key: cord-0899160-64gw7pec
authors: Loske, J.; Röhmel, J.; Lukassen, S.; Stricker, S.; Magalhaes, V. G.; Liebig, J.; Chua, R. L.; Thürmann, L.; Messingschlager, M.; Seegebarth, A.; Timmermann, B.; Klages, S.; Ralser, M.; Sawitzki, B.; Sander, L. E.; Corman, V. M.; Conrad, C.; Laudi, S.; Binder, M.; Trump, S.; Eils, R.; Mall, M.; Lehmann, I.
title: Pre-activated anti-viral innate immunity in the upper airways controls early SARS-CoV-2 infection in children
date: 2021-06-28
journal: nan
DOI: 10.1101/2021.06.24.21259087
sha: 020f1a9d5033640ad3a3f92f60364feed5d8c225
doc_id: 899160
cord_uid: 64gw7pec

Children are consistently reported to have reduced SARS-CoV-2 infection rates and a substantially lower risk for developing severe COVID-19. However, the molecular mechanisms underlying protection against COVID-19 in younger age groups remain widely unknown. Here, we systematically characterized the single-cell transcriptional landscape in the upper airways in SARS-CoV-2 negative and age-matched SARS-CoV-2 positive children (n=42) and corresponding samples from adults (n=44), covering an age range of four weeks to 77 years. Children displayed higher basal expression of the relevant pattern recognition receptor (PRR) pathways in upper airway epithelial cells, macrophages, and dendritic cells, resulting in stronger innate antiviral responses upon SARS-CoV-2 infection compared to adults. We further detected distinct immune cell subpopulations with an overall dominance of neutrophils and a population of cytotoxic T cells occurring predominantly in children. Our study provides evidence that the airway epithelial and mucosal immune cells of children are pre-activated and primed for virus sensing, resulting in a stronger early innate antiviral responses to SARS-CoV-2 infection compared to adults.

It has repeatedly been reported that younger individuals have a substantially lower risk for developing COVID-19, despite a similar risk of infection, as reflected in dramatically increased mortality with increasing age [1] [2] [3] . These observations suggest that children may have a higher capability of controlling SARS-CoV-2 infection. It has been shown that an early cell-intrinsic innate immune response, mediated by pattern recognition receptors (PRR) and the type I and III interferon (IFN) system, are crucial for the successful control of SARS-CoV-2 infection 4 . In line with these observations, recent studies compared adults and children with severe COVID-19 or those presenting to an Emergency Department and described an impaired IFN response in pediatric COVID-19 5, 6 . However, the molecular mechanisms protecting against COVID -19 in younger age groups particularly in those with no or only mild/moderate symptoms remain unknown.

To understand the higher capacity of children for controlling SARS-CoV-2 infection at an early stage we systematically characterized the transcriptional landscape of upper airways, an airway region with high susceptibility for SARS-CoV-2 infection 7 , in SARS-CoV-2 negative and SARS-CoV-2 positive children (n=42) and adults (n=44), comprising 268,745 cells in total (Fig.   1a ). To this end, we included study participants of three different COVID-19 cohorts: the RECAST study focusing on COVID-19 in children and their families, the Pa-COVID-19 and the SC2 study 8, 9 . Samples from the upper airways (nose) were collected from individuals aged 4 weeks to 77 years with a positive SARS-CoV-2 PCR result along with age-matched SARS-CoV-2 negative controls (Suppl. Table 1 , 2). Focusing on early infection only mild/moderate COVID-19 cases were considered for this study (Fig. 1a) . Based on the single cell RNA sequencing data we identified 33 different cell types or states in the upper respiratory tract of these individuals including 21 immune and 12 epithelial cell subtypes (Fig. 1b , Extended Data Fig.1 a, b) . We observed striking differences between the pediatric and adult study participants regarding the composition of the immune cell and epithelial cell compartment in the nasal mucosa. While immune cells were rarely detected in nasal samples from healthy adults, samples from SARS-CoV-2 negative children contained high amounts of almost each immune cell subset with an overall dominance of neutrophils (Fig. 1b, c, Extended Data Fig.2a) . In . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint adults, SARS-CoV-2 infection was associated with immune cell influx, while the proportion of immune and epithelial cells remained nearly stable in children (Fig. 1c, Extended Data Fig.   2a ). Upon infection, children`s neutrophils showed an activated phenotype that was more pronounced than in infected adults, characterized by the enhanced expression of e.g. CCL3

and CXCR1/2 (Suppl. Table 4 ).

