No abstract

The crystal structure of the synthetic DNA dodecamer d (CpGpCpGpApApTpTpCpGpCpG) has been refined to a residual error ofR = 17.8% at 1.9-A resolution (two-o'data In the 28 years since a double helix model for B-DNA was proposed by Watson and Crick (2), direct evidence for its structure has been based on refinement of models having standard bond parameters against x-ray diffraction data from oriented fibers (3-6). This has had two disadvantages: loss of information because of rotational disorder about the fiber axis, and lack of information about the effect ofspecific base sequences, aside from a limited number of experiments on homopolymers and alternating copolymers. Recent advances in triester methods of DNA synthesis have made possible the preparation of molecules ofpredetermined base sequence, in quantities and purities suitable for single-crystal x-ray analysis. This paper presents the results ofa structure analysis and refinement ofa complete turn of right-handed B-DNA, with the sequence d(CpGpCpGpApApTpTpCpGpCpG). A preliminary report of the partially refined structure has appeared (7).Our structure analysis of the dodecamer C-G-C-G-A-A-T-T-C-G-C-G is a logical extension of the earlier analyses of C-G-C-G-C-G (8) and C-G-C-G (9). Both of the latter adopted a lefthanded zigzag or Z helix, a totally unexpected conformation whose relevance to biological DNA of more varied sequence became particularly interesting. In December 1979 we were fortunate in growing large single crystals of dodecamer, synthesized at Pasadena and Duarte. This sequence is ofparticular significance because it contains an EcoRI restriction site, G-A-A-T-T-C, and because it brackets a Z-incompatible A-A-T-T segment with two Z-compatible C-G-C-G ends, offering a test of the tendency of mixed-sequence DNA to adopt the Z conformation. In spite of favorable salt conditions, the central A-A-T-T segment apparently is sufficient to counteract the Z-forming tendencies ofthe tetrameric C-G-C-G ends, resulting in a classical Watson-Crick B helix throughout.Structure analysis and refinementThe dodecamer crystallizes in space group P212121 with a = 24.87 A, b = 40.39 A, c = 66.20 A, and two single strands or one double helix per asymmetric unit. Two isomorphous heavy atom derivatives were obtained by de novo triester synthesis with 5-bromodeoxycytidine at the third position along each strand and by diffusion of the anticancer agent cisplatin [cisdichlorodiamminoplatinum(II)] into pregrown crystals. Phases from this analysis were used to obtain a starting model for restrained least-squares refinement (10).Initial energy parameters for DNA as obtained from Michael Levitt were modified so that conformational energy would restrain the molecule to a sterically acceptable structure but would allow the x-ray data to determine this structure. Internal sugar ring bond angles were set to the average oftheir C2'-endo and C3'-endo values but were left flexible enough that other conformations could be obtained (11). The residual error or R factor for 2...

DNA is probably the most discussed and least observed of all biological macromolecules. Although its role in biology is a central one, with many examples such as operators and restriction sites where specific base sequences have control functions or interact with specific enzymes, the structures that DNA can adopt have been based until now only on sequence-averaged fibre diffraction patterns. Recent improvements in triester synthesis methods have made possible the preparation of sufficient homogeneous DNA of predetermined sequence for crystallization and X-ray structure analysis. We report here the first single-crystal structure analysis of more than a complete turn of right-handed B-DNA, with the self-complementary dodecamer sequence d(CpGpCpGPApApTpTpCpGpCpG) or CGCGAATTCGCG.

No abstract

No abstract

No abstract

No abstract

Recent advances in DNA synthesis methods have made it possible to carry out single-crystal x-ray analyses of double-stranded DNA molecules of predetermined sequence, with 4 to 12 base pairs. At least one example has been examined from each of the three known families of DNA helix: A, B, and Z. Each family has its own intrinsic restrictions on chain folding and structure. The observed solvent positions in these crystal structures have confirmed earlier fiber and solution measurements, and have led to proposals explaining the transitions from B to A and from B to Z helices. Prospects are improving for an understanding of the mode of bending of DNA in chromatin, and the way in which specific DNA sequences are recognized by drug molecules and repressor proteins.