CC BY
75
UDK 577.113.3
SOME RECENT FINDINGS IN THE BIOTECHNOLOGY OF BIOLOGICALLY IMPORTANT NUCLEOSIDES
I. A. MIKHAILOPULO1, A. I. MIROSHNIKOV
institute of Bioorganic Chemistry of Belarus National Academy of Sciences, Minsk, Belarus 2Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry of Russian Academy of Sciences,
Moscow, Russia
Е-mail: [email protected], [email protected]
Received 04.06.2013
Some recent findings in the biotechnology of biologically important nucleosides will be discussed, viz., (i) a new strategy of the cascade one-pot transformation of D-pentoses into nucleosides based on the extension and deepening of the knowledge of the mechanism of functioning of the ribokinase, phosphopento-mutase, and uridine, thymidine and purine nucleoside (PNP) phosphorylases, and the role of different factors (structural, electronic, stereochemical) in the glycoside bond formation, (ii) the modern chemistries of the chemo-enzymatic syntheses of nucleosides, (iii) the transglycosylation reaction using natural and sugar modified nucleosides as donors of carbohydrate residues and heterocyclic bases as acceptors catalyzed by nucleoside phosphorylases (NP).
Key words: nucleosides, bio-mimetic synthesis, chemo-enzymatic synthesis, enzymes of nucleic acid metabolism.
Nucleosides embrace a large family of natural and chemically modified analogues of great structural diversity and a broad spectrum of biological activity. Analogues of natural nucleosides as well as nucleoside antibiotics belong to the most important classes of antiviral drugs, and they are extensively used in the treatment of a variety of cancers. Base and sugar modified nucleosides are very valuable constituents of artificial oligonucleotides of medicinal potential making these oligomers more stable in biological fluids and improving their targeting proper-ties. The chemistry of many antiviral and anticancer drugs, as well as building blocks for oligonucleotide synthesis remains a challenging problem resulting in a high price of the desired compounds preventing them from extensive therapeutic application and use in oligonucleotide business (for a recent reviews, see [1-3].
Analysis of the state of the art of chemo-enzymatic synthesis of nucleosides led us to conclusion that the chemo-enzymatic methodology demonstrates a number of advantages over the chemical methods of nucleoside synthesis, viz., high total yield of desired products, simplicity of work-up of reaction mixtures and isolation of products, conform to
the principles of «green chemistry» to a greater extent vs. the fine organic synthesis [1-3]. Up to the present, a vast majority of the modified nucleosides have been synthesized by chemical methods. Despite the impressive progress achieved in the development of chemical methods, production of many antiviral and anticancer drugs, as well as other biologically active compounds remains a challenge. This leads to high drug costs and, therefore, prevents extensive biological trials and studies, as well as a wide therapeutic application. The need for the development of new strategies became apparent in the late 1970s.
The chemo-enzymatic (biotechnological) strategies are currently displacing multi-stage chemical processes, and this allows performing the key transformations with high selectivity and regio- and s£ereo-specificity. Considerable progress in the production of biologically important analogs of natural nucleosides has been achieved through the rational combination of chemical and biochemical transformations. Use of recombinant nucleoside phosphorylases (NPs) and N-deoxyribosyl transferases (NDTs) as biocatalysts for the synthesis of natural nucleosides and their modified analogs is of considerable importance
for the creation of modern technological processes. Noteworthy that both groups of enzymes complement one another and allow finding out a straightforward route to the desired compound. The use of the chemo-enzy-matic methods allows undoubtedly improve the price-quality ratio in the production of many medical drugs.
The possible areas of application of nucleoside phosphorylases for the synthesis of nucleosides, as well as the limitations of this methodology, have been investigated in detail; however, several very interesting enzymatic synthetic reactions deserve special attention, because they are crucial for understanding the mechanism of synthetic reactions catalyzed by these enzymes and may expand the scope of their practical use. It is well documented that the N7-atom of purines plays a very important role in the phosphorolytic cleavage of the gly-cosyl bond of purine nucleosides and, it seems, in the reversed synthetic reaction catalyzed by E. coli purine nucleoside phosphorylase (PNP; product of deoD gene; EC 2.4.2.1) as well, even though the mechanism of this reaction has not been adequately studied. The finding that 3-deazapurines and 1-deaza-, 3-deaza- and 1,3-dideazapurines (benzimidazoles, including fluoro-, chloro- and bromo-substituted) are good substrates for PNP allows to suggest a key role for two nitrogen atoms of the imidazole ring in the synthetic reaction (for a discussion, see [1]). Namely, one of them is involved in the binding of the heterocyclic base in the enzyme’s active site that may lead to an increase of the nucleophilicity of the second nitrogen atom. This facilitates, in turn, an attack by this atom on the electrophilic C1 car-
bon atom of a-D-pentofuranose-1-phosphate and eventually results in the formation of a glycosidic bond (Scheme 1) [2].
The mechanism of the synthetic reaction catalyzed by nucleoside phosphorylases in general and PNP in particular remains unclear. The participation of two nitrogen atoms of benzimidazole in this reaction seems to be obvious albeit there are two modes A and B of initial binding of the substrate. The Check authors studied substrate properties of a number of purine heterocyclic bases and their aza-and deaza-analogues using partially purified E. coli PNP and gel-entrapped cells of an auxotrophic thymine-dependent strain of E. coli as a biocatalyst for the transfer of the 2' -deoxy-D-ribofuranosyl moiety of 2'-deoxyuridine to bases. The reactions proceeded regio- and stereo-specifically affording purine nucleosides as well as 8-aza-2'-deoxyadenosine and 8-aza-2'-deoxyguanosine but no substrate activity of 7-deazapurines was observed. It was thus concluded that the presence of the nitrogen-7 of purines and their isosteric analogues is a prerequisite for the reaction [4]. However, there are several exceptions, viz., 5-aza-7-deazaguanine (1), 5-aza-7-deazaisoguanine (2) and N-(1,3,4-thiadiazol-2-yl)-cyanamide (3) are substrates for bacterial and mammalian purine nucleoside phosphorylases (PNP’s) (Scheme 2) [2].
These data imply that the mechanism of binding and activation of substrate in the catalytic center of E. coli PNP is not uniform and prompted us to gain insight into the functioning of the enzyme and to search for new substrates and inhibitors.
£ call PNP
A&partk acid 204
Scheme 2
Results and Discussion
The main line of the present study based on the suggested by us new strategy of nucleoside synthesis consisting in the cascade transformation of pentoses into nucleosides through intermediate consecutive formation of a pentofuranose-5-phosphate and a-D-pentofu-ranose-1-phosphate (PF-a1P) catalyzed by recombinant E. coli ribokinase (RK), phospho-pentomutase (PPM) and nucleoside (uridine, thymidine and purine nucleoside) phosphory-lases (Scheme 3) [5]. Before an idea of the cascade transformation — D(L)-pentose + base ^ nucleoside — was proved, the chemical synthesis of PF-a1P was considered as an important supplementary approach to the preparation of sugar and base modified nucleosides for biological and medicinal application. The chemo-enzymatic nucleoside synthesis consisting of the chemical preparation of PF-a1P followed by an enzymatic condensation with heterocyclic bases seemed to be more versatile and rather attractive.
