Lateral motor column axons execute a ternary trajectory choice between limb and body tissues

Abstract

Background: Neuronal topographic map formation requires appropriate selection of axonal trajectories at intermediate choice points prior to target innervation. Axons of neurons in the spinal cord lateral motor column (LMC), as defined by a transcription factor code, are thought to innervate limb target tissues exclusively. Axons of the medial and lateral LMC divisions appear to execute a binary decision at the base of the limb as they choose between ventral and dorsal limb trajectories. The cellular logic that guides motor axon trajectory choices into non-limb tissues such as the ventral flank remains unclear.

Results: We determined the spinal cord motor column origin of motor nerves that innervate ventral flank tissues at hindlimb level. We found unexpectedly that a subset of medial LMC axons innervates ventral non-limb mesenchyme at hindlimb level, rather than entering ventral limb mesenchyme. We also found that in a conditional BmprIa mutant where all ventral hindlimb mesenchyme is converted to a dorsal identity, all medial LMC axons are redirected into the ventral flank, while lateral LMC axons innervate the bidorsal limb.

Conclusion: We have found that medial LMC neurons innervate both ventral flank and limb targets. While normally only a subset of medial LMC axons innervate the flank, all are capable of doing so. Furthermore, LMC axons execute a ternary, rather than binary, choice at the base of the limb between ventral flank, ventral limb and dorsal limb trajectories. When making this choice, medial and lateral LMC axons exhibit different and asymmetric relative preferences for these three trajectories. These data redefine the LMC as a motor column that innervates both limb and body tissues.

Figure 1

Medial LMC axons innervate the ventral flank mesenchyme. The columnar origin of neurons innervating the ventral flank mesenchyme was defined by HRP or RDA retrograde labeling at E13.5. (a) Schematic cross section at hindlimb level, showing limb and flank axonal trajectories and approximate location of the retrograde label injection. Inset: relative positions of motor columns in C-E. (b) Transcription factor combinations expressed by medial LMC (LMC(m)), lateral LMC (LMC(l)) and medial MMC (MMC(m)) neurons. (c,d,e) Representative images showing HRP colocalization with Isl1, and not Lim1 (c) or Lim3 (d), but with FoxP1 (e), in triply immunostained sections. Note that images show representative staining patterns and marker colocalization, and are not intended to be quantitative. Arrowheads: HRP+ neurons colabeled with nuclear markers. Dotted lines (c-e) delineate columnar division borders. n = 5 embryos.

Figure 2

BmprIa^flox/- hindlimbs are bidorsal at and around the time of the motor axon dorsoventral projection choice. The dorsoventral character of limb mesenchyme was defined using molecular markers. Brn4-cre^Tg/-, BmprIa^flox/- (BmprIa^flox/-) mutant hindlimbs are completely dorsalized. (a-d) Upper panels: control hindlimbs. At E11.0 axons have reached the base of the hindlimb (a,b). Plzf is a general limb marker (b), and Lmx1b is a dorsal limb marker (a,b). EphA4 marks proximal dorsal mesenchyme (c) and Σephrin-A marks ventral mesenchyme (d) at E11.5. Note coincidence of proximal expression borders. Lower panels: BmprIa^flox/- mutant hindlimbs. Marker expression demonstrates they are completely dorsalized (Lmx1b+ Σephrin-A-) (a,b,d) and that they are proximally dorsal (EphA4+) (c). (e) Summary of molecular dorsoventral code marker expression in control (upper) or bidorsal BmprIa^flox/- mutant (lower) hindlimb. RNA in situ hybridization (Plzf), immunostaining (Lmx1b, EphA4) and AP fusion protein staining (Σephrin-A) were performed on adjacent transverse cryosections and images were digitally superimposed. n = 12 (a,b), 8 (c), 5 (d) embryos. Top, dorsal; left, medial. Dotted lines: limb outline. NF, neurofilament. Asterisks indicate the position of the plexus.

