TY - JOUR
AU1 - Froim, Doriana
AU2 - Hopkins, Christopher E.
AU3 - Belenky, Alexei
AU4 - Cohen, Aharon S.
AB - Abstract The progress of antisense DNA therapy demands development of reliable and convenient methods for sequencing short single-stranded oligonucleotides. A method of phosphorothioate antisense DNA sequencing analysis using UV detection coupled to capillary electrophoresis (CE) has been developed based on a modified chain termination sequencing method. The proposed method reduces the sequencing cost since it uses affordable CE-UV instrumentation and requires no labeling with minimal sample processing before analysis. Cycle sequencing with ThermoSequenase generates quantities of sequencing products that are readily detectable by UV. Discrimination of undesired components from sequencing products in the reaction mixture, previously accomplished by fluorescent or radioactive labeling, is now achieved by bringing concentrations of undesired components below the UV detection range which yields a ‘clean’, well defined sequence. UV detection coupled with CE offers additional conveniences for sequencing since it can be accomplished with commercially available CE-UV equipment and is readily amenable to automation. Introduction With the increasing number of antisense drugs entering different stages of clinical trials, the demand grows for convenient methods to provide complete sequence information for short single-stranded oligonucleotides. Antisense drugs are short DNA analogs designed primarily as complements to specific regions of the viral mRNA, and the correct sequence of antisense DNA ensures that it will be capable of forming a stable duplex with the target mRNA ( 1 , 2 ). The synthetic DNA sequence is critical for drug performance since even a single base mismatch can prevent specific hybridization and may cause undesired binding outside the target mRNA region. Sequence confirmation analysis is therefore an essential analytical tool required by regulatory agencies for every potential drug to ensure that no error was made during the synthesis. Existing methods of DNA sequencing ( 3 , 4 ) require radioactive or fluorescent labeling of sequencing products as well as expensive and sophisticated equipment for analysis. Both chain termination ( 3 ) and chemical degradation ( 4 ) sequencing methods require modifications to be applied to antisense DNA analysis ( 5–7 ). A method for sequencing of short DNA phosphorothioates by capillary electrophoresis (CE) with laser-induced fluorescence (LIF) detection has been developed and successfully applied in our laboratory ( 6 ). In an improvement to this technique, we have employed ThermoSequenase to generate the DNA extension products by cycle sequencing ( 8 ). Thermocycling permits the incorporation of a high enough number of primer molecules to generate quantities of sequencing fragments readily detectable by UV; therefore, a convenient and affordable setup can now be used for detection. Separation of sequencing products from undesired components of the reaction mixture which previously had been accomplished by fluorescent or radioactive labeling is now achieved because concentrations of undesired components are below the UV detection range. Time-consuming purification procedures previously required for template preparation ( 6 ) have been avoided by thermocycling. The modified method also introduces an additional ligation step which extends the antisense DNA from the 5′-end, allowing a ‘clean’ sequence read for the entire antisense compound. Materials and Methods Materials Primer M13mp18(−21), 5′-auxiliary DNA and 5′-bridge DNA were synthesized in-house on an Expedite nucleic acid synthesis system (PerSeptive Biosystems, Framingham, MA). The 3′-auxiliary DNA and 3′-bridge DNA were custom synthesized by Life Technologies (Rockville, MD). ThermoSequenase (32 U/µl) was obtained from USB/Amersham Life Sciences (Arlington Heights, IL). T4 polynucleotide kinase (10 000 U/ml) and T4 DNA ligase (400 000 U/ml) with 10× T4 DNA ligase buffer (10 mM ATP) were obtained from New England BioLabs (Beverly, MA). dNTPs (100 mM) and