Specific Binding Constant and Stoichiometry ... - ACS Publications

Subscriber access provided by Queen Mary, University of London

Article

Determination of specific binding stoichiometry and affinity constants in free solution by mass spectrometry and capillary electrophoresis frontal analysis Cheng Qian, Hengqing Fu, Kevin A Kovalchik, Huihui Li, and David Da Yong Chen Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b02443 • Publication Date (Web): 06 Aug 2017 Downloaded from http://pubs.acs.org on August 8, 2017

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 free 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 accessible to all readers and 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.

Analytical Chemistry 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 21

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

Analytical Chemistry

Determination of specific binding constant and stoichiometry in free solution by mass spectrometry and capillary electrophoresis frontal analysis Cheng Qian1,2ǂ, Hengqing Fu1ǂ, Kevin A. Kovalchik2, Huihui Li1*, David Da Yong Chen1,2*

1

National and local Joint Engineering Research Center of Biomedical Functional Materials, Jiangsu

Collaborative Innovation Center of Biomedical Functional Materials, School of Chemistry and Materials Science, Nanjing Normal University, Nanjing 210023, P. R. China 2

Department of Chemistry, University of British Columbia, Vancouver V6T 1Z4, BC, Canada

ǂCheng Qian and Hengqing Fu contributed equally to this work. *Correspondence should be addressed to: Huihui Li: [email protected] David D. Y. Chen: [email protected]

ACS Paragon Plus Environment


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

Abstract A free solution method was developed for evaluating specific binding affinity and stoichiometry of small molecules with oligo DNA subsequent to cation induced G quadruplex formation. A nonlinear curve fitting equation capable of extracting specific binding constants in the presence of non-specific binding without the need of reference compounds was proposed and tested. Electrospray ionization mass spectrometry (ESI-MS) was first used to rapidly screen the small molecule candidates, then the stoichiometry and affinity constants of the native state binding pair in solution were obtained with capillary electrophoresis frontal analysis (CE-FA). B cell lymphoma 2 (Bcl-2) oncogene is directly responsible for the expression of Bcl-2 protein, which plays a significant role in cell apoptosis. The binding of G-quadruplex formed in the promoter region of Bcl-2 oncogene with small-molecule could stabilize the quadruplex structure, and potentially regulates the transcription of Bcl-2. Four natural product drug candidates were tested for their activity to bind the Bcl-2 promoter G-quadruplex. Using this reference-free method based on the CE-FA data, jatorrhizine and palmatine were found to bind specifically to the Bcl-2 promoter G-quadruplex with stoichiometries of 4:1 and 3:1, respectively.


Page 2 of 21

Page 3 of 21

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 DNA quadruplex can be formed with one or more DNA strands in the presence of certain cations in biological systems.1 The G-quadruplex has a globular four-stranded structure which is formed by stacked planar tetrads, and its G-quartets formed by Hoogsteen hydrogen bonding of the guanine bases are known to be energetically favored, thus giving G-quadruplexes higher stability than other quadruplex motifs.2-3 Intramolecular G-quadruplexes are more biologically relevant,2 and their structures are formed by winding the GC-rich DNA sequences in physiological condition where monovalent cations (K+ and Na+) are readily available.4 Since its initial discovery in human telomeres in the late 1980s,5-7 the G-quadruplex structure has opened the door to a new era for the DNA-targeted therapeutics. The initial interests were mainly on human telomeric DNA G-quadruplex.8 Neidle et al. found that the forming quadruplex structure could suppress the activation of telomerase, which is found to be overexpressed in 80~85% of the tumor cells.9 DNA G-quadruplexes were also discovered in proximal promoter regions of oncogenes, including c-MYC,10 Bcl-2,11 c-KIT,12 VEGF,13 and KRAS.14 In 2006, Dai et al. discovered an intramolecular G-quadruplex formed in the promoter region of the human Bcl-2 oncogene, and it has become an attractive target for gene therapeutics ever since.11 The Bcl-2 oncogene is directly responsible in the expression of the Bcl-2 oncoproteins, which plays a significant role in prohibiting cell apoptosis and is found overexpressed in a variety of human tumors.1517

With drugs directly bind to the Bcl-2 gene promoter quadruplex, the structure of quadruplex can be

stabilized, eventually regulates the expression of Bcl-2 oncoprotein.3, 18 Porphyrin derivatives19-20 and quindoline derivatives21-22 are among the most reported inhibitors of the Bcl-2 promoter G-quadruplex. The search for Bcl-2 gene promoter quadruplex binding compounds is on-going because compounds with better specificity are always desired. To improve the ability to search for anticancer quadruplextargeting drug candidates, a fast yet more reliable technique is needed to determine the binding interactions. To date, isothermal titration calorimetry (ITC),20, 23 surface plasmon resonance (SPR),21



