Information

What does the term 'epitope mapping' mean?

What does the term 'epitope mapping' mean?


We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

Epitope mapping means identifying the binding site of antibodies on the target antigen. This means that the site to be identified is part of the antigen and not antibody, am I right?


Antibody Defined Linear Epitope Mapping

An epitope, or antigenic determinant, was defined as the site on an antigen at which an antibody binds, by virtue of the antibody's antigen-combining site (called the paratope) 1. The word epitope derives from the Greek epi, meaning "upon", and topos, or "place", and thus it is the place on the antigen upon which the antibody binds.

As antigens can be recognised by two distinct groups of receptor molecules of the immune system, namely antibodies (Ab's) or T-cell receptors (TCR's), we need to distinguish whether we are talking about epitopes defined by antibodies or by TCR's. This article will deal only with antibody-defined epitopes. See T-Cell Epitope Mapping for a guide to mapping of peptide epitopes defined by TCR's.


From genomics to epitope prediction

Immunoinformatics deals with the application of computational methods to immunological problems and is thus considered a part of bioinformatics. Historically, tools for the prediction of HLA-binding peptides were the first tools developed specifically for immunoinformatics applications (Box 1). These tools paved the way for more-complex applications. The development of immunoinformatics tools has been crucial to the availability of sufficient experimental data. High-throughput human leukocyte antigen (HLA) binding assays led to major progress in this area. More recently, next-generation sequencing (NGS) has facilitated many of the novel applications and challenges that we will review here. A first area where the availability of cost-effective sequencing is having a large impact is our knowledge of the major histocompatibility complex (MHC, HLA in human) itself. The number of known HLA alleles, as registered in the International ImMunoGeneTics information system (IMGT) database, has increased from 1000 in 1998 to more than 13,000 in 2015 [1]. Initially tools for prediction of HLA binding (often also — slightly inaccurately — called epitope prediction) were trained on data for each HLA allele independently, but the number of new alleles renders this approach more and more impractical. The development of novel predictors, so-called pan-specific binding predictors, has been necessitated by this development. In general, the availability of large-scale data has improved the performance of immunoinformatics tools, and, for many, although not for all, applications, there is now a wealth of data available. This increase in data volume often translates to an increased accuracy of these tools, primarily because many tools are based on machine learning methods, which profit greatly from additional data. In this context, the availability of comprehensive and well-curated immunological databases is essential.

Here, we will first review how immunoinformatics tools can be used to infer HLA allotypes from NGS data, and then we explain how HLA ligands can be predicted based on this information. There are fundamental differences between the prediction of HLA class I and class II ligands that we will also highlight. Specifically, for HLA class I, we will also discuss the tools available for the prediction of antigen processing [e.g., proteasomal cleavage and transport by transporter associated with antigen processing (TAP)] — although their impact in the field is limited compared with that of tools for HLA binding prediction. Despite all progress in immunoinformatics, prediction of T-cell reactivity, prediction of B-cell epitopes, and large-scale data integration are still major challenges, and we will briefly discuss why and how these could be overcome. Finally, we will consider how the availability of NGS-based data has not only improved the current immunoinformatics tools, but has also paved the way for novel applications of these tools. Most of these applications are centered around the paradigm of epitope-based vaccines. For example, epitope prediction tools can be applied to construct vaccines based only on the genomic sequence of a pathogen [2], and the availability of personal genomic data enables personalized approaches to cancer immunotherapy [3]. It is in these areas that we expect the combination of NGS data and novel computational tools to impact healthcare in a most profound way.


Material and methods

Epitope identification and analysis

Epitope mapping was carried out by means of a library of overlapping peptide fragments corresponding to the complete sequence of Leptospira interrogans serovar Copenhageni LipL32 (http://aeg.lbi.ic.unicamp.br/world/lic/). Peptides were simultaneously synthesized by the SPOT method ( Frank 1992 ) on derivatized cellulose membranes with an Ala-Ala linker, for the preparation of immobilized peptides. Assembly of the peptides was carried out utilizing Fmoc-chemistry essentially, as previously described ( Frank and Overwin 1996 ). The prepared membrane consists of overlapping pentadecapeptides spanning the complete sequence of LipL32 (residues 1–272), with an offset of three amino acid residues. The reactivity of the SPOT membrane was evaluated according to the protocol described by Soutullo et al. (2007) . Bound antibodies were detected with alkaline phosphatase (AP)-conjugated secondary antibody (anti-rabbit IgG Sigma) followed by a colour reaction with 5-bromo-4-choro-indolylphosphate (BCIP) and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide ( Frank 1992 ).

The epitope analysis was carried out with rabbit serum samples, a true negative sample and a pool of three hyperimmune samples against Leptospira Icterohaemorrhagiae serogroup, serovar Copenhageni (strains M20 and RGA) and serovar Icterohaemorrhagiae (strain Ictero no. 1).

After epitope identification, two peptides were chemically synthesized by 9-florenyl-methoxy-carbonyl chemistry and purified by HPLC as described previously ( Tonarelli et al. 2000 ). These two peptides (P1 and P2) were subsequently used as antigens in an ELISA test on human serum samples.

ELISA tests

Three serological tests were carried out in parallel using a blind design, two peptide-based (P1 and P2) and a classical ELISA test.

The protocol used in the peptide-based ELISA was previously described by Lottersberger et al. (2004) . Briefly, the peptides were adsorbed onto polystyrene plates (Costar EIA Microplates Corning Costor, Oneonta, NY) using 100 μl of a solution containing 20 μg ml −1 of peptide in carbonate/bicarbonate buffer (pH 9·6). Microplates were incubated for 60 min at 37°C with 100 μl per well of a 1 : 100 dilution of human sera, followed by incubation with peroxidase-conjugated anti-human IgG (Dako, Glostrup, Denmark) for 30 min at 37°C. The plates were developed by adding 100 μl of a solution containing 3,3′,5,5′-tetramethylbenzidine and H2O2 as substrate. After 15 min, the reaction was stopped by adding 100 μl of 2 N H2SO4 and read at 450/630 nm.

The classical ELISA test was genus specific (whole Leptospira antigen) ( Vanasco et al. 2007 ). Each sample was tested in duplicate, and an average optical density (AOD) for each was obtained. Results were expressed as checked optical density (COD), which was obtained by dividing the AOD by the optical density of the negative control pool of samples. Samples were considered positive when observance readings as a COD result were higher than the cut-off point. The optimal cut-off point was determined by means of a receiver operating characteristic curve using Medcalc ® ( Schoonjans et al. 1995 ).

Serum samples and case definition

One hundred and ten human serum samples of 50 confirmed and 60 unconfirmed cases of leptospirosis (in different stages of illness) were randomly selected from all samples referred to the National Leptospirosis Laboratory of the Instituto Nacional de Enfermedades Respiratorias (INER). Case definition was based on MAT results, clinical presentation, unspecific findings and epidemiological data ( Vanasco et al. 2007 ).


CD8 + T cells

The best understood CD8 + T cells are cytotoxic T lymphocytes (CTLs). They secrete molecules that destroy the cell to which they have bound.

Link to discussion.
This is a very useful function if the target cell is infected with a virus because the cell is usually destroyed before it can release a fresh crop of viruses able to infect other cells.

An example will show the beauty and biological efficiency of this mechanism.

Every time you get a virus infection, say influenza (flu), the virus invades certain cells of your body (in this case cells of the respiratory passages). Once inside, the virus subverts the metabolism of the cell to make more virus. This involves synthesizing a number of different macromolecules encoded by the viral genome.

In due course, these are assembled into a fresh crop of virus particles that leave the cell (often killing it in the process) and spread to new target cells.

Except while in transit from their old homes to their new, the viruses work inside of your cells safe from any antibodies that might be present in blood, lymph, and secretions.

But early in the process, infected cells display fragments of the viral proteins in their surface class I molecules. CTLs specific for that antigen will be able to bind to the infected cell and often will be able to destroy it before it can release a fresh crop of viruses.

In general, the role of the CD8 + T cells is to monitor all the cells of the body, ready to destroy any that express foreign antigen fragments in their class I molecules.


LANDscape MApping of Epitopes and T Cell Receptors for Selected Cancers (LANDMARC)

This is a correlative research project aimed at characterizing the T cell mediated immune responses to hepatocellular carcinoma (HCC), as well as Epstein-Barr virus (EBV)- and human papillomavirus (HPV)-related cancers. This study will enroll approximately 105 patients over 48 months. Of these 105 patients, 30 are EBV-related cancer, 45 are HPV-related cancer, and 30 are HCC. Patients will have blood samples collected one time to identify cancer specific T cells and T cell receptors in their blood. They will also have tissue samples collected one time to study the different types of immune cells, especially the T cells and their receptors.

The 105 patients enrolled in this study will be compared to retrospective samples (N=210 30 from EBV-related cancer cohort, 180 from HPV-related cancer cohort).

The cloning of genes encoding the T cell receptor (TCR), the identification of tumor-associated antigens and the subsequent characterization of the first HLA-restricted T cell-defined antigenic epitope, were key findings illustrating direct recognition of cancer cells by T cells. These discoveries provided a mechanistic foundation for ensuing work examining the dynamic nature of lymphocyte-dependent recognition and elimination of neoplastic cells. Furthermore, preclinical and clinical investigations illustrate an important role for T cell mediated anti-tumor immunity in human disease, and have characterized the complexity of cancer-associated immune responses that are not always sufficient for tumor elimination. Importantly however, therapeutic targeting of the immune system has demonstrated the power of immunomodulatory drugs for the restoration of anti-tumor immune responses for cancer treatment.

Presence of lymphocytes in a variety of human cancers is well documented, and T cells isolated from tumors that recognize cancer antigens can be harnessed for effective treatment. T cells isolated from patient tumors can be ex vivo expanded, and reinfused back into patients in a regime of cellular therapy termed adoptive cell transfer (ACT). This ACT therapy can be further directed to specific tumor antigens with the genetic manipulation of T cells to express TCR recognizing known p-HLA epitopes with dominant expression on cancer cells. However, the peptide-HLA (p-HLA) epitope landscape of tumor associated antigens, and their cognate TCR are not well described, and those which have been described are primarily limited to the class I HLA-A*02:01 allele dominant only in European-Caucasian populations. The purpose of this study is to further document the cancer epitope landscape and provide a comprehensive characterization of TCR specificities in a range of malignancies, for a wide variety of class I and class II HLA alleles. In addition, we aim to elucidate not only TCR repertoires important for anti-tumor immunity, but further clarify the role antigen presenting cells play in shaping these T cell repertoires.

The objectives involve the identification of cancer-associated/specific antigen p-HLA epitopes and their cognate TCR, and the subsequent structural and functional characterization these TCR. To meet this objective, immune cells of T, B and myeloid lineage will be analyzed. Phenotypic characterization of these cell subsets will be performed using standard immunological procedures such as immunofluorescence, immunohistochemistry, ELISA, ELISpot, qRT-PCR, analytic cytometry, CyTOF (cytometry time of flight), and in vitro stimulation. To relate immune responses to cancer cell intrinsic biology, RNA and DNA will be sequenced to identify cancer cell transcriptome and mutations, as well as T cell receptor unique sequences. Results from all laboratory analysis will be combined with relevant clinical data. Any confirmation of diagnosis, tissue types and other clinical data will be provided as available from the pathologists of the relevant disease site at UHN.

An incomplete understanding of T cell responses to cancer impedes the development of more effective immunotherapeutics. Discovery using tumor specimens from cancer patients will clarify how the complexity of the tumor environment shapes T cell specificity to induce effective immune responses and facilitate our development of better immune modulating therapeutics.

Layout table for study information
Study Type : Observational
Estimated Enrollment : 105 participants
Observational Model: Other
Time Perspective: Prospective
Official Title: LANDscape MApping of Epitopes and T Cell Receptors for Selected Cancers
Actual Study Start Date : November 30, 2020
Estimated Primary Completion Date : February 28, 2026
Estimated Study Completion Date : February 28, 2026


ELISA Guide

The enzyme-linked immunosorbent assay (ELISA) is one of the most commonly used labeled immunoassay techniques. It is based on an enzyme-labeled antibody capable of detecting an antigen immobilized to a solid surface, 96-well or 384-well polystyrene plates. A substrate is added to produce either a color change or light signal correlating to the amount of the antigen which presents in the original sample. It is a simple and rapid technique to detect antibodies or antigens attached to a solid surface. Being one of the most sensitive immunoassays, ELISA offers commercial value in laboratory research, diagnostic of disease biomarkers and quality control in various industries.

