Retinoic acid functions as a key GABAergic differentiation signal in the basal ganglia

Abstract

Although retinoic acid (RA) has been implicated as an extrinsic signal regulating forebrain neurogenesis, the processes regulated by RA signaling remain unclear. Here, analysis of retinaldehyde dehydrogenase mutant mouse embryos lacking RA synthesis demonstrates that RA generated by Raldh3 in the subventricular zone of the basal ganglia is required for GABAergic differentiation, whereas RA generated by Raldh2 in the meninges is unnecessary for development of the adjacent cortex. Neurospheres generated from the lateral ganglionic eminence (LGE), where Raldh3 is highly expressed, produce endogenous RA, which is required for differentiation to GABAergic neurons. In Raldh3⁻/⁻ embryos, LGE progenitors fail to differentiate into either GABAergic striatal projection neurons or GABAergic interneurons migrating to the olfactory bulb and cortex. We describe conditions for RA treatment of human embryonic stem cells that result in efficient differentiation to a heterogeneous population of GABAergic interneurons without the appearance of GABAergic striatal projection neurons, thus providing an in vitro method for generation of GABAergic interneurons for further study. Our observation that endogenous RA is required for generation of LGE-derived GABAergic neurons in the basal ganglia establishes a key role for RA signaling in development of the forebrain.

Figure 1. Endogenous RA activity in the developing forebrain.

Tissues were cultured as explants on a monolayer of F9 RARE-lacZ RA-reporter cells, then stained for β-galactosidase activity. (A–C) At E12.5, cortex and LGE were both negative, whereas eye was positive. (D–F) At E13.5, LGE and eye explants induced RA activity in the reporter cells, but not cortex. (G–I) At E14.5, meninges and LGE induced RA activity while the reporter cells surrounding E14.5 cortical explants remained negative. (J–L) For E14.5 Raldh2−/− (KO) tissues, meninges and cortex were negative, while LGE explants remained positive. (M–O) For E14.5 Raldh3−/− (KO) tissues, meninges explants remained positive, while cortex and LGE were negative. For each genotype and stage, tissues from at least three embryos were analyzed with similar results.

Figure 2. Raldh3 is responsible for RA activity in neurospheres derived from the LGE but cortex-derived neurospheres lack RA activity.

(A–D) Neurospheres generated from the LGE of E14.5 wild-type (WT) embryos exhibit Raldh3 immunoreactivity and they induce RA activity when co-cultured with F9-RARE-lacZ RA-reporter cells (n = 4); neurospheres derived from Raldh3−/− LGE always lacked both Raldh3 detection and RA activity (n = 4). (E–H) Neither Raldh3 nor RA activity were detected in cortex-derived neurospheres from either wild-type or Raldh3−/− (KO) embryos (n = 4).

Figure 3. RA induces GABAergic differentiation in neurosphere-derived cells from the LGE.

(A–B) Neurosphere-expanded wild-type (WT) cells derived from the E14.5 LGE give rise to Gad67-positive and Tuj1-positive cells after 1 wk of differentiation; under the same conditions no Gad67-positive cells were detected in differentiating cultures derived from Raldh3−/− (KO) LGEs. (C–D) Gad67 was not detected in differentiating cultures of E14.5 neurospheres from either wild-type or Raldh3−/− cortex. (E–H) Cells derived from E14.5 Raldh3−/− LGE and cortex neurospheres both express nestin and GFAP at indistinguishable levels from cells derived from wild-type neurospheres. (I–J) RA induces GABAergic differentiation of LGE-derived neurospheres; addition of 100 nM RA for 1 wk in the differentiation medium of expanded LGE cells resulted in an increase of Gad67+/Tuj1+ neurons in both wild-type and Raldh3−/− cultures. (K–L) Cortex-derived neurospheres treated with 100 nM RA exhibited no Gad67 detection. (M) Quantification of the percentage of LGE neurosphere-derived Tuj1-positive neurons also positive for Gad67 showed that in differentiating medium containing only fetal calf serum (FCS) with no added RA, the percentage of Gad67+/Tuj1+ cells derived from Raldh3−/− LGE neurospheres was significantly lower than that of wild-type LGE neurospheres. Following addition of 100 nM RA to the differentiation medium for 1 wk, the majority of Tuj1-positive neurons derived from both wild-type and Raldh3−/− LGE neurospheres were also Gad67-positive. The percentage was calculated by dividing the immunopositive cell number with the total number of DAPI-stained nuclei. Values are listed as mean ± SEM; * p<0.05; ** p<0.01; *** p<0.001.

