A number of materials have been proposed for use as biomaterials, including hydrophilic, phase-separated and zwitterionic polymers. The mechanisms responsible for the bio/blood compatibility (bioinertness) of these polymers at the molecular level have not been clearly demonstrated, although many theoretical and experimental efforts have been made to understand these mechanisms. Water interactions have been recognized as fundamental for the biological response to contact with biomaterials. We have proposed the ‘intermediate water’ concept, in which water clearly exhibits defined peaks for cold crystallization in the differential scanning calorimetry chart and presents a strong peak at 3400 cm−1 in a time-resolved infrared spectrum. We found a localized hydration structure consisting of three hydrated waters in poly(2-methoxyethyl acrylate). We hypothesized that intermediate water, which prevents the proteins and blood cells from directly contacting the polymer surface, or non-freezing water on the polymer surface has an important role in the bio/blood compatibility of polymers. We will provide an overview of the recent experimental progress and a theoretical description of the bio/blood compatibility mechanisms as determined by thermal, spectroscopic and surface force measurements.
The variety of polymeric biomaterials with distinct chemical structures that precisely control the molecular architecture underlies the numerous industrial uses of polymers over the past few decades. For example, bio/blood-compatible polymers are used in various bio/medical devices,1, 2, 3 and efforts are continuing to enhance the materials and devices. The recent development of novel polymeric biomaterials, and their application to medical problems, has dramatically improved the treatment of many diseases.4, 5 Although various types of materials have been widely used in nano/medicine, many biomaterials lack the desired functionalities to interface with biological systems and have not been engineered for optimum performance. Therefore, the need to develop novel polymeric materials to address such problems in nano/medicine is increasing. Polymeric materials for medical devices that may come into contact with human blood should have the capacity to resist protein adsorption and blood cell adhesion that can trigger the organism’s defense mechanism.6 Some bio/blood-compatible polymer surfaces have been developed, and these can be classified into the following three categories: (i) hydrophilic surfaces;7 (ii) surfaces with micro-phase-separated domains;8 and (iii) biomembrane-like surfaces,9 including zwitterionic groups.10, 11, 12, 13 The physicochemical properties of surface-bound water, including surface charge, wettability, surface-free energy, stiffness, topography and the presence of specific chemical functionalities, appears to have an instrumental role in the biological response induced by the polymers.14, 15, 16, 17, 18 A new generation polymer, poly(2-methoxyethyl acrylate) (PMEA) (Figure 1), exhibits excellent bio/blood compatibility, and has been approved for medical use by the Food and Drug Administration.19 For example, PMEA-coated circuits significantly reduce blood cell activation when used in cardiopulmonary bypass and catheters. The compatibility of PMEA’s with platelet, leukocyte, erythrocyte, complement and coagulation systems appear to be dictated by the presence of an intermediate water.20, 21, 22
The term ‘Bio/blood compatibility’ is generally used to indicate the properties of polymeric materials that do not cause adverse effects when in contact with components of living organisms, such as proteins, biological cells and tissues. This review deals primarily with the ‘bio/blood compatibility’ of polymeric materials with various biological elements in systems involving direct blood contact.
Protein adsorption on the polymer surfaces
One important property of bio/blood compatibility is the amount of serum proteins adsorbed on the polymer surface. Commonly used methods to determine the amount of adsorption include infrared spectroscopy (IR), ultraviolet spectroscopy, X-ray photoelectron spectroscopy, radioisotope-labeled immunoassay (RI) and circular dichroism.23, 24, 25, 26, 27, 28, 29 In situ studies on adsorption properties, RI-labeling and fluorescent-labeling techniques have been reported. One method that has been successfully applied to gain information on adsorbed proteins is total internal reflection fluorescence spectroscopy. The high sensitivity of fluorescence spectroscopy allows the quantification of small amounts of adsorbed proteins, including their competitive adsorption, interfacial conformation changes and surface mobility. Surface plasmon resonance has been applied to the in situ detection of adsorbed protein. In addition, using a quartz crystal microbalance (QCM) has been recommended as an effective and easy method for analyzing in situ biomolecular interactions.30, 31 QCMs are sensitive devices that can measure mass in air or aqueous solution. The resonance frequency of the QCM electrode decreases linearly with increasing electrode mass due to the adsorption of some compounds and is sensitive at the nanogram level. Using this method, several researchers have reported interactions between polymeric biomaterials and proteins.