ELISA Format

According to the difference of the antigen immobilizing strategy, the antibody labeling strategy, and the type of antibody-antigen reaction (direct recognition or competition), ELISA can be presented in a variety of formats. Each has its own advantages and disadvantages. One can choose an optimal ELISA format flexibly according to the requirements.

This is the simplest form of ELISA (Figure 1). Here an antigen is attached passively to a plastic solid phase by a period of incubation. After a simple washing step, antigen is detected by the addition of an antibody that is linked covalently to an enzyme. After incubation and washing, the test is developed by the addition of a chromogen/substrate whereby enzyme activity produces a color change. Color development is read after a defined time or after enzyme activity is stopped by chemical means at a defined time. Color is read in a spectrophotometer.

Figure 1. The flowchart of direct ELISA.

Direct ELISA is useful for qualitative or quantitative antigen detection in a sample, antibody screening, and epitope mapping since only one antibody is involved. There is no secondary antibody with cross-reactivity and the assay can be performed in less amount of time. However, the Immunoreactivity of the primary antibody might be adversely affected by labeling with enzymes. The labeled primary antibody is not commonly used, so labeling primary antibodies for each specific ELISA system is necessary when use direct ELISA.

The indirect detection method adds a labeled secondary antibody for detection on the basis of direct ELISA and it is the most popular ELISA format. Antigen is passively attached to wells by incubation. After washing, antibodies specific for the antigen are incubated with the antigen. Wells are washed and all bound antibodies are detected by the addition of anti-species antibodies covalently linked to an enzyme. Such antibodies are specific for the species in which the first antibody added were produced. After incubation and washing, the test is developed and can be read as described in figure 2.

Figure 2. The flowchart of indirect ELISA

Similar to direct ELISA, indirect ELISA is useful for antibody screening, epitope mapping, and protein quantification. The secondary antibody serves to enhance the signal of the primary antibody, which makes it more sensitive than direct ELISA. However, it also produces a higher background signal and potentially decreases the overall signal.

Comparison of Direct and Indirect ELISA

Advantages Disadvantages
Direct ELISA ? Quick, only one antibody and fewer steps are used.
? No cross-reactivity of secondary antibody
? Immune reactivity of the primary antibody might be adversely affected by labeling.
? No flexibility in choice of primary antibody label from one experiment to another.
? Minimal signal amplification.
Indirect ELISA ? Versatile because many primary antibodies can be made in one species and the same labeled secondary antibody can be used for detection.
? Maximum immune reactivity of the primary antibody is retained because it is not labeled.
? Sensitivity is increased because each primary antibody contains several epitopes that can be bound by the labeled secondary antibody, allowing for signal amplification.
? Cross-reactivity might occur with the secondary antibody, resulting in nonspecific signal.
? An extra incubation step is required in the procedure.

The sandwich ELISA is one of the most useful immunoassay formats and it is designed for detection of soluble antigens. There are two forms of this ELISA depending on the number of antibodies used. The principle is the same for both instead of adding antigen directly to a solid phase, a capture antibody is immobilized to the solid phase to capture antigen.

For direct sandwich ELISA (figure 3a), capture antibody is attached on the solid phase. After washing away excess unbound antibody, antigen is added and is specifically captured. The antigen is then detected by a second enzyme labeled antibody directly against the antigen. This type of assay is useful where a single species antiserum is available and where antigen does not attach well to plates.

For indirect sandwich ELISA (figure 3b), the antigen is detected with a second unlabeled antibody. This antibody is in turn detected using an anti-species enzyme labeled conjugate. It is essential that the anti-species conjugate does not bind to the capture antibody, therefore the species in which the capture antibody is produced must be different. The same considerations about the need for that at least two antigenic sites allowing the “sandwich” are relevant. The advantage of this system is that a single anti-species conjugate can be used to evaluate the binding of antibodies from any number of samples.

Figure 3. The flowchart of direct sandwich ELISA (a) and indirect sandwich ELISA (b)

These systems are useful when antigens are in a crude form (contaminated with other proteins) or at low concentration. In these cases the antigen cannot be attached directly to the solid phase at a high enough concentration to allow successful assay based on direct or indirect ELISAs. The sandwich ELISAs depend on antigens having at least two antigenic sites so that at least two antibody populations can bind.

The systems described above are the basic configurations of ELISA. All of these can be adapted to measure antigens or antibodies using competitive or inhibition conditions as described in figure 4.

Each assay described above requires pre-reaction of reagents to obtain optimal conditions. These optimal conditions are then challenged either by the addition of antigen (Figure 4a) or antibody (Figure 4b). As the amount of free antigen (antibody) in solution increases, the amount of antibody (antigen) that will bind to the immobilized substrate decreases. After washing step, chromophore substrate is added to generate signal (color change or light). The signal change caused by challenging with antibody/antigen reveals the information about the competitive antigen/antibody.

Figure 4. The flowchart of antigen competition ELISA (a) and antibody competition ELISA (b)

Competition ELISAs are particularly useful for measurements of antigen concentration in complex mixtures when the unknown samples that may contain antigen are compared to similar samples that contain known amounts of purified antigen.

General Protocols for 3 Common ELISA Formats.

In each case, the precise conditions should be optimized for a particular assay.


Fundamentals and Methods for T- and B-Cell Epitope Prediction

Adaptive immunity is mediated by T- and B-cells, which are immune cells capable of developing pathogen-specific memory that confers immunological protection. Memory and effector functions of B- and T-cells are predicated on the recognition through specialized receptors of specific targets (antigens) in pathogens. More specifically, B- and T-cells recognize portions within their cognate antigens known as epitopes. There is great interest in identifying epitopes in antigens for a number of practical reasons, including understanding disease etiology, immune monitoring, developing diagnosis assays, and designing epitope-based vaccines. Epitope identification is costly and time-consuming as it requires experimental screening of large arrays of potential epitope candidates. Fortunately, researchers have developed in silico prediction methods that dramatically reduce the burden associated with epitope mapping by decreasing the list of potential epitope candidates for experimental testing. Here, we analyze aspects of antigen recognition by T- and B-cells that are relevant for epitope prediction. Subsequently, we provide a systematic and inclusive review of the most relevant B- and T-cell epitope prediction methods and tools, paying particular attention to their foundations.

1. Introduction

The immune system is typically divided into two categories, innate and adaptive. Innate immunity involves nonspecific defense mechanisms that act immediately or within hours after a microbe appearance in the body. All multicellular beings exhibit some kind of innate immunity. In contrast, adaptive immunity is only present in vertebrates and it is highly specific. In fact, the adaptive immune system is able to recognize and destroy invading pathogens individually. Moreover, the adaptive immune system remembers the pathogens that fights, acquiring a pathogen-specific long-lasting protective memory that enables stronger attacks each time the pathogen is reencountered [1]. Nonetheless, innate and adaptive immune mechanisms work together and adaptive immunity elicitation is contingent on prior activation of innate immune responses [1].

Adaptive immunity is articulated by lymphocytes, more specifically by B- and T-cells, which are responsible for the humoral and cell-mediated immunity. B- and T-cells do not recognize pathogens as a whole, but molecular components known as antigens. These antigens are recognized by specific receptors present in the cell surface of B- and T-cells. Antigen recognition by these receptors is required to activate B- and T-cells but not enough, as second activation signals stemming from the activation of the innate immune system are also needed. The specificity of the recognition is determined by genetic recombination events that occur during lymphocyte development, which lead to generating millions of different variants of lymphocytes in terms of the antigen-recognizing receptors [1]. Antigen recognition by B- and T-cells differ greatly.

B-cells recognize solvent-exposed antigens through antigen receptors, named as B-cell receptors (BCR), consisting of membrane-bound immunoglobulins, as shown in Figure 1. Upon activation, B-cells differentiate and secrete soluble forms of the immunoglobulins, also known as antibodies, which mediate humoral adaptive immunity. Antibodies released by B-cells can have different functions that are triggered upon binding their cognate antigens. These functions include neutralizing toxins and pathogens and labeling them for destruction [1].

A B-cell epitope is the antigen portion binding to the immunoglobulin or antibody. These epitopes recognized by B-cells may constitute any exposed solvent region in the antigen and can be of different chemical nature. However, most antigens are proteins and those are the subjects for epitope prediction methods.

On the other hand, T-cells present on their surface a specific receptor known as T-cell receptor (TCR) that enables the recognition of antigens when they are displayed on the surface of antigen-presenting cells (APCs) bound to major histocompatibility complex (MHC) molecules. T-cell epitopes are presented by class I (MHC I) and II (MHC II) MHC molecules that are recognized by two distinct subsets of T-cells, CD8 and CD4 T-cells, respectively (Figure 2). Subsequently, there are CD8 and CD4 T-cell epitopes. CD8 T-cells become cytotoxic T lymphocytes (CTL) following T CD8 epitope recognition. Meanwhile, primed CD4 T-cells become helper (Th) or regulatory (Treg) T-cells [1]. Th cells amplify the immune response, and there are three main subclasses: Th1 (cell-mediated immunity against intracellular pathogens), Th2 (antibody-mediated immunity), and Th17 (inflammatory response and defense against extracellular bacteria) [2].

Identifying epitopes in antigens is of great interest for a number of practical reasons, including understanding disease etiology, immune monitoring, developing diagnosis assays, and designing epitope-based vaccines. B-cell epitopes can be identified by different methods including solving the 3D structure of antigen-antibody complexes, peptide library screening of antibody binding or performing functional assays in which the antigen is mutated and the interaction antibody-antigen is evaluated [3, 4]. On the other hand, experimental determination of T-cell epitopes is carried out using MHC multimers and lymphoproliferation or ELISPOT assays, among others [5, 6]. Traditional epitope identification has depended entirely upon experimental techniques, being costly and time-consuming. Thereby, scientists have developed and implemented epitope prediction methods that facilitate epitope identification and decrease the experimental load associated with it. Here, we will first analyze aspects of antigen recognition by T- and B-cells that are relevant for a better understanding of the topic of epitope prediction. Subsequently, we will provide a systematic and inclusive review of the most important prediction methods and tools, paying particular attention to their foundations and potentials. We will also discuss epitope prediction limitations and ways to overcome them. We will start with T-cell epitopes.

2. T-Cell Epitope Prediction

T-cell epitope prediction aims to identify the shortest peptides within an antigen that are able to stimulate either CD4 or CD8 T-cells [7]. This capacity to stimulate T-cells is called immunogenicity, and it is confirmed in assays requiring synthetic peptides derived from antigens [5, 6]. There are many distinct peptides within antigens and T-cell prediction methods aim to identify those that are immunogenic. T-cell epitope immunogenicity is contingent on three basic steps: (i) antigen processing, (ii) peptide binding to MHC molecules, and (iii) recognition by a cognate TCR. Of these three events, MHC-peptide binding is the most selective one at determining T-cell epitopes [8, 9]. Therefore, prediction of peptide-MHC binding is the main basis to anticipate T-cell epitopes and we will review it next.

2.1. Prediction of Peptide-MHC Binding

MHC I and MHC II molecules have similar 3D-structures with bound peptides sitting in a groove delineated by two α-helices overlying a floor comprised of eight antiparallel β-strands. However, there are also key differences between MHC I and II binding grooves that we must highlight for they condition peptide-binding predictions (Figure 3). The peptide-binding cleft of MHC I molecules is closed as it is made by a single α chain. As a result, MHC I molecules can only bind short peptides ranging from 9 to 11 amino acids, whose N- and C-terminal ends remain pinned to conserved residues of the MHC I molecule through a network of hydrogen bonds [10, 11]. The MHC I peptide-binding groove also contains deep binding pockets with tight physicochemical preferences that facilitate binding predictions. There is a complication however. Peptides that have different sizes and bind to the same MHC I molecule often use alternative binding pockets [12]. Therefore, methods predicting peptide-MHC I binding require a fixed peptide length. However, since most MHC I peptide ligands have 9 residues, it is generally preferable to predict peptides with that size. In contrast, the peptide-binding groove of MHC II molecules is open, allowing the N- and C-terminal ends of a peptide to extend beyond the binding groove [10, 11]. As a result, MHC II-bound peptides vary widely in length (9–22 residues), although only a core of nine residues (peptide-binding core) sits into the MHC II binding groove. Therefore, peptide-MHC II binding prediction methods often target to identify these peptide-binding cores. MHC II molecule binding pockets are also shallower and less demanding than those of MHC I molecules. As a consequence, peptide-binding prediction to MHC II molecules is less accurate than that of MHC I molecules.

