Dengue Literature Review

Infection-Triggered Autoimmunity

12 min read

Infection-Triggered Autoimmunity

Overview

Infectious agents — viruses, bacteria, fungi, and parasites — can induce autoantibody production and, in some cases, overt autoimmune disease in individuals without prior autoimmune history. This occurs through distinct but non-mutually exclusive mechanisms: molecular mimicry, bystander activation, and epitope spreading (see Johnson2022 - Infectious Diseases Autoantibodies and Autoimmunity). The resulting autoantibodies may be transient (resolving after pathogen clearance) or persistent (potentially progressing to clinical autoimmune disease), depending on the mechanism and host factors.

This concept is directly relevant to interpreting autoimmune markers — including ANA positivity — detected in patients recovering from dengue (see Autoimmunity in Dengue).

Key Points from Literature

A Fourth Interpretive Layer: Polyreactive Antibodies

A mechanistically distinct category often overlooked in the three-mechanism framework is the normal polyreactive IgM present constitutively in all individuals (see Zhou2007 - Polyreactive Antibodies Natural Antibody Function and Polyreactive Antibodies). These are germline-encoded, low-affinity IgM antibodies that bind structurally unrelated self and non-self antigens — a normal feature of the immune repertoire present from birth.

During acute viral infection, the inflammatory environment (polyclonal B cell stimulation, cytokine amplification, antigen release from damaged tissue) may expand or unmask the polyreactive IgM pool beyond its resting baseline. This would produce a transient, IgM-dominated, broadly self-reactive signal that is:

  • Non-specific by LIA (because LIA tests for defined disease autoantibodies, not polyreactive binding)
  • HEp-2 IIFA-positive (because polyreactive IgM binds nuclear antigens indiscriminately, like any structurally diverse self-antigen)
  • Transient (half-life ~8h; rapidly cleared)
  • Not indicative of antigen-driven autoimmune induction

This framework directly interprets the ~66% IIFA-positive, LIA-negative fraction in acute dengue (Chatterjee2024 - ANA Detection in Dengue Kolkata) and the 80 elevated IgM autoantibodies in Vo2020 - Autoantibody Profiling in Dengue. It is complementary to, not replacing, the molecular mimicry/bystander activation/epitope spreading framework — which accounts for the smaller, antigen-specifically induced autoantibody fraction that carries pathological consequences (anti-platelet IgM in Lin2001; anti-endothelial in Lin2006).

Practical implication: When evaluating infection-associated ANA, the critical question is not simply whether autoantibodies are present — the normal polyreactive IgM guarantees some level of broad self-reactivity in any inflammatory state — but whether the autoantibodies are IgM vs. IgG, germline vs. affinity-matured, and whether they confirm on disease-specific assays (LIA > IIFA). Positive IIFA + negative LIA is the expected signature of polyreactive IgM noise, not disease.

Three Core Mechanisms

Molecular Mimicry

A pathogen-derived antigen shares sufficient sequence or structural similarity with a host self-antigen that anti-pathogen immune responses cross-react with host tissue (see Johnson2022 - Infectious Diseases Autoantibodies and Autoimmunity). Both T cells and B cells may be involved:

  • B cell cross-reactivity: Pathogen antigen activates self-reactive B cells when pathogen epitope resembles a self-epitope, bypassing tolerance
  • T cell cross-reactivity: Viral peptide-MHC complexes activate self-reactive T cells that escaped central tolerance
  • Canonical example: Campylobacter jejuni → anti-ganglioside antibodies (anti-GM1, anti-GD1a, anti-GQ1b) → Guillain-Barré syndrome; ~30–40% of GBS follows C. jejuni infection
  • Relevance to dengue: Dengue NS1 protein shares structural homology with endothelial and platelet proteins — the same mimicry mechanism could, in principle, produce cross-reactive antinuclear antibodies

Bystander Activation

Non-specific polyclonal T and B cell activation driven by cytokines (IFN-α/γ, IL-1, IL-6, TNF) released during infection, without antigen-specific stimulation (see Johnson2022 - Infectious Diseases Autoantibodies and Autoimmunity):

  • IFN-α activates plasmacytoid dendritic cells → polyclonal B cell activation → transient autoantibodies, typically without affinity maturation
  • Critical caveat: Trahtemberg et al. (cited in Johnson2022) found no significant difference in ANA prevalence between COVID-19-positive and COVID-19-negative ICU patients — suggesting that bystander polyclonal activation from severe infection alone is insufficient to consistently elevate ANA
  • Autoantibodies from bystander activation tend to be low-affinity and transient

Epitope Spreading

Initial immune response to a pathogen damages host tissue → releases cryptic self-antigens that were not presented during thymic selection → immune response expands to target additional self-epitopes (see Johnson2022 - Infectious Diseases Autoantibodies and Autoimmunity):

