Neuronal NADPH oxidase is required for neurite regeneration of Aplysia bag cell neurons

Abstract

NADPH oxidase (Nox), a major source of reactive oxygen species (ROS), is involved in neurodegeneration after injury and disease. Nox is expressed in both neuronal and non-neuronal cells and contributes to an elevated ROS level after injury. Contrary to the well-known damaging effect of Nox-derived ROS in neurodegeneration, recently a physiological role of Nox in nervous system development including neurogenesis, neuronal polarity, and axonal growth has been revealed. Here, we tested a role for neuronal Nox in neurite regeneration following mechanical transection in cultured Aplysia bag cell neurons. Using a novel hydrogen peroxide (H₂ O₂ )-sensing dye, 5'-(p-borophenyl)-2'-pyridylthiazole pinacol ester (BPPT), we found that H₂ O₂ levels are elevated in regenerating growth cones following injury. Redistribution of Nox2 and p40^phox in the growth cone central domain suggests Nox2 activation after injury. Inhibiting Nox with the pan-Nox inhibitor celastrol reduced neurite regeneration rate. Pharmacological inhibition of Nox is correlated with reduced activation of Src2 tyrosine kinase and F-actin content in the growth cone. Taken together, these findings suggest that Nox-derived ROS regulate neurite regeneration following injury through Src2-mediated regulation of actin organization in Aplysia growth cones.

Keywords: NADPH oxidase; growth cone; hydrogen peroxide; neurite regeneration; reactive oxygen species.

Figure 1.. Detection of H₂O₂ using BPPT, a novel arylboronate fluorescent probe.

(A) Synthetic scheme of BPPT (1) and its oxidation by H₂O₂ into HPPT (2). (a) N-bromosuccinimide (NBS), azobisisobutryronitrile (AIBN), α,α,α-trifluorotoluene (TFT), 80 ºC (84% yield). (b) 2-pyridylthioamide, ethanol, 90 ºC (microwave) (60% yield). (c) Ethyl iodide, K₂CO₃, acetone, 25 ºC. (d) Tetrakis(triphenylphosphine)palladium (Pd(PPh₃)₄), bis(pinacolato)diboron, KOAc, 1,4-dioxane (70% yield over 2 steps). (B) Normalized optical absorption and photoluminescence (PL) spectra of 1 and 2 (in ethanol). (C) Conversion of 1 (1.25 μM in pH 7.8 buffer; λ_em 472 nm) into 2 (λ_em 540 nm) after 10 min exposure to H₂O₂, as a function of H₂O₂ concentration.

Figure 2.. BPPT is a H₂O₂-sensitive dye that localizes mostly to the cytoplasm.

(A) DIC image of Aplysia bag cell neuronal growth cone, indicating locations of P domain (outlined in yellow), T zone, and C domain. (B,C) Bag cell neuron labeled with BPPT (B) and CRO (C). (D–F) DIC and fluorescence images of neuronal growth cone labeled with BPPT, and DiI. The BPPT and CRO intensities are higher in C domain than P domain, whereas the membrane marker DiI labels C and P domains uniformly. Scale bar: 20 μm.

Figure 3.. Validation of BPPT as H₂O₂-sensitive dye by imaging of live Aplysia bag cell growth cones.

(A–C) DIC and fluorescence images (BPPT and CRO) of bag cell with growth cone (Control; untreated). (D–F) DIC and fluorescence images (BPPT and CRO) of bag cell with growth cone exposed to 1 mM H₂O₂ for 10 minutes. Scale bar: 20 μm. (G) Treatment with 1 mM H₂O₂ increased intracellular H₂O₂ levels by 129% as measured by normalized BPPT/CRO ratios in the P domain of Aplysia bag cell neuronal growth cones. Data for control (n=28) and experiment (n=33) collected from three separate cell platings. (H) Normalized intracellular H₂O₂ levels before and after treatment of growth cones with control solution (n=17) and H₂O₂ at 10 μM (n=19), 100 μM (n=23), and 1 mM H₂O₂ (n=19). Error bars indicate ± SEM. Statistics: D’Agostino–Pearson omnibus normality test; Mann–Whitney non-parametric test (G), ****p< 0.0001; two-tailed paired t-test (H); ***p< 0.001, ****p< 0.0001.

Figure 4.. In vitro transection of Aplysia bag cell neurons followed by neurite regeneration.

(A) Bag cell neuron with neurite just prior to transection by microneedle; location of cut is marked with white arrowhead. (B,C) 2 and 4 hours after transection, respectively; the distal neurite has regrown from the initial cut site. Scale bar: 100 μm. (D) High-magnification (90x) DIC images of regenerating neurite after injury at select timepoints. Cut location is marked with white arrowhead; glass needle tip is marked with a red arrowhead. A new growth cone at the tip of the cut neurite is observed within 10–20 minutes. Scale bar: 10 μm.

