Controlling the Morphology of Aggregates of an Amphiphilic Synthetic

Publication Date (Web): June 13, 2006 ... The interdependence of the aggregation behavior and the host−guest chemistry was studied. ... These result...
0 downloads 0 Views 116KB Size
5994

Langmuir 2006, 22, 5994-5997

Controlling the Morphology of Aggregates of an Amphiphilic Synthetic Receptor through Host-Guest Interactions Chomchai Suksai,†,‡ Sergio Figueiras Go´mez,§ Anjali Chhabra,| Jingyuan Liu,† Jeremy N. Skepper,⊥ Thawatchai Tuntulani,‡ and Sijbren Otto*,† Department of Chemistry, UniVersity of Cambridge, Lensfield Road, Cambridge CB2 1EW, U.K., Multi-Imaging Center, Department of Anatomy, UniVersity of Cambridge, Downing Street, Cambridge CB2 3DY, U.K., and Department of Chemistry, Faculty of Science, Chulalongkorn UniVersity, Bangkok 10330, Thailand ReceiVed April 7, 2006. In Final Form: May 22, 2006 A new amphiphilic receptor containing a macrocyclic anionic headgroup and a single alkyl chain was prepared through an efficient templated synthesis. The interdependence of the aggregation behavior and the host-guest chemistry was studied. In the absence of any guest the terminus of the alkyl chain of the receptor is included inside the hydrophobic cavity of the macrocycle (as evident from 1H NMR studies) leading to self-assembly into micrometer-long nanotubes (as evident from TEM studies). The alkyl chain can be displaced by an acridizinium bromide guest (as evident from 1H NMR and ITC), which leads to a dramatic change in aggregate size and morphology (as evident from DLS). Studies of the solubilization of Nile red suggest that the resulting aggregates are micelles with a cmc of around 35 µM. These results represent a new addition to the still small number of water-soluble amphiphilic receptors and one of the first examples in which specific host-guest chemistry controls the size and shape of nanoscale aggregates.

Creating nanoscale objects through the self-assembly of small molecules is an attractive strategy that brings with it the challenge of controlling the size and structure of the assemblies through interactions on the molecular scale. Much of the work in the area of self-assembly of complex structures from relatively simple synthetic molecules has been inspired by the ability of phospholipids to self-assemble to form biological membranes. Synthetic amphiphiles (low molecular weight or macromolecular) of various architectures have been studied extensively and have produced assemblies with a great variety of shapes and sizes.1 The morphology of these assemblies is often dictated by the structure of the constituent molecules. In some cases changes in morphology can be induced by external factors such as changes in pH,2 concentration,3 or temperature.4 Traditionally, synthetic amphiphiles have been designed to contain nonpolar parts that interact unfavorably with the solvent * To whom correspondence should be addressed. E-mail: [email protected]. Fax: +44 1223 336017. Tel: +44 1223 336509. † Department of Chemistry, University of Cambridge. ‡ Chulalongkorn University. § Current address: Departamento de Quı´mica Orga ´ nica y Unidad Asociada al CSIC, Universidad de Santiago de Compostela, 15706 Santiago de Compostela, Spain. | Current address: Department of Chemistry, Indian Institute of Technology Bombay, India. ⊥ Multi-Imaging Center, Department of Anatomy, University of Cambridge.

together with polar subunits that interact favorably with the solvent. Recently, a new class of more sophisticated synthetic amphiphiles has emerged in which the headgroups have been designed to engage in more specific molecular recognition.5 These include amphiphiles in which the polar and nonpolar parts are held together through host-guest interactions.5h,j Perhaps more interesting are aggregates formed by amphiphilic synthetic receptors, which can provide functional nanoscale surfaces that can in principle bind to specific guests, thus allowing further noncovalent modification and elaboration of the aggregates. Among the small number of such systems that have been reported to date are amphiphilic receptors based on cyclodextrins,5b,c,e-g,j calixarenes,5d and cucurbiturils.5i The host-guest chemistry has been investigated in some of these reports,5b,c,g,i-k in one instance leading to a change in aggregate size upon binding of PEGfunctionalized guests.5j We now report a new synthetic amphiphilic receptor 1b based on a disulfide macrocycle and show how specific host-guest interactions can induce a dramatic change in the morphology of the assemblies formed by this synthetic receptor, from nanotubes in the absence of a guest to much smaller micellar assemblies in the presence of guest 5.

