Search

This review "teaches" researchers how to make their lackluster proteomics data look really impressive, by applying an inappropriate but pervasive strategy that selects features in a biased manner. The strategy is demonstrated and used to build a classification model with an accuracy of 92% and AUC of 0.98, while relying completely on random numbers for the data set. This "lesson" in data processing is not to be practiced by anyone; on the contrary, it is meant to be a cautionary tale showing that very unreliable results are obtained when a biomarker panel is generated first, using all the available data, and then tested by cross-validation. Data scientists describe the error committed in this scenario as having test data leak into the feature selection step, and it is currently a common mistake in proteomics biomarker studies that rely on machine learning. After the demonstration, advice is provided about how machine learning methods can be applied to proteomics data sets without generating artificially inflated accuracies.

Recent studies have highlighted that the proteome can be used to identify potential biomarker candidates for Alzheimer’s disease (AD) in diverse cohorts. Furthermore, the racial and ethnic background of participants is an important factor to consider to ensure the effectiveness of potential biomarkers for representative populations. A promising approach to survey potential biomarker candidates for diagnosing AD in diverse cohorts is the application of machine learning to proteomics datasets. Herein, we leveraged six existing bottom-up proteomics datasets, which included non-Hispanic White, African American/Black, and Hispanic participants, to study protein changes in AD and cognitively unimpaired participants. Machine learning models were applied to these datasets and resulted in the identification of amyloid-β precursor protein (APP) and heat shock protein β-1 (HSPB1) as two proteins that have high ability to distinguish AD; however, each protein’s performance varied based upon the racial and ethnic background of the participants. HSPB1 particularly was helpful for generating high areas under the curve (AUCs) for African American/Black participants. Overall, HSPB1 improved the performance of the machine learning models when combined with APP and/or participant age, and it is a potential candidate that should be further explored in AD biomarker discovery efforts.

The fields of proteomics and machine learning are both large disciplines, each producing well over 5,000 publications per year. However, studies combining both fields are still relatively rare, with only about 2% of recent proteomics papers including machine learning. This review, which focuses on the intersection of the fields, is intended to inspire proteomics researchers to develop skills and knowledge in the application of machine learning. A brief tutorial introduction to machine learning is provided, and research advances that rely on both fields, particularly as they relate to proteomics tools development and biomarker discovery, are highlighted. Key knowledge gaps and opportunities for scientific advancement are also enumerated.

Mass spectrometry data sets from ‘omics studies are an optimal information source for discriminating patients with disease and identifying biomarkers. Thousands of proteins or endogenous metabolites can be queried in each analysis, spanning several orders of magnitude in abundance. Machine learning tools that effectively leverage these data to accurately identify disease states are in high demand. While mass spectrometry data sets are rich with potentially useful information, using the data effectively can be challenging because of missing entries in the data sets and because the number of samples is typically much smaller than the number of features, two challenges that make machine learning difficult. To address this problem, we have modified a new supervised classification tool, the Aristotle Classifier, so that ‘omics data sets can be better leveraged for identifying disease states. The optimized classifier, AC.2021, is benchmarked on multiple data sets against its predecessor and two leading supervised classification tools, Support Vector Machine (SVM) and XGBoost. The new classifier, AC.2021, outperformed existing tools on multiple tests using proteomics data. The underlying code for the classifier, provided herein, would be useful for researchers who desire improved classification accuracy when using their ‘omics data sets to identify disease states.

Uromodulin, also known as the Tamm-Horsfall protein or THP, is the most abundant protein excreted in human urine. It is associated with the progression of kidney diseases; therefore, changes in the glycosylation profile of this protein could serve as a potential biomarker for kidney health. The typical glycomics analysis approaches used to quantify uromodulin glycosylation involve time-consuming and tedious glycoprotein isolation and labeling steps, which limit their utility in clinical glycomics assays, where sample throughput is important. Herein, we introduce a radically simplified sample preparation workflow, with direct ESI-MS analysis, enabling the quantification of N-linked glycans that originate from uromodulin. The method omits any glycan labeling steps but includes steps to reduce the salt content of the samples, thereby minimizing ion suppression. The method is effective for quantifying subtle glycosylation differences of uromodulin samples derived from different biological states. As a proof of concept, glycosylation from samples that differ by pregnancy status were shown to be differentiable.

Transferrins function in iron sequestration and iron transport by binding iron tightly and reversibly. Vertebrate transferrins coordinate iron through interactions with two tyrosines, an aspartate, a histidine, and a carbonate anion, and conformational changes that occur upon iron binding and release have been described. Much less is known about the structure and functions of insect transferrin‐1 (Tsf1), which is present in hemolymph and influences iron homeostasis mostly by unknown mechanisms. Amino acid sequence and biochemical analyses have suggested that iron coordination by Tsf1 differs from that of the vertebrate transferrins. Here we report the first crystal structure (2.05 Å resolution) of an insect transferrin. Manduca sexta (MsTsf1) in the holo form exhibits a bilobal fold similar to that of vertebrate transferrins, but its carboxyl‐lobe adopts a novel orientation and contacts with the amino‐lobe. The structure revealed coordination of a single Fe(3+) ion in the amino‐lobe through Tyr90, Tyr204, and two carbonate anions. One carbonate anion is buried near the ferric ion and is coordinated by four residues, whereas the other carbonate anion is solvent exposed and coordinated by Asn121. Notably, these residues are highly conserved in Tsf1 orthologs. Docking analysis suggested that the solvent exposed carbonate position is capable of binding alternative anions. These findings provide a structural basis for understanding Tsf1 function in iron sequestration and transport in insects as well as insight into the similarities and differences in iron homeostasis between insects and humans.

