![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
1 Biologie Moléculaire du Stress 2 Génomique de la Levure, Centre National de la Recherche Scientifique (CNRS) UMR8541, Ecole Normale Supérieure, 75005 Paris, France; 3 Turku Centre for Biotechnology, University of Turku, Åbo Adademi University, 20520 Turku, Finland; 4 Department of Biology, Åbo Akademi University, 20520 Turku, Finland; 5 Groupe régulateurs et effecteurs de la neurogenèse 6 Groupe de Biologie Cellulaire des Homéoprotéines, Centre National de la Recherche Scientifique (CNRS) UMR8542, Ecole Normale Supérieure, 75005 Paris, France; 7 Institut National de la Santé et de la Recherche Médicale (INSERM) 368, Ecole Normale Supérieure, 75005 Paris, France
 |
Abstract |
|---|
 Top Abstract ResultsDiscussionMaterials and methodsAcknowlegdmentsReferences |
|---|
[Keywords: Corticogenesis; heat-shock factor; p35–Cdk5; radial cortical migration]
Received September 23, 2005; revised version accepted January 31, 2006.
). A unique gene constitutes HSF in yeast, nematode, and fruit fly, whereas a family of four members is present in vertebrates. HSF1 and HSF2 are found in all vertebrate species, while HSF3 is specific for avian species and HSF4 is specific for mammals (Rabindran et al. 1991; Sarge et al. 1991; Schuetz et al. 1991; Nakai and Morimoto 1993; Nakai et al. 1997; Råbergh et al. 2000; Hilgarth et al. 2004; Le Goff et al. 2004). In vertebrates, HSF1 is the stress-responsive prototype, which cannot be substituted by any other HSF in stress-inducible hsp gene expression or in acquired thermotolerance (McMillan et al. 1998; Xiao et al. 1999; Zhang et al. 2002).
A developmental role for the HSFs began to emerge when the Drosophila HSF was found to be required for oogenesis and early larval development (Jedlicka et al. 1997). Strikingly, these developmental effects of Drosophila HSF are not mediated by hsp gene induction. The basal expression levels of hsps during embryonic development in mouse are not affected by the lack of HSF1 (Xiao et al. 1999). Therefore, other target genes are likely to be controlled by HSF1 in development. Recently, binding of HSF1 and HSF4 to the FGF-7 promoter with opposing effects on FGF-7 gene expression suggested a competition between these HSFs during mouse lens development (Fujimoto et al. 2004). HSF4 also binds to the promoter of 
-crystallin, the expression of which is severely decreased in Hsf4-null lens fiber cells, leading to cataract. Inherited cataract in certain Chinese and Danish families has been associated with a missense mutation in the DNA-binding domain of HSF4 (Bu et al. 2002).
HSF2 has been considered an orphan member of the HSF family. HSF2 was found to be abundantly expressed and active for DNA binding in mouse stem cells, embryonic carcinoma cells, and during hemin-mediated differentiation of human K562 erythroleukemia cells (Mezger et al. 1989; Theodorakis et al. 1989; Sistonen et al. 1992, 1994; Murphy et al. 1994). HSF2 does not display significant heat-shock element (HSE)-binding activity in most adult tissues (Fiorenza et al. 1995), but is highly active during mouse embryogenesis. To date, no clear correlation between HSF2 and hsp gene expression has been found during development (Mezger et al. 1994a, b; Murphy et al. 1994; Christians et al. 1997; Rallu et al. 1997; Alastalo et al. 1998; Kallio et al. 2002; Wang et al. 2003). The developmental role of HSF2 has been assessed by several gene inactivation strategies. Although one study did not report any effects in HSF2-deficient mice (McMillan et al. 2002), we and others have reported that HSF2 is important for meiosis in both genders (Kallio et al. 2002; Wang et al. 2003). Hsf2–/– mice also display brain abnormalities characterized by enlarged lateral and third ventricles (Kallio et al. 2002; Wang et al. 2003). The developmental defects of the Hsf2–/– mice do not seem to be due to altered HSP levels, suggesting the importance of other HSF2 target genes.
Mammalian cerebral corticogenesis follows tightly regulated spatial and temporal patterns of neuronal migration. Post-mitotic neurons, generated in the internal ventricular zone (VZ), move along the radial glia guidance fibers toward the external pial surface. The first wave of post-mitotic neurons generates the preplate (PP), into which the next wave of neurons migrates, forming the cortical plate (CP) that consists of an inner subplate (SP) and an outer marginal zone (MZ). Subsequent waves of post-mitotic neurons bypass the SP and any neuronal layers generated earlier, adopting more superficial positions in the CP region formed beneath the MZ. The CP therefore grows in an "inside-out" order, from the innermost layer VI comprising the earliest-born cortical neurons to the outer layer II containing the latest-born neurons (Angevine and Sidman 1961).
