key: cord-0809630-ut6k2oja authors: Robinson, R. T.; Mahfooz, N.; Rosas-Mejia, O.; Liu, Y.; Hull, N. M. title: SARS-CoV-2 disinfection in aqueous solution by UV222 from a krypton chlorine excilamp date: 2021-02-23 journal: nan DOI: 10.1101/2021.02.19.21252101 sha: b264712b44928d0163ea98d5664cb05212e985e8 doc_id: 809630 cord_uid: ut6k2oja There is an urgent need for evidence-based development and implementation of engineering controls to reduce transmission of SARS-CoV-2, the etiological agent of COVID-19. Ultraviolet (UV) light can inactivate coronaviruses, but the practicality of UV light as an engineering control in public spaces is limited by the hazardous nature of conventional UV lamps, which are Mercury (Hg)-based and emit a peak wavelength (254 nm) that penetrates human skin and is carcinogenic. Recent advances in the development and production of Krypton Chlorine (KrCl) excimer lamps hold promise in this regard, as these emit a shorter peak wavelength (222 nm) and are recently being produced to filter out emission above 240 nm. However, the disinfection kinetics of KrCl UV excimer lamps against SARS-CoV-2 are unknown. Here we provide the first dose response report for SARS-CoV-2 exposed to a commercial filtered KrCl excimer light source emitting primarily 222 nm UV light (UV222), using multiple assays of SARS-CoV-2 viability. Plaque infectivity assays demonstrate the pseudo-first order rate constant of SARS-CoV-2 reduction of infectivity to host cells to be 0.64 cm2/mJ (R2 = 0.95), which equates to a D90 (dose for 1 log10 or 90% inactivation) of 1.6 mJ/cm2. Through RT-qPCR assays targeting the nucleocapsid (N) gene with a short (<100 bp) and long (~1000 bp) amplicon in samples immediately after UV222 exposure, the reduction of ability to amplify indicated an approximately 10% contribution of N gene damage to disinfection kinetics. Through ELISA assay targeting the N protein in samples immediately after UV222 exposure, we found no dose response of the ability to damage the N protein. In both qPCR assays and the ELISA assay of viral outgrowth supernatants collected 3 days after incubation of untreated and UV222 treated SARS-CoV-2, molecular damage rate constants were similar, but lower than disinfection rate constants. These data provide quantitative evidence for UV222 doses required to disinfect SARS-CoV-2 in aqueous solution that can be used to develop further understanding of disinfection in air, and to inform decisions about implementing UV222 for preventing transmission of COVID19. (mercury from breaking fragile quartz lamp bulbs is toxic 23 ), (2) the UV dose response 87 kinetics needed to inactivate SARS-CoV-2 are unknown. Should these two challenges 88 be overcome, the use of UV to inactivate SARS-CoV-2 in environments with high 89 potential for transmission (e.g. congregate care facilities, convalescent patient homes, 90 hospital waiting rooms, airplane cabins) would be a practical and readily deployed 91 engineering solution to augment current prophylactic measures (social distancing, face 92 masks, vaccines). Due to a surge in interest and application of UV in various public 93 settings, there is an urgent need to understand the dose response kinetics of SARS-94 CoV-2 to UV radiation to inform decisions which balance the risk to eyes and skin from 95 Krypton and chlorine in KrCl excilamps are much less toxic than mercury, and KrCl 106 excilamps have already been shown to be competitive in terms of electrical efficiency 107 with mercury lamps that have many more years of product development and 108 optimization 30 . Our results demonstrate that when an aqueous solution of pathogenic 109 SARS-CoV-2 is exposed to UV222 light emitted by a Kr-Cl excilamp, its infectivity and 110 integrity is attenuated in a UV dose-dependent manner, as measured by culture and 111 molecular assays. These first UV222 disinfection dose responses demonstrate the 112 METHODS diffusing cosine corrector detector. Raw spectral data from the OceanView software 142 was interpolated to integer wavelengths using the FORECAST function in Microsoft 143 Excel and relativized to peak emission at 222 nm for use in dose calculations (Figures 1 144 and S1). Total