SEARCH

SEARCH BY CITATION

Keywords:

  • numb;
  • notch;
  • neurogenesis;
  • self-renewal;
  • differentiation

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Loss of numb function suggests that numb maintains progenitors in an undifferentiated state. Herein, we demonstrate that numb1 and numb3 are expressed in undifferentiated cortical progenitors, whereas numb2 and numb4 become prominent throughout differentiation. To further assess the role of different numb isoforms in cortical neural development, we first created a Numb-null state with antisense morpholino, followed by the re-expression of specific numb isoforms. The re-expression of numb1 or numb3 resulted in a significant reduction of neural differentiation, correlating with an expansion of the cortical progenitor pool. In contrast, the expression of numb2 or numb4 resulted in a reduction of proliferating progenitors and a corresponding increase in mammalian achete-scute homologue (MASH1) expression, concurrent with the appearance of the mitogen-activated protein-2–positive neurons. Of interest, the effect of numb isoforms on neural differentiation could not be directly related to Notch, because classic canonical Notch signaling assays failed to uncover any differences in the four isoforms to inhibit the Notch downstream events. This finding suggests that numb may have other signaling properties during neuronal differentiation in addition to augmenting notch signal strength. Developmental Dynamics 236:696–705, 2007. © 2007 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The cerebral cortex is formed as a result of precise cell fate decisions, proliferation, and differentiation of multipotential progenitor cells that line the ventricular zone (VZ; Qian et al.,1998). In the mouse cortex, neurogenesis occurs primarily between embryonic day 11 and 17 (E11–E17), using a highly regulated series of intrinsic and extrinsic signals, which establish the proper spatial and temporal generation of the neuronal species (McConnell,1995). Factors that regulate cell fate determination and differentiation in the cortex include members of the Notch signaling pathway and numb isoforms.

The Notch signaling pathway is considered vital to the ontogeny and lineage specification of neural stem cell (NSC) populations (Morrison,2001). In the central nervous system (CNS), Notch uses a two-step process (Grandbarbe et al.,2003) that is required for the maintenance but not the formation of NSCs. Indeed, NSCs have been generated independent of Notch signaling, and embryonic stem cells from RBP-Jκ−/−, Notch1−/−, and Presenilin1−/− mice were quickly depleted in early embryogenesis (Hitoshi et al.,2002). Later in development, Notch signaling is required for astrocytic differentiation of NSC populations (Tanigaki et al.,2001). Beyond its role in stem cell maintenance and astrocytic differentiation, there is evidence that Notch is essential in neuronal maturation (Sestan et al.,1999), learning, and memory (Costa et al.,2005), and may contribute to the etiology of Alzheimer's disease (Chan et al.,2002).

Numb is a critical Notch antagonist and an evolutionarily conserved adapter protein that plays a role in the self-renewal of NSCs and neural differentiation in the CNS (Verdi et al.,1996,1999; Wakamatsu et al.,2000; Zhong et al.,2000; Petersen et al.,2004). Recent studies of Drosophila numb mutants demonstrate hyperproliferation of larval brain tissue that subverts cellular homeostasis (Caussinus and Gonzalez,2005). Loss of numb function in mice results in embryonic lethality at E10.5 and precocious neuronal differentiation at the expense of neural progenitors in the CNS. In addition, the mammalian Numb gene gives rise to at least four alternatively spliced transcripts (Verdi et al.,1999) that produce four protein isoforms, ranging from 65 to 72 kDa. Numb1 or numb3 (i.e., the isoforms containing an insertion in the proline-rich region, PRRL for long) promote the proliferation of neural crest stem cells (NCSCs); whereas ectopic expression of numb2 or numb4 (isoforms without PRR insertions, PRRS for short) lead to an enhanced neuronal differentiation (Verdi et al.,1999). These functions for numb isoforms regulate neurogenesis and neuronal survival in the retina (Dooley et al.,2003) and PC12 cells (Pedersen et al.,2002), respectively. Which numb isoforms are expressed in neural progenitors and whether the numb expression pattern changes during neurogenesis remains to be determined; furthermore, it is essential to understand the mechanism by which numb regulates proliferation and neurogenesis in the cerebral cortex.

