Anion-Induced Self-Assembly of Bis(cyclopeptides) with Rigid Linkers

Abstract

The presence of sulfate anions induces the self-assembly of anion-binding bis(cyclopeptides) in which two cyclopeptide rings are connected via a rigid linker. In this way, 2:2 complexes are formed in which two anions are sandwiched between two bis(cyclopeptide) moieties. Mixed species can be formed if two bis(cyclopeptides) containing different linkers are present and the structural mismatch between the linkers can be compensated for in the self-assembled product. Sulfate complexation seems to proceed with positive cooperativity, leading primarily to the fully formed complexes. As a consequence, these bis(cyclopeptides) represent useful building blocks for the anion-mediated formation of self-assembled products with controllable structural complexity.

1. Introduction

Molecular self-assembly relies on reversible interactions between identical or different building blocks and leads to larger and potentially complex molecular architectures, the structures of which are determined by the number and arrangement of the individual components [1]. The reversibility of the underlying interactions is key for this process to work because it allows error correction should intermediates arise that cannot produce the desired outcome. In addition, reversibility ensures the formation of the thermodynamically most stable product (or mixture of products if some have comparable stability). This product is typically the least strained one, containing the smallest number of building blocks and no vacant binding sites. At high concentrations, the entropic preference for small, discrete products can be overcompensated for, making supramolecular polymers accessible [2]. Based on these rules, the structure of the products can often be reliably predicted by considering the shapes of the building blocks and the arrangement of the binding motifs.

Another crucial aspect for self-assembly to work is the directionality of the interactions. Non-directional interactions such as ion-pairing are generally unsuitable, unless precautions are taken to ensure the proper arrangement of the components [3]. The interactions most frequently used in noncovalently assembled systems are hydrogen bonds [4,5] and halogen bonds [6]. Other useful directional interactions include coordinative interactions in metal–ligand complexes [7,8] and dynamic covalent bonds [9]. Reversible interactions can occur directly between the building blocks or can be mediated by other species, such as the metal ions in coordination compounds. Not only cations can mediate self-assembly but also anions, although they are much less widely used [10]. Important contributions came from Wu and co-workers, who showed that complexes between fully deprotonated orthophosphate ions (PO₄³⁻) and oligoureas can serve as nodes in self-assembled anion-stabilized double or triple helicates and cages [11,12]. These compounds mimic many of the aspects and properties of the complementary metal-containing analogs, including adaptability to complementary guests and stimuli responsiveness. Using related concepts, several other self-assembled architectures driven by anion coordination have been realized [13,14,15,16]. In addition, the synthesis of the mechanically interlocked anion receptors described by Beer and co-workers [17,18,19] and other groups [20] also builds on the self-assembly of suitable precursors.

A simple case of anion-induced self-assembly occurs when the interaction between an anion and a macrocyclic ligand leaves parts of the anion exposed, allowing a second ring to bind. The formation of such sandwich complexes has been observed for several macrocyclic oligoamides [21,22], Flood’s cyanostar [23], which has even been shown to yield stacks containing more than two rings under certain conditions [24], and the anion-binding cyclopeptides developed by our group [25]. The crystal structure of the sandwich complex between the parent cyclic hexapeptide 1, containing an alternating arrangement of L-proline and 6-aminopicolinic acid subunits, and iodide is shown in Figure 1. As can be seen, 1 adopts a folded conformation in which the proline residues and the aromatic residues are arranged on opposite sides of the ring plane. The tertiary amide groups adopt cis conformations, and the aromatic residues are arranged almost parallel to the main axis. The three NH groups converge and protrude into a bowl-shaped cavity. When two of these cyclopeptides self-assemble by interdigitation of the proline rings, an iodide anion can be incorporated between them. Other halides can also be bound, but the most stable complexes are usually formed with sulfate, especially in aqueous methanol or acetonitrile mixtures [25,26]. Anions are necessary for these complexes to form, as there was no evidence of cyclopeptide self-assembly in salt-free solutions.

