Previously, we and others reported that Zfp503 is highly expressed in the LGE (Chang et al., 2004; Ko et al., 2013; Su et al., 2022), but the temporal expression of Zfp503 during striatal development is not fully understood. Firstly, we performed double immunostaining of ZFP503 with the pan-striatal MSN marker BCL11B at E14.5 and E18.5 (Figures 1A,B). Our results showed that approximately 90% of BCL11B positive cells are ZFP503/BCL11B double-positive cells in the subventricular zone (SVZ), but the percentage was decreased to 70% in the mantle zone (MZ) at E14.5 (Figure 1C). Interestingly, nearly all BCL11B-positive cells were co-labelled with ZFP503 in the SVZ/MZ at E18.5 (Figures 1B,C). Next, we performed immunostaining in the developing mouse brain with the ZFP503 antibody. The expression of ZFP503 was detected in the LGE (striatal primordia) at E12.5 and E14.5 (Supplementary Figures S1A,B). A similar expression of ZFP503 was also observed in the LGE at E16.5 (Supplementary Figure S1C). At perinatal period, ZFP503 was present in the entire striatum, including the caudoputamen (CP), nucleus accumbens (NAc) and olfactory tubercle (OT) (Supplementary Figures S1D,E). These results indicated that Zfp503 is highly expressed in the LGE during striatal MSN development.

To examine the role of Zfp503 in regulating the fate of MSNs, we used two Cre lines (Sp9-Cre and Dlx2-Cre) to delete the expression of Zfp503 in the basal ganglia. The phenotype of the Zfp503 mutant mice was analyzed at E18.5. We found that there was no significant difference in BCL11B expression between control mice and Zfp503-CKO mice (Figures 1D,E). These results were further confirmed by our RNA-seq data (Supplementary Table 1) and were consistent with previous research (Chen et al., 2020). Drd1 and Tac1, which are specifically expressed in D1 MSNs were dramatically decreased in the striatum of Zfp503-CKO mice compared to controls. In contrast, Drd2 and Penk, which are specifically expressed in D2 MSNs, were markedly increased in Zfp503-CKO mice (Figures 1D,E). Next, we analysed the phenotype of the Sp9-Cre; Zfp503 ^F/F (Zfp503-SCKO) mice at the transcriptional level. RNA-seq of the entire LGE was performed at E18.5. Again, our results showed that D1 MSN marker genes (Drd1, Tac1, Isl1, Sox8, and Pdyn) were significantly decreased. In contrast, D2 MSN marker genes (Drd2, Penk, Six3, Grik3, and Gucy1a3) were significantly increased (Supplementary Table 1). Taken together, our results indicated that striatal D1 MSNs are converted into D2 MSNs in the absence of Zfp503.

To investigate the mechanism by which Zfp503 control fate switching between D1 and D2 MSNs, scRNA-seq experiments were performed to analyze LGE progenitors undergoing fate switching (Figure 2A). LGE samples from E14.5 Zfp503-DCKO (Dlx2-Cre; Zfp503 ^F/F , n = 3) mice were dissected, dissociated into single-cell suspensions, and sequenced using the 10X genomics platform. We also used the published scRNA-seq data (Li et al., 2022) of the wild type (WT) LGE at E14.5 from our lab. After removing outlier cells that had a high percentage of ribosomal or mitochondrial genes, 14,001 cells in the WT samples and 12,225 cells in the Zfp503-DCKO samples were used for analysis, with an average of 2,107 genes detected per cell (Figure 2B). Our initial clustering analysis revealed 16 (C0-C15) major cell populations with distinct gene expression patterns (Figure 2B; Supplementary Table 2). We used marker genes to define the identities of the 16 clusters more precisely, which resulted in eight discrete populations: RGs, APs, BPs, IPs, Pre-MSNs, imm-D1 MSNs, D1 MSNs, and D2 MSNs (Figure 2C; Supplementary Table 2). According to analysis of marker genes, we determined the identity of the remaining cell clusters: C7: striatal interneurons (Lhx6, Lhx8, Sst, and Npy); C9: cortical projection neurons (Neurod6, Tbr1 and Cux2); C10: intercalated cells (Zic1 and Resp18); C11: ependymal cells (S100a11, Anxa2 and Cryab); C12: endothelial cells (Cdh5 and Cd34); C13: blood cells (Alas2); C14: microglial cells (Bcl2a1b, P2ry12 and Trem2); and C15: oligodendrocyte cells (Olig1/2 and Sox10).

