SeqFeatR for the Discovery of Feature-Sequence Associations

HAMdetector: A Bayesian regression model that integrates information to detect HLA-associated mutations

MOTIVATION

A key process in anti-viral adaptive immunity is that the Human Leukocyte Antigen system (HLA) presents epitopes as Major Histocompatibility Complex I (MHC I) protein-peptide complexes on cell surfaces and in this way alerts CD8+ cytotoxic T-Lymphocytes (CTLs). This pathway exerts strong selection pressure on viruses, favoring viral mutants that escape recognition by the HLA/CTL system. Naturally, such immune escape mutations often emerge in highly variable viruses, e.g. HIV or HBV, as HLA-associated mutations (HAMs), specific to the hosts MHC I proteins. The reliable identification of HAMs is not only important for understanding viral genomes and their evolution, but it also impacts the development of broadly effective anti-viral treatments and vaccines against variable viruses. By their very nature, HAMs are amenable to detection by statistical methods in paired sequence/HLA data. However, HLA alleles are very polymorphic in the human host population which makes the available data relatively sparse and noisy. Under these circumstances, one way to optimize HAM detection is to integrate all relevant information in a coherent model. Bayesian inference offers a principled approach to achieve this.

RESULTS

We present a new Bayesian regression model for the detection of HAMs that integrates a sparsity-inducing prior, epitope predictions, and phylogenetic bias assessment, and that yields easily interpretable quantitative information on HAM candidates. The model predicts experimentally confirmed HAMs as having high posterior probabilities, and it performs well in comparison to state-of-the-art models for several data sets from individuals infected with HBV, HDV, and HIV.

AVAILABILITY

The source code of this software is available at https://github.com/HAMdetector/Escape.jl under a permissive MIT license.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A single residue in influenza virus H2 hemagglutinin enhances the breadth of the B cell response elicited by H2 vaccination

Conserved epitopes on the influenza hemagglutinin (HA) stem are an attractive target for universal vaccine strategies as they elicit broadly neutralizing antibodies. Such antibody responses to stem-specific epitopes have been extensively characterized for HA subtypes H1 and H5 in humans. H2N2 influenza virus circulated 50 years ago and represents a pandemic threat due to the lack of widespread immunity, but, unlike H1 and H5, the H2 HA stem contains Phe45_HA2 predicted to sterically clash with HA stem-binding antibodies characterized to date. To understand the effect of Phe45_HA2, we compared the HA stem-specific B cell response in post hoc analyses of two phase 1 clinical trials, one testing vaccination with an H2 ferritin nanoparticle immunogen ( NCT03186781 ) and one with an inactivated H5N1 vaccine ( NCT01086657 ). In H2-naive individuals, the magnitude of the B cell response was equivalent, but H2-elicited HA stem-binding B cells displayed greater cross-reactivity than those elicited by H5. However, in individuals with childhood H2 exposure, H5-elicited HA stem-binding B cells also displayed high cross-reactivity, suggesting recall of memory B cells formed 50 years ago. Overall, we propose that a one-residue difference on an HA immunogen can alter establishment and expansion of broadly neutralizing memory B cells. These data have implications for stem-based universal influenza vaccination strategies.

Bioinformatics Research on Drug Sensitivity Prediction

Modeling-based anti-cancer drug sensitivity prediction has been extensively studied in recent years. While most drug sensitivity prediction models only use gene expression data, the remarkable impacts of gene mutation, methylation, and copy number variation on drug sensitivity are neglected. Drug sensitivity prediction can both help protect patients from some adverse drug reactions and improve the efficacy of treatment. Genomics data are extremely useful for drug sensitivity prediction task. This article reviews the role of drug sensitivity prediction, describes a variety of methods for predicting drug sensitivity. Moreover, the research significance of drug sensitivity prediction, as well as existing problems are well discussed.

Vaccination in a humanized mouse model elicits highly protective PfCSP-targeting anti-malarial antibodies

Repeat antigens, such as the Plasmodium falciparum circumsporozoite protein (PfCSP), use both sequence degeneracy and structural diversity to evade the immune response. A few PfCSP-directed antibodies have been identified that are effective at preventing malaria infection, including CIS43, but how these repeat-targeting antibodies might be improved has been unclear. Here, we engineered a humanized mouse model in which B cells expressed inferred human germline CIS43 (iGL-CIS43) B cell receptors and used both vaccination and bioinformatic analysis to obtain variant CIS43 antibodies with improved protective capacity. One such antibody, iGL-CIS43.D3, was significantly more potent than the current best-in-class PfCSP-directed antibody. We found that vaccination with a junctional epitope peptide was more effective than full-length PfCSP at recruiting iGL-CIS43 B cells to germinal centers. Structure-function analysis revealed multiple somatic hypermutations that combinatorically improved protection. This mouse model can thus be used to understand vaccine immunogens and to develop highly potent anti-malarial antibodies.

