Abstract Extracellular vesicles (EVs) are nanosized, membrane-bound particles released by all cell types. They carry proteins, nucleic acids, and lipids reflective of their cellular origin and have emerged as promising non-invasive biomarkers for cancer diagnosis and monitoring of therapy response. However, the clinical translation of EV-based assays remains limited by heterogeneous isolation methods, a lack of standardization of the clinical workflows, and insufficient validation in large patient cohorts. To address these challenges, our group at the National Center for Tumor Diseases in Heidelberg, Germany, has developed an EV profiling framework compliant with MISEV2023 recommendations. We systematically benchmark isolation and pre-processing procedures to ensure reproducibility and compatibility with high-throughput liquid biopsy workflows within the prospective EValuate study (S-773/2021). Blood samples are collected via the NCT Cell and Liquid Biobank, processed within 30 minutes, and stored as serum and plasma aliquots at -80 °C for longitudinal analyses. To enable large-cohort EV analyses, we also characterized a miniaturized size-exclusion chromatography protocol using low-volume serum or plasma, which requires no specialized equipment and complements conventional differential centrifugation workflows. To demonstrate the feasibility of the pipeline, we collected and analyzed a total of 125 serum samples - 109 from 24 patients with hepatocellular carcinoma (HCC) undergoing immune checkpoint inhibitor therapy and 16 quality-control samples from four healthy donors. EVs were isolated and characterized by transmission electron microscopy, nanoparticle tracking analysis, Western blotting, quantitative protein assays, and subsequently profiled proteomically to identify EV-derived protein signatures associated with disease course, radiological treatment response (RECIST), and survival. This standardized, high-throughput EV workflow bridges biobanking, analytical validation, and clinical correlation, providing a robust and scalable framework for integrating EV-based liquid biopsy assays into precision oncology. Citation Format: Antonia Schubert, Nadine Winkler, Robert Ihnatko, Joscha Kraske, Sunanjay Bajaj, Michelle Neßling, Karsten Richter, Dirk Jäger, Guy Ungerechts, Oliver Sedlaczek, Jeroen Krijgsveld, Thomas Walle, Michael Boutros. Toward large-scale clinical implementation of extracellular vesicle profiling for precision oncology [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3340.

In a recent publication, Lapitz et al. reported differences in serum levels that predict the development of cholangiocarcinoma (CCA) in patients with primary sclerosing cholangitis prior to clinical manifestation. We examined whether these biomarkers also predict the risk of gallbladder cancer (GBC) in European prospective plasma samples from 24 GBC cases and 90 control individuals. After logarithmic transformation and quantile normalisation of individual protein levels measured with a timsTOF PRO mass spectrometer, we fitted univariate logistic regression models and applied backward model selection to identify the optimal model for GBC risk prediction. CRP and MASP2, previously reported markers of CCA, were found to be predictive for GBC risk as well (p value < 0.05), and complemented by age at blood sampling, provided an area under the receiver operating characteristic curve of 0.80 (95% confidence interval 0.69-0.92) when combined in a prediction model for GBC risk. We further examined the mRNA expression of CRP and MASP2 in serum samples from 82 GBC cases and 79 control subjects from Chile. CRP mRNA levels were elevated in Chilean GBC cases, but MASP2 showed an opposite trend. While there are considerable differences in the design of the discovery study and ours, the finding that circulating levels of CRP and MASP2 are associated with the risk of both CCA and GBC in Europeans highlights the need for further research into potential shared mechanisms and strategies for the prevention of these two aggressive biliary tumours.

p53 plays a central role in the DNA damage response, inducing repair, cell-cycle arrest or apoptosis. Its loss is associated with replication stress and genomic instability. While several underlying mechanisms were suggested, the primary triggers of catastrophic genomic events like chromothripsis, a known driver of tumorigenesis linked with p53 loss, are still unclear. Using p53-depleted epithelial cells and fibroblasts, as well as patient-derived fibroblasts with germline p53 variants that spontaneously undergo chromothripsis, we found that p53 loss causes hypertranscription and increased nucleotide consumption. The resulting nucleotide shortage induces replication stress, causing telomere dysfunction, micronuclei formation, and chromothripsis. These effects were rescued by nucleoside supplementation or normalization of transcription levels, demonstrating a causal link between transcriptional activity, nucleotide availability, and genome stability. Emerging chromothriptic clones displayed restored DNA replication, telomere stabilization, and extrachromosomal DNA, suggesting key features that support clonal selection. We identify nucleotide pool homeostasis as a critical p53 function that suppresses replication stress, prevents chromothripsis, and protects against early tumorigenesis.

SnoRNAs are highly expressed in AML and have implications in leukemogenesis and leukemic maintenance. SnoRNAs can be further processed into snoRNA-derived RNAs (sdRNAs). The role of sdRNAs in AML and healthy hematopoiesis remains largely elusive. We characterized sdRNA and snoRNA levels in hematopoietic stem and progenitor cells (HSPCs), healthy WBCs, and 159 intensively treated AML patient samples at initial diagnosis. HSPCs, healthy WBCs, and AML blasts could be differentiated by their sdRNA expression pattern in a cell-type-specific manner. In AML, high sd3’-RNA/snoRNA-host gene ratios were associated with an inverse patient outcome. Particularly, in NPM1-mutated patients with favorable risk stratification and good initial therapy response, high sd3’-RNA ratios identified a subgroup with inferior outcome. High sd3’-RNA ratios were associated with altered oncogenic, inflammatory, and immune response signaling. Forced expression of single sdRNAs, such as sd3’-SNORD78, sd3’-SNORD76, and sd5’-SNORD93, enhanced clonogenic potential in AML and drove sdRNA-specific gene expression signatures in both AML and healthy HSPCs. Exemplarily, we propose and characterize NUDT21, an important regulator of alternative polyadenylation and oncogenic gene expression, as a downstream target of sd3’-SNORD78 in AML. Our data introduce sdRNAs as standalone regulatory effector molecules in healthy hematopoiesis and AML.