Subsequent work focused on bis(cyclopeptides) in which two cyclopeptide residues were connected covalently by one or more linkers. These linkers were chosen to be sufficiently flexible to enable both cyclopeptide rings to simultaneously interact with an anion. Only recently, we also prepared a bis(cyclopeptide) with a rigid linker derived from a photoswitchable stiff stilbene [27]. Interestingly, we found that the E-isomer of this compound, in which the two cyclopeptide rings were too far apart to bind to the same anion, formed 2:2 complexes containing two bis(cyclopeptides) and two sulfate anions. This result suggested the potential of these bis(cyclopeptides) to serve as building blocks in structurally more complex self-assembled architectures. In an attempt to test this concept, we synthesized two further bis(cyclopeptides) with rigid linkers, one derived from 2,6-naphthalenedicarboxylic acid (2) and one derived from 4,4′-biphenyldicarboxylic acid (3) (Figure 1), and studied their binding properties. The synthesis, structural characterization, and interaction of these bis(cyclopeptides) with sulfate ions are presented here.

2. Results

2.1. Structural Design

The two linkers in bis(cyclopeptides) 2 and 3 were chosen to be rigid enough to prevent the cyclopeptide rings from binding to the same anion. Additionally, they were expected to maintain a sufficient distance between the two rings to prevent steric effects from affecting the synthesis or self-assembly behavior. Preliminary information about the ability of these bis(cyclopeptides) to form 2:2 complexes, as observed previously for a stiff stilbene-containing bis(cyclopeptide) [27], was obtained by molecular modeling. The crystallographically determined coordinates of the sulfate complex of a bis(cyclopeptide) with a flexible linker were used to create the terminal moieties of 2 and 3 [28]. The two linkers were introduced between these residues, and the resulting structures were optimized using Spartan 24 (Wavefunction, Inc., Irvine CA, USA). First, the structures were optimized in the gas phase using the MMFF force field. Then, DFT calculations were performed using the B3LYP functional and the 6-31G* basis set. The resulting structures, which should be regarded as snapshots of flexible systems, are shown in Figure 2.

As can be seen in Figure 2, the linkers in 2 and 3 do not prevent the bis(cyclopeptides) from interacting with anions. In the 2:2 complexes, two pairs of cyclopeptides form the usual sandwich complexes without obvious strain or deformation. The linkers cross each other and are slightly tilted relative to each other, but intermolecular interactions that could help to stabilize the complexes are not immediately evident.

2.2. Synthesis

Bis(cyclopeptides) 2 and 3 were synthesized by coupling the known monofunctionalized analog of 1 [27], which contains a (2S,4R)-4-aminoproline subunit, to either 2,6-naphthalenedicarboxylic acid or 4,4′-biphenyldicarboxylic acid using N,N,N′,N′-tetramethyluronium tetrafluoroborate (TBTU) as the coupling reagent (Figure 3). The products were purified by RP-8 column chromatography, affording 2 and 3 in yields of 63% and 57%, respectively. Both compounds were isolated in analytically pure form and characterized by NMR, MS, and elemental analysis (see electronic supporting information (ESI)).

2.3. NMR Spectroscopic Binding Studies

In the binding studies, we focused on sulfate as the anionic substrate which typically forms the most stable complexes with our cyclopeptides in H₂O/CH₃OH mixtures [25,26]. Preliminary ¹H NMR experiments with NaI indicated that iodide interacted much less efficiently with 2 and 3 under similar conditions, and other anions, such as the other halides, nitrate, phosphate, and perchlorate, are known to bind even more weakly than iodide. Accordingly, we conducted most binding studies using Na₂SO₄.

Since anion binding of our bis(cyclopeptides) has so far primarily been studied in aqueous solvent mixtures and only in a few cases in DMSO [26,27], we focused in this work on the use of H₂O/CH₃OH mixtures. While many of the bis(cyclopeptides) investigated by us were soluble enough in 50 vol% H₂O/CH₃OH, 2 and 3 turned out to be significantly less soluble, requiring a larger methanol content to mediate dissolution. Bis(cyclopeptide) 2 was soluble in salt-free 25 vol% H₂O/CH₃OH so that spectra recorded in the absence and the presence of a salt could be compared. Bis(cyclopeptide) 3, on the other hand, was only sufficiently soluble in this solvent mixture when an appropriate sulfate salt was present, regardless of whether Na₂SO₄ or (NMe₄)₂SO₄ was used. This salt effect on solubility suggested that sulfate binding to the bis(cyclopeptides) indeed occurred.

