Foxp1 and lhx1 coordinate motor neuron migration with axon trajectory choice by gating Reelin signalling

Abstract

Topographic neuronal maps arise as a consequence of axon trajectory choice correlated with the localisation of neuronal soma, but the identity of the pathways coordinating these processes is unknown. We addressed this question in the context of the myotopic map formed by limb muscles innervated by spinal lateral motor column (LMC) motor axons where the Eph receptor signals specifying growth cone trajectory are restricted by Foxp1 and Lhx1 transcription factors. We show that the localisation of LMC neuron cell bodies can be dissociated from axon trajectory choice by either the loss or gain of function of the Reelin signalling pathway. The response of LMC motor neurons to Reelin is gated by Foxp1- and Lhx1-mediated regulation of expression of the critical Reelin signalling intermediate Dab1. Together, these observations point to identical transcription factors that control motor axon guidance and soma migration and reveal the molecular hierarchy of myotopic organisation.

Figure 1. Ventral spinal cord expression of Reelin, VLDLR, ApoER2 and Dab1.

(A–P, U–AB) Protein detection in e11.5 and e12.5 mouse lumbar spinal cord. LMCm (M) and LMCl (L) neurons are identified as Foxp1⁺ Isl1⁺ and Foxp1⁺ Isl1⁻, respectively. Reelin is expressed in a domain medio-dorsal to the LMC at e11.5 (A–C) and expands more ventrally at e12.5 (G; arrowheads). VLDLR expression is stronger in LMCl than in LMCm neurons (arrowheads) at e11.5 (I–K). Arrow in (O) indicates VLDLR protein in neuronal processes extending towards the ventricular zone. Dab1 expression is stronger in LMCl neurons (arrowheads in W and AA) than in LMCm neurons at both e11.5 (U–W) and e12.5 (Y–AA). (Q–S, AC–AE) Detection of mRNA in consecutive lumbar spinal cord sections of an e11.5 (Q–S) and e12.5 (AC–AE) mouse embryo. Isl1 (Q, AC) and Lhx1 (R, AD) expression highlights LMCm and LMCl motor neurons, respectively. ApoER2 mRNA is detected throughout the ventral spinal cord (S), with higher levels of expression in the ventricular zone (VZ). Inset in (S) shows a lower magnification image of the same section, highlighting the specificity of the probe. Dab1 mRNA expression levels are higher in LMCl neurons than in LMCm neurons (AE). (AF) Quantifications of Dab1 mRNA and protein levels in LMCm (Isl1⁺ Foxp1⁺) and LMCl (Isl1⁻ Foxp1⁺) in lumbar spinal cord of e12.5 mouse embryos. Quantifications were gated on Lhx1⁺ LMCl and Isl1⁺ LMCm regions in neighbouring sections and represent a mean intensity value within the gated region with the background (gating on the ventricular zone) subtracted. arb: arbitrary units on a 0 (no signal) to 255 (maximum intensity) scale. Measurements are based on three embryos and six sections. (D, H, L, P, T, X, AB) Schematic representation of LMCm, LMCl, and Reelin signalling components. Blue intensity varies with expression levels. Stippled lines outline the spinal gray matter and LMCl neurons. Scale bar: 65 µm (A–AB), 50 µm (AC–AE).

Figure 2. Impaired LMC position in Dab1 and Reln mutants.

(A–H, J–Q) LMCm (Foxp1⁺ Isl1⁺) and LMCl (Foxp1⁺ Isl1⁻) neurons in the lumbar spinal cord of e12.5 Dab1 mutants (E–H), Reln mutants (N–Q), and wild type littermates (A–D, J–M). In both mutants, LMCl neurons are positioned more ventrally, while LMCm neurons are shifted laterally, relative to control embryos. Some overlap between LMCl and LMCm domain is also evident (M/L in panels E and N). Superimposed ventral spinal cord position (D, H, M, Q) of LMCm (red) and LMCl (green) neurons in several consecutive sections of representative embryos highlights the altered position of LMC neurons. (I, R) Density plots of mediolateral (ML) and dorsoventral (DV) position of LMCm and LMCl neurons as percentage of LMC width and height. See Materials and Methods for details. Mean position (ML, DV) is in brackets. (I) Hotelling's T² test p values: LMCm(wt) versus LMCm(Dab1), p = 0.2925; LMCl(wt) versus LMCl(Dab1), p = 0.0035. n = 4 (Dab1) and 6 (wt) embryos; N>1,500 neurons per genotype. (R) Hotelling's T² test p values: LMCm(wt) versus LMCm(Reln), p = 0.9024; LMCl(wt) versus LMCl(Reln), p = 0.0473. n = 4 embryos per genotype; N>1,500 neurons per genotype. (S–Z) Detection of Foxp1, Lhx1/5, and Pea3 on consecutive sections of lumbar spinal cord of e15.5 Dab1 mutants (W–Z) and control littermate (S–V). In Dab1 mutants, LMCl (Foxp1⁺ Lhx1/5⁺) neurons are positioned medio-ventrally relative to control embryos. The position and clustering of the Pea3⁺ (Isl1⁻ Foxp1⁺) motor pool in the e15.5 lumbar spinal cord of Dab1 mutants (Z) is disrupted compared to control embryos (V). All values are expressed as mean ± s.d. Dashed lines divide the plots into four equal quadrants. Stippled lines outline the spinal grey and LMCl neurons. Scale bar: 77 µm (A–H, J–Q), 50 µm (S–Z).

