James Ye

Retrotransposons are genomic parasites frequently reactivated in cancers, in which their mobility can cause genetic alterations. However, it remains unclear whether their gene products contribute to cancer beyond mutagenesis. In this study, we uncover a chromatin-associated function of RNAs from long interspersed element-1 (LINE-1), the only autonomous retrotransposon in the human genome. Subcellular-resolved transcriptomics revealed that LINE-1 RNAs are primarily nascent transcripts produced by full-length, cell type–specific genomic copies of evolutionarily young subfamilies. Using a long-read chromosome conformation assay, we identified a class of highly interactive LINE-1 loci required for gene expression across cancer subtypes, revealing an unexpected regulatory role for LINE-1 locus transcription in oncogenic gene control. LINE-1 RNA depletion disrupted LINE-1–centric chromatin interactions and downregulated associated genes, whereas genomic insertion of an inducible LINE-1 generated de novo chromatin interactions in a transcription-dependent manner. Therefore, beyond their mutagenic potential, retrotransposons also regulate cancer gene expression by nucleating chromatin architecture through their transcriptional activity.Significance:We developed a long-read chromatin conformation assay capable of resolving repetitive DNA, enabling the identification of highly interactive LINE-1 retrotransposon loci that nucleate chromatin architecture in cancer cells. This activity is driven by chromatin-associated LINE-1 transcripts, offering a biological rationale for the frequent reactivation of LINE-1 transcription in cancer.

Figure S1. Subcellular localization of LINE-1 RNAs in cancer cells. Figure S2. Chromatin-associated LINE-1 RNAs are nascent transcripts associated in cis with their sites of transcription. Figure S3. Chromatin RNA-sequencing of human cancer cells and primary tumors to quantify locusspecific LINE-1 expression. Figure S4. Analysis of uniquely and multi-mapped reads by RNA-seq of LINE-1 expression. Figure S5. Identification of the expression cut-off for RNA-seq analysis of LINE-1 locus expression. Figure S6. Identification of cell type-specific LINE-1 locus expression in human cancer cell lines. Figure S7. Lineage-specific transcription factors and chromatin features regulate LINE-1 locus expression. Figure S8. LINE-1 Capture Pore-C for detecting LINE-1-centric chromatin interactions. Figure S9. Identification of LINE-1-centric chromatin interactions by L1CaP-C. Figure S10. Mappability analysis of L1CaP-C sequencing reads associated with chromatin interactions. Figure S11. LINE-1-centric chromatin interactions identified by L1CaP-C. Figure S12. Identification of highly interacting LINE-1 loci. Figure S13. L1CaP-C identifies pairwise and high-order chromatin interactions. Figure S14. Analysis of chromatin marks at HILL LINE-1 loci and super-enhancers (SEs). Figure S15. Correlation of HILL LINE-1 loci with clinical features in TCGA breast cancer samples. Figure S16. LINE-1 locus expression regulates oncogenic gene expression. Figure S17. LINE-1 depletion impairs cell proliferation and oncogenic gene expression.

Table S1. List of genomic datasets used in this study. Table S2. List of oligonucleotides used in this study. Table S3. List of key reagents and resources used in this study. Table S4. List of full-length LINE-1 loci (>5 kb) annotated in the hg38 human reference genome.

Retrotransposons are genomic parasites frequently reactivated in cancers, in which their mobility can cause genetic alterations. However, it remains unclear whether their gene products contribute to cancer beyond mutagenesis. In this study, we uncover a chromatin-associated function of RNAs from long interspersed element-1 (LINE-1), the only autonomous retrotransposon in the human genome. Subcellular-resolved transcriptomics revealed that LINE-1 RNAs are primarily nascent transcripts produced by full-length, cell type–specific genomic copies of evolutionarily young subfamilies. Using a long-read chromosome conformation assay, we identified a class of highly interactive LINE-1 loci required for gene expression across cancer subtypes, revealing an unexpected regulatory role for LINE-1 locus transcription in oncogenic gene control. LINE-1 RNA depletion disrupted LINE-1–centric chromatin interactions and downregulated associated genes, whereas genomic insertion of an inducible LINE-1 generated de novo chromatin interactions in a transcription-dependent manner. Therefore, beyond their mutagenic potential, retrotransposons also regulate cancer gene expression by nucleating chromatin architecture through their transcriptional activity. Significance: We developed a long-read chromatin conformation assay capable of resolving repetitive DNA, enabling the identification of highly interactive LINE-1 retrotransposon loci that nucleate chromatin architecture in cancer cells. This activity is driven by chromatin-associated LINE-1 transcripts, offering a biological rationale for the frequent reactivation of LINE-1 transcription in cancer.

The Legionella SidE effectors ubiquitinate host proteins independently of the canonical E1-E2 cascade. Here we engineer the SidE ligases to develop a modular proximity ligation approach for the identification of targets of small molecules and proteins, which we call SidBait. We validate the method with known small molecule-protein interactions and use it to identify CaMKII as an off-target interactor of the breast cancer drug ribociclib. Structural analysis and activity assays confirm that ribociclib binds the CaMKII active site and inhibits its activity. We further customize SidBait to identify protein-protein interactions and discover the F-actin capping protein (CapZ) as a target of the Legionella effector RavB during infection. Structural and biochemical studies indicate that RavB allosterically binds CapZ and decaps actin, thus functionally mimicking eukaryotic CapZ interacting proteins. Collectively, our results establish SidBait as a reliable tool for identifying targets of small molecules and proteins.