The En1^cre/+ and Chd7^flox mouse lines have been described (Kimmel et al., 2000; Jones et al., 2015). Both these alleles were backcrossed onto the C57BL/6J background for at least three generations and genotyped according the original publications. Conditional mes/r1-specific Chd7 conditional knockout En1^cre/+;Chd7^flox/flox (En1^cre/+;Chd7^f/f) mice were produced by En1^cre/+;Chd7^flox/+ × Chd7^flox/flox crosses. Mice were bred and maintained in the Biological Services Unit at Guy’s Campus or the Institute of Psychiatry, Psychology and Neuroscience, King’s College London. All procedures involving animals were approved by the local ethical review panel of King’s College London, and the U.K. Home Office Animals Scientific Procedures Act 1986. The work was carried out under licenses (PPL70/6694 and PPL70/7184) and all efforts were made to minimize animal suffering and to reduce the number of animals used.

Brains or embryos were dissected in phosphate buffered saline (PBS), fixed overnight in 4% paraformaldehyde (PFA) at 4°C, dehydrated and embedded in paraffin wax. Serial, sagittal sections were cut at 10 μm and left to dry overnight at 42°C. Sections were stained with Cresyl Violet or processed for immunohistochemistry as described (Whittaker et al., 2017). The following primary antibodies were used: anti-tyrosine hydroxylase (Abcam, ab112; 1:200) and anti-Neurogranin (Millipore, AB5620; 1:500). Primary antibodies were detected using Alexa fluor-conjugated secondary antibodies (Invitrogen; 1:200) or biotinylated secondary antibodies (Dako, E0466; 1:200) with the Vectastain ABC Kit (Vector Laboratories) and visualized using 0.03% diaminobenzidine (DAB; Sigma).

In situ hybridization was performed using standard methods (Basson et al., 2008; Yaguchi et al., 2009). The Fgf8 and Etv5 in situ hybridization probes were reported by Yaguchi et al. (2009) and the Chd7 exon 3 probe by Whittaker et al. (2017).

A total of 45 mice were examined in this study. The mouse numbers used were: 13 Chd7^flox/flox, 12 En1^cre/+, 10 En1^cre/+;Chd7^flox/+ and 10 En1^cre/+;Chd7^flox/flox. All mice were adults (P80–P100). Mice were terminally anesthetized and intracardially perfused with 30 mL of 0.1 M PBS containing 10 U/mL heparin and 2 mM ProHance (Bracco Diagnostics Inc.), a Gadolinium contrast agent followed by 30 mL of 4% PFA containing 2 mM ProHance (Spring et al., 2007; Cahill et al., 2012). Perfusions were performed at a rate of approximately 60 mL/h. After perfusion, mice were decapitated. The brain and remaining skull structures were incubated in 4% PFA + 2 mM ProHance overnight at 4°C then transferred to 0.1 M PBS containing 2 mM ProHance and 0.02% sodium azide for at least 1 month days prior to MRI scanning (De Guzman et al., 2016). A multi-channel 7.0 Tesla MRI scanner (Agilent Inc., Palo Alto, CA, USA) was used to image the brains within skulls. Sixteen custom-built solenoid coils were used to image the brains in parallel (Bock et al., 2005; Lerch et al., 2011). Parameters used in the anatomical MRI scans: T2- weighted 3D fast spin-echo sequence, with a cylindrical acquisition of k-space, and with a TR of 350 ms, and TEs of 12 ms per echo for six echoes, two averages, field-of-view of 20 × 20 × 25 mm³ and matrix size = 504 × 504 × 630 giving an image with 0.040 mm isotropic voxels (Nieman et al., 2005). The current scan time required for this sequence is ~14 h. To visualize and compare any differences in the mouse brains, the images from all brains were linearly (6 parameter followed by a 12 parameter) and non-linearly registered together, and then averaged together to create a population atlas representing the anatomy of the study sample. All registrations were performed with a combination of mni_autoreg tools (Collins et al., 1994) and advanced normalization tools (ANTs; Avants et al., 2008, 2011). The result of this registration is to have all scans deformed into alignment with each other in an unbiased fashion. This allows for the analysis of the deformations, and model how they relate to genotype (Nieman et al., 2006; Lerch et al., 2008). The Jacobian determinants of the deformation fields therefore can be used as measures of volume at each voxel. Significant volume changes were then calculated in two ways: (1) on a region basis; and (2) voxel-wise. Regional volumes are calculated by warping a pre-existing classified MRI atlas onto the population atlas. This atlas is a combination of three different atlases, comprising 159 different structures including, but not limited to, the cortical lobes, large white matter structures (i.e., corpus callosum), ventricles, cerebellum, brain stem and olfactory bulbs (Dorr et al., 2008; Ullmann et al., 2013; Steadman et al., 2014). Using the inverse transformations from the registration process, allows the labeling of the original pre-registration images in order to calculate the volumes of each individual brain region. Significant differences were determined between groups for both the 166 different regions and voxel-wise throughout the brain. Multiple comparisons in this study were controlled for using the False Discovery Rate (Genovese et al., 2002).

