Nanoparticle forming polyelectrolyte complexes derived from well-defined block copolymers

Polymers can be used in nanoparticle associated formulations to encapsulate cytotoxic drugs (e.g., paclitaxel). Polyelectrolyte complexes (PECs) that form drug associated colloids also have potential to form particulate associated formulations. We used RAFT polymerisation to prepare small families of narrow molecular weight distributed (i) methacrylate block co-polymers comprised of oligomeric ethylene glycol, poly(ethylene glycol) methyl ether methacrylate (PEGMA), and dimethyl amino pendent chains, 2-(dimethylamino) ethyl methacrylate (DMAEMA), and (ii) poly(methacrylic acid), PMAA. These polymers were examined for their ability to form PECs capable of drug encapsulation. Optimal control in RAFT polymerisation was confirmed by the linear increase of molecular weight and the narrow dispersity of the polymers (<1.2) as determined by 1 H nuclear magnetic resonance and gel permeation chromatography. Dynamic light scattering and transmission electron microscopy showed formation of well-defined monodispersed nanoparticles with a hydrodynamic diameter of 25 ± 3 nm upon self-assembly of poly(PEGMA 0.23 -b-DMAEMA 0.77 ) 99 and PMAA 75 . These PECs are highly haemocompatible. Thin film hydration was used to encapsulate two hydrophobic drugs, paclitaxel and carmofur, into spherical nanoparticles. The results show that carmofur was encapsulated markedly more effectively than paclitaxel (72 vs 1.5%).


Introduction
Polyelectrolyte complexes (PECs) can be made by mixing two solutions of oppositely charged polymers (polyelectrolytes) that can interact to give nanoparticles with a narrow size distribution, which have potential for use in the formulation sciences [1]. Complexation of polyelectrolytes into PECs in aqueous media is rapid and spontaneous [2]. If one polyelectrolyte is used in excess the resulting PEC will display a net charge associated with the excess polyelectrolyte [3]. Of particular interest is the potential of PECs to encapsulate a drug within a colloidal particle that may display optimal charge and size to reduce systemic distribution of cytotoxic drugs, improve solubility and stability of poorly soluble actives, and avoid their rapid clearance from the blood compartment. Targeted drug delivery can also be achieved, for instance by using PECs with pH-tunable solubility. The PECs sense changes in the pH of their environment and respond by altering their structure, e.g., swelling or dissociation, and thereby freeing a drug cargo [4].
Homo-and block copolymers composed of hydrophilic or hydrophobic blocks and with at least one charged block can form PECs. For example, mixing phosphate buffered solutions of poly(ethylene glycol)-poly(L-lysine) block copolymer, PEG-b-P(Lys), and poly(ethylene glycol)-poly(α,β-aspartic acid) block copolymer, PEG-b-P(Asp) were found to give a monodispersed neutral spherical PEC [5].
The driving force for block copolymer derived PEC formation is due to the electrostatic interaction between the charged segments [6,7]. This results in phase separation and immediate rearrangement into stable PECs, for instance following mixing solutions of poly(acrylic acid) and poly(dimethylaminoethyl methacrylate)-co-poly(glyceryl methacrylate). The neutral block of the copolymer formed the shell of the PECs [8]. The difference in the solubility of each segment of the block copolymers can also influence the resulting PEC, causing polymer micelle formation with the hydrophilic non-charged blocks comprising the outer region of the colloid in aqueous media [9].
Here, we report the fabrication of PEC nanoparticles of less than 100 nm derived from a homopolymer and a block-copolymer, each with a narrow polydispersity index (PDI < 1.2). Both were generated by reversible addition-fragmentation chain-transfer (RAFT) polymerization, which allows preparation of a wide range of functional polymers of various architectures with narrow molecular weight distributions [10][11][12].
A hydrophilic block copolymer designed to have a positive charge at physiological pH was prepared with different block sizes from poly(ethylene glycol) methyl ether methacrylate 1 and dimethylaminoethyl methacrylate 4, giving poly(PEGMA-b-DMAEMA) 5 (Scheme 1a). A PEG-based block was incorporated into the copolymer as PEGylation is known to prolong circulation times and improve therapeutic efficacy by preventing opsonization and recognition by the immune system [13].
A complexing homopolymer, poly(methacrylic acid) 7 (PMAA) was also prepared (Scheme 1b), with a degree of polymerization (DP) to match the DMAEMA block. PMAA has a negative charge at physiological pH and should complex to the positively charged dimethyl amino block in poly(PEGMA-b-DMAEMA) 5 to form PECs designed for drug encapsulation. These polymers were used to fabricate PEC derived nanoparticles and to examine their capacity to encapsulate paclitaxel and carmofur.

