85 mainly Streptococcus pneumoniae, Staphylococcus aureus, Mycoplasma pneumoniae, Haemophilus influenzae and Escherichia coli (particularly antigens from the type III secretion system (T3SS) of serotype O157:H7). Younger individuals also showed more frequent antibody responses against alpha S1 casein proteins. Sex demonstrated 43 significant enrichments (24 for males, 19 for females). Females exhibitedmore frequent antibody-boundpeptides from Lactobacillusacidophilus and Lactobacillus johnsonii, both known inhabitants of the vaginal microbiome.77,78 Antibody-bound peptide responses were particularly directed against Lactobacillus surface proteins, including S-layer proteins (SLPs, e.g. SIpA and SIpX proteins) and the peptidoglycan lysozyme N-acetylmuramidase, reproducing previous findings in.21 Females also demonstrated increased enrichment of EBV and CMV peptides. Males showed higher prevalence of antibody-bound peptides from Haemophilus influenzae bacteria (e.g. serotype Rd KW20 or strain 3179B), also as previously described,20,79 and of several peptides derived from Streptococcus, Staphylococcus, Bacteroides and alpha-herpesviruses (including HSV-1 and varicella zoster virus). Associations between antibody-bound peptides and laboratory cell counts included both cell proportions and absolute cell quantifications, which appeared to be largely driven by antibody-bound peptides from CMV. Lymphocyte counts showed almost exclusively positive associations with CMV, but also some to EBV, whereas the same antibody-bound peptides demonstrated many inverse associations with neutrophil counts. Smoking associations included associations to current smoking status (41) (Figure 4B), ever smoking for at least a year (43) and parental smoking (7). Most associations were related with higher prevalence of peptides belonging to enteroviruses, both rhinovirus and poliovirus. The relationship between smoking and rhinovirus infection has been previously described,80 and thus associations to other viral peptides belonging to enteroviruses could be due to cross- reactivity to homologous proteins. We also observed a consistently higher seroprevalence of EBV in smokers, which might be reactivated by smoking, as shown by an in vitro model.81 In addition, there were increased antibody responses against miscellaneous respiratory pathogens, including several Streptococcus spp. Interestingly, flagellin antibody–bound peptides (Roseburia, Lachnospiraceae, Eubacterium and Clostridiales) show a lower prevalence in smokers, as do Escherichia virulence factors (Figure 3B). We used serological information about the presence of autoantibodies to identify bacterial and allergen peptides linked to the presence of these autoimmune antibodies (Figure 4C). Anti-cyclic citrullinated peptide (anti-CCP) antibody levels, a marker for rheumatoid arthritis, were positively associated with 23 antibody-bound peptides, including peptides derived from Bacteroides, Parabacteroides, Prevotella and Porphyromonas gingivalis bacteria. These findings correspond well with bacterial genera that are known to be altered in the microbiome of patients with anti-CCP-positive rheumatoid arthritis.82 On the other hand, the connective tissue disease (CTD) screen panel, in which total reactivity to a mixture of antigens associated with several autoimmune diseases is measured, was almost exclusively associated with increased antibodybound peptide frequencies of alpha-S1-casein or kappa casein belonging to Bos taurus (cow), Ovis aries (sheep), Bubalus bubalis (buffalo) and Capra hircus (goat). Indeed, several autoimmune diseases such as celiac disease, juvenile idiopathic arthritis and Ehlers-Danlos syndrome have Determinants of the human antibody epitope repertoire
RkJQdWJsaXNoZXIy MjY0ODMw