Adenomatous polyposis coli is required for early events in the normal growth and differentiation of the developing cerebral cortex
- 11k Downloads
Adenomatous polyposis coli (Apc) is a large multifunctional protein known to be important for Wnt/β-catenin signalling, cytoskeletal dynamics, and cell polarity. In the developing cerebral cortex, Apc is expressed in proliferating cells and its expression increases as cells migrate to the cortical plate. We examined the consequences of loss of Apc function for the early development of the cerebral cortex.
We used Emx1 Cre to inactivate Apc specifically in proliferating cerebral cortical cells and their descendents starting from embryonic day 9.5. We observed reduction in the size of the mutant cerebral cortex, disruption to its organisation, and changes in the molecular identity of its cells. Loss of Apc leads to a decrease in the size of the proliferative pool, disrupted interkinetic nuclear migration, and increased apoptosis. β-Catenin, pericentrin, and N-cadherin proteins no longer adopt their normal high concentration at the apical surface of the cerebral cortical ventricular zone, indicating that cell polarity is disrupted. Consistent with enhanced Wnt/β-catenin signalling resulting from loss of Apc we found increased levels of TCF/LEF-dependent transcription and expression of endogenous Wnt/β-catenin target genes (Axin2 (conductin), Lef1, and c-myc) in the mutant cerebral cortex. In the Apc mutant cerebral cortex the expression of transcription factors Foxg1, Pax6, Tbr1, and Tbr2 is drastically reduced compared to normal and many cells ectopically express Pax3, Wnt1, and Wt1 (but not Wnt2b, Wnt8b, Ptc, Gli1, Mash1, Olig2, or Islet1). This indicates that loss of Apc function causes cerebral cortical cells to lose their normal identity and redirect to fates normally found in more posterior-dorsal regions of the central nervous system.
Apc is required for multiple aspects of early cerebral cortical development, including the regulation of cell number, interkinetic nuclear migration, cell polarity, and cell type specification.
KeywordsCerebral Cortex Adenomatous Polyposis Coli Ventricular Zone Cortical Plate Interkinetic Nuclear Migration
Adenomatous polyposis coli
fluorescence activated cell scanning
phosphate buffered saline
quantitative reverse transcriptase PCR
terminal deoxynucleotidyl nick end labelling.
In the mouse, the cerebral cortex develops from anterior neuroepithelium starting around half-way through gestation at embryonic day (E)9.5. During its early development the cerebral cortex is divided into the ventricular zone, the subventricular zone, the intermediate zone, and the cortical plate. Cortical progenitors have radial processes that span the depth of the cerebral cortex and have endfeet at its apical and basal surfaces. Cell division occurs in the ventricular and subventricular zones and daughter cells either undergo more cell divisions or exit the cell cycle and migrate through the intermediate zone to the cortical plate where they undergo further differentiation and project axons into the intermediate zone. Eventually, different regions of the cerebral cortex exhibit distinct cytoarchitecture and function and it is believed that the areal expression of secreted morphogens and transcription factors contribute to setting this up during development. For reviews of these processes, see [1, 2, 3, 4, 5].
APC (Adenomatous polyposis coli) was originally identified as a tumour suppressor gene mutated in familial adenomatous polyposis, an autosomal dominant condition with predisposition to colorectal cancers  and brain tumours . Mutations in APC have also been linked to a case of mental retardation . Mutations in Apc (the murine homologue of APC) have been linked to intestinal neoplasia in mice . Many genes whose dysfunction is associated with tumour formation have important functions in the normal development of the nervous system . Apc is now known to be involved in regulating a variety of cellular processes, including mitosis, cytoskeletal dynamics, axonogenesis, cell polarity and apoptosis [11, 12, 13, 14, 15, 16]. Apc is central to the Wnt signalling pathway, in which it mediates the destruction of cytoplasmic β-catenin protein  unless the cell receives a Wnt signal resulting in the stabilisation of β-catenin and its translocation to the nucleus, where it co-operates with TCF/LEF to activate the transcription of target genes. Wnt/β-catenin signalling has diverse roles in central nervous system development . Given the various functions of Apc, there are many ways in which Apc might influence the development of cerebral cortical cells.
It has previously been shown that Apc mRNA is widely expressed throughout the developing brain. In the cerebral cortex, proliferating cells express Apc mRNA and its expression levels increase as cells migrate to the cortical plate and differentiate . Apc is known to be a regulator of β-catenin activity and the levels of β-catenin signalling change as cells stop dividing and move from the ventricular zone, through the intermediate zone, and into the cortical plate . These observations raise the possibility that Apc protein might be important for proliferation and/or the subsequent differentiation of cells in the cerebral cortex. Apc has also been implicated in cerebral cortical development via its interaction with the cytoskeletal protein Lis1 , indicating that Apc has functions not directly related to Wnt/β-catenin signalling.
In this study we test the hypothesis that Apc is important for early events in the formation of the cerebral cortex using a Cre/LoxP strategy. We use a floxed Apc allele, Apc 580S , in which Cre-induced recombination leads to deletion of Apc exon 14 and a frameshift at codon 580 that has been shown to disrupt the function of Apc . We used the Emx1 Cre allele  to produce this mutation of Apc specifically in cells of the developing cerebral cortex starting at E9.5. Cerebral cortex depleted of Apc exhibits reduced size, a massive disorganisation of cell types, and a profound alteration in the identity of many of its cells. These defects coincide with increased levels of nuclear β-catenin and up-regulation of Wnt/β-catenin target genes as well as alterations to the organisation of the cytoskeleton and cell polarity.
