ABSTRACT
The relative importance of influenza virus transmission via aerosols is not fully understood, but experimental data suggest that aerosol transmission may represent a critical mode of influenza virus spread among humans. Decades ago, prototypical laboratory strains of influenza were shown to persist in aerosols; however, there is a paucity of data available covering currently circulating influenza viruses, which differ significantly from their predecessors. In this study, we evaluated the longevity of influenza viruses in aerosols generated in the laboratory. We selected a panel of H1 viruses that exhibit diverse transmission profiles in the ferret model, including four human isolates of swine origin (referred to as variant) and a seasonal strain. By measuring the ratio of viral RNA to infectious virus maintained in aerosols over time, we show that influenza viruses known to transmit efficiently through the air display enhanced stability in an aerosol state for prolonged periods compared to those viruses that do not transmit as efficiently. We then assessed whether H1 influenza virus was still capable of infecting and causing disease in ferrets after being aged in suspended aerosols. Ferrets exposed to very low levels of influenza virus (≤17 PFU) in aerosols aged for 15 or 30 min became infected, with five of six ferrets shedding virus in nasal washes at titers on par with ferrets who inhaled higher doses of unaged influenza virus. We describe here an underreported characteristic of influenza viruses, stability in aerosols, and make a direct connection to the role this characteristic plays in influenza transmission.
IMPORTANCE Each time a swine influenza virus transmits to a human, it provides an opportunity for the virus to acquire adaptations needed for sustained human-to-human transmission. Here, we use aerobiology techniques to test the stability of swine-origin H1 subtype viruses in aerosols and evaluate their infectivity in ferrets. Our results show that highly transmissible influenza viruses display enhanced stability in an aerosol state compared to viruses that do not transmit as efficiently. Similar to human-adapted strains, swine-origin influenza viruses are infectious in ferrets at low doses even after prolonged suspension in the air. These data underscore the risk of airborne swine-origin influenza viruses and support the need for continued surveillance and refinement of innovative laboratory methods to investigate mammalian exposure to inhaled pathogens. Determination of the molecular markers that affect the longevity of airborne influenza viruses will improve our ability to quickly identify emerging strains that present the greatest threat to public health.
INTRODUCTION
Influenza is a serious respiratory pathogen that causes annual epidemics and occasional pandemics, representing a substantial public health threat worldwide. Transmission of influenza virus occurs through respiratory secretions that can be expelled in a wide range of particle sizes by activities such as coughing, sneezing, talking, and breathing (1), and particle numbers can be enhanced by influenza illness (2). Influenza virus can spread between humans via three, mutually nonexclusive, modes of transmission. First, contact transmission involves direct transfer of infectious respiratory secretions through physical contact between infected and susceptible persons, or indirect transfer of infectious material via contaminated surfaces or objects (3). Second, droplet transmission is associated with particles generally >5 μm in size, which implies short-range transmission through the air since the particles must reach the ocular or respiratory mucosa of a susceptible person before settling (3). However, it was recently shown that droplet transmission may be possible over longer distances than previously estimated. Large droplets expelled from the respiratory tract at high velocities during coughing and sneezing can be carried in turbulent clouds up to 6 m away from the infected person (4). Third, aerosol transmission is mediated by fine particles (<5 μm), including desiccated respiratory droplets referred to as droplet nuclei. These particles remain suspended in the air for prolonged periods of time and can be inhaled deep into the respiratory tracts of susceptible persons (5–7).