Given the relevance of the problem, there are numerous methods to predict peptide-MHC binding. The most relevant with free online use are collected on Table 1. They can be divided in two main categories: data-driven and structure-based methods. Structure-based approaches generally rely on modeling the peptide-MHC structure followed by evaluation of the interaction through methods such as molecular dynamic simulations [8, 13]. Structure-based methods have the great advantage of not needing experimental data. However, they are seldom used as they are computationally intensive and exhibit lower predictive performance than data-driven methods [14].

Data-driven methods for peptide-MHC binding prediction are based on peptide sequences that are known to bind to MHC molecules. These peptide sequences are generally available in specialized epitope databases such as IEDB [15], EPIMHC [16], Antijen [17, 18]. Both MHC I and II binding peptides contain frequently occurring amino acids at particular peptide positions, known as anchor residues. Thereby, prediction of peptide-MHC binding was first approached using sequence motif (SM) reflecting amino acid preferences of MHC molecules at anchor positions [19]. However, it was soon shown that nonanchor residues also contribute to the capacity of a peptide to bind to a given MHC molecule [20, 21]. Subsequently, researchers developed motif matrices (MM), which could evaluate the contribution of each and all peptide positions to the binding with the MHC molecule [22–25]. The most sophisticated form of motif matrices consists of profiles [24–26] that are similar to those used for detecting sequence homology [27]. We would like to remark that motif matrices are often mistaken with quantitative affinity matrices (QAMs) since both produce peptide scores. However, MMs are derived without taking in consideration values of binding affinities and, therefore, resulting peptide scores are not suited to address binding affinity. In contrast, QAMs are trained on peptides and corresponding binding affinities, and aim to predict binding affinity. The first method based on QAMs was developed by Parker et al. [28] (Table 1). Subsequently, various approaches were developed to obtain QAMs from peptide affinity data and predict peptide binding to MHC I and II molecules [29–32].

QAMs and motif matrices assume an independent contribution of peptide side chains to the binding. This assumption is well supported by experimental data but there is also evidence that neighboring peptide residues interfere with others [33]. To account for those interferences, researchers introduced quantitative structure-activity relationship (QSAR) additive models wherein the binding affinity of peptides to MHC is computed as the sum of amino acid contributions at each position plus the contribution of adjacent side chain interactions [34]. However, machine learning (ML) is the most popular and robust approach introduced to deal with the nonlinearity of peptide-MHC binding data [8]. Researchers have used ML for two distinct problems: the discrimination of MHC binders from nonbinders and the prediction of binding affinity of peptides to MHC molecules.

For developing discrimination models, ML algorithms are trained on data sets consisting of peptides that either bind or do not bind to MHC molecules. Relevant examples of ML-based discrimination models are those based on artificial neural networks (ANNs) [35, 36], support vector machines (SVMs) [37–39], decision trees (DTs) [40, 41], and Hidden Markov models (HMMs), which can also cope with nonlinear data and have been used to discriminate peptides binding to MHC molecules. However, unlike other ML algorithms, they have to be trained only on positive data. Three types of HMMs have been used to predict MHC-peptide binding: fully connected HMMs [42], structure-optimized HMMs [43], and profile HMMs [43, 44]. Of these, only fully connected HMMs (fcHMMs) and structure-optimized HMMs (soHMMs) can recognize different patterns in the peptide binders. In fact, profile HMMs that are derived from sets of ungapped alignments (the case for peptides binding to MHC) are nearly identical to profile matrices [45] (Table 1).

With regard to predicting binding affinity, ML algorithms are trained on datasets consisting of peptides with known affinity to MHC molecules. Both SVMs and ANNs have been used for such purpose. SVMs were first applied to predict peptide-binding affinity to MHC I molecules [46] and later to MHC II molecules [47] (Table 1). Likewise, ANNs were also applied first to the prediction of peptide binding to MHC I [48, 49] and later to MHC II molecules [50] (Table 1). Benchmarking of peptide-MHC binding prediction methods appears to indicate that those based on ANNs are superior to those based on QAMs and MMs. However, the differences between the distinct methods are marginal and vary for different MHC molecules [51]. Moreover, it has been shown that the performance of peptide-MHC predictions is improved by combining several methods and providing consensus predictions [52].

A major complication for predicting T-cell epitopes through peptide-MHC binding models is MHC polymorphism. In humans, MHC molecules are known as human leukocyte antigens (HLAs), and there are hundreds of allelic variants of class I (HLA I) and class II (HLA II) molecules. These HLA allelic variants bind distinct sets of peptides [53] and require specific models for predicting peptide-MHC binding. However, peptide-binding data is only available for a minority of HLA molecules. To overcome this limitation, some researchers have developed pan-MHC-specific methods by training ANNs on input data combining MHC residues that contact the peptide with peptide-binding affinity that are capable of predicting peptide-binding affinities to uncharacterized HLA alleles [54, 55].

HLA polymorphism also hampers the development of worldwide covering T-cell epitope-based vaccines as HLA variants are expressed at vastly variable frequencies in different ethnic groups [56]. Interestingly, different HLA molecules can also bind similar sets of peptides [57, 58] and researchers have devised methods to cluster them in groups, known as HLA supertypes, consisting of HLA alleles with similar peptide-binding specificities [59–61]. The HLA-A2, HLA-A3, and HLA-B7 are relevant examples of supertypes 88% of the population expresses at least an allele included in these supertypes [25, 57, 58]. Identification of promiscuous peptide-binding to HLA supertypes enables the development of T-cell epitope vaccines with high-population coverage using a limited number of peptides. Currently, several web-based methods allow the prediction of promiscuous peptide-binding to HLA supertypes for epitope vaccine design including MULTIPRED [62] and PEPVAC [63] (Table 1). A method to identify promiscuous peptide-binding beyond HLA supertypes was developed and implemented by Molero-Abraham et al. [64] with the name of EPISOPT. EPISOPT predicts HLA I presentation profiles of individual peptides regardless of supertypes and identifies epitope combinations providing a wider population protection coverage.

Prediction of peptide binding to MHC II molecules readily discriminate CD4 T-cell epitopes, but cannot tell their ability to activate the response of specific CD4 T-cell subsets (e.g., Th1, Th2, and Treg). However, there is evidence that some CD4 T-cell epitopes appear to stimulate specific subsets of Th cells [65, 66]. Distinguishing the ability of MHC II-restricted epitopes to elicit distinct responses is clearly relevant for epitope vaccine development and has prompted researchers’ attention. A relevant example is the work by Dhanda et al. [67] who generated classifiers capable of predicting potential peptide inducers of interleukin 4 (IL-4) secretion, typical of Th2 cells, by training SVM models on experimentally validated IL4 inducing and noninducing MHC class II binders (Table 1).

2.2. Prediction of Antigen Processing and Integration with Peptide-MHC Binding Prediction

Antigen processing shapes the peptide repertoire available for MHC binding and is a limiting step determining T-cell epitope immunogenicity [68]. Subsequently, computational modeling of the antigen processing pathway provides a mean to enhance T-cell epitope predictions. Antigen presentation by MHC I and II molecules proceed by two different pathways. MHC II molecules present peptide antigens derived from endocyted antigens that are degraded and loaded onto the MHC II molecule in endosomal compartments [69]. Class II antigen degradation is poorly understood, and there is lack of good prediction algorithms yet [70]. In contrast, MHC I molecules present peptides derived mainly from antigens degraded in the cytosol. The resulting peptide antigens are then transported to the endoplasmic reticulum by TAP where they are loaded onto nascent MHC I molecules [69] (Figure 4). Prior to loading, peptides often undergo trimming by ERAAP N-terminal amino peptidases [71].

Proteasomal cleavage and peptide-binding to TAP have been studied in detail and there are computational methods that predict both processes. Proteasomal cleavage prediction models have been derived from peptide fragments generated in vitro by human constitutive proteasomes [72, 73] and from sets of MHC I-restricted ligands mapped onto their source proteins [74–76]. On the other hand, TAP binding prediction methods have been developed by training different algorithms on peptides of known affinity to TAP [77–80]. Combination of proteasomal cleavage and peptide-binding to TAP with peptide-MHC binding predictions increases T-cell epitope predictive rate in comparison to just peptide-binding to MHC I [37, 77, 81–83]. Subsequently, researchers have developed resources to predict CD8 T-cell epitopes through multistep approaches integrating proteasomal cleavage, TAP transport, and peptide-binding to MHC molecules [26, 37, 82–85] (Table 1).

3. Prediction of B-Cell Epitopes

B-cell epitope prediction aims to facilitate B-cell epitope identification with the practical purpose of replacing the antigen for antibody production or for carrying structure-function studies. Any solvent-exposed region in the antigen can be subject of recognition by antibodies. Nonetheless, B-cell epitopes can be divided in two main groups: linear and conformational (Figure 5). Linear B-cell epitopes consist of sequential residues, peptides, whereas conformational B-cell epitopes consist of patches of solvent-exposed atoms from residues that are not necessarily sequential (Figure 5). Therefore, linear and conformational B-cell epitopes are also known as continuous and discontinuous B-cell epitopes, respectively. Antibodies recognizing linear B-cell epitopes can recognize denatured antigens, while denaturing the antigen results in loss of recognition for conformational B-cell epitopes. Most B-cell epitopes (approximately a 90%) are conformational and, in fact, only a minority of native antigens contains linear B-cell epitopes [3]. We will review both, prediction of linear and conformational B-cell epitopes.

3.1. Prediction of Linear B-Cell Epitopes

Linear B-cell epitopes consist of peptides which can readily be used to replace antigens for immunizations and antibody production. Therefore, despite being a minority, prediction of linear B-cell epitopes have received major attention. Linear B-cell epitopes are predicted from the primary sequence of antigens using sequence-based methods. Early computational methods for the prediction of B-cell epitopes were based on simple amino acid propensity scales depicting physicochemical features of B-cellepitopes. For example, Hopp and Wood applied residue hydrophilicity calculations for B-cell epitope prediction [96, 97] on the assumption that hydrophilic regions are predominantly located on the protein surface and are potentially antigenic. We know now, however, that protein surfaces contain roughly the same number of hydrophilic and hydrophobic residues [98]. Other amino acid propensity scales introduced for B-cell epitope prediction are based on flexibility [99], surface accessibility [100], and β-turn propensity [101]. Current available bioinformatics tools to predict linear B-cell epitopes using propensity scales include PREDITOP [102] and PEOPLE [103] (Table 2). PREDITOP [102] uses a multiparametric algorithm based on hydrophilicity, accessibility, flexibility, and secondary structure properties of the amino acids. PEOPLE [103] uses the same parameters and in addition includes the assessment of β-turns. A related method to predict B-cell epitopes was introduced by Kolaskar and Tongaonkar [104], consisting on a simple antigenicity scale derived from physicochemical properties and frequencies of amino acids in experimentally determined B-cell epitopes. This index is perhaps the most popular antigenic scale for B-cell epitope prediction, and it is actually implemented by GCG [105] and EMBOSS [106] packages. Comparative evaluations of propensity scales carried out in a dataset of 85 linear B-cell epitopes showed that most propensity scales predicted between 50 and 70% of B-cell epitopes, with the β-turn scale reaching the best values [101, 107]. It has also been shown that combining the different scales does not appear to improve predictions [102, 108]. Moreover, Blythe and Flower [109] demonstrated that single-scale amino acid propensity scales are not reliable to predict epitope location.

The poor performance of amino acid scales for the prediction of linear B-cell epitopes prompted the introduction of machine learning- (ML-) based methods (Table 2). These methods are developed by training ML algorithms to distinguish experimental B-cell epitopes from non-B-cell epitopes. Prior to training, B-cell epitopes are translated into feature vectors capturing selected properties, such as those given by different propensity scales. Relevant examples of B-cell epitope prediction methods based on ML include BepiPred [110], ABCpred [111], LBtope [112], BCPREDS [113], and SVMtrip [114]. Datasets, training features, and algorithms used for developing these methods differ. BepiPred is based on random forests trained on B-cell epitopes obtained from 3D-structures of antigen-antibody complexes [110]. Both BCPREDS [113] and SVMtrip [114] are based on support vector machines (SVM) but while BCPREDS was trained using various string kernels that eliminate the need for representing the sequence into length-fixed feature vectors, SMVtrip was trained on length-fixed tripeptide composition vectors. ABCpred and LBtope methods consist on artificial neural networks (ANNs) trained on similar positive data, B-cell epitopes, but differ on negative data, non-B-cell epitopes. Negative data used for training ABCpred consisted on random peptides while negative data used for LBtope was based on experimentally validated non-B-cell epitopes form IEDB [15]. In general, B-cell epitope prediction methods employing ML-algorithm are reported to outperform those based on amino acid propensity scales. Nevertheless, some authors have reported that ML algorithms show little improvement over single-scale-based methods [115].

