A deep learning based holistic diagnosis system for immunohistochemistry interpretation and molecular subtyping.

BACKGROUND: Breast cancer in different molecular subtypes, which is determined by the overexpression rates of human epidermal growth factor receptor 2 (HER2), estrogen receptor (ER), progesterone receptor (PR), and Ki67, exhibit distinct symptom characteristics and sensitivity to different treatment. The immunohistochemical method, one of the most common detecting tools for tumour markers, is heavily relied on artificial judgment and in clinical practice, with an inherent limitation in interpreting stability and operating efficiency. Here, a holistic intelligent breast tumour diagnosis system has been developed for tumour-markeromic analysis, combining the automatic interpretation and clinical suggestion. METHODS: The holistic intelligent breast tumour diagnosis system included two main modules. The interpreting modules were constructed based on convolutional neural network, for comprehensively extracting and analyzing the multi-features of immunostaining. Referring to the clinical classification criteria, the interpreting results were encoded in a low-dimensional feature representation in the subtyping module, to efficiently output a holistic detecting result of the critical tumour-markeromic with diagnosis suggestions on molecular subtypes. RESULTS: The overexpression rates of HER2, ER, PR, and Ki67, as well as an effective determination of molecular subtypes were successfully obtained by this diagnosis system, with an average sensitivity of 97.6 % and an average specificity of 96.1 %, among those, the sensitivity and specificity for interpreting HER2 were up to 99.8 % and 96.9 %. CONCLUSION: The holistic intelligent breast tumour diagnosis system shows improved performance in the interpretation of immunohistochemical images over pathologist-level, which can be expected to overcome the limitations of conventional manual interpretation in efficiency, precision, and repeatability.

A pH-Mediated Highly Selective System Enabling Simultaneous Analysis of Circulating RNAs Carried by Extracellular Vesicles and Lipoproteins.

Extracellular vesicles (EVs) and lipoproteins (LPPs) serve as important carriers of circulating miRNAs in peripheral blood, offering immense potential for disease diagnosis and therapeutic interventions. Due to their shared physicochemical attributes, EVs and LPPs are frequently coisolated, potentially leading to misunderstandings regarding their distinct functional roles in physiological and pathological processes. Here, we report a highly selective magnetic system based on the pH-mediated affinity displayed by cibacron blue (CB) toward EVs and LPPs, enabling successful separation and collection of these two nanoparticles without cross-contamination for subsequent circulating RNA analysis. First, we found that CB-modified magnetic beads (CBMBs) exhibit a strong affinity toward LPP particles while displaying little interaction with EVs in standard samples under physiological pH conditions. We further demonstrate that the affinity between CB molecules and bionanoparticles in plasma samples is highly pH-dependent. Specifically, CBMBs show affinities for both LPP and EV particles under neutral and acidic conditions. However, at basic pH levels, CB molecules selectively bind only to LPP particles. Consequently, the remaining EV particles present in plasma are subsequently isolated by using titanium dioxide-modified beads (TiMBs) through phospholipid affinity. The simultaneous analysis of the transcriptomic contents of EV and LPP reveals clear differences in their small RNA profiles, with the differentially expressed RNAs reflecting distinct biological processes. Significantly, in a proof-of-concept study, we successfully demonstrated a strong correlation between miRNAs carried by both EV and LPP particles with the occurrence of ocular neovascularization during the progression of diabetic retinopathy. The involved miRNAs may serve as potential biomarkers for DR diagnostics and severity classification. To sum up, this pH-mediated separation system is not only user-friendly but also highly compatible, rendering it a potent tool for probing the molecular compositions, biomarkers, and underlying biological mechanisms of EVs and LPPs.

Quantitative metabolic analysis of plasma extracellular vesicles for the diagnosis of severe acute pancreatitis.

