Endogenous opioid peptides have been linked to numerous physiological functions, including pain perception and the motivational drive associated with substance use disorders, but many fundamental aspects of transmission remain ambiguous. The kinetics of endogenous opioid peptides are thought to be slower and to last longer than those of more classical small-molecule neurotransmitters, like dopamine; however, a direct comparison of the release and diffusive spread of these molecules in the brain is lacking. Here, fast-scan cyclic voltammetry was coupled with carbon microelectrodes for co-detection of dopamine and met-enkephalin at single recording sites in rat striatal slices. The measurements used a voltammetric waveform that was specifically designed to minimize sensitivity to dopamine, maximize sensitivity to enkephalin, and minimize biofouling. Both neurotransmitter (dopamine) and neuropeptide (met-enkephalin, M-ENK) release scaled with stimulation duration. Interestingly, ENK dynamics in striatum displayed a unique biphasic profile with a significant latency to peak that occurred ∼30 s after stimulation, suggesting a sphere of influence that was ∼3x larger than that of dopamine. Mathematical modeling of the evoked M-ENK concentration profile suggests that multiple forms of ENK were released at once, such that some of the five-amino-acid form of M-ENK was released in exocytosis, and some was generated in the extracellular space by enzymatic cleavage of a larger form of ENK. Finally, a series of experiments combined solid-phase extraction with liquid-chromatography mass spectrometry to independently verify ENK release. The findings provide direct evidence to support widely held assumptions regarding neuropeptide release, and they demonstrate how different classes of signaling molecules can potentially affect distinct cellular populations in striatum─even when released at the same site.

The bacterial surface is rich in diverse molecular features, and fully exploiting these natural recognition mechanisms provides innovative avenues for multimechanism detection of foodborne pathogens. Here, we developed a label-free, dual-modal LFIA platform based on the glycan-cluster effect for the efficient capture of Salmonella. The platform employs dextran-functionalized tungsten diselenide nanosheets (Dex-WSe2) as core probes, where dextran coatings provide antibody-like high-affinity capture through multivalent glycan-bacteria interactions, while WSe2 nanosheets act as dual signal transducers. Benefiting from exciton-plasmon coupling and charge-transfer effects, WSe2 nanosheets not only possess inherent surface-enhanced Raman scattering (SERS) activity but also display distinct visible coloration due to their unique optical properties. Leveraging these dual features, the platform enables highly sensitive bimodal detection, achieving a visual detection limit of 103 CFU/mL and an ultralow SERS detection limit of 52 CFU/mL. Furthermore, machine learning was introduced for multidimensional signal analysis: k-nearest neighbors (KNN) for qualitative concentration classification and random forest (RF) regression for quantitative prediction. The integrated model achieved 100% classification accuracy and an R2 of 0.9977, demonstrating outstanding robustness. By combining glycan-based molecular recognition with machine learning strategies, the Dex-WSe2 probe offers an efficient, stable, and intelligent platform for rapid on-site pathogen detection and food safety monitoring.

Traditional electrochemical analysis with single-working-electrode systems is effective for targeting individual reactions or species but falls short in simultaneously analyzing multiple analytes. Two-working-electrode systems, such as the rotating ring-disk electrode, address this limitation but suffers from several drawbacks, such as low collection efficiency, high background noise, and bulky size, which significantly limit its use toward a wider range. In this study, we proposed the design and fabrication of a simple, low-cost, compact, and user-friendly two-working-electrode system: the micro-ring-disk electrodes (MRDEs). The fabrication process reliably produces high-performance MRDEs using various materials, with disk and ring sizes ranging from 5 to19 μm. Taking advantage of the unique dual-working-electrode structure, the MRDEs enable in situ detection with multiple targets. To demonstrate this capability, the selective detection of dopamine (DA) was performed in the presence of ascorbic acid (AA), a common interference. By holding the microring electrode at a constant potential to oxidize AA, DA was accurately detected at the microdisk electrode, even when AA interference was present at four times the concentration of DA. This work not only provides a practical solution for the easy fabrication of two-working-electrode systems but also offers valuable insights into their application of selective and multitarget detection in complex environments.

