Fabrication of Multi-Responsive Bioactive Nanocapsules through Orthogonal Self-Assembly

Abstract

Multifunctional self-assembled systems present platforms for fundamental research and practical applications, providing tunability of structure, functionality and stimuli response. Pragmatic structures for biological applications have multiple design requirements, including control of size, stability and environmental response. Here we present the fabrication of multifunctional nanoparticle-stabilized capsules (NPSCs) using a set of orthogonal supramolecular interactions. In these capsules, fluorescent proteins are attached to quantum dots through polyhistidine coordination. These anionic assemblies interact laterally with cationic gold nanoparticles that are anchored to the fatty acid core through guanidinium-carboxylate interactions. The lipophilic core then provides a reservoir for hydrophobic endosome-disrupting agents, generating a system featuring stimuli-responsive payload release into the cytosol with fluorescent monitoring.

Multifunctional nanomaterials are important platforms for “smart” applications, combining the properties of their components^{[ 1 ]} to provide synergistic systems^{[ 2 ]} capable of achieving multiple objectives. Through the appropriate choice of attributes, multifunctional materials can overcome challenges that cannot be solved by their individual components,^[3] including the ability to dynamically adjust their properties and functions in response to both endogenous^{[ 4 ]} and external^{[ 5 ]} environmental stimuli. Several construction challenges, however, must be addressed for these systems to be useful, such as control over the size, stability and dynamic properties.^[6]

Many systems used for bionanotechnology employ covalent conjugation approaches to generate stable nanostructures.^{[ 7 ]} However, these building elements are fixed, making it difficult to further modulate the structure and provide stimuli responsiveness. Supramolecular chemistry provides an alternative fabrication strategy that addresses the construction challenges of multifunctional materials through multiple noncovalent interactions to generate controllable and stimuli-responsive structures.^[8] These supramolecular structures are generated through programmed self-assembly of building blocks (e.g., synthetic materials and biomolecules) by using a broad palette of available noncovalent interactions, including hydrogen bonding, hydrophobic interactions and electrostatic affinities to induce biomimetic assembly.^[9] Recent examples of this bioinspired assembly strategy include hybrid assemblies,^{[ 10 ]} peptide amphiphile vesicles^{[ 11 ]} and protocell models.^[12]

Integration of nanoparticles (NPs) into supramolecular structures provides access to the physical^[13] and structural^[14] properties of the particles. In particular, self-assembly of NPs at the liquid-liquid interface provides a straightforward pathway to generate core-shell nanoparticle-stabilized capsules (NPSCs).^{[ 15 ]} Seen from a biomedical perspective, the NPSC platform incorporates the unique physicochemical properties of the NPs into the shell while providing an interior hydrophobic space suitable for the transport of drugs and other functional agents.^[16]

Here we present the integration of multiple functional building blocks into a NPSC-based platform by means of diverse supramolecular interactions (Figure 1). In these constructs, CdSe/ZnS core-shell quantum dots (QDs) are conjugated with polyhistidine-tagged proteins by exploiting the metal affinity of the histidine residues on the QD surface. These anionic QD-protein complexes further interact electrostatically with cationic gold nanoparticles (AuNPs) that themselves are pinned to the capsule surface through a combination of hydrogen bonding and electrostatic interactions with the oil core. The resultant supramolecular nanocapsules demonstrate multifunctional properties in cellular studies, such as fluorescent tracking of intracellular bioconjugates, stimuli-responsive protein release and co-delivery of proteins and hydrophobic endosome-disrupting agents to provide cytosolic protein delivery.

Figure 1.

a) QD_DHLA and hexahistidine-tagged (His₆-tagged) mCherry were conjugated through metal-affinity coordination; this conjugation can be reversed by adding competitive molecules (e.g., histidine). b) Template droplets were generated by agitating AuNP_HKRK and oil (linoleic acid (LA) and decanoic acid (DA)) in 5 mM phosphate buffer (pH 7.4), and then the template droplets were transferred into AuNP_HKRK solution followed by lateral noncovalent crosslinking using anionic QD-mCherry conjugates to form NPSC_QD-mCherry.

