Physics-driven structural docking and protein language models accelerate antibody screening and design for broad-spectrum antiviral therapy (preprint)

Therapeutic antibodies have become one of the most influential therapeutics in modern medicine to fight against infectious pathogens, cancer, and many other diseases. However, experimental screening for highly efficacious targeting antibodies is labor-intensive, which is exacerbated by evolving antigen targets under selective pressure such as fast-mutating viral variants. As a proof-of-concept, we developed a machine learning-assisted antibody generation pipeline that greatly accelerates the screening and re-design of immunoglobulins G (IgGs) against a broad spectrum of coronavirus variants. Using over 1300 IgG sequences derived from patient B cells bound with the viral spike's receptor binding domain (RBD), we first established protein structural docking models in assessing the IgG-RBD-ACE2 interaction interfaces and predicting their viral neutralizing activities with a confidence score. The confidence score is calculated as a fraction of IgG-blocking RBD binding sites versus all ACE2-interacting sites. Additionally, employing Gaussian process regression (also known as Kriging) in a latent space of an antibody language model, we predicted the IgGs' activity profiles against viral strains. Using functional analyses and experimental validations, we subsequently prioritized IgG candidates for neutralizing a broad spectrum of viral variants (wildtype, Delta, and Omicron) and preventing the infection of host cells in vitro and hACE2 transgenic mice in vivo. To further improve the blockade efficacy for the Delta strain (B.1.617), we rationally redesigned the IgG clones with single amino acid substitutions at the RBD-binding interface. Our work expedites applications of artificial intelligence in low-data regimes when limited data is available, including protein language models (using unlabeled data) and Kriging (using few labeled data) for antibody sequence analysis, activity prediction, and efficacy improvement, which are aided by physics-driven protein docking models for antibody-antigen interface structure analyses.

Quantity of SARS-CoV-2 RNA copies exhaled per minute during natural breathing over the course of COVID-19 infection (preprint)

SARS-CoV-2 is spread through exhaled breath of infected individuals. A fundamental question in understanding transmission of SARS-CoV-2 is how much virus an individual is exhaling into the environment while they breathe, over the course of their infection. Research on viral load dynamics during COVID-19 infection has focused on internal swab specimens, which provide a measure of viral loads inside the respiratory tract, but not on breath. Therefore, the dynamics of viral shedding on exhaled breath over the course of infection are poorly understood. Here, we collected exhaled breath specimens from COVID-19 patients and used RTq-PCR to show that numbers of exhaled SARS-CoV-2 RNA copies during COVID-19 infection do not decrease significantly until day 8 from symptom-onset. COVID-19-positive participants exhaled an average of 80 SARS-CoV-2 viral RNA copies per minute during the first 8 days of infection, with significant variability both between and within individuals, including spikes over 800 copies a minute in some patients. After day 8, there was a steep drop to levels nearing the limit of detection, persisting for up to 20 days. We further found that levels of exhaled viral RNA increased with self-rated symptom-severity, though individual variation was high. Levels of exhaled viral RNA did not differ across age, sex, time of day, vaccination status or viral variant. Our data provide a fine-grained, direct measure of the number of SARS-CoV-2 viral copies exhaled per minute during natural breathing, including 312 breath specimens collected multiple times daily over the course of infection, in order to fill an important gap in our understanding of the time course of exhaled viral loads in COVID-19.

