Np

HASa

b

c

Td

e

Sf

g

h

i

a

ARRAA

KlBL

1

tTls[p

aS

(

0h

International Journal of Biological Macromolecules 62 (2013) 180– 187

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules

jo ur nal home p age: www. elsev ier .com/ locate / i jb iomac

anostructure l-asparaginase-fatty acid bioconjugate: Synthesis,reformulation study and biological assessment

ajar Ashrafia,b, Mohsen Aminic, Soliman Mohammadi-Samania,d, Younes Ghasemie,mir Azadi f, Mohammad Reza Tabandehg, Eskandar Kamali-Sarvestanih,i,aeid Daneshamouza,d,∗

Department of Pharmaceutics, Faculty of Pharmacy and Pharmaceutical Sciences Research Center, Shiraz University of Medical Sciences, Shiraz, IranStudent Research Committee, Shiraz University of Medical Sciences, Shiraz, IranDepartment of Medicinal Chemistry and Drug Design and Development Research Centre, Faculty of Pharmacy, Tehran University of Medical Sciences,ehran, IranCenter for Nanotechnology in Drug Delivery, Shiraz University of Medical Sciences, Shiraz, IranDepartment of Pharmaceutical Biotechnology, Faculty of Pharmacy and Pharmaceutical Sciences Research Center, Shiraz University of Medical Sciences,hiraz, IranBiopharmaceutics and Pharmacokinetic Division, Department of Pharmaceutics, Faculty of Pharmacy, Tehran University of Medical Sciences, Tehran, IranDepartment of Biochemistry and Molecular Biology, Faculty of Veterinary Medicine, Shahid Chamran University of Ahvaz, Ahvaz, IranDepartment of Immunology, Shiraz Medical School, Shiraz University of Medical Sciences, Shiraz, IranAutoimmune Diseases Research Center, Shiraz University of Medical Sciences, Shiraz, Iran

r t i c l e i n f o

rticle history:eceived 27 May 2013eceived in revised form 19 August 2013ccepted 19 August 2013vailable online xxx

eywords:-Asparaginase

a b s t r a c t

The present study aims to develop a novel l-asparaginase fatty acid bioconjugates and characterize theirapplicability for intravenous delivery of l-asparaginase. These bioconjugates were achieved by covalentlinkage of fatty acids having different chain lengths (C12, C16 and C22) to the native enzyme. To determinethe optimum conditions of bioconjugation, the effect of lipid:protein ratios, reaction time and mediumcomposition on enzyme activity and conjugation degree were evaluated. The native and bioconjugateshave been characterized by activity, conjugation degree, particle size, and zeta potential. The resultsshowed that bioconjugated l-asparaginase were more resistant to proteolysis, more stable at different

ioconjugationipid–protein drug delivery

pH, and had prolonged plasma half-life, compared to the native form. From partition coefficient study,the modified enzymes showed approximately 15-fold increase in hydrophobicity. Secondary structureanalysis using circular dichroism revealed alteration after lipid conjugation. In addition, the Michaelisconstant of the native enzyme was 3.38 mM, while the bioconjugates showed the higher affinity to thesubstrate l-asparagine. These findings indicate that new lipid bioconjugation could be a very usefulstrategy for intravenous delivery of l-asparaginase.

. Introduction

l-Asparaginase (l-ASNase), an antineoplastic agent, catalyzeshe hydrolysis of the non-essential amino acid l-asparagine [1,2].he leukemic cells are unable to synthesize the adequate levels of-asparagine for their metabolism, due to the lack of l-asparagine

ynthetase; while normal cells can self synthesize l-asparagine3,4]. l-ASNase has sufficient efficacy against leukemia and lym-hosarcoma in clinical experiments, alone or in combination with

∗ Corresponding author at: Department of Pharmaceutics, Faculty of Pharmacynd Pharmaceutical Sciences Research Center, Shiraz University of Medical Sciences,hiraz, Iran. Tel.: +98 711 2424127x258; fax: +98 711 2424126.

E-mail addresses: daneshamuz@sums.ac.ir, daneshamouz@gmail.comS. Daneshamouz).

141-8130/$ – see front matter © 2013 Published by Elsevier B.V.ttp://dx.doi.org/10.1016/j.ijbiomac.2013.08.028

© 2013 Published by Elsevier B.V.

other antitumor agents [5]. The antitumor property of l-ASNaseis not specific, and limited enzyme stability in different biologicalenvironments (i.e. pH, temperature and ionic strength) and rapidclearance rate of l-ASNase by plasma, are the main drawbacksof the use of l-ASNase in clinical practices [6]. Accordingly, needfor repeated administration causes hypersensitivity (mild allergicreactions to anaphylaxis) and immunogenicity [7,8]. Consequently,consideration is being given to prolong the action time of theenzyme, improve its stability in the body, minimize the immuno-genic effect and lower the affinity to natural inhibitors.

One approach to improve the efficacy and stability of the enzymeis chemical modification using various kinds of polymers. Attempts

have been made to render the protein less immunogenic thannative form, increase its biological half-life in the body and improveproteolysis stability such as immobilization of l-ASNase with albu-min [9], dextran [10], polyethylene glycol (PEG) [11], chitosan [12],

Biolog

injyesteittrpichlldantetta

aec(calAhtt

2

2

wpobaop

2b

fatlgggm1

H. Ashrafi et al. / International Journal of

nulin [13], silk fibroin and sericine [3,14]. They include mainlyatural or artificial soluble or insoluble polymers, called biocon-

ugation that have received most extensive attention in recentears. Previously proposed strategies to improve the therapeuticfficacy of proteins included the entrapment of protein into lipo-omes which prevents interaction of preformed antibodies withhe proteins [1,15]. It has been reported that this modified form ofnzyme has improved biological half-life, abrogate the acute tox-city and preserve the in vivo antitumor activity with respect tohe native l-ASNase. Lipid nanoparticles, e.g. solid lipid nanopar-icles, nanostructured lipid carriers and lipid-drug conjugates mayepresent, in fact, promising carriers in drug delivery of protein andeptides [16]. Promising results have been reported concerning the

ncorporation of therapeutically relevant peptides (e.g. calcitonin,yclosporine A, insulin, LHRH, somatostatin), protein antigens (e.g.epatitis B and malaria antigens) and model protein drugs (e.g.-ASNase, bovine serum albumin and lysozyme) into this kind ofipid particulate system [16]. Recently, lipid-drug conjugates areesigned to overcome the limitation of other lipidic structures suchs insufficient loading capacity for most proteins with hydrophilicature. Lipid-drug conjugates are prepared either by salt forma-ion (e.g. with a fatty acid) or by covalent linking (e.g. to esters orthers). In the salt formation process, a suitable solvent is selectedo dissolve the free drug base and fatty acid in it and the solvent ishen being evaporated. In the other procedure, the drug and a fattylcohol react in the presence of a catalyst [17,18].

