Co-assembled Coiled-Coil Peptide Nanotubes with Enhanced Stability and Metal-Dependent Cargo Loading.

Peptide nanotube biomaterials are attractive for their range of applications. Herein, we disclose the co-assembly of coiled-coil peptides, one with ligands for metal ions that demonstrate hierarchical assembly into nanotubes, with spatial control of the metal-binding ligands. Enhanced stability of the nanotubes to phosphate-buffered saline was successfully accomplished in a metal-dependent fashion, depending on the levels and placement of the ligand-containing coiled-coil peptide. This spatial control also allowed for site-specific labeling of the nanotubes with His-tagged fluorophores through the length of the tubes or at the termini, in a metal-dependent manner.

Introduction

Nanotube biomaterials have drawn significant attention for their applications ranging from piezoelectric devices to cargo delivery.[1] While less common than simpler nanofibrils, these three-dimensional materials contain an inner cavity and outer shell to allow for multifunctionality. Peptide-based nanotubes specifically are an attractive choice owing to their tunability and biocompatibility.[2−4] As such, peptide nanotubes have been derived from a diverse collection of structures including dipeptides,[5−9] cyclic peptides,[10−25] peptoids,[26−29] triple helices,[30,31] β-sheets,[32−35] and coiled-coils.[36−42]

The assembly and morphology of peptide nanotubes provide the opportunity for a range of interesting biological applications. For example, early studies with cyclic peptide nanotubes focused on interactions with the cell membrane to mimic membrane-bound proteins and form ion channels.[10,11,19−21,23] Since then, nanotubes have been developed with antibacterial properties, including dipeptide nanotubes clearing biofilms[7] and cyclic peptide nanotubes lysing bacterial membranes.[12,25] β-sheet nanotubes have been used for directed delivery of an anticancer agent to metastatic melanoma,[34] whereas coiled-coil nanotubes, composed of the peptide TriNL (Figure A), for instance, have been loaded with a biopolymer cargo.[42]

Figure 1

Peptide nanotube co-assemblies. (A) Peptide sequences of the coiled-coil (CC) peptides TriNL and p2L. (B) Schematic of co-assembly strategies for incorporating p2L into TriNL nanotubes for potential stabilization with metal ions: forming mixed peptide CC through thermal annealing and using stepwise addition to bring in CC trimers of p2L.

Coiled-coil nanotubes of TriNL have the potential for molecular storage and drug delivery, but instability in phosphate-buffered saline (PBS) limited biological applications. Herein, we demonstrate the co-assembly of TriNL with a variant containing metal-binding ligands that produces nanotubes with enhanced PBS stability in a metal-dependent manner while also promoting His-tagged cargo binding.

Experimental Section

Circular Dichroism

A TriNL-Fl solution (50 μM) was prepared in citrate buffer (10 mM, pH 3.0). The circular dichroism (CD) spectrum was taken at 4 and at 90 °C by averaging three scans between 190 and 260 nm.

Fluorimetry

A 15 μL aliquot of a 550 μM TriNL-Fl solution with varying p2L concentrations was annealed (90 °C for 30 min, 4 °C for >18 h). The samples were diluted with 985 μL of PBS and fluorescence measurements were conducted with an excitation wavelength of 490 ± 2.5 nm and an emission bandwidth of 0.5 nm.

Annealed Assemblies

Peptide solutions were prepared in 15 μL aliquots and annealed (90 °C for 30 min, 4 °C for >18 h). These aliquots were added to 35 μL of 2-(N-morpholino)ethanesulfonic acid buffer (MES, pH 6.0) to create a final buffer concentration of 50 mM and a total peptide concentration of 500 μM. Following a 30 min incubation at room temperature, the assemblies were centrifuged at 10,000g for 3 min where the supernatant was removed and replaced with water. This process was repeated two more times.

Stepwise Assemblies

Assemblies were conducted with a total peptide concentration of 500 μM in a final MES buffer concentration of 50 mM (pH 6.0). Specifically, TriNL and p2L were separately added to the MES buffer in the same buffer ratio as the peptide ratio. After a 2 min incubation at room temperature, p2L was added to TriNL and allowed to assemble for an additional 30 min. The samples were centrifuged at 10,000g for 3 min. The supernatant was removed, replaced with water, and the process was repeated twice.

