New ribose-modified fluorescent analogs of adenine and guanine nucleotides available as subtrates for various enzymes

496

Bwchtmwa et Bwphystca A cta, 742 (1983) 496-508 Elsevier Blomedtcal Press

BBA 31486

NEW RIBOSE-MODIFIED F L U O R E S C E N T ANALOGS OF ADENINE AND GUANINE N U C L E O T I D E S AVAILABLE AS SUBSTRATES FOR VARIOUS ENZYMES TOSHIAKI HIRATSUKA

Department of Chemistry, Asahtkawa Medwal College, Asahtkawa, HokkaMo 078-11 (Japan) (Received September 6th, 1982)

Key words Fluorescence, Adenosme analog, Guanosme derlvattve, Substrate preparatton

The synthesis of fluorescent derivatives of nucleosides and nucleotides, by reaction with isatoic anhydride in aqueous solution at mild pH and temperature, yielding their 3'-O-anthraniloyl derivatives, is here described. The N-methylanthraniloyi derivatives were also synthesized by reaction with N-methylisatoic anhydride. Upon excitation at 330-350 nm these derivatives exhibited maximum fluorescence emission at 430-445 nm in aqueous solution with quantum yields of 0.12-0.24. Their fluorescence was sensitive to the polarity of the solvent; in N,N-dimethylformamide the quantum yields were 0.83-0.93. The major differences between the two fluorophores were the longer wavelength of the emission maximum of the N-methylanthraniloyl group and its greater quantum yield in water. All anthraniloyl derivatives, as well as the N-methylanthraniloyl ones, had virtually identical fluorescent properties, regardless of their base structures. The ATP derivatives showed considerable substrate activity as a replacement of ATP with adenylate kinase, guanylate kinase, glutamine synthetase, myosin ATPase and sodium-potassium transport ATPase. The ADP derivatives were good substrates for creatine kinase and glutamine syntbetase (v-glutamyl transfer activity). The G M P and adenosine derivatives were substrates for guanylate kinase and adenosine deaminase, respectively. All derivatives had only slightly altered K m values for these enzymes. While more fluorescent in water, the N-methylanthraniloyi derivatives were found to show relatively low substrate activities against some of these enzymes. The results indicate that these ribose-modified nucleosides and nucleotides can be versatile fluorescent substrate analogs for various enzymes.

Introduction Nucleotldes play an important role in many metabohc processes. A variety of nucleotxde analogs have been synthesized wluch have proven extremely useful m obtaining informauon about such processes [1] Among various nucleotlde analogs, fluorescent analogs are particularly excellent for obtaining reformation of this sort because of the sens~t~wty to small changes m enwronment as well as the low concentrations sufficient to obtain Abbrevtauon TNP-ATP, 2',3'-O-(2,4,6-trlmtrocyclohexadlenylldene) adenosine 5'-trlphosphate 0167-4838/83/0000-0000/$03 00 © 1983 Elsevier Btomedlcal Press

meamngful results. Several fluorescent nucleot~de analogs have been synthesized. Desirable attributes of a fluorescent analog would be tugh quantum yield, and absorption and ermssion maxima distract from those of proteins and nucleic acids In addition, it is preferable that a fluorescent analog is prepared by an easy one-step synthes~s. 1,N6-Ethenoadenoslne-5'-triphosphate [2] and adenosme-5'-triphosphoro--/- 1-(5-sulfomc acid) naphthylamadate [3] are parucularly useful as fluorescent analogs of ATP m this regard. The former has an altered punne ring structure, wlule the latter has an altered phosphoryl structure However, some nucleot~de-requmng enzymes are sense-

497 B

-O 00I~OH 2 O- O- O- ~ I H 0%0 Rsl~

I

Ant-ATP R=H,B=Ade Ant-GTPR=H,B=Guo I OH Mant-ATPR=CH3, B=Ade Mant-GTPR=CH3. B=Gua

Ftg ] Proposed structures of anthrandoyl (Ant) and methy|antramloy] (Mant) derivatives of ATP and GTP

tlve to alteration m the base or phosphoryl moiety of the nucleotide. For such enzymes the use of nbose-modified nucleotlde analogs is very effective. We have previously synthesized the ribose-modlfied chromophoric and fluorescent analog of ATP, TNP-ATP [4,5]. Many ATP-requlrlng enzymes tolerate chemical modlficatmn at the ribose nng of the ATP molecule [1]. Tlus is also the case for TNP-ATP; TNP-ATP is a good substrate for myosin ATPase [4,6] and adenylate kinase (unpubhshed data), and a potent competmve inhibitor for pyndoxal kinase [7], (Na ÷ + K ÷ )-ATPase [8] and soluble mitochondrial ATPase [9-11]. These resuits encouraged us to synthesize new nbose-modified fluorescent analogs of ATP and other nucleoudes. It has been reported [12,13] that anthramloyl and methylanthraniloyl groups (Fig. 1) are excellent fluorophores. In addmon, they are smaller than most fluorophores, producing only minor perturbations m the enzyme reaction On the other hand, Stinger and Miller [14] have reported the reactmns of lsato~c anhydnde and methyhsatoic anhydride with the hydroxyl group of alcohols, yielding the anthrandoyl and methylanthraniloyl derivatives, respectwely. This suggested to us that it should be possible to prepare the derivatives of nucleosides and nucleoudes containing the group at the nbose ring without alterations in their base moieties. Here we report the synthesis of such fluorescent denvauves, particularly the ATP derivatives (Fig. 1), and their spectroscopic and enzymaUc properties.