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 4 of 21

circular dichroism (CD),19, 24 fluorescence spectroscopy19, capillary electrophoresis with laser induced fluorescence (CE-LIF)25-26 and electrospray ionization mass spectrometry (ESI-MS) 27-28 are some of the most employed techniques for measurements of the DNA quadruplex/small molecule binding interactions. However, the non-covalent nature of G-quadruplex formation made it difficult for methods needing immobilization or other types of modification before affinity evaluation. ESI-MS can provide information on binding stoichiometry in a non-solvated environment. However, because the condition at which the complex molecules observed are dramatically different from that in a solution, e.g., the temperature, solvent environment, and pH, the results given by ESI-MS can be less biologically relevant. In this sense, a solution-based technique needs to be employed to verify the possible drug candidates uncovered by ESI-MS. Capillary electrophoresis (CE), which consumes minimum sample, and is highly automated, can be a good complementary technique to ESI-MS. When used in binding analysis, CE is superior because it does not require immobilization of molecules which could possibly change their conformation, especially in this case where the oligo DNA need to have a conformational change in the free solution. Among all CE-based methods, CE frontal analysis (CE-FA) has many successful applications in studying high-order binding interactions which involve various kinds of bio-molecules (e.g. proteins, nucleic acids). 29-30 Because ESI-MS assay alone suffers from several drawbacks including nonuniform response factor, in-source dissociation and nonspecific binding,31 when used in combination with CE-FA it could improve the specificity and provide more accurate bio-relevant binding parameters. Non-specific binding can add to the complexity of interpreting the data, and usually is evaluated and corrected by a parallel experiment with a reference molecule that is known to not have specific binding with the ligands. In this work, a systematic approach to evaluating molecular binding of DNA G-quadruplex with small molecules is developed. Four small natural product molecules were tested for their affinity to the 23mer Bcl-2 oncogene promoter region G-quadruplex. The compounds were first individually mixed with


Page 5 of 21

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


a solution containing the oligo DNA with G-quadruplex formation, then directly infused into ESI-MS to quickly provide a rough estimate for the binding stoichiometry and binding strength. The candidates showing bound product were further evaluated with CE-FA. Their binding constants with the Gquadruplex in solution were calculated with non-linear regression analysis. We propose for the first time a curve fitting equation capable of extracting specific binding information in the presence of nonspecific binding, and the successful application of this equation is demonstrated.

Experimental section Chemicals and sample preparation Ammonium acetate (Aladdin Industrial Corporation, Shanghai, China) was dissolved with water, and a 20 mM ammonium acetate solution was used as the background electrolyte (BGE) throughout the experiments. High performance liquid chromatography (HPLC) purified wild type 23mer Bcl-2 mid GC-rich sequence (S1, 5’-GGG CGC GGG AGG AAG GGG GCG GG-3’) (m.w. = 7341.8 Da) was purchased from Sangon Biotech Co.Ltd. (Shanghai, China). The DNA powder was firstly dissolved in ultrapure deionized water (18.2 MΩ·cm, The Lab Corporation, Shanghai, China) to the concentration of around 200 µM, desalination and pre-concentration were then performed on this crude DNA solution with Amicon® ultra-centrifugal filters (MWCO 10 kDa, MilliporeSigma, Darmstadt, Germany). The obtained 500 µM of DNA solution was diluted to 100 µM with deionized water, and the obtained solution was quantified by a NanoDrop ND-2000 UV-Vis spectrophotometer (Thermofisher Scientific, Wilmington, DE) to confirm the concentration. This stock solution was then diluted with 20 mM ammonium acetate solution to achieve desired concentration during experiment. Triptolide, Chelidonine, Jatrorrhizine hydrochloride and Palmatine hydrochloride (Tauto Biotech Co.Ltd., Shanghai, China) were each dissolved with the ammonium acetate solution to concentration of 1 mM. The structures of these small molecules are depicted in Figure S1 of the Supporting Information. All solution was filtered with 0.22 µm membrane (ANPEL Laboratory Technologies, Shanghai, China) before experiments unless otherwise stated. ACS Paragon Plus Environment


