a Faculty of Pharmacy, University of Sydney, Sydney, Australia. E-mail: firstname.lastname@example.org
Implantation of foreign materials can elicit an immune response proportional to the cytotoxicity of the material. Surface modification is an important step to improve biocompatibility and biodegradability properties to the surface of foreign materials. In addition, important biological functions can be integrated in such bio-inert materials. Therefore, the immobilization of biological compounds, such as proteins and peptides, is a straightforward method to modify their surfaces. In this chapter we discuss how different protein immobilization methods can influence cellular adhesion, proliferation and differentiation. Furthermore, we describe the chemistry behind the protein immmobilization step; specifically, we describe physisorption, covalent bonding, and ‘linker-free’ immobilization approaches. In each of these sections for protein immobilization, we have included a case study with pros and cons so the reader can follow these protocols and readily apply them in the laboratory.
It is already well-established that cellular responses to foreign materials when implanted in vivo are guided mainly by the surface characteristics. The modulation of characteristics such as roughness, porosity, chemical and biological composition allows the regulation of material integration within the body as well as the guidance of specific responses, e.g., cell adhesion, cell detachment, cell proliferation, differentiation, or metabolic activity. Material integration within the body is critically important for orthopedic implantable devices. Desired integration results in successful single surgery, thus reduces costs and complications related to adverse reactions and implant revisions. In this chapter, we described the methods of surface modification and functionalization, which enable the effective regulation of stem cells behavior for orthopedic applications. In particular, the chapter focuses on immobilization of different signaling molecules such as proteins, peptides, growth factors on the surface and, in three dimensions, to modulate cell adhesion and to trigger and regulate osteogenesis, osteointegration, and osteoregeneration. Furthermore, examination methods for the immobilization of biomolecules on surfaces are described in detail to explain the advantages and disadvantages of each method.
Implantation of exogenous materials into the body triggers common body responses, and in some cases adverse inflammatory reactions. The body tries to ward-off the foreign object and protect from its ‘perceived’ negative influence on healthy cells and tissues. As a consequence the exogenous materials are encapsulated in a vascular, fibrous sacks which often limit the functionality of the implanted material. In the last decades research has been focused on the possibility of addressing this problem and creating surfaces that mimic the body's environment, thus enabling the host cells to accept and interact with the foreign object. Such interactions are facilitated by different cues incorporated into the surface, which includes topographical, mechanical, structural, chemical and biochemical signals. These are capable of stimulating specific cellular responses which include cell adhesion, proliferation, differentiation and, ultimately, cell death. The design of specific cues, single or combined, is guided by the application and tailored for specific needs.
Historically, one of the most successful approaches, which can be called biomimetic, was a deposition of hydroxyapatite coatings on the surface of implants that were placed within the bone environment. Hydroxyapatite, a natural component of bone tissues, when present on surfaces of metallic or polymeric devices allows for significant improvement of implant integration within the body. The most significant limitation is the adhesion of the coating to the substrate. It has to be noted that some of the implants are heavily bent (pre-operatively) and ‘hammered’ into the host tissue. This creates significant sheer and bending stresses that can lead to failure of the coating. Nevertheless, hydroxyapatite coatings have been clinically used and verified.1
The cross-talk between a bioactive layer and the tissue can similarly be achieved using different types of mineral, ceramic, glass–ceramic materials that include tri-calcium phosphate, bioglasses and phosphate glasses. These materials are coated onto the surface or are used as a base component to produce implants. The primary advantage of these materials is that they can be degradable and compounded with different ions (Zn, Ag, Co, Sr), which provides additional biochemical cues for cells at the interface. Zn, Co and Ag have been reported to have antimicrobial activity,2 while Sr and Zn have been demonstrated to support bone formation and are used clinically in the treatment of osteoporosis.2–4 Furthermore, each of these materials interacts with body fluids differently and can guide the adsorption of proteins and growth factors and other components from the body fluids and blood. Subsequently, these biomolecules modulate interactions with cells and tissues. Hence, by a careful selection of the material it is possible to stimulate cellular responses through the structural and chemical composition, and also to guide the adsorption of desired biomolecules that stimulates biological responses, e.g., adsorption of cell adhesive proteins to hydroxyapatite.2,3
