Water‐Soluble Peptoids with Two Different Binding Sites for Strong ATP Chelation

Abstract

Adenosine Triphosphate (ATP) is the activator of many enzymes, including kinases, that play a significant role in various medical conditions such as cancer. Therefore, inhibiting ATP‐dependent enzymes requires to disable its binding to enzymes, and this can be done by developing potent ATP chelators. Currently, there are two main types of chelators: one that binds ATP via zinc complexes of nitrogen‐based ligands, such as 2,2′;6′,2″‐terpyridine (Terpy), which target the phosphate ligands, and another one containing phenylboronic acid (PBA) to bind the diols within the ribose. Herein, we report on a unique chelation approach that combines these two binding strategies in one scaffold; we use peptidomimetic oligomers called peptoids that incorporate both Zn(Terpy) and PBA as strong, water‐soluble ATP binding inhibitors. These peptoids demonstrated high affinity to ATP, where the highest is K _D ‐ATP = 7.416 × 10⁻⁹ M, four times higher than the binding affinity of peptoids targeting only phosphates or only diols, and at least two orders of magnitude higher than known ATP chelators. Structural studies indicated that positioning both Terpy and PBA side chains close together on the scaffold increased ATP binding affinity, compared to spacing them apart.

ATP is key for activating enzymes associated with cancer. A potential therapeutic approach is strong binding of ATP, making it unavailable to enzymes. The water‐soluble peptoid P1‐TB incorporates two synergistic binding sites for ATP: terpyridine‐Zn²⁺ and 4‐CPBA that bind ATP's phosphate and diol groups, respectively, and exhibits an exceptional Kᴅ‐ATP of 7.416 × 10⁻⁹ M, representing a potential ATP chelator.

1. Introduction

Adenosine Triphosphate (ATP) is essential for the catalytic activation of key enzymes in the cell, such as kinases. ^{[ 1} , ² , ³ , ^{4 ]} Hence, ATP is a potential target for enzymatic inhibition toward various applications, including cancer treatment. Targeting ATP toward efficient enzymatic inhibition requires the development of a chelator with a high affinity to ATP, at least two orders of magnitude higher than the affinity of the kinase to the ATP activator, which ranges from K_D = 10⁻⁶–10⁻³ M.^{[ 5 ]} The chemical structure of ATP consists of three structural elements: an adenine base, ribose sugar containing cis‐diols, and a triphosphate moiety. Nowadays, research on ATP binding mainly focuses on small fluorescent sensors, where one of two possible binding approaches is used.^{[ 6} , ^{7 ]} The first approach targets ATP's phosphates using Zn complexes of nitrogen‐based ligands, such as dipycolylamine (DPA)^{[ 8} , ⁹ , ^{10 ]} and 2,2′;6′,2″‐terpyridine (Terpy). ^{[ 11} , ¹² , ¹³ , ¹⁴ , ^{15 ]} Zn‐Terpy‐based probes for phosphate binding are mainly small, organic, aromatic molecules, in which Zn²⁺ is bound in a penta‐coordination geometry where three coordination sites are the nitrogen atoms of Terpy and one or two labile chlorine ions that, upon binding to ATP, are replaced by the oxygen atoms of the phosphates. The binding affinities of these small probes are in the range of K_D (pyridine‐based Zn²⁺ probe‐phosphates) = 10⁻⁷–10⁻³ M. The second approach utilizes the molecular interactions between phenylboronic acid (PBA) and diols to target ATP‘s ribose cis‐diols. Phenylboronic acids (PBAs) are commonly used in medicinal chemistry for their ability to covalently and reversibly interact with cis diols to create stable boronic esters, and because of their low toxicity. The PBA‐diol interactions are used for numerous applications such as molecular recognition, enzymatic inhibition, and saccharide sensing.^{[ 16 ]} Generally, PBA‐diol interactions are relatively weak, with reported binding affinities of PBA derivatives to diols in the range of K_D (PBA‐diol) = 0.001–1.5 M.^{[ 17} , ^{18 ]} These can be slightly improved by aromatic substitutions within PBA, specifically of electron‐withdrawing groups on the ortho/para positions, which activate boron. However, overall, the binding to ATP is still not strong enough for the desired inhibition. In addition to the insufficient binding affinity to ATP demonstrated by both approaches, most of the chelators are small aromatic, hydrophobic molecules with low water solubility, making them less suitable for biological applications. One exception is a single reported example of ATP‐binding peptide,^{[ 19 ]} but its binding to ATP, via electrostatic and pi–pi interactions, is also relatively weak with K_D ATP = 2.2 × 10⁻⁵ M. To increase binding affinities to ATP and overcome the water‐solubility limitation, it is desirable to develop a water‐soluble ATP chelator that will contain several binding sites, and preferably both a Zn complex and PBA. Indeed, it was previously shown that improved affinities of PBA to ATP were achieved when multiple PBA molecules, along with cationic side chains capable of charge interaction, were incorporated within a polymeric scaffold.^{[ 20} , ^{21 ]} In contrast, only two published examples describe the incorporation of both a Zn complex (using the binding ligand DPA in both examples), and PBA on one scaffold. ^{[ 22} , ^{23 ]} The first example describes a small molecule‐based scaffold used for the formation of a combined chelator, but the PBA group did not participate in the binding; thus, there was no synergistic effect, and the binding affinity to ATP was low.^{[ 22 ]} In the second example, cyclodextrin covalently bound to PBA was combined with DPA ligand via a sensitive and conditioned self‐assembly process, and although the combined chelator did show a synergistic effect, its binding affinity to ATP was only 1.3 × 10⁻⁷ M.^{[ 23 ]} Based on these studies we set to design a water‐soluble ATP chelator that combines both Zn‐Terpy and a PBA derivative, which are covalently bound to one scaffold (Scheme 1). To this aim, we sought to use a peptidomimetic scaffold, due to the versatility, turnability, and biocompatibility of such scaffolds.

