Journal of Controlled Release 72 (2001) 85–91www.elsevier.com/ locate / jconrel

Pronounced activity of enzymes through the incorporation intothe core of polyion complex micelles made from charged block

copolymers

*Atsushi Harada, Kazunori KataokaDepartment of Materials Science, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,

Tokyo 113-8656, Japan

Received 12 April 2000; accepted 12 January 2001

Abstract

Compartmentalization of enzymes in the nanometric-scaled container, purposing to improve their stability and availability,has recently attracted a strong interest in the field of pharmaceutics. In this study, the enzymatic activity of lysozyme in thecore of polyion complex (PIC) micelles, which were formed from egg white lysozyme and poly(ethylene glycol)–poly(a,b-aspartic acid) block copolymer (PEG-P(Asp)), was evaluated using a colorimetric method. Apparent enzymatic activity oflysozyme entrapped in the core of PIC micelles remarkably increased compared to that of free lysozyme, which is mainlyattributed to a decrease in the observed Michaelis constant (K ). The reciprocal of the K values nicely correlated tom,obs m,obs

the corona thickness of PIC micelles, suggesting that the corona layer of PIC micelle may act as the reservoir of thesubstrate, p-nitrophenyl penta-N-acetyl-b-chitopentaoside. This result indicates that the enzymatic activity can be controlledby changing the corona thickness of PIC micelles through a variation in the mixing ratio of PEG-P(Asp) to lysozyme. Thistype of PIC micelle system entrapping enzyme in the core might be useful for the design of diagnostic as well as targetabletherapeutic systems of enzyme including antibody-directed enzyme prodrug therapy (ADEPT). 2001 Elsevier ScienceB.V. All rights reserved.

Keywords: Polyion complex micelle; Enzymatic reactor; Poly(ethylene glycol); Block copolymer; Enzyme delivery

1. Introduction stimulated research directed toward the modulationof enzymatic action, both in spatial and chronologi-

Continuous progress over the past decade of cal manners, by a pharmaceutical approach. Thisfundamental as well as applied enzymology has includes controlled release and targeting of enzymesremarkably increased knowledge on the structure and using appropriate carrier systems. Since enzymes arefunction of many clinically useful enzymes, and has large-molecular weight polypeptides with a definite

higher ordered structure, they are susceptible toproteolysis and denaturation [1–5]. In order to

*Corresponding author. Tel.: 181-3-5841-7138; fax: 181-3-improve on their problems, extensive research has5841-8653.been devoted for the development of novel formula-E-mail address: kataoka@bmw.mm.t.u-tokyo.ac.jp (K. Kata-

oka). tions of enzyme, focusing on the achievement of

0168-3659/01/$ – see front matter 2001 Elsevier Science B.V. All rights reserved.PI I : S0168-3659( 01 )00264-4

86 A. Harada, K. Kataoka / Journal of Controlled Release 72 (2001) 85 –91

high enzymatic activity or the accumulation of therapeutics. Notably, the apparent enzymatic activi-therapeutic enzymes to specific tissues. The latter is ty of lysozyme entrapped in the core of PIC micellesespecially a critical issue in antibody-directed en- was found to be twice as high as that of freezyme prodrug therapy (ADEPT) [6,7]. One well- lysozyme, which is due to the apparent decrease indocumented approach in this regard is the enzyme the observed Michaelis constant through the sub-conjugation of water-soluble and non-toxic poly- strate condensation into the corona layer of PICmers, particularly poly(ethylene glycol) (PEG) [8– micelles.10]. L-Asparaginase with anti-tumor activity andadenosine deaminase (ADA), which ameliorates thesevere combined immunodeficiency of ADA-defi-

2. Materials and methodscient patients, were successfully modified with PEGto reduce their immunogenicity as well as to achievea longevity in blood circulation [11]. However, 2.1. MaterialsPEGylation through covalent linkage is not availablefor all enzymes, because some proteins decrease Poly(ethylene glycol)–poly(a,b-aspartic acid)their activities via PEGylation due to such reasons as block copolymer (PEG-P(Asp); PEG M 512 000wsteric hindrance and chemical modulation of active g/mol; polymerization degree of P(Asp)515) wassites [5]. In this regard, complexation of enzyme synthesized by a previously described method [18].with PEG-based polymers via non-covalent linkage Chicken egg white lysozyme was purchased frommay be a promising alternative to construct novel Sigma (St. Louis, MO), and used without furtherdosage forms of therapeutic enzymes. purification. p-Nitrophenyl penta-N-acetyl-b-

