Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2024 Oct 2;146(41):28005–28013. doi: 10.1021/jacs.4c03905

Extended Sulfo-Pillar[6]arenes — a New Host Family and Its Application in the Binding of Direct Oral Anticoagulants

Chelsea R Wilson †,‡, Austia O Puckett †,‡, Isabella M Murray †, Allen G Oliver §, Fraser Hof †,‡,*
PMCID: PMC11487555  PMID: 39356656

Abstract

graphic file with name ja4c03905_0008.jpg

Herein, we report the synthesis of extended sulfo-pillar[6]arenes (sP6), a new host class with a pedigree in salt tolerance and ultrahigh binding affinity toward multiple drug classes. The parent sulfo-pillar[6]arene is a high-affinity host with the potential to act as a supramolecular reversal agent. However, it lacks synthetic diversification of the core scaffold. The new extended sulfo-pillar[6]arenes have either a monodirectional (A1sP6) or bidirectional (A1A2sP6) extension of the hydrophobic cavity. This new functionality enables more noncovalent interactions and strong affinity toward guests, which we demonstrate using the direct oral anticoagulants (DOACs) dabigatran, betrixaban, and edoxaban. DOACs are highly prescribed therapeutics that are underexplored in host–guest chemistry. These agents prevent the formation of blood clots and are prime targets for supramolecular sequestration. This functionalization also introduces new fluorescent properties to the sulfo-pillar[6]arene family via an incorporated p-terphenyl (A1A2sP6). We show that these new hosts have ultrahigh affinity toward dabigatran (Kd = 27 nM, A1A2sP6) in salty solutions and that the A1A2sP6 analogue can bind betrixaban in bovine plasma with a physiologically relevant Kd (7 μM).

Introduction

Antidotes, reversal agents, or rescue drugs are a class of therapeutics that can nullify the effects of another drug by sequestering it or by competing for the target receptor. Naloxone is a notable example of the latter; by directly out-competing opioids for the opioid receptor and thus blocking binding, it reverses their action in vivo.1 Sequestration is an attractive approach to drug reversal. Instead of the rescue drug creating competition at the biological receptor, the synthetic agent binds the drug directly and, in principle, can have fewer unintended side effects.2−6 A need to bind an analyte in biological media is a necessary hurdle to be overcome in the development of new sequestration agents.

Hosts are synthetic concave macrocycles that use noncovalent interactions to recognize and bind a target drug or analyte. These hosts have found many promising applications in health care, where they are primarily used to increase drug solubility or to deliver drugs.7−10 When a host encapsulates a drug in vivo, the host–guest complex is most often removed from general circulation by excretion.3,4,11 A high binding affinity, salt tolerance, and low dissociation rate are required for supramolecular reversal agents to be effective. Hosts can be improved by synthetic functionalization.12,13 For example, sugammadex is a synthetically modified cyclodextrin and is currently the only FDA-approved supramolecular reversal agent. There are other nonclinical examples of macrocycle functionalization leading to introduction of key properties. A sulfonatocalix[4]arene was imbued with the ability to detoxify V-type nerve agents.14 Additionally, the methylation of cucurbit[8]uril drastically improved water solubility, enabling the reversal of PCP-induced hyperlocomotion in mice.15 These examples highlight the promise of developing new host functionalization methods.

Direct oral anticoagulants (DOACs) are a class of drugs in need of reversal agents. In recent years, DOACs have gained popularity for the prevention of blood clotting, as evidenced by their increased prescription rates over classic vitamin K antagonists (VKAs) such as warfarin.16 Relative to VKAs, DOACs have a larger therapeutic window, do not require routine monitoring, have fewer adverse interactions, and have more predictability, which allows for better dosing.17 DOACs are now commonly prescribed for the prevention of deep vein thrombosis, stroke, and pulmonary embolism.18 There are two major groups of DOACs: direct thrombin (Factor IIa) inhibitors such as dabigatran and direct factor Xa inhibitors such as apixaban, rivaroxaban, betrixaban, and edoxaban (Figure 1a). Even though DOACs have many advantages over VKAs, there are still serious concerns around their ability to reverse their action during emergency care.19 Protein-based biologics are the only DOAC reversal agents used clinically. The modified recombinant factor Xa andexanet alfa is currently used as a reversal agent for apixaban and rivaroxaban. However, the drug itself has significant and life-threatening side effects.20 Idarucizumab, an antibody-based treatment for the reversal of dabigatran, comes with a high price tag of > $3500 per dose.21 Cheaper and safer alternatives to DOAC reversal are needed. Cyclodextrin OKL-1111 was recently reported as a broad-spectrum experimental reversal agent for DOACs.22 While this report illustrated the potential of a supramolecular reversal agent, the affinity (Kd of 25 μM – 6.7 mM) of OKL-1111 toward DOACs is weaker than would be expected to lead to success in the clinic. A stronger binding host class is needed.

Figure 1.

Figure 1

Open in a new tab

(a) Chemical structure of select DOACs in charge state present at pH 3.0. (b) Chemical structure of pillar[6]MaxQ (sP6) previously reported by Xue and co-workers,23 with functionalized analogues A1sP6 and A1A2sP6 (this work).

Sulfo-pillar[6]arene (sP6) is an ultrahigh (pM-nM) affinity host. First reported in 2020, sP6 has become an attractive candidate for use in therapeutic applications due to its high affinity to biologically relevant analytes23,24 and was recently shown to act as a broad-spectrum reversal agent for opioid-related overdose25 and neuromuscular blocking agents.26 In addition to its high affinity, sulfo-pillar[6]arene has also been shown to be salt tolerant. The binding affinity of the opioid fentanyl for sP6 only decreased from a Kd of 9.8 to 78 nM when excessive salt was present.25 Additionally, sP6 showed 16 nM affinity toward the peptidic biomarker H3K4Me3 in a salty media (137 mM NaCl).27 With its good biocompatibility, salt tolerance, and high affinity to multiple drug classes like amino acids,27 steroids,28 opioids,25 and neuromuscular blocking agents,23,26sP6 has a bright future in medicinal applications. However, the ability to modify the host scaffold is paramount to manipulating the binding affinity, selectivity, and pharmacological properties of the host. Functionalization methods for pillar[6]arenes are very limited, and the functionalization of sulfo-pillar[6]arenes has been nonexistent until now.

