Super resolution microscopy and deep learning identify Zika virus reorganization of the endoplasmic reticulum

AI analysis of super-resolution microscopy: Biological discovery in the absence of ground truth

Super-resolution microscopy, or nanoscopy, enables the use of fluorescent-based molecular localization tools to study molecular structure at the nanoscale level in the intact cell, bridging the mesoscale gap to classical structural biology methodologies. Analysis of super-resolution data by artificial intelligence (AI), such as machine learning, offers tremendous potential for the discovery of new biology, that, by definition, is not known and lacks ground truth. Herein, we describe the application of weakly supervised paradigms to super-resolution microscopy and its potential to enable the accelerated exploration of the nanoscale architecture of subcellular macromolecules and organelles.

Zika virus replication is impaired by a selective agonist of the TRPML2 ion channel

The flavivirus genus includes human pathogenic viruses such as Dengue (DENV), West Nile (WNV) and Zika virus (ZIKV) posing a global health threat due to limited treatment options. Host ion channels are crucial for various viral life cycle stages, but their potential as targets for antivirals is often not fully realized due to the lack of selective modulators. Here, we observe that treatment with ML2-SA1, an agonist for the human endolysosomal cation channel TRPML2, impairs ZIKV replication. Upon ML2-SA1 treatment, levels of intracellular genomes and number of released virus particles of two different ZIKV isolates were significantly reduced and cells displayed enlarged vesicular structures and multivesicular bodies with ZIKV envelope protein accumulation. However, no increased ZIKV degradation in lysosomal compartments was observed. Rather, the antiviral effect of ML2-SA1 seemed to manifest by the compound's negative impact on genome replication. Moreover, ML2-SA1 treatment also led to intracellular cholesterol accumulation. ZIKV and many other viruses including the Orthohepevirus Hepatitis E virus (HEV) rely on the endolysosomal system and are affected by intracellular cholesterol levels to complete their life cycle. Since we observed that ML2-SA1 also negatively impacted HEV infections in vitro, this compound may harbor a broader antiviral potential through perturbing the intracellular cholesterol distribution. Besides indicating that TRPML2 may be a promising target for combatting viral infections, we uncover a tentative connection between this protein and cholesterol distribution within the context of infectious diseases.

Single-cell analysis reveals an antiviral network that controls Zika virus infection in human dendritic cells

Zika virus (ZIKV) is a mosquito-borne flavivirus that caused an epidemic in the Americas in 2016 and is linked to severe neonatal birth defects, including microcephaly and spontaneous abortion. To better understand the host response to ZIKV infection, we adapted the 10× Genomics Chromium single-cell RNA sequencing (scRNA-seq) assay to simultaneously capture viral RNA and host mRNA. Using this assay, we profiled the antiviral landscape in a population of human monocyte-derived dendritic cells infected with ZIKV at the single-cell level. The bystander cells, which lacked detectable viral RNA, expressed an antiviral state that was enriched for genes coinciding predominantly with a type I interferon (IFN) response. Within the infected cells, viral RNA negatively correlated with type I IFN-dependent and -independent genes (the antiviral module). We modeled the ZIKV-specific antiviral state at the protein level, leveraging experimentally derived protein interaction data. We identified a highly interconnected network between the antiviral module and other host proteins. In this work, we propose a new paradigm for evaluating the antiviral response to a specific virus, combining an unbiased list of genes that highly correlate with viral RNA on a per-cell basis with experimental protein interaction data.

IMPORTANCE

Zika virus (ZIKV) remains a public health threat given its potential for re-emergence and the detrimental fetal outcomes associated with infection during pregnancy. Understanding the dynamics between ZIKV and its host is critical to understanding ZIKV pathogenesis. Through ZIKV-inclusive single-cell RNA sequencing (scRNA-seq), we demonstrate on the single-cell level the dynamic interplay between ZIKV and the host: the transcriptional program that restricts viral infection and ZIKV-mediated inhibition of that response. Our ZIKV-inclusive scRNA-seq assay will serve as a useful tool for gaining greater insight into the host response to ZIKV and can be applied more broadly to the flavivirus field.