Genetic regulation of OAS1 nonsense-mediated decay underlies association with risk of severe COVID-19 (preprint)

Genomic regions have been associated with COVID-19 susceptibility and outcomes, including the chr12q24.13 locus encoding antiviral proteins OAS1-3. Here, we report genetic, functional, and clinical insights into genetic associations within this locus. In Europeans, the risk of hospitalized vs. non-hospitalized COVID-19 was associated with a single 19Kb-haplotype comprised of 76 OAS1 variants included in a 95% credible set within a large genomic fragment introgressed from Neandertals. The risk haplotype was also associated with impaired spontaneous but not treatment-induced SARS-CoV-2 clearance in a clinical trial with pegIFN-{lambda}1. We demonstrate that two exonic variants, rs10774671 and rs1131454, affect splicing and nonsense-mediated decay of OAS1. We suggest that genetically-regulated loss of OAS1 expression contributes to impaired spontaneous clearance of SARS-CoV-2 and elevated risk of hospitalization for COVID-19. Our results provide the rationale for further clinical studies using interferons to compensate for impaired spontaneous SARS-CoV-2 clearance, particularly in carriers of the OAS1 risk haplotypes.

Assessment and Modeling of COVID-19 Outcomes in a Longitudinal Cohort of Hospitalized Adults (preprint)

Background: While several demographic and clinical correlates of Coronavirus Disease 2019 (COVID-19) outcome have been identified, they remain imprecise tools for clinical management of disease. Furthermore, there are limited data on how these factors are associated with virological and immunological parameters over time. Methods and Findings: Nasopharyngeal swabs and blood samples were longitudinally collected from a cohort of 58 hospitalized adults with COVID-19 in Chicago, Illinois between March 27 and June 9, 2020. Samples were assessed for SARS-CoV-2 viral load, viral genotype, viral diversity, and antibody titer. Demographic and clinical information, including patient blood tests and several composite measures of disease severity, were extracted from electronic health records. All parameters were assessed for association with three patient outcome groups: discharge without intensive care unit (ICU) admission (n = 23), discharge with ICU admission (n = 29), and COVID-19 related death (n = 6). Higher age, male sex, and higher body mass index (BMI) were significantly associated with ICU admission. At hospital admission, higher 4C Mortality scores and lactate dehydrogenase (LDH) levels were likewise associated with ICU admission. Longitudinal trends in Deterioration Index (DI) score, Modified Early Warning Score (MEWS), and serum neutrophil count were also associated with ICU admission, though only the retrospectively calculated median DI score was predictive of death. While viral load and genotype were not significantly associated with outcome in this study, viral load did correlate positively with C-reactive protein levels and negatively with D-dimer, lymphocyte count, and antibody titer. Intra-host viral genetic diversity resulted in changes in viral genotype in some participants over time, though intra-host evolution was not associated with outcome. A stepwise-generated multivariable model including BMI, lymphocyte count at admission, and neutrophil count at admission was sufficient to predict outcome with a 0.82 accuracy rate in this cohort. Conclusions: These studies suggest that COVID-19 disease severity and poor outcomes among hospitalized patients are likely driven by dysfunctional host responses to infection and underlying co-morbid conditions rather than SARS-CoV-2 viral loads. Several parameters, including 4C mortality score, LDH levels, and DI score, were ultimately predictive of participant outcome and warrant further exploration in larger cohort studies for use in clinical management and risk assessment. Finally, the prevalence of intra-host diversity and viral evolution in hospitalized patients suggests a mechanism for population-level change, further emphasizing the need for effective antivirals to suppress viral replication and to avoid the emergence of new variants.

Circulating ACE2-expressing Exosomes Block SARS-CoV-2 Virus Infection as an Innate Antiviral Mechanism (preprint)

