Abstract
Transposons, DNA sequences capable of relocating within the genome, make up a significant portion of eukaryotic genomes and are often found in clusters. Within the cell nucleus, the genome is organized into chromatin, a structure with varying degrees of compaction due to three-dimensional folding. Transposon insertion or activation can lead to chromatin decompaction, increasing accessibility and potentially facilitating further nearby insertions. This positive feedback between chromatin unfolding and transposon insertion may result in transposon clustering. Here, we combine bioinformatics with polymer modeling to explore possible mechanisms and conditions that promote clustered transposon insertions. Our analysis of human cell line genomic repeat data reveals extensive clustering of heterochromatic LINE-1 elements and euchromatic Alu elements. For Alu elements, this clustering correlates with increased chromatin accessibility. Both Alu and LINE-1 deviate in their sequence-inherent flexibility from the overall genome, with above-average flexibility for Alu and below-average flexibility for most LINE-1 sequences. Flexibility was highest in young transposons, so that young Alu and LINE-1 exceed overall genome flexibility. We developed an according polymer model of transposon insertion, consisting of a self-attracting chromatin domain. Transposon insertions locally disrupt self-attraction, leading to unfolding of the domain as more transposons are inserted. In simulations where transposons are inserted adjacent to existing ones, we observed gradual unfolding through loop extensions from a folded core. Including transposases as explicit particles, our model shows that adjacent transposon insertion occurs when densely packed chromatin excludes transposases or when insertion rates exceed the thermal equilibration rate of polymer configurations. We conclude that 1) dense chromatin packing that hinders transposase access as well as 2) a local loss of compaction upon transposon insertion favor clustered transposon insertion via loop formation. This biophysical mechanism of clustered insertion site preference would act in combination with selective pressures shaping transposon distribution over evolutionary timescales.