Experimental procedures Matertals ATP and ADP were purchased from Kyowa

Hakko Co. GTP and GDP were from Yamasa Shoyu Co. Adenine, guamne, adenosine, deoxyadenosme, guanosme, urldme, moslne, thyrmdme, AMP and GMP were from Kohjin Co. cAMP, cGMP, dAMP, 2'-O-methyladenoslne, Y-O-methyladenoslne and D-nbose 5-phosphate were from Sigma Chemical Co. Isatolc anhydnde and methyllsatoic anhydride were from Molecular Probes Co. Sihca gel (Sdlca gel 60) and cellulose (Avlcel SF) TLC plates were from Merck and Funakoslu Chemical Co., respectwely. Adenylate lonase (rabbit muscle), creatme kanase (rabbit muscle) and pyruvate klnase (rabbit muscle) were from Boehringer Mannhelm Co. Lactate dehydrogenase (pig heart) was from Oriental Yeast Co. Adenosine dearmnase (calf intestinal mucosa), guanylate kinase (bovine brain), glutanune synthetase (sheep brain), (Na ÷ + K ÷ )-ATPase (hog cerebral cortex) and alkahne phosphatase (E cob) were from Sigma Chermcal Co. Heavy meromyosm ATPase was prepared from rabbit skeletal myosin by dlgestmn with trypsin for 10 mln at 23°C as described previously [15]. Other reagents were of reagent or biochemical research grade. Methods

Analyttcalmethods TLC was performed on SdlCa gel in (A) 1-propanol/NH4OH/water (6 : 3 : 1, v/v, containing 0.5 g/1 EDTA). Fluorescent compounds were detected on chromatograms under a Camag Deluxe ultravaolet lamp (366 nm). TLC was also performed on cellulose In (B) 2-propanol/water/HC1 (65 : 18.4 : 16.6, v/v) and in system A to check contarmnants of the unreacted nucleoslde and nucleotide. For analysis of 2'- and Y-O-methyladenosmes, the cellulose plate was developed m (C) saturated ammonium sulfate/0 1 M CH3COONa (pH 6)/2-propanol (79: 19:2, v/v). The bluish color of fluorescent derivatives on cellulose chromatograms under ultraxaolet light (254 nm) allowed them to be dxstmgulshed from the dark-colored nucleoslde and nucleotlde. The position of the fluorophore of analogs was deterrmned by the monomethylauon study of adenosine according to the procedure of Robins et al. [16]. To a suspension of the adenosine denvatwe (5-7 nmol) m 1,2-dlethoxyethane was added a methanohc solution (0.9 ml) of 1 mM SnCI 2 • 2H20 and 0.5 ml of ethereal solution of dlazomethane

498

(0.43 M) The suspension was stirred overnight at room temperature. The system was dried by evaporation, and the resxdue was &ssolved in 1 ml of 30% ethanol followed by deacylation in 0.1 N N a O H (for 1 h at 25°C) Under these conditions, there was stoichlometrlc release of the fluorophore, while the ether linkage of the O-methyl substltuents was stable. Ahquots of the solution were chromatographed on cellulose than layer in system C The plate was VlSuahzed under ultrawolet hght (254 nm). Each spot was scraped from the plate, Rf values were 0.18 and 0.12 for 2'-Omethyladenosme and 3'-O-methyladenoslne, respectwely The material was eluted from the cellulose with water and the amount was quantltated optically Anthrandoyl-Ado and methylanthranlloyl-Ado, which were products of the action of alkahne phosphatase on the corresponding ATP derivative, were also monomethylated and analyzed in a similar manner. For determination of acld-labde phosphate, the fluorescent derivatives of ATP and G T P were hydrolyzed in l N HC1 at 100°C for 7 nun [17]. The acid-labile phosphate hberated was deterrmned by the method of Flske and SubbaRow [18] Spectral measurements Absorption spectra were measured at room temperature with a Shamadzu double beam spectrophotometer, Model UV-200 Molar absorption coefficients of analogs were deterrmned In 50 m M Trls-HC1 (pH 8 0) Fluorescence ermsslon and excitation spectra were recorded at 25°C in a Hitachi fluorescence spectrophotometer, Model MPF-4, eqmpped with a corrected spectra accessory The absolute quantum yields of analogs were measured by the method of Parker and Rees [19], using quinine sulfate in 0.1 N H2SO 4 as a standard of quantum ywld 0.70 [20] The absorbance m the cell was not allowed to exceed 0.009 at an excitation wavelength to obviate inner filter effect ExcitaUon wavelengths were 330 and 350 nm for anthrandoyl and methylanthrandoyl derivatives, respecuvely. The sht widths on the exotation and emission monochromators were 5 or 10 nm Sodmm acetate (below p H 6.5) and Trls (above p H 6.5) buffers were used at a concentration of 0 2 M for spectrophotometrlc and fluorometnc p H tltratlons of analogs. HC1 and N a O H were used for extremes of p H

Enzyme assays Enzyme assays were performed at 25°C unless otherwise noted Values of Vmax and K m were estimated from a direct linear plot of the enzymauc actw~ty vs. the substrate concentration [21] Prior to kinetic stu&es, activity of normal substrates and analogs were checked by TLC analysis The enzyme concentrations and the incubation Umes were deterrmned from pdot assays to give klneUcally valid data. With adenosine deamlnase the assay procedure was based on the rate of &sappearance of the absorption band of adenosine [22] and anthrandoyl-Ado at 265 nm When anthranlloyl-Ado was converted to anthraniloyl-Ino the absorbancy change was - 7.2 • 103 M - 1. c m - t at p H 7 5 The assay rmxtures (1 ml) contained 40 mM Tns-HC1 ( p H 7.5), 1 6 /xg of the enzyme and 0.002-0.05 m M adenosine or anthrandoyl-Ado. Adenylate kmase and guanylate klnase actlvlt~es were measured with 1 ml of reaction m~xtures containing 50 m M Trls-HC1 (pH 7.5), 0 1 M KC1, 5 m M MgC12 and 0 2 /~g of the enzyme, 1 mM A M P (for adenylate kmase) or 0.33 mM G M P (for guanylate kmase), and 0 01-0.5 m M ATP or the analog. For experiments with G M P analogs as substrates for guanylate klnase, 3.3 m M ATP and 0 0 0 3 - 0 0 3 mM G M P or the analog were used The reaction was stopped by addition of 5% trichloroacetlc acxd. A 20-/xl portion of the reaction mixture was spotted on a TLC plate, on sd~ca gel for systems of the ATP analog + A M P (GMP) and of the G M P analog + ATP, and on cellulose for the ATP + A M P (GMP) system. The plates were developed in system A and then visualized under ultraviolet hght Each spot of products (ADP and fluorescent denvatwates of A D P and G D P ) was scraped from the plate. The product was eluted from the sdica or cellulose with 1 ml Trls-HC1 (pH 8.0) The amount was quantltated spectrophotometrically or fluorometrlcally R f values on sihca gel (for analogs) and cellulose (for ADP) were anthranlloyl-ADP, 0 38, methylanthranlloyl-ADP, 0.44; anthrandoyl-GDP, 0 20, methylanthranlloylG D P , 0.24, ADP, 0 13 Pyruvate kanase activity was measured with 1 ml reaction mtxture containing 50 m M Tris-HCl ( p H 7 5), 0 1 M KC1, 5 m M MgC12, 1 mM phosphoenolpyruvate, 0 1 m M N A D H , 10 umts of lactate dehydrogenase and 0.03 umt of pyruvate