Next, we systematically characterized the gene features of the D1 and D2 MSNs. To date, the generation sequence of D1 and D2 MSNs during striatal development is not fully understood. Interestingly, we found very low expression levels of Drd1 and Drd2 suggesting that the D1 and D2 MSNs were not fully matured at E14.5 (Figure 2D). However, D1 MSN marker genes (Tac1 and Ebf1) were highly expressed. In contrast, D2 MSN marker genes (Penk and Adora2a) were weakly expressed (Figure 2D). Furthermore, the pan-striatal MSN markers (Rarb, Rxrg and Foxp1) were highly expressed in D1 MSNs (C6) but not in D2 MSNs (C8) (Figure 2D). These results indicated that D1 MSN maturation seems faster than D2 MSN maturation. Recent studies indicated that Isl1, Zfhx3 and Zfp503 are expressed in the immature D1 MSNs (Ko et al., 2013; Zhang et al., 2019; Song et al., 2021; Ehrman et al., 2022; Li et al., 2022). Interestingly, we found that Pou3f1 and Aldh1a3 are specifically expressed in immature D1 MSNs (C3). Recently studies (Bocchi et al., 2022; Su-Feher et al., 2022) reported that Pou3f1 and Aldh1a3 are enriched in immature D1 MSNs. Thus, our findings and previous studies suggested that Pou3f1 and Aldh1a3 represent new markers for immature D1 MSNs.

Next, we explored the mechanism underlying fate switching between D1 MSNs and D2 MSNs in Zfp503-DCKO (Dlx2-Cre; Zfp503 ^F/F ) mice. Thus, we focused our further analysis on the major cell types in the LGE at E14.5 (Figure 3A). The radial glia cells (RGs) that expressed Aldh1l1 and Aldoc were not significantly altered between Zfp503-DCKO mice and WT mice (Figures 3B,C,L). Sox2 is highly expressed in the progenitors of the LGE and downregulated in immature neurons. Surprisingly, our results showed that Sox2 expression was upregulated in most cell types, especially in immature MSNs in Zfp503-DCKO mice (Figures 3D,L). Recent studies reported that Ascl1 and Vax1 are highly expressed in basal progenitors (BPs) (Li et al., 2021; Wen et al., 2021; Bocchi et al., 2022), and a study showed that Gadd45g is a direct downstream target of Ascl1 (Huang et al., 2010). These genes were increased in the Zfp503-DCKO mice (Figures 3E–G,L). We also found that the proliferating cells expressing Ki67 (Mki67) and Top2a were increased in the Zfp503-DCKO mice compared to the WT mice (Figures 3H,I,L). Finally, our results showed that Dlx2 and its direct downstream target Gad2 were increased in MSNs (Figures 3J–L). In addition, we performed immunostaining of the progenitor cell markers, including PCNA, ASCL1 and KI67 at E14.5 to further support our findings. Our results showed that the number of the PCNA-, ASCL1-, and KI67-positive cells are increased in the Zfp503-DCKO mice compared to control mice (Supplementary Figures S2A–I). Taken together, these findings indicated that the differentiation of progenitors is blocked in Zfp503-DCKO mice.