Scheme 1.

Illustration of dual‐sited strong ATP chelator design based on covalently bound biocompatible peptoid scaffold.

Peptoids, N‐substituted glycine oligomers are biocompatible peptide mimics with high stability toward proteolytic environments and different pH conditions, and are therefore great candidates for biological applications.^{[ 24} , ^{25 ]} Peptoids can be easily and efficiently synthesized on solid support via the sub‐monomer approach^{[ 26 ]} which incorporates N‐substituted glycine monomers rather than amino acids, and enables the incorporation of various side chains, including metal‐binding ligands^{[ 27} , ²⁸ , ^{29 ]} in specified and preorganized positions along the scaffold. This versatility and tunability of peptoids also facilitates sequence–function relationship studies for optimizing activity and selectivity.^{[ 30} , ³¹ , ^{32 ]} Specifically, the incorporation of Terpy within peptoids was previously demonstrated^{[ 33} , ³⁴ , ^{35 ]} and includes the efficient one‐step synthesis of the primary amine derivative of Terpy.^{[ 36 ]} In addition, different derivatives of phenylboronic acid were previously incorporated within peptoids using click chemistry^{[ 37 ]} or protected PBA incorporation.^{[ 38} , ^{39 ]} However, for the diol binding site of our designed chelator, we chose to use the PBA derivative 4‐carbamoylphenylboronic acid (4‐CPBA) due to its electron‐withdrawing effect that enhances the boronic acid reactivity, and its reported properties for diol^{[ 18 ]} and ATP binding.^{[ 21 ]} This derivative has not yet been used as a peptoid sidechain. Herein, we developed a new synthetic approach for the incorporation of 4‐CPBA to peptoids. Based on this design, we present here for the first time, short water‐soluble peptoids with high affinity of up to 7.4 × 10⁻⁹ M for ATP, via binding by two binding sites covalently incorporated on one scaffold (see Scheme 1).