Materials and methods
Mice harbouring various combinations of the following alleles were used in this study: Apc 580S ; Emx1 Cre ; BAT-gal ; Rosa26R . Timed matings between Emx1Cre/CreApc580S/+or Emx1Cre/+Apc580S/+and Emx1+/+Apc580S/580Sanimals were used to generate experimental embryos with the plug day designated as E0.5. Throughout this manuscript embryos are designated 'control' (Emx1Cre/+Apc580S/+) or 'mutant' (Emx1Cre/+Apc580S/580S). Other genotypes were not used.
Injection of S-phase tracers and estimation of proliferative pool and cell cycle kinetics
For estimation of the proliferative pool size, cumulative bromodeoxyuridine (BrdU) labelling was performed by injecting BrdU (200 μl of 100 μg/ml (in 0.9% NaCl) (Sigma St. Louis, MO, USA) intra-peritoneally to pregnant females every 2 hours over a 12-hour period. Embryos were collected after 12 hours from the first injection. For estimation of cell cycle kinetics, double labelling experiments were done by intra-peritoneal injection of iododeoxyuridine (IdU; 200 μl of 100 μg/ml in 0.9% NaCl (Sigma)) and BrdU (200 μl of 100 μg/ml (in 0.9% NaCl) (Sigma)) 1.5 hours later. Animals were sacrificed 30 minutes after BrdU injection following a protocol previously described . Coronal sections of cerebral cortex were immunostained for BrdU and IdU and cell cycle parameters were calculated as described .
The following pairs of primers were used for PCR genotying of genomic DNA extracted from biopsies: Apc forward 5'-CACTCAAAACGCTTTTGAGGGTTGAAT-3', reverse GTTCTGTATCATGGAAAGATAGGTGGT (product size: 226 base-pairs (wild type allele) 314 base-pairs (580S floxed allele); Emx1 forward 5'-TGGCCCAACTCGGTGTTAGG-3', reverse 5'-CCACCAAGGACTCTATGGTG-3' (product size 260 base-pairs); Cre forward 5'-ACCTGATGGACATGTTCAGGGATC-3', reverse 5'-TCCGGTTATTCAACTTGCACCATG-3' (product size 108 base-pairs); LacZ forward 5'-CGAAATCCCGAATCTCTATCGTGC-3', reverse 5'-GATCATCGGTCAGACGATTCATTGG-3' (product size 400 base-pairs).
In situ hybridisation
Embryonic heads (E13.5) were submerged in a solution of 4% paraformaldehyde, 0.1% Tween20 in phosphate buffered saline (PBS) pH9.5 for 8 to 20 hours at 4°C on a rocking platform, embedded in wax, and cut into sections 10 μm thick. In situ hybridisations for Axin2, Lef1, Wnt2b, Wnt8b, Gli1, and Ptc transcripts were performed as described previously  using digoxygenin-labelled antisense riboprobes.
Embryonic heads (E13.5 to E15.5), or whole embryos (E12.5 and earlier) were submerged in a solution of 4% paraformaldehyde in PBS for 8 to 20 hours at 4°C on a rocking platform. Heads were either: embedded in wax and cut into sections 10 μm thick; frozen and cut into sections 10 μm thick; or embedded in agarose and cut into vibratome sections 100 μm thick. Wax sections were dewaxed in xylene and rehydrated through a series of solutions of descending ethanol concentration to water. Sections were microwaved to unmask the antigen epitope. Chromogenic visualization was done with Envision+ Kit (Dako, Ely, Cambridgeshire, UK). In some cases sections were counterstained with Harris-haematoxylin (Thermo Electron Corporation, Cheshire, UK). For immunofluorescence, following incubation with the primary antibody, sections were incubated with species-specific secondary antibodies conjugated to fluorescent molecules at a 1:200 dilution: goat anti-mouse or goat anti-rabbit Alexafluor-488 or Alexafluor-568 (Invitrogen, Paisley, UK). For detection of IdU/BrdU, a mouse anti-BrdU antibody that recognises both IdU and BrdU was used (clone B44, 1:100 in blocking solution; Becton Dickinson Oxford, UK). Rat anti-BrdU antibody (clone BU1/75, 1:100; Abcam, Cambridge, UK) was used to detect BrdU but not IdU. In case of double labelling, goat anti-mouse highly cross-absorbed antibodies were used to prevent cross-reactivity. Nuclei were counterstained using the DNA dye TO-PRO-3 iodide (Invitrogen) at 1 μM. Sections were mounted under coverslips using Mowiol to prevent fading of fluorescence. The following additional primary antibodies were used at the dilutions stated: Apc(C-20) rabbit, sc-896, 1:400 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) immunohistochemistry on vibratome sections and dissociated cells); Apc (ab15270), rabbit polyclonal, 1:100 (Abcam; immunohistochemistry on wax sections); β-catenin, mouse, 610154, 1:200 (BD Biosciences UK, Oxford, UK) beta-III-tubulin, mouse, Tuj1, 1:800 (Sigma); P21, rabbit, SX118, 1:100 (BD Biosciences); Pax3, mouse, Pax3, 1:400 (DSHB, University of Iowa, Iowa City, USA); Foxg1, rabbit rabbit polyclonal antibody described in , 1:500; C-myc, mouse, 9E10, 1:200 (Roche Diagnostics GmbH, Mannheim, Germany); phosphohistone H3 (PH3), rabbit, H9908, 1:400 (Sigma); WT1, mouse, 6F-H2, 1:1000 (Dako); N-cadherin, mouse, 610920, 1:200 (BD Biosciences); Pericentrin, rabbit, PRB-432C, 1:400 (Covance Emeryville, CA, USA); Tbr1 and Tbr2, rabbit polyclonal antibodies described in , 1:500; Olig2, rabbit polyclonal antibodies described in , 1:10,000; Mash1 mouse monoclonal 24B7.2d11, 1:100 (BD Biosciences); Islet1, mouse 40.2D6, 1:100 (DSHB); Pax6, mouse, Pax6, 1:400 (DSHB).
