nprot.2006.480

5¢ end cDNA amplification using classic RACE Elizabeth Scotto–Lavino1,2, Guangwei Du2 & Michael A Frohman1–3 1Graduate Program in Molecular & Cellular Pharmacology; 2Department of Pharmacological Sciences & Center for Developmental Genetics, Stony Brook University, Stony Brook, New York 11794, U.S.A. 3Correspondence should be addressed to M.A.F. (michael@pharm.stonybrook.edu). Published online 29 December 2006; doi:10.1038/nprot.2006.480 The 5¢ ends of transcripts provide important information about transcription initiation sites and the approximate locations of local cis-acting enhancer elements; it is therefore important to establish the 5¢ ends with some precision. RACE (rapid amplification of cDNA ends) PCR is useful for quickly obtaining full length cDNAs for mRNAs for which only part of the sequence is known and to identify alternative 5¢ or 3¢ ends of fully sequenced genes. The method consists of using PCR to amplify, from complex mixtures of cellular mRNA, the regions between the known parts of the sequence and non-specific tags appended to the ends of the cDNA. Whereas the poly(A) tail serves to provide such a tag at the 3¢ end of the mRNA, an artificial one needs to be generated at the 5¢ end, and various approaches have been described to address this step. The classical scheme for 5¢ RACE described here is simple, suffices in many instances in which RACE is needed and can be performed in 1–3 days. INTRODUCTION Most attempts to identify and isolate a novel cDNA result in the acquisition of clones that represent only a part of the mRNA’s complete sequence. The ever-growing collections of sequenced genomes and high-quality cDNA libraries can often facilitate acquisition of the remainder of the transcript. For less well- characterized organisms, or for low-abundance cDNAs even in well-characterized organisms, such information is often not avail- able, particularly at the 5¢ end of the transcript. Obtaining a full- length cDNA at the 5¢ ensures that the entire protein region has been identified and yields information concerning the transcription initiation site. In some instances, 5¢ untranslated regions encode structural information that is relevant to mRNA stability, restricted subcellular localization or translational efficiency. The missing sequence (cDNA ends) can be cloned by PCR, using a technique variously called rapid amplification of cDNA ends (RACE)1, anchored PCR2 or one-sided PCR3 (Fig. 1). Since the initial reports describing this technique, many labs and com- panies have developed significant improvements on the basic approach4–15. A protocol for 5¢ end cDNA amplification by classic RACE is presented here. 3¢ end cDNA amplification can also be performed using a classic RACE protocol as described in a separate protocol16. A more complex but also more powerful approach (new RACE), which has evolved from the work of a number of laboratories17–24 is also described in a separate proto- col25. Commercial RACE kits and libraries are available from many companies that are more convenient but often not as powerful as the versions described here. Classic RACE PCR is used to amplify partial cDNAs that represent the region between a single point in an mRNA transcript and its 3¢ or 5¢ end (Figs. 1, 2). A short internal stretch of sequence must already be known from the mRNA of interest. From this sequence, gene- specific primers (GSPs) are chosen that are oriented in the direction of the missing sequence. Extension of the partial cDNAs from the unknown end of the message back to the known region is achieved using primers that anneal to the pre-existing poly(A) tail (3¢ end) or an appended homopolymer tail or linker (5¢ end). Using RACE, enrichments in the order of 106–107-fold can be obtained. As a result, relatively pure cDNA ‘ends’ are generated that can be easily cloned or rapidly characterized using conventional techniques1. To generate ‘3¢ end’ partial cDNA clones, mRNA is reverse transcribed using a ‘hybrid’ primer (Qtotal; QT) that consists of two mixed bases (GATC or GAC followed by (T)17 followed by a unique 35-base oligonucleotide sequence (QI–QO; Fig. 2a,c). Amplification is then performed using a primer containing part of this sequence (Qouter, QO), which now binds to each cDNA at its 3¢ end, and using a primer derived from the gene of interest (GSP1). A second set of amplification cycles is then carried out using ‘nested’ primers (Qinner (QI) and GSP2) to quench the amplifica- tion of non-specific products. To generate ‘5¢ end’ partial cDNA clones, reverse transcription (primer extension) is carried out using a gene-specific primer (GSP-RT; Fig. 2b) to generate first-strand products. Following this, a poly(A) tail is appended using terminal deoxynucleotidyl- transferase (Tdt) and dATP. Amplification is then achieved using the hybrid primer QT to form the second strand of cDNA, the QO primer, and a GSP upstream of the