Science of the Total Environment 788 (2021) 147789 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv A rapid test for the environmental detection of pigeon antigen Silvia Sánchez-Díez a,b, María-Jesús Cruz a,b,⁎, Daniel Álvarez-Simón a, Tomás Montalvo c,d, Xavier Muñoz a,b,e, Peter M. Hoet f, Jeroen A. Vanoirbeek f, Susana Gómez-Ollés a,b a Servicio de Neumología, Departamento de Medicina, Hospital Universitario Vall d'Hebron, Universidad Autónoma de Barcelona, Cataluña, Spain b CIBER Enfermedades Respiratorias (Ciberes), Madrid, Spain c Servicio de Vigilancia y Control de Plagas Urbanas, Agencia de Salud Pública de Barcelona, Spain d CIBER de Epidemiología y Salud Pública (Ciberesp), Madrid, Spain e Departamento de Biología Celular, Fisiología e Inmunología, Universidad Autónoma de Barcelona, Cataluña, Spain f Department of Public Health and Primary Care, Centre for Environment and Health, KU Leuven, Leuven, Belgium H I G H L I G H T S G R A P H I C A L A B S T R A C T • Demonstrating antigen exposure is crit- ical in the diagnosis and management of hypersensitivity pneumonitis. • An immunochromatographic test is compared with an ELISA method for the measurement of pigeon antigens in the environment. • The immunochromatographic test de- scribed was highly consistent with the validated ELISA and does not require specific skills. • This method may prove to be a useful tool for predicting the progression of bird- related hypersensitivity pneumonitis. a r t i c l e i n f o a b s t r a c t Article history: Introduction: Avoidance of inhaled bird antigens is essential to prevent hypersensitivity pneumonitis disease pro- Received 11 March 2021 gression. The aim of the present study was to develop a sandwich enzyme link immunoassay (ELISA) and an Received in revised form 11 May 2021 immunochromatographic test (ICT) and compare their ability to detect pigeon antigens in environmental samples. Accepted 11 May 2021 Methods: An amplified sandwich ELISA using pigeon serum as a calibration standard and a ICT using gold-labeled Available online 15 May 2021 anti-pigeon serum antibodies for the rapid detection of pigeon antigens in environmental samples were developed. Editor: Lotfi Aleya Twenty-two different airborne samples were collected and analysed using both methods. Strip density values ob- tained with ICT were calculated and compared with the concentrations determined by the ELISA method for pigeon antigens. Strips results were also visually analysed by five independent evaluators. Keywords: Results: The ELISA method to quantify pigeon antigen had a broader range (58.4 and 10,112.2 ng/ml), compared to Hypersensitivity pneumonitis the ICT assay (420 to 3360 ng/ml). A kappa index of 0.736 (p < 0.0001) was obtained between the observers eval- Immunochromatographic test uating the ICT strips. The results of the ELISA and the relative density of the ICT showed a highly significant correla- Pigeon antigen tion (rs:0.935; p < 0.0001). Bland-Altman plot also confirmed excellent agreement between the two methods Strip assay (mean difference: −1.626; p < 0.0001). Conclusions: Since there was a good correlation between both assays, we can conclude that the rapid and simple ICT assay is a good and valid alternative, which does not require expensive equipment, for the validated ELISA technique. © 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). ⁎ Corresponding author at: Servicio de Neumología, Hospital Universitario Vall d'Hebron, Passeig Vall d'Hebron, 119, 08035 Barcelona, Spain. E-mail address: mj.cruz@vhir.org (M.-J. Cruz). https://doi.org/10.1016/j.scitotenv.2021.147789 0048-9697/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). S. Sánchez-Díez, M.-J. Cruz, D. Álvarez-Simón et al. Science of the Total Environment 788 (2021) 147789 1. Introduction concentration of 15.5 mg/ml determined previously by the Bicinchoninic acid (BCA) method (Pierce Chemical Co., Rockford, USA), was subcutane- Hypersensitivity pneumonitis (HP) is an interstitial lung disease char- ous injected into two New Zealand white rabbits. Three booster injections acterized by inflammation in bronchioles and alveoli (Selman, 2004). It oc- were administered at 4-week intervals. The use of the animals for this curs in some genetically predisposed individuals after recurrent inhalation purpose was approved by the firm's Animal Ethics Committee. of particular environmental antigens, usually avian and fungal proteins. The IgG fraction of the polyclonal antisera of both rabbits was iso- HP is characterized by fever, dyspnea, chest tightness/discomfort and/or lated on immobilized Protein A–agarose columns (Pierce Chemical Co., cough (Quirce et al., 2016), and it has recently been classified into two dif- Rockford, USA), and then eluted to Excellulose columns (Pierce Chemi- ferent forms: an acute form characterized by inflammation and a chronic cal Co., Rockford, USA) to desalt and exchange the buffer to phosphate form characterized by fibrosis, according to the intensity and frequency of buffer saline (PBS). Purified and desalt IgG fractions from each rabbit antigen exposure, clinical evaluation, pulmonary function tests and radio- were named Pol J and Pol K. The protein concentration of the IgG frac- logic findings (Lacasse et al., 2009; Vasakova et al., 2017). tion of the sera was 7.7 mg/ml (Pol J) and 6.4 mg/ml (Pol K), as deter- Bird-related hypersensitivity pneumonitis (BRHP) occurs after inha- mined by the BCA method. lation exposure to avian antigens. It can be caused by either direct expo- A portion of rabbit anti-pigeon polyclonal antibody from one rabbit sure, i.e., through breeding birds, or indirect exposure, i.e., due to the (Pol J) was biotinylated using