Staining for bacterial LacZ (β-galactosidase)
Embryos were dissected in ice cold PBS and fixed with shaking at 4°C for 1 hour in LacZ fixative (PBS containing: paraformaldehyde (4%); NP40 (0.02%), sodium deoxycholate (0.01%); EGTA (5 mM); and MgCl2 (2 mM)). LacZ staining was as described previously .
Detection of apoptosis by TUNEL
To detect apoptotic cells, terminal deoxynucleotidyl nick end labelling (TUNEL) was performed according to the supplier's protocol (Roche). Total numbers of TUNEL positive cells were counted in sections of E13.5 control and mutant embryos.
Primary cell culture and FACS analysis
The telencephalon was isolated and dissected in ice-cold oxygenated Earle's buffered salts solution. The cerebral cortex was separated from the ventral telencephalon. Telencephalic tissue was dissociated using Papain dissociation system kit (Worthington Biochemical Corporation, Lakewood, NJ, USA) according to the supplier's protocol. Cells were resuspended in serum-free medium. Dissociated cells were then either: embedded in collagen  and cultured for 4 hours, after which time they were fixed and processed for Apc immunohistochemistry; or subjected to fluorescence activated cell scanning (FACS) analysis following fixation in -20°C 70% ethanol for ≥ 2 hours on ice and resuspension in PBS containing 0.05 mg/ml of propidium iodide (Sigma) and 0.5 mg/ml of RNAseA (Roche). FACS analysis was performed on FACSCalibur (BD Biosciences), which runs CellQuest software.
Slides were photographed using a Leica DMLB upright compound microscope connected to a Leica DSC480 digital camera with Leica IM50 image management software. Fluorescent staining was imaged using a Leica TCS NT confocal system with associated software. Alexafluor-488 staining was collected in the FITC (green) channel, Alexafluor-568 in the TRITC (red) channel and TO-PRO-3 in the Cy3 (far-red, pseudo-coloured blue) channel.
Statistical analysis and graph plotting
Sigmastat (Systat Software Inc., Richmond, CA, USA) was used for data analysis. Excel (Microsoft) was used to plot data.
cDNA synthesis and quantitative reverse transcriptase PCR
RNA was extracted from the cerebral cortical tissue at E12.5, E13.5, and E15.5 using RNEasy mini kit (Qiagen, Crawley, Sussex, UK) followed by treatment with DNAse (Roche) to eliminate genomic DNA. The SuperScript III First-Strand Synthesis SuperMix kit (Invitrogen) was used to synthesize cDNA from RNA samples. The procedure was performed according to the supplier's protocol using random hexamers. The following primers were used for quantitative reverse transcriptase PCR (qRT-PCR): Apc exon 4 forward 5'-CCGTTCAGGAGAATGCAGTC-3', reverse 5'-TGCCGTCTTGTCATGTCTGT-3'; Apc exon 14 forward 5'-GTGTCCAGCTTGATAGCTAC-3', reverse 5'-CAAGGCTTCCTGGTCTTTAG-3'; Wnt1 forward 5'-ATACGACCCCGTTTCTGCTG-3', reverse 5'-TTCCACTCCCTCACCTCAAAGC-3'; Axin2 forward 5'-ACAGTAGCGTAGATGGAGTC-3', reverse 5'-CTGTGGAACCTGCTGCCTTC-3'; Wnt8b forward 5'-AAGGCTTACCTGGTCTACTC-3', reverse 5'-CAGAGCTGATGGCGTGCACA-3'; GAPDH 5'-GGGTGTGAACCACGAGAAAT-3' and 5'-CCTTCCACAATGCCAAAGTT-3'. qRT-PCR was performed using Qiagen Quantitect SYBR green PCR kit and a DNA Engine Opticon continuous fluorescence detector (Genetic Research Instrumentation, Rayne, Essex, UK) The abundance of each transcript in the original RNA sample was extrapolated from PCR reaction kinetics using Opticon software. Transcript levels are expressed relative to GAPDH.
