**2. Resistance to Trans-Dominant RevM10 Therapy Induces a Conformational Change in the HIV-1 RRE**

The prototype Rev (NL4-3) from HIV clade B is a ~18 kD 116 amino acid phospho-protein [36]. Thee Rev protein comprises several well-characterized functional domains central to Rev-RRE-mediated nucleocytoplasmic transport of unspliced and partially spliced mRNAs [13]. One of these, aa 34–50 constitute the arginine-rich nuclear localization signal (NLS), the primary contact point to the RRE. The NLS is flanked by oligomerization domains that are required for Rev multimerization. Oligomerization is mediated mainly by Leu 12, Val16, Leu18, Leu55, and Leu60 [37]. Towards the carboxy-terminal is a leucine-rich nuclear export signal (NES) (aa 75–84) that mediates interactions with host nuclear export factors (Figure 1B, [38]). During a functional delineation study of Rev by site-directed mutagenesis, Malim et al. [39] constructed a defective variant of Rev, designated RevM10, by replacing two critical NES residues (Figure 1B). When expressed in transfected cells, this mutant Rev protein successfully inhibited the function of its wild-type counterpart. That RevM10 serves as an effective trans-dominant inhibitor of Rev function was further validated by others [40–44] thereby catapulting RevM10-based gene therapy into Phase I and II clinical trials [45]. However, a subsequent study by Hamm et al. [46], involving passage of HIV-1 in a T-cell line constitutively expressing RevM10, reported rapid emergence of a RevM10-resistant virus [46]. Surprisingly, escape mutations were not associated with Rev, but rather the RRE. The same study demonstrated that only two silent mutations (with respect to viral envelop protein) outside the primary Rev-binding region of the RRE, namely, G164 > A164 (at the base of SL-III/IV) and G245 > A245 (in the central loop) sufficed to confer RevM10 resistance, suggesting a conformational change in the RRE was responsible. With this in mind, RRE structures of wild-type and RevM10-resistant HIV variants containing these two RRE mutations (RRE-61) were examined by selective 2- -OH acylation monitored by primer extension (SHAPE) [47]. In this study, the chemical reactivity profile of the wild-type RRE (Figure 2A) predicted the 4 SL conformer (Figure 2B). In contrast, the two silent RRE-61 mutations, inducing RevM10 resistance, introduced significant alterations in chemical reactivity over ~60 nt (Figure 2A), deconvolution of which predicted the 5 SL conformer (Figure 2C). Constructing individually-mutated RRE variants, G164 > A164 (at the base of SL-III/IV) and G245 > A245 (in the central loop), allowed their contribution to be evaluated, where a combination of in vitro replication kinetics, non-denaturing polyacrylamide gel electrophoresis, and chemical footprinting indicated the G245 > A245 mutation sufficed to mimic the structure and activity of RRE-61. Interestingly, in this study, RevM10 resistance could not be attributed to differential/reduced Rev binding to RRE-61. Therefore, it is likely that the stable SL-IV structure in the 5 SL RRE61 provides binding sites for cellular or viral factors that mediate RevM10 resistance (see later). Nevertheless, the study by Legiewicz et al. [47] demonstrated a considerable

degree of conformation flexibility inasmuch as a relatively modest nucleotide change in the RRE central loop had a major impact on its topology.

**Figure 2.** Mutations conferring resistance to trans-dominant RevM10 therapy induce a conformational change in the HIV-1 RRE. (**A**) Structural probing of the wild-type RRE (left) and RevM10-resistant RRE-61 (right). (**C**) Control DNA sequencing lane from which nucleotide numbering was derived. Designations − and + refer to untreated and NMIA-treated RNA, respectively. Major alterations in chemical reactivity are indicated by the bar. (**B**,**C**) Cartoons depicting SHAPE-derived conformations of wild-type RRE and RRE-61. SL, stem-loop. Positions of RRE point mutations, inducing RevM10 resistance, are indicated by asterisks in (**B**). Modified from Reference [47].
