Airways tissue expression of type I interferons and their stimulated genes is higher in children than adults

Abstract

There is emerging evidence that age-dependent differences in susceptibility to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) correlate with stronger innate immune response in the upper respiratory tract in children compared to adults. The efficient induction of interferon (IFN) alpha and beta (α and β) signaling, and interferon-stimulated genes (ISGs) is fundamental to the host antiviral response. In-silico transcriptomic analyses was conducted to determine the expression levels of IFN α/β pathway genes as well as 524 human ISGs in upper and lower airways of children and adults at baseline and post respiratory infections including coronavirus disease 2019 (COVID-19). To validate our in-silico analysis, we conducted qRT-PCR to measure ISGs levels in children and adult's nasal epithelial samples. At baseline, children had significantly higher levels of IFN α/β and ISGs genes compared to adults. More distinction was also seen in bronchial compared to nasal basal levels. Children nasal epithelial cells exhibited superior antiviral IFN α/β and associated ISGs response following ex-vivo poly (I:C) treatment model, and in clinical samples of SARS-CoV-2 infected patients. This was also confirmed in nasal epithelial samples using qRT-PCR validation. No gender-based difference in type I IFN levels across both age groups were observed. Understanding the biological basis for children resistance against severe COVID-19 is a challenge that has substantial clinical importance. More mechanistic studies are needed to carefully quantify how much of early IFN levels is needed to bypass the viral evasion mechanism and prevent its further replication and dissemination to lower airways and the rest of the body.

SARS-CoV-2; IFN alpha; IFN beta; Type I interferon; Interferon-stimulated genes (ISGs); COVID-19; children; adults; Poly (I:C); RSV.

1. Introduction

There is emerging evidence that age-dependent differences in susceptibility to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) correlate with stronger innate immune response in the upper respiratory tract in children compared to adults. The milder presentation of coronavirus disease 2019 (COVID-19) in children is evident from the lower COVID-19 related morbidity and mortality rates [1]. Among hospitalized patients, children and young adolescent had a shorter length of stay, decreased need for mechanical ventilation, and lower death rate [1].

According to the center for disease control and prevention (CDC), USA, as of 13 Nov 2021, the cumulative rates for laboratory-confirmed COVID-19 associated hospitalization was 67.5 per 100,000 for children aged below 18 years, while adults 18 years and above had an overall hospitalization rate of 922.6 per 100,000 population (Data obtained from https://gis.cdc.gov/grasp/COVIDNet/COVID19_3.html). Moreover, as of 04 December 2021, COVID-19 mortality in children accounted only for 0.1 percent of total COVID-19 deaths in USA (644 out of 787,624); and was higher for boys (55% of deaths among boys, 45% among girls) (Data obtained https://data.cdc.gov/NCHS/Provisional-COVID-19-Deaths-by-Sex-and-Age/9bhg-hcku/data).

Several theories were proposed for why children were less prone to develop severe conditions following infection with SARS-CoV-2 [2]. One theory was based on the lower expression of SARS-CoV-2 entry receptor, angiotensin-converting enzyme 2 (ACE2) in children's upper and lower respiratory tract compared to adults. Some studies have confirmed this theory [3, 4, 5], while others have refuted it and reported similar ACE2 expression levels in the two age groups [6, 7]. Another theory was based on pre-existing cross-reactive immunity due to frequent exposure of children to other common cold human coronaviruses (229E, NL63, HKU1 HCOVs). Here also, few reports showed higher cross-reactivity in children [8]while others have reported similar levels of cross-reactivity in children and adults [1].

More efficient antiviral innate immunity in children is one more theory that could explain their milder presentation. Type I interferons (IFNs), namely IFN alpha and beta (α and β), and their downstream effector interferon-stimulated genes (ISGs) are key inflammatory mediators regulating antiviral innate immune response [9, 10]. Age-related dysregulation in frequency and effector function of innate immune cells has been extensively studied previously including the decrease in IFN production [11]. Moreover, lower type I IFN levels have been linked to more severe viral respiratory infections including SARS-CoV-2 infection [10, 12, 13, 14, 15, 16].

