Accurate Measurement of the Effects of All Amino-Acid Mutations on Influenza Hemagglutinin

Age-dependent heterogeneity in the antigenic effects of mutations to influenza hemagglutinin

Human influenza virus evolves to escape neutralization by polyclonal antibodies. However, we have a limited understanding of how the antigenic effects of viral mutations vary across the human population and how this heterogeneity affects virus evolution. Here, we use deep mutational scanning to map how mutations to the hemagglutinin (HA) proteins of two H3N2 strains, A/Hong Kong/45/2019 and A/Perth/16/2009, affect neutralization by serum from individuals of a variety of ages. The effects of HA mutations on serum neutralization differ across age groups in ways that can be partially rationalized in terms of exposure histories. Mutations that were fixed in influenza variants after 2020 cause greater escape from sera from younger individuals compared with adults. Overall, these results demonstrate that influenza faces distinct antigenic selection regimes from different age groups and suggest approaches to understand how this heterogeneous selection shapes viral evolution.

Functional effects of mutations in proteins can be predicted and interpreted by guided selection of sequence covariation information

Predicting the effects of one or more mutations to the in vivo or in vitro properties of a wild-type protein is a major computational challenge, due to the presence of epistasis, that is, of interactions between amino acids in the sequence. We introduce a computationally efficient procedure to build minimal epistatic models to predict mutational effects by combining evolutionary (homologous sequence) and few mutational-scan data. Mutagenesis measurements guide the selection of links in a sparse graphical model, while the parameters on the nodes and the edges are inferred from sequence data. We show, on 10 mutational scans, that our pipeline exhibits performances comparable to state-of-the-art deep networks trained on many more data, while requiring much less parameters and being hence more interpretable. In particular, the identified interactions adapt to the wild-type protein and to the fitness or biochemical property experimentally measured, mostly focus on key functional sites, and are not necessarily related to structural contacts. Therefore, our method is able to extract information relevant for one mutational experiment from homologous sequence data reflecting the multitude of structural and functional constraints acting on proteins throughout evolution.

Natural variation in neuraminidase activity influences the evolutionary potential of the seasonal H1N1 lineage hemagglutinin

The antigenic evolution of the influenza A virus hemagglutinin (HA) gene poses a major challenge for the development of vaccines capable of eliciting long-term protection. Prior efforts to understand the mechanisms that govern viral antigenic evolution mainly focus on HA in isolation, ignoring the fact that HA must act in concert with the viral neuraminidase (NA) during replication and spread. Numerous studies have demonstrated that the degree to which the receptor-binding avidity of HA and receptor-cleaving activity of NA are balanced with each other influences overall viral fitness. We recently showed that changes in NA activity can significantly alter the mutational fitness landscape of HA in the context of a lab-adapted virus strain. Here, we test whether natural variation in relative NA activity can influence the evolutionary potential of HA in the context of the seasonal H1N1 lineage (pdmH1N1) that has circulated in humans since the 2009 pandemic. We observed substantial variation in the relative activities of NA proteins encoded by a panel of H1N1 vaccine strains isolated between 2009 and 2019. We comprehensively assessed the effect of NA background on the HA mutational fitness landscape in the circulating pdmH1N1 lineage using deep mutational scanning and observed pronounced epistasis between NA and residues in or near the receptor-binding site of HA. To determine whether NA variation could influence the antigenic evolution of HA, we performed neutralizing antibody selection experiments using a panel of monoclonal antibodies targeting different HA epitopes. We found that the specific antibody escape profiles of HA were highly contingent upon NA background. Overall, our results indicate that natural variation in NA activity plays a significant role in governing the evolutionary potential of HA in the currently circulating pdmH1N1 lineage.

A chimeric haemagglutinin-based universal influenza virus vaccine boosts human cellular immune responses directed towards the conserved haemagglutinin stalk domain and the viral nucleoprotein

BACKGROUND

The development of a universal influenza virus vaccine, to protect against both seasonal and pandemic influenza A viruses, is a long-standing public health goal. The conserved stalk domain of haemagglutinin (HA) is a promising vaccine target. However, the stalk is immunosubdominant. As such, innovative approaches are required to elicit robust immunity against this domain. In a previously reported observer-blind, randomised placebo-controlled phase I trial (NCT03300050), immunisation regimens using chimeric HA (cHA)-based immunogens formulated as inactivated influenza vaccines (IIV) -/+ AS03 adjuvant, or live attenuated influenza vaccines (LAIV), elicited durable HA stalk-specific antibodies with broad reactivity. In this study, we sought to determine if these vaccines could also boost T cell responses against HA stalk, and nucleoprotein (NP).

METHODS

We measured interferon-γ (IFN-γ) responses by Enzyme-Linked ImmunoSpot (ELISpot) assay at baseline, seven days post-prime, pre-boost and seven days post-boost following heterologous prime:boost regimens of LAIV and/or adjuvanted/unadjuvanted IIV-cHA vaccines.

FINDINGS

Our findings demonstrate that immunisation with adjuvanted cHA-based IIVs boost HA stalk-specific and NP-specific T cell responses in humans. To date, it has been unclear if HA stalk-specific T cells can be boosted in humans by HA-stalk focused universal vaccines. Therefore, our study will provide valuable insights for the design of future studies to determine the precise role of HA stalk-specific T cells in broad protection.

INTERPRETATION

Considering that cHA-based vaccines also elicit stalk-specific antibodies, these data support the further clinical advancement of cHA-based universal influenza vaccine candidates.

FUNDING

This study was funded in part by the Bill and Melinda Gates Foundation (BMGF).

Functional and antigenic characterization of SARS-CoV-2 spike fusion peptide by deep mutational scanning

The fusion peptide of SARS-CoV-2 spike protein is functionally important for membrane fusion during virus entry and is part of a broadly neutralizing epitope. However, sequence determinants at the fusion peptide and its adjacent regions for pathogenicity and antigenicity remain elusive. In this study, we perform a series of deep mutational scanning (DMS) experiments on an S2 region spanning the fusion peptide of authentic SARS-CoV-2 in different cell lines and in the presence of broadly neutralizing antibodies. We identify mutations at residue 813 of the spike protein that reduced TMPRSS2-mediated entry with decreased virulence. In addition, we show that an F823Y mutation, present in bat betacoronavirus HKU9 spike protein, confers resistance to broadly neutralizing antibodies. Our findings provide mechanistic insights into SARS-CoV-2 pathogenicity and also highlight a potential challenge in developing broadly protective S2-based coronavirus vaccines.

Natural variation in neuraminidase activity influences the evolutionary potential of the seasonal H1N1 lineage hemagglutinin

The antigenic evolution of the influenza A virus hemagglutinin (HA) gene poses a major challenge for the development of vaccines capable of eliciting long-term protection. Prior efforts to understand the mechanisms that govern viral antigenic evolution mainly focus on HA in isolation, ignoring the fact that HA must act in concert with the viral neuraminidase (NA) during replication and spread. Numerous studies have demonstrated that the degree to which the receptor binding avidity of HA and receptor cleaving activity of NA are balanced with each other influences overall viral fitness. We recently showed that changes in NA activity can significantly alter the mutational fitness landscape of HA in the context of a lab-adapted virus strain. Here, we test whether natural variation in relative NA activity can influence the evolutionary potential of HA in the context of the seasonal H1N1 lineage (pdmH1N1) that has circulated in humans since the 2009 pandemic. We observed substantial variation in the relative activities of NA proteins encoded by a panel of H1N1 vaccine strains isolated between 2009 and 2019. We comprehensively assessed the effect of NA background on the HA mutational fitness landscape in the circulating pdmH1N1 lineage using deep mutational scanning and observed pronounced epistasis between NA and residues in or near the receptor binding site of HA. To determine whether NA variation could influence the antigenic evolution of HA, we performed neutralizing antibody selection experiments using a panel of monoclonal antibodies targeting different HA epitopes. We found that the specific antibody escape profiles of HA were highly contingent upon NA background. Overall, our results indicate that natural variation in NA activity plays a significant role in governing the evolutionary potential of HA in the currently circulating pdmH1N1 lineage.

