A Chromophoric and Fluorescent Analog of GTP, 2’,3’-0-(2,4,6-Trinitrocyclohexadieny1idene)-GTP, as a Spectroscopic Probe for the GTP Inhibitory Site of Liver Glutamate Dehydrogenase*

The ribose-modified chromophoric and fluorescent analog of ATP, 2’,3’-0-(2,4,6-trinitrocyclohexadien-y1idene)-ATP (TNP-ATP) (Hiratsuka, T., and Uchida, K. (1973) Biochim. Biophys. Acta 320,635-647 and Hiratsuka, T. (1976) Biochim. Biophys. Acta 453, 293-297) has been widely used as an ATP analog for various ATPases. Although the corresponding analog of GTP, 2‘,3‘-0-(2,4,6-trinitrocyclohexadienylidene)-GTP (TNP-GTP) should be useful for the study of various GTP-requiring enzymes, it is difficult to prepare TNP-GTP by the conventional method. In the present study, we succeeded in the synthesis of TNP-GTP with the use of an alternative method. The analogs of GDP, GMP, and guanyl-5’-yl imidodi-phosphate (Gpp(NH)p) were also synthesized. Visible absorption and fluorescent properties of TNP-GTP, TNP-GDP, TNP-GMP, and TNP-Gpp(NH)p were quite similar to those of TNP-ATP. TNP-GTP was found to be able to replace GTP as an inhibitor for

enzymes, i.e. mammalian succinyl-CoA synthetase (l), phosphoenolpyruvate carboxykinase (2) and guanylate cyclase ( 3 , are specific for GTP. One approach to study mechanisms of these biological reactions is the use of a reporter-labeled GTP, i.e. a chromophoric or fluorescent analog of GTP. Ribosemodified GTP analogs (4)(5)(6) are especially useful for GTPrequiring enzymes which are sensitive to alteration in the base or phosphoryl moiety of the nucleotide.
In the present study, we succeeded in the synthesis of TNP-GTP, TNP-GDP, TNP-GMP, and TNP-Gpp(NH)p using TNCB instead of TNBS. Here we describe the synthesis of these analogs and their spectroscopic and biological properties. The studies reported here demonstrate that visible absorption and fluorescent properties of analogs are quite similar to those of TNP-ATP (8), and TNP-GTP can replace GTP as an allosteric inhibitor for bovine liver glutamate dehydrogenase. TNP-GTP, therefore, should be a suitable spectroscopic probe for the study of various GTP-requiring enzymes. ringer Mannheim. TNP-ATP was prepared by the method described previously (7,8). TNCB was from Tokyo Kasei Kogyo Co. and was used without further purification. Silica gel (Silica Gel 60) and cellulose (Avicel SF) TLC plates were from Merck and Funakoshi Chemical Co., respectively. Other reagents were of reagent or biochemical research grade.

TNP-GTP
Bovine liver glutamate dehydrogenase, purchased as a crystalline suspension in ammonium sulfate from Sigma, was dialyzed for 20 h at 4 "C against three changes of 0.1 M potassium phosphate buffer (pH 7.0). The dialyzed material was centrifuged for 15 min at 18,000 rpm to remove insoluble material. The protein concentration was determined with the value = 9.7 (27). The enzyme was used within 4 days. A molecular weight of 56,100 for identical peptide chains was used in the calculations (28).
Enzyme Assays-Glutamate dehydrogenase activity was assayed spectrophotometrically at 340 nm by measuring the oxidation of reduced coenzyme at 25 "C in 10 mM Tris acetate (pH 8.0) and 10 p~ EDTA. The substrate concentrations used were 5 mM a-ketoglutarate, 50 mM ammonium chloride, and 100 p~ NADH. The total volume of the assay solution was 3.0 ml. When the activity was measured in the presence of GTP and the analog as inhibitors, the nucleotide concentrations used were 0.02-1.9 p~ for GTP and 0.4-41 p~ for the analog. The inhibition constant, K,, was taken as the concentration of GTP or the analog required for half-maximum inhibition.