ddNTPs (10 mM) were obtained from USB/Amersham Life Sciences. Template preparation The template for the cycle sequencing reaction was prepared by ligating auxiliary DNAs from both 3′- and 5′-ends of the antisense DNA in one phosphorylation/ligation reaction. To prepare the reaction mixture, the following components were combined in an Eppendorf tube: 150 pmol phosphorothioate antisense DNA, 840 pmol 3′-bridge DNA, 30 pmol 3′-auxiliary DNA, 330 pmol 5′-bridge DNA, 150 pmol 5′-auxiliary DNA and 2× T4 DNA ligase buffer. The final reaction volume after enzyme addition was 20 µl. Phosphorylation/ligation reaction was accomplished by adding 20 U T4 polynucleotide kinase and 800 U T4 DNA ligase and incubating for 3 h at room temperature. Figure 1 View largeDownload slide Template preparation strategy. Extension of antisense DNA from both 3′- and 5′-ends by specific ligation illustrates the rationale for template preparation. The 30mer 3′-auxiliary DNA (5′-AAACCCAAAAAAACTGGCCGTCGTTTCA-3′) is designed to have the priming site for M13mp18(−21) primer (5′-TGTAAAACGACGGCCAGT-3′) as the first 18 bases on the 3′-end. The subsequent 12 bases provide a signal sequence which denotes the beginning of the antisense DNA sequence. The 5′-auxiliary DNA allows a ‘clean’ sequence read to be obtained through the last base on the 5′-end of the antisense DNA, and serves as a signal sequence to denote the end of the antisense DNA sequence. The two bridge DNAs (5′-GGGTTTAGAAGG-3′ and 5′-CGAGAGATCAGT-3′ for the 3′- and 5′-bridge DNAs, respectively) are used to support specific ligation on both 3′- and 5′-ends of antisense oligonucleotide. Figure 1 View largeDownload slide Template preparation strategy. Extension of antisense DNA from both 3′- and 5′-ends by specific ligation illustrates the rationale for template preparation. The 30mer 3′-auxiliary DNA (5′-AAACCCAAAAAAACTGGCCGTCGTTTCA-3′) is designed to have the priming site for M13mp18(−21) primer (5′-TGTAAAACGACGGCCAGT-3′) as the first 18 bases on the 3′-end. The subsequent 12 bases provide a signal sequence which denotes the beginning of the antisense DNA sequence. The 5′-auxiliary DNA allows a ‘clean’ sequence read to be obtained through the last base on the 5′-end of the antisense DNA, and serves as a signal sequence to denote the end of the antisense DNA sequence. The two bridge DNAs (5′-GGGTTTAGAAGG-3′ and 5′-CGAGAGATCAGT-3′ for the 3′- and 5′-bridge DNAs, respectively) are used to support specific ligation on both 3′- and 5′-ends of antisense oligonucleotide. Figure 2 View largeDownload slide UV electropherograms of phosphorylation/ligation reaction before and after addition of enzymes. UV electropherograms show the ligation mixture ( a ) prior to addition of enzymes and ( b ) after incubation with ligase and kinase for 3 h at room temperature. The peak at 26 min on electropherogram (b) is presumed to be the product of the antisense DNA ligation to auxiliary DNAs from both 3′- and 5′-ends, which serves as a template for the cycle sequencing reaction. For details on CE-UV conditions, see the Materials and Methods section. Figure 2 View largeDownload slide UV electropherograms of phosphorylation/ligation reaction before and after addition of enzymes. UV electropherograms show the ligation mixture ( a ) prior to addition of enzymes and ( b ) after incubation with ligase and kinase for 3 h at room temperature. The peak at 26 min on electropherogram (b) is presumed to be the product of the antisense DNA ligation to auxiliary DNAs from both 3′- and 5′-ends, which serves as a template for the cycle sequencing reaction. For details on CE-UV conditions, see the Materials and Methods section. Cycle sequencing reaction To prepare the cycle sequencing reaction mixture, the following components were combined in an Eppendorf tube and put on ice: 100 pmol primer, 0.1 pmol template, 5 µl Tris pH 9 (250 mM), 5 µl MgCl 2 (50 mM), 5 µl dNTP mix (10 mM) and 16 U ThermoSequenase. The total reaction mixture volume was 45 µl. The sequencing reaction was initiated by adding 10 µl of reaction mixture to four 0.2 ml thin-wall reaction tubes (MJ Research, Watertown, MA) containing 1.25 µl of the appropriate ddNTP (0.5 mM) each and by thermocycling in an MJ Research Thermal Cycler (PTC-100) for 200 cycles; each cycle included three consecutive steps: 95°C, 30 s; 50°C, 30 s; 72°C, 30 s. Samples were kept on ice before they were introduced to a preheated block at 95°C and then kept at 95°C for 1 min before the first cycle. CE-UV analysis The only pre-treatment required for cycle sequencing samples before CE-UV analysis is desalting. For this purpose, samples were drop-dialyzed for 1 h on 0.025 µm pore-size VS filters from Millipore (Bedford, MA) and then run on 13% linear polyacrylamide, 35% formamide gel columns [the preparation of gel-filled columns for separation of DNA sequencing samples has been previously described ( 6 )]. The running buffer consisted of 0.1 M Tris-borate, 2 mM EDTA and 7 M Urea. A 30 kV, 500 µA direct current high voltage power supply (Model ER/DM; Glassman, Whitehouse Station, NJ) was used to generate the potential across the capillary. UV detection at 270 nm was accomplished with a Spectra 100 variable wavelength detector (Spectra-Physics, San Jose, CA). Data were acquired and processed using TurboChrom IV software (PE Nelson, Cupertino, CA) through an analog-to-digital converter (Model 970A, PE Nelson) and stored on an Digital Venturis 590 computer (Digital, Maynard, MA). Chromatogram alignment was accomplished using two known peaks at the beginning and at the end of the sequence (for example, the primer peak and one of the 5′-end signal sequence peaks) as two internal standards. Since primer was completely incorporated in the sequencing reactions, samples were spiked with 0.1 pmol primer to provide a reliable internal standard ( Fig. 3 ). The rationale for alignment of CE chromatograms has been previously described ( 6 ). Figure 3 View largeDownload slide UV electropherograms of individual sequencing reactions. Four separate UV electropherograms show the products of four individual sequencing reactions. The first base of antisense DNA (elecropherogram c) is located immediately after the last base of 3′-auxiliary DNA signal sequence (AAAAAACCCAAA), and the last base of antisense DNA (elecropherogram b) is located just before the first base of 5′-auxiliary DNA signal sequence (TAGTCAGTCAGT). The correct antisense DNA sequence is obtained by aligning the four individual chromatograms: 3′-TCTTCCTCTCTCTACCCACGCTCTC-5′. Since primer was depleted in sequencing reactions, samples were spiked with 0.1 pmol primer (peaks marked by arrows) to clearly denote the beginning of 3′-end signal sequence and provide a reliable internal standard for the alignment of electropherograms. A primer spike can be done whenever primer is completely incorporated in the sequencing reactions and an internal standard is needed. See CE-UV analysis in the Materials and Methods section. Figure 3 View largeDownload slide UV electropherograms of individual sequencing reactions. Four separate UV electropherograms show the products of four individual sequencing reactions. The first base of antisense DNA (elecropherogram c) is located immediately after the last base of 3′-auxiliary DNA signal sequence (AAAAAACCCAAA), and the last base of antisense DNA (elecropherogram b) is located just before the first base of 5′-auxiliary DNA signal sequence (TAGTCAGTCAGT). The correct antisense DNA sequence is obtained by aligning the four individual chromatograms: 3′-TCTTCCTCTCTCTACCCACGCTCTC-5′. Since primer was depleted in sequencing reactions, samples were spiked with 0.1 pmol primer (peaks marked by arrows) to clearly denote the beginning of 3′-end signal sequence and provide a reliable internal standard for the alignment of electropherograms. A primer spike can be done whenever primer is completely incorporated in the sequencing reactions and an internal standard is needed. See CE-UV analysis in the Materials and Methods section. Results and Discussion When DNA is sequenced using Sanger's approach, sequence information cannot be directly obtained for the region complementary to the primer. Since we want to determine the sequence of