  • May explain diversity of autoantibody specificities after acute infection, including detection of multiple ANA specificities simultaneously (as seen in HAV patients in Berlin2007 - Autoantibodies in Nonautoimmune Individuals during Infections)
  • Most consistent with persistent post-infectious autoimmunity: if cryptic nuclear antigens are released, secondary ANA responses may continue beyond the acute phase even after pathogen clearance

ANA Prevalence During Acute Infections

Berlin2007 - Autoantibodies in Nonautoimmune Individuals during Infections provides the key empirical data:

Infection TypeANA PrevalenceSignificance vs. Controls
Viral (HAV, HBV, HCV; n=23)21.7%P<0.013
Bacterial (n=41)20.0%P<0.006
Parasitic (n=17)17.6%NS
Rickettsial (n=7)0%—
Healthy controls (n=80)3.8%—

Method: ELISA (ANA 8 Pro, Euroimmun; 8 nuclear antigens; 1:100 dilution). The 3.8% control rate is lower than population IIF-based estimates (13.8–16.1%), reflecting the narrower antigen coverage of the ELISA panel; the fold-change (21.7% vs. 3.8%, ~5.7×) is more meaningful than the absolute rates.

Transience vs. Persistence

A central question is whether infection-triggered autoantibodies are transient or persistent:

  • Evidence for transience: Codes2002 - Autoantibodies in Acute Viral Hepatitis found ANA in 20.5% during acute viral hepatitis (IIF ≥1:40, n=156 prospective, Salvador Brazil), dropping to 6.4% at 90-day convalescence — a rapid resolution with pathogen clearance; also 14.8% → 3.9% ASMA; no severity or chronification association in any hepatitis type
  • Intermediate time point (6 months): Gawali2021 - ANA Prevalence in Seroconverted Dengue Patients reports 18.33% ANA positivity at 6 months post-dengue in IgG+ patients (HEp-2 IIFA, 1:100; Gwalior, India; n=120). No control group limits interpretation, but the figure is consistent with a partial decline from the acute peak (54.8%, Chatterjee2024) toward the 2-year level (23.1%, Garcia2009 — on less sensitive rat liver substrate). If the trajectory is real, dengue-associated ANA persists well into the sub-acute period rather than resolving within weeks.
  • Evidence for persistence: Garcia2009 - Long-term Clinical Symptoms Post-Dengue found 23.1% ANA positivity at 2 years post-dengue — if genuine, this is not transient; it may reflect a different mechanism (epitope spreading, or genetically mediated impaired clearance)
  • Dengue-triggered SLE case reports (multiple): Two case sources document dengue-associated SLE diagnosis with the ANA-positive autoantibody signature:
    • Velazqueza2017 - SLE vs Dengue Case Series (n=2 pediatric; Guadalajara Mexico; ANA 1:1280 + antinucleosome + anti-dsDNA in both; Case 2 SLE diagnosed 2 months after dengue; chronological ambiguity acknowledged; lowest evidence level)
    • Rajadhyaksha2012 - Dengue Evolving into SLE and Lupus Nephritis (n=1 adult 22F; Mumbai India; primary DENV-1 [IgM+/IgG-]; ANA 1:320 + anti-dsDNA 1:80 + anti-cardiolipin IgM+IgG+ + biopsy-confirmed Class IV LN; 4-week dengue→SLE interval; no pre-dengue autoantibody baseline; causal direction unresolved)
    • Together, these cases add to the Talib 2013 case (cited via Palacios2016 - Autoimmunity in Dengue Literature Review) as independent observations of dengue-associated SLE/lupus nephritis across different settings (Mexico, India, Singapore). The consistent features — high-titer ANA, anti-dsDNA, hypocomplementemia, short post-dengue interval — are suggestive of dengue unmasking pre-existing subclinical SLE in genetically susceptible individuals rather than de novo autoimmunity, given the Shih2023 population-level null finding for SLE incidence after dengue.
  • The FcγRIIa-HH genotype (impaired IC clearance) may predispose certain individuals to persistent rather than transient infection-triggered autoimmunity (see FcγRIIa Receptor)

Host Risk Factors

From Johnson2022 - Infectious Diseases Autoantibodies and Autoimmunity:

  • Female sex: TLR7 on X chromosome (double dose), estrogen-driven immune activation; ~20% of healthy women are ANA-positive
  • Genetics: HLA-DR alleles, complement gene deficiencies (C1q/C2/C4 → impaired IC clearance → SLE risk), BANK1 polymorphisms, TLR2/4/7/9 variants
  • Innate immunity: TLR pathway variants modulate both pathogen clearance and autoimmune activation threshold

Specific Pathogen-Autoimmunity Associations

Key examples from Johnson2022 - Infectious Diseases Autoantibodies and Autoimmunity:

