Biology subjects

Huang, A. C.

Publications and source records attributed to Huang, A. C..

14 recordsLinked to original sources

bioRxiv2026

Deep learning representations of human Immune Health for precision immunology

The human immune system is composed of [~]30-50 distinct cell types, each of which can exist in different states of activation or differentiation. Indeed, the mammalian immune system has evolved to sense and respond to infections, cancers, injuries, and changes in tissue or host homeostasis (1). Moreover, an increasingly large fraction of approved drugs target the immune system directly, and/or cause immune changes (2-4). A key feature of the immune system is to store some of this information, for example as innate or adaptive immune memory (5). In addition, rewiring of immune network architecture induced by disease, environmental exposures, drug treatments, and/or chronological age allows the immune system to store information in the pattern of connections and activity across populations of immune cells. This ensemble information storage, in addition to changes to individual cells, functions as a major way the immune system encodes aspects of immune history and future potential. Genetic information can identify inherited risk alleles, but cannot capture the continual remodeling of the immune system shaped by exposures, infection, inflammation, therapy, and aging (6, 7). To define and use such ensemble immunotypes, we developed a self-supervised deep learning framework that transforms high-dimensional immune profiles into representations of immune health. MAESTRO (MAsked Encoding Set TRansformer with self-distillatiOn) encodes a set of cells from an individual into an embedding that captures immune cell population-level organization. Pretrained on 1,792 peripheral blood samples comprising over 418 million immune cells across 13 clinical diagnoses, MAESTRO learns immune fingerprints that are stable within individuals yet diverse across populations, states of health, disease, and treatment, providing a quantitative basis for comparing immune states across individuals and over time. These fingerprints capture immune architecture beyond coarse cell type proportions, enabling patient-efficient clinical prediction using simple task specific models. MAESTRO model embeddings retain a temporal dimension of immune history and potential, reflecting signatures of past exposures and baseline features that predict future immune responses. Finally, we demonstrate a translational precision immunotherapy application by testing this approach in metastatic Pancreatic Ductal Adenocarcinoma (PDAC), where pretreatment immune landscape circuitry maps enable patient stratification and therapeutic response prediction. Overall, we developed a large, attention-based model that captures deep network architecture of immune states through self- supervised representations of immune cytometry data as a reusable foundation for precision immunology, converting immune complexity into clinically actionable embeddings for diagnosis, monitoring, and therapy selection.

immunology↗

bioRxiv2026

Lymphatic egress recycles tumor-experienced effector CD8 T cells to sustain immune surveillance

Successful anti-tumor immune surveillance depends on stem-like CD8+ T cells that are enriched in tumor-draining lymph nodes (LN), but how they are maintained over time remains poorly understood. Here, we identify a continuous lymphatic circuit that sustains stem-like CD8+ T cells. Using photoconversion to fate-map intratumoral T cells we demonstrate that effector cells exit the tumor microenvironment and migrate back to the draining LN. These tumor-specific, migratory effector T cells avoid chronic antigen stimulation, re-express the transcription factor associated with self-renewal, TCF1, and enter a stem-like state in the LN. Antigen presentation in LNs by dendritic cells drives their proliferation thereby inflating the LN stem-like population. Consequently, maintenance of stem-like T cells and ICB response depends on constitutive lymphatic transport, while LN metastasis compromises the stem-like niche, diminishing ICB response. We, therefore, define a continuous, peripheral lymphatic circuit that recycles tumor-experienced effector T cells to fuel durable, systemic immune surveillance.

immunology↗

bioRxiv2026

X-Cell: Scaling Causal Perturbation Prediction Across Diverse Cellular Contexts via Diffusion Language Models

