Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Feb 12.
Published in final edited form as: Neuron. 2009 Feb 12;61(3):359–372. doi: 10.1016/j.neuron.2008.12.022

FARP1 promotes the dendritic growth of spinal motor neuron subtypes through transmembrane Semaphorin6A and PlexinA4 signaling

BinQuan Zhuang 1, YouRong Sophie Su 1, Shanthini Sockanathan 1,✉
PMCID: PMC2654783  NIHMSID: NIHMS96378  PMID: 19217374

Summary

The dendritic morphology of neurons dictates their abilities to process and transmit information; however, the signaling pathways that regulate dendritic growth and complexity are poorly understood. Here we show that retinoids induce the expression of the FERM Rho-GEF protein FARP1 in the developing spinal cord. FARP1 is expressed in subsets of motor neurons and is enriched in dendrites of Lateral Motor Column (LMC) neurons that innervate the limb. FARP1 is necessary and sufficient to promote LMC dendritic growth but does not affect dendrite number or axonal morphology. We show that FARP1 serves as a specific effector of transmembrane Semaphorin6A and PlexinA4 signals to regulate LMC dendritic growth, and that its Rho-GEF domain is necessary for this function. These findings reveal that retinoid and Sema6A/PlexA4 signaling pathways intersect through FARP1 to control dendritic growth, and uncover the existence of subtype-specific signaling networks that control dendritic developmental programs in spinal motor neurons.

Introduction

The ability of neurons to efficiently receive, process, store and transmit information depends upon their polarized architecture that is comprised of arborized dendrites and an extended axon. Dendrites extending from the soma receive synaptic inputs from presynaptic axonal terminals, and exhibit stereotypic and often complex morphologies that are essential for their abilities to receive and process information (Jan and Jan, 2003). Emerging evidence implicates deficits in dendritic growth and complexity in the etiology of many neurological diseases such as epilepsy and autism, underscoring the pivotal role of dendritic structures in nervous system function (Pardo and Eberhart, 2007; Wong, 2008). However, the mechanisms that establish and regulate dendritic development, plasticity and maintenance remain poorly understood.

Several extrinsic signals that regulate dendritic growth and branching have been identified, and these include pathways that regulate axonal growth and guidance. The repellant molecule Slit promotes the dendritic branching and growth of mammalian cortical neurons, a process that is attenuated by the inclusion of a dominant negative version of its receptor, Robo (Whitford et al., 2002). In addition, Semaphorin (Sema) 3A, which can function as a chemoattractant or repulsive axonal guidance cue, regulates the dendritic growth and orientation of subclasses of cortical pyramidal neurons (Polleux et al., 2000; Fenstermaker et al., 2004; Tran et al., 2007; Mann et al., 2007). These signals likely regulate downstream molecules that include the Rho GTPases such as RhoA, Rac and Cdc42, which are critical regulators of actin and microtubule dynamics, key structural dendritic components (Etienne-Manneville and Hall, 2002). While the relationship between the Rho GTPases and the extrinsic signals that trigger axon growth and repulsion is beginning to emerge, the molecular mechanisms that relay signals impinging upon dendritic growth are less clear (Whitford et al., 2002; Jan and Jan, 2003). Indeed, it remains unresolved if axons and dendrites respond to the same growth and guidance signals, or if there are dedicated cues and intracellular signal transduction networks that are specific to one structure and not the other.

The Vitamin A derivative retinoic acid (RA) is a pivotal signaling molecule that regulates multiple processes during nervous system development, including neurite outgrowth (Maden, 2007). RA binds to heterodimers formed between the nuclear receptors RAR and RXR, and directly regulates gene transcription (Aranda and Pascual, 2001). Recent studies demonstrate that neurite outgrowth from explants of dorsal root ganglion (DRG) neurons is mediated by the retinoid receptor RARβ2 and that the lentiviral-mediated introduction of RARβ2 into adult DRGs, which normally lack RARβ2 expression, enables injured DRG axons to cross the dorsal root entry zone (Corcoran et al., 2002; Wong et al., 2006). These observations provide strong evidence that downstream genes regulated by RA signals play pivotal roles in regulating axon outgrowth; however, it remains unclear whether downstream effectors of RA signaling are able to regulate dendritic growth or complexity. Developing spinal motor neurons serve as an informative system to define the role of RA signaling pathways in nervous system development. RA signaling pathways are responsible for specifying motor neuron progenitors, controlling their differentiation into postmitotic motor neurons and regulating the specification, diversification and survival of limb-innervating Lateral Motor Column (LMC) motor neurons (Sockanathan and Jessell, 1998; Diez del Corral et al., 2003; Novitch et al., 2003; Sockanathan et al., 2003; Vermot et al., 2005; Ji et al., 2006). Notably, LMC motor neurons synthesize RA after motor neuron differentiation, specification and axonal targeting have occurred (Sockanathan and Jessell, 1998). This observation, combined with the knowledge that activated retinoid receptors regulate neurite outgrowth, led us to investigate if downstream genes induced by RA signals are involved in regulating the dendritic development of LMC motor neurons.

Here, we identify FARP1 (FERM, RhoGEF, and pleckstrin homology domain protein 1; also called chondrocyte-derived ezrin-like protein (CDEP) in humans) as a gene induced by RA signals that encodes a protein required for regulating the dendritic growth of embryonic LMC motor neurons (Koyano et al., 1997). We show by biochemical and functional analyses that FARP1 relays transmembrane Sema6A signals through the plexin receptor PlexA4 to increase the length of LMC motor neuron dendrites without affecting axons, and that these functions require the integrity of the FARP1 Rho-GEF domain. This study uncovers new roles for RA and Sema6A/PlexA4 signaling pathways in dendritic development, and reveals that they intersect through FARP1 to control the growth of LMC motor neuron dendrites during embryogenesis. Thus, columnar specific signaling programs operate to control the dendritic development of distinct motor neuron subtypes in the developing spinal cord.

Results

Identification and expression of FARP1 in the spinal cord

We isolated FARP1 using neural explant based subtractive screens that were designed to identify genes upregulated by RA signaling in spinal motor neurons (Rao and Sockanathan, 2005). Reverse northern analysis of Hamburger Hamilton stage (St) 14 chick ventral spinal cord explants shows robust induction of FARP1 expression when grown in the presence of retinol, consistent with a role for RA signals in upregulating FARP1 expression in the developing spinal cord (Figure 1A).

Figure 1. FARP1 is induced by RA and is localized to the dendrites of LMC motor neurons.

Figure 1

Open in a new tab

(A) Reverse northern blot shows Farp1, but not control clone 29.1, is induced by addition of retinol (ROL). Two bands are seen in the ROL lane due to traces of plasmid containing FARP1 cDNA (lower band) in addition to the PCR-amplified FARP1 fragment (upper band). (B–E) Right hemisegments of transverse sections of St 31 chick spinal cord. (B, C) In situ hybridization of Farp1 mRNA. Hatched black lines outline the spinal cord. (D, E) Confocal micrographs detecting FARP1 protein (F–H′) Confocal micrographs of ventral right quadrants of St 31 chick spinal cords. Boxes in panels F–H are magnified in panels F′–H′. Arrowheads mark the dendrites with co-localization of FARP1 and MAP2 proteins. * marks FARP1 expression in condensing cartilage.

To examine the spatial and temporal distribution of FARP1 in the chick spinal cord, we carried out a developmental time course of Farp1 mRNA expression from St 20 to St 35. FARP1 is expressed in forelimb spinal motor neurons from St 29 to at least St 35; however, low levels of FARP1 transcripts can be detected in motor neurons as early as St 27 (Figure 1B; Figure S1). At thoracic regions, FARP1 transcripts localize within a mediolateral population of cells that occupy positions characteristic of the preganglionic autonomic motor neurons of the Column of Terni (CT; Figure 1C; Jessell, 2000). To determine the relationship between FARP1 expression and different spinal motor columns, we raised polyclonal antibodies against FARP1 and examined the expression of FARP1 protein with respect to molecular markers that define the major motor neuron subtypes in the spinal cord (Tsuchida et al., 1994). The FARP1 antibody detected a single band corresponding to the predicted molecular weight of FARP1 in spinal cord extracts, and did not cross react with the related protein FARP2 (Figure S2). Motor neurons of the medial division of the Median Motor column (MMCm) are found at all axial levels, innervate axial muscles and coexpress the LIM homeodomain (HD) proteins Lim3 and Islet1/2. Lateral Motor Column (LMC) motor neurons form at limb levels, innervate target muscles in the limb and diversify into medial LMC neurons that express Islet1/2 and lateral LMC subtypes that coexpress Islet2 and Lim1. At thoracic levels, motor neurons of the lateral MMC and the CT coexpress Islet1/2 but lack Lim3 expression. FARP1 is highly expressed in medial LMC neurons and CT neurons, and is consistently detected at lower levels in lateral LMC cells (Figure 1D, E). No expression of FARP1 was found in MMCm or MMCl motor neurons (Figure 1D, E). FARP1 expression in LMC and CT motor neurons initiates after motor neurons have acquired their subtype identities and established their axonal projections, suggesting that FARP1 may regulate later aspects of motor neuron development. To gain deeper insight into the function of FARP1, we examined the subcellular distribution of FARP1 protein in spinal motor neurons. We focused our attention on LMC motor neurons as RA signaling pathways are required for LMC formation and maintenance, suggesting that FARP1 may mediate RA-dependent events critical for LMC function. FARP1 expression overlaps with the dendritic marker MAP2, but is not detected in motor axons (Figure 1F–H′; Figure S3; Tucker et al., 1988). Although the possibility that FARP1 is expressed at low levels in LMC axons cannot be excluded, these observations indicate that FARP1 is dramatically enriched in the dendrites of LMC motor neurons.

