Ultrasensitive Test Systems Are Required to Quantify Interferon Alpha Protein in Serum Samples

Abstract

Selecting a sensitive test system is crucial to measure IFNα in serum samples. Simoa digital ELISA stands out for its high sensitivity and strong correlation with transcript‐based IFN‐induced gene (ISG) scores. In SLE, the ISG score is explained by circulating IFNα levels with negligible contribution from IFNβ or IFNγ. Created in BioRender.

Abbreviations

IFN

interferon

ISG

IFN‐stimulated genes

Simoa

single‐molecule array

SLE

systemic lupus erythematosus

Type I interferons (IFN) in humans are IFNα, IFNβ, IFNκ, IFNε, and IFNω, from which IFNα can be further subdivided into 13 homologous subtypes (IFNα1, IFNα2, IFNα4, IFNα5, IFNα6, IFNα7, IFNα8, IFNα10, IFNα13, IFNα14, IFNα16, IFNα17, and IFNα21), all binding to the same receptor (IFNAR1/2). IFNs upregulate an overlapping set of genes that are vital to antiviral immune responses [1]. Deviations in these pathways can result in inadequate responses against infections or promote autoimmune processes [1].

Although type I interferons were discovered more than half a century ago [2], measuring type I IFN protein levels from serum samples remains challenging. The low sensitivity and unreliability of the commercially available assays may be caused by the low affinity and specificity of the monoclonal antibody pair used and complicated by the relatively low concentrations of IFNα in human serum. Nevertheless, the low, hardly measurable type I IFN concentrations seem to be bioactive [3].

Type I IFNs participate in the pathogenesis of several autoimmune diseases, such as systemic lupus erythematosus (SLE), Sjogren's syndrome, dermatomyositis, and rheumatoid arthritis, but also in type I interferonopathies [4, 5, 6]. This has been deduced by the trace that IFNs leave on the gene expression profile of circulating immune cells—the IFN signature recognized by the upregulation of IFN‐stimulated genes (ISGs). Although a viable alternative for direct protein quantification, it still does not allow dissection of the relative role of different IFN types, as IFNα, IFNβ, and even IFNγ induce overlapping sets of genes.

Our study aimed to compare a variety of commercially available kits to measure IFNα protein levels in human serum samples, and to analyze their correlation with ISG expression levels in a cohort of SLE patients and healthy controls.

We selected different platforms, including multiplexed bead‐based assay, proximity‐based amplification technology, and digital, sandwich, bio‐, or electroluminescent ELISA methods (Table 1). Regarding the well‐established type I IFN signature in up to 80% of SLE patients, we analyzed 63 serum samples from this patient group (demographics in Table S1). To determine the correlation between IFNα serum concentration and ISG score on mRNA level, we used quantitative RT‐PCR (qPCR) to assess the expression of nine ISGs from the matched RNA samples isolated from whole blood. The relative contribution of each gene was normalized using a z‐score transformation of the relative gene expression values, and the overall score was represented as the sum of nine z‐scores. We compared the IFNα levels with the ISG score (Figure 1). Three of the kits demonstrated a good correlation between the measured IFNα concentration and the ISG score, with the highest correlation for Simoa (Rs = 0.91, p < 2.2e‐16), followed by S‐Plex assay (Rs = 0.89, p < 2.2e‐16) and VeriKine‐HS (Rs = 0.72, p = 4.2e‐11). Among the three kits with the highest correlation coefficients, Simoa stood out for its sensitivity, with only one sample with an undetermined value (IFNα detectable in 98% of SLE patients). The number of samples under the detection limit (LOD) was 40 for VeriKine‐HS and 5 for the S‐Plex assay.

TABLE 1.

Characteristics of the test systems.