one used for reverse transcrip- tion. Finally, a second set of PCR cycles is carried out using nested primers (QI and GSP2) to increase the yield of specific product 4. Updated RACE Techniques The most technically challenging step in classic 5¢ RACE is to cajole reverse transcriptase to copy the mRNA of interest in its entirety into first-strand cDNA. Because prematurely terminated first- strand cDNAs are tailed by terminal transferase just as effectively as full-length cDNAs, cDNA populations that are composed largely of prematurely terminated first strands will result primarily in the amplification and recovery of cDNA ends that are not full length (Fig. 3a). This problem is regularly encountered for vertebrate p u or G g n ih si lb uP er u ta N 600 2 © n at ur ep ro to co ls / m oc . er ut a n . w w w//:ptth mRNA Partial cDNA clone 5′ UTR Coding 3′ UTR Figure 1 | A schematic representation of the setting in which Classic RACE is used. The figure shows a mRNA for which only a partial, internal cDNA is available. NATURE PROTOCOLS | VOL.1 NO.6 | 2006 | 2555 PROTOCOL genes, which are often GC-rich at their 5¢ ends and, therefore, often contain sequences that hinder reverse transcription. A number of laboratories and companies have developed steps or protocols that are designed to overcome the problem17–24. New RACE. One approach to force the specific acquisition of full- length 5¢ cDNAs consists of ligating an anchor primer to the 5¢ end of the mRNA before performing the reverse transcription step (Fig. 3b)17. Accordingly, subsequently generated cDNAs that do not extend all the way to the 5¢ end of the transcript fail to incorporate the anchor sequence and do not get amplified in the ensuing PCR mediated by the gene-specific and anchor primers. This method, which is discussed in an accompanying protocol25, is more powerful than the classic 5¢ RACE protocol described here, but is also more challenging to perform. Cap-switching RACE. A simpler method to amplify only full length cDNA ends involves adapter addition during reverse tran- scription (cap-switching RACE; Fig. 4). This method takes advantage of the propensity of Moloney murine leukemia virus reverse transcriptase to add an extra 2–4 cytosines to the 3¢ ends of newly synthesized cDNA strands upon reaching the cap structure at the 5¢ end of the mRNA template26,27. In the presence of a primer terminating in multiple Gs at its 3¢ end, annealing and then complementary copying of the sequence of the annealed oligo takes place, which adds a linker sequence to the cDNA terminus. Because the template-independent addition of cytosines is cap-dependent, the oligo is appended only to full-length cDNA ends. Also, because this method involves fewer steps than classic and new RACE, it is simpler; however, the presence of the dG-terminating (‘switch’) primer can cause problems if it binds to C-rich sequences in the mRNA of interest. Making cDNA ends meet. Another variation of RACE allows for the simultaneous amplification of both ends of a cDNA molecule, eliminating the need for performing two separate 5¢ and 3¢ RACE reactions21,28. One version of this approach28 is achieved through a combination of a standard reverse transcription template switching (TS) reaction — in which a so-called TS-oligo is added that allows the reverse transcriptase enzyme to switch templates from the mRNA to the oligonucleotide, creating a double-stranded molecule — and inverse PCR, with a crucial ligation step between them. The ligation reaction circularizes the double-stranded cDNA, allowing primers that are directed away from the unknown sequence to be used. This straightforward method has been reported to compare very favorably to standard RACE techniques with respect to sensitivity and specificity 28. Commercially available RACE kits and their limitations. Var- ious commercial RACE kits are available, including Clontech’s cap- finding (switching) Smart RACE system, Invitrogen’s 5¢ RACE p u or G g n ih si lb uP er u ta N 600 2 © n at ur ep ro to co ls / m oc . er ut a n . w w w//:ptth mRNA Reverse transcription 1st strand cDNA GSP1 GSP2 First set of amplifications Second set of amplifications "cDNA 3′ End" mRNA Reverse transcription 1st strand cDNA GSP-RT GSP-RT cDNA tailing 1st strand cDNA GSP-RT First set of amplifications Second set of amplifications GSP1 GSP2 "cDNA 5′ End" Xho I Sst I Hind IIIQT QO QI QI QT QO QO-QI- QTQO QTQO QO QI QI * 3′5′ * a b c QI Figure 2 | A schematic representation of Classic RACE. Please see text for details. (a) Amplification of 3¢ partial cDNA ends. (b) Amplification of 5¢ partial cDNA ends. (c) Schematic representation of the primers used in Classic RACE. The 52 nucleotide QT primer (5¢ QO-QI-TTTT 3’) contains a 17-nucleotide oligo-(dT) sequence at the 3¢ end followed by a 35-nucleotide sequence encoding Hind III, Sst I, and Xho I recognition sites. The QI and QO primers overlap by a single nucleotide; the