In order to overcome some of the limitations of native l-ASNasedministration, lipid-drug conjugate (LDC) synthesis was consid-red in current study. The enzyme was modified as lipid-drugonjugate (LDC) by conjugation with lauric acid (LA), palmitic acidPA) and behenic acid (BA). In doing so, we reported the biophysi-al and biochemical characheterization of a series of short, mediumnd long chain fatty acids (C12, C16 and C22) in conjugation with-ASNase. It is proposed that covalent coupling of fatty acids to l-SNase may improve its stability, extend its biological half-life andence biological activity, and thus reduce frequency of administra-ion. We hope all of these modifications could help improve theherapeutic utility of the enzyme.

. Materials and methods

.1. Chemical reagents

Escherichia coli l-asparaginase 140 kDa with 10,000 IU activityas from Medac®, Germany. Palmitic acid and lauric acid wereurchased from Merck Co. (Darmstadt, Germany). Behenic acid wasbtained from Fluka Co. Ltd. (Switzerland). l-Asparagine, trypsin,ovine serum albumin and Trinitrobenzene Sulfonic Acid (TNBS)nd ninhydrin were purchased from Sigma (St. Louis, MO, USA). Allther chemicals and reagents were of analytical grade and wereurchased locally.

.2. Synthesis and purification of l-ASNase-fatty acidioconjugates

The lipid bioconjugates were achieved by covalent linkage ofatty acid with different chain lengths of lauric acid (C12), palmiticcid (C16), and behenic acid (C22) to the native l-ASNase. To findhe optimum condition of bioconjugates synthesis, the effect ofipid:protein ratio, reaction time and reaction medium on conju-ation degree and enzyme activity were investigated. The amine

roup of lysine in l-ASNase was conjugated with the carboxylicroup of fatty acids using carbodiimide activator with someodifications [19]. Briefly, adequate amounts of each fatty acid (6,

2.5, 25, 50, 100 or 200 moles) were dissolved in 5 mL of dimethyl

ical Macromolecules 62 (2013) 180– 187 181

sulfoxide (DMSO) following the addition of 1.1 mole N-(3-dimethylamino propyl)-N-ethyl carbodiimide-hydrochloride (EDC) and1mole N-hydroxysuccinimide (NHS). The mixture was stirred for3 h at 25 ◦C. Subsequently, 1 mole of l-ASNase in 5 mL of aqueousmedium (water, pH 7.4 or Tris buffer, pH 8.6) was added to the mix-ture in dropwise manner and stirred for 24 h. After that, the productwas transferred to dialysis membranes (MW cut-off 12 kDa) against1 L water (24 h at 25 ◦C) for purifying the bioconjugates and remov-ing the un-reacted fatty acid, EDC, NHS and DMSO. Differentialscanning calorimetry (DSC) and colorimetric spectroscopy methods[20] were used to confirm the complete elimination of un-reactedfatty acids form the mixture. The dispersion was finally freeze-driedand a white spongy form of the purified l-ASNase-lipid bioconju-gate was obtained. All experiments were performed in triplicate.

2.3. Lyophilization of the native l-ASNase and bioconjugates

After dialysis, the total volume of Lipid-conjugated l-ASNase(about 10 mL) has been divided into different containers with theliquid depth about 1 cm. The same condition was performed forthe native enzyme as control sample. Lipid-conjugated l-ASNaseand control samples were frozen at −70 ◦C for a minimum of 12 h.In the primary drying, the freeze-dryer chamber was evacuatedand the shelf temperature and chamber pressure were −45 ◦C and0.07 mbar, respectively to sublimate free water out of the system.It took about 30 h. After that, within the secondary drying, by thechamber pressure of 0.04 mbar bound water was removed via des-orption for 18 h. In fact, samples were lyophilized for 48 h using afreeze drier (Christ, �1-4 LD plus, Germany). For enzyme activity,the lyophilized samples were reconstituted with phosphate buffer(0.025 M, pH 8.5) to appropriate concentrations. All experimentswere performed in triplicate.

2.4. Determination of conjugation degree

The degree of lysine residues in the native enzyme and thebioconjugates was evaluated by two colorimetric methods. Nin-hydrin and TNBS can bind to free amine group of lysine of theenzyme, so the degree of chemical modification of the enzymecould be evaluated [14,21,22]. The correlation studies were per-formed between the two colorimetric methods. In this study, theninhydrin colorimetric method was used as a routine process fordetermination of conjugation degree. To prevent of any inaccuracy,in each experiment one sample was evaluated with both techniques(i.e. ninhydrin and TNBS). The conjugation degree was determinedusing the following equation [14]:

Conjugation degree =[

native enzyme concentration − bioconjugates concentrationnative enzyme concentration

]

× 100

2.4.1. Reaction of l-ASNase with ninhydrinIn this regard, 1.8 mL of 0.2% ninhydrin reagent (w/v in N,N-

dimethylformamid) was added to 200 �L of l-ASNase solution(concentration ranging from 1 to 10 mg/mL). The samples wereheated in a water bath at 90 ± 5 ◦C for 5 min. Afterwards, theabsorbance of the solution was measured at 569 nm against theblank solution at room temperature. Finally, the calibration graphwas constructed by plotting the absorbance vs. concentration ofl-ASNase.