incident UV-C irradiance was measured using an International Light 145 Technologies (ILT) 2400 radiometer with a SED 220/U solar blind detector, W Quartz 146 wide eye diffuser for cosine correction, and peak irradiance response NIST-traceable 147 calibration. For irradiance measurement, the peak wavelength calibration value was 148 input manually as the radiometer factor. The incident irradiance was measured with the 149 detection plane of the radiometer centered at the height and location of the sample 150 surface during UV exposures, and corrected for several factors to determine the 151 average irradiance through the sample depth. Spatial nonuniformity of emission was 152 accounted for each test by measuring irradiance at 0.5 cm increments from the center 153 to the edge of the petri dish and relativized to determine a petri factor, which was 154 always > 0.9. The typical detector spectral response was obtained from ILT and used to 155 calculate the radiometer factor integrated over the lamp emission, which was 0.9971. 156 As previously 34 , the reflection factor for water at the 222 nm peak wavelength was 157 assumed to be 0.9726. The divergence factor was determined each experiment day by 158 accounting for the distance between the lamp and the sample surface, and the sample 159 depth and was always > 0.9. The water factor was determined each sample day by the 160 ratio between the incident irradiance and the average irradiance integrated through the 161 sample depth after wavelength-specific absorption. The UV-vis absorbance of virus 162 working stocks (prepared fresh for each test) was measured in the biosafety cabinet 163 using a Nanodrop TM One C spectrophotometer via the microvolume pedestal for 164 wavelengths 200 -295 nm and the 1 cm quartz cuvette for wavelengths above 195 nm. 165 Working stock absorbance spectra for each test are shown in Figures 1 and S1 . After 166 these adjustments to incident irradiance in the center of the sample, the average 167 irradiance was used to calculate exposure times (max: 15 minutes; min: 15 seconds) for 168 pre-determined UV doses (0-40 mJ/cm 2 ) (summarized in Supplementary Table S1). 169 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CoV-2 at ~10 5 PFU/mL in cDMEM was measured for each of three biologically 175 independent Tests for use in UV dose calculations. Expanded emission and 176 absorbance spectra from 200 -800 nm are shown in Supplementary Figure S1 . 177 All UV measurements, sample preparation, UV treatments, and subsequent handling of 179 treated samples were performed in a biosafety cabinet. On the day of each three 180 biologically independent tests while the UV source warmed up and measurements were 181 taken for dose calculations, aliquots of SARS-CoV-2 (previously tittered at 10 7 PFU/mL) 182 were diluted in cDMEM to make a "working stock solution" with a target titer of 10 5 183 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. corresponding to the pre-determined UV dose before replacing the aperture to end the 190 UV exposure. Immediately afterwards, the treated media was transferred to a sterile 15 191 mL polypropylene centrifuge tube (VWR) and used for the assays described below. 192 Working stocks for untreated samples were placed on the stir plate for a representative 193 amount of time with the lamp off before transfer to centrifuge tube (0 mJ/cm 2 ). 194 195 Plaque assays were used to determine PFU/mL of samples before UV treatment (0 196 mJ/cm 2 ) and after UV treatment (all other UV doses). The plaque assay used for this 197 study is a modification of that which was originally developed and reported by Case et 198 al, 31 and is listed here as STEPS 1-5. (STEP 1) At least 18 hours prior to the assay, 12-199 well plates were seeded with a sufficient number of Vero cells so that each well was 200 confluent by the assay start; plates were incubated overnight at 37°C. The virus outgrowth assay used for this study is identical to the plaque assay described 219 above, with the exception that after STEP 4 the virus laden media was replaced with 1 220 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Graphs were prepared using either GraphPad Prism or Microsoft Excel programs; 290 statistical analyses (including regression using the data analysis add-in to determine 291 standard error of regression coefficients) were performed using these programs' 292 bundled software. Log10 Reduction (LR) was calculated as log10(No/N), where N was 293 viral PFU/mL in the plaque assay, N gene copies/µL in qPCR assays for either the short 294 N1 amplicon or the long N1-2 amplicon, or N protein concentration in pg//mL in the 295 ELISA assay after exposure to a given UV222 dose, and No was the initial concentration. 