synthetase. Conversely, an increased glutamine uptake by the liver during metabolic alkalosis was observed due to an inhibition of glutamine synthetase and an activation of glutaminase. However, the two enzyme activities respond differently depending on whether glucagon or ammonium ions are present.
Introduction Depending on the metabolic and experimental conditions, contrasting results were obtained concerning net hepatic uptake or release of glutamine [1-6]. Glutamine synthesis in perfused rat liver or isolated hypatocytes was found to be very low [3,7]. Little is known on regulation of glutamine synthetase in the intact liver cell, although isolated rat liver glutamine synthetase has been well characterized [8,9]. Studies on glutaminase activity in perfused rat liver, isolated hepatocytes and isolated mitochondria showed an activation of the enzyme by ammonium ions [10,11], glucagon [12], cyclic AMP [13] and leucine [13], and an inhibition during acidosis [14,15]. Activation of glutaminase by bicarbonate was shown to occur in isolated 0304-4165/83/0000-0000/$03.00 '& 1983 Elsevier Biomedical Press
hepatocytes [13,16] but not in the intact organ [15]. We have previously shown in experiments with [u-Inc]glutamine that glutaminase and glutamine synthetase are simultaneously active in perfused liver, resulting in a cycling of glutamine at the expense of energy [17]. Therefore, it is necessary to determine separately the metabolic rates through glutaminase and glutamine synthetase, if regulation of glutamine metabolism in the intact organ is studied. In the present work this was performed by use of [l-~4C]glutamine and metabolic rates through glutaminase and glutamine synthetase were studied simultaneously in isolated perfused rat liver under the influence of portal glutamine, ammonium ions, glucagon and pH. The results show that fluxes through glutaminase and glutamine synthetase in the intact organ are of
273 similar magnitude, both enzymes being influenced by effectors on glutamine metabolism. This provides a complex but highly sensitive means for regulation of plasma glutamine concentration by the liver. The utility of so-called futile cycles has recently been pointed out [18]. Materials and Methods
Hemoglobin-free hver perfusion. Livers of male Wistar rats of 100-200 g body weight, fed on stock diet (Altromin) ad libitum, were perfused as described previously [17] without recirculation of the perfusate using the bicarbonate-buffered Krebs-Henseleit solution plus 2.1 mM L-lactate, 0.3 mM pyruvate and 2 mM ornithine as sodium salts. Perfusion fluid was equilibrated with 0 2 / CO 2, 95/5 (v/v). Perfusate flow was approx. 4 ml- min- I • g- ~ and was kept constant throughout the individual perfusion experiment. The temperature was 37°C. Perfusate pH changes were performed by addition of HC1 or NaOH by precision micropumps. Inhibition of glutamine synthetase activity in perfused rat liver was achieved by intraperitoneal injection of methione sulfoximine (5 mg per 100 g body weight) 2-3 h prior to the perfusion experiment and by addition of methionine sulfoximine to the perfusion fluid at a concentration of 0.1-0.2 mM. Assays. The concentrations of glutamine and glutamate in the perfusate were measured based on the procedures described in [19]. Labeled CO 2 production from [1-14C]glutamine by perfusated rat liver was measured as described previously [17]. Perfusate containing [I-14C]glutamine (0.5 nCi/ml) was pooled over 2 rain periods and collected in stoppered conical flasks. CO 2 was trapped in phenylethylamine and radioactivity determined by scintillation spectrometry. Rates of CO 2 release were calculated on the basis of the specific radioactivity of glutamine in the perfusate. Preparation of [l-14C]glutamine. L-[I-14C]Glutamine was prepared from [l-14C]glutamate using a crude rat liver cytosol preparation following the procedures described by Baverel and Lund [13], except that the incubation pH was 7.4. After
incubation and deproteinization, the pH was adjusted to pH 5.7 and [1-tac]glutamine was separated from residual [1-~4C]glutamate by column chromatography with Sephadex QAE A-25, equilibrated with Tris-HCl (40 mM) pH 5.7. As carrier medium 10 mM Tris-HCl, pH 5.7 was used. After separation less than 0.1% glutamate was present in the [ 1-14C]glutamine preparation. Chemicals. All enzymes, pyruvate, oxoglutarate, glucagon and NAD were from Boehringer, Mannheim. [l-14C]Glutamate was from Amersham Buchler, Frankfurt. Glutaminase (grade V) was from Sigma, St. Louis. All other chemicals were from Merck, Darmstadt. Results
Flux through glutaminase under the influence of portal glutamine concentration, ammonium ions and glucagon As shown in Fig. 1A, 14CO2 production from [1-14C]glutamine by perfused rat liver increases with increasing portal glutamine supply, indicating that glutaminase is active also at below-physiological glutamine concentrations. In the experiments presented, a portal glutamine concentration of 1 mM corresponds to a glutamine supply of about 4 #moi/min per g liver. Further, it is shown that 14CO2 production from [l-14C]glutamine is unchanged when glutamine synthetase is inhibited by methionine sulfoximine [20]. In this case, 14CO2 production from [1-14C]glutamine fully accounts for net glutamine uptake by the liver. Similarly, when glutaminase is activated by ammonium ions (Fig. 1B), again there is a close agreement of 14CO2 production from [l-14C]glutamine in absence or presence of methionine sulfoximine and net glutamine uptake by the liver during inhibition of glutamine synthetase by methionine sulfoximine. This indicates that blocking of glutamine synthetase is without effect on glutaminase activity (see also Ref. 17). Flux through giutarninase in perfused liver can be determined as 14CO2 production from [l-14C]glutamine, because net glutamine uptake in presence of methionine sulfoximine fully accounts for 14CO2 production from [114C]glutamine. It may be mentioned that accumulation of glutamate occurs only with rates of glutamine
274
ostc
/
0.40.2-
o,_\ ..//"
0.L-- )
o~ g,
0.2-
B
~
0.1 A/oo~o o 0 ::.._./-" f-J~
G[utomlne Addlt~on (#rno[.mln-%g-~) z~
-0J-
0.3
/
o. f
•
.