BACKGROUND: Severe acute pancreatitis (SAP) is the most common gastrointestinal disease and is associated with unpredictable seizures and high mortality rates. Despite improvements in the treatment of acute pancreatitis, the timely and accurate diagnosis of SAP remains highly challenging. Previous research has shown that extracellular vesicles (EVs) in the plasma have significant potential for the diagnosis of SAP since the pancreas can release EVs that carry pathological information into the peripheral blood in the very early stages of the disease. However, we know very little about the metabolites of EVs that might play a role in the diagnosis of SAP. METHODS: Here, we performed quantitative metabolomic analyses to investigate the metabolite profiles of EVs isolated from SAP plasma. We also determined the metabolic differences of EVs when compared between healthy controls, patients with SAP, and those with mild acute pancreatitis (MAP). RESULTS: A total of 313 metabolites were detected, mainly including organic acids, amino acids, fatty acids, and bile acids. The results showed that the metabolic composition of EVs derived from SAP and MAP was significantly different from those derived from healthy controls and identified specific differences between EVs derived from patients with SAP and MAP. On this basis, we identified four biomarkers from plasma EVs for SAP detection, including eicosatrienoic acid (C20:3), thiamine triphosphate, 2-Acetylfuran, and cis-Citral. The area under the curve (AUC) was greater than 0.95 for both discovery (n = 30) and validation (n = 70) sets. CONCLUSIONS: Our data indicate that metabolic profiling analysis of plasma EVs and the screening of potential biomarkers are of significant potential for improving the early diagnosis and severity differentiation of acute pancreatitis.

Artificial Intelligence-Aided Multiple Tumor Detection Method Based on Immunohistochemistry-Enhanced Dark-Field Imaging.

The immunohistochemical method serves as one of the most practical tools in clinical cancer detection and thus has great application value to overcome the existing limits of the conventional method and further improve the detecting efficiency and sensitivity. This study employed 3,3'-diaminobenzidine (DAB), a conventional color indicator for immunohistochemistry, as a novel high-sensitive scattering reagent to provide a multidimensional image signal varying with the overexpression rate of tumor markers. Based on the scattering properties of DAB aggregates, an efficient and robust artificial intelligence-aided immunohistochemical method based on dark-field imaging has been established, with improvement in both the imaging quality and interpretation efficiency in comparison with the conventional manual-operated immunohistochemical method. Referencing the diagnosis from three independent pathologists, this method succeeded in detecting HER2 overexpressed breast tumors with a sensitivity of 95.2% and a specificity of 100.0%; meanwhile, it was found to be applicable for non-small-cell lung tumors and malignant lymphoma as well. As demonstrated, this study provided an effective and reliable means for making diagnostic suggestions, which exhibited great potential in multiple tumor pathological detection at low cost.

Metabolomic analysis of exosomal-markers in esophageal squamous cell carcinoma.

Esophageal squamous cell carcinoma (ESCC) is a worldwide malignancy with high mortality rates and poor prognosis due to the lack of effective biomarkers for early detection. Exosomes have been extensively explored as attractive biomarkers for cancer diagnosis and treatment. However, little is known about exosome metabolomics and their roles in ESCC. Here, we performed a targeted metabolomic analysis of plasma exosomes and identified 196 metabolites, mainly including lipid fatty acids, benzene, amino acids, organic acids, carbohydrates and fatty acyls. We systematically compared metabolome patterns of exosomes via machine learning from patients with recrudescence and patients without recrudescence and demonstrated a marker set consisting of 3'-UMP, palmitoleic acid, palmitaldehyde, and isobutyl decanoate for predicting ESCC recurrence with an AUC of 98%. These metabolome signatures of exosomes retained a high absolute fold change value at all ESCC stages and were very likely associated with cancer metabolism, which could be potentially applied as novel biomarkers for diagnosis and prognosis of ESCC.

Ultrafiltration combing with phospholipid affinity-based isolation for metabolomic profiling of urinary extracellular vesicles.