Protein adsorption behaviors on various polymer surfaces have been extensively investigated. Research has shown that the important factor in the bio/blood compatibility is not the amount of protein adsorbed on the surfaces but the structure or orientation of the adsorbed proteins.25 Much research has focused on whether the protein adsorption is reversible. Many reports insist on the irreversible adsorption of proteins on the polymer surface,26 whereas other researchers have reported the reversible adsorption.32 Thus, the kinetics of the adsorption of proteins on polymer surfaces must be analyzed in addition to the quantity of protein adsorbed when the bio/blood compatibility of polymers is discussed.27
To explain the excellent bio/blood compatibility of PMEA, the quantity of plasma protein and the kinetics by which it is adsorbed on the PMEA in the early stages of adsorption were investigated along with the secondary structure of the protein. The quantity of protein adsorbed on the PMEA was small and similar to that adsorbed onto poly(2-hydroxyethyl methacrylate) (PHEMA).30, 31 circular dichroism spectroscopy revealed a significant conformational change in the proteins adsorbed on the PHEMA, whereas the conformational change in the proteins adsorbed on the PMEA was small.31 Using QCM measurements, we investigated the adsorption/desorption behavior of proteins on the PMEA surface in terms of their binding constants and association and dissociation rates. The circular dichroism and QCM results suggested that the excellent bio/blood compatibility of PMEA is associated with the low denaturation and the high dissociation rates of the proteins attached to the PMEA.31 A schematic representation of the proposed protein (bovine serum albumin, fibrinogen and Immunoglobulin G) adsorption on PMEA, PHEMA and polypropylene is presented in Figure 2.
The adhesion force between the PMEA and fibrinogen and between the PMEA and bovine serum albumin were measured via atomic force microscopy (AFM; see Self-assembled monolayers (SAMs)). The PMEA surface exhibited almost no adhesion to the native protein molecules.31 The denaturation of the adsorbed protein could lead to platelet activation and subsequent thrombus formation: When the protein molecule that adsorbs onto a polymer surface retains its native conformation, platelets cannot adhere to the surface. A polymer surface that does not denature proteins would be bio/blood compatible.
Structure of water
Although considerable experimental and theoretical efforts during the past few decades have been devoted to clarifying the structure of water, the factors responsible for the bio/blood compatibility of polymers have not yet been elucidated.33 Water molecules serve as a medium for adhesion and have a role in cell morphology and other cellular functions. Water is thought to be a fundamental factor in the biological response to artificial materials. Many researchers have insisted that the structure of water on a polymer surface is a key factor in bio/blood compatibility. However, the proposed structures and/or the functions of water differ in many cases, and little consistency can be found among structures. Detailed studies of the dynamics and structures of hydrated polymers is needed to clarify the underlying mechanism for the bio/blood compatibility of polymers.
The hydrated water in a polymer can be classified into three types: free water (or freezing water), freezing-bound water (or intermediate water), and non-freezing water (or non-freezing-bound water) (Figure 3). The hydrated PMEA possessed a unique water structure, observed via differential scanning calorimetry (DSC) as the cold crystallization of water (Figure 4). Cold crystallization is interpreted as ice formation at temperatures below 0 °C, a characteristic of intermediate water in PMEA. The presence of three types of water in PMEA is supported by attenuated total reflection infrared (ATR-IR) spectroscopy34 and nuclear magnetic resonance (NMR).35 The intermediate water molecules interact weakly with the methoxy group in PMEA (see Spectroscopic analysis). An investigation of the principal factor responsible for the excellent bio/blood compatibility of PMEA must include the intermediate water structure on the polymer surface.
When the polymer surface encounters blood, it first adsorbs water before adsorbing the serum proteins, and a specific water structure is formed on the polymer surface. If the resulting structure is the first layer, then the layers will deposit in the following order: polymer surfacenon-freezing water(intermediate water)free waterbulk water (Figure 5). The free water is unlikely to activate the system and is unable to shield the polymer surface or the non-freezing water on the polymer surface, because free water freely exchanges with bulk water, resulting in a structure similar to bulk water. Because intermediate water is weakly (loosely) bound to the polymer molecule or to non-freezing water (tightly bound water), this layer forms a more stable structure than free water. Thus, when the intermediate water layer becomes sufficiently thick, it prevents the cell or protein from directly contacting the polymer surface or the non-freezing water.
