Physical insights into salicylic acid release from poly(anhydrides)

Salicylic acid (SA) based biodegradable polyanhydrides (PAHs) are of great interest for drug delivery in a variety of diseases and disorders owing to the multi-utility of SA. There is a need for designing SA-based PAHs for tunable drug release optimal for different needs. In this work, we have devised a simple strategy for tuning the release properties and erosion kinetics of a family of PAHs. PAHs incorporating SA were derived from related aliphatic diacids, varying only in the chain length and prepared by simple melt condensation polymerization. Upon hydrolysis induced erosion, the polymer degrades into cytocompatible products including the incorporated bioactive SA and the diacid. The degradation follows first order kinetics with the rate constant varying nearly 25 times between the PAH obtained with adipic acid and dodecanedioic acid. The release profiles have been tailored from 100 % to 50 % SA release in 7 days across the different PAHs. The release rate constants of these semi-crystalline, surface eroding PAHs decreased almost linearly with an increase in the diacid chain length. The release rate constants varied by nearly 40 times between the adipic acid and dodecanedioic acid PAH. The degradation products with SA concentration in a range of 30-350 ppm were used to assess cytocompatibility and showed no cytotoxicity to HeLa cells. This particular strategy is expected to: (a) enable synthesis of application specific PAHs with tunable erosion and release profiles; (b) encompass a large number of drugs that may be incorporated in the PAH matrix. Such a strategy can potentially be extended to the controlled release of other drugs that may be incorporated in the PAH backbone and has important implications for the rational design of drug eluting bioactive polymers.


Introduction
Poly(anhydrides) (PAHs) are a versatile class of polymeric biomaterials that have found extensive use in biomedical applications, particularly in the area of drug delivery. 1-3 These polymers are best suited for short term drug release. 4,5 Controlled release of several drugs has been successfully achieved by loading drugs in PAHs. 6,7 These polymers are typically hydrophobic but they possess hydrolytically labile anhydride linkages. The presence of these linkages qualifies these materials as prospective biodegradable materials. The major benefit of PAHs is that these are typically composed of diacid precursors. Because these precursors are often endogenous to the human body, PAHs do not elicit inflammatory responses when used as biomaterial implants. 8,9 A variety of PAH structures has been synthesized and evaluated to accomplish application specific drug release behavior. Apart from changing the nature/ structure of the polymeric backbone to aliphatic or aromatic, combinations of the two have also been tried. 10,11 PAHs like poly(maleic anhydride) have been synthesized such that anhydride linkages are present in the pendant side chains instead of the primary backbone. 12,13 Owing to the diverse requirements of stimuli-responsiveness 14 in drug release applications, it is important to understand the correlation between the structure of the polymer and its subsequent influence on physical properties such as degradation and drug release. 15 Aliphatic PAHs have created significant interest in the area of short term drug release particularly because of their simple diacid constitution. They are semi-crystalline in nature and typically have melting temperatures higher than the physiological temperature. 16 These are commonly synthesized by simple solvent-free melt polycondensation techniques and may,

Page 3 of 29 Physical Chemistry Chemical Physics
Physical Chemistry Chemical Physics Accepted Manuscript 3 therefore, be considered safe for biomedical applications. Since PAHs come under the class of insoluble, hydrophobic polymers, degradation or mass loss involves an inevitable step of erosion (as a consequence of polymer chain scission). For such hydrolytic chain scission to occur, the bonds present in the vicinity of the cleavable bonds also play a pivotal role. A more hydrophobic chemical structure present in the proximity of the anhydride bond will allow lesser water retention and, subsequently, slower degradation than a hydrophilic structure.
The influence of linker structure on release from PAH backbone has been studied earlier 17 . However, a systematic study of the degradation and release properties of a family of SA based PAHs has not been conducted. In the present work, we use a particular strategy of varying the chain length of the diacid to understand its influence on the degradation/ erosion kinetics and on the SA release behavior. This study thus provides a systematic comparison of SA-loaded polyanhydrides that offers new physical insights into changes that occur in physico-chemical, degradation and release properties resulting from the differences in chain length of the reactant.
We believe that these physical insights can facilitate the rational design of SA loaded polymers for drug delivery.
SA is a potent drug used for its antibacterial, 18,19 anti-inflammatory, 20

