Pyrazolo[4,3-e][1,2,4]triazines: Purine Analogues with Electronic Absorption in the Visible Region

Synthesis of several pryrazolo[4,3-e][1,2,4]-triazines is described. The absorption spectrum of some 5-substituted derivatives was found to extend to the visible region. These compounds were found to inhibit some enzymes of purine metabolism, like xanthine oxidase or bacterial purine-nucleoside phosphorylase with Ki values in the 10-3 – 10-5 M range.


Introduction
Modified purine nucleosides and their analogues have found numerous applications in biological research and pharmacology [1][2][3]. In particular, the isomeric pyrazolopyrimidines and triazolopyrimidines (8-azapurines) exhibit favourable spectroscopic and biochemical properties and have potential for use in cancer and viral chemotherapy [3][4][5][6]. It has been found recently that another class of naturally occurring purine analogues produced by Pseudomonas fluorescens var. pseudoiodininum and Nostoc spongiaeforme such as pseudoiodinine and nostocine A ( Figure 1) display even more interesting spectral characteristics, i.e. their electronic absorption spectra extend well to the visible region [7,8].

Nostocine A Pseudoiodinine
Recent advances in nucleophilic substitution of hydrogen in heteroaromatics [9] and their successful application to the preparation of functionalized 1,2,4-triazines [10,11] prompted us to exploit this approach to make the N-unsubstituted pyrazolo [4,3-e] [1,2,4]triazine 1 [12]. In this paper some derivatives of such a system (compounds 2-9) have been synthesized from the common intermediate 1, as depicted in Scheme 1. Furthermore we show that these compounds both exhibit interesting spectral behaviour and also interact with enzymes of purine metabolism, and therefore may find applications as spectroscopic probes.
To avoid these problems another possibility to synthesize N-unsubstituted pyrazolo [4,3-e][1,2,4]triazines 4 and 7 is to use a N-protecting group. We decided to exploit ethyl vinyl ether as a new protecting group for NH-pyrazoles [14]. Exposure of pyrazolo [4,3- Compound 3 was next reacted with liquid ammonia at -33 o C. The product was cleanly deprotected with concentrated HCl in methanol at room temperature to give the corresponding amino compound 4, as a yellowish crystalline solid, in 80% yield. Similarly, the treatment of 3 with anhydrous hydrazine in tetrahydrofuran provides 5-hydrazino derivative 5 in 90% yield. Oxidation of the latter with yellow mercury (II) oxide results in replacement of the hydrazine group by hydrogen giving compound 6. Deprotection of 6 in methanol with concentrated HCl at room temperature for 12 hours gave the 5-unsubstituted pyrazolo[4,3-e][1,2,4]triazine 7 in excellent yield (see Scheme 1).

Spectral properties
In agreement with previous reports [7,8,15], the electronic absorption spectra of 1-H-pyrazolo-[4,3-e][1,2,4]-triazines differ very markedly from those of the analogous purines. Low-yield fluorescence in the visible region (ca. 500 nm) was observed for some compounds in neutral aqueous medium. The basic spectral characteristics of the synthesized compounds are summarized in Table 1.

