Novelty in the development of biodegradable polymer coatings for biomedical devices: paclitaxel grafting on PDMMLA derivatives

ABSTRACT Biocompatible and biodegradable polymers are widely used in the medical field. In some cases, the biopolymer is accompanied by an active drug, which is delivered locally in a controlled manner in order to improve the healing conditions. Poly([R,S]-3,3-dimethylmalic acid) (PDMMLA) is a synthetic amphiphilic biodegradable polymer, which unlike PLA, can be chemically modified to adapt hydrophilic/hydrophobic balance, degradation kinetics, and physicochemical and biological properties. It may contain a lateral alkyl group or a functional group for coupling bioactive molecules to release during its degradation. In this work, we realized the chemical grafting of paclitaxel (PTX), a microtubule stabilizing anti-cancer agent on PDMMLA derivatives bio-polyesters following a Steglich esterification protocol. 1D and 2D NMR analyses validated the reaction with 10% (using 0.1 equivalent) of PTX on the copolymer PDMMLAH40-co-Hex60 (PDMMLA 40/60) and a maximal PTX grafting rate of 55% on the homopolymer PDMMLAH (PDMMLA 100/0). In vitro adhesion and cytotoxicity assays were carried out on HUVEC cells with PDMMLA 40/60, PDMMLA-PTX 30/10/60 and PLA.


Introduction
Biologically active materials are of great interest in medical fields for the treatment of various diseases. Bioresorbable polymers especially draw attention because of their specific properties. They can be used as suture threads, scaffolds for tissue regeneration and regrowth, or as vectors for local drug delivery. In this case, the polymer is used as a matrix loaded with drugs which is implanted into the area of interest, releasing the drug during the healing period and being degraded naturally at the same time [1,2]. PLA is one of the most used biopolymers in the development of drug-eluting devices, showing interesting biocompatible and bioresorbable properties. However, the slow degradation kinetics of PLA going up to 2 years is considered as a restrictive drawback. In the case of drug-eluting stents (DESs) for example, the durable presence of the polymer coating can lead to a late stent thrombosis [3][4][5].
Here, we study the chemical grafting of paclitaxel (PTX) on new bioresorbable polyesters, PDMMLAs, derived from poly(β-malic acid) (PMLA), which is also a polyester known for its biocompatible and biodegradable properties [6,7]. Previous studies have shown the promising properties of PDMMLA derivatives in terms of mechanical properties [8,9], degradation kinetics [10], and cell response [11]. The most interesting point about PDMMLA is the presence of carboxylic acid groups on its side chain, allowing control of the hydrophilicity levels of the polymer, which affects its degradation kinetics. It also makes possible the functionalization of these groups using potential molecules such as active drugs. It is supposed that when the drug is covalently bonded to the polymer, its release can be slower and controlled compared to the cases where the drug is loaded to the polymer only by physical interactions leading to a burst release [12].
PTX is an FDA approved anti-cancer agent, isolated for the first time from the Pacific yew tree, Taxus brevifolia. As a member of the taxane family, PTX acts through binding to β-tubulin and inhibiting the disassembly of microtubules [13]. This microtubule stabilization during the mitotic stage inhibits cell division, resulting in cell cycle blocking and cell death [14,15]. PTX is widely used for the treatment of ovarian, lung, prostate, breast, and solid tumor cancer. In recent years, it is also being used as an anti-proliferative agent in DESs, since it prevents the migration and the proliferation of smooth muscle cells (SMCs), limiting the neointima formation, and thus intra-stent restenosis [16].
The Steglich esterification protocol is followed in this work, in order to form an ester bond between the hydroxyl group of PTX and the carboxylic acid group of PDMMLA. In this protocol, the carboxylic acid is activated by N,N'dicyclohexylcarbodiimide (DCC), forming an O-acylisourea intermediate [17,18]. The esterification is improved by the addition of a catalytic amount of 4-(dimethylamino)pyridine (DMAP), leading to a more effective reaction [19,20]. According to the literature, by using DCC/DMAP, only the 2'-hydroxyl of the PTX is activated [21,22].
In order to obtain the desired amphiphilic bioactive polymer, the grafting of PTX was first realized on a PDMMLA copolymer containing 40% of carboxylic acid groups and 60% of hexylic ester groups, PDMMLAH 40 -co-Hex 60 (PDMMLA 40/60). Indeed, previous research studies have shown that the amphiphilic copolymer PDMMLAH 30 -co-Hex 70 (PDMMLA 30/70) is the most optimal surface for grafting the desired therapeutic agent. It presents better adhesion and proliferation of human endothelial cells [11], very favorable surface properties [8,9] and adequate degradation rate in physiological conditions [10].
Relying on these results, a new PDMMLA derivative was prepared with 40% of -COOH (PDMMLA 40/60), and 10% of PTX was grafted on these groups in order to recover the initially desired 30% carboxylic acid groups ( Figure 1).
In addition, the hydrophilic homopolymer PDMMLA 100/0, which contains only the repeating unit of -COOH groups on its side chain, is selected in order to determine the maximal grafting percentage of PTX The grafting is confirmed by 1D and 2D NMR analyses. Thereafter, thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) are utilized to study the thermal properties of the polymers.
It is also important to test the bioactive terpolymer PDMMLA-PTX 30/10/60 in vitro and to compare it with the corresponding copolymer PDMMLA 40/60 and the PLA which is the most studied. For this, cell adhesion tests of these three polymers on HUVEC endothelial cells as well as cytotoxicity tests will be carried out.