The En1^cre/+ line has been used previously to efficiently recombine conditional alleles resulting in the loss of gene expression in most cells in the mes/r1 region by the 10 somite stage (~E8.75) (Chi et al., 2003; Basson et al., 2008). We therefore used this line to simultaneously inactivate both Chd7 conditional alleles in the mes/r1 region in En1^cre/+;Chd7^f/f embryos. In situ hybridization to detect transcripts containing Chd7 exon 3 confirmed the ubiquitous embryonic expression of Chd7 at E9.5 (Figure 1A), including the mes/r1 region (Figure 1A′; Randall et al., 2009). Efficient mes/r1-specific deletion of exon 3 was clearly evident in En1^cre/+;Chd7^f/f embryos (Figures 1C,C′). Interestingly, Chd7 expression was also clearly reduced in conditional heterozygous En1^cre/+;Chd7^f/+ embryos (Figures 1B,B′). En1^cre/+, Chd7^f/f and En1^cre/+;Chd7^f/+ animals were born at Mendelian ratios and showed no signs of abnormalities. Although slightly fewer than expected En1^cre/+;Chd7^f/f animals were obtained, this effect was not statistically significant (Table 1).

To determine the consequences of mes/r1-specific Chd7 deletion on brain development, brains were collected from adult animals and examined by structural MRI. We first confirmed that En1^cre/+ animals, where one copy of the En1 gene had been inactivated by Cre insertion and therefore heterozygous for En1 (Kimmel et al., 2000), showed no significant structural brain anomalies, compared to Chd7^f/f controls (Figure 2A). A comparison of absolute volumes of 166 different brain regions in homozygous conditional En1^cre/+, Chd7^f/f mutants with Chd7^f/f controls, we identified striking hypoplasia of the cerebellum and midbrain (Figures 2A–C). Intriguingly, when looking at absolute volumes, the automated MRI analysis (Supplementary Table S1) also identified other brain regions outside the mid-hindbrain region that were hypoplastic in these mutants (Figure 2A). These include the thalamus, hypothalamus and corpus callosum (Figure 2A).

Cerebellar and midbrain sizes, relative to total brain size were significantly reduced by 30.5% and 17.9%, respectively, whilst relative cortical volumes were increased by 6.7%, likely due to the sizeable decrease in cerebellar and midbrain volume (Figures 2E–G). We also found significantly reduced cerebellar (9%) and midbrain (8%) volumes in heterozygous En1^cre/+;Chd7^f/+ animals (Figure 2), suggesting that the reduction in Chd7 expression in the mes/r1 in these embryos (Figures 1B,B′), had significant effects on brain development and that heterozygous Chd7 expression during cerebellar development can be sufficient to cause mild cerebellar hypoplasia.

To further characterize the nature of the cerebellar hypoplasia in the Chd7 conditional mutants, cerebellar structure was visualized in both horizontal and sagittal slices on the MRI images. Compared to Chd7^f/f and En1^cre/+;Chd7^f/+ animals, prominent cerebellar hypoplasia was evident in horizontal images of En1^cre/+;Chd7^f/f brains (Figures 3A–C). and cerebellar vermis tissue was nearly absent in sagittal slices (compare Figure 3C′ with Figures 3A′,B′). The cerebellar hemispheres in these homozygous mutants were also hypoplastic and cerebellar foliation was highly irregular (Figure 3C).