Analyses of human disease genes and spontaneous or engineered mutant mice have uncovered several gene products in the regulation of corticogenesis (for reviews, see Gupta et al. 2002; Götz 2003). Reelin is secreted by Cajal-Retzius cells of the MZ, and mutations in the Reelin gene in humans cause lisencephaly with simplified cortical folding (Curran and D’Arcangelo 1998; Gupta et al. 2002). The spontaneous mouse mutant reeler exhibits an inverted organization of the cortical layers and fails to split the PP (Lambert de Rouvroy and Goffinet 1998). Upon Reelin binding to ApoER2/VLDLR receptors, expressed on migrating neurons, the cytoplasmic adaptor Dab1 becomes phosphorylated on tyrosine residues (Hiesberger et al. 1999; Howell et al. 1999; Bock and Herz 2003). Perinatal lethality due to widespread defects in neuronal migration during CNS development is evident in mice deficient for cyclin-dependent kinase 5, Cdk5 (Ohshima et al. 1996). In contrast to the reeler phenotype, the PP splitting is preserved in Cdk5–/– mice, but later-born neurons stack up in inverted layers under the SP (Gilmore et al. 1998; Kwon and Tsai 1998). The neuron-specific activity of Cdk5 is regulated by its activating subunits p35 and p39 (Tsai et al. 1994). Mice lacking p35 are viable and show lamination defects only in the cerebral cortex, while the p35–/–p39–/– mice are lethal perinatally (Chae et al. 1997; Ko et al. 2001). During the development of cortex, p35 is predominant but p35 and p39 are expressed equally at birth (Ohshima et al. 2001). The p35–Cdk5 complex in neuronal migration phosphorylates several substrates involved in cell–cell adhesion and cytoskeletal dynamics (for review, see Gupta et al. 2002).
In this study, the role of HSF2 in cerebral corticogenesis was examined. Using Hsf2–/– mice, we demonstrate that HSF2 is involved in correct neuronal positioning during cortex formation. Immunohistochemical analyses revealed that HSF2 deficiency has a negative impact on cell populations important for radial neuronal migration. The numbers of radial glia fibers, which provide architectural support for migrating neurons, are reduced in Hsf2–/– cortices. Furthermore, the Hsf2–/– MZ is hypocellular and, correspondingly, the Reelin levels are decreased, affecting the Reelin signaling pathway as evidenced by reduced tyrosine phosphorylation of Dab1. The migrating post-mitotic neurons are affected by the lack of HSF2, as seen by a clear reduction of p35 and p39 leading to lowered Cdk5 activity, all central regulators of radial migration. Using chromatin immunoprecipitation (ChIP), we identified p35 as a direct target gene for HSF2. The multiple levels of radial cortical migration that are affected in the absence of HSF2 suggest novel important functions for this protein in the development of the cortex.
|
Results |
|---|
|
TopAbstractResultsDiscussionMaterials and methodsAcknowlegdmentsReferences |
|---|
We previously reported that HSF2-deficient adult brains display structural abnormalities characterized by enlarged ventricles (Kallio et al. 2002). Figure 1A illustrates the variability in the extent of morphological defects in Hsf2–/– postnatal day 19 (P19) brains. In 
80% of the cases, Hsf2–/– brains show severe enlargement of the hemispheres (Fig. 1A, right panel) compared with the Hsf2+/+ brains (Fig. 1A, left panel), while 20% exhibit a moderate enlargement (Fig. 1A, middle panel).
|
-galactosidase (-gal) reporter gene inserted under the control of the endogenous Hsf2 promoter (Kallio et al. 2002). At embryonic day 15.5 (E15.5), -gal activity mainly resided in the proliferative zones of Hsf2–/– and Hsf2+/– developing cortex, similar to earlier stages (from E9.5 to E13.5) (Kallio et al. 2002; Y. Chang, unpubl.)—that is, the VZ and subventricular zone (SVZ)—and persisted in the SVZ at E17.5 in Hsf2+/– as in Hsf2–/– embryos (Fig. 1B; data not shown). However, in contrast to earlier stages, -gal staining started to be visible at E15.5 in the caudal and medial CP (data not shown) and became more intense at E17.5 (Fig. 1B). The expression of HSF2 in CP was verified in Hsf2+/+ E15.5 embryos by immunohistochemistry in combination with a short pulse of bromodeoxyuridine (BrdU) labeling, which delimited the proliferating zone (Fig. 1C). In conclusion, HSF2 expression persists throughout corticogenesis in the proliferative zones but is also observed in the CP at late gestation stages. This suggests that HSF2 might have a role not only in the proliferative precursors but also in post-mitotic neurons of the CP.
Normal PP formation in Hsf2–/– embryos
To investigate the putative role for HSF2 in corticogenesis, the formation of the PP in E13.5 Hsf2–/– mice was examined. We used two different neuronal markers: The Tuj-1 antibody recognizes class III -tubulin expressed in early differentiated neurons (Geisert and Frankfurter 1989), and necdin is expressed in post-mitotic neurons (Niinobe et al. 2000). Since no gross abnormalities were detected in the expression profiles of either Tuj-1 or necdin in Hsf2–/– compared with Hsf2+/+ cortex at E13.5 by immunohistochemistry (Fig. 2A) or Western blot (data not shown), we concluded that PP formation occurs normally during early corticogenesis in Hsf2–/– embryos.
|
PP splitting occurred normally in the cortex of Hsf2–/– embryos, and the SP was visible by histological analysis at E15.5 in Hsf2–/– as well as in Hsf2+/+ (data not shown). In contrast, SP started to disappear at E16.5 in Hsf2–/– cortices. At E17.5, Hsf2–/– embryos displayed SP disorganization of various degrees (Fig. 2B). The histological sections shown in the right panel of Figure 2B represent a severe phenotype with prominent SP disorganization in the developing Hsf2–/– cortex. Even in severely affected mutants, SP was still visible in the rostral part of the telencephalon (Fig. 2B, magnification 100x, right panel, black arrow), whereas the caudal region was totally devoid of a structure corresponding to SP (Fig. 2B, red arrow, magnification 100x). The more moderate phenotype exhibited a disorganized SP (Fig. 2B, middle panel). Interestingly, numerous trailing cells were observed in the Hsf2–/– cortex at the SP level (Fig. 2B, white arrows, magnification 400x).
