Skip to main content
PMC home page
eLife logoLink to eLife
. 2019 May 17;8:e42012. doi: 10.7554/eLife.42012

FOXP2 exhibits projection neuron class specific expression, but is not required for multiple aspects of cortical histogenesis

Ryan J Kast 1,†,✉, Alexandra L Lanjewar 1,†, Colton D Smith 1, Pat Levitt 1,✉
Editors: Constance L Cepko2, Huda Y Zoghbi3
PMCID: PMC6561705  PMID: 31099752

Abstract

The expression patterns of the transcription factor FOXP2 in the developing mammalian forebrain have been described, and some studies have tested the role of this protein in the development and function of specific forebrain circuits by diverse methods and in multiple species. Clinically, mutations in FOXP2 are associated with severe developmental speech disturbances, and molecular studies indicate that impairment of Foxp2 may lead to dysregulation of genes involved in forebrain histogenesis. Here, anatomical and molecular phenotypes of the cortical neuron populations that express FOXP2 were characterized in mice. Additionally, Foxp2 was removed from the developing mouse cortex at different prenatal ages using two Cre-recombinase driver lines. Detailed molecular and circuit analyses were undertaken to identify potential disruptions of development. Surprisingly, the results demonstrate that Foxp2 function is not required for many functions that it has been proposed to regulate, and therefore plays a more limited role in cortical development than previously thought.

Research organism: Mouse

Introduction

The proper function of the cerebral cortex requires the formation of highly stereotyped circuits during development. These circuits are built through interdependent processes, including proliferation of neural progenitors, migration of neurons to their appropriate positions, morphological and physiological differentiation of diverse neuron subtypes, and the formation of synapses of requisite strength between appropriate pairs of neurons. Impairments in these fundamental aspects of development can lead to lifelong dysfunction of the cortex, which is believed to contribute to core symptoms of many neurodevelopmental disorders (Rubenstein and Rakic, 2013).

The winged helix transcription factor, FOXP2, has been implicated in ontogenetic processes relevant to the development of the cerebral cortex (Vernes et al., 2011; Chiu et al., 2014; Chen et al., 2016), and some studies have directly implicated FOXP2 in cortical ontogeny (Tsui et al., 2013; Garcia-Calero et al., 2016). FOXP2 expression in the developing cortex is restricted to subpopulations of post-mitotic neurons in the deep cortical layers, a pattern that is highly conserved across mammalian species (Ferland et al., 2003; Takahashi et al., 2003; Campbell et al., 2009; Mukamel et al., 2011). Mutations in FOXP2 cause a severe developmental speech and language disorder, known as childhood apraxia of speech (Lai et al., 2001; MacDermot et al., 2005). Human neuroimaging and animal studies have identified alterations in basal ganglia function that could contribute to the clinical disorder symptoms (Vargha-Khadem et al., 1998; Belton et al., 2003; Groszer et al., 2008; French et al., 2012; Chen et al., 2016), but whether changes in cerebral cortical organization and function are critically involved in impairments associated with FOXP2 mutations is currently unknown.

The present study aimed to establish a more detailed understanding of the cell-type identity of FOXP2+ neurons in the developing murine cerebral cortex, using molecular and neuroanatomical phenotyping approaches. Further, this study applied gold-standard conditional mouse genetics to selectively remove Foxp2 from the developing cerebral cortex at different prenatal ages to ascertain its putative roles in the normal histogenic processes that generate the canonical six layers, specific cell types based on gene expression, and basic axon targeting to subcortical structures. The results show that FOXP2 expression is limited to specific corticofugal neuron populations and suggest that the gene plays a more limited role in mouse corticogenesis than previously concluded based on results obtained by other experimental methodologies.

Results

Foxp2 expression is enriched in developing corticothalamic projection neurons

Initial Foxp2 expression mapping studies determined that Foxp2 transcript and protein expression begin prenatally, with the onset of protein expression delayed relative to the mRNA, and with protein present primarily in postmitotic neurons (Ferland et al., 2003). However, other more recent studies have reported that FOXP2 protein is also expressed in mitotic progenitor cells (Tsui et al., 2013), where it regulates cortical neurogenesis. FOXP2 immunohistochemistry of coronal sections of the embryonic forebrain suggested that FOXP2 protein expression begins between embryonic day (E) 14.5 and E16.5 within postmitotic neurons of the infragranular layers (Figure 1A,B). Postnatally, it is well established that Foxp2 expression is limited to glutamatergic neurons of the infragranular cortical layers, with robust expression predominantly in layer 6 (Ferland et al., 2003; Hisaoka et al., 2010; Sundberg et al., 2018). Layer 6 contains many FOXP2+ neurons, whereas layer 5 contains some FOXP2+ neurons that are more abundant in medial cortical areas at early postnatal stages (Ferland et al., 2003; Campbell et al., 2009). Layer 6, where most of the FOXP2+ neurons are located, is comprised of two primary glutamatergic cortical neuron populations, corticothalamic (CT) and corticocortical (CC) neurons (Thomson, 2010; Petrof et al., 2012; Harris and Shepherd, 2015). To determine whether FOXP2 expression is selective to one of these populations or is expressed in both layer 6 projection neuron subtypes during development, on postnatal day (P) 12, the neuroanatomical tracer cholera toxin subunit B (CTB) was injected into primary somatosensory thalamus or primary motor cortex in separate cohorts of mice. Cell bodies of CT or CC neurons residing in layer 6 of primary SSC were retrogradely labeled. Sections through primary SSC were then stained with an anti-FOXP2 antibody and cellular expression of FOXP2 was assessed among retrogradely labeled neurons ipsilateral to the tracer injections. FOXP2 expression was evident in most CT neurons (Mean ± SEM, 78.3 ± 2.9%), but was expressed by very few CC neurons (Mean ± SEM, 6.4 ± 1.7%; Figure 1C–E).

Figure 1. FOXP2 is enriched in corticothalamic neurons during cortical development.

(A) Low magnification (left) and high magnification (right) images of FOXP2 (yellow) immunohistochemical labeling of E14.5 forebrain reveals absence of expression in the dorsal pallium, whereas the developing striatum is robustly labeled at this timepoint (N = 5). (B) Images of FOXP2 immunolabeling at E16.5 demonstrates the presence of FOXP2 expression within the deep layers of the developing cortical plate (N = 3). (C) Retrograde labeling of layer 6 corticothalamic neurons (magenta) by injection of CTB into the ventrobasal thalamus, combined with FOXP2 (yellow) immunohistochemistry at P14. (D) Corticocortical neurons (cyan) labeled by injection of CTB into the ipsilateral primary motor cortex, combined with FOXP2 (yellow) immunohistochemistry at P14. White arrows denote retrogradely labeled projection neurons. (E) Quantification of the percentages of retrogradely labeled corticothalamic (N = 4 mice) and corticocortical (N = 5) neurons that express FOXP2. (F) FOXP2 (yellow) immunohistochemistry in sections of P0, P7, and P14 somatosensory cortex from Ntsr1-cre; tdTomato mice (tdTomato is magenta); white asterisks denote relatively low-level expression in layer 5 at P0. (G) Quantification of the percentages of tdTomato-positive neurons that express FOXP2 at each age (P0, N = 3; P7, N = 3; P14, N = 3). (H) FOXP2 (magenta) immunohistochemistry in sections of P0, P7, and P14 somatosensory cortex from MetGFP (green) mice. Cyan arrowheads denote sparse FOXP2+ and GFP+ double-labeled cells localized to layer 6B/subplate. (I) Quantification of the percentages of GFP+ neurons that co-express Foxp2 at each age (P0, N = 3; P7, N = 4; P14, N = 3). (J) FOXP2 (yellow) colocalizes with PCP4 (magenta), but not ppCCK (cyan) (N = 3). Scale bars: 500 µm, A, B low magnification; 100 µm A, B high magnification; 50 µm C, D, F, H and J.

Figure 1—source data 1. Foxp2 expression among retrogradely labeled and molecularly defined developing layer 6 neuron classes.
elife-42012-fig1-data1.xlsx (17.4KB, xlsx)
DOI: 10.7554/eLife.42012.005

Figure 1.

Figure 1—figure supplement 1. FOXP2 is transiently expressed by a subpopulation of PT neurons.

Figure 1—figure supplement 1.

(A) FOXP2 (red) and CTIP2 (cyan) immunohistochemistry in sections of P0 somatosensory cortex. (B) FOXP2 (red) and CTIP2 (cyan) immunohistochemistry in sections of P7 somatosensory cortex. (C) Quantification of the percentage of CTIP2+ PT neurons in somatosensory cortex that express FOXP2 at P0 (N = 3) and P7 (N = 5). (D) Lack of expression of FOXP2 in most retrogradely (CTB) labeled PT neurons at P14 (N = 5). (E) Quantification of the percentage of CTB-labeled PT neurons that express FOXP2. All scale bars, 100 µm.
Figure 1—figure supplement 1—source data 1. Foxp2 expression among retrogradely and molecularly defined developing pyramidal tract neurons.
elife-42012-fig1-figsupp1-data1.xlsx (11.2KB, xlsx)
DOI: 10.7554/eLife.42012.004
Open in a new tab

Next, FOXP2 expression was assessed among molecularly-defined CT and CC neurons at several postnatal developmental stages. The CT-specific cre-driver mouse, Ntsr1-cre, was crossed with a cre-dependent Rosa-tdTomato reporter line (Ai14) to selectively and comprehensively label layer 6 CT neurons with tdTomato. FOXP2 immunohistochemistry revealed that nearly all tdTomato-positive neurons in the primary SSC co-expressed FOXP2 at P0, P7, and P14 (Figure 1F–G, Mean ± SEM: P0 = 90 ± 2%, P7 = 87 ± 2%, P14 = 91 ± 1%). For further molecular characterization, we also examined FOXP2 overlap with the receptor tyrosine kinase MET, which is expressed in CC neurons of layer 6, but nearly excluded from layer 6 CT neurons in SSC (Kast et al., 2019). Using co-labeling in MetGFP reporter mice (Kast et al., 2019), analysis of FOXP2 expression among GFP+ layer 6 CC neurons revealed relatively few double-positive neurons at P0, P7, and P14 (Figure 1H–I, Mean ± SEM: P0 = 19 ± 2%, P7 = 16 ± 2%, P14 = 5 ± 2%). This finding is consistent with the limited expression of FOXP2 by layer 6 CC neurons observed through retrograde tracing. It is noteworthy, however, that there are FOXP2 and MetGFP double-positive cells positioned in the subplate/layer 6B, similar to the observation of Ntsr1-cre and MetGFP colocalization in deep layer six previously reported (Kast et al., 2019). Notably, the FOXP2 and GFP double-labeled neurons became quite sparse by P14, perhaps due to the downregulation of GFP in the subplate or to the programmed cell death of some subplate neuron populations, as has been reported previously during early postnatal development (Hoerder-Suabedissen and Molnár, 2013). We also examined two other layer 6 neuronal subtype marker genes, PCP4 and ppCCK (Watakabe et al., 2012). There was extensive colocalization between FOXP2 and PCP4, a marker of corticothalamic neurons, in layer 6, but there was minimal colocalization between FOXP2 and ppCCK, a marker of CC neurons (Figure 1J). Together, the molecular and connectivity analysis of FOXP2+ neurons in layer six indicate that FOXP2 expression is highly enriched in CT neurons of primary SSC during postnatal development. This is consistent with analysis in primary visual cortex of adult mice, which showed that Foxp2/FOXP2 expression is nearly exclusive to Ntsr1-cre expressing (CT) neurons in the adult (Tasic et al., 2016; Sundberg et al., 2018). The present study builds on previous findings by demonstrating minimal colocalization with two layer 6 CC markers (MET and CCK). Notably, lower level expression of Foxp2 transcript has been detected in molecularly-defined subcerebral projection neurons by RNA-sequencing at perinatal stages (Molyneaux et al., 2015), consistent with the low-level expression observed in layer 5 at P0 (Figure 1F). Colocalization analysis with the pyramidal-tract (PT) neuron marker gene CTIP2 (Arlotta et al., 2005), confirmed that these layer 5 neurons are PT-type neurons (Figure 1—figure supplement 1). However, in contrast to the temporally stable and relatively uniform expression of FOXP2 by CT neurons in layer 6 of the postnatal SSC (Figure 1), FOXP2 was expressed by a minor subset of CTIP2+ PT neurons at P0 (34 ± 3%) and P7 (10 ± 2%) (Figure 1—figure supplement 1A–C), and was almost completely absent from retrogradely labeled PT neurons in Layer 5 of SSC by P14 (4.4 ± 1.7%) (Figure 1—figure supplement 1D,E).