2. Results and Discussion

2.1. Peptoids Design and Synthesis

The peptoids P1‐TB and P2‐TB were designed to have both Terpy and 4‐CPBA binding sites, either next to each other in the sequence or separated by one N‐(2‐methoxyethyl) glycine monomer, respectively, aiming to probe the effect of the distance between the two binding sites on the binding affinity to ATP. As control peptoids, we designed P3‐T and P4‐B, having only Terpy or 4‐CPBA, respectively, at the same position along the scaffold as in P1‐TB and P2‐TB. In addition to the binding sites, we incorporated at the N‐terminal of all four peptoids one piperidine group, which is required here for synthetic considerations (see below) and for water solubility. We kept the length of all the peptoids at four monomers by incorporating additional N‐(2‐methoxyethyl) glycine as additional water solubilizing group (Figure 1). The novel incorporation method of 4‐CPBA was performed by utilizing Fmoc chemistry without protection and deprotection of PBA (Scheme 2). Initially, commercially available N‐Fmoc‐ethylenediamine hydrochloride, was neutralized to get the free amine. This was used as a monomer within the standard peptoid synthesis method. Following, 20% piperidine solution in DMF was added for 45 minutes to simultaneously deprotect the NH‐Fmoc and cap the N‐terminus to prevent its reactivity during the following step. Finally, the commercially available 4‐carboxyphenylboronic acid was coupled to the free primary amine side chain using 4 equiv. of the boronic acid derivative, 4 equiv. of HCTU^{[ 40 ]} coupling agent, and 12 equiv. of DIPEA^{[ 41 ]} base in DMF for 1 hour. All four peptoids were synthesized via the sub‐monomer approach on a solid support ^{[ 23 ]}, cleaved from the resin, purified by preparative HPLC (>95%), and characterized by analytical electron spray ionization mass spectrometry (ESI‐MS⁺) and high‐performance liquid chromatography (HPLC, Supporting Figures S1–S5). The exact masses measured by the ESI–MS agreed with the expected masses. The water solubility of all the newly synthesized peptoids was 5 × 10⁵ mg/ L.

Figure 1.

Rationally designed ATP binding peptoids; P1‐TB and P2‐TB peptoids with two binding sites. P3‐T and P4‐B control peptoids with one binding site. The terpyridine binding site is marked in red, the 4‐CPBA is marked in blue, and the peptoid backbone and other as well as the non‐binding side chains are marked in black.

Scheme 2.

Illustration of the novel synthetic incorporation of 4‐CPBA derivative into peptoid backbone, utilizing Fmoc‐based chemistry.

2.2. Binding of P1‐TB and P2‐TB to Zn²⁺

The peptoids P1‐TB and P2‐TB were initially examined for their ability to bind Zn²⁺. We used ZnCl₂ for all experiments, aiming to form 1:1 Terpy: Zn complexes^{[ 10} , ¹¹ , ¹² , ¹³ , ^{14 ]}. We performed UV–vis titrations and ESI‐MS⁺ analysis to examine the binding properties of P1‐TB and P2‐TB to Zn²⁺. For the UV–vis experiments, ZnCl₂ (5 mM) was titrated into a solution of the free peptoid (20 uM), and UV–vis spectra were recorded. Metal‐free P1‐TB and P2‐TB exhibited absorption bands near 237 and 277 nm assigned to Terpy. Upon the addition of ZnCl₂, these two bands shifted to 244 and 274 nm, respectively, and two additional new absorption bands appeared near 310 nm and 322 nm (Figure 2). This is consistent with the results obtained with Zn²⁺‐terpy peptoids previously reported by our group,^{[ 29 ]} reflecting the binding of Zn²⁺ to the terpy ligand in both P1‐TB and P2‐TB. To further confirm the formation of the Zn²⁺‐peptoid complexes, ESI‐MS⁺ analysis was performed for both peptoids, after 20 minutes of preincubation with ZnCl₂ in a 1:1 ratio. The obtained masses in water were consistent with the exact mass of the Zn²⁺(peptoid) Cl complexes in the case of both P1‐TB and P2‐TB (Supporting Data, Figures S6, S7), supporting Zn²⁺ binding at the Terpy site of each peptoid in an aqueous environment.

