Nanoparticulate systems for brain delivery of drugs

Advanced Drug Delivery Reviews 64 (2012) 213–222

Contents lists available at SciVerse ScienceDirect

Advanced Drug Delivery Reviews journal homepage: www.elsevier.com/locate/addr

Nanoparticulate systems for brain delivery of drugs☆ Jörg Kreuter ⁎ Institut für Pharmazeutische Technologie, Biozentrum, J.W.Goethe-Universität, D-60439 Frankfurt, Germany

a r t i c l e

i n f o

Available online 13 September 2012 Keywords: Nanoparticles Blood–brain barrier Drug delivery to the brain Brain tumors Glioblastomas

a b s t r a c t The blood–brain barrier (BBB) represents an insurmountable obstacle for a large number of drugs, including antibiotics, antineoplastic agents, and a variety of central nervous system (CNS)-active drugs, especially neuropeptides. One of the possibilities to overcome this barrier is a drug delivery to the brain using nanoparticles. Drugs that have successfully been transported into the brain using this carrier include the hexapeptide dalargin, the dipeptide Kyotorphin, loperamide, tubocurarine, the NMDA receptor antagonist MRZ 2/576, and doxorubicin. The nanoparticles may be especially helpful for the treatment of the disseminated and very aggressive brain tumors. Intravenously injected doxorubicin-loaded polysorbate 80-coated nanoparticles were able to lead to a 40% cure in rats with intracranially transplanted glioblastomas 101/8. The mechanism of the nanoparticle-mediated transport of the drugs across the blood–brain barrier at present is not fully elucidated. The most likely mechanism is endocytosis by the endothelial cells lining the brain blood capillaries. Nanoparticlemediated drug transport to the brain depends on the overcoating of the particles with polysorbates, especially polysorbate 80. Overcoating with these materials seems to lead to the adsorption of apolipoprotein E from blood plasma onto the nanoparticle surface. The particles then seem to mimic low density lipoprotein (LDL) particles and could interact with the LDL receptor leading to their uptake by the endothelial cells. After this the drug may be released in these cells and diffuse into the brain interior or the particles may be transcytosed. Other processes such as tight junction modulation or P-glycoprotein (Pgp) inhibition also may occur. Moreover, these mechanisms may run in parallel or may be cooperative thus enabling a drug delivery to the brain. © 2012 Elsevier B.V. All rights reserved.

Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2. In vivo brain drug delivery with nanoparticles . . . . . . . . 3. In vitro experiments with brain blood vessel endothelial cells . 4. Mechanism of nanoparticle-mediated drug transport to the brain 5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

1. Introduction The blood–brain barrier (BBB) represents an insurmountable obstacle for a large number of drugs, including antibiotics, antineoplastic agents, and a variety of central nervous system (CNS)-active drugs,

☆ PII of original article: S0169-409X(00)00122-8. The article was originally published in Advanced Drug Delivery Reviews 47 (2001) 65–81. ⁎ Tel.: +49 69 798 29682; fax: +49 69 798 29694. E-mail address: [email protected]. 0169-409X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.addr.2012.09.015

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. 213 0 . 214 0 . 216 0 . 218 0 . 221 0 . 221 0 . 221 0

especially neuropeptides. This barrier is formed at the level of the endothelial cells of the cerebral capillaries and essentially comprises the major interface between the blood and the brain. A barrier function also occurs at the arachnoid membrane and in the ependymal cells surrounding the circumventricular organs of the brain [1,2]. It is a vital element in the regulation of the constancy of the internal environment of the brain. The composition of the extracellular fluid of the brain is controlled within precise limits, largely independently of the composition of the circulating blood, to provide a stable environment in which the integrative neuronal functions of the brain can optimally take place [3]. The brain blood vessel endothelial cells are characterized by having

214

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

tight continuous circumferential junctions between them thus abolishing any aqueous paracellular pathways between these cells [4]. The presence of the tight junctions and the lack of aqueous pathways between cells greatly restricts the movement of polar solutes across the cerebral endothelium [3]. Passive diffusion of substances across the brain endothelial cells may occur and is dependent on lipophilicity and molecular weight. However, a large number of drugs that possess a favourable lipophilicity that normally should enable an easy transport across these cells are rapidly pumped back into the blood stream by extremely effective efflux pumps [3]. These pump systems include multiple organic anion transporter (MOAT) and especially P-glycoprotein (Pgp) sometimes referred to as multidrug resistance protein (mdr). A number of attempts have been made to overcome the above barrier including osmotic opening of the tight junctions [5], use of prodrugs or carrier systems such as antibodies [6], liposomes [7–9], and nanoparticles. The opening of the tight junctions by osmotic pressure, however, is a very invasive procedure that also enables the entry of unwanted substances into the brain. Prodrugs take advantage of a higher lipophilicity enabling a better penetration and transport into and across the lipophilic endothelial barrier and/or of a circumvention of the efflux-pump systems, but often the prodrug approach is not possible. The colloidal carriers, on the other hand, may take advantage of the biochemical transport systems that are also present in the BBB: The brain is dependent on the blood to deliver substrates as well as to remove metabolic waste. For this reason, carrier-mediated transport systems exist that enable the entry or the elimination of a variety of compounds including hydrophilic substances such as hexoses, amino acids, purine compounds, and mono-carbonic substances as well as lipoproteins including low density lipoprotein (LDL) [10,11]. Among these systems, for instance, the LDL-receptor and the transferrin transcytosis systems may be employed in the delivery of drugs by the above particulate colloidal drug delivery systems.

2. In vivo brain drug delivery with nanoparticles One of the possibilities to deliver drugs to the brain is the employment of nanoparticles. Nanoparticles are polymeric particles made of natural or artificial polymers ranging in size between about 10 and 1000 nm (1 μm) [12]. Drugs may be bound in form of a solid solution or dispersion or be adsorbed to the surface or chemically attached. Poly(butyl cyanoacrylate) nanoparticles represent the only nanoparticles that were so far successfully used for the in vivo delivery of drugs to the brain. This polymer has the advantage that it is very rapidly biodegradable [13,14]. The first drug that was delivered to the brain using nanoparticles was the hexapeptide dalargin (Tyr-D-Ala-Gly-Phe-Leu-Arg), a Leu-enkephalin analogue with opioid activity [15,16]: The antinociceptive activity after intravenous injection of dalargin-loaded poly(butyl cyanoacrylate) nanoparticles of a size around 250 nm overcoated with polysorbate 80 was demonstrated by the tail-flick test [15,16] as well as by the hot-plate test [17,18]. Both tests showed the same tendencies. All controls including a solution of dalargin, a solution of polysorbate 80, a suspension of poly (butyl cyanoacrylate) nanoparticles, a mixture of dalargin with polysorbate 80, dalargin with nanoparticles or a mixture of all three components, dalargin, polysorbate 80, and nanoparticles, mixed immediately before injection, as well as dalargin bound to nanoparticles without polysorbate 80 overcoating exhibited no effect (Table 1). The maximum of the antinociceptive effect was observed by Alyautdin et al. [15,16] after 30–45 min, but at earlier times by other authors [17–22]. The reason for this difference is not known but may be due to the different strains and source of animals [18,23]. The antinociceptive effect was totally inhibited by injection of the μ-opiate receptor antagonist naloxone 10 min before injection of the nanoparticle preparation, demonstrating that the dalargin-induced analgesia is mediated by a central mechanism [15,16].

