The structure and conformational dynamics of a complex molecular machine: the type 1 DNA restriction enzyme EcoKI

  • Dryden, David (Principal Investigator)

Project Details

Description

The type I restriction enzymes were, as their name suggests, the first restriction enzymes to be analysed genetically and biochemically with work starting in 1953. Analysis of type I restriction enzymes paved the way for the discovery of the type II restriction enzymes and the birth of genetic engineering. The type I enzymes are complex multifunctional enzymes of huge size ( 1/3 the size of the ribosome, molecular weight 440,000, approximately 10nm in diameter) and, in stark contrast to the type II restriction enzymes, the type I enzymes cut DNA into random fragments. Despite their biochemical complexity, the genes for type I restriction systems appear in at least 50% of bacterial genomes so whilst they are not technologically useful, Nature clearly has a use for them. It has been shown that in the correct genetic background that fewer than one in 100 million lambda phage could escape restriction in vivo by the archetypal type I enzyme, EcoKI, the subject of this proposal. Type I restriction enzymes can be very effective bacterial defense systems.

It transpires that type I restriction enzymes are "smart molecular machines" capable of selecting and then performing multiple functions including the maintenance of chromosomal DNA methylation and the manipulation of great lengths of DNA at speeds reaching 1000 base pairs per second and generating forces of up to 5pN. The fragments of DNA produced by these enzymes are likely to be recombinogenic and similarities between various catalytic domains of type I restriction enzymes and the classic RecBCD enzyme, whose structure has recently been solved, are apparent. To perform so many functions in one enzyme requires a complex oligomeric structure which has been largely derived by Dryden and his long-term collaborator Prof. Noreen Murray FRS from extensive genetic, biochemical and biophysical analyses.

However, despite much effort by all of the researchers studying different type I restriction enzymes for the past 30 years, no crystal structure information has ever been achieved for the entire enzyme or for any subunit or domain with any combination of substrates. It is therefore timely to approach the structure problem by modern electron microscopy by establishing a collaboration between two leading groups in these areas.

Excellent preliminary EM images have been obtained showing much structural detail and additional spectroscopic data shows complex behaviour as the enzyme goes through its various reaction cycles (binding, methylation, translocation and cleavage). EM analysis can give structures reaching 7 angstrom resolution at various points during the reaction. Combining this with the extensive biophysical data and the experimentally validated structural models of domains will define the location of domains and subunits within the EM structure and lead to a complete structural model of this complex nanomachine.


Our objectives were achieved at the end of the grant as described below.
1. Determine to better than 1nm resolution the structure of EcoKI.
2. Identify the locations of all five subunits in the EcoKI structure
3. Determine the location of the extended DNA binding site on EcoKI.
4. Determine the location of the binding sites for antirestriction proteins on EcoKI.
5. Identify the location of active domains within the subunits in the EcoKI structure.
6. Time-resolved EM imaging of substrate-induced conformational changes.

The first five objectives have been achieved not only for EcoKI but also for an additional Type I restriction enzyme called EcoR124I. The resolution attained is 1.5 to 2 nm rather than 1nm but the elucidation of several atomic resolution structures of individual subunits of these enzymes and collaboration with a bioinformatics group in Warsaw during the period of this grant, has allowed excellent atomic models of the complete enzymes to be constructed. These models are in agreement with over 40 years of biochemical and genetic experimentation and can be considered to be reliable. Although large scale conformational changes caused by substrate binding were apparent, the Leeds group was unable to pursue the time-resolved aspect of objective 6 as the attainment of objectives 1 to 5 was slower than hoped.

It is worth mentioning that none of the objectives would have been achieved if the grant award had been for anything less than 60 months due to the complexity of these molecular motor systems.

Layman's description

Restriction enzymes are the workhorses of molecular biology, cutting DNA molecules accurately into the precise fragments required for virtually all molecular biology experiments. They were first purified in 1972 and without them modern experimental biology would be very different. However, few people are aware that these enzymes are properly referred to as type II restriction enzymes. What are the type I restriction enzymes?

The type I restriction enzymes were the first restriction enzymes to be analysed genetically and biochemically with work starting in 1953, the same year as Watson and Crick published their famous model for DNA structure. Analysis of type I restriction enzymes paved the way for the discovery of the type II restriction enzymes and the birth of genetic engineering. The relative obscurity of the type I enzymes arises because they are not only complex multifunctional enzymes of huge size (molecular weight 440,000, approximately 10nm in diameter) but also, in stark contrast to the highly commercial and important type II restriction enzymes, the type I enzymes cut DNA into random fragments of no practical use. Despite their biochemical complexity, type I restriction systems appear in at least 50% of bacteria so whilst they are not technologically useful, Nature clearly has a use for them. It has been shown that in the correct genetic background that fewer than one in 100 million lambda phage (viruses) could escape destruction in vivo by the archetypal type I enzyme, EcoKI, the subject of this proposal. Therefore, type I restriction enzymes can be very effective bacterial defense systems.