Recent reports have shown higher basal gene expression levels of IFN signaling pathways [7] and relevant pattern recognition receptors such as IFIH1 and DDX58 [6] in upper airway nasal swab samples of children. In this work, we hypothesized that children may respond to viral infection with a higher level of Type I IFN compared to adults. To investigate that we determined the expression levels of IFN α/β pathway genes as well as 524 human ISGs in upper and lower airways of children and adults at baseline and post COVID-19 infection.

2. Materials and methods

Reactome interferon α/β signaling pathway (R-HSA-909733) and Human ISG library consisting of 524 genes [17, 18] were used to screen for IFN/ISGs levels in nasal and bronchial tissues of children and adults at baseline and following viral infection. To do that we used transcriptomic datasets publicly available at National Center for Biotechnology Information Gene Expression Omnibus (NCBI GEO, http://www.ncbi.nlm.nih.gov/geo) and the European Bioinformatics Institute (EMBL-EBI, https://www.ebi.ac.uk). Studies included used either RNA-sequencing or microarray platforms as shown in Table 1. The details of sample isolation, sequencing, and data processing are available at NCBI GEO, and the studies protocols have been published (Table1).

Table 1.

Gene expression datasets used in this study.

Groups	GEO accession	Platform	Sample	Age group	Condition 1	Condition 2
RNA-seq	GSE166161 [23]	GPL16791	Nasal swabs	Adults	Nasal sample from 13 non-infected controls (Age and gender not available)	Nasal sample from 9 RSV-infected patients (Age and gender not available)
RNA-seq	GSE118761 [24, 64]	GPL11154	Nasal and bronchial epithelial cells	Children	Nasal epithelial samples from 14 children (6 ± 6 years and 6 male) and bronchial epithelial samples from 15 children (7 ± 6 years and 8 males) at baseline
RNA-seq	GSE158752 [25, 65]	GPL18573	Bronchial epithelial cells	Adults	17 Adult Controls (40 ± 13 years and 5 males)
Microarray	GSE117827 [19,66]	GPL23126	Nasal swabs	Children	Nasal samples from 6 non-infected controls (3 ± 3 years and 6 males)	Nasal samples from 6 RSV-infected patients (0.4 ± 0.1 years and 2 males)
Microarray	GSE51392 [20, 35]	GPL13158	Nasal epithelial cells	Adults	Nasal samples from 6 non-infected controls (1 male, age not available)	Nasal samples from 6 non-infected controls stimulated with Poly (I:C) (1male, age not available)
Microarray	GSE153428 [21, 34]	GPL23159	Nasal epithelial cells	Children	Nasal samples from 3 non-infected controls (0.4 ± 0.12 years, gender not available)	Nasal samples from 3 non-infected controls stimulated with Poly (I:C) (0.4 ± 0.12 years, gender not available)
RNA-seq	GSE179277 [26, 36]	GPL24676	Nasopharyngeal swab	Children and Adults	Children: Nasal samples from 34 non-infected controls (6 ± 6 years and 15 males) Adults: Nasal samples from 81 non-infected controls (7 ± 5 years and 23 males)	Children: Nasal samples from 38 COVID-19 patients (62 ± 13 years and 38 males) Adults: Nasal samples from 45 COVID-19 patients (56 ± 12 years and 20 males)

Microarray and RNA-seq data were used. For microarray studies (GSE117827 [19], GSE51392 [20], and GSE153428 [21]), CEL data was pre-processed with Robust Multi-Array Average (RMA) technique using R software [22]. The probe set with the largest interquartile range (IQR) of expression value was selected to represent the gene. For RNA-seq studies (GSE166161 [23], GSE118761 [24], GSE158752 [25], and GSE179277 [26]), raw counts were processed using the Bioconductor package limma-voom [27] and presented as log2 counts per million (log CPM). The generated log-transformed normalized intensities were used in Linear Models for MicroArray data (LIMMA) analyses to identify differentially expressed genes between diseased and control groups [28]. We used the default Benjamini-Horchberg correction for multiple testing [29, 30]. Furthermore, to compare the log fold changes between adults and children, multiple t-test comparison using the method of Benjamini was conducted between the different genes. Statistical analyses were performed using R software (v3.0.2) and Prism (v8; GraphPad). For all analyses, p values < 0.05 were considered significant.