Figure 4. Loss of RA signaling creates a defect in GABAergic differentiation in the LGE.

Immunofluorescence was performed on E14.5 forebrain coronal sections of wild-type (WT) and Raldh3−/− (KO) embryos. (A–B) Raldh3 immunoreactivity in the subventricular zone of the LGE is lost in the mutant. (C–H) Loss of Raldh3 does not affect detection of neural progenitor marker nestin, radial glial marker RC2, or neuronal marker MAP2. (I–L) Detection of both GABA and Gad67 is greatly reduced in the LGE and septum of Raldh3−/− embryos. Similar results were observed for all mutants analyzed (n = 3). LGE, lateral ganglionic eminence; Spt, septum.

Figure 5. RA is required for GABAergic differentiation of striatal projection neurons and interneurons migrating to the cortex and olfactory bulb.

Immunofluorescence was performed on E18.5 forebrain coronal sections of wild-type (WT) and Raldh3−/− (KO) embryos. (A–B) Raldh3 immunoreactivity in the subventricular zone of the LGE and septum is lost in the mutant. (C–F) In Raldh3−/− embryos, detection of both Gad67 and GABA is greatly reduced in the LGE, striatum, and cortex (note arrows for Gad67); septum is unaffected. (G–L) Loss of Raldh3 does not affect detection of neuronal marker MAP2, striatal projection marker Foxp1, or astrocytic marker GFAP. (M–N) Raldh3−/− embryos exhibit a reduction of Gad67-positive neurons migrating from the LGE to the olfactory bulb when compared to wild-type. All mutants analyzed generated similar results (n = 3). iz, intermediate zone; LGE, lateral ganglionic eminence; OB, olfactory bulb; P, pallidum; Spt, septum.

Figure 6. RA induces GABAergic differentiation of human embryonic stem cells.

Embryoid bodies derived from H9 human embryonic stem cells were treated for 3 d with vehicle, 1 µM RA, or 10 µM RA; then neural rosettes differentiating from these cultures were analyzed immunocytochemically 18 d after RA treatment ended. (A–F) RA treatment increased the number of neurons positive for Gad67 and GABA. (G–I) Tuj1 and DCX double-staining shows that RA treatment increases the number of Tuj1-positive neurons, and essentially all are also positive for DCX, which marks migrating neurons. (J–L) RA treatment decreased the number of neurons positive for the neural progenitor marker Pax6. (M) Quantitative analysis of various cell types in differentiating cultures showed that 1 µM RA or 10 µM RA significantly induced the GABAergic neuron phenotype (Gad67-positive) with a concomitant decrease of Pax6-expressing neural progenitors. Percentages were calculated by dividing the immunopositive cell number with the total number of DAPI-stained nuclei. Data are presented as mean ± SEM; * p<0.01 and ** p<0.001 (untreated versus treated).

Figure 7. Cortical expansion appears normal in embryos lacking RA synthesis in meninges.

(A–B) Raldh2−/− embryos at E14.5 exhibit relatively normal craniofacial development while they display stunted forelimbs (f). (C–D) Tuj1 labeling of postmitotic neuronal layer in E14.5 wild-type (WT) and Raldh2−/− (KO) forebrains. Arrows (same length) demonstrate that Raldh2−/− mutant does not exhibit a change in medial-lateral width of the dorsal ventricular zone (vz; labeled with DAPI) compared to wild-type. (E–J) Double-labeling of E14.5 forebrains with Tuj1 (green) and Ki67 (red; proliferative progenitor layer). Arrows (same length) demonstrate that the Raldh2−/− forebrain exhibits no change in radial expansion of the cortex (cx) or ventricular zone (vz) compared to wild-type. Similar results were observed for all mutants analyzed (n = 3).