L
-o:-
-0.2-
-0:-
o\ ~o
-0.3 -
-D'2 t - 0 3
'~
-0.~--0.5
Fig. IA. Effect of portal glutamme addition on 14CO 2 release from [1-14C]glutamlne (closed symbols) and net glutamlne release (open symbols) by perfused rat liver. Data are given for experiments without (O, O) or with (A, zx) methiomne sulfoximlne With methionme sulfoxlmine present, net glutamine release becomes negative, i.e. an uptake. Net glutamine uptake by the liver in presence of methlonlne sulfoxlmme fully accounts for 14CO2 production from [l-J4C]glutamine. Each point represents means of 5 10 separate measurements during metabolic steady states of the indiwdual perfus~on experiment. Data are from I 1 perfuslon e~perlments. On average, a glutamlne addiuon of 4 ~ m o l / m i n per g corresponds to a glutamme concentration of about 1 mM in lnfluent perfusate. B. Effect of portal glutamine addmon on 14C02 release from [l-14C]glutamine and net glutamlne release in presence of ammonium chloride (0.6 mM). Data are from 9 different different perfusion experiments. Symbols as in A. C Effect of portal glutamine addition on 14CO2 release from [l-[4C]glutamine (O) and net glutamine release by the liver (©) in presence of glucagon ( 1 0 - 7 M ) , Data are from 12 different perfusion experiments. Compare with A.
TABLE 1 EFFECT OF G L U C A G O N A N D A M M O N I U M IONS ON FLUX T H R O U G H GLUTAMINASE AND G L U T A M I N E SYNTHETASE IN ISOLATED PERFUSED RAT LIVER Physiological portal glutamme concentration of 0.6 mM led to glutanune additions shown in first column. Glutammase activity was measured as 14CO2 release from [1)4C]glutamine. Activity of glutarmne syntbetase was determined as ~4CO2 release from [1-14C]glutamlne plus net glutamine release by the liver. Net glutamate release by the liver into the perfusate was about 60 nmol/mln per g under all conditions. Data are given as means 5: S.E. and the number of different perfuslon experiments in parentheses. Cycling of glutamine represents flux through the enzyme with the lower flux. With glucagon present, glutamlne synthetase flux is lower than glutaminase flux. However, in one experiment (not included m table) glutamlnase flux was somewhat lower than glutamlne synthetase flux w~th glucagon present. Except for glutamine supply, values are n m o l / m l n per g. Addition of modifier
Glutamlne supply (p~mol/min per g)
Net glutamme release
Glutamlnase activity
Glutamine synthetase actevity
Rate of the cycling of glutamme
Control (9) Glucagon, 10 -7 M (7) NH4C1, 0.6 mM (7) Glucagon, ( 10-7 M) plus NH4CI, 0.6 mM (5) Control (8) Glucagon, 10 -7 M (7) NH4CI, 0.6 mM (8)
2.51 +0.08 2.57+0 12 2.37+0 10
80+ 11 - 147+30 80+ 8
72+ 4 231 + 32 168+__ 9
151__+ 8 84 + 20 248+12
72+ 4 84 + 20 168+ 9
2.47+0.11 none none none
-254+33 147+- 8 145+- 9 381 + 17
379 + 36
122+27
122+__27
275
supply above 10/~mol/min per g, i.e., conditions when ammonia production from glutamine exceeds alanine production from glutamine [15], and even at such high rates 14CO2 production from [lD4C]glutamine accounts for 80-90% of the flux through glutaminase [21]. Concomitant with 14CO2 production from [1~4C]glutamine, there is a net glutamine release by the liver (Fig. IA). This net glutamine release by the liver is due to the activity of glutamine synthetase, because it can be completely blocked by methionine sulfoximine with or without added portal glutamine (Fig. lA, B). When ammonium ions (0.6 mM) are present (Fig. 1B) there is also a net glutamine release by the liver. This is decreased at increasing portal glutamine additions. With glutamine additions of more than 3 /~mol/ min per g there is a net glutamine uptake by the liver.
Glucagon also increases 14CO2 production from [1-14C]glutamine, i.e. glutaminase activity, and similarly there is a net glutamine release by the liver up to glutamine additions of about 2 ftmol/min per g (Fig. 1C). This turns to a net glutamine uptake at higher glutamine concentrations. Thus, in the physiological portal concentration range of glutamine of 0.5-0.6 mM [3,6,22], corresponding to a glutamine addition of about 2-2.4 #mol/min per g in our perfusion experiments, any excess portal glutamine will be removed effectively in presence of glucagon. At a physiological glutamine concentration of 0.6 mM the activating effects of glucagon and ammonium ions on glutaminase activity are additive, suggesting different sites of action of these activators (Table I). (4}
l/,.] O.~0.~- -
u-~'-
E
0.3-
~0.3x (&)
'7
E o O
E
~ ~0.2x O _~n-
0.2IT
5-~ •
O
0J
0.1
Z
E
01
o
0
I
0
1
I
I
2 3 Glutamlne Addition (iamol x mtn-tx g-I )
I
4
Fig. 2. Effect of portal glutamine on flux through glutamine synthetase. Data are given for absence (O) or presence of glucagon (10-7 M ) (A), or addition of ammonium chloride (0.6 mM) O). Flux Is calculated as t4CO2 release from [lt4C]glutaimne plus net glutamine release. Data are gwen as means _+S.E. (number of different experiments m parenthesis).
i Ammontum Chloride Addttlon (pmol x mln -I x g -I )
Fig. 3. Effect of ammonium chloride addition on the metabolism of added portal glutarmn¢ (0.6 mM). Glutaminase flux (O), glutamine synthetase flux (11), net glutamine production by perfused liver (O). Effect of ammonium chloride addition on glutamine production by perfused rat liver in the absence of added portal glutamine (12). Data are given as means + S.E. (number of different perfusion experiments in parenthesis).
