Deep Learning Promotes Profiling of Multiple miRNAs in Single Extracellular Vesicles for Cancer Diagnosis.

Extracellular vesicle microRNAs (EV miRNAs) are critical noninvasive biomarkers for early cancer diagnosis. However, accurate cancer diagnosis based on bulk analysis is hindered by the heterogeneity among EVs. Herein, we report an approach for profiling single-EV multi-miRNA signatures by combining total internal reflection fluorescence (TIRF) imaging with a deep learning (DL) algorithm for the first time. This innovative technique allows for the precise characterization of EV miRNAs at the single-vesicle level, overcoming the challenges posed by EV heterogeneity. TIRF with high resolution and a signal-to-noise ratio can simultaneously detect multi-miRNAs in situ in individual EVs. DL algorithm avoids complicated and inaccurate artificial feature extraction, achieving automated high-resolution image analysis. Using this approach, we reveal that the main variation of EVs from 5 cancer cells and normal plasma is the triple-positive EV subpopulation, and the classification accuracy of single triple-positive EVs from 6 sources can reach above 95%. In the clinical cohort, 20 patients (5 lung cancer, 5 breast cancer, 5 cervical cancer, and 5 colon cancer) and 5 healthy controls are predicted with an overall accuracy of 100%. This single-EV strategy provides new opportunities for exploring more specific EV biomarkers to achieve cancer diagnosis and classification.

Phosphorescence approach based on silica protected carbon dots for autofluorescence interference-free and highly selective detection of fluoride.

BACKGROUND: Fluoride (F-), an anion with the smallest ionic radius and highest charge density, plays an important role in biomedical and environmental processes, making the development of accurate F- detection methods of great importance. Fluorometric methods with simplicity and sensitivity have gained considerable attention in F- detection. However, their accuracy faces challenges due to issues like autofluorescence interference during real-time light excitation and limited selectivity. Therefore, it is important to establish a simple, real-time light excitation-free, and highly selective method for the accurate determination of F- in complicated samples. RESULTS: Herein, a novel phosphorescent approach is developed for the selective and accurate detection of F- in complex samples. Phosphorescence emission CDs@SiO2 is fabricated by confining CDs in a silica protective layer. This design retains the favorable water solubility of silica while benefitting from its inertness, making it resistant to most substances. Furthermore, phosphorescent analysis without real-time light excitation eliminates autofluorescence interference, significantly improving the signal-to-noise ratio (SNR) and simplifying sample pretreatment. The specific interaction between F- and the Si-O bond can lead to the degradation of the silica protective layer, exposing the CDs to the solution, resulting in phosphorescence quenching, achieving the highly accurate and sensitive detection of F- with a linear range of 0.001-4 mM and a limit of detection (LOD) of 1 µM. SIGNIFICANCE: This novel F- phosphorescence method based on the metal-free phosphorescent nanomaterial CDs@SiO2 integrates the benefits of no autofluorescence interference, high selectivity, and full aqueous compatibility, and its combination with a smartphone provides a simple, portable, and cost-effective detection platform for accurate and highly sensitive determination of F- in complex samples.

Phosphorescence, fluorescence, and colorimetric triple-mode sensor for the detection of acid phosphatase and corresponding inhibitor.

Acid phosphatase (ACP) as a clinical diagnostic biomarker for several pathophysiological diseases has aroused widespread interest. Compared to commonly developed single-mode ACP detection technology, the multi-mode detection method with self-validation can provide more reliable results. Herein, we proposed a triple-mode phosphorescence, fluorescence, and colorimetric method for ACP detection in combination with CDs@SiO2. HAuCl4 with oxidase-like activity can catalyze the oxidation of colorless 3,3',5,5'-tetramethylbenzidine (TMB) to the blue oxide TMB (TMBox), offering absorption signals and quenching the phosphorescence and fluorescence of CDs@SiO2 based on the internal filtration effect (IFE). ACP can hydrolyze ascorbic acid 2-phosphate (AAP) to yield ascorbic acid (AA), thereby reducing TMBox to TMB, triggering solution fading and restoring phosphorescence and fluorescence signals. When the ACP inhibitor malathion is present, the reduction of TMBox is hindered, which successively led to the suppression of CDs@SiO2 phosphorescence and fluorescence signal recovery. According to these principles, triple-mode ACP (LOD = 0.0026 mU mL-1) and malathion detections (LOD = 0.039 µg mL-1) with favorable accuracy and sensitivity are realized. With simplicity, robustness, and versatility, the triple-mode sensor can be extended to the detection of the AAP hydrolase family and the screening of corresponding inhibitors.

A machine learning-based colorimetric sensor array for high-precision pathogen identification in household refrigerators.

A precise and convenient sensor was constructed to identify pathogens in household refrigerators (4 °C, RH = 55%) by integrating a volatile organic compound fingerprint-responsive gel-based colorimetric sensor array and a neural network. The platform is expected to be extended to the intelligent food packaging field and has promise for point-of-need monitoring of pathogens.