(1) For a review, see (a) Kunitake, T. Angew. Chem., Int. Ed. Engl. 1992, 31, 709. For selected recent examples, see (b) Abel, E.; De Wall, S. L.; Edwards, W. B.; Lalitha, S.; Covey, D. F.; Gokel, G. W. Chem. Commun. 2000, 433. (c) Hartgerink, J. D.; Beniash, E.; Stupp, S. I. Science 2001, 294, 1684. (d) Zhou, S. Q.; Burger, C.; Chu, B.; Sawamura, M.; Nagahama, N.; Toganoh, M.; Hackler, U. E.; Isobe, H.; Nakamura, E. Science 2001, 291, 1944. (e) Pochan, D. J.; Chen, Z. Y.; Cui, H. G.; Hales, K.; Qi, K.; Wooley, K. L. Science 2004, 306, 94. (f) Boerakker, M. J.; Hannink, J. M.; Bomans, P. H. H.; Frederik, P. M.; Nolte, R. J. M.; Meijer, E. M.; Sommerdijk, N. A. J. M. Angew. Chem., Int. Ed. 2002, 41, 4239. (g) Vriezema, D. M.; Hoogboom, J.; Velonia, K.; Takazawa, K.; Christianen, P. C. M.; Maan, J. C.; Rowan, A. E.; Nolte, R. J. M. Angew. Chem., Int. Ed. 2003, 42, 772. (2) Johnsson, M.; Wagenaar, A.; Engberts, J. B. F. N. J. Am. Chem. Soc. 2003, 125, 757. (3) Hu, Z. J.; Jonas, A. M.; Varshney, S. K.; Gohy, J. F. J. Am. Chem. Soc. 2005, 127, 6526. (4) (a) Yin, H. Q.; Huang, J. B.; Lin, Y. Y.; Zhang, Y. Y.; Qiu, S. C.; Ye, J. P. J. Phys. Chem. B 2005, 109, 4104. (b) Gao, L. C.; Shi, L. Q.; Zhang, W. Q.; An, Y. L.; Liu, Z.; Li, G. Y.; Meng, Q. B. Macromolecules 2005, 38, 4548.

10.1021/la0609470 CCC: $33.50 © 2006 American Chemical Society Published on Web 06/13/2006

Letters

Figure 1. HPLC analysis of the solution obtained 24 h after mixing building blocks 2 (6.67 mM), 3 (3.33 mM), and template 4 (10 mM) in DMSO.