FARP1 increases LMC neurite length in vivo

The enrichment of FARP1 expression within LMC motor neuron dendrites suggests that FARP1 may regulate LMC dendritic development. As a first step to test this hypothesis, we developed a Cre-lox based binary system to trace the neurite morphology of individual motor neurons in intact embryonic chick spinal cords (Figure 2A). We coelectroporated two plasmids, pCAGGS-LSLeGFP and pMN-Cre, into chick spinal cords at St 11, and analyzed electroporated chick spinal cords at different timepoints in development. pCAGGS-LSLeGFP harbors a chick β-actin promoter upstream of a lox-STOP-lox cassette inserted in front of the eGFP open reading frame (ORF), thus rendering the expression of eGFP dependent on Cre recombinase activity. The pMN-Cre construct contains a 250bp element from the mouse HB9 promoter cloned upstream of a minimal CMV promoter driving the expression of Cre, which limits the expression of Cre to postmitotic motor neurons (Lee et al., 2004). Electroporation of a high ratio of pCAGGS-LSLeGFP to pMN-Cre plasmids into chick spinal cords limits the expression of Cre to small numbers of motor neurons, and enables the selective labeling of individual postmitotic motor neurons by Cre-mediated eGFP expression (Figure 2A–C). This strategy combined with the use of molecular markers that distinguish different motor columns allows us to trace the neurite morphology of individual motor neuron subtypes in the spinal cord (Figure 2B, C). Using this approach, we monitored the neurite development of forelimb Isl1/2+Lim3− LMC neurons between St 25 and St 35 in development. At these stages, LMC motor neurons lack complexity and exhibit simple multipolar morphology. At St 25, LMC neurites are relatively short; however, they grow extensively between St 25 and St 31 before reaching a plateau at St 33 (Figure S4). The expression of FARP1 in LMC dendrites spanning the period of robust LMC neurite growth suggests that FARP1 may control the growth of LMC motor neuron dendrites.

Figure 2. FARP1 regulates the growth of LMC neurites in vivo.

Figure 2

Open in a new tab

(A) Schematic depicting strategy to visualize motor neuron dendrites in chick spinal cords. (B, C) Confocal micrographs of ventral right quadrants of St 31 spinal cords. Hatched white lines outline the spinal cord. (D–G; I–L) Confocal micrographs of representative brachial LMC (D–G) and thoracic MMC (I–L) motor neurons electroporated with either the vector, FARP1, short hairpin control (shCtrl), or short hairpin FARP1 (shFarp1) expressing constructs. (D′–G′; I′–L′) Traces of electroporated motor neurons, neurites traced in red were quantified; arrows mark traced neurons (see supplemental methods for details). Individual motor neuron subtypes were identified by costaining with specific molecular markers. (H) Graph of dendrite length of brachial LMC motor neurons; compared to the vector, FARP1 p=0.001, shCtrl p=0.32, shFarp1 p=7×10−15. (M) Graph of dendrite length of thoracic MMC motor neurons; compared to the vector, FARP1 p=0.99, shCtrl p=0.39, shFarp1 p=0.45. All graphs, mean ± s.e.m., two-tailed Student’s t-test, n= at least 50 motor neurons from 6 electroporated embryos.

To test this hypothesis, we utilized our binary Cre-lox dependent tracing system to examine the consequences of FARP1 overexpression on LMC motor neuron neurite length and number. Since dendritic length and morphology are linked to motor neuron subtype identities, we first examined if overexpression of FARP1 altered motor neuron differentiation or specification (Vrieseling and Arber, 2006). Analysis of embryos electroporated with constructs overexpressing FARP1 showed no changes in motor neuron numbers or their subtype specification, although FARP1 was detected throughout the soma, dendrites and axons (Figures S5, S6). However, FARP1 expression in Isl1/2+ Lim3− LMC motor neurons caused an approximately 72% increase in neurite length compared with LMC neurons that had been electroporated with an empty control vector (Figure 2D, E, H). FARP1 overexpression did not alter the number of neurites extended by LMC motor neurons (Figure S6). In contrast, overexpression of FARP1 in MMC motor neurons did not significantly alter the length or number of MMC neurites, suggesting that the cellular machinery involved in mediating the FARP1-dependent increase in neurite length is limited to LMC motor neurons (Figure 2I, J, M; Figure S6). To examine if FARP1 is necessary for regulating the neurite length of LMC neurons, we generated constructs expressing short hairpin RNAs (shRNAs) targeted against the FARP1 ORF (U6-shFARP1). Transfection of U6-shFARP1 constructs into HEK293T cells showed that shFARP1 RNAs efficiently downregulated the expression of heterologous FARP1 expression by 91% compared with unrelated control shRNAs (Figure S7). The electroporation of U6-shFARP1 plasmids into the spinal cord did not compromise motor neuron differentiation or motor column formation (Figure S5). However, the expression of shFARP1 RNAs caused a 39% reduction in the length of LMC neurites compared with control shRNAs (Figure 2F–H). The number of LMC neurites that were formed remained unchanged (Figure S6). Consistent with the lack of FARP1 expression in MMC motor neurons, thoracic MMCm and MMCl neurons did not show any changes in neurite length or number when electroporated with shRNAs against FARP1 (Figure 2K–M; Figure S6).

Taken together, these experiments indicate that FARP1 is necessary and sufficient to regulate the neurite length of LMC motor neurons, but is not involved in the regulation of neurite numbers. These observations correlate with the lack of LMC dendritic complexity before FARP1 is expressed, and the extensive dendritic growth of LMC neurons at the time of FARP1 expression (Figure S4).

FARP1 regulates LMC dendritic length in vitro

Our in vivo analysis provides evidence that FARP1 increases LMC neurite growth; however, it does not distinguish whether FARP1 selectively controls dendritic or axonal growth as both neurite populations are labeled by eGFP expression. To define the specificity of FARP1 function, we dissociated chick spinal cords that had been electroporated with HB9-eGFP plasmids and plasmids that overexpressed FARP1 or shFARP1 RNAs. HB9-eGFP constructs contain a 3kb promoter element from the mouse HB9 promoter cloned upstream of the eGFP ORF, which targets the expression of eGFP to postmitotic motor neurons and thus facilitates their visualization (Figure 3A–D; Lee et al., 2004). Cultured motor neurons retained their molecular profiles of LIM HD protein expression, enabling electroporated LMC motor neurons to be distinguished by their lack of Lim3 expression and their coexpression of eGFP and Islet1/2 (Figure 3A–D). LMC motor neurons in culture developed a typical multipolar morphology in that they extend one long, thin axon lacking MAP2 expression, and several shorter and thicker dendrites that express MAP2 (Figure 3E–E″). Electroporated motor neurons expressed FARP1 throughout the soma, dendrites and axons (Figure S6). Overexpression of FARP1 caused a 28% increase in the length of LMC motor neuron dendrites but did not alter their numbers (Figure 3F, G, J, L). In contrast, increasing FARP1 levels did not appreciably change LMC motor axon length or number (Figure 3K, M). Ablation of FARP1 expression by electroporation of U6-shFARP1 plasmids showed a 23% decrease in LMC dendritic length but did not influence the number of LMC dendrites or the length and branching of LMC motor axons (Figure 3H–M). Taken together, these observations suggest that FARP1 regulates the growth of LMC motor neuron dendrites, but does not control LMC axonal length or LMC axonal and dendritic numbers.

