Layer 6 cortical neurons require Reelin-Dab1 signaling for cellular orientation, Golgi deployment, and directed neurite growth into the marginal zone

Abstract

Background: The secreted ligand Reelin is believed to regulate the translocation of prospective layer 6 (L6) neocortical neurons into the preplate, a loose layer of pioneer neurons that overlies the ventricular zone. Recent studies have also suggested that Reelin controls neuronal orientation and polarized dendritic growth during this period of early cortical development. To explicitly characterize and quantify how Reelin controls this critical aspect of neurite initiation and growth we used a new ex utero explant model of early cortical development to selectively label a subset of L6 cortical neurons for complete 3-D reconstruction.

Results: The total neurite arbor sizes of neurons in Reelin-deficient (reeler mutant) and Dab1-deficient (Reelin-non-responsive scrambler mutant) cortices were quantified and unexpectedly were not different than control arbor lengths (p = 0.51). For each mutant, however, arbor organization was markedly different: mutant neurons manifested more primary processes (neurites emitted directly from the soma) than wild type, and these neurites were longer and displayed less branching. Reeler and scrambler mutant neurites extended tangentially rather than radially, and the Golgi apparatus that normally invests the apical neurite was compact in both reeler and scrambler mutants. Mutant cortices also exhibited a neurite "exclusion zone" which was relatively devoid of L6 neuron neurites and extended at least 15 μm beneath the pial surface, an area corresponding to the marginal zone (MZ) in the wild type explants. The presence of an exclusion zone was also indicated in the orientation of mutant primary neurite and neuronal somata, which failed to adopt angles within ~20˚ of the radial line to the pial surface. Injection of recombinant Reelin to reeler, but not scrambler, mutant cortices fully rescued soma orientation, Golgi organization, and dendritic projection defects within four hrs.

Conclusions: These findings indicate Reelin promotes directional dendritic growth into the MZ, an otherwise exclusionary zone for L6 neurites.

Figure 1

Schematic of explant culture model. (A) Brains from E13 embryos were electroporated ex utero with a GFP expression construct, dissected, and whole hemisphere explants were cultured on collagen-coated filters for 2 DIV. The embryos were genotyped and explants identified as wild type (wt), reeler (r/r,) or scrambler (s/s). After 48 hrs in culture, explants were injected with control conditioned media (Ct-CM) or Reelin conditioned media (Rln-CM), and cultured for an additional 4 hrs before fixation. (B) E13 ex utero electroporation targets prospective L6 neurons. Sections derived from wt explants were immunostained for the layer 5/6 transcription factors Ctip2 (top row) or Tbr1 (bottom row). The majority of the GFP + cells in the CP expressed Ctip2 (93%) and Tbr1 (93%). Scale bars: 20 μm in (B). For Ctip2, n = 363 GFP + positive neurons were totaled from 4 explants across 2 separate litters. For Tbr1, n = 326 GFP + neurons were totaled from 5 explants across 3 separate litters. All GFP + neurons included in these analyses were localized within 50 μm of the pial surface. Abbreviations: MZ, marginal zone; CP, cortical plate; IZ, intermediate zone.

Figure 2

Disorganized L6 neuronal morphology inreelerandscramblerexplants and rescue by Reelin injection. Hoechst nuclear stain (first column) reveals normal CP that forms during the 2 DIV culture period in wt (A) explants but not in r/r (B) or s/s (C) explants that were injected with Ct-CM. GFP expressing neurons (second column) show normal radially oriented neurites in wt (A) explants, but tangentially oriented processes in mutant (B, C) explants. 3-D rendering of imaged neurons (third column) confirms the radial and tangential orientations of L6 neurons observed in wt (A) and mutant (B, C) explants, respectively. Golgi/GM130 immunofluorescent labeling (fourth column) revealed elongated, pia-oriented Golgi colocalized to the apical neurite of L6 neurons in wt explants (yellow arrowheads; A), while condensed juxtanuclear Golgi were observed in both r/r Ct-CM and s/s Ct-CM mutant explants (blue arrowheads; B and C). (D) Within four hrs of exogenous Rln-CM injection into r/r explants, GFP + L6 neurons were oriented towards the pial surface in a clearly definable CP, demonstrated neurite elaboration into the MZ, and revealed elongated, pia-oriented Golgi colocalized to the apical neurite (yellow arrowheads). Purple asterisk represents traced neurons in 3-D rendering (third column). The single green line of the 3-D rendered traced neuron represents the primary neurite, while purple lines represent remaining neurites and corresponding branches. Scale bars: 20 μm in (A-D). Abbreviations: MZ, marginal zone; CP, cortical plate; SPP, superplate; IZ, intermediate zone.

