Biology subjects

Xu, A. M.

Publications and source records attributed to Xu, A. M..

7 recordsLinked to original sources

bioRxiv2026

Spatially resolved transcriptional programs link fallopian tube precursor lesions to immune activation and stromal reorganization

Despite its name, high-grade serous ovarian carcinoma (HGSC) originates in the fallopian tube, not the ovary, arising from a morphologically recognizable precursor lesion, serous tubal intraepithelial carcinoma (STIC). Yet the early cellular and microenvironmental changes driving this transformation remain poorly understood, limiting progress in early detection, interception, and prevention. Here, we generated a Visium HD spatial transcriptomic atlas of fallopian tube carcinogenesis spanning histologically unremarkable fallopian tube epithelium (FTE), STIC, and invasive HGSC. This approach enabled unbiased, tissue-wide, whole-transcriptome mapping at single-cell-level resolution within preserved histologic architecture, providing spatial granularity beyond prior region-of-interest-based platforms. STIC lesions displayed a coordinated epithelial transformation program marked by proliferation, replication stress, DNA repair activation, chromatin remodeling, and induction of tumor-associated antigens, including PRAME and CLDN6, which are emerging targets for vaccine and antigen-directed therapeutic strategies. In contrast, histologically unremarkable FTE contained spatially restricted epithelial defense programs marked by SCGB1A1 and MUC6, suggesting localized protective states that may influence susceptibility to malignant transformation. Using distance- and density-aware spatial analyses, we found that precursor lesions were embedded within immune-enriched, stromal-depleted microenvironments characterized by interferon-dominant immune activation, attenuation of TNF/NF-{kappa}B signaling, macrophage and lymphoid remodeling, and extracellular matrix-associated fibroblast interactions. Computational pathology analysis of collagen architecture confirmed reduced collagen fiber density in STIC-adjacent stroma, linking transcriptomic evidence of stromal remodeling to structural extracellular matrix changes. Together, these data define early epithelial, immune, and stromal programs associated with STIC and identify tumor-associated antigens, epithelial defense states, and immune-stromal niches as candidate targets for HGSC prevention and early interception.

cancer biology↗

bioRxiv2025

Analysis of Large Peptide Collisional Cross Section Dataset Reveals Structural Origin of Bimodal Behavior

Recent advances in ion mobility spectrometry have enabled the measurement of rotationally averaged collisional cross-sectional area (CCS) for millions of peptides, as part of routine proteomic mass spectrometry workflows. One of the most striking finding in recent large ion mobility datasets is that CCS exhibits two distinct modes, most notably for charge 3+ peptides. Here, using classical machine learning approaches, we identify that basic site positioning is a key sequence feature determining if a peptide belongs to the high or low CCS mode. Molecular dynamics simulations suggest that peptides in the high CCS mode tend to adopt more extended conformations and form charge-stabilized helical structures, whereas those in the low CCS mode adopt more compact, globular conformations. Further supporting this structural hypothesis, we provide evidence for preferential protonation near the C-terminus, and uncover multiple position-dependent sequence determinants that all suggest the predominance of helix formation in the high mode. Together, these findings will enable better integration of CCS measurements into protein identification and quantification pipelines, improving the performance of ion mobility-based proteomics.

bioinformatics↗

bioRxiv2023

Generating counterfactual explanations of tumor spatial proteomes to discover effective strategies for enhancing immune infiltration

Cancer progression can be slowed down or halted via the activation of either endogenous or engineered T cells and their infiltration of the tumour microenvironment. Here we describe a deep-learning model that uses large-scale spatial proteomic profiles of tumours to generate minimal tumour perturbations that boost T-cell infiltration. The model integrates a counterfactual optimization strategy for the generation of the perturbations with the prediction of T-cell infiltration as a self-supervised machine-learning problem. We applied the model to 368 samples of metastatic melanoma and colorectal cancer assayed using 40-plex imaging mass cytometry, and discovered cohort-dependent combinatorial perturbations (CXCL9, CXCL10, CCL22 and CCL18 for melanoma, and CXCR4, PD-1, PD-L1 and CYR61 for colorectal cancer) that support T-cell infiltration across patient cohorts, as confirmed via in vitro experiments. Leveraging counterfactual-based predictions of spatial omics data may aid the design of cancer therapeutics.

cancer biology↗

bioRxiv2023

Imaging Mass Cytometry-Driven Spatial Analysis and Biomarker Discovery in Hodgkin Lymphoma

