key: cord-0700174-v9mh7nnt
authors: Thiel, Volker; Rashtchian, Ayoub; Herold, Jens; Schuster, David M.; Guan, Nin; Siddell, Stuart G.
title: Effective Amplification of 20-kb DNA by Reverse Transcription PCR
date: 1997-10-01
journal: Analytical Biochemistry
DOI: 10.1006/abio.1997.2307
sha: b3a4e8c6c078f130d118f060709cadee335b1b61
doc_id: 700174
cord_uid: v9mh7nnt

Abstract Polymerase chain reaction has been applied to the amplification of long DNA fragments from a variety of sources, including genomic, mitochondrial, and viral DNAs. However, polymerase chain reaction amplification from cDNA templates produced by reverse transcription has generally been restricted to products of less than 10 kilobases. In this paper, we report a system to effectively amplify fragments up to 20 kilobases from human coronavirus 229E genomic RNA. We demonstrate that the integrity of the RNA template and the prevention of false priming events during reverse transcription are the critical parameters to achieve the synthesis of long cDNAs. The optimization of the polymerase chain reaction conditions enabled us to improve the specificity and yield of product but they were not definitive. Finally, we have shown that the same reverse transcription polymerase chain reaction technology can be used for the amplification of extended regions of the dystrophin mRNA, a cellular RNA of relatively low abundance.

rose gels. The gels were dried and hybridized with 32 P-(5-end)-labeled oligonucleotide 55, (Table 1, Fig. 2 ) as described (14) .

Fifty to 500 ng of poly(A)-RNA from HCV 229E-infected MRC-5 cells were used for reverse transcription in a volume of 20 ml with 1 ml (200 U) SuperScript II reverse transcriptase (Life Technologies), 4 ml of 51 first-strand buffer (Life Technologies), 2 ml of 10 mM dNTPs, 2 ml of 0.1 M DTT, 0.5 ml of RNasin (50 U/ml) (Pharmacia), and 30-100 ng of oligonucleotide primer (Table 1, Fig. 2 ). The reactions were incubated for 90 min at 42ЊC and then heated for 2 min at 94ЊC and chilled on ice.

First-strand cDNA synthesis with poly(A)-RNA from human heart was done using the SuperScript Preamplification System (Life Technologies), oligo(dT) primer, and 1 mg human heart poly(A)-RNA essentially according to the manufacturer's instructions. The cDNA synthesis reaction was incubated at 45ЊC for 45 min followed by heat inactivation of the reverse transcriptase at 70ЊC for 10 min and removal of the mRNA template by incubation with 2U RNase H for 20 min at 37ЊC (10). Amp PCR System 9600. Aliquots of 0.2-3 ml reverse transcription reaction were used as PCR templates. PCRs were done in a total volume of 50 ml with 1 unit poly(U)-Sepharose (13) . In the second method, the cells Elongase Enzyme Mix (Life Technologies) and final were washed twice with ice-cold PBS and then scraped concentrations of 60 mM Tris-SO 4 (pH 9.1), 18 mM and pelleted in 10 ml ice-cold PBS. The cell pellet was (NH 4 ) 2 SO 4 , 2 mM MgSO 4 , 0.2 mM dNTPs, and 0.2-0.4 resuspended in 1.5 ml lysis buffer (10 mM Tris-HCl, mM oligonucleotide primer (Table 1 , Fig. 2 and 7A). pH 7.5, 140 mM NaCl, 5 mM KCl, 1% NP-40) and incu-Unless indicated otherwise, the PCR cycles were 1 min, bated for 30 s on ice. The cell lysate was centrifuged 94ЊC, followed by 30 cycles of 20 s denaturation at 94ЊC, at 1500g for 1 min and the supernatant was incubated 30 s annealing at 50ЊC, and elongation for 1 min/kb for 5 min at 23ЊC with 5 mg oligo(dT) 25 Dynabeads expected product length at 72ЊC. During the last 18 (Dynal) resuspended in 1.5 ml of 21 binding buffer (20 cycles, the elongation time was increased by 30 s in mM Tris-HCl, pH 7.5, 1 M LiCl, 2 mM EDTA, 1% SDS). each successive cycle. The reaction was terminated by The oligo(dT) 25 magnetic beads were washed twice with a 10-min elongation at 72ЊC. wash buffer (10 mM Tris-HCl, pH 7.5, 150 mM LiCl, 1 mM EDTA) and the bound poly(A)-RNA was eluted in 2 mM EDTA, pH 8.0, for 2 min at 65ЊC.

