Huwe1 ubiquitin ligase is essential to synchronize
neuronal and glial differentiation in the
developing cerebellum
Domenico D’Arcaa, Xudong Zhaoa, Wenming Xua, Nadya C. Ramirez-Martineza, Antonio Iavaronea,b,c,1, and Anna Lasorellaa,b,d,1
aInstitute for Cancer Genetics and Departments ofbPathology,cNeurology, anddPediatrics, Columbia University Medical Center, New York, NY 10032
Edited* by Stephen P. Goff, Columbia University College of Physicians and Surgeons, New York, NY, and approved February 17, 2010 (received for review November 6, 2009)
We have generated a knockout mouse strain in which the gene coding for the ubiquitin ligase Huwe1 has been inactivated in cerebellar granule neuron precursors (CGNPs) and radial glia. These mice have a high rate of postnatal lethality and profound cerebellar abnormalities. The external granule layer of the cerebellum, which contains CGNPs, is expanded and displays aberrant proliferation and impaired differentiation of the progenitor cell population. The uncontrolled proliferation of the CGNPs is associated with accumu-lation of the N-Myc oncoprotein, a substrate of Huwe1, and con-sequent activation of the signaling events downstream to N-Myc. Furthermore, loss of Huwe1 in Bergmann glia leads to extensive disorganization of this cell population with layering aberrations, severe granule neuron migration defects, and persistence of ecto-pic clusters of granule neurons in the external granule layer. Our findings uncover an unexpected role for Huwe1 in regulating Berg-mann glia differentiation and indicate that this ubiquitin ligase orchestrates the programming of the neural progenitors that give rise to neurons and glia in the cerebellum.
HECT domain
|
cell migration|
nervous system|
development|
tumor suppressorD
evelopment of the cerebellum requires spatiotemporal reg-ulation of cell proliferation, differentiation, and migration of neuronal and glial cells. Granule cells are the most abundant neurons of the cerebellum. Progenitors of these neurons actively proliferate during thefirst 2 postnatal weeks in the most super-ficial layer of the developing cerebellar cortex, the external granule layer (EGL). Later, developing cerebellar granule neuron precursors (CGNPs) exit the cell cycle, extend axons, and migrate inward past the Purkinje cell bodies to theirfinal destination, the internal granule layer (IGL). As the inward migration of granule cells progresses to form the IGL, the EGL disappears. Glial cells participate as key cellular components in most of the steps of cerebellar development. Bergmann glia are unipolar cerebellar astrocytes that originate from the radial glia in the ventricular zone. During late embryogenesis, Bergmann glia migrate to the cerebellum and align next to the Purkinje cell layer, emitting radialfibers to the pial surface early after birth (1). The radial fibers of Bergmann glia provide guidance for granule neuron mig-ration and form synapses with Purkinje cell dendrites (2).We have recently shown that Huwe1 deletion in the neural stem/progenitor cell compartment of the embryonic brain leads to profound disorganization of the laminar patterning of the cere-bral cortex. These defects are caused, at least in part, by impaired neurogenesis with uncontrolled expansion of the neural stem cell (NSC) compartment (3). However, deletion of Huwe1 in the nestin-expressing compartment results in neonatal lethality, thus preventing the evaluation of Huwe1 loss in tissues with predom-inant postnatal development, such as the cerebellum.
To investigate the role of Huwe1 in postnatal development of the nervous system, we deleted the HECT domain-containing region of Huwe1 in the mouse brain using the Cre-loxP system
and a GFAP-cre deleter strain, to target Huwe1 deletion to neural and glial precursors (4). We show that the deletion of Huwe1 in neurons and glia leads to severe lethality in the first 3–4 weeks after birth. Huwe1-null cerebellum exhibited defec-tive cell cycle withdrawal and differentiation of CGNPs and disruption of the scaffold organization of radial glia, leading to impaired granule neuron migration and heterotopic granule neuron clusters at the pial surface.
Results
Huwe1 Deletion Results in Postnatal Lethality and Severe Defects in Cerebellar Development. To study the role of the ubiquitin ligase Huwe1 in cerebellar development, we crossed conditional knockout mice for the Huwe1 gene (Huwe1Flox) (3) with mice carrying a human GFAP-Cre transgene (4). When driven by the GFAP pro-moter, expression of the Cre recombinase begins at embryonic day 13 and is detected in cerebellar granule neurons and Bergmann glia but not Purkinje cells (4). Because the Huwe1 gene is X-linked, we performed all our analyses on male offspring in which inactivation of the single Huwe1 allele results in the Huwe1-null genotype. Cross-ings between Huwe1Flox/Xfemales and GFAP-Cre transgenic males generated Huwe1Flox/YGFAP-Cre animals (hereafter referred to as Huwe1F/YGFAP). We observed efficient deletion of Huwe1 in the cerebellum (Fig. 1A). Western blot analysis of Huwe1 protein using two different antibodies, one recognizing the N terminus (N-Ter) and the other the HECT domain at the C terminus (HECT), confirmed that the HECT domain was absent in the cerebellum of Huwe1F/YGFAP mice. Moreover, the mutant Huwe1 protein was markedly reduced, confirming our previous findings in the cortex (Fig. 1B) (3). In the cerebellum at postnatal day 8 (P8), Huwe1 was highly expressed in Purkinje cells (Fig. 1C, yellow arrowheads). In the granule neuron layers, Huwe1 was low in the EGL and increased markedly as neurons migrate into the IGL (Fig. 1C Left). At P8, the Bergmann glia provide essential scaffold functions for migration of postmitotic neurons from the EGL to the IGL (1, 2). Double immunostaining for GLAST, a Bergmann glia specific marker (5), or the Bergmann glia and astrocytic marker GFAP and Huwe1 showed that Bergmann glia cells stain positive for nuclear Huwe1 (Fig. 1C Left, white arrowheads). White matter astrocytes were also positive for Huwe1 (Fig. 1C Left, white arrows). Con-sistent with the reported pattern of expression of the Cre
Author contributions: A.I. and A.L. designed research; D.D., X.Z., W.X., N.C.R.-M., and A.L. performed research; D.D., X.Z., A.I., and A.L. analyzed data; and A.I. and A.L. wrote the paper.
