influence of surface coverage with poly(ethylene oxide) on attachment of sterically stabilized...

Biomoteriols 16 (1995) 427439

0 1995 Elsevier Science Limited Printed in Great Britain. All rights reserved

014%9612/95/$10.00

Influence of surface coverage with poly(ethylene oxide) on attachment of sterically stabilized microspheres to rat Kupffer cells in vitro

Garry R. Harper*, Stanley S. Davis*, Martyn C. Davies*, Maria E. Norman*, Tharwat F. Tadros+, David C. Taylori, Michael P. Irvin&, Julian A. Waters5 and John F. Wattsll *Department of Pharmaceutical Sciences, Nottingham University, University Park, Notfingham NG2 7RD, UK; + ICI Agrochemicals, Jealott’s Hi// Bracknell, Berks RG72 GEY, UK; t ICI Pharmaceuticals, Mereside, Alder/y Edge, Macclesfield SK10 4TG, UK; 5 ICI Paints, Wexham Road, Slough SL2 5DS, UK; 11 Department of Materials Science, University of Surrey, Guildford, Surrey GU2 5KH, UK

The attachment to rat Kupffer cells of polymeric microspheres, sterically stabilized with different

amounts of pendant poly(ethylene oxide) (PEO), was assessed in vitro. Four types of copolymer

polystyrene (PS) microspheres were synthesized by variation of four possible monomer ratios that

included styrene, methoxy-PEO-methacrylate (750 and 2000 mol. wt PEO) and allylurea. This produced

poly(styrene-(methoxy-PEO)methacrylate) microspheres with hydrophilic side-groups of either urea

(PS-U-PEO) and/or mixed molecular weight (750/2000 mol. wt) PEO (PS-U-M-PEO, PS-M-PEO), or

single molecular weight (2000) PEO (PS-PEO) at their surfaces. The hypothesis was tested that

increasing the total content of PEO comprising the steric barrier reduces attachment to cell surfaces.

Attachment of PEO microspheres bearing the urea spacer and/or mixed molecular weight PEO was

found to be intermediate between charge stabilized control PS and PEO (2000 mol. wt) bearing

particles. Post-adsorption of different Poloxamer (PEO-poly(propylene oxide)-PEO) surfactants to the

microspheres further decreased attachment. Significant negative linear correlations between surface

PEO content, measured by electron spectroscopy for chemical analysis (ESCA), and attachment to

Kupffer cells were demonstrated. Decreases in attachment also resulted with all graft PEO particles

bearing adsorbed sodium dodecyl sulphate (SDS), whilst the attachment of SDS-treated PS control

particles increased. It is proposed that trains of adsorbed graft PEO are displaced by the SDS to

increase the effective fraction of graft PEO within the steric layer. Overall, increasing the amount of

hydrophilic PEO in the steric layer, from graft and adsorbed sources, reduces the attachment of these

particles to Kupffer cells in vitro.

Keywords: Poly(ethylene oxide), steric stabilization, electrophoretic mobility, microspheres, Kupffer

cells, drug delivery

Received 6 May 1994; accepted 24 August 1994

Correspondence to Dr G. Harper, 307-89 King Street West, Dundas, Ontario L9H lV1, Canada.

Abbreviations: ADIB, azodiisobutyronitrile; ANOVA, analysis of variance; ESCA, electron spectroscopy for chemical analysis; GBSS, Gey’s balanced salt solution; HISS, heat inactivated swine serum; MEM, minimal essential medium: PCS, photon correlation spectroscopy; pd, polydispersity; PEO, polyfethylene oxide): PPO, poly(propylene oxide);

PS, polystyrene; PS-PEO, poly(styrene-(methoxy-PEO) methacrylate); PS-M-PEO, mixed molecular weight PEO poly(styrene-(methoxy-PEO)methacrylate); PS-U-PEO, poly(styrene-allylurea-(methoxy-PEO)methacrylate); PS-U-M- PEO, mixed molecular weight PEO poly(styrene- allylurea-(methoxy-PEO)methacrylate); RES, reticuloendothe- lial system; SDS, sodium dodecyl sulphate; SIMS, secondary ion mass spectroscopy; TOF, time of flight: XPS, X-ray photoelectron spectroscopy.

It is now well established that the approach of a colloidal particle to another particle or surface can be prevented by the adsorption or grafting of highly solvated polymer chains onto the contacting surfaces. These polymer chains contribute to a strong repulsive barrier, through mixing and entropic effects, that can prevent the close approach of two surfaces. This phenomenon is generally described as steric stabilization’ and has been applied in various fields, including colloid science, biomedical science and drug targeting.

In addition to the prevention of particle flocculation and particle adhesion’, the coating of surfaces with poly(ethylene oxide) (PEO) chains has been shown to make them non-adherent and non-thrombogenic, making them useful as biomaterials3-‘.

427 Biomaterials 1995, Vol. 16 No. 6

428 Varied steric stabilization and cell contact: GA. Harper et al.

Colloidal particles bearing steric barriers are also of interest in parenteral drug delivery, especially in the area of drug targeting. The macrophages of the reticula endothelial system (RES), particularly the Kupffer cells of the liver, normally remove foreign matter from the bloodstream by non-specific uptake. A prerequisite of any site-specific delivery system, whose target is a site other than the liver, is therefore to avoid clearance by these hepatic macrophages. It is now known that a steric coating produced by the adsorption of polymer chains to the particle surface can prevent the attachment and uptake of colloids by the Kupffer cells7. In the longer term the incorporation of specific ligands into the surface with such sterically stabilized colloids might enable targeting to specific cells or tissues to be achieved for drug delivery’.

A systematic investigation of the potential of the steric barrier in preventing the clearance of potential colloidal drug carriers from the bloodstream has been undertaken. The simple adsorption of ABA block copolymer surfactants of the PEO/ oly(propylene oxide) (PPO) type, such as PoloxamersT tJ , onto charged polystyrene (PS) latex particles reduced their uptake by both peritoneal macrophages in vitro and the liver following intravenous injection in rabbits7. The adsorption of plasma proteins to the PS particles was shown to be reduced by the adsorbed layer of copolymerg-‘3 and with other PEO- bearing surfaces**. The chemical structure (PEO:PPO ratio) of the Poloxamer adsorbed has also been shown to be important in influencing the organ distribution of the particles, in viva in the rabbit, once hepatic clearance has been avoided14*‘5. This work has shown that in addition to hepatic clearance being a realistic goal with suitably small particles, carrier targeting to sites such as the bone marrow is achievable in an animal model’5Z16. Species- specific effects of the structure of the polymer layer upon colloid removal from the bloodstream have not yet been fully investigated. It is now known, however, that both the nature of the hydrophilic coating and the particle size are important to combined splenic and hepatic removal from the circulatory system in the rat17, and these may also be true for other species.