2.1. qRT-PCR

Nasopharyngeal swabs were collected from 11 children (mean age of 7 ± 3 years) and 10 adults (mean age of 32 ± 5 years). For poly (I:C) stimulation, nasopharyngeal cells obtained from 8 children and 8 adults were exposed to poly (I:C) (25 $\mathrm{\mu }$g/ml, Sigma-Aldrich, MO, USA; catalog no. P1530) for 4 h. Total RNA was isolated using Trizol reagent according to the manufacturer's instructions (Invitrogen, CA, USA; catalog no. 15596-018) [31]. Complementary cDNA was synthesized from 1μg of RNA using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, CA, USA; catalog no. 4368814) according to the manufacturer's protocol. For cDNA amplification, 5x Hot FirePol EvaGreen qRT-PCR SuperMix (Solis Biodyne, Tartu, Estonia; catalog no. 08-36-00008) was used, and qRT-PCR was performed in QuantStudio 3 Real-Time PCR System (Applied Biosystems) [32]. Primer sequences for MX1, MDA5 (IFIH1), ISG15, and 18s used in qRT-PCR are shown in Table 2. Gene expression was analyzed using the Comparative Ct (ΔΔCt) method upon normalization to the reference gene 18s rRNA [33]. The data was log transformed. Unpaired student t-test was used to compare between the independent groups. (GraphPad Software, San Diego, Calif). For all analyses, P-values < 0.05 were considered significant.

Table 2.

List of primer sequences used in qRT-PCR.

Genes	Forward primer sequence (5′-3′)	Reverse primer sequence (5′-3′)
MX1	GCAAGGTCAGTTACCAGGACTACG	TGATTCCCATTCCTTCCCCGGC
MDA5 (IFIH1)	GGGAGTGGAAAAACCAGAGTGG	GCGGAAGAGCTGTTCAACTAGC
ISG15	TGGGACCTGACGGTGAAGATGC	GCACGCCGATCTTCTGGGTGAT
18s	TGACTCAACACGGGAAACC	TCGCTCCACCAACTAAGAAC

2.2. Ethics statement

All procedures involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Approval for human subject research was granted by the Institutional Review Board of University of Sharjah and included adult informed consent and parental informed consent for children.

3. Results

Using publicly available transcriptomic datasets, we compared the antiviral response between children and adults by determining the expression levels of genes involved in IFN α/β signaling pathway as well as the resultant upregulation of antiviral ISGs. The datasets used in this study are presented in Table 1. IFN α/β production was presented by reactome interferon α/β signaling pathway (R-HSA-909733) consisting of 57 genes while Human ISGs library [17, 18] consisting of 524 genes was used to represent the expression of downstream ISGs. The two lists are presented in Table S1 and S2.

3.1. Higher baseline expression of type I IFN pathway genes and ISGs in children airway tissue compared to adults

To compare the antiviral response between children and adults, we first evaluated the baseline IFN α/β pathway genes expression in nasal and bronchial tissues from children and adults. For children, RNAseq data was obtained from 14 nasal and 15 bronchial tissues using GEO: GSE118761 [24] dataset. For adults, RNAseq of 13 nasal tissues were obtained from GEO: GSE166161 [23] dataset, while data of 17 bronchial tissues were extracted from GEO: GSE158752 [25] dataset.