296 The level of replication in this study was three biologically independent tests, with at 297 least technical duplicates for each assay. 298 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 300 Viral infectivity UV222 dose response was characterized by exponential decay kinetics 301 (Figure 2) . At a mean initial viral titer of 6.51x10 4 PFU/mL, the pseudo first order rate 302 constant for viral disinfection was -1.48 cm 2 /mJ (R 2 = 0.89). When expressed as LR of 303 viral infectivity after exposure to a given UV dose, the linear rate constant was 0.64 304 cm 2 /mJ (R 2 = 0.95), which equates to a D90 (dose for 1 log10 or 90% inactivation) = 1.6 305 mJ/cm 2 . Doses ranges and initial Vero cell confluence were only sufficient in the Test 3 306 experimental replicate to quantify a dose response. However, in Test 2, the mean initial 307 viral titer of 3.54x10 4 PFU/mL in untreated samples was reduced to below detection by 308 the first dose tested of 10 mJ/cm 2 , equivalent to a LR of at least 4.25 logs. These 309 results were also consistent with qualitative results from Test 1, where Vero cells 310 appeared mostly dead in the untreated samples, appeared increasingly healthy through 311 doses 0.7 and 1.4 mJ/cm 2 , and appeared healthy at doses above 2 mJ/cm 2 . 312 313 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21252101 doi: medRxiv preprint exponential decay kinetics ( Figure 3A ). When expressed as LR of N1 copies/µL in 329 qPCR reactions after exposure to a given UV dose, the linear rate constant was 0.069 + 330 0.005 cm 2 /mJ (slope + standard error, R 2 = 0.92). The N1 dose response was modeled 331 using the linear region between 0 -20 mJ/cm 2 to avoid tailing in the dose response. 332 When including only doses up to 10mJ/cm 2 as for the plaque assay, the slope and R 2 of 333 the N1 gene damage dose response was the same as for doses up to 20 mJ/cm 2 . 334 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21252101 doi: medRxiv preprint culture supernatants, and 0.8 + 1.4 copies/L in no template RT-qPCR reaction controls 360 (concentration data and standard curves shown in Supplementary Figures S2 and S3) . 361 Because the long amplicon assay was used to investigate potential for improved 362 measurement of disinfection dose response without culture, no Day 3 samples were 363 analyzed. 364 365 Although no dose response was observed for LR of the N protein versus UV222 dose 366 immediately after treatment for doses up to 40 mJ/cm 2 (0.002 + 0.001 cm 2 /mJ, slope + 367 standard error, R 2 = 0.21), a stronger dose response was observed in Day 3 cell culture 368 supernatants for doses up to 20 mJ/cm 2 (0.243 + 0.028 cm 2 /mJ, slope + standard error, 369 R 2 = 0.21) ( Figure 3C ). Across all tests, the positive signal for SARS-CoV-2 in the N 370 protein assay was 2.69x10 5 + 9. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. One additional limitation of this study related to UV222 application in indoor environments 408 is that the disinfection impact of any ozone production by vacuum UV wavelengths 409 potentially emitted by the KrCl excilamp was not measured, but can likely be neglected 410 due to high airflows in the biosafety cabinet and BSL3 facility. The negative air quality 411 impacts and building material degradation by ozone potentially generated by these 412 lamps, and the potential health hazards and building material solarization from 413 wavelengths below 240 nm and the nonzero emission at wavelengths above 240 nm 414 (Supplementary Figure S1) , should also be considered when weighing the benefits of 415 reducing infectious disease transmission by UV222 for COVID-19 and other infectious 416 diseases. 