Here, we present a protocol to quantify allergen-loaded particles by flow cytometry. Ambient particulate matter particles may act as carriers of adsorbed allergens. We show here that flow cytometry, a method widely used to characterize suspended solids >0.5 µm in diameter, can be used to measure these allergen-loaded particles.
Flow cytometry is a method widely used to quantify suspended solids such as cells or bacteria in a size range from 0.5 to several tens of micrometers in diameter. In addition to a characterization of forward and sideward scatter properties, it enables the use of fluorescent labeled markers like antibodies to detect respective structures. Using indirect antibody staining, flow cytometry is employed here to quantify birch pollen allergen (precisely Bet v 1)-loaded particles of 0.5 to 10 µm in diameter in inhalable particulate matter (PM10, particle size ≤10 µm in diameter). PM10 particles may act as carriers of adsorbed allergens possibly transporting them to the lower respiratory tract, where they could trigger allergic reactions.
So far the allergen content of PM10 has been studied by means of enzyme linked immunosorbent assays (ELISAs) and scanning electron microscopy. ELISA measures the dissolved and not the particle-bound allergen. Compared to scanning electron microscopy, which can visualize allergen-loaded particles, flow cytometry may additionally quantify them. As allergen content of ambient air can deviate from birch pollen count, allergic symptoms might perhaps correlate better with allergen exposure than with pollen count. In conjunction with clinical data, the presented method offers the opportunity to test in future experiments whether allergic reactions to birch pollen antigens are associated with the Bet v 1 allergen content of PM10 particles >0.5 µm.
Air pollution is being considered as an important environmental cause of the increased incidence and severity of respiratory allergies observed in recent decades1-3. Moreover, there was growing interest in the distribution of common allergens in dust4,5.
Birch pollen can provoke hay fever but can also be an important trigger of allergic asthma6-8. Whole birch pollen is not likely to enter the lower respiratory tract or to be found in PM10 as a result of its size (22 µm in diameter). However, birch pollen allergens like Bet v 1, the major birch pollen allergen component, can be released after pollen rupture9 and can bind to ambient air particles10, thus possibly enter the lower respiratory airways. Indeed, it has been shown that PM10 may contain biologically active allergens as demonstrated by in vitro activation of basophiles from a pollen allergic proband11.
Bet v 1 allergen content in PM10 samples has been studied by extracting the respective allergen and subsequent quantification with ELISA12-14. With the ELISA technique, the dissolved allergen was measured, but the amount of allergen-loaded particles still remained unknown. Scanning electron microscopy revealed allergen-loaded particles but did not allow quantification10,15.
This study employs flow cytometry to quantify the proportion of Bet v 1-loaded PM10 particles in ambient air samples. Due to the detection limit of the flow cytometer only particles larger than 0.5 µm can be examined. The >0.5 µm fraction of PM10 will be further referred to as PM10>0.5.
NOTE: This protocol describes the indirect staining of PM10 particles with a monoclonal antibody (monoclonal mouse IgG1 antibody, clone MA-3B4) against Bet v 1, the major birch pollen antigen component, plus an Allophycocyanin (APC)-labeled secondary antibody (anti-Mouse IgG1 antibody, clone A85-1) and the subsequent analysis on a flow cytometer. With appropriate other antibodies available, this method might be extended to the detection of other antigens bound to ambient air particles.
1. PM10 Sampling
Figure 1. Low volume PM10 sampler. Example of a low volume PM10 sampler. Please click here to view a larger version of this figure.
2. PM10 Removal and Particle Count
Figure 2. PM10 removal with an electrical toothbrush. A polytetrafluoroethylene filter with sampled PM10 is placed in a polystyrol Petri dish (A) and is overlaid with 4 ml PBS. Then, PM10 is removed with an electrical toothbrush (B: before brushing and C: after brushing for 1 min). Please click here to view a larger version of this figure.