Immunostaining for calretinin, a marker for pioneering neurons in SP and in the MZ as well as Cajal-Retzius cells in MZ (Fonseca et al. 1995), was dramatically reduced at the SP level in the Hsf2–/– cortex at E18.5 (Fig. 2C). However, weak calretinin immunostaining was still visible in the rostral part of telencephalon near the olfactory bulbs (Fig. 2C, lower right panel). No evidence for massive apoptosis could be detected by a TUNEL assay in the Hsf2–/– SP region at stages E14.5–E17.5 (data not shown). The lack of staining of the SP neurons by calretinin at E18.5 (Fig. 2C) confirms that the SP has already disappeared at E17.5 in Hsf2–/– caudo–medial cortices (Fig. 2B).
The lack of HSF2 affects two cell populations: radial glia cells and Cajal-Retzius cells
Since the disappearance of SP seemed to be associated with the presence of trailing cells in the SP region (Fig. 2B), we suspected that radial migration problems might affect the Hsf2–/– cortices and examined two cell populations involved in this process, radial glia cells and Cajal-Retzius cells. Using the radial glia-specific antibody RC2 (Misson et al. 1988), the Hsf2–/– cortex exhibited a prominent reduction in RC2-positive somae and fibers at E15.5 or E18.5, compared with the Hsf2+/+ cortex (Fig. 3A). Radial glia cells are believed to differentiate into astrocytes at the end of neurogenesis when the neuronal migration is completed (Götz et al. 2002). We labeled E17.5 cortices with the astrocyte-specific marker GFAP (Bignami and Dahl 1974), and Hsf2–/– mutants displayed a significantly lower GFAP labeling than Hsf2+/+ in VZ and SVZ (Fig. 3B). This suggests that, in the absence of HSF2, the number of radial glia cells is specifically reduced in the cortex, which at least partially could account for the reduction in radial glia fibers and fewer astrocytes.
|
). A lower calretinin immunolabeling in MZ of Hsf2–/– embryos was evident as compared with the Hsf2+/+ (Fig. 3D), in agreement with the hypocellularity of the Hsf2–/– MZ. A lower expression of Reelin, which is secreted by Cajal-Retzius cells, was observed in the caudo–medial region of Hsf2–/– cortex at E17.5 (Fig. 3D, lower panel). The reduced Reelin levels were confirmed in Hsf2–/– individual cortices by Western blotting (Fig. 3E, left panel). The extent to which Reelin expression was decreased was in line with the severity of the morphological changes of the dissected mutant brain. The reduction in the number of MZ cells, accompanied by decreased Reelin expression in Hsf2–/– cortex, is consistent with a decrease in Cajal-Retzius cells (Bielle et al. 2005). Next, we investigated the consequences of Reelin reduction on the Reelin signaling pathway by comparing the tyrosine phosphorylation of Dab1 in Hsf2–/– and Hsf2+/+ E17.5 cortices. Decreased levels of Dab1 phosphorylation were observed in mutant cortices (Fig. 3E, right panel), suggesting that the Reelin signaling pathway is negatively affected by the lack of HSF2.
Abnormal positioning of neurons from superficial layers of Hsf2–/– cortices
The effect of HSF2 on the cell populations crucial for radial neuronal migration, radial glia cells, and Cajal-Retzius cells, suggested that radial migration might be affected in Hsf2–/– cortices. To confirm that the initial phases of corticogenesis, PP formation and splitting, were spared in Hsf2–/– cortex (Fig. 2A), we performed BrdU birthdating experiments. The progeny of neural precursors was labeled by BrdU in S phase at E12.5 or E13.5. The positioning of the BrdU-labeled post-mitotic neurons derived from these progenitors at early stages was not affected in Hsf2–/– CPs (data not shown). However, the disappearance of SP and the reduced number of radial glia fibers at E17.5 indicated neuronal positioning problems specifically affecting the most superficial layers. To study the positioning of neurons generated in the late corticogenesis, BrdU was injected into pregnant females at E16.5 and the progeny of the labeled cells was followed 1 mo after birth, after the completion of migration. Hsf2+/+ cortices showed, as expected, that the vast majority of BrdU-labeled cells were found in superficial layers (Fig. 4, left panel). In contrast, although some BrdU-labeled cells were localized in the same layers in Hsf2–/– cortices, a large number of BrdU-positive cells were found deep in the cortex (Fig. 4, right panel), suggesting abnormal positioning of neurons in the Hsf2–/– cortex.
|
We wanted to establish how HSF2 influences the molecular mechanisms regulating neuronal migration. A clue to the nature of these targets was provided by the model system of human K562 cells (see Introduction). K562 human erythroleukemia cells differentiate toward erythroid lineage upon hemin treatment. During this process, HSF2 levels increase and its DNA binding is activated (Sistonen et al. 1992), but HSF2 is not responsible for the expression of the heat-shock genes (Yoshima et al. 1998; P. Östling, unpubl.). To uncover specific HSF2 target genes, we performed cDNA microarray screening with K562 cell lines stably overexpressing HSF2 (Fig. 5A; Leppä et al. 1997; P. Östling, unpubl.). One of the potential target genes identified from this screen was p35, the activator of Cdk5 that is required for cortical lamination (Tsai et al. 1994; Chae et al. 1997). Overexpression of HSF2 was accompanied by a prominent increase in p35 mRNA, which was further enhanced upon hemin treatment (Fig. 5B). Only a minute increase in p35 protein was seen in the hemin-treated cells stably overexpressing HSF2 (Fig. 5C), but it was consistent with elevated levels of the active form of Cdk5 as detected by Tyr 15 phospho-specific antibody (Fig. 5C). The total Cdk5 levels were not markedly altered (Fig. 5C). The increased activity of Cdk5 upon HSF2 overexpression and subsequently elevated p35 expression during hemin treatment was confirmed with a Cdk5 kinase assay (Fig. 5D). Increased phosphorylation of histone H1 was detected in the HSF2-overexpressing cell line as compared with parental K562 cells.