Scanning Electron Microscopy

Aliquots of the assemblies described above (3 μL) were placed on a glass coverslip attached to a metal stub via a copper tape. The samples were either air-dried or lyophilized and platinum-coated. Samples were imaged using an FEI Teneo Volumescope field emission scanning electron microscope.

Degradation Studies

Peptides were assembled using the protocol above. Following washing, the samples were incubated in 1 mM NiCl2 for 1 h at room temperature. The material was centrifuged at 10,000g for 3 min, the supernatant was removed, and the samples were resuspended in 50 μL of 1× PBS for 24 h at room temperature.

Cargo Binding

Peptides were assembled using the protocol above in the presence of 1 mg/mL fluorescein-labeled anionic dextran (MW 40,000). Following washing, the samples were incubated in 1 mM NiCl2 for 1 h at room temperature. The material was centrifuged at 10,000g for 3 min, and the supernatant was removed followed by resuspension in 50 mM MES buffer (pH 6.0) and 20 μM Rh-His6. After a 48 h incubation, the samples were washed as above and imaged on a Nikon A1R multiphoton inverted confocal microscope under a 100× oil objective using 488 and 561 nm excitation lasers.

Results and Discussion

Co-assembled Peptide Nanotube Design

The design of the stabilized nanotubes is based on the leucine-zipper motif of the transcription factor GCN4. A dimeric coiled-coil in its native form and the GCN4 sequence have been modified to form trimeric coiled-coils.[43] This trimeric sequence was the basis of the TriNL peptide (Figure A) that assembles into nanotubes while encapsulating the biopolymer cargo.[42] However, upon exposure to PBS, these coiled-coil peptide tubes were found to rapidly deteriorate. To promote stability, therefore, we hypothesized that the introduction of a coiled-coil peptide with metal-binding ligands as a part of the nanotubes might strengthen the interactions between coiled-coils in a metal-dependent fashion and potentially decrease the rate of tube degradation (Figure B). To this end, we looked at the coiled-coil peptide p2L with di-histidine and nitrilotriacetic acid (NTA) ligands at the C- and N-termini, respectively (Figures A and S1).[44] The p2L peptide alone has been shown to form hexagonal crystals with zinc ions,[44] but here we wished to investigate what levels of p2L could be added to TriNL while retaining the tube morphology and whether stabilized nanotubes would result (Figure B).

In the design of nanotubes comprised of two peptides, heterotrimeric coiled-coil mixtures composed of both TriNL and p2L or individual homotrimers of TriNL and p2L could be used as building blocks (Figure B). Herein, both coiled-coil assemblies were explored, either through prior thermal annealing or sequential addition of the two peptides, to allow for spatial control of p2L within the nanotubes. These co-assemblies, in turn, may allow us to tune nanotube stability and bring in His-tagged cargo in a metal-dependent fashion.

Peptide Synthesis

Peptides were synthesized using standard Fmoc-based solid-phase peptide synthesis on the ChemMatrix Rink Amide resin with hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU) as a coupling reagent. The protected NTA amino acid was synthesized as previously described.[45] The peptides were cleaved from the resin using a trifluoroacetic acid (TFA) cocktail, purified to homogeneity via reverse-phase high-performance liquid chromatography (HPLC), and characterized via matrix-assisted laser desorption ionization time-of-flight (MALDI-ToF) mass spectrometry (Figures S2 and S3).

Intermixed Coiled-Coil Formation: Thermal Annealing

Previous studies have shown that thermal annealing of two different dimeric coiled-coils can give rise to mixed coiled-coils at statistical levels with their homodimeric counterparts.[46] Motivated by this work, and that of others studying mixed coiled-coils,[47−66] we first pursued a strategy to co-assemble p2L and TriNL within nanotubes by intermixing the peptides at the supersecondary structure level using thermal annealing. In this way, we could create a mixture of coiled-coils with some containing both TriNL and p2L.

