sunyu@scau.edu.cn,
Abstract
Rhinitis is one of the most prevalent diseases in the world. Indoor microbiome is confirmed to associate with respiratory diseases such as asthma and infections, but no study reported the association between indoor microbiome and the occurrence of rhinitis. In this study, 370 students were randomly selected from 8 junior schools in Terengganu, Malaysia, and self-administered questionnaire and skin prick tests were conducted to define the allergic and non-allergic rhinitis among students. Vacuum dust was collected from the floor and chair/desk surfaces in the classrooms, and culture-independent high-resolution amplicon sequencing and quantitative PCR were conducted to characterize the absolute concentration of bacterial and fungal species. Hierarchical logistic regression was applied in the association analyses. We found similar microbial associations for the students with allergic and non-allergic rhinitis. The microbial richness in Gammaproteobacteria was protectively associated with allergic and non-allergic rhinitis (p = 0.02 and 0.04), and total fungal richness was positively associated with allergic and non-allergic rhinitis (p = 0.01 and 0.03). The absolute concentration of two bacterial species, Aeromonas enteropelogenesand Brasilonema bromeliae , were associated with both types of rhinitis, and six bacterial and one fungal species was associated with either allergic and non-allergic rhinitis (p < 0.005). Four species previously reported as facultative pathogens, includingA . enteropelogenes , Escherichia fergusonii ,Enterobacter xiangfangensis and Streptococcus salivarius , were protectively (negatively) associated with rhinitis. A higher concentration of two radiation-resistant species, includingDeinococcus gobiensis and Deinococcus grandis , were associated with an increased odds of rhinitis. This is the first health association study conducted between the indoor microbiome and the occurrence of rhinitis, revealing that microbial exposure at school can affect both allergic and non-allergic students.
Keywords
Allergic rhinitis, non-allergic rhinitis, skin prick test, indoor microbiome, high-throughput sequencing
Introduction
Rhinitis is one of the most prevalent diseases in the world, affecting up to 30% of the total population1-3. The prevalence of rhinitis is still increasing in most of the countries in the world, especially in many developing countries2. The disease leads to substantial economic and medical loss as well as many indirect costs for patients, such as absence from work and school and reduced work productivity. The symptoms of rhinitis include rhinorrhea/runny nose, nasal blockage/stuffy nose, itching nose and sneezing. Allergic rhinitis is defined as rhinitis accompanied with allergen-specific IgE production. It is frequently associated with asthma and ocular symptoms, such as allergic rhinoconjunctivitis, and it has been estimated that 10-40% of the patients with rhinitis also have asthma2, 4. Non-allergic rhinitis includes a heterogeneous subgroup of rhinitis without any clinical signs of allergic inflammation, such as rhinitis of the elderly, occupational rhinitis, gustatory rhinitis, hormonal rhinitis and drug-induced rhinitis 1, 5, 6. The mechanisms for non-allergic rhinitis are still mostly unclear. The diagnosis of allergic and non-allergic rhinitis is mainly based on allergen-specific IgE test. The test can be performed either on the skin surface by skin prick tests or in blood serum by enzyme immunoassay or radioimmunoassay 2.
The occurrence of rhinitis is suggested to be associated with genetic and environmental factors. Many environmental factors, such as pollen and fungal allergen, air and traffic pollution, climate change, ozone level, tobacco smoking, fragrance products, industrial chemicals and social class, are also suggested to be associated with the occurrence of rhinitis 7, 8. In recent years, the development of culture-independent high-throughput sequencing technology enables researchers to survey microbiome composition in various human body sites, such as human gut, skin and respiratory tract, as well as indoor environments, such as homes, schools, hospitals, hotels and public transport systems 9-15. These profiling studies showed that microbiome diversity and compositional variation are associated with many metabolic and immune diseases, such as asthma16. Studies identified many potential protective and risk microorganisms for asthma in the home and school environments17-22. Compared with extensive studies in asthma, no studies reported the relationships between indoor microbiome and occurrence of rhinitis, and thus the association pattern between the indoor microbiome and the development of rhinitis is unclear.
In this study, we investigated associations between indoor microbiome exposure and the occurrence of allergic and non-allergic rhinitis. The study was conducted among randomly selected junior high school students in Terengganu, Malaysia, and a self-administered questionnaire and skin prick test were conducted to assess the prevalence of allergic and non-allergic rhinitis. Species-level high-throughput amplicon sequencing was conducted to characterize microbiome composition in the classrooms. A regression model was applied to evaluate the associations between indoor microbiome and disease. The study aims to screen a list of potential protective and risk microorganisms for allergic and non-allergic rhinitis and evaluate the health effects of indoor microbiome exposure.
Materials and methods
Study design and ethical permission
In this study, 8 junior high schools were randomly selected from Terengganu, Malaysia. In each school, 4 classes were randomly selected to collect vacuum dust samples. Dust samples from 2 classrooms were used up in previous chemical analyses, and thus in total 30 samples were sequenced to character indoor microbiome composition. Students were randomly selected to conduct a skin prick test and report their health condition and personal information through the self-administered questionnaires. Vacuum dust, self-administered questionnaires and skin prick test were collected from November 2012 to February 2013. The study was approved by the Medical Research and Ethics Committee at the National University of Malaysia, Malaysia Ministry of Education, Education Department of Terengganu State and principals and teachers of each school. All participants gave their formal written consent, and the records were kept in the National University of Malaysia.