the EZ links Sulfo-NHS-LC-Biotinylation presence of nearby dovecotes, flocks of pigeons in parks, feather duvets, Kit (Pierce Chemical Co.) according to the manufacturer's instructions. and so on (Shirai et al., 2017; Morell et al., 2008). It has been demon- Another portion of Pol J was labeled with colloidal gold by BBI Solutions strated that the most common antigenic sources are droppings, feathers (British Biocell International, Cardiff, UK). Briefly, the antibody was con- and blooms (Vasakova et al., 2019; Raghu et al., 2020). Next to pigeon jugated to a 40 nm gold colloid using passive adsorption and was stable antigens, other types of avian antigen such as those associated with at least for 10 months. Then, the conjugate was blocked with bovine parrots, parakeets, chickens, ducks, etc., can cause BRHP (Kokkarinen serum albumin (BSA) (Pierce Chemical Co., Rockford, USA) and the et al., 1994). Currently, it seems that the prevalence of BRHP is resulting conjugate was concentrated by ultrafiltration to give a final underestimated. A Japanese epidemiological study conducted between optical density (OD) (520 nm) of 10 and suspended in a 2 mM disodium 2000 and 2009 revealed that 60% of chronic cases of HP were caused tetraborate buffer pH 7.2, containing 0.095% sodium azide. by avian antigens (Okamoto et al., 2013), and it has been shown that, in some patients, this disease can be caused by exposure to feather du- 2.2. Sample collection and analysis vets or pillows (Morell et al., 2013). For patients with BRHP, complete avoidance of inhaled avian antigens Twenty-two air samples were collected from two different dove- is essential to improve the respiratory symptoms and prevent disease cotes in the province of Barcelona with a medium-volume automated progression (Vourlekis et al., 2004). Some BRHP cases possibly become air sampler (Flite 3, Vertex technics S.L, Barcelona, Spain) containing ni- chronic because of the persistence of their exposure to low levels of anti- trocellulose filters with a 0.65 μm pore size (Millipore S.A.S., Molsheim, gen (Borderías et al., 2010). To control exposure to avian antigen levels in France). The automated air sampler had a pole 115 cm high with a sup- the environment, air measurements are essential for these patients port at the top for the nitrocellulose filter. Dust was collected at a flow of (Agache and Rogozea, 2013). Previous studies have measured the con- 12 l/min during 1 h, 2 h, 4 h, 8 h, 12 h or 24 h. Air samples of environ- centration of avian antigens in collected dust and air samples by using di- ments without avian antigens were also collected at the same flow for rect, inhibition or sandwich enzyme-linked immunoassays (ELISA) (Craig 1 h, 2 h, 4 h or 8 h, for use as negative controls. Antigens were extracted et al., 1992; Tsutsui et al., 2015), while others used an antigen-capture from the filters in 1.5 ml PBS/0.2% BSA/0.5% Tween-20 (pH 7.4) buffer ELISA with signal amplification in order to improve sensitivity and detect overnight at 4 °C. Filters were discarded and the eluates were stored small amounts of antigen (Kuramochi et al., 2010). All these techniques at −20 °C until being analysed in parallel by the anti-pigeon sandwich are costly and time-consuming and require access to laboratories with so- ELISA and the ICT. phisticated equipment and trained personnel, and so they are not readily adaptable to field use. Therefore, considerable efforts are being made to 2.3. Characterization of pigeon serum and air samples by 1D and 2D develop easy-to-use analytical methods that are able to guide legislation electrophoresis on the concentration of these proteins in the air, and may also allow con- sumers to monitor their levels. For 1D electrophoresis performance, Laemmli sample buffer 4× The immunochromatographic test (ICT) is a rapid semiquantitative (Bio-Rad, Madrid, Spain) was diluted in 2-mercaptoethanol (Sigma-Al- allergen estimation method, easy to perform and interpret, portable, drich Quimica S.L, Madrid, Spain) according to the manufacturer's instruc- and cheap. In this test, a sample with labeled specific antibodies is tions. Pigeon serum (850 μg protein) or air sample (60 ng protein) were mixed and incubated with a strip. Then, the immunocomplexes formed mixed with the previous mixture at a 3:1 proportion and heated for migrate by capillarity through the nitrocellulose membrane and bind to 5 min at 90 °C in the water bath. Subsequently, samples were loaded the line of immobilized specific capture antibodies if the sample con- onto preparative Criterion TGX Stain-Free gels 4–20% precast polyacryl- tains the specific antigen (Bogdanovic et al., 2006; Tsay et al., 2002; amide gel (Bio-Rad, Madrid, Spain) according to the manufacturer's man- Álvarez-Simon et al., 2014). ual, and electrophoresis was performed at 200 V for 40 min. The aim of the present study was to develop and validate a sensitive For 2D electrophoresis, pigeon serum (100 μg protein) was mixed sandwich enzyme link immunoassay technique and a rapid ICT to detect with rehydration solution (pure H2O/8.5 M Urea/2% Chaps/0.5% pigeon antigens in environmental air samples and to determine immobilized pH gradient (IPG) buffer/0.002% Bromophenol blue) and whether the results obtained with the two methods were comparable. loaded onto IPG strips (IPG strip, pH 3–10, 7 cm, General Electric Healthcare, Boston, USA). Isoelectric focusing (IEF) was then carried out 2. Material and methods using the IPGphor IEF system (General Electric Healthcare, Boston, USA) following the manufacturer's instructions. IPG strips were then equili- 2.1. Preparation of rabbit anti-pigeon polyclonal antibodies and colloidal brated and loaded onto Mini-PROTEAN TGX Stain-Free gels 4–20% gold labeling (7 cm, IPG/prep, Bio-Rad, Madrid, Spain) and run at 200 V for 30 min. In both cases, Precision Plus Protein™ Unstained Standards Anti-pigeon polyclonal antibodies were produced by Kaneka (Bio-Rad, Madrid, Spain) were used as molecular weight (MW) markers Eurogentec S.A., Seraing, Belgium. In Brief, a commercial pigeon serum with a MW ranging from 10 to 250 kDa. After 1D and 2D electrophore- (Rockland Immunochemicals Inc., Limerick, Ireland), with a protein sis, gels were imaged in a Gel Doc EZ Imager (Bio-Rad, Madrid, Spain) 2 S. Sánchez-Díez, M.