To further confirm our results regarding D1 and D2 MSN identity at the cell population level, we begin to analyze D1 and D2 MSN clusters in our scRNA-seq data. Previous work and our studies showed that Pou3f1, Isl1 and Zfhx3 are expressed in immature D1 MSNs (Ehrman et al., 2013; Lu et al., 2014; Zhang et al., 2019). We found that the expression of these genes in the immature D1 MSN cluster was not significantly changed between WT mice and Zfp503-DCKO mice (Figures 4A–C,O). However, Ebf1, Ikzf1, Tac1, Rarb, and Foxp2, which were highly expressed in D1 MSNs were reduced in the D1 MSN cluster of Zfp503-DCKO mice compared to WT mice (Figures 4D–H,O). Our previous studies showed that the TFs Sp9 and Six3 are required for the differentiation of D2 MSNs (Zhang et al., 2016; Xu et al., 2018; Song et al., 2021). In the absence of Zfp503, these genes were upregulated in D2 MSNs (Figures 4I,J). Furthermore, Tle4 and Gucy1a3 tend to be expressed in D2 MSNs, and our results showed that these genes were upregulated in D2 MSNs and immature D1 MSNs in Zfp503-DCKO mice (Figures 4K–L). In addition, the specific D2 MSN markers Penk and Adora2a were upregulated in Zfp503-DCKO mice compared to WT mice (Figures 4M–O). Again, our findings indicated that D1 MSNs undergo fate switching into D2 MSNs at the cell population level in Zfp503-DCKO mice.

To gain insights into the mechanism by which Zfp503 regulates the fate switching of striatal MSNs, we performed immunostaining at E16.5 and E18.5. Previous studies reported that TF Sox2 was mainly expressed in the neural precursor cells and regulated multiple aspects of telencephalic development (Cimadamore et al., 2013; Feng et al., 2013; Feng and Wen, 2015). Our previous studies demonstrated that the TFs Sp8/9 and Six3 are required for the differentiation of D2 MSNs (Zhang et al., 2016; Xu et al., 2018; Song et al., 2021). Here, we found that SOX2 expression was significantly upregulated in the whole striatum at E16.5 and E18.5 in Zfp503-CKO mice compared to controls (Figures 5A–D, 6A–D). Considering that striatal MSNs express relatively mature markers, such as Drd2 and Penk (Figure 1D), we speculate that the striatal MSNs is not well differentiated due to blockade of differentiation at the progenitor stage. Moreover, our results showed that the TFs Sp8/9 were dramatically increased in the striatum at E16.5 and E18.5 in the Zfp503-CKO mice compared to controls (Figures 5E–L, 6E–L). Interestingly, Sp8 expression was significantly increased at E16.5 (Figures 5I–L) and E14.5 (data not shown) in both Cre lines, but the expression of Sp8 was not altered at E18.5 in the Dlx2-Cre line (Figures 6I–L). Due to the Dlx2-Cre deleted Zfp503 expression mainly in LGE radial glial cells and the Sp9-Cre deleted Zfp503 expression mainly in progenitors which were located in the LGE SVZ. Therefore, we speculate that Zfp503 may exert its function in radial glial cells. It may be that Zfp503 promotes cell cycle exit of radial glial cells, which leads to Sp8 expression showing the different phenotypes at different time points. In addition, Six3, a direct downstream target of Sp8/9 was significantly increased in Zfp503-CKO mice at E16.5 and E18.5 (Figures 5M–P, 6M–P). Taken together, these findings suggest that Zfp503 represses the D2 MSN identity by repressing the expression of Sp8/9.

In addition to investigating the effects of Zfp503 loss-of-function, we also explored the effects of Zfp503 gain-of-function in striatal MSNs. First, we generated a Zfp503 overexpression allele (Zfp503-OE). A CAG promoter followed by a Zfp503 CDS and three HA-tag sequences was inserted into the Rosa-26 site with the CRISPR/Cas9 strategy (Figure 7A). Then, we crossed Zfp503-OE mice with Dlx2-Cre mice to obtain Zfp503 overexpressing mice (Dlx2-Cre; Zfp503-OE). Dlx2-Cre; Zfp503-OE embryos showed robust Ha-tag signals in Dlx2-expressing regions along the dorso-ventral (DV) axis, including the septum, striatum, and cortex interneuron (Figure 7B). To study the function of Zfp503 in regulating the cell identity of striatal MSNs, we used in situ hybridization to detect the expression of D1 MSN and D2 MSN markers. Consistent with our expectations, the expression of Isl1, Tac1, and Ebf1 was increased in the Zfp503 overexpressing mice ( Figures 7C–E,H ). In contrast, TFs Sp9 and Six3, which control the differentiation of D2 MSNs were reduced in Zfp503 overexpressing mice compared to control mice (Figures 7F–H). In addition, the RNA-seq data showed that the D1 MSN identity was enhanced and the D2 MSN identity was compromised in the Dlx5/6-Cre; Zfp503-OE mice compared to the control mice (Supplementary Table 3). Overall, our findings suggested that Zfp503 promotes the D1 MSN identity and represses the D2 MSN identity.