	Simoa	VeriKine‐HS	ProQuantum	LumiKine	LEGENDplex	S‐Plex	Olink
Assay name	Simoa IFN‐α Multi‐Subtype (MS) Advantage PLUS assay	VeriKine‐HS Human IFN‐α All Subtype ELISA Kit	IFN alpha Human ProQuantum Immunoassay Kit	LumiKine Xpress hIFN‐α 2.0	Human Type 1/2/3 Interferon Panel (5‐plex) with Filter Plate	S‐PLEX Human IFN‐α2a and IFN‐β Kits	Olink Target 96 Inflammation
Company	Quanterix (USA)	PBL Assay Science (USA)	Thermo Fisher Scientific (USA)	InvivoGen (USA)	BioLegend (USA)	Meso Scale Discovery (USA)	Olink Proteomics (Sweden)
Catalog number	Beta version of 103836	41115	A42897	Luex‐hifnav2	740350	K151P3S (IFN‐α2a), K151ADRS (IFN‐β)	91301
Method	Digital ELISA	Sandwich ELISA	Proximity‐based Immunoassay	Bioluminescent ELISA	Bead‐based Multiplex Assay	Electrochemiluminescent ELISA	Proximity Extension Assay
Specificity	Pan‐Alpha	Pan‐Alpha	IFNA1, IFNA2	IFN‐α2, IFN‐α1, IFN‐α6, IFN‐α8, IFN‐α17	IFN‐α2 a	IFN‐α2a, IFN‐β	IFN‐γ
Instrument	HD‐X Automated Immunoassay Analyzer (Quanterix)	Labsystems Multiskan MCC/34	ViiA 7 Real‐Time PCR System (ThermoFisher)	Victor X Multilabel Plate Reader (Perkin Elmer)	ID7000 Spectral Cell Analyzer (Sony Biotechnology)	MSD Instrument	Olink Signature Q100
Data analysis software	GraphPad Prism	GraphPad Prism	ProQuantum Software	GraphPad Prism	LegendPlex Software	Data Analysis Toolbox (MSD)	Performed by OLINK
Sample dilution factor	5 and 20 (for high concentrations)	2	10	2	2	No dilution	No dilution
Range	0.13–160,000 fg/mL	1.95–125 pg/mL	0.064–5,000 pg/mL	5–1,000 pg/mL	1.19–10,000 pg/mL	5.87–67,000 fg/mL (IFN‐α2a), 17–81,000 fg/mL (IFN‐β)	2.51 NPX b
Number of SLE samples under LLOD, n (%)	1 (1.6%)	40 (63.5%)	51 (81%)	60 (95.2%)	10 (16%)	5 (7.9%, IFN‐α2a), 7 (11%, IFN‐β)	0 (0%)

^a IFN‐β, IFN‐λ1, IFN‐λ2/3, and IFN‐γ were measured too, but not included in the results and discussion, as the sensitivity of this assay was low. The percentage of SLE samples under the LLOD was 24% for IFN‐β, 29% for IFN‐γ, 37% for IFN‐λ1, and 27% for IFN‐λ2/3.

^b Only LLOD was specified, NPX—normalized protein expression.

FIGURE 1.

IFNα protein levels in SLE patient serum samples measured with different assays and their correlation to the type I interferon‐stimulated gene (ISG) score. Correlation between IFNα levels (pg/mL) measured by different assays and ISG score calculated based on the relative mRNA expression levels of type I interferon‐induced genes. The dotted line represents the limit of detection (LOD). Values below LOD were expressed as 90% of the LOD. Spearman correlation was applied.

All the analyzed ISGs showed high and significant positive correlations with serum IFNα concentrations in SLE patients (Figure S1), with the highest correlation for IFI44L (Rs = 0.94, p < 2.2e‐16) and the lowest for CXCL10 (Rs = 0.55, p = 5.2e‐6). Among the control samples, IFNα values remained under LOD in five samples (IFNα detectable in 86% of control serum samples). We found moderate to low correlations between ISG expression and IFNα levels in healthy controls (Figure S2), with the highest correlation for IFI44L (Rs = 0.48, p = 0.0025) and the lowest for MX2 (Rs = 0.17, p = 0.31).