499 kinase, following the decrease m absorbance at 340 nm. A D P or anthrandoyl-ADP was used at concentrations varying from 0.04 to 0.5 mM With creatlne kmase the reverse reaction was investigated because of the sensitivity and simplicity of the assays for the release of creatlne from phosphocreatme. Reaction nuxtures contained, in a total volume of 1 ml, 50 mM N-ethylmorphohne (pH 8.0), 5 mM MgC12, 10 mM phosphocreatlne, 0.6 #g of the enzyme and 0 025-0 7 mM A D P or the analog. The reacUon was stopped by addition of 1 4 N N a O H and the amount of creatine was deterrmned by the method of Morrison et al. [23]. Glutamme synthetase actwity was d~termlned in reauon mixtures (1 ml) containing 50 mM lmldazole-HC1 (pH 7 2), 20 mM MgC12, 25 mM 2-mercaptoethanol, 50 mM sodium L-glutamate, 0.1 M hydroxylamlne, 1 5 /~g of the enzyme and 0.5-5 mM ATP or the analog. After incubation at 37°C for 30 nun, 1 5 ml of a solution containing 0.37 M FeC13/0 67 N HC1/0.2 M tnchloroacetic acid was added. The amount of hydroxamate formed was deternuned by the method of Wellner and MeIster [24]. Alternatively, the synthetase activity was assayed by measuring the production of P, according to the method of Fiske and SubbaRow [18]. Results from the two methods were comparable y-Glutamyl transfer activity was determined in reaction mixtures (1 ml) containing 50 mM ]rmdazole-HCl (pH 7 2), 5 mM MnC12, 50 mM L-glutamlne, 20 mM potassium phosphate (pH 7 2), 100 mM hydroxylamlne, 1.5 /~g of the enzyme and 0.5 mM A D P or the analog. After incubation at 37°C for 30 nun, the reaction mixture was treated with the FeC13 reagent, and the amount of hydroxamate formed was deternuned as described above. The heavy meromyosln ATPase acUvlty was measured by assays of P, as described previously [4]. The assay rmxtures (1 ml) contained 50 mM Tns-HC1 (pH 7 5), 0.5 M KC1, 5 mM CaC12, 0.003 mg of the enzyme and 0.001-0.1 mM ATP or the analog. (Na ÷ + K ÷ )-ATPase activity was measured by assays of P, according to the method of Fiske and SubbaRow [18]. The assay nuxtures (1 ml) contained 40 mM Tris-HC1 (pH 7.5), 0.1 mM EDTA, 10 mM KCI, 0.1 M NaCl, 3 mM MgCl2, 0.2 mg of the enzyme and 0.04-2 mM ATP or the analog

The reaction was ternunated by addmon of 5% tnchloroacetic acid. Synthests Anthranlloyl and methylanthranIloyl derivatives of adenosine, guanosme, lnosme and undlne were prepared as follows. The nucleoslde (1 mmol) was dissolved in a mlmmum amount of water (15-35 ml) at 38°C. The pH was adjusted to 9 6 with NaOH. To this solution with continuous stirring a crystalline preparaUon of isatolc anhydride or methyhsatolc anhydride (1 5 mmol) was added The pH was maintained at 9.6 by titration with 2 N N a O H for 2 h. The reacUon mixture was cooled, and the precipitate was collected by centnfugatlon. The denvative was recrystalhzed from water/ethanol/ether mixtures, giving a crystalline product as the analytical sample.

(A) AMP GMP

Ant-AMP Mont-AMP Ant-GMP

137~51 /,a6-s3 13S-52 laa-s4

Mant-GMP

i] 1

I

2O

=

30 )

,

4b

Fr No

50

(Srnl/tube)

6'o

(a) ADP, ATP GDP, GTP

D29,30 Q 26,27

Ant-ADP

Mont-ADP A -ATP

13 " Z

Mont-ATP Ant-GTP Mont-GTP

[39-51 I

I

13s-~s I

2'0

'

3'0

' ' s'o ' F'r No (5ml/tube)

6'o

F~g 2 Grapluc presentatton of chromatographicpurification of anthramloyl (Ant) and methylanthrandoyl(Mant) nucleoudes (A) Reaction products of nucleoslde monophosphates were apphed on a 2 4 × 56 cm column of Sephadex LH-20packed m 30% ethanol (B) Reaction products of nucleoslde dl- and tnphosphates were apphed on a 2 2 × 82 cm column of Sephadex LH-20 packed m water The flow rate was about 40 ml/h, 5-ml fractions were collected The materml shown at the left was eluted m the fracuons indicated

500

The derivatives of AMP, GMP, ADP, ATP and G T P were prepared by reacting the nucleotlde with lsatolc anhydride or methyhsatolc anhydride under the condiuons described above, except that the nucleotlde (1 mmol) was dissolved m 15 ml water. After completion of the reaction, the pH of the reaction mixture was adjusted to 7 0 with 1 N HC1. The reaction products were &rectly placed on a Sephadex LH-20 column (2 4 × 56 cm, packed m 30% ethanol for the monophosphate denvatwes or 2 2 × 82 cm, packed in water for the dl- and tnphosphate derivatives). The column was eluted with the same solvent at a flow rate of about 40 ml/h. Fractions of 5 ml were collected The fluorescence profile of the fractions was obtamed after assays by TLC on slhca gel, ahquots of each fraction (0.2-0 5 /~l) were spotted on the plate. The plate was developed m system A (for R f values of analogs, see Table I). As shown in Fig. 2, the fluorescent analog eluted after the peak of unreacted nucleotlde. Fluorescent by-products, anthranihc actd and N-methylanthramhc acid, eluted after the peak of the fluorescent analog (data not shown). (The elution of the fluorescent analog can be conveniently monitored under an ultraviolet lamp (366 nm) in the dark-room. The analog has brllhant blue fluorescence, while

