Toward precision medicine: Inflammatory nasal epithelial transcriptomic profiles in long COVID.

Background: Little is known about the role of the nasal epithelium in long COVID (LC).

Objective: We sought to assess nasal epithelial transcriptomes of patients with LC to unravel pathophysiological mechanisms for disease management.

Methods: Medical data and transcriptomes were obtained from participants in the Precision Medicine for More Oxygen COVID-19 cohort at 3 to 6 months (n = 40) and at 12 to 18 months (n = 15) post-COVID. Cell-type frequencies were estimated by deconvolution from a single-cell data set. Hierarchical clustering identified transcriptomic clusters and cellular clusters from which differences in gene expression, gene set enrichment, and pulmonary phenotypes were assessed. Functional validation was performed using CRISPR/Cas9 gene editing and in vitro assays in primary mutant nasal epithelium, and gene expression comparisons were made with healthy controls (n = 51).

Results: At 3 to 6 and 12 to 18 months, transcriptomes associated with inflammatory pathways (P < .05). Transcriptomic and cellular clusters were identified and were related to inflammation and ciliogenesis (P < .05). Comparison of transcriptomes of patients with and without pulmonary radiological abnormalities resulted in 613 significant differentially expressed genes (P < .05). Upregulated inflammatory genes were observed in patients with abnormalities. SMURF1 expression was significantly increased in patients with abnormalities compared with those without abnormalities and healthy controls. SMURF1 mutant nasal epithelial cells produced significantly lower levels (P < .05) of proinflammatory cytokines on virus exposure compared with controls.

Conclusions: Nasal epithelium in LC exhibits persistent inflammatory states. SMURF1 upregulation potentially contributes to an exacerbated inflammatory state in nasal epithelium of patients with radiological abnormalities. This study demonstrates the importance of understanding these inflammatory profiles within a clinical context and emphasizes the need for further assessment and validation of SMURF1's role in LC.