The partial purification and properties of pig brain glycogen synthase.

Both the I (independent of glucose 6-phosphate) and D (dependent on glucose 6-phosphate) forms of glycogen synthase (UDP-glucose:glycogen alpha-4-glucosyltransferase EG 2.4.1.11) have been partially purified from pig brain and the kinetic constants of the enzymes have been examined. The Km for UDP-glucose for the I form increased from 0.11 to 0.5 mM when the temperature was raised from 25 to 37 degrees. When glucose 6-phosphate was present, the Km for UDP-glucose was decreased to 0.03 and 0.08 mM at 25 and 37 degrees, respectively. The amount of glucose 6-phosphate required to produce half-maximal stimulation decreased with increasing UDP-glucose concentration at both temperatures but increased with increasing temperature. The Km for glucose 6-phosphate at 0.03 and 0.20 mM UDP-glucose was 0.13 and 0.10 mM, respectively, at 25 degrees. At 37 degrees and 0.125 and 4.0 mM UDP-glucose the Km for glucose 6-phosphate was 0.32 and 0.04 mM, respectively. The Km for UDP-glucose for the D form at 0.75, 2.0, and 10 mM glucose 6-phosphate was 0.71, 0.50, and 0.42 mM at 25 degrees. At higher temperatures the apparent affinity for the substrate was decreased; at 37 degrees, the Km for UDP-glucose at 0.75 and 2.0 nM glucose 6-phosphate was 5.75 and 1.42 mM, respectively. The requirement for glucose 6-phosphate was decreased when UDP-glucose concentrations were increased; at 0.5 and 5.0 mM UDP-glucose concentrations, the Km for glucose 6-phosphate was 22.7 and 1.82 mM at 25 degrees. As was the case with the I form, the apparent Km for glucose 6-phosphate increased at higher temperatures. At 37 degrees, the Km for glucose 6-phosphate at 0.5 and 5.0 mM UDP-glucose was 43.5 and 6.15 mM. The temperature coefficient for the maximum velocity was 10.1% per degree for synthase I and 8.5% per degree for synthase D between 25 and 37 degrees. The D form of synthase was calculated to be virtually inactive under normal physiological conditions with the substrate concentrations found in the brain. The enzymatic activity calculated for synthase I correlates well with the observed rate of incorporation of UDP-[U-14C]glucose into brain glycogen.

Glycogen synthesis from uridine diphosphoglucose was first observed by Leloir and Cardini (1) in a crude homogenate of rat liver. Following this initial observation, the enzyme glycogen synthase (UDP-glucose:glycogen a-4-glucosyltransferase EC 2.4.1.11) has been described in a wide variety of tissues (2). The stimulation of glycogen synthase activity of glucose-6-P was noted originally by Leloir (2) ; two interconvertible forms of glycogen synthase subsequently were described in muscle, one of which was completely dependent on the presence of glucose-6-P (D or b) (3), and the other which was independent of glucose-6-P for catalytic activity (I or a). The two forms have since been shown to exist in a number of tissues including liver (4), heart (5), white blood cells (6), brain (7), spleen (S), adrenal gland (9), adipose tissue (lo), and kidney (11,12).
Kinetic studies have been made with brain glycogen synthase using either crude homogenates (7, 13) or partially purified cellfree preparations (14,15). However, the I and D forms of brain synthase have not been isolated separately nor have any kinetic studies of the separated species been undertaken. In the present study we describe the partial purification of glycogen synthase I and D from pig brain and attempt to define the kinetic properties of the partially purified enzymes.

MATERIALS AND METHODS
Enzymes used for the analysis of glycogen synthase and phosphorylase activity were purchased from Boehringer Mannheim Corp. UDP-[U-Wlglucose (210 mCi per mmol) was purchased from the International Chemical and Nuclear Corp. All other chemicals were purchased from Sigma Chemical Co., St. Louis. Pig brains were purchased from Esskay, Baltimore, Md.  0.4 volume of 1% NaCl and 3.6 volumes of 95% ethanol were added, the precipitate was sedimented by centrifugation at 10,000 rpm for 10 min, the supernatant fluid was decanted, and the precipitate was redissolved in 1 ml of water and reprecipitated with 3 ml of 95oj, ethanol.
The last step was repeated for a total of three times, and the residual glycogen was dissolved in distilled water.
The specific activity of the resultant glycogen ranged from 155 to 1640 cpm per nmol (anhydroglucosyl units).