Preparation of recombinant E. coli riboki-nase (RK), phosphopentomutase (PPM) and nucleoside phosphorylases (NP’s) on the practical level is of utmost importance for the project and this task was realized and all these enzymes were obtained as they spend at work [2, 3, 5-7]. Taking into account the aims of the present study, the preparations of the aforementioned enzymes have been obtained in dif-
ferent forms [lyophilized powders of high enzymatic (>90% after 3-5 years storage at +4 °C) stability], solution of a protein in diverse buffers] and different levels of purity (carefully purified for biochemical studies; sufficient purity for biotechnological studies).
A possibility of cascade one-pot enzymatic transformation of D-ribose or 2-deoxy-D-ribose into nucleosides employing pure recombinant E. coli ribokinase (RK) [D-pentose ^ pentose-5-phosphate (D-PF-5P)], PPM [D-PF-5P ^ a-D-pentofuranose-1-phosphate (D-PF-a 1P)], and nucleoside phosphorylases (NP’s) (D-PF-a1P + heterobase ^ nucleoside) coupled with the appropriate pyrimidine or purine heterobases was demonstrated (Scheme 4) [7]. Preliminary results of a caskade transformation of D-pentoses into nucleosides pointed to a reliability to develop practical methods for the preparation of antileukemic drugs (Cladribine, Fludarabine, Clofarabine, Nelara-bi ne) and a number of biologically important nucleosides.
We noted rather essential differences between the optimal reaction conditions for RK [5], PPM [7] and recombinant E. coli nucleoside phosphorylases [6]. Bearing this in mind, we have optimized the one-pot reaction conditions aiming at the finding out a compromised composition of the substrates allowing satisfactory function of the enzymes under investigation. It was found that D-ribose, 2-deoxy-D-ribose, D-arabinose and 2-deoxy-2-fluoro-D-arabinose are transformed into the respective nucleosides in good yields; D-xylose as well as some 2(3)-deoxyfluoro-D-pentofura-noses and L-pentofuranoses are not involved in the cascade transformation into nucleosides. This study is continued.
OH x HO X
RK
ATP/KCVMgCIj
OH x
D-Ribose: X = OH 2-Deoxy-D-ribose: X = H D-Arabnnose: X = OH 2-Deoxy-2-fluoro-D-arabinose: X = F
Н0^Ч О
HO X
о''Р^он
0O
a-PF-lP
Universal Glycosylating Agents Tor the Enzymatic Synthesis of Pyrimidine and Purine Nucleosides
Thymine/TP
Uraeil/UP,
-phosphate
Purines, 8-Azapurines 8- Aza- 7- d eaz ap urin es/PNP
Recombinant E. coli Enzymes:
RK -Ribokinase
PPM - Phosphopentomutase
TP - Thymidine Phosphorylase
PNP - Purine Nucleoside Phosphorylase
PPM
HO X
Thymidine: R = Me, X = H
2’-Deoxyuridine: R = X = H Ribothymidine: R = Me, X = OH Uridine: R = H
HO-v о 1
Л0,
CO.
HO
Cladribine; Fludarabine Clofarabine; Nelarabine 8-Aza-2’-deoxy-adenosine & -guanosine 8-Aza-7-deazapurine-2'-deoxyribosides
Scheme 4
Our studies unambiguously showed for the first time that 1,6-diphosphates of D-hexo-ses are not necessary for the transformation of 5-phosphates of D-ribose, 2-deoxy-D-ribose, D-arabinose and 2-deoxy-2-fluoro-D-arabi-nose into the corresponding 1-phosphates. This is one of the most important findings because it essentially simplifies the application of PPM as a biocatalyst within the one-pot cascade transformation of D-pentoses into nucleosides as well as for the transformation of chemically prepared 5-phosphates into intermediary 1-phosphates and then into nucleosides (Scheme 4) [7].
The strategy of cascade one-pot synthesis of nucleosides suggested by us [5,7] is of interest for development of practical methods for the preparation of biologically important nucleosides. It should be emphasized that this strategy allows preparing P-D-ribo-, 2-deoxy-P-D-ribo-, P-D-arabino- and 2-deoxy-2-fluoro-P-D-arabino-nucleosides of natural purine and pyrimidine (except for 2-deoxy-2-fluoro-P-D-arabino-nucleosides) bases, as well as base modified nucleosides. Indeed, R&D of this strategy led us to the simple and efficient preparation of antileukemic drugs 2-chloro-
2p-deoxyadenosine (Cladribine) from 2-deoxy-D-ribose and 2-chloroadenine, 9-(P-D-arabinofura-nosyl)-2-fluoroadenine (Fludarabine) (D-arabinose + 2-fluoroadenine), 9-(2-deoxy-2-fluoro-P-D-ara-binofuranosyl)-2-chloroadenine (Clofarabine) (2-deoxy-2-fluoro-D-arabinose + 2-chloroade-nine) and 2-amino-9-(P-D-arabinofuranosyl)-6-methoxypurine (Nelarabine) (D-arabinose + 2-amino-6-methoxypurine). Moreover, preliminary results of the synthesis of a number of biologically important nucleosides [e. g., 2'-deoxyribosides of 8-azapurines and 8-aza-7-deazapurines (vide infra)] have validated this strategy [8].
Transfer of a pentofuranosyl moiety of commercially available nucleosides or prepared by chemical methods to purine or pyrimidine bases catalyzed by nucleoside phosphorylases
(NPs) or N-deoxyribosyltransferases (DRTs) («transglycosylation reaction») was demonstrated to be a very efficient methodology for the synthesis of a plenty of analogues of natural nucleosides of biological and medicinal importance (for recent reviews, see, e.g., [1-3]). The bacterial nucleoside phosphorylases [purine (PNP), thymidine (TP) and uridine (UP)] reversibly catalyze (i) the phospho-
rolysis of nucleosides with an intermediary formation of a-D-pentofuranose-1-phosphate (PF-a1P), and (ii) the synthesis of new nucleosides in the presence of external heterocyclic bases. As distinct from nucleoside phosphorylases, DRTs catalyze the direct transfer of the 2'-deoxyribofuranosyl moiety of donors without intermediary formation of 2-deoxyribofu-ranose-1-phosphate. From the viewpoint of practical synthesis of sugar and base modified nucleosides, NP and DRT complement each other, but the latter have a strict specificity towards substrates thereby limiting their application.
Base halogenated nucleosides of benzimid-azole attract an attention of researchers since the pioneering studies by I. Tamm and his co-workers initiated in the early 1950s of the last century (for a review, see [9]). However, the most important findings from the viewpoint of the biochemical mechanism of antiviral activity of this class modified nucleosides [10] as well as their possible practical application [11] have been published during last two decades (reviewed in [12,13]).
It was earlier shown that benzimidazole (BI) and its derivatives with substituents in the benzene ring are good substrates of E. coli PNP in the transglycosylation reaction [1, 14-17]. In the present project, we studied 5,6-difluorobenzimidazole and its derivatives, one fluorine atom of which is replaced with methoxy, ethoxy, isopropoxy, 4-morpholino and N-methylpiperazino groups, as acceptors of the D-ribofuranose and 2-deoxy-D-ribofura-nose residues in the transglycosylation reaction employing uridne (Urd) and thymidine (Thd) [2'-deoxyuridine (dUrd)], respectively, as the pentofuranose donors and recombinant E. coli UP, TP and PNP nucleoside phosphory-lases as biocatalysts [18].
