Lipophilic Nucleoside Triphosphate Prodrugs of Anti‐HIV Active Nucleoside Analogs as Potential Antiviral Compounds

Abstract Nucleoside analogs require three phosphorylation steps catalyzed by cellular kinases to give their triphosphorylated metabolites. Herein, the synthesis of two types of triphosphate prodrugs of different nucleoside analogs is disclosed. Triphosphates comprising: i) a γ‐phosphate or γ‐phosphonate bearing a bioreversible acyloxybenzyl group and a long alkyl group and ii) γ‐dialkyl phosphate/phosphonate modified nucleoside triphosphate analogs. Almost selective conversion of the former TriPPPro‐compounds into the corresponding γ‐alkylated nucleoside triphosphate derivatives is demonstrated in CEM/0 cell extracts that proved to be stable toward further hydrolysis. The latter γ‐dialkylated triphosphate derivatives lead to the slow formation of the corresponding NDPs. Both types of TriPPPro‐compounds are highly potent in wild‐type CEM/0 cells and more importantly, they exhibit even better activities against HIV‐2 replication in CEM/TK− cell cultures. A finding of major importance is that, in primer extension assays, γ‐phosphate‐modified‐NTPs, γ‐mono‐alkylated‐triphosphates, and NDPs prove to be substrates for HIV‐RT but not for cellular DNA‐polymerases α,γ.

Furthermore, a potential new generation of dialkylphosphate-modified-d4TDPs 8 (CEM, t 1/2 >10 h) and -dialkylphosphonate-modified-d4TDPs 9 (CEM, t 1/2 >15 h) was discovered, that comprised two non-cleavable moieties at the -phosphate group or -phosphonate group, respectively (Scheme 2). [61]It is worth noting that these TriPPPro-compounds 8,9 showed very good antiviral activity (EC 50 : 0.036 μm) in infected wild-type CEM/0 cells that was completely retained (EC 50 : 0.0050 μm, 10 000-fold more active as d4T) in HIV-infected CEM/TK − cells.More interestingly, it was demonstrated that in addition to d4TTP also d4TDP was accepted by HIV-RT as a substrate. [61]uided by the previous results from TriPPPro-d4TTPs 2-9, we synthesized a series of TriPPPro-prodrugs 10 bearing different nucleoside analogs.The chemical formulae of the nucleoside analogs used in our studies are shown in Figure 1.All TriPPProcompounds 10 were studied with regard to their chemical and biological stability.In addition, a few -dialkylphosphate-modified-NTPs 10 and -mono-modified-NTPs 20 were prepared as well for the primer extension experiments and to further study the hydrolysis properties and the delivery mechanism of TriPPProcompounds 10.
the final coupling, TriPPPro-compounds 10 were successfully obtained using method A (Scheme 3) in yields of 21-81%.In addition, -(-cyanoethyl;alkyl-C18 or C4)-NTPs 19 were synthesized as well using the same route.Subsequently, the -cyanoethyl moiety was cleaved under basic conditions to form the corresponding -alkyl-NTPs 20 in yields between 17% and 43% (variation in the nucleoside part; Scheme 4).However, low yields of TriPPPro-compounds 10, such as 10hv, were obtained without good reasons.Thus, method B was used for the synthesis of TriPPPro-compounds 10bv-bx,10ev-ez, and 10hv that gave good yields as shown in Scheme 3. First H-phosphonate 14 was prepared from 9-fluorenylmethanol 13 and diphenyl hydrogen phosphonate (DPP).Next, compound 14 was reacted with NCS to give the phosphorochloridate.Subsequent phosphorylation of NMPs yielded the bis(Fm)-protected diphosphate in compounds 15.Compounds 15 were hydrolyzed to form -Fm-NDPs 16 with NEt 3 (10 min) in CH 3 CN:THF (1:1).Intermediates 16 were isolated by reversed-phase (rp) column chromatography, followed by a deprotection step to form NDPs. The final coupling reaction was accomplished by a stepwise reaction sequence using H-phosph(i)onates 11 with NCS, followed by the addition of NDPs to afford TriPPPro-compounds 10.Notably, the remaining NDP can be recycled using this strategy; thus, a more efficient conversion of the parent nucleoside to the TriPPPro-compounds was achieved.

