Conceived and designed the experiments: SHC SyX. Performed the experiments: SHC LO LH SyX. Analyzed the data: SHC LO LH SyX. Contributed reagents/materials/analysis tools: SHC DO SyX. Wrote the paper: SHC SyX.
Current address: Value Stream Secondary, GlaxoSmithKline Biologicals, Dresden, Germany
SHC SYX DO LH are employees of New England Biolabs (NEB), Inc. This employment does not alter the author's adherence to all the PLoS ONE policies on sharing data and materials. NEB is a commercial source of HpyAV and KpnI restriction endonucleases. LO was a one-year visiting student working on a Master degree thesis (supported by NEB).
We have cloned, expressed and purified a new restriction endonuclease HpyAV from
Some HNH-type endonucleases have unique metal ion cofactor requirement for optimal activities. Homology modeling and site-directed mutagenesis confirmed that HpyAV is a member of the HNH nuclease family. The identification of catalytic residues in HpyAV paved the way for further engineering of the metal binding site. A survey of sequenced microbial genomes uncovered 10 putative R-M systems that show high sequence similarity to the HpyAV system, suggesting lateral transfer of a prototypic HpyAV-like R-M system among these microorganisms.
Restriction-modification (R-M) systems that recognize and cleave DNA in a highly specific manner are ubiquitous in prokaryotic microorganisms (and their viruses)
Genome mining of sequenced microbial genomes has resulted in a wealth of restriction enzymes with new specificities or unique properties (ApeKI (G∧CWGC), PhoI (GG∧CC), CviKI-1 (RG∧CY), NmeAIII (GCCGAG 20–21/18–19)
Restriction mapping and run-off sequencing results indicated that the native HpyAV REase isolated from
The purified recombinant HpyAV exhibited very low cleavage activity on λ DNA in the standard reaction condition containing 4 mM MgSO4 (
HpyAV | KpnI | |||
Sp. Activity (U/mg) |
% | Sp. Activity (U/mg) |
% | |
Mg++ | <<200 | <0.4 |
282000 | 100 |
Ca++ | <<200 | <0.4 |
7000 | 2.5 |
Mn++ | 2000 | 200 | N/Dc | N/Dc |
Co++ | 830 | 83 | 42000 | 15 |
Ni++ | 1000 | 100 | 32000 | 10 |
Cu++ | <200 | <6 |
N/Dc | N/D |
Zn++ | <200 | <0.8 |
28000 | 10 |
Specific activity are average values of triplicate experiments for HpyAV and duplicate experiments for KpnI.
Complete cleavage of substrate DNA was not achieved with the highest available concentration of HpyAV. Percentage activity was estimated by comparing the cleavage pattern of the highest concentration of HpyAV with the indicated metal ion to the matching pattern with Ni++ (data not shown).
The specific activity of KpnI was not determined in Mn++ because star activity appeared before complete cleavage of the substrate DNA was achieved.
Specific activity of KpnI was not determined in Cu++ because the same incomplete cleavage pattern was observed over a 120-fold difference in enzyme concentration.
