ATP-Aptamer or ATP-Aptamer Mutant Complexes by

Subscriber access provided by University of South Dakota

B: Biophysics; Physical Chemistry of Biological Systems and Biomolecules

Probing ATP/ATP-Aptamer or ATP-Aptamer Mutant Complexes by Microscale Thermophoresis and Molecular Dynamic Simulations: Discovery of an ATP-Aptamer Sequence of Superior Binding Properties Yonatan Biniuri, Bauke Albada, and Itamar Willner J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.8b06802 • Publication Date (Web): 06 Sep 2018 Downloaded from http://pubs.acs.org on September 7, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Probing

ATP/ATP-Aptamer

or

ATP-Aptamer

Mutant Complexes by Microscale Thermophoresis and Molecular Dynamic Simulations: Discovery of an ATP-Aptamer Sequence of Superior Binding Properties Yonatan Biniuri†, Bauke Albada‡ and Itamar Willner* †

† Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem, 91904, Israel ‡ Laboratory of Organic Chemistry, Wageningen University & Research, Wageningen, The Netherlands

*E-mail: [email protected]. Tel: 972-2-6585272. Fax: 972-2-6527715.

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 31

ABSTRACT

Micro Scale Thermophoresis (MST) is used to follow the dissociation constants corresponding to ATTO 488-labeled ATP and the ATP-aptamer or ATP-aptamer mutants that include two binding sites for the ATP ligand. A set of eight ATP-aptamer mutants, where the thymidine bases within the reported ATP binding aptamer sites are substituted with cytosine bases are examined. The MST derived dissociation constant of ATP to the reported aptamer is Kd = 31 ± 3 µM, while most of the aptamer mutants show lower affinity (higher Kd values) toward the ATP ligand. One aptamer mutant reveals, however, a higher affinity toward the ATP ligand, as compared to the reported ATP-aptamer. Molecular dynamics and docking simulations identify the structural features that control the affinities of binding of the ATP ligand to the two binding sites associated with the ATP-aptamer or the ATP-aptamer mutants. The simulated structures suggest that H-bonds between the ATP ligand and G9 and G11 bases within one binding domain, and the π-π interactions between G6 and the ATP purine moiety and the pyrimidine ring, in the second binding domain, control the affinity of binding interactions between the ATP ligand and the ATP-aptamer or ATP-aptamer mutants. Very good correlation between the computed Kd values and the MST-derived Kd values is found. The ATP-aptamer mutants (consisting of T4→C, A24→G, T12→C, T14→C and T27→C mutations) reveals superior binding affinities towards the ATP ligands (Kd = 15 ± 1 µM) as compared to the binding affinity of ATP to the reported aptamer. These features of the mutant are supported by molecular dynamics simulations.


2

Page 3 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Micro Scale Thermophoresis (MST) represents a rapidly developing physical method to probe shapes of macromolecules or ligand-receptor dissociation constants.1,2 The MST monitors the fluorescence changes of fluorophore-labeled molecules upon the generation of a temperature gradient in a µm-sized spot. In a typical experiment, an IR-laser generates a temperature gradient of ca. 4 ˚C between the irradiated hot-spot and the bulk environment. The temperature gradient leads to the movement of the labeled molecules from the hot spot to the bulk solution, or to the movement of the molecules from the bulk to the hot-spot (depending on the Soret coefficient of the moving species), and these microscale movements are followed by probing the timedependent fluorescence changes in the hot-spot.3,4 Typical thermophoresis-induced curves corresponding to the time-dependent fluorescence changes are shown in Figure 1(A). Curve (a) shows the time-dependent fluorescence changes of the labeled component. Application of the IR laser yields the temperature gradient and the time-depletion of the fluorescent agent, due to its migration to the bulk. This migration reaches a saturation value when the thermophoretic force is counter-balanced by the Brownian diffusion of the molecule in solution. Upon switching-off the IR laser (time marked with an arrow), the original equilibration of the contents of the irradiated spots is regenerated. The thermophoretic-induced migration of the fluorescent label is controlled by different parameters, such as the mass and shape of the moving fluorescent unit, and environmental effects, such as viscosity or ionic strength of the medium. For example, a binding event between the fluorophore-labeled ligand and a receptor, alters the mass of the moving constituent and this is reflected by changes in the thermophoretic time-dependent fluorescence curve that is controlled by the concentrations of the ligand/receptor complexes.5 As


3


Page 4 of 31

the concentration of the receptor increases, the thermophoretic migration of the resulting complex changes, and the difference between the saturated fluorescence in the absence and presence of the receptor provides a quantitative value for the bound ligand-receptor complex, e.g., curves (b) and (c). By monitoring the normalized fluorescence value, Fnorm, as a function of the ligand concentrations, Figure 1(B), the dissociation constant is derived by applying the law of mass action, eq. 1. In our case, where two ligands bind to the aptamer, the experimental curve, Figure 1(B) is fitted to eq. 2, (Hill equation), which yields the dissociation constant of the ligand/receptor complex. (For a detailed evaluation of eq. 2 see supporting information.) In eq. 1 and eq. 2, [B0] represents the total concentration of binding sites, [L0] stands for the concentration of ligand added at each data point, [BL] is the concentration of formed complex between the ligand and binding sites. Kd is the dissociation constant which describes the binding of the ligand to the receptor. For eq. 2, n stands for the Hill coefficient. Eq. 2 was used where [L] = [BL] approximately. Eq. 1