Nasal tissue direct comparison showed that out of 57 reactome pathway genes, 30 genes were significantly higher in children compared to adults, 6 of which displayed more than 2 log FC difference (Figure 1A and Table S3). This difference was even more noticeable in lower airway bronchial tissue where 41 genes were significantly more expressed in children, 18 of which having greater than 2 log FC difference (Figure 1B and Table S3). Next, we examined the expression of 524 ISGs library in upper and lower airways. Children had significantly higher expression of 242 ISGs (26 ISGs >2 log FC) in nasal tissues, and 262 ISGs (76 ISGs >2 log FC) in bronchial tissues (Figure 1C-D and Table S3) compared to adults. On the other hand, adults had higher expression of 24 ISGs (9 ISGs >2 log FC) in nasal tissues and 101 ISGs (5 > 2 log FC) in bronchial tissues (Figure 1C-D and Table S3) compared to children. Overall, children presented with higher baseline enrichment of both IFN α/β pathway and its downstream stimulated genes. The basal interferon difference was more visible in the lower bronchial tissue than the upper nasal airway.

Figure 1.

Higher baseline expression of IFNα/β pathway genes and ISGs in children airway tissue compared to adults. Basal expression of 57 genes within reactome IFNα/β pathway was examined in (A) nasal tissue and (B) bronchial tissue. Basal expression of 524 ISGs was assessed in (C) upper nasal tissue and (D) bronchial tissue. ISGs that showed more than 2-fold change in children compared to adults were presented. For children, RNAseq data was obtained from 14 nasal tissues and 15 bronchial tissues using GEO: GSE118761 dataset. For adults, RNAseq of 13 nasal tissues were obtained from GEO: GSE166161 dataset, while data of 17 bronchial tissues were extracted from GEO: GSE158752 dataset. Results are presented as log fold change ±SE of gene expression between children and adults. For all analyses, P-values <0.05 were considered significant.

3.2. Stimulation of children nasal epithelial cells with synthetic viral mimic (poly (I:C)) induce higher levels of type I IFN and ISGs responses compared to adults

Next, we examined the antiviral response of children nasal airway epithelial cells to synthetic viral mimic (poly (I:C)) compared to those of adults. We re-analyzed the children dataset deposited by Gustavo Nino (GEO: GSE153428) [21, 34] consisting of three poly (I:C) stimulated versus 3 unstimulated nasal epithelial samples. For adults, we re-analyzed the dataset deposited by Cornelis Drunen (GEO: GSE51392) [20, 35] consisting of 6 poly (I:C) stimulated versus 6 unstimulated nasal epithelial samples.

Poly (I:C) stimulation regulated 39 genes within the IFN α/β pathway in children's cells, while adult's cells displayed change in 49 genes as presented in Figure 2 and Table S4. To compare the upregulation intensities, the 38 shared genes were plotted of which 28 genes were significantly higher in children including USP18 (log FC of 9.3 ± 0.69 vs 5.8 ± 0.52; p-value = 0.0010), ISG15 (log FC of 7.6 ± 0.82 vs 4.5 ± 0.72; p-value = 0.019) and IFNAR2 (log FC of 2.6 ± 0.26 vs 1.3 ± 0.14; p-value = 0.00018).

Figure 2.

Expression of IFNα/β pathway genes following stimulation with synthetic viral mimic in children nasal epithelial cells compared to adults. The significant gene list of children and adults were intersected revealing upregulation of 39 genes in children and 49 genes in adults. Out of these 38 genes were shared and plotted for comparison. Following datasets were used; GSE153428 (n = 3 poly (I:C) stimulated vs n = 3 unstimulated children's nasal epithelial samples), GSE51392 (n = 6 poly (I:C) stimulated vs n = 6 unstimulated adult nasal epithelial samples). Results are presented as log fold change ±SE of gene expression between case and control. Multiple student t-test comparison was used to compare between the independent groups. For all analyses, p < 0.05 was considered significant. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Poly (I:C); Polyinosinic-polycytidylic acid.

