Comparison of Culturable Antibiotic-Resistant Bacteria in Polluted and Non- Polluted Air in Beijing, China T

Environment International 131 (2019) 104936

Contents lists available at ScienceDirect

Environment International

journal homepage: www.elsevier.com/locate/envint

Comparison of culturable antibiotic-resistant bacteria in polluted and non- polluted air in Beijing, China T

Yixin Maoa, Pei Dinga, Youbin Wanga, Cheng Dinga, Liping Wua, Ping Zhenga,b, Xiao Zhanga, ⁎ Xia Lia, Leyao Wangc, Zongke Suna, a Department of Environmental Microbiology, National Institute of Environmental Health, Chinese Center for Disease Control and Prevention, Beijing 100021, China b Department of Women's, Children's, and Adolescents' Environmental Health, National Institute of Environmental Health, Chinese Center for Disease Control and Prevention, Beijing 100050, China c School of Medicine, Institute of Public Health, Washington University, St. Louis, MO 63110, USA

ARTICLE INFO ABSTRACT

Handling Editor: Yong-Guan Zhu Background: Air pollution has been a serious health issue in Beijing for years. Airborne antibiotic-resistant Keywords: bacteria could be a potential health crisis as reserve of antibiotic resistance transmission in environment. The Antibiotic resistance composition and antibiotic resistance pattern of culturable bacterial community and how these are affected by Atmosphere air pollution remain unclear. Culturable bacteria Objectives: This study aimed to compare the compositions and antibiotic resistance patterns of culturable bac- Polluted air teria in polluted and non-polluted weather conditions in Beijing. Beijing smog Methods: Air samples were collected indoors and outdoors during polluted and non-polluted weather using six- Multidrug-resistant stage Andersen Samplers. For each isolated bacterium, the 16S ribosomal RNA gene was amplified, sequenced, and blasted against the National Center for Biotechnology Information database Antibiotic resistance was conducted by antimicrobial susceptibility testing. Results: Bacterial concentration in polluted weather was significantly higher than in non-polluted weather, both indoors and outdoors (P < 0.05). Gram-positive bacteria (GPB) were dominant in both weathers but gram- negative bacteria (GNB) were more abundant in polluted weather than non-polluted weather both indoors and outdoors. Multidrug-resistant (MDR) bacteria occupied 23.7% of all bacterial isolates, 22.4% of isolates from polluted weather and 27.8% of isolates from non-polluted weather. Penicillins were resisted by 72.4% and 83.3% of isolates from polluted and non-polluted weather, respectively. Conclusions: The bacterial concentration was significantly higher in polluted weather, compared to non-polluted weather. Polluted weather is correlated with changes in the bacterial composition in the air, with a greater abundance of GNB. Penicillins was resisted by over 70% of bacterial isolates. The abundance of MDR bacteria suggested potential risks for human health.

1. Introduction 2016). Many studies have been carried out on airborne bacteria in environment (Fang et al., 2007). And it has been well documented that Air pollution is a serious global health issue. Exposure to particulate the air within hospitals or stock farms is highly contaminated with matter (PM) pollutants in the atmosphere increases risks for diseases in bacteria (Clark et al., 1983; Park et al., 2013; Solomon et al., 2017). the respiratory and nervous systems and decreases life expectancy Microorganisms are believed to have varying concentration in the (Allen et al., 2013, 2014; Hunt et al., 2003; Maher et al., 2016; Pope 3rd ambient air depending on the pollution state (Liu et al., 2018; Zhai et al., 2009; Weuve et al., 2012). As one of the major components and et al., 2018). pathogenic agents, airborne microorganisms, such as bacteria, are Bacterial infections are becoming increasingly diﬃcult to treat due considered to be closely related to air pollution that causes human al- to the emergence of antibiotic resistance (AR or ABR), when bacteria lergic responses and respiratory diseases (Kim et al., 2018; Wu et al., evolve in response to antibiotics and make them ineﬀective. ABR is a

⁎ Corresponding author. E-mail addresses: [email protected] (Y. Mao), [email protected] (P. Ding), [email protected] (Y. Wang), [email protected] (C. Ding), [email protected] (L. Wu), [email protected] (P. Zheng), [email protected] (X. Zhang), [email protected] (X. Li), [email protected] (L. Wang), [email protected] (Z. Sun). https://doi.org/10.1016/j.envint.2019.104936 Received 15 February 2019; Received in revised form 15 June 2019; Accepted 15 June 2019 0160-4120/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). Y. Mao, et al. Environment International 131 (2019) 104936

global health crisis that leads to increased morbidity, mortality, and sampling, including temperature, relative humidity, SO2,NO2, CO, O3, medical costs. In China, 92,700 tons of 36 frequently detected anti- PM10 and PM2.5. Real time environmental data were obtained from the biotics were used, according to Zhang's investigation, and 53,800 tons National Urban Air Quality Real-time Distribution website (http://106. of which eventually entered the environment (Zhang et al., 2015). The 37.208.233:20035/), published by China National Environmental development of antibiotic resistance among bacteria has been fa- Monitoring Centre. cilitated by the misuse and abuse of antibiotics in humans and animals. Moreover, antibiotic resistance can be transferred between bacteria in 2.2. Sample culture and counting environment (Allen et al., 2010; Hiltunen et al., 2017; Xie et al., 2019). Large amounts of antibiotic resistant genes were found in air samples Each of the sample plate collected at each stage of the sampler was from Beijing smog (Pal et al., 2016), causing great public concern and immediately incubated at 37 °C for 48 h, allowing the bacterial aerosols panic. The existence of antibiotic resistant genes indicates the potential to grow on the NA plates. The number of colony-forming units (CFUs) for resistance to be transferred between bacteria in air (Xie et al., 2019). was counted manually after 48 h. And concentrations were expressed as Antibiotic-resistant bacteria in the air, particularly pathogenic bacteria, CFU per cubic meter of air (CFU/m3). Because during sampling, there may cause respiratory diseases of a longer duration due to their re- existed a superposition when microbial particles impact the same spot sistance to antibiotic treatment. There have been studies on ABR in food through the same pore, counting results of each sample were statisti- or other environments, such as soil, waste water, or hospital settings cally corrected according to Andersen (Andersen, 1958). Samples from (Gao et al., 2018; Uhrbrand et al., 2017; Zhu et al., 2017). However, 2017/12/29 and 2018/3/1 were selected to represent polluted and few studies have evaluated the changes in the percentage of antibiotic- non-polluted weather, respectively, for further analysis. resistant bacteria in polluted and non-polluted air. The aim of this study was to compare the compositions and anti- 2.3. Bacterial isolation and identification biotic-resistant patterns of bacterial communities in ambient air samples collected under polluted and non-polluted weather conditions in a Each culturable colony from the original sample plates of selected typical site of Beijing. We focused on culturable airborne bacteria days (2017/12/29 and 2018/3/1) was picked using inoculation loop or through culture and antimicrobial susceptibility testing (AST) method, needle onto a Columbia Agar plate adding 5% sheep blood (COS; as the target bacteria is a prompt start of airborne AMR bacteria study BioMerieux) to isolate individual bacteria. These COS plates were in- and these live airborne bacteria could be obtained for further research. cubated at 37 °C for 48 h. The above isolation step was repeated at least twice to ensure that the colonies growing on each of the final culture 2. Material and methods plates had been purified without any other mixed bacteria. Total genomic DNA was extracted from each purified culturable 2.1. Study site and air sampling bacterium using a prepGEM® Bacteria DNA Extraction Kit (ZyGEM, New Zealand) following the manufacturer's instructions. The 16S ribosomal Air samples were collected from the sixth floor of the Experiment RNA (rRNA) gene was then amplified by polymerase chain reaction Building at the National Institute of Environmental Health (39°52′48″N, with the universal primer pair 27F/1492R and the amplified gene was 116°27′26″E; ~18 m above the ground). This site is located in the cloned and sequenced (ABI3730XL, Majorbio, China). Each purified southeast of Beijing between the 2nd and 3rd Ring Roads, with Long culturable bacterium was identified by performing Nucleotide Basic Tan Lake Park and Tian Tan Park to the west, Beijing Railway Station to Local Alignment Search Tool (BLAST) searches against the National the northwest, Cancer Hospital Chinese Academy of Medical Sciences to Center for Biotechnology Information's (NCBI's) 16S rRNA gene se- the southwest, Chui Yang Liu Hospital to the southeast, Fen Zhong quence reference database (https://blast.ncbi.nlm.nih.gov/Blast.cgi) Temple to the south and Beijing University of Technology to the east. and taxonomic information for each identified culturable bacterium Besides, this sampling site is surrounded by over 40 residential com- was obtained from the NCBI Taxonomy database (https://www.ncbi. munities within 2 km. nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi). The database also pro- Two six-stage Andersen Samplers (Thermo-Andersen, USA) were vides references of Gram information. used to collect air samples from inside and outside the Experiment Building separately at the same time, with an average flow rate of 2.4. Antimicrobial susceptibility testing (AST) of bacterial isolates 28.3 L/min for 10 min. The indoor samples were collected in a laboratory that was adjacent to the outdoor sampling site, with a single AST was conducted using the minimal inhibitory concentration closed door between the two sites. The Andersen sampler has six stages (MIC) agar dilution method with AST plates (Biofosun, China). The with different cut-off sizes that represent the human respiratory system wells of each AST plate were covered with specific antibiotics that in- (Table S1), with stages 3–6 representing the hazardous range where cluded nearly all common antibiotic classes for gram-positive bacteria particles have lung penetrability (Andersen, 1958). Nutrient agar (NA) (GPB) and gram-negative bacteria (GNB) (CLSI, 2017; Magiorakos formula (OXOID, England) were added to distilled water and mixed et al., 2012) in a gradient concentration (Tables S4–S6). Negative and well according to manufacturer's instructions. After sterilized by auto- positive control well did not cover with antibiotics. Antibiotic classes claving, culture medium was distributed into plastic culture plates. included penicillins, β-Lactam/β-Lactamase inhibitor combinations, After solidification, the culture plates were directly placed on all six cephems, lipopeptides, glycopeptides, penems, macrolides, lincosa- stages of the sampler to collect air samples. Both the outdoor and in- mides, Macrolides/Lincosamides combinations, pseudomonic acid, door samples were collected on 7 days chosen randomly of polluted and fluoroquinolones, folate pathway inhibitors, tetracyclines, phenicol, non-polluted weather, respectively, during heating season of Beijing and aminoglycosides (Table S6). (through November 2017 to March 2018) (Table S3). Single colonies of bacterial isolates were suspended in Nutrient The ‘polluted weather’ and ‘non-polluted weather’ in this study were Broth (NB) and diluted to a concentration of 1.5 × 108 CFU/mL ac- defined based on the AQI (air quality index) level obtained from China's cording to the McFarland standard. Each bacterial suspension was Ambient Air Quality Standards (GB 3095-2012) (China MEE, 2012a) added to one AST plates except for the negative control well, adding NB and Technical Regulation on Ambient Air Quality Index (HJ 633-2012) without bacterial suspension. The AST plates were then incubated at (China MEE, 2012b). Polluted weather (from Slightly polluted to severe 37 °C for 18–20 h. The plates were read manually according to the polluted level in Table S2) was defined as AQI of > 101, and non-pol- manufacturer's instructions and analyzed based on the MIC values luted weather (excellent weather level in Table S2) was defined as AQI published in the guidelines of the U.S. Clinical and Laboratory of < 50 in this study. Environmental parameters were recorded when Standards Institute (CLSI) (CLSI, 2017). Each bacterial isolate was

