Journal of Nucleic Acids
Volume 2025, Issue 1 5518018
Research Article
Open Access

A Base Pair Outside the Catalytic Core of the I-R3 DNA Enzyme Has a Significant Effect on Its Cleavage Activity: An Improved Catalytic Core Model and an Automated Design Program

Shahidul Islam, 

Shahidul Islam

Department of Biology , Concordia University , Montreal , Quebec , Canada , concordia.ca

Search for more papers by this author
Gabriel Aguiar-Tawil, 

Gabriel Aguiar-Tawil

Department of Biology , Concordia University , Montreal , Quebec , Canada , concordia.ca

Center for Applied Synthetic Biology (CASB) , Concordia University , Montreal , Quebec , Canada , concordia.ca

Search for more papers by this author
Xinxin Yu, 

Xinxin Yu

Faculty of Computer Science , Dalhousie University , Halifax , Nova Scotia , Canada , dal.ca

Search for more papers by this author
Jonathan Ouellet, 

Jonathan Ouellet

Department of Chemistry and Physics , Monmouth University , West Long Branch , New Jersey , USA , monmouth.edu

Search for more papers by this author
Nawwaf Kharma, 

Corresponding Author

Nawwaf Kharma

Department of Biology , Concordia University , Montreal , Quebec , Canada , concordia.ca

Center for Applied Synthetic Biology (CASB) , Concordia University , Montreal , Quebec , Canada , concordia.ca

Department of Electrical and Computer Engineering , Concordia University , Montreal , Quebec , Canada , concordia.ca

Search for more papers by this author
First published: 08 March 2025
https://doi.org/10.1155/jna/5518018
Academic Editor: Ashis Basu

Abstract

The I-R3 DNA enzyme, in its trans-acting form, is capable of cleaving single-stranded DNA (ssDNA) molecules. We have collected all published information on the activity levels of the original I-R3 DNA enzyme and its known variants and embedded that information into a program (we called IR3). The program was applied to the sequences of a set of ssDNA viruses and identified all potential catalytic core substrates (targets) and output optimal I-R3 DNA enzyme sequences for all the targets, along with expected activity levels of the enzymes at those targets. Upon experimentally measuring the in vitro cleavage activities of the I-R3 variants, we found marked differences between the program-predicted and experimentally measured values. This demonstrated the incompleteness of the I-R3 model: The sequence of the nucleotides of the catalytic core is not sufficient to fully determine its activity level. A set of experiments was carried out in which the effect of all possible combinations of Watson–Crick base pairs at two positions near the catalytic core, termed SI and SII, was tested. To confirm a newly formed hypothesis, the nucleotide at the SII position of the enzyme strand was mutated to a G and a T, with the substrate strand mutated accordingly. In every case, this led to an increase in relative activity when changed to a G and a decrease, when changed to a T, of the variant I-R3 DNA enzyme. Clearly, the discovered base pair peripheral to the catalytic core has a substantial effect on cleavage activity. This improves the current model of essential nucleotides, and the IR3 software outputs I-R3 enzyme-sequence recommendations that make them more likely to cleave their targets. The software is available for download at https://github.com/XinxinTree/IR3.

1. Introduction

Since the discovery of catalytic RNAs in the 1980s [1, 2], advances in RNA therapeutics progressed with the discovery of several new nucleic acid–based technologies [3]. Such advances in the area of catalytic nucleic acids include small ribozymes [4, 5], DNAzymes [6, 7], and more recently DNA-cleaving DNA enzyme [8–11]. In 2013, two different structural classes of zinc-dependent DNA-cleaving DNA enzymes were artificially evolved [10]. One such structure was the I-R3. The I-R3 has a simple secondary structure consisting of a long, base-paired stem, which can be shortened, that is interrupted by an asymmetrical bulge, consisting of the catalytic core, where the cleavage occurs (Figure 1). After sequencing the systematic evolution of ligands by exponential enrichment (SELEX)–selected DNA pool and carrying out several site-directed mutagenesis experiments (including compensatory mutations), information on the essential nucleotides of the core was elucidated. Greater complexity of the essential core was determined by massive parallel sequencing of core-degenerated I-R3 cleavage assay [12]. Such complexity included percentage probability for previously invariable core nucleotides that provided a certain level of activity. Moreover, the randomness of the sequences generated by the experiment also provided simultaneous 2 and 3 base mutations within the core.

