Semaphorin 3A inhibits growth of adult sympathetic and parasympathetic neurones via distinct cyclic nucleotide signalling pathways

Abstract

Background and purpose: Semaphorin 3A (Sema3A) is an important secreted repulsive guidance factor for many developing neurones. Sema3A continues to be expressed in adulthood, and expression of its receptor, neuropilin-1 (Nrp-1), can be altered by nerve injury. Autonomic neurones innervating the pelvic viscera are particularly susceptible to damage during pelvic surgical procedures, and failure to regenerate or aberrant growth of sympathetic and parasympathetic nerves lead to organ dysfunction. However, it is not known if adult pelvic neurones are potential targets for Sema3A.

Experimental approach: The effects of Sema3A and activation or inhibition of cyclic nucleotide signalling were assessed in adult rat pelvic ganglion neurones in culture using a growth cone collapse assay.

Key results: Sema3A caused growth cone collapse in both parasympathetic and sympathetic neurones expressing Nrp-1. However, the effect of Sema3A was mediated by distinct cyclic nucleotide signalling pathways in each neurone type. In parasympathetic neurones, cAMP and downstream activation of protein kinase A were required for growth cone collapse. In sympathetic neurones, cGMP was required for Sema3A-induced collapse; cAMP can also cause collapse but was not required. Sema3A-mediated, cGMP-dependent collapse in sympathetic neurones may require activation of cyclic nucleotide-gated ion channels (CNGCs).

Conclusions and implications: We propose that Sema3A is an important guidance factor for adult pelvic autonomic neurones, and that manipulation of their distinct signalling mechanisms could potentially promote functional selective regeneration or attenuate aberrant growth. To our knowledge, this is also the first study to implicate CNGCs in regulating growth cone dynamics of adult neurones.

Figure 1

Sema3A causes growth cone collapse in adult sympathetic and parasympathetic neurones. (A,B) Inverted fluorescence images of cultured sympathetic (A; TH-positive) and parasympathetic (B; NOS-positive) pelvic ganglion neurones. (C, D) Intact growth cones of sympathetic (C) and parasympathetic (D) neurones. (E, F) Collapsed growth cones of sympathetic (E) and parasympathetic (F) neurones (merged images of TH or NOS with actin staining). Note the retraction of F-actin rich filopodia in collapsed growth cones. (G,H) Sema3A increased growth cone collapse in both sympathetic and parasympathetic neurones, at all time points tested (n= 4). Dunnett's test: *P < 0.05, **P < 0.01, ***P < 0.001 versus control. Scale bar = 100 µm in (A and B), 12 µm in (C–F). NOS, nitric oxide synthase; TH, tyrosine hydroxylase.

Figure 2

Sema3A-induced growth cone collapse occurs only in sympathetic and parasympathetic neurones expressing Nrp-1. (A–D) Confocal images of growth cones from sympathetic (TH-positive: A, intact; B, collapsed) and parasympathetic (NOS-positive: C, intact; D, collapsed) pelvic ganglion neurones expressing Nrp-1. Merged images of TH and NOS with Nrp-1 immunostaining are shown in (B) and (D) respectively. Arrowhead in (D) shows an Nrp-1-negative collapsed growth cone in close proximity to an Nrp-1-positive collapsed growth cone. (E, F) Sema3A caused growth cone collapse in sympathetic (E) and parasympathetic (F) neurones with Nrp-1-immunoreactivity (Nrp-1-IR), but did not cause collapse in Nrp-1-negative neurones (n= 3). Tukey's test: ***P < 0.001 versus Sema3A-negative control. Scale bar = 10 µm in (A–D). NOS, nitric oxide synthase; Nrp-1, neuropilin-1; Sema3A, semaphorin 3A; TH, tyrosine hydroxylase.

Figure 3

Adenylyl cyclase-cAMP signalling mediates Sema3A-induced growth cone collapse in parasympathetic neurones. (A) Inhibition of AC (DDA) prevented Sema3A-induced growth cone collapse, and stimulation of AC (FSK) caused growth cone collapse in NOS-positive parasympathetic neurones. (B) Cyclic AMP-dependent phosphodiesterase inhibition (ROL) did not augment Sema3A-induced collapse, but caused growth cone collapse by itself. (C) However, inhibition of sGC (ODQ) did not affect Sema3A-induced growth cone collapse, and stimulation of sGC (SNP) did not cause growth cone collapse. (D,E) Growth cones of parasympathetic neurones identified by VAChT-immunoreactivity (D, intact; E, collapsed). Merged images of VAChT and F-actin staining are shown in (E). Similar to NOS-positive neurones, inhibition of AC (F), but not sGC (G), prevented Sema3A-induced growth cone collapse in VAChT-positive/NOS-negative parasympathetic neurones. In addition, stimulation of AC (F), but not sGC (G), caused growth cone collapse. Tukey's test: *P < 0.05, **P < 0.01, ***P < 0.001 versus control; ##P < 0.01, ###P < 0.001 versus Sema3A (n= 3–6 for all experiments). Scale bar = 10 µm in (D) and (E). DDA, 2′,5′-dideoxyadenosine; FSK, forskolin; ODQ, 1H-[1,2,4]oxadiazole[4,3-a]quinoxalin-1-one; ROL, rolipram; Sema3A, semaphorin 3A; SNP, sodium nitroprusside; TH, tyrosine hydroxylase; VAChT, vesicular acetylcholine transporter.