2 Y. Mao, et al. Environment International 131 (2019) 104936

) 40 3 35

25 Polluted outdoor

20 Polluted indoor

15 Non-polluted outdoor

10 Non-polluted indoor

Colony-forming units(CFU/m Colony-forming 5

0 L1 L2 L3 L4 L5 L6 Stages of Andersen Sampler

Fig. 1. Comparison of Quantity of culturable bacteria in the air in terms of the number of colony-forming units (CFUs). (A) Comparison of the quantity of bacteria under different weather conditions. (B) Comparison of the quantity of bacteria on each stage of the sampler under different weather conditions. categorized as susceptible, intermediate, or resistant to a specific anti- non-polluted weather, respectively. Plates on stage 5 had the highest biotic using MIC breakpoints. We set some breakpoints of antibiotics number of indoor samples in both polluted and non-polluted weather. where they were not included in the guideline, such as erythromycin (ERY) for GNB (Tables S4–S5). The AST results were analyzed using the software provided by the manufacturer. 3.2. Composition of culturable bacteria

A total of 249 and 32 colonies were picked from the original plates 3. Results on the samplers of selected days under polluted and non-polluted weather conditions (2017/12/29 and 2018/3/1), respectively. Of 3.1. Bacterial concentration in different weathers these, 133 isolates (53.4%) from the polluted weather samples and 25 isolates (78.1%) from the non-polluted weather samples were success- The total numbers of CFUs collected during polluted and non-pol- fully cultured. All culturable bacterial isolates were then sequenced and luted weather conditions are shown in Fig. 1A. Real time environmental identified through 16S rRNA sequencing. GPB were dominant in both parameters are shown in Fig. S3. CFUs of the polluted weather samples weather conditions, indoors and outdoors. However, GNB were more were significantly greater than those of the non-polluted weather abundant in polluted than non-polluted weather in both the indoor and samples, both indoors and outdoors (P < 0.05). outdoor samples (Fig. 2). Numbers of CFUs collected by 6 stages of the Andersen Sampler are The culturable bacteria collected in non-polluted weather were shown in Fig. 1B. Plates on stage 4 of the Andersen Sampler had the distributed across 10 genera, with Bacillus, Streptomyces, and highest and second to highest CFUs of outdoor samples in polluted and Rhodococcus being most abundant (Fig. 3). By contrast, in polluted

3 Y. Mao, et al. Environment International 131 (2019) 104936

weather samples and 58 isolates from the polluted weather samples. MDR bacteria were detected in 23.7% (18/76) of all culturable bacteria, 22.4% (13/58) of polluted weather and 27.8% (5/18) of non- polluted weather (Fig. 4A). In non-polluted weather group, 16.67% (3/ 18) of the MDR bacterial isolates were resistant to three antibiotic classes and 11.11% (2/18) were resistant to four classes, whereas in polluted weather group, 8.62% (5/58) of MDR bacterial isolates were resistant to three antibiotic classes and 6.90% (4/58) were resistant to five classes. In GPB group, 16.7% (11/66) MDR bacteria were detected, while in GNB group, 70% (7/10) were detected (Fig. 4B). Among GPB MDR bacterial isolates, 63.64% (7/11) were resistant to three antibiotic classes, 27.27% (3/11) were resistant to four classes, and 9.09% (1/11) were resistant to five classes. By contrast, 42.86% (3/7) of the GNB MDR bacterial isolates were resistant to five antibiotic classes and 28.57% (2/7) were resistant to six classes.