Table 1 Analgesia in male ICR mice (20–22 g) determined by percentage (mean±S.D.) of maximally possible effect (MPE) in the tail flick test 45 min after intravenous injection of dalargin or of excipients in free form or in combination with nanoparticles to mice (n=5)a. Group 1 2 3 4 5 6 7 8 9 10 a b

% MPE Suspension of empty nanoparticles (200 mg/kg) Polysorbate 80 solution (1%, 200 mg/kg) Dalargin (solution 10 mg/ml, 10 mg/kg) Dalargin (10 mg/kg)+polysorbate 80 (1%, 200 mg/kg) Dalargin (10 mg/kg)+empty nanoparticles (200 mg/kg) Dalargin (10 mg/kg)+empty nanoparticles (200 mg/kg) +polysorbate 80 (200 mg/kg) Dalargin-loaded nanoparticles (10 mg/kg) Polysorbate 80-coated and dalargin-loaded nanoparticles (2.5 mg/kg) Polysorbate 80-coated and dalargin loaded nanoparticles (5 mg/kg) Polysorbate 80-coated and dalargin-loaded nanoparticles (7.5 mg/kg)

0.75±3.0 12.0±3.1 9.3±8.7 7.8±2.3 1.5±5.4 12.5±2.0 3.7±1.1 11.6±9.7 36.8±21.5b 51.8±20.2b

Adapted from Kreuter et al. [16]. Pb0.05.

As mentioned earlier, uncoated dalargin nanoparticles exhibited no antinociceptive effect. Overcoating with polysorbate 20, 40, and 60 led to similar although slighter antinociceptive effects as overcoating with polysorbate 80, whereas other surfactants such as poloxamers 184, 188, 338, 407, poloxamine 908, Cremophor® EZ, Cremophor ® RH 40, and polyoxyethylene-(23)-laurylether (Brij® 35) led to no effects [24]. The antinociceptive action of the dalargin nanoparticles was circadian phase-dependent (day time-dependent). The circadian phasedependency of pain as well as of the action of antinociceptive drugs is long known and was previously studied extensively [25]. Accordingly, a circadian-phase dependency of the reaction time in the hot-plate test was observed by Ramge et al. [18] in Balb/c and DBA/2 mice, this effect being more pronounced with DBA/2 mice in the reaction times during the whole day and night. No differences occurred between untreated mice and after intravenous injection of 10 mg/kg dalargin solution. After injection of the dalargin-loaded and polysorbate 80-overcoated nanoparticles not only the reaction times were increased from about 6 to over 25 s, corresponding to a maximally possible effect (MPE) of 80%, but also the maxima and the minima of the pain reaction were shifted by about 10 h compared to the untreated and dalargin solution-treated mice. While with the latter preparations the shortest reaction times corresponding to the highest pain sensitivity were observed at about 6:00 h, i.e. the end of the activity phase of the night active mice, the shortest reaction times with the nanoparticles were seen at about 22:00 h, the beginning of the activity period. Although the reasons for these differences presently are not known, the circadian dependency of cerebral blood flow as well as of the permeability of the blood vessels and of the activity of the endothelial carrier systems probably are responsible for this effect [18]. The antinociceptive effects with the dalargin-loaded polysorbate-coated nanoparticles were determined at two selected times, 08:00 and 20:00 h, at four dalargin concentrations, 2.5, 5, 7.5 and 10 mg/kg by the tail-flick test. A clear, day time-dependent dose-response curve existed again with the longest reaction times corresponding to a higher antinociceptive effect in the morning and a lower effect in the evening. The pharmacokinetics of 3H-labeled dalargin bound to poly(butyl cyanoacrylate) nanoparticles and in free form after intravenous injection to mice was followed by Schroeder et al. [26]. The radioactivity label in the plasma was significantly higher (Pb0.05) up to 20 min with both nanoparticle preparations, uncoated and coated with polysorbate 80, and over 60 min with the latter preparation. In the brain homogenate as well as in the supernatant and in the pellet after centrifugation for 10 min at 10 000×g the 3H concentrations with the coated

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

nanoparticles were higher than with the other two preparations especially shortly (5 min) after injection, whereas no significant differences occurred at later times. Besides the hexapeptide dalargin, the dipeptide Kyotorphin (L-TyrL-Arg) [19] as well as loperamide [27], antinociceptive agents that normally also cannot cross the BBB, were tested. Loperamide was chosen, because it is not a peptide and is in contrast to dalargin very lipophilic. Similar effects as with dalargin were seen, no significant effects with drug solutions and uncoated nanoparticles, but a significant antinociceptive effect after polysorbate 80 and also after polysorbate 85 overcoating. Brain perfusion experiments with tubocurarine were conducted using rats, and the development of epileptic spikes in the EEC was recorded [28]. Spikes occurred following intraventricular injection of tubocurarine but a normal EEC was observed after addition of tubocurarine to the perfusate. Addition of polysorbate 80 or uncoated tubocurarine-loaded nanoparticles did not change the EEC. However, the addition of polysorbate 80-overcoated tubocurarine-loaded nanoparticles to the perfusate led to the development of frequent severe spikes in the EEC as after intraventricular injection of the drug. Another pharmacological effect, prevention of electroshock induced convulsion was investigated using the novel NMDA reception antagonist MRZ 2/576 (8-chloro-4-hydroxy-1-oxo1,2-dihydropyridazino [4,5-b]quinoline-5-oxide choline salt) [29,30]. MRZ 2/576 is a potent but rather short-acting (5–15 min) anticonvulsant following intravenous administration to mice as estimated by the prevention of the maximal electroshock induced convulsions. The short duration is most probably due to a rapid elimination of the drug from the central nervous system by efflux pump-mediated transport processes. These transport processes can be inhibited by probenecid. Accordingly, probenecid pretreatment prolonged the anticonvulsive action of MRZ 2/576 from about 15 to 150 min. Intravenous administration of the drug bound to poly(butyl cyanoacrylate) nanoparticles coated with polysorbate 80, however, prolonged the duration of the anticonvulsive activity in mice up to 210 min and after probenecid pretreatment up to 270 min compared to the 150 min with probenecid and MRZ 2/576 alone. The results of this study demonstrate that polysorbate 80-coated poly(butyl cyanoacrylate) nanoparticles that were so far employed for the delivery of drugs to the brain that cannot freely penetrate the BBB may also be used to prolong the CNS availability of drugs that have a short duration of action. The pharmacokinetics of doxorubicin (5 mg/kg) bound to nanoparticles after intravenous injection to rats was investigated by Gulyaev et al. [31]. Four preparations, doxorubicin solution in saline, doxorubicin solution in saline plus polysorbate 80, doxorubicin bound to poly(butyl cyanoacrylate) nanoparticles, and doxorubicin bound to the nanoparticles and coated with polysorbate 80, were used. The two doxorubicin solutions without and with polysorbate 80 showed very similar concentrations in the plasma as well as in all organs investigated, i.e. brain, liver, lung, spleen, heart, and kidneys. The differences in plasma concentrations between the four preparations also were not large, although sometimes statistically significant. Nevertheless, it is important to note that the highest initial plasma levels were seen with the polysorbate-coated nanoparticles, followed by the solutions and, finally, by the uncoated nanoparticles. This is due to the rapid removal of particles by the reticuloendothelial system (RES) which may be prevented or retarded by the coating with surfactants [32–34]. Consequently, both nanoparticle preparations led to higher drug concentrations in the liver, spleen, and lungs, and, as also expected [34], the polysorbate coating reduced the doxorubicin concentration of the nanoparticle preparation in these organs. After about 120 min, the plasma half-life of the doxorubicin bound to nanoparticles was increased 4-fold compared to the solutions and the mean residence time about 6-fold. More importantly and as already observed by Couvreur et al. [35], both nanoparticle preparations extremely decreased the heart concentrations of doxorubicin and

