|Home | About | Journals | Submit | Contact Us | Français|
Asymmetric cell division of radial glial progenitors produces neurons while allowing self-renewal; however, little is known about the mechanism that generates asymmetry in daughter cell fate specification. Here we found that mammalian partition defective protein 3 (mPar3), a key cell polarity determinant, exhibits dynamic distribution in radial glial progenitors. While it is enriched at the lateral membrane domain in the ventricular endfeet during interphase, mPar3 becomes dispersed and shows asymmetric localization as cell cycle progresses. Either removal or ectopic expression of mPar3 prevents radial glial progenitors from dividing asymmetrically yet generates different outcomes in daughter cell fate specification. Furthermore, the expression level of mPar3 affects Notch signaling, and manipulations of Notch signaling or Numb expression suppress mPar3 regulation of radial glial cell division and daughter cell fate specification. These results reveal a critical molecular pathway underlying asymmetric cell division of radial glial progenitors in the mammalian neocortex.
Radial glial cells constitute a major population of neural progenitor cells that give rise to neurons in the mammalian embryonic neocortex (Anthony et al., 2004; Malatesta et al., 2000; Miyata et al., 2001; Noctor et al., 2001; Noctor et al., 2004; Tamamaki et al., 2001). The division of radial glial progenitors can be either symmetrical or asymmetrical, which is reflected by the fate of the two daughter cells. Prior to the peak phase of neurogenesis (around embryonic day 13 to 18, E13-E18, in mice), radial glial cells largely divide symmetrically to amplify the progenitor cell population. However, during the peak phase of neurogenesis, they predominantly divide asymmetrically to both self-renew, and to produce either a neuron or an intermediate progenitor cell (IPC) (Chenn and McConnell, 1995; Miyata et al., 2004; Noctor et al., 2004; Noctor et al., 2008; Takahashi et al., 1996). While the neurons migrate radially to form the cortical plate (CP) (i.e. the future neocortex), the IPCs undergo additional symmetric division(s) to generate neurons that ultimately migrate into the CP (Haubensak et al., 2004; Miyata et al., 2004; Noctor et al., 2008). Therefore, asymmetric cell division of radial glial cells accounts for nearly all neurogenesis in the developing mammalian neocortex.
Despite its critical importance, the molecular mechanisms that regulate asymmetric cell division of radial glial progenitors are poorly understood. Extensive studies in Drosophila and C. elegans have revealed that a key feature of asymmetric cell division is the unequal distribution and inheritance of cell fate determinants during mitosis, which critically depends on the establishment of cell polarity in dividing progenitor cells (Buchman and Tsai, 2007; Doe et al., 1998; Fishell and Kriegstein, 2003; Jan and Jan, 2001; Knoblich, 2008; Lechler and Fuchs, 2005; Wodarz and Huttner, 2003). In the Drosophila central nervous system, a neuroblast (i.e. neural progenitor cell) delaminates from the neuroepithelium and divides asymmetrically to produce a large cell, which remains a neuroblast, and a small precursor cell, the ganglion mother cell (GMC). GMCs in turn divide asymmetrically to give rise to neurons and glia. It is well-established that the polarized distribution of cell fate determinants in dividing neuroblasts relies on the proper functioning of a group of proteins that include Bazooka (Drosophila Par3, partition defective protein 3, homolog), Par6, atypical protein kinase C (aPKC), Inscuteable, Partner of Inscuteable (Pins) and Gαi. Of these, Bazooka, Par6 and aPKC together make up a core protein complex – the Par protein complex – that is at the top of a genetic hierarchy for specifying the polarity of neuroblasts and ensuring their asymmetric cell division (Johnson and Wodarz, 2003).
The Par protein complex was initially identified in C. elegans (Kemphues, 2000; Kemphues et al., 1988) and found to be highly conserved across species including mammals (Izumi et al., 1998; Joberty et al., 2000; Johansson et al., 2000; Lin et al., 2000). Recently, the mammalian Par (mPar) protein complex has been implicated in regulating neocortical development (Costa et al., 2008; Manabe et al., 2002); however, it is unclear whether this polarity protein complex regulates asymmetric cell division of radial glial progenitors. Furthermore, Notch signaling activity, a key regulator of neocortical neurogenesis (Gaiano et al., 2000; Li et al., 2003; Petersen et al., 2002; Petersen et al., 2004; Yoon and Gaiano, 2005; Zhong et al., 1996), has been recently suggested to be differentially regulated in radial glial progenitors versus differentiating cells in the developing neocortex (Mizutani et al., 2007; Yoon et al., 2008); yet, how this differential regulation of Notch signaling activity comes about is poorly understood. Here, we set out to determine whether mammalian Par3 (mPar3), a key component of the mPar protein complex (Izumi et al., 1998; Joberty et al., 2000; Johansson et al., 2000; Lin et al., 2000), specifies the polarity of dividing radial glial cells and differentially regulates Notch signaling activity in the two daughter cells, thereby ensuring distinct cell fate specification during the peak phase of neocortical neurogenesis.