|
Severe defects in cortical lamination are observed in p35-deficient mice, and since the Cdk5–p35 activity has been shown to be crucial for the migration of later-born neurons, we examined their expression in Hsf2–/– cortex. Despite variations between individual embryos in the telencephalic p35 expression, E17.5 Hsf2–/– telencephalon contained reduced p35 protein levels when compared with Hsf2+/+ (Fig. 6A, left panel). No significant changes in Cdk5 levels were observed. However, the active Tyr 15-phosphorylated form of Cdk5 was markedly diminished in individual Hsf2–/– cortices (Fig. 6A, right panel), demonstrating that this signaling cascade is disturbed by the lack of HSF2. Reduced p35 mRNA levels in Hsf2–/– telencephalon were detected by in situ hybridization on E16.5 Hsf2–/– cortices in the caudo–medial region (Fig. 6B, left panels). The expression of p39 was also decreased in the caudo–medial region in E18.5 Hsf2–/– cortices (Fig. 6B, middle panels). In contrast, the Tbr1 mRNA levels, characteristic for the deep layer VI, were not affected (Fig. 6B, right panels). Quantitative RT–PCR confirmed the reduction of p35 mRNAs (wild-type/mutant ratio 1.19 ± 0.04; n = 6; p = 0.00134) and p39 (wild-type/mutant ratio 3.09 ± 0.35; n = 8; p = 0.00055) in Hsf2–/– CPs. Taken together, these results show that unlike in p35–/– mice (Ko et al. 2001), reduced p35 levels were not compensated by p39 in Hsf2–/– neocortices.
|
The results indicated that p35 mRNA is dependent on the amount of HSF2 in HSF2-overexpressing K562 cells and HSF2-deficient cortices, suggesting that p35 could be a novel target gene for HSF2. As HSF2 was previously found to be active mainly in highly proliferative stem cells (Mezger et al. 1989, 1994a, b; Murphy et al. 1994; Christians et al. 1997; Rallu et al. 1997), we verified that HSF2 species present in the CP were active for DNA binding (Fig. 7A). To investigate whether HSF2 regulates the p35 gene directly by binding to its promoter, we performed ChIP analysis of CP extracts at E16.5. A 200-base-pair (bp) segment within the p35 promoter is sufficient for p35 expression in rat primary cortical neurons or neurons derived from P19 EC cells (Ross et al. 2002). This region, well conserved between mouse and human, contains a putative HSE close to the main transcription initiation site and two GC-boxes (Fig. 7B). EMSA analyses of CP extracts at E17.5 and E18.5 revealed binding activity on these three sites (data not shown). When immunoprecipitating chromatin from E16.5 cortical extracts with an HSF2-specific antibody, a prominent binding to this region was observed in Hsf2+/+ cortex (Fig. 7C,D), demonstrating that the p35 promoter is a direct target for HSF2 binding. Furthermore, binding of Sp3 to this region was detected in vivo, which is in agreement with the EMSA analysis performed by Ross et al. (2002). Interestingly, impaired binding of Sp3 was consistently observed in E16.5 cortical extracts from Hsf2–/– mice (Fig. 7B,C).
|
|
Discussion |
|---|
|
TopAbstractResultsDiscussionMaterials and methodsAcknowlegdmentsReferences |
|---|
; Kallio et al. 2002). Here we show that at the end of gestation, HSF2 is expressed and active for DNA binding in the CP from E15.5 until E18.5. This was unexpected, since previously HSF2 had been found to be active only in undifferentiated proliferative cells during mouse development (Mezger et al. 1989, 1994; Murphy et al. 1994; Christians et al. 1997; Rallu et al. 1997). The timing of HSF2 expression prompted us to investigate how HSF2 might influence cortical development.
The PP formation and the appearance of SP at E14.5 and E15.5 exhibited no gross abnormalities in Hsf2–/– cortices. Interestingly, in reeler mutant mice, PP splitting does not occur (Lambert de Rouvroy and Goffinet 1998), whereas PP splitting seems intact in Hsf2–/– cortices despite reduced Reelin levels. However, Magdaleno et al. (2002) have shown that limited amounts of Reelin (10%–20% of the normal endogenous expression) are sufficient to rescue PP splitting in reeler mice. Therefore, the levels observed in Hsf2–/– cortices could be enough to preserve PP splitting. The finding that early corticogenesis is not notably affected in Hsf2–/– is in good accordance with the fact that HSF2 is not expressed in the PP or in the CP until E15.5. At E16.5, SP becomes highly disorganized, starts to disappear, and is undetectable or vestigial at E17.5 in Hsf2–/– embryos. The disruption does not seem to result from marked apoptotic events, but rather the lack of calretinin-positive cells in Hsf2–/– cortices implicates differentiation problems. This defect is mainly observed in the caudo–medial part of the cortex and is less pronounced in the rostral part, near the olfactory bulbs.