Mutations in Hepatitis D Virus Allow It to Escape Detection by CD8+ T Cells and Evolve at the Population Level

BACKGROUND & AIMS

Hepatitis D virus (HDV) superinfection in patients with hepatitis B virus (HBV) is associated with rapid progression to liver cirrhosis and hepatocellular carcinoma. Treatment options are limited, and no vaccine is available. Although HDV-specific CD8⁺ T cells are thought to control the virus, little is known about which HDV epitopes are targeted by virus-specific CD8⁺ T cells or why these cells ultimately fail to control the infection. We aimed to define how HDV escapes the CD8⁺ T-cell-mediated response.

METHODS

We collected plasma and DNA samples from 104 patients with chronic HDV and HBV infection at medical centers in Europe and the Middle East, sequenced HDV, typed human leukocyte antigen (HLA) class I alleles from patients, and searched for polymorphisms in HDV RNA associated with specific HLA class I alleles. We predicted epitopes in HDV that would be recognized by CD8⁺ T cells and corresponded with the identified virus polymorphisms in patients with resolved (n = 12) or chronic (n = 13) HDV infection.

RESULTS

We identified 21 polymorphisms in HDV that were significantly associated with specific HLA class I alleles (P < .005). Five of these polymorphisms were found to correspond to epitopes in HDV that are recognized by CD8⁺ T cells; we confirmed that CD8⁺ T cells in culture targeted these HDV epitopes. HDV variant peptides were only partially cross-recognized by CD8⁺ T cells isolated from patients, indicating that the virus had escaped detection by these cells. These newly identified HDV epitopes were restricted by relatively infrequent HLA class I alleles, and they bound most frequently to HLA-B. In contrast, frequent HLA class I alleles were not associated with HDV sequence polymorphisms.

CONCLUSIONS

We analyzed sequences of HDV RNA and HLA class I alleles that present epitope peptides to CD8⁺ T cells in patients with persistent HDV infection. We identified polymorphisms in the HDV proteome that associate with HLA class I alleles. Some variant peptides in epitopes from HDV were only partially recognized by CD8⁺ T cells isolated from patients; these could be mutations that allow HDV to escape the immune response, resulting in persistent infection. HDV escape from the immune response was associated with uncommon HLA class I alleles, indicating that HDV evolves, at the population level, to evade recognition by common HLA class I alleles.

Improvement of antibody functionality by structure-guided paratope engraftment

Broadly neutralizing antibodies (bNAbs) represent a promising alternative to antiretroviral drugs for HIV-1 prevention and treatment. Selected antibodies to the CD4-binding site bolster envelope trimer binding via quaternary contacts. Here, we rationally engraft a new paratope, i.e., the extended heavy-chain framework region 3 (FR3) loop of VRC03, which mediates quaternary interaction, onto several potent bNAbs, enabling them to reach an adjacent gp120 protomer. The interactive quaternary surface is delineated by solving the crystal structure of two FR3 loop-chimeric antibodies. Chimerization enhances the neutralizing activity of several potent bNAbs against a majority of global HIV-1 strains. Compared to unmodified antibodies, chimeric antibodies display lower autoreactivity and prolonged in vivo half-life in huFcRn mice and rhesus macaques. Thus, paratope engraftment may be used to expand the epitope repertory of natural antibodies, improving their functionality for disease prevention and treatment.

Quaternary contacts mediated by an extended heavy-chain framework region 3 (FR3) have been shown to improve binding to HIV envelope and virus neutralization for a few antibodies. Here, Liu et al. engraft such an FR3 loop onto several potent broadly neutralizing antibodies, resulting in improved neutralization activity and pharmacokinetics.

