Dynamic Macro- and Microgels Driven by Adenosine Triphosphate-Fueled Competitive Host–Guest Interaction
Supramolecular host–guest systems are generally at thermodynamic equilibrium states or in kinetically trapped states. Herein, we demonstrate the concept of a chemical fuel-driven competitive host–guest assembly featuring autonomous dynamics. The enabling key principle is to design a chemical fuel that possesses a high binding affinity to defeat the guest transiently, undergoes conversion to a waste product, and exhibit weak binding affinity in order to recover the original host–guest pair. By following this principle, adenosine triphosphate (ATP; chemical fuel and competitive guest), biguanidine-functionalized β-cyclodextrin (β-CD; host), and an adamantine species (ADA; guest) have been engineered to construct dynamic adaptive macrogels or chemo-mechanochromic microgels with programmed time domains. Our reported methodology provides temporal control over the host–guest process, representing a conceptually new tool for the design of living supramolecular materials.
Introduction
Host–guest supramolecular interactions are powerful tools that allow access to various multifunctional assemblies and materials. A series of macrocyclic host molecules, including crown ethers, calixarenes, cucurbiturils, pillararenes, and cyclodextrins, have been developed to form stable inclusions with various guest molecules.1–5 In general, incorporating a stimulus-responsive guest molecule can provide a means of tailoring the formation of supramolecular inclusions. A variety of external stimuli, including temperature, pH, redox, and light, have been coupled with the host–guest chemistry to construct stimuli-responsive materials, which have been exploited in a broad range of fields.6–8 Nevertheless, this approach relies on the unique physicochemical properties of guest molecules, hindering the application of many responseless moieties, and thus, represents a bottleneck for diversified host–guest supramolecular materials.
Another common feature of these host–guest supramolecular systems is that they are in a thermodynamic equilibrium state or a formulating kinetically trapped state. Notably, on a conceptual level, these systems are static such that they cannot exhibit self-regulated control of the time domain. Unlike unidirectional stimuli-responsive systems, living systems are chemical fuel-controlled, whereby these dynamic systems can temporarily program their chemical functions and structures.9–11 In recent years, the chemical fuel-driven approach has been successfully coupled to the fields of supramolecular polymerization,12–14 gel materials,15–17 catalysis,18–20 DNA self-assembly,21,22 and colloidal systems.23,24 Besides, chemical fuels have been explored to fascinate assembly with a few macrocyclic hosts to construct macromolecular micelles,25 vesicles,26 and supramolecular crystals,27 performing transient features that are thermodynamically unfavorable. However, using chemical fuels to modulate thermodynamically stable supramolecular host–guest inclusions have never been exploited.
The purpose of this work is to develop a general approach to design dynamic host–guest assemblies using a chemical fuel. Scheme 1 shows this concept of chemical fuel-driven host–guest assembly, compared with the traditional stimuli-responsive host–guest process. For the stimuli-responsive approach, sequential application of “on”- and “off”-triggers is required (Scheme 1a). By contrast, the chemical fuel-driven approach enables the host–guest process to be autonomous and change dynamically, where the chemical fuel functions as a competitive guest (Scheme 1b). The in situ consumption of the fuel leads to dissociation of the transient “occupier” and subsequent recovery of the original supramolecular host–guest pair with time. During this process, the use of stimuli-responsive guests is circumvented by capitalizing on a chemical fuel that exhibits time-programmable competitive binding behavior.
