Characterization of primary biogenic aerosol particles in urban, rural, and high-alpine air by DNA sequence and restriction fragment analysis of ribosomal RNA genes

International audienceThis study explores the applicability of DNA analyses for the characterization of primary biogenic aerosol (PBA) particles in the atmosphere. Samples of fine particulate matter (PM2.5) and total suspended particulates (TSP) have been collected on different types of filter materials at urban, rural, and high-alpine locations along an altitude transect in the south of Germany (Munich, Hohenpeissenberg, Mt. Zugspitze). From filter segments loaded with about one milligram of air particulate matter, DNA could be extracted and DNA sequences could be determined for bacteria, fungi, plants and animals. Sequence analyses were used to determine the identity of biological organisms, and terminal restriction fragment length polymorphism analyses (T-RFLP) were applied to estimate diversities and relative abundances of bacteria. Investigations of blank and background samples showed that filter materials have to be decontaminated prior to use, and that the sampling and handling procedures have to be carefully controlled to avoid artifacts in the analyses. Mass fractions of DNA in PM2.5 were found to be around 0.05% in urban, rural, and high-alpine aerosols. The average concentration of DNA determined for urban air was on the order of ~7 ng m?3, indicating that human adults may inhale about one microgram of DNA per day (corresponding to ~108 haploid bacterial genomes or ~105 haploid human genomes, respectively). Most of the bacterial sequences found in PM2.5 were from Proteobacteria (42) and some from Actinobacteria (10) and Firmicutes (1). The fungal sequences were characteristic for Ascomycota (3) and Basidiomycota (1), which are known to actively discharge spores into the atmosphere. The plant sequences could be attributed to green plants (2) and moss spores (2), while animal DNA was found only for one unicellular eukaryote (protist). Over 80% of the 53 bacterial sequences could be matched to one of the 19 T-RF peaks found in the PM2.5 samples, but only 40% of the T-RF peaks did correspond to one of the detected bacterial sequences. The results demonstrate that the T-RFLP analysis covered more of the bacterial diversity than the sequence analysis. Shannon-Weaver indices calculated from both sequence and T-RFLP data indicate that the bacterial diversity in the rural samples was higher than in the urban and alpine samples. Two of the bacterial sequences (Gammaproteobacteria) and five of the T-RF peaks were found at all sampling locations