key: cord-0771396-iidzrw4i
authors: Hoffman, J. S.; Hirano, M.; Panpradist, N.; Breda, J.; Ruth, P.; Xu, Y.; Lester, J.; Nguyen, B.; Ceze, L.; Patel, S. N.
title: Passively Sensing SARS-CoV-2 RNA in Public Transit Buses
date: 2021-06-04
journal: nan
DOI: 10.1101/2021.06.02.21258184
sha: 7c35080eee2bf57be1b765dfd499defa7a4aff29
doc_id: 771396
cord_uid: iidzrw4i

Affordably tracking the transmission of respiratory infectious diseases in urban transport infrastructures can inform individuals about potential exposure to diseases and guide public policymakers to prepare timely responses based on geographical transmission in different areas in the city. Towards that end, we designed and tested a method to detect SARS-CoV-2 RNA in the air filters of public buses, revealing that air filters could be used as passive fabric sensors for the detection of viral presence. We placed and retrieved filters in the existing HVAC systems of public buses to test for the presence of trapped SARS-CoV-2 RNA using phenol-chloroform extraction and RT-qPCR. SARS-CoV-2 RNA was detected in 14% (5/37) of public bus filters tested in Seattle, Washington, from August 2020 to March 2021. These results indicate that this sensing system is feasible and that, if scaled, this method could provide a unique lens into the geographically relevant transmission of SARS-CoV-2 through public transit rider vectors, pooling samples of riders over time in a passive manner without installing any additional systems on transit vehicles.

The global pandemic of COVID-19 has exceeded 33 million reported cases in the US and 168 million confirmed cases worldwide as of May 28th, 2021. 1,2 The virus causing COVID-19, SARS-CoV-2, is primarily transmitted through airborne respiratory droplets via face-to-face contact 3-5 with asymptomatic or pre-symptomatic infected individuals. 6-11 Therefore, disease monitoring via viral presence testing is essential for managing potential outbreaks. Current disease monitoring is focused primarily on testing individual members of the population.

However, frequent widespread testing across the entire population can be cost-prohibitive in many communities, even with pooled testing. 12 While this resource intensive sampling strategy is useful for capturing the overall presence of a disease, alternative environmental sampling can serve as a warning sign of early-stage disease presence in a community prior to symptomatic patients testing positive. 13, 14 One example of passive viral sensing is testing for SARS-CoV-2 in community wastewater . CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint plants. 13, [15] [16] [17] [18] However, wastewater surveillance methods su↵er from a fixed, coarse granularity since sampling happens far downstream from the individual source. Leveraging multiple environmental sampling techniques through additional infrastructural media, such as public transit, can make viral monitoring more robust. Wastewater sampling has been explored for commercial aircraft and cruise ships, 19 but these approaches cannot be extended to public transport vehicles without wastewater management facilities. Public transit such as buses, light rails, and trains may be valuable targets for surveillance sampling, since they are linked to the population's geospatial mobility. The United Nations estimated >50% global population lives in urban centers, 20 such as Seattle, where nearly 50% of urban commuters use public transit. 21 Viral particles expelled from the respiratory system of an infected individual can circulate through the air into Heating, Ventilation, and Air Conditioning (HVAC) systems, and have been detected in air filters in hospitals treating infected individuals, 22-26 suggesting a similar approach for public transit. Prior work has examined risk of transmission for passengers on buses, trains, and airplanes at local, national, and international scale; 27-39 however, these reports have not leveraged public transportation for community monitoring. One potential reason for this is the cost and time associated with known sampling methods with adequate Limits of Detection (LOD) to sense the low number of copies of virus expected in filters without employing active systems of collection, such as environmental swabbing or vacuum-like Personal Environmental Monitor (PEM) equipment. 40 Rapid and inexpensive RNA extraction methods have detected 10-20 copies/reaction, which may be above the viral copies recovered from passive HVAC systems in non-concentrated settings outside of hospitals. 41 Additionally, virus particles can remain viable for 7 days on porous surfaces, like air filters, and 3 days on non-porous surfaces, like metal hand-grips. 42, 43 Therefore, air filters may accumulate and maintain virus over a longer time than swabbed surfaces, capturing data from more individuals with a single sample, enabling pooled testing.

