Autoantibodies neutralizing type I IFNs in a fatal case of H5N1 avian influenza

Abstract

Avian influenza A virus (IAV) H5N1 is an emerging threat of human pandemic. We describe a 71-year-old man who died of H5N1 pneumonia in Louisiana and whose blood contained autoantibodies neutralizing type I IFNs (AAN-I-IFNs), including the 12 IFN-α subtypes (1-10 ng/ml) and IFN-ω (100 pg/ml). Causality between these AAN-I-IFN and lethal outcome of avian influenza in this patient is based on (1) our previous report that AA-I-IFN underlie about 5% of cases of critical pneumonia triggered by seasonal influenza viruses in three cohorts, (2) the rarity of this combination of AAN-I-FNs in individuals over 70 years old (<1%), and (3) the rarity of lethal avian influenza among infected individuals (<1%). AAN-I-IFNs underlie a growing number of severe viral diseases, from arboviral encephalitis to viral pneumonia, particularly in the elderly. This case suggests they can also underlie life-threatening avian H5N1 influenza. The presence of AAN-I-IFN may facilitate infection, replication, and adaptation of zoonotic IAVs to humans and, therefore, human-to-human transmission.

Graphical abstract

Figure 1.

AAN-I-IFN in the patient’s blood neutralized IFN-α and IFN-ω. (A) A549-IFN-reporter (AIR) cells carrying the ISRE reporter were stimulated with IFN-α, -ω, or -β at the concentration indicated, with or without blood from the patient (pt), an AAN-I-IFN–positive control (positive ctrl), an APS-1 patient, or healthy donors (HD). All samples were diluted 1:20. Renilla luciferase activity was measured 24 h after stimulation. The results are expressed as a percentage of the mean value for HDs. Luciferase activity levels <25% that of HDs were considered to indicate neutralizing activity. Two separate blood draws from the patient were sampled. Experiments were done three times. (B) ELISA plates were coated with 1 μg/ml of the IFN subtypes indicated and incubated with blood samples (diluted 1:50). Anti-human IgG-HRP secondary antibodies were then added, and OD was measured at 450 nm. An OD₄₅₀>0.5 was considered to be a positive results. Three separate blood draws from the patient were sampled. Experiment was done once. (C) Multiplex assay beads were incubated with blood samples (diluted 1:1,000), and the MFI was normalized against a beads-only control. Normalized MFI values > 3 were considered positive. Two separate blood draws from the patient were sampled. Experiment was done once. (D) AAN-I-IFN neutralization tested with patient’s blood collected at the indicated time points as described in A. Three separate blood draws from the patient were sampled. Experiment was done once.

Figure S1.

Characterization of AAN-I-IFN in the patient. (A) AIR cells were stimulated with the 12 subtypes of IFN-α at a concentration of 1 ng/ml, with or without blood from the patient (pt), an AAN-I-IFN–positive control (positive ctrl), an APS-1 patient, or healthy donors (HD). All samples were diluted 1:20. Renilla luciferase activity was measured 24 h after stimulation and expressed as a percentage of the mean value for HDs. Luciferase activity levels <25% of HD values were considered to indicate neutralizing activity. Three separate blood draws from the patient were sampled. Experiment was done once. (B) Multiplex assay beads were incubated with blood samples, and MFI values were normalized against the beads-only control. Normalized MFI values > 3 were considered positive. Two separate blood draws from the patient were sampled. Experiment was done once. AIR, A549-IFN-reporter.

Figure 2.

The patient’s blood is seropositive for H3N2 and blocked IFN-α function in vivo. (A) We tested for HAI assay by treating patient (Pt) and healthy donor (HD) blood with receptor destruction enzyme (RDE) and diluting 1:10 before mixing with the H1N1, H3N2, and H5N1 IAVs at the indicated titers, together with 0.5% Turkey red blood cells. Blood from naive mice and mice immunized with H1N1 (A/Victoria/4897/2022) or H3N2 (A/Thailand/8/2022) was used as negative and positive controls. Experiment was done once. (B) A549 cells were incubated overnight with 50 pg/ml exogenous IFN-α2 with or without anti–IFN-α2 monoclonal antibody (mAb), patient blood (Pt), healthy donor blood (HD1 and HD2), or APS-1 patient blood (APS-1) at indicated dilution, and then infected with influenza A/California/04/2009 virus expressing NS1-mCherry (CalNSmCherry) at an MOI of 1. The percentage of the cells infected was determined 24 h after infection with a Celigo (Nexcelcom) imaging cytometer. The percentage of infection was normalized against cells infected without IFN-α2 treatment. The dotted line at 26.5% indicates the mean percentage of cells infected after treatment with IFN-α2 only. Experiments were done twice and paired t test was performed (P values: ****<0.0001; ***<0.001; **<0.01; *<0.05; ns > 0.05).