Reactive & Functional Polymers 81 (2014) 45–53

Contents lists available at ScienceDirect

Reactive & Functional Polymers

journal homepage: www.elsevier .com/ locate/ react

Poly(b,L-malic acid)/Doxorubicin ionic complex: A pH-dependentdelivery system

http://dx.doi.org/10.1016/j.reactfunctpolym.2014.04.0051381-5148/� 2014 Published by Elsevier B.V.

⇑ Corresponding author. Tel.: +34 934010913; fax: +34 934017150.E-mail address: montserrat.garcia@upc.edu (M. García-Alvarez).

Alberto Lanz-Landázuri, Antxon Martínez de Ilarduya, Montserrat García-Alvarez ⇑,Sebastián Muñoz-GuerraDepartament d’Enginyeria Química, Universitat Politècnica de Catalunya, ETSEIB, Diagonal 647, 08028 Barcelona, Spain

a r t i c l e i n f o a b s t r a c t

Article history:Received 5 March 2014Received in revised form 21 April 2014Accepted 28 April 2014Available online 9 May 2014

Keywords:Poly(malic acid)Ionic complexesDoxorubicin complexpH-dependentDrug-delivery particle

Poly(b,L-malic acid) (PMLA) was made to interact with the cationic anticancer drug Doxorubicin (DOX) inaqueous solution to form ionic complexes with different compositions and an efficiency near to 100%. ThePMLA/DOX complexes were characterized by spectroscopy, thermal analysis, and scanning electronmicroscopy. According to their composition, the PMLA/DOX complexes spontaneously self-assembledinto spherical micro or nanoparticles with negative surface charge. Hydrolytic degradation of PMLA/DOX complexes took place by cleavage of the main chain ester bond and simultaneous release of thedrug. In vitro drug release studies revealed that DOX delivery from the complexes was favored by acidicpH and high ionic strength.

� 2014 Published by Elsevier B.V.

1. Introduction

Biopolymer particles have found important applications as drugdelivery systems and constitute an ideal option for encapsulatingdrugs since their activity, solubility, cell permeability and stabilitycan be tuned by using polymers with appropriate chemical andphysical properties [1,2]. Poly(b,L-malic acid) (PMLA) is a water-soluble, biodegradable, bioabsorbable and non-immunogenic poly-ester [3] that can be produced by either chemical synthesis [4,5] orby fermentation of certain microorganisms [6]. The properties andfunctionality of PMLA are adjustable by chemical derivatization; infact PMLA has been esterified in different degrees to modulate theoverall hydrophobicity of the polymer [7–9] or to introduce bioac-tive ligands [10]. PMLA ionizes readily in water (pKa �3.5) givingrise to a highly soluble polycarboxylate [6]; under these conditionsit is able to be readily coupled with cationic compounds by stableionic interactions.

Doxorubicin (DOX) is an anthracycline antibiotic that has beenused for over 30 years as a potent chemotherapeutic antineoplasticagent to treat a wide spectrum of human cancers, especially breastcancer and lymphoma [11]. However, its therapeutic efficacy is lim-ited because DOX long-term clinical use is compromised by the tox-icity common to anthracycline drugs. Although encapsulation of

DOX in lipid micelles has been used to overcome such shortcomings[12,13], several essential attributes like drug release timing are dif-ficult to control. New alternatives, like pH dependent release, areemerging with the purpose of optimizing the therapeutic actionof DOX [14–16].

The release of the drug in the free form from delivery systems isa prerequisite for the activity of most of the antitumor activeagents. To attain a high antitumor activity along with a satisfactorycell-specifity, the characteristics of the polymer as well as the typeof linkage between drug and carrier have to be properly chosen[17]. Chemical conjugation of DOX to poly(aspartic acid) [18] hasbeen done but it was revealed that was actually the unconjugatedDOX in the micelle which exerted the therapeutic effect [19]. Con-jugation via acid-cleavable hydrazone bond has also been explored[16,20]. Other alternative investigated has been the physicalentrapment of DOX by nanoparticles [9,19] or micelles [21], whichoffers advantages such as easy preparation and low cost but alsosuffers disadvantages such as limited drug loading and difficultyin the drug release control.

Ionic interaction is an effective way to bind a drug to a polymer[22]. Such association is possible between polyelectrolytes andcharged drugs and it is termed as polymer/drug complexation[2]. The pH changes occurring within the body can address theresponse of the complex to a certain tissue or cellular compart-ment [23,24]. Carboxylic polymers, as it is the case of PMLA, aresuitable candidates for pH responsive systems. These polyelectro-lytes usually ionize in the 3–10 pH range to render polyanions able

46 A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53

to bind substantial amounts of cationic drugs owing to their highnegative charge density. DOX complexation with various poly-meric systems like poly(acrylic acid) [25], poly(c-glutamic acid)[2] and dextran sulfate [11] has been earlier reported. In this workwe studied the complexation between DOX and microbially pro-duced PMLA, as an easy and clean option for DOX encapsulationand pH-dependent release. To our knowledge it is the first timethat polymalic acid is used as polyelectrolyte for direct couplingwith a therapeutic drug. The well demonstrated capacity of PMLAto be bioassimilated along with its excellent hydrodegradabilityand biodegradability confer to this kind of complexes an excep-tional interest for building pH-dependent drug delivery systems.

