It's the Wrong Virus: Rapid Adaptation of Operations Inside a Highly Pathogenic Avian Influenza Select Agent Laboratory in Response to the 2019 SARS-CoV-2 Pandemic

Abstract

Background:

The Animal Biosafety Level 3 Enhanced (ABSL-3+) laboratory at St. Jude Children's Research Hospital has a long history of influenza pandemic preparedness. The emergence of SARS-CoV-2 and subsequent expansion into a pandemic has put new and unanticipated demands on laboratory operations since April 2020. Administrative changes, investigative methods requiring increased demand for inactivation and validation of sample removal, and the adoption of a new animal model into the space required all arms of our Biorisk Management System (BMS) to respond with speed and innovation.

Results:

In this report, we describe the outcomes of three major operational changes that were implemented to adapt the ABSL-3+ select agent space into a multipathogen laboratory. First were administrative controls that were revised and developed with new Institutional Biosafety Committee protocols, laboratory space segregation, training of staff, and occupational health changes for potential exposure to SARS-CoV-2 inside the laboratory. Second were extensive inactivation and validation experiments performed for both highly pathogenic avian influenza and SARS-CoV-2 to meet the demands for sample removal to a lower biosafety level. Third was the establishment of a new caging system to house Syrian Golden hamsters for SARS-CoV-2 risk assessment modeling.

Summary:

The demands placed on biocontainment laboratories for response to SARS-CoV-2 has highlighted the importance of a robust BMS. In a relatively short time, the ABSL-3+ was able to adapt from a single select agent space to a multipathogen laboratory and expand our pandemic response capacity.

Introduction

Before SARS-CoV-2, influenza was the virus most associated with pandemics. The 2009 H1N1 pandemic virus was the first pandemic of the 21st century and killed upward of 575,400 people worldwide.¹ The Animal Biosafety Level 3 Enhanced (ABSL-3+) laboratory at St. Jude Children's Research Hospital has a 40-year history facilitating influenza research, including that targeting highly pathogenic avian influenza (HPAI).

The laboratory is part of the Centers of Excellence for Influenza Research and Response (CEIRR) network established by the National Institute of Allergy and Infectious Diseases to study the natural history, transmission, and pathogenesis of influenza and participates in an international research infrastructure to address influenza outbreaks. The laboratory is regulated by the Division of Agricultural Select Agents and Toxins at the U.S. Department of Agriculture's Animal and Plant Health Inspection Service, as governed by the U.S. Federal Select Agent Program (FSAP) regulations (7 Code of Federal Regulations [CFR] Part 331, 9 CFR Part 121.3, 42 CFR Part 73.3).

Before 2020, the focus of the ABSL-3+ facility was exclusively on influenza viruses. The laboratory has 10,000 square feet of space, including six animal rooms containing flexible film isolators (FFIs; primary containment devices used to house open caging), and is equipped to receive emerging influenza strains and perform risk assessment in both in vitro and in vivo models. The laboratory has thrived with a comprehensive biorisk management system (BMS) for HPAI that has been developed over the years in nonpandemic times. As SARS-CoV-2 emerged and rapidly spread into a pandemic, the demand for virus detection, isolation and characterization, and risk assessment in animal models extended to many facilities.

Thus, the release of the U.S. Centers for Disease Control and Prevention guidelines for SARS-CoV-2 that stipulated virus isolation, virus characterization in cell culture, and the manipulation of infectious virus require biosafety level 3 (BSL-3) containment together with site- and activity-specific risk assessment² meant the ABSL-3+ could accommodate the introduction of a novel pathogen without changing the BMS significantly. The facility was already designed with individual room high efficiency particulate air filtration and isolators for use with multiple influenza strains; therefore, engineering controls did not need to be changed to incorporate another pathogen.

The purpose of this report is to outline the biosafety framework and describe the response our laboratory provided to adapt the space to both pathogens. In this study, we describe the three major modifications that were made to accommodate SARS-CoV-2 in our space: administrative changes, including Institutional Biosafety Committee (IBC) protocols, training of staff, physical space constraints, and occupational health additions; inactivation and validation data for both pathogens; and the establishment of a risk-assessment model in hamsters using modified caging inside FFIs.