Figure 5.. H₂O₂ levels in regenerating growth cones are elevated following neurite transection.

(A–C) DIC images of growth cone before (A), immediately after (B), and 20 minutes after neurite transection (C). (D,E) Fluorescence images (BPPT and CRO) of regenerating growth cone shown in (C). (F–H) Images of control growth cone for comparison. Scale bar: 10 μm. (I) Neurite transection significantly increased H₂O₂ levels in the P domain of regenerating growth cones by 45%. Data for control (n=25) and cut neurites (n=20) collected from five separate cell platings. Error bars indicate ± SEM. Statistical test: two-tailed Mann–Whitney; *p< 0.05.

Figure 6.. In vitro injury causes redistribution of p40^phox and Nox2.

(A) DIC image of an uninjured (control, CTL)) growth cone; P domain, T zone, and C domain are labeled. (B,C) immunofluorescence images for Nox2 (B) and p40^phox (C). (C’) Magnified view of region of interest boxed in (C) shown as color overlay of Nox2 (red) and p40^phox (green). Arrows point to foci of co-localization. C domain boundary is marked with a dashed line. (D–F) Images of a regenerating growth cone after in vitro transection. (F’) Magnified view of region of interest boxed in (F) shown as color overlay. Scale bars: 10 μm in (C) and (F); 5 μm in (C’) and (F’). (G,H) Quantification of p40^phox/Nox2 co-localization in the P domain and T zone (G) and C domain (H) by calculating the Pearson correlation coefficient. Increased p40^phox/Nox2 co-localization is observed in the C domain following injury compared to uncut controls, but not in the P domain or T zone. Data shown for control (n=47) and cut neurons (n=23) collected from two separate cell platings. Average values ± SEM are shown. Statistics: Two-tailed unpaired student’s t-test; *p< 0.05.

Figure 7.. Growth rates of regenerating neurites are reduced by Nox inhibition.

Regenerating neurites following mechanical transection grew faster than control neurites that were not transected, following treatment with DMSO alone (p<0.05). Treatment with CST reduced both control and regenerative neurite growth completely at and above 0.5 μM CST. Two-tailed unpaired student’s t-test; *p< 0.05, ****p< 0.0001. Detailed statistics are provided in the supplementary information.

Figure 8.. Nox inhibitor CST reduces F-actin content in Aplysia bag cell growth cone.

(A,B) High magnification DIC and Alexa 568- phalloidin images of a control (CTL) growth cone treated with DMSO. The P domain shows the typical radial array of filopodial F-actin bundles. (C,D) Same set of images of a growth cone treated with 1 μM CST for 20 minutes. CST treatment caused flattening of growth cone T zone and P domain and reduction of F-actin content. (E,F) High magnification DIC and Alexa 568-phalloidin images of a regenerating growth cone following neurite transection. The F-actin in the P domain is not as organized as in untransected control growth cones. (G,H) Same set of images of a regenerating growth cone treated with 1 μM CST for 20 minutes. In vitro injury and CST treatment reduced the F-actin content and affected the F-actin organization in the growth cones. Scale bar: 20 μm.

Figure 9.. Nox inhibitor CST reduces the Src2 activation state in Aplysia growth cone.

(A-C) Control growth cone without transection and treated with vehicle (DMSO) was immunolabeled for pSrc2 (B) and Src2 (C). Outlined regions in (A) represent P domain and T zone (white dashed line) and C domain (orange dashed line). (D–F) Corresponding images of a regenerating growth cone. (G–I) Control growth cone without transection but treated with 1 μM CST for 20 minutes. (J–L) Corresponding images of a regenerating and CST-treated growth cone. Scale bar: 10 μm. (M,N) Quantification of pSrc2/Src2 ratio in the growth cone P domain and T zone (M) and C domain (N). Average values +/– SEM shown for control/DMSO (n=24), cut/DMSO (n=22), control /CST (n=33), and cut/CST (n=25). Data are from three separate cell platings. We found a higher pSrc2/Src2 ratio following injury when compared to uninjured controls in the P domain and T zone (two-tailed unpaired t-test, *p< 0.05). CST treatment reduced pSrc2/Src2 in the growth cone C domain for both uncut control (two-tailed Mann Whitney exact test, *p< 0.05), and injured neurites (two-tailed Mann Whitney exact test, **p< 0.005) compared to the corresponding DMSO-treated control or cut growth cones, respectively.