Previous studies have shown that the RNAse Ha gift from Dr. Chris Gruber and had been isolated deficient reverse transcriptase, SuperScript II, is causing TRIZOL reagent (Life Technologies) and olipable of copying RNAs of at least 7.5 kb (15) . Howgo(dT)-cellulose chromatography according to the manever, there has been no systematic study of the maxiufacturer's instruction. mum length of cDNA synthesis that can be achieved Analysis of Viral RNAs with this enzyme. The availability of HCV 229E genomic RNA, combined with long PCR technology, now Poly(A)-RNA from HCV 229E-infected MRC-5 cells was electrophoresed on 2.2 M formaldehyde-1% aga-makes it possible to study in more detail the capabili- ties of SuperScript II and the parameters that are RT primer. Then we tried to amplify, by PCR, DNA fragments that extended 4.8 kb (PCR primers 16 and critical for its effective use. 89) or 9 kb (PCR primers 147 and 89) upstream of the RT primer binding site. Up to a distance of 4.8 kb from The RNA Template the RT priming site, we were able to obtain the ex-HCV 229E genomic RNA has two major advantages pected PCR products, regardless of the poly(A)-RNA for the studies reported here. First, as a viral RNA, preparation that we used as template for the RT reacit is relatively abundant in the infected cell. Second, tion (Fig. 3, lanes 1 and 3) . However, when we tried to coronaviruses are positive-strand RNA viruses, and the synthesize the 9.0-kb PCR product, we only succeeded genomic RNA has a 3 polyadenylate tract that can be when we did the RT reaction with the poly(A)-RNA used for affinity chromatography (13) . Human epithetemplate prepared by the Dynabeads method (Fig. 3 , lium MRC-5 cells were infected with HCV 229E, and lanes 2 and 4). Increasing the amount of poly(U)-Sephthe poly(A)-RNA was isolated by poly(U)-Sepharose arose-purified poly(A)-RNA up to 1 mg in the RT reacchromatography or chromatography with homopolytion did not lead to the synthesis of the 9-kb PCR prodmeric oligonucleotide (dT) 25 coupled to magnetic beads. uct. This result indicates that the quality of the RNA The poly(A)-RNAs were separated by gel electrophorepreparation, i.e., the integrity of the template rather sis and the viral mRNAs were visualized by hybridizathan its abundance, is a critical factor when longer RTtion analysis using the HCV 229E-specific oligonucleo-PCR products are desired. tide 55 (Fig. 1) . In both RNA preparations it is possible to identify the genomic RNA (27.3 kb) and the six sub-The RT Reaction genomic mRNAs (1.7-6.8 kb) that are characteristic of coronavirus infection. The hybridization analysis indi-Adjusting the amount of RT-primer. First, we established an RT-PCR protocol that enabled us to gener-cates that the material isolated with oligo(dT) 25 magnetic beads (lane 2) is less degraded than material iso-ate a DNA product with a size of 12.6 kb. As RNA template we used the poly(A)-RNA that was prepared lated by poly(U)-Sepharose (lane 1).

In order to generate HCV 229E-specific cDNAs, we from HCV 229E-infected cells using oligo(dT) 25 magnetic beads. We primed the reverse transcription with did reverse transcriptions with HCV 229E-specific oligonucleotides and the RNA templates described above. the oligonucleotide 85 and 3 ml of the RT reaction then served as template for a subsequent PCR reaction with To amplify DNA products from the HCV 229E cDNAs, we inserted aliquots of the RT reaction, without further the primers 159 and 89. As is shown in Fig. 4A (lane 3), this protocol resulted in a PCR product with the purification, into the PCR. In a first series of experiments, we carried out the RT reaction using the HCV expected size of 12.6 kb and an additional smaller product of 3.9 kb. Restriction enzyme analysis revealed that 229E-specific oligonucleotide 85 (Table 1, Fig. 2 ) as the the smaller product was amplified from a region of the transcription reaction, in order to reduce the amount of RT primer carryover into the subsequent PCR. The HCV 229E genome, ranging from approximately nucleotide (nt) 300 to nt 4200. Since the reverse transcrip-result of this experiment is shown in Fig. 4A (lanes 1  and 2) . If we reduced the RT primer concentration from tion reaction was performed at 42ЊC, we reasoned that this product could result from the amplification of a 0.75 to 0.5 mM, the amount of the 3.9-kb PCR product decreases significantly without affecting the yield of cDNA generated during the RT reaction by a ''less stringent'' priming event. Indeed, analysis of the HCV the 12.6-kb product (lane 2). If we reduced the RT primer concentration to 0.25 mM, only small amounts 229E genomic RNA sequence revealed a stretch of 6 nucleotides (nt 4205-4210) that match exactly to the of the 3.9-kb product were synthesized, again without affecting the yield of the 12.6-kb product (lane 1). We 3 end of primer 85 (Fig. 4B) . Thus, a priming event at this position, during the RT reaction, would result also tried to reduce the RT primer concentration to 0. 15 mM but, in this case, the yield of the 12.6-kb product in the synthesis of a cDNA with the primer sequence at its 5 end. Since we inserted aliquots of the reverse decreased significantly (data not shown). transcription reaction directly into the PCR, we also introduce significant amounts of the primer 85, and it The PCR Reaction is likely that during the PCR, the 3 truncated cDNA was amplified with the upstream primer 159 and the Adjusting the amount of PCR template. Our next goal was to establish a RT-PCR protocol that enabled reverse transcription primer 85.

