Nature Biomedical Engineering

Intracortical brain-computer interfaces (BCIs) can restore communication to people with vocal tract paralysis by decoding cortical activity during attempted speech into text. State-of-the-art systems pairing neural-to-phoneme decoders with phoneme-to-word language models have achieved word error rates (WERs) as low as 1%, but only after collecting thousands of sentences of training data. Shortening the data collection process would facilitate scaling this new technology by reducing the time from device implant to high-accuracy communication. Here we introduce a transformer-based decoder model trained jointly across six intracortical speech BCI participants. For every participant -- regardless of sex, disease etiology, or attempted speaking strategy -- a multi-user model decoded speech more accurately (over 50% lower relative WER on average) than models trained on individual users data. Notably, the multi-user model could be finetuned on fewer than 200 sentences from a held-out user to achieve a WER below 7%. These results reveal how to pool intracortical data across people to yield more accurate, generalizable, and rapidly-deployable decoding models.

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.

Gastrointestinal anastomotic leakage (AL) remains a life-threatening complication following gastrointestinal surgery, where outcomes critically depend on timely diagnosis. Current diagnostic strategies rely on periodic sampling and resource-intensive analysis in centralized laboratories. Here, we present a sterilizable, time-integrating hydrogel sensor platform for continuous, infrastructure-free monitoring of the patient's postoperative drain fluid. We introduce enzyme-responsive macromolecular networks for semi-quantitative bedside assessment of leak-associated digestive enzymes. The sensors retain functionality following lyophilization and ethylene oxide sterilization, enabling long-term storage and scalable deployment around the world. In a Swiss clinical cohort of 56 patients, including 19 with gastrointestinal anastomotic leakage, the sensor detected amylase-associated leaks two days (median) prior to clinical diagnosis with a sensitivity of 78% (95% CI 55-91) and a specificity of 95% (95% CI 82-99). The prospective validation in an independent cohort of 37 patients in Colombia, including seven patients with leaks, demonstrated 100% sensitivity (95% CI 64.6-100) and a 100% negative predictive value (95% CI 87.9-100.0), with sensor activation preceding standard clinical diagnosis by a median of five days. By converting episodic biochemical measurements into continuous, cumulative visual records, this infrastructure-free material platform enables close-meshed postoperative monitoring and may facilitate earlier recognition of anastomotic leakage across diverse healthcare settings.

Despite advances in immunotherapy, most solid tumors remain resistant to treatment. Immune cell engagers redirect cytotoxic lymphocytes against cancer, but limited tumor access, immunosuppressive microenvironments and systemic immune activation limit efficacy. Here we develop live immune modulating engagers (LIME), a modular platform where non-pathogenic, tumor-tropic Escherichia coli display tandem single-chain variable fragments targeting a tumor-associated antigen and an activating receptor on T or natural killer cells. LIME bridged effector and tumor cells, induced transcriptional programs of T cell activation, metabolism and proliferation, and enhanced cytotoxicity across cancer cell lines and patient-derived organoids. In mouse models, LIME safely accumulated in tumors, outperformed tarlatamab in small cell lung cancer, and induced durable immunity in lymphoma. RAS inhibition and PD-L1 blockade enhanced LIME activity in pancreatic cancer and induced humoral responses. Multi-lineage immune modulation remained tumor-confined, without organ toxicity. These findings establish LIME as a versatile living therapeutic platform for programmable, tumor-restricted immune orchestration.