anthrandlc acid and N-methylanthranlhc acid show violet fluorescence.) Peak fractions of fluorescent analogs were pooled. In the case of monophosphate derivatives, the preparation required further purification. The pooled fractions were evaporated to dryness in vacuo at 30°C. The residue was &ssolved in a rmnimum amount of water. An excess of cold acetone was added and the precipitate was collected followed by an acetone wash. The material was dried m vacuo at room temperature for analysis. Preparations of dl- and tnphosphate derivatives were used for experiments without further purification after neutrahzauon with HC1. For elementary analysis, ahquots of the preparation of the trlphosphate denvatwe were evaporated to dryness at 30°C. The residue was &ssolved m a minimum amount of water. Repreopltatlon from aqueous ethanol followed by an ethanol wash yielded a pure product. The material was dried over P205, gdvmg an amorphous powder as the analytical sample. Purities of isolated fluorescent analogs were checked by TLC on silica gel and cellulose plates All analogs were chromatographically pure as indicated by a single fluorescent spot, and free from starting matermls and fluorescent by-products.

TABLE I ISOLATED YIELDS A N D R f VALUES OF A N T H R A N I L O Y L A N D M E T H Y L A N T H R A N I L O Y L DERIVATIVES Yield ~s measured on the basis of total nucleoslde and nucleottde m the reaction System A SdlCa gel plate developed m i _ p r o p a n o l / N H 4 O H / w a t e r (6 3 1, v / v , containing 0 5 g / l of EDTA) System B cellulose plate developed m 2.propanol/water/HCl (65 184 16 6, v / v ) Methylanthramloyl derivatives

Anthramloyl derivatives Compound

Ymld

R f m System

Compound

(~)

Anthrandoyl-Ado -Ino -Urd -AMP -GMP -ADP -ATP -GTP Anthramhc acid

40 38 44 57 60 51 51 37 --

Ywld

R f m System

(~) A

B

0 87 0 68 0 69 0 52 0 40 0 38 0 18 0 09

0 27 0 19 0 50 0 30 0 26 0 27 0 32 0 22

Methylanthranfloyl-Ado -Guo -AMP -GMP -ADP -ATP -GTP

0 68

0 69

N-Methylanthramhc a o d

25 37 57 56 48 52 40 --

A

B

0 90 0 68 0 55 0 44 0 44 0 21 0 12

031 0 27 0 36 0 30 0 33 040 0 29

0 72

0 80

501 TABLE II RESULTS OF ELEMENTARY ANALYSIS FOR SOME ANTHRANILOYL AND METHYLANTHRANILOYL DERIVATIVES Compound

Formula ( M r m parenthesis)

Analysis Calcd (%)

Found(%)

C

H

C

H

Anthraniloyl-Ado -Ino -ATP -GTP

CITHIsN6Os 2H20 (422 41) C17HI7N506 3 2H20 (445 01) CITHj7N6014P3Na4 2 3H20 (755 70) ClTHlTN6OisP3Na,t 2 IH20 (768 10)

48 34 45 88 27 02 26 58

5 25 5 30 2 88 2 78

48 85 45 35 27 32 26 93

5 25 5 40 3 02 2 69

Methylanthramloyl-Ado -Guo -ATP -GTP

C is H 20N60s (400 41) CIsHzoN606 0 4H20 (423 61) CtsHI9N6OI4P3Na4 2H20 (764 32) ClaHIgN6OIsP3Na4 2 2H20 (783 92)

54,00 51 04 28 29 27 58

5 03 4 95 3 03 3 01

54 21 51 15 28 50 27 92

5 05 5 06 3 12 3 18

Isolated yields and R f values of analogs are summarlzed m Table I. Purity was confirmed by elementary analysis for some denvatwes (Table II). For routine purpose, all analogs were stored m solution (pH 7.0) at - 2 0 ° C Results and Discussion Charactenzatton The stabihty of anthranlloyl-ATP and methylanthranlloyl-ATP was tested With anthrandoylATP small amounts of the analog were dissolved in (a) 0.1 N NaOH; (b) 0.5 N NaOH; (c) 0.5 N HC1; (d) 0.1 M Tns-HC1 (pH 8.0) IncubaUon was allowed to continue at 25°C, and at 0, 0.5 and 12 h a small ahquot of the reactmn mixture was spotted on cellulose thin layer. TLC was performed in system A and analyzed under an ultraviolet lamp. All four samples immediately upon mixing showed a single clear fluorescent spot with a n R f of 0.44 corresponding to anthraniloyl-ATP and indicating that no hydrolysis had taken place Half an hour later, sample (a) gave one ultraviolet-absorbing spot wtth an R e of 0.10 corresponding to ATP and two fluorescent spots at R e 0.44 and 0.82, which correlated with anthrandoyl-ATP and anthrandic acid Sample (b) was completely hydrolyzed, only ATP and anthramhc acid being detected. Samples (c) and (d) retained resistance to degradation following overnight incubation. With methylanthranlloyl-

ATP similar results were obtamed. At - 2 0 ° C , both analogs were quite stable at neutral pH m solution and could be stored for at least 8 months without slgmflcant degradation. With all analogs, the penodate-benzadme test for glycols [25] gave negative results, indicating that either the 2'- or 3'-hydroxyl group was bound to the fluorophore. Tins agrees with the observation that adenosine, guanoslne and D-ribose 5phosphate reacted with lsatolc anhydride and methylisatolc anhydride, while neither adenine nor guanine did under identical conditions. In addition, the incubation of analogs m 0 1 N N a O H for 1 h at 25°C resulted m the complete disappearance of original materials and the appearance of anthrandlc acid or N-methylanthranlhc acid together with the parent nucleoslde or nucleotlde. Tins is characteristic of the T-(or 3'-)O-acylated nucleoslde derivatives [26,27]. The position of the fluorophore of analogs was determined by the monomethylation study of adenosine [16]. Treatment of adenosine in methanohc solution containlng stannous chloride with a solution of dlazomethane gives a mixture of T-O-methyl (-.~ 40%) and Y-O-methyl ( ~ 60%) isomers [16]. However, the monomethylatlon of anthranlloyl-Ado and methylanthrandoyl-Ado under identical condltions followed by deacylatlon in 0.1 N N a O H gave 2'-O-methyladenoslne exclusively (above 90%). A control experiment with adenosine gave a mixture of T-O-methyl (48%) and 3'-O-methyl

