Fezf2 Expression Identifies a Multipotent Progenitor For

Neuron Report

Fezf2 Expression Identiﬁes a Multipotent Progenitor for Neocortical Projection Neurons, Astrocytes, and Oligodendrocytes

Chao Guo,1,3 Matthew J. Eckler,1,3 William L. McKenna,1 Gabriel L. McKinsey,2 John L.R. Rubenstein,2 and Bin Chen1,* 1Department of Molecular, Cell and Developmental Biology, University of California, Santa Cruz, CA 95064, USA 2Department of Psychiatry, Neuroscience Program, and the Nina Ireland Laboratory of Developmental Neurobiology, University of California, San Francisco, CA 94158, USA 3These authors contributed equally to this work *Correspondence: [email protected] http://dx.doi.org/10.1016/j.neuron.2013.09.037

SUMMARY 1988), and in vitro culture of single RGCs (Shen et al., 2006) suggests that cortical projection neuron subtype is sequen- Progenitor cells in the cerebral cortex sequentially tially determined by birthdate through progressive lineage re- generate distinct classes of projection neurons. striction of a common RGC (Leone et al., 2008). However, Recent work suggests the cortex may contain the identification of early Cux2-expressing (Cux2+) RGCs, intrinsically fate-restricted progenitors marked by which were reported to be intrinsically specified to generate expression of Cux2. However, the heterogeneity late-born, upper-layer neurons (Franco et al., 2012), calls of the neocortical ventricular zone as well as the into question this decades-old model and raises the possibility that deep-layer projection neurons are similarly generated contribution of lineage-restricted progenitors to the from lineage-restricted progenitors (Franco and Mu¨ ller, 2013; overall cortical neurogenic program remains unclear. Marıń, 2012). Here, we utilize in vivo genetic fate mapping to The transcription factor Fezf2 (also known as Fezl and Zfp312) demonstrate that Fezf2-expressing radial glial cells is expressed in early cortical progenitors and deep-layer neurons (RGCs) exist throughout cortical development and and is critical for the fate specification of subcerebral projection sequentially generate all major projection neuron neurons (Chen et al., 2005a, 2005b; Molyneaux et al., 2005). subtypes and glia. Moreover, we show that the vast In Fezf2/ mice, subcerebral projections are absent and majority of CUX2+ cells in the VZ and SVZ are deep-layer neurons instead switch their identity to become migrating interneurons derived from the subcortical corticothalamic or callosal projection neurons (Chen et al., telencephalon. Examination of the embryonic 2005a, 2005b, 2008; Han et al., 2011; McKenna et al., 2011; cortical progenitor population demonstrates that Molyneaux et al., 2005). Several studies suggest that ectopic expression of Fezf2 in late cortical progenitors (Chen et al., Cux2+ RGCs generate both deep- and upper-layer + 2008) or immature neurons (De la Rossa et al., 2013; Rouaux projection neurons. These results identify Fezf2 and Arlotta, 2013) redirects these cells to differentiate into radial glial cells as a multipotent neocortical pro- subcerebral projection neurons. These results indicate that genitor and suggest that the existence, and mole- expression of Fezf2 in cortical progenitors may be sufficient to cular identity, of laminar-fate-restricted RGCs awaits confer a subcerebral neuron identity, and thus Fezf2-expressing further investigation. (Fezf2+) cortical progenitor cells may be lineage restricted to generate deep-layer neurons (Franco and Mu¨ ller, 2013; Wood- worth et al., 2012). INTRODUCTION To investigate the lineage potential of Fezf2+ progenitor cells, we performed in vivo genetic fate mapping using the Fezf2 The neocortex contains six layers of projection neurons and locus. Here we show that Fezf2+ cortical progenitor cells are glia. Projection neurons in each cortical layer display similar RGCs that exist throughout cortical neurogenesis. Temporal morphologies, axonal projections, and gene expression pat- fate mapping demonstrated that Fezf2+ RGCs sequentially terns (Kwan et al., 2012). During development, cortical projec- generate projection neuron subtypes and glia based upon tion neurons are generated from radial glial cells (RGCs) and the birthdate of these cells. Furthermore, Fezf2+ RGCs gener- basal progenitors in an inside-out pattern such that deep-layer ated upper-layer neurons without expressing detectable neurons are generated first, followed by upper-layer neurons levels of CUX2 protein. Finally, we demonstrate that cells (Molyneaux et al., 2007). Three decades of work based labeled by Cux2-Cre and Cux2-CreERT2 generate both upon transplantation experiments (Desai and McConnell, deep- and upper-layer projection neurons. Collectively, these 2000; McConnell, 1985; McConnell and Kaznowski, 1991), results indicate that Fezf2+ RGCs are a multipotent pro- viral lineage tracing (Luskin et al., 1988; Walsh and Cepko, genitor for neocortical projection neurons, astrocytes, and

Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc. 1167 Neuron Fezf2+ RGCs Are Multilineage Cortical Progenitors

oligodendroctytes and suggest that laminar-fate-restricted RGCs remain to be identiﬁed.