The reaction transglycosylation was employed for the synthesis of the P-D-ribo-
and 2-deoxy-P-D-ribonucleosides 4-15. The reaction conditions have been optimized depending on the kind of a ribofuranose donor, the donor/base ratio, and quantity of the recombinant E. coli enzymes and temperature of the reaction. The use of readily available natural purine ribonucleosides as donors of the ribofuranose residue in the transglycosy-lation reaction of BI allows employing PNP as the sole biocatalyst. On the contrary, the use of uridine as a donor of the ribofuranose residue requires two nucleoside phosphory-lases for transglycosylation reactions, viz., the recombinant E. coli uridine phosphorylase (UP) for the intermediary formation of a-D-ribofuranose-1-phosphate (D-Rib-a1P) that is accepted by E. coli PNP for the synthesis of the BI ribosides. The efficiency of both types of donors was tested in the transglycosylation of
5,6-difluorobenzimidazole and it was found that the use of uridine and two nucleoside phosphorylases is preferable in terms of yield of the desired nucleosides.
It was found that the 3:1 to 10:1 molar donor/base ratio and the use of 40 units of uridine phosphorylase (UP) per 1 mmol of uridine (60-160 UP units for 2'-deoxyuridine]) and 155-400 units of PNP are necessary to obtain the ribosides 4-15 in good yields calculated for isolated products (Table 1). Reactions were conducted at 52 °C in the K,Na-phosphate buffer (5-20 mM; pH 7.0) monitoring the formation of the products by HPLC, the conversion of base into nucleoside(s) was 98.5%. It is noteworthy that the synthesis of 2'-deoxyri-bosides was completed in 1-3 h, whereas the trans-ribosylation proceeded much more slowly and 22-28 h required achieving high yields of the reaction products. Similar trend was earlier observed in the case of the ribo and 2'-deoxyribo nucleoside syntheses using the whole E. coli cells as biocatalyst (Scheme 5)
[18].
Table 1. The synthesis of benzimidazole nucleosides
Compd Pfr 1) R Yield for isolated product (%) Ratio of isomers
a b
4 Rib -F 77 -
5 dRib 68 -
6 Rib -OMe 54 95 5
7 dRib 51 41 59
8 Rib -OEt 98 89 11
9 dRib 85 70 30
10 Rib -OiPr 75 100 -
11 dRib 65 96 4
12 Rib 0 75 100 -
13 dRib 80 100 -
14 Rib 79 100 -
15 dRib 68 100 -
Pfr = P-D-Pentofuranosyl (Rib — P-D-ribofuranosyl; dRib — 2'-deoxy-P-D-ribofuranosyl).
Regio-isomeric structure of all isolated nucleosides was proved by scrupulous analysis of the 1H and 13C NMR spectral data (incl. ^H/H] & ['H,13C] 2D: COSY, HSQC, HMBC and NOE spectra).The predominant (OMe and OEt) or exclusive (OiPr, 4-morpholino and N-methylpiperazino) formation of the 5-substi-tuted 6-fluoro-1-(P-D-ribofuranosyl)-benzimi-dazoles was observed. The formation of the regio-isomeric 5-fluoro-1-(2-deoxy-P-D-ribo-furanosyl)-6-methoxy(ethoxy, i-propoxy)ben-zimidazoles was observed in the trans-2-deoxyribosylation reaction of the corresponding bases. The predominant or exclusive formation of the regio-isomeric N1-nucleosides with bulky 5-substituents of 6-fluorobenzimidazole points to a large hydrophobic pocket in the E. coli PNP active site that can accommodate these groups.
During recent years, the use of D-pentofu-ranose-1-phosphates (PF-1Ps) as substrates of an enzymatic synthesis of nucleosides attracts much attention [1-3]. It should be stressed that the enzymatic and chemical syntheses of D-ribofuranose-1-phosphate and its 2-deoxy counterpart have rather reach prehistory. However, only recently a few interesting reports from the point of view of the possible practical application have been published. There are two lines of investigation in this area, viz., (i) biochemical (microbial, enzymatic) re£ro-synthesis of 2'-deoxyribonucleosides employing the triose phosphate isomerase (TRI) and 2-deoxy-D-ribose-5-phosphate aldolase (DERA) enzymes and (ii) chemical
synthesis of D-pentofuranose-1-phosphates (PF-1Ps) followed by the enzymatic condensation with heterocyclic bases.
Within this line of investigation, the laborious preparation of the a-D-pentofuranose-1-phosphates (PF-a1Ps) is a serious bottleneck of this approach. However, despite rather complex synthesis of PF-a1Ps this chemo-enzymatic methodology for the synthesis of biologically valuable nucleosides represents an advisable alternative to an enzymatic transfer of sugar fragment of nucleoside to heterocyclic base («transglycosylation reaction») as well as to the microbial (enzymatic) retro-syn-thesis [1-3].
Scrutiny of the chemical methods for the preparation of pento(hexo)furanose-1-phosphates as well as different methods of an activation of the anomeric carbon atom shows that the most of them are laborious and low-yielding. As might be expected, the formation of anomeric mixtures was usually observed, and only the tedious crystallization-induced asymmetric transformation leads to the predominant formation of the desired dRF-a1P (reviewed in [1-3]). From the standpoint of simplicity, method suggested by MacDonald
[19] for the synthesis of pyranose-1-phos-phates seems to be the most efficient one and prompted us to apply it for the synthesis of a-D-arabinofuranose-1-phosphate (19; AraF-a1P) [20]. We focused our studies on the development of practical chemical synthesis of PF-1Ps and selected D-arabinose as a starting pentose bearing in mind that a plenty of
purine and pyrimidine P-D-arabinofurano-sides manifest antiviral and anticancer activities.
It is remarkable that AraF-a1P synthesized by Wright & Khorana was found to be inactive as a substrate for the pyrimidine deoxyribo-side phosphorylase of E. coli (thymidine phos-phorylase, TP) and for the purine nucleoside phosphorylase (PNP) of fish muscle [21]. These data taken together prompted us to synthesize AraF-a1P and to investigate its substrate properties for recombinant PNP of E. coli.
Treatment of the peracetyl derivative of D-arabinose 16 with anhydrous phosphoric acid under MacDonald’ conditions followed by work-up gave rise to the formation of an amorphous powder or viscous oil consisting of a-D-arabinofuranose-1-phosphate (19; AraF-a 1P) and P-D-arabinopyranose-1-phosphate (20; Arap-P1P) (ca. 50%, combined; the 19:20 ratio was from 1.5:8 to 1:2, according to 1H NMR). The structure of both phosphates was proved by (i) the scrupulous analysis of the 1H and 13C NMR spectra as well as [1H,1H] and [1H,13C] 2D spectra of the mixtures of different anomer ratio, (ii) comparison with the 1H and 13C NMR data for the related 1-phosphates, and (iii) the comparative analysis of the ab initio calculations of 1-phosphates of selected a-D-pentofu-ranoses and two conformers, 1C4 and 1C4, of P-D-arabinopyranose (Table 2).