We then examined the expression of the 524 ISGs post poly (I:C) stimulation. We compared genes that increased by 2 and 4 log FC (Figure 3 and Table S5). For genes increased by log 2 FC, out of the 127 shared genes, higher expression was seen in children for 90 genes while adults showed higher levels for 5 ISGs (Figure S1). In log 4 FC results, after poly (I:C) stimulation, 69 ISGs were specifically increased in children, 10 ISGs were specifically increased in adults, and 49 ISGs were shared (Figure 3 A-C and Table S5). Of the shared signatures, significantly higher expression of 33 genes was seen in children while the level of only 1 ISGs was higher in adults (Figure 3B and Table S5).

Figure 3.

Expression of 524 interferon-stimulated genes (ISGs) following stimulation with synthetic viral mimic in children nasal epithelial cells compared to adults. Differential expression (DE) analyses between the Poly (I:C) stimulated and non-stimulated groups within each age cohort was performed. Gene lists with log fold change of 4 and higher in children and adult comparison were intersected to obtain the specific and shared signatures. Of upregulated ISGs, (A) 69 were specific to children, (B) 49 were shared between children and adults, and (C) 10 ISGs were specific to adults. Following datasets were used; GSE153428 (n = 3 poly (I:C) stimulated vs n = 3 unstimulated children's nasal epithelial samples), GSE51392 (n = 6 poly (I:C) stimulated vs n = 6 unstimulated adult nasal epithelial samples). Results are presented as log fold change ±SE of gene expression between case and control. Multiple student t-test comparison was used to compare between the independent groups. For all analyses, p < 0.05 was considered significant. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Poly (I:C); Polyinosinic-polycytidylic acid.

Higher levels of Type I IFNs and ISGs responses were observed in nasal epithelial cells of COVID-19 children compared to those of COVID-19 adult patients.

To examine the antiviral interferon response in children and adults with SARS-CoV-2 infection, we used a recently deposited dataset of nasal swabs (GEO: GSE179277) [26, 36]; which is composed of 38 COVID-19 and 34 non-infected children samples and 45 COVID-19 and 81 non-infected adult samples. Both COVID-19 children and adults had 33 differentially expressed genes in the Reactome pathway compared to non-infected controls, of which 29 genes were shared. Of the shared genes, the expression levels of 8 genes were higher (IRF7, IFIT2, ISG15, STAT2, IFTIM1, HLA-F, OAS1, and HLA-A), while 2 genes (STAT1 and IFI27) had lower expression levels in children compared to adults (Figure 4 and Table S6).

Figure 4.

Expression of IFNα/β pathway genes in nasal swabs obtained from SARS-CoV-2 infected children and adult. The significant gene list of children and adults were intersected revealing upregulation of 33 genes in children and 33 genes in adults. Out of these 29 genes were shared and plotted for comparison. GSE179277 dataset was used and included nasal swabs from 34 non-infected controls and 38 COVID-19 patients in the children group, and 81 non-infected controls and 45 COVID-19 patients in the adult group. Results are presented as log fold change ±SE of gene expression between case and control. Multiple student t-test comparison was used to compare between the independent groups. For all analyses, p < 0.05 was considered significant. ∗p < 0.05, ∗∗p < 0.01. SARS-CoV-2; severe acute respiratory syndrome coronavirus 2.

We then examined the level of the 524 ISGs and plotted those genes that had more than 2-fold, or 4-fold change compared to non-infected controls (Figure 5, Figure S2, and Table S7). Nighty seven genes in children and 37 in adults were significantly more than 2-fold higher (Figures 5A and 5C), of which 32 genes were shared. Of the shared ISGs, 9 genes showed higher expression in children than adults (Figure 5B). Between the genes that were upregulated for more than 4-folds, 8 ISGs (OASL, SAA1, IL4I1, OASL, IFIT3, CXCL9, IFIT2, and GZMB) were specific to children, one ISG (IFI6) was specific to adults, and 3 ISGs were shared (CXCL10-11 and ISG15) and were significantly higher in children (Figure S2 and Table S7).

Figure 5.