The communication of cells with surfaces is mediated by pre-adsorbed protein layers. The composition of this layer is critical to obtain specific responses such as cell adhesion/repulsion, or differentiation. It is well established that specific proteins, peptides, growth factors and drugs can stimulate cells.5–10 These biomolecules are used to enhance the implant integration and also treat some dysfunction/diseases when released from the surface.11–13 The most common approach to enhance cell adhesion to the surface of implantable materials using biomolecules is to incorporate the cell binding motif Arg-Gly-Asp (RGD).13–15 RGD, together with the integrins that serve as receptors for them, constitute a primary recognition system for cell adhesion. The RGD sequence is the cell attachment site of many adhesive extracellular matrix, blood, and cell surface proteins, and nearly half of the over 20 known integrins recognize this sequence in their adhesion protein ligands.16 The integrin-binding activity of adhesion proteins can be achieved by engineering short synthetic peptides that contain the RGD sequence. Interestingly, it was reported that such peptides promote cell adhesion when immobilized onto a surface, and inhibit it when presented to cells in solution.16 Importantly, integrin-mediated cell attachment regulates further cell migration, growth, differentiation, and apoptosis.14,15 The ability to target cell adhesion receptors, which are responsible not only for cell–matrix adhesion but also for signaling bidirectionally across the cell membrane, provides an opportunity to design new drugs (Figure 1.1). Because integrins are involved in many biological processes such as angiogenesis, thrombosis, inflammation, osteoporosis and cancer, these drugs, which are based on the RGD structure, can be used for the treatment of diseases such as thrombosis, osteoporosis and cancer17–20 (Figure 1.1).
It is well established that tethering of protein and peptides facilitates communication with the cell. In fact, adhesion of cells is a prerequisite for the subsequent proliferation and differentiation of cells.7,21 Recent studies by Webster demonstrated that adsorbed proteins including fibronectin, laminin and vitronectin play an important role in limiting bacteria colonization.22 Hence fibronectin-functionalized surfaces can be considered as multi-functional with a dual purpose: preventing bacteria adhesion and enhancing cell adhesion.
While fibronectin and vitronectin receive the most attention to enhance cell adhesion to surfaces of biomaterials, there are several types of proteins and peptides that have been attempted to use for the same purpose. The use of fibronectin and vitronectin is primarily dictated by their composition, in which the RGD sequence plays a critical role.23 RDG is recognized and bind via integrins (α5β1, αVβ1, αVβ3, αVβ5, αVβ6, αVβ8 and αIIbβ3) cell adhesion receptors that bind to the extracellular matrix (ECM) proteins.16,18,23–25 It has been demonstrated that laminins and collagens also contain RGD sequences but these are inaccessible, thus are not typically used for surface modifications.
Many integrins are expressed in various tissues; however there is some population which are expressed only in a certain type of cell or tissues. The integrin receptors, which were found in human osteoblasts and are typically used to regulate bone cell responses, are the fibronectin receptor (α5β1),16,18,23–27 vitronectin receptor (αVβ3),28 and the type I collagen receptor (α2β1).16,18,23–27 For this reason modification—functionalization—of implant surfaces with biomolecules/protein to mediated cell adhesion to substrates via integrins has recently become one of the most interesting approaches in the development of new biomaterials including drug carriers that are capable of targeting specific sides.15,29–33
It has been already suggested that the RGD peptide, which interacts with the αVβ3 and αVβ5 integrin sub-units, commonly associated with vitronectin, increases biointegration of implants. Matsuura also demonstrated that RGD contributes to the osteoconductive effect of hydroxyapatite more than titanium.34 This phenomenon is associated with higher affinity and adsorption capability of protein to hydroxyapatite surface than to titanium surface.