Results

We prepared sP6 derivatives with an elongated hydrophobic surface area to complement the rod-like structure of the DOACs. We predicted that by extending the hydrophobic surface and cavity depth of sulfo-pillar[6]arene, we would create an increase in binding strength as well as a tuning of guest selectivity through modification of host–guest contacts. We targeted mono- and difunctionalization to create new analogues with an extended aromatic pendent arm on either one face (A1sP6) or both faces (A1A2sP6) of the macrocycle while maintaining the 12 negatively charged sulfate groups (Figure 1b).

The high symmetry of pillar[6]arene renders monofunctionalization a challenging endeavor. The key starting material in functionalized derivatives is an alkoxy-protected pillar[6]arene, most commonly an ethoxypillar[6]arene (EtOP6). We recently reported a protocol that allows for access to ca. 20 g of pure EtOP6 in 1 day without chromatography.29 With ample starting material in hand, the functionalization of EtOP6 can be achieved through two main desymmetrization arteries, deprotection and oxidation.

A multistep synthetic pathway was developed for the A1 and A1A2 functionalization of ethoxypillar[6]arene (Figure 2). The monofunctionalization of 1 of the 12 ethoxy groups is achieved by removing the stated protecting group in the presence of BBr3, exposing one reactive phenol, 1.30 Difunctionalization is done by oxidizing one of the six diethoxybenzenes to a quinone.31−33 Once oxidized, the quinone can be reduced to a hydroquinone to provide two reactive alcohols.31 The now available mono- and direactive sites in the ethoxypillar[6]arene scaffold allow for more elaborate modifications, leading to functionalized water-soluble sulfo-pillar[6]arenes.

Figure 2.

Figure 2

Open in a new tab

Synthetic scheme for the A1 and A1A2 functionalization pathway of sP6.

The mono-oxidation product can be prepared with (NH4)2[Ce(NO3)6] following the literature protocol.32 However, like Zhu and co-workers,33 we also obtained notably lower yields than reported.32 Through optimization, we arrived at an alternative method for the mono-oxidization of ethoxypillar[6]arene. Over the course of 1.5 h, 0.9 M (NH4)2[Ce(NO3)6] (aq) was added via a syringe pump to a solution of EtOP6 in degassed DCM at – 20 °C; reliably obtaining 5 in a 24% yield after flash chromatography. The quinone in 5 was reduced with NaBH4 as in the literature31 to obtain hydroquinone 6. It is important to note that 6 displayed stability issues when stored for more than a few days; therefore, we carried this intermediate forward promptly. Triflation of 1 gave the monotriflated analogue 2 in a 91% yield. Compound 6 was triflated in a similar manner to obtain ditriflate 7 in an 83% yield. The Suzuki–Miyaura coupling protocol was inspired by Demay-Drouhard and co-workers’34 work on rim-differentiated pillar[5]arenes. For the A1 coupling, triflate 2 was coupled with 4-methoxyphenylboronic acid using XPhos Pd G3, providing 3 in a 73% yield. Compound 8 was similarly obtained from 7 in a 69% yield; the nature of the A1A2 extended hydrophobic surface of 8 was confirmed by crystallography (Figures S31–S32).

Water-soluble A1- and A1A2-functionalized pillar[6]arenes were obtained via sulfation. The remaining ethoxy-protecting groups of 3 and 8 were removed through treatment with BBr3 in dry dichloromethane at 0 °C, thus providing 4 (96% yield) and 9 (90% yield). Again, the unprotected phenols created stability problems, and it is recommended to carry these intermediates forward immediately after deprotection. The sulfation protocol of 4 and 9 was adapted from Xue and co-workers’23 work on sP6 with some noted modifications, including rigorous drying of reagents and most notably the achievement of post-workup desalting by dialysis rather than by host precipitation.25 The revised method reliably yields products that meet our standard of ≤50% salt content to allow for subsequent characterization and binding studies. In a typical set of outcomes for this new sulfation protocol, we obtained the novel extended sulfo-pillar[6]arenes A1A2sP6 in a 39% yield (49% salt content, yield corrected for salt) and A1sP6 in a 40% yield (36% salt content, yield corrected for salt). Even lower salt content can be obtained through an additional round of dialysis, leading to ≤26% salt (Figures S13, S28, and S29). Attempting to remove more salt via a third round of dialysis led to instability and significant degradation (Figure S30). We believe that some amount of salt helps to stabilize these highly charged compounds.

We first examined the binding affinity toward 4′,6-diamidino-2-phenylindole (DAPI),27 which is a rod-like dicationic guest. Due to the poor solubility of the DOACs at pH 7.4, we studied the binding of DAPI with the extended sulfo-pillar[6]arenes at both pH 3.0 and 7.4 in citrate buffer saline (CBS) and phosphate buffer saline (PBS), respectfully. The pH and buffer showed minimal effect on the emission and excitation of the host–dye complexes (Figure 3a). When complexed with the hosts, DAPI showed an increase in emission at 450 nm (λex = 360 nm) at both pH 7.4 and 3.0, while the hosts by themselves showed no emission at these wavelengths (Figure S33). Through the direct titration of host into DAPI, the dissociation constant (Kd) for the host–dye complex was determined at both pH values (Figures 3b, S34, and S35). The change in pH and buffer had no effect on the Kd values of the respective host–dye complex, which is consistent with the hosts’ highly acidic sulfate groups remaining anionic even at pH 3.0 (Table 1.). All three hosts showed nanomolar affinity to DAPI with the A1A2sP6 displaying the strongest binding of 15 ± 4 nM. The parent host sP6 and A1sP6 showed comparable strength to each other. The increase in the affinity of DAPI to A1A2sP6 over the other hosts demonstrates the effect of host extension on guest affinity.

Figure 3.

