Polydiacetylene Micelles in Nanomedicine and Beyond

Dedicated to the memory of our former Heads of Department, Dr. Charles Mioskowski (deceased June 2, 2007), and Dr. Bernard Rousseau (deceased April 16, 2021).

Abstract

In this account article, we give an overview of our contribution to the development of stable micellar carriers obtained by self-assembly and photo-polymerization of diacetylenic amphiphiles. The stabilized micelles can be loaded with active substances and used for diagnostic and therapeutic applications, or loaded with a metal catalyst to promote some synthetic transformations in fully aqueous medium.

Table of content

1 Introduction

2 Polydiacetylene Micelles Applied to Nanomedicine

2.1 From Amphiphilic Units to Micelles

2.2 In vivo Behavior of Micelles

2.3 Passive Targeting of Tumors with Micelles

2.4 Drug Delivery with Micelles

2.5 Towards Improved Delivery of Micelles to Tumors Using Sonoporation

2.6 Active Targeting with Micelles

2.7 Behavior of Micelles at the Cellular Level and Potential Cytotoxicity

2.8 Micelles for siRNA Transfection

3 Polydiacetylene Micelles Applied to Catalysis

3.1 Copper Nanoparticles in Micelles

3.2 Copper Salts in Micelles

4 Conclusion

#

Key words

micelle - nanomedicine - polydiacetylene - catalysis - self-assembly

Biographical Sketches

Edmond Gravel (left), Céline Demeese (middle), and Eric Doris (right) belong to the Nanosciences research group of the Alternative Energies and Atomic Energy Commission of France (CEA Paris-Saclay). Their research interests include supramolecular self-assemblies, heterogeneous catalysis, and nanomaterials for biomedical applications.

1

Introduction

The main challenge in nanomedicine is to efficiently transport active molecules across biological barriers to their targets. Some drugs also require particular formulations to compensate for their poor aqueous solubility, in vivo stability and bioavailability. The emergence of nanotechnologies has provided a full range of nanoscale carriers that are capable of safely transporting pharmaceuticals in vivo. However, some issues remain in terms of either biocompatibility, loading capacity, circulation time, biodistribution or pharmacokinetics. Examples of nanoscale delivery systems include polymers, liposomes, lipid nanoparticles, dendrimers, and micelles.[1]

Micelles consist of amphiphilic units that self-assemble into colloidal nanoparticles upon dispersion in water. Micelles have a central hydrophobic core and an outer hydrophilic shell interfacing with the surrounding aqueous environment. The inner core of micelles can serve as a reservoir, while their outer layer contributes to solubility and biocompatibility. When utilized as nanoscale carriers, micelles excel in solubilizing hydrophobic molecules. They offer several advantages over alternative delivery systems, including smaller size for deeper tissue penetration and high loading capacity.[2] However, since the assembly process is reversible, micelles also suffer from poor stability. Micelles can in fact disassemble back to unimers in dilute conditions, affecting their integrity and leading to uncontrolled release of the embarked payload. To address this stability problem, we explored a strategy involving stabilization of micelles by polymerization of the amphiphilic unimers.

Diacetylenic amphiphiles have long been studied in our group for the construction of micellar architectures.[3] Custom-synthesized amphiphilic unimers typically comprise a lipophilic chain incorporating a photo-responsive diacetylene group (DA), and a hydrophilic polar head, the nature of which may vary according to the intended use (Figure [1a]). In water, these amphiphiles self-assemble into supra-molecular micelles with a central hydrophobic domain and an outer hydrophilic shell, a process primarily governed by the critical micelle concentration (CMC, see Figure [1b]). Below the CMC, amphiphiles exist as individual species (unimers), but as concentration increases, unimers aggregate to form supra-molecular colloidal micelles. Under UV irradiation at 254 nm, the diacetylene units constituting the micelle undergo topochemical 1,4-additions, resulting in an ene-yne polymeric network in each individual micelle (Figure [1c]). While non-polymerized micelles are labile and disassemble below the CMC, their polymerized counterparts are stable and remain mostly unaffected by dilution. Depending on the chemical structure of the polar head, polydiacetylene (pDA) micelles with hydrodynamic diameters ranging from 6 to 20 nm are typically produced.

This account article provides an overview of our work on polydiacetylene micelles. The nanoscale carriers were assembled from diacetylenic amphiphiles that were photo-polymerized to stabilize the micellar structure. The polymerized micelles could then be loaded with active pharmaceuticals and used in the biomedical field for diagnostic and therapeutic applications. Micelles were also valorized as nanoreactors encapsulating copper for the catalysis of ‘click’ reactions in water.

Figure 1 (a) Generic structure of a diacetylene amphiphile. (b) Micelle assembly and polymerization. (c) UV-mediated diacetylene polymerization mechanism.

# 2

Polydiacetylene Micelles Applied to Nanomedicine

We initially studied polydiacetylene micelles in interaction with living systems, exploiting their small size, stability, loading capacity, and tunable topography for the targeting and delivery of bioactive payloads to solid tumors.[4] Our approach relied on the peculiar physiology of tumor tissues that allows passive diffusion of intravenously injected nanoscale carriers through the blood vessels irrigating the affected area. In fact, these neo-vessels have a fenestrated endothelium because of their inflammatory nature, making them more permeable, compared to normal blood vessels. As a result, the circulating nanoparticles can diffuse from the bloodstream to the adjacent tumor tissues and accumulate there, a process known as the ‘Enhanced Permeability and Retention’ (EPR) effect.[5] Provided they are not captured by the defense mechanisms of the organism to which they are administered and if they circulate long enough in the blood stream, micelles can passively target tumors. Strategies for the development of stealth and long-circulating micelles include engineering the micellar surface in contact with the biological medium. Micelles can be decorated with biocompatibilizing surface groups such as polyethylene glycol (PEG) and other substitutes to prevent macrophage scavenging and elimination.