3. Bet v 1 Staining
4. Flow Cytometric Analysis
Bet v 1 allergen adsorption to PM10>0.5 particles was quantified by indirect antibody staining and subsequent analysis on a flow cytometer. A PM10 sample from high pollen season served as template. As stated in step 3.1, the negative control consisted of PM10 particles incubated with APC labeled secondary antibody only (Figure 3A). PM10 particles stained with anti-Bet v 1 antibody plus secondary antibody displayed the allergen loaded particles (Figure 3B). As described in step 4.3, two ways of quantifying the allergen load were used: On the one hand, the median value of the APC fluorescence intensity of all particles was analyzed being 137 for the negative control and 904 for the specific sample. On the other hand, the percentage of particles with bound anti-Bet v 1 antibody was determined: A gate was set around the APC positive particles in the negative control and subsequently copied and pasted into the specific sample. In the negative control, 3% of the PM10>0.5 particles were considered APC positive. This percentage of false positive particles was subtracted from the percentage of positive particles in the specific sample thus resulting in 77.5% APC positive PM10>0.5 particles in the specific sample. To prove that the observed binding of the anti-Bet v 1 antibody was specific, binding capacity was blocked with the corresponding antigen prior to staining. This diminished binding of the anti-Bet v 1 antibody by 69%, if quantified by APC fluorescence intensity, and by 84%, if quantified by percentage of Bet v 1 positive PM10>0.5 particles (Figure 3C).
Figure 3. Particle-bound Bet v 1 allergen can be visualized by flow cytometry. APC fluorescence intensity of a PM10 sample from high pollen season stained only with the APC labeled secondary antibody (A), stained with the anti-Bet v 1 primary antibody and subsequently with the APC labeled secondary antibody (B), and after blocking the primary antibody with recombinant Bet v 1 antigen (C). The gate PAPC+ was set around the particles considered APC positive (displayed in red) and respective percentages are given. This figure has been slightly modified from11. Please click here to view a larger version of this figure.
To test whether this method could be adopted to reveal differences in the amount of adsorbed Bet v 1 content of PM10>0.5 samples from high and from low pollen season, 13 PM10 samples from high and 6 PM10 samples from low pollen season were analyzed. Figure 4 depicts significant differences in the APC fluorescence intensity and in the proportion of Bet v 1 positive PM10>0.5 particles of PM10 samples from high pollen season compared to PM10 samples from low pollen season. Both quantification methods hereby showed similar results.
Figure 4. PM10 samples from low and high pollen season differ in their amount of adsorbed Bet v 1. Low pollen season PM10 was sampled in autumn/winter 2013 (n=6), high pollen season PM10 in May 2012 and 2013 (n=13). (A) The APC fluorescence intensity of PM10>0.5 particles from high pollen season was significantly higher than from low pollen season (median/min/max high pollen season: 796/313/1097; median/min/max low pollen season: 197/85/277). (B) PM10 from high pollen season contained significantly more Bet v 1-positive PM10>0.5 particles than PM10 from low pollen season (median/min/max high pollen season: 45.2/18.5/74.5; median/min/max low pollen season: 11.8/4.4/19.8). Box plots show median values (inner line of the box), 25. and 75. percentiles, respectively (lower and upper borders of the box) and minimum and maximum values (whiskers). ***p<0.001 versus low pollen season, Mann Whitney U test. This figure has been slightly modified from11. Please click here to view a larger version of this figure.
A critical step of the protocol is the use of an appropriate filter for the collection of PM10 particles from ambient air (see step 1.1). The filter has to be strong enough to endure brushing with an electrical toothbrush, and not all filter materials fulfill this requirement. The staining protocol was established with a PM10 particle concentration of 8×106 particles per ml. However if the material is limited and pooling of samples is not appropriate, the method will probably function as well, but the antibody concentrations (see steps 3.5 and 3.8) might have to be adjusted.
Bet v 1 staining of PM10 particles did not result in distinct populations of positively and negatively stained particles. This might be caused by the varying amounts of Bet v 1 allergen adsorbed to each of the particles ranging from very little up to a high amount. This could result in the expansion of the APC signal thus shifting the population towards APC positivity. As it is difficult to separate the positive from the negative particles, two quantification methods were used to determine the differences in the Bet v 1 content of the PM10>0.5 specimens: (i) relative quantification by measuring the median APC fluorescence intensity of all particles, and (ii) determining the percentage of APC positive particles. Regarding the Bet v 1 load of particles from low and high pollen season PM10>0.5, both methods revealed similar results. Still, relative quantification by median fluorescence intensity of all particles is recommended as it is independent of placing the gate and therefore probably less error-prone.