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 6 of 21

Instrumentation Electrospray ionization mass spectrometry (ESI-MS) Gas-phase affinity measurements were performed on an Orbitrap Fusion™ Lumos™ Tribrid™ ultra high resolution mass spectrometer (Thermo Fisher Scientific, Waltham, MA) operating in negative ion mode. The original ionization source is replaced by a low dilution flow-through microvial CE-ESI-MS interface developed by our group,32-33 and a standard bare-fused silica capillary (50µm ID, 360µm OD) (Polymicro Technologies, Phoenix, AZ) was inserted into the stainless-steel needle with beveled tip made in house. The solutions containing non-covalent complexes were infused directly into the mass spectrometer at 500 nL/min through the capillary, and a voltage of -3.7 kV was applied on the spray needle to establish the electric field required for ionization. The temperature of ion transfer tube was set to 120 oC, and the radio frequency voltage of ion optics was optimized to 30% for better ion transmission. All sample mixtures tested by ESI-MS were prepared in the spray solvent (50/50 v/v, methanol/50 mM aqueous ammonium acetate, pH = 8.0) for stable ionization. Capillary electrophoresis frontal analysis (CE-FA) The liquid phase affinity measurements were performed on a PA800 Plus capillary electrophoresis system (Sciex, Framingham, MA) with a photodiode array (PDA) detector installed. The capillary had a total length of 50 cm, and its inlet end to detector length (effective length) was 40 cm. To minimize the adsorption of analyte molecules on the inner capillary wall, the capillary interior was coated with hydroxypropyl cellulose (HPC) prior to use.34 During frontal analysis, hydrodynamic injections of sample were performed at 1 psi for 90 s, and the injecting volume was estimated to be around 150 nL. The analyte was separated under voltage of 15 kV, and a pressure of 1 psi was simultaneously applied to the capillary to ensure the detection of all species. Circular dichroism (CD) spectroscopy


Page 7 of 21

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


For purpose of monitoring the DNA structure changes, CD experiments are performed on a Chirascan CD spectrometer (Applied Photophysics Ltd, England). The samples used for CD measurement were prepared by diluting the DNA stock solution to 2.5 µM with 30 mM Tris-HCl (pH = 8.0) or 25 mM ammonium acetate solution (pH = 8.0). During the tests, samples were scanned from 220 nm to 350 nm in a 0.1 cm path-length cuvette, and a spectrum based on the average of three replicates was obtained for each.

Results and discussion Characterization of G-quadruplex structure Circular dichroism (CD) spectroscopy is one of the most common tools used for the characterization of DNA conformational polymorphism in solution, because the two basic types of G-quadruplex each displays its own signature profile.35 In Figure S2 in the Supporting Information, the S1 sequence dissolved in 30 mM Tris-HCl solution (Black trace) does not give any significant absorption in the CD scanning range (200 to 350 nm), and it indicates that the deoxyoligonucleotides mostly existed in a random-coiled form of single strands. When dissolving the S1 sequence in the 25 mM ammonium acetate solution (red trace), very significant positive bands were observed at the wavelengths around 210 nm, 260 nm and 295 nm. This unique peak profile indicates that the mixed parallel /anti-parallel stranded G-quadruplex was formed under the presence of NH4+ cations in solution. This observation is in accordance with that observed by Dai et al.11 ESI-MS was used to study the stability of the G-quadruplex transferred from the solution to gaseous phase. A comparison was done for the mass spectra of S1 when dissolved in a cationic spray solution (50/50, v/v, 50 mM ammonium acetate aqueous solution/methanol) as shown in Figure 1A, and in a H2O/methanol (50/50, v/v) (Figure 1B). While the cations were absent in the spray solution in Figure 1B, a series of deprotonated single stranded ions were observed and the majority of the ions carried 7 to 10 negative charges. After the NH4+ cations were introduced to the solution, most of the ions on the spectra ACS Paragon Plus Environment


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 8 of 21

were with less charges (5 to 6) and the peaks of intramolecular G-quadruplex ions were found at m/z = 1047.74 ([S1+2NH4+-9H+]7-, shown in the spectrum as [G1]7-), m/z=1228.21 ([S1+2NH4+-8H+]6-, shown as [G1]6-), m/z = 1474.06 ([S1+2NH4+-7H+]5- (shown in the spectrum as [G1]5-) and m/z = 1842.82 ([S1+2NH4+-6H+]4-, (shown as [G1]4-). It is known that the G-quadruplex is an inclusion complex formed with the cations placed between the G-tetrad layers.27, 36-37 Thus, it is reasonable to conclude that the G-quadruplex is formed by cation-induced folding of the deoxyoligonucleotides, which is indicated by decreased average charge states of ions in the ammonium acetate solution. According to the (n-1) rule,37 we could also conclude that the formed intramolecular G-quadruplex has three G-tetrad layers from the observation that the most abundant G-quadruplex ions had two NH4+ cations added. It is confirmed by the CD and ESI-MS experiments that the G-quadruplexes were formed in ammonium acetate solution and that they had good stability both in aqueous and gaseous phase.