Enzymatic assays
We have examined several purine metabolism enzymes to check if any of these effectively interact with the new compounds. It was found that enzymatic phosphorolysis of m 7 Guo [16], catalyzed by E. coli purine nucleoside phosphorylase (PNP), was inhibited by selected pyrazolotriazines at concentrations of 30-500 µM. The strongest inhibitor was the 5-methylsulfanyl derivative 1, with an IC 50 of ~40 µM. Compounds methylated on the N-1 pyrazole ring nitrogen exhibited much weaker inhibitory activity, resembling that reported for the analogous pyrazolopyrimidines [17].
Plots of 1/v vs. inhibitor concentration (not shown), obtained at constant substrate concentrations, were in some instances apparently nonlinear, suggesting complex mode(s) of inhibition and/or cooperative effects [18]. The corresponding IC 50 values, given in Table 2, are virtually insensitive to concentration of the m 7 Guo substrate, indicating noncompetitive type of inhibition. This behaviour is not exceptional among enzymes of oligomeric structure, like bacterial PNP's [18].
Calf spleen PNP, examined under identical conditions, was not inhibited by the pyrazolotriazines. This result is not surprising since the bovine enzyme is known to exhibit much higher specificity toward purines and purine moieties of nucleosides than the E. coli enzyme [19].
Since the analogous pyrazolopyrimidines, e.g. allopurinol, have therapeutic applications as known strong inhibitors of the xanthine oxidase (Xox) enzyme, [20], we examined the inhibitory activities of some of pyrazolotriazines toward commercially available Xox from buttermilk. Only weak inhibition was detected in the case of the unsubstituted compound 7, but somewhat surprisingly, moderate inhibitory activity was detected for the 5-thiomethyl derivative 1. The title compound was apparently not a substrate for Xox, at least at the moderate enzyme concentrations employed in this work. 2-Amino substituted pyrazolotriazine 4, which may be regarded as an analogue of guanine, was examined against a rabbit muscle guanine deaminase (GDA), but did not show any activity either as a substrate, nor as an inhibitor, at least under the conditions applied in this experiment (concentrations up to 150 µM, 50 mM phosphate pH 6.1, 25 o C).

Conclusions
Pyrazolotriazines are moderate inhibitors of bacterial (E. coli) PNP, an enzyme employed recently in cancer-oriented gene therapy experiments [21]. There is also a detectable inhibition of xanthine oxidase by the parent 3-methyl-7-azapyrazolo[4,3-d]pyrimidine, as well as its S-methyl derivative. These compounds are unique among purine analogues as having UV absorbtion spectra extending into the visible region (ca. 450 nm), some of them being also weakly fluorescent in aqueous solution (cf. Table 1), and thus are potentially applicable as spectroscopic probes for enzymes of purine metabolism.

General
All melting points are uncorrected and were determined using a Boetius melting point apparatus. Nuclear magnetic resonance ( 1 H-NMR) spectra were recorded on a Varian Gemini 200 MHz spectrometer in a suitable deutered solvent using TMS as internal standard. Mass spectra were obtained on AMD 604 [electron impact (EI)] and API 350 [electrospray ionization (ESI)] spectrometers. IR spectra were measured with a Magna IR-760 spectrophotometer, and UV on a Cary 319. Fluorescence spectra were recorded using a Perkin-Elmer LS-50B spectrofluorometer, equipped with a pulsed xenon light source. Enzymes: xanthine oxidase (Xox) from buttermilk and purine nucleoside phosphorylase (PNP) from calf spleen were purchased from Sigma, guanine deaminase (GDA) was from ICN, and the E. coli PNP was a gift from Dr. W. Koszalka. Hypoxanthine and 7methylguanosine (m 7 Guo) were from Sigma. All other reagents and chemicals were obtained from Aldrich Chemical Company and were used as received, unless otherwise noted. Enzymatic assays and UV spectra were run using a Cary-319 UV spectrophotometer (Varian), equipped with a thermostatic unit. Unless otherwise indicated, all the reactions were carried out in 50 mM phosphate, pH 7.0, at 25˚C. Inhibitor concentrations were evaluated spectrophotometrically, using data from Table 1. It was possible to run assays with inhibitor concentrations up to ~300 µM. Activity of PNP was assayed by following phosphorolysis of m 7 Guo in 50 mM phosphate, pH 7.0 [16], the reaction monitored spectrophoto-metrically at 260 nm. Typical substrate concentration was ca. 9 -135 µM. Xox activity was measured using the hypoxanthine oxidation test, with substrate concentrations 48 -150 µM [22], monitored spectrophotometrically at 300 nm. Activity of GDA was assayed at pH 6.1, using ~50 µM 8-azaguanine as a substrate, the reaction monitored at 265 nm.