Age-specific nasal epithelial responses to SARS-CoV-2 infection

The study addresses why age is the strongest risk factor for COVID-19 severity and mortality, contrasting mild disease in most children with high mortality in older adults. Nasal epithelial cells are primary targets for SARS-CoV-2 and can seed lower airway disease. Prior hypotheses focused on higher expression of viral entry factors ACE2 and TMPRSS2 in adults, but differences remain uncertain. Children may have a pre-activated antiviral state in the upper airway, yet this does not fully explain age-related risk. Most in vivo studies lack precise timing of early infection events. The research aims to define epithelial-intrinsic, age-specific cellular and molecular responses of human NECs to early SARS-CoV-2 infection, independent of immune cells, and to link these to viral replication, tropism and epithelial repair processes.

Background work established that SARS-CoV-2 entry factors are expressed in nasal epithelium and that nasal ciliated cells are major early targets. Reports suggested possible age differences in ACE2/TMPRSS2, and that children may exhibit pre-activated innate antiviral programs in the upper airway. However, early epithelial responses have been difficult to capture in vivo due to unknown infection timing and variable symptom onset. Basaloid-like epithelial states enriched in fibrosis have been described in lung fibrosis datasets, and impaired local intrinsic immunity is associated with severe COVID-19. The role of integrins such as ITGB6 in injury, fibrosis and potential viral interactions is relevant to epithelial repair and EMT during respiratory viral infections.

• Donors and ethics: Healthy paediatric (0–11 years), adult (30–50 years) and older adult (≥70 years) donors recruited in London hospitals (March 2020–Feb 2021); SARS-CoV-2 negative before sampling. Ethics REC 19/NW/0171.
• NEC culture: Primary nasal brushings expanded and differentiated at air–liquid interface (ALI) for 4 weeks to generate NECs.
• Infection: Cultures infected apically with early-lineage SARS-CoV-2 isolate hCoV-19/England/2/2020 at approximately MOI 0.01 (4×10^4 pfu/well; or 4.5×10^4 pfu/ml for 1 h) and analyzed up to 72 h post infection (peak replication at ~72 h). Matched mock controls included.
• Readouts: Single-cell RNA-seq (10x 5′ chemistry; 139,598 high-quality cells across ages, conditions and timepoints 4, 24, 72 h), with viral genome spike-in RT oligo in some samples to enhance viral read detection. Processing included SoupX ambient RNA removal, STARsolo, EmptyDrops, Scrublet, BBKNN batch correction, Leiden clustering, CellTypist annotation, pseudotime (Monocle3; Palantir), and Milo differential abundance testing. Viral reads mapped to concatenated human and multiple viral genomes including SARS-CoV-2.
• Viral kinetics and infectivity: Plaque assays (Vero E6) on combined apical fluid and cell lysates over time; immunofluorescence for dsRNA and spike; TEM for ultrastructure.
• Proteomics and cytokines: Mass spectrometry (UDMSE) of apical secretome and cell lysates; Western blots (e.g., vimentin, ACE2, TMPRSS2, E-cadherin); multiplex cytokine assays for IFN-α, IFN-γ, IFN-λ and others.
• Wound healing assays: Mechanical scratch of ALI cultures, time-lapse microscopy to quantify closure (% per hour), immunostaining for basaloid-like markers (KRT5, ITGB6, vimentin) and infection (dsRNA, spike); comparison of wounded vs non-wounded in mock and infected conditions.
• Viral genomics: Deep sequencing for coverage profiles, subgenomic RNA detection (Periscope; sgRPTL normalization), and intra-host mutation analysis across genome regions.
• In vivo validation: Integrated 8 published single-cell datasets (577k+ cells) from upper/lower airways of healthy and COVID-19 patients across ages using Harmony; identified goblet inflammatory and basaloid-like 2 cell states and compared frequencies with zero-inflated Poisson models.