Figure 2.

 ‐peptoid complex, red = Zn²⁺ ‐peptoid‐ATP).

2.3. Binding of the Peptoids to ATP

2.3.1. ATP Binding Characterization by UV–vis

Following the confirmation of Zn²⁺ binding to P1‐TB and P2‐TB by UV–vis titrations, the binding of both Zn‐peptoids to ATP was characterized by UV–vis. Upon the addition of ATP to a solution of Zn²⁺(peptoid)Cl, a few changes were observed (Figure 2): (i) the intensity of the bands near 310 and 322 nm, which are assigned to Zn²⁺‐Terpy, decreased, the band near 274 overlapped with a new band = near 260 nm, which is attributed to the absorption of the aromatic adenine within ATP, and (iii) the intensity of the band near 244 nm has increased. These observations reflect changes in the Terpy‐Zn²⁺ center and indicate the binding of Zn to phosphate ligands.^{[ 11} , ^{13 ]} In contrast to the binding of Zn²⁺, the binding of the PBA site to ATP's diols is difficult to detect by UV–vis because it is significantly weaker. Thus, we decided to use NMR techniques to further study the binding of ATP to the peptoids, and specifically the binding of the diol to the boronic acid part of the peptoids.

2.3.2. ATP Binding Characterization by NMR

The binding of P1‐TB and P2‐TB, as well as of the control peptoids P3‐T and P4‐B, having only one ATP binding site, was characterized by 1H‐NMR, 31P‐NMR, 2D‐COSY‐NMR and 2D‐NOESY NMR. The 1H‐NMR of all four peptoids showed peaks ranging from 1–5 ppm, which can be assigned to the backbone protons. In addition to these, P3‐T showed peaks in the aromatic region between 7.8‐8.75 ppm that can be assigned to Terpy (Figure S19), and P4‐B exhibited one multiplet peak between 7.57‐7.73 ppm that can be assigned to the protons of 4‐CPBA (Figure S20). For both P1‐TB and P2‐TB, we observed, in addition to the peaks assigned to the backbone protons, peaks ranging from 7.1‐8.75 ppm, assigned to both Terpy and 4‐CPBA (Figure S21‐22). The 1H‐NMR spectrum of ATP, showed two singlet peaks between 8‐8.5 ppm that can be assigned to the aromatic adenosine protons, one doublet peak at 6 ppm that can be assigned to the proton most adjacent to the adenosine, and 4 peaks ranging from 4.1‐4.6 ppm that can be assigned to the protons of the ribose ring (Figure S23). For a detailed assignment of the different protons, 2D‐COSY NMR was performed.

First, we analyzed P3‐T, P4‐B, and ATP by 2D‐COSY‐NMR (400 Hz), to characterize Terpy, 4‐CPBA, and ATP separately, by assigning each group of protons. In the NMR spectrum of P3‐T, the peaks between 7.8–8.8 ppm were assigned to the aromatic protons 3–6 and 3′‐6′ of Terpy (Figure S24). In the spectrum of P4‐B, the peaks at 7.6–7.75 ppm were assigned to the aromatic protons of PBA (Figure S25). Finally, the spectrum of ATP, exhibits peaks between 8.1–8.36 ppm, near 6 ppm and between 4.1–4.6 ppm, which are assigned to the protons of adenine (aromatic protons a, b, Figure S26), to the adjacent protons on the ribose ring (proton g, Figure S26), and to the ribose protons proximate to the diol binding site (protons c, d, e, f, Figure S26), respectively. The spectra of P1‐TB and P2‐TB were similar to those of P3‐T and P4‐B and exhibited peaks between 7.5‐9 ppm and near 7 ppm assigned to the protons of Terpy and 4‐CPBA, respectively (Figures S27, S28). Following the assignment of the relevant groups of protons, we examined the changes in the 1H‐NMR spectra upon ATP binding, focusing on the changes in the chemical shifts of the non‐aromatic ATP protons and of the aromatic peptoid protons, which are indicative of the binding.