Materials
All commercially available chemicals were purchased from Sigma Aldrich (France) and employed as received. Reactions with anhydrous organic solvents were performed under nitrogen atmosphere. THF was distilled on sodium-benzophenone.
FTIR spectra were recorded on AVATAR 370 TF-IR Thermo Nicolet spectrometer using the Nicolet OMNI-Sampler ATR Smart Accessory (Ge, DTGS). Adsorption bands are given in cm −1 .
1 H and 13 C and HMBC NMR spectra were recorded on a Bruker AM-400 MHz spectrometer using deuterated chloroform (CDCl 3 ) or deuterated acetone (CD 3 COCD 3 ) as solvent, depending on the sample solubility. Chemical shifts (δ) are given in ppm. Multiplicity of the groups is given as follows: s (singlet); d (doublet); t (triplet); q (quadruplet); m (multiplet). The absolute molecular weights and weight distributions were determined at room temperature by coupling a high-performance size-exclusion chromatography (HPSEC), a multi-angle laser light scattering detector (MALLS), a viscosimeter, and a differential refractive index (dRI) detector. THF was used as the carrier phase and was filtered through a 0.1 µm filter unit (Millipore, Billerica, USA). It was degassed (DGU-20A3R Shimadzu, Kyoto Japan) and eluted at a 0.5 mL/min flow rate (LC10Ai Shimadzu, Kyoto Japan). 100 µL of a 0.2 µm-filtered sample solution (C = 10 mg/mL) were injected with an automatic injector (SIL-20A HT Shimadzu, Kyoto Japan). The column packing was a divinylbenzene gel. The MALLS photometer, a miniDawn TREOS from Wyatt Technology Inc. (Santa Barbara, CA, USA) was provided with a fused silica cell and a Ga-As laser (λ = 665.8 nm). The whole collected data: light scattering (LS), dRI were analyzed using the Astra v6.0.6 software package. Molar mass was obtained with a Zimm order 1 method. The concentration of each eluted fraction was determined with dRI (RID10A Shimadzu, Kyoto Japan) with a dn/dc value of 0.05. Thermogravimetric analysis (TGA) measurements were carried out on a TGA Q50 analyzer. The temperature range was set from 20°C to 500°C with a heating rate of 10°C/min under N 2 atmosphere. Differential scanning calorimetry (DSC) analyses were carried out on a DSC Q2000 analyzer. Polymers were put in the furnace and heated from −25°C to 200°C with a heating rate of 10°C/min (the final temperature can vary for different polymers regarding their TGA values). The heating cycle was repeated twice, and the T g value was collected from the inclination point on the second heating curve.