In our program on dynamic combinatorial chemistry,6 we have developed a number of disulfide-based receptors7 including macrocycle 1a7a that bind organic ammonium ions such as 4 and 5 with micromolar affinity in water. We have now produced an amphiphilic version of 1a by a templated synthesis starting from 3 and previously described 2.7a The n-dodecyl building block 3 was prepared in six steps from methyl 3,5-dihydroxybenzoate and dodecylamine. Macrocycle 1b was generated in one step by mixing the constituent dithiol building blocks and template 4 in DMSO solution.8 HPLC analysis of the crude product mixture indicated a gratifying 95% conversion into two diastereomers of 1b: 25% of the minor meso isomer (RR/SS) and 70% of the major isomer (RR/RR and SS/SS as a racemic mixture; see Figure 1).9 The major rac isomer was isolated in 47% yield by preparative HPLC. We have studied the aggregation behavior of 1b in water by dynamic light scattering (DLS) and transmission electron miscroscopy (TEM). Analysis of the DLS data of a 0.1 mg/mL solution of 1b in 10 mM borate buffer pH 9 suggested the presence of aggregates with a diameter around 100 nm. Negative staining transmission electron microscopy showed well-defined tubular structures (Figure 2) having a diameter of 60 nm and lengths ranging from 250 nm to 2.5 µm.10 We were rather surprised to observe such large aggregates for an amphiphile with an extensive headgroup and only a single (5) (a) Menger, F. M.; Williams, D. Y. Tetrahedron Lett. 1986, 27, 2579. (b) Petter, R. C.; Salek, J. S. J. Am. Chem. Soc. 1987, 109, 7897. (c) Petter, R. C.; Salek, J. S.; Sikorski, C. T.; Kumaravel, G.; Lin, F. T. J. Am. Chem. Soc. 1990, 112, 3860. (d) Arimori, S.; Nagasaki, T.; Shinkai, S. J. Chem. Soc., Perkin Trans. 2 1995, 679. (e) Ravoo, B. J.; Darcy, R. Angew. Chem., Int. Ed. 2000, 39, 4324. (f) Donohue, R.; Mazzaglia, A.; Ravoo, B. J.; Darcy, R. Chem. Commun. 2002, 2864. (g) Falvey, P.; Lim, C. W.; Darcy, R.; Revermann, T.; Karst, U.; Giesbers, M.; Marcelis, A. T. M.; Lazar, A.; Coleman, A. W.; Reinhoudt, D. N.; Ravoo, B. J. Chem.sEur. J. 2005, 11, 1171. (h) Jeon, Y. J.; Bharadwaj, P. K.; Choi, S. W.; Lee, J. W.; Kim, K. Angew. Chem., Int. Ed. 2002, 41, 4474. (i) Lee, H. K.; Park, K. M.; Jeon, Y. J.; Kim, D.; Oh, D. H.; Kim, H. S.; Park, C. K.; Kim, K. J. Am. Chem. Soc. 2005, 127, 5006. (j) Liu, Y.; Xu, J.; Craig, S. L. Chem. Commun. 2004, 1864. (k) Lim, C. W.; Ravoo, B. J.; Reinhoudt, D. N. Chem. Commun. 2005, 5627. (6) (a) Otto, S. J. Mater. Chem. 2005, 15, 3357. (b) Otto, S. Curr. Opin. Drug DiscoVery DeV. 2003, 6, 5090. (7) (a) Otto, S.; Furlan, R. L. E.; Sanders, J. K. M. Science 2002, 297, 590. (b) Brisig, B.; Sanders, J. K. M.; Otto, S. Angew. Chem., Int. Ed. 2003, 42, 1270. (c) Otto, S.; Kubik, S. J. Am. Chem. Soc. 2003, 125, 7804. (d) Corbett, P. T.; Tong, L. H.; Sanders, J. K. M.; Otto, S. J. Am. Chem. Soc. 2005, 127, 8902. (e) Corbett, P. T.; Sanders, J. K. M.; Otto, S. J. Am. Chem. Soc. 2005, 127, 9390. (f) Vial, L.; Sanders, J. K. M.; Otto, S. New J. Chem. 2005, 29, 1001. (g) West, K. R.; Bake, K. D.; Otto, S. Org. Lett. 2005, 7, 2615. (8) The low water solubility of 3 precludes the use of aqueous conditions. Using DMSO as the solvent as well as the oxidizing agent gave the best results. For references on the DMSO-mediated oxidation of thiols, see (a) Yiannios, C. N.; Karabinos, J. V. J. Org. Chem. 1963, 28, 3246. (b) Wallance, T. J. J. Am. Chem. Soc. 1964, 86, 2018. (c) Wallance, T. J.; Mahon, J. J. J. Am. Chem. Soc. 1964, 86, 4099. (d) Wallance, T. J.; Mahon, J. J. J. Org. Chem. 1965, 30, 1502. (9) The assignment of the stereoisomers is based on structural correlations through reequilibration experiments (Supporting Information). (10) The discrepancy between the results from DLS and electron microscopy is most likely an artifact of the analysis of the DLS data, which is based on the assumption that the aggregates are spherical. Electron microscopy shows that this assumption is not valid in this case.