RESULTS

Lineage-Traced Fezf2-Expressing Progenitor Cells Are RGCs We first characterized Fezf2 expression by in situ hybridization. As previously reported (Hirata et al., 2004), we detected Fezf2 expression in early neocortical progenitors (Figure 1B). Interest- ingly, Fezf2 expression in the ventricular zone (VZ) persisted postnatally, long after deep-layer neuron generation has ceased (Figure S1A available online). This was confirmed by GFP expression in Fezf2-GFP bacterial artificial chromosome (BAC) transgenic mice (Gong et al., 2003; Shim et al., 2012), which revealed GFP+ cells in the VZ during late embryonic and early postnatal stages (Figure S1B). To assess the differentiation potential of Fezf2+ progenitor cells, we generated nine independent Fezf2- CreERT2 BAC transgenic mouse lines (Figure 1A). In situ hybridization for Cre and Fezf2 showed that Cre expression was identical to that of endogenous Fezf2 (Figures 1B and 1C; Fig- ures S1C and S1D). Breeding these mice to three different Cre reporter lines (RCE-GFP, R26R-LacZ,orTauR-mGFP; Friedrich and Soriano, 1991; Hippenmeyer et al., 2005; Sousa et al., 2009) revealed that the fused CreERT2 protein was tightly regu- lated by tamoxifen (Figures S1E–S1F0). Although Cre mRNA was expressed in deep-layer neurons (Figures 1C and S1D), we observed tamoxifen-induced recombination in these neurons with only the TauR-mGFP reporter (Figures S1H–S1I). No recombination was observed in postmitotic neurons upon tamoxifen administration with the Rosa26R-LacZ or RCE-GFP reporters (Figures S1G, S1J, and S1J0). Critically, this allowed us to perform lineage-tracing experiments for Fezf2+ cortical progenitor cells using the RCE-GFP reporter without the ambiguity caused by Cre-mediated recombination in postmitotic neurons. Examination of Fezf2-CreERT2; RCE-GFP mice after tamoxifen induction revealed that recombination specifically marked Fezf2+ RGCs (Figures 1D–1M). Twenty-four hours after tamoxifen administration, 80% of GFP+ cells expressed the RGC marker SOX2, whereas 10% of GFP+ cells expressed the basal progenitor marker TBR2 (Figures 1D, 1E, 1G, 1H, and 1M). The majority of GFP+ cells were located in the VZ; many had both api- cal and basal processes and divided at the ventricular surface (Figures 1G–1I), all of which are characteristic of RGCs. The few TBR2+GFP+ cells were likely basal progenitors newly gener- Fezf2 Figure 1. -Expressing Progenitors Are RGCs ated from Fezf2+ RGCs. Supporting this, 3 days after an embry- (A) Strategy for generation of Fezf2-CreERT2 mice. onic day 13.5 (E13.5) tamoxifen administration (TM at E13.5; (B and C) In situ hybridization for Fezf2 (B) and Cre (C) at E13.5. + + (D–F) Low-magnification images of GFP+ cells in the cortex of Fezf2-CreERT2; E16.5), the fraction of TBR2 GFP cells increased to 39% (Fig- RCE-GFP mice following CRE-mediated recombination. ures 1F and 1J–1M). These results indicate that lineage-traced (G and H) Twenty-four hours after tamoxifen induction, most GFP+ cells ex- Fezf2+ progenitors are RGCs. pressed SOX2 (78% ± 3%) (G), and few cells expressed TBR2 (10% ± 2%) (H). (I) GFP+ cells dividing at the ventricular surface. (J–L) In TM at E13.5; E16.5 brains, Fezf2+ RGCs gave rise to basal progenitors. (M) Quantification of the percentages of GFP+SOX2+ RGCs and GFP+TBR2+ GFP was expressed in both ventricular zone (VZ) and subventricular zone basal progenitors among all the GFP+ cells ± SEM. **p < 0.005, ***p < 0.0001; (SVZ) progenitors (arrowhead), migrating neurons in the intermediate zone CP, cortical plate; Ctx, cerebral cortex; IZ, intermediate zone; LV, lateral (arrow), and cortical neurons (asterisk) (J). GFP+ cells expressed SOX2 ventricle; TM, tamoxifen; SVZ, subventricular zone; VZ, ventricular zone. (54% ± 3%) and showed typical RGC morphology (K). Many GFP+ cells Scale bars: 250 mm (C–F), 10 mm (E, inset), and 25 mm (H–J and L). See also expressed TBR2 (39% ± 2%) (L). Figure S1.

1168 Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc. Neuron Fezf2+ RGCs Are Multilineage Cortical Progenitors

Figure 2. Fezf2+ RGCs Sequentially Generate Projection Neurons and Glia (A–F) Immunohistochemical analysis of TM at E12.5; P21 brains. GFP+ cells were present throughout the cortex (A and B). GFP+ cells expressed CTIP2 (C), SATB2 (D), GFAP (E), or OLIG2 (F). (G–K) Immunohistochemistry on brain sections from TM at E14.5; P21 mice. GFP+ neurons were mostly in layers 2–4 (G and H). GFP+ cells expressed SATB2 (I), GFAP (J), or OLIG2 (K). (L) Percentage of GFP+ neurons in deep and upper layers ± SEM. (M–P) Immunohistochemical analysis of brains from TM at E18.5; P21 mice. GFP+ cells included astrocytes (O) and oligodendrocytes (P). (Q) Percentages of GFP+ cells that were neurons, astrocytes, or oligodendrocytes ± SEM. *p < 0.05, **p < 0.005, ***p < 0.0001. Ctx, cerebral cortex; Hip, hip- pocampus; Thal, thalamus; TM, tamoxifen; WM, white matter. Scale bars: 500 mm (A, G, and M), 250 mm (B, H, and N), and 25 mm (F, K, and P). See also Figure S2.