FOXP2 is not required for normal histogenesis of the cerebral cortex

Given the enrichment of FOXP2 in corticofugal neuron subclasses in SSC, coupled with previous reports of fundamental roles for Foxp2 in cortical neuron development (Clovis et al., 2012; Tsui et al., 2013; Garcia-Calero et al., 2016), we next used a direct genetic deletion strategy to examine putative involvement of FOXP2 in the development of anatomical and molecular properties of CT neurons. For these studies, mice harboring a Foxp2 conditional allele (Foxp2fx) were bred with the CT-specific reporter line Ntsr1-cre; Rosa-tdTomato (Bortone et al., 2014; Kim et al., 2014). Immunohistochemistry verified the elimination of FOXP2 from tdTomato+ neurons of Ntsr1-cre; Foxp2fx/fx mice (Figure 2A, inset). tdTomato-expressing cell bodies remained limited to layer 6 of the cerebral cortex across genotypes, consistent with the absence of gross changes in laminar patterns due to Foxp2 deletion. Confocal microscopy further revealed no overt changes in the distribution of tdTomato-labeled neurites in more superficial layers of cortex, suggesting minimal morphological rearrangement of the CT neuron population in response to Foxp2 deletion. Inspection of tdTomato-labeled efferent axons arising from the deep layer neurons revealed a nearly identical pattern of CT innervation across Foxp2 genotypes (Figure 2A). There was normal fasciculation within the internal capsule and extensive axonal elaboration in the dorsal thalamus and thalamic reticular nuclei of Foxp2 conditional knockout mice, their heterozygous littermates and Ntsr1-cre; Rosa-tdTomato mice on a wild-type C57Bl/6J background. Consistent with the reported cell-type specificity of the Ntsr1-cre driver line, no tdTomato-expressing axons in the corpus callosum or cerebral peduncle were evident for any Foxp2 genotype at the ages examined. This suggests that FOXP2 is dispensable for the typical positioning of CT neurons in layer 6 and the guidance of their axons to reach and ramify in their normal dorsal thalamic targets.

Figure 2. FOXP2 is nonessential for class-specific anatomical and molecular phenotypes of corticothalamic neurons.

(A) At P14 tdTomato (magenta) expression in layer 6 corticothalamic neurons of Ntsr1-cre; Rosa-tdTomato mice reveals similar organization of corticothalamic innervation in Foxp2 conditional knockout mice and heterozygous littermates – boxed inset shows removal of FOXP2 protein (green) from tdTomato+ corticothalamic neurons of Ntsr1-cre; Foxp2Fx/Fx mouse. (B) MetGFP (green) and tdTomato (magenta) label distinct cell populations in Foxp2 conditional knockout mice, heterozygous littermates, and wild-type C57Bl/6J mice. (C) Quantification of co-expression of GFP by tdTomato+ corticothalamic neurons across Foxp2 genotypes (WT = Ntsr1 cre; Rosa-tdTomato, no Flox alleles, N = 3 mice; cHET = Ntsr1 cre; Foxp2Fx/+, N = 2 mice; cKO = Foxp2 Fx/Fx, N = 3 mice). (D) ppCCK expression is excluded from layer 6 corticothalamic neurons across Foxp2 genotypes as indicated by the segregation of tdTomato (magenta) and ppCCK (cyan). (E) Quantification of co-expression of CCK by tdTomato+ corticothalamic neurons (N for each group same as panel C). (F) FOG2 expression by corticothalamic neurons does not require Foxp2, as nearly all tdTomato+ cells express FOG2 (green) across Foxp2 genotypes. (G) Quantification of FOG2 coexpression by tdTomato+ corticothalamic neurons (N for each group same as panel C). All scale bars, 50 µm. Abbreviations: Ctx, cortex; Hpc, hippocampus; Thal, thalamus; TRN, thalamic reticular nucleus.

Figure 2—source data 1. Expression of Layer 6 cell-type markers in Ntsr1-cre; Foxp2Fx mice and controls.
elife-42012-fig2-data1.xlsx (10.1KB, xlsx)
DOI: 10.7554/eLife.42012.008

Figure 2.

Figure 2—figure supplement 1. Developmental timing of Cre-mediated recombination in Ntsr1-cre and Emx1-cre.

Figure 2—figure supplement 1.

(A) Cre-dependent tdTomato (magenta) expression begins at E17.5 (N = 3) in the cortex of Ntsr1-cre mice, when there are abundant FOXP2+ (green) neurons in the subplate and layer 6 that do not yet express tdTomato. (B) Cre-dependent tdTomato (red) expression begins at E10.5 (N = 3) in the cortex of Emx1-cre mice, where the majority of dorsal pallium (DP) progenitors are tdTomato+. (CP, cortical plate; IZ, intermediate zone; Str, striatum; SVZ, subventricular zone; VZ, ventricular zone; WM, white matter). Scale Bars: 100 µm.
Open in a new tab

Recent studies have identified neuronal target genes directly regulated by FOXP2, many of which are directly repressed upon FOXP2 binding (Spiteri et al., 2007; Konopka et al., 2009; Mukamel et al., 2011; Vernes et al., 2011). Two such genes, Cck and Met, display largely non-overlapping expression with FOXP2 within layer 6 (Figure 1H–J). Given the repressive effect of FOXP2 on Met and Cck expression (Spiteri et al., 2007; Mukamel et al., 2011; Vernes et al., 2011), the hypothesis that FOXP2 is required to prevent ectopic expression of Met and Cck among CT neurons was tested. Mice carrying Ntsr1-cre; Rosa-tdTomato and MetGFP reporter alleles were bred with Foxp2fx mice. The fraction of tdTomato+ CT neurons that co-expressed GFP was then quantified in Foxp2 conditional knockout mice, their heterozygous littermates and Ntsr1-cre; Rosa-tdTomato mice on a wild-type C57Bl/6J background at P14. Unexpectedly, the percentage of GFP and tdTomato double-positive neurons was minimal and indistinguishable across genotypes (Figure 2B,C; one-way ANOVA, p=0.441; Ntsr1-cre; Foxp2+/+ (WT) mean ± SEM = 1.9 ± 0.4; Ntsr1-cre; Foxp2Fx/+ (cHET), mean ± SEM = 2.7 ± 0.1; Ntsr1-cre; Foxp2Fx/Fx (cKO), mean ± SEM = 2 ± 0.5), indicating that the exclusion of Met expression from CT neurons occurs independent of transcriptional regulation by FOXP2. Next, to determine whether FOXP2 is required to repress CCK expression in CT neurons, colocalization of CCK and tdTomato was quantified. Despite abundant CCK expression among layer 6 neurons, there was minimal co-expression of CCK by tdTomato+ neurons across Foxp2 genotypes (Figure 2C,D; one-way ANOVA, p=0.5016; Ntsr1-cre; Foxp2+/+ (WT) mean ± SEM = 0.6 ± 0.4; Ntsr1-cre; Foxp2Fx/+ (cHET), mean ± SEM = 1.2 ± 0.4; Ntsr1-cre; Foxp2Fx/Fx (cKO), mean ± SEM = 0.6 ± 0.4), consistent with selective CCK expression by layer 6 CC neurons and exclusion from CT neurons, as previously reported (Watakabe et al., 2012; Kast et al., 2019). These data indicate that FOXP2 is not required for the exclusion of CCK and Met expression from CT neurons. To determine whether molecular markers unique to CT neurons continue to be expressed in their normal pattern in the absence of FOXP2, labeling of FOG2 among tdTomato-expressing CT neurons was assessed in Foxp2 conditional knockouts and their heterozygous littermates. FOG2 immunolabeling was detected in nearly 100% of CT neurons, independent of Foxp2 genotype (Figure 2E,F; one-way ANOVA, p=0.3163; Ntsr1-cre; Foxp2+/+ (WT) mean ± SEM = 97.1 ± 0.1; Ntsr1-cre; Foxp2Fx/+ (cHET), mean ± SEM = 97.5 ± 0.7; Ntsr1-cre; Foxp2Fx/Fx (cKO), mean ± SEM = 96.1 ± 0.7).

Because Ntsr1-cre is selectively expressed in CT neurons, it is likely Cre expression begins postmitotically, but the timing of the developmental onset of Cre-mediated recombination in the Ntsr1-cre mouse line has not been reported. Given the normal development of CT neurons in Ntsr1-cre; Foxp2Fx conditional knockout mice, we reasoned that the potentially late timing of the developmental deletion of Foxp2 could have influenced the lack of abnormal CT phenotypes. This possibility was investigated first by determining the onset of Cre-dependent tdTomato expression in Ntsr1-cre; Rosa-tdTomato embryos. Coronal sections of the embryonic forebrain were analyzed on embryonic days (E)14.5, E16.5, and E17.5. Expression of tdTomato was not detected until E17.5, when it was localized in a sparse population of subplate and layer 6 neurons (Figure 2—figure supplement 1). Thus, cre-mediated recombination in the Ntsr1-cre line does not begin until approximately E17, well after Foxp2 expression is initiated in the cerebral cortex (Figure 2—figure supplement 1; Ferland et al., 2003), and at a developmental time point only shortly preceding the initial innervation of the thalamus by descending CT axons (Grant et al., 2012; Torii et al., 2013). Thus, it is possible that Foxp2 operates in an earlier developmental window, prior to Ntsr1-cre mediated recombination and the subsequent depletion of previously transcribed and translated Foxp2/FOXP2. To determine whether Foxp2 might play a role earlier in the development of cortical neurons, the Emx1-cre driver line was employed, which exhibits cre-mediated recombination in dorsal pallial progenitors beginning at E10.5, when they have just started to proliferate (Gorski et al., 2002). Evaluation of Emx1-cre embryos confirmed cre-dependent tdTomato expression by E10.5 (Figure 2—figure supplement 1B). In situ hybridization revealed the selective removal of the floxed portion of the Foxp2Fx allele, which encodes the DNA-binding domain, at E14.5 (Figure 3A), prior to FOXP2 protein production in the dorsal pallium (Figure 1A,B). FOXP2 immunohistochemistry performed on sections of the E16.5 forebrain demonstrated the absence of FOXP2 protein in Emx1-cre; Foxp2Fx embryos at the earliest timepoint that FOXP2 could be detected in Foxp2Fx embryos (Figure 3B, Figure 3—figure supplement 1B). Thus, the removal of Foxp2 prior to the expression of FOXP2 by any dorsal pallial cells, using Emx1-cre, provided the opportunity to assess the developmental role of Foxp2 function from the earliest stages of cortical development and in the subset of layer 5 PT neurons that normally express FOXP2 (Figure 1—figure supplement 1), but that do not express Ntsr1-cre.

Figure 3. FOXP2 is nonessential for the genesis of cortical neurons and their proper lamination.

