Journal: eBioMedicine

Article Title: Sustained type I interferon activity and N-antigenemia form a vicious cycle during severe COVID-19

doi: 10.1016/j.ebiom.2026.106508

Figure Lengend Snippet: Quantification of N-antigenemia and association with multiple immunological and virological parameters. (A) Comparison of plasma N-antigenemia concentrations among patients with respiratory infections caused by SARS-CoV-2 presenting mild symptoms (n = 100, orange) or severe disease (n = 66, blue, discovery cohort), patients with severe influenza infection (n = 21, white), and non-infected HCs (n = 30, grey). Horizontal continuous black lines represent the median value. The dotted black line represents the positive cut-off and the dotted grey line the LLOQ. Values below the LLOQ but above the LLOD were retained without modification; values below the LLOD were imputed using the formula LLOQ/√2. (B) Graphical representation of the multiparametric analysis comparing N-antigenemia levels in patients with severe COVID-19 with 45 parameters. (C) Correlations in patients with severe COVID-19 between N-antigenemia level and viral load, endothelium and pulmonary damage markers, IgG anti-RBD, mRNA expression level of 26-immune pathways related genes (FilmArray®IPP panel, BioFire Diagnostics, Salt Lake City, UT, USA), circulating pro-inflammatory and immune-related protein concentrations. Significant positive correlations are in purple and negative correlations in orange; non-significant correlations i.e. below the cut-off threshold of 0.05 (discontinuous red line) are in black. Sample size varies for each parameter due to the exclusion of missing values (from n = 37 to n = 65); the exact sample size for each correlation is given in . (D) Correlations of N-antigenemia levels (x-axis) with circulating IFN-α 2 concentration (n = 65); OAS2 (n = 60) and CXCL10 (n = 60) expression; and IgG anti-RBD concentration (n = 53; y-axis). Each dot represents a serum sample from a patient with severe COVID-19, the solid black line represents the linear regression, and the grey area indicates the 95% CI. Comparisons are performed using non-parametric Kruskal–Wallis test with Dunn's post-hoc correction for multiple comparison and correlation using Spearman's rank correlation coefficient (ρ) with Benjamini-Hochberg post-hoc correction for multiple-comparison. ∗∗∗∗ p < 0 . 0001 . All analyses were performed using independent biological replicates. Abbreviations: BAU, binding antibody unit; CI, confidence interval; HCs, Healthy controls; ICU, intensive care unit; IFN, interferon; IgG, immunoglobulin G; LLOD, lower limit of detection; LLOQ, lower limit of quantification; RBD, receptor binding domain.

Article Snippet: SARS-CoV-2 N-antigenemia was quantified from serum samples using the Lumipulse G SARS-CoV-2 Ag assay (Fujirebio Holdings Inc., Tokyo, Japan, 260,340) performed on the Lumipulse G600II automated immunoassay analyser (Fujirebio; RRID: SCR_018018 , ).

Techniques: Comparison, Clinical Proteomics, Infection, Modification, Analysis, Expressing, Concentration Assay, Blood/Plasma/Serum Samples, Binding Assay

Journal: eBioMedicine

Article Title: Sustained type I interferon activity and N-antigenemia form a vicious cycle during severe COVID-19

doi: 10.1016/j.ebiom.2026.106508

Figure Lengend Snippet: SARS-CoV-2 N-antigenemia and IFN-α 2 secretion dynamics over the course of the disease in patients with severe COVID-19. Temporal kinetics of SARS-CoV-2 N- antigenemia (385 longitudinal samples from 134 patients) (A) and circulating IFN-α 2 concentration (395 longitudinal samples from 134 patients) (B); expressed as log 2 pg/mL and fg/mL, respectively, and the sigmoidal relationship between both parameters among positive N-antigenemia and IFN-α 2 concentration levels (239 longitudinal samples from 124 patients) (C), according to the DPSO (days) in patients with severe COVID-19 included in the confirmation cohort. Each circle denotes an individual serum measurement; paired values from the same patient are connected by thin lines. (D) Pseudo-kinetic profiles derived from a k-means algorithm applied to patients with longitudinal samples available between DPSO 0 to 40 only (366 longitudinal samples from 116 patients) are depicted with three independently fitted models focussing on DPSO 0 to 30 to mitigate low counts per cluster at later time points. The continuous curve depicts the nonlinear dynamics of the parameters over the course of the disease, or their sigmoidal relationship, while the shaded area represents the 95% CI, reflecting both fixed-effect estimation uncertainty and variability due to patient-level random effects. All these analyses were performed using generalised additive models including a random intercept to account for inter-individual variability. All analyses were performed using independent biological replicates. Abbreviations: CI, confidence interval; DPSO, days post-symptom onset; IFN, interferon.

