Nature Biomedical Engineering

Synthetic medical data are commonly generated from manually chosen prompts, inverse frequency sampling, or fixed subgroup balancing rules. These strategies increase sample count but do not answer a more consequential question: which missing observations would most improve the current model? We introduce GAP DREAM, a closed-loop framework that treats a medical image generator as a queryable source of candidate evidence rather than an unlimited source of augmentation. At each round, GAP DREAM maps model failures into an evidence gap score that combines scarcity, predictive uncertainty, local representation density, and gradient novelty. A query policy translates high-value gaps into structured clinical requests, a modality specific generator proposes candidate imagetext pairs, and an independent selector retains only diagnostically consistent, diverse, non-duplicative candidates. The downstream model is then updated and the gap map recomputed. The central objective is therefore not image realism in isolation, but reduction of measurable evidence gaps under a fixed synthetic budget. Across a multi-modality protocol spanning long-tailed chest radiography, dermoscopy, clinical dermatology, and histopathology, GAPDREAM reaches stronger held-out real-data performance with substantially fewer generated samples than random,class-balanced, uncertainty-only, and static targeted augmentation. Results show gains in CXRLT tail mAP, HAM10000 macroAUROC, Came lyon17 out-of-domain accuracy, and worst group dermatology accuracy while reducing redundant synthetic samples. GAPDREAM reframes medical generation as active evidence acquisition: generate the examples the learner needs, not simply more examples of what is rare.

Engineered cardiac and skeletal muscle tissues suspended between flexible posts support disease modeling and pharmacology, yet their stimulation, longitudinal imaging and quantitative analysis often remain fragmented and labor-intensive. Here we present Stimulatrix, an open-source platform integrating automated video acquisition within a cell culture incubator, synchronized electrical stimulation and deep learning for longitudinal assessment of contractile tissues in multiwell plates. The platform quantifies tissue compaction, force generation and contraction kinetics with optional cloud processing reducing dependence on local GPU hardware. We demonstrate the workflow in cardiac tissues comprising human induced pluripotent stem cell-derived cardiomyocytes and cardiac fibroblasts, and in primary human skeletal muscle constructs. Force and kinetic measurements were benchmarked against manual annotations. Longitudinal profiling resolved responses to matrix composition and pacing regimens and tracked doxorubicin-associated loss of cardiac contractile force. Stimulatrix provides an accessible workflow for automated functional phenotyping of engineered muscle tissues.

Cytokines can elicit potent antitumor immunity, but their clinical use is often limited by systemic immune activation and dose-limiting toxicity. Protease-responsive masking offers a way to improve selectivity by linking cytokine activity to disease-associated protease cleavage, but current approaches have faced several challenges. Here we describe a de novo protein design strategy for generating protease-activatable cytokine prodrugs that overcomes these challenges. We use this strategy to design and characterize cytokine prodrugs and well-behaved AND-gated split systems that require both target-dependent colocalization and proteolytic unmasking to generate active cytokines. In challenging syngeneic tumor models, our IL-21 prodrugs reduced treatment-associated toxicity while maintaining antitumor efficacy.

Small numbers of cells can determine the fate of an entire organism. For example, metastatic cancer cells seed distant organs, therapy-resistant clones repopulate tumors, and T cells infiltrate and expand in tissues with matching antigens. Visualizing the activity of such cells in living organisms requires highly sensitive, specific and repeatable imaging in optically opaque tissues. Here we introduce cell chirp, an ultrasound imaging method for real-time imaging of single cells in living organs inaccessible to light. Cell chirp exploits the harmonic response of genetically encoded gas vesicles to frequency-coded ultrasound to sensitively and specifically visualize single cells in scattering background. We first develop this method in tissue-mimicking phantoms, demonstrating the ability to quantitatively detect single cells as validated by optical cell counts. We then establish the ability of cell chirp to image thousands of genetically labeled cells within a million-cell background tissue, track reporter gene expression over days, and suppress background signals by more than two orders of magnitude. Finally, we use cell chirp to track the fate of antigen-positive or antigen-negative tumor-cell minorities during weeks-long T-cell engager treatments. Cell chirp reveals the dynamics of sparsely distributed, genetically defined cells in opaque tissues.

