Biology subjects

Rajaram, S.

Publications and source records attributed to Rajaram, S..

7 recordsLinked to original sources

bioRxiv2026

A Visually Interpretable Histopathology-Based Immune Model Predicts T-effector Biology and Response to Immune checkpoint inhibition in Clear Cell Renal Cell Carcinoma Clinical Trial and Contemporary Real-World Datasets

Immune checkpoint inhibitors (ICI) are central to the treatment of metastatic clear cell renal cell carcinoma (ccRCC), yet only a subset of patients derive durable benefit, and clinically deployable predictive biomarkers remain an unmet need. RNA-based T-effector signatures capture cytotoxic immune biology and have been associated with ICI response in clinical trial cohorts; however, their clinical implementation is limited by the marked spatial heterogeneity of ccRCC, as well as cost, long turnaround time, sample quality requirements, and limited accessibility. Here, we developed a visually interpretable deep learning (DL) model that predicts a T-cell-enriched immune score directly from hematoxylin and eosin (H&E)-stained whole-slide images. To overcome the inability of H&E morphology alone to distinguish lymphocyte subsets, we trained the model using multimodal spatial supervision from CD8, PAX8, and ERG IHC, which respectively identified cytotoxic T-cell-rich regions, tumor cells, and endothelial cells, thereby constraining immune predictions to relevant tumor microenvironmental niches. The resulting H&E DL Immune score was validated by pathologist review, comparison with held-out CD8 IHC annotations, and independent datasets. The H&E DL Immune score correlated with T-effector RNA scores across independent institutional and IMmotion150 clinical trial cohorts (spearman correlations of 0.726; p=5.90x10-15 and 0.706; p=4.04x10-19). As a proof of principle, the score was used to characterize associations with key biological features across large cohorts, including sarcomatoid differentiation, BAP1 and PBRM1 mutation status, and additional transcriptomic signatures. In IMmotion150 clinical trial cohort, a median-dichotomized H&E DL Immune score, similar to RNA-based T-effector score, was significantly associated with clinical benefit from atezulumab therapy. In contemporary institutional cohorts of patients treated with frontline ipilimumab plus nivolumab or in initial 3 lines of nivolumab monotherapy, patients in the top quartile of H&E DL Immune score had significantly longer progression-free survival. Collectively, these findings support a scalable and interpretable H&E-based biomarker that captures T-effector biology and can help identify patients with ccRCC more likely to benefit from ICIs.

bioinformatics↗

bioRxiv2025

Resolving phenotyping discordance with SPACEMAP, an integrated machine learning framework

Multiplex imaging technologies have revolutionized our ability to study cellular behavior within the tissue microenvironment. Translating this complex data into meaningful biological insights requires a unified analytical framework. To address this, we developed SPACEMAP (Spatial Phenotyping And Classification with Enhanced Multiplex Analysis Pipeline), a comprehensive Python and Qupath-based platform for multiplex imaging analysis. SPACEMAP integrates image registration, segmentation, artifact removal, tissue and zone classification, spatial feature extraction, and a consolidated phenotyping approach into a single system. A core feature of SPACEMAP is its high-fidelity phenotyping. To evaluate classification performance, we benchmarked our method RESOLVE, against three established approaches, Leiden clustering, Self-Organizing Maps, and SCIMAP revealing substantial disagreement among them. SPACEMAP overcomes this through two complementary workflows: a machine learning model trained on expert-labeled cells, and a consensus classifier that integrates high-confidence cells across methods. Here, we validated SPACEMAP on in-house colorectal cancer samples and a public dataset, demonstrating its robustness.

bioinformatics↗

bioRxiv2025

GPCR-targeted imaging and manipulation of homeostatic microglia in living systems

Microglia, the resident innate immune cells of the brain, are known to perform key roles such as synaptic pruning, apoptotic debris removal, and pathogen defense in the central nervous system. Microglial mutations are directly linked to many neurodevelopmental (e.g., schizophrenia) and neurodegenerative (e.g., Alzheimers disease) disorders, indicating the diagnostic and therapeutic potential of microglia for treating these conditions. Currently, we lack robust molecular tools to specifically image and manipulate microglia in vivo, which presents a major hurdle in our understanding of the brain-wide functions of these cells during the early onset of brain diseases. Here, we describe a molecular technology for imaging and manipulation of homeostatic microglia in live organisms (e.g., in zebrafish and mice) by covalently targeting the purinergic receptor, P2RY12. Using this technology, we imaged microglia-pathogen interactions in the larval zebrafish brain and revealed various morphological states of microglia in the adult mouse brain. We further expanded the microglia labelling approach to single-microglia tracking and microglial surfaceome mapping using photoactivatable fluorophores and photoproximity labelling, respectively. We anticipate the use of this universal tool for studying microglial biology across species to reveal the dynamics and polarization of resting microglia into a reactive state found in many neurodegenerative diseases.