Details are in the caption following the image
Figure 1
Open in figure viewerPowerPoint
Secondary structure of the I-R3 wild type (WT) DNA enzyme–substrate as used in this work. Boxes are drawn to highlight the top base pair of stem PI (red), which precedes the catalytic core and termed SI, while the bottom base pair of stem PII (blue), which follows the catalytic core and is termed SII. The cleavage site is indicated by a dashed line between the two deoxyadenosines.

The catalytic core “bulge” can be designed as a trans-acting system where one DNA strand contains the top and bottom halves of the DNA stem, with an interrupting core of 10 deoxynucleotides. Since that strand is not cleaved, it is considered the enzyme strand (Figure 1). The other strand therefore consists of a substrate that is complementary to the top and bottom halves of the enzyme strand, interrupted by a seven-deoxynucleotide core that contains the cleavage site between two deoxyadenosines. Such a DNA enzyme can cut a DNA substrate requiring only seven conserved consecutive nucleotides. It provides a relatively small number of potential sequences: 47 = 16,384 different 7-nt sequences. As an example, a 4.6 million base pair Escherichia coli genome has a likelihood of having this exact sequence 280 times, and likelihood is even greater than 280 considering the number of tolerated substrate mutations. Moreover, the specificity of cleaving any of those seven-nucleotide long target sequences comes from the elongated base pair stems.

Finding and designing such an enzyme to target a substrate manually is time-consuming and error-prone compared to a bioinformatics-aided approach. In this article, we report on a bioinformatics tool that designs a DNA enzyme strand to target a single-stranded DNA (ssDNA) of interest, making use of the I-R3 DNA enzyme’s slightly variable core. To keep our hypothesis biologically relevant, we used synthetic segments from ssDNA viruses as targets. In several cases, while the software-predicted cleavage activity of some targets was high, we saw an absence of cleavage in our in vitro cleavage assay analyzed by polyacrylamide gel electrophoresis (PAGE) gel [13]. The mutagenesis of each core-adjacent base pair to any new canonical Watson–Crick base pair identified a previously unreported base pair requirement that expands the current model of the I-R3 catalytic core.

2. Materials and Methods

2.1. Materials

All oligonucleotides were purchased from Integrated DNA Technologies (Coralville, Iowa). Oligonucleotides used to study the effect of different bases on either end of the active site were obtained with a standard purity, while those used to study the viral sequences (wild type (WT) and variants) were obtained with a PAGE purity. A 10/60 ssDNA ladder was obtained from Integrated DNA Technologies (Coralville, Iowa). The 10x Tris/borate/EDTA (TBE) buffer was obtained from BioBasic (Markham, Ontario). Tetramethylethylenediamine (TEMED), 40% (19:1) acrylamide:bisacrylamide, ammonium persulfate, urea, 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES), sodium chloride, potassium chloride, and zinc chloride were obtained from BioShop Canada (Burlington, Ontario). The SyberGold DNA stain was obtained from ThermoFisher Scientific (Waltham, Massachusetts). A 2x RNA loading dye was obtained from New England Biolabs (Ipswich, Massachusetts).

2.2. Sample Preparation and Reaction Conditions for Cleavage Activity

The cleavage activity was performed in the same fashion (with small variations) as previously PAGE-based analysis of I-R3 cleavage activity [13]. A total of 2.5 μM DNA enzyme and 1 μM DNA substrate strands were annealed in 1x annealing buffer (50 mM HEPES pH 7.05, 500 mM NaCl, and 500 mM KCl) using a thermal cycler by heating at 95°C for 5 min then decreasing the temperature by 5°C every 3 min until it reached 20°C. Reaction buffer (40 mM ZnCl2, 50 mM HEPES, 500 mM NaCl, and 500 mM KCl, pH 7.05) was then added, and the mixture was kept at 30°C for 30 min in 20 mM ZnCl2. The reaction was stopped by the addition of 2x denaturing RNA loading dye and subsequently heated and kept at 95°C for 5 min, before being loaded onto a 20% denaturing polyacrylamide gel containing 8 M urea. Samples were run at 20 W for 2 h. The subsequent gel was stained with SyberGold diluted in 1x TBE buffer in the dark, while shaking it at 150 rpm for 30 min (SI4: control DNA enzyme digestion).