The positive effects of the RGD peptide on regulation of cell adhesion have been confirmed in many in vitro as well as in vivo studies. Elmengaard35 has shown that RGD peptide-coated porous-coated titanium implant significantly increased bone formation on and around the implant. In this study a cyclic RGD (Figure 1.2), which interacts with both αVβ3 and αVβ5 integrin sub-units has been immobilized on unloaded press-fit titanium implants. Not only was the increase in bone formation observed, but also a significant reduction in presence of fibrous tissue around the implant was evidenced.35 Because the RGD coating can be relatively easily applied to the surface it offers a cost-effective way to enhance the early osseo-integration of press-fitted clinical implants.
One of the interesting features of this approach was the use of cyclic peptide. Cells typically show higher integrin binding affinity to the cyclic RGD sequence and previous studies have shown that cyclic peptides are more stable when immobilized to complex and three-dimensional substrates.39,40 Another important feature of cyclic RGD peptides is their high selectiveness toward the αV integrin, which regulates activation of the osteoblast-specific transcription factor (core binding factor alpha-1/runt related transcription factor 2 during osteogenesis).41 These peptides show also more resistance to enzymatic cleavage,42 and may also promote a more stable cell–ligand bond. The integrin affinity and specificity to the RGD peptide is affected by both steric conformation and the amino-acid sequences flanking the RGD peptide.43,44 Mooney40 has also demonstrated that a linear RGD peptide promoted osteogenic differentiation of preosteoblasts (MC3T3-E1) but did not induce differentiation of hBMSCs or D1 stem cells. At the same time matrices that presented the higher-affinity cyclic form of this adhesion ligand enhanced osteoprogenitor differentiation in three dimensions.
In vivo studies by Elmengaard35 on press-fit titanium alloy implants coated with cyclic RGD and its effect on bone ongrowth showed that after 4 weeks, cyclic RGD coating significantly stimulated bone formation directly at the interface. A two-fold increase in bone growth was found for RGD-coated implants compared to the uncoated. Importantly, RGD-coated implants showed less fibrous tissue around the implant, which is considered very positive. However, a significantly higher amount of bone for RGD coated implants was observed mostly at the interface (0–100 μm), and the difference in bone volume between both coated and uncoated groups gradually decreased with distance from the surface. There were no significant differences between both groups at a distance of 750 μm from the implant surface. Furthermore, an increase in the mechanical fixation was also observed; apparent shear stiffness was significantly higher for RGD-coated implants. A similar study on the osteoconductive hydroxyapatite coating with titanium alloy implants using the same press-fit implant model showed only a significant increase in bone ongrowth but no difference in mechanical fixation.45
To immobilize RGD on amino-functionalized glass surfaces Dechantsreiter et al.43 suggested the use of isothiocyanate-terminated peptides. It was demonstrated that RGD peptides were fused to an isothiocyanate anchor during synthesis and bound to amino-terminated surfaces. Importantly, two types of linear peptides and one cyclic peptide were investigated and were shown to enhance cell spreading and induce the formation of focal adhesions in murine fibroblasts. The authors also concluded that formed adhesions were specific because cells did not recognize the corresponding negative control peptides and did not spread in the presence of soluble H-RGDS-OH peptide. This coupling method was shown to be effective also for patterning, where cells selectively recognized areas coated with RGD-containing peptides.
Better integration and deposition of the bone on the surface is directly linked to pullout strength of the implants. O'Toole suggested the use of bone sialoproteins (BSPs) to improve the pullout strength.46 In his approach two scenarios were considered: (1) BSPs were coated on the surface and (2) BSPs were not placed directly on the surface but BSP-containing gelatin was used as a plug where the implant was placed. Results failed to demonstrate the benefits of the BSP. In general, BSP-coated acid-etched implants perform more poorly, mechanically, than do uncoated implants. It was concluded that BSP forms an insulating barrier on the implant surface, preventing the direct apposition of the bone on the implants. At the same time, histology tests showed that the BSP coating failed at the BSP–implant interface but osteoinductive behaviors were confirmed at the BSP–bone interface. It was postulated that a better method of coating is required to allow the formation of bone at the interface with the implant. Similar results were found for the second group where BSP-containing gelatin was used. These implants performed poorly both histologically and mechanically. Histologically, osteoid, osteoblasts and osteocytes are stimulated by the presence of BSP, but it was not observed at the surface. This distant osteoinduction does not correspond with better mechanical performances when implants are subjected to pullout testing analysis.