Nanomolar anti-SARS-CoV-2 Omicron activity of the host-directed TMPRSS2 inhibitor N-0385 and synergistic action with direct-acting antivirals

SARS-CoV-2 Omicron subvariants with increased transmissibility and immune evasion are spreading globally with alarming persistence. Whether the mutations and evolution of spike (S) Omicron subvariants alter the viral hijacking of human TMPRSS2 for viral entry remains to be elucidated. This is particularly important to investigate because of the large number and diversity of mutations of S Omicron subvariants reported since the emergence of BA.1. Here we report that human TMPRSS2 is a molecular determinant of viral entry for all the Omicron clinical isolates tested in human lung cells, including ancestral Omicron subvariants (BA.1, BA.2, BA.5), contemporary Omicron subvariants (BQ.1.1, XBB.1.5, EG.5.1) and currently circulating Omicron BA.2.86. First, we used a co-transfection assay to demonstrate the endoproteolytic cleavage by TMPRSS2 of spike Omicron subvariants. Second, we found that N-0385, a highly potent TMPRSS2 inhibitor, is a robust entry inhibitor of virus-like particles harbouring the S protein of Omicron subvariants. Third, we show that N-0385 exhibits nanomolar broad-spectrum antiviral activity against live Omicron subvariants in human Calu-3 lung cells and primary patient-derived bronchial epithelial cells. Interestingly, we found that N-0385 is 10-20 times more potent than the repositioned TMPRSS2 inhibitor, camostat, against BA.5, EG.5.1, and BA.2.86. We further found that N-0385 shows broad synergistic activity with clinically approved direct-acting antivirals (DAAs), i.e., remdesivir and nirmatrelvir, against Omicron subvariants, demonstrating the potential therapeutic benefits of a multi-targeted treatment based on N-0385 and DAAs.

Single cell analysis reveals an antiviral network that controls Zika virus infection in human dendritic cells

Zika virus (ZIKV) is a mosquito-borne flavivirus that caused an epidemic in the Americas in 2016 and is linked to severe neonatal birth defects, including microcephaly and spontaneous abortion. To better understand the host response to ZIKV infection, we adapted the 10x Genomics Chromium single cell RNA sequencing (scRNA-seq) assay to simultaneously capture viral RNA and host mRNA. Using this assay, we profiled the antiviral landscape in a population of human moDCs infected with ZIKV at the single cell level. The bystander cells, which lacked detectable viral RNA, expressed an antiviral state that was enriched for genes coinciding predominantly with a type I interferon (IFN) response. Within the infected cells, viral RNA negatively correlated with type I IFN dependent and independent genes (antiviral module). We modeled the ZIKV specific antiviral state at the protein level leveraging experimentally derived protein-interaction data. We identified a highly interconnected network between the antiviral module and other host proteins. In this work, we propose a new paradigm for evaluating the antiviral response to a specific virus, combining an unbiased list of genes that highly correlate with viral RNA on a per cell basis with experimental protein interaction data. Our ZIKV-inclusive scRNA-seq assay will serve as a useful tool to gaining greater insight into the host response to ZIKV and can be applied more broadly to the flavivirus field.

Membrane contact site detection (MCS-DETECT) reveals dual control of rough mitochondria-ER contacts

Identification and morphological analysis of mitochondria-ER contacts (MERCs) by fluorescent microscopy is limited by subpixel resolution interorganelle distances. Here, the membrane contact site (MCS) detection algorithm, MCS-DETECT, reconstructs subpixel resolution MERCs from 3D super-resolution image volumes. MCS-DETECT shows that elongated ribosome-studded riboMERCs, present in HT-1080 but not COS-7 cells, are morphologically distinct from smaller smooth contacts and larger contacts induced by mitochondria-ER linker expression in COS-7 cells. RiboMERC formation is associated with increased mitochondrial potential, reduced in Gp78 knockout HT-1080 cells and induced by Gp78 ubiquitin ligase activity in COS-7 and HeLa cells. Knockdown of riboMERC tether RRBP1 eliminates riboMERCs in both wild-type and Gp78 knockout HT-1080 cells. By MCS-DETECT, Gp78-dependent riboMERCs present complex tubular shapes that intercalate between and contact multiple mitochondria. MCS-DETECT of 3D whole-cell super-resolution image volumes, therefore, identifies novel dual control of tubular riboMERCs, whose formation is dependent on RRBP1 and size modulated by Gp78 E3 ubiquitin ligase activity.

