J. Mol. Biol. (1992) 228, 322-326

An Accurate Method for Determining the Helical Repeat of DNA in Solution Reveals Differences to the Crystal Structures of Two B-DNA Decamers Michael Niederweis, Thomas Lederer and Wolfgang Hillen? Lehrstuhl

fiir

Mikrobiologie, Institut fiir Mikrobiologie Universittit Erlangen-Niirnberg, Staudtstr. (Received 27 April

und Biochemie der Friedrich-Alexander 5, 8520 Erlangen, Germany

1992; accepted 5 August

1992)

Many DNA sequences have been studied by X-ray crystallography with the goal of deciphering a sequence-structure code. We have determined the helical repeats of two B-type DNA decamers in solution employing an electrophoretic method based on phasing of bent segments. The decamers contain recognition sites for the dcm methyltransferase and for the restriction nuclease NarI with a mutational hotspot. Their helical repeats are 1059( + 0.05) bp and 10*52( + @03) bp, respectively, whereas crystallographic analysis yielded 100 bp in the solid state. This difference is greater than that for the transition Thus, reliable information about, the between B- and A-type DNA in solution. polymorphism of DNA in solution must be based on both X-ray and solution data. We describe a generally applicable approach to accurately determine helical repeats of small DNA duplexes in solution.

Keywords: DNA;

solution

structure;

helical

repeat’; gel electrophoresis:

bent DNA

result is in contrast to the structures of two A-type DNA octamers in differently packed crystals (Jain & Sundaralingam, 1989; Shakked et al.: 1989) and has been taken as an indication that the conformat’ion of this B-type DNA decamer is mainly determined by its sequence and would not> depend on packing forces (Heinemann & Alings. 1991). d(CCGGCGCCGG) contains t]hr recognition sequence for the restriction enzyme &‘a,rl (underlined), which is also a hot spot for chemically induced -2 frameshift mutagenesis (Timsit et al.. 1989). Therefore, the determination of its structure has received much attention. The octamer d(GGGCGCCC) has been crystallized as A-form DNA with a helical repeat of 11.4 bp (Rabinovich et al.. 1988). The dodecamer d(ACCGGCGCCACA) (Timsit et al., 1989) and the decamer d(CCGGCGC’CGG) (Heinemann et al.. 1992) have been crystallized as B-type DNA with helical repeats of 10.3 bp and l@O bp, respectively. These length variations of t,hr oligonucleotides embedding this sequence resulted in remarkably different st’ructures in the solid state. In contrast to the crystal struct’ures all helical repeats det’ermined so far for a number of various DNA sequences in solution vary between 10.5 bp and 10.6 bp, except for poly dA. dT, which has 10.0 bp (Peck & Wang, 1981; Rhodes & Klug. 1981). This apparent discrepancy was extensively discussed (Rhodes & Klug, 1980). To examine

Local DNA structure is considered to play an important role for sequence specific recognition by proteins (Travers, 1989; Harrison, 1991; Yanagi et al., 1991) and drugs (Yanagi et al., 1991). Therefore. many DNA sequences have been studied by X-ray crystallography with the goal of deciphering a sequence-structure code (Kennard & Hunter, 1989; Yanagi et al., 1991). We have focused our study of potential DNA structure-function relationships on two sequences with interesting biological properties, which show helical repeats in the solid state differing from the 10.6 bp$ commonly found for various DNA sequences in solution (Rhodes & Klug, 1980, 1981; Peck & Wang, 1981; Tullius & Dombrowski, 1985). The decamer d(CCAGGCCTGG) contains two recognition sites for the Escherichia coli dcm methyltransferase and its nonmethylated and hemi-methylated forms have been crystallized as B-type DNA in two different packing arrangements (Heinemann & Alings, 1989, 1991). Thus, for the first time, DNA conformations in the biologically predominating B-form could be compared in different crystal environments and turned out to be remarkably similar with helical repeats of l@O bp (Heinemann & Alings, 1991). This I Author to whom all correspondence should be addressed. $ Abbreviations used: bp. base pair(s). 322 0022%2836/92/22032245

$08.00/O

I(:) 1992 Acadt=mic

Press Limited

323

Communications

r---,--------T AGTGGTAA

GGATCCTC;TilGA(STCGAC

01

1H

2w 3H 4-

1 bp distance variations. The organization of the constructed distance variants is depicted in Figure 1. These fragments with differently phased bends show a sinusoidal mobility variation in polyacrylamide gels (Fig. 2). The reduced mobility of bent DNA fragments in polyacrylamide gels has been correlated to their shortened end-to-end distances (Crothers et al., 1990). Thus, the maximal mobility for the distance variant 0 indicates the trans configuration of both bent segments and the minimal mobility for the distance variant 6 the cis configuration.