Here, we explore the feasibility of passive surveillance sampling in public buses by in-3 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint stalling fabric sensors in vehicle air filtration systems. We demonstrate that sensitive methods of detection can be used to detect small copy numbers from samples collected from bus filters, using viral lysis, RNA extraction, and RNA detection via reverse transcription quantitative polymerase chain reaction (RT-qPCR) in a combination not proven in prior literature.

Although prior work has demonstrated species extraction from building air filters, 26,44 our method is a novel, passive strategy for local SARS-CoV-2 surveillance monitoring in urban transit, which shows high analytical sensitivity and specificity for low concentration environments. We evaluated this in-house method in samples collected from actively circulating buses to demonstrate the detection of SARS-CoV-2 RNA in real-world environments, and we present herein an analysis of how this method can relate to citywide cases for future disease monitoring use cases.

Between August 2020 and March 2021, environmental samples were collected from 15 actively deployed buses in the Seattle King County Metro fleet ( Figure 1A ). Bus selection was narrowed down to the main bus depot that serviced the Downtown Seattle area, which has the highest ridership. Individual buses were selected to be sampled via a convenience sampling approach based on which buses could be made available at the depot on a regular basis between 7:00-9:00 AM for sample retrieval. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint Figure 1 : Detection of the samples collected from the metro bus using our inhouse extraction protocol. (A) Workflow for passive sensing SARS-CoV-2 RNA including sample collection, sample transfer from papers or swabs, RNA extraction, and RT-qPCR for detection. (B) Sampling occurred via two methods in di↵erent areas of the bus. We collected supplementary pre-filters after more than 7 days of being installed inside the HVAC systems of actively-used metro buses (blue). We also swabbed commonly-touched surfaces on the bus (red). (C) Sample types and collection methods used during the course of the study. (D) Positivity rate breakdown by collection material and location. Sampling from both air filters as well as surfaces returned traces of SARS-CoV-2. Swabs from bus handholds made up the majority of SARS-CoV-2 detections with 42% positivity rate (13/31), while materials placed in air filters had the lowest positivity rate at 11% (5/45).

. CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint materials were also tested: PolyPro fabric, microporous paper separators, mixed cellulose ester filters, and EnviroMax Swabs. Swabbing was performed by running one swab across common hand-hold areas ( Figure 1B in red). When EnviroMax swabs were available, a second swab of the front face of the main bus filter ( Figure 1B in blue) was performed.

All bagged samples were placed in a plastic secondary container, which was wiped with bleach-based disinfectant, and transported to an approved lab facility. All procedures involving the untreated filter material were performed in a BSL2-certified Class II A2 biosafety cabinet. All types of filters that were used are shown in Figure 1C . Due to safety-related lab space and chemical SOP limits for phenol-chloroform isoamyl extraction, a maximum of n=6 buses (2 replicates for each of the two methods -filter and swab) could be tested in a single experiment.

Sample extraction for testing was performed within the same day of the sample collection from metro buses. Detection of SARS-CoV-2 RNA consisted of the following steps: viral extraction and lysis, RNA isolation via phenol-chloroform isoamyl extraction, and RNA detection via RT-qPCR ( Figure 1A ).

Filters collected from buses were cut into 2-cm 2 pieces. Two pieces, considered sample replicates, were placed into microcentrifuge tubes containing 200 µL lysis bu↵er (50mM EDTA pH 8.0, 250mM Tris-HCl pH 8.0, 50mM NaCl, 1% (w/v) SDS). 45 The tubes were placed on a foam tube rack attached to a vortexer and agitated for 15 minutes, at high speed, at room temperature. After vortexing, 600 µL TRIzol was added to each tube, pipette-mixed 10 times, and then the resultant 800 µL solution was transferred into a new tube. The solutions were incubated at room temperature for 5 minutes to allow complete dissociation of viral particles into the upper media and inactivation of any potentially remaining active virus in 6 . CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint the solution. The RNA was then isolated from protein and DNA following the standard TRIzol phase separation procedure. 46 Precipitation of RNA was carried out by adding 1 mL 200-proof ethanol and 1 µL RNA-grade glycogen (R0551, ThermoFisher) to each tube, followed by 1-minute vortexing. Each tube was then incubated overnight at 20 C. Following overnight precipitation, the supernatant was discarded and residual ethanol was allowed to evaporate. The RNA pellet was washed following the standard TRIzol RNA wash procedure and subsequently re-suspended in 8 µL nuclease-free water.