Apc expression in developing wild-type and mutant cerebral cortex
We used Cre recombinase driven by the endogenous Emx1 promoter [24, 33] to inactivate Apc expression from a floxed Apc allele in neural cells in the developing cerebral cortex starting at E9.5 when the cerebral cortex is starting to form. LacZ staining of embryos harbouring Emx1 Cre and R26R reporter alleles confirms the ability of Cre recombinase to recombine the floxed-stop LacZ reporter allele in the cerebral cortex of these embryos (Figure 1Q, R). Emx1 is not expressed in the most dorsal medial telencephalon (the cortical hem region) [33, 34] and this region escapes Cre-mediated recombination (Figure 1Q, R). The efficiency of Cre recombination (judged from the intensity of LacZ staining) appears to diminish towards the lateral edges of the cerebral cortex (Figure 1Q, R) and, for this reason, we concentrated our subsequent analysis of the Apc mutant phenotype on the central portion of the cerebral cortex. We used the Apc 580S allele in which exon 14 of Apc is flanked by LoxP sites . Transcripts produced from the Apc 580S allele following Cre-mediated recombination lack exon 14, causing a frameshift, and translation of these transcripts is predicted to produce non-functional Apc. In order to confirm that Apc levels were reduced in the cerebral cortex of our mutant, we performed qRT-PCR using primers to exon 4 (present in transcripts produced from the floxed Apc both before and after Cre-mediated recombination) and exon 14 (only present in transcripts produced from un-recombined Apc) on RNA extracted from dorsal telencephalon of control and mutant embryos at E12.5, E13.5, and E15.5. The ratio of Apc transcripts containing exon 14 to those containing exon 4 provides a measure of the efficiency with which wild-type Apc transcripts were eliminated from the mutant cortex. At all ages examined we found that the exon 14:exon 4 ratio was reduced in the mutant compared to the controls (Figure 1S), indicating successful reduction in the levels of wild-type Apc transcripts in the mutant. The low levels of wild-type transcripts in RNA samples from the mutant cortex might reflect subpopulations of cells within the cerebral cortex that escaped recombination in the mutant (for example blood vessels) and/or persistence of wild-type Apc transcripts after recombination of the genomic locus. Comparison of Apc immunofluorescence on sections and collagen-embedded dissociated cells from control (Figure 1B, E, G, J, L) and mutant (Figure 1C, H, K, M) cerebral cortex showed that the intensity of immunostaining was much lower in the mutant cells, indicating successful depletion of Apc protein. The intense spots of Apc staining in Figure 1H, K correspond to non-specific staining of blood vessels by the fluorescent secondary antibody (blood vessels are strongly stained in controls reacted with the fluorescent secondary antibody alone; Figure 1N, O). Apc immunohistochemistry on dissociated cells indicates that some cells in the mutant cerebral cortex retain Apc protein, albeit at reduced levels to that seen in control cerebral cortex (compare Figure 1L to 1M). In summary, mutant embryos are unable to express Apc mRNA or Apc protein at the same levels as their wild-type counterparts and do not upregulate Apc/Apc expression as occurs during normal cerebral cortical development.
Neuronal differentiation is disrupted when Apc is lost from cerebral cortex
Loss of Apc leads to a reduction in the proliferative pool
The small size of the mutant cerebral cortex prompted us to ask whether there was a defect in proliferation. Continuous cumulative injection of the thymidine analogue BrdU labels all cells that are undergoing S-phase during the pulse .
Loss of Apc causes increased apoptosis
Disruption to spatial organisation of the cell cycle and cell polarity
β-Catenin is disregulated following loss of Apc
Activation of Wnt/β-catenin target gene expression
Loss of Apc function causes cerebral cortical cells to adopt alternative fates
In the preceding section we demonstrate that loss of Apc function results in increased nuclear β-catenin and aberrant activation of Wnt/β-catenin target gene expression in the cerebral cortex. Wnt signalling is important for specifying cell identity in the embryonic central nervous system, so we next examined whether cerebral cortical cells alter their identity following loss of Apc.
Cells in the mutant cerebral cortex exhibit gene expression patterns normally diagnostic of elevated Wnt/β-catenin signalling (Figure 8) in combination with reduced levels of the transcription factor Foxg1 (Figure 9). These molecular properties are characteristic of cells in the cortical hem region and presumptive hippocampus of the medial dorsal telencephalon [59, 60, 61], prompting us to speculate that cells in the mutant cerebral cortex had adopted the fate of this region. To test this hypothesis, we used in situ hybridisation to examine the expression of Wnt2b, which is normally expressed at high levels in the cortical hem region, and Wnt8b, which is normally expressed in the cortical hem region with expression extending dorsally into the presumptive hippocampus [59, 62]. Contrary to our hypothesis, we found that neither Wnt2b (compare Figure 11K and 11L) nor Wnt8b (compare Figure 11M and 11N) were ectopically expressed in the mutant cerebral cortex (compare staining intensity in the cerebral cortex to that in the cortical hem/hippocampal region in Figure 11K–N). qRT-PCR on RNA extracted from control and mutant dorsal telencephalon revealed a twofold increase in Wnt8b expression (Figure 11O). As Wnt8b expression had not extended into the mutant cerebral cortex, this modest increase might be a local up-regulation of Wnt8b expression within the cortical hem region or reflect the reduced size of the mutant cerebral cortex relative to the cortical hem region. The Emx1 Cre allele we used to inactivate Apc is not active in the cortical hem itself (Figure 1) , so it is not surprising that the molecular characteristics of the cortical hem are not affected in Apc mutants (compare cortical hem region expression of Axin2 and Lef1 expression in Figure 8 and Wnt2b and Wnt8b expression in Figure 11 between control and mutant embryos). The mutant cerebral cortex does not, therefore, correspond to a lateral expansion of cortical hem or hippocampal identity.