(A) Foxp2 in situ hybridization based on the BaseScope method reveals expression of Foxp2 transcript (Red) in Foxp2Fx/Fx embryos (N = 4) and selective removal of exons 12–14 (DNA-binding domain) from the dorsal pallium including the cortical plate (white arrowhead) of Emx1-cre; Foxp2Fx/Fx mice (N = 6) by E14.5. (B) Immunohistochemical analysis of FOXP2 protein in E16.5 embryos (N = 3 each genotype) demonstrates selective elimination of FOXP2 protein (green) from the infragranular layers of the dorsal pallium of Emx1-cre;Foxp2Fx/Fx mouse embryos. (C) FOXP2 immunohistochemistry on coronal sections of P0 Foxp2Fx/Fx and Emx1-cre; Foxp2Fx/Fx mice reveals absence of FOXP2 (black) in the infragranular cortical layers (red bracket) of Emx1-cre; Foxp2Fx/Fx mice.Inset (red outline) shows selective loss of FOXP2 in the cortex at higher magnification. (D) FOG2 (green) and CTIP2 (magenta) immunohistochemistry in coronal sections of the primary somatosensory cortex of Foxp2Fx/Fx and Emx1-cre; Foxp2Fx/Fx mice reveals similar distributions of laminar specific markers at P0. (E) Quantification of FOG2+ cells in layer 6 across genotypes at P0 (Fx/Fx, Foxp2fx/fx, N = 7; cHET, Emx1-cre; Foxp2fx/+, N = 6; cKO, Emx1-cre;Foxp2fx/fx, N = 7; Emx1-cre, N = 6). (F) Quantification of CTIP2+/Fog2- cells in layer 5 across genotypes at P0 (N for each group, same as panel E). (G) ppCCK (cyan) and PCP4 (magenta) immunohistochemistry in coronal sections of conditional knockout and control littermates at P14. (H) Quantification of ppCCK+ cells in layer 6 of SSC across genotypes at P14 (Foxp2Fx/Fx, N = 3 mice; Emx1-cre; Foxp2Fx/Fx, N = 4 mice). (I) Quantification of FOG2+ cells in layer 6 of SSC across genotypes at P14 (Foxp2Fx/Fx, N = 3 mice; Emx1-cre; Foxp2Fx/Fx, N = 4 mice). (J) DAPI-staining of coronal sections of Foxp2Fx/Fx and Emx1-cre; Foxp2Fx/Fx mice reveals similar size of cortex, including the thickness of primary somatosensory cortex (indicated by cyan bracket). (K) Quantification of somatosensory cortex thickness across genotypes at P14 (Foxp2Fx/Fx, N = 3 mice; Emx1-cre; Foxp2Fx/Fx, N = 4 mice). Scale Bars: A(inset), 100 µm; B, E, 50 µm; H, 500 µm.

Figure 3—source data 1. Quantification of cortical cell type numbers in Emx1-cre; Foxp2Fx mice and controls.
elife-42012-fig3-data1.xlsx (16.3KB, xlsx)
DOI: 10.7554/eLife.42012.011

Figure 3.

Figure 3—figure supplement 1. Elimination of Foxp2 transcript and protein from the forebrain of Emx1-cre; Foxp2Fx/Fx embryos.

Figure 3—figure supplement 1.

(A) BaseScope RNAscope demonstrating selective removal of Foxp2 exons 12–14 (DNA-binding domain) from the dorsal pallium at E14.5 (note that high magnification insets are the same images as Figure 3A). (B) FOXP2 immunohistochemistry demonstrates selective removal of FOXP2 protein from the infragranular layers (red bracket) of the dorsal pallium at the earliest timepoint that it can be detected, E16.5. Scale bar = 500 µm.
Open in a new tab

Foxp2 is reportedly important for cortical neurogenesis and cell migration (Tsui et al., 2013; Garcia-Calero et al., 2016); these ontogenetic events thus were evaluated in Emx1-cre; Foxp2Fx/Fx conditional knockout mice. FOXP2 immunohistochemistry validated the removal of Foxp2 from the cortex of Emx1-cre; Foxp2Fx/Fx mice at E16.5 (Figure 2—figure supplement 1C,D), at P0 (Figure 3A), as well as P4 and P14 (data not shown). Immunohistochemistry of the layer-specific markers FOG2, CTIP2 (Figure 3D) and DARPP-32 (data not shown) at P0 revealed normal patterns of lamination and cell-density in conditional knockout mice. There also were no overt differences in the thickness of layers as revealed by DAPI staining at P0 or P14 (Figure 3C,J). In addition, high-level expression of CTIP2 remained restricted to layer 5, and neurons expressing FOG2, a marker of CT neurons that are enriched for Foxp2, were found in their normal position, in layer 6 (Figure 3D). The numbers of CTIP2+/FOG2- cells in layer five were not different between genotypes at P0 (Figure 3E Kruskal-Wallis test, p=0.). The numbers of FOG2+ cells in layer 6 were comparable across genotypes (Figure 3F; Kruskal-Wallis test, p=0.0437, Foxp2Fx, mean ± SEM = 416 ± 28; cHET, mean ± SEM = 440 ± 68; cKO, mean ± SEM = 357 ± 37; Emx1-cre, mean ± SEM = 430 ± 52), as no statistically significant pairwise differences were observed between genotypes (Dunn’s Multiple Comparisons test: cKO vs. Foxp2Fx, p=0.3688; cKO vs. Emx1-cre, p=0.1353; cKO vs. cHET, p=0.0674; cHET vs. Foxp2Fx, p=0.9999; cHET vs. Emx1-cre, p=0.9999; Foxp2Fx vs. Emx1-cre, p=0.9999). Importantly, at P14, the numbers of FOG2+ layer 6 CT cells were similar in Emx1-cre; Foxp2Fx/Fx and Foxp2Fx/Fx littermates (Figure 3I, unpaired two-tailed t-test, p=0.6912 Foxp2Fx/Fx, mean ± SEM = 209 ± 21; Emx1-cre; Foxp2Fx/Fx (cKO), mean ± SEM = 199 ± 13), as were the numbers of CTIP2+/FOG2- cells in layer 5 (unpaired two-tailed t-test, p=0.99; Foxp2Fx/Fx, mean ± SEM = 77 ± 14; Emx1-cre; Foxp2Fx/Fx (cKO), mean ± SEM = 79 ± 7). These data demonstrate that the corticofugal populations that express FOXP2 during development do not require FOXP2 for their proper specification. Additionally, the number of ppCCK+ layer 6 CC cells was indistinguishable between conditional knockout and control groups (Figure 3G,H. unpaired two-tailed t-test, p=0.5178; Foxp2Fx/Fx, mean ± SEM = 139 ± 7; Emx1-cre; Foxp2Fx/Fx (cKO), mean ± SEM = 147 ± 9). Thus, the infragranular layers, which contain the neurons that express FOXP2, develop their normal complement of diverse projection neuron subtypes in normal numbers in the absence of Foxp2. Finally, the radial thickness of the SSC was indistinguishable across genotypes at P14 (Figure 3J,K, unpaired two-tailed t-test, p=0.583), suggesting that FOXP2 function also is dispensable to produce a histogenically and architecturally normal SSC.

Next, to evaluate the putative role of FOXP2 in axon guidance (Vernes et al., 2011), Emx1-cre; Rosa-tdTomato mice were crossed with Foxp2Fx mice. In agreement with the results from the Ntsr1-cre; Rosa-tdTomato experiments, tdTomato-labeled subcortical innervation revealed nearly identical patterns in the internal capsule, thalamus, cerebral peduncle, and pyramidal decussation across Foxp2 genotypes at P0 (data not shown) and P4 (Figure 4). This result, using a genetic deletion strategy prior to cortical neurogenesis, indicates that Foxp2 is dispensable for proper specification of the cortical neuron subtypes that normally express FOXP2 in SSC, and the appropriate guidance and targeting of their efferent axons.

Figure 4. FOXP2 is not required for proper corticofugal axon pathfinding.

Figure 4.

Open in a new tab

(A) tdTomato reporter (black) reveals similar patterns of corticofugal axon growth in sagittal sections of Emx1-cre;Foxp2+/+ (WT, top panel, N = 4) Emx1-cre; Foxp2Fx/+ (middle, N = 3) and Emx1-cre; Foxp2Fx/Fx (bottom, N = 3). Note the fasciculation of axons in the internal capsule (blue asterisks) and the comparable growth of axons into the thalamus (blue arrows) and pyramidal decussation (red arrowheads). (B) Higher magnification images of the corticothalamic innervation patterns in each genotype. Scale Bars = 500 µm.

Discussion

The current study focused on putative roles for FOXP2 in murine cortical histogenesis, using conditional mouse genetics. The results demonstrate that Foxp2 is not required for establishing basic developmental organization, molecular phenotypes or efferent connectivity of Foxp2-expressing neurons in SSC of mice. The role of FOXP2 in neural development has been of significant interest following the identification of FOXP2 mutations that cause developmental apraxia of speech in humans (Lai et al., 2001; MacDermot et al., 2005). Diverse methods and genetic models have been used to interrogate FOXP2 function in a variety of brain areas and species (French and Fisher, 2014). Notably, recent studies primarily in mice have implicated Foxp2 in many developmental processes including cortical neurogenesis (Tsui et al., 2013), neuronal migration (Garcia-Calero et al., 2016), neuron subtype specification (Chiu et al., 2014), neural tissue patterning (Ebisu et al., 2017), neurite outgrowth and axon guidance (Vernes et al., 2011), synapse formation (Chen et al., 2016), and synaptic plasticity (Groszer et al., 2008). However, assessment of Foxp2 function during cerebral cortical development by means of conditional mouse genetics had not been thoroughly pursued. Here, Foxp2 was deleted at different prenatal ages using dorsal pallial- and cell-type specific Cre-recombinase mouse lines.

FOXP2 and cortical projection neuron phenotypes

FOXP2 expression is enriched in the deepest layers of the developing and mature neocortex of mammalian species ranging from mice to humans (Ferland et al., 2003; Campbell et al., 2009; Mukamel et al., 2011), suggesting conservation of the cortical cell types that utilize the gene. The connectivity and molecular phenotyping data generated here demonstrate that, in mice, FOXP2 expression within the infragranular layers of the developing SSC is present in nearly all CT neurons, excluded from most layer 6 CC neurons, and transiently expressed by a minor subset of PT neurons. This is consistent with observations of enrichment of Foxp2 in CT neurons in the primary visual cortex of adult mice by single cell RNA-sequencing and immunohistochemistry (Tasic et al., 2016; Sundberg et al., 2018), and previous findings that FOXP2 is expressed by subsets of layer five neurons (Ferland et al., 2003; Hisaoka et al., 2010; Molyneaux et al., 2015), which we show here are PT-type neurons. Analysis of SSC at three postnatal ages demonstrated that expression of FOXP2 by CT-type neurons is stable, whereas PT-type neurons express FOXP2 transiently, with no detectable FOXP2 expression in most PT-type neurons at P14. The developmental cell-type selectivity of FOXP2 expression raises important questions regarding the function of FOXP2.

With the enrichment of FOXP2 in CT neurons, and very limited expression in CC neurons, the current study addressed whether Foxp2 expression is required to repress expression of two putative target genes, Cck and Met, which are generally excluded from CT neurons. Such a role would be consistent with the non-overlapping expression patterns of these genes with Foxp2 in layer 6, as well as previous reports of direct repressive control by FOXP2 in vitro (Spiteri et al., 2007; Mukamel et al., 2011; Vernes et al., 2011). However, deletion of Foxp2 failed to alter expression patterns of Met or Cck, suggesting other transcriptional mechanisms may mediate their cell type-specific expression in infragranular layers in vivo. Additionally, two important molecular features unique to FOXP2+ neurons, FOG2 and DARPP-32 expression, are unchanged following either very early or late prenatal deletion of Foxp2. Much broader molecular profiling is warranted, but the data indicate that Foxp2 is neither required for the specification of some of the unique molecular features of FOXP2+ layer 6 CT neurons, nor for the regulation of alternate cell-type molecular signatures that were predicted from previous analysis of FOXP2 transcriptional regulatory targets.