Understanding the mechanisms underlying neuronal fate acquisition is a major challenge in developmental neurobiology, because there are many distinct regions and different progenitors in the mammalian telencephalon, particularly in the subpallium (Yun et al., 2001; Corbin et al., 2003; Stenman et al., 2003; Flames et al., 2007; Wen et al., 2021). More recent evidence has shown that TF Zfp503 expression is enriched in the ventral parts of the vLGE (Chen et al., 2020), while the TF Six3 expression preferentially present in the dorsal parts of the vLGE at an early stage (Xu et al., 2018). Furthermore, the differentiation of D1 MSNs and D2 MSNs is completely blocked in Zfp503-CKO and Six3-CKO mice, respectively (Xu et al., 2018; Soleilhavoup et al., 2020). These results support that D1 MSNs are generated from the ventral part of the vLGE, in contrast to D2 MSNs which are derived from the dorsal part of the vLGE.

Retinoic acid (RA) has been implicated as an extrinsic signal regulating the differentiation of distinct cell types, including specific neuronal subtypes (Toresson et al., 1999; Chatzi et al., 2011; Cunningham and Duester, 2015; Rataj-Baniowska et al., 2015; Podleśny-Drabiniok et al., 2017). RA is derived from vitamin A through a two-step enzymatic process, employing retinol dehydrogenase (Rdh10) for the oxidation of retinol to retinaldehyde, and the retinaldehyde dehydrogenases Aldh1a1, Aldh1a2, and Aldh1a3 for the oxidation of retinaldehyde to RA, which then functions as a ligand for nuclear RA receptors (Chatzi et al., 2011). During striatal development, Aldh1a3 is predominantly expressed in LGE (striatal primordium) progenitors and is nearly absent in adjacent structures of the medial ganglionic eminence and cerebral cortex (Toresson et al., 1999; Molotkova et al., 2007). Its expression is regulated by Gsx2 and Meis2 (Waclaw et al., 2004; Su et al., 2022). In fact, the expression of the Aldh1a3 is located in the ventral vLGE (Li et al., 2000; Smith et al., 2001; Waclaw et al., 2004; Chatzi et al., 2011). The roles of RA in the LGE have been investigated, including promoting striatal neurogenesis and neuronal differentiation (Molotkova et al., 2007; Liao et al., 2008; Urban et al., 2010; Al Tanoury et al., 2014; Rataj-Baniowska et al., 2015). The absence of Aldh1a3 leads to abnormal differentiation of striatal MSNs (Chatzi et al., 2011). Interestingly, RA can promote the expression of Zfp503 in the LGE at the early stage of striatal development (Urban et al., 2010). These studies support that RA can induce Zfp503 expression in ventral vLGE progenitors at an early stage. Therefore, we speculate that the progenitors located in the ventral vLGE tend to give rise to D1 MSNs. In contrast, the progenitors located in the dorsal vLGE tend to generate D2 MSNs.