The functional human immunodeficiency virus (HIV-1) envelope glycoprotein (Env) trimer [(gp120/gp41)(3)] is produced by cleavage of a conformationally flexible gp160 precursor. gp160 cleavage or the binding of BMS-806, an entry inhibitor, stabilizes the pretriggered, “closed” (state 1) conformation recognized by rarely elicited broadly neutralizing antibodies. Poorly neutralizing antibodies (pNAbs) elicited at high titers during natural infection recognize more “open” Env conformations (states 2 and 3) induced by binding the receptor, CD4. We found that BMS-806 treatment and cross-linking decreased the exposure of pNAb epitopes on cell surface gp160; however, after detergent solubilization, cross-linked and BMS-806-treated gp160 sampled non-state-1 conformations that could be recognized by pNAbs. Cryo-electron microscopy of the purified BMS-806-bound gp160 revealed two hitherto unknown asymmetric trimer conformations, providing insights into the allosteric coupling between trimer opening and structural variation in the gp41 HR1(N) region. The individual protomer structures in the asymmetric gp160 trimers resemble those of other genetically modified or antibody-bound cleaved HIV-1 Env trimers, which have been suggested to assume state-2-like conformations. Asymmetry of the uncleaved Env potentially exposes surfaces of the trimer to pNAbs. To evaluate the effect of stabilizing a state-1-like conformation of the membrane Env precursor, we treated cells expressing wild-type HIV-1 Env with BMS-806. BMS-806 treatment decreased both gp160 cleavage and the addition of complex glycans, implying that gp160 conformational flexibility contributes to the efficiency of these processes. Selective pressure to maintain flexibility in the precursor of functional Env allows the uncleaved Env to sample asymmetric conformations that potentially skew host antibody responses toward pNAbs. IMPORTANCE The envelope glycoprotein (Env) trimers on the surface of human immunodeficiency virus (HIV-1) mediate the entry of the virus into host cells and serve as targets for neutralizing antibodies. The functional Env trimer is produced by cleavage of the gp160 precursor in the infected cell. We found that the HIV-1 Env precursor is highly plastic, allowing it to assume different asymmetric shapes. This conformational plasticity is potentially important for Env cleavage and proper modification by sugars. Having a flexible, asymmetric Env precursor that can misdirect host antibody responses without compromising virus infectivity would be an advantage for a persistent virus like HIV-1.

MALDI-TOF MS has shown great utility for rapidly identifying microbial species. It can be used to successfully type bacteria and fungi from a variety of sources more rapidly and cost-effectively than traditional methods. One area where improvements are necessary is in the typing of highly similar samples, such as those samples from the same genus but different species or samples from within a single species but from different strains. One promising way to address this current limitation is by using advanced machine learning techniques. In this work, we adapt a newly developed machine learning tool, the Aristotle Classifier, to bacterial classification of MALDI-TOF MS data. This tool was originally developed for classifying glycomics and glycoproteomics data, so we modified it to be well-suited for assigning mass spectral data from bacterial proteins. The classifier exceeds existing benchmarks in classifying bacteria, and it shows particularly strong performance when the samples to be identified are highly similar. The combination of mass spectrometry data and tools like the Aristotle Classifier could ameliorate the ambiguities associated with challenging bacterial classification problems.

Glycopeptide analysis is a growing field that is struggling to adopt effective, automated tools. Many creative workflows and software apps have emerged recently that offer promising capabilities for assigning glycopeptides to MS data in an automated fashion. The effectiveness of these tools is best measured and improved by determining how often they would select a glycopeptide decoy as a spectral match, instead of its correct assignment; yet generating the appropriate number and type of glycopeptide decoys can be challenging. To address this need, we have designed DecoyDeveloper, an on-demand decoy glycopeptide generator that can produce a high volume of decoys with low mass differences. DecoyDeveloper has a simple user interface and is capable of producing large sets of decoys containing complete, biologically relevant glycan and peptide sequences. We demonstrate the tool's efficiency by applying it to a set of 80 glycopeptide targets. This tool is freely available and can be found at http://glycopro.chem.ku.edu/J1.php .

Due to the important pathological roles of the HIV-1 gp120, the protein has been intensively used in the research of HIV. However, recombinant gp120 preparation has proven to be difficult because of extremely low expression levels. In order to facilitate gp120 expression, previous methods predominantly involved the replacement of native signal peptide with a heterologous one, resulting in very limited improvement. Currently, preparation of recombinant gp120 with native glycans relies solely on transient expression systems, which are not amendable for large scale production. In this work, we employed a different approach for gp120 expression. Besides replacing the native gp120 signal peptide with that of rat serum albumin and optimizing its codon usage, we generated a stable gp120-expressing cell line in a glutamine synthetase knockout HEK293T cell line that we established for the purpose of amplification of recombinant gene expressions. The combined usage of these techniques dramatically increased gp120 expression levels and yielded a functional product with human cell derived glycan. This method may be applicable to large scale preparation of other viral envelope proteins, such as that of the emerging SARS-CoV-2, or other glycoproteins which require the presence of authentic human glycans.

The SARS-CoV-2 coronavirus, the etiologic agent of COVID-19, uses its spike (S) glycoprotein anchored in the viral membrane to enter host cells. The S glycoprotein is the major target for neutralizing antibodies elicited by natural infection and by vaccines. Approximately 35% of the SARS-CoV-2 S glycoprotein consists of carbohydrate, which can influence virus infectivity and susceptibility to antibody inhibition. We found that virus-like particles produced by co-expression of SARS-CoV-2 S, M, E and N proteins contained spike glycoproteins that were extensively modified by complex carbohydrates. We used a fucose-selective lectin to purify the Golgi-modified fraction of a wild-type SARS-CoV-2 S glycoprotein trimer, and determined its glycosylation and disulfide bond profile. Compared with soluble or solubilized S glycoproteins modified to prevent proteolytic cleavage and to retain a prefusion conformation, more of the wild-type S glycoprotein N-linked glycans are processed to complex forms. Even Asn 234, a significant percentage of which is decorated by high-mannose glycans on other characterized S trimer preparations, is predominantly modified in the Golgi compartment by processed glycans. Three incompletely occupied sites of O-linked glycosylation were detected. Viruses pseudotyped with natural variants of the serine/threonine residues implicated in O-linked glycosylation were generally infectious and exhibited sensitivity to neutralization by soluble ACE2 and convalescent antisera comparable to that of the wild-type virus. Unlike other natural cysteine variants, a Cys15Phe (C15F) mutant retained partial, but unstable, infectivity. These findings enhance our understanding of the Golgi processing of the native SARS-CoV-2 S glycoprotein carbohydrates and could assist the design of interventions. IMPORTANCE The SARS-CoV-2 coronavirus, which causes COVID-19, uses its spike glycoprotein to enter host cells. The viral spike glycoprotein is the main target of host neutralizing antibodies that help to control SARS-CoV-2 infection and are important for the protection provided by vaccines. The SARS-CoV-2 spike glycoprotein consists of a trimer of two subunits covered with a coat of carbohydrates (sugars). Here, we describe the disulfide bonds that assist the SARS-CoV-2 spike glycoprotein to assume the correct shape, and the composition of the sugar moieties on the glycoprotein surface. We also evaluate the consequences of natural virus variation in O-linked sugar addition and in the cysteine residues involved in disulfide bond formation. This information can expedite the improvement of vaccines and therapies for COVID-19.

