Tree Allergen Pollen-Related Content as Pollution Source in the City of Ourense (NW Spain)

Allergies became a major public health problem, identified as an important global pandemic with a considerable impact on the worldwide economy. In addition, a higher prevalence of pollen Type I sensitization cases in urban environments in comparison with the rural territories was detected. Our survey sought to assess the main biological pollution episodes caused by the aeroallergens of the major allergenic tree species in urban environments. A Hirst-type volumetric device was used for pollen sampling and a Burkard Cyclone sampler for the detection of tree atmospheric allergens over two years. The main allergens of Alnus, Fraxinus, Betula, Platanus and Olea, were detected in the atmosphere. Three peaks of important pollen concentrations were recorded throughout the year. The developed regression equations between pollen counts and allergen proteins registered great R2 values. The number of days with probability of allergenic symptoms was higher when the pollen and allergen data were assessed altogether. Fraxinus allergens in the atmosphere were detected using Ole e 1 antibodies and the Aln g 1 allergens with Bet v 1 antibodies, demonstrating the cross-reaction processes between the principal allergenic proteins of the Oleaceae and Betulaceae families. Long Distance Transport processes (LDT) showed that pollen from Betula populations located in mountainous areas increased the secondary peaks of pollen and allergen concentrations, and air masses from extensive olive orchards of North-Eastern Portugal triggered the highest concentrations in the atmosphere of Olea pollen and Ole e 1 allergens.


Introduction
An ongoing global intensification of the incidence of pollen allergy diseases over the last half century was observed [1]. Allergies became a major public health problem in the urban atmosphere of the industrialized and emergent countries, recognized as an important global pandemic with a considerable impact on the worldwide economy [1,2]. In Europe, it was estimated that 20% of the citizens suffer from pollen Type I sensitive reactions [3], which increased in the most developed countries with incidences above 30% [4,5]. In addition, a higher prevalence of pollen-related sensitization cases in urban environments in comparison with rural spaces was detected [6,7]. Among the possible causes that have enlarged the allergenic content in the air of the urbans areas, the "heat island effect" of the cities prompts an increase of plants' pollen production and shifts pollen seasons to an earlier onset and lengthier durations [8]. Furthermore, urban atmospheric chemical contaminants favor a higher vegetal biomass growth and the increase of the allergenic protein content on pollen grains [9][10][11]. The greater occurrence and intensity of pollen allergic symptomatology in hypersensitized people during recent years was related with the increase of pollen production by plants [1,10]. Additionally, this situation is

Materials and Methods
The research was conducted in the city of Ourense located in the North-Western part of the Iberian Peninsula, an altitude of 454 m a.s.l. and a geographical location 42 • 20 N-7 • 52 W. The climate of this area is described as Oceanic with Mediterranean features, with an annual average temperature of 14 • C and a total precipitation of 772 mm in a year [40].
Aerobiological sampling of tree pollen and allergens during the years 2017 and 2018 was conducted using two volumetric traps placed on the roof of the Science Faculty building, approximately at 15 m above the ground level and near to the town center. Pollen was monitored using a Hirst-type Lanzoni VPPS-2000 volumetric sampler (Lanzoni s.r.l., Bologna, Italy) [41] with a pressure flow rate of 10 L/min, simulating the human breathing. Melinex adhesive tape was used as a pollen grain capture surface. Pollen quantification was conducted applying the methodology proposed by the Spanish Aerobiological Network (REA) [42], based on four longitudinal transects along the slides. The Main Pollen Season (MPS) was stablished using the Andersen method [43], which defines the MPS as the period from the day 2.5% of total annual pollen concentrations were reached to the date when 97.5% is accomplished. The classification recommended by the REA [30] was followed to categorize the pollen concentrations, as well as for the calculation of the thresholds of allergy hazard. For the quantification of the allergenic fraction, a Burkard Multi-Vial Cyclone Sampler (Hertfordshire, UK) with 16.5 L/min of aspiration flow rate was used. The bioaerosol particles were sampled into Eppendorf vials every 24 h and analyzed with the Takahashi et al. method [44] modified by the Moreno-Grau et al. method [45]. The 2-site ELISA methodology was used for the quantification of the aeroallergen content in the bioaerosol samples in four steps [12,46]. The antibodies Ole e 1 and Pla a 1 (Roxall S.A) were used for the determination of the allergen content of Fraxinus, Olea and Platanus allergens, and the Bet v 1 specific monoclonal antibody (ALK-Abelló) was used to quantify the Betula and Alnus allergen content in the aerosol. The absorbance was measured at 492 nm.
Meteorological data were acquired from the Galician Institute for Meteorology and Oceanography METEOGALICIA "Ourense" station, placed at 300 m of the pollen and allergen samplers. The measured parameters were temperature ( • C), relative humidity (%), precipitation (mm) and wind speed (km/h) ( Figure 1).
Spearman's non-parametric correlation test and Principal Component analysis (PCFA) were applied to evaluate the association between the pollen and the allergen concentrations in the air with the main weather parameters. A regression equation between the pollen and allergen data was conducted in order to obtain the aeroallergens thresholds and the amount of days with potential hazard of allergy prompted, both for pollen and allergens. The STATISTICA 7 program was used for the statistical analysis.
HYSPLIT back trajectories were assessed to study the daily pollen and allergen higher concentrations. The models led us to stablish the provenance (latitude, longitude and elevation) in the selected days of air masses using meteorological data at the 250, 500 and 700 m heights from the earth surface [47]. Spearman's non-parametric correlation test and Principal Component analysis (PCFA) were applied to evaluate the association between the pollen and the allergen concentrations in the air with the main weather parameters. A regression equation between the pollen and allergen data was conducted in order to obtain the aeroallergens thresholds and the amount of days with potential hazard of allergy prompted, both for pollen and allergens. The STATISTICA 7 program was used for the statistical analysis.
HYSPLIT back trajectories were assessed to study the daily pollen and allergen higher concentrations. The models led us to stablish the provenance (latitude, longitude and elevation) in the selected days of air masses using meteorological data at the 250, 500 and 700 m heights from the earth surface [47].

