Abstract Body

HIV-1 splicing produces more than 50 alternatively spliced RNAs. The initial product is a full-length transcript and all other RNAs are made by splicing. Short 1.8 kb mRNAs remove all introns, but long 4 kb mRNAs retain the env intron. Unspliced mRNAs retain all introns. Transcripts with introns require Rev for nuclear export. Splicing has been studied as a possible theraputic target, but transmitted/founder viruses tolerate wide variation among spliced transcript levels. HIV-1 is unique in that the majority of transcripts retain introns, and viral infection was shown to cause intron retention in transcripts of cellular genes (Sherrill-Mix 2015). This suppression of splicing is unique to HIV-1 and thus a potential theraputic target. We present evidence of early and complete splicing suppression, mutations that override suppression, and possible roles of Rev in splicing suppression.

Our Primer ID tagged deep sequencing assay quantifies HIV-1 spliced transcripts within the two RNA size classes. A second assay uses a Primer ID-tagged random reverse primer coupled with a fixed forward primer upstream of the major splice donor to detect changes in the amounts of full-length, long spliced transcripts (containing the env intron), and short fully spliced transcripts.

Mutations to exonic splicing silencers activated the upstream proximal acceptor site. They also lifted suppression of splicing (i.e. activated splicing) at the major 5′ splice donor D1 but not at splice donor D4. Activating mutations were made both to known silencers, and novel silencer elements were discovered by synonymous mutagenesis of the HIV-1 genome. We observed that the absence of Rev decreases but does not eliminate unspliced or partially spliced transcripts. We looked for but did not find evidence of transcripts that fail to use D1 (retain the gag/pro/pol intron) but use D4 to splice out the env intron. Additionally, we found no transcripts that used D2 or D3 without first using D1, making D1 suppression a marker for suppression of all splice donors.

Silencer elements that activate acceptor sites can also lift suppression of splicing from D1. As splicing is cotranscriptional and efficient, the absence of spliced gag/pro/pol transcripts suggests unspliced transcripts are suppressed early and completely from all splicing, long before the other downstream splice donors or the Rev Response Element (RRE) are transcribed. The role of Rev in splicing suppression remains to be fully studied.