Synthesis of poly(PEGMA)
For the synthesis of poly(PEGMA) with DP of 20, PEGMA (1.00 g, 3.30 mmol, 20 eq.), 4-cyano-(phenyl-carbonothioylthio) pentanoic acid (46.60 mg, 0.16 mmol, 1 eq.) and AIBN (2.74 mg, 0.016 mmol, 0.1 eq.) were dissolved in 1,4-dioxane (7 ml) in a 25 ml single neck round bottom flask. The mass ratio of CTA to initiator used in the synthesis of all the polymers was 10:1. Details are given in Table S1. The reaction mixture was sealed with a rubber septum and purged using argon for 30 min;   the flask was then heated at 70 ˚C for 17 h under magnetic stirring inside a fume cupboard with the sash in the down position. The reaction was stopped by exposing the solution to air via a needle and the polymer was precipitated twice under mild stirring using 70 ml of cooled hexane (10× the volume of dioxane). The precipitated sample was washed with acetone (6 ml) and dried using a rotatory evaporator under vacuum. The same procedure was repeated with 10, 30, 50, 60, and 70 equivalents of PEGMA.

Synthesis of poly(PEGMA-b-DMAEMA)
Poly(PEGMA) (0.20 g, 0.028 mmol, 1 eq.), 2-(dimethylamino) ethyl methacrylate (DMAEMA) (0.34 g, 2.14 mmol, 77 eq.) and AIBN (0.46 mg, 0.0028 mmol, 0.1 eq.) were dissolved in 1,4-dioxane (3.78 ml) in a 10 ml single neck round bottom flask. Details can be found in Table S2. The reaction mixture was sealed with a rubber septum and purged using argon for 30 min; the flask was then heated at 70 ˚C for 17 h under magnetic stirring. Experiments were performed in a fume hood with the sash down. The reaction was stopped by exposing the solution to air via a needle and the polymer was precipitated using 38 ml of cooled hexane (10× the volume of dioxane) under mild stirring. The precipitated sample was washed with acetone (4 ml) and dried using a rotatory evaporator under vacuum. The same procedure was repeated to prepare copolymers with total DP of 100 by copolymerising DMAEMA with the different poly(PEGMA) blocks. This was done by using the correct amount of macroRAFT agents (PEGMA homopolymers) and DMAEMA monomers.

Synthesis of PMAA
The synthesis of PMAA with DP of 85 is described as representative example. Methacrylic acid (1.00 g, 11.60 mmol, 85 eq.), 4-cyano-(phenyl-carbonothioylthio) pentanoic acid (38 mg, 0.14 mmol, 1 eq.) and AIBN (2.24 mg, 0.014 mmol, 0.1 eq.) were dissolved in methanol (7 ml) in a 25 ml single neck round bottom flask. Details can be found in Table S3. The reaction mixture was sealed with a rubber The obtained polymer was then washed with methanol (6 ml) and dried using rotatory evaporator and vacuum. The same procedure was repeated for synthesis of methacrylic acid polymers with degrees of polymerisation of 18, 28, 41, 50, 58, and 76.

Nuclear magnetic resonance (NMR)
Solution state 1 H NMR spectra were recorded in d 6 -DMSO using Bruker Avance 400 MHz NMR spectrometer, and were analysed using the Topspin software. The average length of polymer molecules and the average molecular weight of the polymers and copolymers were calculated by end-group analysis. Briefly, the proton signals for the end-group and the repeating units were identified, integrated, and calibrated. The degree of polymerisation was calculated by dividing the normalised integrated value by the number of protons in the repeating unit.

Nanoparticle preparation
Solutions of 1 mg ml -1 of poly(PEGMA-b-DMAEMA) and PMAA with similar chain lengths of the charged segments were prepared in water. Nanoparticles of 1:1 polycation to polyanion molar ratio were prepared by mixing solutions of the cationic copolymer and anionic homopolymer. For instance, 1 ml of a polycation solution (poly(PEGMA 0.23 -b-DMAEMA 0.77 ) 99 ) and 0.364 ml of a polyanion solution (PMAA 75 ) were mixed together to prepare PEC nanoparticles. The mixture was then filtered using 0.45 µm filters. The impact of different molar ratios (0.25, 0.5, 1, 2, and 4) of the cationic copolymer CP2 to the anionic homopolymer P2 (C/A) on the formation of PECs and their hydrodynamic size and zeta potential was also investigated.

