Ferrets were experimentally inoculated with SARS-CoV-2 (severe acute respiratory syndrome (SARS)-related coronavirus 2) to assess infection dynamics and host response. During the resulting subclinical infection, viral RNA was monitored between 2 and 21 days post-inoculation (dpi), and reached a peak in the upper respiratory cavity between 4 and 6 dpi. Viral genomic sequence analysis in samples from three animals identified the Y453F nucleotide substitution relative to the inoculum. Viral RNA was also detected in environmental samples, specifically in swabs of ferret fur. Microscopy analysis revealed viral protein and RNA in upper respiratory tract tissues, notably in cells of the respiratory and olfactory mucosae of the nasal turbinates, including olfactory neuronal cells. Antibody responses to the spike and nucleoprotein were detected from 21 dpi, but virus-neutralizing activity was low. A second intranasal inoculation (re-exposure) of two ferrets after a 17-day interval did not produce re-initiation of viral RNA shedding, but did amplify the humoral response in one animal. Therefore, ferrets can be experimentally infected with SARS-CoV-2 to model human asymptomatic infection.

SARS-CoV-2 virus was first detected in late 2019 and circulated globally, causing COVID-19, which is characterised by sub-clinical to severe disease in humans. Here, we investigate the serological antibody responses to SARS-CoV-2 infection during acute and convalescent infection using a cohort of (i) COVID-19 patients admitted to hospital, (ii) healthy individuals who had experienced ‘COVID-19 like-illness’, and (iii) a cohort of healthy individuals prior to the emergence of SARS-CoV-2. We compare SARS-CoV-2 specific antibody detection rates from four different serological methods, virus neutralisation test (VNT), ID Screen^® SARS-CoV-2-N IgG ELISA, Whole Antigen ELISA, and lentivirus-based SARS-CoV-2 pseudotype virus neutralisation tests (pVNT). All methods were able to detect prior infection with COVID-19, albeit with different relative sensitivities. The VNT and SARS-CoV-2-N ELISA methods showed a strong correlation yet provided increased detection rates when used in combination. A pVNT correlated strongly with SARS-CoV-2 VNT and was able to effectively discriminate SARS-CoV-2 antibody positive and negative serum with the same efficiency as the VNT. Moreover, the pVNT was performed with the same level of discrimination across multiple separate institutions. Therefore, the pVNT is a sensitive, specific, and reproducible lower biosafety level alternative to VNT for detecting SARS-CoV-2 antibodies for diagnostic and research applications. Our data illustrate the potential utility of applying VNT or pVNT and ELISA antibody tests in parallel to enhance the sensitivity of exposure to infection.

This study addressed the need in Great Britain for supplementary blood tests for deer and pig herds under movement restrictions due to confirmed Mycobacterium bovis infection—to enhance the overall sensitivity and reliability of tuberculosis (TB) testing and contribute to an exit strategy for these herds. We evaluated four antibody tests (lateral flow DPP VetTB Assay for Cervids, M. bovis IDEXX ELISA, Enferplex Cervid and Porcine antibody tests and an in-house comparative PPD ELISA) using serum samples from defined cohorts of TB-infected and TB-free deer and pigs. TB-infected deer included two separate cohorts; farmed deer that had received a tuberculin skin test less than 30 days prior, and park deer that had received no prior skin test. In this way, we were able to assess the effect of the skin test anamnestic boost upon antibody test sensitivity. We tested a total of 402 TB-free pigs and 416 TB-free deer, 77 infected farmed deer and 105 infected park deer, and 29 infected pigs (including 2 wild boar). For deer, we found an equivalent high performance of all four tests: specificity range 98.8–99.5% and sensitivity range 76.6–85.7% for skin test-boosted infected deer, and 51.4–58.1% for non-boosted infected deer. These data suggest an overall approximate 25% increase in test sensitivity for infected deer following a skin test boost. For pigs, the tests again had equivalent high specificity of 99–99.5% and a sensitivity range of 62.1–86.2%, with substantial agreement for three of the four tests. Retrospective application of the ELISA tests to individual culled park deer and wild boar that showed no obvious evidence of TB at larder inspection identified a significant seropositivity within wild boar suggestive of low-level M. bovis infection that would otherwise not have been detected. Overall this investigation provided a robust evaluation of four antibody tests, which is essential to generate confidence in test performance before a wider deployment within TB control measures can be considered.