Figure 3. FARP1 regulates dendritic growth of LMC motor neurons in vitro.

Figure 3

Open in a new tab

(A–E″) Confocal micrographs of cultured chick spinal neurons. White arrows in A–D mark the same neuron in panels of merged and split detection channels. Hatched box in E is magnified in panels E′ and E″. White asterisk marks the axon, white arrows mark dendrites. (F–I) Confocal micrographs of representative cultured LMC motor neurons electroporated with either vector, FARP1, shCtrl, or shFarp1. Red arrowheads mark the axons. (J) Graphs of dendrite length; compared to the vector, FARP1 p=0.0007, shCtrl p=0.28, shFarp1 p=6×10−7. (K) Graphs of axon length; compared to the vector, FARP1 p=0.48, shCtrl p=0.69, shFarp1 p=0.81. (L) Graphs of dendrite number; compared to the vector, FARP1 p=0.93, shCtrl p=0.85, shFarp1 p=0.76. (M) Graphs of axon branches, compared to the vector, FARP1 p=0.76, shCtrl p=0.97, shFarp1 p=0.29. All graphs, mean ± s.e.m; two-tailed Student’s t-test =dendrite and axon length, chi-square test =dendrite number and axon branch.

FARP1 interacts with PlexA4 independently of Neuropilins

FARP2, a protein closely related to FARP1, binds to Class A plexin receptors and triggers the repulsion of DRG axons in response to Sema3A binding (Toyofuku et al., 2005). To examine if FARP1 regulates the growth of LMC dendrites as a function of semaphorin-plexin signaling pathways, we first examined the distribution of chick Class A plexins and their coreceptors, the Neuropilins (Npn), in developing LMC neurons. Literature and database searches indicate that Npn1, Npn2, PlexA1, PlexA2 and PlexA4 are encoded by the chick genome (Mauti et al., 2006). In situ hybridization analyses of chick spinal cords show that at forelimb regions Npn1, Npn2, PlexA1 and PlexA4 overlap with FARP1 expression in LMC neurons, while PlexA2 is not expressed in spinal motor neurons (Figure 4A–F). Since Npn coreceptors interact directly with only a very limited number of downstream signaling components, we examined if FARP1 is capable of interacting with PlexA1 and PlexA4 by co-immunoprecipitation (IP), using tagged versions of these proteins in transfected HEK293T cells (Tran et al., 2007). FARP1 interacts strongly with PlexA4, but associates very weakly with PlexA1 (Figure 4G). Co-IP assays using extracts from HEK293T cells transfected with constructs expressing PlexA4 and the individual FERM, DH and PH domains of FARP1 show that the FERM and PH domains of FARP1 interact with PlexA4 (Figure S8). In contrast to FARP2, FARP1 does not require Npn1 or Npn2 to interact with PlexA4 or PlexA1, and the presence of Npn1 or Npn2 does not significantly change the association between FARP1 and PlexA4 or PlexA1 (Figure 4G). Taken together, these results indicate that although Npn1, Npn2, PlexA1 and PlexA4 are expressed in LMC neurons, FARP1 preferentially interacts with PlexA4 and this interaction is independent of Npn binding. This finding suggests a model where FARP1 mediates LMC dendritic growth through PlexA4 directed signaling.

Figure 4. FARP1 preferentially interacts with PlexA4 independently of Npns.

Figure 4

Open in a new tab

(A–F) In situ hybridization on transverse sections of St 31 right hemisegments of chick spinal cords. Hatched black lines outline LMC motor neurons. (G) Western blot analysis of co-IP experiments using HEK293T cells.

PlexA4 increases LMC dendritic length through FARP1 function

To investigate if PlexA4 is required for the dendritic growth of LMC motor neurons, we examined the consequences of increasing PlexA4 levels in motor neurons using our assay to visualize motor neuron dendrites in the chick spinal cord. Overexpression of PlexA4 increased the length of LMC motor neuron dendrites by approximately 31% but did not affect LMC dendritic numbers, a phenotype similar to when FARP1 levels were elevated (Figure 5A, C, F, G). In contrast, overexpression of PlexA1 did not alter the length or number of LMC motor neuron dendrites (Figure 5A, B, F, G). The specificity of PlexA4 function correlates with the ability of PlexA4 to interact with FARP1, and further supports the model that FARP1 mediates PlexA4 receptor signaling to regulate LMC dendritic growth. To determine if PlexA4 is necessary for LMC dendritic growth, we generated constructs expressing shRNAs designed against the PlexA4 ORF (U6-shPlexA4). In transfected HEK293T cells, expression of shPlexA4 RNAs downregulated heterologous PlexA4 expression by 84% compared with control unrelated shRNAs (Figure S7). Electroporation of U6-shPlexA4 into embryonic chick spinal cords caused a 37% reduction in the length of LMC motor neuron dendrites but did not change their numbers, demonstrating a requirement for PlexA4 in promoting LMC dendritic extension (Figure 5F–H). As an alternative approach to disrupt endogenous PlexA4 signaling, we expressed a dominant-negative version of PlexA4 (A4ΔCyto) that lacks the intracellular cytoplasmic domain, in embryonic spinal cords (Takahashi and Strittmatter, 2001). Electroporation of plasmids expressing A4ΔCyto caused a similar decrease in LMC dendritic length obtained by ablating PlexA4 using shPlexA4 RNAs (Figure 5D, F). However, overexpression of A4ΔCyto caused a minor decrease in the numbers of LMC dendrites (Figure 5G). A4ΔCyto does not interact with FARP1, and fails to bind to other known transducers of PlexA signals (data not shown); thus, the effects of A4ΔCyto on dendritic numbers may reflect the blockade of other PlexA or Npn pathways that are FARP1-independent. Deletion of the extracellular domain of PlexA4 renders PlexA4 ligand-insensitive and causes it to become constitutively active (A4ΔEcto; Takahashi and Strittmatter, 2001). Surprisingly, electroporation of A4ΔEcto into chick spinal cords did not change LMC dendritic length or number (Figure 5E–G). This result suggests that PlexA4 dependent regulation of LMC motor neuron dendrite length requires ligand activation, and implies that the mechanism of PlexA4 function in dendritic growth may differ from its roles in axonal repulsion.

Figure 5. PlexA4 regulates dendritic growth of LMC motor neurons through FARP1.

Figure 5

Open in a new tab

(A–E, H–K) Micrographs of representative electroporated brachial LMC motor neurons. (A′–E′; H′-K′) Traces of electroporated motor neurons, neurites traced in red were quantified; arrows mark traced neurons. (F) Graph depicting dendrite length of brachial LMC motor neurons; compared to the vector; PlexA1 p=0.75, PlexA4 p=4×10−5, shPlexA4 p=9×10−12, A4ΔCyto p=5×10−6, A4ΔEcto p=0.06.(G) Graph of dendrite number of brachial LMC motor neurons; compared to the vector; PlexA1 p=0.59, PlexA4 p=0.11, shPlexA4 p=0.53, A4ΔCyto p=0.01, A4ΔEcto p=0.52. (L, M) Graphs showing dendrite length of brachial LMC motor neurons; non-significant (N.S.) between shPlexA4 and shPlexA4+ FARP1, p= 0.64; N.S. between shFarp1 and shFARP1+PlexA4, p=0.27. All graphs, mean ± s.e.m., two-tailed Student’s t-test= dendrite length, chi-square test= dendrite number.

Taken together, these results show that PlexA4 is necessary and sufficient to regulate the growth of LMC motor neuron dendrites. To test if FARP1 mediates this process, we overexpressed PlexA4 in the context of ablating FARP1 expression in the chick spinal cord using shFARP1 RNAs. Overexpression of PlexA4 failed to increase the length of LMC motor neuron dendrites when FARP1 expression is reduced (Figure 5J, K, M). These results provide strong evidence that FARP1 functions as a downstream effector of PlexA4 signaling to mediate the growth of LMC motor neuron dendrites. Interestingly, FARP1 overexpression did not increase LMC dendritic length when PlexA4 levels were reduced (Figure 5H, I, L). This result suggests that FARP1 requires PlexA4 to stimulate dendritic growth, and implies that association between FARP1 and PlexA4 is necessary to promote the dendritic growth of LMC motor neurons.