Our results demonstrate that two cell populations are affected in the Hsf2–/–neocortices. Immunodetection using RC2 revealed a reduction in the radial glia somae and fibers at E15.5 and E18.5. The reduced number of radial glia cells was surprising, since these cells represent a major pool of neuron progenitors (Götz et al. 2002), and no gross abnormalities in neuron generation were detected by either Tuj1 or necdin antibodies at E13.5 or E17.5. This observation was, however, supported by the fact that Hsf2–/– cortices displayed a similar reduction in the number of GFAP-positive astrocytes, and that astrocytes are derived from radial glia cells at the end of neurogenesis. Future experiments will determine whether the decrease in radial glia cells is due to the lack of HSF2 in this major pool of progenitors in the VZ. One possibility, which is under investigation, would be that other types of neuronal precursors compensate for the reduction of radial cell progenitors to generate cortical neurons. Moreover, MZ appeared hypocellular, and accordingly the expression of calretinin in the MZ was diminished. The expression of Reelin, which is secreted by Cajal-Retzius cells in MZ, was attenuated in Hsf2–/– cortices, suggesting that the lack of HSF2 affects the Cajal-Retzius cell population in the caudo–medial region. This result together with the diminished tyrosine phosphorylation of a downstream effector Dab1 in Hsf2–/– cortices suggests that the Reelin signaling pathway is compromised in our mutants. Reelin signaling directly affects the morphology of radial glia cells, and reeler mutants display a reduction in the extension of radial fibers (Hartfuss et al. 2003). The reduction of radial glia fibers in our mutants might at least partly result from reduced Reelin levels. The reduction of cell populations important for radial neuronal migration, in addition to SP disorganization, prompted us to investigate whether neuronal positioning was disturbed in mutant cortices. BrdU birthdating experiments revealed incorrect positioning of neurons that should have settled in the superficial layers but instead were found in the deep cortical layers. Taken together, our results show that the lack of HSF2 induces incorrect positioning of superficial neurons during cortex formation, suggesting altered migration.
cDNA microarray analysis of K562 cells stably overexpressing HSF2 revealed an increase in p35 mRNA levels, which in turn were down-regulated in Hsf2–/– cortex at E16.5. The minor increase in p35 protein levels in hemin-treated K562 stably overexpressing HSF2 resulted in an increased activity of Cdk5. Therefore, it was conceivable that HSF2 influences Cdk5 activity through induction of p35, which could partly explain incorrect positioning of HSF2-deficient neurons from superficial layers. Indeed, we found that the levels of p35, but also p39 and, consequently, the activity of Cdk5 activity were reduced in Hsf2–/– cortices. However, these alterations do not recapitulate the entire p35–/– phenotype. Although PP splits normally in p35–/– mice, the subsequent waves of neurons are roughly inverted (Chae et al. 1997; Kwon and Tsai 1998). In Hsf2–/– cortices, PP splitting is intact, but in contrast to p35–/–, only superficial layers are affected. This is in agreement with HSF2 expression starting at E15.5 in post-mitotic neurons of the CP. In addition, p35–/– mice show more severe defects in the rostral–medial part of the brain, whereas in Hsf2–/–mice p35 levels are reduced mainly in the caudo–medial portion of the developing cortex, where the Hsf2–/– phenotype is most severe in terms of SP disorganization, calretinin expression, and abnormal neuron positioning. The reduction of p35 due to the lack of HSF2 can occur only in the area of HSF2 expression, which appears in the caudo–medial region but is absent from the rostral area at E15.5 until E17.5 (data not shown).
We identified within the p35 promoter a putative HSE downstream of two GC-boxes, previously shown to be important for expression in cortical primary neurons (Ross et al. 2002). Sp3 and Sp4 are the major transactivators binding to these two GC boxes in cortical primary neurons. ChIP analysis of E16.5 CP revealed that HSF2 binds to this p35 promoter region in vivo, identifying p35 as a direct target gene for HSF2 in brain development. When examining the mouse and human p39 promoters, we could not find conserved HSEs, suggesting that the mechanism of regulation of p39 by HSF2 is different from that of p35. Recently, HSF2 was reported to regulate chromatin structure by recruiting PP2A, which dephosphorylates and inactivates the condensin complex, to prevent compaction of the hsp70 promoter during mitosis (Xing et al. 2005). It will be interesting to determine whether HSF2 has a similar mechanism of action in the context of chromatin structure within the p35 promoter. The identification of p35 as the first HSF2 target gene in brain development helps us in our attempts to understand the mechanism by which HSFs regulate gene expression during development. HSF2 not only plays a role in radial glia and MZ cell populations, but also has effects on the Reelin and Cdk5 signaling cascades that regulate migration of post-mitotic neurons. The Reelin and p35–Cdk5 pathways seem to operate synergistically in the correct positioning of cortical neurons (Ohshima et al. 2001; Beffert et al. 2004). This requires a tight coordination of signaling molecules, to which HSF2 and possibly also other HSFs might contribute by fine-tuning their expression.
|
Materials and methods |
|---|
|
TopAbstractResultsDiscussionMaterials and methodsAcknowlegdmentsReferences |
|---|
Hsf2 heterozygous mice described in Kallio et al. (2002) were maintained in a C57Bl/6 N background (backcross 8). Noon on the day of the vaginal plug was considered as E0.5.
Cell culture and experimental treatments
Human K562 erythroleukemia cells were cultured in RPMI 1640 with 10% fetal calf serum. K562 cells stably overexpressing HSF2- isoform or the empty vector control (Mock) were maintained as parental K562 cells, with neomycin (G418, 500 µg/mL; Leppä et al. 1997). Hemin (Fluka) was used at a final concentration of 40 µM for 16 h.
-galactosidase staining, immunohistochemistry, and histological analysis
–galactosidase activity was detected as described (Kallio et al. 2002). For histology, sections were stained with hematoxylin or toluidine blue. For IHC, embryos were fixed in Bouin’s fixative for 5–12 h, or in PFA overnight at 4°C, embedded in paraffin, and cut into 10-µm sections. Serial sections were incubated with antibodies for 1 h at room temperature or overnight at 4°C: rabbit anti-GFAP 1:100 (Dako), mouse anti-Tuj-1 1:100, rabbit anti-necdin 1:100, rat BrdU antibody 1:100 (Sigma). For immunostaining on cryosections, brains were fixed in 4% PFA-PBS overnight at 4°C, embedded in 0.5% gelatine/30% sucrose solution, and cut into 14-µm sections. Serial sections were incubated overnight at 4°C with mouse anti-RC2 1:10 (IgM, Hybridoma Bank) or anti-calretinin 1:2000 (Swant).