Langmuir, Vol. 22, No. 14, 2006 5995

Figure 2. Transmission electron micrograph of negatively stained (2% phosphotungstic acid (PTA)) assemblies of amphiphilic receptor 1b (0.1 mg/mL) in 10 mM borate buffer pH 9. The scale bar represents 1 µm.

Figure 3. Part of the 1H NMR spectrum of amphiphilic receptor 1b in (a) D2O/CD3CN (1:1); (b) D2O (pD 8.7); and (c) D2O (pD 8.7)/MeOD-d4 (2:1) at room temperature.

alkyl chain, which are characteristics that usually give rise to micellar aggregates. NMR investigations shed some light on this unexpected behavior. Figure 3 shows the 1H NMR of 1b in D2O/CD3CN (panel a) and in D2O at pD 8.7 (panel b). A comparison of the two spectra showed that in D2O the protons of the alkyl chain are shifted upfield (up to -1.6 ppm), presumably as a result of the aromatic-ring currents that they experience inside the hydrophobic cavity of 1b. We have found that at higher temperatures (not shown) or upon addition of CD3OD the spectrum sharpened up and small changes in chemical shifts occurred, which allowed us to observe all 11 methylene groups as separate signals (Figure 3c). A TOCSY spectrum recorded in the presence of methanol (Supporting Information) enabled us to assign all signals of the alkyl chain. Correlation with the spectrum in D2O indicated that in this solvent the strongest shifts are experienced by the six methylene groups b-g. Thus, in the aggregates, the terminus of the alkyl chain of 1b is included inside the hydrophobic cavity of the macrocycle.

5996 Langmuir, Vol. 22, No. 14, 2006

Letters

Figure 4. Cartoon representation of different modes of inclusion of the alkyl chain in the cavity of 1b. Table 1. Thermodynamic Dataa for the Binding of Receptors 1a and 1b with Guests 4 and 5 in 10 mM Borate Buffer pH 9 at 298 K 1a

K [M-1] ∆G° [kJ/mol] ∆H° [kJ/mol] T∆S° [kJ/mol]

4

5

2.5 × 105 b -30.8 -41.6 -10.8

6.4 × 105 c -33.1 -40.6 -7.5

Kapp [M-1]12 2.4 × 104 ∆G° [kJ/mol] no binding -25.0 ∆H° [kJ/mol] detectable -8.3 T∆S° [kJ/mol] 16.7 a Equilibrium constants (K), Gibbs energies (∆G°), enthalpies (∆H°), and entropies (T∆S°) of the binding of guests 4 and 5 to receptors 1a and 1b rac were determined using isothermal titration calorimetry. b Reference 7a. c Reference 7b. 1b

One possibility would be that the alkyl chain would loop back into the cavity of the same molecule (Figure 4a). Alternatively, the alkyl chain could insert itself into the cavity of a neighboring receptor, leading to dimerization (Figure 4b) or the formation of “daisychains” (Figure 4c). Discriminating between these arrangements (if indeed one of these arrangements is preferred) is challenging, and experiments in this direction have thus far been inconclusive. However, irrespective of the exact mechanism of inclusion, we speculate that the result will be an arrangement in which the cross-sectional area of the alkyl chain region is effectively doubled, either by looping back (Figure 4a and c) or by interdigitation (Figure 4b) of the chains. Such an arrangement is more likely to result in a bilayer-type organization rather than in micelles. The inclusion of the alkyl chain inside the hydrophobic cavity can be expected to reduce the affinity of the receptor toward guest molecules. Binding studies using isothermal titration calorimetry (ITC) confirmed this (Table 1). Indeed, we failed to detect any binding of 4 in aqueous solutions containing 0.075 mM of 1b, despite the fact that the same guest binds strongly to 1a (K ) 2.5 × 105 M-1).11 However, we were able to measure the binding of the more hydrophobic guest 5 to 1b (Kapp ) 2.4 × 104 M-1),12 but also for this guest, binding to 1b is much less (27 times) efficient than binding to 1a. Our binding data suggests that guest 5 is able to displace the alkyl chain from the cavity of host 1b. This was confirmed by the changes in the 1H NMR spectrum of the alkyl chain upon titrating 5 into a solution of 1b (Figure 5). The upfield-shifted methylene resonances disappeared and were replaced by resonances at 1 to 2 ppm, which corresponds to the normal region for unsubstituted alkanes (cf. Figure 3a). An analysis of the resulting solution by DLS indicated that after the addition of 5 the size of the aggregates was reduced to (11) The fact that 4 is able to template the formation of 1b in DMSO solution is probably because in this solvent the alkyl chain is not occupying the cavity of the receptor. (12) The binding of 5 to 1b is accompanied by a change in aggregate morphology. The apparent binding constant (Kapp) therefore contains contributions from the interactions between host and guest as well as contributions associated with the reorganization of the aggregates.