-J. Cruz, D. Álvarez-Simón et al. Science of the Total Environment 788 (2021) 147789 and proteins were stained with Coomassie brilliant blue (Bio-Rad, by the BCA method (Pierce Chemical Co., Rockford, USA). Characteriza- Madrid, Spain) or blotted to a membrane. tion of the different avian sera was performed by 1D and 2D electropho- resis following the protocol described above. 2.4. Immunoblotting of pigeon serum 2.7. Immunochromatographic strip test 1D and 2D gels of pigeon serum were blotted onto polyvinylidene difluoride (PVDF) membranes using the Trans-Blot Turbo transfer sys- The capture antibody Pol K (960 ng/strip) was applied to direct cast tem (Bio-Rad, Madrid, Spain). Then, membranes were washed with backed nitrocellulose nitrate membranes (Unisart 1UN95E CN95, PBS and blocked with 5% nonfat dry milk (Bio-Rad, Madrid, Spain) for Sartorius Stedim Biotech, Goettingen, Germany) in a line format at 1 h at room temperature (RT) under gentle shaking. After three washing 3 mm from the application end of the strip (Test line). At 6 mm, a steps with PBS/0.05 T (PBS with 0.05% Tween-20), membranes were in- goat antirabbit IgG antibody (Southern Biotechnology Associates, Inc., cubated with rabbit anti-pigeon polyclonal antibody Pol J or Pol K Birmingham, USA) was bound at 600 ng/strip to provide a positive control diluted 1/5000 in PBS for 2 h at RT with gentle shaking. After three (Control line). Membranes were dried for 1 h at 37 °C and assembled as washing steps with PBS/0.05 T, membranes were incubated with goat follows: on an adhesive support, an absorbent pad was layered and the anti-rabbit IgG-HRP (H + L, SouthernBiotech, Birmingham, USA) di- membrane was placed overlapping it by 2 mm and cut into individual luted 1/20,000 in PBS for 1 h at RT. Membranes were then washed strips of 4 mm in width. three times with PBS/0.05 T and incubated with the enzyme substrate After optimization, 20 μl of environmental samples or pigeon serum for 5 min at RT (Kit Clarity Western ECL Substrate, Bio-Rad, Madrid, standards were mixed 1:1 in a microtiter plate well with gold-labeled Spain), according to the supplier's recommendations. Image acquisition anti-pigeon Pol J antibody solution with an OD of 4. After forward pipet- was carried out with the Odyssey Fc Imaging System and the Image Stu- ting, the gold-labeled Pol J antibody was allowed to react during 10 min dio Lite software (LI-COR, Nebraska, USA). at RT with pigeon antigens of the samples to form a complex. Then, a test strip was dipped in each well for 30 min at RT to allow antigen- 2.5. Sandwich ELISA for pigeon antigens antibody complexes to diffuse across the nitrocellulose membrane. These complexes reacted with the specific anti-pigeon Pol K antibodies Wells of high-binding microtiter plates (Costar, Cambridge, USA) immobilized on the test line forming a pink-purple line. The excess of were coated overnight at 4 °C with Pol K (0.1 μg protein/well) in PBS colloidal dye antibody conjugate migrated further and reacted with buffer (pH 7.4). The wells were blocked (200 μl/well) with PBS/1% the goat anti-rabbit IgG antibody in the control line. BSA/0.05% Tween-20 for 1 h at 37 °C. After washing three times with Strips were scanned with a scanner (HP Scanjet G4050, Hewlett- PBS/0.1 T (PBS with 0.1% Tween 20), plates were incubated for 1 h at Packard Enterprise, California, USA) and test line intensities were 37 °C with 100 μl/well of filter elutes, negative controls (eluate of a analysed by densitometry with ImageJ 1.52a (National Institutes of white filter) and pigeon antigen standards containing 6.7–27,500 ng Health, USA). To determine the limit of detection of the strip assay, pi- protein/ml. After three washes with PBS/0.1 T, plates were incubated geon serum (Rockland Immunochemicals Inc., Limerick, Ireland) with for 1 h at 37 °C with 100 μl/well of Pol J biotinylated at 1:15,000 dilution a protein concentration of 27.5 mg/ml, previously determined by the in PBS/1% BSA/0.05% Tween-20. Streptavidin (Sigma-Aldrich Quimica S. BCA method, was used in a series of concentrations from 52.5 to 6720 L, Madrid, Spain) at 1/10,000 dilution in PBS/1% BSA/0.05% Tween-20 ng/ml and tested in triplicate. In parallel, test lines of environmental was added (100 μl/well) after washing again three times with PBS/0.1 samples were also read with the naked eye by five independent ob- T. After streptavidin incubation for 1 h at 37 °C, wells were washed servers who classified the strips as negative (52.5 ng/ml), positive three times and the reaction was developed with 30% H2O2 and (420 ng/ml) and double positive (3360 ng/ml) comparing their color in- 3,3′,5,5′-tetramethylbenzidine in acetate buffer (pH 5.5) at 25 °C in tensity with a 3-point standard curve (Fig. 1). the dark. After 20 min, the reaction was stopped with 2 M H2SO4 and read at 450 nm with a plate reader (Infinite M Nano, Tecan, Männedorf, 2.8. Statistical analysis Switzerland). The standard curve was constructed from seven data points using a four-parameter logistic curve fit. Results are expressed The Shapiro-Wilk normality test was calculated