Biology subjects

Chang, H. Y.

Publications and source records attributed to Chang, H. Y..

At least 19 recordsLinked to original sources

bioRxiv2026

A mitotic restriction point for extrachromosomal DNA hubs in cancer cells

Extrachromosomal DNA (ecDNA) is widespread in human cancers and promotes tumour evolution, intratumour heterogeneity and poor clinical outcomes. ecDNA molecules tend to cluster into micron-scale nuclear hubs that enhance oncogene transcription. However, how these higher-order assemblies form and persist remains largely unknown. Here we discover mitosis as the critical window for ecDNA hub organisation across cell generations. We develop an optical screen that identifies the mitotic-entry regulator WEE1 as an important regulator of hub integrity and oncogene expression. Analysis of WEE1 perturbation reveals that CDK1- and CDK2-regulated programmes cooperate to carry unresolved damage into mitosis, promoting reorganisation of ecDNA tethered to mitotic chromosomes. Long-term live-cell imaging shows that, in unperturbed cells, ecDNA hub architecture is remodelled primarily during mitosis and subsequently maintained throughout interphase. This coupling between mitotic and interphase organisation is conserved across diverse cancer types and ecDNA-associated oncogenes. Together, these findings define a mitotic restriction point that governs ecDNA hub organisation, oncogene expression and ecDNA inheritance through successive cell generations, revealing a potential window for therapeutic intervention in ecDNA-driven cancers.

cancer biology↗

bioRxiv2026

Inverted logic of ecDNA localization converts nuclear periphery into oncogenic transcription hubs in cancer

Extrachromosomal DNA (ecDNA) is a major driver of cancer pathogenesis, however, its spatial organization in the nucleus is not currently understood. Here we show that ecDNAs, in contrast to chromosomal DNA, preferentially localize to the nuclear periphery yet resist peripheral gene silencing. Single-molecule multi-omic sequencing and imaging-based CRISPR screening demonstrate enhanced ecDNA hub formation, H3K27ac and H3K9me3 bivalency, late DNA replication, long-range CpG hypomethylation, and increased transcriptional output when ecDNAs localize to the nuclear periphery. EcDNA anchors to the nuclear lamina through H3K9me3-dependent interactions with the Lamin B Receptor (LBR), which if interrupted, results in inward migration of ecDNA particles, reduced ecDNA congregation, and decreased transcription of ecDNA-encoded genes. These results reveal inherent advantages of the spatial organization of ecDNA that are not available to chromosomal DNA, enhancing ecDNA congregation and amplifying oncogenic transcriptional output, by localizing to the nuclear periphery of cancer cells.

cancer biology↗

bioRxiv2026

Dynamic optimization of extrachromosomal DNA copy number drives tumour evolution

Extrachromosomal DNA (ecDNA) is common in human cancers and is associated with poor clinical outcomes, yet how ecDNA-driven genetic heterogeneity is translated into functional heterogeneity remains unclear. Using single-cell multiomics sequencing and multiplexed IF-FISH, we show that asymmetric inheritance of ecDNA generates copy number heterogeneity that propagates to gene expression programs, including oncogenic signaling and cellular stress responses. Transgenerational live-cell lineage tracking directly shows that ecDNA heterogeneity arises within only a few cell divisions and modulates daughter cell division timing in a copy number-dependent manner, a property not observed for evenly inherited chromosomal amplicons. We identify an optimal middle ecDNA copy number range that maximizes proliferative fitness at baseline, while drug selection pressure induced by low-dose CHK1 inhibition selects for cells with a new optimal range at low ecDNA copy numbers. These low ecDNA copy number cells pre-exist in the population and can be generated de novo, driving copy number shifts promoting drug resistance. In vivo experiments further demonstrate that shifts toward ecDNA copy numbers that are optimal under the tumour microenvironment enhance tumourigenicity. Together, these findings establish ecDNA copy number plasticity as a central driver of tumour evolution.