2.3. Data Analysis

Gels were subsequently imaged using a Syngene G:BOX EF2 gel documentation system, then analyzed using GelAnalyzer version 23.1.1. Images were uploaded as uncompressed.tif files. In GelAnalyzer, under “Detect Lines,” tilted lines were turned on, equal width was turned off, and the threshold was set to 3. “Baseline detection method” was set to “valley to valley”. Under “Detection Parameters,” minimal height was set to 3, minimal slope was set to 15, slope break threshold was set to 20, and profile smoothing rad was set to 1. Raw volume for both product bands (P1 and P2) and the remaining substrate band (S) in each lane was extracted from GelAnalyzer and used in the following equation to calculate cleavage activity.
()

Relative activity was calculated as the ratio of the given reaction activity to the WT I-R3 DNA enzyme/substrate reaction activity, measured under the same conditions and in the same laboratory.

3. Results and Discussion

3.1. Software

A software program named IR3 was custom-made to scan DNA sequences (presented 5⁣′–3⁣′) to locate potential substrate sequences for a trans-acting I-R3 DNA enzyme. For each substrate sequence, it presents all potential enzyme sequences and their respective relative I-R3 cleavage activity. The level of cleavage activity of an I-R3 DNA enzyme acting at a given substrate sequence was presumed to be identical to the reported level of activity for the same enzyme–substrate pair, as published [12]. Therefore, a relative activity was calculated with the ratio of the known extent of cleavage of a given variant over the extent of cleavage of the WT sequence (SI1: mutations and subsequent activity ratios from literature and SI6: mutation table). No attempt was made to extrapolate or make an intelligent prediction of what the relative activity could be for unpublished I-R3 sequences.

The program implemented the algorithm is described in detail in the supporting information (SI3: program). Below, we describe the algorithm in brief and in a form that facilitates its understanding by the casual computer user.
  • Step 1.

    Read the sequence of a ssDNA (presented 5⁣′–3⁣′) from a fasta text file.

  • Step 2.

    Identify potential seven nucleotide substrate sequences (i.e., sequences belonging to the substrate side of the catalytic core of a complete I-R3 DNA enzyme).

  • Step 3.

    For every substrate sequence, find the enzyme sequence that maximizes the relative activity at that site based on the activity ratios (SI1: mutations and subsequent activity ratios from literature and SI6: mutation table).

  • Step 4.

    Save the substrate sequence, the sequence of the chosen “optimal” I-R3 enzyme, the position of the substrate in the ssDNA, and the expected relative activity level.

  • Step 5.

    Upon reaching the end of the target substrate, generate the results in the following fashion:

  • 5.1.

    Extend the identified substrate sequence by 13 nucleotides so it now includes the full length of the substrate.

  • 5.2.

    Extend the I-R3 enzyme sequence, using Watson–Crick complementation, to generate complete PI and PII stems once annealed, while removing three nucleotides to the 5⁣′ end and one nucleotide to the 3⁣′ end of the substrate strand to provide greater separation on denaturing gels.

  • Step 6.

    Format all the results to output them in the form shown in Figure 2.

Details are in the caption following the image
Figure 2
Open in figure viewerPowerPoint
Sample output from the IR3 program on the human parvovirus 4 G1 sequence (GenBank Accession Number AY622943.1). In addition to the algorithm described above, it incorporates one additional (but critical) step that represents a significant result of our study. This step recommends a G-C or an A-T base pair for SII, which is at the bottom of the stem PII.