3.4. Antibiotic drug resistance in diﬀerent weathers

Percentage of all tested antibiotics resisted by culturable bacteria in different sample groups were calculated and shown in Fig. 5. Fig. 2. Percentage of gram-positive bacteria (GPB) and gram-negative bacteria Penicillins were resisted by most of the culturable bacterial isolates (GNB) occupation in the indoor and outdoor samples in different weathers. in polluted (72.4%, 42/58) and non-polluted (83.3%, 15/18) weather. Furthermore, penicillins were resisted by 100% (11/11) and 85.7% (6/ weather, the bacterial isolates were distributed across 43 genera, with 7) culturable bacterial isolates in MDR of GPB and GNB group, re- Bacillus, Kocuria, and Streptomyces being the top three most dominant spectively. genera. Streptomyces, Bacillus, Micrococcus, and Microbacterium were in Cephems were resisted by 29.3% (17/58) and 33.3% (6/18) cul- the top 10 most abundant genera in both weathers. Five genera de- turable bacterial isolates in polluted and non-polluted weather, re- tected in non-polluted weather (Rhodococcus, Curtobacterium, Iso- spectively. 90.0% (10/11) and 100.0% (10/10) culturable MDR bac- ptericola, Planococcus, and Variovorax) were not detected in polluted terial isolates resisted cephems in GPB and GNB group, respectively. weather samples. At the phylum level, Actinobacteria were dominant (68.4% in pol- 4. Discussion luted weather and 68.0% in non-polluted weather). At the order level, Micrococcales, Streptomycetales, and Corynebacteriales in Here, the quantity and composition of culturable airborne bacteria Actinobacteria, Bacillales in Firmicutes, and Burkholderiales in in the ambient air along with their antibiotic resistance patterns in Proteobacteria were all detected in both weathers and were most polluted and non-polluted weathers were surveyed in Beijing. Many abundant in their corresponding classes (Fig. S1). studies including airborne bacteria have been carried out in different environments in the past (Cao et al., 2014; de Rooij et al., 2019; Fang 3.3. Culturable multidrug-resistant (MDR) bacteria et al., 2006, 2007; Gao et al., 2018; He et al., 2017; Xie et al., 2019). Yet this is the first observation of the existence and composition analysis of In total, 76 bacterial isolates out of 158 (48.1%) were successfully culturable antibiotic-resistant bacteria in the ambient air from polluted tested through AST, including 18 isolates from the non-polluted and non-polluted weather conditions in Beijing using bacterial culture

Polluted Non-polluted

4% 12% 4% 4% 24% 4%

35% 12% 4%

10% 8%

2% 24% 3% 8% 6% 16% 6% 6%

Bacillus Kocuria Bacillus Streptomyces Streptomyces Arthrobacter Rhodococcus Cellulosimicrobium Microbacterium Micrococcus Micrococcus Curtobacterium Pseudomonas Ornithinimicrobium Isoptericola Microbacterium Moraxella Other 34 genus Planococcus Variovorax

Fig. 3. Proportion of top 10 culturable bacteria in air samples at the genus level in polluted and non-polluted weather.

4 Y. Mao, et al. Environment International 131 (2019) 104936

A B

35% 80%

30% 70% 60% 25% 50% 20% 40% 15% Polluted GPB 30% Percentage Percentage 10% Non-polluted GNB 20% 5% 10% 0% 0% Polluted Non-polluted GPB GNB Weather Gram group

Fig. 4. Comparison of proportion of the culturable multidrug-resistant (MDR) bacteria between (A) polluted and non-polluted weather conditions, and (B) gram- positive bacteria (GPB) and gram-negative bacteria (GNB) groups.

and outdoor environments have been associated with reduced health, including allergies, asthma, infectious diseases, such as severe acute respiratory syndrome, and even cancer (Douwes et al., 2003; Kim et al., 2018; Yoo et al., 2017), which may have high socio-economic impacts (Ghosh et al., 2015). Bacteria from environment, especially from ambient air, can be vulnerable through medium culture process. After several generations, those bacteria may become too weak to grow continually. In this case, a total of 56.2% bacterial isolates (53.4% from the polluted weather samples and 78.1% from the non-polluted weather samples) were found to be culturable airborne bacteria. GPB were dominant under both weather conditions, which was in agreement with other studies (Brandl et al., 2014; Fang et al., 2006; Górny and Dutkiewicz, 2002; Górny et al., 1999; Liang et al., 2013). However, in both the indoor and outdoor samples, GNB had a higher relative abundance in polluted weather than in non-polluted weather, which indicated a potential correlation between pollutants and GNB. Actinobacteria were most abundant at the phylum level, and multiple genera within Actinobacteria were soil inhabitants (e.g., Streptomyces)(Ventura et al., 2007), indicating that soil inhabitants may be one of the key components of the airborne bacterial Fig. 5. Percentage of culturable bacteria resisted to antibiotics in different community, which was also found by Cao et al. during a severe smog sample groups P: Polluted weather; NP: Non-polluted weather; GPB: gram-po- event (Cao et al., 2014). Bacteria has correlation with the ground dis- sitive bacteria; GNB: gram-negative bacteria; MDR: multidrug-resistant bac- turbance caused by human activities, which can stir up the dust and terial strains; All: All bacterial strains. small soil particles from the ground. The relationship between air pollutants and GNB need further research to explore. methods. To the best of our knowledge, this is also the first observation of the Results revealed that bacterial concentrations were significantly existence of culturable MDR bacteria in the ambient air in Beijing using higher in polluted weather compared to non-polluted weather, re- bacterial culture methods. MDR was defined as acquired non-suscept- gardless of whether they were collected indoors or outdoors. This could ibility to at least one agent in ≥3 antimicrobial categories (Magiorakos be attributed to the environmental conditions, sampling time, and et al., 2012). And ≥10 categories of antibiotics and their combinations bacteria's capability of adhering to PM in the ambient air. Air sampling were used to detect MDR bacteria in this study. A total of 23.7% of all using different sampler could bring various difference to bacterial tested culturable bacteria were MDR, which raises the question as to concentration. The limitation of using Andersen Samplers is that it whether these bacteria could serve as reservoirs of resistant genes in the could damage cells during sampling process which may cause missing air and could pass on this resistance to more bacteria, including pa- of a fraction of culturable bacteria. Still, it is a convenient and efficient thogenic species. MDR bacteria occupied 70% in GNB, compared to method for air sample collection. Bacteria concentrations in indoor air 17% in GPB. It is worth noticing that compare to GNB, GPB dominant were lower than in outdoor air under both weather conditions. Similar much more in ambient air, which we believe may influence this dif- findings have been reported (Shelton et al., 2002; Emerson et al., 2017). ference of MDR occupation percentage to a certain extent. Still, the In addition, results suggested that in polluted weather, most bacteria situation of MDR bacteria occupation in the ambient air was serious, collected outdoor and indoor are likely to influence the secondary approaching the percentage of airborne MDR bacteria that have been bronchi and terminal bronchi, respectively, indicating that people who isolated from swine feeding operations (Chapin et al., 2005) or waste- remain inside during polluted weather may still be at risk of bacterial water treatment plants (Teixeira et al., 2016). MDR bacteria occupied infection, particularly of the lower respiratory tract. Therefore, re- approximately one forth in both polluted or non-polluted weather, sidents should better use health protective equipment (e.g., an air which also suggest the serious situation of airborne MDR bacteria. And purifier) if they stay at home in polluted weather. It has previously been interestingly, the distribution of MDR bacteria is not dependent on the demonstrated that microbes can become attached to ambient particles weather conditions. and have significant effects on climate change (Brodie et al., 2007), However, it should be noticed that the difference in the proportion increasing the risk to human health. Biological agents in both indoor of MDR bacteria between different groups may influenced by the