Dihydrolipoic acid functionalized CdSe/ZnS core-shell QDs (QDs_DHLA) generate stable assemblies with polyhistidine-tagged proteins through metal-affinity coordination.^[17] For our studies, hexahistidine-tagged red fluorescent protein mCherry (λ_em= 610 nm) was conjugated onto green-emitting QD_DHLA (λ_em= 550 nm; Figure 1a). The emission spectrum of the QD was overlapped with the absorption spectrum of mCherry (Supporting Information, Figure S1a), enabling the use of fluorescence resonance energy transfer (FRET) to track the association process. The FRET studies of QD-mCherry conjugates were performed in phosphate-buffered saline (PBS, pH 7.4). The loss of the QD emission was accompanied by an increase in the mCherry emission as the number of mCherry molecules per QD (N) increased (Supporting Information, Figure S1b). The FRET phenomena of the QD-mCherry conjugates were maintained in PBS that contained 10% serum (Figure 2a), demonstrating the stability of QD-mCherry conjugates in the presence of serum proteins.

Figure 2.

a) Stability of QD-mCherry conjugates in 10% serum-supplemented PBS for 4 h. b) Histidine disrupted the QD-mCherry interactions (QD-mCherry complexes were at a fixed conjugation ratio N= 6). c) TEM image of NPSC_QD-mCherry. d) Hydrodynamic diameter of NPSC_QD-mCherry was measured by DLS before and after 24 h incubation in 10% serum-supplemented media.

An important aspect of metal-coordination conjugation is that the complexes can be disrupted by competing molecules. The release of polyhistidine-tagged protein from the particle surface can be achieved by the treatment of the conjugates with histidine or its analogues (e.g., imidazole).^{[ 18 ]} Here we used histidine as a biocompatible external stimulus to investigate the binding events in the QD-mCherry conjugates. After addition of histidine to the solution of QD-mCherry conjugates (N= 6, where the fluorescence of QD was quenched by mCherry), the fluorescence recovery of QDs was accompanied by the decreasing fluorescence of mCherry (Figure 2b). The results indicated that histidine can disrupt the FRET process in the QD-mCherry conjugates and release mCherry from the QD surface in a concentration-dependent manner.

We hypothesized that negatively charged QD-mCherry conjugates (zeta potential of the conjugates was ca. −30 mV, Supporting Information, Figure S2) could be used as noncovalent crosslinkers to provide lateral stability to the NPSC structure. Anionic QD-mCherry conjugates and cationic AuNPs were self-assembled at the water-oil interface by means of electrostatic interactions to form NPSC_QD-mCherry (Figure 1b). NPSC_QD-mCherry was characterized by using transmission electron microscopy (TEM; Figure 2c) and dynamic light scattering (DLS; Figure 2d). NPSC_QD-mCherry presented a hydrodynamic diameter of 120 ± 50 nm and a higher stability in serum-supplemented media than the NPSC (Figure 2d and Supporting Information, Figure S3). The formation of stable NPSC_QD-mCherry confirmed the noncovalent crosslinking between anionic QD-mCherry conjugates and cationic AuNPs at the oil-water interface.

The high stability of NPSC_QD-mCherry in serum-supplemented media makes this capsule potentially useful in biological applications. The utility of NPSC_QD-mCherry was tested through uptake studies in HeLa cells with cellular uptake and distribution analyzed using confocal laser scanning microscopy (CLSM) and flow cytometry. NPSC_QD-mCherry presented a dominant red fluorescence (Figure 3a) inside the cells compared to the control experiments, which included mCherry and QD-mCherry conjugates (Supporting Information, Figure S4a and S4b). No green fluorescence was observed in the confocal images, indicating the FRET efficiency and hence stability of the QD-mCherry complexes. Higher intracellular fluorescence intensity was also observed with the NPSC_QD-mCherry relative to controls using flow cytometry (Supporting Information, Figure S4c).