502

(52%) isomers. These results indicate that the 3'hydroxyl group of the analog ts protected by the fluorophore. Similar results were obtained with anthrandoyl-Ado and methylanthranlloyl-Ado prepared from the corresponding ATP derivative Thus, the presence of excluswely the 3'-0- derivative (Fig, 1) rather than 2'-0- denvatwe was proposed. Isatolc anhydride and methyhsatoxc anhydnde were allowed to react with adenosine, guanoslne, mosme, urldme, cytldme, deoxyadenosme and thymldlne and their nucleotlde denvatwes including cAMP and cGMP. Details of the preparation and properues of anthranlloyl and methylanthrandoyl derivatives of cAMP and cGMP are presented elsewhere [39]. Estenficatmns of the nbonucleoslde and nbonucleotlde result in the formatmn of either 2' or 3' isomer. In the case of anuno acid esters of nucleotldes, such ~somers can be resolved chromatographically. In the early stages of the reaction of nucleot~des with lsatolc anhydride or methyhsatolc anhydnde, one minor fluorescent component with an R f value higher (by a factor of about 1.2) than that of the major fluorescent component (Table I) was detected by TLC on slhca gel in system A. The manor component was converted to the major component as the reaction proceeded. By a combination of ultraviolet spectra and the penodate-benzadlne test for glycols [25], it was concluded that the component of tugher R f was the 2' ]somer Tlus conclusmn is supported by the observation that such a minor fluorescent component was not detected m reactmns of dAMP, cAMP and cGMP. The poss~bdlty of conversxon of the 2' to the 3' ~somer by acyl rmgratlon ~s well supported by a number of studies [28,29] Evidence in the hterature indicates that the T-hydroxyl groups of both adenosine and undxne are more open to electropluhc attack than the Y-hydroxyl substitution [30]. Tlus was also the case for the reaction with lsatolc anhydride or methyhsatotc anhydride. For example, the reaction of Y-O-methyladenosme with lsatolc anhydride proceeded better than that of 2'-O-methyladenosme. Thus the T-hydroxyl group ~s kmetlcaUy more reacUve for substitution. However, such substitution is relatively less stable in comparison to the thermally more favorable estenflcatmn at the Y-hydroxyl group [29]. The

present results indicate the 3' isomer to be the specific components Isolated under our synthetic con&tlons. Further support for ttus conclusion is seen in observations that estenficatmns of adenosine with fluorosulfonylbenzyl chloride [31] and 5-(&methylanuno)naphthalene- 1-sulfonyl chloride [32], and ATP with N-(4-azado-2-nltrophenyl)-flalanlne (27) and 4-benzoylbenzolc a o d [33] result m preferential formaUons of 3'-0- derivatives.

Absorption and fluorescent propertws Absorption spectra and molar absorption coefficients of anthramloyl-ATP and methylanthranlloyI-ATP at pH 8.0 are shown m Fig. 3A and T a b l e III, respectively The spectrum of anthramloyl-ATP exhibits two maxima at 252 nm ( e = 2 0 200 M - l - c m -1) and 332 nm ( e = 4 7 0 0 M - l. c m - t). A broad band centered at 332 nm is associated with the anthranlloyl group [12] The spectrum of methylanthramloyl-ATP is smular to

(A)

(B)

2

7 E

E u :E

¢2

v

v

'-,1" I0 X LU

uJ

I

t\

\ 250

300

350 nrrl

400

250

300

350

40O

nm

Fig 3 Absorpuon spectra of anthrandoyl and methylanthramloyl denvatwes of ATP and GTP (A), anthrandoylATP ( ) and methylanthramloyl-ATP ( - - - - - - ) (B), anthranlloyl-GTP ( ) and methylanthranlloyl-GTP ( - - - - - - ) All spectra were measured m 50 mM Tns-HCI (pH 8

0)

503 TABLE III ABSORPTION PROPERTIES OF A N T H R A N I L O Y L A N D M E T H Y L A N T H R A N I L O Y L DERIVATIVES IN 50 mM TRIS-HCI (pH 8 0) c values are based on three determlnat]ons The values for denvattves of ATP and GTP were estimated on the basts of actd-labfle phosphate [4,17] Anthranfloyl derivatives Compound

Methylanthranlloyl denvattves ~ × I O- 3 (M - 1 c m - l )

Compound

(nm)

~' max (nm)

c × 1O- 3 (M - 1 cm - i )

252 332

20 3 46

Methylanthramloyl-Guo

252 350

23 2 57

-Ino

247 331

20 8 46

-ATP

255 356

23 3 58

-ATP

252 332

20 2 47

-GTP

252 350

22 6 57

-GTP

249 332

20 7 46

k max

Anthramloyl-Ado

that of anthrandoyl-ATP, but there are some sigmficant differences. An absorption band of the methylanthraniloyl group is centered at approx 350 nm [13] compared with 332 nm for the anthraniloyl group. In addition, the molar absorption coefficients of methylanthranlloyl-ATP are higher than those of anthranlloyl-ATP; e = 23 300 M - I • c m - 1 at 255 nm and e = 5800 M - 1 c m - 1 at 356 nm. The spectrum of the GTP analog is similar to that of the corresponding ATP analog, except that it exhibits a distinct shoulder around 280 nm (Fig. 3B), wluch is characteristic of guanine derivatives. Molar absorption coefficients at two maxima are essentially identical with those of the corresponding ATP derivative (Table III). As summanzed in Table III, all anthramloyl derivatives including anthraniloyl-Ino, as well as methylanthranlloyl derivatives, have similar molar absorptmn coefficients at two maxtma, regardless of their base structures. These results are also consistent with an anthrandoyl derivative containing one anthrandoyl group, since the molar absorpuon coefficient of a model compound, methyl anthranllate, in a variety of solvents ranges from 3800-5400 M -1- cm -1 at 327-336 nm [12] Both anthraniloyl-ATP and methylanthrandoylATP fluoresce strongly in the range 410-445 nm when excited with hght m the 330-nm or 350-nm

regions. Fig. 4 shows corrected fluorescence emission spectra of anthraniloyl-ATP (A) and methylanthranlloyl-ATP (B) in ethanol/water mixtures at pH 8.0. Although the data are not shown, two excitauon maxima (corrected) were observed in water (pH 8.0) at 247 nm and 332 nm for