Methods
Polymer synthesis: PDMMLAs were prepared in anhydrous THF solution by ring-opening polymerization of racemic β-lactones following the previously reported protocol, using tetraethylammonium benzoate as initiator [23]. β-lactones were also synthesized following the previously described procedure with different functional groups to bring the hydrophilic and hydrophobic properties to the polymer [24].
Lactate dehydrogenase (LDH) colorimetric assay for cytotoxicity measurement: 25,000 HUVECs were cultured on different polymer films (PLA, PDMMLA 40/ 60 or PDMMLA-PTX 30/10/60) with complete culture media. After 24 h, 48 h, and 72 h of cell culture, the cytotoxicity was determined using colorimetric LDH cytotoxicity assay following the manufacturer's instructions (LDH, CyQUANT™, Invitrogen France) using spectrophotometer at 490 nm. The experiment was produced four times at different cell passage numbers (n = 4).
Adherence assay: 3000 HUVECs were plated on Labtek® (Thermo Fisher Scientific, Brébières, France) coated with fibronectin (control) or polymer films (PLA, PDMMLA 40/60 or PDMMLA-PTX 30/10/60) with complete culture media. After 15 minutes of incubation, the cell culture media was removed and the cells were washed twice with PBS. Nuclei were stained with DAPI (dilution 1/1000, Invitrogen, France) and the cytoskeleton (F-actin) was labelled with Alexa fluor 546 phalloidin (dilution 1/200, Invitrogen, France). A fluorescence microscope and Archimed® software were used, x40 magnification to analyze cell adhesion. The experiment was produced three times at different cell passage numbers (n = 3).
Statistical analysis: Results are expressed by mean ± SEM. Data were analyzed by ANOVA test p < 0.05 represent statistically significant difference.
The chemical structure of the polymers was confirmed by 1 H and 13 C NMR. The co-monomers relative content (40%) is obtained by 1 H NMR using the integration ratio of peak at 4.15 ppm corresponding to -CH 2 -O-hexyl and at 5.20 ppm corresponding to -CH 2 -O-benzyl. The final polymers with -COOH groups were obtained after a catalytic hydrogenolysis using palladium on charcoal (Scheme 1), confirmed by the disappearance of benzylic peaks at 5.2 and 7.4 ppm on the 1 H-NMR spectrum for both PDMMLA 40/60 and PDMMLA 100/0.

Covalent grafting
The approach was to synthesize a bioresorbable polymer with a covalently grafted active drug. The reaction was realized via the Steglich esterification  protocol, which forms a new ester bond between the 2'-OH group of PTX and the -COOH groups of the polymers (Scheme 2).
Final polymers obtained after PTX grafting on PDMMLA 40/60 and PDMMLA 100/0, are the terpolymer PDMMLAH 30 -ter-PTX 10 -ter-Hex 60 (PDMMLA-PTX 30/10/60) and the copolymer PDMMLAH 45 -co-PTX 55 (PDMMLA-PTX 45/55), respectively (Figure 3). The new terpolymer PDMMLA-PTX 30/10/60 is particularly important since its development aims biomedical applications. Figure 4 shows the 1 H and 13 C-NMR spectra of the terpolymer PDMMLA-PTX 30/10/60 with the assignments of the peaks, and Figure 5 shows the comparative 1 H-NMR spectra of PTX, PDMMLA 40/60 and PDMMLA-PTX 30/10/ 60. Chemical shifts and integrals show that the grafting by esterification has taken place involving the 2'-hydroxyl group of PTX and the carboxylic acid group of the copolymer. The peak at 3.55 ppm on the spectrum 'a' corresponding to the proton of the alcohol on C 2ʹ of PTX ( Figure 5 -a) disappears on the spectrum 'c' (Figure 5-c) (after grafting), indicating that this functional group has been transformed into an ester. Using the integral values of the peak at 5.34 ppm (integrating for 1.32, proton b on the polymer chain) and the peak at 6.20 ppm (integrating for 0.12, proton H 2' ), one can deduce the ratio of the PTX-ester groups and the remaining carboxylic acid groups on the side chain of the polymer. Calculations confirm that 10% of PTX has been grafted on PDMMLA 40/60. The desired proportion of PTX and the new terpolymer are thus successfully obtained: 10% of PTX, terpolymer PDMMLA-PTX 30/10/60 (Figure 3-a).
In addition to 1 H and 13 C spectra, a heteronuclear multiple-bond correlation (HMBC) of the PDMMLA-PTX 30/10/60 was recorded in order to demonstrate the formation of a new bond ( Figure 6). An HMBC spectrum  represents the correlations between two different nuclei (here, carbons, and protons) through several bonds [26]. In our case, the terpolymer shows a correlation between H 2' and C a , which are through three bonds ( 3 J) and a correlation between H b and C d ( 3 J). This method confirms with evidence the formation of the ester bond by demonstrating the correlation between the groups directly involved in this new bond, H 2' and C a ( 3 J).
Regarding the homopolymer, NMR signals (Figure 7) of the same functional moieties show that there is 45% of free carboxylic acid groups on the PTX-grafted PDMMLA, compared to the PDMMLA 100/0. This indicates that the maximal grafting percentage is limited to around 55%, which can be explained by the relatively large structure of PTX and its consequent steric hindrance. The obtained copolymer is thus the PDMMLAH 45 -co-PTX 55 or PDMMLA-PTX 45/55 (Figure 3-b).