Sections of the ¹H NMR spectra of 2 in 25 vol% D₂O/CD₃OD in the absence and the presence of one molar equivalent of Na₂SO₄ are shown in Figure 4a,b. Due to the lack of a symmetry axis in the cyclopeptide rings, protons in equivalent positions produced three signals in the spectrum. However, the two cyclopeptide subunits were equivalent and produced the same sets of signals. For example, the three signals at 5.59, 5.63, and 5.92 ppm corresponded to the α-protons in the six proline subunits (H^α). Three sets of signals belonging to the protons in the 6-aminopicolinic acid subunits were visible in the aromatic region of the spectrum. In addition, the proton signals at 8.20, 7.89, and 7.79 ppm, belonging to the naphthalene linker, were clearly separated and could be assigned unambiguously.

Pronounced signal shifts, affecting almost all signals, occurred upon the addition of Na₂SO₄. The H^α signal shifts were particularly strong. These signals moved downfield by more than 1 ppm, which is a typical indication of anion complexation [25]. The reason for this effect is the close proximity of the bound anion not only to the cyclopeptide NH groups but also to the proline H^α protons in the binding site. Through-bond effects affected the resonances of the 6-aminopicolinic acid protons upon complex formation, and additional major shifts were observed in the signals of the linker protons, which moved upfield by approximately 1 ppm. The shielding of the linker protons during complex formation was a strong indication for the formation of a complex in which two linkers approached each other, causing their protons to experience the shielding effects of the neighbor. Thus, the spectral changes observed in the spectrum of 2 after the addition of the sulfate salt not only confirmed anion binding but also provided information about the structure of the complexes.

In the case of 3, the solubility was insufficient to obtain a ¹H NMR spectrum in the absence of an anion. However, in the presence of Na₂SO₄, solubility improved, allowing the measurement of a ¹H NMR spectrum. This spectrum was very similar to the spectrum of 2 recorded under the same conditions (Figure 4c). In particular, the signals of the proline H^α protons had similar chemical shifts, appearing at 6.66, 6.69, and 6.97 ppm. Moreover, the signals of the linker protons had chemical shifts of 7.28 and 7.07 ppm. In the spectrum of 3 in DMSO-d₆, these signals appeared at 7.87 and 7.77 ppm (see ESI). Thus, a shielding effect upon complex formation, which suggested the formation of a 2:2 complex, could also be observed for this bis(cyclopeptide).

We attempted to obtain further structural information about these complexes from ROESY spectra. In particular, we hoped to obtain evidence of the spatial proximity of the linker moieties in the 2:2 complexes. Only 2 was useful for this purpose because, according to the calculated structure, self-assembly induced a spatial proximity between protons H¹ and H³ in the different naphthalene units, which are too far apart to produce intramolecular cross-peaks (see Figure 4 for the assignment). In contrast, the chemical non-equivalent protons in the linker of 3 should produce cross-peaks in the ROESY spectrum regardless of whether the bis(cyclopeptide) is monomeric or self-assembled. Nevertheless, we recorded the ROESY spectra of both bis(cyclopeptides). Unfortunately, the ROESY spectrum of an equimolar mixture of 2 and Na₂SO₄ only contained cross-peaks arising from the intramolecular proximity of protons H¹ and H⁴, as well as H³ and H⁴, but none clearly indicating intermolecular proximity of two naphthalene units (Figure S2). Interestingly, a weak cross-peak was visible between the H^α signals of the substituted proline residue and an unsubstituted one, which was absent in the spectrum of the free receptor (Figure S1). This cross-peak was even stronger in the ROESY spectrum of 3 in the presence of Na₂SO₄ (Figure S3). It resulted from the spatial proximity of H^α protons in the sandwich complex and therefore provided evidence for the proposed binding mode.

2.4. Mass Spectrometric Binding Studies

Electrospray ionization mass spectrometry (ESI-MS) was used to obtain information about the complexes between sulfate and 2 or 3. To this end, solutions of the bis(cyclopeptides) in 25 vol% H₂O/CH₃OH containing one molar equivalent of Na₂SO₄ were analyzed. In the spectrum of 2, two peaks were visible, of which the most prominent one appeared at an m/z ratio of 804.5, while the smaller peak had an m/z ratio of 1080.4 (Figure 5a).