Recent years have seen the field of extracellular vesicle (EV) studies burgeoning. This is mainly because EV constituents including nucleic acid, proteins, lipids, and metabolites are promising sources towards disease biomarker discovery. However, EV study remains challenging due to the complexity of biofluids as well as technical limitations during sample preparation. Here, we proposed a simple method combing ultrafiltration (UF) and phospholipid affinity to collect high purity EVs from 30 mL of urine sample for their metabolomic profiling. Ultracentrifugation (UC) for EV isolation was applied as a reference method. Western blot (WB) analysis, nanoparticles tracking analysis (NTA) and electron microscopy (EM) were used to assess EV protein markers and to characterize vesicle size and morphology. The results revealed that more than 1010 EV particles could be isolated from a 30 mL urine sample by the proposed method, and the resulting EVs carry specific protein markers and had a typical "cup shape" morphology. This suggests that our method is suitable for EV isolation and can provide sufficient EV quantity to ensure downstream analysis. Further untargeted metabolomic profiling of isolated EVs by UHPLC-QTOF-MS detected 433 metabolites by our methods and 432 metabolites by UC with a MS/MS similarity score greater than 0.7. We then applied the lipid metabolites-targeted method using UHPLC-QTrap-MS with the MRM mode, which successfully detected 467 compounds from urine EVs. This indicates that UF integrating phospholipid affinity is a reliable method for metabolic analysis of urinary EVs, which holds the potential for EV clinical application towards biomarker investigation from their metabolites.

Exosome detection via the ultrafast-isolation system: EXODUS.

Exosomes have shown great potential in disease diagnostics and therapeutics. However, current isolation approaches are burdensome and suffer from low speed, yield and purity, limiting basic research and clinical applications. Here, we describe an efficient exosome detection method via the ultrafast-isolation system (EXODUS) that allows automated label-free purification of exosomes from varied biofluids. We obtained the ultra-efficient purification of exosomes by negative pressure oscillation and double coupled harmonic oscillator-enabled membrane vibration. Our two coupled oscillators generate dual-frequency transverse waves on the membranes, enabling EXODUS to outperform other isolation techniques in speed, purity and yield. We demonstrated EXODUS by purifying exosomes from urine samples of 113 patients and validated the practical relevance in exosomal RNA profiling with the high-resolution capability and high-throughput analysis.

Deep dive on the proteome of salivary extracellular vesicles: comparison between ultracentrifugation and polymer-based precipitation isolation.

Salivary extracellular vesicles (EVs), as novel functional carriers and potential biomarkers, are usually obtained by ultracentrifugation (UC) and polyethylene glycol (PEG)-based precipitation methods. However, salivary EVs obtained by these two methods have not been systematically compared. Here, we perform an in-depth analysis on EVs isolated by these two methods using proteomics. Both methods obtain EVs ranging from 40 to 210 nm, with the PEG method resulting in a wider size distribution. PEG-separated products were irregularly shaped and aggregated, while UC-separated ones were monodispersed and teacup-shaped. Additionally, the expression of EV-specific markers was higher in UC-separated EVs. Using tandem mass spectrometry proteomics, we identified and quantified 1217 kinds of saliva exosomal proteins and 361 kinds of differential proteins, showing that UC can isolate more EV-related proteins. These results offer some guidance for EV separating and provide potential direction for the use of EVs in non-invasive diagnosis.

An integrative microfluidic device for isolation and ultrasensitive detection of lung cancer-specific exosomes from patient urine.