In consent with experimental results, the ab initio calculations point to a higher stability of both possible conformers of P-D-ara-binopyranose-1-phosphate, viz., 1C4 and 1C4, vs a-D-arabinofuanose-1-phosphate [AE = E(AP-P1P; 4C1) — E(AP-P1P; 4C1) = -11.7 kcal/mol; [AE = E(AP-P1P;4C1) — E(AF-a1P; O4-exo) = -26.9 kcal/mol] (Table 2). It was previously shown that the ratio of a- and P-anomers
strongly depends on the reaction conditions implying the a/P-anomerization to the thermodynamically more stable anomer and the possibility of furanose/pyranose isomerization during the treatment of peracyl sugars with anhydrous phosphoric acid cannot be excluded. In toto, it appears to be rather likely the formation of P-D-arabinopyranose-1-phos-phate along with the desired a-D-arabinofura-nose-1-phosphate using peracetyl D-arabino-furanose as the starting compound. In this context, it is noteworthy that the MacDonald method was up to present study successfully employed for the synthesis pyranose-1-phos-phates [19]. To escape or diminish the formation of the pyranose-1-phosphates, we focused our further studies on the preparation of new starting sugars, primary hydroxyl group of which would be protected with an acid resistant function. However, methyl 2,3,5-tri-O-benzoyl-a-D-arabinofuranoside (17) was stable under MacDonald’s reaction conditions and was recovered unchanged from the reaction mixture; on the contrary, 1-O-acetyl-2,3,5-tri-
O-benzoyl-D-arabinofuranose (18) showed good reactivity, but gave a mixture of the 1-phosphates 19 and 20 similar to that obtained from the peracetyl derivative of D-arabinose. In toto,
The mixture of isomeric phosphates AraF-a 1P and Arap-P 1P of different ratios was tested in the reaction with 2-fluoroadenine and 2-amino-6-methoxypurine catalyzed by the recombinant E. coli purine nucleoside phos-phorylase (PNP). It was found that the pyra-nose phosphate Arap-P1P did not interfere with the reaction of the furanose phosphate AraF-a1P with purine bases. Moreover, the rate of formation of 9-(P-D-arabinofuranosyl)-2-fluoroadenine (Fludarabine) under optimum conditions [water solution (pH = 7.0); 55 °C,
Table 2. The ab initio geometry optimization of a(P)-D-pentofuranose(pyranose)-1-phosphates (as mono sodium salts) (HyperChem, 8.1; in vacuo, basis set; medium 6-31G*)
Compound (1-Phosphate) Positive partial charge at the C1 carbon atom Total (binding) energy kcal/mol Conformation of the pento-furanose(pyranose) ring
Ribo (RF- a 1P) 0.425 -808 850.3 C1-exo
2-Deoxyribo (dRF- a 1P) 0.454 -762 140.7 C3-endo
Arabino (AF- a 1P) a) 0.464 -808 841.6 O4-exo
в-D-Arabinopyranose 0.410 -808 868.5 4C1(more stable)
(AP-P1P) 0.451 -808 856.8 4C1(less stable)
a)a-D-arabinofuranose-1-phosphate is thermodynamically less stable vs both conformers of P-D-ara-binopyranose-1-phosphates, viz., 4C1 and 4C1; note that among the two pyranose conformers the former is more stable than the latter. Colored data are for isomeric compounds with analogous elemental composition.
1-3 h] was found to be similar to that of the PNP catalyzed condensation of a-D-ribofura-nose-1-phosphate with 2-fluoroadenine. Reaction of furanose phosphate AraF-a1P with
2-fluoroadenine is shifted towards the formation of Fludarabine and ca. 5% of the initial base remained after 3 h in the reaction mixture, keeping of which at 14 °C for 24 h resulted in crystallization of Fludarabine in 77% yield (Scheme 6) [20].
As distinct from 2-fluoroadenine, the enzymatic reaction of 2-amino-6-methoxy-purine with phosphate Ara-a1P reached equilibrium at ca. equimolar ratio of a base and its nucleoside, 2-amino-9-(P-D-arabinofurano-syl)-6-methoxypurine (Nelarabine). Notably that the reaction transglycosylation of 2-amino-6-methoxypurine using 1-(P-D-arabino-furanosyl)uracil as a donor of the arabinofura-nose residue and the E. coli uridine and purine nucleoside phosphorylases as biocatalysts gave rise to the preparation of Nelarabine in 53% yield [22]. Unexpectedly, 2-amino-6-methoxypurine revealed lower substrate activity compared to that of 2-fluoroadenine, whereas the respective nucleoside Nelarabine showed higher substrate activity for PNP vs Fludarabine. As a consequence, the reaction reached equilibrium after 36 h at ca. equimolar concentration of base and its nucleoside. It is noteworthy that very similar results were observed in the case of the trans-ribosylation
[uridine as a donor of the ribofuranose residue; 23 °C; 5 mM KPB (pH 7.0)] and the trans-arabinosylation [1-(P-D-arabinofura-nosyl)-uracil as a donor of the arabinofuranose residue; 45 °C; 5 mM KPB (pH 7.0)] of 2-amino-6-methoxypurine employing UP and PNP as biocatalysts [20].
Unexpectedly, the enzymatic reaction of hypoxanthine and a mixture of phosphates 19 and 20 (ca. 1:2) in the presence of PNP (172 units) in water solution (55 °C) proceeded very slowly. After 7 days, in the reaction mixture remained ca. 10% of the starting base (HPLC) and 9-(P-D-arabinofuranosyl)-hypoxanthine (Ara-Hyp) was isolated in 80% yield after standard work-up and chromatography.
The synthesis of 1-(P-D-arabinofura-nosyl)thymine (ara-Thy) from thymine and the phosphate 19 (as a ca. 1:2 mixture of phosphates 19 and 20) in water solution (55 °C) in the presence of the recombinant E. coli thymidine (TP) and uridine (UP) phosphorylases was studied. As expected (cf. [21]), TP did not catalyze the formation of the nucleoside. The formation of ara-Thy in the presence of E. coli UP (144 units) proceeded smoothly and ca. 70% of the base was transformed into ara-Thy after 7 days that was isolated in 61% yield [20].
In continuation of these studies, we investigated the chemical synthesis of 2-deoxy-2-fluoro-a-D-arabinofuranose-1-phosphate (2FAra-a1P) and its substrate properties for
the recombinant E. coli nucleoside phosphory-lases, as well as the cascade transformation of
2-deoxy-2-fluoro-D-arabinose, D-arabinose, D-xylose and D-ribose into the corresponding P-D-pentofuranosides of 2-chloroadenine. The modified MacDonald method was employed for the synthesis of 2FAra-a1P and its use as a universal glycosylating agent for the nucleoside synthesis was studied in comparison with a-Ara-1P. It was found that the phosphate a-2FAra-1P, like a-Ara-1P, is good substrate of the recombinant E. coli purine nucleoside phosphorylase (PNP) and can be used for the synthesis of a number of purine nucleosides (e.g., Clofarabine and related nucleosides of
2,6-diaminopurine and hypoxanthine) as well as base modified derivatives [e.g., 5-aza-7-deaza-9-(2-deoxy-2-fluoro-P-D-arabinofura-nosyl)guanine]. Unexpectedly, the phosphate a-2FAra-1P, unlike a-Ara-1P, showed no substrate activity for the recombinant E. coli uridine phosphorylase (UP); both phosphates devoid substrate activity towards the recombinant E. coli thymidine phosphorylase (TP) (for a preliminary report, see [23]).
Recently we have described an enzymatic synthesis of nucleosides of N6-benzoyladenine and N2-acetylguanine using the respective acylated bases as acceptors of the pentofura-nose residues and recombinant E. coli PNP as a biocatalyst [24]. Based on this finding, a new approach for the synthesis of orthogonally protected nucleosides was suggested and verified by the preparation of N6-benzoyl-2' ,3'-dideoxy-3' -Fmocaminoadenosine (23). Commercially available 3'-amino-3'-deoxythy-midine (21) and N6-benzoyladenine were used as substrates of the enzymatic coupling catalyzed by E. coli PNP to give 3'-amino-2',3'-dideoxy-N6-benzoyladenosine (22) in high yield. Standard treatment of the latter with Fmoc-OSU yielded the desired nucleoside 23 with orthogonally protected amino groups of
3' -amino-2' ,3' -dideoxyadenosine (Sche-
me 7) [24].