2.4.2. Reaction of l-ASNase with TNBSTo 1 mL of l-ASNase solution with different concentrations

(0.1–1 mg/mL), 1 mL of 4% NaHCO3 and 1 mL of 0.1% TNBS were

1 Biolog

a11atb

2

AWtosawattttFpt

2

B0twFroutt

2

rswCtfaub5m

2

2

fv

wM

82 H. Ashrafi et al. / International Journal of

dded. The solutions were allowed to react at 40 ◦C for 2 h; then, mL of 10% sodium lauryl sulfate (SLS) and subsequently 0.5 mL of N, HCl were added. The absorbance of the solution was measuredt 335 nm against the treated blank with 1 mL of water instead ofhe l-ASNase solution. Finally, the calibration graph was preparedy plotting the absorbance vs. concentration of l-ASNase.

.5. Determination of l-ASNase activity

The enzyme activity for the native and lipid-conjugated l-SNase were determined using Nessler’s reagent according to theriston and Yellin method [23] with slight modifications. The reac-

ion mixture consisted of 0.1 mL of l-asparagine (186 mM), 1.0 mLf 0.05 M Tris buffer (pH 8.5), 0.9 mL water and 0.1 mL of l-ASNaseolution with different concentration in the range of 1–10 �g/mLnd the mixture was incubated for 30 min at 37 ◦C. The reactionas terminated by the addition of 0.1 mL of 1.5 M trichloroacetic

cid solution and then, centrifuged to separate the precipitated pro-ein from the solution. Activity of the enzyme was determined byhe colorimetric method at 436 nm using a UV–visible spectropho-ometer. A quantitative standard curve was plotted accordingo the coupled librated ammonium ion using Nessler’s reagent.or lyophilized bioconjugated samples having unknown activity,roper concentrations were prepared which has the absorbance inhe linear range analyzed.

.6. Determination of protein content

Protein concentration was determined spectroscopically usingradford protein assay [24], for the concentration range of.25–20 �g/mL. In doing so, 100 �L of the solutions was added intohe test tubes and the volume was adjusted to 800 �L with distilledater. After that, 3.2 mL of Bradford protein reagent was added.

inally, the absorbance at 595 nm was measured against a blankeagent prepared from 800 �L of the appropriate buffer and 3.2 mLf Bradford protein assay reagent. Calibration curves were plottedsing standard native l-ASNase and BSA. The concentration of pro-ein was plotted against the corresponding absorbance resulting inhe standard curve.

.7. Determination of molecular weight

The native l-ASNase and the bioconjugates were evaluated byeducing sodium dodecyl sulfate-polyacrylamide gel electrophore-is (SDS-PAGE), based on the method described by Laemmli [25],ith 12% acrylamide gel and 5% condensing gel stained with 0.25%oomassie Brilliant Blue R-250 (Aldrich, Milwaukee, USA). The elec-rophoresis analysis was performed for the molar ratio of 100:1or all bioconjugated l-ASNase samples that was then used forll subsequent reactions (lane 1, the native l-ASNase with sub-nit molecular weight about 38 kDa; lane 2, M.W. markers; lane 3,ehenic acid bioconjugate; lane 4, palmitic acid bioconjugate; lane, lauric acid bioconjugate). The gels were loaded with an averageass of 100 �g per well and run at 100 mA constant current.

.8. Characterization of the bioconjugates

.8.1. Kinetic measurementsIn kinetic study, the affinity of the native and modified l-ASNase

or the substrate was calculated by the Michaelis constant (Km)alue from the Lineweaver–Burk equation [26].

1v

= Km

Vmax× 1

C+ 1

Vmax

here Vmax is the maximum elimination rate and Km is theichaelis constant that reflects the capacity of the enzyme system,

ical Macromolecules 62 (2013) 180– 187

C is the various concentration of drug and � is a series of reactionrates.

The Km value was calculated by measuring the enzyme cat-alytic activity at different asparagine concentrations (0.4–2 mM)in 0.05 M Tris buffer, pH 8.5. The smaller value of the Km indicatedas higher affinity of the enzyme to the substrate. All experimentswere performed in triplicate.

2.8.2. Determination of partition coefficientTo measure the probable increment of hydrophobicity after

lipid-conjugation of l-ASNase, equivalent volumes (2.5 mL) ofwater and octanol were mixed vigorously using a vortex to obtainan emulsion. Five hundred microgram of the native or lipid-conjugated l-ASNase was added to the mixture. The emulsionswere stirred for 2 h and then, centrifuged at 5000 × g for 5 min.The aqueous phase was analyzed for protein content (as describedin Section 2.6). Similar procedure was performed in the absenceof protein for the preparation of blank sample. The partition coef-ficient in octanol/water was calculated as the ratio between theprotein concentration in octanol and water phases.

2.8.3. Determination of pH-activity profileIn order to achieve the optimum medium having pH of

maximum activity, the enzyme activities of the native and lipid-conjugated l-ASNase were determined by the incubation of eachenzyme in phosphate buffer (0.05 M) in different ranges of pH from3 to 12. All experiments were performed in triplicate.

2.8.4. Determination of time-activity profile in PBSEnzyme activities of the native and lipid-conjugated l-ASNase

were determined in PBS (pH 7.4) at 37 ◦C according to the incuba-tion time (0, 1, 2, 4, 8, 10, 12, 24, 48, 72, 96, 168 and 336 h intervals).To determine the activity of the enzyme in PBS, equivalent prepa-rations containing 1 mg of the native or conjugated l-ASNase wereloaded in 2 mL of respective medium. After mild mixing, sampleswere incubated at 37 ◦C for the mentioned time intervals. Prepa-rations were analyzed for catalytic activity and enzyme in vitrohalf-life by taking 0.1 mL of each sample. All experiments wereperformed in triplicate.

2.8.5. Determination of time-activity profile in plasmaEquivalent preparations containing 1 mg of native and modi-

fied l-ASNase were loaded in 2 mL of plasma. After mild mixing,samples were incubated at 37 ◦C for 0, 1, 2, 4, 8, 10, 12, 24, 48, 72,96, 168 and 336 h. Preparations were analyzed for catalytic activityand in vitro half-life in plasma by taking 0.1 mL of each sample forthe mentioned time intervals. All experiments were performed intriplicate.

2.8.6. Stability to proteolysisEquivalent amounts of the native or lipid-conjugated l-ASNase

were added to 2 mL phosphate buffer solution (pH 8.5) containing50 IU trypsin. Samples were incubated at 37 ◦C for 4 h and resid-ual activity of the enzyme was measured in a time interval of 10,20, 30, 40 and 60 min. All data were average values of triplicatemeasurements.