This hypothesis is supported by several reports demonstrating the formation of cold crystallizable water (in intermediate water) in well-known bio/blood compatible synthetic and natural polymers such as poly(ethylene glycol), polyvinylpyrrolidone, poly(methylvinyl ether), poly(2-methacryloyloxyethyl phosphorylcholine), poly(tetrahydrofurfuryl acrylate), poly(2-(2-ethoxyethoxy)ethyl acrylate), other biocompatible synthetic polymers, and gelatin, albumin, cytochrome c and various polysaccharides, including hyaluronan, alginate, gum and other natural polymers.36, 37, 38, 39, 40, 41 By contrast, no cold crystallization of water was observed in hydrated PMEA-analogous polymers, which do not exhibit excellent bio/blood compatibility. Based on these findings, the intermediate water, which prevents the biocomponents from directly contacting the polymer surface or the non-freezing water on the polymer surface, has an important role in the excellent bio/blood compatibility of PMEA. We proposed an ‘intermediate water’ concept: the water exhibited both a clear peak for cold crystallization in the DSC chart and a strong peak at 3400 cm−1 in the time-resolved IR spectrum; the localized hydration structure consisted of the three hydrated waters in PMEA.
Spectroscopic methods using electromagnetic waves in different wavelength regions have been applied to explore the structures, functions and properties of various types of molecules, including bio-related molecules and medical polymers.42, 43, 44, 45, 46, 47, 48 This section provides a review of the spectroscopic analyses of biomaterials, especially those on hydration structures at biointerfaces. As previously described, sorbed water in a PMEA matrix can be classified into three different types via DSC.20, 21, 22 However, detailed hydration structures of each water molecule on the polymer chain at a functional group level cannot be revealed by this method due to thermal analysis. Spectroscopic methods are promising for the investigation of such molecular structures and the molecular interactions among the molecules and functional groups in the system.
The crystalline structures of hydrated water in a polymer matrix below 0 °C have been investigated using X-ray diffraction and IR spectroscopy. Kishi et al.49 demonstrated the simultaneous use of DSC and X-ray diffraction with hydrated biomaterials. Freezing-bound water (intermediate water) in PMEA exhibited the growth of a hexagonal ice structure at the cold crystallization temperature during the heating process.49 Although their detailed interpretations differ from ours, Gemmei-Ide et al.50, 51, 52, 53 reported the temperature-dependent IR spectra of hydrated polymer solids at similar temperature variations to the DSC measurements. No ice formation occurred during the cooling process, and an ice-like structure appeared during the process of heating the hydrated PMEA with rapid temperature variations of 5.0 K min−1.51
The dynamic behavior of both the hydrating water molecules and the hydrated polymer chains have been explored using NMR spectroscopy.54, 55 Results of 2H-NMR and 13C-NMR on hydrated PMEA and PHEMA have been reported by Miwa et al.35, 56 and are as follows: Non-freezing water exhibits low mobility due to a strong interaction with the polymer chain. By contrast, free water has high mobility due to its location far from the polymer chain. Freezing-bound water (intermediate water) has intermediate mobility compared with the other two types of water in PMEA. The flexibility of bound water or the PMEA chain is relatively high compared with that of PHEMA, although the EWC (equilibrium water content) for PMEA is smaller than that for PHEMA. One can conclude from the NMR spectroscopy that the flexible water loosely bound to the polymer chain prevents protein adsorption at the biointerface.