antisense DNA from the first to the last base, the antisense oligonucleotide is extended from the 3′-end with 3′-auxiliary DNA to provide a priming site outside the antisense DNA. The 3′-auxiliary DNA is designed to have a region complementary to the sequence of the primer, followed by a 12 base signal sequence region which denotes the beginning of the sequence of interest. Extension of antisense DNA from the 5′-end is introduced to solve the problem of last peak identification, which is usually complicated by non-templated addition of nucleotide catalysed by the sequencing enzyme ( 9–11 ). Once the 5′-end of antisense DNA is ligated to the 5′-auxiliary DNA, a ‘clean’ sequence read is obtained over the last base of antisense DNA and into the 5′-auxiliary DNA. Since the sequence of 5′-auxiliary DNA is known, it also serves as a 5′-end signal sequence denoting the end of the antisense DNA. Figure 4 View largeDownload slide Computer overlay of individual UV electropherograms. Four individual UV electropherograms were aligned using two known peaks from the 3′-and 5′-end signal sequences as internal standards and overlayed to obtain an ‘easy-to-read’ sequence. Figure 4 View largeDownload slide Computer overlay of individual UV electropherograms. Four individual UV electropherograms were aligned using two known peaks from the 3′-and 5′-end signal sequences as internal standards and overlayed to obtain an ‘easy-to-read’ sequence. Additional 12mer bridge DNAs are used to hybridize the auxiliary DNAs and antisense oligonucleotide into a substrate for the ligation reaction ( Fig. 1 ). This limits the method to sequence confirmation rather than general sequencing since the first six bases from the 3′-end of the antisense oligomer must be known prior to using the bridge ligation procedure. Should we need to analyze antisense DNA of unknown sequence, the blunt ligation procedure with T4 RNA ligase can be utilized ( 6 ). Under the appropriate conditions, T4 RNA ligase can ligate any two unknown sequences of ssDNA ( 12 ). Since blunt ligation using T4 RNA ligase can overcome the limitations of the bridge approach, this method can be used for sequence determination of unknown DNA. The electropherograms of Figure 2 show the ligation mixture prior to addition of enzymes and after incubation with ligase and kinase for 3 h at room temperature. The peak at 26 min is presumed to be the product of antisense DNA ligation to auxiliary DNAs from both 3′- and 5′-ends which serves as a template for the cycle sequencing reaction. The previously required template purification ( 6 ) becomes unnecessary with the introduction of cycle sequencing. The 12mer bridge melting temperature is well below the annealing temperature employed in this protocol, and under the experimental conditions the bridge cannot re-hybridize to the complementary site on the ligated DNA. As a result, there is no bridge/template hybrid to interfere with primer extension ( 6 ) or to generate extension products primed from the bridge DNA. Sequencing fragments are generated across the antisense DNA of interest and for both signaling regions of 3′- and 5′-end auxiliary DNAs, giving a complete sequence read for the DNA of interest from the first to the last base ( Figs 3 and 4 ). The first base of the antisense DNA is identified as the first base immediately after the signal sequence of the 3′-end auxiliary DNA ( Fig. 3c ). The last base on the 5′-end is identified as the last base before the start of the signal sequence of the 5′-end auxiliary DNA ( Fig. 3b ). In the first base position, an interfering peak appears as the result of sequencing non-ligated 3′-end auxiliary DNA, followed by non-templated nucleotide addition catalysed by ThermoSequenase ( 9–11 ). This peak can be reduced to noise level by optimizing the ligation procedure toward complete consumption of 3′-end auxiliary DNA. The first base identity is confirmed however by the successful 3′-end ligation which would not occur in the case of a first base mismatch. A clean sequence read is obtained for the last base on the 5′-end of antisense DNA. In addition to ddNTP terminated sequencing products, chromatograms on Figure 3 show dNTP terminated products of primer extension as sequencing reaction noise most noticeable immediately after the primer. This noise can be significantly reduced by lowering the primer concentration to force dNTP terminated fragments to act as primers and subjecting them to further elongation. With the current reaction protocol, complete primer incorporation can be observed with significantly fewer cycles; however, less cycling increases dNTP terminated reaction noise. Components of the ligation mixture such as antisense DNA, 3′-end auxiliary DNA, and ligation products do not appear on the chromatogram as interfering peaks since they are present in the reaction mixture in amounts below the UV detection range. The cycle sequencing products, on the other hand, are generated in quantities an order of magnitude higher than the template DNA and are readily detected by UV. Thus, the separation of sequencing products from other components of the reaction mixture, previously achieved by fluorescent labeling, is now accomplished by reducing the concentrations of undesired components while increasing the concentrations of the sequencing products. This composition of the final sequencing reaction mixture makes UV detection of the sequencing products possible without any additional labeling or purification. Conclusions Optimization of the cycle sequencing reaction protocol toward incorporation of high quantities of primer has made it possible to generate sufficient quantities of sequencing products for low sensitivity UV detection. The proposed method of sample preparation can be used with UV detection instrumentation available commercially and can be automated. UV detection of sequencing products reduces the cost of sequencing since it does not require labeling and any additional sample processing beyond thermocycling and desalting. A key to the successful application of UV detection is the correct composition of the final reaction mixture which must discriminate against possible interference peaks. Although this method was originally designed to sequence short phosphorothioate antisense oligonucleotides, the proposed sample preparation procedure does not discriminate between phosphorothioate and phosphodiester; therefore, it can be successfully used for sequencing of short synthetic phosphodiester DNAs as well. References 1 Zamecnik P.. ,  Antisense Nucleic Acid Drug Dev. ,  1997, vol.  7 (pg.  199- 202) CrossRef Search ADS PubMed  2 Uhlmann E.,  Peyman A.. ,  Chem. Rev. ,  1990, vol.  90 (pg.  544- 583) CrossRef Search ADS   3 Sanger F.,  Nicklen S.,  Coulson A.R.. ,  Proc. Natl. Acad. Sci. USA ,  1977, vol.  93 (pg.  709- 713) 4 Maxam A.M.,  Gilbert W.. ,  Proc. Natl. Acad. Sci. USA ,  1977, vol.  74 (pg.  560- 564) CrossRef Search ADS   5 Wyrzykiewicz T.K.,  Cole D.L.. ,  Nucleic Acids Res. ,  1994, vol.  22 (pg.  2667- 2669) CrossRef Search ADS PubMed  6 Belenky A.,  Smisek D.L.,  Cohen A.S.. ,  J. Chromatogr. A ,  1995, vol.  700 (pg.  137- 149) CrossRef Search ADS PubMed  7 Tang J.,  Roskey M.,  Agrawal S.. ,  Anal. Biochem. ,  1993, vol.  212 (pg.  134- 137) CrossRef Search ADS PubMed  8 Tabor S.,  Richardson C.. ,  Proc. Natl. Acad. Sci. USA ,  1995, vol.  92 (pg.  6339- 6343) CrossRef Search ADS   9 Hu G.. ,  DNA Cell Biol. ,  1993, vol.  12 (pg.  763- 770) CrossRef Search ADS PubMed  10 Clark J.M.. ,  Nucleic Acids Res. ,  1988, vol.  16 (pg.  9677- 9686) CrossRef Search ADS PubMed  11 Brownstein M.J.,  Carpten J.D.,  Smith J.R.. ,  Biotechniques ,  1996, vol.  20 (pg.  1004- 1006) PubMed  12 Tessier D.C.,  Brousseau R.,  Vernet T.. ,  Anal. Biochem. ,  1986, vol.  158 (pg.  171- 178) CrossRef Search ADS PubMed  © 1997 Oxford University Press
TI - Method for phosphorothioate antisense DNA sequencing by capillary electrophoresis with UV detection
JF - Nucleic Acids Research
DO - 10.1093/nar/25.21.4219
DA - 1997-11-01
UR - https://www.deepdyve.com/lp/oxford-university-press/method-for-phosphorothioate-antisense-dna-sequencing-by-capillary-uwA7VhjjnV
SP - 4219
EP - 4223
VL - 25
IS - 21
DP - DeepDyve
ER -