Discovery of a nuclease-resistant, non-natural dinucleotide that inhibits HIV-1 integrase

Article abstract of DOI:10.1016/S0960-894X(01)00231-1

Integration of HIV viral DNA into human chromosomal DNA catalyzed by HIV integrase is essential for the replication of HIV. Discovery of novel inhibitors of HIV integrase is of considerable significance in approaches to the development of therapeutic agents against AIDS. We have synthesized a new dinucleotide 1 with an internucleotide phosphate bond that is unusually resistant to exonucleases. This compound exhibits potent anti-HIV-1 integrase activity.

Full text of DOI:10.1016/S0960-894X(01)00231-1

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1433–1435
Discovery of a Nuclease-Resistant, Non-natural
Dinucleotide that Inhibits HIV-1 Integrase
Michael Taktakishvili,a Nouri Neamati,b,y Yves Pommierb and Vasu Naira,*
aDepartment of Chemistry, The University of Iowa, Iowa City, IA 52242, USA
bLaboratory of Molecular Pharmacology, National Cancer Institute, NIH, Bethesda, MD 20892, USA
Received 5 January 2001; accepted 3 April 2001
Abstract—Integration of HIV viral DNA into human chromosomal DNA catalyzed by HIV integrase is essential for the replication
of HIV. Discovery of novel inhibitors of HIV integrase is of considerable significance in approaches to the development of ther-
apeutic agents against AIDS. We have synthesized a new dinucleotide 1 with an internucleotide phosphate bond that is unusually
resistant to exonucleases. This compound exhibits potent anti-HIV-1integrase activity. # 2001Elsevier Science Ltd. All rights
reserved.
The retroviral enzyme, HIV-1integrase, incorporates
HIV double helical DNA (the replicative form of the
HIV genome synthesized by HIV reverse transcriptase)
into host chromosomal DNA. This viral enzyme first
catalyzes the excision of two terminal nucleotides of the
30-end of each strand of viral DNA (endonuclease activ-
ity, 30-processing) leaving recessed ends that terminate
with xxCA-OH (Fig. 1). In the next steps (strand trans-
fer, integration) nucleophilic attack of the terminal 30-
OH of the tailored HIV DNA on a specific internucleo-
tide phosphodiester functionality results in cleavage of
host DNA and subsequent integration of the tailored
HIV DNA into host DNA.1ꢀ4 The integration process is
essential for the replication for HIV and there is no
functional equivalent of HIV integrase in human cells.
Figure 1. Tailoring of HIV double helical DNA prior to integration
into host DNA.
integrase activity.5 We have designed and synthesized a
conceptually new dinucleotide with con-
1
a
formationally unusual internucleotide phosphate bond
and overall conformation that exhibits complete resis-
tance to mammalian 30- and 50-exonucleases and that
has potent anti-HIV-1integrase activity. This dinucleo-
tide is superior in activity and stability toward nucleases
than a compound previously reported by us.7
Some oliognucleotides of natural origin are capable of
interfering with the integration process by competing
with viral DNA for binding to HIV integrase.5 This type
of protein–nucleotide binding is important in other
steps in the replication cycle of HIV, for example, in the
recognition and binding of Tat protein to HIV-1TAR
RNA.6 However, small oligonucleotides are rapidly
cleaved by cellular nuclease activity. In addition,
increasing nuclease resistance by chemical alteration of
the internucleotide phosphate bond results in decreased
The preparation of dinucleotide 1 (Scheme 1) utilized
the deoxyisouridine 28,9 as the starting point of the
synthesis. This precursor can be synthesized in several
steps from ribose.8 Compound 2 can be tailored for
coupling by its conversion to 3 in three steps (50-dim-
ethoxytritylation, 30-acetylation, and 50-detritylation).
Treatment of 3 with 4, the protected 20-deoxycytidine
phosphate, in the presence of TPS-TAZ and pyridine
followed by 50-deprotection with dichloroacetic acid
gave the coupled compound 5 (two diastereoisomeric
forms) in 72% overall yield. In the next step, 50-phos-
phorylation was carried out with our new phosphoryl-
ating reagent, 2-O-(4,40-dimethoxytrityl) sulfonyldiethanol
*Corresponding author. Tel.:+1-319-335-1364; fax: +1-319-353-
2621; e-mail: vasu-nair@uiowa.edu
yCurrent address: School of Pharmacy, University of Southern Cali-
fornia, Los Angeles, CA 9008, USA.
0960-894X/01/$ - see front matter # 2001Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00231-1