Spectral Measurements-Absorption spectra were measured at room temperature with a Shimadzu double-beam spectrophotometer, Model UV-200. Fluorescence emission spectra (uncorrected) were recorded at 25 "C in a thermostated Hitachi fluorescence spectrophotometer, Model MPF-4, as described previously (8). Excitation wavelengths were 408 or 470 nm. The slit widths on the excitation and emission monochromators were 5 or 10 nm.
TNP-GTP Binding Measured by Fluorescence Titration-The binding of TNP-GTP to glutamate dehydrogenase was measured at 25 "C in the presence and absence of 100 p~ NADH in 10 mM Tris acetate (pH 8.0) containing 7 mM potassium phosphate and 10 p~ EDTA. Samples were excited at 470 nm to avoid errors arising from absorption by NADH, and emission was measured at 548 and 545 nm in the presence and absence of the nucleotide coenzyme, respectively. Samples were allowed to incubate for 5 min prior to fluorescence measurements in order to attain equilibrium.
The dissociation constants and number of TNP-GTP binding sites on the enzyme were determined as follows. The fluorescence of TNP-GTP was measured in the presence, F, and absence, Fo, of enzyme (1.0 p~) , and the ratio of fluorescence (F/F,J was used to calculate [TNP-GTP]b,.d from where Q is the enhancement of TNP-GTP fluorescence for bound ligand. The enhancement factor, Q, was measured by titrating a fixed amount of TNP-GTP (2.7 p~) with increasing amounts of the enzyme in a concentration range of 0.5-4 p~. A double-reciprocal plot of total protein concentration versus observed fluorescence was extrapolated to infinite protein concentration in order to determine the value of Q. Enhancement factors, Q, in the presence and absence of NADH were 2.7 and 5.6, respectively. The amount of free TNP-GTP is obtained from the difference of the total TNP-GTP and calculated bound TNP-GTP. The data were analyzed in terms of the Scatchard equation: where [L] is the free TNP-GTP concentration, B and ET are amounts of bound TNP-GTP and total enzyme, respectively, n is the number of binding sites, and KD is the dissociation constant for the enzyme.

3'-O-(N-Methylanthraniloyl)-Gpp(NH)p
was prepared by the procedure similar to that for 3'-O-(N-methylanthraniloyl)-GTP (4), except that Gpp(NH)p (0.16 mmol, lithium salt) was dissolved in 2 ml of water and allowed to react with N-methylisatoic anhydride (40 mg) by titrating with 1 N LiOH instead of NaOH. The reaction products were chromatographed on a column of Sephadex LH-20 (2.2 X 83 cm) eluted with water (4). Fractions of 2.5 ml were collected. 3'-0-(N-Methylanthraniloy1)-Gpp(NH)p was eluted after the peak of unreacted Gpp(NH)p. Peak fractions of the analog were pooled and evaporated to dryness at 25 "C after neutralization (pH 6.8) with 1 N HC1. The residue was dissolved in 1 ml of water. An excess of cold acetone (15 ml) was added, and the resultant precipitate was collected and washed with ether (20 ml). The material was dried on PnOs in uacuo overnight. The analog was chromatographically pure as indicated by a single fluorescent spot, and free from starting materials and degradation products. The RF values in Solvent I on cellulose TNP-analogs of GTP, GDP, and GMP were prepared as follows. The nucleotide (0.5 mmol, sodium salt) was dissolved in 3 ml of water. The pH was adjusted to 9.6 with 4 N LiOH. To this solution, with continuous stirring, a crystalline preparation of TNCB (1 mmol) was added. The pH was maintained at 9.6 by titration with LiOH for 6 h at 38 "C. After completion of the reaction, the pH of the reaction mixture was adjusted to 7.0 with concentrated HCI. The reaction mixture was left for 15 min in ice. The resultant suspension was centrifuged at 18,000 rpm for 10 min. An excess of cold ethanol (20 ml) was added to the supernatant fluid. After cooling in ice for 10 min, the precipitate was collected by centrifugation, dissolved in 4 ml of water, and the solution was neutralized with 1 N HC1. The solution was then placed on a Sephadex LH-20 column (2.5 X 68 cm, packed in water) (7,8). This column was developed with water at a flow rate of about 60 ml/h. Fractions of 5 ml were collected. A separation profile was obtained after assays by TLC on cellulose; portions of each fraction (0.2-0.5 pl) were spotted on the plate. The plate was developed in Solvent I (for RF values of analogs, see Table I): As shown in Fig. 2, the TNP analog eluted after the unreacted nucleotide. A byproduct, picric acid, eluted after the analog (not shown in Fig.  2). Peak fractions of TNP analogs were pooled, evaporated to dryness at 30 "C, and the residue was dissolved in a minimum amount of water. An excess of cold ethanol (20 ml) was added, and the resultant precipitate was collected and washed with ether (20 ml X 2). The material was dried over PZO, in uacuo overnight. Yields based on the nucleotide were 28-45% (Table I).