The functional human immunodeficiency virus (HIV-1) envelope glycoprotein (Env) trimer [(gp120/gp41)3] is produced by cleavage of a conformationally flexible gp160 precursor. Gp160 cleavage or the binding of BMS-806, an entry inhibitor, stabilizes the pre-triggered, “closed” (State-1) conformation recognized by rarely elicited broadly neutralizing antibodies. Poorly neutralizing antibodies (pNAbs) elicited at high titers during natural infection recognize more “open” Env conformations (States 2 and 3) induced by binding the receptor, CD4. We found that BMS-806 treatment and crosslinking decreased the exposure of pNAb epitopes on cell-surface gp160; however, after detergent solubilization, crosslinked and BMS-806-treated gp160 sampled non-State-1 conformations that could be recognized by pNAbs. Cryo-electron microscopy of the purified BMS-806-bound gp160 revealed two hitherto unknown asymmetric trimer conformations, providing insights into the allosteric coupling between trimer opening and structural variation in the gp41 HR1N region. The individual protomer structures in the asymmetric gp160 trimers resemble those of other genetically modified or antibody-bound cleaved HIV-1 Env trimers, which have been suggested to assume State-2-like conformations. Asymmetry of the uncleaved Env potentially exposes surfaces of the trimer to pNAbs. To evaluate the effect of stabilizing a State-1-like conformation of the membrane Env precursor, we treated cells expressing wild-type HIV-1 Env with BMS-806. BMS-806 treatment decreased both gp160 cleavage and the addition of complex glycans, implying that gp160 conformational flexibility contributes to the efficiency of these processes. Selective pressure to maintain flexibility in the precursor of functional Env allows the uncleaved Env to sample asymmetric conformations that potentially skew host antibody responses toward pNAbs.IMPORTANCEThe envelope glycoprotein (Env) trimers on the surface of human immunodeficiency virus (HIV-1) mediate the entry of the virus into host cells and serve as targets for neutralizing antibodies. The functional Env trimer is produced by cleavage of the gp160 precursor in the infected cell. We found that the HIV-1 Env precursor is highly plastic, allowing it to assume different asymmetric shapes. This conformational plasticity is potentially important for Env cleavage and proper modification by sugars. Having a flexible, asymmetric Env precursor that can misdirect host antibody responses without compromising virus infectivity would be an advantage to a persistent virus like HIV-1.

Background: African American/Black adults have a disproportionate incidence of Alzheimer’s disease (AD) and are underrepresented in biomarker discovery efforts. Objective: This study aimed to identify potential diagnostic biomarkers for AD using a combination of proteomics and machine learning approaches in a cohort that included African American/Black adults. Methods: We conducted a discovery-based plasma proteomics study on plasma samples (N = 113) obtained from clinically diagnosed AD and cognitively normal adults that were self-reported African American/Black or non-Hispanic White. Sets of differentially-expressed proteins were then classified using a support vector machine (SVM) to identify biomarker candidates. Results: In total, 740 proteins were identified of which, 25 differentially-expressed proteins in AD came from comparisons within a single racial and ethnic background group. Six proteins were differentially-expressed in AD regardless of racial and ethnic background. Supervised classification by SVM yielded an area under the curve (AUC) of 0.91 and accuracy of 86%for differentiating AD in samples from non-Hispanic White adults when trained with differentially-expressed proteins unique to that group. However, the same model yielded an AUC of 0.49 and accuracy of 47%for differentiating AD in samples from African American/Black adults. Other covariates such as age, APOE4 status, sex, and years of education were found to improve the model mostly in the samples from non-Hispanic White adults for classifying AD. Conclusion: These results demonstrate the importance of study designs in AD biomarker discovery, which must include diverse racial and ethnic groups such as African American/Black adults to develop effective biomarkers.

Here we present a plasma proteomics dataset that was generated to understand the importance of self-reported race for biomarker discovery in Alzheimer's disease. This dataset is related to the article “Why inclusion matters for Alzheimer's disease biomarker discovery in plasma” [1]. Plasma samples were obtained from clinically diagnosed Alzheimer's disease and cognitively normal adults of African American/Black and non-Hispanic White racial and ethnic backgrounds. Plasma was immunodepleted, digested, and isobarically tagged with commercial reagents. Tagged peptides were fractionated using high pH fractionation and resulting fractions analysed by liquid chromatography – mass spectrometry (LC-MS/MS & MS3) analysis on an Orbitrap Fusion Lumos mass spectrometer. The resulting data was processed using Proteome Discoverer to produce a list of identified proteins with corresponding tandem mass tag (TMT) intensity information.

