Native Purification and Analysis of Long RNAs

Isabel Chillón*,†; Marco Marcia*,1; Michal Legiewicz*,2; Fei Liu*; Srinivas Somarowthu*; Anna Marie Pyle*,†,‡,3    * Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, Connecticut, USA
† Howard Hughes Medical Institute, Chevy Chase, Maryland, USA
‡ Department of Chemistry, Yale University, New Haven, Connecticut, USA
3 Corresponding author: email address: anna.pyle@yale.edu
1 Present address: European Molecular Biology Laboratory, Grenoble Outstation, 71 Avenue des Martyrs, 38042 Grenoble Cedex 9, France
2 Present adress: School of Life Sciences, University of Warwick, Coventry, United Kingdom

1 Introduction

Biochemical and biophysical studies on RNA molecules in vitro require that target molecules are synthesized and purified as homogeneous, functional species. For catalytic RNAs, enzymatic assays can confirm that the target molecules are correctly folded (Russell et al., 2002; Wan, Mitchell, & Russell, 2009). For noncatalytic RNAs, folding homogeneity needs to be assessed by nonenzymatic, biochemical, or biophysical assays (Woodson, 2011).

Traditionally, two folding procedures have been successfully used in biochemical and biophysical studies of large RNAs, such as group I and II introns, other ribozymes, riboswitches, signal recognition particle RNA, and tRNAs. The first approach involves heat denaturation and refolding of the RNA after in vitro transcription (Fedorova, Waldsich, & Pyle, 2007; Lomant & Fresco, 1975; Takamoto, He, Morris, Chance, & Brenowitz, 2002; Walstrum & Uhlenbeck, 1990; Zhang & Ferre-D'Amare, 2013), whereas the second approach consists of cotranscriptional folding without any denaturation step (Batey, 2014; Toor, Keating, Taylor, & Pyle, 2008). The latter method preserves the secondary and/or tertiary structure adopted by the RNA during transcription, which can be important considering that functional RNA structures are formed cotranscriptionally in vivo (Frieda & Block, 2012; Heilman-Miller & Woodson, 2003; Lai, Proctor, & Meyer, 2013).

Long noncoding RNAs (lncRNAs) are involved in a staggering diversity of fundamental cellular functions and they represent important subjects for ongoing research (Gutschner & Diederichs, 2012). Most lncRNAs do not have ribozyme activity, but they are essential in development, transcription, and epigenetic processes (Necsulea et al., 2014). These RNAs, which can reach tens of kilobases in length, appear to have encountered weaker evolutionary selection constraints than protein-coding genes, thus accumulating repetitive sequences (Derrien et al., 2012). LncRNAs are therefore challenging molecules for biophysical analysis because they can form alternative conformations in the absence of any structural constraint (Huthoff & Berkhout, 2002). Given these properties, cotranscriptional native purification may be particularly useful for lncRNA targets.

Here, we describe a native purification protocol that results in pure, homogeneous lncRNA preparations amenable for biochemical and biophysical studies. While other purification methods make use of affinity tags suitable to extract RNA from in vivo sources (Batey & Kieft, 2007; Said et al., 2009), our protocol does not necessarily require tags. The use of tags, generally added to the 3′ end of target RNAs, ensures capturing a homogeneous population of full-length molecules. In our method, we achieve a similar goal by using a T7 polymerase construct that rarely produces short abortive transcripts (Tang et al., 2014), and centrifugal filtration and size-exclusion chromatography (SEC) as final polishing steps in purification. Not using tags may simplify cloning design and avoid inclusion of nonnative sequences that may interfere with structure formation of the target RNA. However, if tags are useful for downstream applications, their inclusion is compatible with our protocol.

We additionally describe methods to study lncRNA folding based on sedimentation velocity analytical ultracentrifugation (SV-AUC), analytical SEC, and chemical probing. These analytical techniques are provided as examples, as many other techniques (such circular dichroism, small angle X-ray scattering, etc.) can be used to monitor homogeneity and oligomeric state of long RNAs (Behrouzi, Roh, Kilburn, Briber, & Woodson, 2012; Pan & Sosnick, 1997; Rambo & Tainer, 2010). SV-AUC and analytical SEC allow one to monitor global compaction of RNA preparations in the presence of divalent ions, under equilibrium conditions (Cole, Lary, Moody, & Laue, 2008; Mitra, 2014), and the equipment required is commonly available. Chemical probing facilitates the determination of lncRNA secondary structure (Athavale et al., 2013; Novikova, Hennelly, & Sanbonmatsu, 2012), and it also utilizes reagents that are available to most investigators. In this review, we describe protocols for selective 2′-hydroxyl acylation analyzed by primer extension (SHAPE) and dimethyl sulfate (DMS) chemical probing, as they have been applied for studying lncRNAs (Novikova et al., 2012; Watts et al., 2009). Again, these represent only a subset of available techniques for mapping RNA structure in solution. The SHAPE and DMS methods employ an electrophilic reagent that reacts selectively with flexible, accessible sites on ribonucleotides, facilitating detection of loops and other single-stranded regions within lncRNA molecules.

2 Native Purification of Long Noncoding RNAs

In this section, we outline the procedures for transcription and purification of lncRNAs following nondenaturing methods. These protocols are based on those developed in our laboratory in recent years (Fedorova, Su, & Pyle, 2002; Fedorova et al., 2007; Marcia & Pyle, 2012; Toor et al., 2008), which have been modified and updated to fit the idiosyncrasies of large noncoding RNAs that range from one to several kilobases in length (up to 4 kb in our hands). The pipeline of the procedure includes the design of the construct, linearization of the DNA template and transcription of the RNA, digestion of DNA, enzyme proteolysis, buffer exchange, and finally, filtration and purification using SEC (Fig. 1A).

2.1 Construct design

The target lncRNA should be inserted into a suitable vector (e.g., pBluescript, Life Technologies) such that it is immediately flanked at the 5′ end by T7 promoter sequence and at the 3′ end by a unique and efficient restriction site (e.g., BamHI). In addition, we recommend digesting the template DNA with a restriction enzyme that produces blunt or...