Expression of 524 interferon-stimulated genes (ISGs) in nasal swabs obtained from SARS-CoV-2 infected children and adult. Differential expression (DE) analyses between the SARS-CoV-2 infected and non-infected controls within each age cohort was performed. Gene lists with log fold change of 2 and higher in children and adult comparison were intersected to obtain the specific and shared signatures. Of upregulated ISGs, (A) 65 were specific to children, (B) 32 were shared between children and adults, and (C) 5 ISGs were specific to adults. GSE179277 dataset was used and included nasal swabs from 34 non-infected controls and 38 COVID-19 patients in the children group, and 81 non-infected controls and 45 COVID-19 patients in the adult group. Results are presented as log fold change ±SE of gene expression between case and control. Multiple student t-test comparison was used to compare between the independent groups. For all analyses, p < 0.05 was considered significant. ∗p < 0.05, ∗∗p < 0.01. SARS-CoV-2; severe acute respiratory syndrome coronavirus 2.

3.3. Similar increase in type I IFN gene expression in children compared to adults was observed following RSV infections

We next examined whether the observed differential Type I IFN responses between children and adults following SARS-CoV-2 infection can be also observed following infections with another respiratory virus such as RSV. For this we used transcriptomic data of nasal epithelial cells obtained from 9 RSV infected and 13 non-infected adults (GSE166161 [23] dataset); and 6 RSV infected and 6 non-infected children using GSE117827 [19] dataset (Table S8). Differential expression analyses was performed between infected and non-infected controls within each age cohort. The significant gene lists were intersected showing 16 common dysregulated genes in the four groups of RSV and COVID-19 infected children and adults. Overall, SARS-CoV-2 infection induced higher levels of IFNα and β genes upregulation; this was more noticeable in children compared to adults (Figure 6).

Figure 6.

Common differentially expressed IFNα/β signatures in nasal swabs obtained from RSV and SARS-CoV-2 infected children and adults. Differential expression (DE) analyses between the infected and non-infected controls within each age cohort was performed. (A) The significant gene lists were intersected showing 16 common dysregulated genes in the four groups of RSV and COVID-19 infected children and adults. Type I IFN gene expression in children compared to adults following (B) RSV infection, and (C) COVID-19. The Multiple student t-test comparison was used to compare between the independent groups. Results are presented as log fold change ±SE of gene expression between case and control. For all analyses, p < 0.05 was considered significant. ∗p < 0.05, ∗∗p < 0.01. RSV; Respiratory syncytial virus (RSV), SARS-CoV-2; severe acute respiratory syndrome coronavirus 2.

3.4. Similar type I IFN gene expression levels in males and females comparisons for both children and adults

We next performed sub-group analyses for each age group to compare the IFN signaling between the males and females at baseline and following COVID-19 infections. These analyses did not reveal any gender-based difference in type I IFN levels across both age groups (Table S9).

3.5. Validation of the observed ISGs differential expression using Poly (I:C) treated nasal epithelial cells

To validate the in-silico analyses, we examined the mRNA expression of MDA5 (IFIH1), MX1 and ISG15 at baseline and post-treatment with poly (I:C) by means of qRT-PCR. The ISGs were significantly higher in children at baseline and following stimulation, especially the downstream effectors, MX1 and ISG15 (Figure 7).

Figure 7.

qRT-PCR Validation in Poly (I:C) treated nasal epithelial cells. Higher baseline expression of MDA5 (IFIH1), MX1 and ISG15 in children's nasal epithelial cells (n = 11) as compared to adults (n = 10). Children nasal epithelial cells induce higher levels of MDA5 (IFIH1), MX1 and ISG15 following stimulation with synthetic viral mimic (poly (I:C)). Nasopharyngeal cells obtained from 8 children and adults were exposed to poly (I:C) (25 μg/ml) for 4 h. Gene expression was analyzed using the Comparative Ct (ΔΔCt) method upon normalization to the reference gene 18s rRNA. Two-way analysis of variance (ANOVA) and post hoc Tukey multiple comparison analyses were applied to compare between the independent groups. ∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

