High expression of interleukin-18 receptor alpha correlates with severe respiratory viral disease and defines T cells with reduced cytotoxic signatures

Abstract

Hyperactivated immunity underpins severe outcomes of respiratory viral infections, yet specific immune perturbations are ill-defined. Our recent findings identified OLAH (oleoyl-ACP-hydrolase) as a driver of life-threatening viral diseases. In the same patient cohorts, we now identify the gene encoding IL-18Rα chain (IL18R1), as being highly expressed in life-threatening influenza, COVID-19, RSV and multisystem inflammatory syndrome in children (MIS-C) and demonstrate markedly elevated surface protein IL-18Rα expression on CD8 T cells in these infections. Using a mouse model of severe influenza, we further show that high IL-18Rα expression on effector T cells is associated with increased disease severity. We find that IL-18Rα expression on CD8 T cells is inversely associated with cytotoxicity-related genes, including granzyme A, granzyme B, perforin, Eomes, and KLRG-1. Our study demonstrates that IL-18Rα is associated with severe and fatal respiratory disease outcomes and proposes the use of IL-18Rα as a potential biomarker for severe respiratory viral disease.

Fig. 1. High IL18R1 expression in patients with fatal H7N9 infection and with life-threatening complications of SARS-CoV-2 and RSV infection.

a Summary of disease cohorts, group sample sizes, age range and generated data used for analysis. b Volcano plot showing DEGs between fatal (left) and recovery (right) groups from whole blood microarray of early infection phase samples. c IL18R1 transcript expression levels (n = 4 per group at early and late timepoint, mean ± SEM) across early and late time points in A/H7N9 patients. (d) Bulk RNA sequencing on whole blood from healthy volunteers and hospitalized patients with acute COVID-19 or MIS-C, aged <21 years (n = 143). IL18R1 expression levels in healthy individuals (n = 22) and hospitalized SARS-CoV-2 patients: no/minimal respiratory dysfunction (n = 43), moderate/severe (n = 25), life-threatening n = 23), diagnosed with MIS-C, as described (n = 30). P values were obtained from a model that controlled for days since symptoms onset, sex, whether a patient was previously healthy, steroid administration prior to sampling, bacterial co-infection, age, race and ethnicity, and were adjusted for multiple comparisons. e IL18R1 expression in single-cell RNA transcriptomic data acquired from tracheal aspirate samples obtained from RSV-infected children with no/mild (n = 5) or moderate/severe pARDS (n = 7) (left), and children with non-RSV-related infection but with moderate/severe pARDS (n = 5) as well as control children without acute LRTI or lung injury (n = 6) (right). Box plot hinges are the first and third quartiles (the 25th and 75th percentiles). The line between hinges corresponds to the media. Whiskers show ± 1.5 * IQR from the hinge (where IQR is the inter-quartile range) f IL18R1 expression across time for human challenge models of mild respiratory infections (H1N1 DEE3 n = 477, H3N2 DEE2 n = 355, HRV Duke n = 471, RSV DEE4 n = 420). Smooth curves fit using a LOESS model. Shaded bands show the 95% confidence interval. Linear mixed models were used to test for effects of time, infection status, and the interaction of time with infection status, with subject included as a random effect. Infection status was not significant for any study.

Fig. 2. High surface expression of IL-18Rα on human CD8 T cells, CD4 T cells and tetramer⁺ CD8 T cells during severe respiratory disease.

IL-18Rα on human a CD8 T cells, b CD4 T cells and c tetramer⁺ CD8 T cells. a, b Representative FACS plots of surface IL-18Rα expression on CD8 T cells and CD4 T cells in a healthy individual and in hospitalized RSV, COVID-19 and influenza A patients are shown; d: days post disease onset. Graphed IL-18Rα expression in healthy individuals (n = 17) and patients hospitalized with influenza A (n = 37), influenza B (n = 1), RSV (n = 4) and COVID-19 (n = 2) at all visit (V) time points, grouped by ICU and ward or oxygen support is shown. IL-18Rα expression in healthy individuals versus influenza A patients is shown at hospital visit 1 (V1). Bars indicate median. Statistical significance was analysed using a two-tailed Mann-Whitney or Kruskal-Wallis. c Representative FACS plots of surface IL-18Rα expression on unenriched CD8 T cells (n = 13), tetramer-enriched influenza-specific A2/M1₅₈⁺CD8 T cells (n = 12) and unenriched CMV-specific A2/pp65₄₉₅⁺CD8 T cells (n = 9) in influenza A patients. Graphed IL-18Rα expression in HLA-A2⁺ patients hospitalized with influenza A at acute timepoints (mean ± SD, two-tailed Wilcoxon matched-pairs signed rank test). Patient demographics are outlined in Supplementary Tables 1 and 2.

Fig. 3. High surface expression of IL-18Rα on T cells in a severe influenza A virus patient case.

a Timeline of influenza A infection, showing hospital admission to the Emergency Department (ED) and intensive care unit (ICU) on day 17 post disease onset (V1), recovery and admission to General Ward on day 29 and discharge on day 31. Blood samples were obtained on days 17 (V1), 21 (V2) and 24 (ICU, V3) and day 31 (General Ward, V4). b Proinflammatory cytokines and chemokines in patient sera. surface IL-18Rα expression by flow cytometry, shown as c %IL-18Rα expression and d mean fluorescence intensity (MFI) on CD8 T cells, CD4 T cells and monocytes in the influenza A patient.