Recently, we found the formation of a novel and chitopentaoside was purchased from Seikagakukogyonarrowly distributed supramolecular assembly, poly- (Osaka, Japan) and used without further purification.ion complex (PIC) micelle, from the mixture of Na HPO ?12H O and NaH PO ?2H O were reagent2 4 2 2 4 2oppositely charged pair of block ionomers having grade and were purchased from Wako Pure ChemicalPEG as one segment and synthetic and natural Industries (Osaka, Japan).polyelectrolytes, including antisense ODN, plasmidDNA and enzyme as the other [12–18]. Thesepolyelectrolytes were successfully compartmen- 2.2. Preparation of PIC micelles entrappingtalized in the core of a PIC micelle having a PEG lysozyme in the coreouter-shell with a mesoscopic size range (severaltens of nanometers). As for the system entrapping PIC micelles were prepared at various mixingenzyme, egg white lysozyme was intensively used in ratios (r51.000, 1.600, 2.000 and 2.667), where rour research, since its structural and biological was defined as a ratio of the number of aspartic acidproperties are well-characterized and its pI (|11) is residues in the PEG-P(Asp) to the number of ar-high enough to expect complex formation with ginine and lysine residues in lysozyme. Detailedanionic block ionomer, in this case, poly(ethylene mixing procedure is as follows: 100.0 mg of lyso-glycol)–poly(a,b-aspartic acid) block copolymer zyme were dissolved in 12.5 ml of sodium phosphate(PEG-P(Asp)). The detailed physicochemical prop- buffer (PBS, 10 mM, pH 7.4; Na HPO ?12H O,2 4 2

erties including average diameter, core size, corona 2.865 g/ l; NaH PO ?2H O, 0.312 g/ l) to make PBS2 4 2

thickness and the association number were deter- solution of lysozyme. Similarly, PEG-P(Asp) wasmined for PIC micelles formed from lysozyme and dissolved in PBS at various concentration (8.8–23.8PEG-P(Asp) at various mixing ratios [18,23]. mg/ml). An equal volume of lysozyme and PEG-

In this study, the activity of lysozyme entrapped in P(Asp) solutions thus prepared were mixed afterthe PIC micelles was thoroughly investigated to filtration through a 0.1-mm filter (Millex-VV, Milli-explore the potential utility of enzyme-entrapped PIC pore, Bedford, MA) to remove dust. Mixed solutionmicelles as nanometric-scaled enzymatic reactors, was stored overnight at room temperature prior towhich might be useful in the field of diagnostics and following characterization.

A. Harada, K. Kataoka / Journal of Controlled Release 72 (2001) 85 –91 87

2.3. Evaluation of enzymatic activity of lysozyme PIC micelles. On the other hand, it is likely that asubstrate with a considerably low molecular weight

The enzymatic activity of lysozyme was may diffuse into the core of the micelle to receive ancolorimetrically evaluated using p-nitrophenyl- enzymatic reaction. This feasibility was indeed dem-penta-N-acetyl-b-chitopentaoside (NP-(GlcNAc) ) onstrated in Fig. 1 by using p-nitrophenyl-penta-N-5

as the substrate at 25.08C. Different amounts (80, acetyl-b-chitopentaoside (NP-(GlcNAc) ) as a5

120, 160 and 200 mmol /ml) of the substrate solution model substrate. NP-(GlcNAc) releases p-nitro-5

in PBS were prepared as the stock solutions. The phenol through lysozyme-catalyzed cleavage of b-PIC micelle solutions prepared at various mixing 1,4-glycosidic bonds [21,22]. The cumulativeratios, free lysozyme solution and the stock solution amount of released p-nitrophenol was calculatedof the substrate were separately incubated at 25.08C from the change in the absorbance at 340 nm usingfor 30 min before the activity evaluation. The stock molar absorptivity. Obviously, when NP-(GlcNAc)5

