Part:BBa_K5334019

CU2#19 aptamer for Bevacizumab

Notice: Functional DNA

This part is a sequence of a functional ssDNA. It is only active as single-stranded DNA. It can not be cloned into a plasmid. For use order it as a DNA oligo.

Aptamers are single-stranded nucleic acid ligands with high affinity and specificity to target molecules[1]. This is an anti-idiotype aptamer (Figure 1) that recognizes the CDRs of bevacizumab, an anti-vascular endothelial growth factor (VEGF) mAb, which is used in lateral-flow detection system for bevacizumab. The aptasensor allows the detection of bevacizumab in anti-cancer drugs (Avastin) in a wide concentration range, comprised between 0.05 and 5.0 μg/mL, with the limit of detection was 2.09 ng/mL

Usage and Biology

Aptamers, which include RNA, single-stranded DNA (ssDNA), and peptide molecules, exhibit high target specificity and binding strength due to their distinct three-dimensional structures. Research on the binding of nucleic acids with proteins started in the 1980s, largely from studies on HIV and adenovirus, where small structured RNAs were found to bind viral or cellular proteins with precision [6]. For example, the HIV TAR element RNA binds with the viral Tat protein, while adenovirus VA-RNA controls translation [7]. The major breakthrough in aptamer technology came with the 1990 development of SELEX (Systematic Evolution of Ligands by EXponential enrichment), which made it possible to select aptamers in vitro for a range of targets, from small molecules to cells [8].

Since then, aptamers have been explored in fields like diagnostics, therapeutics, biosensors, and drug delivery [8]. They offer advantages over traditional antibodies, such as being highly stable at elevated temperatures, simpler and more cost-effective to produce, and less likely to trigger immune responses [9]. For instance, Macugen, an aptamer targeting VEGF, was approved by the FDA in 2004 for the treatment of wet age-related macular degeneration [9], marking a key milestone for aptamer applications. Given their versatility and superior characteristics, aptamers are seen as a viable alternative to antibodies across various biological and medical fields.

Source of the part

Given the project timeline and efficacy constraints, the SELEX process was deemed infeasible. Consequently, we opted to model an existing aptamer with a demonstrated high affinity for bevacizumab: 5′-GCGGTTGGTGGTAGTTACGTTCGC-3′ [5]. Utilizing the structural ideogram of the previously characterized aptamer [3], we aimed to engineer interactions involving residues that were not previously engaged in antibody binding: T5, T6, G7, G8, T9, G10, G11, T12, A13, G14, T15, T16, A17, C18, G19, T20, T21.

Modeling

Docking

2D structure

Consequently, we opted to model an existing aptamer with a demonstrated high affinity for bevacizumab: 5′-GCGGTTGGTGGTAGTTACGTTCGC-3′[5]. Utilizing the structural ideogram of the previously characterized aptamer A14#1(BBa_K5334000)[3], we aimed to engineer interactions involving T20 via changing to G20 (5′-GCGGTTGGTGGTAGTTACGGTCGC-3′-aptamer CU2#19) due to T20 not previously engaged in antibody binding. The DNA secondary structures were predicted using the RNAfold web server [16] with the default parameters (Figure 1). Energy parameters were rescaled to 37 °C and 1.021 M salt concentration. The color represents the base pairing/unpairing probabilities with red being the most plausible.

Figure 1 | 2D structure of CU2#19 aptamer.

3D structure

The coordinates of the initial complex (7V5N, 1.70 Å) were taken from the Protein Data Bank [3]. To predict aptamer coordinates 3dDNA web server was used for folding [10] (Figure 2 2).

Figure 2 | 3D structure of CU2#19 aptamer and bevacizumab. .

For virtual screening, an approach from [11] was adopted. Normalized results of 3 different models (Autodock Vina, PatchDock, and Hex), were used. Smina [12] which is a modification of Autodock vina used for docking as it is capable of setting up the simulation box automatically according to the initial ligand position. The grid box was chosen with 8 Å buffer space in each direction to ensure that different folds would fit. The exhaustiveness was 4000 and 10 best docking poses were taken for further evaluation. Hex 8.0 [13] was used with default parameters and consequent minimization using the OPLS force field. PatchDock [14] docking was performed locally with 4.0 Å RMSD clustering threshold. Therefore, we adopted the methodology outlined by Ahirwar et al.[15] For binding affinity predictions, we employed three distinct docking platforms: Autodock Vina, PatchDock, and Hex (Table 1), and took a total score as characteristic of binding.

