Assays and enumeration of bioaerosols-traditional approaches to modern practices

Abstract

In recent years, much attention has been drawn to the exposure to bioaerosols in both occupational and indoor environments due to adverse effects on human health. Exposure to these agents may cause infectious diseases, allergic diseases, acute toxic effects, respiratory diseases, neurological effects, and cancer. Bioaerosols play a significant role in air quality studies, as well as in industrial and agricultural regulations. There is a diversity of sources for bioaerosol exposures, which include occupational activities such as waste disposal, sorting and composting, agricultural and food processing, livestock production and handling, healthcare, and fungal growth following flooding. Bioaerosol monitoring includes the measurement of viable (culturable and nonculturable) and nonviable microorganisms in both indoor and outdoor environments. The development of standardized methods for the detection and quantitation of bioaerosols has become an important issue. Technologies, including microbial plating, physical-chemical assays, and molecular techniques, have been adapted for the assessment of collected bioaerosol particles. In this review, we will discuss traditional and modern assays that have been applied in bioaerosol enumeration based on culturability, optical properties, immunoreactions, and nucleic acid amplification and sequencing of biological particles. The analysis of additional characteristics of bioaerosols including microbiome and antimicrobial resistance is also discussed. Finally, concluding thoughts are offered regarding the challenges and perspectives in the field.

Copyright © 2020 American Association for Aerosol Research

EDITOR:

• Tiina Reponen

1. Brief history and summary of topic

In recent years, a growing number of studies have focused on the assessment of exposure to biological agents in occupational and indoor environments due to their various negative effects on human health (Qian et al. 2012; Van Leuken et al. 2016). Although there has been a recent expansion of access to data, models, and management guidelines via the Internet, there is still a need for updating specific studies of aerobiology (Kakde 2012).

In 1861, French scientist Louis Pasteur recorded the first measurements of airborne organisms (Eduard et al. 2012). A century later, the inhalation of spores from thermophilic microorganisms was identified in the development of farmers’ lung, while endotoxins from Gram-negative bacteria were connected to byssinosis in cotton workers (Eduard et al. 2012). From 1985 to 2003, the International Commission on Occupational Health (ICOH) Organic Dust Committee organized four meetings which made important contributions to recognizing organic dust-related diseases by identifying the differences between the potentially disabling disease, hypersensitivity pneumonitis, and the benign organic dust toxic syndrome (Rylander, Peterson, and Donham 1990).

The health effects of the bioaerosols and the nature of the disease (infections vs. allergies vs. toxic effects) play a significant role in determining the analytical method to be used. Fungi and bacteria have been found to produce diverse human health problems including irritations, infections, and allergies (Kim, Kabir, and Jahan 2018). Symptoms of fatigue, cognitive difficulties or memory loss as well as common diseases such as allergy, asthma, and hypersensitivity pneumonitis are caused by fungal exposure (Siersted and Gravesen 1993; Selman et al. 2010; Lee et al. 2006). There is a vast diversity of fungi, with more than an estimated one million species that produce airborne spores, conidia, hyphae, and other fragments that can affect human health (Lee et al. 2010). According to Green, Tovey, et al. (2006), about 112 fungal genera, belonging to three taxonomic groups known as Ascomycota, Basidiomycota, and Deuteromycota, release allergens. In past studies, high concentrations of Aspergillus niger have been found in aerosol samples from homes where asthmatic children reside (Mousavi et al. 2016). Other fungal species, such as Aspergillus ochraceus, Aspergillus unguis, and Penicillium variabile, have also been found in dust associated with childhood asthma (Reponen et al. 2012). Disruption of fungi caused by daily activities usually causes the production of aerosolized fungal fragments, which are broken or fractured conidia and hyphae (Green, Tovey, et al. 2006). Bacterial cell wall components, including endotoxins in Gram-negative bacteria and the less studied peptidoglycans in Gram positive bacteria may induce respiratory symptoms (Douwes et al. 2003).

Viruses, just like any microorganism, can become airborne. They are among the smallest of bioaerosols, with diameters in the nanoregime (>20 nm) (Duan 2006). However, viruses are more likely attached to other airborne particles (e.g., Yang, Elankumaran, and Marr 2011). They can remain infectious outside their hosts for prolonged periods of time (Pirtle and Beran 1991), leading to infections by indirect contact with surfaces that became contaminated by deposited infectious droplets (Verreault, Moineau, and Duchaine 2008).

Allergens are different macromolecular structures comprising mostly of proteins of biological origin, including IgE binding enzymes from fungi and bacteria (Cullinan et al. 2000). Plant pollens are similarly IgE binding allergens that may cause allergic reactions in greenhouse workers (Sinha et al. 2014). Animal proteins, including allergens from dust mite, cat, mouse and rat can also trigger immune responses (Platts-Mills and Woodfolk 2011; Kuehn and Hilger 2015).

When studying bioaerosols at field sites it is useful to monitor environmental factors as well, to fully understand the source of bioaerosols (Mandal and Brandl 2011). Stetzenbach, Buttner, and Cruz (2004) concluded that there is evidence for the association between bioaerosol levels and certain environmental factors, such as temperature and relative humidity. Seasonal variations significantly affect indoor microbial exposures, which are influenced by temperature, relative humidity, and air exchange rates (Talley, Coley, and Kursar 2002; Frankel et al. 2012; Dedesko and Siegel 2015). Everyday ordinary human activities such as talking, walking, sneezing, showering, and cleaning floors also contribute to the generation of bioaerosols (Chen and Hildemann 2009, Estrada-Perez et al. 2018).

There are various levels that bioaerosols can be characterized or quantified. Bioaerosols are categorized into two main groups: viable and nonviable (Table 1), where viable organisms are able to reproduce, whereas nonviable cannot reproduce. Viable bioaerosols can be further categorized as either culturable or nonculturable. Culturable bioaerosols, including fungi, bacteria, and viruses are microorganisms that are able to reproduce in a controlled environment (Jensen and Schafer 1998).

Table 1. Comparison of air sampler performances for sample analysis. Optimal performance is indicated by light gray color, acceptable by medium gray and low by dark gray.

Recently, Pearson et al. (2015) identified a diversity of sources for bioaerosol exposures, including occupational activities such as waste sorting and composting, agricultural and food processing, livestock handling, healthcare and others. In addition, reports indicate that workers who are susceptible to such exposure exhibit symptoms of various respiratory diseases or allergies (Beck, Young, and Huffnagle 2012; Rohr et al. 2015; Sanchez-Monedero and Stentiford 2003).

2. Collection of environmental samples

The preparation of an environmental bioaerosol sample involves collecting air particles on a solid or in a liquid medium and further testing the collected particles for the presence of bioaerosols. Several particle collection methods can be used, with major methods being impaction, impingement, and filtration, while alternative methods include gravitational settling, electrostatic precipitation, and cyclone separation.

Impaction is widely used due to its low cost and is beneficial for direct collection on a solid growth medium with reduced post-sampling processing. However, the sampler can be overloaded if collecting a sample in an area of highly concentrated bioaerosols, making this method more applicable in low concentration areas for enumeration only (Vaughan 1989; King and McFarland 2012a; Ghosh, Lal, and Srivastava 2015). The method involves drawing in air and changing the flow direction causing particles with higher inertia to be impacted on the collecting surface. If the velocity of impact increases, some particles may richochet off the medium resulting in some microorganisms to be damaged and less viable (Ghosh, Lal, and Srivastava 2015). In a study by King and McFarland (2012b), a high airflow bioaerosol sampler with a foam collection surface was used for the impaction of aerosolized E. coli and washed between tests to collect the sample. The viability and DNA integrity of the collected bioaerosols was reduced as compared to the collection by liquid impingement using a wetted wall cyclone collector.