The relative importance of each of the three modes of transmission is not fully understood and is the subject of some debate. Mathematical modeling studies (8) and detection of influenza RNA in the air collected from spaces occupied by infected individuals (9–16), or in air directly exhaled from patients (17–24), support a role for aerosol transmission in the spread of seasonal influenza virus. Subjects with detectable influenza virus RNA in their breath can exhale more than 1,000 copies per min (25). Quantification of infectious influenza virus expelled from the respiratory tracts of infected individuals presents additional challenges because of the variability from one person to the next and the low concentrations of virus. It was reported that the number of infectious virus particles in exhaled breath may be 2 to 4 orders of magnitude less than the viral RNA copy number (21, 26). Animal studies revealed that ≤11 PFU of infectious human influenza virus was exhaled from ferrets during 30 min of normal breathing, resulting in a rate of about 3 PFU/min (27). Although these levels are low, findings from human and ferret studies demonstrated that ≤5 50% tissue culture infective doses (TCID50) or PFU of influenza virus is sufficient to initiate an infection (28, 29). In order for inhalation of airborne influenza virus to result in a productive infection, virus infectivity must be preserved as aerosol particles travel from one host to another. While prototypical H1N1 influenza viruses, A/WS/1933 and A/PR/8/1934, were shown decades ago to remain stable in aged aerosols (30–34), little is known about the stability of contemporary influenza viruses that are suspended in aerosols for prolonged periods of time.
The H1N1 viruses that are currently circulating among humans emerged during the 2009 pandemic and were caused by a swine-origin quadruple reassortant virus possessing genome segments from avian, North American swine, Eurasian swine, and human seasonal influenza viruses. The 2009 pandemic H1N1 viruses continue to evolve in swine populations by antigenic drift and reassortment with other swine H1N1, H1N2, and H3N2 viruses (35). Swine-origin influenza viruses that cross the species barrier and infect humans are referred to as variant influenza viruses. Since 2005, more than 400 variant influenza viruses have been isolated from humans, and because they are often antigenically different from the influenza virus vaccine strains, they pose considerable risk to human populations and need to be continuously monitored for their pandemic potential (36, 37).
In this study, we selected seasonal and swine-origin H1 variant influenza viruses that display diverse respiratory droplet transmissibility profiles in the ferret model and assessed their stability in aerosols aged under typical indoor conditions. While the levels of infectious influenza virus declined over time, due to both particle settling and inactivation, viable influenza virus was still detectable after 30 min of suspension in an aerosol state. Sufficient levels of infectious influenza virus remained in the air to cause a productive infection in ferrets that were exposed to the aged aerosols. Notably, influenza viruses capable of efficient respiratory droplet transmission among ferrets displayed higher stability in aerosols than the variant viruses that transmitted with less efficiency. This research provides evidence that contemporary influenza viruses can remain infectious in aerosols for prolonged periods. We propose that the longevity of influenza viruses in an aerosol state may represent a key determinant of efficient airborne transmission. Understanding the molecular markers affecting the stability of influenza viruses that persist long after release from an infected individual will improve our ability to recognize emerging strains capable of efficient airborne transmission, a critical property for a pandemic virus.
RESULTS
Particle size distribution and settling of aerosolized influenza virus aged in an enclosed chamber.Although prototypical laboratory viruses were previously shown to have the ability to persist in aerosols for prolonged periods, it is not clear whether contemporary seasonal and swine-origin variant influenza viruses maintain infectivity over time while in aerosols. Previously established methods and an aerosol management platform were used in this study (29). Following nebulization of virus into the primary exposure chamber, airflow was interrupted while the aerosol particles were aged before being evacuated into a secondary exposure chamber housing a sedated ferret and then collected by a sampler (Fig. 1). To validate the system, particle size distribution was assessed following nebulization of either virus diluent alone (phosphate-buffered saline [PBS] supplemented with bovine serum albumin [BSA]) or a representative virus, A/MN/45/2016 H1N2 (MN/45), in diluent. Samples were collected at three time points: 0, 15, and 30 min postnebulization. A total of 2.45 × 106 to 2.69 × 106 particles were counted during each run. Individual counts for particles ranging between 0.5 to 20 μm were determined (Fig. 2A). Overall, a shift in particle size distribution was observed when MN/45 virus was added to the diluent. Within 30 min of aerosol aging in the exposure chamber, the count median aerodynamic diameter (CMAD) for diluent only changed from 0.924 ± 1.520 μm to 0.858 ± 1.440 μm, while the CMAD for samples containing MN/45 virus changed from 1.030 ± 1.530 μm to 0.970 ± 1.480 μm (Fig. 2B). The shift from large to small particles observed with both types of samples was likely due to the desiccation of particles, and the loss of larger particles was likely due to settling in the chamber.