2.8.7. Circular dichroismFar-UV circular dichroism (CD) analysis was performed on the

reconstituted protein solutions to compare the secondary struc-tural characteristics of the native and lipid-conjugated l-ASNase.Data were collected on a Jasco J-810CD Spectropolarimeter (Möl-

ndal, Sweden) in the 190–240 nm range using a cell path lengthof 1 mm at 25 ◦C and a protein concentration of 0.1 mg/mL. Jasco’sSpectra ManagerTM software (Jasco Secondary Structure Estima-tion version 1.0 and Jasco Spectra Analysis version 1.53.02) is

Biological Macromolecules 62 (2013) 180– 187 183

uf(

2

oiJqp5pppe

objMs

3

3

cbltlpNCtpfi

alalNaIufsgtaN2to

gicioaa

H. Ashrafi et al. / International Journal of

sed to analyze the structure. The reported spectra were correctedor buffer contribution and expressed as mean residue ellipticitymdeg) as a function of wavelength (nm).

.8.8. Particle size, size distribution and zeta potentialThe statistical central and dispersion indices of the particle sizes

f bioconjugates were determined using a laser-diffraction basednstrument Particle Size Analyzer (Shimadzu, model SALD-2101,apan). The particle size measurements were performed using auartz cell in the manual mode. Samples were diluted to appro-riate concentrations (0.1 mg/mL) with distilled/filtered water (pH.5–6). Measurements were collected at 25 ◦C. The instrumentarameters were set as follows: refractive index of material (i.e.rotein): 1.45; dispersant viscosity (i.e. water): 0.8872; and dis-ersant refractive index: 1.33. Analyses were done in triplicate forach sample.

Since, the surface zeta-potential of the bioconjugates is onef the major parameters with remarkable impact on the in vivoehaviors of the bioconjugates, the zeta-potential of the biocon-

ugates was measured using a zetameter (Zetasizer V-R 3000-HS,alvern Instruments, UK), working based on photon correlation

pectroscopy technique.

. Results and discussion

.1. Synthesis of l-ASNase-fatty acid bioconjugates

In this study, l-ASNase was conjugated with three differenthain lengths of fatty acid including lauric acid, palmitic acid andehenic acid (C12, C16 and C22). Primary structure of one subunit of-ASNase isolated from E. coli contained 22 lysine residues, whilehe active site characterization of the native enzyme did not revealysine residues [27]. Thus, the lipid-conjugated l-ASNase was pre-ared by the direct reaction of carboxyl group of fatty acids toH2 groups of lysine and N-terminal amino groups of the enzyme.onjugation of lysine group has been reported for the modifica-ion of E. coli and Erwinia l-ASNase using oxidized dextran [10],olyethylene glycol [11], oxidized inulin [13], silk sericine andbroin [14,28].

Lipid-conjugated l-ASNase with different degrees of fattycid attachment was obtained by changing the molar ratio ofipid:protein in the reaction medium and the results of enzymaticctivities and conjugation degrees are summarized in Fig. 1. As theipid:protein mole ratio in the reaction mixture was increased, theH2 groups of the native enzyme have been modified progressivelynd the conjugation degree increased significantly (P-value < 0.05).n general, the catalytic activity of the enzyme gently decreasedpon modification [3,13,14]. This trend has been shown in Fig. 1or lipid conjugation of l-ASNase and could be explained as: (a) theteric hindrance due to conjugation of fatty acid with free aminoroups of the l-ASNase can prevent l-asparagine from approachinghe enzyme active site and (b) the effect of reaction conditions suchs organic solvent like DMSO, and other reagents including EDC andHS can affect the affinity; so, the experiments discussed in Section.2 on the native enzyme treated without fatty acid addition (con-rol sample) indicate that reaction condition had the major effectsn activity reduction.

In previous studies, the best report in modification of the aminoroups of l-ASNase was due to conjugation with dextran, levan andnulin with 15–33% [10,13,29] while in the present study, higheronjugation degrees were achieved along with expected reduction

n activity. This condition may be as a result of the participationf more reactive carboxyl groups at higher concentration of fattycid along with increasing steric inhibition by conjugated fattycid and may interfere with enzyme and substrate reaction. Thus,

Fig. 1. Conjugation degree and activity of different lipid:l-asparaginase mole ratios(6:1, 12.5:1, 25:1, 50:1, 100:1, and 200:1) in the reaction mixture; (A) behenic acid,(B) palmitic acid and (C) lauric acid.

we conducted the following experiment by modified l-ASNase atmolar ratio of 100:1 of lipid:protein for further optimization, whilethe modification of amino groups have accepted residual activ-ity (about 52–57% activity of the native enzyme, Fig. 1) whichis in agreement with other studies [13]. Similar findings werereported by different molecules binding to l-ASNase such as dex-tran sulfate [10], chitosan [12], polyethylene glycol [11], levan [29],oxidized inulin [13] and silk sericin [30]. The obtained results inthis study, based on the calculation of the free amino groups ofthe enzyme, indicate modification of about 12 lysine residues inaverage (54.35 ± 10.74%, 53.21 ± 6.61% and 48.74 ± 2.80% % conju-gation degree for BA, PA and LA-bioconjugates, respectively) for thelipid:protein molar ratio of 100:1.

In this study, due to the differences in conjugation chemistrywith other investigations on l-ASNase, the optimization of reactioncondition should be considered. To this end, the effects of reac-tion time and medium on activity and conjugation degree were

184 H. Ashrafi et al. / International Journal of Biological Macromolecules 62 (2013) 180– 187

Table 1Effect of reaction time and medium on conjugation degree and activity of behenic acid:l-asparaginase conjugated form.

