[43 Detection of DNA Base Mismatches Using DNA Intercalators

By Elizabeth M. Boon, Jennifer L. Kisko, and Jacqueline K. Barton Introduction

The most prevalent mechanisms leading to mutations in DNA are direct mis-incorporation of bases during replication and sustained chemical damage. Under normal circumstances, the cell corrects these problems using DNA polymerase proofreading mechanisms as well as the complex repair machinery of the cell. In certain tissues that contain mismatch repair deficiencies, DNA mispairs may accumulate.1"5 Even in healthy cells, however, mismatches and lesions can sometimes go unchecked, resulting in permanent alterations in the gene sequence for subsequent generations. Identification of genetic variations (single-nucleotide polymorphisms, SNPs) among individuals and populations has implications in understanding human disease and treatment, as well as the interaction of the environment and multiple genes during evolution.6 Once these SNPs are identified and understood, rapid and reliable detection of them will be critical for the study, diagnosis, and treatment of genetically linked disease.

SNPs are detected as mismatches in heteroduplexes formed from a known copy of a gene and the test gene. Several procedures have been described to search duplex DNA for mismatches using chemical, enzymatic, and differential hybridization techniques.7"15 None of these, however, are as selective, inexpensive,

4 R. D. Kolodner, Trends Biochem. Sci. 20, 397 (1995).

5 B. A. Jackson and J. K. Barton, "Current Protocols in Nucleic Acid Chemistry," p. 6.2.1. John Wiley

7 D. H. Geschwind, R. Rhee, and S. F. Nelson, Genet. Anal. 13,105 (1996).

8 J.-F. Xu, Q.-P. Yang, J. Y.-J. Chen, M. R. van Baalen, and I.-C. Hsu, Carcinogenesis 17,321 (1996).

9 K. Yamana, Y. Ohashi, K. Nunota, and H. Nakano, Tetrahedron 153,4265 (1997).

10 N. C. Nelson, P. Hammond, E. Matsuda, A. A. Goud, and M. Becker, Nucleic Acids Res. 24, 4998 (1996).

11 A. Ganguly and D. J. Prockop, Electrophoresis 16, 1830 (1995).

12 R. G. H. Cotton, N. R. Rodrigues, andR. D. Campbell, Proc. Natl. Acad. Sci. U.S.A. 85,4397 (1998).

13 R. G. H. Cotton, H.-H. M. Dahl, S. Forrest, D. W. Howells, S. J. Ramus, R. E. Bishop, I. Dianzani, J. A. Saleeba, E. Palombo, M. J. Anderson, C. M. Milner, and R. D. Campbell, DNA Cell Biol. 12, 945 (1993).

14 D. H. Johnston, K. C. Glasgow, and H. H. Thorp, J. Am. Chem. Soc. 117, 8933 (1995).

15 G. Rowley, S. Saad, F. Giannelli, and P. M. Green, Genomics 30, 574 (1995).

and reliable as would be necessary for widespread application. Consequently, new strategies for mismatch detection are needed.

In this review, we describe approaches for mismatch discovery and diagnosis using intercalative DNA-binding molecules. Our approach to mismatch discovery utilizes a metal complex that binds by intercalation selectively to mismatches in duplex DNA due to the local thermodynamic and kinetic destabilization associated with these sites. On irradiation, this metal complex cleaves the DNA at its site of intercalation, the mismatched site. In this approach, the gene sequence does not have to be known, and we can search for mismatches on a genome-wide basis. During mutation diagnosis, we look for mismatches in short heteroduplexes formed from a known sequence and test oligonucleotides. In this assay, we take advantage of a distinctly different characteristic associated with mismatches, the perturbation of the electronic structure of DNA at the mismatched site. Figure 1 schematically illustrates mismatch discovery and diagnosis using these strategies.

mismatch mismatch mismatch

mismatch strand break mismatch mismatch

genomic processing, hybridization to DNA surface, electrochemistry strand break

Fig. 1. Schematic representation of our approaches to mismatch discovery by [Rh(bpy)2(chry si)]3+ photochemistry and mismatch diagnosis by electrochemistry. Both assays feature DNA intercalators as probe molecules.