Figure 3

Open in a new tab

Fluorescence assays determine dissociation constants for DAPI and DOACs. (a) Emission (λex = 360 nm, solid lines) and excitation (λem = 450 nm, dashed lines) of DAPI (100 nM), sP6–DAPI (1 μM; 100 nM), A1sP6–DAPI (1 μM; 100 nM), and A1A2sP6–DAPI (1 μM; 100 nM) at pH 3.0 (10 mM CBS) and pH 7.4 (10 mM PBS). (b) Exemplary direct binding titration of DAPI (50 nM) into A1A2sP6 (1 μM – 0.5 nM) at pH 3.0 (10 mM CBS) (λex = 360 nm and λem = 450 nm). Reported dissociation constants and R2 are for the exemplary set of triplicates. (c) Exemplary indicator displacement titration of dabigatran (10 μM – 5 nM) into A1A2sP6–DAPI (62.5; 50 nM) at pH 3 (10 mM CBS) (λex = 360 nm and λem = 450 nm). Reported dissociation constants and R2 are for the exemplary set of triplicates. (d) Representation of the displacement of DAPI from the host–DAPI complex by a DOAC guest, leading to a decrease in emission; negative charges represent sulfate groups.

Table 1. Dissociation Constants (Kd, nM) Determined by Fluorescence-Based Assaya.

  sP6 A1sP6 A1A2sP6
DAPIb 40 ± 10 50 ± 10 17 ± 5
DAPIc 40 ± 8 35 ± 7 15 ± 4
dabigatranc 21 ± 2 32 ± 4 27 ± 3
apixabanc,d,e n.b. n.b. n.b.
betrixabanc,d 1400 ± 200 1000 ± 300 230 ± 40
edoxabanc,d 2000 ± 300 800 ± 200 1300 ± 200
Open in a new tab
a

Binding constants were determined assuming a 1:1 stoichiometry. All values reported are the average of six measurements with propagated standard error, unless stated otherwise.

b

Assays performed at pH 7.4 in PBS (10 mM).

c

Assays performed at pH 3.0 in CBS (10 mM).

d

≤2.5% DMSO used.

e

No binding observed at 500 μM guest concentration, triplicate measurement.

The dissociation constants of DOACs were determined via an indicator displacement assay using DAPI (Table 1 and Figure 3c–d). Due to previously stated solubility issues, the DOAC competition experiments were conducted at pH 3.0 in 10 mM CBS buffer with ≤2.5% DMSO (Figures S36–S38). Out of the five DOACs studied, sulfo-pillar[6]arene and its extended analogues showed low micromolar to nanomolar affinity toward dabigatran, betrixaban, and edoxaban. Apixaban showed no binding at the tested concentration range (Kd > 500 μM), and rivaroxaban was not soluble in the assay conditions. Based on the similar chemical structures of apixaban and rivaroxaban, it is unlikely that rivaroxaban would be a strong binder. The strongest affinity was observed for dabigatran, which displayed comparable nanomolar binding to the three hosts. The increasing hydrophobic area altered the binding to betrixaban and edoxaban. Betrixaban had the strongest binding to the A1A2 extended cavity with a Kd = 230 ± 40 nM, a 6- and 4-fold increase over the parent host and the A1 extension, respectively. The A1 extension led to a greater than 2.5-fold increase in affinity toward edoxaban, relative to the parent sulfo-pillar[6]arene.

The recognition of DOACs by extended sulfo-pillar[6]arenes is enthalpically driven. The thermodynamic parameters of pillararene–DOAC binding were determined by using isothermal titration calorimetry (ITC). Each sulfo-pillar[6]arene was titrated into the DOAC in 10 mM CBS buffer (pH 3.0); to aid in solubility, 0.5% DMSO was used for edoxaban and betrixaban, in these instances, a control for the DMSO was run and subtracted from the respective thermograms (Figure S39). Apixaban and rivaroxaban were excluded from this study as they showed either no binding (Table 1) or solubility issues in the competitive displacement assay. Sulfo-pillar[6]arene and its extended analogues showed low micromolar to high nanomolar affinity for all three DOACs tested (Table 2, Figures 4a–b, and S40–S42).

Table 2. Thermodynamic Parameters of DOACs and Sulfo-Pillar[6]arene Recognition Determined by Isothermal Calorimetrya.

  host Kd (nM) ΔG (kcal/mol) ΔH (kcal/mol) –TΔS (kcal/mol)
betrixabanb sP6 3000 ± 400 –7.7 ± 0.1 –7.5 ± 0.1 –0.2 ± 0.2
  A1sP6 2000 ± 100 –7.90 ± 0.06 –7.73 ± 0.06 –0.17 ± 0.09
  A1A2sP6 1500 ± 100 –8.08 ± 0.09 –7.7 ± 0.1 –0.3 ± 0.1
edoxabanb sP6 2900 ± 200 –7.67 ± 0.07 –7.39 ± 0.07 –0.3 ± 0.2
  A1sP6 1370 ± 70 –8.14 ± 0.05 –8.82 ± 0.05 0.69 ± 0.07
  A1A2sP6 3800 ± 300 –7.52 ± 0.08 –5.78 ± 0.07 –1.7 ± 0.1
dabigatran sP6 150 ± 20 –9.5 ± 0.1 –11.67 ± 0.09 2.2 ± 0.2
  A1sP6 150 ± 10 –9.47 ± 0.07 –10.24 ± 0.05 0.77 ± 0.08
  A1A2sP6 370 ± 40 –8.93 ± 0.09 –9.96 ± 0.08 1 ± 0.1
Open in a new tab
a

All assays were performed in CBS solution (10 mM, pH 3.0) at 30 °C. One site binding model was used to fit the data. All values are reported as the average of two replicates with propagated standard error.

b

Guest contains 0.5% DMSO.

Figure 4.