Results

Administrative Changes

Work began with the creation of four new IBC protocols for working with SARS-CoV-2, each focused on distinct studies. Once the protocols were in place, our Environmental Health and Safety (EH&S) staff provided agent-specific training to >40 individuals about the hazards and risk mitigation strategies when working with SARS-CoV-2. The staff already working with HPAI in the ABSL-3+ were instructed to perform all procedures in the same way as HPAI, with two exceptions: First, the three procedure rooms used for in vitro work were segregated into either a SARS room or an HPAI room.

Equipment and consumables were moved and often duplicated to ensure each room could operate independently; second, an activation/deactivation checklist was developed to change an animal room from select agent space to nonselect agent space. The checklist contained five categories for operations: (1) Communication to the Alternate Responsible Official and Responsible Official (RO) of the intent to change, (2) experiment review by EH&S, (3) activation of the specific room for select agent/nonselect agent work, (4) termination/extension of experiments, and (5) deactivation of the specific room for select agent/nonselect agent work.

After RO review, the room was released to the investigator for use. Importantly, staff never moved between rooms active with either HPAI or SARS-CoV-2 without a full body shower and complete change of personal protective equipment. In addition, no non-security risk assessment cleared staff were allowed to work in the facility. In addition to the current influenza testing and monitoring program, the Occupational Health staff performed COVID-19 testing for any individuals working in the ABSL-3+ and full contact tracing in the event they developed symptoms after working with SARS-CoV-2.

In addition, upon the availability of the Pfizer (BioNTech), Moderna, and Johnson and Johnson (Janssen) COVID-19 vaccines, the IBC quickly instituted a vaccine mandate for laboratory workers handling SARS-CoV-2. Every step of the administrative response to the addition of SARS-CoV-2 into select agent space was reviewed and approved by the St. Jude IBC (Figure 1).

Figure 1.

Graphic depiction of administrative changes that occurred to adapt the ABSL-3+ to incorporate SARS-CoV-2. Several parts of the BMS (Environmental Health and Safety, Research Staff, Animal Core, Occupational Health) had diverse final goals, yet the systems were all funneled through the IBC for consistent oversight. ABSL-3+, Animal Biosafety Level 3 Enhanced; BMS, Biorisk Management System; IBC, Institutional Biosafety Committee; NSA, non select agent; SA, select agent.

SARS-CoV-2 Inactivation Conditions and Testing

The primary driver for validated methods of sample inactivation in biocontainment laboratories is the investigative need of the users. The initial studies working with SARS-CoV-2 would involve whole genome sequencing of clinical samples and viral isolates, high-throughput wild-type virus microneutralization assays for the testing of serum, and risk-assessment studies in a Syrian golden hamster model using modified caging for aerosol transmission. All of these would require movement of samples outside of the ABSL-3+; however, there was not complete overlap with the downstream needs for HPAI samples.

To comply with FSAP regulations, inactivation and validation data were required for the select agent (HPAI), even if the procedure was only going to be used to inactivate SARS-CoV-2. Although some validation procedures for HPAI were performed before the introduction of SARS-CoV-2, the rationale for this was to ensure the inactivation procedure would be validated for the select agent in the unlikely event of cross-contamination. Therefore, out of an abundance of caution, inactivation methods were validated for both HPAI and SARS-CoV-2 (for detailed methods of buffer preparation, sample processing, and readout, see Supplementary Data).

Inactivation can be achieved by chemical or physical procedures, many of which are well established.^3–6 We validated inactivation of HPAI and SARS-CoV-2 virus stocks by QIAGEN RLT, TRIzol^® LS, formalin/formaldehyde, paraformaldehyde (PFA), and heat. We used the highly pathogenic A/Vietnam/1203/2004 (H5N1) influenza virus and 2019-nCoV/USA-WA1/2020 (SARS-CoV-2) virus because they can be grown to high viral titers, as needed for this analysis (Figure 2). We chose to use one strain of each virus because FSAP regulations state that the validation of inactivation procedures only needs to be performed using a single member of any given family of viruses.

