ASP conceived the screen. ASP, KZ, SRM, and YS designed and performed the experiments. KZ, SRM, YS, JLRR, and ASP analyzed the data. ME contributed reagents/materials/analysis tools. ASP, KZ, and SRM wrote the paper.
The authors have declared that no conflicts of interest exist.
Although the mechanisms that regulate development of the cerebral cortex have begun to emerge, in large part through the analysis of mutant mice (
A novel screen uncovers thirteen mutants with defects ranging from the establishment of epithelial polarity in the cortex to the invasion of axons from the thalamus.
The cerebral cortex is the seat of consciousness and the means by which we carry out abstract reasoning. Understanding how the cortex is assembled during embryonic development will give deeper insights into how this marvelous machine functions and provide the basis for therapy and repair. Although a diversity of approaches will be needed to answer all of our questions, an important starting point in studying events in development is often the careful analysis of mutant phenotypes. Much of what we know about cortical development has emerged through the study of mutations in mice and humans. For example, spontaneous mutations in mice such as
Forward genetic approaches in the mouse, although technically feasible for many years, have become increasingly attractive with the availability of a dense genetic map and a nearly complete genomic sequence. These tools allow the process of gene identification, which was once very cumbersome, to be relatively straightforward. With the initial resurgence of interest in genetic screens, large-scale screens aimed at identifying mutations in broadly defined phenotypic categories were established at mutagenesis centers in Germany, the United Kingdom, the United States, and other countries (
Here we describe a focused genetic screen in the mouse that takes advantage of a transgenic reporter line that labels the ganglionic eminences of the ventral forebrain and migrating cortical interneurons with β-galactosidase. The use of this reporter gene allows the isolation of mutations that alter growth and morphogenesis of the cortex very efficiently and, more specifically, allows the identification of mutations disrupting interneuron migration from the ganglionic eminences into the cortex. In this screen, we isolated four mutations that affect tangential migration. We also isolated nine novel mutations affecting other aspects of cortical development; these mutations include three that represent novel alleles of genes that have been shown to have a role in cortical development. These results illustrate the power of a forward genetic approach, borrowed from other model organisms, that can be applied to various questions of mammalian biology.
The expression of Dlx family homeodomain transcription factors is necessary for differentiation of neurons in the striatum and for migration of many, if not all, of the interneuron precursors that arise in the ganglionic eminences of the embryonic telencephalon (
(A) Migrating interneuron precursors appear to migrate tangentially (yellow arrows) through both the MZ and the IZ/SVZ. Precursors also migrate radially (pink arrows) to reach the developing cortical plate (CP).
(B) Migrating interneuron precursors expressing β-galactosidase can be seen in a whole-mount preparation of an E14.5 Dlx-LacZ mouse as diffuse cortical staining.
(C) In coronal section, the densely labeled SVZ of the LGE can be seen, as can the streams of migrating interneuron precursors in the cortex.
(D) To identify recessive mutations, male C57BL6/J mice were treated with ENU and then crossed to FVBN/J females that were homozygous for the
During the course of screening we identified eight mutants with defective growth or patterning of the cerebral cortex, four mutants with defects in the migration of interneuron precursors into the cortex, and one mutant in which thalamocortical axons fail to invade the cortex (
(A and B) Comparison of E12.5 WT (A) and mutant (B) embryos showing the severe craniorachischisis of line 90 embryos.
(C) Sequencing chromatograms show the thymine in the WT sequence (top) that is an adenine in line 90 (bottom). This thymine-to-adenine substitution changes an isoleucine codon to a lysine codon within a conserved LRR domain of SCRIBBLE.
(D and E) Ribbon diagrams of an LRR fold, illustrating the predicted position (red arrowheads) of the line 90 amino acid substitution. The substitution is in a linker region between two stretches of β-sheet. An LRR fold has a β-sheet on the inside, and linker regions on the outside, of a broad half-circular curve.
