Abstract
Mounting evidence indicates that SARS-CoV-2 can infect multiple systemic tissues, but few studies have evaluated SARS-CoV-2 RNA dynamics in multiple specimen types due to their reduced accessibility and diminished performance of RT-qPCR with non-respiratory specimens. Here, we employed an ultrasensitive CRISPR-RT-PCR assay to analyze longitudinal mucosal (nasal, buccal, pharyngeal, and rectal), plasma, and breath samples from SARS-CoV-2-infected non-human primates (NHPs) to detect dynamic changes in SARS-CoV-2 RNA level and distribution among these specimens. We observed that CRISPR-RT-PCR results consistently detected SARS-CoV-2 RNA in all sample types at most time points post-infection, and that SARS-CoV-2 infection dose and administration route did not markedly affect the CRISPR-RT-PCR signal detected in most specimen types. However, consistent RT-qPCR positive results were restricted to nasal, pharyngeal, and rectal swab samples, and tended to decrease earlier than CRISPR-RT-PCR results, reflecting lower assay sensitivity. SARS-CoV-2 RNA was detectable in both pulmonary and extrapulmonary specimens from early to late infection by CRISPR-RT-PCR, albeit with different abundance and kinetics, with SARS-CoV-2 RNA increases detected in plasma and rectal samples trailing those detected in upper respiratory tract samples. CRISPR-RT-PCR assays for SARS-CoV-2 RNA in non-respiratory specimens may thus permit direct diagnosis of suspected COVID-19 cases missed by RT-PCR, while tracking SARS-CoV-2 RNA in minimally invasive alternate specimens may better evaluate the progression and resolution of SARS-CoV-2 infections.