3.1Increase of cell proliferation in the Vax1-/- subpallium V-SVZ
The Ventral Homeodomain Protein 1 (VAX1) is highly expressed in the subpallium progenitors from embryonic day 12.5 (E12.5) to adult would suggest Vax1 plays an important role in forebrain development (See Supplementary fig. S1a-f, Additional File 1). To investigate whether Vax1 regulates cell proliferation, we first performed 30-min BrdU pulse labeling at E16.5 and BrdU immunofluorescence staining. The number of BrdU+ cells in the Vax1-/- LGE gradually increased in a rostral to caudal pattern, compared to controls (Fig. 1a-c’, j). This means the number of cells in S-phase of the cell cycle was increased after Vax1 deletion. Since Cdca7 is a cell division cycle associated gene and Ccnd2 promotes cell cycle progression from G1 to S phase, we then checked these genes at E16.5. In situ RNA hybridization of Cdca7 and Ccnd2 showed that Cdca7-mRNA+ and Ccnd2-mRNA+ cells were also significantly increased (Fig. 1d-i’, k, l). The above results illustrated that there is an increase of cell proliferation in the Vax1-/- subpallium. This is consistent with previous in vitro neurosphere culture results .
3.2Vax1 promotes neuronal differentiation in the subpallium V-SVZ
The bHLH transcription factor, Ascl1, is heavily expressed in ventral forebrain progenitors, and Ascl1 is crucial for neurogenesis . At E16.5, the number of Ascl1+ cells in the Vax1-/- subpallium VZ (ventricular zone) was significantly increased, compared to controls (Fig. 2a-c’, m). Dlx1 and Dlx2 are mainly expressed in the VZ and SVZ of the subpallium, Dlx5 appears later than Dlx1/2, and is expressed in more differentiated neurons in the SVZ (subventricular zone) and MZ (mantle zone). Our results showed that Dlx1+, Dlx2+ and Dlx5+ cells in the Vax1-/- subpallium were increased markedly (Fig. 2d-l’, m). Accumulation of Ascl1+, Dlx1+, Dlx2+ and Dlx5+ immature neurons in the Vax1-/- VZ and SVZ suggests that neural progenitors in the subpallium could not normally exit the cycle and differentiate. Therefore, Vax1 promotes neuronal differentiation in the subpallium VZ-SVZ.
3.3 The expansion of the dLGE domain in the Vax1-/- mice
The progenitors in the dLGE mainly give rise to olfactory bulb interneurons. Strong expression of Gsx2 in the dLGE promotes the activation and lineage progression of OB interneuron progenitors. Sp8 is expressed in migrating neuroblasts in the embryonic dLGE, postnatal SVZ and most OB interneurons, and the OB interneuron defects in conditional inactivation of Sp8 mice . Sp8 regulates the expression of Tshz1 and Prokr2 . Prokr2 and Tshz1 are expressed in the SVZ-RMS-OB, severe tangential, and radial migration defects of neuroblasts in Prokr2-KO mice [24, 29]. Etv1+ cells in the dLGE and postnatal SVZ travel through the RMS into the OB where they mature into OB interneurons. At E16.5, we found that these dLGE markers (Gsx2, Etv1, Sp8, Tshz1 and Prokr2) were significantly increased in the Vax1-/- mice (Fig. 3a-o’). Gsx2 was heavily expressed in progenitors of the dLGE and its expression showed a ventral-low to dorsal-high gradient in the subpallium of the controls whereas, more GSX2+ cells were seen in the VZ/SVZ of the vLGE and MGE. The obvious ventral to dorsal gradient in GSX2+ cell numbers in the Vax1-/- mice disappeared (Fig. 3a-c’). Even more remarkably, Prokr2, Tshz1, Etv1, and Sp8 showed an increased gradient expression in the dLGE (Fig. 3d-o’). Gsx2 is expressed in proliferating progenitors, Sp8 is found in some dividing cells, and Prokr2 is observed in post-mitotic immature interneurons, suggesting the blockage of neuronal differentiation in the dLGE after Vax1 gene deletion. The expression of these markers (Gsx2, Sp8 and Prokr2) was also significantly increased in the Vax1-/- mice at P0 (data not shown). Taken together, the progenitors and immature neurons in the dLGE are significantly increased in the Vax1-/- mice.