FOXP2 and early cortical neuron development

Using the early expressing Emx1-cre driver line (E10.5), the data show that Foxp2 deletion does not disrupt the normal generation and migration of neurons in SSC. Thus, unlike other transcriptional regulators of cell-type identity (e.g. CTIP2, FEZF2, TBR1), for which dramatic changes in cell type numbers and projection phenotypes develop upon mutation (Arlotta et al., 2005; Chen et al., 2005; Han et al., 2011; McKenna et al., 2011; Molyneaux et al., 2005), FOXP2 appears dispensable for the general production of cortical neurons and the specification of the specific projection populations that normally express FOXP2. This conclusion is consistent with the absence of developmental defects recently reported in Nex1-cre; Foxp2Fx mice, in which Foxp2 is deleted from postmitotic excitatory neurons of the dorsal pallium (Medvedeva et al., 2018). These results were unexpected given the data in previous studies using in utero electroporation (IUEP)-mediated Foxp2 shRNA knockdown, which demonstrated atypical cortical neurogenesis and migration (Tsui et al., 2013; Garcia-Calero et al., 2016). In fact, using IUEP and identical shRNA reagents, we observed similar phenotypes to those previously reported (data not shown) (Tsui et al., 2013). Several explanations, based on distinct methodologies, could account for the discrepant findings using IUEP knockdown compared to hetero- and homozygous genetic deletion. Recent studies have revealed compensatory transcriptional responses by orthologous transcripts in some mutant mouse lines, and thus it is possible that other Foxp family members could compensate for Foxp2 removal in our studies (discussed more below). Additionally, when crossed with the Foxp2Fx conditional allele, Emx1-cre leads to uniform removal of Foxp2 from dorsal pallial progenitors, whereas in utero electroporation reduces gene expression in a much smaller subpopulation of cells. This generates a mosaic of Foxp2-positive and negative cells. The altered neurogenesis phenotype observed following IUEP mediated-knockdown could result from the removal of Foxp2 in a mosaic fashion in the dorsal pallium resulting in atypical interactions between neighboring FOXP2+ and FOXP2- cells. A similar mosaic effect could explain the altered migration observed following IUEP of Foxp2 shRNA (Tsui et al., 2013). Mosaic expression of mutant and wild-type alleles can influence cortical development, shown recently in mouse models of X-linked Pcdh19 epilepsy (Pederick et al., 2018). Resolving whether mosaic Foxp2 expression can disrupt neurogenesis and migration will require many additional studies using methods distinct from shRNA knockdown, such as in utero electroporation of Cre-expressing plasmid constructs into Foxp2Fx/Fx embryos. We note, however, the importance of determining the relevance of such mosaic effects, may depend on the identification of a natural context, in humans or developing mouse models, in which mosaic Foxp2 function occurs in the cortex.

An alternative explanation is that the Foxp2-targeting shRNA used in ours and previous studies may lead to incomplete reduction of Foxp2 expression. This would in turn result in different adaptive responses compared to complete deletion of Foxp2 genetically. Similar mechanistic explanations have been posed for the discrepant observations of germline versus IUEP-mediated manipulation of doublecortin (Bai et al., 2003). Noteworthy is the finding that ectopic overexpression of Foxp2 results in a paradoxically similar arrest in the radial migration of cortical neurons caused by Foxp2 shRNA knockdown (Clovis et al., 2012). While no direct evidence currently exists, non-physiological mosaic reduction or overexpression of FOXP2 could create an imbalance that itself disrupts cortical development, while genetic disruptions fail to produce the same phenotypes. Finally, shRNA knockdown does have the technical caveat of potential ‘off-target’ effects, which cannot be ruled out unequivocally. For example, it is possible that the Foxp2 shRNAs also impact the expression of other genes including other Foxp family members, such as Foxp1, which are expressed in the cortex and could hypothetically compensate for Foxp2 function in our knockout studies. However, in vitro analysis of the specificity of the shRNA constructs suggested that they did not alter the levels of Foxp1 or Foxp4 expression (Tsui et al., 2013). Additionally, it is noteworthy that, in a previous study, quantification of Foxp1 and Foxp4 transcript levels demonstrated that these genes were expressed at similar levels in WT and Foxp2 conditional knockout embryos (French et al., 2007). Thus, recombination of the Foxp2Fx allele does not necesarrily trigger genetic compensation by other Foxp family members. Although unidentified compensatory mechanisms may mask a dispensable role that Foxp2 plays in the core aspects of cortical development studied here, based on results obtained following complete genetic removal of Foxp2 function, we conclude that Foxp2 does not play an essential role in murine SSC neurogenesis, neuron migration, subtype specification or axonal pathfinding, contrary to conclusions of other studies.

Functional implications of FOXP2 in developmental disorders

The lack of overt changes in the generation, migration, differentiation, or axon pathfinding of SSC neurons following conditional Foxp2 deletion, using two different Cre driver lines, is important to consider in several contexts. The results may have implications for understanding the involvement of the cerebral cortex in speech and language impairments associated with FOXP2 mutations in humans (Vargha-Khadem et al., 2005). The results suggest that loss of FOXP2 function likely does not contribute to these deficits through altered cortical histogenesis. However, it is important to note that the present studies were carried out in mice, and thus it remains possible that novel species-specific roles for Foxp2 may have been acquired in the human lineage. Nonetheless, the results provide foundational knowledge that will be essential when designing studies to further address the role of FOXP2 in the development and function of specific neuronal cell types in the cortex.

The normal development of cortical phenotypes in Foxp2 conditional knockout mice is consistent with conventional magnetic resonance imaging of brain structure in patients with FOXP2 mutations, which did not identify substantive alterations in gray and white matter structure of the cerebral cortex (Vargha-Khadem et al., 2005). However, more refined analysis using voxel-based morphometric analysis identified spatially restricted, minor alterations in the gray matter in perisylvian cortical areas of patients with FOXP2 mutations (Belton et al., 2003). Area-restricted deficits in the development of cerebral cortical anatomy may occur in Foxp2 conditional knockout mice, but were not detected in the present study due to the focus of the analysis on primary SSC. More expansive studies will need to be pursued. In addition, given the lack of speech and language homologous regions in the cerebral cortex of mice, discovery of regional disruptions relevant to humans with FOXP2 mutations may not be possible.

Here, the genetic deletion of Foxp2 in mice is distinct from the most common mutations observed in human patients with inherited speech and language abnormalities (Lai et al., 2001; MacDermot et al., 2005). Cre-mediated recombination of the conditional Foxp2 allele produces a nonsense mutation that eliminates DNA-binding ability and causes near complete loss of FOXP2 protein (French et al., 2007). Functionally analogous, truncating FOXP2 mutations have been identified in some patients with speech and language disorders (MacDermot et al., 2005). However, missense mutations like the one in the KE pedigree are more common and are the best characterized in terms of their associated brain abnormalities (Vargha-Khadem et al., 1998; Lai et al., 2001). Thus, the conditional knockout mice used here may not fully recapitulate aberrant FOXP2 functions caused by single amino acid changes, which are proposed to elicit dominant-negative functions that could be distinct from the simple loss-of-function caused by conditional Foxp2 deletion (Tsui et al., 2013). For example, missense FOXP2 mutations could lead to gain-of-function impairments by influencing the activity of other transcription factors. Importantly, transgenic mouse models that carry the same missense mutations as those observed in human populations have been generated and functionally, but not developmentally, characterized (French and Fisher, 2014). Irrespective of the differences in the genetic strategies used to disrupt Foxp2, the present results strongly suggest that the histogenesis of murine SSC does not depend on transcriptional regulation by FOXP2.

Finally, in Foxp2 constitutive knockout mice, medium spiny neurons of the striatum display decreased mEPSC frequency, decreased dendritic spine density, and increased mEPSC amplitudes, whereas the macro-level organization and cell-type composition of the striatum remains intact (Chen et al., 2016). It would be interesting to investigate whether the cortical neurons that express Foxp2 display similar synaptic abnormalities in Foxp2 knockout mice. Altered excitability of CT neurons could contribute to atypical activity within cortico-striato-thalamocortical loops that are important for motor control and information processing, which could significantly alter speech related functions. The demonstration that the histological organization of the somatosensory cortex is unaffected by the removal of Foxp2 warrants more detailed characterization of cortical circuit function in Foxp2 mutant mice.

Materials and methods

Key resources table.

Reagent type
(species) or
resource		 Designation Source or reference Identifiers Additional
information
Genetic reagent (M. musculus) Foxp2Fx (Foxp2tm1.1Sfis) French et al., 2007 MGI Cat# 3800702, RRID:MGI:3800702
Genetic reagent (M. musculus) Rosa-TdTomato (Ai14) Jackson Laboratory IMSR Cat# JAX:007914, RRID:IMSR_JAX:007914
Genetic reagent (M. musculus) Ntsr1-cre (GN220) MMRRC MMRRC Cat# 030648-UCD, RRID:MMRRC_030648-UCD
Genetic reagent (M. musculus) Emx1-cre Gorski et al., 2002 IMSR Cat# JAX:005628, RRID:IMSR_JAX:005628
Genetic reagent (M. musculus) MetEGFP BAC (MetGFP) Gong et al., 2003 MGI:6144427
Antibody Goat anti-Foxp2 (polyclonal) Santa Cruz Biotechnology Cat# sc-21069, RRID:AB_2107124 IHC (1:100)
Antibody Chicken anti-GFP (polyclonal) Abcam Cat# ab13970, RRID:AB_300798 IHC (1:500)
Antibody Rat anti-Ctip2 (monoclonal) Abcam Cat# ab18465, RRID:AB_2064130 IHC (1:500)
Antibody Guinea Pig anti-ppCCK Watakabe et al., 2012 Dr. Takeshi Kaneko (University of Tokyo) IHC (1:500)
Antibody Rabbit anti-PCP4(PEP-19) Dr. James Morgan (St. Jude's Research Hospital) IHC (1:3000)
Antibody Rabbit anti-Fog2 (polyclonal) Santa Cruz Biotechnology Cat# sc-10755, RRID:AB_2218978 IHC (1:250)
Antibody Rabbit anti-DARPP-32 (monoclonal) Cell Signaling Technology Cat# 2306, RRID:AB_823479 IHC (1:500)
Antibody AlexaFluor F(AB')2 488- or 594- or 647- secondaries Jackson Immunoresearch Laboratories, Inc IHC (1:500)
Peptide, recombinant protein Cholera Toxin Subunit B (CTB) Invitrogen Cat. #: C-34776, C-34778 Alexa Fluor Conjugate (555, 647)
Commercial assay or kit BaseScope assay Advanced Cell Diagnostics Cat. #: 323971
Software, algorithm IMARIS Imaris (http://www.bitplane.com/imaris/imaris) RRID:SCR_007370 Microscopy Image Analysis Software
Software, algorithm Adobe Photoshop Adobe Photoshop (https://www.adobe.com/products/photoshop.html) RRID:SCR_014199
Software, algorithm Adobe Illustrator (CS6) Adobe Illustrator (http://www.adobe.com/products/illustrator.html) RRID:SCR_010279
Software, algorithm Code used for nuclear immunofluorescent quantification This paper custom written for ImageJ macros (Source Code 1; Source Code 2)
Software, algorithm ImageJ ImageJ (http://imagej.nih.gov/ij/) RRID:SCR_003070
Software, algorithm GraphPad Prism 6 GraphPad Prism (https://graphpad.com) RRID:SCR_015807 Version 6
Open in a new tab

Animals

All animal procedures used in this study were in accordance with the guidelines of the Institutional Animal Care and Use Committee at Children’s Hospital Los Angeles. Mice were housed on a 13:11 hr light-dark cycle, and were provided with food and water ad libitum. Mice harboring the conditional Foxp2 allele (Foxp2Fx; French et al., 2007), Rosa-TdTomato allele (Ai14), the Ntsr1-Cre transgene (GN220), or the Emx1-Cre transgene (B6.129S2-Emx1tm1(cre)Krj/J; obtained from Jackson Laboratories) were maintained on an isogenic C57Bl/6J background. MetEGFP BAC transgenic (MetGFP) mice were re-derived on the FVB background using the BX139 BAC from the GENSAT collection (Gong et al., 2003). Founder mice were backcrossed to C57Bl/6J for at least two generations prior to experimental breeding, such that experiments involving MetGFP mice were carried out on a mixed C57Bl/6J x FVB background. Emx1-Cre first exhibits recombination of floxed loci at embryonic day (E) 10 in mice (Gorski et al., 2002), as confirmed in the present study. Based on data reported here, Ntsr1-Cre exhibits recombination initially at E17.

Retrograde tracing

On postnatal day (P) 12, mice were anesthetized with vaporized isoflurane (5% induction, 1.5–2% maintenance) and stabilized in a Narishige SG-4N small animal head holder. Mice were maintained at 37 ˚C for the duration of the surgical procedure through a TCAT-2 temperature control device (Physitemp Intruments, Inc) and respiratory rate was continuously monitored to assess depth of anesthesia. Through stereotaxic guidance, a picospritzer connected to a pulled borosilicate pipette (28 µm tip diameter) was used to inject 50–100 nl of Cholera Toxin Subunit B, Alexa Fluor Conjugate (Invitrogen) into the desired cortical or subcortical target. To minimize contamination of unintended brain regions along the needle tract, the pipette was left in place for 5 min before being slowly retracted. Mice received a subcutaneous injection of the non-steroidal anti-inflammatory drug (NSAID) ketoprofen (5 mg/kg) immediately before the surgery and provided with ibuprofen (0.2 mg/mL) in the drinking water until the end of the experiment. After 2 days of recovery, on P14, mice were transcardially perfused with 4% paraformaldehyde dissolved in 1X phosphate buffered saline (PBS) and tissue was processed for immunohistochemical analysis as described below. Stereotaxic coordinates used for P12 mice are as follows: ventrobasal thalamus, AP −1.7, ML 1.3 mm, Depth 3.15 mm; primary motor cortex, AP 0.25 mm, 1.5 mm, 0.9 mm; cerebral peduncle, AP −3.5 mm, 1 mm, 4.8 mm.