2. Experimental

2.1. Materials

The PMLA used in this work was biotechnologically producedby aerobic cultivation of the fungi Physarum polycephalum, whichwas isolated and purified as described elsewhere [6]. Purity ofthe sample was ascertained by 300 MHz 1H NMR and it had aweight-averaged molecular weight of 34 kDa with a dispersity of1.08 as determined by GPC. Doxorubicin ((7S,9S)-7-[(2R,4S,5S,6S)-4-amino-5-hydroxy-6-methyloxan-2-yl]oxy-6,9,11-trihydroxy-9-(2-hydroxyacetyl)-4-methoxy-8,10-dihydro-7H-tetracene-5,12-dione) hydrochloride (DOX) was obtained from AKSci (Union City,CA, USA).

2.2. PMLA/Doxorubicin ionic complexes synthesis

Ionic complexes of PMLA with DOX with different molar ratioswere obtained by a simple and clean method. Briefly, to 1 mmol(repeating units) of PMLA dissolved in 4 mL of milli-Q water, pH�6.0 (final pH �4), 0.1, 0.25 or 0.5 mmol of DOX dissolved in2 mL of the same solvent were added dropwise under magneticstirring, and left to react overnight in the dark to prevent DOX pho-todegradation. Non-attached DOX was removed by extensive dial-ysis against deionized water for 48 h using a cellulose membraneof 8 kDa molecular weight cut-off. The PMLA/DOX ionic complexesprecipitated from solution and were lyophilized for recovery andstorage. The complexation degree attained was determined by 1HNMR.

2.3. Particles characterization

The particles formed upon complex formation were examinedby scanning electron microscopy (SEM). SEM images were takenwith a field-emission JEOL-JSM-7001F instrument (JEOL, Japan)from platinum/palladium coated samples. The hydrodynamicdiameter and the f potential (surface charge) of the particles weredetermined by dynamic light scattering. Particle size measure-ments were performed with a ZetaSizer NS (Malvern Instruments,UK) from particles suspended in deionized water.

2.4. Hydrolytic degradation and drug release studies

The hydrolytic degradation of PMLA/DOX ionic complexes wasstudied in citrate buffer pH 5.0 or PBS pH 7.4 at 37 �C by analyzingthe soluble degraded compounds by 1H NMR. For this, about 10 mgof sample were immersed in the buffer solution, placed in NMRtubes and spectra recorded from the supernatant at scheduledtimes.

In vitro DOX release was evaluated by the dialysis method as afunction of pH and ionic strength. Briefly, 10 mg of freeze-driedPMLA/DOX ionic complex particles were resuspended in 1 mL of

buffer, either at pH 5.0 and 150 mM NaCl ionic strength or at pH7.4 and 75, 150 or 300 mM NaCl ionic strength, and transferredinto a dialysis tube with 8 kDa molecular weight cut-off. The tubewas then immersed into 20 mL of the corresponding buffer and leftunder mild stirring at 37 �C. 2 mL aliquots of the releasing mediumwere taken at scheduled times and the drawn volume replaced byfresh buffer.

Drug concentration was determined by UV spectroscopy at480 nm and cumulative drug release was calculated as a functionof time.

2.5. Measurements

NMR spectra of complexes were recorded on a Bruker AMX-300instrument from samples dissolved in methanol-d6 containingminor amounts of trifluoroacetic acid. In order to detect any spe-cific interaction, spectra of malic acid/DOX mixtures with differentmolar ratios, dissolved in D2O at 0.025 mM concentrations wererecorded. Degradation products were analyzed in the deuteratedaqueous buffer using for incubation. 1H NMR spectra wererecorded at 25 �C operating at 300.1 MHz; 128 scans with 32 Kdata points and relaxation delays of 2 s were acquired. FT-IR spec-tra were registered in a Perkin–Elmer Frontier FT-IR spectrometerprovided with a universal ATR sampling accessory. UV measure-ments were performed using a UV–visible spectrophotometerCECIL CE 2021.