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor. Freely available online through the PNAS open access option.
1To whom correspondence may be addressed. E-mail: [email protected] or al2179@
columbia.edu.
This article contains supporting information online atwww.pnas.org/cgi/content/full/ 0912874107/DCSupplemental.
DEVELOPMENTA
L
recombinase in the GFAP-Cre transgenic mouse, Huwe1 immunostaining with antibodies targeting the HECT domain revealed that the intact Huwe1 protein was lost in the EGL, IGL, Bergmann glia, and astrocytes but not Purkinje cells of Huwe1F/YGFAP mice (Fig. 1 C and D Right).
Huwe1F/YGFAP mice were born at the expected ratio. However, more than 50% Huwe1F/YGFAP mice died within 4 weeks of post-natal age (Fig. 1E). Before the weaning age, the Huwe1F/YGFAP mice became significantly smaller than control littermates and began to show ataxic symptoms such as abnormal posture and hyperextension of the hind limbs (Fig. 1F). When suspended by the tail, Huwe1F/Y GFAP mice displayed limb withdrawal and digit clasping. Histological analysis revealed that the Huwe1F/YGFAP cerebellum was similar to control at P0 (Fig. 2A). However, a markedly smaller mutant cerebellum was observed at P8. At this age, Huwe1 mutant cerebella had a conserved pattern of foliation with relatively enlarged EGL and underdeveloped IGL (Fig. 2B). The Purkinje cell layer was also abnormal with crowding and mis-alignment of Purkinje cells in mutant mice as shown by immuno-staining for calbindin, a Purkinje cell-specific marker (Fig. 2C). As Huwe1 is not deleted in Purkinje neurons, the described defects are likely to be secondary to abnormalities of granule neurons or Bergmann glia. Although attenuated, the defect in the EGL per-sisted at P15, suggesting a delay in the development of the cer-ebellar cortex in the mutant mouse (Fig. 2D).
Severe Defects of Cell Cycle Withdrawal and Differentiation of
Granule Neuron Precursors in Huwe1F/YGFAP Mice. During
post-natal development, the EGL is divided into two sublayers: the granule cell progenitors continue to proliferate in the outer sub-layer, whereas postmitotic granule cells extend bipolar axons and move horizontally in the inner part of the EGL (IEGL), prior to extending a third process perpendicular to the bipolar axon and initiate the migration into the IGL (6). To assess how loss of Huwe1 activity affected these crucial steps of granule cell development, we evaluated proliferation and differentiation of granule cells. To study cell proliferation, we performed quantitative analysis of the proliferation marker Ki67 at P8, the time when proliferative expansion of granule precursors is at peak, and P15, the time when the EGL is no longer evident, as postmitotic granule cells have completed migration into the IGL. When compared with litter-mate controls, we found a higher number of proliferating cells both
A
B
C
D
E
F
25 50 1.0 0.8 0.6 0.4 0.2 0.0 0 2 4 6 8 10 12 ML IGL ML IGL (n =10) (n =32) EGL ML IGL EGL ML IGL EGL ML IGL EGL ML IGLFig. 1. Ataxic phenotype and postnatal lethality in Huwe1F/YGFAP mice. (A) The upper three panels are results from tail DNA genotyping for Huwe1 loxP allele, GFAP-Cre, and Sex-determining Region Y, SRY in representative P5 mice from heterozygousfloxed Huwe1 females bred onto GFAP-Cre heterozygous males (lanes 4 and 5). Bottom panel is genomic PCR for Huwe1 using DNA from the cerebella of mice in the Upper panels. Lane 1 is PCR control for the Huwe1 WT allele. Lane 2 is PCR control for the recombined Huwe1 allele. Lane 3 is PCR control for Huwe1 loxP plus Huwe1 recombined alleles. (B) Western blot analysis of cer-ebellum lysates from mice in lanes 4 and 5 of A using an antibody against the HECT domain of the Huwe1 protein (HECT) or the N-terminal region (N-Ter).α-Tubulin is shown as control for loading. (C) Expression of Huwe1 protein in the cerebellum of Huwe1F/YGFAP and control mice at P8. Sagittal sections were immunostained using Huwe1 N-Ter antibody and antibodies against GLAST (Upper), or GFAP (Lower). Besides granule neurons, Bergmann glia, GLAST/GFAP-positive cells (white arrowhead), and GFAP-postive astrocytes in the IGL (arrows) were also positive for Huwe1 in control but not mutant cerebellum. Purkinjie cells (yellow arrowheads) expressed high levels of Huwe1 in wild-type and mutant cerebella. (D) Immunofluorescence using the Huwe1-HECT antibody confirmed that the HECT domain of Huwe1 was efficiently deleted in the Huwe1F/YGFAP granule neurons, Bergmann glia, and astrocytes but not Purkinjie cells (P15). (E) Kaplan-Meier curves of wild-type and Huwe1F/YGFAP mice. (F) Growth retardation and ataxia in Huwe1F/YGFAP mice. EGL, external granule layer; ML, molecular layer; IGL, internal granule layer.