Treatment with chimeric antigen receptor (CAR) T cells has emerged as a promising immune therapy for relapsed and refractory hematologic malignancies. The CAR T cells are manufactured in a series of steps that involve isolating T cells from the patients leukapheresis product, genetically modifying them to express the CAR against the target antigen, and reinfusing them into the patient. Efficient T-cell enrichment from leukapheresis products is critical to the success of these therapies. Current methods for T-cell sorting on a clinical scale involve several washing steps to remove red blood cells and platelets, followed by T-cell selection and activation. These multi-step processes result in cell loss during processing and involve several handling steps. Here, we utilize fluidically assembled micromagnetic lenses to develop a high-throughput, continuous-flow microfluidic T-cell sorter, designated as the T-Chip, for sorting magnetic bead-labeled CD3+ T cells in a single step. Our approach allows direct sorting of T cells in expansion media from leukopaks without any washing steps, effectively removing 99.999% of RBCs and platelets from the leukapheresis product. A single 1-inch x 3-inch T-Chip can process leukapheresis product at a throughput of 60 mL/hr and 2.56 {+/-} 0.12 billion cells/hr. Using this optimized workflow, we demonstrate clinical-scale enrichment of highly pure CD3+ T cells (97.7 {+/-} 1.3%) with high viability (97.0 {+/-} 1.1%) and recovery (87.3 {+/-} 14.8%) in a functionally closed manner. Downstream processing of T cells isolated using the T-Chip yielded potent anti-mesothelin CAR T cells with demonstrated anti-tumor efficacy. Overall, by exploiting precisely engineered magnetic forces and laminar flow, the microfluidic T-Chip overcomes bottlenecks caused by low throughput and enables single-step large-scale T-cell purification for the rapid development of CAR T cells.

Recent studies show that pallidal and subthalamic local field potentials (LFPs) encode locomotor state and can guide adaptive deep brain stimulation (DBS) for gait impairment in Parkinson's disease. Here, in one participant implanted with the Picostim DyNeuMo-2c, we demonstrate a simpler and more direct approach for inferring locomotor state using the device's onboard accelerometer. Triaxial acceleration was classified independently on each axis to select among preconfigured stimulation programs. Using a cranially mounted digital twin, we characterized inertial signatures across medication and activity states, developed a classifier that distinguished walking from rest while rejecting tremor, and verified the intended stimulation switches during walking. In an exploratory comparison, a gait-adaptive program improved objective gait measures relative to open-loop stimulation optimised for resting tremor. These findings provide a first-in-human demonstration of the feasibility of device-embedded inertial sensing for gait-responsive DBS. They establish a practical framework for further evaluation in larger cohorts.

Precise correction of pathogenic mutations remains challenging for therapeutic applications because genome editing within protein-coding exons can generate unintended insertion-deletion byproducts that disrupt coding integrity and protein function. To address this limitation, we developed the Spliceable Editable Microexon Element (SEME), an intron-targeting platform that enables programmable microexon incorporation through endogenous splicing. In induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs), SEME corrected aberrant splicing caused by the dilated cardiomyopathy-associated FLNC c.2003A>G variant by restoring the five nucleotides missing from exon 12, thereby recovering filamin C expression. Together, these findings establish programmable microexon incorporation as a proof-of-concept strategy for correcting diverse disease-causing transcript defects.

Scalable human models that capture interactions between distinct tissues remain limited, constraining mechanistic insight and therapeutic prediction. Here, we established a scalable, automation-compatible human neuromuscular organoid (NMO) platform that enables integrated analysis of neuronal and muscle lineages in spinal muscular atrophy (SMA). Patient-derived NMOs reproducibly self-organise into spinal cord and skeletal muscle compartments and form functional neuromuscular circuits. SMA NMOs recapitulate early disease features, including reduced survival motor neuron (SMN) protein levels and impaired neuromuscular junction (NMJ) maturation. Single-nucleus RNA sequencing identifies lineage-specific transcriptional changes across neuronal and muscle compartments preceding functional deficits. Using this platform, we compared two clinically relevant SMN2 splicing modulators and observed distinct, cell-type-dependent responses. While both compounds increased SMN levels and NMJ number, only one enhanced myofiber growth and improved contractile function. These findings highlight muscle maturation, rather than NMJ number alone, as a key determinant of functional recovery and establish NMOs as a scalable system for studying cell-type-specific therapeutic responses.