(F) The predicted domain structure of SCRIBBLE indicating where the line 90 and the
(G and H) Dorsal views of the cortex of E17.5 WT (G) and line 267 mutant (H) embryos stained to reveal the expression of the
(I) The choroid plexus persists in its hypertrophic state after birth and can be seen as a pinkish lump on the head of this postnatal day 14 (P14) pup.
(J) A coronal section through the dorsal portion of the diencephalon of a P0 pup.
(K) Sequencing chromatograms show the thymine in the WT sequence (top) that is an adenine in line 267 (bottom).
(L) The predicted domain structure of MEGALIN indicating where the new stop codon is introduced by the line 267 mutation.
(M and N) Dorsal views of the cerebral cortex of E14.5 WT (M) and line 269 mutant (N) embryos stained to reveal the expression of the
(O and P) Coronal sections of E14.5 WT (O) and line 269 mutant (P) forebrains. Red arrowheads indicate the apparent span over which dorsal midline structures are lost in the mutant.
(Q) Sequence chromatograms showing the thymine in the WT sequence (top) that is a cytosine in line 269 (bottom).
(R) The predicted domain structure of RFX4 indicating where the line 269 substitution alters the protein relative to the DNA binding and the extended (B, C, and Dim) dimerization domains.
a See supporting information
b Scored on the basis of polydactyly
c These two mutants fail to complement and are likely to be allelic
d Scored on the basis of migration defects
e These two mutants have very different phenotypes. Since the interval is large the two mutants seem likely to represent two different genes
The three mutations in previously characterized loci produce alleles that differ from the existing ones in informative ways (
In
We also identified a mutation that caused an enlarged cortex, hypertrophy of the choroid plexus of the third ventricle, and abnormalities in the dorsal diencephalon (
A morphologically identifiable dorsal midline is absent from the cerebral cortex of line 269 homozygotes (
Four mutants were identified in which the morphology of the cortex was normal at E14.5, but in which the distribution of LacZ-expressing cells in the cortex was altered. Three mutants, 154, 239, and 275, were identified on the basis of defects that are apparent in whole-mount preparations. Defects in the fourth mutant, 251, were only apparent upon examining sections. All four mutations are, for unknown reasons, perinatal lethal when homozygous, and this has prevented us from analyzing adult phenotypes. In addition, the line 154, 239, and 275 mutations appear to cause an increase in spontaneous seizures and mortality in young adult heterozygotes. Because the seizures are sporadic, their basis has not yet been studied.
Line 154 has the most severe defects. Affected embryos often have fewer labeled cells in the cortex except in the most rostral regions, where they form abnormal aggregates (
(A) Disseminated immature interneurons are seen as diffuse cortical (Cx) staining in WT embryos. Subcortical expression in the SVZ of the LGE can be seen as a darkly stained, inverted crescent.
(B) Embryos homozygous for the line 154 mutation have little or no cortical staining, and the subcortical staining has aberrant streaks (pink arrowheads) and spots (white arrowhead), particularly in the frontal cortex.
(C, E, and G) In this rostral-to-caudal series of coronal sections from WT embryos, the normal MZ and IZ/SVZ migratory streams are diffusely labeled (red arrowheads), and a sharp cortical-subcortical boundary (black arrowheads) is marked by the abrupt transition between the densely stained SVZ of the LGE and the diffuse staining of the migrating interneuron precursors.
(D, F, and H) A rostral-to-caudal series of coronal sections from a line 154 mutant embryo shows the rostral spots (white arrowheads) visible in whole mount to be aggregates of cells in the cortex, and the streaky subcortical staining to be radially directed linear aggregates (pink arrowheads). The SVZ of the LGE is also noticeably thinner, and there is not a well-defined cortical-subcortical transition in the staining pattern.
(I and J) Rostral views of
Lines 239, 251, and 275 share features that suggest that the loci involved may have roles in a common regulatory process. All three mutations produce defects that are most apparent in the rostral cortex at E14.5, and all three cause anterior polydactyly (
Anterior is to the right for all limbs.
