Interaction of pHPMA–pLMA Copolymers with Human Blood Serum

Article pubs.acs.org/molecularpharmaceutics

Interaction of pHPMA−pLMA Copolymers with Human Blood Serum and Its Components Mirjam Hemmelmann,†,‡ Kristin Mohr,‡,§ Karl Fischer,§ Rudolf Zentel,*,† and Manfred Schmidt*,§ †

Institute of Organic Chemistry, Johannes Gutenberg University Mainz, Duesbergweg 10-14, 55099 Mainz, Germany Institute for Physical Chemistry, University of Mainz, Welder Weg 11, D-55099 Mainz, Germany

§

S Supporting Information *

ABSTRACT: Immediately after administration, polymer therapeutics are exposed to complex biological media like blood which may influence and alter their physicochemical properties due to interactions with proteins or serum components. Among such interactions those leading to larger sized aggregates can be sensitively detected by dynamic light scattering (DLS) as a pre in vivo screening method. Random copolymers from N-(2-hydroxypropyl)methacrylamide and lauryl methacrylate p(HPMA-co-LMA) and copolymers loaded with the model drug domperidone were characterized by DLS in isotonic salt solution and in blood serum. The bare amphiphilic copolymer micelles (Rh = 30 nm in isotonic salt solution) formed large aggregates in serum of over 100 nm radius which were shown to originate from interactions with very low density lipoproteins (VLDLs). Encapsulation of the hydrophobic drug domperidone resulted, at first, in drug− copolymer formulations with larger hydrodynamic radii (39 nm < Rh < 49 nm) which, however, did not induce aggregate formation in human serum. Since p(HPMA-co-LMA) copolymers were demonstrated to have a high potential for drug delivery into the brain, the knowledge of serum−copolymer interactions provides a better understanding of their function in the biological context. KEYWORDS: HPMA based copolymers, polymer therapeutics, amphiphilic copolymers, blood polymer interactions, dynamic light scattering, protein polymer complex, human serum

■

INTRODUCTION In the past several decades the fast advancing field of nanomedicines has shown great potential for diagnostic and therapeutic applications.1,2 Many polymeric materials for drug delivery have been developed and explored, especially as polymer−drug conjugates, a concept that was first introduced by Helmut Ringsdorf in the 1970s.3 The general idea is that water-soluble, nontoxic, and nonimmunogenic polymers can improve the body distribution and biocompatibility of low molecular weight drugs, which are either encapsulated or covalently bound to them. A number of polymeric therapeutics, especially those for the treatment of cancer are already in clinical use or in different stages of preclinical and clinical evaluation.2,4 For a rational design it should be considered that polymer therapeutics face many challenges during their lifecycle beginning with administration, which leads to immediate contact with complex biological media, and proceeding to body distribution, metabolism, and elimination of the polymer.5 Thus a basic understanding and control over the interactions of polymeric systems with blood and its componentsand especially the formation of larger aggregatesis one of the major challenges, which polymer therapeutics face for successful clinical use.6 In 2009, the FDA stated the high importance to sensitively detect physical properties of nanomedicines upon absorption (administration), distribution, © 2013 American Chemical Society

metabolism, and excretion (ADME model). The necessity to characterize polymeric therapeutics in physiologically relevant media is discussed by the FDA, since major changes in particle size and surface properties can be induced by interactions with surrounding highly concentrated protein solutions.2 In general, great care is taken to control particle size, size distribution, shape, surface properties, and aggregation state of polymeric drug delivery systems in aqueous solutions, since the close relationship between body distribution, interaction with the immune system, and physicochemical properties of polymer therapeutics has become increasingly apparent.5,7,8 In addition, it is important to consider that these physicochemical properties may be altered due to various and subtle interactions with proteins or other components of the respective body fluids depending on the route of administration. Among those interactions, the ones leading to larger size aggregates are considered to be particularly important. For many medical applications in therapy and diagnosis, polymers are administered intravenously and thereby enter the bloodstream directly.5 Other administration routes, like intraperitoneal or subcutaneous injections, also result in the release Received: Revised: Accepted: Published: 3769

May 1, 2013 July 16, 2013 August 19, 2013 September 4, 2013 dx.doi.org/10.1021/mp400254b | Mol. Pharmaceutics 2013, 10, 3769−3775