Gel permeation chromatography (GPC) was done with a Waters515 HPLC pump provided with a Waters 410 differential refrac-tometer as detector and a Waters Styragel HR 5E column(7.8 � 300 mm) (Waters, Massachusetts, USA). Samples were chro-matographed using 0.05 M sodium trifluoroacetate in hexafluoro-isopropanol at 0.5 mL min�1 flow rate. Chromatograms werecalibrated against poly(methyl methacrylate) standards (Varian,USA). DSC traces were recorded using a Perkin–Elmer Pyris 1 appa-ratus under a nitrogen flow from 2 to 4 mg samples at heating andcooling rates of 10 �C min�1. Indium and zinc were used as stan-dards for calibration. TGA measurements were performed with aPerkin–Elmer TGA6 thermobalance under a nitrogen flow from10 to 15 mg samples at a heating rate of 10 �C min�1.

3. Results and discussion

3.1. Synthesis and characterization

Poly(b,L-malic acid) (PMLA) ionizes readily in water giving riseto a highly soluble polyanion. The pKa of biologically producedPMLA has been determined to take values within the 3.4–3.6 rangeso it is extensively charged under physiological conditions [6]. Onthe other hand, DOX is a positively charged amphoteric drug con-taining one protonable amino group in the sugar moiety with apKa = 8.6 and two deprotonable phenolic groups in the aglyconepart of the molecule; thus an equilibrium exists between the pos-itively charged, negatively charged and neutral species of DOXdepending on pH. In the 0–6 pH range, the amino group in DOXis protonated as NH3

+ which makes possible its electrostatic bindingto negatively charged PMLA [2]; in fact, non-stoichiometric ioniccomplexes with polymer/drug molar ratios of 10:1, 4:1 and 2:1were successfully obtained by precipitation upon mixing the twocomponents in an aqueous medium that was initially set at pH4.0 (Scheme 1). Complexes precipitated from the solution displac-ing the equilibrium towards their continuous formation.

The complexes were formed with high efficiency for whichevercomposition with more than 90% of the added drug becoming ion-ically coupled to PMLA (Table 1). Such loading efficiency was up tothree times higher than that attained in the physical entrapment ofDOX in esterified PMLA derivatives using the emulsion-evapora-

O

OOMe

OH

OH O

OH

OHO

O Me

O

OOHO

O

OO-O

H3N+

OH

x y

Scheme 1. Non-stoichiometric ionic complex PMLA/DOX.

A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53 47

tion [9] or precipitation-dialysis methods [19]. The partial chargeneutralization taking place upon complexation resulted in theinstantaneous precipitation of a reddish product in the form of par-ticles. For PMLA/DOX-(10:1) and PMLA/DOX-(4:1), particlesremained in a colloidal suspension, while those made of PMLA/DOX-(2:1) tend to aggregate and settle down onto the bottom ofthe container.

Since aqueous PMLA/DOX mixtures evolve with immediate pre-cipitation of the complex, the occurrence of interactions betweenthe two components in solution is difficult to investigate. Toapproximate the question 1H NMR spectra were taken from aque-ous solutions of malic acid/DOX mixtures over a wide range ofcomposition ratios, and compared to the spectra of the pure com-ponents. No differences was detected for any composition, whichcan be taken as indicative of complete absence of specific interac-tion between MLA and DOX in aqueous solution (1H NMR spectrashown in the Supplementary material file linked to this paper,Fig. S1).

Conversely the formation of the ionic PMLA/DOX complex inthe solid state could be ascertained by infrared spectroscopy. TheFT-IR profiles registered from the 2:1 complex and from its compo-nents, both separately and physically mixed are compared in Fig. 1,and characteristic absorption bands are listed in Table 2. Mainchanges detected in the spectrum of the ionic complex comparedto that registered from the physical mixture or PMLA and DOXare the disappearance of the 3528 and 1521 cm�1 bands, as wellas the decrease in intensity of the 869 and 802 cm�1 bands; allthese bands are associated to the DOX NH bond vibrations (stretch-ing, bending or wagging), and their absence indicates that the DOXamine group must be directly involved in the association of thedrug with the polyacid, most probably through ionic interaction.Similar changes in the DOX infrared spectrum were reported inthe work of Kayal and Ramanujan [26], in which they demon-strated the attachment of DOX to PVA coated iron oxide nanopar-ticles via amine-hydroxyl interactions. In order to support ourinterpretation the solid product resulting from evaporating anaqueous solution containing equimolar amounts of malic acidand DOX was also examined by FTIR. The spectrum from thecompound showed also a large decreasing of the 1521 cm�1 DOXband according to what should be expected for the interaction of

Table 1Coupling reaction characterization.

Compound Feed ratio mol:mol Complex compositio

PMLA – –PMLA/DOX-(10:1) 1:0.10 1:0.09PMLA/DOX-(4:1) 1:0.25 1:0.23PMLA/DOX-(2:1) 1:0.50 1:0.46

the amine group of DOX with the carboxylic group of malic acid(Fig. S2).