2.1

From Amphiphilic Units to Micelles

The amphiphilic units that were developed in our team all share the same C₂₅-lipophilic backbone that is derived from commercially available pentacosa-10,12-diynoic acid. Using different synthetic approaches, amphiphiles incorporating variable biocompatibilizing groups have been synthesized, assembled into the corresponding micelles, and polymerized under UV irradiation. For example, diacetylenic amphiphiles made of PEG₃₅₀ (DA-PEG₃₅₀), PEG₂₀₀₀ (DA-PEG₂₀₀₀), or sulfobetain zwitterion (DA-Zwitt) polar heads afforded 8, 13, and 9 nm polydiacetylene micelles (pDA micelles), respectively (Figure [2]). PEG-coating of the micelles contributes to steric protection against plasma proteins (opsonins), which classically bind to objects that are of foreign origin (a process known as ‘opsonization’) and trigger phagocytosis by macrophages. On the other hand, zwitterions display a non-fouling effect attributed to the scrambling of the ion pairing between proteins and the surface of the micelles. These properties enable micelles to delay their uptake by macrophages and circulate longer in the blood stream.

Figure 2 Selected examples of diacetylene amphiphiles with different polar head groups and their corresponding polydiacetylene micelles

# 2.2

In vivo Behavior of Micelles[6] ^, [7]

As mentioned above, one of the prerequisites for the targeting of tumors by a nanoscale carrier is its ability to circulate long enough in vivo. To assess the impact of surface chemistry on the in vivo fate of pDA micelles, non-invasive near infrared (NIR) imaging was employed. In fact, NIR light (i.e., 700–1000 nm) has little interaction with biological tissues (low absorbance and negligible auto-fluorescence) and can penetrate to depths of several centimeters. Polydiacetylene micelles with three different surface coatings (PEG₃₅₀, PEG₂₀₀₀, or zwitterionic) were made trackable by encapsulating a NIR-emitting probe (i.e., lipophilic DiR carbocyanine) in their core. The encapsulation process involved simple ultrasonication of the micellar solution in the presence of the dye. DiR-loaded micelles (typically 1 wt% dye loading) were investigated in a murine model as regards their pharmacokinetic properties. Blood residence time of the micelles was evaluated after intravenous injection to nude mice, by collecting blood samples at different time points, and measuring the remaining fluorescence in the plasma (Figure [3]).

Figure 3 Pharmacokinetic profiles of polydiacetylene micelles with different surface chemistries

For all three micelle types, a drop in blood concentration was observed during the first two hours, followed by a more stable phase. Throughout the first phase, the signal of pDA-PEG₃₅₀ micelles decreased abruptly, reaching ca. 25% of the initial fluorescence after 1 hour. As a comparison, pDA-PEG₂₀₀₀ micelles and pDA-Zwitt micelles had an overall slower clearance from the blood, with 60% and 45% of the micelles left circulating after 1 hour, respectively. Over a 24-hour period, virtually all the fluorescence associated with pDA-PEG₃₅₀ micelles had vanished, indicating full elimination from the blood stream, whereas approximately 15% remained circulating for pDA-PEG₂₀₀₀ micelles and pDA-Zwitt micelles, highlighting the protection provided by higher-molecular-weight PEG and zwitterion against opsonization. This permitted prolonged circulation of the corresponding micelles in the blood, compared with low-molecular-weight PEG₃₅₀.

# 2.3

Passive Targeting of Tumors with Micelles

The DiR-loaded micelles were then injected intravenously to mice bearing subcutaneous tumors on their backs (MDA-MB-231 cells, derived from human breast cancer). The biodistribution and accumulation of micelles in the tumor area were monitored using planar NIR-imaging. Over an initial period of 24 hours, the fluorescence signal increased in the whole body, suggesting diffusion of the micelles in tissues, and progressively decreased in the ensuing days, indicating slow excretion of the micelles. A signal was also detected in the liver and spleen, but no fluorescence was observed in the bladder and kidneys, indicating hepato-biliary, rather than urinary, excretion of the micelles. Dorsal view fluorescence images, recorded 24 hours post-injection, showed a marked contrast between the tumor area and the surrounding healthy tissue (Figure [4]). NIR imaging confirmed effective passive accumulation in the tumor via the EPR effect for all micelle types. However, in the case of pDA-PEG₃₅₀ micelles, the uptake was moderate compared to pDA-PEG₂₀₀₀ micelles and pDA-Zwitt micelles that had a more significant accumulation, which gradually rose and peaked after 48 hours. Of note, the signal from pDA-PEG₂₀₀₀ micelles remained constant in the tumor area over a week, while that from pDA-Zwitt micelles slowly faded.