To test this hypothesis, and to try to circumvent such us to amplify cDNAs longer than 12.6 kb. On the basis of the results described above, we looked for an RT problems, we performed two further experiments. First we changed the temperature of the reverse transcrip-primer that would fulfill two criteria. First, in order to synthesize cDNAs over 12.6 kb in length, the priming tion reaction from 42 to 45ЊC in order to minimize the occurrence of ''less stringent'' priming events. However, site should be located at least 20 kb downstream from the 5 end of the HCV 229E genomic RNA. Second, this when we used this approach, we only obtained small amounts of the PCR product with the expected size of primer has to be highly specific, in order to minimize the occurrence of ''less stringent'' priming events dur-12.6 kb without diminishing the amount of the 3.9-kb PCR product (data not shown). The second strategy ing the reverse transcription reaction. Therefore, we chose RT-primer 32 at a concentration of 0.25 mM. The was to reduce the primer concentration in the reverse Optimization of the PCR cycle conditions. In parallel to the experiments described above, we set out to optimize the PCR cycle conditions, in order to improve the yield of long RT-PCR products. In most DNA template PCR protocols that aim to produce longer products, the elongation step is performed at 68ЊC. Therefore, we repeated the RT-PCR reactions described above (Fig. 5, lanes 2-4) , but changed the elongation temperature from 72 to 68ЊC. With this change, the amount of full-length (16.7 kb) PCR product synthesized is significantly increased (Fig. 5, lanes 5-7) . This result encouraged us to believe that it should be possible to effectively synthesize RT-PCR products longer than 16.7 kb.

In an attempt to effectively amplify DNA of more than 20 kb by RT-PCR, we used the protocol described above with RT primer 32 and PCR primers 159 and 11 (Table 1, Fig. 2) . In contrast to our standard protocol, the PCR cycle conditions for this exper- 2 and 4) . 5ml aliquots of the PCR reactions were separated on a 0.6% agarose gel. Lane M shows 400 ng of HindIII-EcoRI-digested lDNA.

subsequent PCR was done with 2 ml of the RT reaction as template and oligonucleotides 127 and 11 as upstream and downstream PCR primers, respectively (Table 1, Fig. 2 ). As is shown in Fig. 5 , it was possible to amplify a DNA fragment with the expected size of 11.5 kb (lane 1) using this procedure. However, when we did the PCR with primers 147 and 11 (Table 1, Fig. 2), we obtained not only the desired product with the expected size of 16.7 kb but also a second product with a size of approximately 4.9 kb (lane 2). Again, by restriction enzyme analysis, we identified this 4.9-kb fragment as an amplification product of the HCV 229E genomic RNA. Our interpretation of this result was as described above, so again we tried to lower the RT primer concentration in the reverse transcription reac- length RT-PCR product (Fig. 5, lanes 2-4) .

The experiments described above were all done with poly(A)-RNA from HCV 229E-infected cells as RNA template in the RT reaction. In this material, the viral RNAs constitute 10-20% of the total RNA and the 27 kb viral genomic RNA represents approximately 4% of the viral RNAs or 0.4-0.8% of the total RNA (S. Siddell, unpublished) . Obviously, we were interested to know if the RT-PCR protocols described above could also be used for less abundant RNAs, for example, cellular mRNAs.