Behenic acid:l-asparaginase ratio Conjugation media Time of conjugation (h) Activity (%) Conjugation degree (%)

100:1 Water pH 7.4 12 h 66.78 ± 0.59 37.2 ± 1.74100:1 Water pH 7.4 24 h 63.32 ± 2.62 52.1 ± 0.54

8 h

4 h

4 h

iwajwacsatd2

pidoieodtfieosAYc

tuDttbttbs

3

mu0matr10rTa

been even conjugated. The band broadenings were well shown inlanes 3–5. In fact, the absence of the sharp band in lanes 3–5, mightbe due to the different number of fatty acid molecules attached toeach molecules of l-ASNase. Also it must be mentioned that the

100:1 Water pH 7.4 4100:1 Water pH 7.4 2100:1 Tris buffer pH 8.6 2

nvestigated (Table 1). In previous studies, reaction times 24–48 here selected for the conjugation of l-ASNase with oxidized inulin

nd silk sericine [13,14]. In this experiment, the activity and con-ugation degree of bioconjugates (at optimum mole ratio of 100:1)

ere quantified in 12, 24 and 48 h. The activity of bioconjugatesfter 12 h of reaction was significantly higher than that of the bio-onjugates after 24 and 48 h; while their conjugation degrees wereignificantly lower than those for two other samples. Moreover, thectivity of the bioconjugates after 48 h was significantly lower thanhat of the bioconjugates after 24 h reactions, but the conjugationegree was not increased significantly, as predicted. Thereafter, the4 h were set as the optimum time for the conjugation.

The effects of pH and medium (Tris buffer (0.05 M, pH 8.6) andurified water pH 7.4) on the activity and conjugation degree were

nvestigated. Activity and conjugation degree were not significantlyifferent in these two media, even decreased in Tris buffer. Thisbservation could be related to the R COO− tendency for beingonized at alkaline pH, considering its pKa. Moreover, the interfer-nce of the excess OH− in alkaline medium can interrupt the actionf EDC during conjugation and so on the activity and conjugationegree at alkaline pH were a little but not significantly lower thanhat in neutralized water. Thus, the simple aqueous medium (puri-ed water, pH 7.4) was selected as the reaction medium in thisxperiment. On the other hand, the conjugation process was carriedut at alkaline medium for l-ASNase to get better yield in previoustudies. Tabandeh et al. [13] and Martins et al. [8] performed the l-SNase conjugation process at alkaline medium (pH = 9) but Zhangu-Qing et al. used the phosphate buffer (pH = 7.4) for l-ASNaseonjugation [3].

Differential scanning calorimetry (DSC) and colorimetric spec-roscopy (CS) methods were used to confirm the elimination ofn-reacted fatty acids from the bioconjugates after dialysis. TheSC thermograms of unreacted fatty acids after dialysis were

he same as that of the respective standard lipid. Consequently,his demonstrated qualitatively that unreacted fatty acid coulde removed by dialysis. Furthermore, quantitative determina-ion of the fatty acid using colorimetric spectroscopy, confirmedhat three kinds of unreacted fatty acid can be separated fromioconjugates more than 90% by dialysis during 10 h (data nothown).

.2. Detection methods

Two colorimetric methods (ninhydrin-based and TNBS-basedethods) were used to determine the conjugation degree. Val-

es obtained from these methods had a correlation (R) value of.9997 (confidence interval95% = −0.066 to 0.3). The slopes of twoethods were not significantly different from unity. It means there

re high correlations between the two methods for measuring ofhe conjugation degree. The ninhydrin method produced linearesponses throughout the l-ASNase concentration range from 1 to0 mg/mL with r2, slope and intercept values of 0.9996, 0.0867 and

.0121, respectively, for a typical daily calibration curve. The linearesponses for l-ASNase concentration range from 0.1–1 mg/mL forNBS method has r2, slope and intercept values of 0.9996, 0.7849nd 0.0024, respectively.

52.14 ± 1.56 50.9 ± 0.9466.67 ± 2.02 51.2 ± 1.0263.40 ± 1.71 48.5 ± 1.02

Typical daily calibration curve of l-ASNase activity (Nesslermethod) was linear within the l-ASNase concentration range of1–10 �g/mL, and r2, slope and intercepts values of 0.998, 0.0598and 0.0156, respectively were achieved.

In Bradford colorimetric protein assay, the absorbance of thenative l-ASNase with the concentration range of 0.25–20 �g/mLwas determined at 595 nm; r2, slope and intercept values were0.9991, 0.0359 and 0.0007, respectively.

3.3. Determination of molecular weight

Herein, the formation of modified l-ASNase and the relativemolecular mass of the modified protein were confirmed by SDS-PAGE with 12% gel as Laemmli reported in his research [25]. Asshown in Fig. 2, the sharp band about 35–38 kDa stands for themolecular weight of the pure enzyme (lane 1) due to the sub-unit of l-ASNase. Moreover, broad molecular weight bands withwide distribution in molecular weight achieved by the conjugatedsamples (lanes 3–5). This wide distribution is due to the differ-ent number of fatty acid molecules attached to the native enzyme.The SDS-PAGE data of the conjugated samples in lanes 3–5 showsthe different range in the degree of conjugation. Hence, molecularweight range of 35–38 kDa for conjugated samples can be referredto the low conjugation degree or the native enzyme that has not

Fig. 2. SDS-PAGE data of the native l-asparaginase and lipid-bioconjugates achievedin optimum experiment condition with 100:1 mole ratio of lipid:protein; lane 1,the native asparaginase with subunit molecular weight about 38 kDa; lane 2, M.W.markers; lane 3, Behenic acid bioconjugate; lane 4, Palmitic acid bioconjugate; lane5, Lauric acid bioconjugate.

H. Ashrafi et al. / International Journal of Biological Macromolecules 62 (2013) 180– 187 185

Table 2Effect of freeze drying on the activity of the native and lipid-conjugated l-asparaginase.

l-Asparaginase/lipid-bioconjugate Activity % (before lyophilization) Activity % (after lyophilization) Activity recoveryb (%)

Controla 74.2 ± 4.2 69.5 ± 1.6 93.79 ± 3.59Behenic acid bioconjugate 77.0 ± 3.5 74.9 ± 1.8 97.40 ± 2.30Palmitic acid bioconjugate 81.9 ± 2.5 76.8 ± 1.0 93.79 ± 2.66Lauric acid bioconjugate 78.8 ± 3.2 78.8 ± 2.8 100.08 ± 4.58

a Control sample is the native enzyme under the reaction condition without fatty acid.b Activity recovery = (activity after lyophilization/activity before lyophilization) × 100.