The amount of drug bound to the polymer was quantified byNMR spectroscopy. The 1H NMR spectra of complexes and theircomponents are compared in Fig. 2. Calculations for drug contentwere done using the signal at 1.55 ppm arising from the methylgroup attached to the pyrano moiety in DOX and the signal locatedat 5.6–5.9 ppm, which arises from the PMLA main chain CHa andone CH of DOX. The PMLA/DOX molar ratios in the complexes asdetermined by this method are listed in Table 1, which were foundto be very close to the feed.

3.2. Thermal characterization

A DSC analysis of the complexes was carried out in order toappraise their thermal behavior. The DSC traces of PMLA, DOX,their physical mixture with a 2:1 ratio, and the ionic PMLA/DOX-(2:1) complex are compared in Fig. 3. The DSC trace of PMLAdisplayed a wide melting peak at 210–215 �C, indicative of thesemicrystalline nature of this polymer, and the trace of DOX pre-sented a well-defined melting peak at 231 �C characteristic ofhighly crystalline material. Conversely the trace registered fromthe PMLA + DOX physical mixture is essentially similar to that ofPMLA with the melting peak broadened and displaced downwardsdue to the presence of DOX, which presumably was dispersed inmolten PMLA. On the contrary, the heat exchange detectable inthe trace of the PMLA/DOX-(2:1) complex was almost impercepti-ble indicating that crystallinity of PMLA was largely suppresseddue to the ionic interaction taking place between the drug andthe polymer. Furthermore, the total absence of the characteristicDOX melting peak on the complex may be taken as demonstrativeof that no pure DOX precipitated separately during complexation.

The thermal stability of DOX/PMLA complexes was evaluated bythermogravimetry. The TGA traces of the complexes and theircomponents are compared in Fig. 4. The initial weight loss of about5% seen on the TGA traces of samples containing PMLA is attrib-uted to their moisture content. More than 90% of weight loss ofPMLA occurred over the 200–250 �C range, which is consistentwith the degradation signal that is observed by DSC overlappingthe melting of the polymer. On the other hand, DOX degradationwas found to happen along three steps that started at 237, 320and 400 �C, respectively, and that entailed a total weight loss of50% of the initial mass. The complexes also presented a degrada-tion in three steps but at lower onset temperatures and entailingsmaller weight losses and that are depending on composition.Remaining weights after heating at 800 �C were 20, 27 and 38%,for 10:1, 4:1 and 2:1 PMLA/DOX complexes respectively. In thethree cases thee second degradation step happened about 325 �Cwith a weight loss corresponding to the DOX degradation secondstep.

3.3. Particle formation and characterization

Particles formed during drug–polymer complexation wereexamined by scanning electron microscopy. Since the ionic com-plexes are amphiphilic they are expected to self-assemble in water

n (mol:mol) Efficiency (%) Yield (%) Mw (g mol�1)

– – 32,00099 85.7 29,30092 83.3 29,80092 80.5 29,600

Fig. 1. FTIR spectra of the PMLA/DOX-(2:1) complex, the PMLA + DOX (2:1) physical mixture, DOX, and PMLA.

Table 2FT-IR of PMLA and DOX separately, their physical mixture (2:1), and their ioniccomplex (2:1).

Absorption bands (cm�1) Assignment

PMLA DOX PMLA + DOX PMLA/DOX

3528 3528 m(OAH) free3316 3316 m(HAO) bonded3160–2300 3160–2300 m(NH3

+)2897 2897 m(CAH)

1730 1730 1730 1730 m(C@O)a

1615 1615 1615 m(C@O)b

1580 1580 1580 m(C@C)1521 1521 d(NAH)1413 1413 1413 d(CH)12,83,989 12,83,989 12,83,989 m(CAOAC)

1157 1157 m(OACAC)1071 1071 1071 m(CAO)

1044 m(CAO), m(CaACb)869/802 869/802 869 x(NAH)

a Band from ketone in DOX.b Bands from quinone carbonil associated by intramolecular hydrogen bonds.

48 A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53

to form spherical particles, presumably with the hydrophobicDOX-coupled PMLA units integrating the inner particle core andthe remaining ionized polymer segments preferentially locatednear to the surface. In fact, individually dispersed microspheresfor 10:1 and 4:1 complexes, and aggregated nanospheres for 2:1complex were observed (Fig. 5). The hydrodynamic average diam-eter of microparticles determined by LS was around 1.5 lm whilein nanospheres it was ten times smaller, i.e. 150 nm approximately.

The result is comparable to that reported for the chemical conjuga-tion of DOX with poly(a-aspartic acid) which changed the initialhydrophilic units of the polyacid into hydrophobic blocks by ioniccoupling with the drug. The coupled DOX will self-assembly toform a hydrophobic micelle core [18]. A somewhat simple modelillustrating the possible structure of the particles is depicted inFig. 6. The role of the ionic interactions as the main cohesive forcesoperating in the particles was evidenced by appraising the effectthat the ionic strength exerts on their stability. In fact, decomposi-tion of the complex with subsequent liberation of DOX to the med-ium took place in particles suspended in NaCl aqueous solution assoon as the salt concentration came up around 0.1 M (Fig. 7).