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes the coronavirus disease 2019 (COVID-19) with innate and adaptive immune response triggered in such patients by viral antigens. Both convalescent plasma and engineered high affinity human monoclonal antibodies have shown therapeutic potential to treat COVID-19. Whether additional antiviral soluble factors exist in peripheral blood remain understudied. Herein, we detected circulating exosomes that express the SARS-CoV-2 viral entry receptor angiotensin-converting enzyme 2 (ACE2) in plasma of both healthy donors and convalescent COVID-19 patients. We demonstrated that exosomal ACE2 competes with cellular ACE2 for neutralization of SARS-CoV-2 infection. ACE2-expressing (ACE2+) exosomes blocked the binding of the viral spike (S) protein RBD to ACE2+ cells in a dose dependent manner, which was 400- to 700-fold more potent than that of vesicle-free recombinant human ACE2 extracellular domain protein (rhACE2). As a consequence, exosomal ACE2 prevented SARS-CoV-2 pseudotype virus tethering and infection of human host cells at a 50-150 fold higher efficacy than rhACE2. A similar antiviral activity of exosomal ACE2 was further demonstrated to block wild-type live SARS-CoV-2 infection. Of note, depletion of ACE2+ exosomes from COVID-19 patient plasma impaired the ability to block SARS-CoV-2 RBD binding to host cells. Our data demonstrate that ACE2+ exosomes can serve as a decoy therapeutic and a possible innate antiviral mechanism to block SARS-CoV-2 infection.

Ca2+-dependent mechanism of membrane insertion and destabilization by the SARS-CoV-2 fusion peptide (preprint)

Cell penetration after recognition of the SARS-CoV-2 virus by the ACE2 receptor, and the fusion of its viral envelope membrane with cellular membranes, are the early steps of infectivity. A region of the Spike protein (S) of the virus, identified as the "fusion peptide" (FP), is liberated at its N-terminal site by a specific cleavage occurring in concert with the interaction of the receptor binding domain of the Spike. Studies have shown that penetration is enhanced by the required binding of Ca2+ ions to the FPs of corona viruses, but the mechanisms of membrane insertion and destabilization remain unclear. We have predicted the preferred positions of Ca2+ binding to the SARS-CoV-2-FP, the role of Ca2+ ions in mediating peptide-membrane interactions, the preferred mode of insertion of the Ca2+-bound SARS-CoV-2-FP and consequent effects on the lipid bilayer from extensive atomistic molecular dynamics (MD) simulations and trajectory analyses. In a systematic sampling of the interactions of the Ca2+-bound peptide models with lipid membranes SARS-CoV-2-FP penetrated the bilayer and disrupted its organization only in two modes involving different structural domains. In one, the hydrophobic residues F833/I834 from the middle region of the peptide are inserted. In the other, more prevalent mode, the penetration involves residues L822/F823 from the LLF motif which is conserved in CoV-2-like viruses, and is achieved by the binding of Ca2+ ions to the D830/D839 and E819/D820 residue pairs. FP penetration is shown to modify the molecular organization in specific areas of the bilayer, and the extent of membrane binding of the SARS-CoV-2 FP is significantly reduced in the absence of Ca2+ ions. These findings provide novel mechanistic insights regarding the role of Ca2+ in mediating SARS-CoV-2 fusion and provide a detailed structural platform to aid the ongoing efforts in rational design of compounds to inhibit SARS-CoV-2 cell entry. STATEMENT OF SIGNIFICANCESARS-CoV-2, the cause of the COVID-19 pandemic, penetrates host cell membranes and uses viral-to-cellular membrane fusion to release its genetic material for replication. Experiments had identified a region termed "fusion peptide" (FP) in the Spike proteins of coronaviruses, as the spearhead in these initial processes, and suggested that Ca2+ is needed to support both functions. Absent structure and dynamics-based mechanistic information these FP functions could not be targeted for therapeutic interventions. We describe the development and determination of the missing information from analysis of extensive MD simulation trajectories, and propose specific Ca2+-dependent mechanisms of SARS-CoV-2-FP membrane insertion and destabilization. These results offer a structure-specific platform to aid the ongoing efforts to use this target for the discovery and/or of inhibitors.