Circular dichroism (CD) was initially used to study coiled-coil folding by monitoring the two negative absorption bands at 222 and 208 nm. CD spectra were taken at 4 and 90 °C using a 2:1 ratio of TriNL and p2L (500 μM total peptide concentration, Figure S4). Elevated temperature (90 °C) led to a decrease in helical content from 90 to 58%, and subsequent cooling to 4 °C promoted refolding to 97% α-helicity. It is worth noting that in the low-temperature spectra the magnitude of the absorption at 222 nm is higher than that at 208 nm (θ222/θ208 ∼ 1.5), which has been shown to be indicative of the presence of a coiled-coil.[67] At 90 °C, however, this ratio decreased to ∼1.1, demonstrating lower levels of the coiled-coil fold.

With CD supporting the refolding of the coiled-coils through thermal annealing, the formation of heterotrimeric coiled-coils was verified using the fluorometric assay of Xiao and co-workers.[68] This assay relies on fluorescence self-quenching of fluorophores at the termini of peptides’ trimeric assemblies. With this in mind, fluorescein was installed at the N-terminus of TriNL (TriNL-Fl, Figures S5 and S6). If unlabeled p2L is introduced into the labeled coiled-coil through thermal annealing, an increase in fluorescence due to less fluorescein self-quenching should result. The coiled-coil refolding process was monitored by measuring the fluorescence of TriNL-Fl (8.5 μM) with varying amounts of p2L before and after thermal annealing (Figures and S7). All solutions exhibited similar initial fluorescence readings prior to annealing. Following annealing, however, an increase in fluorescence was observed relative to the amount of p2L in solution, with up to a 1.9-fold increase at a ratio of 2:1 TriNL-Fl:p2L (Figure ). Additionally, the fluorescence of TriNL-Fl did not increase when annealed in the absence of p2L, further demonstrating that any observed increase in fluorescence was due to intermixed coiled-coil formation rather than the annealing process.

Figure 2

Fluorescence spectra of TriNL-Fl (8.5 μM) before and after annealing in the presence of increasing levels of p2L.

Nanotube Formation: Thermally Annealed Samples

With intermixed coiled-coils available through thermal annealing, nanotube formation was investigated. All assembly experiments were conducted with a total peptide concentration of 500 μM in MES buffer (50 mM, pH 6.0) at room temperature for 30 min, with a range of TriNL:p2L ratios (1:1–50:1). Upon assembly, the precipitate was collected via centrifugation and washed. Peptide levels within the assemblies were determined by first dissolving the material with aqueous HCl, followed by quantitation using reverse-phase UPLC (Figure S8 and Table S1). The experimentally determined ratios of TriNL and p2L in the assemblies were quite similar to starting ratios of the two peptides in solution, indicating that both of the peptides entered the materials at levels close to the added amounts.

Scanning electron microscopy (SEM) was used to determine if the TriNL nanotube morphology (Figure A) was maintained with the added p2L. Indeed, upon increasing the amount of p2L up to a ratio of 10:1 TriNL:p2L (Figure B–D), we observed definitive nanotubes, with lengths in the 10–15 μm range and inner diameters in the range of 400–800 nm (n = 10). From these experiments and the UPLC data, we can confirm that up 9% incorporation of p2L (10:1) still allows the nanotubes to form. Interestingly, assemblies that form with 17–33% of p2L (5:1 and 2:1) have a hexagonal rod morphology (Figure E,F), with an inner channel in the case of the 5:1 ratio. A 1:1 ratio of peptides provided similar rods as the 2:1 mixtures (Figure S9), although lower levels of assembly were observed. Although no metals are used in these experiments, this observed morphology is reminiscent of the assemblies formed with p2L and metal ions, albeit the current assemblies have an increased aspect ratio (∼3.5-times increase in the length/diameter).[44] It is important to note that p2L does not assemble on its own, without metal ions, under these conditions.

Figure 3

SEM visualization of assemblies of (A) TriNL and at different ratios of TriNL:p2L. (B–F) Annealed samples and (G, H) stepwise addition of the peptide. Insets show the interior morphology of the assemblies (inset scale bar: 500 nm).