QI primer contains all three of the restriction enzyme recognition sites. Optionally, two additional nucleotides can be added to the 3¢ end of QT to force it to bind to the junction of the cDNA and the poly(A) tail: (G, A or C, followed by G, A, T or C). Primers: QT: 5¢- CCAGTGAGCAGAGTGACGAGGACTCGAGCTCAAGCTTTTTTTTTTTTTTTTT-3¢ QO: 5¢- CCAGTGAGCAGAGTGACG-3¢ QI: 5¢-GAGGACTCGAGCTCAAGC-3¢ GSP1, gene-specific primer 1; GSP2, gene-specific primer 2; GSP-RT, gene-specific primer, used for reverse transcription; *-, GSP-Hyb/Seq (a gene-specific primer for use in hybridization and sequencing reactions). Classic RACE New RACE mRNA mRNA Reverse transcription GSP-RT Ligation of RNA oligo mRNA Reverse transcription GSP-RT Poly(A) tailing * * * * * * a b Figure 3 | The advantage of new RACE over classic RACE. (a) In classic RACE, premature termination in the reverse transcription step results in polyadenylation of less-than-full-length first-strand cDNAs, all of which can be amplified using PCR to generate less-than-full length cDNA 5¢ ends. The asterisk indicates cDNA ends that will be amplified in the subsequent PCR. (b) In new RACE, less-than-full-length cDNAs are also created, but only full-length molecules are terminated by the RNA oligonucleotide (the anchor sequence) and hence amplified in the subsequent PCR. 2556 | VOL.1 NO.6 | 2006 | NATURE PROTOCOLS PROTOCOL system, and Ambion’s First-Choice RLM-RACE kit. The Ambion kit has been a popular choice for validating micro RNA (miRNA) cleavage sites29,30. Commercial systems are often geared toward the construction of universal pools of full-length cDNAs (Fig. 4), in which all of the mRNAs in the starting material become converted to cDNA. The value of this approach is that a single reverse- transcription pool can, in theory, be used to obtain the 5¢ end of any transcript. By contrast, non-commercial versions of RACE have emphasized the use of a GSP to generate the first-strand cDNA templates. Although it lacks universality, the latter approach is more powerful because the reverse transcription step starts closer to the 5¢ end of transcript, and the relative frequency of the desired cDNA is increased 450-fold in the resulting pool. This greatly increases the chances of the desired 5¢ end being present in sufficient quantity to be amplified using standard PCR methods. Which approach should investigators choose? For the one-time user or for those with limited molecular biology experience, the most practical approach would be to obtain a commercial system and, if possible, a pre-made pool of reverse-transcribed cDNAs. Pools representing many human tissues are available — for example, from Clontech or Origene. Failing that, using a GSP-RT primer with the commercial kits will overcome the limitation described above. The Clontech and Ambion systems are relatively powerful and easy to use (both are variations on new RACE); however, they may not be optimal for every purpose. Invitrogen’s system is simpler and less powerful (a variation on classic RACE), but may suffice for many needs. In addition, because the commer- cial kits are relatively expensive, investigators who plan to use RACE regularly will achieve substantial savings if they prepare the reagents themselves. Experimental design considerations for 5¢ RACE Reverse transcription reaction. In 5¢ end cDNA amplification, the efficiency of cDNA extension is crucial. In the classic 5¢ procedure, each specific cDNA, no matter how short, is tailed and becomes a potential target for amplification (Fig. 2a). Thus, the quality of the final PCR products directly reflects that of the reverse transcription reaction. The length of the first-strand cDNA can be maximized by using clean, intact RNA, and by selecting a reverse transcriptase primer that anneals near to the 5¢ end of a region of known sequence. Improvements can also be made, at least in theory, by using a combination of SuperScript II and heat-stable reverse transcriptase at multiple temperatures. At increased tem- peratures the amount of secondary structure encountered in GC- rich regions of the mRNA should be reduced. Incorporation of cyclic compatible solutes such as homoectoine can also improve the generation of first-strand cDNA or the subsequent PCR amplifica- tion steps 31,32. Poly(A) tailing reaction. To attach a known sequence to the 5¢ end of the first-strand cDNA, a homopolymeric tail is appended using Tdt. It is preferable to add poly(A) tails4 rather than poly(C) tails2 for a number of reasons. First, the 3¢ end strategy is based on the naturally occurring poly(A) tail; adding a poly(A) tail to the 5¢ end allows the same adapter primer to be used for both ends, which simplifies the protocol and reduces the cost. Second, because A:T binding is weaker than G:C binding, longer stretches of A residues (approximately two times longer) are required before the oli- go(dT)-tailed QT primer will bind to the