Figure 4

Open in a new tab

Thermodynamic parameters of sulfo-pillar[6]arene–DOAC binding determined by isothermal calorimetry. (a) Plot of DP versus time from the titration of A1sP6 (250 μm) into dabigatran (25 μm) in CBS (10 mM, pH 3.0) at 30 °C. (b) Plot of ΔH versus molar ratio of A1sP6 to dabigatran; thermodynamic parameters of the respective replicate. (c) Cartoon representation of the interaction of extended sulfo-pillar[6]arenes and dabigatran; negative charges represent sulfate groups.

Compared to the dissociation constants obtained from the competitive displacement assays, the values from the ITC experiments are 1–7-fold weaker; however, the general trend in the host–DOAC preference is the same. This decrease in affinity might be in part due to the increased concentration of the host during ITC vs fluorescence experiments (250 – 500 μM vs 62.5 – 125 nM), which may lead to aggregation of the titrant. Overall, all pillararene–DOAC complexes were enthalpically driven with minor entropic contributions.

The A1 extension and the parent sP6 had identical affinity for dabigatran but different entropic and enthalpic contributions. The parent sP6 had an unfavorable entropic contribution of 2.2 ± 0.2 kcal·mol–1; the A1 analogue reduced this unfavorability to 0.77 ± 0.08 kcal·mol–1. The increased hydrophobic surface area increased the entropic favorability by 1.4 kcal·mol–1, equally decreasing the enthalpic contribution. This apparent enthalpy–entropy compensation might be due to the increased depth of the A1 host, which prevents optimal alignment for cation–anion interactions between the benzimidazole moiety of dabigatran and the sulfates on the upper rim of A1sP6 (Figure 4c). The A1A2 extension had over 2-fold weaker binding to dabigatran, with an enthalpic loss of 1.7 kcal·mol–1 and an entropic gain of 1.2 kcal·mol–1 with respect to parent sP6. Similar to A1sP6, the enthalpic loss maybe due to the less-than-optimal alignment of cation–anion interactions (Figure 4c).

As observed in the indicator displacement assay, edoxaban and betrixaban showed preferential binding to A1sP6 and A1A2sP6, respectively. The binding of edoxaban to sulfo-pillar[6]arene (Kd = 2900 ± 200 nM) increased 2-fold with the A1 increased surface area (Kd = 1370 ± 70 nM) but decreased with the A1A2 (Kd = 3800 ± 300 nM). This increase in affinity toward A1sP6 is due to the increased enthalpic contribution of −8.82 ± 0.05 kcal·mol–1, which is 1.43 and 3.04 kcal·mol–1 greater than that of sP6 and A1A2sP6, respectively. While the A1sP6 host had the most enthalpic contribution, it also had the most unfavorable entropy (0.69 ± 0.07 kcal·mol–1). Betrixaban favored A1A2sP6 (Kd = 1500 ± 100 nM), with the A1 and A1A2 functionalization leading to a subtle increase in enthalpy relative to the parent sP6 and no notable change in entropy with the reported values being within error. These subtle changes in betrixaban binding indicate that the DOAC binds the three hosts in a similar fashion.

Extended sulfo-pillar[6]arenes form a complex with the benzamidinium moiety of betrixaban. The binding orientation of betrixaban and the extended sulfo-pillar[6]arenes was determined using 1H NMR by titrating the host into 0.5 mM betrixaban in deuterated CBS buffer at pD 3.0 with 1% DMSO, monitoring the chemical shift change of betrixaban (Figure 5). The NMR signals were assigned from 2D correlation spectroscopy (COSY) (Figures 5a–b, and S44–S46). When 0.5 equiv (0.25 mM) sP6 is added, the two singlets of the benzamidinium methyl resonances at 3.32 and 3.10 ppm converge and move upfield to 1.68 ppm, indicating the inclusion of the amidinium moiety in the hydrophobic cavity of sulfo-pillar[6]arene (Figures 5c, and S47). In addition to the shielding effect observed with the methyl protons, the signals of the aromatic doublet (b) move upfield from 7.70 to 5.88 ppm, further confirming the inclusion in the cavity of sP6. However, the aromatic doublet (c) only moves 0.36 ppm upfield, showing a much smaller shielding effect relative to that of (b) (1.82 ppm shift). When 1 equiv (0.5 mM) of sP6 is added, the amidinium methyl resonances move further upfield by 0.23 ppm. While the aromatic protons (b) adjacent to the amidinium move an additional 0.30 ppm, the more distant aromatic protons (c) move only 0.06 ppm upfield. Minor downfield shifts are noted in the remaining protons of betrixaban, but their assignment is ambiguous in the pillararene complex, except for the methoxy singlet at 3.95 ppm, which has a very minor downfield shift (0.02 ppm) in the presence of sP6. From the chemical shift changes, it is clear that sP6 encapsulates the cationic benzamidinium moiety of betrixaban and that it forms a 1:1 complex.

Figure 5.

Figure 5

Open in a new tab

(a) Chemical structure of betrixaban with key residues labeled. (b) COSY data of 1.0 mM sP6 0.5 mM betrixaban in 20 mM deuterated CBS buffer (pD 3.0) with 1% DMSO. Red box indicates the correlation between aromatic protons b and c of betrixaban. (c). 1H NMR titration of sP6 (0 – 1 mM) into betrixaban (0.5 mM) in 20 mM deuterated CBS buffer (pD 3.0) with 1% DMSO. Citrate and D2O signals are removed for clarity; see Figure S47 for the full spectrum.