solutions of the substrate with varying concentrations was used as the substrate, lysozyme showed appreci-were then mixed with an equal volume of the micelle able enzymatic activity in spite of its segregation intoor free lysozyme solution to start enzymatic reaction. the core of PIC micelles. This suggests that NP-Lysozyme concentration in the reaction system was (GlcNAc) should smoothly diffuse into the core of5

constant (2.0 mg/ml), while substrate concentration PIC micelles and was cleaved by entrapped lyso-was varied by 40, 60, 80 and 100 mmol /ml. The zyme. The core of PIC micelles in this regarddecrease in the absorbance at 340 nm due to the provides nanometric-scaled field for enzymatic re-hydrolysis of the nitrophenyl ester group of the action. Further, the initial hydrolysis rates weresubstrate was monitored at 25.08C using a V500 significantly higher for entrapped lysozyme com-UV/VIS spectrophotometer (Jasco, Tokyo, Japan). pared to the one in the free form. The apparent

relative activity was then defined as the ratio of an2.4. Circular dichroism measurements initial rate of entrapped lysozyme to that of free

Circular dichroism (CD) measurements were car-ried out using a 1-mm cell (GL Science, Tokyo,Japan) and a J-600 spectropolarimeter (Jasco) atroom temperature. Samples were prepared in PBS tohave 0.01 wt.% of lysozyme.

3. Results and discussion

It is well-known that lysozyme hydrolyses theb-1,4-glycosidic bond between N-acetylmuramicacid and N-acetylglucosamine [19]. This is the basisof the lytic activity of lysozyme toward microorga-nisms such as Micrococcus luteus, which has theb-1,4-glycosidic bond between N-acetylmuramicacid and N-acetylglucosamine at the surface of itscell wall. Our previous study revealed that lysozyme-induced lysis of Micrococcus luteus can be success-

Fig. 1. Release profile of p-nitrophenol due to the hydrolysis offully regulated in an on–off manner, synchronizingNP-(GlcNAc) by lysozyme. (s) Free lysozyme; (d, m, j, ♦)5with the micellization and dissociation of lysozymelysozyme incorporated into PIC micelles of r51.000, 1.600, 2.000

with PEG-P(Asp) [20]. This is because a cellular and 2.667, respectively; temperature, 25.08C; NP-(GlcNAc)5substrate such as Micrococcus luteus is not able to concentration, 100 mmol /ml; lysozyme concentration, 0.14 mmol /directly access lysozyme entrapped in the core of ml; monitoring wavelength of absorbance, 340 nm.

88 A. Harada, K. Kataoka / Journal of Controlled Release 72 (2001) 85 –91

Fig. 3. The double reciprocal plots of the initial rate of p-nitrophenol release and the substrate concentration for free and

Fig. 2. Apparent enzymatic activity of free and micelle-incorpo-micelle-incorporated lysozyme. (s) Free lysozyme; (d, m, j, ♦)

rated lysozyme. The apparent activity is defined as the relativelysozyme incorporated into PIC micelles of r51.000, 1.600, 2.000

initial rate of p-nitrophenol release shown in Fig. 1, taking theand 2.667, respectively; temperature, 25.08C.

initial rate for free lysozyme as control (r50.000); temperature,25.08C.

the slope corresponds to K /V . Then, K can bem max m

lysozyme, and is given in Fig. 2 for the system with calculated from the ratio of the slope to the intercept.varying mixing ratio, r. It is obvious to see an Observed K and V (K and V ) werem max m,obs max,obs

increased apparent enzymatic activity compared to summarized in Table 1. Although the apparentfree lysozyme for all of the PIC micelle system enzymatic activity of lysozyme as shown in Fig. 2regardless of r value. There is a trend of increasing remarkably increased by the incorporation into theactivity with an increase in r value, and particularly, core of PIC micelles, V increased only 10%max,obs

lysozyme in the core of PIC micelles prepared at compared to free lysozyme. On the other hand,r52.667 exhibited twice-higher activity compared to K drastically decreased through the incorpora-m,obs

free lysozyme. tion into the core of PIC micelles. This resultMore detailed experiments were then carried out suggests that a change in K value may play am,obs

in order to get an insight into the mechanism of anTable 1increase in apparent activity of lysozyme with theThe observed kinetic constants for hydrolysis of NP-(GlcNAc)5incorporation into the core of PIC micelle. Toby free and micelle-incorporated lysozyme

determine kinetic constants of enzymatic reaction,a b a b cMixing K (relative) V (relative) km,obs max,obs 3the double reciprocal plot was used as follows: 24 28 24ratio (r) (10 mol / l) (10 mol / s) (10 l / s)