Immunogenicity of chimeric hemagglutinins delivered by an orf virus vector platform against swine influenza virus

Orf virus (ORFV) is a large DNA virus that can harbor and efficiently deliver viral antigens in swine. Here we used ORFV as a vector platform to deliver chimeric hemagglutinins (HA) of Influenza A virus of swine (IAV-S). Vaccine development against IAV-S faces limitations posed by strain-specific immunity and the antigenic diversity of the IAV-S strains circulating in the field. A promising alternative aiming at re-directing immune responses on conserved epitopes of the stalk segment of the hemagglutinin (HA2) has recently emerged. Sequential immunization with chimeric HAs comprising the same stalk but distinct exotic head domains can potentially induce cross-reactive immune responses against conserved epitopes of the HA2 while breaking the immunodominance of the head domain (HA1). Here, we generated two recombinant ORFVs expressing chimeric HAs encoding the stalk region of a contemporary H1N1 IAV-S strain and exotic heads derived from either H6 or H8 subtypes, ORFVΔ121cH6/1 and ORFVΔ121cH8/1, respectively. The resulting recombinant viruses were able to express the heterologous protein in vitro. Further, the immunogenicity and cross-protection of these vaccine candidates were assessed in swine after sequential intramuscular immunization with OV-cH6/1 and OV-cH8/1, and subsequent challenge with divergent IAV-S strains. Humoral responses showed that vaccinated piglets presented increasing IgG responses in sera. Additionally, cross-reactive IgG and IgA antibody responses elicited by immunization were detected in sera and bronchoalveolar lavage (BAL), respectively, by ELISA against different viral clades and a diverse range of contemporary H1N1 IAV-S strains, indicating induction of humoral and mucosal immunity in vaccinated animals. Importantly, viral shedding was reduced in nasal swabs from vaccinated piglets after intranasal challenge with either Oh07 (gamma clade) or Ca09 (npdm clade) IAV-S strains. These results demonstrated the efficiency of ORFV-based vectors in delivering chimeric IAV-S HA-based vaccine candidates and underline the potential use of chimeric-HAs for prevention and control of influenza in swine.

Efficient evolution of human antibodies from general protein language models

Natural evolution must explore a vast landscape of possible sequences for desirable yet rare mutations, suggesting that learning from natural evolutionary strategies could guide artificial evolution. Here we report that general protein language models can efficiently evolve human antibodies by suggesting mutations that are evolutionarily plausible, despite providing the model with no information about the target antigen, binding specificity or protein structure. We performed language-model-guided affinity maturation of seven antibodies, screening 20 or fewer variants of each antibody across only two rounds of laboratory evolution, and improved the binding affinities of four clinically relevant, highly mature antibodies up to sevenfold and three unmatured antibodies up to 160-fold, with many designs also demonstrating favorable thermostability and viral neutralization activity against Ebola and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pseudoviruses. The same models that improve antibody binding also guide efficient evolution across diverse protein families and selection pressures, including antibiotic resistance and enzyme activity, suggesting that these results generalize to many settings.

Advancing Antibody Engineering through Synthetic Evolution and Machine Learning

Abs are versatile molecules with the potential to achieve exceptional binding to target Ags, while also possessing biophysical properties suitable for therapeutic drug development. Protein display and directed evolution systems have transformed synthetic Ab discovery, engineering, and optimization, vastly expanding the number of Ab clones able to be experimentally screened for binding. Moreover, the burgeoning integration of high-throughput screening, deep sequencing, and machine learning has further augmented in vitro Ab optimization, promising to accelerate the design process and massively expand the Ab sequence space interrogated. In this Brief Review, we discuss the experimental and computational tools employed in synthetic Ab engineering and optimization. We also explore the therapeutic challenges posed by developing Abs for infectious diseases, and the prospects for leveraging machine learning-guided protein engineering to prospectively design Abs resistant to viral escape.

Combinatorial immune refocusing within the influenza hemagglutinin RBD improves cross-neutralizing antibody responses

The receptor-binding domain (RBD) of influenza virus hemagglutinin (HA) elicits potently neutralizing yet mostly strain-specific antibodies. Here, we evaluate the ability of several immunofocusing techniques to enhance the functional breadth of vaccine-elicited immune responses against the HA RBD. We present a series of "trihead" nanoparticle immunogens that display native-like closed trimeric RBDs from the HAs of several H1N1 influenza viruses. The series includes hyperglycosylated and hypervariable variants that incorporate natural and designed sequence diversity at key positions in the receptor-binding site periphery. Nanoparticle immunogens displaying triheads or hyperglycosylated triheads elicit higher hemagglutination inhibition (HAI) and neutralizing activity than the corresponding immunogens lacking either trimer-stabilizing mutations or hyperglycosylation. By contrast, mosaic nanoparticle display and antigen hypervariation do not significantly alter the magnitude or breadth of vaccine-elicited antibodies. Our results yield important insights into antibody responses against the RBD and the ability of several structure-based immunofocusing techniques to influence vaccine-elicited antibody responses.

Age-dependent heterogeneity in the antigenic effects of mutations to influenza hemagglutinin

Human influenza virus evolves to escape neutralization by polyclonal antibodies. However, we have a limited understanding of how the antigenic effects of viral mutations vary across the human population, and how this heterogeneity affects virus evolution. Here we use deep mutational scanning to map how mutations to the hemagglutinin (HA) proteins of the A/Hong Kong/45/2019 (H3N2) and A/Perth/16/2009 (H3N2) strains affect neutralization by serum from individuals of a variety of ages. The effects of HA mutations on serum neutralization differ across age groups in ways that can be partially rationalized in terms of exposure histories. Mutations that fixed in influenza variants after 2020 cause the greatest escape from sera from younger individuals. Overall, these results demonstrate that influenza faces distinct antigenic selection regimes from different age groups, and suggest approaches to understand how this heterogeneous selection shapes viral evolution.

ProteinGym: Large-Scale Benchmarks for Protein Design and Fitness Prediction

Predicting the effects of mutations in proteins is critical to many applications, from understanding genetic disease to designing novel proteins that can address our most pressing challenges in climate, agriculture and healthcare. Despite a surge in machine learning-based protein models to tackle these questions, an assessment of their respective benefits is challenging due to the use of distinct, often contrived, experimental datasets, and the variable performance of models across different protein families. Addressing these challenges requires scale. To that end we introduce ProteinGym, a large-scale and holistic set of benchmarks specifically designed for protein fitness prediction and design. It encompasses both a broad collection of over 250 standardized deep mutational scanning assays, spanning millions of mutated sequences, as well as curated clinical datasets providing high-quality expert annotations about mutation effects. We devise a robust evaluation framework that combines metrics for both fitness prediction and design, factors in known limitations of the underlying experimental methods, and covers both zero-shot and supervised settings. We report the performance of a diverse set of over 70 high-performing models from various subfields (eg., alignment-based, inverse folding) into a unified benchmark suite. We open source the corresponding codebase, datasets, MSAs, structures, model predictions and develop a user-friendly website that facilitates data access and analysis.