Hydration structures of water molecules bound to polymer chains at a functional group level have been investigated using vibrational spectroscopy, especially IR spectroscopy combined with quantum chemical calculations.57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67 To obtain the IR spectra of hydrated water in a polymer matrix excluding the information on the bulk water contacting a polymer surface, Morita et al.34 applied ATR-IR spectroscopy. Figure 6 provides a schematic illustration of a custom in situ ATR-IR flow trough cell designed by our research group. A polymer film is coated on the flat surface of a hemispherical prism with a large refractive index, for example, ZnSe, Si or Ge via a solvent-casting method. The thickness of the film is controlled so that it is thicker than the penetration depth of the near field light (evanescent wave) generated at the prism/film interface. The time-resolved ATR-IR spectra of the sorption process of water into the polymer film were recorded. Before measurement, the film sample was sufficiently dried using a nitrogen gas flow into the cell. After initiating the time-resolved acquisition, water vapor or liquid water was introduced into the cell.
Figure 7 shows the time-resolved ATR-IR spectra in the O-H stretching region of the sorption process for liquid water into a PMEA film collected every 1.86 s. A gradual increase is observed in the 3700–3000 cm−1 region with a complex spectral shape variation. This broad feature with several peaks overlapping in the O-H stretching region is due to the hydrogen-bonds network among the water molecules.68, 69 To deconvolute the information in the complicated spectral variation, computational analyses based on the multivariate analysis of chemometrics70 and perturbation-correlation moving-window two-dimensional correlation spectroscopy71, 72 were applied to obtain the time-resolved spectra.73, 74 Figure 8 shows the pure spectra of non-freezing water, freezing-bound water and freezing water, and their concentration profiles calculated from the time-resolved spectra shown in Figure 7 using multivariate curve resolution analysis with an alternating least squares technique. Non-freezing water exhibits a relatively high wavenumber contribution at approximately 3600 cm−1. This high wavenumber contribution in the O-H stretching region is generally observed for isolated molecules without hydrogen-bonds network or for molecules that hydrogen-bond to carbonyl groups.75 By contrast, free water has a broad feature centered around 3400–3200 cm−1 similar in spectral shape to bulk water. The freezing-bound water demonstrates an intermediate vibrational frequency at 3400 cm−1 as a peak maximum in a broad feature, implying that freezing-bound water possesses a smaller water cluster than bulk water because of its higher wavenumber contribution. These describe that water molecules bound to the polymer chain are characterized by O-H stretching bands in the IR spectra. By contrast, information regarding hydrated polymer chains can be interpreted from the IR bands assigned to the polymer chains, such as the C=O stretching band at 1730 cm−1. The C=O stretching shows two contributions of free C=O that do not hydrogen-bond and hydrogen-bonded C=O. The band assigned to hydrogen-bonded C=O increases with time similar to that assigned to non-freezing water, which implies that the non-freezing water interacts with the C=O group in the PMEA side chain. The band assigned to the O-CH3 rocking band in the methoxy moiety also exhibits a peak position shift over time that is associated with the intensity variation of the freezing-bound water, which demonstrates that the freezing-bound water interacts with the polymer chain at the methoxy moiety in the side chain terminal. Figure 9 illustrates the hydration structures of three types of hydrated water on the PMEA chain revealed using time-resolved ATR-IR spectroscopy. Many non-freezing water molecules (85.6%) interact with two carbonyl groups in a C=O···HOH···O=C type of hydrogen-bonding interaction. Thus, non-freezing water does not freeze even below −100 °C. However, this water is easily dehydrated in a nitrogen atmosphere, although water tightly bound to the ionomer dehydrates only slightly under these conditions.76
This result leads to a simple question: Why does PMEA exhibit excellent bio/blood compatibility, even though the EWC for the polymer is only 9%? To clarify this point, the hydration structure of a PMEA model monomer of 2-methoxyethyl acetate (MEAc) was investigated using ATR-IR spectroscopy.77 Figure 10 illustrates the hydration structure of MEAc compared with that of PMEA as determined using ATR-IR spectroscopy. The concentration-dependent ATR-IR spectra of water dissolved in MEAc show a spectral variation similar to the time-resolved ATR-IR spectra of the water sorption process into a PMEA matrix. An approximately 40 wt% solution of water in MEAc yields a band shape in the O-H stretching region similar to that of 9 wt% in PMEA. The hydrating water molecules on the PMEA chain at an EWC of 9 wt% have hydration structures at the functional group level similar to those of water molecules homogeneously mixed in MEAc at 40 wt%. Thus, the water in PMEA can be considered inhomogeneous, and the localized water cluster is approximately four times more concentrated than the homogeneous water in MEAc. In addition, the hydrated region in the PMEA matrix is phase separated from the dehydrated (that is, segment rich) region. Our recent study using sum-frequency generation spectroscopy revealed that hydrated water at the PMEA/water interface is more concentrated than that in the PMEA bulk. One could expect that hydrated water in PMEA is four times as concentrated at the biointerface with a small water cluster and flexible mobility.