Figure 2.

Starting viral titers of samples used for the validation of inactivation procedures. (A) Amount of HPAI (A/Vietnam/1203/2004; H5N1) and SARS-CoV-2 (2019-nCoV/USA-WA1/2020) before and after purification on Amicon columns. Viral endpoint titers for the untreated or Amicon-treated samples were determined by EID₅₀ for HPAI or TCID₅₀ for SARS-CoV-2. Data are expressed as Log₁₀ EID₅₀/mL or Log₁₀ TCID₅₀/mL (SARS-CoV-2). (B) Amount of virus measured in HPAI-infected animal tissues before inactivation procedures. Data are expressed as mean EID₅₀/mL ± standard deviation. Numbers above the error bars represent the mean. EID₅₀, 50% egg infectious dose; HPAI, highly pathogenic avian influenza; TCID₅₀, 50% Tissue Culture Infectious Dose.

Removing the toxic effects of QIAGEN RLT, TRIzol LS, and formalin/formaldehyde/PFA from virus stock cultures and infected cell monolayers was done by buffer exchange using 10 kDa Amicon purification columns. To demonstrate no significant loss of virus using these columns, we used untreated infected samples to measure virus titers after three phosphate buffered saline changes (Figure 2A). There was no loss of titer with HPAI, and less than one log reduction with SARS-CoV-2. Virus was also measured in tissue samples from HPAI-infected animals to determine the relative titers in each tissue that was to be inactivated before the addition of 10% neutral buffered formalin (NBF) (Figure 2B).

All tissues tested from infected chickens had high viral titers (>6.5 log₁₀ EID₅₀). Complete inactivation was achieved with both QIAGEN RLT and TRIzol LS nucleic acid extraction buffers, 4% formaldehyde, 4% PFA, 10% NBF, and heat (100°C) (Table 1), demonstrating greater than six log reductions in viral titer. Importantly, we were not able to detect live virus in these treated samples, even after two serial passages in eggs and long incubation times in cell culture (14 days).

Table 1.

Inactivation of highly pathogenic avian influenza and SARS-CoV-2 samples after treatment with lysis buffers, formalin fixation, and heat

Virus	Treatment	Sample typea
		Virus stock		Cell monolayer		Animal tissue
		Untreated control	Treated	Untreated control	Treated	Untreated control	Treated
HPAI	QIAGEN RLTb	2/2	0/5	2/2	0/5	2/2	0/5
	TRIzol® LSb	2/2	0/5	2/2	0/5	2/2	0/5
	Formaldehyde/formalin/PFAc	ND	ND	2/2	0/5	2/2	0/5
	Heat (100°C)d	2/2	0/5	2/2	0/5	ND	ND
SARS-CoV-2	QIAGEN RLT	2/2	0/5	2/2	0/5	ND	ND
	TRIzol®LS	2/2	0/5	2/2	0/5	ND	ND
	Formaldehyde/PFA	ND	ND	2/2	0/5	ND	ND
	Heat (100°C)	2/2	0/5	2/2	0/5	ND	ND

^a Duplicate viral stock cultures, infected cell monolayers, or infected 5 mm pieces of homogenized animal tissue. For HPAI, inactivated samples were injected into embryonated chicken eggs one time, then pooled allantoic fluid was injected into a second set of five eggs for a blind passage. Data are displayed as the number of hemagglutinin positive eggs/number of eggs injected. For SARS-CoV-2, samples that had been inactivated were used to infect Vero E6 TMPRSS2 cells and CPE was monitored for 14 days. Data are displayed as the number of cultures positive for CPE/number of cultures infected for SARS-CoV-2. Control samples were treated with PBS then treated the same way as the inactivated samples before testing.

^b Samples were combined with QIAGEN RLT or TRIzol^® LS according to the manufacturer's instructions, and the volume was purified on Amicon columns to remove toxicity before testing.