The human cortex acquires its advanced cognitive capacity through tightly regulated developmental programs, disruption of which underlies neurodevelopmental disorders such as Schaaf-Yang syndrome (SYS) and Prader-Willi syndrome (PWS). While SYS results from pathogenic variants in the imprinted gene MAGEL2, PWS arises from chromosomal deletions, imprinting defects or uniparental disomy encompassing the MAGEL2 locus. However, the contribution of MAGEL2 to disease pathogenesis and human corticogenesis is not fully understood. Here, we performed integrated transcriptomic, proteomic, and ubiquitinomic profiling of cortical neurons derived from CRISPR/Cas9-engineered isogenic human pluripotent stem cells (hiPSC) modeling SYS and PWS. Beyond PWS-specific signatures including dysregulated ribosomal processes, we identified MAGEL2-dependent defects shared across both disorders. These include reduced progenitor proliferation, accelerated neuronal maturation, impaired migration and adhesion, as well as abnormal synaptic development, collectively linking PWS and SYS at the level of cortical development. Notably, these phenotypes partially overlap with those observed in other neurodevelopmental disorders, suggesting that MAGEL2 governs core pathways broadly vulnerable in disease. Together, our findings establish MAGEL2 as a key regulator of human cortical development, provide a unifying mechanistic framework for SYS and PWS, accessible via a web-based platform.

Abstract The vascular endothelium forms a systemically disseminated organ that translates micromilieu factors into instructive cues, designated as angiocrine signalling. While vascular control of organ function mechanisms have been discovered in essentially all major organs, the mechanisms of translating milieu factors into angiocrine signalling mechanisms remain largely elusive. Here, focussing on the well-established angiocrine Wnt signalling-regulated liver metabolic zonation paradigm, we identified blood flow-induced haemodynamic stress as a biophysical sensor that governs the angiocrine expression profile of liver sinusoidal endothelial cells (LSEC). Combining single-cell RNA sequencing with spatial proteomics, we generated a high-resolution crosstalk map of LSEC and hepatocytes from LSEC Wnt-deficient mutant mice and littermate controls. Intriguingly, vascular Wnt receptors, in parallel with angiocrine Wnt ligands, were specifically enriched in pericentral LSEC, enforcing the spatially coordinated vascular Wnt functions. Consequently, vascular Wnt further modulated angiocrine gene signatures, LSEC morphology, and the expression of gap junctional molecules in an autocrine manner. Taken together, the data define LSEC as a dynamic cellular decoder that translates biophysical forces into instructive angiocrine signals via activating promiscuous vascular Wnt factors.

SnoRNAs are highly expressed in AML and play a role in leukemogenesis and leukemic maintenance. SnoRNAs can be further processed into snoRNA-derived RNAs (sdRNAs). Expression and implications of sdRNAs in AML and healthy hematopoiesis, however, remain largely elusive. We characterized sdRNA and snoRNA levels in hematopoietic stem cells (HSCs), healthy peripheral blood cells, and 159 AML patient samples at initial diagnosis. HSCs, healthy WBCs and AML blasts could be differentiated by their sdRNA expression pattern in a cell-type specific manner. In AML, high sd3’-RNA/snoRNA-hostgene ratios were associated with inverse patient outcome. Particularly, in NPM1-mutated patients with favorable risk stratification and good initial therapy response, high sd3’-RNA ratios identified a subgroup with inferior outcome, and could therefore represent biomarkers to identify those at-risk patients. High sd3’-RNA ratios were associated with clear alterations in oncogenic, inflammatory and immune response signalling. Forced expression of single sdRNAs, such as sd3’-SNORD78 and sd5’-SNORD93, enhanced clonogenic potential in AML. Total proteome and transcriptome analyses suggested NUDT21, a reported tumor suppressor with implications in inflammatory and immune response signalling, as novel target of sd3’-SNORD78. Our data introduces sdRNAs as effector molecules in healthy hematopoiesis and AML with mechanistic, diagnostic, as well as potential prognostic and therapeutic implications.

Abstract Cancer cells can evade immune surveillance by triggering inhibitory checkpoint responses in tumor-associated T cells through the expression of immunosuppressive surface molecules. While therapeutic blockade of such receptors has emerged as a pillar of cancer therapy, tumor cell-intrinsic mechanisms controlling their expression remain incompletely understood. Fluorescence-activated cell sorting (FACS)-based genetic screens can be used to decipher regulatory pathways, but conventional screening approaches are biased towards regulators that are dispensable for cancer cell proliferation and survival. Here, we used a tetracycline-inducible Cas9 system enabling fully time-controllable CRISPR-based mutagenesis to gain a more comprehensive and comparative survey of regulators controlling the expression of four major immunosuppressive surface molecules, PD-L1 (CD274), CD47, CD276 and HLA-E, as well as CD151, a candidate surface target associated with tumor growth and invasion. As a prominent hit, our screens identify the membrane-trafficking factor DNAJC13 as a regulator of PD-L1 and CD276. Among DNAJC13-controlled surface proteins, we identify other known and proposed immune-checkpoint molecules. Based on this function, suppression of DNAJC13 strongly increases the sensitivity of human cancer cells to T-cell attack in vitro and prolongs survival of mice bearing pancreatic tumors. Together, our study establishes regulatory maps of major immune-modulatory surface molecules and identifies DNAJC13 as a potential target for the coordinated inhibition of multiple immunosuppressive signals.