Patterns and persistence of SARS-CoV-2 IgG antibodies in a US metropolitan site (preprint)

BackgroundEstimates of seroprevalence to SARS-CoV-2 vary widely. We ascertained IgG levels across a single US metropolitan site, Chicago, over the 2020 summer, a period when restrictions on activities had been lifted. MethodsWe utilized a self-sampled dried blood spot assay to quantitatively monitor antibodies to the receptor binding domain (RBD) of the spike glycoprotein of SARS-CoV-2 in 1545 participants, with return of blood spot cards either by mail or in-person drop-off. ResultsSeroprevalence was 19.8%, with no significant difference between method of contact, or between essential and non-essential workers. Only a small number (1.2%) of participants reported having had a diagnosis of COVID-19 based on virus detection, consistent with a 16-fold greater exposure to SARS-CoV-2 measured by serology than detected by viral testing. Only a modest correlation was observed between having antibodies to SARS-CoV-2 nucleocapsid compared to RBD, with many only having detectable anti-RBD antibodies. From a subset of those who participated in repeat testing, three-quarters of seropositive individuals retained detectable antibodies for at least 120 days. One seropositive individual experienced a strong boost in IgG levels following a symptomatic illness, suggestive of potential re-exposure. ConclusionsThese data underscore the importance of a self-collected, quantitative assay with adequate sensitivity to detect antibodies at the lower levels among non-hospitalized persons with community-acquired exposure to COVID-19.

Sequential ER stress and inflammatory responses are induced by SARS-CoV-2 ORF3 through ERphagy (preprint)

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections have resulted in a number of severe cases of COVID-19 and deaths worldwide. However, knowledge of SARS-CoV-2 infection, diseases and therapy remains limited, underlining the urgency of fundamental studies and drug development. Studies have shown that induction of autophagy and hijacking of autophagic machinery are essential for infection and replication of SARS-CoV-2; however, the mechanism of this manipulation and function of autophagy during SARS-CoV-2 infection remain unclear. In the present study, we identified ORF3 as an inducer of autophagy and revealed that ORF3 localizes to the ER and induces FAM134B-related ERphagy through the HMGB1-Beclin1 pathway. As a consequence, ORF3 induces ER stress and inflammatory responses through ERphagy and sensitizes cells to ER stress-induced cell death, suggesting that SARS-CoV-2 ORF3 hijacks ERphagy and then harms ER homeostasis to induce inflammatory responses through excessive ER stress. These findings reveal a sequential induction of ERphagy, ER stress and acute inflammatory responses during SARS-CoV-2 infection and provide therapeutic potential for ERphagy and ER stress-related drugs for COVID-19 treatment and prevention. ImportanceSARS-CoV-2 infection and replication require autophagosome-like double-membrane vacuoles. Inhibition of autophagy suppresses viral replication, indicating the essential role of autophagy in SARS-CoV-2 infection. However, how SARS-CoV-2 hijacks autophagy and the function of autophagy in the disease progression remain unknown. Here, we reveal that SARS-CoV-2 ORF3 induces ERphagy and consequently induces ER stress to trigger acute inflammatory responses and enhance sensitivity to ER stress-induced apoptosis. Our studies uncover ERphagy-induced inflammatory responses during SARS-CoV-2 infection and provide a promising therapeutic approach for treating SARS-CoV-2 infection and inflammatory responses in COVID-19 by manipulating autophagy and ER stress.

Single-cell Transcriptome of Bronchoalveolar Lavage Fluid Reveals Dynamic Change of Macrophages During SARS-CoV-2 Infection in Ferrets (preprint)

Although the profile of immune cells changes during the natural course of SARS-CoV-2 inflection in human patients, few studies have used a longitudinal approach to reveal their dynamic features. Here, we performed single-cell RNA sequencing of bronchoalveolar lavage fluid cells longitudinally obtained from SARS-CoV-2-infected ferrets. Landscape analysis of the lung immune microenvironment showed dynamic changes in cell proportions and characteristics in uninfected control, at 2 days post-infection (dpi) (early stage of SARS-CoV-2 infection with peak viral titer), and 5 dpi (resolution phase). NK cells and CD8+ T cells exhibited activated subclusters with interferon-stimulated features, which were peaked at 2 dpi. Intriguingly, macrophages were classified into 10 distinct subpopulations, and their relative proportions changed over the time. We observed prominent transcriptome changes among monocyte-derived infiltrating macrophages and differentiated M1/M2 macrophages, especially at 2 dpi. Moreover, trajectory analysis revealed gene expression changes from monocyte-derived infiltrating macrophages toward M1 or M2 macrophages and identified the distinct macrophage subpopulation that had rapidly undergone SARS-CoV-2-mediated activation of inflammatory responses. Finally, we found that different spectrums of M1 or M2 macrophages showed distinct patterns of gene modules downregulated by immune-modulatory drugs. Overall, these results elucidate fundamental aspects of the immune response dynamics provoked by SARS-CoV-2 infection.