In situ hybridization

The BaseScope assay (Advanced Cell Diagnostics) was performed on embryonic brain sections prepared from Foxp2Fx/Fx and Emx1-cre; Foxp2Fx/Fx embryos that were harvested at embryonic day (E) 14.5, with noon on the day of vaginal plug (identified following overnight mating) designated as E0.5. Briefly, the pregnant dam was deeply anesthetized with saturated isoflurane vapor, and cervical dislocation ensured euthanasia prior to embryo dissection. Embryos were decapitated and the brains were submerged in optimal cutting temperature (OCT, Tissue-Tek) compound and frozen in a pre-chilled dry ice and isopropanol slurry, and subsequently stored at −80 ˚C until cryosectioning. 20 µm coronal cryosections were collected onto SuperFrost Plus (Fisher Scientific) microscope slides, and then stored at −80 ˚C until in situ hybrization procedures. Slide-mounted sections were removed from −80 ˚C and immediately fixed by submerging in prechilled 4% PFA in 1X DEP-C PBS for 30 min on ice with gentle agitation. Sections were dehydrated in 50%, 70%, and two 100% EtOH washes at room temperature for 5 min each. Tissue was stored in 100% EtOH overnight at −20°C. Sections were pretreated with RNAscope Hydrogen Peroxide for 10 min at room temperature and then with RNAscope Protease IV for 15 min at room temperature. Tissue was washed in 1X DEP-C PBS at room temperature. A custom Foxp2 BaseScope probe (that hybridizes to the floxed region, bases 1832–1977 of Foxp2 transcript variant 2 (Refseq ID NM_212435.1), of the Foxp2Fx allele) was hybridized for 2 hr at 40°C. Sections were washed twice with 1X wash buffer. Amplification and signal detection steps followed the protocol provided in the BaseScope users manual. Slides were counterstained in 25% Hematoxylin solution modified according to Gill III for 2 min at room temperature. Slides were washed in H2O, and then in 0.02% ammonia for 15 s, and H2O once more. Slides were then incubated at 55°C for 15 min and then mounted in VectaMount (Vector Laboratories). Slides were stored at −20 ˚C until imaged with aLeica DFC295 color camera using brightfield microscopy through a 20x objective lens.

Immunohistochemistry

Somatosensory cortex (SSC) was the focus of all data analyses reported in the present study. Embryos were decapitated and brains were either immediately submerged in OCT and frozen in a pre-chilled dry ice and isopropanol slurry (‘fresh frozen’) and stored at −80 ˚C until cryosectioning, or brains were transferred to 4% paraformaldehyde dissolved in PBS (pH 7.4) and incubated at 4 ˚C for 12–18 hr. Early postnatal (P0) mouse brains were dissected in room temperature PBS, transferred to 4% paraformaldehyde dissolved in PBS (pH 7.4) and incubated at 4 ˚C for 12–18 hr. Mice aged P4 or older were perfused transcardially with 4% paraformaldehyde dissolved in PBS. Following perfusion, brains were immediately removed, transferred to 4% paraformaldehyde and incubated at 4 ˚C for 12–18 hr. Following overnight fixation, brains were incubated sequentially in 10%, 20% and 30% sucrose dissolved in PBS for 12–24 hr each. Next, brains were embedded in Clear Frozen Section Compound (VWR International) and placed on a weigh boat floating in liquid nitrogen. Once frozen, embedded brains were stored at −80 ˚C until cryosectioning. Twenty µm coronal or sagittal cryosections were cut and collected on SuperFrost Plus slides (Fisher Scientific) at −20 ˚C (PFA fixed tissue) or −15 ˚C (fresh frozen), and then stored at −80 ˚C until immunohistochemical analysis. Before immunostaining, fresh frozen sections were thawed to room temperature, fixed in 4% paraformaldehyde dissolved in PBS at room temperature with agitation for 1 hr and 40 min, and then washed in PBS three times for five minutes each. For immunostaining, sections were warmed at room temperature for 10 min, dried in a hybridization oven at 55 ˚C for 15 min, and then incubated in PBS for 10 min. Blocking and permeabilization were done by incubating sections in PBS containing 5% normal donkey serum and 0.3% Triton X-100 for 1 hr at room temperature. Sections were incubated subsequently in primary antibodies diluted in 0.1% Triton X-100 in PBS overnight at room temperature. Sections were washed five times for five minutes each with 0.2% Tween 20 in PBS. Sections were incubated in Alexa Fluor conjugated secondary antibodies (1:500) diluted in 0.1% Triton X-100 in PBS for 1 hr at room temperature. Sections were washed three times for five minutes each with 0.2% Tween 20 in PBS. Sections were then incubated in 950 nM DAPI in PBS for 8 min, and then subjected to two additional five minute PBS washes. Sections were mounted in Prolong Gold antifade reagent (Life Technologies), and the mounting medium cured for at least 24 hr before collecting confocal microscopy images. Primary antibodies used were as follows: Goat anti-Foxp2 (1:100; Santa Cruz sc-21069), Chicken anti-GFP (1:500; Abcam #ab13970), Rat anti-Ctip2 (1:500; Abcam # ab18465), Guinea Pig anti-ppCCK (1:500; T. Kaneko Lab), Rabbit anti-PCP4 (1:3000; J. Morgan Lab), Rabbit anti-Fog2 (1:250; Santa Cruz sc-10755), Rabbit anti-DARPP-32 (1:500; Cell Signaling #2306).

Co-localization analyses

Co-localization analysis in SSC was performed as described previously (Kast et al., 2019). Briefly, confocal images were collected through a 20x/0.8NA Plan-APOCHROMAT objective lens mounted on a Zeiss LSM 700 confocal microscope with refractive index correction. Optical sections were collected at 1 A.U. and 2 µm z-steps through the entire thickness of each 20 µm section. Colocalization analysis was performed in three-dimensional renderings of each confocal z-stack using IMARIS software (Bitplane).

Cortical thickness measurements

Fluorescent images of DAPI-stained sections containing SSC were collected through a 5x objective lens mounted on an Axionplan II upright fluorescent microscope (Zeiss), an Axiocam MRm camera (Zeiss) and Axiovision software 4.1 (Zeiss). The images were opened in ImageJ and three lines separated by ≥50 µm were drawn in the posteromedial barrel subfield from the pia to the white matter. The length of the nine lines (3 lines x three images) were averaged to give a value for the radial thickness of SSC for each mouse.

Cell type quantification

Maximum Z-projections were created from confocal z-stacks and custom written ImageJ macros were run to quantify the number of FOG2 +and CTIP2+/FOG2- nuclei at P0. The numbers of FOG2+ and CTIP2+/FOG2- cells for each animal were averaged from three 300 µm wide fields (each separated rostrocaudally by ≥200 µm) of the cortex representing the anterior, middle and posterior portions of SSC. The numbers of CCK+ cells within layer 6 at P14 were manually counted by an observer blind to genotype, as the punctate and discontinuous distribution of the immunofluorescent ppCCK signal prevented accurate automated quantitation. Similarly, numbers of FOG2+ and CTIP2+/FOG2- nuclei at P14 were manually quantified by an observer blind to genotype, due to challenges in automated detection of the lower level expression at this time point.

Image adjustments and figure preparation

Figures were prepared using Adobe Photoshop and Adobe Illustrator (CS6). Only linear adjustments (i.e. gamma = 1.0) were made to the contrast and signal levels of fluorescence microscopy images, and were done in an identical manner across genotypes.

Experimental design and statistics

Numbers of biological replicates (number of animals) for each experiment are included in the figure legends. Numbers of animals in each group were chosen in accordance with group numbers in previous publications reporting differences in murine cortical phenotypes similar to those measured in the current study (Han et al., 2011; Woodworth et al., 2016). Summary statistics and specific statistical tests used are described in the Results section. Parametric tests were used in some cases, although tests for normality were not possible given the modest sample sizes. Statistical analyses were performed using Prism 6 (GraphPad).

Acknowledgements

We thank Dr. Simon Fisher and Dr. Catherine French for generously sharing the Foxp2Fx/Fx mice. We thank Dr. James Morgan and Dr. Takeshi Kaneko for providing anti-PCP4 and anti-CCK antibodies, respectively. We thank Amanda Whipple for assistance with animal husbandry, and Dr. Esteban Fernandez of the Children’s Hospital Los Angeles Cellular Imaging Core for assistance with image acquisition and analysis.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Contributor Information

Ryan J Kast, Email: rkast@mit.edu.

Pat Levitt, Email: plevitt@med.usc.edu.

Constance L Cepko, Harvard Medical School, United States.

Huda Y Zoghbi, Texas Children's Hospital, United States.

Funding Information

This paper was supported by the following grants:

  • National Institute of Mental Health MH067842 to Ryan J Kast, Alexandra L Lanjewar, Colton D Smith, Pat Levitt.

  • Children's Hospital Los Angeles to Ryan J Kast.

  • Simms/Mann Institute and Foundation to Pat Levitt.

  • National Institutes of Health T32GM113859 to Alexandra L Lanjewar.

Additional information

Competing interests

No competing interests declared.

Author contributions

Conceptualization, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Visualization, Writing—original draft, Writing—review and editing.

Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—review and editing.

Validation, Visualization, Writing—review and editing.

Conceptualization, Resources, Data curation, Supervision, Funding acquisition, Project administration, Writing—review and editing.

Ethics

Animal experimentation: All animal procedures used in this study were in strict accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee at Children's Hospital Los Angeles (Protocol #374-15).

Additional files

Source code 1. Maximum Projection and Cropping Macro.
elife-42012-code1.ijm (761B, ijm)
DOI: 10.7554/eLife.42012.013
Source code 2. Cortical Cell Type Autocounting Macro for nuclear markers.
elife-42012-code2.ijm (1.4KB, ijm)
DOI: 10.7554/eLife.42012.014
Transparent reporting form
elife-42012-transrepform.docx (256KB, docx)
DOI: 10.7554/eLife.42012.015

Data availability

All numbers relating to quantitative experiments have been uploaded to Dryad (https://dx.doi.org/10.5061/dryad.6hd7bf7).

The following dataset was generated:

Kast R, Lanjewar A, Smith C, Levitt P. 2018. Data from: FOXP2 exhibits neuron class specific expression, but is not required for multiple aspects of cortical histogenesis. Dryad Digital Repository.