A
B
C
D
1000 50 1000 50 100 50 1000 EGL ML EGL ML IGL IGL ML EGL ML EGLFig. 2. Histological analysis of Huwe1F/YGFAP cerebellum. (A) Hematoxylin and eosin staining of sagittal sections from P0 Huwe1F/Yand Huwe1F/YGFAP cer-ebella. (B) Hematoxylin and eosin staining from P8 Huwe1F/Yand Huwe1F/YGFAP cerebella. Higher magnificaton view (Middle and Lower) of framed area in the Upper panels shows an enlarged EGL, decreased thickness of the ML, and reduced cell density in the IGL in Huwe1F/YGFAP compared to Huwe1F/Y. (C) Immunofluorescence for the Purkinje cell marker calbindin in Huwe1F/Yand Huwe1F/YGFAP P8 mice. (D) Hematoxylin and eosin staining from P15 Huwe1F/Y and Huwe1F/YGFAP cerebella. (Lower) Higher magnificaton view of framed area in Upper. P15 Huwe1F/YGFAP mice exhibit a residual EGL and migrating granule cells are visible in the ML (arrows).
in P8 and P15 mutant mice (Fig. 3 A and B andFig. S1AUpper; P8,
P = 1.85293E-07; P15, P = 0.000975). Consistently, mutant granule cell precursors exhibited a 2-fold increase in phosphorylated his-tone H3 (pHH3), a marker of mitosis (Fig. 3 C and D andFig. S1A
Lower). Ki67-positive granule cells forming heterotopic clusters at the pial surface were identified until P25 in Huwe1F/YGFAP mice (Fig. S2A). In the surviving adult Huwe1F/YGFAP mice (12 months) the ectopic granule cell clusters persisted, although they had lost Ki67 immunoreactivity and expressed the neuronal dif-ferentiation marker NeuN (Fig. S2 B and C).
Previous studies have shown that Huwe1 targets N-Myc for ubiquitination and degradation in neural cells (3, 7). N-Myc activity is necessary for proliferation of granule cell precursors (8, 9) and down-regulation of N-Myc is required for their ter-minal cell cycle exit and differentiation (10). When compared with controls, P8 Huwe1F/YGFAP EGL exhibited an enlarged N-Myc-positive compartment and concordant increased expression of cyclin D2, a transcriptional target of N-Myc in CGNPs (11) (Fig. 3 E and F). A similar pattern of expression of N-Myc and cyclin D2 was observed in P15 cerebella and paralleled the his-tological defects at this age (Fig. S1B). Conversely, the cyclin-dependent kinase (cdk) inhibitor p18Ink4C, whose expression is derepressed in the N-myc-null cerebellum (11), was markedly reduced in Huwe1F/YGFAP cerebella (Fig. 4A). Furthermore, the expression of the cdk inhibitor p27Kip1, which marks the
post-mitotic granule cells in the premigratory EGL and the IGL in control cerebella (12), was also reduced in mutant mice, being restricted to a limited strip of the IEGL and displaying a het-erogeneous pattern in the IGL (Fig. 4B). Accordingly, the ratio of p27Kip1-positive to total nuclei was significantly reduced in Huwe1F/YGFAP mice (Fig. 4B, P = 4.85935E-05).
Besides proliferation, Huwe1 was proposed as direct regulator of programmed cell death (13–15). However, the deregulated activity of Myc oncoproteins is generally sufficient to trigger secondary apoptosis (16). To establish the exact relationship between the two phenotypic events, we conducted a parallel and quantitative analysis of proliferation (Ki67) and apoptosis (cleaved caspase3) in Huwe1F/YGFAP mice at two stages of cerebellar development (P0 and P8,Fig. S3). At P0, proliferation was already elevated in the Huwe1 mutant cerebellum but apoptosis was unaffected (Fig. S3 A and B). However, the hyperproliferative activity in the mutant cerebellum at P8 (Fig. 3 A–D) was also accompanied by elevated numbers of apoptotic cells (Fig. S3C). Together, thesefindings indicate that the pri-mary consequence of deleting Huwe1 in the mouse cerebellum is enhancement of cell proliferation, and persistent proliferative stress eventually triggers secondary apoptosis.
The morphology and behavior of CGNPs in the mutant cer-ebella raised the possibility that differentiation might be dis-turbed. To test this hypothesis, we examined the pattern of expression of doublecortin (DCX). DCX is a molecular marker of migrating and differentiating neuroblasts in the central and peripheral nervous system (17, 18). In control cerebella at P8, DCX was highly expressed in the premigratory region of the EGL (IEGL) and identified T-shaped axons. DCX was also detected in granule cells migrating radially in the ML (Fig. 4C, white arrow-heads, trailingfibers), in postmitotic granule cells migrating tan-gentially (Fig. 4C Left; white arrows, spindle-shaped cells; yellow arrowheads, parallelfibers), and in several neurons in the IGL (Fig. 4C Lower Left). The premigratory region of Huwe1F/YGFAP cerebella expressed lower amounts of DCX and contained only round-shaped cells, which lacked parallelfibers and the perpen-dicular process (Fig. 4C Upper Right, arrows). The population of DCX-positive neurons in the IGL was also markedly reduced in Huwe1F/YGFAP cerebella (Fig. 4C Lower Right). Immunostaining for the mature neuronal marker NeuN confirmed that a reduced number of cells in the IEGL had acquired differentiation features and fewer mature granule cells had reached the IGL in the mutant cerebella compared with controls (Fig. 4D). Together, these results indicate a delay in granule cell differentiation and suggest that Huwe1 functions as an intrinsic factor necessary for executing terminal cell cycle exit and initiating the differentiation programs in the external granule precursor cells.