The loss of Apc function therefore causes dorsal telencephalic cells to adopt molecular characteristics normally associated with more posterior-dorsal parts of the central nervous system.
Shh signalling is not upregulated following loss of Apc
Discussion and conclusion
In this study we have examined the consequences of removing Apc protein from the developing mouse cerebral cortex after E9.5, an age by which it has already acquired molecular characteristics of the cerebral cortex. Apc is a multifunctional protein with roles in cell proliferation, differentiation, migration, and apoptosis , so it is perhaps not surprising that we find disruption of a multitude of developmental processes in Apc mutants.
Disruption to β-catenin function in the Apc mutant
It is likely that many of the defects observed in conditional Apc mutants stem from a disregulation of β-catenin function, although loss of Apc function may also have consequences independent of β-catenin. The earliest abnormality we detected was an increase in nuclear β-catenin at E10.5 followed by up-regulation in the expression of β-catenin/TCF/Lef transcriptional targets, including the BAT-gal reporter transgene and endogenous targets c-myc, Lef1 and Axin2. A key role of Apc is to mediate the destruction of β-catenin, a process that is blocked when the cell receives a Wnt signal . Loss of Apc function might, therefore, be expected to produce similar effects to Wnt signalling and/or β-catenin stabilisation. Transgenic mice in which a stabilised β-catenin mutant protein was expressed in central nervous system progenitor cells under the control of a nestin promoter had shown dramatically enlarged cerebral cortex surface area, although lamination was not affected [74, 75]. Increased brain size was explained by expansion of the progenitor pool due to cells re-entering the cell cycle instead of differentiating. In contrast, our Apc mutants exhibit a massive reduction in the size of the cerebral cortex, resulting from a combination of reduced proliferative pool size and increased cell death by apoptosis, and lamination is disrupted. These differences can likely be explained by differences in the timing of β-catenin stabilisation since the timing of Wnt/β-catenin signalling is known to be critical for neural cell fate decisions, and stabilising β-catenin during early cerebral cortical development promotes proliferation while at later stages it promotes neural differentiation [76, 77, 78, 79, 80]. The nestin promoter used by Chenn and Walsh results in expression of stabilised β-catenin in neural precursors  before they become committed to a cerebral cortical fate and start to express Emx1 at E9.5, whereas in our model nuclear β-catenin is first detected at E10.5 after the onset of Emx1 expression.
The magnitude of Wnt/β-catenin signalling experienced by cells is tightly regulated during development by the combined action of Wnt proteins and their antagonists. A number of Wnt proteins are secreted by cells in the cerebral cortex and by the cortical hem in the medial telencephalon [59, 82]. Wnt antagonists – for example, Sfrp and Dickkopf proteins – are expressed in complex spatiotemporal patterns during cerebral cortical development [83, 84, 85, 86, 87]. By disrupting Apc we short-circuited these regulatory mechanisms as β-catenin is stabilised, and translocated to the nucleus, regardless of the presence of a Wnt signal. This results in the expression of Wnt/β-catenin target genes at higher levels than normally experienced by cerebral cortical cells. The severe consequences for cerebral cortical development emphasise the importance of maintaining the correct levels of Wnt/β-catenin signalling for the acquisition of the correct fate by cells in the developing cerebral cortex. One of our major findings is that many cells in the Apc mutant stop expressing genes normally expressed by cerebral cortex and start to express genes normally expressed elsewhere, indicating altered identity (see below). Critically, these changes occur after the cerebral cortex has become committed to its fate as Apc is inactivated after the expression of Emx1. Our results indicate, therefore, that acquisition of cerebral cortical neural fates is reversible. An interesting possibility is that Apc is required to stabilise the acquisition of cerebral cortical cell fates and that loss of Apc allows cells to redirect to alternative fates.
Apc and cell polarity
A striking feature of Apc mutants is the disorganisation of the cerebral cortex. In wild-type embryos the nuclei of cells undergoing M-phase line up along the ventricular surface and then travel deeper into the ventricular zone before entering S-phase. When Apc is inactivated the molecular and cellular properties of the ventricular zone are severely affected. Pericentrin, β-catenin, and N-cadherin proteins are normally concentrated at the apical surface of the ventricular zone and this normally polarised organisation is lost in the Apc mutants. The normal positioning of M-phase nuclei at the apical surface and S-phase nuclei away from the apical surface is lost in the mutants. This suggests that Apc is required for the establishment or maintenance of apical identity and indeed we observed particularly high levels of Apc protein at the apical surface of the ventricular zone from E11.5 to E14.5. Apc might function by polarising the response to Wnt signalling within the cell and/or by tethering proteins (including pericentrin, β-catenin, and N-cadherin) and regulating their polarised distribution within the cell. N-cadherin is a cell adhesion molecule that functions by coupling to the cytoplasmic catenin proteins (including β-catenin). Developing cerebral cortex lacking N-cadherin  exhibits many defects that resemble those seen when Apc is inactivated (present study), including disrupted interkinetic nuclear migration, loss of cell polarity, and disorganisation of cerebral cortical layers. A conditional β-catenin loss of function mutant  also exhibited disrupted interkinetic nuclear migration and cerebral cortical disorganisation reminiscent of both our Apc mutant and the N-cadherin mutant. This provides support for the hypothesis that Apc, β-catenin, and N-cadherin co-operate genetically during normal cerebral cortical development and that loss of Apc function causes loss of N-cadherin and β-catenin functions. Mechanistically, this could occur at the level of disruption to cytoskeletal protein interactions and/or by means of altered gene expression in the Apc mutant. In the developing limb bud, nuclear β-catenin down-regulates expression of cadherin  and the altered N-cadherin distribution seen in our mutant could be caused by a down-regulation in gene expression caused by elevated levels of nuclear β-catenin.