Fezf2+ RGCs Sequentially Generate Deep-Layer and Indeed, GFP-labeled axon tracts demonstrated that Fezf2+ Upper-Layer Cortical Projection Neurons and Glia RGCs generated callosal, subcerebral, and corticothalamic To assess the lineage potential of Fezf2+ RGCs, we administered projection neurons (Figures S2D–S2F). In addition to projection tamoxifen to Fezf2-CreERT2; RCE-GFP mice at different neurons, 15% of GFP+ cells showed an astrocytic morphology embryonic stages. In TM at E12.5; P21 brains, GFP+ cells were and expressed GFAP (Figures 2E and 2Q), whereas 10% of detected throughout the cortical plate and included cortical GFP+ cells expressed the oligodendrocyte marker OLIG2 projection neurons (75%) and glia (25%; Figures 2A–2F, 2Q, (Figures 2F, 2Q, and S2A). These results demonstrate that at and S2A). GFP+ neurons were present in both deep (36%) and E12.5, Fezf2+ RGCs are multipotent. upper (64%) layers (Figures 2A–2D and 2L). Those in layer In TM at E14.5; P21 brains, GFP+ cells included projection 5 expressed high levels of CTIP2 (also known as BCL11B), neurons, astrocytes, and oligodendrocytes (Figures 2G–2K, indicating a subcerebral neuron identity (Figure 2C). Many 2Q, and S2B). However, 94% of GFP+ neurons were present in GFP+ cells expressed the callosal neuron marker SATB2 and upper layers, and many expressed SATB2 (Figures 2G–2I and were observed in both deep and upper layers (Figure 2D). 2L). Retrograde tracing from the contralateral cortical

Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc. 1169 Neuron Fezf2+ RGCs Are Multilineage Cortical Progenitors

Figure 3. Clonal Analysis of Fezf2+ RGCs (A–D) Fezf2-CreERT2; Confetti mice enabled clonal anaylys of RGCs based upon ﬂourescent protein expression. (A) TM at E12.5; P10 brain demonstrating that clones included both deep- and upper-layer neurons and glia. (B) We occasionally observed clones that contained only glia. (C–D) TM at E14.5; P10 brains contained clones with only upper-layer neruons and glia. TM, tamoxifen. Scale Bars: (D) 100 mm, (D4) 10 mm. See also Figure S3.

from Fezf2+ RGCs expressed CUX2 (Figures 4A and 4B). Since Cux2+ projection neurons were reported to arise from Cux2+ RGCs + + hemisphere showed that GFP SATB2 cells projected callosal (Franco et al., 2012), we next investigated whether Fezf2+ axons (Figures S2G–S2O). In TM at E18.5; P21 brains, only 2% RGCs transit through a CUX2+ RGC stage to generate upper- + of GFP cells were projection neurons (Figure 2Q). Instead, the layer neurons. In agreement with previous reports (Cobos + GFP cells consisted of astrocytes (53%) and oligodendrocytes et al., 2006; Cubelos et al., 2008a, 2008b; Franco et al., 2011, (45%) (Figures 2M–2Q and S2C). Collectively, these results 2012; Nieto et al., 2004; Zimmer et al., 2004), we detected robust + indicate that Fezf2 RGCs are multipotent and sequentially CUX2 expression beginning at E14.5, including expression in generate neocortical projection neurons and glia according to neocortical interneurons (Figure S4). We examined the relation- their birthdate. ship between Fezf2+ RGCs and CUX2+ progenitors using TM at E10.5; E15.5 Fezf2-CreERT2; RCE-GFP brains, since the Clonal Analysis of Fezf2+ RGCs Cux2+ RGC population was reported to peak at this age (Franco To further examine the lineage of Fezf2+ RGCs, we performed et al., 2012). We found that although some CUX2+ cells resided in vivo clonal analysis. Low-efficiency tamoxifen induction at in the VZ, the majority were in the SVZ (Figure 4C). We rarely E12.5 using the RCE-GFP reporter labeled putative RGC clones, observed GFP+CUX2+ cells in the VZ/SVZ (Figure 4E), and which consisted of neurons in both deep and upper layers, these cells were SOX2 (Figure 4D), suggesting they were not astrocytes, and oligodendrocytes (Figure S3A). To extend our RGCs, but likely basal progenitors generated from Fezf2+ clonal analysis, we crossed Fezf2-CreERT2 mice to the Confetti RGCs. Taken together, these results indicate that Fezf2+ RGCs reporter line (Snippert et al., 2010), which enabled the identifi- that generate upper-layer neurons do not express significant cation of individual RGC clones based upon the exclusive levels of CUX2 protein. expression of CFP, GFP, YFP or RFP. Early (TM at E12.5) The lack of GFP+CUX2+ RGCs in the VZ/SVZ of TM at E10.5; Fezf2+ RGCs produced clones that consisted of neurons in E15.5 brains, prompted us to investigate the identity of CUX2+ deep and upper layers and glia (Figures 3A and 3B and Figures cells in the neocortical VZ/SVZ. Since CUX2 was previously re- S3B and S3C). In contrast, when tamoxifen was administered ported to be expressed in interneurons (Zimmer et al., 2004), at E14.5, clones consisted of only upper-layer neurons and glia we determined the percentage of CUX2+ cells in the E15.5 (Figures 3C and 3D and Figures S3D and S3E). Of note, we neocortical VZ/SVZ that originated from the ventral telenceph- occasionally observed clones consisting of only glia (Figure 3E), alon. To label cortical interneurons, we crossed either the consistent with previous viral-based clonal analysis of cortical Dlx1/2-Cre (Potter et al., 2009) or the Nkx2.1-Cre (Xu et al., progenitors (Luskin et al., 1988; Walsh and Cepko, 1988). These 2008) alleles to mice harboring the Ai14 reporter (Madisen results support the conclusion that Fezf2+ RGCs sequentially et al., 2010). In the mice carrying both Cre and Ai14 alleles, generate cortical projection neuron subtypes and glia. cortical interneurons were labeled by TdTomato expression (Figures 4G and 4I). Examination of E15.5 Dlx1/2-Cre; Ai14 or CUX2+ Cells in the VZ and SVZ Are Migrating Nkx2.1-Cre; Ai14 brains revealed that the vast majority (94% or Interneurons Derived from the Ventral Telencephalon 98%, respectively) of CUX2+ cells in the neocortical VZ/SVZ The finding that Fezf2+ RGCs contribute substantially to upper- were TdTomato+ (Figures 4F–4J). This result demonstrates layer neurogenesis is consistent with the classic progressive that in the developing neocortex, nearly all CUX2+ cells in the restriction model (Leone et al., 2008). However, this is in contrast VZ/SVZ are immature interneurons derived from the ventral to a newly proposed model that suggests all upper-layer neurons telencephalon. are generated from Cux2+ RGCs (Franco and Mu¨ ller, 2013). To resolve this difference, we explored the relationship between Cux2-Cre and Cux2-CreERT2 Label RGCs that Generate Fezf2+ RGCs and CUX2+ cortical cells using two different Both Deep- and Upper-Layer Neurons CUX2 antibodies that recognize different regions of the CUX2 The lack of CUX2+ RGCs in the developing neocortex seemed protein (Conforto et al., 2012; Iulianella et al., 2008; Laz et al., to contradict the previous report that Cux2+ RGCs, marked by 2007). We found that 72% of upper-layer neurons generated Cux2-Cre and Cux2-CreERT2 alleles, generate upper-layer