215

yielded levels below the detection limit after 2 h. This finding is very important because the use of doxorubicin is limited by its high heart toxicity. The most significant finding, however, was that only polysorbate 80-coated nanoparticles produced very considerable (6 μg/g) doxorubicin concentrations in the brain whereas with all the three other preparations concentrations were below the detection limit of 0.1 μg/g; (Fig. 1). Although it can be argued that the method of measuring the doxorubicin concentration in the brain homogenate as in the above study does not discriminate between the drug that has crossed the BBB and the drug remaining in the blood vessels, it has to be considered that the blood capillaries represent only 1% of the brain volume and the brain blood vessel endothelial cells only 0.1%. If the doxorubicin would have remained in these compartments, the concentrations would have corresponded to 600 or 6000 μg/g, respectively, concentrations that would have been absolutely toxic for the animals leading to their immediate death. For this reason, it is most likely that the drug penetrated into the brain. Accordingly, intravenously injected doxorubicin-loaded polysorbate 80-coated nanoparticles were able to lead to a 40% cure in rats with intracranially transplanted glioblastomas 101/8 that typically kill the rats within 10–20 days [36]. The dose schedule in these experiments was rather conservative, i.e. 3×1.5 mg/kg. Cure was proved by histology following sacrifice of these animals after 6 months. In the seven control groups only one other rat survived (doxorubicin in saline plus polysorbate 80). A present ongoing study with a higher doxorubicin dose supports these findings. The pharmacokinetics of another drug, amitriptyline, a tricyclic antidepressant that normally, however, is able to penetrate the BBB, also showed an improvement in brain AUC following intravenous injection of polysorbate 80-coated nanoparticles whereas their serum AUC was reduced [19]. Schroeder et al. [20] also showed that polysorbate 85-coated poly(butyl cyanoacrylate) nanoparticles may even enable a brain transport after oral administration. The antinociceptive effect with dalargin obtained by this delivery route was not as pronounced although rather prolonged. Magnetic neutral dextran (MDM) and cationic aminodextran (MADM) microspheres of a size of 1–2 μm (average diameter 1.2± 0.39 μm) were produced by emulsification of the aqueous dextran and magnetite (Fe3O4) suspension in cottonseed oil using polysorbate 80 as the emulsifier and crosslinking by cyanogen bromide [37]. The particles were precipitated and washed with ethanol. For this reason, the residual polysorbate 80 content is not known. The microspheres were then suspended in 0.1% polysorbate containing saline and injected into the carotid arteries of normal and rat glioma-2 (RG-2)-bearing rats. Although the polysorbate 80 concentration was 10 times lower than in the above experiments, it can be assumed that a large portion of this amphiphile also adsorbed on the particle surface. The tissue distribution of these microspheres was determined in male Fisher 344 rats bearing RG-2 tumors as well as normal animals with a magnetic field of 0 or 0.6 T applied to the brain for 30 min. Animals were sacrificed at 30 min or 6 h post-injection after which the microspheres were recovered from various tissues and analyzed for magnetite (Fe3O4) content by atomic absorption. Overall, administration of cationic MADM and neutral MDM particles in normal animals resulted in low brain tissue concentrations with the highest concentrations observed in lung and spleen tissue. In contrast, studies in the brain tumor-bearing animals resulted in cationic MADM particles concentrating in brain tumor at levels significantly higher than neutral MDM particles. Cationic particles were also retained in brain tissue over a longer period of time compared to neutral particles with MADM tumor concentrations decreasing only 4% after 6 h compared with a 32% decrease for MDM. Application of a magnetic field failed to produce any significant effect on tissue distribution due to high variability in these groups, but generally resulted in increased brain concentrations and decreased non-target tissue concentrations.

216

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

Fig. 1. Doxorubicin brain concentration after intravenous injection of one of the following preparations: ●, Doxorubicin [5 mg/kg] solution in saline; ■, doxorubicin [5 mg/kg] solution in saline plus 1% polysorbate 80; ▲, doxorubicin [5 mg/kg] bound to poly(butyl cyanoacrylate) nanoparticles; w, doxorubicin [5 mg/kg] bound to poly(butyl cyanoacrylate) nanoparticles plus 1% polysorbate 80. Reprinted from Gulyaev et al. [31] with permission of the copyright holder.

Transmission electron microscopy analysis of brain tissue sections in tumor animals also revealed differences in particle distribution with MADM particles observed in the interstitial space and MDM particles trapped in the vasculature. 3. In vitro experiments with brain blood vessel endothelial cells In vitro experiments using brain blood vessel endothelial cells were conducted to gain insight in the quantitation and possible mechanism of the nanoparticle-mediated transport of drugs into the brain. The first in vitro experiments employed poly(methyl-[2-14C]-methacrylate) nanoparticles [38]. These particles were overcoated with several surfactants and their uptake by bovine brain microvessel endothelial cells (BMEC) was measured. The BMECs were isolated from the grey matter of the cerebral cortex. The cultures were grown in 24-well culture plates, where they formed confluent monolayers 10–12 days after seeding. The nanoparticle suspensions were then incubated with the cell cultures at 37°C and the radioactivity within the cell cultures was determined after 30 min, 2 h and 6 h. The highest and fastest uptake (>300% compared to uncoated controls) was observed after coating with polysorbate 80. Poloxamer 407 also yielded a very high uptake that was, however, delayed by several hours. Poloxamers 184 and 188, polysorbate 20, and polyoxyethylene-23-laurylether showed an intermediate uptake enhancement (>100%) whereas poloxamer 338 and poloxamine 908 only yielded a minimal insignificant uptake enhancement. More recently, rat brain endothelial cells of the RBE4 cell line and poly(butyl cyanoacrylate) nanoparticles were used to study the nanoparticle uptake by these cells [39]. As mentioned before, poly(butyl cyanoacrylate) particles are very rapidly biodegradable [13,14], and are slightly more hydrophilic than the above poly(methyl methacrylate) particles in the experiments by Borchard et al. [38]. The poly(butyl cyanoacrylate) nanoparticles in the RBE4 cell experiments were produced

using fluoresceine isothiocyanate (FITC) dextran 70 000 instead of the normal dextran as in most of the above in vivo experiments. Untreated RBE4 cells were non-fluorescent and only visible by transmission light microscopy. After incubation for 45 min with FITC-dextran 70 000 labeled nanoparticles overcoated with polysorbate 80, cells showed a punctate appearance of fluorescence concentrated within the cells. In contrast, after treatment with nanoparticles at the same concentration without a polysorbate 80-coating, no fluorescence was observable within the cells, even after the addition of a 10-fold higher concentration of nanoparticles. A strong fluorescence was apparent in the surrounding medium. After the addition of polysorbate 80-coated nanoparticles, fluorescence in the cells appeared rapidly after 10 min and reached a maximum after 48 min. After this time the fluorescence began to diminish. In none of the above experiments did the addition of polysorbate 80-coated or uncoated nanoparticles appear to damage the RBE4 cells. In order to determine whether the polysorbate 80-coated nanoparticles were taken up and internalized or whether they were simply adsorbed to the cell surface, cells exposed to coated nanoparticles were rotated employing the image analysis software of the confocal microscope. This treatment clearly showed that the fluorescence was intracellular. Similar experiments were conducted in cultured human, bovine, and murine BMEC [40,41]. This time the poly(butyl cyanoacrylate) nanoparticles were labeled with Rhodamine 6G. The brain endothelial cells were characterized by immunostaining for von Wilebrand factor and LDC and electron microscopy. The uptake was followed by fluorescent as well as by laser confocal microscopy. Uptake of the surfactant-coated nanoparticles was far more pronounced compared to the uncoated particles, even though in the bovine cells a slight increase in uptake of uncoated particles was observed with increasing time of incubation. Using image analysis software, a 20-fold increase in uptake of coated with respect to uncoated nanoparticles was observed