Detection of SARS-CoV-2 RNA Using RT-qPCR Each 8 µL TRIzol isolation product was assayed with TaqPath 1-step RT-qPCR (A15299, ThermoFisher Scientific) in 20 µL reactions. We used probes from the CDC SARS-CoV-2 qPCR probe assay targeting two regions in the N gene, designated N1 and N2 (10006713, Integrated DNA Technologies), one for each sample replicate. To avoid cross-contamination, the reactions were loaded into non-adjacent wells in a 96-well plate on ice at a separate bench from where the RNA isolation step was performed. Wells also were covered with Parafilm between loading samples. RT-qPCR was carried out on a Quantstudio 3 (ThermoFisher) using the CDC-recommended protocol. 47

Positive results were determined by amplification before a specified PCR cycle threshold (Table S2 ). This indicates that the viral particles may not be distributed evenly across the sampling material or that the sample methods may be sensing viral presence from di↵erent signal sources (i.e. riders who breathe may not touch the railing). Most positive samples had SARS-CoV-2 RNA near or below the LOD of the RT-qPCR assay ( Figure S6) and thus were confirmed correct product sizes by fragment analysis ( Figure S7 ).

We compared our TRIzol-based RNA extraction method with a more commonly used column-based RNA extraction from Qiagen. After dividing five filters in half and processing in parallel, we found a bus positivity rate of 60% (3/5) and 80% (4/5) (Table S3) in TRIzolbased and column-based methods, respectively. Interestingly, our TRIzol-based method did not yield positive results from any samples collected by EnviroMax swabs, which were positive with the column-based method. We observed black particle residues in samples using EnviroMax swabs ( Figure S8 ), which were filtered out by the column-based extraction. These residues ended up in the RT-qPCR reactions when EnviroMax samples were extracted by the TRIzol-based method, which may have inhibited the RT-qPCR reaction. On the other hand, the column-based method displayed 0% positivity rate on all air filter samples, while the Trizol-based method detected 40% (2/5) positivity on the same air filters. We hypothesize that the debris broken from filters could interfere with the binding of SARS-CoV-2 RNA to the silica membrane in the columns or the SARS-CoV-2 RNA might remain trapped in the columns. These filters are made from mixed cellulose ester which have di↵erent surface properties and porous structures from those of polyurethane foam structures (EnviroMax 8 . CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint Figure 2 : Top chart shows total riders per bus per 7-day period of filter installation before sampling. Each color denotes a unique bus that week. The color of associated circles denotes a positive result from that bus. An empty circle denotes a positive sample for a bus with 0 riders during that week. Bottom chart shows new cases of SARS-CoV-2 in King County (blue) superimposed with the proportion of buses sampled that week returning positive results (orange).

. CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint swabs) which reportedly had a high release e ciency even with minimal agitation. 48 All air filters were installed on buses for more than 7 days, and thus can represent pooled samples of all riders for the prior 7 days (Figure 2) . One exception is that, for one sampling date on October 14, 2021, filters were installed and collected in one day. In one-day testing, 0 filters returned positive, indicating that one day may not be enough filter exposure time to build up a detectable viral load. However, a relatively high rate (60%) of swab samples returned positive, which may be attributable to lack of surface decontamination mid-day.

Metro cleans buses nightly, and the morning sampling period for all 2-week samples occurred the morning following the decontamination, in between which no riders would have ridden the bus. 

. CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint

Here we show that passive infrastructure-mediated sensing of viral presence may be feasible.

By leveraging a passive viral sensing method such as ours in parallel with other environmental and individual testing methods, epidemiologists could inexpensively monitor a community to identify locales of transmission and estimate case numbers within local regions. Our method may be more valuable when cases within the general community are low, leveraging the fabric sensors to passively pool respiratory droplet samples from riders temporally over the filter installation period and spatially over bus routes. This study is limited by its sample size (n=39 total buses). Sample filters were placed and recovered manually by the research team and metro collaborators, which could be scaled by larger research teams (Table S4) . False negatives may have been caused by the limited coverage of the sample media over the air vent and air currents diverting around the sample filter media. In addition, mask mandates were in e↵ect for riders during the sample period, likely reducing the number of viral particles from infected riders landing on the filters. Considering these e↵ects, the small viral loads ( Figure S6 ) of some samples are unsurprising, but may fall below the typical LOD for many commercial PCR kits, including our chosen Taqpath 1-step kit. 47 Thus, false negatives may occur due to loss in the extraction and PCR steps. We also note that our method does not necessarily identify the risk to bus riders, but rather the presence of inactive SARS-CoV-2 RNA. Viral viability tests and more frequent sampling are needed to understand the risk to riders.