Apc and cell identity
The position of cells in the developing cerebral cortex is normally tightly regulated and it is possible that the failure of cells to adopt their correct positions in the Apc mutant might affect their molecular environment and this might cause them to adopt alternative fates. It has recently been shown that conditional deletion of the small Rho-GTPase cdc42 in the developing cerebral cortex disrupts interkinetic nuclear migration and cell polarity, with fewer cells undergoing mitosis at the apical surface of the ventricular zone . The corresponding increase in basally located mitosis correlates with an increase in the number of cells acquiring the fate of basal progenitors and expressing the transcription factor Tbr2. Although the cdc42 mutant resembles our Apc mutant with respect to the shift in balance from apical to basal mitosis, we actually see a decrease in the numbers of Tbr2 expressing cells (Figure 10). Executing mitosis away from the ventricular surface in Apc mutants is, therefore, not sufficient to direct cells to a basal cerebral cortical Tbr2 expressing fate. The Apc mutant is not the simplest model in which to study correlations between basal mitosis and post-mitotic Tbr2 expression as the loss of cerebral cortical identity found in Apc mutants might independently contribute to the reduction in the numbers of cells expressing Tbr2. It is extremely unlikely that the alterations to cell identity in the Apc mutant stem from an ectopic upregualtion of the Shh pathway as cells in the Apc mutant cerebral cortex do not ectopically express the Shh target genes Gli1, Ptc, or Olig2.
There is accumulating evidence that high levels of Wnt/β-catenin signalling during early vertebrate development promote posterior neural cell fates at the expense of anterior neural cell fates and that the formation of anterior head structures requires the active suppression of Wnt/β-catenin signalling [91, 92, 93, 94]. Wnt/β-catenin signalling is also tightly regulated along the dorsal-ventral axis such that higher levels of Wnt/β-catenin activity generally map to dorsal central nervous system structures .
The failure of the Apc mutant cerebral cortex to express transcription factors Pax6, Foxg1, Tbr1, and Tbr2 indicates that they lose their cerebral cortical identity, raising the question of what identity they acquire. One possibility is that ectopic activation of Wnt/β-catenin signalling in the cerebral cortex of the Apc mutant might redirect cerebral cortical cells to adopt fates normally associated with more posterior-dorsal central nervous system structures (such as dorsal diencephalon, dorsal midbrain, dorsal hindbrain and dorsal spinal cord) where Wnt/β-catenin signalling is high. To test this hypothesis, we examined the expression of three genes (Pax3, Wnt1, and Wt1) normally expressed in posterior/dorsal brain structures. Pax3 is normally expressed dorsally in the epithalamus, in the ventricular zone at the mesencephalic-rhombencephalic border, in the dorsal part of the ventricular zone and the roof plate of the medulla oblongata, and the dorsal spinal cord [54, 95]. Wnt1 expression is normally restricted to the dorsal and ventral midline of the midbrain and caudal diencephalons, a narrow ring rostral to the midbrain-hindbrain junction and the roof plate of the spinal cord [56, 57, 58]. The transcription factor Pax3 can regulate the expression of Wnt1 by binding to DNA sequences in its promoter , which may account for the similarities in their expression domains. Wt1 is normally expressed in the roofplate of the midbrain and in the spinal cord . Conversely, Gli1, Ptc, Mash1, Olig2, and Islet1 are normally expressed in ventral regions of the central nervous system where Wnt/β-catenin signalling is generally low . Our finding that ectopic expression of Pax3, Wnt1, and Wt1, but not Gli1, Ptc, Mash1, Olig2, or Islet1, occurs in the Apc mutant cerebral cortex is consistent with the idea that loss of Apc simulates enhanced Wnt/β-catenin signalling and pushes cerebral cortical cells towards posterior-dorsal, but not ventral, fates. Following this reasoning, it might be predicted that the Apc mutant cerebral cortex would ectopically express molecular characteristics of the cortical hem region and presumptive hippocampus, dorsal-medial telencephalic structures adjacent to the cerebral cortex and characterised by high levels of Wnt/β-catenin signalling and expression of Wnt genes, including Wnt2b and Wnt8b and low expression of Foxg1 [25, 59, 60]. However, we found no ectopic expression of Wnt2b or Wnt8b in the Apc mutant cerebral cortex, raising the possibility that some factor in the cerebral cortex in the Apc mutant somehow protects it from lateral to medial re-patterning and acquiring cortical hem region identity. Foxg1 [50, 60] is a good candidate for this factor. Conditional Foxg1 mutants in which Foxg1 is deleted after E13.5 do not exhibit an enlarged cortical hem region [97, 98], although Foxg1-/- null embryos do exhibit large scale lateral to medial re-patterning of the telencephalon . Our Apc mutants express Foxg1 normally at E10.5 but subsequently lose expression in the cerebral cortex with almost none left at E13.5. One possibility, therefore, is that early expression of Foxg1 is sufficient to prevent the acquisition of dorsal-medial telencephalic fate by the more lateral cerebral cortex. Other possibilities are that the levels of β-catenin transcriptional activation in our mutants are not appropriate to specify cortical hem region identity on the cerebral cortex or that there is a requirement for Apc itself in this process.