Differential escape of HCV from CD8+ T cell selection pressure between China and Germany depends on the presenting HLA class I molecule

Adaptation of hepatitis C virus (HCV) to CD8⁺ T cell selection pressure is well described; however, it is unclear if HCV differentially adapts in different populations. Here, we studied HLA class I-associated viral sequence polymorphisms in HCV 1b isolates in a Chinese population and compared viral substitution patterns between Chinese and German populations. We identified three HLA class I-restricted epitopes in HCV NS3 with statistical support for selection pressure and found evidence for differential escape pathways between isolates from China and Germany depending on the HLA class I molecule. The substitution patterns particularly differed in the epitope VTLTHPITK_1635-1643 , which was presented by HLA-A*03 as well as HLA-A*11, two alleles with highly different frequencies in the two populations. In Germany, a substitution in position seven of the epitope was the most frequent substitution in the presence of HLA-A*03, functionally associated with immune escape and nearly absent in Chinese isolates. In contrast, the most frequent substitution in China was located at position two of the epitope and became the predominant consensus residue. Moreover, substitutions in position one of the epitope were significantly enriched in HLA-A*11-positive individuals in China and associated with different patterns of CD8⁺ T cell reactivity. Our study confirms the differential escape pathways selected by HCV that depended on different HLA class I alleles in Chinese and German populations, indicating that HCV differentially adapts to distinct HLA class I alleles in these populations. This result has important implications for vaccine design against highly variable and globally distributed pathogens, which may require matching antigen sequences to geographic regions for T cell-based vaccine strategies.

Complete functional mapping of infection- and vaccine-elicited antibodies against the fusion peptide of HIV

Eliciting broadly neutralizing antibodies (bnAbs) targeting envelope (Env) is a major goal of HIV vaccine development, but cross-clade breadth from immunization has only sporadically been observed. Recently, Xu et al (2018) elicited cross-reactive neutralizing antibody responses in a variety of animal models using immunogens based on the epitope of bnAb VRC34.01. The VRC34.01 antibody, which was elicited by natural human infection, targets the N terminus of the Env fusion peptide, a critical component of the virus entry machinery. Here we precisely characterize the functional epitopes of VRC34.01 and two vaccine-elicited murine antibodies by mapping all single amino-acid mutations to the BG505 Env that affect viral neutralization. While escape from VRC34.01 occurred via mutations in both fusion peptide and distal interacting sites of the Env trimer, escape from the vaccine-elicited antibodies was mediated predominantly by mutations in the fusion peptide. Cryo-electron microscopy of four vaccine-elicited antibodies in complex with Env trimer revealed focused recognition of the fusion peptide and provided a structural basis for development of neutralization breadth. Together, these functional and structural data suggest that the breadth of vaccine-elicited antibodies targeting the fusion peptide can be enhanced by specific interactions with additional portions of Env. Thus, our complete maps of viral escape both delineate pathways of resistance to these fusion peptide-directed antibodies and provide a strategy to improve the breadth or potency of future vaccine-induced antibodies against Env’s fusion peptide.

A major goal of HIV-1 vaccine design is to elicit antibodies that neutralize diverse strains of HIV-1. Recently, some of us elicited such antibodies in animal models using immunogens based on the epitope of a broad antibody (VRC34.01) isolated from an infected individual. Further improving these vaccine-elicited antibody responses will require a detailed understanding of how the resulting antibodies target HIV’s envelope protein (Env). Here, we used mutational antigenic profiling to precisely map the epitopes of two vaccine-elicited antibodies and the template VRC34.01 antibody. We did this by quantifying the effect of all possible amino acid mutations to Env on antibody neutralization. Although all antibodies target a similar region of Env, we found clear differences in the functional interaction of Env with these vaccine- and infection-elicited antibodies. We combined these functional data with structural analyses to identify antibody–Env interactions that may contribute to the relatively greater breadth of the infection-elicited antibody and could improve the breadth of vaccine-elicited antibodies. These data thereby help to refine vaccination schemes to achieve broader responses.

Genotypic Prediction of Co-receptor Tropism of HIV-1 Subtypes A and C

Antiretroviral treatment of Human Immunodeficiency Virus type-1 (HIV-1) infections with CCR5-antagonists requires the co-receptor usage prediction of viral strains. Currently available tools are mostly designed based on subtype B strains and thus are in general not applicable to non-B subtypes. However, HIV-1 infections caused by subtype B only account for approximately 11% of infections worldwide. We evaluated the performance of several sequence-based algorithms for co-receptor usage prediction employed on subtype A V3 sequences including circulating recombinant forms (CRFs) and subtype C strains. We further analysed sequence profiles of gp120 regions of subtype A, B and C to explore functional relationships to entry phenotypes. Our analyses clearly demonstrate that state-of-the-art algorithms are not useful for predicting co-receptor tropism of subtype A and its CRFs. Sequence profile analysis of gp120 revealed molecular variability in subtype A viruses. Especially, the V2 loop region could be associated with co-receptor tropism, which might indicate a unique pattern that determines co-receptor tropism in subtype A strains compared to subtype B and C strains. Thus, our study demonstrates that there is a need for the development of novel algorithms facilitating tropism prediction of HIV-1 subtype A to improve effective antiretroviral treatment in patients.