cancer biology↗

bioRxiv2026

scAmp analyzes focal gene amplifications at single-cell resolution

Oncogene amplification on extrachromosomal DNA (ecDNA) is a common driver of tumor progression and is associated with acquired drug resistance and poor patient survival. While whole genome sequencing (WGS) studies have revealed the landscape of genes amplified on ecDNA in tumors, it remains challenging to study the subclonal heterogeneity and functional (e.g., transcriptomic) consequences of ecDNA on tumors. To address this, we introduce scAmp: a probabilistic algorithm for detecting and analyzing ecDNA from single-cell datasets. We demonstrate scAmps improved accuracy over WGS approaches on well-characterized cell-lines and its applicability to clinical histopathology. We further showcase scAmp by analyzing 73 patient tumors profiled with single-cell ATAC-seq, where we analyze the subclonal evolution of ecDNA+ subclones and identify the effect of ecDNA amplifications on the chromatin accessibility landscape of cancer cells. Together, we anticipate that scAmp will broadly enable further studies - both retrospective and prospective - that dissect critical questions of how ecDNA affect cancer cells and the tumors in which they reside.

cancer biology↗

bioRxiv2026

Deep learning-guided design of cell type-specific AAV promoters

Precise cell type targeting is critical for both clinical and experimental applications of adeno-associated viral (AAV) vectors, yet engineering vectors with cell type-specific activity remains a challenge. Here, we compared three strategies leveraging single-cell chromatin accessibility data to design cell type-specific AAV promoters, including a deep learning-based method to generate de novo regulatory sequences. When applied to target retinal ganglion cells or horizontal cells in mouse retina, deep learning-guided design consistently outperformed rational approaches, yielding synthetic promoters with stronger and more specific expression in vivo. Synthetic AAV promoters supported diverse transgenes, enabling the recording and ablation of targeted cells. Promoter activity was also maintained in human retinal organoids, suggesting that deep learning-designed sequences may be suitable for translation. Our findings highlight the potential of deep learning to synthesize cell type-specific AAV promoters and establish a versatile platform for cell type targeting with broad implications for gene therapy and basic research.

synthetic biology↗

bioRxiv2025

Genetic elements promote retention of extrachromosomal DNA in cancer cells

Extrachromosomal DNA (ecDNA) is a prevalent and devastating form of oncogene amplification in cancer1,2. Circular megabase-sized ecDNAs lack centromeres and segregate stochastically during cell division3-6 yet persist over many generations. EcDNAs were first observed to hitchhike on mitotic chromosomes into daughter cell nuclei over 40 years ago with unknown mechanism3,7. Here we identify a family of human genomic elements, termed retention elements, that tether episomes to mitotic chromosomes to increase ecDNA transmission to daughter cells. We develop Retain-seq, a genome-scale assay that reveals thousands of human retention elements conferring generational persistence to heterologous episomes. Retention elements comprise a select set of CpG-rich gene promoters and act additively. Live-cell imaging and chromatin conformation capture show that retention elements physically interact with mitotic chromosomes at regions which are mitotically bookmarked by transcription factors and chromatin proteins, intermolecularly recapitulating promoter-enhancer interactions. Multiple retention elements are co-amplified with oncogenes on individual ecDNAs in human cancers and shape their sizes and structures. CpG-rich retention elements are focally hypomethylated; targeted cytosine methylation abrogates retention activity and leads to ecDNA loss, suggesting that methylation-sensitive interactions modulate episomal DNA retention. These results highlight the DNA elements and regulatory logic of mitotic ecDNA retention. Amplifications of retention elements promote the maintenance of oncogenic ecDNA across generations of cancer cells, revealing the principles of episome immortality intrinsic to the human genome.

cancer biology↗

bioRxiv2025

PLI Analyzer for Data-driven Validation of AI Predicted Biomolecular Interfaces

PLI-Analyzer is a computational tool for detailed analysis of protein, RNA, and DNA interaction interfaces, combining atomic-level contact detection with domain annotations from UniProt and InterPro. It offers customizable interaction thresholds, residue-level output, and compatibility with AI-generated structures, enabling precise validation and interpretation of contact dynamics. We applied PLI-Analyzer to evaluate high-quality protein-protein complexes predicted by AlphaFold-3 and Boltz-2, revealing notable differences in predicted interfaces and confidence scores. These findings highlight current limitations in generative structure models specifically for macromolecular complexes and underscore the need for robust, domain-informed evaluation frameworks in structural bioinformatics.

