Hypertensive individuals are concurrently at risk of heavy metal exposure, particularly lead (Pb). However, it remains unclear whether individuals with hypertension are more sensitive to the neurotoxicity induced by Pb exposure. Here, we established a hypertension model of mice with Pb exposure for 2, 4, 8, 12, and 24 weeks to detect the temporal and spatial changes in microglial polarization. Pb-exposed and hypertensive mice exhibited a notably higher proportion of M1 as early as 4 and 8 weeks, but a lower proportion of M2 as early as 4 and 12 weeks. Notably, hypertension exacerbated Pb-induced microglial polarization, which reached a plateau at 12 weeks. Elevated inducible nitric oxide synthase and reduced arginase-1 protein expressions were observed in the prefrontal cortex, hippocampus, and hypothalamus of the Pb, hypertension (HTN), and Pb+HTN groups, with the most pronounced alterations occurring in the prefrontal cortex. Furthermore, treatment with minocycline (microglial polarization inhibitor) markedly alleviated the depression-like behavior of mice in the Pb+HTN group. Bioinformatics analysis revealed that the differentially expressed genes of microglia following Pb and angiotensin II combined exposure were mainly enriched in the cyclic GMP-AMP synthase/stimulator of interferon genes (cGAS/STING) pathway. In addition, the protein expression of the cGAS/STING pathway was significantly upregulated both in vitro and in vivo, especially in the Pb+HTN group. Inhibition of cGAS significantly alleviated the increase in M1-type microglia and the decrease in M2-type microglia in hypertensive mice with Pb exposure. Meanwhile, mitochondrial DNA (mtDNA) leakage of microglia was observed under Pb and angiotensin II exposure, and the activation of the cGAS/STING pathway by mtDNA was proven to play an important role in microglial polarization. Adenine nucleotide translocase 1 (ANT1) was identified as the pivotal protein regulating mtDNA leakage via bioinformatics analyses and was proven by in vitro and in vivo studies. Collectively, hypertensive mice are more sensitive to Pb-induced neurotoxicity, which is mediated by the ANT1-regulated mtDNA/cGAS/STING signaling pathway, which then subsequently induces microglial polarization.

As terahertz (THz) science and technology continue to advance, THz biophysics and chemistry have emerged as a pivotal interface, linking molecular-scale insights with transformative innovations in biological, physical, and chemical sciences. This review elucidates the fundamental mechanisms underlying THz–matter interactions, emphasizing the relationship between THz wave physics and molecular motions. We highlight molecular-level effects of THz radiation on water, ion channels, DNA, and proteins, and summarize its biological impacts spanning molecular, cellular, and neurological scales. These mechanistic insights underpin diverse applications, including the control of water transport and molecular dynamics, diagnostics, neurobiology, and emerging therapeutic strategies. Furthermore, we examine advanced methods for detection and characterization in THz biophysics and chemistry, including THz spectroscopy, imaging, and sensing, which bridge fundamental principles with practical implementations, positioning THz technology as a versatile platform for probing and manipulating complex biological and chemical systems. By integrating these recent advances, this review aims to catalyze further research in this rapidly evolving field.

Generative artificial intelligence (AI) is rapidly transforming medical imaging by enabling capabilities such as data synthesis, image enhancement, modality translation, and spatiotemporal modeling. This review presents a comprehensive and forward-looking synthesis of recent advances in generative modeling—including generative adversarial networks (GANs), variational autoencoders (VAEs), diffusion models, and emerging multimodal foundation architectures—and evaluates their expanding roles across the clinical imaging continuum. We systematically examine how generative AI contributes to key stages of the imaging workflow, from acquisition and reconstruction to cross-modality synthesis, diagnostic support, treatment planning, and prognosis prediction. Emphasis is placed on both retrospective and prospective clinical scenarios, where generative models help address longstanding challenges such as data scarcity, standardization, and integration across modalities. To promote rigorous benchmarking and translational readiness, we propose a 3-tiered evaluation framework encompassing pixel-level fidelity, feature-level realism, and task-level clinical relevance. We also identify critical obstacles to real-world deployment, including limited generalization under domain shift, risks of hallucinated or unreliable features, data scarcity and privacy concerns, as well as stringent regulatory and ethical constraints. Finally, we explore the convergence of generative AI with large-scale foundation models, highlighting how this synergy may enable the next generation of scalable, reliable, and clinically integrated imaging systems. By charting technical progress and translational pathways, this review aims to guide future research and foster interdisciplinary collaboration at the intersection of AI, medicine, and biomedical engineering.