To investigate mPar3 function, we first examined the subcellular distribution of mPar3 in radial glial cells during the peak phase of neurogenesis using an antibody raised against mPar3 (Supplementary Fig. 1). As previously reported (Costa et al., 2008; Kosodo et al., 2004; Manabe et al., 2002), mPar3 was concentrated at the luminal surface of the ventricular zone (VZ) (Fig. 1A, left arrow). Interestingly, when examined en face from the ventricle, mPar3 showed a distinct, ring-like expression pattern (Fig. 1A, right). Radial glial cells in the VZ undergo interkinetic nuclear migration as the cell cycle progresses (Sauer, 1935). During interphase, their cell bodies are located away from the luminal surface of the VZ with two processes extending out, the radial glial fiber that reaches the pia and the ventricular endfoot that reaches the luminal surface of the VZ. Evidence from previous studies suggest that the ventricular endfeet of neighboring radial glial cells form junctions at the contacts of the lateral membrane domains, which are manifested in ring-like structures at the luminal surface of the VZ (Aaku-Saraste et al., 1996; Astrom and Webster, 1991; Chenn et al., 1998; Manabe et al., 2002; Rasin et al., 2007; Shoukimas and Hinds, 1978). To test whether mPar3 is localized at the junctions, we examined the co-localization of mPar3 with several junctional proteins including ZO-1 (Fig. 1B), β-catenin (Fig. 1C), and N-cadherin (data not shown). We found that mPar3 was co-localized with ZO-1, but to a much less extent with N-cadherin or β-catenin. These results indicate that mPar3 is enriched at the ZO-1-expressing lateral membrane domain that is apical to the region where β-catenin and N-cadherin are localized in the ventricular endfeet of interphase radial glial cells (Fig. 1D).
We next assessed the subcellular distribution of mPar3 in dividing radial glial cells, which is more critical for regulating daughter cell fate specification. During G2 phase, the cell bodies of mitotically active radial glial cells descend through the VZ; once they arrive at the luminal surface of the VZ, radia glial cells enter mitosis. To precisely locate mPar3 in dividing radial glial cells, we took advantage of the antibody against phosphorylated Vimentin (Ph-Vim), which specifically recognizes dividing progenitor cells in the developing neocortex and reveals their morphology (Kamei et al., 1998; Noctor et al., 2002; Rasin et al., 2007) (Fig. 2A). Interestingly, whereas mPar3 showed preferential localization at ring-like structures in Ph-Vim-negative cells, it exhibited more diffuse distribution in Ph-Vim-positive cells, as evident from the ventricle en face images (Fig. 2B) and the cross-section images (Fig. 2 C and D). These results suggest that mPar3 is delocalized from the ZO-1-expressing junction and becomes dispersed as radial glial cells enter mitosis. This distinct pattern of mPar3 distribution in radial glial cells during interphase and mitosis was also observed when recombinant mPar3 proteins tagged with enhanced green fluorescence protein (EGFP), EGFP-mPar3, was transiently expressed in the developing neocortex (Supplementary Fig. 2). It is worth noting that while the recombinant EGFP-mPar3 protein exhibits overall similar subcellular distribution in radial glial progenitor cells during interphase and mitosis as the endogenous mPar3 protein, it leads to a change in the daughter cell fate specification (see later results), thereby precluding its use in recapitulating the normal segregation of mPar3 in dividing radial glial cells during mitosis.
Does dispersed mPar3 in dividing radial glial cells become polarized as the cell cycle progresses and is it consequently differentially inherited by the two daughter cells? To fully answer these questions, given that the mitotic spindle of dividing radial glial cells rocks in metaphase and the orientation of the cleavage plane is not set until anaphase/telophase (Adams, 1996; Haydar et al., 2003; Sanada and Tsai, 2005), we examined mPar3 distribution selectively in anaphase/telophase radial glial cells. In five E14.5 cortical sections, a total of 51 radial glial cells in anaphase/telophase located at the luminal surface of the VZ were identified (Fig. 2D). In addition to a cytoplasmic/cortical localization, mPar3 was found to be enriched around the cleavage furrow (Fig. 2D, arrows). Interestingly, similar observation has been recently reported for Par3 in dividing neural progenitors in the developing zebrafish neural tube (Tawk et al., 2007). We next quantitatively analyzed the subcellular distribution of mPar3 in the dividing radial glial cells with respect to the cleavage plane (Supplementary Fig. 3). Of these 51 cells, 26 cells (50.9%) were found to possess asymmetrically distributed mPar3, while no clear asymmetry in DNA or Ph-Vim was observed (Fig. 2E and Supplementary Fig. 3). These results suggest that in about half of dividing radial glial cells around E14.5, mPar3 is asymmetrically distributed and in turn differentially inherited by the two daughter cells. This ratio of asymmetric mPar3 distribution and inheritance coincides well with the previously estimated fraction of asymmetric cell division at this developmental stage (Takahashi et al., 1996). Furthermore, we found a similar (50-60%) percentage of radial glial cells with a vertical, an intermediate, or a horizontal cleavage plane that displayed mPar3 asymmetry (Fig. 2F), indicating that the asymmetric segregation and inheritance of mPar3 occurs with equal probability among radial glial cells that divide in different orientations. Moreover, while mPar3 did not show any preference to be localized at either the apical or the basal side of the dividing radial glial cells, the axis of mPar3 asymmetry was always perpendicular to the orientation of the cleavage plane, indicating that mPar3 asymmetry influences the spindle orientation and the cleavage plane in these cells.
