Resistance of Aerosolized Bacterial Viruses to Relative Humidity and Temperature

Abstract

The use of aerosolized bacteriophages as surrogates for hazardous viruses might simplify and accelerate the discovery of links between viral components and their persistence in the airborne state under diverse environmental conditions. In this study, four structurally distinct lytic phages, MS2 (single-stranded RNA [ssRNA]), ϕ6 (double-stranded RNA [dsRNA]), ϕX174 (single-stranded DNA [ssDNA]), and PR772 (double-stranded DNA [dsDNA]), were nebulized into a rotating chamber and exposed to various levels of relative humidity (RH) and temperature as well as to germicidal UV radiation. The aerosolized viral particles were allowed to remain airborne for up to 14 h before being sampled for analysis by plaque assays and quantitative PCRs. Phages ϕ6 and MS2 were the most resistant at low levels of relative humidity, while ϕX174 was more resistant at 80% RH. Phage ϕ6 lost its infectivity immediately after exposure to 30°C and 80% RH. The infectivity of all tested phages rapidly declined as a function of the exposure time to UVC radiation, phage MS2 being the most resistant. Taken altogether, our data indicate that these aerosolized phages behave differently under various environmental conditions and highlight the necessity of carefully selecting viral simulants in bioaerosol studies.

INTRODUCTION

Human populations are constantly exposed to viral particles, whether through direct or indirect contacts with an infected individual or through contaminated environments. Despite precautions, we remain at risk of exposure to infective viral particles, particularly through the airborne route. This mode of transmission is difficult to control in our everyday lives due to the ubiquitous nature of airborne particles, which may harbor infectious materials. Although the airborne route is not the most effective mode of transmission for the majority of known human pathogens, many viruses may be transmitted through this route (1). For example, measles, varicella zoster (2), and variola (3) viruses are naturally transmitted by aerosols. Other viruses, such as Newcastle disease virus, are particularly resistant to aerosolization and may potentially cause infections by the aerosol route (4, 5). On the other hand, the importance of aerosol transmission in the spread of some viruses, such as the influenza virus, is still a subject of debate (6).

Aerosolized particles may be involved in viral transmission at short range through contamination of fomites by the rapid deposition of large droplets. However, true aerosol dissemination implies that sufficiently small infectious particles remain airborne for a prolonged period (2). Particles smaller than 5 μm in aerodynamic diameter have the potential to travel long distances, as they sediment more slowly. However, these smaller particles harbor fewer viruses than larger particles but also less material that might protect the viruses in the airborne state. Indeed, viral resistance to aerosolization is partly dependent upon the composition of the droplet or droplet nuclei (5, 7, 8). Furthermore, the resistance of viruses to aerosolization appears to be unique to each virus (5, 9, 10), although, based on very limited data, some similarities, such as the presence or absence of an envelope, exist between viruses with similar structural components (11).

Laboratory work with pathogens requires appropriate bioconfinement procedures, depending on the biosafety classification of each virus. When pathogens are nebulized in high concentrations, additional safety precautions need to be implemented, adding complexity to the studies. The use of nonpathogenic surrogate viruses might help facilitate aerosol studies. Although it is clear that no viral surrogate can mimic with perfect accuracy the reactivities of all airborne viruses to their environment, the characterization of a panel of surrogates might help establish some general guidelines to help predict the reactivity of some airborne viruses.

Bacterial viruses, or phages, have been used in a variety of fields as viral models, but their potential in aerobiology has been poorly exploited. Airborne phages have been studied mostly for filter testing (12), for phage therapy (13), and as surrogates in biodefense research (14). Phage MS2 has been the most used in these virus aerosol studies. Although phages are specific to their bacterial host, they have some similarities with eukaryotic viruses. Namely, they can be enveloped or not and can possess a single- or double-stranded RNA or DNA genome, which may be segmented, linear, or circular, and the viral capsids exist in a multitude of shapes and sizes (15). Phages can be amplified safely to high concentrations at low cost. Interestingly, phages are even accepted by the Parenteral Drug Association (PDA) and the Virus Filter Task Force as published in the 2008 update of the PDA's 41st technical report (16).

In a previous study, we investigated the resistance to aerosolization and air sampling of several phages (MS2, PR772, ϕ6, ϕX174, and PM2), which were chosen because of some similarities (virion size, nucleic acid composition, and envelope) with pathogenic viruses (5). Here, we investigated the infectivity of phages following exposure to environmental stress in the airborne state using a newly designed rotating environmental chamber (17). Four phages were aerosolized at various temperatures and levels of relative humidity (RH) and exposed to UV radiation.

MATERIALS AND METHODS

Bacteriophages.

The phages used in this study are described in Table 1. Phages ϕ6 (HER 102), ϕX174 (HER 36), and PR772 (HER 221) and their respective host bacterial strains, Pseudomonas syringae HER 1102, Escherichia coli HER 1036, and E. coli HER 1221, were provided by the Félix d'Hérelle Reference Center for Bacterial Viruses (www.phage.ulaval.ca). Phage MS2 (ATCC 15597-B1) and its host, E. coli (ATCC 15597), were obtained through the American Type Culture Collection (ATCC).

TABLE 1.

Description of the tailless phages used in this study

Phage	Phage family	Enveloped	Genome characteristicsa	Bacterial host	Incubation temp (°C)
MS2	Leviviridae	No	ssRNA, linear, 3,569 nt	E. coli	37
ϕ6	Cystoviridae	Yes	dsRNA, linear, segmented, 13,385 bp	P. syringae	25
PR772	Tectiviridae	No	dsDNA, linear, 14,492 bp	E. coli	37
ϕX174	Microviridae	No	ssDNA, circular, 5,386 nt	E. coli	37

^assRNA, single-stranded RNA; dsRNA, double-stranded RNA; dsDNA, double-stranded DNA; ssDNA, single-stranded DNA; nt, nucleotides.

Phages MS2 and ϕX174 were propagated in liquid cultures on their respective hosts as described previously (9, 10). Phage PR772 was propagated on its host on tryptic soy broth (TSB) supplemented with agarose (0.75%), as reported elsewhere (18). Phage ϕ6 was propagated on its bacterial host on TSB soft agar (0.75%), as described previously (10) but with minor modifications. Briefly, a liquefied preparation of TSB supplemented with 0.75% agar was inoculated with an overnight culture of P. syringae. A stock suspension of ϕ6 was added to the inoculated medium and poured over tryptic soy agar (TSA) plates. The plates were then incubated overnight at 25°C, and those with nearly confluent PFU were selected for phage extraction. The soft agar was scraped from the plates and transferred into tubes containing 5 ml of phage buffer (20 mM Tris-HCl [pH 7.4], 100 mM NaCl, 10 mM MgSO₄). The tubes were placed under slow agitation for 6 h. Large debris were removed by centrifugation at 2,800 × g for 10 min, and the supernatant was filtered (0.45 μm) and kept at 4°C until use. Phage amplifications produced approximately 10¹⁰ PFU/ml, as determined by plaque assays (19).

Aerosol chamber.

The specifications and a detailed description of the aerosol chamber used in this study are available elsewhere (17). Briefly, the aerosol chamber is a 55.5-liter aluminum rotating drum sheltered inside an insulated temperature-controlled enclosure. The temperature inside the rotating chamber is modulated by controlling the temperature inside the insulated enclosure with two thermoelectric assemblies (model INB340-24-AA; Watronix, Inc., West Hills, CA) for either cooling or heating. Both ends of the cylindrical chamber are closed with custom-made caps mounted on double-sealed ball bearings. The interior sections of the bearings remain stationary during drum rotation and hold protruding aluminum rods with multiple ports, which are accessible from the exterior of the insulated enclosure. The left cap holds a 254-nm UVC band light (model GCL356T5L/4P; Light Sources, Inc., Orange, CT) at the center of rotation as well as a UV probe (model UV-Air; sglux SolGel Technologies GmbH, Berlin, Germany) placed 5 cm from the light source. A temperature and relative humidity (RH) probe (model RH-USB; Omega) placed inside the rotating chamber through an access port and an identical probe placed inside the insulated box were used to monitor and record temperatures and RH levels in real time throughout the experiments. The speed of rotation of the chamber was set at 1 rotation per minute (rpm) throughout the experiments.

Aerosol generation and sampling protocol.

Phage buffer was used for the preparation of the aerosol generation fluid. Fresh lysates of the four phages were added to the buffer at a final titer of 10⁸ to 10⁹ PFU/ml. Five microliters of concentrated Antifoam A (A5633; Sigma-Aldrich) was added to the final volume of 50 ml of each suspension prepared for aerosol generation. The aerosolization was carried out using a 6-jet Collison nebulizer (BGI Inc., Waltham, Mass.) supplied with medical-grade filtered dry air at a rate of 12 liters/min and a pressure of 20 lb/in² gauge for 10 min.

Immediately after nebulization, an aerodynamic particle sizer (APS) (model 3321; TSI, Inc.) equipped with a 1/100 diluter (TSI model 3302A) was used to acquire data at a flow rate of 5 liters/min for 20 s. The APS pump was then turned off, and the rotating chamber was sealed. A BioSampler (SKC, Inc., Eighty Four, PA) filled with 20 ml of phage buffer was used to sample the aerosols from the drum at a rate of 12.5 liters/min for 20 min. Essentially, the entire aerosol content of the chamber was concentrated into a single sample, maximizing the detection of infective viral particles. Each aerosol sample collected corresponded to a separate nebulization protocol. For each environmental condition described below, aerosol samples were taken after 5 min, 6 h, and 14 h of suspension time, and the air inside the rotating chamber was flushed after each sample. The sample at the time point corresponding to the zero hour of suspension was taken immediately after the APS acquisition. Samplings at time points 0, 6, and 14 h were always carried out sequentially using the same aerosol generation fluid. Each environmental condition was assayed in triplicate or more. The aerosol samples were kept at 4°C until analysis, and the infectivity assays were performed within 3 h following sampling.

Effects of relative humidity and temperature on airborne phages.

To evaluate the effects of temperature and relative humidity on the infectivity of airborne phages, RH levels of 20% (low) and 50% (medium) were assayed at 18°C, whereas an RH of 80% was assayed at 18°C and 30°C. The desired levels of RH were obtained by passing the aerosol through various desiccant-filled pipes (17). At 18°C, 20% RH was obtained by passing the aerosol through 60 in. of desiccant, 50% RH was obtained with 36 in., and 80% RH was obtained with 12 in. In order to attain 80% RH at 30°C, an extra source of humidified air was used instead of the desiccators. The air was humidified inside the rotating chamber through a separate inlet port concomitantly with virus nebulization.

Effects of UVC exposure on airborne phages.

The 254-nm UVC low-pressure mercury lamp inside the drum was used to assess the effects of UV radiation on the integrity of airborne viruses. As indicated above, a UV sensor probe was placed 5 cm from the source to ensure repeatability between experiments. Phage preparations were nebulized into the drum as described above. The chamber was sealed, and the aerosol was allowed to stabilize inside the rotating chamber for 15 min. The temperature inside the chamber was kept at 18°C with a relative humidity of 20%. The UVC light was then turned on for a period of 3, 6, or 10 s; controls without UVC exposure were also performed. Aerosol samples were taken 1 min after the UV exposure with a BioSampler as described above.

Plaque assays.

Plaque assays were performed according to standard protocols using liquefied TSB (0.75% agar). The BioSampler liquid was assayed undiluted and serially diluted in phage buffer. Each aerosol sample was assayed on all four bacterial hosts and incubated overnight at the optimal temperature of each bacterial strain. The infective titers of viruses were calculated as the number of PFU per milliliter of BioSampler collection liquid.

RNA extraction and cDNA synthesis.

RNA extractions from phages MS2 and ϕ6 were carried out using the QIAamp viral RNA minikit (Qiagen Canada Inc., Mississauga, ON, Canada) without the carrier RNA, as described elsewhere (10). Prior to the synthesis of the cDNA, the extracted RNA samples were heat treated at 110°C for 5 min and placed on ice; this step was performed to denature double-stranded RNA segments (10). Then, cDNA was synthesized using the iScript cDNA synthesis kit (Bio-Rad Laboratories, Hercules, CA), according to the manufacturer's protocol. The reaction mix consisted of 5 μl of the RNA sample, 4 μl of 5× iScript reaction mix (Bio-Rad), 1 μl of iScript reverse transcriptase (Bio-Rad), and 10 μl of nuclease-free water.

qPCR analysis.

The primers and probes used for quantitative PCR (qPCR) analysis are described in Table 2. All qPCRs were performed with the DNA Engine Opticon 2 real-time PCR detection system (Bio-Rad). Samples were analyzed using Opticon Monitor software (version 2.02.24; Bio-Rad). The PCR mix contained a 1× final concentration of iQ Supermix (Bio-Rad) and 1 μM forward and reverse primers. The probe concentrations were 150 nM, 300 nM, 200 nM, and 200 nM for MS2 (10), ϕ6 (10), ϕX174 (9), and PR772 (5), respectively. This master mix solution was distributed in aliquots of 23 μl/well for ϕ6 and MS2 phages and aliquots of 20 μl/well for PR772 and ϕX174, in 96-well plates. Two-microliter test samples were added for ϕ6 and MS2 and 5-μl samples for PR772 and ϕX174, for a final volume of 25 μl. The qPCR protocol for ϕX174, PR772, and MS2 was 94°C for 3 min (hot start) followed by 40 cycles of 95°C for 15 s and 60°C for 60 s. The protocol for ϕ6 was 94°C for 3 min (hot start) followed by 40 cycles of 95°C for 20 s and 60°C for 60 s. All qPCR assays were performed in duplicate. A specific plasmid was prepared for each phage; the targeted gene segment was cloned into TOPO (Invitrogen, Carlsbad, CA) and purified using the QIAprep Spin miniprep kit (Qiagen, Inc.) (5, 9, 10). Plasmids were quantified by determining the optical density at 260 nm, and serial dilutions of known concentrations were used to prepare the standard curves.

TABLE 2.

Primers and probes used for qPCR and real-time qPCR

Phage	Primer type and probe	Sequencea	Reference
MS2	Forward	5′-GTCCATACCTTAGATGCGTTAGC-3′	10
	Reverse	5′-CCGTTAGCGAAGTTGCTTGG-3′
	Probe	5′-FAM-ACGTCGCCAGTTCCGCCATTGTCG-BHQ_1-3′
ϕ6	Forward	5′-TGGCGGCGGTCAAGAGC-3′	10
	Reverse	5′-GGATGATTCTCCAGAAGCTGCTG-3′
	Probe	5′-HEX-CGGTCGTCGCAGGTCTGACACTCGC-BHQ-3′
PR772	Forward	5′-CCTGAATCCGCCTATTATGTTGC-3′	5
	Reverse	5′-TTTTAACGCATCGCCAATTTCAC-3′
	Probe	5′-FAM-CGCATACCAGCCAGCACCATTACGCA/IABlk_FQ-3′
ϕX174	Forward	5′-ACAAAGTTTGGATTGCTACTGACC-3′	9
	Reverse	5′-CGGCAGCAATAAACTCAACAGG-3′
	Probe	5′-FAM-CTCTCGTGCTCGTCGCTGCGTTGA/BHQ_1-3′

^aFAM, 6-carboxyfluorescein; BHQ, black hole quencher dye; IABlk_FQ, Iowa black forward quencher; HEX, hexachlorofluorescein.

Analysis of infective virus recovery.

The number of PFU found in each sample was divided by the number of phage genomes found in the same sample volume. The resulting fraction was multiplied by 100 to obtain the percentage of viral genomes associated with a PFU.

Statistical analysis.

The results for quantitative and nominal variables are expressed as means ± standard deviations (SD) and percentages, respectively. Two-way analysis of variance (ANOVA) was performed for comparison between groups and temperature or various RHs. From residuals of the statistical model, the normality assumption was verified with the Shapiro-Wilk test, and the Brown and Forsythe variation of Levene's test statistic was used to verify the homogeneity of variances. For all variables, the graphical analyses of residuals with predicted values revealed a relationship between the variances of the observations and the means for these variables. To estimate the form of the required transformation associated with these variables, a regression approach between the logarithm of the standard deviations and the logarithm of the means from different conditions was used. The logarithm transformation was the appropriate one, and statistical results from these parameters are expressed with the log-transformed values. When these assumptions were not fulfilled after a log transformation, an alternative procedure that does not depend on these assumptions was done. In this procedure, the observations were replaced by their rank, called rank transformation, and the ordinary F test from the two-way ANOVA was applied. This technique produces approximate results, but it has good statistical properties compared to those of exact tests. When both procedures (log-transformed data and the ranks) gave similar results, results from the standard analysis were retained. When the results from both procedures differed, the rank transformation was preferred. When necessary, a posteriori comparisons were performed using the Tukey technique.

The effect of UV exposure on the infectivity of aerosolized bacteriophages was analyzed using a mixed model on log-transformed data. Three experimental factors, one associated with phages (a fixed factor with four levels), one being samples (a random factor), and one linked to time (a fixed factor with three levels), were defined. The last was analyzed as a repeated-measure factor with the use of a symmetric covariance structure. The data were analyzed using a linear mixed model with an interaction term between the fixed factors. The statistical model was fitted to compare phages with heterogeneous variances and was tested as to whether it could be reduced to a statistical model with the same variance across phages. As the effect that specifies heterogeneity in the covariance structure was significant (heteroscedasticity) compared to the effect of the same variance among phages, the statistical analyses were performed using a separate residual covariance structure per phage. The residual maximum likelihood was used as the method of estimation, and the Kenward-Roger method was used to estimate denominator degrees of freedom for the test of fixed effects. A posteriori comparisons were performed using the Tukey method. The normality assumption was verified with the Shapiro-Wilk tests on the error distribution from the Cholesky factorization of the statistical model. The results were considered significant if P values were ≤0.05. The data were analyzed using the statistical package program SAS (version 9.4; SAS Institute, Inc., Cary, NC).

RESULTS

Physical characterization of aerosols.

Particle concentrations following aerosol generation were between 1.1 × 10⁵ and 1.7 × 10⁵ particles/cm³ as determined by the aerodynamic particle sizer. The mass median aerodynamic diameter (MMAD) was 1.12 ± 0.05 μm immediately after nebulization. The variations in particle concentrations and MMADs over time inside the chamber were described previously (17).

Temperature and relative humidity measurements.

The effects of RH (20%, 50%, and 80%) on airborne phages were evaluated at 18°C. The recorded temperature was 17.6 ± 0.3°C, and the RH levels varied slightly within and between experiments. The RH levels were kept between 19.6% and 25.0% for the 20% RH experiments, between 46.6% and 52.7% for the 50% RH conditions, and between 75.0% and 80.3% for the 80% RH conditions. When an intended temperature of 30°C and RH level of 80% were evaluated, the actual temperature was 29.3 ± 0.6°C and the RH varied between 79.1% and 85.0%. To simplify the text, the relative humidity levels will be referred to as 20% RH, 50% RH, and 80% RH and temperatures will be referred to as 18°C and 30°C.

Effect of relative humidity on the infectivity of airborne phages.

Aerosolized particles were maintained in the rotating aerosol chamber for periods of 0, 6, and 14 h at 18°C and RH levels of 20%, 50%, or 80%. The aerosol samples were analyzed by plaque assays and qPCRs, and the percentages of viral genomes associated with infective viral particles were calculated. The four phages were infective at 0 h of exposure at all RHs. The effects of RH on the infectivities of the four phages tested are illustrated in Fig. 1.

FIG 1.

Effects of relative humidity and time of aerosol suspension on phage infectivity. Experiments were conducted at 18°C with various RHs. The infectious ratios after 6 h and 14 h of exposure were compared to the infectious ratio at 0 h to calculate the relative infectious ratio. The dotted line indicates the reference value at time zero. An asterisk indicates a significant difference from the reference value (P < 0.05), and a and b indicate significant differences between relative infectious ratios at different RHs. The reference value (time zero) was below the detection limit for phage ϕ6 at 20% and 80% RH for one of three replicates and at 50% for two of four replicates.

Phage ϕX174 displayed a better resistance at 80% RH, at both 6 and 14 h of exposure, than at 20% RH and 50% RH (P < 0.05). Phage PR772 needs high levels of RH to resist in an airborne state, as no infectious PR772 particles were observed after 6 h and 14 h of exposure at 20% RH and after 14 h of exposure at 50% RH. There was a significant difference between the 6-h and 14-h exposures and the 0-h exposure reference point for all levels of RH. On the other hand, phage ϕ6 resists better at the lower RH (20%, P < 0.05) that at the higher RHs (50% and 80%). Infectious ϕ6 phage particles were not detected after 6 h and 14 h of exposure to 50% RH and after 14 h of exposure to 80% RH. Finally, phage MS2 was highly stable under all RH conditions tested but was significantly more resistant in the aerosol state at 20% RH (P < 0.05).

Effect of temperature on the infectivities of airborne phages.

Phage-containing aerosols were exposed for 0, 6, or 14 h at 18°C or 30°C and at a constant RH of 80% (Fig. 2). Plaques were observed for all phages at time zero, with the exception of ϕ6, which did not form any plaque after aerosolization at 30°C. Phage ϕX174 was resistant to both temperatures but resisted better at 18°C (P < 0.05), although there was a decrease in ϕX174's infectivity after 14 h of exposure (P < 0.05). Phage PR772 was sensitive to both temperatures, as its relative infectivity after 6-h and 14-h exposures was significantly lower than that at the reference time, 0 h. At 80% RH, phage ϕ6 was remarkably unstable after 0 h, 6 h, or 14 h at both temperatures. Finally, phage MS2 was stable during 6 h of exposure at both temperatures, but its infectivity decreased after 14 h of exposure at 30°C (P < 0.05).

FIG 2.

Effects of temperature and time of aerosol suspension on phage infectivity. Experiments were conducted at 80% RH under two temperatures. The infectious ratios after 6 h and 14 h of exposure were compared to the infectious ratio at 0 h to calculate the relative infectious ratio. The dotted line indicates the reference value at time zero. An asterisk indicates a significant difference from the reference value at time zero (P < 0.05), a and b indicate significant differences between relative infectious ratios at different temperatures, and £ and § indicate significant effects of exposure time. The reference value (time zero) for phage ϕ6 was below the detection limit at 18°C for one replicate out of three experiments. All data points (including the reference value) at 30°C for phage ϕ6 were below the detection limit.

Resistance of airborne phage to UV.

Phages were aerosolized at 18°C and 20% RH and exposed to UV light for 0, 3, 6, or 10 s (Fig. 3). A significant effect was noticeable starting at 3 s of UV exposure for all phages (P < 0.0001). Phages displayed various resistance levels to UV. MS2 was significantly more resistant than the others (P < 0.005), while ϕX174 was the most sensitive.

FIG 3.

Effect of UV exposure on infectivity of aerosolized phages. Experiments were conducted at 18°C and 20% RH. The infectious ratios at 3 s, 6 s, and 10 s of exposure were compared to the infectious ratio at 0 s of exposure to calculate the relative infectious ratio. The dotted line indicates the reference value at time zero. a, b, and c indicate significant differences between levels of phage resistance to UV (P < 0.05), and § and # indicate significant effects of UV exposure time.

DISCUSSION

The aim of this study was to compare the effects of environmental conditions on the integrity of four structurally and genetically distinct nonpathogenic phages in aerosols. Four tailless phages, including MS2, which is frequently used as a surrogate for eukaryotic viruses, were nebulized into a rotating chamber and exposed to various environmental conditions. Their infectivity was measured immediately after nebulization as well as after 6 h and 14 h of exposure to the controlled conditions of temperature, relative humidity, and UV radiation. All four phages were nebulized and sampled simultaneously, thereby limiting the variables to the nature of the viruses. In order to consider time or airborne state as a variable factor, the infectious ratios obtained after 6 and 14 h of aerosol suspension were compared to the infectious ratio obtained at time zero. Levels of sensitivity to environmental conditions varied significantly from one phage to another, and their behavior is summarized in Table 3. Phage ϕ6 was highly sensitive to several tested conditions. Consequently, few data points were obtained, and the results must be interpreted with caution.

TABLE 3.

Behavior of the aerosolized phages under the various environmental conditions used^a

Phage	Behavior at a temp of:		Behavior at an RH of:			Behavior under UV light
	18°C	30°C	20%	50%	80%
MS2	+	−	+	+	+	+++
ϕ6	−	ND	+	ND	−	++
PR772	−	−	ND	−	−	++
ϕX174	+	−	−	−	+	+

^a+, resistant; −, sensitive; ND, not detected.

MS2 is overall the most resistant phage used in this study. This phage was stable at all levels of relative humidity and temperatures tested. It also demonstrated a higher level of resistance to germicidal UV radiation. MS2 is widely used because it is nonpathogenic and easy to use and can reach high titers in little time. The sturdiness of this virus makes it a good choice for a variety of studies where highly stable viruses are needed. Previous studies have demonstrated that the infectivity of phage MS2 is not significantly affected by nebulization or sampling (5, 20, 21). While it may be a good surrogate candidate for some aerosol studies, its resistance to some stress factors, such as UV irradiation, also indicates that it may not always be the best representative. For example, adenoviruses have been shown to be more resistant to UV irradiation than MS2 (22, 23). Therefore, other phage models should be tested to identify appropriate surrogates. It is noteworthy that the exposure of airborne viruses to germicidal UV radiation can help in understanding how efficient UV air treatment systems are for inactivating viral particles. However, since these wavelengths are not naturally present on the Earth's surface, a different UV lamp would be required in order to characterize the stability of airborne viruses in the outdoor environment.

The enveloped phage ϕ6 was more resistant to lower levels of RH, which is consistent with some observations suggesting that enveloped viruses, such as influenza virus, are more resistant to drier and cooler conditions (11). Nevertheless, similar structural components do not predict viral resistance. Phage MS2, which does not have an envelope, was also stable at lower RH levels. This highlights the multifactorial nature of viral resistance to environmental stresses.

The nebulization liquid can also have an impact on the outcome of aerosol resistance studies. It has been suggested that the presence of proteins in the nebulization medium might protect the integrity of airborne viruses, at least for phages MS2 (21), ϕ6, and PR772 (5). Here, we used a phage buffer made of a mixture of salts. The protein content was limited to the phage protein and mostly the residual proteins contained in the filtered phage lysates used to prepare the nebulization sample. Thus, we assume that this protective effect was limited in our study due to the low protein content. Since the viruses used were nebulized simultaneously, the airborne particles were all exposed to the same concentrations of salts and proteins. The data obtained are thus specific to the phage and conditions tested.

In addition to the nebulization liquid, other possible factors that might affect viral infectivity in aerosol studies are the nebulization and the sampling procedures. Although the infectivity of phage MS2 was not significantly affected by nebulization and sampling (5, 19, 21), this may not be representative of all viruses. Indeed, the differences in structural components, including the presence or absence of an envelope or of protruding structures, may have a significant impact on the vulnerability of airborne viruses to their environment. Differences in sensitivity have been reported between subtypes of influenza virus (24), suggesting that even minor structural differences may have an impact on viral infectivity. Our results clearly show that phages may present different levels of resistance to various environmental conditions. With this in mind and with consideration of the availability of phage collections, it is highly possible to find a phage or a set of phages with an airborne behavior similar to that of the virus of interest.

There is increasing interest in the short- and long-term fate of airborne viruses and their potential to cause infection after a prolonged exposure to environmental stress factors. One of the major pitfalls for the generation of a sufficient amount of data to start raising some general outlines linking viral structures to their airborne stability is the difficulty in studying pathogenic viruses under standard laboratory conditions. In our opinion, the use of phages is a good compromise to establish some general guidelines without the need for advanced biosecurity precautions. Furthermore, phages have the additional advantage of being less costly and less labor-intensive to amplify than eukaryotic viruses. At this time, the panel of airborne viruses studied is limited and allows only very conservative predictions of viral sensitivity to the conditions encountered in the aerosol state. A better understanding of the conditions affecting viral infectivity as a function of their structural components will help in the elaboration of algorithms for the evaluation of the potential of a given virus to withstand various environmental conditions. It will also certainly lead to the discovery of novel solutions to efficiently eliminate airborne viruses. This will allow faster undertaking of the appropriate preventive measures to minimize the risks associated with the spread of airborne viral particles. Taken together, the results of this study provide a framework for, as well as robust insights into, the study of the resistance of airborne viruses.