Schematic representation of aerosol management platform. Influenza virus was aerosolized into the primary exposure chamber, where it was aged and then passed through the secondary exposure chamber, which, during exposure experiments, contained a sedated ferret, and then was collected in a BioSampler for titration. The lower image shows an enlargement of the secondary exposure chamber. The procedure and image are modified from Gustin et al. (29).
Aerosol particle counts following aging in an enclosed chamber. Diluent (black) or A/Minnesota/45/2016 H1N2v virus (107 PFU/ml) in diluent (red) was aerosolized into a chamber, and the aerodynamic diameters of the particles inside the chamber were measured 0, 15, and 30 min after aging using an aerodynamic particle sizer. (A and B) Particle size distributions (A) and count median aerodynamic diameter (CMAD) with geometric standard deviation (GSD) (B) in the range of 0.5 to 20 μm are shown. Data sets represent averages from at least three independent experiments.
Because aerosol particles undergo gravitational settling in still air at a velocity proportional to the square of the particle diameter (5), we expected to detect less virus in the aged aerosol samples. To determine the proportion of virus that is lost during aging due to settling, samples collected at different time points postnebulization were analyzed by real-time quantitative reverse transcription-PCR (qRT-PCR). Five contemporary viruses were tested, including a human seasonal H1N1 virus (Bris/59), H1N1 variant viruses (OH/09, IA/39), and H1N2 variant viruses (MN/45, WI/71). A decline in the total viral RNA recovered at the 15- and 30-min time points for each of the five H1 viruses tested was observed compared to the baseline time point (0 min). On average, viral RNA recovery was reduced to about 67% after 15 min and to 51% after 30 min of aging for all of the virus strains tested, demonstrating the degree to which viral material in aerosols settles in an enclosed environment (Fig. 3). Although half of the aerosolized viral RNA was still present in the chamber air after 30 min, this measurement does not reflect the amount of infectious virus still present in the aerosols.
Settling of influenza virus aerosol particles in an enclosed chamber. A/Minnesota/45/2016 H1N2 (red), A/Ohio/09/2015 H1N1 (blue), A/Iowa/39/2015 H1N1 (black), A/Wisconsin/71/2016 H1N2 variant (purple), and A/Brisbane/59/2007 H1N1 seasonal (green) viruses (106 to 107 PFU/ml) were aerosolized into an exposure chamber. Aerosols were aged for 0, 15, or 30 min inside the sealed chamber and then collected using a BioSampler. Viral RNA copy numbers in each sample were determined by real-time qRT-PCR. The data shown represent the percentages of viral RNA recovered compared to the 0-min time point, which is set at 100%. Each point represents an average from at least three biological replicates. Error bars show ± the standard deviations (SD).