Shiga toxin type 2 (Stx2) is a risk factor for acute bloody diarrhea and hemolytic uremic syndrome, which has posed a major threat to global public health. Currently, there is a lack of effective strategies to combat Stx2 infection. Herein, according to the structural characteristics and infection mechanism of Stx2, we have designed a structure mediated by self-assembling DNA tetrahedral nanoframes (TDN), which can orderly assemble three neutral aptamers (TDN-S) to prevent Stx2 from entering cells and inhibit Stx2 infection. The neutralizing aptamers SA, SB, and SC can respectively bind to the three sites (sites 1, 2, and 3) on the Stx2 B subunit, which coincidentally occupy the binding sites of Stx2 and Gb3 receptor. Benefiting from the simultaneous capture of three active sites, spatial matching, and the steric hindrance effect, TDN-S exhibits excellent neutralization efficiency against Stx2. The inhibition ability of Stx2 infection in vitro was evaluated by cell viability, intracellular fluorescence imaging, and ROS levels, and the inhibition rate was as high as 80.44%. The TDN-S significantly reduces the damage to the liver, kidneys, and intestines and the inflammatory response in vivo. Overall, the neutralization strategy with multimodal interaction provides a new way to alleviate the Stx2 toxicity and serves as a model for other wide-range toxin-like inhibitions.

Vascular endothelial growth factor (VEGF) is a key signaling protein in the regulation of angiogenesis, and the dysregulation of its expression is associated with the development of various diseases such as tumorigenesis. The traditional fluorescence detection methods for the VEGF are prone to photobleaching and are easily interfered by autofluorescence of the sample, resulting in insufficient detection stability and accuracy. In order to develop a stable and accurate method for VEGF detection, an ultrasensitive surface-enhanced Raman scattering biosensor was constructed in this study by designing a scalable Au/DNA/SiO2 nanoprobe (named ADSNP). In the ADSNP, we labeled the Prussian blue (PB) Raman molecules, whose characteristic peaks are in the silent region, effectively avoiding signal interference from the biomolecules. Meanwhile, the distance modulation between Raman molecules and Au nanoparticles was achieved by using stretchable DNA nanostrands, which in turn responded to different concentrations of the VEGF. Assisted by the aptamer nucleic acid amplification strategy, trace VEGF would cause a high Raman signal response. In addition, the ADSNP can image the VEGF in living cancer cells. Therefore, this study provides a feasible method for the identification and detection of cancer markers in cells, as well as a potential value for observing cellular biochemical responses.

Pure shift NMR spectra, in which multiplet structure is suppressed, are widely used but exact a high price in sensitivity. Here we present CUPID (Computer-assisted Undiminished-sensitivity Protocol for Ideal Decoupling), which uses parametric estimation to produce pure shift NMR spectra from easily acquired 2D J-resolved (2DJ) data sets. Unlike previous practical methods for broadband pure shift NMR, it makes use of all of the available signal. CUPID is therefore effective even at sample concentrations where current methods are too insensitive to yield usable spectra. As an additional benefit, the estimation method used allows the extraction of individual multiplet structures from overlapping spectra. CUPID is freely available through NMR Estimation in Python (NMR-EsPy), an open-source package with a simple-to-use API, and comes with a graphical user interface that is accessible via Topspin, a widely used NMR software platform.

Natural products (NPs) are a treasure trove of drug discovery, yet their structural complexity and extreme data scarcity critically hinder AI-driven exploration. To address this challenge, we present MSformer, a transformer-based architecture that bridges this gap by leveraging molecule fragments to systematically encode NP chemical space. These fragments were generated by a mass spectrometry-inspired fragmentation algorithm, termed meta-structures. Unlike chemical models pretrained on comprehensive molecule databases, MSformer is totally pretrained on very limited NP data set by deconstructing 400,000 NPs into 234 million meta-structures. This design enables MSformer to capture the structural richness and drug-like relevance of NPs. Evaluated on 14 tasks across MoleculeNet and the Therapeutics Data Commons data sets, MSformer outperforms state-of-the-art models, demonstrating superior generalizability in property prediction. The abundant meta-structures enable MSformer hierarchical interpretability that reveals task-specific structural determinants and successfully deconstructing approved drugs into bioactive fragments. By integrating domain knowledge with deep learning, MSformer establishes a transformative paradigm for NP-based drug discovery, offering a scalable framework to navigate nature's underexplored chemical repertoire and accelerate the identification of bioactive candidates.