Molecular Pharmaceutics

Article

−18 °C. 2,20-Azobis(isobutyronitrile) (AIBN) was recrystallized from diethyl ether and stored at −18 °C. Human albumin was obtained from Sigma Aldrich. Characterization. 1H NMR spectra were obtained by a Bruker AC 300 spectrometer at 300 MHz. 19F NMR analysis was carried out with a Bruker DRX-400 at 400 MHz. All measurements were carried out at room temperature, and spectroscopic data were analyzed using ACDLabs 9.0 1D NMR manager and MestReNova. The reactive ester polymers were dried overnight at 40 °C under vacuum and afterward analyzed via gel permeation chromatography (GPC). GPC was performed in tetrahydrofuran (THF) as solvent and equipped with pump PU 1580, autosampler AS 1555, UV detector UV 1575, and RI detector RI 1530 from Jasco as well as a miniDAWN Tristar light scattering detector from Wyatt. Columns used were from MZ Analysentechnik: MZ-Gel SDplus 106 Å 5 μm, MZ-Gel SDplus 104 Å 5 μm, and MZGel SDplus 102 Å 5 μm. GPC data were evaluated using the software PSS WinGPC Unity from Polymer Standard Service Mainz. The flow rate was set to 1 mL/min with a temperature of 25 °C. General Synthetic Route of Statistic Copolymers. The synthesis of the here described polymers was performed in analogy to the literature18,16,20 with some alterations (see Supporting Information). Preparation of Blood Serum. The human blood serum used for this study was obtained from the University Clinic of Mainz (Germany). It was prepared according to the standard guidelines of the University Clinic Mainz. Because of the high variation of protein composition of different patients12 a pool of serum obtained by the mixture of serum of seven healthy donors was used for all measurements. Preparation of VLDL and LDL Solutions. Human low density lipoproteins (LDL) (1.025 < d < 1.050 g/mL) and very low density lipoproteins (VLDL) (d < 1.006 g/mL) were isolated from human plasma of normolipidemic adults by sequential ultracentrifugation.19 Preparation of Samples. All solutions for light scattering experiments are prepared in a dust free flow box. Cylindrical quartz cuvettes (20 mm diameter, Hellma, Müllheim) are cleaned by dust-free distilled acetone. Serum solutions are filtered through Milex GS filters, 220 nm pore size (Millipore). After testing of several filters, concentration losses of serum proteins by filtration with Milex GS filters are negligible.12 The polymer formulations are prepared by dissolving polymer and domperidone (insoluble in water) in DMSO and shaking the solution for five minutes. The mixture is then added dropwise to the buffer solution and purified by GPC (Sephadex, water). Analysis via GPC revealed thereafter that there is no unbound domperidone left. We also observe an immediate precipitation of domperidone when we overload (more than 60 wt %) the copolymer with this drug. For light scattering the p(HPMA-co-LMA) as well as the p(HPMA-coLMA) domperidone complex is prepared in DPBS buffer solution (GIBCO, Invitrogen). The polymer is filtered through an LCR450 nm filter in the light scattering cuvette. For the measurements of p(HPMA-co-LMA) with and without domperidone, serum and polymer dissolved in DPBS buffer (GIBCO, Invitrogen) are given subsequently into the light scattering cuvette. Then, the cuvettes are incubated for 20 min on a shaker at room temperature before measurement. For the comparability of light scattering data with in vivo experiments,