Figure 5. Part of the 1H NMR spectrum of amphiphilic receptor 1b upon addition of guest 5 at 330 K in D2O pD 8.7.

Figure 6. (a) Absorption spectra of Nile red in a solution of 0.1 mM 1b in the absence and presence of 0.4 mM 5. (b) Emission intensity of different saturated solutions of Nile red at 640 nm (λex ) 570 nm) versus concentration of 1b‚5 in 10 mM borate buffer pH 9.

around 10 nm, which is close to the size expected for micellar aggregates. Further support for a guest-induced transition to micelles was obtained by studying the solubilization of the hydrophobic dye Nile red.13 Hydrophobic species are much more readily solubilized in micelles than in bilayer-type aggregates. Figure 6a shows the absorbance spectrum of Nile red (λmax ) (13) (a) Jiang, J. Q.; Tong, X.; Zhao, Y. J. Am. Chem. Soc. 2005, 127, 8290. (b) Goodwin, A. P.; Mynar, J. L.; Ma, Y. Z.; Fleming, G. R.; Frechet, J. M. J. J. Am. Chem. Soc. 2005, 127, 9952.

Letters

570 nm) in a solution of 1b, 5, and 1b‚5. Obviously, the dye is only significantly solubilized in the solution of 1b‚5. We have studied the solubilization of Nile Red as a function of the concentration of 1b‚5 in order to determine the critical micelle concentration (cmc). The curve in Figure 6b shows the fluorescence intensity of different solutions of 1b‚5, which were stirred for 2 h in the presence of Nile red and then filtered to remove unsolubilized dye. The cmc of this system was estimated to be around 35 µM. In conclusion, we have synthesized a new amphiphilic synthetic receptor that in the absence of any guest aggregates into nanotubes as a result of the self-inclusion of the alkyl part of the molecule inside the nonpolar cavity of the polar headgroup. While this leads to a reduction in the efficiency with which these receptors bind their guests, complexation of a suitable guest molecule is still achievable. This creates the possibility of controlling aggregate morphology through specific host-guest interactions. Indeed, we have demonstrated that the addition of a guest induces

Langmuir, Vol. 22, No. 14, 2006 5997

a dramatic change in the morphology of aggregates formed by the amphiphilic receptor. Studies are underway to elaborate on the resulting organized receptor assemblies for the construction of well-defined nanosized objects through noncovalent surface functionalization including the use of multivalent interactions. Acknowledgment. This work was supported by the Thailand Research Fund (PHD/00123/2545), the EPSRC, and the Royal Society. We are grateful to Marc Stuart for his help with the analysis of the electron microscopy data and to Jeremy Sanders for continuous support. Supporting Information Available: Procedures for the synthesis of 1b and 3, HPLC/ESI-MS analyses of the mixture in Figure 1, TOCSY spectrum of 1b, DLS, TEM, dye entrapment, and reequilibration experiments. This material is available free of charge via the Internet at http://pubs.acs.org. LA0609470