Nuclear-Import Receptors Reverse Aberrant Phase Transitions of RNA-Binding Proteins with Prion-like Domains.

RNA-binding proteins (RBPs) with prion-like domains (PrLDs) phase transition to functional liquids, which can mature into aberrant hydrogels composed of pathological fibrils that underpin fatal neurodegenerative disorders. Several nuclear RBPs with PrLDs, including TDP-43, FUS, hnRNPA1, and hnRNPA2, mislocalize to cytoplasmic inclusions in neurodegenerative disorders, and mutations in their PrLDs can accelerate fibrillization and cause disease. Here, we establish that nuclear-import receptors (NIRs) specifically chaperone and potently disaggregate wild-type and disease-linked RBPs bearing a NLS. Karyopherin-ß2 (also called Transportin-1) engages PY-NLSs to inhibit and reverse FUS, TAF15, EWSR1, hnRNPA1, and hnRNPA2 fibrillization, whereas Importin-α plus Karyopherin-ß1 prevent and reverse TDP-43 fibrillization. Remarkably, Karyopherin-ß2 dissolves phase-separated liquids and aberrant fibrillar hydrogels formed by FUS and hnRNPA1. In vivo, Karyopherin-ß2 prevents RBPs with PY-NLSs accumulating in stress granules, restores nuclear RBP localization and function, and rescues degeneration caused by disease-linked FUS and hnRNPA2. Thus, NIRs therapeutically restore RBP homeostasis and mitigate neurodegeneration.

MicroRNA directly enhances mitochondrial translation during muscle differentiation.

MicroRNAs are well known to mediate translational repression and mRNA degradation in the cytoplasm. Various microRNAs have also been detected in membrane-compartmentalized organelles, but the functional significance has remained elusive. Here, we report that miR-1, a microRNA specifically induced during myogenesis, efficiently enters the mitochondria where it unexpectedly stimulates, rather than represses, the translation of specific mitochondrial genome-encoded transcripts. We show that this positive effect requires specific miR:mRNA base-pairing and Ago2, but not its functional partner GW182, which is excluded from the mitochondria. We provide evidence for the direct action of Ago2 in mitochondrial translation by crosslinking immunoprecipitation coupled with deep sequencing (CLIP-seq), functional rescue with mitochondria-targeted Ago2, and selective inhibition of the microRNA machinery in the cytoplasm. These findings unveil a positive function of microRNA in mitochondrial translation and suggest a highly coordinated myogenic program via miR-1-mediated translational stimulation in the mitochondria and repression in the cytoplasm.

The tumor suppressor PTEN has a critical role in antiviral innate immunity.

The gene encoding PTEN is one of the most frequently mutated tumor suppressor-encoding genes in human cancer. While PTEN's function in tumor suppression is well established, its relationship to anti-microbial immunity remains unknown. Here we found a pivotal role for PTEN in the induction of type I interferon, the hallmark of antiviral innate immunity, that was independent of the pathway of the kinases PI(3)K and Akt. PTEN controlled the import of IRF3, a master transcription factor responsible for IFN-ß production, into the nucleus. We further identified a PTEN-controlled negative phosphorylation site at Ser97 of IRF3 and found that release from this negative regulation via the phosphatase activity of PTEN was essential for the activation of IRF3 and its import into the nucleus. Our study identifies crosstalk between PTEN and IRF3 in tumor suppression and innate immunity.

Stress Induces Dynamic, Cytotoxicity-Antagonizing TDP-43 Nuclear Bodies via Paraspeckle LncRNA NEAT1-Mediated Liquid-Liquid Phase Separation.

Despite the prominent role of TDP-43 in neurodegeneration, its physiological and pathological functions are not fully understood. Here, we report an unexpected role of TDP-43 in the formation of dynamic, reversible, liquid droplet-like nuclear bodies (NBs) in response to stress. Formation of NBs alleviates TDP-43-mediated cytotoxicity in mammalian cells and fly neurons. Super-resolution microscopy reveals distinct functions of the two RRMs in TDP-43 NB formation. TDP-43 NBs are partially colocalized with nuclear paraspeckles, whose scaffolding lncRNA NEAT1 is dramatically upregulated in stressed neurons. Moreover, increase of NEAT1 promotes TDP-43 liquid-liquid phase separation (LLPS) in vitro. Finally, we discover that the ALS-associated mutation D169G impairs the NEAT1-mediated TDP-43 LLPS and NB assembly, causing excessive cytoplasmic translocation of TDP-43 to form stress granules, which become phosphorylated TDP-43 cytoplasmic foci upon prolonged stress. Together, our findings suggest a stress-mitigating role and mechanism of TDP-43 NBs, whose dysfunction may be involved in ALS pathogenesis.