Causal models of cellular systems hold the promise to empower broad biological discovery, including the systematic identification of novel targets for drug discovery. Predicting how genetic and pathway perturbations reshape gene expression across diverse cellular contexts is a prerequisite for building generalizable cellular foundation models. However, current methods typically fail to extrapolate beyond their training distributions because they rely predominantly on observational expression atlases rather than interventional perturbation data. We present X-Atlas/Pisces, the largest genome-wide CRISPRi Perturb-seq compendium to date, comprising 25.6 million perturbed single-cell transcriptomes across 16 biologically diverse contexts, including widely used cell lines, induced pluripotent stem cells (iPSCs), resting and CD3/CD28 activated Jurkat T lymphoma cells, and multi-lineage differentiating iPSCs. Leveraging this resource, we develop X-Cell, a diffusion language model that predicts perturbation responses by iteratively refining control-to-perturbed state transitions through cross-attention to multi-modal biological priors derived from natural language, protein language models, interaction networks, genetic dependency maps, and morphological profiles. X-Cell outperforms existing state-of-the-art models by up to five-fold on key metrics such as Pearson{Delta} (correlation between predicted and observed perturbation-induced log-fold changes), and demonstrates zero-shot prediction of T cell inactivating perturbations in stimulated Jurkat cells. We scale X-Cell to 4.9 billion parameters (X-Cell-Ultra), the largest causal perturbation model to date. We demonstrate for the first time that perturbation prediction follows power-law scaling with an exponent matching large language models. X-Cell-Ultra demonstrates zero-shot generalization to novel biological contexts, including unseen iPSC-derived melanocyte progenitors and primary human CD4+ T cells from multiple donors, and outperforms all baselines after self-supervised test-time adaptation. These results demonstrate that coordinated scaling of causal perturbation data and model capacity yields foundation models capable of generalizable perturbation prediction across cellular contexts, with potential applications for improving computational target identification, validation, and context-specific therapeutic prioritization.

systems biology↗

bioRxiv2026

Harnessing Inflammatory Monocytes to Overcome Resistance to Anti-PD-1 Immunotherapy

BackgroundResistance to immune checkpoint inhibitors represents a major therapeutic challenge, as less than 50% of patients with melanoma achieve long-term response to immune checkpoint inhibitor therapy. One mechanism of acquired resistance involves somatic mutations, such as loss of beta-2 microglobulin (B2m), that enable tumor cells to evade T cell-mediated killing. MethodsThis study used single-cell RNA-seq, flow cytometry, and ex vivo functional assays to characterize tumor-infiltrating immune cells in antigen presentation-deficient tumors. Tumor-bearing mice were treated with anti-PD-1 or CD40 agonist antibodies and cell depletion or cytokine blocking antibodies to define mechanisms of action. Analysis of published human RNA-seq datasets was performed to dissect the contributions of inflammatory monocytes to patient outcomes. ResultsWe found an increase in immunosuppressive macrophages in B2m-null tumors. We hypothesized that repolarizing myeloid cells may restore control of tumor growth. Treatment with CD40 agonist antibody, which promotes differentiation of monocytes and macrophages towards a proinflammatory phenotype, reduced tumor growth and improved survival in B2m-null melanoma and colorectal cancer models. Unexpectedly, both CD8+ T cells and NK cells, but not CD4+ T cells, were required for the efficacy of CD40 agonist, even though CD8+ T cells cannot directly recognize antigen presentation-deficient tumor cells. Instead, these lymphocytes control tumor growth via secretion of IFN{gamma}, as depletion of IFN{gamma} inhibited the therapeutic effect of CD40 agonist. IFN{gamma} receptor (Ifngr1) expression was required on host cells, not tumor cells, for CD40 agonist-mediated tumor control. Single-cell analysis identified a distinct population of inflammatory monocytes that were enriched for an IFN{gamma} response signature in CD40 agonist-treated tumors, suggesting that these cells may be important for tumor control. Analysis of human bulk and single-cell RNA-seq datasets demonstrated that an inflammatory monocyte signature derived from our data was associated with improved patient outcomes and response to immune checkpoint inhibitors. ConclusionsThese data demonstrate that CD8+ T cells contribute to tumor control even in the absence of direct antigen presentation by tumor cells. More broadly, our work suggests that strategies to activate the effector functions of inflammatory monocytes may limit tumor growth and overcome acquired resistance to immune checkpoint inhibitors.