Results and Discussion
As a proof of concept, adenosine triphosphate (ATP), polymers bearing biguanidine-functionalized β-cyclodextrin (PCD; Mn = 1.4 × 104, Ɖ = 1.3, three receptors per chain), and adamantine guests (PADA; Mn = 9 × 103, Ɖ = 1.05, eight donors per chain) were selected (Figure 1a and Supporting Information Figures S1–S6). The two-dimensional (2D) nuclear Overhauser effect spectroscopy (NOESY) technique was used to elucidate the competitive supramolecular process. As shown in Supporting Information Figure S7, the spectrum exhibited apparent nuclear Overhauser effect (NOE) correlations between the protons of β-CD (d = 3.80–3.95 ppm) and the adamantine (ADA) species moiety (d = 1.6 ppm, 2.0–2.2 ppm), suggesting that the ADA moiety was deeply included in the cavity of β-CD. After the addition of ATP, the NOE correlations between the protons of β-CD and the adamantyl moiety disappeared, whereas new cross-correlation peaks appeared, assigned to the adenine moiety of ATP and the β-CD moiety (Figure 1b). In contrast, the introduction of adenosine monophosphate (AMP) did not impact the NOE correlations between the protons of β-CD and the ADA moiety, and no NOE correlations were found between the adenine moiety of AMP and the β-CD moiety. Isothermal titration calorimetry (ITC) analysis revealed that the binding affinity of ATP/PCD was as high as ∼106 M−1 (Figure 1c and Supporting Information Figure S8), which is much larger than that of adamantine (∼104 M−1). Importantly, its hydrolysate (AMP) showed a much lower affinity to the host (∼102 M−1), allowing the thermodynamically stable supramolecular inclusion to be reformed. Such a discrepancy in association constants can be ascribed to the decreased supramolecular interaction, as we have discussed previously.26,27 Therefore, we deduced that ATP/AMP/PCD/PADA might represent a potential candidate model to demonstrate the concept of a chemical fuel-driven competitive host–guest process as the binding affinity (Figure 1d), following the order KATP/PCD ≫ KPADA/PCD > KADP/PCD > KAMP/PCD.
With increasing concentrations of PCD and PADA, a supramolecular gel can be obtained due to the host–guest interaction. Notably, potato apyrase was pre-embedded, which did not affect the rheological properties ( Supporting Information Figure S9). Upon the addition of 40 μM ATP, a gel–sol transition was observed in <30 s from the tube inversion test (Figure 2a). At this time, the sample existed as a viscous sol with nonexistent or low yield stress. It took ∼140 min for this sol to revert to the gel state, reminiscent of thixotropy. Unlike many current chemical fuel-driven sol–gel–sol systems,15–17 the ATP-driven competitive host–guest approach afforded an unusual gel–sol–gel supramolecular system. The kinetics of the gel–sol–gel transition were followed by time sweep rheology. From Figure 2b, the disassembly process was rapid, with storage modulus (G′) decreasing from 86 to 0.3 Pa (<1 min), while the re-assembly process lasted a long time (∼140 min). Interestingly, the final G′ value was ∼132.6 Pa, which was 1.5 times that of the original stiffness. In the control experiment, the addition of AMP or Pi induced only a slight decrease in gel strength due to the salt-out effect ( Supporting Information Figure S9). Compared with the fast gelation found in heterogeneous structures, the homogeneous hydrogelation of the network with time may contribute to enhanced stiffness, analogous to annealing.28,29 Consequently, the chemical fuel-driven competitive approach may become a powerful tool to adapt or anneal the properties of host–guest materials, which offer more advantages for transport, pumping, and injection ( Supporting Information Figure S10). Frequency sweep experiments further demonstrated that the initial G′ value had a substantial elastic response and was always higher than the loss modulus (G″) value over the entire range of frequencies (Figure 2c). With ATP addition (∼1 min), G′ was always lower than G″, both of which were frequency dependence, implying a sol state. After a period of ∼140 min, the sample autonomously reverted to the gel state in a frequency-independent manner. Moreover, neither the enzymatic hydrolysate (ADP and AMP) nor other analogs (phosphoric acid, triphosphoric acid, and phytic acid) could trigger the gel–sol–gel transition (Figure 2d).
With a continuous fuel supply, the system can exhibit repetitive gel–sol–gel transitions, and the cycle could be repeated more than 10 times (only four cycles are shown), akin to a sinusoidal wave fashion (Figure 2e). In contrast to other chemical fuel-driven systems, the waste exerts few effects on the cycles. We speculated that the selective recognition nature of the host–guest interaction might reduce the damage of waste products (AMP, Pi) to the cycles.26,27 Moreover, the lifetime of each cycle can be tuned by the concentration of ATP or potato apyrase. As shown in Figure 2f, at higher potato apyrase concentrations, the lifetimes shortened dramatically. The influence of the potato apyrase concentration on the backward contraction process was much more significant, and the whole pulsating period was prolonged. In comparison, the lifetimes were roughly proportional to the amount of fuel added (Figure 2g). As such, the lifetimes showed a negative correlation with the enzyme potato apyrase concentration but a positive correlation with ATP concentration.