The Forkhead box P2 (FOXP2) has been identified as a gene related to neurodevelopmental disorders, and it was also the first gene to be associated with vocal functions in songbirds and rodents (Bacon and Rappold, 2012; Chiu et al., 2014). Indeed, mutations in the FOXP2 gene have been identified in patients with a severe speech and language disorders (Lai et al., 2001). Naturally, Foxp2 is enriched in the striatum and controls the differentiation of striatal MSNs (Lai et al., 2001; Chiu et al., 2014; Chen et al., 2016; Fong et al., 2018; van Rhijn et al., 2018). The mechanisms, however, of regulating Foxp2 expression in the striatum are largely unknown. More recent evidence supports that Foxp2 is highly expressed in D1 MSNs and weakly expressed in the D2 MSNs (Liu et al., 2018; Tinterri et al., 2018; van Rhijn et al., 2018; Wei et al., 2019). In this study, we found that the expression of Foxp2 was significantly reduced in Zfp503-DCKO mice (Figure 4H; Supplementary Table 1). In contrast, the expression of Six3 was present in the whole striatum (Figures 6M–O). Six3 is known as a key factor for the generation of D2 MSNs (Xu et al., 2018; Song et al., 2021). Thus, we speculated that Six3 represses the expression of Foxp2 in D2 MSNs. Actually, in our recent works and unpublished data, we found that Foxp2 expression is upregulated in Six3 mutant mice and downregulated in Six3 overexpressing mice (Xu et al., 2018; Song et al., 2021). Therefore, our findings suggested that Foxp2 expression is closely regulated by Six3 in D2 MSNs.

Overall, our findings have demonstrated that Zfp503 is critical for the fate acquisition of D1 versus D2 MSNs during striatal development. Zfp503 appears to repress the expression of the TF Sp8/9 in precursor cells to control the fate of D1 MSNs. The results of this work will deepen our understanding of fate determination in striatal MSNs and open up new opportunities for translational investigations of degenerative diseases, such as PD and HD.

All experiments conducted in this study were in accordance with guidelines from Fudan University, China. Zfp503 ^F/+ (Su et al., 2022), Sp9-Cre (Zhang et al., 2016), and Dlx5/6-CIE (Liu et al., 2018) mice have been previously described. Dlx2-Cre mice were generated via the CRISPR/Cas9 strategy. IRES elements and Cre sequences were knocked into the coding region of exon 3 of Dlx2. Genotyping of the Dlx2-Cre constitutive knock-in (KI) allele was performed by PCR using the following primers:

F1: 5′- CCT​CGG​CCT​TTC​TGG​GAA​ACT​AC-3′;

R1: 5′- CTT​GCA​GGT​ACA​GGA​GGT​AGT​CC-3′;

R2: 5′- TTG​GCA​CTA​AAG​GAT​CCC​ACG​AG-3′. These primers yielded bands of 390 and 501 bp for the WT and knock-in (KI) alleles, respectively. Rosa26-Zfp503-OE/+ mice that conditionally overexpressed Zfp503 were generated by CRISPR/Cas9 techniques (Figure 7A). An expression vector containing CAG-promoter-Flox-PolyA STOP-Flox-Kozak-mouse Zfp503 CDS-3 XHA tag-PolyA was knocked into the Rosa26 locus. With this strategy, the Zfp503-HA tag fusion protein was continuously expressed in CRE recombinase-positive cells. In cells after CRE recombination, both the HA tag and the Zfp503 protein were specifically detected. Genotyping of the Zfp503-OE allele was performed by PCR using the following primers:

WT-F3: 5′-CAC​TTG​CTC​TCC​CAA​AGT​CGC​TC-3′;

WT-R3: 5′-ATA​CTC​CGA​GGC​GGA​TCA​CAA-3′;

OE-F5: 5′-GCA​TCT​GAC​TTC​TGG​CTA​ATA​AAG-3′. These primers yield bands of 453 and 633 bp for the WT and overexpression alleles, respectively. All mice were maintained in a mixed genetic background of C57BL/6J and CD1. The day of vaginal plug detection was considered E0.5, and the day of birth was defined as P0. Both male and female mice were used in all experiments.

Tissue preparation was performed as previously described (Yang et al., 2021).