bioinformatics↗

bioRxiv2025

Local translation of circular RNAs is required for synaptic activity and memory

Circular RNAs (circRNAs) are enriched in synapses and implicated in cognitive processes, and recent studies have shown that circRNAs can encode micropeptides, which suggests that there may be novel synaptic proteoforms in the brain that await discovery. Here we report widespread learning-induced local circRNA translation in the prefrontal cortex of male C57BL/6 mice. More than 1500 synapse-enriched circRNAs contain active IRES elements, with 842 interacting with the ribosome and 241 exhibiting direct evidence of activity-induced translation. We discovered a synapse-enriched micropeptide (P1) that is derived from a single exon circRNA, the mRNA host of which encodes an enzyme associated with protein repair. Although P1 is only a third of the size of the full-length protein, it is locally expressed, enzymatically active, and interacts with plasticity-related proteins, including CaMKII. In addition to direct effects on synaptic activity, targeted P1 knockdown impairs whereas its overexpression enhances fear extinction memory. These findings shed new light on the dark proteome in the brain and reveal local circRNA translation as a novel mechanism of plasticity and memory.

neuroscience↗

bioRxiv2025

Accurate Prediction of ecDNA in Interphase Cancer Cells using Deep Neural Networks

Oncogene amplification is a key driver of cancer pathogenesis and is often mediated by extrachromosomal DNA (ecDNA). EcDNA amplifications are associated with increased pathogenicity of cancer and poorer outcomes for patients. EcDNA can be detected accurately using fluorescence in situ hybridization (FISH) when cells are arrested in metaphase. However, the majority of cancer cells are non-mitotic and must be analyzed in interphase, where it is difficult to discern extrachromosomal amplifications from chromosomal amplifications. Thus, there is a need for methods that accurately predict oncogene amplification status from interphase cells. Here, we present interSeg, a deep learning-based tool to cytogenetically determine the amplification status as EC-amp, HSR-amp, or not amplified from interphase FISH images. We trained and validated interSeg on 652 images (40,446 nuclei). Tests on 215 cultured cell and tissue model images (9,733 nuclei) showed 89% and 97% accuracy at the nuclear and sample levels, respectively. The neuroblastoma patient tissue hold-out set (67 samples and 1,937 nuclei) also revealed 97% accuracy at the sample level in detecting the presence of focal amplification. In experimentally and computationally mixed images, interSeg accurately predicted the level of heterogeneity. The results showcase interSeg as an important method for analyzing oncogene amplifications.

cancer biology↗

bioRxiv2025

Oncogene Silencing via ecDNA Micronucleation

Extrachromosomal DNA (ecDNA) contributes to cancer genome instability by enabling high-copy oncogene amplification, intratumoural heterogeneity and rapid genetic change. Micronuclei (MN) are frequently observed in chromosomally unstable cancers, yet their origins and relevance in ecDNA-driven tumours remain incompletely understood. Here, we investigate the relationship between ecDNA segregation errors and MN formation. We find that ecDNA frequently localizes to MN and represents a prominent source of MN content in ecDNA-positive cancer cells. Mitotic clustering of oncogene-bearing ecDNAs is associated with asymmetric inheritance and mis-segregation into MN. Transfer of oncogenes from ecDNA into MN is accompanied by reduced transcriptional output. Single-MN sequencing shows that individual MN are enriched for ecDNA to a degree that exceeds expectations from stochastic mis-segregation and that in MN oncogenes originating from multiple distinct ecDNAs coalesce. Using live-cell imaging, we observe that cells inheriting ecDNA-positive MN show limited proliferative capacity and an increased likelihood of cell death. In neuroblastoma patients with MYCN-amplified ecDNA, higher frequencies of ecDNA-positive MN at diagnosis are associated with improved event-free and overall survival. Together, these findings link ecDNA mis-segregation to MN formation and reduced cellular fitness, suggesting that ecDNA-positive MN may reflect a state of impaired oncogenic ecDNA function with potential relevance for clinical outcome.

cancer biology↗

bioRxiv2025

Single-cell multiome and enhancer connectome of human retinal pigment epithelium and choroid nominate pathogenic variants in age-related macular degeneration