The theoretical efficiency of grafted PEO or PEO- containing diblock copolymers in sterically stabilizing surfaces against adhesion and protein deposition has been modelled’8-20. The importance of these various parameters are largely untested in vitro, with the exception of Norman et al.” with microspheres and Needham et aJ.‘l with PEO graft liposomes. It is not known whether the architecture of the steric barrier, or subsequent opsonization by protein (specific or non- specific) with displacement of Poloxamer surfactant, or a combination of such factors are responsible for the differences observed in biodistribution of colloids bearing different adsorbed Poloxamers.

To eliminate the possibility of stabilizing Poloxamer desorption playing a role in particle biodistribution, we have produced colloids in which PEO chains were covalently grafted onto the particle surface”. These were demonstrated to reduce their non-specific uptake by rat Kupffer cells, in vitro, compared to controls of charge stabilized PS particles”. The stability of covalently attached PEO systems in viva has subsequently been corroborated by other researchers23-25.

In a previous study we found that a further reduction in the uptake of PEO-grafted particles by Kupffer cells under in vitro conditions could be achieved by the post-synthetic adsorption of Poloxamer surfactant”. It was suggested that the grafted PEO chains on the surface of the particles may not have achieved a maximal packing density, and that Poloxamer surfactant could thereby adsorb to hydrophobic patches between the grafted chains. Such a change in the surface architecture of the particles following Poloxa- mer adsorption could have altered particle contact with the cell surface and attachment to the phagocyte.

We report here the results of experiments designed to test the hypothesis that the structure of the hydrodynamic layer (the region of hydrated polymers extending from the surface), determined in part by both surface graft density of PEO chains and molecular weight of the PEO, influences contact between such particles and the phagocytic cell surface.

Copolymeric PS microspheres were synthesized to incorporate up to three different monomers bearing hydrophilic side-groups into the polymer chain, and add to the surface active portion that partitions into the particle surface. Two of the monomers contribute side- chains of 750 or 2000 molecular weight PEO. A third monomer was used to contribute urea side-groups, which should become inserted within the polymer chain, independently spacing the PEO-bearing and styrene components. The urea also contributes to the hydrophilicity of the polymer and promotes its selective segregation at the particle surface. Urea was chosen since, as a free non-ionic solute in plasma, it is physiologically abundant and relatively inert26.

An in vitro model was used as an indicator of the potential for the PEO-grafted particles to attach to the phagocytic cells of the RES in vivo. Attachment to primary cultures of rat Kupffer cells was measured for five particle types. These were synthesized with a 2000 molecular weight pendant PEO barrier from poly(styrene-methacrylate(methoxy-PEO)) copolymer, a 2000 molecular weight PEO barrier ‘spaced’ along the polymer chain through random copolymerization with an additional monomer bearing a urea side-chain, a ‘mixed’ (750/2000 molecular weight) PEO barrier from a random copolymerization of PS including two PEO-bearing acrylate monomers, and a ‘mixed’ and ‘spaced’ barrier from a random polymerization of all four monomers. Charge stabilized PS particles bearing sulphate groups were used as a control.

Attachment of the particles to Kupffer cells, in vitro, was also assessed after adsorption of two non-ionic PEO- PPO-PEO block surfactants that possess different PPO chain lengths. These may differ in their ability to adsorb to the surface of PEO-grafted particles. Sodium dodecyl sulphate (SDS) was used as a control to assess the effect of adsorbing an anionic surfactant to particle attachment.

EXPERIMENTAL

Preparation of microspheres

Charge stabilized PS microspheres These controls were prepared by the aqueous emulsion polymerization of styrenez7 using a potassium persul-

Biomaterials 1995, Vol. 16 No. 6

Varied steric stabilization and cell contact: GM. Harper et al. 429

phate initiator, as described previously”. Such particles are charged stabilized through the surface segregation of terminal sulphate groups in the polymer chainz7. The microspheres were cleaned by coarse filtration, repeated centrifugation and resuspension in double distilled water.

PEU-bearing PS copolymer microspheres (PS-PEO) A number of methods have been reported for the incorporation of PEO chains into the surface6’14,27-2g. For the present studies poly(styrene-methacry- latefmethoxy-PEO)) copolymer microspheres were prepared by dispersion polymerization2’. Briefly, methoxy-PEO-methacrylate (CH=C(CH,)COO (CH2 CHsO),OCHs) (0.33 g, R = 44, 2000 mol. wt PEO; ICI Paints, Slough) was polymerized with styrene (10 g) in a water-ethanol (105:30 w/w) mixture using azodiisobutyronitrile (ADIB; BDH, Poole, UK; 0.5 g) as an initiator in a reflux apparatus for 6 h at 70°C. Divinylbenzene (Aldrich, Milwaukee, WI, USA) was used at 0.05% as a cross-linking agent.

Mixed molecular weight PEO-bearing PS microspheres [PS-M-PEO) Poly(styrene-methacrylate(PE0)) microspheres were prepared by dispersion polymerization using two methoxy-PEO-methacrylates of PEO molecular weight 2000 and 750 (n = 44 and 17; ICI Paints). The ratios of 0.1 and 6.3 g respectively were added to 10 g styrene in the method described for PS-PEO above.

PEO-urea bearing PS microspheres (PS-U-PEO) Poly(styrene-methacrylate(PEO)-allylurea) microspheres were also prepared by dispersion polymerization. In this case allylurea monomer (H2C= CHCH,NHC=O.NH,; Sigma, St Louis, MO, USA) was incorporated (0.3 g) into the monomer mixture used to make the particles as described for PS-PEO above.

Mixed molecular weight PEO-urea-bearing PS microspheres (PS-U-M-PEO) Poly(methacrylate-styrene-allylurea) microspheres with mixed molecular weight methoxy-PEO were also prepared by dispersion polymerization. The monomer feed mixture contained styrene (10 g), methoxy-PEO- methacrylate (750 mol. wt, 0.3 g; 2000 mol. wt, 0.1 g) and allylurea (0.3 g) in the standard ethanol-water mixture as cited for PS above.

Microsphere clean-up

All PEO-bearing lattices were cleaned by coarse filtration, and heated at 60°C for 6 h in a volume change of the solvent mixture, as used in the synthesis, to remove oligomers and unreacted monomers. Particles were separated from the solvent by centrifugation. Three cleaning cycles were carried out. Prior to a Kupffer cell attachment experiment microspheres were further cleaned by repeated centrifugation and resuspension in a series of ethanol-water mixtures up into double distilled water. The removal of oligomers from the cleaned preparations was monitored by UV spectrophotometry (180-260 nm; Uvikon, Kontron) of the supernatant after centrifugation.

Non-ionic surfactants

Poloxamers 238 and 407 (BASF, NY) and SDS (Biochemical Grade; BDH) were used without further purification and dissolved separately in distilled water (2% w/v). The hydrophilic PEO chain lengths of the two Poloxamer surfactants (HO(CH,CH,O),(CH(CH,) CH20),(CH2CH20),H) are identical (n = 98), whereas the hydrophobic anchoring blocks of PPO are of different length (m is 39 and 67 for Poloxamers 238 and 407 respectively).