immunology↗

bioRxiv2026

Precision CRISPR annotation of the functional enhancer landscape in primary human T cells

Precise modulation of T cell function through engineering the non-coding genome holds great promise for advancing next-generation immunotherapies. However, robust high-throughput approaches to annotate functional cis-regulatory elements (CRE) in human T cells remain limited. Here, we developed a simple and highly efficient CRISPR interference (CRISPRi) perturbation platform to systematically annotate CREs in human primary T cells. Using this platform, we identified novel CREs controlling PDCD1, HAVCR2, and TBX21 expression. Combinatorial CRE perturbations revealed synergistic CRE pairs that fine-tune PDCD1 and HAVCR2 expression, while Cas9-indel-based mutagenesis pinpointed the critical nucleotides within each enhancer that are essential for their activity. Functional experiments demonstrated that CRE-edited HAVCR2 outperformed conventional total gene knockout in enhancing CAR T cell anti-tumor efficacy. Moreover, CRE editing of PDCD1 and HAVCR2 repressed PD-1 and TIM-3 expression in human tumor-infiltrating lymphocyte CD8 T cells, highlighting regulatory role of these CREs in disease relevant exhausted T cells. Together, this approach offers a compact CRISPRi platform that enables high-throughput dissection of functionally relevant non-coding genomic regions in T cells, providing insights for mechanistic studies and precision genome engineering of advanced cellular therapies.

immunology↗

bioRxiv2025

Cellular interactions in the sentinel lymph node predict melanoma recurrence

Melanoma outcomes have dramatically improved over the past decade, but some patients still experience disease recurrence, particularly those who present at later stage of disease. Prior studies identified an altered immune microenvironment in the sentinel lymph node (SLN) including the presence of dysfunctional CD8 T cells and CD4 T regulatory cells (Tregs), but the spatial organization of the SLN and the interactions of individual cell types have not been extensively studied. To understand how spatial organization of immune cells in the SLN is related to outcomes, we performed spatial proteomic profiling of Stage I and II melanoma SLN to generate an atlas of cell types. Following multiplexed immunofluorescence imaging, deep learning-based segmentation and clustering, we identified 33 subsets of T cells, B cells and Tumor cells. Lymphoid region analyses established a foundational spatial map of the melanoma SLN, revealing higher regional frequencies of activated and memory CD4 T cells in Stage I SLN and consistent positioning of T cells at functional spatial locations. To better understand cellular interactions, we evaluated the immediate neighbors around each index cell using Effect Size Interaction mapping (ESI-map), a novel computational toolkit for understanding spatial interactions. Nearest-neighbor analyses across biological replicates revealed rewiring in cell-cell interacting pairs across disease progression and patient outcomes, particularly with respect to exhausted TOX+ CD8 T cells. Stage II patients who experienced disease recurrence had notable enrichment of cellular interactions between Tregs and TOX+ CD8 T cells that was associated with reduced expression of granzyme B and interferon-{gamma} that was most pronounced in exhausted CD8 T cells. Together, these data demonstrate that the specific spatial interactions between Tregs and exhausted CD8 T cells in the SLN provide a novel immune signature for understanding melanoma outcomes. One Line SummaryTreg-CD8 T cell interactions in the sentinel lymph node predict recurrence outcomes in Stage II SLN+ melanoma patients.

immunology↗

bioRxiv2025

Divergent ontogeny of Tissue Resident Memory and Tissue Resident Exhausted CD8+ T cells underlies distinct functional potential