^c Infected cell monolayers were fixed with 4% formaldehyde or 4% PFA for 30 min, and the volume was purified on Amicon columns to remove toxicity before testing. Animal tissues were fixed with 10% neutral buffered formalin for 24 h or greater, then perfused through PBS for 24 h to remove toxicity before testing.

^d Heat treatment of samples was at ≥100°C for 10 min.

CPE, cytopathic effect; HPAI, highly pathogenic avian influenza; ND, not done; PBS, phosphate buffered saline; PFA, paraformaldehyde.

Implementing a Viable Animal Model for SARS-CoV-2 Studies in an Established ABSL-3+ Animal Facility

The Syrian golden hamster was determined early in SARS-CoV-2 research to be a desirable model because of its high susceptibility to infection and efficient transmission between animals.^7,8 As part of the CEIRR network SARS-CoV-2 response, we were tasked with providing information on vaccine efficacy in this model. Hamsters were housed in modified rodent microisolation caging stacked on stainless steel shelving within FFIs. The requirement for hamsters to be housed individually⁹ presented a unique challenge for experimental design. To integrate this model into the FFI, we used the NexGen 1800 multispecies cage system, engineered by Allentown with biosafety and security as a key priority.

The cages incorporate positive/negative air control, T-99 cage filters, and an optional cage lock into its design (Figure 3). With respect to SARS-CoV-2, Allentown developed a unique cage divider that permitted hamsters to be individually housed, yet still be utilized for aerosol transmission. They also provided various hole pattern and gap distance options for review before deciding on a design/pattern that worked best. With this adapted cage system, our program was able to house hamsters safely and securely in enough animal numbers to ensure statistical relevance regarding the experimental study design.

Figure 3.

Photograph of the Allentown NexGen 1800 cage system designed to house individual hamsters. The center of the cage has a perforated metal divider, which allowed for SARS-CoV-2 aerosol transmission.

Discussion

In nonpandemic times, the role of the IBC at U.S. institutions is often perceived as a committee with monthly duties that primarily consist of protocol review. It is grossly underappreciated how much additional and immediate work IBCs, as well as occupational health departments, had to undertake after the emergence of SARS-CoV-2. The urgency with which investigations needed to happen inside the ABSL-3+ required review of many new protocols and an occupational health plan for potential worker exposures. Our IBC committee had several emergency meetings, phone calls, and extra reviews during the spring and summer of 2020.

Once these were complete, the EH&S staff worked with Occupational Health, the animal care staff, and the research staff to provide timely and appropriate training. Furthermore, obtaining permission to receive live SARS-CoV-2 on campus for research use involved the creation of a new system within EH&S and the St. Jude Materials Management Department for tracking of packages containing such material. The coordinated efforts of the IBC, EH&S, the animal research group, and the research staff was a testament to the effective response that can occur under increased and immediate demand.

It was expected that the outcome of both physical and chemical inactivation methods used against HPAI and SARS-CoV-2 would be similar since both are enveloped viruses, but methods need to be validated “in house” because there are no entity or inter-agency standards available. In addition, each institution has unique needs for the inactivated products. Therefore, buffers, chemicals, and equipment are varied and sourced from assorted vendors making any kind of standardized system between entities difficult to operationalize. There are a variety of nucleic acid extraction buffers available, and to meet the needs of our downstream applications we used two common buffers QIAGEN RLT and TRIzol LS.

Both buffers have been used extensively for inactivation of viruses,^4,10–12 and in our hands we observed similar results. It should be noted, however, that several attempts were made to inactivate virus by the addition of QIAGEN RLT to virus stocks, in the absence of EtOH. We could not consistently inactivate virus using this approach (data not shown) and is consistent with other reports when using a starting material that has a high concentration of protein or complex matrix in combination with a guanidine isothiocyanate buffer.¹¹ Therefore, it was necessary to add 70% EtOH as part of the inactivation process. This methodology resulted in complete viral inactivation in all tested samples.