Having found that mPar3 shows polarized distribution in dividing radial glial cells (Fig. 2), we asked whether mPar3 is essential for differential daughter cell fate specification and asymmetric cell division. To address this, we perturbed the function of mPar3 in radial glial cells either by depleting or by ectopically expressing mPar3. Seven RNA interference (RNAi) constructs containing short hairpin RNA sequences against mPar3 (mPar3 shRNAs) were developed and tested (Supplementary Fig. 4 and 5). We found that six out of the seven mPar3 shRNAs specifically suppressed endogenous mPar3 expression in radial glial cells (Supplementary Fig. 4). Moreover, when introduced into the developing neocortex, these different mPar3 shRNAs caused similar defects in neocortical neurogenesis (Supplementary Fig. 5A), which could be rescued by co-expression of the shRNA-resistant wild type mPar3 protein (Supplementary Fig. 5D). These results strongly suggest that these mPar3 shRNAs specifically knockdown endogenous mPar3, which regulates neocortical neurogenesis. In addition, we also developed and tested ectopic expression constructs of mPar3 (e.g. EGFP-mPar3 and DsRedexpress-mPar3) (Shi et al., 2003) (data not shown).
To analyze the function of mPar3 in controlling the mode of division of radial glial cells, we adopted an assay that has been widely used for explicitly assessing symmetric versus asymmetric cell division of neural progenitor cells – the clonal pair-cell assay (Li et al., 2003; Sanada and Tsai, 2005; Shen et al., 2002) (Fig. 3A). While a substantial population of radial glial cells expressing the control construct divided asymmetrically as previously described (Li et al., 2003; Sanada and Tsai, 2005; Shen et al., 2002), the fraction of radial glial cells expressing either mPar3 shRNA (Fig. 3B) or EGFP-mPar3 (Fig. 3C) that divided asymmetrically was drastically reduced. Interestingly, even though both manipulations of mPar3 expression augmented symmetric division, the outcome in the daughter cell fate specification was rather different. While shRNA-mediated depletion of mPar3 led to a depletion of mPar3 from both the daughter cells (Supplementary Fig. 4) and an increase in neuron production (Fig. 3B), ectopic expression of mPar3 resulted in the presence of mPar3 in both the daughter cells (Fig. 3C, left) and promoted radial glial progenitor cell fate (Fig 3C, right). These results suggest that mPar3 is essential for asymmetric cell division of radial glial cells and that the inheritance level of mPar3 influences daughter cell fate specification.
We next examined whether mPar3 regulates asymmetric cell division of radial glial cells in situ. To achieve this, we developed an assay for analyzing the mode of division of radial glial cells in brain slices (Fig. 4A). In this assay, individual dividing radial glial cells in slices expressing either control or mPar3 shRNAs were traced with time-lapse imaging. The fate of the two daughter cells originating from the same radial glial cell was then analyzed. Recent studies indicate that during the peak phase of neurogenesis, a majority of radial glial cells in situ divide asymmetrically to generate a radial glial cell and an IPC, which in turn divides symmetrically to produce neurons (Noctor et al., 2008). Consistent with this, we found that most of the dividing radial glial cells expressing control shRNA divided at the luminal surface of the VZ and produced two daughter cells, only one of which expressed the IPC marker Tbr2 (designated Tbr2+ and Tbr2-respectively) (Fig. 4 B and D). Moreover, the Tbr2- daughter cell adopted bipolar cell morphology, suggesting a radial glial cell fate. In contrast, nearly all dividing radial glial cells expressing mPar3 shRNA divided away from the luminal surface of the VZ and generated two daughter cells that were both either Tbr2- or Tbr2+, and adopted multipolar morphology (Fig. 4 C and D), indicating symmetric daughter cell fate specification. These results suggest that suppression of mPar3 expression in radial glial cells in situ causes a switch from asymmetric to symmetric cell division, and further support a critical role for mPar3 in regulating distinct daughter cell fate specification during the peak phase of neurogenesis.