Health data and skin prick test
A self-administered questionnaire in the Malay language was distributed to students to survey the personal information, including age, gender, current smoking and parental asthma, and the occurrence of rhinitis. The questionnaire was also used in previous rhinitis studies in Sweden, China and Malaysia 23, 24. One question asked whether the students had a stuffy nose or runny nose in the last 3 months, and 4 alternatives were available to choose, including (A) “Yes, everyday”, (B) “Yes, 1-4 times/week”, (C) “Yes, 1-3 times/month” and “No, never”. The students chose (A), (B) or (C) were defined as having rhinitis. One question asked whether the students had respiratory infections in the last 3 months, and 4 alternatives were available to choose, including (A) “Yes, everyday”, (B) “Yes, 1-4 times/week”, (C) “Yes, 1-3 times/month” and “No, never”. The students chose (A), (B) or (C) were defined as having recent respiratory infections, and these students were excluded in the further rhinitis analyses. The questionnaire was answered at home with the help of the students’ parents. Medical staff from our group went through the questionnaire and had a face-to-face interview with students to clarify the uncertainties about the questions. When they answer the questionnaire, the students had no information regarding the data collected in the classrooms.
To define allergic and non-allergic rhinitis, we conducted skin prick tests for the selected students25. The tests were performed by experienced hospital nurses, and the participants were asked about their recent medication intake before the tests. Allergen extract was purchased from ALK-Abello, Horsholm, Denmark, including pollen (Birch mix, Phleum pretense and Artemisia vulgaris ), animal epidermal (Equus caballus and Felis domesticus ), house dust mite (Dermatophagoides farinae andDermatophagoides pteronyssinus ), molds (Cladosporium cladosporioides and Alternaria alternate /Alternaria tenuis ). Each bottle contained allergen extract in 10 HEP (histamine equivalent). A standard saline solution and histamine solution (10 mg/mL) were used as negative and positive controls. Skin prick tests were performed on the forearm with a disposable ALK skin lancet. The skin surface was observed after 15 minutes, and a positive reaction was defined as wheal > 3mm in diameter.
Vacuum dust sampling, DNA extraction and sequencing
A vacuum cleaner (400W) was used to collect dust in classrooms. A dust sampler (ALK Abello, Copenhagen, Denmark) with a Millipore filter was equipped in the vacuum cleaner. The Millipore filter was made of cellulose acetate. The pore size of the filter was 6 µm, which retains 74% particles of 0.3-0.5 µm, 81% particles of 0.5-1.0 µm, 95% particles of 1-10 µm and 100% particles larger than 10 µm according to manufacturer’s instruction. Each selected classroom was divided into two parts with approximately the same size; one near the corridor and one near the window. Dust sample was collected from each part by 4 minutes vacuum sampling with 2 minutes on the floor surface and 2 minutes on the upper surfaces of desks and chairs. Two dust samples in each classroom were combined and sieved through a 0.3-mm mesh screen to obtain the fine dust, which was further stored in a -80 ˚C freezer until DNA extraction and sequencing.
DNA extraction and multiplex amplicon sequencing were performed by Personal Biotechnology Co., Ltd (Shanghai, China). First, the total bacterial and fungal DNA was extracted from vacuum dust by DNA kit D5625-01 (Omega Bio-Tek, Inc., Norcross, GA, USA) and Fast DNA SPIN kit (MP Biomedicals, Santa Ana, CA, USA). The quality and quantity of the extracted DNA were assessed by agarose gel electrophoresis and NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The full-length of bacterial 16S rRNA gene was amplified by the forward primer 27F (5’-AGAGTTTGATCMTGGCTCAG-3’) and the reverse primer 1492R (5’-ACCTTGTTACGACTT-3’), and the fungal ITS region was amplified by the forward primer ITS1F (5’-CTTGGTCATTTAGAGGAAGTAA-3’) and the reverse primer LR3 (5’-CCGTGTTTCAAGACGGG-3’). Then sample-specific tag sequences (16 bp) were incorporated into the PCR amplicons by the Barcoded Univ F/R Primers Plate-96v2 kit (Pacific Biosciences, Menlo Park, CA, USA). The PCR amplicons were purified by Agencourt AMPure Beads (Beckman Coulter, Indianapolis, IN) and further quantified by PicoGreen dsDNA Assay Kit (Invitrogen, Carlsbad, CA, USA). The amplicons were sequenced by the Single-Molecule Real-Time (SMRT) technology combined with circular consensus sequencing (CCS) technique at the PacBio Sequel platforms 26, 27.