Table 3The kinetic parameter of the native l-asparaginase and lipid-bioconjugates. The con-centrations of the substrate l-asparagine were between 0.4 and 2 mM. All data wereaverage value of triplicate measurement.

l-Asparaginase/lipid-bioconjugate Equation fromregression

Km (mM)

Native l-asparaginase y = 0.224x + 0.0663 3.38Behenic acid bioconjugate y = 0.1495x + 0.0778 1.92

mmdiS6ntnbtoiig

3

mst(tdb

3

3

raLtwhtLbtts

Table 4The particle sizes, zeta potential and partition coefficient of the native L-asparginaseand lipid-bioconjugates.

l-Asparaginase/lipid-bioconjugate

Particlesize (�m)

Zeta potential(mV)

Log (octanol/water)

Native l-asparginase 0.038 ± 0.02 −21.2 −0.26 ± 0.002Behenic acidbioconjugate

0.905 ± 0.06 −15.1 1.42 ± 0.013

Palmitic acidbioconjugate

0.790 ± 0.19 −13.7 1.17 ± 0.003

Laurie acid 0.961 ± 0.10 −14.2 l.11 ± 0.007

with previous studies based on chemical modification of l-ASNasethrough conjugation with oxidized inulin [13] and silk fibroin [3].

Palmitic acid bioconjugate y = 0.1889x + 0.0615 3.07Lauric acid bioconjugate y = 0.1673x + 0.0771 2.17

olecular weights of the attached fatty acids are not as high as theolecular weight changes significantly. This result emphasizes the

ecreases in electrophoretic mobility of the enzyme after the lip-dation and increases in molecular weight of the l-ASNase. Also,DS-PAGE result shows additional band in molecular weight about6 kDa in the bioconjugated samples. It is quite visible that theative and bioconjugates are different in their SDS-PAGE result inhis area. Therefore, due to the conjugation process this new phe-omenon is created. With regard to the molecular weight, it canecause of the some possible association followed by the conjuga-ion procedure. One of the associations that are also mentioned inther studies may due to the enzyme dimerization [31–33]. Dimer-zation of two enzyme subunit can lead to the observed results. Its notable that, this is just a hypothesis which needs more investi-ation.

.4. Lyophilization of the native l-ASNase and the bioconjugates

Lyophilization has been performed to convert the enzyme inore stable form prior to parenteral administration [34,35]. As

hown in Table 2, the l-ASNase-lipid bioconjugates could remainheir activity during lyophilization better than the control groupi.e. native enzyme). Lipid modification of protein plays an impor-ant role in stabilizing the proteins and hindering the proteinsegradation [36,37]. Higher activity value after lyophilization forioconjugate forms could be resulted from this effect of lipidation.

.5. Characheterization of the native and modified l-ASNase

.5.1. Kinetic measurementsThe Lineweaver–Burk plots based on the results obtained from

elative rate (V−1) vs. substrate concentration [S−1] for the nativend modified l-ASNase are shown in Table 3. The concentration of-asparagine was between 0.4 and 2 mM. Four linear equations forhe native and modified l-ASNase (BA, PA and LA bioconjugates)ere obtained from the plots. The smaller value of the Km indicatesigher affinity of the enzyme to the substrate. As shown in Table 3,he K values of the native and modified l-ASNase (BA, PA and

m

A bioconjugates) were 3.38, 1.92, 3.07, and 2.17, respectively. PAioconjugate has the Km value close to the native enzyme whilehe BA bioconjugate and LA bioconjugate showed smaller value ofhe Km, compared to the native form and higher affinity for theubstrate.

bioconjugate

3.5.2. Determination of partition coefficientTable 4 shows the octanol/water partition coefficient for the

native l-asparaginase and lipid-bioconjugates. Results show anincrease in the affinity of the bioconjugate to octanol being morethan 15-folds. These results reveal a significant increase in the affin-ity of the bioconjugate to hydrophobic medium. As expected, thepartitioning of the bioconjugates to the octanol phase considerablyraised with increasing the length of hydrocarbon chain of fatty acid.

3.5.3. Determination of pH-activity profilesThe effect of pH on the activities of native l-ASNase and the bio-

conjugates was studied by changing the pH of phosphate bufferfrom 3.0 to 12.0, as shown in Fig. 3. The optimum pH range for bio-conjugates (pH value from 6.5 to 9) was significantly broader thanthat for the native enzyme (pH value 7.5–8.5). Active site preser-vation could be the possible mechanism for protein stability in awide range of the pH. The active sites in native l-ASNase are moresensitive at lower pH, thus, without protecting effect of fatty acidschains, active site structure irreversibly changes in the presenceof higher [H+] concentration [13]. These results were consistent

Fig. 3. Effect of different pH of conjugation medium (between 3 and 12) on theactivity of the native and lipid-conjugated l-asparaginase. The highest activity ofnative and modified l-asparaginase has been set as 100% activity.

186 H. Ashrafi et al. / International Journal of Biological Macromolecules 62 (2013) 180– 187

Table 5The in vitro half-lives of the native l-asparaginase and the lipid-bioconjugates inhuman plasma and PBS (0.150 mM, pH 7.4). The highest activity of native and con-jugated l-asparaginase has been set as 100%.

l-Asparaginase/lipid-bioconjugate

PBS Plasma

t1/2 (h) K (h−1) t1/2 (h) K (h−1)

Native l-asparaginase 21.32 0.033 19.04 0.036Behenic acid bioconjugate 27.18 0.026 46.2 0.015

3

oTPpiAtce

3

aogwt4fmtgIcscrtotwp

Ftm

Fig. 5. Circular dichroic spectra of the native l-asparaginase, control (the

Palmitic acid bioconjugate 19.36 0.036 38.72 0.018Laurie acid bioconjugate 21.93 0.032 25.86 0.027

.5.4. Determination of time-activity profile in PBS and plasmaFirst order model was used to determine the in vitro half- life

f the native and bioconjugates l-ASNase as data presented inable 5. The half-life of the native l-ASNase was 21 and 19 h inBS and plasma, respectively. The half-life lengths of the behenic,almitic and lauric-bioconjugate were 46, 38 and 25 h, respectively

ndicating that lipid–protein conjugation can lead to increase in l-SNase in vitro half-life. One of the important methods to improve

he stability of l-ASNase is bioconjugation [2,3] and the fatty acidonjugation of l-ASNase could improve the stability, as reportedlsewhere [1,15].