Herein, we demonstrate that numb switches from isoforms containing the PRR insert (numb1 and numb3) to those lacking this insert (numb2 and numb4) during cortical development in vivo. This switch in numb isoform expression is replicated in vitro throughout differentiation of the E13 cortical progenitors and P19 cells. The change in isoform expression occurs concurrent with the onset of neuronal differentiation and the induction of mammalian achete-scute homologue (MASH1) expression. Using numb-specific antisense morpholino oligonucleotides, we were successful in knocking down numb expression in cortical progenitors and P19 cells to mimic the Numb-null state. The re-expression of numb2 or numb4 in the numb depleted cortical progenitors resulted in precocious and enhanced MASH1 expression, leading to a higher efficiency of neuronal differentiation. Furthermore, the overexpression of numb4 led to an enhanced neurite growth. In contrast, the re-expression of numb1 or numb3 resulted in an expansion of the progenitor pool. Interestingly, the effect of individual numb isoforms on neural differentiation occurred independent of the Notch pathway, as all isoforms demonstrate similar influence on modulating the Notch signaling events. This study demonstrated that the switch in numb isoforms is a critical step in the differentiation process that does not require changes in Notch activation. It also suggests that numb has multiple functions facilitating neurodevelopment, and underscores the distinct roles of individual numb isoforms in cell fate and differentiation.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

A Switch in Numb Isoform Expression Occurs During Neural Differentiation

The numb proteins are asymmetrically distributed within multipotential neural progenitors to establish distinct progeny upon mitosis (Spana and Doe,1996; Conboy and Rando,2002). However, it still remains to be determined whether neural progenitors express specific numb isoforms as they differentiate and whether this pattern is maintained throughout the differentiation process. To address this issue, mouse cortical lysates were generated from 10 days postcoitus (dpc) embryos through adulthood and subjected to immunoblot analyses. There was a shift in numb protein expression from those isoforms containing the PRR insert (PRRL; numb1 and numb3; 71–72 kDa) in embryonic day 10 (E10) embryos, to isoforms lacking the PRR insert (PRRS; numb2 and numb4; 65–66 kDa) in postnatal day 2 (P2) mice (Fig. 1A, left). numb2 and numb4 remained the predominant isoforms expressed into adulthood. To better understand and potentially manipulate the switch in isoform expression during cortical development, we used two in vitro models of neurogenesis: neural progenitors isolated from the VZ of 13.5 dpc embryos (Mehler et al.,2000) and the embryonal P19 cells (McBurney et al.,1982). A switch in numb isoform expression from the 71- to 72-kDa to the 65- to 66-kDa isoforms was recapitulated in both cell systems, as cells underwent neuronal differentiation in vitro (Fig. 1A, right, and 1B). To further establish a link between the switch in numb isoforms and cell fate in the cortex and P19 cells, we used the well-established neural progenitor marker SOX2 as a marker for undifferentiated neural progeny (Bani-Yaghoub et al.,2006). Both VZ neural progenitors and P19 cells were Sox2-positive when the expression of numb1 and numb3 were prominent. In contrast, differentiated neurons isolated from the cortex of P0 mice and differentiated P19 cells were mitogen-activated protein-2–positive, after the switch to numb2 and numb4 expression (Supplementary Figure S1, which can be viewed at http://www.interscience.wiley.com/jpages/1058-8388/suppmat).

thumbnail image

Figure 1. A switch in numb isoform expression occurs during neural development. A,B: Western analyses of cortical (A) and P19 (B) cells demonstrate a switch in numb expression from the isoforms containing the PRR insert (71–72 kDa) to those lacking this insert (65–66 kDa) throughout differentiation. A, left: Cortical cell lysates were directly isolated from embryonic days 10–17 (E10–E17), postnatal day 2 (PN2) and adult mouse cortices. A, right: Protein lysates were prepared from the E13 cortical progenitors and 7 days in vitro cortical neurons. B: P19 cells were grown for 8 days in vitro (0–8); day 0, undifferentiated; days 1–4, retinoic acid (RA) treatment of cell aggregates; days 5–8, neurons examined. C,D: Reverse transcriptase-polymerase chain reaction analyses confirm that the switch in numb isoform expression occurs at the transcriptional level, as cortical (C) and P19 (D) cells differentiate into neurons.