Having shown that the chemical fuel-driven competitive host–guest approach can readily construct dynamic macrogels, we further extended this approach to fabricate chemo-mechanochromic microgels. The microgel was synthesized by inverse emulsion polymerization using acrylamide (AM), force-responsive spiropyran-based mechanophore cross-link (SP), adamantyl monomer (ADA), and β-CD monomer (Figure 3a). Owing to the absence of an electron-withdrawing group, the SP mechanophore could sense changes in force but insensitive to light,30,31 serving as a real-time reporter in the chemical fuel-driven process. In addition to chemical cross-links, supramolecular inclusions (ADA/β-CD) incorporated within microgels could serve as additional physical cross-links. These physical cross-links exerted little effect on the original size of the microgels ( Supporting Information Figure S11 and Table S1) but influenced their swelling behavior to a large extent. Herein, the swelling ratios were defined as α = (D1/D0)3, where “D0” and “D1” were the hydrodynamic diameters of the microgels in hexane and water, respectively. From the dynamic light scattering (DLS) study, the hydrodynamic diameters of these microgels with different molar fractions of supramolecular cross-links, D0, were ∼80 ± 15 nm, with a relatively uniform size distribution ( Supporting Information Table S1). After redispersion in water, these microgels with fewer supramolecular cross-links absorbed water and expanded drastically (Figure 3b). In the absence of supramolecular cross-links, microgels with 0.5 mol % chemical cross-links exhibited a 6.0-fold size increase to 480 ± 10 nm within 10 min. Concomitantly, a dramatic color change from translucent to pink was observed visually. The appearance of color changes could be attributed to the SP-to-merocyanine (MC) transition within the microgels due to mechanical swelling.30,31 However, the swelling ratio became smaller with increasing molar ratios of the supramolecular or chemical cross-links. Specifically, microgels with the highest contents of supramolecular cross-links (∼10 mol %) or chemical cross-links (∼6.5 mol %) only expanded in size to 110 ± 10 nm, a 1.5-fold increase (Figures 3b and 3c). No color changes were detected for these microgels with a high density of cross-links.
The time-dependent competitive role of ATP in these microgels was then identified. DLS measurements indicated that with the addition of ATP, a remarkable increase in the swelling ratio was observed for the microgels with different supramolecular cross-links (Figures 3b and 3c). For example, microgels with ∼6 mol % supramolecular cross-links expanded to 420 ± 20 nm with 20 μM ATP addition, accompanied by color changes (Figure 3). Notably, after ATP treatment, the maximum swollen size of these microgels decreased as a function of the molar ratios of the supramolecular cross-linkers. This trend was attributed to the hydrophobic effect of the adamantyl-moieties upon dissociation of the supramolecular cross-links.7 In the control experiments, neither ADP nor AMP could trigger a size increase of the microgels. Transmission electron microscopy (TEM) also validated the significant size change elicited by ATP (Figure 3d), revealing that the initial microgels were spherical structures with a mean size of 70 ± 10 nm ( Supporting Information Figure S11). After the addition of ATP, the maximum size reached ∼380 nm. UV–vis spectroscopy further confirmed the competitive supramolecular process. From Figure 3e, after ATP treatment, the absorption band centered at 540 nm became noticeably more intense, corresponding to an increasing amount of the opened MC form. There was no apparent change in the absorption band at ∼500–600 nm when ADP or AMP was added.