Sema6A signals regulate LMC dendritic length

Known ligands that activate PlexA4 signaling pathways include the secreted Class 3 semaphorins (Sema3) and the transmembrane Class 6 semaphorins (Sema6; Tran et al., 2007). Npn-PlexA4 holoreceptor complexes bind Sema3 proteins, while PlexA4 associates directly with Sema6 ligands independently of Npns (Suto et al., 2005; Yaron et al., 2005). Sema3A is implicated in increasing the dendritic length of cortical neurons and is expressed in LMC motor neurons, making it a good candidate for the signal that triggers the PlexA4-FARP1-dependent growth of LMC dendrites (Shepherd et al., 1996; Polleux et al., 2000; Fenstermaker et al., 2004). To determine if Sema3A can increase the length of LMC motor neuron dendrites, we generated secreted alkaline phosphatase fusions of chick Sema3A (Sema3A-AP) and applied it to dissociated spinal cords that had been electroporated with HB9-eGFP plasmids. Addition of Sema3A-AP fusion proteins effectively decreased eGFP+ LMC motor axon length by 25%, consistent with previous work demonstrating functions for Sema3A in axon repulsion (Figure S9; Shepherd et al., 1996; Huber et al., 2005). However, the application of Sema3A-AP did not alter the length or number of eGFP+ LMC motor neuron dendrites (Figure S9); thus ruling out Sema3A as the signal that triggers the PlexA4-FARP1 dependent growth of LMC dendrites. This observation is consistent with our finding that Npns, which are required for PlexA receptors to associate with Sema3 ligands, do not influence the interaction between PlexA4 and FARP1 (Figure 4G).

To examine if the transmembrane Sema6 proteins are relevant ligands that signal through PlexA4 and FARP1 to regulate the growth of LMC motor neuron dendrites, we first determined the number and identity of Sema6 proteins in the chick. In mammals, Sema6A and Sema6B are ligands for PlexA4, while Sema6C and Sema6D constitute ligands for PlexA1 (Suto et al., 2005; Yoshida et al., 2006). Only full-length Sema6A and Sema6D ORFs are annotated in chick genomic databases; Sema6B sequences are incomplete while Sema6C is absent. We cloned full-length chick Sema6A and Sema6D ORFs by RT-PCR and examined their mRNA distribution in the developing spinal cord. At forelimb levels, both Sema6A and Sema6D transcripts are distributed throughout the spinal cord and overlap with FARP1 and PlexA4 expression in LMC neurons (Figure 6A, B). To determine if chick Sema6A and Sema6D exhibit the same receptor specificity as their mammalian counterparts, we performed receptor-ligand binding experiments in transfected COS-7 cells. Addition of soluble Sema6 ligands fused to human IgG Fc segments (Sema6A-ectoFc; Sema6D-ectoFc) to COS-7 cells transfected with constructs expressing either chick PlexA4 or PlexA1, showed that Sema6A bound to PlexA4, while Sema6D interacted specifically with PlexA1 (Figure 6C–H). The specific binding of Sema6A to PlexA4 suggests the possibility that Sema6A is the relevant ligand for controlling the PlexA4-FARP1 dependent growth of LMC motor neuron dendrites.

Figure 6. Sema6A signals through PlexA4 and FARP1 to regulate dendritic growth of LMC motor neurons.

Figure 6

Open in a new tab

(A, B) In situ hybridization on transverse sections of right hemisegments of St 31 chick spinal cords. (C–H) Ligand-receptor binding experiments in COS−7 cells. (I, J) Micrographs of representative electroporated brachial LMC motor neurons. Boxed insets show traces of electroporated neurons marked by arrows. (K, L) Graphs of dendrite length (p=6×10−12) and number (p=0.44) of brachial LMC motor neurons. (M–O, R–U) Confocal micrographs of representative cultured LMC motor neurons treated with ligand combined with electroporation of shFarp1 or shPlexA4. Red arrowheads mark the axons. (P, V, W) Graphs of dendrite length of cultured LMC motor neurons; compared to the Fc, Sema6A p=0.004, Sema6D p=0.08, and shFarp1 N.S. p=0.15, shPlexA4 N.S. p=0.46. (Q) Graph of dendrite number of cultured LMC motor neurons; compared with Fc, Sema6A p=0.19, Sema6D p=0.34. For all graphs, mean ± s.e.m., two-tailed Student’s t-test = dendrite number, chi-square test = dendrite length.

To test the requirement for Sema6A in LMC dendritic growth, we used our in vivo dendritic tracing assay to determine if ablating Sema6A expression in the spinal cord affects the length of LMC dendrites. ShRNAs against the Sema6A ORF (shSema6A) proved effective in decreasing heterologous Sema6A expression by approximately 87% in transfected HEK293T cells (Figure S7). Electroporation of U6 plasmids that drive the expression of shSema6A RNAs in cells throughout the dorsal and ventral spinal cord caused a 39% decrease in the length of eGFP+ Isl1/2+Lim3− LMC motor neuron dendrites compared to unrelated shRNAs, while the number of LMC dendrites remained unchanged (Figure 6I–L). These results suggest that Sema6A is required to regulate the length of LMC motor neuron dendrites in vivo. To determine if Sema6A application can increase LMC dendritic length, we prepared dissociated cultures from dissected spinal cords electroporated with HB9-eGFP plasmids. Addition of Sema6A-ectoFc to cultured eGFP+ motor neurons increased the length of LMC motor neuron dendrites by approximately 33%, but did not change the number of extended dendrites (Figure 6M, N, P, Q). In contrast, addition of Fc fragments alone or Sema6D-ectoFc, the ligand for PlexA1, had no effects on LMC motor neuron dendrite length or number, consistent with our earlier observations that FARP1 binds very weakly to PlexA1, and PlexA1 does not influence LMC dendritic growth (Figure 6M, O, P, Q). These observations indicate that Sema6A is necessary and sufficient to regulate the growth of LMC motor neuron dendrites.

Sema6A signals through PlexA4 and FARP1 to regulate the length of LMC dendrites

To determine if Sema6A regulates LMC dendritic growth through PlexA4, we assayed if the ability of Sema6A to increase dendritic growth was affected by the loss of PlexA4 expression. We prepared dissociated neuronal cultures from embryonic chick spinal cords that had been coelectroporated with HB9-eGFP and shPlexA4 expressing plasmids, and incubated them with Sema6A-ectoFc, or Fc alone. Knockdown of PlexA4 blocked the increase in LMC dendritic length normally elicited by the application of Sema6A-ectoFc (Figure 6R, S, W). This observation suggests that the ability of Sema6A to regulate the growth of LMC motor neuron dendrites depends upon PlexA4, and supports the model that the Sema6A-PlexA4 receptor pair is the signaling complex involved in regulating LMC dendritic growth. To determine if Sema6A-PlexA4 signaling requires FARP1, we repeated the experiment, this time knocking down FARP1 instead of PlexA4 expression by electroporating shFARP1 RNA expressing plasmids into the spinal cord. Knockdown of FARP1 abrogated the ability of Sema6A-ectoFc to increase LMC dendritic length (Figure 6T, U, V). Taken together, these results suggest that FARP1 acts downstream of Sema6A binding to PlexA4 to regulate the growth of LMC motor neuron dendrites.

MMC motor neurons lack FARP1 and PlexA4 expression, consistent with our observations that ectopic expression of PlexA4 or FARP1 alone does not increase the length of MMC dendrites (Figure 7A, B). However, MMC motor neurons and adjacent areas express high levels of Sema6A (Figure 7C). To test the sufficiency of the PlexA4-FARP1 complex in promoting motor neuron dendritic growth, we overexpressed PlexA4 and FARP1 in MMC motor neurons, and utilized our in vivo dendritic tracing assay to examine the consequences on MMC dendritic length and number. Electroporation of vector, FARP1 or PlexA4 alone failed to promote MMC dendritic growth; however, coexpression of FARP1 and PlexA4 caused a 47% increase in the length of MMC dendrites (Figure 7D–F). In all cases, the number of MMC motor neuron dendrites was not changed (Figure 7G). Taken together, these observations provide evidence that PlexA4 and FARP1 are sufficient to regulate motor neuron dendritic growth in the presence of Sema6A. In addition, they imply that the effectors downstream of FARP1 involved in triggering dendritic growth are shared by LMC and MMC motor neurons.

Figure 7. The FARP1-PlexA4 complex is sufficient to promote the dendritic growth of MMC motor neurons.