Download figure to PowerPoint

Because numb2 and numb4 are expressed in differentiating neuronal cells, we also determined which numb isoforms were expressed in the terminally differentiated neurons by isoform-specific reverse transcriptase-polymerase chain reaction (RT-PCR) analysis after 10 days of in vitro differentiation. The numb2 and numb4 transcripts (PRRS) were detectable in the neurons derived from cortical progenitors and P19 cells after 25 cycles of amplification (Fig. 1C,D). However, even after prolonged amplification (40 cycles), only very faint amplification of numb1 and numb3 transcripts (PRRL) was observed in neurons (data not shown). In both systems, the transcripts encoding numb4, which does not contain the 11 amino acid insertion in the phosphotyrosine binding domain (PTBs), were 8–10 times more abundant than the PTBL transcripts, as determined by real-time PCR (2[−ΔΔCT] = 23.403 ± .362) (Livak and Schmittgen,2001). Despite an equal distribution of numb1 and numb3 transcripts in early undifferentiated proliferating progenitors (2[−ΔΔCT] = 20.518 ± .032), there is a clear enrichment specific to numb4 expression in mature neurons.

Specific Numb Isoforms Mediate Discrete Facets of Neural Differentiation

The lethality of Numb-null mice (10 dpc) precedes the switch in numb isoform expression. This problem precluded in vivo assessment of the functions that individual numb protein isoforms serve in cortical development in these studies (Petersen et al.,2002,2004). To circumvent this obstacle and to mimic the null phenotype, antisense morpholino (MO) targeting the 5′-untranslated region of murine numb plus the first eight nucleotides of the coding region were used to knock down the expression of all numb protein species in neural progenitors isolated from the VZ of 13.5 dpc embryos and P19 cells. Then, specific numb isoforms were re-expressed in the numb-depleted neural progenitors cells (Fig. 2A) to assess the role of individual numb isoforms in neural development (Supplementary Figure S2). The addition of a randomized numb morpholino (control MO) had no effect on the differentiation process as compared with progenitors transduced with enhanced green fluorescent protein (EGFP) in terms of total cell number, nestin-positive progenitors, or TUJ1-positive cells after 3 days of treatment (Fig. 2B). Similar results were observed using a control morpholino provided by Gene Tools (data not shown). Because the numb morphlino-treated cells did not achieve the robust cell numbers that EGFP-transduced or control morpholino-treated cells reached, the significance of numb in regulating total cell number, the number of neural progenitors, and the number of cells undergoing differentiation is apparent. To determine the possibility of cell death, random fields of cells treated with numb morpholino (numb knockdown) and cells infected with numb isoforms (numb isoform-specific replacement) were examined for apoptosis. This was attempted by staining with 4′,6-diamidine-2-phenylidole-dihydrochloride (DAPI) for fragmented nuclei and by terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) analysis. Neither assay provided evidence for altered survival in either the numb knockdown or numb replacement groups as compared with controls (data not shown). The re-expression of numb1 in numb morpholino-treated cultures (Fig. 2B, middle panel, dark red bars) resulted in an increase in the percentage of Nestin-positive cells compared with EGFP-transfected (green bars), numb4-expressing (purple bars), and numb-depleted (pink bars) cells (P ≤ 0.05; N = 4) within 24 hr, which was maintained during the 72 hr of examination (Fig. 2B, P ≤ 0.025; N = 4). However, despite a significant rise (P ≤ .025 numb1 vs. GFP, data not shown) in the total cell number that was not maintained for time course of the experiment (Fig. 2B, top panel), there was no corresponding rise the percentage of differentiating neuronal progeny even after 72 hr, as measured by the expression of TUJ1 and MAP2 (Lo et al.,1994; Fig. 2B, bottom panel). Starting at 24 hr, the percentage of numb4-expressing cells that became TUJ1+ neurons always was significantly greater than cells expressing numb1, or EGFP. In addition, after 48 hr of differentiation, the percentage of numb1-expressing cells was significantly lower than EGFP-expressive cells and comparable to the levels obtained in numb-depleted cells. Similar results to numb 1 were observed using the numb3 isoform (data not shown). These results for numb1 and numb3 are consistent with the hypothesis that numb functions to maintain progenitors in a stem cell-like state (Petersen et al.,2002,2004). However, in a sharp opposition to this hypothesis, the re-expression of numb4 (Fig. 2B, solid red bars) or numb2 (data not shown) resulted in (1) significantly fewer nestin-positive cells (only 12% of total infected cells at 72 hr, **P ≤ 0.025; N = 4; Fig. 2C–E), and (2) a marked enhancement of neuronal differentiation (74% of total infected cells at 72 hr; Fig. 2B, red bar, **P ≤ 0.025; Fig. 2F–H) compared with EGFP-expressing or numb-depleted cells. Together, these data suggest that numb regulates the differentiation of cortical progenitors in an isoform-dependent manner.

