UHH > Faculty of Biology > Highlights of Biochemistry > Holliday junction

DNA replication with 100% fidelity is a nice feature to keep offspring in just the genetic background of the species. But to get there, or to evolve further, requires genetical changes, one of which results from recombination of (near) homologous parts of DNA. The nature of structural changes in DNA neccessary to result in homologous genetic recombination were layed out by Holliday in 1964, and in subsequent years the crossover-structures were visualized by electron microscopy. The actual conformation of a DNA crossover was speculated to be a four-way-junction with separate DNA helices, or with stacked helices in either a parallel or an antiparallel orientation of the helices. The models had to allow for branch migration, else no exchange of genetic material would happen.

branch migration models
in parallel aligned helices	in antiparallel aligned helices

During branch migration hydrogen bonds between paired bases have to be broken and others reformed instead. On average the energy for braking and reforming these bonds will cancel each other - but in real existing DNA not all base pairs are created equal. This calls for the action of enzymes to overcome the neccessary activation energy. And enzymes are needed anyway to resolve the four-way-junctions into separate helices. In E. coli. e.g. there exists an enzyme system (RuvABC) the components of which hold the Holliday junction (RuvA), swivel the DNA strands to enable branch migration (RuvB) and finally cut the junction (RuvC). A DNA ligase restores intact double helices.

Homologous genetic recombination is a highly dynamic process, in contrast to X-ray crystallography relaying on static structures. So it took to the end of the previous millenium to get an atomic detail view of relevant structures. You may see here the structure of a four-way Holliday-junction formed by homologous DNA strands, a RuvA-tetramer complexed to a static Holliday-junction, the motor driving branch migration, and a Holliday-junction resolving enzyme.

The Ruv-System of E. coli is in itself a dynamic complex. During branch migration two tetramers of RuvA hover on both sides of a cruciform DNA, with multimeric RuvB clamping two of the DNA strands to wind them. This complex is not accessible for RuvC. In order for the resolvase to act, one of the RuvA tetramers has to be dissociated so that one side of the DNA junction is amenable to strand separation. In vitro the tetramer-octamer-equilibrium is subject to the salt concentration of the buffer. Conditions neccessary for crystallisation of the complex resulted in tetrameric RuvA complexed to the DNA.

Literature:
R Holliday, A mechanism for gene conversion in fungi, Genet. Res. 5 (1964) 282-304
DJM Lilley, Structures of helical junctions in nucleic acids, Quart. Rev. Biophys. 33 (2000) 109-159
PS Ho & BF Eichman, The crystal structures of DNA Holliday junctions, Curr. Opin. Struct. Biol. 11 (2001) 302-308
GJ Sharples, The X philes: structure-specific endonucleases that resolve Holliday junctions, Mol. Microbiol. 39 (2001) 823-834