How does mPar3 go about regulating daughter cell fate specification? Given that a role in cell fate determination usually involves transcriptional activity (Ross et al., 2003), mPar3 itself is an unlikely candidate. We therefore sought to identify downstream effectors of mPar3 that regulate daughter cell fate specification. One attractive candidate is the Notch signaling pathway, which plays a key role in neocortical neurogenesis (Chenn and McConnell, 1995; Gaiano et al., 2000; Li et al., 2003; Mizutani et al., 2007; Petersen et al., 2002; Yoon et al., 2008). To test this, we first examined whether mPar3 affects Notch signaling activity in the developing neocortex. We took advantage of a well-characterized Notch signaling activity reporter that contains the canonical Notch effector C-promoter binding factor 1 (CBF1) response element upstream of EGFP (CBFRE-EGFP) (Mizutani et al., 2007). Using this reporter, the endogenous Notch signaling activity can be inferred based on EGFP expression. Should the Notch signaling pathway be a downstream effector of mPar3, either knockdown or ectopic expression of mPar3 would change endogenous Notch signaling activity and in turn, alter the expression of EGFP in the developing neocortex. Indeed, we found that when compared with control, ectopic expression of mPar3 led to an increase in CBFRE-EGFP expression (Fig. 5 A and B), indicative of an enhanced Notch signaling activity. Conversely, expression of mPar3 shRNA resulted in a decrease in CBFRE-EGFP expression (Fig. 5 C and D), indicative of a reduced Notch signaling activity. These results suggest that the expression level of mPar3 regulates endogenous Notch signaling activity. While cells with a high level of mPar3 expression develop high Notch signaling activity, cells with a low level of mPar3 expression harbor low Notch signaling activity. Given that Notch signaling activity controls cell fate specification in the developing neocortex (Gaiano et al., 2000; Yoon and Gaiano, 2005), these results raise the possibility that mPar3 acts through the Notch signaling pathway in generating the asymmetry in radial glial daughter cell fate specification.
To test the possibility that mPar3 acts through the Notch signaling pathway, we examined whether Notch signaling activity is required for mPar3-mediated regulation of radial glial cell division and daughter cell fate specification. It is known that upon activation, the Notch receptor is cleaved to produce the intracellular domain of Notch (NICD), which then enters the nucleus and cooperates with the DNA-binding protein CBF1 and its co-activator Mastermind (MAM) to promote transcription. Based on this, two mutant proteins have been widely used to manipulate Notch signaling activity; NICD for constitutively active Notch signaling and the dominant negative mastermind-like (DN-MAML) for inhibiting Notch signaling (Mizutani et al., 2007; Weng et al., 2003; Yoon et al., 2008). To test the functional relationship between mPar3 and Notch signaling activity, we examined the effect of ectopic mPar3 expression on neocortical neurogenesis in the presence or absence of the Notch signaling inhibitor, DN-MAML (Fig. 6 and and7).7). We found that, while ectopic expression of mPar3 restricted cells to the VZ and promoted a radial glial cell fate (Pax6+) (Fig. 6 A and B), co-expression of DN-MAML suppressed this effect of mPar3 ectopic expression (Fig. 7 A and B). Conversely, we examined the effect of suppression of mPar3 expression on neocortical neurogenesis in the presence or absence of constitutively active Notch signaling NICD (Fig. 6 and and7).7). We found that while depletion of mPar3 causes cells to exit the VZ and to adopt a neuronal fate (TUJ1+) (Fig. 6 C and D), co-expression of NICD largely eliminated this effect of mPar3 depletion on neocortical neurogenesis (Fig. 7 C and D). These results strongly support the notion that the Notch signaling pathway is downstream of mPar3 and is required for mPar3 function in regulating radial glial cell division and daughter cell fate specification.
In Drosophila, Numb is a key player that functions between Par3 (i.e. Bazooka) and Notch signaling in regulating asymmetric cell division of neuroblasts and sensory organ precursors (SOPs) in the nervous system (Knoblich et al., 1995; Rhyu et al., 1994). In mammals, two Numb homologues, Numb (Nb) and Numb-like (Nbl), have been identified (Zhong et al., 1996). Genetic deletion studies demonstrate that Nb/Nbl is essential for the mammalian nervous system development (Li et al., 2003; Petersen et al., 2002; Petersen et al., 2004; Zhong et al., 2000; Zilian et al., 2001). To test whether mammalian Nb and Nbl are required for mPar3 regulation of Notch signaling and radial glial progenitor cell division, we suppressed the expression of Nb and Nbl using specific shRNA sequences against Nb and Nbl as previously published (Rasin et al., 2007) and examined their effect on mPar3 regulation of Notch signaling and neocortical neurogenesis (Fig. 8). Should the regulation of Notch signaling by mPar3 (Fig. 5) depend on Nb and Nbl, we would expect that depletion of Nb and Nbl abolishes the effect of manipulating mPar3 expression on Notch signaling activity. Indeed, we found that in the presence of Nb and Nbl shRNAs removal of mPar3 failed to reduce Notch signaling activation (Fig. 8 A and B), suggesting that Nb and Nbl are required for mPar3 regulation of Notch signaling. Consistent with the notion that Nb and Nbl antagonize Notch signaling, removal of Nb and Nbl led to an increase in Notch signaling activity (Fig. 8 A and B).
