New Flexible Analogues of 8-Aza-7-deazapurine Nucleosides as Potential Antibacterial Agents

A variety of ribo-, 2′-deoxyribo-, and 5′-norcarbocyclic derivatives of the 8-aza-7-deazahypoxanthine fleximer scaffolds were designed, synthesized, and screened for antibacterial activity. Both chemical and chemoenzymatic methods of synthesis for the 8-aza-7-deazainosine fleximers were compared. In the case of the 8-aza-7-deazahypoxanthine fleximer, the transglycosylation reaction proceeded with the formation of side products. In the case of the protected fleximer base, 1-(4-benzyloxypyrimidin-5-yl)pyrazole, the reaction proceeded selectively with formation of only one product. However, both synthetic routes to realize the fleximer ribonucleoside (3) worked with equal efficiency. The new compounds, as well as some 8-aza-7-deazapurine nucleosides synthesized previously, were studied against Gram-positive and Gram-negative bacteria and M. tuberculosis. It was shown that 1-(β-D-ribofuranosyl)-4-(2-aminopyridin-3-yl)pyrazole (19) and 1-(2′,3′,4′-trihydroxycyclopent-1′-yl)-4-(pyrimidin-4(3H)-on-5-yl)pyrazole (9) were able to inhibit the growth of M. smegmatis mc2 155 by 99% at concentrations (MIC99) of 50 and 13 µg/mL, respectively. Antimycobacterial activities were revealed for 4-(4-aminopyridin-3-yl)-1H-pyrazol (10) and 1-(4′-hydroxy-2′-cyclopenten-1′-yl)-4-(4-benzyloxypyrimidin-5-yl)pyrazole (6). At concentrations (MIC99) of 40 and 20 µg/mL, respectively, the compounds resulted in 99% inhibition of M. tuberculosis growth.


Introduction
To date, great success has been achieved in the treatment of infectious diseases through antibiotic therapies [1].Despite this, the rapid development of drug-resistant strains [2] of pathogenic microorganisms creates a need for the new and more effective antibiotics.The problem of drug resistance is more acute than ever in the case of tuberculosis, as it is one of the factors that reduces the effectiveness of many treatments [3,4].
In 2021, 10.6 million people were diagnosed with tuberculosis worldwide, and a total of 1.6 million deaths (including 187,000 people living with HIV) [4] have been attributed to it.Tuberculosis is the 13th leading cause of death globally and is the second leading cause of infectious death after COVID-19 (ahead of HIV/AIDS) [5].The high incidence of tuberculosis in the global population is a serious threat to society [4].In addition, over the past few years, the world has faced unprecedented challenges associated with the emergence of new pathogens such as SARS-CoV-2 [6], among other emerging and reemerging infectious diseases.This underscores the critical need for new and more effective therapeutic drugs.
During decades of searching for molecules active against new and emerging bacterial infections, drugs belonging to various chemical classes of compounds have been discovered.Unfortunately, again, the development of resistance has limited their utility [7].One potential solution is to explore new viral and bacterial targets, as well as classes of drugs not previously considered [8].Another is the development of compounds that exhibit broadspectrum activity, or the use of combination therapies, thereby reducing susceptibility to the development of resistance [9].
In terms of the search for new compounds with broad spectrum activity, a very promising group of purine nucleoside analogues are the fleximers [30,31].It has been shown that these unique compounds, where the purine base is split into its two separate components, have exhibited a wide spectrum of biological activity [30].The additional flexibility of such nucleoside analogues is a significant advantage and in the case of resistant pathogens makes it possible for a potential drug to "adjust" to the binding site containing point mutations and retain inhibitory activity [30].