Figure 1. ESI-MS spectra of 5 µM S1 in (A) spray solution (50 mM ammonium acetate solution/methanol, 50/50, v/v); (B) 5 µM S1 in H2O/methanol (50/50, v/v) ACS Paragon Plus Environment

Page 9 of 21

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


Rapid ESI-MS screening for small molecules binding with G-quadruplex To quickly screen for candidate molecules with binding interaction with our target DNA G-quadruplex (G), ESI-MS was used to examine the binding affinity and stoichiometry. The binding interactions were measured by mixing 5 µM of G with each of the ligands (Ligand 1 to 4) separately in a 1:4 molar ratio, and the samples were directly infused to ESI-MS. Figure 2 shows the obtained mass spectra with the m/z range from 1200 to 1800, with all peaks of over 2% relative abundance. We can quickly conclude that triptolide (L2), chelidonine (L1), palmatine hydrochloride (L4), jatorrhizine hydrochloride (L3) showed that they could interacted with the target G-quadruplex under the stoichiometries of 1:2, 1:2, 1:3 and 1:4, respectively.

Figure 3. ESI-MS spectra of 5 µM G-quadruplex (G) binding with (A) 20 µM of triptolide; (B) 20 µM of chelidonine; (C) 20 µM of jatorrhizine hydrochloride; (D) 20 µM of palmatine hydrochloride. All samples were dissolved in a spray solution (50 mM ammonium acetate solution/methanol, 50/50, v/v), and all mass spectra were obtained in negative ion mode based on average of 10 scans. ACS Paragon Plus Environment


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 10 of 21

To evaluate the relative binding strength between various ligands and the G-quadruplex, a parameter IRa was introduced.38-39 n−

∑ Ir[G + mLx] IRa = ∑ Ir[G + mLx] + ∑ Ir[G] n−

In Eq 1, m = 1 to 4; n = 5 to 6; and x = 1 to 4.

n−

∑ Ir[G + mLx]

n−

(1)

was used to denote the arithmetic sum of

the relative intensity of G-quadruplex/ligand complex ions of all charge states,

∑ Ir[G]

n−

denoted the

arithmetic sum of relative intensity of unbound G-quadruplex ions of all charge states. Thus, IRa represents the ratio of bound G-quadruplex concentration to total G-quadruplex concentration. The relative abundances of all complex species are tabulated in Table 1. Table 1. Relative intensity of all complex species formed by G-quadruplex and ligands* Ligand Abbr. L1 Triptolide L2 Chelidonine L3 Jatrorrhizine L4 Palmatine *N.D.= not detectable

IRa[G+Lx] 0.32 0.24 0.34 0.37

IRa[G+2Lx] 0.08 0.03 0.22 0.15

IRa[G+3Lx] N.D. N.D. 0.10 0.04

IRa[G+4Lx] N.D. N.D. 0.03 N.D.

IRa 0.40 0.27 0.69 0.56

Under the same experimental condition, a ligand with a higher IRa value towards a certain host typically suggests stronger affinity. Since the same molar ratio of DNA to ligand was used through all tests, the order of ligand binding strength is suggested as L3 > L4 > L1 > L2 based on the IRa value. When mixed with L4 at the ratio of 1:4, close to 70% of the DNA quadruplex was bound with one or more ligand molecules. For further evaluation of the binding constants, the following equation was used to calculate the binding affinity. 31, 40-41 n

∑ Ab(GL

n+

R1 =

) =

n

∑ Ab(G

n+

)

[G + L ] (2) [G ]


Page 11 of 21

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


In Eq 2, R1 represents the abundance ratio of complex [G+L] to unbound quadruplex [G] when the reaction has a 1:1 stoichiometry. The abundance of ions must include all charge states when doing the calculation. The affinity K1 is calculated from R1 with Eq 3:

K1 =

R1 R1 R1 = = (3) [ L] [ L]0 − [G + L] [ L] − R1 [G ] 0 0 1 + R1

On the right-hand side of the equation, [G]0 and [L]0 denote the total concentration of DNA quadruplex and ligand molecules added in the mixture, respectively. If a DNA quadruplex molecule had multiple binding sites (2, 3...n), then a general form of Equation 3 is available to calculate the binding constant of each step of reaction.

Rm

km =

∑ nR − 1+ ∑ R

(4)

n

Rm −1 *([ L]0

n =1

[G ]0 )

n

n =1

In Eq 4, km denotes the binding constant of step m and Rm denotes the abundance ratio of complex with multiple ligand [G+mL] (m≤n) to unbound quadruplex [G]. When m=1, then Rm-1=R0, and it represents the ratio of Ab(G)/Ab(G) which obviously equals to the value of 1. With Eq 4, the binding constants of all the ligands were calculated and the results were listed in Table 2. Table 2. Summary of the binding affinity between ligands and the Bcl-2 DNA quadruplex Ligand Triptolide