TNP-Gpp(NH)p was prepared as follows. Gpp(NH)p (0.16 mmol, lithium salt) was dissolved in 1.5 ml of water and allowed to react with TNCB (0.32 mmol) as described above. After neutralization with HCl and cooling in ice, the resultant suspension was centrifuged, and the supernatant solution was directly placed on a Sephadex LH-20 column (2.5 X 44 cm, packed in water). The column was developed with water as described above. Fractions of 3 ml were collected. Peak

FIG. 2.
Graphic presentation of chromatographic purification of TNP-guanine nucleotides. Reaction products were applied on a 2.5 X 68-cm column of Sephadex LH-20 packed in water. The flow rate was 60 ml/h; 5-ml fractions were collected. The material shown at the left was eluted in the fractions indicated.
fractions of TNP-Gpp(NH)p, which eluted after the unreacted Gpp(NH)p, were pooled, evaporated to dryness at 25 "C, and the residue was dried over PZOS in vacuo overnight.
Purities of TNP-analogs were checked by TLC on silica gel and cellulose plates (Table I). All analogs were chromatographically pure as indicated by a single orange spot, and free from starting materials and a byproduct, picric acid. Spots of TNP analogs fluoresced under ultraviolet light. Purity was confirmed by elemental analysis (Table   I). For routine purposes, all analogs were stored in solution (pH 6. .0) at -20 "C.

RESULTS
Characterization of TNP-Guanine Nucleotides-Like TNP-ATP (7,8), the analogs are quite stable at neutral pH and can be stored for several months at -20 "C without significant degradation.' Proposed structures of the analogs are shown in Fig. 1. The assignment of these structures is based on the following observations: 1) On treating with 1 N HC1 for 4 h at 100 "C, all analogs were hydrolyzed to give picric acid and guanine. This is characteristic of 0-TNP-guanosine derivatives; by this acid Less than 3% decomposition in 2 months was observed with preparations of TNP-Gpp(NH)p and 3'-0-(A"methylanthraniloyl)-Gpp(NH)p as judged from TLC using two solvent systems given in Table I 2) By the action of 1 N KOH for 18 h at 37 "C, none of the analogs was degraded, suggesting 0-TNP-guanosine derivatives; N-TNP-guanine derivatives are almost completely degraded on this treatment (25).
The molar absorption coefficients (e) for TNP-GTP (26,500 M" .cm-' at 408 nm, and 18,300 M" .cm" at 470 nm) are essentially identical with those of TNP-ATP (7, 8), which implies that the base moiety of TNP-GTP has little effect on the chromophore (Fig. 3). The spectrum also exhibits an absorption maximum at 252 nm ( E = 24,100 M".cm") and a distinct shoulder at 280 nm, which is characteristic of guanine derivatives. There was no significant difference in absorption spectra of TNP analogs of GMP, GDP, GTP, and Gpp(NH)p (not shown).