Stability of H1 viruses in aged aerosols.The five H1 viruses included in this study exhibit diverse transmissibility profiles in the ferret model, and the stability of these viruses in aerosols may play a key role in their ability to transmit efficiently through the air. To determine whether influenza viruses aged in an aerosol state remain infectious over time, the collected aerosol samples were further subjected to infectivity testing. Each virus was diluted in PBS with BSA and aerosolized into the primary chamber, and air samples were collected at 0-, 15-, and 30-min time points. All samples were analyzed for the total virus RNA copy number by real-time qRT-PCR and for infectious virus by plaque assay (Fig. 4A to E). A decrease in both RNA copy number and PFU was observed over time. The ratios of RNA copies to PFU were calculated to gauge how much of the observed decrease in infectious virus was due to virus inactivation as opposed to aerosol settling (Fig. 4F to J). First, we measured the ratio found in the virus-diluent solution prior to nebulization. Comparison of these ratios to those of samples collected just after nebulization (0 min) revealed no significant loss of infectious virus caused by the aerosolization process. With aging, the ratios increased, indicating that lower levels of infectious virus were recovered from the aerosol samples relative to viral genetic material. Similar trends were observed for all of the tested H1 viruses; however, statistically significant differences between the ratios of baseline (0 min) and aged samples were observed for WI/71, OH/09, and IA/39, suggesting that these viruses may be less stable in aerosols compared to Bris/59 and MN/45 viruses. Interestingly, Bris/59 (seasonal H1N1) and MN/45 (variant H1N2) viruses were previously shown to transmit through the air among ferrets with 100% efficiency, while WI/71, OH/09, and IA/39 viruses transmitted with 33 to 66% efficiency (38–40).
Influenza virus stability in aerosols. H1 viruses were aerosolized into an enclosed chamber and air samples were collected at 0, 15, and 30 min after nebulization. (A to E) The samples from at least three independent experiments were analyzed in duplicate by a plaque assay to determine the infectious virus titer (left axis/bars) and by real-time qRT-PCR to determine viral RNA copies (right axis/lines). (F to J) Ratios were calculated by dividing RNA copy number values by PFU values and are shown as the log10 (RNA copies/PFU ± SD). Nebulizer samples were also collected prior to the aerosolization process. The statistical significance of the ratio comparisons was determined using two-way ANOVA, followed by Tukey’s posttest (*, P < 0.05; **, P < 0.01; ***, P < 0.001).
H1 variant influenza virus replication and infectivity in ferrets exposed to aged aerosols.Although we have shown that infectious virus recovered from aged aerosols can infect cultured cells, it is not clear whether these airborne influenza viruses are also capable of causing a productive infection when inhaled by ferrets. Because swine-origin H1 variant viruses have only been shown to infect ferrets after intranasal inoculation (39), we first wanted to confirm that aerosol delivery of unaged H1 variant virus would result in a similar course of disease in ferrets. To this end, a representative variant virus, MN/45, diluted in PBS, was aerosolized into a primary chamber, and then the air was immediately evacuated into a secondary chamber housing a sedated ferret. Three groups of ferrets were exposed to aerosolized virus (the presented doses ranged from ≤5 to 435 PFU). The presented dose is an estimate of the amount of virus inhaled by the animal. For comparison, an additional group of ferrets was inoculated via the intranasal route (standard dose of 1 × 106 PFU). Similar to published findings (39), all three ferrets that were intranasally inoculated shed virus in nasal washes that peaked on day 1 postinoculation at 1.4 × 107 PFU/ml (Fig. 5A). Aerosol inhalation challenges with presented doses of 361 to 435 PFU or 6 to 41 PFU of unaged influenza virus also resulted in the productive infection of all of the animals with mean peak nasal wash titers on days 3 to 5 postexposure at 3.9 × 106 and 5.8 × 106 PFU, respectively (Fig. 5B and C). In comparison to the intranasal route of delivery, the ferrets that were challenged via aerosol inhalation had decreased, although not statistically significant, and delayed peak virus titers in nasal washes; they also did not always exhibit the same level of morbidity, which was likely due to the lower-exposure doses (29) (Table 1). No clear indication of dose dependency on clinical disease or virus shedding was observed among ferret groups. However, when ferrets were presented with unaged aerosol doses of ≤5 PFU, influenza disease was observed in two of three ferrets, as evidenced by nasal wash viral titers that peaked between days 3 and 7 postexposure and averaged 2.1 × 106 PFU/ml, indicating that ferrets can become productively infected by aerosolized H1 variant virus at very low presented doses.