The peak at 804.5 could be assigned to either a twofold-charged 1:1 complex between 2 and a sulfate ion (C₇₈H₇₂N₂₀O₁₄⋅SO₄^2–: m/z calcd. = 804.3) or a fourfold-charged 2:2 complex with the composition (2)₂⋅(SO₄^2–)₂, which has the same m/z ratio. The peak of the 1:1 complex should exhibit an isotope pattern with the individual signals separated by 0.5 m/z units, whereas the distance should be 0.25 units in the peak of the 2:2 complex. Unfortunately, the resolution of the spectrum was insufficient to make this distinction. However, using 33 vol% H₂O/CH₃OH as solvent greatly improved the quality of the spectrum, clearly demonstrating that sulfate binding involved the formation of a 2:2 complex (Figure S4). The presence of the peak at 1080.4 in the spectrum in Figure 5a also provided evidence for this stoichiometry because this m/z ratio was consistent with that of a Na⁺ adduct of the 2:2 complex ((C₇₈H₇₂N₂₀O₁₄)₂⋅(SO₄^2–)₂⋅Na⁺: m/z calcd. = 1080.3).

The corresponding ESI-MS spectrum of 3 contained a single peak at m/z = 817.5, which could be attributed to either the 1:1 or the 2:2 complex (Figure 5c). Again, the resolution of the peak was too low to make an assignment. However, we also measured an ESI-MS spectrum of an equimolar mixture of 2 and 3 containing an equivalent amount of Na₂SO₄ (Figure 5b). In this case, three signals appeared in the spectrum, the two previously observed ones and one at an m/z ratio of 811.0. This peak, which has the m/z ratio of a 2:2 heterodimer containing 2, 3, and two sulfate ions (2⋅3⋅(SO₄^2–)₂; C₇₈H₇₂N₂₀O₁₄⋅C₈₀H₇₄N₂₀O₁₄⋅(SO₄^2–)₂: m/z calcd. = 811.3), strongly supported the 2:2 stoichiometry because mixed species require the presence of two bis(cyclopeptides).

2.5. UV-Vis and C.d. Spectroscopic Binding Studies

We also measured the UV-vis and circular dichroism (c.d.) spectra of solutions of 2 in the absence and presence of Na₂SO₄ (the low solubility of 3 again prevented its use in these measurements). The spectra were compared to those of the monotopic cyclopeptide 1, which was present at twice the concentration of 2 to account for the different numbers of cyclopeptide residues. In this way, information about the contribution of the naphthalene units to the absorbance and chiroptical properties of 2 should be obtained.

The UV-vis spectra of both 1 and 2 contained three bands with λ_max values of 209, 246, and 286 nm (Figure 6). The UV-vis spectrum of 2,6-napthalenedicarboxylic acid exhibited bands within the same spectral regions (Figure S5). Therefore, the minor absorption between 325 and 350 nm and the greater intensity of the bands of 2 were the only indications of the presence of the naphthalene unit. The band intensities changed upon the addition of Na₂SO₄ for both 1 and 2, but the effects were minor and not very specific.

The c.d. spectra of 1 and 2 had similar structures, with the bands between 250 and 300 nm resembling those typically observed for α-helical peptides [29]. A major difference was the positive band at 212 nm in the spectrum of 2 that was absent in the spectrum of 1. The presence of Na₂SO₄ mainly caused the bands at 261 and 282 nm to grow for both 1 and 2. A band appeared at approximately 235 nm in the spectrum of 2 that was not visible in the absence of the salt. However, the same band was present in the spectrum of 1, so a clear assignment to the linker was not possible. Unfortunately, more pronounced spectral changes that would account for a proximity and chiral arrangement of the two naphthyl chromophores could not be observed.

2.6. Fluorescence Spectroscopic Binding Studies

Fluorescence measurements should reveal whether complex formation affects the emission of the naphthalene linker in 2. These measurements were performed similar to the UV-vis spectroscopic investigations using an excitation wavelength of 280 nm.

In the spectrum of 2, intensive emission bands with maxima at 362 nm and 376 nm, and a shoulder at 396 nm were observed (Figure 7a), which were absent in the emission spectrum of 1 when the same excitation wavelength was used (Figure 7b). These bands were attributed to the naphthalene linker. Indeed, emission bands occurred in the same wavelength region in the fluorescence spectrum of 2,6-naphthalenedicarboxylic acid, although the bands of 2 had a different intensity (Figure S6).

