Kinetic studies on muscle glycogen synthase.

Using the I form of rabbit muscle glycogen synthase essentially free of glycogen, the kinetics and mechanism of action was investigated. No evidence for an exchange between [14C]UDP and UDP-glucose was found. The bisubstrate kinetics of the enzyme for UDP-glucose and glycogen, as well as for UDP-glucose and maltose, was determined. An intersecting pattern in the double reciprocal plot (velocity versus substrate concentration) suggestive of a sequential mechanism (ordered or random) was found in all cases. The K-m for UDP-glucose (45 to 48 mM) was the same with either maltose or glycogen as acceptor. The K-m for maltose (230 mM) and for glycogen (1.5 mug/ml) differed.

EDUARDO SALSAS$ AND JOSEPH LARNER From the Department of Pharmacology, University of Virginia School of Medicine, Charlottesville, Virginia 22903 Using the Z form of rabbit muscle glycogen synthase essentially free of glycogen, the kinetics and mechanism of action was investigated.
No evidence for an exchange between [W]UDP and UDP-glucose was found. The bisubstrate kinetics of the enzyme for UDP-glucose and glycogen, as well as for UDP-glucose and maltose, was determined. An intersecting pattern in the double reciprocal plot (velocity uersus substrate concentration) suggestive of a sequential mechanism (ordered or random) was found in all cases. The K, for UDP-glucose (45 to 48 mM) was the same with either maltose or glycogen as acceptor. The K, for maltose (230 mM) and for glycogen (1.5 &ml) differed.
Rabbit skeletal muscle glycogen synthase was described as having a bisubstrate kinetic pattern suggesting a ping-pong mechanism (1). It was pointed out that the data were compatible with a stable enzyme substrate intermediate being formed during the reaction and that therefore an exchange between [W JUDP and UDP-glucose should be tested in the absence of glycogen.
Using rabbit skeletal muscle glycogen synthase, we have found no evidence for such an exchange. Accordingly, the bisubstrate kinetics of this enzyme for UDP-glucose and glycogen was reinvestigated and we report here the presence of an intersecting pattern in the double reciprocal plot of velocity versus concentration of substrates suggestive of a sequential mechanism.
In addition, to avoid the problem of using glycogen as one of the substrates since it is not molecularly defined, we have used maltose as acceptor with essentially the same results. Purification of Glycogen Synthase-Fully converted glycogen synthase Z form was purified from rabbit skeletal muscle by the procedure described previously (2), with the following modifications.
The high speed centrifugation step normally following the first ethanol precipitation was omitted. In order to prevent the aggregation of the enzyme when it is free of

Materials-Maltose
(HIO) grade HHH from Hayashibara Co., Ltd. (Japan) was a gift of Dr. S. Hizukuri. All other materials were purchased from conventional sources.
temperature, as described previously, but in the presence of 25% glycerol.
The presence of glycerol (25 or 50%) and/or a high salt concentration (0.5 M KCl) effectively prevents the loss of enzyme activity due to the aggregation and may even partially reverse it (Table I).
The last step in the purification procedure was Sepharose 4B gel filtration.
We have now found that this can be substituted by hydrophobic chromatography using w-aminoalkyl Sepharose according to the method of Shaltiel and Er-El (8). We have used two w-aminoalkyl Sepharoses, namely the 4-and B-carbon derivatives.
With the purified enzyme, the Sepharose-NH-(CHJ,-NH,, rather than the 4-carbon analog, was found to be most suitable. With the 4-carbon derivative, the enzyme was only retarded ( of glycogen synthase Z, heavily aggregated after digestion of the glycogen with a-amylase precipitated on standing, and its activity is shown as control. This precipitated preparation was incubated at room temperature for 30 hours with 50 mM Tris, 5 mM EDTA, and 100 mM mercaptoethanol containing (a) 0.5 M KC1 and 25% glycerol, (5) 0.5 M KCl, and (c) 50% glycerol. The activity was assayed after HO dilution with 50 mM Tris, 5 mM EDTA, 50 rnM mercaptoethanol, f 1 mg/ml of glycogen & 0.5 M KC1 (diluting buffer). The activities are given as percentage of the activity found in the control.