.5.5. Stability to proteolysisResults from stability of native l-ASNase and lipid bioconjugates

gainst trypsin digestion are illustrated in Fig. 4. The resistancef the modified l-ASNase to trypsin digestion was considerablyreater than that of the native enzyme. The stability of free enzymeas very low. l-ASNase lost half of its original activity after the reac-

ion was performed for 10 min and completely lost its activity after0 min. While 13.8% of initial activity is maintained after 30 minor native enzyme, the activity of the modified enzyme retained

ore than 70% of its original activity. Even though the incubationime in trypsin solution was lengthened to 60 min, the bioconju-ates had maintained more than half of their original activities.n the previous studies on l-ASNase conjugation with inulin [13],hitosan [12] and silk fibroin [3] similar increasing in proteolysistability has been achieved. It has been known that trypsin specifi-ally hydrolyzes the peptide bonds consisting of lysine or arginineesidues [12]. The increase in proteolysis stability may be caused bywo factors: (a) the steric effect of the lipids hindered the contact

f protease with modified enzyme, and thus, prevented the pro-eolysis reaction; (b) the lysine residues of l-ASNase was modifiedith fatty acids, thus, prevented specific proteolysis reaction on theeptide bonds consisting of lysine residues.

ig. 4. The relative activity of the native and lipid conjugated l-asparaginase afterrypsin digestion (pH 8.5, 50 IU trypsin, 37 ◦C). The highest activity of native and

odified l-Asparaginase has been set as 100%.

native enzyme under conjugation process without lipid attendance) and lipid-bioconjugates. The concentration of the enzyme was 0.1 mg/mL at a path lengthof 1.0 mm.

3.5.6. Circular dichroismThe most widely used applications of protein CD are deter-

mining the secondary structure and folding properties of proteins.There have been many efforts to correlate the shape of circulardichroic spectra at particular wave lengths with the folding of thepolypeptide chain. Fig. 5 shows the circular dichroic spectrum ofl-ASNase, compared with the spectrum of the control and lipid-bioconjugated enzymes. The secondary structure of l-ASNase andfatty acid bioconjugates is investigated by far-UV CD in order tostudy the effect of lipidation. Peaks present in the 200–250 nmwavelength range (‘far-UV’) are generally a ‘w’-shaped spectra withtroughs around 222 and 208 nm being indicative of the presence of�-helical structures, and a ‘v’-shaped spectra with a trough around217–220 nm being indicative of �-sheet structures [38].

It was shown that the CD signals decrease dramatically upon lip-idation. From structural analysis, 30.39% of the polypeptide chainof the native l-ASNase is in the �-helix form while the controlsample (the native enzyme under reaction condition without fattyacids) revealed 23.11%. The �-helix contents of the BA, PA and LAbioconjugates were 28.64%, 27.63% and 35.87%, respectively. The�-turns in the native enzyme were 24.21%, for control sample (thenative enzyme treated as discussed in Section 2.2, without addi-tion of fatty acids) was 18.27% and for BA, PA and LA bioconjugateswere 17.82%, 11.07% and 00.03%, respectively. Consequently, thischange in secondary structures might be the cause of bioconjugatesactivity reduction after conjugation. As discussed in Section 3.1, theobserved activity reduction in lipidation, may have been because ofthe reaction condition, such as organic solvent and other reagents.Finally, as shown in the results, conjugation with BA, compared tothe PA and LA, has a minimum effect on secondary structure (espe-cially � turns) of the protein, and this may be the result of the sterichindrance of BA chain length. Similar results have been reported byPlesner et al. for GlycoPEGylated recombinant human factor VIIa[39].

Table 6 presents the percentages of secondary structural com-ponents for the deconvolution results.

3.5.7. Particle size, size distribution and zeta potentialLight scattering is the most commonly method used to ana-

lyze the size of nanoparticles. The surface charge and particle sizeof the bioconjugates are listed in Table 4. The surface charges of−21.2, −15.1, −13.7 and −14.2 (mV) are measured from the native,BA, PA and LA bioconjugates, respectively. The average particle

sizes after reconstitution of the BA, PA and LA bioconjugates were0.905 ± 0.06, 0.790 ± 0.19 and 0.961 ± 0.10 �m, respectively, witha narrow size distribution. The particles obtained in the presentstudy were uni-disperse (uni-modal curves) in terms of both central

H. Ashrafi et al. / International Journal of Biological Macromolecules 62 (2013) 180– 187 187

Table 6Circular dichroism deconvolution results: the percentages of secondary structural components of the native l-asparginase, control group (the native enzyme under conjugationprocess without lipid attendance) and lipid-bioconjugates (The concentration of the enzyme was 0.1 mg/mL at a path length of 1.0 mm).

Control Nativel-asparaginase

Lauric acidbioconjugate

Palmitic acidbioconjugate

Behenic acidbioconjugate

Helix 23.11% 30.39% 35.87% 27.63% 28.64%Beta 18.27% 24.21% 0.03% 11.07% 17.82%Turn 35.26% 14.37% 27.72% 21.49% 19.11%

3

10

tcccpaatihh

4

ccesabmmcus

A

BPSsCVD

R

[[

[

[[

[

[[

[[

[[[[

[[[

[[[[

[[

[[

[[

Random 23.36% 31.03%

Total 100.00% 100.00%

endency indices and dispersity indices. In this study, a chemicalonjugate between hydrophilic protein and hydrophobic fatty acidhain was considered which was able to form nanostructured parti-le in aqueous solution. Even so, the increase in biological half- life,roteolysis stability and pH stability of the bioconjugates may bes a result of this nanostructured system. Moreover, because of themphiphilic structure of the bioconjugates, they have the potentialo self-assemble as a micelle in an aqueous media, as reported fornsulin in conjugation with deoxycholic acid [40]. In doing so, theydrophobic fatty acid could constitute the inner core, while theydrophilic part exists as a surrounding corona.