In this study, 6-μm and 12-μm thick frozen sections were used for immunostaining. The sections were washed with 0.05 M Tris buffered saline (TBS) for 10 min, (If necessary, the sections were immersed in antigenic repair solution(pH = 6.0) and boiled before the Triton step, and cooled naturally to room temperature (RT) incubated in Triton X-100 (0.5% in 0.05 M TBS) for 30 min at RT, and then incubated with blocking solution (10% donkey serum + 0.5%; Triton-X-100 in 0.05 M TBS, pH 7.2) for 2 h at RT. The primary antibodies were diluted in 10% donkey serum blocking solution, incubated overnight at 4°C and then rinsed three times with 0.05 M TBS. Secondary antibodies matching the appropriate species were added and incubated for 2-4 h at RT. Fluorescently stained sections were then washed three times with 0.05 M TBS. This was followed by 4′,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich, 200 ng/ml) staining for 5 min, and the sections were then cover-slipped with Gel/Mount (Biomeda).

In situ hybridization (ISH) was performed on 20-μm cryosections using digoxigenin riboprobes as previously described (Su et al., 2022). Probes were made from P0 WT mouse brain cDNA amplified by PCR.

LGE samples of E16.5 Zfp503-OE, E18.5 Zfp503-CKO and littermate control mice were dissected (n > = 3 for each group). Total RNA was purified with a Mini RNA Isolation Kit (Zymo) followed by library generation according to the manufacturer’s protocol (Illumina TruSeq Stranded Total RNA Library Prep Kit with Ribo-Zero Mouse). Fragment size distribution was assessed using a Bioanalyzer 2100. The concentration of the libraries was measured using a Kapa Library Quantification Kit. The purified libraries were sequenced on a HiSeq 4000 platform. Gene expression levels were reported as fragments per kilobase of exon per million fragments mapped (FPKM) values. Genes with a p value < 0.05 were considered to be differentially expressed.

Embryonic mouse (E14.5) brains were quickly collected and placed in HBSS. The LGE samples were dissected and incubated in 1 mg/ml papain in HBSS for 20 min at 37°C. The tissues were gently dissociated into a single-cell suspension by pipetting. Cells were centrifuged and washed twice, filtered through Flowmi Tip 40 μM strainers, and resuspended in HBSS+ 0.04% BSA. Cell viability was assessed by trypan blue exclusion.

The Chromium Droplet-based Sequencing Platform (10X Genomics) was used to generate scRNA-seq libraries, following the manufacturer’s instructions (manual document part number: CG00052 Rev C). The cDNA libraries were purified, quantified using an Agilent 2100 Bioanalyzer, and sequenced on an Illumina HiSeq4000. High quality sequences (clean reads) were obtained by removing low quality sequences and joints. The clean reads were then processed with Cell Ranger software to obtain quantitative information on gene expression. Genes expressed in < 3 cells and cells with < 750 detected genes were filtered out. Cells with >10% mitochondrial genes were also filtered out. The global-scaling normalization method “Log Normalize” was applied to the raw read counts generated by 10X Cell Ranger to normalize the gene expression measurements for each cell based on the total expression. The log-transformed normalized single-cell expression values were used for differential expression tests. Potential sources of variation, including technical noise, batch effects, and biological sources of variation such as cell-cycle stage, were removed to improve downstream dimensionality reduction and clustering. We regressed gene expression on the number of detected molecules per cell and the cell-cycle stage score. The scaled z-scored residuals were used for principal component analysis (PCA). Statistically significant principal components determined by a resampling test were kept for uniform manifold approximation and projection (UMAP) analysis. Differentially-expressed genes (DEGs) among clusters were identified by comparing cells in each cluster against all other cells with the likelihood-ratio test. Gene A was defined as a biomarker of cluster X if it was detected in ≥25% cells, and had an adjusted p-value < 5%, and fold change ≥2 between cells of cluster X and all other cells. All these analyses were performed in Seurat v3.2 (https://satijalab.org/seurat/).

Images for quantitative analyses were acquired using an Olympus VS 120 microscope with a 20X objective. Bright-field images were acquired using an Olympus VS 120 microscope with a 10X objective. Images were merged, cropped and optimized in Adobe Photoshop CC without distorting the original information. Analyses were performed using GraphPad Prism 6.0, Microsoft Excel and the R language. Unpaired two-tailed t-test or one-way ANOVA followed by the Tukey–Kramer post-hoc test was used to determine statistical significance. All quantification results are presented as the mean ± SEM. Differences with p-values < 0.05 were considered significant.