Characterization of the microspheres

Within the constraints of successful particle synthesis, the extent and composition of polymer moieties incorporated into the surface will likely be different from that comprising the bulk of the particle. For this reason different methods of surface analysis were used to characterize the particle surfaces, including light microscopy, photon correlation spectroscopy (PCS), electrophoretic mobility, secondary ion mass spectroscopy (SIMS) and electron spectroscopy for chemical analysis (ESCA). A Malvern 4700 PCS instrument (Malvern Instruments, Malvern, UK) and a Malvern Zetasizer II were used to measure particle size and electrophoretic mobility respectively, as described previously”. The former enabled the selection of different particle types of similar size for uptake experiments with Kupffer cells: PS-PEO, 1040 nm diameter (polydispersity, pd, 6.17); PS, 1620 nm (pd 6.90); PS-U-PEO, 1290 nm (pd 0.15); PS-M-PEO, 1090 nm (pd 0.07); PS- U-M-PEO, 1139 nm (pd 0.12).

PS-PEO particles

Efficiency of the clean-up of the PS-PEO microspheres was also monitored by pH profiles of electrophoretic mobility and time of flight (TOF) SIMS was performed before and after the clean-up cycle described above. Electrophoretic mobility profiles were performed on the Malvern Zetasizer II using a constant ionic strength citrate/phosphate-buffered pH series (McIlvaine)30. TOF SIMS was performed using a VG instrument at the University of Manchester Institute of Science and Technology (UMIST).

Surface analysis by ESCA

ESCA allows quantitative analysis of certain functional groups at the outermost (10 nm) layer of surfaces and has already been used to characterize PEO at the surface of biomaterials14. In this instance the predominant PS backbone should be detected as single C-C and C-H binding energies, whilst grafted PEO should be detected as C-O- (ether) binding energy. Thus, the relative contri- butions of C-O:C-H in the Cls ESCA peak should give a relative measure of the PEO/PS composition of the surface of the various particles studied.

Examination of the surface characteristics of the all microsphere types by ESCA (X-ray photoelectron spectroscopy, XPS) was performed using the VG instruments ESCALAB Mk II electron spectrometer at the University of Surrey. Samples were prepared by depositing a packed layer of latex onto pre-wetted 6.2 pm membrane filters (Type NC, Millipore Corp.,


430 Varied steric stabilization and cell contact: G.R. Harper et al.

MA) by filtration of a dilute suspension and allowing the sample to dry in air followed by freeze-drying. The membranes were then attached to aluminium specimen stubs and analysed with Mg Ka target- derived X-rays (hv = 1254 eV). Full survey (50 eV pass energy) and narrow scans of Cls and 01s spectra (at 20 eV pass energy) were acquired. The XPS results were analysed with a VS 5000 datasystem on a DEC PDP 11/73 computer. The curve-fitting software (based on the Chi-square statistic) was employed to compare the ratio of the Cls binding energy peaks correspond- ing to C-H and C-O groups (eV separation and peak width constant), relative to C-H/C-C referenced to 285 eV. This was interpreted as providing a measure of the surface abundance of ethylene oxide (-CH2CH2 0-), (CHCO, 1:l) relative to propylene oxide (-CH(CH,) CH,O-), (CH:CO, 2:l) and styrene (-CH&H(C,H,)-) (CH only). The CO:CH ratio should thus be a function of the surface coverage by PEO.

Similarly, the presence of a urea group (-NHC= O.NHJ at the particle surface might be inferred from the N:Ols peak ratio (2:1), and allow differentiation relative to the other nitrogen content that could be expected at the particle surface, such as any ADIB initiator-derived cyanoisopropyl (NC-C(CH&-) (N:Ols, 2:0) end-groups.

Isolation and culture of rat Kupffer cells

Kupffer cells were isolated for cell culture through pronase (Sigma) perfusion of rat liver in situ to release parenchymal cells, and centrifugal elutriation to separate the Kupffer cell fractions, as described previouslyz2. After isolation, Kupffer cell numbers were estimated from a 0.5 ml sample by Coulter Counter (TAll) or haemocytometer (Improved Neumbauer, Northern Media). The cells were plated at 1.5 million cells well-’ into 30-mm diameter six-well tissue culture plates (Cel-cult). After incubation at 37°C in 5% COJ95% air for 24 h to allow cell adherence, the medium was replaced with 10% heat inactivated swine serum (HISS)/89% minimal essential medium (MEM)/l% L-glutamine and incubated for a further 24 h. Trypan Blue exclusion confirmed 95% cell viability with this method.

Adsorption of polymers to microspheres

The Poloxamer solutions were each equilibrated for 5 h at room temperature with a sample (5% w/v) of PS, PS- PEO, PS-M-PEO, PS-U-PEO or PS-U-M-PEO particles (1:l v/v). In addition, SDS (1% w/v) was incubated with samples (10 ml, 1:l v/v) of PS-PEO, PS-M-PEO, PS-U-M-PEO and PS. Excess unadsorbed Poloxamer or SDS was removed by repeated centrifugation (3500g), removal of the supernatant and resuspension of the particles in distilled water. Removal of surfactants in the supernatant was monitored by UV spectrophotometry (180 nm) and conductivity for Poloxamers and SDS, respectively.

Equilibrium adsorption to microspheres

The equilibrium adsorption of SDS to PS, PS-U-PEO and PS-PEO particles was assessed on 0.5 ml of 1%

(w/v) solids in 20 ml for a dilution series of the surfactant (~30 mM). The adsorption was determined by difference using a conductivity bridge (Wayne-Kerr) to measure the concentration of initial surfactant and the supernatant, after removal of the particles by centrifugation.

Attachment assay

Attachment experiments involved incubation of particles with Kupffer cells at a constant particle:cell number ratio (1O:l) for 1 hz2. Particle numbers were determined by Coulter Counter, with predilution to equivalent particle number in Gey’s balanced salt solution (GBSS; Gibco) buffer for each treatment before in vitro use. For each particle treatment, four replicate culture wells were prepared. Particle dilutions were prepared to add 2 ml of MEM/HISS (Gibco) per well. After incubation at 37°C (5% COJair) for 1 h the plates were washed with three changes of GBSS to remove non-adherent particles, fixed in 50% methanol for 10 min and stained with prefiltered Giemsa (50% v/v in water; BDH). Excess stain was removed with three washes of distilled water and the plates left to dry in air.

Attachment of microspheres by Kupffer cells was assessed by light microscopy”. The number of particles associated with each of at least 40 cells taken from a transect of each of the four replicate plates was determined by direct counting using a Nikon Optiphot at xl000 with an oil immersion lens directly onto the well. Distribution histograms of particle uptake were determined to assess any bimodality in the uptake behaviour of the Kupffer cell populations with different particles. Analysis of variance (ANOVA) was performed to factorially compare treatments, using SUPERAnova software (Abacus Concepts Inc., Berkeley, CA, USA) on an Apple Quadra computer. This allowed a comparison of the significance of differences and interactions of experimental factors between treatments within each of the separate experiments.