thumbnail image

Figure 2. Numb isoforms have differing abilities to regulate neuronal differentiation. E14 cortical progenitors or P19 cells were treated with antisense mouse numb morpholino for 72 hr and then infected with individual human numb cDNA isoforms. A: Western blot analysis of numb demonstrates the knockdown efficiency after numb morpholino treatment (lane 3). Note that the cell lysate solely expresses the introduced human numb1-myc after retroviral infection and morpholino treatment (lane 5). B: Numb1 and numb4 demonstrate opposite effects on neurogenesis, as evidenced by the percentage of Nestin-positive neural progenitors or TUJ1-positive neurons over the total number of infected cells. The data are presented as mean ± SEM (N = 4). C–H: Representative images of numb1–enhanced green fluorescent protein (EGFP; C–E) and numb4-EGFP (F–H) cortical cultures show the efficiency of infection and the MAP2-positive neurons generated in each culture after 7 days. Scale bar = 20 μm.

Download figure to PowerPoint

Differences in Numb Actions May Be Independent of Notch Inhibition

There is evidence that mammalian numb, like its Drosophila homologue, antagonizes the Notch-mediated signaling events (Sestan et al.,1999; Wakamatsu et al.,2000). Because Notch regulates many facets of neurodevelopment, one explanation for the above findings is that intrinsic differences exist between the numb isoforms to inhibit the Notch signaling strength. Therefore, using cortical progenitors and P19 cells, Notch transactivation assays were conducted to determine the effect of specific numb isoforms in regulating Notch-mediated signaling. The cultures were maintained at a high density to rule out the possibility of examining the effects of numb in the absence of Notch signaling or at best a very low level of Notch activity. The experiments were also conducted in aggregates to enhance the three-dimension input and effects on notch signaling. As anticipated, numb morpholino treatment resulted in increased endogenous Notch signaling that was more evident upon stimulation with 5 μM Delta-Fc, a soluble form of the Notch ligand (Fig. 3). However, using either Hes1-Luciferase (Sestan et al.,1999; Fig. 3, left graph) or CSL-Luciferase reporters (Small et al.,2001; Fig. 3), there was no difference among specific numb isoforms to inhibit endogenous (Fig. 3, solid bars) or Delta-Fc–activated (Fig. 3, speckled bars) Notch signaling events. This finding was true whether the cells were tested in two-dimensional cultures or from three-dimensional aggregates. These two Notch signaling readouts showed that there is no hierarchy in the ability of numb isoforms to augment the Notch signaling responsible for differences in the biological readouts. Therefore, it stands to reason that any differences in numb protein-specific functions may lie in the ability of specific isoforms to mediate distinct biological endpoints.

thumbnail image

Figure 3. Numb isoforms display no difference in their ability to inhibit Notch. Numb isoforms show insignificant differences to inhibit Notch signaling, using transactivation assays. The numb antisense morpholino (MO) -treated P19 cultures were cotransfected with individual numb isoforms, the CSL-luciferase (right) or HES1-luciferase (left) reporter constructs, and renillia luciferase. Twenty-four hours later, cells were treated with Delta-Fc– (speckled bars) or Fc-conditioned medium for 10 min, lysed, and examined by dual luciferase assay. Data are presented as mean ± SD of three experiments performed in quadruplicate, normalized to the enhanced green fluorescent protein (EGFP) -transfected cultures. *P ≤ 0.05. The horizontal line signifies no change in luciferase expression compared with controls.