Severe Bergmann Glia Abnormalities inHuwe1F/YGFAP Mice.Analysis of Huwe1 expression indicated that Huwe1 was efficiently deleted in cerebellar glial cells in Huwe1F/YGFAP cerebella (Fig. 1C). Therefore, we asked whether defects of the cerebellar glia are also implicated in the phenotype of Huwe1F/YGFAP mice. The Bergmann glia are the major glial cell type in the cerebellar cortex. During postnatal life Bergmann glia undergo significant morphological changes to support granule neuron differentiation and migration (1, 2). The protein BLBP is not only a useful marker for radial glia but it is also an important protein for granule cell migration, given the observation that anti-BLBP antibodies block the migration of granule cells along radial glia fibers (19). Bergmann glia cell bodies align nearly in a single layer adjacent to Purkinje cells and exhibit unipolarfibers, which extend and contact the pial surface. In P8 wild-type cerebella, immunostaining with the anti-BLBP antibody identified the classical palisade of Bergmann glia extensions in the ML. Scat-tered BLBP-positive cells were also present in the IGL, marking astrocytes (Fig. 5A Left). In mutant mice, the pattern of BLBP
A
B
C
D
E
F
ML ML IGL IGL IGL IGL IGL IGL 50 50 50 50Fig. 3. Huwe1F/YGFAP mice display increased proliferation and accumulation of N-Myc and cyclin D2 in the EGL of postnatal cerebella. (A) Sagittal sections of the cerebellum of wild-type and Huwe1F/YGFAP mice at P8 were immu-nostained with anti-Ki67 antibody. (B) Histograms are quantification of Ki67+ cells normalized by area (150μm2) in P8 (Left) and P15 (Right) cerebella. Bars indicate mean± SD n = 3 for Huwe1F/Yand n = 3 for Huwe1F/YGFAP; P = 1.85293E-07 (P8) and P = 0.000975 (P15). (C) Sagittal sections of cerebella from P8 Huwe1F/Yand Huwe1F/YGFAP mice were immunostained for the mitotic marker pHH3. (D) Histograms are quantification of pHH3+cells normalized by area (150μm2) in P8 (Left) and P15 (Right) cerebella. Bars indicate mean± SD n = 3 for Huwe1F/Yand n = 3 for Huwe1F/YGFAP; P = 7.59553E-05 (P8) and P = 0.001447 (P15). Sagittal sections of cerebella from P8 Huwe1F/Y and Huwe1F/YGFAP mice were immunostained for N-Myc (E) and cyclin D2 (F).
DEVELOPMENTA
L
immunostaining demonstrated the disorganized microarchitecture of the cerebellar folia. Bergmann glia cell bodies were not aligned but BLBP-positive cells were irregularly located in both the EGL and IGL (Fig. 5A Right). Immunostaining for the Bergmann glia marker GLAST confirmed that most mutant Bergmann glia cells had rudimentary or irregularfibers that lacked contact with the pial surface and were completely misaligned (Fig. 5B). Moreover, GFAP-positive, mature astrocytes were reduced in the deep white matter of mutant cerebella (Fig. 5C).
Loss of Huwe1 Impairs Migration of Granule Neurons. Mature
Bergmann glia provide the indispensable scaffold for cerebellar neuron migration and positioning into the IGL. Given the pro-found defect of the Bergmann glia scaffold in Huwe1F/YGFAP cerebellum, we predicted that radial migration of granule cells might be perturbed in mutant mice. To test this hypothesis, we pulse labeled in vivo granule neurons born at P6 with bromo-deoxyuridine (BrdU) and analyzed them after 20 h and 110 h, respectively. Fig. 6A shows immunostaining for BrdU to identify the labeled granule cells and calbindin to identify Purkinje cells, thus marking the ML/IGL boundary. After 20 hours from labeling, the number of BrdU-positive cells that entered the IGL was small in both controls and mutant cerebella (Fig. 6A Upper). However, evaluation of the ratio of positive cells in the ML to BrdU-positive cells in the IGL suggested that a defective migration was already detectable in mutant mice at this early time (Fig. 6B). After 110 h, a vast majority of BrdU-positive granule cells had migrated into the IGL in control cerebella. Conversely, mutant cerebella retained more than 80% of the BrdU-positive cells in the ML and only a small number of BrdU-labeled cells reached the IGL (Fig. 6A Lower and Fig. 6B). The failure of granule cell migration reflects the dual defect of differentiation in granule and glial cells that is manifested by the collapse of the intricate neuron–glia interactions that regulate cerebellar development (Fig. 6C).
A
B
C
D
Fig. 4. Defective differentiation of granule neurons in Huwe1F/YGFAP mice. (A) Sagittal sections of cerebella from P8 Huwe1F/Yand Huwe1F/YGFAP mice were immunostained for p18Ink4C. (B Left) Sagittal sections of cerebella from P8 Huwe1F/Yand Huwe1F/YGFAP mice were immunostained for p27Kip1(red). Nuclei were counterstained with DAPI (blue). (B Right) Histogram shows the percentage of p27Kip1-positive cells in the EGL. Bars indicate mean± SD n = 3 for Huwe1F/Y and n = 3 for Huwe1F/YGFAP; P = 4.85935E-05. (C) Immunofluorescence using DCX, a marker of migrating neuroblasts in P8 wild-type and Huwe1F/YGFAP cerebella. Spindle-shaped cells (arrows in Upper Left) are present in the premigratory area of EGL presumably migrating along parallelfibers (yellow arrowheads). Huwe1 mutant EGL shows accumulation of round cells (arrows in Upper Right). Migratory leading processes perpendicular to the pia (white arrowheads in Upper Left) are identified in controls but are rudimentary or absent in the Huwe1-null granule cells. DCX-positive cells are present in the IGL of control mice but are markedly reduced in Huwe1F/YGFAP mice (Lower). (D) Immunofluorescence using NeuN (red), a marker of terminally differentiated neurons, in P8 wild-type and Huwe1F/YGFAP cerebella. Nuclei were counterstained with DAPI (blue). In mutant cerebella NeuN immunostaining exhibits an irregular pattern in the premigratory region of the EGL (IEGL, Upper) and IGL, in which cell density is markedly reduced (Lower).