3.4 The vLGE derived striatal MSNs were reduced in Vax1 mutant mice
We showed above that the dLGE was expanded, next we wanted to know whether the vLGE and MGE-derived neurons changed or not. Isl1 was highly expressed in D1-type MSNs [30, 31]. At E16.5, Isl1+ cells were significantly reduced in the LGE SVZ and MZ of the Vax1-/- mice (Fig. 4a-c’, m). More than 90% striatal MSNs are expressed Foxp1. The number of FOXP1+ cells in the Vax1-/- mouse SVZ and MZ was greatly reduced at E16.5 (Fig. 4d-f’, m). Striatal projection neurons were mainly composed of dopamine D1 receptor (Drd1) expressing MSNs and dopamine D2 receptor (Drd2) expressing MSNs. In situ hybridization of Drd1-mRNA and Drd2-mRNA showed that Drd1+ and Drd2+ MSNs were also reduced in Vax1 mutant mice (Fig. 4g-l’). These results suggest striatal MSNs are decreased in Vax1-/- mice at E16.5.
At P0, we can see a thicker SVZ in Vax1-/- mice from the nuclear staining of DAPI (See Supplementary Fig. S2a, b, Additional File 1). The partial sequences of exon 2 and exon 3 of the Vax1 gene were replaced by a LacZ reporter gene in the Vax1-/- mice. In situ hybridization of LacZ-mRNA showed a marked increase of LacZ+ cells, suggesting that most Vax1 mutant cells were still there (See Supplementary Fig. S2 c, d, Additional File 1). Ki67 is expressed in dividing cells and the KI67+ region was the thicker LacZ+ SVZ region in Vax1-/- mice (See Supplementary Fig. S2e, f, Additional File 1), whereas, FOXP1+ cells were rarely detected (See Supplementary Fig. S2g-h’, Additional File 1). Altogether, these results indicated the blockage of striatal MSN differentiation in the Vax1-/- mice.
Only about 6% Vax1-/- mice can survive to P20, and we were fortunate to have collected two P20 Vax1-/- mouse brains. We then performed DAPI, and in situ hybridization staining of Drd1 and Drd2. We observed the absence of the septum and preoptic area in the Vax1 mutant mice, the change in the shape of the striatum, and that Vax1 mutants had enlarged lateral ventricles (See Supplementary Fig. S3a, a’, Additional File 1). The Vax1 mutants had a low density of Drd1-mRNA+ and Drd2-mRNA+ in the medial striatum (See Supplementary Fig. S3b-c’, Additional File 1). Taken together, our results showed the number of vLGE derived striatal MSNs was reduced in the Vax1-/- mice.
3.5 The MGE domain had defects in the Vax1-/- mice
Since the dLGE domain is enlarged and the vLGE domain is atrophic in the Vax1-/- mice, we wanted to further investigate the change of the MGE domain in the Vax1 mutants. We then performed in situ hybridization staining of Nkx2.1, Lhx6 and Lhx8 at E13.5. The induction and regional pattern formation of MGE are dependent on the transcription factor Nkx2.1 . Nkx2.1 plays an important role in maintaining the specificity and differentiation of MGE progenitors [12, 32]. The Nkx2.1+ MGE domain in the Vax1-/- mice was significantly smaller than that in the controls (Fig. 5a-c’). LIM-homeodomain transcription factors Lhx6, and Lhx8, are expressed in the MGE SVZ, and are crucial for the migration and differentiation of MGE-derived neurons. In Vax1-/- mice, the expression of Lhx6 and Lhx8 was greatly reduced (Fig. 5d-i’). Thus, the interneurons in the neocortex were reduced by 30%-44% after Vax1 gene deletion , possibly due to the smaller MGE domain in the Vax1-/- mice. The progenitors in the MGE not only give rise to cortical interneurons, but also striatal interneurons and globus pallidus neurons. Many transcription factors have been found to be expressed in the globus pallidus, such as: Nkx2.1, Gbx1, Gbx2, Arx, Dlx1, Etv1, Lhx6, and Lhx8 . In order to know clearly the changes in the globus pallidus, we performed in situ hybridization staining of Nkx2.1, Gbx1, and Gbx2 at E12.5 in the Vax1-/- and control mice. We also detected Etv1, Nkx2.1, Lhx6 and Lhx8 at E18.5. There was a reduction in globus pallidus neurons after Vax1 deletion (See Supplementary Fig. S4, Fig.S5, Additional File 1). In summary, the MGE domain gets smaller and MGE-derived neurons are decreased in the Vax1-/- mice.