of drugs from the injection site into the bloodstream, as was shown for p(N-(2-hydroxypropyl)methacrylamide) (pHPMA) based copolymers.9,10 Consequently, it is highly important to investigate the interactions of polymers with the multicomponent system blood11,12 and in particular the potential formation of larger aggregates as a consequence of polymer− serum interactions. As the majority of nanomaterials are immediately coated by proteins upon contact with biological matrices such as blood, it is indispensable to study how the amount and identity of proteins binding to the polymer influence the pharmacokinetics, body distribution, blood circulation time, and toxicity of the investigated systems.11 An example for the influence of interactions with blood components on organ distribution is the adsorption of apolipoproteins (Apo) on polybutylcyanoacrylate nanoparticles (PBCA NPs) coated with Tween80.13 It was shown that a preferential adsorption of ApoE and most likely ApoA-I and ApoA-IV enables passage over the blood brain barrier (BBB) via receptor mediated endocytosis.14 As shown recently, we have developed a novel method to sensitively detect the change of size of polymers and polymer formulations in human blood serum.12 Via dynamic light scattering (DLS) the size distribution of human blood serum could be measured and changes in protein and polymer size could be identified. By applying several fractionation techniques isolated protein fraction solutions may also be investigated in respect to their ability to form aggregates in contact with the polymeric systems. The identification of the aggregation inducing serum components may also allow elucidation of the targeting profile of nanoparticles in biological systems. In recent years, p(N-(2-hydroxypropyl)methacrylamide) (pHPMA) based copolymers modified with randomly distributed hydrophobic lauryl methacrylate (LMA) were evaluated toward their potential for medical applications.15,16 There is evidence that amphiphilic p(HPMA-co-LMA) copolymers are able to mediate transport of drugs which are substrates of efflux pumps into the brain in vitro17 as well as in vivo.18 For the in vivo study domperidone, a dopamine D2 antagonist that cannot pass the BBB served as a model drug to demonstrate the transport of copolymer encapsulated drug across the BBB. In the present study, the aggregation behavior of p(HPMAco-LMA) copolymers and copolymers loaded with the hydrophobic model drug domperidone is investigated as a result of the interaction with human serum and with isolated serum components such as lipoproteins or albumin. DLS measurements in human serum12 are utilized to monitor aggregate formation as one of the major consequences of protein adsorption on the biophysical properties of p(HPMA-co-LMA) copolymers. The aggregation profile of the polymer with human blood serum was found to change upon encapsulation of a hydrophobic drug into p(HPMA-co-LMA) copolymer micelles. Furthermore, the serum components, which induce the formation of aggregates, were identified by control experiments with single serum components.

■

EXPERIMENTAL SECTION Materials. All chemicals/solvents were of reagent/analytical grade, as obtained from Sigma Aldrich and Acros Organics. Pentafluorophenol was obtained from Fluorochem (Great Britain, U.K.). Dioxane was distilled over a sodium/potassium composition prior to use. Lauryl methacrylate was distilled under reduced pressure to remove the stabilizer and stored at 3770

dx.doi.org/10.1021/mp400254b | Mol. Pharmaceutics 2013, 10, 3769−3775

Molecular Pharmaceutics

Article

Figure 1. Molecular structure of p(HPMA-co-LMA) copolymer.

corona.15,20 In the hydrophobic compartments hydrophobic drugs like domperidone, which is not water-soluble, can be encapsulated and thereby solubilized in aqueous solutions.18 Analysis via GPC (Sephadex, water) revealed that there is no unbound domperidone. Two different polymer formulations with either 10 wt % or 50 wt % domperidone relative to the weight of the copolymer were investigated. Aggregation Induced by Interaction with Human Blood Serum. As described previously, aggregation (and its change) of polymeric nanosystems in contact with human serum may be monitored by DLS.12 The electric field autocorrelation functions (ACF) of the complex mixture of proteins and solutes in the human serum are well described by the sum of three exponentials:

the ratio between number of particles and total protein content was kept constant. Light Scattering Apparatus. All light scattering experiments were performed with an instrument consisting of a HeNe laser (632.8 nm, 25 mW output power), an ALV-CGS 8F SLS/DLS 5022F goniometer equipped with eight simultaneously working ALV 7004 correlators, and eight QEAPD Avalanche photodiode detectors. All correlation functions were measured in the cross-correlation mode at scattering angles between 30° and 150° in steps of 30°.

■

RESULTS Amphiphilic p(HPMA-co-LMA) copolymers show good potential as a polymer carrier system for brain drug delivery. They are synthesized by controlled radical polymerization of reactive ester monomers with lauryl methacrylate (LMA) monomers (see Supporting Information). In a second step, the reactive ester groups are aminolyzed to yield the amphiphilic copolymer p(HPMA-co-LMA)16,18,20 with 10 mol % of hydrophobic lauryl side chains. Their structure, molecular weight, and polydispersity are shown in Figure 1 plus Table 1,

⎛ t ⎞ ⎛ t ⎞ ⎟⎟ g1,s(t ) = a1,s exp⎜⎜ − ⎟⎟ + a 2,s exp⎜⎜ − ⎝ τ1,s ⎠ ⎝ τ2,s ⎠ ⎛ t ⎞ + a3,s exp⎜⎜ − ⎟⎟ ⎝ τ3,s ⎠

with the amplitudes ai and the decay times τi = 1/(q2Di), q being the absolute value of the scattering vector (q = 4πn sin(θ/2)/λ0) and Di the Brownian diffusion coefficient of component i. The obtained ACFs for the p(HPMA-co-LMA) copolymers and for the two formulations with either 10 wt % or 50 wt % domperidone encapsulated by the copolymer micelles, respectively, in aqueous solutions are well fitted by a sum of two exponentials.

