Contact Guidance of Connective Tissue Fibroblasts on Submicrometer Anisotropic Topographical Cues Is Dependent on Tissue of Origin, β1 Integrins, and Tensin-1 Recruitment

Abstract

The substratum topography of both natural and synthetic materials is a prominent regulator of cell behaviors including adhesion, migration, matrix fibrillogenesis, and cell phenotype. Connective tissue fibroblasts are known to respond to repeating groove topographical modifications by aligning and exhibiting directed migration, a phenomenon termed contact guidance. Although both reside in collagen rich connective tissues, dermal and gingival fibroblasts are known to exhibit differences in phenotype during wound healing, with gingival tissue showing a fetal-like scarless response. Differences in adhesion formation and maturation are known to underlie both a scarring phenotype and cell response to topographical features. Utilizing repeating groove substrates with periodicities of 600, 900, and 1200 nm (depth, 100 nm), we investigated the roles of integrins αvβ3 and β1 associated adhesions on contact guidance of human gingival (HGFs) and dermal fibroblasts (HDFs). HGFs showed a higher degree of orientation with the groove long axis than HDFs, with alignment of both vinculin and tensin-1 evident on 600 and 900 nm periodicities in both cell types. Orientation with grooves of any periodicity in HGFs and HDFs did not alter the adhesion number or area compared to smooth control surfaces. Growth of both cell types on all periodicities reduced fibronectin fibrillogenesis compared to control surfaces. Independent inhibition of integrin αvβ3 and β1 in both cell types induced changes in spreading up to 6 h and reduced alignment with the groove long axis. At 24 h post-seeding with blocking antibodies, HGFs recovered orientation, but in HDFs, blocking of β1, but not αvβ3 integrins, inhibited alignment. Blocking of β1 and αvβ3 in HDFs, but not HGFs, inhibited tensin-1-associated fibrillar adhesion formation. Furthermore, inhibition of β1 integrins in HDFs, but not HGFs, resulted in recruitment of tensin-1 to αvβ3 focal adhesions, preventing HDFs from aligning with the groove long axis. Our work demonstrates that tensin-1 localization with specific integrins in adhesion sites is an important determinant of contact guidance. This work emphasizes further the need for tissue-specific biomaterials, when integration into host tissues is required.

Keywords: directed migration; fibrillar adhesion; fibroblasts; submicrometer topography; wound healing.

Figure 1

Atomic force microscopy (dynamic mode) images of the ridge/peak profiles of repeating groove topographical cues fabricated in fused silica.

Figure 2

HDFs and HGFs orient with respect to the long axis of the groove topographies. HDF and HGF cell populations were cultured for 24 h upon smooth and groove surfaces with periodicities of 600, 900, and 1200 nm. Vinculin (green), F-actin (red), and nuclei (blue) are identified through immunofluorescent staining. The direction of the underlying nanogroove long axis is indicated with white arrows.

Figure 3

Nanogrooves affect the adhesion site arrangement of HGFs and HDFs. Immunofluorescent staining of vinculin (green), tensin-1 (red), and nuclei (blue) in HDF and HGF populations following a 24 h timepoint upon smooth and groove surfaces. The direction of the underlying nanogroove long axis is indicated with white arrows.

Figure 4

Submicrometer grooves orient HGF and HDF adhesion proteins. Directionality of vinculin and tensin-1 in HDF and HGF fibroblasts at 24 h post-seeding. Mean values established through Fourier transform analysis are displayed in 10° increments ranging from 0 to 360°. The greatest standard deviation at any point of the distribution is displayed. Data were analyzed using the Kolmogorov–Smirnov test of the smooth controls and each experimental surface for both HDF and HGF populations (N = 3; *P < 0.05).

Figure 5

Submicrometer grooves orient HGF and HDF cytoskeletal proteins. Directionality of F-actin and the cellular long axis in HDF and HGF fibroblasts at 24 h post-seeding. Mean values established through Fourier transform analysis are displayed in 10° increments ranging from 0 to 360°. The greatest standard deviation at any point of the distribution is displayed. Data were analyzed using the Kolmogorov–Smirnov test of the smooth controls and each experimental surface for both HDF and HGF populations (N = 3; *P < 0.05).

Figure 6

HDF and HGF fibronectin fibril assembly is disrupted by anisotropic repeating groove topographies. Immunofluorescent staining of fibronectin (green), F-actin (red), and nuclei (blue) within HDF and HGF populations following a 24 h timepoint upon the smooth and groove surfaces.

Figure 7

Association of α5 and β1 integrin dimers is disrupted on anisotropic repeating groove topographies. Immunofluorescent staining of integrin β1 (green), integrin α5 (red), and nuclei (blue) within HDF and HGF populations following a 24 h timepoint upon the smooth and groove surfaces.

Figure 8

Inhibition of integrin αvβ3 and integrin β1 has a differential effect upon HGF and HDF spreading. Immunofluorescent staining of (A) HDF and (B) HGF vinculin (green), F-actin (red), and nuclei (blue) following 1, 3, and 6 h upon a 900 nm periodicity groove surface. The direction of the underlying nanogroove long axis is indicated by white arrows. Scale bar: 20 μm.

Figure 9

HDF and HGF cell area and circularity are affected by αvβ3 and integrin β1 blocking antibodies The average HDF (A) and HGF (C) area and HDF (B) and HGF (D) circularity in the presence of integrin αvβ3 and β1 blocking antibodies at 1, 3, and 6 h were quantified. Data is expressed as mean ± standard deviation. Data was analyzed using two-way ANOVA, followed by Bonferroni post-tests (N = 3; *P < 0.05).

Figure 10

Independent inhibition of integrin αvβ3 and integrin β1 demonstrates contrast in alignment between HDF and HGF. (A) Immunofluorescent staining of F-actin in HDFs and HGFs at 24 h on a 900 nm periodicity groove surface. (B) Directionality of F-actin expression of integrin αvβ3 and integrin β1 inhibited HDF and (C) HGF following a 24 h timepoint. Mean values are displayed in 10° increments ranging from 0 to 360°. The greatest standard deviation at any point of the distribution is displayed. Data was analyzed using the Komolgorov–Smirnov test between each experimental condition and the control IgG treatment for both HDF and HGF populations (N = 3; *P < 0.05). The direction of the underlying nanogroove long axis is indicated by white arrows.

Figure 11

Independent inhibition of integrin αvβ3 and integrin β1 demonstrates a functional disparity among (A) HDF and (B) HGF. Immunofluorescent staining of vinculin (green), tensin-1 (red), and nuclei (blue) within HDF and HGF populations, following a 24 h timepoint, upon a 900 nm periodicity nanogroove surface. The direction of the underlying nanogroove long axis is indicated by white arrows.

Figure 12

Inhibition of integrin β1 results in tensin-1 recruitment to αvβ3 peripheral adhesion sites in (A) HDFs, but not (B) HGFs. Immunofluorescent staining of tensin-1 (green), integrin β1 (red), and nuclei (blue) within HDF and HGF populations following a 24 h timepoint. The direction of the underlying nanogroove long axis is indicated by white arrows.