Implant‐bone‐interface: Reviewing the impact of titanium surface modifications on osteogenic processes in vitro and in vivo

Abstract Titanium is commonly and successfully used in dental and orthopedic implants. However, patients still have to face the risk of implant failure due to various reasons, such as implant loosening or infection. The risk of implant loosening can be countered by optimizing the osteointegration capacity of implant materials. Implant surface modifications for structuring, roughening and biological activation in favor for osteogenic differentiation have been vastly studied. A key factor for a successful stable long‐term integration is the initial cellular response to the implant material. Hence, cell–material interactions, which are dependent on the surface parameters, need to be considered in the implant design. Therefore, this review starts with an introduction to the basics of cell–material interactions as well as common surface modification techniques. Afterwards, recent research on the impact of osteogenic processes in vitro and vivo provoked by various surface modifications is reviewed and discussed, in order to give an update on currently applied and developing implant modification techniques for enhancing osteointegration.


| FOREWORD AND REVIEW SCOPE
Titanium (Ti)-commercially pure titanium and its alloys, usually grade 5 Ti6Al4V-are commonly used for dental and orthopedic implant applications due to their excellent resistance to corrosion, biocompatibility properties, mechanical strength and elastic modulus, which is closer to bone compared to other metals. [1][2][3] As bones have a major functional importance including structural composition of the skeleton, load bearing, and motion support of the human body, a skeletal impairment or disease greatly affects the quality of life of a patient. 4 Therefore, it is of great importance to maintain bone function throughout life and in the case of terminal disease stage or severe injury, bone replacement by implants is the primary choice for treatment. Dental implants composed of titanium are widely used and show excellent long-term results. In orthopedics, titanium is used for uncemented implants, which are in direct contact to the bone tissue. Cementless fixation requires bone tissue to attach to the implant surface to secure the integration of the implant. For that reason, cementless implants are primarily used for bones of good tissue quality, such as in healthy young patients, and are not suitable for bones with lower mineral density, such as in aged and osteoporotic patients.
However, developing an implant that allows cementless fixation also in compromised bone would offer clear benefits, such as protection of native bone tissue and avoidance of incorporation of body foreign substances (bone cement). In addition, bone implants, despite the fact that they are well established, still face the problem of implant failure due to two leading reasons-implant loosening owing to insufficient bone integration and/or the production of fibrous tissue or infection.
Therefore, there is a continuous scientific effort toward the development of innovative implant materials (surfaces) that can (i) stimulate healing and enhance osteointegration, independently of the bone quality, (ii) act inhibitory for infections, and (iii) prolong the longevity of an implant. 5 Osteointegration arises from the physical and chemical interaction between the implant surface and the bone tissue. [6][7][8] Evaluating the biological responses triggered by surface modifications can be used to guide the cellular response at the bone implant interface for achieving implant surfaces with augmented osteointegration. [6][7][8] Thus, nature-inspired implant surfaces that are very similar to the native bone tissue topography at the macro-and nano-scale as well as that can be further functionalized to simulate the bone biochemical milieu are of great interest to the field. [9][10][11] In this review, we start with a foreword on titanium implants and the review scope, followed by a synopsis on the discrete interactions between cells and biomaterials and an overview of surface modifications enhancing osteogenic differentiation. Next, literature on recent research regarding implant surface modifications and their impact on osteogenic processes in vitro and in vivo is discussed in detail.
Surface modifications for improved implant performance is a vastly studied area. We were particularly interested to obtain the latest information of research, focusing on biological assessment of implant surface modification techniques with the overall aim to enhance osteointegration. Literature search was conducted via the National Center for Biotechnology Information (NCBI) database.
For the informational chapters 1-3 and Table 1, articles (approximately 50, many of them review articles) dealing with general information on titanium implants and types of surface modifications and cell to material interactions and integrin signaling were selected.

| DISCRETE INTERACTIONS BETWEEN CELLS AND MATERIAL SURFACES
The surface of an implant is in direct contact with the host tissue, for example, bone tissue. Therefore, the surface properties are a main determining factor for the subsequent complex cell behavior at the bone-implant interface in vivo as well as for the cell response in vitro ( Figure 1). 12 Different parameters, for instance surface topography, chemistry, charge and culture conditions (in vitro) or physiological environment (in vivo), impact the discrete interactions between cells and the biomaterial.