276 t3)
0.2-
A
0.3-
D~D C~"O
:
-g f71
X
'5,;-E_0.1-
17"7 cn
1'-I
~.;~ 0.2
81151
c
c~-~
13
0
[31
-5 ~ O.l
o
~
~
;
L
Glulamme Addfflon (pmol x mln -1 x g-1 )
Fig. 4. Effect of glutamme o n the cychng of glutamme without ( © ) or with glucagon (10 v M) (zx) or a m m o m u m chloride (0.6 mM) (e). Data are from the experiments shown in Fig. 2.
o
E~ ~5 5~
~°
\
J i
i
7.0
7.2
i
~ I
7.z~ 7.~
I
i
7.8 8.0 pH ,n Effluent
",,..o -0.1
These data (Fig. IA-C, Table I) show that slight changes in portal glutamine concentration or release of glucagon into portal circulation will cause significant changes of net glutamine metabolism by the liver. This provides a basis for explanatton of the conflicting results obtained in the literature concerning portal-hepatic vein concentration differences of glutamine [1-6] and thus the role of the liver in glutamine metabolism. Flux through glutamme svnthetase under the influence of portal glutamine concentration, ammomum tons and glucagon Flux through glutamine synthetase ,s calculated as net glutamine release by the liver plus flux through glutaminase assayed as ~4CO2 production from [l-14C]glutamine. As shown in Table I, at a physiological glutamine concentration of 0.6 mM and in absence of further effectors on glutamine metabolism, flux through glutamine synthetase is about twice as high as that through glutaminase. Further, glutamine synthetase is activated by ammonium ions and inhibited by glucagon. However, the effects of glucagon and ammonium ions on glutamine synthetase are dependent on portal glutamme concentration (Fig. 2). Variation of portal glutamine concentration in the range of 0-1 mM is nearly without effect on glutamine synthetase flux in the absence of glucagon or ammonium ions. In presence of glucagon, however. there is an activation of glutamine synthetase at
B 0.3.-z o
1:3
u_.
~x
~_c ~E c .
Q
0.2-
0,1
E 0
pH tn Effluent
_~ c~_0.1 o
o
c_~5 Eu~ - 0 . 2 (.9 Z
-0.3
F,g. 5. Effect of perfusate pH on net glutamme release (©1. and flux through glutaminase (*) or glutamme synthetase (1:3) A. LLvers perfused with 0.6 mM glutamine plus 0.6 mM ammonium chloride. Perfusate changes were performed by addition of HCI or NaOH. (pH,. n - p H e m ) was 0.18 on average m stmdar experiments. B. Livers perfused with 0.6 mM glutamme plus glucagon (10 -~ M). Each point represents 5-9 separate measurements during metabohc steady states, Data are from six different perfuslon experiments m each of the A and B.
277 low portal glutamine additions, but a strong inhibition of the enzyme when portal glutamine concentration is raised to above-physiological. With no added glutamine, however, glucagon is without effect on glutamine synthesis (Fig. 2, Table I). With added ammonium ions, flux through glutamine synthetase is substantially increased, but the effect is more pronounced at low portal glutamine concentrations (Fig. 2). This may be due to product inhibition of glutamine synthetase, as was found for isolated rat liver glutamine synthetase in presence of manganese ions [9]. Portal ammonia concentration is reported to be about 0.2-0.3 mM [3,23] and may rise up to 1 mM in ammonia toxicity. Therefore, the effect of portal ammonia concentration on the activities of glutaminase and glutamine synthetase was studied with and without added glutamine at 0.6 mM (Fig. 3). With no added glutamine, there is a sigmoidal increase of glutamine synthesis upon addition of ammonium ions at 0.5-4 ~ m o l / g per min, possibly reflecting a competition of urea and glutamine synthesis for ammonia. Glutamine synthesis is maximal at addition of ammonium ions above 3 t~mol/min per g, i.e. 0.7-0.8 mM in influent. However, with 0.6 mM glutamine added, a maximal glutamine synthetase activity of about 0.25 /~mol/min per g is already observed at an addition of ammonium ions of about 0.4/~mol/min per g. This shows that supply of ammonium ions for glutamine synthesis is not rate-limiting with physiological portal ammonia and glutamine concentrations. Further it is shown in Fig. 3 that glutaminase activity is also increased by ammonium ions. However, this effect is half-maximal at ammonium chloride addition of about 1 /~mol/min per g, and maximal at 2.5-3 ~ m o l / m i n per g. Despite the activity of glutaminase, net glutamine production by the liver with or without added glutamine will account for 50% of ammonia uptake by the liver, if portal ammonia concentration is in the range of about 0.1 mM (Fig. 3).
Cycling of glutamine m the glutaminase / glutamme synthetase system The simultaneous flux through glutaminase and glutamine synthetase in the liver results in a cycling of glutamine at the expense of energy. An
increase of O z uptake by the liver after addition of glutamine, not accounted for energy requirements for urea synthesis from glutamine, has been demonstrated [17]. The rate of this cycle will be determined by flux through the enzyme with the lower activity. Fig. 4 shows that the rate of cycling of glutamine depends on portal glutamine addition and on glucagon and ammonium ions. Interestingly, the rate of cycling is near maximal within the physiological concentration range of glutamine of 0.5-0.6 mM but decreases at higher concentration in the case of glucagon.
Effect of perfusate pH on the simultaneous activities of glutaminase and glutamine synthetase Experimental acidosis, induced by addition of HC1 to the perfusion fluid, leads to an increased net glutamine release by the liver (Fig. 5A, B). With ammonium ions present, this net glutamine release is due to an inhibition of glutaminase activity, whereas glutamine synthetase activity is almost unchanged. In presence of glucagon, however, net glutamine release by the liver is matched by an inhibition of glutaminase as well as an activation of glutamine synthetase. Conversely, increased net glutamine uptake during metabolic alkalosis is due to an inhibition of glutamine synthetase in presence of ammonium ions. With glucagon, again activity changes of both enzymes to a similar extent but in opposite directions bring about increased net glutamine uptake. From Fig. 5A, B it is also evident that maximum rates of the cycling of glutamine are observed at physiological pH values. Discussion
Regulation of flux through glutaminase and glutamine synthetase in the intact liver The flux through glutaminase and glutamine synthetase can be monitored simultaneously in perfused rat liver by measuring 14CO2 release from [1-¿4C]glutamine and glutamine release (when negative:uptake) by the liver. The flux throt~gh both enzymes is effectively modified by the portal glutamine and ammonia concentrations, by glucagon and perfusate pH. The most marked effects are observed at physiological concentrations of these modifiers.
278
The observed response of the two enzymes to acidosis should favour renal acid excretion, by increasing the supply with glutamine from liver, in addition to that coming from muscle [24].
Posstble slgntficance of the cychng of glutamine The simultaneous activities of glutaminase and glutamine synthetase result in a cycling of glutamine at the expense of energy. Such cycles are called futile cycles and have been established to occur in glucose metabolism and their significance for regulatory problems has been elucidated (see Refs. 18, 25). A significance of the glutamine cycling for effective regulation of blood glutamine concentration is considered likely, because maximum rates of the cycle are obtained in presence of a physiological glutamine concentration and at physiological pH values. Glutamine cycling may have a useful purpose because the combined action of glutamine synthetase, a cytosolic enzyme [9], and glutaminase, a mitochondrial enzyme [26], will bring about a net transport of glutamate and ammonia from the cytosolic into the mitochondrial compartment. This is supported by the experiments with and without methionine sulfoximine (Fig. IA, B) showing no utilization of labeled glutamate derived [l-14C]glutamine in the mitochondria by the cytosolic glutamine synthetase reaction. Although two potent glutamate transport systems across the mitochondrial membrane are known (for review see Ref. 27), glutamine cycling could provide a mechanism of glutamate transport from the cytosolic into the mitochondrial space at the expense of ATP, regardless of the mitochondrial/cytosolic glutamate concentration gradient. Interestingly, the highest rates of the cycle are observed during ureogenesis from ammonium ions. As shown in Table I addition of ammonium ions does not affect net glutamine release by the liver and thus hepatic vein glutamine concentration. The rate of cycling, however, is increased about 2-3-fold. Acknowledgements
Expert technical assistance was provtded by H. Weigand and P. Graf. This study was supported by Deutsche Forschungsgemeinschaft, Forschergruppe "Leber", Grant De 113/26-3, and by Schwerpunktsprogramm "Regulationsmechanis-
men des Kohlenhydrat- und Lipidstoffwechsels", Grant Si 255/6-2. References 1 Schrock. H. and Goldstem, L. 11981) Am. J. Physiol. 240, E 519-E 525 2 Lund, P (1980) FEBS Lett. 117, K 8 6 - K 92 3 Lund, P and Wafford, M. (19761 m The Urea Cycle (Gnsolia, S. Baguena. R and Mayor, F., eds ), pp. 479-488, Wiley & Sons, London 4 Atkawa, T., Matsutaka, H., Yamamoto, H., Okuda, T., Ishikawa, E., Kawano. T. and Matsumura, E. (1973) J. Blochem. 74, 1003-1017 5 R~m6sy, C., Demign~, C and Aufrere. J. 119781 Blochem. J. 170, 321-329 6 Yamamoto, H., Alkawa, T., Matsutaka, H., Okuda, T. and Ishlkawa, E. (19741 Am. J. Physiol. 226, 1428-1433 7 Lurid, P. 11971) Btochem. J. 124, 653-660 8 Tare, S.S. and Metster, A 11971) Proc. Natl. Acad Sc~ U S A. 68, 781-785 9 Deuel, T . F , Lome, M. and Lerner, A. 11978) J. B~ol. Chem 253, 6111-6118 10 Charles, R. 11968) Ph D Thesis, Amsterdam, Rototype, Amsterdam I1 Haussinger, D. and Stes, H. (1975) Abstr. Commun. 10th Meet. Fed. Eur Blochem. Soc. No 1497 12 Joseph, S.K and McGlvan, J.D. (1978) Btochlm Blophys Acta 543, 16-28 13 Baverel, G. and Lund, P 11979) Blochem. J. 184. 599 606 14 Lueck, J D and Miller, L.L (19701 J. Biol. Chem. 245, 5491 - 5497 15 Haussmger, D., Akerboom, T.P.M. and Stes, H. 119801 Hoppe-Seyler's Z. Phystol Chem. 361,995-1001 16 Joseph, S K. and McGivan, J.D 11978) Blochem. J 176. 837-844 17 Haussmger, D. and Sles. H 119791 Eur. J. Btochem 101, 179 184 18 Reich, J.G. and Sel'Kov, E E (19811 Energy Metabolism of the Cell, Academic Press, London 19 Lund, P. (1974) m Methoden der Enzymatlschen Analyse (Bergmeyer, H.U., ed.I, 3rd Edn, pp 1767-1769, Verlag Chemle, Wemhe~m 20 Melster, A (1974) m The Enzymes (Boyer, P.D., ed.), 3rd Edn., Vol. 10, pp. 699-754, Academic Press, New York 21 Haussmger, D., Gerok, W. and Stes, H. 119821 Eur J B~ochem. 126, 69-76 22 Du Rmsseau, J.P., Greenstem, J.P. Wtmtz, M. and Blrnbaum, S.M. (1975) Arch. Biochem. Btophys. 68, 161-171 23 / u n d , P., Brosnan, J T and Eggleston, L.V. (19701 m Essays m Cell Metabohsm (Bartley, W., Kornberg, H.L. and Quayle. J.R., eds.), pp 167 188, Wdey lntersclence. London 24 Goldstem, L. (1976) Int. Rev. Physto[ I1,283-316 25 Hue, L. 11982) m Metabohc Compartmentatton (Stes, H , ed.), pp. 71 97, Acadenuc Press, London 26 Kalra, J. and Brosnan, J.T. (1973) FEBS Lett. 37, 325-328 27 La Noue, K.F. and Schoolwerth, A C. 119791 Annu Rev B~ochem 48, 871 922
272
Elsevier Biomedical Pre,~ BBA 21355
REGULATION O F FLUX T H R O U G H GLUTAMINASE AND G L U T A M I N E S Y N T H E T A S E IN ISOLATED PERFUSED RAT LIVER DIETER H A U S S I N G E R a. W O L F G A N G G E R O K a and H E L M U T SIES b
" Mediztmsche Khnik, Hugstetterstrasse 55, Unlversttat Fretburg, D- 7800 Fretburg, and h Instttut fur Ph~ stologt~che Chemle 1, UmverTttat Diisseldorf, Dasseldorf (F R. G.) (Received August 25th, 1982)
Key words. Glutammase," Glutamlne synthetase, Glutamme fluff, (Rat hver)
1. Glutaminase and glutamine synthetase are simultaneously active in the intact liver, resulting in an energy consuming cycling of glutamine at a rate up to 0.2 pmol per g per min. 2. An increase in portal glutamine concentration was followed by an increased flux through glutaminase, but flux through glutamine synthetase remained unchanged. Glutaminase flux was also increased by ammonium ions or glucagon; these effects were additive. 3. Glutamine synthetase flux was increased by ammonium ions, but this activation was partly overcome by increasing portal glutamine concentrations. Glutamine synthetase flux was slightly increased by glucagon at portal glutamine concentrations of about 0.2-0.3 mM, but was strongly inhibited above 0.6 mM. 4. During experimental metabolic acidosis there was an increased net release of glutamine by the liver, being due to opposing changes of flux through glutaminase and glutamine synthetase. Conversely, an increased glutamine uptake by the liver during metabolic alkalosis was observed due to an inhibition of glutamine synthetase and an activation of glutaminase. However, the two enzyme activities respond differently depending on whether glucagon or ammonium ions are present.