Persistent antigen stimulation promotes differentiation of exhausted CD8+ T (TEX) cells. TEX cells are distinct from circulating memory T (TCIRCM) cells but share many features with tissue-resident memory (TRM) cells established following infection resolution. CD8+ T cells co-expressing residency- and exhaustion-associated molecules in chronic diseases often correlate with clinical outcomes. However, the relationship between these cells and conventional TRM or TEX cells remains unclear. Here, we show that chronic antigen stimulation drives development of tissue-resident TEX (TR-TEX) cells that are ontologically and functionally distinct from TRM cells generated after antigen clearance. TR-TEX phenotypically resembled TRM cells but were regulated by distinct transcriptional networks and were uniquely dependent on Tox for residency programming. Although TEX progenitor cells acquired residency features upon entering chronically infected tissues, they failed to generate conventional TRM cells after antigen withdrawal. Conversely, TRM cells were able to differentiate into TEX cells during chronic antigen stimulation. Deriving cell-state specific transcriptional signatures revealed a selective association of TR-TEX cells with patient responses to immune checkpoint blockade, and only TR-TEX but not TRM cells responded to PD-1 pathway inhibition in vivo. These data suggest that TR-TEX and TRM cells are developmentally distinct cell types that share a tissue-residency program but have distinct roles in disease control.

immunology↗

bioRxiv2025

X-Atlas/Orion: Genome-wide Perturb-seq Datasets via a Scalable Fix-Cryopreserve Platform for Training Dose-Dependent Biological Foundation Models

The rapid expansion of massively parallel sequencing technologies has enabled the development of foundation models to uncover novel biological findings. While these have the potential to significantly accelerate scientific discoveries by creating AI-driven virtual cell models, their progress has been greatly limited by the lack of large-scale high-quality perturbation data, which remains constrained due to scalability bottlenecks and assay variability. Here, we introduce "Fix-Cryopreserve-ScRNAseq" (FiCS) Perturb-seq, an industrialized platform for scalable Perturb-seq data generation. We demonstrate that FiCS Perturb-seq exhibits high sensitivity and low batch effects, effectively capturing perturbation-induced transcriptomic changes and recapitulating known biological pathways and protein complexes. In addition, we release X-Atlas: Orion edition (X-Atlas/Orion), the largest publicly available Perturb-seq atlas. This atlas, generated from two genome-wide FiCS Perturb-seq experiments targeting all human protein-coding genes, comprises eight million cells deeply sequenced to over 16,000 unique molecular identifiers (UMIs) per cell. Furthermore, we show that single guide RNA (sgRNA) abundance can serve as a proxy for gene knockdown (KD) efficacy. Leveraging the deep sequencing and substantial cell numbers per perturbation, we also show that stratification by sgRNA expression can reveal dose-dependent genetic effects. Taken together, we demonstrate that FiCS Perturb-seq is an efficient and scalable platform for high-throughput Perturb-seq screens. Through the release of X-Atlas/Orion, we highlight the potential of FiCS Perturb-seq to address current scalability and variability challenges in data generation, advance foundation model development that incorporates gene-dosage effects, and accelerate biological discoveries.

genomics↗

bioRxiv2025

DR5 CAR-T cells target solid tumors and suppress MDSCs with minimal toxicity

Chimeric antigen receptor (CAR) T cell therapies have poor efficacy in solid tumors due to limited target specificity and an immunosuppressive tumor microenvironment. We investigated death receptor 5 (DR5) as a CAR target based on its high expression in both solid tumors and myeloid- derived suppressor cells (MDSCs). We engineered agonistic DR5-specific CAR constructs and evaluated their activity in multiple models, demonstrating DR5-expression-dependent tumor killing, confirmed by knockout and overexpression experiments. DR5-targeting single-chain variable fragments retained their pro-apoptotic activity when expressed on non-effector cells or extracellular vesicles. Among multiple CAR designs, we identified a construct with optimized binding affinity that maintained T cell viability while preserving strong tumor and MDSC-killing potency. To assess safety and efficacy in an immunocompetent setting, we also developed a murine DR5-targeted CAR. In multiple xenograft and syngeneic mouse models, DR5 CAR-T cells reduced tumor growth, prolonged survival, and did not cause detectable toxicity. In patient- derived organoids and tissue slices, DR5 CAR-T cells infiltrated tumor tissues, reduced MDSCs, boosted CD8+ T cell activity, and inhibited tumor growth. These findings support DR5-targeted CAR-T therapy as a promising strategy for treating solid tumors, which combines direct tumor cytotoxicity with immune activation and minimizes off-target effects. TeaserDR5 CAR-T cells eliminate tumor cells and MDSCs and activate tumor-resident CD8+ T cells within the TME.