4. Discussion

Herein in-silico transcriptomic datasets of children and adults were used to determine the airway expression of IFN α/β signaling genes and the downstream effector ISGs at baseline and following viral infection. To our knowledge, this is the first report to analyze the differential expression between children and adults of almost all genes involved in type I IFN production and signaling. A remarkable difference was seen at baseline levels where children had noticeable higher expression levels of IFN α/β genes and ISGs (Figure 1). More distinction was observed in bronchial rather than nasal basal IFN/ISGs levels. Children nasal epithelial cells exhibited superior antiviral IFN α/β and associated ISGs responses following ex-vivo poly (I:C) stimulation model (Figures 2 and 3); and in clinical samples of SARS-CoV-2 infected patients (Figures 4 and 5). We have validated the in-silico findings using nasal epithelial cells isolated form children and adults; and where, once again, children showed higher expression of ISGs at baseline and following poly (I:C) stimulation (Figure 7).

The efficient induction of IFN α/β signaling and ISGs in virus-infected cells is fundamental to the antiviral response of the host. Recently, our group has contributed to the discovery that genetic defect in type I IFN production including TLR3, and IRF7, or presence of neutralizing autoantibody against type I IFNs could cause life-threatening COVID-19 disease [37, 38, 39, 40]. Knowing that SARS-CoV-2 virus is a poor inducer of IFN cytokines further suggest that even a small decrease in type I IFNs levels could increase host vulnerability to severe COVID-19 development [41].

Type I interferon cytokines include more than 13 IFNα subtypes, a single IFNβ, and other IFNs including IFN-ω, -ε, and -κ [42]. These antiviral proteins bind to a dimeric receptor composed of IFNAR1 and IFNAR2 subunits and result in formation of two main transcription complexes, namely IFN-stimulated gene factor 3 (ISGF3) and IFNA-activated-factor (AAF) [43, 44, 45]. Upon IFN signaling, ISGF3, which consists of phosphorylated STAT1/STAT2 and IRF9, migrates to nucleus, binds to interferon-stimulated response elements (ISREs), and activates transcription of several IFN-inducible genes including IRF1-9, OAS proteins, Mx GTPases, IFTIM1-3, and IFITs [43]. The other non-canonical STAT-based signaling pathway involves AAF which is a homodimer of STAT1 [43]. AAF translocate to nucleus and activates IRF1 and IRF1 inducible genes including ISG15, ISG54, and IFI6 [46]. Both the canonical and non-canonical IFN α/β signaling genes are presented in Reactome pathway which was adapted in this study. IFN α/β signatures were found to be upregulated at baseline and following SARS-CoV-2 infection in children compared to adults. At baseline, more than half of the IFN signature genes were higher in children in both nasal and bronchial tissues (Figure 1 A-B). Following SARS-CoV-2 infection, the expression of 8 genes was higher (IRF7, IFIT2, ISG15, STAT2, IFTIM1, HLA-F, OAS1, and HLA-A), while that of 2 genes (STAT1 and IFI27) was lower in children compared to adults (Figure 4).

ISGs are antiviral proteins that target diverse steps in the viral replication cycle, including viral entry (IFITM1 and IFITM3), nuclear import (MX1 and MX2), viral mRNA synthesis (IFI6 and MX1), protein synthesis (IFIT1, IFIT2), viral replication (IFI6), and viral degradation (ISG20, OAS1-3) [47]. All these genes were significantly upregulated in children at baseline (Figure 1C-D). Following SARS-CoV-2 infection, IFITM1, IFIT2, and OAS1 were significantly higher, IFTIM3, IFIT1, ISG20 had an increasing trend, while IFI6 had lower expression in children compared to adults (Figure 5 and Figure S2). Another group of ISGs are negative regulators of IFN signaling and could promote viral replication [47]. This group includes SOCS1, USP18, ADAR, and ABCE1 [47,48]. Among these, the basal expression levels of SOCS1 and USP18 were higher in children nasal and bronchial tissue compared to adults. Interestingly, the gene expression levels of a small number of IFNs/ISGs including IRF9 and JAK1, were lower at baseline but higher or similar following either stimulation with Poly I:C or viral infection in children (Figures 2 and 4). Upon Type I interferon signaling, the stimulated ISGs includes inhibitory regulators such as USP18 which then upregulate negative feedback inhibition of JAK-STAT and prevent prolongation of ISGF3-independent IFNAR signaling [49, 50, 51]. Therefore, considering the upstream position of IRF9 and JAK1, the observed downregulation could possibly be due to the negative feedback mechanism following production of downstream inhibitory ISGs; USP18 and SOCS1. Additional studies are needed to validate the level of these genes at baseline and following infection, and to determine their contribution to the net Type I IFN signaling outcome.