A
B
C
Fig. 5. Defects in Bergman glia scaffold in Huwe1F/YGFAP mice. (A) Immu-nofluorescence for BLBP on sagittal sections from Huwe1F/Yand Huwe1F/YGFAP P8 mice. (B) Immunofluorescence for GLAST on sagittal sections from Huwe1F/Y and Huwe1F/YGFAP P8 mice. (C) Immunofluorescence for GFAP on sagittal sections from Huwe1F/Yand Huwe1F/YGFAP P8 mice. In control littermates, Bergmann glia are aligned with Purkinje cell bodies and extend processes to the pial surface. Huwe1 mutant Bergmann glia are misplaced, Bergmann glia fibers are enlarged, and most of them lack their contacts to the pial surface (arrows in C). An increased number of BLBP-positive cells occupy the mutant IGL (arrowheads in A Right).
Discussion
In this study, we demonstrated that Huwe1 plays a key role in differentiation of CGNPs and maturation of Bergmann glia during the postnatal development of the cerebellum. Detailed analysis showed that the proliferative window of Huwe1-null granule neurons was prolonged and was associated with elevated N-Myc and cyclin D2 and reduced p18Ink4Cand p27Kip1. The examination of several differentiation markers of granule neurons and Berg-mann glia in the developing cerebellum revealed remarkable changes in these cell populations after Huwe1 deletion, supporting the notion that the interplay between the Bergmann glial scaffold and migrating granule neurons is vital for granule neuron migra-tion and formamigra-tion of thefinal structure of the IGL. Besides the phenotypes in granule neurons and Bergmann glia, Purkinje cells were misaligned and abnormally crowded. A recent study sug-gested that maturation and alignment of Purkinje cells requires support from extended Bergmannfibers (5, 20). Indeed, the lack of GFAP-Cre expression and Huwe1 deletion in Purkinje cells indi-cates that phenotypic alterations of Purkinje cells in Huwe1-null cerebella are likely secondary to the defective Bergmann glial scaffold. Together with our recent demonstration that Huwe1 is required for timely cell cycle withdrawal, differentiation, and patterning of the developing cortex, the present study establishes the essential role of the Huwe1 ubiquitin ligase in coordinating maturation of neurons and glia in the nervous system.
The development of laminar structure of the cerebellum is orchestrated by signals from diverse cell types. The intimate structural association between granule neurons and Bergmann glialfibers led to the long-standing notion that elongated and anchored radial Bergmann glial cells serve as a scaffold that is crucial for granule neuron migration (21). Genetic models have indicated that a number of cell intrinsic and extrinsic signals con-tribute to Bergmann glia maturation and cerebellar lamination (22–29). One of the intrinsic factors is the PTEN tumor suppressor. It has been reported that GFAP-Cre-mediated deletion of PTEN
causes defects of the Bergmann glia similar to those we describe in the Huwe1 knockout cerebellum (30). However, an important difference among these two mouse strains (Huwe1 null and PTEN null) is the absence of neuronal abnormalities in the PTEN mutant cerebellum, whereas we found that granule neuron differentiation defects are present in the Huwe1 mutant cerebellum. This differ-ence points to two important conclusions: (i) exclusive Bergmann glia defects (as seen in PTEN knockout) result in impaired neu-ronal migration but do not affect the intrinsic ability of neurons to progress through the programmed differentiation steps; and (ii) dual neuron–glia differentiation abnormalities (as in Huwe1 knockout) not only cause a defective neuronal migration but also enhance the basal proliferative capacity of granule precursors and delay cell cycle withdrawal leading to the formation of ectopic clusters of granule cells. Whether direct or indirect interactions exist between the Huwe1-N-Myc and the PTEN-AKT pathways in Bergmann glia will have to be elucidated by future studies. How-ever, the similarity of thefindings suggests that the physiological role of Huwe1 in cerebellar development may be linked to the prodifferentiation function of a tumor suppressor protein.