Figure 1. Physical structures of 3 isomeric sets of DNA fragments with varied phasing of 2 bent segments. The upper portion shows the DNA fragment used to construct systematically phased bends. The heavy lines represent the DNA, the black boxes indicate the 2 intrinsically bent DNA segments, each of which contains 4 dA, alternating with 3 dA, tracts properly phased to give a macroscopic curvature of about 140”. The phased dA tracts originate from plasmid pK5/6T217 (Diekmann, 1987). The distance between the curved segments is varied over 9 bp at, the site indicated by the bidirectional arrow by base-pair-wise insertion of the sequence given in bold print below. The length of the lines correlates the inserted sequence to the distance designations (0 to 8). The 3 sets of distance variants 10 to 18 were constructed by insertion of 1 of the decamers given on the left side as a synthetic BamHI-Sal1 fragment in each of the distance variants 0 the original sequences unchanged. to 8, leaving Eon-methylated sequences of d(CCAGGCCTGG) were prepared from a dcm strain of E. coli. The methylation status was verified by restriction with StuI, which is blocked by overlapping dcm methylation.

whether the two decamers have indeed a 10.0 bp repeat in solution, we developed a gel electrophoresis assay based on an approach first published by Zinkel & &others (1987). A similar analysis was used to determine an averaged periodicity of RNA (Bhattacharyya et al., 1990; Tang & Draper, 1990), which could, however, not be assigned to a defined sequence. We have used intrinsically bent DNA designed for maximal curvature separated by

Insertion

of the decamers

in this set of DNA

fragments results in a displacement of the extrema (Fig. 2(a) and (b)). Provided that the inserted DNA is not bent, this would be a direct measure of its helical repeat. The electrophoretic mobilities of each set of distance variants have been fitted to third order polynomials. Inspection of the results in Figure 2(a) and (b) shows that the corresponding fits for each set of decamer insertions have identical shapes in their overlapping parts when compared to the DNAs without insertion. Considering the high sensitivity of this assay for small intrinsic curvature (Zinkel & Crothers, 1987), this demonstrates that the inserted decamers are not bent. This confirms structures in which the respective crystal d(CCAGGCCTGG) (Heinemann & Alings, 1989, 1991) and d(CCGGCGCCGG) (Heinemann et al., 1992) are also straight B-type DNA. Furthermore, it allows the precise determination of the helical repeats of the decamers from the displacements of minimal mobilities. As it has been shown that the mobilities of DNA fragments with phased bends can depend on the gel matrix (Drak & Crothers, 1991), we have measured the displacements of minimal mobilities for polyacrylamide concentrations between 6% and 12 %. The results are shown in Table 1 and demonstrate that the helical repeats of both decamers in solution can be very accurately determined by this method and lie between 10.5 bp and 1@6 bp independently of the polyacrylamide matrix. This differs from the repeats adopted by the decamers in the crystal structures. The fully methylated decamer d(CCAGGCCTGG) yielded

Table 1 Helical repeats of three decamers determined in polyacrylamide different concentrations

gels of

Helical repeat (bp) Acrylamide concentration

(O/c)

d(AAAAAAAAAA)

d(CCAGGCCTGG)

d(CCGGCGCCGG)

1060_+017 1@60~0~07

1047_+0.06 1055*0.07

10 12

10.55 * 0.05

1@47+0.06

10.60+_010

1057+_0.06

993 5 0.06 9.92 * 008 938 +_@05

Mean

10.59 + @05

10.52 k 0.03

9.90 f 0.04

6 8

988_+0.15

The migration distances were fitted by a spline function. The average values from at least 3 independent experiments with their standard deviations are given for each polyacrylamide concentration. The bisacrylamide concentration was 2%. 8% gels with other acrylamide to bisacrylamide ratios (39:1, 29:1, 19:l) gave the same helical repeats for d(CCAGGCCTGG). The means of the results at different polyacrylamide concentrations are given in the bottom line with their average errors determined by propagation of errors.