With respect to proteins, peptides benefit from a lower immunogenic activity as well as from the ability to be synthesized and handled. In addition, over the last decade, highly active and ανβ3- and ανβ5-integrin selective cyclic pentapeptide ligands such as cyclo(-RGDfX) have been developed.13 It was thus demonstrated that in addition to the RGD binding sequence a d-amino acid, especially d-Phe following the Asp residue in the cycle, is essential for high activities and αν selectivity. Cyclic pentapeptides with d-amino acid in other positions and/or a non-hydrophobic amino acid following Asp as well as linear peptides have lower activity and are less selective towards αν integrins. Our results clarify that osteoprogenitor cells exhibit a differential binding to RGD peptides displaying a specific conformation (linear RGDC and cyclo-DfKRG). Such a behavior has to be related to the different signal transduction pathways implied by both peptides. Indeed, some of us14 have shown that linear GRGDSPC peptide interacts preferentially with αχβ1 integrins while cyclo-DfKRG peptide interacts with ανβ3 and ανβ5 integrins. Consequently, different cell adhesion at 24 h seeding may be linked to the different cellular activity, as extracellular proteins synthesis, implied by the signal transduction pathways that both RGD-containing peptides induce, respectively. In addition, cell adhesion behavior has to be related to the accessibility of integrins receptors by peptides displaying conformation. However, both cyclo-DfKRG and linear RGDC appear to be good candidates for developing hybrid biomaterials made of titanium alloys and human osteogenic cells.
Besides the improvement of cell adhesion by means of adhesive proteins or peptides, huge efforts are currently directed toward the development of osteoconductive/osteoinductive materials. Osteo-conductive materials are able to enhance the formation of new bone, i.e., accelerating the bone development or improving it from a physiological point of view. Osteo-inductive materials trigger the osteogenic process without external stimuli (e.g., supplements in osteogenic media) and are therefore more promising for a rapid healing process. To assess these properties in vitro, stem cells are commonly cultured in basal and osteogenic medium and several markers of differentiation are evaluated. For in vivo studies, they are tested by implantation in ectopic and orthotopic bone site and assessing the bone formation. Several directions can be followed for obtaining an osteoinductive material. Few research groups have successfully modified the physical and chemical properties of an inert material, such as composition, elasticity and topography, in order to provide the inert material with osteoinductive capabilities.47–56 Another approach is based on the delivery of biomolecules, mainly growth factors, integrated into the scaffold and able to control osteogenesis, bone tissue regeneration and ECM formation through specific cellular pathways in the implant site.57 For example, insulin-like growth factors (IGFs) and transforming growth factor-β (TGF-β) affect the migration and recruitment of different bone cells involved in bone healing, respectively.58,59 Bone morphogenetic protein (BMP) influences the osteogenic differentiation of progenitor bone cells, and vascular endothelial growth factor (VEGF) enhances the formation of functional blood vessels. Recently, small peptides resembling the functionality of physiologically active proteins have grown an increasing interest. Therapeutic peptides overcome many drawbacks of macromolecules, such as long synthesis time, high production costs, low stability, limited long-term efficiency, or risk of unforeseen side effects. Small peptides carrying the active site of several growth factors have already been developed. For example, Tanihara and co-workers found that a 20 amino acid peptide sequence from the ‘knuckle’ epitope of BMP-2 shows the biological activity of the full-length BMP-2 protein.60 Similarly, a VEGF-mimetic peptide has been synthesized and has shown retaining the biological activity of VEGF protein.61 Moreover, several peptides able to inhibit bone resorption have been synthesized after the identification of osteoprotegerin (OPG) and RANKL, important elements in the bone remodeling pathway.62
During the last few years orthopedic researchers aiming to regenerate and repair bone have realized that the appropriate function of engineered bone substitutes cannot rely on the sole diffusion of nutrients and oxygen.63 The lack of a functional vascular network results in the formation of necrotic cores within the bone substitute.64,65 Therefore, engineered bone scaffolds with clinically relevant sizes require a vascular network for the successful engraftment and survival of the bone substitute. However, host-derived vascularization of implanted constructs is largely limited by the capacity of host cells to invade and form in situ capillaries.66 Creating functional and perfusable microvascular systems, replicating the structure, biological properties, heterotypic cell interactions and biomechanics of native microvascular environment, represents nowadays an important bottleneck of bone tissue engineering strategies. Below we describe important progress in strategies to create vascularized tissues using biological cues.