1434
M. Taktakishvili et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1433–1435
(Fig. 2) that was consistent with these results. This
unusual structure of the internucleotide sugar phosphate
diester linkage and the overall molecular conformation
of 1 appears to manifest itself in the total resistance of
the internucleotide bond of 1 towards hydrolytic cleav-
age catalyzed by exonucleases (PDE I and II from
bovine intestinal mucosa and bovine spleen, respec-
tively). Interestingly, even the ‘precursor’ dinucleotide 6
(i.e., compound 1 without the 50-phosphate group) was
totally resistant to cleavage by PDE II and almost so by
PDE I (Vmax/Km=0.61% of dCpdU). A comparison of
these data for PDE I (50-exonuclease) for 6 and the
natural counterpart (dCpdU) is illustrated in Figure 3.
Integrase inhibition assays were conducted with purified
recombinant HIV-1integrase using a 21-mer oligonu-
cleotide substrate as described by Pommier and co-
workers.10 The data showed that compound 1 has
strong inhibitory activity against recombinant wild-type
HIV integrase in reproducible assays [IC50=7.5 mM (30-
processing); IC50=5.9 mM (strand transfer)]. The data
for this compound are superior to that obtained for
mononucleotides7,11,12 and are similar to that of the
closest known natural dinucleotide, pdCpdT (IC50=8
and 6 mM).5 However, pdCpdT and other natural dinu-
cleotides are rapidly cleaved by 50-exonucleases while 1
is totally resistant to both 30- and 50-exonucleases. The
inhibition data and the nuclease resistance properties
for 1 are also more compelling than that exhibited by
the non-natural compound, pIsodApdC (IC50=19 and
25 mM), previously reported by us.7 Compound 1
inhibited all enzymatic activities of integrase and DNA–
integrase cross-linking in the same range as indicated
Figure 2. A preferred conformation of pdCpIsodU.
phosphate7 in the presence of TPS-TAZ and pyridine
followed by deprotection. Following purification by
HPLC, target compound 1 was obtained in 79% yield.
Its structure was confirmed by multinuclear NMR data
(1H, 13C, 31P, COSY, HMQC, and HMBC), quantita-
tive UV and CD spectra, and electrospray HRMS data
[calculated for 1: 596.0795 (M–H)ꢀ, found: 596.0790].
The CD and UV data [lmax 268 (e 15,900)] and observed
hypochromicity suggested the presence of base stacking in
preferred conformations and implied the presence of a
conformationally unusual internucleotide phosphate
bond because of the spatial arrangement of the two
sugar rings to accommodate base stacking. The mole-
cular modeling data showed a preferred conformation
Scheme 1. Key steps in the synthesis of dinucleotide 1.

M. Taktakishvili et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1433–1435
1435
Acknowledgements
The authors gratefully acknowledge support of this
research work by the National Institutes of Health
(NIAID) through AI43181. Our thanks are also expres-
sed to Dr. Travis Mickle and Dr. Suresh Pal of the Nair
Research Group at the University of Iowa for technical
assistance with molecular modeling and PDE studies.
References and Notes
1. Frankel, A. D.; Young, J. A. T. Annu. Rev. Biochem. 1998,
67, 1 .
2. Katz, R. A.; Skalka, A. M. Annu. Rev. Biochem. 1994, 63,
133.
3. Dyda, F.; Hickman, A. B.; Jenkins, T. M.; Engelman, A.;
Craigie, R.; Davies, D. R. Science 1994, 266, 1981.
4. Mazumder, A.; Neamati, N.; Ojwang, J. O.; Sunder, S.;
Rando, R. F.; Pommier, Y. Biochemistry 1996, 35, 13762.
5. Mazumder, A.; Uchida, H.; Neamati, N.; Sunder, S.;
Jaworska-Maslanka, M.; Wickstrom, E.; Zeng, F.; Jones,
R. A.; Mandes, R. F.; Chenault, H. K.; Pommier, Y. Mol.
Pharmacol. 1997, 51.
6. Tao, J.; Frankel, A. D. Proc. Natl. Acad. Sci. U.S.A. 1992,
89, 2723.
7. Taktakishvili, M.; Neamati, N.; Pommier, Y.; Pal, S.; Nair,
V. J. Am. Chem. Soc. 2000, 122, 5671.
8. (a) Purdy, D. F.; Zintek, L. B.; Nair, V. Nucleoside
Nucleotides 1994, 13, 109. (b) Nair, V.; Nuesca, Z. M. J. Am.
Chem. Soc. 1992, 114, 7951.
Figure 3. Lineweaver–Burk plots (1/V vs 1/[S], mMꢀ1) of the substrate
activities of the natural compound, dCpdU (top) and the novel non-
natural compound, dCpIsodU (bottom) with bovine 50-exonuclease
(PDE I).
9. Nair, V.; St. Clair, M.; Reardon, J. E.; Krasny, H. C.;
Hazen, R. J.; Paff, M. T.; Boone, L. R.; Tisdale, M.; Najera,
I.; Dornsife, R. E.; Everett, D. R.; Borroto-Esoda, K.; Yale,
J. L.; Zimmerman, T. P.; Rideout, J. L. Antimicrob. Agents
Chemother. 1995, 39, 1993.
10. Mazumder, A.; Neamati, N.; Sundar, S.; Owen, J.; Pom-
mier, Y. In Methods in Cellular and Molecular Biology: Anti-
viral Evaluation; Kinchington, D., Schinazi, R., Eds.;
Humana: Totowa, 1998.
11. Mazumder, A.; Neamati, N.; Sommadossi, J.-P.; Gosselin,
G.; Schinazi, R. F.; Imbach, J.-L.; Pommier, Y. Mol. Phar-
macol. 1996, 49, 621.
12. Mazumder, A.; Cooney, D.; Agbaria, R.; Gupta, M.;
Pommier, Y. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 5771.
above when Mn+2 or Mg+2 was used as a cofactor.
This suggests that the inhibitor binds to the catalytic
core of integrase and the inhibition is metal-indepen-
dent. The base sequence recognition (CU) by HIV inte-
grase is significant in that both bases C and U appear to
be recognized. This recognition of U in other small oli-
gonucleotide inhibitors has not been previously repor-
ted. The remarkable nuclease stability and the strong
anti-HIV integrase activity suggests that this compound
is worthy of further investigation as a potential antiviral
therapeutic agent and this and related studies are cur-
rently in progress.