The total bacterial and fungal concentration in dust was quantified by quantitative PCR. First, the bacterial and fungal DNA was extracted from 10 mg dust, respectively. Universal primers targeted bacterial 16S rRNA gene (forward: 5’-GCAGGCCTAACACATGCAAGTC-3’ and reverse: 5’-CTGCTGCCTCCCGTAGGAGT-3’) and fungal ITS 1 regions (forward primer 5.8F1: 5’-AACTTTCAACAACGGATCTCTTGG-3’, the reverse primer is 5.8R1: 5’-GCGTTCAAAGACTCGATGATTCAC-3’) were used for the quantitative PCR23, 28.
Bioinformatics and association analyses
The raw sequence data were deposited in the Genome Sequence Archive29 with accession numbers CRA002825 and CRA002876. The sequenced circular consensus sequences were first processed by the PacBio SMRT Link portal (version 5.0.1.9585) with minfullpass as 3 and minPredictedAccuracy as 99. Sequences with predicted accurate < 99% were removed from further analyses. The Quantitative Insights Into Microbial Ecology (QIIME, v1.8.0) pipeline was used to process the filtered data30. Sequences were first assigned to the corresponding samples based on the sample-specific tag information. In each sample, sequences were clustered into operational taxonomic units (OTUs) with 97% of sequence similarity by UCLUST (v5.2.236) 31, and a representative sequence from each OUT was picked and annotated by searching against the Silva (release 115) and UNITE database (release 5) for bacteria and fungi32, 33. For each sample, an OTU table was generated to record the abundance of all characterized OTUs. OTUs counted for < 0.001% of total sequences were removed from further analyses. All samples were rarefied at even depth for the comparative analyses. Microbial compositional variation was visualized by NMDS (non-metric multidimensional scaling) analysis34 based on Bray-Curtis distance metrics.
We performed three-level (class and school as second and third levels) hierarchical logistic regression between microbial richness/absolute concentration and occurrence of allergic/non-allergic rhinitis by StataSE 15.0 (StataCorp LLC). Gender, current smoking and parental asthma were included as adjustments in the regression analyses. The microbial richness was represented as the number of OTUs in bacteria and fungi. The absolute concentration of microbial species was calculated as multiplying the relative abundance of species with the total bacterial/fungal concentration from the qPCR35. The absolute concentration was first transformed in the logarithmic scale (log10 (absolute concentration + 1)) and then analyzed in the logistic regression model. The association between the absolute microbial concentration and allergic/non-allergic rhinitis were conducted for species presented in > 10 classrooms with average relative abundance > 0.1%. Only associations with p-value < 0.005 were considered as significant throughout the study. The false discovery rate (FDR) was calculated by the p.adjust function with the Benjamini-Hochberg procedure in R.
Results
Demographics and prevalence of rhinitis
In this study, we collected vacuum dust from 30 classrooms of 8 junior high schools in Terengganu, Malaysia. Four schools were from the urban area and four schools from the rural area. Self-administered questionnaires and skin prick test results were obtained from 370 students in the selected classes. Among them, 23 students reported recent respiratory infections in the last 3 months and thus were removed from the analyses. In total, 347 students were included in the final dataset, included 129 males and 218 females. Based on health information from questionnaire and skin prick tests, we defined 107 students as healthy subjects, 110 students with allergic rhinitis and 130 students with non-allergic rhinitis. The prevalence of house dust mite and cat allergy was high among students (38.3% and 11.0%; Table S1). There was no difference between male and female students in the prevalence of allergic and non-allergic rhinitis (p = 0.98; Table 1). Ten students reported current smoking, and there was no difference between smoking and non-smoking students in the prevalence of allergic and non-allergic rhinitis (p = 0.38).
Species-level microbiome composition and variation
The abundance of microbial taxa at the taxonomical class and species level were analyzed first. Alphaproteobacteria (17.2%), Gammaproteobacteria (15.9%), Cyanobacteria (14.3%), Actinobacteria (12.9%), Bacilli (10.5%), Deinococci (7.2%) and Betaproteobacteria (7.2%) were the high abundant bacterial classes in the classrooms.Acinetobacter johnsonii (4.0%), Aliterella Antarctica(2.7%), Klebsiella pneumonia (2.0%), Methylobacterium radiotolerans (1.9%) and Saccharopolyspora halotolerans (1.9%) were the high abundant bacterial species (Figure 1A; Table S2). For fungi, Dothideomycetes (30.7%), Eurotiomycetes (20.6%), Wallemiomycetes (12.2%), Agaricomycetes (11.9%), Sordariomycetes (7.8%) and Agaricostibomycetes (5.8%) were the high abundant fungal classes in the classrooms. Wallemia sebi (10.9%), an unidentified Pleosporales (7.1%), Sterigmatomyces halophilus(5.8%), Aspergillus penicillioides (5.7%), Eupenidiella venezuelensis (5.3%) and Bambusaria bambusae (5.2%) were the high abundant fungal species in the junior high school of Terengganu (Figure 1B; Table S3).
We further characterize and visualized the overall microbial compositional variation among classrooms by non-metric multidimensional scaling (NMDS) analysis. The bacterial and fungal composition in urban schools did not differ from rural schools (Figure 2A and 2B). Samples collected in the same school were compositionally more similarity compared with samples collected in different schools (Mann-Whitney U test on Bray-Curtis distance matrix, bacteria p = 0.003, fungi p < 0.001). However, there were also exceptions that samples from the same school had a low compositional similarity. For example, the bacterial composition in T7.3 (school 7 class 3) differed drastically from the bacterial composition in T7.1, T7.2, and T7.4, due to high abundance of K. pneumoniae in T7.3 (46.2%; Figure 1A and 2A).