H1 variant influenza virus infectivity in ferrets. (A to F) Ferrets were inoculated with A/Minnesota/45/2016 H1N2 variant virus by the intranasal (IN) route (A) or by aerosol inhalation (AR) using aerosols aged for 0 min (B, C, and D), 30 min (E), or 15 min (F). Each bar represents the nasal wash titer for an individual ferret expressed as the log10 PFU/ml, and inoculum doses are listed in the same order to reflect the respective doses received by each ferret. The limit of detection was 1 log10 PFU/ml.
Comparison of clinical signs observed in ferrets challenged with A/Minnesota/45/2016 virus
We next addressed the question of whether H1 variant influenza viruses that have been kept in an aerosol state can cause a productive infection in ferrets. First, MN/45 virus was nebulized into the primary chamber at approximately the same concentrations and under the same conditions as in the experiment shown in Fig. 5B, at which point MN/45 virus-containing aerosols were aged in the primary chamber for 30 min, evacuated into the secondary chamber holding a sedated ferret, and then collected for virus titration. As a consequence of particle settling during the 30-min aging period, the expected baseline (i.e., unaged) RNA exposure dose of about 8.9 × 104 to 1.1 × 105 RNA copies (Fig. 5B) was reduced to 1.1 × 103 to 1.3 × 104 RNA copies (Fig. 5E). In addition to settling, virus inactivation contributed to the decrease of the expected baseline dose of about 361 to 435 PFU at 0 min postnebulization to about 0.7 to 17 PFU observed 30 min postnebulization. Two of the three exposed ferrets had detectable virus in nasal washes; these levels peaked on day 3 postexposure at an average titer of 7.0 × 106 PFU/ml. Interestingly, all three animals, including the one that did not shed virus in nasal washes, seroconverted with hemagglutination inhibition (HI) titers of ≥5,120 (Table 1). Next, a lower quantity of virus was nebulized to mimic the experiment in Fig. 5C, and the aerosols were aged for 15 min. All ferrets exposed to aged aerosols containing MN/45 virus at a presented dose of 1.7 to 10 PFU became infected, as evidenced by viral titers in nasal washes (peak mean titer, 6.0 × 106 PFU/ml) and seroconversion (HI titer, ≥5,120; Table 1). Clinical signs and symptoms of infection were similar to those observed in ferrets challenged with unaged virus (Table 1). Collectively, these data identify the exceptionally high infectivity of a variant influenza virus in ferrets following aerosolization, even after aging in an enclosed space for prolonged durations.
DISCUSSION
In addition to annual outbreaks of seasonal influenza viruses, zoonotic influenza viruses occasionally cross the species barrier and cause human infection. However, unlike seasonal influenza viruses which are able to efficiently transmit among people, zoonotic influenza viruses generally lack specific adaptations facilitating transmission. Because human populations may not have preexisting immunity to zoonotic influenza viruses, each such event carries a risk for the accumulation of mutations that increase mammalian transmissibility, which could trigger a pandemic (41). Factors such as receptor binding specificity, the ability to replicate in the human respiratory tract, and the hemagglutinin (HA) and neuraminidase balance have all been shown to be critical for transmission (42). Furthermore, compelling evidence exists that supports the importance of adaptations linked to HA activation and the threshold pH of this property (43–45). Since one of the modes of virus spread among humans is aerosol transmission, a key parameter that could limit influenza virus transmissibility through the air may be the ability to maintain infectivity in an aerosol state. In this study, we selected H1 influenza viruses with different respiratory droplet transmissibility profiles in the ferret model, including four swine-origin viruses isolated from humans and one human seasonal virus, to assess and compare stability and infectivity in aged aerosols.
Using previously described methods (29) that were modified to study aged aerosolized virus, we generated particles of ≤20 μm, with the greatest numbers of particles in the respirable range of ≤0.5 to 1.5 μm. This particle size distribution is comparable to that generated by humans, where the majority of expelled particles were shown to be <1 μm (23, 46) and <1 to 4 μm for particles containing influenza virus (14, 18). A reduction in the CMAD was observed when aerosols were aged; this is likely attributable to the loss of larger particles due to settling and desiccation (5), a scenario that could emulate conditions found in poorly ventilated spaces. In our case, settling accounted for about 49% of the loss of viral RNA copies after 30 min of aging, regardless of the particular influenza virus.