Results
Two periods of important tree pollen concentrations were recorded throughout the studied year ( Figure 2). The first was consequence of the Alnus and Fraxinus blooms during January and February. The second, the quantitatively most important period, was mainly due to the pollination of Betula, Platanus and Olea during the spring months. The last period had a great impact on sensitization processes due to the flowering of trees with a high recognized allergy potential mainly in urban environments.

Results
Two periods of important tree pollen concentrations were recorded throughout the studied year ( Figure 2). The first was consequence of the Alnus and Fraxinus blooms during January and February. The second, the quantitatively most important period, was mainly due to the pollination of Betula, Platanus and Olea during the spring months. The last period had a great impact on sensitization processes due to the flowering of trees with a high recognized allergy potential mainly in urban environments.   The occurrence of Alnus pollen in the Ourense atmosphere was observed from the second fortnight of January to the end of February. We registered a total airborne pollen of 2692 and 6368 pollen grains in 2017 and 2018, respectively, during a MPS with a length of 40 and 52 days. The highest alder pollen value was recorded on 24 January with 867 pollen/m 3 . The total annual integral of Aln g 1 was 9.200 ng and 7.386 ng in 2017 and 2018, respectively, with a peak observed on 1 February 2017 with 1.868 g/m 3 , one day after the pollen peak. Although both peaks were recorded during a period of lack of rainfall, the Aln g 1 peak was observed during an increase (2.5 • C) in maximum temperatures (Table 1, Figure 2 (Table 1, Figure 2). Table 1. Date of the start, end and length of the main pollen season (MPS) (days), mean pollen (pollen/m 3 ), date of the pollen peak (day), pollen (pollen) and allergen, allergen peak of the MPS (ng/m 3 ), date of the allergen peak (day) and Pollen Allergen Potency (AP) (ng/pollen). With the aim of determining the effect of the meteorological parameters in the pollen and allergen airborne content, a non-parametric Spearman's correlation test was conducted. Generally, spring-flowering trees showed positive correlations between airborne pollen and allergen concentrations and temperature, and negative with relative humidity (p < 0.01) ( Table 2). Overall, the highest significant correlation coefficients were obtained among the allergen or the pollen and the average temperatures, with the greatest positive degree of association for the Olea pollen and the Ole e 1 allergen (p < 0.01). In addition, negative significant correlations between pollen concentrations with mean, maximum and minimum temperatures and positive with relative humidity were recorded with Alnus and Fraxinus pollen, as well as with Aln g 1 and Fra e 1 allergens, the winter bloom trees. For the rest of the studied parameters, we obtained positive significant correlations between pollen and allergens of Betula, Olea and Platanus and wind speed ( Table 2). Furthermore, a principal component analysis (PCFA) was conducted with the aim to better understand the meteorological influence in the pollen and allergen airborne concentrations as PCFA showed the influence of all consider weather variables as a whole. The purpose of the analysis is to obtain a small number of linear combinations of the selected variables which account for most of the variability in the data. Three components have been extracted for each taxon, since they had eigenvalues greater than or equal to 1.0 and they account for together between 79% and 84% of the variability in the original data. The three PCs were correlated as follows: Component 1 temperatures and relative humidity, Component 2 pollen and allergen values and Component 3 wind speed and rainfall ( Figure 3). To better understand the relationship between pollen/allergens and meteorological parameters, a plot with the PC explaining the higher variability PC1 vs. PC2 was conducted ( Figure 3). The results obtained reinforced the correlation analysis results registering a high positive degree of association between the pollen counts and the allergen levels ( Figure 3).     The Pollen Allergen Potency (AP) index was calculated for each taxon, which represented the rate between the allergen and pollen grain concentrations. The highest value was 0.004 ng/pollen registered for Fraxinus and the lowest for Olea with 0.0004, both during the first year of study (Table 1).