N-linked Fc galactosylation drives variability in therapeutic mAb quality, affecting complement-dependent cytotoxicity and ADCC. Current control strategies suffer from narrow ranges, productivity loss, or unwanted glycoforms. Here, we combine metabolic and cellular glycoengineering by feeding 2-deoxy-2-fluorogalactose (2FG) to CHO cells engineered for hypergalactosylation via COSMC knockout and {beta}4-galactosyltransferase overexpression (C /GT+), achieving broad, tuneable control of Fc {beta}4-galactosylation. In DP12 C-/GT+ cells, a split-feed regimen spanned a wide galactosylation range (95-46%) but generated substantial aglycosylated and Man5 glycoforms. The same regimen in a second host, VRC01 C-/GT+, gave only a narrow range, indicating 2FG dosing does not transfer directly across cell lines. Based on a proposed mechanism where 2FG is metabolised via the Leloir pathway into inhibitory fluorinated nucleotide-sugar intermediates, we developed a cell-normalised, multi-bolus feeding strategy that substantially broadened control in VRC01 C-/GT+ cells while suppressing aglycosylation and Man5 formation. In a fed-batch Ambr(R) 250 process, this strategy achieved comparable galactosylation reduction with roughly ten-fold less 2FG, showing that the fraction of mAb secreted after feeding is a key determinant of control strategy efficiency. Across the resulting glycoform range, Bio-Layer Interferometry identified Fc {beta}4-galactosylation as the strongest correlate of Fc{gamma}RIIIA (CD16a) binding affinity (R2 = 0.92). Together, these findings establish combined metabolic and cellular glycoengineering as a platform for controlling mAb galactosylation across a broad range, with potential to fine-tune downstream pharmacological activity.

Genetically modified immune effector cells, including chimeric antigen receptor (CAR) T cells, have emerged as transformative anti-cancer agents. However, these autonomous, living therapies are limited by difficult-to-control hyperactivation toxicities. To achieve their full potential as therapies for a broader range of diseases, CAR T cells must target novel antigens and integrate further potency enhancements that raise the possibility of novel and intensified toxicities. High performance safety systems are needed to restrain these highly active next-generation gene- and cell-based therapies. Cell therapy suicide switches remain infrequently used, and are currently limited by inaccessible controller drugs, incomplete cell depletion, fitness cost, and immunogenicity. To address these issues, we engineered a fully human sequence-based cell therapy suicide switch based on the overexpression of Caspase-Activated DNase (CAD) and its stoichiometric chaperone/inhibitor ICAD fused to a lenalidomide-inducible degron. Acute destabilization of ICAD with the broadly available, off-patent molecular glue degrader drug lenalidomide uncaged CAD to induce DNA damage and cell death. By this degradable cage mechanism, the ICAD-CAD suicide switch achieved long-term expression in primary human T cells without detectable leakiness, and deep drug-induced cell depletion. The ICAD-CAD switch was generalizable to alternative acute destabilization systems. Genome-wide CRISPR screening established the necessity of CAD and the lenalidomide-interacting E3 ubiquitin ligase substrate adaptor CRBN for switch function. When co-delivered with a CAR, ICAD-CAD CAR T cells retained effector functions and could be rapidly depleted with lenalidomide.