Interestingly, many of the epithelial cell populations showed a clear age dependency with, e.g., goblet cells decreasing and ciliated cells increasing with age (Fig. 1d, e) . A recent complementary study analyzed the cell composition of the nasal mucosa in healthy and SARS-CoV-2 infected children based on bulk RNA-Seq and cell deconvolution methods. They were unable to identify children-specific goblet cells, but rather described that samples from healthy children were dominated by a ciliated cell signature highlighting the limitations of bulk RNA approaches 10 .

The expression of the SARS-CoV-2 entry receptor ACE2 and the entry-associated proteases TMPRSS2, FURIN, CTSB, CTSL, and CTSV was similar between children and adults and not up-regulated by mild/moderate COVID-19 compared to the uninfected status (Extended Data Fig.2b) . Hence, these viral entry factors cannot explain the differences in SARS-CoV-2 pathophysiology between children and adults.

SARS-CoV-2 is a positive-strand RNA virus with a very high rate of replication 11, 12 . Hence, the control of SARS-CoV-2 infection requires an optimal early and coordinated innate antiviral immunity. This response is activated by various PRRs. Recently, mounting evidence has been generated in support of MDA5 (IFIH1) as the major PRR for SARS-CoV-2 in epithelial cells with RIG-I (DDX58) possibly playing an additional, but minor role 13, 14 (own unpublished data).

An important enhancer of viral RNA sensing by MDA5 is LGP2 (DHX58) 15 . Importantly, PRRs, in particular MDA5 and LGP2, are only weakly expressed in many epithelial cell types but are profoundly upregulated by positive feedback regulation upon viral infection of the cell or by paracrine exposure to type I or III interferon (IFN). The dynamics of this feedback regulation are crucial for the successful control of an infecting virus (Fig. 2a) . The importance of the PRR/IFN axis for the successful resolution of SARS-CoV-2 infection was recently . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint 5 demonstrated by clinical studies finding a strong association between genetic polymorphisms at various loci of the PRR/IFN system with an increased risk of severe COVID-19 16 . Similarly, and affecting even a much broader fraction of patients, autoantibodies directed against type I IFNs have been shown to occur at a remarkably high frequency in patients suffering from severe COVID-19 17 .

Notably, we found a significantly higher basal expression of the genes coding for RIG-I, MDA5

and LGP2 in epithelial cells in the upper respiratory tract of healthy children as compared to adults (Fig. 2b) . This result suggests an increased ability of the respiratory mucosa of children to respond to viral infections, which is further supported by the highly increased amounts of (day 0 -4, referred to as early phase) and sustained at lower levels in the later disease phase (days 5 to 12 post symptom onset, referred to as late phase). It can be assumed that higher basal expression of these PRRs would permit immediate sensing of SARS-CoV-2 by MDA5/LGP2 in infected epithelial cells (Fig. 2a) . Strikingly, children's airway epithelial cells displayed increased expression of these PRR genes compared to the expression level of these genes in epithelial cells in SARS-CoV-2 positive adults (Fig. 2b) , in particular in the early disease phase after symptom onset. From day 5 onwards, virus sensing is largely comparable between children and adults ( Fig.2b , lower plot).

Following virus sensing, signaling through IRF3/NFkB leads to the expression of primary antiviral effectors, as well as antiviral cytokines such as IFN-β and IFN-λ (Fig. 2a) . IFNs act on epithelial cells in an auto-and paracrine manner, further increasing MDA5/LGP2 responsiveness in the tissue and inducing a broad range of interferon-stimulated genes (ISG).

While we were not able to detect the expression of type I and type III interferons themselves, ISGs showed an impressive activation pattern in epithelial cells of SARS-CoV-2 positive children, including many genes previously shown to exhibit strong antiviral activity against SARS-CoV-2, such as LY6E 18 , IFITM2, and BST2 19 . In all epithelial cells, and in particular in . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. and monocyte-derived macrophages (moMa) as well as CD11c + dendritic cells (CD11c_mDCs) were most interactive (Fig. 3a) . These immune cell subsets showed a higher activation status in children as demonstrated by an increased expression level of several cytokine and chemokine-coding genes such as IL1B, IL8, TNF, CCL3 and CCL4 (Fig. 3b) . We furthermore observed an enhanced expression of IFIH1 in moMas, nrMas, and CD11c_mDCs in SARS-CoV-2 infected children and a significant increase of TLR2 in moMa, nrMa in the early phase of infection, suggesting that these cells might play an additional role in virus sensing and IFN production. This is further underlined by the fact that moMa, nrMa and CD11c_mDCs of SARS-CoV-2 negative children expressed IFIH1 and TLR2 at higher levels than adults.