The strong stimulation effect of Ni++ on endonuclease activity is unique to HpyAV. It led us to examine the HpyAV amino acid sequence in more details. HpyAV is not homologous to any known REases except its isoschizomer Hin4II (data not shown). Manual examination of the amino acid sequence of HpyAV revealed a HNH catalytic motif highly homologous to that of colicin E9. Homology modeling of amino acid (aa) residues 287–325 of HpyAV to the ββα-Me motif of ColE9 (aa 95–131) resulted in a model free of clashes and with all the conserved catalytic residues (His102, His103, Asn118 and His127 in ColE9; H294, H295, N311, and H320 in HpyAV) structurally aligned to the HNH endonucleases including I-HmuI and Hpy99I (
From biochemical and structural studies of colicin E9, His103 acts as the general base to deprotonate a water molecule for the hydrolysis of the scissile phosphodiester bond. His102 and His127 coordinate the single divalent metal ion for transition state stabilization. Asn87 of I-HmuI and Asn118 of ColE9 are proposed to form a hydrogen bond to the general base His and increase its pKa for the activation of the nucleophilic water
Purified protein of WT, H294D, H295A and H320A were assayed as described in
KpnI is an HNH Type IIP REase that has been shown to be active with Mg++ or Ca++
Ten units of KpnI or EcoRI were incubated with 1 µg of pXba or λ DNA, respectively, as described in
HpyAV is an isoschizomer of Hin4II. M.HpyAV contains a C5 cytosine and a N6 adenine MTase domain highly homologous to M1 and M2.Hin4II (
REase |
Length (aa) | % ID | MTase1 | Length (aa) | % ID | Organism |
HpyAV | 419 | 100 | M. HpyAV | 823 | 100 | |
HpyPORF48P | 423 | 92 | M.HpyPORF48P | 822 | 98 | |
HpyGORF49P |
419 | 96 | M.HpyGORF49P | 823 | 97 | |
CupORF1468P | 339 | 50 | M.CupORF1468P | 817 | 64 | |
SeqZORF1536P | 417 | 50 | M. SeqZORF1536P | 810 | 52 | |
Hin4II | 418 | 48 | M1.Hin4II M2.Hin4II | 387 445 | 57 |
|
BhyWAORF699P | 386 | 43 | M.BhyWAORF699P | 812 | 43 | |
Nme180ORF295P | 296 | 31 | M.Nme18ORF295P | 862 | 42 | |
MmyLCORFBP | 330 | 40 | M. MmyLCORFBP | 834 | 55 | |
YkrORF13790P | 426 | 39 | M.YkrORF13790P | 826 | 44 | |
Bst43183ORF2897P | 390 | 33 | M.Bst43183ORF2897P | 856 | 43 | |
VbaORF22060P | 326 | 30 | M.VbaORF22060P | 820 | 43 |
Names as in REBASE. All REases, except HpyAV and Hin4II, are putative (names end with P).
In the genomic sequence, HpyGORF49P contains a deletion that introduces a stop codon within the HNH catalytic site. The reported length is a theoretical number based on the re-introduction of the deleted nucleotide to the genomic sequence.
Sequence identity for M1 and M2.Hin4II were calculated based on pair-wise alignment of segments of M.HpyAV that can be aligned to M1 and M2.Hin4II, respectively (
It has been well-documented that for restriction endonucleases (REases) with the canonical PD-X-(D/E)XK catalytic site, Mg++ and Mn++ support catalysis but Ca++ only supports DNA binding. One to two coordinated metal ions have been observed in the catalytic site in the crystal structures of REases in the presence of substrate DNA. Catalytic mechanisms for one- and two- ion-induced cleavage have been proposed
Endonucleases that contain the ββα-Me motif, on the other hand, are only observed with one coordinated divalent metal ion in their catalytic sites. The ββα-Me motif adopts a V-shape conformation consisting of two beta-strands connected by a loop in one arm followed by a helix that constitutes the other arm. The ββα-Me motif is present in non-specific endonucleases such as
In this study, we demonstrated the importance of conserved catalytic residues H294, H295, N311, and H320 by site-directed mutagenesis. HpyAV variants H294D, N311A, and H320A lack any detectable catalytic activity (less than 0.1% activity). Substitution of the general base His295 by Ala displays impaired cleavage activity only (H295A retains approximately 1% of wt activity), suggesting that an alternative weaker general base exists in the catalytic site when the general base His295 is mutated. In addition, positively charged aa substitution by Lys in H295K did not rescue the catalytic activity, indicating Lys cannot replace His as the general base in HpyAV. Other charged aa residue substitutions at the 295 position (HpyAV variants H295N, H295D, and H295E) failed to show any detectable cleavage activity. The observation that deletion of the proposed histidine general base in the HpyAV active site yields an endonuclease with reduced activity, rather than completely inactivating the enzyme, is unusual for the HNH superfamily, but not unprecedented. Mutation of the active histidine general base in the I-PpoI homing endonuclease also results in reduced activity