Machine learning for cross-scale microscopy of viruses

Despite advances in virological sciences and antiviral research, viruses continue to emerge, circulate, and threaten public health. We still lack a comprehensive understanding of how cells and individuals remain susceptible to infectious agents. This deficiency is in part due to the complexity of viruses, including the cell states controlling virus-host interactions. Microscopy samples distinct cellular infection stages in a multi-parametric, time-resolved manner at molecular resolution and is increasingly enhanced by machine learning and deep learning. Here we discuss how state-of-the-art artificial intelligence (AI) augments light and electron microscopy and advances virological research of cells. We describe current procedures for image denoising, object segmentation, tracking, classification, and super-resolution and showcase examples of how AI has improved the acquisition and analyses of microscopy data. The power of AI-enhanced microscopy will continue to help unravel virus infection mechanisms, develop antiviral agents, and improve viral vectors.

Quantitative Single-Molecule Localization Microscopy

Super-resolution fluorescence microscopy allows the investigation of cellular structures at nanoscale resolution using light. Current developments in super-resolution microscopy have focused on reliable quantification of the underlying biological data. In this review, we first describe the basic principles of super-resolution microscopy techniques such as stimulated emission depletion (STED) microscopy and single-molecule localization microscopy (SMLM), and then give a broad overview of methodological developments to quantify super-resolution data, particularly those geared toward SMLM data. We cover commonly used techniques such as spatial point pattern analysis, colocalization, and protein copy number quantification but also describe more advanced techniques such as structural modeling, single-particle tracking, and biosensing. Finally, we provide an outlook on exciting new research directions to which quantitative super-resolution microscopy might be applied.

A novel fluorescent endoplasmic reticulum marker for super-resolution imaging in live cells

Endoplasmic reticulum (ER) is a highly complicated and dynamic organelle that actively changes its shape and communicates with other organelles. Visualization of ER in live cells is of great importance to understand cellular activities. Here, we designed a novel ER marker, RR-mNeonGreen, which comprised an N-terminal ER retention signal, a bright fluorescent protein (mNeonGreen), and a C-terminal transmembrane region. Colocalization of RR-mNeonGreen with mCherry-KDEL verified that RR-mNeonGreen perfectly labeled the ER. RR-mNeonGreen showed better continuity of ER tubules when imaged by super-resolution microscopy. Moreover, RR-mNeonGreen is competent for live-cell imaging of ER dynamics and tracing of the interaction between ER and mitochondria at high spatiotemporal resolution. In summary, RR-mNeonGreen is a novel ER marker for super-resolution live-cell imaging with multiple merits.

Generating post-hoc explanation from deep neural networks for multi-modal medical image analysis tasks

Explaining model decisions from medical image inputs is necessary for deploying deep neural network (DNN) based models as clinical decision assistants. The acquisition of multi-modal medical images is pervasive in practice for supporting the clinical decision-making process. Multi-modal images capture different aspects of the same underlying regions of interest. Explaining DNN decisions on multi-modal medical images is thus a clinically important problem. Our methods adopt commonly-used post-hoc artificial intelligence feature attribution methods to explain DNN decisions on multi-modal medical images, including two categories of gradient- and perturbation-based methods. • Gradient-based explanation methods - such as Guided BackProp, DeepLift - utilize the gradient signal to estimate the feature importance for model prediction. • Perturbation-based methods - such as occlusion, LIME, kernel SHAP - utilize the input-output sampling pairs to estimate the feature importance. • We describe the implementation details on how to make the methods work for multi-modal image input, and make the implementation code available.

Guidelines and evaluation of clinical explainable AI in medical image analysis

Explainable artificial intelligence (XAI) is essential for enabling clinical users to get informed decision support from AI and comply with evidence-based medical practice. Applying XAI in clinical settings requires proper evaluation criteria to ensure the explanation technique is both technically sound and clinically useful, but specific support is lacking to achieve this goal. To bridge the research gap, we propose the Clinical XAI Guidelines that consist of five criteria a clinical XAI needs to be optimized for. The guidelines recommend choosing an explanation form based on Guideline 1 (G1) Understandability and G2 Clinical relevance. For the chosen explanation form, its specific XAI technique should be optimized for G3 Truthfulness, G4 Informative plausibility, and G5 Computational efficiency. Following the guidelines, we conducted a systematic evaluation on a novel problem of multi-modal medical image explanation with two clinical tasks, and proposed new evaluation metrics accordingly. Sixteen commonly-used heatmap XAI techniques were evaluated and found to be insufficient for clinical use due to their failure in G3 and G4. Our evaluation demonstrated the use of Clinical XAI Guidelines to support the design and evaluation of clinically viable XAI.

