ARCHIVES
Vol.
OF BIOCHEMISTRY
284, No. 1, January,
AND
BIOPHYSICS
pp. 143-150,
1991
Bovine Pancreatic Asparagine Synthetase Explored with Substrate Analogs and Specific Monoclonal Antibodies Paige
M. Mehlhaff*
and Sheldon
M. Schustert,’
TDepartment of Biochemistry and Molecular Biology, University of Florida, Gainesville, and *Department of Pathology, University of California, La Jolla, California 92093
Received
May
17, 1990, and in revised
form
August
L-Asparagine synthetase (EC 6.3.1.1) is the enzyme responsible for the production of asparagine from aspartate and a nitrogen source. In mammalian tissues the enzyme can utilize either glutamine or ammonia as the nitrogen
0003.9861/91$3.00
Copyright All rights
0
1991 by of reproduction
Academic in any
Press, Inc. form
reserved.
32610;
28, 1990
Several substrate analogs were tested for their ability to inhibit bovine pancreatic asparagine synthetase. Of the substrate analogs tested both 6-diazo-5-oxo-L-norleucine (DON) and 5-chloro-4-oxo-L-norvaline (CONV) were shown to inhibit the enzyme strongly. DON inhibited the glutaminase and glutamine-dependent asparagine synthetase activities and CONV inhibited the ammonia-dependent activity as well. Both of these inhibitors appeared to be relatively tight binding since desalting failed to remove the inhibition. The inactivation of bovine pancreatic asparagine synthetase by DON is accompanied by a shift from a 47,000 molecular weight monomer to a 96,000 molecular weight dimer as observed by HPLC gel filtration chromatography. This DON-induced shift is prevented by the presence of the substrate glutamine. A monoclonal antibody known to inhibit specifically the ammonia-dependent and glutamine-dependent asparagine synthetase activities but not glutaminase (monoclonal antibody 2B4) binds to both the monomer and the dimer forms of untreated enzyme, as well as to the dimer form of the DON-inactivated enzyme. On the other hand, a monoclonal antibody known to inhibit specifically the glutaminase and glutamine-dependent activities and not the ammonia-dependent asparagine synthetase (monoclonal antibody 5A6) binds to both forms of untreated enzyme but cannot bind to the DON-inactivated enzyme. These data are used to describe the relation of regions of the active site of asparagine synthetase in relation to ano 1991 Academic press, I~C. tibody binding sites.
I To whom correspondence should be addressed University of Florida, Gainesville, Fla 32611.
Florida
at 1301 Fifield
Hall,
source in the ATP-dependent synthesis of asparagine. The glutamine-dependent activity can be depicted as L-glutamine + L-aspartate + Mg-ATP
+
L-asparagine + AMP + PPi + Mg2+, and the ammonia-dependent reaction as (1, 2) NH3 + L-aspartate + Mg-ATP
+
L-asparagine + AMP + PPi + Mg’+. In addition, asparagine synthetase can function as a glutaminase in the absence of Mg-ATP and aspartate (3-5), L-glutamine + L-glutamate + NH3. In earlier studies from this laboratory it was shown that when glutamine is provided as the nitrogen source, bovine pancreatic asparagine synthetase exhibits an ordered binding mechanism where glutamine binds first followed by Mg-ATP. Glutamate is then released, aspartate is allowed to bind, and the rest of the products are released in the order of pyrophosphate, AMP, and finally asparagine. Conversely when ammonia is the nitrogen source, both Mg-ATP and ammonia bind in a random order, then aspartate binds and the products are then released in the same order (6). In previous work several monoclonal antibodies (Mabs)2 were produced against bovine pancreatic asparagine synthetase. One group of Mabs binds to the enzyme but does not affect activity. Mabs from this group were used to purify the enzyme. Another set of Mabs were * Abbreviations used: CONV, 5-chloro-4-oxo-L-noraline; DON, 6diazo-5-oxo-L-norleucine; DTT, dithiothreitol; Mab, monoclonal antibody; PBS, phosphate-buffered saline; PP, pyrophosphate; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; FGAR, formylglycinamide ribonucleotide amidotransferase. 143
144
MEHLHAFF
AND
shown to be inhibitory, particularly Mab 5A6, which inhibited glutamine-dependent asparagine synthetase and glutaminase activity, but not ammonia-dependent asparagine synthesis. Mab 2B4 and Mab 3F3 inhibited both the glutamine-dependent and the ammonia-dependent asparagine synthetase activities but not the glutaminase activity. This differential inhibition by the Mabs showed that there are two distinct amino acid binding domains on the enzyme-one for asparagine (and aspartate) and the other most likely for glutamine (and glutamate). Additional studies with the Mabs also suggested that the binding sites are topographically separated from each other on the enzyme surface (7). Several important questions remain regarding the relation of the binding sites for the antibodies and the sites for amino acid binding. These center around the key question of whether or not there is a region of contiguous primary sequence of the asparagine synthetase that is specifically involved in the various portions of the amino acid binding sites. The binding of the antibodies previously characterized has already been shown to be related to the binding domains in several distinct antigenic regions (7, 8). However, it could not be determined if the binding and subsequent inhibition was due to complexation of the glutamine binding site or was due to conformational effects caused by interaction with a site distinct from the glutamine binding site. If other inhibitorspreferentially covalent, specific modifiers-could be identified and kinetically characterized, and their effect on antibody binding determined, then it would be possible to distinguish between these possibilities. In addition, if the interaction were directly to the glutamine binding site, then we could begin determining the region of primary sequence essential for amino acid binding. To this date, however, little is known regarding inhibitors of asparagine synthetase, and nothing is known regarding how any might effect the binding of antibodies known to inhibit specific functions of the enzyme. This work reports attempts to probe the structurefunction relationship of the enzyme and the various antibodies available with substrate analogs. One particular substrate analog, DON, is shown to be an apparently covalent modifier of asparagine synthetase from beef pancreas that specifically effects only glutamine-dependent function. The effect of the binding of DON on antibody binding is also described. It has been known that asparagine synthetase is a dimer of identical subunits (9, lo), but no information has been presented in the literature regarding the interaction of the subunits, and the relation of these interactions to the catalytic function of the enzyme. These studies also provide information regarding the quaternary structure of asparagine synthetase. The major conclusion, however, is that the data are consistent with the hypothesis that the binding of Mab 5A6 to asparagine synthetase is directly to the glutamine binding site.
SCHUSTER
EXPERIMENTAL
PROCEDURES
Monoclonal antibodies. The production and growth of hybridomas producing monoclonal antibodies specific for bovine pancreatic asparagine synthetase has been described previously (8). Antibodies were purified from ascites fluids by precipitation to 45% with saturated ammonium sulfate, followed by a second similar precipitation. Asparagine synthetase purification. Bovine pancreas was obtained from Spencer Beef (Schuyler, NE) or Iowa Beef Packers (West Point, NE) and purified as previously described (8) with modifications as indicated by Pfeiffer et al. (7). Briefly, the purification involves homogenizing the pancreas in Tris buffer at pH 8.0 (50 mM Tris-Cl, pH 8.0, 0.5 mM EDTA, 1.0 mM dithiothreitol). After centrifugation, the homogenate is applied to an immunoaffinity column, the column is washed with PBS, and the enzyme is eluted with sodium carbonate at pH 10.6. The eluted enzyme is precipitated by adding solid ammonium sulfate to 50% saturation, the precipitate is collected by centrifugation, and the pellet is resuspended in Tris buffer. Before use, the enzyme preparation is desalted on a G-25 column equilibrated in the same Tris buffer. The enzyme prepared in this manner is extremely pure as judged on acrylamide gels stained by either Coomassie blue or silver stain (7). All experiments reported use only enzyme that yields a single band on SDSPAGE monitored by silver staining. Zmmunoufinity the asparagine Mab 3F3.