The impingement method uses a liquid medium for collection, where the liquid used can vary depending on whether it is an isotonic or buffered solution (Jensen and Schafer 1998; Mandal and Brandl 2011). Air is drawn through a narrow inlet into the liquid and the collected samples are enumerated by microbial plating or other quantitative methods. Higher total recovery was found for airborne yeast collected by the AGI-30 impinger compared to Nuclepore and gelatin filtration methods where dehydration effects were significant (Lin 1999).

Filtration is a frequently used collection method due to its simplicity, low cost, and effectiveness (Cox and Wathes 1995). For bioaerosols, the most beneficial filter material is gelatin as it offers a better growth environment for plating and doesn’t inhibit molecular analysis (Ghosh, Lal, and Srivastava 2015; Wu, Shen, and Yao 2010). However, because of the filter’s small size and the limitation in flow rate they are generally used in personal samplers (Ghosh, Lal, and Srivastava 2015). When used as a personal sampler, porous media submerged in a liquid layer and plated after collection showed detection limits of 15 bacterial CFU/m³ and above 5 CFU/m³ for fungi at a sampling time of 8 h (Agranovski et al. 2002; Xu et al. 2011). Similarly to impaction samplers, filtration samplers used for enumeration in high concentration environments can become overloaded. Samples are removed from the filter and prepared for analysis through resuspension after collection (Hinds 1982). Losses can occur during the collection and extraction as some bacterial cells collected cannot survive this process, depending on the sampling time, temperature, and relative humidity or remain attached to the filter (Wang et al. 2001). However, vortexing the filters can help in the recovery of bacterial cells from environmental samples.

Gravitational settling is an alternative method of collection where an agar medium is exposed to the environment and airborne microorganisms are allowed to settle onto it due to gravity (Grinshpun, Buttner, and Willeke 2007). Post-sampling preparation is not needed; the sample is collected directly on the medium and can be enumerated with minimal interference. This method is only qualitative since the volume of the air is not known, and the sample would be biased to larger particles, which settle more quickly; thus, not giving an accurate idea of all the microorganisms in the environment (Pasquarella, Pitzurra, and Savino 2000).

Electrostatic precipitation based bioaerosol collectors when compared to the Biosampler (SKC Inc., Eighty Four, PA, USA), may produce more concentrated culturable samples for bacteria, but lower concentrations of fungal spores. The electric charge applied to the particles can cause some microorganisms to become nonviable (Mainelis et al. 2001; Han, Thomas, and Mainelis 2018; Bond and Russel 2000).

Due to the limitations of traditional approaches that can negatively impact microbial viability, cyclone-based bioaerosol samplers including the BioSampler®, NIOSH two-stage cyclone, Coriolis® and the wetted wall cyclone (WWC) were developed (Willeke, Lin, and Grinshpun 1998; Chen et al. 2004; Lindsley, Schmechel, and Chen 2006; Carvalho et al. 2008; McFarland et al. 2010). Cyclones capture the airborne microorganisms in a liquid by centrifugal forces applied to the particles as the airstream rotates around the cyclone at increasing velocities (Sigaev et al. 2006). With WWCs, real-time samples can be collected while maintaining the culturability and DNA integrity of the collected bioaerosols due to continuous liquid flow in the system (Hu and McFarland 2007; McFarland et al. 2010; Hubbard et al. 2011; King and McFarland 2012b). In addition to continuous wetted wall cyclones, there are batch wetted wall cyclones (BWWC), where the particles are collected in liquid batches that are replaced on pre-set time intervals. Similarly to the WWC, BWWC have near-continuous water input that forms a thin film of liquid on the cyclone wall, allowing the air particles to be deposited and transported to the external liquid collection system (McFarland et al. 2010; King et al. 2009). BWWC systems are advantageous because they support long sampling periods, which allow for more concentrated samples, and unlike the impingement method, the sampler accounts for evaporation (King et al. 2009; Yooseph et al. 2013).

3. Analysis of culturable bioaerosols

3.1. Fungi

Most culture-based methods are biased to rapidly growing and highly concentrated species; reducing detection of non-culturable and non-viable spores. There is no ideal medium for growth of all fungi (Keswani, Kashon, and Chen 2005); consequently, a wide variety of media formulations is used since physical and chemical factors have a large effect on the diagnostic characteristics of fungi (Sharma and Pandey 2010). Lignocellulose agar (LCA) was found to be the most suitable for heavy sporulation while potato dextrose agar (PDA) reproduced the most visible colony morphology from decaying vegetable fungi. In most studies, malt extract agar (MEA) and dichloran glycerol 18 (DG18) agar are used for fungal quantification by cultivation (Jürgensen and Madsen 2016). In earlier studies, DG18 agar was recommended for the identification and enumeration of fungi, including xerophilic fungi (Jensen and Schafer 1998). In traditional studies, MEA and rose bengal agar (RBA) are recommended for collection and enumeration of fungi because of the broad spectrum of fungal growth (Jensen and Schafer 1998). However, RBA may make the medium toxic to some fungi depending on whether it is exposed to natural or artificial light. To prevent the growth of bacteria, antibiotics should be added to the medium, such as streptomycin (Jensen and Schafer 1998). Certain conditions, such as the temperature of incubation, can produce morphological changes in dimorphic fungi. For example, Histoplasma capsulatum, a thermally dimorphic human pathogen, exists as a spore or mycelial form below 25 °C, but higher temperatures have shown to cause a transition to the yeast form (Jensen and Schafer 1998; Sil and Andrianopoulos 2015). In a study conducted by Keswani, Kashon, and Chen (2005), Aspergillus fumigatus was cultured on MEA, and Stachybotrys chartarum was cultured on PDA at 25 °C for 10 days to induce sporulation. In a separate study conducted by Górny and Ławniczek-Wałczyk (2012), three isolated fungal species, Aspergillus versicolor, Cladosporium cladosporioides, and Penicillium chrysogenum, commonly found indoors, were grown on MEA for 14 days at 22 °C. In most cases, the incubation temperature remains at a constant 25 °C with varying incubation time depending if sporulation is desired.

Exposure to airborne β(1→3)-glucans, that are glucose polymers originating from most fungi, some bacteria and plants (Stone and Clarke 1992) may induce inflammatory responses and has been studied extensively (Rylander 1996; Kozajda et al. 2017; Adhikari et al. 2009).

3.2. Bacteria

The most used analytical technique for monitoring airborne bacteria in both indoor and outdoor environments is the culture-based colony counting method (Jung and Lee 2016; Byeon et al. 2008; Dong and Yao 2010; Park et al. 2015). Bioaerosol sampling with a one-stage Andersen impactor was found to have higher recoveries for vegetative bacteria than those of MAS-100 and Burkard samplers (Li and Lin 1999). Another advantage of this direct sampling method is the lack of need for post-collection sample processing. However, culture-based colony counting generally requires several days for colony formation following sampling. Additionally, this technique only applies to culturable microbes that are able to divide at a sufficient rate to form colonies. It has been suggested that only 1–2% of all environmental bacteria are culturable (Sharma et al. 2005). As a result, this method may underestimate the number of cells due to the presence of viable but non-culturable (VBNC) cells which proliferate under certain conditions (Park et al. 2015). Correction factors have been developed to reduce counting errors due to colony masking (Chang et al. 1994). A statistical “positive hole correction” table was developed for the well-known six-stage Andersen sampler to evaluate highly loaded plates (Andersen 1958). A method was provided to estimate size distributions for both culturable particles (CP) and culturable organisms (CO) from a single sample collected with an Andersen bioaerosol impactor (King and McFarland 2012a).

Airborne bacteria experience high stress from the lack of optimal conditions, warm temperatures, and hydration. However, they can grow resistance to certain stresses, due to a stress response gene. This can result in them living longer within areas that have low relative humidity, in the range of 30–40% (Ng, Chan, and Lai 2017). The stress response gene increases the survivability of the bacteria in certain conditions; thus, collection and sample preparation is very important for viable bioaerosol samples (Hoisington et al. 2014).

It is essential to obtain pure bacterial colonies in most studies in order to perform enumeration, accurate genome sequencing, antibiotic susceptibility analysis, and bacterial gene manipulation and transformation. Nonselective media, such as tryptic soy agar (TSA), casein soy peptone agar (CSPA) and nutrient agar (NA), permit the growth of most microorganisms since no inhibitors are present (Jensen and Schafer 1998; Lagier et al. 2015). According to Jensen and Schafer (1998), it is preferred to use more than one type of culture medium to collect aerosolized microorganisms; therefore, a nonselective culture medium should be used initially to cultivate all viable and culturable bacteria from a bioaerosol sample. After observation of bacterial growth on a nonselective culture medium, a selective medium is often used to select for specific microorganisms of interest. A selective medium usually contains indicators, specifically of pH, that allow for the recognition of the microorganism of interest due to its particular metabolic activities resulting in pH changes (Atlas 2005). A culture-based enumeration method is solely applicable to culturable bacteria, which usually underestimates the number of cells since many bacteria cells are classified as VBNC.

The time and temperature of incubation used for bacterial plate counts are significant variables that affect bacterial growth. The amount of time needed to grow a microcolony of bacteria can range from hours for a fast-growing bacterium to weeks for a drug resistant microorganism (Jensen and Schafer 1998). Bacteria found in ambient environmental conditions (e.g., outdoors) are usually incubated at a range of 25 to 30 °C, while those sourced from humans are incubated at 35 to 37 °C. Thermophilic bacteria such as Actinomycetes, are incubated at 50 to 56 °C (Jensen and Schafer 1998).

The study by Mirhoseini et al. (2016) utilized two common biosamplers, a single-stage Andersen impactor and an all-glass impinger (AGI), to evaluate levels of airborne bacteria and fungi in indoor environments. Using the colony-forming unit method with TSA and MEA plates at different temperatures, the study showed that particle counting was not an adequate substitute for bioaerosol measurements but should be combined with bioaerosol sampling for rapid detection. The results of Chang and Wang (2014) show that the BioSampler performed better than the 1-stage Andersen and AGI‐30 and using Tween mixture was superior to Phosphate Buffer Saline (PBS) and distilled water for the collection of Staphylococcus aureus bioaerosols.

3.3. Viruses

Cultivation methods for the quantitation of viral bioaerosols includes either human embryonic kidney (HEK) or human embryonic fibroblast tissue cultures (Couch et al. 1966). Although culture is often used to determine virus concentrations, most aerosol sampling methods affect viral infectivity, making culture inadequate for calculating the concentrations of airborne viruses (Verreault, Moineau, and Duchaine 2008). In laboratory studies bacteriophages of E. coli or other bacteria have been used as virus surrogates, enabling the quantitation by double agar overlay plaque assay (Dubovi and Akers 1970; Tseng and Li 2005; Tseng and Li 2006). The overlay assay has been used to determine the stability of spray dried bacteriophage aerosols that have been the focus of numerous studies due to their potential application for the treatment of pulmonary infections (Pabary et al. 2016; Leung et al. 2016; Chang et al. 2017).

3.4. Enumeration of culturable bioaerosols

Enumeration and calculation of the concentration of pathogenic bioaerosols are crucial to risk assessment (Garre et al. 2019). The exposure limit set by the World Health Organization (World Health Organization [WHO]) 1990, 2002) for total bacteria bioaerosols is 500 CFU/m³. Fungal and bacterial concentrations greater than 10⁴ CFU/m³ are considered a health threat (Heida, Bartman, and van der Zee 1995).

A standard method for enumerating bacteria from a sample after collection is to inoculate a small portion on either Luria-Bertani or tryptic soy agar (Missiakas and Schneewind 2013; Novoa Rama et al. 2018). A study completed to compare different agar types for the drug resistant pathogen Staphylococcus aureus found that TSA had a higher recovery of S. aureus colonies, after being aerosolized, compared to mannitol salt agar, CHROMagar, Chapman stone medium, and Baird–Park agar (Chang and Wang 2015). Usually, serial dilutions are plated (Mitsuboshi et al. 2016) and incubated at ∼30 °C for 2 days (Gutarowska et al. 2015) or at 37 °C for 1 day (Chuang et al. 2013), depending on the rate of colony growth (Chuang et al. 2013). Optimal incubation time and temperature are culture-specific. One study found that after 3 h of incubation, the CFU count of Lactococcus lactis was smaller than at lower temperatures. However, after 3–12 h the colony count was larger at the lower temperature (Yang and Moon 2018). In another study, E. coli was incubated at 37 °C for only 16 h (Huang et al. 2018).

Colonies should be counted only for plates with 30–300 CFUs (Suaifan, Alhogail, and Zourob 2017). In the study of Kaur et al. (2015), plates with <300 colonies were counted and considered for concentration calculations using a food microbiology standard to calculate the CFU/ml or CFU/m³. Colony-forming units (CFU) of M. tuberculosis measured directly in cough aerosols produced by patients with pulmonary tuberculosis (TB) was found to predict quantitatively IGRA (Interferon gamma release assay) readouts on secondary infection and disease in household contacts of positive TB cases (Acuña-Villaorduña et al. 2018). Contacts of high aerosol cases had greater IGRA readouts (median 4.6 IU (International Unit)/mL) when compared to those with low (0.8, 0.2–10) or no aerosol (0.1, 0–3.7; p = 0.08).

3.5. Identification of culturable bioaerosols

While researchers continue to investigate the best methods of identification, a general method for identifying culturable bioaerosols is first collecting air samples, followed by a culturing step to isolate single colonies, then taking another culturing step to grow the different single colonies, and finally identifying the colonies (Duquenne 2018). Researchers can identify bioaerosols by macroscopic and microscopic phenotyping (Dawson et al. 2015; Xu and Yao 2011). For culturable fungal aerosols, microscopy method can be used for the identification of fungal genera based on visual comparison of their images with published fungal morphologies (Bold, Alexopoulos, and Delevoryas 1980). Macroscopic phenotyping can give insight into the growth conditions of a microorganism, for example, agar nutrients needed, temperature, light and oxygen requirements (Allen and Waclaw 2019). Microscopic phenotyping can give information such as cell shape, color, size, and Gram stain association. Based on microscopy, bacteria can be grouped into Gram-positive cocci, Gram-positive spore bacillus, Gram-positive bacteria, Gram-negative non-enteric bacteria, and Gram-negative intestinal bacteria (Liang et al. 2013; Yang, Blair, et al. 2016). Gram-positive bacteria have thicker cell walls and Gram-negative cells are covered by an outer membrane outside of the thin cell wall that can be distinguished by staining (Coico 2005; Bayraktar et al. 2019). This information is important for classifying bacteria but also in treating infections (Kapoor, Saigal, and Elongavan 2017).

Several types of biochemical tests, such as catalase, oxygen tolerance, etc., are used to further classify microorganisms by uncovering how a microorganism oxidizes or metabolizes. In one study, a biochemical API 20 A test (www.biomerieux.com) was used for the identification of culturable microorganisms by analyzing the ability of anaerobic bacteria to break down organic substrates and detecting the appropriate metabolites synthesized by the reactions (Duquenne 2018). For this purpose, biochemical metabolism tests such as API 20 CH or the API 50 CH are used in clinical settings (Cyprowski et al. 2018). Another study used McIlvaine buffer to show the pH influence of the culture medium on the Gluconacetobacter physiology (Yassine et al. 2016).

4. Analysis of non-culturable bioaerosols

VBNC and nonviable and nonculturable bioaerosols are classified through different methods than used for culturable microorganisms. These techniques include traditional microbiology methods, microscopy, immunoassays, molecular techniques and mass spectrometry (Table 2). Microscopy provides visual images and limited biological information of the collected particles that have to be collected undamaged. Immunoassays based on the analysis of specific or bulk proteins provide quantification of the biological composition (Womiloju et al. 2003). Nucleic acid based molecular techniques have been used for the identification of organisms at the species level (Hua and Tong 1992). Williams, Ward, and McCartney (2001) have applied conventional polymerase chain reaction (PCR) to analyze air samples for the presence of airborne mycobacteria and fungi commonly associated with adverse health effects. A PCR assay allows for the detection and identification of non-culturable airborne microorganisms (Peccia and Hernandez 2006) but does not allow for distinguishing between non-viable and viable microorganisms. To cover the gap between traditional microbial and molecular techniques for bioaerosol monitoring the culture-based analysis was combined with molecular analysis to increase the observed bacterial diversity (Hubad and Lapanje 2013). Currently, viability real-time PCR (RT-PCR) analysis is capable of accurate measurements of total microorganism concentrations in environmental samples with the ability to discriminate between live and dead cells by using propidium monoazide (PMA) (Nocker and Camper 2009). However, the toxicity of PMA at higher concentrations presents a limitation of the method (Taylor, Bentham, and Ross 2014). The advantage of RT-PCR is the ability of rapid sample quantification and species-specific identification (An, Mainelis, and White 2006). Microarrays, based on multiplexed 16S PCR reactions can identify a large number of genes if their sequence is known (Brodie et al. 2006). Whole genome sequencing and metagenomic analysis offer broad information about a bioaerosol sample (Boissy et al. 2014). The molecular assays offer the potential of real-time identification of single bioaerosol particles, however, the complexity of assays makes it difficult to detect single particles.

Table 2. Current practices for the assays and enumeration of bioaerosols.

	Culturable Bioaerosols	Viable but Non-Culturable Bioaerosols	Non-Viable and Non-Culturable Bioaerosols	Non-Viable and Non-Culturable Bioaerosols (without Intact DNA)
Microbial Enumeration Culture Techniques	X
Biochemical Assay Adenosine Triphosphate (ATP) Based Bioluminescence	X	X
Microscopy Bright Field or Light, Phase Contrast Fluorescence, Epifluorescence Electron Microscopy (SEM or TEM)	X	X	X	X
Endotoxin assay	X	X	X	X
Immunoassays Radioimmunoassay (RIA) Fluorescent Immunoassay (FIA) Enzyme Immunoassay (EIA) Enzyme-Based Halogen Immunoassay (HIA)	X	X	X	X
Molecular Techniques Synthetic Oligonucleotide DNA Probes Nick-translated DNA Probes Polymerase Chain Reaction (PCR)  -Conventional PCR  -Real-Time (quantitative) PCR  -Viability PCR  -Digital PCR  -Convective PCR Microbiome Analysis (Pyrosequencing, Illumina Sequencing)	X	X	X
Mass Spectrometry	X	X	X	X

Endotoxin assays, where the major cell wall component of gram-negative bacteria, the lipopolysaccharide (LPS) is detected based on a colorimetric method, can provide researchers with valid endotoxin exposure estimates (Zhang and Ghosh 2000). Although the Limulus Amebocyte Lysate (LAL) assay has been used most extensively to quantify endotoxins (Milton et al. 1990), other methods including the whole blood assay (Liebers et al. 2009) and the toll-like receptor (TLR4) assay (Peters, Fritz, and Bufe 2012) are also available (Duquenne, Marchand, and Duchaine 2013).

4.1. Microscopy

Bright-field microscopy is the oldest and best-known microscope technique, using a simple microscope with front or backlighting with basic slides. Standard bright-field microscopy uses dyes to enhance the contrast between the specimen and its surroundings and by observing characteristics of the cells, help identify them. This technique is limited by the optical resolution of the microscope (no less than 200 nm) and the contrast method applied.

Further research is being conducted to use machine learning and visual processing to identify microorganisms on a slide or other medium. The simple collection, mounting, and sensing requirements for bright-field microscopy mean that more effort can be spent on the computing requirements, rather than equipment setup. During an intentional release of a pathogen, topological similarities can be quickly examined for a monodisperse bioaerosol and identified with high accuracy (Wagner and Macher 2012). For mixed bioaerosols such as pollen and spores, image recognition is less precise but still useful. A holographic technique was demonstrated with up to 96% precision when identifying pollens and mold spores with the highest consistency for the largest particles (Wu et al. 2018).

Phase-contrast microscopy is a variant of bright-field microscopy that takes advantage of the variation in the refractive index between a microorganism and its surrounding medium to enhance contrast and provide easier viewing (Zernike 1942; Morris 1995). Phase-contrast microscopy can be used to observe live microorganisms, but is particularly useful for imaging low-contrast specimens. Fluorescence is a powerful tool that can be used in microscopy to examine particular structures or molecules in a microorganism or its surroundings. Fluorescence can be achieved either with fluorescent dyes or with cells that are naturally or artificially fluorescent. A variant of fluorescent microscopy is confocal microscopy, which allows very precise resolution and location of fluorescent molecules in a cell. Epifluorescence microscopy has been used to determine viral abundance in bioaerosols (Michaud et al. 2018).

Visible light microscopy as described in the previous sections has a resolution limit of approx. 200 nm. Any structural detail smaller than this scale cannot be observed (Morris 1995).

Electron microscopy is a technique that uses a beam of electrons to observe a target, and is capable of considerably higher magnifications, with a practical limit around 0.5 nm. Because an electron beam is dispersed by air, the highest pressure an electron microscope can operate at is about 2.7 kPa.

Scanning electron microscopy is currently the most useful electron microscopy method for identifying and counting bioaerosols. Because it observes the surface of an object and does not require thin-slicing like transmission microscopy, it is also more accessible. SEM analysis enables the morphological classification of bioaerosols (Vestlund et al. 2014), however, the assignment of a large number of collected particles can be challenging (Wittmaack et al. 2005).

Recently, cryogenic transmission electron microscopy (cryo-TEM) has been used for structural studies and identification of biological particles (Patterson et al. 2016).

There is no currently available electron microscopy method to observe live microorganisms, but collected particles can be examined for classification, counting, and physical characteristics. Due to the significantly higher magnification available from an electron microscope, even very small particles can be identified with high confidence (Straumfors et al. 2014).

4.2. Endotoxin and β-D-glucan assays

The chemical structure and toxicity of endotoxins differ across species of Gram-negative bacteria. Using endotoxin from E. coli is typically used as a reference to assess the combined activity of endotoxins with the Limulus amebocyte lysate (LAL) assay using an enzyme system from the horseshoe crab (Spaan et al. 2007). However, the LAL method is prone to interlaboratory variations and can be used only for water-soluble endotoxins (Chun et al. 2006). Monoclonal antibody-based methods have also been developed but are less sensitive than the LAL assay (Eduard et al. 2012). Endotoxins can also be estimated by gas chromatography-mass spectrometry using 3-hydroxy fatty acids as chemical markers (Spaan et al. 2008).

Two methods have been described for the measurement of (1→3)β-D-glucans, that originate from most fungi, based on the LAL assay (Aketagawa et al. 1993) and an enzyme immunoassay (Douwes et al. 1996).

4.3. Bioluminescence—ATP assay

One of the potential methods to reduce processing time in bioaerosol analysis is the bioluminescence-based technique that detects the presence of adenosine triphosphate (ATP), the basic energy molecule present in all types of living organisms. The method uses the firefly enzyme luciferase to catalyze a reaction between its substrate D-luciferin and ATP, causing luciferin to emit photons in the 500 nm range (Karl 1980). Since the intensity of produced light is directly proportional to the ATP content (which is proportional to biomass), it is possible to quantify microbial biomass by measuring the ATP content using bioluminescence. The ATP method was used successfully for the rapid characterization of bioaerosol sampling devices when collecting bacterial aerosols in various environments (Seshadri et al. 2009; Park et al. 2014).

4.4. Immunoassays

Immunoassays are based on the specific reaction between antigens and antibodies. Monoclonal or polyclonal antibodies can be employed to detect the collected bioaerosols by binding to antigens, which are usually the proteins or polysaccharides, on the cell surface of the organisms. Many immunoassays are commercially available, allowing the rapid detection and identification of the microbes regardless of whether they are culturable. Some of the more widely used formats are detailed below (Jensen and Schafer 1998):

4.4.1. Radioimmunoassay (RIA)

The assay is based on the competitive binding of the unlabeled antigen (in the collected bioaerosol) to the specific antibody in the presence of the quantitatively added radiolabeled antigen. Higher concentrations of unlabeled antigen in the sample will result in lower levels of radiolabeled antigen-antibody complexes. A calibration curve created with known concentrations of the unlabeled antigen enables the assessment of the amount of an unknown, unlabeled antigen present in a sample (Garvey, Cremer, and Sussdorf 1977; Douwes et al. 2003).

4.4.2. Fluorescent immunoassay (FIA)

Fluorescent immunoassays utilize fluorescent-labeled antibodies to detect bacterial antigens either by directly detecting the antigen (cell-bound) using fluorescent antibody or by indirectly using antibody and fluorescent anti-gamma globulin antibody. In indirect FIA, serum antibody is detected using antigen, serum, and fluorescent antibody. A fluorescent microscope is used to evaluate the samples and to count the number of fluorescent viruses and microorganisms (Garvey, Cremer, and Sussdorf 1977; Hemmillä 1985).

4.4.3. Enzyme immunoassay (EIA)

In enzyme immunoassays, the antibody is coupled with an enzyme, such as horseradish peroxidase (HRP) or alkaline phosphatase (AP), which, in the presence of a chromogen, produces a colored end-product that can be quantitated by spectrophotometry. Most commercially available EIAs are enzyme-linked immunosorbent assays (ELISAs) where the antibody is coated onto the surfaces of test tubes, or wells of a microtiter plate, enabling rapid detection and speciation of the collected bioaerosols (Douwes et al. 2003).

4.4.4. Enzyme-Based halogen immunoassay (HIA)

The halogen immunoassay relies on the detection of airborne particles by capturing them onto protein-binding membranes by volumetric air sampling (Tovey et al. 2000). Eluted antigens are either detected with monoclonal antibodies (mAbs) to identify specific antigens of interest (environmental monitoring) or immunostained with human serum IgE to identify the patient’s sensitization patterns (serological monitoring) (Green, Millecchia, et al. 2006).

4.5. Molecular methods

Molecular methods have the potential to quantify exposure to airborne microbes independently of culturability and with high specificity. DNA extraction includes lysis to separate microbial cell contents, followed by precipitation and purification (Gorny et al. 2018). Rittenour et al. (2012) compared different types of DNA extraction kits for DNA extraction starting with a lysis buffer and using a supernatant mix with a binding buffer to wash the DNA precipitate. A vortex step and a heating step are recommended to reduce the amount of DNA that will bind to filters. These steps have led to a 10-fold increase in DNA yield, allowing for more analysis of air samples (Dommergue et al. 2019).

The molecular level identification methods include PCR, mass spectrometry and sequencing. PCR analysis detects and amplifies a target segment of DNA (Stewart and Dowhanick 1996). Mass spectrometry measures the protein mass of isolated cultures and compares the results to a database of species’ profiles. Mass spectrometric analysis of bioaerosols has been described in another review in this special issue (Huffman et al. 2019). Sequencing is the most expensive route of identification (Rodrigues et al. 2017). Sequencing technologies use universal 16S rRNA to encode the genes and match the bacterial DNA sequence to a known profile. After PCR amplification the pure product of the 16S gene is sequenced, and compared to bacterial DNA segments in a database to identify the bacteria (Barghouthi 2011). Analysis of the fungal community in a complex sample is performed by sequencing the internal transcribed spacer (ITS) in the fungal rRNA operon (McTaggart et al. 2019). Species-specific primers can be used to identify the collected microorganisms (Novoa Rama et al. 2018).

Nebulization and sampling of aerosols may cause stress on the microorganisms, hindering the direct representation of the biodiversity at the sample site. Finding ways to represent biodiversity and analyze microbes from larger air volumes will provide a more accurate insight into what humans, animals, and other living organisms are exposed to (Duquenne 2018; Adhikari et al. 2017). Some DNA extraction methods can create PCR inhibitors and it is crucial to uncover these inhibitors for future research (Ferguson et al. 2019).

In a study involving students with influenza-like symptoms, cough-generated aerosols were collected using a NIOSH two-stage bioaerosol cyclone sampler or an SKC BioSampler. Quantitative real-time reverse-transcription PCR (qPCR) analysis of the viral RNA in the samples indicates that the cough-aerosols contain influenza virus and viral RNA within particles in the respirable size range (Lindsley et al. 2010). The number of viral aerosol studies in swine facility surveillance has significantly increased in recent years (Anderson et al. 2017). A study by Lauterbach et al. (2018) on the molecular detection and recovery of influenza A virus from both the air and surfaces collected at agricultural fairs provides evidence for potential intra- and inter-species viral transmission through these routes.

4.5.1. Polymerase chain reaction (PCR)

4.5.1.1. Conventional PCR

Conventional polymerase chain reaction (PCR) is a system performed for DNA amplification, allowing the detection of specific nucleic acid sequences of DNA found in both culturable and non-culturable microorganisms. According to Alvarez, Buttner, and Stetzenbach (1995) solid-phase PCR amplification detects airborne bacterial cells when a traditional culture cannot. Therefore, conventional PCR can be used as an alternative method for total bacterial count assessment and has been applied to analyze air samples for endemic microorganisms such as airborne mycobacteria and fungi (Li, Gao, and Liu 2011). During PCR amplification the desired gene or part of a gene in the collected bioaerosol is detected by amplifying the relevant fragment in a DNA template using a thermostable Taq polymerase enzyme and specific DNA primers that will recognize and bind only to the target DNA sequence (Alvarez, Buttner, and Stetzenbach 1995).

DNA sample preparation used for PCR amplification should be optimized for the successful amplification of increasingly longer targets from genomic DNA (Cheng et al. 1995). Therefore, before the PCR process, the intactness of the genomic DNA should be maintained during the collection and isolation process to create reproducible amplification of fragments with sizes >1.3 kb (Deagle, Eveson, and Jarman 2006). A larger target size can increase the probability of producing an unusable DNA template strand by randomly introducing a single-stranded (ss) nick or double-stranded (ds) break within the target sequence (Zhou, Pape, and Schwartz 2008). The DNA sample preparation varies for each experiment. Genomic DNA is isolated from the sample’s cultured cells by various methods such as boiling in the presence of a chelating resin, using alkaline lysis, or phenol extraction (Cheng et al. 1995). A standard reaction mixture contains the DNA template along with the specified forward and reverse primers, deoxyribonucleotide triphosphate (dNTP) mix, buffer with Mg⁺⁺ for the specific DNA polymerase, and DNA polymerase. The PCR process typically utilizes thermal cycling, which consists of an initial denaturing step and a series of 20 to 40 cycles of denaturing, annealing and extension steps (Svabenska 2012; Liang and Johnson 1988; Sambrook, Fritsch, and Maniatis 1989).

The product of the PCR process is usually analyzed with gel electrophoresis to verify if the single product of the correct size was formed. Recently nano-materials (metallic and non-metallic) have been introduced to enhance PCR amplification by effectively reducing inhibition and increasing detection efficiencies (Xu and Yao 2013). Although colony counting is not as efficient as PCR, a combination of colony counts and conventional PCR can provide more detailed information when monitoring and verifying the presence of bioaerosols (Li, Gao, and Liu 2011).

PCR techniques are highly sensitive to bioaerosol detection and identification, making it independent of culturing. The advantage of using the PCR method is that it can be applied to most biological matter with nucleic acids, such as fungi, viruses, bacteria, and various allergens (Peccia and Hernandez 2006).

A disadvantage of using PCR is that air samples commonly contain compounds that could inhibit the amplification assay, therefore environmental interferences should be assessed before utilizing PCR for field screening (Alvarez, Buttner, and Stetzenbach 1995). Additional steps may be required to remove or inactivate PCR inhibitors (Chen et al. 2010; Schrader et al. 2012). For example, when using an impactor for bioaerosol collection there is an increased likelihood for the presence of EDTA in the impactor substrate wash liquid (King and McFarland 2012b). High concentrations of EDTA inhibit PCR (Huggett et al. 2008), therefore a procedure should be used to reduce the EDTA content before the PCR process.

4.5.1.2. Real-time (quantitative) PCR

Real-time quantitative polymerase chain reaction (qPCR) is a widely used method for the detection, quantification, and typing of different microbial agents (He and Yao 2011; Kralik and Ricchi 2017). It is also used to compare gene transcription levels (Saint-Marcoux et al. 2015). By simple definition, qPCR is the process of amplifying DNA, using fluorescent (SYBR® Green) dye to detect the dsDNA PCR product as it accumulates during PCR and intercalates with the dye. However, these dyes detect the accumulation of both specific and nonspecific PCR products. To improve the specificity of the detection, fluorogenic-labeled probes (TaqMan) have been introduced that use the 5’ nuclease activity of Taq DNA polymerase (Holland et al. 1991; Georgakopoulos et al. 2009). A probe containing a fluorescent reporter dye on the 5’ end and a quencher dye on the 3’ end anneals downstream from one of the primer sites and is cleaved by the 5’ nuclease activity of Taq DNA polymerase as this primer is extended. The cleavage of the probe separates the reporter dye from the quencher dye, increasing the fluorescence emitted by the reporter dye. Due to its simpler process and lower cost, SYBR^® Green assays are more applied (Libert et al. 2015). The steps of qPCR are the same as for conventional PCR, including DNA denaturation, primer annealing, and extension (Kralik and Ricchi 2017).

Although PCR is a common method for amplifying DNA, it is also applicable for RNA templates by using reverse transcriptase to synthesize complementary DNA (cDNA) (Saint-Marcoux et al. 2015). Reverse transcription is performed by adding a primer or random hexamer primers and transcriptase before amplification (Lauterbach et al. 2018). DNA is less likely to degrade than RNA samples (Bustin et al. 2009). A small volume of the prepared cDNA sample can be used for further amplification (Saint-Marcoux et al. 2015).

Controls, positive and negative, and quantification calibrators are needed in all qPCR reactions. Positive controls are taken from the nucleic acids of the original sample to monitor assay variation and needed for the calibration curves (Bustin et al. 2009). Quantification calibrators need to be diluted and may act as the negative controls (Bustin et al. 2009). Multiple cycles, 20 and above, are applied to amplify the DNA properly. Amplicons containing 80 to 150 nucleotides in length can be used to reduce primer mismatch leading to low amplification efficiency and amplification bias (Saint-Marcoux et al. 2015). After performing the amplifications and comparing against amplicons, a calibration curve is prepared for final quantification either by using dilutions of the target genomic nucleic acid or plasmid standards (Kralik and Ricchi 2017). An alternative method for increased quantification is using whole-cell PCR, where bacterial cells followed by a 15 min heat lysis are added to the reaction as a DNA template (Hospodsky, Yamamoto, and Peccia 2010). Fresh mid-log phase bacteria can be used for the calibration curve (King and McFarland 2012b).

The four factors to consider in assessing the performance of a PCR are the limit of detection, the limit of quantification, precision, and PCR efficiency. Limit of detection, or LOD, is the lowest amount of analyte that can be detected in a 95% confidence, but not the exact value (Kralik and Ricchi 2017; Bustin et al. 2009; Forootan et al. 2017). This value can be estimated from the analysis of replicate standard curve[s] (Forootan et al. 2017). No more than 5% of the positive control samples can fail for the PCR reaction to be considerable. Limit of quantification, or LOQ, similar to LOD, can be estimated from the standard curve (Forootan et al. 2017). LOQ is the smallest amount of analyte which can be measured and quantified with defined precision and accuracy, with a value proposed to be fixed under 25% (Pavšic, Žel, and Milavec 2016).

The precision of qPCR can be affected by temperature differences affecting the completion of annealing and denaturation, concentration differences introduced by pipetting errors, and stochastic variation (Bustin et al. 2009). The repeatability of the qPCR is dependent on intra- and inter-assay variation describing the variability of replicates in the same experiment and the variability of different experiments conducted on different days (Kralik and Ricchi 2017). The amplification efficiency is determined by the slope of the log-linear portion of the calibration curve (Bustin et al. 2009).

After completing the qPCR amplification cycles, a melting curve can be generated for the resulting amplicon presenting a negative first derivative plot to indicate the loss in fluorescence over a temperature range of 55–92 °C. The desired amplicon will appear as a peak with a specific melting temperature (Tm). Potential contamination due to nonspecific binding, or the presence of primer dimers will have lower Tm (An, Mainelis, and White 2006).

Normalization of results from qPCR corrects the biological and experimental variability (Saint-Marcoux et al. 2015). In addition, it allows for comparisons of data from different RNA concentrations (Bustin et al. 2009). Normalization could be done by comparing transcript levels of the gene of interest to those of a reference gene similar to that being tested (Bustin et al. 2009). Qualitative assays of the qPCR data should also be done to verify the validity of the results (Bustin et al. 2009).

4.5.1.3. Viability PCR

Because of the amplification capabilities of PCR reactions, it is possible to design qPCR assays for every microorganism (Kralik and Ricchi 2017). The main advantages of this method are its fast and high-throughput detection and quantification of target DNA in different sequences. In addition, qPCR avoids cross-contamination, has a wide dynamic range for the quantification and allows for multiplexing which is essential for detection and quantification (Kralik and Ricchi 2017). However, the biggest disadvantage of qPCR is the inability to tell the difference between live and dead cells (Kralik and Ricchi 2017), leading to the development of viability qPCR (Nocker and Camper 2009). The application of a simple pretreatment of the sample by using specific intercalating photo-reactive reagents enables the neutralization of the DNA of dead cells. As a result, only DNA from live cells will be detected by PCR. Propidium monoazide (PMA) has been used as an effective intercalating reagent (Nocker, Cheung, and Camper 2006). Although the efficiency of photoactivation has been greatly improved, the viability PCR method is still under development.

4.5.1.4. Digital PCR

Digital PCR (dPCR) carries out a single reaction within the sample, however, it is separated into a large number of partitions (droplets) and the reaction is carried out in each partition individually, allowing a more reliable collection and sensitive, digital measurement of nucleic acid concentration. By measuring the number of droplets that are fluorescing due to amplification, dPCR provides absolute quantification while qPCR relies on the standard curve. The method has been routinely used for clonal amplification of samples for next-generation sequencing. Damit (2017) developed a droplet microfluidics-based bioaerosol detector and analyzed bioaerosols impinged directly into the droplet.

4.5.1.5. Convective PCR

PCR by thermal convection is an alternative to other PCR processes. Convective PCR is shown to be a faster, near-instantaneous method for amplification (Wheeler et al. 2011). In this method, the bottom surface of the sample is heated, then cooled by the surrounding air creating a temperature gradient so PCR can be completed by natural convection (Chou et al. 2011). The thermal gradient through natural convection allows the sample to go through all three steps of PCR. That temperature gradient must be maintained throughout the experiment for it to be successful (Chou et al. 2011). In convective PCR, the temperature during annealing/extension and denaturation must be monitored and can be optimized depending on which gene pair is being amplified. If the temperature is not properly controlled, the process can fail (Zhang and Xing 2009). Overall, PCR by natural convective heat transfer is a simpler, and faster method of amplification.

The sample for the convective PCR is prepared similarly to the conventional PCR (Muddu, Hassan, and Ugaz 2011, Zhang and Xing 2009). The thermocycler, which facilitates convective PCR, had set conditions similar to conventional or real-time PCR technologies (Chou et al. 2011) and may be connected to impingement based continuous bioaerosol collectors via microfluidics. The convective PCR product is separated by agarose electrophoresis (Muddu, Hassan, and Ugaz 2011; Chou et al. 2011). The recently developed capillary version is able to monitor the fluorescence in situ with a smartphone camera, enabling real-time amplification (Qiu et al. 2019). In some cases, additional analysis may be done to validate the amplification, such as melting curve analysis (Chou et al. 2011).

4.5.2. Microbiome analysis

More detailed identification of biodiversity and microbiome is possible with the advancement and combination of powerful tools. The most commonly used method to identify bacterial taxonomy is the 16S rRNA gene sequencing as this gene is present in most bacteria, its function has not changed over time, and it has a fairly long base pair of 1500 bp for analysis (Janda and Abbott 2007). This is advantageous in many ways, such as identifying the species of ambiguous isolates, but it also has its limitations. Some identified species strains are outdated and may not be accurate anymore, the sequencing resolution does not clearly distinguish between certain bacterial strains, and there are still some errors in public databases of sequences. Databases, including the open access GenBank (National Center for Biotechnology Information [NCBI]), the European Nucleotide Archive, the DNA Data Bank of Japan, and the UniProt, Protein Data Bank, Ensembl and InterPro are powerful tools that allow the search and alignment of nucleotide or protein sequences (Boratyn et al. 2013; Wheeler and Bhagwat 2007). Shotgun sequencing is another method of high-throughput sequencing that uses untargeted sequencing over a microbial genome to compile whole genomes. Although it has some limitations, such as high cost, lack of valid annotations, and genome biases, shotgun sequencing results in metagenomics that enables complex microorganisms to be identified and analyzed (Quince et al. 2017). Next-generation sequencing (NGS) platforms enable other types of microbiome analysis and identification, such as whole-genome sequencing, total RNA sequencing, and methylation sequencing (Illumina 2016). The NGS technologies enable the generation of millions of reads from aerosol samples to analyze the genomics and transcriptomics (for RNA analysis), and have been successfully used in several bioaerosol studies (Womack et al. 2015; Shin et al. 2015). Metagenomic approaches enable comprehensive determination of the diversity and metabolic potential of the collected airborne organisms.

Various computer programs and platforms are available for the analysis of genomic sequences. Typically, analyzing the 16S rRNA gene data for bacterial and archaeal communities starts by clustering sequences into operational taxonomic units (OTUs). A similar approach can be used for the processing of other marker genes, including 18S rRNA gene sequences for eukaryotic analyses or ITS sequences for fungal studies (Khot, Ko, and Fredricks 2009). An open-source software package called Mothur combines algorithms from previous tools to perform tasks such as aligning sequences, assigning sequences to molecular operational taxonomic units (OTUs), calculating α and β diversity in samples, and calculating pairwise distances with and without trimming sequences (Schloss et al. 2009, Edgar 2018).

Another software called the Quantitative Insights Into Microbial Ecology (QIIME) and QIIME2 can be used for sequencing 16S rRNA gene fragments and identifying their taxonomy classification (Lawley and Tannock 2016; Hall and Beiko 2018). However, it has been reported that amplicons from Illumina sequencing sometimes contain errors in quality filtering and constructing OTUs for clusters of sequencing reads that differ by less than a fixed dissimilarity threshold. This problem can be overcome by using another open-source R package called DADA2, which contains an algorithm that can correct these errors enabling amplicon sequence variants (ASVs) to be resolved to the level of single-nucleotide differences over the sequenced gene region (Callahan et al. 2016). Recent debates focus on phasing out the traditional OTUs of marker gene sequences and instead delineating microbial taxa using exact sequence variants (ESVs) (Callahan, McMurdie, and Holmes 2017; Glassman and Martiny 2018). This approach avoids clustering sequences and instead uses only unique, identical 16S rRNA sequences for downstream community analyses that could differ by only one base pair. Another error-prone process is the conversion of genome abundances to cell numbers that requires taxa-specific information about the number of genome copies in the cell (Bonk et al. 2018). However, this is a useful way to normalize NGS data (Dannemiller et al. 2014). A number of additional tools are available to ‘de-noise’ and identify ESVs including Deblur, oligotyping, and UNOISE2 (Amir et al. 2017). Recent studies have found that archaea and bacteria can have less or more than 10 genome copies per cell, making it hard to correct for ploidy (Soppa 2014).