Dynamic light scattering (DLS)
DLS size measurements and zeta-potentials were obtained using a Malvern Zetasizer Nano-ZS instrument. One ml of the nanoparticle solution was pipetted into a 1.6 ml disposable cuvette for size measurements, and transferred into a folded capillary cell for zeta potential measurements.

Transmission electron microscopy (TEM)
TEM images were recorded using a Philips/FEI CM120 Bio Twin transmission microscope.
Nanoparticle samples in water (1 mg ml -1 ) were directly dispensed onto TEM grids and left to dry, in some cases also with staining using an aqueous 3% phosphotungstic acid (PTA) solution.
pH Responsiveness of PEC nanoparticles pH responsivity of the PEC nanoparticles was investigated by observing the changes in their hydrodynamic size in different pH environments (7.4, 6.5, and 5.5) at 37 ˚C, using DLS.

Haemolysis assay
The haemolytic activity of the polymers and nanoparticles was assessed using freshly obtained blood from adult female Wistar rats. All animal handling was performed by licensed researchers. Red blood cells (RBC) were separated from the plasma by centrifugation and washed with Dulbecco's phosphate buffered saline (DPBS, Sigma Aldrich) three times. The RBC suspension was diluted with DPBS to a total volume of 50 ml.
The RBC suspension (180 μl) was then added to the wells of a clear 96-well plate (Corning). Polymer solutions and nanoparticle suspensions (20 μl) in DPBS at four different concentrations (1 mg ml -1 , 500 μg ml -1 , 100 μg ml -1 , and 20 μg ml -1 ) were added to the wells to give a final polymer or nanoparticle concentration of 100, 50, 10, and 2 μg ml -1 . For negative and positive control measurements, DPBS and a solution of Triton-X in DPBS (10% v/v) were used. The plates were incubated at 37 °C for 1 hour.
Next, the plates were centrifuged and 100 μl of the supernatant from each well was removed and deposited into a clean microplate. The procedure was carried out three times, and in each independent plate three wells were used for each concentration. Absorbance was read at 540 nm using a SpectraMax M2e microplate reader (Molecular Devices). The extent of haemolysis was calculated using the following equation: (1) (%) = × 100 Drug encapsulation A 0.2 mg ml -1 solution of drug in methanol was prepared. 10 ml of this stock solution was added to a vial containing 10 mg of the cationic copolymer or the anionic homopolymer to prepare 1 mg ml -1 solutions in separate vials. The appropriate volume of PMAA or PMAA/drug solution was added to 1 ml of the copolymer solution to give PECs with 1:1 molar ratio of polycation to polyanion. Methanol was evaporated using a rotatory evaporator until a thin film was formed in the vial followed by addition of water. The samples were then sonicated for 2 minutes and passed through a 0.45 µm filter. The filtered samples were centrifuged in Vivaspin 6 centrifugal concentrators with MWCO of 3000 at 10000 rpm for 15-20 minutes at room temperature. Following centrifugation, the supernatant was collected and the entrapment efficiency (EE%) and drug loading capacity (DLC%) of the concentrated nanoparticles were determined using an Agilent Cary 60 UV-Vis spectrophotometer and calculated using equations (2) and (3). UV absorbance of paclitaxel and carmofur was measured at 227 and 246 nm, respectively.

Polymer synthesis
RAFT polymerisation allows for the synthesis of polymers and copolymers with target DPs and narrow dispersity. Poly(PEGMA) 3 was first synthesised by RAFT polymerisation using 4-cyano-(phenylcarbonothioylthio) pentanoic acid 2 as the chain transfer agent (CTA). This thiocarbonylthio CTA 2 has previously been reported to be suitable for RAFT polymerisation of methacrylic monomers, providing good control over their polymerisation [14][15][16][17].      [18,19]. Linear block copolymers of DMAEMA and poly(2-hydroxyethyl methacrylate (PHEMA), an uncharged non-toxic hydrophilic polymer, synthesised via sequential RAFT polymerisation, were reported to provide better accessibility for charge-to-charge interaction with DNA compared to random copolymers [20]. Likewise, in this study, the aim was to synthesise block copolymers that allow charge-to-charge interaction and complexion with an oppositely charged polymer.
Samples of poly(methacrylic acid) (PMAA) with similar chain length as the DMAEMA blocks were prepared (Figure 3). High monomer conversions (84-90%; see Figure S2) and low dispersity (1.05-1.11) were obtained for the RAFT polymerisation of methacrylic acid. Comparable results have been reported previously in the literature [21][22][23]. The chain lengths of the polymers were estimated by comparing the integrals of the peak of CTA to that of the -CH 3 group of the polymers at δ H ∼ 0.94. A summary of the results is listed in Table 2.