Introduction Depending on the metabolic and experimental conditions, contrasting results were obtained concerning net hepatic uptake or release of glutamine [1-6]. Glutamine synthesis in perfused rat liver or isolated hypatocytes was found to be very low [3,7]. Little is known on regulation of glutamine synthetase in the intact liver cell, although isolated rat liver glutamine synthetase has been well characterized [8,9]. Studies on glutaminase activity in perfused rat liver, isolated hepatocytes and isolated mitochondria showed an activation of the enzyme by ammonium ions [10,11], glucagon [12], cyclic AMP [13] and leucine [13], and an inhibition during acidosis [14,15]. Activation of glutaminase by bicarbonate was shown to occur in isolated 0304-4165/83/0000-0000/$03.00 '& 1983 Elsevier Biomedical Press
hepatocytes [13,16] but not in the intact organ [15]. We have previously shown in experiments with [u-Inc]glutamine that glutaminase and glutamine synthetase are simultaneously active in perfused liver, resulting in a cycling of glutamine at the expense of energy [17]. Therefore, it is necessary to determine separately the metabolic rates through glutaminase and glutamine synthetase, if regulation of glutamine metabolism in the intact organ is studied. In the present work this was performed by use of [l-~4C]glutamine and metabolic rates through glutaminase and glutamine synthetase were studied simultaneously in isolated perfused rat liver under the influence of portal glutamine, ammonium ions, glucagon and pH. The results show that fluxes through glutaminase and glutamine synthetase in the intact organ are of
273 similar magnitude, both enzymes being influenced by effectors on glutamine metabolism. This provides a complex but highly sensitive means for regulation of plasma glutamine concentration by the liver. The utility of so-called futile cycles has recently been pointed out [18]. Materials and Methods
Hemoglobin-free hver perfusion. Livers of male Wistar rats of 100-200 g body weight, fed on stock diet (Altromin) ad libitum, were perfused as described previously [17] without recirculation of the perfusate using the bicarbonate-buffered Krebs-Henseleit solution plus 2.1 mM L-lactate, 0.3 mM pyruvate and 2 mM ornithine as sodium salts. Perfusion fluid was equilibrated with 0 2 / CO 2, 95/5 (v/v). Perfusate flow was approx. 4 ml- min- I • g- ~ and was kept constant throughout the individual perfusion experiment. The temperature was 37°C. Perfusate pH changes were performed by addition of HC1 or NaOH by precision micropumps. Inhibition of glutamine synthetase activity in perfused rat liver was achieved by intraperitoneal injection of methione sulfoximine (5 mg per 100 g body weight) 2-3 h prior to the perfusion experiment and by addition of methionine sulfoximine to the perfusion fluid at a concentration of 0.1-0.2 mM. Assays. The concentrations of glutamine and glutamate in the perfusate were measured based on the procedures described in [19]. Labeled CO 2 production from [1-14C]glutamine by perfusated rat liver was measured as described previously [17]. Perfusate containing [I-14C]glutamine (0.5 nCi/ml) was pooled over 2 rain periods and collected in stoppered conical flasks. CO 2 was trapped in phenylethylamine and radioactivity determined by scintillation spectrometry. Rates of CO 2 release were calculated on the basis of the specific radioactivity of glutamine in the perfusate. Preparation of [l-14C]glutamine. L-[I-14C]Glutamine was prepared from [l-14C]glutamate using a crude rat liver cytosol preparation following the procedures described by Baverel and Lund [13], except that the incubation pH was 7.4. After
incubation and deproteinization, the pH was adjusted to pH 5.7 and [1-tac]glutamine was separated from residual [1-~4C]glutamate by column chromatography with Sephadex QAE A-25, equilibrated with Tris-HCl (40 mM) pH 5.7. As carrier medium 10 mM Tris-HCl, pH 5.7 was used. After separation less than 0.1% glutamate was present in the [ 1-14C]glutamine preparation. Chemicals. All enzymes, pyruvate, oxoglutarate, glucagon and NAD were from Boehringer, Mannheim. [l-14C]Glutamate was from Amersham Buchler, Frankfurt. Glutaminase (grade V) was from Sigma, St. Louis. All other chemicals were from Merck, Darmstadt. Results
Flux through glutaminase under the influence of portal glutamine concentration, ammonium ions and glucagon As shown in Fig. 1A, 14CO2 production from [1-14C]glutamine by perfused rat liver increases with increasing portal glutamine supply, indicating that glutaminase is active also at below-physiological glutamine concentrations. In the experiments presented, a portal glutamine concentration of 1 mM corresponds to a glutamine supply of about 4 #moi/min per g liver. Further, it is shown that 14CO2 production from [l-14C]glutamine is unchanged when glutamine synthetase is inhibited by methionine sulfoximine [20]. In this case, 14CO2 production from [1-14C]glutamine fully accounts for net glutamine uptake by the liver. Similarly, when glutaminase is activated by ammonium ions (Fig. 1B), again there is a close agreement of 14CO2 production from [l-14C]glutamine in absence or presence of methionine sulfoximine and net glutamine uptake by the liver during inhibition of glutamine synthetase by methionine sulfoximine. This indicates that blocking of glutamine synthetase is without effect on glutaminase activity (see also Ref. 17). Flux through giutarninase in perfused liver can be determined as 14CO2 production from [l-14C]glutamine, because net glutamine uptake in presence of methionine sulfoximine fully accounts for 14CO2 production from [114C]glutamine. It may be mentioned that accumulation of glutamate occurs only with rates of glutamine
274
ostc
/
0.40.2-
o,_\ ..//"
0.L-- )
o~ g,
0.2-
B
~
0.1 A/oo~o o 0 ::.._./-" f-J~
G[utomlne Addlt~on (#rno[.mln-%g-~) z~
-0J-
0.3
/
o. f
•
.