for pigeon in ng/ml, referring to the protein content of the standard preparation. aeroallergen levels measured by the ELISA method and for relative den- Magellan software (Tecan Austria GmbH, Grödig, Austria) was used to sity determined from immunochromatographic strips. Spearman's rank calculate the results. correlation coefficient (rs) was calculated between pigeon aeroallergen The lower limit of detection (LOD) was defined as the concentration of levels and strip density values since all data showed a non-normal dis- the lowest point on the standard curve, with the optical density (OD) tribution. Bland-Altman analysis was performed to determine the value at this point having to be at least 0.05 above the blank. The lower agreement between the two assays. After logit transformation of the limit of quantification (LOQ) was defined, according to the curve fitting data, the difference between the paired measurements was plotted function, as the concentration that gave the OD of the lowest point on against the mean of the measurements. The immunochromatographic the standard curve plus 20% of the highest point on the standard curve. strip results were visually analysed by five independent evaluators, The upper LOQ was defined as the concentration that gave 80% of the and Fleiss' Kappa test was used to calculate the degree of agreement be- OD at the highest point on the standard curve. Following these premises, tween the different raters. the determinations were largely limited to the log-linear part of the curve. Statistical analyses were performed using GraphPad Prism 6 for The inter-assay coefficient of variation (CV) was determined by Windows (version 6.01, GraphPad Software Inc., San Diego, California, analysing two points of the standard curve in duplicate on 10 different USA) and IBM SPSS Statistics (version 26, IBM Corporation, Armonk, days. Thus, the CV of each point was calculated as the standard deviation New York, USA). Differences with a p-value <0.05 (two-tailed) were of the mean of 10 measurements in duplicate (SD/mean x 100). considered to be significant. 2.6. Specificity of the ELISA 3. Results Cross-reactivity of the ELISA assay towards other types of bird pro- 3.1. Characterization of pigeon serum and immunoblotting tein was tested using serum of small parrot, parakeet and parrot. Sera from those birds were obtained by centrifugation of collected blood Fig. 2A shows the different protein bands separated by 1D electro- samples. Subsequently, protein concentration of sera was determined phoresis from pigeon serum (Fig. 2A, lane 2) and one of the air sample 3 S. Sánchez-Díez, M.-J. Cruz, D. Álvarez-Simón et al. Science of the Total Environment 788 (2021) 147789 C T Flow (Fig. 2A, lane 3) collected from the dovecotes. Different bands were ob- served with molecular weights ranging from 10 to 250 KDa. Some bands - were found in common between the serum and the air sample, mostly between 25 and 65 kDa. Pixel value sum Fig. 2B (lane 2, 3 and 4) shows different spots of proteins from pi- geon serum separated by molecular weight and isoelectric point in two dimensions. The observed proteins were mostly in the range of 20 to 25 KDa and 50 to 75 KDa. C Immunoblotting in one and two dimensions showed that the poly- (52.5 ng/ml) T clonal antibodies Pol K and Pol J were able to detect a wide range of pro- Distance teins from pigeon serum (Fig. 2A, lane 4 and 5, and 2 B, lane 3 and 4) and + from the air sample (Fig. 2A, line 6 and 7). Serum from a patient diag- Pixel value sum nosed with BRHP due to pigeon exposure showed a similar pattern of recognition to those of the polyclonal antibodies Pol K and Pol J (Fig. 2A, line 8 and 9). C 3.2. Characterization of the ELISA method (420 ng/ml) T The sandwich ELISA against pigeon proteins was able to detect a Distance range of concentrations between 6.7 and 27,500 ng/ml in the standard ++ curve. However, the lower and upper LOQ determined the working Pixel value sum range of concentrations between 58.4 and 10,112.2 ng/ml. The standard curve was formed by seven different concentrations of pigeon serum proteins that followed a sigmoidal shape (Fig. 3A. The inter-CV using two different points of the standard curve was 2.76% and 7.35% respec- C tively. In the negative controls the antigen concentration was undetect- (3360 ng/ml) T able. As regards the specificity of the method, the sandwich ELISA Distance showed some cross-reactivity with small parrot, parakeet and parrot. Fifty percent of the OD of the highest point on the pigeon standard Fig. 1. Three-point standard curve of pigeon serum proteins to evaluate the test line of the curve was reached at protein concentrations of 54,344, 74,806 and environmental strips as negative (52.5 ng/ml), positive (420 ng/ml) or double positive 172,175 ng/ml for small parrot, parakeet and parrot respectively (3360 ng/ml). The figure shows the scanning of each standard curve strip and their (Fig. 3A). One and two-dimensional electrophoresis showed that these densitometry analyses. C, control line; T, test line. birds had some protein bands/spots of similar size and/or charge with pigeon that might explain this cross-reactivity (Fig. 3B). A 1 2 3 4 5 6 7 8 9 B 1 2 3 4 pH 3 pH 10 pH 3 pH 10 pH 3 pH 10 Fig. 2. Protein composition of pigeon serum and an air sample eluate collected from the dovecote by one and two dimensional electrophoresis and Western blots with antibodies Pol K and Pol J. 2A) Lane 1, a molecular weight size marker in KDa; lane 2, Coomassie-stained 1D gel of pigeon serum; lane 3, Coomassie-stained 1D gel of an 8 h air sample; lane 4 and 5, immunoblots with Pol K and Pol J against pigeon serum, respectively; lane 6 and 7, 8 h air sample immunoblot with Pol K and Pol J, respectively; lane 8 and 9, pigeon immunoblot and air sample immunoblot, respectively, with serum of a patient with BRHP due to pigeon exposure. 