Fluorescent Properties of TNP-GTP-Upon excitation with light in the 410-or 470-nm regions, TNP-GTP fluoresced similarly to TNP-ATP (8,9). The uncorrected fluorescence emission of TNP-GTP in aqueous buffer forms a single band with a maximum at 552 nm (Fig. 4). The potential usefulness of the analog as a fluorescent probe of hydrophobic microen- vironments is indicated by the fact that the position of emission maximum and the fluorescence intensity of the analog vary significantly with solvent polarity. As shown in Fig. 4, the intensity of TNP-GTP increases 5-and 24-fold in going from water to 40% and to 80% N,N-dimethylformamide, respectively. At the same time, the emission maximum is shifted to blue by 8 and 20 nm, respectively. Quantum yields of TNP-GTP were identical with those of TNP-ATP within the experimental error (&5%) in 0-80% N,N-dimethylformamide. Although data are not shown, all TNP-guanine nucleotides, including TNP-Gpp(NH)p, had similar fluorescent properties.
Kosower (29) has introduced an empirical polarity scale, the 2 value, based on transition energies of a pyridiniumiodine complex in various solvents. We have also observed a good correlation between the 2 value and fluorescent properties of TNP-guanine nucleotides. Results obtained with TNP-GTP are shown in Fig. 5 . As expected, large changes in the fluorescence intensity and the emission maximum were observed with changes in the ethanol composition of ethanol/ water mixtures. A good correlation was found between the emission maximum, as well as the fluorescence intensity, and a polarity scale, the 2 value. These results allow ready use of these TNP-guanine nucleotides, as well as TNP-adenine nucleotides (8,9), as fluorescent probes for hydrophobic regions of proteins.
Effect of TNP-GTP on Catalytic Activity of Glutamate Dehydrogenase-GTP is the natural inhibitor of glutamate dehydrogenase, which is the strongest and the most specific for this enzyme (30, 31). Our first interest was to determine whether the structural features of TNP-GTP resemble GTP sufficiently to allow acceptance by glutamate dehydrogenase as an inhibitor. As shown in Fig. 6A, TNP-GTP inhibits the enzymatic activity to a maximum extent of 54% at saturating When the terminal P-0-P moiety of TNP-GTP was replaced by a P-NH-P group, there was decreased binding of the analog to the enzyme, as evidenced by a 3.1-fold increase in the KI value (8.4 p~) , while the maximum extent of inhibition scarcely changed (i.e. TNP-Gpp(NH)p).
To determine whether TNP-GTP and GTP compete kinetically for the enzyme, the enzymatic activity as a function of GTP concentration was measured in the presence and absence of a constant TNP-GTP concentration (Fig. 6B). In the absence of TNP-GTP, the enzyme was inhibited by GTP to a maximum extent of 92% at saturating concentrations of the nucleotide with a KI of 0.07 p~. In the presence of 8.3 p~ TNP-GTP, a KI for the enzyme-GTP complex was measured to be 0.25 p~, a value 3.6-fold higher than the actual KI of 0.07 p~ measured in the absence of TNP-GTP. Saturating levels of GTP caused a 92% inhibition of the enzymatic activity, which is the same as that observed in the absence of TNP-GTP. These results suggest that TNP-GTP competes kinetically with GTP for the enzyme, and that TNP-GTP bound to the enzyme can be totally displaced by excess GTP.
Other ribose-modified analogs of GTP and Gpp(NH)p were also used to test the role of hydroxyl groups of the ribose moiety and the triphosphate moiety in GTP for inhibition of the enzymatic activity of glutamate dehydrogenase. These results are summarized in Table 11. For comparative purposes, the results obtained with GTP, TNP-GTP, and TNP-Gpp(NH)p are also listed. 3'-O-Anthraniloyl-GTP, 3'-0-(N-methylanthraniloy1)-GTP, and 3'-O-(N-methylanthra-niloy1)-Gpp(NH)p, in which only the 3'-hydroxyl group of the ribose moiety is modified (4), showed a greater extent of maximum inhibition than the corresponding T N P analog, in which both the 2'-and 3"hydroxyl groups are modified (Fig.  1). Taking into account that 3'-O-anthraniloyl-GTP showed an inhibitory effect of 90% of that of GTP at saturating levels of the analog, the 2'-hydroxyl group, rather than the 3'hydroxyl one in GTP seems to be required for proper positioning of the GTP molecule to inhibit the enzyme completely. On the other hand, replacement of the P-0-P moiety of the G T P analog with the P-NH-P group resulted in little or no change in the maximum extent of inhibition, but there was a 2.5-%fold increase in the KI value (i.e. TNP-Gpp(NH)p and

3'-O-(N-methylanthraniloyl)-Gpp(NH)p).