3.2. Testing of Software Results

To test the IR3 program, the sequences of various single-stranded DNA viruses were used as test cases for the software. These included the complete genomes of human parvovirus 4 G1 (HParvoV) (GenBank: AY622943.1), human bocavirus 2c isolate PK-5510 (HBocaV) (GenBank: FJ170278.1), and human circovirus 1 (HCircV) (GenBank Accession ON677309.1) (Supporting Information 2). Oligonucleotides were generated reflecting the identified target sequences and the corresponding enzyme strands (SI2: IR3-scanned ssDNA viruses and SI5: sequences used in this work). Sequences selected for in vitro cleavage intentionally included ones with predicted high, medium, and low relative activities. These oligonucleotides were incubated together in the reaction buffer and the subsequent products were run on a denaturing polyacrylamide gel and analyzed [13]. The initial results suggest that the existing model is insufficient for describing the activity of mutations within the I-R3 DNA enzyme (Figures 3(a) and 3(b)).

Details are in the caption following the image
Figure 3 (a)
Open in figure viewerPowerPoint
Cleavage assay of various ssDNA viruses from the IR3-generated corresponding DNA enzymes. (a) Separation of the enzyme, substrate, and product DNAs of various viral sequences on a 20% denaturing 8 M urea polyacrylamide gel after a 30-min incubation in a buffered 20 mM zinc solution. (b) Grey bars represent the relative activity calculated from the denaturing gel analysis (Equation (1)); flat black bars represent the predicted relative activity determined using the software. Error bars represent the standard error of the mean with n = 3.
Details are in the caption following the image
Figure 3 (b)
Open in figure viewerPowerPoint
Cleavage assay of various ssDNA viruses from the IR3-generated corresponding DNA enzymes. (a) Separation of the enzyme, substrate, and product DNAs of various viral sequences on a 20% denaturing 8 M urea polyacrylamide gel after a 30-min incubation in a buffered 20 mM zinc solution. (b) Grey bars represent the relative activity calculated from the denaturing gel analysis (Equation (1)); flat black bars represent the predicted relative activity determined using the software. Error bars represent the standard error of the mean with n = 3.

3.3. Effect of Certain Nucleotides Peripheral to the Active Site

Given that the existing model of essential nucleotides is based solely on the catalytic core (bulge) residues, the results in Figures 3(a) and 3(b) strongly suggest that there are essential nucleotides that are not included in the existing model. To investigate this, base pairs SI and SII (red and blue boxes in Figure 1, respectively) were changed to every combination of canonical base pairs across both base pairs. This led to a total of 16 combinations for both base pairs, including that of the WT sequence (Figures 4(a) and 4(b)). At the SI base pair, it appears that the nucleotide position does not seem to have a major effect as A-T, T-A, G-C, and C-G enzyme substrate base pairs all showed high activity (Figure 4(b), Bars 1–4 and 9–12). However, at the SII base pair, G-C (WT) and A-T showed high activity (Bars 1–4 and 9–12) while the C-G and T-A base pairs had activities practically abolished (Figure 4(b), Bars 5–8 and 13–16). From this, it was found that the SII position plays a substantial role in its activity. On the enzyme strand of position SII, a base pair starting with a purine (with G preferable to A) results in a high level of cleavage activity. In contrast, a pyrimidine at the same position is highly detrimental to cleavage activity.

Details are in the caption following the image
Figure 4 (a)
Open in figure viewerPowerPoint
Cleavage assay of the 16 combinations of canonical base pairs at SI and SII. (a) Separation of the enzyme, substrate, and product DNAs on a 20% denaturing 8 M urea polyacrylamide gel after a 30-min incubation in a buffered 20 mM zinc solution. (b) Relative activity calculated from the denaturing gel analysis. Error bars represent the standard error of the mean with n = 3. In (a) and (b), the first base in the sample name refers to the base in the enzyme-SI position, while the second base refers to the base on the enzyme-SII position. For example, the WT CG is a C-G base pair at SI and a G-C base pair at SII, with colors reflecting what is shown in Figure 1.
Details are in the caption following the image
Figure 4 (b)
Open in figure viewerPowerPoint
Cleavage assay of the 16 combinations of canonical base pairs at SI and SII. (a) Separation of the enzyme, substrate, and product DNAs on a 20% denaturing 8 M urea polyacrylamide gel after a 30-min incubation in a buffered 20 mM zinc solution. (b) Relative activity calculated from the denaturing gel analysis. Error bars represent the standard error of the mean with n = 3. In (a) and (b), the first base in the sample name refers to the base in the enzyme-SI position, while the second base refers to the base on the enzyme-SII position. For example, the WT CG is a C-G base pair at SI and a G-C base pair at SII, with colors reflecting what is shown in Figure 1.