Birthdating experiments
Pregnant females (two to three) were injected at E12.5, E13.5, or E16.5 with 70 µg of BrdU per gram of body weight. E15.5 or P31 brains were fixed overnight in Bouin fixative, paraffin embedded, and cut into 10-µm parasagittal sections.
IP and Western blot analysis
Tissue extracts (TE) and whole-cell extracts (WCE) from K562 cells were prepared as described by Mosser et al. (1988) or Loones et al. (2000). Approximately 20–36 µg of TE or WCE per sample was used for SDS-PAGE. Antibodies: mouse anti-Reelin 1:1000 (clone G10, Calbiochem), mouse monoclonal anti-Dab1 1:400 (H3, gift of A. Goffinet, Universitáe Catholique de Louvain, Brussels, Belgium), mouse monoclonal anti-phosphotyrosine 1:2500 (4G10, Upstate Cell Signaling Solutions), rabbit anti-phosphoserine 1:2500 (Chemicon International), rat anti-Hsc70 1:5000 (SPA-815, Stressgen), rabbit anti-Cdk5 1:200 (C-8, sc-173, Santa Cruz Biotechnology Inc.), rabbit anti-p-Cdk5 1:200 (sc-12918-R, Santa Cruz Biotechnology Inc.), and rabbit anti-p35 1:200 (N-20, Santa Cruz Biotechnology Inc.). Dab1 was immunoprecipitated from 150 µg of total extracts from E17.5 caudo–medial neocortex with 10 µL of Dab1 antibody. Conversely, phosphoproteins were immunoprecitated with 4 µL of anti-phosphotyrosine antibody (4G10) and blotted against anti-Dab1. The pull-down was performed using Bio-Ademtech magnetic beads coupled to protein G according to the manufacturer’s instructions (Ademtech SA).
Northern blot analysis
mRNA was isolated from control or hemin-treated K562 cells with the QuickPrep Micro mRNA purification kit (Amersham Biosciences) and separated on an agarose–formaldehyde gel and hybridized with [
-32P]dCTP-labeled probes: a 775-bp NotI–EcoRI human p35 insert (GenBank accession no. AA442853
[GenBank]
) and a PstI-linearized plasmid for rat GAPDH (pGAPDH) (Fort et al. 1985).
Cdk5 kinase assay
The IP was performed using 500 µg of cell lysate and 5 µL of anti-Cdk5 antibody (clone C-8, sc-173, Santa Cruz Biotechnology Inc.) pulled down with Protein A-coated sepharose beads. The immunocomplexes were washed twice with IP buffer and twice with kinase buffer (see Supplemental Material). The Sepharose A pellets were resuspended in 20 µL of kinase buffer containing histone H1 (100 µg/mL). The kinase assay was performed with 5 µL of 5x ATP mix (1 mCi of [32P] per milliliter and 125 µM cold ATP in kinase buffer) for 20 min at 30°C and stopped by the addition of 3x Laemmli buffer. The radioactive signal was detected with a PhosphorImager. The measured activities were normalized to the endogenous Cdk5 levels as detected by Western blotting. The relative Cdk5 activity of K562 cells was arbitrarily given the value 1.
In situ hybridization
Embryos were fixed in 4% PFA and embedded in 0.5% gelatine/30% sucrose solution and cut into 150-µm sections by a vibratome (Leica). Digoxygenin-labeled riboprobes synthesized from pcDNA3-p35, pcDNA3-p39 (Ko et al. 2001), or pBSE61-Tbr-1 (Hevner et al. 2001) were incubated on sections as described by Wilkinson and Nieto (1993), using digoxygenin alkaline phosphatase-coupled antibodies (Roche) and BM purple AP revelation substrate (Roche).
RT–PCR analysis
RNAs from the caudo–medial region of E16.5 cortices were purified using RNAqueous-Micro kit (Ambion). Reverse transcription was performed from 500 ng of RNAs using SuperScript first strand Synthesis system for RT–PCR (Life Technologies). Quantitative analysis of the LightCycler data were performed using the Quantitect SYBR green PCR kit (Qiagen) and LightCycler analyzing software. Three genes—the highly expressed -tubulin, the moderately expressed cyclophilin B, and PPOX with low expression—were used for normalization of the results (Dauphinot et al. 2005). For p35 and p39 expression analysis, six and eight independent sets (wild-type and mutant couples), respectively, were analyzed. One set of primers for p39 and two different sets of primers for p35 were used (Harada et al. 2001), adapted to murine p35 sequence (Griffin et al. 2004; Ledee et al. 2005). For primer sequences, see Supplemental Material. P value was determined by Student’s t-test.
