Modifying Polydiacetylene Vesicle Compositions to Reduce Non-Specific Interactions

Abstract

Polydiacetylene (PDA) vesicles provide useful stimuli-responsive behavior as well as by the modular structure afford a means for the design of sensing and delivery systems with tunable target specificity. To reduce inherent non-specific interaction with either anionic or cationic formulations of polydiacetylene vesicles, we explored the use of various lengths of poly(ethylene glycol) (PEG) amphiphiles for integration and polymerization within PDA vesicles. Our results established that as little as 1% of polyethylene glycol amphiphile integration into anionic vesicles was sufficient to significantly reduce non-specific association with mammalian cells. Similarly integrating a low percent of PEG amphiphile content within cationic vesicles could also significantly reduce non-specific cell association, and moreover reduced cytotoxicity. These results may be prove useful in augmenting PDA vesicles formulations for reduced non-specific interaction which is of particularly interest to enhancing selectivity in vesicles designed with integrated targeting moieties for sensing and drug delivery applications.

Graphical Abstrcat

1. Introduction

There has been extensive work in the use of diacetylene containing amphiphiles for the self-assembled formation of polydiacetylene (PDA) stimuli-responsive vesicles,^1-5 and due to their interesting colorimetric and fluorimetric properties,^6-8 PDA vesicles have been applied as sensing,^5,9-16 drug delivery,^14,17,18 and display technologies among others.^19-22 It is critical in many vesicle applications to minimize non-specific interactions, which represents promiscuous attachment with a broad range of surfaces by having favorable interaction energies that are not strongly influenced by geometric consideration of forces at the interface.^{23, 24} It is well-known that non-specific interaction can lead to undesirable outcomes such as vesicle-vesicle aggregates resulting in suspension instability²⁵ or unwanted interactions with non-target molecules that can result in false positive signals for the case of sensing applications^26-28 or lack of target-tissue localization for drug delivery applications.^29-31 Therefore, reducing non-specific interactions for PDA vesicles is a critical consideration for successful formulation, development, and implementation of PDA-based vesicle technologies.

A number of prior works have shown that to overcome non-specific interactions, poly(ethylene glycol) (PEG) can be displayed on the surface of vesicles.^32-36 This has proven very effective at excluding the presence of potential interacting component and is believed to be due to the high flexibility of the polymer chains, for example PEG-2000 which shows repulsive forces.^37-39 It has also been shown that the surface charge of cationic or anionic vesicles can be either retained or alternatively neutralized depending on the molar fraction of displayed PEG, where in some cases using only small fractions of PEG can provide stable vesicles without shielding the surface charge.⁴⁰ In this case, charge-induced interactions can be modulated by controlling PEG content.³⁷ Vesicle formulations have previously incorporated PEGylated amphiphiles showing improved vesicle stability when using PEG components in the range of 1 to 5 kDa and molar fractions of only 0.2 to 10%.^40,41 Researchers have also previously examined the chromic properties of polydiacetylene micelles (~10 nm in size) with displayed PEG-2000 moieties and have also found the PEG units help to extend the in vivo stability of the micelles.^42-44 PDA vesicles (~200 nm in size) displaying PEG-2000, as an amphiphile of stearate rather than an integrated polymerizable diacetylene amphiphile, have also been studied in regard to their physicochemical stability and release profile of encapsulated drug.¹⁸ In building on these foundational works, here we examine different formulations of polydiacetylene vesicles with integrated polymerizable bis-diacetylene amphiphiles of PEG-3400 or PEG-8000 in order to assess their impact on colloidal stability and non-specific interactions for both cationic and anionic vesicles.