Age-related macular degeneration (AMD) is a leading cause of vision loss worldwide. Genome-wide association studies (GWAS) of AMD have identified dozens of risk loci that may house disease targets. However, variants at these loci are largely noncoding, making it difficult to assess their function and whether they are causal. Here, we present a single-cell gene expression and chromatin accessibility atlas of human retinal pigment epithelium (RPE) and choroid to systematically analyze both coding and noncoding variants implicated in AMD. We employ HiChIP and Activity-by-Contact modeling to map enhancers in these tissues and predict cell and gene targets of risk variants. We further perform allele-specific self-transcribing active regulatory region sequencing (STARR-seq) to functionally test variant activity in RPE cells, including in the context of complement activation. Our work nominates new pathogenic variants and mechanisms in AMD and offers a rich and accessible resource for studying diseases of the RPE and choroid.

genomics↗

bioRxiv2025

Autoantibody hotspots reveal origin and impact of immunogenic XIST ribonucleoprotein complex

Four out of five patients with autoimmune diseases are women. The XIST ribonucleoprotein (RNP) complex, comprising the female-specific long noncoding RNA XIST and over 100 associated proteins, may drive several autoimmune diseases that disproportionately affect women, who have elevated levels of autoantibodies against the XIST RNP. However, the structural distribution, potential origin, and clinical significance of XIST RNP autoantibodies remained unexplored. Here, we find that XIST RNP is associated with autoantigens associated with six female-biased autoimmune conditions. Mapping autoantibody targets to their occupancy sites on XIST shows that these autoantigens are concentrated at discrete "hotspots" along the XIST lncRNA, notably the A-repeat. Cell type-specific protein expression data nominated neutrophils as a predominant source of hotspot antigens, and we confirmed the presence of both XIST and hotspot antigens in neutrophil extracellular traps during NETosis, an immunogenic programmed cell death pathway triggered by neutrophil activation upon which neutrophils extrude their nuclear content. Furthermore, we found that levels of autoantibodies against a top hotspot antigen, SPEN, that binds the A-repeat, correlate with severe digital ischemia in systemic sclerosis in two independent cohorts. Together, these data show a plausible mechanism for the origin of AXA, guided by RNA structure and RNA-protein interactions, and show that antibodies to XIST RNP holds promise for disease endotyping and prognostication in female-biased autoimmune conditions. One Sentence SummaryNovel autoantibodies target hotspots on XIST ribonucleoprotein complex in female-biased autoimmune diseases.

immunology↗

bioRxiv2024

Extrachromosomal DNA driven oncogene spatial heterogeneity and evolution in glioblastoma

Oncogene amplification on extrachromosomal DNA (ecDNA) is strongly associated with treatment resistance and shorter survival for patients with cancer, including patients with glioblastoma. The non-chromosomal inheritance of ecDNA during cell division is a major contributor to intratumoral genetic heterogeneity. At present, the spatial dynamics of ecDNA, and the impact on tumor evolutionary trajectories, are not well understood. Here, we investigate the spatial-temporal evolution of ecDNA and its clinical impact by analyzing tumor samples from 94 treatment-naive human IDH-wildtype glioblastoma patients. We developed a spatial-temporal computational model of ecDNA positive tumors ( SPECIES) that integrates whole-genome sequencing, multi-region DNA FISH, and nascent RNAscope, to provide unique insight into the spatial dynamics of ecDNA evolution. Random segregation in combination with positive selection of ecDNAs induce large, predictable spatial patterns of cell-to-cell ecDNA copy number variation that are highly dependent on the oncogene encoded on the circular DNA. EGFR ecDNAs often reach high mean copy number (mean of 50 copies per tumor cell), are under strong positive selection (mean selection coefficient, s > 2) and do not co-amplify other oncogenes on the same ecDNA particles. In contrast, PDGFRA ecDNAs have lower mean copy number (mean of 15 copies per cell), are under weaker positive selection and frequently co-amplify other oncogenes on the same ecDNA. Evolutionary modeling suggests that EGFR ecDNAs often accumulate prior to clonal expansion. EGFR structural variants, including vIII and c-terminal deletions are under strong positive selection, are found exclusively on ecDNA, and are intermixed with wild-type EGFR ecDNAs. Simulations show EGFRvIII ecDNA likely arises after ecDNA formation in a cell with high wild-type EGFR copy number (> 10) before the onset of the most recent clonal expansion. This remains true even in cases of co-selection and co-amplification of multiple oncogenic ecDNA species in a subset of patients. Overall, our results suggest a potential time window in which early ecDNA detection may provide an opportunity for more effective intervention. HighlightsO_LIecDNA is the most common mechanism of focal oncogene amplification in IDHwt glioblastoma. C_LIO_LIEGFR and its variants on ecDNA are particularly potent, likely arising early in tumor development, providing a strong oncogenic stimulus to drive tumorigenesis. C_LIO_LIWild-type and variant EGFR ecDNA heteroplasmy (co-occurrence) is common with EGFRvIII or c-terminal deletions being derived from EGFR wild-type ecDNA prior to the most recent clonal expansion. C_LIO_LITumors with ecDNA amplified EGFR versus PDGFRA exhibit different evolutionary trajectories. C_LIO_LISPECIES model can infer spatial evolutionary dynamics of ecDNA in cancer. C_LIO_LIA delay between ecDNA accumulation and subsequent oncogenic mutation may give a therapeutic window for early intervention. C_LI