Dynamics and Patterning of 5-Hydroxytryptamine 2 Subtype Receptors in JC Polyomavirus Entry

The organization and dynamics of plasma membrane receptors are a critical link in virus-receptor interactions, which finetune signaling efficiency and determine cellular responses during infection. Characterizing the mechanisms responsible for the active rearrangement and clustering of receptors may aid in developing novel strategies for the therapeutic treatment of viruses. Virus-receptor interactions are poorly understood at the nanoscale, yet they present an attractive target for the design of drugs and for the illumination of viral infection and pathogenesis. This study utilizes super-resolution microscopy and related techniques, which surpass traditional microscopy resolution limitations, to provide both a spatial and temporal assessment of the interactions of human JC polyomavirus (JCPyV) with 5-hydroxytrypamine 2 receptors (5-HT2Rs) subtypes during viral entry. JCPyV causes asymptomatic kidney infection in the majority of the population and can cause fatal brain disease, and progressive multifocal leukoencephalopathy (PML), in immunocompromised individuals. Using Fluorescence Photoactivation Localization Microscopy (FPALM), the colocalization of JCPyV with 5-HT2 receptor subtypes (5-HT2A, 5-HT2B, and 5-HT2C) during viral attachment and viral entry was analyzed. JCPyV was found to significantly enhance the clustering of 5-HT2 receptors during entry. Cluster analysis of infected cells reveals changes in 5-HT2 receptor cluster attributes, and radial distribution function (RDF) analyses suggest a significant increase in the aggregation of JCPyV particles colocalized with 5-HT2 receptor clusters in JCPyV-infected samples. These findings provide novel insights into receptor patterning during viral entry and highlight improved technologies for the future development of therapies for JCPyV infection as well as therapies for diseases involving 5-HT2 receptors.

Dengue virus NS4B protein as a target for developing antivirals

Dengue virus is an important pathogen affecting global population while no specific treatment is available against this virus. Effort has been made to develop inhibitors through targeting viral nonstructural proteins such as NS3 and NS5 with enzymatic activities. No potent inhibitors entering clinical studies have been developed so far due to many challenges. The genome of dengue virus encodes four membrane-bound nonstructural proteins which do not possess any enzymatic activities. Studies have shown that the membrane protein-NS4B is a validated target for drug discovery and several NS4B inhibitors exhibited antiviral activities in various assays and entered preclinical studies.. Here, we summarize the recent studies on dengue NS4B protein. The structure and membrane topology of dengue NS4B derived from biochemical and biophysical studies are described. Function of NS4B through protein-protein interactions and some available NS4B inhibitors are summarized. Accumulated studies demonstrated that cell-based assays play important roles in developing NS4B inhibitors. Although the atomic structure of NS4B is not obtained, target-based drug discovery approach become feasible to develop NS4B inhibitors as recombinant NS4B protein is available.

A TMPRSS2 inhibitor acts as a pan-SARS-CoV-2 prophylactic and therapeutic

The COVID-19 pandemic caused by the SARS-CoV-2 virus remains a global public health crisis. Although widespread vaccination campaigns are underway, their efficacy is reduced owing to emerging variants of concern1,2. Development of host-directed therapeutics and prophylactics could limit such resistance and offer urgently needed protection against variants of concern3,4. Attractive pharmacological targets to impede viral entry include type-II transmembrane serine proteases (TTSPs) such as TMPRSS2; these proteases cleave the viral spike protein to expose the fusion peptide for cell entry, and thus have an essential role in the virus lifecycle5,6. Here we identify and characterize a small-molecule compound, N-0385, which exhibits low nanomolar potency and a selectivity index of higher than 106 in inhibiting SARS-CoV-2 infection in human lung cells and in donor-derived colonoids7. In Calu-3 cells it inhibits the entry of the SARS-CoV-2 variants of concern B.1.1.7 (Alpha), B.1.351 (Beta), P.1 (Gamma) and B.1.617.2 (Delta). Notably, in the K18-human ACE2 transgenic mouse model of severe COVID-19, we found that N-0385 affords a high level of prophylactic and therapeutic benefit after multiple administrations or even after a single administration. Together, our findings show that TTSP-mediated proteolytic maturation of the spike protein is critical for SARS-CoV-2 infection in vivo, and suggest that N-0385 provides an effective early treatment option against COVID-19 and emerging SARS-CoV-2 variants of concern.