Increasing the aromatic surface area increases the host–guest interactions. In the 1H NMR titrations, A1sP6 and A1A2sP6 overall have more broadening of the host and guest signals upon complexation. This causes chemical shift tracking to be more obscure (Figures 6a, S48–S50). However, it is clear from the upfield shift of an aromatic proton (b) and the methyl groups of amidinium (a) that all three hosts bind in a similar fashion. In the case of A1sP6, the singlet from the amidinium methyl moieties (a) is shifted less upfield than with sP6 (Figure 6a); this would indicate that the amidinium is not as shielded by the A1sP6 cavity. The aromatic proton (b) moves very minorly upfield, likely experiencing a similar environment between A1sP6 and the sP6. This is also seen with A1A2sP6, where with respect to sP6, the amidinium methyl resonances (a) are less upfield, but the aromatic proton (b) is only minorly affected (Figure 6a). From this, we propose that the amidinium moiety of betrixaban is inside the cavity and interacts with the sulfates on the bottom rim of the host (Figure 6b). This would cause the remaining aromatic residues of betrixaban to have more interaction with the hosts. Due to broadening, these signals could not be assigned. However, there is an overall upfield movement in the functionalized host relative to that of sP6 (Figure 6a, gold stars), possibly due to increased interactions. This proposed binding motif is supported by the general trend observed in binding affinity between the three hosts, where increasing surface area (A1A2sP6) increased the affinity 6-fold (Table 1). Rudimentary molecular modeling was used to visualize the sulfo-pillar[6]arene complexes with betrixaban. Betrixaban was oriented in the host cavity based off the respective NMR titration data, and the complex was minimized in Maestro 13.8 using OPLS-2005 (Figures 6c and S51). The minimized structures show an increased interaction between the pendant arm of A1sP6 and A1A2sP6 with betrixaban as proposed.

Figure 6.

Figure 6

Open in a new tab

Functionalized analogues have more interactions with betrixaban. (a) 1H NMR (bottom to top): betrixaban (0.5 mM), betrixaban (0.5 mM) with sP6 (1 mM), betrixaban (0.5 mM) with A1sP6 (1 mM), and betrixaban (0.5 mM) with A1A2sP6 (1 mM). All in 20 mM deuterated CBS buffer (pD 3.0) with 1% DMSO. Citrate and D2O signals are removed for clarity; see Figure S50 for the full spectrum. (b) Perspective drawing of potential increased interactions between betrixaban and functionalized sulfo-pillar[6]arenes; negative charges represent sulfate groups. (c) Molecular models of (left to right) sP6–betrixaban, A1sP6–betrixaban, and A1A2sP6–betrixaban. Models made in Maestro 13.8, structures minimized with OPLS-2005.

The host A1A2sP6 is inherently fluorescent and shows a dose-dependent response to betrixaban binding. The incorporated terphenyl moiety gives rise to a substantial increase in fluorescence relative to that of sP6 and the A1 derivative (Figures 7a and S52). Using this new property, we sought to explore how the binding of A1A2sP6 to betrixaban was affected by the presence of a complex biological medium, such as plasma. Betrixaban was titrated into A1A2sP6 (Figures 7b and S53a–b) in dilute bovine plasma causing the emission of A1A2sP6 to decrease. A control experiment in the absence of A1A2sP6 (Figure S53c–d) showed no significant change in the autofluorescence signal of the plasma blank, indicating that A1A2sP6 is the fluorescent component being monitored. A dose–response was plotted using the emission at 330 nm (Figures 7c and S54a), and from this, a dissociation constant was estimated to be 7 ± 2 μM. The control titration in the absence of A1A2sP6 (Figures 7d and S54b) showed no significant response.

Figure 7.

Figure 7

Open in a new tab

Dose-dependent decrease in emission of A1A2sP6 toward betrixaban in bovine plasma. (a) Emission spectra of sP6, A1sP6, and A1A2sP6 (10 μM in 10 mM CBS, pH 3.0, λex = 270 nm) collected on a BioTek Cytation-5 multimode reader plate reader. (b) Direct titration of betrixaban (180 μM, black trace) into A1A2sP6 (90 μM, red trace) in dilute bovine plasma (10%, blue trace). A1A2sP6 was held constant and betrixaban was varied (180–0.7 μM, gray traces). All contained 1.8% DMSO to aid in solubility; emission was monitored at λex = 250 nm. A total of four technical replicates were collected. Data are reported as mean values with error bars (standard deviation). (c) Dose-dependent response on A1A2sP6 emission (λex = 250 nm) by betrixaban at λem = 330 nm. The data were fitted to the 1:1 binding isotherm to determine the binding constant. (d) Control titration in the absence of A1A2sP6 (from duplicate titrations).

Discussion

Herein, we report the first functionalized derivatives of sP6 and their molecular recognition properties toward DOACs. A1sP6 and A1A2sP6 are the first reported mono- or difunctionalized water-soluble pillar[6]arenes. Different water-solubilizing groups have been utilized to functionalize pillar[6]arene in a global way such that all 12 positions are modified uniformly.35 Additionally, a pyrene has been incorporated into the methylene bridge of water-soluble pillar[6]arene in a 3-fold symmetrical fashion.36 However, there are no reports of a water-soluble pillar[6]arene that is not symmetrical (i.e., contains a mono- or difunctionalization in its core scaffold). This vacancy in the field is also present in water-soluble pillar[5]arene research where examples of mono- or difunctionalization is limited.37−42 A water-soluble meso-TPE-functionalized pillar[5]arene diversified the core scaffold through a McMurray coupling at the methylene bridge.37 Additionally, a water-soluble pillar[4]arene[1]quinone had a dissymmetrical scaffold.38 The remaining reported functionalized water-soluble pillar[5]arenes have the diversifying element on an elongated alkyl-based appendage, removed from the core structure.39−42 With these novel extended sulfo-pillar[6]arenes, we sought to explore the change in the molecular recognition properties relative to the unfunctionalized sulfo-pillar[6]arene.

The elongated hydrophobic surface of the extended sulfo-pillar[6]arenes resulted in stronger affinity toward the DOACs edoxaban and betrixaban. Betrixaban bound A1A2sP6 with a 6-fold stronger binding affinity than the unfunctionalized parent sP6, and similarly, edoxaban bound A1sP6 2.5-fold stronger than sP6. The preferential binding of edoxaban to A1sP6 was due to a more favorable enthalpy, likely due to improved size complementarity with the A1 extension. All host–guest complexes were enthalpically driven with minor entropic contributions. The high binding affinities we observed (Tables 1 and 2) demonstrate that the collapsed state seen in the crystal structure for intermediate 8 (Figures S27 and S28) is unlikely to exist in solution for A1A2sP6; this makes sense since the finished host is decorated with 12 mutually repulsive anionic sulfates in a way that would prevent host collapse.