ሾ஻௅ሿ ሾ஻బ ሿ

= (ሾ‫ܮ‬଴ ሿ + ሾ‫ܤ‬଴ ሿ + ‫ܭ‬ௗ ) − ሾ஻௅ሿ

Eq. 2

ሾ஻బ ሿ

=

ඥ((ሾ௅బ ሿାሾ஻బ ሿା௄೏ )మ ିସ∗ሾ௅బ ሿ∗ሾ஻బ ሿ ଶሾ஻బ ሿ

ଵ

(಼ ೙ ቀଵା ሾಽሿ೏ ቁ

Aptamers are sequence-specific single-stranded nucleic acids that reveal selective binding properties toward low-molecular-weight or macromolecular ligands.6,7 The aptamers are elicited by the “systematic evolution of ligands by exponential amplification”, SELEX procedure.8,9 The binding affinities of aptamers were extensively used for the development of different sensing platforms10–12, nucleic acid-based catalysts13–15, DNA-based machines16–18, and stimuli-responsive drug carriers.19–21 In addition, aptamer-ligand complexes were used to trigger logic gate operations22,23, to assemble programmed DNA structures18,24, and to use the complexes for nanomedical applications.25,26


4

Page 5 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Microscale thermophoresis has been recently applied as a physical method to characterize structural changes of proteins1 and ligand-receptor binding interactions27, such as the binding of the inhibitor, quercetin, to a cAMP kinase.28 Specifically, Microscale thermophoresis was, also, used to follow the formation of ligand-aptamer complexes.5,29 In the present study, we apply microscale thermophoresis to follow the affinity binding interactions, (dissociation constants, Kd) between ATP and a series of mutants of the ATP-aptamer. We further apply molecular dynamic (MD) simulations to identify the minimum-energy structures of the different ATP/aptamer mutant complexes and their computed Kd values. We find very good correlation between the experimental Kd values and the MD simulations. We use the MD simulations to rationalize the parameters that control the binding features of the aptamer to the ATP ligands. Surprisingly, we find that one of the aptamer mutants reveals a higher binding affinity toward ATP, as compared to the reported binding affinity of ATP to the ATP-aptamer.5

EXPERIMENTAL SECTION Aptamer and aptamer mutant sequences The nucleic acid sequences used in the study were ordered from Integrated DNA Technologies (IDT), Israel, and were dissolved overnight at 20 mM HEPES buffer, pH 7.2, that included 200 mM NaCl and 1mM MgCl2. Sample concentrations were determined using a Biotek Take 3 quartz microdot carrier.


5


Page 6 of 31

Evaluation of the dissociation constants of the ATP-aptamer and ATP-aptamer mutants to Fl-ATP using MST A 1:1 serial dilution of 16 samples of the ATP-aptamer or mutant with 20 mM HEPES buffer, pH 7.2, that included 200 mM NaCl and 1mM MgCl2 was prepared, covering a concentration range between 1 mM to 33 nM of the aptamer. To 18 µl of each of the serial dilutions, 2 µl of 1.25 µM stock of ATTO-488 functionalized ATP (γ-(6-Aminohexyl)-ATP - ATTO-488, (Jena Bioscience GmbH, Jena, Germany) were added and allowed to incubate for 10 minutes at room temperature. Subsequently, each sample was loaded onto a Nanotemper (Munich, Germany) KM-022 capillary tube and mounted on a Nanotemper Monolith Nt.115 capillary tray. Each measurement was tested at an LED power of 20% at the “Blue” excitation wave-length setting, with a cold region of 5 seconds, followed by a hot region of 30 seconds, until the new hot equilibrium state was observed, followed by 5 more seconds of cold region to observe the regeneration of the fluorescence signal characteristic to the cold region value. For each MST measurement, each capillary was subjected to an IR-laser power at 20% or 40% of its power to record the resulting thermophoretic curves. Each of these measurements included three repeat experiments.

Evaluation of the dissociation constants of the ATTO-488-ATP-aptamer and ATTO-488ATP-aptamer mutants using MST A 1:1 serial dilution of 16 samples of ATP (Sigma-Aldrich) was made using 20 mM HEPES buffer, pH 7.2, that included 200 mM NaCl and 1mM MgCl2, covering a concentration range between 1 mM to 33 nM of Ligand. To 10 µl of each of the serial dilutions, 10 µl of 500 nM stock of ATTO-488 functionalized at the 5’ of the ATP-aptamer or ATP-aptamer mutant


6

Page 7 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(Integrated DNA Technologies (IDT), Israel) were added and allowed to equilibrate incubate for 10 minutes at room temperature. Subsequently, each sample was loaded onto a Nanotemper (Munich, Germany) KM-022 capillary tube and mounted on a Nanotemper Monolith Nt.115 capillary tray. Measurements were carried as described above. A minimum of 3 repeat experiments were performed for each aptamer or aptamer mutant.

