Metagenomic Study Of The Pathogens In Unexplained Pneumonia Cases

Many viruses and bacteria could lead unexplained pneumonia (UP). UP was also observed in many emerging diseases. An accurate identification and detailed characterization of pathogens caused UP is critical for targeted therapy in clinic and epidemics surveillance in public health. However, we still knew few in the etiology of UP, as well as the roles of viruses and bacteria played in the pathogenesis of UP.Metagenomic assay could provide an unbiased investigation of all the pathogens in clinical samples. Compared to common PCR-based tools, the metagenomic method need no priori sequence information of the pathogen, and could perform a comprehensive identification of all the pathogens in a clinical sample. Meanwhile, detailed genomic information of the pathogens could be achieved using metagenomic assay, which made metagenomic method preferable in pathogen study of the UP cases. Moreover, the newly developed next generation sequencing (NGS) technology has greatly facilitated the metagenomic studies, which substantially promoted its application in clinic. However, several challenges, such as biased amplification of sample nucleic acid, high proportion of host sequences in the resulting data, and bioinformatic analysis obstacle, must been addressed before its popular application in clinic.Our study pretreated the clinical samples with low-speed centrifugation, filtration, and nuclease digestion to deplete host sequences in the NGS output and achieved a good performance. We also evaluated two isothermal amplification methods, the multiple displacement amplification (MDA) and single primer isothermal amplification (SPIA), for their ability to detect respiratory viruses in clinical samples. Our results showed the SPIA performed better on pathogen detection and uniform RNA amplification, which was chosen as the nucleic acid amplification method in the following NGS sequencing.We applied this NGS protocol to pathogen identification of 33 oropharyngeal swabs collected from unexplained pneumonia cases in Guizhou province in 2013-2014. After sample pretreatment and SPIA amplification, these samples were subjected to RNA-Seq using the Ion torrent PGM or Illumina MiSeq NGS platform, in which 13 viruses were detected in 16 samples by metagenomic analysis, and human rhinovirus C was the most detected pathogen. Metagenomic method could also obtain sequence information of the detected pathogen, which could be used for further genotype characterization.DNA-Seq was applied to 17 samples with no respiratory virus detected in the RNA-Seq, but still detected no respiratory virus in these samples. However, microbiome distribution analysis indicated a high proportion of Streptococcus in the clinical samples of UP patients. While in the genus Streptococcus, some opportunistic bacteria, such as Streptococcus pneumonia, possessed a high proportion.Regular antibiotic treatments were applied to all the UP cases but no improvement was observed, which led us conduct an antibiotic-resistance analysis of the microbes in the UP cases. The DNA-Seq result was aligned to a comprehensive antibiotic-resistance database to observe the antibiotic-resistance profile of UP microbes in clinical samples. In most of the UP cases, we could detect a large amount of antibiotic-resistance genes, most of which targeted against beta-lactam, tetracycline, and macrolide. These antibiotic-resistance genes existed in microbes belonging to diverse orders in taxonomy, among which Bacteroidales, Lactobacillales, Pseudomonadales were the main taxa. In some samples, high proportion of Acinetobacter baumannii and streptococcus was correlated with the high abduance of antibiotic-resistance genes they possessed, indicating a possible correlation in the abduance between microbes and AR genes itself contianed.In conclusion, we established an unbiased pathogen identification method based on the newly developed NGS technologies. With this method, we conducted DNA- and RNA-Seq of clinical samples from UP patients in Guizhou province in 2013-2014. We found common respiratory viruses, especially human rhinovirus C, were the most main viruses detected in the clinical samples of UP cases. Furthermore, substantial opportunistic pathogenic bacteria existed in in the samples of UP patients. Meanwhile, we found a large amount of antibiotic-resistance genes in the microbes of UP cases. The larger amout of antibiotic-resistance genes in streptococcus and Acinetobacter may contribute their predominant abundance in the UP clinical samples. Our study provided preliminary information of virome, microbiome, and antibiotic-resistance profile in the clinical samples of UP cases, which set a reference in methodology and etiology for the diagnosis of UP cases in clinic.