Figure 4 Impact of surface chemistry on the tumor accumulation of fluorescently labeled micelles (images were recorded 24 hours post-injection)

# 2.4

Drug Delivery with Micelles

As per the preliminary in vivo evaluation, the best tumor targeting properties were associated with pDA-PEG₂₀₀₀ micelles, with ca. 3% of the injected dose that accumulated in the tumor area. pDA-PEG₂₀₀₀ micelles thus appeared as promising candidates for local delivery of anticancer drugs, and were selected as carriers. Paclitaxel (PTX), which is the active pharmaceutical ingredient of Taxol^®, an established drug used in the treatment of various cancers, was encapsulated in pDA-PEG₂₀₀₀ micelles. The encapsulation process involved ultrasonication of the preformed micelle with the lipophilic PTX, resulting in 10 wt% drug loading. Paclitaxel-loaded micelles were stable with no signs of precipitation, even after storage for several months.

Potency of the paclitaxel formulation based on pDA-PEG₂₀₀₀ micelles was first evaluated in vitro and its biological activity (cytotoxicity) was compared with that of the commercial Taxol^® formulation. Cytotoxicity assays were conducted on the MDA-MB-231 cell line and showed comparable half-maximal inhibitory concentration (IC₅₀) values, in the tens of μM range. This result indicated that paclitaxel retained its cytotoxic potency upon encapsulation in micelles. As a control experiment, empty pDA-PEG₂₀₀₀ micelles showed no interference with the biological system.

We next proceeded with the evaluation of the therapeutic efficacy of paclitaxel-loaded micelles in vivo (Figure [5]).[6] Tumor-bearing mice were split into three groups and treated with either: (i) paclitaxel-loaded micelles, (ii) Taxol^®, or (iii) 0.9% NaCl (control). Treatments were administered by intra-peritoneal injections at a dose corresponding to 5 mg kg^–1 of PTX (two injections/week for two months). During the experiments, we monitored tumor volumes and observed a significant decrease in tumor growth in the groups treated with Taxol^® and paclitaxel-loaded micelles, compared to the control group. After two months, mice treated with Taxol^® and PTX-loaded micelles had tumors 3.3 and 4.5 times smaller than control mice, respectively. These results demonstrate that pDA-PEG₂₀₀₀ micelles could perform satisfactorily for the targeted delivery of paclitaxel to tumors, and were even more potent than the reference formulation of the drug (i.e., Taxol^®) in reducing tumor growth in vivo.

Figure 5 Anti-cancer activity of PTX@pDA-PEG₂₀₀₀ micelles: (a) Intra-peritoneal injection of PTX-loaded micelles to tumor bearing mice. (b) In vivo tumor growth inhibition (MDA-MB231 xenografts).

In the same vein, we also valorized pDA-PEG₂₀₀₀ micelles as carrier for drug delivery to inflammatory atherosclerotic lesions.[8] In fact, passive targeting could similarly be achieved as the microvessels irrigating the atherosclerotic plaque also have a disrupted endothelium. This altered functionality favors diffusion of the micelles from microvessels and their accumulation in atherosclerotic plaques. The nuclear liver X receptors agonist drug GW3965 was encapsulated in the core of the micelle but we observed rapid leaching of the drug. The origin of the instability of the micellar formulation was attributed to the presence of the hydrophilic carboxyl group of GW3965. It was therefore decided to adjust the strategy by synthesizing a more lipophilic prodrug analogue of GW3965 (Figure [6a]) with a hydrophobic alkyl ester chain in lieu of the anionic carboxylate. As the ester bond is labile and liable to react under hydrolytic conditions or with esterase enzymes, the prodrug should retain the pharmacological properties of GW3965. In vivo evaluation (mouse model of atherosclerosis, i.e., LDLr^–/– mice) of the GW3965 analogue formulated in pDA-PEG₂₀₀₀ micelles proved effective in targeting the atherosclerotic plaque region (Figures [6b] and 6c) and activating LXR receptors with no adverse effect on circulating lipid levels and hepatic lipid metabolism, a phenomenon classically encountered with the molecular GW3965 drug.

Figure 6 (a) Structure of GW3965 and its ester prodrug analogue. (b) Schematic injection protocol. (c) Ex vivo fluorescence imaging of the whole aorta, 72 hours after intravenous injection.

# 2.5

Towards Improved Delivery of Micelles to Tumors, Using Sonoporation[9]

As previously stated, the primary mechanism underlying the uptake of micellar nanocarriers in tumors is the EPR effect. However, the effectiveness of EPR depends of several factors, including tumor type, anatomical location, physicochemical nature of the carriers, and level of blood perfusion in tumor tissues. In an attempt to increase micellar uptake by tumors, while overcoming the above limitations, we sought to take advantage of non-invasive focused sonoporation involving ultrasonic waves associated with injected microbubbles. The resulting acoustic phenomenon transiently enhances the permeability of the vasculature, which could be beneficial for the translocation of micelles from the blood compartment to tumor tissues. For this study, in addition to optical imaging, we introduced a positron emission tomography (PET) imaging modality to track and quantify micelles in vivo. Micelles could be labeled with radioactive zirconium-89 by mixing 10% of a tailor-made diacetylenic deferoxamine amphiphile (acting as zirconium ligand) to the classical PEG₂₀₀₀ construct before polymerization (Figure [7a]). The ⁸⁹Zr-labeled-pDA micelles were injected intravenously and evaluated in a subcutaneous glioblastoma (U87-MG) mouse model, treated with or without sonoporation (Figure [7b]). Although sonoporation did not significantly increase the overall accumulation of micelles in the tumor, it did lead to a clear improvement in the homogeneity of their distribution within the tumor volume. In fact, under sonoporation we observed a reduced standard deviation in the micelle accumulation in the tumor, compared to the tumor not in contact with the ultrasound source (Figure [7c]). This observation suggests that sonoporation could help in delivering nanocarriers more homogenously inside of the tumors, thus limiting the inter-individual variability often observed with EPR-based therapeutic interventions.

