Search for new restriction enzymes in clinical bacterial strains

Restriction enzymes and the corresponding modification methylase (R-M systems) are unique to bacteria. Based on the cutting mode, restriction endonucleases are divided into three types (I to III). Among those, type II restriction enzymes are the most well studied and approximately 3,000 have been identified. Some of these have become essential tools for gene manipulation. However, our knowledge about the distribution of type I enzymes is limited. Less than thirty type I enzymes have been reported so far from mainly E. coli and S. typhimurium. Recently, the complete DNA sequences for a few hundred bacteria were revealed and many type I restriction enzyme homologues were found. These results suggest a wider spread of type I restriction enzymes.

To facilitate the process of identification of restriction enzymes in bacteria, a new simple method based on plasmid transformation was recently developed in our laboratory. This method is based on a plasmid R-M (restriction and modification) test. A computer program, RM search, was also developed to identify the recognition sequence of both type I and type II systems. Many enteric bacterial (especially, E. coli) strains have been collected from our medical center and the search for restriction enzymes is underway. Our preliminary results suggest there are many new restriction enzymes yet to be discovered in E. coli strains.

DNA recognition sequences of the type I restriction enzymes predicted from bacterial genome projects.

The elucidation of the DNA recognition sequences for type I enzymes is difficult and involves extensive enzyme purification, biochemical assays, DNA labeling, and the use of a computer program. As mentioned above, we have developed a new and simple in vivo method based on plasmid R-M tests and a computer program. Since many type I restriction enzyme homologues were identified from sequenced bacterial genomes, it is useful to apply our method to elucidate the DNA recognition sequences of putative type I enzymes. Those type I enzymes predicted from bacterial genome projects will be first cloned in E. coli using PCR. When expression is confirmed, the DNA recognition sequence will be elucidated using our plasmid R-M tests. This project is a joint project with Dr. Richard Roberts (1993 Nobel Laureate) in New England Biolabs who specializes in bioinformatics in restriction enzymes.

Regulation of type I restriction enzymes in enteric bacteria

The genes coding for type I restriction endonucleases can be transferred to bacteria in various methods such as conjugation and transformation, even if they do not contain any corresponding methylase activity. Since the restriction endonuclease transferred to the recipient should cleave the unmodified recipient chromosome, this successful "establishment" of the type I restriction enzymes poses an intriguing question about how restriction and modification genes are expressed after gene transfer. Using E. coli cells, we have shown that the expression is controlled by regulating the timing of the expression of restriction activity. We have also shown that the level of restriction subunits is controlled at the post translational level. A key factor was identified by identifying a mutant which does not have this control. We named this key gene “hsdC.” We then found that this key mutation is suppressed by a gene product of ClpX, a serine protease. The mechanism of this suppression is now under the investigation.