Loss of Dynamic RNA Interaction and Aberrant Phase Separation Induced by Two Distinct Types of ALS/FTD-Linked FUS Mutations.

FUS is a nuclear RNA-binding protein, and its cytoplasmic aggregation is a pathogenic signature of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). It remains unknown how the FUS-RNA interactions contribute to phase separation and whether its phase behavior is affected by ALS-linked mutations. Here we demonstrate that wild-type FUS binds single-stranded RNA stoichiometrically in a length-dependent manner and that multimers induce highly dynamic interactions with RNA, giving rise to small and fluid condensates. In contrast, mutations in arginine display a severely altered conformation, static binding to RNA, and formation of large condensates, signifying the role of arginine in driving proper RNA interaction. Glycine mutations undergo rapid loss of fluidity, emphasizing the role of glycine in promoting fluidity. Strikingly, the nuclear import receptor Karyopherin-ß2 reverses the mutant defects and recovers the wild-type FUS behavior. We reveal two distinct mechanisms underpinning potentially disparate pathogenic pathways of ALS-linked FUS mutants.

Resolving Cell Fate Decisions during Somatic Cell Reprogramming by Single-Cell RNA-Seq.

Somatic cells can be reprogrammed into induced pluripotent stem cells (iPSCs), which is a highly heterogeneous process. Here we report the cell fate continuum during somatic cell reprogramming at single-cell resolution. We first develop SOT to analyze cell fate continuum from Oct4/Sox2/Klf4- or OSK-mediated reprogramming and show that cells bifurcate into two categories, reprogramming potential (RP) or non-reprogramming (NR). We further show that Klf4 contributes to Cd34+/Fxyd5+/Psca+ keratinocyte-like NR fate and that IFN-γ impedes the final transition to chimera-competent pluripotency along the RP cells. We analyze more than 150,000 single cells from both OSK and chemical reprograming and identify additional NR/RP bifurcation points. Our work reveals a generic bifurcation model for cell fate decisions during somatic cell reprogramming that may be applicable to other systems and inspire further improvements for reprogramming.

Hierarchically structured diamond composite with exceptional toughness.

The well known trade-off between hardness and toughness (resistance to fracture) makes simultaneous improvement of both properties challenging, especially in diamond. The hardness of diamond can be increased through nanostructuring strategies1,2, among which the formation of high-density nanoscale twins - crystalline regions related by symmetry - also toughens diamond2. In materials other than diamond, there are several other promising approaches to enhancing toughness in addition to nanotwinning3, such as bio-inspired laminated composite toughening4-7, transformation toughening8 and dual-phase toughening9, but there has been little research into such approaches in diamond. Here we report the structural characterization of a diamond composite hierarchically assembled with coherently interfaced diamond polytypes (different stacking sequences), interwoven nanotwins and interlocked nanograins. The architecture of the composite enhances toughness more than nanotwinning alone, without sacrificing hardness. Single-edge notched beam tests yield a toughness up to five times that of synthetic diamond10, even greater than that of magnesium alloys. When fracture occurs, a crack propagates through diamond nanotwins of the 3C (cubic) polytype along {111} planes, via a zigzag path. As the crack encounters regions of non-3C polytypes, its propagation is diffused into sinuous fractures, with local transformation into 3C diamond near the fracture surfaces. Both processes dissipate strain energy, thereby enhancing toughness. This work could prove useful in making superhard materials and engineering ceramics. By using structural architecture with synergetic effects of hardening and toughening, the trade-off between hardness and toughness may eventually be surmounted.

Poly(ADP-Ribose) Prevents Pathological Phase Separation of TDP-43 by Promoting Liquid Demixing and Stress Granule Localization.

