Synthesis of potential inhibitors of a novel flavin-dependent thymidylate synthase in Mycobacterium tuberculosis

SummaryHuman tuberculosis is a contagious infectious disease and is regarded as a major threat to global health. It is estimated that one third of the world population is infected with Mycobacterium tuberculos is, the causative agent of this disease. The emergence of drug-resistant strains of M. tuberculosis in synergy with the rise of the HIV/AIDS epi demic has further severely exacerbated effective TB-control. Drug-suscep tible TB is treated with so-called first-line drugs isoniazid, rifampici n, pyrazinamide, and ethambutol over a period of six months. Multi-drug resistant TB is defined as resistance to isoniazid and rifampicin and it s effective treatment requires the inclusion of second-line drugs which are often less efficacious, but more toxic and considerably more costly. The prolonged course of treatment of 18 to 24 months and their severe s ide effects led to a high proportion of non-compliance patients, thereby giving rise to the selection of extensively drug-resistant (XDR) strain s. XDR-TB strains are defined as being resistant to isoniazid and rifamp icin, to all the fluoroquinolones and to at least one of the injectable drugs (capreomycin, kanamycin, amikacin). XDR-TB is very difficult to tr eat and comprises the use of other, less common second-line drugs with e ven more severe adverse effects. Consequently, there is an urgent demand for drugs acting on novel drug targets to avoid cross-resistance with o ther drugs. Preferentially, these drugs should also allow shortening of treatment courses.Until about a decade ago, the only known pathway f or the de novo synthesis of 2 -deoxythymidine-5 -monophosphate (dTMP), o ne of the four essential building blocks of DNA, was by thymidylate synt hase (TS or ThyA). This enzyme catalyzes the reductive methylation of 2 -deoxyuridine-5 -monophosphate (dUMP) through the use of methylenetetrah ydrofolate (CH2H4folate) as both, carbon source and reductant with forma tion of dihydrofolate. Given the importance of the catalytic activity of this enzyme for replication, the inhibition of ThyA proved to be a vali dated strategy for the development of novel drugs that inhibit cellular proliferation.Genomic studies have revealed that several microorgani sms lack the genes encoding for ThyA which led to the discovery of the T hyX protein, a novel class of flavin-dependent thymidylate synthases (FD TS). ThyX uses CH2H4folate only as a one-carbon donor, whereas reduced f lavin adenine dinucleotide (FADH2) fulfills the role as a reductant. A n umber of human pathogens, including M. tuberculosis, carry the genes for both thymidylate synthases, but to date, the precise metabolic role of each of these two enzymes remains to be elucidated. However, thyX was id entified in a screen for essential genes in M. tuberculosis. It has been postulated that either thyX or thyA might be preferentially activated u nder different growth conditions.The crystal structure of Mtb ThyX a nd mechanistic studies demonstrated that ThyX shares no structural and sequence homology with ThyAand differs considerably in view of its bioche mical reaction mechanism. This should facilitate the design of selective inhibitors of either of the two enzymes as it does not have to rely on small structural differences between ThyX and ThyA proteins.Herein, we describe our pioneering efforts towards the design, synthesis, and bi ological evaluation of novel dUMP analogues as selective inhibitors of m ycobacterial FDTS. The results obtained in a recent study prompted us to evaluate 5-alkynyl dUMP and araUMP analogues in the first part of this thesis. Starting from suitably protected or unprotected 5-iodo nucleosid es, a number of long-chain acetylenic moieties have been introduced via palladium-catalyzed Sonogashira cross-coupling as key step. Further synt hetic modifications, followed by phosphorylation yielded the desired der ivatives Ia-b, IIa-b, IIIa-d, and IVa-b. Biological evaluation has shown that introduction of a long-chain acetylenic moiety at C-5 of dUMP is a promising strategy for the development of selective mycobacterial ThyX- inhibitors with IC50 values ranging from 0.9 to 28 µM. On the other hand, araUMP analogues Ib and IIb completely lacked inhibitory activity against ThyX. Moreover, none of the synthesized derivatives exhibited s ignificant levels of ThyA inhibition, thereby highlighting their selecti vity for mycobacterial ThyX.The second project of this thesis deals with the synthesis and biological evaluation of 6-aryl and alkenyl dUMP analogues Va-f and VIa-c. These compounds were accessible via regioselec tive lithiation, followed by quenching with a suitable electrophile. To gain access to 6-aryl dUMP derivatives Va-c and 6-styryl derivative Vd, a base-free Suzuki-Liebeskind cross-coupling methodology has been succes sfully employed. Unfortunately, the desired vinylic congeners VIa-c coul d not be obtained due to problems in the phosphorylation step. None of t he synthesized monophosphates Va-f displayed any significant level of ei ther ThyX or ThyA inhibitory activity. The last part of this thesis describes the synthesis and biological eval uation of 6-aza dUMP analogues VIIa-d and VIIIa-c as mechanism-based inh ibitors of mycobacterial ThyX. Key step in the synthetic pathway is a Vo rbrüggen-type glycosylation reaction of a silylated 6-azauracil-derived nucleobase with Hoffer s chlorosugar. The glycosylation reaction proceed ed with only moderate stereoselectivity in favour of the ß-ano mer. However, both isomers could be easily separated by recrystallizatio n or silica gel column chromatography. Final deprotection and Yoshikawa- phosphorylation afforded the desired 6-aza dUMP analogues. The results f rom the inhibition assays revealed that the parent nucleotide 6-aza dUMP VIIa exhibited weak levels of inhibitory activity (34% inhibition at 50 µM). In order to increase activity, a number of alkyl and aryl gro ups (methyl, i-propyl, n-octyl, and 4-fluorophenyl) were introduced at p osition 5. These compounds were totally devoid of ThyX inhibitory activi ty. On the other hand, compound VIIIb, the 6-aza congener of compound II Ia, displayed, albeit weak, ThyX inhibitory activity (40% inhibition at 50 µM). The hydrated form of VIIIb, VIIIc, however, completely lack ed any mycobacterial ThyX-inhibition. It highlights the particular role of a propargylamide side chain to confer inhibitory activity against myc obacterial ThyX. Notably, all synthesized 6-aza dUMP analogues did not e xhibit any significant level of mycobacterial ThyA inhibitory activity.status: publishe