NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Committee on Animal Models for Testing Interventions Against Aerosolized Bioterrorism Agents. Overcoming Challenges to Develop Countermeasures Against Aerosolized Bioterrorism Agents: Appropriate Use of Animal Models. Washington (DC): National Academies Press (US); 2006.

Cover of Overcoming Challenges to Develop Countermeasures Against Aerosolized Bioterrorism Agents

Overcoming Challenges to Develop Countermeasures Against Aerosolized Bioterrorism Agents: Appropriate Use of Animal Models.

Show details

3Generation and Characterization of Aerosolized Agents

Testing of aerosolized bioterrorism agents for research purposes does not require strict compliance with Good Laboratory Practice regulations (21 CFR 58), which require that investigators establish standard operating procedures for their testing procedures. However, if testing of aerosolized agents is performed in a standardized way—especially those procedures that may lead to application to the FDA for approval of therapeutics or vaccines—then laboratories can share their findings and compare outcomes. Judging by a number of aerosol generators and aerosol-characterization procedures that were described during the Workshop, it appears that they have not been standardized among laboratories. This lack of standardization is of concern, as it could lead to a high degree of variability between laboratories in terms of their results.

In this section of the report, several ways of generating aerosols are described and their advantages and disadvantages enumerated: so that scientists working in this field can be aware of the usefulness of these techniques. This section also offers a number of recommendations for addressing the various barriers to testing the effects of aerosolized bioterrorism agents, as well as methodological criteria that should be included in reports by laboratories working in this field.


Bioterrorism agents can be formulated and generated as liquid aerosols or as dry powder aerosols. Monodisperse aerosols (aerosols having a narrow size distribution) are most useful for calibrating laboratory instruments or for answering basic questions related to aerosol deposition, but they may also be useful in bioterrorism research. Monodisperse aerosol generators include spinning disk, vibrating orifice and condensation devices. On the other hand, polydisperse aerosols (aerosols that consist of particles of various sizes) usually more closely resemble what humans inhale and are often most relevant for countermeasure studies.

Liquid bioaerosols are usually generated by air-blast nebulizers (also known as compressed-air or jet nebulizers) or by ultrasonic nebulizers. The ultrasonic nebulizer produces aerosol particles through the vibration of a piezoelectric crystal, which forms a fountain of liquid that emits droplets from its tip. Although ultrasonic nebulizers produce a large number of droplets per liter of air, they are less applicable to the testing of bioterrorism agents than are air-blast nebulizers because the droplets tend to be larger—too large, in fact—and the heat produced during aerosolization can lead to the degradation of proteins that may be present in viral and bacterial agents. With the air-blast nebulizer, a liquid stream is drawn from a reservoir into the path of a jet of air that is under high pressure. As a result, the liquid shatters into large and small particles. These smaller particles exit the nebulizer and can be inhaled, while the larger particles impact on surfaces within the nebulizer and recirculate into its liquid reservoir. (Descriptions of the basic operation of many air-blast nebulizers can be found in Phalen [1984] and in Moss and Cheng [1995a; 1995b].)

The most commonly used aerosol generator for generating bioaerosols including bacteria, viruses and toxins at USAMRIID and Ft. Detrick has been the Collision nebulizer (Hartings and Roy 2004; Jahrling and others 2004; Roy and others 2003; Zaucha and others 2001; Pitt and others 2001; Johnson and others 1995; Larson and others 1980; May 1973; Henderson 1952). This generator produces droplet aerosols with mass median aerodynamic diameters of 1 to 3 μm. Other aerosol generators, such as the spinning disk aerosol generator, have also been used in some studies (Roy and others 2003). In the case of viruses, bovine serum at concentrations up to 10 percent have usually been added to the generator solutions as a stabilizer (Jahrling and others 2004; Zaucha and others 2001).

Nebulization is an extremely useful method for aerosolizing many substances and is therefore valuable for studying the consequences of inhaling aerosolized bioterrorism agents. Nevertheless, it presents a number of challenges to such studies. First, nebulization can lead to high shear stress levels, which may result in fragmentation and deactivation of bacteria and viruses. Air-blast nebulization can also result in the recirculation of the media and the formation of particle aggregates that are less inhalable than naturally occurring particles. In addition, during air-blast nebulization, the concentration of particles in the aqueous solution steadily increases. This means that, over time, the animal could be exposed to aerosols that contain increasingly concentrated amounts of the agent, resulting in an unrepresentative response. To circumvent these difficulties, researchers have used the following techniques: (1) placing the nebulizer reservoir on ice to reduce evaporative losses and help maintain the aqueous concentration at consistent levels; (2) nebulizing for short periods of time to keep the concentration more consistent and reduce the effects of shear stress; (3) increasing the reservoir volume to reduce the change in concentration over time; and (4) using continuous fluid feed or recirculation with a large external fluid volume. An example of a continuous feed nebulizer is the TSI Model 3076 (TSI Inc., St. Paul, MN).

Generation over a short time period will not eliminate recirculation of the media or all of its consequences. Investigators need to consider: (1) using a single-pass bubbling aerosol generator that does not recirculate the media and therefore may be a good alternative to the air-blast nebulizer (Mainelis and others 2005); or (2) continuously feeding the reservoir with fresh liquid to reduce the artifacts associated with recirculation. Continuously feeding the reservoir is particularly important for long exposures and when using generators that are not single-pass.

A final challenge is that freshly nebulized particles may be electrically charged, which affects deposition within the respiratory tract. In order to reduce this variability, investigators need to consider including a final discharge step for neutralizing the electrical charge (Ji and others 2004; Hinds and Kennedy 2000), unless charged particles are required to produce a specific disease state.

Liquid aerosols can also be generated by devices that utilize spinning disk or vibrating-orifice technology (Rubsamen 1997; Swift 1993) and a number of these devices are commercially available for the administration of therapeutic aerosols. In terms of testing for the effects of aerosolized bioterrorism agents, use of these devices could be an improvement over air-blast nebulization, as they allow for custom-design of the aerosol particle size. Thus they could be used to generate test aerosols of different particle sizes that could be compared in terms of their infectivities. Electrohydrodynamic spraying has also been used to disperse fine droplets (Chen and others 1995).

Solid-particle aerosols are typically generated by the pneumatic redispersion of a dry powder (“dry dispersion”). The output concentration of dry-dispersion generators ranges from milligrams per cubic meter to greater than 100 g/m3. The basic requirements for all dry-dispersion generators are: (1) a means of continuously metering a powder into the generator at a constant rate; and (2) a means of dispersing the powder to form an aerosol. Dispersability of the powder depends on the powder material, particle size and size range, particle shape, and moisture content. To fully disperse a powder, it is necessary to supply sufficient energy to overcome the attractive forces between the particles. Several dry-dispersion aerosol generators are commercially available. For more information about them, see Hinds (1999).

Additional technologies for aerosol generation, resulting from the need for improved therapeutic medications, have led to new formulations of dry powders and new devices for their aerosolization (Laube 2005; Rubsamen 1997). It is possible that these same approaches could be used to improve the animal testing of certain biological agents that can exist in dry-powder form. For descriptions of the basic operation of many dry-powder-inhaler designs, see Dunbar and others (1998).

In the case of microorganisms, the Committee recommends that investigators determine whether the generator fragments or inactivates the organism of interest. If so, this reduced viability would clearly impact the animal’s response to the exposure; the researchers should then select another type of generator.

The Committee also recommends that investigators have some knowledge of the cellular targets of the agent under test within the human airways. They may then more appropriately choose an aerosol generator that will produce aerosol particles capable of reaching the appropriate anatomical regions in the animal model. For example, the cellular targets of a given bioterrorism agent are likely to be located in different regions of the respiratory tract (e.g., the tracheobronchial region versus the alveolar region). How much of the aerosol will be inhaled by a given animal and where in the respiratory tract it will deposit is a function of the aerosol’s particle-size distribution. Techniques for characterizing the aerosol particle-size distribution from a particular aerosol generator will be discussed in detail in the next section of this chapter.

If one chooses the mouse as the animal model for early tiered testing of the effects of aerosolized bioterrorism agents, it is important to keep in mind that an aerosol consisting of 1-μm particles or smaller has the highest probability of depositing in the tracheobronchial and alveolar regions of the lungs of mice that are exposed during nose-only breathing (Raabe and others 1988). Similar results have been shown for Golden Syrian hamsters, Fischer 344 rats, Hartley guinea pigs, and New Zealand rabbits during nose-only breathing exposures (Raabe and others 1988). Therefore, if one is testing an inhaled biological agent in animal models, the Committee recommends that the investigator choose a generator that can produce an aerosol that will reach the intended anatomical sites, which may be the nasopharyngeal, tracheobronchial or pulmonary regions.


Certain properties of inhaled particles, and of the exposure environment, can contribute to the agent’s biological effects. The most important particle properties are its aerodynamic size, size distribution, geometric size and shape, electrical charge, chemical composition, irritancy, and mass concentration (mass of particles per unit mass of air). The most important properties of the exposure environment are its temperature, relative humidity, osmolarity, airflow, and uniformity of the exposure in the breathing zone. Because one laboratory’s well-characterized aerosols and exposure environments may not be readily comparable with those of another, the Committee recommends that these properties be quantified during each inhalation experiment and reported in all publications resulting from the work.

It is convenient to express the size of an irregularly shaped particle by an equivalent spherical dimension—its “aerodynamic diameter” (Dae)—which is defined as the diameter of a unit-specific-gravity sphere having the same settling velocity as the particle being studied. This dimension encompasses the particle’s shape, density, and physical size.

A population of particles can be defined in terms of the mass carried in each particle-size range, and a measure called mass median aerodynamic diameter (MMAD) essentially divides that distribution of the mass in half. The distribution around the MMAD is expressed in terms of the geometric standard deviation (GSD). Aerosols with GSD >1.2 are considered to be polydispersed (i.e., the particles vary significantly in size).

In humans, particles in the range of 1-10 μm or <0.5 μm are most likely to deposit in the tracheobronchial and pulmonary regions of the lungs. However, particles ≤5 μm are thought to have the highest probability of entering the lower airways of the average adult during oral inhalation. Because the human nose is a much more efficient filter of particles compared to the mouth, only particles ≤3 μm have a high probability of entering the lower airways during nose breathing. As shown in Figure 3-1, maximal pulmonary deposition occurs when particles have diameters between 1 and 3 μm or <0.5 μm (NCRP 1997; ICRP 1995). However, there is great variation in the efficiencies and locations of particle deposition in humans, given individuals’ diversity of airway sizes and breathing patterns.

FIGURE 3-1. Fractional regional deposition of inhaled particles in the human (Snipes 1994).


Fractional regional deposition of inhaled particles in the human (Snipes 1994). Reprinted with permission from Medical Physics Publishing, Madison, Wisconsin.

A filter is the most common means of collecting an aerosol sample for assessment. That assessment might include gravimetric weighing on an analytical balance before and after sampling; or it might consist of visual characterization using an optical or electron microscope and a variety of analytical, chemical, or microbiological techniques. Membrane filters can retain particles effectively on their surfaces (good for microscopy and measurements of geometric size and shape), whereas fibrous filters provide in-depth particle collection and a high-load-carrying capacity (good for gravimetric assessment) (Vincent 1995). Specific filter requirements may be identified, depending on the characteristics chosen. For example, weight stability is important for gravimetric assessment and durability is needed for various extraction processes. Filters used for particle counting by optical microscopy need to be capable of being rendered transparent, and filters used for electron microscopy analyses need to allow good transmission of the electron beam.

Measurements of the MMAD and size distribution of a bio-aerosol can be achieved with a cascade impactor, wherein one determines the collected mass on each impactor stage as well as on the back-up filter (Lodge and Chan 1986). This approach, which can be done gravimetrically or by chemical analysis, requires time to obtain the sizing information. Other analyzers utilize optical techniques to provide particle-size information in real time. Such counters may be specifically designed for light scattered at low angles (low-angle devices), light scattered in the generally forward direction below 90° (forward-scattering devices), or light scattered at angles close to or beyond 90° (large-angle devices or devices that use aerodynamic properties of particles). Examples include Climet optical counters (Climet Instruments Company, Redlands, Calif.) and the Aerodynamic Particle Sizer (APS) (TSI Inc., St. Paul, Minn.).

Unfortunately, none of these optical particle counters can differentiate between those particles that contain microorganisms and those that do not, much less which particles contain viable versus nonviable organisms. Thus, in the case of exposures that involve the inhalation of microorganisms, the Committee recommends that if investigators use an instrument that provides real-time particle-size information, they should also provide information regarding the size distribution of particles that contain microorganisms.

In addition, it is important to obtain assessments of organism viability within the size distributions that are critical to deposition inside the lung. For these assessments, survival of the collected particles in their original airborne state is an important consideration in selecting the size analyzer. One factor that can adversely affect survival is desiccation, which usually occurs during particle sizing with a cascade impactor. An alternative to the impactor is the liquid impinger, though it has diminished collection efficiency for particles smaller than 1μm.

It is clear that several different approaches and instruments can be utilized when characterizing particle size and size distribution of inhaled bioterrorism agents. For more information concerning aerosol-sizing instruments, see Phalen (1984) and Lalor and Hickey (1997). For more information about aerosol sampling, see Vincent (1995).

To standardize aerosol characterization, the Committee recommends that measurement of certain parameters be incorporated into standard operating procedures to be adopted by researchers (Table 3-1). The Committee also recommends that specific information (Table 3-2) be included in reports of studies in order to standardize the reporting of aerosol characterization between laboratories.

TABLE 3-1. Recommended Measurements to be Included in Standard Operating Procedures for Generation of Aerosols.


Recommended Measurements to be Included in Standard Operating Procedures for Generation of Aerosols.

TABLE 3-2. Recommended Information to be Included in Scientific Reports on Aerosols.


Recommended Information to be Included in Scientific Reports on Aerosols.

Copyright © 2006, National Academy of Sciences.
Bookshelf ID: NBK20351
PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (929K)
  • Disable Glossary Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...