Freshly collected, wholemount images taken of P21 cerebella confirmed the striking hypoplasia of the cerebellar vermis, as well as the abnormal cerebellar foliation in the hypoplastic hemispheres in En1^cre/+;Chd7^f/f animals (Figures 3D–F). Histological sections taken at different medio-lateral positions along the cerebellum showed the near-absence of cerebellar tissue at the midline in En1^cre/+;Chd7^f/f animals (Figure 3F′), and mild vermis hypoplasia in En1^cre/+;Chd7^f/+ animals (Figure 3E′) compared to controls (Figure 3D′). Sections through the paravermis revealed striking hypoplasia in the homozygous mutants compared to the other genotypes (Figures 3D″–F″). Hypoplasia of the cerebellar hemispheres and abnormal foliation was clearly evident in En1^cre/+;Chd7^f/f mutants (compare Figure 3F″′ with Figures 3D″′,E″′).

To determine whether midbrain hypoplasia in the conditional mutants (Figures 2C,F) is associated with the loss of specific midbrain structures, we examined sections from newborn animals when individual midbrain structures are easily identified. This analysis confirmed the striking cerebellar vermis hypoplasia in the homozygous, conditional mutants and revealed an abnormally-shaped midbrain (compare Figure 4A with Figure 4B). Despite these structural changes, both anterior (superior colliculus, SC) and posterior (inferior colliculus, IC) midbrain structures were present in the conditional mutants, in both medial (Figures 4A,B) and lateral (Figures 4C,D) sections. The identity of the IC was confirmed by Neurogranin immunostaining, which again confirmed the abnormally-shaped IC (Figure 4F), compared to controls (Figure 4E). Tyrosine hydroxylase immunostaining was used to visualize ventral brain structures in the diencephalon, mid- and hindbrain. In medial sections, the ventral tegmental area (VTA) was clearly present in the conditional mutants (Figures 4G,H). Immunostaining of more lateral sections (Figures 4I,J) identified the substantia nigra (SN) and the locus coeruleus (LC). Although both these structures could be identified in conditional mutants, the TH staining intensity was reduced in both (n = 3 mutants, compared to littermate controls), suggesting that the development of cells in these more lateral midbrain areas is affected by Chd7 deletion from the mes/r1 region.

Given the phenotypic similarities between En1^cre/+;Chd7^f/f mutants and mutants with reduced FGF signaling (Chi et al., 2003; Basson et al., 2008), we predicted that Fgf8 expression and signaling would be reduced in the mid-hindbrain region. Indeed, Fgf8 expression was slightly reduced in heterozygous En1^cre/+;Chd7^f/+ mutants at E9.5 compared to Chd7^f/f controls (Figures 5A,B), and reduced even further in homozygous En1^cre/+;Chd7^f/f embryos (Figure 5C). The Fgf8 expression pattern observed from a dorsal view, showed evidence for a slightly expanded roof plate at the midline of the IsO in both the heterozygous and homozygous mutants (Figures 5A′–C′), characteristic of embryos with reduced FGF signaling at the IsO (Basson et al., 2008).

An examination of Fgf8 expression in embryos a day later (E10.5), still showed reduced Fgf8 expression in the IsO of homozygous mutants (compare Figure 5F with Figure 5D), whilst heterozygous mutants appeared to have normal levels of Fgf8 expression (compare Figure 5E with Figure 5D), although subtle changes in Fgf8 expression cannot be ruled out based on in situ hybridization experiments. Furthermore, viewing the dorsal aspect of r1 in these embryos, revealed a morphology typical of embryos with reduced FGF expression in homozygous mutants (Figure 5F′), compared to heterozygous and control embryos (Figures 5D′,E′). These embryos were characterized by an expanded roof plate in anterior r1, apparently at the expense of the cerebellar vermis progenitor domain, as previously reported for embryos with reduced FGF signaling in mes/r1 (Basson et al., 2008). In addition to having normal levels of Fgf8 expression, Chd7 heterozygous embryos also had a normal morphology, consistent with the very mild vermis hypoplasia displayed by these mutants (Figures 3E,E′).