Figure 7 Overview of the evaluation sequence for assessing the impact of sonoporation on EPR-mediated accumulation of pDA-PEG₂₀₀₀ micelles

# 2.6

Active Targeting with Micelles

Despite the advantages offered by the EPR effect, which enables micelles and their payload to reach the tumor area, the so-called ‘passive’ targeting is only tissue-specific. EPR effect does not bring any selectivity in the distribution pattern at the cellular level, once micelles have diffused into the tumor. To achieve ‘active’ targeting and specific internalization into a particular subset of cells, micelles can be functionalized with ligands capable of interacting with receptors overexpressed at the surface of the targeted cells. Ligand–receptor interaction should enable selective cellular uptake.

Functionalization of pDA micelles with a ligand proved straightforward thanks to the versatility of the micellar construct. We investigated the assembly of ‘clickable’ micelles and their post-functionalization. We chose to install alkyne groups at the surface of the carriers to achieve conjugation via a ‘click’ reaction with an azido-terminated ligand (Figure [8]). Micelles were assembled from a mixture of simple DA-PEG₂₀₀₀ amphiphiles and ‘functional’ DA-PEG₂₀₀₀ amphiphiles carrying a primary alkyne group at the end of the PEG unit. The resulting micelles were, as previously, polymerized by UV irradiation and further functionalized with targeting ligands. Two families of ligands were investigated and clicked to micelles, namely a small molecule (e.g., biotin)[10] and an aptamer oligonucleotide (e.g., ACE4 aptamer – an Annexin A2 ligand).[11] While the incorporation of biotin to micelles was direct (reaction with azido-biotin in the presence of copper sulfate and sodium ascorbate), that of ACE4 aptamer proceeded in two steps: ‘click’ conjugation of the micelle with an azido-oligomer (spacer G), followed by hybridization of the newly ‘clicked’ spacer G with the targeting ACE4 aptamer. Micelles were made fluorescent by encapsulation of a dye (DiO) and evaluated individually in interaction with MCF-7 cells, which advantageously overexpress both the biotin and Annexin A2 receptors. It was found that the ligand-functionalized micelles were more readily internalized than ‘naked’ micelles and that the internalization process was mediated by the tethered ligands. As an example, surface functionalization with biotin (Figure [8a]) enabled accelerated micelle internalization in MCF-7 cells. DiO-loaded biotinylated micelles were incubated with cells and the internalization rate was measured by fluorescence-assisted cell sorting (FACS). The accelerated internalization effect was dependent on the grafting level of biotin and best results were obtained for micelles bearing 25% biotin at their surface. The uptake was found to be 10 times faster than that of non-biotinylated micelles (Figure [8b]), and the addition of competing free biotin dramatically decrease micelle uptake, suggesting a specific receptor-mediated process.

# 2.7

Behavior of Micelles at the Cellular Level and Potential Cytotoxicity

The influence of the surface chemistry of pDA micelles on their behavior at the cellular level was studied by fluorescence microscopy and fluorescence-assisted cell sorting. Pegylated (PEG₂₀₀₀) diacetylene amphiphiles with either cationic, anionic, or neutral terminal groups (i.e., ammonium, carboxylate, or methoxy, respectively) were assembled into pDA micelles, labeled with a fluorescent dye, and their cellular uptake, internalization pathway and intracellular fate were assessed.[12] Our investigations revealed that the surface charge of pDA micelles strongly influenced the kinetics of the uptake (cationic > neutral > anionic pDA micelles) but had little effect on internalization pathway of the micelles, which proceeded mainly via a caveolae route. Surface chemistry also had a minor impact on the intracellular fate of micelles, all of which ended up in the lysosomal compartments.

Figure 8 Strategy for the surface functionalization of pDA micelles with active targeting ligands using ‘click’ chemistry

Compared to their non-polymerized counterparts, pDA-PEG micelles exhibit much reduced cytotoxicity with little to no effect on mitochondrial activity, apoptosis/necrosis, secretion of proinflammatory cytokines or genotoxicity.[13] The core polymerization of the micelles not only confers greater stability to the assembly but also reduces the interference with cell membranes, as observed with non-polymerized micelle that induce membrane permeabilization. In addition, pDA-PEG micelles do not activate the complement component 1q (C1q), a protein complex involved in the innate immune response, suggesting no influence of pDA micelles on the various physiological processes mediated by C1q.[14]

# 2.8

Micelles for siRNA Transfection[15]

In addition to conventional chemotherapeutics, small interfering RNAs (siRNAs) and their gene-silencing properties have recently attracted attention for the treatment of diseases such as cancer. Therapeutic efficacy of siRNAs depends largely on their active transfection, as free nucleic acid do not freely pass cell membranes. Transfection can be triggered by complexing siRNA with a nanocarrier that provides protection and promotes cellular uptake and delivery. We became interested in investigating polydiacetylene micelles as transfection tools since their surface can be tuned as needed to adjust electrostatic binding properties to siRNAs. When we initiated this work, we had only a limited understanding of how the cationic regions of the carriers would influence the overall transfection efficacy. Four different micelles were thus constructed from diacetylenic amphiphiles having the same backbone, yet incorporating variable alkylation patterns at the terminal amino-group position, leading to primary, secondary, tertiary and quaternary ammonium micelles (pDA-AM micelles, Figure [9]). The cationic micelles were systematically studied with respect to transfection potency to decipher the role played by the complexing ammonium group substitution on the binding, transport, and release of siRNA.