Figure 3. Dab1 is sufficient to specify LMC neuron position.

(A–C, E–G) GFP expression in LMCm (Foxp1⁺Isl1⁺) and LMCl (Foxp1⁺Isl1⁻) neurons in lumbar spinal cord of chick HH St 29 embryos electroporated with GFP (A–C) or Dab1::GFP expression plasmids (E–G). Arrowheads point to electroporated LMCm neurons shown in insets at higher magnification. (D, H) Superimposed ventral spinal cord location of electroporated LMCm (red) and LMCl (green) neurons in several consecutive sections of representative embryos highlighting the laterally shifted position of Dab1-overexpressing LMCm neurons. Blue dashed line indicates the 50% ML value. (I) Density plots of mediolateral (ML) and dorsoventral (DV) position of LMCm and LMCl neurons as percentage of LMC width and height. Mean position (ML, DV) is indicated in brackets. Hotelling's T² test p values: LMCm(GFP) versus LMCm(Dab1::GFP), p = 0.0165; LMCl(GFP) versus LMCl(Dab1::GFP), p = 0.9019. n = 4 (GFP), 5 (Dab1::GFP) embryos; N>60 neurons per embryo per experimental condition. Dashed lines divide LMC in four equal quadrants. All values are expressed as mean ± s.d.. Stippled lines outline the spinal gray matter. Scale bar: 20 µm.

Figure 4. Loss of Reelin signalling disrupts LMC myotopy.

(A–D) Axonal neurofilament (NF) and LacZ protein in a representative forelimb section of e11.5 control Lhx1^tlz/+ littermate (A–B) and Dab1 ^−/−; Lhx1^tlz/+ embryos (C–D). (E) Quantification of LacZ⁺ axons within dorsal (d) and ventral (v) limb nerves expressed as a percentage of total LacZ⁺ signal within both limb nerves. Student's t test p value: Dab1^+/+; Lhx1^tlz/+ versus Dab1 ^−/−; Lhx1^tlz/+, p≥0.5. n = 3 embryos per genotype; N>8 sections quantified per embryo. (F–I) Neurofilament (NF) immunodetection and alkaline phosphatase (PLAP) enzymatic detection in axons of a representative forelimb section of e11.5 control hCrest/Isl1-PLAP littermate (F–G) and Dab1 ^−/−; hCrest/Isl1-PLAP embryos (H–I). PLAP enzymatic reaction signal images (G, I) were colour-inverted and overlaid with same section NF signal images in (F, H). (J) Quantification of PLAP⁺ axons within dorsal (d) and ventral (v) limb nerves expressed as percentage of total PLAP⁺ signal within both limb nerves. Student's t test p value: Dab1^+/+; hCrest/Isl1-PLAP versus Dab1 ^−/−; hCrest/Isl1-PLAP, p = 0.335. n = 3 embryos per genotype; N>6 sections quantified per embryo. Scale bar: 90 µm.

Figure 5. Foxp1 regulates Dab1 expression.

(A–L) Dab1 mRNA and protein expression in the cervical spinal cord of e12.5 Hb9::Foxp1 transgenic (D–F), Foxp1 mutant (J–L), and control embryos (A–C, G–I). (M) Dab1 mRNA and protein mean pixel intensity measurement within Foxp1⁺ Lhx3⁻ LMC regions (fine stippled outline; D–F, G–I) or the corresponding Foxp1⁻ Lhx3⁻ region (arrowheads; B–C, K–L). arb: arbitrary units on a 0 (no signal) to 255 (maximum intensity) scale. Student's t test p values for mRNA quntifications: Hb9::Foxp1 versus wt, p = 0.002, n = 3 embryos per genotype analysed; Foxp1 ^−/− versus Foxp1^{+/ −}, p<0.001, n = 3 embryos per genotype analysed; Student's t test p values for protein quantification: Hb9::Foxp1 versus wt, p<0.001, n = 4 embryos per genotype analysed; Foxp1 ^−/− versus Foxp1^{+/ −}, p<0.001, n = 3 embryos per genotype analysed. All sections of the same genotypes are consecutive. Arrows: Foxp1⁺ interneurons (INs); LMC*: ectopic LMC; bracket: expanded MMC/HMC*. Thicker stippled lines outline the spinal gray matter. Scale bar: 50 µm.