ChIP
The ChIP protocol was modified from Takahashi et al. (2000). E16.5 cortical tissue was cross-linked with a final concentration of 1% formaldehyde. Quenching was performed with a final concentration of 125 mM glycine. After cortical cell dissociation, three to five embryonic CPs were lysed in 1 mL of lysis buffer. Fragmentation of the chromatin samples was performed by sonication with Bioruptor (Diagenode) to an approximate size of 500 bp. IP was performed after preclearing with a 50% slurry of protein G-coated Sepharose beads containing bovine serum albumin (100 µg/mL, Amersham Biosciences) overnight at 4°C. The following antibodies were used: HSF2 (Sarge et al. 1993), Sp3 (clone D-20, SC-644X Santa Cruz Biotechnology Inc.), and normal rabbit serum (NRS, Jackson Immuno Research Laboratory). Washing of immunocomplexes was performed three times with wash buffer 1, twice with wash buffer 2, and three times in wash buffer 3 (for washing buffers, see Supplemental Material). Cross-links were reversed by incubating the samples overnight at 65°C. DNA was purified, and PCR analysis was performed on 1:10 of each ChIP sample using puRe Taq Ready-to-go PCR Beads (Amersham Biosciences). For primer sequences, see Supplemental Material.
|
Acknowlegdments |
|---|
|
TopAbstractResultsDiscussionMaterials and methodsAcknowlegdmentsReferences |
|---|
|
Footnotes |
|---|
E-MAIL mezger{at}biologie.ens.fr; FAX 33-1-44-32-39-41. 
8 These authors contributed equally to this work.
Supplemental material is available at http://www.genesdev.org.
10 Present address: Institut National de la Santé et de la Recherche Médicale (INSERM) U362, Institut Gustave Roussy, 94805 Villejuif, France.
Article and publication are at http://www.genesdev.org/cgi/doi/10.1101/gad.366906
|
References |
|---|
|
TopAbstractResultsDiscussionMaterials and methodsAcknowlegdmentsReferences |
|---|
Angevine J.B. and Sidman R.L. 1961. Autoradiographic study of cell migration during histogenesis of the cerebral cortex in the mouse. Nature 192: 766–768.[Medline]
Beffert U., Weeber E.J., Morfini G., Ko J., Brady S.T., Tsai L.H., Sweatt J.D., Herz J. 2004. Reelin and cyclin-dependent kinase 5-dependent signals cooperate in regulating neuronal migration and synaptic transmission. J. Neurosci. 24: 1897–1906.
Bielle F., Griveau A., Narboux-Neme N., Vigneau S., Sigrist M., Arber S., Wassef M., Pierani A. 2005. Multiple origins of Cajal-Retzius cells at the borders of the developing pallium. Nat. Neurosci. 8: 1002–1012.[CrossRef][Medline]
Bignami A. and Dahl D. 1974. Astrocyte-specific protein and radial glia in the cerebral cortex of newborn rat. Nature 252: 55–56.[CrossRef][Medline]
Bock H.H. and Herz J. 2003. Reelin activates SRC family tyrosine kinases in neurons. Curr. Biol. 13: 18–26.[CrossRef][Medline]
Bu L., Jin Y., Shi Y., Chu R., Ban A., Eiberg H., Andres L., Jiang H., Zheng G., Qian M. et al. 2002. Mutant DNA-binding domain of HSF4 is associated with autosomal dominant lamellar and Marner cataract. Nat. Genet. 31: 276–278.[CrossRef][Medline]
Chae T., Kwon Y.T., Bronson R., Dikkes P., Li E., Tsai L.H. 1997. Mice lacking p35, a neuronal specific activator of cdk5, display cortical lamination defects, seizures, and adult lethality. Neuron 18: 29–42.[CrossRef][Medline]
Christians E., Michel E., Adenot P., Mezger V., Rallu M., Morange M., Renard J.P. 1997. Evidence for the involvement of mouse heat shock factor 1 in the atypical expression of the HSP70.1 heat shock gene during mouse zygotic genome activation. Mol. Cell. Biol. 17: 778–788.[Abstract]
Curran T. and D’Arcangelo G. 1998. Role of Reelin in the control of brain development. Brain Res. Brain Res. Rev. 26: 285–294.[CrossRef][Medline]
Dauphinot L., Ryle R., Rivals I., Tran Dang M., Moldrich R.X., Golfier G., Ettwiller L., Toyama K., Rossier J., Personnaz L. et al. 2005. The cerebellar transcriptome during postnatal development of the Ts1Cje mouse, a segmental trisomy model for Down syndrome. Hum. Mol. Genet. 14: 373–384.
Fiorenza M.T., Farkas T., Dissing M., Kolding D., Zimarino V. 1995. Complex expression of murine heat shock transcription factors. Nucleic Acids Res. 23: 467–474.
Fonseca M., Del Rio J.A., Martinez A., Gomez S., Soriano E. 1995. Development of calretinin immunoreactivity in the neocortex of the rat. J. Comp. Neurol. 361: 177–192.[CrossRef][Medline]
Fort P., Marty L., Piechaczyk M., El Sabrouty S., Dani C., Jeanteur P., Blanchard J.M. 1985. Various rat adult tissues express only one major mRNA species from the glyceraldehyde-3-phosphate-dehydrogenase multigenic family. Nucleic Acids Res. 13: 1431–1442.
Fujimoto M., Izu H., Seki K., Fukuda K., Nishida T., Yamada S., Kato K., Yonemura S., Inouye S., Nakai A. 2004. HSF4 is required for normal cell growth and differentiation during mouse lens development. EMBO J. 23: 4297–4306.[CrossRef][Medline]
Geisert Jr. , E.E. , Frankfurter A. 1989. The neuronal response to injury as visualized by immunostaining of class III -tubulin in the rat. Neurosci. Lett. 102: 137–141.[CrossRef][Medline]
Gilmore E.C., Ohshima T., Goffinet A.M., Kulkarni A.B., Herrup K. 1998. Cyclin-dependent kinase 5-deficient mice demonstrate novel developmental arrest in cerebral cortex. J. Neurosci. 18: 6370–6377.
