The holoimmunity hypothesis (Root-Bernstein, BioEssays 2016) proposes that T-cell receptors (TcRs) and the microbiome co-evolved to mimic each other, so that immune tolerance extends to a composite 'holobiont self', and that loss of this tolerance produces disease-specific 'holoautoimmunity' in which autoreactive TcRs attack both host tissues and the microbial taxa whose antigens mimic them. Its original support came from proteomic similarity searches on small public TcR-sequence datasets (roughly 68–114 TcRs per group) showing, for example, excess mimicry of coxsackie B virus and Clostridia in type 1 diabetes and of Candida albicans, Enterobacteriaceae, and Pseudomonas in Crohn's disease. The open problem is the validation program the source itself states is required: determining, in large independent cohorts across autoimmune diseases, whether disease-specific TcR mimicry signatures are robust and predictive; whether TcR mimicry can prospectively classify which microbial genera are tolerated versus targeted; and whether the predicted microbial combinations can induce the corresponding disease in animal models.

Background

A holobiont is a host organism together with its associated microbial communities (the microbiome), treated as a combined biological unit. T cells carry T-cell receptors (TcRs), surface proteins that recognize peptide fragments displayed on major histocompatibility complex (MHC) molecules; the TcR repertoire is the full diversity of TcR sequences in an individual, now measurable at scale by immune repertoire sequencing. Molecular mimicry denotes similarity between a microbial antigen and a host antigen, classically invoked to explain how infections can trigger autoimmunity. A specific established result in this area: in a spontaneous mouse model of autoimmune uveitis, activation of retina-specific autoreactive T cells depended on gut commensal microbiota and occurred through the autoreactive TcR responding to non-cognate intestinal antigen, independent of the retinal autoantigen (Horai et al., Immunity, 2015) — evidence that commensal microbes can activate autoreactive TcRs in vivo. In 2016 Robert Root-Bernstein proposed a broader framework built on two ideas: (i) TcRs are selected to avoid strong cross-reactivity with the host's own proteins, so the TcR repertoire statistically 'mimics' genetic self; and (ii) the microbiome has evolved to mimic genetic self so as to share tolerance, so the TcR repertoire also mimics microbial antigenicity. He called the resulting shared tolerance to the holobiont self 'holoimmunity', and predicted that autoimmune attack on host antigens would cross-react with mimicking microbiome components (and conversely), producing 'holoautoimmunity'. His operational tool is nonstandard: proteomic similarity searches in which TcR hypervariable (V-beta) sequences are compared against microbial and human protein databases, with the frequency of 'TcR mimics' of each microbial taxon compared across patient and control repertoires (note that in conventional immunology a TcR binds a peptide-MHC complex, so TcR-sequence-to-antigen similarity is the hypothesis's own construct, not a standard measure of antigen specificity). Applying this to small public datasets, he reported that TcRs from Crohn's disease patients (CD in the source means Crohn's disease, not celiac disease) showed excess mimicry of C. albicans, Corynebacterium, Enterobacteriaceae, mycobacteria and Pseudomonas, and TcRs from type 1 diabetes mellitus (T1DM) patients excess mimicry of coxsackie B virus and Clostridia, proposing that combined infections (e.g., coxsackie B virus together with Clostridia for T1DM) trigger these diseases. The paper's own conclusion was that these six tests 'provide initial data confirming the plausibility' but that 'far more data-intensive tests are clearly required to validate the details', and that no one had yet studied whether anti-TcR antibodies and anti-idiotypic TcRs (receptors or antibodies directed against the variable regions of other TcRs, whose existence in autoimmune patients was known) also recognize host and microbiome antigens. Commentaries by Tauber (2016) and Swiatczak and Tauber (2018) discussed the hypothesis but did not test it.

Problem Statement

Validate or refute the holoimmunity/holoautoimmunity hypothesis through the criteria its source sets out, with the common objective of determining whether TcR mimicry of microbial antigens is a real, disease-specific mechanism of autoimmune triggering rather than a statistical artifact of small datasets. Concretely: (1) determine whether disease-specific TcR mimicry signatures — statistically significant, disease-associable excesses of TcR sequences showing proteomic similarity to particular microbial taxa — are robust and predictive when TcR repertoires are sequenced and analyzed from substantially larger, independent cohorts of healthy individuals and patients across multiple autoimmune diseases (the source names rheumatoid arthritis, lupus, myocarditis, and thyroiditis as additional targets beyond the originally analyzed Crohn's disease and type 1 diabetes); (2) test whether TcR mimicry can prospectively classify which microbial genera are tolerated versus targeted in a given autoimmune disease, rather than only discriminating retrospectively; (3) test experimentally whether the predicted combinations of microorganisms — e.g., coxsackie B virus together with Clostridium species for type 1 diabetes mellitus, or Candida albicans together with Enterobacteriaceae and Pseudomonas species for Crohn's disease — can induce the corresponding autoimmune disease in animal models. These three components jointly support one objective, because each is a different evidential leg (statistical robustness, prospective predictivity, and causal inducibility) of the same framework. A resolution requires an explicit assessment of all three components against the available evidence — stating, for each, whether the evidence supports or undermines the framework — and admits three outcomes: validation if all three are supported, refutation if any one of them is falsified, and a finding that the decisive evidence is not yet available for one or more components, naming which remain untested. All three components must be assessed. Linked secondary predictions of the same framework that also remain untested (optional extensions, not required deliverables): whether B-cell receptor and antibody (particularly IgA, the mucosal antibody class) repertoires parallel the TcR mimicry signatures and mediate mucosal tolerance to the microbiome; whether anti-TcR antibodies and anti-idiotypic TcRs found in autoimmune patients also recognize host and microbiome antigens as part of an idiotypic regulatory network; whether the microbiome functions as a 'secondary immune system' through molecular mimicry of host-protective antigens, and how antibiotics, diet, or microbiome-altering cancer therapies affect this; and whether plant and other non-host-interacting microbes are immunological 'non-antigens', which requires direct challenge studies testing for a complete lack of physiological and immunological interaction.