Table 1. Characterization of p(HPMA-co-LMA) Copolymers HPMA:LMA ratioa

Mn(RE) (g/mol)b

90:10

23000

a

Mw(RE) (g/mol)b

Mn (g/mol)c

PDIb

27000

14000

1.18

1

Monomer ratio determined by H NMR spectroscopy after postpolymerization modification. bDetermined by GPC in THF as solvent for the reactive ester polymers. cCalculated from the molecular weights of the reactive ester precursor polymers as determined by GPC in THF as solvent.

⎛ t ⎞ ⎛ ⎞ ⎟⎟ + a 2,np exp⎜⎜ − t ⎟⎟ g1,np(t ) = a1,np exp⎜⎜ − ⎝ τ1,np ⎠ ⎝ τ2,np ⎠

and the respective characterization of the copolymer and copolymer−domperidone formulation is given in Table 2. In aqueous solutions the amphiphilic copolymers of hydrophilic pHPMA modified with randomly distributed hydrophobic lauryl methacrylate (LMA) self-assemble into aggregates with hydrophobic compartments stabilized by a hydrophilic

Rh (nm)a

P(HPMA-co-LMA) P(HPMA-co-LMA) 50 w% DOM P(HPMA-co-LMA) 10 w% DOM

5 ± 0.2 (Rh1 = 4; Rh2 = 30) 49 ± 2 39 ± 2

(2)

The inverse z-average hydrodynamic radii, Rh = (⟨1/Rh⟩z)1/2, of the pHPMA-co-pLMA copolymer as well as of the pHPMA-copLMA domperidone complexes are listed in Table 2. Usually, a cumulant fit or the initial slope of a biexponential fit may be utilized to derive the z-average diffusion coefficient for monomodal broad distributions of decay times as well as for bimodal distributions. Both cases can be distinguished on the basis of DLS correlation functions only, if the decay times of the bimodal distribution are sufficiently different, where the quantification of “sufficiently” may vary for different investigators. Being conservative in the present work a bimodal distribution is only evident for bare nonloaded copolymer micelles for which two q2-dependent relaxation processes are clearly visible by eye yielding Rh = 4 nm expected for the radius of single copolymer chains and a micelle or aggregate radius of Rh = 30 nm. It may be anticipated that an equilibrium exists between single polymer chains and bigger aggregates, which are

Table 2. Hydrodynamic Radii of p(HPMA-co-LMA) Copolymer and the Formulations with 50 and 10 wt % Domperidone in DPBS polymer formulation

(1)

a Inverse z-average hydrodynamic radii (Rh = (⟨1/Rh⟩z)1/2 in DPBS solution with c(polymer) = 0.1 g/L for all formulations and c(domperpidone) = 0.05 g/L (50 wt %) and 0.01 g/L (10 wt %).

3771

dx.doi.org/10.1021/mp400254b | Mol. Pharmaceutics 2013, 10, 3769−3775

Molecular Pharmaceutics

Article

formed due to hydrophobic interactions as was postulated in a recent study on the aggregation behavior of amphiphilic p(HPMA-co-LMA) copolymers.20 For the copolymer domperidone formulation hydrodynamic radii of Rh = 49 nm (50 wt % Dom) and Rh = 39 nm (10 wt % Dom) were found and no single polymer chains were detectable anymore. Most likely, the binding of domperidone increases the hydrophobicity of the copolymers, thus inducing more extensive micelle formation, i.e., the equilibrium is shifted toward the formation of aggregates With the correlation functions of the serum components (eq 1) and the respective copolymer or copolymer formulation with domperidone as reference fit functions, the correlation functions of the polymer serum mixtures were analyzed. If no or negligible aggregation takes place, the resulting ACF of the serum copolymer mixture should be perfectly fitted by the sum of the individual correlation functions, with the known parameters of the two components kept fixed and the intensity contribution of serum and polymer, fs and f np, being the only fit parameters (eq 3). g1,m(t ) = fs g1,s(t ) + fnp g1,np(t )

Figure 2. ACF of p(HPMA-co-LMA) in serum, violet line, fit with eq 3 and the resulting residue. Green line, fit with eq 4 and the resulting residue, ■ data points of the ACF. Scattering angle 60°.