2B) Lane 1, molecular weight marker in KDa; lane 2, Coomassie-stained 2D gel of pigeon serum; lane 3 and 4, Pol K and Pol J immunoblots against pigeon serum, respectively. 4 B Small parrot MW 1D MW 2D S. Sánchez-Díez, M.-J. Cruz, D. Álvarez-Simón et al. A 1 .2 Parakeet 1 .0 MW 1D MW 2D 0 .8 Pigeon Small parrot 5 0 .6 Parakeet Parrot OD (450 nm) 0 .4 0 .2 Parrot MW 1D MW 2D 0 .0 10 0 10 1 10 2 10 3 10 4 10 5 10 6 10 7 Protein concentration (ng/ml) Fig. 3. Protein composition of bird sera (small parrot, parakeet and parrot serum) by one and two-dimensional electrophoresis (3A) and cross-reactivity of these sera with a sandwich ELISA for pigeon antigen detection (3B). Science of the Total Environment 788 (2021) 147789 S. Sánchez-Díez, M.-J. Cruz, D. Álvarez-Simón et al. Science of the Total Environment 788 (2021) 147789 0 52,5 105 210 420 840 1680 3360 6720 ng/ml A 3.5. Comparison between ELISA and immunochromatographic strip assay Test line 䊻 A strong correlation was observed between the concentration deter- mined by the ELISA method and the relative density results of the strip B Standard curve assay with a rs coefficient of 0.935 (CI 0.843–0.974; p < 0.0001) 12000 (Fig. 5A). The Bland-Altman plot also confirmed a good agreement be- Relative density (Test line) tween the methods with a mean difference of −1.626 (CI -1.988 to 10000 −1.264; p < 0.0001) (Fig. 5B). 8000 4. Discussion 6000 We developed and validated two sensitive methods to detect pigeon antigens in environmental air samples: an ELISA and an 4000 immunochromatographic strip assay. Although other studies describing the quantification of avian antigens have been reported (Craig et al., 2000 1992; Tsutsui et al., 2015; Kuramochi et al., 2010), to our knowledge, this is the first in which an ICT is presented and compared with an 0 ELISA method for the measurement of pigeon antigens in the environ- Blank 2 3 4 ment. The study demonstrates that the methods have enough sensitiv- Log [Concentration (ng/ml)] ity, specificity, reproducibility, and applicability for the detection of environmental exposure to pigeon antigens. Fig. 4. Immunochromatographic strip assay standard curve to determine the lower and Chronic HP is difficult to identify and is often misdiagnosed as idio- upper limits of detection. 4A) Scanning of the test line of each standard curve strip. 4B) pathic pulmonary fibrosis (Raghu et al., 2020; Morell et al., 2013). Graph of relative density results (mean ± SD) at each concentration of pigeon serum Therefore, demonstrating antigen exposure and evaluating the effects proteins using a four parameter logistic curve. of allergen avoidance measures are critical in the diagnosis and manage- ment of this disease. Tsutsui et al. reported that the concentration of avian antigen present in household dust correlated with the prognosis 3.3. Characterization of the immunochromatographic strip assay of chronic bird-related HP (Tsutsui et al., 2015). Chronic HP has been as- sumed to progress as a result of persistent exposure to very small The rapid test was able to detect a range of concentrations from 420 amounts of antigen. Moreover, avian antigens were reported to persist to 3360 ng/ml (Fig. 4). The lower LOD (420 ng/ml) was considered as in the patient's house six months after removal of all birds (Morell the minimum amount of pigeon serum proteins from the standard et al., 2008; Craig et al., 1992). Therefore, it is important to examine curve giving a well-defined pink-purple band at the test line (Fig. 4A). the environmental levels of antigen after diagnosis and during antigen The upper LOD (3360 ng/ml) was considered as the maximum amount avoidance in order to facilitate the control of this disease. of pigeon serum proteins producing a more intense pink-purple band at Environmental monitoring of airborne antigens is a time-consuming the test line compared to the previous concentration (Fig. 4A). In the process that commonly comprises three steps: sampling aeroallergens negative controls the antigen concentration was undetectable. on a filter, elution of the allergens, and determination of allergen levels by an allergen-specific ELISA (Gómez-Ollés et al., 2006). Here, we first obtained polyclonal antibodies with high specific reactivity to pigeon 3.4. Inter-reader agreement of immunochromatographic strip assay results serum antigens and developed a validated ELISA method using those antibodies. Second, we developed and validated a novel ICT using the According to the observers' interpretations of the strips as negative, same antibodies used in the ELISA method. Previous studies have pro- positive or double positive, a kappa index of 0.574 (CI 0.479–0.668; p < posed that ICT may be a valid alternative to the ELISA, as it speeds up 0.0001) was obtained. However, when considering positive and double the last step of the process (Bogdanovic et al., 2006; Tsay et al., 2002; positive results as a single category the kappa index increased (0.736, CI Álvarez-Simon et al., 2014). 0.605–0.867, p < 0.0001). The observers were able to detect visually pi- The antigenic components associated with the development of BRHP geon allergens in a concentration range of 141 to 3360 ng/ml in air sam- was described using biochemical analyses and immunoassays. The main ple eluates. antigenic substances suspected were immunoglobulins, serum albumin, A B Correlation Bland-Altman plot Difference (log(ELISA), log(ICT)) 20000 Relative density (Test line) -1.0 15000 Upper limit r s =0.9345 -1.5 10000 IC (0.8425-0.9735) Mean p<0.0001 5000 -2.0 Lower limit 0 0 100 200 300 400 2.0 2.5 3.0 3.5 ELISA concentration (ng/ml) Average (log(ELISA), log(ICT)) Fig. 5. Comparison of pigeon aeroallergen concentration determined by ELISA method and density values calculated with the immunochromatographic strip assay. 