• Cell composition and entry factors: Healthy paediatric cultures had more goblet cells (goblet 2), adults/older adults had higher basal/progenitor subtypes; total ACE2/TMPRSS2 protein levels did not differ by age, but paediatric goblet cells expressed higher TMPRSS2 and ACE2 mRNA, while adult/older adult expression enriched in secretory/basal 2 cells.
• Viral tropism and replication: At 24 h p.i., fewer cell types were infected in paediatric (3/24) vs adult (7/24) and older adult (11/24); by 72 h p.i. infection broadened in all. dsRNA spread at 72 h p.i. was higher in older adult cultures (mean 16.1% ± 9.5) vs paediatric (3.8% ± 3.1). Infectious titres at 72 h p.i. were >800-fold higher in older adults (mean 1.64×10^7 ± 3.94×10^7 pfu/well, n=8) than paediatric (1.71×10^1 ± 3.20×10^0 pfu/well, n=13), P=0.04.
• Interferon response in children: Infection induced a paediatric-specific goblet 2 inflammatory cell state with strong ISG expression (e.g., CXCL10, IFIT1, IFIT3) absent in mock and rare in older ages; these cells also expressed IFNL1/2 and IFNA1, suggesting autocrine IFN. Paediatric cultures had higher interferon protein (IFN-α/γ/λ) in apical supernatant at 72 h.
• Incomplete replication in paediatric goblet inflammatory cells: Viral reads in paediatric ciliated cells biased to 3′ (active replication), whereas goblet inflammatory cells showed 5′ bias (failed replication). Non-canonical sgRNAs (e.g., spike, ORF7a) were more abundant in paediatric/adult vs older adult samples (P=0.042). Paediatric cultures displayed more low-frequency and fixed mutations, particularly <16 kb (pre-RdRp), consistent with pressure and defective viral genomes.
• Older adult EMT/repair phenotype: Infection decreased culture thickness and TEER in adult/older adult, increased basal cell mobilization and epithelial protrusion/shedding. Older adult cultures showed expansion of basaloid-like 2 cells expressing ITGB6, VIM, KRT5 and elevated ITGB6/TMPRSS2 in the secretome; vimentin was upregulated (P<0.05). Vimentin cage-like structures and ITGB6 colocalized with spike in infected cells.
• Wound-repair links to infection: Basaloid-like 2 cells enriched for ECM organization, wound response and migration. Wound assays increased KRT5, VIM, and ITGB6 around wound sites (all P<0.002–0.001). Infected older adult cultures healed faster than paediatric (8.01% ± 2.67 vs 3.67% ± 2.78 per hour; P=0.01). Wounding increased dsRNA+ area (4.09% ± 3.61 to 9.69% ± 9.04; P=0.03) and boosted infectious titres in initially low-producer cultures (P=0.006).
• In vivo validation: Integration of patient datasets showed goblet inflammatory cells induced across ages, most robust in paediatric COVID-19; basaloid-like 2 cells enriched in COVID-19 and most abundant in older adults, and also increased in pulmonary fibrosis datasets.

The findings address age-related COVID-19 pathogenesis by revealing epithelial-intrinsic differences in early nasal responses. In children, infection elicits goblet 2 inflammatory cells with strong interferon responses that limit productive replication, generate sgRNAs and likely defective viral genomes, and reduce infectious virus output despite high viral RNA reads. In older adults, infection promotes epithelial damage, cell shedding, and a repair program with expansion of basaloid-like 2 cells expressing ITGB6 and EMT markers, which may facilitate viral spread and fibrogenic remodeling. Wound-associated repair processes correlate with greater susceptibility to infection and increased viral production, particularly in older adults. These age-specific cellular states and processes likely contribute to lower disease severity in children and enhanced pathogenesis in older adults. Integration with in vivo datasets supports the relevance of these states in patients.

This work demonstrates age-specific SARS-CoV-2 tropism and epithelial responses in human nasal epithelium: goblet cells are key early targets in children with a potent interferon response and incomplete viral replication, whereas secretory and basaloid-like repair-associated cells predominate in older adults, enabling higher infectious virus production and signs of EMT and fibrosis-related pathways. The in vitro NEC model recapitulates early in vivo epithelial responses and identifies basaloid-like 2 and goblet inflammatory states as age-associated features of COVID-19. Future research should dissect molecular drivers of the goblet inflammatory state and defective genome formation, define mechanisms by which ITGB6/integrins and EMT pathways enhance viral spread, evaluate therapeutic modulation of epithelial repair, and test whether targeting integrin signaling or enhancing mucosal IFN responses can mitigate disease in older adults.

• In vitro ALI NEC cultures lack immune cells and systemic factors; epithelial-intrinsic responses may differ in the full mucosal environment.
• Time course limited to early infection (up to 72 h p.i.); later dynamics and resolution were not assessed.
• Basaloid-like 2 cells appear less abundant in vivo than in vitro, possibly due to sampling sites, dissociation protocols, or technical differences.
• Donor numbers for scRNA-seq were modest (n=11 donors across age groups), though multiple timepoints and validations increased power.
• A single early-lineage SARS-CoV-2 isolate was used; variant-specific differences were not studied.
• Some viral read enhancement used a spike-in RT oligo; analyses controlled for potential biases, but residual technical effects cannot be entirely excluded.