1170 Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc. Neuron Fezf2+ RGCs Are Multilineage Cortical Progenitors

(legend on next page)

Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc. 1171 Neuron Fezf2+ RGCs Are Multilineage Cortical Progenitors

projection neurons (Franco et al., 2012). To resolve this discrep- and oligodendrocytes. Late Fezf2+ RGCs generate upper-layer ancy, we obtained Cux2-Cre and Cux2-CreERT2 mice and neurons, astrocytes, and oligodendrocytes. Thus, they sequen- analyzed them using RCE-GFP, Ai14, and Ai9 (Madisen et al., tially generate projection neurons and glia in accordance with 2010) reporter lines. In the brains of E15.5 Cux2-Cre; RCE-GFP the classic model of cortical neurogenesis (Leone et al., 2008). and Cux2-Cre; Ai14 embryos, many cells were labeled by GFP However, our results do not exclude the existence of intrinsically or TdTomato (Figures S5E and S5F). Labeled cells in the neocor- fate-restriced RGCs. Rather, because Cux2-Cre- and Cux2- tical VZ/SVZ expressed the RGC marker PAX6 (Figure S5G) or CreERT2-labeled neocortical RGCs generate both deep- and the basal progenitor marker TBR2 (Figure S5H), consistent with upper-layer neurons, this suggests that laminar-fate-restricted the Cux2 in situ hybridization results (Figures S5A–S5D; see RGCs cannot be identiﬁed by Cux2 expression alone. also Franco et al., 2012). Immunohistochemistry using CUX2 A comparison between Fezf2-CreERT2 mice and the Cux2-Cre and CRE antibodies revealed that most of the lineage-traced and Cux2-CreERT2 mice used by Franco et al. (2012) reveals cells in the VZ/SVZ did not express CUX2 or CRE protein (Figures possible reasons for the divergent conclusions reached by these S5J and S5K), indicating that the Cre reporter system may be two studies. Upon tamoxifen administration, the Fezf2-CreERT2 more sensitive than immunohistochemistry for detecting low allele induces recombination of the RCE-GFP reporter only in levels of Cux2 expression. Aside from cortical progenitor cells, progenitor cells and not in postmitotic neurons. This was Cux2-Cre also labeled many neurons in the cortical plate that observed when tamoxifen or 4-hydroxytamoxifen was adminis- expressed CTIP2, a marker for early-born subcortical projection tered from E12 to adult stages. However, the Cux2-Cre and neurons (Figure S5I). Cux2-CreERT2 alleles induced recombination in both progenitor We next investigated the lineage potential of Cux2+ RGCs cells and postmitotic neurons (Franco et al., 2011, 2012), using Cux2-CreERT2; Ai9 mice. In TM at E10.5; P0 brains, possibly masking the true lineage potential of Cux2+ progenitor TdTomato+ cells were observed throughout the brain (Figure 4K) cells. Furthermore, Fezf2 is not expressed in neocortical inter- including in both deep and upper cortical layers (Figure 4L). neurons or their progenitors. In contrast, Cux2 is expressed at Indeed, within layers 2–6 of the neocortex, 29% and 26% of high levels in interneurons that are born in the subpallium and TdTomato+ cells expressed the deep-layer markers CTIP2 and migrate into the neocortex (Cobos et al., 2006). TBR1, respectively, whereas 30% expressed the upper-layer marker CUX1 (Figures 4M–4S and S5N). CTIP2 expression in Fezf2 Expression in Neocortical Progenitors TdTomato+ cells in P0 Cux2-CreERT2; Ai9 brains is consistent Fezf2 is essential for subcerebral neuron development (Chen with the observed expression in labeled cells of E15 Cux2-Cre et al., 2005a, 2005b, 2008; Han et al., 2011; McKenna et al., brains (Figure S5I). In addition to CTIP2 and TBR1, TdTomato+ 2011; Molyneaux et al., 2005). Indeed, overexpression of Fezf2 cells also expressed NFIB and SOX5 (Figures S5L and S5M), in late cortical progenitors redirects these cells to generate both of which mark early-born subcortical projection neurons subcerebral projection neurons (Chen et al., 2005b, 2008; at P0 (Betancourt et al., 2013; Kwan et al., 2008; Lai et al., Molyneaux et al., 2005). Here, we demonstrate that, during 2008; McKenna et al., 2011). Collectively, these results demon- normal brain development, Fezf2 is expressed in late RGCs as strate that early Cux2-Cre- and Cux2-CreERT2-labeled neocor- these cells are generating upper-layer projection neurons and tical RGCs are not lineage restricted, in that they generate both glia. This suggests it is not simply Fezf2 expression, but rather deep- and upper-layer projection neurons. its expression level, in RGCs that inﬂuences the differentiated cell types that are produced. Going forward it will be important DISCUSSION to understand the mechanisms that precisely regulate Fezf2 transcription levels across different expression domains in order to Sequential Generation of Projection Neurons functionally dissect its involvement in neocortical development. and Glia by Multipotent RGCs Our lineage analysis demonstrates that both early and late Heterogeneity of Embryonic Neural Stem Cells Fezf2+ RGCs are multipotent. Early Fezf2+ RGCs generate all Previous studies suggest that neocortical RGCs are a heteroge- major cortical projection neuron subtypes as well as astrocytes neous cell population. Subsets of RGCs were reported to