5. Antimicrobial resistance

The use of antimicrobials in agriculture, human medicine, and veterinary medicine has become a driving force in the evolution of drug-resistant microorganisms (Blair et al. 2015). In the past antimicrobial agents have been overused to treat infections, this overuse has led to an increase of selected mutated bacteria that have resistant genes (Karakonstantis and Kalemaki 2019; Schwarz, Loeffler, and Kadlec 2017).

Over 100,000 people die every year in US hospitals due to untreatable infections. Transmissible airborne pathogens are a possible cause for patient readmittance and are difficult to control due to air movement in the facility. When confined to the hospital setting, community-associated methicillin-resistant S. aureus (CA-MRSA) has now emerged as an important pathogen in the medical community (Le et al. 2017). The main drivers of antibiotic resistance include the overuse of antibiotics, in humans and animals, and the advantage of over three billion years of existence. Microorganisms use different methods to resist antimicrobials such as through mutations or through acquiring outside DNA by horizontal gene transfer. When an antibiotic is used, the susceptible bacteria will die off creating a concentration of growing colonies of resistant species (Munita and Arias 2016). Microorganisms can use chemical or genotypic changes, a decrease in membrane penetration, an increase in efflux of drugs, and changes in target sites as mechanisms for drug resistance (Blair et al. 2015; Knopp and Andersson 2018). Streptococcus pneumoniae, a multidrug-resistant bacterium (Nguyen et al. 2019) is the leading cause of lower respiratory infection morbidity and mortality globally (Troeger et al. 2018). Multidrug-resistant strains that cause nosocomial infections can be highly concentrated, with Acinetobacter baumannii and Staphylococci; some causes of infections included mechanical ventilation and time (Li, Sun, et al. 2018; Kim, Kim, and Kim 2010).

Antibiotics, metals and pollutants may be the main drives for antimicrobial resistance (AR) (Warnes, Highmore, and Keevil 2012; Antonova and Hammer 2011; Hastings, Rosenberg, and Slack 2004), however, other environmental stressors may contribute to the development of resistance by a mechanism that is not well studied. The high amount of antibiotics released into the wastewater may promote the selection of antimicrobial-resistant bacteria which find their way into natural environments. Numerous studies focus on emerging microbial pathogens with multidrug resistance in municipal wastewater treatment plants (Li et al. 2016; Barancheshme and Munir 2018). From the bioaerosol and liquid samples collected at the plants, qPCR assays identified 44 antibiotic-resistant genes (ARGs) that confer resistance to a wide range of antibiotics (Gaviria-Figueroa et al. 2019). A comparison of the ARG profiles across samples showed that the downwind bioaerosol profile was 68% similar to the profile found in liquid sludge samples. Animal operations are another source of ARGs (Zhou et al. 2016), that were detected both within and 150 m downwind from swine production facilities (Gibbs et al. 2004; Gibbs et al. 2006), posing a threat to workers and residents nearby. Arfken, Song, and Sung (2015) found that fecal spray fields could be a source of antibiotic-resistant bacteria into the environment.

The widely applied Kirby Bauer disk diffusion method (Bauer et al. 1966) uses Mueller-Hinton agar (Oxoid, Hampshire, England) to analyze antibiotic resistance. Article discs impregnated with different antimicrobials are placed on inoculated plates and incubated overnight at 37 °C to assess the sensitive, intermediate and resistant cells based on the clearance of growth of the bacterial lawn around the discs (Bacto-Dickinson, England). Resistance data are interpreted according to National Committee for Clinical Laboratory Standards (NCCLS). For quantitative data, ETEST strips, which are calibrated with drug gradients, can be used to determine the Minimum Inhibitory Concentration (MIC) of the antibiotics (BioMerieux, Durham, NC, USA).

For the quantification of the genomic copy number (GCN) RT-PCR can be used. Stress from aerosol formation could lead to genome rearrangements or favor mutations that enhance antibiotic resistance. Furthermore, the selection of ARGs conferring resistance to β-lactams and aminoglycosides are often reported to be present on mobile genetic elements that are able to move within or between DNA strands, which could promote the acquisition and spread of resistance genes (Devarajan et al. 2015; Partridge et al. 2018). To identify molecular signatures for elevated antibiotic resistance, target genes that regulate multi-drug resistance functions can be analyzed from the collected aerosols (Griffith et al. 2018). Specific target genes may include marR and rfaC, which acquire two of the most prevalent alleles that affect the regulation of several genes involved in antibiotic resistance (Weiss et al. 2016). Alternate target genes include katE, sodC, and mdtM, which also have mutant alleles associated with enhanced resistance (Weiss et al. 2016). For each aerosol sample, DNA extraction can be performed with three replicate samples using standard methods (Zhou, Yang, and Jong 1990). Following isolation and purification, DNA from the replicates and time points for each aerosol sample can be prepared as barcoded libraries and sequenced on an Illumina HiSeq 2000. The sequence outputs can be analyzed using QIIME software to determine the allele distributions among the bacteria before and after aerosolization (Paulson et al. 2013). As an alternative approach, apparent resistant bacteria can be isolated from the Kirby-Bauer test plates and the presence of resistance alleles can be determined using PCR amplification of the target genes followed by standard Sanger sequencing.

6. Limitations, problems and perspectives

6.1. Bioaerosols

Most bioaerosol samples are a mixture of living, damaged, dead cells and inerts. Depending on environmental factors and collector devices, the ratio of living to the dead can be highly variable (Table 1). In non-viable bioaerosols with damaged cell membrane, the DNA is leaking out, unfolding and becoming hydrated and oxidized. However, in viable bioaerosols, the DNA often remains inside the cell, in a compact, folded form. The bioaerosols can also be non-viable, albeit with intact cell membranes; and viable, when the organism is metabolically active, but not culturable. As viability enhances a wide range of analysis, the airborne microorganisms should remain viable and culturable, maintaining intact cell wall and DNA integrity during collection. Preserving viability during collection may improve sensitivity of detection and identification assays and provide more information about agents of interest or concern. The increasingly wide-spread use of DNA-based molecular techniques requires the maintenance of intact cell wall to protect the integrity of the nucleic acids in the collected bioaerosols.

6.2. Bioaerosol detection

There is an increasing demand in the food processing industry for portable collector units with analytical capabilities to enumerate and identify the bioaerosol particles on-site, in real-time. Currently, the analysis of the samples requires a dedicated laboratory with trained personnel. Another challenge is the in-line, on-site collection and detection of nanosize bioaerosols. Viral epidemics including the recent Ebola infections necessitate the efficient sampling of large air volumes for the timely detection of a potential outbreak. However, a collector system coupled with a detector via microfluidics for the analysis of bioaerosols, even less for nanosize particles in room-size spaces is not readily available.

Acknowledgments

The authors acknowledge organizers of the special issue “Bioaerosol Research: Methods, Challenges, and Perspectives,” including Shanna Ratnesar-Shumate and Alex Huffman, based on requests from the Bioaerosol Working Group and after discussion during the Bioaerosol Standardization Workshop at the International Aerosol Conference in St Louis, Missouri in September 2018.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.