Titration of TNP-GTP Binding Sites on Glutamate Dehydrogenase-The binding of TNP-GTP to glutamate dehydrogenase was reflected in the emission spectrum of the analog. The addition of 2.4 p~ enzyme to a 2.7 p~ TNP-GTP solution resulted in a 4.3-fold enhancement in the fluorescence ( b in

TABLE I1
Inhibition of glutamate dehydrogenase by G T P and GTP analogs The enzymatic activity was assayed at 25 "C in 10 mM Tris acetate (pH 8.0), 10 p~ EDTA, 5 mM a-ketoglutarate, 50 mM ammonium chloride, and 100 p~ NADH. . ' " K , represents the concentration of GTP or the analog required for half-maximum inhibition. The maximum extent of inhibition at saturating concentrations of GTP was taken as 100%. Fig. 7). Under these conditions of measurement, a blue shift in the emission maximum from 552 to 545 nm was observed. The enhancement factor of the bound analog extrapolated to infinite protein concentration was found to be 5.6. This fluorescence enhancement was reversed or prevented by addition of 5 mM GTP or 250 mM ammonium chloride, which competes kinetically with GTP (32). This observation suggests that TNP-GTP binds to the GTP inhibitory site of the enzyme.
In the presence of NADH, the fluorescence of a solution of TNP-GTP and enzyme was less enhanced. When 100 p~ NADH was added to a 2.7 p~ TNP-GTP solution containing 2.4 p~ enzyme, the enhancement decreased from 4.3-to 2.3fold (d in Fig. 7). At the same time, only a small red shift in the emission maximum from 545 to 548 nm was observed. The limiting enhancement of the bound analog in the presence of NADH was measured to be 2.7. On the other hand, the addition of NADH to a free TNP-GTP solution resulted in no change in the fluorescence ( c in Fig. 7). These results suggest that, in the NADH. TNP-GTP. enzyme complex, the T N P moiety of the analog is either less buried or is less rigidly positioned than in the TNP-GTP .enzyme complex. Fig. 7 (inset) shows a typical TNP-GTP fluorescence titration in the presence and absence of enzyme. In the presence of 1 p~ enzyme, a large fluorescence enhancement was observed as described above. This enhancement appears to saturate since the slope of the plot of fluorescence uersus TNP-GTP concentration in the presence of enzyme approaches that in the absence of enzyme. The concentration of bound TNP-GTP is calculated from Equation 1 and used to determine B/ET in Equation 2. The dissociation constants for enzyme.TNP-GTP complexes and number of binding sites in the absence and presence of NADH were calculated by Scatchard analysis from Equation 2.
In the absence of NADH, glutamate dehydrogenase exhibits a linear Scatchard plot (Fig. 8A)

DISCUSSION
Okuyama and Satake (33) have reported that the TNP group of TNCB is introduced into hydroxyl groups of Tyr, Ser, Thr, and alcohols more rapidly than that of TNBS. This is also the case with the modification of guanosine derivatives. The use of TNCB instead of TNBS is very effective for the synthesis of 0-TNP-guanine nucleotides (Fig. 1) as revealed in the present paper. In contrast, the reaction of guanosine derivatives with TNBS yields the mixture of N-TNP, 0-TNP, and N,O-bis-TNP derivatives (25). These results suggest that TNCB is more reactive than TNBS toward hydroxyl groups of the ribose moiety of guanosine derivatives.