serum is very high, only significant changes may be accessible by DLS. Identification of the Aggregation Inducing Serum Components. In order to determine the serum component that induces the aggregation with the p(HPMA-co-LMA) copolymers, DLS was performed in solutions of isolated human serum albumin (HSA), very low density lipoprotein (VLDL), low density lipoproteins (LDL), and VLDL depleted human serum. No aggregate formation could be detected with albumin, the most abundant protein in human serum, and with LDL lipoprotein (see Supporting Information). However, incubation of the copolymer in a concentrated VLDL lipoprotein solution resulted in aggregate formation with a size of Rh = 109 nm (Figure 4a). It is believed that amphiphilic copolymers are attracted to the carriers of fatty acids and cholesterol most probably via hydrophobic interactions. In a control experiment, the copolymer was incubated in VLDL depleted serum and no aggregation was observed (Figure 4b). The respective experiments with the p(HPMA-co-LMA) domperidone formulations revealed no aggregation with any of the serum components and the loaded polymers. This is consistent with the lack of aggregation in human serum as described above (see Supporting Information).

(3)

If additional larger sized aggregates are being formed by interactions of drug particle and serum protein components, eq 3 has to be modified. This would be typically indicated by an additional longer ACF relaxation time related to the size of these aggregates (eq 4). g1,m(t ) = fs g1,s(t ) + fnp g1,np(t ) + fagg g1,agg (t )

(4)

with fagg the intensity contribution of the aggregates and the unknown correlation function of the aggregates (eq 5). ⎛ t ⎞⎟ g1,agg = a1,agg exp⎜⎜ − ⎟ ⎝ τ1,agg ⎠

(5)

The concentration of the polymer in human serum with 0.1 mg of copolymer in 1 mL of serum is chosen to be comparable to the calculated values in recently published in vivo studies, in which the copolymer showed good potential as drug carrier for drug delivery over the BBB.18 The autocorrelation function (ACF) of the copolymer− serum mixture indeed exhibits an additional relaxation process at longer times indicating the formation of larger aggregates with a radius of Rh,agg = 160 nm (Figure 2). In contrast, neither of the two p(HPMA-co-LMA) domperidone loaded micelles with either 10 wt % or 50 wt % domperidone show aggregation formation in serum (Figure 3). Even after 24 h no aggregation in human serum could be detected (see Supporting Information). It should be noted that the applied DLS method is extremely sensitive for the detection of high molar mass or larger sized aggregates being formed in complex multicomponent serum mixtures. The term “larger sized” refers to sizes larger than the largest component present in either pure serum or polymer solution. Changes within the size distribution of serum and polymer are detectable only if the amplitudes (i.e., the intensity fraction) of the newly formed sizes are sufficiently large. Preliminary tests (data not shown) typically revealed intensity fractions between 3% and 20% of the newly formed particles to be necessary in order to become detectable by the fitting procedure described above. Since the scattering intensity of

■

DISCUSSION DLS experiments in aqueous media show that amphiphilic p(HPMA-co-LMA) copolymers self-assemble into aggregates with hydrodynamic radii around Rh = 30 nm composed of hydrophobic patches surrounded by a hydrophilic shell of pHPMA which coexist with single dissolved polymer chains with Rh = 4 nm. Upon encapsulation of hydrophobic domperidone into the copolymer the detected hydrodynamic radius of the aggregate increases and no single polymer chains are detectable anymore (Table 2). Probably the hydrophobic interactions between domperidone and the hydrophobic lauryl side chains stabilize the formed particles and all polymer chains are recruited to solubilize the hydrophobic drug. The difference between the amphiphilic copolymer alone and the copolymer−domperidone formulations (filling of the hydrophobic patches of the copolymer aggregates with a hydrophobic drug) results in completely different aggregation profiles with selected serum components. In human serum p(HPMA-co-LMA) copolymers clearly exhibit strong interactions with serum components leading to additional 3772

dx.doi.org/10.1021/mp400254b | Mol. Pharmaceutics 2013, 10, 3769−3775

Molecular Pharmaceutics

Article

Figure 3. ACF of p(HPMA-co-LMA) loaded with (a) 50% domperidone and (b) 10% domperidone in serum, violet line, fit with eq 3 and the resulting residue, ■ data points of the ACF. Scattering angle 60°.

Figure 4. ACF of p(HPMA-co-LMA) (a) in concentrated VLDL solution, violet line, fit with eq 3 and the resulting residue. Green line, fit with eq 4 and the resulting residue, ■ data points of the ACF. Scattering angle 60°. (b) In VLDL depleted human plasma, violet line, fit with eq 3 and the resulting residue, ■ data points of the ACF. Scattering angle 60°.

Table 3. Overview of Interactions of pHPMA-co-pLMA Copolymers and Its Formulation with 50 and 10 wt % Domperidone with HSA, VLDL, LDL, and VLDL Depleted Serum and the Respective Aggregation Sizes polymer formulation

aggregate seruma

aggregate VLDLb

aggregate LDLb

aggregate VLDL depl seruma

aggregate albuminb

HPMA-co-LMA HPMA-co-LMA 50 w% DOM HPMA-co-LMA 10 w% DOM

yes (160 nm) none none

yes (109 nm) none none

none none none

none none none

none none none

a

Inverse z-average hydrodynamic radii in 28 g/L serum with c(polymer) = 0.1 g/L for all formulations and c(domperidone) = 0.05 g/L (50 wt %) and 0.01 g/L (10 wt %). bCopolymer concentration is 0.1 g/L in the respective concentrated protein solution.

aggregates with Rh = 160 nm (Table 3, Figure 2). In contrast, the copolymer−domperidone formulations with either 10 wt % or 50 wt % of the drug did not show any aggregate formation in human blood serum (Figure 3). This does not exclude “some minor interaction”, like the adsorption of a single shell of small plasma proteins. However it excludes a denaturation of the adsorbed proteins, because then inevitably aggregation would result. In the mixture of single polymer chains and core−shell aggregates of the amphiphilic copolymer probably not all hydrophobic lauryl side chains are strongly confined to the hydrophobic patches. Due to the dynamic nature of the copolymer conformation even in the core−shell aggregates steric constraints force some hydrophobic moieties closer to the surface, thereby facilitating hydrophobic interactions with serum components. Thus interactions with the hydrophobic

domains of VLDL are possible and obviously result in the formation of high molecular weight aggregates in human serum as well as in the isolated VLDL solution. In contrast, encapsulation of domperidone into p(HPMA-co-LMA) copolymers stabilizes the hydrophobic domains of the amphiphilic copolymer. As a consequence, the copolymer micelles increase in size (Table 2). Furthermore fewer lauryl side chains are available for protein association. In addition, no dissolved single polymer chains are detectable anymore by DLS which may strongly interact with serum proteins. Consequently, no aggregate formation is detectable anymore and the copolymer−domperidone formulations are stable over 24 h in human serum. In a second step serum was fractionated and the copolymers were incubated with several isolated solutions of serum components to identify aggregation inducing components. 3773

dx.doi.org/10.1021/mp400254b | Mol. Pharmaceutics 2013, 10, 3769−3775

Molecular Pharmaceutics

Article

detectable, which result in the formation of high molecular weight aggregates. By serum fractionation also the respective aggregate inducing species may be identified, thus confirming dynamic light scattering to be a powerful tool for preclinical evaluation of polymer therapeutics with respect to their behavior in complex biological media. Here, the aggregation profile of amphiphilic p(HPMA-coLMA) copolymers was shown to completely change when hydrophobic domperidone is loaded into the copolymer micelles. The pure copolymers/copolymer micelles form complexes with VLDL, whereas the copolymer−domperidone formulations do not change their size in human serum for up to 24 h. The interactions of p(HPMA-co-LMA) copolymers with the ApoE carrying VLDL lipoproteins are especially interesting since ApoE adsorption onto NP surfaces was found to mediate receptor mediated endocytosis over the BBB. The formation of p(HPMA-co-LMA)−VLDL complexes may therefore be interpreted as a hint that ApoE and VLDL might be involved in the mechanism of p(HPMA-co-LMA) mediated drug delivery over the BBB. Our results demonstrate the biomedical relevance of polymer/nanoparticle−serum aggregation and of the identification of the aggregation inducing serum components.

The majority of water-soluble or suspended polymers can interact with components in the blood serum by either electrostatic or hydrophobic interactions.10 Albumin for example associates with anionic or cationic polymers,21 binds, and transports hydrophobic substances via a fatty acid binding site. Several poly(sodium acrylate) polymers modified with alkyl chains22 and stearyl- or oleyl-poly(ethylene glycol)ester block copolymers23 revealed association with albumin. Therefore aggregate formation between the most abundant serum protein albumin and the polymer were investigated. Although the studied p(HPMA-co-LMA) copolymers are carrying hydrophobic lauryl side chains which could lead to binding of albumin via the fatty acid binding site, no formation of copolymer−albumin complexes was detected via DLS in a concentrated albumin solution (Table 3, Supporting Information). Lipoproteins are also known to associate with hydrophobic drugs and amphiphilic polymers due to their amphiphilic nature with a hydrophobic core and a more hydrophilic shell. They are macromolecular vehicles responsible for the transport of hydrophobic lipids, fatty acids, and cholesterol throughout the aqueous environment of the systemic circulation.24VLDL is primarily composed of triacylglycerol and apoliproteins like ApoB and ApoE. Upon enzymatic release of triacylglycerol as free fatty acids, increase in cholesterol esters, and continued loss of ApoE, VLDL is converted to LDL.24 DLS experiments reveal no aggregation of the copolymer with isolated LDL particles but strong association with VLDL leading to copolymer−VLDL complexes with Rh = 109 nm (Table 3, Figure 4a). Copolymer−domperidone formulations did not exhibit any aggregation with either of the solutions of isolated serum components consistent with the lack of aggregate formation in full serum. In order to ensure that the observed aggregation of the bare copolymers in serum can only be attributed to the VLDL, VLDL was removed from serum via ultracentrifugation. In VLDL depleted serum no aggregation between the copolymer and serum components was detectable, thus confirming that interaction of p(HPMA-co-LMA) copolymers with VLDL causes aggregation in human serum. The association with VLDL is most likely due to hydrophobic interactions between the lauryl side chains and the hydrophobic core of the lipoproteins. It is interesting to note that the copolymer interacts obviously more strongly with VLDL than with LDL particles. The major difference between these two types of lipoproteins is the amount of ApoE bound to the particle surface,24 and ApoE was found to mediate brain drug delivery of PBCA NPs.13 Since the p(HPMA-co-LMA) copolymers can mediate drug delivery over the BBB, a possible interpretation could be that ApoE has a high affinity toward the hydrophobically modified polymer. p(HPMA-co-LMA)−domperidone complexes that have lost most of their cargo of domperidone may associate with VLDL. These complexes may be involved in the mechanism of brain drug delivery.

■

ASSOCIATED CONTENT

S Supporting Information *

Information concerning the synthesis of the copolymer and dynamic light scattering. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*M.S.: e-mail, [email protected]; fax, +49 (0) 61313923768; tel, +49 (0) 6131-3923769. *R.Z.: e-mail, [email protected]; fax, +49-6131-3924778; tel, +49-6131-3920361. Author Contributions ‡

M.H. and K.M. contributed equally.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the Max Planck Graduate Center at the University of Mainz (MPGC, grant for M.H.). We are grateful to Dr. Roland Conradi, University Clinic, Mainz, for fruitful discussions and the preparation of several human blood serum samples.

■

REFERENCES

(1) Ferrari, M. Cancer nanotechnology: opportunities and challenges. Nat. Rev. Cancer 2005, 5 (3), 161−71. (2) Zolnik, B. S.; Sadrieh, N. Regulatory perspective on the importance of ADME assessment of nanoscale material containing drugs. Adv. Drug Delivery Rev. 2009, 61 (6), 422−7. (3) Ringsdorf, H. Structure and properties of pharmacologically active polymers. J. Polym. Sci. 1975, 51, 135−153. (4) Duncan, R. Polymer conjugates as anticancer nanomedicines. Nat. Rev. Cancer 2006, 6 (9), 688−701. (5) Markovsky, E.; Baabur-Cohen, H.; Eldar-Boock, A.; Omer, L.; Tiram, G.; Ferber, S.; Ofek, P.; Polyak, D.; Scomparin, A.; SatchiFainaro, R. Administration, distribution, metabolism and elimination of polymer therapeutics. J. Controlled Release 2012, 161 (2), 446−60.

■

CONCLUSION Applying DLS to characterize size distributions in human serum has only recently been established. The technique provides a very sensitive method and efficient tool to monitor aggregate formation of polymer therapeutics in contact with complex biological fluids, as demonstrated above. Furthermore, extending the method to single serum components, interactions between polymers and specific serum components are 3774