Associations between indoor microbial richness/concentration and allergic/non-allergic rhinitis
In total, 4,977 bacterial OTUs (operational taxonomic units) were identified in this study. The overall bacterial richness (represented as the number of OTUs) in each classroom was not associated with the occurrence of rhinitis in the classroom (p > 0.05; Table 2). The same test was also conducted for the major bacterial classes (> 200 OTUs). The richness in Gammaproteobacteria was protectively/negatively associated with allergic and non-allergic rhinitis (p = 0.02 and 0.04). Similarly, 2,632 fungal OTUs were identified in this study, and the overall fungal richness was positively associated with allergic and non-allergic rhinitis (p = 0.01 and 0.03). We did not find significant associations between specific fungal class and the occurrence of rhinitis (p > 0.05). The richness in Agaricomycetes had a small trend to be associated with allergic rhinitis (p = 0.06) and non-allergic rhinitis (p = 0.10).
We further analyzed the associations between the absolute concentration of indoor bacterial and fungal species and the occurrence of allergic and non-allergic rhinitis. To control the number of tests, only species presented in at least 10 classrooms and mean abundance > 0.1% were included in the regression analyses. Brasilonema bromeliae from the phylum of Cyanobacteria and Aeromonas enteropelogenes and Escherichia fergusonii from the class of Gammaproteobacteria were protectively associated with allergic rhinitis (p < 0.005, FDR = 0.17; Table 3). These species were presented in 21, 10 and 23 classrooms out of the total 30 classrooms, and the geometric mean absolute concentration was 4.1*105, 1.2*105 and 2.6*105 copies per gram dust. Deinococcus grandis from the class of Deinococci andChaetomium grande from the fungal class Sordariomycetes were positively associated with allergic rhinitis (p < 0.005, FDR = 0.09 and 0.18). D . grandis was presented in 15 classrooms with a geometric mean concentration as 1.2*105 copies per gram dust. C . grande was presented in 11 classrooms with a geometric mean of concentration as 399 copy per gram dust. The results indicate that the concentration of health-associated bacterial species is three orders of magnitude higher than the fungal species in the classrooms of Terengganu.
Interestingly, we also found B. bromeliae and A. enteropelogenes were protectively associated with non-allergic rhinitis (p = 0.002 and 0.003, FDR = 0.07; Table 4). Enterobacter xiangfangensis from the class of Gammaproteobacteria,Streptococcus salivarius from the class of Bacilli andPatulibacter minatonensis from the class of Thermoleophilia were also protectively associated with non-allergic rhinitis (p < 0.05, FDR < 0.1). Deinococcus gobiensis was the only bacterial species that were positively associated with non-allergic rhinitis. These species were presented in 10 to 23 classrooms in Terengganu with a geometric mean concentration from 1.2*105 to 6.0*105 copy per gram dust. No fungal species was associated with the occurrence of non-allergic rhinitis.
No association between environmental characteristics and rhinitis-related species
We further tested the associations between environmental characteristics and the identified rhinitis-related species. Environmental characteristics were tested separately with the rhinitis-related species, including location of schools (urban/rural), the concentration of house dust mite (Der p 1 and Der f 1) and cockroach allergen (Bla g 1 and Per a 1), indoor CO2, NO2 levels and relative humidity and indoor visible dampness and mold. We did not find any significant associations between the environmental characteristics and the rhinitis-related species (p > 0.05).
Discussion
Strengths and limitations of the study
Here, we report the first association study between indoor microbiome exposure and the occurrence of rhinitis. With this approach, we identified a list of potential health-related species and found similar association patterns between allergic rhinitis and non-allergic rhinitis, which can be further tested in future studies. Another strength is that we applied high-resolution amplicon sequencing (PacBio) to characterize the bacterial and fungal composition and diversity. The species-level taxonomic resolution facilitates accurate microbial diversity estimation compared to standard second-generation sequencing, which can only resolve taxa at the genus level. Also, we applied absolute concentration rather than the relative abundance of microbial species in the association analysis. Previous studies confirmed that absolute concentration is a superior approach compared with the relative abundance in terms of finding the correct microorganism-phenotype associations 35.
There are also limitations in this study. The cross-sectional study design limits the causal inference in the relationship between indoor microbial exposure and rhinitis. The cross-sectional study design is widely used in epidemiological studies, as it is an economical approach to obtain health associations among populations. However, this study design cannot draw the causal relationships, and thus future longitudinal studies are needed to verify the patterns we observed in this study. Also, we applied amplicon sequencing to characterize microbiome composition in this study. Since only one marker gene was sequenced, this approach can provide microbial taxonomic information rather than accurate functional inference, which can be achieved by the shotgun metagenomics sequencing. We conducted the false discovery rate correction in this study, which is one of the widely used approaches to adjust the type I error in null hypothesis testing when conducting multiple comparisons36. The associations were not significant after adjustment with the Benjamini-Hochberg (BH) procedure (0.06 < FDR < 0.19). However, the BH procedure was confirmed to be over-conservative in microbiome analysis with much-reduced power in detecting the accurate significant findings. To solve the problem, a new approach DS-FDR has been proposed to improve the sensitivity of detection37. However, the new approach can only be applied in the microbiome dataset with relative abundance, and thus it is not applicable in this study. Future approaches dealing with microbiome dataset with transformed absolute concentration are needed to solve the issue.