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

in 2 h. Human cells also exhibited the enhanced uptake of the surfactantcoated nanoparticles compared to uncoated particles. In this case the accumulation of particles inside the cells even increased after 4 h of incubation [40]. Similar results were obtained with the murine brain endothelial cells [41]. The toxicity of poly(butyl cyanoacrylate) nanoparticles was determined in the MTT-test [42]. In this test the viability of the cells is measured by the metabolic activity towards the tetrazolium salt MTT (3-[4,5-dimethylthiazol-2-yl] 2,5-diphenyl tetrazolium bromide). The same cells (RBE4, bovine, and human brain microvessel endothelial cells) as in the previous experiments were used. The highest toxicity was observed in the RBE4 cells. At a poly (butyl cyanoacrylate) nanoparticle concentration of 10 μg/ml no decrease in viability was seen, but at the other concentration used, 100 μg/ml, the viability decreased to about 20%. Polysorbate 80-coating reduced the viability slightly. The toxicity in the bovine brain cells is shown in Fig. 2 and that in the human cells in Fig. 3. No significant toxicity was observed up to 20 μg/ml. At 100 μg/ml the viability decreased to 40% in the bovine cells with no apparent difference in toxicity between polysorbate 80-coated and uncoated nanoparticles. In the human cells, at 100 μg/ml, the viability was about 45% with the polysorbate-coated and about 75% with uncoated particles. Permeability experiments towards 14C-sucrose and 3H-inulin after addition of the poly (butyl cyanoacrylate) nanoparticles were conducted in transwell systems by Ramge [43], Steiniger et al. [44], and Olivier et al. [22]. Ramge used the Cecchelli-Model in cooperation with Dr. B. Engelhard and Mr. S. Hamm at the Max-Planck-Institut in Bad Nauheim. The Cecchelli-Model consists of cloned bovine brain capillary endothelial cells seeded on Millicell-CM 0.4 filters cocultured with rat astrocytes in multiwell systems [11,45,46]. At the two nanoparticle concentrations employed, 10 and 20 μg/ml, no influence on the 14C-sucrose or on the 3H-inulin permeability was observable (Figs. 5 and 6). Steiniger et al. [44] cultivated bovine brain endothelial cells (no coculture with astrocytes) originating from the grey matter onto precoated Transwell® inserts with a supplemented DMEM/Ham’s F12 medium. After incubation with 10 μg/ml nanoparticles the 14C-sucrose flux increased 2-fold with uncoated and 6.5-fold with polysorbate 80-coated nanoparticles. Steiniger et al. [44] also measured the transendothelial resistance (TEER) in these experiments. After incubating the cells with the nanoparticles a decrease in TEER of more than 70% with the 5 μg/ml concentration and 90% with the 10 μg/ml concentration was observed with the coated nanoparticles.

217

Fig. 3. Viability of in vitro cultured human brain blood vessel endothelial cells in percent of untreated control cells by the MTT test after addition of poly(butyl cyanoacrylate) nanoparticle (NP) preparations or of polysorbate 80 alone: 100 oP=100 μg/ml NP without polysorbate 80; 100 mP=100 μg/ml NP with polysorbate 80; 10 oP=10 μg/ml NP without polysorbate 80; 10 mP=10 μg/ml NP with polysorbate 80; and P 80=100 μg/ml polysorbate 80. Reprinted from Ramge [42] with permission of the copyright holder.

Nanoparticles without coating also showed a TEER-decrease. However, this effect was delayed, and the decrease was observed only with the 10 μg/ml concentration. Pure polysorbate 80 solution without particles had no effect. Two to 3 h after incubating the cells, the TEER-values started raising and reached their starting level after 24 h. Olivier et al. [22] also used the Cecchelli-Model and observed an over 10-fold increase in the sucrose and inulin fluxes. However, these authors did not add any serum to the cell coculture, thus enabling an easy disruption of the integrity of their cell layer. These experiments giving three different results demonstrate that slight changes in the in vitro models of the BBB can lead to totally different results. The performance with these models also depends on the experience and scrutiny of the investigators. The experiments clearly show that the in vitro models represent a useful tool but still are no substitutes for the in vivo studies. Fenart et al. [47] used the Cecchelli-Model to investigate the effect of charge and lipid coating on the ability of malto-dextrin 60 nm

Fig. 2. Viability of in vitro cultured bovine brain blood vessel endothelial cells in percent of untreated control cells by the MTT test after addition of poly(butyl cyanoacrylate) nanoparticle (NP) preparations or of polysorbate 80 alone: 100 oP=100 μg/ml NP without polysorbate 80; 100 mP=100 μg/ml NP with polysorbate 80; 20 oP=20 μg/ml NP without polysorbate 80; 20 mP=20 μg/ml NP with polysorbate 80; 10 oP=10 μg/ml NP without polysorbate 80; 10 mP=10 μg/ml NP with polysorbate 80; P 1000=1000 μg/ml polysorbate 80; P 100=100 μg/ml polysorbate 80; and P 10=10 μg/ml polysorbate 80. Reprinted from Ramge [42] with permission of the copyright holder.

218

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

nanoparticles to cross the brain capillary endothelial cell layer of the model. Neutral, anionic, and cationic crosslinked malto-dextrin nanoparticles were derivatized or not (neutral) with phosphates (anionic), or quaternary ammonium (cationic) ligands. The particles were used uncoated or coated with a lipid bilayer made of dipalmitoyl phosphatidyl choline and cholesterol. Lipid coating of ionically charged nanoparticles was able to increase endothelial cell layer crossing 3- or 4-fold compared with uncoated particles, whereas coating of neutral particles did not significantly alter their permeation characteristics across the endothelial cell monolayer. Lipid-coated nanoparticles were non-toxic toward the endothelial cell integrity, and crossed this layer by transcytosis without any degradation. Furthermore, a 27-fold increase in albumin transport was observed when albumin had previously been loaded in the cationic lipid-coated nanoparticles. The influence of red blood cells was studied; a marked inhibition of the transport was observed, probably due to strong interaction between nanoparticles and red blood cells. 4. Mechanism of nanoparticle-mediated drug transport to the brain A number of possibilities exist that could explain the mechanism of the delivery of the above mentioned substances across the BBB: 1. An increased retention of the nanoparticles in the brain blood capillaries combined with an adsorption to the capillary walls. This could create a higher concentration gradient that would enhance the transport across the endothelial cell layer and as a result the delivery to the brain. 2. A general surfactant effect characterized by a solubilization of the endothelial cell membrane lipids that would lead to membrane fluidization and an enhanced drug permeability through the BBB. 3. The nanoparticles could lead to an opening of the tight junctions between the endothelial cells. The drug could then permeate through the tight junctions in free form or together with the nanoparticles in bound form. 4. The nanoparticles may be endocytosed by the endothelial cells followed by the release of the drugs within these cells and delivery to the brain. 5. The nanoparticles with bound drugs could be transcytosed through the endothelial cell layer. 6. The polysorbate 80 used as the coating agent could inhibit the efflux system, especially P-glycoprotein (Pgp). All these mechanisms also could work in combinations. Among these mechanisms, mechanisms 1 and 2 are unlikely to contribute to the observed nanoparticle-mediated drug delivery to the brain to a major degree: loperamide, doxorubicin, and tubocurarine are known substrates for Pgp [3]. It is very unlikely that the creation of a higher local concentration in the brain capillaries alone (mechanism 1) would be sufficient to overcome the action of the very effective efflux pumps located in the luminal membrane of the endothelial cells such as Pgp. A general surfactant effect (mechanism 2) can be ruled out by the experiments of Kreuter et al. [24] in which the dalargin nanoparticles were overcoated with a number of different surfactants. Only overcoating with polysorbate 20, 40, 60, and 80 induced an antinociceptive effect, whereas the other surfactants, although being very effective solubilizing agents, were not able to transport dalargin into the brain in pharmacologically sufficient concentrations. The opening of the tight junctions (mechanism 3) can be investigated by the measurement of the so-called inulin spaces by the Oldendorf method [48]. A significant opening of the inulin spaces for instance by osmotic methods increases the space by factors of 10–20 (1000– 2000%) [49,50]. The in vivo determination of the inulin spaces in defined brain regions of rats by Alyautdin et al. [39] confirmed that the control inulin spaces obtained in that study were similar to the comparable vascular spaces obtained for the rat in other studies and species [51,52]. In rats treated with polysorbate 80-coated nanoparticles the