It is well established that type I and II IFN induce similar transcriptional changes in immune cells, and therefore, their specific roles can be challenging to dissect based on gene expression patterns [7]. To evaluate which IFNs potentially have the most significant impact on ISG score, we measured IFNβ and IFNγ serum levels (Table 1) in the same cohort of SLE patients (Figure S3) and healthy controls (Figure S4). Notably, the correlations between these IFNs and ISG scores were only moderate in SLE (IFNγ Rs = 0.43, p = 0.00051 and IFNβ Rs = 0.36, p = 0.0038), suggesting that ISGs are primarily induced by IFNα. In healthy controls, IFNα and IFNγ showed equal correlations with ISG scores (IFNα Rs = 0.34, p = 0.035 and IFNγ Rs = 0.45, p = 0.0062), whereas IFNβ levels had no association (Rs = 0.071, p = 0.82) (Figure S4). These results indicate that in SLE, the IFN signature is primarily caused by IFNα but not by IFNβ or IFNγ. The tonic IFN signature in healthy controls can be influenced by both circulating IFNα and IFNγ.

To evaluate the proportion of SLE patients with increased type I IFN levels, we calculated the upper normal range according to the values from healthy controls for the ISG score and for IFNα serum levels obtained using the Simoa assay. A total of 46 out of 63 patients (73%) had positive ISG scores (ISG score > 15.4), and 44 out of 63 patients (70%) had IFNα values over the cut‐off value (IFNα levels > 0.054 pg/mL). This validates the common practice of using IFN signatures to indicate the elevated IFNα values in SLE [4]. Even though ISG score correlated strongly with IFNα serum levels, the correlations of these markers with SLEDAI score, measuring the SLE disease activity, were moderate (IFNα Rs = 0.46, p = 2e‐4 and ISG score Rs = 0.3, p = 0.019), depicting the complexity of this disease (Figure S5). Although IFNα serum levels and ISG score are strongly correlated in SLE, the score can be influenced by IFNγ in other autoimmune disorders, for example, in Sjögren's syndrome [7] or dermatomyositis [8]. Quantifying IFNα protein levels would be a more rapid way to help diagnose or stratify patients, monitor and predict disease progress, and provide more insight into treatment options. As neutralizing autoantibodies against IFNα have been described in SLE [9], simultaneous screening of IFNα serum and autoantibody levels would give valuable information, as the presence of these autoantibodies can reduce the biological activity of IFNα and cause a sudden drop in IFNα protein to undetectable levels [9, 10, 11]. Importantly, circulating autoantibodies to IFNα can interfere with the performance of some of the IFNα assays [10]. Screening of IFNα serum and autoantibody levels in patients with severe forms of viral infections [12, 13] could help to choose the best treatment strategy and dissect the individual contribution of IFNs in the course of these viral infections.

To conclude, while ISG expression scores have served as valuable surrogate biomarkers for IFN pathway activity, direct protein quantification offers a more immediate and comprehensive assessment, refining diagnostic and prognostic strategies. Furthermore, considering the existence of multiple IFN subtypes with distinct biological activities and varying cellular sensitivities, ultra‐sensitive methods are vital for unraveling their specific contributions within the broader IFN family.

Conflicts of Interest

Kai Kisand and Pärt Peterson are among the authors of the patent “Human anti‐ifn‐alpha antibodies” WO2015001013A2. The remaining authors declare no conflicts of interest.

Supporting information

Supporting File 1:eji70100 sup 0001 SuppMat.docx.

Haljasmägi L., Meisalu S., Bondet V., Peterson P., Duffy D., and Kisand K., “Ultrasensitive Test Systems Are Required to Quantify Interferon Alpha Protein in Serum Samples.” European Journal of Immunology 55, no. 12 (2025): e70100. 10.1002/eji.70100

Data Availability Statement

All the data supporting the findings of this study are available in the published article and its online supplemental materials.