columns. The immunoaffinity synthetase was made as previously
column used to purify described (7,8) using
Asparugine synthetase assay. The glutamineand ammonia-dependent synthesis of asparagine was assayed as previously described (11). In this assay either glutamine or ammonia, Mg-ATP and L-[414C]aspartate are incubated with the enzyme at 37°C for the desired time period. The reaction is stopped by pipetting the reaction mixture into hot sodium acetate. The radioactive label in the excess aspartate is removed in a decarboxylation reaction using pyridoxal and A&SO,. The label remaining in the asparagine is then quantitated using a Beckman liquid scintillation counter and the amount of asparagine is calculated by comparison to a nondecarboxylated L-[4-“Claspartate control performed with each experiment. Zeroes for each data point are obtained in the same manner except that the reaction is stopped immediately upon addition of the enzyme to the substrate mixture. The glutaminase activity was assayed as described by Pfeiffer et al. (8), where the formation of glutamate was coupled to the reduction of p-iodonitrotetrazolium violet and the absorbance monitored at 510 nm. Protein assay. The protein concentration zyme was determined using the Bio-Rad Richmond, CA).
in each preparation protein assay (Bio-Rad
of enCo.,
Gel filtration HPLC. The gel filtration HPLC was performed using a Bio-Rad TSK 250 gel filtration column and a Beckman HPLC pump equipped with a uv detector. Unless otherwise indicated, the column was equilibrated in 20 mM Bis-Tris at pH 6.5 with 150 mM Na$O,. Enzyme samples were mixed with control buffers, inhibitors, substrates, antibodies, etc. and then 50 to 100 aI of the mixture was immediately injected onto the column and the absorbance was monitored at 280 nm. The corresponding chromatograms and the integration of the peaks were obtained using an Apple IIE computer. Molecular weights were obtained by calculation by comparison to commercially available standards (Bio-Rad).
RESULTS Several substrate analogs were tested for their ability to inhibit bovine pancreatic asparagine synthetase. In most cases the inhibitor was incubated with the enzyme for 30 min at the designated concentration (preassay concentration) and then the enzyme-inhibitor mixture was added to the substrate mixture for assay. Thus the con-
SUBSTRATE
ANALOGS
AND ANTIBODIES
ON BOVINE
TABLE
ASPARAGINE
145
SYNTHETASE
I
Effect of Several Compoundson the Glutamine and Ammonia-Dependent Activity of AsparagineSynthetase Concentration Inhibitor Albizzin
Azaserine
CONV DON Maleimide
Preassay 10 10 30 10 10 30 10 5 40
Substrates
(mM)
Assay
Saturated
0.83 0.83 10 0.83 0.83 10 0.42
X
10
X
centration of the inhibitor during the assay was lessthan during the incubation time. Since saturating concentrations of substrates mask the action of a simple competitive inhibitor, some of the assays were performed at low or nonsaturating substrate concentrations in order to observe more readily any effect. The effects on asparagine synthetase activity of some of the compounds tested are shown in Table I. Albizzin, a glutamine analog, inhibited the glutamine-dependent asparagine synthesis activity of asparagine synthetase to 37% of the control activity only when the inhibitor concentration was high (10 mM during the assay) and the substrate concentrations were low (nonsaturating). At lower concentrations of albizzin there were only slight effects whether or not the substrates were in saturating concentrations. Azaserine, another glutamine analog, tested under the same conditions as albizzin showed no significant inhibition of either ammonia- or glutamine-dependent activity. At high concentrations (10 mM), 5-chloro-4-oxo-L-norvaline (CONV) inhibits both the glutamine- and the ammonia-dependent asparagine synthesis by bovine pancreatic asparagine synthetase. At 5 mM, DON totally inhibits the glutamine-dependent activity but has only a slight effect on the ammonia-dependent activity. Maleimide, a sulfhydryl modifier, exhibits significant inhibition of the glutamine-dependent asparagine synthetase activity (36% of the control) and a slight activation (118%) of the ammonia-dependent activity. Of the compounds tested, DON and CONV appeared to have the greatest effect on the pancreatic enzyme. In addition, in other caseswhere these compounds have been used, they were shown to be covalent, and highly specific to glutaminase functional sites (12). Therefore, DON was used for further studies of pancreatic asparagine synthetase. The concentration dependence of the DON inhibition of bovine pancreatic asparagine synthetase was examined by adding increasing amounts of DON to a preparation of asparagine synthetase and measuring the three activ-
% Activity Nonsaturated
X X X X
X X X
remaining
Gln
NH3
105 91 37 105 91 95 0 1 36
111 95 90 113 95 100 0 86 118
ities relative to a control without DON. As seen in Fig. 1, in the presence of DON, the glutamine-dependent asparagine synthetase activity rapidly decreased with increasing concentrations of DON until 2 mM, where the activity was virtually zero. The glutaminase activity was inactivated even more rapidly and was completely abolished at 1 mM DON. On the other hand, the ammoniadependent activity was not inhibited at all in the presence of DON and appeared to be slightly increased, even at concentrations as high as 20 mM. In order to determine if the inactivation of asparagine synthetase was due to covalent modification of the enzyme, the DON-inactivated asparagine synthetase was passedthrough a desalting column (Bio-Gel P-GDG, BioRad) and reassayed for activity. There was very little change in the extent of inhibition after passagethrough the column, indicating that the DON molecule is not readily removed from the enzyme (see Table II).
0
IO
20
30
DON (mt-l) FIG. 1. DON inhibition of asparagine synthetase. The percentage activity remaining relative to a sample with no DON, for the ammoniadependent synthesis (m), the glutamine-dependent synthesis (O), and the glutaminase (A) activities is plotted versus increasing DON concentrations (0, 1.0, 2.0, 5.0, 10.0, and 25.0 mM).
146
MEHLHAFF
TABLE Activity
of DON-Inactivated Passage through
AND
II
Asparagine Synthetase a G-25 Column
after
Specific activity of asparagine synthetase (nmol/min/mg) Sample Initial enzyme activity DON-inactivated enzyme DON-inactivated enzyme after G-25 chromatography
Glutamine dependent
Ammonia dependent
156 15
167 188
25
185
Since the asparagine synthetase is capable of three different enzymatic activities, it is of interest to determine if the interaction of the subunits affects any of the various catalytic properties of the enzyme. Indeed it is crucial to know if the effects of DON, for example, could be due to changes in quaternary structure. However, there is no basic information regarding the quaternary structure of the enzyme. For a better understanding of the enzymesubstrate interaction, some studies were done to examine some of the factors responsible for the interaction of the subunits, especially in regard to substrates, antibodies, and substrate analogs. A purified preparation of bovine pancreatic asparagine synthetase was subjected to gel filtration chromatography, and the chromatogram shown in Fig. 2A was obtained. It should be noted that the buffer used contained 150 mM NaPSO to maintain enzyme solubility. In buffers of lower ionic strength, protein precipitation is observed. It should also be noted
that the absorbance
peaks observed
SCHUSTER
monia-dependent (-lo-fold greater) or the glutaminedependent activity (-20-fold greater). After storage at -20°C there is a marked decrease in all three activities with the greatest loss in the glutamine-dependent synthetase activity followed by the glutaminase activity with the ammonia-dependent synthetase activity exhibiting the greatest stability (data not shown). As noted above, it was crucial to determine if the effects of DON on activity are due to changes in subunit interactions. When the enzyme preparation shown in Fig. 2A was treated with DON at a final concentration of 1 mM and subsequently chromatographed on the gel filtration column, the result is that shown in Fig. 2B. As can be seen, the presence of DON appears to shift the enzyme to its dimer form. Again, the species eluting after 13 min are DON and storage buffer components. The shift to the dimer form as a function of DON concentration is shown in Fig. 3. The integrated area corresponding to both the dimer and the monomer are almost equal when no DON is present but as DON is added the area corresponding to the monomer rapidly decreases to a minimum value
in these
elutions after about 13 min are due to the absorbance or refraction of salts and solutes present in the enzyme storage. Controls done without enzyme but with storage buffer components (Tris, EDTA, DTT, etc.), show these same peaks. There are two major forms of the enzyme present under these conditions, a high molecular weight form (96,000), corresponding to the dimer form, and a smaller (47,000) form, corresponding to the monomer form. These two forms are present in all enzyme preparations tested although the relative amounts vary from preparation to preparation. The two peaks have been collected separately and assayed, and both exhibit all three activities i.e., glutaminase and glutamine- and ammonia-dependent asparagine synthesis (data not shown). In all casesthe enzyme prior to chromatographic separation has higher amounts of ammonia-dependent activity than glutaminedependent activity (usually by a factor of two). After gel filtration chromatography the ratio of ammonia-dependent to glutamine-dependent activity is closer to one for the dimer and less than one for the monomer. The glutaminase activity is much greater than either the am-
I
10 Elution
Time
15 (min)
FIG. 2. Chromatogram of asparagine synthetase using HPLC gel filtration. A is untreated enzyme, showing the dimer (peak 1) and the monomer (peak 2) forms. B is the same enzyme after treatment with 1.0 mM DON.