50 80

(A) 15

400

450nm500

550

400

450 nm500

550

Fig 4 Fluorescence emission spectra of anthrandoyl-ATP (A) and methylanthrandoyl-ATP (B) m water/ethanol mixtures All samples (1 #M) contamed 50 mM Tns-HCl (pH 8 0) The percentages of ehtanol ( v / v ) are m&cated on the curves Excited at 330 nm and 350 nm for anthramloyl-ATP and methylanthramloyl-ATP, respecttvely

504

anthrandoyl-ATP, and at 253 nm and 357 nm for methylanthranlloyl-ATP. To be used as an environmental probe for protems, the molecule must be lughly senslttve to some Indicator of local environment. The potential usefulness of these analogs as such fluorescent probes of hydrophoblc mlcroenvlronments is indicated by the fact that the position of emission maxima and quantum yields of analogs vary significantly with solvent polarity. As can be seen, quantum y~elds of anthrandoyl-ATP (Fig 4A) and methylanthranlloyl-ATP (Fig. 4B) increase approximately 5and 4-fold, respectively, in going from water to 80% ethanol. At the same time, emission maxima are shifted to blue by about 10 nm All anthrandoyl derlvatwes, as well as methylanthranlloyl derlvatwes, had slnular fluorescent properties (data not shown) Table IV contains the enusston maxima and quantum y~elds of the analogs of adenosme, ATP and GTP m selected solvents All analogs fluoresce in the range of 410-445 nm m water and organtc solvents. Their quantum ytelds are relatwely low (0.12-0.24) m water, but significantly high (0.65-0.93) m orgamc solvents These data also tmply that phosphate and base portions of anthranlloyl and methylanthrandoyl denvattves have ltttle effect on the fluorophore The major differences between the two fluorophores are the

longer wavelength of the emission maximum of the methylanthramloyl group and tts greater quantum yield in water. Haugland and Stryer [12] have reported the detaded fluoresence study on the anthrandoyl group, which ~s utdlzed to obtain information about the acttve site of a-chymotrypsin On the other hand, Trzos and Reed [13] have recently reported the synthesis and fluorescent properties of the fluorescent procaine analog, the methylanthrandoyl derivative of N,N-dlethylethanolarmne. Fluorescent properties of anthranyloyl and methylanthrantloyl dertvatwes of nucleos~des and nucleottdes were essentmlly the same as those reported by them As the pH was decreased, gradual decreases in absorption m the region of 300-400 nm and in fluorescence were observed for all analogs (data not shown) No shift in the posttlon of the absorption and emission maxama occurred, suggesting that only one structure, the deprotonated form of anthramloyl and methylanthranlloyl groups, is the origin of the absorption and the fluorescence Spectrophotometrlc pH t~tratlon of analogs yielded ground-state pK a values of 2.1-2 4 These values were m fair agreement with the excited state values obtained fluorometncally. It ~s ewdent from these results that absorption m the region of 300-400 nm and fluorescence of these analogs are mvarlant in the physiological pH regton

T A B L E IV F L U O R E S C E N T PROPERTIES OF A N T H R A N I L O Y L A N D M E T H Y L A N T H R A N I L O Y L DERIVATIVES All samples (1 # M ) contained 50 m M Tns-HC1 (pH 8 0) Excited at 330 and 350 n m for anthramloyl and methylanthramloyl derivatives, respectively W, water, E, ethanol, D, N,N-d~methylformarmde Q u a n t u m yield values are based on three determinations Anthranlloyl denvat~ves

Methylanthramloyl denvatwes

Compound

Solvent

~ max (nm)

Quantum yield

Compound

Solvent

~km a x (nm)

Quantum yield

Anthramloyi-Ado

W E D

429 415 408

0 12 0 71 0 88

Methylanthrandoyl-Ado

W E D

446 431 423

0 20 0 73 0 86

-ATP

W E D

428 417 410

0 14 0 65 0 86

-ATP

W E D

446 432 427

0 22 0 69 0 85

-GTP

W E D

428 417 410

0 14 0 66 0 83

-GTP

W E D

442 432 426

0 24 0 66 0 93

505

Enzymatw propertws Anthraniloyl and methylanthramloyl derivatives were examaned for their ablhty to act as substrates for various enzymes. Our first interest was to determine whether the structural features of anthranlloyl-Ado resembled adenosine sufficiently to allow acceptance by adenosine deamlnase. 1,N6-Ethenoadenosme derivatives have been successfully utlhzed in the study of a number of enzymes [2] Whde l'nghly fluorescent, I , N 6ethenoadenosme presents a modification of adenosine at the 6-amino group, winch is the reaction site of adenosine dearmnase, and thus may not be a suitable fluorescent analog for the study of tins enzyme. Recently, the nbose-modxfied fluorescent analog of adenosine, Y-O-[5-(&methylammo)naphthalene-l-sulfonyl] adenosine, was synthesized to use as a substrate for adenosine deammase [32]. However, tins analog was not a substrate for the enzyme. The presence of the

relatwely bulky fluorophore, which seems to be m close proxanuty to the adenine moiety [32], may alter the positioning of the 6-amino group m the catalytic center of the enzyme. In contrast to tins, not only did anthrandoyl-Ado funcUon as a substrate with adenosine deamlnase, w~th a g m of 10 /tM compared to 34 /~M for adenosine, but the Vmax was 16% of that of adenosine (Table V), making anthrandoyl-Ado a good substitute compared to other fluorescent analogs. Guanylate klnase shows that the nucleoslde monophosphate binding site is inghly speofic for the guanine moiety [34]. Among the naturally occurring monophosphate nucleot~des tested, only G M P and d G M P exinblted significant actwlty [34]. Thus a fluorescent G M P analog w~th an altered base structure should not be a statable substrate for the enzyme Fortunately, anthrandoyl and methylanthrandoyl denvatwes of G M P have unaltered guanine structures, showing considerable

TABLE V K I N E T I C DATA F O R A N T H R A N I L O Y L (Ant) A N D M E T H Y L A N T H R A N I L O Y L (Mant) DERIVATIVES K m for normal substrate is m parentheses Vmax is measured relative to normal substrate n d , not determined Enzyme

Substrate

K m (~M)

Vma~

Adenosine dearmnase Guanylate kmase

Ant-Ado Ant-GMP Mant-GMP

10 9 3

(34) (3) (3)