Synthesis
Step I: Synthesis of SA based diacids SA (1 mol. equiv. 7 g, 50.7 mmol) was dissolved completely in THF (10 mL). Pyridine (2 mol. equiv., 8.3 mL, 101.4 mmol) was added and allowed to stir in this solution for 15 min at 35 ˚C. The respective diacyl chlorides (0.5 mol. equiv.) were added slowly at 0 ˚C (ice bath) to the stirred solution. The ice bath was removed after addition and the reaction mixture was stirred for 12 to 16 h at 37 ˚C. After completion of the reaction, hydrochloric acid was added to the reaction mixture kept at 0 ˚C (ice bath) to precipitate the diacid. Acid was added until a pH of 2 was reached. The obtained precipitate (white) was washed repeatedly with water. The precipitate was then vacuum filtered and dried to constant weight before further reaction with anhydride in the next step. This step was based on a previously reported work. 23 Step III: Polymerization of acetylated diacid The anhydride precursors were polymerized by a simple melt condensation process. The

FTIR spectroscopy
The synthesized PAHs were chemically characterized by FTIR spectroscopy. Spectra were recorded on a Perkin-Elmer Frontier FT-NIR/MIR spectrometer under universal attenuated total reflectance (uATR-FTIR) mode. The recording range extended from 4000-650 cm −1 and the spectrum was an average of 32 scans with a resolution of 4 cm −1 .

1 H NMR spectroscopy
Proton-nuclear magnetic resonance (NMR) spectroscopy was also used for chemical characterization. The spectra were recorded on a 400 MHz Bruker Avance NMR spectrometer. 2 mg of the prepolymer was dissolved in 500 μL of CDCl 3 (for P4ASA and P6ASA) and d6-DMSO (for P8ASA and P10ASA) with 0.03 % (v/v) tetramethylsilane as internal reference (Deutero, Germany).

Molecular weight determination
The molecular-weight distribution (MWD) was obtained by gel permeation chromatography (GPC, Waters, Milford, USA). The setup was maintained at 50°C and consisted of an isocratic pump, three size-exclusion columns (Waters Styragel HR 4, HR 3, and HR 0.5 columns (300 7.5 mm)), a differential refractive index detector (Waters, R415) and data acquisition system. The eluent used in the system was THF and run at a flow rate of 0.9 mL/min.
were used. Samples were injected in a Rheodyne valve with a 100 μL sample loop. The chromatograph was converted to MWD using the universal calibration with polystyrene

Differential scanning calorimetry
Synthesized PAHs were thermally characterized on a differential scanning calorimeter (DSC, TA Instruments, Q2000). PAH samples weighing 5-7 mg were placed in a copper pan and crimped. The samples were then exposed to a uniform temperature program from 50 to 200 ˚C with a temperature ramp of 5 ˚C/ min.

Surface water wettability
The wettabilities of the samples were analyzed by measuring the water contact angle with a contact angle goniometer (Dataphysics). The static contact angles were evaluated by placing a 1 μL droplet of ultrapure water (Sartorius) on polymer discs (10 mm in diameter). The contact angle was measured after the drop equilibrated ~10 s after it was placed. The data is represented as mean standard deviation (S.D) based on three independent measurements.

Hydrolytic degradation studies
The samples were melted and made into discs of 5 mm diameter. These samples were then placed in 20 mL phosphate buffered saline (PBS) of pH 7.4. The samples were maintained at 37 ˚C in an incubator shaker maintained under shaking conditions (100 rpm). The samples are removed from PBS at specified time points and weighed after drying to constant weight in a vacuum desiccator. The mass loss of the polymer was measured using the following formula: (1)

Physical Chemistry Chemical Physics Accepted Manuscript
In equation (1), M o and M t indicate the initial and final weights of the PAH discs, respectively. All samples were made in triplicate and the mass loss data are represented as mean standard deviation.

SA release studies
SA release from the PAH matrices was evaluated to compare the release kinetics of the different polymers. Discs, as prepared for degradation studies were immersed in 20 mL PBS at 37 ˚C and 100 rpm and allowed to degrade over time. The release media was collected and the UV spectrum was recorded in a UV-vis spectrophotometer (Shimadzu, UV-1700 PharmaSpec).SA has characteristic absorption maxima at 297 nm. The SA concentration was determined by comparing with the standard calibration curve.

