Extremely potent human monoclonal antibodies from COVID-19 convalescent patients.

Human monoclonal antibodies are safe, preventive, and therapeutic tools that can be rapidly developed to help restore the massive health and economic disruption caused by the coronavirus disease 2019 (COVID-19) pandemic. By single-cell sorting 4,277 SARS-CoV-2 spike protein-specific memory B cells from 14 COVID-19 survivors, 453 neutralizing antibodies were identified. The most potent neutralizing antibodies recognized the spike protein receptor-binding domain, followed in potency by antibodies that recognize the S1 domain, the spike protein trimer, and the S2 subunit. Only 1.4% of them neutralized the authentic virus with a potency of 1-10 ng/mL. The most potent monoclonal antibody, engineered to reduce the risk of antibody-dependent enhancement and prolong half-life, neutralized the authentic wild-type virus and emerging variants containing D614G, E484K, and N501Y substitutions. Prophylactic and therapeutic efficacy in the hamster model was observed at 0.25 and 4 mg/kg respectively in absence of Fc functions.

Antibody design using deep learning: from sequence and structure design to affinity maturation.

Deep learning has achieved impressive results in various fields such as computer vision and natural language processing, making it a powerful tool in biology. Its applications now encompass cellular image classification, genomic studies and drug discovery. While drug development traditionally focused deep learning applications on small molecules, recent innovations have incorporated it in the discovery and development of biological molecules, particularly antibodies. Researchers have devised novel techniques to streamline antibody development, combining in vitro and in silico methods. In particular, computational power expedites lead candidate generation, scaling and potential antibody development against complex antigens. This survey highlights significant advancements in protein design and optimization, specifically focusing on antibodies. This includes various aspects such as design, folding, antibody-antigen interactions docking and affinity maturation.

Wall Street QE vs. Main Street Lending.

Monetary and fiscal authorities reacted swiftly to the COVID-19 pandemic by purchasing assets (or "Wall Street QE") and lending directly to non-financial firms (or "Main Street Lending"). Our paper develops a new framework to compare and contrast these different policies. For the Great Recession, characterized by impaired balance sheets of financial intermediaries, Main Street Lending and Wall Street QE are perfect substitutes and both stimulate aggregate demand. In contrast, for the COVID-19 recession, where non-financial firms faced significant cash flow shortages, Wall Street QE is almost completely ineffective, whereas Main Street Lending can be highly stimulative.

Exploiting generative self-supervised learning for the assessment of biological images with lack of annotations.

MOTIVATION: Computer-aided analysis of biological images typically requires extensive training on large-scale annotated datasets, which is not viable in many situations. In this paper, we present Generative Adversarial Network Discriminator Learner (GAN-DL), a novel self-supervised learning paradigm based on the StyleGAN2 architecture, which we employ for self-supervised image representation learning in the case of fluorescent biological images. RESULTS: We show that Wasserstein Generative Adversarial Networks enable high-throughput compound screening based on raw images. We demonstrate this by classifying active and inactive compounds tested for the inhibition of SARS-CoV-2 infection in two different cell models: the primary human renal cortical epithelial cells (HRCE) and the African green monkey kidney epithelial cells (VERO). In contrast to previous methods, our deep learning-based approach does not require any annotation, and can also be used to solve subtle tasks it was not specifically trained on, in a self-supervised manner. For example, it can effectively derive a dose-response curve for the tested treatments. AVAILABILITY AND IMPLEMENTATION: Our code and embeddings are available at https://gitlab.com/AlesioRFM/gan-dl StyleGAN2 is available at https://github.com/NVlabs/stylegan2 .

Deep-learning image analysis for high-throughput screening of opsono-phagocytosis-promoting monoclonal antibodies against Neisseria gonorrhoeae.