Synthesis of the Target Compounds
We have previously compared the chemical and chemoenzymatic methods for the synthesis of fleximer adenosine analogues [32].It was shown that fleximer analogues of 8-aza-7-deazaadenine are useful substrates for purine nucleoside phosphorylase E. coli (PNP E. coli) in the synthesis of corresponding modified nucleosides.Fleximer 8-aza-7deazaadenine ribonucleosides can be obtained through both methods with comparable effectiveness.In contrast, the chemical synthesis of 2 -deoxynucleoside analogues resulted in mixtures of αand β-anomers, so an additional purification step is needed.Thus, enzymatic transglycosylation proved to be the preferred route for the 2 -deoxyribonucleosides [32].
In an effort to explore these routes for other fleximer aza/deaza purines, we have compared the chemical and chemo-enzymatic methods of 8-aza-7-deazainosine fleximer synthesis.In that regard, a series of experiments was carried out to determine the substrate specificity of E. coli purine nucleoside phosphorylase to heterocyclic bases 1a and 1b.In the case of 1a, the transglycosylation reaction (Scheme 1) proceeded with the formation of side products for both the riboside and 2 -deoxyriboside (see Supplementary file).This occurs due to the availability of several glycosylation sites.Base 1a contains four nitrogen atoms and lacks functional group protection.In the case of the fleximer base 1b, the reaction proceeds selectively with the formation of just one product, as observed by chromatography (see Supplementary file).The presence of benzyl protecting groups creates steric hindrance for the enzyme for glycosylation of the pyrimidine, thereby ensuring the regioselectivity of the reaction.Thus, we chose to use the chemoenzymatic glycosylation method for the synthesis of the nucleoside analogues of the 8-aza-7-deazahypoxanthine fleximer using the protected flex-base 1b.
8-aza-7-deazaadenine are useful substrates for purine nucleoside phosphorylase E. coli (PNP E. coli) in the synthesis of corresponding modified nucleosides.Fleximer 8-aza-7-deazaadenine ribonucleosides can be obtained through both methods with comparable effectiveness.In contrast, the chemical synthesis of 2′-deoxynucleoside analogues resulted in mixtures of α-and β-anomers, so an additional purification step is needed.Thus, enzymatic transglycosylation proved to be the preferred route for the 2′-deoxyribonucleosides [32].
In an effort to explore these routes for other fleximer aza/deaza purines, we have compared the chemical and chemo-enzymatic methods of 8-aza-7-deazainosine fleximer synthesis.In that regard, a series of experiments was carried out to determine the substrate specificity of E. coli purine nucleoside phosphorylase to heterocyclic bases 1a and 1b.In the case of 1a, the transglycosylation reaction (Scheme 1) proceeded with the formation of side products for both the riboside and 2′-deoxyriboside (see Supplementary file).This occurs due to the availability of several glycosylation sites.Base 1a contains four nitrogen atoms and lacks functional group protection.In the case of the fleximer base 1b, the reaction proceeds selectively with the formation of just one product, as observed by chromatography (see Supplementary file).The presence of benzyl protecting groups creates steric hindrance for the enzyme for glycosylation of the pyrimidine, thereby ensuring the regioselectivity of the reaction.Thus, we chose to use the chemoenzymatic glycosylation method for the synthesis of the nucleoside analogues of the 8-aza-7-deazahypoxanthine fleximer using the protected flex-base 1b.Protype reactions were carried out to determine the best donor of ribose and 2′-deoxyribose for base 1b.In the first series of experiments, uridine, guanine, adenine, and inosine were chosen as the potential carbohydrate donors.In the second series, 2′-deoxyuridine, 2′-deoxyguanosine, 2′-deoxyadenosine, and 2′-deoxyinosine were chosen.The reactions were carried out in a phosphate buffer pH 7.0 at 50 °C.Interestingly, there was no difference in glycosylation efficiency when the natural purine nucleosides inosine, guanosine, and adenosine were used as donors.However, for uridine the formation of the target nucleoside proceeded faster (conversion 97% after 1 h).As a result, uridine and 2′-deoxyuridine, respectively, were chosen as the ribose/deoxyribose donors.
Nucleoside analogues 2 and 3 were then synthesized chemoenzymatically using uridine phosphorylase (UP E. coli), PNP E. coli, and 1-(4-benzyloxypyrimidin-5-yl)pyrazole 1b as the base.Removal of the benzyl protective group by a hydrogenation reaction on a palladium catalyst (Scheme 1) afforded fleximer nucleosides 4 and 5 in 25-26% yield.Protype reactions were carried out to determine the best donor of ribose and 2deoxyribose for base 1b.In the first series of experiments, uridine, guanine, adenine, and inosine were chosen as the potential carbohydrate donors.In the second series, 2deoxyuridine, 2 -deoxyguanosine, 2 -deoxyadenosine, and 2 -deoxyinosine were chosen.The reactions were carried out in a phosphate buffer pH 7.0 at 50 • C. Interestingly, there was no difference in glycosylation efficiency when the natural purine nucleosides inosine, guanosine, and adenosine were used as donors.However, for uridine the formation of the target nucleoside proceeded faster (conversion 97% after 1 h).As a result, uridine and 2 -deoxyuridine, respectively, were chosen as the ribose/deoxyribose donors.
The 5 -norcarbocyclic analogue of 8-aza-7-deazainosine 9 was synthesized by oxidation of the double bond of compound 6 with osmium tetraoxide in dioxane-water (10:1), and subsequent removal of the benzyl protecting group of 8 with palladium on carbon under a hydrogen atmosphere (Scheme 2).Product 9 was obtained in a 38% yield in two steps.An attempt to directly remove the benzyl group from compound 6 through hydrogenation produced compound 7 in a 63% yield.
product with 75% yield.Removal of the acetyl and benzyl protecting groups led to the target fleximer analogue 5 (18% yield).
The 5′-norcarbocyclic analogue of 8-aza-7-deazainosine 9 was synthesized by oxidation of the double bond of compound 6 with osmium tetraoxide in dioxane-water (10:1), and subsequent removal of the benzyl protecting group of 8 with palladium on carbon under a hydrogen atmosphere (Scheme 2).Product 9 was obtained in a 38% yield in two steps.An attempt to directly remove the benzyl group from compound 6 through hydrogenation produced compound 7 in a 63% yield.