The deregulated proliferation conferred by loss of Huwe1 ulti-mately leads to severe perturbation of CGNP differentiation and disorganization of the cerebellar architecture. Within the prolonged proliferative window of Huwe1-null CGNPs, the N-Myc oncoprotein, a Huwe1 substrate, accumulated aberrantly with consequent activa-tion of cyclin D2 and repression of p18Ink4Cand p27Kip1, downstream events triggered by N-Myc in CGNPs (9, 31–33). Although our work does not rule out the possibility that the accumulation of other, yet unknown Huwe1 substrates in granule neurons (and Bergmann glia) may also participate in the manifestation of the Huwe1-null pheno-type, the importance of N-Myc inactivation for coordinated cell cycle withdrawal and differentiation of CGNPs has beenfirmly established in previous studies (8, 10, 34). One interestingfinding of our study was the observation of ectopic clusters of granule neurons at the pial surface in adult mutant mice. These clusters originated from CGNPs with extended proliferation window, which were able eventually to 50
25
A
B
C
Fig. 6. Huwe1F/YGFAP mice display profound de-fects of granule neuron migration. (A) Granule cell migration was analyzed by BrdU pulse labeling. Mice at P6 were injected intraperitoneally with BrdU and distribution of BrdU-labeled neurons (red) was determined at 20 h (Upper) and 110 h (Lower) after injection. Calbindin immunostaining (green) was used to mark the border between ML and IGL. Compared with littermate controls, Huwe1 mutant mice exhibited increased number of BrdU-positive granule cells in the ML and correspondingly reduced number of BrdU-positive granule cells in the IGL. (B) Bar graph shows quantification of BrdU-labeled neurons expressed as the ratio between BrdU-labeled neurons in ML and IGL. (20 h, n = 3; P = 0.007055; 110 h, P = 0.000699). (C) Double immuno-fluorescence for DCX (neurons) and GFAP (glia) shows that most of the interactions among granule neurons and radial glia are lost in mutant cerebella (white arrows, migrating granule cells; white arrowheads, radial trailing fibers; yellow arrow-heads, parallelfibers).
DEVELOPMENTA
L
execute a normal program of differentiation and did not progress to a preneoplastic or tumorigenic state. This is consistent with data obtained in many systems, which indicate that N-Myc overexpression alone is not sufficient to transform granule precursor cells but cooperates with different genetic lesions to promote medulloblas-toma, the tumor derived from granule precursors. Among these lesions is loss of the Sonic hedgehog (Shh) antagonist Ptc (35, 36), which constitutively activates the signaling pathway that promotes proliferation of the CGNPs in the EGL during cerebellar develop-ment and cell cycle regulators such as the cyclin-dependent kinase inhibitor Ink4c and p27Kip1(37–41). It remains to be tested whether Huwe1 deletion cooperates with such lesions to promote medullo-blastoma in the mouse and whether, as found in human gliomedullo-blastoma, Huwe1 is also inactivated in medulloblastoma (3).
Materials and Methods
Generation ofHuwe1FloxGFAP Mice. To generate Huwe1Flox/YGFAP-Cre ani-mals heterozygous or homozygousfloxed Huwe1 females (Huwe1Flox/Xor Huwe1Flox/Flox) were bred onto GFAP-Cre heterozygous males. Huwe1 mutants were genotyped by PCR of genomic DNA prepared from tail using primers described previously (3). All animal experiments were approved by and performed in accordance with the guidelines of the International Agency for Research on Cancer’s Animal Care and Use Committee. Tissue Preparation, Histology, and Immunostaining. Histological analysis was carried out on 5-μm paraffin sections stained with hematoxylin and eosin (H&E). For granule neuron migration experiments, 50μg/g body weight bromodeox-yuridine (BrdU, Sigma-Aldrich) was injected intraperitoneally in P6 mice. Mice
were killed after 20 h and 110 h. The primary antibodies used were: anti-Huwe1 HECT-domain (Lifespan Biosciences), anti-Huwe1-Lasu1 (Bethyl Laboratories), anti-GFAP (Cell Signaling Technology), anti-GLAST (Millipore), anti-BLBP (Abcam), anti-Calbindin (Chemicon), anti-p18Ink4C(Santa Cruz), anti-p27Kip1 (Thermo Scientific), anti-Doublecortin (Santa Cruz), anti-NeuN (Calbiochem), anti-BrdU (Roche), anti-Ki67 (Novocastra), anti-N-Myc (Calbiochem), anti-Cyclin D2 (NeoMarkers), and anti-cleaved caspase3 (Cell Signaling). Fluorescent detection was performed using Cy3 or FITC-labeled secondary antibodies (Jackson Immunoresearch Laboratories) or using biotin-conjugated secondary antibodies (Vector Laboratories) followed by Vectastain Elite ABC development and the Tyramide amplification system (Perkin-Elmer).
Western Blot. Cerebella were triturated using RIPA lysis buffer. Equal amounts of total protein were separated on SDS/PAGE gel. After blocking in 5% dry milk in PBS, blots were incubated with primary antibodies against Huwe1-Lasu1 N terminus (Bethyl Laboratories), Huwe1 HECT-domain (Lifespan Biosciences), or α-tubulin (Sigma) before exposure to peroxidase-conjugated secondary anti-bodies and detection using ECL enhanced chemiluminescence (Amersham).
Statistical Analysis. Each experiment was performed with samples from at least three animals from two independent litters. In histograms, values rep-resent the mean values; error bars are standard deviations. Statistical sig-nificance was determined by t test (with Welch’s correction) using Prism 4.0 software (GraphPad).
ACKNOWLEDGMENTS. We thank Angelica Castano for technical assistance with immunohistochemistry.This work was supported by grants from the National Institutes of Health–National Cancer Institute to A.L. (R01CA131126) and A.I. (R01CA085628). D.D. is supported by fellowships from Provincia di Benevento/Ministero del Lavoro, Italy.
1. Chanas-Sacre G, Rogister B, Moonen G, Leprince P (2000) Radial glia phenotype: Origin, regulation, and transdifferentiation. J Neurosci Res 61:357–363.
2. Yamada K, Watanabe M (2002) Cytodifferentiation of Bergmann glia and its relationship with Purkinje cells. Anat Sci Int 77:94–108.
3. Zhao X, et al. (2009) The N-Myc-DLL3 cascade is suppressed by the ubiquitin ligase Huwe1 to inhibit proliferation and promote neurogenesis in the developing brain. Dev Cell 17:210–221.
