Elsevier

Biomaterials

Volume 23, Issue 9, May 2002, Pages 2043-2056
Biomaterials

Effects of cloud-point grafting, chain length, and density of PEG layers on competitive adsorption of ocular proteins

https://doi.org/10.1016/S0142-9612(01)00334-9Get rights and content

Abstract

The effects of pinning density, chain length, and ‘cloud point’ (CP) versus non-CP grafting conditions have been studied on the ability of polyethylene glycol (PEG) layers to minimize adsorption from a multicomponent (lysozyme, human serum albumin (HSA), IgG and lactoferrin) protein solution. Methoxy-terminated aldehyde-PEG (M-PEG, MW 5000) and dialdehyde-PEG (PEG(ald)2, MW 3400) were grafted by reductive amination onto two surfaces of different amine group density, generated by radio-frequency glow discharge (r.f.g.d.) deposition of n-heptylamine (HA) (low density) or allylamine (AlA) (high density) r.f.g.d. polymer layers. The PEG graft density was varied also by increasing the temperature and salt (K2SO4) content of the grafting solution; it reached a maximum at the CP of the PEGs. The CP reaction conditions were critical for producing PEG layers capable of minimizing protein adsorption. X-ray photoelectron spectroscopy (XPS) showed that under these conditions, PEG(ald)2 produced a thick linear PEG layer, most likely by aldol condensation. Protein adsorption was assessed using XPS and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-ToF-MS) in the surface mode (Surface-MALDI-MS). Coatings grafted at non-CP conditions showed some protein adsorption, as did the HA/M-PEG layer grafted at the CP. On the other hand, no protein adsorption was detected on the HA/PEG(ald)2, AlA/M-PEG, and AlA/PEG(ald)2 surfaces when grafted at the CP. Thus, the effects of pinning density and chain length are interrelated, but the key factor is optimization of PEG chain density by use of the CP conditions, provided that a sufficient density of pinning sites exists.

Introduction

Non-specific bioadhesion (protein adsorption, bacterial adhesion, platelet deposition) continues to hinder biomaterials utilization because of the failure of existing and emerging biomedical device materials to prevent such uncontrolled accumulation of biological material at the surface [1]. Attempts at combating protein adsorption and cellular adhesion by surface modification include the use of polysaccharides [2], [3], [4], phospholipids [5], [6], and many others [7], [8]. We have recently reviewed some of the most recent attempts at preventing bioadhesion [8]; surface modification with polyethylene glycol (PEG) or polyethylene oxide (PEO) continues to provide the most promising results in terms of reducing undesirable protein adsorption and cell adhesion [8], [9].

The key properties that determine the protein-repellent effects of PEG coatings have been the subject of much discussion. Reported experimental results are at times inconsistent; one challenge is to verify complete PEG coverage and thereby exclude artefactual protein adsorption at coating defects. Molecular models have been proposed which predict the necessary interfacial properties of PEG layers required to prevent protein adsorption, such as the effects of chain length and graft density of end-attached PEGs [10], [11], or the requirement of a combined strong interaction between the PEG chains and the substrate and high chain density [12], [13]. An essential finding of these theoretical models is that the PEG layer (whether in a brush or mushroom conformation) should provide an interfacial barrier to prevent the protein from interacting with the underlying substrate. Recent experimental evidence is supportive of this notion [14], [15], [16], [17], [18], [19], with the interfacial graft density proving to be the most crucial property of a PEG layer for minimizing protein adsorption. However, comparative studies of ethylene oxide (EO) terminated self-assembled monolayers (SAMs) on Au and Ag suggest that adsorption of fibrinogen can occur when the graft density becomes too high: when EO groups are too tightly packed on the surface they lose their helical character, a condition for maintaining a stable hydration shell, and adsorb protein [20].

Many studies of PEG coating have used carefully constructed model systems such as the SAMs or grafting onto surface-modified silicon wafers; the transferability of these approaches to biomedical devices is, however, not assured. In order to apply PEG coatings onto contact lenses and similar devices, we have chosen an approach based on thin interfacial bonding layers deposited from radio-frequency glow discharges (r.f.g.d.'s) established in vapours of organic molecules. This r.f.g.d. polymerization method enables for instance the deposition of thin polymeric coatings with surface amine groups [21] which can subsequently be used for the covalent immobilization of bioactive molecules or suitably functionalized PEGs. By appropriate choice of the ‘monomer’ vapour and r.f.g.d. deposition conditions, it is possible to produce thin uniform coatings with a range of densities of surface amine groups [22], [23], [24]. Another advantage of r.f.g.d. polymerization is the strong adhesion of r.f.g.d. deposited polymeric coatings on a variety of substrates. Furthermore, the deposition method is usually readily transferable from model substrates, suited to coating characterizations, to real-life devices such as contact lenses, and thereby allows application of coatings optimized in vitro to devices for use in vivo. Thin amine r.f.g.d. interlayers have previously been used for the covalent immobilization of polysaccharides onto highly oxygen permeable contact lenses [25].

