Elevated nucleoprotein-induced interferon-γ release in COVID-19 patients detected in a SARS-CoV-2 enzyme-linked immunosorbent spot assay.

Dear Editor,

The mechanisms underlaying the host defense against SARS-CoV-2 remain largely unknown. The relative contribution and importance of the innate, humoral and cellular immune response have to be elucidated to improve our understanding of coronavirus disease 2019 (COVID-19) pathogenesis and to aid vaccine development. In a recent article in the Journal, Demey and colleagues presented data on four lateral flow assays (LFIA) for the detection of antibodies directed against SARS-CoV-2. They assessed the kinetics of antibody appearance using these assays in 22 patients after they were tested positive by RT-PCR. They reported a sensitivity of 100% on day 15 post onset of symptoms.

In addition, various other studies suggest a suppressed T-cell immunity in patients with severe COVID-19 based on decreased T-cell numbers or abnormal interferon gamma (IFN-γ) expression by T-lymphocytes detected by flowcytometry.2, 3, 4

The objective of the present study was to determine the cellular and humoral immunity of cases with various levels of COVID-19 disease severity. The functional T-cell responses to two SARS-CoV-2 antigens (the mosaic surface protein and the nucleoprotein) were measured by using an inhouse enzyme-linked immunosorbent spot (ELISpot) interferon-γ release assay (see supplementary information for method), in 27 patients with confirmed COVID-19   and 16 healthy controls. Patienst were confirmed by using reverse transcriptase polymerase chan reaction (RT-PCR). Of the 27 COVID-19 patients, nine were included from the intensive care unit (ICU) and 18 from the pulmonary ward. The moment of sampling varied from six to 32 days post onset of symptoms. In addition, the concomitant humoral immune response was assessed by detection of SARS-CoV-2 specific IgA and IgG antibodies, directed against the structural protein (S1 domain) of SARS-CoV-2, with a commercial enzyme-linked immunosorbent assay (ELISA) (EUROIMMUN Medizinische Labordiagnostika AG, Lϋbeck, Germany).

The SARS-CoV-2-specific T-cell response measured in the ELISpot induced by the mosaic surface protein and the nucleoprotein showed different patterns. In all but one of the 27 COVID-19 cases, the T-cell response against the mosaic surface protein was absent or weak, with ELISpot results with less than 20 spot forming cells (SFC). The outlier was recruited from the pulmonary ward and exhibited 45 SFC (Fig. 1 a). In contrast, the T-cell response against the nucleoprotein measured by the ELISpot assay was elevated (10–150 SFC) in 12 of 19 patients (63%) that were sampled at ≥14 days post onset of symptoms (Fig. 1b). A subgroup of 9 (Fig. 1b, red oval) showed a delayed or reduced T-cell response against the nucleoprotein, compared to the other patients. Five of these showed practically no response, and four showed a weak response (10–20 SFC) at 18–32 days post onset of symptoms. Absolute lymphocyte numbers loaded in the SARS-CoV-2 ELISpot were not significantly different between the normal and the delayed or reduced responders (data not shown). However, the number of spot forming cells following stimulation with the mitogen control was also significant lower in the delayed or reduced responders (P < 0.001) (see supplementary Figure 1). Moreover, SARS-CoV-2 IgA and IgG antibody levels did not differ between the normal and delayed or reduced responders (data not shown).

Fig. 1

SARS-CoV-2 ELISpot mosaic surface protein (a) and nucleoprotein (b) IFN-γ spot forming cells (SFC) in relation to days post onset of symptoms. SARS-CoV-2 specific IgG antibody response in COVID-19 patients versus days post onset of symptoms (c). Open and closed circles represent COVID-19 patients from the ICU and the pulmonary ward, respectively (a–c). The red oval encloses patients which seem to have a delayed or reduced T-cell response (b). Correlation between T-cell reactivity (SFC) against the SARS-CoV-2 mosaic surface protein (d) and the SARS-CoV-2 nucleoprotein (e) and concomittant SARS-CoV-2 antibody responses (IgA open symbols, IgG closed symbols) in COVID-19 patients (black symbols) and healthy controls (green symbols). The broken line represents the cut-off of the SARS-CoV-2 antibody ELISA. Fig. 1f depicts the ROC analyses of the SARS-CoV-2 nucleoprotein ELISpot results in COVID-19 patients at >7 days, >14 days and >21 days post the onset of symptoms versus healthy controls.

Serology showed a sigmoidal pattern, with a sharp increase in specific IgA (see supplementary Figure 2) and IgG antibodies (Fig. 1c) against the structural protein (S1 domain) of SARS-CoV-2 around 14 and 15 days post onset of symptoms, respectively. Most of the healthy controls showed antibody levels below the cut-off. Except four controls, who showed detectable anti-SARS-CoV-2 IgA antibody levels, while anti-SARS-CoV-2 IgG antibodies were negative (see supplementary Figure 3).

Figs. 1d and 1e depict the combined T- and B-cell response in COVID-19 patients and healthy controls. Thirteen (48%) of the 27 COVID-19 patients had 10 or more SFC in response to stimulation with the nucleoprotein, whereas none of the healthy controls reached that level. One COVID-19 case and two healthy controls showed a strong T-cell reactivity (of more than 30 SFC) against the mosaic surface protein as measured in the ELISpot assay. This COVID-19 case also showed the highest T-cell reactivity against the nucleoprotein (146 SFC) as measured in the ELISpot assay. SARS-CoV-2 IgA and IgG antibodies were positive in the COVID-19 case and negative in the two healthy controls.

As none of the healthy controls had more than nine SFC specific for the SARS-CoV-2 nucleoprotein in the ELISpot assay, 10 or more spots was determined to be indicative for COVID-19 disease. Prolonged illness, i.e. when sampled more days post onset of symptoms, increased the chance of finding higher numbers of spots. Receiver operating characteristc (ROC) analysis was performed for the nucleoprotein ELISpot results at >7, >14 and >21 days post onset of symptoms. All ROC analyses showed significant areas under the ROC curve, respectively 0.77 (p = 0.004), 0.82 (p = 0.001) and 1 (p = 0.002) for detection of COVID-19 disease (Fig. 1f).

Interestingly, in a recent study published by Grifoni et al. SARS-CoV-2 epitope pools were used to probe CD4+ T-cell responses. They found that M, spike and N proteins were co-dominant, and that each protein was recognized by 100% of the 20 COVID-19 cases studied. With respect to SARS-CoV-2 CD8+ T-cell responses, the spike protein was less dominant, while significant reactivity was noted for M, N and other antigens. Similarly, in our study T-cell reactivity was detected in the SARS-CoV-2 ELISpot assay against the nucleoprotein in the majority of patients. In contrast however, the mosaic surface protein, consisting of exposed extracellular domains of the SARS-CoV-2 spike, envelope and membrane proteins generated only a very modest T-cell response in most patients. As these are all trans-membrane proteins and therefore poorly soluble in aqueous solutions, the use of native proteins was not technically feasible. It is therefore possible that the mosaic nature of the recombinant protein and the production of the protein in E. coli could have affected the potential of this protein to elicit spot formation in the ELISpot.

Further studies investigating the association between SARS-CoV-2 neutralizing antibodies and ELISpot reactivity might reveal whether patients develop a protective immunity after COVID-19.

Authors’ contributions

ST, GL, AB,MH and RK designed and performed the study and/or analyzed data. ST and MH wrote the manuscript. HG, RK, CR, KK, GL and AB provided intellectual input and advice on study design and analysis.

Declaration of Competing Interests

none