Effect of temperature and relative humidity on ultraviolet (UV 254) inactivation of airborne porcine respiratory and reproductive syndrome virus.

The objective of this research was to estimate the effects of temperature and relative humidity on the inactivation of airborne porcine reproductive and respiratory syndrome (PRRS) virus by ultraviolet light (UV(254)). Aerosols of PRRS virus were exposed to one of four doses of UV(254) under nine combinations of temperature (n=3) and relative humidity (n=3). Inactivation constants (k), defined as the absolute value of the slope of the linear relationship between the survival fraction of the microbial population and the UV(254) exposure dose, were estimated using the random coefficient model. The associated UV(254) half-life dose for each combination of environmental factors was determined as (log(10)2/k) and expressed as UV(254) mJ per unit volume. The effects of UV(254) dose, temperature, and relative humidity were all statistically significant, as were the interactions between UV(254) dose × temperature and UV(254) dose × relative humidity. PRRS virus was more susceptible to ultraviolet as temperature decreased; most susceptible to ultraviolet inactivation at relative humidity between 25% and 79%, less susceptible at relative humidity ≤ 24%, and least susceptible at ≥ 80% relative humidity. The current study allows for calculating the dose of UV(254) required to inactivate airborne PRRS virus under various laboratory and field conditions using the inactivation constants and UV(254) half-life doses reported therein.

Introduction

Initially described in the late 1980s, porcine reproductive and respiratory syndrome (PRRS) is characterized by reproductive failure in sows, poor growth performance in growing pigs, and respiratory disease in pigs of all ages (Zimmerman et al., 2012). Since its emergence as a clinical entity in the late 1980s, PRRS virus has proven to be a persistent threat to the health and productivity of pig herds and the economic well-being of pig producers. Neumann et al. (2005) estimated the annual cost of PRRS to U.S. pig producers at $560.32 million per year. By comparison, prior to eradication, annual losses in the U.S. to classical swine fever (hog cholera) and pseudorabies virus were estimated at $364.09 million (Wise, 1981) and $36.27 million (Hallam et al., 1987), respectively (adjusted to year 2004 dollars).

Since the beginning of the PRRS virus pandemic in the 1980s, movement of the virus between neighboring herds in the apparent absence of direct contact (“area spread”) has been often reported (Robertson, 1991). Several epidemiological investigations showed that proximity to infected herds increased the risk of a herd acquiring PRRS virus. In France, Le Potier et al. (1997) found that 45% of herds suspected to have become infected through area spread were located within 500 m (0.3 miles) of the postulated source herd and only 2% were 1 km from the initial outbreak. In Denmark it was observed that the likelihood of herd positivity increased as the density and proximity of PRRS virus-positive neighboring herds increased (Mortensen et al., 2002).

In the last decade, researchers have confirmed the occurrence of airborne transmission of PRRS virus over significant distances. Initially, Dee et al. (2005b) demonstrated that infectious airborne PRRS virus could travel over distances ≥150 m. Thereafter, using a source population of 300 PRRS virus-infected pigs, Dee et al. (2009) demonstrated long-distance airborne movement by the successful recovery of infectious virus up to 4.7 km from the source. Subsequently, Otake et al. (2010) recovered infectious airborne PRRS virus at distance of 9.1 km from the source. Cumulatively, the epidemiological and experimental data suggest that airborne PRRS virus is a significant, and perhaps the primary, means of area spread.

Protection of pig barns from airborne spread of infectious agents is a recent concept. To date, researchers have primarily focused on preventing the introduction of airborne PRRS virus into barns using commercially available air filters, e.g., high-efficiency particulate air (HEPA) filters, minimum efficiency rating value (MERV) filters of various efficiencies, and fiberglass pre-filters. Overall, the results demonstrated that filtering incoming air with HEPA filters and MERV filters ≥14 prevented the transmission of PRRS virus (Dee et al., 2005, Dee et al., 2006a, Dee et al., 2006b, Dee et al., 2010).

Ultraviolet inactivation of PRRS virus may also offer promise. Wheeler et al. (1945) reported the use of ultraviolet to inactivate airborne rubella virus and Streptococcus pyogenes in Army and Navy barracks. Likewise, Perkins et al. (1947) reduced the spread of airborne viral pathogens (“measles”) in school classrooms using ultraviolet. Riley (1961) demonstrated that ventilated air from hospital tuberculosis wards produced tuberculosis in guinea pigs, but not when the air was irradiated with ultraviolet light. In recent years, UV254 emitters have been engineered into areas where people congregate either by placing UV254 light tube grids into existing ventilation ductwork or by installing free standing UVC emitters (Brickner et al., 2003, Dumyuahn and First, 1999, McDevitt et al., 2008, Menzies et al., 1999, Noakes et al., 2006). Ultraviolet technology is appealing due to its low cost as compared to HEPA filtration (Brickner et al., 2003). However, effective implementation must be based on achieving a dose of UV254 sufficient to inactivate the target. No estimates of the effect of UV254 on airborne PRRS virus have been published. Therefore, the objective of this experiment was to evaluate the capability of ultraviolet (UV254) to inactivate airborne PRRS virus under varying conditions of temperature and relative humidity.

Materials and methods

Experimental design

The objective of this study was to evaluate the effect of temperature and relative humidity on the inactivation of airborne PRRS virus by ultraviolet irradiation (UV254). Aerosols of PRRS virus were exposed to four levels of UV254 under nine defined conditions of temperature and relative humidity (Table 1 ). Each combination of temperature and relative humidity was replicated 3 times. Samples of air collected after UV254 treatment were titrated for infectious PRRS virus and the data used to calculate the UV254 inactivation constants (k) and UV254 half-life (T 1/2) exposure doses for each combination of temperature and relative humidity.

Table 1

Temperature and relative humidity combinations of PRRS virus aerosols.

Temperature	Relative humidity
	≤24%	25–79%	≥80%
≤15 °C16–29 °C≥30 °C	Airborne PRRS virus received 4 levels of UV254 treatment at each combination of temperature and relative humidity. Each combination was replicated 3 times.

Porcine reproductive and respiratory syndrome virus

A type 2 PRRS virus isolate, MN-184 (kindly provided by Dr. Scott Dee, University of Minnesota, MN, USA) was propagated on MARC-145 cells, a clone of the African monkey kidney cell line MA-104 (Kim et al., 1993). Cells were grown in 162 cm2 flasks (Corning Incorporated, Corning, NY, USA) using growth media; Dulbecco's Modified Eagles Medium (DMEM), (Mediatech Inc., Manassas, VA, USA) supplemented with 0.25 μg/ml Amphotericin B (Sigma Chemical Co., St. Louis, MO, USA), 50 μg/ml gentamicin (Sigma), 0.5 M l-glutamine (Fisher Scientific, Hampton, NH, USA), 300 international units (IU) per ml penicillin (Sigma), 300 μg/ml streptomycin (Sigma), 1.0% nonessential amino acids (HyClone, Logan, UT, USA), 25 mM HEPES buffer (Sigma Chemical Co.) and 10.0% heat-inactivated fetal bovine serum (FBS) (Sigma Chemical Co.). When cells were confluent (72–84 h), the medium was discarded and the flasks inoculated with 5 ml DMEM (without l-glutamine) containing PRRS virus isolate MN-184 at a virus titer of 1 × 103.5 median tissue culture infective dose (TCID50) per ml. Flasks were placed on a rocking platform in a 37 °C humidified 5% CO2 incubator for 90 min and then 40 ml of supplemented DMEM (Mediatech Inc.) growth medium (now with 4% FBS) was added and the flasks returned to the incubator for 36 h. Cell culture supernatant was harvested by flask freeze-thaw and centrifugation (3000 ×  g for 20 min at 4 °C). Virus stock was stored in 25 ml aliquots at −80 °C.

Experimental procedures

Overview

The system was constructed such that, throughout the 45 min experiment, aerosolized PRRS virus continuously flowed from Reservoir One to Reservoir Two and then across a UV254 exposure field. A manifold in Reservoir Two functioned to equally distribute aerosolized PRRS virus into four quartz tubes placed parallel to each other in the field of ultraviolet irradiation. Each quartz tube represented a different level of UV254 treatment.

Airflow through the system was impelled by negative pressure generated by four AGI-30 glass impingers (Ace Glass, Vineland, NJ, USA), each operating at a flow rate of 12.5 l/min. Thus, air flow through the system totaled 50 l/min. Temperatures in Reservoir One and Reservoir Two were adjusted to achieve targeted temperatures and relative humidities at the UV254 irradiation field. Air samples collected by the impingers downstream of the UV254 exposure field were titrated for infectious PRRS virus.

Aerosolized PRRS virus

The virus suspension to be aerosolized consisted of 25 ml of stock PRRS virus (1 × 107  TCID50/ml), 50 ml of sterile 1× PBS (Thermo Fisher, Rockford IL, USA), and 0.1% (v/v) antifoam A emulsion (Sigma Chemical Co., A5758). Previous research showed that antifoam A emulsion innocuous for cultured cells and PRRS virus (Hermann et al., 2006). The virus suspension was maintained on ice and shielded from light until nebulization. The solution was aerosolized using a 24-jet Collison nebulizer (BGI Inc., CN60, Waltham, MA, USA) operating on compressed air at 1.55 kg/cm2 (22 psi). Under these parameters approximately 1.0 ml of virus solution was nebulized each minute into particles of approximately 1.9 μm diameter (Hermann et al., 2007).

Relative humidity and temperature

For each 45 min replicate, virus was continuously nebulized into Reservoir One (16 l; aerosol residence time ∼19 s). Reservoir One was housed in a chamber capable of maintaining temperatures between −20 °C and room temperature (Carroll Coolers Inc., Carroll, IA). Air was drawn from Reservoir One into Reservoir Two (13 l; aerosol residence time ∼16 s). Reservoir Two was housed in a chamber capable of maintaining temperatures between 0 °C and 40 °C (SS Series 600, 1695-03-36231, Terra Universal Inc., Anaheim, CA, USA).

Based on the rate of nebulization (1.0 ml/min at 1.55 kg/cm) and the temperature capabilities of Reservoirs One and Two, psychrometric calculations determined that the system was theoretically capable of producing an air flow with the RH (%) values listed in Table 2 , with temperature at the UV254 field of exposure approximating the temperature in Reservoir Two.

Table 2

Calculated relative humidity at the UV254 field of exposure.a

	Reservoir One (°C)
	−20	−15	−10	0	5	10	15	20	25	30
Relative humidity (%)	20	31	47	100	–	–	–	–	–	–	0	Reservoir Two (°C)
	14	22	32	70	100	–	–	–	–	–	5
	10	16	23	50	71	100	–	–	–	–	10
	7	11	17	36	50	72	100	–	–	–	15
	5	8	13	26	37	53	73	100	–	–	20
	4	6	9	20	28	38	54	74	100	–	25
	3	5	7	14	21	30	40	55	75	100	30

Air temperature approximated by temperature of Reservoir Two.

For each replicate, the actual temperature and relative humidity of the aerosol was measured (Vaisala, HMI41 indicator and HMP46 temperature probe, Helsinki, Finland) inside a manifold located within Reservoir Two and immediately upstream of the UV254 exposure field. Temperature and relative humidity measurements were taken prior to starting the replicate and then at the beginning, midpoint and end of each replicate to confirm that the system operated at targeted parameters. All targeted (Table 1) combinations of temperature (n  = 3) and relative humidity (n  = 3) were conducted in triplicate.

Ultraviolet (UV254) field of exposure

The UV254 emitting apparatus consisted of six low-pressure, mercury-vapor discharge lamps 28.8 cm in length (American UV Company, Lebanon, IN, USA). Lamps were mounted in three 2-lamp, reflective, surface-mounted, 110 V, fixtures (American UV Company). To avoid fluctuation in UV254 intensity, lamps were operated at their maximum UV254 emission capacity for 10 min before the start of each experiment. UV254 emission was measured using three calibrated radiometers (Model 1700, International Light Inc., Newburyport, MA, USA; VLX3W Technika, Phoenix, AZ, USA).

Exposure of airborne PRRS virus to UV254 was done by passing the airborne virus through 4 quartz tubes (10 mm internal diameter × 12.75 external diameter × 14 cm in length) connected to a manifold within Reservoir Two. Quartz tubes were placed 25 cm from, and parallel to, the UV254 emitters. Between each replicate, quartz tubes were cleaned with commercial quartz cleaner following the manufacturer's recommendations (Hellmanex® II, Hellma GmbH & Co., Mülheim, Germany). Quartz tubing was evaluated for cleanliness and UV254 absorbance prior to each replicate by measuring UV254 intensity beneath and beside the tubing with the impingers in operation (mock aerosol). The UV254 exposure dose measured directly below the quartz tubing.

Ultraviolet (UV254) treatment of PRRS virus aerosol

Each of the four quartz tubes delivered a different UV254 treatment. This was achieved by shielding all but a specific length of each tube, i.e., 1.3 cm (residence time of 0.07 s); 3.2 cm (residence time of 0.14 s); and 5.2 cm (residence time of 0.25 s). A completely shielded tube served as a non-exposed (positive) control. Treatment (shielding) was randomized to quartz tubes before each replicate. UV254 irradiance was measured at the unshielded area of each quartz tube at the start, midpoint, and end of each replicate and averaged. For each treatment, the UV254 dose delivered to airborne PRRS virus was calculated as:where D is the ultraviolet dose (mJ/cm2), I is irradiance (mW/cm2), and T is residence time(s). I was calculated as the average of the irradiance measured at the start, midpoint, and end of each replicate.

Based on the parameters described above, the mean UV254 exposure doses across all replicates for the four treatments were calculated as: zero for the non-exposed (positive) control, 0.05 (S.D. 0.009) mJ/cm2; 0.12 (S.D. 0.016) mJ/cm2, and 0.20 (0.039) mJ/cm2 .

Sampling of UV254-treated airborne PRRS virus

Each quartz tube was independently connected to an all-glass impinger (AGI-30, Ace Glass Inc., Vineland, NJ, USA) containing 25 ml of 1× PBS and shielded against ultraviolet. Impingers were placed on ice throughout the 45 min experiment to avoid dessication of PBS and preserve virus viability. Impingers were operated at a constant flow rate of 12.5 l/min. Vacuum pressure was maintained using oil-less pumps (Fisher Scientific, S413801, Hampton, NH). All connections were sealed and checked for leakage prior to each replicate. Air from the system was exhausted through a biosafety level 2 (BL2) cabinet (NuAire Laboratory Equipment Supply, Plymouth, MN, USA).

PRRS virus microinfectivity assay (TCID50)

Impinger collection fluid was assayed for the concentration of infectious PRRS virus immediately following each replicate. Virus titrations were done on confluent monolayers of MARC-145 cells in 96-well plates (Corning Incorporated). Cell monolayers were prepared by inoculating 100 μl of cells suspended in growth medium into each well, and then incubating the plates at 37 °C in a humidified 5% CO2 incubator for 24 h. Each sample was serially 10-fold diluted in DMEM, with five wells were inoculated with 100 μl of each sample dilution. Thereafter, plates were incubated at 37 °C in a humidified 5% CO2 incubator for 2 h, after which the inoculum was discarded and 100 μl per well of DMEM supplemented with 4% FBS was added. Plates were incubated at 37 °C in a humidified 5% CO2 incubator for 24 h, after which the cells were fixed with aqueous 80% acetone solution and stained with a fluorescein isothiocyanate-conjugated monoclonal antibody specific for PRRS virus (SDOW17, Rural Technologies Inc., Brookings, SD, USA). Virus titers were calculated using the Spearman–Kärber method on the basis of the number of wells showing PRRS virus-specific fluorescence at each dilution and expressed as tissue culture infection dose 50 (TCID50)/ml of impinger fluid.

Statistical analysis

To study the main effects (UV254 dose, temperature, and relative humidity) and their interactions, the TCID50 data were log10-transformed and analyzed using a random-coefficient ANCOVA (analysis of covariance) model with the quantitative explanatory variable “UV254 dose” and categorical explanatory variables of “temperature” and “relative humidity” in SAS® Version 9.2 (SAS® Institute Inc., Cary, NC, USA). Inactivation constants (k), defined as the absolute value of the slope describing the linear relationship between the survival fraction of the microbial population and the UV254 exposure dose (Goldberg et al., 1958), were estimated using the random coefficient model. The associated UV254 half-life (T 1/2) dose for each combination of environmental factors was determined as (log10  2/k) and expressed as UV254 mJ per unit volume.

Results

The effects of UV254 dose, temperature, and relative humidity on the concentration of PRRS virus (TCID50) recovered in air samples were all statistically significant (p  < 0.001), as were the interactions between UV254 dose × temperature (p  = 0.0475), and UV254 dose × relative humidity (p  = 0.0204). Inactivation constant estimates and UV254 half-life dose estimates are given in Table 3 . Comparisons among the three temperature groups detected a significant difference in the UV254 inactivation constant for PRRS virus at low vs. high temperatures (p  = 0.0167), but not between low vs. medium (p  = 0.1044) or medium vs. high temperatures (p  = 0.4635). Comparisons among the three levels of relative humidity revealed a statistically significant difference in inactivation constants between medium vs. high relative humidity (p  = 0.0060), but not between and low vs. medium (p  = 0.0700) or low vs. high relative humidity (p  = 0.3291).

Table 3

Inactivation constants and UV254 half-life doses by temperature and relative humidity.

*Inactivation constant and (half-life). Inactivation constant (k) is the absolute value of the slope of the PRRS virus survival fraction and the UV254 exposure dose. Larger k values indicate more rapid PRRS virus inactivation. Half-life dose is expressed as UV254 mJ per unit volume.

Discussion

In this study, UV254 inactivation constants for PRRS virus were derived for three temperature ranges, three relative humidity ranges, and their combinations. The effects of temperature and relative humidity on the UV254 inactivation of PRRS virus were statistically significant, but the interaction of temperature and relative humidity was not. Ultraviolet inactivation constants decreased in a linear fashion as temperature increased, with statistically significant differences in inactivation constants detected in low vs. high temperatures, but not between low vs. medium or medium vs. high temperatures. The effect of relative humidity on UV254 inactivation was more complex. For any temperature, the rate of PRRS virus inactivation was highest at relative humidity between 25% and 79% and lowest at relative humidity ≥80%. These observations were reflected in statistically significant differences in inactivation constants in medium vs. high relative humidity, but not between other comparisons.

A search of the refereed literature found no publications describing UV254 inactivation of airborne viruses under varying conditions of relative humidity and temperature and only three publications on UV254 inactivation of airborne viruses under varying conditions of relative humidity. In agreement with the results of this study, Tseng and Li (2005) reported that UV254 inactivation of four bacteriophages (MS2, phi X174, phi 6, T7) decreased as relative humidity increased at temperatures of 25 °C to 28 °C and speculated that decreased UV254 susceptibility under higher relative humidity conditions resulted from attenuation of UV254 by water sorption onto the viral surface. In contrast, in a study involving bacteriophage MS2, respiratory adenovirus serotype 2, and mouse hepatitis virus (coronavirus), Walker and Ko (2007) reported that UV254 inactivation increased as relative humidity increased (temperature conditions not reported). This same general trend was reported by McDevitt et al. (2007) who stated there was an increase in UV254 susceptibility with an increase in relative humidity (temperature conditions not reported). Walker and Ko (2007) hypothesized that increased UV254 susceptibility at higher relative humidity could have been a function of larger droplet size at higher relative humidities. Given the overall paucity of data, fruitful hypothesis generation regarding the mechanisms underlying the interactions between UV254 inactivation, temperature, and relative humidity must await additional data on the ultraviolet inactivation of a greater diversity of micro-organisms.

The current study allows for calculating the dose of UV254 required to inactivate airborne PRRS virus under various laboratory and field conditions using the inactivation constants and their associated UV254 half-life doses from Table 3. Since inactivation constants vary by temperature and relative humidity, a conservative estimate of the necessary UV254 dose can be made using the smallest inactivation constant (k  = 2.49, Table 3) and its associated UV254 half-life dose. The percent of the infectious viral population remaining after n UV254 half-life doses may be described as (1/2); thus, the percent of infectious PRRS virus remaining after 10 UV254 doses may be calculated as (1/210) = 0.1%. From this, the dose of UV254 required to inactivate 99.9% of airborne PRRS virus may be calculated using the value from Table 3 as (0.121 mJ/area2/half-life) × (10 half-lifes) = 1.21 mJ/area2. This exposure dose may be achieved under different conditions of ultraviolet intensity and airflow. As given by the Bunsen–Roscoe Law of Reciprocity, if achieved, this dose will be effective regardless of UV254 intensity, distance, or residence time (Riley and Kaufman, 1972).

Conflict of interest statement

The authors declare no conflict of interest.