There is limited literature precedent for the direct binding affinity of a DOAC to a host. Researchers have looked at the effect β-cyclodextrin has on the binding between human serum albumin and the pro-drug dabigatran etexilate.43 In the presence of β-cyclodextrin, the affinity for human serum albumin toward dabigatran etexilate decreases from 5.95 × 103 M–1 (Kd ∼ 0.2 mM) to 1.02 × 103 M–1 (Kd ∼ 1 mM), indicating that β-cyclodextrin was competing in the binding event. However, the equilibrium constant for the β-cyclodextrin–dabigatran etexilate complex was not reported.43 In 2023, the modified procoagulant β-cyclodextrin, OKL-1111, was reported as a potential reversal agent for DOACs.22 The mechanism of the reversal is unclear, and as stated by the researchers, it is unlikely that it acts through the sequestration of circulating DOAC in the blood stream. This is based on the low affinity of the DOACs for OKL-1111, which binds with dissociation constants (Kd) of >6 mM for dabigatran, >2 mM for apixaban, >50 μM for rivaroxaban, and 25 μM for exodaban. Betrixaban was not included in this study. Our extended sulfo-pillar[6]arene hosts range from 25 to 200,000 times stronger than OKL-1111 at binding DOAC targets. Additionally, the affinities we determined for extended sulfo-pillar[6]arenes are obtained in the presence of 137 mM NaCl, which closely resemble the concentration of salt circulating in the blood stream; the reported OKL-1111 affinities were obtained in the absence of physiological salt concentrations.

A1A2sP6 binds the DOAC betrixaban in biological media. Bovine plasma contains a mixture of proteins, salts, small molecules, lipids, and various other species that can perturb the sequestration of an analyte by a host molecule. By using the inherent fluorescence of A1A2sP6, we have been able to show that A1A2sP6 binds betrixaban in bovine plasma with a physiologically relevant Kd (7 μM). This illustrates the potential that A1A1sP6 has in both sequestration- and fluorescence-based studies.

Conclusions

Sulfo-pillar[6]arene was an ultrahigh-affinity, salt-tolerant, supramolecular host that previously lacked synthetic functionalizability, an important factor in tuning affinities and in pursuing applications. Through the development of an A1 and A1A2 synthetic pathway, we modified the host cavity in ways that directly impact guest binding and selectivity. A1sP6 and A1A2sP6 offer a larger surface area, allowing more host–guest interactions for guests with complementary shapes and functionality. The A1A2 functionalization had a 6-fold improved potency for betrixaban, which had been previously unexplored in host–guest chemistry. Previous work has highlighted the potential for sulfo-pillar[6]arene as a supramolecular reversal agent for neuromuscular blocking agents and opioids.23,25,26 We have shown that these hosts have high affinity toward an underexplored drug class, DOACs, in biologically relevant salty media and in bovine plasma. Through synthetic functionalization, the A1A2sP6 analogue has opened the door to fluorescent applications of sulfo-pillar[6]arenes. These developed methods can easily be adapted in future endeavors to diversify the parent host and allow researchers to explore the full potential of sulfo-pillar[6]arenes.

Acknowledgments

The authors thank the Natural Sciences and Engineering Research Council of Canada (NSERC, RGPIN-2019-04806) for financial support. C.R.W. thanks NSERC for the Postgraduate Scholarship – Doctoral (PGS D). We thank Alisdair Boraston for providing access to an isothermal titration calorimeter and Cornelia Bohne for providing access to a fluorometer and spectrophotometer. We thank Christopher Barr for helpful discussion on NMR. We also thank CAMTEC for the use of shared facilities. We thank Jayachandran Kizhakkedathu and Iren Constantinescu for helpful discussion and supplementary investigation. Funding for the Bruker Venture was supported by NSF MRI Award CHE-2214606.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.4c03905.

  • Synthesis and characterization data, crystallography, fluorescence experiments, ITC, NMR titrations, and molecular modeling (PDF)

Author Contributions

All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.

Supplementary Material

ja4c03905_si_001.pdf (7MB, pdf)