Enzvme incubation medium
Composition of diluting buffer  (Fig. 2). The secondary plot of the intercepts and slopes (Fig. 3) obtained at the different glycogen concentrations is linear and allows the calculation of a K, value of 1.5 pg/ml for glycogen. Calculation of the K, for glycogen in terms of the nonreducing end termini, assuming that 9% of the glucosyl units are present as nonreducing end termini, gives the value of 8 x lo-' M.
The Adsorption and elution of glycogen synthase to w-aminoalkyl Sepharoses. A, a preparation of glycogen synthase was digested with a-amylase at room temperature in the presence of 25% glycerol. After centrifugation at 5000 rpm (Sorvall SS-34 rotor), the supematant was dialyzed against 50 mM &glycerophosphate, 1 mM EDTA, and 25% glycerol (pH 7.0), and applied to a Sepharose IB-NH-(CH,),-NH, column (8 x 1.4 cm) previously equilibrated with the same buffer. When the absorbance decreased below 0.02 unit, the buffer was changed to 50 mM )9-glycerophosphate, 1 mM EDTA, 0.5 M KCl, and 25% glycerol (pH 7.0). Fractions were collected, and the glycogen synthase activity (-O-O-) as well as absorbance were monitored (-).
B, a pool of Fractions 3 to 21 from the experiment of Fig. 1, containing 0.125 M KC1 was applied to a Sepharose 4B-NH-(CH,),-NH, column (7 x 1.4 cm) previously equilibrated with the same buffer. Three stepwise elutions with 50 mM fi-glycerophosphate, 1 mre EDTA, and 25% glycerol (pH 7.0) containing 0.4 M KCl, 0.6 M KCl, and 1.0 M KC1 were done. Fractions were collected, and the glycogen synthase activity (-O-O-) as well as absorbance (-) were monitored. Isotopic exchange between ["C ]lJDP and UDP-glucose An incubation mixture containing 1.5 nh4 [IC]UDP (0.2 j&i), 1 ELM UDP-glucose, 50 milliunits of glycogen synthase Z essentially free of glycogen, and 50 rnM 2-(N-morpholine)ethane sulfonic acid (pH 6.5) in a total volume of 1 ml was incubated at 30" during 2 or 4 hours. Trichloroacetic acid was then added to a final concentration of 5%, the incubation reaction mixture was centrifuged, the supematant was extracted with ether several times until the pH reached 5.5, and the mixture was lyophilized. The lyophilized material was redissolved in water, UDP was added to a final concentration of 1 mM, and the mixture was chromatographed on Whatman No. 3MM paper in ammonium acetate-ethanol (3:7, v/v, pH 7.5) containing 20 mM EDTA. The radioactivity found in the UDP and UDP-glucose positions is shown. series of intersecting lines (Fig. 4). The secondary plot of the intercepts and slopes obtained at the different UDP-glucose concentrations is also linear, and the V,., obtained is in good agreement with the one obtained in Fig. 3. A K, of 45 PM for UDP-glucose can be calculated. The problem with enzymes acting on polysaccharides is the difficulty in defining molecularly the substrate (in this case, glycogen) and the fact that the reaction always proceeds in the presence of the product of the reaction which also serves as substrate.
In the case of phosphorylase, it has been indicated (10) that it would be necessary to consider two complexes of enzyme and polysaccharide.
In one complex, the binding would be for degradation, whereas in the other complex, the binding would be for chain elongation.
In such a case, the rate equations are somewhat different from the common two substrate systems, and for some mechanisms could present a pattern of lines in the double reciprocal plot (10) different from the usual patterns (5, 11).
In the case of glycogen synthase, the product again is expected to form a complex with the enzyme, but depending on the nature of the kinetic mechanism, this complex can either immobilize the enzyme in a non-useful form or be useful for the next catalytic cycle, and therefore not inhibit the reaction. Particularly, in an ordered sequential mechanism with UDPglucose binding first and in a ping-pong mechanism (for which the only possibility is UDP-glucose binding first), a complex of enzyme and glycogen (involving in any way the active center) will produce an inhibition.
The extent of this inhibition will depend on the relative values of the kinetic constants of the steps involved. On the other hand, in a random mechanism or in an ordered sequential mechanism with glycogen binding first and the glycogen product released last, most likely no inhibition will be seen because the enzyme is not able to bind the glycogen in any way different from that for elongation.
Initial Velocity Experiments with UDP-Glucose and Maltose as Substrates-A simple and expeditive way to circumvent the outlined difficulties encountered with glycogen is to use small oligosaccharides such as maltose and maltotriose as acceptors. Both of them, as well as glucose (12), can be acceptors for the glycogen synthase catalyzed reaction, although at a much lower rate (13). If the concentration of acceptor is high enough, the reaction is single step; thus when maltose is used, only maltotriose is found as product; whereas when maltotriose is used, only maltotetraose is found as product. All of the radioactivity is found at the nonreducing termini. This fact has been proven in the case of maltotriose formation by means of metaperiodate oxidation and separation of the formic acid formed. Of the total radioactivity, one-sixth is recovered as formic acid. 1 When the bisubstrate kinetic analysis is done using UDP-glucose and maltose, the double reciprocal plots show intersecting patterns (Figs. 5 and 6). As expected, the concentrations of acceptor necessary to achieve measurable velocities are much higher than those of glycogen in terms of nonreducing ends. Saturation with the maltose acceptor was not achieved, the main limitation being the viscosity and ultimately the solubility of the maltose. The secondary plots are linear, and a K, of 230 mM for maltose and a K, of 48 PM for UDP-glucose can be calculated from them. Therefore, while the K, for the acceptor changes markedly with its complexity, the K, for UDP-glucose is the same with maltose or glycogen as acceptors.