Co-evolution of immunity and seasonal influenza viruses

Seasonal influenza viruses cause recurring global epidemics by continually evolving to escape host immunity. The viral constraints and host immune responses that limit and drive the evolution of these viruses are increasingly well understood. However, it remains unclear how most of these advances improve the capacity to reduce the impact of seasonal influenza viruses on human health. In this Review, we synthesize recent progress made in understanding the interplay between the evolution of immunity induced by previous infections or vaccination and the evolution of seasonal influenza viruses driven by the heterogeneous accumulation of antibody-mediated immunity in humans. We discuss the functional constraints that limit the evolution of the viruses, the within-host evolutionary processes that drive the emergence of new virus variants, as well as current and prospective options for influenza virus control, including the viral and immunological barriers that must be overcome to improve the effectiveness of vaccines and antiviral drugs.

The effect of single mutations in Zika virus envelope on escape from broadly neutralizing antibodies

IMPORTANCE

The wide endemic range of mosquito-vectored flaviviruses-such as Zika virus and dengue virus serotypes 1-4-places hundreds of millions of people at risk of infection every year. Despite this, there are no widely available vaccines, and treatment of severe cases is limited to supportive care. An avenue toward development of more widely applicable vaccines and targeted therapies is the characterization of monoclonal antibodies that broadly neutralize all these viruses. Here, we measure how single amino acid mutations in viral envelope protein affect neutralizing antibodies with both broad and narrow specificities. We find that broadly neutralizing antibodies with potential as vaccine prototypes or biological therapeutics are quantifiably more difficult to escape than narrow, virus-specific neutralizing antibodies.

Deep mutational scanning reveals the functional constraints and evolutionary potential of the influenza A virus PB1 protein

IMPORTANCE

The influenza virus polymerase is important for adaptation to new hosts and, as a determinant of mutation rate, for the process of adaptation itself. We performed a deep mutational scan of the polymerase basic 1 (PB1) protein to gain insights into the structural and functional constraints on the influenza RNA-dependent RNA polymerase. We find that PB1 is highly constrained at specific sites that are only moderately predicted by the global structure or larger domain. We identified a number of beneficial mutations, many of which have been shown to be functionally important or observed in influenza virus' natural evolution. Overall, our atlas of PB1 mutations and their fitness impacts serves as an important resource for future studies of influenza replication and evolution.

Functional and antigenic characterization of SARS-CoV-2 spike fusion peptide by deep mutational scanning

The fusion peptide of SARS-CoV-2 spike protein is functionally important for membrane fusion during virus entry and is part of a broadly neutralizing epitope. However, sequence determinants at the fusion peptide and its adjacent regions for pathogenicity and antigenicity remain elusive. In this study, we performed a series of deep mutational scanning (DMS) experiments on an S2 region spanning the fusion peptide of authentic SARS-CoV-2 in different cell lines and in the presence of broadly neutralizing antibodies. We identified mutations at residue 813 of the spike protein that reduced TMPRSS2-mediated entry with decreased virulence. In addition, we showed that an F823Y mutation, present in bat betacoronavirus HKU9 spike protein, confers resistance to broadly neutralizing antibodies. Our findings provide mechanistic insights into SARS-CoV-2 pathogenicity and also highlight a potential challenge in developing broadly protective S2-based coronavirus vaccines.

Accurate prediction of functional effect of single amino acid variants with deep learning

The assessment of functional effect of amino acid variants is a critical biological problem in proteomics for clinical medicine and protein engineering. Although natively occurring variants offer insights into deleterious variants, high-throughput deep mutational experiments enable comprehensive investigation of amino acid variants for a given protein. However, these mutational experiments are too expensive to dissect millions of variants on thousands of proteins. Thus, computational approaches have been proposed, but they heavily rely on hand-crafted evolutionary conservation, limiting their accuracy. Recent advancement in transformers provides a promising solution to precisely estimate the functional effects of protein variants on high-throughput experimental data. Here, we introduce a novel deep learning model, namely Rep2Mut-V2, which leverages learned representation from transformer models. Rep2Mut-V2 significantly enhances the prediction accuracy for 27 types of measurements of functional effects of protein variants. In the evaluation of 38 protein datasets with 118,933 single amino acid variants, Rep2Mut-V2 achieved an average Spearman's correlation coefficient of 0.7. This surpasses the performance of six state-of-the-art methods, including the recently released methods ESM, DeepSequence and EVE. Even with limited training data, Rep2Mut-V2 outperforms ESM and DeepSequence, showing its potential to extend high-throughput experimental analysis for more protein variants to reduce experimental cost. In conclusion, Rep2Mut-V2 provides accurate predictions of the functional effects of single amino acid variants of protein coding sequences. This tool can significantly aid in the interpretation of variants in human disease studies.

Learning from prepandemic data to forecast viral escape

Effective pandemic preparedness relies on anticipating viral mutations that are able to evade host immune responses to facilitate vaccine and therapeutic design. However, current strategies for viral evolution prediction are not available early in a pandemic-experimental approaches require host polyclonal antibodies to test against1-16, and existing computational methods draw heavily from current strain prevalence to make reliable predictions of variants of concern17-19. To address this, we developed EVEscape, a generalizable modular framework that combines fitness predictions from a deep learning model of historical sequences with biophysical and structural information. EVEscape quantifies the viral escape potential of mutations at scale and has the advantage of being applicable before surveillance sequencing, experimental scans or three-dimensional structures of antibody complexes are available. We demonstrate that EVEscape, trained on sequences available before 2020, is as accurate as high-throughput experimental scans at anticipating pandemic variation for SARS-CoV-2 and is generalizable to other viruses including influenza, HIV and understudied viruses with pandemic potential such as Lassa and Nipah. We provide continually revised escape scores for all current strains of SARS-CoV-2 and predict probable further mutations to forecast emerging strains as a tool for continuing vaccine development ( evescape.org ).

Viral Evolution Shaped by Host Proteostasis Networks

Understanding the factors that shape viral evolution is critical for developing effective antiviral strategies, accurately predicting viral evolution, and preventing pandemics. One fundamental determinant of viral evolution is the interplay between viral protein biophysics and the host machineries that regulate protein folding and quality control. Most adaptive mutations in viruses are biophysically deleterious, resulting in a viral protein product with folding defects. In cells, protein folding is assisted by a dynamic system of chaperones and quality control processes known as the proteostasis network. Host proteostasis networks can determine the fates of viral proteins with biophysical defects, either by assisting with folding or by targeting them for degradation. In this review, we discuss and analyze new discoveries revealing that host proteostasis factors can profoundly shape the sequence space accessible to evolving viral proteins. We also discuss the many opportunities for research progress proffered by the proteostasis perspective on viral evolution and adaptation.

The effect of single mutations in Zika virus envelope on escape from broadly neutralizing antibodies

Zika virus and dengue virus are co-circulating flaviviruses with a widespread endemic range. Eliciting broad and potent neutralizing antibodies is an attractive goal for developing a vaccine to simultaneously protect against these viruses. However, the capacity of viral mutations to confer escape from broadly neutralizing antibodies remains undescribed, due in part to limited throughput and scope of traditional approaches. Here, we use deep mutational scanning to map how all possible single amino acid mutations in Zika virus envelope protein affect neutralization by antibodies of varying breadth and potency. While all antibodies selected viral escape mutations, the mutations selected by broadly neutralizing antibodies conferred less escape relative to those selected by narrow, virus-specific antibodies. Surprisingly, even for broadly neutralizing antibodies with similar binding footprints, different single mutations led to escape, indicating distinct functional requirements for neutralization not captured by existing structures. Additionally, the antigenic effects of mutations selected by broadly neutralizing antibodies were conserved across divergent, albeit related, flaviviruses. Our approach identifies residues critical for antibody neutralization, thus comprehensively defining the as-yet-unknown functional epitopes of antibodies with clinical potential.