References

  1. Kim H. K.; Nelson L. S. Reducing the harm of opioid overdose with the safe use of naloxone: a pharmacologic review. Expert Opin. Drug Saf. 2015, 14 (7), 1137–1146. 10.1517/14740338.2015.1037274. [DOI] [PubMed] [Google Scholar]
  2. Yin H.; Zhang X.; Wei J.; Lu S.; Bardelang D.; Wang R. Recent advances in supramolecular antidotes. Theranostics 2021, 11 (3), 1513–1526. 10.7150/thno.53459. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Deng C.-L.; Murkli S. L.; Isaacs L. D. Supramolecular hosts as in vivo sequestration agents for pharmaceuticals and toxins. Chem. Soc. Rev. 2020, 49 (21), 7516–7532. 10.1039/D0CS00454E. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Welliver M.; McDonough J.; Kalynych N.; Redfern R. Discovery, development, and clinical application of sugammadex sodium, a selective relaxant binding agent. Drug Des. Dev. Ther. 2009, 2, 49–59. 10.2147/dddt.s2757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bom A.; Bradley M.; Cameron K.; Clark J. K.; Van Egmond J.; Feilden H.; MacLean E. J.; Muir A. W.; Palin R.; Rees D. C.; Zhang M.-Q. A Novel Concept of Reversing Neuromuscular Block: Chemical Encapsulation of Rocuronium Bromide by a Cyclodextrin-Based Synthetic Host. Angew. Chem., Int. Ed. 2002, 41 (2), 266–270. . [DOI] [PubMed] [Google Scholar]
  6. Hristovska A.-M.; Duch P.; Allingstrup M.; Afshari A. Efficacy and safety of sugammadex versus neostigmine in reversing neuromuscular blockade in adults. Cochrane Database Syst. Rev. 2017, 8 (9), CD012763. 10.1002/14651858.CD012763. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Geng W.-C.; Sessler J. L.; Guo D.-S. Supramolecular prodrugs based on host–guest interactions. Chem. Soc. Rev. 2020, 49 (8), 2303–2315. 10.1039/C9CS00622B. [DOI] [PubMed] [Google Scholar]
  8. Yu G.; Chen X. Host-Guest Chemistry in Supramolecular Theranostics. Theranostics 2019, 9 (11), 3041–3074. 10.7150/thno.31653. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Webber M. J.; Langer R. Drug delivery by supramolecular design. Chem. Soc. Rev. 2017, 46 (21), 6600–6620. 10.1039/C7CS00391A. [DOI] [PubMed] [Google Scholar]
  10. Ma X.; Zhao Y. Biomedical Applications of Supramolecular Systems Based on Host-Guest Interactions. Chem. Rev. 2015, 115 (15), 7794–7839. 10.1021/cr500392w. [DOI] [PubMed] [Google Scholar]
  11. Haerter F.; Simons J. C. P.; Foerster U.; Moreno Duarte I.; Diaz-Gil D.; Ganapati S.; Eikermann-Haerter K.; Ayata C.; Zhang B.; Blobner M.; Isaacs L.; Eikermann M. Comparative Effectiveness of Calabadion and Sugammadex to Reverse Non-depolarizing Neuromuscular-blocking Agents. Anesthesiology 2015, 123 (6), 1337–1349. 10.1097/ALN.0000000000000868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Yin H.; Cheng Q.; Bardelang D.; Wang R. Challenges and Opportunities of Functionalized Cucurbiturils for Biomedical Applications. JACS Au 2023, 3 (9), 2356–2377. 10.1021/jacsau.3c00273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Nimse S. B.; Kim T. Biological applications of functionalized calixarenes. Chem. Soc. Rev. 2013, 42 (1), 366–386. 10.1039/C2CS35233H. [DOI] [PubMed] [Google Scholar]
  14. Schneider C.; Bierwisch A.; Koller M.; Worek F.; Kubik S. Detoxification of VX and Other V-Type Nerve Agents in Water at 37 °C and pH 7.4 by Substituted Sulfonatocalix[4]arenes. Angew. Chem., Int. Ed. 2016, 55 (41), 12668–12672. 10.1002/anie.201606881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Murkli S.; Klemm J.; Brockett A. T.; Shuster M.; Briken V.; Roesch M. R.; Isaacs L. In Vitro and In Vivo Sequestration of Phencyclidine by Me4Cucurbit[8]uril. Chem. Eur. J. 2021, 27 (9), 3098–3105. 10.1002/chem.202004380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ho K. H.; van Hove M.; Leng G. Trends in anticoagulant prescribing: a review of local policies in English primary care. BMC Health Serv. Res. 2020, 20 (1), 279. 10.1186/s12913-020-5058-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Shaw J. R.; Siegal D. M. Pharmacological reversal of the direct oral anticoagulants-A comprehensive review of the literature. Res. Pract. Thromb. Haemostasis. 2018, 2 (2), 251–265. 10.1002/rth2.12089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Chen A.; Stecker E.; Warden B. A. Direct Oral Anticoagulant Use: A Practical Guide to Common Clinical Challenges. J. Am. Heart Assoc. 2020, 9 (13), e017559 10.1161/jaha.120.017559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Rose D. K.; Bar B. Direct Oral Anticoagulant Agents: Pharmacologic Profile, Indications, Coagulation Monitoring, and Reversal Agents. J. Stroke Cerebrovasc. Dis. 2018, 27 (8), 2049–2058. 10.1016/j.jstrokecerebrovasdis.2018.04.004. [DOI] [PubMed] [Google Scholar]
  20. Cuker A.; Burnett A.; Triller D.; Crowther M.; Ansell J.; Van Cott E. M.; Wirth D.; Kaatz S. Reversal of direct oral anticoagulants: Guidance from the Anticoagulation Forum. Am. J. Hematol. 2019, 94 (6), 697–709. 10.1002/ajh.25475. [DOI] [PubMed] [Google Scholar]
  21. Tangedal K.; Semchuk W.; Bolt J. How Should Canadian Pharmacy Departments Utilize Idarucizumab?. Can. J. Hosp. Pharm. 2016, 69 (5), 429–430. 10.4212/cjhp.v69i5.1601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Meijers J. C. M.; Bakhtiari K.; Zwiers A.; Peters S. L. M. OKL-1111, A modified cyclodextrin as a potential universal reversal agent for anticoagulants. Thromb. Res. 2023, 227, 17–24. 10.1016/j.thromres.2023.05.003. [DOI] [PubMed] [Google Scholar]
  23. Xue W.; Zavalij P. Y.; Isaacs L. Pillar[n]MaxQ: A New High Affinity Host Family for Sequestration in Water. Angew. Chem., Int. Ed. 2020, 59 (32), 13313–13319. 10.1002/anie.202005902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Isaacs L. D., Xue W.. Sulfated pillararenes, methods of making same, and uses thereof. U.S. Patent 17/822, 863, 2023. https://patents.google.com/patent/US20230174468A1/en.