5 Y. Mao, et al. Environment International 131 (2019) 104936 relatively small sample size of non-polluted weather and GNB. Also, between bacteria. (4) Penicillins was resisted by over 70% of bacterial some bacteria were unable to survive the entire AST process, which we isolates from the ambient air. These findings emphasized the im- believe was due to the following reasons: a) some airborne bacteria portance of studying the antibiotic resistance of bacteria in the air. cannot be cultured in NB; b) environmental bacteria can be vulnerable Future studies may focus on the pathway of AMR (and MDR) in the during bacterial culture, purification, isolation, and the AST process; c) human respiratory tract by obtaining a larger number of air samples the bacterial suspension used during the AST process was diluted to a more frequently and by designing cross-sectional or case-control studies certain concentration according to the manufacturer's instructions, to determine whether the isolation and virulence of pathogens from the which may not be suitable for certain bacteria; and d) there were fewer air are associated with human or clinical diseases. original bacteria in the non-polluted weather samples than in the polluted weather samples, and fewer original GNB in the air than GPB. Funding Nevertheless, MDR bacteria were detected in both non-polluted and polluted weather, and the actual proportion of MDR bacteria in the This work was supported by the National Key Research and atmosphere would be no less than we detected in the present study. Development Program of China (No. 2017YFC0702800); and the Penicillins were found being resisted by > 70% airborne culturable Health Impact of Air Pollution Program of Chinese Center for Disease bacteria in both weathers in this study, followed by cephems being Control and Prevention. resisted by approximately 30% bacteria. These two classes of antibiotics were also resisted by > 85% MDR bacteria. These findings are con- Declaration of Competing Interest sistent with β-lactams (including penicillins and cephems) being one of the most highly used antibiotics in China. And Li et al. have also found The authors declare they have no actual or potential competing fi- the β-lactams resistance gene being most abundant in air of a city in nancial interests. China (Li et al., 2018). The main groups of antibiotics for human use in China are similar to those used in other countries (Zhang et al., 2015); Acknowledgments moreover, penicillins, along with macrolides and fluoroquinolones, are the main antibiotic active ingredients sold in veterinary medicinal We thank all participants. We thank professor S Tong who provided products in the US. People should be aware of this alarming phenom- help and assistance during our research and manuscript writing. We enon and reduce the unnecessary usage of antibiotics. thank J Ban and Y Hu who provided suggestions during our manuscript This study evaluated the airborne antibiotic-resistant bacteria pat- writing. We also thank the editor and reviewers for improving the terns in different weather conditions in Beijing. Airborne AMR bacteria quality of our work. could be causing health problems by contributing to the conservation or expansion of the antibiotic resistance and resistant gene pool in the Appendix A. Supplementary data environment, thereby increasing the risk of resistance transfer to pathogenic microorganisms. It has been proven that resistance genes are Supplementary data to this article can be found online at https:// capable of moving from animals to the environment and from the en- doi.org/10.1016/j.envint.2019.104936. vironment to the clinic (Cabello, 2006; Forsberg et al., 2012), with the latter representing a great threat to human health. Furthermore, References transmission from the clinic to the environment can also occur, causing resistance to arise in environmental microorganisms, which may cause Allen, H.K., Donato, J., Wang, H.H., Cloud-Hansen, K.A., Davies, J., Handelsman, J., them to emerge as pathogens that continue to transfer resistance genes 2010. Call of the wild: antibiotic resistance genes in natural environments. Nat. Rev. Microbiol. 8, 251–259. PMID: 20190823. https://doi.org/10.1038/nrmicro2312. to other microorganisms. The antibiotic/antibiotic resistance phenom- Allen, J.L., Conrad, K., Oberdörster, G., Johnston, C.J., Sleezer, B., Cory-Slechta, D.A., enon and ARGs have been present in the environment since long before 2013. Developmental exposure to concentrated ambient particles and preference for their discovery by humans, because microorganisms produce anti- immediate reward in mice. Environ. Health Perspect. 121, 32–38. PMID: 23063827. https://doi.org/10.1289/ehp.1205505. biotics to defend themselves against competing organisms. It is widely Allen, J.L., Liu, X., Pelkowski, S., Palmer, B., Conrad, K., Oberdörster, G., et al., 2014. known that resistant bacteria can be selected when high concentrations Early postnatal exposure to ultrafine particulate matter air pollution: persistent of antibiotic drugs are used therapeutically, particularly when these are ventriculomegaly, neurochemical disruption, and glial activation preferentially in – misused or abused. Subtherapeutic application of antibiotics can also male mice. Environ. Health Perspect. 122, 939 945. PMID: 24901756. https://doi. org/10.1289/ehp.1307984. promote the development of antibiotic resistance (Lu et al., 2005). Andersen, A.A., 1958. New sampler for the collection, sizing, and enumeration of viable However, it has also been shown that even very low concentrations of airborne particles. J. Bacteriol. 76, 471–484 (PMID: 13598704). antibiotics are sufficient to provide a selective advantage for resistant Brandl, H., Fricker-Feer, C., Ziegler, D., Mandal, J., Stephan, R., Lehner, A., 2014. Distribution and identification of culturable airborne microorganisms in a Swiss milk microorganisms (Gullberg et al., 2011). Furthermore, the presence of processing facility. J. Dairy Sci. 97, 240–246. PMID: 24210492. https://doi.org/10. heavy metals and biocides could co-select antibiotic resistance in the 3168/jds.2013-7028. environment. Resistance genes from this increasing environmental re- Brodie, E.L., DeSantis, T.Z., Parker, J.P.M., Zubietta, I.X., Piceno, Y.M., Andersen, G.L., 2007. Urban aerosols harbor diverse and dynamic bacterial populations. Proc. Natl. servoir can then be transferred to pathogenic bacteria (Allen et al., Acad. Sci. 104, 299–304. https://doi.org/10.1073/pnas.0608255104. 2010). While the results of this study are essential for public health, Cabello, F.C., 2006. Heavy use of prophylactic antibiotics in aquaculture: a growing regretfully, it cannot be confirmed whether this bacterial resistance problem for human and animal health and for the environment. Environ. Microbiol. 8, 1137–1144. PMID: 16817922. https://doi.org/10.1111/j.1462-2920.2006. originated from the clinic or the environment. Therefore, further re- 01054.x. search is required to expand our understanding of MDR in the air Cao, C., Jiang, W., Wang, B., Fang, J., Lang, J., Tian, G., et al., 2014. Inhalable micro- around China and the world. organisms in Beijing's pm2.5 and pm10 pollutants during a severe smog event. Environ. Sci. Technol. 48, 1499–1507. PMID: 24456276. https://doi.org/10.1021/ In conclusion, an investigation of bacterial composition and AMR es4048472. pattern of culturable bacteria in the ambient air in polluted and non- Chapin, A., Rule, A., Gibson, K., Buckley, T., Schwab, K., 2005. Airborne multidrug-re- polluted weathers was conducted in Beijing. Results indicated that (1) sistant bacteria isolated from a concentrated swine feeding operation. Environ. – bacterial concentration was significantly higher in polluted weather, Health Perspect. 113, 137 142. PMID: 15687049. https://doi.org/10.1289/ehp. 7473. compared to non-polluted weather. (2) Polluted weather is correlated China MEE, 2012a. Ambient Air Quality Standards (GB 3095-2012). Ministry of Ecology with changes in the bacterial composition in the air, with a greater and Environment of The People's Republic of China, Beijing, China. abundance of GNB. (3) Approximately one-fourth of bacteria in the China MEE, 2012b. Technical Regulation on Ambient Air Quality Index (on Trial) (hj 633- 2012). Ministry of Ecology and Environment of The People's Republic of China, ambient air were MDR, regardless of the weather conditions, which Beijing, China. may serve as a potential pathway for antibiotic resistance transfer Clark, S., Rylander, R., Larsson, L., 1983. Airborne bacteria, endotoxin and fungi in dust