cancer biology↗

bioRxiv2024

Spatiotemporal lineage tracing reveals the dynamic spatial architecture of tumor growth and metastasis

Tumour progression is driven by dynamic interactions between cancer cells and their surrounding microenvironment. Investigating the spatiotemporal evolution of tumours can provide crucial insights into how intrinsic changes within cancer cells and extrinsic alterations in the microenvironment cooperate to drive different stages of tumour progression. Here, we integrate high-resolution spatial transcriptomics and evolving lineage tracing technologies to elucidate how tumour expansion, plasticity, and metastasis co-evolve with microenvironmental remodelling in a Kras;p53-driven mouse model of lung adenocarcinoma. We find that rapid subclonal expansion contributes to a hypoxic, immunosuppressive, and fibrotic microenvironment that is associated with the emergence of pro-metastatic cancer cell states. Furthermore, metastases arise from spatially-confined subclones of primary tumours and remodel the distant metastatic niche into a fibrotic, collagen-rich microenvironment. Together, we present a comprehensive dataset integrating spatial assays and lineage tracing to elucidate how sequential changes in cancer cell state and microenvironmental structures cooperate to promote tumour progression.

cancer biology↗

bioRxiv2024

A STAG2-PAXIP1/PAGR1 axis suppresses lung tumorigenesis

The cohesin complex is a critical regulator of gene expression. STAG2 is the most frequently mutated cohesin subunit across several cancer types and is a key tumor suppressor in lung cancer. Here, we coupled somatic CRISPR-Cas9 genome editing and tumor barcoding with an autochthonous oncogenic KRAS-driven lung cancer model and show that STAG2 is uniquely tumor suppressive among all core and auxiliary cohesin components. The heterodimeric complex components PAXIP1 and PAGR1 have highly correlated effects with STAG2 in human lung cancer cell lines, are tumor suppressors in vivo, and are epistatic to STAG2 in oncogenic KRAS-driven lung tumorigenesis in vivo. STAG2 inactivation elicits changes in gene expression, chromatin accessibility and 3D genome conformation that impact cancer cell state. Gene expression and chromatin accessibility similarities between STAG2- and PAXIP1-deficient neoplastic cells further relates STAG2-cohesin to PAXIP1/PAGR1. These findings reveal a STAG2-PAXIP1/PAGR1 tumor-suppressive axis and uncover novel PAXIP1-dependent and PAXIP1-independent STAG2-cohesin mediated mechanisms of lung tumor suppression. SUMMARYSTAG2 is a frequently mutated cohesin subunit across several cancers and one of the most important functional suppressors of lung adenocarcinoma. Our findings underscore important roles of STAG2 in suppressing lung tumorigenesis and highlight a STAG2-PAXIP1/PAGR1 tumor-suppressive program that may transcend cancer type.

