Dengue Literature Review

Cheng2015 - NS1 P311-330 Anti-PDI Autoantibodies in DHF

6 min read

Cheng2015 - NS1 P311-330 Anti-PDI Autoantibodies in DHF

Full citation: Cheng HJ, Luo YH, Wan SW, Lin CF, Wang ST, Hung NT, Liu CC, Ho TS, Liu HS, Yeh TM, Lin YS. Correlation between serum levels of anti-endothelial cell autoantigen and anti-dengue virus nonstructural protein 1 antibodies in dengue patients. Am J Trop Med Hyg. 2015. doi:10.4269/ajtmh.14-0162.

Raw file: [[raw/cheng2015.pdf]]

Summary

This study from the Yee-Shin Lin laboratory at National Cheng Kung University (NCKU), Tainan, Taiwan — the same group behind Lin2001, Lin2006, Lin2011, and Wan2012 — investigates the clinical correlates of anti-endothelial cell autoantibodies in dengue haemorrhagic fever (DHF). Patient sera from a Vietnamese DHF/DF paediatric cohort (Ho Chi Minh City) were characterised for antibody reactivity against four endothelial autoantigens (PDI, HSP60, vimentin, ATP synthase β-chain) previously identified as cross-targeted by anti-NS1 antibodies. The study additionally maps the responsible NS1 epitope for PDI cross-reactivity to the 20-amino acid peptide P311–330 (aa 311–330).

Anti-PDI, anti-HSP60, and anti-vimentin IgM and IgG levels were all significantly elevated in DHF sera versus normal controls. Anti-PDI and anti-HSP60 IgM (but not anti-vimentin IgM) correlated positively with both anti-endothelial cell IgM and anti-DENV NS1 IgM titers — suggesting these two are the principal NS1-mimicry-driven autoantibodies. The P311–330 peptide-specific antibodies were elevated in DHF vs. both controls and DF patients, and anti-PDI levels (both IgM and IgG) correlated with anti-P311–330 levels — identifying P311–330 as the major NS1 epitope for PDI cross-reactivity. Notably, anti-HSP60 did not correlate with anti-P311–330, indicating HSP60 cross-reactivity is mediated by a different NS1 epitope.

A key serological finding: no autoantibody (anti-PDI, anti-HSP60, anti-vimentin, anti-P311–330) differed significantly between primary and secondary DHF patients. Only anti-DENV NS1 IgG was elevated in secondary vs. primary infection (P = 0.0075), consistent with the expected anamnestic IgG response. The autoimmune component of DHF pathogenesis thus appears infection-order independent.

Study Design

  • Type: Cross-sectional clinical correlation study with in vitro antibody assays
  • Sample size: 15 DHF patients (43 serial sera) + 2 DF patients (5 sera) + 10 normal controls from Tainan (dengue-endemic area)
  • Setting: Patient sera from Children’s Hospital No. 1, Ho Chi Minh City, Vietnam; laboratory work at NCKU Medical College, Tainan, Taiwan; samples collected Days 3–10 post-fever onset
  • Population: Vietnamese paediatric DHF/DF patients (1997 WHO criteria); majority secondary infections (13/15 DHF were secondary; 1/2 DF secondary). Serotypes in DHF: DENV-1 (2), DENV-2 (1), DENV-3 (7), DENV-4 (3), unknown (2). Primary DHF: 2 patients (DENV-4 + unknown serotype).

Key Findings

  • Anti-PDI, anti-HSP60, and anti-vimentin IgM and IgG significantly higher in DHF sera vs. normal controls (all P < 0.05 to P < 0.001)
  • Anti-PDI IgM correlated with anti-endothelial cell IgM (r = 0.649, P < 0.0001) and anti-NS1 IgM (r = 0.364, P = 0.017)
  • Anti-HSP60 IgM correlated with anti-endothelial cell IgM (r = 0.610, P < 0.0001) and anti-NS1 IgM (r = 0.443, P = 0.003)
  • Anti-vimentin IgM did NOT correlate with anti-endothelial cell IgM (r = 0.280, NS) or anti-NS1 IgM (r = 0.101, NS) — despite being elevated in DHF vs. controls; suggests vimentin cross-reactivity is driven by a non-NS1 or non-P311–330 antibody population
  • None of the IgG levels (anti-PDI, anti-HSP60, anti-vimentin) correlated with anti-endothelial cell IgG or anti-NS1 IgG
  • Anti-P311–330 IgM and IgG were significantly higher in DHF vs. controls (IgM P < 0.001, IgG P < 0.01); anti-P311–330 IgM was also significantly higher in DHF vs. DF (P < 0.05) — anti-P311–330 distinguishes DHF from DF whereas anti-PDI/HSP60 do not
  • Anti-PDI IgM and IgG correlated with anti-P311–330 IgM and IgG respectively (r = 0.377 IgM P = 0.013; r = 0.732 IgG P < 0.0001) — confirming P311–330 as the PDI epitope
  • Anti-HSP60 did NOT correlate with anti-P311–330 (both IgM and IgG NS) — HSP60 cross-reactivity uses a distinct, unidentified NS1 epitope
  • Anti-PDI IgG correlation with anti-P311–330 IgG (r = 0.732, P < 0.0001) is the strongest quantitative relationship in the paper, suggesting an epitope-specific antibody class switch from IgM to IgG for the PDI autoantigen
  • No significant differences in any autoantibody between primary and secondary DHF; only anti-NS1 IgG was higher in secondary infection (P = 0.0075)
  • No significant variation in antibody levels by collection day (D3–D4, D5–D6, D7–D8, D9–D10) within the acute phase window