6 Y. Mao, et al. Environment International 131 (2019) 104936

in poultry and swine confinement buildings. Am. Ind. Hyg. Assoc. J. 44, 537–541. Magiorakos, A.-P., Srinivasan, A., Carey, R.B., Carmeli, Y., Falagas, M.E., Giske, C.G., PMID: 6613856. https://doi.org/10.1080/15298668391405265. et al., 2012. Multidrug-resistant, extensively drug-resistant and pandrug-resistant CLSI (Clinical and Laboratory Standards Institute), 2017. Performance standards for an- bacteria: an international expert proposal for interim standard definitions for ac- timicrobial susceptibility testing. In: CLSI supplement M100, 27th ed. Clinical and quired resistance. Clin. Microbiol. Infect. 18, 268–281. https://doi.org/10.1111/j. Laboratory Standards Institute, Wayne, PA, USA. 1469-0691.2011.03570.x. de Rooij, M.M.T., Hoek, G., Schmitt, H., Janse, I., Swart, A., Maassen, C.B.M., et al., 2019. Maher, B.A., Ahmed, I.A., Karloukovski, V., MacLaren, D.A., Foulds, P.G., Allsop, D., Insights into livestock-related microbial concentrations in air at residential level in a et al., 2016. Magnetite pollution nanoparticles in the human brain. Proc. Natl. Acad. livestock dense area. Environ. Sci. Technol. https://doi.org/10.1021/acs.est. Sci. U. S. A. 113, 10797–10801. PMID: 27601646. https://doi.org/10.1073/pnas. 8b07029. PMID: 31081619. 1605941113. Douwes, J., Thorne, P., Pearce, N., Heederik, D., 2003. Bioaerosol health effects and Pal, C., Bengtsson-Palme, J., Kristiansson, E., Larsson, D.G., 2016. The structure and di- exposure assessment: progress and prospects. Ann. Occup. Hyg. 47, 187–200 (PMID: versity of human, animal and environmental resistomes. Microbiome 4, 54. PMID: 12639832). 27717408. https://doi.org/10.1186/s40168-016-0199-5. Emerson, J.B., Keady, P.B., Clements, N., Morgan, E.E., Awerbuch, J., Miller, S.L., et al., Park, D.U., Yeom, J.K., Lee, W.J., Lee, K.M., 2013. Assessment of the levels of airborne 2017. High temporal variability in airborne bacterial diversity and abundance inside bacteria, gram-negative bacteria, and fungi in hospital lobbies. Int. J. Environ. Res. single-family residences. Indoor Air 27, 576–586. PMID: 27743387. https://doi.org/ Public Health 10, 541–555. PMID: 23435586. https://doi.org/10.3390/ 10.1111/ina.12347. ijerph10020541. Fang, Z.G., Ouyang, Z.Y., Zhao, J.Z., Wang, X.K., Zheng, H., 2006. Community compo- Pope 3rd, C.A., Ezzati, M., Dockery, D.W., 2009. Fine-particulate air pollution and life sition and dynamics of airborne bacteria in Beijing. Acta Microbiol Sin. 46, 618–623 expectancy in the United States. N. Engl. J. Med. 360, 376–386. PMID: 19164188. (PMID: 17037066). https://doi.org/10.1056/NEJMsa0805646. Fang, Z., Ouyang, Z., Zheng, H., Wang, X., Hu, L., 2007. Culturable airborne bacteria in Shelton, B.G., Kirkland, K.H., Flanders, W.D., Morris, G.K., 2002. Profiles of airborne outdoor environments in Beijing, China. Microb. Ecol. 54, 487–496. PMID: fungi in buildings and outdoor environments in the United States. Appl. Environ. 17308950. https://doi.org/10.1007/s00248-007-9216-3. Microbiol. 68, 1743–1753. PMID: 11916692. https://doi.org/10.1128/AEM.68.4. Forsberg, K.J., Reyes, A., Wang, B., Selleck, E.M., Sommer, M.O., Dantas, G., 2012. The 1743-1753.2002. shared antibiotic resistome of soil bacteria and human pathogens. Science 337, Solomon, F.B., Wadilo, F.W., Arota, A.A., Abraham, Y.L., 2017. Antibiotic resistant air- 1107–1111. PMID: 22936781. https://doi.org/10.1126/science.1220761. borne bacteria and their multidrug resistance pattern at university teaching referral Gao, X.L., Shao, M.F., Wang, Q., Wang, L.T., Fang, W.Y., Ouyang, F., et al., 2018. Airborne hospital in south Ethiopia. Ann. Clin. Microbiol. Antimicrob. 