3.4. Modulating Activity of I-R3 With Mutations at the SII Position

The emerging hypothesis was that having a purine on the enzyme side of the SII Watson–Crick base pair is essential for activity. The ssDNA virus sequences used in Figure 3 were mutated, changing the substrate SII nucleotide to a C and the corresponding enzyme SII nucleotide to a G. This results with a G-C base pair at the SII position (Figures 5(a) and 5(c)). In these cases, the activity level was substantially increased across all mutants. For example, HCircVa, b, and c; HBocaVa, b, and c; and HParvoVb showed little to no activity with their original purine nucleotide, while their relative activities now range between 0.59 and 1.06. Only HParvoVa showed a high activity before the mutation, which nevertheless was boosted from 0.82 to 1.06. This is likely because the original sequence has a T at the SII position of the substrate strand. The substantial gain of activity suggests that a C at the SII position of the substrate strand and a G in the enzyme strand are essential for significant cleavage activity.

Details are in the caption following the image
Figure 5 (a)
Open in figure viewerPowerPoint
Cleavage assay of SII-modified virus targets. (a) Separation of the enzyme, substrate, and product DNAs of viruses with C-substitution at the substrate-SII position and G at the enzyme-SII position on a 20% denaturing 8 M urea polyacrylamide gel, after a 30-min incubation in a buffered 20 mM zinc solution. (b) Separation of the enzyme, substrate, and product DNAs of viruses with A substitutions at the substrate-SII position and T on the enzyme-SII position on a 20% denaturing 8 M urea acrylamide gel. (c) Relative activity calculated from the denaturing gels in A and B. The black bar represents the wild-type I-R3 DNA enzyme. The blue bars represent the relative activities of all viral sequences with a C substitution at the substrate-SII position. The red bars represent the relative activities of all viral sequences with an A substitution at the substrate-SII position. Error bars represent the standard error of the mean with n = 3.
Details are in the caption following the image
Figure 5 (b)
Open in figure viewerPowerPoint
Cleavage assay of SII-modified virus targets. (a) Separation of the enzyme, substrate, and product DNAs of viruses with C-substitution at the substrate-SII position and G at the enzyme-SII position on a 20% denaturing 8 M urea polyacrylamide gel, after a 30-min incubation in a buffered 20 mM zinc solution. (b) Separation of the enzyme, substrate, and product DNAs of viruses with A substitutions at the substrate-SII position and T on the enzyme-SII position on a 20% denaturing 8 M urea acrylamide gel. (c) Relative activity calculated from the denaturing gels in A and B. The black bar represents the wild-type I-R3 DNA enzyme. The blue bars represent the relative activities of all viral sequences with a C substitution at the substrate-SII position. The red bars represent the relative activities of all viral sequences with an A substitution at the substrate-SII position. Error bars represent the standard error of the mean with n = 3.
Details are in the caption following the image
Figure 5 (c)
Open in figure viewerPowerPoint
Cleavage assay of SII-modified virus targets. (a) Separation of the enzyme, substrate, and product DNAs of viruses with C-substitution at the substrate-SII position and G at the enzyme-SII position on a 20% denaturing 8 M urea polyacrylamide gel, after a 30-min incubation in a buffered 20 mM zinc solution. (b) Separation of the enzyme, substrate, and product DNAs of viruses with A substitutions at the substrate-SII position and T on the enzyme-SII position on a 20% denaturing 8 M urea acrylamide gel. (c) Relative activity calculated from the denaturing gels in A and B. The black bar represents the wild-type I-R3 DNA enzyme. The blue bars represent the relative activities of all viral sequences with a C substitution at the substrate-SII position. The red bars represent the relative activities of all viral sequences with an A substitution at the substrate-SII position. Error bars represent the standard error of the mean with n = 3.