Figure 3

Reelin-Dab1 signaling is required for normal neurite organization and branching but does not control total neurite length. (A) L6 neurons in r/r Ct-CM and s/s Ct-CM explants had the same total neurite arbor length as wt neurons, despite longer primary neurites observed in mutant neurons. (B) Mutant neurons also displayed higher numbers of primary neurites (emitted from cell soma), but a decrease in higher order branch number as compared to wt controls. Rln-CM injection into r/r mutant explants rescued primary neurite length and branching patterns. (C) Categorization of neurons by primary neurite number and treatment condition. (D) 3-D Sholl analysis of traced GFP + neurons revealed consistent profiles of neurite crosses vs. radius for r/r Rln-CM neurons and wt neurons at all radial domains (a-d). Error bars denote standard error of the mean. Kruskal-Wallis one-way ANOVA on ranks with post-hoc Dunn tests were performed between treatment conditions. * p < 0.05, as compared to wt controls. wt analysis: 209 neurons reconstructed from 22 explants across 9 litters. r/r Ct-CM analysis: 105 neurons reconstructed from 7 explants across 5 separate. s/s Ct-CM analysis: 49 neurons reconstructed from 3 explants across 3 litters. r/r Rln-CM analysis: 86 neurons reconstructed from 8 explants across 4 separate litters.

Figure 4

Reelin-Dab1 signaling is required for orientation of neuronal somata and primary neurites. (A) The angles of the neuronal soma (θ) and primary neurite (φ) with respect to the shortest line from the somal tip to the pial surface (red line in A) were measured in 3-D using a custom MATLAB script. Only GFP + neurons within 50 μm of the pial surface were included for analysis. Scatter and box plots of somal orientation (B) and primary neurite orientation (C) in all four explant conditions. Both r/r Ct-CM and s/s Ct-CM neurons demonstrated significant somal and primary neurite misorientation compared to wt neurons. Injection of Rln-CM significantly rescued both somal and primary neurite orientation angles in r/r mutant explants. Dashed lines on both scatter and box plots represent mean values; solid lines on box blots represent median values. Box plots display upper and lower quartiles; whiskers represent 90th and 10th percentiles. Kruskal-Wallis one-way ANOVA on ranks with post-hoc Dunn tests were performed between treatment conditions. * p < 0.05, as compared to wt controls. wt analysis: 209 neurons from 22 explants across 9 litters. r/r Ct-CM analysis: 105 neurons from 7 explants across 5 separate litters. s/s Ct-CM analysis: 49 neurons from 3 explants across 3 litters. r/r RlnCM analysis: 86 neurons from 8 explants across 4 separate litters.

Figure 5

L6 neurons in mutant explants have decreased neurite elaboration into the MZ. (A) Representation of threshold analysis for the measurement of neurite elaboration into the MZ. Threshold was determined for the flattened, 8-bit grayscale image, and the ratio of GFP + pixels in the MZ (a; ~ 15 x 100 μm ROI) to GFP + pixels in the CP (b; ~ 35 x 100 μm ROI) was determined for each treatment condition. (B) Both r/r Ct-CM and s/s Ct-CM L6 neurons revealed decreased GFP + signal (i.e. neurite elaboration) into the MZ as compared to wt controls and r/r Rln-CM explants. Ratios were compared across images acquired from 18 wt explants, 8 r/r Ct-CM explants, 4 s/s Ct-CM explants, and 8 r/r Rln-CM explants. Error bars denote standard error of the mean. One-way ANOVA with post-hoc Bonferroni t-tests were performed between treatment conditions. *** p < 0.001, as compared to wt controls.

Figure 6

Reelin-Dab1 signaling is required for Golgi extension. Sections from explants were immunolabeled with the cis-Golgi marker GM130 and imaged. (A) Scatter plots of proximal-to-distal Golgi lengths measured in each experimental condition. L6 neuron Golgi lengths were compact in both r/r Ct-CM and s/s Ct-CM explants as compared to elongated wt and partially rescued (i.e. elongated) Golgi in r/r Rln-CM explants. Rln-CM injection into s/s or wt explants(s/s Rln-CM and wt Rln-CM) had no effect on Golgi length in comparison to Ct-CM injections. (B) Bar graph representation of measured Golgi lengths. Dashed lines in scatter plots represent mean values; error bars denote standard error of the mean. Golgi lengths were compared across images acquired from 18 wt explants, 7 r/r Ct-CM explants, 5 s/s Ct-CM explants, 9 r/r Rln-CM explants, 5 s/s Rln-CM explants, and 6 wt Rln-CM explants. One-way ANOVA with post-hoc Bonferroni t-tests were performed between treatment conditions. *** p < 0.001, as compared to wt controls.

Figure 7

Reelin-Dab1 signaling in L6 neuron orientation and dendritogenesis. In the absence of Reelin signaling (r/r explant, lower panel), L6 neurons had the same total neurite arbor length as wild type (wt explant, upper panel). However, neurons in the mutant had more primary neurites that were longer and less branched than control primary neurites. Additionally, the primary neurites and neuronal somata in the mutant were oriented tangentially rather than radially. Notably, L6 neuron neurites failed to penetrate the MZ in the mutant despite their normal length. These observed defects are hypothesized to be consistent with the operation of an inhibitory signal (purple zone) present in the MZ that is counteracted by Reelin signaling.