16, 29. https://doi.org/ microbial communities in the atmospheric environment of urban hospitals in China. 10.1186/s12941-017-0204-2. J. Hazard. Mater. 349, 10–17. PMID: 29414740. https://doi.org/10.1016/j.jhazmat. Teixeira, J.V., Cecílio, P., Gonçalves, D., Vilar, V.J., Pinto, E., Ferreira, H.N., 2016. 2018.01.043. Multidrug-resistant enterobacteriaceae from indoor air of an urban wastewater Ghosh, B., Lal, H., Srivastava, A., 2015. Review of bioaerosols in indoor environment with treatment plant. Environ. Monit. Assess. 188, 388. PMID: 27260528. https://doi. special reference to sampling, analysis and control mechanisms. Environ. Int. 85, org/10.1007/s10661-016-5382-4. 254–272. PMID: 26436919. https://doi.org/10.1016/j.envint.2015.09.018. Uhrbrand, K., Schultz, A.C., Koivisto, A.J., Nielsen, U., Madsen, A.M., 2017. Assessment of Górny, R.L., Dutkiewicz, J., 2002. Bacterial and fungal aerosols in indoor environment in airborne bacteria and noroviruses in air emission from a new highly-advanced hos- central and eastern European countries. Ann. Agric. Environ. Med. 9, 17–23 (PMID: pital wastewater treatment plant. Water Res. 112, 110–119. PMID: 28153697. 12088392). https://doi.org/10.1016/j.watres.2017.01.046. Górny, R., Dutkiewicz, J., Krysińska-Traczyk, E., 1999. Size distribution of bacterial and Ventura, M., Canchaya, C., Tauch, A., Chandra, G., Fitzgerald, G.F., Chater, K.F., et al., fungal bioaerosols in indoor air. Ann. Agric. Environ. Med. 6, 105–113 (PMID: 2007. Genomics of Actinobacteria: tracing the evolutionary history of an ancient 10607991). phylum. Microbiol Mol Biol Rev 71, 495–548. PMID: 17804669. https://doi.org/10. Gullberg, E., Cao, S., Berg, O.G., Ilbäck, C., Sandegren, L., Hughes, D., et al., 2011. 1128/MMBR.00005-07. Selection of resistant bacteria at very low antibiotic concentrations. PLoS Pathog. 7, Weuve, J., Puett, R.C., Schwartz, J., Yanosky, J.D., Laden, F., Grodstein, F., 2012. e1002158 PMID:. 21811410. https://doi.org/10.1371/journal.ppat.1002158. Exposure to particulate air pollution and cognitive decline in older women. Arch. He, C., Mackay, I.M., Ramsay, K., Liang, Z., Kidd, T., Knibbs, L.D., et al., 2017. Particle Intern. Med. 172, 219–227. PMID: 22332151. https://doi.org/10.1001/ and bioaerosol characteristics in a paediatric intensive care unit. Environ. Int. 107, archinternmed.2011.683. 89–99. PMID: 28692913. https://doi.org/10.1016/j.envint.2017.06.020. Wu, S., Ni, Y., Li, H., Pan, L., Yang, D., Baccarelli, A.A., et al., 2016. Short-term exposure Hiltunen, T., Virta, M., Laine, A.L., 2017. Antibiotic resistance in the wild: an eco-evo- to high ambient air pollution increases airway inflammation and respiratory symp- lutionary perspective. Philos. Trans. R. Soc. Lond. Ser. B Biol. Sci. 372https://doi. toms in chronic obstructive pulmonary disease patients in Beijing, China. Environ. org/10.1098/rstb.2016.0039. PMID: 27920384. Int. 94, 76–82. PMID: 27209003. https://doi.org/10.1016/j.envint.2016.05.004. Hunt, A., Abraham, J.L., Judson, B., Berry, C.L., 2003. Toxicologic and epidemiologic Xie, J., Jin, L., He, T., Chen, B., Luo, X., Feng, B., et al., 2019. Bacteria and antibiotic