Thermal properties
TGA and DSC analyses were performed in order to determine the thermal properties of the polymers. The TGA measurements provide information about the thermal stability and the degradation temperature T d of the polymers [27]. The DSC technique determines mainly the glass transition temperature T g and melting temperature T m of the polymer [28,29].    The DSC curves of the polymers are shown in Figure 9. Results show that the grafting of 10% of PTX on the copolymer PDMMLA 40/60 increases the T g from 23.7°C to 34.7°C. In contrast, in the case of the homopolymer PDMMLA-H, the T g undergoes a slight decrease from 69.9°C to 64.5°C [11]. On the other hand, a T g value of 34.7°C is obtained for the terpolymer PDMMLA-PTX 30/ 10/60. A T g below 37°C indicates that the polymer will be in its rubbery state at physiological temperature and consequently, will not undergo a change after implantation in the body.

Cell response
Adhesion: Cell adhesion assay was performed in vitro to study the attachment of HUVECs on different polymer layers (PLA, PDMMLA 40/60 or PDMMLA-PTX 30/10/60) and compared to fibronectin. Consequently, after 15 minutes of HUVEC incubation, we noticed an important attachment of HUVECs on PDMMLA-PTX 30/10/60 layer, which was similar to cells attached on fibronectin. In contrast, less of HUVECs were attached on PLA and PDMMLA 40/60 ( Figure 10).

Conclusion
The tailored synthesis of PDMMLA derivatives allows to obtain amphiphilic copolymers with hexylic ester (-COOHex) and carboxylic acid (-COOH) side groups. The presence of chemically modifiable -COOH groups makes possible a covalent grafting of molecules on PDMMLA, such as active drugs. The covalent grafting would improve the release profile and kinetics of the active drug during the healing period.
In this work, the chemical grafting of PTX, an FDA approved active drug, was carried out on two PDMMLA derivatives following the Steglich esterification protocol. The grafting was implemented on the copolymer PDMMLA 40/60 using 0.1 equivalent of PTX per repetition unit, allowing to obtain the terpolymer PDMMLA-PTX 30/10/60, which is the product of interest for future biomedical applications, such as cardiovascular stent coating. 1D and 2D NMR analyses confirm the reaction and the formation of a new bond between PTX and the copolymer.
Following the same protocol, the grafting was also carried out on the homopolymer PDMMLA-H, using 1 equivalent of PTX in order to determine the maximal grafting percentage. NMR analyses of the products show that the grafting of PTX on PDMMLA-H is limited to 55% due to its steric hindrance.
Thermogravimetric analyses of the polymers before and after PTX grafting gave high T d values for all polymers and a T g value below 37°C for the terpolymer PDMMLA-PTX 30/10/60 (34.7°C), which is suitable for biomedical applications.
For the cell adhesion test, we observed the more efficient binding of the HUVECs on the PDMMLA-PTX 30/10/60 layer. Cytotoxicity analyses for 3 days (24 h, 48 h, and 72 h) showed the lowest cytotoxicity for HUVECs cultured on PDMMLA-PTX 30/10/60. All these positive results lead to the conclusion that the PTX grafting reaction has made it possible to develop a new biodegradable and non-cytotoxic terpolymer to be used as a future coating for endovascular stents.