Pancreatic ductal adenocarcinoma (PDAC) arises within an extensively remodeled microenvironment, yet tissue markers that distinguish invasive carcinoma from precursor lesions remain limited. Integrating four single cell RNA sequencing atlases comprising 324,662 cells from 26 patients, we identified cancer associated fibroblast expansion, accumulation of C1QA+ immunoregulatory macrophages, and depletion of a dendritic cell compartment expressing plasmacytoid markers. Fibroblasts were enriched for extracellular matrix and PI3K AKT programs and macrophages for an antigen-presentation program. We then examined the protein level correlates of these populations in a tissue proteomic reference framework of 356 samples spanning non diseased pancreas, normal-adjacent tissue, low and high grade intraductal papillary mucinous neoplasms (IPMNs), IPMN associated invasive carcinoma, and PDAC, prioritizing proteins that changed monotonically across the six histopathologic reference groups or marked tumor enriched cell types. Despite depletion of plasmacytoid dendritic cells, the principal source of type I interferon, the interferon stimulated protein ISG15, expressed by both fibroblasts and macrophages, increased monotonically toward PDAC. In exploratory cross-validated benchmarking of PDAC versus low and high grade IPMN, excluding IPMN-associated invasive carcinoma (145 versus 56 samples), CATB+EPIPL achieved an AUC of 0.98, with a sensitivity of 92% at an achieved specificity of 92%, exceeding protein abundance models of diagnostic IHC-associated markers (S100P+IMP3+maspin+MSLN, 0.79) and refitted published IPMN progression combinations (<0.91). Against non diseased pancreas, candidate and comparator panels performed similarly (AUC 0.99 1.00). These biologically informed tissue protein panels warrant independent validation, including immunohistochemical evaluation of EUS guided fine needle biopsy specimens and measurement in blood.

Fluorescence microscopy is a central tool in biomedical image analysis, but its practical use is often constrained by incomplete and heterogeneous data. This is especially true in rare disease research, where patient cohorts are small, biological material is limited, and complex study designs frequently yield images with missing, degraded, or low-quality channels. We introduce Reliability-Mask-Conditioned DDPM (RM-DDPM), a conditional diffusion model that formulates multichannel image generation, missing-channel imputation, and degraded-channel restoration as a single reliability-guided inference problem. The model is trained on a curated subset of complete, high-quality multichannel images using a self-supervised masking objective to learn rich cross-channel dependencies from the available channels. At inference time, reliable observations condition the reverse diffusion process, allowing the same trained model to solve multiple reconstruction tasks without task-specific retraining. We evaluate RM-DDPM against task-specific baselines on one public and two in-house fluorescence microscopy datasets, comprising a rare-disease preclinical model study and a clinical patient cohort. Across these settings, RM-DDPM matches the specialized baselines in reconstruction quality, hallucinates fewer false structures, and remains more robust as degradation severity increases, while repeated sampling provides per-pixel uncertainty maps that flag unreliable reconstructions.

Reconstituting organ development in vitro requires reproducing dynamic developmental transitions rather than only terminal tissue structures. Here, we developed MoSAIC (Modular Spatial Assembly and Integration of Cues), a modular platform that decomposes complex culture operations and subsequently integrates them to reconstruct multimodal mechano-chemical spatial cues. Using human induced pluripotent stem cells, MoSAIC enabled independent control of tissue geometry, extracellular matrix localization, physical boundaries, and morphogen gradients, while also permitting defined mechanical and chemical perturbations during development. Their integration guided a continuous ectodermal sheet through neural induction, tissue bending, and bifurcation into neural and epidermal tissues, followed by formation of a polarized neural tube structure. Spatially controlled morphogen exposure further established dorsoventral patterning, while the resulting tissue exhibited posterior neural tube identity. Using the developed model recapitulating the transition of human neurulation, MoSAIC further enables non-genetic perturbation of mechanical and chemical cues to interrogate the mechanisms governing neural tube closure.

Functional ultrasound (fUS) imaging provides deep, wide-field access to brain function, but instrument cost and complex acquisition and analysis workflows restrict its wider adoption. Here we introduce @fUS, an agentic transcranial imaging platform with an open architecture that integrates purpose-built hardware with executable experimental skills and persistent structured memory. The system combines 128-channel, 16-bit acquisition at 125 MHz with GPU-based processing and enables functional imaging through the intact scalp and skull of mice, with approximately 100-m spatial resolution and 10-Hz temporal sampling, providing a basis for longitudinal studies without cranial-window surgery. Its agent connects scientific objectives to experimental design, direct instrument control and data analysis, retaining context across interactions and supporting inspectable workflows through text and hands-free voice control. Neuroscience trainees with limited engineering experience independently configured the system within 30 min. Agent-assisted exploration of whisker-stimulation datasets revealed low-frequency vascular changes beyond conventional response mapping, supported by complementary two-photon measurements of single-vessel diameter. By combining transcranial imaging with accessible instrument control and analysis, @fUS provides a foundation for wider adoption and larger, more diverse neuroscience datasets. More broadly, it offers a framework for scientific instruments in which measurement, computation and experimental reasoning are developed as parts of the same system.