References

  1. Arlotta P, Molyneaux BJ, Chen J, Inoue J, Kominami R, Macklis JD. Neuronal subtype-specific genes that control corticospinal motor neuron development in vivo. Neuron. 2005;45:207–221. doi: 10.1016/j.neuron.2004.12.036. [DOI] [PubMed] [Google Scholar]
  2. Bai J, Ramos RL, Ackman JB, Thomas AM, Lee RV, LoTurco JJ. RNAi reveals doublecortin is required for radial migration in rat neocortex. Nature Neuroscience. 2003;6:1277–1283. doi: 10.1038/nn1153. [DOI] [PubMed] [Google Scholar]
  3. Belton E, Salmond CH, Watkins KE, Vargha-Khadem F, Gadian DG. Bilateral brain abnormalities associated with dominantly inherited verbal and orofacial dyspraxia. Human Brain Mapping. 2003;18:194–200. doi: 10.1002/hbm.10093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bortone DS, Olsen SR, Scanziani M. Translaminar inhibitory cells recruited by layer 6 corticothalamic neurons suppress visual cortex. Neuron. 2014;82:474–485. doi: 10.1016/j.neuron.2014.02.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Campbell P, Reep RL, Stoll ML, Ophir AG, Phelps SM. Conservation and diversity of Foxp2 expression in muroid rodents: functional implications. The Journal of Comparative Neurology. 2009;512:84–100. doi: 10.1002/cne.21881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chen B, Schaevitz LR, McConnell SK. Fezl regulates the differentiation and axon targeting of layer 5 subcortical projection neurons in cerebral cortex. PNAS. 2005;102:17184–17189. doi: 10.1073/pnas.0508732102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen YC, Kuo HY, Bornschein U, Takahashi H, Chen SY, Lu KM, Yang HY, Chen GM, Lin JR, Lee YH, Chou YC, Cheng SJ, Chien CT, Enard W, Hevers W, Pääbo S, Graybiel AM, Liu FC. Foxp2 controls synaptic wiring of corticostriatal circuits and vocal communication by opposing Mef2c. Nature Neuroscience. 2016;19:1513–1522. doi: 10.1038/nn.4380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chiu YC, Li MY, Liu YH, Ding JY, Yu JY, Wang TW. Foxp2 regulates neuronal differentiation and neuronal subtype specification. Developmental Neurobiology. 2014;74:723–738. doi: 10.1002/dneu.22166. [DOI] [PubMed] [Google Scholar]
  9. Clovis YM, Enard W, Marinaro F, Huttner WB, De Pietri Tonelli D. Convergent repression of Foxp2 3'UTR by miR-9 and miR-132 in embryonic mouse neocortex: implications for radial migration of neurons. Development. 2012;139:3332–3342. doi: 10.1242/dev.078063. [DOI] [PubMed] [Google Scholar]
  10. Ebisu H, Iwai-Takekoshi L, Fujita-Jimbo E, Momoi T, Kawasaki H. Foxp2 regulates identities and projection patterns of thalamic nuclei during development. Cerebral Cortex. 2017;27:3648–3659. doi: 10.1093/cercor/bhw187. [DOI] [PubMed] [Google Scholar]
  11. Ferland RJ, Cherry TJ, Preware PO, Morrisey EE, Walsh CA. Characterization of Foxp2 and Foxp1 mRNA and protein in the developing and mature brain. The Journal of Comparative Neurology. 2003;460:266–279. doi: 10.1002/cne.10654. [DOI] [PubMed] [Google Scholar]
  12. French CA, Groszer M, Preece C, Coupe AM, Rajewsky K, Fisher SE. Generation of mice with a conditional Foxp2 null allele. Genesis. 2007;45:440–446. doi: 10.1002/dvg.20305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. French CA, Jin X, Campbell TG, Gerfen E, Groszer M, Fisher SE, Costa RM. An aetiological Foxp2 mutation causes aberrant striatal activity and alters plasticity during skill learning. Molecular Psychiatry. 2012;17:1077–1085. doi: 10.1038/mp.2011.105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. French CA, Fisher SE. What can mice tell us about Foxp2 function? Current Opinion in Neurobiology. 2014;28:72–79. doi: 10.1016/j.conb.2014.07.003. [DOI] [PubMed] [Google Scholar]
  15. Garcia-Calero E, Botella-Lopez A, Bahamonde O, Perez-Balaguer A, Martinez S. FoxP2 protein levels regulate cell morphology changes and migration patterns in the vertebrate developing telencephalon. Brain Structure and Function. 2016;221:2905–2917. doi: 10.1007/s00429-015-1079-7. [DOI] [PubMed] [Google Scholar]
  16. Gong S, Zheng C, Doughty ML, Losos K, Didkovsky N, Schambra UB, Nowak NJ, Joyner A, Leblanc G, Hatten ME, Heintz N. A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature. 2003;425:917–925. doi: 10.1038/nature02033. [DOI] [PubMed] [Google Scholar]
  17. Gorski JA, Talley T, Qiu M, Puelles L, Rubenstein JL, Jones KR. Cortical excitatory neurons and glia, but not GABAergic neurons, are produced in the Emx1-expressing lineage. The Journal of Neuroscience. 2002;22:6309–6314. doi: 10.1523/JNEUROSCI.22-15-06309.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Grant E, Hoerder-Suabedissen A, Molnár Z. Development of the corticothalamic projections. Frontiers in Neuroscience. 2012;6:53. doi: 10.3389/fnins.2012.00053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Groszer M, Keays DA, Deacon RM, de Bono JP, Prasad-Mulcare S, Gaub S, Baum MG, French CA, Nicod J, Coventry JA, Enard W, Fray M, Brown SD, Nolan PM, Pääbo S, Channon KM, Costa RM, Eilers J, Ehret G, Rawlins JN, Fisher SE. Impaired synaptic plasticity and motor learning in mice with a point mutation implicated in human speech deficits. Current Biology. 2008;18:354–362. doi: 10.1016/j.cub.2008.01.060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Han W, Kwan KY, Shim S, Lam MM, Shin Y, Xu X, Zhu Y, Li M, Sestan N. TBR1 directly represses Fezf2 to control the laminar origin and development of the corticospinal tract. PNAS. 2011;108:3041–3046. doi: 10.1073/pnas.1016723108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Harris KD, Shepherd GM. The neocortical circuit: themes and variations. Nature Neuroscience. 2015;18:170–181. doi: 10.1038/nn.3917. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hisaoka T, Nakamura Y, Senba E, Morikawa Y. The forkhead transcription factors, Foxp1 and Foxp2, identify different subpopulations of projection neurons in the mouse cerebral cortex. Neuroscience. 2010;166:551–563. doi: 10.1016/j.neuroscience.2009.12.055. [DOI] [PubMed] [Google Scholar]
  23. Hoerder-Suabedissen A, Molnár Z. Molecular diversity of early-born subplate neurons. Cerebral Cortex. 2013;23:1473–1483. doi: 10.1093/cercor/bhs137. [DOI] [PubMed] [Google Scholar]
  24. Kast RJ, Wu HH, Levitt P. Developmental connectivity and molecular phenotypes of unique cortical projection neurons that express a Synapse-Associated receptor tyrosine kinase. Cerebral Cortex. 2019;29:189–201. doi: 10.1093/cercor/bhx318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kim J, Matney CJ, Blankenship A, Hestrin S, Brown SP. Layer 6 corticothalamic neurons activate a cortical output layer, layer 5a. Journal of Neuroscience. 2014;34:9656–9664. doi: 10.1523/JNEUROSCI.1325-14.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Konopka G, Bomar JM, Winden K, Coppola G, Jonsson ZO, Gao F, Peng S, Preuss TM, Wohlschlegel JA, Geschwind DH. Human-specific transcriptional regulation of CNS development genes by FOXP2. Nature. 2009;462:213–217. doi: 10.1038/nature08549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lai CS, Fisher SE, Hurst JA, Vargha-Khadem F, Monaco AP. A forkhead-domain gene is mutated in a severe speech and language disorder. Nature. 2001;413:519–523. doi: 10.1038/35097076. [DOI] [PubMed] [Google Scholar]
  28. MacDermot KD, Bonora E, Sykes N, Coupe AM, Lai CS, Vernes SC, Vargha-Khadem F, McKenzie F, Smith RL, Monaco AP, Fisher SE. Identification of FOXP2 truncation as a novel cause of developmental speech and language deficits. The American Journal of Human Genetics. 2005;76:1074–1080. doi: 10.1086/430841. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. McKenna WL, Betancourt J, Larkin KA, Abrams B, Guo C, Rubenstein JL, Chen B. Tbr1 and Fezf2 regulate alternate corticofugal neuronal identities during neocortical development. Journal of Neuroscience. 2011;31:549–564. doi: 10.1523/JNEUROSCI.4131-10.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Medvedeva VP, Rieger MA, Vieth B, Mombereau C, Ziegenhain C, Ghosh T, Cressant A, Enard W, Granon S, Dougherty JD, Groszer M. Altered social behavior in mice carrying a cortical Foxp2 deletion. Human Molecular Genetics. 2018;28:701–717. doi: 10.1093/hmg/ddy372. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Molyneaux BJ, Arlotta P, Hirata T, Hibi M, Macklis JD. Fezl is required for the birth and specification of corticospinal motor neurons. Neuron. 2005;47:817–831. doi: 10.1016/j.neuron.2005.08.030. [DOI] [PubMed] [Google Scholar]
  32. Molyneaux BJ, Goff LA, Brettler AC, Chen HH, Hrvatin S, Rinn JL, Arlotta P. DeCoN: genome-wide analysis of in vivo transcriptional dynamics during pyramidal neuron fate selection in neocortex. Neuron. 2015;85:275–288. doi: 10.1016/j.neuron.2014.12.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mukamel Z, Konopka G, Wexler E, Osborn GE, Dong H, Bergman MY, Levitt P, Geschwind DH. Regulation of MET by FOXP2, genes implicated in higher cognitive dysfunction and autism risk. Journal of Neuroscience. 2011;31:11437–11442. doi: 10.1523/JNEUROSCI.0181-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Pederick DT, Richards KL, Piltz SG, Kumar R, Mincheva-Tasheva S, Mandelstam SA, Dale RC, Scheffer IE, Gecz J, Petrou S, Hughes JN, Thomas PQ. Abnormal cell sorting underlies the unique X-Linked inheritance of PCDH19 epilepsy. Neuron. 2018;97:59–66. doi: 10.1016/j.neuron.2017.12.005. [DOI] [PubMed] [Google Scholar]
  35. Petrof I, Viaene AN, Sherman SM. Two populations of corticothalamic and interareal corticocortical cells in the subgranular layers of the mouse primary sensory cortices. The Journal of Comparative Neurology. 2012;520:1678–1686. doi: 10.1002/cne.23006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Rubenstein J, Rakic P. Neural Circuit Development and Function in the Brain: Comprehensive Developmental Neuroscience. Academic Press; 2013. [DOI] [Google Scholar]
  37. Spiteri E, Konopka G, Coppola G, Bomar J, Oldham M, Ou J, Vernes SC, Fisher SE, Ren B, Geschwind DH. Identification of the transcriptional targets of FOXP2, a gene linked to speech and language, in developing human brain. The American Journal of Human Genetics. 2007;81:1144–1157. doi: 10.1086/522237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sundberg SC, Lindström SH, Sanchez GM, Granseth B. Cre-expressing neurons in visual cortex of Ntsr1-Cre GN220 mice are corticothalamic and are depolarized by acetylcholine. Journal of Comparative Neurology. 2018;526:120–132. doi: 10.1002/cne.24323. [DOI] [PubMed] [Google Scholar]
  39. Takahashi K, Liu FC, Hirokawa K, Takahashi H. Expression of Foxp2, a gene involved in speech and language, in the developing and adult striatum. Journal of Neuroscience Research. 2003;73:61–72. doi: 10.1002/jnr.10638. [DOI] [PubMed] [Google Scholar]
  40. Tasic B, Menon V, Nguyen TN, Kim TK, Jarsky T, Yao Z, Levi B, Gray LT, Sorensen SA, Dolbeare T, Bertagnolli D, Goldy J, Shapovalova N, Parry S, Lee C, Smith K, Bernard A, Madisen L, Sunkin SM, Hawrylycz M, Koch C, Zeng H. Adult mouse cortical cell taxonomy revealed by single cell transcriptomics. Nature Neuroscience. 2016;19:335–346. doi: 10.1038/nn.4216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Thomson AM. Neocortical layer 6, a review. Frontiers in Neuroanatomy. 2010;4:13. doi: 10.3389/fnana.2010.00013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Torii M, Hackett TA, Rakic P, Levitt P, Polley DB. EphA signaling impacts development of topographic connectivity in auditory corticofugal systems. Cerebral Cortex. 2013;23:775–785. doi: 10.1093/cercor/bhs066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Tsui D, Vessey JP, Tomita H, Kaplan DR, Miller FD. FoxP2 regulates neurogenesis during embryonic cortical development. Journal of Neuroscience. 2013;33:244–258. doi: 10.1523/JNEUROSCI.1665-12.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Vargha-Khadem F, Watkins KE, Price CJ, Ashburner J, Alcock KJ, Connelly A, Frackowiak RS, Friston KJ, Pembrey ME, Mishkin M, Gadian DG, Passingham RE. Neural basis of an inherited speech and language disorder. PNAS. 1998;95:12695–12700. doi: 10.1073/pnas.95.21.12695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Vargha-Khadem F, Gadian DG, Copp A, Mishkin M. FOXP2 and the neuroanatomy of speech and language. Nature Reviews Neuroscience. 2005;6:131–138. doi: 10.1038/nrn1605. [DOI] [PubMed] [Google Scholar]
  46. Vernes SC, Oliver PL, Spiteri E, Lockstone HE, Puliyadi R, Taylor JM, Ho J, Mombereau C, Brewer A, Lowy E, Nicod J, Groszer M, Baban D, Sahgal N, Cazier JB, Ragoussis J, Davies KE, Geschwind DH, Fisher SE. Foxp2 regulates gene networks implicated in Neurite outgrowth in the developing brain. PLOS Genetics. 2011;7:e1002145. doi: 10.1371/journal.pgen.1002145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Watakabe A, Hirokawa J, Ichinohe N, Ohsawa S, Kaneko T, Rockland KS, Yamamori T. Area-specific substratification of deep layer neurons in the rat cortex. The Journal of Comparative Neurology. 2012;520:3553–3573. doi: 10.1002/cne.23160. [DOI] [PubMed] [Google Scholar]
  48. Woodworth MB, Greig LC, Liu KX, Ippolito GC, Tucker HO, Macklis JD. Ctip1 regulates the balance between specification of distinct projection neuron subtypes in deep cortical layers. Cell Reports. 2016;15:999–1012. doi: 10.1016/j.celrep.2016.03.064. [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision letter