Cytocompatibility studies
HeLa, a cervical carcinoma cell line was used for assessing the cytocompatibility of the Cytocompatibility was estimated using an indirect method where cells were treated with conditioned media. Conditioned media were prepared by immersing 5 mm discs (thickness 1 mm) of P4ASA, P6ASA, P8ASA and P10ASA and allowed to undergo hydrolytic degradation for 24 h in 5 mL complete media (DMEM+FBS). HeLa cells were exposed to this preconditioned media and allowed to grow in its presence. Cell viability was monitored at specific significances were considered significant for p<0.05.

Anti-bacterial activity
To assess the pharmacological activity of the drug released from the PAHs, the effectiveness of the polymers were evaluated against S. aureus (ATCC 25923).Luria Bertani (LB) broth medium was used to culture S. aureus. Bacteria were allowed to grow overnight in an incubator shaker (100 rpm) and maintained at a temperature of 37 ˚C. The overnight culture

Physical Chemistry Chemical Physics Accepted Manuscript
(200 μL) was added to 5 mL of fresh LB medium and grown for 2 h. PAHs were dispersed in the LB medium (3 mg/mL) and the bacterial cultures were added to obtain a final optical density of 0.1 at 600 nm (OD 600 ). Bacterial growth was monitored by obtaining OD 600 at 6 h and 24 h after initial incubation. The readings from the replicates are reported as mean ± S.D. for n= 3.

RESULTS AND DISCUSSION
The synthesized PAHs have molecular weights ranging from 8300-49000 Da (as shown in Table 1) with a polydispersity index of 1.4. The molecular weights increased as the chain length of the diacid precursor increases.

FTIR spectroscopy
All the acids (AC, SC, SeC and DC) show the characteristic IR peak at 1690 cm -1 due to the -C=O stretching of the carboxylic chloride. After the synthesis of the diacid in the first step, a peak appears at the 1750 cm -1 indicating the formation of the ester bond at the -OH group of SA. The peak at 1690 cm -1 is still present indicating the presence of the free -COOH group. After the reaction with AAn and subsequent polymerization under vacuum, two peaks are observed in FTIR. The peaks at 1815 cm -1 and 1740 cm -1 indicate the formation of anhydride bond. The anhydride bond shows two characteristic -C=O stretching vibrations and is, therefore, characterized by two peaks (Figure 1). All peaks match previously reported literature. 25

NMR spectroscopy
NMR spectroscopy ( Figure S1, see electronic supplementary information) confirms the chemical characterization as observed in FTIR. All peaks in the 1 H NMR spectra match previously reported data in literature. 26 whereas P10ASA contains protons attached to α, β, γ, δ and ε carbons. This confirms the structure of the PAHs. However, the formation of the anhydride is confirmed by the 13 C NMR that confirms the presence of a peak at 170 ppm corresponding to the carbon of the anhydride bond. This is present for all the synthesized PAHs.

Differential scanning calorimetry
Thermal characterization ( Figure  P4ASA, P6ASA, P8ASA and P10ASA, respectively. Correspondingly, the T g ranges from 27 to 3 ˚C and scales with the diacid chain length. The increase in T g may be attributed to the higher degree of polymerization in the higher chain length PAH as compared to their shorter chain counterparts (as shown in Table 1). The longer chain restricts the rotational flexibility of the polymer chains, thereby, causing an increase in the T g . However, since all the synthesized PAHs have T g lower (and T m higher) than the physiological temperature (37 ˚C), this family of PAHs is suitable for biomedical applications. The ends of a polymer chain have more freedom of motion as compared to the inner segments. Low molecular weight polymers have more chain ends per unit volume and, therefore, higher free volume and lower T g . PAHs with higher loading concentration of SA (~15 %) were synthesized and it was observed that these polymers had T m values very close to 37 ˚C. This is undesirable as drug delivery is considerably affected if T m 37 ˚C. Therefore, all further synthesis and characterization were performed on PAHs with loading < 5 %.