Since surface charge character can give rise to attractive or repulsive electrostatic forces, we tested the impact of PEGylation on both anionic vesicles (assembled using 10,12 pentacosadiynoic amphiphile, herein referred to as PCDA) and cationic vesicles (assembled using PCDA conjugated with ethylenedioxy-bis-ethylamine, herein referred to as PCDA-EDEA). To characterize vesicle formulations in terms of their non-specific interactions with components of biomedical relevance, we examined the PEGylated vesicles propensity for association with cultured mammalian cells (HEK293). Due to the inherent fluorescence of the polydiacetylene vesicles, we could directly observe their non-specific association with the cells by fluorescence microscopy and determine the extent of vesicles attached to the cells by fluorescence activated cell sorting (FACS). As detailed in the following section, variations in vesicle PEGylation utilizing PEG3400-bis-PCDA and PEG8000-bis-PCDA were found to influence the degree of shielding from non-specific interactions depending on the molar ratios of PEG amphiphile displayed as well as the surface charge of the vesicles. The stability of the distinct formulations are also reported.

Continued work in understanding the effect of vesicle surface character and functionality on their stability and susceptibility to non-specific binding are important to the advancement of polydiacetylene vesicle technologies. The findings from this study hold significance in informing the design of polydiacetylene vesicles in which polymerizable PEG amphiphiles are incorporated to reduce non-specific interactions. Future works will continue to shed light on other formulations that may further improve the performance of polydiacetylene vesicles including biological applications of diagnostic and drug delivery.

2. Results and discussion

In this study, PEGylated amphiphiles comprised PEG3400-bis-PCDA and PEG8000-bis-PCDA which were synthesized as follows. A detail methodology is provided in the supporting information, but in brief PCDA was reacted under agitation with N-(3-dimethylaminopropyl)-ethylcarbodiimide (EDC), and N-hydroxysuccinimide (NHS) in dichloromethane (DCM) at room temperature to form PCDA-NHS. After purification, PCDA-NHS was used for three different subsequent coupling reactions with either EDEA or PEG-bis(amine) having M_n of 3,400 or 8,000, and the amphiphile product were purified by silica gel column chromatography. Before studying the impact of the molar ratio and size of PEGylated monomers with respect to reducing non-specific binding of vesicles, we first examined which PDA formulations incorporating the PEG-bis-PCDA amphiphiles could yield stable polymerized vesicle suspensions. To generate anionic (or cationic vesicles), PCDA (or PCDA-EDEA) amphiphiles were dissolved in dimethylsulfoxide (DMSO) at different proportions with PEG3400-bis-PCDA and PEG8000-bis-PCDA (1, 5, or 10% PEG with a consistent total molar concentration of diacetylene units) and sonicated in HEPES followed by overnight self-assembly at 4 °C for vesicle formation. Prepared samples were then exposed to 254 nm UV light for polymerization of the self-assembled vesicles, wherein appearance of a blue colored suspension is indicative of polymerization. The use of carboxylate terminated PCDA or amine moiety terminated PCDA-EDEA allowed us to impart negative or positive surface charge character to the vesicles, respectively. Sample formulations that yielded stable suspensions were further characterized in terms of the hydrodynamic radius, zeta potential, and spectral absorption. We observed that from 0 to 10% PEG8000 content, the vesicle suspension remained stable; however; at compositions greater than 10% the samples became aggregated within several days after UV polymerization. For the remainder of the study, we thus used vesicles with PEG contents of 10% or less.

In comparing the size of the resulting stable vesicles having different ratios of PEG-bis-PCDA to PCDA, it was found that the lognormal distribution of the hydrodynamic diameter of the vesicles was consistent for vesicles having between 1% to 10% content of the same PEG-bis-PCDA amphiphile. In contrast, a noticeable increase in the hydrodynamic diameter was observed in accordance with increased length of the incorporated PEG segment (PCDA vesicles with no PEG: 175 ± 2 nm; PEG3400 containing PCDA vesicles: 222 ± 2 nm; PEG8000 containing PCDA vesicles: 263 ± 6 nm). From zeta potential measurements, the anionic PCDA vesicles and cationic PCDA-EDEA vesicles did see a degree of neutralization in surface charge, albeit a minimal effect, when incorporating high proportions of PEG8000-bis-PCDA. From Figure 1, the zeta potential magnitude decreased to nearly 10 mV, which may perhaps explain the poor colloidal stability observed when incorporating vesicles with greater than 10% of PEGylated amphiphile. When evaluating the visible absorption of the polymerized vesicles, we observed a characteristic blue phase polydiacetylene spectra (Figure S1) having a maximum absorption of ~645 nm for each of the PCDA vesicle formulations having 0% to 10% PEG3400-bis-PCDA or PEG8000-bis-PCDA content.