SUBSTRATE
ANALOGS
AND
ANTIBODIES
80
E; 0 0
I
60
.z g
40
+ -Y 2
20
0
0
L
IO
20
30
DON (mM) FIG. 3. Area of dimer thetase after inactivation
(m) and monomer (0) peaks of asparagine synwith 0, 1.0, 2.0, 5.0, 9.3, and 25 mM DON.
while the area corresponding to the dimer increases to a maximum and then levels off. It is of interest to determine if DON binding competes for antibody binding to asparagine synthetase, especially since so much is known regarding antibodies affecting specific function. Two of the monoclonal antibodies previously shown to inhibit the asparagine synthetase activities differentially were tested for their ability to bind in the presence and absence of the DON. The binding of the monoclonal antibody to the enzyme can be visualized on the gel filtration column by adding increasing amounts of antibody to the enzyme and observing the disappearance of the enzyme peaks and the concomitant emergence of a very large molecular weight peak corresponding to the antibody-enzyme complex. Figure 4 shows three of the gel filtration chromatograms obtained when increasing amounts of Mab 2B4 are added to a preparation of asparagine synthetase. Panel A shows the enzyme alone with no antibody added. Both the dimer and the monomer forms of the enzyme are clear. When a 0.1 mg/ml solution of Mab 2B4 is added after mixing with asparagine synthetase (B) both of the enzyme peaks are markedly reduced and a new, large molecular weight peak emerges.When the antibody concentration is further increased (C), the two enzyme peaks completely disappear and two large molecular weight peaks remain. It is likely that the earliest eluting peak represents the antibodyenzyme complex, while the other is in the position where free Mab 2B4 elutes, and thus represents excess antibody. When Mab 2B4 is added to an enzyme preparation previously inactivated with DON the results shown in Fig. 5 are obtained. Panel A shows the DON-inactivated enzyme in the absence of antibody. As before, there is only the dimer form of the enzyme in the presence of DON. The addition of increasing amounts of Mab 2B4 (B-D) results in the disappearance of the enzyme peak and the appearance of the large molecular weight species as occurred in the absence of DON. The same experiment was
ON
BOVINE
ASPARAGINE
147
SYNTHETASE
performed with Mab 5A6, which inhibits glutamine-dependent asparagine synthesis and glutaminase activities, but not ammonia-dependent asparagine synthesis. As can be seen in Fig. 6, the monomer and dimer forms of the enzyme are removed by Mab 5A6. When Mab 5A6 was added to DON-inactivated enzyme, somewhat different results were obtained. As shown in Fig. 7, the DON-inactivated enzyme is present only in dimer form in the absence of antibody (panel A). As increasing amounts of Mab 5A6 are added (B-D) there is an increase in the amount of the peak eluting at approximately 8 min, but no noticeable decrease in the dimer peak. Even at very high concentrations of Mab 5A6 (Fig. 7D) there is only a slight decrease in the area corresponding to the dimer of asparagine synthetase. Again, in all of these figures, the material eluting after about 13 min was shown to be buffer components. The ability of the monoclonal antibodies to bind to the DON-inactivated enzyme is presented graphically in Fig. 8. As the amount of antibody is increased, the area of the dimer decreases in a linear fashion for both Mabs 2B4 and 3F3 (Mab 3F3 also inhibits the ammonia-dependent and glutamine-dependent asparagine synthetase activities and not the glutaminase activity as does Mab 2B4 but not as strongly), whereas for Mab 5A6 almost no decrease 1
A
ALJL i
5
Elution
10
Time
(min)
FIG. 4. Chromatograms from gel filtration of asparagine synthetase with Mab 2B4. A is asparagine synthetase alone, with peak 1 the dimer form and peak 2 the monomer form. In B a 0.1 mg/ml solution of Mab 2B4 was mixed with asparagine synthetase, then added. Peak 1 and 2 are the dimer and monomer forms, respectively, and peak 3 is the antigen-antibody complex. In C, more Mab 2B4 has been added (a 0.1 mg/ml solution); peak 4 represents excess uncomplexed antibody.
148
MEHLHAFF
AND
SCHIJSTER
of asparagine synthetase are present. When smaI1 amounts of DON are added, there is a 51% increase in the area corresponding to the dimer peak and simuitaneously a 56% decrease in the area corresponding to the monomer. The presence of MgATP has no effect on the shift from the monomer to the dimer forms, and ammonia and aspartate have only small effects. Only g1utamine appears to have a significant effect, since there is no increase in area corresponding to the dimer with the addition of DON if glutamine is present, indicating that the glut~ine may actually prevent DON from binding to the enzyme.
5
Elution
FIG. 5. Chromatogram from gel paragine synthetase with Mab 2B4. the dimer form (peak 1). In B a 0.1 added with peak 3 again being the had 0.25 mg/ml Mab 2B4 added and the excess uncomplexed Mab 2B4.
Asparagine synthetase is one of many enzymes comprising the class of glutamine amidotransferases. The g~utamine amidotransferases exhibit a wide variety of functions including biosynthesis of amino acids, nucleotides, and coenzymes among others (18). Many of these amidotransferases have also been studied using substrate analogs. Albizzin was shown by Schroeder et ai. specifically to inhibit formylglycinamide ribonucleotide amidotransferase (FGAR) (23). As with asparagine synthetase, albizzin inhibited only the g~utamine-dependent activity of FGAR and not the ammonia-dependent activity. Nonetheless it did so at a much lower concentration than was needed for inhibition of the asparagine synthe10
15
Time fminl filtration of DON-inactivated asA is DON-inactivated enzyme in mg/ml solution of 934 had been ant.igen-antibody complex. C has D 0.5 mg/mI Mab 284. Peak 4 is
in dimer area is seen over the same range of antibody concentration. Since the shift from monomer to dimer upon addition of DON occurs simultaneously with the inactivation of the glutamine-dependent asparagine synthetase activity, monitoring the relative amounts of dimer and monomer is a convenient way of determining the extent of inactivation and also the ability of various compounds to protect the enzyme from inactivation by DON. In order to determine if the substrates normally utilized by asparagine synthetase could protect the enzyme from inactivation and the concurrent shift to the dimer form, the various substrates were added to the enzyme at concentrations normally used for assaying activity at V,,,. DON was then added, the samples were immediately put on the gel filtration column, and the area corresponding to the dimer and monomer peaks was obtained. These results are shown in Table III. In the absence of DON, significant amounts of both dimer and monomer forms
5
10
Elution Time (mid FIG. 6, Chromatugram from a gel fiitration column of asparagine synthetase with Mab 5A6. A has no Mab added and peak t and 2 are the dimer and monomer forms, respectively. B has had a 0.25 mg/ml solution of Mab 5A6 added and C a 0.5 mg/ml solution of Mab 5A6.
SUBSTRATE
ANALOGS
AND
ANTIBODIES
ON
BOVINE
c
ASPARAGINE
0.0
149
SYNTHETASE
0.2
0.4
0.6
0.8
1.o
I .2
mg Mab per mg Enzyme
1 4
FIG. 8.
Plot of the dimer peak of asparagine synthetase with respect to the concentration of Mab 5A6 (W), Mab 3F3 (Cl), and Mab 2B4 (A).
A 4
D
1
~
~ 5
10
El&ion
Time (mid
1
15
FIG. 7.