Apart from the up-regulated cell-intrinsic antiviral capacity of airway epithelial cells, macrophages and DCs, we found specific patterns of immune cell subpopulations in children . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint vs. adults. Among others we identified a subpopulation of KLRC1 (NKG2A) + cytotoxic T cells (CTL2) occurring predominantly in children (Fig. 3c ). NKG2A is a lectin-like inhibitory receptor on cytotoxic T cells playing a role in limiting excessive activation, preventing apoptosis and sustaining the virus-specific CD8+ T cell response 21 Taken together our data provide clear evidence that the epithelial and immune cells of the upper airways (nose) of children are pre-activated and primed for virus-sensing. This is likely a general feature of the children's mucosal immune response, but of particular relevance for SARS-CoV-2. Very recently, scRNASeq of Chikungunya virus-infected fibroblasts showed an extremely narrow window of opportunity for the cells to express IFNs before viral protein production shuts the antiviral system off 22 . This likely also explains the differences between

CoV-1 in terms of the induced host response. SARS-CoV-2 is characterized by extensive intracellular replication and a remarkable absence of IFN-production and -secretion. On the other hand, SARS-CoV-2 is highly sensitive to treatment with IFNs prior to or after infection as shown in lung epithelial cells, even more so than SARS-CoV-1 20, 23 . Primed virus sensing and a pre-activated innate immune response in children leads to efficient early production of IFNs . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint 8 in the infected airways, likely mediating substantial antiviral effects mirroring those observed in vitro in IFN-(pre)treated cells. Ultimately, this may lead to reduced virus replication and faster clearance in children. In fact, several studies already showed that children are much quicker in eliminating SARS-CoV-2 compared to adults consistent with the concept that they shut down viral replication earlier 24-27 . For other respiratory viruses, such as RSV and IAV, that more efficiently induce an IFN response by themselves, a pre-activated innate immune response may be less relevant. The enhanced innate antiviral capacity in children together with the high IFN sensitivity of SARS-CoV-2 may explain why children are better able to control early-stage infection as compared to adults and therefore have a lower risk of developing severe COVID-19.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint

Individuals of three different cohorts were included in this study. Patients of the prospective observational cohort study Pa-COVID-19 28 and its study arm RECAST (Understanding the 

Sample processing, single-cell and library preparation, and data analysis were performed as documented previously 8, 9 . Briefly, fresh nasopharyngeal swabs were transferred into cold DMEM/F12 medium (Gibco, 11039) and within one hour processed further. Under biosafety S2 an equal volume of 13 mM DTT (AppliChem, A2948) was added to each sample. To achieve higher cell numbers, the solution was slowly pipetted up and down and the swab was dipped roughly 20 times into the medium. Following incubation at 37°C, 500 rpm for 10 minutes on a thermomixer, samples were centrifuged at 350xG at 4°C for 5 minutes and the supernatant slowly removed. If the pellet showed any sign of red blood cells (RBC), it was resuspended in 1x PBS (Sigma-Aldrich, D8537), treated with RBC Lysis Buffer (Roche, 11814389001) at 25°C for 10 minutes and centrifuged at 350xG at 4°C for 5 minutes. If samples were not processed immediately, the cell pellet was resuspended in DMEM/F12 supplemented with 20% FBS (Gibco, 10500) and 10% DMSO (Sigma-Aldrich, D8418) and frozen at -80°C. For the library preparations cells were thawed at 37°C, centrifuged at 350xG at 4°C for 5 minutes and further processed according to the protocol. To obtain a single cell suspension Accutase (Thermo Fisher, 00-4555-56) was added to the pellet and the solution incubated at room temperature for 10 minutes with carefully pipetting the cells after 5 minutes.

The incubation was stopped by adding DMEM/F12 supplemented with 10% FBS and centrifugation at 350xG at 4°C for 5 minutes. Subsequently, the supernatant was removed and the cell pellet was resuspended in 1x PBS (volume was adjusted to the size of the cell pellet).

The suspension was cleared of any cell debris using a 35 μm cell strainer (Falcon, 352235) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint epithelial 33-35 and immune 36-38 cell populations were assigned to cell types/ stages according to the expression level of different marker genes (Extended Data Fig. 1, Extended Data Fig.   4 ). The T cell and macrophage/DC clusters were sub-clustered and then further refined manually.

The object was stored as h5ad, converted to h5seurat using SeuratDisk version 0.0.0.9014 and imported back into R.

In total, 268,745 cells were included in the data set. Cell numbers between the groups were to its study group were compared using a Kruskal-Wallis test, which did not indicate significant differences between groups (p=0.2).

To enable visual comparisons between UMAPs of different groups, equal numbers of cell (45,000) per group were randomly sampled using SubsetData function in Seurat.

