DENS (differential extension with nucleotide subsets): application to the sequencing of human genomic DNA and cDNA.

Here we describe further development of our method of DNA sequencing by Differential Extension with Nucleotide Subsets (DENS) and its application to the sequencing of human genomic DNA and full-insert cDNA. Essentially, DENS is primer walking without custom primer synthesis; instead, DENS uses a presynthesized library of octamer primers degenerate in two positions (4,096 tubes/sequences for a complete library). DENS converts an octamer selected from this library into a long primer on the template, at the intended site only. This is done using a two-step procedure which starts with a limited extension of the octamer (at 20 degrees C) in the presence of only two of the four possible dNTPs. The primer is extended by five bases or more at the intended priming site, which is deliberately selected to maximize the extension length (as are the two-dNTP set and the primer itself). The subsequent termination reaction at 60 degrees C then accepts the primer extended at the intended site, but not at alternative sites, where the initial extension (if any) is generally much shorter. This paper presents a set of rules for selection of DENS priming sites. We also compare different ways of template preparation for DENS sequencing. The data were obtained from primer walking on three human genomic DNA subclones of 3 to 4 kbp and four cDNA clones containing inserts of 1.9, 2.3, 3.8, and 4.9 kbp. Full-length sequences were obtained from both strands of each subclone by automated dye-terminator fluorescent DNA sequencing using DENS with degenerate octamer primers. We compared the following types of DNA templates: single-stranded and double-stranded phagemid DNA, double-stranded PCR products, asymmetric PCR products, and single-stranded DNA produced by digestion with Lambda Exonuclease of double-stranded PCR product phosphorylated at one end (Exo-PCR). While all of the preps were found to work, the best results were obtained with Exo-PCR and phagemid single-stranded DNA. Exo-PCR directly from overnight bacterial culture with no plasmid prep of any kind yielded templates for DENS as good as Exo-PCR from purified DNA. We found that the Tm of the differentially extended octamers is an important factor in the success of DENS. Clustering of successful reactions was clearly distinguished in the Tm range of 50-66 degrees C, with success rates of 70% for Exo-PCR and 65% for ss phagemid templates.

Interdependence between DNA template secondary structure and priming efficiencies of short primers.

Here we analyze the effect of DNA folding on the performance of short primers and describe a simple technique for assessing hitherto uncertain values of thermodynamic parameters that determine the folding of single-stranded DNA into secondary structure. An 8mer with two degenerate positions is extended simultaneously at several complementary sites on a known template (M13mp18) using one, two or three (but never all four) of the possible dNTPs. The length of the extension is site specific because it is limited by the first occurrence in the downstream template sequence of a base whose complementary dNTP is not present. The relative priming efficiencies of different sites are then ranked by comparing their band brightnesses on a gel. The priming efficiency of a short primer (unlike conventional long primers) depends dramatically on the secondary structure of the template at and around the priming site. We calculated the secondary structure and its effect on priming using a simple model with relatively few parameters which were then optimized to achieve the best match between the predictions and the actual rankings of the sites in terms of priming efficiency. This work introduces an efficient and conceptually novel approach that in the future can make use of more data to optimize a larger set of DNA folding parameters in a more refined model. The model we used, however crude it may be, significantly improved the prediction of priming efficiencies of 8mer primers and appreciably raised the success rate of our DNA sequencing technique (from 67 to 91% with a significance of P < 7 x 10(-5)), which uses such primers.

DNA sequencing using differential extension with nucleotide subsets (DENS).

Here we describe template directed enzymatic synthesis of unique primers, avoiding the chemical synthesis step in primer walking. We have termed this conceptually new technique DENS (differential extension with nucleotide subsets). DENS works by selectively extending a short primer, making it a long one at the intended site only. The procedure starts with a limited initial extension of the primer (at 20-30 degrees C) in the presence of only two out of the four possible dNTPs. The primer is extended by 6-9 bases or longer at the intended priming site, which is deliberately selected, (as is the two-dNTP set), to maximize the extension length. The subsequent termination reaction at 60-65 degrees C then accepts the extended primer at the intended site, but not at alternative sites, where the initial extension (if any) is generally much shorter. DENS allows the use of primers as long as 8mers (degenerate in two positions) which prime much more strongly than modular primers involving 5-7mers and which (unlike the latter) can be used with thermostable polymerases, thus allowing cycle-sequencing with dye-terminators compatible with Taq DNA polymerase, as well as making double-stranded DNA sequencing more robust.

On the mechanism of the modular primer effect.

Modular primers are strings of three contiguously annealed unligated oligonucleotides (modules) as short as 5- or 6-mers, selected from a presynthesized library. It was previously found that such strings can prime DNA sequencing reactions specifically, thus eliminating the need for the primer synthesis step in DNA sequencing by primer walking. It has remained largely a mystery why modular primers prime uniquely, while a single module, used alone in the same conditions, often shows alternative priming of comparable strength. In a puzzling way, the single module, even in a large excess over the template, no longer primes at the alternative sites, when modules with which it can form a contiguous string are also present. Here we describe experiments indicating that this phenomenon cannot be explained by cooperative annealing of the modules to the template. Instead, the mechanism seems to involve competition between different primers for the available polymerase. In this competition, the polymerase is preferentially engaged by longer primers, whether modular or conventional, at the expense of shorter primers, even though the latter can otherwise prime with similar or occasionally higher efficiency.

DNA sequencing: modular primers for automated walking.

Here we describe our progress in the development of the technology of DNA sequencing by primer walking based on the modular primer approach, which eliminates the primer synthesis bottleneck. Modular primers are assembled from 5-mers, 6-mers, or 7-mers selected from a presynthesized library of as few as 1000 oligonucleotides. This technology is expected to reduce the time per walk by a factor of 20 to 50, and the cost of DNA sequencing by a factor of 5 to 15. Both time and expense will be saved not only on the primer synthesis per se but, more importantly, as a result of the closed-end automation of the complete cycle of walking sequencing, where no human intervention will be required between the walks. Such closed-end automation has until now been precluded due to the need to synthesize a new primer for each walk. As a part of the closed cycle automation development, this report deals with modular primers used for sequencing with fluorescent dye terminators. The success rate and the quality of automated sequencing with modular primers are found similar to those with the conventional 17-20-base-long primers. One of the advantages of the described technique is simple purification without any proteins that need to be removed.

DNA sequencing: modular primers assembled from a library of hexamers or pentamers.

Here we report a striking effect displayed by "modular primers," which consist of hexamer or pentamer oligonucleotide modules base-stacked to each other upon annealing to a DNA template. Such a combination of modules is found to prime DNA sequencing reactions uniquely, unlike either of the modules alone. We attribute this effect in part to the increase in the affinity of an oligonucleotide for the template in the presence of an adjacent module. All possible pentamer (or hexamer) sequences total 1024 (or 4096) samples, a manageable size for a presynthesized library. This approach can replace the synthesis of primers, which is the current bottleneck in time and cost of the primer walking sequencing, and can allow full automation of the closed cycle of walking.

DNA trapping electrophoresis.

Attempts to improve the size separation of single-stranded DNA in polyacrylamide gels by field-inversion gel electrophoresis (FIGE) have met with limited success. Here we show that attaching a neutral globular protein, streptavidin, to one end of a single-stranded DNA molecule profoundly alters the DNA mobility pattern and increases the band separation manyfold within a size range controlled by voltage and pulse cycle. In constant field, short modified fragments are only slightly retarded but long molecules are retarded dramatically above a 'threshold size' of 0.6 kilobases at 60 V per cm. At this voltage, molecules above a 1.2-kilobase 'cut-off' do not enter the gel. Both the threshold and the cut-off sizes decrease as the voltage increases. In FIGE, the longer the reverse pulse, the larger the modified fragments that enter the gel. We interpret these results as the trapping by the gel matrix of the protein attached to the DNA. The probability of release then depends on the balance between the electric field and thermal motion: the larger the DNA and the higher the voltage, the harder it is to release.

Estimation of wedge components in curved DNA.

There are growing indications that the inherent curvature of DNA is important in protein-DNA recognition. The 10.5-base-pair (bp) periodicity of some dinucleotides first found in eukaryotic DNA sequences was interpreted as the expression of curvature of periodic segments of double-stranded DNA, the curvature resulting from co-orientation of periodically spaced 'wedges' between stacked base pairs. The wedge can be decomposed into roll and tilt components, opening towards a groove and a backbone respectively, both contributing to DNA curvature. The largest wedge was estimated to belong to the AA-TT dinucleotides. Recent work provided new experimental data on synthetic curved DNA. The authors tried to apply the wedge model to their results and met problems in doing so. We have found that taking into account both roll and tilt components of the AA-TT wedge, in the correct ratio, leads to remarkable consistency between the wedge model and the data.

Curved DNA: design, synthesis, and circularization.

Curved DNA molecules and unusually small circles have been obtained by ligation of synthetic 21-base precursors: (sequence in text). The ligation resulted in the formation of double-stranded oligo-(precursor)s possessing a strong 10.5-base-pair (bp) periodicity of the runs of adenines. Two-dimensional polyacrylamide gel electrophoresis of the ligation products showed two distinct families of spots: (i) noncircular oligo(precursor)s of 21 to 231 bp (1- to 11-mers) and (ii) four circles from 105 to 168 bp (eluted and analyzed by denaturing gel electrophoresis). The noncircular oligomers exhibited anomalously slow migration, as if they were as much as three times longer than they actually are. The amount of circular products peaked sharply at approximately equal to 126 bp, near which size the circles have been estimated to be nonconstrained both torsionally and in terms of bending. The nonconstrained circularization provides a technique for the direct measurement of the inherent curvature of DNA in solution. From the size of the circles, an estimate of 8.7 degrees is obtained for the absolute value of the AA X TT wedge angle (roll and tilt combined).

Superhelicity of nucleosomal DNA changes its double-helical repeat.

Winding DNA in a superhelix can be considered a process consisting of two smooth deformations: bending and twisting. The extra twist angle introduced by winding DNA into the nucleosomal superhelix is calculated by means of the Crick formula to be -0.5 degrees per base pair (bp). This is equivalent to a change of -0.15 +/- 0.015 bp in the DNA double-helical repeat. Free DNA in solution is known to have a helical repeat of 10.55 +/- 0.1 bp. On the other hand, a weighted average of various estimates of the DNA repeat in the nucleosome is 10.38 +/- 0.02. The difference happens to be perfectly accounted for by the superhelicity of the nucleosomal DNA. This implies that the latter is essentially nonconstrained .