Structure-based network analysis predicts mutations associated with inherited retinal disease

With continued advances in gene sequencing technologies comes the need to develop better tools to understand which mutations cause disease. Here we validate structure-based network analysis (SBNA)1,2 in well-studied human proteins and report results of using SBNA to identify critical amino acids that may cause retinal disease if subject to missense mutation. We computed SBNA scores for genes with high-quality structural data, starting with validating the method using 4 well-studied human disease-associated proteins. We then analyzed 47 inherited retinal disease (IRD) genes. We compared SBNA scores to phenotype data from the ClinVar database and found a significant difference between benign and pathogenic mutations with respect to network score. Finally, we applied this approach to 65 patients at Massachusetts Eye and Ear (MEE) who were diagnosed with IRD but for whom no genetic cause was found. Multivariable logistic regression models built using SBNA scores for IRD-associated genes successfully predicted pathogenicity of novel mutations, allowing us to identify likely causative disease variants in 37 patients with IRD from our clinic. In conclusion, SBNA can be meaningfully applied to human proteins and may help predict mutations causative of IRD.

Genotype-phenotype landscapes for immune-pathogen coevolution

Our immune systems constantly coevolve with the pathogens that challenge them, as pathogens adapt to evade our defense responses, with our immune repertoires shifting in turn. These coevolutionary dynamics take place across a vast and high-dimensional landscape of potential pathogen and immune receptor sequence variants. Mapping the relationship between these genotypes and the phenotypes that determine immune-pathogen interactions is crucial for understanding, predicting, and controlling disease. Here, we review recent developments applying high-throughput methods to create large libraries of immune receptor and pathogen protein sequence variants and measure relevant phenotypes. We describe several approaches that probe different regions of the high-dimensional sequence space and comment on how combinations of these methods may offer novel insight into immune-pathogen coevolution.

satmut_utils: a simulation and variant calling package for multiplexed assays of variant effect

The impact of millions of individual genetic variants on molecular phenotypes in coding sequences remains unknown. Multiplexed assays of variant effect (MAVEs) are scalable methods to annotate relevant variants, but existing software lacks standardization, requires cumbersome configuration, and does not scale to large targets. We present satmut_utils as a flexible solution for simulation and variant quantification. We then benchmark MAVE software using simulated and real MAVE data. We finally determine mRNA abundance for thousands of cystathionine beta-synthase variants using two experimental methods. The satmut_utils package enables high-performance analysis of MAVEs and reveals the capability of variants to alter mRNA abundance.

An RNA-based system to study hepatitis B virus replication and evaluate antivirals

Hepatitis B virus (HBV) chronically infects an estimated 300 million people, and standard treatments are rarely curative. Infection increases the risk of liver cirrhosis and hepatocellular carcinoma, and consequently, nearly 1 million people die each year from chronic hepatitis B. Tools and approaches that bring insights into HBV biology and facilitate the discovery and evaluation of antiviral drugs are in demand. Here, we describe a method to initiate the replication of HBV, a DNA virus, using synthetic RNA. This approach eliminates contaminating background signals from input virus or plasmid DNA that plagues existing systems and can be used to study multiple stages of HBV replication. We further demonstrate that this method can be uniquely applied to identify sequence variants that confer resistance to antiviral drugs.

Maximal interferon induction by influenza lacking NS1 is infrequent owing to requirements for replication and export

Influenza A virus exhibits high rates of replicative failure due to a variety of genetic defects. Most influenza virions cannot, when acting as individual particles, complete the entire viral life cycle. Nevertheless influenza is incredibly successful in the suppression of innate immune detection and the production of interferons, remaining undetected in >99% of cells in tissue-culture models of infection. Notably, the same variation that leads to replication failure can, by chance, inactivate the major innate immune antagonist in influenza A virus, NS1. What explains the observed rarity of interferon production in spite of the frequent loss of this, critical, antagonist? By studying how genetic and phenotypic variation in a viral population lacking NS1 correlates with interferon production, we have built a model of the "worst-case" failure from an improved understanding of the steps at which NS1 acts in the viral life cycle to prevent the triggering of an innate immune response. In doing so, we find that NS1 prevents the detection of de novo innate immune ligands, defective viral genomes, and viral export from the nucleus, although only generation of de novo ligands appears absolutely required for enhanced detection of virus in the absence of NS1. Due to this, the highest frequency of interferon production we observe (97% of infected cells) requires a high level of replication in the presence of defective viral genomes with NS1 bearing an inactivating mutation that does not impact its partner encoded on the same segment, NEP. This is incredibly unlikely to occur given the standard variation found within a viral population, and would generally require direct, artificial, intervention to achieve at an appreciable rate. Thus from our study, we procure at least a partial explanation for the seeming contradiction between high rates of replicative failure and the rarity of the interferon response to influenza infection.

Mutational fitness landscape and drug resistance

Robust technology has been developed to systematically quantify fitness landscapes that provide valuable opportunities to improve our understanding of drug resistance and define new avenues to develop drugs with reduced resistance susceptibility. We outline the critical importance of drug resistance studies and the potential for fitness landscape approaches to contribute to this effort. We describe the major technical advancements in mutational scanning, which is the primary approach used to quantify protein fitness landscapes. There are many complex steps to consider in planning and executing mutational scanning projects including developing a selection scheme, generating mutant libraries, tracking the frequency of variants using next-generation sequencing, and processing and interpreting the data. Key experimental parameters impacting each of these steps are discussed to aid in planning fitness landscape studies. There is a strong need for improved understanding of drug resistance, and fitness landscapes provide a promising new approach.

Mutational fitness landscape of human influenza H3N2 neuraminidase

Influenza neuraminidase (NA) has received increasing attention as an effective vaccine target. However, its mutational tolerance is not well characterized. Here, the fitness effects of >6,000 mutations in human H3N2 NA are probed using deep mutational scanning. Our result shows that while its antigenic regions have high mutational tolerance, there are solvent-exposed regions with low mutational tolerance. We also find that protein stability is a major determinant of NA mutational fitness. The deep mutational scanning result correlates well with mutational fitness inferred from natural sequences using a protein language model, substantiating the relevance of our findings to the natural evolution of circulating strains. Additional analysis further suggests that human H3N2 NA is far from running out of mutations despite already evolving for >50 years. Overall, this study advances our understanding of the evolutionary potential of NA and the underlying biophysical constraints, which in turn provide insights into NA-based vaccine design.

Identification of ADAR1 p150 and p110 Associated Edit Sites

Adenosine deaminase acting on RNA 1 (ADAR1) catalyzes adenosine-to-inosine editing on double-stranded RNA molecules and is involved in regulating cellular responses to endogenous and exogenous RNA. ADAR1 is the primary A-to-I editor of RNA in humans, and the majority of edit sites are found in a class of short interspersed nuclear elements called Alu elements, many of which are located in introns and 3' untranslated regions. Two ADAR1 protein isoforms, p110 (110 kDa) and p150 (150 kDa), are known to be coupled in expression, and decoupling the expression of these isoforms has revealed that the p150 isoform edits a broader range of targets compared to p110. Numerous methods for identification of ADAR1-associated edits have been developed, and we present here a specific method for identification of edit sites associated with individual ADAR1 isoforms.