Nanotube Stability Studies: Thermally Annealed Samples

Nanotubes formed only with TriNL exhibit limited stability in the presence of PBS buffer with degradation of the tubes at the termini and dissolution into shard-like tube fragments (Figure A).[42] The introduction of the ligand-containing p2L peptide, through thermal annealing, allows for the potential of using metal–ligand interactions to increase the stability of the tubes. With this in mind, the degradation of the nanotube co-assemblies in PBS was investigated at different TriNL:p2L ratios with added Ni(II), a metal ion that has been used extensively for His-tags with NTA. Preformed nanotubes were first incubated in NiCl2 (1 mM, 1 h) and subsequently resuspended in PBS (10 mM phosphate, 137 mM NaCl, pH 7.4) for 24 h. The resulting materials were imaged using SEM. With low levels of p2L (2–5%), there was distinct erosion at the ends into the center of the nanotubes upon exposure to PBS (Figure B,C). Increasing the p2L content to 9%, however, resulted in intact nanotubes (Figure D). In the absence of Ni(II), the co-assembled nanotubes were found to degrade more substantially than with TriNL alone, with only small pieces of nanotubes remaining (Figure S10). The addition of p2L may be destabilizing the tubes due to the charge ligands at each terminus. Overall, these data support the idea that the observed nanotube stabilization is derived from metal–ligand interactions from p2L within the nanotubes.

Figure 4

Stability of nanotubes formed with TriNL or TriNL:p2L ratios after a 24 h incubation with PBS. (A) TriNL alone; (B–D) pre-annealed TriNL:p2L at a (B) 50:1 ratio, (C) 20:1 ratio, and (D) 10:1 ratio; and (E, F) stepwise addition of TriNL:p2L at a (E) 10:1 ratio and (F) 2:1 ratio.

Nanotube Cargo Loading: Thermally Annealed Samples

The presence of ligands for metal ions within the mixed coiled-coil nanotubes presents the opportunity of using these interactions to facilitate the binding of the cargo functionalized with metal-binding ligands. Previously, the TriNL nanotubes were shown to encapsulate fluorescently labeled anionic dextrans within their interior.[42] Similarly, the co-assembled nanotubes resulting from thermally annealed peptide mixtures also encapsulated anionic, fluorescein-labeled dextrans (Figure A). However, after treating both of these tubes with NiCl2 (1 mM, 1 h), followed by treatment with His-tagged rhodamine (Rh-His6, 20 μM, 48 h), only the nanotubes co-assembled with p2L were found to have rhodamine fluorescence throughout the tubes (Figure B vs S11). Also, without the addition of Ni(II), no Rh-His6 association with the mixed coiled-coil tubes was observed (Figure S12). These data demonstrate the importance of metal-loaded ligands within the nanotubes for His-tagged fluorophore binding and suggest that p2L is distributed throughout the assembly.

Figure 5

Confocal microscopy of co-assembled TriNL:p2L nanotubes. (A,B) Annealed nanotubes with encapsulated fluorescein-labeled dextran (1 mg/mL, 40 kD) (A) alone and (B) with pretreated NiCl2 (1 mM) followed by Rh-His6 (20 μM). (C,D) Nanotubes from stepwise addition with encapsulated fluorescein-labeled dextran (1 mg/mL, 40 kD) (C) alone and (D) with pretreated NiCl2 (1 mM) followed by Rh-His6 (20 μM). (B) and (D) Overlay of both the red and green channels.

Nanotube Formation and Stability: Stepwise Addition

Motivated by the results of nanotube cargo loading and stability after including p2L into the TriNL tubes using thermal annealing, we sought an additional way to control the location of the ligand-containing p2L peptide within the nanotubes. To this end, we investigated the addition of individual coiled-coils of TriNL and p2L in a stepwise fashion at 10:1 and 2:1 ratios. Attempts to simultaneously add the two peptides led to results that were difficult to replicate. Therefore, we chose to add the p2L coiled-coils to TriNL trimers after a 2 min incubation of the latter peptide in MES buffer (50 mM, pH 6.0). The combined peptides (500 μM total) were then incubated at room temperature for 30 min, and the precipitates were collected and washed. SEM analysis of the assemblies demonstrated morphologies that were similar to those obtained with the thermally annealed, co-assembled nanotubes. The 10:1 TriNL:p2L ratio provided nanotubes (Figure G), whereas the 2:1 ratio led to hexagonal rods with an aspect ratio of ∼9 (length/diameter) (Figure H). UPLC analysis was used to determine the levels of p2L in the tubes/rods after acid treatment to dissolve the assemblies. As opposed to the thermally annealed samples, the assemblies formed from stepwise addition contained substantially lower levels of the ligand-containing p2L (10:0.3 and 2:0.5 TriNL:p2L ratios, respectively) (Table S2).