Chromatogram from a gel filtration column of aaparagine synthetase with Mab 5A6. A is DON-inactivated enzyme in the dimer form (peak 1). In B a 0.1 mg/ml solution of 5A6 had been added. C has had 0.25 mg/ml Mab 5A6 added and panel D 0.5 mg/ml Mab 5A6. Peak 4 is the excess uncomplexed Mab 2B4.
tase (23). Azaserine also inhibited FGAR and French et al. provided evidence that azaserine alkylates a particular sulfhydryl group on the enzyme (24). But as shown in Table I, azaserine had no effect on any of the asparagine synthetase activities. Although azaserine had no effect, another sulfhydryl modifying reagent, maleimide, did inhibit bovine pancreatic asparagine synthetase. Milman and Cooney have shown that maleimide produced a 94% inhibition of glutamine-dependent asparagine synthetase from L5178,fAR and a 47% inhibition of ammonia-dependent asparagine synthetase activity from the same source (25). A similar inhibition pattern was observed in asparagine synthetase from mouse pancreas and the authors propose that maleimide reacts with the enzymes via a sulfhydryi alkylation (25). Unlike the results of Milman and Cooney a similar concentration of maleimide inhibits only the glutamine-dependent activity in bovine pancreatic asparagine synthetase (see Table I). The glutamine analog DON specifically inhibits the glutaminase and glutamine-dependent activity of bovine
pancreatic asparagine synthetase, and therefore could be proposed to bind to the glutamine binding site. This is supported strongly by the observation that the ammoniadependent asparagine synthesis activity is completely uninhibited under conditions where glutaminase and glutamine-dependent synthetase activities are totally eliminated. Further support comes from the fact that glutamine was the only substrate to reduce the DON-dependent shift to dimer, and reduce the effectiveness of DON as an inhibitor. There is a great deal of precedent from work with many other glutamine amido transferases that shows that DON is specific for glutamine binding sites (13-15, 17, 19), so inhibition by DON is consistent with DON binding to the glutamine site. Therefore, the observation presented here that DON eliminates binding by Mab 5A6 is highly suggestive that the two are competing for the same site, i.e., the glutamine binding site. It must also be noted, however, that the dimerization caused by DON
TABLE
III
Substrate Protection of the DON-Dependent Shift to Dimer
Substrate
Dimer area % control
Monomer area % control
Control (no DON) DON only Glutamine + DON Ammonia + DON Aspartate + DON Mg-ATP + DON
100 151 90 135 122 151
100 44 91 37 57 40
Note. Saturating concentrations of substrates were added to the enzyme, DON was then added to a final concentration of 0.5 mM, and the sample was immediately chromatographed on a gel filtrat,ion HPLC column.
150
MEHLHAFF
AND
modification could also effectively block access of antibodies to sites at or near subunit-subunit interfaces. It is also noteworthy that DON could be inducing some conformational changes that preclude antibody binding at sites distinct from the glutamine binding sites. This is especially difficult to rule out since DON affects the quaternary structure of asparagine synthetase. Other studies have shown that DON may bind in specific ratios, e.g., Hartman showed that in 5-phosphoribosyl-pyrophosphate amidotransferase, one mole of DON binds per mole of tetramer (17), but no other studies have shown that DON can actually be responsible for a change in the quaternary structure as seen here when DON causes a shift from the low molecular weight (monomer) form to the high molecular weight (dimer) form of asparagine synthetase. Observing the disappearance of the enzyme peak from the chromatogram of the gel filtration column is an effective way of monitoring the binding of an antibody. Since the presence of DON effectively prevents the binding of Mab 5A6 and not Mab 2B4 or others, this suggests, but does not prove, that Mab 5A6 inhibits glutaminase and glutamine-dependent activity by actually binding to the glutamine amino acid binding site and not by steric or conformation effects. If this is actually the case, Mab 5A6 could theoretically be used to pick out the fragment corresponding to the glutamine binding site from a sample of chemically or enzymatically cleaved enzyme. Once the correct fragment had been obtained using Mab 5A6 then the corresponding nucleotide sequencecan be determined and sites for mutagenesis chosen. These types of studies require further exploration. Nonetheless, it is apparent from these studies that asparagine synthetase may undergo substrate-dependent changes in quaternary structure. Both of these observations will provide great insight into the mechanism and regulation of this metabolically important enzyme. ACKNOWLEDGMENT This stitute,
work was supported by a grant from the National Cancer Department of Health and Human Services (CA 28725).
In-
REFERENCES 1. Prager,
M. D., and Bachynsky,
127,645-654.
N. (1968)
Arch.
Biochem.
Biophys.
SCHUSTER 2. Mehlhaff, istry 24,
P. M., Luehr, 1104-1110.
3. Meister, 561-580,
A. (1974) Academic
C. A., and Schuster,
in The Enzymes (Boyer, Press, New York.
4. Hongo, S., Matsumoto, Acta 522, 258-266.
T., and Sato,
5. Luehr, C. A., and Schuster, 237, 335-346. 6. Markin, R. S., Luehr, 20, 7226-7232. I. Pfeiffer, N. P., (1987) J. Biol. 8. Pfeiffer, N. P., (1986) J. Biol. 9. Markin, Commun. 10. Markin,
S. M.
S. M. (1985) P. D., Ed.),
T. (1978)
(1985)
Arch.
C. A., and Schuster,
Rio&emVol. 10, pp.
Biochim.
Biophys.
Biochem.
Biophys.
S. M. (1981)
Biochemistry
Mehlhaff, P. M., Wylie, D. E., and Schuster, Chem. 262, 11,565-11,570. Mehlhaff, P. M., Wylie, D. E., and Schuster, Chen. 261, 1914-1919.
R. S., and Schuster, S&583-588. R. S., and Schuster,
11. Luehr, C. A., and Schuster, ods 3, 151-161.
S. M. (1979) S. M. (1979) S. M. (1980)
Biochem.
S. M.
Biophys.
Metabolism J. Biochem.
S. M.
Res.
33,891-896. Biophys.
Meth-
12. Khedouri, E., Anderson, P. M., and Meister, A. (1966) Biochemistry 5, 355223557. L. J., and Zalkin, H. (1979) J. Biol. Chem. 254, 338213. Messenger, 3392. 14. Pinkus, L. M. (1977) in Methods in Enzymology (Jakoby, W. B., and Wilchek, M., Eds.), Vol. 46, pp. 414-435, Academic Press, New York. J. M. (1973) in Advances in Enzymology and Related 15. Buchanan, Area on Molecular Biology (Meister, A. Ed.), Vol. 39, pp. 91-183, Wiley, New York. 16. Pinkus, 6127. 17. Hartman,
L. M.,
and Meister,
S. C. (1963)
J. Biol.
A. (1972) Chem.
J. Biol. 238,
Chem.
247,
6119-
3036-3047.
18. Zalkin, H. (1985) in Methods in Enzymology (Meister, A. Ed.), Vol. 113, pp. 2633305, Academic Press, New York. H. N., Cooney, D. A., Milman, H. A., Homan, E. R., and 19. Jayaram, Rosenbluth, R. J. (1976) Biochem. Pharmacol. 25, 1571-1582. 20. Rosenbluth, and Homan,
R. J., Cooney, D. A., Jayaram, H. N., Milman, H. A., E. R. (1976) Biochem. Pharmacol. 25, 1851-1858.
21. Cooney, D. A., Jayaram, H. N., Milman, H. A., Homan, E. R., Pittillo, R., Geran, R. I., Ryan, J., and Rosenbluth, R. J. (1976) Biochem. Pharmacol. 25, 1859-1870. B., and Meister, A. (1972) J. Biol. Chem. 247, 670822. Horowitz, 6719. D. D., Allisone, A. J., and Buchanan, J. M. (1969) J. 23. Schroeder, Biol. Chem. 244, 5856-5865. 24. French, T. C., Dawid, I. B., and Buchanan, J. M. (1963) J. Biol. Chem. 238,2186-2193. 25. Milman, H. A., and Cooney, D. A. (1978) Biochem. Pharmacol. 27, 151-166.