Antimicrobial resistance (AMR) is nowadays a global health concern as bacterial pathogens are increasingly developing resistance to antibiotics. Monoclonal antibodies (mAbs) represent a powerful tool for addressing AMR thanks to their high specificity for pathogenic bacteria which allows sparing the microbiota, kill bacteria through complement deposition, enhance phagocytosis or inhibit bacterial adhesion to epithelial cells. Here we describe a visual opsono-phagocytosis assay which relies on confocal microscopy to measure the impact of mAbs on phagocytosis of the bacterium Neisseria gonorrhoeae by macrophages. With respect to traditional CFU-based assays, generated images can be automatically analysed by convolutional neural networks. Our results demonstrate that confocal microscopy and deep learning-based analysis allow screening for phagocytosis-promoting mAbs against N. gonorrhoeae, even when mAbs are not purified and are expressed at low concentration. Ultimately, the flexibility of the staining protocol and of the deep-learning approach make the assay suitable for other bacterial species and cell lines where mAb activity needs to be investigated.

Development of a visual Adhesion/Invasion Inhibition Assay to assess the functionality of Shigella-specific antibodies.

Introduction: Shigella is the etiologic agent of a bacillary dysentery known as shigellosis, which causes millions of infections and thousands of deaths worldwide each year due to Shigella's unique lifestyle within intestinal epithelial cells. Cell adhesion/invasion assays have been extensively used not only to identify targets mediating host-pathogen interaction, but also to evaluate the ability of Shigella-specific antibodies to reduce virulence. However, these assays are time-consuming and labor-intensive and fail to assess differences at the single-cell level. Objectives and methods: Here, we developed a simple, fast and high-content method named visual Adhesion/Invasion Inhibition Assay (vAIA) to measure the ability of anti-Shigellaantibodies to inhibit bacterial adhesion to and invasion of epithelial cells by using the confocal microscope Opera Phenix. Results: We showed that vAIA performed well with a pooled human serum from subjects challenged with S. sonnei and that a specific anti-IpaD monoclonal antibody effectively reduced bacterial virulence in a dose-dependent manner. Discussion: vAIA can therefore inform on the functionality of polyclonal and monoclonal responses thereby supporting the discovery of pathogenicity mechanisms and the development of candidate vaccines and immunotherapies. Lastly, this assay is very versatile and may be easily applied to other Shigella species or serotypes and to different pathogens.

Systems analysis of human responses to an aluminium hydroxide-adsorbed TLR7 agonist (AS37) adjuvanted vaccine reveals a dose-dependent and specific activation of the interferon-mediated antiviral response.

The candidate Adjuvant System AS37 contains a synthetic toll-like receptor agonist (TLR7a) adsorbed to alum. In a phase I study (NCT02639351), healthy adults were randomised to receive one dose of licensed alum-adjuvanted meningococcal serogroup C (MenC-CRM197) conjugate vaccine (control) or MenC-CRM197 conjugate vaccine adjuvanted with AS37 (TLR7a dose 12.5, 25, 50 or 100 µg). A subset of 66 participants consented to characterisation of peripheral whole blood transcriptomic responses, systemic cytokine/chemokine responses and multiple myeloid and lymphoid cell responses as exploratory study endpoints. Blood samples were collected pre-vaccination, 6 and 24 h post-vaccination, and 3, 7, 28 and 180 days post-vaccination. The gene expression profile in whole blood showed an early, AS37-specific transcriptome response that peaked at 24 h, increased with TLR7a dose up to 50 µg and generally resolved within one week. Five clusters of differentially expressed genes were identified, including those involved in the interferon-mediated antiviral response. Evaluation of 30 cytokines/chemokines by multiplex assay showed an increased level of interferon-induced chemokine CXCL10 (IP-10) at 24 h and 3 days post-vaccination in the AS37-adjuvanted vaccine groups. Increases in activated plasmacytoid dendritic cells (pDC) and intermediate monocytes were detected 3 days post-vaccination in the AS37-adjuvanted vaccine groups. T follicular helper (Tfh) cells increased 7 days post-vaccination and were maintained at 28 days post-vaccination, particularly in the AS37-adjuvanted vaccine groups. Moreover, most of the subjects that received vaccine containing 25, 50 and 100 µg TLR7a showed an increased MenC-specific memory B cell responses versus baseline. These data show that the adsorption of TLR7a to alum promotes an immune signature consistent with TLR7 engagement, with up-regulation of interferon-inducible genes, cytokines and frequency of activated pDC, intermediate monocytes, MenC-specific memory B cells and Tfh cells. TLR7a 25-50 µg can be considered the optimal dose for AS37, particularly for the adjuvanted MenC-CRM197 conjugate vaccine.