The biology of tumors is suffused with spatial interactions, such as tumor-immune signaling through localized cytokine/ligand secretion, cell-cell contacts, and checkpoint ligand/receptor signaling. Hodgkin Lymphoma (HL) can serve as a study paradigm for tumor microenvironment (TME) architecture as the defining pathological feature is the scarcity of the malignant Hodgkin and Reed Sternberg (HRS) cells, leaving a diverse and predominantly immune cell rich tumor microenvironment (TME) with complex tumor-immune interactions. Previous studies have identified TME features that are prognostic and predictive, however these studies did not consider the entirety of TME cellular ecosystems, including precisely defined immune cell subsets with opposing inflammatory and immune-suppressive effects, as a determinant for differential clinical course of HL patients. Here we use Imaging Mass Cytometry (IMC) with 42 antibody markers to profile tumors from 93 patients with HL. Our cohort consists of relapsed/refractory HL with matched diagnostic and relapsed biopsies, and we present a bioinformatic pipeline to profile 10 major cell lineages and their subtypes including spatial interaction mapping. Our pipeline identifies putative biomarker candidates with a focus on "rosettes" - local aggregates of immune cells around single tumor cells. In addition to validating existing biomarkers centered on CD68+ macrophages, GranzymeB+CD8+ T cells, and others in HL, we propose new biomarkers based on localized interactions between HRS cells and aggregating CD4+ and CD8+ T cells and macrophages involving the immune checkpoints PD1/PDL1, LAG3, and Galectin9. This study serves as a broad tissue imaging resource for multi-timepoint biopsies in HL, and a computational resource and pipeline for users of IMC and other multiplexed imaging studies to perform tissue analysis and biomarker candidate testing with any tissue type.

cancer biology↗

bioRxiv2023

A high-throughput approach reveals distinct peptide charging behaviors in electrospray ionization mass spectrometry

Electrospray ionization is a powerful and prevalent technique used to ionize analytes in mass spectrometry. The distribution of charges that an analyte receives (charge state distribution, CSD) is an important consideration for interpreting mass spectra. However, due to an incomplete understanding of the ionization mechanism, the analyte properties that influence CSDs are not fully understood. Here, we employ a machine learning-based high-throughput approach and analyze CSDs of hundreds of thousands of peptides. Interestingly, half of the peptides exhibit charges that differ from what one would naively expect (number of basic sites). We find that these peptides can be classified into two regimes--undercharging and overcharging--and that these two regimes display markedly different charging characteristics. Strikingly, peptides in the overcharging regime show minimal dependence on basic site count, and more generally, the two regimes exhibit distinct sequence determinants. These findings highlight the rich ionization behavior of peptides and the potential of CSDs for enhancing peptide identification.

systems biology↗

bioRxiv2019

MATE-Seq: Microfluidic Antigen-TCR Engagement Sequencing

Adaptive immunity is based on peptide antigen recognition. Our ability to harness the immune system for therapeutic gain relies on the discovery of the T cell receptor (TCR) genes that selectively target antigens from infections, mutated proteins, and foreign agents. Here we present a method that selectively labels peptide antigen-specific CD8+ T-cells in human blood using magnetic nanoparticles functionalized with peptide-MHC tetramers, isolates these specific cells within an integrated microfluidic device, and directly amplifies the TCR genes for sequencing. Critically, the identity of the peptide recognized by the TCR is preserved, providing the link between peptide and gene. The platform requires inputs on the order of just 100,000 CD8+ T cells, can be multiplexed for simultaneous analysis of multiple peptides, and performs sorting and isolation on chip. We demonstrate 1000-fold sensitivity enhancement of antigen-specific T-cell receptor detection and simultaneous capture of two virus antigen-specific T-cell receptors from samples of human blood.

immunology↗

bioRxiv2017

Cellular reprogramming of human monocytes is regulated by time-dependent IL4 signalling and NCOR2

The clinical and therapeutic value of human in vitro generated monocyte-derived dendritic cell (moDC) and macrophages is well established. However, in line with recent findings regarding myeloid cell ontogeny and due to our limited understanding of their physiological counterparts, transcriptional regulation and heterogeneity, the full potential of these important cellular systems is still underestimated.\n\nIn this study, we use cutting edge high-dimensional analysis methods to better understand the transcriptional organization, phenotypic heterogeneity and functional differences between human ex vivo isolated and in vitro generated mononuclear phagocytes with the aim to better realize their full potential in the clinic.\n\nWe demonstrate that human monocytes activated by MCSF or GMCSF most closely resemble inflammatory macrophages identified in vivo, while IL4 signalling in the presence of GMCSF generates moDCs resembling inflammatory DCs in vivo, but not steady state cDC1 or cDC2. Moreover, these reprogramming regimes lead to activated monocytes that present with profoundly different transcriptomic, metabolic, phenotypic and functional profiles. Furthermore, we demonstrate that CD14+ monocytes are integrating multiple exogenous activation signals such as GMCSF and IL4 in a combinatorial and temporal fashion, resulting in a high-dimensional cellular continuum of reprogrammed monocytes dependent on the mode and timing of cytokine exposure. Utilizing nanostraw-based knockdown technology, we demonstrate that the IL4-dependent generation of moDCs relies on the induction, nuclear localization and function of the transcriptional regulator NCOR2.\n\nFinally, we unravel unappreciated heterogeneity within the clinically moDCs population and propose a novel high-dimensional phenotyping strategy to better tailor clinical quality control strategies for patient need and culture conditions to enhance therapeutic outcome.

immunology↗