Overview

This study, co-authored by Jungmok Seo (then in the School of Electrical and Electronic Engineering, Yonsei University), who now leads the BLISS Lab at Yonsei University, built graphene oxide (GO) culture substrates that combine microgrooves, ridges and nanoscale roughness, and tested their effect on human neural stem cells (hNSCs). On substrates with 5 μm grooves, 37.5% of the recorded cells generated sodium currents, compared with 10% on flat substrates and 25.0% on unpatterned GO.

Background

Topographical features of culture substrates can regulate stem cell fate through integrin clustering, focal adhesion assembly and cytoskeletal reorganization, and graphene-based surfaces have been reported to promote neural differentiation. Most earlier topographical studies used simple patterns and did not assess whether the differentiated cells were electrophysiologically functional. The authors tested whether hierarchical GO topography of specific dimensions could strengthen focal adhesion signaling in hNSCs and yield functional neuron-like cells.

Design and approach

Photoresist line patterns with groove widths of 5, 10 and 20 μm (spacing ratio of 1) were formed on SiO2, functionalized with a positively charged 3-aminopropyltriethoxysilane (APTES) monolayer, and coated with negatively charged GO by electrostatic attachment. Annealing and lift-off left GO lines with ridges at both edges and a rough GO surface between them. The resulting GO-based patterned substrate (GPS) therefore presents three length scales: microgrooves that guide cell elongation, ridges with an average height of 179 ± 19 nm, and nanoroughness that offers contact points for integrins. Unpatterned GO-coated substrates (GS) showed roughness with a root-mean-square value of 3.86 nm.

How it was evaluated

hNSCs derived from human fetal brain were cultured on fibronectin-coated flat SiO2 (FS), GS and GPS without the mitogens basic fibroblast growth factor and leukemia inhibitory factor, so that differentiation occurred spontaneously. After 5 days, the authors assessed integrin clustering, F-actin alignment and focal adhesion proteins by immunostaining, quantitative real-time PCR and western blotting, and neuronal and astrocytic markers by immunostaining and PCR. Mechanisms were probed with β1 integrin antibodies and inhibitors of myosin II (blebbistatin), ROCK (Y27632) and MEK-ERK (U0126). Whole-cell patch clamp with tetrodotoxin (TTX) was used to test electrophysiological function.

Key results

GO nanoroughness alone increased β1 integrin clustering relative to FS, and the GPS with 5 μm grooves produced the most extensive clustering among all substrates. F-actin was randomly extended on FS and GS but aligned along the grooves and ridges on GPS. Gene expression of vinculin, paxillin and FAK was highest on the 5 μm GPS, and phosphorylation of FAK at Y397 and of paxillin at Y118 was increased on that substrate.

Neuronal gene expression (Tuj1 and MAP2) was highest on the 5 μm GPS compared with FS, GS and GPS with 10 or 20 μm grooves, whereas GFAP and nestin expression did not differ significantly among groups. The Tuj1-positive population increased and the GFAP-positive population decreased on the 5 μm GPS, and these cells showed the most extensive neurite formation.

Blocking β1 integrin, or inhibiting myosin II or ROCK, disrupted alignment, reduced focal adhesion and decreased neuronal differentiation on the GPS, and U0126 completely reversed the increase in neuronal lineage cells. These treatments did not significantly change mitochondrial metabolic activity. Fibronectin adsorption was higher on GO substrates than on FS (p < 0.05) but did not differ between GS and the 5 μm GPS, so the authors suggested that the additional effect of GPS might be mainly due to its hierarchical topography.

Cells differentiated on the 5 μm GPS showed voltage-dependent sodium currents and action potentials that disappeared after treatment with 0.5 μM TTX. Among 8 to 10 analyzed cells per group, the proportion generating sodium currents was 37.5% on the 5 μm GPS, 25.0% on GS and 10% on FS.

Open questions

The authors noted that GO may influence differentiation through protein adsorption, altered surface properties or graphene-inherent electrical effects, and that these contributions need to be clarified in future studies.

Related research

RESEARCH NEWS

Research in the news

Sources used in this summary

This summary is based on the publisher PDF of the article, including figure legends; the Supporting Information (Live/Dead staining, fibronectin adsorption and inhibitor viability data) was not reviewed.

DOI: 10.1021/acsami.6b01804