Capturing Coronaviruses: Comparison of Recovery Methods for Infectious Coronaviruses From Environmental Surfaces

Abstract

The emergence of novel, zoonotic betacoronavirus, SARS‐CoV‐2, demands quantification of infectious virus rather than viral RNA to provide more accurate assessment of transmission risk for COVID‐19. To maximize investigator safety, similar in morphology and inactivation profile, human betacoronavirus OC43 (OC43) was used as a surrogate to refine infectious virus recovery methods for SARS‐CoV‐2, a biosafety level (BSL)‐3 pathogen. Using OC43 and SARS‐CoV‐2, we examined the ability of InnovaPrep Mano Surface Samplers (MANOs), made of hydrophobic, open‐cell polymeric foam, to recover virus from large stainless‐steel surfaces hypothesizing that they would outperform hydrophilic cellulose sponges (sponges), used for environmental sampling.In triplicate, 1.0 x 105 TCID50 of OC43 and SARS‐CoV‐2 were applied to the inner surface of a biological safety cabinet (BSC) within the literature or vendor specified optimal surface area, 1267.36 cm2 (MANO) and 100 cm2 (sponges). The areas were sampled with eluant presoaked MANOs and sponges, and samples aliquoted and stored at ‐80ºC until batch titration by indirect immunofluorescence TCID50 assay, using human ileocecal colorectal adenocarcinoma cells (HRT‐18 CCL‐244, ATCC) for OC43 samples and crystal violet detection‐based TCID50 using African green monkey kidney epithelial cells (VeroE6 CRL‐1586, ATCC) for SARS‐CoV‐2. Tested eluants were a beef extract buffer, pH 7 (BEB7), BEB7 with 0.05% Tween‐20 (BEB7/T) and (MANOs only) phosphate buffered saline with 0.05% Tween‐20 (PBS/T), the vendor default. Additionally, an expanded surface area (0.77m2 for MANOs, 300cm2 for sponges) was tested with OC43 as described above, except with 1.0x104 TCID50 inoculum.For OC43, recovery for MANOs with BEB7 was 2.66x105 TCID50 compared to PBS/T’s 4.40x104 TCID50. BEB7 MANOs had a higher average percent recovery (102%) than BEB7 sponges (84%). By one‐way ANOVA (α=0.05), MANO BEB7’s and BEB7/T’s average recovery percentages were significantly greater than PBS/T’s. Samples taken from the larger surface areas for both samplers were below the assay’s limits of detection for BEB7; however, OC43 eluted with BEB7/T from sponges was detectable (3.47x102 TCID50, 32% recovery). For SARS‐CoV‐2, MANO BEB7’s average recovery percentage (27%) was again shown to be significantly greater than PBS/T’s (0.38%). However, decreased recovery of infectious virus was seen across both tools and all eluants, perhaps indicating a difference between SARS‐CoV‐2 and OC43 for BEB7.Use of BEB7 improved OC43 and SARS‐CoV‐2 recovery from optimal surface areas with MANOs. Surprisingly, when the surface area was enlarged, all samples except those from BEB7/T sponges were below the limit of detection. This might be explained by additional eluant loss on the larger surface during mechanical recovery or virus inactivation due to desiccation from BSC airflow or other environmental factors. Similar factors might explain the SARS‐CoV‐2 results. Quantification by RT‐qPCR or flow virometry, and planned experiments in BSL‐3 agricultural space, where the room is primary containment, may provide further insights.