5A) Spearman's rank correlation. 5B) Bland-Altman plot. 6 S. Sánchez-Díez, M.-J. Cruz, D. Álvarez-Simón et al. Science of the Total Environment 788 (2021) 147789 and intestinal mucin, from droppings, blooms and pigeon sera (Shirai et al. (2002) graded the line intensity of some samples analysed as me- et al., 2017; de Beer et al., 1990; Rouzet et al., 2017). Owing to the diffi- dium or high even although the allergen levels analysed by ELISA were culty of obtaining antigenic extracts from intestinal mucin or bloom, pi- below the theoretical LOD of the rapid assay. Similarly, in the studies by geon serum is the antigen source most often used to measure specific Bogdanovic et al. (2006) and Álvarez-Simon et al. (2014), some samples antibodies in this type of HP (Rodrigo et al., 2000). The immunoblotting with allergen levels below the sensitivity of the rapid test were classi- results of the present study demonstrated that the polyclonal antibodies fied as positive. Moreover, visual interpretation of rapid assays is inher- used in the methods developed were able to detect the relevant pigeon ently subjective, and depends on factors such as the rater's color antigens and that most of the antigenic proteins identified were also abun- perception. Therefore, it is important to assess inter-rater agreement. dant in droppings and blooms (Rouzet et al., 2017). Rouzet et al. (2017) In this study, the agreement was graded as substantial, according to found proteins with a molecular weight range between 20 and 200 kD the classification recommended by Landis and Koch (1977). in bird dropping extracts, and that IGLL-1 (Immunoglobulin lambda-like The main advantage of this ICT is the speed of the analysis, and the polypeptide-1, 24.5 KDa) and ProE (Proproteinase E, 20.5 KDa) were the fact that it can be performed in anon-laboratory environments. Al- antigenic proteins with the most clinical interest as biomarkers for BRHP though speed may not be decisive in environmental monitoring, it diagnosis. In this regard, Shirai et al. (2017) in an immunoblot analysis of may constitute a real advantage for industrial hygiene monitoring, as sera from patients with BRHP, identified multiple immunoreactive bands some of the sampling procedures widely used in occupational hygiene in pigeon serum at various molecular weights, especially IGLL-1, similar take only a short time (for example, surface wipe sampling, dust sam- to those obtained in the present study. pling or bulk sampling). In this context, some studies have assessed Moreover, the cross-reactivity study with other birds demonstrated the environmental measurement of avian antigen in dust (Craig et al., elevated specificity (Fig. 3). Sera of small parrot, parakeet and parrot 1992; Kuramochi et al., 2010; Sema et al., 2018; Curtis et al., 2002). seem to cross-react slightly with pigeon serum, probably because they For example, Craig et al. (1992) reported that avian antigen could be de- have common sera proteins (as observed in the 2D electrophoresis). An- tected in dust samples for prolonged periods of time after bird removal other possible reason was suggested by Rouzet et al. (2014). in a study and environmental clean-up. Curtis et al. (2002) also reported that pi- designed to determine the specificity of immunogenic proteins of the geon antigen was found in dust and air samples from a pigeon- droppings of three bird species; those authors reported that although infected school. no proteins were found common to all extracts, cross-reactions to One limitation of these methods is the problem of discriminating be- avian antigens from the three species of bird were observed. tween safe and unsafe environments. In this sense, data on environ- The sandwich ELISA using these polyclonal antibodies has a high mental levels of pigeon antigen to which patients with bird-related sensitivity and specificity and has demonstrated its value for measuring hypersensitivity pneumonitis are exposed were lacking in the present pigeon antigens, even at relatively low concentrations. In this study the study. It may be difficult to establish thresholds for allergen exposures, levels of pigeon antigens were assessed with respect to a reference due to inter-individual variations in susceptibility to both sensitization preparation (pigeon antigen standard) containing 6.7–27,500 ng of pro- and disease elicitation. These differences may be caused by genetic tein/ml. This standard proved to be a suitable reference preparation for variations, age-dependent effects, or lifestyle factors (Maciag and the measurement of pigeon exposure, as shown by the results of the Phipatanakul, 2020). Smoking, for example, may promote sensitization SDS-PAGE gel (Fig. 2) and the comparison of the proteins of this stan- due to an adjuvant effect (Carlos et al., 2009). Moreover, different pro- dard with the air sample eluates collected from the dovecote. These as- teins have different sensitization potencies (Krutz et al., 2020). In any sessments demonstrated that this standard is composed of the proteins case, avoiding exposure to causal agents is crucial in HP management contained in the air. and is a decisive factor in prognosis, because progression can broadly Although the ELISA is a very useful technique for detecting soluble be prevented with appropriate antigen avoidance. Moreover, identifica- antigens and meets most of the main requirements for analysing pigeon tion of