(A–C) Left hindlimbs are shown from E14.5 WT (A), line 251 (B), and line 239 (C). Yellow arrowheads point to extra digits on the anterior (thumb) side of the limb in mutants. Line 275 mutants also have anterior polydactyly. The mutants have a slight developmental delay that causes some differences in appearance of the limb buds at E14.5, when the limbs are growing rapidly.
(D) This diagram illustrates components of the
(E–J) In situ hybridization on E11.5 limb buds. Unlike in WT (E), expression of the posterior patterning gene
To identify the developmental basis of the limb defects, we examined the expression of genes involved in limb patterning. A regulatory network involving
The 239, 251, and 275 mutants have clusters of LacZ-expressing cells in the IZ and/or SVZ of the rostral cortex, whereas more caudally, defects appear milder (
(A) Frontal view of an E14.5 WT embryo.
(B) Frontal view of an E14.5 embryo homozygous for the line 239 mutation. The stream of cells leaving the SVZ of the LGE is streaky and aggregated in the rostral cortex of the mutant.
(C–E) A rostral-to-caudal series of coronal sections from an E14.5 WT embryo. WT embryos have diffuse cortical staining and a sharp boundary (black arrowheads) between the subcortical and the cortical telencephalon.
(F–H) Coronal sections from E14.5 line 239 mutant forebrains. Yellow circles indicate the area of the cortical-subcortical boundary where a large excess of migrating cells can be seen in the IZ/SVZ area. Orange arrowhead indicates aggregated cells in the IZ/SVZ of the lateral wall of the cortex. White arrowheads indicate aggregates in the medial wall. The staining in the MZ appears normal. Defects become less apparent in the more caudal sections.
(I–K) Sections from an E14.5 line 251 mutant embryo. Orange arrowheads indicate aggregates in the lateral wall of the cortex. In (I), a linear aggregate of stained cells can be seen extending from the IZ/SVZ to the MZ (red arrowhead). The cortical-subcortical boundary is well defined in line 251 mutants.
(L–N) Sections from an E14.5 line 275 mutant embryo. Yellow circles indicate aberrant staining in the cortical-subcortical boundary region. The white arrowhead indicates aggregates in the medial wall of the cortex, and the red arrowhead points to a radially directed aggregate of cells extending from the IZ/SVZ to the MZ.
To study the effects of the mutations at later stages of cortical interneuron development, we examined near-term (E18.5) animals (
(A) Coronal section through the forebrain of an E18.5 WT embryo stained for β-galactosidase and counterstained with nuclear fast red.
(B) A similar section from a line 251 mutant embryo. Unusual accumulations of stained cells can be seen in the cortex just dorsal to the striatum (white arrowheads). Aggregates of cells in the IZ/SVZ can also be seen (black arrowhead).
(C and D) Sections adjacent to those in (A) and (B) hybridized with a probe for
(E–H) Higher-magnification views of the dorsal portions of the sections in (A–D).
(I and J) Higher-magnification view of cortex. In the WT cortex (I), cells can be seen dispersed through the cortical plate (yellow arrowheads) and scattered through the MZ (blue arrowheads). Similar distributions of labeled cells can be seen in the mutant cortex (J). In contrast to the WT, however, aggregates of cells are found in IZ/SVZ (black arrowheads).
In addition to the seven mutants described in the previous sections, six other mutants were isolated. Most of these have not been characterized in detail, but one, line 412, illustrates the range of phenotypes that the
(A–D) Coronal hemisections of E15.5 WT (A and C) and line 412 (B and D) mutant embryos stained for the
(E and F) Immunostaining for L1 antigen labels the thalamocortical fibers in coronal hemisections from E14.5 WT (E) and mutant (F) embryos. The striatum and corticostriatal areas are shown. In the WT section, the thalamocortical fibers can be seen traversing the striatum through the internal capsule and coursing into the cortex. In the mutant section, a few fibers enter the striatum but do not traverse it to reach the cortex. Corticostriatal delamination can be seen as a hole in the right side of the section.
(G and H) Immunostaining for TAG1 antigen (blue arrowheads) reveals corticofugal fibers.