Scalable Functional Assays for the Interpretation of Human Genetic Variation

Scalable sequence-function studies have enabled the systematic analysis and cataloging of hundreds of thousands of coding and noncoding genetic variants in the human genome. This has improved clinical variant interpretation and provided insights into the molecular, biophysical, and cellular effects of genetic variants at an astonishing scale and resolution across the spectrum of allele frequencies. In this review, we explore current applications and prospects for the field and outline the principles underlying scalable functional assay design, with a focus on the study of single-nucleotide coding and noncoding variants.

The evolutionary potential of influenza A virus hemagglutinin is highly constrained by epistatic interactions with neuraminidase

Antigenic evolution of the influenza A virus (IAV) hemagglutinin (HA) gene limits efforts to effectively control the spread of the virus in the population. Efforts to understand the mechanisms governing HA antigenic evolution typically examine the HA gene in isolation. This can ignore the importance of balancing HA receptor binding activities with the receptor-destroying activities of the viral neuraminidase (NA) to maintain viral fitness. We hypothesize that the need to maintain functional balance with NA significantly constrains the evolutionary potential of the HA. We use deep mutational scanning and show that variation in NA activity significantly reshapes the HA fitness landscape by modulating the overall mutational robustness of HA. Consistent with this, we observe that different NA backgrounds support the emergence of distinct repertoires of HA escape variants under neutralizing antibody pressure. Our results reveal a critical role for intersegment epistasis in influencing the evolutionary potential of the HA gene.

Intragenic compensation through the lens of deep mutational scanning

A significant fraction of mutations in proteins are deleterious and result in adverse consequences for protein function, stability, or interaction with other molecules. Intragenic compensation is a specific case of positive epistasis when a neutral missense mutation cancels effect of a deleterious mutation in the same protein. Permissive compensatory mutations facilitate protein evolution, since without them all sequences would be extremely conserved. Understanding compensatory mechanisms is an important scientific challenge at the intersection of protein biophysics and evolution. In human genetics, intragenic compensatory interactions are important since they may result in variable penetrance of pathogenic mutations or fixation of pathogenic human alleles in orthologous proteins from related species. The latter phenomenon complicates computational and clinical inference of an allele's pathogenicity. Deep mutational scanning is a relatively new technique that enables experimental studies of functional effects of thousands of mutations in proteins. We review the important aspects of the field and discuss existing limitations of current datasets. We reviewed ten published DMS datasets with quantified functional effects of single and double mutations and described rates and patterns of intragenic compensation in eight of them.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12551-022-01005-w.

Predicting Permissive Mutations That Improve the Fitness of A(H1N1)pdm09 Viruses Bearing the H275Y Neuraminidase Substitution

Oseltamivir-resistant influenza viruses arise due to amino acid mutations in key residues of the viral neuraminidase (NA). These changes often come at a fitness cost; however, it is known that permissive mutations in the viral NA can overcome this cost. This result was observed in former seasonal A(H1N1) viruses in 2007 which expressed the H275Y substitution (N1 numbering) with no apparent fitness cost and lead to widespread oseltamivir resistance. Therefore, this study aims to predict permissive mutations that may similarly enable fit H275Y variants to arise in currently circulating A(H1N1)pdm09 viruses. The first approach in this study utilized in silico analyses to predict potentially permissive mutations. The second approach involved the generation of a virus library which encompassed all possible NA mutations while keeping H275Y fixed. Fit variants were then selected by serially passaging the virus library either through ferrets by transmission or passaging once in vitro. The fitness impact of selected substitutions was further evaluated experimentally. The computational approach predicted three candidate permissive NA mutations which, in combination with each other, restored the replicative fitness of an H275Y variant. The second approach identified a stringent bottleneck during transmission between ferrets; however, three further substitutions were identified which may improve transmissibility. A comparison of fit H275Y variants in vitro and in experimentally infected animals showed a statistically significant correlation in the variants that were positively selected. Overall, this study provides valuable tools and insights into potential permissive mutations that may facilitate the emergence of a fit H275Y A(H1N1)pdm09 variant.

IMPORTANCE Oseltamivir (Tamiflu) is the most widely used antiviral for the treatment of influenza infections. Therefore, resistance to oseltamivir is a public health concern. This study is important as it explores the different evolutionary pathways available to current circulating influenza viruses that may lead to widespread oseltamivir resistance. Specifically, this study develops valuable experimental and computational tools to evaluate the fitness landscape of circulating A(H1N1)pmd09 influenza viruses bearing the H275Y mutation. The H275Y substitution is most commonly reported to confer oseltamivir resistance but also leads to loss of virus replication and transmission fitness, which limits its spread. However, it is known from previous influenza seasons that influenza viruses can evolve to overcome this loss of fitness. Therefore, this study aims to prospectively predict how contemporary A(H1N1)pmd09 influenza viruses may evolve to overcome the fitness cost of bearing the H275Y NA substitution, which could result in widespread oseltamivir resistance.

Influenza A virus undergoes compartmentalized replication in vivo dominated by stochastic bottlenecks

Transmission of influenza A viruses (IAV) between hosts is subject to numerous physical and biological barriers that impose genetic bottlenecks, constraining viral diversity and adaptation. The bottlenecks within hosts and their potential impacts on evolutionary pathways taken during infection are poorly understood. To address this, we created highly diverse IAV libraries bearing molecular barcodes on two gene segments, enabling high-resolution tracking and quantification of unique virus lineages within hosts. Here we show that IAV infection in lungs is characterized by multiple within-host bottlenecks that result in “islands” of infection in lung lobes, each with genetically distinct populations. We perform site-specific inoculation of barcoded IAV in the upper respiratory tract of ferrets and track viral diversity as infection spreads to the trachea and lungs. We detect extensive compartmentalization of discrete populations within lung lobes. Bottleneck events and localized replication stochastically sample individual viruses from the upper respiratory tract or the trachea that become the dominant genotype in a particular lobe. These populations are shaped strongly by founder effects, with limited evidence for positive selection. The segregated sites of replication highlight the jackpot-style events that contribute to within-host influenza virus evolution and may account for low rates of intrahost adaptation.

Transmission of influenza A viruses (IAV) between hosts and replication within host impose genetic bottlenecks, constraining viral diversity and adaptation. Here, Amato et al. perform site-specific inoculation of barcoded IAV of ferrets and track viral diversity as infection spreads to the lower respiratory tract and conclude that narrow population bottlenecks are an important feature of the within-host infection dynamics.

The 2009 Pandemic H1N1 Hemagglutinin Stalk Remained Antigenically Stable after Circulating in Humans for a Decade

An H1N1 influenza virus caused a pandemic in 2009, and descendants of this virus continue to circulate seasonally in humans. Upon infection with the 2009 H1N1 pandemic strain (pH1N1), many humans produced antibodies against epitopes in the hemagglutinin (HA) stalk. HA stalk-focused antibody responses were common among pH1N1-infected individuals because HA stalk epitopes were conserved between the pH1N1 strain and previously circulating H1N1 strains. Here, we completed a series of experiments to determine if the pH1N1 HA stalk has acquired substitutions since 2009 that prevent the binding of human antibodies. We identified several amino acid substitutions that accrued in the pH1N1 HA stalk from 2009 to 2019. We completed enzyme-linked immunosorbent assays, absorption-based binding assays, and surface plasmon resonance experiments to determine if these substitutions affect antibody binding. Using sera collected from 230 humans (aged 21 to 80 years), we found that pH1N1 HA stalk substitutions that have emerged since 2009 do not affect antibody binding. Our data suggest that the HA stalk domain of pH1N1 viruses remained antigenically stable after circulating in humans for a decade. IMPORTANCE In 2009, a new pandemic H1N1 (pH1N1) virus began circulating in humans. Many individuals mounted hemagglutinin (HA) stalk-focused antibody responses upon infection with the 2009 pH1N1 strain, since the HA stalk of this virus was relatively conserved with other seasonal H1N1 strains. Here, we completed a series of studies to determine if the 2009 pH1N1 strain has undergone antigenic drift in the HA stalk domain over the past decade. We found that serum antibodies from 230 humans could not antigenically distinguish the 2009 and 2019 HA stalk. These data suggest that the HA stalk of pH1N1 has remained antigenically stable, despite the presence of high levels of HA stalk antibodies within the human population.