RESULTS AND DISCUSSION In the present study we have examined the application of the Microscale Thermophoresis technique to examine the dissociation constants of the fluorophore (Atto-488)-modified ATP to the reported ATP-aptamer30 and to a series of mutated aptamer sequences. The ATP-aptamer (1) and the binding affinities of ATP to the aptamer were characterized by NMR spectroscopy.31 The structural features of the two ATP ligands that bind to the aptamer were elucidated in detail, Figure 2. The reported ATP-aptamer contains two similar binding sites for ATP, composed each of a three-nucleotide motif that includes a GA sequence and an oppositely-positioned G base31. Also, the ATP-aptamer sequence contains five thymidine bases: two Watson-Crick base pairing thymidine bases, T4 and T27 at the short duplex region of the aptamer, and three other thymidine residues that are not involved in the classical Watson-Crick base pairing, T12, T14 and T15, positioned at the hairpin region of the aptamer. Following the significance of the GA/G binding motifs in the aptamer binding sites, and realizing the potential cooperative stabilization of the complexes by the thymidine bases in the duplex region and the loop domain of the aptamer, we decided to probe the binding affinities of aptamer mutants that include thymidine base mutations in these domains. Accordingly, we examined the binding features of the fluorophore-modified ATP to the reported ATP-aptamer, and to the series of mutated aptamers (2)-(9) (Table 1). The


7


Page 8 of 31

mutated aptamers included the substitution of an individual thymidine unit associated with the binding domain of the aptamer, (2)-(6), where the bases mutated are as follows: (2), T14→C, (3), T15→C, (4), T12→C, (5), T4→C and A24→G and (6), T27→C and A1→G. A further mutant includes the substitution of one thymidine unit in the hairpin domain and two thymidine units associated with the short duplex structure of the aptamer complex, (7), T27→C, A24→G, T12→C, T4→C and A1→G. In addition, two mutants, (8), T27→C, A24→G, T14→C, T12→C, T4→C and A1→G and (9), T27→C, A24→G, T15→C, T12→C, T4→C and A1→G, that include the substitution of four thymidine units; e.g. two thymidine bases in the hairpin domain and two thymidine bases in the duplex structure of the ATP-aptamer complex, while preserving one of the thymidine residues in the hairpin domain. Note that, mutants (8) and (9) differ by the different thymidine units that were replaced in the ATP-aptamer’s hairpin region. By mutating the residues that surround the binding domain, which is formed by the GAGGG box, we probe whether changes in the sphere surrounding the ligand-binding domain, i.e. the secondary binding sphere, affect the binding properties of the aptamer. Figure 3(A) shows the microscale thermophoretic curves corresponding to the treatment of the fluorophore-modified ATP with variable concentrations of the reported ATP-aptamer (1). As the concentration of the ATP-aptamer is elevated, the changes in thermophoretic curve characteristics increase, implying the binding of the ATP ligand to the aptamer. Figure 3(B) shows the normalized fluorescence changes as a function of the logarithmic concentrations of the ATP-aptamer. Figure 3(C) exemplifies the thermophoretic curves corresponding to the treatment of the fluorophore-modified ATP with variable concentrations of the aptamer or aptamer mutant. The solid line represents the binding curve according to eq. 2, from which the fluorophoreATP/(1) complex dissociation constant, Kd = 31 ± 3 µM was derived. Evidently, the dissociation


8

Page 9 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


constant of the fluorophore-ATP/(2) is higher as compared to the fluorophore-ATP/(1) complex, consistent with the expectation that mutations in the aptamer sequence would lower the binding affinity of the fluorophore-ATP ligand to the mutated sequence.

Figure S1 to Figure S9,

supporting information, depict the microscale thermophoretic curves corresponding to the treatment of the fluorophore-modified ATP with variable concentrations of all other mutants. The respective normalized fluorescence changes as a function of the concentrations of the respective mutants are also provided.

Table 2 summarizes the dissociation constants, Kd,

corresponding to the binding of the fluorophore-modified ATP to the ATP-aptamer, (1), and the different mutated aptamers. From Table 2 we realize that the Kd values for the fluorophore-ATP complex with the ATP-aptamer (1) and aptamer mutants (2), (3), (4), (5), (6), (8) and (9) are similar or higher than the Kd value of the ligand with the reported aptamer (1), implying that the introduction of mutations decreases the affinity of binding of the fluorophore-ATP ligand to the mutated aptamers. It is, however surprising to note that the mutant (7) reveals a lower Kd value as compared to the binding of the fluorophore-ATP ligand to (1), implying higher binding affinity of the ligand to the mutated aptamer. It should be noted that the binding features of bare ATP to the ATP-aptamer are similar to those of the ATTO-488 functionalized ATP. For these measurements the aptamer (1) or mutant (7) were modified at their 5’-end with ATTO-488 and the dissociation constants of base ATP to the aptamers were evaluated by MST, Figure S10 and Figure S11. The dissociation constants of ATP to the fluorophore-functionalized aptamer (1) corresponds to 25 ± 5 µM, consistent with the literature values29, and similar to the Kd value of ATTO-488-ATP to (1). In addition, the Kd value of ATP to the ATTO-488-modifed mutant (7) corresponds to 14 ± 3 µM µM. These results further confirm that the mutant (7) exhibits higher affinity toward ATP, as compared to the binding affinity of ATP to the reported ATP-aptamer.


9


Page 10 of 31

Figure S12 depicts a simulated structure of the ATTO-488-modified-ATP-aptamer/ATP complex. It should be noted that the MST method is well-suited to derive the Kd values of the ATP ligand to the respective aptamers under different environmental conditions such as pH changes, addition of ions or ionic strength of the systems. For example, lowering the pH of the solution to pH 3.5 prohibits the formation of the complex between ATP and its aptamer, while elevating the pH of the solution to pH 5.5 yields a Kd value between ATP and its aptamer, similar to the Kd value of the complex at pH 7.2. Also, we find that in the absence of Mg2+ ions, no binding between ATP and its aptamer could be detected. Furthermore, we note that the MST method provides a useful technique to evaluate the Kd values between macromolecules (proteins) and their ligands. In the next step we performed molecular dynamics and docking simulations that follow the binding of ATP to the ATP-aptamer or to its mutants. Recent research efforts were directed to apply molecular dynamic simulations to identify the structures and binding affinities of ligandaptamer complexes.32,33 Among the molecular dynamic simulations programs, the YASARA software is especially attractive since it provides a toolbox to systematically and stepwise model the resulting structures and to evaluate the minimum energy configurations of biomaterials.34,35 In fact, the YASARA structure software is well-suited to follow supramolecular interactions and structures of nucleic acids. The general molecular dynamic simulation process of aptamer-ligand complexes involves the initial identification of experimentally suggested structure(s) of the aptamer-ligand complex (e.g., by 1H-NMR), followed by the docking of the ligand to the reported structure. After identifying a possible configuration of the aptamer-ligand complex, the free aptamer is subjected to a run of energy minimizing steps, followed by the docking of the ligand to the optimized aptamer structure. Following these steps, the program provides the


10

Page 11 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


dissociation constant and geometrical features of the aptamer/ligand complex(es). Indeed, the YASARA software package and procedures were successfully applied to derive simulated Kd values and suggested structures of arginineamide and its derivatives to the arginineamide binding aptamer36. The molecular dynamic simulations of the ATP-aptamer complex involve, however, increased complexity since two ATP ligands bind to two different domains of oligonucleotides associated with the aptamer.

Accordingly, the docking simulations included a two-step

interaction of two sequestered ATP ligands to the aptamer structures of ATP/ATP-aptamer ligand complex, using the following procedure: (i) The YASARA-built in AutoDock Lamarckian Genetic Algorithm (LGA) and the AMBER03 force field were applied to probe the docking of the ATP ligand to the set of 1H-NMR solved of PDB deposited structures of the aptamer, without any energy minimization. Out of 100 docking simulations on each of these structures, two ligand-aptamer clusters that included the ATP in the 1H-NMR predicted binding site (with an RMSD value of 5 Å) were identified, with a low rate of 5-7 hits out of the 100 simulations. It should be noted that other ligand clusters of the ligand/aptamer were identified in non-binding domains with a RMSD value of 5 Å and very low binding constants. These clusters were ignored, and only the two complexes that included the ligand in the appropriate 1H-NMR predicted binding site were used for the subsequent simulations and for the refinement of the simulations. The low reproduction hit rates of the LGA docking simulations is attributed to mismatches between the different force fields used to identify the 1H-NMR structures and the resulting docked configurations. (ii) The low-hit ATP-aptamer structures were subsequently subjected to a second round of docking, preceded by energy minimization simulation. In this step, the original aptamer structure was subjected to an energy minimization step followed by the docking of ATP to the energy minimized structure. This process resulted in an increased hit-rate


11


Page 12 of 31

of the ATP binding to the aptamer binding site that corresponded to 35-40%. (iii) The resulting ATP-aptamer complex was subjected to a dynamic energy minimization step that led to a slightly higher hit-rate (ca. + 5%) and a substantially lower dissociation constant.

The molecular

dynamic-simulated dissociation constant corresponds to Kd = 29 µM while the MST-derived dissociation constant for the ATTO 488-labeled ATP corresponded to Kd = 31 ± 3 µM. Figure 4 shows the optimized simulated structure of the ATP-aptamer complex. It should be noted that molecular dynamic simulation on the binding of ATTO 488-labeled ATP to the aptamer yielded similar results to the non-modified ATP. (For further discussion on the Kd values of ATP to its aptamer, vide infra.) Following these molecular dynamic simulations on the ATP/ATP-aptamer complex we performed similar simulations on the set of different aptamer mutants characterized by the MSTderived Kd values of ATP to the ATP-aptamer or ATP-aptamer mutants aimed to identify the respective Kd values of the resulting predicted structures by the molecular simulations. The procedure described for the molecular dynamic simulations of the ATP-aptamer complex was adapted to characterize the Kd values and the geometrical features of the set of the ATP-aptamer mutants. Table 2 summarizes the Kd values of ATP to the ATP-aptamer and the aptamer mutants derived by the molecular dynamic simulations and the MST experiments. The Kd values derived from the YASARA software package simulations follow the experimental trend of the Kd values derived from the MST measurements. That is, the aptamer mutants (2), (3), (4), (5), (6), (8) and (9) reveal similar or higher Kd values as compared to the Kd value of ATP to the reported ATP-aptamer. Interestingly, however, the molecular dynamics simulations predict a lower Kd value between ATP and the mutant (7) (higher binding affinity), consistent with the experimental MST derived Kd value. To account for the parameters that