cancer biology↗

bioRxiv2024

Enhancer activation from transposable elements in extrachromosomal DNA

Extrachromosomal DNA (ecDNA) drives oncogene amplification and intratumoral heterogeneity in aggressive cancers. While transposable element (TE) reactivation is common in cancer, its role on ecDNA remains unexplored. Here, we map the 3D architecture of MYC-amplified ecDNA in colorectal cancer cells and identify 68 ecDNA-interacting elements (EIEs)--genomic loci enriched for TEs that are frequently integrated onto ecDNA. We focus on an L1M4a1#LINE/L1 fragment co-amplified with MYC, which functions only in the ecDNA amplified context. Using CRISPR-CATCH, CRISPR interference, and reporter assays, we confirm its presence on ecDNA, enhancer activity, and essentiality for cancer cell fitness. These findings reveal that repetitive elements can be reactivated and co-opted as functional rather than inactive sequences on ecDNA, potentially driving oncogene expression and tumor evolution. Our study uncovers a mechanism by which ecDNA harnesses repetitive elements to shape cancer phenotypes, with implications for diagnosis and therapy.

genetics↗

bioRxiv2024

Deciphering regulation of FOXP3 expression in human conventional T cells

FOXP3 is a lineage-defining transcription factor (TF) for immune-suppressive regulatory T cells (Tregs). While mice exclusively express FOXP3 in Tregs, humans also transiently express FOXP3 in stimulated conventional CD4+ T cells (Tconvs). Mechanisms governing these distinct expression patterns remain unknown. Here, we performed CRISPR screens tiling the FOXP3 locus and targeting TFs in human Tregs and Tconvs to discover cis-regulatory elements (CREs) and trans-regulators of FOXP3. Tconv FOXP3 expression depended on a subset of Treg CREs and Tconv-selective positive (TcNS+) and negative (TcNS-) CREs. The CREs are occupied and regulated by TFs we identified as critical regulators of FOXP3. Finally, mutagenesis of murine TcNS- revealed that it is critical for restriction of FOXP3 expression to Tregs. We discover CRE and TF circuitry controlling FOXP3 expression and reveal evolution of mechanisms regulating a gene indispensable to immune homeostasis. HighlightsO_LIComprehensive CRISPR maps of CREs and TFs controlling FOXP3 in human Tregs and Tconvs C_LIO_LIKey TFs that control FOXP3 directly occupy and regulate CREs forming TF-CRE circuits C_LIO_LIA previously unknown negative CRE stringently restricts FOXP3 to Tregs in mice C_LI

immunology↗

bioRxiv2024

Rampant transcription replication conflict creates therapeutic vulnerability in extrachromosomal DNA containing cancers

Extrachromosomal DNA (ecDNA) presents a major challenge for precision medicine, contributing to poor survival for patients with oncogene-amplified tumours. EcDNA renders tumours resistant to targeted treatments by facilitating massive transcription of oncogenes and rapid genome evolution. At present, there are no ecDNA- specific treatments. Here we show that enhancing transcription replication conflict enables targeted elimination of ecDNA-containing cancers, exposing an actionable vulnerability. Stepwise analyses of ecDNA transcription reveal landscapes of pervasive RNA transcription and associated single-stranded DNA, leading to excessive transcription replication conflicts and replication stress (RS) compared to chromosomal loci. Nucleotide incorporation onto growing DNA strands is markedly slower on ecDNA, and RS is significantly higher in ecDNA-containing tumours regardless of cancer type or oncogene cargo. Replication Protein A2 phosphorylated on serine 33, a mediator of DNA damage repair that binds single-stranded DNA, shows elevated localization on ecDNA in a transcription dependent manner, along with increased DNA double strand breaks, and activation of the S-phase checkpoint kinase, CHK1. Genetic or pharmacological CHK1 inhibition abrogates the DNA replication check point, causing extensive and preferential tumour cell death in ecDNA-containing tumours as they enter S-phase. To exploit this vulnerability, we develop a highly selective, potent, and bioavailable oral CHK1 inhibitor, BBI-2779, and demonstrate that it preferentially kills ecDNA-containing tumour cells. In a gastric cancer model containing FGFR2 on ecDNA, BBI-2779, suppresses tumour growth and prevents ecDNA-mediated acquired resistance to the pan-FGFR inhibitor infigratinib, resulting in potent and sustained tumour regression in mice. These results reveal transcription-replication conflict as an ecDNA-generated vulnerability that can be targeted as an ecDNA-directed therapy and suggest that synthetic lethality of excess can be exploited as a strategy for treating cancer.

cancer biology↗