M. Niederweis

C

+ d(CCAWCM)

A

0 1234 56 76 \

5

;

0

n

I)

m

0 1 2 3 4 5 6 7 6

m

101112131416161716

II)

n

._-

* r: Z

+ W,,)

101112l3l46l617l6

- 0-L --

et al.

-0”

4

I

I\\ I

15

5 10 distance [bp]

0

0

0 1 2 3 4 5 6 7 6

1

,

0

1

I

I

I

,‘I

I

I

5 distance

101112131415161716

,

(

I

”

I

”

”

5 distance

+ d(CC6QCtNCGf3)

B

”

I

10

I

I

,

1’1

I

I

15 [ bp]

results identical to the non-methylated form (data not shown). Thus, methylation at the two dcm sites neither changes the twist nor introduces a bend in this DNA in solution. This is consistent with the results obtained from the crystal structures (Heinemann & Alings, 1989, 1991). As both decamers examined with this method in

I

““I”’

10

15 [bp]

Figure 2. Gel mobility analyses of the sets of distance variants with (a) the dcm, (b) the NarI and (c) the dA,, . dT,, sequence. For each of the 3 decamers analyzed in this study a representative 8% (w/v) polyacrylamide gel is shown in the upper part, of the Figure. The numbering of lanes corresponds to the distance designation as defined in Fig. 1. The sequences of the inserted decamers are given above the respective gel pictures. Gels with polyacrylamide concentrations varying from 6% to 12% (49:l acrylamide to bisacrylamide) were prerun in 89 mM-Tris-borate (pH 8.3), 2 mix-EDTA at 20°C unt,il the current was constant. The DNA fragments were prepared by digestion with EcoRI, purified, endlabeled and electrophoresed at 20°C to an average migration distance of about 35 cm, while the electrode buffers were circulated to prevent gel anomalies. The sketches on the left sides of the autoradiographs indicate the cis (upper) and bans (lower) configurations. The analysis of the autoradiographs is illustrated in the lower half of the Figures. Filled circles indicate the distance variants 0 to 8 and filled triangles the distance variants 10 to 18. The average migration distance of each set has been defined to be 0. The migration differences were then normalized to 100 cm. Thus, the results of all gel percentages were comparable. The migration distances were fitted to a third-order polynomial and plotted against the relative distance in bp. The vertical broken lines indicate the minimal mobilities that were calculated from the 1st derivatives of the polynomials. The results for the inserted sequences are averages from at least 5 independent runs of 8% polyacrylamide gels and yield helical repeats of 1967 bp, 1062 bp and 991 bp for (a) the dcm. (b) the NarI and (c) the dA,,.dTio sequence, respeatively. Small errors of these determinations arise from the choice of the polynomials and lie in the range of those given in Table 1.