Figure 4. CUX2+ Cells in VZ/SVZ Are Migrating Interneurons and Cux2-Cre/CreER T2-Labeled RGCs Generate Both Deep- and Upper-Layer Projection Neurons (A and B) Fezf2+ RGCs generated CUX2+ upper-layer neurons. (C–E) Immunohistochemical analysis of TM at E10.5; E15.5 Fezf2-CreERT2; RCE-GFP brains. Few CUX2+ cells in the SVZ expressed SOX2, and these cells were located at the VZ/SVZ boundary (C and D). Rare CUX2+GFP+ cell in the VZ/SVZ (E). These CUX2+GFP+ cells did not express SOX2 (D and E). The arrowheads in (C)–(E) point to the CUX2+SOX2+ cells in VZ/SVZ, and the arrows point to a rare GFP+CUX2+ cell. (F–J) Dlx1/2-Cre; Ai14 and Nkx2.1-Cre; Ai14 mice revealed that the majority of CUX2+ cells in the E15.5 neocortical VZ/SVZ were interneurons generated from Dlx1/2 (F and G) or Nkx2.1 (H and I) lineages. Quantiﬁcation of the percentage of CUX2+TdTomato+ cells among all CUX2+ cells in the VZ/SVZ ± SEM (J). (K and L) TM at E10.5; P0 Cux2-CreERT2; Ai9 brains contained TdTomato+ cells in all cortical layers. (M–O) TdTomato+ cells in deep layers expressed CTIP2. (P–R) Upper-layer TdTomato+ cells expressed CUX1. (S) Percentage of TdTomato+ cells in layers 2–6 that expressed CTIP2, TBR1, or CUX1 ± SEM. Ctx, cerebral cortex; LV, lateral ventricle; SVZ, subventricular zone; Thal, thalamus; TM, tamoxifen; VZ, ventricular zone. Scale bars: 25 mm (B, E, G, I, O, and R), 500 mm (K), 200 mm (L), and 10 mm (close-ups from O and R). See also Figures S4 and S5.

1172 Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc. Neuron Fezf2+ RGCs Are Multilineage Cortical Progenitors