Multiple sclerosis (MS) is an autoimmune disease of the central nervous system characterized by neuroinflammation, demyelination, and neurodegeneration, associated with a complex interplay between the innate and adaptive immune systems. Currently, no cure is available for MS, and personalized disease-modifying treatment remains largely limited, partially due to the lack of preclinical human models that can faithfully recapitulate disease pathology and evaluate treatment responses at the individual-patient level. Here, we report a human myelin microphysiological system (myelin MPS) platform that recaptures disease phenotype and treatment response for testing patient treatment response. By culturing neural organoids on 3D-printed devices containing directional microfibers, followed by coculture with oligodendrocyte progenitor cells, 96 myelin MPS models can be generated within a conventional well plate. Using this myelin MPS platform, autologous T cells and monocytes from MS patients induced substantially greater demyelination than those from healthy donors, accompanied by expansion of proinflammatory T-cell subsets and increased myelin uptake by monocytes/macrophages after coculture with these healthy myelinating neural tissues. Integration of imaging and flow-cytometric features distinguished healthy-donor, untreated-MS, responder, and nonresponder profiles following treatment with prednisone, glatiramer acetate, interferon {beta}-1a, or dimethyl fumarate. Thus, the myelin MPS platform provides a scalable, human pathophysiology-relevant platform for functional phenotyping and individualized treatment-response evaluation in MS.

CAR-T-cell-associated on-target off-tumor (OTOT) toxicity represents a major safety concern, as recognition of target antigens on healthy tissues can trigger severe and potentially life-threatening complications. Predicting OTOT toxicity remains a challenge because current preclinical models fail to capture the complexity of native human tissues. Here, we developed a human organotypic tissue platform that enables functional assessment of CAR-T-cell activity in intact human tissues across organ-specific and inflammatory contexts. Using a panel of clinically relevant CAR-T-cell products with known OTOT toxicities, we demonstrate that the platform faithfully recapitulates clinically observed tissue-specific toxicity profiles. CAR-T cells targeting EGFR, HER2, and mesothelin induced inflammatory and cytotoxic responses in healthy human lung tissue, whereas CD19 CAR-T cells remained inactive. We further show that OTOT toxicity cannot be reliably predicted from antigen abundance alone but instead results from the integration of multiple target-dependent determinants, including CAR affinity, inflammatory context, antigen accessibility, and effector-cell dose. The platform also enables quantitative assessment of inflammatory and cytotoxic responses and supports evaluation of pharmacological and CAR design-based strategies to mitigate toxicity. Together, this work establishes the first human organotypic platform for functional modeling of CAR-T-cell-associated OTOT toxicity, providing a clinically relevant framework for preclinical safety evaluation and the rational development of safer engineered cell therapies. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=167 SRC="FIGDIR/small/740527v1_ufig1.gif" ALT="Figure 1"> View larger version (38K): org.highwire.dtl.DTLVardef@2b47e2org.highwire.dtl.DTLVardef@17c7ce2org.highwire.dtl.DTLVardef@1eaf9feorg.highwire.dtl.DTLVardef@1c295bb_HPS_FORMAT_FIGEXP M_FIG C_FIG

Continuous neural decoding is fragile under nonstationary neural recordings because unconstrained sequence regressors can turn small mapping errors into temporally inconsistent and physically implausible motion. We propose Neural State-Space Dynamic Movement Primitives (Neural SS-DMP), which shifts the inductive bias from the neural encoder to the decoded output space: instead of directly predicting kinematics, the model infers low-dimensional movement-primitive controls and realizes them through a differentiable second-order dynamical generator. This reframes decoding as structured control inference, shrinking the set of admissible trajectories while retaining expressivity through learned forcing inputs. Because a universal motor prior cannot capture subject-specific movement dynamics, we form a personalized generator by blending the base DMP dynamics with behavior-derived subject dynamics estimated solely from training kinematics. Across ECoG and multi-session spiking benchmarks, Neural SS-DMP improves strong offline baselines in accuracy, consistently improves trajectory smoothness, and shows slower degradation on chronologically held-out sessions under an offline window-causal protocol.

Delivery remains the main obstacle to the development of in vivo genome editing therapies. CRISPR ribonucleoproteins confer high editing activity with transient exposure but lack intrinsic cell entry and targeting. Here we introduce PERCEPT, a delivery platform featuring reversible, covalent modification of CRISPR enzymes. PERCEPT enables modular installation of shielding polymers, amphiphilic delivery peptides and targeting ligands, allowing traceless cytosolic release of the native editor. A formulation incorporating an amphiphilic delivery peptide and the neuron-targeting ligand TET1 edited approximately 56% of striatal volume and 78% of neurons within edited regions after local striatal administration. In the R6/2 Huntington's disease model, PERCEPT mediated targeted editing of the mutant human HTT transgene, reducing mutant huntingtin aggregate burden and shifting local transcriptional programs away from inflammatory and injury-associated states. Tissue-adapted formulations edited 53% of Muller glia following intravitreal delivery, increased skeletal muscle reporter fluorescence tenfold versus unconjugated control following intramuscular injection, and enabled lung airway epithelial cell that persisted for three months following intranasal delivery. Rapid screening revealed that local anatomical and cellular barriers require distinct surface display for optimal editing. These results establish reversible chemical modification as a general strategy for adapting CRISPR enzyme delivery across multiple target tissues in vivo.