Antibodies elicited in the course of an immune response are generally of a given isotype that determines their biological function. A recent advance in B-cell epitope prediction is the development of a method by Gupta et al. [116] that allows the identification of B-cell epitopes capable of inducing specific class of antibodies. This method is based on SMVs trained on a dataset that includes linear B-cell epitopes known to induce IgG, IgE, and IgA antibodies.

3.2. Prediction of Conformational B-Cell Epitopes

Most B-cell epitopes are conformational and yet, prediction of conformational B-cell epitopes has lagged behind that of linear B-cell epitopes. There are two main practical reasons for that. First of all, prediction of conformational B-cell epitopes generally requires the knowledge of protein three-dimensional (3D) structure and this information is only available for a fraction of proteins [117]. Secondly, isolating conformational B-cell epitopes from their protein context for selective antibody production is a difficult task that requires suitable scaffolds for epitope grafting. Thereby, prediction of conformational B-cell prediction is currently of little relevance for epitope vaccine design and antibody-based technologies. Nonetheless, prediction of conformational B-cell epitopes is interesting for carrying structure-function studies involving antibody-antigen interactions.

There are several available methods to predict conformational B-cell epitopes (Table 2). The first to be introduced was CEP [118], which relied almost entirely on predicting patches of solvent-exposed residues. It was followed by DiscoTope [119], which, in addition to solvent accessibility, considered amino acid statistics and spatial information to predict conformational B-cell epitopes. An independent evaluation of these two methods using a benchmark dataset of 59 conformational epitopes revealed that they did not exceed a 40% of precision and a 46% of recall [120]. Subsequently, more methods were developed, like ElliPro [121] that aims to identify protruding regions in antigen surfaces and PEPITO [122] and SEPPA [123] that combine single physicochemical properties of amino acids and geometrical structure properties. The reported area under the curve (AUC) of these methods is around 0.7, which is indicative of a poor discrimination capacity yet better than random. Though, in an independent evaluation, SEPPA reached an AUC of 0.62 while all the mentioned methods had an AUC around 0.5 [124]. ML has also been applied to predict conformational B-cell epitopes in 3D-structures. Relevant examples include EPITOPIA [125] and EPSVR [126] which are based on naïve Bayes classifiers and support vector regressions, respectively, trained on feature vectors combining different scores. The reported AUC of these two methods is around 0.6.

The above methods for conformational B-cell epitope prediction identify generic antigenic regions regardless of antibodies, which are ignored [127]. However, there are also methods for antibody-specific epitope prediction. This approach was pioneered by Soga et al. [128] who defined an antibody-specific epitope propensity (ASEP) index after analyzing the interfaces of antigen-antibody 3D-structures. Using this index, they developed a novel method for predicting epitope residues in individual antibodies that worked by narrowing down candidate epitope residues predicted by conventional methods. More recently, Krawczyk et al. [129] developed EpiPred, a method that uses a docking-like approach to match up antibody and antigen structures, thus identifying epitope regions on the antigen. A similar approach is used by PEASE [130], adding that this method utilizes the sequence of the antibody and the 3D-structure of the antigen. Briefly, for each pair of antibody sequence and antigen structure, PEASE uses a machine learning model trained on properties from 120 antibody-antigen complexes to identify pair combination of residues from complementarity-determining regions (CDRs) of the antibody and the antigen that are likely to interact.

Another approach to identify conformational B-cell epitopes in a protein with a known 3D-structure is through mimotope-based methods. Mimotopes are peptides selected from randomized peptide libraries for their ability to bind to an antibody raised against a native antigen. Mimotope-based methods require to input antibody affinity-selected peptides and the 3D-structure of the selected antigen. Examples of bioinformatics tools for conformational B-cell epitope prediction using mimotopes include MIMOX [131], PEPITOPE [132], EPISEARCH [133], MIMOPRO [134], and PEPMAPPER [135] (Table 2).

As remarked before, methods for conformational B-cell epitope prediction generally require the 3D-structure of the antigen. Exceptionally, however, Ansari and Raghava [136] developed a method (CBTOPE) for the identification of conformational B-cell epitope from the primary sequence of the antigen. CBTOPE is based on SVM and trained on physicochemical and sequence-derived features of conformational B-cell epitopes. CBTOPE reported accuracy was 86.6% in crossvalidation experiments.

4. Concluding Remarks

Currently, T-cell epitope prediction is more advanced and reliable than that of B-cell prediction. However, while it is possible to confirm experimentally the predicted binding to MHC molecules of most peptides predicted, only

10% of those are shown to be immunogenic (able to elicit a T-cell response) [68]. Such a low T-cell epitope discovery rate is due to the fact that we do not have adequate models for predicting antigen processing yet [68]. The economic toll of low T-cell epitope discovery rate can be overcome, at least in part, by prioritizing protein antigens for epitope prediction [137–139]. For T-cell epitope vaccine development, researchers can also resort to experimentally known T-cell epitopes, available in epitope databases, selecting through immunoinformatics those that provide maximum population protection coverage [64, 140, 141]. In any case, T-cell epitope prediction remains an integral part of T-cell epitope mapping approaches. In contrast, B-cell epitope prediction utility is currently much more limited. There are several reasons to that. First of all, prediction of B-cell epitopes is still unreliable for both linear and conformational B-cell epitopes. Secondly, linear B-cell epitopes do usually elicit antibodies that do not crossreact with native antigens. Third, the great majority of B-cell epitopes are conformational and yet predicting conformational epitopes have few applications, as they cannot be isolated from their protein context. Under this scenario, the key is not only to improve current methods for B-cell epitope prediction but also to develop novel approaches and platforms for epitope grafting onto suitable scaffolds capable of replacing the native antigen.

To conclude, we wish to make two final remarks that are relevant for epitope vaccine design. First of all, it is that epitope prediction methods can provide potential epitopes from any given protein query but not all the antigens are equally relevant for vaccine development. Therefore, researchers have also developed tools to identify vaccine candidate antigens [142, 143], those likely to induce protective immunity, which can then be targeted for epitope prediction and epitope vaccine design. Second, it should be borne in mind that epitope peptides exhibit little immunogenicity and need to be used in combination with adjuvants, which increase immunogenicity by inducing strong innate immune responses that enable adaptive immunity [144–146]. Consequently, the discovery of new adjuvants is particularly relevant for epitope-based vaccines [146] and to that end, Nagpal et al. [147] developed a pioneered method that can predict the immunomodulatory activity of RNA sequences.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

Jose L. Sanchez-Trincado and Marta Gomez-Perosanz contributed equally to this work.

Acknowledgments

The authors wish to thank Inmunotek, SL and the Spanish Department of Science at MINECO for supporting the Immunomedicine group research through Grants SAF2006:07879, SAF2009:08301, and BIO2014:54164-R to Pedro A. Reche. The authors also wish to thank Dr. Esther M. Lafuente for critical reading and corrections.

References

  1. W. E. Paul, Fundamental Immunology, Lippincott Williams & Wilkins, 2012.
  2. B. Sun and Y. Zhang, “Overview of orchestration of CD4+ T cell subsets in immune responses,” Advances in Experimental Medicine and Biology, vol. 841, pp. 1–13, 2014. View at: Publisher Site | Google Scholar
  3. M. H. Van Regenmortel, “What is a B-cell epitope?” Methods in Molecular Biology, vol. 524, pp. 3–20, 2009. View at: Publisher Site | Google Scholar
  4. J. Ponomarenko and M. Van Regenmortel, “B-cell epitope prediction,” in Structural Bioinformatics, pp. 849–879, John Wiley & Sons, Inc, 2009. View at: Google Scholar
  5. T. A. Ahmad, A. E. Eweida, and L. H. El-Sayed, “T-cell epitope mapping for the design of powerful vaccines,” Vaccine Reports, vol. 6, pp. 13–22, 2016. View at: Publisher Site | Google Scholar
  6. L. Malherbe, “T-cell epitope mapping,” Annals of Allergy, Asthma & Immunology, vol. 103, no. 1, pp. 76–79, 2009. View at: Publisher Site | Google Scholar
  7. R. K. Ahmed and M. J. Maeurer, “T-cell epitope mapping,” Methods in Molecular Biology, vol. 524, pp. 427–438, 2009. View at: Publisher Site | Google Scholar
  8. E. M. Lafuente and P. A. Reche, “Prediction of MHC-peptide binding: a systematic and comprehensive overview,” Current Pharmaceutical Design, vol. 15, no. 28, pp. 3209–3220, 2009. View at: Publisher Site | Google Scholar
  9. P. E. Jensen, “Recent advances in antigen processing and presentation,” Nature Immunology, vol. 8, no. 10, pp. 1041–1048, 2007. View at: Publisher Site | Google Scholar
  10. L. J. Stern and D. C. Wiley, “Antigenic peptide binding by class I and class II histocompatibility proteins,” Structure, vol. 2, no. 4, pp. 245–251, 1994. View at: Publisher Site | Google Scholar
  11. D. R. Madden, “The three-dimensional structure of peptide-MHC complexes,” Annual Review of Immunology, vol. 13, no. 1, pp. 587–622, 1995. View at: Publisher Site | Google Scholar
  12. D. R. Madden, D. N. Garboczi, and D. C. Wiley, “The antigenic identity of peptide-MHC complexes: a comparison of the conformations of five viral peptides presented by HLA-A2,” Cell, vol. 75, no. 4, pp. 693–708, 1993. View at: Publisher Site | Google Scholar
  13. D. V. Desai and U. Kulkarni-Kale, “T-cell epitope prediction methods: an overview,” Methods in Molecular Biology, vol. 1184, pp. 333–364, 2014. View at: Publisher Site | Google Scholar
  14. A. Patronov and I. Doytchinova, “T-cell epitope vaccine design by immunoinformatics,” Open Biology, vol. 3, no. 1, article 120139, 2013. View at: Publisher Site | Google Scholar
  15. R. Vita, J. A. Overton, J. A. Greenbaum et al., “The immune epitope database (IEDB) 3.0,” Nucleic Acids Research, vol. 43, D1, pp. D405–D412, 2015. View at: Publisher Site | Google Scholar
  16. M. Molero-Abraham, E. M. Lafuente, and P. Reche, “Customized predictions of peptide-MHC binding and T-cell epitopes using EPIMHC,” Methods in Molecular Biology, vol. 1184, pp. 319–332, 2014. View at: Publisher Site | Google Scholar
  17. C. P. Toseland, D. J. Clayton, H. McSparron et al., “AntiJen: a quantitative immunology database integrating functional, thermodynamic, kinetic, biophysical, and cellular data,” Immunome Research, vol. 1, no. 1, p. 4, 2005. View at: Publisher Site | Google Scholar
  18. S. P. Singh and B. N. Mishra, “Major histocompatibility complex linked databases and prediction tools for designing vaccines,” Human Immunology, vol. 77, no. 3, pp. 295–306, 2016. View at: Publisher Site | Google Scholar
  19. J. D'Amaro, J. G. A. Houbiers, J. W. Drijfhout et al., “A computer program for predicting possible cytotoxic T lymphocyte epitopes based on HLA class I peptide-binding motifs,” Human Immunology, vol. 43, no. 1, pp. 13–18, 1995. View at: Publisher Site | Google Scholar
  20. M. Bouvier and D. Wiley, “Importance of peptide amino and carboxyl termini to the stability of MHC class I molecules,” Science, vol. 265, no. 5170, pp. 398–402, 1994. View at: Publisher Site | Google Scholar
  21. J. Ruppert, J. Sidney, E. Celis, R. T. Kubo, H. M. Grey, and A. Sette, “Prominent role of secondary anchor residues in peptide binding to HLA-A2.1 molecules,” Cell, vol. 74, no. 5, pp. 929–937, 1993. View at: Publisher Site | Google Scholar
  22. M. Nielsen, C. Lundegaard, P. Worning et al., “Improved prediction of MHC class I and class II epitopes using a novel Gibbs sampling approach,” Bioinformatics, vol. 20, no. 9, pp. 1388–1397, 2004. View at: Publisher Site | Google Scholar
  23. H. G. Rammensee, J. Bachmann, N. P. N. Emmerich, O. A. Bachor, and S. Stevanovic, “SYFPEITHI: database for MHC ligands and peptide motifs,” Immunogenetics, vol. 50, no. 3-4, pp. 213–219, 1999. View at: Publisher Site | Google Scholar
  24. P. A. Reche, J. P. Glutting, and E. L. Reinherz, “Prediction of MHC class I binding peptides using profile motifs,” Human Immunology, vol. 63, no. 9, pp. 701–709, 2002. View at: Publisher Site | Google Scholar
  25. P. A. Reche and E. L. Reinherz, “Definition of MHC supertypes through clustering of MHC peptide-binding repertoires,” Methods in Molecular Biology, vol. 409, pp. 163–173, 2007. View at: Publisher Site | Google Scholar
  26. P. A. Reche, J. P. Glutting, H. Zhang, and E. L. Reinherz, “Enhancement to the RANKPEP resource for the prediction of peptide binding to MHC molecules using profiles,” Immunogenetics, vol. 56, no. 6, pp. 405–419, 2004. View at: Publisher Site | Google Scholar
  27. M. Gribskov and S. Veretnik, “Identification of sequence pattern with profile analysis,” Methods in Enzymology, vol. 266, pp. 198–212, 1996. View at: Publisher Site | Google Scholar
  28. K. C. Parker, M. A. Bednarek, and J. E. Coligan, “Scheme for ranking potential HLA-A2 binding peptides based on independent binding of individual peptide side-chains,” The Journal of Immunology, vol. 152, no. 1, pp. 163–175, 1994. View at: Google Scholar
  29. H. H. Bui, J. Sidney, B. Peters et al., “Automated generation and evaluation of specific MHC binding predictive tools: ARB matrix applications,” Immunogenetics, vol. 57, no. 5, pp. 304–314, 2005. View at: Publisher Site | Google Scholar
  30. M. Nielsen, C. Lundegaard, and O. Lund, “Prediction of MHC class II binding affinity using SMM-align, a novel stabilization matrix alignment method,” BMC Bioinformatics, vol. 8, no. 1, p. 238, 2007. View at: Publisher Site | Google Scholar
  31. B. Peters and A. Sette, “Generating quantitative models describing the sequence specificity of biological processes with the stabilized matrix method,” BMC Bioinformatics, vol. 6, no. 1, p. 132, 2005. View at: Publisher Site | Google Scholar
  32. T. Sturniolo, E. Bono, J. Ding et al., “Generation of tissue-specific and promiscuous HLA ligand databases using DNA microarrays and virtual HLA class II matrices,” Nature Biotechnology, vol. 17, no. 6, pp. 555–561, 1999. View at: Publisher Site | Google Scholar
  33. B. Peters, W. Tong, J. Sidney, A. Sette, and Z. Weng, “Examining the independent binding assumption for binding of peptide epitopes to MHC-I molecules,” Bioinformatics, vol. 19, no. 14, pp. 1765–1772, 2003. View at: Publisher Site | Google Scholar
  34. P. Guan, I. A. Doytchinova, C. Zygouri, and D. R. Flower, “MHCPred: a server for quantitative prediction of peptide–MHC binding,” Nucleic Acids Research, vol. 31, no. 13, pp. 3621–3624, 2003. View at: Publisher Site | Google Scholar
  35. M. Milik, D. Sauer, A. P. Brunmark et al., “Application of an artificial neural network to predict specific class I MHC binding peptide sequences,” Nature Biotechnology, vol. 16, no. 8, pp. 753–756, 1998. View at: Publisher Site | Google Scholar
  36. V. Brusic, G. Rudy, G. Honeyman, J. Hammer, and L. Harrison, “Prediction of MHC class II-binding peptides using an evolutionary algorithm and artificial neural network,” Bioinformatics, vol. 14, no. 2, pp. 121–130, 1998. View at: Publisher Site | Google Scholar
  37. P. Donnes and O. Kohlbacher, “Integrated modeling of the major events in the MHC class I antigen processing pathway,” Protein Science, vol. 14, no. 8, pp. 2132–2140, 2005. View at: Publisher Site | Google Scholar
  38. M. Bhasin and G. P. S. Raghava, “SVM based method for predicting HLA-DRB1