Putative cell-cell interactions were quantified using CellPhoneDB version 2.1.2 using default settings 39 . In order to reduce the influence of individual samples contributing a larger number of cells and to speed up computation, we capped the number of cells per sample at randomly sampled 2,000 cells. This was done using the SubsetData function in Seurat.

For the analysis of PRR/IFN responses, a gene set of the most prominent ISGs expressed by lung epithelial cells was assembled. As described previously 9 , we treated A549 epithelial cells with a mix of IFN-β and IFN-λ for 2, 8 or 24 h, and analyzed transcript levels by microarray analyses using the Illumina Human HT-12 Expression Beadchip platform at the genomic and proteomics core facility at DKFZ. We identified ISGs as exhibiting a log-2-fold-change > 0.8 at any time point, yielding 183 genes. We further included ISGs described to exhibit strong anti-. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint SARS-CoV-2 activity (65 top-scoring genes) if not already included in our list. This eventually yielded a gene set of 217 genes also expressed in our scRNA-Seq.

A549 cells stably transduced with a lentiviral vector expressing human IFIH1 under the control of the murine ROSA26 promoter (termed A549 MDA5 high in Fig. 2e) were kindly provided by Nadine Gillich and Ralf Bartenschlager. We transduced A549 (termed MDA5 low ) and A549 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint

To test whether elevated MDA5 levels in A549 cells significantly increased IFNB and ISG induction upon infection, an unpaired one-tailed Student-t-test of three biologically independent repetitions was performed (GraphPad Prism v9.1).

Due to potential risk of de-identification of pseudonymized RNA sequencing data the raw data will be available under controlled access in the EGA repository, [will be added upon manuscript acceptance]. Count and metadata tables (patient-ID, sex, age, cell type, QC metrics per cell)

can be found at FigShare: [will be added upon manuscript acceptance]. In addition, these data can be further visualized and analyzed in the Magellan COVID-19 data explorer at https://digital.bihealth.org [will be publicly available upon manuscript acceptance].

No custom code was generated/used during the current study.

We thank all patients of the RECAST, Pa-COVID-19, and SC2 studies for kindly donating nasal . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. IL6   IL8   IL18   TNF   IL1B   IFNAR1  IFNAR2  IFNGR1  IFNGR2  TLR1  TLR2  TLR3  TLR7  MYD88  STAT1  IRF7  DHX58  IFIH1  DDX58 IFNG expression CCL5 expression *** *** ***

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection in Children and Adolescents: A Systematic Review

On the effect of age on the transmission of SARS-CoV-2 in households, schools and the community

Age-specific mortality and immunity patterns of SARS-CoV-2

Type I and Type III Interferons -Induction, Signaling, Evasion, and Application to Combat COVID-19

Natural mucosal barriers and COVID-19 in children

The local and systemic response to SARS-CoV-2 infection in children and adults

SARS-CoV-2 Reverse Genetics Reveals a Variable Infection Gradient in the Respiratory Tract

COVID-19 severity correlates with airway epithelium-immune cell interactions identified by single-cell analysis

Hypertension delays viral clearance and exacerbates airway hyperinflammation in patients with COVID-19

Immune response to SARS-CoV-2 in the nasal mucosa in children and adults

Coronavirus biology and replication: implications for SARS-CoV-2

Virological assessment of hospitalized patients with COVID-2019

SARS-CoV-2 triggers an MDA-5-dependent interferon response which is unable to control replication in lung epithelial cells

MDA5 Governs the Innate Immune Response to SARS-CoV-2 in Lung Epithelial Cells

The innate immune sensor LGP2 activates antiviral signaling by regulating MDA5-RNA interaction and filament assembly

Inborn errors of type I IFN immunity in patients with life-threatening COVID-19

Auto-antibodies to type I IFNs can underlie adverse reactions to yellow fever live attenuated vaccine

LY6E impairs coronavirus fusion and confers immune control of viral disease

Functional landscape of SARS-CoV-2 cellular restriction

Imperfect innate immune antagonism renders SARS-CoV-2 vulnerable towards IFN-γ and -λ. bioRxiv

The Inhibitory Receptor NKG2A Sustains Virus-Specific CD8⁺ T Cells in Response to a Lethal Poxvirus Infection

Single-cell transcriptomic analysis of antiviral responses and viral Authors contributions

SvL recruited the patients, provided the human specimens, clinical data and annotation of the patients

RLCh performed scRNA experiments

VGM and MB performed and analyzed the in vitro experiments

IL wrote the manuscript, all authors read, revised, and approved the manuscript

late phase early phase negative negative

All other authors do not declare any competing interest.. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)The copyright holder for this preprint this version posted June 28, 2021. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 28, 2021. ; https://doi.org/10.1101/2021.06.24.21259087 doi: medRxiv preprint