Prevalence of allergic/non-allergic rhinitis and environmental stimuli
Rhinitis is one of the most prevalent diseases in the world. It is estimated that allergic rhinitis affects approximately 10%-40% of the population in different countries3. The international collaborated ISAAC and ECRHS studies reported a list of countries with high prevalence in allergic rhinitis, including Nigeria (>35%), Paraguay (30-35%), Sweden (31%), Hong Kong (25-30%) Australia (15-20%), New Zealand (15-20%), the United Kingdom (15-20%%) 2, 38, 39. In this study, we reported that the prevalence of allergic rhinitis was 31.7% among junior high schools students in Terengganu, Malaysia, which is comparable with the countries with high prevalence. The prevalence of non-allergic rhinitis is less well defined, and no international collaboration has been conducted to survey the prevalence among different countries systematically. One study estimated that in the United States, 17 million people had non-allergic rhinitis and 22 million people had mixed rhinitis, a combination of allergic and non-allergic rhinitis40. Another study reported that the prevalence of non-allergic rhinitis was 9.6% among high school students in Belgium41. The prevalence of non-allergic rhinitis is higher in our study (37.5%) compared with previous studies. In this study, we applied skin prick tests to define allergic and non-allergic rhinitis, by using nine allergen extracts from plants, animals, mites and molds. The allergic reaction for students in Terengganu was mainly from the sensitization of house dust mites and cat allergen, which is consistent with a previous study in Malaysia that approximately 50% and 25% office workers had mites and cat allergy42.
The mechanisms for allergic and non-allergic rhinitis are suggested to be different. Allergic rhinitis is related to the nasal inflammation caused by IgE-mediated allergy. Exposure to the indoor allergen and environmental stimuli lead to overproduction of IgE (immunoglobulin E), which can further activate the receptors on the surface of cells and lead to the allergic and inflammatory symptoms43. Compared with allergic rhinitis, the mechanisms for non-allergic rhinitis are less clear. Classic T-cell and cytokine inflammation are suggested to be the major mechanisms for non-allergic rhinitis1. Neurogenic dysregulation is another proposed pathway that leads to nasal obstruction and rhinorrhea1, 44. Rhinitis of the elderly is suggested to be mainly associated with the neural dysregulation.
Although the mechanisms for allergic and non-allergic rhinitis are quite different, the environmental factors associate with the development of rhinitis could be similar. For example, environmental and traffic pollution, including NOx, SOx and ozone, environmental tobacco smoking, mold growth and chemical exposures, including volatile organic compounds (VOCs) are related to both allergic and non-allergic rhinitis7, 8, 45, 46. In this study, we found that the association pattern for indoor microbiome was similar between the two types of rhinitis. The phenomenon was supported by analysis in microbial richness and absolute concentration. For example, the richness (number of observed OTUs) in Gammaproteobacteria and in fungi was associated with both allergic and non-allergic rhinitis. The absolute concentration of B. bromeliae and A. enteropelogenes and two species from the genus of Deinococcus were also associated with both allergic and non-allergic rhinitis. The detailed mechanisms for how microorganisms affect nasal inflammation still need to be explored. It is possible that certain microorganisms can stimulate the inflammatory symptoms by multiple pathways, either with allergen sensitization or disturb the epithelium barrier. Future studies can explore the mechanisms in depth by using the microorganisms identified in this study. Epidemiological surveys should also consider to routinely survey the indoor microbiome exposure, together with other environmental factors, to add a new dimension to the exposure science.
Health effects of the microbial species
As this is the first association study between indoor microbiome exposure and rhinitis, we cannot directly compare our results with studies with the same experimental design, but there are some studies available with similar framework. One study in junior high school of Johor Bahru, Malaysia reported that the levels of total endotoxin (lipopolysaccharide, LPS) was negatively associated with rhinitis47. LPS is a bacterial cell-wall component and is widely used as a chemical marker for Gram-negative bacteria. In this study, we found that the diversity of Gammaproteobacteria, one of the most diverse classes of Gram-negative bacteria, was protectively associated with rhinitis. Another study reported that the total fungal DNA, assessed by quantitative PCR, was a potential risk factor for rhinitis in the indoor environment23. This is also consistent with our finding in this study that total fungal diversity was positively associated with the occurrence of rhinitis.
