Gapped DNA and cyclization of short DNA fragments

Abstract

We use the cyclization of small DNA molecules, approximately 200 bp in length, to study conformational properties of DNA fragments with single-stranded gaps. The approach is extremely sensitive to DNA conformational properties and, being complemented by computations, allows a very accurate determination of the fragment's conformational parameters. Sequence-specific nicking endonucleases are used to create the 4-nt-long gap. We determined the bending rigidity of the single-stranded region in the gapped DNA. We found that the gap of 4 nt in length makes all torsional orientations of DNA ends equally probable. Our results also show that the gap has isotropic bending rigidity. This makes it very attractive to use gapped DNA in the cyclization experiments to determine DNA conformational properties, since the gap eliminates oscillations of the cyclization efficiency with the DNA length. As a result, the number of measurements is greatly reduced in the approach, and the analysis of the data is greatly simplified. We have verified our approach on DNA fragments containing well-characterized intrinsic bends caused by A-tracts. The obtained experimental results and theoretical analysis demonstrate that gapped-DNA cyclization is an exceedingly sensitive and accurate approach for the determination of DNA bending.

FIGURE 1

Conformations of a short DNA fragment in DNA-protein complexes. (Top) A short DNA fragment wrapped around a protein. If the bend angle, α, is close to 180°, the cyclization perturbs the DNA conformation in the complex. (Bottom) A gapped-DNA fragment of the same length can be closed without perturbing its conformation and, therefore, provides a more accurate determination of α.

FIGURE 2

Sequences of DNA fragments used in the study. (A) 200-bp-long fragment. The recognition sites for different endonucleases are underlined. The arrows show positions of nicks produced by nicking enzymes. The shadowed region is removed from the fragment resulting in a gap. (B) 42-bp segments with two and four A₅-tracts, which were used to replace the segment of the same length located between XhoI sites in panel A.

FIGURE 3

Analysis, in 8% nondenaturing PAGE, of creating a gap in the 200-bp-long DNA fragment shown in Fig. 2. Lane 1, DNA size marker; lane 2, a plasmid carrying the fragment digested by HindIII; lane 3, the same as in lane 2 but also treated by the nicking enzyme N.BbvC IA; lane 4, the same as in lane 3 but additionally treated by the N.BstNB I nicking enzyme; lane 5, the plasmid treated by the nicking enzymes, then by ligase followed by the HindIII digestion; and lane 6, the same as in lane 5 but with purification from the tetranucleotide before the treatment by ligase.

FIGURE 4

j-factor determination from the ligation time course. (A) Typical result of agarose gel electrophoresis shows separated bands of the linear monomers (LM), the reaction substrate, linear dimers (LD), circular monomers (CM), and circular dimers (CD). (B) PhosphorImager scan of the bands for t = 8′. (C) The ratio 2M₀C(t)/D(t) is extrapolated to zero ligation time to obtain the j-factor value. Both linear and circular dimers were included in D(t). The data are for a 200-bp-long fragment carrying four phased A₅-tracts.

FIGURE 5

Dependence of j-factors on the length of DNA fragments. The j-factor values were measured for intact DNA fragments (○) and for same fragments with a 4-nt-long gap (•). The sequence of the fragments is shown in Fig. 2 except for a few point mutations/deletions that were introduced to change the fragment length. The straight line corresponds to the theoretical j-factor values calculated for the fragments with gaps. The gap bending rigidity, g_gap, was adjusted to yield the best fit between the theoretical curve and experiment (solid line). To illustrate the sensitivity of the theoretical dependence to g_gap, we performed the computations for two more values of g_gap(dotted lines), and where is the best-fit value. The computation assumes that all torsional orientations of the fragment ends are equally probable (see Materials and Methods for details). The experimental data for intact fragments (without gaps) were approximated by the theoretical dependence (Shimada and Yamakawa, 1984) by adjusting the values of three parameters, a, γ, and C. The best fit (solid line) corresponds to a = 48.5 nm, γ = 10.49 bp/(helix turn), and C = 2.4 × 10⁻¹⁹ erg × cm.

FIGURE 6

The j-factor values for 200-bp-long fragments carrying A₅-tracts. The experimental data obtained for gapped-DNA fragments with zero, two, and four phased A₅-tracts (•, ○) are shown together with the theoretical j-factor values. Solid and open circles correspond to two different orientations of the segment with four A₅-tracts in the 200-bp fragment. The theoretical values were calculated for the bend angle of 13° (lower dotted line), 15° (dashed line), and 17° (upper dotted line) per A-tract.

FIGURE 7

Dependence of the j-factor value and the relative error of the bend angle determination on the bend angle. The data are calculated for a 200-bp-long fragment carrying a 4-nt-long gap on the basis of the theoretical model with the value of the gap rigidity constant determined from data in Fig. 5. (A) Theoretical dependence of the j-factor on the bend angle. (B) The relative error of the angle determination that corresponds to 15% error in the j-factor measurements.