Desorption electrospray ionization-mass spectrometry imaging (DESI-MSI) is among the most powerful techniques for visualizing the spatial distribution of small organic molecules, particularly lipids, on tissue surfaces. Conventional DESI-MSI analysis typically involves sectioning fresh-frozen tissues or, less commonly, embedding samples in matrices specifically formulated to preserve the tissue integrity for multifunctional analyses. In this study, we present an optimized sucrose-gelatin hydrogel matrix compatible with DESI-MSI, using mouse brain tissue as a model system. The method involves low-temperature embedding of frozen specimens into the hydrogel matrix, followed by snap-freezing at -160 °C. This matrix formulation ensures minimal background interference and prevents metabolite delocalization, thereby preserving the native molecular composition of the tissue. Notably, sucrose-derived adduct ions restricted to the embedding medium serve as stable internal reference signals in both positive and negative ionization modes. These signals enable continuous lock-mass correction throughout acquisition, offering a new solution to the unresolved challenge for accurate mass-based measurements in DESI-MSI without an infusion of exogenous calibration standards. Complementary DESI-MS/MS analyses further facilitate confident lipid identification and resolve structural ambiguities. Moreover, the sucrose-gelatin embedding medium provides excellent preservation of tissue morphology and antigenicity, supporting subsequent histological and immunohistochemical analyses. Overall, this sucrose-based hydrogel embedding protocol offers a robust, reproducible, and multimodal platform for molecular tissue imaging by DESI-MSI, especially in delicate biological specimens with broad translational potential across preclinical and clinical research domains.

The current limited understanding of small bioactive peptide metabolism hinders effective doping control. A primary challenge lies in distinguishing suspicious features from extensive mass spectral datasets contaminated by biological matrix interference and background noise. In this study, we introduce a strategy integrating in silico prediction and nontargeted data mining to achieve more comprehensive metabolic profiling through ultrahigh-performance liquid chromatography-high-resolution mass spectrometry (UHPLC-HRMS). The strategy operates by identifying and applying chemical similarity (CSIM) rules of peptides (such as LC/MS behaviors and specific biotransformation) to mine unknown metabolites. With this strategy, a semiautomated workflow utilizing computational software and custom script was constructed and applied to the human liver microsomal metabolism of two significant kisspeptin analogues with doping potential (TAK-448 and TAK-683). The characteristic behaviors induced by CSIM enabled effectively in-depth data mining from redundant background signals, leading to the identification of 13 metabolites (three were validated via synthetic standards) and two uncommon biotransformation pathways (N-terminal vinylation and N-terminal carboxylation) for both investigated compounds. Notably, the two biotransformation pathways offered an innovative perspective on small peptide metabolism, which was further confirmed in rats, and produced two promising long-term metabolites for monitoring doping abuse. Further drug activity evaluation indicated higher or retained performance-enhancing effects of these metabolites, alerting doping control subjects to pay attention to more information. The study provided the first comprehensive characterization of the metabolic profiles of TAK-448 and TAK-683, while also offering an effective tool for metabolism research on small peptide doping agents.

Mass spectrometry imaging (MSI) is a powerful multiplexed biochemical imaging modality. It relies on raster scanning for localized data acquisition, which can be time-consuming, limiting applications of high-resolution tissue mapping and 3D reconstruction. This work presents a computational framework that integrates a raster scanning forward model with a deep learning prior to reconstruct high-resolution ion images from sparsely sampled pixels. The deep learning prior, implemented as a pretrained network-based denoiser, is incorporated into a plug-and-play-based iterative reconstruction algorithm without retraining for different acquisition settings. We show that our method can reconstruct high-fidelity ion images from sparse data acquired with different MSI instruments, acquisition settings, and tissue types without requiring additional training. Notably, our approach generalizes robustly to biologically and structurally distinct tissues, such as from brain to kidney sections, highlighting its potential for broad deployment in various experimental MSI workflows.