RESULTS
The purification of a representative preparation of the I form of synthase is outlined in Table I   sociated from glycogen at the calcium phosphate gel step, the use of amylase as described by others (62) was not necessary. The ratio of synthase I to glycogen in four subfractions is shown in Table II. The ratios in the succeeding subfractions ivere not calculated because glycogen was too low to measure; in these latter subfractions, however, synthase activity was greatly reduced unless glycogen primer was added during a preincubation step (16). The ratio of synthase I activity with and without preincubation with glycogen primer in the first five subfractions (Table I) was 1.12, 1.08, 4.4, 2.8, and 6.3.
The purification of synthase D from pig brain is summarized in Table III. The procedure of purification was similar to that of the I enzyme except that an I to D conversion was carried out in a large volume at Step 2 and repeated a second time in a smaller volume at Step 6. Synthase D was purified almost 200fold and was 60% as pure of the I form from brain. The 4% present at Step 1 increased to 99% D after treatment to effect the I to D conversion (as determined by measurements in the presence and absence of 5 mM glucose-6-P). In subsequent steps there was a decrease in the percentage of the D form, in spite of the presence of KF to inhibit phosphatase activity. Since protein kinase had co-purified with D through Step 5 (Table III) (16), a second incubation with ATP was carried out. The resulting preparation was 98% in the D form which persisted through column chromatography.
The over-all recovery was 2%.
E$ect of Glycogen Primer on Enzyme Activity-It has been shown that incubation of purified glycogen synthase with glycogen results in an activation of the enzyme (16, 23). The effect of incubation with glycogen on both I and D synthase is shown in Table IV. Further experiments were carried out to determine whether the incubated enzyme was affected by additional glycogen in the assay system. After incubation with 25 mM glycogen (anhydroglucosyl units) the enzyme was diluted in the assay system so that the final glycogen concentration was 0.07 to 0.14 mM. Addition of 10 mM glycogen to the assay system increased the reaction velocities of both synthase I and D only slightly, with similar results over a 35-fold range of UDP-glucose concentrations. Neither the maximum velocities without glucose-6-P nor the affinity for UDP-glucose were substantially affected by the presence of glycogen added to the assay system.
Incubation of the D form of synthase with 25 mM glycogen, and subsequent analyses with and without added glycogen showed results similar to those for the I form of the enzyme. The reaction velocites were slightly greater in the presence of 10 mM than in 0.07 to 0.14 mM glycogen. The maximum velocities were 730 and 768 nmol per ml per min and K, values for UDP-glucose were 0.416 and 0.415 mM in the presence and absence of added glycogen, respectively.

Kinetic
Studies of Synthase I Form-In another series of experiments, the K, for UDP-glucose of the I form of synthase was measured at 25" in the presence and absence of glucose-6-P (Fig. 1). The enzyme was preincubated routinely for 60 min at 25" in the presence of 25 mM glycogen, and 10 mM glycogen was present in the assay medium. The apparent K, for UDP-glucose in the absence of glucose-6-P (according to least squares regression analysis) was 0.113 mM and the maximum velocity was 1.2 pmol per ml per min, or 0.088 unit per mg of the protein. In the presence of 5 mM glucose-6-P the apparent K, for UDPglucose was reduced by two-thirds to 0.0333 mM; the V,,,,, was essentially unchanged at 1.0 pmol per ml per min. At 37" the K, for UDP-glucose in the absence of glucose-6-P was increased to 0.508 mM and the V,,, to 4.0 pmol per ml per min. In the 90 (without glycogen) presence of 5 InM glucose-6-P, the K, was reduced to one-sixth,  1. Activity of glycogen synthase I from pig brain as a function of UDP-glucose (UDPG) concentration. The velocities were measured at 25" in the one-step assay described under "Materials and Methods." Velocities are expressed at micromoles per ml per min. Glucose-6-P (G-6-P) when present was 5 mM. Least squares regression lines are drawn in.  2 (left). The activity of glycogen synthase I at 25" at two concentrations of UDP-glucose (UDPG) and varying concentrations of glucose-6-P. The assays were conducted as described under "Materials and Methods" and Fig. 1.  FIG. 3 (right). The effect of glucose-6-P (G-6-P) concentration on the activity of brain glycogen synthase I at two UDP-glucose (UDPG) concentrations at 37". The assays were conducted as described under "Materials and Methods" and Fig. 1. 0.082 mM, and the Vm,, was somewhat less, 3.0 pmol per ml per min.