Figure 6. Impaired LMC migration in Lhx1 mutants.

(A–H) Foxp1, Isl1, and Lhx1/5 expression identifies LMCm (Foxp1⁺ Isl1⁺), LMCl (Foxp1⁺ Isl1⁻ Lhx1/5⁺) and LMCl* (Foxp1⁺ Isl1⁻ Lhx1/5⁻) neurons in the lumbar spinal cord of e12.5 Lhx1 mutants (E–H), and heterozygous littermates (A–D). In Lhx1 mutants, LMCl* cells are found more medially and intermingled with LMCm neurons (arrowheads). (I) Proportion of LMCl motor neuron in Lhx1 mutants and control embryos expressed as percentage of Isl1⁻ Foxp1⁺ motor neurons (37.3%±9% versus 95.2%±2%; p<0.001, Student's t test). (J) Density plots of mediolateral (ML) and dorsoventral (DV) position of LMCm and LMCl neurons as percentage of LMC width and height. Mean position (ML, DV) is indicated in brackets. Hotelling's T² test p values: LMCm(Lhx1^{+/ −}) versus LMCm(Lhx1^COND), p = 0.1613; LMCl*(Lhx1^COND) versus LMCl(Lhx1^COND), p = 0.2825; LMCl*(Lhx1^COND) versus LMCl(Lhx1^{+/ −}), p = 0.9886. (K) Proportion of LMCl and LMCl* neurons in quadrants 1–3 and 2–4 of Lhx1^{+/ −} and Lhx1^COND embryos expressed as percentage of all neurons analysed. Student's t test p value: LMCl(Lhx1^{+/ −}) versus LMCl*(Lhx1^COND), p = 0.008. Dashed lines in (J) divide the plots into four equal quadrants. Stippled lines outline the spinal gray matter. n = 7 (Lhx1^COND) and 5 (Lhx1^{+/ −}) embryos; N>1,400 neurons per genotype. Scale bar: 50 µm.

Figure 7. Lhx1 controls the differential expression of Dab1 in LMC neurons.

(A–N) e12.5 lumbar spinal cord Dab1 protein (A–H) and mRNA (I–N) expression in LMCl (Isl1⁻ Foxp1⁺) is decreased in Lhx1^COND mutants (E–H, L–N) compared to control embryos (A–D, I–K). (I–N) Detection of Lhx1, Isl1, and Dab1 mRNA in consecutive sections of e12.5 Lhx1^COND and control lumbar spinal cord. (O) Quantification of Dab1 protein and mRNA levels in LMC (Foxp1⁺), LMCm (Isl1⁺ Foxp1⁺), and LMCl (Isl1⁻ Foxp1⁺) spinal cord area normalized to Dab1 protein or mRNA levels in LMC of littermate control embryos. Dab1 protein expression in LMC of Lhx1 mutants was 80%±16% of the expression level of control littermate embryos (p = 0.038; Student's t test; n = 4 embryos per genotype analysed). In heterozygous embryos LMCm Dab1 protein expression was 39%±4% and in Lhx1 mutants was 42%±5% (p>0.5; Student's t test; n = 4 embryos per genotype analysed), while Dab1 mRNA expression in Lhx1^COND mutants was 29%±6% compared to 30%±6% in heterozygous embryos (p>0.5; Student's t test; n = 3 embryos per genotype analysed). In heterozygous embryos, LMCl Dab1 protein expression (61%±4%) and mRNA (70%±6%) was significantly different from Lhx1^COND mutants (protein 47%±8%; p = 0.016, Student's t test; n = 4 embryos per genotype analysed; mRNA 41%±2%; p = 0.01, Student's t test; n = 3 embryos per genotype analysed). All values are expressed as mean ± s.d. Yellow ovals highlight LMCl neurons; yellow lines outline the spinal gray. Scale bar: 66 µm (A–H), 50 µm (I–N).

Figure 8. Model of transcriptional control of LMC myotopy.

(A) Foxp1, Isl1, and Lhx1 control Dab1 and Eph receptor expression in LMC neurons and thus coordinate LMC neuronal migration and axon projections. (B) Molecular hierarchy of transcription factor control of LMC myotopy. See discussion for details.