Small-molecule viral entry inhibitors, such as BMS-626529 (BMS-529), allosterically block CD4 binding to HIV-1 envelope (Env) and inhibit CD4-induced structural changes in Env trimers. Here, we show that the binding of BMS-529 to clade C soluble chimeric gp140 SOSIP (ch.SOSIP) and membrane-bound trimers with intact transmembrane domain (gp150) prevented trimer conformational transitions and enhanced their immunogenicity. When complexed to BMS-529, ch.SOSIP trimers retained their binding to broadly neutralizing antibodies (bNAbs) and to their unmutated common ancestor (UCA) antibodies, while exposure of CD4-induced (CD4i) non-bNAb epitopes was inhibited. BMS-529-complexed gp150 trimers in detergent micelles, which were isolated from CHO cells, bound to bNAbs, including UCA and intermediates of the CD4 binding site (bs) CH103 bNAb lineage, and showed limited exposure of CD4i epitopes and a glycosylation pattern with a preponderance of high-mannose glycans. In rabbits, BMS-529-complexed V3 glycan-targeting ch.SOSIP immunogen induced in the majority of immunized animals higher neutralization titers against both autologous and select high mannose-bearing heterologous tier 2 pseudoviruses than those immunized with the noncomplexed ch.SOSIP. In rhesus macaques, BMS-529 complexed to CD4 bs-targeting ch.SOSIP immunogen induced stronger neutralization against tier 2 pseudoviruses bearing high-mannose glycans than noncomplexed ch.SOSIP trimer immunogen. When immunized with gp150 complexed to BMS-529, rhesus macaques showed neutralization against tier 2 pseudoviruses with targeted glycan deletion and high-mannose glycan enrichment. These results demonstrated that stabilization of Env trimer conformation with BMS-529 improved the immunogenicity of select chimeric SOSIP trimers and elicited tier 2 neutralizing antibodies of higher potency than noncomplexed trimers. IMPORTANCE Soluble forms of HIV-1 envelope trimers exhibit conformational heterogeneity and undergo CD4-induced (CD4i) exposure of epitopes of non-neutralizing antibodies that can potentially hinder induction of broad neutralizing antibody responses. These limitations have been mitigated through recent structure-guided approaches and include trimer-stabilizing mutations that resist trimer conformational transition and exposure of CD4i epitopes. The use of small-molecule viral inhibitors that allosterically block CD4 binding represents an alternative strategy for stabilizing Env trimer in the pre-CD4-triggered state of both soluble and membrane-bound trimers. In this study, we report that the viral entry inhibitor BMS-626529 restricts trimer conformational transition and improves the immunogenicity of select Env trimer immunogens.

One unifying challenge when classifying biological samples with mass spectrometry data is overcoming the obstacle of sample-to-sample variability so that differences between groups, such as between a healthy set and a disease set, can be identified. Similarly, when the same sample is re-analyzed under identical conditions, instrument signals can fluctuate by more than 10%. This signal inconsistency imposes difficulties in identifying subtle differences across a set of samples, and it weakens the mass spectrometrist’s ability to effectively leverage data in domains as diverse as proteomics, metabolomics, glycomics, and imaging. We selected challenging data sets in the fields of glycomics, mass spectrometry imaging, and bacterial typing to study the problem of within-group signal variability and adapted a 30-year-old statistical approach to address the problem. The solution, “local-balanced model,” relies on using balanced subsets of training data to classify test samples. This analysis strategy was assessed on ESI-MS data of IgG-based glycopeptides and MALDI-MS imaging data of endogenous lipids, and MALDI-MS data of bacterial proteins. Two preliminary examples on non-mass spectrometry data sets are also included to show the potential generality of the method outside the field of MS analysis. We demonstrate that this approach is superior to simple normalization methods, generalizable to multiple mass spectrometry domains, and potentially appropriate in fields as diverse as physics and satellite imaging. In some cases, improvements in classification can be dramatic, with accuracy escalating from 60% with normalization alone to over 90% with the additional development described herein.

As the field of mass spectrometry imaging continues to grow, so too do its needs for optimal methods of data analysis. One general need in image analysis is the ability to classify the underlying regions within an image, as healthy or diseased, for example. Classification, as a general problem, is often best accomplished by supervised machine learning strategies; unfortunately, conducting supervised machine learning on MS imaging files is not typically done by mass spectrometrists because a high degree of specialized knowledge is needed. To address this problem, we developed a fully open-source approach that facilitates supervised machine learning on MS imaging files, and we demonstrated its implementation on sets of cancer spheroids that either had or had not undergone chemotherapy treatment. These supervised machine learning studies demonstrated that metabolic changes induced by the monoclonal antibody, Cetuximab, are detectable but modest at 24 hours, and by 72 hours, the drug induces a larger and more diverse metabolic response.

In the RV144 HIV-1 phase III trial, vaccine efficacy directly correlated with the magnitude of the variable region 2–specific (V2-specific) IgG antibody response, and in the presence of low plasma IgA levels, with the magnitude of plasma antibody-dependent cellular cytotoxicity. Reenrollment of RV144 vaccinees in the RV305 trial offered the opportunity to define the function, maturation, and persistence of vaccine-induced V2-specific and other mAb responses after boosting. We show that the RV144 vaccine regimen induced persistent V2 and other HIV-1 envelope–specific memory B cell clonal lineages that could be identified throughout the approximately 11-year vaccination period. Subsequent boosts increased somatic hypermutation, a critical requirement for antibody affinity maturation. Characterization of 22 vaccine-induced V2-specific mAbs with epitope specificities distinct from previously characterized RV144 V2-specific mAbs CH58 and CH59 found increased in vitro antibody-mediated effector functions. Thus, when inducing non-neutralizing antibodies, one method by which to improve HIV-1 vaccine efficacy may be through late boosting to diversify the V2-specific response to increase the breadth of antibody-mediated anti–HIV-1 effector functions.

Human lysyl oxidase-like 2 (hLOXL2), a glycoprotein implicated in tumor progression and organ fibrosis, is a molecular target for anticancer and antifibrosis treatment. This glycoprotein contains three predicted N-linked glycosylation sites; one is near the protein's active site, and at least one more is known to facilitate the protein's secretion. Because the glycosylation impacts the protein's biology, we sought to characterize the native, mammalian glycosylation profile and to determine how closely this profile is recapitulated when the protein is expressed in insect cells. All three glycosylation sites on the protein, expressed in human embryonic kidney (HEK) cells, were characterized individually using a mass spectrometry-based glycopeptide analysis workflow. These data were compared to the glycosylation profile of the same protein expressed in insect cells. We found that the producer cell type imparts a substantial influence on the glycosylation of this important protein. The more-relevant version, expressed in HEK cells, contains large, acidic glycoforms; these glycans are not generated in insect cells. The glycosylation differences likely have structural and functional consequences, and these data should be considered when generating protein for functional studies or for high-throughput screening campaigns.