Figure 9 Complexation of siRNA at the surface of cationic pDA micelles

The overall cationic character of the micelles was confirmed by zeta potential measurements (comprised between +15 and +30 mV), and their electrostatic adherence to siRNA was evaluated by agarose gel electrophoresis, which allowed micelle-bound siRNAs and freestanding ones to be distinguished. We next evaluated cellular delivery ability, using a blend of siRNAs (AllStars Death Control) that target essential genes for cell survival. siRNA transfection was evaluated by measuring the proliferation/survival of HeLa cells and we observed a potent effect of the micelles incorporating primary and secondary amines. Performances of the two cationic micelles were in the same range as those of a commercially available carrier system (Lipofectamine). We rationalized the superior activity of these micelles by the enhanced stability of their complexation with siRNA, as well as their ability to release the oligonucleotidic payload intracellularly, which is mediated by the so-called ‘proton sponge’ effect. The latter is responsible for the escape of siRNAs from lysosomes in which they are trapped after cellular internalization of the siRNA-micelle complex. The buffering capacity of free amines induces an increase in lysosomal pH along with an influx of Cl^– ions. The resulting osmotic swelling triggers disruption of the lysosome and release of siRNAs into the cytoplasm.

In addition to the polydiacetylene micelles accounted herein, our group has also contributed to the development of other micelles types (non-polymerized), mostly for biomedical applications. Selected examples include: (i) activatable micelles capable of releasing an active payload upon triggering of a photolabile nitrobenzyl linkage connecting the hydrophilic and hydrophobic regions of the amphiphile;[16] (ii) perfluorinated micelles that can be visualized in vivo by ¹⁹F-magnetic resonance imaging and used for diagnostic purposes;[17] and (iii) micelle-forming amphiphilic drug conjugates, made from biologically active epipodophyllotoxin (topoisomerase inhibitor) that was covalently inserted in the amphiphilic unimer.[18]

As discussed throughout the above section, the core of the micelles allowed the encapsulation of hydrophobic pharmaceuticals and their dispersion in aqueous media. Applied to the biomedical field, this feature was exploited by us for the development of drug delivery systems and imaging tools. The central micellar domain can also accommodate other species such as an active catalyst payload and the resulting hybrid micelle used as a nanoreactor, in which poorly water-soluble reagents accumulate transiently and react, enabling reactions in a 100% aqueous medium.

#
# 3

Polydiacetylene Micelles Applied to Catalysis

The alkyne-azide 1,3-dipolar cycloaddition was initially introduced by Huisgen in 1963 as a thermally activated cycloaddition reaction. Advances by Meldal and Sharpless led to the more recent discovery of copper-catalyzed azide–alkyne cycloadditions (CuAACs) that offer milder conditions, better regioselectivity, and better functional groups tolerance. However, CuAACs traditionally require Cu(I) catalysts in a mixture of aqueous/organic solvents, and the catalysts are difficult to recover. Bertozzi also introduced copper-free alternatives but they involve complex cyclooctynes dipolarophiles.

We became interested in sustainable ‘click’ chemistry approaches by trying to promote copper-catalyzed cycloaddition reactions in pure water, while offering the possibility of recovering and recycling the active catalytic species. With these considerations in mind, our strategy was to encapsulate copper into stabilized polydiacetylene micelles in such a way to convert the active catalyst into a stable and semi-heterogeneous species. In fact, micelles have previously been shown to be beneficial to organic transformations run in aqueous media, as the lipophilic core of micelles can accommodate hydrophobic reagents, enabling their dispersion and concentration.[19] In our approach, the ‘nanoreactor’ properties of the micelles were simply combined with copper encapsulation to produce a semi-heterogeneous nanohybrid system. The colloidal catalysts were valorized in effecting CuAAC reactions in water. We investigated two different but closely related approaches, based on the incorporation of copper nanoparticles or copper salts into polydiacetylene micelles to provide access to catalytically active nanohybrid systems.

3.1

Copper Nanoparticles in Micelles[20]

Cuprous oxide nanoparticles (Cu₂O NPs) were readily prepared by heating copper(I) acetate in the presence of oleic acid and trioctylamine. The process led to the formation of 9 nm spheroidal nanoparticles coated with hydrophobic oleic acid ligands. In parallel to the synthesis of Cu₂O NPs, polydiacetylene micelles were assembled from amphiphilic units made of a C₂₅-diacetylene chain and a polyethyleneglycol (550 Da) polar head. As before, amphiphilic units were dispersed in water and stabilized by photo-polymerization. The oleic acid-coated Cu₂O NPs were then encapsulated in the core of the micelles by ultrasonication in water, affording a stable dark-green colloid. Copper concentration of the micellar suspension was measured by inductively coupled plasma mass spectrometry ([Cu] = 7 mM) and the hydrodynamic diameter of the final nanohybrid assembly was approximately 30 nm (Figure [10]).

Figure 10 Encapsulation of copper oxide nanoparticles (coated with oleic acid) in pDA-PEG₅₅₀ micelles and application to the promotion of Huisgen cycloadditions in water.