Methods Used

  • ELISA (anti-PDI, anti-HSP60, anti-vimentin, anti-NS1, anti-P311–330, anti-P211–225; 1:25 dilution; HRP-conjugated; TMB/ABTS substrate)
  • Flow cytometry antibody-cell-binding assay (HMEC-1 human microvascular endothelial cell line; 1:50 serum dilution; FITC-conjugated anti-IgM/IgG; FACSCalibur)
  • RT-PCR (one-step SYBR Green I real-time RT-PCR; DENV serotype differentiation in acute-phase samples; Qiagen QuantiTect kit)
  • E/M-specific capture IgM ELISA and NS1 serotype-specific IgG ELISA (primary vs. secondary infection classification)
  • Linear mixed model (SAS PROC MIXED; intra-individual correlation correction for repeated samples)
  • Spearman’s rank correlation (antibody-antibody correlation analysis)

Entities Mentioned

Concepts Addressed

Relevance & Notes

This paper is the fourth empirical contribution from the Yee-Shin Lin/NCKU group to this wiki (after Lin2001, Lin2006, Lin2011/Wan2012), representing a longitudinal refinement of the NS1 molecular mimicry model. Its primary structural contribution is resolving the PDI-specific epitope within NS1 to the 20-amino acid P311–330 region — narrowing the previously established “C-terminal region (aa 311–352)” to a specific sub-domain. This has direct vaccine implications: P311–330-bearing NS1 constructs will generate anti-PDI cross-reactive antibodies; P311–330-depleted constructs may reduce this risk.

The finding that anti-HSP60 and anti-vimentin are also elevated in DHF (though without direct NS1 IgM correlation and P311–330 correlation respectively) suggests heterogeneity in the endothelial autoantibody population: (1) some anti-NS1 antibodies target PDI through P311–330; (2) others target HSP60 through an unidentified NS1 epitope; (3) anti-vimentin elevation may involve a non-NS1, non-P311–330 mechanism.

The primary/secondary independence of anti-endothelial autoantibodies (confirmed here for anti-PDI, anti-HSP60, anti-vimentin, anti-P311–330) is a convergent finding with Lin2001 (anti-platelet IgM in primary infection) and Saito2004 (PAIgM/PAIgG in secondary via IC mechanism). Together these paint a picture: NS1 molecular mimicry autoantibodies are generated in both primary and secondary DHF at comparable levels; the secondary-infection escalation of severity is driven by ADE/immune complex mechanisms, not by a quantitative increase in NS1-driven autoantibodies.

Limitation: Only 2 primary DHF patients — too small to confirm the primary/secondary equivalence claim definitively.

Questions Raised

  • What is the NS1 epitope responsible for HSP60 cross-reactivity? Identification could clarify whether anti-HSP60 independently contributes to endothelial damage or merely tracks with the anti-PDI response at lower magnitude.
  • Anti-vimentin IgM is elevated in DHF vs. controls but does not track with anti-NS1 IgM or anti-EC IgM — what drives anti-vimentin elevation if not NS1 P311–330? Could another DENV protein (e.g., capsid, prM-cross-reactive HSP60 via Dejnirattisai2010) or non-NS1 mechanism be involved?
  • PDI facilitates DENV entry on endothelial cells (Wan SW et al. 2012). Does anti-PDI autoantibody have any antiviral effect by blocking PDI-mediated integrin activation, thereby reducing viral entry — a potential self-limiting mechanism?
  • Does the primary/secondary equivalence of anti-endothelial autoantibodies hold in larger cohorts with adequate primary DHF representation? The n=2 primary DHF group here is insufficient for confident confirmation.

Graph View

Backlinks