In cell mutational interference mapping experiment (in cell MIME) identifies the 5' polyadenylation signal as a dual regulator of HIV-1 genomic RNA production and packaging

Abstract

Non-coding RNA regulatory elements are important for viral replication, making them promising targets for therapeutic intervention. However, regulatory RNA is challenging to detect and characterise using classical structure-function assays. Here, we present in cell Mutational Interference Mapping Experiment (in cell MIME) as a way to define RNA regulatory landscapes at single nucleotide resolution under native conditions. In cell MIME is based on (i) random mutation of an RNA target, (ii) expression of mutated RNA in cells, (iii) physical separation of RNA into functional and non-functional populations, and (iv) high-throughput sequencing to identify mutations affecting function. We used in cell MIME to define RNA elements within the 5' region of the HIV-1 genomic RNA (gRNA) that are important for viral replication in cells. We identified three distinct RNA motifs controlling intracellular gRNA production, and two distinct motifs required for gRNA packaging into virions. Our analysis reveals the 73AAUAAA78 polyadenylation motif within the 5' PolyA domain as a dual regulator of gRNA production and gRNA packaging, and demonstrates that a functional polyadenylation signal is required for viral packaging even though it negatively affects gRNA production.

Figure 1.

(A) The HIV-1 5′UTR folds into a series of structural domains that control key steps of the HIV-1 life cycle including transcription, translation, export, packaging and reverse transcription. From 5′ to 3′ these structural domains are: transactivation response (TAR) for transcription; PolyA stem loop for polyadenylation; the primer binding site (PBS) for reverse transcription; SL1 promotes gRNA dimerization; SL2 contains the major splice donor (SD) site; SL3 has historically been considered the major packaging signal (Psi); the sequences surrounding the AUG start codon are thought to be involved in a base-pairing interaction with the upstream U5 region. (B) In cell Mutational Interference Mapping Experiment (in cell MIME). The proviral genome is randomly mutated using error prone PCR, and subsequently cloned into a gRNA expression vector. The structural and enzymatic proteins, Gag and Gag-Pol are expressed from a separate expression plasmid. Co-transfection of the mutant library and Gag/Gag-Pol expression plasmid into 293T cells leads to the transcription of mutant RNAs and subsequent sorting of functional and non-functional RNA populations by the viral and cellular machinery. Viral RNA present in cells and virus is reverse transcribed. Viral cDNA and the input DNA plasmid is amplified, fragmented, barcoded, sequenced on an Illumina HiSeq2500, and analysed using the MIMEAnTo software.

Figure 2.

In cell Mutational Interference Mapping Experiment (in cell MIME) discovery of RNA motifs regulating HIV-1 gRNA production (A) Log2 K_prod showing the maximal effect of mutations on RNA production in cells with the HIV-1 5′ UTR and Gag coding region (smoothed with a linear, two-sided convolution filter of width 2). Functional domains are indicated with coloured boxes below the graph. Positions with significant effects on RNA production are indicated by black triangles above the graph. Three regions with significant (P < 0.05) and strong (log2 K_prod ≥ 1 or ≤ –1; gray dotted line) effects on gRNA production are highlighted with red circles. (B to D) Mutations with maximal effect on log2 K_prod mapped on RNA structure. Positions impairing RNA production are shown in red. Positions improving RNA production shown in blue. Box and whisker plots show effect of each class of mutation on log2 K_prod. Black dot shows median, box shows quartiles (25% and 75%) and whiskers show extremes (excluding outliers beyond 1.5× IQR). Mutation classes are colour coded: red mutated to A; green mutated to C; blue mutated to G; yellow mutated to U. (B) Effect of mutations on gRNA production (log2 K_prod) mapped to TAR. (C) Effect of mutations on gRNA production (log2 K_prod) mapped to 5′ PolyA. All mutations to AAUAAA sequence improve gRNA production except for a single A to U mutation. (D) Effect of mutations on gRNA production (log2 K_prod) mapped to SL2. Mutations impairing gRNA production cluster to the U1 snRNA binding site.

Figure 3.

In cell Mutational Interference Mapping Experiment (in cell MIME) discovery of RNA motifs regulating HIV-1 gRNA packaging. (A) Log2 K_pack showing the maximal effect of mutations on RNA packaging with the HIV-1 5′ UTR and Gag coding region (smoothed with a linear, two-sided convolution filter of width 2). Functional domains are indicated with coloured boxes. Positions with significant effects on gRNA packaging are indicated by black triangles. Two regions with significant (P < 0.05) and strong (log2 K_pack ≥ 1; gray dotted line) effects on gRNA packaging are highlighted with dot red line/circle. (B and C) Mutations with maximal effect on log2 K_pack represented on RNA structure. Positions impairing gRNA packaging are shown in red. Positions improving gRNA packaging are shown in blue. Box and whisker plots show effect of each class of mutation on log2 K_pack. Black dot shows median, box shows quartiles and whiskers show extremes (excluding outliers beyond 1.5× IQR). Mutation classes are colour coded: red mutated to A; green mutated to C; blue mutated to G; yellow mutated to U. (B) Effect of mutations on gRNA packaging expressed as Log2 K_pack mapped to 5′ PolyA. All mutations to AAUAAA sequence impair gRNA packaging except for a single A to U mutation. (C) Effect of mutations on gRNA packaging expressed as log2 K_pack mapped to RNA structure in the region SL1–SL3. (D) Qualitative comparison between the significant effects of mutations on Pr55^Gag binding determined by in vitro MIME (upper portion, green) and the effects of mutations on gRNA packaging by in cell MIME (lower portion, blue). Sites significantly affecting both are pictured red. Color-coded arrows below (for in cell) and above (for in vitro) indicate the affected functional domain (colored boxes on the bottom). Filled arrows show significant effects at sites in both in vitro and in cell.

Figure 4.

Role of the AAUAAA polyA motif in gRNA production and packaging. (A) 5′PolyA mutants contain point mutations or deletions to the AAUAAA sequence. SL2 mutant containing substitutions within the U1snRNA binding site. (B) Production of gRNA and spliced viral RNA (mRNA Tat) for 5′ polyA and SL2 mutants. Bar charts represent six independent experiments. (C) Relative packaging efficiency of gRNA and spliced viral RNA into viral particles, expressed as a virus/cellular RNA (36). Bar charts represent 3 independent experiments. Statistical tests were carried out using ANOVA corrected for multiple comparisons.

Figure 5.

Five regulatory elements controlling HIV-1 replication. gRNA production is positively regulated by sequences within TAR and by the U1snRNP binding site within SL2. gRNA production is negatively regulated by the AAUAAA motif in 5′ polyA. The U1snRNP binding site is required for repression of 5′ polyadenylation. gRNA packaging into virions requires both the Pr55^Gag binding site (SL1-SL3), and the AAUAAA motif in 5′ PolyA. Positive regulatory elements are highlighted in green. Negative regulatory elements are highlighted in red.