Figure 3.

a) NPSC_QD-mCherry (N= 6) was incubated with HeLa cells for 1 h. a1) Bright field image, fluorescence images of a2) QD (λ_ex= 458 nm, λ_em= BP 505–530 nm), a3) mCherry (λ_ex= 543 nm, λ_em= LP 650 nm) and a4) FRET signals (λ_ex= 458 nm, λ_em= LP 650 nm). Emission filters: BP= band pass, LP= high pass. b) Co-localization studies of green-emitting lysotracker and NPSC_QD-mCherry (N= 6). NPSC_QD-mCherry was incubated with HeLa cells for 1 h, and green-emitting lysotracker (100 nM) was incubated with cells for 30 min prior to the microscopy experiments. b1) Bright field image, fluorescence images of b2) green-emitting lysotracker, b3) mCherry and b4) merged fluorescence images. To examine the effect of the external triggers, NPSC_QD-mCherry (N= 6) was first incubated with HeLa cells for 1 h, and then treated with c) histidine (100 mM), d) CQ (100 μM) and e) co-treated with histidine and CQ for another 1 h. f) Schematic illustration of the endocytosis of NPSC_QD-mCherry and cytosolic release of mCherry upon the co-treatment of histidine and CQ.

The intracellular distribution of NPSC_QD-mCherry was further investigated using green-emitting lysotracker due to the negligible fluorescence of QDs in the green channel (Figure 3a2). A significant co-localization of red fluorescent NPSC_QD-mCherry and green fluorescent lysotracker was observed after the incubation with HeLa cells for 1 h (Figure 3b) and 4 h (Supporting Information, Figure S5), which indicated that NPSC_QD-mCherry was taken up as expected via endocytosis^[19] and subsequently trapped in endosomes/lysosomes.

Endosomal/lysosomal entrapment limits the utility of delivered biomacromolecules.^{[ 20 ]} We investigated the abilities of two orthogonal chemical reagents, histidine and chloroquine (CQ), to serve as stimuli to regulate the intracellular distribution and binding phenomena of QD-mCherry conjugates. Histidine can be used to disassemble the QD-mCherry conjugates due to the competitive binding toward the hexahistidine-tags on mCherry; CQ is a weak base that can destabilize endosomal/lysosomal membranes.^{[ 21 ]} NPSC_QD-mCherry was incubated with HeLa cells for 1 h before treatment with each reagent. Punctate fluorescence was observed for mCherry in the cells that had internalized NPSC_QD-mCherry after treatment with histidine or CQ alone (Figure 3c and 3d). In contrast, mCherry was distributed throughout the cytosol after the co-treatment with histidine and CQ (Figure 3e). This outcome was verified through two channel studies, allowing tracking of QDs and mCherry simultaneously (Supporting Information, Figure S6). These results demonstrated a simple version of an AND logic gate,^[22] where both histidine-induced disassembly of QD-mCherry conjugates and CQ-mediated release of the free mCherry from the endosome/lysosome were required for cytosolic access. This release of mCherry was presumably facilitated by high concentration of intracellular proteins that would be expected to displace mCherry from the capsules. The uptake mechanism of NPSC_QD-mCherry and cytosolic release of mCherry have been summarized in Figure 3f.

The above experiments demonstrated the combined effects of histidine and CQ as external stimuli in modulating the cytosolic release of mCherry. We next used the hydrophobic core of the NPSC as a reservoir for CQ to form NPSC-CQ_QD-mCherry. After the internalization of NPSC-CQ_QD-mCherry into the cells, punctate fluorescence of mCherry was observed (Figure 4a, see Supporting Information, Figure S7a for color image), indicating that as above that dual triggering was required. As expected, disperse fluorescence of mCherry was observed in the cytosol after delivery of NPSC-CQ_QD-mCherry followed by treatment with histidine (Figure 4b, see Supporting Information, Figure S7b for color image). Significantly, these studies demonstrated the dual-delivery capability of NPSC-CQ_QD-mCherry; it can deliver both the hydrophilic proteins and hydrophobic endosome-disrupting agents by making use of both its exterior and interior simultaneously.