0 16 0 41 0 11

Ant-ATP Mant-ATP

50 40

(120) (120)

0 53 0 32

Ant-ATP Mant-ATP

120 90

(91) (91)

0 70 0 71

Pyruvate klnase

Ant-ADP

310

(290)

0 10

Creatme klnase

Ant-ADP Mant-ADP

150 50

(70) (70)

10 11

Adenylate kmase

Glutamme synthetase (glutawane synthetase acuvlty) ('t-glutamyl transfer acuvlty)

Ant-ATP Mant-ATP

2900 (2800) nd

0 56 0 81 a

Ant-ADP Mant-ADP

nd nd

0 68 b 0 71 b

Heavy meromyosm ATPase

Ant-ATP Mant-ATP

7 5

(13) (13)

0 73 0 68

(Na + + K + )-ATPase

Ant-ATP Mant-ATP

900 540

(480) (480)

0 20 0 11

a Substrate concentration, 5 mM b Substrate concentraUon, 0 5 mM

506 substrate activity against the enzyme K m values were 9 # M for anthraniloyl-GMP and 3 /~M for methylanthranlloyl-GMP and GMP. Vmax values were 41% and 11% of G M P for anthranlloyl-GMP and m e t h y l a n t h r a n i l o y l - G M P , respectively (Table V) This result implies a wide applicability of anthraniloyl and methylanthraniloyl analogs not only of G M P but of other guanine nucleotldes to investigate structures of enzymes, which are highly specific for the guanine moiety The enzyme also exhibits a high degree of specificity with regard to the phosphoryl donor; only ATP and dATP serve as effective donors among several nucleoside trlphosphates tested [34] Both anthranlloyl-ATP and methylanthraniloyl-ATP showed considerable substrate activity with K m values of 50/~M and 40 btM compared to 120 /tM for ATP, and Vm~x values 53% and 32% that of ATP, respectively (Table V) It should be emphasized that both methylanthraniloyl-GMP and methylanthraniloyl-ATP showed substrate activity lower than the corresponding anthranlloyl derivative With adenylate klnase, activity was found in the system anthraniloyl-ATP (methylanthranlloylA T P ) + A M P In contrast to guanylate klnase, there was no significant difference i n Vmax values between anthraniloyl-ATP and methylanthraniloyl-ATP, K m values of 120 #M and 90 #M, respectively, compared to 91 /~M for ATP, and a Vm~x about 70% that of ATP (Table V) However, the system anthraniloyl-AMP (methylanthraniloylAMP) + ATP was devoid of activity These results suggest that the ATP site is not too specific, readily accepting these ATP analogs, while the AMP site is rather more stringent Slrmlar results have been reported with 1,N6-ethenoadenosxne derivatives of ATP and AMP [2] Anthramloyl-ADP acts as a nucleotlde substrate for pyruvate lonase with a K m comparable to that of ADP (290/~M) and a Vm~x 10% that of ADP (Table V) This relative Vmax value is nearly equal to that of dADP (11%) [35]. However, methylanthraniloyl-ADP was a poor substrate for the enzyme, showing a Vm,x less than 2% that of ADP (data not shown) This result suggests that the presence of the relativity bulky methylanthranlloyl group at the ribose moiety significantly affects the ability of ADP to serve as a substrate On the other hand, the enzyme is not sensitive to modifl-

cation of the adenine ring, readily accepting IDP, G D P [35] and 1,N6-etheno-ADP [2] in place of ADP With creatlne klnase lllmted variation of the sugar of nucleotide substrate has a relatively small effect on either the Vmax o r the K m [36]. This was also the case with anthranlloyl-ADP and methylanthraniloyl-ADP, showing K m values of 150/~M and 50 #M, respectively, compared to 70/xM for ATP, and a Vmax value similar to that of ADP (Table V). These results allow ready use of these ADP analogs for the enzyme in place of ADP Glutarmne synthetase requires ATP and ADP in the biosynthetic reaction of glutarmne and in the ),-glutamyl transfer reaction, respectively. The enzyme shows that nucleotlde-bmdlng sites are highly specific for the adenine moiety [24]. Among the nucleotides tested only ATP, dATP, ADP and dADP exhibited activity in the reaction As shown in Table V, anthraniloyl-ATP was 56% as effective as the natural substrate ATP in the synthetic reaction with a K m comparable to that of ATP (2 8 mM) Although the K m was not determined, methylanthranlloyl-ATP also exhablted a relative activity of 81%. Slrmlarly, both anthranlloyl-ADP and methylanthranlloyl-ADP activated the 3,-glutamyl transfer reaction, giving about 70% of the effect observed with ADP (Table V) These results indicate that ribose-modlfied ADP and ATP are suitable fluorescent substrates for the study of the enzyme. The specificity of myosin ATPase for cleavage of ATP has been widely studied with ATP analogs [37] It has been shown that there is a wide tolerance to modification of the ribose ring of ATP for cleavage of the "t-phosphate. This was also the case with b o t h a n t h r a n i l o y l - A T P and methylanthraniloyl-ATP, showing K m values of 7 /~M and 5 #M, respectively, compared to 13 # M for ATP, and a VmaXabout 70% that of ATP (Table V) These values may be compared with those of another ribose-modifled ATP analog, TNP-ATP [4], a K m of 4 #M and a relative Vmax of 76%. Anthrandoyl-ATP was also hydrolyzed by (Na + + K +)-ATPase at a relatively slow rate compared to ATP, showing a K m of 900 #M compared to 480 #M for ATP, and a relative Vmax of 20% (Table V) In contrast to myosin ATPase, methylanthraniloyl-ATP was hydrolyzed by the enzyme

507 at a m a x i m u m rate of a b o u t one-half (11%) of that for a n t h r a n i l o y l - A T P , with a K m of 540 # M . These results suggest that the A T P a s e is more sensitive to m o d i f i c a t i o n of the rmbose moiety of A T P compared to m y o s i n ATPase. This conclusion is supported by the fact that the A T P analog with modified 2'- a n d 3'-hydroxyl groups, T N P - A T P , is n o t a substrate for (Na ÷ + K + ) - A T P a s e [8] b u t is a good substrate for m y o s i n A T P a s e [4]. A l k a l i n e phosphatase hydrolyzes a wide variety of phosphate esters [38]. All a n t h r a n l l o y l a n d m e t h y l a n t h r a n d o y l nucleotldes were also degraded b y this enzyme at a rate Sllmlar to that of the p a r e n t nucleotlde in 40 m M T n s - H C I (pH 7.5), p r o d u c i n g the c o r r e s p o n d i n g nucleosmde (data not shown)