ENU overwhelmingly induces single-basepair substitution mutations. The mutant alleles that are produced often have relatively selective effects on protein function, and so provide valuable probes of the function of different protein domains. This is illustrated by two of the alleles described here,
The SCRB1 protein has two sets of recognizable domains: three amino-terminal LRRs and four, more carboxy-terminal, PDZ domains. The previously described allele of
The missense allele of
The
In each of these three cases, the ENU-induced allele provides novel information about the role of the gene in cortical development and suggests avenues for further exploration. As we determine the molecular nature of the other mutations, it is likely that the selective nature of the ENU-induced alleles will provide important insights into the function of other proteins that regulate cortical development.
The migration of immature interneurons has been followed using lipophilic dyes, tissue chimeras, transfection with a green fluorescent protein (GFP) expression vector, and, recently, using a
Three out of the four tangential migration defects also result in anterior polydactyly. The association of limb patterning and neuronal migration defects has not been previously reported and, given that a great deal more is known about limb patterning than about tangential migration, it is tempting to speculate on what this may imply about the migration mutants. One possibility is that shared mechanisms are used to pattern the telencephalon and the limb. It is equally possible that genes that regulate patterning in the limb regulate cell migration decisions more directly. Mutations in an
The work described here demonstrates clearly that genetic screening strategies in the mouse need not be limited to general or broad-based phenotyping approaches; a focused genetic screening strategy can provide a powerful means of dissecting a specific aspect of mammalian biology. The strategy of broad-based approaches has been an effective one for the past several years as the collection of mutants using chemically induced mutations resurged in popularity. In other genetic systems, strategies have evolved from general toward focused screens that allow mutations affecting a specific process to be identified, and it is likely that mouse genetics will progress in a similar fashion. An additional concern with mice, however, is the cost of breeding and housing, which is higher than that for other model genetic organisms. We have shown, nonetheless, that a laboratory-based, focused screening strategy is a productive pursuit. The costs of such a screen could be shared by the careful combination of reporters so that screening for several distinct processes could be carried out at once. This general idea is highlighted by our isolation of mutations in which the growth and patterning of the cortex is defective. With the exception of line 90, the identification of these mutants benefited from the easy visualization of cortical size and structure by expression of the transgene. In several cases, the interneuron migration mutants being a good example, the identification of the mutant phenotype without the transgene would have been very unlikely. Independent reporters could be used to pursue the simultaneous identification of several different classes of mutants more directly. For example, reporters that label migrating neurons with GFP, and thalamocortical axons with β-galactosidase or alkaline phosphatase, would allow both migration and axonal pathfinding mutations to be identified in the same screen. Given the large number of reporter and indicator strains that have been made over the last few years and the powerful tools that exist for gene identification, it is clear that focused screens in the mouse will provide the resources to address many questions in the coming years.
Male C57BL/6J mice were obtained from Jackson Laboratories (Bar Harbor, Maine, United States) and treated with three intraperitoneal injections of 100 mg/kg ENU (Sigma, St. Louis, Missouri, United States) spaced at 7-d intervals. Eight weeks after the last injection, the ENU-treated males were set up in breeding pairs with FVB/NJ females homozygous for the
Embryos were dissected in phosphate-buffered saline (PBS) and fixed for 45 min at room temperature in 4% paraformaldehyde (PFA) in PBS. Subsequently, embryos were washed in detergent rinse (0.1 M phosphate buffer [pH 7.3], 2 mM MgCl2, 0.01% sodium deoxycholate and 0.02% Nonidet P-40). Embryos were then stained for 48–72 h at room temperature on a rocking platform using X-gal as a substrate for the detection of β-galactosidase activity. Staining was terminated after visual inspection by repeated washing in PBS. Embryos were then fixed again and stored in 4% PFA in PBS until further examined. All litters were examined in whole mount and approximately one-half were selected for sectioning on a vibratome (VT1000S; Leica, Wetzlar, Germany). For sectioning, the heads of all pups from a litter were separated from the torsos and mounted aligned in the same orientation in an agarose block. Sections 100 μm thick were collected in PBS and mounted in Kaiser's glycerol gelatin (Merck, Darmstadt, Germany) on slides. All phenotypes reported here were seen in more than ten litters resulting from both G1 male × G2 female and G2 male × G2 female crosses.