Figure 4.

a, b) NPSC-CQ_QD-mCherry (N= 6) was incubated with HeLa cells for 1 h followed by treatment with histidine (100 mM, 1 h). c) Fluorescence intensity profiles along with the lines drawn in a) and b). d) Pixel-count histograms of the fluorescence inside the region of interest as marked with a yellow line in a) and b). e) Ratio of cytosolic pixel counts to endosomal/lysosomal pixel counts of intracellular mCherry under different experimental conditions: NPSC_QD-mCherry only, NPSC_QD-mCherry with histidine, NPSC_QD-mCherry with CQ, NPSC_QD-mCherry with co-treatment of histidine and CQ, NPSC-CQ_QD-mCherry only and NPSC-CQ_QD-mCherry with histidine. (See Supporting Information, Figure S7 for color image). *p ≤ 0.05, *** p ≤ 0.001

The results showed qualitatively that the proteins can be released into the cytosol using appropriate triggers. Fluorescence intensity analysis was used to quantify the cytosolic and vesicular distribution of the intracellular fluorescent proteins.^{[ 23 ]} The cytosolic and endosomal/lysosomal fluorescence intensity can be determined by analyzing the fluorescence intensity profiles along with the line in Figure 4a and 4b. Pixels with high fluorescence intensities (200–255) in the punctate fluorescent spots were considered to originate from endosomal/lysosomal fluorescence, while pixels with low fluorescence intensities (30–85) observed in the disperse fluorescent area were considered to be due to cytosolic fluorescence (Figure 4c).

The fluorescence intensity profile of a single cell was measured to allow the analysis of the frequency distribution of pixels with high/low fluorescence intensities. As shown in Figure 4d, cells that had internalized NPSC-CQ_QD-mCherry showed a majority of pixels with very low fluorescence intensities. After treatment with histidine, however, a broad distribution of pixel intensities was observed, in agreement with the reported fluorescence signal distributions of proteins that escape from endocytic vesicles to the cytosol.^[23]

The frequency distribution of fluorescence intensities can be used to quantitatively compare the degree of fluorescence distribution in the endosome/lysosome and cytosol.^{[23a], [24]} Cytosolic fluorescent pixels (fluorescence intensities 30–85) in the whole cell were counted to determine the percentage of cytosolic mCherry when NPSC_QD-mCherry and NPSC-CQ_QD-mCherry were internalized into cells. Cytosolic fluorescence of mCherry was noticeable in NPSC_QD-mCherry upon co-treatment with histidine and CQ as well as NPSC-CQ_QD-mCherry with treatment by histidine alone (Supporting Information, Figure S8). The ratio of cytosolic pixel counts (fluorescence intensities 30–85) to endosomal/lysosomal pixel counts (fluorescence intensities 200–255) under different experimental conditions was shown in Figure 4e. The results indicated that the NPSC-CQ_QD-mCherry internalized into cells treated with histidine presented a significantly higher level of cytosolic fluorescence of mCherry compared to cells that had internalized NPSC_QD-mCherry without any treatment (ca. 10-fold). Moreover, a substantially higher degree of endosomal/lysosomal release was observed with the encapsulated CQ relative to the externally added CQ, presumably due to higher local concentration of CQ in the endosome/lysosome.

In summary, we have fabricated NPSC structures that use multiple supramolecular interactions to provide multifunctional materials. These nanocapsules are stabilized by combinations of supramolecular interactions that impart multifunctionality to the capsules. In particular, these capsules feature a dual release “AND gate” mechanism, where two stimuli are required for release of the cargo protein to the cytosol. These studies demonstrate that utilization of multiple supramolecular interactions in nanocapsules provides a potent strategy for the creation of responsive systems that serve as potential platforms for next-generation smart materials for biological and functional materials applications.