Figure 1.

Zeta potential measurements of PCDA and PCDA-EDEA vesicles produced with different mole % of PEG8000-bis-PCDA amphiphile integrated within the vesicles.

In order to determine if the PEG3400 or PEG8000 displayed on the vesicles could provide a “shielding effect” in preventing non-specific interactions with mammalian cells, we applied the different vesicle suspensions to cultured HEK293 cells. Because the polydiacetylene vesicles possess an innate red fluorescence, we were able to examine the extent of vesicle co-localization with the cells by fluorescent microscopy as an indicator of non-specific interaction. In first looking at the anionic PCDA vesicles, we could observe a substantial amount of vesicle association with the HEK293 cells for the case of non-PEGylated vesicles (Figure 2) as revealed by the bright red fluorescent signal over-laying the location of each cell. In contrast, vesicles containing 1, 5, or 10% PEG8000-bis-PCDA had a dramatic reduction in their interaction with cells and almost no association could be observed. To determine if a shorter length PEG component would be able to invoke a similar protective effect in preventing non-specific interactions with cells, we examined the HEK293 cultures in the presence of 1, 5, or 10% PEG3400-bis-PCDA vesicles. As seen in Figure S2, the shorter chain PEG3400 displayed on the surface also nearly eliminated vesicle interactions with HEK293 and was comparable to the protective effect seen for PEG3400-bis-PCDA containing vesicles. No observable difference in the associated cell fluorescence was identified between the 1, 5, and 10% PEG3400-bis-PCDA or 1, 5, and 10% PEG8000-bis-PCDA containing vesicles.

Figure 2.

Phase contrast (bottom) and fluorescence microscopy (top) images of HEK293 cells cultured with (a) PCDA vesicles without PEG, (b) PCDA vesicles with 1% PEG8000-bis-PCDA, (c) PCDA vesicles with 5% PEG8000-bis-PCDA, and (d) PCDA vesicles with 10% PEG8000-bis-PCDA.

These result pointing to the ability of the PEGylated surface to reduce non-specific interactions for anionic PCDA vesicles becomes more interesting when taken in the context with the subsequent examination of the PCDA-EDEA vesicle behavior with cultured cells. We previously have shown that PCDA-EDEA vesicles, which possess a dense positive surface charge, will be taken up into cells and are highly cytotoxic.²⁵ In Figure 3(a), we can see this clearly from the high levels of vesicle association with the cells as revealed by the clustered fluorescence within the cells. Examining the impact of using 1, 5, or 10% PEG8000-bis-PCDA incorporated into the PCDA-EDEA vesicles, we can see from Figure 3(b)-(d) that there is a substantial decrease in non-specific interactions resulting from the PEGylation, as the extent of vesicle association with the cells is significantly decreased when the PCDA-EDEA vesicles are integrated with the PEG but most notably reduced when using 10% PEG8000-bis-PCDA vesicles. To quantitatively confirm these observations, cells cultured with the distinct vesicle formulation underwent FACS in order to individually count each cell and measure the magnitude of red fluorescence arising from its associated vesicles, wherein a higher fluorescent intensity reflects a more vesicles interacting with a given cell. From Figure 4, we can see the validation of our fluorescent microscopy interpretations in that PCDA vesicles with 1, 5, or 10% PEGylation provided a significant decrease in cell associated fluorescence, while PCDA-EDEA vesicles saw prevention of non-specific interaction to a greater extent when using 10% PEGylation than 1 or 5%. While the FACS results are in agreement with our microscopy results, it is important to note that flow cytometry does not distinguish between intracellular and membrane-associated vesicles. Nonetheless, the FACS results in combination with our fluorescence microscopy studies provide important information about the vesicle interactions with the cells. It has been noted in an interesting study that PEGylated surfaces that the normally repulsive forces of PEG can become attractive when PEG surfaces are forced into close proximity to a protein or membrane that is exposing proteins (such as a cell surface).^38,39

Figure 3.