inulin spaces were increased by 10% after 10 min and 99% after 45 min. This increase would suggest that the coated nanoparticles were increasing the volume available to the intravascular inulin but were not significantly disrupting the blood-brain barrier. This increase could be due to a slight opening of the tight junctions, an upfolding of the cell membrane due to endocytotic events, or to an increase in fluid phase endocytosis of inulin associated with the internalisation of the nanoparticles. In this context it is interesting to note that the inulin spaces in rat brain structures are larger than the plasma space determined in previous studies [51,52], suggesting that inulin finds a space in the brain greater than that available to larger tracers such as albumin and erythrocytes. In this study the fact that the inulin space also increased in non-bloodbrain barrier structures such as the pituitary and the choroid plexus is a further indication that the increase in the spaces is the result of inulin entry into cells possibly by endocytosis. On the other hand, the rapid onset of the antinociceptive effects after injection of the polysorbate 80-coated nanoparticles observed by Ramge et al. [21]; Schroeder et al. [17,19,20], and Olivier et al. [22] may be explained by transport through at least partially opened tight junctions. However, this effect was not constantly observed since the onset in the experiments of Alyautdin et al. [15] and Kreuter et al. [16] was delayed reaching significant antinociception only after 30 min and a maximum after 45 min. At present the most likely mechanism of nanoparticle-mediated transport of drugs to the brain is mechanism 4, endocytosis of the nanoparticles and release of the drug in these cells. Endocytosis of polysorbate 80-coated poly(butyl cyanoacrylate) nanoparticles with a variety of different labels by brain blood vessel endothelial cells has frequently been observed [16,38–41] and in vitro been demonstrated by fluorescence and laser confocal microscopy. At an incubation temperature of 37°C only polysorbate overcoating led to a significant and rapid uptake, whereas without coating this uptake was minimal. The uptake of the polysorbate 80-coated nanoparticles was inhibited at 4°C or by a pretreatment with cytochalasin B, a potent phagocytic uptake inhibitor [40,41]. Lück [53] observed an adsorption of apolipoprotein E (apo E) on the surface of polysorbate 20, 40, 60 or 80-coated nanoparticles after incubation for 5 min in human citrate-stabilized plasma at 37°C. The particles were separated from the serum by centrifugation and the adsorbed plasma proteins desorbed and analyzed by 2-D PAGE (two dimensional polyacrylamide gel electrophoresis). Only after polysorbate 20, 40, 60, or 80 overcoating an apo E adsorption was detected (Table 2) whereas no apo E adsorption resulted after incubation of uncoated nanoparticles or overcoating with poloxamers 338, 407, Cremophor ® EL, or Cremophor ® RH40. These results corroborate the in chapter 2 mentioned in vivo findings where only polysorbate 20, 40, 60, or 80 induced an antinociceptive effect with dalargin nanoparticles whereas the other surfactants including those of the in vitro plasma incubation study (Table 3) were unable to induce a significant effect. Moreover, in apo E-deficient ApoEtm1Unc mice that were derived from C57BL/6J mice the antinociceptive effect was reduced after injection of the polysorbate 80-coated nanoparticles compared to the latter ones (Fig. 4) and to other mice [21]. These results are a strong indication that apo E is involved in the nanoparticlemediated drug transport to the brain. Apo E does play an important role in the transport of LDL into the brain. Lipoproteins are of critical importance for the delivery of essential lipids to this organ. The presence of an LDL receptor in the BBB has Table 2 Cerebral cortex inulin spaces in rats after i.v. injection of polysorbate 80-coated poly(butyl cyanoacrylate) nanoparticlesa. Control (n=6) 10 min exposure 45 min exposure a

After Alyautdin et al. (submitted [39]).

1.03±0.07 1.14±0.04 n.s. 2.05±0.16 Pb0.001

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222 Table 3 Amount of apolipoprotein E (apo E) adsorbed on the surface of surfactant-coated poly(butyl cyanoacrylate) nanoparticles in percent of the total amount of adsorbed plasma proteinsa. Surfactant

Percentage (%)

Uncoated Polysorbate 20 Polysorbate 40 Polysorbate 60 Polysorbate 80 Poloxamer 338 Poloxamer 407 Cremophor® EL Cremophor® RM40

0 21.6 29.7 13.9 14.6 0 0 0 0

a

From M. Lück [53].

been demonstrated by a number of authors [11,54–58], and apo E and apo A-I containing particles have been detected in human cerebrospinal fluid [59]. It is very possible that the polysorbate-coated nanoparticles mimic LDL-particles after apo E adsorption following injection into the blood. In this way they could act as Trojan horses by interacting with the LDL receptor leading to their uptake by endocytic processes. The role of polysorbates would be that of an anchor for apo E: Although preincubation of uncoated nanoparticles with apo E before the in vitro incubation in human plasma also led to a significant adsorptive apo E coating of the nanoparticles [53], without a polysorbate anchor this apo E layer was to a large degree displaced by other plasma components during plasma incubation. Accordingly, in vivo the antinociceptive effect of the dalargin nanoparticles was significantly lower if the nanoparticles were incubated in apo E alone without polysorbate prior to injection (R. Alyautdin, personal communication), but it remained similar to coating with polysorbate 80 alone if the particles were coated with the polysorbate before the incubation with apo E. Interestingly and in contrast to what would be expected from these considerations, preincubation of the bovine brain endothelial cells used by Ramge with lipoprotein-deficient fetal calf serum increased the uptake in comparison to control cells with the polysorbate 80-coated nanoparticles [40,41]. This poses the question, why was the uptake of the polysorbate 80-coated nanoparticles in vitro enhanced and not reduced after incubation of these cells with lipoprotein-deficient calf serum? Probably the cells were starved by

219

this treatment leading to an up-regulation of the LDL-receptor. Residual lipoprotein could adsorb to the nanoparticles and possibly, as a result, the cells would take up these particles to a much higher degree. Little is presently known about mechanism 5, transytosis of the nanoparticles. Of course, it is possible that after their endocytotic uptake the particles can be transcytosed through the brain blood vessel endothelial cells. In vitro transcytosis of LDL across the BBB was observed in the Cecchelli-Model by Dehouck et al. [11]. This process was totally blocked by the C7 monoclonal antibody that is known to interact with the LDL receptor. Furthermore, cholesterol depletion upregulated the expression of the LDL receptor in this model. Using the same in vitro model, transcytosis of dipalmitoyl phosphatidyl choline and cholesterol coated ionic crosslinked malto-dextrin nanoparticles of a size of about 60 nm was observed by Fenart et al. [47]. Hence, it is possible that the polysorbate-coated poly(butyl cyanoacrylate) nanoparticles also can be transcytosed. The material used in most of the above experiments for overcoating, polysorbate 80, was shown to be able to inhibit Pgp [60,61]. As mentioned above, this glycoprotein is present in the brain endothelial cells [62], and it is responsible for the multidrug resistance which represents a major obstacle to cancer chemotherapy [60,61]. Hence, inhibition of this efflux pump located in the brain blood vessel endothelial cell also could be responsible for the nanoparticle-mediated transport of drugs to the brain. However, as shown in the in vivo experiments, polysorbate 80 added to the drug solutions in concentrations of 1% as used with the nanoparticles had no effect. On the other hand, the surfactant, of course, may be delivered to the brain endothelial cells more efficiently if it is adsorbed to the nanoparticles. Thus it may augment some of the above mechanisms. At present the significance of the inhibition of the efflux pumps by polysorbate 80 during nanoparticle-mediated drug delivery to the brain is not known. Nevertheless, the author of this review is of the opinion that blockage of the efflux system by polysorbate 80 may contribute to brain drug delivery of nanoparticles but that other processes, especially endocytosis, play a larger role. The main reason for this assumption is the observation that in the extensive and detailed pharmacokinetic experiment with doxorubicin [31] in contrast to other organs significant brain concentrations were only obtained after 2–4 h. Such a delayed response seems to be a reflection of time consuming processes, possibly endocytosis or also transcytosis. Olivier et al. [22] hypothesise that the pharmacological effects on the brain observed with the poly(butyl cyanoacrylate) nanoparticles overcoated with polysorbate 80 were due to the toxicity of these

Fig. 4. Antinociceptive effect expressed as % MPE by the tail-flick test after intravenous administration of dalargin-loaded [10 mg/kg] polysorbate 80-coated poly(butyl cyanoacrylate) nanoparticles to ○ C57BL/6J mice, or to ● apo E-deficient ApoEtm1Unc mice. Reprinted from Ramge [21] with permission of the copyright holder.