Biological techniques for microvessel formation rely on heterotypic cell–cell interactions and the cross-communication of cells with encapsulated growth factors in the extracellular microenvironment.63 Endothelial cells co-cultured with human mesenchymal cells (hMSCSs) or perivascular cells have resulted in the formation of microvascular networks.63,66–71 This tendency becomes more pronounced when endothelial and progenitor cells are encapsulated in ECM-like hydrogels. These types of hydrogels can be chemically modified to allow homing mechanisms to cells and facilitate the cross-talk of cell and the encapsulating matrix. Recently, Chen developed vascularized methacrylated gelatin hydrogels by encapsulating blood-derived endothelial colony forming cells (ECFCs) and bone marrow-derived MSCs in a hydrogel matrix.71 The results demonstrated an extensive formation of capillary-like networks with lumens within the hydrogel structure. Pre-vascularized hydrogels have the potential to be combined with osteoconductive materials to trigger angiogenesis in bone substitutes.
Presenting angiogenic growth factors and signaling markers such as VEGF can trigger the microvessel formation process. D'Andrea et al. reported the development of a synthetic short peptide of 15 amino acids called QK peptide which contained a VEGF binding domain to activate VEGF receptors in endothelial cells. This peptide was used to induce endothelial proliferation and VEGF signaling mechanisms on a Matrigel substrate.61 Synthetic small peptides have been further modified to incorporate acrylate motifs to facilitate the functionalization of VEFG-like peptides in PEG hydrogels.72 Additionally, Chiu et al. guided endothelial cell proliferation in a microfabricated chitosan–collagen hydrogel rich in Tβ4, an angiogenic and cardio-protective peptide that enhances cardiomyocyte survival.73,74 This approach enabled capillary-mediated anchorage and anastomosis of arteries and veins.73 In a recent study, Kim et al. integrated a microfluidic chip perfused with a fibrin gel to perform a comprehensive angiogenesis study by endothelial cell-mediated formation of perfusable microcapillaries.75 Also, microvessels generated in a microfluidic chip presented morphological and biochemical cues replicating those in native blood vessels and capillaries niches.75 The combination of these technologies offers the potential to create bone replacement scaffolds in the future in a cost-effective and reproducible manner. Biological techniques for pre-vascularization of bone substitutes hold a great promise for fabrication of functional large bones in vitro. However, there are some challenges that have limited the progress of these methods. Challenges associated with the fabrication of vascular networks are the inability to form three-dimensional (3D) biomimetic vascular networks in long bones, and the corresponding slow processing time of these networks.
Over the years, many functionalization strategies have been developed and optimized in order to add new functionalities to bio-inert materials and broaden their applicability while overcoming any important material deficiency and preserving their qualities. Approaches for protein immobilization can generally be divided in two groups: (1) surface functionalization, when only the surface is involved in the modification process, commonly through a post-processing step; and (2) bulk functionalization, when the material is homogenously modified, usually before or during the processing. Here, we will focus on these two approaches for introducing biomolecules (proteins and peptides) to inert materials, having the creation of innovative bio-active materials as final goal. Protocols for the visualization and immobilization of proteins are given in Section 1.3.
Surface functionalization strategies aim to introduce new biologically relevant properties to inert materials without undermining its bulk material properties. Surface properties are able to affect the overall performance of materials, control the first phase of material–cells interaction and trigger specific biological responses.
Generally, surface functionalization approaches can be classified into three main categories (physical, chemical and biological). However, this distinction is not always strict and multiple modifications can be employed. The choice of the right route is determined by either the material chemistry or the envisioned application and every choice presents advantages and drawbacks. Here, we will describe biological functionalization methods (physisorption and covalent bonding), overlooking physical (e.g., polishing, grinding, laser treatment) and chemical (e.g., ion bombardment, acid/alkaline treatment, sol–gel, small molecule grafting) approaches.
