Complementary oligonucleotides regulate induced fit ligand binding in duplexed aptamers

Hybridizing a complementary oligonucleotide to an ATP aptamer is shown to functionally regulate a newly revealed induced fit ligand-binding pathway.


Fig. S1
Graphical representation and modeling of observed binding affinities for cocaine DAs.      for 5 min, and then dried using filtered nitrogen gas. All washing, deactivation and prehybridization steps were carried out at room temperature in 100 mL of the respective buffer.
Surface-based Fluorescence Assay. 250 µL of 0.1 µM Cy3-labeled aptamer in Hybridization Buffer was heated to 72˚C for 5 min, 41˚C for 5 min, and then incubated at room temperature (25˚C) for 45 min. ACE microarrays were hybridized for 18 h at room temperature using custom hybridization chambers in a dark and humidity-saturated container. After hybridization, arrays were dissembled in 4xSSC, followed by washes with 2xSSC for 3 min and 1xSSC for 1 min, all at room temperature. Finally, microarrays were dried using filtered nitrogen. Slides were scanned (Calibration Scan), and stored dry under nitrogen in the dark at room temperature until carrying out an assay.
Hybridized slides were split into identical and addressable sub-arrays using a 16-well gasket (ProPlate 16-well slide module, Grace Bio-Labs, Bend, Oregon, USA). To carry out a surface-based fluorescence assay, ATP (or GTP) was serially diluted in Assay Buffer, and 110 µL of ligand (or buffer only) was added to each sub-array (generating test subarrays and "Buffer" subarrays), except for any "Blank" subarrays, which were not incubated with any buffer. After sample addition, wells were sealed and the slide was incubated for 1 h at room temperature. After sample incubation, slides were dissembled in room temperature 2xSSC and washed in room temperature 1xSSC for 1 min, dried using filtered nitrogen, and scanned (Assay Scan).
In support of the high accuracy, specificity and quantitative nature of measurements made using the surface-based fluorescence assay developed here, it is important to note that the performance of the current assay is superior to a previously published DA for ATP, in which DAs constructed on a glass surface were unable to provide quantitative readouts for ATP 4 . To our knowledge, the study by Carrasquilla et al. 4  step. This second normalization allows for relative changes in fluorescence caused by the presence of a target molecule for each ACE-based DA to be independently assessed, and also normalizes the dataset for any handlingspecific changes to DA fluorescence over the slide surface. In this manner, it is possible to quantitatively assess the loss of aptamers from the surface that arise from incubation in buffer-only or target-containing conditions. Finally, for some microarrays a Cy3-labeled control oligonucleotide was also printed in each sub-array, to verify that no differences in inter-subarray fluorescence losses occurred over the course of a surface-based fluorescence assay. Accordingly, in this work we have chosen to model DAs on the basis of K d Apt , and not K d Int . This is reflected in Fig. 2.

2) Four state model of DA ligand binding.
The four-state, two-pathway thermodynamic cycle for DAs shown in Fig. 2 can also be shown in compact form: Here, A represents aptamer molecules, C represents ACE molecules, and T represents ligand (ATP) molecules.

I.F.
The affinity constants for this cycle are defined as: At equilibrium, the induced fit and conformational selection pathways in this cycle must have equal equilibrium constants 7 , leading to the following equality: The apparent affinity (K d app ) of a system is defined as the amount of ligand needed to obtain a half-maximal increase in observed signal from baseline, and can be defined as for a typical affinity constant 7 . For a model of DAs considering only conformational selection, we can define the C.S.-only observed affinity (which we term K d C.S. , Fig.   3c black solid lines) using the effective concentration of reactants (the concentration of A and AC species before T addition, and (ii) the amount of T added) and products (i.e. (iii) the concentration of AT) to arrive at: Substituting the affinities defined in Eq.1, we have: Likewise, we can consider a model of DAs with only induced fit ligand binding. Here, we use the same logic used to define Eq.3 to define the I.F.-only observed affinity (which we term K d I.F. , Fig. S4b grey solid lines): (Eq. 5) Substituting the affinity constants defined in Eq.1, we arrive at: Finally, we can derive the apparent affinity for a complete model of DAs in which both conformational selection and induced fit are possible. In developing a model of the predicted observed binding affinity of DAs (which we term K d Obs ), we can define K d Obs as: Substituting the affinity constants defined in Eq.1, we can derive the following equation for K d Obs : (Eq. 8) By substituting for the affinity constants in the equality defined in Eq.2, Eq.8 can be used to predict K d Obs as a function of either K Hyb or K Hyb * for known values of K d Apt and K Fit , as done in Fig. 3c and S4a,b (dotted lines).
Tab. S1 Measured and simulated thermodynamic properties of ACEs tested in this study.  * For the simulated data, as well as the model presented in Fig. S4a, an additional correction value of 7.6 kJ mole -1 , a value derived for FAM:BHQ1 interactions 6 , is applied to the simulated ΔG 25˚C which is used to calculate K Hyb (K Hyb = e -∆G/RT ) for quencher labeled ACEs, since a Cy3:BHQ2 pair has been found to stabilize duplexes to a similar extent 8,9 .
** For the simulated data presented here, as well as the model presented in Fig. S4b, an additional correction value of 7.6 kJ mole -1 for the stabilizing Cy3:BHQ2 pair is applied to ΔG 25˚C for calculating K * Hyb for all ATP-disrupted DAs.
*** k * off values were calculated using k * off ≡ K * Hyb × k * on , and assuming a k * on of 10 7 M -1 s -1 given the relatively high salt conditions used here (e.g. 10,11 ). K * Hyb was predicted using DINAMelt with a correction of 3.8 kJ mole -1 applied to account for the stabilizing effect of the 5'-Cy3 moiety on the duplex (half the correction applied for a Cy3:quencher pair).
Additional Note: These values represent a lower estimate of duplex stabilities. In particular, in the surface-based assay, the effective salt, aptamer and ACE concentrations are higher, resulting in more stable duplexes on the microarray surface, leading to increased T m and ∆G values and lower k * off rates for surface-based DAs.
Tab. S2 Complete sequences of DNA oligonucleotides used in this study.