Fixed-backbone sequence discovery, or inverse folding, is a critical recurring task in the development of new polypeptide therapeutics. Once promising backbones are established for a target pocket, computational inverse folding methods greatly help accelerate generation of candidate sequences. Such methods are mature for the traditional case of limiting to the fixed twenty-letter canonical vocabulary; however, they cannot access the broader space of non-canonical amino acids (NCAAs). This design constraint is exacerbated for peptide binders, a fast-growing modality that readily incorporates NCAAs, though in practice non-canonical design frequently depends on laborious medicinal-chemistry campaigns. An extension of inverse folding to NCAAs is thus critical to accelerating design of novel therapeutic peptides. AtomWeaver uses a joint all-site, atom-level generative scheme that does not restrict side-chain categorical assignment by either predetermined or co-resolving residue identity. Conditioned only on a fixed peptide backbone and its target protein, its multi-component flow guides side-chain atoms as unlabeled points in R3 from a nested shell prior to a variable-count final atom cloud. Identity is then read by matching each predicted cloud against a reference library of canonical and non-canonical templates. Since identity is decided only at decode time, the addressable vocabulary is a property of the library rather than of the trained weights: a new NCAA costs one reference structure and no retraining, and the model can select residues it was never prompted for and never saw in training. On a deep mutational scan of two peptide-target systems, AtomWeaver's canonical readout shows high observed mean agreement with experimental values among the compared inverse-folding methods. In the mixed canonical-noncanonical setting that canonical-only baselines cannot support at all, it likewise retains ranking signal across both systems. AtomWeaver also displayed self-consistent designs on de novo binder backbones, with the highest interface confidence among compared methods. Notably, it reached these metrics while achieving broad empirical coverage of our 300-residue vocabulary, including four non-canonical types never visible in training. AtomWeaver thus serves canonical and non-canonical peptide design alike, while transforming residue vocabulary to an expandable inference-time choice.

The eradication of solid tumors by chimeric antigen receptor (CAR) T cells requires dynamic therapies capable of outlasting an immune suppressive tumor microenvironment (TME). However, biological barriers--including rapid exhaustion, poor expansion, and loss of stem-like memory--quickly neutralize these therapies. Because single-technology interventions often introduce unacceptable tradeoffs between efficacy and safety, durable remission demands a paradigm where multiple engineered solutions work in concert. To holistically address these mechanisms, we rationally designed a single-vector transgene that intrinsically drives CAR T functional persistence. The platform integrates four synergistic technologies: a high-avidity mesothelin (MSLN)-targeting CAR optimized to resist shed decoy antigens, a T-cell activation-responsive promoter (OUTLAST OP1) resisting exhaustion, a CD8-targeted designed IL-2 cytokine (OUTSMART dIL-2) driving intratumoral CAR-T expansion, and an EGFRopt safety switch. In lung and ovarian cancer models, this rational integration was required to drive antigen-dependent T cell expansion and eradicate established tumors at extremely low CAR T doses. Furthermore, engineered cells established a self-renewing pool of stem-like memory T cells capable of rejecting tumor rechallenge months later. Ultimately, this work demonstrates that intrinsic T cell dysfunction and extrinsic tumor-derived barriers can be simultaneously overcome by integrating synergistic technologies.

Stereotactic needle-based delivery remains the standard for local administration of Advanced Therapy Medicinal Products (ATMPs) to the human brain. ATMP administration typically involves multiple trajectories, presenting cumulative risks and prolonging surgery. Reflux-prone, patchy therapy coverage compromises clinical results. We demonstrate a novel approach, deploying a helical delivery catheter via a single access trajectory per target, referred to as SpiraMed. Helix retraction is synchronised with therapy delivery, enabling comprehensive target coverage in seconds. Helix pitch and diameter can be precisely tailored to patient-specific target volume and vascular anatomy, as part of the preoperative stereotactic surgical planning process. Testing in agarose phantoms, live sheep and cadaveric human brain confirms enhanced therapy delivery volume, delivery speed and target coverage, with reductions in reflux and predicted risk of bleeding complication. SpiraMed represents a new paradigm, promising to help deliver on the transformative potential of cell and gene therapies across the spectrum of human CNS disease.