Future research into scalable, sensitive viral detection for environmental samples would enhance this approach. Studies evaluating filter placement and size, as well as control experiments in simulated environments, could further validate the sensitivity of the method. Citywide deployments enabled by scalable detection methods, such as rapid diagnostic lateral flow detection for on-site detection enabled by miniaturized PCR devices 41 or sequencers, 49 could gather more data, enabling network analysis techniques to study probability of SARS-CoV-2 transmission on a neighborhood level. This method could be adapted and deployed to pro-11 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

The authors declare no competing financial interest.

(1) Centers for Disease Control and Prevention, CDC Covid Data Tracker. 2020; https: //covid.cdc.gov/covid-data-tracker/.

(2) World Health Organization and others, Coronavirus disease (COVID-2019) situation reports. 2020; https://www.who.int/emergencies/diseases/ novel-coronavirus-2019/situation-reports/.

12 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 14 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint

Detection of SARS-CoV-2 RNA in Commercial Passenger Aircraft and Cruise Ship Wastewater: A Surveillance Tool for Assessing the Presence

Infected Travellers

How Urbanization A↵ects the Epidemiology of Emerging Infectious Diseases

Clinical Infectious Diseases: An O cial Publication of the Infectious Diseases Society of America

Air, Surface Environmental, and Personal Protective Equipment Contamination by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) From a Symptomatic Patient

Aerosol and Surface Distribution of Severe Acute Respiratory Syndrome Coronavirus 2 in Hospital Wards

Can Air-Conditioning Systems Contribute to the Spread of SARS/MERS/COVID-19 Infection? Insights from a Rapid Review of the Literature

Identification of SARS-CoV-2 RNA in Healthcare Heating, Ventilation, and Air Conditioning Units

The Risk of COVID-19 Transmission in Train Passengers: An Epidemiological and Modelling Study

Spatial Transmission of COVID-19 via Public and Private Transportation in China

The Association between Domestic Train Transportation and Novel Coronavirus (2019-nCoV) Outbreak in China from 2019 to 2020: A Data-Driven Correlational Report

The Roles of Transportation and Transportation Hubs in the Propagation of Influenza and Coronaviruses: A Systematic Review

Early Release -Asymptomatic Transmission of SARS-CoV-2 on Evacuation Flight

In-Flight Transmission Cluster of COVID-19: A Retrospective Case Series

Clusters of 2019 Coronavirus Disease (COVID-19) Cases in Chinese Tour Groups

An Outbreak of Covid-19 on an Aircraft Carrier

Longitudinal Monitoring of SARS-CoV-2 RNA on High-Touch Surfaces in a Community Setting

Tracing Surface and Airborne SARS-CoV-2 RNA inside Public Buses and Subway Trains

Simpler and faster Covid-19 17 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) testing: Strategies to streamline SARS-CoV-2 molecular assays

Stability of SARS-CoV-2 and other coronaviruses in the environment and on common touch surfaces and the influence of climatic conditions: A review

E↵ect of environmental conditions on SARS-CoV-2 stability in human nasal mucus and sputum

Diversity of DNA and RNA Viruses in Indoor Air As Assessed via Metagenomic Sequencing

Development of an E↵ective Method for Recovery of Viral Genomic RNA from Environmental Silty Sediments for Quantitative Molecular Detection

Purification of RNA using TRIzol (TRI reagent)

Prevention, CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time RT-PCR Diagnostic Panel

Swab sample transfer for point-of-care diagnostics: characterization of swab types and manual agitation methods

. CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. 

. CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

. CC-BY-NC-ND 4.0 International license It is made available under a 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 June 4, 2021. Figure S1 -5 Amplification curves of RT-qPCR from extraction controls, RNA standards, and bus samples.2. Figure S6 Threshold cycles (CT) of RT-qPCR from extraction controls, RNA standards, and bus samples.3. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)The copyright holder for this preprint this version posted June 4, 2021. ; https://doi.org/10.1101/2021.06.02.21258184 doi: medRxiv preprint