The ARA theory of compartmental autoimmunity: Access–Recognition–Amplification as a sequential architecture of autoimmune phenotypes

"Autoimmunity" groups together, under a single term, mechanistically heterogeneous processes that existing models of molecular mimicry and loss of tolerance fail to explain through a shared mechanism. It is proposed that autoimmunity is not a mechanistically homogeneous natural kind, but a common clinical phenotype reachable by divergent etiological pathways. It is argued that the primary causal hypothesis is not molecular resemblance between an exogenous antigen and a self-antigen, but access: the arrival of a damaging supramolecular architecture — of microbial origin, iatrogenic origin, or both at once — in a protected anatomical compartment, and the consequent exposure of an autoantigen normally hidden or tolerized. This proposal is organized under the Access–Recognition–Amplification (ARA) framework: mimicry is only one of the possible modalities by which the immune response comes to recognize and fix upon a self antigen, not the only one, and treating it as synonymous with "recognition" obscures mechanisms — such as bystander co-presentation, enzymatic neoepitope generation, or dual BCR–TLR activation by nucleic-acid complexes — that require no structural resemblance at all. An eight-step sequential architecture (the Compartmental Sequence) is developed that mechanistically instantiates the three stages of the ARA framework; its submodels are specified (VECT — Vascular Escape–Corona–Transcytosis — for excipient-mediated access; seven modalities of Access, including apoptotic relocalization of autoantigen; six modalities of Recognition grouped into three families, plus a hapten-carrier submechanism formalized in celiac disease; two mechanisms of Amplification, including a cross-cutting axis of biopersistent-load reactivation); and it is applied to a bank of more than thirty cases — bacterial, iatrogenic, and idiopathic autoimmune — evaluated under explicit causality standards (Oldstone [1]; Witebsky/Rose-Bona [2]) and with associated falsifiable predictions. The bank also includes boundary cases that test the framework at its extremes: phenotypes with demonstrated Access but no established Recognition (ALS), phenotypes with established causal Recognition but no identified pathological Access (idiopathic pemphigus vulgaris, anti-AChR myasthenia gravis, anti-MuSK myasthenia gravis, and AQP4-IgG-positive NMOSD); a further, amplification-first boundary profile in which Amplification is well specified while Access and Recognition remain unresolved (idiopathic dermatomyositis); and a paraneoplastic and pharmacological family of candidate exceptions to Access-necessity, assessed in §9.1 against an explicit threshold for when such a candidate counts as genuine (checkpoint-inhibitor-associated ANCA vasculitis; cancer-associated anti-TIF1-γ dermatomyositis). It is further proposed that this architecture is of general scope — not restricted to the cases addressed here — but that universality is explicitly put forward as a prediction of the theory, open to refutation, not as an already-demonstrated premise; one boundary case developed in this document, thymoma-associated myasthenia gravis, is the strongest positive candidate identified to date for such a refutation — a demonstrated primary failure of central tolerance, not yet shown to form a sufficient causal chain to pathogenic Recognition — and §9.1 specifies in advance the evidence that would elevate it to a confirmed exception. The framework distinguishes nominal exposure from effective compartmental Access, formalizes Access imprinting and phase-dependent partial or full autonomy, and treats host state as a modifier of each transition and of reversibility. It also makes explicit a functional-configuration layer between Recognition and Amplification: a persistent or semi-persistent compartmental state that can remain subclinical yet condition the response to later inputs. This layer is a state, not a fourth ARA operation, and therefore does not alter the acronym. ARA is situated as the autoimmune sister grammar of ADA, and as a methodological comparator for the independently developed ACA neurodevelopmental framework, within the provisional broader program of Spatially Compartmentalized Disorders, while retaining its own mechanisms, scope and evidentiary standards.