We next asked whether Nb and Nbl are required for mPar3 regulation of neocortical neurogenesis. While suppression of mPar3 expression led to a premature depletion of cells from the VZ (Fig. 6 C and D, Fig. 7 C and D, Fig. 8 C and D), depletion of Nb and Nbl largely eliminated this effect (Fig. 8 C and D), suggesting that Nb and Nbl are required for mPar3 regulation of neocortical neurogenesis. Recently, a direct interaction between mPar3 and Nb has been reported (Nishimura and Kaibuchi, 2007). To test whether this mode of action between mPar3 and Nb is critical, we examined the effect of overexpression of the region in mPar3 (amino acids 937-1038) that binds to Nb, mPar3(937-1038), on neocortical neurogenesis (Fig. 8 E and F). Interestingly, we found that overexpression of the Nb-binding region of mPar3 resulted in a premature depletion of cells from the VZ (Fig. 8 E and F), similar to removal of mPar3 (Fig. 6 C and D, Fig. 7 C and D, and Fig. 8 C and D). These results further support a role of Nb/Nbl in mediating the function of mPar3 in regulating Notch signaling and neocortical neurogenesis.
In summary, the results presented here demonstrate that the evolutionarily conserved cell polarity protein mPar3 and the Notch signaling pathway act together to regulate the asymmetric cell division of radial glial progenitor cells in the developing neocortex (Fig. 9). Mammalian Par3 is not statically restricted to the apical membrane domain of radial glial cells as previously suggested (Costa et al., 2008; Kosodo et al., 2004); instead, its distribution is dynamic depending on the cell cycle progression. It is selectively localized to the ZO-1- expressing lateral membrane domain in the ventricular endfeet during interphase and then becomes dispersed during mitosis. This dynamic distribution of mPar3 can lead to asymmetric inheritance of mPar3 by the two daughter cells, which results in differential Notch signaling activation that depends on Numb/Numb-like and distinct daughter cell fate specification. While the daughter cell that inherits a greater amount of mPar3 develops high Notch signaling activity and remains a radial glial cell, the daughter cell that inherits less mPar3 harbors low Notch signaling activity and adopts either a neuronal or an IPC fate.
The dynamic nature of mPar3 subcellular localization in radial glial progenitor cells has not been shown previously. In fact, the distribution of mPar3 in dividing radial glial progenitor cells has not been rigorously examined. A recent study suggests that the mPar protein promotes the proliferation of progenitor cells (Costa et al., 2008). However, it is unclear whether the mPar protein regulates asymmetric radial glial cell division. Precisely determining the subcellular distribution of mPar3 in dividing radial glial cells is of critical importance to understanding its function and the molecular control of asymmetric cell division. Given the enrichment of mPar3 in interphase radial glial cells at the luminal surface of the VZ (Fig. 1), where the cell bodies of scarce dividing radial glial cells are located (Fig. 2A), it is rather challenging to distinguish mPar3 in the cell bodies of dividing radial glial cells from that in the ventricular endfeet of interphase radial glial cells. To overcome this difficulty, we took advantage of the phospho-Vimentin antibody, which selectively labels radial glial cells in mitosis (Fig. 2). Moreover, the cytoplasmic labeling seen with this antibody helps to define the cell contour and its cleavage furrow, thereby facilitating the determination of the precise distribution of mPar3 and the cleavage plane of individual dividing radial glial cells (Fig. 2 and Supplementary Fig. 3). We found that at E14.5 in about half of radial glial cells with a defined cleavage plane (i.e. in anaphase/telophase), mPar3 shows asymmetric distribution and the axis of the mPar3 asymmetry is perpendicular to the cleavage plane; this would result in a preferential segregation of mPar3 into one of the two future daughter cells.
Previous studies showed that about half of the divisions in the VZ of the developing mouse cortex at this developmental stage are asymmetric and neurogenic (Takahashi et al., 1996). Although our analysis of mPar3 asymmetry in dividing radial glial cells is likely an underestimation, these data suggest that the subcellular distribution of mPar3 (i.e. symmetric vs. asymmetric) may be critical for determining the mode of division of radial glial cells. Indeed, we found that disrupting mPar3 asymmetry in radial glial cells either by depletion or by ectopic expression of mPar3 prevents asymmetric cell division and promotes symmetric cell division (Fig. 3 and and4).4). While the precise mechanisms underlying the establishment of the mPar3 asymmetry remain to be uncovered, our findings strongly suggest that mPar3 and its subcellular distribution regulate the mode of radial glial cell division and daughter cell fate specification in the developing neocortex.