For samples that required chemical fixation, we utilized NBF, formaldehyde, and PFA that are commonly used for preservation of cells for microscopic analysis, and routinely used to inactivate enveloped, negative sense RNA viruses, including select agents requiring BSL-3 containment, such as HPAI.^13–16 Formalin has also been routinely used to inactivate vaccine viruses.¹⁷ Formalin and formaldehyde fixation of cells and tissues has been a historic and reliable form of viral inactivation,¹⁸ and our testing provides further validation that this method is both safe and effective, even in complex tissue matrices containing high viral load.

When assessing viral inactivation using heat, we avoided the use of detergents for validation since results can vary based on their type and concentration,¹⁹ because sodium dodecyl sulfate (SDS) is toxic to cells and, therefore, requires removal of this chemical before being able to test its inactivating potential. Since the combination of heat and SDS is commonly used to prepare cell lysates for Western blot analysis, which was our desired final application, we focused on validation with heat alone because that would be a final required step. Importantly, application of heat alone has been demonstrated to successfully inactivate negative sense RNA viruses.^20,21

There have been other thorough reports of inactivation methods and validation from biocontainment laboratories,^4,5 and our data do not conflict with the published work. Although it might seem redundant to publish similar chemical and physical inactivation of enveloped viruses, we would argue that this should be encouraged from institutions, simply because of the lack of operational standards between entities. Inactivation failures are rare, and investigations determine human error is often the root cause and not the method itself.²² The more data that are available only strengthen the overall message that hazardous work done in high-containment laboratories, when conducted in compliance with accurate and effective standard operating procedures, is both safe and secure.

Syrian hamsters are recognized as a valuable model for studying emerging and acute human viral diseases caused by RNA viruses.^7,8 These include paramyxoviruses (Nipah, Hendra), flaviviruses (West Nile virus, Zika, Yellow Fever), and filoviruses (Ebola). The utility of hamsters in SARS-CoV-2 research has extended far beyond susceptibility to infection, but also to pathology of disease, vaccine efficacy, testing of therapeutics, production of antibodies, providing clues as to why there is a loss of taste and smell during infection, and even for direct public health benefit such as showing the effectiveness of mask material to stop transmission.²³

Since our containment facility had not used hamsters before, adopting a new cage system into a biocontainment laboratory required a risk assessment for both the pathogen and the animal model. Our biggest challenge was deciding whether to put the new cages inside or outside the FFI. The NexGen 1800 has features that provide efficient filtration and security; however, we chose to keep them inside the isolators because of concurrent HPAI studies in other animal rooms. The directional air flow inside the ABSL-3+ facility goes from the corridor into the animal room; thus, in the event of a loss of containment in one of the HPAI rooms, there was concern contaminated air might go into a room with active SARS-CoV-2.

Keeping the hamsters inside the FFIs would prevent any cross-contamination of agent in the unlikely event of an air-flow loss or reversal. Health observations were performed directly in the FFI, and since the NexGen 1800 fit inside our biosafety cabinets any procedures that used live virus or sharps, and euthanasia could be performed inside the biosafety cabinets.

In conclusion, we have highlighted three major areas of operational response as it relates to the incorporation of a new pathogen into an ABSL-3+ laboratory select agent space. Although our systems were designed for pandemic response to an influenza virus, the experience with SARS-CoV-2 has left our ABSL-3+ BMS better prepared for response to any emerging respiratory pathogen. The administrative controls, space segregations, validated inactivation methods, and adoption of new caging have enhanced our capabilities in new and unexpected ways.

Ethical Approval Statement

Animal experiments were approved by St. Jude Children's Research Hospital Institutional Animal Care and Use Committee in accordance with the Animal Welfare Act (U.S. Department of Agriculture) and the Guide for the Care and Use of Laboratory Animals.

Authors' Disclosure Statement

No competing financial interests exist.

Funding Information

This project has been funded in whole or in part with Federal funds from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services, under Contract Nos. 75N93021C00016 and HHSN272201400006C.

Supplementary Material

Supplementary Data