Similarly, loss of activity was explored using the same viral I-R3 substrates. Here, the critical nucleotide on the substrate strand was similarly replaced with an A (Figures 5(b) and 5(c)). A remarkable decrease in HParvoVa activity was observed when compared to Figure 3, supporting the hypothesis that a T-A enzyme–substrate base pair at the position SII is highly detrimental. The activities of the other sequences were also not improved, if not completely diminished, compared to those in Figure 3. However, it is worth noting that the activity levels of the WT sequences were low to begin with. In addition, the WT sequence for HCircVb already contained an A in the SII position of the substrate strand but was included with the other samples for consistency. The activities of WT virus sequences, along with C and A mutants, are summarized in Table 1.

Table 1. A comparison of cleavage activity without and with SII substitutions. Data in the estimated column comes from the output of the IR3 program; the wild-type viral sequences are represented in the WT column; C represents the results of viruses with a C mutation, leading to a G-C enzyme–substrate base pair at the SII position; and the A column represents the results of viruses with a T mutation, leading to a T-A enzyme–substrate base pair at the SII position. The error values reported for the experimental data are the standard error of the mean with n = 3.
Sample Estimated WT C A
HCircVa 1.00 0.07 ± 0.01 1.05 ± 0.01 0.06 ± 0.03
HCircVb 0.56 0.00 ± 0.00 1.04 ± 0.02 0.00 ± 0.00
HCircVc 0.13 0.00 ± 0.00 0.59 ± 0.08 0.00 ± 0.00
HBocaVa 0.56 0.00 ± 0.00 0.70 ± 0.03 0.03 ± 0.00
HBocaVb 0.06 0.01 ± 0.01 1.06 ± 0.03 0.00 ± 0.00
HBocaVc 0.56 0.07 ± 0.04 0.76 ± 0.06 0.00 ± 0.00
HParvoVa 1.00 0.82 ± 0.04 1.06 ± 0.04 0.12 ± 0.02
HParvoVb 0.09 0.08 ± 0.02 0.96 ± 0.04 0.00 ± 0.00

4. Conclusions

We developed a computer program, IR3, that scans the sequence of any given ssDNA strand, identifies all potential target sites, and outputs the most promising I-R3 DNA enzyme for every target site.

Using this program and comparing its predictions to the results of wet-lab experiments (that we carried out), we were able to show that the existing catalytic core model for I-R3 DNA enzyme activity is incomplete. Furthermore, our work demonstrates that the identity of a certain base pair, termed SII, at the immediate periphery of the catalytic core has a substantial effect on cleavage activity. A purineenzyme–pyrimidinesubstrate base pair at SII substantially increases cleavage activity, while the reversed base pair yields minimal activity.

Given these findings, it is recommended that one employs a purine (preferably a G) at the DNA enzyme side of the SII base pair, if one wishes to maximize cleavage activity of an I-R3 DNA enzyme. We believe that this aspect of the DNA enzyme’s activity is key in its use in future applications, as it expands the variety of available target sequences. Potential applications of I-R3-based systems include the targeting and cleavage of various single-stranded DNA viruses, such as parvoviruses, in therapeutic or biosensing contexts. They can also serve roles in dynamic DNA systems, such as recycling-based amplification systems in aptamer-based small-molecule biosensors. The latest version of the program includes this recommendation in its output. The IR3 program is free and publicly available for download at https://github.com/XinxinTree/IR3.

Conflicts of Interest

The authors declare no conflicts of interest.

Author Contributions

Shahidul Islam and Gabriel Aguiar-Tawil contributed equally.

Funding

This study was funded by Concordia University’s Applied AI Institute under grant number 300010761.

Acknowledgments

Thanks are due to Phylicia Ma, former student at High Technology High School, Middletown, New Jersey, who performed initial relative rate compilation during a summer internship sponsored by the Monmouth University Summer Research Program.

    Data Availability Statement

    The data that supports the findings of this study are available in the supporting information of this article (supporting data).

      The full text of this article hosted at iucr.org is unavailable due to technical difficulties.