Abbr. L1

k1/ M-1 3.1×104

k2/ M-1 1.4×104

k3/ M-1

k4/ M-1

——

——

Chelidonine Jatrorrhizine Palmatine

L2 L3 L4

1.8×104 9.7×104 5.2×104

5.8×103 6.0×104 2.6×104

—— 4.0×104 1.5×104

—— 2.8×104


——


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 12 of 21

From Table 2, the order of calculated binding constant is in good agreement with IRa values, and L3 showed the greatest affinity to the Bcl-2 DNA quadruplex. According to Rosu et al.,37 k1 = 4k2 when the two binding sites are equivalent and independent in a reaction with 1:2 binding stoichiometry. If k1 < 4k2, then either positive cooperative binding is suggested or the binding sites are not equivalent. Based on this theory, the ESI-MS data indicated that the binding sites on the DNA quadruplex are not equivalent, and the first bound ligand molecule helps to stabilize the quadruplex structure which leads to further binding interaction. Further examination of target molecules with capillary electrophoresis frontal analysis Although the ESI-MS results had provided useful information towards the binding interaction, it remained unclear if the binding interaction also occurs in liquid phase, which is more critical for drug discovery. Thus, further tests on the binding pairs in aqueous environment are necessary. Since all the ligand candidates showed positive affinity to the target DNA quadruplex in the initial ESI-MS measurements, CE-FA experiments were performed on all of them for further investigation and specificity determination。

In CE-FA experiments, a relatively large quantity (ca 100 nL) of pre-equilibrated mixture of DNA quadruplex and ligand was injected to the capillary prior to separation. Upon applying voltage on the capillary, different species of analyte in the sample will separate due to their different electrophoretic mobility. Because the ligand are much smaller molecules (m.w. = 300~400 Da) compared to the DNA quadruplex (m.w. > 7000 Da), they have very different electrophoretic mobilities. During the separation, we observed a plateau-shaped signal of the ligand molecules which migrate out from the sample plug of DNA quadruplex and the complex molecules. The height of this plateau is directly related to the unbound ligand concentration in the equilibrated mixture, and the unbound ligand concentration can be determined with the help of a standard calibration curve.


Page 13 of 21

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. Electropherogram obtained from the CE-FA experiments. Sample mixture of 50 µM of palmatine hydrochloride and 10 µM of Bcl-2 promoter DNA quadruplex was separated under 15 kV.

To correlate the concentration of unbound ligand to the stoichiometry and the binding constant, a method described by Ying etc42 was used. For a binding interaction with 1:1 binding stoichiometry,

I=

k [ L] f [ L]b [GL] = = [G ]t [GL] + [G ]u 1 + k[ L] f

(5)

The subscript b, u and t in Eq 5 are used to denote the bound, unbound and total concentration of the corresponding species in solution, respectively. In the experiments, the term [L]b is determined by subtracting the unbound ligand concentration from total ligand concentration. Thus, term I on the left side of equation which represents the average binding ratio, can be determined if the total concentration of G-quadruplex is known. If an assay is designed to change the input total concentration of ligand [L]t while keeping [G]t constant, then the binding constant k can be determined by a nonlinear regression analysis of the I-[L]t relationship with equation 5. For a binding interaction with unknown stoichiometry that possibly involves more ligands, the value of I also serves as a good indicator of the stoichiometry if the ligand concentration is high enough to saturate all binding sites on the G-quadruplex. In those situations, a more general form of equation is used for the binding constant determination. ACS Paragon Plus Environment


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

∑ nk [ L] I= 1 + ∑ k [ L]

Page 14 of 21

n u

n

n =1

n

n u

(6)

n =1

In eq 6, n represents the total number of binding sites on the quadruplex. The binding constants kn for each step of the interaction can be determined simultaneously by performing nonlinear regression analysis on the experimental data. The CE-FA results showed 3 of the 4 ligands (Ligand 1,3,4) has affinity to the DNA G-quadruplex. Detailed data obtained from the measurement are listed in the Tables S1, S2, and S3 of the Supporting Information. Although the quadruplex-L2 complex was observed in ESI-MS measurement, no sign of binding interaction was observed in solution. Since chelidonine (L2) was negatively charged under experimental condition, it is possible that the formation of quadruplex-L2 complexes could enhance the ionization in the negative ESI mode. Although the G-quadruplex also carries negative charges in gas phase, the hydrophobic interaction can keep the complex intact. However, the stability of such complexes could not hold in solution, and it is determined to be a false-positive result. The rest of the 3 ligands showed binding activity, and the relationships of I versus [L]t were plotted in Figure 4. Quite noticeable from the graph, two distinct binding profiles were discovered from the three remaining ligands. As the total input concentration of ligand [L]t increases, the increasing of I showed a linear trend for ligand 1, shown in Figure 4A, which very likely suggested non-specific binding.43 For ligand 3 and 4, the I value increased faster initially when the molar ratio of ligand to G-quadruplex was low. After more ligands were put into the mixture, the binding curves gradually flattened and eventually showed a linear trend at higher ligand concentration. Since specific binding sites are generally more effective than nonspecific binding sites, the specific binding tends to dominate when the ligand concentration is relatively low.44-45 Thus, the binding curves of ligand 3 and 4 are estimated to be contributed by both specific binding and non-specific binding, simultaneously.