immunology↗

bioRxiv2025

Autonomous learning of pathologists' cancer grading rules

Deep learning reveals that tissue morphology contains rich pathophysiological information beyond human understanding. However, approaches to convert these spatially distributed signals into subcellular insights informing disease mechanisms are lacking. We introduce Delta-Marches, an interpretability-first approach that nominates distinguishing morphological features rather than explaining existing models' decisions. Delta-Marches simulates idealized morphological changes between classes by coupling latent-space traversals with generative AI. Comparing each image to its class-shifted counterpart allows feature extractors to infer the most affected aspects, reducing sample-to-sample variability and resolving transformations at subcellular resolution. Prototyped in renal carcinoma grading, Delta-Marches generates realistic grade transitions and pinpoints tumor-cell nuclear phenotypes as key determinants. It also reveals reduced vasculature with increasing grade, a known pattern absent from standard rubrics. Applied to dysplastic progression in colorectal tissue, it recovers glandular remodeling and goblet-cell loss, generalizing across tissue types and scales. These results show Delta-Marches parses complex image phenotypes and catalyzes hypothesis generation.

bioinformatics↗

bioRxiv2024

Caveolin-1 regulates context-dependent signaling and survival in Ewing Sarcoma.

Plasticity is a hallmark function of cancer cells, yet the mechanisms that enable dynamic switching between survival states remain incompletely understood. Here, we identify Caveolin-1, a membrane-domain scaffolding protein, as a context-dependent regulator of survival signaling in Ewing sarcoma (EwS). Single-cell analyses reveal a distinct subpopulation of EwS cells marked by high CD99 and elevated Caveolin-1 expression. These CD99High cells exhibit unique morphology, transcriptional programs, and markedly enhanced survival both under chemotherapeutic challenge and in vivo. Importantly, CD99High and CD99Low states are reversible, providing EwS cells with a flexible route to survival-oriented plasticity. Mechanistically, we show that Caveolin-1 in CD99High cells orchestrates PI3K/AKT survival signaling by modulating the spatial organization of PI3K activity on the plasma membrane. We propose that the CD99High state establishes a Caveolin-1-driven signaling architecture that supports survival through mechanisms distinct from those used by CD99Low cells. These findings uncover a dynamic state transition in EwS cells and position Caveolin-1 as a key driver of context-specific survival signaling.

cancer biology↗

bioRxiv2021

Natural killer cell immunosuppressive function requires CXCR3-dependent redistribution within lymphoid tissues

Natural killer (NK) cell suppression of T cells is a key determinant of viral pathogenesis and vaccine efficacy. This process involves perforin-dependent elimination of activated CD4 T cells during the first three days of infection. Although this mechanism requires cell-cell contact, NK cells and T cells typically reside in different compartments of lymphoid tissues at steady state. Here, we show that NK-cell suppression of T cells is associated with a transient accumulation of NK cells within T cell-rich sites of the spleen during lymphocytic choriomeningitis virus infection. The chemokine receptor CXCR3 is required for relocation to T-cell zones and suppression of antiviral T cells. Accordingly, this NK-cell migration is mediated by type I interferon (IFN)-dependent promotion of CXCR3 ligand expression. In contrast, adenoviral vectors that weakly induce type I IFN and do not stimulate NK-cell inhibition of T cells also do not promote measurable redistribution of NK cells to T-cell zones. Provision of supplemental IFN could rescue NK-cell migration during adenoviral vector immunization. Thus, type I IFN and CXCR3 are critical for properly positioning NK cells to constrain antiviral T-cell responses. Development of strategies to curtail migration of NK cells between lymphoid compartments may enhance vaccine-elicited immune responses.

immunology↗

bioRxiv2019

Reduced transfer coefficient of carbon monoxide in pulmonary arterial hypertension implicates rare protein-truncating variants in KDR

BackgroundApproximately 25% of patients with pulmonary arterial hypertension (PAH) have been found to harbor rare mutations in disease-causing genes. To identify missing heritability in PAH we integrated deep phenotyping with whole-genome sequencing data using Bayesian statistics. MethodsWe analyzed 13,037 participants enrolled in the NIHR BioResource - Rare Diseases (NBR) study, of which 1,148 were recruited to the PAH domain. To test for genetic associations between genes and selected phenotypes of pulmonary hypertension (PH), we used the Bayesian rare-variant association method BeviMed. ResultsHeterozygous, high impact, likely loss-of-function variants in the Kinase Insert Domain Receptor (KDR) gene were strongly associated with significantly reduced transfer coefficient for carbon monoxide (KCO, posterior probability (PP)=0.989) and older age at diagnosis (PP=0.912). We also provide evidence for familial segregation of a rare nonsense KDR variant with these phenotypes. On computed tomographic imaging of the lungs, a range of parenchymal abnormalities were observed in the five patients harboring these predicted deleterious variants in KDR. Four additional PAH cases with rare likely loss-of-function variants in KDR were independently identified in the US PAH Biobank cohort with similar phenotypic characteristics. ConclusionsThe Bayesian inference approach allowed us to independently validate KDR, which encodes for the Vascular Endothelial Growth Factor Receptor 2 (VEGFR2), as a novel PAH candidate gene. Furthermore, this approach specifically associated high impact likely loss-of-function variants in the genetically constrained gene with distinct phenotypes. These findings provide evidence for KDR being a clinically actionable PAH gene and further support the central role of the vascular endothelium in the pathobiology of PAH.

genetics↗