To date many studies examine the allergen content in ambient air particulate matter by extracting the respective allergen and subsequent quantification with ELISA5,12-14,17. There is a fundamental difference between the procedure described here and the quantification with ELISA: ELISA quantifies the extracted and dissolved antigen, while flow cytometry analyzes the particle-bound antigen. By means of ELISA the Bet v 1 load of the tested PM10 samples (n=8) was below the detection limit of 1.2 ng/ml (data not shown). Similarly, Buters and others identified no Bet v 1 in the PM <2.5 µm fraction and only about 7% in the 10 µm >PM >2.5 µm fraction, but more than 93% in the PM >10 µm fraction of ambient air13. The contrasting results of the ELISA on the one hand and the FACS analysis on the other hand, may be caused by differences in the detection method in conjunction with divergent sensitivity. Further research however is needed to fully understand this difference.
A method to visualize particle-bound antigen is scanning electron microscopy10,14. By scanning electron microscopy, Ormstad et al. visualized Bet v 1 on the surface of suspended particulate matter soot particles sampled in the high pollen season and to a lesser extent on particles sampled in the low pollen season15. Additionally, allergens from pollen, latex and also β-glucans were found to be adsorbed to combustion particles in ambient air10. This method, however, does not allow quantification of the particle-bound allergen.
By use of flow cytometry, particle-bound Bet v 1 allergen could be quantified. Thus, flow cytometry may offer a new way to characterize the 10 to 0.5 µm biological fraction of PM10 as with other suitable antibodies on hand, this method might be extended to the detection of other antigens on ambient air particles, e.g., mold, dust mite allergens or LPS. As PM10 particles adsorb not only biological material, but also chemicals and metals quite easily, unspecific binding of antibodies could, however, pose a problem. If a new antibody is tested, a critical step is to prove specific binding. This can be done by, for example, blocking the binding capacity of the specific antibody with the corresponding antigen prior to staining11.
As Bet v 1 content of ambient air can differ from birch pollen count12,13,18, allergic symptoms might perhaps correlate better with allergen level than with pollen count14,18. Hence, the presented method in conjunction with clinical data enables to examine in future experiments whether allergic reactions to birch correspond to the Bet v 1 allergen load of PM10>0.5.
The authors have nothing to disclose.
The authors would like to thank Katrin Bossmann, Anett Neumann and Eike Wolter (German Environment Agency) for their valuable preparatory work.
Teflon filter | Pall Life Sciences, USA | R2PL047 | 47 mm, 1.0 µm |
low volume sampler | Sven Leckel Ingenieur Büro GmbH, Germany | LVS3 | air flow of 2.3 m3/h |
Phosphate-buffered saline | Biochrom, Germany | L1825 | without Ca/Mg, low endotoxin |
electrical toothbrush | Braun, Germany | Oral-B Vitality Sensitive | |
Casy cell counter | Schärfe System GmbH, Germany | Model TTC | range of detectable particle size: 0.7 µm to 45 µm |
FACSCanto II | Becton Dickinson, USA | 3-laser, 8-color (4-2-2) | |
FACS Diva Software v6.1.3 | Becton Dickinson | ||
bovine serum albumin (BSA) | Sigma-Aldrich, USA | A2153-10G | |
monoclonal mouse IgG1 antibody against Bet v 1 | Indoor Biotechnologies, UK | MA-3B4 | clone MA-3B4 |
APC (Allophycocyanin)-labeled secondary anti-Mouse IgG1 antibody | Becton Dickinson | 560089 | clone A85-1 |
SPSSTM software version 18 | PASW Statistics 18, Hongkong, China | ||
Petri Dish | Gosselin, France | BP50-02 | D 55mm, H 15mm |
FACS Tube | Becton Dickinson, USA | REF 352054 | 5ml Polystyrene |
CASYton | Roche Germany |
REF 05651808001 | |
Matrix Blank Tubes | Thermo Scientific, USA | 4140 | 1,4 ml, PP |
Centrifuge | Heraeus, Thermo Scientific | Megafuge 40R | |
Vacuum Pump | INTEGRA Biosciences AG, Switzerland | Model 158 320 | Inetrgra Vacusafe |
recombinant Bet v 1a antigen | Indoor Biotechnologies, UK | LTR-BV1A-1 | Concentration: 2.0 mg/ml |