Editor: Constance L Cepko1
Reviewed by: Myriam Heiman2

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "FOXP2 exhibits neuron class specific expression, but is not required for multiple aspects of cortical histogenesis" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Huda Zoghbi as the Senior Editor. The following individual involved in review of your submission has agreed to reveal her identity: Myriam Heiman (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

Interest in Foxp2 derives from human genetics, where it is linked to developmental language defects. Given the paucity of disease genes linked to language, the original finding of Foxp2 defects in humans was very exciting. Studies in the mouse suggested that it is important in development of the cerebral cortex, specifically layer 6 cortical neurons. Here, the authors use mouse genetics to delete FoxP2 from the developing mouse cortex. They did not find defects in cortical neuron genesis, migration, or corticothalamic connectivity. Other defects may be present, but were not characterized. These data are important to the field, since language and cognitive development of humans is such an interesting, but difficult, topic to address outside of animal models. As the current manuscript shows convincing evidence (but needing more documentation of Foxp2 loss) that the previous studies from mouse are incorrect, these data are important to set things right.

Essential revisions:

1) Loss of Foxp2 needs to be definitively established along with the timing of the loss. RNA ISH, using RNAscope, or the new SABER FISH method, can give quantitative data on RNA level or isoform changes. If IHC of IF are used, controls must be convincing.

2) The numbers of animals used for various experiments needs to be provided, and that those numbers should be >2. For negative findings, it is important that we are not looking at some biological oddities.

3) Assess the expression of Foxp1 following the KO of Foxp2. Again, can be done by ISH. If not ISH, then again any protein detection must be done in a convincing and quantitative manner.

Reviewer #1:

This is an important study utilizing mouse genetics to assess the requirement of Foxp2 in cortical development. The findings are predominantly negative building upon previous loss and gain of function studies, which suggested an important role for Foxp2 in layer 5/6 cortical development. Surprisingly, the findings from the current study instead suggest that Foxp2 does not appear to be required for deep layer neuron generation, specification, or axonal pathfinding.

Importantly, I think these negative results would be a valuable contribution to the field and would be received with much interest. The study was well-executed and if a few more experiments were performed, I think it could increase confidence in the results. After completion of additional experiments, it should be accepted for publication.

Overall critiques/questions:

To look at the early specification of CT neurons, the authors utilize Emx1-cre tdTomato reporter. In their conditional FOXP2 deletion, is the total number or overall size of the cortex affected? In the Emx1-cre Foxp2fl/fl mice, for example, it seems as though the overall brain size is affected, at least transiently during development? Figure 2—figure supplement 1 is a good example of that. I can see that at adult stages, the total number of Fog2 positive cells is comparable, but is there a chance that during development there are differences in the timing of their generation that might affect brain function?

A related concern about experimental design, the authors use foxp2fl/fl as a control to foxp2fl/fl emx1-cre, to report that the number of Fog2 cells in layer 6 are reduced, although not significantly. One issue is that the control is not really an appropriate control.

To investigate the early phenotypes of FOXP2, the authors use Emx1-cre and find early evidence of recombination, but only report loss of protein at E16.5, which seems pretty late in layer 6 differentiation. It would be important to investigate further loss of Foxp2 protein becomes apparent.

The authors emphasize Foxp2 mutations in speech deficits and forebrain development, which may have an impact in human cortical neurogenesis. However, the studies were performed in mouse. The text should clarify that this is mouse forebrain development in the Abstract/Discussion and Foxp2 requirement may be different in different species.

Foxp2 may be very important for circuit formation and function which may explain the disease phenotypes in humans. What are the effects of loss of Foxp2 on connectivity? Synapse formation? Functional activity?

Could some differences be identified if you looked earlier in development during embryogenesis? Potentially Foxp2 loss may affect timing of generation, differentiation, maturation etc. which would provide insight into its developmental role.

Although the authors speculate reasons for previously reported findings on Foxp2, it is difficult to reconcile with the current study without a more thorough characterization/description of Foxp2 activity. Are there other genes known to compensate for Foxp2? It would be helpful to have some data reconciling the results of this study with previous findings.

Reviewer #2:

Recent studies have suggested that FOXP2, a gene that in human is linked to developmental speech impairments, may have a role in cortical neurogenesis and migration. Although a study in the developing mouse brain by the Walsh group demonstrated that cortical Foxp2 expression is absent from the ventricular zone and restricted to stages of late neuronal migration or during neuronal differentiation, and MRI studies of brain structure in patients with FOXP2 mutations showed no gross defects of the cerebral cortex, mouse overexpression and shRNA knockdown studies have suggested an important role for Foxp2 in mouse cortical neurogenesis. To more fully investigate the proposed role of Foxp2 in cortical histogenesis, Kast et al. have used molecular, neuroanatomical, genetic, and circuit level approaches. The data together convincingly show, with the use of two independent conditional genetic approaches, that murine Foxp2 is not required for cortical neurogenesis or migration. The results are of great importance to the field to clarify the function of FOXP2.

One major point is that loss of Foxp2 with the conditional genetic approaches is shown only with the use of a goat Foxp2 polyclonal antibody. It would be important to present RNAScope (or similar in situ with cell type-specific resolution) and global qPCR assessment to show Foxp2 mRNA depletion based on both conditional approaches.

Reviewer #3:

This interesting paper essentially reports negative results: despite a good deal of previous data that supported a role for FoxP2 as being an important transcriptional regulator of layer 6 cortical neuron development, two different conditional knockout strategies reported here both lack any detectable phenotypes. Looking at neuronal projections (corticothalamic) and protein expression in layer 5 and 6 neurons (by immunofluorescence) in both models, no differences with controls were observed. To control for timing of Cre-mediated recombination, the study was carried out using Cre lines with early (Emx1-Cre) and late (Ntsr1-Cre) expression, and presumably recombination.

Overall, this is a thorough and well-reported study. One potential area of concern is that the paper is completely dependent on using immunofluorescence in sections for FoxP2, using a commercial polyclonal antibody, to confirm the timing and degree of loss of FoxP2 function. In terms of major comments, this is the one area that is essential to address with a technically independent method: confirming that either the mRNA or protein are completely absent from the knockout cortex, and when this is the case.

Related to this, the authors speculate a good deal as to why acute knockdown with shRNA would lead to detectable phenotypes, whereas the genetic knockout does not. Aside from the additional experiment suggested above to confirm that this is a true null, the lack of discussion of possible redundancy with or compensation by FoxP1 is a puzzling omission. FoxP1 expression overlaps with FoxP2 in many regions/times – some comment by the authors in the discussion on whether this may or may not compensate for FoxP2 would be helpful.

The conclusion of the paper lacks balance: the authors do not detect gross phenotypes in cell specification or projections, but there are many other features that are not studied here, including neuronal maturation, dendrite development, synaptogenesis etc. Given that FoxP2 is largely expressed post-mitotically in migratory neurons, the title and main conclusions do not accurately reflect the field.

eLife. 2019 May 17;8:e42012. doi: 10.7554/eLife.42012.020

Author response

Essential revisions:

1) Loss of Foxp2 needs to be definitively established along with the timing of the loss. RNA ISH, using RNAscope, or the new SABER FISH method, can give quantitative data on RNA level or isoform changes. If IHC of IF are used, controls must be convincing.

We have generated new in situ hybridization (ISH) data using a new method (BaseScope) that demonstrates the loss of the floxed region of the Foxp2 transcript (exons 12-14) that encodes the DNA binding domain of Foxp2. The loss of the gene region encoding the DNA-binding domain demonstrate that the protein product cannot be functional. The data were generated using Emx1-cre; Foxp2Fx/Fxand control Foxp2Fx/Fxlittermates on embryonic day (E)14.5, a timepoint when FOXP2 protein cannot be detected by immunohistochemistry in WT animals. Importantly, by E16.5, when FOXP2 protein first can be detected by immunohistochemistry in Foxp2Fx/Fxcontrols, there is no FOXP2 protein in the cortex of conditional knockouts.

2) The numbers of animals used for various experiments needs to be provided, and that those numbers should be >2. For negative findings, it is important that we are not looking at some biological oddities.

We have increased sample size (N) for all experiments to achieve at least N=3, and typically greater for most experiments reported.

3) Assess the expression of Foxp1 following the KO of Foxp2. Again, can be done by ISH. If not ISH, then again any protein detection must be done in a convincing and quantitative manner.

We respectfully disagree with this request. Hisaoka et al., 2010, reported using double labeling methods that Foxp1 (concentrated in layers 2-5) and Foxp2 (concentrated in layer 6, with some expression in layer 5) are expressed almost exclusively in different neuronal populations in the cerebral cortex. There is minor overlap in deep layer 5. There is, however, substantial overlap of Foxp1 and Foxp2 expression in the developing striatum, and yet Vernes et al., 2011 nor French et al., 2007 , did not report Foxp1 expression changes in striatum or whole brain at E16.5, well after Foxp2 positive neurons are generated. Given the lack of significant overlap in the cerebral cortex, and the hundreds of other genes identified by Vernes et al. that exhibit changes in expression in mice carrying a Foxp2 mutation, we believe the rationale is weak for doing further studies of Foxp1 expression. We do believe it would be of interest to do an in depth profiling of gene expression using single cell RNA sequencing (scRNAseq) of Foxp2-expressing layer 6 and 5 neurons, but that profiling would require follow-up functional studies to determine the relevant developmental adaptations, which we argue is well beyond the scope of the current report. Finally, as we note below in response to reviewer #1, changes in gene expression of other Foxp family members, or any other genes, due to Foxp2 loss, would not provide conclusive mechanistic evidence for the lack of a phenotype, nor would it change the main conclusion of this report that Foxp2 is not required for cortical development phenotypes that have been assigned previously to Foxp2.

Reviewer #1:

[…] Overall critiques/questions:

To look at the early specification of CT neurons, the authors utilize Emx1-cre tdTomato reporter. In their conditional FOXP2 deletion, is the total number or overall size of the cortex affected? In the Emx1-cre Foxp2fl/fl mice, for example?

The numbers of CT neurons at P0 were quantified in cKO (Emx1-cre; Foxp2Fx) and animals of 3 other genotypes (Emx1-cre; Foxp2Fx/Fx (cHET), Emx1-cre on WT C57Bl6 background, and Foxp2Fx/Fx), which did not contain tdTomato reporter. The results (Figure 3C, D) demonstrated that the number of CT neurons were not statistically significantly different across genotypes. Additionally, the radial thickness of the cortex was measured at P14 and found to be indistinguishable across genotypes (Figure 3H, I).

It seems as though the overall brain size is affected, at least transiently during development. Figure 2—figure supplement 1 is a good example of that.

We thank the reviewer for the observation. We have looked more carefully at this and included comparisons of very closely matched (in terms of anterior-posterior level) coronal sections from cKO and control littermates. The new data show that the size of the brains are not different across genotypes at E14, E16, P0, or P14. We have replaced the images in Figure 2—figure supplement 1 with these new images at E16. These sections also were stained with an optimized immunohistochemical procedure to convincingly show the presence and absence of cortical FOXP2+ cells in the control and cKO brains, respectively.