  25. Brockett A. T.; Xue W.; King D.; Deng C.-L.; Zhai C.; Shuster M.; Rastogi S.; Briken V.; Roesch M. R.; Isaacs L. Pillar[6]MaxQ: A potent supramolecular host for in vivo sequestration of methamphetamine and fentanyl. Chem 2023, 9 (4), 881–900. 10.1016/j.chempr.2022.11.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Zhang W.; Bazan-Bergamino E. A.; Doan A. P.; Zhang X.; Isaacs L. Pillar[6]MaxQ functions as an in vivo sequestrant for rocuronium and vecuronium. Chem. Commun. 2024, 60 (32), 4350–4353. 10.1039/D4CC00772G. [DOI] [PubMed] [Google Scholar]
  27. King D.; Wilson C. R.; Herron L.; Deng C.-L.; Mehdi S.; Tiwary P.; Hof F.; Isaacs L. Molecular recognition of methylated amino acids and peptides by Pillar[6]MaxQ. Org. Biomol. Chem. 2022, 20 (37), 7429–7438. 10.1039/D2OB01487D. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. King D.; Deng C.-L.; Isaacs L. Pillar[n]MaxQ sensing ensemble: Application to world anti-doping agency banned substances. Tetrahedron 2023, 145, 133607. 10.1016/j.tet.2023.133607. [DOI] [Google Scholar]
  29. Wilson C. R.; Chen E. F. W.; Puckett A. O.; Hof F. Ethoxypillar[6]arene. Org. Synth. 2022, 99, 125–138. 10.15227/orgsyn.099.0125. [DOI] [Google Scholar]
  30. Ogoshi T.; Kayama H.; Yamafuji D.; Aoki T.; Yamagishi T.-A. Supramolecular polymers with alternating pillar[5]arene and pillar[6]arene units from a highly selective multiple host–guest complexation system and monofunctionalized pillar[6]arene. Chem. Sci. 2012, 3 (11), 3221–3226. 10.1039/c2sc20982a. [DOI] [Google Scholar]
  31. Ogoshi T.; Yamafuji D.; Kotera D.; Aoki T.; Fujinami S.; Yamagishi T.-A. Clickable di- and tetrafunctionalized pillar[n]arenes (n = 5, 6) by oxidation-reduction of pillar[n]arene units. J. Org. Chem. 2012, 77 (24), 11146–11152. 10.1021/jo302283n. [DOI] [PubMed] [Google Scholar]
  32. Rashvand Avei M.; Etezadi S.; Captain B.; Kaifer A. E. Visualization and quantitation of electronic communication pathways in a series of redox-active pillar[6]arene-based macrocycles. Commun. Chem. 2020, 3 (1), 117. 10.1038/s42004-020-00363-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Zhu Y.; Escorihuela J.; Wang H.; Sue A. C.-H.; Zuilhof H. Tunable Supramolecular Ag+-Host Interactions in Pillar[n]arene[m]quinones and Ensuing Specific Binding to 1-Alkynes. Molecules 2023, 28 (20), 7009. 10.3390/molecules28207009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Demay-Drouhard P.; Du K.; Samanta K.; Wan X.; Yang W.; Srinivasan R.; Sue A. C.-H.; Zuilhof H. Functionalization at Will of Rim-Differentiated Pillar[5]arenes. Org. Lett. 2019, 21 (11), 3976–3980. 10.1021/acs.orglett.9b01123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Tang R.; Ye Y.; Zhu S.; Wang Y.; Lu B.; Yao Y. Pillar[6]arenes: From preparation, host-guest property to self-assembly and applications. Chin. Chem. Lett. 2023, 34 (3), 107734. 10.1016/j.cclet.2022.08.014. [DOI] [Google Scholar]
  36. Ling L.; Zhao Z.; Mao L.; Wang S.; Ma D. Water-soluble pillar[6]arene bearing pyrene on alternating methylene bridges for direct spermine sensing. Chem. Commun. 2023, 59 (95), 14161–14164. 10.1039/D3CC05094G. [DOI] [PubMed] [Google Scholar]
  37. Tian X.; Zuo M.; Niu P.; Velmurugan K.; Wang K.; Zhao Y.; Wang L.; Hu X.-Y. Orthogonal Design of a Water-Soluble meso-Tetraphenylethene-Functionalized Pillar[5]arene with Aggregation-Induced Emission Property and Its Therapeutic Application. ACS Appl. Mater. Interfaces 2021, 13 (31), 37466–37474. 10.1021/acsami.1c07106. [DOI] [PubMed] [Google Scholar]
  38. Wang J.; Cen M.; Wang J.; Wang D.; Ding Y.; Zhu G.; Lu B.; Yuan X.; Wang Y.; Yao Y. Water-soluble pillar[4]arene[1]quinone: Synthesis, host-guest property and application in the fluorescence turn-on sensing of ethylenediamine in aqueous solution, organic solvent and air. Chin. Chem. Lett. 2022, 33 (3), 1475–1478. 10.1016/j.cclet.2021.08.044. [DOI] [Google Scholar]
  39. Zhang Y.; Chen L.; Du X.; Yu X.; Zhang H.; Meng Z.; Zheng Z.; Chen J.; Meng Q. Selective Fluorescent Sensing for Iron in Aqueous Solution by A Novel Functionalized Pillar[5]arene. ChemistryOpen 2023, 12 (10), e202300109 10.1002/open.202300109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sun J.; Shao L.; Zhou J.; Hua B.; Zhang Z.; Li Q.; Yang J. Efficient enhancement of fluorescence emission via TPE functionalized cationic pillar[5]arene-based host–guest recognition-mediated supramolecular self-assembly. Tetrahedron Lett. 2018, 59 (2), 147–150. 10.1016/j.tetlet.2017.12.009. [DOI] [Google Scholar]
  41. Shurpik D. N.; Aleksandrova Y. I.; Mostovaya O. A.; Nazmutdinova V. A.; Zelenikhin P. V.; Subakaeva E. V.; Mukhametzyanov T. A.; Cragg P. J.; Stoikov I. I. Water-soluble pillar[5]arene sulfo-derivatives self-assemble into biocompatible nanosystems to stabilize therapeutic proteins. Bioorg. Chem. 2021, 117, 105415. 10.1016/j.bioorg.2021.105415. [DOI] [PubMed] [Google Scholar]
  42. Wu X.; Ni M.; Xia W.; Hu X.-Y.; Wang L. A novel dynamic pseudo[1]rotaxane based on a mono-biotin-functionalized pillar[5]arene. Org. Chem. Front. 2015, 2 (9), 1013–1017. 10.1039/C5QO00159E. [DOI] [Google Scholar]
  43. Naik R. S.; Pawar S. K.; Tandel R. D.; Seetharamappa J. Insights in to the mechanism of interaction of a thrombin inhibitor, dabigatran etexilate with human serum albumin and influence of β-cyclodextrin on binding: Spectroscopic and computational approach. J. Mol. Liq. 2018, 251, 119–127. 10.1016/j.molliq.2017.12.056. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ja4c03905_si_001.pdf (7MB, pdf)

Articles from Journal of the American Chemical Society are provided here courtesy of American Chemical Society

RESOURCES