differentially express markers such as RC2, GLAST, and BLBP acquisition. This work was funded by grant R01MH094589 from the National (Hartfuss et al., 2001). In addition, in vivo lineage analysis Institutes of Health (to B.C.), New Faculty Award RN1-00530-1 from the of neocortical progenitors using retroviral vectors produced California Institute of Regenerative Medicine (CIRM; to B.C.), Training Grant TG2-01157 from CIRM (to C.G. and W.L.M.), and grant R01MH049428 from clones that consisted of only neurons or glia (Luskin et al., the National Institutes of Health (to J.L.R.R.). 1988; Walsh and Cepko, 1988), suggesting the existence of neuron- or glia-restricted progenitor cells. Furthermore, recent Accepted: September 10, 2013 work indicates that the neocortical progenitor pool contains Published: December 4, 2013 both outer radial glia cells (Hansen et al., 2010; Wang et al., 2011) as well as short neural precursors (Gal et al., 2006; Stancik REFERENCES et al., 2010; Tyler and Haydar, 2013) that are molecularly and morphologically distinct from ventricular zone RGCs. Betancourt, J., Katzman, S., and Chen, B. (2013). Nuclear factor one b Results from our study indicate that progressive lineage regulates neural stem cell differentiation and axonal projection of corticofugal restriction of multipotent RGCs is a common mechanism neurons. J. Comp. Neurol. Published online June 8, 2013. http://dx.doi.org/ 10.1002/cne.23373. for generating cellular diversity during cortical development. However, the finding that Cux2-Cre- and Cux2-CreERT2-labeled Chen, B., Schaevitz, L.R., and McConnell, S.K. (2005a). Fezl regulates the differentiation and axon targeting of layer 5 subcortical projection neurons in RGCs are not upper-layer, fate-restricted progenitors does not cerebral cortex. Proc. Natl. Acad. Sci. USA 102, 17184–17189. exclude the possibility that some cortical projection neurons Chen, J.G., Rasin, M.R., Kwan, K.Y., and Sestan, N. (2005b). Zfp312 is may originate from fate-restricted RGCs. Indeed, our clonal required for subcortical axonal projections and dendritic morphology of + analysis of Fezf2 RGCs produced clones that consisted of deep-layer pyramidal neurons of the cerebral cortex. Proc. Natl. Acad. Sci. only glial cells (Figure 3B), suggesting the possibility of USA 102, 17792–17797. + lineage-restricted subpopulations within the Fezf2 RGC Chen, B., Wang, S.S., Hattox, A.M., Rayburn, H., Nelson, S.B., and McConnell, population. Moreover, it is possible that Fezf2+ RGCs are a het- S.K. (2008). The Fezf2-Ctip2 genetic pathway regulates the fate choice of erogeneous population in that some Fezf2+ RGCs may express subcortical projection neurons in the developing cerebral cortex. Proc. Natl. Cux2 mRNA or protein at levels below the detection threshold of Acad. Sci. USA 105, 11382–11387. in situ hybridization or immunohistochemistry. Many genes Cobos, I., Long, J.E., Thwin, M.T., and Rubenstein, J.L. (2006). Cellular pat- expressed in RGCs have been identified (Molyneaux et al., terns of transcription factor expression in developing cortical interneurons. 2007; Woodworth et al., 2012). Genetic fate mapping using Cereb. Cortex 16 (Suppl 1 ), i82–i88. these loci will further expand our knowledge of progenitor Conforto, T.L., Zhang, Y., Sherman, J., and Waxman, D.J. (2012). Impact of cell heterogeneity within the neocortical progenitor pool and CUX2 on the female mouse liver transcriptome: activation of female-biased genes and repression of male-biased genes. Mol. Cell. Biol. 32, 4611–4627. uncover the mechanisms that generate projection neuron ´ diversity. Cubelos, B., Sebastian-Serrano, A., Kim, S., Moreno-Ortiz, C., Redondo, J.M., Walsh, C.A., and Nieto, M. (2008a). Cux-2 controls the proliferation of neuronal intermediate precursors of the cortical subventricular zone. Cereb. Cortex 18, EXPERIMENTAL PROCEDURES 1758–1770. Cubelos, B., Sebastia´ n-Serrano, A., Kim, S., Redondo, J.M., Walsh, C., and All experiments were carried out in accordance with protocols approved by Nieto, M. (2008b). Cux-1 and Cux-2 control the development of Reelin the IACUC at University of California at Santa Cruz and were performed in expressing cortical interneurons. Dev. Neurobiol. 68, 917–925. accordance with institutional and federal guidelines. Fezf2-CreERT2 mice De la Rossa, A., Bellone, C., Golding, B., Vitali, I., Moss, J., Toni, N., Lu¨ scher, were generated according to previously established strategies (Lee et al., C., and Jabaudon, D. (2013). In vivo reprogramming of circuit connectivity in 2001) by modifying the RP23-141E17 BAC. postmitotic neocortical neurons. Nat. Neurosci. 16, 193–200. In situ hybridization and immunohistochemistry were performed as previously described (Eckler et al., 2011). The rabbit anti-CUX2 antibody Desai, A.R., and McConnell, S.K. (2000). Progressive restriction in fate (Conforto et al., 2012; Laz et al., 2007) was a gift from Dr. David Waxman. potential by neural progenitors during cerebral cortical development. We also confirmed CUX2 expression using a second rabbit anti-CUX2 anti- Development 127, 2863–2872. body (Iulianella et al., 2008, 2009) kindly provided by Angelo Iulianella. Eckler, M.J., McKenna, W.L., Taghvaei, S., McConnell, S.K., and Chen, B. (2011). Fezf1 and Fezf2 are required for olfactory development and sensory SUPPLEMENTAL INFORMATION neuron identity. J. Comp. Neurol. 519, 1829–1846. Franco, S.J., and Mu¨ ller, U. (2013). Shaping our minds: stem and progenitor Supplemental Information includes Supplemental Experimental Procedures cell diversity in the mammalian neocortex. Neuron 77, 19–34. and five figures and can be found with this article online at http://dx.doi.org/ Franco, S.J., Martinez-Garay, I., Gil-Sanz, C., Harkins-Perry, S.R., and Mu¨ ller, 10.1016/j.neuron.2013.09.037. U. (2011). Reelin regulates cadherin function via Dab1/Rap1 to control neuronal migration and lamination in the neocortex. Neuron 69, 482–497. ACKNOWLEDGMENTS Franco, S.J., Gil-Sanz, C., Martinez-Garay, I., Espinosa, A., Harkins-Perry, S.R., Ramos, C., and Mu¨ ller, U. (2012). Fate-restricted neural progenitors in We thank Drs. Lindsay Hinck, Susan McConnell, and Jennifer Betancourt for the mammalian cerebral cortex. Science 337, 746–749. critical reading of the manuscript. We thank Drs. David Waxman and Tara L. Conforto (Boston University) and Dr. Angelo Iulianella (Dalhousie University) Friedrich, G., and Soriano, P. (1991). Promoter traps in embryonic stem cells: for providing CUX2 antibodies. Drs. Guo-Ping Feng, Xue-wen Cheng, and a genetic screen to identify and mutate developmental genes in mice. Genes Zhi-qi Xiong kindly provided the pERFN vector. We thank Drs. Ulrich Mueller, Dev. 5, 1513–1523. Susan McConnell, and Pushkar Joshi for providing Cux2-Cre and Cux2- Gal, J.S., Morozov, Y.M., Ayoub, A.E., Chatterjee, M., Rakic, P., and Haydar, CreERT2 mice and Dr. Gord Fishell for providing the RCE-GFP mice. Dr. Ben T.F. (2006). Molecular and morphological heterogeneity of neural precursors Abrams and the UCSC Microscopy Center provided support with image in the mouse neocortical proliferative zones. J. Neurosci. 26, 1045–1056.

Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc. 1173 Neuron Fezf2+ RGCs Are Multilineage Cortical Progenitors

Gong, S., Zheng, C., Doughty, M.L., Losos, K., Didkovsky, N., Schambra, U.B., McConnell, S.K., and Kaznowski, C.E. (1991). Cell cycle dependence of Nowak, N.J., Joyner, A., Leblanc, G., Hatten, M.E., and Heintz, N. (2003). laminar determination in developing neocortex. Science 254, 282–285. A gene expression atlas of the central nervous system based on bacterial McKenna, W.L., Betancourt, J., Larkin, K.A., Abrams, B., Guo, C., Rubenstein, artificial chromosomes. Nature 425, 917–925. J.L., and Chen, B. (2011). Tbr1 and Fezf2 regulate alternate corticofugal Han, W., Kwan, K.Y., Shim, S., Lam, M.M., Shin, Y., Xu, X., Zhu, Y., Li, M., and neuronal identities during neocortical development. J. Neurosci. 31, 549–564. Sestan, N. (2011). TBR1 directly represses Fezf2 to control the laminar origin Molyneaux, B.J., Arlotta, P., Hirata, T., Hibi, M., and Macklis, J.D. (2005). Fezl and development of the corticospinal tract. Proc. Natl. Acad. Sci. USA 108, is required for the birth and specification of corticospinal motor neurons. 3041–3046. Neuron 47, 817–831. Hansen, D.V., Lui, J.H., Parker, P.R., and Kriegstein, A.R. (2010). Neurogenic Molyneaux, B.J., Arlotta, P., Menezes, J.R., and Macklis, J.D. (2007). Neuronal radial glia in the outer subventricular zone of human neocortex. Nature 464, subtype specification in the cerebral cortex. Nat. Rev. Neurosci. 8, 427–437. 554–561. Hartfuss, E., Galli, R., Heins, N., and Go¨ tz, M. (2001). Characterization of Nieto, M., Monuki, E.S., Tang, H., Imitola, J., Haubst, N., Khoury, S.J., CNS precursor subtypes and radial glia. Dev. Biol. 229, 15–30. Cunningham, J., Gotz, M., and Walsh, C.A. (2004). Expression of Cux-1 and Cux-2 in the subventricular zone and upper layers II-IV of the cerebral cortex. Hippenmeyer, S., Vrieseling, E., Sigrist, M., Portmann, T., Laengle, C., Ladle, J. Comp. Neurol. 479, 168–180. D.R., and Arber, S. (2005). A developmental switch in the response of DRG neurons to ETS transcription factor signaling. PLoS Biol. 3, e159. Potter, G.B., Petryniak, M.A., Shevchenko, E., McKinsey, G.L., Ekker, M., and Rubenstein, J.L. (2009). Generation of Cre-transgenic mice using Dlx1/Dlx2 Hirata, T., Suda, Y., Nakao, K., Narimatsu, M., Hirano, T., and Hibi, M. (2004). enhancers and their characterization in GABAergic interneurons. Mol. Cell. Zinc finger gene fez-like functions in the formation of subplate neurons and Neurosci. 40, 167–186. thalamocortical axons. Dev. Dyn. 230, 546–556. Iulianella, A., Sharma, M., Durnin, M., Vanden Heuvel, G.B., and Trainor, P.A. Rouaux, C., and Arlotta, P. (2013). Direct lineage reprogramming of post- (2008). Cux2 (Cutl2) integrates neural progenitor development with cell-cycle mitotic callosal neurons into corticofugal neurons in vivo. Nat. Cell Biol. 15, progression during spinal cord neurogenesis. Development 135, 729–741. 214–221. Iulianella, A., Sharma, M., Vanden Heuvel, G.B., and Trainor, P.A. (2009). Cux2 Shen, Q., Wang, Y., Dimos, J.T., Fasano, C.A., Phoenix, T.N., Lemischka, I.R., functions downstream of Notch signaling to regulate dorsal interneuron Ivanova, N.B., Stifani, S., Morrisey, E.E., and Temple, S. (2006). The timing of formation in the spinal cord. Development 136, 2329–2334. cortical neurogenesis is encoded within lineages of individual progenitor cells. Nat. Neurosci. 9, 743–751. Kwan, K.Y., Lam, M.M., Krsnik, Z., Kawasawa, Y.I., Lefebvre, V., and Sestan, N. (2008). SOX5 postmitotically regulates migration, postmigratory differen- Shim, S., Kwan, K.Y., Li, M., Lefebvre, V., and Sestan, N. (2012). Cis-regulatory tiation, and projections of subplate and deep-layer neocortical neurons. control of corticospinal system development and evolution. Nature 486, Proc. Natl. Acad. Sci. USA 105, 16021–16026. 