0.000 1.464 (1.000) 4.958 (1.000) 3.541 /v 5 (K /V ) (1 / [s]) 1 1/V (1)m max max1.000 0.6368 (0.434) 5.439 (1.097) 3.891.600 0.4390 (0.300) 5.433 (1.096) 3.88

where K is Michaelis constant, V is maximumm max 2.000 0.3734 (0.255) 5.431 (1.095) 3.88velocity, v is initial rate and [s] is the concentration 2.667 0.2912 (0.199) 5.429 (1.095) 3.88of the substrate. Fig. 3 shows the double reciprocal a These values were determined from Fig. 3 based on Eq. (1).

bplots for free and micelle-incorporated lysozyme. In These values were relative values to free lysozyme.cthese plots, the vertical intercept gives 1 /V and These values were calculated from V using Eq. (4).max,obsmax

A. Harada, K. Kataoka / Journal of Controlled Release 72 (2001) 85 –91 89

critical role in an increased enzymatic activitythrough micellization.

This apparent correlation between a decreasedK value and an incorporated enzymatic activitym,obs

upon the micellization is further confirmed byanalyzing the kinetics of enzymatic reaction based onthe following equation:

k1 k3→E 1 S ES →E 1 products (2)←k2

where E is enzyme, S is substrate and ES is thecomplex of E and S. In this mechanism, the K andm

V values are expressed as follows:max

K 5 (k 1 k ) /k (3)m 2 3 1

V 5 k [E] (4)max 3

where [E] is the whole concentration of enzyme inthe system, which is the summation of the con-centration of free enzyme and the complexed enzymein Eq. (2). As summarized in Table 1, there wasnegligible change in V for lysozyme incorpo-max,obs

rated into the core of PIC micelles. Further, theconcentration of lysozyme kept constant (2.0 mg/ml(0.14 mmol /ml)) in this experimental condition.Consequently, the k values for micelle-incorporated3

lysozyme should be independent on the mixing ratio(r) as from Eq. (4), and were calculated to have a

24constant value of 3.88310 l / s as shown in Table1 using Eq. (4) with lysozyme concentration of 0.14mmol /ml and the corresponding V values givenmax,obs

in the table. Since the k value was constant, the3

change in the K value should be mainly due tom,obs

the change in the k and/or k values according to1 2

Eq. (3). If this is the case, there should be aconsiderable change in the structure of lysozymethrough the complexation with PEG-P(Asp). Circulardichroism (CD) measurements were then carried out Fig. 4. Circular dichroism spectra for free (a) and micelle-in-

corporated (b) lysozyme. PIC micelles were prepared at r52.667.to compare the secondary structure of free andmicelle-incorporated lysozyme, allowing to estimateany change in lysozyme structure caused by an Thus, CD spectra solely reflect the structure ofinteraction with PEG-P(Asp). Fig. 4 shows CD lysozyme, and were identical for micelle-incorpo-spectra of free and micelle-incorporated lysozyme rated and free lysozymes. This indicates that the(r52.667). Note that P(Asp) segment in PEG-P(Asp) incorporation into the core of PIC micelles inducedused here has no detectable band in CD spectrum no detectable change in the secondary structure of(data not shown), because it is racemized and lysozyme, suggesting a negligible change in the realconsists of a and b isomers of aspartic acid units. K value. Consequently, it is unlikely that a changem

90 A. Harada, K. Kataoka / Journal of Controlled Release 72 (2001) 85 –91

in the K value through micellization is derivedm,obs

from an essential change in the kinetic constants (k1

and/or k values).2

Another feasible reason for a decreased Km,obs

value in the PIC micelle system is that the con-centration of the local substrate may be higher in themicelle phase, because the corona of the micelleconsists of highly dense PEG strands, which mayhave intermolecular interaction with NP-(GlcNAc)5

including hydrogen bonding. Obviously, a condensa-tion of the substrate in the vicinity of enzyme resultsin a shift in the apparent kinetic parameters, in thiscase, a decrease in K values.m,obs