In amyotrophic lateral sclerosis (ALS) and frontotemporal degeneration (FTD), cytoplasmic aggregates of hyperphosphorylated TDP-43 accumulate and colocalize with some stress granule components, but how pathological TDP-43 aggregation is nucleated remains unknown. In Drosophila, we establish that downregulation of tankyrase, a poly(ADP-ribose) (PAR) polymerase, reduces TDP-43 accumulation in the cytoplasm and potently mitigates neurodegeneration. We establish that TDP-43 non-covalently binds to PAR via PAR-binding motifs embedded within its nuclear localization sequence. PAR binding promotes liquid-liquid phase separation of TDP-43 in vitro and is required for TDP-43 accumulation in stress granules in mammalian cells and neurons. Stress granule localization initially protects TDP-43 from disease-associated phosphorylation, but upon long-term stress, stress granules resolve, leaving behind aggregates of phosphorylated TDP-43. Finally, small-molecule inhibition of Tankyrase-1/2 in mammalian cells inhibits formation of cytoplasmic TDP-43 foci without affecting stress granule assembly. Thus, Tankyrase inhibition antagonizes TDP-43-associated pathology and neurodegeneration and could have therapeutic utility for ALS and FTD.

FUS Regulates Activity of MicroRNA-Mediated Gene Silencing.

MicroRNA-mediated gene silencing is a fundamental mechanism in the regulation of gene expression. It remains unclear how the efficiency of RNA silencing could be influenced by RNA-binding proteins associated with the microRNA-induced silencing complex (miRISC). Here we report that fused in sarcoma (FUS), an RNA-binding protein linked to neurodegenerative diseases including amyotrophic lateral sclerosis (ALS), interacts with the core miRISC component AGO2 and is required for optimal microRNA-mediated gene silencing. FUS promotes gene silencing by binding to microRNA and mRNA targets, as illustrated by its action on miR-200c and its target ZEB1. A truncated mutant form of FUS that leads its carriers to an aggressive form of ALS, R495X, impairs microRNA-mediated gene silencing. The C. elegans homolog fust-1 also shares a conserved role in regulating the microRNA pathway. Collectively, our results suggest a role for FUS in regulating the activity of microRNA-mediated silencing.

LncRNA CamK-A Regulates Ca2+-Signaling-Mediated Tumor Microenvironment Remodeling.

Cancer cells entail metabolic adaptation and microenvironmental remodeling to survive and progress. Both calcium (Ca2+) flux and Ca2+-dependent signaling play a crucial role in this process, although the underlying mechanism has yet to be elucidated. Through RNA screening, we identified one long noncoding RNA (lncRNA) named CamK-A (lncRNA for calcium-dependent kinase activation) in tumorigenesis. CamK-A is highly expressed in multiple human cancers and involved in cancer microenvironment remodeling via activation of Ca2+-triggered signaling. Mechanistically, CamK-A activates Ca2+/calmodulin-dependent kinase PNCK, which in turn phosphorylates IκBα and triggers calcium-dependent nuclear factor κB (NF-κB) activation. This regulation results in the tumor microenvironment remodeling, including macrophage recruitment, angiogenesis, and tumor progression. Notably, our human-patient-derived xenograft (PDX) model studies demonstrate that targeting CamK-A robustly impaired cancer development. Clinically, CamK-A expression coordinates with the activation of CaMK-NF-κB axis, and its high expression indicates poor patient survival rate, suggesting its role as a potential biomarker and therapeutic target.

Efficient C-N coupling in the direct synthesis of urea from CO2 and N2 by amorphous SbxBi1-xOy clusters.

Although direct generation of high-value complex molecules and feedstock by coupling of ubiquitous small molecules such as CO2 and N2 holds great appeal as a potential alternative to current fossil-fuel technologies, suitable scalable and efficient catalysts to this end are not currently available as yet to be designed and developed. To this end, here we prepare and characterize SbxBi1-xOy clusters for direct urea synthesis from CO2 and N2 via C-N coupling. The introduction of Sb in the amorphous BiOx clusters changes the adsorption geometry of CO2 on the catalyst from O-connected to C-connected, creating the possibility for the formation of complex products such as urea. The modulated Bi(II) sites can effectively inject electrons into N2, promoting C-N coupling by advantageous modification of the symmetry for the frontier orbitals of CO2 and N2 involved in the rate-determining catalytic step. Compared with BiOx, SbxBi1-xOy clusters result in a lower reaction potential of only -0.3 V vs. RHE, an increased production yield of 307.97 µg h-1 mg-1cat, and a higher Faraday efficiency (10.9%), pointing to the present system as one of the best catalysts for urea synthesis in aqueous systems among those reported so far. Beyond the urea synthesis, the present results introduce and demonstrate unique strategies to modulate the electronic states of main group p-metals toward their use as effective catalysts for multistep electroreduction reactions requiring C-N coupling.