Physical immobilization is the simplest bio-functionalization method and consists just in dipping the material in a solution containing the target biomolecules. Physisorption (i.e., physical adsorption) relies on electrostatic interactions, van der Waals forces, hydrogen bonds and/or hydrophobic interactions. It can be applied to most types of surfaces and, in general, does not require any surface pre-treatment. Although it is considered a gentle method (i.e., non-destructive), the biomolecule–surface randomly-oriented interaction can lead to conformational changes and subsequent loss of functionality. In addition, due to the presence of weak interactions the binding stability of adsorbed molecules is dramatically affected by environmental conditions, such as pH, ionic strength and biomolecule concentration.76
Recently, a few approaches have been developed to introduce biomolecules by physisorption in a more controlled way. For example, Messersmith's group at Northwestern University introduced a fast, reproducible, versatile method to coat many different organic and inorganic surfaces by exploiting alkaline oxidative polymerization of dopamine.77 Having a thickness ranging from few nanometers (nm) to >100 nm, the polydopamine (pD) coating does not affect the pristine material properties. The coating can be exploited to introduce biomolecules through a single step,78i.e., with the biomolecules entrapped in the pD layer, or two separated steps, with the molecules attached on top of the pD layer.79
Another interesting physisorption-based approach for ceramic materials employed as bone substitutes involves the use of short (10–30 amino acids) modular peptides having two domains: (1) a domain able to bind strongly the surface through electrostatic interactions; and (2) a domain that introduces a biologically relevant function (e.g., cell adhesive peptides or growth factor-derived moieties). The mechanism behind the surface functionalization mimics the natural mechanism by which human bone extracellular matrix proteins bind hydroxyapatite (HA) into the body. Different sequences have indeed been identified as responsible of this interaction in osteonectin and bone sialoprotein,80,81 osteocalcin82 and salivary statherin.83 Moreover, following this approach, phage display technology has been employed for discovering new peptides able to bind ceramic surfaces.84,85 Modular peptides have been reported for functionalizing several calcium phosphate-based materials, such as HA,80,86 HA–titanium,87 β-TCP,88 HA–polymer composites,89 native bone grafts90 and allografts.91
In a series of recent studies, Rohanizadeh and his team investigated protein adsorption onto hydroxyapatite (HA) surface, factors affecting this phenomenon and techniques to modulate it. Using bovine serum albumin (BSA) and cytochrome c as model proteins, the effect of HA crystallinity on its protein adsorptive capacity was investigated by this team.1 They concluded that regardless of the total surface charge of protein, the adsorption of proteins onto HA was directly influenced by HA crystallinity, where higher crystallinity resulted in a lower protein adsorption rate. The crystallinity of HA also affected proteins release profile, in which HA particle with higher crystallinity showed lower protein release kinetics. In addition, the study demonstrated that the surface charge of HA particles influence protein adsorption, where BSA, an acidic protein, was adsorbed at higher rate on a negatively charged surface compared to cytochrome c, a basic protein.
In another study, the team altered the HA surface charge by immobilization of different amino acids during HA precipitation.2 Four amino acids with different isoelectric points were used in this study: neutral (serine, Ser; and asparagines, Asn), acidic (aspartic acid, Asp) and basic (arginine, Arg). In order to evaluate the affinity of amino acid-functionalized HA (AA-HA) to proteins, BSA and lysozyme were used respectively as an acidic and basic model protein. The protein adsorption onto the surface of AA-HA depended on the surface charges of HA particles, whereby BSA demonstrated higher affinity towards positively charged Arg-HA. Alternatively, lysozyme, a positively charged protein showed greater tendency to attach to negatively charged Asp-HA. The AA-HA particles that had higher proteins adsorption demonstrated a lower protein release rate. Thus, the results demonstrated that selective immobilization of amino acid onto HA could provide a strategy to tailor the adsorptive capacity of surface to a specific protein.