Then the microgel system was pretreated with the enzyme potato apyrase, followed by the addition of the chemical fuel. The microgel system turned from colorless to pink within 30 min after introducing 40 μM ATP, and then they faded back to colorless in ∼210 min (Figure 4a). Laser confocal microscopy (LCM) was employed to track the real-time changes in fluorescence in the system. Before the ATP treatment, no fluorescent microgels (only black background in Figure 4a) could be found, implying that the chemical cross-links were in the form of colorless SP. When ATP was added, a substantial quantity of fluorescent microgels appeared in a short period (red dots), indicating the chemical fuel-driven SP-to-MC transition. Subsequently, these fluorescent red dots vanished gradually over time. The DLS results further revealed that the size evolution process could be divided into two distinct regimes: in the first regime (0–30 min), all microgels self-dilated, increasing from the initial size of 110 ± 20 nm to a maximum value of 410 ± 30 nm; in the second regime, the microgels experienced self-shrinkage back to their initial size at much longer time of ∼210 min. The color change, LCM, and DLS results suggested that there was a time-orchestrated supramolecular closed loop inside the system: first, ATP destroyed the supramolecular inclusion of β-CD/ADA to decrease the density of cross-links, giving rise to swelling; then the ATP consumption process dominated, with a reformation of the β-CD/ADA supramolecular inclusions, accompanied by contraction of the microgels.
Since we considered the “swollen” and “shrunken” states of the transient signals as part of one cycle, we wondered whether a new cycle could restart by resupplying chemical energy. Therefore, the kinetics of system evolution was monitored by DLS and fluorescence emission (FM) studies. As shown in Figure 4b, the variations in the microgel size plotted versus ATP incubation time were similar to a pulse wave. The size cycle could be refueled under chemical energy influx, demonstrating that these microgels automatically exhibited persistent reciprocating “breathing” behavior. The FM curve also assumed a sinusoidal wave fashion by continuous ATP supply, in line with the DLS results ( Supporting Information Figure S12). After several cycles, the minimum fluorescence intensity and hydrodynamic size slightly diverged from those of the original state due to the accumulation of waste (AMP and 2Pi).
DLS was further employed to study the influence of the concentration of building blocks on the lifetimes of the breathing behavior. Similar to that of the macrogel system, at higher ATP concentration, the forward expansion time of the microgel was shortened, while the backward contraction time lengthened (Figure 4c). Fuel influence on the backward contraction process was much more significant than the whole prolonged lifetime. As expected, the lifetime showed a negative correlation with the enzyme potato apyrase concentration (Figure 4d). The higher the concentration of potato apyrase added to the reaction, the larger the lifetime value became. Notably, the molar fraction of the chemical cross-links was critical in adjusting the periodicity and amplitude of the microgels. As shown in Supporting Information Figure S13, when the molar ratio of chemical cross-links was set as 6.5 mol %, neither size increase nor detectable color change was noted due to the high density of the cross-links. It was only by decreasing the content of cross-links (0.5–3.5 mol %) that the fuel-driven self-regulated process could be initiated. For these microgels with moderate cross-link density, the periodicity and amplitude showed positive correlations with the molar fraction of the chemical cross-links.
Conclusion
Overall, we have demonstrated a conceptually different approach for temporal control over the host–guest process, whereby a dynamic adaptive macrogel or a chemo-mechanochromic microgel could be achieved. The dual roles of ATP in this study, chemical fuel and competitive guest, contributed to assembling homogeneous macrogels with improved mechanical properties and endowed the chemo-mechanochromic microgels with a breathing feature. This strategy could be viewed as an essential advancement relative to classical stimuli-responsive host–guest assembly, as it provides strategies for more life-like materials with autonomous behavior. We envision that this approach might serve as a generic method that could be extended to orchestrate the lifetimes of other macrocyclic host-based assemblies, including calixarenes, cucurbiturils, pillararenes, and so on.
Supporting Information
Supporting Information is available and includes detailed synthetic methods and characterizations of the hosts and guests, NMR, ITC, DLS, TEM, and rheological data.
Conflict of Interest
There is no conflict of interest to report.
