Microtopographical features generated by photopolymerization recruit RhoA/ROCK through TRPV1 to direct cell and neurite growth

Abstract

Cell processes, including growth cones, respond to biophysical cues in their microenvironment to establish functional tissue architecture and intercellular networks. The mechanisms by which cells sense and translate biophysical cues into directed growth are unknown. We used photopolymerization to fabricate methacrylate platforms with patterned microtopographical features that precisely guide neurite growth and Schwann cell alignment. Pharmacologic inhibition of the transient receptor potential cation channel subfamily V member 1 (TRPV1) or reduced expression of TRPV1 by RNAi significantly disrupts neurite guidance by these microtopographical features. Exogenous expression of TRPV1 induces alignment of NIH3T3 fibroblasts that fail to align in the absence of TRPV1, further implicating TRPV1 channels as critical mediators of cellular responses to biophysical cues. Microtopographic features increase RhoA activity in growth cones and in TRPV1-expressing NIH3T3 cells. Further, Rho-associated kinase (ROCK) phosphorylation is elevated in growth cones and neurites on micropatterned surfaces. Inhibition of RhoA/ROCK by pharmacological compounds or reduced expression of either ROCKI or ROCKII isoforms by RNAi abolishes neurite and cell alignment, confirming that RhoA/ROCK signaling mediates neurite and cell alignment to microtopographic features. These studies demonstrate that microtopographical cues recruit TRPV1 channels and downstream signaling pathways, including RhoA and ROCK, to direct neurite and cell growth.

Keywords: Cell signaling; Micropatterning; Nerve guide; Photopolymerization; Spiral ganglion neuron; Surface topography.

Figure 1

Method of measuring neurite and cell alignment to micropatterned surfaces. A. The overall length of the longest neurite extending from the cell body and the end-to-end distance of the micropattern are measured in using the segmented line tool in ImageJ. Neurite alignment is defined as the ratio of total neurite length (T_L) to aligned length (A_L) i.e. the end-to-end distance. Values close to unity imply strong alignment to the pattern. B. Cell alignment is determined by determining the angle (θ) of an ellipse drawn around the long axis of the cell to the horizontal using ImageJ.

Figure 2

Photopolymerized micropatterns direct neurite growth. A. Schematic demonstrating micropattern fabrication via photopolymerization. Pre-polymer mixture of HMA, HDDMA and photoinitiator is UV cured beneath 50-µm periodicity glass-chrome Ronchi rule photomasks for patterned samples (upper row), or with glass microscope slides for unpatterned samples (lower row). Pattern feature height is tuned by modulating total exposure time. B-C. Representative SEM micrographs of top-down (B) and angled cross-section (C) views of micropatterned substrates. D-E. SGN neurites grow in random directions on unpatterned substrates (D) but orient to the direction of the pattern (horizontal) on micropatterned surfaces (E).

Figure 3

Inhibition of TRP channels decreases neurite alignment to micropatterned surfaces. A, C-F. SKF96365 (15 µM, n = 75), Ruthenium Red (2 µM, n = 105) and gentamicin (200 µM, n = 110) each significantly decrease SGN neurite alignment on micropatterned substrates compared to control cultures (n = 99). B. GsTMx-4 (15 µM, n = 57), a peptide inhibitor of mechanosensitive ion channels, has no significant effect on neurite alignment (n = 64 in control group). SGN neurite alignment to micropatterns is similar for cultures maintained with or without GsTMx-4. *p < 0.05, Kruskal-Wallis one way ANOVA on ranks followed by Dunn’s method (A). n value in this figures are the number of neurites scored in each condition. Data represent means ± SEM.

Figure 4

TRPV1 mediates neurite and NIH3T3 cell alignment to micropatterned surfaces. A-C. Expression of TRPV1 in SGN growth cones immunostained with anti-NF200 (A, green) and anti-TRPV1 (B, red) antibodies with combined labeling (C). D. Relative TRPV1 gene expression determined by real-time RT-PCR in cultures transfected with scrambled or TRPV1-targeted DsiRNA oligonucleotides. E. Transfection of cultures with TRPV1-targeted DsiRNA oligonucleotides significantly decreases SGN neurite alignment on micropatterned substrates compared to transfection of a scrambled oligonucleotide (n = 97 in each group). F, G. Images of SGN neurite alignment in cultures treated with scrambled (F) or TRPV1-targeted DsiRNA oligonucleotides (G). H. Transfection of NIH3T3 cells with a TRPV1 expression vector (n = 120) significantly increases 3T3 alignment to micropatterns compared to cells transfected with an empty plasmid (n = 155). I, J. Images of NIH3T3 cells cultured on micropatterns and co-transfected with empty and GFP expression plasmids (I) or TRPV1 and GFP expression plasmids (J). *p < 0.05, Student t-test (D), or Mann-Whitney test (E, H). Data represent means ± SEM.

