Steric inhibition of human immunodeficiency virus type-1 Tat-dependent trans-activation in vitro and in cells by oligonucleotides containing 2'-O-methyl G-clamp ribonucleoside analogues

Abstract

We report the synthesis of a novel 2'-O-methyl (OMe) riboside phosphoramidite derivative of the G-clamp tricyclic base and incorporation into a series of small steric blocking OMe oligonucleotides targeting the apical stem-loop region of human immunodeficiency virus type 1 (HIV-1) trans- activation-responsive (TAR) RNA. Binding to TAR RNA is substantially enhanced for certain single site substitutions in the centre of the oligonucleotide, and doubly substituted anti-TAR OMe 9mers or 12mers exhibit remarkably low binding constants of <0.1 nM. G-clamp-containing oligomers achieved 50% inhibition of Tat-dependent in vitro transcription at approximately 25 nM, 4-fold lower than for a TAR 12mer OMe oligonucleotide and better than found for any other oligonucleotide tested to date. Addition of one or two OMe G-clamps did not impart cellular trans-activation inhibition activity to cellularly inactive OMe oligonucleotides. Addition of an OMe G-clamp to a 12mer OMe-locked nucleic acid chimera maintained, but did not enhance, inhibition of Tat-dependent in vitro transcription and cellular trans-activation in HeLa cells. The results demonstrate clearly that an OMe G-clamp has remarkable RNA-binding enhancement ability, but that oligonucleotide effectiveness in steric block inhibition of Tat-dependent trans-activation both in vitro and in cells is governed by factors more complex than RNA-binding strength alone.

Figure 1

(a) Sequence of the HIV-1 TAR RNA 39mer model stem–loop showing the region targeted by oligonucleotides and (b) sequences of the 2′-O-methyl oligoribonucleotides containing single G-clamp analogues.

Figure 2

Polyacrylamide gel mobility shift analysis showing binding of oligonucleotide 12TAR OMe and G-clamp-containing 12TAR OMe P5 to ³²P-labelled TAR RNA in TK80 buffer.

Figure 3

Oligonucleotide analogue inhibition of Tat-dependent in vitro transcription. Graphs show the amount of full-length transcript as a function of oligomer concentration: (a) all single G-clamp-containing 2′-O-methyl oligonucleotides; (b) oligonucleotide containing two G-clamp analogues compared with 12TAR OMe and 12TAR OMe P5.

Figure 4

Effect of oligonucleotide analogues delivered by the cationic surfactant GS11 on HeLa Tet-Off/Tat/luc-f/luc-R cells. 12TAR 7OMe–5LNA, 12TAR 6OMe–5LNA+P4 and 12TAR 7OMe–4LNA+P5 concentrations are shown as a black wedge from left to right: 1000, 500, 250, 125, 62.5 and 0 nM, respectively.

Scheme 1. (A) Markiewicz reagent, pyridine (95% yield). (B) (i) TMSCl, Et₃N, DCM; (ii) MesCl, Et₃N, DMAP; (iii) o-nitrophenol, DBO; (iv) 2% tosic acid, DCM (83%). (C) MeI, Ag₂O, acetone (92%). (D) 4-Nitrobenzaldoxime, tetramethylguanidine, dioxane/water (76%). (E) (i) CCl₄, PPh₃, DCM; (ii) 2-aminoresorcinol (86%). (F) Benzyl-N-(2-hydroxyethyl)carbamate, DEAD, PPh₃ (70%). (G) KF, EtOH (64%). (H) (i) H₂, Pd/C, DMF; (ii) ethyltrifluoroacetate, DMAP (72%). (I) DmtCl, pyr. (85%). (J) Phosphitylating agent, DIPEA, DCM (90%).