Funding Information
This project was funded by the Beijing Natural Science Foundation (grant no. 6214041), Fundamental Research Funds for the Central Universities (grant no. BLX201918), China Postdoctoral Science Foundation (grant no. 2020M670176), and the Project of State Key Laboratory of Molecular Engineering of Polymers of Fudan University (grant no. K2021-21).
References
- 1. Liu Z.; Nalluri S. K. M.; Stoddart J. F.Surveying Macrocyclic Chemistry: From Flexible Crown Ethers to Rigid Cyclophanes.Chem. Soc. Rev.2017, 46, 2459–2478. Google Scholar
- 2. Lee J. H.; Park J.; Park J.-W.; Ahn H.-J.; Jaworski J.; Jung J. H.Supramolecular Gels with High Strength by Tuning of Calix[4]arene-Derived Networks.Nat. Commun.2015, 6, 6650. Google Scholar
- 3. Liu J.; Tan C. S. Y.; Yu Z.; Lan Y.; Abell C.; Scherman O. A.Biomimetic Supramolecular Polymer Networks Exhibiting Both Toughness and Self-Recovery.Adv. Mater.2017, 29, 1604951. Google Scholar
- 4. Ogoshi T.; Yamagishi T.-A.; Nakamoto Y.Pillar-Shaped Macrocyclic Hosts Pillar[n]arenes: New Key Players for Supramolecular Chemistry.Chem. Rev.2016, 116, 7937–8002. Google Scholar
- 5. Yin Z.; Song G.; Jiao Y.; Zheng P.; Xu J.-F.; Zhang X.Dissipative Supramolecular Polymerization Powered by Light.CCS Chem.2019, 1, 335–342. Link, Google Scholar
- 6. Yamaguchi H.; Kobayashi Y.; Kobayashi R.; Takashima Y.; Hashidzume A.; Harada A.Photoswitchable Gel Assembly Based on Molecular Recognition.Nat. Commun.2012, 3, 603. Google Scholar
- 7. Miyamae K.; Nakahata M.; Takashima Y.; Harada A.Self-Healing, Expansion–Contraction, and Shape-Memory Properties of a Preorganized Supramolecular Hydrogel through Host–Guest Interactions.Angew. Chem. Int. Ed.2015, 54, 8984–8987. Google Scholar
- 8. Wang S.; Xu Z.; Wang T.; Xiao T.; Hu X.-Y.; Shen Y.-Z.; Wang L.Warm/Cool-Tone Switchable Thermochromic Material for Smart Windows by Orthogonally Integrating Properties of Pillar[6]arene and Ferrocene.Nat. Commun.2018, 9, 1737. Google Scholar
- 9. Ragazzon G.; Prins L. J.Energy Consumption in Chemical Fuel-Driven Self-Assembly.Nat. Nanotechnol.2018, 13, 882–889. Google Scholar
- 10. Leng Z.; Peng F.; Hao X.Chemical-Fuel-Driven Assembly in Macromolecular Science: Recent Advances and Challenges.ChemPlusChem2020, 85, 1190–1199. Google Scholar
- 11. Deng J.; Walther A.ATP‐Responsive and ATP‐Fueled Self‐Assembling Systems and Materials.Adv. Mater.2020, 32, 2002629. Google Scholar
- 12. Dhiman S.; Jain A.; Kumar M.; George S. J.Adenosine-Phosphate-Fueled, Temporally Programmed Supramolecular Polymers with Multiple Transient States.J. Am. Chem. Soc.2017, 139, 16568–16575. Google Scholar
- 13. Deng J.; Walther A.Pathway Complexity in Fuel-Driven DNA Nanostructures with Autonomous Reconfiguration of Multiple Dynamic Steady States.J. Am. Chem. Soc.2020, 142, 685–689. Google Scholar
- 14. Sorrenti A.; Leira-Iglesias J.; Markvoort A. J.; de Greef T. F. A.; Hermans T. M.Non-Equilibrium Supramolecular Polymerization.Chem. Soc. Rev.2017, 46, 5476–5490. Google Scholar