Article Snippet: SARS-CoV-2 N-antigenemia was quantified from serum samples using the Lumipulse G SARS-CoV-2 Ag assay (Fujirebio Holdings Inc., Tokyo, Japan, 260,340) performed on the Lumipulse G600II automated immunoassay analyser (Fujirebio; RRID: SCR_018018 , ).

Techniques: Concentration Assay, Derivative Assay

Journal: eBioMedicine

Article Title: Sustained type I interferon activity and N-antigenemia form a vicious cycle during severe COVID-19

doi: 10.1016/j.ebiom.2026.106508

Figure Lengend Snippet: Distinct trajectories of SARS-CoV-2 N-antigenemia and association with impaired or sustained IFN-I response. (A) Longitudinal plasma SARS-CoV-2 N-antigenemia according to DPSO, using 60 serial blood samples of the 22 patients with severe COVID-19 from the validation cohort. Clusters obtained in the confirmation cohort dataset were transferred to the validation cohort using a random forest model, enabling the identification of patients with sustained high N-antigenemia (red; 4/22) and low or decreasing N-antigenemia over the course of the disease (grey; 18/22). Group trajectories are shown with linear mixed-effects regression lines, and 95% CI. The dotted line indicates the positivity threshold. (B) Heatmap of virological and IFN-related parameters over time in the two N-antigenemia clusters. One individual, with two unique samples at day 18 and 21 was excluded (60 longitudinal samples from 21 patients). Rows represent the median of plasma N-antigenemia, plasma IFN-α2 concentration, nasal viral load, blood IFN-I score, and nasal IFN-I score. Columns correspond to DPSO intervals (days 1–3, 4–6, 7–9, and ≥10). Values are scaled per row using min–max normalisation and displayed with increasing intensity of colour from low to high. Line plots on the right illustrate temporal trends (median value) for each parameter in both clusters. All analyses were performed using independent biological replicates. Abbreviations: AANs, neutralising autoantibodies; DPSO, days post-symptom onset; IFN, interferon; ISG, interferon-stimulated genes.

Article Snippet: SARS-CoV-2 N-antigenemia was quantified from serum samples using the Lumipulse G SARS-CoV-2 Ag assay (Fujirebio Holdings Inc., Tokyo, Japan, 260,340) performed on the Lumipulse G600II automated immunoassay analyser (Fujirebio; RRID: SCR_018018 , ).

Techniques: Clinical Proteomics, Blood/Plasma/Serum Samples, Biomarker Discovery, Concentration Assay, Autoantibodies

Journal: eBioMedicine

Article Title: Sustained type I interferon activity and N-antigenemia form a vicious cycle during severe COVID-19

doi: 10.1016/j.ebiom.2026.106508

Figure Lengend Snippet: Proportion and impact of neutralising anti–IFN-α2 autoantibodies (AAN--IFN-α 2 ) on N-antigenemia and IFN-α 2 levels. (A) Comparison of IFN-α 2 levels according to the presence (n = 10) or absence (n = 210) of AAN-IFN-α 2 , assessed in 220 samples collected at admission of patients with severe COVID-19 across the discovery, validation, and confirmation cohorts, plus five additional patients with severe COVID-19. Comparisons were performed using the non-parametric Wilcoxon test. ∗∗ p < 0.01 and ∗∗∗∗ p < 0.001 . (B) Proportion of patients with severe COVID-19 with detectable N-antigenemia according to the presence of AAN-IFN-α 2 within the same patients. Horizontal continuous black lines represent the median value. Stacked bars show the proportions of N-antigenemia–positive and –negative patients. Groups are compared using Fisher's exact test ( p . value shown). (C) Medical history and clinical course of patient #2, adapted from Bussy et al. a patient suffering from severe COVID-19 with a severe hypoxaemic state and positive for AAN-IFN-β. SARS-CoV-2 viral load and the PaO2/FiO2 ratio were obtained from the NPS sample, whereas N-antigenemia and IFN-α 2 concentrations were measured in blood samples. Discontinuous line indicates positive cut-off or LLOQ. IFN-α2 are expressed as fg/mL and N-antigenemia as pg/mL. All analyses were performed using independent biological replicates. Abbreviations: AAN, neutralising autoantibodies; ATB, antibiotic; FiO2, fraction of inspiratory oxygen; ICU, intensive care unit; IFN, interferon; LLOD, lower limit of detection; LLOQ, lower limit of quantification; NS, not significant; NPS, nasopharyngeal swab; PaO2, partial pressure of oxygen.

Article Snippet: SARS-CoV-2 N-antigenemia was quantified from serum samples using the Lumipulse G SARS-CoV-2 Ag assay (Fujirebio Holdings Inc., Tokyo, Japan, 260,340) performed on the Lumipulse G600II automated immunoassay analyser (Fujirebio; RRID: SCR_018018 , ).

Techniques: Autoantibodies, Comparison, Biomarker Discovery, Blood/Plasma/Serum Samples