Figure 1.
Organization of EIAV genome, transcript splicing patterns, and location of the ERRE sequence.
(A) Schematic view of the EIAV genome showing the locations of open reading frames and alternatively spliced mRNA transcripts generated from the EIAV genome: mRNA transcript (a) encodes both the Tat (T) and Rev (r, rev) proteins. In the presence of Rev protein, EIAV exon 3 is skipped and the Tat (T) protein is produced from mRNA (b). mRNA (c) encodes Ttm, a protein of unknown function. Structural and enzymatic proteins are translated from mRNAs (d) and (e). Unspliced mRNA (e) corresponds to progeny RNA that is packaged to produce infectious virus. (B) Sequence of the exonic splicing enhancer (ESE) within the 555 nt ERRE (genomic location, nt 5280–5834), previously designated as ERRE-1 [34]. Boxed sequences represent two purine-rich sequence stretches (PuA and PuB) previously reported to interact with both the EIAV Rev protein and host protein SF2-ASF [29], [32].
Figure 2.
RNA secondary structural models for the ERRE.
(A) The lowest free energy secondary structure of the ERRE, predicted by Mfold without incorporatating experimental constraints [36]. (B) Lowest free energy RNA secondary structure models for the ERRE, generated by 4 different algorithms, all using chemical probing results as experimental constraint input: Mfold, Sfold [38], RNAStructure [39], and RNAfold [40]. SL-X, -Y and -Z, stem-loop structures common to all four models, are highlighted; the exonic splicing enhancer (ESE) is indicated by heavy lines (see Materials and Methods for details).
Table 1.
Oligonucleotide primers used in the primer extension analysis of ERRE
Figure 3.
Chemical probing results mapped onto the RNA secondary structure of ERRE.
RNAalifold [41] was used to generate an optimized RNA secondary structure of the ERRE based on a combination of thermodynamic considerations, experimental constraints, and sequence covariation information derived from multiple sequence alignment of a collection of 140 ERRE sequence variants. Arrowheads denote ribonucleotides modified by chemical probing reagents: kethoxal (red squares) and DMS (green circles), with the relative extent of modification represented by either two (strong) or one (weak) symbol. SL-X, -Y and -Z are stem-loop structures also shown in Figure 2.
Figure 4.
Two distinct regions of the ERRE undergo structural transitions in the presence of bound EIAV Rev protein.
Consensus chemical modification patterns, based on at least 3 experiments in which several different primers were used to probe the complete ERRE RNA, mapped onto the RNA secondary structure model shown in Figure 3. Ribonucleotides that were consistently displayed enhanced modification with either kethoxal or DMS upon Rev binding are circled: bold circle (strong) and thin circle (mild). Regions protected from hydroxyl radical cleavage in the presence of Rev are denoted by a thick line. Purine-rich motifs are highlighted in green.
Figure 5.
Representative footprinting results for RBR-1.
(nt 5449–5539, genomic location). (A) Representative gels from primer extension analysis of chemical probing experiments using kethoxal (a) and DMS (b). Similar experiments were performed using the hydroxyl radical cleavage reagent, Fe-EDTA (data not shown). Circled ribonucleotides denote positions with enhanced reactivity in the presence of bound EIAV Rev protein (“footprints”). Lanes A & G, Dideoxy sequencing markers; lane K, control, unmodified ERRE, in the absence of Rev; lane M, ERRE modified in the absence of Rev; (lanes ▴) ERRE modified in the presence of increasing amounts of Rev protein (1∼30 fold molar excess). (B) Consensus chemical modification patterns in RBR-1, based on several experiments similar to those illustrated in part (A), are mapped onto the corresponding portion of the RNA secondary structure of the ERRE (from Figure 3). Ribonucleotides that consistently display enhanced modification with either kethoxal or DMS upon Rev binding are circled: bold circle (strong) and thin circle (mild). Regions protected from hydroxyl radical cleavage in the presence of Rev are denoted by a thick line. Purine-rich motifs are highlighted in green.
Figure 6.
Representative Rev footprinting results in RBR-2.
(nt 5639–5749, genomic location). (A) Representative gels from primer extension analysis of chemical probing experiments using hydroxyl radical cleavage reagent, Fe-EDTA (•OH) (a) and DMS (b). Similar experiments were performed using kethoxal (data not shown). Circled ribonucleotides denote positions with enhanced reactivity in the presence of bound EIAV Rev protein (“footprints”). Lanes A & G, Dideoxy sequencing markers; lane K, control, unmodified ERRE, in the absence of Rev; lane M, ERRE modified in the absence of Rev; (lanes ▴) ERRE modified in the presence of increasing amounts of Rev protein (1∼30 fold molar excess). (B) Consensus chemical modification patterns in RBR-2, based on several experiments similar to those illustrated in part (A), are mapped onto the corresponding portion of the RNA secondary structure of the ERRE (from Figure 3). Ribonucleotides that consistently display enhanced modification with either kethoxal or DMS upon Rev binding are circled: bold circle (strong) and thin circle (mild). Regions protected from hydroxyl radical cleavage in the presence of Rev are denoted by a thick line. Purine-rich motifs are highlighted in green.
Figure 7.
Conservation of RNA sequences in the gp90 (SU) gene of EIAV.
Conservation of RNA sequences in EIAV env gene (gp90) was assessed by evaluating information content at each nucleotide position in a CLUSTALW-generated multiple sequence alignment of 140 gp90 sequence variants (see Materials and Methods for details). Information content (I) is plotted against nucleotide position, numbered from the first ribonucleotide in the ERRE. The gp90 gene begins at position 35. The maximum information content value is 2, which corresponds to 100% conservation at a particular position in this alignment. The locations of gp90 hypervariable regions, identified in a previous analysis of SU variants [42], are indicated by horizontal blue bars above the graph. Horizontal bars indicate the locations of the ESE (pink), EIAV Rev binding regions 1 and 2 (RBR-1 & RBR-2) (red), and EIAV Rev exon 1 (green).
Figure 8.
A conserved RNA structural motif identified in the high affinity Rev-binding sites of HIV-1 and EIAV is found within the RREs or env genes of diverse lentiviruses.
(A) Two representative structures (#1 and #11) from an ensemble of 20 pairs of similar structures are shown. These motifs were identified in 100 nt regions containing high affinity Rev binding sites from HIV-1 or EIAV, which were extracted and structurally aligned using Dynalign [51] to generate ensembles of structures. (B) Similar motifs were identified in genomic RNA sequences of 8 additional lentiviruses. The RNA structural motifs shown in (A) were used to scan the complete genomic RNA sequences of additional lentiviruses: HIV-2, SIV (simian immunodeficiency virus), RELIK (rabbit endogenous lentivirus type K), FIV (feline immunodeficiency virus), OLV (ovine lentivirus), JDV (jembrana disease virus), BIV (bovine immunodeficiency virus), CAEV (caprine arthritis-encephalitis virus), and Visna (visna virus). The motifs shown are examples of those similar to structure #11 in Figure 8A (see Materials and Methods and Table S1 for details).