Recombination of Alu and L1 elements generates somatic complexity in human genomes

Giovanni Pascarella (RIKEN Center for Integrative Medical Sciences, Japan) kicked off by outlining how recombination of repeat elements in human genomes is usually a consequence of the repair of DNA double strand breaks. Several structural variants can arise from this process - including inversion, deletions, and duplication – which have the potential to disrupt genomic integrity and function. Giovanni suggested that the known Alu and L1 element recombinations that lead to disease phenotypes are just the tip of the iceberg in terms of the amount of recombination of these elements in the genome. To test this hypothesis, the team analysed post-mortem tissues from 10 neurotypical donors. Samples were obtained from three sites within the brain plus matching liver and kidney samples from each individual. An antibody specific for neurons was used to separate neuronal and non-neuronal nuclear fraction, samples were then enriched for Alu/L1 using sequence capture enrichment, prior to analysis using short-read sequencing technology.

The TE-reX pipeline was utilised to identify recombination events in the captured libraries. Alu recombination was the predominant recombination type found across all different sample types and fractions. In general, a higher number of recombination events per genome were identified in kidney and liver samples; however, intra-chromosomal recombination was found to be enriched in brain samples.  Interestingly, most recombination occurred between repeat elements that were either close or far apart from each other. Recombination hotspots were shown to be enriched in cancer genes.

They then expanded this dataset to incorporate 10 samples each for sporadic cases of Parkinson’s disease (PD) and Alzheimer’s disease (AD). For both diseases they found an enrichment in intrachromosomal recombination rate in frontal cortex and parietal cortex when compared to control samples.

Next, the team wanted to determine if they could detect somatic recombination of Alu and L1 in capture-free and PCR-free conditions, for which they utilised nanopore sequencing on the PromethION. For this study, the temporal cortex neural fraction taken from the AD and control samples were analysed. In these 20 samples, over 30,000 recombination events were identified. Again, Alu elements were shown to be the major class of repeats.

Comparing the repeat events identified using capture-free, PCR-free nanopore sequencing with capture-based short-read sequencing technology revealed 1,700 shared repeats and 33,555 repeats unique to the nanopore dataset.

Summarising, Giovanni pointed delegates to his bioRxiv paper, where they can find much more data from this study: https://www.biorxiv.org/content/10.1101/2020.07.02.163816v1