The team further investigated the effects of side chain length of amino acids on protein adsorption onto amino acid-treated HA.3 The results showed that immobilization of amino acids with longer side chains decreased the crystallinity and increased the negative value of the surface charge of HA particles. This could be due to the steric hindrance of a ‘bulky’ side chain, disturbing the arrangement of HA crystals, due to lowered crystallinity of AA-HA. Furthermore, manipulation of external factors, namely pH and ionic strength, during proteins adsorption can significantly improve HA binding affinity to proteins. The protein adsorption rate in AA-HA increased with decreasing the pH, while reverse trend obtained in unmodified HA.
HA consisting of carboxyl rich (COO–) groups on its surface was also synthesized by Boccaccini an his team.4 HA particles were precipitated in presence of different concentrations of citric acid (CA). Results showed that CA-HA displayed significantly higher affinity towards lysozyme. Using the optimized parameters obtained from lysozyme adsorption (as a model protein for BMP-2), the results demonstrated that HA particles immobilized with COO– significantly increased the HA loading capacity for bone morphogenic protein-2 (BMP-2) and prolonged BMP-2 release profile. The in vitro results showed that the prepared CA-HA was not toxic towards human osteoblasts and indeed citric ions present on HA surface shifted the surface charge towards negative value, which promoted cell proliferation on HA.
The pros and cons of physisorption are: Pros Applicable to the broad range of natural or synthetic materials No pre-treatment of the target molecules is necessary Mild conditions Usually a very quick and easy approach Cons The coating is not stable for long-term application and molecules detach quickly from the surface with an unpredictable profile Molecules can lose their functionality upon binding
Covalent immobilization of biomolecules is usually chosen when possible over physisorption, as the attachment is controlled and stable, and the kinetics behind biomolecule release can be predicted and calculated. Numerous immobilization protocols have been optimized for different materials and different cross-linker molecules have been used to graft biomolecules to specific, surface functional groups. The most popular routes utilize glutaraldehyde (GA) or carbodiimide (EDC) chemistry to immobilize proteins and peptides as well as oligonucleotides.76 GA can be used to activate the target surface and attach biomolecules irreversibly on the activated surface through an amino group (biomolecule)–aldehyde (surface) reaction. Due to prevalent cytotoxicity effects reported for GA,92 EDC chemistry has become very common due to its ‘zero-length property’. In other words, EDC facilitates the reaction between a carboxylic acid group and an amine group, leaving no residues on the new bond.
In polymers, the functional groups required for GA/EDC coupling reactions93 can be exposed either naturally or by quick surface modification (e.g., oxygen plasma treatment). For ceramic materials, pre-treatments are necessary to introduce such functional groups to the surface calcium phosphate surface. Silanization is the most widely used strategy to introduce functional groups on inorganic surfaces.76 Silanization is extremely versatile in terms of materials to be used with and does not require extreme conditions (i.e., mild pH, room temperature) or expensive equipment. In general, silane molecules are activated by hydrolysis and then condensation between Si–OH groups of the silanol and the OH present on the surface occurs. This activation leads to the exposure of functional groups (from the silanol molecules) on the ceramic surface. In this way amine, carboxylic acid or sulfur groups are introduced on the surface material and are often used for further functionalization reactions through GA or EDC. Though simple, several problems can occur such as layer stability, silane reactivity towards water, possible toxicity of free silanes or leachable products, complete hiding of the bulk material, and potential problem for bioactive ceramics.94 Besides silanization, the formation of a pD coating layer can be a successful strategy for introducing functional amine groups, useful for other functionalization steps.
