A sensitive assay for measuring whole-blood responses to type I IFNs

Abstract

Human inborn errors of the type I IFN response pathway and auto-Abs neutralizing IFN-α, -β, and/or -ω can underlie severe viral illnesses. We report a simple assay for the detection of both types of condition. We stimulate whole blood from healthy individuals and patients with either inborn errors of type I IFN immunity or auto-Abs against type I IFNs with glycosylated human IFN-α2, -β, or -ω. As controls, we add a monoclonal antibody (mAb) blocking the type I IFN receptors and stimulated blood with IFN-γ (type II IFN). Of the molecules we test, IP-10 (encoded by the interferon-stimulated gene (ISG) CXCL10) is the molecule most strongly induced by type I and type II IFNs in the whole blood of healthy donors in an ELISA-like assay. In patients with inherited IFNAR1, IFNAR2, TYK2, or IRF9 deficiency, IP-10 is induced only by IFN-γ, whereas, in those with auto-Abs neutralizing specific type I IFNs, IP-10 is also induced by the type I IFNs not neutralized by the auto-Abs. The measurement of type I and type II IFN-dependent IP-10 induction therefore constitutes a simple procedure for detecting rare inborn errors of the type I IFN response pathway and more common auto-Abs neutralizing type I IFNs.

Keywords: ELISA; IP-10; auto-antibodies; diagnostic test; type I interferons.

Fig. 1.

IP-10 (CXCL10) induction after the stimulation of whole blood from healthy donors with type I IFNs. (A and B) IP-10 induction, assessed with plasma supernatants after the stimulation of whole blood with various concentrations of glycosylated IFN-α2 for 16 h, as measured by LEGENDplex™. Blood samples were collected 8 to 24 h before stimulation. Three (A) or nine (B) healthy donors were tested once for each set of conditions. (C) CXLC10 mRNA induction after the stimulation of PBMCs from three healthy donors with 1 ng/mL glycosylated IFN-α2 for 6 h, as measured by RT-qPCR. (D) Fold-induction of IP-10 after the stimulation of whole blood with 1 ng/mL glycosylated IFN-α2 for 16 h, as measured by LEGENDplex™. Nine healthy donors were tested once each. Blood samples were collected 8 to 24 h before stimulation. (E and F) IP-10 induction, assessed with plasma supernatants after the stimulation of whole blood with various concentrations of glycosylated IFN-β (E) or glycosylated IFN-ω (F) for 16 h, as measured by LEGENDplex™. Blood samples were collected 8 to 24 h before stimulation. Five healthy donors were tested once for each set of conditions.

Fig. 2.

Whole-transcriptome analysis after PBMC stimulation with IFN-α2. (A) Volcano plot analysis of bulk RNAseq performed on total mRNA extracted from the PBMCs of three healthy donors after stimulation with 1 ng/mL glycosylated IFN-α2 for 6 h. The labeled genes are the top 20 genes displaying the highest levels of induction relative to nonstimulated conditions. (B–D) Volcano plot analysis showing the transcripts induced in cDC2 cells (B), classical monocytes (C), and nonclassical monocytes (D), as determined by scRNAseq (Parse Bioscience). PBMCs of three healthy donors were stimulated with 1 ng/mL glycosylated IFN-α2 for 6 h and whole mRNA was extracted. (E) Single-cell transcriptom analysis. PBMCs from three HD were analyzed after IFN-α2 stimulation vs unstimulated. Clustering analysis. After batch correction with Harmony, celltypes were identified manually. Feature plots show the induction of CXCL10 (encoding IP-10).

Fig. 3.

