Restriction-Modification (R-M) systems in E. coli
evolved to protect bacteria from invading foreign DNA such as bacteriophage genomes. These R-M systems can be split into two broad classes.
- Those that protect bacterial DNA from restriction (degradation) by modification (methylation) of specific sequences that are recognized by the corresponding restriction enzymes of the R-M system.
- Those that only cut DNA bearing foreign modifications. The host DNA is then protected from cleavage by the restriction enzymes of this system by virtue of it not being modified.
The first type include the classic R-M systems as exemplified by the type I and type II restriction and modification systems. Type II include the typical restriction enzymes used in molecular biology. However, for cloning of foreign DNA sequences in E. coli, the most important is the type I system encoded by the hsdR, hsdM and hsdS genes. The products of all three genes form a multimeric enzyme capable of cleaving or methylating a particular target sequence (5´-AAC[N6]GTGC-3´ in E. coli K12 strains) (1)(2). If the target sequence is hemimethylated (only one strand is methylated as occurs during DNA replication of the host chromosome), then the enzyme acts as a methylase thus protecting the sequence from cleavage. If the sequence is not methylated on any strand, then the enzyme acts as a restriction endonuclease and cuts the target sequence. Typically DNA from another organism will not be methylated at this target sequence and will be degraded when introduced into a strain that is wildtype for all three genes. As the hsdR gene is required only for endonuclease cleavage of the target sequence (hsdM and hsdS are necessary and sufficient for methylation of the target sequence), E. coli strains that are mutated for hsdR have the so-called restriction minus, modification plus phenotype (r–, m+). This means they can be used to clone foreign DNA that is not methylated at the target sequence. In fact, when such DNA is propagated in this genetic background, the target sequence becomes methylated, which then allows it to be grown in a strain that is wildtype for all three hsd genes. Examples of E. coli strains available from Promega with this phenotype (r–, m+) include JM109 (Glycerol Stock: Cat.# P9751; Competent Cells: Cat.# L1001 and L2001); glycerol stocks of: JM109(DE3) (Cat.# P9801)and LE392 (Cat.# K9981). Mutations in either the hsdM or hsdS genes result in a restriction-modification minus phenotype (r–, m–). DNA propagated in this genetic background is not subject to endonuclease cleavage but similarly is not methylated. Consequently, plasmid DNA isolated from such a strain cannot be introduced into a strain that is wildtype for hsdR, hsdM and hsdS (r+, m+) as it will be degraded. It first has to be passed through a restriction minus modification plus strain such as JM109 to protect it from cleavage. An example of where this can be an issue is the GeneEditor™ in vitro Site-Directed Mutagenesis System. The mismatch repair-deficient strain used in this system (BMH 71-18 mutS) is wildtype for hsdR, hsdM and hsdS genes and therefore will degrade any plasmid DNA that has been grown in a methylation minus strain of E. coli such as HB101. Plasmids grown in strains that are wildtype for methylation (such as JM109 and DH5α™) are protected from such degradation. Thus, before transforming a plasmid into a different strain of E. coli, it is important to consider the hsdR, hsdM and hsdS genotype of the strain that the plasmid was grown in as well as the genotype of the strain into which you plan to transform it.
The Dam (DNA adenine methyltransferase) and Dcm (DNA cytosine methyltransferase) modification systems methylate adenines and cytosines located within specific recognition sequences (5´-GATC-3´ for Dam and the second cytosine of 5´-CCA/TGG-3´ for Dcm) (3). The biological role of these modifications does not appear to be to protect bacterial chromosomal DNA from a corresponding restriction endonuclease within the bacterium (such as with the class I R-M systems). However, the effect of these modifications is to prevent certain restriction enzymes used in molecular biology from cutting their corresponding target sequence in plasmid DNAs. This can happen if the restriction enzyme recognition sequence encompasses the methylation sequence or overlaps the methylation sequence (the methylated adenine or cytosine usually residing within the restriction enzyme recognition sequence). For example, the restriction enzyme XbaI will not cut its recognition sequence (5´-TCTAGA-3´) when the last adenine is methylated. For obvious reasons this can be a problem if one is not able to cut a particular site. This can be overcome by transforming the plasmid into a strain that is mutated for dam, or dcm or both genes.
The second type of restriction-modification system consists of those that direct endonuclease cleavage to DNA targets that are methylated on certain sequences (2). In E. coli there are two methyl-cytosine restricting systems called McrA and McrBC. In addition, there is also a methylated adenine recognition and restriction system called Mrr. Neither of these systems cleaves DNA that has been methylated by the type I (encoded by the hsdRMS genes), Dcm or Dam systems. Both the McrA and McrBC systems restrict hemimethylated DNA as well as DNA methylated on both strands. DNA that has been methylated by the SssI methylase (methylates the cytosine of the CG dinucleotide) and the HpaII methylase (methylates the second cytosine of the sequence 5´-CCGG-3´) are restricted by the McrA system (4)(5). The McrBC system cleaves DNA methylated at the cytosines of the sequence (G/A)C (6). As mammalian DNA tends to be methylated at cytosines in CG sequences and plant DNA at cytosines in CNG sequences, it is recommended to use strains that are mutated for these systems to increase the recovery of genomic DNA propagated in E. coli during cloning experiments. The Mrr system cleaves DNA methylated at adenines and has also been reported to restrict DNA methylated at cytosines. However, there is no apparent consensus sequence for recognizing these methylated bases by this system. As with the McrA and McrB systems, strains that are mutated in the Mrr system should be considered when cloning DNA from other organisms as plants and mammals usually contain some degree of methylation that can be recognized by these systems.