Dual Functional Full-Color Carbon Dot-Based Organelle Biosensor Array for Visualization of Lipid Droplet Subgroups with Varying Lipid Composition in Living Cells.

In situ visualization of lipid composition diversity in lipid droplets (LDs) is essential for decoding lipid metabolism and function. However, effective probes for simultaneously localizing and reflecting the lipid composition of LDs are currently lacking. Here, we synthesized full-color bifunctional carbon dots (CDs) that can target LDs as well as respond to the nuance in internal lipid compositions with highly sensitive fluorescence signals, due to lipophilicity and surface state luminescence. Combined with microscopic imaging, uniform manifold approximation and projection, and sensor array concept, the capacity of cells to produce and maintain LD subgroups with varying lipid composition was clarified. Moreover, in oxidative stress cells, LDs with characteristic lipid compositions were deployed around mitochondria, and the proportion of LD subgroups changed, which gradually disappeared when treated with oxidative stress therapeutics. The CDs demonstrate great potential for in situ investigation of the LD subgroups and metabolic regulations.

Nanozyme Sensor Array Plus Solvent-Mediated Signal Amplification Strategy for Ultrasensitive Ratiometric Fluorescence Detection of Exosomal Proteins and Cancer Identification.

Tumor exosomes with molecular marker-proteins inherited from their parent cells have emerged as a promising liquid biopsy biomarker for cancer diagnosis. However, facile, robust, and sensitive detection of exosomal proteins remains challenging. Therefore, a nanozyme sensor array is constructed by using aptamer-modified C3N4 nanosheets (Apt/C3N4 NSs) together with a solvent-mediated signal amplification strategy for ratiometric fluorescence detection of exosomal proteins. Three aptamers specific to exosomal proteins are selected to construct Apt/C3N4 NSs for high specific recognition of exosomal proteins. The adsorption of aptamers enhances the catalytic activity of C3N4 NSs as a nanozyme for oxidation of o-phenylenediamine (oPD) to 2,3-diaminophenazine (DAP). In the presence of target exosomes, the strong affinity between aptamer and exosome leads to the disintegration of Apt/C3N4 NSs, resulting in a decrease of catalytic activity, thereby reducing the production of DAP. The ratiometric fluorescence signal based on a photoinduced electron transfer (PET) effect between DAP and C3N4 NSs is dependent on the concentration of DAP generated, thus achieving highly facile and robust detection of exosomal proteins. Remarkably, the addition of organic solvent-1,4-dioxane can sensitize the luminescence of DAP without affecting the intrinsic fluorescence of C3N4 NSs, achieving the amplification of the aptamer-exosome recognition events. The detection limit for exosome is 2.5 × 103 particles/mL. In addition, the accurate identification of cancer can be achieved by machine learning algorithms to analyze the difference of exosomal proteins from different patients' blood. We hope that this facile, robust, sensitive, and versatile nanozyme sensor array would become a promising tool in the field of cancer diagnosis.

MnO2-graphene oxide hybrid nanomaterial with oxidase-like activity for ultrasensitive colorimetric detection of cancer cells.

Robust and sensitive cell-based enzyme-linked immunosorbent assay (CELISA) is of great significance in the diagnosis and screening of cancer. However, the method is limited by the high rate of negative results attributed to the instability of horseradish peroxidase (HRP), H2O2, and antibody. Here, we construct a folic acid-functionalized in situ-grown MnO2 nanosheet/graphene oxide hybrid (FA-MnO2/GO) with oxidase-like activity instead of the anti-folate receptor antibody in traditional CELISA to resist the possible negative interference arising from unstable HRP, H2O2, and antibodies for more robust colorimetric detection of cancer cells. The functionalization of FA enables the selective binding between hybrid and cancer cells through the over-expressed folate receptor, and then the binding events are converted into quantitative colorimetric signals though the oxidation of the chromogenic substrate TMB catalyzed by MnO2, allowing the detection of cancer cells with colorimetric method. Moreover, the construction of MnO2/GO hybrid can synergistically enhance the oxidase-like activity of MnO2 and promote its dispersion in water, further ensuring the accuracy and sensitivity of the detection. A detection limit of 20 cancer cells is obtained by a plate reader, which is lower than those obtained by most reported CELISA methods for cancer cell detection, and as few as 75 cancer cells can be identified by the naked eye. This study not only provides a multifunctional sensing platform for robust and sensitive cancer cell detection, but also offers a promising oxidase-like mimic in the field of bioanalysis.

Integral Multielement Signals by DNA-Programmed UCNP-AuNP Nanosatellite Assemblies for Ultrasensitive ICP-MS Detection of Exosomal Proteins and Cancer Identification.