The human dystrophin gene was used to study the long RT-PCR amplification of cellular mRNAs. The exon structure of this gene has been characterized and it has been shown to transcribe a 14-kb mRNA with a 4 and 7) of the RT reaction were then used for PCR reactions with the oligonucleotides 127 and 11 (lane 1) or oligonucleotides 147 and 11 (lanes 2 to 7). Elongation temperatures were 72ЊC (lanes 1 to 4) or 68ЊC (lanes 5 to 7). 5-ml aliquots of the PCR reactions were separated on a 0.6% agarose gel. Lane M shows 400 ng of HindIII-EcoRI-digested lDNA.

iment were 12 cycles of 20 s at 94ЊC, 30 s at 50ЊC, and 27 min at 68ЊC, followed by 18 cycles where the elongation time was increased by 30 s in each successive cycle. The result is shown in Fig. 6 . Using this protocol, it was possible to synthesize a DNA product with the expected size of 20.3 kb (lane 1). The identity of this product was confirmed by restriction analysis, however, the yield of the product was relatively poor. Also, we observed an unacceptable amount of heterogeneous DNA products that were clearly visible in agarose gel electrophoresis. Therefore, we did a series of experiments to optimize the PCR conditions by varying the cycle profile. As is shown in 20.3-kb RT-PCR shown in Fig. 6 (lane 2) . 2.7-kb region of 3 untranslated sequence (16; Fig. 7A ). (17, 18) , and, thus, it serves as a good test system for the RT-PCR amplification of low abundance cellular The dystrophin mRNA is expressed in low amounts, approximately 0.01-0.001% of the mRNA in muscle mRNAs.

Nucleic Acids

Poly(A)-RNA from human heart (1 mg) was used as Finally, as is the case for all PCRs, the cycle conditions have to be optimized according to the amount of template for an oligo(dT)-primed reverse transcription reaction as described under Materials and Meth-template, the PCR primers, and the cycle profile.Having shown that it is possible to amplify 20-kb ods. Subsequently, two different sets of PCR primers were tested for the amplification of cDNA segments DNAs by RT-PCR, it is interesting to briefly consider the possible applications of this technique. Our own using the following cycle profile: 94ЊC, 15 s; 62ЊC, 15 s; 68ЊC, 12 min for 35 cycles with an initial 1-min major interest is the RT-PCR amplification of viral RNA sequences and the generation of infectious cDNA denaturation at 94ЊC. The PCR-primers 421U and 10358L produced a 10.0-kb DNA fragment from the clones. Previous studies with picornaviruses, alphaviruses, and flaviviruses have demonstrated how infec-dystrophin cDNA, while the PCR primers 2784U and 13207L amplified a 10.4-kb segment of the cDNA tious cDNA clones can greatly facilitate studies on the molecular biology and pathogenesis of important hu- (Fig. 7B, lanes 1 and 2) . These results show that cDNAs of at least 13.5 kb (including the 3 untrans-man viruses (19) (20) (21) (22) . It would be highly desirable to extend this approach to RNA viruses with genomes lated region) can be effectively synthesized by reverse transcription of the dystrophin mRNA and that of more than 10 kb, for example, coronaviruses and arteriviruses. The feasibility of the RT-PCR approach this cDNA can be readily amplified by PCR.in the generation of an infectious cDNA has already DISCUSSION been demonstrated for the potato virus Y (9) and for hepatitis A virus (6). In this study we have adapted the concept of long PCR technology to reverse transcription PCR. The na-Finally, another important application of long RT-PCR technology will be the amplification of cellular ture of RT-PCR requires the synthesis of a cDNA by reverse transcription prior to its amplification in the mRNAs, particularly those of low abundance. We have shown that RT-PCR can be used for the amplification PCR reaction. We have demonstrated that there is no limitation concerning the ability of reverse tran-of 10-kb products from dystrophin mRNA and we see no reason why the same technology could not be applied, scriptase to synthesize cDNAs of up to 20 kb in sufficient amounts. However, to achieve this goal, a number perhaps in combination with nested-set PCR, to other, perhaps even less abundant, cellular mRNAs. This type of critical parameters have to be kept in mind. First, the integrity of the RNA template is important. De-of application could provide important information in relation to diagnostic procedures and will also find a pending on the source of the RNA template, a method of preparation should be chosen that minimizes degra-wide variety of uses in the areas of molecular and cellular biology. dation of the RNA. In our hands, cell or tissue lysis can be achieved by detergent (NP-40) or TRIZOL (Life Technologies) reagent, as appropriate, and oligo(dT)-ACKNOWLEDGMENT based affinity chromatography with magnetic beads, or cellulose matrix, has proven to be reliable for the plification.