A single-shot adenoviral vaccine provides hemagglutinin stalk-mediated protection against heterosubtypic influenza challenge in mice

Conventional influenza vaccines fail to confer broad protection against diverse influenza A viruses with pandemic potential. Efforts to develop a universal influenza virus vaccine include refocusing immunity towards the highly conserved stalk domain of the influenza virus surface glycoprotein, hemagglutinin (HA). We constructed a non-replicating adenoviral (Ad) vector, encoding a secreted form of H1 HA, to evaluate HA stalk-focused immunity. The Ad5_H1 vaccine was tested in mice for its ability to elicit broad, cross-reactive protection against homologous, heterologous, and heterosubtypic lethal challenge in a single-shot immunization regimen. Ad5_H1 elicited hemagglutination inhibition (HI+) active antibodies (Abs), which conferred 100% sterilizing protection from homologous H1N1 challenge. Furthermore, Ad5_H1 rapidly induced H1-stalk-specific Abs with Fc-mediated effector function activity, in addition to stimulating both CD4+ and CD8+ stalk-specific T cell responses. This phenotype of immunity provided 100% protection from lethal challenge with a head-mismatched, reassortant influenza virus bearing a chimeric HA, cH6/1, in a stalk-mediated manner. Most importantly, 100% protection from mortality following lethal challenge with a heterosubtypic avian influenza virus, H5N1, was observed following a single immunization with Ad5_H1. In conclusion, Ad-based influenza vaccines can elicit significant breadth of protection in naive animals and could be considered for pandemic preparedness and stockpiling.

Evolutionary Landscapes of Host-Virus Arms Races

Vertebrate immune systems suppress viral infection using both innate restriction factors and adaptive immunity. Viruses mutate to escape these defenses, driving hosts to counterevolve to regain fitness. This cycle recurs repeatedly, resulting in an evolutionary arms race whose outcome depends on the pace and likelihood of adaptation by host and viral genes. Although viruses evolve faster than their vertebrate hosts, their proteins are subject to numerous functional constraints that impact the probability of adaptation. These constraints are globally defined by evolutionary landscapes, which describe the fitness and adaptive potential of all possible mutations. We review deep mutational scanning experiments mapping the evolutionary landscapes of both host and viral proteins engaged in arms races. For restriction factors and some broadly neutralizing antibodies, landscapes favor the host, which may help to level the evolutionary playing field against rapidly evolving viruses. We discuss the biophysical underpinnings of these landscapes and their therapeutic implications.

Limits to detecting epistasis in the fitness landscape of HIV

The rapid evolution of HIV is constrained by interactions between mutations which affect viral fitness. In this work, we explore the role of epistasis in determining the mutational fitness landscape of HIV for multiple drug target proteins, including Protease, Reverse Transcriptase, and Integrase. Epistatic interactions between residues modulate the mutation patterns involved in drug resistance, with unambiguous signatures of epistasis best seen in the comparison of the Potts model predicted and experimental HIV sequence “prevalences” expressed as higher-order marginals (beyond triplets) of the sequence probability distribution. In contrast, experimental measures of fitness such as viral replicative capacities generally probe fitness effects of point mutations in a single background, providing weak evidence for epistasis in viral systems. The detectable effects of epistasis are obscured by higher evolutionary conservation at sites. While double mutant cycles in principle, provide one of the best ways to probe epistatic interactions experimentally without reference to a particular background, we show that the analysis is complicated by the small dynamic range of measurements. Overall, we show that global pairwise interaction Potts models are necessary for predicting the mutational landscape of viral proteins.

The endoplasmic reticulum proteostasis network profoundly shapes the protein sequence space accessible to HIV envelope

The sequence space accessible to evolving proteins can be enhanced by cellular chaperones that assist biophysically defective clients in navigating complex folding landscapes. It is also possible, at least in theory, for proteostasis mechanisms that promote strict quality control to greatly constrain accessible protein sequence space. Unfortunately, most efforts to understand how proteostasis mechanisms influence evolution rely on artificial inhibition or genetic knockdown of specific chaperones. The few experiments that perturb quality control pathways also generally modulate the levels of only individual quality control factors. Here, we use chemical genetic strategies to tune proteostasis networks via natural stress response pathways that regulate the levels of entire suites of chaperones and quality control mechanisms. Specifically, we upregulate the unfolded protein response (UPR) to test the hypothesis that the host endoplasmic reticulum (ER) proteostasis network shapes the sequence space accessible to human immunodeficiency virus-1 (HIV-1) envelope (Env) protein. Elucidating factors that enhance or constrain Env sequence space is critical because Env evolves extremely rapidly, yielding HIV strains with antibody- and drug-escape mutations. We find that UPR-mediated upregulation of ER proteostasis factors, particularly those controlled by the IRE1-XBP1s UPR arm, globally reduces Env mutational tolerance. Conserved, functionally important Env regions exhibit the largest decreases in mutational tolerance upon XBP1s induction. Our data indicate that this phenomenon likely reflects strict quality control endowed by XBP1s-mediated remodeling of the ER proteostasis environment. Intriguingly, and in contrast, specific regions of Env, including regions targeted by broadly neutralizing antibodies, display enhanced mutational tolerance when XBP1s is induced, hinting at a role for host proteostasis network hijacking in potentiating antibody escape. These observations reveal a key function for proteostasis networks in decreasing instead of expanding the sequence space accessible to client proteins, while also demonstrating that the host ER proteostasis network profoundly shapes the mutational tolerance of Env in ways that could have important consequences for HIV adaptation.

The host cell’s endoplasmic reticulum proteostasis network has a profound, constraining impact on the protein sequence space accessible to HIV’s envelope protein, which is a major target of the host’s adaptive immune system; in particular, upregulation of stringent quality control pathways appears to restrict the viability of destabilizing envelope variants.

Broadly neutralizing antibodies target a haemagglutinin anchor epitope

Broadly neutralizing antibodies that target epitopes of haemagglutinin on the influenza virus have the potential to provide near universal protection against influenza virus infection1. However, viral mutants that escape broadly neutralizing antibodies have been reported2,3. The identification of broadly neutralizing antibody classes that can neutralize viral escape mutants is critical for universal influenza virus vaccine design. Here we report a distinct class of broadly neutralizing antibodies that target a discrete membrane-proximal anchor epitope of the haemagglutinin stalk domain. Anchor epitope-targeting antibodies are broadly neutralizing across H1 viruses and can cross-react with H2 and H5 viruses that are a pandemic threat. Antibodies that target this anchor epitope utilize a highly restricted repertoire, which encodes two public binding motifs that make extensive contacts with conserved residues in the fusion peptide. Moreover, anchor epitope-targeting B cells are common in the human memory B cell repertoire and were recalled in humans by an oil-in-water adjuvanted chimeric haemagglutinin vaccine4,5, which is a potential universal influenza virus vaccine. To maximize protection against seasonal and pandemic influenza viruses, vaccines should aim to boost this previously untapped source of broadly neutralizing antibodies that are widespread in the human memory B cell pool.