Copyright

Copyright © 2017 Jose L. Sanchez-Trincado et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Fine epitope signature of antibody neutralization breadth at the HIV-1 envelope CD4-binding site

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by Sundling, C. in: JCI | PubMed | Google Scholar | />

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by Donofrio, G. in: JCI | PubMed | Google Scholar

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by Karlsson Hedestam, G. in: JCI | PubMed | Google Scholar

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by Bonsignori, M. in: JCI | PubMed | Google Scholar | />

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by von Reyn, C. in: JCI | PubMed | Google Scholar

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by Zolla-Pazner, S. in: JCI | PubMed | Google Scholar | />

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by Bailey-Kellogg, C. in: JCI | PubMed | Google Scholar

1 Thayer School of Engineering and

2 Molecular and Cellular Biology Program, Dartmouth College, Hanover, New Hampshire, USA.

3 Faculty of Medicine and Biomedical Sciences, University of Yaoundé 1, Yaoundé, Cameroon.

4 Department of Immunology and Microbiology, The Scripps Research Institute, La Jolla, California, USA.

5 Unit of Infectious Diseases, Department of Medicine, Solna, Karolinska Institute, Stockholm, Sweden.

6 US Military HIV Research Program, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, Maryland, USA.

7 Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden.

8 Duke Human Vaccine Institute, Durham, North Carolina, USA.

9 Department of Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire, USA.

10 DarDar Health Programs, Dar es salaam, Tanzania.

11 Tokyo Medical and Dental University, Tokyo, Japan.

12 Department of Pathology, NYU School of Medicine, New York, New York, USA.

13 Departments of Medicine and Microbiology, Icahn School of Medicine at Mount Sinai, New York, New York, USA.

14 Ragon Institute of MGH, MIT, and Harvard University, Cambridge, Massachusetts, USA.

15 Howard Hughes Medical Institute, Chevy Chase, Maryland, USA.

16 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA.

17 Department of Computer Science, Dartmouth College, Hanover, New Hampshire, USA.

Address correspondence to: Margaret E. Ackerman, 14 Engineering Drive, Hanover, New Hampshire 03755, USA. Phone: 603.646.9922 Email: [email protected]

Find articles by Ackerman, M. in: JCI | PubMed | Google Scholar

Major advances in donor identification, antigen probe design, and experimental methods to clone pathogen-specific antibodies have led to an exponential growth in the number of newly characterized broadly neutralizing antibodies (bnAbs) that recognize the HIV-1 envelope glycoprotein. Characterization of these bnAbs has defined new epitopes and novel modes of recognition that can result in potent neutralization of HIV-1. However, the translation of envelope recognition profiles in biophysical assays into an understanding of in vivo activity has lagged behind, and identification of subjects and mAbs with potent antiviral activity has remained reliant on empirical evaluation of neutralization potency and breadth. To begin to address this discrepancy between recombinant protein recognition and virus neutralization, we studied the fine epitope specificity of a panel of CD4-binding site (CD4bs) antibodies to define the molecular recognition features of functionally potent humoral responses targeting the HIV-1 envelope site bound by CD4. Whereas previous studies have used neutralization data and machine-learning methods to provide epitope maps, here, this approach was reversed, demonstrating that simple binding assays of fine epitope specificity can prospectively identify broadly neutralizing CD4bs–specific mAbs. Building on this result, we show that epitope mapping and prediction of neutralization breadth can also be accomplished in the assessment of polyclonal serum responses. Thus, this study identifies a set of CD4bs bnAb signature amino acid residues and demonstrates that sensitivity to mutations at signature positions is sufficient to predict neutralization breadth of polyclonal sera with a high degree of accuracy across cohorts and across clades.

Advancements in donor identification, experimental methods, and the generation of better reagents for cloning pathogen-specific antibodies have been realized in the past decade ( 1 – 3 ). These innovations have led to rapid growth in the number of broadly neutralizing antibodies (bnAbs) identified against clinically relevant pathogens, including HIV-1 ( 4 – 11 ). However, the ability to unambiguously translate binding to recombinant protein in biophysical assays into an understanding of in vivo activity against HIV-1 has lagged behind. In particular, there is a lack of concordance between antibody recognition of recombinant antigen in biophysical assays and broad antiviral neutralization activity in cell- and pseudovirus-based functional assays in the setting of humoral responses to HIV-1 infection and vaccination. Consequently, identification of subjects and mAbs with potent antiviral activity has generally remained reliant on empirical evaluation of neutralization potency and breadth. Such distinctions suggest that further work aimed at refining the properties of the antigen-derived protein probes used to characterize binding patterns of humoral responses may contribute to more efficient and effective identification of both protective mAbs and polyclonal antibody (pAb) responses.

HIV-1 bnAbs are of high interest due to their ability to prevent infection in animal models ( 12 – 16 ) and influence viral loads and host responses in humans ( 17 – 21 ). Numerous such bnAbs have been isolated and mapped to different regions of the HIV-1 spike, including the V1/V2 loop region, the V3 loop region, the membrane proximal extracellular region, the gp120-gp41 interface, and the CD4 receptor–binding site (CD4bs) ( 22 – 27 ). Among these, the CD4bs is of particular interest as a target for both mAb therapy and prophylaxis as well as for vaccine design. Despite the high sequence variability of the virus envelope, the CD4bs is functionally conserved and less masked by the glycan shield, and numerous potent bnAbs have been mapped to this region. Intriguingly, while many infected and vaccinated subjects raise CD4bs-specific antibodies, only a subset is broadly or potently neutralizing ( 28 ). Among these, convergent binding modes and common maturation pathways have been observed ( 29 – 31 ). Strikingly, differences in the angle of approach of broad or nonneutralizing CD4bs antibodies have been noted, as have differences in the fine epitope footprints ( 32 , 33 ). Taken together, these findings suggest an association between CD4bs antibody fine epitopes and the ability to broadly and potently neutralize HIV-1.

Here, we undertook a study to evaluate the fine epitope specificity of a panel of CD4bs mAbs using a designed library of YU2 gp120 core amino acid point mutants ( 34 ), and, in doing so, we define the molecular recognition features of functionally potent humoral responses to the HIV-1 envelope CD4bs. Whereas previous studies have used neutralization data to delineate antibody epitopes ( 35 – 40 ), here that process is reversed to allow prediction of neutralization potency from epitope-mapping experiments. Employing both computational models and experimentation, we demonstrate that biophysical definition of antibody fine epitope specificity can contribute to prospective identification of broadly neutralizing CD4bs–specific mAbs. Building on this result, we further show that epitope mapping and predictions of neutralization breadth can be accomplished in the setting of polyclonal serum responses.

Fine CD4bs epitope specificity predicts neutralization breadth. To define antibody epitopes, we mapped a set of approximately 30 single amino acid substitution variants of yeast-displayed YU2 gp120 core ( 34 ) via flow cytometry for disruption of binding across a panel of CD4bs antibodies. We evaluated 9 antibodies with narrow and/or weak neutralizing activity (wnAbs) ( 41 ), including b6 and F105 6 vaccine-induced non-bnAbs from macaques (viAbs) ( 42 ) 9 antibodies with more broad and potent neutralization activity (bnAbs b12, VRC01, PGV04, NIH45-46 G45W, CH31-CH34, and CH103) and the natural ligand CD4. This set of CD4bs mAbs, as well as CD4, generally shared a common pattern of sensitivity toward 4 residues, K282, D368, G473, and R476, all located along the CD4bs ridge facing the inner core domain (representative examples, Figure 1A). However, hierarchical clustering demonstrated that the epitope maps of CD4bs bnAbs could be differentiated from the non–broadly neutralizing (non-bnAb) CD4bs viAbs and wnAbs (Figure 1B), indicating a common difference in CD4bs recognition that was associated with neutralization breadth. The group of clustered bnAbs was distinguished by sensitivity to point mutations deeper within the CD4bs (Figure 1A), associated with S365K, T455E, and G459E substitutions (Figure 1B). Thus, both visual inspection and hierarchical clustering indicated that epitope maps could be used to discriminate neutralization breadth among CD4bs mAbs.