L
-o:-
-0.2-
-0:-
o\ ~o
-0.3 -
-D'2 t - 0 3
'~
-0.~--0.5
Fig. IA. Effect of portal glutamme addition on 14CO 2 release from [1-14C]glutamlne (closed symbols) and net glutamlne release (open symbols) by perfused rat liver. Data are given for experiments without (O, O) or with (A, zx) methiomne sulfoximlne With methionme sulfoxlmine present, net glutamine release becomes negative, i.e. an uptake. Net glutamine uptake by the liver in presence of methlonlne sulfoxlmme fully accounts for 14CO2 production from [l-J4C]glutamine. Each point represents means of 5 10 separate measurements during metabolic steady states of the indiwdual perfus~on experiment. Data are from I 1 perfuslon e~perlments. On average, a glutamlne addiuon of 4 ~ m o l / m i n per g corresponds to a glutamme concentration of about 1 mM in lnfluent perfusate. B. Effect of portal glutamine addmon on 14C02 release from [l-14C]glutamine and net glutamlne release in presence of ammonium chloride (0.6 mM). Data are from 9 different different perfusion experiments. Symbols as in A. C Effect of portal glutamine addition on 14CO2 release from [l-[4C]glutamine (O) and net glutamine release by the liver (©) in presence of glucagon ( 1 0 - 7 M ) , Data are from 12 different perfusion experiments. Compare with A.
TABLE 1 EFFECT OF G L U C A G O N A N D A M M O N I U M IONS ON FLUX T H R O U G H GLUTAMINASE AND G L U T A M I N E SYNTHETASE IN ISOLATED PERFUSED RAT LIVER Physiological portal glutamme concentration of 0.6 mM led to glutanune additions shown in first column. Glutammase activity was measured as 14CO2 release from [1)4C]glutamine. Activity of glutarmne syntbetase was determined as ~4CO2 release from [1-14C]glutamlne plus net glutamine release by the liver. Net glutamate release by the liver into the perfusate was about 60 nmol/mln per g under all conditions. Data are given as means 5: S.E. and the number of different perfuslon experiments in parentheses. Cycling of glutamine represents flux through the enzyme with the lower flux. With glucagon present, glutamlne synthetase flux is lower than glutaminase flux. However, in one experiment (not included m table) glutamlnase flux was somewhat lower than glutamlne synthetase flux w~th glucagon present. Except for glutamine supply, values are n m o l / m l n per g. Addition of modifier
Glutamlne supply (p~mol/min per g)
Net glutamme release
Glutamlnase activity
Glutamine synthetase actevity
Rate of the cycling of glutamme
Control (9) Glucagon, 10 -7 M (7) NH4C1, 0.6 mM (7) Glucagon, ( 10-7 M) plus NH4CI, 0.6 mM (5) Control (8) Glucagon, 10 -7 M (7) NH4CI, 0.6 mM (8)
2.51 +0.08 2.57+0 12 2.37+0 10
80+ 11 - 147+30 80+ 8
72+ 4 231 + 32 168+__ 9
151__+ 8 84 + 20 248+12
72+ 4 84 + 20 168+ 9
2.47+0.11 none none none
-254+33 147+- 8 145+- 9 381 + 17
379 + 36
122+27
122+__27
275
supply above 10/~mol/min per g, i.e., conditions when ammonia production from glutamine exceeds alanine production from glutamine [15], and even at such high rates 14CO2 production from [lD4C]glutamine accounts for 80-90% of the flux through glutaminase [21]. Concomitant with 14CO2 production from [1~4C]glutamine, there is a net glutamine release by the liver (Fig. IA). This net glutamine release by the liver is due to the activity of glutamine synthetase, because it can be completely blocked by methionine sulfoximine with or without added portal glutamine (Fig. lA, B). When ammonium ions (0.6 mM) are present (Fig. 1B) there is also a net glutamine release by the liver. This is decreased at increasing portal glutamine additions. With glutamine additions of more than 3 /~mol/ min per g there is a net glutamine uptake by the liver.
Glucagon also increases 14CO2 production from [1-14C]glutamine, i.e. glutaminase activity, and similarly there is a net glutamine release by the liver up to glutamine additions of about 2 ftmol/min per g (Fig. 1C). This turns to a net glutamine uptake at higher glutamine concentrations. Thus, in the physiological portal concentration range of glutamine of 0.5-0.6 mM [3,6,22], corresponding to a glutamine addition of about 2-2.4 #mol/min per g in our perfusion experiments, any excess portal glutamine will be removed effectively in presence of glucagon. At a physiological glutamine concentration of 0.6 mM the activating effects of glucagon and ammonium ions on glutaminase activity are additive, suggesting different sites of action of these activators (Table I). (4}
l/,.] O.~0.~- -
u-~'-
E
0.3-
~0.3x (&)
'7
E o O
E
~ ~0.2x O _~n-
0.2IT
5-~ •
O
0J
0.1
Z
E
01
o
0
I
0
1
I
I
2 3 Glutamlne Addition (iamol x mtn-tx g-I )
I
4
Fig. 2. Effect of portal glutamine on flux through glutamine synthetase. Data are given for absence (O) or presence of glucagon (10-7 M ) (A), or addition of ammonium chloride (0.6 mM) O). Flux Is calculated as t4CO2 release from [lt4C]glutaimne plus net glutamine release. Data are gwen as means _+S.E. (number of different experiments m parenthesis).
i Ammontum Chloride Addttlon (pmol x mln -I x g -I )
Fig. 3. Effect of ammonium chloride addition on the metabolism of added portal glutarmn¢ (0.6 mM). Glutaminase flux (O), glutamine synthetase flux (11), net glutamine production by perfused liver (O). Effect of ammonium chloride addition on glutamine production by perfused rat liver in the absence of added portal glutamine (12). Data are given as means + S.E. (number of different perfusion experiments in parenthesis).