Although the nanotubes obtained from the two different assembly techniques at a 10:1 ratio of peptides looked similar at the micron scale, the two sets of tubes behaved quite differently when treated first with NiCl2 and then with PBS. The nanotube morphology was retained in the thermally annealed co-assemblies in PBS after metal ion treatment (Figure D), whereas the sequential assemblies showed notable disintegration of the tube interior with PBS, with the ends of the tubes remained somewhat intact (Figure E). This latter PBS degradation pattern was also observed after Ni(II) treatment of the nanorods derived from the sequential assembly of the 2:1 ratio of peptides (Figure F). Regions of the interior of the hexagonal rods were eaten away leaving hollow shells with defined ends. These data may indicate that the ends of these nanotubes/rods are rich in p2L coiled-coils and addition of metal ions could stabilize these regions. To determine if these terminal ends contain more p2L than the entire structure, the 10:1 and 2:1 stepwise assemblies and their corresponding PBS-degraded materials (Figure E,F) were dissolved with aqueous HCl, and their peptide contents were quantified via UPLC (Table S2). We observed an increase in the p2L levels after PBS treatment of 2.7- and 6.8-fold, respectively, compared to the intact nanotubes/rods. These data indicate that the percentage of p2L is higher at the ends than in the center of the material. It is likely that substantial nanotube/rod formation occurs with TriNL in the 2 min prior to p2L addition, thereby leading to an abundance of p2L near the ends of the assemblies.

Nanotube Cargo Loading: Sequential Coiled-Coil Addition

As an additional means to verify the location of the metal-binding p2L, we also investigated where His-tagged rhodamine binds to the co-assembled nanotubes formed from sequential addition. Preformed nanotubes (10:1) with encapsulated dextran (Figure C) were subjected to Ni(II) incubation followed by Rh-His6. Even after extended periods of time (48 h), rhodamine fluorescence was only observed at the termini of the nanotubes (Figure D). In contrast to the annealed assemblies where the p2L peptide is likely distributed throughout the tubes, all of the above observations provide support for the p2L coiled-coils being substantially localized near the end of the nanotubes when added sequentially (Figure ).

Figure 6

Schematic of PBS stability results for the two strategies of p2L incorporation into TriNL nanotubes. (A) Annealed samples led to p2L distribution throughout the tubes via intermixed coiled-coils resulting in metal ion stabilization throughout the length of the nanotubes. (B) Stepwise addition of individual coiled-coils led to the concentration of p2L near the ends of the nanotubes, followed by treatment with metal ions, resulting in PBS degradation of the center of the tubes.

Conclusions

Peptide-based nanotubes hold great promise for molecular storage and drug delivery. Coiled-coil nanotubes derived from the peptide TriNL have been shown to encapsulate biopolymers, but instability to PBS has limited their bioapplications.[42] Herein, we have demonstrated a means of incorporating the coiled-coil forming peptide p2L, which contains ligands for metal ions at both termini, into nanotube assemblies with TriNL. It was possible to tune the position of p2L in the assemblies, either distributed throughout the tubes or localized near the ends, depending on whether intermixed or individual coiled-coils of p2L and TriNL were used. Indeed, this spatial control of the position of p2L has led to both stabilized nanotubes or tubes with stabilized ends, upon addition of metal ions (Figure ). Metal-charged ligands derived from p2L were also harnessed to bring His-tagged fluorophores to the full length or the termini of the nanotubes, providing an additional readout for the placement of the p2L peptide in the assemblies. The encapsulated cargo within the nanotubes should have differing release profiles depending on the stability of the tubes. It is likely that metal ion-based stabilization will slow down the kinetics of the release of the cargo from the nanotubes, since the tubes are more intact. Future experiments will focus on the range of cargoes that may be placed on and within the structure of these coiled-coil nanotubes.