In this study, we have built onto previous findings that the graft density of PEG chains and their length (molecular weight) together are important for achieving high protein repellency. Utilizing the ability of r.f.g.d. interlayers to provide surfaces with varying densities of amine pinning sites for covalent PEG anchoring, we extend previous work by studying how the density and the PEG chain length interrelate with the density of surface pinning groups. Using varying grafting conditions we also compared grafting under conditions ranging from good solubility of PEG chains to marginal solvation (‘cloud point’, CP). The latter condition results in reduced chain repulsion during the grafting reaction and hence greater chain packing density at the interface. We conjectured that grafting under marginal solvation conditions therefore may require a higher density of surface amine sites. By using surfaces with different densities of pinning groups, we aimed to study whether the density of pinning sites is important, and if so, how the effects of pinning site density, PEG chain length, and chain solvation would be interrelated. We conjectured that surfaces with less dense pinning sites would require longer ‘brush’ or well-hydrated ‘mushroom’ graft chains to yield a coating that shields the underlying r.f.g.d. surface from irreversible protein adsorption. Thus, we studied protein adsorption onto coatings containing various combinations of surface pinning densities, PEG chain lengths, and chain densities (modulated by solvation conditions during grafting and possibly also by the surface pinning density). For protein adsorption we used a multicomponent solution containing four proteins that are used as constituents in artificial tear fluid [26]. Elucidation of which of the proteins, if any, would succeed in competitive adsorption might yield information on the mechanisms of protein adsorption onto the coatings.

Section snippets

Plasma polymer surfaces

Perfluorinated poly(ethylene-co-propylene) tape (Teflon FEP 100 Type A; DuPont) and poly(ethylene terephthalate) (PET, DuPont) sheets were coated with thin amine polymer layers deposited using a r.f.g.d. struck in ‘monomer’ vapour of either n-heptylamine (HA; Aldrich, 99%) or allylamine (AlA; Fluka, 99%) in a custom-built r.f.g.d. apparatus [27], as previously described [20], [21]. X-ray photoelectron spectroscopy (XPS) was used to assess the N content of the coatings and high-resolution N 1s

Amine pinning layers

Previous reports from this laboratory [21], [22] have detailed the chemical composition and properties of the n-heptylamine (HA) and allylamine (AlA) r.f.g.d. polymer coatings produced on the r.f.g.d. reactor used. On freshly deposited coatings, the XPS N 1s BE of 399.1–399.2 eV indicates that the N is predominantly in the form of amine groups, as expected. However, when the r.f.g.d. coatings are exposed to air for periods of days to weeks, there is substantial oxygen uptake, and a shift in the

Discussion

In the present study, we have varied both the density of surface pinning groups and the length of PEG chains, and we have also used varying grafting conditions to study which factors and structures of PEG coating matter most.

By using a methoxylated PEG and amine surfaces of different densities of pinning sites, we fabricated coatings with a constant length of the PEG chain but different surface coverages due both to varying grafting conditions and different densities of reactive surface pinning

Conclusions

XPS and Surface-MALDI-MS analyses of PEG coatings exposed to a multicomponent protein solution consisting of lysozyme, lactoferrin, albumin, and IgG have shown that linear PEG coatings can prevent protein adsorption through specific tailoring of the coating structure. PEG coatings are protein resistant (within the sensitivity of our techniques; a few ng/cm2 of protein coverage) only if fabricated under suitable conditions, requiring as a minimum the use of marginal solvation as obtained by CP

Acknowledgments

We acknowledge partial support for this research by the Australian Government under the Generic Technology component of the Industry Research and Development Act 1986 (GIRD Project ‘Polymer Surface Engineering’).

References (39)

  • B.D. Ratner

    New ideas in biomaterials science—a path to engineered biomaterials

    J Biomed Mater Res

    (1993)
  • E. Osterberg et al.

    Protein-rejecting ability of surface-bound dextran in end-on and side-on configurationscomparison to PEG

    J Biomed Mater Res

    (1995)
  • K. Ishihara et al.

    J Biomed Mater Res

    (1995)
  • Ratner BD, Castner DG, editors. Surface modification of polymeric biomaterials. New York: Plenum Press,...
  • Kingshott P, Griesser HJ. Surfaces that resist bioadhesion. In: Ratner BD, Hench L, editors. Curr Opinions Sol State...
  • Harris JM, editor. Poly(ethylene glycol) chemistry: biotechnical and biomedical applications. New York: Plenum Press,...
  • T. McPherson et al.

    Prevention of protein adsorption by tethered poly(ethylene oxide) layersexperiments and single-chain mean-field analysis

    Langmuir

    (1998)
  • C.G. Gölander et al.

    Properties of immobilised PEG films and the interaction with proteinsexperiments and modelling

  • K. Emoto et al.

    Grafting of poly(ethylene glycol) epoxide to amino-derivatised quartzeffect of temperature and pH on grafting density

    Anal Chem

    (1996)
  • Cited by (517)

    View all citing articles on Scopus
    View full text