4. Zhuo L, et al. (2001) hGFAP-cre transgenic mice for manipulation of glial and neuronal function in vivo. Genesis 31:85–94.
5. Yamada K, et al. (2000) Dynamic transformation of Bergmann glialfibers proceeds in correlation with dendritic outgrowth and synapse formation of cerebellar Purkinje cells. J Comp Neurol 418:106–120.
6. Hatten ME, Heintz N (1995) Mechanisms of neural patterning and specification in the developing cerebellum. Annu Rev Neurosci 18:385–408.
7. Zhao X, et al. (2008) The HECT-domain ubiquitin ligase Huwe1 controls neural differentiation and proliferation by destabilizing the N-Myc oncoprotein. Nat Cell Biol 10:643–653.
8. Kenney AM, Cole MD, Rowitch DH (2003) Nmyc upregulation by sonic hedgehog signaling promotes proliferation in developing cerebellar granule neuron precursors. Development 130:15–28.
9. Knoepfler PS, Cheng PF, Eisenman RN (2002) N-myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Genes Dev 16:2699–2712.
10. Sjostrom SK, Finn G, Hahn WC, Rowitch DH, Kenney AM (2005) The Cdk1 complex plays a prime role in regulating N-myc phosphorylation and turnover in neural precursors. Dev Cell 9:327–338.
11. Knoepfler PS, Kenney AM (2006) Neural precursor cycling at sonic speed: N-Myc pedals, GSK-3 brakes. Cell Cycle 5:47–52.
12. Miyazawa K, et al. (2000) A role for p27/Kip1 in the control of cerebellar granule cell precursor proliferation. J Neurosci 20:5756–5763.
13. Bernassola F, Karin M, Ciechanover A, Melino G (2008) The HECT family of E3 ubiquitin ligases: Multiple players in cancer development. Cancer Cell 14:10–21. 14. Chen D, et al. (2005) ARF-BP1/Mule is a critical mediator of the ARF tumor suppressor.
Cell 121:1071–1083.
15. Zhong Q, Gao W, Du F, Wang X (2005) Mule/ARF-BP1, a BH3-only E3 ubiquitin ligase, catalyzes the polyubiquitination of Mcl-1 and regulates apoptosis. Cell 121:1085–1095. 16. Hoffman B, Liebermann DA (2008) Apoptotic signaling by c-MYC. Oncogene 27:
6462–6472.
17. Francis F, et al. (1999) Doublecortin is a developmentally regulated, microtubule-associated protein expressed in migrating and differentiating neurons. Neuron 23:247–256. 18. Gleeson JG, Lin PT, Flanagan LA, Walsh CA (1999) Doublecortin is a
microtubule-associated protein and is expressed widely by migrating neurons. Neuron 23:257–271. 19. Feng L, Hatten ME, Heintz N (1994) Brain lipid-binding protein (BLBP): A novel
signaling system in the developing mammalian CNS. Neuron 12:895–908. 20. Lordkipanidze T, Dunaevsky A (2005) Purkinje cell dendrites grow in alignment with
Bergmann glia. Glia 51:229–234.
21. Rakic P (1971) Neuron-glia relationship during granule cell migration in developing cerebellar cortex. A Golgi and electronmicroscopic study in Macacus Rhesus. J Comp Neurol 141:283–312.
22. Adams NC, Tomoda T, Cooper M, Dietz G, Hatten ME (2002) Mice that lack astrotactin have slowed neuronal migration. Development 129:965–972.
23. Rio C, Rieff HI, Qi P, Khurana TS, Corfas G (1997) Neuregulin and erbB receptors play a critical role in neuronal migration. Neuron 19:39–50.
24. Eiraku M, et al. (2005) DNER acts as a neuron-specific Notch ligand during Bergmann glial development. Nat Neurosci 8:873–880.
25. Lutolf S, Radtke F, Aguet M, Suter U, Taylor V (2002) Notch1 is required for neuronal and glial differentiation in the cerebellum. Development 129:373–385.
26. Schmid RS, et al. (2003) Neuregulin 1-erbB2 signaling is required for the establishment of radial glia and their transformation into astrocytes in cerebral cortex. Proc Natl Acad Sci USA 100:4251–4256.
27. Weller M, Krautler N, Mantei N, Suter U, Taylor V (2006) Jagged1 ablation results in cerebellar granule cell migration defects and depletion of Bergmann glia. Dev Neurosci 28:70–80.
28. Gaiano N, Nye JS, Fishell G (2000) Radial glial identity is promoted by Notch1 signaling in the murine forebrain. Neuron 26:395–404.
29. Lin Y, et al. (2009) Neuron-derived FGF9 is essential for scaffold formation of Bergmann radialfibers and migration of granule neurons in the cerebellum. Dev Biol 329:44–54. 30. Yue Q, et al. (2005) PTEN deletion in Bergmann glia leads to premature
differentiation and affects laminar organization. Development 132:3281–3291. 31. Ciemerych MA, et al. (2002) Development of mice expressing a single D-type cyclin.
Genes Dev 16:3277–3289.
32. Nakamura M, Matsuo T, Stauffer J, Neckers L, Thiele CJ (2003) Retinoic acid decreases targeting of p27 for degradation via an N-myc-dependent decrease in p27 phosphorylation and an N-myc-independent decrease in Skp2. Cell Death Differ 10:230–239.