For quantification of BCL11B⁺ and ZFP503⁺ cells in the striatum of Wild type at E14.5 and E18.5, four anatomically matched 6-μm thick coronal sections were selected (n = 3 mice per group). We counted BCL11B⁺ and ZFP503⁺ cells in the striatum under a 20X objective lens. The striatum was delineated by DAPI staining. The numbers of BCL11B⁺ and ZFP503⁺ cells per section for each LGE SVZ were presented as the relative to these in the control group.

For quantification of BCL11B⁺ cells in the striatum of control and two Zfp503-CKO mice at E18.5, four anatomically matched 12-μm thick coronal sections were selected (n = 3 mice per group). We counted BCL11B⁺ cells in the striatum under a 20X objective lens. The striatum was delineated by DAPI staining. The numbers of BCL11B⁺ cells per section for each striatum were presented as the relative to these in the control group.

For quantification of Drd1 ⁺, Tac1 ⁺ , Penk ⁺ , and Drd2 ⁺ cells in the striatum of control and two Zfp503-CKO mice at E18.5, four anatomically matched 20-μm thick coronal sections were selected (n = 3 mice per group). We counted Drd1 ⁺ , Tac1 ⁺ , Penk ⁺ , and Drd2 ⁺ cells in the striatum under a 10X objective lens. The numbers of Drd1 ⁺ , Tac1 ⁺ , Penk ⁺ , and Drd2 ⁺ cells per section for each striatum were presented as the relative to these in the control group.

For quantification of SIX3⁺, SOX2⁺, and SP9⁺ cells in the striatum of control and two Zfp503-CKO mice at E16.5 and E18.5, four anatomically matched 12-μm thick coronal sections were selected (n = 3 mice per group). We used the histogram tool in Adobe Photoshop CC to calculate the fluorescence intensity of SIX3⁺, SOX2⁺, and SP9⁺ cells in the striatum. Data were presented in proportion to the control group.

For quantification of SP8⁺ cells in the striatum of control and two Zfp503-CKO mice at E16.5 and E18.5, four anatomically matched 12-μm thick coronal sections were selected (n = 3 mice per group). We counted SP8⁺ cells in the striatum under a 20X objective lens. The striatum was delineated by DAPI staining. The numbers of SP8⁺ cells per section for each striatum were presented as the relative to these in the control group.

For quantification of Ebf1 ⁺, Isl1 ⁺, Six3 ⁺, and Sp9 ⁺ cells in the striatum of control and Dlx2-Cre; Zfp503-OE mice at E16.5, four anatomically matched 12-μm thick coronal sections were selected (n = 3 mice per group). We used the Adobe Photoshop CC to invert the image, and then used histogram tool in Adobe Photoshop CC to calculate the fluorescence intensity of Ebf1 ⁺ , Isl1 ⁺ , Six3 ⁺ , and Sp9 ⁺ cells in the striatum. Data were presented in proportion to the control group.

For quantification of Tac1 ⁺ cells in the striatum of control and Dlx2-Cre; Zfp503-OE mice at E16.5, four anatomically matched 12-μm thick coronal sections were selected (n = 3 mice per group). We counted Tac1 ⁺ cells in the striatum under a 10X objective lens. The numbers of Tac1 ⁺ cells per section for each striatum were presented as the relative to these in the control group.

For quantification of PCNA⁺, KI67⁺, and ASCL1⁺ cells in the striatum of control and Zfp503-DCKO mice at E14.5, four anatomically matched 12-μm thick coronal sections were selected (n = 3 mice per group). We used the histogram tool in Adobe Photoshop CC to calculate the fluorescence intensity of PCNA⁺, KI67⁺, and ASCL1⁺ cells in the striatum. The striatum was delineated by DAPI staining. Data were presented in proportion to the control group.