Developing a GNN-based AI model to predict mitochondrial toxicity using the bagging method.

Mitochondrial toxicity has been implicated in the development of various toxicities, including hepatotoxicity. Therefore, mitochondrial toxicity has become a major screening factor in the early discovery phase of drug development. Several models have been developed to predict mitochondrial toxicity based on chemical structures. However, they only provide a binary classification of positive or negative results and do not provide the substructures that contribute to a positive decision. Therefore, we developed an artificial intelligence (AI) model to predict mitochondrial toxicity and visualize structural alerts. To construct the model, we used the open-source software library kMoL, which employs a graph neural network approach that allows learning from chemical structure data. We also utilized the integrated gradient method, which enables the visualization of substructures that contribute to positive results. The dataset used to construct the AI model exhibited a significant imbalance, with significantly more negative than positive data. To address this, we employed the bagging method, which resulted in a model with high predictive performance, as evidenced by an F1 score of 0.839. This model can also be used to visualize substructures that contribute to mitochondrial toxicity using the integrated gradient method. Our AI model predicts mitochondrial toxicity based on chemical structures and may contribute to screening mitochondrial toxicity in the early stages of drug discovery.

Inducing Melanoma Cell Apoptosis by ERp57/PDIA3 Antibody in the Presence of CPI-613 and Hydroxychloroquine.

The combination of the cancer mitochondrial metabolic inhibitor CPI-613 and hydroxychloroquine has tumor-suppressive effects on clear cell sarcoma, which shares pathobiological properties with melanoma. Therefore, we intended to examine the effects of a combination of CPI-613 and hydroxychloroquine on the growth of melanoma cells in the present study. However, cell death was not induced in melanoma cells. Therefore, a monoclonal antibody, ICT, that induced apoptosis in melanoma cells in combination with CPI-613 and hydroxychloroquine was developed. Immunoprecipitation, mass spectrometry, and small interfering RNA (siRNA)-mediated gene silencing demonstrated that ICT targeted Endoplasmic Reticulum Resident Protein 57/ Protein Disulfide Isomerase Family A Member 3 (ERp57/PDIA3), which was first identified as being upregulated by metabolic depletion stress and is localized on the cell surface during immunogenic cell death. The combination of CPI-613 and hydroxychloroquine enhanced the localization of ERp57/PDIA3 to the surface of melanoma cells. siRNA-mediated downregulation of ERp57/PDIA3 did not significantly induce ICT-mediated apoptosis in melanoma cells in the presence of CPI-613 and hydroxychloroquine. Therefore, the ICT antibody acts as a tumor suppressor in melanoma cells by targeting the cell membrane ERp57/PDIA3, expression of which was enhanced by the combination of CPI-613 and hydroxychloroquine.

Low pH condition impairs BP-IgG binding to the basement membrane zone.

Bullous pemphigoid (BP), an autoimmune subepidermal blistering disease, shows tense blisters associated with urticarial erythema. Tissue-bound Immunoglobulin G (IgG) at the basement membrane zone (BMZ) detected by direct immunofluorescence (DIF) is strong evidence for a diagnosis of BP. The sensitivity of DIF is higher in complement component 3 (C3) than in IgG, but the reason for this different sensitivity is not fully understood. In this study, we performed several ex vivo studies to investigate the possible mechanism of IgG negativity and C3 positivity at the BMZ by DIF in some BP cases. First, sera from BP patients showing IgG negativity by DIF were found to clearly react to the BMZ in their own DIF skin samples. Next, indirect immunofluorescence (IIF) was performed using sera diluted with different pH phosphate-buffered saline (PBS), pH 7.4, 6.0, and 3.0. Patients' sera diluted with pH 7.4 PBS showed linear staining at the BMZ, but sera diluted with pH 6.0 PBS and pH 3.0 PBS showed lower fluorescence intensities. Finally, sections of skin from BP patients were pre-incubated with different pH PBS (pH 3.0, 6.0, and 7.4), followed by staining with anti-human IgG and C3. The fluorescence intensities were notably lower for IgG and C3 that had been pre-incubated with pH 3.0 PBS and pH 6.0 PBS than for IgG and C3 that had been pre-incubated with pH 7.4 PBS. These results suggest that a low pH condition hinders the binding of autoantibodies to the BMZ, that is, the drop in tissue pH induced by inflammation inhibits autoantibodies from depositing at the BMZ. Furthermore, the drop in tissue pH causes tissue-bound autoantibodies to detach from the BMZ. Complement fragments are activated not only on IgG but also on the cell surface of cells close to IgG during complement activation. IgG may detach from the BMZ under low pH condition induced by inflammation, but some complement fragments remain at the BMZ. These phenomena may help to explain why C3 is more sensitive than IgG when DIF is used to diagnose BP.

Feature extraction of particle morphologies of pharmaceutical excipients from scanning electron microscope images using convolutional neural networks.

Scanning electron microscopy (SEM) images are the most widely used tool for evaluating particle morphology; however, quantitative evaluation using SEM images is time-consuming and often neglected. In this study, we aimed to extract features related to particle morphology of pharmaceutical excipients from SEM images using a convolutional neural network (CNN). SEM images of 67 excipients were acquired and used as models. A classification CNN model of the excipients was constructed based on the SEM images. Further, features were extracted from the middle layer of this CNN model, and the data was compressed to two dimensions using uniform manifold approximation and projection. Lastly, hierarchical clustering analysis (HCA) was performed to categorize the excipients into several clusters and identify similarities among the samples. The classification CNN model showed high accuracy, allowing each excipient to be identified with a high degree of accuracy. HCA revealed that the 67 excipients were classified into seven clusters. Additionally, the particle morphologies of excipients belonging to the same cluster were found to be very similar. These results suggest that CNN models are useful tools for extracting information and identifying similarities among the particle morphologies of excipients.

Diacron-Reactive Oxygen Metabolites Levels Are Initially Elevated in Patients with Bullous Pemphigoid.

ROS are involved in the pathogenesis of bullous pemphigoid (BP), but this involvement has not been fully elucidated. In this study, to further elucidate the pathogenic role of ROS in BP, we examined the results of the diacron-reactive oxygen metabolite test and the biological antioxidant potential test for 16 patients with BP who visited our hospital before being treated with systemic corticosteroids. In the patients with BP, the average diacron-reactive oxygen metabolite levels, expressed in Carratelli units, were significantly reduced at 1 month of treatment (from 335.6 ± 40.3 Carratelli units to 224.7 ± 61.6 Carratelli units, P < .001). Bullous Pemphigoid Disease Area Index (erosions/blisters) scores correlated with diacron-reactive oxygen metabolite levels (r = 0.51), suggesting that those levels reflect the disease severity. We also performed staining of 3,5-dibromotyrosine in skin tissues. The 3,5-dibromotyrosine is expected to be a marker of tissue damage related to inflammation and allergies. The 3,5-dibromotyrosine was stained in infiltrated cells around the dermis, throughout the blister fluid, and at the basement membrane within the blister. It is considered that tissue destruction caused by the myeloperoxidase released from neutrophils and by eosinophil peroxidase released from eosinophils is involved in blister formation. The results suggest that ROS play a role in BP.