The discovery of a substance in serum with the ability to protect against disease dates back to Emil von Behring and Shibasaburo Kitasato (1); just a year later, Paul Ehrlich made the first reference to ‘Antikörper’, or antibodies, in an 1891 report (2). No less important in retrospect was Karl Landsteiner’s discovery 50 years later that antisera contain specificities to multiple antigens (3); this may be viewed as the discovery of an antibody repertoire. The serological repertoire is comprised of a diverse combination of immunoglobulins secreted by B cells in various compartments including peripheral blood, bone marrow, and mucosal sites (4, 5). From initial exposures to exogenous and endogenous (in the case of cancer and autoimmune disease) antigens, the antibody repertoire is established and constantly reshaped through subsequent exposures to a multitude of different antigens during one’s lifetime (6, 7).

Characterization of serum antibodies has traditionally relied on serological assays that determine the total abundance, binding specificity, and neutralizing activity of polyclonal antibodies using various techniques, including enzyme-linked immunosorbent assays (ELISA) as well as neutralization and immunofluorescence assays (8, 9). Though serology remains essential in the present-day study of antibody responses (10) and gives critical insights into the global immune repertoire, it does not inform on the traits of individual constituent antibody molecules. More recently, single B cell sequencing has allowed recombinant expression and characterization of monoclonal antibodies (mAbs), leading to functional delineation of antibody responses at the single-mAb level and discovery of numerous mAbs with potent therapeutic activity (11). However, as some B cells do not secrete antibodies, B-cell-based studies are often unable to accurately determine the relative abundance of each identified mAb, or its relevance to serological protection. As protective antibody molecules that circulate in blood or coat mucosal surfaces are the key correlate of humoral immunity to various diseases (12), proteomic studies of abundant immunoglobulins are critical to in-depth analysis of the antibody landscape.

Over the last two decades, mass spectrometry (MS) has found increasing use in the analysis of complex protein samples (13); more recently, it has been applied to profiling of polyclonal antibody mixtures, giving rise to next-generation serology (14–17). The proteomic deconvolution of antigen-specific serum antibody mixtures, pioneered by the Georgiou group, is known as Ig-Seq (14, 15, 18) ( Figure 1 ). This method has allowed identification, quantification, and longitudinal tracking of antibody lineages at the molecular level. Ig-Seq is a bottom-up proteomic approach involving affinity purification of antibodies against a target antigen, followed by analysis via a liquid-chromatography-tandem-MS (LC-MS/MS) system. Generated peptide spectra are then matched to a sequence database (19), often constructed by high-throughput B cell sequencing [BCR-Seq, reviewed in (20–22)] to identify serum antibodies and enable their subsequent recombinant expression as mAbs for further functional study (15, 23–25). Technical advances and challenges in mass-spectrometry based antibody sequencing, from sample preparation to computational pipelines, have been recently reviewed comprehensively by Greiff and colleagues (26). In this mini review, we present a survey of the various pathologies explored to date using antibody mass spectrometry, highlighting unique insights into the characteristics of the disease-specific immune repertoire and the therapeutic molecules or strategies which may arise from these studies. We emphasize the implications of disease-specific insights in combating infectious disease, autoimmunity, and cancer.