GENERAL DISCUSSION
The Z form of glycogen synthase more readily aggregates than the D form and is more cold-sensitive when freed of glycogen (9). We have found that glycerol and high salt concentrations prevent and even partially reverse this aggregation. The absence of isotopic exchange between [14C]UDP and UDP-glucose suggests that no glucosyl enzyme intermediate is formed. This fact rules out a Ping Pong Bi Bi mechanism for glycogen synthase. It is worthwhile to note that the test for the back reaction was run at pH 6.7 instead of pH 7.8. It has been pointed out (14) that the synthase reaction liberates a proton, the neutralization of which displaces the equilibrium further to the right. We considered it necessary to minimize this factor to have a better chance of seeing the exchange reaction should it occur.
In agreement with the lack of a demonstrable exchange reaction, when a careful reinvestigation of the bisubstrate kinetics was undertaken, we found an intersecting pattern, compatible with a sequential (ordered or random) mechanism. Nevertheless, in certain experiments (not shown) a curvature in the double reciprocal plot was apparent. In such cases, an apparent parallelism of the lines was seen near the y axis. This parallelism led us to consider as a working hypothesis that the pattern of addition of glycosyl residues to the glycogen nonreducing terminals could change in relation to the relative concentrations of both substrates, glycogen and UDP-glucose. In that way, at high UDP-glucose concentrations, the pattern would be close to a single chain elongation.
The glycogen chain would not need to leave the active center of the enzyme, and the release of 1 or more UDP residues before all the molecules of UDP-glucose were added would give a pattern close to parallelism.
We have chemically proven' that this is not the case. Thus, only 1 or 2 glucose residues added successively to the same branch, and most probably only 1, is compatible with these results.
Another explanation for the curvature found only with some preparations of enzyme would be the presence of two forms of the enzyme (I5-17), one of them arising from limited proteolysis of the other. In fact, evidence has been found that indicates that this actually happens during the standard purification procedures of both forms of the enzyme. p In any event, the present studies indicate that the kinetic mechanism is compatible with a sequential mechanism for glycogen synthase acting either on glycogen or maltose as acceptor. After these studies were completed, two publications on the subject have appeared. Plesner et al. (18) working with glycogen synthase D from human polymorphonuclear leukocytes showed that the enzyme has a rapid equilibrium random Bi Bi mechanism, in which the attachment of the activator glucose-6-P was a necessary prerequisite for the addition of the substrate UDP-glucose. Huang and Cabib (19) observed biphasic curves in the double reciprocal plots of reaction rate uersus UDP-glucose concentration. Convergent patterns of lines on double reciprocal plots was observed in all cases. Thus, these results are in general agreement with those we obtained with the rabbit muscle enzyme.