Experimental Methods for Mismatch Discovery by a Designed Synthetic Intercalative Complex: Detection Based on Thermodynamic and Kinetic Destabilization

Metallointercalators of the type [RhL2(phi)]3+ have been shown to be effective in binding DNA via insertion of the 9,10-phenanthrene quinone diimine (phi) intercalator between base pairs in the DNA duplex.1617 For example, the complex A-a-[Rh(/?,fl)-Me2trien](phi)]3+ [Me2trien = (2/?,9/i)-diamino-4,7-diazadecane] has been used to target the sequence 5'-TGCA-3', where functionalization of the ancillary ligands with methyl groups leads to specific van der Waals interactions.18 High-resolution nuclear magnetic resonance (NMR) studies19'20 and a crystal structure21 of the metal complex bound to its target site have confirmed such sequence-specific interactions. The crystal structure, in which A-a-[Rh(i?,/?)-Me2trien](phi)]3+ is bound to a DNA octamer, shows the phi ligand to be deeply inserted and base stacked within the duplex from the major groove side.

There is widespread interest in site-specific recognition of DNA base pair mismatches. Strategies have exploited mismatch recognition proteins, differential chemical cleavage, hybridization of fluorescent conjugates, and DNA chip methodologies.7-15 NMR studies have shown that, when a phi ligand intercalates between two base pairs, the phenanthrene ring stacks snugly into the base step and there is little unoccupied space left over (Fig. 2).19'20 This suggested a mechanism for recognition in which a sterically demanding intercalating ligand, too large to fit readily into standard B-DNA, might recognize destabilized regions of the helical core. Our laboratory has reported the construction of a novel mismatch recognition agent, bis(2,2'-bipyridyl) (5,6-chrysenequinone diimine)rhodium(III), [Rh(bpy)2(chrysi)]3+, which has the differentiating characteristic of a four-, rather than three-, ringed intercalating ligand, chrysi.22-25 The sterically bulky chrysi intercalating ligand is too wide to intercalate readily into B-form DNA, and therefore binding is restricted to destabilized regions at or near base pair mismatches. Thus, the complex is able to specifically target the duplex at or near mismatch sites and, on photoactivation, cleave the DNA backbone. Therefore, rather than

16 K. E. Erkkila, D. T. Odom, and J. K. Barton, Chem. Rev. 99, Till (1999).

17 T. Johann and J. K. Barton, Phil. Trans. R. Soc. Lond. A 354, 299 (1996).

18 A. H. Krotz, B. P. Hudson, and J. K. Barton, J. Am. Chem. Soc. 115, 12577 (1993).

19 B. P. Hudson and J. K. Barton, J. Am. Chem. Soc. 120, 6877 (1998).

20 B. P. Hudson, C. M. Dupureur, and J. K. Barton, J. Am. Chem. Soc. 117, 9379 (1995).

21 C. L. Kielkopf, K. E. Erkkila, B. P. Hudson, J. K. Barton, and D. C. Rees, Nat. Struct. Biol. 7, 117 (2000).

22 H. Murner, B. A. Jackson, and J. K. Barton, Inorg. Chem. 37, 3007 (1998).

23 B. A. Jackson and J. K. Barton, J. Am. Chem. Soc. 119, 12986 (1997).

24 B. A. Jackson, V. Y. Alekseyev, and J. K. Barton, Biochemistry 38,4655 (1999).

25 B. A. Jackson and J. K. Barton, Biochemistry 39, 6176 (2000).

Three-ringed phi intercalator fits snugly in B-form DNA

Four-ringed chrysi ligand is too wide to fit in B-form DNA, but can fit into mismatches

Fig. 2. Comparison of steric width of tricyclic phi and tetracyclic chrysene ligands reveals that the chrysene ligand would demand an additional 2 A of space on intercalation into DNA. As shown in the space-filling model above the phi ligand, just enough room is afforded in normal DNA to allow phi intercalation. Thus, intercalation of the chrysene ligand preferentially targets DNA sites that are in some way distorted, or more easily distorted than normally base-paired sequences; simply put, the chrysene ligand is too sterically demanding for intercalation into normal DNA.

selecting for exposed functionalities of bases as with previously described metallo-intercalators, specificity for mismatch sites is enabled by selecting for the thermodynamic constraints associated with the mispair. [Rh(bpy)2chrysi]3+ was found to induce strand scission at approximately 80% of all mismatch sites in all possible surrounding sequence contexts and recognition correlated generally with the local helical destabilization associated with the mismatch.25

As A-[Rh(bpy)2(chrysi)]3+ has shown that a comparably sterically bulky intercalating ligand can enable binding specificity for mismatches on small, synthetic oligonucleotides, the next challenge was to test its ability to specifically recognize mispairs on long DNA polymers of biological origin.24 Thus, an experiment using a mixture of two plasmids was performed; the plasmids differed in sequence at a single base position. Each plasmid was cut with Seal to produce linear 2725-mers with the C/G variable site 975 base pairs from one end (Fig. 3). Equal amounts of the C- and G-containing plasmids were combined, resulting in a

0 0

Post a comment