Interestingly, the same basic substrate can provoke different cell responses when exhibiting different nanostructures, leading, for example, to modulations in cell adhesion, motility and signaling pathways. 13 It is important to understand the dynamic interactions between biomaterials and adhering cells, as this affects cell proliferation, differentiation, migration and consequently, the integration of the biomaterial into the host tissue. 14 Bone tissue has a mineralized macroporous structure with nano-scale components that determine its strength. Inorganic hydroxyapatite (HAp) constitutes the major part of the mineralized component. The organic extracellular matrix (ECM) predominantly consists of collagen type I and the bone cells-osteogenic progenitor cells, osteoblasts, osteocytes, and osteoclasts. Naturally, the hierarchical structure of the bone (from nanolevel, e.g., collagen molecules, minerals, to microlevel, e.g., the osteon) guides the bone cells in their tissue specific behavior. 13,15 Thus, titanium implant surfaces should ideally have characteristics similar to the native bone topography in order to facilitate the desired cell responses which in turn enable osteointegration. 10,13,15 In this manner, it may be possible that even aged and osteoporotic cells could be stimulated and have an enhanced osteogenic differentiation potential.
After an implant or biomaterial is exposed to biofluids, the adsorption of water, serum molecules, proteins and cells ( Figure 2, step 1) is determined by the physicochemical state of the surface, mainly its chemistry and charge. 16 Following their adsorption to the surface, proteins adapt to a specific conformation, which depends upon the surface properties. The initial cell linkage to the material is governed by composition, density and conformation of the adsorbed proteins. Subsequently, cells close to the surface start filopodial sensing via integrins Interactions between integrins and ECM proteins occurs via recognition of amino acid sequence domains (e.g., RGD (Arg Gly Asp)) that is found in fibronectin, osteoprotegerin and bone sialoprotein, or GFOGER (glycine-phenylalanine-hydroxyproline-glycine-glutamate-arginine) for collagen type I. The impact of surface characteristics on cell morphology and differentiation is mediated via integrins, as surface properties interfere with integrins and influence interactions between integrins and their ligands. The integrin signaling cross-talks with signaling pathways of growth factors, guiding the behavioral pattern of MSCs and bone cells. 12,14,17,18 For example, fibronectin, an adhesive protein considered as pro-osteogenic, interacts with cells via integrin focal adhesion points. Thereby, it is controlling cell activity and promotes osteogenic differentiation of MSCs. Osteoblasts were shown to attach to 2D surfaces in vitro via integrins, whereby the focal adhesion site formation relied on the integrin activation state. 12,14,[19][20][21] Biomaterials devoid of surface roughness in the micro-and nanoscale range have shown to hinder cell osteogenic differentiation.
Rougher surfaces (mean average roughness R a > 0.5 μm) were correlated to increased bone to implant contact (BIC) and described to be preferred by bone cells compared to smooth surfaces. Figure 3 graphically depicts major differences between smooth and roughened surfaces. On smooth surfaces, less pro-osteogenic but rather fibrotic cells attach and proliferate, which can result in fibrous tissue formation and implant loosening in vivo. 22 However, such surfaces have been shown to achieve sufficient osteointegration in dentistry. 23 In general, pro-osteogenic cells are more favorable to attach, proliferate and differentiate on rough nano-patterned surfaces, thereby reducing the risk of undesirable fibrosis.
Cells exposed to roughened biomaterials exhibit more focal contact points, cell adhesion and increased proliferation. These differences in cell F I G U R E 1 Visualization of the interrelation of biomaterial properties and the biological (osteogenic) response. The interrelationship of surface characteristics of a biomaterial and the cell response is a complex mechanism dependent on numerous factors that are accountable for successful osteointegration. (1) Various surface properties, ranging from topographical to chemical features, affect (2) the biological and cellular response to biomaterials (e.g., ligand density, protein adsorption, cell adhesion, cell signaling) and finally (3) determine the biological outcome of an implant (surface) in terms of osteogenic differentiation and osteointegration F I G U R E 2 Cartoon depicting the cell receptor recognition of biomaterials. The initial response of cells to biomaterials occurs via surface receptors, such as integrins. (1) First, water, other solubles of the biofluid (not depicted), and proteins (depicted in green) attach to the implant surface and (2) adopt a certain conformation depending on the surface properties. (3) Cells are able to sense and attach to the proteins, and form focal adhesions on the surface. Source: Adapted from Kim et al. 16 response also rely on the integrin reaction to the surface topography, which is determined by the structure (roughness, size, morphology) and the mechanical properties (stiffness, deformity, rigidity, elasticity).