Interestingly, while both suppression of mPar3 expression and ectopic mPar3 expression impair asymmetric radial glial cell division, their effects on daughter cell fate specification are rather different. Ectopic mPar3 expression promotes radial glial cell fate (Fig. 3, ,4,4, ,66 and and7),7), whereas suppression of mPar3 expression facilitates neuronal production (Fig. 3, ,4,4, and and66--8).8). These results indicate that the inheritance level of mPar3 influences daughter cell fate specification, although mPar3 itself being an unlikely cell fate determinant. Intriguingly, we found that the expression level of mPar3 affects Notch signaling activity, a key cell fate regulator required for proper neocortical neurogenesis (Gaiano et al., 2000; Yoon and Gaiano, 2005). While a high level of mPar3 expression leads to high Notch signaling activity, a low level of mPar3 expression results in low Notch signaling activity. Previous studies have shown that Notch signaling activity is high in radial glial progenitor cells, but low in differentiating cells such as neurons (Gaiano and Fishell, 2002; Gaiano et al., 2000; Mizutani et al., 2007; Yoon and Gaiano, 2005; Yoon et al., 2008). However, it is unclear how differential regulation of Notch signaling activity is initialized in the daughter cells of dividing radial glial progenitors. Here, we found that asymmetric segregation of mPar3 can lead to differential Notch signaling activity in the two daughter cells.
In Drosophila neuroblasts, the asymmetric localization of Numb, a negative regulator of Notch signaling, is fundamental for differential Notch signaling activity and cell fate diversity in the central nervous system (Jan and Jan, 2001; Roegiers and Jan, 2004; Wodarz and Huttner, 2003). Furthermore, this asymmetry in Numb distribution depends on the asymmetric segregation of Bazooka, the mPar3 ortholog in Drosophila. In mammals there are two Numb homologues, Numb and Numb-like (Zhong et al., 1996). Previous studies suggest that Numb is essential for the proper development of the mammalian brain (Zhong et al., 2000; Zilian et al., 2001). However, the correlation between Numb protein segregation and asymmetric daughter cell fate specification has not been definitively established (Li et al., 2003; Petersen et al., 2002). In addition, recent studies suggest that Numb is involved in trafficking and proper localization of the junctional protein cadherin in radial glial cells and thereby functions in maintaining the tissue architecture of the developing neocortex (Kuo et al., 2006; Rasin et al., 2007). Here, we found that mPar3 acts through Numb and Numb-like in regulating Notch signaling activity. Moreover, our data suggest that a direction interaction between mPar3 and Numb is critical. Despite that it is unclear whether Numb is asymmetrically distributed in dividing radial glial progenitor cells, these findings suggest that asymmetric inheritance of mPar3, which interacts with Numb/Numb-like, results in differential activation of Notch signaling in the two daughter cells of asymmetrically dividing radial glial progenitors in the developing neocortex. Moreover, a recent study showed that removal of Cdc42 in the developing neocortex leads to mislocalization of mPar3 and defects in neocortical neurogenesis (Cappello et al., 2006). Given that mPar3 and activated Cdc42 interact with each other, our findings coupled with these observations suggest that the mPar protein complex and its interacting proteins, such as Cdc42 and Lgl (Klezovitch et al., 2004; Vasioukhin, 2006), likely represent an essential molecular pathway that regulates Notch signaling activity and asymmetric cell division of radial glial progenitor cells in the mammalian neocortex.
Timed pregnant CD-1 mice at the appropriate embryonic age were anesthetized. Embryos were removed and transcardially perfused with ice-cold PBS followed by 4% paraformaldehyde (PFA). Brains were dissected out and coronal sections were prepared using a cryostat or a vibratome (Leica Microsystems). Tissue sections were incubated for one hour at room temperature in a blocking solution (10% normal goat or donkey serum as appropriate, 0.1% Triton X-100, and 0.2% gelatin in PBS), followed by incubation with the primary antibody overnight at 4°C. Sections were then washed in 0.1% Triton X-100 in PBS and incubated with the appropriate fluorescence conjugated secondary antibody for one to two hours at room temperature. For Pax6 and Tbr2 staining, an antigen retrieval procedure was performed before the blocking step by incubating tissue sections in 0.1 M sodium citrate for 5 minutes in a microwave pressure cooker at 50% power. The polyclonal mPar3 antibody was generated by immunizing rabbits with the purified recombinant protein GST-mPar3 (1-229) (Cocalico Biologicals, Inc.) and affinity-purified before use. Primary antibodies used were: Rabbit polyclonal anti-mPar3 (home-made, 1:300), Rabbit polyclonal anti-mPar3 (1:300, Millipore/Upstate), mouse monoclonal anti-β-catenin (1:100, Zymed), mouse monoclonal anti-phosphorylated Vimentin (clone 4A4; MBL, 1:500), mouse monoclonal anti-β-III tubulin (clone TUJ1) (Covance, 1:500), mouse monoclonal anti-Nestin (Millipore/Chemicon, 1:200), rabbit polyclonal anti-MAP2 (Millipore/Chemicon, 1:1000), rabbit polyclonal anti-Pax6 (Covance, 1:100), rabbit polyclonal anti-Tbr2 (kindly provided by Dr. Robert Hevner and Millipore/Chemicon, 1:500), chicken polyclonal anti-GFP (Aves, 1:400), mouse monoclonal ZO-1 (Zymed, 1:250), rabbit monoclonal ZO-1 (Invitrogen, 1:200) and rat anti-prominin-1 (eBioScience, 1:100). Secondary antibodies used were: goat or donkey anti-mouse or anti-rabbit Cy5 or Alexa 568 conjugated antibodies (1:500, Invitrogen/Molecular Probes) and goat anti-chicken Alexa 488 conjugated antibody (1:500, Invitrogen/Molecular Probes). DNA was stained with Propidium Iodide, 4′,6-diamidino-2-phenylindole (DAPI), or Syto-63 (Molecular Probes). Images were acquired with a Zeiss LSM Pascal or an Olympus FV1000 confocal microscope, and analyzed with LSM (Zeiss), Fluoview (Olympus), Volocity (ImproVision), and Photoshop (Adobe Systems).