- 15. Boekhoven J.; Hendriksen W. E.; Koper G. J.; Eelkema R.; van Esch J. H.Transient Assembly of Active Materials Fueled by a Chemical Reaction.Science2015, 349, 1075–1079. Google Scholar
- 16. Hu K.; Sheiko S. S.Time Programmable Hydrogels: Regulating the Onset Time of Network Dissociation by a Reaction Relay.Chem. Commun.2018, 54, 5899–5902. Google Scholar
- 17. Wojciechowski J. P.; Martin A. D.; Thordarson P.Kinetically Controlled Lifetimes in Redox-Responsive Transient Supramolecular Hydrogels.J. Am. Chem. Soc.2018, 140, 2869–2874. Google Scholar
- 18. Colomer I.; Morrow S. M.; Fletcher S. P.A Transient Self-Assembling Self-Replicator.Nat. Commun.2018, 9, 2239. Google Scholar
- 19. Biagini C.; Fielden S. D. P.; Leigh D. A.; Schaufelberger F.; Di Stefano S.; Thomas D.Dissipative Catalysis with a Molecular Machine.Angew. Chem. Int. Ed.2019, 58, 9876–9880. Google Scholar
- 20. Afrose S. P.; Bal S.; Chatterjee A.; Das K.; Das D.Designed Negative Feedback from Transiently Formed Catalytic Nanostructures.Angew. Chem. Int. Ed.2019, 58, 15783–15787. Google Scholar
- 21. Heinen L.; Walther A.Programmable Dynamic Steady States in ATP-Driven Nonequilibrium DNA Systems.Sci. Adv.2019, 5, eaaw0590. Google Scholar
- 22. Deng J.; Walther A.Fuel-Driven Transient DNA Strand Displacement Circuitry with Self-Resetting Function.J. Am. Chem. Soc.2020, 142, 21102–21109. Google Scholar
- 23. Che H.; Buddingh B. C.; van Hest J. C. M.Self-Regulated and Temporal Control of a “Breathing” Microgel Mediated by Enzymatic Reaction.Angew. Chem. Int. Ed.2017, 56, 12581–12585. Google Scholar
- 24. van Ravensteijn B. G. P.; Hendriksen W. E.; Eelkema R.; van Esch J. H.; Kegel W. K.Fuel-Mediated Transient Clustering of Colloidal Building Blocks.J. Am. Chem. Soc.2017, 139, 9763–9766. Google Scholar
- 25. Hao X.; Sang W.; Hu J.; Yan Q.Pulsating Polymer Micelles via ATP-Fueled Dissipative Self-Assembly.ACS Macro Lett.2017, 6, 1151–1155. Google Scholar
- 26. Hao X.; Chen L.; Sang W.; Yan Q.Periodically Self-Pulsating Microcapsule as Programmed Microseparator via ATP-Regulated Energy Dissipation.Adv. Sci.2018, 5, 1700591. Google Scholar
- 27. Choi S.; Mukhopadhyay R. D.; Kim Y.; Hwang I.-C.; Hwang W.; Ghosh S. K.; Baek K.; Kim K.Fuel-Driven Transient Crystallization of a Cucurbit[8]uril-Based Host–Guest Complex.Angew. Chem. Int. Ed.2019, 58, 16850–16853. Google Scholar
- 28. Panja S.; Adams D. J.Gel to Gel Transitions by Dynamic Self-Assembly.Chem. Commun.2019, 55, 10154–10157. Google Scholar
- 29. Zhang X.; Wang K.; Hu J.; Zhang Y.; Dai Y.; Xia F.Role of a High Calcium Ion Content in Extending the Properties of Alginate Dual-Crosslinked Hydrogels.J. Mater. Chem. A.2020, 8, 25390–25401. Google Scholar
- 30. Li M.; Lei L.; Zhang Q.; Zhu S.CO2-Breathing Induced Reversible Activation of Mechanophore within Microgels.Macromol. Rapid. Commun.2016, 37, 957–962. Google Scholar
- 31. Li M.; Zhang Q.; Zhu S.Photo-Inactive Divinyl Spiropyran Mechanophore Cross-Linker for Real-Time Stress Sensing.Polymer2016, 99, 521–528. Google Scholar