Phase-dependent visual and sensorimotor integration of features for grasp computations before and after effector specification.

The simple act of viewing and grasping an object involves complex sensorimotor control mechanisms that have been shown to vary as a function of multiple object and other task features such as object size, shape, weight, and wrist orientation. However, these features have been mostly studied in isolation. In contrast, given the nonlinearity of motor control its computations require multiple features to be incorporated concurrently. Therefore, the present study tested the hypothesis that grasp computations integrate multiple task features superadditively in particular when these features are relevant for the same action phase. We asked male and female human participants to reach-to-grasp objects of different shapes and sizes with different wrist orientations. Also, we delayed movement onset using auditory signals to specify which effector to use. Using electroencephalography (EEG) and representative dissimilarity analysis to map the time course of cortical activity we found that grasp computations formed superadditive integrated representations of grasp features during different planning phases of grasping. Shape-by-size representations and size-by-orientation representations occurred before and after effector specification, respectively, and could not be explained by single-feature models. These observations are consistent with the brain performing different preparatory, phase-specific computations; visual object analysis to identify grasp points at abstract visual levels and downstream sensorimotor preparatory computations for reach-to-grasp trajectories. Our results suggest the brain adheres to the needs of nonlinear motor control for integration. Furthermore, they show that examining the superadditive influence of integrated representations can serve as a novel lens to map the computations underlying sensorimotor control.Significance Statement The nonlinearity of the sensorimotor control of grasping should require computations to incorporate multiple task features such as object shape, size, and orientation concurrently. However, grasp research so far has primarily investigated the influences of task features in isolation. In contrast, integrated representations of task features have been studied in cognitive paradigms showing that multiple visual and action features are joined together in abstract representations based on working memory or priming effects called events files. Using multivariate analysis of EEG, here we observe a new form of integrated representations of task features for grasping that cannot be explained by single-feature models or event files. Our approach offers novel insights into the preparatory processes of sensorimotor grasp control.

Computation on Demand: Action-Specific Representations of Visual Task Features Arise during Distinct Movement Phases.

The intention to act influences the computations of various task-relevant features. However, little is known about the time course of these computations. Furthermore, it is commonly held that these computations are governed by conjunctive neural representations of the features. But, support for this view comes from paradigms arbitrarily combining task features and affordances, thus requiring representations in working memory. Therefore, the present study used electroencephalography and a well-rehearsed task with features that afford minimal working memory representations to investigate the temporal evolution of feature representations and their potential integration in the brain. Female and male human participants grasped objects or touched them with a knuckle. Objects had different shapes and were made of heavy or light materials with shape and weight being relevant for grasping, not for "knuckling." Using multivariate analysis showed that representations of object shape were similar for grasping and knuckling. However, only for grasping did early shape representations reactivate at later phases of grasp planning, suggesting that sensorimotor control signals feed back to the early visual cortex. Grasp-specific representations of material/weight only arose during grasp execution after object contact during the load phase. A trend for integrated representations of shape and material also became grasp-specific but only briefly during the movement onset. These results suggest that the brain generates action-specific representations of relevant features as required for the different subcomponents of its action computations. Our results argue against the view that goal-directed actions inevitably join all features of a task into a sustained and unified neural representation.

Rapid growth of Moso bamboo (Phyllostachys edulis): Cellular roadmaps, transcriptome dynamics, and environmental factors.

Moso bamboo (Phyllostachys edulis) shows remarkably rapid growth (114.5 cm/day), but the underlying biological mechanisms remain unclear. After examining more than 12,750 internodes from more than 510 culms from 17 Moso populations, we identified internode 18 as a representative internode for rapid growth. This internode includes a 2-cm cell division zone (DZ), a cell elongation zone up to 12 cm, and a secondary cell wall (SCW) thickening zone. These zones elongated 11.8 cm, produced approximately 570,000,000 cells, and deposited ∼28 mg g-1 dry weight (DW) lignin and ∼44 mg g-1 DW cellulose daily, far exceeding vegetative growth observed in other plants. We used anatomical, mathematical, physiological, and genomic data to characterize development and transcriptional networks during rapid growth in internode 18. Our results suggest that (1) gibberellin may directly trigger the rapid growth of Moso shoots, (2) decreased cytokinin and increased auxin accumulation may trigger cell DZ elongation, and (3) abscisic acid and mechanical pressure may stimulate rapid SCW thickening via MYB83L. We conclude that internode length involves a possible tradeoff mediated by mechanical pressure caused by rapid growth, possibly influenced by environmental temperature and regulated by genes related to cell division and elongation. Our results provide insight into the rapid growth of Moso bamboo.