Fraxinus
Regression equations were performed to identify the aeroallergen thresholds for low, moderate and high hazard of symptomatology appearance on sensitized people (Table 3). The pollen threshold concentrations suggested by the REA were followed to obtain the equivalent aeroallergens thresholds ( Table 4). Values of the adjusted R 2 coefficients in the performed equations oscillated between 0.289 for Alnus and 0.737 for Olea (Table 3) The obtained thresholds were applied in order to ascertain the number of days with possible allergy hazard for sensitive patients. Considering the pollen data, the taxa that registered a higher quantity of days with moderate potential hazard for allergenic suffers were Alnus and Platanus with a sum of 22 and 27 days in both years, respectively (Table 4). In the case of the allergen airborne moderate levels, the great amount of days was registered for Platanus and Betula with a total of 30 and 31 days, respectively, during the two years of study (Table 4). Some discordances were also observed in the case of the episodes of high potential hazard on sensitive patients. Alnus and Betula registered the most important quantity with a sum of 55 and 58 days detected for the Aln g 1 and Bet v 1 allergen concentration during the period of the study (Table 4). The Pollen Allergen Potency (AP) index was calculated for each taxon, which represented the rate between the allergen and pollen grain concentrations. The highest value was 0.004 ng/pollen registered for Fraxinus and the lowest for Olea with 0.0004, both during the first year of study (Table 1).
Regression equations were performed to identify the aeroallergen thresholds for low, moderate and high hazard of symptomatology appearance on sensitized people (Table 3). The pollen threshold concentrations suggested by the REA were followed to obtain the equivalent aeroallergens thresholds ( Table 4). Values of the adjusted R 2 coefficients in the performed equations oscillated between 0.289 for Alnus and 0.737 for Olea (Table 3) The obtained thresholds were applied in order to ascertain the number of days with possible allergy hazard for sensitive patients. Considering the pollen data, the taxa that registered a higher quantity of days with moderate potential hazard for allergenic suffers were Alnus and Platanus with a sum of 22 and 27 days in both years, respectively (Table 4). In the case of the allergen airborne moderate levels, the great amount of days was registered for Platanus and Betula with a total of 30 and 31 days, respectively, during the two years of study (Table 4). Some discordances were also observed in the case of the episodes of high potential hazard on sensitive patients. Alnus and Betula registered the most important quantity with a sum of 55 and 58 days detected for the Aln g 1 and Bet v 1 allergen concentration during the period of the study (Table 4). Furthermore, a back-trajectory analysis was conducted to explain the timing discrepancies observed between the pollen and allergen peaks for all taxa. Only special situations were observed in the case of Betula and Olea (Figure 4). The analysis led us to detect that the second pollen and allergen Betula peaks were coincident with air masses from the high mountainous areas around the city of Ourense. In the case of Olea and Ole e 1, the pollen and allergen maximum peaks were influenced by continental air masses coming from the North of Portugal (Figure 4). Furthermore, a back-trajectory analysis was conducted to explain the timing discrepancies observed between the pollen and allergen peaks for all taxa. Only special situations were observed in the case of Betula and Olea (Figure 4). The analysis led us to detect that the second pollen and allergen Betula peaks were coincident with air masses from the high mountainous areas around the city of Ourense. In the case of Olea and Ole e 1, the pollen and allergen maximum peaks were influenced by continental air masses coming from the North of Portugal (Figure 4).