template. Internal poly(A) tracts are rare so the chance of non-specific binding and the production of truncated amplification products is reduced. Third, vertebrate coding sequences and 5¢ untranslated regions tend to be biased toward G/C residues; therefore, use of a poly(A) tail further decreases the likelihood of inappropriate amplification. Unlike many other applications that use homopolymeric tails, the actual length of the tail added here is unimportant, as long as it exceeds 17 nucleotides. This is because the oligo(dT)-tailed primer binds at the junction of the appended poly(A) tail and the cDNA transcript. The conditions described in the procedure result in the addition of 30-400 A residues. Many of the remarks made above apply also to the protocol on amplifying 3¢-end partial cDNAs16 and should be noted. There is, however, one major difference. The annealing temperature in the first step of 5¢ RACE (48 1C) is lower than that used in successive cycles (52–68 1C). This is because cDNA synthesis during the first round depends on the interaction of the appended poly(A) tail and the oligo(dT)-tailed QT primer. In all subsequent rounds, ampli- fication can proceed using the QO primer, which is composed of B60% GC and which can anneal to its complementary target at a p u or G g n ih si lb uP er u ta N 600 2 © n at ur ep ro to co ls / m oc . er ut a n . w w w//:ptth Reverse transcription BiotinPi PO-5′ PTotal BiotinPi PO-5′ Template switch Biotin 3′ -RACE GSP1 GSP2 GS-Hyb PO Pi 5′ -RACE RT 5′-UO Ui 3′ 5′-UO Ui 3′ Cap 5′-UO Ui 3′ 3′-UO Ni Cap UO Ui GSP-Hyb RGSP1 RGSP2 Cap Pi PO-5′ - a b c Figure 4 | Schematic representation of cap-switching RACE. (a) Reverse transcription, template switch and incorporation of adaptor sequences at the 3’-end of first strand of cDNA. Biotin-labeled primer Ptotal is used to initiate reverse transcription through hybridization of the poly(dT) tract with the mRNA poly(A) tail. After reaching the 5¢ end of the mRNA, oligo(dC) is added by reverse transcriptase in a cap-dependent manner. Following this, through template switch via base-pairing between the oligo(dC) and the oligo(dG) at the end of cap finder Adaptor, the reverse complementary sequence of the cap finder primer is incorporated into the first strand of the cDNA. Dotted line, mRNA; solid line, cDNA; rectangle, primer. The bracket indicates the known region. (b) The first round of PCR uses primer Uo and RGSP1 (reverse gene-specific primer 1), the 2nd round, Ui and RGSP2. GSP-Hyb is also within the known region, and it can be used to confirm the authenticity of the RACE product. (c) 3¢-RACE. Reprinted with permission from ref. 35. NATURE PROTOCOLS | VOL.1 NO.6 | 2006 | 2557 PROTOCOL much higher temperature. In practice, the ideal melting tempera- tures will vary with individual PCR machines made by different companies. Here we provide a detailed protocol for classic RACE, describing the reverse transcription steps, addition of the poly(A) tail and the subsequent rounds of PCR amplification. MATERIALS REAGENTS .dNTP solution (containing all four dNTPs, each at 10mM) .DTT (0.1 M) .Tris–EDTA solution (10 mM Tris-HCl [pH 7.5], 1 mM EDTA [pH 8.0]) .Reverse transcription buffer, 5x (as supplied by manufacturer) .RNase H .RNasin .SuperScript II reverse transcriptase (Invitrogen) .Gene-specific primer, used for reverse transcription (GSP-RT primer; 100 ng/ml) .Poly(A)+ RNA, or total RNA. Poly(A)+ RNA is used in preference to total RNA for reverse transcription to reduce background, but it is unnecessary to prepare it if only total RNA is available .CoCl2 (25 mM) .dATP solution (1 mM) .Terminal deoxynucleotidyltransferase (Tdt, Invitrogen or Boehringer Mannheim) .Tailing buffer, 5x (125 mM Tris-HCl, pH 6.6, 1 M potassium cacodylate, 1250 mg/ml BSA) .Hercules Hot-Start polymerase buffer (10x) m CRITICAL If the buffer contains dNTPs already, do not add additional nucleotides to the mixture .Common oligonucleotide primers (see REAGENT SETUP for primer design details); for example:QT: 5¢- ccagtgagcagagtgacgaggactcgagctcaagcttttttttttttttt tt-3¢ .QO: 5¢- ccagtgagcagagtgacg-3¢ .QI: 5¢-gaggactcgagctcaagc-3¢ .Gene-specific oligonucleotide primers (user-specific, see REAGENT SETUP for primer design details) EQUIPMENT .Water baths or heating blocks preset to 371, 421, 501, 651, 701 and 80 1C .QIAquick DNA clean-up spin columns (Qiagen) or equivalent .Programmable thermal cycler REAGENT SETUP Primer design for 5¢ RACE QT is a multipurpose primer. It contains binding sites for two, mostly non-overlapping smaller primers (QO (Qouter) and QI (Qinner)) and an oligo dT sequence capable of annealing to the appended poly(A) tail, terminated by a non-A nucleotide to force the primer to set at the junction of the appended poly(A) tail and the bona fide cDNA sequence. The oligo-dT needs to be at least 17 nucleotides in length to anneal at 48 1C. The QO and QI primers should be designed to work well when