Figure 3

Both nerve branches contain lateral LMC axons in BmprIa^flox/- hindlimbs. Comparison of anterior plexus axonal projection patterns in control and mutant embryos. A Lim1^tlzallele was used to label lateral LMC axons with LacZ. (a,b) E11.75 control hindlimb; overlays of two adjacent sections. (a) NF+ nerves form three branches at the base of the hindlimb, with only the dorsal branch entering Lmx1b+ dorsal limb mesenchyme. (b) Only the dorsal limb branch contains LacZ+ lateral LMC axons. (c,d) E11.75 mutant hindlimb; in consecutive coimmunostained sections, dorsal, ventral and flank branches can be followed. (c) NF+ nerves form three branches, two of which enter Lmx1b+ limb mesenchyme. (d) Both limb nerve branches contain LacZ+ lateral LMC axons, while the ventral flank branch does not. (e) LacZ immunoreactivity was quantified and normalized for neurofilament immunoreactivity in each nerve branch; low-level signal in ventral limb and flank branches likely represents weak non-specific staining. Values are presented in relative units that represent the normalized LacZ signal in each branch. The relative signals in normal limbs (white bars) were (mean ± SEM): 5.3 ± 0.9 (dorsal), 1.1 ± 0.2 (ventral), 1 ± 0.1 (flank); n = 5 embryos. The relative signals in mutant limbs (black bars) were: 4.8 ± 1 (dorsal), 4.6 ± 0.9 (ventral), 1 ± 0.2 (flank); n = 5 embryos. P = 0.00005 for ventral limb branches, two-tailed t test. *P < 0.00005. D, dorsal limb nerve branch; F, ventral flank nerve branch; V, ventral limb nerve branch. Arrows: LacZ+ nerves. Dotted lines: limb outlines. Boxed areas in (a,c) are enlarged in (b,d).

Figure 4

Medial LMC axons innervate the BmprIa^flox/- ventral flank mesenchyme. The columnar origin of neurons innervating the ventral flank mesenchyme at E13.5 was defined by HRP or RDA retrograde labeling. (a) Schematic of experiments shown in (b-d). (b) Flank nerves originate from medial LMC neurons in normal and mutant embryos, as Isl1+ medial LMC cells are readily HRP-labeled in triple coimmunostained sections. (c) Flank nerves in normal and mutant do not originate from medial MMC cells, as Lim3+ medial MMC cells are rarely HRP+ in triple coimmunostained sections. (d) Flank nerves in normal and mutant originate from medial LMC but not medial MMC as coimmunostained HRP+ Isl1+ FoxP1+ medial LMC cells were readily detected. In (b-d), upper panels are controls and lower panels are BmprIa^flox/- mutants. Arrowheads indicate representative HRP+ neurons colabeled with nuclear markers and dotted lines in (b-d) indicate columnar division outlines. (e) Quantification of labeling data shows that axons projecting to ventral flank mesenchyme in normal and BmprIa^flox/- mutant embryos are from Isl1+ Lim1- Lim3- FoxP1+ medial LMC neurons. The percentages of HRP+ or RDA+ cells were as follows. Isl1+ Lim1-: normal, 94 ± 1.2, n = 5 embryos, 648 neurons counted; mutant, 93 ± 2.1, n = 5, 388 neurons; P = 0.83. Lim3+: normal, 3.5 ± 1.8, n = 5, 515 neurons; mutant: 2.4 ± 1.4, n = 4, 283 neurons; P = 0.62. FoxP1: normal, 95 ± 1.5, n = 5, 479 neurons; mutant, 96 ± 1.5, n = 4, 262 neurons; P = 0.63). (f) Anterior-posterior length of the LMC is similar in normal and mutant E13.5 embryos, while the extent of the LMC labeled from the ventral flank is significantly increased in mutant embryos. LMC lengths (Isl1+/Hb9-GFP+ sections) were: normal, 1,658 ± 23 μm, n = 6 embryos; mutant, 1,690 ± 113 μm, n = 5; P = 0.745. Tracer extents (HRP+ or RDA+ sections) were: normal, 407 ± 96 μm, n = 5 embryos; mutant, 864 ± 116 μm, n = 5; P = 0.008.