MnlI and I-PpoI are the only ββα-Me endonuclease whose metal ion preference has been systematically studied. In the presence of 1 mM M++, the order of MnlI activity was shown to be Mg++ > Ni++ = Co++ > Mn++ > Ca++ > Zn++
HpyAV is unique in that it is the only HNH-type endonuclease characterized so far to have higher double-strand cleavage activity with transition metals (Mn++, Ni++ and Co++) than with Mg++. It has been argued that in ββα-Me REases, the metal ion is not involved in the coordination of the nucleophilic water but only interacts with the phosphoanion transition state and the leaving group through a water molecule, therefore allowing for a less stringent metal ion requirement for catalysis
The loss of fidelity of KpnI in the presence of Mn++ has also been observed in EcoRI
In addition to Hin4II, a BLASTP search of GenBank database discovered 10 putative R-M systems highly homologous to the HpyAV system. Interestingly, these R-M systems are mostly carried by infectious microorganisms of human or mammalian hosts closely associated with humans (
The HpyAV R-M system was identified in
The over-expression strain of HpyAV was cultured in LB medium containing 100 µg/ml ampicillin and 30 µg/ml kanamycin at 30°C and 200 rpm overnight (∼15 h). Ten milliliters of the overnight culture was inoculated into 1 L of LB medium containing the same antibiotics and cultured at 30°C and 200 rpm to log phase. The culture was cooled down to 25°C before IPTG was added to a final concentration of 0.25 mM. Growth was then continued at 25°C for ∼15 h and the cultures were harvested by centrifugation. The cell pellet was resuspended in 100 ml of 20 mM Tris-HCl, pH 8.0, 50 mM NaCl, 1 mM EDTA (Buffer A) supplemented with 1% PMSF and sonicated on ice. After centrifugation, the supernatant was loaded onto a Heparin HiTrap column (5 ml; GE Life Sciences). Peak fractions from a linear elution gradient of 0.05–1 M NaCl in Buffer A was diluted 4-fold in Buffer A and loaded onto a HiTrap SP HP column (5 ml; GE Life Sciences). Peak fractions from a 0.05–1 M NaCl gradient were pooled and concentrated by Vivaspin 15 (10 kDa MWCO; Sartorius). An equal volume of 60% glycerol was added to the concentrated protein for storage at −20°C.
The DNA cleavage activity of the crude extract or purified HpyAV was assayed in 50 µl reactions containing 20 mM Tris-HCl, pH 7.9, 200 mM NaCl supplemented with the indicated concentrations of MgSO4, CaCl2, MnCl2, CoSO4, NiSO4, Cu(OAc)2 or Zn(OAc)2 and 1 µg of λ DNA at 37°C for 1 h. KpnI was assayed in 20 mM Tris-HCl, 50 mM NaCl, pH 7.9 with the same battery of salts using 1 µg of pXba DNA (a 10 kb XbaI fragment of adenovirus DNA inserted into pUC19; NEB). EcoRI activity assay was carried out in the same buffer using 1 µg of λ DNA. The cleavage reactions were then analyzed by 1.2% agarose gel electrophoresis in 1x TBE. One enzyme unit is defined as the amount of enzyme needed to cleave the 1 µg of the designated DNA completely at 37°C in 1 h. Specific activity is defined as the number of units per mg of enzyme. Specific activity was determined in duplicate (KpnI) or triplicate (HpyAV) by titrating the enzymes (in steps of 2-fold dilution) in their respective reaction buffer in the presence of 2 mM MgSO4, CaCl2, MnCl2, CoSO4 or NiSO4, or 0.1 mM Cu(OAc)2 or Zn(OAc)2. KpnI and EcoRI were from NEB. All reaction buffers and metal ion solutions were prepared using MilliQ water.
Amino acid residues 281 to 360 of HpyAV were modeled to the ββα-Me motif of ColE9 (mutant H103A; PDB: 1V14) by homology modeling using SWISS-MODEL
We thank Rich Roberts, Bill Jack, and Elisabeth Raleigh for critical comments and discussions; Barry Stoddard for informing us of the HpyAV homologs; Jim Ellard and Don Comb for support and encouragement; NEB DNA sequencing lab for sequencing the HpyAV expression clone and HpyAV variants.