. Conclusion

The aim of this study was to improve the biophysical and bio-hemical characteristics of anti-leukemic enzyme, l-ASNase, byonjugation with different chain length fatty acids. The modifiednzyme showed increased in vitro half-life, promoted proteolysistability and extended the range of optimum pH of activity. Theffinity between l-ASNase and its substrate has increased afterioconjugation. Based on the possible formation of self assemblyicelles by bioconjugates, the nanoscale size is achieved. The CDeasurements suggest that the secondary structure of l-ASNase is

hanges upon lipidation. This method of l-ASNase modification bysing fatty acid indicates a promising stabilized product that mayerve as a new candidate for medical purposes in future.

cknowledgements

The authors would like to thank Dr. M.R. Rouini, Professor ofiopharmaceutics and Pharmacokinetic division, Department ofharmaceutics, Faculty of Pharmacy, Tehran University of Medicalciences for his kindly supports. This work is a part of the Ph.D. the-is in Faculty of Pharmacy and Pharmaceutical Sciences Researchenter, Shiraz University of Medical Sciences (Grant No. 90-5980ice-Chancellery of Research and Technology). Authors also thankr. Hassan Khajeie (Ph.D.), for copy editing of the manuscript.

eferences

[1] M.E.M. Cruz, M.M. Gaspar, F. Lopes, J.S. Jorge, R. Perez-Soler, Int. J. Pharm. 96(1993) 67–77.

[[[[

6.38% 39.81% 34.43%

0.00% 100.00% 100.00%

[2] H.J. Müller, J. Boos, Crit. Rev. Oncol. Hematol. 28 (1998) 97.[3] Y.Q. Zhang, W.L. Zhou, W.D. Shen, Y.H. Chen, X.M. Zha, K. Shirai, K. Kiguchi, J.

Biotechnol. 120 (2005) 315–326.[4] E.E. Haley, G.A. Fischer, A.D. Welch, Cancer Res. 21 (1961) 532–536.[5] R. Wang, B. Xia, B.J. Li, S.L. Peng, L.S. Ding, S. Zhang, Int. J. Pharm. 364 (2008)

102–107.[6] M. Wolf, M. Wirth, F. Pittner, F. Gabor, Int. J. Pharm. 256 (2003) 141–152.[7] A.I. Fernandes, G. Gregoriadis, Int. J. Pharm. (2001) 215–224.[8] M.B.F. Martins, J.C.S. Jorge, M.E.M. Cruz, Biochimie 72 (1990) 671–675.[9] M.J. Poznansky, M. Shandling, M.A. Salkie, Cancer Res. 42 (1982)

1020–1025.10] T.E. Wileman, Adv. Drug Deliv. Rev. 6 (1991) 167–180.11] A.L. Soares, G.M. Guimarães, B. Polakiewicz, R.N.M. Pitombo, J. Abrahão-Neto,

Int. J. Pharm. 237 (2002) 163–170.12] Q. Guoqiang, Z. Juyan, M. Jianbiao, H. Binglin, W. Daobin, Sci. China 40 (1997)

337–341.13] M.R. Tabandeh, M. Aminlari, J. Biotechnol. 141 (2009) 189–195.14] Y.Q. Zhang, M.L. Tao, W.D. Shen, J.P. Mao, Y. Chen, J. Chem. Technol. Biotechnol.

81 (2006) 136–145.15] J.C.S. Jorge, R. Perez-Soler, J.G. Morais, M.E.M. Cruz, Cancer Chemother. Phar-

macol. 34 (1994) 230–234.16] J. António, E. Almeida, Souto, Adv. Drug Deliv. Rev. 59 (2007) 478–490.17] S.A. Wissing, O. Kayser, R.H. Muller, Adv. Drug Deliv. Rev. 56 (2004)

1257–1272.18] A. Seyfoddin, J. Shaw, A.R. Al-Kassas, Drug Deliv. 17 (2010) 467–489.19] G.T. Hermanson, Bioconjugate Techniques, 2nd ed., Elsevier Inc., London, 2008,

pp. 215–220.20] K. Itaya, J. Lipid Res. 18 (1977) 663–665.21] A.F.S.A. Habeeb, Anal. Biochem. 14 (1966) 328–336.22] H.E. Abdellatef, H.M. Khalil, J. Pharm. Biomed. Anal. 31 (2003) 209–214.23] J.C. Wriston Jr., T.O. Yellin, Adv. Enzymol. Relat. Areas Mol. Biol. 39 (1973)

185–248.24] M.M. Bradford, A. Rapid, Anal. Biochem. (1976) 248–254.25] U.K. Laemmli, Nature 227 (1970) 680–685.26] S.S. Jambhekar, P.J. Breen, Basic Pharmacokinetics, 2nd ed., Pharmaceutical

Press, London, 2009.27] T. Maita, K. Morokuma, G. Mastsuda, J. Biochem. 76 (1974) 1351–1354.28] B. Wang, Y. Cao, Eng. Life Sci. 11 (2010) 44–51.29] A. Karsakevich, M. Bekers, J. Mol. Catal. B: Enzym. 11 (2001) 551–558.30] Y.Q. Zhang, M.L. Tao, W.D. Shen, Y.Z. Zhou, Y. Ding, Y. Ma, W.L. Zhou, Biomate-

rials 25 (2004) 3751–3759.31] G.A. Kotzia, K. Lappa, N.E. Labrou, Biochem. J. 404 (2) (2007) 337–343.32] M. Sanches, S. Krauchenco, I. Polikarpov, Curr. Chem. Biol. 1 (2007)

75–86.33] K. Michalska, M. Jaskolski, Acta Biochim. Pol. 53 (2006) 627–640.34] H.R. Costantino, K. Griebenow, P. Mishra, R. Langer, A. Klibanov, Biochim. Bio-

phys. Acta-Protein Struct. Molec. Enzym. 1 (1995) 69–74.35] J.F. Carpenter, M.C. Manning, Pharm. Biotechnol. 13 (2002).36] D.S. Pisal, M.P. Kosloski, S.V. Balu-Iyer, J. Pharm. Sci. 6 (2010) 2557–2575.

37] M.J. Nadolski, M.E. Linder, FEBS J. 274 (2007) 5202–5210.38] B. Ranjbar, P. Gill, Chem. Biol. Drug Des. 74 (2009) 101–120.39] B. Plesner, P. Westh, A.D. Nielsen, Int. J. Pharm. 406 (2011) 62–68.40] S. Lee, K. Kim, T.S. Kumar, J. Lee, S.K. Kim, D.Y. Lee, Y.-K. Lee, Y. Byun, Bioconjug.

Chem. 16 (2005) 615–620.