The surface charge of the particles, reflected by their f potential,had approximately the same negative value for the three com-plexes confirming the outer location of the non-coupled carboxylicgroups (Table 3). These values are significantly lower than thatreported for neat PMLA [27]. It was surprised to see an increasein the negative zeta-potentials after loading DOX, a certainly strik-ing result for which we have not explanation. It should be saidhowever that a similar result has been reported by Manocha andMargaritis in the study of DOX-poly(c-glutamic acid) [2]. Althoughpositively charged particles are preferred for cell uptake, it isknown that negative particles are also able to undergo endocytosisby adsorption at the positively charge cell sites [28]. In fact a com-bination of factors including size, shape and surface chemistry isactually governing cellular uptake. Regarding the zeta potential,it is accepted that particles may be internalized if values of f arewithin the range of tens of �mV and +mV [29].

1.21.62.02.42.83.23.64.04.44.85.25.66.0(ppm)

Fig. 2. 1H NMR spectra of PMLA, DOX and PMLA/DOX ionic complexes.

120 130 140 150 160 170 180 190 200 210 220 230 2400

20

40PMLA/DOX-(2:1)

PMLA+DOX

PMLA

Hea

t Flo

w U

p En

do (m

W)

Temperature (°C)

DOX

Fig. 3. DSC traces of the PMLA/DOX-(2:1) complex, the PMLA + DOX (2:1) physicalmixture, DOX, and PMLA.

0 100 200 300 400 500 600 700 8000

20

40

60

80

100

Wei

ght (

%)

Temperature (°C)

Dox PMLA/DOX (2:1) PMLA+DOX (2:1) PMLA

Fig. 4. TGA traces of the PMLA/DOX-(2:1) complex, the PMLA + DOX (2:1) physicalmixture, DOX, and PMLA.

A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53 49

3.4. Hydrolytic degradation mechanism

The hydrolytic degradation of the complexes was followed by1H NMR analysis of the products released to the incubating med-ium (Fig. 8). The supernatant of PMLA/DOX-(10:1) incubated atpH 7.4 and 37 �C showed signals arising from the methylenegroups of PMLA after 3 days of incubation which are attributedto the presence of partially solubilized large size oligomers. Thepresence of short oligomers was later evidenced by the appearance

of a signal at 2.90 ppm arising from the b-CH2 of the PMLA end-chain unit. The intensity of this signal increased continuously alongthe two first weeks of incubation to be finally replaced by the2.70 ppm methylene signal from free malic acid. Weak signalscharacteristic of DOX were detected from the third day of incuba-tion (Fig. 8a) but their intensity remained essentially unchangeddue to the instability of this compound in the incubating medium.In fact it has been reported that DOX in aged aqueous solutiontends to form clusters with water included between the stacked

Fig. 5. SEM micrographs of PMLA/DOX complex particles. (a) PMLA/DOX-(10:1), (b) PMLA/DOX-(4:1) and (c) PMLA/DOX-(2:1).

Fig. 6. Schematic model of the interaction mechanisms operating in the PMLA/DOX particle formation and drug loading and release.

0.0 0.1 0.2 0.3 0.4 0.50.5

0.6

0.7

0.8

0.9

1.0

A 480PM

LA-D

OX/

A 480D

OX

NaCl [M]

PMLA/DOX (10:1) PMLA/DOX (4:1) PMLA/DOX (2:1)

Fig. 7. Normalized absorbance of DOX released from PMLA/DOX complexes as afunction of NaCl concentration.

50 A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53

aromatic sheets which are excluded from solution in form of gel[30,31]. We have experimentally checked this behavior by NMRand results have been included in the SM file as Fig. S3. The NMRanalysis of PMLA/DOX-(2:1) incubated under similar conditionsafforded the same signal changing pattern but with lower peakintensity and a delay of one week in the appearance of new signals,in agreement with what should be expected from its lower contentin non-complexed PMLA residues.

The 1H NMR spectra recorded from the incubation medium atpH 5.0 were more difficult to analyze due to the fact that the sig-nals of the citric acid used for pH buffering overlapped with themethylene signals of the free malic acid and terminal malate units.Nevertheless, a close inspection of the spectra shown in Fig. 8brevealed that they followed an evolution pattern similar to thatobserved at pH 7.4 but at a higher rate. Degradation of PMLA/DOX-(10:1) at pH 5.0 clearly showed signals from terminal b-CH2

of the PMLA chain after 3 days of incubation, and at difference withwhat happened at pH 7.4, signals arising from the CH end groupswere also detected at this time (Fig. 8b). 1H NMR spectra recordedat different pH for PMLA/DOX-(2:1) showed slighter differencesthan for PMLA/DOX-(10:1). Although less soluble products and a

Table 3PMLA/DOX ionic complex particles characterization.