Discussion
In this work, we continue to study the importance and biological scope of these flexible nucleoside analogues known as fleximers.Earlier [32], we have shown that fleximer analogues of 8-aza-7-deazaadenine are useful substrates for purine nucleoside phosphorylase E. coli (PNP E. coli).It also provided the opportunity to further investigate and compare the two routes of synthesis using chemical methods and enzymatic transglycosylation, in this case for the 8-aza-7-deazainosine fleximer scaffold.It was found during the experiments, through the determination of the substrate specificity of E. coli PNP to heterocyclic bases 1a and 1b, that the benzyl protective group is required for ensuring the regioselectivity of the reaction.As a result, protected flex-base 1b was chosen for enzymatic glycosylation as well as for the chemical synthesis of ribonucleoside 3 using the classical Vorbruggen procedure.In the case of the synthesis of the 8-aza-7-deazainosine fleximer 3 starting from 1-(4-benzyloxypyrimidin-5-yl)pyrazole 1b, both methods worked with equal efficiency.
In addition to fleximers with traditional ribose or 2 -deoxyribose, we synthesized derivatives of 8-aza-7-deazahypoxanthine bearing a 5 -norcarbocyclic fragment as the sugar moiety.5 -Norcarbocyclic nucleoside analogues [39] have a number of advantages.One of them is the absence of 5 -CH 2 group, which prevents phosphorylation thereby resulting in decreasing cytotoxicity.At the same time, biological properties not associated with phosphorylation are retained.This feature also helps to exclude classical nucleoside polymerases inhibition [40], but at the same time several 5 -norcarbocyclic nucleoside analogues have been shown to act as nonnucleoside inhibitors of viral RNA polymerases [41] or reverse transcriptase [42,43].
Antibacterial screening has shown that flexible analogues of 8-aza-7-deazapurine nucleosides, compounds 9 and 19, inhibited the growth of M. smegmatis mc2 155 in 13 and 50 µg/mL, respectively.Analogues 6 and 10 caused 99% inhibition of M. tuberculosis H37Rv, at concentrations of 20 and 40 µg/mL, respectively.It was not surprising that different compounds were active against M. smegmatis and M. tuberculosis, since they are not closely related groups of mycobacteria.M. tuberculosis is slow-growing mycobacteria, and M. smegmatis is fast-growing one, so these two types of microorganisms may have differential sensitivities to the various fleximer analogues.Nevertheless, the data obtained will help guide for further optimization of new fleximer scaffolds in the search for new and more effective antibacterial agents.
Liquid chromatography mass spectrometry was performed using an Agilent 6210 TOF LC-MS system (Agilent Technologies, Santa Clara, CA, USA).
High-resolution mass spectra (HRMS) were obtained on a Bruker Daltonics micrOTOF II instrument using electrospray ionization (ESI).The measurements were acquired in a positive ion mode with the following parameters: interface capillary voltage-4500 V; mass range from m/z 50 to 3000; internal calibration (ESI Tuning Mix, Agilent); nebulizer pressure-0.3Bar; flow rate-3 µL/min; dry gas nitrogen (4.0 L/min); interface temperature was set at 180 • C. Syringe injection was used.

General Procedure for the Enzymatic Synthesis of Fleximer Nucleosides
The flex-base 1b and 2 -deoxyuridine/uridine at ratios of 1:2 were dissolved in 70 mL 10 mM potassium phosphate buffer (pH 7.0) at 40-50 • C. The enzymes 3.2 e.u./mL PNP and 4.0 e.u./mL UP E. coli were added.The reaction mixtures were incubated at 50    (20 mL).The reaction mixture was kept at 36 • C for 3 h.Purification with preparative chromatography on silica gel glass plate in chloroform/methanol (95:5) system, gave the riboside (3) as a white powder (71 mg) with 95% yield.