Extensive T-Cell Epitope Repertoire Sharing among Human Proteome, Gastrointestinal Microbiome, and Pathogenic Bacteria: Implications for the Definition of Self

Abstract

T-cell receptor binding to MHC-bound peptides plays a key role in discrimination between self and non-self. Only a subset, typically a pentamer, of amino acids in a MHC-bound peptide form the motif exposed to the T-cell receptor. We categorize and compare the T-cell exposed amino acid motif repertoire of the total proteomes of two groups of bacteria, comprising pathogens and gastrointestinal microbiome organisms, with the human proteome and immunoglobulins. Given the maximum 205, or 3.2 million of such motifs that bind T-cell receptors, there is considerable overlap in motif usage. We show that the human proteome, exclusive of immunoglobulins, only comprises three quarters of the possible motifs, of which 65.3% are also present in both composite bacterial proteomes. Very few motifs are unique to the human proteome. Immunoglobulin variable regions carry a broad diversity of T-cell exposed motifs (TCEMs) that provides a stratified random sample of the motifs found in pathogens, microbiome, and the human proteome. Individual bacterial genera and species vary in the content of immunoglobulin and human proteome matched motifs that they carry. Mycobacteria and Burkholderia spp carry a particularly high content of such matched motifs. Some bacteria retain a unique motif signature and motif sharing pattern with the human proteome. The implication is that distinguishing self from non-self does not depend on individual TCEMs, but on a complex and dynamic overlay of signals wherein the same TCEM may play different roles in different organisms, and the frequency with which a particular TCEM appears influences its effect. The patterns observed provide clues to bacterial immune evasion and to strategies for intervention, including vaccine design. The breadth and distinct frequency patterns of the immunoglobulin-derived peptides suggest a role of immunoglobulins in maintaining a broadly responsive T-cell repertoire.

BrucellaPeptide Cross-Reactive MHC I Presentation Activates SIINFEKL-Specific TCR Expressing T Cells

“…Our results underscore the complexity and ubiquity of molecular mimicry in T cell recognition. The potential for extensive sharing of nonamers between pathogens, gastrointestinal microbiome and human proteome has been demonstrated, both for MHC I and MHC II (51). This may enable microbes to escape immune surveillance by presenting peptides similar to the host and may also lead to microbial exposure cross-activating autoreactive T cells.…”

“…The genome of Brucella melitensis is predicted to encode for 3197 ORFs distributed over two circular chromosomes (62). However, even with this level of complexity, microbe/human commonality is extremely high with 99.7% of human proteins containing bacterial pentapeptides (51, 63). Furthermore, while this study addresses continuous pentamers which are recognized by CD8+ T cells, such peptides are overlaid by the discontinuous pentamers presented by MHC II and recognized by CD4 T cells, with a similar degree of potential cross reactivity (51).…”

“…However, even with this level of complexity, microbe/human commonality is extremely high with 99.7% of human proteins containing bacterial pentapeptides (51, 63). Furthermore, while this study addresses continuous pentamers which are recognized by CD8+ T cells, such peptides are overlaid by the discontinuous pentamers presented by MHC II and recognized by CD4 T cells, with a similar degree of potential cross reactivity (51). Although chicken ovalbumin is not a host protein of mouse or human, we have demonstrated here that there is enough commonality for cross reactivity of several putative Brucella peptides with SIINFEKL.…”

Linking microbiota and respiratory disease

“…Furthermore, lung epithelial and resident myeloid cells in the lung may sense microbiota‐derived products and promote cell autonomous antimicrobials defense mechanisms such as mucus production , AMP secretion or MX protein expression, as known from the gut . The human microbiome shares T cell epitopes not only with pathogens but also with host proteins providing a good reason to believe that commensal bacteria can also shape T cell responses . Commensals may thereby also influence vaccine efficacy by either priming or tolerizing the immune system against vaccine proteins.…”

Holoimmunity Revisited

“…[1] A recent report offers support for a large shared antigen pool: Calculations of the putative number of combinations of amino acids that TCRs can bind is 3.2 million, and 75.4% of these possible combinations are present in the human proteome and as many as 91.4% of them can be found on commensal bacteria as well. [2] This assessment has possible significance inasmuch as microbial peptides similar to human peptides are generally less immunogenic than other microbial molecules [3] and thus peripheral immune-modulation is likely. Indeed, the functional relevance of a large shared antigen pool is suggested by a study of the role of T-cell-exposed integrase motifs expressed by Bacteroides commensals in non-obese diabetic mice.…”