Götz M. 2003. Doublecortin finds its place. Nat. Neurosci. 6: 1245–1247.[CrossRef][Medline]
Götz M., Hartfuss E., Malatesta P. 2002. Radial glia cells as neuronal precursors: A new perspective on the correlation of morphology and lineage restriction in the developing cerebral cortex of mice. Brain Res. Bull. 57: 777–788.[CrossRef][Medline]
Griffin S.V., Hiromura K., Pippin J., Petermann A.T., Blonski M.J., Takahashi S., Kulkarni A.B., Shankland S.J. 2004. Cyclin-dependent kinase 5 is a regulator of podocyte differentiation, proliferation, and morphology. Am. J. Pathol. 165: 1175–1185.
Gupta A., Tsai L.H., Wynshaw-Boris A. 2002. Life is a journey: A genetic look at neocortical development. Nat. Rev. Genet. 3: 342–355.[CrossRef][Medline]
Harada T., Morooka T., Ogawa S., Nishida E. 2001. ERK induces p35, a neuron-specific activator of Cdk5, through induction of Egr1. Nat. Cell Biol. 3: 453–459.[CrossRef][Medline]
Hartfuss E., Förster E., Bock H.H., Hack M.A., Leprince P., Luque J.M., Herz J., Frotscher M., Götz M. 2003. Reelin signaling directly affects radial glia morphology and biochemical maturation. Development 130: 4597–4609.
Hevner R.F., Shi L., Justice N., Huesh Y., Sheng M., Smiga S., Bulfone A., Goffinet A.M., Campagnoni A.T., Rubenstein J.L. 2001. Tbr1 regulates differentiation of the preplate and layer 6. Neuron 29: 309–311.[CrossRef][Medline]
Hiesberger T., Trommsdorff M., Howell B.W., Goffinet A., Mumby M.C., Cooper J.A., Herz J. 1999. Direct binding of Reelin to VLDL receptor and ApoE receptor 2 induces tyrosine phosphorylation of disabled-1 and modulates tau phosphorylation. Neuron 24: 481–489.[CrossRef][Medline]
Hilgarth R.S., Murphy L.A., O’Connor C.M., Clark J.A., Park-Sarge O.-K., Sarge K.D. 2004. Identification of Xenopus heat shock transcription factor-2: Conserved role of sumoylation in regulating deoxyribonucleic acid-binding activity of heat shock transcription factor-2 proteins. Cell Stress Chaperones 9: 214–220.[CrossRef][Medline]
Howell B.W., Herrick T.M., Cooper J.A. 1999. Reelin induced tyrosine phosphorylation of disabled 1 during neuronal positioning. Genes & Dev. 13: 643–648.
Jedlicka P., Mortin M.A., Wu C. 1997. Multiple functions of Drosophila heat shock transcription factor in vivo. EMBO J. 16: 2452–2462.[CrossRef][Medline]
Kallio M., Chang Y., Manuel M., Alastalo T.-P., Rallu M., Gitton Y., Pirkkala L., Loones M.-T., Paslaru L., Larney S. et al. 2002. Brain abnormalities, defective meiotic chromosome synapsis and female subfertility in HSF2 null mice. EMBO J. 21: 2591–2601.[CrossRef][Medline]
Ko J., Humbert S., Bronson R.T., Takahashi S., Kularni A.B., Li E., Tsai L.H. 2001. p35 and p39 are essential for cyclin-dependent kinase 5 function during neurodevelopment. J. Neurosci. 21: 6758–6771.
Kwon Y.T. and Tsai L.H. 1998. A novel disruption of cortical development in p35–/– mice distinct from reeler. J. Comp. Neurol. 395: 510–522.[CrossRef][Medline]
de Lambert Rouvroy C. and Goffinet A. 1998. The reeler mouse as a model of brain development. Adv. Anat. Embryol. Cell Biol. 150: 1–106.[Medline]
Ledee D.R., Gao C.Y., Seth R., Fariss R.N., Tripathi B.K., Zelenka P.S. 2005. A specific interaction between muskelin and the cyclin-dependent kinase 5 activator p39 promotes peripheral localization of muskelin. J. Biol. Chem. 280: 21376–21383.
Le Goff P., Le Drean Y., Le Peron C., Le Jossic-Corcos C., Ainouche A., Michel D. 2004. Intracellular trafficking of heat shock factor 2. Exp. Cell Res. 294: 480–493.[CrossRef][Medline]
Leppä S., Pirkkala L., Saarento H., Sarge K.D., Sistonen L. 1997. Overexpression of HSF2- inhibits hemin-induced heat shock gene expression and erythroid differentiation in K562 cells. J. Biol. Chem. 272: 15293–15298.
Loones M.-T., Chang Y., Morange M. 2000. The distribution of heat shock proteins in the nervous system of the unstressed mouse embryo suggests a role in neuronal and non-neuronal differentiation. Cell Stress Chaperones 5: 291–305.[CrossRef][Medline]
Magdaleno S., Keshvara L., Curran T. 2002. Rescue of ataxia and preplate splitting by ectopic expression of Reelin in reeler mice. Neuron 33: 573–586.[CrossRef][Medline]
McMillan D.R., Xiao X., Shao L., Graves K., Benjamin I.J. 1998. Targeted disruption of heat shock transcription factor 1 abolishes thermotolerance and protection against heat-inducible apoptosis. J. Biol. Chem. 273: 7523–7528.
McMillan D.R., Christians E., Forster M., Xiao X., Connell P., Plumier J.C., Zuo X., Richardson J., Morgan S., Benjamin I.J. 2002. Heat shock transcription factor 2 is not essential for embryonic development, fertility, or adult cognitive and psychomotor function in mice. Mol. Cell. Biol. 22: 8005–8014.