3.6 RNA-Seq analysis provided further molecular evidence for defects in Vax1 mutant mice
To characterize the molecular changes in Vax1-/- mice, we performed RNA-Seq analysis. Gene expression profiles from the embryonic day (E) 13.5 and E16.5 ganglionic eminence (GE) were analyzed. Changed expression of genes revealed by the RNA-Seq was consistent with our staining results, above. At E13.5, there was a remarkable increased expression of Gsx2 and Sp8 in the GE, with significantly decreased Isl1, Ebf1, Foxp1, Tac1, Otx2, Lhx8, Gbx2, and Gbx1 (Table1). At E16.5, the dLGE markers (Sp8, Etv1, Tshz1) were up-regulated and the vLGE- derived neuronal markers (Foxp1, Foxo1, Ikzf1, Ebf1, Gpr88, Tac1, Sox8, Zfp503, Drd1, Ppp1r1b, Gpr6, Isl1, Zcchc12, Adora2a) were down-regulated, with an upregulation of genes that promote cell cycle and maintenance of the progenitor state (Table2). The combined with the above staining results, compared with controls in the Vax1-/- mice, showed that the dLGE cells invaded ventrally into the vLGE, causing the dLGE region to get larger while the vLGE and MGE regions get smaller.
3.7 Increased expression of Gsx2 in the vLGE and MGE in Vax1 mutant mice
Previous studies showed that Vax1 is a fundamental regulator of ventral identity for retinal ganglion cells and Gsx2 is a key regulator of dLGE identity for OB interneurons [4, 34-37]. In an attempt to understand the relationship between Vax1 and Gsx2, we performed immunofluorescence or in situ RNA hybridization staining of Gsx2 or Vax1 from immediate adjacent 20 μm sections in WT mice at E11.5 and E16.5. We observed Gsx2 expression as a ventral-low to dorsal-high gradient in the subpallium (Fig. 6a-c). In contrast, Vax1 was expressed in a ventral-high to dorsal-low gradient along the subpallium (Fig. 6b-d). The expression pattern of Vax1 is largely complementary to that of Gsx2, supporting our hypothesis that Vax1 may be crucial for subpallium regionalization. To test this, we performed immunofluorescence or in situ RNA hybridization of Gsx2 at early stages, and our results showed that the expression pattern of Gsx2 (ventral-low to dorsal-high) was disrupted in the Vax1-/- mice at E11.5 and E13.5 (Fig. 7a-d). Gsx2 was highly expressed in the whole subpallium VZ and SVZ in the Vax1-/- mice suggesting that Vax1 may be as an inhibitor to suppress the expression of Gsx2 in the subpallium VZ and SVZ. Altogether, these results suggest that Vax1 may regulate subpallium regionalization by repressing Gsx2.
3.8 Vax1inhibits the expression of Gsx2 in the progenitor cells.
To further investigate whether Vax1 inhibits the expression of Gsx2 during forebrain development, we over-expressed Vax1 in the lateral neural progenitors by electroporation and analyzed the impact on Gsx2 expression. A Vax1 expression plasmid (pCAG-Vax1-GFP) or a control vector (pCAG-GFP) was electroporated into the lateral ventricular wall at P0 in the WT mice. Three days later, animals were sacrificed and the number of total GFP+ or GFP/GSX2 double positive cells was measured in the dorsal-lateral region (Fig. 8a-b). Our results showed no difference in the total number of GFP+ cells between control and pCAG-Vax1-GFP mice (Fig. 8c). However, the number of GFP/GSX2 double positive cells showed a significant decrease (5-fold) in the pCAG-Vax1-GFP mice compared to control mice (Fig. 8a’-b’ and d). Thus, we conclude that in progenitor cells, Vax1 has the capacity to act as a negative regulator of Gsx2 expression.