Surface water wettability
The surface water wettability pertains to the hydrophobicity of the polymeric material.
The hydrophobicity increases in the following trend: P4ASA< P6ASA< P8ASA< P10ASA. The values of the water contact angles are tabulated in Table 1. The increase in the contact angle and subsequent decrease in the wettability may be attributed to the number of methylene (CH 2 ) units present in the parent diacid. Longer aliphatic chains are more hydrophobic than their shorter counterparts. The increased hydrophilicity may also be attributed to the higher loading % of SA in the shorter chain diacids than their longer chain counterparts. SA is hydrophilic and its higher inclusion in the matrix leads to increased hydrophilicity of the PAH.

Page 13 of 29 Physical Chemistry Chemical Physics
Physical Chemistry Chemical Physics Accepted Manuscript 13

Hydrolytic degradation studies
The hydrolytic degradation rate increases as the chain length decreases and follows: P4ASA>P6ASA>P8ASA>P10ASA. P4ASA, P6ASA, P8ASA and P10ASA degrade 100, 90, 50 and 28 % in 48 h, respectively (Figure 2a). The degradation data show that the PAHs degrade between 30-100 % in the first 48 h of hydrolysis. This fast degradation may limit their uses in biomedical applications. However, the need for antimicrobial, anti-inflammatory drugs is most often immediate after an implantation. The incorporated SA is reported to possess both of these features and may, therefore, be considered suitable for therapeutic applications.
The higher hydrophobic chains of the longer chain acids allow lesser water permeation than the shorter chain diacids. The solubility of adipic acid in water is considerably higher than that of the other diacids. 28 The hydrolytically labile anhydride bonds are located in the proximity of the aliphatic chains. As shown in the contact angle studies, the hydrophobicity of these chains increases with the diacid chain length. Owing to this, the water available at the vicinity of the anhydride bond decreases as the chain length increases. Lesser water results in lower hydrolysis of the anhydride bonds and, consequently, lower mass loss.
-Hydrolytic degradation causes chain scission in the backbone of the polymer due to bulk or surface erosion. 29,30 In most cases it occurs as a simultaneous bulk or surface phenomenon, where one mechanism is more predominant. The anhydride bonds are extremely labile to hydrolysis, and thus the degradation of the surface anhydrides takes place much faster than its penetration into the bulk of the matrix. 31 Since the mass loss due to degradation is accompanied by a concomitant decrease in the sample dimensions, it is likely that the erosion takes place layer  29 This is further corroborated by the kinetics of mass loss, as discussed below.

Kinetics of degradation
The degradation kinetics of these PAHs may be modelled using the following equation: In equation 2, k is the degradation rate constant and n is the order of degradation. First order release kinetics is followed in all four PAHs as is evident by fitting the degradation data to the following equation: factor driving the release of SA from these polymeric matrices is because these aliphatic PAHs are generally hydrophobic in nature. 29 Thus, there is a tendency of the hydrophilic SA to be released in the media.

Kinetics of release
The kinetics of release follow the first order degradation of the PAH matrices. The SA release may be modeled using the following equation, In as compared to P8ASA. Higher hydrophobicity of P8ASA compared to P4ASA (Table 1)

Cytocompatibility studies
The cell studies showed that cell proliferation was not affected by the degradation products of the PAHs. In spite of the decreased pH of the medium due to the fast degradation of  (Figure 4). There was no statistically significant difference observed across the samples.

Antibacterial studies
The antibacterial studies of the PAHs show that all four PAHs are active against S. aureus. This study also shows that the released SA is pharmacologically active. The polymers did not show any significant difference from the blank (untreated) after 6 h. However, after 24 h of growth in the presence of the PAHs, the PAHs show a significant drop in the OD 600 value indicating much lesser bacterial growth. All the PAHs showed similar antibacterial effect. In order to find significant differences between the polymers, paired t-test with p=0.05 was

Conclusions
We have successfully synthesized a family of PAHs for SA delivery. We have investigated in detail their degradation and drug release kinetics by varying the chain length of the precursor diacid. The chain length of the precursor plays an important role in controlling the rate of drug release from the polymeric system and may, therefore, be effectively used as a strategy to tailor drug release from aliphatic PAHs for application specific purposes. The released drug showed pharmacological activity against bacterial cells. Owing to their cytocompatibility, these SA-based polymers may find uses for biomedical applications.

Acknowledgements
The