The stability of virus in aerosols can be influenced by multiple environmental factors, such as the composition of the vehicle (respiratory secretion or diluent), temperature, and humidity. Experimental data suggest that manipulation of the salt and protein content of virus diluent alters the stability of influenza virus (47, 48). The composition of respiratory secretions are not homogenous but have unique characteristics depending on the anatomical origin of the secretion and the state of infection of the individual (49). To avoid the introduction of additional variables into our experimental design, all of the viruses were diluted in a previously described virus diluent (29), which allowed us to focus on the virus as the only variable. Due to the inherent complexity, respiratory fluids are difficult to recreate in the laboratory but with future efforts to address this challenge, our current findings will provide a baseline for comparisons with diluents that more closely represent human respiratory secretions. Higher temperatures and moderate relative humidity (about 50%) have been shown to correlate with lower virus stability, while lower temperatures and relative humidity correlate with the highest level of stability (30–34). In agreement with this, the transmission of influenza viruses between ferrets or guinea pigs was found to vary at different temperatures and under different relative humidity conditions (50, 51). However, while the recovery of viable influenza virus from a residential space occupied by infected individuals had been previously documented (13), there is a paucity of experimental data evaluating virus stability under the ambient conditions commonly found indoors, where people tend to congregate more during cold and flu season. Our results showed a time-dependent loss of infectious virus at 23 to 24°C and 45 to 55% relative humidity. The most stable influenza viruses tested were also the most transmissible through the air in our ferret transmission model (39, 40), while the influenza viruses exhibiting reduced stability were unable to transmit as well in our model (38, 39). These results reveal a potential link between influenza virus transmissibility and virus stability in aerosols and warrants additional study.
While our ultimate goal is to identify a mechanism to account for the difference in influenza virus stability that we have observed and identify molecular markers for this trait, it is critical to establish a model for testing the effects of genotypic manipulations on phenotypic outcomes. To this end, we characterized the disease resulting from the inhalation of aerosolized MN/45 variant virus, which was previously shown to efficiently replicate in a human bronchial epithelial cell line, as well as in the respiratory tracts of mice and ferrets (39). We found that despite the delayed onset of symptoms and peak nasal wash titers in ferrets presented with low doses of unaged aerosolized virus, robust infection was observed with comparable titers in nasal washes between all infected ferrets. Low doses of influenza viruses were previously shown to be sufficient to infect humans in one study; H2N2 virus delivered via aerosols to human subjects displayed a human 50% infectious dose (ID50) of approximately 3 TCID50 (28). Here, exposure to ≤5 PFU of MN/45 virus resulted in the infection of two of three exposed ferrets, suggesting a low ID50 for this virus, which is similar to what was observed for human seasonal H3N2 and avian H5N1 viruses following aerosol inhalation administration to naive ferrets (ferret ID50s of 1.9 and 4 PFU, respectively) (29). Aging of MN/45 virus in aerosols reduced the amount of infectious virus available for inhalation, as evidenced by a decrease in ferret presented doses by approximately 1 to 2 orders of magnitude. Nevertheless, all animals exposed to aged aerosols became infected. These findings demonstrate that, even after 30 min of aging, sufficient influenza virus remains viable and suspended in the respirable range of aerosols to cause infection.