Table 1. | Z-score of aptamers from three different docking platforms (Autodock Vina, PatchDock, and Hex) and their cumulative sum for aptamers A14#1 and CU2#19.

#	Subsequence	AutoDock Vina	PatchDock	Hex	Sum
A14#1	5′-GCGGTTGGTGGTAGTTACGTTCGC-3′	-2.224	-1.280	-1.814	-5,318
CU2#19	5′-GCGGTTGGTGGTAGTTACGGTCGC-3′	-2.674	-1.240	-1.692	-5.607

Aptamer CU2#19 showed better results than initial aptamer A14#1. As a result, aptamer CU2#19 was taken for calculation in MD.

Molecular Dynamics

We initially considered the total binding score from docking as a primary characteristic of molecular dynamics (MD). MD simulations of protein-DNA complexes were performed in GROMACS 2021.1 software [16]. The system was described by the amber14sb_ol21 force field [17]. The initial structure was solvated in a rectangular box with a 10 Å margin of TIP3P water. Na+ and Cl- ions were added up to 0.15 M concentration. Non-bonded interactions were computed with a cutoff distance of 10 Å, while long-range electrostatics were addressed using the Particle-Mesh Ewald method. The SHAKE algorithm was employed to constrain all bonds involving hydrogen during the simulations. The time step used was 2 fs. The system was minimized for 8000 steps using the steepest descent algorithm to avoid steric clashes. Then it was gradually heated up from 0 to 300 K in the NVT ensemble during 1 ns and equilibrated at 1 bar and 300 K in NPT for the same time applying constraints on the protein-aptamer complex. The production run was performed for 200 ns and the last 150 ns were used for analysis. The protein-aptamer binding affinity was assessed using MM/PBSA approach implemented in gmx_MMPBSA software [18] which allows using the original MMPBSA.py script implemented in AmberTools in GROMACS. 50 frames were extracted from each of the trajectories for analysis time averaged values and their standard deviation were an estimate of the binding energetics. Subsequently, we employed aptamers with enhanced binding affinity compared to existing ones in GROMACS to determine the time-averaged total energy (Table 2)[19,20]. The P-value was calculated based on the average MM/PBSA, standard deviation (SD), and a number of frames via a two-tailed student's t-test.

Table 1. | Z-score of aptamers from three different docking platforms (Autodock Vina, PatchDock, and Hex) and their cumulative sum for aptamers A14#1 (BBa_K5334000) and CU2#19.

#	Subsequence	MM/PBSA	Standard deviation	p-value
A14#1	5′-GCGGTTGGTGGTAGTTACGTTCGC-3′	-142.260	7.150	-
CU2#19	5′-GCGGTTGGTGGTAGTTACGGTCGC-3′	-146.170	6.750	p<0.01

The molecular dynamics results indicated that the initial aptamer A14#1 exhibits a lower affinity for bevacizumab compared to aptamer CU2#19 with p<0.01. Results showed that aptamer CU2#19 formed hydrogen bonds with T28, N31, G33, N35, N52, H101, and Y103 of bevacizumab (Figure 3). Consequently, for wet-lab experiments and the development of the LFA device, the decision was made to focus on aptamer CU2#19 as a main one.

Figure 3 | Detailed interaction between aptamer CU2#19 and bevacizumab.

Characterisation

MicroScale Thermophoresis (MST)

The dissociation constant Kd for aptamer CU2#19 was determined to be 10.4 ± 1.5 nM under pH 6.3, mean standard deviations were calculated based on at least three independent experiments (Fig. 4). Although the Kd of aptamer CU2#19 is better than that of the А14#1, the difference is not statistically significant. Therefore, we decided to use ELONA as another method to evaluate the binding of the aptamers to bevacizumab [21,22].

Figure 4 | (A) Relative Fluorescence vs time was determined for 16 capillaries with investigating solutions. (B) Fraction Bound vs bevacizumab concentration of 50 nM of FAM-labeled aptamer CU2#19 and 1250 to 0.153 nM of bevacizumab.