Vol.
OF BIOCHEMISTRY
284, No. 1, January,
AND
BIOPHYSICS
pp. 143-150,
1991
Bovine Pancreatic Asparagine Synthetase Explored with Substrate Analogs and Specific Monoclonal Antibodies Paige
M. Mehlhaff*
and Sheldon
M. Schustert,’
TDepartment of Biochemistry and Molecular Biology, University of Florida, Gainesville, and *Department of Pathology, University of California, La Jolla, California 92093
Received
May
17, 1990, and in revised
form
August
L-Asparagine synthetase (EC 6.3.1.1) is the enzyme responsible for the production of asparagine from aspartate and a nitrogen source. In mammalian tissues the enzyme can utilize either glutamine or ammonia as the nitrogen
0003.9861/91$3.00
Copyright All rights
0
1991 by of reproduction
Academic in any
Press, Inc. form
reserved.
32610;
28, 1990
Several substrate analogs were tested for their ability to inhibit bovine pancreatic asparagine synthetase. Of the substrate analogs tested both 6-diazo-5-oxo-L-norleucine (DON) and 5-chloro-4-oxo-L-norvaline (CONV) were shown to inhibit the enzyme strongly. DON inhibited the glutaminase and glutamine-dependent asparagine synthetase activities and CONV inhibited the ammonia-dependent activity as well. Both of these inhibitors appeared to be relatively tight binding since desalting failed to remove the inhibition. The inactivation of bovine pancreatic asparagine synthetase by DON is accompanied by a shift from a 47,000 molecular weight monomer to a 96,000 molecular weight dimer as observed by HPLC gel filtration chromatography. This DON-induced shift is prevented by the presence of the substrate glutamine. A monoclonal antibody known to inhibit specifically the ammonia-dependent and glutamine-dependent asparagine synthetase activities but not glutaminase (monoclonal antibody 2B4) binds to both the monomer and the dimer forms of untreated enzyme, as well as to the dimer form of the DON-inactivated enzyme. On the other hand, a monoclonal antibody known to inhibit specifically the glutaminase and glutamine-dependent activities and not the ammonia-dependent asparagine synthetase (monoclonal antibody 5A6) binds to both forms of untreated enzyme but cannot bind to the DON-inactivated enzyme. These data are used to describe the relation of regions of the active site of asparagine synthetase in relation to ano 1991 Academic press, I~C. tibody binding sites.
I To whom correspondence should be addressed University of Florida, Gainesville, Fla 32611.
Florida
at 1301 Fifield
Hall,
source in the ATP-dependent synthesis of asparagine. The glutamine-dependent activity can be depicted as L-glutamine + L-aspartate + Mg-ATP
+
L-asparagine + AMP + PPi + Mg2+, and the ammonia-dependent reaction as (1, 2) NH3 + L-aspartate + Mg-ATP
+
L-asparagine + AMP + PPi + Mg’+. In addition, asparagine synthetase can function as a glutaminase in the absence of Mg-ATP and aspartate (3-5), L-glutamine + L-glutamate + NH3. In earlier studies from this laboratory it was shown that when glutamine is provided as the nitrogen source, bovine pancreatic asparagine synthetase exhibits an ordered binding mechanism where glutamine binds first followed by Mg-ATP. Glutamate is then released, aspartate is allowed to bind, and the rest of the products are released in the order of pyrophosphate, AMP, and finally asparagine. Conversely when ammonia is the nitrogen source, both Mg-ATP and ammonia bind in a random order, then aspartate binds and the products are then released in the same order (6). In previous work several monoclonal antibodies (Mabs)2 were produced against bovine pancreatic asparagine synthetase. One group of Mabs binds to the enzyme but does not affect activity. Mabs from this group were used to purify the enzyme. Another set of Mabs were * Abbreviations used: CONV, 5-chloro-4-oxo-L-noraline; DON, 6diazo-5-oxo-L-norleucine; DTT, dithiothreitol; Mab, monoclonal antibody; PBS, phosphate-buffered saline; PP, pyrophosphate; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; FGAR, formylglycinamide ribonucleotide amidotransferase. 143
144
MEHLHAFF
AND
shown to be inhibitory, particularly Mab 5A6, which inhibited glutamine-dependent asparagine synthetase and glutaminase activity, but not ammonia-dependent asparagine synthesis. Mab 2B4 and Mab 3F3 inhibited both the glutamine-dependent and the ammonia-dependent asparagine synthetase activities but not the glutaminase activity. This differential inhibition by the Mabs showed that there are two distinct amino acid binding domains on the enzyme-one for asparagine (and aspartate) and the other most likely for glutamine (and glutamate). Additional studies with the Mabs also suggested that the binding sites are topographically separated from each other on the enzyme surface (7). Several important questions remain regarding the relation of the binding sites for the antibodies and the sites for amino acid binding. These center around the key question of whether or not there is a region of contiguous primary sequence of the asparagine synthetase that is specifically involved in the various portions of the amino acid binding sites. The binding of the antibodies previously characterized has already been shown to be related to the binding domains in several distinct antigenic regions (7, 8). However, it could not be determined if the binding and subsequent inhibition was due to complexation of the glutamine binding site or was due to conformational effects caused by interaction with a site distinct from the glutamine binding site. If other inhibitorspreferentially covalent, specific modifiers-could be identified and kinetically characterized, and their effect on antibody binding determined, then it would be possible to distinguish between these possibilities. In addition, if the interaction were directly to the glutamine binding site, then we could begin determining the region of primary sequence essential for amino acid binding. To this date, however, little is known regarding inhibitors of asparagine synthetase, and nothing is known regarding how any might effect the binding of antibodies known to inhibit specific functions of the enzyme. This work reports attempts to probe the structurefunction relationship of the enzyme and the various antibodies available with substrate analogs. One particular substrate analog, DON, is shown to be an apparently covalent modifier of asparagine synthetase from beef pancreas that specifically effects only glutamine-dependent function. The effect of the binding of DON on antibody binding is also described. It has been known that asparagine synthetase is a dimer of identical subunits (9, lo), but no information has been presented in the literature regarding the interaction of the subunits, and the relation of these interactions to the catalytic function of the enzyme. These studies also provide information regarding the quaternary structure of asparagine synthetase. The major conclusion, however, is that the data are consistent with the hypothesis that the binding of Mab 5A6 to asparagine synthetase is directly to the glutamine binding site.
SCHUSTER
EXPERIMENTAL
PROCEDURES
Monoclonal antibodies. The production and growth of hybridomas producing monoclonal antibodies specific for bovine pancreatic asparagine synthetase has been described previously (8). Antibodies were purified from ascites fluids by precipitation to 45% with saturated ammonium sulfate, followed by a second similar precipitation. Asparagine synthetase purification. Bovine pancreas was obtained from Spencer Beef (Schuyler, NE) or Iowa Beef Packers (West Point, NE) and purified as previously described (8) with modifications as indicated by Pfeiffer et al. (7). Briefly, the purification involves homogenizing the pancreas in Tris buffer at pH 8.0 (50 mM Tris-Cl, pH 8.0, 0.5 mM EDTA, 1.0 mM dithiothreitol). After centrifugation, the homogenate is applied to an immunoaffinity column, the column is washed with PBS, and the enzyme is eluted with sodium carbonate at pH 10.6. The eluted enzyme is precipitated by adding solid ammonium sulfate to 50% saturation, the precipitate is collected by centrifugation, and the pellet is resuspended in Tris buffer. Before use, the enzyme preparation is desalted on a G-25 column equilibrated in the same Tris buffer. The enzyme prepared in this manner is extremely pure as judged on acrylamide gels stained by either Coomassie blue or silver stain (7). All experiments reported use only enzyme that yields a single band on SDSPAGE monitored by silver staining. Zmmunoufinity the asparagine Mab 3F3.