Abstract Background AXIN1 is a central regulatory hub of many oncogenic pathways in colorectal cancer (CRC). As the main scaffold protein and least abundant component of the beta-catenin destruction complex, changes in AXIN1 levels tightly control Wnt signaling activity. How other cancer pathways beyond Wnt signaling regulate cellular AXIN1 levels is incompletely understood. Methods Colorectal cancer cell lines, murine and patient-derived intestinal and cancer organoids were used as model systems. Changes in AXIN1 levels upon drug perturbation were profiled by immunoblot, qPCR and RNA-seq. Ubiquitin-affinity immunoprecipitation assays and mass spectrometry were used to determine mechanisms of AXIN1 loss. To characterize effects on protein synthesis, we performed polysome and ribosome profiling (Ribo-seq). Results We show that targeting the Ras-MAPK pathway using clinically approved MEK1/2 inhibitors induces AXIN1 loss across a panel of CRC cell lines and patient-derived organoids. In contrast to GSK3 inhibitors, MEK1/2 inhibition neither affects protein stability nor post-translational modifications of AXIN1 and only caused a minor reduction of AXIN1 transcript levels. Co-treatment with tankyrase inhibitors could partially prevent loss of AXIN1 upon MEK1/2 inhibition. Using isogenic CRC cell lines and murine intestinal organoids, we show that APC truncations strongly reduce basal cellular AXIN1 levels, but do not alter dynamics of AXIN1 loss after MEK1/2 inhibition. Polysome profiling and Ribo-seq revealed that MEK1/2 inhibitors reduce global protein synthesis via an mTOR associated pathway. This translational repression is sufficient to cause significant AXIN1 loss, as treatment with mTOR or S6K inhibitors phenocopies the effect of MEK1/2 inhibitors. Conclusion Our study demonstrates that AXIN1 protein homeostasis is critically controlled by Ras-MAPK signaling at the level of protein synthesis, and that MEK1/2 inhibitors cause AXIN1 loss by global translational repression.

A significant obstacle in treating brain tumors is the limited drug penetration across the blood-brain barrier (BBB), characterized by an interplay of endothelial tight junctions and efflux pumps. Brain tumors can alter BBB characteristics; however, there is limited understanding in ependymoma (EPN), the third most common pediatric brain tumor. To this end, we characterized EPN tumor (n = 364) and healthy brain tissues (n = 225) at RNA level and identified a distinct EPN group-specific BBB transcriptional pattern. Analyses of public datasets from Aubin and Gojo as well as a validation single-cell dataset (n = 8) could further specify a novel BBB signature expressed in an endothelial subpopulation. Clinically relevant drugs (n = 3) that were effective against EPN in vitro were further evaluated for BBB penetration in our subtype-specific patient-derived xenograft (PDX) models. Idasanutlin showed overall low brain-to-plasma ratios, while the P-glycoprotein (PGP) substrates temsirolimus and etoposide accumulated slightly more in zinc finger translocation associated (ZFTA)-fusion positive EPN than in PFA tumors and adjacent brain. These differences align with modestly lower PGP levels in ZFTA PDX, although expression does not necessarily reflect transporter activity and was not consistently observed in patient tumors. Despite these differences, all tested drugs remained below their effective in vitro levels. In summary, multi-omics analyses of BBB characteristics improve the understanding of drug penetrance and may potentially guide treatment choices in the context of molecular EPN groups within upcoming clinical trials.

The development of advanced cell culture models has overcome the limitations of conventional monolayer cultures, still doubts remain about the reliability of data obtained using traditional systems, as well as the comparability of results from different models. This is highly relevant to preclinical drug analysis, where in vitro studies determine the fate of molecules. We investigated the molecular mechanisms that regulate the activity of cannabidiolic acid (CBDA) in a 3D model of a glioblastoma cell line, comparing the results with those obtained on a conventional monolayer model of the same cells. CBDA targeted the translation initiation factor EIF2A in both tested models. However, the downstream consequences of the CBDA-EIF2A interaction differed between cells cultured in 3D and 2D, since the biological functions and interactome of EIF2A were found to change significantly. Overall, this study sheds new light on the difficulties associated with comparing results obtained using different in vitro models.

Advancing MS-based proteomics toward clinical applications evolves around developing standardized start-to-finish and fit-for-purpose workflows for clinical specimens. Steps along the method design involve the determination and optimization of several bioanalytical parameters such as selectivity, sensitivity, accuracy, and precision. In a joint effort, eight proteomics laboratories belonging to the MSCoreSys initiative including the CLINSPECT-M, MSTARS, DIASyM, and SMART-CARE consortia performed a longitudinal round-robin study to assess the analysis performance of plasma and serum as clinically relevant samples. A variety of LC-MS/MS setups including mass spectrometer models from ThermoFisher and Bruker as well as LC systems from ThermoFisher, Evosep, and Waters Corporation were used in this study. As key performance indicators, sensitivity, precision, and reproducibility were monitored over time. Protein identifications range between 300 and 400 IDs across different state-of-the-art MS instruments, with timsTOF Pro, Orbitrap Exploris 480, and Q Exactive HF-X being among the top performers. Overall, 71 proteins are reproducibly detectable in all setups in both serum and plasma samples, and 22 of these proteins are FDA-approved biomarkers, which are reproducibly quantified (CV < 20% with label-free quantification). In total, the round-robin study highlights a promising baseline for bringing MS-based measurements of serum and plasma samples closer to clinical utility.

Supplementary Figure S1 shows enrichment of FBL in LSC, correlation of FBL with LSC signature and rRNA 2'-O-Me in healthy hematopoietic cells. Supplementary Figure S2 shows the association of rRNA 2’-O-Me with LSC and hematopoietic differentiation signatures. Supplementary Figure S3 shows strategy for LSC sorting and distribution of LSC methylation sites on ribosomes. Supplementary Figure S4 shows the effect of FBL knockdown on rRNA 2’-O-methylation and in vitro proliferation of leukemia cells. Supplementary Figure S5 show the effect of FBL overexpression on in vivo engraftment of primary AML cells. Supplementary Figure S6 shows the effect of FBL knockdown on nascent proteome and metabolism of AML cells. Supplementary Figure S7 shows the ribosome footprinting analysis after FBL knockdown. Supplementary Figure S8 shows the association of Gm1447 with LSC signature and the effect of Gm1447 supression on cellular amino acid levels. Supplementary Figure S9 shows regulatory effect of Gm1447 on in vivo leukemic self-renewal.

Targeted protein degradation (TPD) has emerged as a powerful strategy to selectively eliminate cellular proteins using small-molecule degraders, offering therapeutic promise for targeting proteins that are otherwise undruggable. However, a remaining challenge is to unambiguously identify primary TPD targets that are distinct from secondary downstream effects in the proteome. Here we introduce an approach for selective analysis of protein degradation by mass spectrometry (DegMS) at proteomic scale, which derives its specificity from the exclusion of confounding effects of altered transcription and translation induced by target depletion. We show that the approach efficiently operates at the timescale of TPD (hours) and we demonstrate its utility by analyzing the cyclin K degraders dCeMM2 and dCeMM4, which induce widespread transcriptional downregulation, and the GSPT1 degrader CC-885, an inhibitor of protein translation. Additionally, we apply DegMS to characterize a previously uncharacterized degrader, and identify the zinc-finger protein FIZ1 as a degraded target.

Human plasma is routinely collected during clinical care and constitutes a rich source of biomarkers for diagnostics and patient stratification. Liquid chromatography-mass spectrometry (LC-MS)-based proteomics is a key method for plasma biomarker discovery, but the high dynamic range of plasma proteins poses significant challenges for MS analysis and data processing. To benchmark the quantitative performance of neat plasma analysis, we introduce a multispecies sample set based on a human tryptic plasma digest containing varying low level spike-ins of yeast and E. coli tryptic proteome digests, termed PYE. By analysing the sample set on state-of-the-art LC-MS platforms across twelve different sites in data-dependent (DDA) and data-independent acquisition (DIA) modes, we provide a data resource comprising a total of 1116 individual LC-MS runs. Centralized data analysis shows that DIA methods outperform DDA-based approaches regarding identifications, data completeness, accuracy, and precision. DIA achieves excellent technical reproducibility, as demonstrated by coefficients of variation (CVs) between 3.3% and 9.8% at protein level. Comparative analysis of different setups clearly shows a high overlap in identified proteins and proves that accurate and precise quantitative measurements are feasible across multiple sites, even in a complex matrix such as plasma, using state-of-the-art instrumentation. The collected dataset, including the PYE sample set and strategy presented, serves as a valuable resource for optimizing the accuracy and reproducibility of LC-MS and bioinformatic workflows for clinical plasma proteome analysis.

Anticipated reactions to stressful situations are vital for the survival and well-being of organisms, and abnormal reactions results in stress-related disorders. The neuropeptide oxytocin is a key modulator ensuring well-adapted stress responses. Oxytocin acts on both neurons and astrocytes, but the molecular and cellular mechanisms mediating stress response remain poorly understood. Here, we focus on the amygdala, a crucial hub that integrates and processes sensory information through oxytocin-dependent mechanisms. Using an acute stress paradigm in mice, genetic and pharmacological manipulations combined with proteomic, morphological, electrophysiological and behavioral approaches, we reveal that oxytocinergic modulation of the freezing response to stress is mediated by transient Gαi-dependent retraction of astrocytic processes, followed by enhanced neuronal sensitivity to extracellular potassium in the amygdala. Our findings elucidate a pivotal role for astrocytes morphology-dependent modulation of brain circuits that is required for proper anticipated behavioral response to stressful situations. The role of oxytocin in modulating astrocytes during stress behaviour is not fully understood. Here the authors show that in the amygdala, oxytocin modulates stress related behaviour by transient Gαi-dependent retraction of astrocytic processes, followed by enhanced neuronal sensitivity to extracellular potassium.

Desmoplastic small round cell tumor (DSRCT) is an aggressive cancer that predominantly affects adolescents and young adults, typically developing at sites lined by mesothelium [1, 2]. DSRCT is genetically defined by a chromosomal translocation that fuses the N-terminus of EWS RNA binding protein 1 (EWSR1) to the C-terminus of Wilms tumor protein (WT1), forming EWSR1::WT1 [3]. This fusion encodes a potent transcription factor and is the only known driver of oncogenic transformation in DSRCT [4]. The lack of a comprehensive understanding of DSRCT biology parallels its dismal survival rate (5%-20%) [1]. These challenges are exacerbated by the absence of clinical trials, the limited systematic collection and analysis of DSRCT biomaterial [1], and the notable lack of specific diagnostic markers, necessitating resource-intensive molecular testing for an accurate diagnosis. Here we first focused on identifying promising candidates for validation as single, fast, and reliable diagnostic DSRCT markers. For this, we performed differential gene expression (DEG) analysis on datasets comprising patient samples from 32 DSRCT and 20 morphological mimics, identifying 23 genes overexpressed in DSRCT (log2 fold change (log2FC) > 2.5; adjusted P-value (Padj) < 0.01; Figure 1A, Supplementary Figure S1A). Secondly, we analyzed EWSR1::WT1 binding sites derived from chromatin immunoprecipitation followed by sequencing (ChIP-seq) data [5] obtained from the JN-DSRCT-1 cell line, identifying 2,065 genomic loci likely regulated by EWSR1::WT1 (Figure 1A). Third, we established JN-DSRCT-1 and SK-DSRCT2 cell lines expressing doxycycline (DOX)-inducible short hairpin RNA (shRNA)-mediated EWSR1::WT1 knockdown (KD) (Supplementary Figure S1B). Differential protein expression (DEP) analysis of these cells identified 104 proteins consistently regulated across both cell lines (log2FC > 1.0 and Padj < 0.01; Figure 1A, Supplementary Table S1). The intersection of these analyses revealed calcium voltage-gated channel auxiliary subunit alpha2delta 2 (CACNA2D2) and IQ motif containing G (IQCG) as potential DSRCT biomarkers (Figure 1A). CACNA2D2 was selected for validation due to its significantly higher expression in DSRCTs compared to IQCG (P < 0.001; Figure 1A). Indeed, DSRCT exhibited the highest expression of CACNA2D2 among all studied morphological mimics and normal tissues (P < 0.001; Supplementary Figures S1C-D). Further ChIP-seq data and motif analyses of EWSR1::WT1 binding coordinates and histone marks in JN-DSRCT-1 and four DSRCT patient samples [5, 6] suggested a direct regulatory role of EWSR1::WT1 through an enhancer interaction at the CACNA2D2 locus (Figure 1B). Notably, KD of EWSR1::WT1 in JN-DSRCT-1 resulted in a loss of the EWSR1::WT1 signal and Histone H3 lysine 27 acetylation (H3K27ac) enhancer marks at the CACNA2D2 locus (Figure 1B). Additionally, chromatin interaction data [6] revealed 19 loops connecting the EWSR1::WT1 binding site to the transcription start site of CACNA2D2, which were depleted upon KD of EWSR1::WT1 (Figure 1C). Super enhancer (SE) analysis further demonstrated that the EWSR1::WT1-bound enhancer exhibited a characteristic SE H3K27ac profile in JN-DSRCT-1, which was lost upon EWSR1::WT1 KD (Figure 1D, Supplementary Table S2). Post-transcriptional and post-translational KD of EWSR1::WT1 in three DSRCT cell line models expressing different EWSR1::WT1 isoforms (Supplementary Figure S2A) resulted in a significant reduction in CACNA2D2 expression (Figures 1E–F, Supplementary Figure S1B, Supplementary Figures S2B–F). Additionally, ChIP-seq data derived from MeT-5A mesothelial cells [6] – the potential cell of origin of DSRCT [7, 8] – ectopically expressing different EWSR1::WT1 isoforms (-KTS, +KTS, or -KTS/+KTS) suggested direct regulation, as evidenced by the co-occurrence of H3K27ac signals and signals for V5- or HA-tagged EWSR1::WT1 isoforms at the CACNA2D2 enhancer region (Supplementary Figure S2G). Notably, MeT-5A cells transfected with a control vector showed no substantial signal at this locus (Supplementary Figure S2G). Publicly available RNA-sequencing (RNA-seq) data from MeT-5A cells [6] expressing different EWSR1::WT1 isoforms showed that CACNA2D2 was differentially expressed in the presence of EWSR1::WT1 (4.1 ≤ log2FC ≤ 5.9, Padj < 0.001) (Supplementary Figure S2H). Finally, quantitative polymerase chain reaction (qPCR) analysis of MeT-5A cells stably expressing a DOX-inducible ectopic EWSR1::WT1 expression cassette confirmed that upon EWSR1::WT1 induction, CACNA2D2 was significantly and highly overexpressed (Supplementary Figure S2I). Taken together, these results emphasize that EWSR1::WT1 is sufficient to drive CACNA2D2 expression. SE analysis of MeT-5A-derived data strikingly showed that the CACNA2D2 enhancer bound by EWSR1::WT1 became a SE upon ectopic expression of EWSR1::WT1− KTS + KTS (Supplementary Figure S2J). To explore whether CACNA2D2 could serve as a surrogate indicator of oncogenic EWSR1::WT1 transformation, we defined a CACNA2D2 gene set and gene signature by performing a correlation analysis of gene expression data from 32 DSRCT patient samples (Supplementary Figure S3A, Supplementary Tables S3-S4). Next, an EWSR1::WT1 signature was computed by performing a combined DEG analysis of newly generated in vivo and in vitro [4] material derived from three DSRCT cell lines upon EWSR1::WT1 KD (Supplementary Figure S3A, Supplementary Table S4). Notably, fast gene set enrichment analysis (fGSEA) of the CACNA2D2 gene set demonstrated a highly significant (Padj < 0.001) and strong positive enrichment for the EWSR1::WT1 signature (normalized enrichment score, NESEWSR1::WT1 = 3.6). Moreover, single sample gene set enrichment analysis (ssGSEA) of expression data from 32 DSRCT patient samples confirmed that the EWSR1::WT1 signature significantly correlated with that of CACNA2D2 (r = 0.75), highlighting a transcriptional interconnection between CACNA2D2 and EWSR1::WT1 in situ (Figure 1G). These observations were further supported by single-cell (sc)-derived signatures from orthotopically-generated tumors using two DSRCT cell lines with DOX-inducible KD of EWSR1::WT1 at primary (n = 221) and metastatic (n = 221) locations (Figure 1G, Supplementary Table S4). Indeed, ssGSEA of our single-cell data showed highly significant correlation between the NES of our generated EWSR1::WT1 and CACNA2D2 signatures (Figure 1G), regardless of tumor location, implying that CACNA2D2-associated genes are also characteristic features of metastasized DSRCT cells (Supplementary Figure S3B). To delineate the specificity of the interaction between CACNA2D2 and EWSR1::WT1 in DSRCT, we performed ssGSEA using our EWSR1::WT1 and CACNA2D2 signatures on expression data from 20 DSRCT morphological mimics (Figure 1H). Here, non-DSRCT cancer entities showed significantly lower NES and correlation strength for all signatures compared to DSRCT (Supplementary Figures S3C-D). These results further emphasized the high specificity of the CACNA2D2 and EWSR1::WT1 interplay in DSRCT. Moreover, both bulk- and sc-derived CACNA2D2 signatures precisely distinguished DSRCT cell clusters from non-tumor cells in single-cell RNA-sequencing (scRNA-seq) data from four DSRCT patients (n = 11 samples) [9] (Figure 1I, Supplementary Figure S3E). Concordantly, all predicted normal cell types within these tumors exhibited low enrichment of both CACNA2D2 signatures (Supplementary Figures S3F-G). Further, dimensional reduction of CACNA2D2-associated CpG sites in 24 DSRCT patient samples, compared with 192 samples from 13 morphological mimics [10] revealed distinct clustering of all DSRCT samples, which was unique to CACNA2D2 compared to other described EWSR1::WT1-regulated genes or IQCG (Figure 1A, Supplementary Figures S4A-B). Additionally, these CACNA2D2-associated CpG sites exhibited significant (P < 0.001) and specific hypomethylation in DSRCT patient samples, collectively suggesting that the CACNA2D2-associated methylation signature is a distinct and specific feature of DSRCT (Supplementary Figure S4C). To assess the diagnostic utility of CACNA2D2, we optimized a staining protocol for DSRCT cell line xenografts, achieving consistent and robust membranous or cytoplasmatic staining, even uncovering micrometastases (Figure 1J, Supplementary Figure S4D). Finally, we assembled the largest collection of fresh-frozen and paraffin-embedded DSRCT patient samples analyzed to date (n = 61), comprising primary, metastatic, and post-treatment samples, and supplemented it with 249 patient samples from 18 different DSRCT morphological mimics (Supplementary Table S5). CACNA2D2 immunoreactivity was evaluated using a modified Immune Reactive Score (IRS) (Supplementary Material and Methods). Excitingly, DSRCT tumor sections exhibited the highest IRS for CACNA2D2 (IRSmean = 10.5, 6 ≤ IRSDSRCT ≤ 12, P < 0.001) (Supplementary Figure S4E-F), with specificity reaching 98% when applying a cutoff of IRS > 1 (Figure 1K-M, Supplementary Figure S4E). Indeed, even samples derived from CIC- and BCOR-rearranged sarcomas, as well as fusion-positive alveolar rhabdomyosarcomas, showed negligible mean protein expression compared to DSRCT (IRSCIC = 0.21, IRSBCOR = 0, IRSfp-ARMS = 0.56). Furthermore, 100% sensitivity was achieved when applying an IRS cutoff of ≤ 6, implying that DSRCT samples consistently displayed strong staining for CACNA2D2 (Figure 1M). Thus, we recommend a single CACNA2D2 staining for clinically and histologically compatible DSRCT differential diagnosis. If IRSCACNA2D2 ≤ 1, the diagnosis should be reconsidered or re-evaluated using molecular diagnostic procedures (such as fluorescence in situ hybridization, qRT-PCR, or next-generation sequencing), if available (Figure 1N). Conversely, if IRSCACNA2D2 > 1, a diagnosis of DSRCT may be established. Also, CACNA2D2 staining may be used to rule out DSRCT within the broad spectrum of small-round-blue-cell tumors, potentially offering extensive diagnostic utility. Finally, the high, specific, and homogenous membranous expression of CACNA2D2 in DSRCT, combined with the highly specific antibody described here, makes CACNA2D2 an ideal candidate for targeted therapeutic approaches, including drug delivery using antibody-drug conjugates or CAR-T cell therapy. Future studies should investigate the precise role of CACNA2D2 in DSRCT biology, with a focus on its potential contributions in tumor cell fitness, differentiation, and tumorigenic potential. In conclusion, here we developed an extensive toolset for DSRCT research (Supplementary Figure S4G), a validated blueprint for how such resources could be harnessed in other cancer entities, and identified CACNA2D2 as a singular, powerful DSRCT biomarker. Florian Henning Geyer, Florencia Cidre-Aranaz, and Thomas Georg Phillip Grünewald conceived the study. Florian Henning Geyer and Florencia Cidre-Aranaz wrote the paper and drafted all figures and tables. Florian Henning Geyer carried out all in vitro and in vivo experiments and performed all bioinformatic and statistical analyses. Florian Henning Geyer, Alina Ritter, and Thomas Georg Phillip Grünewald performed immunohistochemical evaluation and scoring of tumor samples and TMAs. Florencia Cidre-Aranaz, Roland Imle, and Ana Banito performed and/or coordinated in vivo experiments. Olivier Delattre provided microarray expression data. Seneca Kinn-Gurzo performed in vitro experiments on BER cell lines. Tobias Faehling and Clémence Henon performed single-cell bioinformatic analyses. Karim Aljakouch and Azhar Orynbek performed MassSpec and analyzed MassSpec data. Alina Ritter, Jing Li, Endrit Vinca, Laura Romero-Perez, Martin Sill, and Shunya Ohmura contributed to experimental procedures. Wolfgang Hartmann and Benjamin Friedrich Berthold Mayer provided clinical and/or histological guidance. Enrique De Álava, Juan Díaz-Martín, Stefanie Bertram, Sophie Postel-Vilnay, Martin Ebinger, Monika Sparber-Sauer, Daniel Baumhoer, Carine Ngo, David Horst, Yvonne Versleijen-Jonkers, Armin Jarosch, Sabine Stegmaier, and Thomas Kirchner provided clinical samples. Patrick Joseph Grohar, Thomas Georg Phillip Grünewald, and Jeroen Krijgsveld provided laboratory infrastructure. Florencia Cidre-Aranaz and Thomas Georg Phillip Grünewald supervised the study and data analysis. All authors read and approved the final manuscript. We would like to thank Nadine Gmelin, Stefanie Kutschmann, and Felina Zahnow for their expert technical assistance, and Claudia Schmidt from the Light Microscopy Facility (German Cancer Research Center (DKFZ), Heidelberg, Germany) for her meticulous work in conducting immunohistochemical stainings. We thank the Microarray Core Facility (German Cancer Research Center (DKFZ)) for providing the Gene Expression Arrays and related services. We thank