Comparison of heat-inactivated and infectious SARS-CoV-2 across indoor surface materials shows comparable RT-qPCR viral signal intensity and persistence

Abstract

Environmental monitoring in public spaces can be used to identify surfaces contaminated by persons with COVID-19 and inform appropriate infection mitigation responses. Research groups have reported detection of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) on surfaces days or weeks after the virus has been deposited, making it difficult to estimate when an infected individual may have shed virus onto a SARS-CoV-2 positive surface, which in turn complicates the process of establishing effective quarantine measures. In this study, we determined that reverse transcription-quantitative polymerase chain reaction (RT-qPCR) detection of viral RNA from heat-inactivated particles experiences minimal decay over seven days of monitoring on eight out of nine surfaces tested. The properties of the studied surfaces result in RT-qPCR signatures that can be segregated into two material categories, rough and smooth, where smooth surfaces have a lower limit of detection. RT-qPCR signal intensity (average quantification cycle (Cq)) can be correlated to surface viral load using only one linear regression model per material category. The same experiment was performed with infectious viral particles on one surface from each category, with essentially identical results. The stability of RT-qPCR viral signal demonstrates the need to clean monitored surfaces after sampling to establish temporal resolution. Additionally, these findings can be used to minimize the number of materials and time points tested and allow for the use of heat-inactivated viral particles when optimizing environmental monitoring methods.

Figure 1:

Scatterplots showing the average Cq of RT-qPCR viral gene calls for corresponding heat-inactivated viral spike-in over seven days. Viral spike-in concentrations reported as GE’s from ddPCR. Linear regressions of average Cq on days since inoculation per spike-in were overlaid on the measured data. Average decay slope (m-bar) reported alongside each surface type.

Figure 2:

(A-C) 3D scatterplots showing distribution of average Cq of viral gene calls over seven days for nine different surfaces inoculated with 5×10⁵ GEs (nine surfaces for heat-inactivated virus [circles], two (acrylic and olefin carpet) for infectious [diamonds]). The distribution of Cq’s differs significantly across surface types (B), but not across days since inoculation (C). (D) Clustermap of the U statistic from pairwise Mann-Whitney U tests between surface types. (E-F) Standard curves relating surface viral load (spike-in) to average Cq across all time-points for smooth (E) and rough (F) surface types.