Recent Progress of Amorphous Nanomaterials.

Amorphous materials are metastable solids with only short-range order at the atomic scale, which results from local intermolecular chemical bonding. The lack of long-range order typical of crystals endows amorphous nanomaterials with unconventional and intriguing structural features, such as isotropic atomic environments, abundant surface dangling bonds, highly unsaturated coordination, etc. Because of these features and the ensuing modulation in electronic properties, amorphous nanomaterials display potential for practical applications in different areas. Motivated by these elements, here we provide an overview of the unique structural features, the general synthetic methods, and the potential for applications covered by contemporary research in amorphous nanomaterials. Furthermore, we discussed the possible theoretical mechanism for amorphous nanomaterials, examining how the unique structural properties and electronic configurations contribute to their exceptional performance. In particular, the structural benefits of amorphous nanomaterials as well as their enhanced electrocatalytic, optical, and mechanical properties, thereby clarifying the structure-function relationships, are highlighted. Finally, a perspective on the preparation and utilization of amorphous nanomaterials to establish mature systems with a superior hierarchy for various applications is introduced, and an outlook for future challenges and opportunities at the frontiers of this rapidly advancing field is proposed.

HISTONE DEACETYLASE 9 transduces heat signal in plant cells.

Heat stress limits plant growth, development, and crop yield, but how plant cells precisely sense and transduce heat stress signals remains elusive. Here, we identified a conserved heat stress response mechanism to elucidate how heat stress signal is transmitted from the cytoplasm into the nucleus for epigenetic modifiers. We demonstrate that HISTONE DEACETYLASE 9 (HDA9) transduces heat signals from the cytoplasm to the nucleus to play a positive regulatory role in heat responses in Arabidopsis. Heat specifically induces HDA9 accumulation in the nucleus. Under heat stress, the phosphatase PP2AB'ß directly interacts with and dephosphorylates HDA9 to protect HDA9 from 26S proteasome-mediated degradation, leading to the translocation of nonphosphorylated HDA9 to the nucleus. This heat-induced enrichment of HDA9 in the nucleus depends on the nucleoporin HOS1. In the nucleus, HDA9 binds and deacetylates the target genes related to signaling transduction and plant development to repress gene expression in a transcription factor YIN YANG 1-dependent and -independent manner, resulting in rebalance of plant development and heat response. Therefore, we uncover an HDA9-mediated positive regulatory module in the heat shock signal transduction pathway. More important, this cytoplasm-to-nucleus translocation of HDA9 in response to heat stress is conserved in wheat and rice, which confers the mechanism significant implication potential for crop breeding to cope with global climate warming.

The synthesis and application of crystalline-amorphous hybrid materials.

Crystalline-amorphous hybrid materials (CA-HMs) possess the merits of both pure crystalline and amorphous phases. Abundant dangling bonds, unsaturated coordination atoms, and isotropic structural features in the amorphous phase, as well as relatively high electronic conductivity and thermodynamic structural stability of the crystalline phase simultaneously take effect in CA-HMs. Furthermore, the atomic and bandgap mismatch at the CA-HM interface can introduce more defects as extra active sites, reservoirs for promoted catalytic and electrochemical performance, and induce built-in electric field for facile charge carrier transport. Motivated by these intriguing features, herein, we provide a comprehensive overview of CA-HMs on various aspects-from synthetic methods to multiple applications. Typical characteristics of CA-HMs are discussed at the beginning, followed by representative synthetic strategies of CA-HMs, including hydrothermal/solvothermal methods, deposition techniques, thermal adjustment, and templating methods. Diverse applications of CA-HMs, such as electrocatalysis, batteries, supercapacitors, mechanics, optoelectronics, and thermoelectrics along with underlying structure-property mechanisms are carefully elucidated. Finally, challenges and perspectives of CA-HMs are proposed with an aim to provide insights into the future development of CA-HMs.