276 t3)
0.2-
A
0.3-
D~D C~"O
:
-g f71
X
'5,;-E_0.1-
17"7 cn
1'-I
~.;~ 0.2
81151
c
c~-~
13
0
[31
-5 ~ O.l
o
~
~
;
L
Glulamme Addfflon (pmol x mln -1 x g-1 )
Fig. 4. Effect of glutamme o n the cychng of glutamme without ( © ) or with glucagon (10 v M) (zx) or a m m o m u m chloride (0.6 mM) (e). Data are from the experiments shown in Fig. 2.
o
E~ ~5 5~
~°
\
J i
i
7.0
7.2
i
~ I
7.z~ 7.~
I
i
7.8 8.0 pH ,n Effluent
",,..o -0.1
These data (Fig. IA-C, Table I) show that slight changes in portal glutamine concentration or release of glucagon into portal circulation will cause significant changes of net glutamine metabolism by the liver. This provides a basis for explanatton of the conflicting results obtained in the literature concerning portal-hepatic vein concentration differences of glutamine [1-6] and thus the role of the liver in glutamine metabolism. Flux through glutamme svnthetase under the influence of portal glutamine concentration, ammomum tons and glucagon Flux through glutamine synthetase ,s calculated as net glutamine release by the liver plus flux through glutaminase assayed as ~4CO2 production from [l-14C]glutamine. As shown in Table I, at a physiological glutamine concentration of 0.6 mM and in absence of further effectors on glutamine metabolism, flux through glutamine synthetase is about twice as high as that through glutaminase. Further, glutamine synthetase is activated by ammonium ions and inhibited by glucagon. However, the effects of glucagon and ammonium ions on glutamine synthetase are dependent on portal glutamme concentration (Fig. 2). Variation of portal glutamine concentration in the range of 0-1 mM is nearly without effect on glutamine synthetase flux in the absence of glucagon or ammonium ions. In presence of glucagon, however. there is an activation of glutamine synthetase at
B 0.3.-z o
1:3
u_.
~x
~_c ~E c .
Q
0.2-
0,1
E 0
pH tn Effluent
_~ c~_0.1 o
o
c_~5 Eu~ - 0 . 2 (.9 Z
-0.3
F,g. 5. Effect of perfusate pH on net glutamme release (©1. and flux through glutaminase (*) or glutamme synthetase (1:3) A. LLvers perfused with 0.6 mM glutamine plus 0.6 mM ammonium chloride. Perfusate changes were performed by addition of HCI or NaOH. (pH,. n - p H e m ) was 0.18 on average m stmdar experiments. B. Livers perfused with 0.6 mM glutamme plus glucagon (10 -~ M). Each point represents 5-9 separate measurements during metabohc steady states, Data are from six different perfuslon experiments m each of the A and B.
277 low portal glutamine additions, but a strong inhibition of the enzyme when portal glutamine concentration is raised to above-physiological. With no added glutamine, however, glucagon is without effect on glutamine synthesis (Fig. 2, Table I). With added ammonium ions, flux through glutamine synthetase is substantially increased, but the effect is more pronounced at low portal glutamine concentrations (Fig. 2). This may be due to product inhibition of glutamine synthetase, as was found for isolated rat liver glutamine synthetase in presence of manganese ions [9]. Portal ammonia concentration is reported to be about 0.2-0.3 mM [3,23] and may rise up to 1 mM in ammonia toxicity. Therefore, the effect of portal ammonia concentration on the activities of glutaminase and glutamine synthetase was studied with and without added glutamine at 0.6 mM (Fig. 3). With no added glutamine, there is a sigmoidal increase of glutamine synthesis upon addition of ammonium ions at 0.5-4 ~ m o l / g per min, possibly reflecting a competition of urea and glutamine synthesis for ammonia. Glutamine synthesis is maximal at addition of ammonium ions above 3 t~mol/min per g, i.e. 0.7-0.8 mM in influent. However, with 0.6 mM glutamine added, a maximal glutamine synthetase activity of about 0.25 /~mol/min per g is already observed at an addition of ammonium ions of about 0.4/~mol/min per g. This shows that supply of ammonium ions for glutamine synthesis is not rate-limiting with physiological portal ammonia and glutamine concentrations. Further it is shown in Fig. 3 that glutaminase activity is also increased by ammonium ions. However, this effect is half-maximal at ammonium chloride addition of about 1 /~mol/min per g, and maximal at 2.5-3 ~ m o l / m i n per g. Despite the activity of glutaminase, net glutamine production by the liver with or without added glutamine will account for 50% of ammonia uptake by the liver, if portal ammonia concentration is in the range of about 0.1 mM (Fig. 3).
Cycling of glutamine m the glutaminase / glutamme synthetase system The simultaneous flux through glutaminase and glutamine synthetase in the liver results in a cycling of glutamine at the expense of energy. An
increase of O z uptake by the liver after addition of glutamine, not accounted for energy requirements for urea synthesis from glutamine, has been demonstrated [17]. The rate of this cycle will be determined by flux through the enzyme with the lower activity. Fig. 4 shows that the rate of cycling of glutamine depends on portal glutamine addition and on glucagon and ammonium ions. Interestingly, the rate of cycling is near maximal within the physiological concentration range of glutamine of 0.5-0.6 mM but decreases at higher concentration in the case of glucagon.
