The viability of SARS-CoV-2 on sports equipment is limited and depends on the composition of the materials
The sporting goods were selected from the Sports Technology Institute, Loughborough University collection of goods (Table 1). These included commonly handled objects from popular sports, such as rugby balls, soccer balls and tennis balls. All materials were unused prior to testing. Materials frequently shared among participants were prioritized due to greater potential as transmission vectors. The materials were cut into discs 2 cm in diameter using a metal punch. The red cricket ball was not cut into discs, but used whole, due to difficulties in maintaining surface integrity when removed. Materials were not sterilized prior to inoculation, so as not to affect surface coatings, and antibiotics in the media were used to avoid contamination during cell culture. Steel discs were purchased (Lasermaster, UK) for use as a control surface.
Materials were inoculated with a 40 μl droplet of Dulebecco’s Modified Eagle’s Medium (DMEM) containing a high concentration (5.4 × 104 plaque forming units (PFU)) or low (5.4 × 102 PFU) live SARS-CoV-2 virus concentration quantified (isolate REMRQ0001/Human/2020/Liverpool). Inoculum concentrations were chosen to represent the upper and lower quartile of viral loads in symptomatic patients15. All work with live viruses took place under BSL3 conditions in a Class 2 biological safety cabinet. Materials were inoculated on the outward-facing surface, and precautions were taken to ensure that the inoculum did not flow out of the material during inoculation. Triplicate pieces of each material were inoculated at the following time points: 1, 5, 15, 30, 90 min. At each time point, the materials were swabbed using dry cotton swabs (Copan, Italy) and added to 400 μl of DMEM. A standardized swab technique was used for each sample to reduce variation, with the swab pulled up for two seconds and sideways for two seconds. The tubes containing the swabs were vortexed for 5 s, then serially diluted 10-fold for three dilutions, in DMEM. During the study period, the laboratory temperature and humidity were 22.1°C ± 1.6 and 52% relative humidity ± 3.8%.
Cultures of VERO E6 cells (C1008; African green monkey kidney cells, European Collection of Authenticated Cell Cultures 85,020,206) were maintained in T75 cell culture flasks (Corning, USA) in modified Eagle’s medium from Dulbecco (DMEM) supplemented with 4.5 g/L of glucose and I-Glutamine (Lonza, US), 10% fetal bovine serum (Sigma, US) and 50 units per ml penicillin/streptomycin (Gibco, US), at 37.5°C + 5% CO2. Cells for plaque assays were detached from the monolayer using 2 mL of 1 × Trypsin-EDTA (Sigma, USA) and 500 μL were seeded into 24-well microtiter plates (Corning, USA) at a density of 250,000 cells/ml. Plates were incubated for 24 h at 37.5°C + 5% CO2 and used for downstream plaque assays if judged to be >95% confluent by microscopy.
Viral plaque tests
Viral plaque assays were performed using VERO E6 cells in 24-well microtiter plates. Medium was aspirated from the microtiter plates and 40 µl of medium from the swabs and further dilutions were added to the wells in triplicate with 160 µl of DMEM 2% FBS. Microtiter plates were incubated for 1 hour at 37.5°C + 5% CO2 to ensure viral infection. The plates were then removed from the incubator and covered with a 1.1% cellulose suspension (Sigma, UK) in DMEM 2% FBS, and incubated again under the same conditions for 72 h. The plates were then removed from incubation, fixed with 100% formaldehyde for one hour and stained with 1 ml/well of 0.25% crystal violet solution. After staining for 1 min, plates were washed gently with water, then air-dried for >3 h. Viral plaques were then visually counted for each well. The total amount of PFU extracted from the swab was then recalculated from the number of viral plaques from the 40 µl of medium tested. A schematic of the methods is shown in Fig. S1 (produced using BioRender, https://biorender.com/).
All individual plaque counts for each time point material/swab were used to analyze the viral recovery of each material. Readings below the limit of quantification (BLQ) were assumed equal to BLQ/2 levels. To characterize SARS-CoV-2 retained on different materials, a dynamic approach was used to measure the temporal evolution of the virus. This was achieved by estimating the viral decay half-life using linear models to record the transformed PFU data on each material. The model assumed a single-phase decay profile on all materials over time. Each material was assumed to have a different intercept and slope, as the materials varied greatly in the observed initial levels of virus in the first minute.
The linear fit was achieved using the I am function in R. After generating a slope and an intercept for each material, simulations were performed based on the estimates of each of these parameters and their covariance using R. 500 simulations were performed for each material and their 5 to 95 percentiles were plotted against the observed value. The data. To estimate overall virus exposure over time for each material, the area under the curve (AUC) of the simulated time-virus profile was estimated for each of the 500 simulations and compared as the primary metric describing exposure. aggregate of each material to the virus over time. The simulated AUCs for each material were compared statistically using nonparametric pairwise comparisons using Wilcoxon’s rank sum test. Materials were ranked by AUC versus control material on all graphs reported. Simulated half-life and AUC data are presented as box and whisker plots displaying the 5 to 95 percentiles of the 500 simulated profiles for each material.
Statement of Patient and Public Involvement
As this was a laboratory study only, we did not involve any patients or members of the public in the design of the study.