º£½ÇÆƽâ°æ

Existing antibody language models (LMs) are pre-trained using a masked language modeling (MLM) objective with uniform masking probabilities. While these models excel at predicting germline residues, they often struggle with mutated and non-templated residues, which are crucial for antigen-binding specificity and concentrate in the complementarity-determining regions (CDRs). Here, we demonstrate that preferential masking of the non-templated CDR3 is a compute-efficient strategy to enhance model performance. We pre-trained two antibody LMs (AbLMs) using either uniform or preferential masking and observed that the latter improves residue prediction accuracy in the highly variable CDR3. Preferential masking also improves antibody classification by native chain pairing and binding specificity, suggesting improved CDR3 understanding and indicating that non-random, learnable patterns help govern antibody chain pairing. We further show that specificity classification is largely informed by residues in the CDRs, demonstrating that AbLMs learn meaningful patterns that align with immunological understanding.

This study sheds light on the pivotal role of the oncoprotein DEK in B-cell lymphoma. We reveal DEK expression correlates with increased tumor proliferation and inferior overall survival in cases diagnosed with low-grade B-cell lymphoma (LGBCL). We also found significant correlation between DEK expression and copy number alterations in LGBCL tumors, highlighting a novel mechanism of LGBCL pathogenesis that warrants additional exploration. To interrogate the mechanistic role of DEK in B-cell lymphoma, we generated a DEK knockout cell line model, which demonstrated DEK depletion caused reduced proliferation and altered expression of key cell cycle and apoptosis-related proteins, including Bcl-2, Bcl-xL, and p53. Notably, DEK depleted cells showed increased sensitivity to apoptosis-inducing agents, including venetoclax and staurosporine, which underscores the therapeutic potential of targeting DEK in B-cell lymphomas. Overall, our study contributes to a better understanding of DEK¡¯s role as an oncoprotein in B-cell lymphomas, highlighting its potential as both a promising therapeutic target and a novel biomarker for aggressive LGBCL. Further research elucidating the molecular mechanisms underlying DEK-mediated tumorigenesis could pave the way for improved treatment strategies and better clinical outcomes for patients with B-cell lymphoma.

Epstein-Barr Virus (EBV) is associated with numerous cancers including B cell lymphomas. In vitro, EBV transforms primary B cells into immortalized Lymphoblastoid Cell Lines (LCLs) which serves as a model to study the role of viral proteins in EBV malignancies. EBV induced cellular transformation is driven by viral proteins including EBV-Nuclear Antigens (EBNAs). EBNA-LP is important for the transformation of na?ve but not memory B cells. While EBNA-LP was thought to promote gene activation by EBNA2, EBNA-LP Knockout (LPKO) virus-infected cells express EBNA2-activated cellular genes efficiently. Therefore, a gap in knowledge exists as to what roles EBNA-LP plays in na?ve B cell transformation. We developed a trans-complementation assay wherein transfection with wild-type EBNA-LP rescues the transformation of peripheral blood- and cord blood-derived na?ve B cells by LPKO virus. Despite EBNA-LP phosphorylation sites being important in EBNA2 co-activation; neither phospho-mutant nor phospho-mimetic EBNA-LP was defective in rescuing na?ve B cell outgrowth. However, we identified conserved leucine-rich motifs in EBNA-LP that were required for transformation of adult na?ve and cord blood B cells. Because cellular PPAR-g coactivator (PGC) proteins use leucine-rich motifs to engage transcription factors including YY1, a key regulator of DNA looping and metabolism, we examined the role of EBNA-LP in engaging transcription factors. We found a significant overlap between EBNA-LP and YY1 in ChIP-Seq data. By Cut&Run, YY1 peaks unique to WT compared to LPKO LCLs occur at more highly expressed genes. Moreover, Cas9 knockout of YY1 in primary B cells prior to EBV infection indicated YY1 to be important for EBV-mediated transformation. We confirmed EBNA-LP and YY1 biochemical association in LCLs by endogenous co-immunoprecipitation and found that the EBNA-LP leucine-rich motifs were required for YY1 interaction in LCLs. We propose that EBNA-LP engages YY1 through conserved leucine-rich motifs to promote EBV transformation of na?ve B cells. Author summaryEpstein-Barr Virus (EBV) is associated with various B cell lymphomas, particularly in immunosuppressed individuals. In the absence of a functional immune system, viral latency proteins, including EBV Nuclear Antigens (EBNAs) act as oncoproteins to promote tumorigenesis. EBNA-LP is one of the first viral proteins produced after infection and is important for the transformation of na?ve B cells. However, the roles of EBNA-LP during infection are largely undefined. In this study, developed an assay in which the role of wild type and mutant EBNA-LP could be investigated in the context of primary na?ve B cells infected with an EBNA-LP Knockout virus. Using this assay, we identified highly conserved leucine-rich motifs within EBNA-LP that are important for transformation of EBV-infected na?ve B cells. These conserved motifs associate with the cellular transcription factor YY1, an important transcriptional regulator in B cell development and in many cancers, that we now show is essential for outgrowth of EBV infected B cells. Our study provides further insights into the mechanisms by which EBV transforms na?ve B cells.