PEC preparation
The association of poly(PEGMA-b-DMAEMA) block copolymers 5 (CP1-CP6) to PMAA homopolymers 7 (P1-P6) to fabricate PEC nanoparticles was examined. The PMAA samples was matched to have similar DP to that of DMAEMA block of the copolymers. The DLS derived size and zeta potential measurements of the six different nanoparticle combinations prepared by using 1:1 molar ratios of the copolymer and PMAA are listed in Table 3. The sizes and zeta potentials of the polymers alone in solution were also determined (see Table S4). These nanoparticles (NP2) had hydrodynamic diameter of 25 ± 3 nm and PDI of 0.08 ± 0.01, and ζ-potential of -23.2 ± 1.4 mV (Figure 4a). In contrast the mean hydrodynamic diameter of NP3, made from mixing CP2 and P6 with a 1:1 molar ratio, was 91 ± 19 nm with PDI of 0.38 ± 0.14 and ζ-potential of -5.4 ± 1.2 mV. Therefore, the DP of PMAA was selected to match the DP of the DMAEMA block as closely as possible. NP4, NP5, and NP6 were prepared using copolymers with longer PEGMA blocks and shorter polyelectrolyte blocks. These PECs were polydisperse, with PDI values above 0.29.
Use of PEG in low molar fraction in a polymer mixture has been reported to avoid steric hindrance between PEG strands, allowing the formation of stable nanoparticles [24]. Therefore, the smaller size NP2 nanoparticles formed in water could be due to the longer chain length of the ionic segments of the PEC nanoparticles and the lower weight fraction of PEGMA in the mixture compared to the other copolymers. A 1:1 molar mixture of CP2 and P2 was chosen as the optimal NP formulation given the PEC's small hydrodynamic size and PDI. TEM images of NP2 (Figure 4b) were broadly in agreement with the DLS data (Figure 4a; Table 3). According to DLS measurements, a filtered sample of 1 mg ml -1 copolymer CP2 in water  The molar ratio of polycation to polyanion can impact the size and zeta potential of PECs significantly [26].   The mean diameters of the polymers on their own are higher than those of the PECs at all five concentrations. This is likely due to the polymers being in coil form. There is also the possibility that polymer molecules could become entangled and form aggregates. At 25 °C, the mean diameter of the PEC nanoparticles remained the same at all polymer concentrations (Figure 6a) and no clear trend was seen in the zeta potential data (Figure 6b). An increase in the average diameter of the polymers was observed when the temperature was raised to 37 °C. An increase of the average end-to-end length per polymer coil is expected in good solvents as is the case here: water is a good solvent for all the examined polymers. The average diameter of the PECs, however, remained unchanged (Figure 6c). This is because the PECs are tightly bound due to the interactions between the oppositely charged polymers.
The PECs also had considerably lower dispersity compared to the polymers. Again, no clear trend was seen in the zeta potential values at 37 °C (Figure 6d).

pH responsiveness of PEC nanoparticles
The changes in hydrodynamic size of PEC nanoparticles in PBS at different pH values was monitored (Figure 7). PBS was used as the solvent to help maintain a constant pH during the experiments. PBS 99.8% at pH 7.4). Similar behaviour was observed in a PEG-based micellar drug delivery system by Xiao et al. [27]. The instability of the micelles at low pH was attributed to partial PEG chain shedding and formation of aggregates. These results suggest that the PECs are stable in blood at physiological conditions, whereas they become unstable in slightly acidic environments. This means that they could potentially be used for anti-cancer drug delivery, given the acidic nature of the tumour microenvironment.