Katharina Bauer, Denise Keitel and Jan-Philipp Mallm from the Single-cell Open Lab (German Cancer Research Center (DKFZ)) for expert support in the preparation of single-cell libraries. We thank the Flow Cytometry Facility team (German Cancer Research Centre (DKFZ)) for their support with cell sorting. We thank Dr. Marc Ladanyi for sharing the SK-DSRCT2 cell line. The authors declare no competing interests. The laboratory of Thomas Georg Phillip Grünewald is supported by grants from the Matthias-Lackas Foundation, the Dr. Leopold und Carmen Ellinger Foundation, the European Research Council (ERC CoG 2023 #101122595), the Deutsche Forschungsgemeinschaft (DFG 458891500), the German Cancer Aid (DKH-70112257, DKH-7011411, DKH-70114278, DKH-70115315), the Dr. Rolf M. Schwiete foundation, the SMARCB1 association, the Ministry of Education and Research (BMBF; SMART-CARE and HEROES-AYA), and the Barbara and Wilfried Mohr foundation. The research team of Florencia Cidre-Aranaz was supported by the German Cancer Aid (DHK-70114111), and the Dr. Rolf M. Schwiete Stiftung (2020-028 and 2022-31). In addition, this work was delivered as part of the PROTECT team supported by the Cancer Grand Challenges partnership funded by Cancer Research UK, the National Cancer Institute, the Scientific Foundation of the Spanish Association Against Cancer And KiKa (Children Cancer Free Foundation). Florian Henning Geyer, Tobias Faehling, Endrit Vinca, and Alina Ritter were supported by the German Academic Scholarship Foundation. In addition, Endrit Vinca was supported by scholarships from the Heinrich F.C. Behr foundation and the Rudolf and Brigitte Zenner foundation, Tobias Faehling by the Heinrich F.C. Behr foundation, and Florian Henning Geyer and Alina Ritter are supported by the German Cancer Aid through the 'Mildred-Scheel-Doctoral Program' (DKH-70114866). This project is co-funded by the European Union (ERC, CANCER-HARAKIRI, 101122595). All views and opinions expressed are however those of the authors only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them. In vivo experiments were approved by the government of North Baden and conducted in accordance with ARRIVE guidelines and recommendations of the European Community (86/609/EEC) and UKCCCR (guidelines for the welfare and use of animals in cancer research). Open slides or tissue-microarrays from human formalin-fixed, paraffin-embedded or cryopreserved tissue samples were retrieved from the archives of the Institute of Pathology of the LMU Munich, the Charité Berlin, The Biobank of the Hospital Universitario Virgen del Rocío of Seville, the Hospital Gustave Roussy (Villejuif), the Bone Tumor Reference Center at the University of Basel, the University of Essen, the Cooperative Weichteilsarkom Studiengruppe (CWS) study center, the Klinikum Stuttgart (ethics committee from the Medical Faculty of the Eberhard-Karls University and University Hospital of Tübingen, approval no. 207/2022BO2), the Radboud University Medical Center, the Pathology Institute of the LMU Munich (approval no. 550-16 UE), and the University of Heidelberg (approval no. S-211/2021). The microarray data are deposited at the National Center for Biotechnology Information (NCBI) GEO database with accession codes GSE273438 and GSE273441. All proteomic data is deposited at the PRoteomics IDEntifications database with accession code PXD053786. All other data supporting the findings of this study are available within the article and its supplementary information files, or from the corresponding author upon reasonable request. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Prostate carcinoma (PCa) is the most common cancer of men, associated with a still unresolved issue of accurate risk-stratification. While recent advances in omics technologies have provided clues as to how molecular changes shape the onset and progression of PCa, it remains largely unclear whether germline variants and somatic mutations cooperate to contribute to PCa progression and outcome. Thus, we explored whether oncogenic cooperation between regulatory germline variants and somatic driver mutations can help explain why some PCa patients develop a more aggressive phenotype, which may have implications for risk-adapted medical treatment. Here, by employing an integrative functional genomics approach, we identified receptor-type protein-tyrosine phosphatase kappa ( PTPRK ) as a TMPRSS2::ERG (TE)-modulated gene associated with PCa progression whose expression is controlled by cooperation of the TE-fusion with a regulatory single nucleotide polymorphism (SNP). Analysis of available clinically annotated patient cohorts demonstrated that PTPRK is overexpressed in TE-positive PCa tumors and associated with higher Gleason scores and metastatic disease. TE knockdown in PCa cell lines reduced PTPRK expression, while ectopic overexpression of the fusion in TE-negative PCa cell lines and prostatic epithelium cells induced its expression. Functionally, PTPRK silencing inhibited cellular proliferation, cell cycle progression, and clonogenic growth of PCa cells, which was mirrored by dysregulation of corresponding gene and protein signatures in global transcriptomic and phospho-proteomic analyses after PTPRK knockdown. Analysis of TE ChIP-Seq and Hi-C data from PCa cells highlighted a proximal TE-bound DNA element whose TE-dependent enhancer activity was validated in reporter assays and which could be abrogated by a regulatory SNP. Collectively, our results provide evidence of how exploration of oncogenic cooperation may help to identify novel biomarkers and potentially druggable pathways and highlight the role of the regulatory genome in PCa progression. ### Competing Interest Statement The authors have declared no competing interest. * CV : crystal violet CDF : chip description files cDNA : complementary DNA DEGs : differentially expressed genes DOX : doxycycline FC : fold change GS : Gleason score GWAS : genome-wide association study HRP : horseradish peroxidase NES : normalized enrichment score PCa : prostate cancer PI : propidium iodide PTPRK : receptor-type protein tyrosine phosphatase kappa qRT-PCR : quantitative real-time polymerase chain reaction RMA : robust multi-array analysis SNP : single nucleotide polymorphism STR : short tandem repeat TE : TMPRSS2::ERG German Cancer Aid, DKH-70114278, DKH-70114285, DKH-70114286 Matthias-Lackas foundation