The physicochemical properties including averagediameter, the critical association concentration, coresize and corona thickness of PIC micelles preparedfrom lysozyme and PEG-P(Asp) were previouslyreported in our publication [23]. Worth mentioning isthat the PIC micelles from lysozyme and PEG- Fig. 5. Relationship between the 1/K values and the coronam,obsP(Asp) satisfy the stoichiometry at the mixing ratio thickness of PIC micelles. The values of the corona thickness were(r) of unity: 1 to 1 molar ratio of the number of refereed from our previous paper [23].

aspartic acid residues in the PEG-P(Asp) and thenumber of arginine and lysine residues in lysozyme.Micelles at this stoichiometric condition were shown function of the mixing ratio (r) of PEG-P(Asp) toto have ca. 50 nm of the average diameter with lysozyme. From this result, it was suggested thatextremely narrow size distribution. The average remarkable increase in apparent enzymatic activitydiameter increased from 50 to 70 nm with an might be induced by a control of affinity between theincrease in the r value from 1.000 to 2.667, while the substrate and the corona-forming segment.number of lysozyme in the core as well as the coreradius kept constant: ca. 50 molecules of lysozymein the core with ca. 7 nm of core radius. Eventually, 4. Conclusionthe increment of average diameter with an increasedr value should be mainly due to an increase in the The enzymatic activity of lysozyme located in thecorona thickness accompanied with an increased core of PIC micelles was evaluated by a colorimetricassociation number of PEG segments, inducing the method using NP-(GlcNAc) as the substrate. The5

more stretched conformation of PEG segments in the apparent enzymatic activity of lysozyme in the corecorona layer [23]. was remarkably higher than that of free lysozyme.

Of interest, as seen in Fig. 5, the corona thickness, By the determination of kinetic constants of en-which was evaluated using light scattering tech- zymatic reaction, it was confirmed that such increaseniques, was found to have a nice linear correlation in the enzymatic activity was induced by the sub-with 1/K suggesting that K would be de- strate condensation into the corona layer. Thus, them,obs m,obs

termined by the corona thickness due to the substrate apparent activity of lysozymes was able to becondensation into the corona phase. Further, since an controlled by the corona thickness. This result ofincrease in apparent enzymatic activity shown in Fig. controlling enzymatic activity through the formation2 was in line with the change in K , it is assumed of supramolecular assembly might provide a basis tom,obs

that the apparent enzymatic activity of PIC micelle design delivery system of enzymes, including forsystem might be controlled by the substrate con- ADEPT, and novel diagnostic formulation of en-densation in the corona phase, whose thickness is a zymes.

A. Harada, K. Kataoka / Journal of Controlled Release 72 (2001) 85 –91 91

charged block copolymers with poly(ethylene glycol) seg-Acknowledgementsments, Macromolecules 18 (1995) 5294–5299.

[13] A. Harada, K. Kataoka, Formation of stable and monodisper-This work was financially supported by a Grant- sive polyion complex micelles in aqueous medium from

in-Aid for Scientific Research on Priority Area (A) poly(L-lysine) and poly(ethylene glycol)-poly(aspartic acid)‘Molecular Synchronization for Design of New block copolymer, J. Macromol. Sci. Pure Appl. Chem. A34

(1997) 2119–2133.Materials System’ (No. 11167210) from the Ministry[14] A. Harada, K. Kataoka, Chain length recognition: core-shellof Education, Science, Sports and Culture, Japan.

supramolecular assembly from oppositely charged blockcopolymers, Science 283 (1999) 65–67.

[15] K. Kataoka, H. Togawa, A. Harada, K. Yasugi, T. Mat-References sumoto, S. Katayose, Spontaneous formation of polyion

complex micelles with narrow distribution from antisenseoligonucleotide and cationic block copolymer in physiologi-[1] M.C. Manning, K. Patel, T. Borchardt, Stability of proteincal saline, Macromolecules 29 (1996) 8556–8557.pharmaceuticals, Pharm. Res. 6 (1989) 903–918.