The PB1 gene from H9N2 avian influenza virus showed high compatibility and increased mutation rate after reassorting with a human H1N1 influenza virus

Background

Reassortment between human and avian influenza viruses (AIV) may result in novel viruses with new characteristics that may threaten human health when causing the next flu pandemic. A particular risk may be posed by avian influenza viruses of subtype H9N2 that are currently massively circulating in domestic poultry in Asia and have been shown to infect humans. In this study, we investigate the characteristics and compatibility of a human H1N1 virus with avian H9N2 derived genes.

Methods

The polymerase activity of the viral ribonucleoprotein (RNP) complex as combinations of polymerase-related gene segments derived from different reassortment events was tested in luciferase reporter assays. Reassortant viruses were generated by reverse genetics. Gene segments of the human WSN-H1N1 virus (A/WSN/1933) were replaced by gene segments of the avian A2093-H9N2 virus (A/chicken/Jiangsu/A2093/2011), which were both the Hemagglutinin (HA) and Neuraminidase (NA) gene segments in combination with one of the genes involved in the RNP complex (either PB2, PB1, PA or NP). The growth kinetics and virulence of reassortant viruses were tested on cell lines and mice. The reassortant viruses were then passaged for five generations in MDCK cells and mice lungs. The HA gene of progeny viruses from different passaging paths was analyzed using Next-Generation Sequencing (NGS).

Results

We discovered that the avian PB1 gene of H9N2 increased the polymerase activity of the RNP complex in backbone of H1N1. Reassortant viruses were able to replicate in MDCK and DF1 cells and mice. Analysis of the NGS data showed a higher substitution rate for the PB1-reassortant virus. In particular, for the PB1-reassortant virus, increased virulence for mice was measured by increased body weight loss after infection in mice.

Conclusions

The higher polymerase activity and increased mutation frequency measured for the PB1-reassortant virus suggests that the avian PB1 gene of H9N2 may drive the evolution and adaptation of reassortant viruses to the human host. This study provides novel insights in the characteristics of viruses that may arise by reassortment of human and avian influenza viruses. Surveillance for infections with H9N2 viruses and the emergence of the reassortant viruses in humans is important for pandemic preparedness.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12985-022-01745-x.

Predicting and interpreting large-scale mutagenesis data using analyses of protein stability and conservation

Understanding and predicting the functional consequences of single amino acid changes is central in many areas of protein science. Here, we collect and analyze experimental measurements of effects of >150,000 variants in 29 proteins. We use biophysical calculations to predict changes in stability for each variant and assess them in light of sequence conservation. We find that the sequence analyses give more accurate prediction of variant effects than predictions of stability and that about half of the variants that show loss of function do so due to stability effects. We construct a machine learning model to predict variant effects from protein structure and sequence alignments and show how the two sources of information support one another and enable mechanistic interpretations. Together, our results show how one can leverage large-scale experimental assessments of variant effects to gain deeper and general insights into the mechanisms that cause loss of function.

H3N2 Influenza Viruses with 12- or 16-Amino Acid Deletions in the Receptor-Binding Region of Their Hemagglutinin Protein

Human influenza viruses evade host immune responses by accumulating mutations around the receptor-binding region of the hemagglutinin (HA) protein, which is composed of three key elements, the 130-loop, the 190-helix, and the 220-loop. Here, we characterized two human H3N2 influenza viruses with 12- and 16-amino acid deletions around the HA receptor-binding site that were isolated after antigenic selection of mutated H3N2 viruses. Structural modeling suggested that the 12-amino acid deletion eliminated the 190-helix. The 16-amino acid deletion comprises two stretches of 11- and 5-amino acid deletions. As the result of a frameshift, “novel” amino acids (not found in wild-type HA at these positions) are encoded between the deleted regions. Interestingly, structural modeling predicted that the novel sequence forms a structure resembling the 190-helix. However, compared to wild-type HA, the 16-amino acid deletion mutant lacks two antiparallel beta-sheets that connect the 190-helix and the 220-loop in wild-type HA. Nonetheless, both HA deletion mutants replicated in mammalian cells, and the 16-amino acid deletion mutant (with a remodeled 190-helix) also replicated in Syrian hamsters, albeit at low titers. Wild-type virus bound preferentially to α2,6-linked sialic acids, whereas both mutants gained affinity for α2,3-linked sialic acids. Moreover, the 12- and 16-amino acid deletions may affect the antigenic properties of the viruses. Thus, viruses with sizeable deletions around the HA receptor-binding site are viable but may display altered sialic acid preferences, altered antigenic properties, and attenuated replicative ability in cultured cells and virulence in Syrian hamsters.

Library-based analysis reveals segment and length dependent characteristics of defective influenza genomes

Found in a diverse set of viral populations, defective interfering particles are parasitic variants that are unable to replicate on their own yet rise to relatively high frequencies. Their presence is associated with a loss of population fitness, both through the depletion of key cellular resources and the stimulation of innate immunity. For influenza A virus, these particles contain large internal deletions in the genomic segments which encode components of the heterotrimeric polymerase. Using a library-based approach, we comprehensively profile the growth and replication of defective influenza species, demonstrating that they possess an advantage during genome replication, and that exclusion during population expansion reshapes population composition in a manner consistent with their final, observed, distribution in natural populations. We find that an innate immune response is not linked to the size of a deletion; however, replication of defective segments can enhance their immunostimulatory properties. Overall, our results address several key questions in defective influenza A virus biology, and the methods we have developed to answer those questions may be broadly applied to other defective viruses.

Influenza Immunization in the Context of Preexisting Immunity

Although we develop influenza immunity from an early age, it is insufficient to prevent future infection with antigenically novel strains. One proposed way to generate long-term protective immunity against a broad range of influenza virus strains is to boost responses to the conserved epitopes on the hemagglutinin, the major surface glycoprotein on the influenza virus. Influenza-specific humoral immunity comprises a large fraction of the overall immune memory in humans, and it has been long recognized that preexisting immunity to influenza shapes the response to subsequent influenza infections and vaccinations. However, the mechanisms by which preexisting immunity modulates the response to influenza vaccination are still not completely understood. Using a set of mathematical models, we explore several hypotheses that may contribute to diminished boosting of antibodies to conserved epitopes after repeated vaccinations.

AMaLa: Analysis of Directed Evolution Experiments via Annealed Mutational Approximated Landscape

We present Annealed Mutational approximated Landscape (AMaLa), a new method to infer fitness landscapes from Directed Evolution experiments sequencing data. Such experiments typically start from a single wild-type sequence, which undergoes Darwinian in vitro evolution via multiple rounds of mutation and selection for a target phenotype. In the last years, Directed Evolution is emerging as a powerful instrument to probe fitness landscapes under controlled experimental conditions and as a relevant testing ground to develop accurate statistical models and inference algorithms (thanks to high-throughput screening and sequencing). Fitness landscape modeling either uses the enrichment of variants abundances as input, thus requiring the observation of the same variants at different rounds or assuming the last sequenced round as being sampled from an equilibrium distribution. AMaLa aims at effectively leveraging the information encoded in the whole time evolution. To do so, while assuming statistical sampling independence between sequenced rounds, the possible trajectories in sequence space are gauged with a time-dependent statistical weight consisting of two contributions: (i) an energy term accounting for the selection process and (ii) a generalized Jukes-Cantor model for the purely mutational step. This simple scheme enables accurately describing the Directed Evolution dynamics and inferring a fitness landscape that correctly reproduces the measures of the phenotype under selection (e.g., antibiotic drug resistance), notably outperforming widely used inference strategies. In addition, we assess the reliability of AMaLa by showing how the inferred statistical model could be used to predict relevant structural properties of the wild-type sequence.