The need to investigate the fragmentation of fucosylated glycopeptides is driven by recent work showing that at least one, and perhaps many, glycopeptide analysis scoring algorithms are less effective at identifying fucosylated glycopeptides than non-fucosylated glycopeptides. Herein, we study the CID fragmentation characteristics of fucosylated glycopeptides and the scoring rules of the glycopeptide analysis software, GlycoPep Grader, in an effort to improve automated assignments of these important glycopeptides. We identified some prominent product ions from a common fragmentation pathway of fucosylated glycopeptides that were not accounted for in the scoring rules. Based on this finding, we propose new scoring rules for fucosylated glycopeptides that can be incorporated into GlycoPep Grader and other similar analysis software tools to more accurately identify these species. The approach used here, to improve one particular scoring algorithm, could henceforth be used to improve any other algorithm that assigns glycopeptides based on their MS/MS data.

The glycopeptide analysis field is tightly constrained by a lack of effective tools that translate mass spectrometry data into meaningful chemical information, and perhaps the most challenging aspect of building effective glycopeptide analysis software is designing an accurate scoring algorithm for MS/MS data. Herein, we provide the glycoproteomics community with two tools to address this challenge. The first tool, a curated set of 100 expert-assigned CID spectra of glycopeptides, contains a diverse set of spectra from a variety of glycan types; the second tool, Glycopeptide Decoy Generator, is a new software application that generates glycopeptide decoys de novo. We developed these tools so that emerging methods of assigning glycopeptides’ CID spectra could be rigorously tested. Software developers or those interested in developing skills in expert (manual) analysis can use these tools to facilitate their work. We demonstrate the tools’ utility in assessing the quality of one particular glycopeptide software package, GlycoPep Grader, which assigns glycopeptides to CID spectra. We first acquired the set of 100 expert assigned CID spectra; then we used the Decoy Generator (described herein) to generate 20 decoys per target glycopeptide. The assigned spectra and decoys were used to test the accuracy of GlycoPep Grader’s scoring algorithm; new strengths and weaknesses were identified in the algorithm using this approach. Both newly-developed tools are freely available to interested parties. The software can be downloaded at http://glycopro.chem.ku.edu/GPJ.jar

Summary: Induction of broadly neutralizing antibodies (bnAbs) that target HIV-1 envelope (Env) is a goal of HIV-1 vaccine development. A bnAb target is the Env third variable loop (V3)-glycan site. To determine whether immunization could induce antibodies to the V3-glycan bnAb binding site, we repetitively immunized macaques over a 4-year period with an Env expressing V3-high mannose glycans. Env immunizations elicited plasma antibodies that neutralized HIV-1 expressing only high-mannose glycans—a characteristic shared by early bnAb B cell lineage members. A rhesus recombinant monoclonal antibody from a vaccinated macaque bound to the V3-glycan site at the same amino acids as broadly neutralizing antibodies. A structure of the antibody bound to glycan revealed that the three variable heavy-chain complementarity-determining regions formed a cavity into which glycan could insert and neutralized multiple HIV-1 isolates with high-mannose glycans. Thus, HIV-1 Env vaccination induced mannose-dependent antibodies with characteristics of V3-glycan bnAb precursors. : Most bnAb epitopes on HIV-1 Envelope include host glycans, but previous Env vaccines have not induced glycan-dependent antibodies. Saunders et al. describe here the ontogeny, crystal structure with glycan, and virion Man9GlcNAc2-dependent neutralization for glycan-reactive antibodies induced by envelope vaccination. Keywords: HIV, V3 glycan, vaccination, glycan, long-term immunization

Glycomic-based approaches to discover potential biomarkers have shown great promise in their ability to distinguish between healthy and diseased individuals; these methods can identify when aberrant glycosylation is significant, but they cannot practically be adapted into widely implemented diagnostic assays because they are too complex, expensive, and low-throughput. We have developed a new strategy that addresses challenges associated with sample preparation, sample throughput, instrumentation needs, and data analysis to transfer the valuable knowledge provided by protein glycosylation into a clinical environment. Notably, the detection limits of the assay are in the single-digit picomole range. Proof of principle is demonstrated by quantifying the changes in the sialic acid content in fetuin. As the sialic acid content in proteins varies in a number of disease states, this example demonstrates the utility of the method for biomarker analysis. Furthermore, the developed method can be adapted to other biologically important saccharides, affording a broad array of quantitative glycomic analyses that are accessible in a high-throughput, plate-reader format. These studies enable glycomic-based biomarker discovery efforts to transition through the difficult landscape of developing a potential biomarker into a clinical assay.

“The totality is not, as it were, a mere heap, but the whole is something besides the parts.” --Aristotle. We built a classifier that uses the totality of the glycomic profile, not restricted to a few glycoforms, to differentiate samples from two different sources. This approach, which relies on using thousands of features, is a radical departure from current strategies, where most of the glycomic profile is ignored in favor of selecting a few features, or even a single feature, meant to capture the differences in sample types. The classifier can be used to differentiate the source of the material; applicable sources may be different species of animals, different protein production methods, or, most importantly, different biological states (disease vs. healthy). The classifier can be used on glycomic data in any form, including derivatized monosaccharides, intact glycans, or glycopeptides. It takes advantage of the fact that changing the source material can cause a change in the glycomic profile in many subtle ways; some glycoforms can be upregulated; some downregulated; some may appear unchanged, yet their proportion – with respect to other forms present – can be altered to a detectable degree. By classifying samples using the entirety of their glycan abundances, along with the glycans’ relative proportions to each other, The Aristotle Classifier is more effective at capturing the underlying trends than standard classification procedures used in glycomics, including PCA (Principal Components Analysis). It also outperforms workflows where a single, representative glycomic-based biomarker is used to classify samples. We describe The Aristotle Classifier and provide several examples of its utility for biomarker studies and other classification problems using glycomic data from several sources.