The ventral tegmental area (VTA) is a crucial brain region for dopamine synthesis and reward processing, yet the molecular diversity and functional roles of nondopaminergic VTA neurons remain poorly characterized. While ARVCF (a member of ARVCF delta catenin family) has been implicated in dopamine release and reward-related behaviors (e.g., nicotine/alcohol and natural rewards), its influence on VTA neuronal subpopulations at single-cell resolution is unknown. Based on the Arvcf-knockout (Arvcf-KO) mouse model and nicotine exposure, we constructed a reward behavior-related profile at the single-cell transcriptome level and explored the cell subpopulations associated with reward behavior in VTA, as well as how these populations communicate with dopaminergic neurons. Following single-nucleus RNA sequencing (snRNA-seq) from VTA, we obtained 96,240 cells of wild-type and Arvcf-KO mice with and without nicotine treatment. Subsequent cell-type abundance analysis revealed a significant reduction in the abundance of neuronal populations upon Arvcf-KO (FDR < 0.05). By integrating single-cell transcriptomics, neurology proteomics, and multiplex immunofluorescence imaging, we discovered a subpopulation of glutamatergic–dopaminergic combinatorial neurons, which is significantly associated with reward. Further cell communication analysis indicated that Arvcf-KO reduced the signal output from these neurons to dopaminergic neurons, represented by Wnt signaling. Finally, integrative analyses of metabolite detection and 2-way analysis of variance based on snRNA-seq indicated that the glutamatergic properties represent a key regulator of reward learning behaviors induced by nicotine and related stimuli. Taken together, our single-cell analysis identified that VTA combinatorial neurons are crucial for nicotine-induced reward through cellular signaling and glutamatergic properties, implying potential new therapeutic targets for addiction treatment.

The mechanism of and effective treatment for the headache, with the global prevalence of 52%, and its common anxiety comorbidity remain elusive. Here, we found that chronic isosorbide dinitrate (ISDN) injections induce c-fos expression in the anterior insula (AI), prelimbic cortex (PrL), and oval nucleus of the bed nucleus of the stria terminalis (ovBNST), suggesting their contributions to headache and anxiety comorbidity. This hypothesis is substantiated by our findings that chronic ISDN injection-induced headache and anxiety are blocked by inhibition of ventral AI (vAI)-PrL and dorsal AI (dAI)-ovBNST circuits, respectively. Headache and anxiety stimuli in chronic ISDN-injected mice markedly increase endocannabinoid (eCB) release at both glutamatergic vAI-PrL synapses and dAI-ovBNST synapses, indicating the role of eCB signaling in modulating headache and anxiety. Indeed, presynaptic knockdown of eCB hydrolase or presynaptic activation of cannabinoid type 1 receptors (CB1Rs) in vAI-PrL and dAI-ovBNST circuits separately alleviates headache and anxiety. A systemic application of eCB degradation enzyme inhibitors blocks chronic ISDN-induced headache and anxiety comorbidity, which are separately blocked by CB1R antagonist application in PrLs and ovBNSTs. Our findings reveal divergent counteracting effects of elevated eCB signaling in vAI-PrL and dAI-ovBNST circuits on comorbid headache and anxiety.