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.

Multibodies, or "two-in-one" Immunoglobulin G (IgG) antibodies, are standard symmetrical IgG molecules engineered to competitively bind more than one antigen within a single variable fragment (Fv) binding surface. This format merges the functional advantages of bispecifics, such as multi-target binding and dynamic adaptation to target concentrations, with the superior manufacturing, developability, pharmacokinetics, and avidity of monospecific IgGs. Moreover, the co-accommodation of multiple paratopes on a single set of 6 CDRs introduces new functional possibilities that can improve efficacy and safety. Multibodies can, therefore, be thought of as force multipliers: for any format of antibodies, or fragments thereof, multibodies can bind double the number of epitopes compared to standard antibodies. While these advantages were recognized more than 15 years ago, the systematic design of multibodies has been intractable due to the challenge of optimizing two binding specificities into one Fv region, without having one of them compromising the other and without inducing poly-reactivity. To overcome this engineering barrier, we have developed an artificial intelligence (AI)-assisted computational platform that enables the design of functional multibodies against virtually any pair of targets. We applied the platform to design nine multibodies combining 15 different unrelated targets. We obtained therapeutic-grade multibodies that bind each desired pair of targets. We demonstrate that the generated multibodies possess excellent developability, high affinity, and stringent specificity, comparing favorably to clinical monospecific benchmarks. Critically, we show that these multibodies exhibit superior functional activity across a diverse range of mechanisms of action (MOAs), including internalization, T-cell engagement, and immune system modulation. This capability to reliably engineer versatile multibodies opens a new domain in antibody therapeutics, enabling complex multipharmacology and novel functions within a natural, cost-effective, and highly developable format. Two of these multibodies are currently in IND enabling studies, with first in human studies expected in 2026. The timeline from idea to a fully optimized, developable, lead candidate, ready for IND enabling studies, is 9 months.

The most accurate neural decoder on held-out trials is not necessarily the most useful for brain-computer interfaces or neural population analysis. In practical use, neural decoders may also need to remain robust to noisy neural inputs, satisfy calibration or deployment constraints, and produce comparable representations across recordings. We introduce BEND-BCI, an open-source benchmark of 23 neural decoding methods on motor, visual, speech and spatial decoding tasks across 16 real or synthetic neural recordings. BEND-BCI compares decoders across held-out prediction, robustness to input perturbation, computational cost and cross-recording latent consistency. These additional axes frequently changed decoder rankings: held-out accuracy did not reliably identify the most robust, efficient or cross-recording-consistent models. Simpler baselines were also competitive with, and in some cases outperformed, more heavily parameterized deep neural networks. Diagnostic analyses based on explainable machine learning further showed that decoder performance was associated with the use of expected neural features and could be improved by selecting high-quality training trials. BEND-BCI reframes neural-decoder selection from an accuracy leaderboard into a constrained decision over task, resource, representation and diagnostic goals.

Deep brain stimulation (DBS) is transforming from a static therapy toward adaptive systems that adjust stimulation based on neural biomarkers. However, the detection of reliable biomarkers that capture the multi-dimensional nature of complex symptoms is often challenging. Here we demonstrate volitional DBS (vDBS)--a paradigm in which patients use brain-computer interface (BCI) training to learn self-regulation of a neural signal that then controls closed-loop DBS. Two patients with Parkinson's disease implanted with sensing-enabled neurostimulators completed chronic, at-home BCI training by playing an airplane simulation game. Through training, they were able to effectively down-regulate their cortical beta signal (p's < 1e-10), represented as the real-time position of a plane in the BCI game. Following training, this cortical beta signal served as the input to a closed-loop DBS algorithm. By modulating their beta signal to cross personalized thresholds, patients voluntarily increased or decreased neurostimulation amplitude at will, in the absence of physical movement (p's < 1e-10). This proof-of-principle demonstration establishes that volitional control of intracranial neurostimulation is achievable without the need of an externalized manual controller. BCI-vDBS could potentially be used for a range of neuropsychiatric conditions and brain rehabilitation to support personalized control of neurostimulation.

