Thermo-switchable activity of the restriction endonuclease SsoII achieved by site-directed enzyme modification

Authors

  • Liudmila A. Abrosimova,

    1. Department of Bioengineering and Bioinformatics, M. V. Lomonosov Moscow State University, Moscow, Russian Federation
    2. Department of Chemistry, and Belozersky Institute of Physico-Chemical Biology, M. V. Lomonosov Moscow State University, Moscow, Russian Federation
    Search for more papers by this author
  • Mayya V. Monakhova,

    1. Department of Chemistry, and Belozersky Institute of Physico-Chemical Biology, M. V. Lomonosov Moscow State University, Moscow, Russian Federation
    Search for more papers by this author
  • Anzhela Yu. Migur,

    1. Department of Bioengineering and Bioinformatics, M. V. Lomonosov Moscow State University, Moscow, Russian Federation
    2. Department of Chemistry, and Belozersky Institute of Physico-Chemical Biology, M. V. Lomonosov Moscow State University, Moscow, Russian Federation
    Search for more papers by this author
  • Wende Wolfgang,

    1. Institute of Biochemistry, Department of Biology and Chemistry, Justus-Liebig University, Giessen, Germany
    Search for more papers by this author
  • Alfred Pingoud,

    1. Institute of Biochemistry, Department of Biology and Chemistry, Justus-Liebig University, Giessen, Germany
    Search for more papers by this author
  • Elena A. Kubareva,

    1. Department of Chemistry, and Belozersky Institute of Physico-Chemical Biology, M. V. Lomonosov Moscow State University, Moscow, Russian Federation
    Search for more papers by this author
  • Tatiana S. Oretskaya

    Corresponding author
    1. Department of Chemistry, and Belozersky Institute of Physico-Chemical Biology, M. V. Lomonosov Moscow State University, Moscow, Russian Federation
    • Address correspondence to: Tatiana S. Oretskaya, Department of Chemistry, M. V. Lomonosov Moscow State University, Leninskie Gory 1, Moscow 119234, Russian Federation. Fax: +74959393181. E-mail: oretskaya@belozersky.msu.ru

    Search for more papers by this author

Abstract

In this work, the possibility of constructing a thermo-switchable enzyme according to the “molecular gate” strategy is demonstrated. The approach is based on the covalent attachment of oligodeoxyribonucleotides to cysteine residues of an enzyme adjacent to its active center to form a temporal barrier for enzyme–substrate complex formation. The activity of the modified enzyme that had been studied here—the restriction endonuclease SsoII (R.SsoII)—was diminished by a factor of 180 at 25 °С that almost abolished the enzymatic activity when compared with the unmodified enzyme. However, heating of the modified enzyme to 45 °С resulted in a 30-fold increase of activity. The activity of unmodified R.SsoII also increased on heating from 25 to 45 °; however, the difference did not exceed a factor of 3–4. The changes in enzymatic activity observed were shown to be reversible for both the unmodified and the modified R.SsoII. Variation of the length and the sequence of the attached oligodeoxyribonucleotides might allow greater modulation of the activity of DNA–protein conjugates. © 2013 IUBMB Life, 65(12):1012–1016, 2013

Introduction

All organisms have to adapt to changing environmental conditions. Sensing environmental temperature is one of the most fundamental functions of living creatures. Temperature dramatically affects the vitality of all organisms. Transcription factors, for example, in plants, alter their ability to bind to DNA in response to temperature changes [1-3]. These transcription factors can regulate the biosynthesis of secondary metabolites, such as anthocyanins, influence fruit ripening, and alter amino acid metabolism [4, 5]. Expression of Cer1, a retrotransposon of C. elegans, has the highest level at 15 °С and is not detectable at 25 °С [6]. Temperature sensitivity of influenza virus mutants is mediated by the functioning of a viral temperature-sensitive endonuclease [7]. Nevertheless, the mechanism of the influence of temperature on protein activity remains enigmatic and is intensively studied at present [8-11].

The conjugation of proteins with oligodeoxyribonucleotides often considerably changes their properties. For example, conjugates of oligonucleotides with cell-penetrating peptides were reported to be useful agents for gene silencing applications [12]. Covalent linking of an antisense nonadeoxyribonucleotide to E. coli ribonuclease HI alters its kinetic parameters [13] and allows cleavage of the oligoribonucleotide without addition of an antisense oligodeoxyribonucleotide. Fusion with oligonucleotides can also extend the specificity of restriction endonucleases, which is a necessary prerequisite for targeting individual genes in complex genomes [14].

The aim of this work was to construct an enzyme with pronounced thermo-dependent properties that could later serve as a model to understand the mechanism of naturally occurring thermo-dependent processes. Keeping in mind that transcription factors usually function as homodimers, we chose a well-studied homodimeric DNA-binding protein as a model system—the restriction endonuclease SsoII (R.SsoII). Conjugation of R.SsoII with DNA seemed to be a promising approach for constructing an enzyme with the desired thermo-switchable properties.

Experimental Procedures

Software

The length of the dekadeoxyribonucleotide with the maleimide group was assessed using Accelrys DS Visualizer.

Synthesis

Synthesis of self-complementary dekadeoxyribonucleotide 1 [5′-NH2-(CH2)6p-CACACGTGTG-3′] and dekathymidylate 2 [5′-NH2-(CH2)6p-ТТТТТТТТТТ-3′] was performed by standard procedures using commercial reagents and solvents.

Oligodeoxyribonucleotide Modification by N-(γ-Maleimidobutyryloxy)succinimide Ester

About 30 µL of 0.28 mM oligodeoxyribonucleotide solution in phosphate buffered saline (PBS) buffer was mixed with 20 µL of 0.21 M N-(γ-maleimidobutyryloxy)succinimide ester (GMBS) dissolved in dimethyl sulfoxide (DMSO). The mixture was stirred for 1 h at room temperature. The yield of the reaction as analyzed by ion-pair RP-HPLC was ∼95%. R.SsoII variant was purified according to the protocol described in ref. [15].

Modification of the R.SsoII Variant by Polyethylene Glycol–Maleimide

About 10 µL of a 10 µM enzyme solution (calculated for the monomer) in PBS buffer was mixed with 0.5 µL of 5 mM polyethylene glycol–maleimide (PEG–Mal, Mr = 5 kDa) dissolved in DMSO. The mixture was incubated for 15 min at room temperature and analyzed by SDS-PAGE.

Modification of the R.SsoII Variant by Oligodeoxyribonucleotide–Maleimide

About 3 µL of a 36 µM R.SsoII(S171C) solution was added to 7 µL of a 305 µM oligodeoxyribonucleotide–maleimide solution in PBS buffer. The mixture was incubated for 1 h at room temperature. The yield of the conjugate as determined by SDS-PAGE was about 90%. The unreacted oligodeoxyribonucleotide was not removed from the reaction as its presence does not influence the enzymatic activity of the conjugate (data not shown).

Hydrolysis of DNA Substrate by R.SsoII(S171C) and Its Conjugates with Oligodeoxyribonucleotides

A 30-bp DNA duplex containing a single recognition site for R.SsoII and a [32Р]-label at its 5′-end was used as a substrate (5′-GTATGAAGCTAGAGCCAGGTTGGCAGCATC-3′/3′-CATACTT CGATCTCGGTCCAACCGTCGTAG-5′). For the analysis of the cleavage reactions, 20 nM DNA duplex and 100 nM R.SsoII(S171C) or its conjugates (calculated for the monomer) were incubated in 10 mM Tris-HCl (pH 7.5), 15 mM MgCl2, 50 mM NaCl, 1 mM DTT, and 0.1 mg/mL BSA for 30 min. The reaction products were analyzed by polyacrylamide gel electrophoresis with 7 M urea in Tris/borate/EDTA buffer.

Results and Discussion

R.SsoII is a part of the restriction modification system found in Shigella sonnei 47 [16]. The protein consists of 305 amino acids (Mr = 35.9 kDa). R.SsoII is a mesophilic enzyme [16]. It recognizes a pentanucleotide sequence in double-stranded DNA and hydrolyzes the phosphodiester bonds next to the recognition site (marked with arrows): 5′-↓CCNGG-3′/3′-GGNCC↑-5′, where N is A, G, C, or Т. R.SsoII has an isoschizomer—R.Ecl18kI. The two enzymes differ only slightly: V232 of R.Ecl18kI corresponds to I232 of R.SsoII. The conclusions about DNA binding drawn from the X-ray structure analysis of R.Ecl18kI in complex with its substrate (PDB code 2fqz; Fig. 1) are in full agreement with our biochemical data for R.SsoII DNA binding [17-19]. Thus, the structural data obtained for R.Ecl18kI can be used to choose amino acids of R.SsoII for further modification.

Figure 1.

Structure of the R.SsoII homodimer (based on the X-ray data for R.Ecl18kI). Green color indicates amino acid residue substituted by cysteine. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

To regulate the activity of DNA-binding proteins by means of an external signal, we suggest using the “molecular gate” strategy [20]. The strategy is based on the fact that many DNA-binding proteins function as homodimers and their active site is located close to the interface of the homodimer. The majority of enzymes undergo conformational changes during interaction with a substrate that favor DNA binding in the active center. To create a physical obstacle for DNA penetration into the active center one could modify the protein by different compounds near the DNA-binding center [21]. We chose to modify R.SsoII by oligodeoxyribonucleotides that are able to form a duplex at the entrance of the DNA binding site. Because R.SsoII functions as a homodimer, it is necessary to use self-complementary oligodeoxyribonucleotides for its modification. The double-strand formation of oligodeoxyribonucleotides is reversible and cooperative, thus allowing switching of the enzymatic activity on and off multiple times (Scheme 1).

Scheme 1.

The principle of the “molecular gate” strategy. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

The duplex is suggested to be stable at room temperature, thus preventing penetration of the substrate into the active center (the molecular gate is closed). Increasing the temperature will cause duplex dissociation and recovery of enzymatic activity (the molecular gate is opened). In addition, a structure adopted at intermediate temperature is also shown, in which the single-stranded oligodeoxyribonucleotide adopts a hairpin in which the “gate” is partly closed.

Previously, single-stranded oligodeoxyribonucleotides with a terminal maleimide group were used to assess the redox status of cysteinyl thiol groups [22]. We suggest using oligodeoxyribonucleotides with a terminal maleimide group for protein modification that in our case will result in the formation of a molecular gate at the entry of the R.SsoII DNA-binding center. Hence for implementation of the molecular gate strategy, it is necessary to have single-cysteine variants of the protein. Such variants of R.SsoII were obtained previously in our laboratory [23]. According to the cocrystal structure of R.Ecl18kI in complex with its substrate, R.SsoII contains external and internal “clamps” that surround the specifically bound DNA substrate. The R.SsoII(S171C) variant containing a single cysteine residue in the external clamp was used in this study.

The accessibility of R.SsoII(S171C) was probed using PEG–Mal (Mr = 5 kDa). The reaction of PEG–Mal with R.SsoII(S171C) had a quantitative yield (Fig. 2) that confirmed the high reactivity of the two symmetry-related Cys residues. Therefore, the variant R.SsoII(S171C) was used for further modification by oligodeoxyribonucleotides containing a maleimide group.

Figure 2.

Analysis of the ability of the Cys residues in R.SsoII(S171C) to react with the maleimide group of the modified oligodeoxyribonucleotide. Lanes: marker proteins, MW (kDa) is indicated on the left [1]; R.SsoII(S171C) [2]; and R.SsoII(S171C) after reaction with PEG–Mal [3] and with the modified oligodeoxyribonucleotides 1 and 2 (4 and 5). Silver staining of a 12% SDS gel is shown. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

The design of the maleimide-modified oligodeoxyribonucleotide for protein modification included several steps. First, the sequence of the oligodeoxyribonucleotide had to be self-complementary to modify the two identical subunits of the R.SsoII(S171C) homodimer. Second, the oligodeoxyribonucleotide had to be sufficiently long enough to close the active center of the enzyme by forming a stable duplex at room temperature, but to be able to dissociate at a temperature at which the enzyme is still active.

The distance between the sulfur atoms of the two Cys 171 residues in the R.SsoII(S171C) homodimer is 49 Å, and the length of the dekadeoxyribonucleotide modified by maleimide group is ∼54 Å. Hence, a dekadeoxyribonucleotide is presumably able to block the entry of the substrate into the active center of the enzyme. For protein modification, a self-complementary dekadeoxyribonucleotide 1 and dekathymidylate 2 were synthesized. Both oligodeoxyribonucleotides contained an amino linker at their 5′-end for subsequent conjugation with the protein. DNA–protein conjugation was performed with GMBS. The maleimide group of GMBS reacts with a protein thiol group, whereas its succinimide group is able to react with a terminal amino group of a modified oligodeoxyribonucleotide [14].

Two DNA–protein conjugates were obtained: the R.SsoII(S171C) conjugate with oligodeoxyribonucleotide 1 [R.SsoII(S171C)−1, Mr = 39.3 kDa] and 2 [R.SsoII(S171C)−2, Mr = 39.2 kDa] (Fig. 3). A 5′-32P-labeled 30-bp DNA duplex was used as a substrate in a hydrolysis reaction to assess the enzymatic activity of the unmodified and modified proteins. The activity of both DNA–protein conjugates were shown to be considerably lower when compared with the activity of the unmodified R.SsoII(S171C) (Fig. 3). This effect can be explained by the covalent attachment of oligodeoxyribonucleotides that creates steric hindrance for DNA substrate binding by the enzyme. However, it cannot be excluded that such a decrease of enzymatic activity may also be due in part to the effect of the modification on the local protein structure itself.