Adding one molar equivalent of Na₂SO₄ to the solution of 2 caused a notable reduction in the emission intensity of the bands between 360 and 400 nm, an effect that was not observed for 1 but also for 2,6-naphthalenedicarboxylic acid (Figure S6). This might be related, at least in part, to the change in the ionic strength of the solution. More importantly, a broad band appeared in the fluorescence spectrum of 2 at 404 nm, extending to wavelengths >450 nm, which is typically assigned to an excimer produced by stacked naphthalene units [30]. The appearance of this band thus suggests that the two linkers of 2 come close upon sulfate binding, and fluorescence spectroscopy therefore also provided evidence for the self-assembly of the bis(cyclopeptide).

2.7. Calorimetric Binding Studies

Quantitative information about complex stability was obtained using isothermal titration calorimetry (ITC). The measurements were performed by titrating a solution of Na₂SO₄ in 25 vol% H₂O/CH₃OH to a solution of a bis(cyclopeptide) in the same solvent mixture. The thermograms showed that complex formation was exothermic, as also observed for sulfate complexation of other bis(cyclopeptides) in H₂O/CH₃OH mixtures [26]. The corresponding binding isotherms had sigmoidal shapes, indicating strong binding, and an inflection point at a 1:1 molar ratio of bis(cyclopeptide) and anion, i.e., the expected ratio for a 2:2 complex (Figure 8).

We used SupraFit (2.5.120) [31] for the nonlinear regressions because it enabled us to fit the obtained isotherms to a 2:2 binding model. Since nine parameters had to be optimized for a complete evaluation, namely, four binding constants, four enthalpies, and the n value, the results were very sensitive to the starting conditions. Therefore, we focused on determining just the log K_a and enthalpy associated with the formation of the fully formed complex, neglecting other microscopic binding constants. In addition, we experimentally determined the heats of dilution by titrating the salt solution to the pure solvent and subtracted them from the heat pulses in the actual measurements. In this way, heat changes at the end of the titration were corrected and excellent fits of the binding isotherms were obtained, with repeated measurements yielding comparable results.

The log K_a values associated with sulfate complexation of the bis(cyclopeptides) 2 and 3 were the same (14.1 ± 0.1), showing that the linker structure had no measurable effect on the overall stability of the complexes. Sulfate binding of 3 was slightly more exothermic (ΔH⁰ = −21.1 ± 0.2 kJ mol⁻¹) than that of 2 (ΔH⁰ = −18.3 ± 0.3 kJ mol⁻¹). To relate this result to the anion affinity of other cyclopeptides, we also performed titrations with the monocyclic peptides 1 and 4 (Figure 1), of which 1 forms a 2:1 sandwich complex with sulfate [25] and 4 forms a 1:1 complex [32].

In the case of 4, small heats of complex formation and pronounced heats of dilution made it difficult to determine K_a under the conditions used for the other receptors. At the higher concentrations necessary to obtain more reliable results, Na₂SO₄ was insufficiently soluble in 25 vol% H₂O/CH₃OH. Therefore, we changed the salt to (NMe₄)₂SO₄. Previous work had demonstrated that the counterion has a negligible effect on the anion affinity of our cyclopeptides [26], and reference measurements with 1 and 2 yielded results analogous to those obtained under the original conditions. Thus, the results obtained for the different salts were comparable. The log K_a of 3.1 ± 0.2 (ΔH⁰ = −6.8 ± 2.1 kJ mol⁻¹) obtained for the 1:1 sulfate complex of 4 in 25 vol% H₂O/CH₃OH was one order of magnitude greater than the value obtained previously under more competitive conditions in 80 vol% D₂O/CD₃OD [32].

Due to sufficiently large heat pulses, the titration of 1 could be performed under the usual conditions. However, the binding isotherms also lacked a sigmoidal shape, which is why we again restricted the fitting to the overall binding constant log K_a and the corresponding heat of complex formation. These values amounted to 7.6 ± 0.1 and −20.7 ± 2.5 kJ mol⁻¹, respectively, demonstrating that the formation of the 2:1 complex of 1 had more than twice the log K_a associated with the formation of the 1:1 complex of 4. Thus, the second binding step proceeded with positive cooperativity [32], but the extent of cooperativity was lower than previously observed in media with a higher water content [28]. The reason for this is presumably the smaller contribution of the hydrophobic effect to complex formation in the solvent mixture with a higher methanol fraction. More importantly, the log K_a values of the bis(cyclopeptide)–sulfate complexes investigated here were less than twice the log K_a obtained for 1, suggesting that the linkers have a small destabilizing effect on complex stability.