Figure 7

Open in a new tab

(A–C) In situ hybridization on transverse sections of right hemisegments of St 31 chick spinal cords. Hatched black lines outline MMC motor neurons. (D, E) Confocal micrographs of representative electroporated thoracic MMC motor neurons. (F) Graphs of dendrite length of thoracic MMC motor neurons; compared to the vector, FARP1 p=0.99, PlexA4 p=0.76, FARP1+PlexA4 p=9×10−5. (G) Graphs of dendrite number of thoracic MMC motor neurons; compared to the vector, FARP1 p=0.29, PlexA4 p=0.16, FARP1+PlexA4 p=0.79. All graphs, mean ± s.e.m., two-tailed Student’s t-test = dendrite number, chi-square test = dendrite length.

FARP1 is constitutively associated with PlexA4 and requires the DH domain

To investigate the mechanism by which FARP1 mediates Sema6A-PlexA4 signaling, we first examined if the interaction between FARP1 and PlexA4 is altered in the presence of Sema6A ligand. We performed co-IP assays using extracts of HEK293T cells cotransfected with PlexA4 and FARP1 in the presence or absence of Sema6A-ectoFc. Addition of Sema6A-ectoFc did not disrupt or enhance PlexA4-FARP1 complex formation, suggesting that the interaction between PlexA4 and FARP1 is maintained in the presence of ligand (Figure 8A). This result is consistent with our earlier observation that FARP1 requires PlexA4 to promote LMC dendritic growth (Figure 5H, I, L).

Figure 8. FARP1 function requires the DH domain.

Figure 8

Open in a new tab

(A) Western blot of co-IP experiments using HEK293T cells. (B, C) Confocal micrographs of representative LMC motor neurons electroporated with FARP1ΔDH cultured with Sema6A. Red arrowheads mark the axons. (D) Graph of dendrite length of cultured LMC motor neurons, N.S. p=0.62 (E–G) Confocal micrographs of representative electroporated brachial LMC motor neurons. (H) Graph of dendrite length of brachial LMC motor neurons; compared to the vector, FARP1ΔDH p=3×10−19, FARP1ΔDH+PlexA4 p=5×10−17, and N.S. p=0.41, all graphs, mean ± s.e.m., two-tailed Student’s t-test. (I) Model of FARP1 effector function. RA= retinoic acid. Although Sema6A and PlexA4 are shown acting in trans, it is possible that they function in cis.

Previous studies have shown that Rho-GEF activity is involved in modulating cytoskeletal dynamics, raising the possibility that the DH Rho-GEF domain of FARP1 is required for FARP1 to increase LMC dendritic length (Rossman et al., 2005). To test this possibility, we overexpressed versions of FARP1 that lacked the DH Rho-GEF region (FARP1ΔDH) in developing chick spinal cords. Overexpression of FARP1ΔDH failed to increase LMC dendritic length and instead decreased the extent of LMC dendrites by 44% (Figure 8, E, F, H). These results suggest that the DH Rho-GEF domain is critical for FARP1 activity, and that the FARP1ΔDH molecule acts as a dominant negative molecule to interfere with endogenous FARP1 function. To examine if the FARP1 DH domain is necessary for its effector function in transducing Sema6A-PlexA4 signals to increase LMC dendritic growth, we coelectroporated constructs expressing FARP1ΔDH and PlexA4 into developing chick spinal cords. Expression of FARP1ΔDH abrogated the ability of PlexA4 to increase LMC dendritic length in vivo, suggesting that the FARP1 Rho-GEF domain is critical for mediating PlexA4-dependent dendritic growth (Figure 8G, H). To investigate the consequences of FARP1ΔDH on Sema6A-induced increases in LMC dendritic length, we prepared dissociated neuronal cultures from chick spinal cords that had been coelectroporated with FARP1ΔDH and HB9-eGFP expression plasmids, and incubated them with Sema6A-ectoFc ligands. FARP1ΔDH expression abrogated the ability of Sema6A to increase the length of LMC dendrites (Figure 8B–D). Taken together, these results imply that FARP1 Rho-GEF activity is critical for mediating the Sema6A-PlexA4-dependent growth of LMC motor neuron dendrites.

Discussion

The development of dendritic arbors depends upon a series of carefully regulated events that control overall dendritic growth, targeted extension and branching (Jan and Jan, 2003). Here we identify FARP1 as a key regulator of the growth of LMC motor neuron dendrites during the peak period of motor neuron dendritic extension in embryonic development. Our data are consistent with a model where FARP1 binds to the PlexA4 receptor in LMC motor neuron dendrites and transduces transmembrane Sema6A signals to stimulate dendritic extension (Figure 8I). The Rho-GEF domain of FARP1 is an essential component in the mechanism of FARP1 function, suggesting that FARP1 increases dendritic length by modulating the activity of small GTPase proteins. We discuss below how the properties of FARP1 confer its specificity in transducing semaphorin-plexin signals within dendrites, and the implications for how these programs are utilized to regulate the morphology of motor neuron dendrites in the developing spinal cord.

FARP1 binding properties dictate receptor specificity

The semaphorin-plexin signaling pathway has diverse biological functions. In the developing nervous system, their functions include axon outgrowth, axon guidance, cell migration and dendritic growth, guidance and branching (Tran et al., 2007; Mann et al., 2007). The diversity of semaphorin-plexin function is elicited by the multitude of different semaphorin receptor complexes, the presence of transmembrane and secreted semaphorin ligands, and the triggering of distinct intracellular signaling components. In the developing spinal cord, LMC motor neurons contain multiple components of the semaphorin-plexin signaling pathway. Both medial and lateral divisions of the LMC express Npn1, Npn2, PlexA1, and PlexA4 receptors and transmembrane Sema6 ligands, while secreted Sema3 ligands are expressed in LMC neurons and tissues surrounding LMC axons (Shepherd et al., 1996; Huber et al., 2005; Cohen et al., 2005; Moret et al., 2007). Thus, LMC neurons contain the components to generate secreted and transmembrane ligand receptor complexes for both PlexA1 and PlexA4 (Tran et al., 2007; Mann et al., 2007).

We show here that FARP1 is necessary and sufficient to promote LMC dendritic growth and show that it is a specific effector of Sema6A/PlexA4 signaling pathways in the developing vertebrate spinal cord. Gain of function experiments where PlexA1 is expressed throughout motor neurons, and global exposure of motor neurons to Sema3A and Sema 6D, provide evidence that transmembrane and secreted semaphorin-PlexA1 signaling complexes do not play roles in the regulation of LMC dendritic growth. In contrast, overexpression of PlexA4 or application of Sema6A promotes the dendritic growth of LMC motor neurons, while knockdown of these components effectively decreases dendrite extension. We find that FARP1 associates strongly with PlexA4 relative to PlexA1, suggesting that the specificity of plexin receptor function in this context is imposed by the different binding capabilities of FARP1. The related molecule FARP2 is also an effector of semaphorin-plexin signaling and regulates the collapse of DRG axons in response to Sema3A signals (Toyofuku et al., 2005). However, DRG neurons express PlexA1, A2, A3 and A4, and FARP2 binds indiscriminately to all Class A plexin receptors, raising the intriguing question as to how and if specificity in this system is achieved (Toyofuku et al., 2005; Yoshida et al., 2006). FARP1 and FARP2 share a similar domain structure but employ different mechanisms to transduce specific semaphorin-plexin signals. One conceivable outcome of these differences is that they trigger distinct downstream cascades that ultimately have important consequences for remodeling the cytoskeleton of axons and dendrites.

Asymmetric compartmentalization of FARP1 and other signaling components confers the dendritic specificity of the Sema6A/PlexA4 pathway

The specific enrichment of FARP1 within motor neuron dendrites contributes to restricting the effector function of FARP1 to this cellular compartment, and is consequently a critical factor in imposing the specificity of FARP1 function. Interestingly, the overexpression of FARP1 in LMC motor neurons causes a specific increase in dendritic length without altering axonal growth or branching even though FARP1 in this context is distributed throughout the soma and axon (Figure S6). These collective observations suggest that one or more upstream or downstream components of the Sema6A/PlexA4 signaling pathway itself must be restricted to the dendrites of LMC motor neurons. Global application of Sema6A elicits the repulsion of DRG and sympathetic ganglia neurons but does not elicit a response in motor neurons (Mauti et al., 2007). Furthermore, knockdown of PlexA4 expression in LMC motor neurons causes a 27% increase in axonal length and Sema6A ectodomain binding studies detects Sema6A receptors in motor axons, consistent with a role for PlexA4 in regulating motor axon growth (B.Z. and S.S., unpublished observations; Mauti et al., 2007). These observations together suggest that components downstream of FARP1 are likely to be localized asymmetrically to the dendrites of LMC motor neurons.