Apc and cell differentiation
The activation of Wnt/β-catenin mediated gene expression varies during the division and maturation of cerebral cortical cells . Proliferating cells in the ventricular zone exhibit high levels of β-catenin signalling, which is down-regulated as they exit the cell cycle and start their radial migration to the cortical plate and then up-regulated as they reach their destinations in the cortical plate . Loss- and gain-of-function experiments have identified roles for β-catenin in neuronal differentiation as well as proliferation [20, 76, 77, 78, 79, 80]. In this study we identified an up-regulation of Apc/Apc during the early development of the cerebral cortex that coincides with increasing numbers of post-mitotic cortical plate cells expressing high levels of Apc. This raised the possibility that Apc is required for the proper differentiation of cells within the cortical plate. Consistent with this idea, we showed that the transcription factors Tbr1 and Tbr2 do not exhibit their normal laminar distribution in the Apc mutant cerebral cortex. The failure of mutant cells to occupy their normal locations indicates defects in cell adhesion or migration consistent with previously described roles for Apc [99, 100].
We found that the Apc mutant cerebral cortex has lost many aspects of its molecular identity by E13.5. The severe disruption to multiple aspects of cerebral cortical development we see in Apc mutants by E13.5 makes it difficult to address the primary function(s) of Apc in the differentiation of more mature cell types generated at later stages because it becomes increasingly difficult to disentangle the primary consequences of Apc disruption from the secondary consequences of major tissue insult and re-specification. In order to directly address the primary functions of Apc in the differentiation of cerebral cortical cells, it will be necessary to manipulate Apc function in these cells at later stages of development than in the current Emx1Cre/+Apc580S/580Stransgenic model.
We are grateful to Rowena Smith for help with in situ hybridizations, Linda Wilson and Trudi Gillespie for help with confocal imaging (IMPACT facility, Edinburgh), Jan Vrana (Institute for Stem Cell Research, Edinburgh) for help with FACS analysis, and Colin Smith (Western General Hospital, Edinburgh) for help with antibodies. We thank Tetsuo Noda (Tokyo Cancer Institute, Japan) for the Apc 580S mice, Stefano Piccolo (University of Padua, Italy) for the BAT-gal mice, Takuji Iwasato and Shigeyoshi Itohara (RIKEN Brain Sciences Institute, Japan) for the the Emx1 Cre mice, and Ian Simpson for help with characterising the Emx1 Cre mice. We thank Yoshiki Sasai, Robert Hevner, and David Rowitch for kindly providing antibodies. We are grateful to the Developmental Studies Hybridoma Bank (Iowa, USA) for supplying antibodies. We thank Vassiliki Fotaki and Thomas Theil for helpful discussions. This project was funded by MRC, Wellcome Trust, and BBSRC grants to JOM and DJP. YC is funded by a University of Edinburgh Scholarship and an ORSAS. UI was funded by a PhD Scholarship from The Darwin Trust of Edinburgh.