Assessment of cytotoxicity

Since any traces of leachate from cleaned particles may contribute to the results of the particle attachment assay, a cytotoxicity experiment was carried out to determine the influence of contaminants added back to the dispersing media on the attachment of cleaned PS or PS-PEO particles to Kupffer cells.

The experiment assessed the addition of increasing amounts of either ethanol, methoxy-PEO-acrylate (2000 mol. wt) or styrene to a standard particle attachment assay for both PS and PS-PEO particles. These materials were added to the particle/GBSS suspension to give final concentrations of 1.3, 4 and 40 ppm.

Attachment of particle types

Experiment design Two experiments on the uptake of sterically stabilized PS particles by rat Kupffer cells were performed. Because of the variability in particle uptake between different cell preparations taken from different rats,


Varied steric stabilization and cell contact: G.R. Harper et a/. 431

ANOVA experiments were designed to compare treatments within an individual experiment”.

Experiment 1 The first experiment investigated the attachment to Kupffer cells of microspheres with PS-PEO and also microspheres with a 2000 molecular weight PEO graft which had been ‘spaced’ by the terpolymerization of allylurea (PS-U-PEO) to include a hydrophilic side- group pendant to the polymer chains. Charge stabilized PS particles were part of the experiment as an internal control of attachment function.

In addition, the effect upon Kupffer cell uptake of the adsorption of either Poloxamer 238 or 407 surfactant onto the particle surface was studied, as a factorial level, for all three particle types.

Experiment 2 The second experiment investigated the attachment to Kupffer cells of the four microsphere types: PS-M-PEO, PS-U-M-PEO, PS-PEO and P.S. In addition, the effect of the pre-adsorption onto all four particle types of each of either Poloxamer 238, Poloxamer 407 or SDS on particle attachment to Kupffer cells was studied. Electrophoretic mobility measurements and surface characterization by ESCA were performed on the particle types used in the experiment.

RESULTS

Removal of macromonomers

An example of the spectrophotometric monitoring of the sequential serum replacements during cleaning of PEO-bearing microspheres is displayed in Figure I.

Electrophoretic mobility of PS-PEO particles with clean-up

PS-PEO particles demonstrate a small negative electrophoretic mobility. The clean-up procedure actually resulted in an increase in measured mobility (Figure 2). The form of the curve changes on clean-up of the particles and the new inflection at pH 3.5 possibly represents the pK, of acidic groups partitioned at the particle surface, exposed after removal of weakly attached macromonomers.

The reduction in electrophoretic mobility of PS control, between the two experiments reported here, was probably due to continued leaching of charged macromonomer from these particles with serial cleaning, since these were the only particles not cross- linked by divinylbenzene during synthesis. Nonethe- less, the advantage of experimental design allows consideration of the variables of particle type within each discrete experiment.

TOF SIMS of PS-PEO and PS microspheres

The positive ion results from TOF SIMS of PS-PEO particles before and after cleaning are summarized in Figure 3a and b. The results were referenced as the relative intensity (RI) to the greatest peak in each of the three classes of 100 mass/charge (m/z) units

Relative Absorbance

1.00

0.50

0.00

I 2:0 GO A0 31!0 nm

Figure 1 Spectrophotometric monitoring of the sequential serum replacement of PS-PEO microspheres during clean-

up.

Mobility

(~10~~ m2N/S)

-l.O-

10

Figure 2 Electrophoretic mobility of PS-PEO particles versus pH: 0, before and 0, after cleaning.

(ranges O-100, 101-200, 201-300). For simplicity, peaks below 0.1 RI within a given range were not included unless they corresponded to assigned peaks of the published reference spectra. Peaks were allocated to published PS, PEO and poly(methy1 methacrylate) (methacrylate was the monomer unit of the graft PEO) spectra of the homopolymers31.3z. Major assignments for PEO were expected at 45, 59,

Biomaterials 1995. Vol. 16 No. 6


1.0 2

0.8

0.6

0.4

0.2

0.0 I

I

0

a

1.0 27

0.8

0.6

0.4

0.2

0.0 !

I

0

b

I I I,

Figure 3 Redrawn positive ion time of flight secondary ion mass spectra for PS-PEO particles: a, before clean-up; b, after clean-up. Scaled ranges O-100, 101-200 and 201- 300 mlz.

89, 175 and 195, but only 45 produced a significant peak. Peaks assignable to PS were detected at 39, 51,

77, 91 (C&H;), 103 (CgH;), 105, 115, 128, 152, 178

and 193. In addition, a significant number of peaks consistent with assignment to a methacrylate-type polymer were 41, 55, 59 (shared with PEO), 69, 101, 107, 115, 121 and 126. It should be noted that the PEO (45) to PS (91) ratio increases on clean-up, providing evidence that some PS-dominant macromonomer is stripped from the surface. Although PS, polyacry- late and PEO all share a fragment at 27 (C,HT), this can be used to compare relative changes in the PS content with clean-up by comparison to its other dominant fragment at 91.

It can be seen that there are substantial unassigned peaks present that were much reduced by cleaning. These major peaks are at 154, 172, 219, 263, 276, 281

and 290 m/z. The initiator ADIB produces cyanoisopropyl (NC-C(CH,),-) groups at the terminal ends of the polymer chain in free radical reactions33, and this could be a source of additional functionality at the particle surface. Inspection of the unassigned peak spacings shows that the assignment of a terminal cyanoisopropyl (CIP) moiety (68 mol. wt) of the initiator to major PS fragments could possibly account for some of the major peaks lost through clean-up: (CIP) + 103 + 1 = 172, 172 + 91 = 263,

263+12 = 276, 276+ 14 = 290. All these peaks decline in intensity on clean-up. This path would account for four of the unassigned peaks. Overall, these results confirm that a latex has been synthesized that bears graft PEO at the surface.

Surfactant adsorption to microspheres

The equilibrium adsorption of SDS surfactant to the PS, PS-U-PEO and PS-PEO particles, measured from the change in supernatant conductivity, is displayed in Figure 4.

Cytotoxicity assessment

Figure 5 shows the effect on PS and PS-PEO particle attachment of the contaminants methoxy-PEO-acrylate, ethanol and styrene, respectively. A common feature is that introduced contaminants result in an increase in attachment at intermediate concentrations and inhibition of attachment at the highest level (40 ppm).

Particle type attachment

Hypothetical surface structures of the particles used in each experiment are summarized in Figure 6.

Experiment 1

PS, PS-U-PEO and PS-PEO particle attachment to Kupffer cells The results of Kupffer cell experiments with PS, PS- PEO and PS-U-PEO microspheres are summarized in Figure 7. PS-PEO microsphere attachment was less than that of control PS microspheres by a factor of about three. This reduction in attachment is consistent with out previously published results that showed that the ratio of the uptake of PS/PS-PEO particles may be in the range 2.3-30.1 (mean 11.8)‘l. By contrast, PS-U- PEO particles showed more attachment to Kupffer cells than PS-PEO microspheres (P = 0.001, x 1.8), but less than PS particles. Histograms that show the

30-

SDS Binding

QJMok?/nl*)

Equilibrium Solute Concentration ( SDS mMole)

Figure 4 Equilibrium sodium dodecyl sulphate (SDS) binding to particle types of experiment 1. 0, PS; A, PS-U- PEO; 0, PS-PEO.

