Human Mesenchymal Stem Cell Morphology and Migration on Microtextured Titanium

Abstract

The implant used in spinal fusion procedures is an essential component to achieving successful arthrodesis. At the cellular level, the implant impacts healing and fusion through a series of steps: first, mesenchymal stem cells (MSCs) need to adhere and proliferate to cover the implant; second, the MSCs must differentiate into osteoblasts; third, the osteoid matrix produced by the osteoblasts needs to generate new bone tissue, thoroughly integrating the implant with the vertebrate above and below. Previous research has demonstrated that microtextured titanium is advantageous over smooth titanium and PEEK implants for both promoting osteogenic differentiation and integrating with host bone tissue; however, no investigation to date has examined the early morphology and migration of MSCs on these surfaces. This study details cell spreading and morphology changes over 24 h, rate and directionality of migration 6-18 h post-seeding, differentiation markers at 10 days, and the long-term morphology of MSCs at 7 days, on microtextured, acid-etched titanium (endoskeleton), smooth titanium, and smooth PEEK surfaces. The results demonstrate that in all metrics, the two titanium surfaces outperformed the PEEK surface. Furthermore, the rough acid-etched titanium surface presented the most favorable overall results, demonstrating the random migration needed to efficiently cover a surface in addition to morphologies consistent with osteoblasts and preosteoblasts.

Keywords: PEEK; cell–material interactions; regenerative medicine; spinal implant; titanium (alloys).

Figures

Figure 1

Surface morphology of PEEK and titanium samples. (A) PEEK, (B) smooth titanium, and (C) rough, acid-etched endoskeleton surface.

Figure 2

Morphological changes of mesenchymal stem cells analyzed at 2, 6, and 24 h post-seeding. (A) Area, (B) circularity, and (C) aspect ratio measurements were taken. The results indicate that stem cells on the acid-etched endoskeleton surface spread the most over 24 h. The circularity of the three surfaces began dissimilar, but converged at 24 h. The aspect ratio of stem cells initially began close to 1, but over 24 h, the smooth surfaces, Ti and PEEK, increased significantly higher than the rough, acid-etched endoskeleton surface. Taken together, the aspect ratio and circularity indicate that stem cells on smooth surfaces move toward a spindle or fibroblastic morphology, whereas those on the rough, acid-etched endoskeleton surface moved toward a stellate or star-like morphology. Within a single time point, * indicates significance, p < 0.05 between acid-etched Ti and PEEK, † indicates significance between acid-etched Ti and Ti, and § indicates significance between PEEK and Ti. Color-coded bars demonstrate significance between time points for a single surface.

Figure 3

Representative morphologies of MSCs. (A) PEEK, (B) smooth titanium, and (C) rough, acid-etched endoskeleton surface, at 24 h. Immunofluorescence was carried out to examine the focal adhesion protein vinculin (green), the actin cytoskeleton (red), and the cell nuclei (blue). Additionally, a gray scale depiction of the surface was obtained with reflected DIC. The results demonstrated the trends observed in Figure 2 with cells on the smooth surfaces moving toward an elongated spindle-shaped morphology, whereas the cells on the rough surface demonstrated a range of morphologies from spindle-shaped cells to cuboidal and stellate-shaped cells. In particular, the cuboidal and stellate cells in C. are representative of morphologies expected of osteoblastic differentiation. Scale bar indicating 50 μm applies to (A–C).

Figure 4

Stem cell migration on each surface was assessed from 6 to 18 h post-seeding. The results demonstrate random migration on the PEEK and acid-etched endoskeleton surfaces indicated by the rose plots in (A), a histogram of the angle of migration for each cell monitored in (B), and the graph of directionality in (C), which demonstrates significance between PEEK and acid-etched endoskeleton surfaces when compared to the smooth titanium surface. Furthermore, the non-random migration on smooth titanium followed the grooves created by milling the surface, and this non-random migration resulted in an expected velocity increase, which was significantly higher than both the PEEK and acid-etched endoskeleton surfaces. Between the two surfaces demonstrating random migration, the MSCs on the acid-etched endoskeleton surface demonstrated a significantly higher velocity than those on PEEK. Significance, p < 0.05, is demonstrated by bars between groups in (B,C).

Figure 5

Early differentiation marker alkaline phosphatase (ALP) and osterix (OSX), a transcription factor significant for osteoblast differentiation, were normalized to dsDNA. (A) Early differentiation marker, ALP, is increased on the smooth Ti surface, while the mid-differentiation marker OSX increased on the acid-etched endoskeleton surface. The PEEK surface fell short for both ALP and OSX. This suggests that hMSC differentiation is moving toward bone formation for the Ti surfaces. (B) Seeding densities for all samples were equal. The dsDNA value for the acid-etched endoskeleton surface is the highest amongst the surfaces, which suggests that there is improved cell attachment and/or proliferation for the acid-etched endoskeleton surface compared to the smooth Ti and PEEK substrates.

Figure 6

Nuclear morphology was examined to assess the general cell morphology after 7 days when the populations were confluent and cell borders were difficult to identify. The nuclear morphology on PEEK and Ti surfaces were very similar in regards to axial ratio (A), whereas the nuclei on the rough acid-etched endoskeleton surface had a significantly lower axial ratio than either the PEEK or smooth Ti surface indicating more circular nuclei on the acid-etched endoskeleton surface. The nuclear area (B) followed a similar trend to axial ratio with the smooth surfaces demonstrating significantly more nuclear area than the rough acid-etched endoskeleton surface. Finally, the orientation of nuclei (C) was assessed establishing 0° as the average orientation direction. The inset provides a plot of the cumulative distribution and clearly demonstrates that PEEK and smooth Ti surfaces were different than the rough, acid-etched endoskeleton surface. Nuclei on PEEK and smooth Ti were grouped very close to 0° indicating that most cells presented an elongated nucleus in the same direction; however, on the acid-etched endoskeleton surface, the nuclei were randomly oriented with only one range, 70–90°, demonstrating a slight increase. The black dotted line in the inset of (C) provides the expected cumulative distribution for random orientation; p values were calculated for each of the three samples with a χ2 test and yielded p values of 10−10, 10−19, and 1.0 for PEEK, Ti, and acid-etched endoskeleton, respectively. Taken together, these results indicated that the aligned spindle morphology observed early on the PEEK and smooth Ti surfaces persists when the stem cells are confluent, and likewise, the random cuboidal/stellate morphology on the acid-etched endoskeleton surface also persists to the confluent cell layer observed after 7 days.

Figure 7

Representative images of confluent cells stained for actin (red) and the cell nuclei (blue). (A) PEEK, (B) smooth titanium, and (C) rough, acid-etched endoskeleton surface, after 7-day culture. The cells on PEEK and smooth titanium demonstrate an elongated morphology in a uniform direction, whereas cells on the acid-etched endoskeleton surface demonstrate a branched random morphology. Scale bar indicating 200 μm applies to (A–C).