In our MST assays, this unspecific binding was a significant factor contributing to fluctuations in binding measurements for the aptamers. Such variability resulted in false positives and reduced assay sensitivity, complicating the interpretation of experimental results [23]. To mitigate these effects and evaluate Kd correctly, we optimized the assay, made independent experiments, and excluded results that were outliers. This approach is consistent with the paper of Chen et al., who demonstrated similar optimization by presenting only 8 data points on a graph using a machine with 16 capillaries [24].

Enzyme-Linked Oligonucleotide Assay (ELONA)

In our results, we observed that DNA aptamers, unlike monoclonal antibodies (mAbs) commonly used in ELISA methods, demonstrate high stability against factors such as heat, organic solvents, and acids. The chemical modification of the DNA aptamers on the microplate surface remained stable as well. By maintaining proper washing protocols and preserving the three-dimensional structure of the aptamers, strong linearity of the calibration curves was achieved (r² > 0.9847)(Fig.5).

Figure 5 | Calibration curve of colorimetric ELONA for bevacizumab samples.

After obtaining data from the initial aptamer A14#1, we repeated the procedure with the optimized aptamer CU2#19. The optimizations were aimed at improving its binding affinity and specificity through molecular docking and dynamics simulations. The graph below (Fig. 6) presents the results of the ELONA assay, demonstrating the enhanced performance of CU2#19 in detecting Bevacizumab, showing improved sensitivity and precision across the tested concentration range.

Figure 6 | Lower limit of detection of Bevacizumab in samples (CU2#19)

The graph for A14#1 (Fig. 6) shows a relatively linear trend but demonstrates a limited dynamic range and sensitivity across the tested Bevacizumab concentrations. The normalized absorbance values for A14#1 exhibit larger error bars, indicating higher variability between replicates, and a less distinct trend in response to increasing Bevacizumab concentrations. The confidence interval is also wider, which suggests lower precision and possibly weaker binding interaction at lower concentrations.

Figure 6 | Lower limit of detection of Bevacizumab in samples (A14#1)

In contrast, the graph for CU2#19 shows a significant improvement in both the trendline and overall sensitivity. The normalized absorbance values for CU2#19 increase more sharply with increasing Bevacizumab concentrations, demonstrating better discrimination between lower and higher concentrations. Additionally, the trendline is more consistent, and the error bars are smaller, reflecting improved reproducibility and tighter binding between the aptamer and the target.

The 95% confidence interval for CU2#19 is also narrower than for A14#1, indicating higher precision in the measurements. CU2#19’s improved detection limits and reduced variability suggest that the molecular docking and dynamics optimizations successfully enhanced the aptamer's binding affinity and specificity, particularly at lower Bevacizumab concentrations. The improvements observed with CU2#19 in both sensitivity and precision indicate that the optimizations applied to A14#1 were effective. CU2#19 demonstrates superior performance in detecting Bevacizumab across the concentration range, making it a more reliable and sensitive tool for this application. This result highlights the potential of using in silico techniques, such as molecular docking and dynamics, to improve aptamer design for enhanced biosensor performance.

Statistical analysis. Statistical analysis was performed with unpaired two-tailed t-test using GraphPad Prism Software (GraphPad Software Inc., CA) with p < 0.05 considered as significant (* p < 0.05). Error bars represent standard error of the mean (SEM) of at least three independent reproducible biological experiments (Fig. 7).

Figure 7 | Normalized Absorbance for A14#1 and CU2#2 for concentration of Bevacizumab 1 μg/ml. Mean±SEM. (* p<0.05; two-tailed student's t-test).

References

[1] Ellington AD, Szostak JW. In vitro selection of RNA molecules that bind specific ligands. Nature, 346(6287):818-22, (1990).

[2] Nonaka, Y., Sode, K., & Ikebukuro, K. Screening and Improvement of an Anti-VEGF DNA Aptamer. Molecules, 15(1), 215–225, (2010).

[3] Saito, T. et al. Development of a DNA aptamer that binds to the complementarity-determining region of therapeutic monoclonal antibody and affinity improvement induced by pH-change for sensitive detection. Biosens Bioelectron 203, 114027 (2022).

[4] Hasegawa, H., Sode, K., & Ikebukuro, K. Selection of DNA aptamers against VEGF165 using a protein competitor and the aptamer blotting method. Biotechnology Letters, 30(5), 829–834, (2008).

[5] Yamada, T., Saito, T., Hill, Y., Shimizu, Y., Tsukakoshi, K., Mizuno, H., … Todoroki, K. (2019). High-throughput bioanalysis of bevacizumab in human plasma based on enzyme-linked aptamer assay using anti-idiotype DNA aptamer. Analytical Chemistry (2019).