As this is the first association study for indoor microbiome and rhinitis, all these species were first time reported to be associated rhinitis. Several of the identified species were from the outdoor environment. Two of the environmental species, including B. bromeliae , P. minatonensis , were protectively (negatively) associated with rhinitis, and two of the environmental species, including D. gobiensis , D. grandis , were positively associated with rhinitis. Interestingly, the two potential risk species,D. gobiensis and D. grandis were both extremely radiation-resistant bacteria48, 49. These species are physically protected from the UV damages due to their efficient DNA repair system50. No study reported their health effects for humans, and it is unclear whether the radiation-resistant feature of these bacteria is related to the potential adverse health effect. Four rhinitis-related bacteria species, including E. fergusonii , A. enteropelogenes , E. xiangfangensis and S. salivarius , were protectively associated the rhinitis, and interestingly, these bacteria were all reported to be facultative pathogens that can cause various diseases in human and animals, such as respiratory infections, diarrhea and neutropenia51-54. AlthoughS. salivarius is a facultative pathogen that can infect immunocompromised patients, the species can also be used as probiotic that could reduced 80% of nasal streptococcus infections in adults55. The results suggest complex health effects of the species. The “hygiene hypothesis” proposes that having respiratory infections could reduce the occurrence of allergic diseases for children56. The role of infectious pathogens and the development of asthma in children have also been fiercely debated over the past years57, 58. Some studies suggest that rhinovirus infections and respiratory syncytial virus (RSV) infection were associated with increased odds of wheeze and asthma57, 59, but other population-based studies reported that exposure to infections diseases could lead to reduced atopic disorders 60. The potential protective effects of childhood infections are suggested to relate to the training and maturation of the immune system from the exposure 58. A recent study in university dormitories reported that exposed to facultative pathogens, including Haemophilus , Klebsiella ,Buttiauxella , and Raoultella , increased the odds of having respiratory infections61. Overall, more studies are needed to verify the pattern we observed in this study and further disentangle the health effects of the indoor microbiome on different respiratory diseases.
Acknowledgements
We thank the Natural Science Foundation of Guangdong Province (2020A1515010845) and the Department of Education of Guangdong Province (2018KTSCX021) for financial support. We thank Personalbio (www.personalbio.cn) for help in sequencing analysis.
References
1. Hellings PW, Klimek L, Cingi C, Agache I, Akdis C, Bachert C, et al. Non-allergic rhinitis: Position paper of the European Academy of Allergy and Clinical Immunology. Allergy 2017; 72:1657-65.
2. Bousquet J, Khaltaev N, Cruz AA, Denburg J, Fokkens WJ, Togias A, et al. Allergic Rhinitis and its Impact on Asthma (ARIA) 2008*. Allergy 2008; 63:8-160.
3. Brożek JL, Bousquet J, Agache I, Agarwal A, Bachert C, Bosnic-Anticevich S, et al. Allergic Rhinitis and its Impact on Asthma (ARIA) guidelines—2016 revision. Journal of Allergy and Clinical Immunology 2017; 140:950-8.
4. Togias A, Gergen PJ, Hu JW, Babineau DC, Wood RA, Cohen RT, et al. Rhinitis in children and adolescents with asthma: Ubiquitous, difficult to control, and associated with asthma outcomes. Journal of Allergy and Clinical Immunology 2019; 143:1003-11.e10.
5. Hox V, Steelant B, Fokkens W, Nemery B, Hellings PW. Occupational upper airway disease: how work affects the nose. Allergy 2014; 69:282-91.
6. Kowalski ML, Asero R, Bavbek S, Blanca M, Blanca-Lopez N, Bochenek G, et al. Classification and practical approach to the diagnosis and management of hypersensitivity to nonsteroidal anti-inflammatory drugs. Allergy 2013; 68:1219-32.
7. Shusterman D. Nonallergic Rhinitis: Environmental Determinants. Immunology and Allergy Clinics of North America 2016; 36:379-99.
8. Dunlop J, Matsui E, Sharma HP. Allergic Rhinitis: Environmental Determinants. Immunology and Allergy Clinics of North America 2016; 36:367-77.
9. Fu X, Li Y, Yuan Q, Cai G-h, Deng Y, Zhang X, et al. Continental-Scale Microbiome Study Reveals Different Environmental Characteristics Determining Microbial Richness, Composition, and Quantity in Hotel Rooms. mSystems 2020; 5:e00119-20.
10. Sun Y, Fu X, Li Y, Yuan Q, Ou Z, Lindgren T, et al. Shotgun Metagenomics Of Dust Microbiome From Flight Deck And Cabin In Civil Aviation Aircraft. Indoor Air 2020; 00:1-14.
11. Gilbert JA, Blaser MJ, Caporaso JG, Jansson JK, Lynch SV, Knight R. Current understanding of the human microbiome. Nat Med 2018; 24:392-400.
12. Lloyd-Price J, Mahurkar A, Rahnavard G, Crabtree J, Orvis J, Hall AB, et al. Strains, functions and dynamics in the expanded Human Microbiome Project. Nature 2017; 550:61-6.
13. Lax S, Sangwan N, Smith D, Larsen P, Handley KM, Richardson M, et al. Bacterial colonization and succession in a newly opened hospital. Sci Transl Med 2017; 9.
14. Lax S, Smith DP, Hampton-Marcell J, Owens SM, Handley KM, Scott NM, et al. Longitudinal analysis of microbial interaction between humans and the indoor environment. Science 2014; 345:1048-52.
