Hong Zhang

The rapid and accurate identification of nucleic acids (NAs) is essential to early diagnosis and treatment of disease in healthcare. Paper‐based analytical devices (PADs) offer a preferred alternative to laborious, expensive, and time‐consuming conventional nucleic acid testing (NAT) methods with considerable success. However, their development has been constrained by challenges in NA recognition and signal amplification within a restricted paper fiber network. Recently developed PADs, which establish a three‐dimensional environment for NA reactions by the integration of porous fibers and modified substrate materials, are limited by the precision of base identification and user convenience. In this review, we focus on paper‐based analytical devices with enhanced point‐of‐care applicability for NA detection and discuss the strategies for enhanced NA cascade signal amplification and background noise suppression. Emphasis is placed on the integrated detection of NA molecules from input to output, highlighting the effective signal amplification and rapid signal output of low‐concentration NA molecules of the paper‐based assay in complicated detection systems.

High-fidelity profiling of intracellular biomarkers is crucial for deciphering tumorigenesis and progression, as well as classifying cell types. Conventional techniques suffer from high background interference and poor stability due to the scattered, point-like signals. Herein, we present a DNA tile–driven in vivo logical encoding self-assembly (TILES) strategy for robust and signal-amplified imaging of multiple biomarkers in cancer cells and tumor-bearing mice. Capitalizing on AND-gated logic operations, the TILES assay enables confined recognition of dual biomarkers to initiate the intracellular self-assembly of DNA scaffolds, yielding nanotube-like signals. Owing to the compact nucleic acid framework, DNA scaffolds exhibit an over sevenfold nuclease resistance enhancement than double-stranded DNA probes. Moreover, the structural output, characterized by high-density fluorophore architecture and defined morphology, substantially boosts the visualization and precise classification of diverse cancer types. The proposed strategy provides a robust platform to enable the high-fidelity, multiplexed profiling of intracellular biomarkers and dynamic gene regulatory networks analysis in living organisms.

Abnormal glycemic management in diabetic patients may lead to severe acute consequences such as hypoglycemia and hyperosmolar coma. Although continuous glucose monitoring (CGM) allows timely detection of dysglycemia, conventional electrochemical sensors are often limited by narrow detection ranges and insufficient sensitivity for reliable long‐term in vivo use. To overcome these limitations, we developed glucose oxidase (GOx)‐functionalized Ti ₃ C ₂ Tx MXene‐Au microneedles (GTA‐MNs) for real‐time interstitial glucose sensing, achieving nanomolar sensitivity and stable operation for over 10 days. Upon skin insertion and contact with glucose, the Ti ₃ C ₂ Tx MXene nanocomposite activates a dual electron‐transfer pathway. It simultaneously facilitates direct electron transfer through the enzymatic FAD/FADH ₂ center of GOx and catalyzes the indirect oxidation of enzymatically generated H ₂ O ₂ . This synergistic mechanism enables a broad detection range from 10 nM to 50 mM and an ultralow detection limit of 1.444 nM. A support vector regression calibration model was implemented to dynamically align sensor signals with blood glucose levels, thereby compensating for physiological lag. Comprehensive in vitro and in vivo validations confirmed the system's high accuracy, stability, and biocompatibility. Furthermore, we engineered a user‐friendly home‐diagnostic system that wirelessly pairs the wearable patch with a smartphone, enabling CGM in freely moving subjects to support personalized diabetes management.

Sepsis-associated secondary infection often leads to a high mortality rate. Dysfunctional macrophages are primary contributors to inadequate antimicrobial defense in patients with sepsis-associated immunosuppression. Rejuvenating macrophage antibacterial capacity is beneficial for host defense against secondary infection. Here, we developed “BATMAN” (bacteria-targeted transformable macrophage nanorejuvenator), a self-assembling peptide nanoparticle to tackle sepsis-associated secondary infection by coordinating the arrest of invasive bacteria and rejuvenation of dysfunctional macrophages. BATMAN comprises a bacteria-targeting ubiquicidin peptide domain, bacterial lipase–sensitive cholesteryl hemisuccinate, an assembly-driving FFVLK domain, and the immunoglobulin G–derived tuftsin peptide. Upon activation by bacterial lipase, the particles undergo an inside-out transformation and assembly to expose and cluster the concealed tuftsin peptides for interaction with macrophage Fcγ receptors. Interaction of tuftsin clusters with macrophage Fcγ receptors enhanced bacterial phagocytosis and drove macrophage repolarization. In a cecal slurry–induced septic mouse model with secondary pulmonary infection, BATMAN treatment improved survival rates and rejuvenated the sepsis-compromised immune response to address polymicrobial- and multidrug-resistant pathogen–induced pulmonary infections. These findings suggest that BATMAN holds promise for further development as a therapeutic alternative for sepsis-associated secondary infection.

Sepsis is a leading cause of hospital mortality, and its significant heterogeneity complicates prognosis and stratification. To address this challenge, we developed an explainable artificial intelligence prognostic model (SepsisFormer, a transformer-based neural network) and an automated risk-stratification tool (SMART) for sepsis. In a multi-center retrospective study of 12,408 sepsis patients, SepsisFormer achieved high predictive accuracy (AUC: 0.9301, sensitivity: 0.9346, and specificity: 0.8312). SMART (AUC: 0.7360) surpassed most established scoring systems. Seven coagulation-inflammatory routine laboratory measurements and patient age were identified to classify patients’ four risk levels (mild, moderate, severe, dangerous) and two subphenotypes (CIS1 and CIS2), each with distinct clinical characteristics and mortality rates. Notably, patients with moderate/severe levels or CIS2 derive more significant benefits from anticoagulant treatment. Our work, therefore, offers a set of simple, real-time executable tools for sepsis heterogeneity, demonstrating the potential to enhance sepsis clinical practice globally, particularly in resource-constrained healthcare settings.