Brain-derived neurotrophic factor regulation of retinal growth cone filopodial dynamics is mediated through actin depolymerizing factor/cofilin

Abstract

The molecular mechanisms by which neurotrophins regulate growth cone motility are not well understood. This study investigated the signaling involved in transducing BDNF-induced increases of filopodial dynamics. Our results indicate that BDNF regulates filopodial length and number through a Rho kinase-dependent mechanism. Additionally, actin depolymerizing factor (ADF)/cofilin activity is necessary and sufficient to transduce the effects of BDNF. Our data indicate that activation of ADF/cofilin mimics the effects of BDNF on filopodial dynamics, whereas ADF/cofilin inactivity blocks the effects of BDNF. Furthermore, BDNF promotes the activation of ADF/cofilin by reducing the phosphorylation of ADF/cofilin. Although inhibition of myosin II also enhances filopodial length, our results indicate that BDNF signaling is independent of myosin II activity and that the two pathways result in additive effects on filopodial length. Thus, filopodial extension is regulated by at least two independent mechanisms. The BDNF-dependent pathway works via regulation of ADF/cofilin, independently of myosin II activity.

Figure 1.

BDNF regulates filopodial dynamics through a ROCK-dependent pathway. A-H, Phalloidin staining of F-actin in example growth cones after treatment with three ROCK inhibitors. Inhibition of ROCK with 500 nm H-1152 or 50 μm HA1077 for 30 min, or 10 μm Y-27632 for 1 hr (C, E, G), enhanced filopodial length by 29-37% relative to control (A). The addition of 160 nm BDNF for 30 min to drug-treated cultures had no additive effect on filopodial length (B, D, F, H, I). Scale bar, 10 μm. J, ROCK inhibitors have a minor effect on filopodial number, but BDNF treatment of drug-treated cultures further enhanced the number to levels similar to BDNF treatment alone. Data are mean ± SEM from a minimum of three independent experiments. ^*p < 0.001; ^**p < 0.005; statistical difference relative to control. For all treatments in I, p < 0.001 indicates statistical significance relative to control (two-sample t test).

Figure 5.

Inhibition of myosin II enhanced filopodial length, but growth cones exhibit morphological differences compared with BDNF treatment. A-D, (S)-(-)-Blebbistatin had a dose-dependent effect on filopodial length and number of retinal growth cones stained using Alexa568-conjugated phalloidin. E, A 50 μm concentration of blebbistatin increased filopodial length by ∼47%. F, A 50 μm concentration of blebbistatin had a small effect on filopodial number, whereas lower concentrations had no effect. G, Time-lapse images of a growth cone treated with 20 μm blebbistatin. The growth cone displayed normal motility before the addition of blebbistatin (-2:00). After the addition of blebbistatin (0:00), the lamellipodia retracted, whereas filopodia increased in length (4:00). There was a tendency for filopodia to detach and buckle after treatment with blebbistatin (arrows). H, Blebbistatin increased filopodia extension rates while causing a gradual reduction in the retraction rates. I, Time-lapse images of a BDNF-treated growth cone. Four minutes after the addition of 160 nm BDNF (4:00), filopodia length and number increased, whereas the lamellipodia remained intact. J, BDNF increased extension rates while having no effect on the retraction rates. K, Filopodia treated with blebbistatin exhibited a twofold reduction in the motility index, whereas BDNF-treated filopodia exhibited no change in filopodial motility. Filopodia extension and retraction dynamics were pooled into 2 min time bins. Scale bars, 10 μm. Data are percentage of control ± SEM from four independent experiments. Videomicroscopy results were obtained from at least six growth cones for each treatment. ^*p < 0.001; statistical significance relative to control (two-sample t test).

Figure 2.

ADF/cofilin is necessary and sufficient to transduce the effects of BDNF on filopodial dynamics. A, XAC(3A) was sufficient to enhance filopodial length to levels that mimicked BDNF treatment. BDNF had no additive effect on length on XAC(3A)-loaded growth cones. XAC(3E) and XAC(KK95,96QQ) alone had no effect on filopodial length but both blocked the stimulatory effects of BDNF. B, BDNF, XAC(3A), and XAC(3A) plus BDNF increased all filopodia lengths as indicated by a rightward shift of the length distribution curve. C, Although XAC(3A) alone did not enhance filopodial number to levels similar as BDNF treatment alone, XAC(3E) and XAC(KKQQ) blocked the BDNF-induced increases on filopodial number. D, E, BDNF-induced regulation of ADF/cofilin is mediated through p75 ^NTR. CHEX-induced increases on filopodial length and number were blocked by XAC(3E). Data were presented as percentage of control ± SEM from a minimum of four independent experiments. ^*p < 0.001; statistical difference relative to control (two-sample t test).

Figure 3.

BDNF causes a reduction in pAC levels on retinal growth cones that is mediated through ROCK. Growth cones treated with BDNF for up to 30 min displayed a gradual reduction in pAC staining levels (A, C, E, G). Growth cones were counterstained using Alexa568-conjugated phalloidin (B, D, F, H). I, Thirty-minute treatment with BDNF caused a 49% reduction in pAC levels relative to control treatment. J-O, ROCK inhibition with H-1152 or Y-27632 mimicked the effects of BDNF treatment on pAC levels. P, Treatment with H-1152 for 30 min or Y-27632 for 1 hr produced a 47 and 45% reduction, respectively, in pAC levels. Measurements were taken from at least 77 growth cones from a minimum of three independent experiments. Scale bars, 10 μm. ^*p < 0.001; ^**p < 0.005; statistical significance relative to control (two-sample t test).

Figure 4.

The 14-3-3ζ protein blocked the effects of BDNF on filopodial dynamics and pAC levels. The BDNF-induced increases on filopodial length (A) and number (B) were blocked by 14-3-3ζ. C, Although 14-3-3ζ alone had no effect on pAC levels, 14-3-3ζ blocked the BDNF-mediated reduction in pAC levels. Data in A and B are percentage of control ± SEM from six independent experiments. Data in C were taken from at least 84 growth cones from four independent experiments. ^*p < 0.001; statistical difference relative to control (two-sample t test).

Figure 6.

BDNF does not signal through myosin II to regulate filopodial dynamics. BDNF has additive effects on filopodial length (A) and number (B) on blebbistatin-treated growth cones. C, A 10 μm concentration of Y-27632 or 20 μm blebbistatin caused an ∼31 and ∼49% increase on filopodial length, respectively, whereas treatment with both Y-27632 and blebbistatin produced an 86% increase on filopodial length. D, Y-27632 and blebbistatin enhanced filopodial number by ∼25 and 12%, respectively. Treatment with both Y-27632 and blebbistatin enhanced filopodial number to similar levels as Y-27632 treatment alone. E, The effects of blebbistatin do not require ADF/cofilin activity. Contrary to BDNF treatment, XAC(3E) had no effect on blebbistatin-induced increases on length. In addition, treatment of XAC(3A) cultures with blebbistatin exhibited additive effects on length. F, Blebbistatin treatment of XAC(3E)-loaded growth cones caused a similar increase in filopodial number as blebbistatin treatment alone, whereas blebbistatin had additive effects on the number of filopodia on XAC(3A)-treated growth cones. Data are percentage of control ± SEM from a minimum of four independent experiments. ^*p < 0.001; ^**p < 0.005; statistical difference (two-sample t test).