Characterizing the small RNA transcriptome associated with COPD and ILD using next-generation sequencing

Abstract

Introduction: Despite the increasing public health burden associated with COPD and ILD, the molecular mechanisms associated with disease pathogenesis remain unclear and a limited number of studies have been performed examining role of small noncoding RNA in regulating these disease processes. The goal of this study is to comprehensively profile the lung small RNA transcriptome via next generation sequencing and elucidate the roles of microRNA in COPD and ILD. Methods: As part of the Lung Genomics Research Consortium, we deep-sequenced the small RNA in lung tissue samples from patients with COPD (n=150) or ILD (n=149) and from normal lung tissue (n=65). 319 lung tissue samples were sequencing via multiplexing on the Illumina HiSeq 2000 (10 samples/lane) and 45 samples were sequenced on the Illumina GAIIx (1 sample/lane). Reads were trimmed using the FASTX toolkit and aligned to hg19 using Bowtie. Negative binomial generalized linear models were used to identify microRNAs differentially expressed between phenotypes. Results: An average of 26.3 and 7.1 million reads were sequenced per sample using the singleplex and multiplex protocols, respectively. An average of 73% of reads had at least one alignment to the human genome with one or fewer mismatches at 10 or fewer locations. Across all samples, we found that 2.4 billion reads mapped to microRNA (84% of aligned reads), 85.5 million reads mapped to tRNA (3%), 1.7 million reads mapped to snoRNA (<1%), 19.0 million reads mapped to rRNA (<1%), 49.2 million reads mapped to gene exons (2%), and 162.7 million reads mapped to gene introns or UTRs (6%). 287 novel microRNA precursors were predicted using the miRDeep algorithm. 797 microRNAs were detected with a median absolute deviation greater than one. Of these, 194 were differentially expressed between ILD and controls while 29 varied between COPD and controls (FDR < 0.05). When comparing ILD samples to controls, 62 microRNAs displayed 1 - 2 bp differences in the start position of their 5' ends (FDR < 0.05). These differences in start positions potentially produce microRNA with alternate seeds which may result in the targeting of different sets of mRNA transcripts. These potential disease-specific alterations in microRNA function could contribute to disease pathogenesis. Conclusions: Our results demonstrate the power of deep sequencing to reveal disease-associated patterns of small RNA expression that may provide insights into the molecular pathogenesis of chronic lung diseases, novel biomarkers for disease activity, and novel targets for therapy.