大气PM2.5污染是北京市典型环境问题之一，PM2.5中微生物和抗生素抗性基因(ARGs)等生物组分对人体健康产生的负面影响日渐引起公众关注。目前关于PM2.5中微生物、ARGs受季节和典型天气共同影响的研究较少，而大气ARGs可能的宿主微生物和微生物的潜在来源尚未明确。本研究在北京城区典型大气监测站开展了为期一年的大气PM2.5采样，采集了四季常规天气和雾霾、降雨、降雪、沙尘暴等典型天气的样品，结合分子生物学和多元统计学方法对PM2.5的化学组分、微生物、ARGs的变化及其响应互作关系开展了研究，并结合HYSPLIT模型对PM2.5中微生物的潜在来源进行分析。

首先对大气PM2.5样品的无机离子、金属元素等化学组分进行检测，结合采样期间的气象因素、大气污染物等环境因子信息进行分析。常规天气下，PM2.5中NO3-/SO42-的值大于1，说明北京城区大气污染主要受汽车尾气排放影响；但在降雨、降雪、雾霾、沙尘暴期间，该值小于1，此时大气污染主要受煤炭燃烧影响。相关性分析、因子分析表明北京城区PM2.5中无机离子、金属元素最主要的来源分别是二次源和汽车排放源。

基于高通量测序分析了北京城区大气微生物群落的结构和功能，并采用实时荧光定量PCR检测了PM2.5中4大类14种ARGs的相对丰度，对不同天气的微生物群落和ARGs进行对比分析，发现：PM2.5污染水平和微生物群落丰富度、多样性都在春、冬两季偏高；在PM2.5中发现了43个菌门，优势类群为：Proteobacteria、Firmicutes、Actinobacteriota、Cyanobacteria和Bacteroidota；发现了1344个菌属，四季优势菌属和特征菌属各不相同，春季为Chloroplast和Mitochondria，夏季为Acidibacter和Paracoccus，秋季为Ammoniphilus和Bacteria，冬季为Sphingomonas和Pseudomonas；不同典型天气也有各自的特征菌属，雾霾天气特征菌属为Diaphorobacter，降雨天气为Delftia，降雪天气为Terrisporobacter，沙尘暴天气为Cytophagaceae。非度量多维尺度分析结果表明大气微生物群落在夏季较稳定，其他季节降雨、降雪、雾霾、沙尘暴等典型天气会对微生物群落产生较大的扰动。采用典型关联分析研究环境因子对群落的影响，发现温度、相对湿度、PM2.5和二次离子是关键环境影响因子。基于BugBase的功能预测发现，大气微生物主要以好氧菌、革兰氏阴性菌为主，Acidibacter是四季最丰富的潜在致病菌，夏季大气微生物的潜在致病风险较大。

整体上看，大气ARGs污染在冬季最严重，在夏季转移风险最大，PM2.5中磺胺类抗性基因sul3和β-内酰胺类抗性基因blaTEM的转移风险最大。大环内酯类抗性基因ermB和四环素类抗性基因tetW与总ARGs相关性较高，是北京城区大气ARGs的标志基因。采用相关性分析、Network网络分析研究大气ARGs可能的宿主微生物，其中四环素类抗性基因tet32的宿主菌属为Corynebacterium和Streptococcus; β-内酰胺类抗性基因blaTEM的宿主菌属为Acidovorax和Sphingomonas; 氨基糖苷类抗性基因aph(3‘)-Ⅲa的宿主菌属为Streptococcus; 万古霉素类抗性基因VanB的宿主菌属为Methylobacterium-Methylorubrum和Acinetobacter，大多数宿主微生物为致病菌。

使用HYSPLIT模型对大气PM2.5进行后向轨迹分析，结合指示微生物分析发现：夏季的海洋气团、其他季节的内陆气团运输以及周边环境的植被、土壤对大气微生物的来源影响较大。气团运输对特殊天气的大气微生物来源影响较大，沙尘暴期间大气微生物主要受蒙古国的长距离气团跨境运输影响，沙尘暴特征菌属Cytophagaceae常见于内蒙古土壤环境，和轨迹分析结果吻合。