the offending antigen source at work is important for the deter- antigens (i.e., reproducibility, good specificity and acceptable sensitiv- mination of antigen-specific IgG antibodies. ity), it is a time-consuming process that requires skilled personnel and In conclusion, we developed an ICT to monitor pigeon aeroallergen expensive laboratory equipment. Therefore, cost-efficient, less labor- levels. The strip assay described was rapid, simple, sensitive, highly con- intensive technological procedures that can be performed by non- sistent with the validated ELISA and does not require expensive equip- specialist laboratory staff are needed in order to monitor pigeon antigen ment or specific skills. This method for monitoring levels of pigeon levels. ICT can provide fast qualitative or even semi-quantitative results antigen may prove to be a useful and versatile tool for predicting the within minutes with very good analytical sensitivity. Furthermore, the progression of bird-related hypersensitivity pneumonitis. characteristics of their design and construction make them especially suitable for field testing and for use by non-laboratory-skilled personnel Funding (Bogdanovic et al., 2006; Tsay et al., 2002). Fast antigen monitoring sys- tems has been shown to be highly beneficial to diminishing exposure SSD is researcher supported by CIBER (Instituto de Salud Carlos III, (Tsutsui et al., 2015; Kuramochi et al., 2010), therefore, rapid immuno- Spain), and MJC is a researcher supported by the Miguel Servet pro- assay for the assessment of pigeon antigens would be a very useful tool. gramme of the Instituto de Salud Carlos III (MSII17/00025). This project The visual interpretation of the ICT showed high levels of sensitivity was supported by the Fundació Catalana de Pneumologia (FUCAP), and specificity, comparable to those recorded with other rapid assays Sociedad Española de Patología Respiratoria (SEPAR), FIS PI15/01954 measuring other aeroallergens (Bogdanovic et al., 2006; Tsay et al., (Instituto de Salud Carlos III), Fondo Europeo de Desarrollo Regional 2002; Álvarez-Simon et al., 2014; Koets et al., 2011). Furthermore, the (FEDER) and European Union's Horizon 2020 research and innovation line intensity (analysed by densitometry, an objective measurement) programme under grant agreement No 874707. The funders had no of samples categorized as negative was significantly lower than that of role in study design, data collection and analysis, decision to publish positive and double positive samples. The LOD, determined as the min- or preparation of the manuscript. imum amount of allergen capable of producing a clearly visible signal in the test line of the standard in the rapid immunoassay, was 420 ng/ml. CRediT authorship contribution statement Unexpectedly, in the visual analysis of our test results, the independent observers detected pigeon antigen concentrations below the theoretical Silvia Sánchez-Díez: Conceptualization, Formal analysis, Methodol- LOD determined for the assay; depending on the rater, the actual LOD of ogy, Writing – original draft, Writing – review & editing. María-Jesús the assay with real field samples ranged from 141 to 416 ng/ml. This ob- Cruz: Conceptualization, Funding acquisition, Project administration, servation has also been reported in other studies: for example, Tsay Writing – original draft, Writing – review & editing. Daniel Álvarez- 7 S. Sánchez-Díez, M.-J. Cruz, D. Álvarez-Simón et al. Science of the Total Environment 788 (2021) 147789 Simón: Methodology, Writing – review & editing. Tomás Montalvo: Landis, J.R., Koch, G.G., 1977. The measurement of observer agreement for categorical data. Biometrics. 33 (1), 159. https://doi.org/10.2307/2529310. Conceptualization, Methodology, Writing – review & editing. Xavier Maciag, M.C., Phipatanakul, W., 2020. Prevention of asthma: targets for intervention. Muñoz: Writing – review & editing. Peter M. Hoet: Methodology, Chest. 158 (3), 913–922. https://doi.org/10.1016/j.chest.2020.04.011. Writing – review & editing. Jeroen A. Vanoirbeek: Methodology, Writ- Morell, F., Roger, À., Reyes, L., Cruz, M.J., Murio, C., Muñoz, X., 2008. Bird fancier’s lung: a series of 86 patients. Medicine (Baltimore) 87 (2), 110–130. https://doi.org/10.1097/ ing – review & editing. Susana Gómez-Ollés: Conceptualization, MD.0b013e31816d1dda. Funding acquisition, Writing – original draft, Writing – review & editing. Morell, F., Villar, A., Montero, M.-Á., et al., 2013. Chronic hypersensitivity pneumonitis in patients diagnosed with idiopathic pulmonary fibrosis: a prospective case-cohort study. Lancet Respir. Med. 1 (9), 685–694. https://doi.org/10.1016/S2213-2600(13) Declaration of competing interest 70191-7. Okamoto, T., Miyazaki, Y., Ogura, T., et al., 2013. A nationwide epidemiological survey of The authors declare that they have no known competing financial chronic hypersensitivity pneumonitis in Japan. Respir. Investig. 51 (3), 191–199. interests or personal relationships that could have appeared to influ- https://doi.org/10.1016/j.resinv.2013.03.004. Quirce, S., Vandenplas, O., Campo, P., et al., 2016. Occupational hypersensitivity pneumo- ence the work reported in this paper. nitis: an EAACI position paper. Allergy. 71 (6), 765–779. https://doi.org/10.1111/ all.12866. References Raghu, G., Remy-Jardin, M., Ryerson, C.J., et al., 2020. Diagnosis of hypersensitivity pneu- monitis in adults. An official ATS/JRS/ALAT clinical practice guideline. Am. J. Respir. Agache, I.O., Rogozea, L., 2013. Management of hypersensivity pneumonitis. Clin. Transl. Crit. Care Med. 202 (3), e36–e69. https://doi.org/10.1164/rccm.202005-2032ST. Allergy. 