12

Page 13 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


control the binding affinity of ATP to the reported ATP-aptamer and its mutant, we tried to identify characteristic structural features in the simulated ATP/ATP-aptamer complex or the ATP/ATP-aptamer mutant complexes (that are guided by 1H-NMR data) that could provide an insight to the parameters controlling the stabilities of the of the different complexes. Toward this goal, we identify one structural binding motif consisting of the three H-bonds between ATP and the reported aptamer, that includes two H-bonds between G9 and ATP and an additional single H-bond between G11 and ATP in one of the ATP binding sites associated with the aptamer, cf. Figure 5. H-bonds were, also, identified by 1H-NMR as key interactions that bind ATP to the ATP-aptamer’s binding site.31 The second apparent binding motif of ATP to the second ATPaptamer binding site involves π-π overlap interactions between ATP and G6 associated with the second site, cf. Figure 6. It should be noted that for the first binding site motif we find differences in the geometrical features of the H-bonds within the set of complexes. In turn, for the second binding motif, we find two unchanged H-bond interactions for binding ATP to the ATP-aptamer mutants, but we identify differences of the π-π interactions between the ATP ligand and the G6 base associated with the mutants. Figure 5(A) exemplifies the suggested simulated structures of ATP to the reported aptamer (1), ATP to the mutant (2), ATP to the mutant (4), and ATP to mutant (7). Table 3 shows the characteristic structural features of the three H-bonds in the different structures. These data provide the following possible conclusions: (i) In some of the mutants, the H-bond between G11 and ATP are depleted and provide one possible parameter for determining the binding affinity of the mutants (mutants (2), (3) and (4)). (ii) The angle between the two H-bonds generated between G9 and ATP is close to 180° for the ATP/ATP-aptamer and ATP/ATP mutant (7) and is substantially lower for the other ATP/ ATPaptamer mutants, Figure 5(B). As linear coupled H-bond interactions are energetically


13


Page 14 of 31

preferred37, the planarity of the H-bonds in ATP/ATP-aptamer and ATP/ATP mutant (7) contributes to the high affinity towards ATP. (Interestingly, the angles in the H-bonds in ATP/ATP mutant (7) is 171-179° as compared to 163-171° in the ATP/ATP-aptamer (1), consistent with the higher experimental binding affinity of ATP to mutant (7)). The π-π interactions that control, in our opinion, the binding of ATP to the second binding site, are shown in Figure 6. Since the simulated structures for ATP to the ATP-aptamer or of ATP to all other mutants indicate unchanged H-bond geometries between ATP and the nucleotide bases associated with second ATP binding site, we conclude that the H-bonds do not significantly contribute to the difference in affinities of ATP to the second ATP binding site of the ATPaptamer or ATP-aptamer mutants. We note, however, that the distance separating the π-π domains of G6 and ATP and the spatial orientation of the two π-backbones may contribute to the observed differences in the Kd values of ATP to the aptamer and its mutants. This is exemplified in Figure 6 that shows the simulated structures of ATP to the aptamer mutant (2), that reveals the lowest affinity towards ATP, Figure 6(A), and of the simulated structure of ATP to the aptamer mutant (7) that shows the highest affinity towards ATP, Figure 6(B). For the ATP/ATP-aptamer mutant (2), that reveals low binding affinity, the distance separating the π domains of ATP and G6 is 4.64 Å and the π-perimeters are disturbed by a twisting angle of 17°. In turn, the distance separating the π-π domains of ATP and G6 in ATP/ATP mutant (7) corresponds to 3.59 Å, and the overlap between the π array is higher (a twist angle of ca. 10°), resulting in a higher binding affinity of ATP to the aptamer mutant (7). It should be noted that the YASARA structure software package allows to perform MD simulations at different pH values, and variable salt concentrations, thus allowing the correlation between experimental Kd values and computed results under different conditions.


14

Page 15 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


CONCLUSIONS In conclusion, in the present study we applied microscale thermophoresis to characterize the dissociation constants between ATTO-488-functionalized ATP and the ATP binding aptamer and a series of ATP-aptamer mutants. The dissociation constants of the fluorophore functionalized ATP/ATP-aptamer complex are in the range of the Kd values derived by other physical methods, for the ATP-ligand/aptamer complex. As expected, seven out of eight aptamer mutants reveal similar or higher Kd values for the resulting fluorophore-ATP/ATP-aptamer mutant complexes (similar or lower affinity of the ligand to the aptamer mutant). Surprisingly, however, one of the mutants revealed a lower dissociation constant of the fluorophore-modifiedATP to the mutant, as compared to the Kd of ATP to the reported aptamer sequence. To further understand the effects controlling the Kd values of the fluorophore-modified ATP to the ATPaptamer, we performed molecular dynamics and docking simulations that provide energeticallystabilized structures between the ATP ligand and its aptamer (or aptamer mutants). The computationally simulated structures suggest that two motifs may control the stability of the resulting complexes, upon binding of two ATP-ligands to the binding sites associated with the aptamers: (i) Two H-bonds, and particularly the angle separating the two H-bonds generated by guanosine 9 (G9) and the purine’s pyrimidine ring associated with the ATP or fluorophoremodified ATP ligands control the binding site of ATP to one binding site. (ii) The π-π complexes generated between G6 of the ATP-aptamer and the pyrimidine ring of the ATP ligand, and particularly the distances separating the two π-systems and the relative spatial overlaps of the two π networks control the binding of ATP to the second binding site. The ATP/ATP mutant (7) complex revealed a lower Kd value, (Kd = 15 ± 1 µM, higher affinity) as compared to the


15


Page 16 of 31

ATP/ATP-aptamer complex. Consistently we identify shorter π-π interactions and a smaller bending angle in hydrogen bonds formed between the ATP ligand and ATP-aptamer mutant (7), as composed to the values of these regulating parameters in the ATP/ATP-aptamer complex. The results demonstrate that MST is a rapid and effective technique to elucidate binding constants between ligands and their aptamers. Furthermore, the study demonstrated that point mutations on structurally-defined aptamers may lead to superior binding affinities of the ligands to the aptamer mutants. The lack of capturing of the mutated aptamer during the primary selection process of the non-existent sequence of the mutant in the primary selection library might be the origin for not eliciting the preferred sequence. In addition, we highlight the application of the YASARA software suite as a powerful computational method to construct energetically-stabilized ligands/aptamer complexes. The results suggest that computational simulations of point mutations of structurally-defined ligand/aptamer complexes may lead to aptamer sequences exhibiting superior affinities towards their ligands. Such predictive simulations could then guide the experimentalists for designing improved aptamers. Nonetheless, in order to generate meaningful seed structures of the ligand-aptamer complexes for the molecular dynamic simulations, it is essential to have an insight for the dominant bases of the aptamer that interact with the bound ligand. Such knowledge (that can be obtained by NMR studies) leads to high-yield seed structures for the simulations and provide guidelines for the possible mutation of bases in the aptamer binding domains.


16

Page 17 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. (A) Typical MST curves corresponding to an unbound fluorophore ligand, (a), a partially bound fluorophore-labeled ligand, (b), and a saturated fluorophore-labeled ligand to its respectable receptor, (c). (B) Analysis of the thermophoretic curves in terms of normalized fluorescence vs. the concentration of ligand.


17


Page 18 of 31

Figure 2. Schematic presentation of the ATP aptamer structure derived from the respective NMR studies (adapted with permission from ref. 31). The structure includes the two binding sites of the aptamer, the enlarged binding domains of the H-bond stabilized binding domain I, and the π-π stabilized binding domain II. Inset: Schematic diagram displaying the GA/counter G domains associated with two binding sites of the aptamer.


18

Page 19 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. (A) Representative MST curves corresponding to: (a) ATTO-488 functionalized ATP only. (b) ATTO-488 functionalized ATP complex with an intermediary concentration of the ATP-aptamer, 30 µM. (c) ATTO-488 functionalized ATP in the presence of a relatively high concentration of the ATP-aptamer, 1 mM. Fluorescence between blue lines represents Fcold while the fluorescent signal between the red lines represents Fhot. Fnorm is the normalized fluorescence signal represented by Fnorm = Fhot/Fcold. (B) Normalized fluorescence values of the MST measurements of the ATTO-488 functionalized ATP in the presence of variable concentrations of ATP-aptamer. (C) Binding curves representing the occupancy of the aptamer receptor with ATTO-488 functionalized ATP as a function of the ATP-aptamer or ATP-aptamer mutants: (a) Reported aptamer binding curve, (b), Aptamer mutant (2), (c), Aptamer mutant (7). The derived


19


Page 20 of 31

Kd values correspond to the 50% occupancy of the respective aptamers. For all MST curves corresponding to all aptamers, their normalized fluorescence curves and the respective binding curves, see supporting information.

Figure 4. Energy minimized structure of the ATP/ATP-aptamer complex. The structure was derived using the NMR elucidated seed structure followed by molecular dynamics and docking simulations (see details in text).


20

Page 21 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. (A) Molecular dynamic simulated structures depicting the H-bonds stabilizing the ATP/ATP-aptamer or aptamer mutant complexes. For the complexes between ATP and aptamer mutant (7) and the ATP/ATP-aptamer complexes, three H-bonds are formed (2xG9 and 1xG11). For the complexes generated between ATP/ATP mutant (1) and ATP/ATP mutant (4) only two H-bonds are formed (2xG9). (B) Schematic description of the H-bond donor-acceptor angle corresponding with ATP mutant (7) (violet), ATP-aptamer (1) (orange) and ATP mutant (2) (indigo).


21


Page 22 of 31

Figure 6. Molecular dynamics simulated structures displaying the π-π interactions (distances) between G6 and the ATP ligand corresponding to: (A) ATP/ATP mutant (7), (B) the ATP/ATPaptamer (1) complex.


22

Page 23 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Names and nucleic acid sequences of the different ATP-aptamer mutants and ATPaptamer a. Sequence Name

Sequence

ATP-aptamer (1)

ACC TGG GGG AGT ATT GCG GAG GAA GGT

ATP-aptamer mutant (2)

ACC TGG GGG AGT ACT GCG GAG GAA GGT


ACC TGG GGG AGT ATC GCG GAG GAA GGT


ACC TGG GGG AGC ATT GCG GAG GAA GGT


ACC CGG GGG AGT ATT GCG GAG GAG GGT


GCC TGG GGG AGT ATT GCG GAG GAA GGC


GCC CGG GGG AGC ATT GCG GAG GAG GGC


GCC CGG GGG AGC ACT GCG GAG GAG GGC


GCC CGG GGG AGC ATC GCG GAG GAG GGC

a

Underlined bases represent point mutations in the reported ATP-aptamer sequence (1).


23


Page 24 of 31

Table 2. Dissociation constants of the ATP/ATP-aptamer or ATP/ATP-aptamer mutants evaluated by MST and complementary molecular dynamics and docking simulations. ATP Complex

MST Kd(µM)

AutoDock LGA Kd (µM)

Simulated Binding energy (Kcal/mole)

ATP-aptamer (1)

31 ± 3

29

6.228


76 ± 2

155

5.699


62 ± 4

98

5.818


71 ± 3

80

5.736


35 ± 1

30

6.147


25 ± 5

26

6.469


15 ± 1

18

6.661


26 ± 2

27

6.406


33 ± 1

36

6.186


24

Page 25 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 3. Simulated H-bond parameters between the ATP ligand at site (2) and the aptamer or aptamer mutant bases. Mutant

G9-H1

G9-H2

G11-H3

(Å)

Angle (°) (Å)

Angle (°) (Å)

2.23

169

2.05

171

1.63

ATP-aptamer mutant (2) 2.02

159

1.93

160

-


160

1.96

173

1.96


144

1.97

162

-


168

1.97

174

-


171

2.22

170

1.83


172

2.05

179

1.70


170

1.96

168

1.80


167

1.94

178

1.62

ATP-aptamer (1)

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Tel: 972-2-6585272. Fax: 972-2-6527715. ACKNOWLEDGMENT We thank Prof. G. Haran and H. Mazal from the Weizmann Institute of Science, Rehovot, Israel, for their assistance in the initial MST experiments. Funding This research is supported by the Volkswagen Foundation, Germany, and the Minerva Center for Complex Bio-hybrid Systems.


25


Page 26 of 31

ASSOCIATED CONTENT Supporting Information. Fluorescence capillary scan values of ATTO-488 functionalized ATP in different ATP aptamer or aptamer mutant concentrations, Thermophoresis curves of aptamer and aptamer mutants, Fnorm values as a function of ATP-aptamer or aptamer mutants concentration, Thermophoresis curves of ATTO-488 labeled aptamer and aptamer mutant (7), Simulated structures of the ATP-aptamer/γ-(6-Aminohexyl)-ATP - ATTO-488. This information is available free of charge via the Internet at http://pubs.acs.org.

REFERENCES

(1)

Breitsprecher, D.; Schlinck, N.; Witte, D.; Duhr, S.; Baaske, P.; Schubert, T. Nucleic

Acid Aptamers. Methods in Molecular Biology; Mayer G.; Humana Press, New York, 2016, Vol. 1380, pp 99-111. (2)

Braun, D.; Libchaber, A. Trapping of DNA by Thermophoretic Depletion and

Convection. Phys. Rev. Lett. 2002, 89, 2–5. (3)

Duhr, S.; Braun, D. Thermophoretic Depletion Follows Boltzmann Distribution. Phys.

Rev. Lett. 2006, 96, 168301. (4)

Duhr, S.; Braun, D. Why Molecules Move Along a Temperature Gradient. Proc. Natl.

Acad. Sci. 2006, 103, 19678–19682.


26

Page 27 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(5)

Baaske, P.; Wienken, C. J.; Reineck, P.; Duhr, S.; Braun, D. Optical Thermophoresis for

Quantifying the Buffer Dependence of Aptamer Binding. Angew. Chem. Int. Ed. 2010, 49, 2238– 2241. (6)

Tuerk, C.; Gold, L. Systematic Evolution of Ligands by Exponential Enrichment: RNA

Ligands to Bacteriophage T4 DNA Polymerase. Science. 1990, 249, 505–510. (7)

Sharma, T. K.; Bruno, J. G.; Dhiman, A. ABCs of DNA Aptamer and Related Assay

Development. Biotechnol. Adv. 2017, 35, 275–301. (8)

Ellington, a D.; Szostak, J. W. In Vitro Selection of RNA Molecules that Bind Specific

Ligands. Nature 1990, 346, 818–822. (9)

Sefah, K.; Shangguan, D.; Xiong, X.; O’Donoghue, M. B.; Tan, W. Development of

DNA Aptamers Using Cell-SELEX. Nat. Protoc. 2010, 5, 1169–1185. (10) Willner, I.; Zayats, M. Electronic Aptamer-Based Sensors. Angew. Chem. Int. Ed. 2007, 46, 6408–6418. (11) Famulok, M.; Mayer, G. Aptamer Modules as Sensors and Detectors. Acc. Chem. Res. 2011, 44, 1349–1358. (12) Wang, F.; Lu, C. H.; Willner, I. From Cascaded Catalytic Nucleic Acids to Enzyme-DNA Nanostructures: Controlling Reactivity, Sensing, Logic Operations, and Assembly of Complex Structures. Chem. Rev. 2014, 114, 2881–2941. (13) Albada, H. B.; Golub, E.; Willner, I. Rational Design of Supramolecular Hemin/GQuadruplex-Dopamine Aptamer Nucleoapzyme Systems with Superior Catalytic Performance. Chem. Sci. 2016, 3092–3101.