For both P1‐TB and P2‐TB, the observed changes were similar; upon mixing of 1 equiv. of Zn²⁺(peptoid)Cl complex with 1 equiv. of ATP in D₂O for 1 hour, several shifts in the 1H‐NMR spectrum were observed. The first is a significant up‐field shift (∆0.2 ppm) of all protons proximate to the boronic ester formed during the binding of ATP and the peptoids. For ATP we observed upfield shifts for all these proximate protons; for example, proton “e”, which is the closest to the boronic ester creation site, was shifted upon binding to Zn²⁺(peptoid)Cl from 4.635 ppm to 4.473 ppm (Figure 3). This up‐field shift of ATP protons is consistent with literature reports on PBA‐diol binding,^{[ 20 ]} and can be explained by the shielding effect upon boronic ester formation.

Figure 3.

2D‐COSY NMR of ATP protons (blue bundles) and the evident shift upon binding to P1‐TB (red bundles). An upfield shift is observed upon binding (black arrows). The axis projections correspond to the free ATP 1H‐NMR.

The second is an up‐field shift, assigned to the PBA aromatic protons of the peptoids, from 7.25 ppm (center of the bundle) to 7.17 ppm (Figure 4). This shift as well stems from the electronic change within the boronic ester, causing shielding, which is also consistent with the literature.^{[ 42} , ^{43 ]} Finally, the Terpy protons ranging from 7.8ppm to 8.8 ppm, were shifted downfield upon binding to ATP (Figure 4). Protons 6,6′, which are the most proximate to the binding nitrogen of Terpy and thus are the most downfield in the free peptoid, were shifted even more downfield from 8.67 ppm to 8.83 ppm. This shift is commonly observed due to the binding of Zn²⁺ ions to terpy and the formation of Zn²⁺(terpy) complexes.^{[ 22} , ^{31 ]} To verify the participation of the phosphates in the binding, 31P‐NMR was performed for free ATP, Zn²⁺(P1‐TB)ATP, and Zn²⁺(P2‐TB)ATP. The 31P‐NMR revealed a downfield shift of the β‐phosphate from ‐23.1‐22.97 ppm (Figure 4), suggesting that the ATP is mainly bound to the terpy site via the β‐phosphate.

Figure 4.

(A) 31P‐NMR spectra comparing free ATPs’ phosphates and in a mixture of Zn²⁺(P1‐TB)ATP, (B) 1H‐NMR of free P1‐TB, ATP, and Zn²⁺(P1‐TB)ATP. The arrows indicate the shifts in the spectra for each binding site upon ATP binding. (C) 2D COSY‐NMR for the aromatic region of both free P1‐TB (bundles in blue) and Zn²⁺(P1‐TB)ATP (bundles red). The separation of the two binding sites is visible.

2.3.3. ATP Binding Characterization by ESI–MS

To further confirm the binding of ATP by P1‐TB and P2‐TB, we performed an ESI–MS analysis of the peptoids after ATP binding. For negatively charged molecules like ATP, ESI‐MS⁻ (negative mode) is commonly used.^{[ 44 ]} Thus, we analyzed a mixture containing either the control peptoid complex Zn(P3‐T) or the control peptoid P4‐B and ATP by ESI‐MS⁻. In both cases, the resulted chromatograms exhibited masses identical to the calculated full masses of Zn(P3‐T)ATP, and (P4‐B)ATP (Figures S9–S12). When analyzing the mixtures of Zn(P1‐TB) or Zn(P2‐TB), and ATP, the obtained ESI–MS⁻ spectra showed a mass of 1307.4 m/z, which matched the calculated mass for (P1‐TB)ATP and (P2‐TB)ATP. The fragments obtained in the ESI‐MS⁻ spectra matched the fragments of ATP presented in the literature ^{[ 45 ]} (Figures S13–S16). However, when the mixture solutions of Zn(P1‐TB) or Zn(P2‐TB) and ATP were analyzed by ESI‐MS⁺ the obtained chromatogram showed a mass of 936.4 m/z in both cases, which matched the calculated mass for Zn²⁺(P1‐TB)Cl and Zn²⁺(P2‐TB)Cl complexes (Figures S17–S18). The combination of all the above analyzed ESI‐MS⁻ and ESI‐MS⁺ results supported the binding of ATP by P1‐TB and P2‐TB at both binding sites.