220

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

preparations. They did not perform any thorough toxicological investigations but instead base their speculations primarily on a decrease in locomotive activity of the mice after injection of the dalargin nanoparticles, secondly, on the number of mice (30–40%) that died after they injected them intravenously, and thirdly, on their in chapter 3 mentioned observation that in their in vitro model of the BBB an enhanced transendothelial flux of sucrose was observable. Furthermore, they suggest that this hypothetic toxicity is caused by the polymer or its degradation products. However, based on the present literature data the conclusions of Olivier at al. are not at all justifiable. As the most significant sign of toxicity Olivier et al. took their observation of a decrease in locomotive activity preceded by a short hyperactivity of the mice. If Olivier et al. would have read their pharmacological textbooks carefully, they would have realized that besides antinociception this is a typical reaction of opioid drugs like dalargin. (Hence the name morphine, from Morpheus=god of the dreams and the occurrence of a so-called flash that is craved for by junkies.) Secondly, all other groups have not experienced an unusual enhanced toxicity with the polysorbate 80-coated nanoparticles (less than 2%) despite the high numbers of animals they worked with during their experiments (Drs. Alyautdin, Schroeder, Ramge, Friese and Gelperina, personal communications). As for the third observation, i.e. the enhanced flux across the endothelial layer, Olivier et al. did not find any difference between uncoated and coated particles whereas in their vivo experiments only the latter exhibited any effects. Although those authors acknowledge this important discrepancy, they totally neglect this paramount fact in their conclusions and postulate without any more evidence that a disruption of the BBB had occurred in vivo and reject other BBB transport possibilities such as of endocytic uptake of the particles. As mentioned above, Ramge [43] in cooperation with the very experienced group at the Max-Planck-Institute did not detect an increased sucrose or inulin flux in the Cecchelli coculture model (Figs. 5 and 6) while Steiniger [44] observed a slight increase in his model consisting only of endothelial cells. In addition, no toxicity was observable in cultured brain endothelial cells by the MTT test [42]. Olivier et al. did not add serum during

their transport studies in their in vitro model. This probably led to the disruption of the integrity of their barrier and to the huge increase in sucrose flux. Last but not least, Olivier et al. claimed that poly(butyl cyanoacrylate) toxicity would be a well-documented issue. They oversaw that poly(isohexyl cyanoacrylate) nanoparticles made of a very similar polymer performed very well in a clinical phase I study [63] and are according to our information in clinical phase II. It is worth mentioning that the latter polymer leads to similar antinociceptive effects as poly(butyl cyanoacrylate) after coating with polysorbate 80 (Dr. Alyautdin, personal communication). Moreover, in newer experiments a much higher LD50 of above 600 mg/kg was determined compared to 230 mg/kg in the earlier experiments cited by Olivier et al. (Dr. Gelperina, personal communication). Consequently, the above postulation, that the observed pharmacological effects in mice or rats with polysorbate 80-coated poly(butyl cyanoacrylate) nanoparticles are due to a toxic action on the brain endothelium is not justified: The decrease in locomotive activity which Olivier et al. interpreted as a prime sign of toxicity is a typical pharmacological effect of an opiod like dalargin. Secondly, the enhanced mortality in the experiments of these authors probably seems to be caused by the lack of experimental experience. Thirdly, the in vitro results of the authors cannot be directly extrapolated to the in vivo situation since in vitro uncoated and coated nanoparticles exhibited the same effects whereas in vivo only the latter exhibited an effect. The above differences in the results with different BBB models demonstrate that an extrapolation of in vitro results is always dangerous. To summarize, there are a number of possibilities that can explain the mechanism of transport of drugs across the BBB using nanoparticles. Some of these processes may run in parallel or may be cooperative. The brain typically uses two or more processes to maintain its homeostasis. Electron microscopy may help in the detection of these processes and may identify the most important ones. Electron microscopy, however, is very prone to artefacts. In addition, it is very difficult to unambiguously label the nanoparticles for intratissuelar electron microscopy. Moreover,

Fig. 5. Sucrose permeability in the Cecchelli coculture model after incubation with poly(butyl cyanoacrylate) nanoparticle preparations or controls. ●, Filter holder without cells; ○, brain endothelial cell coculture with astrocytes without addition of nanoparticles; ▽, addition of 10 μl/ml poly(butyl cyanoacrylate) nanoparticles; ▼, addition of 20 μl/ml poly(butyl cyanoacrylate) nanoparticles; □, addition of 10 μl/ml poly(butyl cyanoacrylate) nanoparticles overcoated with polysorbate 80; and ■, addition of 20 μl/ml poly(butyl cyanoacrylate) nanoparticles overcoated with polysorbate 80. Reprinted from Ramge [43] with permission of the copyright holder.

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

221

Fig. 6. Inulin permeability in the Cecchelli coculture model after incubation with poly(butyl cyanoacrylate) nanoparticle preparations or controls. ●, Filter holder without cells; ○, brain endothelial cell coculture with astrocytes without addition of nanoparticles; ▽, addition of 10 μl/ml poly(butyl cyanoacrylate) nanoparticles; ▼, addition of 20 μl/ml poly(butyl cyanoacrylate) nanoparticles; □, addition of 10 μl/ml poly(butyl cyanoacrylate) nanoparticles overcoated with polysorbate 80; and ■, addition of 20 μl/ml poly(butyl cyanoacrylate) nanoparticles overcoated with polysorbate 80. Reprinted from Ramge [43] with permission of the copyright holder.

minor contributors to the overall transport processes may be difficult to identify, let alone quantitate by this or also other methods. 5. Conclusions Nanoparticles represent a tool to transport essential drugs across the BBB that normally are unable to cross this barrier. Drugs that have successfully been transported into the brain using this carrier include the hexapeptide dalargin, the dipeptide kytorphin, loperamide, tubocurarine, the NMDA receptor antagonist MRZ 2/576, and doxorubicin. The nanoparticles may be especially helpful for the treatment of the disseminated and very aggressive brain tumors. First promising results in this direction were already obtained in rats. The mechanism of the nanoparticle-mediated transport of the drugs across the BBB at present is not fully elucidated. The most likely mechanism seems to be endocytosis by the endothelial cells lining the brain blood capillaries. Nanoparticle-mediated drug transport to the brain depends on the overcoating of the particles with polysorbates, especially polysorbate 80. This material seems to act as an anchor for apolipoprotein E (apo E) or other substances following injection into the blood stream. The adsorption of apo E from blood plasma onto the nanoparticle surface has been detected in vitro. The particles then seem to mimic LDL particles and could interact with the LDL receptor leading to their uptake by the endothelial cells. After this the drug may be released in these cells and diffuse into the brain interior or the particles may be transcytosed. Other processes such as tight junction modulation or Pgp inhibition also may occur. Moreover, these mechanisms may run in parallel or may be cooperative thus enabling a drug delivery to the brain. Acknowledgements The author wishes to thank Dr. Gelperina, Institute for Medical Ecology, Moscow, Russia, for carefully reading the manuscript and for valuable suggestions.