Ig-Seq is a bottom-up proteomic method involving injection of proteolytically-digested peptides into the LC-MS system (85); trypsin is a good choice for Ig-Seq studies due to the prevalence of arginine or lysine residues flanking the antibody CDR-H3 region. However, proteolytic digestion with trypsin or any other proteases may lead to loss of detectable CDR-H3 peptides and reduce the quantity of identifiable CDR-H3 sequences. To address this, other groups have used top-down or middle-down proteomic approaches, which preserve sequence coverage at the expense of resolution; nevertheless, these methods have resulted in useful insight regarding the antibody repertoire [reviewed in (26)]. Bottom-up approaches have also sought to increase coverage by digesting samples with complementary proteases and using computational pipelines to construct intact antibody sequences (86). Separately, Ig-Seq relies on reference antibody databases to deconvolute peptide spectra; the quality, size and source of the database can significantly impact the results and accuracy of antibody identification (26). Several groups are attempting to overcome this dependence using germline gene sequence databases available through online resources, such as IMGT (72, 87, 88). However, database matching cannot discern exact sequences of antibodies, as their CDR-H3’s specifically are formed only after somatic hypermutation in their secreting B cells. Reference-free sequencing of antigen-specific immunoglobulins from biological fluids via mass spectrometry would address this limitation, though this remains difficult due to the complexity and high variability of antibody variable regions. Recent methods have enabled sequencing of full-length purified mAbs at close to 100% accuracy (89), implementing error correction with database homology searches and introducing mutations and post-translational modifications to converge on exact sequences. Some groups have combined similar pipelines with database matching and have been successful in sequencing antibodies from highly restricted populations, as in autoimmune disease (90). More recent work suggests that antibodies will soon be sequenced directly from repertoires with enough precision to allow recombinant expression (82, 91).

Numerous studies have utilized Ig-Seq to identify individual antibodies in serum and track how their abundances change over time in the contexts of various pathologies (34, 39, 42, 54, 55, 57, 66, 76). This is of particular interest in viral diseases, as there is increasing evidence that exposure history plays a decisive role in the antibody response to infection or vaccination (39, 54, 92, 93). Though the effects of immune imprinting have been most extensively studied in influenza (34, 39, 42), recent work in HuNoV has revealed boosting of persistent cross-reactive antibodies in response to vaccination (54). Further, emerging research in SARS-CoV-2 indicates that the antibody repertoire may be shaped by previous infection with endemic coronaviruses, suggesting that exposure history may play a critical role in other viral immune responses (92, 93). Tracking individual antibody abundances over longitudinal samples will allow us to measure the relative contributions to protection made by pre-existing and de novo antibodies. Describing both of these antibody populations on a disease-specific basis will be necessary for understanding the immunopathogenesis of infection and evaluating the effectiveness of vaccines.

The Ig-Seq pipeline has yielded high-affinity, broadly neutralizing antibodies reactive to SARS-CoV-2, influenza, HIV, and HuNoV, demonstrating the broad potential of the method in discovering novel therapeutics (54, 62, 64). Notably, this method has allowed groups to identify immune ‘signatures’ of exposure, such as the presence of broadly-reactive antibodies against select influenza (34) or norovirus strains (54), or non-protective antibodies to a conserved epitope on SARS-CoV-2 (93). These and similar data may inform the design of next-generation vaccines customized for an individual’s exposure history. For example, a novel vaccine may present epitopes known to boost pre-existing, broadly neutralizing antibodies and avoid presenting epitopes known to be associated with previous non-protective immune responses. It is worth noting, however, that this strategy cannot predict immune escape mutations, which may occur even on conserved epitopes.

Ig-Seq holds promise as a means to profile the secreted antibody repertoire at mucosal surfaces, including the respiratory and intestinal tracts (94–98) ( Figure 1 ). Functional characterizations of site-specific antibody repertoires in diseases occurring at mucosal surfaces are highly important for treatment of individuals with respiratory conditions, such as cystic fibrosis and lung cancer. In addition, novel vaccines are being developed to target mucous membranes (99), and quantitative profiling of the mucosal immune repertoire will be essential to understand their effectiveness and function. As the source of the antibody transcript sequence database influences the quality of Ig-Seq data, subsequent work will require generating more comprehensive donor-specific reference databases using B cells from various compartments. Alternatively, substantial advances in reference-free sequencing may soon enable identification of serum antibodies without the need for such databases.

While mass spectrometric studies of the antibody repertoire span a wide breadth of disease types, more depth of study is needed to obtain additional data for developing novel therapeutic strategies in each disease case, and to validate the protective effects of those strategies. Ig-Seq will allow us to interrogate the overall immune response in unprecedented detail and use site-specific characteristics of the antibody repertoire to design novel therapeutics and vaccine strategies.

SI and JL wrote and revised the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by the National Institutes of Health [P20GM113132] and the Cystic Fibrosis Foundation [STANTO19R0].

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.