dx.doi.org/10.1021/mp400254b | Mol. Pharmaceutics 2013, 10, 3769−3775

Molecular Pharmaceutics

Article

(6) Walkey, C. D.; Chan, W. C. Understanding and controlling the interaction of nanomaterials with proteins in a physiological environment. Chem. Soc. Rev. 2012, 41 (7), 2780−99. (7) Takakura, Y.; Hashida, M. Macromolecular carrier systems for targeted drug delivery: pharmacokinetic considerations on biodistribution. Pharm. Res. 1996, 13 (6), 820−31. (8) Haag, R.; Kratz, F. Polymer therapeutics: concepts and applications. Angew. Chem., Int. Ed. 2006, 45 (8), 1198−215. (9) Seymour, L. W.; Duncan, R.; Strohalm, J.; Kopecek, J. Effect of molecular weight (Mw) of N-(2-hydroxypropyl)methacrylamide copolymers on body distribution and rate of excretion after subcutaneous, intraperitoneal, and intravenous administration to rats. J. Biomed. Mater. Res. 1987, 21 (11), 1341−58. (10) Xue, W. Interactions between solubilized polymer molecules and blood components. J. Controlled Release 2012, 160, 14−24. (11) Aggarwal, P.; Hall, J. B.; McLeland, C. B.; Dobrovolskaia, M. A.; McNeil, S. E. Nanoparticle interaction with plasma proteins as it relates to particle biodistribution, biocompatibility and therapeutic efficacy. Adv. Drug Delivery Rev. 2009, 61 (6), 428−37. (12) Rausch, K.; Reuter, A.; Fischer, K.; Schmidt, M. Evaluation of Nanoparticle Aggregation in Human Blood Serum. Biomacromolecules 2010, 11 (11), 2836−9. (13) Kreuter, J. Nanoparticulate systems for brain delivery of drugs. Adv. Drug Delivery Rev. 2001, 47 (1), 65−81. (14) Gessner, A.; Olbrich, C.; Schroder, W.; Kayser, O.; Muller, R. H. The role of plasma proteins in brain targeting: species dependent protein adsorption patterns on brain-specific lipid drug conjugate (LDC) nanoparticles. Int. J. Pharm. 2001, 214 (1−2), 87−91. (15) Barz, M.; Tarantola, M.; Fischer, K.; Schmidt, M.; Luxenhofer, R.; Janshoff, A.; Theato, P.; Zentel, R. From defined reactive diblock copolymers to functional HPMA-based self-assembled nanoaggregates. Biomacromolecules 2008, 9 (11), 3114−8. (16) Barz, M.; Luxenhofer, R.; Zentel, R.; Kabanov, A. V. The uptake of N-(2-hydroxypropyl)-methacrylamide based homo, random and block copolymers by human multi-drug resistant breast adenocarcinoma cells. Biomaterials 2009, 30 (29), 5682−90. (17) Hemmelmann, M.; Metz, V. V.; Koynov, K.; Blank, K.; Postina, R.; Zentel, R. Amphiphilic HPMA-LMA copolymers increase the transport of Rhodamine 123 across a BBB model without harming its barrier integrity. J. Controlled Release 2012, 163 (2), 170−7. (18) Hemmelmann, M.; Knoth, C.; Schmitt, U.; Allmeroth, M.; Moderegger, D.; Barz, M.; Koynov, K.; Hiemke, C.; Rosch, F.; Zentel, R. HPMA based amphiphilic copolymers mediate central nervous effects of domperidone. Macromol. Rapid Commun. 2011, 32 (9−10), 712−7. (19) Havel, R. J.; Eder, H. A.; Bragdon, J. H. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J. Clin. Invest. 1955, 34 (9), 1345−53. (20) Hemmelmann, M.; Kurzbach, D.; Koynov, K.; Hinderberger, D.; Zentel, R. Aggregation Behavior of Amphiphilic p(HPMA)-cop(LMA) Copolymers Studied by FCS and EPR Spectroscopy. Biomacromolecules 2012, 13, 4065−74. (21) Klajnert, B.; Stanislawska, L.; Bryszewska, M.; Palecz, B. Interactions between PAMAM dendrimers and bovine serum albumin. Biochim. Biophys. Acta 2003, 1648 (1−2), 115−26. (22) Audobert, P. Association between hydrophobically modified polyanions and negatively charged bovine serum albumine. J. Phys. Chem. B 1998, 102, 1327−1333. (23) Asadi, A.; Saboury, A. A.; Moosavi-Movahedi, A. A.; Divsalar, A.; Sarbolouki, M. N. Interaction of bovine serum albumin with some novel PEG-containing diblock copolymers. Int. J. Biol. Macromol. 2008, 43 (3), 262−70. (24) Wasan, K. M.; Brocks, D. R.; Lee, S. D.; Sachs-Barrable, K.; Thornton, S. J. Impact of lipoproteins on the biological activity and disposition of hydrophobic drugs: implications for drug discovery. Nat. Rev. Drug Discovery 2008, 7 (1), 84−99.

3775

dx.doi.org/10.1021/mp400254b | Mol. Pharmaceutics 2013, 10, 3769−3775