Bone marrow mesenchymal stromal cells (MSCs) are a major source of secreted factors that control hematopoietic stem and progenitor cell (HSPC) function. We previously reported the generation of revitalized MSCs (rMSCs), which more effectively support HSPCs in culture. In a secretome screen using rMSCs, we identified semaphorin 3A (SEM3A) as a secreted factor upregulated as part of a pro-inflammatory signature that may contribute to HSPC expansion by rMSCs. We show that recombinant SEM3A acts directly on HSPCs to inhibit their cycling ex vivo . Analysis of a SEM3A loss of function mutation in vivo revealed hematopoietic progenitor expansion and accelerated recovery after myeloablation, consistent with a role for SEM3A in regulating HSPCs at steady state and during hematopoietic stress. This work highlights proteomic screening using rMSCs as a method to identify novel secreted niche factors and uncovers a novel role for SEM3A in controlling HSPC proliferation in stress hematopoiesis. Summary Borger et al. characterize the secretome of revitalized bone marrow stromal cells and identify a novel role of the protein semaphorin 3A in regulating hematopoietic stem and progenitor cell proliferation in steady state and stress conditions. ### Competing Interest Statement P.S.F. served as a consultant for Pfizer, received research funding from Ironwood Pharmaceuticals outside the submitted work, and was a shareholder of Cygnal Therapeutics. K.G. has received research funding from ADC Therapeutics and iOnctura outside the submitted work. All other authors declare no competing interests. The raw LC-MS/MS data, process report files, and extracted peptide features for the pSILAC experiments have been deposited in the ProteomeXchange Consortium via the PRIDE partner repository under the accession code PXD065939 . The raw LC-MS/MS data and annotated spectra for the unlabeled proteomic experiments are available under the accession code PXD065232 . Code used for analysis and visualization of transcriptomic and proteomic data is available upon reasonable request. National Institutes of Health, DK056638, DK130895, F30HL154749, S10OD026833, S10OD032169, P30CA013330 New York State Department of Health, https://ror.org/04hf5kq57, C029154, C029570 National Institute of General Medical Sciences, https://ror.org/04q48ey07, 5T32GM007491, 5T32GM007288