The catalytic properties of the colloidal catalyst (i.e., Cu₂O nanoparticles encapsulated in polydiacetylene micelles) were then assessed on a model reaction involving benzyl azide and phenylacetylene, with 0.35 mol% Cu-loading in water and at room temperature. Under the above reaction conditions, we were pleased to observe the clean formation of the expected triazole unit in 99% yield after 24 hours. As a comparison, running the reaction under the same conditions but in the presence of the heterogeneous ‘micelle-free’ Cu₂O NPs led to a sluggish reaction (15% yield after 24 hours), and the classical copper sulfate/sodium ascorbate ‘click’ conditions led only to 55% conversion after 24 hours.

We also investigated the possibility to recycle and reuse our nanohybrid catalyst. Five successive cycloaddition reactions were conducted with the same catalyst batch, which was recovered after each run and subsequently reused in the next reaction. Upon completion of the reaction, the aqueous phase was simply extracted with ether, allowing the recycling of the aqueous phase containing the colloidal catalyst. No loss in catalytic performance was detected over the five consecutive cycles.

The potency of the nanohybrid-catalyzed dipolar cycloaddition reaction was then looked at on a range of other alkynes and azides. Various azido derivatives were reacted with phenylacetylene, and other alkynes with azidomethyl phenyl sulfide. All these reactions gave access to the expected triazole products in very good yields and 100% regioselectivity after 24 hours reaction time. For selected examples, see Table [1].

Table 1 Examples of Cycloadducts Obtained from Cu₂O@pDA-PEG₅₅₀ Micelle-Catalyzed Huisgen Reactions in Water

Although Cu₂O nanoparticles have already been reported as ‘click’ reaction promoters in aqueous media, these systems classically afford lower yields and require higher concentrations of active copper catalyst. In addition, inert atmosphere, organic solvents, and heating were sometimes needed to obtain satisfactory conversions. The micellar nanohybrid combines here an active colloidal copper catalyst center and a micellar environment behaving as a nanoreactor. The latter offers some beneficial effects in that it promotes transient aqueous dispersion of the reagents and their concentration at the vicinity of the catalytically active copper sites. The system operates efficiently with a variety of organic compounds in water, under ambient air conditions, at room temperature, and with catalyst loadings as low as 0.35 mol%.

# 3.2

Copper Salts in Micelles[21]

Pursuing the use of micelle-encapsulated copper for the promotion of semi-heterogeneous ‘click’ reactions in water, we have more recently developed an alternative system. The latter still relies on the use of polydiacetylene micelles, but with two main differences: (i) copper salts were used as the source of the catalyst instead of Cu₂O nanoparticles, and (ii) the active metal was incorporated into the micelles by complexation with a custom-made diacetylene amphiphilic unit.

Our design of the amphiphilic units to be assembled into micelles and photopolymerized took into account the prerequisite that it should also behave as a ligand for copper salts. We thus proceeded with the strategic inclusion of an amino-triazole group between the hydrophobic diacetylene chain and hydrophilic PEG₅₅₀ polar head. The embedded amino-triazole group is expected to create a favorable environment for copper complexation. The amphiphile incorporating the amino-triazole motif was then treated with stoichiometric amounts of CuCl₂ to afford the Cu-amphiphile complex. The copper-containing amphiphile was thereafter assembled into the corresponding [Cu]pDA-PEG₅₅₀ micelle by dispersion in water, and irradiated under UV light to trigger topochemical polymerization of the diacetylene groups (Figure [11]). The micellar solution was finally subjected to dialysis against water to remove unbound amphiphiles and free copper salts. DLS analysis of the polymerized micelles containing copper indicated an average hydrodynamic diameter of 9 nm and the copper concentration was found to be 1 mM.

Figure 11 Assembly of copper-complexing DA-amphiphiles into [Cu]pDA-PEG₅₅₀ micelle used for the promotion of Huisgen cycloadditions in water

Micelles encapsulating copper salts were then evaluated in the catalysis of the Huisgen 1,3-dipolar cycloaddition. In contrast to previously documented methods, which typically require the addition of a reductant (e.g., sodium ascorbate) to trigger the reduction of Cu(II) to Cu(I), we initially opted for a more direct approach as we wanted to exploit the inherent one-electron transition from Cu(II) to Cu(I) that occurs spontaneously upon homo-coupling of terminal alkynes. This reaction, known as the ‘Glaser–Eglinton–Hay coupling’, leads to the formation of dimeric alkynes with concomitant reduction of Cu(II) to Cu(I). As model substrates, we selected benzyl azide and phenylacetylene and the reaction was carried out under air, at room temperature, and in pure water as the solvent. To promote in situ generation of the catalytically active Cu(I) species, cycloaddition reactions were conducted with an excess (2 equivalents) of the alkyne partner.

When benzyl azide and phenylacetylene were reacted with 0.2 mol% of micellar copper in water, a nearly quantitative conversion was observed within 24 hours, yielding 98% of the expected triazole. These results have to be compared to those obtained under conventional ‘click’ reaction conditions, i.e., copper sulfate/sodium ascorbate, which led only to 48% conversion after the same time. Additional control experiments, including the use of CuCl ₂ alone, displayed limited catalytic efficacy.

