89 by PhIP-Seq profiling,94 while the respective bacteria have been cleared from the gut. Second, there may have been a lack of resolution in the microbiome data. For example, some bacterial species commonly detected by metagenomics may have been accompanied by higher detection thresholds in PhIP-Seq, whereas highly immunogenic antigen peptides may not be frequently detected by metagenomics sequencing.21 In addition, the use of fecal microbiota as a proxy for the gut microbiota limits the characterization of local immune-microbiota interactions. Profiling mucosa-attached microbiota rather than fecal microbiome could have improved the antibodybacteria concordance as locally residing (mucosal) microbial communities may elicit stronger immune responses that may also depend on the anatomical location within the intestines.95 We also explored the relationship between peptide prevalence and various morphological, biochemical and lifestyle factors. Our observations reveal a number of interesting associations. For instance, EBV and CMV were associated with lymphocyte and neutrophil counts. These findings are in accordance with observations of absolute lymphocytosis and neutropenia that constitute characteristic laboratory findings in individuals affected by EBV (infectious mononucleosis)96,97 or CMV infections98,99 which may translate into altered immune cell proportions on the longer-term. We also identified a series of associations of allergies and allergens. Allergies are normally triggered by the epitope interaction with IgE antibodies. However, in this study, we mainly used IgG for immunoprecipitation since IgE are found in small amounts in serum and bind with relatively low affinity to the protein A/G coated magnetic beads employed for the immunoprecipitation. Previous studies have shown that allergens have the chance to bind both to IgG and IgE, although they might have different epitope preferences.100 Thus, the allergen associations presented here should be interpreted with caution as they may differ from the classical pathway involved in allergy. Using co-occurrence networks, we identified different peptide groups that normally belonged to the same taxa or orthologous structures in different taxa. However, the existence of modules with apparently unrelated peptides may indicate either interesting biological phenomena or technical factors that we are not accounting for. For instance, H. pylori peptides were observed to occur with the prevalence of antibody-bound peptides for a couple of phages. The disruption of the gut barrier by this pathogen101 could potentially explain the translocation of those phages to blood and the generation of mucosal and systemic immune responses. On the other hand, phenotypic associations also allow us to conjecture about observed cryptic peptide co-occurrence. For instance, CMV peptides were observed to co-occur with several bacterial and plant peptides. Most of those peptides were associated to the same phenotypes, mainly blood cell leukocyte and granulocyte counts, age and sex, meaning that the co-occurrence could be driven by those factors, or that those phenotypes may mediate their co-occurrence. Widespread antibody screenings will be of great importance for the immunology field. Large longitudinal studies will enable us to go from association to causality, for instance uncovering factors that influence the development of autoimmune102 or common allergies.103 Studies like this will enable the development of personalized treatments, e.g. through vaccination strategies.104 Determinants of the human antibody epitope repertoire
RkJQdWJsaXNoZXIy MjY0ODMw