Biomaterials 1995. Vol. 16 No. 6

Varied steric stabilization and cell contact: G.R. Harper et al. 433

hkthoxy-PEO-Acryfate 61

Ethanol 14 7

Mean Particle ’ AttachmentlCell

6

Styrene

Contamination (ppm)

Figure 5 Effect of different contaminants on polystyrene (0) and PS-PEO (0) particle attachment to Kupffer cells.

frequency distribution of particles per cell for each particle type (Figure 8) reveal no bimodality in particle attachment.

The effect of surfcrctant adsorption on particle uptake The adsorption of either Poloxamer 238 or Poloxamer 407 to the surface of all three particle types reduced attachment to Kupffer cells in all cases (Figure 7).

PS PS-PEO PS-U-PEO

PI ‘MIXED’ ‘MIXED & SPACED’

I31 P407 P238 SDS

Figure 6 Schematic of hypothetical surface structures of: 1, the three particle types used in experiment 1; 2, the four particle types of experiment 2; 3, the three surfactants used. -, PEO; B, PPO; - , hydrocarbon; 0, sulphate; [?, urea: q , PS.

Attachment ““1 T

(Particles/Cell)

PS PS-U-PEO

Particle Type

PS-PEO

Figure 7 Experiment 1, effect of particle type and post- treatment with Poloxamer 238 or 407 on attachment to Kupffer cells in vitro. n , particle alone; ETj, adsorbed Poloxamer 407; HI adsorbed Poloxamer 238.

Adsorption of either Poloxamer 238 or 407 to charge stabilized PS particles also results in less attachment to Kupffer cells than particles not treated (P = 0.001). However, there was no difference between surfactant type in reducing PS particle attachment (t = 2.28, d.f. = 7).

The adsorption of either Poloxamer type to PS-PEO microspheres also produced a further reduction (t = 1.65) in attachment to Kupffer cells.

The adsorption of Poloxamers 238 and 407 to PS-U- PEO microspheres further reduced attachment to Kupffer cells compared to PS-U-PEO particles alone. In this case P238 is more effective than P407 (P = 0.02). The attachment of PS-U-PEO particles with Poloxa- mers was no different from graft PS-PEO (t = 1.34) and not as low as the attachment observed for PS-PEO with Poloxamers adsorbed (P = 0.001).


434 Varied steric stabilization and cell contact: GM. Harper et al.

-PEO

40

30

20

10

0 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40

Figure 8 Histograms of the frequency distribution of particles per cell for the microsphere types PS, PS-U-PEO and PS-PEO. q , Zero attachment; n , class interval 5.

Characterization of the particle suq4ace: electrophoretic mobility The electrophoretic measurements on the three particle types with and without adsorbed Poloxamer 238 or 407

are given in Table Z. The PS-U-PEO microspheres have an electrophoretic mobility (-2.64 x 10m4 m* V-l s-l) which is intermediate between that of charge stabilized PS microspheres (-6.37 x 10e4 m2 V-' s-l) and PS-PEO particles (-0.89 x 10m4 mz V-‘S-~).

The results in Table I also show that the adsorption of Poloxamers 238 and 407 to all three particle types reduces electrophoretic mobility. The adsorption of either Poloxamer produces a similar reduction in electrophoretic mobility of PS-PEO particles.

ESCA studies Control PS particles were found to have a low C-O:C-H ratio in the ESCA experiment, whereas this ratio was higher for PS-U-PEO and PS-PEO particles. The C-O: C-H ratio increased upon adsorption of Poloxamers (Figure 9). Using the results of the Kupffer cell experiments, a plot of particle attachment per cell against ESCA C-O:C-H ratio gives a significant negative correlation (correlation coefficient r = 0.97; Figure ~0).

Experiment 2

PS, PS-M-PEO and PS-PEO attachment to Kupffer cells The results of Kupffer cell attachment experiments are summarized in Figure 1 I. In this experiment the PS-M- PEO particles show a similar amount of attachment to Kupffer cells as PS-PEO particles (t = 0.29). PS controls show attachment that is comparable to PS-U-

M-PEO particles (t = 0.46), whilst both PS-PEO and PS-M-PEO particles show less uptake than PS particles (P > 0.05 in both cases).

The effect of adsorbed surfactant on particle uptake As in experiment 1, the adsorption of either Poloxamer 238 or Poloxamer 407 to the surface of PS-U-M-PEO and PS-M-PEO reduced attachment to Kupffer cells compared to similar particles without adsorbed surfactant. For PS-M-PEO particles, the adsorption of either Poloxamer inhibited attachment to Kupffer cells as effectively as PS-PEO particles not receiving Poloxamer.

The PS-PEO and PS-M-PEO particles in this experiment showed, respectively, only 44% and 39% reduction in mean attachment relative to PS controls (P = 0.05 for both). Adsorption of Poloxamer 407 was more effective than Poloxamer 238 in reducing attachment of the PS-PEO particles (P = 0.05).

Conversely, Poloxamer 238 was apparently more effective than Poloxamer 408 in reducing attachment of charge stabilized PS particles (P = 0.001). Such particles with Poloxamer showed no significant difference in attachment compared to PS-PEO particles (P > 0.05).

The adsorption of Poloxamers 238 and 407 to PS-U- M-PEO particles resulted in a reduction in attachment equivalent to PS-PEO with no added Poloxamer.

Adsorption of the anionic surfactant SDS to all three PEO-grafted particle types (PS-PEO, PS-M-PEO and PS- U-M-PEO) resulted in further reduction in attachment to Kupffer cells (Figure 12). This reduction was statistically significant for both PS-PEO and PS-M-PEO particles (P = 0.05). The attachment of these three SDS-treated particle types to Kupffer cells was the same irrespective of the structure/density of the particle type.

However, the effect of SDS adsorption produced a reduction in attachment that was not as large as PS-PEO with adsorbed Poloxamer 407 (P = 0.01). The differences between adsorption of SDS or Poloxamer were not significant for the other particle types bearing PS-M-PEO or PS-U-M-PEO. Conversely, PS control particles bearing SDS or Poloxamers showed increased and reduced attachment, respectively (P = 0.001 for both).