Ageing affects innate immune cells frequency, migration and effector functions [11]. Plasmacytoid DCs (pDCs) which are the main interferon producing cells decreases in number with ageing [52, 53]. Furthermore, pDCs have displayed age-associated decrease in type I IFN production levels which correlated with decrease in IRF7 nuclear translocation [11, 54, 55]. Age-associated impairment in pDCs migration was also reported and could potentially affect their antiviral response [55]. Therefore, this suggests that age-associated innate immune dysregulation resulting in lower basal IFN expression and weaker antiviral response could contribute to the severe clinical presentation of the disease in older adults; while children with higher basal IFN levels better control SARS-CoV-2 infection.

High basal expression of ISGs in human bronchial epithelial (BEAS-2B) cells induced resistance to influenza A virus infection [56]. More recently, basal expression levels of ISGs including IFITM3, MX1, and OAS3 were associated with strong IFN antiviral response post-influenza infection, and may contribute to individual disparity to infection susceptibility [57]. In addition, higher basal expression of IFIH1 and DDX58 in upper airway epithelial cells, macrophages, and dendritic cells of children as compared to adults was correlated to more efficient antiviral response to SARS-CoV-2 infection and milder presentation of the disease [6]. Our analysis also showed that children displayed higher basal expression of type I IFNs and their associated ISGs both in lower and upper airways, although the difference was more noticeable in bronchial tissue (Figure 1).

SARS-CoV-2's hallmark mechanism of immune evasion includes its effects on type I and III IFNs at multiple levels which results in a delayed or muted interferon response promoting viral replication and quick dissemination of the virus from upper respiratory tract to lower lungs [58, 59]. Understanding the biological basis for children resistance against severe COVID-19 is a challenge that has substantial clinical importance. Multiple factors could contribute to the significant disparity between children and adults COVID-19 outcomes. Out of those, higher basal IFN expression levels could be helpful in scenarios of SARS-CoV-2 induced IFN suppression and host delayed immune response. Higher basal IFN/ISGs level in children and their mild COVID-19 presentation encourage consideration of earlier interferon treatment in adults, especially those at higher risk. Multiple studies have now shown reduced mortality following earlier initiation of interferon therapy in COVID-19 patients [60, 61, 62, 63]. More mechanistic studies are needed to carefully quantify how much of early IFN levels is needed to bypass the viral evasion mechanism and prevent its further replication and dissemination to lower airways and the rest of the body.

Declarations

Author contribution statement

Rabih Halwani: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper.

Narjes Saheb Sharif-Askari, Fatemeh Saheb Sharif-Askari: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Qutayba Alheialy, Mohamed Temsah, Saleh Almuhsen, Habiba S Alsafar: Conceived and designed the experiments; Wrote the paper.

Shirin Haghani, Zaina Kalaji: Performed the experiments; Wrote the paper.

Funding statement

Rabih Halwani was supported by the University of Sharjah (COV19-0307) and Al Jalila Foundation (AJF202019).

Data availability statement

Data associated with this study has been deposited at “National Center for Biotechnology Information Gene Expression Omnibus” under the accession number GSE166161, GSE118761, GSE158752, GSE117827, GSE51392, GSE153428, GSE179277.

Declaration of interest's statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.

Appendix A. Supplementary data

The following is the supplementary data related to this article: