Sampling methodologies and dosage assessment techniques for submicrometre and ultrafine virus aerosol particles C.J. Hogan Jr1,2, E.M. Kettleson1,2, M.-H. Lee1,2, B. Ramaswami2, L.T. Angenent2,3 and P. Biswas1,2,3 1Aerosol and Air Quality Research Laboratory, 2Environmental Engineering Science Program and 3Department of Chemical Engineering, Washington University in St Louis, St Louis, MO, USA 2005/0258: received 10 March 2005, revised 21 May 2005 and accepted 22 May 2005 ABSTRACT C.J. HOGAN JR, E.M. KETTLESON, M.-H. LEE, B. RAMASWAMI, L.T. ANGENENT AND P. BISWAS. 2005. Aims: The aerosolization and collection of submicrometre and ultrafine virus particles were studied with the objective of developing robust and accurate methodologies to study airborne viruses. Methods and Results: The collection efficiencies of three sampling devices used to sample airborne biological particles ��� the All Glass Impinger 30, the SKC BioSampler�� and a frit bubbler ��� were evaluated for submicrometre and ultrafine virus particles. Test virus aerosol particles were produced by atomizing suspensions of single-stranded RNA and double-stranded DNA bacteriophages. Size distribution results show that the fraction of viruses present in typical aqueous virus suspensions is extremely low such that the presence of viruses has little effect on the particle size distribution of atomized suspensions. It has been found that none of the tested samplers are adequate in collecting submicrometre and ultrafine virus particles, with collection efficiencies for all samplers below 10% in the 30���100 nm size range. Plaque assays and particle counting measurements showed that all tested samplers have time-varying virus particle collection efficiencies. A method to determine the size distribution function of viable virus containing particles utilizing differential mobility selection was also developed. Conclusions: A combination of differential mobility analysis and traditional plaque assay techniques can be used to fully characterize airborne viruses. Significance and Impact of the Study: The data and methods presented here provide a fundamental basis for future studies of submicrometre and ultrafine airborne virus particles. Keywords: aerosol sampling, biological nanoparticles, differential mobility analysis, dosage, virus aerosols. INTRODUCTION Airborne viruses, such as poxviruses and influenza, are of particular concern because of their ability for rapid infection via the respiratory system. Most viruses are on the order of 25���400 nm (Madigan et al. 1997) in characteristic length, and are known to associate with larger particles and aggregate in natural systems (Hull et al. 1970 Hirst and Pons 1973 Aller et al. 2005). This will lead to a virus particle (defined as a particle containing at least one virus) size distribution which spans the ultrafine (100 nm), submicrometre (1 lm) and micrometre (1 lm) size ranges. Airborne virus particle size distributions are rarely reported, and the samplers commonly used to collect virus particles have been designed for the collection of microme- tre-sized particles (Grinshpun et al. 1997 Willeke et al. 1998). Therefore, previous virus aerosol studies using such samplers were limited to the study of micrometre-sized particles (Trouwborst and Kuyper 1974 Trouwborst et al. 1974 Ijaz et al. 1994 Brooks et al. 2005 Tseng and Li 2005). However, it is unknown as to whether the majority of virus particles in the ambient air are in the micrometre size range, or if a substantial fraction of virus particles are in the submicrometre and ultrafine size ranges. Submicrometre Correspondence to: Pratim Biswas, Department of Chemical Engineering, Box 1180, Washington University in St Louis, St Louis, MO 63130, USA (e-mail: pratim.biswas@wustl.edu). �� 2005 The Society for Applied Microbiology Journal of Applied Microbiology 2005, 99, 1422���1434 doi:10.1111/j.1365-2672.2005.02720.x

and ultrafine particles containing viable viruses would be particularly harmful because of the ability of particles in these size ranges to diffuse through alveolar membranes and rapidly enter the blood stream. Recent research has shown that the size of inhaled particles greatly determines the toxicological and immunological effects the particles have, and in general, the effects are much greater for submicro- metre and ultrafine particles (Cassee et al. 2002 Esmen et al. 2002 Daigle et al. 2003). Therefore, knowledge of the size distribution of airborne virus particles in ambient studies and control of the size distribution of aerosolized virus particles in laboratory studies are of the utmost importance because the health effects of inhaled particles are particle size dependent. For laboratory scale evaluation and animal respiratory challenges, biological aerosols have been aerosolized almost exclusively using collison nebulizers (Ijaz et al. 1994 Lin et al. 2000 Agranovski et al. 2002 Bray et al. 2002 Mainelis et al. 2002a Tseng and Li 2005). However, most virus propagation media foam during the nebulization process, thus additional sample preparation, such as dialysis or centrifugation, is usually required in order to use collison nebulizers to aerosolize biological particles. Airborne biological particles have traditionally been sampled using liquid impingers, which rely on inertial collection mechanisms to collect particles (Terzieva et al. 1996 Tseng and Li 2005). Liquid impingers have a distinct advantage in the collection of biological particles in that most biological analyses require samples contained in liquid media (Terzieva et al. 1996 Lin et al. 2000). In many commercial impingers, sampling liquid loss through eva- poration and the reaerosolization of collected particles greatly reduces the collection efficiency and viability of particles (Lin et al. 1997, 1999, 2000). The SKC BioSam- pler�� (SKC Inc., Eighty Four, PA, USA), also called the swirling aerosol collector, was designed by Willeke et al. (1998) to prevent the loss of sampling liquid during operation and to prevent damage to bacterial cells during collection. It, along with other commercial impingers, has been characterized with regard to the collection of airborne bacterial cells and spores (Lin et al. 1999, 2000). Despite full-scale evaluation for the collection of micrometre-sized particles, to the best of our knowledge, the physical collection efficiency of liquid impingers has not been evaluated for ultrafine and submicrometre particles with diameters 300 nm. Here, the aerosolization of submicrometre and ultrafine virus particles and subsequent collection in three samplers which utilize liquid impingment are evaluated. The aeros- olization method used is similar to that used in respiratory challenges (Roy et al. 2003a,b Roy and Hartings 2003). The collection efficiency of the All Glass Impinger 30 (ACE Glass Inc., Vineland, NJ, USA), the SKC BioSampler��, and a frit bubbler (ACE Glass Inc.) for submicrometre and ultrafine particles is examined as a function of particle size, sampler flow rate, and sampling time, using single-stranded RNA (ssRNA) and double-stranded DNA (dsDNA) bacte- riophages as test viruses. The size distributions of aeroso- lized virus suspensions are presented with discussion on methods to control virus particle size distributions. As many studies use the aforementioned samplers, a strategy to estimate the correction factors necessary to find the true airborne virus concentration (or airborne virus dosage) is provided. A method utilizing differential mobility selection to determine the size distribution function of particles which contain viable viruses is also developed and used on the aerosolized virus suspensions in this study. Overall, this study provides a fundamental background for future studies of submicrometre and ultrafine airborne virus particles. MATERIALS AND METHODS Virus suspensions Bacteriophages MS2 (ATCC 15597-B1) and T3 (ATCC 11303-B3) were used as test viruses in this study. MS2 bacteriophage is an ssRNA icosahedral virus. A single virion has an approximate diameter of 27��5 nm (Golmohammadi et al. 1993). T3 is a multi-subunit dsDNA virus with a small tail subunit and a spherical head that is approx. 45 nm in diameter. MS2 bacteriophages were propagated in bacterial host Escherichia coli (ATCC 15597) in a glucose and thiamine minimal media to a titre of 2 �� 1010 PFU ml)1. For aerosolization, filtered stock suspensions of MS2 were diluted with filtered deionized water (Milli-Q Ultrapure Water Purification System Millipore, Billerica, MA, USA) to a titre of 5 �� 109 PFU ml)1. T3 bacteriophage stock suspensions were prepared in a similar fashion using bacterial host E. coli (ATCC 11303) in an LB broth, calcium chloride and magnesium sulfate solution. The filtered and diluted T3 suspension had a titre of 2��3 �� 108 PFU ml)1. Liquid impingement samplers We evaluated the virus particle collection efficiency of three different samplers, the All Glass Impinger 30 (AGI-30), the SKC BioSampler�� and a frit bubbler. Drawings of the samplers are shown in Fig. 1. All three samplers utilize a collecting liquid to capture and retain aerosol particles. Aerosol particles enter the AGI-30 and flow through a 1 mm diameter nozzle, which is 30 mm above the base of the impinger. When filled with 20 ml of liquid, the nozzle outlet is 10 mm above the resting liquid surface. The sampler outlet is above the nozzle outlet, thus there is a sharp turn in flow streamlines at the nozzle outlet, just above the liquid VIRUS AEROSOL DETECTION 1423 �� 2005 The Society for Applied Microbiology, Journal of Applied Microbiology, 99, 1422���1434, doi:10.1111/j.1365-2672.2005.02720.x