PM2.5 pollution is one of the typical environmental problems in Beijing. The negative impact of biological components such as microorganisms and antibiotic resistance genes (ARGs) in PM2.5 on human health has increasingly attracted public attention. At present, there are few studies on the combined effects of seasons and weather conditions on microorganisms and ARGs in PM2.5, and the possible host microorganisms of ARGs and potential sources of microorganisms are not yet clear. In this study, a one-year PM2.5 sampling was carried out at a typical atmospheric monitoring station in the urban area of Beijing, and samples were collected from regular weather in four seasons, haze, rainfall, snowfall, sandstorm, combined with molecular biology methods and multivariate statistical methods. The changes of chemical components, microorganisms, and ARGs of PM2.5 and their response interactions were studied, and the potential sources of microorganisms in PM2.5 were analyzed combined with the HYSPLIT model.

Firstly, the chemical composition of PM2.5 was analyzed, and environmental factors such as meteorological factors and atmospheric pollutants during the sampling period were obtained. Under normal weather, the value of NO3-/SO42- in PM2.5 is greater than 1, indicating that the air pollution in Beijing urban area is mainly due to vehicle exhaust emissions; but during rainfall, snowfall, smog, and sandstorms, the value is less than 1. At this time, the air pollution mainly comes from coal burning. Correlation analysis and factor analysis showed that the main sources of inorganic ions and metal elements in PM2.5 in Beijing urban areas were secondary sources and automobile emission sources, respectively.

Based on high-throughput sequencing, the structure and function of the airborne microbial community in Beijing urban area were analyzed, and the relative abundance of 14 ARGs in 4 categories in PM2.5 was detected by real-time quantitative PCR, and the microbial community and ARGs in different weather were compared. Analysis showed that: PM2.5 pollution level and microbial richness and diversity were higher in spring and winter; 43 phyla were found in PM2.5, and the dominant groups were: Proteobacteria, Firmicutes, Actinobacteriota, Cyanobacteria and Bacteroidota ; 1344 genera were found, with different dominant and characteristic genera in different seasons, Chloroplast and Mitochondria in spring, Acidibacter and Paracoccus in summer, Ammoniphilus and Bacteria in autumn, Sphingomonas and Pseudomonas in winter; Diaphorobacter in haze, Delftia in rainfall, Terrisporobacter in snowfall, and Cytophagaceae in sandstorm. The results of non-metric multi-dimensional scaling analysis showed that the airborne microbial community was relatively stable in summer, and the weather in other seasons such as rainfall, snowfall, haze, and sandstorm would have a greater disturbance to the microbial community. The effects of environmental factors on the community were studied by canonical correlation analysis, and it was found that temperature, relative humidity, PM2.5 and secondary ions were the key environmental factors. Based on the functional prediction of BugBase, it was found that the airborne microorganisms were mainly aerobic bacteria and gram-negative bacteria, and Acidibacter was the most abundant potential pathogenic bacteria in the four seasons, and the potential pathogenic risk of airborne microorganisms was greater in summer.

On the whole, airborne ARGs pollution was the most serious in winter, and the risk of transfer was the greatest in summer. The sul3 and the blaTEM had the greatest transfer risk in PM2.5, while the macrolide resistance gene ermB and the tetracycline resistance gene tetW were highly correlated with total ARGs. They were the marker genes of airborne ARGs in Beijing urban area. Correlation analysis and Network analysis were used to study the possible host microorganisms of airborne ARGs. The tet32 was hosted by Corynebacterium and Streptococcus; blaTEM was hosted by Acidovorax and Sphiningomonas; The host of aph(3')-Ⅲa was Streptococcus; the hosts of VanB were Methylobacterium-Methylorubrum and Acinetobacter, and most of the host microorganisms were pathogenic bacteria.

Using the HYSPLIT model to analyze the backward trajectory of PM2.5, combined with the analysis of indicator microorganisms, it was found that the marine air mass in summer, the transport of inland air mass in other seasons, and the vegetation and soil of the surrounding environment had a greater impact on the source of airborne microorganisms. Air mass transport has a greater impact on the source of airborne microorganisms in special weather. During sandstorms, airborne microorganisms are mainly affected by long-distance air mass transboundary transport from Mongolia. Cytophagaceae is characteristic genus of sandstorms, which commonly found in soil environments in Inner Mongolia, which is consistent with the trajectory analysis results.