Axon guidance in the developing ocular motor system and Duane retraction syndrome depends on Semaphorin signaling via alpha2-chimaerin

Abstract

Eye movements depend on correct patterns of connectivity between cranial motor axons and the extraocular muscles. Despite the clinical importance of the ocular motor system, little is known of the molecular mechanisms underlying its development. We have recently shown that mutations in the Chimaerin-1 gene encoding the signaling protein α2-chimaerin (α2-chn) perturb axon guidance in the ocular motor system and lead to the human eye movement disorder, Duane retraction syndrome (DRS). The axon guidance cues that lie upstream of α2-chn are unknown; here we identify candidates to be the Semaphorins (Sema) 3A and 3C, acting via the PlexinA receptors. Sema3A/C are expressed in and around the developing extraocular muscles and cause growth cone collapse of oculomotor neurons in vitro. Furthermore, RNAi knockdown of α2-chn or PlexinAs in oculomotor neurons abrogates Sema3A/C-dependent growth cone collapse. In vivo knockdown of endogenous PlexinAs or α2-chn function results in stereotypical oculomotor axon guidance defects, which are reminiscent of DRS, whereas expression of α2-chn gain-of-function constructs can rescue PlexinA loss of function. These data suggest that α2-chn mediates Sema3-PlexinA repellent signaling. We further show that α2-chn is required for oculomotor neurons to respond to CXCL12 and hepatocyte growth factor (HGF), which are growth promoting and chemoattractant during oculomotor axon guidance. α2-chn is therefore a potential integrator of different types of guidance information to orchestrate ocular motor pathfinding. DRS phenotypes can result from incorrect regulation of this signaling pathway.

Fig. 1.

Schematic diagram of ocular motor development, and expression patterns of Sema3A, Sema3C, PlexinA1, and PlexinA2. (A–C) Diagrams of development of the ocular motor system in the chicken embryo at E4–E6 in a lateral view of a chick head (from ref. 7). Abducens nerve (VI) is shown in green, oculomotor nerve (III) in blue, trochlear nerve (IV) in purple, and muscles are shown in red. Muscle abbreviations are as presented in the text. MB, midbrain; HB, hindbrain. (D–G) Transverse sections through E5–E6 chicken midbrain (stages as labeled), showing in situ hybridization for PlexinA1 or PlexinA2 as labeled or PlexinA2/Islet1 double in situ hybridization (Islet1 mRNA detection, in red). (H–S) In situ hybridization for Sema3A or Sema3C, as labeled, in the chick head at E4 or E6. (H and N) Sagittal sections through the head at E4. (M and T) Transverse sections at rhombomere 2/3 level. (I–L and O–S) Transverse sections through the periocular region at E6, showing in situ hybridization either singly or in combination with immunostaining for nerves (neurofilament H; green) and muscles (sarcomeric myosin; red). (Scale bar, 150 μm in D–G; 400 μm in H, M, N, and T; and 80 μm in all remaining panels.)

Fig. 2.

In vivo phenotypes in the ocular motor system following electroporation of shRNAs for PlexinA1 and α2-chn. Confocal montages at E6 of the periocular regions of chicken embryos electroporated at E2 with control plasmid (A) or with shRNAs to knock down expression of PlexinA1 (B), α2-chn (C), or both (D). Whole mounts were immunostained to show axons in green and muscles in red. Midbrain is Left and ventral at Bottom in all panels. White arrows show defasciculating or overshooting axons; arrowheads show axons ectopically directed toward the LR muscle. Asterisk shows ciliary ganglion; muscle abbreviations are as presented in text. (Scale bar, 200 μm in A–D.)

Fig. 3.

α2-chn and PlexinAs mediate growth cone collapse (GCC) and axon outgrowth in oculomotor neurons in vitro. (A–D) Illustrative examples of entire oculomotor neurons (A), and normal (B), or collapsed growth cones (C), visualized using an antibody to Islet1/2 or phalloidin to label F-actin. (Scale bar, 20 μm.) (D) Histograms showing percentage of collapsed growth cones in cultures transfected with either control GFP or G228S-α2-chn plasmids, in response to treatment with control (Fc) or Sema3A. Data were analyzed by two-way ANOVA. Post hoc analysis shows that Sema3A induces GCC in control (GFP transfected) neurons (*P < 0.01) and G228S-α2-chn also increases the percentage of GCC in absence of any treatment: ^#P < 0.01. (E) Histograms showing collapsed growth cones in oculomotor cultures transfected with scrambled control (SCR), PlexinA1, -A2, or α2-chn shRNA. Cultures were challenged with either control (Fc reagent), Sema3A, or Sema3C. Data were analyzed by two-way ANOVA. Post hoc analysis shows that Sema3A or Sema3C induce GCC in control (SCR shRNA transfected) neurons (*P < 0.001) and PlexinA1/A2 and α2-chn shRNA reduce Sema3A/3C-induced GCC to control levels; in all cases P > 0.05. # denotes that all these conditions are significantly different from Sema3A/C-induced GCC (P < 0.05). (F and G) Histograms showing outgrowth of oculomotor neurons transfected with either a control plasmid (SCR shRNA or GFP), α2-chn shRNA (F), or G228S-α2-chn (G) and treated with CXCL12 or HGF. Data were analyzed by a pairwise multiple comparison procedure (Holm–Sidak method). (F) In SCR shRNA-expressing neurons, both CXCL12 and HGF increase outgrowth relative to controls (*P < 0.001) and α2-chn shRNA-transfection abrogates this increase (^#P < 0.001). (G) In GFP-transfected neurons both CXCL12 (ψ P = 0.044) and HGF (ψ P = 0.012) increase outgrowth relative to controls. Neurons transfected with G228S-α2-chn display an increase in outgrowth in absence of treatment (^@P < 0.001). This increase is similar in CXCL12 and HGF-treated neurons. All histograms show mean ± SEM.

Fig. 4.

Coelectroporation of PlexinA2 shRNA and G228S-α2-chn produces a normal oculomotor axon projection in vivo. Confocal image montages at E6 of the periocular regions of chicken embryos electroporated at E2 either with 1 μg/μL (A) or 0.5 μg/μL (B) of G228S-α2-chn, and combinatorial electroporation of G228S-α2-chn together with PlexinA2 shRNA (C) or PlexinA1 shRNA (D). Immunostaining was with antibodies to GFP for electroporated axons (green) or to sarcomeric myosin for muscles (red). Midbrain is Left, ventral at Bottom in all panels. White solid arrow shows region where the majority of oculomotor axons stall. Asterisk shows ciliary ganglion; muscle abbreviations are as in text. (Scale bar, 200 μm).

Fig. 5.

Schematic diagram illustrating in vivo phenotypes resulting from manipulations in the chicken embryo. (A) Phenotypes resulting from abrogation of PlexinA or α2-chn signaling, expression of α2-chn gain-of-function mutant construct, or rescue experiment with Plexin loss of function, α2-chn gain of function. (B) Model from E4–E6, based on timing of axon projections shown in Fig. 1 A–C. Model of the role of chemoattraction and repulsion in targeting axon projections with color-coded cues; hatching represents expression of multiple cues.