Exosomes are expected to be used as cancer biomarkers because they carry a variety of cancer-related proteins inherited from parental cells. However, it is still challenging to develop a sensitive, robust, and high-throughput technique for simultaneous detection of exosomal proteins. Herein, three aptamers specific to cancer-associated proteins (CD63, EpCAM, and HER2) are selected to connect gold nanoparticles (AuNPs) as core with three different elements (Y, Eu, and Tb) doped up-conversion nanoparticles (UCNPs) as satellites, thereby forming three nanosatellite assemblies. The presence of exosomes causes specific aptamers to recognize surface proteins and release the corresponding UCNPs, which can be simultaneously detected by inductively coupled plasma-mass spectrometry (ICP-MS). It is worth noting that rare earth elements are scarcely present in living systems, which minimize the background for ICP-MS detection and exclude potential interferences from the coexisting species. Using this method, we are able to simultaneously detect three exosomal proteins within 40 min, and the limit of detection for exosome is 4.7 × 103 particles/mL. The exosomes from seven different cell lines (L-02, HepG2, GES-1, MGC803, AGS, HeLa, and MCF-7) can be distinguished with 100% accuracy by linear discriminant analysis. In addition, this analytical strategy is successfully used to detect exosomes in clinical samples to distinguish stomach cancer patients from healthy individuals. These results suggest that this sensitive and high-throughput analytical strategy based on ICP-MS has the potential to play an important role in the detection of multiple exosomal proteins and the identification of early cancer.

One-Step Synthesis of Carbon Nanoparticles Capable of Long-Term Tracking Lipid Droplet for Real-Time Monitoring of Lipid Catabolism and Pharmacodynamic Evaluation of Lipid-Lowering Drugs.

Lipid droplets (LDs) are intracellular lipid-rich organelles, which not only serve as neutral lipid reservoirs but also involve in many physiological processes and are associated with a variety of metabolic diseases and cancers. Long-term tracking of the state and behavior of LDs is of great significance but challenging. The difficulty is largely due to the lack of low cytotoxicity, high photobleaching resistance, and long intracellular retention probes that are capable of long-term tracking LDs. Herein, we report the discovery of two amphiphilic LD-targeting carbon nanoparticles (CNPs, i.e., CPDs and CDs) prepared by one-step room-temperature and hydrothermal methods. Their high lipid-water partition coefficient (log P > 2.13) and strong positive solvatochromism property ensure the quality of LD imaging. Especially, CDs exhibit favorable biocompatibility (2 mg mL-1, cell viability >90%), excellent photostability (after continuous laser irradiation on a confocal microscope for 2 h, relative FL intensity >85%), and superior intracellular retention ability, thereby enabling long-term tracking of LDs in hepatocytes for up to six passages. Based on the excellent long-term tracking ability, CDs are successfully applied to observe autophagy in a typical catabolic process and to evaluate the effect of a commonly used lipid-lowering drug atorvastatin on hepatocyte lipid uptake.

A carbon-based polymer dot sensor for breast cancer detection using peripheral blood immunocytes.

We constructed a carbon-based polymer dot (CPD) sensor to detect breast cancer based on the differences of peripheral blood cells, providing a new minimally invasive method for cancer diagnosis. This simple and extensible system exhibits clinically relevant accuracy in terms of cancer identification, making it an attractive strategy for diagnosis and prognosis.

Discrimination of antibiotic-resistant Gram-negative bacteria with a novel 3D nano sensing array.

A novel sensor array based on a (+)AuNP/AuNC nanocomposite was constructed for the selective discrimination of 10 types of Gram-negative bacteria (including 3 types of antibiotic-resistant strains) at a low concentration level of OD600 = 0.015. By recognizing the triple optical patterns of Gram-negative bacteria with the assistance of LDA, the sensor array is able to group the bacteria with respect to their species to each other.

A Novel Three-Dimensional Nanosensing Array for the Discrimination of Sulfur-Containing Species and Sulfur Bacteria.

The discrimination of various sulfur -containing species helps us to deeply understand how sulfur affects cellular signaling and other physiological events. Herein, we present a three-dimensional sensor array based on simultaneous variation of the optical properties (fluorescence, light scattering, and UV-vis absorption) of gold-silver alloy nanocluster (AuAgNCs)-gold nanoparticle (AuNPs) composite for the rapid identification of 13 sulfur-containing species and sulfur-oxidizing bacteria. The sensor array is fabricated based on the strong coordination interactions between sulfur-containing compounds and AuAgNCs on the surface of AuNPs. The sulfur species of interest exhibit different affinities toward AuAgNCs and generate unique optical properties. These response patterns could divide the analytes into three categories including organic sulfide, inorganic sulfide, and thiols. Thirteen types of sulfur species including cystine, methionine, GSSG, S2-, SO32-, S2O32-, S2O72-, S2O82-, S4O62-, GSH, N-acetyl-l-cysteine, homocysteine, and cysteine can be well distinguished by the sensor array with principal component analysis at 0.5 µM. Moreover, sulfur-oxidizing bacteria and nonsulfur bacteria can also be well discriminated at a level of OD600 = 0.005.