15. Huttenhower C, Gevers D, Knight R, Abubucker S, Badger JH, Chinwalla AT, et al. Structure, function and diversity of the healthy human microbiome. Nature 2012; 486:207-14.
16. Bello MGD, Knight R, Gilbert JA, Blaser MJ. Preserving microbial diversity. Science 2018; 362:33-4.
17. Fu X, Norbäck D, Yuan Q, Li Y, Zhu X, Hashim JH, et al. Indoor microbiome, environmental characteristics and asthma among junior high school students in Johor Bahru, Malaysia. Environment International 2020; 138:105664.
18. Kirjavainen PV, Karvonen AM, Adams RI, Täubel M, Roponen M, Tuoresmäki P, et al. Farm-like indoor microbiota in non-farm homes protects children from asthma development. Nature Medicine 2019; 25:1089-95.
19. Pekkanen J, Valkonen M, Täubel M, Tischer C, Leppänen H, M. Kärkkäinen P, et al. Indoor bacteria and asthma in adults: a multicentre case–control study within ECRHS II; 2018.
20. Dannemiller KC, Gent JF, Leaderer BP, Peccia J. Indoor microbial communities: Influence on asthma severity in atopic and nonatopic children. Journal of Allergy and Clinical Immunology 2016; 138:76-83.e1.
21. Ege MJ, Mayer M, Normand AC, Genuneit J, Cookson WO, Braun-Fahrlander C, et al. Exposure to environmental microorganisms and childhood asthma. N Engl J Med 2011; 364:701-9.
22. O’Connor GT, Lynch SV, Bloomberg GR, Kattan M, Wood RA, Gergen PJ, et al. Early-life home environment and risk of asthma among inner-city children. Journal of Allergy and Clinical Immunology 2018; 141:1468-75.
23. Norback D, Hashim JH, Cai GH, Hashim Z, Ali F, Bloom E, et al. Rhinitis, Ocular, Throat and Dermal Symptoms, Headache and Tiredness among Students in Schools from Johor Bahru, Malaysia: Associations with Fungal DNA and Mycotoxins in Classroom Dust. PLoS One 2016; 11:e0147996.
24. Wang J, Engvall K, Smedje G, Norbäck D. Rhinitis, asthma and respiratory infections among adults in relation to the home environment in multi-family buildings in Sweden. PloS one 2014; 9:e105125-e.
25. Ma’pol A, Hashim JH, Norback D, Weislander G, Hashim Z, Isa ZM. FeNO level and allergy status among school children in Terengganu, Malaysia. J Asthma 2019:1-8.
26. Eid J, Fehr A, Gray J, Luong K, Lyle J, Otto G, et al. Real-time DNA sequencing from single polymerase molecules. Science 2009; 323:133-8.
27. Travers KJ, Chin C-S, Rank DR, Eid JS, Turner SW. A flexible and efficient template format for circular consensus sequencing and SNP detection. Nucleic acids research 2010; 38:e159-e.
28. Nadkarni MA, Martin FE, Jacques NA, Hunter N. Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 2002; 148:257-66.
29. Wang Y, Song F, Zhu J, Zhang S, Yang Y, Chen T, et al. GSA: Genome Sequence Archive. Genomics Proteomics Bioinformatics 2017; 15:14-8.
30. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods 2010; 7:335-6.
31. Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 2010; 26:2460-1.
32. Quast C, Pruesse E, Gerken J, Peplies J, Yarza P, Yilmaz P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Research 2012; 41:D590-D6.
33. Koljalg U, Nilsson RH, Abarenkov K, Tedersoo L, Taylor AF, Bahram M, et al. Towards a unified paradigm for sequence-based identification of fungi. Mol Ecol 2013; 22:5271-7.
34. Kruskal JB. Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika 1964; 29:1-27.
35. Dannemiller KC, Lang-Yona N, Yamamoto N, Rudich Y, Peccia J. Combining real-time PCR and next-generation DNA sequencing to provide quantitative comparisons of fungal aerosol populations. Atmospheric Environment 2014; 84:113-21.
36. Noble WS. How does multiple testing correction work? Nature Biotechnology 2009; 27:1135-7.
37. Jiang L, Amir A, Morton JT, Heller R, Arias-Castro E, Knight R. Discrete False-Discovery Rate Improves Identification of Differentially Abundant Microbes. mSystems 2017; 2:e00092-17.
38. Asher MI, Montefort S, Bjorksten B, Lai CK, Strachan DP, Weiland SK, et al. Worldwide time trends in the prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and eczema in childhood: ISAAC Phases One and Three repeat multicountry cross-sectional surveys. Lancet 2006; 368:733-43.
39. Bjerg A, Ekerljung L, Middelveld R, Dahlén S-E, Forsberg B, Franklin K, et al. Increased prevalence of symptoms of rhinitis but not of asthma between 1990 and 2008 in Swedish adults: comparisons of the ECRHS and GA²LEN surveys. PloS one 2011; 6:e16082-e.