These studies were continued using N2-acetyl-O6-[2-(4-nitrophenyl)ethyl]guanine (24) [25] as an acceptor in (i) the transglycosylation reaction employing thymidine (25) and its 3'-aminodeoxy derivative (26) as donors of the pentofuranose residues and recombinant E. coli TP and PNP as biocatalysts (Scheme 9, path A), and (ii) the synthesis of the ribo- and arabino-nucleosides 29 and 30 in the cascade one-pot transformation of the corresponding D-ribose (27) and D-arabinose (28) into the nucleosides in the presence of the recombinant E. coli RK, PPM and PNP (Scheme 9, path B).
The base 24 is poorly soluble in the phosphate buffer and a 3:2 (vol) mixture of K-phos-phate buffer (20 mM; pH 7.1) and DMSO was, therefore, used in the studied reactions. It was found that the base 24 displays satisfactory substrate activity for PNP and the nucleosides 29 and 30 form in the reaction mixture in 64 and 42% (HPLC data), respectively, after 30 h at 50 °C (Scheme 8, A). Both individual nucleosides have been isolated by silica gel column chromatography in 50 and 34% yields, respectively. The use of the acceptor/donor ratio of 1.0:1.5 (mol) gave rise to the formation of the nucleosides 29 and 30 in 78 and 60% yields (HPLC), respectively, after 96 h at 50 °C [26].
A good substrate activity of the base 24 for PNP prompted us to test it in the one-pot synthesis of nucleosides from D-pentoses in the cascade transformation in the presence of RK, PPM and PNP (Scheme 8, B). It was found that the riboside 33 forms in the reaction mixture in 65% yield after incubation for 48 h at 48 °C. Under similar reaction conditions, the formation of the arabinoside 34 proceeds more slowly affording 10% of the product after 48 h.
The finding that base 24 with voluminous NPE group shows good substrate activity points to a large hydrophobic pocket in PNP
TP/Pi
Me.
YjT H0“YV
\ / -thymine
X
25, X = OH
26, X = NH2
ONPE
V'OH
©o
24/PNP
■Pi
27, X = OH
28, X= NH2
CO.
O^N^N^NHAc
29, X = OH
30, X = NH2
Reaction conditions:
A - K-phosphate Duffer (20 mmol pH 7 1J/DMSO, 3 2 (vol)
50°C; 25,26/24 ratio 1.2:1 0 (mol), 30 h isolated yields: 29 (50%); 30 (34%) j: !■ ' ■ -■ ■ ; ■■ . ■■■ _
(0.02 mmoO, ATP (0.013 mmol), 48°C; 31,32/24 ratio 1.2:1.0 (mol) yields by HPLC: 33 (65%); 34 (10%)
- „O HO—V o 24 HO—V n ,B
ho'T^Uqh y-OH -------------------------► Y y
I x Enzymes (units/mmol of 24):
OH HO X RK (10), PPM (8) & PNP (12) HO X
31, X = ribo-OH (D-ribose) 33, X = ribo-OH
32 ' f-f ■■■•■ :>■ .iM: jl 34 • r r '.'A-
Scheme 8
active site that can accommodate this group. The dimensions of this pocket are illustrated by the overlay of imidazole fragments of the geometry optimized structures of N2-acetyl-O6-[2-(4-nitrophenyl)ethyl]guanine and 5-morpholino-6-fluorobenzimidazole (vide supra) (Fig. 1).
Remarkably, the pocket in the E. coli PNP active site exceeds the dimension of purine base and allows suggesting a minimal contribution of the C6 amino/carbonyl groups and N1 atom to the substrate binding.
Nucleosides of 8-azapurines and 8-aza-7-deazapurines (purine numbering throughout) are applicable as drugs and tools in chemistry, chemical biology and molecular diagnostics (see, e.g., [27-31]). Chemical synthesis of these nucleoside shape mimics suffers from the formation of mixtures of regio-isomers and a/P-anomers in the case of the synthesis 2-deoxy-D-pentofuranosides and arabinosides. This makes it necessary to conduct time-consuming separation. As a result, the desired nucleosides are obtained in moderate or low yields (e.g., [29,31]).
We applied the enzyme catalyzed glycosy-lation to diversity of heterocyclic bases belonging to the classes of 8-azapurines (35 & 36) and 8-aza-7-deazapurines (37-40) using various donors of the 2'-deoxy-D-ribofura-nosyl residue and recombinant E. coli PNP as a biocatalyst (transglycosylation reaction; Scheme 9; path A) and the enzymatic cascade
transformation of 2-deoxy-D-ribose into the nucleosides (Scheme 10; path B) using recombinant E. coli RK, PPM and PnP (one-pot synthesis). The enzymatic synthesis of N9-2'-deoxy-P-D-ribonucleosides of a number of 8-azapurines (44, 45) and 8-aza-7-deaza-purines (46-49) was studied (Scheme 9) [8].
At first, the substrate activity of bases 35-
39 was tested in the transglycosylation reaction performed under standard conditions using 2'-deoxyguanosine (41) as a glycosyl donor (Scheme 10, A). All these bases are satisfactory substrates and the respective nucleosides were obtained in good yields. The structure of the nucleosides 44-48 was confirmed by 1H- and 13C-NMR spectroscopy data and UV spectra as well by comparison with already published spectral data. We have not observed the formation of the regio-isomeric nucleosides.
In the next series of experiments, we studied the one-pot synthesis of nucleosides 44-48 with 2-deoxy-D-ribose and heterocyclic bases 35-39 in the presence of recombinant RK, PPM and PNP (Scheme 10; path j). Under reaction conditions employed, the formation of 8-aza-2'-deoxyadenosine (44) and 8-aza-2'-deoxyguanosine (45) proceeded slowly affording the nucleosides in moderate yields. On the contrary, 8-aza-7-deazapurines 37-39 showed satisfactory substrate activity and the respective nucleosides 46-48 were formed in yields more than 50% after 20 h [8].
fTfjf
VV4
35. X = NH2, Y = H
36, X = OH, Y = NH2
37, X = NH2, Y = H
38, X = OMe, Y = NH2
39, X = OH, Y = H
40, X = CI, Y = NH2
H0—\ r, в 'HO—ч jn J
1 \У PNP/P, N ц>0-^0Н
не/ -BH . Н9 0О
41, В = guanin-9-yl 42, В = cytosin-1-yl 43, B = uracil-1-yl 26, В = thymin-1-yl 27 Т PPM
PNP/Base
-P,
B
2-Deoxy-D-ribose
RK/ATP
I
Л^он
44, В = 8-azaadenin-9-yl
45, В = 8-azaguanin-9-yl
46, В = 8-aza-7-deazaadenin-9-yl
47, В = 8-aza-7-deaza-C^-methylguanin-9-yl 46, В = 8-aza-7-deazahypoxanthm-9-yl
49, В = 2-amino-8-aza-6-chloro-7-deazapurin-9-yl
50
Scheme 9
Recombinant E. coli enzymes:
PNP - purine nucleoside phosphorylase
RK ■ ribokinase
PPM - phosphopentomutase
Fig. 1.