Exosomes, derived from various biofluids, may serve as potential biomarkers for cancer early detection. Nevertheless, exosome clinical translation remains challenging due to the lack of reliable isolation and detection methods. Herein, we present a novel integrated microfluidic device specifically designed for isolation and in-situ detection of lung cancer-specific exosomes collected from patient's urine. The new device has been fabricated using polymethyl methacrylate (PMMA) and a nanoporous gold (Au) nanocluster membrane modified with the capture antibody. The second antibody-conjugated Au nanorod probe was then loaded to identify and quantify lung cancer-specific exosomes using a dark field microscope. AuNC-Exosome-AuR complex produces a significant scattering wavelength shift and an improved scattering intensity due to resonance Rayleigh scattering, which enables the ultrasensitive detection of exosomes with a LOD below 1000 particles/mL. The proteomic analysis revealed that the high-purity exosomes has been isolated by the device. We then validated this method with 500 µL urine samples from lung cancer patients and controls, which showed great promise for differentiating early-stage lung cancer patients from healthy individuals. Taken together, the presented method is fast and ultrasensitive and can be easily adapted for the isolation and detection of cancer specific exosomes from other malignant tumors.

High-Efficiency Separation of Extracellular Vesicles from Lipoproteins in Plasma by Agarose Gel Electrophoresis.

Isolation and purification of extracellular vesicles (EVs) from plasma is essential to understand the EV circulation mechanism and discover biomarkers for the early detection of diseases. However, the size range of lipoprotein particles such as high density lipoprotein (HDL), low density lipoprotein (LDL), and very low density lipoprotein (VLDL) overlap that of EVs, making it difficult to remove lipoproteins from EVs. Here, we propose a method for the high efficiency separation of EVs in plasma using agarose gel electrophoresis based on their differences in size and zeta potential properties. Electrophoresis track assays revealed that EVs propagate more slowly than HDL but more quickly than LDL and VLDL in 1% agarose gel with pH 7.4 Tris-Acetate-EDTA (TAE) buffer. The size and morphology of the electrophoresis-recovered products were characterized to be consistent with typical EVs. In addition, the biological function of recovered EVs was investigated with cell uptake tests. The feasibility of this method was further verified with human plasma samples. In summary, this technique has the potential to become a convenient and efficient approach for high-purity EV separation.

Antibody-Oriented Strategy and Mechanism for the Preparation of Fluorescent Nanoprobes for Fast and Sensitive Immunodetection.

Nanoprobes have been widely used in biomedical engineering. However, antibodies are generally conjugated onto nanoparticles disorderly, which reduces their antigen recognition ability. The existing antibody orientation approaches are usually complex. Here, we developed and demonstrated a simple antibody-oriented strategy for the lateral flow immunoassay of cardiac troponin I by conjugating antibodies onto polystyrene nanospheres at the optimal pH. The binding amount and orientation of antibodies as well as the detection sensitivity were significantly improved. Although pH regulation is commonly used to optimize antibody conjugation, this paper illustrates the mechanism of its antibody orientation enhancement ability for the first time and reveals the important influences of the density, the charge distribution and hydrophilicity of the antibody, the control of the velocities of physical adsorption and chemical coupling, and other factors on antibody orientation. It is of great significance to understand and regulate antibody conjugation on the surface of micro- or nanospheres to construct high-performance probes for in vitro diagnosis applications.

A Rapid Test Strip for Diagnosing Glycosylated Hemoglobin (HbA1c) Based on Fluorescent Affinity Immunochromatography.

In this research, we developed a rapid and easy-to-operate point-of-care testing (POCT) strip based on fluorescent affinity immunochromatography to quantitatively determine HbA1c concentrations in whole blood. This assay, based on a sandwich method performed on test strips, effectively utilized the principle of an affinity chromatography column, which was commonly used in the detection of HbA1c, and the technology of traditional fluorescence immunochromatographic test strips (FICTS) were combined. In our test strips, the test line of traditional FICTS was transformed into the region of affinity chromatography, while improving the linearity and reducing the interference of the precursor of HbA1c and hemoglobin variants. The test strips could quantitatively detect HbA1c over a wide range (3 - 13.8%) with excellent linearity (R2 > 0.99), and the assay accuracy was demonstrated by comparing with high-performance liquid chromatography (HPLC) (R2 > 0.95). The simple, rapid, effective and quantitative strips will provide a novel method for the detection of HbA1c in clinical diagnosis.