INTRODUCTION A major goal of HIV-1 vaccine development is the design of immunogens that induce broadly neutralizing antibodies (bnAbs). However, vaccination of humans has not resulted in the induction of affinity-matured and potent HIV-1 bnAbs. To devise effective vaccine strategies, we previously determined the maturation pathway of select HIV-1 bnAbs from acute infection through neutralizing antibody development. During their evolution, bnAbs acquire an abundance of improbable amino acid substitutions as a result of nucleotide mutations at variable region sequences rarely targeted by activation-induced cytidine deaminase, the enzyme responsible for antibody mutation. A subset of improbable mutations is essential for broad neutralization activity, and their acquisition represents a key roadblock to bnAb development. RATIONALE Current bnAb lineage–based vaccine strategies can initiate bnAb lineage development in animal models but have not specifically elicited the improbable mutations required for neutralization breadth. Induction of bnAbs requires vaccine strategies that specifically engage bnAb precursors and subsequently select for improbable mutations required for broadly neutralizing activity. We hypothesized that vaccination with immunogens that bind with moderate to high affinity to bnAb B cell precursors, and with higher affinity to precursors that have acquired improbable mutations, could initiate bnAb B cell lineages and select for key improbable mutations required for bnAb development. RESULTS We elicited serum neutralizing HIV-1 antibodies in human bnAb precursor knock-in mice and wild-type macaques vaccinated with immunogens designed to select for improbable mutations. We designed two HIV-1 envelope immunogens that bound precursor B cells of either a CD4 binding site or V3-glycan bnAb lineage. In vitro, these immunogens bound more strongly to bnAb precursors once the precursor acquired the desired improbable mutations. Vaccination of macaques with the CD4 binding site–targeting immunogen induced CD4 binding site serum neutralizing antibodies. Antibody sequences elicited in human bnAb precursor knock-in mice encoded functional improbable mutations critical for bnAb development. In bnAb precursor knock-in mice, we isolated a vaccine-elicited monoclonal antibody bearing functional improbable mutations that was capable of neutralizing multiple HIV-1 global isolates. Structures of a bnAb precursor, a bnAb, and the vaccine-elicited antibody revealed the precise roles that acquired improbable mutations played in recognizing the HIV-1 envelope. Thus, our immunogens elicited antibody responses in macaques and knock-in mice that exhibited the mutational patterns, structural characteristics, or neutralization profiles of nascent broadly neutralizing antibodies. CONCLUSION Our study represents a proof of concept for targeted selection of improbable mutations to guide antibody affinity maturation. Moreover, this study demonstrates a rational strategy for sequential immunogen design to circumvent the difficult roadblocks in HIV-1 bnAb induction by vaccination. We show that immunogens should exhibit differences in affinity across antibody maturation stages where improbable mutations are necessary for the desired antibody function. This strategy of selection of specific antibody nucleotides by immunogen design can be applied to B cell lineages targeting other pathogens where guided affinity maturation is needed for a protective antibody response.

We investigated whether there is any association between a native-like conformation and the presence of only the canonical (i.e., native) disulfide bonds in the gp120 subunits of a soluble recombinant human immunodeficiency virus type 1 (HIV-1) envelope (Env) glycoprotein. We used a mass spectrometry (MS)-based method to map the disulfide bonds present in nonnative uncleaved gp140 proteins and native-like SOSIP.664 trimers based on the BG505 env gene. Our results show that uncleaved gp140 proteins were not homogeneous, in that substantial subpopulations (20 to 80%) contained aberrant disulfide bonds. In contrast, the gp120 subunits of the native-like SOSIP.664 trimer almost exclusively retained the canonical disulfide bond pattern. We also observed that the purification method could influence the proportion of an Env protein population that contained aberrant disulfide bonds. We infer that gp140 proteins may always contain a variable but substantial proportion of aberrant disulfide bonds but that the impact of this problem can be minimized via design and/or purification strategies that yield native-like trimers. The same factors may also be relevant to the production and purification of monomeric gp120 proteins that are free of aberrant disulfide bonds. IMPORTANCE It is widely thought that a successful HIV-1 vaccine will include a recombinant form of the Env protein, a trimer located on the virion surface. To increase yield and simplify purification, Env proteins are often made in truncated, soluble forms. A consequence, however, can be the loss of the native conformation concomitant with the virion-associated trimer. Moreover, some soluble recombinant Env proteins contain aberrant disulfide bonds that are not expected to be present in the native trimer. To assess whether these observations are linked, to determine the extent of disulfide bond scrambling, and to understand why scrambling occurs, we determined the disulfide bond profiles of two soluble Env proteins with different designs that are being assessed as vaccine candidates. We found that uncleaved gp140 forms heterogeneous mixtures in which aberrant disulfide bonds abound. In contrast, BG505 SOSIP.664 trimers are more homogeneous, native-like entities that contain predominantly the native disulfide bond profile.

The use of monoclonal antibodies (mAbs) for the manufacture of innovator and biosimilar biotherapeutics has increased tremendously in recent years. From a structural perspective, mAbs have high disulfide bond content, and the correct disulfide connectivity is required for proper folding and to maintain their biological activity. Therefore, disulfide linkage mapping is an important component of mAb characterization for ensuring drug safety and efficacy. The native disulfide linkage patterns of all four subclasses of IgG antibodies have been well established since the late 1960s. Among these IgG subtypes, disulfide mediated isoforms have been identified for IgG2 and IgG4, and to a lesser extent in IgG1, which is the most studied IgG subclass. However, no studies have been carried out so far to investigate whether different IgG3 isoforms exist due to alternative disulfide connectivity. In an effort to investigate the presence of disulfide-mediated isoforms in IgG3, we employed a bottom-up mass spectrometry approach to accurately determine the disulfide bond linkages in endogenous human IgG3 monoclonal antibody, and our results show that no such alternative disulfide bonds exist. While many antibody-based drugs are developed around IgG1, IgG3 represents a new, and in some cases, more desirable drug candidate. Our data represent the first demonstration that alternative disulfide bond arrangements are not present in endogenous IgG3; and therefore, they should not be present in recombinant forms used as antibody-based therapeutics.

Interactions between innate antiviral factors at mucosal surfaces and HIV-1 virions contribute to the natural inefficiency of HIV-1 transmission and are a platform to inform the development of vaccine and nonvaccine strategies to block mucosal HIV-1 transmission. Tenascin-C (TNC) is a large, hexameric extracellular matrix glycoprotein identified in breast milk and genital fluids that broadly neutralizes HIV-1 via interaction with the HIV-1 Envelope (Env) variable 3 (V3) loop. In this report, we characterize the specific determinants of the interaction between TNC and the HIV-1 Env. We observed that TNC binding and neutralization of HIV-1 is dependent on the TNC fibrinogen-like globe (fbg) and fibronectin-type III (fn) domains, oligomerization, and its newly-mapped glycan structure. Moreover, we observed that TNC-mediated neutralization is also dependent on Env V3 residues 321/322 and 326/327, which surround the IGDIR motif of the V3 loop, as well the N332 glycan, which is critical to the broadly neutralizing activity of glycan-dependent V3-specific antibodies such as PGT128. Our results demonstrate a striking parallel between innate and adaptive immune mechanisms of broad HIV neutralization and provide further insight into the host protein-virus interactions responsible for the natural inefficiency of mucosal HIV-1 transmission.

