Genome size expansions are common among eukaryotic lineages. Enlarged genomes can be bioenergetically demanding, and active mobile elements can trigger chromosomal rearrangements and loss of gene function. What triggers genome size expansions remains largely unexplored in many biological clades, particularly within the fungal kingdom. Activation of large transposable elements (TEs), such as long-terminal repeats (LTRs), is a common contributor. Yet the mechanisms of LTR activation remain poorly understood. Here, we focus on the fungal genus Pseudocercospora and closely related species with known variation in genome size. In using an assembly-free approach, we found that TE content is highly variable among species, with species-specific retrotransposon families being the main drivers of independent genome expansions. We further focused on the two species with the most expanded genomes and reference-quality genomes, P. fijiensis and P. ulei. We found that the P. ulei genome is compartmentalized, with highly variable TE densities among chromosomal regions, and a striking reduction in pathogenicity-associated genes. Overall, our study indicates that species of Pseudocercospora originally had reduced genome sizes, and genome expansions are species-specific, driven by heterogeneous sets of TE families. We discuss what might have caused TE activation and subsequent proliferation in the genus, including stress conditions and host adaptation. Surveys of clades with highly dynamic genome sizes are crucial for the investigation of causal factors driving long-term TE dynamics.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13100-026-00396-x.

Transposable elements (TEs) constitute a substantial fraction of the human genome and contribute to gene regulatory programs. However, systematic analysis of TEs at the individual locus level remains technically challenging, particularly in single-cell contexts. While single-cell technologies have advanced the study of cellular heterogeneity, most analytical frameworks remain gene-centric. Existing TE-focused approaches are largely restricted to transcriptional profiling using scRNA-seq data, while analyses of single-cell chromatin accessibility have focused primarily on aggregate or family-level TE signals rather than individual loci. Consequently, no dedicated computational framework exists for quantifying chromatin accessibility at individual TE loci from scATAC-seq data, limiting investigation of locus-specific TE regulatory activity at single-cell resolution. scTELL (single-cell Transposable Element Locus-Level analysis) is a computational framework that quantifies TE accessibility at individual loci from scATAC-seq data using a distance-weighted scoring scheme. We applied scTELL to diverse biological systems, including healthy peripheral blood mononuclear cells (PBMCs), clear cell renal cell carcinoma (ccRCC), and breast cancer (BC). In PBMCs, scTELL identified distinct cell-type-specific TE accessibility patterns with clustering performance comparable to established gene activity scoring approaches, and validated key TE accessibility patterns using bulk ATAC-seq data from sorted immune cell populations. Motif enrichment analyses of TE-associated accessible regions revealed distinct TF motif landscapes, including family-level motif signatures, within-family locus heterogeneity across cell types, and motifs enriched in TE-associated regions relative to gene promoters. In cancer contexts, scTELL identified heterogeneity-associated TE loci and observed clinically associated accessibility patterns, including an L1PA2 locus in ccRCC associated with progression-free interval, and survival-associated TE loci in BC. scTELL provides a much-needed and robust tool to investigate the locus-specific regulatory landscape of TEs at single-cell resolution. Our findings demonstrate that this approach can uncover previously unrecognized cell-type-specific and disease-associated TE accessibility. The scTELL framework offers a new layer of biological insight, complementing existing single-cell analysis protocols and enabling the discovery of candidate biomarkers from a vast, understudied portion of the genome. While these associations are reproducible across datasets, prospective validation and functional studies will be required to establish clinical utility and to determine whether any locus has a causal role or therapeutic relevance.

Processed pseudogenes and retrogenes are defined by their RNA-mediated origin and, by virtue of this origin-based definition, are often interpreted as discrete genomic insertions. The completion of telomere-to-telomere (T2T) reference assemblies has substantially improved the resolution of segmental duplication architectures and centromeric satellite sequences that were previously inaccessible, allowing genomic structural contexts that were effectively invisible in earlier references to be directly examined. Using the SEPTIN14P-CICP locus family as a case study, chain-based comparative analyses showed that a genomic window spanning the SEPTIN14 3′ terminal exon and the adjacent processed pseudogene CICP12 is dispersed into multiple segmental duplication-associated units across great apes, rather than being maintained as a single orthologous locus. Genome-wide analyses further indicated that annotated CICP loci preferentially localize within segmental duplication blocks and accumulate near pericentromeric or subtelomeric regions. Despite this duplication-associated dispersion, codon-based selection analyses revealed pervasive purifying selection acting on the full-length SEPTIN14 coding sequence and its 3′ terminal exon, arguing against a model in which the terminal exon was newly formed through segmental duplication. Together, these results show that when highly conserved, strongly constrained coding regions are embedded within segmental duplication-rich regions, co-dispersed processed pseudogene copies can be interpreted as distinct from independently generated LINE-1-mediated insertions and as reflecting secondary structural propagation. When considered in light of origin-based definitions of processed pseudogenes and retrogenes, and specifically within duplication-rich and structurally unstable genomic regions resolved by T2T-level assemblies, these results suggest that multiple annotated loci can arise through secondary propagation of a single RNA-derived insertion. Under such contexts, incorporation of selective constraint and cross-species conservation enables more reliable distinction between source insertions and their secondarily propagated copies. This case study highlights a limitation of current annotation frameworks and demonstrates the need for more precise annotation that incorporates evolutionary and structural context in the T2T era.

The seventh Japanese meeting on host–transposon interactions, titled “Biological Function and Evolution through Interactions between Hosts and Transposable Elements,” was held on September 1st and 2nd, 2025, at the National Institute of Genetics, as well as online. This meeting was supported by the National Institute of Genetics and aimed to bring together researchers studying the diverse roles of transposable elements (TEs) in genome function and evolution, as well as host defense systems against TE mobility, TE bursts during evolution, and intron mobility in mammals, insects, land plants, fungi, and protozoa. Here, we present the highlights of these discussions.

How endogenous retroviral elements (ERVs), a family of transposable elements, may promote tumor progression is not well understood. Tripartite motif-containing 28 (TRIM28/TIF1b/KAP1) is a key transcriptional co-repressor protein that represses ERV expression in many cell types including embryonic stem cells, neural progenitor cells, differentiated adult cells, and cancer cells. In this study, we investigated the effect of Trim28 deletion on the expression of ERVs using an immune competent genetically engineered mouse model for prostate cancer. We found Trim28 deletion in prostate tumors led to the expression of ERVs in prostates from both hormonally intact and castrated mice. ERVs can regulate the expression of neighboring genes, and we detected increased expression of several protein-coding genes near overexpressed ERVs. Our data suggest that Trim28 deletion in prostate tumor epithelial cells may promote an innate immune response. However, Trim28 deletion also led to excessive deposition of tumor extracellular matrix (ECM). Our findings suggest that ECM alterations downstream of ERV derepression could affect immune cells in the tumor microenvironment and may promote tumor progression.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13100-025-00382-9.