Epitope mapping of CD4bs antibodies. A panel of gp120 core point mutants was used to epitope map CD4bs antibodies (n = 25). Representative mAbs from 4 groups, including weakly neutralizing (wnAbs), vaccine-induced, and broadly neutralizing (bnAb) antibodies were evaluated. (A) The sensitivity of various mAbs to mutation of core residues is plotted on a structural representation. The CD4bs is colored green, and tolerated point mutations are colored black, while substitutions driving reduced (<80% light blue to red) or strongly enhanced (>160% blue) binding to the core relative to WT are indicated. (B) Heatmap representation of the epitope-mapping results observed for the set of CD4bs mAbs. Hierarchical clustering identifies major subgroups of CD4bs mAbs that are associated with neutralization breadth and potency. The color bar at top indicates the class of core: core variants with substitutions made in CD4bs residues are indicated in green, core variants with substitutions made in other sites on the core indicated are indicated in black, and the unmutated WT core are indicated in gray. The color bar at the left indicates the mAb class, with bnAbs indicated in black and non-bnAbs indicated in shades of gray (vaccine-induced antibodies in light gray, infection-induced antibodies in dark gray).

Validation of bnAb epitope signatures to predict CD4bs antibody neutralization breadth. To further investigate this observation, a random forest model ( 43 ), which corrects for decision trees’ propensity to overfit data, was trained to classify the CD4bs mAbs by neutralization breadth (bnAb versus viAb and wnAbs) via the generation of an ensemble of decision trees, using epitope-mapping data as input. Perfect classification accuracy was achieved across the set of mAbs in the setting of leave-one-out cross-validation (Table 1). The relative importance of each mutant in the panel to predictively classifying neutralization breadth across the forest of decision trees identified positions S365, G459, and T455 (Figure 2A) as contributing the most toward classifier predictions. As might be expected, the 10 mutated CD4bs residues were generally ranked higher than mutations at other positions in the gp120 core. The widely used CD4bs-defining D386R mutation ( 44 ) ranked among the bottom half of variants, providing essentially no contribution to neutralization breadth classification, which is consistent with the sensitivity of all CD4bs mAbs tested to this mutation. Similar performance and contributing features were observed using elastic net classification as an alternative modeling approach, supporting the generalized ability of this epitope recognition signature to predict neutralization breadth. Finally, permutation tests, in which models were learned from data when the neutralization class labels had been scrambled, further established model robustness.

Classification of CD4bs mAb neutralization breadth. (A) A random forest approach was used to classify mAb neutralization breadth using epitope maps. The relative importance of each point mutant to the classification models is presented in decreasing order, as ranked based on mean decrease in Gini index. CD4bs residues are colored in green, and other core residues are colored black. A structural model of the core, denoting the locations of the top 3 positions (S365, T455, G459) utilized by the classifier in red, is shown. (B) Benchmarking against D368R. The residues most important to the classifier were mutated to generate a triple mutant probe (STG). The binding of each CD4bs mAb (n = 26) relative to the WT gp120 core is presented for STG and the CD4bs probe D368R across individual mAbs and when grouped according to neutralization breadth. Data are represented as median and interquartile range.

Classification confusion matrix of mAb neutralization breadth

Based on these models, a triply mutated variant of the gp120 core, denoted as STG and consisting of S365K, T455L, and G459E substitutions, was generated by site-directed mutagenesis in order to evaluate bnAb signature substitutions in combination. The sensitivity of each CD4bs mAb to the substitutions present in the STG mutant and to the D368R substitution was evaluated relative to WT core. Whereas all CD4bs mAbs were sensitive to substitution at the D368 position, as evidenced by an average 75% decrease in binding signal for the D368R variant relative to WT core, only bnAb recognition was dramatically affected by the STG triple mutation (Figure 2B). While the D368R mutation did tend to be more disruptive to bnAbs than non-bnAbs, recognition by more than half of the non-bnAbs was reduced by a factor equivalent to that of the bnAbs. In contrast, antibody binding to the STG mutant was generally unaffected for non-bnAbs, maintaining an average of 97% of WT signal, whereas it was dramatically decreased among bnAbs, averaging 3% of WT signal (P = 3.1 × 10 –12 by 2-tailed Mann-Whitney test). Thus, consistent with the hierarchical clustering and random forest classification results, in which bnAbs and non-bnAbs could be discriminated statistically with the mutant panel using models weighted heavily on the 3-position signature, evaluation of the STG triple mutant confirmed robust discrimination capacity experimentally. A limited set of other substitutions was made at these positions and demonstrated similar binding profiles, indicating that bnAb sensitivity was likely more position rather than substitution specific.

Epitope mapping of polyclonal sera. Based on excellent discrimination among mAbs, we next evaluated whether the mutant panel could be used to epitope map pAb samples derived from serum. A set of 121 samples from chronically infected subjects from Boston area cohorts was screened for binding to yeast-displayed YU2 core, and the 99 samples with significant YU2 core reactivity were further evaluated across the entire mutant panel (Figure 3A). While essentially all serum samples with core-specific antibodies were highly sensitive to the D368R mutation and moderately sensitive to substitutions at G473 and R476, they did not bear appreciable sensitivity to the neutralization breadth signature residues. When clustered alongside the maps from the CD4bs mAb panel (Figure 3A), the pAb-binding signatures were clearly more similar to the wnAbs and viAbs than the group of bnAbs, which formed their own branch on the dendrogram. Indeed, because neutralizing responses were expected among chronically infected subjects ( 45 ), the lack of any CD4bs bnAb-like epitope signatures suggested that either the mAb-derived signature was not a general feature of pAb responses or that more prevalent responses toward the lower CD4bs ridge might impede detection of less common antibodies able to interact with the CD4bs neutralization signature residues.

Polyclonal sera epitope-mapping results. (A) Heatmap comparing epitope-mapping data from polyclonal sera from Boston area (Ragon Institute) HIV-infected subjects positive for YU2 gp120 core-specific antibodies clustered together with the mAb epitope maps. The vertical color bar indicates sample type: pAb samples are shown in white (n = 99), bnAbs in black (n = 10), vaccine-induced weakly neutralizing antibodies (viAbs, n = 6) in light gray, and infection-induced weakly neutralizing antibodies (wnAbs, n = 9) in dark gray. (B) Serum from a donor with a known bnAb (PGV04) was evaluated for binding to the epitope-mapping panel. The ability of the STG mutant to deplete core-specific antibodies without broad neutralization and to facilitate identification of the presence of CD4bs antibodies with broad neutralization was determined for the PGV04 donor serum and for PGV04-spiked HIVIG. Error bars, shown only for the enriched PGV04 serum, indicate SD observed between duplicate measurements. (C) Heatmap of polyclonal sera samples from a Tanzanian cohort (n = 10) before and after enrichment of potential bnAbs using the STG mutant-based depletion. Horizontal color bars indicate the class of core: core variants with substitutions made in CD4bs residues are indicated in green, core variants with substitutions made in other sites on the core are indicated in black.

To investigate the hypothesis that CD4bs non-bnAbs may mask the detection of CD4bs bnAbs, serum from the CD4bs bnAb PGV04 donor was epitope mapped (Figure 3B). Indeed, despite the presence of a bnAb, serum from the PGV04 donor failed to recapitulate characteristics of the PGV04 mAb and appeared similar to the other serum samples evaluated: sensitivity to the bnAb signature residues was not observed, but reduced binding to D368R and R476V mutants was noted. Similar results were observed when known concentrations of the PGV04 mAb were spiked into HIV immune globulin (HIVIG), confirming the hypothesis that nonneutralizing core-specific antibodies can impede detection of bnAbs in mapping approaches that rely on the gp120 core. This observation suggested that depleting pAb samples of non-bnAbs using the STG mutant might result in the enrichment of CD4bs bnAbs. This possibility was experimentally tested by passing PGV04-spiked HIVIG through a spin column containing a bed of STG-displaying yeast and epitope mapping the column flow through. Similar serum adsorption steps have contributed to previous gp120 epitope-mapping studies ( 46 ). This enrichment step enabled identification of sensitivity toward substitution of critical residues recognized by the PGV04 mAb, demonstrating that this flow-through fraction was enriched for antibodies with the PGV04 phenotype. Similarly, though insufficient PGV04 donor serum was available to evaluate the STG-enriched fraction against the whole epitope panel, a subset of the panel was analyzed for binding to the enriched fraction. Sensitivity to S365K, T455E, G459E, I467K, and G473R, key features of the PGV04 mAb that were not observed in whole PGV04 donor serum, was observed following enrichment (Figure 3B).

Inspired by the observation that STG-based enrichment could unmask the presence of a CD4bs bnAb, we extended this enrichment process across a set of 10 serum samples from a Tanzanian cohort. In a pilot experiment, enriched sera (the flow-through fraction that results when the pool is depleted by passage through a column of STG yeast cells) were analyzed for differential binding to a subset of the mutant panel, and the majority of these demonstrated differences in epitope-mapping results before and after enrichment (Figure 3C). The difference between before and after enrichment maps that was most apparent was increased sensitivity to K282V, S365K, and G473R substitutions. Limited sensitivity to T455E was noted, and none of these samples demonstrated sensitivity to G459E. When the neutralization activity of this limited panel of serum samples was determined, only weak and narrow responses were noted, consistent with the lack of responses to the substitutions characteristic of CD4bs bnAbs. While the Tanzanian serum samples were relatively insensitive to T455E and G459E, and were also nonneutralizing, suggesting the utility of this mapping approach, we wished to more meaningfully explore the utility of this epitope signature by applying it to samples with known neutralization profiles. To this end, we applied the enrichment process to a set of 16 serum samples from the RV217 cohort that were known to be broadly neutralizing but for which epitope mapping based on neutralization data ( 40 ) suggested a wide range of epitope specificities. As positive and negative controls, mock samples consisting of VRC01, PGV04, or F105 mAbs spiked into HIVIG (5% by mass) were generated and subjected to the same enrichment process in parallel. Diverse titration patterns were observed across the set of 16 sera, with some subjects exhibiting depletion patterns similar to that of VRC01-spiked HIVIG and others more similar to the F105-spiked sample (Figure 4A). To quantitate these differences, the area under the titration curve (AUC) was calculated, and all samples were ranked by the ratio of enriched WT-AUC over enriched STG-AUC (Figure 4B), as sera samples with a higher ratio would be predicted to be more likely to contain CD4bs bnAbs. Notably, among these samples, one subject demonstrated a binding profile superior to that observed for either the VRC01 or PGV04 spike-in samples, and a number of subjects exhibited profiles intermediate between these bnAbs and the F105-spiked sample. These results suggest the potential utility of the STG mutant in B cell cloning and bnAb isolation efforts as well as in assessing bnAb prevalence more precisely in heterogeneous populations.

Inferring the presence of CD4bs bnAbs via STG-based enrichment and serum profiling. A set of 16 samples from the RV217 HIV cohort were titrated for binding to WT (red) and STG (blue) core before and following enrichment of CD4bs bnAbs via STG-based depletion to enable inference of the presence or absence of CD4bs bnAbs. VRC01, PGV04, or F105 were spiked into HIV immune globulin (HIVIG) as positive and negative controls. (A) Titration curves of VRC01-spiked and F105-spiked HIVIG (top), and two RV217 subjects (bottom) before (broken line) and after (solid line) enrichment. The dotted line indicates the signal baseline used for AUC calculations. (B) The AUC for all sample titrations against WT and STG cores after enrichment was calculated and samples were plotted by rank in the ratio of the AUC for WT relative to STG. The mAb-spiked samples are highlighted in gray. Samples with no measurable binding were assigned a value of 100.

To link single substitution-level epitope maps of enriched pAb fractions to pAb neutralization activity, samples with known neutralization potency from a second, independent Boston area cohort were evaluated. Recognition profiles before (Figure 5A) and after (Figure 5B) CD4bs bnAb enrichment using the STG mutant were defined for a set of 10 serum samples. Clustering subjects based on the maps following enrichment identified a subset of individuals with broad reactivity against the CD4bs, including sensitivities to bnAb signature residues that were only apparent following enrichment.