What is the ligand specificity that triggers the ability of FARP1 to promote dendritic extension? Genetic and in vitro evidence indicate that Sema3A constitutes a ligand for PlexA4 in the context of PlexA4/Npn1 holoreceptors (Yaron et al., 2005; Suto et al., 2005). Since LMC neurons express Sema3A, Npn1 and PlexA4, and FARP1 can bind PlexA4 in the presence of Npn1, why are the effects of FARP1 on dendritic growth specifically elicited by Sema6A and not Sema3A? One possibility is the availability of Sema3A ligand in dendrites; however, bath application of Sema3A does not affect motor neuron dendritic length and LMC neurons express high levels of Sema3A (Moret et al., 2007). Another potential explanation is that PlexA4/FARP1 complexes in LMC dendrites are assembled in a form more amenable to transducing Sema6A signals than Sema3A signals, as compared with LMC motor axons. Although Npn1 does not alter the association of FARP1 with PlexA4, Npn1 might directly influence FARP1 activity in the context of Npn1/PlexA4 complexes to prevent it from triggering LMC dendritic growth. Alternatively, Npn1 may be preferentially expressed in LMC axons but excluded from LMC dendrites. Indeed, analyses of Npn1−/− mice reveal deficits in LMC axonal growth and guidance; however, whether Npn1 function in LMC neurons is restricted solely to axons awaits further investigation (Huber et al., 2005).

The mechanism of FARP1 function

Our structure-function analysis of FARP1 reveals that the integrity of its Rho-GEF domain is critical for its ability to promote the growth of LMC motor dendrites. In vitro studies show that FARP1 functions as a guanine nucleotide exchange factor (GEF) for RhoA by catalyzing the exchange of GDP for GTP, consistent with a function for FARP1 in promoting the activity of downstream small GTPases (Koyano et al., 1997). Components of Rho GTPase signaling pathways are established modulators of actin filaments and microtubules, which make up the primary structural components of dendrites. Thus, it is plausible that FARP1 Rho-GEF function ultimately regulates cytoskeletal dynamics, leading to the growth of LMC motor dendrites (Etienne-Manneville and Hall, 2002). Interestingly, activation of Rho GTPases by other Semaphorin/plexin signaling pairs is normally associated with growth cone collapse rather than extension (Tran et al., 2007). However, these observations apply to axons and not dendrites, raising the possibility that the modulatory effects of activated Rho on the cytoskeleton downstream of semaphorin/plexin signals are highly context-dependent. In support of this, we find that the mechanism of PlexA4 function in dendrites is likely different to its roles in axonal repulsion as deletion of the extracellular domain has no effect on dendritic growth, whereas it renders PlexA4 constitutively active in promoting growth cone collapse. In cortical neurons, Sema3A exerts repulsive effects on axons but promotes the apical growth of dendrites through mechanisms involving the asymmetric localization of guanylate cyclase (Polleux et al., 2000). Whether cyclic nucleotides are similarly involved in regulating LMC dendritic growth through FARP1 mediation of semaphorin/plexin signals awaits further investigation.

How does Sema6A/PlexA4 signaling lead to FARP1 effector function? In contrast to FARP2, our data show that FARP1 requires the presence of PlexA4 to increase dendritic length and that FARP1 remains associated with PlexA4 in the presence of Sema6A, indicating that dissociation of the FARP1/PlexA4 complex is not a prerequisite for triggering LMC dendritic growth. Analyses of Rho-GEF proteins show that highly variable sequences N-terminal to the DH domain function as intramolecular negative regulators of Rho-GEF activity (Rossman et al., 2005). We thus speculate that FARP1 Rho-GEF activity is not constitutively active, but is positively regulated by Sema6A-PlexA4 signaling. Precedent for this mechanism comes from studies of PlexB1, where Sema4D binding to PlexB1 does not change this plexin’s ability to complex with the Rho-GEF protein PDZ-RhoGEF/LARG, but stimulates LARG activity to activate RhoA (Swiercz et al., 2002). One potential advantage in maintaining the association between plexin receptors and Rho-GEF proteins is that this constitutive association acts as a molecular tether to generate a functional microdomain at the cell membrane that may be critical for effective and rapid modulation of the cytoskeleton.

Regulating dendritic growth and morphology in motor neurons

The dendritic morphology of a particular neuron dictates its ability to receive, process and integrate synaptic inputs (Jan and Jan, 2003). Our data are consistent with a model in which FARP1 mediates Sema6A/PlexA4 signaling to control the early dendritic growth of LMC motor neurons (Figure 8I). At the developmental times we analyzed, Sema6A is expressed throughout the ventral-lateral spinal cord at relatively uniform levels. It remains unclear whether the function of Sema6A is cell- or non cell-autonomous; nonetheless, the widespread expression of Sema6A suggests that FARP1 mediates general, rather than targeted dendritic growth during embryogenesis. Thus, our model suggests that the FARP1 effector function contributes to the formation of early dendritic arbors on which later programs of targeted growth and pruning act to sculpt final patterns of dendritic morphology. Notably, the columnar specific expression of FARP1 implies that there are dedicated programs controlling dendritic growth of specific motor columns, and suggests that these early programs controlling dendrite extension may directly influence the final dendritic patterns of specific motor neuron subtypes. We show here that upstream signaling pathways mediating the induction of FARP1 include RA, and this correlates well with the sustained expression of the RA synthetic enzyme RALDH2 in LMC motor neurons, and the established roles for RA in promoting neurite outgrowth in a variety of cellular paradigms (Sockanathan and Jessell, 1998; Maden, 2007). However, FARP1 is also expressed in CT neurons, a population of visceral motor neurons that lack RALDH2 expression and are located in the thoracic cord, which expresses RA catabolic enzymes such as Cyp26 (Abu-abed et al., 2002). These observations suggest that other signaling pathways in addition to RA induce the CT-specific expression of FARP1. Interestingly, overexpression of FARP1 does not increase the dendritic length of CT neurons although PlexA4 is expressed in these cells. This finding suggests other functions for FARP1 in the development of CT neurons (B.Z and S. S., unpublished observations).

The cell bodies of motor neurons resident within motor columns cluster into distinct pools that reflect their axonal projections to discrete target muscles (Jessell, 2000). Pool-specific motor neurons exhibit distinctive dendritic arbors that are important for their ability to receive and process information (Landmesser, 1978; Okado et al., 1990; Vrieseling and Arber, 2006). Our discovery that Sema6A-PlexA4 signaling regulates columnar-specific dendritic growth through FARP1 raises the possibility that semaphorin-plexin pathways may also contribute to refining the distinct dendritic morphologies of motor pools. This possibility is particularly compelling given that the expression of RALDH2 and different classes of secreted semaphorins, Npns and plexins overlaps in distinct motor pools (Sockanathan and Jessell, 1998; Cohen et al., 2005). It will be of great interest to define their function in motor dendrite formation and determine if semaphorin-plexin pathways intersect with RA-responsive effector molecules to shape the pool-specific dendritic morphology of motor neurons.

Experimental procedures

RT-PCR, plasmids and reverse northerns

Coding sequences of all genes were isolated using RT-PCR from total RNA extracted from stage 31 chick spinal cords with Trizol (Invitrogen). Reverse Northern blot analysis was performed as described previously (Rao and Sockanathan, 2005), with the modification that the FARP1 ORF was first amplified by PCR from FARP1 plasmids prior to electrophoresis and blotting. Short hairpin (sh) RNAs were cloned into the pSilencer1.0-U6 vector (Ambion). ShRNA target sequences were as follows: AACCACCTGAACCTTGTTGAA and AAGTAGCGACCACAGAACGAA for Farp1, AATGCTCCCCATAGACTACAA and AAGATCCGTGTGGATGGCACC for PlexA4, AAGCTGATGTAGACACATGCA and AAGTGCAGTTACGATGGCATG for Sema6A, and AATTCGCGCCTAGGTCCGAAC and AACTAGGTCGTTCGACGTAAG for unrelated control short hairpins.

Co-IP assay

HEK293T cells (ATCC) were transfected and lysed after 24 hours using standard protocols. Protein lysates were incubated with 4μg antibody and GammaBind Sepharose beads (GE Healthcare) overnight at 4°C. The beads were washed with lysis buffer, boiled in sample buffer and used for Western blot analysis. Antibodies used were mouse anti-HA (Roche; HRP-conjugated, Santa Cruz), mouse anti-myc 9E10 (Developmental Studies Hybridoma Bank; HRP-conjugated, Abcam), and mouse anti-FLAG (Sigma).