Page 15 of 21

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 4. Binding isotherms of DNA G-quadruplex with ligands obtained from CE-FA. (A) triptolide (L1); (B) jatrorrhizine (L3); (C) palmatine (L4). Closed squares represent the apparent binding isotherms of ligands obtained directly from the experimental data, open triangles represent the linear non-specific binding which is deduced from the slope of the linear part of the apparent binding curve, and open circles represent the specific binding which is obtained by the subtraction of non-specific binding from total binding of ligands.



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 16 of 21

Extraction of specific binding information Specific bindings were crucial to the drug research and characterization.44, 46-47 To account only for the specific binding in a binding interaction which consists of both specific and non-specific portion, a more reliable method is needed. The most common approach to solve this issue is to introduce a control group,31, 41-42, 48 and the molecules selected as the control are required to have no specific binding, and they normally share a similar structure with the either the target ligands or host molecules. However, sometimes those control groups can be hard to find in real situations due to the uniqueness of some ligands or hosts. Thus, we are proposing a novel reference-free method which is simple and effective, for the extraction of the specific binding information. Non-specific binding is often an electro-static interaction in nature.49-50 Since non-specific binding does not require certain “pockets” in the host structure, more ligand molecules can bind non-specifically to the host especially when the size of host is much larger than ligand molecules (e.g. protein with small drug molecules). While specific binding becomes minimal after nearly all specific binding sites are occupied, the continued increase in binding which extends the binding curve to much higher ligand concentrations is due to non-specific binding. As long as it is in the linear range, the non-specific binding should always be proportional to the free ligand concentration in the solution. Thus, we can establish an equation which accounts for both specific and non-specific binding:

∑ nk [ L] I= 1 + ∑ k [ L] n u

n

n =1

n

n u

+ λ[ L ]u

(7)

n =1

On the right side of eq 7, the first part describes non-specific binding while the second part describes the non-specific binding, with a non-specific factor λ multiplies the [L]u. In order to determine the maximum number of specific binding sites, the non-specific binding factor λ should be estimated first. A rough estimation can be done by measuring the slope of the linear part of binding curve which is dominated by non-specific binding. In Figure 4B and 4C, the estimated non-specific binding curves are ACS Paragon Plus Environment

Page 17 of 21

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


represented with the dashed line. To obtain the extent of specific binding, we need to calculate the height difference between the total binding curve and non-specific binding curve. After the non-specific bound ligand [L]bn deducted from the bound ligand [L]b, the specific binding isotherms of ligand 3 and ligand 4 were established and represented as open circles in Figure 4B and 4C. It was revealed that jatrorrhizine and palmatine reacts with the bcl-2 DNA G-quadruplex at 4:1 ratio and 3:1 ratio, respectively. To quantitatively provide the accurate binding constants, eq 7 is used to fit the total binding isotherm in Figure 5, and related results are listed in Table 3.

Figure 5. Total binding isotherms of DNA G-quadruplex to jatrorrhizine (A) and palmatine (B) . The curves (solid line) represent the nonlinear regression results with Eq 7.



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 18 of 21

Table 3. Binding constants of DNA G-quadruplex to jatrorrhizine and palmatine

Jatrorrhizine

Palmatine

1st binding constant k1/M-1

2.0×106

8.4×106

2nd binding constant k2/M-1

2.2×105

2. 5×105

3rd binding constant k3/M-1

1.1×105

3.5×104

4th binding constant k4/M-1

9.4×104

__

Non-specific binding factor λ/M-1

2.4×104

3.0×104

R square of fitting

0.99

1.00

While the binding stoichiometries are in accordance with the initial ESI-MS results, the obtained strength of affinity is quite different in gaseous phase and aqueous phase. Orders of magnitude difference in binding strengths determined from different phases has been mentioned and well discussed previously.40, 51 Due to the lack of solvent molecules in gaseous phase, H-bond and electrostatic force play the most important roles in binding interactions without existence of solvent. While in aqueous environment, the hydrophobic interactions between ligands and hosts also contribute to the strength of binding. This effect is especially significant in the G-quadruplex-small molecule binding interactions, since it has been commonly accepted that drug candidates with planar aromatic rings binds to the quadruplex preferably through π-π stacking.1