A densely vascularized tissue can still be starved of oxygen. Tissue oxygenation depends heavily on vessel arrangement, blood flow distributions, and cellular metabolism. Resolving these effects from individual constituent vessels to whole organs has historically been computationally prohibitive. We introduce the Cardiovascular Analysis and Simulation of Circulatory Advection, Diffusion, and Exchange (CASCADE) toolkit, which makes vessel-resolved oxygen transport tractable at the organ scale through a novel screened Green's-function formulation. On laptop-class hardware, CASCADE simulates networks containing thousands of vessels in under one second and a full biventricular human heart model in about two minutes. We validate spatial oxygen and cell-survival predictions against experiments in perfused engineered tissues and matched finite-element simulations. In engineered tissues, sweeps across thousands of network designs and physiological conditions reveal when vessel density, oxygen supply, or network architecture limits viability, establishing design rules for sustaining thick constructs. In a human heart model, CASCADE translates coronary occlusion, anemia, polycythemia, and hypoxemia into distinct myocardial oxygenation patterns, revealing where oxygen delivery fails and how much tissue is at risk.

Capturing fast hemodynamics in deep organs is essential for understanding microvascular regulation of organ responses to physiological demands and pathological stress. However, non-invasive three-dimensional (3D) imaging of these microscale processes remains challenging due to trade-offs between spatial resolution, imaging speed, and penetration depth. Here, we present fluctuation-based high-order super-resolution acoustic microscope (FLAME), a tracking-free 3D ultrasound imaging technology capable of fast microvascular angiography and flow measurement. Using as few as 30 volumes, FLAME improves 3D ultrasound resolution by 8-fold (~50 m) and shortens data acquisition time by over two orders of magnitude, from tens of seconds to tens of milliseconds, compared with conventional tracking-based super-resolution approaches1-6. This high-throughput capability supports high-fidelity imaging of tissue functions under flexible experimental settings. FLAME achieves an unprecedented 3D super-resolution frame rate of ~40 Hz, capturing transient hemodynamic responses to various physiological and pathological stimuli in different mouse organs, from the microvascular level to the whole-body scale. With open-sourced implementation, FLAME provides an accessible platform for real-time, in-depth hemodynamic monitoring across diverse biomedical applications and clinical translation.

The dense glycocalyx of cancer cells can restrict immune-cell access to surface antigens and limit CAR-T cell activity. Here, we show that mucin density and epitope position determine how glycocalyx remodeling affects CAR-T cell recognition and killing. We identify KLK5 as a human protease that cleaves tumor-associated mucins, increases access to membrane-proximal antigens, and enhances CAR-T cell function. We then engineer CAR-T cells to display or secrete KLK5, enabling remodeling of the tumor glycocalyx during antigen recognition. KLK5-engineered CAR-T cells improved tumor control across multiple xenograft models, and KLK5-secreting MUC17 CAR-T cells produced the strongest in vivo benefit, prolonging survival compared with conventional MUC17 CAR-T cells. These findings show that CAR-T cells can be engineered to breach the mucin-rich glycocalyx while preserving accessible target epitopes.

Phosphorylation is a fast, reversible layer of post-translational control that directly governs protein activity, signaling flux, and cell state. Kinases and phosphatases collectively establish these phosphorylation states, but their dysregulation drives cancer, neurodegenerative diseases, and developmental disorders. Selectively manipulating the phosphorylation of individual proteins in cells remains challenging, particularly for disordered substrates. Here, we introduce Peptide-Retargeted Phosphatases (PepRePs), a genetically encoded system that fuses a peptide binder to a serine/threonine phosphatase subunit to localize phosphatase activity to a target protein. We first establish the PepReP architecture using characterized tau-binding peptides, then apply protein language model-guided peptide design to target Rab8A, PTEN, and STAT3. Across endogenous and exogenous targets, PepRePs reduce phosphorylation, with downstream effects on tau aggregation, Rab8A-associated protein interactions, and cellular proliferation following PTEN and STAT3 targeting. Sequence-based peptide design enables targeting without structural inputs, supporting application to substrates with disordered regions that challenge conventional structure-based binder design. By coupling programmable peptide binders with modular phosphatase subunits, PepRePs enable targeted control of protein phosphorylation to probe cellular signaling and develop strategies for correcting disease-associated protein dysfunction.