Particles State Sizea (nm) Pd.I.b f Potential (mV)

PMLA Solution �22.9 ± 1.7PMLA/DOX-(10:1) Micro Suspension 1613 0.544 �26.0 ± 5.1PMLA/DOX-(4:1) Micro Suspension 1496 0.447 �28.0 ± 4.8PMLA/DOX-(2:1) Nano Aggregate/Precipitation 150 0.114 �26.9 ± 4.0

a Hydrodynamical average diameter.b Polydispersity index of particle size.

Fig. 8. 1H NMR spectra of the degradation media of ionic complexes PMLA/DOX-(10:1) and PMLA/DOX-(2:1) at different pH: (a) 7.4 and (b) 5.0. D = DOX, �solvent traces.

A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53 51

lower rate of disappearance of the main chain signals wereobserved in this case, DOX signals started to be observed for bothcomplexes at the third day of incubation.

3.5. In vitro drug release

It is known that pH and ionic strength play a key role in the lib-eration of drugs that are immobilized to a polymer matrix by ioniccoupling interactions. Drug release from PMLA/DOX complexeswas assessed as a function of releasing media conditions concern-ing pH and ionic strength. The physiological pH 5.0 and 7.4 were

assayed; pH 5 emulates inside mature lysosome environmentwhereas pH 7.4 imitates plasma conditions. The cumulative DOXreleasing profiles observed for the three studied complexes as afunction of time are comparatively plotted in Fig. 9a for the twoassayed pH values. DOX release appeared to be markedly pHdependent since at pH 5 the delivery rate was four times fasterthan at pH 7.4 regardless complex composition. In fact, the cumu-lative amount of DOX released after 25 days of incubation at pH 5was 35–40% of the initial load whereas at pH 7.4 it hardly reached15%. Furthermore, the cumulative release at pH 7.4 tends to a sta-bilized on a plateau while at pH 5 the release rate was almost con-

0.0

0.1

0.2

0.3

0.4

0.5 pH 5 - (10:1) pH 5 - (4:1) pH 5 - (2:1) pH 7.4 - (10:1) pH 7.4 - (4:1) pH 7.4 - (2:1)

Cum

ulat

ive

rele

ase

(nor

mal

ized

)

Days

a

0 10 20 30

0 10 20 300.0

0.1

0.2

75 mM - (10:1)150 mM - (10:1)300 mM - (10:1)

75 mM - (4:1)150 mM - (4:1)300 mM - (4:1)

75 mM - (2:1)150 mM - (2:1)300 mM - (2:1)

Cum

ulat

ive

rele

ase

(nor

mal

ized

)

Days

b

Fig. 9. Doxorubicin release from PMLA/DOX ionic complexes as a function of pH (a)and ionic strength at pH 7.4 (b).

52 A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53

stant along the whole incubation time. A moderate burst is noticedat either pH and at any ionic strength. The small fraction of DOXinitially delivered at high rate would be that located nearly the sur-face. The release of inner located drug will be somewhat dependingon polymer degradation and its delivery rate is expected to decay.Both H-bonding PMLA/DOX interactions and p–p stacking DOX/DOX interactions would be responsible for the further delayed oreven incomplete drug release observed in these complexes.

The pH dependence observed for the PMLA/DOX complexes iscertainly a relevant result since drug release during blood trans-port would be prevented whereas drug discharge would take placeat the target compartment where the pH is expected to be around5.0. Tumor tissue has an extracellular pH of 6.5–7.2, slightly lowerthan the 7.4 value present in normal tissues, whereas inside thelysosomes, where the microparticles would access via phagocyto-sis, the pH is 4.5–5.0. Furthermore the hydrolytic enzymes thereinpresent could enhance polymer degradation and therefore drugrelease. Since residence times of a few hours should be expectedfor a carrier in the blood stream, DOX losses from microparticlesmade of PMLA/DOX complexes are expected to be negligible dur-ing transport. This is a highly appreciated advantage for cytotoxic

Table 4Comparative of DOX release by different PMLA-based delivery systems.

Method of encapsulation DOX content (

PMLA/DOX (10:1) Simple mix – precipitation 29.7 ± 2.4PMLA/DOX (4:1) Simple mix – precipitation 107.6 ± 1.9PMLA/DOX (2:1) Simple mix – precipitation 234.0 ± 3.1PGGA Simple mix – precipitation 99.0 ± 0.06DS/DOX Simple mix – precipitation 60.0PAALM-1 Emulsion – solvent evaporation 4.2 ± 0.6PAALM-L60 Precipitation – dialysis 8.5 ± 1.2

drugs like DOX, although only an efficient delivery at the site ofaction will allow to taking full benefit [2].