Conclusion A systematic approach to studying binding interactions between large biomolecules and small ligands is developed. The method utilizes ESI-MS to first rapidly screen for possible ligands, then uses CE-FA to confirm solution state interaction and determine the reaction stoichiometry and stepwise binding ACS Paragon Plus Environment

Page 19 of 21

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


constants. A new equation is introduced to make this reference-free method based on CE-FA possible with the consideration of both specific and non-specific binding interactions. With this method, we discovered that jatorrhizine and palmatine showed good binding strength and specificity when reacting with the Bcl-2 promoter gene G-quadruplex. Therefore, they could act as potential inhibitors to regulate cell apoptosis. The combination of ESI-MS and CE-FA not only increases the speed of analysis, but also improves accuracy and specificity compared to conventional binding analysis. In the future, we believe that the method described in this paper can be used in the screening of other molecular interactions for drug candidate screening.

Acknowledgement The work is supported by grants from the National Natural Science Foundation of China (Grant Number 21475061), Natural Science Foundation of the Higher Education Institutions of Jiangsu Province (15KJB150017), the Priority Academic Program Development of Jiangsu Higher Education Institution and Nanjing Normal University (NJNU). CQ was supported by a Mitacs Globalink Award for a Research Visit at NJNU.

Supporting Information Chemical structures of ligands and the CD spectrum of S1 DNA CE-FA data used for plotting Figures 4 and 5, as well as for curve fittings.

Reference 1. Yang, D.; Okamoto, K., Future medicinal chemistry 2010, 2, 619-646. 2. Chen, Y.; Yang, D., Sequence, Stability, and Structure of G-Quadruplexes and Their Interactions with Drugs. In Current Protocols in Nucleic Acid Chemistry, John Wiley & Sons, Inc.: 2012. ACS Paragon Plus Environment