References

  1. Robert Root-Bernstein (2016), 'Autoimmunity and the microbiome: T-cell receptor mimicry of "self" and microbial antigens mediates self tolerance in holobionts', BioEssays 38(11):1068–1083. DOI: 10.1002/bies.201600083, https://doi.org/10.1002/bies.201600083
  2. Alfred I. Tauber (2016), 'A hypothesis: Establishing the microbiome through immune mimicry (comment on DOI 10.1002/bies.201600083)', BioEssays 38(11):1062. DOI: 10.1002/bies.201600173, https://doi.org/10.1002/bies.201600173
  3. Bartlomiej Swiatczak, Alfred I. Tauber (2018), 'Holoimmunity Revisited', BioEssays 40(11):e1800117. DOI: 10.1002/bies.201800117, https://doi.org/10.1002/bies.201800117
  4. R. Root-Bernstein (2017), 'Human Immunodeficiency Virus Proteins Mimic Human T Cell Receptors Inducing Cross-Reactive Antibodies', International Journal of Molecular Sciences 18(10):2091. DOI: 10.3390/ijms18102091, https://doi.org/10.3390/ijms18102091
  5. Tatsuo Ichinohe, Takahiko Miyama, Takakazu Kawase, Yasuko Honjo, Kazutaka Kitaura, Hiroyuki Sato, Tadasu Shin-I, Ryuji Suzuki (2018), 'Next-Generation Immune Repertoire Sequencing as a Clue to Elucidate the Landscape of Immune Modulation by Host–Gut Microbiome Interactions', Frontiers in Immunology 9:668. DOI: 10.3389/fimmu.2018.00668, https://doi.org/10.3389/fimmu.2018.00668
  6. Robert Root-Bernstein, Elizabeth Churchill, Shelby Oliverio (2023), 'T Cell Receptor Sequences Amplified during Severe COVID-19 and Multisystem Inflammatory Syndrome in Children Mimic SARS-CoV-2, Its Bacterial Co-Infections and Host Autoantigens', International Journal of Molecular Sciences 24(2):1335. DOI: 10.3390/ijms24021335, https://doi.org/10.3390/ijms24021335
  7. Robert Root-Bernstein, Kaylie Chiles, Jack Huber, Alison Ziehl, Miah Turke, Maja Pietrowicz (2023), 'Clostridia and Enteroviruses as Synergistic Triggers of Type 1 Diabetes Mellitus', International Journal of Molecular Sciences 24(9):8336. DOI: 10.3390/ijms24098336, https://doi.org/10.3390/ijms24098336
  8. Robert Root-Bernstein (2024), 'T-Cell Receptor Sequences Identify Combined Coxsackievirus–Streptococci Infections as Triggers for Autoimmune Myocarditis and Coxsackievirus–Clostridia Infections for Type 1 Diabetes', International Journal of Molecular Sciences 25(3):1797. DOI: 10.3390/ijms25031797, https://doi.org/10.3390/ijms25031797
  9. Xin Ma, Jian Zhang, Qianling Jiang, Yong-Xin Li, Guan Yang (2025), 'Human microbiome-derived peptide affects the development of experimental autoimmune encephalomyelitis via molecular mimicry', EBioMedicine 111:105516. DOI: 10.1016/j.ebiom.2024.105516, https://doi.org/10.1016/j.ebiom.2024.105516
  10. Mana Iizuka-Koga, Minako Ito, Noriko Yumoto, Setsuko Mise-Omata, Taeko Hayakawa, Kyoko Komai, Shunsuke Chikuma, Satoru Takahashi, Isao Matsumoto, Takayuki Sumida, Akihiko Yoshimura (2024), 'Reconstruction of Sjögren's syndrome-like sialadenitis by a defined disease specific gut-reactive single TCR and an autoantibody', Clinical Immunology 264:110258. DOI: 10.1016/j.clim.2024.110258, https://doi.org/10.1016/j.clim.2024.110258
  11. Robert Root-Bernstein (2026), 'Type 1 Diabetes-Associated Autoantibodies, Sera and T-Cell Receptor Sequences Recognize Myelin Basic Protein: A Novel Peripheral Diabetic Neuropathy–Multiple Sclerosis Connection', Cells 15(15):1384. DOI: 10.3390/cells15151384, https://doi.org/10.3390/cells15151384
  12. Reiko Horai, Carlos R. Zárate-Bladés, Patricia Dillenburg-Pilla, Jun Chen, Jennifer L. Kielczewski, Phyllis B. Silver, Yingyos Jittayasothorn, Chi-Chao Chan, Hidehiro Yamane, Kenya Honda, Rachel R. Caspi (2015), 'Microbiota-Dependent Activation of an Autoreactive T Cell Receptor Provokes Autoimmunity in an Immunologically Privileged Site', Immunity 43(2):343–353. DOI: 10.1016/j.immuni.2015.07.014, https://doi.org/10.1016/j.immuni.2015.07.014