27


Page 28 of 31

(14) Golub, E.; Albada, H. B.; Liao, W.-C.; Biniuri, Y.; Willner, I. Nucleoapzymes: Hemin/GQuadruplex DNAzyme-Aptamer Binding Site Conjugates with Superior Enzyme-Like Catalytic Functions.. J. Am. Chem. Soc. 2016, 164-172. (15) Biniuri, Y.; Albada, B.; Wolff, M.; Golub, E.; Gelman, D.; Willner, I. Cu2+ or Fe3+ Terpyridine/Aptamer Conjugates: Nucleoapzymes Catalyzing the Oxidation of Dopamine to Aminochrome. ACS Catal. 2018, 1802-1809. (16) Elbaz, J.; Wang, Z. G.; Orbach, R.; Willner, I. pH-Stimulated Concurrent Mechanical Activation of Two DNA "Tweezers". A "SET-RESET" Logic Gate System. Nano Lett. 2009, 9, 4510–4514. (17) Yan, H.; Zhang, X.; Shen, Z.; Seeman, N. C. A Robust DNA Mechanical Device Controlled by Hybridization Topology. Nature 2002, 415, 62–65. (18) Teller, C.; Willner, I. Functional Nucleic Acid Nanostructures and DNA Machines. Curr. Opin. Biotechnol. 2010, 21, 376–391. (19) Jo, H.; Her, J.; Ban, C. Dual Aptamer-Functionalized Silica Nanoparticles for the Highly Sensitive Detection of Breast Cancer. Biosens. Bioelectron. 2015, 71, 129–136. (20) Chen, W. H.; Yu, X.; Liao, W. C.; Sohn, Y. S.; Cecconello, A.; Kozell, A.; Nechushtai, R.; Willner, I. ATP-Responsive Aptamer-Based Metal–Organic Framework Nanoparticles (NMOFs) for the Controlled Release of Loads and Drugs. Adv. Funct. Mater. 2017, 27, 1702102. (21) Liao, W. C.; Lu, C. H.; Hartmann, R.; Wang, F.; Sohn, Y. S.; Parak, W. J.; Willner, I. Adenosine Triphosphate-Triggered Release of Macromolecular and Nanoparticle Loads from Aptamer/DNA-Cross-Linked Microcapsules. ACS Nano 2015, 9, 9078–9086.


28

Page 29 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(22) Okamoto, A.; Tanaka, K.; Saito, I. DNA Logic Gates. J. Am. Chem. Soc. 2004, 126, 9458–9463. (23) Stojanovic, M. N.; Stefanovic, D. A Deoxyribozyme-Based Molecular Automaton. Nat. Biotechnol. 2003, 21, 1069–1074. (24) Krishnan, Y.; Simmel, F. C. Nucleic Acid Based Molecular Devices. Angew. Chem. Int. Ed. 2011, 50, 3124–3156. (25) Campolongo, M. J.; Tan, S. J.; Xu, J.; Luo, D. DNA Nanomedicine: Engineering DNA as a Polymer for Therapeutic and Diagnostic Applications. Adv. Drug Deliv. Rev. 2010, 62, 606– 616. (26) Gallas, A.; Alexander, C.; Davies, M. C.; Puri, S.; Allen, S. Chemistry and Formulations for siRNA Therapeutics. Chem. Soc. Rev. 2013, 42, 7983-7997. (27) Arbel, N.; Ben-Hail, D.; Shoshan-Barmatz, V. Mediation of the Antiapoptotic Activity of Bcl-xL Protein Upon Interaction with VDAC1 Protein. J. Biol. Chem. 2012, 287, 23152–23161. (28) Wienken, C. J.; Baaske, P.; Rothbauer, U.; Braun, D.; Duhr, S. Protein-Binding Assays in Biological Liquids Using Microscale Thermophoresis. Nat. Commun. 2010, 1, 100. (29) Entzian, C.; Schubert, T. Studying Small Molecule-Aptamer Interactions Using MicroScale Thermophoresis (MST). Methods 2016, 97, 27–34. (30) Huizenga, D. E.; Szostak, J. W. A DNA Aptamer that Binds Adenosine and ATP. Biochemistry 1995, 34, 656–665.


29


Page 30 of 31

(31) Lin, C. H.; Patel, D. J. Structural Basis of DNA Folding and Recognition in an AMPDNA Aptamer Complex: Distinct Architectures But Common Recognition Motifs for DNA and RNA Aptamers Complexed to AMP. Chem. Biol. 1997, 4, 817–832. (32) Jeddi, I.; Saiz, L. Three-Dimensional Modeling of Single Stranded DNA Hairpins for Aptamer-Based Biosensors. Sci. Rep. 2017, 7, 1178. (33) Ahirwar, R.; Nahar, S.; Aggarwal, S.; Ramachandran, S.; Maiti, S.; Nahar, P. In Silico Selection of an Aptamer to Estrogen Receptor Alpha Using Computational Docking Employing Estrogen Response Elements as Aptamer-Alike Molecules. Sci. Rep. 2016, 6, 1285. (34) Krieger, E.; Darden, T.; Nabuurs, S. B.; Finkelstein, A.; Vriend, G. Making Optimal Use of Empirical Energy Functions: Force-Field Parameterization in Crystal Space. Proteins Struct. Funct. Genet. 2004, 57, 678–683. (35) Caulfield, T.; Devkota, B. Motion of Transfer RNA from the A/T State Iito the A-Site Using Docking and Simulations. Proteins Struct. Funct. Bioinforma. 2012, 80, 2489–2500. (36) Albada, H. B.; Golub, E.; Willner, I. Computational Docking Simulations of a DNAAptamer for Argininamide and Related Ligands. J. Comput. Aided. Mol. Des. 2015, 29, 643– 654. (37) Riek, R. Characterization of Hydrogen Bond Lengths in Watson-Crick Base Pairs by Cross-Correlated Relaxation. J. Magn. Reson. 2001, 149, 149–153.


30

Page 31 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC Graphic


31

ATP-Aptamer or ATP-Aptamer Mutant Complexes by

Recommend Documents