Whole-mount in situ hybridization was carried out according to
Immunohistochemistry using anti-TAG1 (Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, Iowa, United States) and anti-L1 (Chemikon, Mönchengladback, Germany) antibodies was performed on 100-μm free-floating sections following the instructions and using the reagents of the Vectastain ABC Kit (Vector). Primary antibodies were used at a dilution of 1:500. Horseradish peroxidase activity from secondary antibodies was revealed by diaminobenzidine as a substrate in staining buffer (0.5 mg/ml diaminobenzidine, 20 mM sodium cacodylate, and 30 mN acetic acid).
Initial linkage was established using 12 DNA samples from both carriers and mutant embryos. We scored a set of 82 simple sequence repeat markers. The panel of markers was selected from the set at the Center for Inherited Disease Research Website (
For sequencing of
(A) Dorsal views of the cortex of WT (left) and line 152 mutant (right) embryos stained for expression of the Dlx-LacZ transgene. The line 152 mutation reduces the size of the cortex.
(B and C) Coronal sections through the cortex of E14.5 WT (B) and line 351 mutant (C) embryos. The cortex has a characteristic high-domed shape and is thinner in the mutant.
(D and E) Dorsal (D) and ventral (E) views of adult brains of WT (left) and line 351 mutant (right) brains. The cortex is overall larger in the mutant than in the WT. The olfactory bulbs are present in the mutant but are tucked under the cortex, making them less visible.
(9.1 MB TIF).
(A) Lateral view of an E14.5 WT embryo.
(B) The cleft upper jaw of a line 366 mutant is visible at the left. The telencephalon, including the cortex, is significantly shortened along the rostrocaudal axis.
(C and D) Lateral and front views of the cleft and reduced upper jaw of an E18.5 embryo homozygous for the line 366 mutation.
(E and F) Lateral views of WT (E) and line 357 homozygote (F) embryos at E13.5 showing the reduced telencephalon and relatively expanded midbrain.
(G) A front view of the embryo in (F) shows the cleft upper jaw.
(H and I) Sagittal sections through E13.5 WT (H) and line 357 (I) embryos. The overgrown midbrain in the mutant embryo has forced the neuroepithelium into folds.
(9.9 MB TIF).
(A) Lateral view of an E14.5 mutant embryo showing edema and hemorrhage suggestive of vascular defects. Frontal views of WT (B) and mutant (C) embryos illustrate the narrowed frontonasal process, maxilla, and mandible of the mutant. (D) shows coronal hemisections in a WT embryo. (E) shows the accumulation of Dlx-LacZ–positive cells in a SVZ-like area dorsal to the LGE.
(4.3 MB TIF).
Accession numbers of the genes discussed in this paper are available at LocusLink (
We thank Dr. Amir Ashique for permission to cite unpublished work, Dr. Viktor Kharazia for help and advice with histological analysis, and Marlene Lane for help with many aspects of animal breeding and husbandry and for help with the early stages of the screen. We also thank Deanna Louie, Kristina Bray, Tyler Wai Chan, and William Tseng for their assistance.
This research was supported in part by funds provided by the National Institutes of Health/National Institute of Mental Health, award number R21 MH068525 (ASP); a Human Frontiers Science Program Grant, number RG0160/2000-B (ASP); a Wills Foundation Fellowship (KZ); and a Broad Foundation Fellowship (SM). It was also supported in part by funds provided by the State of California for medical research on alcohol and substance abuse through the University of California, San Francisco.
apical ectodermal ridge
anterior-posterior
embryonic day [number]
ethyl-nitroso-urea
green fluorescent protein
intermediate zone
lateral ganglionic eminence
leucine-rich repeat
marginal zone
postnatal day [number]
phosphate-buffered saline
paraformaldehyde
wild-type