Multipartite entanglement enables secure group key distribution among multiple users while providing immunity against hacking attacks targeting source devices, thereby realizing source-independent quantum conference key agreement (SI-QCKA). However, previous experimental demonstrations of SI-QCKA have encountered substantial technical challenges, primarily due to the low efficiency and scalability limitations inherent in the generation and distribution of multipartite entanglement. Here, we experimentally demonstrate a scalable and efficient SI-QCKA protocol using polarization-entangled photon pairs in a 3-user star network, where Greenberger–Horne–Zeilinger correlations are realized via a post-matching method. We achieve a secure group key rate of 2.11 × 10⁴ bits/s under the single-user channel transmission of 1.64 × 10⁻¹ in a symmetric channel loss network. Additionally, we conduct 6 sets of experiments to investigate the impact of varying channel transmission and random basis selection probabilities on secure key rates. Our work establishes an efficient pathway for SI-QCKA and demonstrates potential scalability for future large-scale multiuser quantum networks.

Organelle morphology and dynamics are closely linked to cellular function and fate, yet their relationships remain poorly defined across physiological and pathological contexts. Live-cell imaging enables the visualization of subcellular structures and dynamic processes but often requires extensive manual analysis, introducing variability and limiting reproducibility and throughput. Image segmentation partitions digital images into meaningful regions, facilitating the quantification of organelle morphology and molecular behavior for precise subcellular analysis. Herein, this review surveys recent advances in live-cell imaging segmentation algorithms across diverse organelles, from traditional thresholding-based methods to deep learning approaches that enhance accuracy and adaptability in complex biological environments. We discuss key challenges, including 3-dimensional imaging, multi-organelle segmentation, and generalization across diverse imaging modalities. We also highlight label-efficient strategies, synthetic data, and physics-guided modeling that reduce reliance on manual annotations and large annotated datasets. By advancing generalist models, these innovations improve quantitative cell biology, accelerate disease research, and drive therapeutic discovery, underscoring the transformative role of artificial intelligence in biomedical microscopy.

MgH₂ (magnesium hydride) has attracted attention as a potential hydrogen storage material owing to its rich availability and high theoretical hydrogen capacity. Nevertheless, its practical utilization is restricted by its high thermodynamic stability and slow hydrogen sorption kinetics. Recent advancements have demonstrated that incorporating various catalytic systems—such as transition metals, metal oxides, metal halides, metal sulfides, and carbon-supported materials—effectively improves hydrogen dissociation, diffusion, and Mg–H bond modulation. Structural transformations, interfacial interactions, and synergistic effects in multicomponent systems substantially enhance MgH₂ performance. Furthermore, cutting-edge computational techniques such as DFT (density functional theory) and ML (machine learning) have become indispensable for expediting catalyst development and forecasting their performance. These computational techniques enable high-throughput screening, provide atomic-scale insights into catalytic mechanisms, and substantially reduce experimental workloads. This review systematically summarizes recent progress in catalytic modifications of MgH₂, elucidates the underlying enhancement mechanisms, highlights the contributions of DFT and ML methodologies, and discusses future directions such as nanostructuring, multifunctional composite catalysts, and computational-driven rational catalyst design for next-generation hydrogen storage technologies.

Probing early-stage crystallization in hydrated environments is crucial for elucidating the microscopic mechanism of crystal growth. However, capturing these processes remains challenging because of the nanometric dimensions of nanocrystals and the dynamic role of water in solvation and ion–ion interactions. Here, we employ a cryogenic scanning probe microscopy platform, which integrates qPlus-type atomic force microscopy with a frozen-solution preparation technique, to directly visualize hydrated sodium chloride (NaCl) nanocrystals at atomic resolution. We observe double- to 5-stranded ionic chain structures, which are hydrated by water dimers. Those structures promote the anisotropic growth of NaCl nanocrystals. Density functional theory calculations reveal that the water dimer can substantially stabilize the chain-shaped configurations by optimizing the water–water and water–ion interactions. In contrast, larger crystals favor isotropic crystalline lattices due to dominant bulk ionic interactions. These findings highlight the unique role of water dimers in the initial crystallization process of salts. Furthermore, this work demonstrates the potential of cryogenic scanning probe microscopy as a powerful tool for probing the crystallization process at atomic resolution, particularly in hydrated environments.