clues from the characterization of the 1952 London smog fine particulate matter in resistance genes (ARGs) in PM2.5 from China: implications for human exposure. archival autopsy lung tissues. Environ. Health Perspect. 111, 1209–1214. PMID: Environ. Sci. Technol. 53, 963–972. PMID: 30525504. https://doi.org/10.1021/acs. 12842775. https://doi.org/10.1289/ehp.6114. est.8b04630. Kim, K.H., Kabir, E., Jahan, S.A., 2018. Airborne bioaerosols and their impact on human Yoo, K., Lee, T.K., Choi, E.J., Yang, J., Shukla, S.K., Hwang, S.I., et al., 2017. Molecular health. J. Environ. Sci. 67, 23–35. PMID: 29778157. https://doi.org/10.1016/j.jes. approaches for the detection and monitoring of microbial communities in bioaer- 2017.08.027. osols: a review. J. Environ. Sci. 51, 234–247. PMID: 28115135. https://doi.org/10. Li, J., Cao, J., Zhu, Y.G., Chen, Q.L., Shen, F., Wu, Y., et al., 2018. Global survey of 1016/j.jes.2016.07.002. antibiotic resistance genes in air. Environ. Sci. Technol. 52, 10975–10984. PMID: Zhai, Y., Li, X., Wang, T., Wang, B., Li, C., Zeng, G., 2018. A review on airborne micro- 30043612. https://doi.org/10.1021/acs.est.8b02204. organisms in particulate matters: composition, characteristics and influence factors. Liang, R., Xiao, P., She, R., Han, S., Chang, L., Zheng, L., 2013. Culturable airborne Environ. Int. 113, 74–90. PMID: 29421410. https://doi.org/10.1016/j.envint.2018. bacteria in outdoor poultry-slaughtering facility. Microbes Environ. 28, 251–256. 01.007. PMID: 23474646. https://doi.org/10.1264/jsme2.ME12178. Zhang, Q.Q., Ying, G.G., Pan, C.G., Liu, Y.S., Zhao, J.L., 2015. Comprehensive evaluation Liu, H., Zhang, X., Zhang, H., Yao, X., Zhou, M., Wang, J., et al., 2018. Effect of air of antibiotics emission and fate in the river basins of China: source analysis, multi- pollution on the total bacteria and pathogenic bacteria in different sizes of particulate media modeling, and linkage to bacterial resistance. Environ. Sci. Technol. 49, matter. Environ. Pollut. 233, 483–493. PMID: 29101891. https://doi.org/10.1016/j. 6772–6782. PMID: 25961663. https://doi.org/10.1021/acs.est.5b00729. envpol.2017.10.070. Zhu, B., Chen, Q., Chen, S., Zhu, Y.G., 2017. Does organically produced lettuce harbor Lu, L., Hume, M.E., Pillai, S.D., 2005. Autoinducer 2-like activity in poultry-associated higher abundance of antibiotic resistance genes than conventionally produced? enteric bacteria in response to subtherapeutic antibiotic exposure. Avian Dis. 49 (1), Environ. Int. 98, 152–159. PMID: 27823798. https://doi.org/10.1016/j.envint.2016. 74–80. PMID: 15839416. https://doi.org/10.1637/7212-05200R. 11.001.