Figure 3.

Temperature dependence of the hydrolysis of a 30-bp DNA substrate by unmodified R.SsoII(S171C) (black squares) and its two conjugates: R.SsoII(S171C)−1 (red circles) and R.SsoII(S171C)−2 (green triangles) at different temperatures after 30 min. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

The introduction of a molecular gate at the entry to the active center almost completely abolished the enzymatic activity at 25 °С, whereas the DNA cleavage activity of the unmodified R.SsoII(S171C) was almost 180 times higher at the same temperature. Moreover, the temperature optimum of R.SsoII(S171C) is 37–40 °С, whereas the modified proteins have the highest activity at 45 °С. The shift of the optimal temperature suggests that at 25 °С, there is steric hindrance for substrate binding induced by the oligodeoxyribonucleotide at the entry of the DNA binding site that can be relieved by increasing the temperature, which results in the opening of the molecular gate and thereby in the increase of enzymatic activity.

It is worth noting that the effect of thermoregulation is more substantial for the R.SsoII(S171C)−1 conjugate than for the R.SsoII(S171C)−2 conjugate as the activity of R.SsoII(S171C)−1 varies in a broader range. This fact can be explained by the ability of the self-complementary oligodeoxyribonucleotide 1 to form a hairpin structure that is more compact than the single-stranded oligodeoxyribonucleotide and partly “opens” the gate for the substrate. The melting temperature of the hairpin formed by oligodeoxyribonucleotide 1 is about 51 °C. Melting of the hairpin structure at this temperature leads to the closure of the active center and a decrease of the enzymatic activity (Scheme 1). Oligodeoxyribonucleotide 2 does not adopt any secondary structure under the same conditions and thus appears to hinder the entry of the DNA substrate into the active center to a greater extent. At 25 °С, both oligodeoxyribonucleotides conjugated with the enzyme (in double- and single-stranded forms) presumably do not have much conformational freedom at this temperature, thus blocking substrate binding by the enzyme. According to these results, modification by the self-complementary oligodeoxyribonucleotide seems to be particularly effective to construct a thermo-switchable enzyme. Nevertheless, non-self-complementary oligodeoxyribonucleotide of the same length can also act as molecular gates decreasing the enzymatic activity. Increasing the temperature enlarges the conformational freedom of the oligodeoxyribonucleotides making the catalytic center more accessible for the substrate.

The ability to reversibly switch the enzymatic activity was studied in the temperature interval between 25 and 45 °С. For this comparison, the same amount of enzyme was incubated at different temperatures for 30 min, and then the initial rates of substrate hydrolysis were determined (Figs. 4a and 4b). The temperature was changed in the following order: 25 °С to 37 °С to 45 °С to 37 °С to 25 °С to 37 °С to 45 °С to 37 °С to 25 °С.

Figure 4.

Comparison of the initial hydrolysis rates determined for unmodified R.SsoII(S171C) (a) and its two conjugates (b) during two cycles of heating–cooling. The red (left) columns correspond to conjugate R.SsoII(S171C)−1, and the green ones (right) correspond to conjugate R.SsoII(S171C)−2. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Unmodified R.SsoII(S171C) was shown to retain its activity during two heating–cooling cycles. The activity of conjugates R.SsoII(S171C)−1 and R.SsoII(S171C)−2 was gradually decreasing, which could be due to an instability of the modified protein. Despite this, the overall tendency of changing the enzymatic activity was the same during two heating–cooling cycles.

Conclusion

Our molecular gate approach is based on DNA–protein conjugation and proposes a new way to switch on/off an enzymatic activity by temperature. The molecular gate strategy can be used in vitro for controlling enzyme activity by temperature. Our approach—introduced here as proof of principle—can also be applied for modification of other proteins that accept their substrate in a cleft, for example, some chaperones or transmembrane channels with a pore-like entrance, thus making these proteins thermosensitive and active in a desirable range of temperatures.

Acknowledgements

The plasmid used in this work was kindly provided by Dr. Karyagina (N. F. Gamaleya Institute of Epidemiology and Microbiology, Russia). This work was supported by grants RFBR-DFG 11-04-91338 and GRK 1384.

Ancillary