The regio-specific 2' -deoxy-D-ribosylation of 8-azaadenine (35) and 8-azaguanine (36) catalyzed by recombinant E. coli PNP is in line with published data [32]. On the contrary, gly-cosylation of anions of 8-azapurines with 2-deoxy-3,5-di-O-(4-toluoyl)-a-D-erythro-pentofuranosyl chloride gave a very complex mixture of regio-isomers and their a,P-anomers [28, 29].
From a chemical point of view, the formation of the glycosyl bond results from a nucle-ophilic attack of a nitrogen atom of base at the electrophilic C1-carbon atom of 2-deoxy-D-ribofuranose-1-phosphate (27). The regio-selectivity of the enzymatic glycosylation is governed by the binding mode of base in the catalytic center of the enzyme enabling a nucleophilic attack of a nitrogen atom of the sp2 hybridized imino -C=N9- tautomer at the electrophilic center of co-substrate (Scheme 1). A similar type of binding of the substrate
can be tolerated with a high probability for natural purine bases as well as for 8-aza-purines 35 and 36 leading to the exclusive formation of N9-glycosides. However, the formation of nucleosides 46-48 was rather unexpected in the light of the earlier data pointing to the crucial importance of the nitrogen-7 of purines and their isosteric analogues in the synthetic reaction catalyzed by E. coli PNP [2, 3, 32].
Detailed analysis of the mechanism of the phosphorolysis of purine nucleosides by E. coli PNP showed that the base binding site is formed mainly by Asp204 and to some extent by Phe159 [33]. Taking into account that this is an equilibrium reaction, one can expect that the same amino acid residues make the main contribution in binding of substrate in the synthetic reaction. The Asp204 interacts with the nitrogen-7 and likely the C6 substituents of natural purine bases giving rise, in all likelihood, to the proper base orientation and to enhancement of the nucleophilic properties of the nitrogen-9, i.e., activation of the substrate. It is, therefore, surprising that replacement of the nitrogen-7 with CH group did not abolish the substrate activity of bases 37-39 pointing to the rather efficient contribution of the Asp204-C6 substituent interaction in the correct binding and activation [1]. To prove the role of such interaction, we investigated the substrate properties of base 40, which has no groups (NH2, carbonyl, OH or
OR) that could imitate an interaction of natural purine substrates with Asp204. It was surprising to found that base 40 still retains moderate substrate activity despite the absence of any interactions with Asp204 and the extremely low solubility.
Analysis of the crystal structure of the ternary complex of hexameric E. coli PNP with Formycins A and B showed that the Ser90 is involved in binding of the bases [33a,b]. Moreover, crystallographic data for the adenine binding to the active site of E. coli PNP clearly showed a close proximity of Ser90OY to the carbon-8 of the base [33c,d]. These data together with the aforementioned considerations suggest the possible explanation for the good substrate properties of bases 46-48 and moderate activity of base 49, viz., Ser90OY is hydrogen bonded to nitrogen-8 (purine numbering) of the bases giving rise to the acceptable base orientation in the catalytic site of E. coli PNP followed by activation of the nitrogen-9 in productive complex 27 (Fig. 2). Contribution of Phe159 of E. coli PNP in both processes, binding and activation, appears to be analogous to that of the natural bases. Thus, in the case of 8-aza-7-deaza purine analogs Ser90 residue of the catalytic site of E. coli PNP takes effect of Asp204 in the case of natural purine substrates.
The spatial tautomeric structures of base
40 in complex with Ser90OY have been analyzed by the restricted Hartree-Fock (RHF) ab initio method using basis set of 6-31** FIREFLY QC package, which is partially based on the GAMESS (US) source code. The files of MOPAC format containing Z-matrix of internal coordinates obtained by the PM3 geometry optimization was used as starting approximation for the ab initio calculations. The following main dimensions were obtained for the respective structure E. coli PNP/base 40 — Ser90-OY-H — N8 0.97 & 1.86 E; Ser90-OY •••H-N9 2.39 & 0.99 E; =N8-N9(H)- 1.39 E; 27 — Ser90-OY •••H-N8 2.37 & 0.99 E; Ser90-OY-H — N9 0.96 & 1.83 E;
-N8(H)-N9 = 1.37 E; these data are in fair agreement with strong hydrogen bonding of the base enabling the correct binding followed by activation [8].
To prove this suggestion, the Ser90Ala mutant of E. coli PNP was prepared and its catalytic activity in the synthesis of purine and 8-aza-7-deazapurine nucleosides studied.
The transribosylation of hypoxanthine and
8-aza-7-deazahypoxanthine (allopurinol) using uridine and the recombinant E. coli uridine phosphorylase for the generation of
Fig. 2. Schematic presentation of productive E. coli PNP/base 40 complex
intermediary a-D-ribofuranose-1-phosphate revealed the differences in rates of the formation of inosine and allopurinol riboside (AR) along with high yields of both nucleosides. In the presence of the native E. coli PNP, the formation of (i) inosine reached 90% after ca. 30 min and after next 30 min established an equilibrium at ca. 2:8 base — nucleoside ratio in the reaction mixture, and (ii) 8-aza-7-deaza-9-(P-D-ribofuranosyl)purine (AR) proceeded somewhat slowly achieving a 95% conversion of allopurinol to its N9-riboside 3a after 24 h, witnessing to non-critical role of the N7-nitro-gen atom in the binding and activation of heterocyclic substrate in the synthesis of nucleosides catalyzed by the native E. coli PNP. Replacement of the native PNP with the mutated enzyme in the synthetic reactions resulted in (i) the slowing down of the rate of the inosine formation (35% yield of inosine after 1 h; an equilibrated ca. 2:8 mixture of base — nucleoside after 24 h), and (ii) dramatic reduction of the rate of the AR formation attaining ca. 20% yield after 48 h. These data allow implying the moderate contribution of the Ser90 residue of the catalytic center of the native E. coli PNP in the binding of a-D-Rib-1P. Dramatic reduction of the rate of the allopurinol riboside formation points to an unique importance of Ser90 residue of the native E. coli PNP in the binding and activation of allopurinol as well as other 8-aza-7-deazapurines [especially 2-amino-8-aza-6-chloro-7-deazapurine (40)] in the enzymatic synthesis of their nucleosides (for a preliminary report, see [34].
Thus, the pure recombinant E. coli riboki-nase (RK), phosphopentomutase (PPM) and nucleoside phosphorylases [uridine (UP), thymidine (TP) and purine nucleoside (PNP)] were prepared on a multigram level [5-7].
A new strategy for the synthesis of nucleosides consisting in one-pot enzymatictransfor-mation of D-ribose, 2-deoxy-D-ribose, D-arabi-nose and 2-deoxy-2-fluoro-D-arabinose in the presence of heterocyclic bases into the pyrimidine and purine nucleosides was suggested and validated. It consists in consecutive transformation of D-pentoses into nucleosides under the action of recombinant E. coli ribokinase (RK) [D-pentose pentose-5-phosphate (D-PF-5P)], PPM [D-PF-5P a-D-pentofuranose-1-
phosphate (D-PF-a1P)], and nucleoside phos-phorylases (NPs) (D-PF-a1P + heterobase nucleoside). It was unambiguously shown for the first time that 1,6-diphosphates of D-hex-oses are not necessary for the transformation of 5-phosphates of D-pentoses into the corresponding 1-phosphates. Practical methods for the synthesis of antileukemic drugs (Cladribine, Fludarabine, Nelarabine, Clofara-bine) and a number of biologically important nucleosides were developed [7].