Chemical Stability and Enzymatic Activation of TriPPPro-Compounds 10 and 20
The hydrolysis properties of TriPPPro-compounds 10 and 20 were evaluated in phosphate buffer saline (PBS, pH 7.3), pig liver esterase (PLE) or CEM/0 cell extracts.In both cases, hydrolysis products were analyzed by means of analytical RP18-HPLC and the hydrolysis half-lives (Table 1, t 1/2 ) of TriPPPro-compounds 10 and 20 were calculated after complete consumption of the starting materials.
It has been reported that abacavir acts as a prodrug for the toxic carbovir (CBV). [67,68]However, after intracellular monophosphorylation ABCMP is converted by adenosine monophosphate deaminase (AMPDA) into carbovir-monophosphate (CB-VMP), which is then processed into its triphosphate. [67,68]Thus, the active form of abacavir is the HIV-RT inhibitor carbovirtriphosphate (CBVTP).However, for the TriPPPro-compounds disclosed here, it was shown that the ABCDP is formed that cannot be converted into carbovir-diphosphate (CBVDP).This suggests that the active form in our case is indeed ABCTP or ABCDP itself.
Most TriPPPro-compounds 20 bearing a long lipophilic alkyl chain (C18) attached to the -phosphate moiety were endowed with moderate antiviral activity in the cell assay using CEM/TK − cells.It seems that TriPPPro-compounds 20 were at least in part able to cross the cell membrane and deliver phosphorylated metabolites presumably NDPs, or are used as substrates in their mono-alkylated form.Nevertheless, some TriPPProcompounds 10 comprising non-bioreversible and hydrolytically stable alkyl groups, such as 10hv (EC 50 = 0.0048 μm/HIV-2), had much higher (33-fold) activity as compared to the corresponding TriPPPro-compounds 20h (EC 50 = 0.16 μm/HIV-2), proving the advantage of prodrug strategy.Notably, TriPPPro-compounds 10 and 20 did not show a significant increase in cytotoxicity as compared to their parent nucleosides.

Primer Extension Assays
We examined TriPPPro-prodrugs 10 and 20 in primer extension assays and investigated their suitability to act as substrates for the HIV-RT as compared to two different human DNA poly-merases ,.In these primer extension assays, the four canonical NTPs were added to the polymerases (positive control (+ lane)) or were added in the absence of the polymerase (negative control (−lane)).TTP, dCTP, and dATP were used as the reference compounds because they were substrates for HIV-RT and DNA polymerases ,.
Human DNA polymerases  (Figure 7) and  (Figure 8) were also tested to ensure n+1 band incorporation.As compared to HIV-RT, no incorporation was detected in primer extension assays using human DNA polymerases  or  for TriPPProcompounds 10 and 20.Thus, these experiments proved that the double alkylated TriPPPro-prodrugs 10 and single alkylated compounds 20 were not substrates for human DNA polymerases  or .

Conclusion
In summary, the synthesis of a series of TriPPPro-compounds 10 and 20 bearing different nucleoside analogs is described here, demonstrating the applicability of the TriPPPro-strategy.TriPPPro-NTPs 10 as well as -mono-masked triphosphates 20 were prepared by using the new H-phosphonate route and Hphosphinate route with modest to very good yields.
The approach also points to interesting details on the activation of uracil-bearing nucleoside analogs.This was shown for FddClU 1d, FddU 1e which were active in the wild-type cells but inactive in the TK-deficient cells.The conversion of these nucleosides into their triphosphate prodrug form restored the antiviral activity pointing to at least a contribution of thymidine-kinase to the metabolic phosphorylation of uracil comprising nucleosides.Interestingly, the approach also converted BVDU, an anti-VZV, and HSV-1 (herpes viruses) active nucleoside analog, into a powerful anti-HIV active compound and thus broadens the antiviral spectrum of the parent 1g.
In conclusion, it was convincingly shown that this TriPPProapproach provides high potential for further antiviral chemotherapies.Highly active TriPPPro-prodrugs may be developed for the treatment of infections by not only HIV but also for SARS-CoV-2 and other RNA viruses in the future.

Figure 1 .
Figure 1.Chemical structures of nucleoside analogs used in this study.

Figure 4 .
Figure 4. Stack plot of the HPLC chromatograms of 10bx after incubation in CEM/0 cell extracts.

Figure 5 .
Figure 5. Stack plot of the HPLC chromatograms of 10by after incubation in CEM/0 cell extracts.