columns. The immunoaffinity synthetase was made as previously
column used to purify described (7,8) using
Asparugine synthetase assay. The glutamineand ammonia-dependent synthesis of asparagine was assayed as previously described (11). In this assay either glutamine or ammonia, Mg-ATP and L-[414C]aspartate are incubated with the enzyme at 37°C for the desired time period. The reaction is stopped by pipetting the reaction mixture into hot sodium acetate. The radioactive label in the excess aspartate is removed in a decarboxylation reaction using pyridoxal and A&SO,. The label remaining in the asparagine is then quantitated using a Beckman liquid scintillation counter and the amount of asparagine is calculated by comparison to a nondecarboxylated L-[4-“Claspartate control performed with each experiment. Zeroes for each data point are obtained in the same manner except that the reaction is stopped immediately upon addition of the enzyme to the substrate mixture. The glutaminase activity was assayed as described by Pfeiffer et al. (8), where the formation of glutamate was coupled to the reduction of p-iodonitrotetrazolium violet and the absorbance monitored at 510 nm. Protein assay. The protein concentration zyme was determined using the Bio-Rad Richmond, CA).
in each preparation protein assay (Bio-Rad
of enCo.,
Gel filtration HPLC. The gel filtration HPLC was performed using a Bio-Rad TSK 250 gel filtration column and a Beckman HPLC pump equipped with a uv detector. Unless otherwise indicated, the column was equilibrated in 20 mM Bis-Tris at pH 6.5 with 150 mM Na$O,. Enzyme samples were mixed with control buffers, inhibitors, substrates, antibodies, etc. and then 50 to 100 aI of the mixture was immediately injected onto the column and the absorbance was monitored at 280 nm. The corresponding chromatograms and the integration of the peaks were obtained using an Apple IIE computer. Molecular weights were obtained by calculation by comparison to commercially available standards (Bio-Rad).
RESULTS Several substrate analogs were tested for their ability to inhibit bovine pancreatic asparagine synthetase. In most cases the inhibitor was incubated with the enzyme for 30 min at the designated concentration (preassay concentration) and then the enzyme-inhibitor mixture was added to the substrate mixture for assay. Thus the con-
SUBSTRATE
ANALOGS
AND ANTIBODIES
ON BOVINE
TABLE
ASPARAGINE
145
SYNTHETASE
I
Effect of Several Compoundson the Glutamine and Ammonia-Dependent Activity of AsparagineSynthetase Concentration Inhibitor Albizzin
Azaserine
CONV DON Maleimide
Preassay 10 10 30 10 10 30 10 5 40
Substrates
(mM)
Assay
Saturated
0.83 0.83 10 0.83 0.83 10 0.42
X
10
X
centration of the inhibitor during the assay was lessthan during the incubation time. Since saturating concentrations of substrates mask the action of a simple competitive inhibitor, some of the assays were performed at low or nonsaturating substrate concentrations in order to observe more readily any effect. The effects on asparagine synthetase activity of some of the compounds tested are shown in Table I. Albizzin, a glutamine analog, inhibited the glutamine-dependent asparagine synthesis activity of asparagine synthetase to 37% of the control activity only when the inhibitor concentration was high (10 mM during the assay) and the substrate concentrations were low (nonsaturating). At lower concentrations of albizzin there were only slight effects whether or not the substrates were in saturating concentrations. Azaserine, another glutamine analog, tested under the same conditions as albizzin showed no significant inhibition of either ammonia- or glutamine-dependent activity. At high concentrations (10 mM), 5-chloro-4-oxo-L-norvaline (CONV) inhibits both the glutamine- and the ammonia-dependent asparagine synthesis by bovine pancreatic asparagine synthetase. At 5 mM, DON totally inhibits the glutamine-dependent activity but has only a slight effect on the ammonia-dependent activity. Maleimide, a sulfhydryl modifier, exhibits significant inhibition of the glutamine-dependent asparagine synthetase activity (36% of the control) and a slight activation (118%) of the ammonia-dependent activity. Of the compounds tested, DON and CONV appeared to have the greatest effect on the pancreatic enzyme. In addition, in other caseswhere these compounds have been used, they were shown to be covalent, and highly specific to glutaminase functional sites (12). Therefore, DON was used for further studies of pancreatic asparagine synthetase. The concentration dependence of the DON inhibition of bovine pancreatic asparagine synthetase was examined by adding increasing amounts of DON to a preparation of asparagine synthetase and measuring the three activ-
% Activity Nonsaturated
X X X X
X X X
remaining
Gln
NH3
105 91 37 105 91 95 0 1 36
111 95 90 113 95 100 0 86 118
ities relative to a control without DON. As seen in Fig. 1, in the presence of DON, the glutamine-dependent asparagine synthetase activity rapidly decreased with increasing concentrations of DON until 2 mM, where the activity was virtually zero. The glutaminase activity was inactivated even more rapidly and was completely abolished at 1 mM DON. On the other hand, the ammoniadependent activity was not inhibited at all in the presence of DON and appeared to be slightly increased, even at concentrations as high as 20 mM. In order to determine if the inactivation of asparagine synthetase was due to covalent modification of the enzyme, the DON-inactivated asparagine synthetase was passedthrough a desalting column (Bio-Gel P-GDG, BioRad) and reassayed for activity. There was very little change in the extent of inhibition after passagethrough the column, indicating that the DON molecule is not readily removed from the enzyme (see Table II).
0
IO
20
30
DON (mt-l) FIG. 1. DON inhibition of asparagine synthetase. The percentage activity remaining relative to a sample with no DON, for the ammoniadependent synthesis (m), the glutamine-dependent synthesis (O), and the glutaminase (A) activities is plotted versus increasing DON concentrations (0, 1.0, 2.0, 5.0, 10.0, and 25.0 mM).
146
MEHLHAFF
TABLE Activity
of DON-Inactivated Passage through
AND
II
Asparagine Synthetase a G-25 Column
after
Specific activity of asparagine synthetase (nmol/min/mg) Sample Initial enzyme activity DON-inactivated enzyme DON-inactivated enzyme after G-25 chromatography
Glutamine dependent
Ammonia dependent
156 15
167 188
25
185
Since the asparagine synthetase is capable of three different enzymatic activities, it is of interest to determine if the interaction of the subunits affects any of the various catalytic properties of the enzyme. Indeed it is crucial to know if the effects of DON, for example, could be due to changes in quaternary structure. However, there is no basic information regarding the quaternary structure of the enzyme. For a better understanding of the enzymesubstrate interaction, some studies were done to examine some of the factors responsible for the interaction of the subunits, especially in regard to substrates, antibodies, and substrate analogs. A purified preparation of bovine pancreatic asparagine synthetase was subjected to gel filtration chromatography, and the chromatogram shown in Fig. 2A was obtained. It should be noted that the buffer used contained 150 mM NaPSO to maintain enzyme solubility. In buffers of lower ionic strength, protein precipitation is observed. It should also be noted
that the absorbance
peaks observed
SCHUSTER
monia-dependent (-lo-fold greater) or the glutaminedependent activity (-20-fold greater). After storage at -20°C there is a marked decrease in all three activities with the greatest loss in the glutamine-dependent synthetase activity followed by the glutaminase activity with the ammonia-dependent synthetase activity exhibiting the greatest stability (data not shown). As noted above, it was crucial to determine if the effects of DON on activity are due to changes in subunit interactions. When the enzyme preparation shown in Fig. 2A was treated with DON at a final concentration of 1 mM and subsequently chromatographed on the gel filtration column, the result is that shown in Fig. 2B. As can be seen, the presence of DON appears to shift the enzyme to its dimer form. Again, the species eluting after 13 min are DON and storage buffer components. The shift to the dimer form as a function of DON concentration is shown in Fig. 3. The integrated area corresponding to both the dimer and the monomer are almost equal when no DON is present but as DON is added the area corresponding to the monomer rapidly decreases to a minimum value
in these
elutions after about 13 min are due to the absorbance or refraction of salts and solutes present in the enzyme storage. Controls done without enzyme but with storage buffer components (Tris, EDTA, DTT, etc.), show these same peaks. There are two major forms of the enzyme present under these conditions, a high molecular weight form (96,000), corresponding to the dimer form, and a smaller (47,000) form, corresponding to the monomer form. These two forms are present in all enzyme preparations tested although the relative amounts vary from preparation to preparation. The two peaks have been collected separately and assayed, and both exhibit all three activities i.e., glutaminase and glutamine- and ammonia-dependent asparagine synthesis (data not shown). In all casesthe enzyme prior to chromatographic separation has higher amounts of ammonia-dependent activity than glutaminedependent activity (usually by a factor of two). After gel filtration chromatography the ratio of ammonia-dependent to glutamine-dependent activity is closer to one for the dimer and less than one for the monomer. The glutaminase activity is much greater than either the am-
I
10 Elution
Time
15 (min)
FIG. 2. Chromatogram of asparagine synthetase using HPLC gel filtration. A is untreated enzyme, showing the dimer (peak 1) and the monomer (peak 2) forms. B is the same enzyme after treatment with 1.0 mM DON.