The enzymatic trans-ribosylation and trans-2-deoxyribosylation was used for the synthesis of the corresponding nucleosides of
5,6-difluorobenzimidazole and its derivatives, one fluorine atom of which is replaced with methoxy, ethoxy, isopropoxy, 4-morpholino and N-methylpiperazino groups aiming at search of new biologically active nucleosides. A large pocket close to the area corresponding to the purine N1 atom was disclosed [18].
The MacDonald’ method was studied for the synthesis of a-D-arabinofuranose-1-phos-phate (AraF-a1P) using diverse starting pera-cyl derivatives of D-arabinose. Mixtures of different ratios of AraF-a1P and P-D-ara-binopyranose-1-phosphate (AraP-P1P) were obtained and they will be used in the enzymatic condensations with purine and pyrimidine
bases giving rise to the nucleosides, incl. Fludarabine, Nelarabine and 1-(P-D-arabino-furanosyl)-thymine. Similar approach was studied for the synthesis of antileukemic drug
9-(2-deoxy-2-fluoro-P-D-arabinofuranosyl)-2-chloroadenine (Clofarabine) [20,23].
An enzymatic synthesis of 2 -deoxyribonu-cleosides of 8-azapurines and 8-aza-7-deaza-purines has been studied using the transglyco-sylation reaction and cascade one pot synthesis from 2-deoxy-D-ribose and nucleobases. Good substrate activity of 8-aza-7-deazapurinesto-wards recombinant E. coli purine nucleoside phosphorylase (PNP) was disclosed and mechanism of binding and activation of these analogs in the catalytic site of E. coli PNP was studied. The participation of Ser90OY of E. coli PNP in the binding of 8-aza-7-deazapurines in the catalytic center of PNP followed by the formation of productive complex and glyco-sidic bond was suggested [8] and proved by the preparation of the Ser90Ala mutant of E. coli PNP and investigation of its substrate properties [34].
The financial support of this study by the International Science and Technology Center (www.istc.ru; project #B-1640) is gratefully acknowledged.
The authors are deeply indebted to Prof. Dr. Frank SEELA (Laboratory of Bioorganic Chemistry and Chemical Biology, Center for Nanotechnology, Heisenbergstrafle 11, D-48149 Mbnster, Germany) for supplying most of heterocyclic bases used in the present work.
IAM is thankful to the Alexander von Humboldt-Stiftung (Bonn — Bad-Godesberg, Germany) for partial financial support of this study.
REFERENCES
1. Mikhailopulo I. A. Biotechnology of Nucleic Acid Constituents — State of the Art and Perspectives // Current Org. Chem. — 2007. — V. 11. — P. 317.
2. Mikhailopulo I. A.,Miroshnikov A. I. New Trends in Nucleoside Biotechnology // Acta Naturae. — 2010. — V. 2, N 5. — P. 36.
3. Mikhailopulo I. A., Miroshnikov A. I. Biologically important nucleosides: modern trends in biotechnology and application // Mendeleev Communs. — 2011. — V. 21. — P. 57.
4. Doskocil J., Holy A. // Coll. Czech. Chem. Communs. — 1977. — V. 42. — P. 370 (see also the works cited therein); Votruba I., Holy A., Dvorakova H. et al. // Ibid. — 1994. — V. 59. — P. 2303.
5. Chuvikovsky D. V., Esipov R. S., Skoblov Y. S. et al. // Bioorg. Med. Chem. — 2006. — V. 14. — P. 6327.
6. Esipov R. S., Gurevich A. I., Chuvikovsky D. V. et al. // Protein Expes. Purif. — 2002. — V. 24. — P. 56.
7. Miroshnikov A. I., Esipov R. S., Konstantino-vaI. D. et al. // The Open Conference Proc. J. — 2010.- V. 1. — P. 98.
8. Stepchenko V. A., Seela F., Esipov R. S. et al. Enzymatic Synthesis of 8-Azapurineand 8-Aza-7-deazapurine 2’-Deoxyribonucleosides // SynLett. — 2012. — V. 23. — P. 1541.
9. De Clercq E. // Adv. Virus Res. — 2009. — V. 73. — P. 1.
10. Krosky P. M., Borysko K. Z., Nassiri M. R. et al. // Antimicrob. Agents Chemother. — 2002. — V. 46. — P. 478; Evers D. L., Koma-zin G., Shin D. et al. // Antiviral Res. —
2002. — V. 56. — P. 61; Gudmundsson K. S., Tidwell J., Lippa N. et al. // J. Med. Chem. — 2000. — V. 43. — P. 2464.
11. Prichard M.N. // Rev. Med. Virol. — 2009. — V. 19. — P. 215; Williams S. L., Hartline C. B., Kushner N. L. et al. // Antimicrob. Agents Chemother. — 2003. — V. 47. — P. 2186; Hwang J. S., Kregler O., Schilf R. et al. // Virol. — 2007. — V. 81. — P. 11604; Hwang J. S., Schilf R., Drach J. C. et al. // Antimicrob. Agents Chemother. — 2009. — V. 53. — P. 5095.
12. Trofe J., Pote L., Wade E. et al. // Ann. Phar-macother. — 2008. — V. 42. — P. 1447; Drugs R. D. Maribavir: 1263W94, Benzimi-davir, GW 1263, GW 1263W94, VP41263. — 2007. — V. 8 — P. 188; Chou S., Marousek G. I. // J. Virol. — 2008. — V. 82. — P. 246.
13. Galmarini C. M., Popowycz F., Joseph B. // Curr. Med. Chem. — 2008. — V. 15. — P. 1072; Mathe C., Gosselin G. // Antiviral Res. — 2006. — V. 71. — P. 276; Biron K. K. // Antiviral Res. — 2006. — V. 71. — P. 154.
14. Mikhailopulo I. A., Zinchenko A. I., Bokut S. B. et al. // Biotechnol. Letters. — 1992. — V. 14. — P. 885.
15. Mikhailopulo I. A., Kazimierczuk Z., Zinchenko A. I. et al. // Nucl. Acids Symp. Ser. — 1994. — N 31. — P. 83.
16. Mikhailopulo I. A., Kazimierczuk Z., Zinchenko A. I. et al. // Nucleosides Nucleotides. — 1995. — V. 14. — P. 477.
17. Bentancor L., Trelles J. A., Nybile M. et al. // J. Mol. Cat. B: Enzymatic. — 2004. — V. 29. — P. 3.
18. Konstantinova I. D., Selezneva O. M., Fate-ev I. V. et al. // Synthesis. — 2013. — V. 45. — P. 272.
19. (a) MacDonald D. L. // J. Org. Chem. — 1962. — V. 27. — P. 1107; (b) MacDonald D- L. // Carbohyd. Res. — 1966. — V. 3. — P. 117; (c) MacDonald D. L. // Carbohyd. Res. — 1968. — V. 6. — P. 376; (c) Mendicino J., Hanna R. // J. Biol. Chem. — 1970. — V. 245. — P. 6113; (d) Chittenden G- J- F. // Carbohyd. Res. — 1972. — V. 25. — P. 35.