SummaryThe human immunodeficiency virus (HIV-1) envelope glycoprotein (Env) trimer mediates viral entry into cells and is the major target for the host antibody response. In infected cells, the mature Env [(gp120/gp41)3] is produced by cleavage of a trimeric gp160 precursor. Proteolytic cleavage decreases Env conformational flexibility, allowing the mature Env to resist antibody binding to conserved elements. The conformational plasticity of the Env precursor skews the humoral immune response towards the elicitation of ineffectual antibodies, contributing to HIV-1 persistence in the infected host. The structural basis for the plasticity of the Env precursor remains elusive. Here we use cryo-electron microscopy to visualize two coexisting conformational states of the full-length Env precursor at nominal resolutions of 5.5 and 8.0 Å. The State-P2 conformation features a three-helix bundle of the gp41 heptad repeat region in the core, but has disordered membrane-interactive regions. State-P1 trimers lack the three-helix bundle and instead retain ordered transmembrane and membrane-proximal external regions embracing a central cavity. Our structural data shed light on the unusual plasticity of the Env precursor and provide new clues to Env immunogen discovery.

Recombinantly expressed proteins are susceptible to oxidation during expression, purification, storage, and analysis; the residue most susceptible to oxidation is methionine. Methionine oxidation can be overestimated using current quantitative analysis methods because oxidation can occur during sample preparation, and researchers often do not use methods that account for this possibility. An experimental strategy had been developed previously to solve this problem through the use of an 18O-labeled hydrogen peroxide reagent. However, the method did not address the analysis of peptides that contained multiple methionine residues. Herein, we develop and validate a new analysis method that uses theoretical isotope distributions and experimental spectra to quantify methionine oxidation that is present prior to sample preparation. The newly described approach is more rapid than the previously described method, and it needs only half the amount of protein for analysis. This method was validated using model proteins; then, it was applied to the analysis of recombinant HIV-1 Env, the key protein in HIV vaccine candidates. While Met oxidation of this protein could not be analyzed using previous methods, the approach described herein was useful for determining the oxidation state of HIV-Env. Graphical Abstract ᅟ.

Glycosylation on proteins adds complexity and versatility to these biologically vital macromolecules. To unveil the structure-function relationship of glycoproteins, glycopeptide-centric analysis using mass spectrometry (MS) has become a method of choice because the glycan is preserved on the glycosylation site and site-specific glycosylation profiles of proteins can be readily determined. However, glycopeptide analysis is still challenging given that glycopeptides are usually low in abundance and relatively difficult to detect and the resulting data require expertise to analyze. Viewing the urgent need to address these challenges, emerging methods and techniques are being developed with the goal of analyzing glycopeptides in a sensitive, comprehensive, and high-throughput manner. In this review, we discuss recent advances in glycoprotein and glycopeptide analysis, with topics covering sample preparation, analytical separation, MS and tandem MS techniques, as well as data interpretation and automation.

We describe the properties of BG505 SOSIP.664 HIV‐1 envelope glycoprotein trimers produced under current Good Manufacturing Practice (cGMP) conditions. These proteins are the first of a new generation of native‐like trimers that are the basis for many structure‐guided immunogen development programs aimed at devising how to induce broadly neutralizing antibodies (bNAbs) to HIV‐1 by vaccination. The successful translation of this prototype demonstrates the feasibility of producing similar immunogens on an appropriate scale and of an acceptable quality for Phase I experimental medicine clinical trials. BG505 SOSIP.664 trimers are extensively glycosylated, contain numerous disulfide bonds and require proteolytic cleavage, all properties that pose a substantial challenge to cGMP production. Our strategy involved creating a stable CHO cell line that was adapted to serum‐free culture conditions to produce envelope glycoproteins. The trimers were then purified by chromatographic methods using a 2G12 bNAb affinity column and size‐exclusion chromatography. The chosen procedures allowed any adventitious viruses to be cleared from the final product to the required extent of >12 log10. The final cGMP production run yielded 3.52 g (peptidic mass) of fully purified trimers (Drug Substance) from a 200 L bioreactor, a notable yield for such a complex glycoprotein. The purified trimers were fully native‐like as judged by negative‐stain electron microscopy, and were stable over a multi‐month period at room temperature or below and for at least 1 week at 50°C. Their antigenicity, disulfide bond patterns, and glycan composition were consistent with trimers produced on a research laboratory scale. The methods reported here should pave the way for the cGMP production of other native‐like Env glycoprotein trimers of various designs and genotypes.

Developing effective high-throughput screening (HTS) methods is of paramount importance in the early stage of drug discovery. While rugged and robust assays may be easily developed for certain enzymes, HTS assays designed to identify ligands that block protein binding are much more challenging to develop; attenuating the number of false positives and false negatives under high-throughput screening conditions is particularly difficult. We describe an MS-based HTS workflow that addresses these challenges. The assay mitigates false positives by selectively identifying positive hits exclusively when a ligand at the binding site of interest is displaced; it mitigates false negatives by detecting a reporter compound that ionizes well, not by detecting the ligand binder, which may not ionize. The method was validated by detecting known binders of three proteins, pepsin, maltose binding protein (MBP), and carbonic anhydrase (CA) in the presence of hundreds of non-binders. We also identified a novel CA binder, pifithrin-µ, which could not have been identified by any other MS-based assay because of its poor ionization efficiency. This new method addresses many of the challenges that are currently encountered during high-throughput screening.