Fibrotic skin diseases are characterized by persistent fibroblast activation and extracellular matrix remodeling, yet the mechanisms governing fibroblast state transitions remain incompletely understood. Here, we established a human iPSC-derived skin organoid model of fibrosis through chronic TGF-{beta} stimulation. Single-cell RNA sequencing combined with immunofluorescence-based spatial analysis revealed dynamic fibroblast state transitions, spatial reorganization, and expansion of activated fibroblast populations during fibrotic remodeling. Integration with human scleroderma single-cell datasets demonstrated conserved fibroblast states and transcriptional programs between organoids and patient tissues. We further identified broad induction of RUNX2 in the dermal compartment during fibrosis, and RUNX2 depletion attenuated fibrotic marker expression. CUT&RUN profiling revealed RUNX2 occupancy at fibrosis-associated loci, including RUNX1 and LOXL2. Using a machine learning-guided screening approach, we identified F0565-0303, a small molecule that suppressed RUNX2-dependent fibrotic programs in vitro and reduced fibrosis in a bleomycin-induced mouse model. Together, these findings establish human skin organoids as a platform for modeling fibrosis and nominate RUNX2 as a potential therapeutic target.

Inflammatory bowel disease (IBD) is a heterogeneous chronic disease driven by dysregulated mucosal immunity and impaired epithelial barrier function. Although biologics have improved disease management, they are frequently associated with systemic immunosuppression and adverse effects, highlighting the need for localized therapeutic strategies that both control inflammation and promote tissue repair. Here, we developed a protein bispecific termed 7A2-IgG4-IL22, composed of a human IgG4-Fc domain displaying an antagonistic anti-human MAdCAM-1 single chain (sc)-Fv and a human interleukin (IL-)22. The anti-MAdCAM-1 scFv retained the functional activity of the parental monoclonal antibody, inhibiting T cell activation, expansion and differentiation from naive precursors. Blockade of the MAdCAM-1 signaling axis also reduced production of pro-inflammatory cytokines relevant to IBD pathogenesis, including IFN{gamma} and TNF. On the epithelial side, the IL-22 cargo induces robust signaling in epithelial cells, promoting the expression of IL-22 response genes associated with antimicrobial defense, mucosal homeostasis, as well as IL-10 and CXCL1 expression. This effect contributes to immune cell trafficking to the intestinal mucosa. Together, this bispecific provides a localized dual-mechanism strategy for restoring intestinal immune homeostasis.

How ecological architectures within the gut microbiome convert complex inputs into specific host physiological outcomes remains poorly understood. We used CDD-2101, a multi-component botanical drug operating under an FDA (U.S. Food and Drug Administration) Investigational New Drug program, as a defined ecological perturbation in functional constipation (FC). Integrating a randomized, double-blind, placebo-controlled clinical trial with genome-resolved metagenomics, targeted metabolomics, staged prediction modeling, and receptor-level validation, we show that clinical efficacy of CDD-2101 depends on remodeling a function-specific substructure of the stable Two Competing Guilds (TCG) architecture. We term this substructure the FC-TCG, demonstrate its role along the gut-motility axis, and confirm its effect in three independent gut hypomotility cohorts. The two guilds responded asymmetrically: the intervention selectively suppressed the C1B guild (the pathobiont guild) while largely sparing the C1A guild, the foundation guild that anchors the core gut community, restoring its ecological dominance, producing a coordinated metabolic shift that elevates lithocholic acid and propionic acid. Through gnotobiotic transplantation and receptor antagonism, we demonstrate that lithocholic acid and propionic acid restore gut motility via concurrent engagement of Takeda G protein-coupled receptor 5 (TGR5) and G-protein coupled receptor 43 (GPR43). These findings identify microbial guild architecture as a function-resolved signal-transducing layer that converts multi-component botanical intervention into multi-receptor-mediated gut motility restoration, reframing the gut microbiome from a compositional system into a structural transducer between complex environmental inputs and host physiology.