Differentiation of Cytopathic Effects (CPE) induced by influenza virus infection using deep Convolutional Neural Networks (CNN)

Therapeutic potential of mesenchymal stem cell-derived extracellular vesicles in SARS-CoV-2 and H1N1 influenza-induced acute lung injury

Mesenchymal stem cell (MSC)-derived extracellular vesicles (EVs) have shown anti-inflammatory potential in multiple inflammatory diseases. In the March 2022 issue of the Journal of Extracellular Vesicles, it was shown that EVs from human MSCs can suppress severe acute respiratory distress syndrome, coronavirus 2 (SARS-CoV-2) replication and can mitigate the production and release of infectious virions. We therefore hypothesized that MSC-EVs have an anti-viral effect in SARS-CoV-2 infection in vivo. We extended this question to ask whether also other respiratory viral infections could be treated by MSC-EVs. Adipose stem cell-derived EVs (ASC-EVs) were isolated using tangential flow filtration from conditioned media obtained from a multi-flask cell culture system. The effects of the ASC-EVs were tested  in Vero E6 cells in vitro. ASC-EVs were also given i.v. to SARS-CoV-2 infected Syrian Hamsters, and H1N1 influenza virus infected mice. The ASC-EVs attenuated SARS-CoV-2 virus replication in Vero E6 cells and reduced body weight and signs of lung injury in infected Syrian hamsters. Furthermore, ASC-EVs increased the survival rate of influenza A-infected mice and attenuated signs of lung injury. In summary, this study suggests that ASC-EVs can have beneficial therapeutic effects in models of virus-infection-associated acute lung injury and may potentially be developed to treat lung injury in humans.

High-Throughput Determination of Infectious Virus Titers by Kinetic Measurement of Infection-Induced Changes in Cell Morphology

Infectivity assays are the key analytical technology for the development and manufacturing of virus-based therapeutics. Here, we introduce a novel assay format that utilizes label-free bright-field images to determine the kinetics of infection-dependent changes in cell morphology. In particular, cell rounding is directly proportional to the amount of infectious virus applied, enabling rapid determination of viral titers in relation to a standard curve. Our kinetic infectious virus titer (KIT) assay is stability-indicating and, due to its sensitive readout method, provides results within 24 h post-infection. Compared to traditional infectivity assays, which depend on a single readout of an infection endpoint, cumulated analysis of kinetic data by a fit model results in precise results (CV < 20%) based on only three wells per sample. This approach allows for a high throughput with ~400 samples processed by a single operator per week. We demonstrate the applicability of the KIT assay for the genetically engineered oncolytic VSV-GP, Newcastle disease virus (NDV), and parapoxvirus ovis (ORFV), but it can potentially be extended to a wide range of viruses that induce morphological changes upon infection. The versatility of this assay, combined with its independence from specific instruments or software, makes it a promising solution to overcome the analytical bottleneck in infectivity assays within the pharmaceutical industry and as a routine method in academic research.

A versatile automated pipeline for quantifying virus infectivity by label-free light microscopy and artificial intelligence

Virus infectivity is traditionally determined by endpoint titration in cell cultures, and requires complex processing steps and human annotation. Here we developed an artificial intelligence (AI)-powered automated framework for ready detection of virus-induced cytopathic effect (DVICE). DVICE uses the convolutional neural network EfficientNet-B0 and transmitted light microscopy images of infected cell cultures, including coronavirus, influenza virus, rhinovirus, herpes simplex virus, vaccinia virus, and adenovirus. DVICE robustly measures virus-induced cytopathic effects (CPE), as shown by class activation mapping. Leave-one-out cross-validation in different cell types demonstrates high accuracy for different viruses, including SARS-CoV-2 in human saliva. Strikingly, DVICE exhibits virus class specificity, as shown with adenovirus, herpesvirus, rhinovirus, vaccinia virus, and SARS-CoV-2. In sum, DVICE provides unbiased infectivity scores of infectious agents causing CPE, and can be adapted to laboratory diagnostics, drug screening, serum neutralization or clinical samples.

COVID-19 on Oral Health: A New Bilateral Connection for the Pandemic

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and transmission are generally known to be produced by respiratory droplets and aerosols from the oral cavity (O.C.) of infected subjects, as stated by the World Health Organization. Saliva also retains the viral particles and aids in the spread of COVID-19. Angiotensin-converting enzyme Type 2 (ACE2) and transmembrane serine protease 2 (TMPRSS2) are two of the numerous factors that promote SARS-CoV-2 infection, expressed by O.C. structures, various mucosa types, and the epithelia of salivary glands. A systemic SARS-CoV-2 infection might result from viral replication in O.C. cells. On the other hand, cellular damage of different subtypes in the O.C. might be associated with various clinical signs and symptoms. Factors interfering with SARS-CoV-2 infection potential might represent fertile ground for possible local pharmacotherapeutic interventions, which may confine SARS-CoV-2 virus entry and transmission in the O.C., finally representing a way to reduce COVID-19 incidence and severity.

