Special Section | Microbial Genomics | Methodology Open Access Determining RNA quality for NextGen sequencing: some exceptions to the gold standard rule of 23S to 16S rRNA ratio § Arvind A Bhagwat 1* , Z. Irene Ying 1 , Jef Karns 1 and Allen Smith 2 *Correspondence: Arvind.bhagwat@ars.usda.gov § Mention of brand or irm name does not constitute an endorsement by the U.S. Department of 13 Agriculture above others of a similar nature not mentioned. 1 Environmental Microbial & Food Safety Laboratory and 2 Diet Genomics and Immunology Laboratory, Henry A. Wallace Beltsville Agricultural Research Center, Agricultural Research Service, USDA, 10300 Baltimore Avenue, BARC-E, Beltsville, MD 20705-2350, USA. Abstract Background: Using next-generation-sequencing technology to assess entire transcriptomes requires high quality starting RNA. Traditionally, the ratios of 23S to 16S ribosomal RNA bands from agarose gel electrophoresis have been used to judge integrity of RNA. Currently, RNA quality is routinely judged using automated microluidic gel electrophoresis platforms and associated algorithms. Findings: Here we report that the two most popular automated platforms (i.e., BioAnalyzer and Experion systems of Agilent Technology and Bio-Rad Laboratories respectively) and their associate algorithms are based on limited data sets of model organisms. he systems perform data interpretation with presumption that prokaryotic rRNA molecules are eluted in two unique peaks corresponding to 23S and 16S molecules. However, certain microorganisms carry intervening sequences in their rRNA structural genes that are subsequently excised during ribosome formation. In such instances, the 23S and 16S rRNA components are eluted in multiple peaks. As a result, current algorithms used by microluidic platforms read such samples as ‘degraded’ and assign them poor RNA quality scores. We observed RNA isolated from several Citrobacter and Salmonella isolates generated false quality scores and low 23S to 16S ribosomal RNA ratios. Conclusions: For RNA-sequencing projects involving non-model organisms, relying solely on automated algorithms for ‘quality control’ of RNA could be misleading. Multiple peaks corresponding to 23S or 16S RNA could be due to occurrence of multiple intervening sequences in rRNA genes. Keywords: RNA-Seq, RNA quality, salmonella transcriptome, agarose gel electrophoresis © 2013 Bhagwat et al; licensee Herbert Publications Ltd. his is an Open Access article distributed under the terms of Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0). his permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background The decreasing cost and increasing availability of next- generation sequencing (NGS) technology will lead the way to its wide ranging applications across the fields of microbiology [9]. One powerful application of NGS technology is sequencing an organism’s entire mRNA population, allowing for high- resolution gene expression studies. Obtaining high quality RNA preparations is extremely important in such experiments. Researchers have relied on guidelines such as a high ratio of A 260 /A 280 (>2.0) to measure RNA purity, and a high ratio of 23S rRNA to 16S rRNA (~1.5) to measure RNA integrity [14]. More recently, gel electrophoresis profiles from automated microfluidic platforms are used to determine RNA integrity [15,17]. However, interpretation of RNA quality by automated methods is limited by the associated algorithms, which are constructed from a limited number of biological samples and biased for eukaryotic RNA analyses. A number of prokaryotes do not possess intact 23S rRNA molecules [7], which is not accounted for by platforms using computer algorithms. Here we report several Salmonella sp. isolates that do not carry intact 23S rRNA, and as a result, their RNA preparations may be erroneously judged as degraded or having low RNA integrity. Methods Bacterial strains and growth conditions Salmonella enterica serovar strains paratyphi B, paratyphi A, Montevideo, and Panama were from Salmonella Genetic Stock Center (SARB Collection, University of Calgary, Canada). Salmonella enterica Serovar Typhimurium strain SL1344, Escherichia coli O157:H7 EDL933, Escherichia albertii USDA 181, and Citrobacter rodentium ATCC 51459 have been described earlier [2,3,8,11,16]. Bacterial cultures were streaked on Luria-Bertani (LB) agar plates from freezer stocks, and a single colony was inoculated in LB broth and grown at 37 o C in a shaker incubator for 18-20 h. RNA isolation procedure RNA was isolated from stationary phase cells grown in LB broth for 18-20 h. Cells were stabilized with RNA stabilization reagent (RSR-3) [5] and RNA was isolated using Qiagen RNeasy kit as described before [1,5]. RNA was also isolated by suspending cells (after RNA stabilization) in hot Trizol reagent (65 o C) and purified using RNeasy silica columns and on-column DNase treatment (Qiagen). After elution from Qiagen column, RNA was treated with DNase (New England Biolabs, USA) and repurified with a RNeasy column. DNA contamination was checked by Microbiology Discovery ISSN 2052-6180