SUBSTRATE
ANALOGS
AND
ANTIBODIES
80
E; 0 0
I
60
.z g
40
+ -Y 2
20
0
0
L
IO
20
30
DON (mM) FIG. 3. Area of dimer thetase after inactivation
(m) and monomer (0) peaks of asparagine synwith 0, 1.0, 2.0, 5.0, 9.3, and 25 mM DON.
while the area corresponding to the dimer increases to a maximum and then levels off. It is of interest to determine if DON binding competes for antibody binding to asparagine synthetase, especially since so much is known regarding antibodies affecting specific function. Two of the monoclonal antibodies previously shown to inhibit the asparagine synthetase activities differentially were tested for their ability to bind in the presence and absence of the DON. The binding of the monoclonal antibody to the enzyme can be visualized on the gel filtration column by adding increasing amounts of antibody to the enzyme and observing the disappearance of the enzyme peaks and the concomitant emergence of a very large molecular weight peak corresponding to the antibody-enzyme complex. Figure 4 shows three of the gel filtration chromatograms obtained when increasing amounts of Mab 2B4 are added to a preparation of asparagine synthetase. Panel A shows the enzyme alone with no antibody added. Both the dimer and the monomer forms of the enzyme are clear. When a 0.1 mg/ml solution of Mab 2B4 is added after mixing with asparagine synthetase (B) both of the enzyme peaks are markedly reduced and a new, large molecular weight peak emerges.When the antibody concentration is further increased (C), the two enzyme peaks completely disappear and two large molecular weight peaks remain. It is likely that the earliest eluting peak represents the antibodyenzyme complex, while the other is in the position where free Mab 2B4 elutes, and thus represents excess antibody. When Mab 2B4 is added to an enzyme preparation previously inactivated with DON the results shown in Fig. 5 are obtained. Panel A shows the DON-inactivated enzyme in the absence of antibody. As before, there is only the dimer form of the enzyme in the presence of DON. The addition of increasing amounts of Mab 2B4 (B-D) results in the disappearance of the enzyme peak and the appearance of the large molecular weight species as occurred in the absence of DON. The same experiment was
ON
BOVINE
ASPARAGINE
147
SYNTHETASE
performed with Mab 5A6, which inhibits glutamine-dependent asparagine synthesis and glutaminase activities, but not ammonia-dependent asparagine synthesis. As can be seen in Fig. 6, the monomer and dimer forms of the enzyme are removed by Mab 5A6. When Mab 5A6 was added to DON-inactivated enzyme, somewhat different results were obtained. As shown in Fig. 7, the DON-inactivated enzyme is present only in dimer form in the absence of antibody (panel A). As increasing amounts of Mab 5A6 are added (B-D) there is an increase in the amount of the peak eluting at approximately 8 min, but no noticeable decrease in the dimer peak. Even at very high concentrations of Mab 5A6 (Fig. 7D) there is only a slight decrease in the area corresponding to the dimer of asparagine synthetase. Again, in all of these figures, the material eluting after about 13 min was shown to be buffer components. The ability of the monoclonal antibodies to bind to the DON-inactivated enzyme is presented graphically in Fig. 8. As the amount of antibody is increased, the area of the dimer decreases in a linear fashion for both Mabs 2B4 and 3F3 (Mab 3F3 also inhibits the ammonia-dependent and glutamine-dependent asparagine synthetase activities and not the glutaminase activity as does Mab 2B4 but not as strongly), whereas for Mab 5A6 almost no decrease 1
A
ALJL i
5
Elution
10
Time
(min)
FIG. 4. Chromatograms from gel filtration of asparagine synthetase with Mab 2B4. A is asparagine synthetase alone, with peak 1 the dimer form and peak 2 the monomer form. In B a 0.1 mg/ml solution of Mab 2B4 was mixed with asparagine synthetase, then added. Peak 1 and 2 are the dimer and monomer forms, respectively, and peak 3 is the antigen-antibody complex. In C, more Mab 2B4 has been added (a 0.1 mg/ml solution); peak 4 represents excess uncomplexed antibody.
148
MEHLHAFF
AND
SCHIJSTER
of asparagine synthetase are present. When smaI1 amounts of DON are added, there is a 51% increase in the area corresponding to the dimer peak and simuitaneously a 56% decrease in the area corresponding to the monomer. The presence of MgATP has no effect on the shift from the monomer to the dimer forms, and ammonia and aspartate have only small effects. Only g1utamine appears to have a significant effect, since there is no increase in area corresponding to the dimer with the addition of DON if glutamine is present, indicating that the glut~ine may actually prevent DON from binding to the enzyme.
5
Elution
FIG. 5. Chromatogram from gel paragine synthetase with Mab 2B4. the dimer form (peak 1). In B a 0.1 added with peak 3 again being the had 0.25 mg/ml Mab 2B4 added and the excess uncomplexed Mab 2B4.
Asparagine synthetase is one of many enzymes comprising the class of glutamine amidotransferases. The g~utamine amidotransferases exhibit a wide variety of functions including biosynthesis of amino acids, nucleotides, and coenzymes among others (18). Many of these amidotransferases have also been studied using substrate analogs. Albizzin was shown by Schroeder et ai. specifically to inhibit formylglycinamide ribonucleotide amidotransferase (FGAR) (23). As with asparagine synthetase, albizzin inhibited only the g~utamine-dependent activity of FGAR and not the ammonia-dependent activity. Nonetheless it did so at a much lower concentration than was needed for inhibition of the asparagine synthe10
15
Time fminl filtration of DON-inactivated asA is DON-inactivated enzyme in mg/ml solution of 934 had been ant.igen-antibody complex. C has D 0.5 mg/mI Mab 284. Peak 4 is
in dimer area is seen over the same range of antibody concentration. Since the shift from monomer to dimer upon addition of DON occurs simultaneously with the inactivation of the glutamine-dependent asparagine synthetase activity, monitoring the relative amounts of dimer and monomer is a convenient way of determining the extent of inactivation and also the ability of various compounds to protect the enzyme from inactivation by DON. In order to determine if the substrates normally utilized by asparagine synthetase could protect the enzyme from inactivation and the concurrent shift to the dimer form, the various substrates were added to the enzyme at concentrations normally used for assaying activity at V,,,. DON was then added, the samples were immediately put on the gel filtration column, and the area corresponding to the dimer and monomer peaks was obtained. These results are shown in Table III. In the absence of DON, significant amounts of both dimer and monomer forms
5
10
Elution Time (mid FIG. 6, Chromatugram from a gel fiitration column of asparagine synthetase with Mab 5A6. A has no Mab added and peak t and 2 are the dimer and monomer forms, respectively. B has had a 0.25 mg/ml solution of Mab 5A6 added and C a 0.5 mg/ml solution of Mab 5A6.
SUBSTRATE
ANALOGS
AND
ANTIBODIES
ON
BOVINE
c
ASPARAGINE
0.0
149
SYNTHETASE
0.2
0.4
0.6
0.8
1.o
I .2
mg Mab per mg Enzyme
1 4
FIG. 8.
Plot of the dimer peak of asparagine synthetase with respect to the concentration of Mab 5A6 (W), Mab 3F3 (Cl), and Mab 2B4 (A).
A 4
D
1
~
~ 5
10
El&ion
Time (mid
1
15
FIG. 7.