3 (1), 5. https://doi.org/10.1186/2045-7022-3-5. Rodrigo, M.J., Benavent, M.I., Cruz, M.J., et al., 2000. Detection of specific antibodies to pi- Álvarez-Simon, D., Cruz, M.J., Untoria, M.D., et al., 2014. A rapid test for soy aeroallergens geon serum and bloom antigens by enzyme linked immunosorbent assay in pigeon exposure assessment. PLoS One 9 (2). https://doi.org/10.1371/journal.pone.0088676. breeder’s disease. Occup. Environ. Med. 57 (3), 159–164. https://doi.org/10.1136/ Bogdanovic, J., Koets, M., Sander, I., et al., 2006. Rapid detection of fungal α-amylase in the OEM.57.3.159. work environment with a lateral flow immunoassay. J. Allergy Clin. Immunol. 118 Rouzet, A., Reboux, G., Rognon, B., et al., 2014. Immunogenic proteins specific to different (5), 1157–1163. https://doi.org/10.1016/j.jaci.2006.07.004. bird species in bird fanciers lung. J. Toxicol. Environ. Heal - Part A Curr. Issues 77 (12), Borderías, L., Morell, F., Vera, J., Briz, H., Muñoz, X., Cruz, M.J., 2010. Neumonitis por 724–730. https://doi.org/10.1080/15287394.2014.889616. hipersensibilidad crónica por estorninos. Exposición antigénica no masiva pero Rouzet, A., Reboux, G., Dalphin, J.-C., et al., 2017. An immunoproteomic approach revealed persistente. Arch. Bronconeumol. 46 (11), 607–609. https://doi.org/10.1016/j. antigenic proteins enhancing serodiagnosis performance of bird fancier’s lung. arbres.2009.10.009. J. Immunol. Methods 450, 58–65. https://doi.org/10.1016/j.jim.2017.07.012. Carlos, E.B.C., Gómezc, R.M., Rodrigo, B.C., Canonicab, G.W., 2009. Impact of environmental Selman, M., 2004. Hypersensitivity pneumonitis: a multifaceted deceiving disorder. Clin. tobacco smoke and active tobacco smoking on the development and outcomes of Chest Med. 25 (3), 531–547. https://doi.org/10.1016/j.ccm.2004.04.001. asthma and rhinitis. Curr. Opin. Allergy Clin. Immunol. 9 (2), 136–140. https://doi. Sema, M., Miyazaki, Y., Tsutsui, T., Tomita, M., Eishi, Y., Inase, N., 2018. Environmental org/10.1097/ACI.0b013e3283294038. levels of avian antigen are relevant to the progression of chronic hypersensitivity Craig TJ, Hershey J, Engler RJ, Davis W, Carpenter GB, Salata K. Bird antigen persistence in pneumonitis during antigen avoidance. Immunity, Inflamm. Dis. 6 (1), 154–162. the home environment after removal of the bird. Ann. Allergy. 1992;69(6):510–512. https://doi.org/10.1002/iid3.202. http://www.ncbi.nlm.nih.gov/pubmed/1471783. Accessed May 2, 2019. Shirai, T., Furusawa, H., Furukawa, A., et al., 2017. Protein antigen of bird-related hyper- Curtis L, Lee B-S, Cai D, et al. Pigeon allergens in indoor environments: a preliminary sensitivity pneumonitis in pigeon serum and dropping. Respir. Res. 18 (1), 65. study. Allergy. 2002;57(7):627–631. http://www.ncbi.nlm.nih.gov/pubmed/ https://doi.org/10.1186/s12931-017-0555-4. 12100304. Accessed May 2, 2019. Tsay, A., Williams, L., Mitchell, E.B., Chapman, M.D., Multi-Centre Study Group, 2002. A de Beer, P.M., Bouic, P.J., Joubert, J.R., 1990. Identification of a “disease-associated” antigen rapid test for detection of mite allergens in homes. Clin. Exp. Allergy 32 (11), in pigeon breeder’s disease by western blotting. Int. Arch. Allergy Immunol. 91 (4), 1596–1601. http://www.ncbi.nlm.nih.gov/pubmed/12569980 (Accessed May 2, 343–347. https://doi.org/10.1159/000235139. 2019). Gómez-Ollés, S., Cruz, M.J., Renström, A., Doekes, G., Morell, F., Rodrigo, M.J., 2006. An am- Tsutsui, T., Miyazaki, Y., Kuramochi, J., Uchida, K., Eishi, Y., Inase, N., 2015. The amount of plified sandwich EIA for the measurement of soy aeroallergens. Clin. Exp. Allergy 36 avian antigen in household dust predicts the prognosis of chronic bird-related hyper- (9), 1176–1183. https://doi.org/10.1111/j.1365-2222.2006.02542.x. sensitivity pneumonitis. Ann. Am. Thorac. Soc. 12 (7), 1013–1021. https://doi.org/ Koets, M., Renström, A., Zahradnik, E., Bogdanovic, J., Wouters, I.M., Van Amerongen, A., 10.1513/AnnalsATS.201412-569OC. 2011. Rapid one-step assays for on-site monitoring of mouse and rat urinary aller- Vasakova, M., Morell, F., Walsh, S., Leslie, K., Raghu, G., 2017. Hypersensitivity pneumoni- gens. J. Environ. Monit. 13 (12), 3475–3480. https://doi.org/10.1039/c1em10658a. tis: perspectives in diagnosis and management. Am. J. Respir. Crit. Care Med. 196 (6), Kokkarinen, J., Tukiainen, H., Seppä, A., Terho, E.O., 1994. Hypersensitivity pneumonitis 680–689. https://doi.org/10.1164/rccm.201611-2201PP. due to native birds in a bird ringer. Chest. 106 (4), 1269–1271. https://doi.org/ Vasakova, M., Selman, M., Morell, F., Sterclova, M., Molina-Molina, M., Raghu, G., May 10.1378/CHEST.106.4.1269. 2019. Hypersensitivity pneumonitis: current concepts of pathogenesis and potential Krutz, N.L., Kimber, I., Maurer-Stroh, S., Gerberick, G.F., 2020. Determination of the relative targets for treatment. Am. J. Respir. Crit. Care Med. https://doi.org/10.1164/ allergenic potency of proteins: hurdles and opportunities. Crit. Rev. Toxicol. 50 (6), rccm.201903-0541PP rccm.201903-0541PP. 521–530. https://doi.org/10.1080/10408444.2020.1793895. Vourlekis, J.S., Schwarz, M.I., Cherniack, R.M., et al., 2004. The effect of pulmonary fibrosis Kuramochi, J., Inase, N., Takayama, K., Miyazaki, Y., Yoshizawa, Y., 2010. Detection of in- on survival in patients with hypersensitivity pneumonitis. Am. J. Med. 116 (10), door and outdoor avian antigen in management of bird-related hypersensitivity 662–668. https://doi.org/10.1016/j.amjmed.2003.12.030. pneumonitis. Allergol. Int. 59 (2), 223–228. https://doi.org/10.2332/allergolint.09- OA-0161. Lacasse, Y., Selman, M., Costabel, U., et al., 2009. Classification of hypersensitivity pneu- monitis. Int. Arch. Allergy Immunol. 149 (2), 161–166. https://doi.org/10.1159/ 000189200. 8