Figure 5

Medial LMC axons in the BmprIa^flox/- mutant are redirected from the ventral limb into the ventral flank. Scip immunostaining and retrograde labeling (HRP or RDA) of nerves at E13.5 was used to define the contribution of Scip+ neurons to each nerve branch. (a) Retrograde labeling from normal dorsal or ventral limb mesenchyme. Dorsal nerves originate from Scip- lateral LMC neurons (a, i-vi). HRP-labeled cells do not express Scip and are Isl1- Lim1+ lateral LMC cells (white arrows). Scip+ medial LMC cells contribute to ventral nerves (a, vii-xii). Many HRP-labeled cells express Scip and all Scip+ cells are Isl1+ Lim1- medial LMC neurons (white arrows). (b) Retrograde labeling from normal or mutant ventral flank mesenchyme. In normal and mutant embryos retrogradely labeled Scip- cells are readily detected in sections of rostral spinal cord (b, i-vi). Labeled cells are detected in caudal lumbar spinal cord only in BmprIa^flox/- mutants (b, vii-xii, white arrows). Many of these labeled neurons are also Scip+. Similar staining patterns were observed in three embryos for each labeling experiment. Representative sections are shown of labeled motor columns co-immunostained for HRP and Scip. Experiments in (a,b) are diagrammed schematically to the right, and boxed areas show the regions of the images. (c) Quantification of flank retrograde labeling data. The fraction of HRP+ or RDA+ flank-labeled cells that are also Scip+ differs significantly (P < 0.02) between normal and mutant (normal: white bar, 6.4% ± 2.6%, n = 6 embryos, N = 1,018 labeled neurons; mutant: black bar, 23.6% ± 4.7%; n = 5 embryos, N = 784 labeled neurons). (d) The anteroposterior span of the Scip+ pool within the E13.5 lumbar LMC does not differ significantly between normal and mutant embryos (normal: 701 ± 26 μm, n = 6 embryos; mutant: 592 ± 94 μm, n = 5; P = 0.219).

Figure 6

Medial LMC axons enter only the ventral flank nerve branch in the BmprIa^flox/- hindlimb. Anterograde labeling was used to determine the trajectories of axons originating within the medial LMC. (a) Schematic showing the dorsal approach used for the anterograde HRP labeling of spinal cord neurons at E12.5. Regions of interest in (b,c) are boxed. (b) Triple co-immunostaining showing spinal cords in which the vast majority of HRP+ cells are Isl1+ Lim1- medial LMC neurons. (c) In normal hindlimbs, both ventral flank and ventral limb branches are labeled with HRP (n = 5 embryos, label detected in ventral flank branch = 5/5, ventral limb branch = 5/5, dorsal limb branch = 0/5), while neurofilament labeling identifies all three nerve branches. In mutant hindlimbs, only the ventral flank branch is HRP+ (n = 5 embryos, ventral flank branch = 5/5, ventral limb branch = 0/5, dorsal limb branch = 0/5). Adjacent sections from anterior plexus are shown for both normal and mutant. Arrowheads: representative HRP+ Isl1+ neurons. Arrows: HRP+ nerve branches. Dotted lines: outlines of lateral and medial LMC. D, dorsal limb nerve branch; F, ventral flank nerve branch; V, ventral limb nerve branch.

Figure 7

Lumbar spinal cord motor axon projections to the ventral flank. (a-d) E13.5 normal and BmprIa^flox/- embryos carrying an Hb9-GFP transgene imaged via indirect fluorescence. The embryos were eviscerated and the ventral abdominal wall was reflected to allow examination from a peritoneal aspect. Images of rostral (a,c) and caudal (b,d) regions are of different embryos, and in the caudal images the more rostral (for example, genitofemoral) nerves were removed during the dissection. In control embryos three major motor nerves (white arrows) project abdominally via the ventral flank from rostral segments of the lumbar spinal cord (a), while none project abdominally from more caudal segments (b). The limb nerves (f, femoral nerve; o, obturator nerve; S, sciatic plexus) turn along dorsal or ventral limb trajectories into the plane of the image. In mutant embryos, additional flank-projecting nerves (yellow arrows) are present at both rostral (c) and caudal (d) levels. (e) Quantification of nerve trajectories at E11.5–11.75. Transverse sections of normal and BmprIa^flox/- embryos carrying an Hb9-GFP transgene were scored for the presence of a GFP+ dorsal limb, ventral limb or ventral flank nerve branch. The combined continuous anteroposterior extent of sections containing nerves following each trajectory is presented. gf, genitofemoral nerve; ih, iliohypogastric nerve; il, ilioinguinal nerve. M, medial; L, lateral; A, anterior; P, posterior.