Applicability of the Cu-micelle-catalyzed dipolar cycloaddition was then evaluated on a range of other organic alkynes and azides. Under our refined reaction conditions, all the investigated substrates underwent smooth 1,3-dipolar cycloaddition in very good yields. However, when we turned our attention to catalyst recycling, we were unpleasantly surprised to find out that slow deactivation of the catalyst had taken place. In fact, after completion of the reaction, triazole products were extracted from the aqueous phase using an organic solvent, and the aqueous phase containing the colloidal catalyst was reused as is. Although the catalytic system retained some activity, we observed a gradual decrease in performance, leading to extended reaction times (from 24 to 48 hours). These findings were unexpected since ICP-MS analysis of the organic extracts did not show any copper that had leached from the aqueous phase. Therefore, the gradual deactivation of the catalyst was tentatively attributed aerobic oxidation of copper. Our attempts to shield the catalyst from oxidation by conducting reactions under an inert argon atmosphere were unsuccessful, leading to sluggish cycloadditions. Given the suboptimal catalyst recycling performance, we decided to explore the potential advantage of introducing some reducing sodium ascorbate to the reaction mixture.

Under the revised reaction conditions (i.e., addition of two equivalents of sodium ascorbate), we were able to run five consecutive cycloaddition reactions between benzyl azide and phenylacetylene using the same batch of the Cu-micelle catalyst. Gratifyingly, not only could the catalyst be recycled and reused without any loss in catalytic performance, but reactions were also completed in just 2 hours (compared with 24 hours for the ascorbate-free conditions). Triazoles were systematically obtained with yields ≥95%. Notably, only one equivalent of the alkyne partner was required in the case of the ascorbate reaction, instead of two equivalents under ascorbate-free conditions. Taken together, these observations suggest that the deactivation detected under ascorbate-free conditions was a result of in situ aerobic oxidation of the active copper species.

The ascorbate/Cu-micelle conditions were thereafter applied to more complex substrates than benzyl azide and phenylacetylene, and performances were compared to those of the initial ascorbate-free conditions (Figure [12]). Coumarin-derived triazole (from azido-coumarin and phenylacetylene) and biotin triazole (from benzyl azide and biotin-alkyne) were obtained in nearly quantitative yield via the ascorbate/Cu-micelle procedure within 2 hours. In contrast, the ascorbate-free process required 36 hours to produce the same products in satisfactory yields. These results underscore not only the effectiveness of ascorbate-mediated ‘click’ reactions but also the mildness of the operating conditions (with and without ascorbate), which allowed for efficient access to triazoles in pure water.

Figure 12 Examples of cycloadducts obtained from [Cu]pDA-PEG₅₅₀ micelle-catalyzed reactions with or without sodium ascorbate. Reaction conditions with ascorbate: azide (1 equiv), alkyne (1 equiv), sodium ascorbate (2 equiv), H₂O, Cu-micelle (0.2 mol% of Cu), room temp. Reaction conditions without ascorbate: azide (1 equiv), alkyne (2 equiv), H₂O, Cu-micelle (0.2 mol% of Cu), room temp.

Again, the micellar architecture played here a pivotal role in facilitating the aqueous dispersion of the reagents and their concentration at the vicinity of the active catalytic sites. Micelles not only served as a colloidal stabilizer for copper salts but also functioned as a nanoreactor. In addition, polymerization of the micelles contributed favorably to the overall performances of the nanohybrid catalyst as we observed a dramatic increase in reaction times when micelles were not polymerized. In the absence of ascorbate, the non-polymerized micelle containing copper promoted the click reaction between benzyl azide and phenylacetylene but required 40 hours reaction time. As a comparison, the polymerized micelle catalyzed the same transformation in 24 hours only. The same trend was observed in the presence of ascorbate, since the non-polymerized micelle required 6 hours to catalyze triazole formation, while the polymerized micelles needed just 2 hours. Polymerized micelles hence outperformed their non-polymerized counterparts likely because of the intrinsic dynamic nature of the latter. In non-polymerized micelles, an exchange of copper-containing amphiphiles between micellar arrays can take place. This phenomenon reduces the time spent by the reagents in close contact with the metal centers in the micelles, thereby increasing overall reaction times.

#
# 4

Conclusion

The various investigations carried out in our group demonstrated that polydiacetylene micelles are versatile by nature, as their engineering is modular, allowing great flexibility in the construction of tailor-made nanocarriers with the desired properties. The presence of a photoreactive group in the starting amphiphilic units enables the supramolecular assembly to be stabilized by simple UV irradiation, making polydiacetylene micelles more robust than their non-polymerized counterparts.

In the context of biomedical applications, we have exploited the biocompatible micelles to load biologically active compounds (molecular drugs, biologics) that could be efficiently delivered to their site of action, both in vitro and in vivo (targeted delivery). Imaging agents (fluorescent, radioactive) have also been associated with micelles so that the biodistribution and the fate of the nanocarrier could be monitored using imaging techniques. The ability to combine drug delivery and imaging in a single object makes pDA micelles potential theranostic tools for diagnosis, therapy, and real-time monitoring.

In the context of catalytic applications, the modularity of the micellar construct allowed the encapsulation of copper in two different forms (i.e., nanoparticle and salt), and the copper-loaded micelles were used in the semi-heterogeneous catalysis of ‘click’ reactions in pure water. Micelles acted as nanoreactors, favoring aqueous dispersion of the reagents, their concentration in close vicinity to the active sites of the catalyst, and minimizing the interference of the aqueous medium on the reaction outcome. Triazole products were consistently obtained more satisfactorily than under conventional CuSO₄/ascorbate conditions, and the colloidal catalyst was easily recovered and reused.

In the aforementioned examples, polydiacetylene micelles benefit from some key advantages such as improved stability, compactness, load-bearing capacity and, above all, a modular structure that can readily be adapted to the foreseen application. As such, polydiacetylene micelles provide a valuable and polyvalent platform for further applications in nanomedicine and beyond.