20. Konstantinova I. D., Antonov K. V., Fateev I. V. et al. // Synthesis. — 2011. — V. 10. — P. 1555.
21. Wright R. S., Khorana H. G. // J. Am. Chem. Soc. — 1958. — V. 80. — P. 1994.
22. Averett D. R., Koszalka G. W., Fyfe J. A. // Antimicrob. Agents Chemother. — 1991. — V. 35. — P. 851.
23. Fateev I. V., Antonov K. V., Konstantinova I. D. et al. The Chemo-Enzymatic Synthesis of
Clofarabine and Related Nucleosides. The Role of Electronic and Stereochemical Factors of Substrates in Reactions Catalyzed by E. coli Nucleoside Phosphorylases. Federation of European Biochemical Societies. CONGRESS 2013 «Mechanisms in Biology», July 6-11, 2013. — Sainct Petersburg, Russia. Abstracts.
24. Roivainen J., Elizarova T., Lapinjoki S. et al. // Nucleosides, Nucleotides & Nucl. — Acids. — 2007. — V. 26. — P. 90б.
2б. a) Himmerlsbach F., Shulz B. S., Trichtinger T. et al. // Tetrahedron. — 1984. — V. 40. — P. б9; b) Zhou J., Tsai J.-Y., Bouhadir K., SlevinP.B. // Synthetic Commun. — 1999. — V. 29.— P. 3003.
26. Kulikovskiy A. I., Esipov R. S., Miroshni-kov A. I., Mikhailopulo I. A. An Enzymatic Synthesis of Nucleosides of N2-Acetyl-O6-[2-(4-nitrophenyl)ethyl]guanine // Collection Symp. Series. — 2011. — V. 12. — P. 397.
27. Elion G. B. // Cancer Res. — 1969. — V. 29. — P. 2448.
28. (a) Seela F., Pujari S. // Bioconjugate Chem. — 2010. — V. 21. — P. 1629; (b) Jiang D., Seela F. // J. Am. Chem. Soc. — 2010. — V. 132. — P. 4016.
29. Seela F., Jiang D., Xu K. // Org. Biomol. Chem. — 2009. — V. 7. — P. 3463, and references cited therein.
30. (a) Pokharel S., Jayalath P., Maydanovych O. et al. // J. Am. Chem. Soc. — 2009. — V. 131. — P. 11882; (b) Veliz E. A., Easterwood L. M., Beal P. A. // Ibid. —
2003. — V. 12б. — P. 10867; (c) Wier-zchowski J., Ogiela M., Iwanska B., Shugar D. // Anal. Chim. Acta. — 2002. — V. 472. — P. 63.
31. (a) Elliot R. D., Montgomery J. A. // J. Med. Chem. — 1976. — V. 19. — P. 1186; (b) Kazimierczuk Z., Cottam H. B., Revankar G. R., Robins R.K. // J. Am. Chem. Soc. — 1984. — V. 106. — P. 6379, and references therein.
32. (a) Doskocil J., Holy A. // Coll. Czech. Chem. Communs. — 1977. — V. 42. — P. 370 (see also the works cited therein); (b) Wierzchowski J., Wielgus-Kutrowska B., Shugar D. // Biochim. Biophys. Acta. — 1996. — V. 1290. — P. 9; (c) Votruba I., Holy A., Dvorakova H. et al. // Coll. Czech. Chem. Communs. — 1994. — V. б9. — P. 2303.
33. (a) Koellner G., Luic M., Shugar D. et al. // J. Mol. Biol. — 1998. — V. 280. — P. 1б3; (b) Koellner G., Bzowska A., Wielgus-Kut-rowska B. et al. // J. Mol. Biol. — 2002. — V. 31б. — P. Зб1; (c) Bennett E. M., Li C., Allan P. W. et al. // J. Biol. Chem. — 2003. — V. 278. — P. 47110; (d) Zhang Y., Parker W. B., Sorscher E. J., Ealick S. E. // Curr. Topics Med. Chem. — 200б. — V. б. — P. 12б9; (e)
Mikleusevic G., Stefanic Z., Narczyk M. et al. // Biochimie. — 2011. — V. 93. — P. 1610.
34. Fateev I. V., Esipov R. S., Antonov K. V. et al. The Role of the Ser90 Residue of the Catalytic Site of E. coli PNP in the Synthesis
of Purine and 8-aza-7-deazapurine Nucleosides. Federation of European Biochemical Societies. CONGRESS 2013 «Mechanisms in Biology»; July 6-11, 2013. — Sainct Petersburg, Russia. Abstracts.
ДЕЯКІ НОВІ ДАНІ БІОТЕХНОЛОГІЇ БІОЛОГІЧНО АКТИВНИХ НУКЛЕОЗИДІВ
Мuхайлопуло І. А. М^ошшков А. І.
1Інститут біоорганічної хімії Національної академії наук Білорусі, Мінськ, Білорусь.
E-mail: [email protected]
2Інститут біоорганічної хімії ім. академіків М. М. Шемякіна
і Ю. А. Овчинникова Російської академії наук, Москва, Росія
НЕКОТОРЫЕ НОВЫЕ ДАННЫЕ БИОТЕХНОЛОГИИ БИОЛОГИЧЕСКИ АКТИВНЫХ НУКЛЕОЗИДОВ
Михайлопуло И. А. Мирошников А. И.
1Институт биоорганической химии Национальной академии наук Беларуси, Минск, Беларусь
E-mail: [email protected]
2 Институт биоорганической химии им. академиков М. М. Шемякина и Ю. А. Овчинникова Российской академии наук, Москва, Россия
E-mail: [email protected]
E-mail: [email protected]
Обговорено нещодавні відкриття в галузі біотехнології біологічно активних нуклеози-дів, до яких належить низка лікарських препаратів проти лейкемії (кладрибін, флудара-бін, неларабін, клофарабін). Докладно розглянуто: нову стратегію каскадної «one-pot» трансформації D-пентоз у нуклеозиди, що ґрунтується на розширенні та поглибленні знань про механізм функціонування рибокіна-
зи, фосфопентомутази, уридин-, тимідин- і пуриннуклеозидфосфорилаз, а також роль різних чинників (структурних, електронних, стереохімічних) у формуванні глікозидного зв'язку; сучасні методи хемоензиматичного синтезу нуклеозидів; реакцію трансглікози-лювання, що каталізується нуклеозидфосфо-рилазами, за якої донорами карбогідратних залишків виступають природні та модифіковані за цукрами нуклеозиди, а акцепторами — гетероциклічні основи.
Ключові слова: нуклеозиди, біоміметичний синтез, хемоензиматичний синтез, ензими метаболізму нуклеїнових кислот.
Обсуждаются недавние открытия в области биотехнологии биологически активных нук-леозидов, к которым относится ряд лекарственных препаратов против лейкемии (кладри-бин, флударабин, неларабин, клофарабин). Подробно рассмотрены: новая стратегия каскадной «one-pot» трансформации D-пентоз в нуклеозиды, основанная на расширении и углублении знаний о механизме функционирования рибокиназы, фосфопентомутазы, ури-дин-, тимидин- и пуриннуклеозидфосфорилаз, а также роль различных факторов (структурных, электронных, стереохимических) в формировании гликозидной связи; современные методы хемоэнзиматического синтеза нуклео-зидов; катализируемая нуклеозидфосфорила-зами реакция трансгликозилирования, при которой в качестве доноров карбогидратных остатков выступают природные и модифицированные по сахарам нуклеозиды, а акцепторов — гетероциклические основания.
Ключевые слова: нуклеозиды, биомиметичес-кий синтез, хемоэнзиматический синтез, энзимы метаболизма нуклеиновых кислот.