Integrins, plasma membrane receptors, can sense the biomechanical niche and initiate biochemical signaling cascades regulating cell behavior. [24][25][26] The exact degree of nano-scale influence on the cell response, however, depends on the cell type. 13,22 The biomaterial nano-scale features can enrich protein adsorption and modulate the arrangement of the cytoskeleton ( Figure 2, step 3) leading to an improved osteogenic stimulation of cells. 16,27 For example, osteoblasts exhibit an enhanced collagen production and calcification processes when cultured on rough surfaces. 28,29 It has also been shown that the combination of multiple length-scale features of the implant topography correlates with increased osteoblast differentiation. 30 Biomaterials incorporated in the bone tissue form the so-called bone-implant interface at the implant site.  15,31 Taken together, biomaterials and their surface properties influence cell behavior. The processes of cell-material interaction along with bone healing around an implant displays complex interactions between the material, different cell types and signaling pathways. 14,18 It is essential to be conscious about these processes when designing an implant surface.
Understanding the discrete cell responses can help modulating the surface features in order to steer the cell toward the desirable biological response.

| IMPLANT SURFACE MODIFICATION TECHNIQUES
This chapter provides a short synopsis on surface modification techniques, for detailed reviews on methodologies, please refer to other reviews, for example, Refs. 32-35 . Combined effects of the surface chemistry, topography and the resulting surface energy play essential roles, especially during the early phases of the biological response, and influence the subsequent osteointegration of the implant. 36,37 The surface properties of a metallic implant material are essentially characterized by its inherent chemical composition and the surface's physical and or biochemical modification(s). 38 As mentioned above, the topography describes the biomechanical and structural characteristics of the surface. In general, the roughness of a surface on the micro-scale has been classified into smooth (average roughness R a < 0.5 μm), machined/minimal (R a = 0.5-1 μm), moderate (R a = 1-2 μm) and rough (R a > 2 μm). 39,40 Overall, surface modifications increasing hydrophilicity and roughness exert positive effects on osteogenic differentiation of cells and enhance osteointegration of implants. 41,42 Hydrophilic and roughened surfaces support cell attachment while roughness at the macroand micrometer scale improve mechanical anchorage of the implant in the bone tissue. 43,44 Roughening produces an enlarged surface area leading to a broader territory for cell adhesion, bone-implant-contact and thus better biomechanical integrity after the bone-implantinterface is filled with new bone matrix. 12 Table 1. For the generation of the basic implant surface roughness, physical (e.g., grinding or laser texturing) and chemical (e.g., acid etching) modification techniques are applied. Figure 5 exemplarily shows titanium surfaces modified with different techniques. Chemical modification techniques, such as acid etching, are more likely to alter the chemical surface composition than physical methods. For example, acid etching of titanium with HCl and H 2 SO 4 was shown to lead to hydrogen adsorption and formation of stable titanium hydride on the surface. 46,47 Interestingly, titanium surfaces roughened with physical methods often demonstrate the formation of the so-called TiO 2 passivation layer. [48][49][50] In addition to appropriate macro-and micro-features of an implant, nano-patterning has been reckoned to play a crucial role for the biological response. 9,27,51 Despite that the sand blasting and acid etching (SLA)-treated implants are commonly used in clinics, there are indications that laser texturing provides a more suitable nano-topography compared to the rather sharp-edged morphology after SLA treatment. Comparing a scanning electron microscope (SEM) image of a laser textured surface to a SEM image of bone tissue surface, shows their great resemblance ( Figure 5). Laser texturing is one of the latest and promising technologies for metal implant surface structuring that allows to design a desired, controlled and reproducible surface geometry at different length-scales. 52,53,54 During the manufacturing process, no additional chemicals, which might be harmful, are incorporated into the surface layer. Moreover, in a stochastic manner, laser texturing automatically creates metal nanodroplets on the implant surface, thereby generating a nano-roughened topography with a foamy, roundly shaped nano-features. 53,54,51 To further enhance the bioactivity of a titanium implant surface, additional ion and molecular functionalization (Table 1) can be carried out with the goals of (1) eliminating proteins which would lead to attachment of unspecific cells, resulting in fibrotic tissue formation or bacterial adhesion; (2) boosting the adherence of desired cell types, that is, osteogenic progenitor cells and osteoblasts; (3) guiding responses of immune cells modulating inflammation during the process of bone healing. 12 The functionalization is based on the incorporation or binding of inorganic ions or molecules (e.g., magnesium (Mg), calcium (Ca) and strontium (Sr)) and organic molecules (e.g., peptides, proteins and drugs). 55,11,56 HAp has been investigated as a coating substance for a long time and is still frequently chosen. Its deposition can promote better BIC and bone formation, and is already in clinical use. [57][58][59][60] The deposition of coating molecules is performed with various methods including plasma spraying, electrochemical/micro-arc/ anodic oxidation, immersion, acid etching and laser ablation (Table 1).
Either, molecules are formed automatically but uncontrolled on the surface (indirect coating, e.g., anodic oxidation or immersion); or the molecules are directly deposited on the surface in a controlled density (e.g., plasma spraying, laser ablation).
There has been an enormous advancement in new methods for texturing and biofunctionalizing implants. However, to estimate the translational power of novel surface modifications, thoughtful assessment of the complex cellular and tissue responses is required. Therefore, the following chapters will focus on the output of surface modification techniques on osteogenic processes in vitro and in vivo.  The appropriate selection and combination of surface modification techniques affects its cellular biocompatibility and influence.
For example, an apatite coated titanium dioxide (TiO 2 , titania) surface produced by blasting, performed better in terms of cellular adhesion and proliferation than an apatite coated TiO 2 surface fabricated by flame spraying. 71 Moreover, the blasting method achieved increased cellular alkaline phosphatase activity and expression of essential cell-cell and cell-matrix adhesion proteins (e.g., fibronectin and E-cadherin), indicating enhanced osteogenic ability. 71 Mariscal-Munoz et al. found that the micro-to-nano surface roughness generated by laser ablation, augmented osteoblast differentiation and matrix mineralization, alongside an increased expression of bone specific genes. 48 Table 3 gives an overview on the included research articles, the utilized surface modifications and achieved outcome. In the subsequent sections, the included studies are discussed in more detail.

| Effect of micro-nano-scale surface roughening of titanium implants
At the micrometer scale, moderately rough sandblasted and acidetched titanium surfaces inserted into the tibia of rabbits showed considerably higher RTVs at a later stage of the remodeling process.
However, no difference between the modified and machined surface was observed regarding the BIC. 74,49 The combination of surface roughness at different length scales    ness. This indicates that the nano-roughness of a surface is a highly significant factor for implant performance, which may be more significant than mimicking the bone micro-scale. In particular for laser texturing, the created nano-scale features appear in a foamy, roundly shaped morphology and have greater similarity to bone tissue, which is different to the SLA treated surfaces resembling rather sharp-edged morphology ( Figure 5). 51 In accordance, Souza et al. concluded that proper nanotexturing leads to a faster osteointegration process and furthermore, can reduce the risk of bacterial contamination. 27

| Effect of additional functionalization and bioactive coating
Further approaches that have been successfully applied to enhance the bioactive properties of a titanium implant surface are the functionalization or coating with specific molecules.
One method of functionalization is photo functionalization via ultraviolet light immediately prior to implantation. For example, this approach provoked an increased amount of bone mineralization and osteoblast proliferation at the early stages of healing compared to the standard SLA. 95 Interestingly, the UV treatment in addition to increasing surface roughness, also led to the formation of superhydrophilic surface characteristics that promoted beneficial physicochemical changes and increased bone healing. Likewise, UV-treated microfiber implants inserted into the rat femur promoted better implant anchorage and bone formation after 4 weeks compared to the non-UVtreated control group. 96 Implantation of HAp and bioactive glass coated implants into human jawbones showed better biocompatibility with the surrounding tissue when compared to machined implants. These findings indicated that an improved surface hydrophilicity positively impacts the surface energy, thereby promoting the adhesion and proliferation of osteoblasts and relevant growth factors required for bone formation. 97 Certain metallic ions such as calcium, magnesium, sodium and strontium have also demonstrated synergistic effects on osteogenesis.
For example, the incorporation of calcium ions (Ca 2+ ) into the titanium surface enabled the conversion of passive oxide into a bioactive oxide (CaTiO 3 ), which is more favorable for biological interaction. Wang et al.
reported excellent biocompatibility and osteointegration effects of nano-bioactive CaTiO 3 coated screws produced via treatment with NaOH and CaCl 2 . 98 The results after 12 weeks of implantation were comparable to HAp-coated and superior to uncoated implants. 98 Ca 2+ deposition in a nano-porous Ti alloy equally resulted in improved osteoconductivity and overall bone formation at week four and eight after implantation in a rat femur compared to Na + incorporation. The divalent Ca 2+ incorporates deeper into the layer of the nano-porous structure, enabling a consistent and sustained release over time, leading to a superior bioactivity and increased trabecular bone formation. 99 Like calcium, magnesium is also vital in the process of bone regeneration, it promotes osteogenic differentiation, as well as angiogenesis.
The integration or Mg 2+ into Ti surfaces has led to an increased surface bioactivity and osteointegration. Interestingly, Mg released from mesoporous titanium films significantly supported bone formation after 7 days of implantation into the tibia and femora of osteoporotic rats. In addition, a positive osteogenic effect of Mg 2+ doped surfaces compared to uncoated could be demonstrated by a 3-fold higher expression of BMP6, a key growth factor involved in bone formation. 100  Dopamine coating facilitated bone formation by inhibiting the expression of genes associated with osteoclast differentiation. 105 Similarly, titanium implant surface coating with antimicrobial agents, such as the bactericidal cationic peptide GL13K, not only inhibited microbial activity but also promoted osteointegration after 6 weeks of implantation in a rabbit femur model. 106 The addition of silicon-substituted nano-HAp to the surface of a selective laser structured titanium implant, inserted into the rabbit femur, promoted more organized bone formation, especially at the later stages of bone healing compared to implants without additional chemical treatment. 107  In vitro priming of implant surfaces with living cells that are present in bone tissue can be the next step of surface functionalization further mimicking the native bone environment. Due to more elaborate ethical and preparatory processes prior to implantation, this approach will require a lot more investigation before application in clinical daily life.
To summarize this chapter, the key factors in all reviewed experiments were the optimization of coating techniques and the combinations with structuring methods to ensure the optimal contribution of various bioactive agents to osteointegration improvement. The combination of chemical treatment with other surface topography modification techniques has led to the development of novel titanium-based implant surfaces with improved micro-to-nano hybrid topographies.
Their enhanced bioactive properties facilitate earlier bone regeneration and could lead to improved osteointegration at the bone-implant interface in both healthy and compromised bone.

| CONCLUSION
The current research on the osteointegration capacity of titanium implants reports promising enhancement strategies via increasing porosity, hydrophilicity and nano-structuring of the surface, frequently using a combination of roughening techniques and bioactive substance coatings.
In general, hydrophilic surfaces show improved osteoinduction and decreased inflammatory response, and when combined with nano-patterning, augmented osteointegration can be achieved. HAp, the primary inorganic component of bone tissues, has been investigated as a coating material for a long time and is still frequently chosen. Its deposition can promote better BIC, as well as bone tissue formation and is already in clinical use for cementless fixated implants.
However, new coating compositions, such as calcium titanate or bioactive glass, arise as promising candidates for implant surface modification. In sum, creating a rough, nano-textured surface and sequential application of various techniques to further biofunctionalize the implant is desired. Next to coating with bioactive molecules, another interesting approach is surface loading with cells. This type of functionalization has not been vastly studied, as its clinical translation is more challenging due to the cell preparatory requirements and regulations. Still, this approach can potentially gain more attention in the future, alongside the progression of cell-based therapy and personalized medicine in many other clinical areas.
Nano-structuring of titanium surfaces (e.g., via laser texturing), is a very attractive and expanding area, which should be further explored in great detail, as it holds the potential to induce high osteointegration and biomechanical anchorage without additional coatings. Micro-and nano-porous titanium substrates are able to achieve the same repair capacity as porous HAp constructs, with titanium having more suitable biomechanical features, suggesting that the surface nanostructure is of great importance for proper bone formation. Hence, in the future, even more attention will be paid not only to the micro-scale modifications, but also to the nano-patterning of novel implants for augmented osteointegration.
In the process of developing next-generation-implants, it will be of great importance on behalf of the biological assessment, as well as cross-study comparability, to improve certain evaluation parameters. Kov ařík, data sharing is not applicable as this is a literature review article and no datasets were generated or analysed. Copyrights for images in Figure 5f (Science Photo Library/Science Source/Nano Creative) were purchased from Science Photolibrary, Munich, Germany. Open Access funding enabled and organized by Projekt DEAL.

CONFLICT OF INTEREST
The authors have no conflicts of interest to declare.