Differential responses of osteoblast lineage cells to nanotopographically-modified, microroughened titanium-aluminum-vanadium alloy surfaces

Abstract

Surface structural modifications at the micrometer and nanometer scales have driven improved success rates of dental and orthopaedic implants by mimicking the hierarchical structure of bone. However, how initial osteoblast-lineage cells populating an implant surface respond to different hierarchical surface topographical cues remains to be elucidated, with bone marrow mesenchymal stem cells (MSCs) or immature osteoblasts as possible initial colonizers. Here we show that in the absence of any exogenous soluble factors, osteoblastic maturation of primary human osteoblasts (HOBs) but not osteoblastic differentiation of MSCs is strongly influenced by nanostructures superimposed onto a microrough Ti6Al4V (TiAlV) alloy. The sensitivity of osteoblasts to both surface microroughness and nanostructures led to a synergistic effect on maturation and local factor production. Osteoblastic differentiation of MSCs was sensitive to TiAlV surface microroughness with respect to production of differentiation markers, but no further enhancement was found when cultured on micro/nanostructured surfaces. Superposition of nanostructures to microroughened surfaces affected final MSC numbers and enhanced production of vascular endothelial growth factor (VEGF) but the magnitude of the response was lower than for HOB cultures. Our results suggest that the differentiation state of osteoblast-lineage cells determines the recognition of surface nanostructures and subsequent cell response, which has implications for clinical evaluation of new implant surface nanomodifications.

Copyright © 2012 Elsevier Ltd. All rights reserved.

Figures

Figure 1

SE images and image analyses of the Ti alloy surfaces used for in vitro cell studies. (A) Microsmooth (sTiAlV) and (B) microrough (rTiAlV) surfaces were relatively smooth at the nanoscale, with some sub-microscale features. After the nanomodification oxidation treatment for 45 minutes, (C) NMsTiAlV and (D) NMrTiAlV surfaces possessed high and homogeneous surface area coverage of nanostructures. Image analyses of the (E) NMsTiAlV and (F) NMrTiAlV surfaces revealed that the nanostructure diameter (when viewed from above by SEM analyses) ranged between 20 and 180 nm, with average values of 73 nm and 61 nm, respectively.

Figure 2

(A) Optical and (B-D) SE images of the surface nanostructural modification applied to clinically relevant Ti alloy spine implants. (A, B) Low magnification images show the complex design of the device. (C, D) High magnification images of the unmodified implant reveal that the surface was relatively smooth at the micro- and nanoscales. Conversely, (E, F) high magnification images of the nanomodified implant surface display homogeneous coverage of nanostructures throughout exposed and non-line-of-sight areas.

Figure 3

(A) Surface roughness average (Sa) of sTiAlV and NMsTiAlV surfaces measured by laser confocal microscopy (LCM, black bars) and atomic force microscopy (AFM, orange bars). AFM scans were not possible on microrough specimens, rTiAlV and NMrTiAlV, due to z-height tool limitations. * refers to a statistically-significant p value below 0.05 vs. sTiAlV; # refers to a statistically-significant p value below 0.05 vs. NMsTiAlV. (B) TEM evaluation of a NMsTiAlV surface cross-section prepared by focused ion beam (FIB) milling. The cross-sectional TEM image of the NMsTiAlV specimen reveals a conformal oxide layer that possesses pores and has a thickness of up to 1600 nm. (C) X-ray diffraction (XRD) patterns obtained from sTiAlV and NMsTiAlV specimens. The original sTiAlV specimen only exhibited peaks for α- and β-titanium, while the nanomodified NMsTiAlV specimen exhibited peaks for α-titanium, rutile and anatase TiO2. (D) Sessile-drop water contact angles on the surfaces of sTiAlV, NMsTiAlV, rTiAlV and NMrTiAlV specimens.

Figure 4

Elemental compositions of the sTiAlV, NMsTiAlV, rTiAlV and NMrTiAlV specimens measured by XPS. All samples were mainly composed of Ti, Al, and O, with C also highly present on the surface. N was also present at low levels on the sTiAlV surfaces, while NMsTiAlV and rTiAlV surfaces only had traces (T) and on the NMrTiAlV surfaces it was not detectable (ND). * refers to a statistically-significant p value below 0.05 vs. unmodified control; # refers to a statistically-significant p value below 0.05 vs. micro-smooth control.

Figure 5

Effects of micro- and nanoscale surface modifications on human primary osteoblasts (A-E) and human MSCs (F-J). Osteoblasts and MSCs were plated on sTiAlV, NMsTiAlV, rTiAlV and NMrTiAlV surfaces and grown to confluence. The nanomodification involves surface oxidation in flowing synthetic air for 45 minutes at 740 °C. At confluence, (A, F) cell number, (B, G) alkaline phosphatase specific activity, (C, H) OCN, (D, I) OPG, and (E, J) VEGF levels were measured. Data represented are the mean ± SE of six independent samples. * refers to a statistically-significant p value below 0.05 vs. sTiAlV; # refers to a statistically-significant p value below 0.05 vs. NMsTiAlV; $ refers to a statistically-significant p value below 0.05 vs. rTiAlV.