A rapid, high-throughput, viral infectivity assay using automated brightfield microscopy with machine learning

Infectivity assays are essential for the development of viral vaccines, antiviral therapies, and the manufacture of biologicals. Traditionally, these assays take 2-7 days and require several manual processing steps after infection. We describe an automated viral infectivity assay (AVIATM), using convolutional neural networks (CNNs) and high-throughput brightfield microscopy on 96-well plates that can quantify infection phenotypes within hours, before they are manually visible, and without sample preparation. CNN models were trained on HIV, influenza A virus, coronavirus 229E, vaccinia viruses, poliovirus, and adenoviruses, which together span the four major categories of virus (DNA, RNA, enveloped, and non-enveloped). A sigmoidal function, fit between virus dilution curves and CNN predictions, results in sensitivity ranges comparable to or better than conventional plaque or TCID50 assays, and a precision of ∼10%, which is considerably better than conventional infectivity assays. Because this technology is based on sensitizing CNNs to specific phenotypes of infection, it has potential as a rapid, broad-spectrum tool for virus characterization, and potentially identification.

Characterizing the Pathogenicity and Immunogenicity of Simian Retrovirus Subtype 8 (SRV-8) Using SRV-8-Infected Cynomolgus Monkeys

Simian retrovirus subtype 8 (SRV-8) infections have been reported in cynomolgus monkeys (Macaca fascicularis) in China and America, but its pathogenicity and immunogenicity are rarely reported. In this work, the SRV-8-infected monkeys were identified from the monkeys with anemia, weight loss, and diarrhea. To clarify the impact of SRV-8 infection on cynomolgus monkeys, infected monkeys were divided into five groups according to disease progression. Hematoxylin (HE) staining and viral loads analysis showed that SRV-8 mainly persisted in the intestine and spleen, causing tissue damage. Additionally, the dynamic variations of blood routine indexes, innate and adaptive immunity, and the transcriptomic changes in peripheral blood cells were analyzed during SRV-8 infection. Compared to uninfected animals, red blood cells, hemoglobin, and white blood cells were reduced in SRV-8-infected monkeys. The percentage of immune cell populations was changed after SRV-8 infection. Furthermore, the number of hematopoietic stem cells decreased significantly during the early stages of SRV-8 infection, and returned to normal levels after antibody-mediated viral clearance. Finally, global transcriptomic analysis in PBMCs from SRV-8-infected monkeys revealed distinct gene expression profiles across different disease stages. In summary, SRV-8 infection can cause severe pathogenicity and immune disturbance in cynomolgus monkeys, and it might be responsible for fatal virus-associated immunosuppressive syndrome.

The ratio of nicotinic acid to nicotinamide as a microbial biomarker for assessing cell therapy product sterility

Controlling microbial risks in cell therapy products (CTPs) is important for product safety. Here, we identified the nicotinic acid (NA) to nicotinamide (NAM) ratio as a biomarker that detects a broad spectrum of microbial contaminants in cell cultures. We separately added six different bacterial species into mesenchymal stromal cell and T cell culture and found that NA was uniquely present in these bacteria-contaminated CTPs due to the conversion from NAM by microbial nicotinamidases, which mammals lack. In cells inoculated with 1 × 10⁴ CFUs/mL of different microorganisms, including USP <71> defined organisms, the increase in NA to NAM ratio ranged from 72 to 15,000 times higher than the uncontaminated controls after 24 h. Importantly, only live microorganisms caused increases in this ratio. In cells inoculated with 18 CFUs/mL of Escherichia coli, 20 CFUs/mL of Bacillus subtilis, and 10 CFUs/mL of Candida albicans, significant increase of NA to NAM ratio was detected using LC-MS after 18.5, 12.5, and 24.5 h, respectively. In contrast, compendial sterility test required >24 h to detect the same amount of these three organisms. In conclusion, the NA to NAM ratio is a useful biomarker for detection of early-stage microbial contaminations in CTPs.

Detection of Cytopathic Effects Induced by Influenza, Parainfluenza, and Enterovirus Using Deep Convolution Neural Network

The isolation of a virus using cell culture to observe its cytopathic effects (CPEs) is the main method for identifying the viruses in clinical specimens. However, the observation of CPEs requires experienced inspectors and excessive time to inspect the cell morphology changes. In this study, we utilized artificial intelligence (AI) to improve the efficiency of virus identification. After some comparisons, we used ResNet-50 as a backbone with single and multi-task learning models to perform deep learning on the CPEs induced by influenza, enterovirus, and parainfluenza. The accuracies of the single and multi-task learning models were 97.78% and 98.25%, respectively. In addition, the multi-task learning model increased the accuracy of the single model from 95.79% to 97.13% when only a few data of the CPEs induced by parainfluenza were provided. We modified both models by inserting a multiplexer and de-multiplexer layer, respectively, to increase the correct rates for known cell lines. In conclusion, we provide a deep learning structure with ResNet-50 and the multi-task learning model and show an excellent performance in identifying virus-induced CPEs.