Haemolytic activity
Prior to testing the encapsulation abilities of the PECs, their blood compatibility was assessed.
Haemolysis can lead to life threatening conditions such as anaemia and renal failure, therefore blood compatibility is a vital requirement for drug delivery systems developed for intravenous administration [28]. Red blood cells (RBCs) were exposed to different concentrations of polymers and PECs. RBCs were also incubated with PBS and Triton X as negative control and positive control, respectively.
Incubation of red blood cells with different concentrations of polymers and PECs did not result in any observable haemolysis (Figure 9). Triton X, on the other hand, caused 100% haemolysis. It can be concluded that the PECs are safe to be used for intravenous administration of drugs.   Figure 9. Percent haemolysis of RBCs incubated with the cationic copolymer, anionic homopolymer, and NP2 PECs (2-100 μg ml -1 ) for 1 hour at 37 °C. RBCs were also incubated with Triton X as a positive control and PBS as a negative control.

Drug encapsulation
To assess the ability of the optimal formulation (NP2) to encapsulate hydrophobic anticancer drugs, a hydrophobic dye was first used in a preliminary study. Phthalocyanine is an organic dye with a strong absorption in the far-red region (670 nm) and is often used as a photosensitizer in photodynamic therapy studies [29]. The PECs were able to encapsulate the hydrophobic dye with encapsulation efficiency (EE) of 71 ± 2 % and loading capacity of 11 ± 2 % when a 0.2 mg ml -1 phthalocyanine solution was used for the encapsulation process. The individual polymers however were unable to encapsulate the dye, resulting in the precipitation of the dye in the vials.
We next looked at loading two exemplar drugs into the PECs. The first, paclitaxel, is a poorly water-soluble anti-neoplastic drug that has long been used clinically and in recent years has been reformulated in macromolecule derived nanoparticles (e.g. albumin, PLGA-PEG) and are used for treatment of various types of cancer [30]. The second, carmofur, is a derivative of 5-fluorouracil that is clinically registered to treat breast and colorectal cancer. Carmofur is poorly water-soluble and is reported to cause leukoencephalopathy. We characterised the encapsulation of paclitaxel and carmofur in our poly(PEGMA-b-DMAEMA)/PMAA PECs.
Paclitaxel stock solutions (0.2 mg ml -1 ) were used to prepare paclitaxel loaded PECs with total polymer concentration of 1 mg ml -1 . An EE of 1.5% and drug loading capacity of 0.3% were recorded.
Increasing the concentration of paclitaxel resulted in lower EE% and LC%. However, increasing the total polymer concentration in solution from 1 to 2 mg ml -1 , while maintaining the 0.2 mg/ml concentration of paclitaxel, resulted in a slight increase in EE (to 2.0 %). Even though the solubility of paclitaxel in aqueous media was improved when encapsulated in the PECs, the results clearly indicate the inability of the nanoparticles to encapsulate paclitaxel.
In contrast, an EE of 72% and loading capacity of 12.6% were recorded when carmofur was incorporated into the PECs. It was possible to load more carmofur into the PEC nanoparticles, likely due to the carbonyl group and the fatty acid moiety of carmofur, which allows carmofur to be loaded into the core of the nanoparticles and also exist in the hydrophilic segment of the PECs. In contract, the hydrophobic paclitaxel can only reside in the hydrophobic core of the particles. Given the promising encapsulation efficiency, a preliminary study was performed to look at the release of carmofur. These assays showed a burst release (>80%) within the first hour at pH 7.4 and 5.5, indicating that the PECs are able to effectively disassemble and free their drug cargo in the tumor microenvironment.

Conclusions
In this study, RAFT polymerisation was used to synthesise polymers and copolymers with predetermined degrees of polymerisation. The synthesised polymers and copolymers have a narrow distribution of molecular weights with dispersity ranging from 1.05 to 1.20. This proves that polymers of similar chain length were synthesised following RAFT polymerisation. Polyelectrolyte complex nanoparticles with a narrow size distribution were formed upon mixing aqueous solutions of poly(PEGMA-b-DMAEMA) and PMAA. These were found to be non-hemolytic and to effectively encapsulate the anticancer drug carmofur. The PECs are also sensitive to temperature and pH, breaking up more rapidly at reduced pH values or increased temperatures.

Declaration of competing interest
The authors declare no conflict of interest.

Funding
This work was part-supported by the Engineering & Physical Sciences Research Council (grant EP/M014649/1).

Abbreviations
The following abbreviations are used in this manuscript:

Data availability
The raw data required to reproduce these findings are available from the authors on request. The processed data required to reproduce these findings are available from the authors on request.

Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
☐ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: