Massively parallel jumping assay decodes Alu retrotransposition activity

Abstract

The human genome contains millions of copies of retrotransposons that are silenced but many of these copies have potential to become active if mutated, having phenotypic consequences, including disease. However, it is not thoroughly understood how nucleotide changes in retrotransposons affect their jumping activity. Here, we develop a massively parallel jumping assay (MPJA) that tests the jumping potential of thousands of transposons en masse. We generate a nucleotide variant library of four Alu retrotransposons containing 165,087 different haplotypes and test them for their jumping ability using MPJA. We found 66,821 unique jumping haplotypes, allowing us to pinpoint domains and variants vital for transposition. Mapping these variants to the Alu-RNA secondary structure revealed stem-loop features that contribute to jumping potential. Combined, our work provides a high-throughput assay that assesses the ability of retrotransposons to jump and identifies nucleotide changes that have the potential to reactivate them in the human genome.

Fig. 1. Massively parallel jumping assay.

a Schematic showing the random mutagenesis strategy to generate Alu variant libraries using error-prone PCR. b Retrotransposition assay and Alu integration into the genome with the help of L1 transposase machinery. The transposition vector contains RNA PolIII 7SL driving an Alu-Neo cassette that is spliced and complexed with the ORF2 transposase machinery from the helper plasmid. The Alu-Neo cassette gets reverse transcribed and integrates randomly into the genome, allowing the neomycin resistance gene to be expressed through the RNA PolII SV40 promoter, thus conferring G-418 sulfate-resistant colonies. c Retrotransposed Alu resurrection and retrieval from the genome and sequence library generation. Alu-Neo integrations were selected using neomycin-specific and Alu-specific primers that generate a 1 kb PCR product if neomycin is spliced due to retrotransposition. Alu-specific primers are then used to amplify the integrated Alus, which are then processed for sequencing.

Fig. 2. Alu selection and variants calling in the mutagenized library.

a Sequence similarity (percent sequence identity), the number of sequence differences, and the relative jumping potential to AluYa5 (positive control element considered to have activity score of 1) for all four tested Alu sequences. b Pie charts showing the percent of haplotypes detected in the Alu-Mut jumping libraries from total haplotypes detected in Alu-Mut plasmid libraries. c Variant calling at each position across the 280 bp full-length Alu sequence in the Alu-mutagenized jumping/plasmid library and haplotype calling depicting only significant high-jumping and low-jumping haplotypes beyond a cutoff of ±2.5 log2 fold-change (Log2FC). Significant jumping was defined via the DESeq2 package using a Wald-test (two-sided) p value threshold of 10⁻⁵. The lower panel shows a schematic depicting the major structural features in Alu-RNA.

Fig. 3. Haplotypes of dominant jumping effects.

a–c Violin (left panel) and Volcano (right panel) plots showing fold-change differences in high jumpers, low jumpers, and possible non-jumping Alu haplotypes with cutoff of ±2 log2 fold-change (Log2FC) for Alu14B-Mut (a), Alu6B-Mut (b), AluSx-Mut (c), respectively. Significant effects were defined using the DESeq2 package with a Wald-test (two-sided) and p value threshold of 10⁻⁵. Classes were defined as significant high jumpers in red (Log2FC > 2), significant low jumpers in blue (Log2FC < −2, jumping counts > 0), and non-jumpers in purple (Log2FC < −2, jumping counts = 0 and plasmid count > 50). Non-significant haplotypes are shown in gray. Plots are normalized to a zero Log2FC of the reference or wild-type sequence (in green) of each element. Horizontal lines in the Violin plot are the median. d The number of nucleotide changes observed in the library with respect to the reference or wild-type sequence for high jumper (top) and low/non-jumper haplotypes (bottom). Box and whisker plots include the median line, the box denotes the interquartile range (IQR), whiskers denote the rest of the data distribution, and outliers are denoted by points greater than ±1.5 × IQR.

Fig. 4. Mutations that affect jumping are associated with SRP binding domains.

a–c In the top panel, 5bp-sliding-window analyses of Alu6B (a), Alu14b (b), and AluSx (c) are shown along with the Alu-RNA folding structure in the panels below. The folding structure is reverse mountain coded with different colors for each secondary structure: stem-loop-unhybridized in teal, hybridized left strand of the stem in dark blue, and hybridized right arm of the stem in yellow. The upper right panel shows predicted RNA secondary structures with the left SRP and right arm SRP binding folds highlighted in red. The right lower panels show hypergeometric analyses (Hypergeometric test one-sided) with the enrichment of variants of dominant jumping effect in SRP binding regions (red bar) versus the enrichment of variants of negative jumping effect in SRP binding regions (blue bar). Next to these, sequence logos show DNA nucleotide composition observed in high jumper (top) and low jumper (bottom) haplotypes at select positions.

Fig. 5. Comparison of endogenous human genome AluS sequences and Alu-MPJA haplotypes.

a 852 genomic AluS from the human genome was compared to the haplotypes found in the mutagenized libraries of each element. We detected haplotypes that are at least seven mismatches away from the genomic AluS in AluSx and Alu6B mutagenized libraries and at least eleven mismatches away for Alu14B haplotypes. b Box and whisker chart showing the mismatch distance of high (top panel) or low (bottom panel) jumpers in the MPJA from genomic AluS sequences. Box and whisker plots include the median line, the box denotes the interquartile range (IQR), whiskers denote the rest of the data distribution, and outliers are denoted by points greater than ±1.5 × IQR. c Dot plot showing fifteen identified AluS elements from the human genome (Supplementary Data 8) that require minimum number of nucleotide changes to match with the high jumper haplotypes.