HIV-1 envelope glycoprotein (Env) glycosylation is important because individual glycans are components of multiple broadly neutralizing antibody epitopes, while shielding other sites that might otherwise be immunogenic. The glycosylation on Env is influenced by a variety of factors, including the genotype of the protein, the cell line used for its expression, and the details of the construct design. Here, we used a mass spectrometry (MS)-based approach to map the complete glycosylation profile at every site in multiple HIV-1 Env trimers, accomplishing two goals. (i) We determined which glycosylation sites contain conserved glycan profiles across many trimeric Envs. (ii) We identified the variables that impact Env's glycosylation profile at sites with divergent glycosylation. Over half of the gp120 glycosylation sites on 11 different trimeric Envs have a conserved glycan profile, indicating that a native consensus glycosylation profile does indeed exist among trimers. We showed that some soluble gp120s and gp140s exhibit highly divergent glycosylation profiles compared to trimeric Env. We also assessed the impact of several variables on Env glycosylation: truncating the full-length Env; producing Env, instead of the more virologically relevant T lymphocytes, in CHO cells; and purifying Env with different chromatographic platforms, including nickel-nitrilotriacetic acid (Ni-NTA), 2G12, and PGT151 affinity. This report provides the first consensus glycosylation profile of Env trimers, which should serve as a useful benchmark for HIV-1 vaccine developers. This report also defines the sites where glycosylation may be impacted when Env trimers are truncated or produced in CHO cells. IMPORTANCE A protective HIV-1 vaccine will likely include a recombinant version of the viral envelope glycoprotein (Env). Env is highly glycosylated, and yet vaccine developers have lacked guidance on how to assess whether their immunogens have optimal glycosylation. The following important questions are still unanswered. (i) What is the “target” glycosylation profile, when the goal is to generate a natively glycosylated protein? (ii) What variables exert the greatest influence on Env glycosylation? We identified numerous sites on Env where the glycosylation profile does not deviate in 11 different Env trimers, and we investigated the impact on the divergent glycosylation profiles of changing the genotype of the Env sequence, the construct design, the purification method, and the producer cell type. The data presented here give vaccine developers a “glycosylation target” for their immunogens, and they show how protein production variables can impact Env glycosylation.

N-linked glycans are required to maintain appropriate biological functions on proteins. Underglycosylation leads to many diseases in plants and animals; therefore, characterizing the extent of glycosylation on proteins is an important step in understanding, diagnosing, and treating diseases. To determine the glycosylation site occupancy, protein N-glycosidase F (PNGase F) is typically used to detach the glycan from the protein, during which the formerly glycosylated asparagine undergoes deamidation to become an aspartic acid. By comparing the abundance of the resulting peptide containing aspartic acid against the one containing non-glycosylated asparagine, the glycosylation site occupancy can be evaluated. However, this approach can give inaccurate results when spontaneous chemical deamidation of the non-glycosylated asparagine occurs. To overcome this limitation, we developed a new method to measure the glycosylation site occupancy that does not rely on converting glycosylated peptides to their deglycosylated forms. Specifically, the overall protein concentration and the non-glycosylated portion of the protein are quantified simultaneously by using heavy isotope-labeled internal standards coupled with LC-MS analysis, and the extent of site occupancy is accurately determined. The efficacy of the method was demonstrated by quantifying the occupancy of a glycosylation site on bovine fetuin. The developed method is the first work that measures the glycosylation site occupancy without using PNGase F, and it can be done in parallel with glycopeptide analysis because the glycan remains intact throughout the workflow.

Studying protein O-glycosylation remains an analytical challenge. Different from N-linked glycans, the O-glycosylation site is not within a known consensus sequence. Additionally, O-glycans are heterogeneous with numerous potential modification sites. Electron transfer dissociation (ETD) is the method of choice in analyzing these glycopeptides since the glycan side chain remains intact in ETD, and the glycosylation site can be localized on the basis of the c and z fragment ions. Nonetheless, new software is necessary for interpreting O-glycopeptide ETD spectra in order to expedite the analysis workflow. To address the urgent need, we studied the fragmentation of O-glycopeptides in ETD and found useful rules that facilitate their identification. By implementing the rules into an algorithm to score potential assignments against ETD-MS/MS data, we applied the method to glycopeptides generated from various O-glycosylated proteins including mucin, erythropoietin, fetuin, and an HIV envelope protein, 1086.C gp120. The site-specific O-glycopeptide composition was correctly assigned in every case, proving the merits of our method in analyzing glycopeptide ETD data. The algorithm described herein can be easily incorporated into other automated glycomics tools.

Electron transfer dissociation (ETD) is commonly used in fragmenting N-linked glycopeptides in their mass spectral analyses to complement collision-induced dissociation (CID) experiments. The glycan remains intact through ETD, while the peptide backbone is cleaved, providing the sequence of amino acids for a glycopeptide. Nonetheless, data analysis is a major bottleneck to high-throughput glycopeptide identification based on ETD data, due to the complexity and diversity of ETD mass spectra compared to CID counterparts. GlycoPep Detector (GPD) is a web-based tool to address this challenge. It filters out noise peaks that interfere with glycopeptide sequencing, correlates input glycopeptide compositions with the ETD spectra, and assigns a score for each candidate. By considering multiple ion series (c-, z-, and y-ions) and scoring them separately, the software gives more weighting to the ion series that matches peaks of high intensity in the spectra. This feature enables the correct glycopeptide to receive a high score while keeping scores of incorrect compositions low. GPD has been utilized to interpret data collected on six model glycoproteins (RNase B, avidin, fetuin, asialofetuin, transferrin, and AGP) as well as a clade C HIV envelope glycoprotein, C.97ZA012 gp140ΔCFI. In every assignment made by GPD, the correct glycopeptide composition earns a score that is about 2-fold higher than other incorrect glycopeptide candidates (decoys). The software can be accessed at http://glycopro.chem.ku.edu/ZZKHome.php .

Glycopeptide-based analysis is used to inform researchers about the glycans on one or more proteins. The method's key attractive feature is its ability to link glycosylation information to exact locations (glycosylation sites) on proteins. Numerous applications for glycopeptide analysis are known, and several examples are described herein. The techniques used to characterize glycopeptides are still emerging, and recently, research focused on facilitating aspects of glycopeptide analysis has advanced significantly in the areas of sample preparation, MS fragmentation, and automation of data analysis. These recent developments, described herein, provide the foundation for the growth of glycopeptide analysis as a blossoming field.