33. Zindy F, et al. (2006) N-Myc and the cyclin-dependent kinase inhibitors p18Ink4c and p27Kip1 coordinately regulate cerebellar development. Proc Natl Acad Sci USA 103:11579–11583. 34. Kenney AM, Widlund HR, Rowitch DH (2004) Hedgehog and PI-3 kinase signaling
converge on Nmyc1 to promote cell cycle progression in cerebellar neuronal precursors. Development 131:217–228.
35. Lum L, Beachy PA (2004) The Hedgehog response network: Sensors, switches, and routers. Science 304:1755–1759.
36. Rohatgi R, Scott MP (2007) Patching the gaps in Hedgehog signalling. Nat Cell Biol 9: 1005–1009.
37. Browd SR, et al. (2006) N-myc can substitute for insulin-like growth factor signaling in a mouse model of sonic hedgehog-induced medulloblastoma. Cancer Res 66:2666–2672. 38. Zindy F, et al. (2007) Genetic alterations in mouse medulloblastomas and generation of tumors de novo from primary cerebellar granule neuron precursors. Cancer Res 67: 2676–2684.
39. Kessler JD, et al. (2009) N-myc alters the fate of preneoplastic cells in a mouse model of medulloblastoma. Genes Dev 23:157–170.
40. Ayrault O, Zindy F, Rehg J, Sherr CJ, Roussel MF (2009) Two tumor suppressors, p27Kip1 and patched-1, collaborate to prevent medulloblastoma. Mol Cancer Res 7:33–40. 41. Uziel T, et al. (2005) The tumor suppressors Ink4c and p53 collaborate independently
Supporting Information
D
’Arca et al. 10.1073/pnas.0912874107
Ki67/DAPI Ki67/DAPI
N-Myc/DAPI
Cyclin D2/DAPI Cyclin D2/DAPI
N-Myc/DAPI A B 50 ˩m 50 ˩m pHH3/DAPI pHH3/DAPI
Huwe1F/Y Huwe1F/YGfapa
Huwe1F/Y Huwe1F/YGfap
Fig. S1. P15 Huwe1F/YGFAP cerebella display increased proliferation, accumulation of N-Myc and cyclin D2 in the EGL. (A) Sagittal sections of the cerebellum of wild-type and Huwe1F/YGFAP mice at P15 were immunostained with anti-Ki67 antibody (red, Upper) and anti-pHH3 antibody (red, Lower). (B) Sagittal sections of cerebella from P15 Huwe1F/Yand Huwe1F/YGFAP mice were immunostained for N-Myc antibody (red, Upper) and anti-cyclin D2 antibody (red, Lower).
A
Ki67/DAPI Ki67/DAPI
Ki67/DAPI Ki67/DAPI 100 ˩m 25 ˩m B 1000 ˩m 25 ˩m NeUN NeUN C NeUN NeUN 100 ˩m 25 ˩m
Huwe1F/Y Huwe1F/YGfap Huwe1F/Y Huwe1F/YGfap
Huwe1F/Y Huwe1F/YGfap
Fig. S2. Analysis of cerebellum of adult Huwe1F/YGFAP mice. (A) Sagittal sections of the cerebellum of wild-type and Huwe1F/YGFAP mice at P25 were im-munostained with anti-Ki67 antibody (red) and counterstained with DAPI (blue). Note the heterotopic clusters of granule cells containing Ki67 positive cells. (B) Hematoxylin and eosin staining shows clusters of granule cells at the pial surface in a representative 12-month old Huwe1F/YGFAP mouse (Right). (C) Im-munofluorescence using NeuN (red), marker of terminally differentiated neurons, demonstrated that in 12-month old Huwe1F/YGFAP mice the ectopic cell clusters are composed of differentiated cells.
35 Huwe1
F/YGFAP
Huwe1FY
*
Huwe1
F/YHuwe1
F/YGFAP
Ki67/DAPI Ki67/DAPI
A
10 15 20 25 30 Ki67 + cells /150 µµm * 0 5 KCaspase3/DAPI Caspase3/DAPI 20
Huwe1F/YGFAP
Huwe1FY
Huwe1
F/YHuwe1
F/YGFAP
B
50 μm p p Caspase3 + cells 4 8 12 16Caspase3/DAPI Caspase3/DAPI
0 4
100 Huwe1***F/YGFAP
Huwe1FY
Huwe1
F/YHuwe1
F/YGFAP
C
50 μm 20 40 60 80 pase3 + cells/folium 0 Cas 50 μmFig. S3. Induction of apoptosis in Huwe1F/YGFAP cerebella is secondary to aberrant proliferation. (A) Sagittal sections of the cerebellum of wild-type and Huwe1F/YGFAP mice at P0 were immunostained with anti-Ki67 antibody (red) and counterstained with DAPI (blue). Histogram is quantification of Ki67+cells normalized by area (150μm2). Bars indicate mean± SD n = 3 for Huwe1F/Yand n = 3 for Huwe1F/YGFAP; P = 0.0015832. (B) Sagittal sections of cerebella from P0 Huwe1F/Yand Huwe1F/YGFAP mice were immunostained for the apoptotic marker cleaved caspase3 (red) and counterstained with DAPI (blue). Histogram is quantification of cleaved caspase3+cells in the entire cerebellum. Bars indicate mean± SD n = 3 for Huwe1F/Yand n = 3 for Huwe1F/YGFAP; P = 0.373861. (C) Immunofluorescence using the apoptotic marker cleaved caspase3 (red) in P8 Huwe1F/Yand Huwe1F/YGFAP mice. Nuclei were counterstained with DAPI (blue). Histogram is quantification of cleaved caspase3+cells in each folium. Bars indicate mean± SD n = 3 for Huwe1F/Yand n = 3 for Huwe1F/YGFAP; P = 0.0004755.