Characterization of the particle suzface: electrophoretic mobility The results of electrophoretic mobility studies with the particles used in Kupffer cell experiment 2 are shown in Table 2. The electrophoretic mobility of PS-PEO particles used in this experiment was found to be small (-0.49 x 10m4 m2 V-l s-l, respectively) and was reduced even further by adsorption of Poloxamers 238

Table 1 Experiment 1, effect of non-ionic surfactants on particle electrophoretic mobility (units ~10~~ m2 V-’ s-‘)

Particle type Particles alone + Poloxamer 238 + Poloxamer 407

PS -6.37 -0.54 -1.39 (1.52) (0.01) (0.01)

PS-U-PEO -2.64 -0.45 -0.45 (0.08) (0.02) (0.02)

PS-PEO -0.89 -0.25 -0.26 (0.02) (0.01) (0.01)



ESCA C-OIC-H Ratio

PS PSU-PEO Particle Type

PSPEO

Figure 9 Experiment 1, effect of particle composition and Poloxamer adsorption on electron spectroscopy for chemical analysis C-O:C-H ratio of particle surface.

n v Particle alone; q , adsorbed Poloxamer 407; q , adsorbed Poloxamer 238.

30-

Attachment (ParticlelCell)

zo-

lo-

y = 37.7 - 168x r = 0.977

0.a 0.1 a.2 0.3 C-O/C-H

Figure 10 Experiment 1, regression of mean particle attachment to Kupffer cells in vitro on electron spectroscopy for chemical analysis C-O:C-H ratio of particle surface. l , PS; 0, PS-U-PEO; 0, PS-PEO.

and 407. These results are consistent with the results of experiment 1 and previous work”.

PS-M-PEO and PS-U-M-PEO particles were found to have higher electrophoretic mobilities than PS-PEO particles (0.84 and -1.04 x lop4 mz V-’ s-l respectively).. These results are consistent with PS-PEO particles having a larger volume fraction occupation of the steric barrier by PEO than either PS-M-PEO or PS- U-M-PEO as discussed above, given the assumption that these particles carry Stern layers of approximately equivalent charge density.

The adsorption of either Poloxamer 238 or 407 to PS- M-PEO and PS-U-M-PEO reduces electrophoretic mobility of both particle types.

As mentioned above, the adsorption of SDS to PS- PEO, PS-M-PEO and PS-U-M-PEO also results in a further reduction in attachment to Kupffer cells. From our previous work, we would predict that this reduction is produced by a steric barrier that masks the electrokinetic layer, and this can be demonstrated by a measured reduction in particle electrophoretic mobility. Table 2 shows that this is not the case for adsorption of the anionic surfactants SDS to the

10 Attachment

(Particles/Cell)

6

PS PS-U-M-PEO PS-M-PEO

Particle Type

PS-PEO

Figure 11 Experiment 2, effect of adsorption of Poloxamers to particles on particle attachment to Kupffer cells in vitro. n , Particle alone; q , adsorbed Poloxamer 407; IB, adsorbed Poloxamer 238.

Attachment (Particles/Cell)

PS PS-U-M-PEO PS-M-PEO Particle Type

PSPEO

Figure 12 Experiment 2, effect of adsorption of sodium dodecyl sulphate to particles on particle attachment to Kupffer cells in vitro. n , Particle alone; LJ, added sodium dodecyl sulphate.

particles. Indeed, both PS-U-PEO and PS-U-M-PEO particle types show a significant increase in electrophoretic mobility (P < 0.05). The adsorption of SDS to the particle surface must result in the anionic head- group occupying some position in the particle- aqueous medium interphase. This must make an increased contribution to the total negative charge density on the particle surface so that an increased electrophoretic mobility should be expected. Indeed, the increased mobility and attachment of PS particles bearing adsorbed SDS support this conclusion (P = 0.001).

ESCA studies The particles used in experiment 2 were studied by ESCA, as described above. In general, there was a correlation observed between the ESCA C-O:C-H ratio and particle attachment to Kupffer cells. However, the correlation was not as strong as in the first experiment (Figure 23), but was still statistically significant



Table 2 Experiment 2, effect of Poloxamer and sodium dodecyl sulphate on particle electrophoretic mobility (units ~10~~ m2 V-’ s-‘)

Particle type Particles alone + Poloxamer 238 + Poloxamer 407 + SDS

PS -1.06 -0.54 -0.61 -6.75 (0.11) (0.02) (0.04) (0.01)

PS-M-PEO -1.04 -0.18 -0.08 -1.20 (0.01) (0.01) (0.02) (0.03)

PS-U-M-PEO -0.84 -0.15 -0.23 -0.95 (0.01) (0.01) (0.01) (0.04)

PS-PEO -0.50 -0.14 -0.11 -0.45 (0.01) (0.01) (0.03) (0.02)

a-

Attachment (Particles/Cell)

6-

y = 16.3 - 75.3x r = 0.636

O.i2 o.i4 O.i6 o.ia o.io o.i2 C-O/C-H

Figure 13 Experiment 2, regression of mean particle attachment to Kupffer cells in vitro on electron spectroscopy for chemical analysis C-OC-H ratio of particle surface. 0, PS; IJ, PS-M-PEO; a, PS-U-M-PEO; O,PS-PEO.

(r = 0.836, P = 0.001). The differences in N:Ols ratios of PS, PS-PEO, PS-LJ-PEO and PS-U-M-PEO particles are shown in Figure 14. The ratios are slightly larger for those particles synthesized with allylurea.

DISCUSSION

Effect of contaminants

A consistent effect was observed when individual precursors of particle synthesis were added back with cleaned particles when exposed to Kupffer cells. Particle attachment is enhanced relative to the normal clean particles up to an optimal concentration, beyond which there is decreasing particle attachment that we can attribute to impaired cell function or even cell death. Thus, in our attachment studies the effect of any trace contaminants of particle synthesis introduced with the clean particles would tend to sensitize the Kupffer cells to promote attachment. In all attachment experiments we have demonstrated reduced attachment of cleaned PS-PEO particles, with or without added surfactant, relative to controls. Any trace contamination could not approach the levels achieved experimentally in the cytotoxicity study.

Effect of PEO graft density on particle attachment

Previously, we have shown that the grafting of PEO chains onto the surface of PS microspheres and the subsequent addition of Poloxamer 238 reduces the electrophoretic mobility of these particle?. We have also discussed the mechanism of the change in electro-

0.45

N/Ols Ratio

0.40

0.35

0.30 -# PS PS-PEO PS-U-M-PEO PS-U-PEO

Figure 14 Electron spectroscopy for chemical analysis measured N:Ols ratio for the particle types PS, PS-PEO, PS-U-M-PEO and PS-U-PEO.

phoretic mobility in terms of the pendant chains extending beyond the slipping plane of the particle and thus obscuring the charge layer”. In addition, a thicker layer of adsorbed steric barrier also results in a greater reduction in electrophoretic mobility24,34-36. Since the steric barrier structure is a function of volume fraction of stabilizing chains, it is dependent on both the molecular weight and surface graft density of the stabilizing molecule. Consequently, changes in electrophoretic mobility are influenced by the volume fraction of the steric barrier occupied by the pendant PEO chains and are thereby an indirect indicator of the hydrodynamic thickness of PEO screening the charge layer beneath.