Methodologies for Generating Brain Organoids to Model Viral Pathogenesis in the CNS

(1) Background: The human brain is of interest in viral research because it is often the target of viruses. Neurological infections can result in consequences in the CNS, which can result in death or lifelong sequelae. Organoids modeling the CNS are notable because they are derived from stem cells that differentiate into specific brain cells such as neural progenitors, neurons, astrocytes, and glial cells. Numerous protocols have been developed for the generation of CNS organoids, and our goal was to describe the various CNS organoid models available for viral pathogenesis research to serve as a guide to determine which protocol might be appropriate based on research goal, timeframe, and budget. (2) Methods: Articles for this review were found in Pubmed, Scopus and EMBASE. The search terms used were "brain + organoid" and "CNS + organoid" (3) Results: There are two main methods for organoid generation, and the length of time for organoid generation varied from 28 days to over 2 months. The costs for generating a population of organoids ranged from USD 1000 to 5000. (4) Conclusions: There are numerous methods for generating organoids representing multiple regions of the brain, with several types of modifications for fine-tuning the model to a researcher's specifications. Organoid models of the CNS can serve as a platform for characterization and mechanistic studies that can reduce or eliminate the use of animals, especially for viruses that only cause disease in the human CNS.

Are the Organoid Models an Invaluable Contribution to ZIKA Virus Research?

In order to prevent new pathogen outbreaks and avoid possible new global health threats, it is important to study the mechanisms of microbial pathogenesis, screen new antiviral agents and test new vaccines using the best methods. In the last decade, organoids have provided a groundbreaking opportunity for modeling pathogen infections in human brains, including Zika virus (ZIKV) infection. ZIKV is a member of the Flavivirus genus, and it is recognized as an emerging infectious agent and a serious threat to global health. Organoids are 3D complex cellular models that offer an in-scale organ that is physiologically alike to the original one, useful for exploring the mechanisms behind pathogens infection; additionally, organoids integrate data generated in vitro with traditional tools and often support those obtained in vivo with animal model. In this mini-review the value of organoids for ZIKV research is examined and sustained by the most recent literature. Within a 3D viewpoint, tissue engineered models are proposed as future biological systems to help in deciphering pathogenic processes and evaluate preventive and therapeutic strategies against ZIKV. The next steps in this field constitute a challenge that may protect people and future generations from severe brain defects.

Towards a Quantitative Single Particle Characterization by Super Resolution Microscopy: From Virus Structures to Antivirals Design

In the last year the COVID19 pandemic clearly illustrated the potential threat that viruses pose to our society. The characterization of viral structures and the identification of key proteins involved in each step of the cycle of infection are crucial to develop treatments. However, the small size of viruses, invisible under conventional fluorescence microscopy, make it difficult to study the organization of protein clusters within the viral particle. The applications of super-resolution microscopy have skyrocketed in the last years, converting this group into one of the leading techniques to characterize viruses and study the viral infection in cells, breaking the diffraction limit by achieving resolutions up to 10 nm using conventional probes such as fluorescent dyes and proteins. There are several super-resolution methods available and the selection of the right one it is crucial to study in detail all the steps involved in the viral infection, quantifying and creating models of infection for relevant viruses such as HIV-1, Influenza, herpesvirus or SARS-CoV-1. Here we review the use of super-resolution microscopy (SRM) to study all steps involved in the viral infection and antiviral design. In light of the threat of new viruses, these studies could inspire future assays to unveil the viral mechanism of emerging viruses and further develop successful antivirals against them.

Challenges facing quantitative large-scale optical super-resolution, and some simple solutions

Optical super-resolution microscopy (SRM) has enabled biologists to visualize cellular structures with near-molecular resolution, giving unprecedented access to details about the amounts, sizes, and spatial distributions of macromolecules in the cell. Precisely quantifying these molecular details requires large datasets of high-quality, reproducible SRM images. In this review, we discuss the unique set of challenges facing quantitative SRM, giving particular attention to the shortcomings of conventional specimen preparation techniques and the necessity for optimal labeling of molecular targets. We further discuss the obstacles to scaling SRM methods, such as lengthy image acquisition and complex SRM data analysis. For each of these challenges, we review the recent advances in the field that circumvent these pitfalls and provide practical advice to biologists for optimizing SRM experiments.