Figure 8

Only lateral LMC axons innervate BmprIa^flox/- hindlimb mesenchyme. HRP retrograde labeling at E13.5 identified spinal cord neurons that contribute to each limb nerve branch. (a) Dorsal limb nerve originates from the lateral LMC in normal and mutant as HRP+ Lim1+ lateral LMC cells are readily detected, while HRP+ Isl1+ cells are not, in triple co-immunostained sections. (b) Schematic of the experiment in (a). (c) Ventral limb nerve originates from the medial LMC in normal embryo as HRP+ cells are also Isl1+ Lim1-. In contrast, this nerve originates from lateral LMC neurons in the mutant as HRP+ cells are Lim1+ Isl1-. (d) Schematic of the experiment in (c). In (a-d) the upper panels show control embryos and the lower panels show BmprIa^flox/- mutants. Arrowheads indicate representative HRP+ Lim1+ or HRP+ Isl1+ neurons. Dotted lines in (a,c) indicate lateral and medial LMC outlines. Boxed areas in (b,d) indicate regions of interest in (a,c). (e,f) Quantification of (e) dorsal and (f) ventral retrograde labeling data. Lim1+ Isl1- HRP+ lateral LMC or Lim1- Isl1+ HRP+ medial LMC cells were counted in triply immunostained sections after retrograde labeling of control and BmprIa^flox/- hindlimbs, and are shown as percentages of total HRP+ cells. Dorsal limb HRP injection does not label significantly different populations in normal and mutant (normal: lateral LMC 97% ± 3%, n = 3 embryos, N > 150 HRP+ neurons; mutant: lateral LMC 95% ± 2%, n = 8 embryos, N > 350 HRP+ neurons; P = 0.6). Ventral limb HRP injection does label significantly different cell populations (normal: medial LMC 95% ± 1%, n = 6 embryos, N > 300 HRP+ neurons; mutant: medial LMC 6% ± 2%, n = 9 embryos, N > 400 HRP+ neurons; P < 0.00001). *P < 0.00001.

Figure 9

The motor innervation of Hb9^cre/+, BmprIa^flox/- hindlimbs is normal. Retrograde labeling was used to determine the columnar origin of axons innervating Hb9^cre/+, BmprIa^flox/- hindlimbs. (a) Dorsal nerves originate from Isl1- Lim1+ lateral LMC neurons in normal and Hb9^cre/+, BmprIa^flox/- mutant embryos. (b) Schematic of the experiment in (a). (c) Ventral nerves originate from Isl1+ Lim1- medial LMC neurons in normal and Hb9^cre/+, BmprIa^flox/- mutant embryos. (d) Schematic of the experiment in (c). (e,f) Quantification of dorsal (e) and ventral (f) retrograde labeling data. Lateral LMC cells were Lim1+ Isl1- HRP+; medial LMC cells were Lim1- Isl1+ HRP+. The percentage of labeled lateral or medial LMC cells marked by either dorsal or ventral limb HRP injection does not differ significantly between normal and mutant. Dorsal: lateral LMC 93% ± 0.9%, n = 8 embryos, N > 750 HRP+ neurons; mutant lateral LMC 93% ± 0.5%, n = 3 embryos, N > 450 HRP+ neurons; P = 0.79. Ventral: medial LMC 95% ± 1.5%, n = 5 embryos, N > 350 HRP+ neurons; mutant medial LMC 96% ± 0.9%, n = 3 embryos, N > 240 HRP+ neurons; P = 0.60.

Figure 10

Summary of motor projections to bidorsal hindlimbs. (a) Observed trajectories of motor axons. In Brn4-cre^Tg/-, BmprIa^flox/- mutant embryos with bidorsal hindlimbs, medial LMC axons do not invade the limb mesenchyme but are deflected to the ventral flank. Lateral LMC axons innervate both the dorsal and the ventral half of the bidorsal limb. (b) Trajectory preferences. LMC axons normally choose among three mesenchymal trajectories at the base of the limb, where they have different relative trajectory preferences (I > II > III for each branch; as it is unclear if lateral LMC axons ever enter the ventral flank, this branch is represented as a broken line). Medial LMC: ventral limb > ventral flank > dorsal limb. Lateral LMC: dorsal limb > ventral limb > ventral flank. (c) Model of guidance interactions. All medial LMC axons express receptor for a repellant guidance cue. A rostral subset of axons expresses high levels of receptor. In normal embryos, target mesenchyme expresses a step gradient of repellant cues: high levels in the dorsal limb, intermediate levels in the ventral limb, low levels in the ventral flank. Medial LMC axons expressing receptor at low levels are directed to the ventral limb by the high level of dorsal limb repellant, while the medial LMC axons expressing the receptor at high levels are directed to the flank by the intermediate levels of ventral limb repellant. All medial LMC axons also express receptor for an attractant cue expressed in limb mesenchyme. In bidorsal embryos, both limb halves express the repellant cue at high levels. Consequently, both subpopulations of medial LMC axons are directed to the flank. Blue (Y): receptor for repellant cue. (-): repellant guidance cue. (+): attractant guidance cue. Attractant receptor expressed on all medial LMC axons is not shown. Red axons: medial LMC. Green axons: lateral LMC.