Phase contrast (bottom) and fluorescence microscopy (top) images of HEK293 cells cultured with (a) PCDA-EDEA vesicles without PEG, (b) 1% PEG8000-bis-PCDA, (c) 5% PEG8000-bis-PCDA, (d) 10% PEG8000-bis-PCDA.

Figure 4.

FACS histograms revealing the relative amount of red fluorescent vesicles associated with cells for (A) PCDA and (B) PCDA-EDEA vesicles comprised of varying ratios (0, 1, 5, or 10%) of PEG-bis-PCDA amphiphile.

It is well known that cationic particles are taken up into certain cells, but less agreed upon are the role in which surface charge of particles regulate their mechanism of cell uptake.⁴⁵ It is known that many mammalian cells possess a significant negative surface charges arising from sialic acid displayed on the membrane which helps preventing cell aggregation.^46,47 Spontaneous translocation by cationic nanoparticles may occur by generation of transient holes in response to a strong local potential difference as one such means for particle uptake.⁴⁵ Nanoparticle cytotoxicity studies have attributed the high toxicity of cationic nanoparticles to their ability to cause lysosome swelling and membrane rupture due to the proton sponge,⁴⁸ where cytotoxicity increases with cationic charge density.⁴⁹ We examined here if PEGylation could mitigate the cytotoxicity that we had previously observed for PCDA-EDEA vesicles.²⁵ In looking at the MTT assay results for PCDA-EDEA vesicles with varying amounts of PEG8000-bis-PCDA, we find in Figure S3 that there is high cell viability for the HEK293 cultured with either PCDA or PCDA-EDEA vesicles when containing 1, 5, or 10% PEG amphiphile, which indicates reduction in cytotoxicity of the cationic PCDA-EDEA vesicles with PEGylations. This is in stark contrast to the PCDA-EDEA vesicles without PEG which are highly cytotoxic. This can be expected given that PEGylation marginally neutralized the surface charge of the PCDA-EDEA vesicles and also reduced non-specific interactions. In looking at these results in the context of existing literature, liposomes comprised of DOTAP and DOPS have been found to maintain their cationic and anionic surface charge character when using 0.2% PEG component but at 2% shifted the zeta potential to near-neutral (that is in the range of −10 to +10 mV).⁴⁰ This would indicate that charge screening by integration of PEG components is largely dependent on the vesicle or liposomal system and other groups have similarly reported that results are highly dependent on the amphiphiles used.⁵⁰ While research has shown that PEGylation of vesicles can decrease cellular uptake many of such studies had also incorporated high levels of PEGylation resulting in charge neutralization.^51,52

In summary, we found that incorporating PEG amphiphiles into PCDA vesicles caused a small increase in size concomitant with length of the PEG component and provided a small neutral shift in charge. Examining the vesicles in terms of their interaction with cultured mammalian cells we found that PEGylation was effective in preventing non-specific cell interactions to a greater extent for the anionic PCDA vesicles than the case of the cationic PCDA-EDEA vesicles. This may suggest that the PEG amphiphile components integrated into the vesicles could effectively prevent non-specific interactions arising from certain short range van der Waals forces or hydrogen bonding but are potentially less capable of inhibiting longer-range electrostatic forces as that seen for attractive forces between the negatively charged cell surface and the positively charge PCDA-EDEA vesicles. With our findings having established that even a small percentage of PEG amphiphile integrated in the anionic PCDA vesicles or PCDA-EDEA vesicles is sufficient to reduce non-specific binding for the case of the HEK293 cell surface, we hope to extend this toward future research in assessing vesicle formulation with integrated targeting moieties to provide highly selective vesicles for sensing and drug delivery applications.