It transpires that type I restriction enzymes are "smart molecular machines" capable of selecting and then performing multiple functions including the maintenance of chromosomal DNA methylation and the manipulation of great lengths of DNA at speeds reaching 1000 base pairs per second and generating forces of up to 5pN. The fragments of DNA produced by these enzymes are likely to be recombinogenic (i.e. they can change the genetics of the bacterium) and similarities between various catalytic domains of type I restriction enzymes and the classic recombination enzyme, RecBCD, whose structure has recently been solved, are apparent. To perform so many functions in one enzyme requires a complex oligomeric structure.

However, despite much effort by all of the researchers studying type I restriction enzymes for the past 30 years, no detailed atomic structure has ever been achieved for the entire enzyme or for any subunit or domain with any combination of substrates. It is therefore timely to approach the structure problem by modern electron microscopy by establishing a collaboration between two leading groups in these areas. Electron microscopy is perfectly suited to finding the structure of a large machine such as EcoKI. Although the resolution of electron microscopy is limited, it can be combined with all of the current knowledge to produce a high resolution structure and to help us understand how one of the first nanomachines operates.

Key findings

This grant supplied the raw materials for the PI, John Trinick at Leeds, to solve the structures of Type I restriction endonucleases (published) and Type I DNA methyltransferases (published).

These 0.5 MDa structures for two Type I enzymes have now been solved with and without DNA bound and with a DNA-mimic antirestriction protein bound revealing how these complex DNA-translocating motors operate to control horizontal gene transfer in bacteria.

We were also able to solve the atomic structures of three antirestriction proteins (all published), ArdA found on the mobile genetic elements spreading antibiotic resistance genes, and two variants of ArdB another protein found on mobile elements but operating in an unknown way.

The ArdB protein fold was a novel fold while the ArdA protein was highly elongated and mimicked the shape and charge distribution of 42 base pairs of duplex DNA making it the largest DNA mimic yet discovered.


Structure and operation of the DNA-translocating Type I DNA restriction enzymes. Christopher K Kennaway, James E Taylor, Chun Feng Song, Wojciech Potrzebowski, William Nicholson, John H White, Anna Swiderska, Agnieszka Obarska-Kosinska, Philip Callow, Laurie P Cooper, Gareth A Roberts, Jean-Baptiste Artero, Janusz M Bujnicki, John Trinick, G Geoff Kneale and David T F Dryden. Genes and Development (2012), 26, 92-104.

Extensive DNA mimicry by the ArdA antirestriction protein and its role in the spread of antibiotic resistance. Stephen A. McMahon, Gareth A. Roberts, Kenneth A. Johnson, Laurie P. Cooper, Huanting Liu, John H. White, Lester G. Carter, Bansi Sanghvi, Muse Oke, Malcolm D. Walkinshaw, Garry W. Blakely, James H. Naismith, David T.F. Dryden. Nucleic Acids Res. (2009) 37, 4887-4897.

The structure of the KlcA and ArdB proteins reveals a novel fold and antirestriction activity against Type I DNA restriction systems in vivo but not in vitro. Dimitra Serfiotis-Mitsa, Andrew P Herbert, Gareth A Roberts, Dinesh C Soares, John H White, Garry W Blakely, Dusan Uhrin, David TF Dryden Nucleic Acids Res. (2010) 38, 1723–1737.

The structure of M.EcoKI Type I DNA methyltransferase with a DNA mimic antirestriction protein. Christopher K Kennaway, Agnieszka Obarska-Kosinska, John H White, Irina Tuszynska, Laurie P Cooper, Janusz M Bujnicki, John Trinick and David T F Dryden. Nucleic Acids Research (2009) 37, 762-70.

The Scottish Structural Proteomics Facility: targets, methods and outputs. Muse Oke, Lester S. G. Carter, Kenneth A. Johnson, Haunting Liu, Stephen A. McMahon, Xuan Yan, Kerou Melina, Nadine D. Weikart, Nadia Kadi, Md Arif Sheikh, Stefan Schmelz, Mark Dorward, Michal Zawadzki, Christopher Cozens, Helen Falconer, Helen Powers, Ian M Overton, C. A. Johannes van Niekerk, Xu Peng, Roger A. Garrett, David Prangishvili, Catherine H. Botting, Peter J. Coote, David T. F. Dryden, Geoffrey J Barton, Ulrich Schwarz-Linek, Gregory L. Challis, Garry L.Taylor, Malcolm F. White & James H. Naismith. Journal of Structural and Functional Genomics. (2010) 11, 167–180.
StatusFinished
Effective start/end date1/12/0530/11/10

Funding

  • BBSRC: £123,543.00