Aging of aerosols is typically conducted in a rotating drum, which provides a sealed environment to maintain aerosols in suspension for extended periods to assess the long-term aging effects on viruses. However, humans are frequently exposed to virus-laden aerosols as they settle in still air. Understanding differences in how the settling process may affect virus stability and virus infectivity in aerosols is a critical and understudied area of research. Our finding that aged aerosols in an enclosed chamber are infectious at low doses to mammals, despite differences in virus viability and particle size distribution compared to unaged aerosols, highlights the heterogeneity of aerosolized virus and the need to study virus-laden aerosols under many discrete conditions. Extension of this research employing a rotating drum would maintain the suspension of aerosols and facilitate the aging of bio-aerosols for longer periods of time (53), allowing for the further study of influenza virus stability in aerosols and evaluate the differences in stability of various virus strains and subtypes.
Although multiple factors may influence the stability of infectious influenza virus while in an aerosol state, one potential mechanism is the ability of the virus to withstand the pH changes encountered in aerosols. Following expulsion from the respiratory tract, aerosol particles quickly undergo desiccation and condensation while equilibrating to the conditions outside the human body. During this process, the chemical microenvironment that influenza virions encounter can fluctuate considerably; the pH of droplets shift based on size, and even within individual droplets, which can affect virus stability (54–56). Typically, human influenza viruses have lower pH thresholds for fusion than swine or avian influenza viruses; hence, mutations leading to a lower HA activation pH (<5.5) are important for human adaptation of zoonotic viruses (43). Mutations that result in acid stabilization of the HA protein were previously associated with adaptation of H5 viruses to the upper respiratory tracts of ferrets and enhanced transmissibility (45, 57). Adaptation of swine-origin 2009 pandemic H1N1 viruses to humans also coincided with stepwise acid stabilization of the HA from a pH threshold of 5.5 to 6.0 observed for precursor 2009 pandemic H1N1 viruses to a pH threshold of 5.2 to 5.4 for isolates that were adapted to humans (43). The human seasonal H1N1 virus, Bris/59, exhibits fusion at a pH of 5.3 to 5.4, whereas the swine-origin variant H1 viruses studied here demonstrate HA activation at pH 5.5 to 5.6 for MN/45 virus and at pH 5.6 to 5.7 for OH/09, IA/39, and WI/71 viruses (39). The HA activation pH has known roles in infectivity and transmissibility of influenza; further studies are needed to better understand the possible link between acid stability and virus stability in aerosols.
In recent years, advances in aerobiology have contributed greatly to our understanding of influenza virus transmissibility. Our finding that aerosolized seasonal and variant H1 influenza viruses lose infectivity over time, but are nonetheless capable of causing productive infections in ferrets exposed to very low doses of aged virus particles, highlights the contribution that aerobiology-based research can add to traditional risk assessment-based studies in the laboratory. Extension of this work examining virus stability in aerosols among other zoonotic viruses with pandemic potential, as well as identification of molecular markers affecting this trait, will improve our knowledge of the role aerosols play in conferring a transmission advantage in humans and will further our understanding of the public health threat posed by novel influenza viruses as they continue to cross the species barrier to cause human infection.
MATERIALS AND METHODS
Viruses.Stocks of A/Minnesota/45/2016 (MN/45) H1N2v (GISAID sequence accession number 221825), and A/Wisconsin/71/2016 (WI/71) H1N2v (GISAID 230639) viruses were propagated in MDCK cells at 37°C for 48 h. Stocks of A/Ohio/09/2015 (OH/09) H1N1v (GISAID 178264), A/Iowa/39/2015 (IA/39) H1N1v (GISAID 194942), and A/Brisbane/59/2007 (Bris/59) H1N1 (GISAID 71838) viruses were propagated in the allantoic cavity of 10-day-old embryonated hens’ eggs at 37°C for 48 h (58). Cell supernatant or allantoic fluid was pooled and clarified by centrifugation, and virus stocks were stored in aliquots at −80°C until use. Each stock was sequenced and exclusivity tested for the presence of other strains, and titers were determined by a standard plaque assay in MDCK cells (38). All work with variant viruses (MN/45, WI/71, OH/09, and IA/39) was conducted in a BSL2-enhanced or higher laboratory setting.
Aerosol experiments.Aerosol exposure experiments were conducted inside a class II biological safety cabinet using the AeroMP aerosol management platform (Biaera Technologies) and methods as previously described with modifications (29). Briefly, virus was suspended in a solution of PBS–0.3% BSA and was aerosolized using a three-jet Collison nebulizer (BGI, Inc.) into a primary chamber (0.04 m3) at a rate of 20 liters/min for 3 min. Airflow through the system was then interrupted, and the chamber was sealed for 0, 15, or 30 min until the airflow was reestablished; air then passed from the chamber through a secondary chamber (for ferret inoculations) and was collected at a rate of 12.5 liters/min for 5 min using a BioSampler (SKC, Inc.) (59) containing Dulbecco modified Eagle medium supplemented with 0.3% BSA and 0.01% antifoaming agent Y-30 (Sigma) (Fig. 1). Air from the chamber entered the BioSampler inlet, passed through the liquid medium (for sample deposition), and was then routed through the outlet. An aerodynamic particle sizer (TSI, Inc.) was used to determine the aerodynamic diameter size distribution of particles in the collected aerosol samples. All aerosol experiments were conducted at 23 to 24°C and at 45 to 55% relative humidity to be consistent with the conditions used during previous transmission assessments of the tested viruses using the ferret model (38–40).
Ferret experiments.Animal experiments were performed under the guidance of the Centers for Disease Control and Prevention’s Institutional Animal Care and Use Committee and were conducted in an Association for Assessment and Accreditation of Laboratory Animal Care International-accredited animal facility. Serologically negative for currently circulating influenza viruses, 11-month-old male Fitch ferrets (Triple F Farms, Sayre, PA) were used for this study. All ferrets were housed in Duo-Flo Bioclean mobile units (Lab Products Incorporated, Seaford, DE) during experimentation. Aerosol inhalation inoculations were conducted as previously described, with modifications (29). Briefly, an anesthetized ferret was placed in a disposable Tyvek sleeve to prevent contamination of the fur and positioned in a secondary exposure chamber with the animal’s snout placed in an airflow inlet cone (Fig. 1). Each exposure was conducted at 23 to 24°C and at 45 to 55% relative humidity for 5 min, while the air inside the primary chamber was evacuated through the secondary chamber and then ultimately collected in a BioSampler for quantification. The presented dose, which represents the amount of virus inhaled, was estimated as previously described by multiplying the aerosolized virus concentration, respiratory minute volume of the animal, and exposure time (29). Control groups of ferrets that were inoculated via the intranasal route received 106 PFU of MN/45 virus diluted in 1 ml of PBS. After intranasal inoculation or aerosol exposure, all ferrets were observed daily for clinical signs of infection for 2 weeks (61). Nasal wash samples were collected every 2 days for virus titer determination using plaque assay (62). Convalescent serum was collected from all ferrets 2 to 3 weeks postinoculation, and seroconversion was assessed by HI assay using homologous virus and 0.5% turkey red blood cells (63).
Real-time quantitative RT-PCR.Total RNA was extracted from each aerosol sample collected by the BioSampler using an RNeasy minikit (Qiagen). Real-time quantitative RT-PCR was performed with a SuperScript III Platinum One-Step qRT-PCR system (Invitrogen) in duplicate reactions using an influenza A virus M1 gene primer and probe set (64). Influenza virus M gene RNA copy numbers were extrapolated using a standard curve based on samples of known M gene copy number.
Statistical analysis.Experimental results were analyzed by two-way analysis of variance (ANOVA), followed by a Tukey’s posttest. Analyses were performed using GraphPad Prism 6.0 software.
ACKNOWLEDGMENTS
We thank the Comparative Medicine Branch for excellent care of the animals used in this study and Alissa Eckert for graphical assistance.
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention or the Agency for Toxic Substances and Disease Registry.
FOOTNOTES
- Received 24 January 2019.
- Accepted 12 March 2019.
- Accepted manuscript posted online 15 March 2019.
- This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.