In this study, particles bearing a hydrophilic side- chain of urea in addition to grafted PEO show Kupffer cell attachment, electrophoretic mobility and C-O:C-H ESCA ratio that is intermediate between charge stabilized controls and particles bearing only grafted PEO. This is consistent with an increased spacing of



the PEO chains at the surface of the sterically stabilized particle. This ‘spaced’ PEO layer on the surface of PS particles (PS-U-PEO) is not as effective as in ‘unspaced’ PS-PEO particles in terms of preventing attachment to Kupffer cells. We interpret this as showing that the graft density of PEO and, by implica- tion, the volume fraction of the hydrodynamic profile, determine the extent of interaction between sterically stabilized particles and cell surfaces in vitro.

PS particles with surface grafted ‘mixed’ PEO chains of 750 and 2000 molecular weights (PS-M-PEO) are as effective as PS-PEO particles in their avoidance of attachment to Kupffer cells in vitro. PS-U-M-PEO particles in which the ‘mixed’ barrier is further ‘spaced’ by the presence of urea are less effective than PS-PEO and PS-M-PEO in avoiding attachment to Kupffer cells in vitro. All of these particle types would appear to have a submaximal packing of PEO chains at their surface (see next section).

Further PEO chains were introduced into the surface layer of all the particle types studied by adsorption of PEO-bearing Poloxamer surfactants37. Their presence was inferred from the results of ESCA studies and the observed reduction in particle electrophoretic mobility, and implied from the measurement of equilibrium SDS adsorption to similar particles. This confirms that the graft PEO particles possess subopti- ma1 amounts of PEO stabilizer in their steric barrier. The Poloxamer must adsorb to hydrophobic ‘gaps’ between the graft sites of the PEO chains, increasing the effective number of PEO chains in the steric barrier. This may also change the conformation of any grafted PEO chains that are weakly adsorbed, from ‘trains’ to the extended ‘tail’ and ‘loop’ conformations. This would effectively allow the PEO chains to extend the hydrodynamic layer. This mechanism for displacement of grafted PEO chains is equivalent to the theoretical approach of Silberberg3*, who proposed two main energy states for segments of polymers adsorbed at interfaces, namely ‘trains’ lying in the layer immedi- ately adjacent to the particle surface and ‘loops’ extending directly towards the bulk solvent phase. Hoffman3’ has proposed, for surfaces stabilized by adsorbed polymers, that as additional polymer molecules adsorb to the particle the proportion of the pendant polymer chains in the ‘train’ conformation will decrease, and this is accompanied by a simultaneous increase in profile depth (Delta). These theories are to some extent consistent with our experimental data.

The adsorption of Poloxamer to the surface of PS- PEO, PS-U-PEO and PS-U-M-PEO also results in a further reduction in particle attachment to Kupffer cells in vitro. This and the electrophoretic mobility results for experiment z agree with our previous findings that steric stabilization effectively masks the charged double layer or slipping plane, measured here as reduced electrophoretic mobility. The reduced attachment to Kupffer cells is evidence of increased steric stabilization.

Indeed, increases in the amount of surface PEO measured by ESCA correlate with reduced particle attachment to Kupffer cells in both experiments. Extrapolation from Figure 30 indicates that complete inhibition of particle attachment by cultured Kupffer

Biomaterials 1995, Vol. 16 NO. 6

cells in vitro might be predicted using particles having pendant PEO (mol. wt 2000) chains giving rise to an ESCA C-O:C-H ratio of approximately 0.223. Polyur- ethane with grafted PEO chains having a similar surface ESCA 0:C ratio has been found to be effective in influencing the surface adsorption of vitronectin in competition with other opsonizing proteins4’. This may have wider relevance to the role of opsonins in particle recognition and modification of adsorption of such proteins to surfaces bearing grafted PEO. Indeed, Norman et al.ll have demonstrated such differences with adsorbed Poloxamer systems.

Unexpectedly, adsorption of SDS to the mixed and singular molecular weight PEO graft particles results in reduced attachment to Kupffer cells and increased electrophoretic mobility. In contrast, PS controls show an increase in both electrophoretic mobility and attachment to Kupffer cells. It is likely that the SDS adsorbs to hydrophobic domains on the particle surface that occur between the pendant graft PEO chains, as with the non- ionic Poloxamer surfactants. However, no additional PEO chains can be contributed by the SDS to the hydrodynamic profile. The explanation proposed is that any weakly adsorbed PEO ‘trains’ are displaced by the surfactant increasing the thickness of the hydrodynamic layer, whilst the SDS contributes to the increased surface charge density and thus electrophoretic mobility of the particle (Figure 1.5).

CONCLUSIONS

The experimental results support our hypothesis that variation in the density of PEO chains on the surface of polymeric microspheres is relevant to their reduced

Loops & Tails ,

j Extension ,Fraction Volume)

Adsorbed Poloxamer Surfactant

Adsorbed SDS Surfactant

Figure 15 Hypothetical effect of surfactant adsorption in displacing trains of adsorbed PEO and extending steric barrier profile.


attachment to Kupffer cells in vitro. The results of these particle-cell interactions are consistent with a process involving some form of steric stabilization between heterogeneous surfaces. Results from the addition of PEO to the hydrodynamic layer from graft, Poloxamer and SDS surfactant adsorption are consistent with the achievement of different suboptimal levels of steric stabilization with the graft particles. The reduced attachment of microspheres to Kupffer cells increases as the total volume fraction (a function of surface density and molecular weight) of the pendant PEO chains forming the steric layer is increased by adsorption of PEO-bearing surfactants. Adsorption of the anionic surfactant SDS produces a similar effect that is likely to involve displacement of PEO trains from the particle surface. Furthermore, an upper limit to the efficacy of steric stabilization appears to exist near the highest PEO surface densities achieved for this in vitro system.

These observations on the effect of surface density of stabilizer on attachment to Kupffer cells in vitro are important in developing an understanding of the stability of colloidal systems within the systemic circulation, and to the endogenous defence mechan- isms responsible for their non-specific clearance from the vascular compartment. To further understand the potential role of steric stabilization in the development of site-specific drug delivery vehicles, it is important to test these systems in the further biological context of opsonization and in vivo

studies.

ACKNOWLEDGEMENTS

Dr Harper was in receipt of a SERC/ICI collaborative Postdoctoral Research Fellowship (1986-1989).

REFERENCES

Napper DH. Polymeric Stabilisation of Colloidal Disper- sions. London: Academic Press, 1983. Tadros F, Warszyski P, Zembala M. The influence of polymer adsorption on desorption kinetics of colloid particles. II. Experimental studies. Colloids Surfaces 1989: 39: 93-105. Lee JH, Kopecek J, Andrade JD. Protein-resistant surfaces prepared by PEO-containing copolymer surfactants. JBiomed Mater Res 1989; 23: 351-368. Lee JH, Kopeckova P, Kopecek J, Andrade JD. Surface properties of copolymers of alkyl methacrylates with methoxy(polyethylene oxide) methacrylates and their application as protein resistant coatings. Biomuterials 1990; 11: 455-464. Nagaoka S, Nakao A. Clinical applications of antithrom- bogenic hydrogel with long poly(ethylene oxide) chains. Biomaterials 1990; 11: 119-121. Han DK, Jeong SY, Kim YH, Min BG, Cho HI. Negative cilia concept for thromboresistance: synergistic effect of PEO. J Biomed Mater Res 1991; 25: 561-575. Illum L, Hunneyball IM, Davis SS. The effect of hydrophilic coatings on the uptake of colloidal particles by the liver and by peritoneal macrophages. Int JPharm 1986; 29: 53-65.

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

Torchilin VP, Klibanov AL, Huang L, O’Donnell S, Nossiff ND, Khaw BA. Targeted accumulation of polyethylene glycol-coated immunoliposomes in infarcted rabbit myocardium. FASEB J 1992; 6: 2716- 2719. Norman ME, Harper GR, Williams P, Illum L. Opsonisa- tion of sterically stabilised colloidal carriers and their interaction in-vitro with Kupffer cells. J Pharm Pharma- co1 1989; 41 (suppl.): 127P. O’Mullane JE, Davidson CJ, Petrak K, Tomlinson E. Adsorption of fibrinogen on to polystyrene latex coated with the non-ionic surfactant, poloxamer 338. Biomateriuls 1989; 9: 203-204. Norman ME, Williams P, Illum L. Human serum albumin as a probe for surface conditioning (opsonization) of block copolymer-coated microspheres. Biomaterials 1992; 13: 841-849. Lee J, Martic PA, Tan JS. Protein adsorption on Pluronic copolymer-coated polystyrene particles. 1 Colloid Interface Sci 1989; 131: 252-266. Tan JS, Martic PA. Protein adsorption and conforma- tional change on small polymer particles. J Colloid Interface Sci 1990; 136: 415431. Desai NP, Hubbell JA. Solution technique to incorporate polyethylene oxide and other water soluble polymers into surfaces of polymeric biomaterials. Biomaterials 1991; 12: 144-153. Illum L, Davis SS, Muller RH, Mak E, West P. The organ distribution and circulation time of intravenously injected colloidal carriers sterically stabilised with a block copolymer-poloxamine 908. Life Sci 1987; 40: 1553-1560. Illum L, Jacobsen LO, Muller RH, Mak E, Davis SS. Surface characteristics and the interaction of colloidal particles with mouse peritoneal macrophages. Biomaterials 1987; 8: 113-117. Moghimi SM, Porter CJH, Muir IS, Illum L, Davis SS. Non-phagocytic uptake of intravenously injected microspheres in rat spleen: influence of particle size and hydrophilic coatings. Biochem Biophys Res Commun 1991; 177: 861-866. Jeon SI, Lee LH, Andrade JD, De Gennes PG. Protein- surface interactions in the presence of polyethylene oxide. J Colloid Inteqface Sci 1991; 142: 149-158. Jeon SI, Andrade JD. Protein-surface interactions in the presence of polyethylene oxide. II. Effect of protein size. J Colloid Interface Sci 1991; 142: 159-166. Rodrigues K, Mattice WL. Simulation of the steric stabilisation of polymer colloids by diblock copolymers./ ChemPhysl991;94:761-766. Needham D, McIntosh TJ, Lasic DD. Repulsive interactions and mechanical stability of polymer- grafted lipid membranes. Biochim Biophys Acta 1992;llOB: 40-48. Harper GR, Davies MC, Davis SS et al. Steric stabilisation of microspheres with grafted polyethylene oxide reduces phagocytosis by rat Kupffer cells in vitro. Biomaterials 1991; 12: 695-700. Brindley A, Davies MC, Davis SS. The surface characterization of model polymer colloids for potential drug targeting systems. Abstr ACS (Coil) 1992; 203: 20. Davis SS, Illum L, Moghimi SM et al. Microspheres for targeting drugs to specific body sites. 1 Contr Rel 1993; 24: 157-163. Gref R, Minamitaka Y, Peracchia MT, Trubetskoy V, Torchilin V, Langer R. Biodegradable long-circulating polymeric nanospheres. Science 1994; 263: 1600-1603. Smith ST. In: Bishop ML, Duben-Engelkirk JL, Fody ED, eds. Clinical Chemistry: Principles, Procedures, Correla- tions, 2nd edn. Philadelphia: Lippincott, 1991: 436.



27

28

29

30

31

32

33

34

Goodwin JW, Hearn J, Ho CC, Ottewill RH. Studies on the preparation and characterisation of monodisperse polystyrene latices. III. Preparation without added surface active agents. Co11 Polym Sci 1974; 252: 464- 471.

Desai NP, Hubbell JA. Solution technique to incorporate polyethylene oxide and other water-soluble polymers into surfaces of polymeric biomaterials. Eiomateriols 1991; 12: 144-153.

Jialanella GL, Piirma I. Synthesis and characterisation of poly-(2-methyl-co-Z-hydroxypropylmethacrylate)-gra~- polyoxyethylene. J Appl Polym Sci 1991; 43: 1577- 1583. Perrin DD, Dempsey B. Buffers for pH and Metal Zen Control. London: Chapman, 1974. Briggs D, Brown A, Vickerman JC. Handbook of Static Secondary Ion Mass Spectrometry (SIMS). New York: Wiley, 1989. Newman JG, Carlson BA, Michael RS, Moulder JF, Hohlt TA, eds. Static SIMS Handbook of Polymer Analysis. Minnesota: Perkin-Elmer, 1991. Mishra MK. Graft copolymerization of vinyl monomers onto and cellulosic materials. Rev Macromol Chem Phys 1983; C22: 471-513. Garvey MJ, Tadros ThF, Vincent B. Comparison of the

35

36

37

38

39

40

adsorbed layer thickness obtained by several techni- ques of various molecular weight fractions of poly(viny1 alcohol) on aqueous polystyrene latex particles. J Colloid Interface Sci 1976; 55: 440-453. Oshima H, Kondo T. Electrophoretic mobility and Donnan potential of large colloidal particles. J Colloid Integace Sci 1987; 116: 305-311. Oshima H, Kondo T. Approximate analytic expression for the electrophoretic mobility of colloidal particles with surface-charge layers. J Colloid Znteqface Sci 1989; 130: 281-282. Kayes JB, Rawlins DA. Adsorption characteristics of certain polyoxyethylene-polyoxypropylene block copolymers. Co11 Polym Sci 1979; 257: 622-629. Silberberg A. The adsorption of flexible macromole- cules. Part 1. The isolated macromolecule at a plane interface. J Phys Chem 1962; 60: 1872. Hoffman RL. Discontinuous and dilatant viscosity behavior in concentrated suspensions. II. Theory and experimental tests. J Colloid Intedace Sci 1974; 46: 491. Fabrizius-Homan DJ, Cooper SL. Competitive adsorption of vitronectin with albumin, fibrinogen, and fibronectin on polymeric biomaterials. J Biomed Mater Res 1991; 25: 953-971.


influence of surface coverage with poly(ethylene oxide) on attachment of sterically stabilized...

Documents