#
#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

Contributions from our biologist/physicist/physicochemist colleagues to this work are gratefully acknowledged. Special thanks to Dr. Frédéric Ducongé, Dr. Guillaume Pinna, Dr. Charles Truillet, Prof. Elias Fattal, Dr. Wai-Li Ling, Dr. Sébastien Mériaux, and Dr. Laurent Devel for their invaluable help in the realization of our common research projects. The ‘Service de Chimie Bioorganique et de Marquage’ (SCBM) is a partner of NOMATEN, a Centre of Excellence in Multifunctional Materials for Industrial and Medical Applications (EU H2020 Teaming #857470).

References
1 Ogier J, Arnauld T, Doris E. Future Med. Chem. 2009; 1: 693

Crossref PubMed Search in Google Scholar
2 Gravel E, Doris E. Future Med. Chem. 2018; 10: 1137

Crossref PubMed Search in Google Scholar
3 Ogier J, Arnauld T, Carrot G, Lhumeau A, Delbos J.-M, Boursier C, Loreau O, Lefoulon F, Doris E. Org. Biomol. Chem. 2010; 8: 3902

Crossref PubMed Search in Google Scholar
4 Gravel E, Ogier J, Arnauld T, Mackiewicz N, Ducongé F, Doris E. Chem. Eur. J. 2012; 18: 400

Crossref PubMed Search in Google Scholar
5 Wu J. J. Pers. Med. 2021; 11: 771

Crossref PubMed Search in Google Scholar
6 Mackiewicz N, Gravel E, Garofalakis A, Ogier J, John J, Dupont DM, Gombert K, Tavitian B, Doris E, Ducongé F. Small 2011; 7: 2786

Crossref PubMed Search in Google Scholar
7 Theodorou I, Anilkumar P, Lelandais B, Clarisse D, Doerflinger A, Gravel E, Ducongé F, Doris E. Chem. Commun. 2015; 51: 14937

Crossref PubMed Search in Google Scholar
8 Jamgotchian L, Devel L, Thai R, Poupel L, Huby T, Gautier E, Le Goff W, Lesnik P, Gravel E, Doris E. Nanoscale 2023; 15: 18864

Crossref PubMed Search in Google Scholar
9 Porret E, Hoang S, Denis C, Doris E, Hrubý M, Novell A, Gravel E, Truillet C. Nanoscale 2023; 15: 12574

Crossref PubMed Search in Google Scholar
10 Doerflinger A, Quang NN, Gravel E, Pinna G, Vandamme M, Ducongé F, Doris E. Chem. Commun. 2018; 54: 3613

Crossref PubMed Search in Google Scholar
11 Doerflinger A, Quang NN, Gravel E, Ducongé F, Doris E. Int. J. Pharm. 2019; 565: 59

Crossref PubMed Search in Google Scholar
12 Gravel E, Thézé B, Jacques I, Anilkumar P, Gombert K, Ducongé F, Doris E. Nanoscale 2013; 5: 1955

Crossref PubMed Search in Google Scholar
13 Costamagna F, Hillaireau H, Vergnaud J, Clarisse D, Jamgotchian L, Loreau O, Denis S, Gravel E, Doris E, Fattal E. Nanoscale 2020; 12: 245

Crossref PubMed Search in Google Scholar
14 Thielens NM, Belime A, Gravel E, Ancelet S, Caneiro C, Doris E, Ling WL. Int. J. Pharm. 2018; 537: 434

Crossref PubMed Search in Google Scholar
15 Hoang M.-D, Vandamme M, Kratassiouk G, Pinna G, Gravel E, Doris E. Nanoscale Adv. 2019; 1: 4331

Crossref PubMed Search in Google Scholar
16 Anilkumar P, Gravel E, Theodorou I, Gombert K, Thézé B, Ducongé F, Doris E. Adv. Funct. Mater. 2014; 24: 5246

Crossref Search in Google Scholar
17 Jamgotchian L, Vaillant S, Selingue E, Doerflinger A, Belime A, Vandamme M, Pinna G, Ling WL, Gravel E, Mériaux S, Doris E. Nanoscale 2021; 13: 2373

Crossref PubMed Search in Google Scholar

18a Alliot J, Theodorou I, Nguyen DV, Forier C, Ducongé F, Gravel E, Doris E. Nanoscale 2019; 11: 9756

Crossref PubMed Search in Google Scholar
18b Alliot J, Theodorou I, Ducongé F, Gravel E, Doris E. Chem. Commun. 2019; 55: 14968

Crossref PubMed Search in Google Scholar

19 Banerjee M, Panjikar PC, Bhutia ZT, Bhosle AA, Chatterjee A. Tetrahedron 2021; 88: 132142

Crossref Search in Google Scholar
20 Clarisse D, Prakash P, Geertsen V, Miserque F, Gravel E, Doris E. Green Chem. 2017; 19: 3112

Crossref Search in Google Scholar
21 Kumar RA, Jawale DV, Oheix E, Geertsen V, Gravel E, Doris E. Adv. Synth. Catal. 2020; 362: 4425

Crossref Search in Google Scholar

Corresponding Authors

Edmond Gravel

Université Paris-Saclay, CEA, INRAE, Département Médicaments et Technologies pour la Santé (DMTS), SCBM

91191 Gif-sur-Yvette

France

Email: edmond.gravel@cea.fr

Eric Doris