Shifts in amino acid preferences as proteins evolve: A synthesis of experimental and theoretical work

Amino acid preferences vary across sites and time. While variation across sites is widely accepted, the extent and frequency of temporal shifts are contentious. Our understanding of the drivers of amino acid preference change is incomplete: To what extent are temporal shifts driven by adaptive versus nonadaptive evolutionary processes? We review phenomena that cause preferences to vary (e.g., evolutionary Stokes shift, contingency, and entrenchment) and clarify how they differ. To determine the extent and prevalence of shifted preferences, we review experimental and theoretical studies. Analyses of natural sequence alignments often detect decreases in homoplasy (convergence and reversions) rates, and variation in replacement rates with time-signals that are consistent with temporally changing preferences. While approaches inferring shifts in preferences from patterns in natural alignments are valuable, they are indirect since multiple mechanisms (both adaptive and nonadaptive) could lead to the observed signal. Alternatively, site-directed mutagenesis experiments allow for a more direct assessment of shifted preferences. They corroborate evidence from multiple sequence alignments, revealing that the preference for an amino acid at a site varies depending on the background sequence. However, shifts in preferences are usually minor in magnitude and sites with significantly shifted preferences are low in frequency. The small yet consistent perturbations in preferences could, nevertheless, jeopardize the accuracy of inference procedures, which assume constant preferences. We conclude by discussing if and how such shifts in preferences might influence widely used time-homogenous inference procedures and potential ways to mitigate such effects.

ECNet is an evolutionary context-integrated deep learning framework for protein engineering

Machine learning has been increasingly used for protein engineering. However, because the general sequence contexts they capture are not specific to the protein being engineered, the accuracy of existing machine learning algorithms is rather limited. Here, we report ECNet (evolutionary context-integrated neural network), a deep-learning algorithm that exploits evolutionary contexts to predict functional fitness for protein engineering. This algorithm integrates local evolutionary context from homologous sequences that explicitly model residue-residue epistasis for the protein of interest with the global evolutionary context that encodes rich semantic and structural features from the enormous protein sequence universe. As such, it enables accurate mapping from sequence to function and provides generalization from low-order mutants to higher-order mutants. We show that ECNet predicts the sequence-function relationship more accurately as compared to existing machine learning algorithms by using ~50 deep mutational scanning and random mutagenesis datasets. Moreover, we used ECNet to guide the engineering of TEM-1 β-lactamase and identified variants with improved ampicillin resistance with high success rates.

In Vivo Therapy with M2e-Specific IgG Selects for an Influenza A Virus Mutant with Delayed Matrix Protein 2 Expression

The ectodomain of matrix protein 2 (M2e) of influenza A viruses is a universal influenza A vaccine candidate. Here, we report potential evasion strategies of influenza A viruses under in vivo passive anti-M2e IgG immune selection pressure in severe combined immune-deficient (SCID) mice. A/Puerto Rico/8/34-infected SCID mice were treated with the M2e-specific mouse IgG monoclonal antibodies (MAbs) MAb 65 (IgG2a) or MAb 37 (IgG1), which recognize amino acids 5 to 15 in M2e, or with MAb 148 (IgG1), which binds to the invariant N terminus of M2e. Treatment of challenged SCID mice with any of these MAbs significantly prolonged survival compared to isotype control IgG treatment. Furthermore, M2e-specific IgG2a protected significantly better than IgG1, and even resulted in virus clearance in some of the SCID mice. Deep sequencing analysis of viral RNA isolated at different time points after treatment revealed that the sequence variation in M2e was limited to P10H/L and/or I11T in anti-M2e MAb-treated mice. Remarkably, in half of the samples isolated from moribund MAb 37-treated mice and in all MAb 148-treated mice, virus was isolated with a wild-type M2 sequence but with nonsynonymous mutations in the polymerases and/or the hemagglutinin genes. Some of these mutations were associated with delayed M2 and other viral gene expression and with increased resistance to anti-M2e MAb treatment of SCID mice. Treatment with M2e-specific MAbs thus selects for viruses with limited variation in M2e. Importantly, influenza A viruses may also undergo an alternative escape route by acquiring mutations that result in delayed wild-type M2 expression. IMPORTANCE Broadly protective influenza vaccine candidates may have a higher barrier to immune evasion compared to conventional influenza vaccines. We used Illumina MiSeq deep sequence analysis to study the mutational patterns in A/Puerto Rico/8/34 viruses that evolve in chronically infected SCID mice that were treated with different M2e-specific MAbs. We show that under these circumstances, viruses emerged in vivo with mutations in M2e that were limited to positions 10 and 11. Moreover, we discovered an alternative route for anti-M2e antibody immune escape, in which a virus is selected with wild-type M2e but with mutations in other gene segments that result in delayed M2 and other viral protein expression. Delayed expression of the viral antigen that is targeted by a protective antibody thus represents an influenza virus immune escape mechanism that does not involve epitope alterations.

Stabilization of the SARS-CoV-2 Spike Receptor-Binding Domain Using Deep Mutational Scanning and Structure-Based Design

The unprecedented global demand for SARS-CoV-2 vaccines has demonstrated the need for highly effective vaccine candidates that are thermostable and amenable to large-scale manufacturing. Nanoparticle immunogens presenting the receptor-binding domain (RBD) of the SARS-CoV-2 Spike protein (S) in repetitive arrays are being advanced as second-generation vaccine candidates, as they feature robust manufacturing characteristics and have shown promising immunogenicity in preclinical models. Here, we used previously reported deep mutational scanning (DMS) data to guide the design of stabilized variants of the RBD. The selected mutations fill a cavity in the RBD that has been identified as a linoleic acid binding pocket. Screening of several designs led to the selection of two lead candidates that expressed at higher yields than the wild-type RBD. These stabilized RBDs possess enhanced thermal stability and resistance to aggregation, particularly when incorporated into an icosahedral nanoparticle immunogen that maintained its integrity and antigenicity for 28 days at 35-40°C, while corresponding immunogens displaying the wild-type RBD experienced aggregation and loss of antigenicity. The stabilized immunogens preserved the potent immunogenicity of the original nanoparticle immunogen, which is currently being evaluated in a Phase I/II clinical trial. Our findings may improve the scalability and stability of RBD-based coronavirus vaccines in any format and more generally highlight the utility of comprehensive DMS data in guiding vaccine design.

Comprehensive Profiling of Mutations to Influenza Virus PB2 That Confer Resistance to the Cap-Binding Inhibitor Pimodivir

Antivirals are used not only in the current treatment of influenza but are also stockpiled as a first line of defense against novel influenza strains for which vaccines have yet to be developed. Identifying drug resistance mutations can guide the clinical deployment of the antiviral and can additionally define the mechanisms of drug action and drug resistance. Pimodivir is a first-in-class inhibitor of the polymerase basic protein 2 (PB2) subunit of the influenza A virus polymerase complex. A number of resistance mutations have previously been identified in treated patients or cell culture. Here, we generate a complete map of the effect of all single-amino-acid mutations to an avian PB2 on resistance to pimodivir. We identified both known and novel resistance mutations not only in the previously implicated cap-binding and mid-link domains, but also in the N-terminal domain. Our complete map of pimodivir resistance thus enables the evaluation of whether new viral strains contain mutations that will confer pimodivir resistance.