Regarding compositions it was observed that release rate differ-ences between complexes at pH 5 were small, whereas they weresignificant at pH 7.4. Apparently the DOX release rate is dependenton the complex composition provided that incubation is performedunder neutral conditions. Probably such a different behavior isrelated to the effect of pH on degradation rate of PMLA. Also theeffect of the ionic strength on DOX release was studied at pH 7.4for I values of 75, 150 and 300 mM (Fig. 9b). A clearly enhancingeffect of the ionic strength on the DOX release rate was observedfor the three complexes, although in not so much extent as forpH changes. The maximum release rate observed at pH 7.4 wasfor the PMLA/DOX (10:1) complex at 300 mM ionic strength; thedrug accumulated in the incubation medium after 26 days of resi-dence was 17% of the initial loaded amount. The important conclu-sion drawn from these results is that the drug release rateincreasing with ionic strength supports the concept of a drug deliv-ery mechanism mainly determined by the dissociation of the ioniccomplex, which is also in agreement with is shown in Fig. 7.

The release kinetics of DOX physically entrapped in nanoparti-cles made of PMLA derivatives has been recently studied [9,15].DOX release rates at pH 7.4 from PMLA/DOX complexes measuredin this work are proven to be up to three times slower than therelease reported for physical entrapped DOX (Table 4). It shouldbe remarked that the drug release behavior observed in thesePMLA nanoparticles is quite different to that reported for poly(-methyl malate) nanoparticles loaded with DOX [9]. In the esterifiedpolymalic acid ionic interactions are not possible and only hydro-phobic] interactions may be invoked to explain the retention ofthe drug by the polymer. On the other hand, in the work of Mano-cha and Margaritis on the ionic complex made of poly(c-glutamicacid) and DOX [2], the release of the drug was reported to takeplace initially following a profile similar to that reported for phys-ically entrapped systems but reaching a plateau at the end of theincubation period. This is according to expectations since releaseof the physically entrapped drug is mainly depending on diffu-sion/degradation factors whereas the releasing from ionically cou-pled systems, as it is the case of polyacid/DOX complexes, is largelydetermined by the environment conditions required to cleave theionic link between the drug and the polymer.

4. Conclusions

Result presented in this work show that the biopolymer PMLAcan be used as a carrier for targeted delivery of DOX, a hydrophobicdrug widely used in cancer therapy. In aqueous solution ionizedPMLA interacted efficiently with DOX to form stable ionic com-plexes. The complexes tend to self-assemble in micro or nano-spheres according to their polymer/drug ratio. These particlesunderwent hydrolysis at a rate that was dependent on pH andcomplex composition. The ability of PMLA/DOX complex particlesto operate as a pH-sensitive drug delivery system has been evi-denced. The DOX release rate from the complexes was enhanced

w/w%) Release at 24/240 h (%) Reference

2.7/8.3 This work2.1/7.1 This work2.0/5.4 This work8.5/�18 Manocha and Margaritis, [2]14.0/29.0 Yousefpour et al., [11]14.3/27.4 Lanz et al., [9]5.2/30.2 Lanz et al., [19]

A. Lanz-Landázuri et al. / Reactive & Functional Polymers 81 (2014) 45–53 53

by decreasing pH and, in less degree, by increasing the ionicstrength of the medium. Furthermore the release rate could bemodulated by adjusting the composition of the complexes. It canbe concluded therefore that PMLA/DOX complexes afford a goodpotential for drug delivery in cancer therapy.

Acknowledgements

Financial support for this work was provided by MICINN (Spain)with Grant MAT2012-38044-C03-03, and by AGAUR (Catalonia)2009SGR1469 Grant to ALL.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.reactfunctpolym.2014.04.005.

References

[1] B. Manocha, A. Margaritis, Crit. Rev. Biotechnol. 28 (2008) 83–99.[2] B. Manocha, A. Margaritis, J. Nanomater. 2010 (2010) 1–9.[3] B.S. Lee, M. Vert, E. Holler, Water-soluble aliphatic polyesters: Poly(malic

acid)s, in: Y. Doi, A. Steinbuechel (Eds.), Biopolymers, Wiley-VCH, New York,2002, pp. 75–103.

[4] O. Coulembier, P. Degée, J.L. Hedrick, P. Dubois, Prog. Polym. Sci. 31 (2006)723–747.

[5] M. Vert, Polym. Degrad. Stab. 59 (1998) 169–175.[6] E. Holler, Poly(malic acid) from natural sources, in: N.P. Cheremisinoff (Ed.),

Handbook of Engineering Polymeric Materials, Marcel Dekker, USA, 1997, pp.93–103.