Ion channels remain notoriously recalcitrant drug targets despite their extensive pathophysiological significance. Acid-sensing ion channels (ASICs) are proton-gated cation channels that cause neuronal death when excessively activated, such as during ischaemic stroke. This acidosis-driven injury pathway remains unchallenged by existing therapeutics. Here we generate two de novo designed hASIC1a-specific miniproteins by developing a platform that integrates computational miniprotein design, high-throughput automated patch clamp screening, biophysical and structural characterisation, and in vivo validation. We identify a subtype-selective hASIC1a inhibitor, Denasin1, which achieves near-complete inhibition with nanomolar potency. Denasin1 is disordered in solution but adopts the computationally predicted helix-turn-helix fold when bound to the extracellular acidic pocket of hASIC1a. Crucially, Denasin1 reduces infarct volume by > 50% in a murine model of ischaemic stroke. We thereby establish our de novo design platform as a route to develop potent and selective ion channel modulators.

Adoptive cellular therapies can be effective against solid tumors but rely on the identification and expansion of rare tumor-reactive T cells (TRTs). Here, we utilize antigen-presenting cell-mimetic scaffolds (APC-ms), a modular biomaterial platform that presents defined peptide-HLA and co-stimulatory ligands on a fluid lipid bilayer to selectively recover rare antigen-specific T cells from peripheral blood. Independent variation of peptide-HLA, co-stimulatory ligands, and cytokine support defined a formulation that enriched MART-1-specific T cells to above 2% of CD8+ T cells in 35 of 50 HLA-A*02:01+ healthy donors from low or undetectable baseline frequencies. Compared with mo-DCs, APC-ms achieved greater fold expansion of MART-1-specific T cells with increased functional avidity and preservation of an early differentiation phenotype, and generated a higher number of MART-1-specific clonotypes. In five of eight patients with treatment-naive melanoma, APC-ms could enrich MART-1-specific T cells to a median of 13.7% of CD8+ cells (range 0.30 - 57.2%) Moreover, APC-ms could expand T cells specific for an HLA-A*24:01-restricted personal neoantigen that were previously undetectable in peripheral blood of a patient with melanoma. These findings establish APC-ms as a tunable platform for the recovery of rare TRTs from peripheral blood, supporting its further development for therapeutic TCR discovery and personalized adoptive cell therapy.

Pancreatic ductal adenocarcinoma (PDAC) is characterized by a dense desmoplastic microenvironment that promotes therapeutic resistance and disease progression. Here, we investigated the impact of pulsed field ablation (PFA), a non-thermal electroporation-based therapy, on tumor and stromal remodeling in a syngeneic murine PDAC model. PFA induced rapid apoptotic and necrotic tumor cell death, transient neutrophil infiltration, and durable suppression of tumor growth. Single-cell RNA sequencing following repeated PFA treatments revealed enrichment of a classical secretory tumor cell state associated with complement and stress-response pathways and reduced epithelial-to-mesenchymal transition (EMT) programs. Although overall immune composition was largely unchanged, PFA markedly remodeled the stromal compartment, decreasing SMA-positive activation and shifting cancer-associated fibroblasts (CAFs) from tumor-promoting myofibroblastic CAFs toward Dpp4+ complement-secreting CAFs. This transition was accompanied by increased IL-1 signaling potential, reduced Tgfb1 expression, and diminished tumor cell TGF-{beta} response signatures. Serial PFA treatments further suppressed EMT-associated programs, identifying stromal reprogramming as a potential mechanism through which PFA enhances tumor control.