key: cord-0741304-74h61sa3
authors: Ong, Qunxiang; Ronnie Teo, J. W.; Cruz, Joshua Dela; Wee, Elijah; Wee, Winson; Han, Weiping
title: Irradiation of UVC LED at 277 nm inactivates coronaviruses by photodegradation of spike protein
date: 2021-05-31
journal: bioRxiv
DOI: 10.1101/2021.05.31.446403
sha: da095cd894d0eadc1029deeb3a474e584600ccd1
doc_id: 741304
cord_uid: 74h61sa3

To interrupt SARS-CoV-2 transmission chains, Ultraviolet-C (UVC) irradiation has emerged as a potential disinfection tool to aid in blocking the spread of coronaviruses. While conventional 254-nm UVC mercury lamps have been used for disinfection purposes, other UVC wavelengths have emerged as attractive alternatives but a direct comparison of these tools is lacking with the inherent mechanistic properties unclear. Our results using human coronaviruses, hCoV-229E and hCoV-OC43, have indicated that 277-nm UVC LED is most effective in viral inactivation, followed by 222-nm far UVC and 254-nm UVC mercury lamp. While UVC mercury lamp is more effective in degrading viral genomic content compared to 277-nm UVC LED, the latter results in a pronounced photo-degradation of spike proteins which potentially contributed to the higher efficacy of coronavirus inactivation. Hence, inactivation of coronaviruses by 277-nm UVC LED irradiation constitutes a more promising method for disinfection.

The novel coronavirus SARS-CoV-2 has precipitated into the COVID-19 pandemic, and at the 28 time of writing, resulted in more than 171 million infections and 3 million deaths. The actual numbers 29 should be much higher than reported, given the high incidence of asymptomatic cases escaping the Numerous studies have studied the sensitivity of different microbes to UVC wavelengths, including human 

(MW403500.1). We found that their overall base composition (Fig. 1C) and adjacent base arrangements 73 ( Fig. 1D) to be similar. We therefore hypothesize that the kinetics of UVC inactivation of hCoV-OC43 and 74 hCoV-229E to be roughly similar.

Inactivation of human coronaviruses after exposure to different UVC wavelengths

To examine the inactivation efficacy of UVC on hCoV-OC43 and hCoV-229E, virus was placed on plastic 77 petri dishes and expose to various UVC wavelengths of 73 µW/cm 2 for different timings ranging from 30 to 78 300 seconds (Fig. S1-S2) . The reduction in infectivity of hCoV-OC43 ( Fig. 2A, Table S1 ) and hCoV-229E 79 (Fig. 2B, Table S2 ) can be observed after exposure to different UVC irradiation. The 277-nm UVC LED 80 was most effective in carrying out the inactivation and achieved 3-log inactivation at 22 mJ/cm 2 for both 81 human coronavirus strains, whereas 254-nm UV lamp achieved only 2-log inactivation for hCoV-OC43 and 1-log inactivation for hCoV-229E with the same dosage. Two-way ANOVA analyses of both sets of data of UVC sources. 

To understand if the inactivation efficacy comes from RNA damage, we performed quantitative RT-PCR to 92 examine the copy number of hCoV-OC43 after UVC irradiation. We observed that the copy number of 93 hCoV-OC43 to be unperturbed after 300 seconds of 222-nm far UVC irradiation, while 254-nm UV lamp 94 exerted the largest decrease in copy number followed by 277-nm UVC LED (Fig. 3A) .

Photodegradation of hCoV-OC43 spike protein under 277-nm UVC LED

We next hypothesized that molecular components other than nucleic acids could be implicated, and the 97 spike protein is an especially attractive target to pursue given that it facilitates viral transmission by binding 98 to the host receptors(Shang et al., 2020). To this end, we subjected hCoV-OC43 to different duration of overall amount of protein loaded in each lane (Fig. 3B) . To further confirm that spike protein is indeed 102 degraded by 277-nm UVC LED, we performed UVC illumination on purified hCoV-OC43 spike proteins in vitro. Silver staining as depicted in Fig. 3C shows that hCoV-OC43 spike protein presents at a lower First, we exposed SARS-CoV-2 spike protein S1 subunit to varied durations of 254-nm UVC lamp and 277protein (Fig. 4A) . Absorbance spectroscopy (Fig. 4B ) analysis showed the UV absorbance profile of 254-112 nm lit proteins to be relatively unchanged while there is an increase of absorbance profile in the 250-300 113 nm region for 277-nm lit proteins. Western blot (Fig. 4C ) analysis further revealed a reduction in the SARSspike protein in the form of dimers and trimers could be seen in a dose-dependent manner.

In the search for potential mechanisms that drive the absorption of 277-nm wavelengths and degradation / 

In this study, we focused on human coronaviruses hCoV-OC43 and hCoV-229E, which belong to the genus SARS-CoV-2. On the other hand, hCoV-229E resembles the viruses that causes the common cold. The 

Due to the lack of access to BSL-3 facilities, we were unable to perform the viral infectivity tests on coronavirus hCoV-OC43, which provides a close approximation towards SARS-CoV-2 with the relevant 156 structures of the spike proteins being relatively similar. 

Images were acquired with a 40x objective, with the scale bars at 50 µm. 

(e) Western blot analysis of SARS-CoV-2 spike S1 RBD proteins revealed the differential rate of 205 aggregation and degradation amongst the different mutants.

(f) Quantification of relative monomer fraction for wild type and mutant RBD proteins. The fraction is 207 calculated as a function of (intensity at 35kDa) / (overall intensity across the whole lane).

(g) Absorbance spectra of the RBD proteins reveal little changes in 250-300 nm UVC absorbance for 209 W436R compared to wild type and Y453F mutant, indicating potentially that W436R spike protein 210 is less susceptible to 277nm UVC LED treatment.

Supporting Information. The following files are available free of charge.

215 Figure S1 . Wavelength spectrum of different UVC light sources 216 Figure S2 . Schematic diagram of UVC enclosure loci of cells. The cells are incubated at 37C in a 5% CO2 incubator for 3 days before the liquid overlay 338 medium is aspirated and fixation with 4% paraformaldehyde is performed at room temperature for an hour.

Staining with 0.5% crystal violet is then conducted and the plaques are then quantified.

Immunofluorescence experiments

To assess whether 5 minutes of various UVC illumination reduces the number of infected cells,

immunostaining was performed to detect the presence of OC43 viral particles in the host human cells. 

Protein structure studies

The Desktop PyMOL 2.4 was used to visualize the protein structure of SARS-CoV-2 S glycoprotein (Protein

Further evidence that far-UVC for disinfection is unlikely