Tonsils are major sites of persistence of SARS-CoV-2 in children

Abstract

In the present study, we show that SARS-CoV-2 can infect palatine tonsils, adenoids, and secretions in children without symptoms of COVID-19, with no history of recent upper airway infection. We studied 48 children undergoing tonsillectomy due to snoring/OSA or recurrent tonsillitis between October 2020 and September 2021. Nasal cytobrushes, nasal washes, and tonsillar tissue fragments obtained at surgery were tested by RT-qPCR, immunohistochemistry (IHC), flow cytometry, and neutralization assay. We detected the presence of SARS-CoV-2 in at least one specimen tested in 27% of patients. IHC revealed the presence of the viral nucleoprotein in epithelial surface and in lymphoid cells in both extrafollicular and follicular regions, in adenoids and palatine tonsils. Also, IHC for the SARS-CoV-2 non-structural protein NSP-16 indicated the presence of viral replication in 53.8% of the SARS-CoV-2-infected tissues. Flow cytometry showed that CD20⁺ B lymphocytes were the most infected phenotypes, followed by CD4+ lymphocytes and CD123 dendritic cells, CD8+ T lymphocytes, and CD14+ macrophages. Additionally, IF indicated that infected tonsillar tissues had increased expression of ACE2 and TMPRSS2. NGS sequencing demonstrated the presence of different SARS-CoV-2 variants in tonsils from different tissues. SARS-CoV-2 antigen detection was not restricted to tonsils but was also detected in nasal cells from the olfactory region. Palatine tonsils and adenoids are sites of prolonged RNA presence by SARS-CoV-2 in children, even without COVID-19 symptoms. IMPORTANCE This study shows that SRS-CoV-2 of different lineages can infect tonsils and adenoids in one quarter of children undergoing tonsillectomy. These findings bring advancement to the area of SARS-CoV-2 pathogenesis, by showing that tonsils may be sites of prolonged infection, even without evidence of recent COVID-19 symptoms. SARS-CoV-2 infection of B and T lymphocytes, macrophages, and dendritic cells may interfere with the mounting of immune responses in these secondary lymphoid organs. Moreover, the shedding of SARS-CoV-2 RNA in respiratory secretions from silently infected children raises concern about possible diagnostic confusion in the presence of symptoms of acute respiratory infections caused by other etiologies.

Keywords: COVID-19; adenoid; children; otorhinolaryngology; palatine tonsil; pediatric infectious disease; persistent infection; respiratory viruses; viral infection; virus persistence.

Fig 1

SARS-CoV-2 viral loads and antigen detection. (A) Quantification of SARS-CoV-2 RNA genome copies in palatine tonsil and adenoid tissues, nasal washes, and nasal cytological brushes from patients with chronic adenotonsillar disease (mean ± standard deviation). (B) Representative section of an adenoid with positive staining for SARS-CoV-2 NP in the pseudo-stratified ciliated epithelium. (b’) Higher magnification of the inset (dashed square) in panel B. (C) Representative section of a palatine tonsil with positive staining for SARS-CoV-2 NP in the inter-follicular area. (c’) Higher magnification of the inset (dashed square) in panel C. (D) Representative section of palatine tonsil with positive staining for SARS-CoV-2 NP in follicular lymphoid cells. (d’). Higher magnification of the inset (dashed square) in panel D. (E) Representative section of an adenoid with positive staining for SARS-CoV-2 NSP-16 in cells in the epithelial region. (e’). Higher magnification of the inset (dashed square) in panel E. (F) Representative section of a palatine tonsil with positive staining for SARS-CoV-2 NSP-16 in cells in the inter-follicular area. (f’). Higher magnification of the inset (dashed square) in panel F. (G) Fluorescent staining of palatine tonsil section for ACE2 (green), TMPRSS2 (red), and the S protein of SARS-CoV-2 (gray). (H) Mean fluorescence intensity for ACE2 and TMPRSS2 in SARS-CoV-2-infected and control tonsils. Scale bar: 100 µm.

Fig 2

Immune phenotyping of SARS-CoV-2 NP-positive tonsillar cells by flow cytometry. (A) Representative gating illustrating the infected population in SARS-CoV-2-negative (Control) and SARS-CoV-2-positive (Patient) tonsils. (B) Frequencies of infected immune cells in adenoids and palatine tonsils positive for SARS CoV-2 NP (±SEM). Each circle represents one subject donor of tonsillar tissue.

Fig 3

Immunofluorescence for SARS-CoV-2 in cytobrush preparations. Representative fields of cytobrush preparations from SARS-CoV-2-negative (A–C) and SARS-CoV-2-positive children (D–G), showing positivity for NP protein in some cells. The inset on F is enlarged on G. Scale bar: 25 µm.

Fig 4

Schematic representation of the evolution of SARS-CoV-2 clades. Each clade is represented by a color circle, and the samples sequenced in the present cohort with the mutational patterns represented by gray rectangles containing the following information, from top to bottom: sample ID, the clade, and the Pangolin lineage assigned by NextClade. The solid rectangles represent SARS-CoV-2 genomes assembled with higher coverage, and the dashed ones represent those with lower coverage but carrying enough defining mutations to enable lineage assignment (AD, adenoid; PT, palatine tonsil; NC, nasal cytobrush).