Prediction of polyclonal sera neutralization breadth across diverse subject cohorts. Heatmap of the relative binding of sera before (A) and after (B) STG-based enrichment for a set of 10 samples from a second Boston area (BIDMC) cohort against CD4bs mutant cores (green), the WT core, and the triple mutant STG core (both in gray). (C) Neutralization potency (ID50) of the 10 sera samples across a panel of 18 tier 2 virus strains (each dot represents 1 unique virus strain) with subjects aligned as in A and B. Interquartile ranges and Tukey whiskers are shown mean ID50 values across the virus panel for each subject were compared between the STG sensitive (green) and insensitive (black) group, with significance defined by Mann-Whitney test. (D) A classifier was trained to predict the neutralization breadth of polyclonal samples from HIV-infected subjects in the Boston area (BIDMC) cohort (n = 10), a South African cohort (n = 19), or both, based on epitope-mapping input of sera or the STG-enriched IgG fraction. Dashed lines represent classifier performance when the complete mutant panel (red), CD4bs residues only (green), or the S, T, and G positions alone (blue) were used in model training. Black lines represent classifier performance when permuted neutralization class assignments were predicted using the whole panel or CD4bs and S, T, and G mutant subsets. Classification accuracy of models learned from considering S, T, and G positions only are noted at the top left of each panel in blue.

The neutralization potency of these samples across a panel of 18 tier 2 viruses was tested (Figure 5C). Those samples that clustered together as being most strongly disrupted by CD4bs substitutions following STG-based enrichment exhibited potentiated neutralization capacity (as defined by the mean ID50 across the virus panel for each subject) relative to the cluster of samples that were relatively insensitive to CD4bs substitutions (P < 0.01 by 2-tailed Mann-Whitney test).

Prediction of serum neutralization breadth using the bnAb epitope signature. The presence of the post hoc association described suggested that epitope maps following STG-based enrichment might support predictive models of neutralization breadth in polyclonal samples. Therefore, we again utilized the random forest machine-learning method to train a classifier to distinguish neutralization breadth. We evaluated two cohorts, the 10 Boston area subjects described above and an additional cohort of 19 South African subjects to investigate whether this approach could be generalized to clade C infection. Predictions of neutralization breadth class (high versus low) were made using all 31 mutations, the 10 CD4bs residues, or the 3 bnAb signature positions S, T, and G (Figure 5D and Table 2). Because the neutralization data for pAbs were collected from a limited panel of pseudoviruses and for a limited number of subjects, neutralization potency was divided into two groups: broadly neutralizing and non–broadly neutralizing, rather than across a continuum of activities. As with any such classification exercise, results of this analysis are expected to be somewhat dependent on the specific boundary used to define the two groups.

Classification model accuracies

Permutation tests were conducted to evaluate the classification error distribution when neutralization class labels were scrambled. These tests established a baseline performance expectation that served as a negative control. When native serum maps were used, the performance of predictive models was essentially random, demonstrating accuracies of approximately 0.5 in defining the two classes. This performance was indistinguishable from the error distribution when permuted neutralization class assignments were predicted. In contrast, when the epitope-mapping data following the STG-based enrichment were used, all residue sets demonstrated accuracies that were significantly better than predicted by chance alone (Table 2). Strikingly, the best predictive performance was observed when only the 3 bnAb signature positions were utilized (Figure 5D), demonstrating that this motif was sufficient to effectively classify the breadth and potency of neutralization activity present. When these signature positions were excluded from consideration, model performance was significantly degraded, demonstrating that beyond being sufficient, these positions were also necessary for best results. Furthermore, both cohorts, which represent subjects infected primarily with either clade B or clade C viruses, demonstrated similar performance and reliance on the S, T, and G positions, suggesting that predictive models were neither cohort nor clade specific, but were generalizable. Indeed, when cohorts were combined, similar performance accuracy was observed, suggesting that similar decision trees effectively classified subjects from both cohorts. Collectively, these results suggest that STG-based enrichment was necessary, and that this enrichment approach was sufficient to enable prediction of neutralization breadth of polyclonal sera across cohorts and across clades with a high degree of accuracy.

Considerable effort has been expended to identify and characterize antibodies with the ability to neutralize diverse HIV-1 variants. Such bnAbs are not only potentially useful in prevention and treatment, but they also offer a model for vaccine design in that they encapsulate the natural developmental history and illuminate features of potentially protective humoral immune responses. Since the first identification of HIV-1 bnAbs in the early 1990s, almost 100 bnAbs have been described ( 47 ), and a new wave of clinical studies evaluating their utility has been initiated. In particular, a number of CD4bs bnAbs are currently being investigated for antiviral therapy and prophylaxis ( 17 – 21 ). Similarly, diverse vaccine design strategies have focused on the CD4bs, involving approaches that have matured from obscuring other surface residues ( 48 ), to minimizing the presence of other epitopes ( 49 ), to grafting the CD4bs onto alternative scaffolds ( 50 ), to designing modified CD4bs epitopes favoring interactions with specific germline B cell receptors ( 51 ), to selecting immunogens based on B cell and viral coevolution ( 52 ). These and other examples highlight the importance of the CD4bs in the design of both HIV-1 vaccines and antiviral antibody drugs.

The promising new work devoted to these efforts provides reason to be optimistic that eliciting CD4bs bnAbs by vaccination is possible and suggests that effective characterization of the antibody responses being raised by novel vaccine candidates may serve a critical role in vaccine research and development. While other antibody phenotypes, such as loop length and CDR composition, have been used as surrogates to suggest progress toward neutralization breadth, such proxies are unlikely to be either strongly or universally associated with neutralization. Similarly, given the insufficiency of CD4 or antibody competition experiments to fully reflect neutralization potency, we expect that more sophisticated epitope-mapping probes and strategies, such as that defined here, may be central to ongoing efforts to profile mAbs and pAb responses ( 53 , 54 ). These methods may help in particular by providing more fine-grained information about modes of recognition that may then better inform the design of future immunogens or immunization sequencing regimens.

Here, we show that prior knowledge as to function-phenotype linkages can be used to develop robustly predictive models of the neutralization phenotype observed for samples based on antigen recognition profiles. While inferences of neutralization potency made from patterns of binding recognition learned from examples could likewise be made from epitope information gained from other experimental methods, such as cocrystallization, hydrogen-deuterium exchange mass spectrometry, and scanning and shotgun mutagenesis, the yeast display-based approach has considerable experimental advantages. Similarly, as appropriate sets of antibodies become available, defined antigen variants could be developed to distinguish neutralization breadth and potency at other epitopes of interest.

Whereas the current state of the art in pAb epitope mapping involves using experimental neutralization data to predict HIV-1 antibody epitopes ( 35 – 39 ), here the directionality of inference is reversed, and, to our best knowledge, this is the first time that polyclonal HIV-1 antibody epitope-mapping data were used to predict neutralization potency. While numerous laboratories have established high-throughput means to conduct neutralization assays, biophysical probes bear advantages in cost, ease, and safety ( 55 ). Significantly, distillation of a panel of single amino acid point mutants into a single, triply mutated variant enabled highly efficient experimental discrimination by simply evaluating binding to WT versus triple mutant core. Collectively, these data also suggest that B cell sorting of B cell receptors that prefer WT to the STG mutant could contribute to efforts to identify and clone CD4bs bnAbs.

However, generalization of this epitope-based approach to deduce neutralization has a number of limitations. First, it relies on the availability of antibodies with known activity profiles and thus has little relevance to settings in which such rich knowledge does not yet exist. Second, since it is built on prior knowledge, it is expected to have a limited ability to identify antibodies with novel recognition modes. If there are no examples in the class of antibodies used to train models and build a discriminatory epitope probe, the designed probe will presumably have no ability to recognize new antibodies that may exhibit the same functional phenotype but rely on a different biophysical one.

We also note that while the STG probe was effective in distinguishing neutralization potency among mAbs and for the cohorts (Beth Israel Deaconess Medical Center [BIDMC] and South Africa) in which neutralization data were available, additional data are needed to further validate these observations. Relevant to this concern, epitope maps (before and/or after enrichment) from other cohorts (Ragon Institute and Tanzania) differed from those observed among subjects for which neutralization data were available in several ways. For example, prior to enrichment, D368R was strongly disruptive in the majority of subjects from the Ragon Institute cohort this phenotype was considerably less prevalent among the BIDMC (and Tanzanian) subjects. Additionally, after enrichment, the BIDMC cohort samples with broad neutralization potency exhibited multiple sites of substitution sensitivity, and this broad pattern of disruption across the CD4bs was distinct from that observed among Tanzanian subjects after enrichment. Such differences in both the before and after enrichment epitope maps among the samples from which neutralization data were available could be potentially explained by a number of factors, including time since infection, concomitant disease burdens, or differences in recognition of YU2 core associated with diversity in the infecting virus, among others. By extension, we cannot exclude that other factors such as these may influence the utility of this approach to predict neutralization activity.

Our results also suggest that, in general, the bulk of the antibody response in infected subjects that recognizes the gp120 core may not be relevant to virus neutralization. Indeed, it has long been known that much of the humoral response to HIV-1 infection is directed to “viral debris” ( 56 ). Our observation that non-bnAb responses can mask the presence of bnAbs has several important potential biological implications. First, this result is consistent with numerous previous studies reporting the inability of gp120-binding titers to strongly correlate with neutralization breadth and potency and with reports of limited success in isolating gp120-reactive antibodies with potent neutralizing activity ( 32 , 57 ). That most core-specific antibody responses are apparently irrelevant to neutralization has implications as to the usefulness of gp120 or gp120 cores as vaccine immunogens. This observation also suggests that there could be competition between neutralizing core-specific antibodies and nonneutralizing core-specific antibodies and that binding of non-bnAbs might block bnAb activity. While we did not evaluate this possibility intensively, the inability of non-bnAbs to compete with nAbs has been long established ( 58 ). Further, in this study, neutralization enhancement was not observed when a subset of sera was evaluated for neutralization activity before and after STG-based depletion of non-bnAbs. Finally, and in contrast to our biophysical assays, the presence of PGV04 in the PGV04 donor serum was readily apparent in neutralization assays. Thus, these results indicate that, while non-bnAbs can mask biophysical detection of bnAbs, they likely do not mask bnAb neutralization activity either in vitro or in vivo. Finally, and perhaps surprisingly, we could observe good predictions based on assessment of core-binding profiles alone. This result suggests that CD4bs-specific antibodies may represent a dominant mode of neutralization or that they are correlated with the induction of bnAbs directed at other sites, at least among the samples evaluated here.

Overall, this study demonstrates that antibody fine epitope specificity can serve as a powerful tool in neutralization breadth prediction for both mAbs and polyclonal sera. We envision that the application of this method to additional panels of antibodies and antigen variants may bring more insights into future vaccine design against HIV-1 as well as other viruses and that the STG probe in particular may prove useful in ongoing efforts to evaluate humoral responses to candidate vaccines and to identification and cloning of novel bnAbs of the CD4bs class.

mAbs. Six viAbs, including GE121, GE125, GE136, GE137, GE143, and GE148, were sourced from the Karolinska Institute ( 42 ). Seven narrowly and/or weakly neutralizing antibodies, including 448D, 559-64D, 654-30D, 1008-30D, 1202-30D, and 1263D, were sourced from the NYU School of Medicine ( 41 ). Five antibodies, including F105, b12, VRC01, and NIH45-46 G45W, were acquired from the NIH AIDS Reagent Program. Six antibodies, including CH31, CH32, CH33, CH34, CH98, and CH103, were provided by the Duke Human Vaccine Institute. The remaining CD4bs mAbs were provided by The Scripps Research Institute. The same set of mAbs was used in Figure 1, Figure 2, and Figure 3A.

Clinical samples. Serum samples from 176 HIV-infected subjects from cohorts, including chronically infected individuals from cohorts established by the Ragon Institute (n = 121) chronically infected individuals from the greater Boston area (n = 10) available from BIDMC antiretroviral drug-naive HIV-1 clade C chronically infected individuals from the Southern African National Blood services (n = 19) individuals subsequently purified for immunoglobulin at the DarDar Study in Dar es Salaam, Tanzania (n = 10) ( 59 ) and the RV217 early capture HIV cohort study (n = 16) (ECHO) collected 1–3 years after infection and prior to ART ( 60 ) were evaluated. Cohorts were not controlled for variation in viral load, CD4 count, time from infection, ART therapy, age, sex, or other factors that may influence antibody responses. IgG present in sera from HIV-infected donors was purified using Pierce Melon Gel according the manufacturer’s instructions. A 5% Triton X-100 solution in PBS was used to bring each sample to 0.5% Triton X-100 before heating at 37°C for 1 hour to inactivate virus. Neutralization activity for the Boston and African sample sets was determined using the TZM.bl assay across an 18-virus panel as previously described ( 61 ). Neutralization class identity was determined by mean value of log-transformed ID50 across the panel of HIV-1 strains.