Fig. 4. High IL-18Rα expression on αβ T cells correlates with increased influenza disease severity.

a Mean frequency of IL-18Rα⁺ cells on innate and adaptive immune cells in the lung of naïve mice (N, n = 5) and on 1, 3, 6, 10, and 28 dpi in mild (n = 5) and severe groups (n = 4–5). b Correlation between body weight loss and IL-18Rα expression on CD4 and CD8 T cells in the lung, 6 dpi (grey bands show 95% CI). r² = Pearson’s correlation coefficient. Data pooled from 2 independent experiments. c IL-18Rα⁺ T cell frequencies in lungs at 3 dpi (n = 5 per group, mean ± SEM, two-tailed unpaired t-test). d Correlation between body weight loss and IL-18Rα on CD8 T cells based on CD44 and CD62L expression (Pearson’s correlation, 95% CI). e PD-1 and CD38 on IL-18Rα⁺ and IL-18Rα^- CD8 T cells in mild (n = 9) and severe (n = 8) groups (two-tailed paired t-test) (left). Frequencies of IL-18Rα⁺PD-1⁺ CD8 T cells in the lung of mild (n = 9) and severe (n = 8) groups (mean ± SEM, two-tailed Mann-Whitney). f Co-expression of IL-18Rα and PD-1 in lungs of mild (n = 8) and severe group (n = 9) on day (mean ± SEM, two-tailed Mann-Whitney). g Correlation between body weight loss and frequencies of IL-18Rα⁺PD-1⁺ cells (Pearson’s correlation, 95% CI). h Representative histograms showing IL-18Rα on influenza-specific and non-specific CD8 T cells during acute (10 dpi) and memory (28 dpi) phases of primary IAV infection (2×10⁴ pfu A/HKx31), and on 8 dpi following secondary challenge (10⁵ pfu A/PR8). i IL-18Rα on D^bPA₂₂₄ and D^bNP₃₆₆ CD8 T cells and non-specific (tetramer^-) CD8 T cells (RM one-way ANOVA, Geisser-Greenhouse and Tukey’s correction, n = 7–9, mean ± SEM). j Frequency of naïve cells within D^bNP₃₆₆ and D^bPA₂₂₄-tetramer negative CD8 T cells (n = 7–9, mean ± SEM, Holm-Sidak’s one-way ANOVA). IL-18Rα on different memory CD8 T cell subsets in the lung, 28 dpi. IL-18Rα⁺ k frequencies and l geometric mean fluorescence intensity (gMFI) in naïve (T_naive, CD44^-, CD62L⁺), central memory (T_CM, CD44⁺CD62L⁺, light red) and effector memory (T_EM, CD44⁺CD62L^-, yellow) CD4 and CD8 T cell subsets in lung and spleen, 28 dpi (n = 4 per group, RM two-way ANOVA, Geisser-Greenhouse and Tukey’s correction, mean ± SEM). Data shown on 6, 10, and 28 dpi are from two independent experiments.

Fig. 5. CD8 T cells expressing high and low levels of IL-18Rα possess distinct transcriptomic profiles.

a 1 × 10⁵ CD45.1⁺ OT-I cells adoptively transferred into C57BL/6 mice. Mice were infected with 10⁵ pfu of A/HK x 31-OVA i.n. and lungs were collected on 6 dpi and sorted for cells expressing high (IL-18Rα^hi) and low-to-intermediate levels of IL-18Rα (IL-18Rα^lo). Created in BioRender. Cabug, A. (2025) https://BioRender.com/q5po0oj. b Antigen-experienced OT-I cells were sorted as CD44⁺CD45.1⁺TCRVα2⁺, then sorted into IL-18Rα^hi and IL-18Rα^lo subpopulations. c Number of differentially expressed genes (DEG) between IL-18Rα^hi and IL-18Rα^lo OT-I cells. d Volcano plot showing the DEGs between IL-18Rα^hi and IL-18Rα^lo cells. Horizontal line indicates -log₁₀(P = 0.05). e Heatmaps showing expression levels of select genes grouped according to functional clusters. Each square represents an individual mouse where the blue and orange bars represent paired IL-18Rα^lo and IL-18Rα^hi cells, respectively (n = 6). Z-score values are shown. f DEGs in each subpopulation were analysed for protein-protein interaction based on the calculation of a STRING interaction score. Bubble size is FDR and colour intensity represents the magnitude of the difference in gene expression between IL-18Rα^hi and IL-18Rα^lo subsets (STRING score>0.7). Red networks represent genes more highly expressed in IL-18Rα^hi and blue gene networks are those more highly expressed in IL-18Rα^lo. g DEGs were compared to the MSigDB hallmark gene sets. Bubble size corresponds to gene overlap and represents the number of DEGs enriched in gene set. Bubble colours show the -log₁₀(adjusted P-value). h IFNγ intracellular cytokine staining of IL-18Rα^lo and IL-18Rα^hi CD8 T cells from lungs on day 6 following influenza virus infection following IL-12, IL-18 or IL-12 + IL-18 stimulation. Unstimulated negative control and PMA/I stimulated positive control were included. IFNγ production by IL-18Rα^hi CD8 T cells (top); IFNγ production by IL-18Rα^lo CD8 T cells (bottom), (n = 5, mean ± SD, one-way ANOVA, Kruskal-Wallis, Dunn’s multiple comparisons).

