Protein structure and oligomerization are important for the formation of export-competent HIV-1 Rev-RRE complexes

Abstract

The translation of the unspliced and partially spliced viral mRNAs that encode the late, structural proteins of HIV-1 depends on the viral-protein Rev. Oligomeric binding of Rev to the Rev response element (RRE) in these mRNAs promotes their export from the nucleus and thus controls their expression. Here, we compared the effects of hydrophobic to hydrophilic mutations within the oligomerization domain of Rev using assays for oligomeric RNA binding, protein structure, and export from the nucleus. Oligomeric RNA binding alone does not correlate well with RNA transport activity in the subset of mutants. However, protein structure as judged by CD spectroscopy does correlate well with Rev function. The oligomeric assembly of Rev-L18T is impaired but exhibits minor defects in structure and retains a basal level of activity in vivo. The prevalence of L18T in infected individuals suggests a positive selection mechanism for L18T modulation of Rev activity that may delay the onset of AIDS.

Figure 1.

Model for oligomeric assembly of Rev on the RRE. (A) A model of stepwise binding was proposed based on the observation of oligomerization deficient mutants (Jain and Belasco 2001). These states of the model include free Rev, Rev monomer bound, Rev dimer bound, and Rev oligomer bound RRE. Tail mutants arrested at the monomer bound state and were proposed to interrupt the tail oligomerization interface (dark gray). Head mutants arrested at the dimer bound state and were proposed to interrupt the head oligomerization interface (white). (B) Tail and head mutations were mapped onto a helical wheel model of the N-terminal helix-turn-helix motif of the oligomerization domain.

Figure 2.

In vitro assembly of HIV-1 Rev variants on the RRE. (A) Radiolabeled RRE RNA was titrated with purified variants of HIV-1 Rev at six different concentrations (left to right 6, 18, 60, 166, 500, and 1500 nM). Free RNA (green), monomeric Rev (black), dimeric Rev (red), and oligomeric Rev (purple) bound RRE radioactive bands are labeled. (B) Nonlinear least squares fits for the binding data to a model of stepwise binding are also shown. Fraction of RRE RNA bound by Rev at each binding step is plotted as a function of the log Rev concentration. Data points represent the measured percent of the total RRE RNA in each state. Error bars on the data points represent the standard deviation from at least three experiments. Solid curves represent the fits to the data. Each state is color coded to match the corresponding band on the gels. Systematic errors in the residuals result from the error introduced by correcting for background over gel areas of multiple sizes.

Figure 3.

In vivo activity of putative tail and head mutations of HIV-1 Rev. (A) Diagrammatic representation of the 3-RRE reporter system: The reporter contains subgenomic fragments of HIV-1 and a CMV promoter, as indicated. Translation of Gag in transiently transfected HeLa cells from the reporter is dependent on interactions between Rev and the included RRE sequence. Splice donor site is indicated by the upward pointing arrow. (B) HeLa cells were cotransfected with wild-type or mutant pRev, p3-RRE, and pEGFPN1. Immunostaining for Gag expression (red) allowed for estimation of Rev export activity, and GFP fluorescence (green) provided a measure of transfection efficiency. (C) Rev expression in the transfected cells was detected by Western blot. (D) Nuclear export activity of wild-type and mutant Rev proteins in HeLa cells transfected with p3-RRE plasmid. Rev mutant activities are expressed as a percentage of wild-type Rev activity (set to 100%). Error bars represent the standard deviation of three independent experiments.

Figure 4.

CD of HIV-1 Rev variants. All CD spectra of 30 μM wild-type and mutant Rev proteins were collected at 5°C. The double minima at 208 nm and 222 nm, characteristic of a helical protein, is apparent in all of the spectra. Visual inspection of the spectra demonstrates that the minimum at 208 is shallower for the L60R, I55N, and V16D mutations relative to the other Rev variants. The 222 nm/208 nm ratio for each variant are summarized in Table 1.

Figure 5.

Comparison of Rev sequences. Rev sequences isolated from two long-term nonprogressors (01zatm45 and 99zatm10) were aligned with wild-type Rev, the consensus sequence for HIV-1C (Novitsky et al. 2002), and the HIV-1 Rev consensus sequence calculated from the multiple sequence alignment from the Los Alamos Sequence database. The putative helix-turn-helix motif of the oligomerization/RBD is indicated graphically. An arrow marks the oligomerization domain mutations considered in this study, and their positions in the wild-type sequence are highlighted by a black oval. The consensus sequence represents the most probable character at each portion of the alignment. Sequence numbering does not include spaces where a gap is the most common character. A color-coded conservation score is depicted as a box under each position. Point mutations within nonprogressor sequences that are known to reduce Rev activity, and the additional inserts within the nonprogressor sequences are highlighted in gray.

Figure 6.

Summary of effects of mutations to the oligomerization domain of Rev. For each type of measurement, a qualitative grayscale scheme is used to depict the change for each mutation relative to wild type. In general, the mutations cause a range of structural and functional changes with mutation to position 18 being the least disruptive.