Figure 5

Micropatterned surfaces activate RhoA to mediate neurite alignment. A-B. Interaction of DRGN growth cones with pattern edges increases RhoA activity and decreases Cdc42/Rac activity detected by in situ Rho GTPase activity assay. The average RBD-GST labeling (A), indicating RhoA activity, is significantly higher in growth cones interacting with sloping transitions (edge) of the pattern (n = 25) compared with growth cones in relatively flat regions (n = 23). The extent of RBD-GST labeling of growth cones interacting with the edge (n = 25) is comparable to growth cones in flat regions that are in cultures treated with a RhoA activator, RAII (1 µg/ml, n = 12). Conversely, PBD-GST labeling (B), indicating Cdc42/Rac activity, is higher in growth cones remaining in the flat regions of the pattern (n = 36) compared to those at the pattern edge (n = 49). C-E. RhoA activity detected by a FRET based biosensor is significantly higher in growth cones on patterned substrates (D, n = 26) compared to unpatterned controls (C, n = 25). *p < 0.05, one way ANOVA with post hoc Tukey (A), or Student t-test (B,E). Data represent means ± SEM.

Figure 6

RhoA activity facilitate neurite alignment to micropatterned surfaces. A. Treatment of cultures with RAII (1, 2 µg/ml) significantly increases neurite alignment on patterns of 1 µm channel depth compared to cultures without RAII (n = 100 in each group). B-C. Representative images of cultures in the absence (B) or presence (C) of RAII and immunostained with anti-neurofilament 200 (NF200). D. Treatment of cultures with Rho inhibitor C3 Transferase (C3T, 1, 2 µg/ml)) significantly reduces neurite alignment to micropatterns of 2–3 µm channel depth compared to cultures without C3T (n = 100 in each group). E-F. Images of cultures in the absence (E) or presence of C3T (F). *p < 0.05, Kruskal-Wallis one way ANOVA on ranks followed by Dunn’s method (A, D). Data represent means ± SEM.

Figure 7

ROCK activity facilitates neurite alignment to micropatterned surfaces. A-C. pROCK immunofluorescence intensity is significantly higher in SGN distal neurites and growth cones on micropatterned substrates (C) compared to those on unpatterned surfaces (B) (n = 42 in each group). D-I. SGN neurite alignment on micropatterned substrates is significantly reduced in cultures treated with ROCK inhibitors H1152 (0.1, n = 250; 1 µM, n = 178) or Y27632 (10, n = 149; 100 µM, n = 129) compared to control (n = 120 and 132, respectively). J. Transfection of spiral ganglion cultures with DsiRNA oligonucleotides targeting ROCK1 or ROCK2 significantly reduces mRNA expression as determined by real-time RT-PCR compared to cultures transfected with a scrambled, non-targeted oligonucleotide. K-N. Neurite alignment in cultures transfected with ROCK1 DsiRNA (n = 81) or ROCK2 DsiRNA (n = 83) is significantly decreased compared to alignment when treated with a scrambled oligonucleotide (n = 80) on micropatterned substrates. *p < 0.05, Student t-test (A, J) or Kruskal-Wallis one way ANOVA on ranks followed by Dunn’s method (D, E, K). Data represent means ± SEM.

Figure 8

Inhibition of ROCK disrupts spiral ganglion Schwann cells (SGSCs) alignment to micropatterned surfaces. A. Treatment of SGSC cultures with ROCK inhibitor H1152 significantly reduces SGSC alignment to micropatterns compared to SGSCs maintained in the absence of H1152 (n = 50 in each group). SGSC alignment in cultures treated with H1152 and grown on micropatterned surfaces is nearly random, comparable to SGSC grown on unpatterned surfaces. B-C. Images of SGSC alignment in cultures plated on micropatterned surfaces and maintained in the absence (B) or presence (C) of H1152. *p < 0.05, Kruskal-Wallis one way ANOVA on ranks followed by Dunn’s method. Data represent means ± SEM.

Figure 9

TRPV1 transfection of 3T3 cells significantly increases 3T3 alignment to micropatterns by facilitating RhoA/ROCK activity. A. Western blot of protein lysates from NIH3T3 and NIH3T3-TRPV1 cells probed with anti-TRPV1 antibody. The blots were stripped and reprobed with anti-β-actin antibody to verify equal protein loading. B. RhoA activity in NIH3T3 cells on patterned (Pat) substrates or NIH3T3-TRPV1 cells on unpatterned (Unpat) or patterned substrates normalized to 3T3 cells on unpatterned substrates. *p < 0.05, one way ANOVA followed by Tukey post hoc test. C,D. NIH3T3 or NIH3T3-TRPV1 cells on patterned substrates and labeled with Alexa Fluor 568 phalloidin. E. Alignment of NIH3T3 or NIH3T3-TRPV1 cells on unpatterned or patterned substrates in the presence or absence of H1152 (0.1 µM) (n = 100 in each group). **p < 0.05, Kruskal-Wallis one way ANOVA on ranks followed by Dunn’s method compared to all other conditions. Data represent means ± SEM.