References [1] H. Davson, M. Segal, Physiology of the CSF and Blood–Brain Barrier, CRC Press, Boca Raton, FL, 1996. pp. 1–192. [2] D. Begley, J. Kreuter, Do ultra-low frequency (ULF) magnetic fields affect the blood-brain barrier, in: M.F. Holick, E.G. Jung (Eds.), Biologic Effects of Light 1998, Kluwer Academic Publishers, Boston, MA, 1999, pp. 297–301. [3] D. Begley, The blood-brain barrier: Principles for targeting peptides and drugs to the central nervous system, J. Pharm. Pharmacol. 48 (1996) 136–146. [4] M. Brightman, Ultrastructure of the brain endothelium, in: M.W.B. Bradbury (Ed.), Physiology and Pharmacology of the Blood–Brain Barrier. Handbook of Experimental Pharmacology, Vol. 103, Springer, Berlin, 1992, pp. 1–22. [5] M.K. Gummerloch, E.A. Neuwalt, Drug entry into the brain and its pharmacologic manipulation, in: M.W.B. Bradbury (Ed.), Physiology and Pharmacology of the Blood–Brain Barrier. Handbook of Experimental Pharmacology, Vol. 103, Springer, Berlin, 1992, pp. 525–542. [6] W.M. Pardridge, J.L. Buciak, P.M. Friden, Selective transport of an anti-transferrin receptor antibody through the blood–brain barrier in vivo, J. Pharmacol. Exp. Ther. 259 (1991) 66–70. [7] X. Zhou, L. Huang, Targeted delivery of DANN by liposomes and polymers, J. Control. Rel. 19 (1992) 269–274. [8] D. Chen, K.H. Lee, Biodistribution of calcitonin encapsulated in liposomes in mice with particular reference to the central nervous system, Biochem. Biophys. Acta 1158 (1993) 244–250. [9] J. Huwyler, D. Wu, W.M. Pardridge, Brain drug delivery of small molecules using immunoliposomes, Proc. Natl. Acad. Sci. USA 93 (1996) 14164–14169. [10] S.I. Rapoport, Modulation of the blood–brain barrier permeability, J. Drug Target. 3 (1996) 417–425. [11] B. Dehouck, L. Fenart, M.-P. Dehouck, A. Pierce, G. Torpier, R. Cecchelli, A new function for the LDL receptor: Transcytosis of LDL across the blood–brain barrier, J. Cell Biol. 138 (1997) 877–889. [12] J. Kreuter, Nanoparticles, in: J. Swarbrick, J.C. Boylan (Eds.), Encyclopedia of Pharmaceutical Technology, Vol. 10, Marcel Dekker, New York, 1994, pp. 165–190. [13] L. Grislain, P. Couvreur, V. Lenaerts, M. Roland, D. Deprez-Decampeneere, P. Speiser, Pharmacokinetics and distribution of a biodegradable drug-carrier, Int. J. Pharm. 15 (1983) 335–345. [14] P. Couvreur, L. Grislain, V. Lenaerts, F. Brasseur, P. Guiot, A. Biernacki, Biodegradable polymeric nanoparticles and drug carriers for antitumor agents, in: P. Guoit, P. Couvreur (Eds.), Polymeric Nanoparticles and Microspheres, CRC Press, Boca Raton, FL, 1986, pp. 27–93. [15] R. Alyautdin, D. Gothier, V. Petrov, D. Kharkevich, J. Kreuter, Analgesic activity of the hexapeptide dalargin adsorbed on the surface of polysorbate 80-coated poly(butyl cyanoacrylate) nanoparticles, Eur. J. Pharm. Biopharm. 41 (1995) 44–48. [16] J. Kreuter, R.N. Alyautdin, D.A. Kharkevich, A.A. Ivanov, Passage of peptides through the blood–brain barrier with colloidal polymer particles (nanoparticles), Brain Res. 674 (1995) 171–174.

222

J. Kreuter / Advanced Drug Delivery Reviews 64 (2012) 213–222

[17] U. Schroeder, B.A. Sabel, Nanoparticles, a drug carrier system to pass the blood–brain barrier, permit central analgesic effects of i.v. dalargin injections, Brain Res. 710 (1996) 121–124. [18] P. Ramge, J. Kreuter, B. Lemmer, Circadian phase-dependent antinociceptive reaction in mice after i.v. injection of dalargin-loaded nanoparticles determined by the hot-plate test and the tail-flick test, Chronobiol. Int. 17 (1999) 767–777. [19] U. Schroeder, P. Sommerfeld, S. Ulrich, B.A. Sabel, Nanoparticle technology for delivery of drugs across the blood-brain barrier, J. Pharm. Sci. 87 (1998) 1305–1307. [20] U. Schroeder, P. Sommerfeld, B.A. Sabel, Efficacy of oral dalargin-loaded nanopartcle delivery across the blood–brain barrier, Peptides 19 (1998) 777–780. [21] P. Ramge, Untersuchungen zur Überwindung der Blut-Hirn-Schranke mit Hilfe von Nanopartikeln, Shaker Verlag, Aachen, 1998. pp. 123–139. [22] J.-C. Olivier, L. Fenart, R. Chauvet, C. Pariat, R. Cecchelli, W. Couet, Indirect evidence that drug brain targeting using polysorbate 80-coated polybutylcyanoacrylate nanoparticles is related to toxicity, Pharm. Res. 16 (1999) 1836–1842. [23] C. Castellano, S. Puglisi-Allegra, P. Renzi, Genetic differences in daily rhythms of pain sensitivity in mice, Pharmacol. Biochem. Behav. 23 (1985) 91–92. [24] J. Kreuter, V.E. Petrov, D.A. Kharkevich, R.N. Alyautdin, Influence of the type of surfactant on the analgesic effects induced by the peptide dalargin after its delivery across the blood–brain barrier using surfactant-coated nanoparticles, J. Control. Rel. 49 (1997) 81–87. [25] R.C.A. Frederickson, V. Burgis, J.D. Edwards, Hyperalgesia induced by naloxone follows diurnal rhythm in responsitivity to painful stimuli, Science 198 (1977) 756–758. [26] U. Schroeder, H. Schroeder, B.A. Sabel, Body distribution of 3H-labelled dalargin bound to poly(butyl cyanoacrylate nanoparticles) after i.v. injections to mice, Life Sci. 66 (2000) 495–502. [27] R.N. Alyautdin, V.E. Petrov, K. Langer, A. Berthold, D.A. Kharkevich, J. Kreuter, Delivery of loperamide across the blood–brain barrier with polysorbate 80-coated polybutylcyanoacrylate nanoparticles, Pharm. Res. 14 (1997) 325–328. [28] R.N. Alyautdin, E.B. Tezikov, P. Ramge, D.A. Kharkevich, D.J. Begley, J. Kreuter, Significant entry of tubocurarine into the brain of rats by absorption to polysorbate 80-coated polybutylcyanoacrylate nanoparticles: an in situ brain perfusion study, J. Microencapsul. 15 (1998) 67–74. [29] A. Friese, E. Seiler, G. Quack, B. Lorenz, J. Kreuter, Enhancement of the duration of the anticonvulsive activity of a novel NMDA receptor antagonist using poly(butylcyanoacylate) nanoparticles as a parenteral controlled release delivery system, Eur. J. Pharm. Biopharm. 49 (2000) 103–109. [30] A. Friese, Kleinpartikuläre Trägersysteme (Nanopartikel) als ein parenterales Arzneistofftransportsystem zur Verbesserung der Bioverfügbarkeit ZNS-aktiver Substanzen dargestellt am Beispiel der NMDA-Rezeptor-Antagonisten MRZ 2/576 und MRZ 2/596, Ph.D. Thesis, J.W. Goethe-Universität, Frankfurt, 2000. [31] A.E. Gulyaev, S.E. Gelperina, I.N. Skidan, A.S. Antropov, G.Y. Kivman, J. Kreuter, Significant transport of doxorubicin into the brain with polysorbate 80-coated nanoparticles, Pharm. Res. 16 (1999) 1564–1569. [32] L. Illum, S.S. Davis, Effect of the nonionic surfactant poloxamer 338 on the fate and deposition of polystyrene microspheres following intravenous administration, J. Pharm. Sci. 72 (1983) 1086–1089. [33] L. Illum, S.S. Davis, R.H. Müller, E. Mak, P. West, The organ distribution and circulation time of intravenously injected colloidal carriers sterically stabilized with a blockcopolymer-poloxamine 908, Life Sci. 40 (1987) 367–374. [34] S.D. Tröster, U. Müller, J. Kreuter, Modification of the body distribution of poly(methyl methacrylate) nanoparticles in rats by coating with surfactants, Int. J. Pharm. 61 (1990) 85–100. [35] P. Couvreur, B. Kante, L. Grislain, M. Roland, P. Speiser, Toxicity of polyalkylcyanoacrylate nanoparticles II: Doxorubicin-loaded nanoparticles, J. Pharm. Sci. 71 (1982) 790–792. [36] S.E. Gelperina, Z.S. Smirnova, A.S. Khalanskiy, I.N. Skidan, A.I. Bobruskin, J. Kreuter, Chemotherapy of brain tumours using doxorubicin bound to polysorbate 80-coated nanoparticles, Proceedings of the 3rd World Meeting APV/APGI, Berlin, 2000. 3/6th April 2000. pp 441–442. [37] S.K. Pulfer, J.M. Gallo, Enhanced brain tumor selectivity of cationic magnetic polysaccharide microspheres, J. Drug Target. 6 (1998) 215–227. [38] G. Borchard, K.L. Audus, F. Shi, J. Kreuter, Uptake of surfactant-coated poly(methyl methacrylate)-nanoparticles by bovine brain microvessel endothelial cell monolayers, Int. J. Pharm. 110 (1994) 29–35. [39] R.N. Alyautdin, A. Reichel, R. Löbenberg, P. Ramge, J. Kreuter, D.J. Begley, Interaction of poly-(butylcyanoacrylate) nanoparticles with the blood–brain barrier in vivo and in vitro, J. Drug Target (2001) in press. [40] P. Ramge, R.E. Unger, J.B. Oltrogge, D. Begley, H. von Briesen, J. Kreuter, Polysorbate 80-coating enhances uptake of polybutylcyanoacrylate (PBCA)-nanoparticles by