[6] Dollins CM, Nair S, Sullenger BA. Aptamers in immunotherapy. Hum Gene Ther, 19(5):443-50, (2008).

[7] Sullenger BA, Gallardo HF, Ungers GE, Gilboa E. Overexpression of TAR sequences renders cells resistant to human immunodeficiency virus replication. Cell. (1990)

[8] Han K, Liang Z, Zhou N. Design strategies for aptamer-based biosensors. Sensors (Basel). 10(5):4541-57, (2010).

[9] Mascini M. Aptamers and their applications. Anal Bioanal Chem;390(4):987-8, (2008).

[10] Zhang, Y., Xiong, Y. & Xiao, Y. 3dDNA: A Computational Method of Building DNA 3D Structures. Molecules 27, 5936 (2022).

[11] Ahirwar, R. et al. In silico selection of an aptamer to estrogen receptor alpha using computational docking employing estrogen response elements as aptamer-alike molecules. Sci Rep 6, 21285 (2016).

[12] Koes, D. R., Baumgartner, M. P. & Camacho, C. J. Lessons Learned in Empirical Scoring with smina from the CSAR 2011 Benchmarking Exercise. J Chem Inf Model 53, 1893–1904 (2013).

[13] Ghoorah, A. W., Devignes, M., Smaïl‐Tabbone, M. & Ritchie, D. W. Protein docking using case‐based reasoning. Proteins: Structure, Function, and Bioinformatics 81, 2150–2158 (2013).

[14] Duhovny, D., Nussinov, R. & Wolfson, H. J. Efficient Unbound Docking of Rigid Molecules. in 185–200 (2002).

[15] Ahirwar, R. et al. In silico selection of an aptamer to estrogen receptor alpha using computational docking employing estrogen response elements as aptamer-alike molecules. Sci Rep 6, 21285 (2016).

[16] Burford, A. et al. Ookami: Deployment and Initial Experiences. in ACM International Conference Proceeding Series (Association for Computing Machinery, 2021).

[17] Zgarbová, M., Luque, F. J., Šponer, J., Otyepka, M. & Jurečka, P. A Novel Approach for Deriving Force Field Torsion Angle Parameters Accounting for Conformation-Dependent Solvation Effects. J Chem Theory Comput 8, 3232–3242 (2012).

[18] Valdés-Tresanco, M. S., Valdés-Tresanco, M. E., Valiente, P. A. & Moreno, E. gmx_MMPBSA: A New Tool to Perform End-State Free Energy Calculations with GROMACS. J Chem Theory Comput 17, 6281–6291 (2021).

[19] Martin, D. R. et al. Molecular modeling and simulation studies of SELEX-derived high-affinity DNA aptamers to the Ebola virus nucleoprotein. J Biomol Struct Dyn 1–18 (2024)

[20] Sabri, M. Z., Abdul Hamid, A. A., Sayed Hitam, S. M. & Abdul Rahim, Mohd. Z. In Silico Screening of Aptamers Configuration against Hepatitis B Surface Antigen. Adv Bioinformatics 2019, 1–12 (2019).

[21] Toh, S. Y., Citartan, M., Gopinath, S. C. B. & Tang, T.-H. Aptamers as a replacement for antibodies in enzyme-linked immunosorbent assay. Biosens Bioelectron 64, 392–403 (2015).

[22] Song, S., Wang, L., Li, J., Fan, C. & Zhao, J. Aptamer-based biosensors. TrAC Trends in Analytical Chemistry 27, 108–117 (2008).

[23] Yoo, H., Jo, H. & Oh, S. S. Detection and beyond: challenges and advances in aptamer-based biosensors. Mater Adv 1, 2663–2687 (2020).

[24] Chen, K. et al. Aptamers as Versatile Molecular Tools for Antibody Production Monitoring and Quality Control. J Am Chem Soc 142, 12079–12086 (2020).

Sequence and Features

Assembly Compatibility:

• 10
  COMPATIBLE WITH RFC[10]
• 12
  COMPATIBLE WITH RFC[12]
• 21
  COMPATIBLE WITH RFC[21]
• 23
  COMPATIBLE WITH RFC[23]
• 25
  COMPATIBLE WITH RFC[25]
• 1000
  COMPATIBLE WITH RFC[1000]