Overall, covalent immobilization procedures require laborious multi-functionalization work, can produce potentially toxic by-products, and the stability of linkers and bioconjugates during and after the immobilization reactions can be an issue.95
The pros and cons of the covalent bonding method are: Pros Applicable to a broad range of natural or synthetic polymers Stable functionalization over time and usually a predictable release profile Cons Some polymers need to be activated (e.g., grafting reactive groups on the polymer chain) before functionalization By-products from the functionalization step can remain and show cytotoxicity
Simplicity and cost-effectiveness, besides biological efficacy, are two of the factors that remain critically important in the development of new functionalization technologies. The immobilization of biomolecules without chemical linkers—linker-free—addresses common problems related to physisorption and chemical linking. Linker-free immobilization was developed by Kondyurin and is currently used by producers of implantable devices (LfC Z.o.o).37,38,96 This approach is particularly useful for polymeric devices and also has been translated to metallic devices by Chrzanowski.13 In this approach biomolecules are immobilized via radicals created at the interfaces by energetic ion-implantation. Implanted ions disrupted the primary bonds of the polymer and results in the formation of a carbonized interface that contains highly reactive radicals, species with unpaired electrons.97 The radicals are stabilized by delocalization on π-electron clouds of the condensed aromatic structures. The stability increases with larger conjugated areas in the aromatic structure. In addition to the chemical activity that they bring to the surface, the unpaired electrons typically react with oxygen in the environment98 and create an elevated negative potential on the surface. This effect generates a strong double layer in solution and attracts charged protein molecules to the surface. The radicals in the interface also allow the covalent binding of biomolecules without the need for any additional linkers.38 It is believed that the reactive radicals act as anchors to immobilize biomolecules through the substitution of mobile hydrogen atoms bonded to carbon, nitrogen or oxygen atoms in amino acids on the outer surface of the protein molecule. At the same time, the radicals and surface oxygen groups interact with water molecules due to strong hydrogen bonds providing a hydrophilic surface. It is known that an adsorbed water layer prevents direct physical contact of the protein molecule with the surface creating a water shell around the protein and maintaining the natural conformation of the protein molecule despite the covalent bonding with the surface.99 This approach has been proven effective for both polymers and metals and immobilization of different types of proteins, peptides and yeast has been demonstrated.
The pros and cons of the linker-free approach are: Pros Single step method Applicable to the broad range of biomolecules Applicable for all types of polymers Significantly increases the amount of immobilized protein Cons Not applicable for 3D structures Activity of the surface decreases with time Requires a specialized facility
The determination of successful covalent functionalization of growth factors or peptides in a 3D microenvironment (synthetic or naturally derived scaffolds) is not easy. As explained above, growth factors and small peptides can elicit specific cellular responses in encapsulated cells to recapitulate a particular tissue, and promote their proliferation or differentiation. Additionally, growth factors and peptides show short half-lives when injected in vivo, and their functionalization within a 3D matrix extends their functionality and protects them from degradation enzymes in the host environment.100,101 In practice, we have found that the assessment of covalent functionalization of growth factors or peptides in naturally derived scaffolds made of type I collagen, gelatin, alginate, hyaluronic acid, or silk fibroin cannot be done using traditional analytical tools such as 1H nuclear magnetic resonance (1H NMR) or Fourier transform infrared spectroscopy (FTIR). Thus, the functionalization of fluorescent moieties to growth factors and small peptides is a practical approach to determine the extent of covalent functionalization of small molecules in a (hard) scaffold or hydrogel. When choosing the fluorescent tag, the researcher has to take into consideration the molecular mass of the fluorescent tag to attach to the growth factor or peptide. Fluorescent tags such as FITC (5,6-fluorescein isothiocyanate, MW = 389.38) or N-hydroxysuccinimide (NHS)–fluorescein (MW = 473.39) could be used as they are small compared to proteins, peptides, and growth factors, thus minimizing the impact in the biological function of the growth factor or peptide.102
A brief description of the process of determining the functionalization of growth factors and peptides in naturally derived hydrogels or scaffolds is outlined below. We need to clarify that no covalent functionalization is 100% effective and part of the growth factor or peptide will eventually leach out of the hydrogel or scaffolds. Thus, we present a simple and straightforward method to characterize the diffusivity of the growth factor or peptide out of the matrix. We will use FITC as the model fluorescent tag attached to a short peptide such as the QK peptide, mentioned previously, functionalized to a type I collagen hydrogel.
The pros and cons of the fluorescence method are: Pros Applicable for 3D structures Applicable to the broad range of natural or synthetic hydrogels Applicable for all types of polymers Cons The use of large fluorescent tags is not recommended Activity of the functionalized 3D structure may decrease with time Functionalization will be limited by diffusivity of species into hydrogels
© The Royal Society of Chemistry 2015 (2014)