TNP-ATP, in which adenine replaces the guanine in TNP-GTP, has been previously synthesized (7,8). To be used as an "environmental probe" for proteins, the molecule must be highly sensitive to some indicator of local environment, e.g. polarity and viscosity. Previous studies (7)(8)(9) clearly indicated that the TNP group in TNP-adenine nucleotides is very useful not only as a chromophoric but also as a fluorescent probe for proteins. Because of these spectroscopic properties of the TNP group, TNP-ATP has been widely used in the study of various ATP-requiring systems (7,(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24). Especially, TNPadenine nucleotides are useful for the study of nucleotide binding because fluorescence of the analog is greatly enhanced upon binding to proteins: 11-fold for (Na+ + K+)-ATPase (12), 9-16-fold for aspartokinase (21), 7-fold for mitochondrial ATPase (16) and (Ca" + MP)-ATPase (13), and 4-5-fold for glutamate dehydrogenase (34,35). The present study revealed that spectroscopic properties of TNP-GTP are quite similar to those of TNP-ATP, and binding of TNP-GTP to glutamate dehydrogenase results in a 2.7-and 5.6-fold fluorescence enhancement in the presence and absence of NADH, respectively. TNP-GTP, therefore, should be a suitable spec-troscopic probe for the study not only of glutamate dehydrogenase but also of various GTP-requiring enzymes.
Bovine liver glutamate dehydrogenase is an allosteric enzyme composed of six identical subunits. The enzyme contains several nucleotide sites/subunit, including sites for the activator ADP and the inhibitor GTP (30, 31). Although the average distances between the ADP and GTP sites, and the ADP or GTP and catalytic sites have been recently measured by fluorescence energy transfer (34, 35), the number of GTP binding sites on the enzyme has been very controversial. Frieden and Colman (36) and Dieter et al. (37) have indicated the existence of one GTP binding site/subunit. In contrast, Bell et al. (38) have recently reported the existence of two GTP binding sites. Pal and Colman (39) have also demonstrated that the second GTP site is demasked by NADH binding. Our present study revealed the binding of 2 mol of TNP-GTP/mol of enzyme subunit. The relationship of TNP-GTP to the GTP inhibitory site on the enzyme has been established under "Results." The enzymatic activity was inhibited to a maximum extent of 54% by TNP-GTP, and through kinetic measurements, competition with the natural inhibitor, GTP, was demonstrated.
NADH enhances the binding of only 1 of the 2 mol of TNP-GTP bound to glutamate dehydrogenase. Pal and Colman (39) have also reported a similar result in the binding study of GTP in the presence of NADH although, unlike TNP-GTP, GTP is accessible to only one of the two sites in the absence of coenzyme. Therefore, there are high and low affinity sites for GTP or TNP-GTP on the enzyme in the presence of NADH. However, affinity of TNP-GTP for both sites appears to be higher than GTP as judged from the dissociation constants: 0.61 and 18 p~ for GTP (39), which were measured at the lowest enzyme concentration level used (0.2 mg/ml), and 0.11 and 0.71 p~ for TNP-GTP. This is also the case with TNP-adenine nucleotides. Several investigators have previously reported the higher affinity of TNP-adenine nucleotide over the corresponding natural nucleotide for various ATPases (7,8,12,13,16) and glutamate dehydrogenase (34). Thus, increased affinity of TNP-GTP may result from electrostatic interaction of the negative charge on the TNP moiety ( Fig. 1) with a cationic amino acid residue at or near the location of ribose in the GTP inhibitory site.
Our results also revealed that fluorescence of TNP-GTP is less enhanced by the enzyme in the presence of NADH than in the absence of the coenzyme. This suggests that the microenvironment of the TNP moiety becomes less hydrophobic or less rigid upon binding of the coenzyme to the TNP-GTP . enzyme complex. On the basis of the maximal extent of fluorescence enhancement and the data shown in Fig. 5, the hydrophobicity of microenvironment around the TNP moiety may be considered as identical to 40 and 24% ethanol for the TNP-GTP .enzyme and NADH. TNP-GTP .enzyme complexes, respectively.
In conclusion, TNP-guanine nucleotides, as well as TNPadenine nucleotides (7)(8)(9), are useful as chromophoric and fluorescent probes for hydrophobic regions in proteins. Especially, TNP-GTP is expected to have a wide applicability in probing the nucleotide sites of GTP-requiring enzymes. Alternatively, TNP-GTP may have a wide application in fluorescence energy transfer studies to investigate the structure of proteins. Although relatively low quantum yields of the analog (8)  .