Effect of perfusate pH on the simultaneous activities of glutaminase and glutamine synthetase Experimental acidosis, induced by addition of HC1 to the perfusion fluid, leads to an increased net glutamine release by the liver (Fig. 5A, B). With ammonium ions present, this net glutamine release is due to an inhibition of glutaminase activity, whereas glutamine synthetase activity is almost unchanged. In presence of glucagon, however, net glutamine release by the liver is matched by an inhibition of glutaminase as well as an activation of glutamine synthetase. Conversely, increased net glutamine uptake during metabolic alkalosis is due to an inhibition of glutamine synthetase in presence of ammonium ions. With glucagon, again activity changes of both enzymes to a similar extent but in opposite directions bring about increased net glutamine uptake. From Fig. 5A, B it is also evident that maximum rates of the cycling of glutamine are observed at physiological pH values. Discussion
Regulation of flux through glutaminase and glutamine synthetase in the intact liver The flux through glutaminase and glutamine synthetase can be monitored simultaneously in perfused rat liver by measuring 14CO2 release from [1-¿4C]glutamine and glutamine release (when negative:uptake) by the liver. The flux throt~gh both enzymes is effectively modified by the portal glutamine and ammonia concentrations, by glucagon and perfusate pH. The most marked effects are observed at physiological concentrations of these modifiers.
278
The observed response of the two enzymes to acidosis should favour renal acid excretion, by increasing the supply with glutamine from liver, in addition to that coming from muscle [24].
Posstble slgntficance of the cychng of glutamine The simultaneous activities of glutaminase and glutamine synthetase result in a cycling of glutamine at the expense of energy. Such cycles are called futile cycles and have been established to occur in glucose metabolism and their significance for regulatory problems has been elucidated (see Refs. 18, 25). A significance of the glutamine cycling for effective regulation of blood glutamine concentration is considered likely, because maximum rates of the cycle are obtained in presence of a physiological glutamine concentration and at physiological pH values. Glutamine cycling may have a useful purpose because the combined action of glutamine synthetase, a cytosolic enzyme [9], and glutaminase, a mitochondrial enzyme [26], will bring about a net transport of glutamate and ammonia from the cytosolic into the mitochondrial compartment. This is supported by the experiments with and without methionine sulfoximine (Fig. IA, B) showing no utilization of labeled glutamate derived [l-14C]glutamine in the mitochondria by the cytosolic glutamine synthetase reaction. Although two potent glutamate transport systems across the mitochondrial membrane are known (for review see Ref. 27), glutamine cycling could provide a mechanism of glutamate transport from the cytosolic into the mitochondrial space at the expense of ATP, regardless of the mitochondrial/cytosolic glutamate concentration gradient. Interestingly, the highest rates of the cycle are observed during ureogenesis from ammonium ions. As shown in Table I addition of ammonium ions does not affect net glutamine release by the liver and thus hepatic vein glutamine concentration. The rate of cycling, however, is increased about 2-3-fold. Acknowledgements
Expert technical assistance was provtded by H. Weigand and P. Graf. This study was supported by Deutsche Forschungsgemeinschaft, Forschergruppe "Leber", Grant De 113/26-3, and by Schwerpunktsprogramm "Regulationsmechanis-
men des Kohlenhydrat- und Lipidstoffwechsels", Grant Si 255/6-2. References 1 Schrock. H. and Goldstem, L. 11981) Am. J. Physiol. 240, E 519-E 525 2 Lund, P (1980) FEBS Lett. 117, K 8 6 - K 92 3 Lund, P and Wafford, M. (19761 m The Urea Cycle (Gnsolia, S. Baguena. R and Mayor, F., eds ), pp. 479-488, Wiley & Sons, London 4 Atkawa, T., Matsutaka, H., Yamamoto, H., Okuda, T., Ishikawa, E., Kawano. T. and Matsumura, E. (1973) J. Blochem. 74, 1003-1017 5 R~m6sy, C., Demign~, C and Aufrere. J. 119781 Blochem. J. 170, 321-329 6 Yamamoto, H., Alkawa, T., Matsutaka, H., Okuda, T. and Ishlkawa, E. (19741 Am. J. Physiol. 226, 1428-1433 7 Lurid, P. 11971) Btochem. J. 124, 653-660 8 Tare, S.S. and Metster, A 11971) Proc. Natl. Acad Sc~ U S A. 68, 781-785 9 Deuel, T . F , Lome, M. and Lerner, A. 11978) J. B~ol. Chem 253, 6111-6118 10 Charles, R. 11968) Ph D Thesis, Amsterdam, Rototype, Amsterdam I1 Haussinger, D. and Stes, H. (1975) Abstr. Commun. 10th Meet. Fed. Eur Blochem. Soc. No 1497 12 Joseph, S.K and McGlvan, J.D. (1978) Btochlm Blophys Acta 543, 16-28 13 Baverel, G. and Lund, P 11979) Blochem. J. 184. 599 606 14 Lueck, J D and Miller, L.L (19701 J. Biol. Chem. 245, 5491 - 5497 15 Haussmger, D., Akerboom, T.P.M. and Stes, H. 119801 Hoppe-Seyler's Z. Phystol Chem. 361,995-1001 16 Joseph, S K. and McGivan, J.D 11978) Blochem. J 176. 837-844 17 Haussmger, D. and Sles. H 119791 Eur. J. Btochem 101, 179 184 18 Reich, J.G. and Sel'Kov, E E (19811 Energy Metabolism of the Cell, Academic Press, London 19 Lund, P. (1974) m Methoden der Enzymatlschen Analyse (Bergmeyer, H.U., ed.I, 3rd Edn, pp 1767-1769, Verlag Chemle, Wemhe~m 20 Melster, A (1974) m The Enzymes (Boyer, P.D., ed.), 3rd Edn., Vol. 10, pp. 699-754, Academic Press, New York 21 Haussmger, D., Gerok, W. and Stes, H. 119821 Eur J B~ochem. 126, 69-76 22 Du Rmsseau, J.P., Greenstem, J.P. Wtmtz, M. and Blrnbaum, S.M. (1975) Arch. Biochem. Btophys. 68, 161-171 23 / u n d , P., Brosnan, J T and Eggleston, L.V. (19701 m Essays m Cell Metabohsm (Bartley, W., Kornberg, H.L. and Quayle. J.R., eds.), pp 167 188, Wdey lntersclence. London 24 Goldstem, L. (1976) Int. Rev. Physto[ I1,283-316 25 Hue, L. 11982) m Metabohc Compartmentatton (Stes, H , ed.), pp. 71 97, Acadenuc Press, London 26 Kalra, J. and Brosnan, J.T. (1973) FEBS Lett. 37, 325-328 27 La Noue, K.F. and Schoolwerth, A C. 119791 Annu Rev B~ochem 48, 871 922