The effect of glucose-6-P concentration on glycogen synthase I activity at 25" with two UDP-glucose concentrations is shown in Fig. 2. There was a marked effect of glucose-6-P in both cases (UDP-glucose was below saturation in each instance) ; however, as would be expected from the affinity of the enzyme for UDPglucose, much larger effects were seen at lower concentrations of UDP-glucose. A 4.5-fold increase in synthase I activity was observed at 0.03 IrIM UDP-glucose. When the UDP-glucose concentration was increased to 0.21 mM, the maximum effect of glucose-6-P was a 1.3.fold increase in velocity.
Because glycogen synthase I is active in the absence of glucose-6-P, the data were analyzed for AV, the difference between the velocity in the presence and absence of 5 mM glucose-6-P. If one plots l/AV against l/glucose-6-P the amount of glucose-6-P required for half-maximal stimulation at 0.03 and 0.21 mM UDPglucose is 0.13 and 0.1 mM, respectively. Similar analyses were made for the effect of glucose-6-P at 37" (Fig. 3). At 4 mM UDP-glucose, a near-saturating concentration of substrate, only 1.2-fold increase in velocity is seen at high levels of glucose-6-P. When the UDP-glucose concentration is reduced to 0.125 mM, glucose-6-P increases the reaction velocity 2.8-fold. If one plots the l/AV versus l/glucose-6-P, the concentration of glucose-6-P required to produce half-maximal stimulation at 4 and 0.125 mM UDP-glucose are 0.037 and 0.317 mM, respectively.
The effect of ATP on synthase I activity was minimal. When 5 mM ATP was tested with 1 mM UDP-glucose at 25", the enzyme velocity was 80% of control in the absence of glucose-6-P, and 75% of control in the presence of glucose-6-P. The lack of effect of ATP may be attributed in part to the relatively alkaline pH of the assay system as shown by Piras et al. (24).
Kinetic Studies of Synthme D Form-Unlike the I form, synthase D is completely inactive in the absence of glucose-6-P (in the present preparation the synthase was 97% in the D form, Table III). At 25", when glucose-6-P is 0.75 mM, the K, for UDP-glucose is 0.71 mM and the V,,, 0.303 pmol per ml per min (0.05 unit per mg of protein) (Fig. 4). When glucose-6-P is increased to 2 mM the K, for UDP-glucose is lowered to 0.50 mM and the TImax is increased to 0.440 pmol per ml per min. At 10 mM glucose-6-P the K, is reduced to 0.42 mM and the V,,, is increased to 0.768 pmol per ml per min.
The effect of glucose-6-P on the K, for UDP-glucose is more marked at 37". When glucose-6-P concentration is 0.75 mM the apparent K, for UDPglucose is 5.76 mM and the Vmax is 1.4 pmol per ml per min. When glucose-6-P concentration is increased to 2 mM, the K, for UDP-glucose is decreased to onefourth (1.42 mM), and the V,,, is increased to 2.04 pmol per ml per min. Because the D form of brain synthase was totally inactive in the absence of glucose-6-P, it was possible to use a Lineweaver-Burk plot with l/V plotted against l/glucose-6-P (Fig. 5). At high glucose-6-P concentrations there appeared to be some inhibition and these points, although shown in the figure, were not used in the regression analysis. At 25", the apparent Km for glucose-6-P at 0.5 mM UDP-glucose was 22.7. When UDP-glucose was increased to a near saturating level, 5 mM, the Km for glucose-6-P was reduced to 1.82 mM. The extrapolated maximum velocity was lower with 5 mM than with 0.5 mM UDP-glucose. However, since all of the experimental points with 5 mM UDP-glucose show higher velocity, the extrapolated points may not be valid.
At 37", the Km for glucose-6-P at 0.5 mM UDP-glucose was 43.5 mM and the V,,,,, was 1.04 pmol per ml per min. At near saturating levels of UDP-glucose (5 mM) the Km for glucose-6-P was reduced to 6.15 mM and the V,,, remained unchanged. As at 25", the velocities at the highest glucose-6-P concentrations (4 and 7 mM) were decreased and therefore not considered in the regression analysis. Whether the decreased velocities were a result of an impurity in glucose-6-P which inhibits synthase activity, or inhibition due to excess substrate, was not investigated further.