2.3.4. ATP Binding Affinity Characterization by Nano ITC

To determine the K_D ‐dissociation constant and ∆G values for the ATP binding event, we used the Nano Isothermal Titration Calorimetry (Nano ITC) technique. This method enables to obtain the thermodynamic parameters by titrating a solution of the substrate (ATP in this case) into a solution of the chelator (peptoid or peptoid‐metal in this case). Following the titration, the data is fitted to a specific binding model based on computational data sets, to give the experimental thermodynamic parameters. The details of the Nano ITC experiment are described in Figures S31–S33.

The Nano‐ITC results for P3‐T revealed a K_D (Zn²⁺(P3‐T)ATP) of 8.347 × 10⁻⁵ M, when fitted to the “independent model” (95% fit precision, see S31). This model applies to a binding event where the binding site operates independently, without any interactions with other sites. This K_D value is within the range reported in the literature for the Zn²⁺(terpy) phosphate binding site.^{[ 11} , ¹² , ¹³ , ¹⁴ , ^{15 ]} The ITC heat changes for P4‐B were not significant, and thus the model fit for the P4‐B peptoid containing 4‐CPBA binding site was poor, hence, the resulting K _D value was not considered. This result might indicate weaker binding of P4‐B toward ATP, which cannot be detected by Nano‐ITC due to its sensitivity limitations (>K_D‐10⁻³ M). This result is consistent with the reported values for this type of 4CPBA‐diol binding, stated to be in the range of K_D‐10⁻³‐10⁻¹ M.^{[ 18 ]} To proceed and determine the binding constant of P4‐B to ATP using a different method, Alizarin Red S (ARS) three‐component competition experiment was performed.^{[ 18 ]} ARS is a diol‐containing fluorescent dye that exhibits a low signal in its free form but an evident signal when bound to boronic acid. This enables the detection of the binding event. In this experiment, initially, the binding constant between the ARS indicator and the PBA‐carrying peptoid P4‐B was determined. Next, a competition experiment between two diol‐containing molecules, ARS and target ATP, was performed and monitored by changes in the fluorescent signal. If the ATP binds the peptoid and replaces the prebound ARS, the ARS fluorescent signal will quench as it becomes free. The fluorescent signal change can be used to calculate the K _D of P4‐B to ATP. The competition experiment resulted in K_D (P4‐B‐ATP) = 0.135 M (Figure S34). This binding affinity is very low compared with the binding range reported in the literature for PBA‐diol interactions^{[ 17 ]} and PBA‐containing peptoids^{[ 39 ]} but supports our assumption that the surrounding moieties and scaffold of PBA and its derivative have a great effect on its diol binding affinity. The Nano‐ITC performed for P1‐TB and P2‐TB revealed binding affinities of K_D (Zn²⁺(P1‐TB)ATP) = 7.416 × 10⁻⁹ M and K_D (Zn²⁺(P2‐TB)ATP) = 6.776 × 10⁻⁸ M when fitted to the “Multiple sites” model (95% fit precision, see S32‐S33); this model is compatible with a biding event where different sites have effect on one another. Comparing the K _D for both P1‐TB and P2‐TB with that of the control peptoids, we clearly see a synergistic effect of the two binding groups, where the binding of each group separately is weaker than that of both. Moreover, P1‐TB showed higher affinity than previously reported probes, either with one of the binding sites or with both (Table 1), indicating that incorporating the two binding sites within one covalent scaffold is beneficial. Interestingly, the binding affinity of P1‐TB to ATP is higher than that of Firefly luciferase,^{[ 45 ]} which is an enzyme used to quantify the amount of ATP in cells, indicating the possible ATP extraction and, thus, enzymatic inhibition by peptoid P1‐TB. From the Nano‐ITC measurements, we also elucidated the ∆G for the ATP binding event. The calculated binding event ∆G for both P1‐ TB and P2‐TB was calculated to be ∆G_Zn ²⁺ _{( P1‐TB )ATP} = −46.4 Kj/mol and ∆G_Zn ²⁺ _{( P2‐TB )ATP}  = −40.9 Kj/mol, respectively.