Fluorescent Nanoprobes with Oriented Modified Antibodies to Improve Lateral Flow Immunoassay of Cardiac Troponin I.

Performance of nanoprobes can often determine the detection level of Lateral immunochromatography. Traditional probes were limited by the quantity and orientation of antibodies, immune activity of the Fab region or binding strength between protein and substrate. This study developed a new efficient and robust technology to construct fluorescent nanoprobes with oriented modified antibodies, based on specific binding of the Fc region of antibody with streptococcal protein G (SPG) on the surface of polystyrene microspheres (MS) and subsequent covalent cross-linking at binding sites to firm them. Lateral flow immunoassay using these probes was applied for the detection of cardiac troponin I (cTnI). The significantly improved detection sensitivity demonstrated that antibody orientation on MS surfaces effectively enhanced immunological activities of probes compared with random immobilizing methods. Furthermore, performance evaluation results of lateral flow test strips met clinical requirements perfectly, including limit of detection (0.032 ng/mL), linearity ( R > 0.99), repeatability (CV < 10%), correlation ( R > 0.99), and heat aging stability. This research also employed heterophilic blocking reagent (HBR) to actively block redundant binding sites of SPG for the first time in order to eliminate false positive interferences, improving the sensitivity and precision of test results further.

Peroxidase-Like Activity of Gold Nanoparticles and Their Gold Staining Enhanced ELISA Application.

We found that the peroxidase-like activity of gold nanoparticles (GNPs) followed the Michaelis-Menten kinetic model and was dependent on environmental pH and temperature, which was very similar to natural Horseradish Peroxidase (HRP). However, unlike HRP, which needs a lower H2O2 concentration with a very narrow range to reach a maximum reaction rate and avoid enzyme poisoning, GNPs have very high activity, even at an H2O2 concentration two orders of magnitude higher than HRP. It was demonstrated that H2O2 treatment could enhance the peroxidase-like activity of GNPs, resulting thus in the activity increase in a circular catalytic reaction by the reduplicative use of GNPs. It was also found that the peroxidase-like activity of GNPs responded sensitively to nanoparticle size and surface modifications. When used in an immunoassay, GNPs were generally conjugated with antibody and blocked with hydrophilic macromolecules to construct a nanoprobe. This strongly reduced the peroxidase-like activity and detection sensitivity of GNPs, therefore, restricting their use as peroxidase mimetics. We presented a novel strategy that combined the nanoprobes with gold staining to expose fresh catalytic gold surfaces and obtained a great increase in detection sensitivity.

Enzyme catalysis enhanced dark-field imaging as a novel immunohistochemical method.

Conventional immunohistochemistry is limited to subjective judgment based on human experience and thus it is clinically required to develop a quantitative immunohistochemical detection. 3,3'-Diaminobenzidin (DAB) aggregates, a type of staining product formed by conventional immunohistochemistry, were found to have a special optical property of dark-field imaging for the first time, and the mechanism was explored. On this basis, a novel immunohistochemical method based on dark-field imaging for detecting HER2 overexpressed in breast cancer was established, and the quantitative analysis standard and relevant software for measuring the scattering intensity was developed. In order to achieve a more sensitive detection, the HRP (horseradish peroxidase)-labeled secondary antibodies conjugated gold nanoparticles were constructed as nanoprobes to load more HRP enzymes, resulting in an enhanced DAB deposition as a dark-field label. Simultaneously, gold nanoparticles also act as a synergistically enhanced agent due to their mimicry of enzyme catalysis and dark-field scattering properties.

Rituximab-Au nanoprobes for simultaneous dark-field imaging and DAB staining of CD20 over-expressed on Raji cells.

A novel dual-modal cell immunodetection method based on both dark-field imaging and catalysis functions of gold nanoparticles has been established, where the Rituximab-Au conjugates were used as nanoprobes to label and image specifically the CD20 overexpressed on the surface of malignant lymphoma cells of Raji with high affinity.