3. Discussion

NMR, UV-vis, c.d., and fluorescence spectroscopy, as well as MS, and ITC provided evidence that bis(cyclopeptides) 2 and 3 interacted with sulfate ions in an aqueous solvent mixture. The most conclusive information on the composition of the complexes came from MS. These measurements demonstrated that complex formation was very specific, resulting in only a single detectable species in the mass spectra. In addition, the presence of the signal at an m/z ratio of 1080.4 in the spectrum of 2 indicated that the complex contained two bis(cyclopeptides) and two anions. Even more important evidence for this stoichio-metry came from the presence of a peak in the mass spectrum belonging to a mixed complex when the solution contained both 2 and 3. Although the formation of this heterocomplex is statistically favored over that of the homocomplexes by a factor of two, its peak was smaller than the peaks of the homocomplexes. Peak intensities in mass spectra do not necessarily correlate with the concentration of the corresponding species, but because of the narrow m/z range in which the three peaks appeared and the identical charges of the complexes, the observed intensity distribution could indicate that the heterocomplex is less stable than the homocomplexes. Unfortunately, the stability of the heterocomplex could not be determined independently to support this assumption, because three complexes formed simultaneously when 2 and 3 were present in solution. However, considering the size mismatch of the two ligands, which should create strain in the heterocomplex, a lower stability for the heterocomplex is probably reasonable. This mismatch is reflected in the distance between the two carboxylate carbon atoms of the linkers. According to the results of the calculations shown in Figure 2, this distance amounts to 8.0 Å in 2 and 10.1 Å in 3. Despite this difference of slightly more than 2 Å, the structures seem to be flexible enough to allow the formation of a 2:2 complex, as demonstrated by the DFT-optimized structure of the heterocomplex shown in Figure 9.

Despite numerous attempts, we did not succeed in obtaining crystals suitable for X-ray crystallography to confirm our structural assignments. However, several spectroscopic methods provided strong support. For instance, complex formation induced pronounced upfield shifts in the linker signals in the ¹H NMR spectrum, which demonstrated that the linkers approach each other in the complexes, causing their protons to mutually experience the shielding effects of their aromatic systems. In addition, changes in the fluorescence spectrum of 2 upon the addition of Na₂SO₄ also suggested the formation of a 2:2 complex.

ITC qualitatively demonstrated that sulfate binding is exothermic in the aqueous solvent mixture. In addition, the steep steps of the binding isotherms indicated positive cooperativity, primarily resulting in fully formed complexes, and the inflection points were consistent with 2:2 receptor/substrate ratios. To avoid fitting these isotherms to too many parameters, we focused on determining just the overall stability constants. This approach also allowed us to compare the sulfate affinity of the bis(cyclopeptide) complexes with that of related receptors, such as the monocyclic 4, which forms a 1:1 complex with sulfate with a log K_a of 3 in the chosen solvent mixture. The stability of the corresponding 2:1 sandwich complex of 1 turned out to be more than twice as high as a result of the cooperativity of the binding process, as also observed in other studies [28,32]. The stability of the bis(cyclopeptide) complexes should be twice this log K_a value if the effect of the linkers on complex stability were negligible. However, although the log K_a values determined for the bis(cyclopeptides) are indeed several orders of magnitude greater than the log K_a value of the sulfate complex of 1, reflecting the fact that they are a product of four binding constants (K_a has units of M⁻⁴), their actual values are slightly smaller than the expected value of 15.2, regardless of the linker. Therefore, the linkers in 2 and 3 apparently slightly weaken the cyclopeptide–anion interactions.

4. Conclusions

Unlike bis(cyclopeptides) with flexible linkers that have been shown to form 1:1 complexes with anions in which the anion is sandwiched between two cyclopeptide rings, bis(cyclopeptides) with rigid linkers self-assemble in the presence of suitable anions. In this way, complexes are formed in which two anions are bound between two bis(cyclopeptide) moieties. Complex formation takes place in aqueous solvent mixtures and proceeds with positive cooperativity, causing the fully formed products to be the dominant species in solution. Thus, such cyclopeptide–anion complexes could serve as binding nodes in anion-induced self-assembled products. While the structural complexity of the complexes studied in this work is low, it can be increased by incorporating ligands between two or more cyclopeptide rings that induce the formation of products with higher structural complexity. Work in this direction is currently underway.