74–79. Kwan, K.Y., Sestan, N., and Anton, E.S. (2012). Transcriptional co-regulation Snippert, H.J., van der Flier, L.G., Sato, T., van Es, J.H., van den Born, M., of neuronal migration and laminar identity in the neocortex. Development Kroon-Veenboer, C., Barker, N., Klein, A.M., van Rheenen, J., Simons, B.D., 139, 1535–1546. and Clevers, H. (2010). Intestinal crypt homeostasis results from neutral Lai, T., Jabaudon, D., Molyneaux, B.J., Azim, E., Arlotta, P., Menezes, J.R., competition between symmetrically dividing Lgr5 stem cells. Cell 143, and Macklis, J.D. (2008). SOX5 controls the sequential generation of distinct 134–144. corticofugal neuron subtypes. Neuron 57, 232–247. Sousa, V.H., Miyoshi, G., Hjerling-Leffler, J., Karayannis, T., and Fishell, G. Laz, E.V., Holloway, M.G., Chen, C.S., and Waxman, D.J. (2007). (2009). Characterization of Nkx6-2-derived neocortical interneuron lineages. Characterization of three growth hormone-responsive transcription factors Cereb. Cortex 19 (Suppl 1 ), i1–i10. preferentially expressed in adult female liver. Endocrinology 148, 3327–3337. Stancik, E.K., Navarro-Quiroga, I., Sellke, R., and Haydar, T.F. (2010). Lee, E.C., Yu, D., Martinez de Velasco, J., Tessarollo, L., Swing, D.A., Court, Heterogeneity in ventricular zone neural precursors contributes to neuronal D.L., Jenkins, N.A., and Copeland, N.G. (2001). A highly efficient Escherichia fate diversity in the postnatal neocortex. J. Neurosci. 30, 7028–7036. coli-based chromosome engineering system adapted for recombinogenic Tyler, W.A., and Haydar, T.F. (2013). Multiplex genetic fate mapping reveals targeting and subcloning of BAC DNA. Genomics 73, 56–65. a novel route of neocortical neurogenesis, which is altered in the Ts65Dn Leone, D.P., Srinivasan, K., Chen, B., Alcamo, E., and McConnell, S.K. (2008). mouse model of Down syndrome. J. Neurosci. 33, 5106–5119. The determination of projection neuron identity in the developing cerebral Walsh, C., and Cepko, C.L. (1988). Clonally related cortical cells show several cortex. Curr. Opin. Neurobiol. 18, 28–35. migration patterns. Science 241, 1342–1345. Luskin, M.B., Pearlman, A.L., and Sanes, J.R. (1988). Cell lineage in the cerebral cortex of the mouse studied in vivo and in vitro with a recombinant Wang, X., Tsai, J.W., LaMonica, B., and Kriegstein, A.R. (2011). A new subtype retrovirus. Neuron 1, 635–647. of progenitor cell in the mouse embryonic neocortex. Nat. Neurosci. 14, 555–561. Madisen, L., Zwingman, T.A., Sunkin, S.M., Oh, S.W., Zariwala, H.A., Gu, H., Ng, L.L., Palmiter, R.D., Hawrylycz, M.J., Jones, A.R., et al. (2010). A robust Woodworth, M.B., Custo Greig, L., Kriegstein, A.R., and Macklis, J.D. (2012). and high-throughput Cre reporting and characterization system for the whole SnapShot: cortical development. Cell 151, 918–918.e1. mouse brain. Nat. Neurosci. 13, 133–140. Xu, Q., Tam, M., and Anderson, S.A. (2008). Fate mapping Nkx2.1-lineage cells Marıń, O. (2012). Brain development: The neuron family tree remodelled. in the mouse telencephalon. J. Comp. Neurol. 506, 16–29. Nature 490, 185–186. Zimmer, C., Tiveron, M.C., Bodmer, R., and Cremer, H. (2004). Dynamics McConnell, S.K. (1985). Migration and differentiation of cerebral cortical neu- of Cux2 expression suggests that an early pool of SVZ precursors is fated rons after transplantation into the brains of ferrets. Science 229, 1268–1271. to become upper cortical layer neurons. Cereb. Cortex 14, 1408–1420.

1174 Neuron 80, 1167–1174, December 4, 2013 ª2013 Elsevier Inc.