Discussion
The occurrence of pollen grains in the atmosphere was recognized as a cause of important pollution problems such as allergies to human health [48]. The most important tree allergenic taxa of the city, which constituted the 50% of the total of airborne pollen, were studied to determine their importance in the atmospheric allergenic load. Regarding the pollen season, Alnus and Fraxinus are arboreal species that flower in winter whereas Betula, Olea and Platanus with spring flowering. Several studies reported similar findings for Alnus in the same area and for other countries regarding the start, end date and duration of the MPS, but with differences in the total amount of annual pollen [10,49,50]. Some authors found that meteorological parameters, as mean temperature during the previous months to pollination, affect the annual airborne pollen sum [51]. We registered a shorter birch pollen season duration with higher total pollen amount than those reported by several studies for the same area [10,49] or in other European countries like Portugal [47], Poland [51], Sweden [52] and Romania [53]. In the case of Olea, a similar MPS duration was registered regarding the reported by studies carried out in the same area [54] or in Portugal [47,55]. However, longer MPS with higher pollen number was observed in Mediterranean areas because of the olive crops [56].
The classically information for hyper sensitized patients is the concentration of pollen grains in the atmosphere and their timing [24]. Nevertheless, in the last years the period of pollen exposure often did not coincide with the symptoms' appearance in different regions [16,25]. In the present study we detected a high atmospheric allergenic load [10] in the atmosphere coinciding with low levels of airborne pollen because the presence of pollen allergens in the air. Although pollen allergens are firstly carried in the atmosphere by pollen grains [37], they may also could be detected in the microaerosol suspension which could remain for longer periods in the atmosphere [57]. Our results reinforced the fact that pollen concentration data must be combined with the aeroallergen detection in order to determine the real load of atmospheric allergenic particles and the development of complete and lasting systems aimed to observe and gather environmental pollution information [11,47]. In spite of regression equations between pollen counts and allergen protein registered high R 2 , several discordances were detected between the days of aeroallergen and pollen allergy risk, as in some cases did not coincide. Considering both together, the Alnus pollen and allergen data, the number of days with high allergenic hazard raised to 64 compared with their assessment separately. For Fraxinus, when the pollen and allergen concentrations were combined, the number of episodes with potential hazard raised to 16 for high risk of symptomatology appearance. Additionally, lower discrepancies were observed in the case of Platanus and Betula as the number of days with high potential hazard of allergy increased to 34 and 59, respectively, when the pollen and allergen data