Figure 4

Protein kinase A signalling mediates Sema3A-induced growth cone collapse in nitrergic (NOS-positive) parasympathetic neurones. (A) The PKA inhibitor, KT-5720, prevented Sema3A-induced growth cone collapse in NOS-positive neurones. (B) Similar results were found with Rp-cAMPS. (C, D) In contrast, two separate inhibitors of PKG (C, KT-5823; D, Rp-8-Br-cGMPS) did not significantly affect Sema3A-collapse, but both caused growth cone collapse on their own. Tukey's test: *P < 0.05, **P < 0.01, ***P < 0.001 versus control; #P < 0.05, ###P < 0.001 versus Sema3A (n= 3–4 for all experiments). KT-5720, (9R,10S,12S)-2,3,9,10,11,12-hexahydro-10-hydroxy-9-methyl-1-oxo-9,12-epoxy-1H-diindolo[1,2,3-fg:3′,2′,1′-kl]pyrrolo[3,4-i][1,6]benzodiazocine-10-carboxylic acid, hexyl ester; KT-5823, (9S,10R,12R)-2,3,9,10,11,12-hexahydro-10-methoxy-2,9-dimethyl-1-oxo-9,12-epoxy-1H-diindolo[1,2,3-fg:3′,2′,1′-kl]pyrrolo[3,4-i][1,6]benzodiazocine-10-carboxylic acid, methyl ester; Rp-cAMPS, Rp-adenosine-3′,5′-cyclic monophosphorothioate; Rp-8-Br-cGMPS, Rp-8-bromoguanosine-3′,5′-cyclic monophosphorothioate; Sema3A, semaphorin 3A.

Figure 5

Soluble guanylyl cyclase-cGMP signalling mediates semaphorin 3A-induced growth cone collapse in sympathetic neurones. (A) Inhibition of sGC (ODQ) prevented Sema3A-induced growth cone collapse and stimulation of sGC (SNP) also caused collapse. (B) In contrast to parasympathetic neurones, inhibition of AC (DDA) did not prevent Sema3A-induced growth cone collapse in sympathetic neurones. However, stimulation of AC (FSK) did cause growth cone collapse. (C) Neither cAMP- nor cGMP-dependent phosphodiesterase inhibition (ROL and ZAP, respectively) augmented Sema3A-induced growth cone collapse, nor did these inhibitors significantly affect collapse on their own. However, when added together, ROL and ZAP caused growth cone collapse. Tukey's test: *P < 0.05, **P < 0.01, ***P < 0.001 versus control; ###P < 0.001 versus Sema3A (n= 4 for all experiments). DDA, 2′,5′-dideoxyadenosine; FSK, forskolin;; ROL, rolipram; Sema3A, semaphorin 3A; SNP, sodium nitroprusside; ZAP, zaprinast.

Figure 6

Cyclic nucleotide-gated ion channels, not PKA or PKG signalling, mediates Sema3A-induced growth cone collapse in sympathetic neurones. (A, B) Inhibition of PKG with KT-5823 (A) or Rp-8-Br-cGMPS (B) did not significantly alter Sema3A-induced growth cone collapse in sympathetic neurones, but both inhibitors caused growth cone collapse on their own. (C, D) Similarly, inhibition of PKA (C, KT-5720; D, Rp-cAMPS) did not alter Sema3A-collapse, but both of these inhibitors caused growth cone collapse. (E,F) However, inhibition of CNGCs (E, DCB; F, LCD) prevented Sema3A-collapse in sympathetic neurones. Furthermore, Sema3A prevented the collapse caused by DCB and LCD on their own. Tukey's test: *P < 0.05, **P < 0.01, ***P < 0.001 versus control (n= 3–5 for all experiments). **P < 0.01, ***P < 0.001 versus control; ##P < 0.01 versus Sema3A; ⁺P < 0.05, ⁺⁺P < 0.01 versus DCB or LCD. DCB, 2′,4′-dichlorobenzamil; KT-5720, (9R,10S,12S)-2,3,9,10,11,12-hexahydro-10-hydroxy-9-methyl-1-oxo-9,12-epoxy-1H-diindolo[1,2,3-fg:3′,2′,1′-kl]pyrrolo[3,4-i][1,6]benzodiazocine-10-carboxylic acid, hexyl ester; KT-5823, (9S,10R,12R)-2,3,9,10,11,12-hexahydro-10-methoxy-2,9-dimethyl-1-oxo-9,12-epoxy-1H-diindolo[1,2,3-fg:3′,2′,1′-kl]pyrrolo[3,4-i][1,6]benzodiazocine-10-carboxylic acid, methyl ester; LCD, L-cis-diltiazem; Rp-cAMPS, Rp-adenosine-3′,5′-cyclic monophosphorothioate; Rp-8-Br-cGMPS, Rp-8-bromoguanosine-3′,5′-cyclic monophosphorothioate; Sema3A, semaphorin 3A.

Figure 7

Proposed model for cyclic nucleotide-dependent signalling of growth cone collapse in adult autonomic neurones in response to Sema3A. In parasympathetic neurones, Sema3A binds to Nrp-1 and activates adenylyl cyclase, leading to the production of cAMP. In turn, cAMP activates PKA, which mediates growth cone collapse. However, cGMP may positively regulate growth cone dynamics independent of Sema3A signalling, since inhibition of PKG caused collapse. In sympathetic neurones, Sema3A binds Nrp-1 and stimulates cGMP production via soluble guanylyl cyclase. In turn, cGMP preferentially activates cyclic nucleotide-gated channels (CNGCs) (1), which mediate growth cone collapse. However, if CNGCs are inhibited, cGMP activates PKG (2), which does not cause collapse. Growth cone collapse is also caused by cAMP in sympathetic neurones, but this occurs independently of Sema3A and PKA signalling. AC, adenylyl cyclase; CNGCs, cyclic nucleotide-gated ion channel; Nrp-1, neuropilin-1; PKA, protein kinase A; PKG, protein kinase G; Sema3A, semaphorin 3A; sGC, soluble guanylyl cyclase.