The effects of alternate polypurine tracts (PPTs) and mutations of sequences adjacent to the PPT on viral replication and cleavage specificity of the Rous sarcoma virus reverse transcriptase

Abstract

We previously reported that a mutant Rous sarcoma virus (RSV) with an alternate polypurine tract (PPT), DuckHepBFlipPPT, had unexpectedly high titers and that the PPT was miscleaved primarily at one position following a GA dinucleotide by the RNase H of reverse transcriptase (RT). This miscleavage resulted in a portion of the 3' end of the PPT (5'-ATGTA) being added to the end of U3 of the linear viral DNA. To better understand the RNase H cleavage by RSV RT, we made a number of mutations within the DuckHepBFlipPPT and in the sequences adjacent to the PPT. Deleting the entire ATGTA sequence from the DuckHepBFlipPPT increased the relative titer to wild-type levels, while point mutations within the ATGTA sequence reduced the relative titer but had minimal effects on the cleavage specificity. However, mutating a sequence 5' of ATGTA affected the relative titer of the virus and caused the RNase H of RSV RT to lose the ability to cleave the PPT specifically. In addition, although mutations in the conserved stretch of thymidine residues upstream of the PPT did not affect the relative titer or cleavage specificity, the mutation of some of the nucleotides immediately upstream of the PPT did affect the titer and cleavage specificity. Taken together, our studies show that the structure of the PPT in the context of the cognate RT, rather than a specific sequence, is important for the proper cleavage by RSV RT.

FIG. 1.

Sequences of the DuckHepBFlipPPT mutants, their relative titers and 2-LTR circle junction analysis. (A) The PPT (bold) and flanking sequences of the WT RSV (RSV PPT) and the various DuckHepBFlipPPT mutants are shown. The specific nucleotide changes are underlined in the sequences of the DuckHepBFlipPPT mutants. The relative titers of the vectors encoding the mutations are shown on the right. The relative titers were determined by infecting 293-tva cells and normalizing the titer based on the amount of p27 antigen (capsid) present in the viral stock used for the infection. The results were normalized to the WT RSV. Shown are the averages of three independent infections ± the standard deviation. (B) 2-LTR circle junction data for the DuckHepBFlipPPT mutants. Also shown are the percentages of the different classes of sequences detected in the 2-LTR circle junction assay. Consensus sequences represent a joining of the correct ends of the linear viral DNA. Simple PPT insert mutations ranged from 1 nucleotide to the insertion of the entire PPT. The PPT sequences could be inserted either into a consensus sequence junction or into a junction in which a portion of the U5 sequence was deleted. PPT plus short flank inserts were insertions of the PPT with no more than 10 nucleotides from the segment immediately upstream of the PPT. The PPT plus long flank inserts were identical to the PPT plus short flank inserts, except that the insertions contained more than 10 nucleotides from the flanking region immediately 5′ of the PPT. The tRNA insertions consisted of 1 or more nucleotides from the tryptophan tRNA primer inserted at the circle junction. The tRNA plus PPT inserts were the same as the tRNA inserts, except that they also contained either a portion of PPT or the entire PPT. Small-deletion mutants had deletions of no more than 10 nucleotides in U5 and/or U3. Large-deletion mutants had deletions larger than 10 nucleotides in the U5 and/or U3. Insertion mutants had nucleotide insertions at the circle junction of sequences that were not derived from either the PPT or tRNA primer. In the cases in which there were both insertions of nucleotides not derived from the PPT or the tRNA and deletions of either U5 and/or U3, the mutants were classified based on whether the deletion or the insertion involved more nucleotides. The data represent the average of two independent experiments.

FIG. 2.

Cleavages of the DuckHepBFlipPPT mutants by the RNase H of RSV RT. The arrows depict the positions of the PPT miscleavages, which result in insertions of part or all of the PPT (simple PPT inserts) or small deletions of the U3. The positions of the cleavages are inferred from the analysis of the 2-LTR circle junctions. The arrows at the top represent the results from one experiment, and the arrows at the bottom represent the results from a second, independent experiment. The sizes of the arrows are related to the number of events (denoted by a number above and/or below the arrows). The line delineates the original border between the PPT and the U3 and, in the case of the DuckHepBFlipPPT4 mutant, where there were cleavages at this junction. The numbers above and below the line are the numbers of consensus sequences found in two experiments. The underlined nucleotides were mutated.

FIG. 3.

Sequence of PPT-U3 junctions from passaged DuckHepBFlipPPT viruses and their cleavage by RNase H. (A) The sequences from viruses at passage 3 (P3), passage 4 (P4), and passage 12 (P12) are shown. The PPTs from the passaged viruses are shown in bold, and the dashed lines represent the deleted nucleotides. (B) Cleavage of the passaged DuckHepBFlipPPT by the RNase H of RSV RT. The arrows represent the position of the (mis)cleavages of the PPT by RT. The preferential cleavage creates an insertion, and a smaller number of miscleavages causes deletions of U3. The number of events is denoted above the arrows and by the sizes of the arrows. The line represents the position of the PPT-U3 junction in WT RSV.

FIG. 4.

(Mis)cleavages of the RSV U3 and RSV PPT (A to T) mutants by the RNase H of RSV RT and the relative titer. (A) The arrows depict the positions of the PPT (mis)cleavages, shown the same way as in Fig. 2. (B) 2-LTR circle junction analysis of RSV U3 and RSV (A to T) mutants. The categories of the 2-LTR circle junctions are the same as those in Fig. 1B.

FIG. 5.

Relative titers of RSV upstream PPT and poly(U) tract mutants. The relative titers obtained from infection of 293-tva cells and normalized to the amount of p27 (capsid) measured by p27 antigen-capture ELISA are shown. The results were normalized to the WT RSV. The values are the averages of three independent experiments ± standard deviation.

FIG. 6.

2-LTR circle junction analysis of mutants 5′ of the PPT. (A) 2-LTR circle junction data. The data are the average of two independent experiments. The categories of the 2-LTR circle junctions are the same as those in Fig. 1B. (B) (Mis)cleavages of the PPT-U3 junctions derived from the mutants 5′ of the PPT. Shown are the PPT (bold), sequences 5′ of the PPT, and the 5′ end of U3. The line delineates the PPT-U3 junction, and the underlined nucleotides indicate the changes made to the WT sequence. The arrows depict the positions of the PPT (mis)cleavages, shown the same way as in Fig. 2.