40. Settipane RA, Lieberman P. Update on nonallergic rhinitis. Ann Allergy Asthma Immunol 2001; 86:494-507; quiz -8.
41. Bachert C, Van Cauwenberge P, Olbrecht J, Van Schoor J. Prevalence, classification and perception of allergic and nonallergic rhinitis in Belgium. Allergy 2006; 61:693-8.
42. Lim FL, Hashim Z, Than LTL, Md Said S, Hisham Hashim J, Norbäck D. Asthma, Airway Symptoms and Rhinitis in Office Workers in Malaysia: Associations with House Dust Mite (HDM) Allergy, Cat Allergy and Levels of House Dust Mite Allergens in Office Dust. PloS one 2015; 10:e0124905-e.
43. Poole JA, Rosenwasser LJ. The role of Immunoglobulin E and immune inflammation: Implications in allergic rhinitis. Current Allergy and Asthma Reports 2005; 5:252-8.
44. Baroody FM. Nonallergic Rhinitis: Mechanism of Action. Immunol Allergy Clin North Am 2016; 36:279-87.
45. Junker MH, Danuser B, Monn C, Koller T. Acute sensory responses of nonsmokers at very low environmental tobacco smoke concentrations in controlled laboratory settings. Environmental health perspectives 2001; 109:1045-52.
46. Andersen I, Lundqvist GR, Mølhave L, Pedersen OF, Proctor DF, Vaeth M, et al. Human response to controlled levels of toluene in six-hour exposures. Scand J Work Environ Health 1983; 9:405-18.
47. Norbäck D, Hashim JH, Markowicz P, Cai G-H, Hashim Z, Ali F, et al. Endotoxin, ergosterol, muramic acid and fungal DNA in dust from schools in Johor Bahru, Malaysia — Associations with rhinitis and sick building syndrome (SBS) in junior high school students. Science of The Total Environment 2016; 545-546:95-103.
48. Satoh K, Onodera T, Omoso K, Takeda-Yano K, Katayama T, Oono Y, et al. Draft Genome Sequence of the Radioresistant Bacterium &lt;span class=&quot;named-content genus-species&quot; id=&quot;named-content-1&quot;&gt;Deinococcus grandis&lt;/span&gt;, Isolated from Freshwater Fish in Japan. Genome Announcements 2016; 4:e01631-15.
49. Yuan M, Zhang W, Dai S, Wu J, Wang Y, Tao T, et al. Deinococcus gobiensis sp. nov., an extremely radiation-resistant bacterium. International Journal of Systematic and Evolutionary Microbiology 2009; 59:1513-7.
50. Kawaguchi Y, Yang Y, Kawashiri N, Shiraishi K, Takasu M, Narumi I, et al. The possible interplanetary transfer of microbes: assessing the viability of Deinococcus spp. under the ISS Environmental conditions for performing exposure experiments of microbes in the Tanpopo mission. Orig Life Evol Biosph 2013; 43:411-28.
51. Sanders WE, Jr., Sanders CC. Enterobacter spp.: pathogens poised to flourish at the turn of the century. Clinical microbiology reviews 1997; 10:220-41.
52. Janda JM, Abbott SL. The Genus Aeromonas: Taxonomy, Pathogenicity, and Infection. Clinical Microbiology Reviews 2010; 23:35.
53. Tunkel AR, Sepkowitz KA. Infections caused by viridans streptococci in patients with neutropenia. Clin Infect Dis 2002; 34:1524-9.
54. Gaastra W, Kusters JG, van Duijkeren E, Lipman LJA. Escherichia fergusonii. Veterinary Microbiology 2014; 172:7-12.
55. Di Pierro F, Adami T, Rapacioli G, Giardini N, Streitberger C. Clinical evaluation of the oral probiotic Streptococcus salivarius K12 in the prevention of recurrent pharyngitis and/or tonsillitis caused by Streptococcus pyogenes in adults. Expert Opin Biol Ther 2013; 13:339-43.
56. Strachan DP. Hay fever, hygiene, and household size. BMJ (Clinical research ed.) 1989; 299:1259-60.
57. Eder W, Ege MJ, von Mutius E. The asthma epidemic. N Engl J Med 2006; 355:2226-35.
58. von Mutius E. Infection: friend or foe in the development of atopy and asthma? The epidemiological evidence. European Respiratory Journal 2001; 18:872.
59. Régnier SA, Huels J. Association between respiratory syncytial virus hospitalizations in infants and respiratory sequelae: systematic review and meta-analysis. Pediatr Infect Dis J 2013; 32:820-6.
60. Smits HH, Hartgers FC, Yazdanbakhsh M. Helminth infections: Protection from atopic disorders. Current Allergy and Asthma Reports 2005; 5:42-50.
61. Fu X, Li Y, Meng Y, Yuan Q, Zhang Z, Norbäck D, et al. Associations between respiratory infections and bacterial microbiome in student dormitories in Northern China. Indoor Air 2020; 00:1-11.
Figure 1. Relative abundance of top (A) bacteria and (B) fungal species level in the classrooms.
Table 1. Prevalence of allergic and non-allergic rhinitis among junior high school students (N=347) in Terengganu, Malaysia. P-values were calculated by Chi-square test.