  • EBV: Molecular mimicry and epitope spreading → Sjögren’s syndrome, SLE, multiple sclerosis
  • Influenza: Anti-phospholipid antibodies → type 1 diabetes, GBS, anti-phospholipid syndrome
  • SARS-CoV-2: Bystander polyclonal activation → broad but non-specific autoantibodies; ANA not significantly increased vs. non-COVID ICU patients
  • HBV: CNS demyelination via molecular mimicry
  • C. jejuni: Anti-ganglioside → GBS (canonical molecular mimicry example)

The Lin/Lei group at NCKU provides the most mechanistically detailed dengue-specific example, spanning two reviews (see Lin2011 - Molecular Mimicry Virus Host Dengue Pathogenesis Table 1, which also catalogues Coxsackievirus B/Type I DM, HCV/cryoglobulinemia, Rubella/Type I DM, HCMV/SLE, HSV/stromal keratitis, Parvovirus B19/SLE+RA).

Population-Level Constraint on Post-Dengue Autoimmune Disease

Shih2023 - Autoimmune Disease Risk After Dengue provides the largest epidemiological test of whether dengue-triggered autoimmunity leads to clinical autoimmune disease. Key findings relevant to the transience/persistence debate:

  • Over a mean 4.57-year follow-up, 14 autoimmune disease categories were examined in 63,814 lab-confirmed dengue patients. After Bonferroni correction, only autoimmune encephalomyelitis (ADEM) was significantly elevated (aHR 2.72; P < 0.0001), and only in the first month after infection.
  • Diseases with strong prior case-report support (SLE, GBS, post-infectious arthritis) were statistically non-significant at the population level.
  • This finding constrains the interpretation of dengue-associated autoimmunity: even if molecular mimicry or bystander activation produce transient autoantibodies, these do not commonly result in sustained clinical autoimmune disease. The Shih2023 data supports the “transient autoimmunity” interpretation of the mechanistic literature.
  • ADEM after dengue fits the classical post-infectious autoimmunity model: transient molecular mimicry or autoreactive T-cell activation targeting CNS myelin, confined to the acute/post-acute period (days 3–19 after symptom onset, consistent with the first-month risk window). See Dengue Neurological Complications.

Dengue NS1 — A Detailed Molecular Mimicry Case

Dengue provides one of the best-characterised examples of acute-phase molecular mimicry leading to direct tissue damage (see NS1 Molecular Mimicry in Dengue):

  • Anti-NS1 Abs cross-react with PDI, vimentin, ATP synthase β-chain, and HSP60 on platelets and endothelial cells; responsible domain mapped to C-terminal NS1 aa 311–352 (see Lin2011 - Molecular Mimicry Virus Host Dengue Pathogenesis)
  • Anti-NS1 causes complement-mediated platelet lysis (IgM), inhibits platelet aggregation via PDI, and induces endothelial apoptosis and NF-κB inflammatory activation (see Lin2006 - Autoimmune Pathogenesis in Dengue Virus Infection)
  • A WGNGCG motif on the E protein is homologous to coagulation factors; present in haemorrhagic flaviviruses (JEV, WNV, YFV) but not HCV — a molecular basis for flavivirus-specific haemorrhagic phenotype
  • Critically distinctive: dengue autoimmunity manifests during the acute infection phase, not post-infectious — this contrasts with the C. jejuni/GBS paradigm and most EBV-associated autoimmune diseases, which develop after pathogen clearance; it makes dengue uniquely suited to study the initiation of autoimmunity in real time

Contradictions & Debates

  • Biomarker elevation vs. clinical disease: Garcia2009 found elevated ANA, IC, and CRP at 2 years post-dengue in symptomatic patients. Yet Shih2023 found no elevated incidence of clinical autoimmune diseases over 4.57 years in 63,814 dengue patients. These are not necessarily contradictory — ANA positivity and clinical autoimmune disease are distinct endpoints, and most ANA-positive individuals do not develop clinical disease — but the discrepancy highlights that biological autoimmune activation does not reliably translate into clinical outcome.
  • Bystander activation vs. ANA: The COVID-19 data (Johnson2022) suggests bystander activation alone does not explain ANA elevation, yet severe dengue involves a similar cytokine storm. This leaves the mechanism of dengue-associated ANA (Garcia2009) unclear — molecular mimicry or epitope spreading may be more relevant than bystander activation.
  • ELISA vs. IIF: Berlin2007’s 3.8% control rate (ELISA, 8 antigens) vs. 13.8–16.1% (IIF, all antigens; Satoh2012/Dinse2022) highlights that infection-triggered ANA detected by ELISA may represent a narrow, disease-associated subset, while IIF captures a broader range including non-pathogenic nuclear reactivities.
  • Transience timeline: The evidence for transience (Codes2002 prospective hepatitis cohort, 90-day follow-up; also cited in Berlin2007) and persistence (Garcia2009, 2 years post-dengue) may reflect genuinely different mechanisms or simply different time points; longitudinal data specifically for dengue ANA is absent.

Sources

Graph View

Backlinks