BAP1 loss and PRAME expression converge to remodel the tumor-immune ecosystem during uveal melanoma progression

Dollar, J. J.; Liu, X.; Singh, M.; Sorouri, M.; Kuznetsoff, J. N.; Salazar, S.; Cruz, A.; Decatur, C. L.; Correa, Z. M.; Harbour, J. W.

doi:10.64898/2026.09.23.752454

bioRxiv · 10.64898/2026.09.23.752454

BAP1 loss and PRAME expression converge to remodel the tumor-immune ecosystem during uveal melanoma progression

Dollar, J. J.·Liu, X.·Singh, M.·Sorouri, M.·Kuznetsoff, J. N.·Salazar, S.·Cruz, A.·Decatur, C. L.·Correa, Z. M.·Harbour, J. W.

View original source ↗Export metadata ↓

Abstract

Uveal melanoma (UM) is characterized by a small number of recurrent genetic alterations that determine metastatic propensity. BAP1 loss and PRAME expression define the dominant prognostic axes in UM, yet how they promote malignant progression remains unclear. We profiled 190,535 cells from normal uvea, uveal nevus, primary and metastatic UM using single-cell transcriptomics, T cell receptor sequencing, spatial transcriptomics and isogenic perturbation models. Normal melanocytes, nevus cells and UM cells formed a transcriptional continuum marked by loss of differentiation and emergence of neural crest-like, stress-responsive, hypoxic-glycolytic and immune-interacting states. BAP1 loss and PRAME expression imposed distinct but convergent immunoregulatory programs, inducing interferon and TNF-NFkB signaling and MHC-I expression, with HLA-E showing the strongest response. These alterations were accompanied by macrophage and CD8+ T cell remodeling. PRAME-enriched tumor regions formed spatially organized niches enriched for macrophages and plasma cells. These findings define BAP1 loss and PRAME expression as distinct but convergent axes of tumor-immune coevolution and nominate HLA-E as a candidate mediator of immune resistance.

Explore related subjects

cancer biology Preprint

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Dollar, J. J., Liu, X., Singh, M., Sorouri, M., Kuznetsoff, J. N., Salazar, S., Cruz, A., Decatur, C. L., Correa, Z. M., Harbour, J. W.. 2026-09-28. BAP1 loss and PRAME expression converge to remodel the tumor-immune ecosystem during uveal melanoma progression. https://doi.org/10.64898/2026.09.23.752454

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Tissue resident CD4+ memory T-cells mark response to immune checkpoint inhibition in high-grade glioma

Background: Immune checkpoint inhibitors (ICI) are efficacious in many solid tumors, but response in glioma is restricted to a small subgroup. The determinants of response and resistance to ICI remain poorly understood. Methods: Here we exploit a syngeneic hypermutated high-grade glioma model with dichotomous response to combined PD-1 and CTLA-4 inhibition to unravel determinants of tumor-infiltrating T-cells driving response. Tumor-infiltrating T-cells from ICI-responsive and non-responsive tumors were analyzed by single-cell RNA and T-cell receptor sequencing and tumor-reactive T-cell receptor clonotypes were functionally validated to characterize their transcriptional phenotypes. We verify our findings in IDH1 wildtype glioblastoma patients treated with neoadjuvant pembrolizumab. Results: ICI response was associated with intratumoral clonal expansion of tumor-reactive cytotoxic T-cells and increased infiltration of CXCR6+ CD4+ tissue resident memory T-cells (Trm). CD4 stem-like memory T-cells in responding tumors demonstrated elevated interferon responses, following trajectories toward clonally expanded Trm, versus trajectories toward exhaustion in non-responsive tumors. In responsive tumors, CD4+ Trm interacted with infiltrating CXCR3+ tumor-reactive and clonally expanded, yet transcriptionally versatile cytotoxic T-cells. Probing the post neoadjuvant ICI high-grade glioma patient tissue dataset, we confirmed increased CXCR6 expression in CD4+ T cells and the association of CD4+ Trm with prolonged overall survival. Conclusion: These findings identify CD4 tissue-resident memory T-cells as determinants of ICI response in IDH1 wildtype high-grade glioma and warrant their further investigation to improve immunotherapy outcomes.

cancer biology↗

bioRxiv2026

Circadian gene-network distortion in high-risk neuroblastoma across multiple biological reference contexts

Background: The circadian clock regulates cellular homeostasis, and its disruption has been implicated in aggressive neuroblastoma, particularly in tumours harbouring MYCN- amplification. However, it remains unclear whether alterations are restricted to individual clock genes or extend to circadian gene network coordination. We therefore examined circadian clock network disruption in adverse neuroblastoma across multiple biological contexts. Methods: We estimated circadian gene network dysregulation using Delta-CCD in tumours from two neuroblastoma cohorts (SEQC n=498 and Kocak n=649), comparing clinical features associated with outcome across canonical, adrenal-tissue matched, and developmental references. Robustness was assessed by cross-cohort meta-analysis and leave-one-gene-out analyses. Cox proportional hazards models adjusted for clinical covariates assessed association between individual clock gene expression patient outcome. Results: Delta-CCD was highest in tumours classified as high-risk (study-specific definition) across reference contexts in both cohorts. MYCN-amplified tumours showed a more reference-dependent pattern, strongest in adrenal context, while stage 4 tumours showed a similar but weaker pattern. Additional analyses supported the high-risk signal as a distributed network-level alteration rather than a single-gene phenomenon. Conclusions: High-risk neuroblastoma is characterised by robust disruption of coordinated clock gene network organisation across canonical and tissue-matched references, extending beyond individual clock genes. The extent of circadian dysregulation depends on the reference state used.

cancer biology↗

bioRxiv2026

A plasma metabolomics workflow for breast cancer detection using quantitative GC/MS and machine learning

Blood-based metabolomic profiling has been widely investigated for breast cancer (BC) detection; however, clinical implementation remains limited due to variability in sample handling, analytical reproducibility, and overfitting during statistical analysis. We established a plasma GC/MS metabolomics workflow for discriminating BC from healthy controls (HC) using conventional machine-learning algorithms. Plasma samples (n = 360; BC = 180, HC = 180) were collected prospectively under standardized preanalytical conditions before surgery and the initiation of systematic anticancer therapy and analyzed using a quantitative GC/MS platform with automated derivatization. Feature selection and model development were conducted using three machine-learning (ML) algorithms (Lasso logistic regression (LR), random forest classifier (RFC), and support vector machine (SVM)). A total of 45 metabolite candidate biomarkers were identified, and the optimal number of metabolite features for each algorithm was estimated by a recursive feature elimination (RFE)-based strategy. The best-performing models achieved area under the ROC curve values (AUC) of 0.910 (LR), 0.893 (RFC), and 0.843 (SVM). We selected prioritizing candidate biomarkers consistently expressed across the multi-algorithm pipeline. A bagging ensemble model improved stability (AUC = 0.911) and reduced false-positive predictions in the independent HC dataset. In addition, model stability with respect to false-positive predictions was assessed using an independent HC cohort (n = 15) that was collected at a separate institution. These results indicate that a plasma metabolomics workflow combined with conventional multi-algorithm ML, algorithm-specific feature selection, and independent assessment provides stable discrimination between BC and HC in a moderately sized cohort.

cancer biology↗