Table 1.

Calculated K _D  values for the binding of all four designed peptoids to ATP as obtained from Nano‐ITC in 0.1 M PBS, pH = 7.4, compared with reported ATP binding molecules.

Molecule	Binding sites‐ Terpy, PBA, or Both	Binding Affinity to ATP (M)	Refs.
Firefly Luciferase	–	2 × 10−5	45
Terpy‐based phosphate probe	Zn2+ Terpy	9.43 × 10−6	11
Polymeric PBA probe	PBA	3.73 × 10−5	19
Cyclodextrin‐based dual ATP probe	Two sites	1.3 × 10−7	22
P1‐TB	Two sites	7.416 × 10−9
P2‐TB	Two sites	6.776 × 10−8
P3‐T	Terpy‐Zn2+	8.347 × 10−5
P4‐B	4‐CPBA	0.135[ [a]

^[a] The K _D for P4‐B was not obtained by Nano‐ITC, but rather by ARS competitive experiment

These results indicate that the formation of the complex Zn²⁺(P1‐TB)ATP is energetically more preferable than the formation of Zn²⁺(P2‐TB)ATP, suggesting that the peptoid sequence, and specifically here the methoxyethyl side chain spacing between terpy and 4‐CPBA, has a role in this difference.

2.3.5. Structural Effect Characterization by 2D‐NMR

To further probe the effect of the distance between the two binding sites on the binding affinity for ATP, we conducted 2D‐NOESY NMR. While the spectrum of free P1‐TB revealed that the two binding sites 4‐CPBA and Terpy are proximate in space, the spectrum of P2‐TB suggests no proximity between these two groups (Figure  5). As there is a difference of about 6 Kj/mol, which is in the same range as the reported energy required for peptoid scaffold rearrangement,^{[ 35 ]} we can suggest that the proximity between Terpy and 4‐CPBA is crucial for the synergy between them and for high binding affinity. We also suggest that the energy difference between the two peptoids is utilized for the rearrangement of the P2‐TB peptoid backbone toward the proximity of the two binding sites, which is required to promote a stronger synergistic binding toward ATP.

Figure 5.

2D‐NOESY NMR of (A) free P1‐TB, and (B) free P2‐TB. The proximate protons in space for P1‐TB are circled in black.

3. Conclusions

Herein, we report the first example of ATP‐binding peptoids, which are also a novel example of scaffolds containing both a Terpy‐Zn²⁺ complex and a boronic acid derivative for binding two sites of ATP. Specifically, the boronic acid derivative 4‐CPBA was incorporated within the peptoids for the first time, using a novel synthetic approach. The two water‐soluble peptoid complexes Zn²⁺(P1‐TB)Cl and Zn²⁺(P2‐TB)Cl demonstrated a synergistic binding effect between the sites; these two peptoids presented higher affinity toward ATP than peptoids with only one of the binding sites. Moreover, P1‐TB exhibited the highest binding constant ever reported for synthetic chelators. The binding of the peptoids to ATP was characterized using NMR techniques, and together with Nano‐ITC, revealed that the high binding affinity stems from the proximity of the two binding sites. These initial findings set the path toward the design of biocompatible, tunable, and rationally designed peptoids that can be used as ATP‐binding inhibitors for biological applications. Further studies will target modifying the peptoids to achieve their specificity for ATP over other substrates such as Adenosine diphosphate (ADP) and Pyrophosphate (PPi), and their delivery and detection in the cell.

Supporting Information

Additional experimental data of NMR, ESI‐MS, and Nano‐ITC can be found in the supporting information file.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supporting Information

Data Availability Statement

The data that support the findings of this study are available in the supporting information of this article.