Fig. 6. IL-18Rα^lo influenza-specific CD8 T cells are KLRG1^hi exhibit increased production of cytotoxic effectors.

a C57BL/6 mice were infected with 2 × 10⁴ pfu of A/HKx31, and IL-18Rα^hi (orange) and IL-18Rα^lo (blue) influenza tetramer-specific CD8 T cells were analysed on 10 dpi. b Representative histograms (left) showing the expression of Eomes and TCF1 in IL-18Rα^hi and IL-18Rα^lo D^bNP₃₆₆-specific CD8 T cells. gMFI values shown on the right. Representative FACS plots and graphs comparing the frequency of IL-18Rα^hi and IL-18Rα^lo tetramer-specific cells expressing (c) KLRG1 and (d) NRP1. e CX3CR1 expression was compared between tetramer-specific IL-18Rα^hi and IL-18Rα^lo cells. Representative populations showing expression of f granzyme A, g granzyme B and h perforin compared between IL-18Rα^hi and IL-18Rα^lo groups. i Frequencies of IL-18Rα^hi and IL-18Rα^lo tetramer-specific CD8 T cells expressing cytotoxic molecules in the lungs. Data show pooled D^bPA₂₂₄ and D^bNP₃₆₆ specificities. j gMFI of cytotoxic mediators expressed by IL-18Rα^hi and IL-18Rα^lo tetramer-specific CD8 T cells in the lungs. Data show pooled D^bPA₂₂₄ and D^bNP₃₆₆ specificities. k Polyfunctional profiles (granzyme A, granzyme B, perforin) of pooled tetramer-specific IL-18Ra^lo and IL-18Ra^hi CD8 T cells in the lungs and spleen. Data were analysed via permutation test. l Volcano plot summarising differences in expression of all transcription factors, surface markers, and cytotoxic effectors (total n = 68 parameters) analysed. Data were analysed via multiple paired t-test with Holm-Sidak’s correction for multiple comparisons. The horizontal dotted line is equivalent to p = 0.05 and the vertical lines denote a log₂(fold change) of ±0.5. For (a–l), data were analysed using a two-tailed paired t-test where n = 7–9 mice per group. Data shown are from two independent repeats.

Fig. 7. IL-18Rα^hi CD8 T cells are highly polyfunctional.

a Frequency of IFNγ-secreting IL-18Rα^hi and IL-18Rα^lo CD8 T cells in the lung and spleen was analysed ex vivo 10 dpi after peptide stimulation with immunodominant NP₃₆₆ and PA₂₂₄ peptides. b Comparison of TNF⁺ producers in IFNγ⁺IL-18Rα^hi and IL-18Rα^lo CD8 T cells populations. c Frequencies of IL-2⁺ of IFNγ⁺TNF⁺CD8⁺IL-18Rα^hi and IL-18Rα^lo T cells. Data were analysed using a two-tailed paired t-test (n = 15). d Polyfunctional profiles (IFNγ, TNF, IL-2) of IL-18Rα^lo and IL-18Rα^hi CD8 T cells after peptide stimulation in the lungs and spleen. Data were analysed via a permutation test. Data shown are from three independent experiments.

Fig. 8. Inverse correlations between IL-18Rα and cytotoxic T cell signatures associated with severe disease outcomes in human A/H7N9 and COVID-19 cohorts.

Correlations between IL18R1 expression and GZMA (Granzyme A), GZMB (Granzyme B), PRF1 (Perforin 1), EOMES, KLRG1 and CX3CR1 were analysed in our A/H7N9 and COVID-19 datasets. a, b Associations between GZMA (p = 8.5e-09), GZMB (p = 0.03), PRF1 (p = 1.4e-06), EOMES (p = 2.2e−16), KLRG1 (p = 6.1e−10) and CX3CR1 (p = 1.5e−08) expression levels with disease severity was also analysed. c Significant correlations with IL18R1 expression and disease severity for differentially expressed genes from our OT-I experiment (Fig. 4), namely IL1R1L1, TIGIT, IL10RA and XCL1. R = spearman’s correlation (two-tailed), where * = p < 0.05 l;** = p < 0.01; *** = p < 0.001. Box plot hinges are the first and third quartiles (the 25th and 75th percentiles). The line between hinges corresponds to the median. The upper whisker extends from the hinge to the largest value no further than 1.5 * IQR from the hinge (where IQR is the inter-quartile range, or distance between the first and third quartiles). The lower whisker extends from the hinge to the smallest value at most 1.5 * IQR of the hinge.