Conclusion A n t h r a m l o y l a n d m e t h y l a n t h r a n i l o y l derivatives of nucleosides a n d nucleotides are useful as fluorescent substrates for various enzymes. The following chenucal, enzymatic a n d spectral propertles are factors that suggest a wide applicability of these analogs to investigate structures of nucleoslde- a n d n u c l e o t l d e - r e q u m n g enzymes: (a) they are easily synthesized, a n d stable for long periods of time; (b) the biological activity of nucleoslde a n d nucleotide is preserved to a considerable extent in the analogs with an only slightly altered K m value for various enzymes; (c) they a b s o r b a n d emit i n the region far from that of p r o t e i n , (d) their q u a n t u m yields are rather high in various solvents, a n d ermsslon is sensitive to solvent polarity, (e) their a b s o r p t i o n a n d fluorescent propertles are i n v a n a n t in the physiological p H region, (f) they can be extended to m a n y nucleosldes a n d nucleotides without alterations in their base structures to prepare various substrate analogs.

Acknowledgements I am i n d e b t e d to Professor K U c h i d a for his e n c o u r a g e m e n t a n d support. I a m grateful to Professor K Yagl, F a c u l t y of Science, H o k k a l d o university, for e l e m e n t a r y analysis, a n d K. N a k a m u r a for typing the m a n u s c r i p t

References 1 Yount, R G (1975) in Advances m Enzymology (Melster, A, ed ), Vol 43, pp 1-56, John Wdey and Sons, New York 2 Secnst, J A , Barrio, J R, Leonard, N J and Weber, G (1972) Biochemistry 11, 3499-3506 3 Yarbrough, L R, Schlageck, J G and Baughman, M (1979) J Blol Chem 254, 12069-12073 4 Hiratsuka, T and Uchlda, K (1973) Bloclum Blophys Acta 320, 635-647 5 Hlratsuka, T (1976) Blochlm Blophys Acta 453, 293-297 6 Hlratsuka, T, Uchlda, K and Yagl, K (1977) J Btochem (Tokyo) 82, 575-583 7 Churchach, JE and Wu, C (1981)J Blol Chem 256, 780-784 8 Moczydlowskl, EG and Fortes, P A G (1981) J Blol Chem 256, 2357-2366 9 Grubmeyer, C and Penefsky, H S (1981) J Blol Chem 256, 3718-3727 10 Kormer, Z, Kozlov, I A, Mdgrom, Y M and Novlkova, I Y (1982)Eur J Blochem 212, 451-455 11 Schafer, G (1982) FEBS Lett 139, 271-275 12 Haugland, R P and Stryer, L. (1967) in Conformation of Biopolymers (Ramachandran, G N, ed), Vol 1, pp 321-335, Academic Press, New York 13 Trzos, W and Reed, J K (1981) FEBS Lett 127, 196-200 14 Stalger, R P and Mdler, E B (1959)J Org Chem 24, 1214-1219 15 Mueller, H and Perry, SV (1962) Blochem J 85, 431-439 16 Robins, M J, Nalk, SR and Lee, A S K (1974) J Org Chem 39, 1891-1899 17 Umbrelt, W W, Bums, R H and Stauffer, J F (1972) Manometric and Biochemical Techniques, 5th Edn, pp 267-268, Burgess, Mlnneapohs 18 Flske, C H and SubbaRow, Y (1925)J B,ol Chem 66, 375-400 19 Parker, C A and Rees, W T (1960) Analyst 85, 587-600 20 Scott, T G, Spencer, R D, Leonard, N J and Weber, G (1970) J Am Chem Soc 92, 687-695 21 Etsenthal, R and Cormsh-Bowden, A (1974)Blochem J 139, 715-720 22 Zlelke, C L and Suelter, C H (1971) m The Enzymes (Boyer, P D, ed) 3rd Edn, Vol 4, pp 47-78, Academic Press, New York 23 Mornson, J F, O'Sulhvan, W J and Ogston, A G (1961) BlochJm Blophys Acta 52, 82-96 24 Wellner, V P, and Melster, A (1966) Biochemistry 5, 872-879 25 Vlscontml, M, Hoch, D and Karrer, P (1955) Helv Chlm Acta 38, 642-645 26 Falbrmrd, J G, Posternak, T and Sutherland, E W (1967) Blochlm Blophys Acta 148, 99-105 27 Gudlory, R J and Jeng, S J (1977) Methods Enzymol 46, 259-288 28 McLaughhn, C S and lngram, V M (1965) Bzochemlstry4, 1448-1456 29 Zamecmk, P C (1962) Blochem J 85, 257-264

508 30 Griffin, B E, Jarman, M, Reese, C B, Sulston, J E and Trentham, D R (1966) Blochenustry 5, 3638-3649 31 Pal, P K, Wechter, W.J and Colman, R F (1975) Blochem~stry 14, 707-715 32 Skorka, G , Shuker, P, Gdl, D, Zablcky, J and Parola, A H (1981) B~ochenustry 20, 3103-3109 33 Wdhams, N. and Coleman, PS (1982) J Btol Chem 257, 2834-2841 34 Anderson, E P (1973) m The Enzymes (Boyer, P D, ed ), 3rd Edn, Vol 9, pp 49-96, Academic Press, New York

35 Bergmeyer, H U (1974) Methods of Enzymauc Analysis, Vol 1, pp 509-510, Acadermc Press, New York 36 Watts, D C (1973) m The Enzymes (Boyer, P D , ed ), 3rd Edn, Vol 8, pp 383-455, Aeadenuc Press, New York 37 Tonomura, Y, Imamura, K, Ikehara, M, Uno, H and Harada, F (1967) J Blochem (Tokyo) 61,460-472 38 Reid, T W and Wdson, 1 B (1971) m The Enzymes (Boyer, P D, ed ), 3rd Edn, Vol 4, pp 373-415, Acaderrnc Press, New York 39 Hlratsuka, T (1982)J Blol Chem 257, 13354-13358