I can see that at adult stages, the total number of Fog2 positive cells is comparable, but is there a chance that during development there are differences in the timing of their generation that might affect brain function?

We have quantified the numbers of layer 6 neuron subtypes at P0 and P14 and found no differences across genotypes at either age. So, any developmental differences would be limited to the late prenatal period. We have not performed an exhaustive analysis of cell types across genotypes during prenatal development, so we cannot rule out such a change. The hypothesized rate changes of neuronal subtype differentiation would need to be limited to a 72-hour developmental window, between E16.5 (FOXP2 onset) and P0.5 (when we observe no differences in cell numbers). We note that addressing this hypothesis would require analysis at many prenatal time points (to account for variation in developmental timing due to the imprecision in determining fertilization), and thus many more months of experimentation. While possible, we believe rate changes in differentiation are unlikely, and moreover would not explain the end result of normal subpopulation representations and other normal phenotypes in the Foxp2 cKO.

A related concern about experimental design, the authors use foxp2fl/fl as a control to foxp2fl/fl emx1-cre, to report that the number of Fog2 cells in layer 6 are reduced, although not significantly. One issue is that the control is not really an appropriate control.

We have updated the text to properly emphasize that there is no statistical difference in the numbers of Fog2 cells across genotypes. We share the reviewer’s emphasis on properly controlling for genotypes in these experiments, which is why 4 genotypes were included in experiments that suggested a trending effect when comparing Emx1-cre; Foxp2Fx/Fxand Foxp2Fx/Fxlittermates (e.g. Fog2 counts at P0). We note that most developmental studies of similar type to the report here do not include 4 relevant genotypes, as we have done. In experiments in which initial comparisons of Emx1-cre; Foxp2Fx/Fxand Foxp2Fx/Fx groups did not show any differences (i.e. cell type counts and cortical thickness measurements at P14), we chose not to include a Cre-only control group as there was no effect of Foxp2 removal. We added the Cre-only control group in cases where initial data indicated that there might be a trend. We have repeated the experiment by increasing the sample sizes (N) of each group in the P0 cell type counting experiment:

1 – Foxp2Fx/Fx from n=5 to n=7

2 – cHet from n=5 to n=6

3 – cKO from n=4 to n=7

4 – Emx1-cre only from n=4 to n=6

The results of the experiment, and the application of appropriate non-parametric statistics, with correction for multiple comparisons, continue to support our conclusion that the numbers of Fog2 cells are not different across genotypes, and thus Foxp2 does not influence the specification of this cortical cell type.

To investigate the early phenotypes of FOXP2, the authors use Emx1-cre and find early evidence of recombination, but only report loss of protein at E16.5, which seems pretty late in layer 6 differentiation. It would be important to investigate further loss of Foxp2 protein becomes apparent.

We have added ISH data based on the BaseScope(Advanced Cell Diagnostics) technique, which demonstrates that in Emx1-cre; Foxp2Fx/Fxmice the floxed Foxp2 allele undergoes recombination by E14.5, thus leading to the complete absence of the mRNA segment encoding the Foxp2 DNA-binding domain. This precedes the expression of FOXP2 protein, which we could not detect through IHC until E16.5. We note that our assessment of initial expression of FOXP2 protein is consistent with that reported by Ferland et al., 2003, and in the in situ hybridization data publicly available in the Allen Brain Institute database for developing mice.

The authors emphasize Foxp2 mutations in speech deficits and forebrain development, which may have an impact in human cortical neurogenesis. However, the studies were performed in mouse. The text should clarify that this is mouse forebrain development in the Abstract/Discussion and Foxp2 requirement may be different in different species.

We have updated the text in the Abstract and the Discussion to emphasize that our work has been done in mice and leave open the possibility that there may be specific functions of Foxp2 related to cortical development in other species.

Foxp2 may be very important for circuit formation and function which may explain the disease phenotypes in humans. What are the effects of loss of Foxp2 on connectivity? Synapse formation? Functional activity?

We agree with this statement. For example, it is clear that disruption of Foxp2 expression in the striatum has developmental and functional outcomes. However, we wish to re-emphasize that the results of experiments probing later aspects of brain development and function would be difficult to interpret in the absence of the foundational data provided in the present study. Moreover, in the context of the dramatic effects attributed to Foxp2 in previous publications, such experiments would seem somewhat distant from the primary action of Foxp2. Given the lack of overt histological changes, we agree that as a follow-up of the current detailed analysis, it would be of interest to investigate the role of Foxp2 in later aspects of brain development, plasticity and function. We believe the current study will assist in focusing those in the field on other epochs of development that speak to the issue raised by this reviewer.

Could some differences be identified if you looked earlier in development during embryogenesis? Potentially Foxp2 loss may affect timing of generation, differentiation, maturation etc. which would provide insight into its developmental role.

This issue is related to our third response to reviewer 1 above and we have addressed the challenges of the hypothesis of evaluating timing of maturation in a short window from the onset of Foxp2 expression to our initial quantitative analysis (72 hr). As noted for #6, we believe follow-up studies are merited, but are beyond the scope of the present study. We have attempted to be very careful about the conclusions drawn with regard to speaking only to the time periods and measures that we have conducted.

Although the authors speculate reasons for previously reported findings on Foxp2, it is difficult to reconcile with the current study without a more thorough characterization/description of Foxp2 activity. Are there other genes known to compensate for Foxp2? It would be helpful to have some data reconciling the results of this study with previous findings.

This is an interesting and valid point. Indeed, pioneering work by Edward Morrisey and colleagues demonstrated the functional cooperativity of Foxp family members including Foxp1, 2, and 4 in the development of anterior gut derived structures including the esophagus and airway (Li et al., 2004; Shu et al., 2007). These studies demonstrated that compound Foxp2-/- and Foxp1-/+ mutants display more dramatic lung development defects than Foxp2-/- single mutants, but in the case of the lung, even Foxp2-/- single mutants showed defects in anterior gut developmental phenotypes. The present neurodevelopmental study aimed to build on previous publications that concluded that Foxp2 plays a critical role in corticogenesis. However, we found that previous conclusions drawn about the role of Foxp2 in cortical development (suggested to occur independent of manipulating other Foxp family members using shRNA technology) are inconsistent with the genetic approaches used here, with 2 different Cre driver lines, performed in the present study. Genetic manipulations are typically taken as the gold standard for determining gene function, although other approaches can be valuable. Thus, while the possibility still exists that compensation for the loss of Foxp2 function [by any mechanism, including Foxp1 upregulation, or upregulation of the hundreds of transcripts reportedly regulated by Foxp2 (Vernes et al., 2011)] might mask a valid, but nonessential, role for Foxp2 in the development of the cortex, the main conclusion drawn here – that Foxp2 is not required for basic cortical histogenic events that include cell type production in which Foxp2 is expressed, and fundamental cortical axon extension and targeting to the dorsal thalamus and pyramidal tract – is supported firmly by the present data. Developmental manipulations can also be assessed for potential compensatory mechanisms, because phenotypes could be a sum of direct loss of gene activity coupled with adaptations. These types of studies are of great interest, and will require significant new detailed molecular, developmental and functional experiments that will likely take years, given the number of genes that are regulated putatively by Foxp family members. We believe this is well beyond the scope of the present study, and will not impact the conclusions drawn that Foxp2 is not required for specific aspects of cortical development in mice.

Reviewer #2:

[…] One major point is that loss of Foxp2 with the conditional genetic approaches is shown only with the use of a goat Foxp2 polyclonal antibody. It would be important to present RNAScope (or similar in situ with cell type-specific resolution) and global qPCR assessment to show Foxp2 mRNA depletion based on both conditional approaches.

We have added new ISH data based on a new method (BaseScope) that demonstrates complete absence of Foxp2 exons 12-14 (floxed region, encodes DNA-binding domain) in the dorsal pallium of Emx1-cre; Foxp2Fx/Fxmice from at least E14.5 onward due to the recombination of the floxed Foxp2 allele in the cortex. We note that we were unable to detect FOXP2 protein until E16.5, and thus genetic recombination occurs prior to FOXP2 protein production.

Reviewer #3:

[…] One potential area of concern is that the paper is completely dependent on using immunofluorescence in sections for FoxP2, using a commercial polyclonal antibody, to confirm the timing and degree of loss of FoxP2 function. In terms of major comments, this is the one area that is essential to address with a technically independent method: confirming that either the mRNA or protein are completely absent from the knockout cortex, and when this is the case.

As described above, we’ve added new ISH data based on a new method (BaseScope) that demonstrates complete absence of Foxp2 exons 12-14 (floxed region, encodes DNA-binding domain) in the dorsal pallium of Emx1-cre; Foxp2Fx/Fxmice. Please see response to reviewer 2.

Related to this, the authors speculate a good deal as to why acute knockdown with shRNA would lead to detectable phenotypes, whereas the genetic knockout does not. Aside from the additional experiment suggested above to confirm that this is a true null, the lack of discussion of possible redundancy with or compensation by FoxP1 is a puzzling omission. FoxP1 expression overlaps with FoxP2 in many regions/times – some comment by the authors in the Discussion on whether this may or may not compensate for FoxP2 would be helpful.

As noted above (Essential revisions, comment #3), Foxp2 and Foxp1 are expressed mostly in non-overlapping neuronal populations in the developing and adult cerebral cortex. We have added discussion to address this. Also, as noted, given the plethora of genes with Foxp family member DNA binding sites and that are hypothetically regulated by Foxp2, compensation is certainly possible, and we note this.

The conclusion of the paper lacks balance: the authors do not detect gross phenotypes in cell specification or projections, but there are many other features that are not studied here, including neuronal maturation, dendrite development, synaptogenesis etc. Given that FoxP2 is largely expressed post-mitotically in migratory neurons, the title and main conclusions do not accurately reflect the field.

We have edited the manuscript to ensure accuracy in the conclusions drawn from these studies. While limited to the late prenatal and early postnatal periods, we do not believe that the phenotypes we studied are gross in nature. Cell type specification is an essential histogenic event for which connectivity, circuit formation and function all depend upon. We agree that doing studies later in development may be useful for follow-up, but would not change the careful conclusions that we have drawn.

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Data Citations

    1. Kast R, Lanjewar A, Smith C, Levitt P. 2018. Data from: FOXP2 exhibits neuron class specific expression, but is not required for multiple aspects of cortical histogenesis. Dryad Digital Repository. [DOI] [PMC free article] [PubMed]

    Supplementary Materials

    Figure 1—source data 1. Foxp2 expression among retrogradely labeled and molecularly defined developing layer 6 neuron classes.
    elife-42012-fig1-data1.xlsx (17.4KB, xlsx)
    DOI: 10.7554/eLife.42012.005
    Figure 1—figure supplement 1—source data 1. Foxp2 expression among retrogradely and molecularly defined developing pyramidal tract neurons.
    elife-42012-fig1-figsupp1-data1.xlsx (11.2KB, xlsx)
    DOI: 10.7554/eLife.42012.004
    Figure 2—source data 1. Expression of Layer 6 cell-type markers in Ntsr1-cre; Foxp2Fx mice and controls.
    elife-42012-fig2-data1.xlsx (10.1KB, xlsx)
    DOI: 10.7554/eLife.42012.008
    Figure 3—source data 1. Quantification of cortical cell type numbers in Emx1-cre; Foxp2Fx mice and controls.
    elife-42012-fig3-data1.xlsx (16.3KB, xlsx)
    DOI: 10.7554/eLife.42012.011
    Source code 1. Maximum Projection and Cropping Macro.
    elife-42012-code1.ijm (761B, ijm)
    DOI: 10.7554/eLife.42012.013
    Source code 2. Cortical Cell Type Autocounting Macro for nuclear markers.
    elife-42012-code2.ijm (1.4KB, ijm)
    DOI: 10.7554/eLife.42012.014
    Transparent reporting form
    elife-42012-transrepform.docx (256KB, docx)
    DOI: 10.7554/eLife.42012.015

    Data Availability Statement

    All numbers relating to quantitative experiments have been uploaded to Dryad (https://dx.doi.org/10.5061/dryad.6hd7bf7).

    The following dataset was generated:

    Kast R, Lanjewar A, Smith C, Levitt P. 2018. Data from: FOXP2 exhibits neuron class specific expression, but is not required for multiple aspects of cortical histogenesis. Dryad Digital Repository.

    Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd

    RESOURCES