Discussion
The occurrence of pollen grains in the atmosphere was recognized as a cause of important pollution problems such as allergies to human health [48]. The most important tree allergenic taxa of the city, which constituted the 50% of the total of airborne pollen, were studied to determine their importance in the atmospheric allergenic load. Regarding the pollen season, Alnus and Fraxinus are arboreal species that flower in winter whereas Betula, Olea and Platanus with spring flowering. Several studies reported similar findings for Alnus in the same area and for other countries regarding the start, end date and duration of the MPS, but with differences in the total amount of annual pollen [10,49,50]. Some authors found that meteorological parameters, as mean temperature during the previous months to pollination, affect the annual airborne pollen sum [51]. We registered a shorter birch pollen season duration with higher total pollen amount than those reported by several studies for the same area [10,49] or in other European countries like Portugal [47], Poland [51], Sweden [52] and Romania [53]. In the case of Olea, a similar MPS duration was registered regarding the reported by studies carried out in the same area [54] or in Portugal [47,55]. However, longer MPS with higher pollen number was observed in Mediterranean areas because of the olive crops [56].
The classically information for hyper sensitized patients is the concentration of pollen grains in the atmosphere and their timing [24]. Nevertheless, in the last years the period of pollen exposure often did not coincide with the symptoms' appearance in different regions [16,25]. In the present study we detected a high atmospheric allergenic load [10] in the atmosphere coinciding with low levels of airborne pollen because the presence of pollen allergens in the air. Although pollen allergens are firstly carried in the atmosphere by pollen grains [37], they may also could be detected in the microaerosol suspension which could remain for longer periods in the atmosphere [57]. Our results reinforced the fact that pollen concentration data must be combined with the aeroallergen detection in order to determine the real load of atmospheric allergenic particles and the development of complete and lasting systems aimed to observe and gather environmental pollution information [11,47]. In spite of regression equations between pollen counts and allergen protein registered high R 2 , several discordances were detected between the days of aeroallergen and pollen allergy risk, as in some cases did not coincide. Considering both together, the Alnus pollen and allergen data, the number of days with high allergenic hazard raised to 64 compared with their assessment separately. For Fraxinus, when the pollen and allergen concentrations were combined, the number of episodes with potential hazard raised to 16 for high risk of symptomatology appearance. Additionally, lower discrepancies were observed in the case of Platanus and Betula as the number of days with high potential hazard of allergy increased to 34 and 59, respectively, when the pollen and allergen data were considered together. No differences were observed in the case of Olea. The developed innovative tools for quantification of the atmospheric allergenic load are especially useful and necessary for complementing the classic pollen counts to attain an improvement and optimization of the clinical allergy treatments administration and a decrease in medication consumption by the sensitized-to-pollen population [11,47]. The establishment of new periods of allergen presence in the atmosphere will highlight novel perspectives in the epidemiological study of respiratory allergy-related disorders and the biological pollution.
In addition, a great prevalence of pollen related incidence of allergy in urban environments compared with the rural areas was detected [6]. It is noteworthy that the allergy incidence was prompted by an inaccurate planning and design of the urban tree vegetation, with several plants of the same family that can develop cross-reactions processes between their allergens, enhancing sensitizations in sensitive people [12]. One of the major results of our study was the detection of the Fraxinus and Alnus pollen related allergens in the air by using antibodies from another genus. This finding evidences the cross-reactivity between the principal allergens of the Oleaceae and Betulaceae tree families, referred as the capacity of several IgE antibodies to recognize diverse antigens [58]. Due to their resilience and tolerance to adapt to urban conditions, both tree families are broadly planted as tree ornamental vegetation in green areas of urban settings [59]. The urban allergic population to the Betula and Olea pollen could display allergy responses in the winter months (as a consequence of ash and alder pollen allergens), in the early spring because the Bet v 1 allergens or late spring due to Ole e 1 tree pollen related allergens. Some authors pointed out that the people with oral allergy syndrome (OAS) showed oral symptoms before Betula pollinosis symptoms [32], and this symptomatology may lead to people think that with a lower concentration of Bet v 1 allergen there are symptoms. Patients allergic to birch have previously suffered from the symptoms during the flowering of the alder, which is known as priming effect.
Moreover, Long Distance Transport processes (LDT) could also justify that airborne pollen concentrations are not always related with the actual exposure to their main allergens [60,61]. In the case of Betula, the first peak of the pollen curve matched with the flowering of the birch populations in the surroundings of the city, whilst the secondary was registered when the nearest Betula tree had finished its flowering period [62]. Topographical characteristics must also be taken into account when considering pollen transport. The HYSPLIT models showed that pollen originating from Betula populations located in mountainous areas even at some distance from the city, which flowers some weeks later due to its higher altitudinal distribution, could be transported through the channels formed by the river crossing the Ourense city, increasing the secondary peaks of pollen and allergen concentration. During the year 2017, the second peak pollen potency seems higher compared to the first peak, possibly related to a possible transport of high-potency birch pollen from the most elevated areas of the region, as it was observed for the olive pollen in South of Europe [60]. In addition, the back trajectories during the maximum concentrations of olive and Ole e 1 in the air noticed the influence of the air masses from the widespread olive orchards of North-Eastern Portugal in the amount of olive related bio particles in the atmosphere. Betula and Olea pollen morphology favors its transport over medium or long distances [60,63,64].
The analysis of the main meteorological variables, pollen and aeroallergen concentrations showed different results depending on the taxa. In the spring flowering trees, a statistically significant positive correlation between pollen or allergen occurrence and temperature [54] and wind speed was observed. On the contrary, the association degree was positive with relation to humidity in the case of the winter flowering trees and negative with temperatures, which presented negative correlations with temperatures. Therefore, weather-related factors, such as mild temperatures, influenced the dispersion of spring and fall pollen and allergens, as previously described by other authors [10,65,66].

Conclusions
The major allergens of Alnus, Fraxinus, Betula, Platanus and Olea were detected in the atmosphere of the Ourense city. Two peaks of important pollen concentrations were recorded throughout the year. One of the major findings of our study was the detection of the Fraxinus and Alnus pollen related allergen proteins in the air using antibodies from another genus, demonstrating the cross-reactivity processes between the principal allergenic proteins of the Oleaceae and Betulaceae families. The developed regression equations between pollen counts and allergen proteins registered high R 2 values. We observed high atmospheric allergenic load in the atmosphere coinciding with low levels of airborne pollen because of the presence of pollen allergens in the air. The number of days with a moderate and high hazard of allergy was higher when the pollen and allergen data were assessed together. Considering the pollen data individually, the number of episodes of high allergy symptomatology hazard were understated. The combination of pollen and allergen information should be evaluated to ascertain the real biological pollution in the atmosphere and the actual potential risk episodes for the sensitized population. Long Distance Transport processes (LDT) also explain that airborne pollen levels could not appropriately represent the exposure to their main allergens. The applied back trajectory analysis showed that pollen from Betula populations located in mountainous areas increased their secondary peaks of pollen and Bet v 1 concentrations and southern air masses from intensive plantations caused the highest airborne Olea concentrations.