Amorphous Bismuth-Tin Oxide Nanosheets with Optimized C-N Coupling for Efficient Urea Synthesis.

Closing the carbon and nitrogen cycles by electrochemical methods using renewable energy to convert abundant or harmful feedstocks into high-value C- or N-containing chemicals has the potential to transform the global energy landscape. However, efficient conversion avenues have to date been mostly realized for the independent reduction of CO2 or NO3-. The synthesis of more complex C-N compounds still suffers from low conversion efficiency due to the inability to find effective catalysts. To this end, here we present amorphous bismuth-tin oxide nanosheets, which effectively reduce the energy barrier of the catalytic reaction, facilitating efficient and highly selective urea production. With enhanced CO2 adsorption and activation on the catalyst, a C-N coupling pathway based on *CO2 rather than traditional *CO is realized. The optimized orbital symmetry of the C- (*CO2) and N-containing (*NO2) intermediates promotes a significant increase in the Faraday efficiency of urea production to an outstanding value of 78.36% at -0.4 V vs RHE. In parallel, the nitrogen and carbon selectivity for urea formation is also enhanced to 90.41% and 95.39%, respectively. The present results and insights provide a valuable reference for the further development of new catalysts for efficient synthesis of high-value C-N compounds from CO2.

High-Valence Cu Induced by Photoelectric Reconstruction for Dynamically Stable Oxygen Evolution Sites.

Oxygen vacancies are generally considered to play a crucial role in the oxygen evolution reaction (OER). However, the generation of active sites created by oxygen vacancies is inevitably restricted by their condensation and elimination reactions. To overcome this limitation, here, we demonstrate a novel photoelectric reconstruction strategy to incorporate atomically dispersed Cu into ultrathin (about 2-3 molecular) amorphous oxyhydroxide (a-CuM, M = Co, Ni, Fe, or Zn), facilitating deprotonation of the reconstructed oxyhydroxide to generate high-valence Cu. The in situ XAFS results and first-principles calculations reveal that Cu atoms are stabilized at high valence during the OER process due to Jahn-Teller distortion, resulting in para-type double oxygen vacancies as dynamically stable catalytic sites. The optimal a-CuCo catalyst exhibits a record-high mass activity of 3404.7 A g-1 at an overpotential of 300 mV, superior to the benchmarking hydroxide and oxide catalysts. The developed photoelectric reconstruction strategy opens up a new pathway to construct in situ stable oxygen vacancies by high-valence Cu single sites, which extends the design rules for creating dynamically stable active sites.

A non-invasive artificial intelligence model for identifying axillary pathological complete response to neoadjuvant chemotherapy in breast cancer: a secondary analysis to multicenter clinical trial.

BACKGROUND: This study aims to develop a stacking model for accurately predicting axillary lymph node (ALN) response to neoadjuvant chemotherapy (NAC) using longitudinal MRI in breast cancer. METHODS: We included patients with node-positive breast cancer who received NAC following surgery from January 2012 to June 2022. We collected MRIs before and after NAC, and extracted radiomics features from the tumour, peritumour, and ALN regions. The Mann-Whitney U test, least absolute shrinkage and selection operator, and Boruta algorithm were used to select features. We utilised machine learning techniques to develop three single-modality models and a stacking model for predicting ALN response to NAC. RESULTS: This study consisted of a training cohort (n = 277), three external validation cohorts (n = 313, 164, and 318), and a prospective cohort (n = 81). Among the 1153 patients, 60.62% achieved ypN0. The stacking model achieved excellent AUCs of 0.926, 0.874, and 0.862 in the training, external validation, and prospective cohort, respectively. It also showed lower false-negative rates (FNRs) compared to radiologists, with rates of 14.40%, 20.85%, and 18.18% (radiologists: 40.80%, 50.49%, and 63.64%) in three cohorts. Additionally, there was a significant difference in disease-free survival between high-risk and low-risk groups (p < 0.05). CONCLUSIONS: The stacking model can accurately predict ALN status after NAC in breast cancer, showing a lower false-negative rate than radiologists. TRIAL REGISTRATION NUMBER: The clinical trial numbers were NCT03154749 and NCT04858529.