cancer biology↗

bioRxiv2023

Prior vaccination enhances immune responses during SARS-CoV-2 breakthrough infection with early activation of memory T cells followed by production of potent neutralizing antibodies

SARS-CoV-2 infection of vaccinated individuals is increasingly common but rarely results in severe disease, likely due to the enhanced potency and accelerated kinetics of memory immune responses. However, there have been few opportunities to rigorously study early recall responses during human viral infection. To better understand human immune memory and identify potential mediators of lasting vaccine efficacy, we used high-dimensional flow cytometry and SARS-CoV-2 antigen probes to examine immune responses in longitudinal samples from vaccinated individuals infected during the Omicron wave. These studies revealed heightened Spike-specific responses during infection of vaccinated compared to unvaccinated individuals. Spike-specific CD4 T cells and plasmablasts expanded and CD8 T cells were robustly activated during the first week. In contrast, memory B cell activation, neutralizing antibody production, and primary responses to non-Spike antigens occurred during the second week. Collectively, these data demonstrate the functionality of vaccine-primed immune memory and highlight memory T cells as rapid responders during SARS-CoV-2 infection.

immunology↗

bioRxiv2022

Enhanced STAT5a activation rewires exhausted CD8 T cells during chronic stimulation to acquire a hybrid durable effector like state

Rewiring exhausted CD8 T cells (TEX) towards more functional states is a major goal of cancer immunotherapy but has proven challenging due to the epigenetic stability of TEX. Indeed, TEX are epigenetically programmed by the transcription factor Tox. However, epigenetic changes continue to occur as TEX transition from progenitor (TEXprog), to intermediate (TEXint) and terminal (TEXterm) subsets, suggesting potential developmental flexibility in mature TEX subsets. By examining the transition of TEXprog into TEXint cells, we discovered a reciprocally antagonistic circuit between Stat5a and Tox in TEX cells. Stat5-activity controlled TEXint development, antagonized Tox, and instigated partial effector biology. Stat5 was also essential for TEX reinvigoration by PD-1 blockade. Indeed, temporal induction of Stat5-activity in TEX using an orthogonal IL-2/IL2R{beta}-pair fostered TEXint cell accumulation and synergized with PD-L1 blockade. Constitutive Stat5a activity (STAT5CA) antagonized Tox-dependent TEX epigenetic programming to generate a durable hybrid effector/NK-like population with enhanced tumor control. Finally, enforcing Stat5-signals in established TEXprog partially rewired the TEX epigenetic landscape towards the effector/memory lineage. Together, these data highlight therapeutic opportunities of manipulating Stat5 to rewire TEX towards a durably protective hybrid program.

immunology↗

bioRxiv2022

Induction of a CD8 T cell intrinsic DNA damage and repair response is associated with clinical response to PD-1 blockade in uterine cancer.