Binding Experiments-The [r4C]glycogen prepared in our laboratory ("Materials and Methods") was polydisperse in the ultracentifuge, and consequently not suitable for binding experiments, so we attempted to make a more uniform preparation in an effort to determine the molar ratio of synthase bound to the glycogen. The glycogen was centrifuged in a 10 to 30% sucrose gradient, using the SW 39 rotor in a Spinco model L centrifuge. The heaviest fractions (0.6 ml) were combined and concentrated by vacuum dialysis to a final volume of 0.2 ml. The molecular weight of the main component was found to be 4 x lo6 using keyhole limpet hemocyanin as a marker (25). A portion of this glycogen (250 nmol of anhydroglucosyl units, 155 cpm per nmol) was incubated for 60 min at room temperature with 15 pmol of glycogen synthase. Three 10 to 30% sucrose gradients of 5 ml were prepared and layered on top with 60 ~1 of either [14C]glycogen, glycogen synthase, or the [14C]glycogen-enzyme mixture. The tubes were centrifuged 16 hours at 25,000 rpm. Twenty-five 200+1 fractions were collected and portions of each were assayed for glycogen synthase activity, [14C]glycogen radioactivity, and glycogen content. The position of the glycogen synthase peak was not markedly affected by incubation with [14C]glycogen, being displaced only one fraction toward the bottom (Fig. 6). It is clear that, in this preparation at least, the synthase binds preferentially to low molecular weight glycogen. Other binding experiments gave essentially the same results.
Using phosphorylase a as a marker (molecular weight, 400,000 (26)) the molecular weight of the synthase was calculated to be FIG. 6. The binding of glycogen synthase (I form) to glycogen. The samples were prepared and centrifuged in a sucrose density gradient as described in the text. Enzyme activities are given on the left ordinate, and counts per min of the "C-labeled glycogen on the right ordinate. A, 14C-labeled glycogen alone; 0, enzyme activity of glycogen synthase not incubated with glycogen; l , enzyme activity of glycogen synthase incubated with 14C-labeled glycogen; 0, counts per min of 14C-labeled glycogen incubated with glycogen synthase. The bottom of the gradient is shown on the right (Fraction 25). a The V,., in these instances is the maximum velocity with added glucose-6-P at the given UDP-glucose level. gradient centrifugation, found the molecular weight of the rabbit muscle glycogen synthase subunit to be 20,000 to 100,000 and suggested that the native enzyme is a tetramer. Our results agree reasonably well if one assumes the lower molecular weight of the 311,000. Soderling et al. (22), using gel electrophoresis and density monomer. The molecular weight is also in close agreement with Since brain tissue is 10% protein the calculated amount of total synthase in pig brain from Tables I and III is 150 pmol per min per kg wet weight. Assuming Michaelis-Menten kinetics and given the concentrations of glucose-6-P and UDP-glucose in brain tissue (100 and 80 pmol per kg, respectively), synthase D would operate at less than 0.20/, of capacity. In rapidly frozen brains, about 80% of the synthase is in the D form (7, 13) ; thus, the maximum rate of the D synthase activity would be 0.24 pmol per min per kg. Synthase I, on the other hand, could function at about 20% of its I',,,,,. Thus, the calculated velocity of the I form would be in the neighborhood of 6 pmol per min per kg. While these figures are necessarily approximations, the calculated rate of synthase activity is of the same order as the observed rate of incorporation of UDP-[ U-%]glucose into brain glycogen (30). Glycogen concentration need not be considered, since the I form of brain synthase is apparently bound to glycogen in the native state, and its activity independent of added glycogen primer (13).
The total activity of glycogen synthase in the brain is very low compared to that in liver and muscle, and also low compared to that of brain glycogen phosphorylase, which is 3300 pmol kg-' mine1 (35). Brain glycogen stores are only 10 y0 as large as those of resting muscle and 1% of those of liver from a fed animal. These meager stores are not metabolically inert (30), but the concentration of glycogen remains remarkably constant except under conditions of extreme stress such as anoxia (32)) prolonged anesthesia (13, 33), and injury (34). Glycogen in the brain serves as an emergency ration rather than as glucose stores as in liver, or an energy source as in muscle. Although potential phosphorylase activity in brain far exceeds that of synthase (3300 versus 150 pmol kg-1 min-I), the kinetic properties of brain phosphorylase are such that the glycogen is virtually inaccessible for phosphorolysis (20). Under ordinary physiological conditions glycogen stores in the brain are maintained by a balance between low glycogen synthase activity and an effective "kinetic inhi-