Ligand production and ligand receptor binding

The sequences corresponding to the ectodomains of chick Sema6A and Sema6D were cloned into the pFcTag vector. Plasmids expressing alkaline phosphatase (AP) or the AP fusions of chick Sema3A (Sema3A-AP) were gifts from Dr. Jonathan Raper. HEK293T cells transfected with these plasmids were cultured in OptiMEM-I medium (Invitrogen), and conditioned medium containing secreted ligands were collected and concentrated 40-fold with Centricon filters. COS-7 cells (ATCC) transfected with myc-PlexA1 or myc-PlexA4 were cultured for 2 days. 10μg of prepared ligands were pre-incubated with 3μg alkaline phosphatase-conjugated sheep anti human antibody (JacksonImmuno Research) in 500μl OptiMEM-I at room temperature for 30 minutes, and the ligand/antibody mixture was applied to COS-7 cells for 2 hours at room temperature. The cells were rinsed with 1×PBS, fixed with 4% PFA, and developed for AP activity with BM-purple AP substrate (Roche Diagnostics) supplemented with 1mM levamisole (Vector Laboratories, Inc.).

In ovo electroporation and tissue preparation

Stage 13–15 chick embryos were electroporated as described (Nakamura and Funahashi, 2001), incubated and then collected, eviscerated and fixed in 4% PFA for 2 hours at 4°C. Prepared embryos were embedded with TissueTek (Sakura Finetek USA, Inc.) and sectioned.

In situ hybridization and immunohistochemistry

In situ hybridization was performed as described (Schaeren-Wiemers and Gerfin-Moser, 1993). Brightfield images were acquired with a Zeiss Axioskop2 microscope. Immunohistochemical staining was performed as described (Ji et al., 2006). Fluorescent 29 images were acquired with a Zeiss LSM Pascal microscope. Antibodies used were as described in Ji et al (2006) with the exception of guinea pig anti-FARP1 (1:5000), rabbit anti-GFP (1:1000, Invitrogen), mouse anti-NF (1:50, 3A10, DHSB), mouse anti-MAP2 (1:1000, Sigma), and mouse anti-Tau1 (1:2000, Chemicon Mab 3420). Secondary antibodies (JacksonImmuno Research) were used according to the supplier’s protocol. The FARP1 antibody is an affinity-purified guinea pig polyclonal antibody raised against a 14 amino acid C-terminal peptide (Affinity BioReagents).

Primary neuronal cultures and motor neuron analysis

Primary neuronal cultures from spinal cords were performed as described previously (Kuhn, 2003). Spinal neurons were cultured in growth medium on PDL/laminin-coated glass chamber slides (Nalge Nunc International) for 2 days at 37°C. Cultured neurons were rinsed with 1×PBS, fixed in 4% PFA for 10 mins at room temperature, washed with 1×PBS, and used for immunohistochemical analysis. Dendrites and axons of motor neurons were traced and measured with NeuronJ (copyright by Erik Meijering) plug-in of ImageJ (NIH) [see supplemental methods for details]. At least 50 motor neurons from at least 6 electroporated spinal cords were analyzed in each group. Motor neuron counts were confined to the brachial regions corresponding to the forelimbs of chick embryos.