- 11.Dikovskaya D, Schiffmann D, Newton IP, Oakley A, Kroboth K, Sansom O, Jamieson TJ, Meniel V, Clarke A, Nathke IS: Loss of APC induces polyploidy as a result of a combination of defects in mitosis and apoptosis. J Cell Biol. 2007, 176: 183-195. 10.1083/jcb.200610099.PubMedCentralCrossRefPubMedGoogle Scholar
- 13.Mogensen MM, Tucker JB, Mackie JB, Prescott AR, Nathke IS: The adenomatous polyposis coli protein unambiguously localizes to microtubule plus ends and is involved in establishing parallel arrays of microtubule bundles in highly polarized epithelial cells. J Cell Biol. 2002, 157: 1041-1048. 10.1083/jcb.200203001.PubMedCentralCrossRefPubMedGoogle Scholar
- 16.Louie RK, Bahmanyar S, Siemers KA, Votin V, Chang P, Stearns T, Nelson WJ, Barth AI: Adenomatous polyposis coli and EB1 localize in close proximity of the mother centriole and EB1 is a functional component of centrosomes. J Cell Sci. 2004, 117: 1117-1128. 10.1242/jcs.00939.PubMedCentralCrossRefPubMedGoogle Scholar
- 22.Shibata H, Toyama K, Shioya H, Ito M, Hirota M, Hasegawa S, Matsumoto H, Takano H, Akiyama T, Toyoshima K, Kanamaru R, Kanegae Y, Saito I, Nakamura Y, Shiba K, Noda T: Rapid colorectal adenoma formation initiated by conditional targeting of the Apc gene. Science. 1997, 278: 120-123. 10.1126/science.278.5335.120.CrossRefPubMedGoogle Scholar
- 25.Maretto S, Cordenonsi M, Dupont S, Braghetta P, Broccoli V, Hassan AB, Volpin D, Bressan GM, Piccolo S: Mapping Wnt/beta-catenin signaling during mouse development and in colorectal tumors. Proc Natl Acad Sci USA. 2003, 100: 3299-3304. 10.1073/pnas.0434590100.PubMedCentralCrossRefPubMedGoogle Scholar
- 30.Englund C, Fink A, Lau C, Pham D, Daza RA, Bulfone A, Kowalczyk T, Hevner RF: Pax6, Tbr2, and Tbr1 are expressed sequentially by radial glia, intermediate progenitor cells, and post-mitotic neurons in developing neocortex. J Neurosci. 2005, 25: 247-251. 10.1523/JNEUROSCI.2899-04.2005.CrossRefPubMedGoogle Scholar
- 37.Hasegawa S, Sato T, Akazawa H, Okada H, Maeno A, Ito M, Sugitani Y, Shibata H, Miyazaki Ji J, Katsuki M, Yamauchi Y, Yamamura Ki K, Katamine S, Noda T: Apoptosis in neural crest cells by functional loss of APC tumor suppressor gene. Proc Natl Acad Sci USA. 2002, 99: 297-302. 10.1073/pnas.012264999.PubMedCentralCrossRefPubMedGoogle Scholar
- 39.Venesio T, Balsamo A, Scordamaglia A, Bertolaso M, Arrigoni A, Sprujevnik T, Rossini FP, Risio M: Germline APC mutation on the beta-catenin binding site is associated with a decreased apoptotic level in colorectal adenomas. Mod Pathol. 2003, 16: 57-65. 10.1097/01.MP.0000042421.83775.0E.CrossRefPubMedGoogle Scholar
- 45.Yan D, Wiesmann M, Rohan M, Chan V, Jefferson AB, Guo L, Sakamoto D, Caothien RH, Fuller JH, Reinhard C, Garcia PD, Randazzo FM, Escobedo J, Fantl WJ, Williams LT: Elevated expression of axin2 and hnkd mRNA provides evidence that Wnt/beta -catenin signaling is activated in human colon tumors. Proc Natl Acad Sci USA. 2001, 98: 14973-14978. 10.1073/pnas.261574498.PubMedCentralCrossRefPubMedGoogle Scholar
- 46.Wetering van de M, Sancho E, Verweij C, de Lau W, Oving I, Hurlstone A, Horn van der K, Batlle E, Coudreuse D, Haramis AP, Tjon-Pon-Fong M, Moerer P, Born van den M, Soete G, Pals S, Eilers M, Medema R, Clevers H: The beta-catenin/TCF-4 complex imposes a crypt progenitor phenotype on colorectal cancer cells. Cell. 2002, 111: 241-250. 10.1016/S0092-8674(02)01014-0.CrossRefPubMedGoogle Scholar
- 71.Mizuguchi R, Sugimori M, Takebayashi H, Kosako H, Nagao M, Yoshida S, Nabeshima Y, Shimamura K, Nakafuku M: Combinatorial roles of olig2 and neurogenin2 in the coordinated induction of pan-neuronal and subtype-specific properties of motoneurons. Neuron. 2001, 31: 757-771. 10.1016/S0896-6273(01)00413-5.CrossRefPubMedGoogle Scholar
- 79.Machon O, Backman M, Machonova O, Kozmik Z, Vacik T, Andersen L, Krauss S: A dynamic gradient of Wnt signaling controls initiation of neurogenesis in the mammalian cortex and cellular specification in the hippocampus. Dev Biol. 2007, 311: 223-237. 10.1016/j.ydbio.2007.08.038.CrossRefPubMedGoogle Scholar
- 91.Heisenberg CP, Houart C, Take-Uchi M, Rauch GJ, Young N, Coutinho P, Masai I, Caneparo L, Concha ML, Geisler R, Dale TC, Wilson SW, Stemple DL: A mutation in the Gsk3-binding domain of zebrafish Masterblind/Axin1 leads to a fate transformation of telencephalon and eyes to diencephalon. Genes Dev. 2001, 15: 1427-1434. 10.1101/gad.194301.PubMedCentralCrossRefPubMedGoogle Scholar
- 93.Mukhopadhyay M, Shtrom S, Rodriguez-Esteban C, Chen L, Tsukui T, Gomer L, Dorward DW, Glinka A, Grinberg A, Huang SP, Niehrs C, Belmonte JC, Westphal H: Dickkopf1 is required for embryonic head induction and limb morphogenesis in the mouse. Dev Cell. 2001, 1: 423-434. 10.1016/S1534-5807(01)00041-7.CrossRefPubMedGoogle Scholar
- 99.Watanabe T, Wang S, Noritake J, Sato K, Fukata M, Takefuji M, Nakagawa M, Izumi N, Akiyama T, Kaibuchi K: Interaction with IQGAP1 links APC to Rac1, Cdc42, and actin filaments during cell polarization and migration. Dev Cell. 2004, 7: 871-883. 10.1016/j.devcel.2004.10.017.CrossRefPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.