Epitope mapping. A panel of gp120 core mutants was induced and displayed on Saccharomyces cerevisiae strain EBY100 as previously described ( 34 , 62 ). Amino acid substitutions S365K, T455L, and G459E were sequentially introduced using site-directed mutagenesis and confirmed by sequencing. Titrations were performed for each antibody sample in order to determine, first, whether a YU2 core-specific response was present and, second, to identify the dose-response inflection point. The concentration at which signal is half-maximal represents an optimal concentration for epitope mapping, at which binding to the core was most sensitive to concentration and good signal to noise resolution is observed. For both titrations and epitope maps, approximately 1 × 10 5 yeasts displaying wild-type gp120 4G core were combined with 200 μl PBS + 0.1% BSA (PBSB) per well and centrifuged in 96-well plates at 3,200 g for 4 minutes. Supernatants were removed by aspiration, and cells were resuspended in 50 μl antibody solution containing titrations of mAbs or polyclonal sera and 1:400 mouse anti-HA tag antibody (Covance) and incubated with shaking for 1 hour at room temperature for mAbs or overnight at 4°C for human sera. Yeasts were washed twice with PBSB and stained for 20 minutes at room temperature with a 1:200 solution of fluorescent goat anti-mouse and anti-human/rhesus antibodies (1:200 each) to enable detection of surface displayed core and bound core-specific antibody, respectively. Plates were washed and resuspended in 200 μl PBSB, and data were acquired on a MACSQuant Analyzer (Miltenyi Biotec). The mean fluorescent intensity (MFI) of gp120 core-displayed yeast was determined by gating on the HA tag-positive cells, and normalized MFIs were determined by determining the ratio of test antibody signal relative to the level of core display (test antibody MFI/HA tag MFI). Assays were generally performed in singlicate, as variability in the epitope mapping assay was previously evaluated, and found to exhibit an inter-study coefficient of variation (%CV) of generally less than 10% and intra-study %CVs under 5% ( 53 , 54 ).

bnAb enrichment. The STG-based bnAb enrichments were performed by depleting polyclonal pools of core-specific antibodies that recognized epitopes other than the 3-residue bnAb signature. Briefly, approximately 1.4 × 10 9 yeasts were washed with PBS before being gently pelleted for 2 minutes at 500 g in a cellulose acetate filter column (Pierce). A small volume of PBS was added without disturbing the pellet, and a 30-μm polyethylene filter was placed on top of the pellet as a frit (Pierce). PBS was removed from above the frit, and pAb sample was added to the column and centrifuged until the solution passed through the column at 250 g (approximately 15 minutes). The flow through was collected and reapplied to the column and centrifugation repeated at 500 g for approximately 10 minutes. This process was repeated on a total of 3 yeast-based affinity columns to ensure complete depletion. Enriched pAbs were either evaluated by epitope mapping using the yeast-displayed core mutants as described above or evaluated by assessing binding to STG and WT-conjugated fluorescently coded magnetic beads. Soluble STG gp120 core and WT gp120 core protein were expressed by HEK cells (Invitrogen) and purified using standard Ni-NTA chromatography. STG gp120 and WT gp120 proteins were then conjugated to the magnetic beads via primary amines through the NHS-EDC chemistry, as previously described ( 63 , 64 ). To detect pAb bound to STG/WT gp120, 30 μl enriched and unenriched pAbs were serially diluted in black, clear-bottom 384-well plates (Greiner Bio One) with a dilution factor of 4. For each specificity (STG and WT), 500 beads were added in a 20-μl volume to each well, followed by 1-hour incubation on a plate shaker at 1,050 rpm at room temperature. Plates were then washed with PBS with 1% BSA and 0.05% Tween and incubated in 50 μl anti-human IgG Fc-PE (Southern Biotech) at 650 ng/ml as a secondary antibody to detect bound pAbs for 30 minutes. Finally, plates were washed and beads were resuspended in 35 μl Luminex sheath fluid buffer. The net MFI was detected and reported by a FlexMap 3D (Luminex, Bio-Plex Manger 5.0, Bio-Rad). Area under the curve was calculated in GraphPad Prism.

Data analysis and visualization. Surface representations of YU2 gp120 core mutants and mapped epitopes were generated using PyMOL and were colored based on a modeled gp120 core structure as described previously ( 34 ). Heatmaps were plotted by the gplots package in R.3.1.0 with the heatmap.2 function and dendrograms were generated by hierarchical clustering (Euclidean distance). Classification models were built using the random forest decision tree package “randomForest” in R.3.1.0 ( 43 ). 20,000 decision trees were built for each binomial trainer per run. The relative importance of each epitope-mapping measurement to classification models was evaluated by the mean decrease in Gini index. Predictive accuracy was assessed by leave-one-out cross-validation. Model quality was also assessed by permutation tests, in which neutralization class identity was randomized with a fixed ratio between bnAbs and nnAbs and the classifier performance was evaluated over 1,000 different permutations. Classification models with accuracies greater than 2 SDs above than the permuted model mean were regarded as significant. As methodological alternatives, elastic net models were also built using the Glmnet package ( 65 ), with an elastic net mixing parameter at 0.4, and use of leave-one-out cross-validation to determine the value of the tuning parameter lambda, such that minimum cross-validated mis-classification error was observed. Similar classification performance and common contributing features were observed across random forest and elastic net methods. Results from random forest models were selected for presentation, as this method is robust to outliers, computationally efficient, and resistant to overfitting ( 66 ).

Statistics. Subjects were clustered into groups defined by CD4bs epitope-mapping data, and the median neutralization ID50 values observed for individuals from each group were compared. A P value of less than 0.01 was considered significant. Classification models were considered to perform significantly better than expected at random if their accuracies were greater than 2 SDs above the mean accuracy observed from models learned from permuted data. Comparison of neutralization ID50 values was conducted by Mann-Whitney test in R 3.3.1 with the wilcox.test function.

Study approval. All subjects were adults, and they provided written informed consent. The Dartmouth College Committee for the Protection of Human Subjects approved the study.

HDC, SKG, MSAG, and MEA conceived of and designed the study. SJK, MSS, CBK, and MEA supervised experimental and statistical analysis. HDC, SKG, MSAG, LCG, and GD performed experiments. DS, CS, GBKH, MB, BFH, TPL, IM, CFVR, MKG, SZP, BDW, GA, DRB, MLR, and MSS provided critical reagents and reviewed data resulting from their use. HDC and MEA wrote the manuscript. All authors critically reviewed the manuscript text.

These studies were conducted with support from the NIH/National Institute Allergy and Infectious Disease (1R01AI102691, 1P01AI120756, UM-1 AI100645) Division of AIDS Center for HIV/AIDS Vaccine Immunology & Immunogen Discovery (1UM1AI100663) the NIH-funded Center for AIDS Research (P30 AI060354), which is supported by the following NIH cofunding and participating institutes and centers: the National Institute of Allergy and Infectious Diseases, National Cancer Institute, National Institute of Child Health and Human Development, National Heart, Lung, and Blood Institute, National Institute on Drug Abuse, National Institute of Mental Health, National Institute on Aging, Fogarty International Center, and the Office of AIDS Research the Bill and Melinda Gates Foundation (OPP1114729, OPP1066973, OPP1146433, and OPP1146996) and the NIH and Fogarty International Center for The Dartmouth/Boston University AIDS (TRIM-TB D43TW009573). LCG was supported by the European Union’s FP7 PRD COLLEGE Science Exchange Programme. The authors thank Bonnie Slike for technical support and Ann Hoen for statistical support. The views expressed are those of the authors and should not be construed to represent the positions of the US Army or the Department of Defense. This work was supported by a cooperative agreement (W81XWH-11-2-0174) between the Henry M. Jackson Foundation for the Advancement of Military Medicine Inc. and the US Department of Defense.

Conflict of interest: Authors of this publication report pending or awarded intellectual property related to antibodies whose epitopes are mapped in this study.


Antigenic Determinants and Processing Pathways

Antigen epitopes make it possible for the immune system to recognize pathogens.

Learning Objectives

Describe antigenic determinants and pathways of processing

Key Takeaways

Key Points

  • An epitope (also known as an antigenic determinant) is part of an antigen that is recognized by the immune system, specifically by antibodies and B and T cells. Other immune cells like APCs cannot recognize epitopes (only PAMPS and DAMPS).
  • Antigenic determinants (epitopes) are divided into conformational epitopes and linear epitopes.
  • Antigen processing occurs within a cell and results in fragmentation of proteins, association of the fragments with MHC molecules, and expression of the peptide -MHC molecules at the cell surface where they can be recognized by the T cell receptor on a T cell.
  • Antigen processing may be done through either the endogenous pathway (viral proteins from within an infected cell) or through the exogenous pathway (engulfing a pathogen and isolating its antigen from within the APC).
  • The endogenous pathway uses MHC class I and binds to cytotoxic T cells, while the exogenous pathway uses MHC class II and binds to helper T cells.
  • Some viruses can prevent antigen processing by disrupting movement of MHC within the cell.

Key Terms

  • Linear epitopes: These consist of the primary amino acid structure of a protein that makes up the larger antigen.
  • The Exogenous Pathway: Phagocytized pathogens are broken down from within the cell and their broken-down antigens bind with MHC II, which then is expressed on the surface of the antigen-presenting cell.

An epitope, also known as an antigenic determinant, is the part of an antigen that is recognized by the immune system, specifically by antibodies, B cells, and T cells. The latter can use epitopes to distinguish between different antigens, and only binds to their specific antigen. In antibodies, the binding site for an epitope is called a paratope. Although epitopes are usually derived from non-self proteins, sequences derived from the host that can be recognized are also classified as epitopes. Epitopes determine how antigen binding and antigen presentation occur.

Types of Antigenic Determinants

The epitopes of protein antigens are divided into two categories based on their structures and interaction with the paratope.

  • A conformational epitope is composed of discontinuous sections of the antigen’s amino acid sequence. These epitopes interact with the paratope based on the 3-D surface features and tertiary structure (overall shape) of the antigen. Most epitopes are conformational.
  • Linear epitopes interact with the paratope based on their primary structure (shape of the protein’s components). A linear epitope is formed by a continuous sequence of amino acids from the antigen, which creates a “line” of sorts that builds the protein structure.

Antigenic determinants recognized by B cells and the antibodies secreted by B cells can be either conformational or linear epitopes. Antigenic determinants recognized by T cells are typically linear epitopes. T cells do not recognize polysaccharide or nucleic acid antigens. This is why polysaccharides are generally T-independent antigens and proteins are generally T-dependent antigens. The determinants need not be located on the exposed surface of the antigen in its original form, since recognition of the determinant by T cells requires that the antigen be first processed by antigen presenting cells. Free peptides flowing through the body are not recognized by T cells, but the peptides associate with molecules coded for by the major histocompatibility complex (MHC). This combination of MHC molecules and peptide is recognized by T cells.

Antigen-Processing Pathways

Antigen-Binding Site of an Antibody: Antigen-binding sites can recognize different epitopes on an antigen.

In order for an antigen-presenting cell (APC) to present an antigen to a naive T cell, it must first be processed so itacan be recognized by the T cell receptor. This occurs within an APC that phagocytizes an antigen and then digests it through fragmentation (proteolysis) of the antigen protein, association of the fragments with MHC molecules, and expression of the peptide-MHC molecules at the cell surface. There, they are recognized by the T cell receptor on a T cell during antigen presentation. MHC molecules must move between the cell membrane and cytoplasm in order for antigen processing to occur properly. However, the pathway leading to the association of protein fragments with MHC molecules differs between class I and class II MHC, which are presented to cytotoxic or helper T cells respectively. There are two different pathways for antigen processing:

  • The endogenous pathway occurs when MHC class I molecules present antigens derived from intracellular (endogenous) proteins in the cytoplasm, such as the proteins produced within virus-infected cells. Generally, proteosomes are used to break up the viral proteins and combine them with MHC I.
  • The exogenous pathway occurs when MHC class II molecules present fragments derived from extracellular (exogenous) proteins that are located within the cell. First, pathogens are phagocytized, then endosomes within the cell break down antigens with proteases, which then combine with MHC II.

Some viral pathogens have developed ways to evade antigen processing. For example, cytomegalovirus and HIV-infected cells sometimes disrupt MHC movement through the cytoplasm, which may prevent them from binding to antigens or from moving back to the cell membrane after binding with an antigen.


Watch the video: isotype allotype idiotype (May 2022).