Replicating bacterium-vectored vaccine expressing SARS-CoV-2 Membrane and Nucleocapsid proteins protects against severe COVID-19 disease in hamsters (preprint)

An inexpensive readily manufactured COVID-19 vaccine that protects against severe disease is needed to combat the pandemic. We have employed the LVS {Delta}capB vector platform, previously used successfully to generate potent vaccines against the Select Agents of tularemia, anthrax, plague, and melioidosis, to generate a COVID-19 vaccine. The LVS {Delta}capB vector, a replicating intracellular bacterium, is a highly attenuated derivative of a tularemia vaccine (LVS) previously administered to millions of people. We generated vaccines expressing SARS-CoV-2 structural proteins and evaluated them for efficacy in the golden Syrian hamster, which develops severe COVID-19 disease. Hamsters immunized intradermally or intranasally with a vaccine co-expressing the Membrane (M) and Nucleocapsid (N) proteins, then challenged 5-weeks later with a high dose of SARS-CoV-2, were protected against severe weight loss and lung pathology and had reduced viral loads in the oropharynx and lungs. Protection by the vaccine, which induces murine N-specific interferon-gamma secreting T cells, was highly correlated with pre-challenge serum anti-N TH1-biased IgG. This potent vaccine against severe COVID-19 should be safe and easily manufactured, stored, and distributed, and given the high homology between MN proteins of SARS-CoV and SARS-CoV-2, has potential as a universal vaccine against the SARS subset of pandemic causing {beta}-coronaviruses.

A Unique Clade of SARS-CoV-2 Viruses is Associated with Lower Viral Loads in Patient Upper Airways (preprint)

BackgroundThe rapid spread of SARS-CoV-2, the causative agent of Coronavirus disease 2019 (COVID- 19), has been accompanied by the emergence of distinct viral clades, though their clinical significance remains unclear. Here, we aimed to investigate the phylogenetic characteristics of SARS-CoV-2 infections in Chicago, Illinois and assess their relationship to clinical parameters. MethodsWe performed whole-genome sequencing of SARS-CoV-2 isolates collected from COVID-19 patients in a Chicago healthcare system in mid-March, 2020. Using these and other publicly available sequences, we performed phylogenetic, phylogeographic, and phylodynamic analyses. Patient data was assessed for correlations between demographic or clinical characteristics and virologic features. FindingsThe 88 SARS-CoV-2 genome sequences in our study separated into three distinct phylogenetic clades. Clade 1 was most closely related to viral sequences from New York, and showed evidence of rapid expansion across the US, while Clade 3 was most closely related to those in Washington state. Clade 2 was localized primarily to the Chicago area with limited evidence of expansion elsewhere. At the time of diagnosis, patients infected with Clade 1 viruses had significantly higher average viral loads in their upper airways relative to patients infected with Clade 2 viruses, independent of time to symptom onset and disease severity. InterpretationThese results show that multiple variants of SARS-CoV-2 are circulating in the Chicago area that differ in their relative viral loads in patient upper airways. These data suggest that differences in virus genotype impact viral load and may in turn influence viral transmission and spread. FundingDixon Family Translational Research Award, Northwestern University Clinical and Translational Sciences Institute (NUCATS), National Institute of Allergy and Infectious Diseases (NIAID)