Compared with polymers, SAMs formed on solid surfaces have not found many practical applications. However, SAMs have provided model platforms to investigate interactions of organic surfaces with biomolecules, living cells and tissues because of their high ordering and well-defined structures and the ease of controlling the physicochemical properties of their surfaces (Figure 11).78 In particular, SAMs have enabled us to explore the structure and dynamics of interfacial water because the SAM–water interface is rigorously defined compared with the interface of polymer systems.79, 80 Herein, we review articles on the behavior of water near protein- or cell-resistant (nonfouling) SAMs.
Several types of nonfouling SAMs have been reported, for example, oligo(ethylene glycol) (OEG)-81, 82 and poly(ethylene glycol)-terminated83, 84, 85 SAMs, and zwitter ionic86, 87, 88, 89 SAMs (we cite only a few representative articles). Although the chemical structures of the terminal groups of the molecules constituting these SAMs are different, they all exhibit nonfouling behavior. However, the physics underlying their bioinertness has remained a matter of intense debate. Thus far, the origin of the bioinertness of surfaces covered with hydrophilic polymers is thought to be due to an elastic effect arising from the flexibility of the polymer chains and an osmotic effect arising from the tight bonding of the water molecules tightly bound to the polymer chains (so-called steric repulsion).90 In water, the polymer chains immobilized on the substrate are fully hydrated, and the degree of freedom of the polymer chain is high. When biomolecules (or cells) approach the substrate to adsorb onto it, the biomolecules must push the polymer chains aside, partially dehydrate the polymer chains and approach the surface. Such a process is entropically and enthalpically unfavorable. Apart from the poly(ethylene glycol)-terminated SAMs, the molecules constituting the above-mentioned nonfouling SAMs are densely packed with low conformational freedom in their monolayers. Therefore, the idea of steric repulsion cannot be applied to SAM systems. By combining the previous experimental and theoretical findings, we have reached one conclusion: the nonfouling behavior of the SAMs cannot be explained by electrostatic repulsion; rather, the structures and dynamics of the interfacial water, which never appear in macroscopic surface wettabilities, must be involved.80 We review previous attempts to explore the interfacial behavior of water in the vicinity of SAMs, in particular, those relevant to the mechanism of nonfouling. We summarize recent research in this field in Table 1.
SAMs of OEG-terminated alkanethiols on gold (hereafter denoted as OEG-SAMs) are the most widely used nonfouling SAMs used for protein and cell patterning and for the enhancement of biosensor selectivity since they were first reported by Prime and Whitesides.81, 82, 91 Although many reports have discussed the origin of the bioinertness of OEG-SAMs, the underlying mechanism has remained unclear. The protein resistance of the OEG-SAM (in this case HS-(CH2)11-(O-CH2-CH2)3-OCH3: EG3-OMe) was first discovered to strongly depended on the substrate, which determined the density of the thiolate molecules in the monolayer, that is, a densely packed Ag-supported monolayer adsorbed proteins, whereas an Au-supported monolayer with low molecular density deterred protein adsorption.92 Feldman et al.93 investigated the interaction of an AFM probe functionalized with fibrinogen against the Ag- and Au-supported SAMs. They observed strong adhesion between the probe and the Ag-supported SAM, whereas a strongly repulsive force was observed in both the approaching and retracting curves with the Au-supported SAM.
Since that time, other experimental investigations have indicated remarkable differences in the interfacial behavior of molecules. Harder et al.92 found that the OEG moieties in EG3-OMe molecules adopted an all-trans conformation on an Ag substrate, whereas the OEG moieties in the Au-supported monolayer adopted a helical conformation. This finding was supported by the results of a computer simulation performed by Pertsin et al.,94 which also suggested that water molecules penetrated into the SAM and were trapped by the oxygen atoms of the OEG moieties via hydrogen bonds. By contrast, the Ag-supported SAM did not allow penetration by the water because of its high molecular density in the monolayer. Herrwerth et al.95 systematically altered the hydrophilicity of the terminal groups and the number of EG units and found that protein resistance is largely governed by the molecular density in the SAMs, that is, the accommodation of water in OEG-SAMs.
Although the previously mentioned studies clearly indicate that water molecules at the SAM–water interface have a pivotal role, a detailed description of the role of water in the protein resistance of nonfouling SAMs has not yet been presented. Computer simulations by Jiang and colleagues provided direct evidence of the water-induced repulsion operating between lysozymes and several protein-resistant monolayers, including OEG, zwitterionic and sugar-based SAMs.96, 97, 98, 99, 100, 101, 102 These simulations determined that interfacial water molecules in the vicinity of these SAMs prevent the adsorption of lysozyme as a physical barrier. More importantly, the repulsion of proteins is induced not only by the tight binding of the water molecules to the terminal groups of the thiolate molecules but also by the layer of water molecules in the interfacial region (its thickness is 2–3 nm, roughly 6–10 layers of water molecules). Moreover, the dynamics of the water is approximately seven times slower than that of the other protein-adsorbing hydrophilic SAMs and of bulk water.101
As previously discussed, there is no reason to doubt the assertion that the structure and dynamics of the interfacial water is a key factor in a polymer’s bioinertness. However, we are still far from a full understanding of the behavior of interfacial water, partially because differences in structure and dynamics of interfacial water are not reflected in macroscopic surface wettability (for example, the water contact angle).95 To study the water structure at the microscopic level, rapid progress has recently been made in the frequency modulation non-contact AFM (NC-AFM) techniques to reveal the hydration structure of hydrophilic surfaces ranging from mica103, 104 to lipid bilayers105, 106, 107 and to SAMs.108 The hydration structures found in these systems were quite similar, although the surfaces exhibit different protein-adsorbing characteristics and bio/blood compatibility. At most, only three hydration layers were observed using NC-AFM.
We performed surface force measurements via AFM using colloid-type probes and measured the interaction between OEG-SAMs to investigate the behavior of the interfacial water and ions.109 Water-induced repulsion operated between the OEG-SAMs from SAM-to-SAM distances smaller than 4–6 nm (Figures 12 and 13), whereas no such repulsion was observed in the other SAMs, which indicates that a stable layer of structured water molecules with a thickness of 2–3 nm exists in the vicinity of the OEG-SAM (Figure 14). This finding is in good agreement with the results of the computer simulation by Jiang and Cao17 Surprisingly, the repulsion largely depended on the number of EG units (Figure 13). In addition, a clear correlation was found between the SAM–SAM interaction and the platelet adhesion. These results re-emphasized the differences in the interactions on the nanoscale and that anti-platelet adhesion cannot be predicted from water contact angles.
These results also imply that NC-AFM might not detect the behavior of interfacial water molecules higher than the second or third hydration layer, which does not appear in the local density profile of water, although the results of surface force measurements and computer simulation suggest that a water layer with a thickness of 2–3 nm (6–10 layers of water) is necessary for bioinertness (Figures 12 and 13). Our recent vibrational analysis of water in the vicinity of a SAM using ATR Fourier transform infrared spectroscopy, which will be published elsewhere, revealed that the shapes of the O-H stretching bands of the interfacial water observed in the spectra of OEG-SAMs were obviously different from those observed for other SAMs (SAMs of the CH3-, OH-, NH2- and COOH-terminated alkanethiols). The behavior of the interfacial water specifically responsible for the bioinertness is beginning to be understood.
In this manuscript, we summarized the findings of recent research into the mechanism underlying the nonfouling of SAMs. Further analytical work using interface-sensitive methods (for example, surface force measurements and vibrational spectroscopy) and computer simulations will allow a deeper understanding of the mechanism and lead to directional design of polymeric biomaterials in the future.
This work is supported by Grants-in-Aid and Special Coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science and Technology, Japan. We gratefully acknowledge the financial support from the Funding Program for Next Generation World-Leading Researchers (NEXT Program, Japan).
About this article
Annals of Biomedical Engineering (2017)