Evaluation of technical parameters and assessment of the activity of other cytokines (human IFN-ε, IFN-κ, and IFN-γ) on IP-10 induction. (A) IP-10 induction after stimulation with 100 U/mL or 1,000 U/mL recombinant human IFN-ε, IFN-κ, or IFN-γ for 16 h. Three healthy donors were tested for each set of conditions. IP-10 levels were then assessed in plasma supernatants by LEGENDplex™. (B) IP-10 induction after stimulation with 100 U/mL or 1,000 U/mL recombinant human IFN-ε, IFN-κ, or IFN-γ for 16 h. Three healthy donors were tested for each set of conditions. IP-10 levels were then assessed in plasma supernatants by ELISA. (C) Effect of time between collection and stimulation on IP-10 induction. IP-10 levels were assessed in plasma supernatants by LEGENDplex™ after the stimulation of whole blood with 1 ng/mL IFN-α2, -β, -ω, or -γ for 16 h. Three healthy donors were tested for each set of conditions. (D) Effect of the matrix on which blood is collected. Lithium heparin, EDTA, and citrate tubes were used to collect blood from three healthy donors. Whole-blood samples were stimulated with 1 ng/mL IFN-α2, -β, -ω, or -γ for 16 h. IP-10 levels were then assessed in plasma supernatants by LEGENDplex™. (E) Kinetics of IP-10 induction (effect of stimulation time). Whole blood from two healthy donors was stimulated with the indicated concentrations of IFNs for 2, 4, 6, 8, 10, 12, 14, and 16 h. IP-10 levels were then assessed in plasma supernatants by LEGENDplex™. (F) Effect of the presence or absence of 5% CO₂ on IP-10 production after the stimulation of whole blood from three healthy donors with 1 ng/mL IFN-α2, -β, -ω, or -γ for 16 h. −CO₂: stimulation at 37 °C without CO₂ supplementation, +CO₂: stimulation at 37 °C with 5% CO₂ supplementation. (G) IP-10 induction after the stimulation of whole blood from two HDs with type I IFNs and with preincubation with 3 µg/mL anti-IFNAR2 neutralizing mAb, isotype antibody, or a cocktail of antibodies neutralizing all type I IFNs (dilution 1:50). The mAbs were incubated with whole blood for 30 min before stimulation with IFNs.

Fig. 4.

IP-10 induction after the stimulation of whole blood from patients with impaired type I IFN-dependent immunity. (A) IP-10 induction after the stimulation of whole blood from five APS-1 patients and nine healthy donors with 10 ng/mL glycosylated IFN-α2, IFN-β, or IFN-ω for 16 h. IP-10 levels were measured in plasma supernatants by ELISA. Whole blood from three APS-1 patients and five healthy donors was also stimulated with 1,000 U/mL IFN-γ as a control. IP-10 levels were compared between the HD and APS-1 groups by implementing nonparametric Mann–Whitney tests in GraphPad Prism. Ns: not significant; **P-value < 0.001. (B) IP-10 induction after the stimulation of whole blood from five APS-1 patients and nine healthy donors with 1 ng/mL glycosylated IFN-α2, IFN-β, or IFN-ω for 16 h. IP-10 levels were measured in plasma supernatants by ELISA. Whole blood from three APS-1 patients and five healthy donors was also stimulated with 100 U/mL IFN-γ as a control. Multiple Mann–Whitney tests were performed to compare the HD and APS-1 groups for each set of stimulation conditions. IP-10 levels were compared between the HD and APS-1 groups in nonparametric Mann–Whitney tests implemented in GraphPad Prism. Ns: not significant; **P-value < 0.001. (C) IP-10 induction in plasma supernatants after the stimulation of whole blood from a patient neutralizing low concentrations of IFN-α2 (1 ng/mL to 100 pg/mL) in a luciferase assay. Whole-blood samples from the patient and three healthy donors were stimulated with glycosylated type I (10 ng/mL) or type II (1,000 U/mL) IFNs for 16 h. IP-10 levels were then assessed in plasma supernatants by ELISA. (D) IP-10 induction in plasma supernatants after the stimulation of whole blood from a patient neutralizing low concentrations of IFN-α2 (1 ng/mL to 100 pg/mL) in a luciferase assay. Whole-blood samples from the patient and three healthy donors were stimulated with glycosylated type I (1 ng/mL) or type II (100 U/mL) IFNs for 16 h. IP-10 levels were then assessed in plasma supernatants by ELISA. (E) IP-10 induction in plasma supernatants after the stimulation of whole blood from a patient with a NFKB2 p52^LOF/IκBδ^GOF mutation, a female patient with NEMO deficiency (incontinentia pigmenti), and a patient with auto-Abs neutralizing type I IFNs with no genetic diagnosis. TC: Travel control. Whole-blood samples were stimulated with glycosylated type I (1 ng/mL) or type II (100 U/mL) IFNs for 16 h. IP-10 levels were then assessed in plasma supernatants by ELISA. (F) IP-10 induction in plasma samples from patients with autosomal recessive IFNAR1, IFNAR2, TYK2, or IRF9 deficiency. Whole-blood samples from the patients and three healthy donors were stimulated with glycosylated type I (1 ng/mL) or type II (100 U/mL) IFNs for 16 h. IP-10 levels were then assessed in plasma supernatants by ELISA. A TC blood sample from a healthy individual was available for one of the IFNAR2^−/− patients. These samples were stimulated >48 h after blood sampling. The blood of the TYK2^−/− patient was stimulated 24 h after blood sampling. The blood of the IRF9^−/− patient was stimulated 6 h after blood sampling.