Chromatogram from a gel filtration column of aaparagine synthetase with Mab 5A6. A is DON-inactivated enzyme in the dimer form (peak 1). In B a 0.1 mg/ml solution of 5A6 had been added. C has had 0.25 mg/ml Mab 5A6 added and panel D 0.5 mg/ml Mab 5A6. Peak 4 is the excess uncomplexed Mab 2B4.
tase (23). Azaserine also inhibited FGAR and French et al. provided evidence that azaserine alkylates a particular sulfhydryl group on the enzyme (24). But as shown in Table I, azaserine had no effect on any of the asparagine synthetase activities. Although azaserine had no effect, another sulfhydryl modifying reagent, maleimide, did inhibit bovine pancreatic asparagine synthetase. Milman and Cooney have shown that maleimide produced a 94% inhibition of glutamine-dependent asparagine synthetase from L5178,fAR and a 47% inhibition of ammonia-dependent asparagine synthetase activity from the same source (25). A similar inhibition pattern was observed in asparagine synthetase from mouse pancreas and the authors propose that maleimide reacts with the enzymes via a sulfhydryi alkylation (25). Unlike the results of Milman and Cooney a similar concentration of maleimide inhibits only the glutamine-dependent activity in bovine pancreatic asparagine synthetase (see Table I). The glutamine analog DON specifically inhibits the glutaminase and glutamine-dependent activity of bovine
pancreatic asparagine synthetase, and therefore could be proposed to bind to the glutamine binding site. This is supported strongly by the observation that the ammoniadependent asparagine synthesis activity is completely uninhibited under conditions where glutaminase and glutamine-dependent synthetase activities are totally eliminated. Further support comes from the fact that glutamine was the only substrate to reduce the DON-dependent shift to dimer, and reduce the effectiveness of DON as an inhibitor. There is a great deal of precedent from work with many other glutamine amido transferases that shows that DON is specific for glutamine binding sites (13-15, 17, 19), so inhibition by DON is consistent with DON binding to the glutamine site. Therefore, the observation presented here that DON eliminates binding by Mab 5A6 is highly suggestive that the two are competing for the same site, i.e., the glutamine binding site. It must also be noted, however, that the dimerization caused by DON
TABLE
III
Substrate Protection of the DON-Dependent Shift to Dimer
Substrate
Dimer area % control
Monomer area % control
Control (no DON) DON only Glutamine + DON Ammonia + DON Aspartate + DON Mg-ATP + DON
100 151 90 135 122 151
100 44 91 37 57 40
Note. Saturating concentrations of substrates were added to the enzyme, DON was then added to a final concentration of 0.5 mM, and the sample was immediately chromatographed on a gel filtrat,ion HPLC column.
150
MEHLHAFF
AND
modification could also effectively block access of antibodies to sites at or near subunit-subunit interfaces. It is also noteworthy that DON could be inducing some conformational changes that preclude antibody binding at sites distinct from the glutamine binding sites. This is especially difficult to rule out since DON affects the quaternary structure of asparagine synthetase. Other studies have shown that DON may bind in specific ratios, e.g., Hartman showed that in 5-phosphoribosyl-pyrophosphate amidotransferase, one mole of DON binds per mole of tetramer (17), but no other studies have shown that DON can actually be responsible for a change in the quaternary structure as seen here when DON causes a shift from the low molecular weight (monomer) form to the high molecular weight (dimer) form of asparagine synthetase. Observing the disappearance of the enzyme peak from the chromatogram of the gel filtration column is an effective way of monitoring the binding of an antibody. Since the presence of DON effectively prevents the binding of Mab 5A6 and not Mab 2B4 or others, this suggests, but does not prove, that Mab 5A6 inhibits glutaminase and glutamine-dependent activity by actually binding to the glutamine amino acid binding site and not by steric or conformation effects. If this is actually the case, Mab 5A6 could theoretically be used to pick out the fragment corresponding to the glutamine binding site from a sample of chemically or enzymatically cleaved enzyme. Once the correct fragment had been obtained using Mab 5A6 then the corresponding nucleotide sequencecan be determined and sites for mutagenesis chosen. These types of studies require further exploration. Nonetheless, it is apparent from these studies that asparagine synthetase may undergo substrate-dependent changes in quaternary structure. Both of these observations will provide great insight into the mechanism and regulation of this metabolically important enzyme. ACKNOWLEDGMENT This stitute,
work was supported by a grant from the National Cancer Department of Health and Human Services (CA 28725).
In-
REFERENCES 1. Prager,
M. D., and Bachynsky,
127,645-654.
N. (1968)
Arch.
Biochem.
Biophys.
SCHUSTER 2. Mehlhaff, istry 24,
P. M., Luehr, 1104-1110.
3. Meister, 561-580,
A. (1974) Academic
C. A., and Schuster,
in The Enzymes (Boyer, Press, New York.
4. Hongo, S., Matsumoto, Acta 522, 258-266.
T., and Sato,
5. Luehr, C. A., and Schuster, 237, 335-346. 6. Markin, R. S., Luehr, 20, 7226-7232. I. Pfeiffer, N. P., (1987) J. Biol. 8. Pfeiffer, N. P., (1986) J. Biol. 9. Markin, Commun. 10. Markin,
S. M.
S. M. (1985) P. D., Ed.),
T. (1978)
(1985)
Arch.
C. A., and Schuster,
Rio&emVol. 10, pp.
Biochim.
Biophys.
Biochem.
Biophys.
S. M. (1981)
Biochemistry
Mehlhaff, P. M., Wylie, D. E., and Schuster, Chem. 262, 11,565-11,570. Mehlhaff, P. M., Wylie, D. E., and Schuster, Chen. 261, 1914-1919.
R. S., and Schuster, S&583-588. R. S., and Schuster,
11. Luehr, C. A., and Schuster, ods 3, 151-161.
S. M. (1979) S. M. (1979) S. M. (1980)
Biochem.
S. M.
Biophys.
Metabolism J. Biochem.
S. M.
Res.
33,891-896. Biophys.
Meth-
12. Khedouri, E., Anderson, P. M., and Meister, A. (1966) Biochemistry 5, 355223557. L. J., and Zalkin, H. (1979) J. Biol. Chem. 254, 338213. Messenger, 3392. 14. Pinkus, L. M. (1977) in Methods in Enzymology (Jakoby, W. B., and Wilchek, M., Eds.), Vol. 46, pp. 414-435, Academic Press, New York. J. M. (1973) in Advances in Enzymology and Related 15. Buchanan, Area on Molecular Biology (Meister, A. Ed.), Vol. 39, pp. 91-183, Wiley, New York. 16. Pinkus, 6127. 17. Hartman,
L. M.,
and Meister,
S. C. (1963)
J. Biol.
A. (1972) Chem.
J. Biol. 238,
Chem.
247,
6119-
3036-3047.
18. Zalkin, H. (1985) in Methods in Enzymology (Meister, A. Ed.), Vol. 113, pp. 2633305, Academic Press, New York. H. N., Cooney, D. A., Milman, H. A., Homan, E. R., and 19. Jayaram, Rosenbluth, R. J. (1976) Biochem. Pharmacol. 25, 1571-1582. 20. Rosenbluth, and Homan,
R. J., Cooney, D. A., Jayaram, H. N., Milman, H. A., E. R. (1976) Biochem. Pharmacol. 25, 1851-1858.
21. Cooney, D. A., Jayaram, H. N., Milman, H. A., Homan, E. R., Pittillo, R., Geran, R. I., Ryan, J., and Rosenbluth, R. J. (1976) Biochem. Pharmacol. 25, 1859-1870. B., and Meister, A. (1972) J. Biol. Chem. 247, 670822. Horowitz, 6719. D. D., Allisone, A. J., and Buchanan, J. M. (1969) J. 23. Schroeder, Biol. Chem. 244, 5856-5865. 24. French, T. C., Dawid, I. B., and Buchanan, J. M. (1963) J. Biol. Chem. 238,2186-2193. 25. Milman, H. A., and Cooney, D. A. (1978) Biochem. Pharmacol. 27, 151-166.