417 418 Considering these limitations, these data provide a strong foundation for future 419 development and application of UV222 for reducing airborne viral transmission. UV222 is 420 both 4.2 times safer for human exposure (the threshold limit values for human UV 421 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21252101 doi: medRxiv preprint exposure are 25 mJ/cm 2 and 6 mJ/cm 2 at 222 and 254 nm, respectively 41 ) and at least 422 1.3 times as effective at disinfecting SARS-CoV-2 (the D90 we observed for UV222 (1.6 423 mJ/cm 2 ) is lower than recently predicted by genomic modeling for UV254 (2.15 424 mJ/cm 2 ) 42 ). A recent study applying continuous UV222 at doses below these threshold 425 limit values to treat other airborne coronaviruses demonstrated multiple logs of 426 inactivation within minutes 43 . This low wavelength advantage for SARS-CoV-2 427 disinfection is consistent with a study where UV222 was more than twice as effective as 428 growth media, viruses were disinfected below detection in plaque assays, indicating that 448 aggregation did not interfere with complete viral inactivation. We did not observe a 449 strong relationship between the kinetics of N gene damage (measured by qPCR with a 450 short and long amplicon) and disinfection, which could reflect that protein damage 451 contributes more to disinfection than genome damage for SARS-CoV-2. One study of 452 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted February 23, 2021. ; https://doi.org/10.1101/2021.02.19.21252101 doi: medRxiv preprint Airborne Transmission of SARS-CoV-2: What We Know Randomized Trial of Convalescent Plasma in Covid-19 Severe Pneumonia Convalescent Plasma Antibody Levels and the Risk of Death from Covid-19 Antibody Cocktail, in Outpatients with Covid-19 SARS-CoV-2 Neutralizing Antibody LY-CoV555 in Outpatients with Covid-19 Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report Efficacy 533 of Tocilizumab in Patients Hospitalized with Covid-19 Mammalian Skin Safety of 222-Nm UV Light Evaluation of Acute Corneal Damage Induced by 222-Nm and 254-Nm Ultraviolet Action 580 Spectra for Validation of Pathogen Disinfection in Medium-Pressure Ultraviolet 581 (UV) Systems Comparison of UV-Induced Inactivation and RNA Damage in MS2 Wavelength-Dependent Molecular Indications of Protein Damage in 591 Adenoviruses after UV Disinfection Synergy of MS2 Disinfection by Sequential Exposure to 594 Detection, Quantification, and Inactivation of SARS-CoV-2. Virology Standardization of Methods for Fluence Estimating Effective Germicidal Dose from Medium Synergy of MS2 Disinfection by Sequential Exposure to 606 Real-time RT-PCR Primers and Probes for COVID-19 | CDC UV-Induced Inactivation Rates for Airborne 612 Effect of Ultraviolet Germicidal Irradiation on Viral Aerosols Effects of Relative Humidity on the 618 Ultraviolet Induced Inactivation of Airborne Bacteria Ultraviolet Germicidal Irradiation Handbook: UVGI for Air and 621 Surface Disinfection Inactivation of Virus-Containing Aerosols by Ultraviolet 624 A Critical Review on Ultraviolet Disinfection Systems 627 against COVID-19 Outbreak: Applicability, Validation, and Safety Considerations Genomic Modeling as an 630 Approach to Identify Surrogates for Use in Experimental Validation of SARS-CoV-631 2 and HuNoV Inactivation by UV-C Treatment Ultraviolet Irradiation Doses 637 for Coronavirus Inactivation -Review and Analysis of Coronavirus 638 Effectiveness of 222-Nm Ultraviolet Light on Disinfecting SARS-CoV-2 Surface Contamination Persistence of SARS-CoV-2 in Water and Wastewater Recovery of Enveloped Viruses in Untreated Municipal 650 Long-Range Quantitative PCR for Determining Inactivation of Adenovirus 2 by Ultraviolet Light Wavelength-Dependent Cell Entry 659 Mechanisms of SARS-CoV-2 Proteins, and Lipids with Free Chlorine and UV254 Site-Specific Cleavage in Viral Proteins Virus Disinfection Mechanisms: The Role of Virus 673 Composition, Structure, and Function. Current Opinion in Virology Virus 676 Inactivation Mechanisms: Impact of Disinfectants on Virus Function and Structural Robinson 1,2 , Najmus Mahfooz 1 Infectious Diseases Institute