For quantification of mPar3 immunofluorescence in dividing radial glial cells at the VZ surface, confocal images were taken to cover the entire cell. A contour was drawn around individual dividing cells based on the phospho-Vimentin (Ph-Vim) immunofluorescence in each confocal section, and the cell was then divided along the cleavage plane inferred from the constriction sites corresponding to the mitotic cleavage furrow revealed by the Ph-Vim immunofluorescence and the DNA labeling. Total or average fluorescence intensity for mPar-3, Ph-Vim, or DNA inside the two daughter cell areas over all z-sections was then measured (Supplementary Fig. 3). To quantitatively describe the distribution of mPar3, the normalized ratio of mPar3 immunofluorescence between the two future daughter cells was then calculated as , where “0” indicates perfect symmetry and “1” indicates absolute asymmetry. Similar measurements were also conducted for DNA staining, which served as a control. For quantification of CBFRE-EGFP expression, the EGFP fluorescence intensity in the cell bodies of transfected cells was measured and EGFPhigh cell was defined as the top 10% of the population under control conditions. For quantification of the distribution of EGFP-expressing cells in the developing neocortex, different cytoarchitectural regions of the neocortex (i.e. VZ, SVZ, IZ, and CP) were distinguished based on the cell organization and density reflected by the DNA labeling, and the number of EGFP-expressing cells with their nuclei located in these areas was then counted. Nearly identical areas in the neocortex of individual brains were chosen for analysis.
Data are presented as mean±s.e.m. and nonparametric tests (Mann-Whitney-Wilcoxon test for two groups of data and Kruskal-Wallis test for three or more groups of data) were used for statistical significance estimations.
Seven shRNA sequences against mPar3 were designed as follows: mPar3 shRNA-a (ACAGACTGGTAGCAGTAT), -b (ATGAAAACTACAGAAGCC), -c (CTATGCGTGCGCGTGTCA), -d (CTGAAGATGAGGACGTTG), -e (TCCTACGACAAGCCCATGG), -f (GAGAACCCCAGGTATTCCAG) and -g (GCTGAGCAAGAAAACCTT). Four shRNA sequences against mPar6 were designed as follows: mPar6 shRNA-1 (AGCAAATTTGACGCCGAG), -2 (GTGACTCGAGTGGCCTGGC), -4 (AGCAAGTTTGGAGCTGAG) and -12 (CCAACTGTTCCATCCGTG). The Numb- and Numb-like shRNA sequences were designed as previously described (Rasin et al., 2007). All sense and anti-sense oligos were purchased from Sigma. Annealed oligos were cloned into the HpaI and XhoI sites of the Lentiviral vector pLL3.7, which contains a separate CMV promoter that drives the expression of EGFP (Rubinson et al., 2003). In this study, mPar3 shRNA-a was primarily used after extensive characterization (Supplementary Fig. 3 and 4), while non-specific shRNA constructs or mPar3 shRNA constructs that failed to suppress mPar3 expression were used as control. EGFP-mPar3, CBFRE-GFP, Notch NICD, and DN-MAML constructs were generated as previously described (Mizutani et al., 2007; Shi et al., 2003; Weng et al., 2003). The mPar3(1-712), mPar3(1117-1232) (kinesin-II-binding domain) and mPar3(937-1038) (Numb-binding domain) fragments were generated by PCR and cloned into the EcoRI and XhoI sites of pCAG-IRES-EGFP. All plasmids were confirmed by sequencing.