325

Communications solution had a helical repeat of around I@6 bp, we asked whether helical repeats differing from 10.6 bp can be determined with our method. We therefore inserted the dA 10 *dT,, duplex in the set of distance variants as is also shown in Figure 1. Poly dA *dT is the only sequence for which determination of the helical repeat in solution and in fibres yielded the same value of 10.0 bp (Peck & Wang, 1981; Rhodes & Klug, 1981; Alexeev et al., 1987). The results depicted in Figure 2(c) show the smaller helical repeat, of 9*9(IfI@O4) bp for dA,,*dTlo. Thus, the method employed here can detect smaller helical repeat’s. We conclude that the other two decamers indeed have different helical repeats in the solid state and in solution and that crystallization imposes a torsional constraint on DNA that is not present in solution. As d(CCAGGCCTGG) in the non-methylated and methylated forms exhibits the same condifferent crystal lattices formation in two (Heinemann & Alings, 1989, 1991), another factor than packing forces may be responsible for the deviation of helical repeats in that case. We have also employed the topoisomer method as the only other method available to measure helical repeats of short DNA sequences in solution (Wang, 1979; Peck & Wang, 1981). The mobility differences between pairs of topoisomers with and without the inserted decamers were very small and did not exceed the error range (data not shown). Heinemann & Alings (1991) have reported the of failure to crystallize sequence variants d(CCAGGCCTGG) and suggested that sequencespecific deviations from an average twist of 36”, which would facilitate end-to-end stacking of the decamers in the crystal, might prevent their formation. Our results raise the hypothesis that only sequences for which the twist can be deformed to 36” may crystallize as B-type DNA. This could be also true for the dodecamer crystals of the “Dickerson-Drew-family”, which form columns of helices with two-base-pair overlapping ends but also have an average twist of 36” in the solid state (Yanagi et al., 1991). In the light of this hypothesis the surprisingly limited number of single crystal B-type DNA structures (Heinemann & Alings, 1989; Kennard & Hunter, 1989; Yanagi et aE., 1991) might be due to the fact, that not many sequences can be deformed to a helical repeat near ten base-pairs. If this holds true, the crystal structures of DNA provide information about the potential of the examined sequence to be deformed to the solid state structure rather than information about the dominating structure in solution. It has recently been pointed out by Schultz et al. (1991) that even the structures of DNA sequences complexed with their binding proteins in co-crystals may be influenced by packing forces. The only B-type DNA sequence with deviations from the helical repeat of l@O bp in crystals is the NarI recognition sequence in the dodeca,mer. Its crystal exhibits a very unusual helical packing of the DNA, which may be the reason for the reduced repeat (Timsit et al., 1989).

Furthermore, when crystallized as an octamer this sequence assumes the A-form. This could suggest an enhanced conformational flexibility of this sequence or could just be a matter of chance as it is difficult to rationalize what sequence specific differences may lead to the A-form in the octamer and the B-form in the decamer. However, this sequence also exhibits the common helical repeat near 1@6 bp in solution. The atomic resolution of crystallographic studies allows the elucidation of local structural variation within DNA. The results presented here suggest that twist, which determines the helical repeat of DNA, and its correlated parameters rise, cup and roll (Yanagi et al., 1991) cannot be simply added up to yield the average solution structure. The average twist values of B-type DNA in the solid and solution states differ by about 2” per bp, which is more than the difference of 1.5” per bp determined for the transition of B- to A-type DNA (Krylov et al., 1990). Thermal fluctuations of the twist angles of DNA occur in the same range indicating the flexibility of DNA. Crystallization. may select for an conformation individual out of this range. Therefore, the use of twist angles from crystal structures to elucidate a sequence-structure code (Yanagi et aE., 1991) is hampered twofold: (1) the twist difference of the solution and crystal structures may not be evenly distributed over the entire length of the oligonucleotide; and (2) crystallographic data for base pairs in a sequence context, which is unable to undergo the structural change required in the solid state, may never be obtained. One of the decamers has been crystallized in two different space groups. It may therefore be suggested that even this elaborate procedure (Privk et al.. 1991) does not lead to the solution structure. Thus, the discussion of polymorphism of DNA should include structural parameters determined in solution. The method used in this study provides a sensitive approach to compare DNA structures in crystals and in solution. We thank Dr U. Heinemann for fruitful discussions and for providing results prior to publication, C. Berens and Dr C. Gatz for critically reading the manuscript and F. Schirmer for constructing a part of the plasmids. This work was supported by the Deutsche Forschungsgemeinschaft and the Fonds der chemischen Industrie.

References Alexeev, D. G., Lipanov, A. A. & Skuratovskii, I. Y. (1987). Poly(dA).poly(dT) is a B-type double helix with a distinctively narrow minor groove. Nature (London), 325, 821-823. Bhattacharyya, A., Murchie, A. I. H. & Lilley, D. M. J. (1990). RNA bulges and the helical periodicity of double-stranded RNA. Nature (London), 343. 484487. Crothers, D. M., Haran, T. E. & Nadeau, J. G. (1990). Intrinsically bent DNA. J. Biol. Chem. 265, 7093-7096. Diekmann, S. (1987). Temperature and salt dependence of the gel migration anomaly of curved DNA fragments. Nucl. Acids Res. 15, 247-265.