Despite the success of immune checkpoint blockade (ICB), many patients fail to achieve durable clinical benefit, and the underlying immunological mechanisms remain poorly understood. Here, we investigated immune reinvigoration by ICB in advanced or recurrent hypermutated or microsatellite instability-high, mismatch repair deficient (MSI-H/MMRd) uterine cancer patients treated with anti-PD-1 (nivolumab). CD8 T cells underwent rapid pharmacodynamic proliferation 2-4 weeks after initiating PD-1 blockade. This immunological response, however, did not correlate with clinical response. We hypothesized that the T cell-intrinsic response to proliferative and genotoxic stress might contribute to the disparity between immunological and clinical response. We developed a high-dimensional single cell cytometric platform to simultaneously analyze T cell differentiation with changes in DNA damage and repair (DDR) pathways. This DDR-Immune platform revealed T cell subset-specific patterns of DDR, and distinct DDR pathways induced by different types of DNA damage. Applying this platform to MSI-H/MMRd or hypermutated uterine cancer patients revealed a signature of DDR exemplified by rapid increase in phosphorylated-ATM (pATM) intrinsic to CD8 T cells proliferating in response to PD-1 blockade that distinguished clinical responders and non-responders. ATM regulated transcriptional circuits in T cells were associated with better clinical response to PD-1 blockade. These findings highlight a previously unrecognized role for CD8 T cell-intrinsic DDR as a potential determinant of immune fitness and clinical outcome of PD-1 blockade therapy. STATEMENT OF SIGNIFICANCEUsing high-dimensional immune and DNA damage response (DDR) profiling in T cell subsets we identified distinct signatures in anti-PD-1 treated MSI-H/hypermutated uterine cancer patients associated with clinical response versus non-response. Thus, T cell-intrinsic DDR is a potential determinant of immune responsiveness and clinical outcome to PD-1 blockade therapy in cancer.

immunology↗

bioRxiv2021

Signaling through FcγRIIA and the C5a-C5aR pathway mediates platelet hyperactivation in COVID-19

Patients with COVID-19 present with a wide variety of clinical manifestations. Thromboembolic events constitute a significant cause of morbidity and mortality in patients infected with SARS-CoV-2. Severe COVID-19 has been associated with hyperinflammation and pre-existing cardiovascular disease. Platelets are important mediators and sensors of inflammation and are directly affected by cardiovascular stressors. In this report, we found that platelets from severely ill, hospitalized COVID-19 patients exhibit higher basal levels of activation measured by P-selectin surface expression, and have a poor functional reserve upon in vitro stimulation. Correlating clinical features to the ability of plasma from COVID-19 patients to stimulate control platelets identified ferritin as a pivotal clinical marker associated with platelet hyperactivation. The COVID-19 plasma-mediated effect on control platelets was highest for patients that subsequently developed inpatient thrombotic events. Proteomic analysis of plasma from COVID-19 patients identified key mediators of inflammation and cardiovascular disease that positively correlated with in vitro platelet activation. Mechanistically, blocking the signaling of the Fc{gamma}RIIa-Syk and C5a-C5aR pathways on platelets, using antibody-mediated neutralization, IgG depletion or the Syk inhibitor fostamatinib, reversed this hyperactivity driven by COVID-19 plasma and prevented platelet aggregation in endothelial microfluidic chamber conditions, thus identifying these potentially actionable pathways as central for platelet activation and/or vascular complications in COVID-19 patients. In conclusion, we reveal a key role of platelet-mediated immunothrombosis in COVID-19 and identify distinct, clinically relevant, targetable signaling pathways that mediate this effect. These studies have implications for the role of platelet hyperactivation in complications associated with SARS-CoV-2 infection. Cover illustration O_FIG_DISPLAY_L [Figure 1] M_FIG_DISPLAY C_FIG_DISPLAY One-sentence summaryThe Fc{gamma}RIIA and C5a-C5aR pathways mediate platelet hyperactivation in COVID-19

immunology↗

bioRxiv2020

Deep immune profiling of COVID-19 patients reveals patient heterogeneity and distinct immunotypes with implications for therapeutic interventions

COVID-19 has become a global pandemic. Immune dysregulation has been implicated, but immune responses remain poorly understood. We analyzed 71 COVID-19 patients compared to recovered and healthy subjects using high dimensional cytometry. Integrated analysis of [~]200 immune and >30 clinical features revealed activation of T cell and B cell subsets, but only in some patients. A subgroup of patients had T cell activation characteristic of acute viral infection and plasmablast responses could reach >30% of circulating B cells. However, another subgroup had lymphocyte activation comparable to uninfected subjects. Stable versus dynamic immunological signatures were identified and linked to trajectories of disease severity change. These analyses identified three "immunotypes" associated with poor clinical trajectories versus improving health. These immunotypes may have implications for therapeutics and vaccines.

immunology↗