[7] J. Portilla-Arias, M. García-Alvarez, J.A. Galbis, S. Muñoz-Guerra, Macromol.Biosci. 8 (2008) 551–559.

[8] J. Portilla-Arias, R. Patil, J. Hu, H. Ding, K.L. Black, M. García-Alvarez, S. Muñoz-Guerra, J.Y. Ljubimova, E. Holler, J. Nanomater. (2010), http://dx.doi.org/10.1155/2010/825363.

[9] A. Lanz-Landázuri, M. García-Alvarez, J. Portilla-Arias, A. Martínez de Ilarduya,R. Patil, E. Holler, J.Y. Ljubimova, S. Muñoz-Guerra, Macromol. Biosci. 10 (2011)1370–1377.

[10] B.S. Lee, M. Fujita, N.M. Khazenzon, K.A. Wawrowsky, S. Wachsmann-Hogiu,D.L. Farkas, K.L. Black, J.Y. Ljubimova, E. Holler, Bioconjugate Chem. 17 (2006)317–326.

[11] P. Yousefpour, F. Atyabi, E.V. Farahani, R. Sakjtianchi, R. Dinarvand, Int. J.Nanomed. 6 (2011) 1487–1496.

[12] F.M. Muggia, J.D. Hainsworth, S. Jeffers, P. Miller, S. Groshen, M. Tan, L. Roman,B. Uziely, L. Muderspach, A. García, A. Burnett, F.A. Greco, C.P. Morrow, L.J.Paradiso, L.J. Liang, J. Clin. Oncolo. 15 (1997) 987–993.

[13] A. Gabizon, S. Amselem, D. Goren, R. Cohen, S. Druckmann, I. Fromer, R. Chisin,T. Peretz, A. Sulkes, Y. Barenholz, J. Liposome Res. 1 (1990) 491–502.

[14] J. Park, P.M. Fong, J. Lu, K.S. Russell, C.J. Booth, W.M. Saltzman, T.M. Fahmy,Nanomed 5 (2009) 410–418.

[15] X. Hu, S. Liu, Y. Huang, X. Chen, X. Jing, Biomacromolecules 11 (2010). 2094–2012.

[16] H. Guan, M.J. McGuire, S. Li, K.C. Brown, Bioconjugate Chem. 19 (2008) 1813–1821.

[17] T. Etrych, P. Chytil, M. Jelínková, B. Ríhová, K. Ulbrich, Macromol. Biosci. 2(2002) 43–52.

[18] M. Yokoyama, S. Fukushima, R. Uehara, K. Okamoto, K. Kataoka, Y. Sakurai, T.Okano, J. Controlled Release 50 (1998) 79–92.

[19] A. Lanz-Landázuri, M. García-Alvarez, J. Portilla-Arias, A. Martínez de Ilarduya,E. Holler, J.Y. Ljubimova, S. Muñoz-Guerra, Macromol. Chem. Phys. 213 (2012)1623–1631.

[20] H.S. Yoo, E.A. Lee, T.G. Park, J. Controlled Release 82 (2002) 17–27.[21] X. Shuai, H. Ai, N. Nasongkla, S. Kim, J. Gao, J. Controlled Release 98 (2004)

415–426.[22] P. Kolhe, E. Misra, R.M. Kannan, S. Kannan, M. Lieh-Lai, Int. J. Pharm. 259 (2003)

143–160.[23] D. Schmaljohann, Adv. Drug Delivery Rev. 58 (2006) 1655–1670.[24] X. Liu, R. Ma, J. Shen, Y. Xu, Y. An, L. Shi, Biomacromolecules 13 (2012) 1307–

1314.[25] Y. Tian, L. Bromberg, S.N. Lin, T.A. Hatton, K.C. Tam, J. Controlled Release 121

(2007) 137–145.[26] S. Kayal, R.V. Ramanujan, Mater. Sci. Eng. C 30 (2010) 484–490.[27] R. Patil, J.A. Portilla-Arias, H. Ding, S. Inoue, B. Konda, J. Hu, K.A. Wawrowsky,

P.K. Shin, K.L. Black, E. Holler, J.Y. Ljubimova, Pharm. Res. 27 (2010) 2317–2329.

[28] S. Honary, F. Zahir, Trop. J. Pharm. Res. 12 (2013) 255–264.[29] L. Hu, Z. Mao, C. Gao, J. Mater. Chem. 19 (2009) 3108–3115.[30] X. Li, D.J. Hirsh, D. Cabral-Lilly, A. Zirkel, S.M. Gruner, A.S. Janoff, W.R. Perkins,

Biochim. Biophys. Acta 1415 (1998) 23–40.[31] G. Haran, R. Cohen, L.K. Bar, Y. Barenholz, Biochim. Biophys. Acta 1151 (1993)

201–215.