The Holliday Model of Genetic Recombination
This model of recombination was first proposed by Robin Holliday in 1964 and re-established by David Dressler and Huntington Potter in 1976 who demonstrated that the proposed physical intermediates existed.
The basic (simple) model
Align two homologous DNA molecules.
Nick the DNA at the same place on the two molecules.
This must happen in strands with the same polarity.
Exchange strands and ligate.
The intermediate that is formed is called a Holliday intermediate or Holliday structure. The shape of this intermediate in vivo is similar to that of the greek letter chi, hence this is also called a chi form.
Visualization of the next step is made easier if one molecule is now rotated through 180š with respect to the other. This also helps to emphasize the chi-shape of the intermediate:
Resolve the structure.
There are two ways in which this can happen:
- If the same strands are cleaved a second time then the original two DNA molecules are generated:
- If the other strands are cleaved, then recombinant molecules are generated:
A more realistic model
The above model is too simple and does not explain a number of genetic results, including the occurence of two different recombinant bacteriophage in a single plaque in the Meselson-Weigle experiment.
These can be explained by modifying the model slightly. As before, two homologous DNA molecules must be aligned and nicked at the same place. Following strand exchange the intermediate Holliday structure is formed. At this stage a new step is introduced:
Branch migration.
Migration of the branch can occur over many nucleotides in either direction. The result is a physical transfer of part of one of the strands of one molecule with that of the other:
Once again, visualization of the next step is made easier if one molecule is now rotated through 180š with respect to the other.
Resolve the structure.
There are still two ways in which this can happen, however, the consequences are different:
- If the same strands are cleaved a second time then nonrecombinant DNA molecules are generated but they each contain a region of heteroduplex DNA that spans the region of branch migration:
- If the other strands are cleaved, then recombinant molecules are generated as before, however, each will also contain a region of heteroduplex DNA that spans the region of branch migration:
 |
View an Animated Model of a Holliday Junction |  |
View a 3D Model of a Holliday Junction |
| These models were prepared by Dr. Bill Engels at the University of Wisconsin | |||
Potter & Dressler's evidence for the Holliday Model
In 1976, David Dressler and Hunt Potter published the results of a series of experiments that demonstrated the validity of the Holliday model of recombination.
They used E. coli cells containing the colicin E1 derived plasmid, pMB9. This plasmid was one of the very earliest plasmids developed for cloning in Herbert Boyer's laboratory. Normally, E. coli contain about 20 copies of this plasmid per cell. However, if the cells are exposed to chloramphenicol then, although chromosomal replication stops, plasmid replication does not and the number of plasmid molecules increases to 1000 copies per cell. With so many more copies of the plasmid in the cell, the chances of recombination increase as does the probability of observing a recombination intermediate.
When plasmid was isolated from the cells, purified by CsCl gradient centrifugation, and observed in the electron microscope, a number of candidates for intermediates were observed. These all had the appearance of "figure 8" structures. However, there are 3 possible ways such structures might arise:
- as a double-sized circular plasmid twisted over on itself.
- as two interlocking circular plasmid molecules.
- as a genuine recombination intermediate.
In order to distinguish between the three possibilities, Potter and Dressler digested their plasmid preparations with EcoRI. This enzyme will generate monomer sized linear molecules from either of the first two possible structures. However, it will generate unique chi-shaped structures from the third.
When they did this, Potter and Dressler found that between 0.5% and 3% of the molecules they observed were chi-shaped structures. The molecules were symmetrical in that the opposite arms were identical lengths and had identical denaturation patterns. Finally, they saw no such structures if they prepared their plasmids from recA- strains of E. coli.
From this evidence they concluded:
. . . the intermediates we have observed in the electron microscope provide physical evidence in support of the recombination intermediate postulated by Holliday on genetic grounds. [S32-4]
Other Models of Genetic Recombination
Although, the basic features of the Holliday model are well-established, the model does have flaws. For example, the mechanism by which two homologous regions of DNA are paired and then nicked is not well explained. In addition, the model does not explain all of the observed results in different recombination systems.
A central feature of the Holliday model as outlined above is that the heteroduplex regions in recombinant molecules will be identical in length and position. Experimental results suggest, however, that this is not necessarily so.
In order to explain such discrepancies, Matt Meselson and Charles Radding proposed a modified model. In their model, only one strand in one of the two paired homologous DNA molecules is nicked. This strand is then displaced by DNA polymerase I and invades the homologous chromosome. A Holliday intermediate is eventually formed and resolved as above. In this model, the extent of heteroduplex DNA in the resolved products, whether recombinant or nonrecombinant, will be different.
[MVH25-22]
Finally, gene conversion (which is a type of recombination event) in yeast follows a mechanism, proposed by Szostak, that requires double-strand breaks in one of the recombining molecules. See also the Meselson-Radding Model - also at the University of Arizona, ALL you could want to know about General Homologous Recombination at the University of Saskatchewan, and Recombination: A molecular perspective at the University of Illinois.
|
RESOURCE MATERIAL |
|
| VOET, VOET & PRATT |
|
| STRYER |
|
| LEHNINGER |
|
| TAMARIN |
|
| WEB SITES |
|
|
|