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 20 of 21

3. Balasubramanian, S.; Hurley, L. H.; Neidle, S., Nat Rev Drug Discov 2011, 10, 261-275. 4. Sen, D.; Gilbert, W., Nature 1990, 344, 410-414. 5. Henderson, E.; Hardin, C. C.; Walk, S. K.; Tinoco, I.; Blackburn, E. H., Cell 1987, 51, 899-908. 6. Williamson, J. R.; Raghuraman, M. K.; Cech, T. R., Cell 1989, 59, 871-880. 7. Sundquist, W. I.; Klug, A., Nature 1989, 342, 825-829. 8. Neidle, S., FEBS Journal 2010, 277, 1118-1125. 9. Parkinson, G. N.; Lee, M. P. H.; Neidle, S., Nature 2002, 417, 876-880. 10. Siddiqui-Jain, A.; Grand, C. L.; Bearss, D. J.; Hurley, L. H., Proceedings of the National Academy of Sciences 2002, 99, 11593-11598. 11. Dai, J.; Dexheimer, T. S.; Chen, D.; Carver, M.; Ambrus, A.; Jones, R. A.; Yang, D., Journal of the American Chemical Society 2006, 128, 1096-1098. 12. Rankin, S.; Reszka, A. P.; Huppert, J.; Zloh, M.; Parkinson, G. N.; Todd, A. K.; Ladame, S.; Balasubramanian, S.; Neidle, S., Journal of the American Chemical Society 2005, 127, 10584-10589. 13. Sun, D.; Guo, K.; Rusche, J. J.; Hurley, L. H., Nucleic Acids Research 2005, 33, 6070-6080. 14. Cogoi, S.; Xodo, L. E., Nucleic acids research 2006, 34, 2536-2549. 15. Tsujimoto, Y.; Cossman, J.; Jaffe, E.; Croce, C. M., Science 1985, 228, 1440-1444. 16. Vaux, D.; Cory, S.; Adams, J., Nature 1988, 335, 440-2. 17. Adams, J. M.; Cory, S., Science 1998, 281, 1322-1326. 18. Brooks, T. A.; Kendrick, S.; Hurley, L., FEBS Journal 2010, 277, 3459-3469. 19. Nagesh, N.; Buscaglia, R.; Dettler, J. M.; Lewis, E. A., Biophysical Journal 2010, 98, 2628-2633. 20. Le, V. H.; Nagesh, N.; Lewis, E. A., PLOS ONE 2013, 8, e72462. 21. Wang, X.-D.; Ou, T.-M.; Lu, Y.-J.; Li, Z.; Xu, Z.; Xi, C.; Tan, J.-H.; Huang, S.-L.; An, L.-K.; Li, D.; Gu, L.-Q.; Huang, Z.-S., Journal of Medicinal Chemistry 2010, 53, 4390-4398. 22. Xiong, Y.-X.; Chen, A.-C.; Yao, P.-F.; Zeng, D.-Y.; Lu, Y.-J.; Tan, J.-H.; Huang, Z.-S.; Ou, T.-M., Biochemistry and Biophysics Reports 2016, 5, 346-352. 23. DuPont, J. I.; Henderson, K. L.; Metz, A.; Le, V. H.; Emerson, J. P.; Lewis, E. A., Biochimica et Biophysica Acta (BBA) - General Subjects 2016, 1860, 902-909. 24. Dexheimer, T. S.; Sun, D.; Hurley, L. H., Journal of the American Chemical Society 2006, 128, 5404-5415. 25. Pavski, V.; Le, X. C., Current opinion in biotechnology 2003, 14, 65-73. 26. Zhou, W.; Tomer, K. B.; Khaledi, M. G., Analytical Biochemistry 2000, 284, 334-341. 27. Li, H.; Yuan, G.; Du, D., Journal of the American Society for Mass Spectrometry 2008, 19, 550-559. 28. Casagrande, V.; Alvino, A.; Bianco, A.; Ortaggi, G.; Franceschin, M., Journal of Mass Spectrometry 2009, 44, 530-540. 29. Østergaard, J.; Heegaard, N. H. H., ELECTROPHORESIS 2003, 24, 2903-2913. 30. Fermas, S.; Gonnet, F.; Varenne, A.; Gareil, P.; Daniel, R., Analytical chemistry 2007, 79, 4987-4993. 31. Kitova, E. N.; El-Hawiet, A.; Schnier, P. D.; Klassen, J. S., Journal of The American Society for Mass Spectrometry 2012, 23, 431-441. 32. Maxwell, E. J.; Zhong, X.; Zhang, H.; van Zeijl, N.; Chen, D. D. Y., ELECTROPHORESIS 2010, 31, 1130-1137. 33. Zhong, X.; Maxwell, E. J.; Chen, D. D. Y., Analytical chemistry 2011, 83, 4916-4923. 34. Shen, Y.; Xiang, F.; Veenstra, T. D.; Fung, E. N.; Smith, R. D., Analytical chemistry 1999, 71, 5348-5353. 35. Kypr, J.; Kejnovská, I.; Renčiuk, D.; Vorlíčková, M., Nucleic Acids Research 2009, 37, 1713-1725. 36. Rosu, F.; Gabelica, V.; Houssier, C.; Colson, P.; Pauw, E. D., Rapid Communications in Mass Spectrometry 2002, 16, 1729-1736. 37. Rosu, F.; De Pauw, E.; Gabelica, V., Biochimie 2008, 90, 1074-1087. 38. Li, H.; Hai, J.; Zhou, J.; Yuan, G., Rapid Communications in Mass Spectrometry 2016, 30, 407-414. 39. Wan, K. X.; Shibue, T.; Gross, M. L., Journal of the American Chemical Society 2000, 122, 300-307. 40. Hofstadler, S. A.; Sannes-Lowery, K. A., Nat Rev Drug Discov 2006, 5, 585-595. 41. Kitova, E. N.; Soya, N.; Klassen, J. S., Analytical chemistry 2011, 83, 5160-5167. 42. Sun, Y.; Cressman, S.; Fang, N.; Cullis, P. R.; Chen, D. D. Y., Analytical chemistry 2008, 80, 3105-3111. 43. Olson, S. T.; Halvorson, H.; Björk, I., Journal of Biological Chemistry 1991, 266, 6342-6352.


Page 21 of 21

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


44. LONDON, E. D.; COYLE, J. T., Molecular Pharmacology 1979, 15, 492-505. 45. Ha, J.-H.; Capp, M. W.; Hohenwalter, M. D.; Baskerville, M.; Record, M. T., Journal of Molecular Biology 1992, 228, 252-264. 46. SUDLOW, G.; BIRKETT, D. J.; WADE, D. N., Molecular Pharmacology 1976, 12, 1052-1061. 47. Taylor, J. A.; Pastrana, C. L.; Butterer, A.; Pernstich, C.; Gwynn, E. J.; Sobott, F.; Moreno-Herrero, F.; Dillingham, M. S., Nucleic Acids Research 2015, 43, 719-731. 48. El-Hawiet, A.; Kitova, E. N.; Arutyunov, D.; Simpson, D. J.; Szymanski, C. M.; Klassen, J. S., Analytical chemistry 2012, 84, 3867-3870. 49. Gao, J.; Mammen, M.; Whitesides, G. M., Science 1996, 272, 535-537. 50. Mulgrew-Nesbitt, A.; Diraviyam, K.; Wang, J.; Singh, S.; Murray, P.; Li, Z.; Rogers, L.; Mirkovic, N.; Murray, D., Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 2006, 1761, 812-826. 51. Nesatyy, V. J., International Journal of Mass Spectrometry 2002, 221, 147-161.


Specific Binding Constant and Stoichiometry ... - ACS Publications

Recommend Documents