In utero electroporation was performed as previously described (Tabata and Nakajima, 2001). In brief, a timed pregnant CD-1 mouse at 13 days of gestation (E13) was anesthetized, the uterine horns were exposed, and ~1 μl of plasmid DNA (1-3 μg/μl) mixed with Fast green (Sigma) was microinjected through the uterus into the lateral ventricle manually using a beveled and calibrated glass micropipette (Drummond Scientific). For electroporation, five 50 ms pulses of 40-50 mV with a 950 ms interval were delivered across the uterus with two 9-mm electrode paddles positioned on either side of the head (BTX, ECM830). During the procedure, the embryos were constantly bathed with warm PBS (pH 7.4). After electroporation, the uterus was placed back in the abdominal cavity and the wound was surgically sutured. After surgery, the animal was placed in a 28°C recovery incubator under close monitoring until it recovered and resumed normal activity. All procedures for animal handling and usage were approved by the institutional research animal resource center (RARC).
Pair-cell analysis was performed as described (Shen et al., 2002). In brief, embryos were electroporated with DNA constructs at E13 and the brains were dissected out and sectioned using a vibratome (Leica Microsystems) at E14. Cortical sections containing EGFP-expressing cells in the VZ were isolated, incubated in a protease solution containing 10 units/ml papain (Fluka), 1000 units/ml DNAse I (Roche) and 5 mM L-cysteine in DMEM (Gibco), and triturated using a fire-polished Pasteur pipette to create a single-cell suspension. Cells were resuspended in a culture medium containing DMEM, glutamine, penicillin/streptomycin, sodium pyruvate (Gibco), 1 mM N-acetyl-L-cysteine (Sigma), B27, N2 and 10 ng/ml bFGF2 (Gibco) and plated onto coverslips coated with poly-L-lysine (Sigma) in 24-well plates at clonal density. The cultures were maintained in a humidified incubator at 37 °C with constant 5% CO2 supply. About 24 hours later, the cultures were fixed and immunostained with anti-GFP (1:500, Aves Labs), anti-TUJ1 (1:500, Covance) and anti-Pax6 (1:250, Covance) antibodies, and counterstained with the DNA dye. Images were acquired on an inverted epifluorescence microscope (Axiovert 2000, Carl Zeiss) equipped with a cooled CCD (Orca ER, Hamamatsu Photonics) and analyzed with Axiovision (Zeiss) and Photoshop (Adobe System).
Twelve hours after in utero electroporation, embryos were removed and the brains were extracted into ice-cold artificial cerebro-spinal fluid (ACSF) containing (in mM): 125 NaCl, 5 KCl, 1.25 NaH2PO4, 1 MgSO4, 2 CaCl2, 25 NaHCO3 and 20 glucose; pH 7.4, 310 mOsm/L. Brains were embedded in 4% low-melting agarose in ACSF and sectioned at 400 μm using a vibratome (Leica microsystems). Brain slices that contained EGFP-expressing cells were then transferred onto a slice culture insert (Millicell) in a glass-bottom petri dish (MatTek Corporation) with culture medium containing (by volume): 66% BME, 25% Hanks, 5% FBS, 1% N-2, 1% Penicillin/Streptomycin/Glutamine (all from Gibco) and 0.66% D-(+)-glucose (Sigma). Cultures were maintained in a humidified incubator at 37°C with constant 5% CO2 supply. Twelve hours after plating, petri dishes with slice cultures were transferred to an inverted confocal microscope (LSM 5 Pascal, Carl Zeiss) and time-lapse images of EGFP-expressing cells were acquired every three to four hours for about 24 hours. Transmitted-light images were also taken at each time point to monitor the movement of EGFP-expressing cells in relation to the ventricular surface. Slices were kept in the incubator between time points. After imaging, slices were rinsed once in PBS and fixed in 4% PFA for 24 hours, and processed for immunohistochemistry using an anti-Tbr2 antibody (Millipore/Chemicon). EGFP-expressing cells that divided unambiguously during the imaging period were identified for the analysis of cell division location and mode.
We thank Drs. A. Hall, A.L. Joyner, S.C. Noctor, J. Kaltschmidt, Y. Chin and L.A. McDowell for comments on the manuscript and members of our laboratories for thoughtful discussions and valuable inputs, Dr. N. Gaiano for providing CBFRE-GFP construct, and Dr. Y. Li for providing Notch NICD and DN-MAML constructs. We thank Xin Pei, Michelle Wei, She Chen, Cathy S. Young, William Walantus, Jeanelle Agudelo, and Joy Mirjahangir for technical assistance. This work is supported by grants from March of Dimes Birth Defects Foundation, Whitehall Foundation, DANA Foundation, Autism Speaks Foundation, The Esther A. & Joseph Klingenstein Foundation, NARSAD (to S.-H.S.), and NIH (to S.-H.S. and A.R.K.). L.Y.J and Y.-N.J. are investigators of Howard Hughes Medical Institute.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.