[16] S. Katayose, K. Kataoka, Water-soluble polyion complex[2] Y.J. Wang, A. Hanson, Parenteral formulations of proteinsassociates of DNA and poly(ethylene glycol)-poly(L-lysine)and peptides: stability and stabilizers, J. Parent. Sci. Technol.block copolymer, Bioconjug. Chem. 8 (1997) 702–707.Suppl. 42 (1988) S3–S26.

[17] S. Katayose, K. Kataoka, Remarkable increase in nuclease[3] A.K. Banga, Y.W. Chien, Systematic delivery of therapeuticresistance of plasmid DNA through supramolecular assemblypeptides and proteins, Int. J. Pharm. 48 (1988) 15–50.with poly(ethylene glycol)-poly(L-lysine) block copolymer, J.[4] S.S. Davis, Delivery systems for biopharmaceuticals, J.Pharm. Sci. 87 (1998) 160–163.Pharm. Pharmacol. 44 (1992) 186–190.

[18] A. Harada, K. Kataoka, Novel polyion complex micelles[5] W.R. Gombotz, D.K. Pettit, Biodegradable polymers forentrapping enzyme molecules in the core: preparation ofproteins and peptides drug delivery, Bioconjugate Chem. 6narrowly-distributed micelles from lysozyme and poly-(1995) 332–351.(ethylene glycol)-poly(aspartic acid) block copolymer in[6] R.F. Sherwood, Advanced drug delivery reviews:enzymeaqueous medium, Macromolecules 31 (1998) 288–294.prodrug therapy, Adv. Drug Deliv. Rev. 22 (1996) 269–288.

[19] A.R. Rees, M.J.E. Sternberg, in: From Cells to Atoms,[7] I. Niculescu-Duvaz, C.J. Springer, Antibody-directed en-Blackwell Scientific Publications, London, 1984.zyme prodrug therapy (ADEPT): a review, Adv. Drug Deliv.

[20] A. Harada, K. Kataoka, On-off control of enzymatic activityRev. 26 (1997) 151–172.synchronized with reversible formation of supramolecular[8] M.J. Knauf, D.P. Bell, P. Hirtzer, Z.P. Luo, J.D. Young, N.V.assembly from enzyme and charged block copolymers, J.Katre, Relationship of effective molecular size to systemicAm. Chem. Soc. 121 (1999) 9241–9242.clearance in rats of recombinant interleukin-2 chemically

[21] T. Osawa, Y. Nakazawa, Lysozyme substrates. Chemicalmodified with water-soluble polymers, J. Biol. Chem. 263synthesis of p-nitrophenyl O-(2-acetamido-2-deoxy-b-D-(1988) 15064–15070.glucopyranosyl)-(1→4)-O-(2-acetamido-2-deoxy-b-D-gluco-[9] N.V. Katre, Immunogenicity of recombinant IL-2 modified bypyranosyl)-(1→4)-2-acetamido-2-deoxy-b-D-glucopyrano-covalent attachment of polyethylene glycol, J. Immunol. 144side and its reaction with lysozyme, Biochim. Biophys. Acta(1989) 209–213.130 (1966) 56–63.[10] C. Delgado, G.E. Francis, D. Fisher, The uses and properties

[22] M.A. Raftery, T. Rand-Meir, Distinguishing between pos-of PEG-linked proteins, Crit. Rev. Ther. Drug Carrier Syst. 9sible mechanistic pathways during lysozyme-catalyzed cleav-(1992) 249–304.age of glycosidic bonds, Biochemistry 7 (1968) 3281–3289.[11] Y. Kodera, T. Sekine, T. Yasukohchi, Y. Kiriu, M. Hiroto, A.

[23] A. Harada, K. Kataoka, Novel polyion complex micellesMatsushita, Y. Inada, Stabilization of L-asparaginase modifiedentrapping enzyme molecules in the core. 2. Characterizationwith comb-shaped poly(ethylene glycol) derivatives, in vivoof the micelles prepared at nonstoichiometric mixing ratios,and in vitro, Bioconjug. Chem. 5 (1994) 283–286.Langmuir 15 (1999) 4208–4212.[12] A. Harada, K. Kataoka, Formation of polyion complex

micelles in an aqueous milieu from a pair of oppositely