Supplementary Material

01

Acknowledgments

We thank Thomas Jessell, Alex Kolodkin, Sam Pfaff and Jonathan Raper for reagents; ChangHee Lee for help with software; Zachary Bitzer and Shaneka Lawson for technical assistance; Alex Kolodkin, ChangHee Lee, Marianeli Rodriguez and Fengquan 30 Zhou for critical comments on the manuscript, and members of the Sockanathan lab for discussions. This work was funded by grants from the NINDS (NIH), the Packard center for ALS Research and the Muscular Dystrophy Association.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errorsmaybe discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Abu-Abed S, MacLean G, Fraulob V, Chambon P, Petkovich M, Dollé P. Differential expression of the retinoic acid-metabolizing enzymes CYP26A1 and CYP26B1 during murine organogenesis. Mech Dev. 2002;110:173–177. doi: 10.1016/s0925-4773(01)00572-x. [DOI] [PubMed] [Google Scholar]
  2. Aranda A, Pascual A. Nuclear hormone receptors and gene expression. Physiol Rev. 2001;81:1269–1304. doi: 10.1152/physrev.2001.81.3.1269. [DOI] [PubMed] [Google Scholar]
  3. Cohen S, Funkelstein L, Livet J, Rougon G, Henderson CE, Castellani V, Mann F. A semaphorin code defines subpopulations of spinal motor neurons during mouse development. Euro J Neurosci. 2005;21:1767–76. doi: 10.1111/j.1460-9568.2005.04021.x. [DOI] [PubMed] [Google Scholar]
  4. Corcoran J, So PL, Barber RD, Vencent KJ, Mazarakis ND, Mitrophanous KA, Kingsman SM, Maden M. Retinoic acid receptor beta2 and neurite outgrowth in the adult mouse spinal cord in vitro. J Cell Sci. 2002;115:3779–86. doi: 10.1242/jcs.00046. [DOI] [PubMed] [Google Scholar]
  5. Diez del Corral R, Olivera-Martinez I, Goriely A, Gale E, Maden M, Storey K. Opposing FGF and retinoid pathways control ventral neural pattern, neuronal differentiation, and segmentation during body axis extension. Neuron. 2003;40:65–79. doi: 10.1016/s0896-6273(03)00565-8. [DOI] [PubMed] [Google Scholar]
  6. Etienne-Manneville S, Hall A. Rho GTPases in cell biology. Nature. 2002;420:629–35. doi: 10.1038/nature01148. [DOI] [PubMed] [Google Scholar]
  7. Fenstermaker V, Chen Y, Ghosh A, Yuste R. Regulation of dendritic length and branching by semaphorin 3A. J Neurobiol. 2004;58:403–12.31. doi: 10.1002/neu.10304. [DOI] [PubMed] [Google Scholar]
  8. Huber AB, Kania A, Tran TS, Gu C, De Marco Garcia N, Lieberam I, Johnson D, Jessell TM, Ginty DD, Kolodkin AL. Distinct roles for secreted semaphorin signaling in spinal motor axon guidance. Neuron. 2005;48:949–64. doi: 10.1016/j.neuron.2005.12.003. [DOI] [PubMed] [Google Scholar]
  9. Jan YN, Jan LY. The control of dendrite development. Neuron. 2003;40:229–42. doi: 10.1016/s0896-6273(03)00631-7. [DOI] [PubMed] [Google Scholar]
  10. Jessell TM. Neuronal specification in the spinal cord: Inductive signals and transcriptional codes. Nat Rev Genet. 2000;1:20–29. doi: 10.1038/35049541. [DOI] [PubMed] [Google Scholar]
  11. Ji SJ, Zhuang B, Falco C, Schneider A, Schuster-Gossler K, Gossler A, Sockanathan S. Mesodermal and neuronal retinoids regulate the induction and maintenance of limb innervating spinal motor neurons. Dev Biol. 2006;297:249–61. doi: 10.1016/j.ydbio.2006.05.015. [DOI] [PubMed] [Google Scholar]
  12. Koyano Y, Kawamoto T, Shen M, Yan W, Noshiro M, Fujii K, Kato Y. Molecular cloning and characterization of CDEP, a novel human protein containing the ezrin-like domain of the band 4.1 superfamily and the Dbl homology domain of Rho guanine nucleotide exchange factors. Biochem Biophys Res Communi. 1997;241:369–75. doi: 10.1006/bbrc.1997.7826. [DOI] [PubMed] [Google Scholar]
  13. Kuhn TB. Growing and working with spinal motor neurons. Methods in Cell Bio. 2003;71:67–87. doi: 10.1016/s0091-679x(03)01005-7. [DOI] [PubMed] [Google Scholar]
  14. Landmesser L. The development of motor projection patterns in the chick hind limb. J Physiol. 1978;284:391–414.32. doi: 10.1113/jphysiol.1978.sp012546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Lee SK, Jurata LW, Funahashi J, Ruiz EC, Pfaff SL. Analysis of embryonic motoneuron gene regulation: derepression of general activators function in concert with enhancer factors. Development. 2004;131:3295–306. doi: 10.1242/dev.01179. [DOI] [PubMed] [Google Scholar]
  16. Maden M. Retinoic acid in the development, regeneration, and maintenance of the nervous system. Nat Rev Neurosci. 2007;8:755–65. doi: 10.1038/nrn2212. [DOI] [PubMed] [Google Scholar]
  17. Mann F, Chauvet S, Rougon G. Semaphorins in development and adult brain: Implication for neurological diseases. Prog Neurobiol. 2007;82:57–79. doi: 10.1016/j.pneurobio.2007.02.011. [DOI] [PubMed] [Google Scholar]
  18. Mauti O, Sadhu R, Gemayel J, Gesemann M, Stoeckli ET. Expression patterns of plexins and neuropilins are consistent with cooperative and separate functions during neural development. BMC Dev Biol. 2006;6:32. doi: 10.1186/1471-213X-6-32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mauti O, Domanitskaya E, Andermatt I, Sadhu R, Stoeckli ET. Semaphorin6A acts as a gate keeper between the central and the peripheral nervous system. Neural Develop. 2007;2:28. doi: 10.1186/1749-8104-2-28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Moret F, Renaudot C, Bozon M, Castellani V. Semaphorin and neuropilin coexpression in motoneurons sets axon sensitivity to environmental semaphorin sources during motor axon pathfinding. Development. 2007;134:4491–501. doi: 10.1242/dev.011452. [DOI] [PubMed] [Google Scholar]
  21. Nakamura H, Funahashi J. Introduction of DNA into chick embryos by in ovo electroporation. Methods. 2001;24:43–8. doi: 10.1006/meth.2001.1155. [DOI] [PubMed] [Google Scholar]
  22. Novitch BG, Wichterle H, Jessell TM, Sockanathan S. A requirement for retinoic acid-mediated transcriptional activation in ventral neural patterning and motor neuron specification. Neuron. 2003;40:81–95. doi: 10.1016/j.neuron.2003.08.006. [DOI] [PubMed] [Google Scholar]
  23. Okado N, Homma S, Ishihara R, Kohno K. Distribution patterns of dendrites in motor neuron pools of lumbosacral spinal cord of the chicken. Anat and Embryol. 1990;182:113–21. doi: 10.1007/BF00174012. [DOI] [PubMed] [Google Scholar]
  24. Pardo CA, Eberhart CG. The neurobiology of autism. Brain Path. 2007;17:434–47. doi: 10.1111/j.1750-3639.2007.00102.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Polleux F, Morrow T, Ghosh A. Semaphorin 3A is a chemoattractant for cortical apical dendrites. Nature. 2000;404:567–73. doi: 10.1038/35007001. [DOI] [PubMed] [Google Scholar]
  26. Rao M, Sockanathan S. Transmembrane protein GDE2 induces motor neuron differentiation in vivo. Science. 2005;309:2212–2215. doi: 10.1126/science.1117156. [DOI] [PubMed] [Google Scholar]
  27. Rossman KL, Der CJ, Sondek J. GEF means go: turning on RHO GTPases with guanine nucleotide-exchange factors. Nat Rev Mol Cell Biol. 2005;6:167–180. doi: 10.1038/nrm1587. [DOI] [PubMed] [Google Scholar]
  28. Shaeren-Wiemers N, Gerfin-Moser A. A single protocol to detect transcripts of various types and expression levelsin neural tissue and cultured cells: in situ hybridization using digoxigenin-labelled cRNA probes. Histochemistry. 1993;100:431–440. doi: 10.1007/BF00267823. [DOI] [PubMed] [Google Scholar]
  29. Shepherd I, Luo Y, Raper JA, Chang S. The distribution of collapsin-1 mRNA in the developing chick nervous system. Dev Biol. 1996;173:185–199. doi: 10.1006/dbio.1996.0016. [DOI] [PubMed] [Google Scholar]
  30. Sockanathan S, Jessell TM. Motor neuron-derived retinoid signaling specifies the subtype identity of spinal motor neurons. Cell. 1998;94:503–14. doi: 10.1016/s0092-8674(00)81591-3. [DOI] [PubMed] [Google Scholar]
  31. Sockanathan S, Perlmann T, Jessell TM. Retinoid receptor signaling in postmitotic motor neurons regulates rostrocaudal positional identity and axonal projection pattern. Neuron. 2003;40:97–111. doi: 10.1016/s0896-6273(03)00532-4. [DOI] [PubMed] [Google Scholar]
  32. Suto F, Ito K, Uemura M, Shimizu M, Shinkawa Y, Sanbo M, Shinoda T, Tsuboi M, Takashima S, Yagi T, Fujisawa H. Plexin-A4 mediates axon-repulsive activities of both secreted and transmembrane semaphorins and plays roles in nerve fiber guidance. J Neurosci. 2005;25:3628–37. doi: 10.1523/JNEUROSCI.4480-04.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Swiercz JM, Kuner R, Behrens J, Offermanns S. Plexin-B1 directly interacts with PDZ-RhoGEF/LARG to regulate RhoA and growth cone morphology. Neuron. 2002;35:51–63. doi: 10.1016/s0896-6273(02)00750-x. [DOI] [PubMed] [Google Scholar]
  34. Takahashi T, Strittmatter SM. PlexinA1 autoinhibition by the plexin sema domain. Neuron. 2001;29:429–39. doi: 10.1016/s0896-6273(01)00216-1. [DOI] [PubMed] [Google Scholar]
  35. Toyofuku T, Yoshida J, Sugimoto T, Zhang H, Kumanogoh A, Hori M, Kikutani H. FARP2 triggers signals for Sema3A-mediated axonal repulsion. Nat Neurosci. 2005;8:1712–9. doi: 10.1038/nn1596. [DOI] [PubMed] [Google Scholar]
  36. Tran TS, Kolodkin AL, Bharadwaj R. Semaphorin regulation of cellular morphology. Ann Rev Cell Dev Biol. 2007;23:263–92. doi: 10.1146/annurev.cellbio.22.010605.093554. [DOI] [PubMed] [Google Scholar]
  37. Tsuchida T, Ensini M, Morton SB, Baldassare M, Edlund T, Jessell TM, Pfaff SL. Topographic organization of embryonic motor neurons defined by expression of LIM homeobox genes. Cell. 1994;79:957–970. doi: 10.1016/0092-8674(94)90027-2. [DOI] [PubMed] [Google Scholar]
  38. Tucker RP, Binder LI, Matus AI. Neuronal microtubule-associated proteins in the embryonic avian spinal cord. J Comp Neurol. 1988;271:44–55. doi: 10.1002/cne.902710106. [DOI] [PubMed] [Google Scholar]
  39. Vermot J, Schuhbaur B, Le Mouellic H, McCaffery P, Garnier JM, Hentsch D, Brûlet P, Niederreither K, Chambon P, Dollé P, Le Roux I. Retinaldehyde dehydrogenase 2 and Hoxc8 are required in the murine brachial spinal cord for the specification of Lim1+ motoneurons and the correct distribution of Islet1+ motoneurons. Development. 2005;132:1611–21. doi: 10.1242/dev.01718. [DOI] [PubMed] [Google Scholar]
  40. Vrieseling E, Arber S. Target-induced transcriptional control of dendritic patterning and connectivity in motor neurons by the ETS gene Pea3. Cell. 2006;127:1439–52. doi: 10.1016/j.cell.2006.10.042. [DOI] [PubMed] [Google Scholar]
  41. Whitford KL, Dijkhuizen P, Polleux F, Ghosh A. Molecular control of cortical dendrite development. Ann Rev Neurosc. 2002;25:127–149. doi: 10.1146/annurev.neuro.25.112701.142932. [DOI] [PubMed] [Google Scholar]
  42. Wong LF, Yip PK, Battaglia A, Grist J, Corcoran J, Maden M, Azzouz M, Kingsman SM, Kingsman AJ, Mazarakis ND, McMahon SB. Retinoic acid receptor beta2 promotes functional regeneration of sensory axons in the spinal cord. Nat Neurosci. 2006;9:243–50. doi: 10.1038/nn1622. [DOI] [PubMed] [Google Scholar]
  43. Wong M. Stabilizing dendritic structure as a novel therapeutic approach for epilepsy. Exp Rev Neurotherap. 2008;8:907–915. doi: 10.1586/14737175.8.6.907. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Yaron A, Huang PH, Cheng HJ, Tessier-Lavigne M. Differential requirement for Plexin-A3 and -A4 in mediating responses of sensory and sympathetic neurons to distinct class 3 Semaphorins. Neuron. 2005;45:513–23. doi: 10.1016/j.neuron.2005.01.013. [DOI] [PubMed] [Google Scholar]
  45. Yoshida Y, Han B, Mendelsohn M, Jessell TM. PlexinA1 signaling directs the segregation of proprioceptive sensory axons in the developing spinal cord. Neuron. 2006;52:745–6. doi: 10.1016/j.neuron.2006.10.032. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS96378-supplement-01.pdf (9.9MB, pdf)

RESOURCES