[41] [42] [43] [44]

[45]

[46]

[47]

[48] [49]

[50] [51] [52]

[53]

[54] [55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

human, bovine and murine primary brain capilary endothelial cells, Eur. J. Neuro. 12 (2000) 1935–1940. P. Ramge, Untersuchungen zur Überwindung der Blut-Hirn-Schranke mit Hilfe von Nanopartikeln, Shaker Verlag, Aachen, 1998. pp. 99–110. P. Ramge, Untersuchungen zur Überwindung der Blut-Hirn-Schranke mit Hilfe von Nanopartikeln, Shaker Verlag, Aachen, 1998. pp. 93–98. P. Ramge, Untersuchungen zur Überwindung der Blut-Hirn-Schranke mit Hilfe von Nanopartikeln, Shaker Verlag, Aachen, 1998. pp. 111–118. S. Steiniger, D. Zenker, H. von Briesen, D. Begley, J. Kreuter, The influence of polysorbate 80-coated nanoparticles on bovine brain capillary endothelial cells in vitro, Proc. Int. Symp. Control. Rel. Bioact. Mater. 26 (1999) 789–790. N.P. Dehouck, S. Méresse, P. Delorme, J.C. Fruchart, R. Cecchelli, An easier, reproducible, and mass-production method to study the blood brain barrier in vitro, J. Neurochem. 54 (1990) 1798–1801. M.P. Dehouck, P. Jolliet-Riant, F. Brée, J.C. Fruchart, R. Cecchelli, J.P. Tillement, Drug transfer across the blood–brain barrier: Correlation between in vitro and in vivo models, J. Neurochem. 58 (1992) 1790–1797. L. Fenart, A. Casanova, B. Dehouck, C. Duhem, S. Slupek, R. Cecchelli, D. Betbeder, Evaluation of effect of charge and lipid coating on ability of 60-nm nanoparticles to cross an in vitro model of the blood–brain barrier, J. Pharmacol. Exp. Ther. 291 (1999) 1017–1022. W.B. Sisson, W.H. Oldendorf, Brain distribution spaces of mannitol-3H, inulin-14C, and dextran-14C in the rat, Am. J. Physiol. 221 (1971) 214–217. J.D. Fenstermacher, C.S. Patlak, The exchange of materials between cerebrospinal fluid and brain, in: H.F. Cserr, J.D. Fenstermacher, V. Fencl (Eds.), Fluid Environment of the Brain, Academic Press, New York, 1975. J.A. Kessler, J.D. Fenstermacher, E.S. Owens, Spinal subarachnoid perfusion of rhesus monkeys, Am. J. Physiol. 230 (1976) 614–618. J.E. Cremer, M.P. Seville, Regional brain blood low, blood volume, and haematocrit values the adult rat, J. Cereb. Blood Flow Metab. 3 (1983) 254–256. J.D. Fenstermacher, P. Gross, N. Sposito, V. Acuff, S. Pettersen, K. Gruber, Structural and functional variations in capillary systems within the brain, Ann. N.Y. Acad. Sci. 529 (1988) 21–30. M. Lück, Plasmaproteinadsorption als möglicher Schlüsselfaktor für eine kontrollierte Arzneistoffapplikation mit partikulären Trägern, Ph.D. Thesis, Freie Universität, Berlin, 1997. pp. 14–24 and 137–154. S. Méresse, C. Delbart, J.G. Fruchart, R. Cecchelli, Low-density lipoprotein receptor on endothelium of brain capillaries, J. Neurochem. 53 (1989) 340–345. H.E. De Vries, J. Kuiper, A.G. de Boer, T.J. van Berkel, D.D. Breimer, Characterization of the scavenger receptor on bovine cerebral endothelial cells in vitro, J. Neurochem. 61 (1993) 1813–1821. B. Dehouck, M.P. Dehouck, J.C. Fruchart, R. Cecchelli, Upregulation of the low density lipoprotein receptor at the blood–brain barrier: intercommunications between brain capillary endothelial cells and astrocytes, J. Cell Biol. 126 (1994) 465–473. M. Lucarelli, M. Gennarelli, P. Cardelli, G. Novelli, S. Scarpa, B. Dallapicolla, R. Strom, Expression of receptors for native and chemically modified low-density lipoproteins in brain microvessels, FEBS Lett. 401 (1997) 53–58. R. Cecchelli, L. Fenart, L. Descamps, G. Torpier, M.P. Dehouck, Receptor-mediated transcytosis of blood-borne molecules through blood–brain barrier endothelial cells: Drug targeting by endogenous transport routes, Proc. Intl. Symp. Control. Rel. Bioact. Mater. 24 (1997) 221–222. R.E. Pitas, J.K. Boyles, D.H. Lee, D. Foss, R.W. Maley, Astrocytes synthesize apolipoprotein E and metabolize apolipoprotein E-containing lipoproteins, Biochem. Biophys. Acta 917 (1987) 148–161. D.M. Woodcock, M.E. Linsenmeyer, G. Chrojnowski, A.B. Kriegler, V. Nink, L.K. Webster, W.H. Sawyer, Reversal of multidrug resistance by surfactants, Br. J. Cancer. 66 (1992) 62–68. M.M. Nerurkar, P.S. Burto, R.T. Borchardt, The use of surfactants to enhance the permeability of peptides through Caco-2 cells by inhibition of an apically polarized efflux system, Pharm. Res. 13 (1996) 528–534. C. Cordon-Cardo, J.P. O’Brien, D. Casals, L. Rittmann-Grauer, J.L. Biedler, M.R. Melamed, J.R. Bertino, Multidrug resistance gene (P-glycoprotein) is expressed by endothelial cells at blood–brain barrier sites, Proc. Natl. Acad. Sci. USA 86 (1989) 695–698. J. Kattan, J.-P. Droz, P. Couvreur, J.-P. Marino, A. Boutan-Laroze, P. Rougier, P. Brault, H. Vranckx, J.-M. Grognet, X. Morge, H. Sancho-Garnier, Phase I clinical trial and pharmacokinetic evaluation of doxorubicin carried by poly(isohexyl cyanoacrylate) nanoparticles, Invest. New Drugs 10 (1992) 191–199.