M. Niederweis et al.

326

Drak, J. & Crothers, D. M. (1991). Helical repeat and chirality effects on DNA gel electrophoretic mobility. Proc. Nat. Acad. Sci., U.S.A.

88, 3074-3078.

Harrison, 8. C. (1991). A structural taxonomy of DNA-binding domains. Nature (London), 353, 715719. Heinemann, U. & Alings, C. (1989). Crystallographic study of one turn of G/C-rich B-DNA. J. MoZ. Biol. 210, 369381. Heinemann, U. & Alings, C. (1991). The conformation of a B-DNA decamer is mainly determined by its sequence and not by crystal environment. EMBO J. 10, 3543. Heinemann, U., Alings, C. & Bansal, M. (1992). Double helix conformation, groove dimensions and ligand binding potential of a G/C stretch in B-DNA. EMBO J. 11, 1931-1939. Jain, S. & Sundaralingam, M. (1989). Effect of crystal packing environment on conformation of the DNA duplex. J. Biol. Chem. 264, 12780-12784. Kennard, 0. $ Hunter, W. N. (1989). Oligonucleotide structure: a decade of results from single crystal X-ray diffraction studies. Quart. Rev. Biophys. 22, 327-379.

Krylov, D. Y.. Makarov, V. L. & Ivanov, V. I. (1990). The B-A transition in superhelical DNA. Nucl. Acids Res. 18, 75S761.

Peck, L. J. & Wang, J. C. (1981). Sequence dependence of the helical repeat of DNA in solution. Nature (London),

292, 375-378.

Prive, G. G., Yanagi, K. & Dickerson. R. E. (1991). Structure of the B-DNA decamer CCAACGTTGG comparison with isomorphous decamers and CCAAGATTGG and CCAGGCCTGG. J. Mol. Biol. 217, 177-199. Rabinovich, D.. Haran, T., Eisenstein, M. & Shakked, Z.

(1988). Structures of the mismatched duplex d(GGGTGCCC) and one of its Watson-Crick analogues d(GGGCGCCC). J. Mol. Biol. 200. 151-161. Rhodes, D. & Klug, A. (1980). Helical periodicity of DNA determined by enzyme digestion. Nature (London). 286, 573-578.

Rhodes. D. & Klug, A. (1981). Sequence-dependent helical periodicity of DNA. Nature (London), 292, 378-380. Schultz, S. C., Shields, G. C. & Steitz, T. A. (1991). Crystal structure of a CAP-DNA complex--The DNA is bent by 96 degrees. Science, 253, 1001-1007. Shakked, Z., Guerstein-Guzikevich, G., Eisenstein. M.. Frolow, F. & Rabinovich, D. (1989). The conformation of the DNA double helix in the cryst,al is depenNature (London), 342. dent on its environment. 456-460.

Tang, R. S. & Draper, 1). E. (1999). Bulge loops used to measure the helical twist of RNA in solmion. Biochemistry,

29, 5232-5237.

Timsit, Y., Westhof, E., Fuchs, R. P. I’. c\t Moras, 1). (1989). Unusual packing in crystals of DNA bearing a mutation hot spot. Nature (London). 341. 459-462. Travers. A. A. (1989). DNA conformation and protein binding. Annu. Reo. Biochem. 58, 4277452. Tullius, T. D. & Dombrowsky, B. A. (1985). Iron(H) EDTA used to measure the helical twist, along an) DNA molecule. Science, 230. 679681. Wang, J. C. (1979). Helical repeat of DNPr’Bin solution. Proc. Nat. Acad. &i., U.S.A. 76, 200-203. Yanagi? K.. PrivC. G. G. & Dickerson, R. E. (1991). Analysis of local helix geometry in three B-DNA decamers and eight dodecamers. J. Mol. Biol. 217. 201-214.

Zinkel. S. S. & Crothers, D. M. (1987). DNA bend dire?tion by phase sensitive detection. Nature (London). 328, 178-181.