Download figure to PowerPoint

Expression of Numb PRRs Isoform Correlates With the Onset of MASH1 Expression

Two downstream targets of Notch signaling are the basic helix–loop–helix (bHLH) proteins, Hes and Hey, which function as transcriptional regulators that antagonize pro-neural genes such as MASH1. Because numb1–4 were almost equally capable of antagonizing Notch signaling, all four isoforms were expected to show increases in MASH1 expression during differentiation, if numb solely functions as a Notch antagonist. However, if numb serves additional functions during differentiation, the differences in MASH1 expression might be observed. Indeed, the percentage of MASH1-expressing cortical cells increased from 21 ± 8% to 66 ± 12% (P ≤ 0.05; N = 3) within 12 hr of numb4-EGFP expression (a total of 24 hr after isolation). More strikingly, MASH1 was detected in the stable numb4-EGFP P19 cells (Supplementary Figure S2) before the initiation of the differentiation process (Fig. 4A,B). This acceleration of a neurogenic program by numb4 overexpression was not restricted to MASH1. Overexpression of numb4 or numb2 also enhanced the levels of Delta1 (Fig. 4A) and TUJ1 (data not shown). What is interesting is that, once the cells were removed from the plate, numb4-expressing cells showed a transient reduction in MASH1 expression. Whether this finding is a result of the aggregation or addition of retinoic acid or even the possibility that neuronal differentiation had concluded and gliogenic differentiation was commencing was not determined. However, in all numb1-EGFP (five of five), numb3-EGFP (four of four), or EGFP (two of two) clones, MASH1 induction was absent before the initiation of differentiation and only appeared 48–72 hr after retinoic acid (RA) was added (Supplementary Fig. 3). Using numb-depleted P19 cells, we then assessed the level and rate of MASH1 appearance as a function of numb isoform expressed. Each of numb2 and numb4 induced MASH1 expression more rapidly and in a higher percentage of cells than numb1 or numb3 (Fig. 4B). Furthermore, the expression of MASH1 in numb1 or numb3 P19 cultures required RA addition and cell aggregation.

thumbnail image

Figure 4. The switch in numb isoform expression is correlated with MASH1 induction. The P19 numb1-4–enhanced green fluorescent protein (EGFP) clones were differentiated as aggregates in the presence of 10−7M retinoic acid (RA; see Supplementary Figure S1). A: Western blot analyses show that MASH1 expression precedes the RA-induced differentiation process in P19 numb4-EGFP cells. Moreover, the level of Delta1 is enhanced in P19 numb4-EGFP cultures. B: MASH1 is increased in the P19 cells stably expressing human numb2-EGFP or numb4-EGFP and treated with antisense murine numb morpholino and RA (data are presented as mean ± SD; N = three independent experiments). The inset shows the effect of morpholino on numb expression.

Download figure to PowerPoint

To further determine a hierarchical role of numb isoforms, numb1 and numb4 were simultaneously re-expressed in the numb morpholino-treated P19 cells. However, little to no MASH1 induction was observed (Fig. 5A). Based on earlier observations, we tested whether the ratio of the two numb isoforms determined the differentiated state of the cell. Adherent P19 cells subjected to 72 hr of numb morpholino treatment were transfected with a fixed amount of numb1 cDNA and increasing amounts of numb4 cDNA. The induction of MASH1 expression was assessed 24 hr later by immunoblotting. Only when the level of the numb4 isoform was in a 2 molar-excess was MASH1 expression induced. The induction of MASH1 was amplified as the ratio of numb4 to numb1 increased. In addition, there was a corresponding decrease in the percentage of cells incorporating bromodeoxyuridine (BrdU), as numb4 levels increased in comparison to numb1 (Fig. 5B). This reduction in the BrdU-positive cells is consistent with our previous reports in which PRRL isoforms increased the proliferation of undifferentiated progenitors in neural crest stem cells (Verdi et al.,1999).

thumbnail image

Figure 5. MASH1 induction is correlated with a change in the ratio of numb isoforms. The ratio of numb4 and numb1 were altered by transfection, and the cells were either lysed for Western blotting or the bromodeoxyuridine (BrdU) cell counts after 24 hr. A,B: The increase in numb4 coincided with an up-regulation of MASH1 (A) and a corresponding decrease in proliferation (B). C: Western blot shows that the switch in numb isoforms still occurs in the P19 cells expressing a dominant-negative MASH1, after cell aggregates are treated with retinoic acid (RA).

Download figure to PowerPoint

Because the transition or forced switch in numb isoforms was closely linked with the induction of MASH1 expression, we could not determine whether the expression of numb2 and numb4 was a causal event in triggering the induction of MASH1 expression and overt neuronal differentiation. To address this issue, a dominant-negative MASH1 construct, coding only for the DNA binding domain of MASH1, was overexpressed in P19 cells that underwent differentiation in the presence of RA. In the absence of functional MASH1, the switch in numb isoforms occurred, suggesting that the switch in numb expression is not contingent upon MASH1 (Fig. 5C). Based on this result, it stands to reason that MASH1 induction is a later or concurrent event of neuronal differentiation and not an inducer of the switch itself.

Numb PRRs Enhances Neurite Growth in Cortical Pyramidal-Like Neurons

Both Notch and numb appear to play important roles in modulating neurite growth (Redmond et al.,2000; Huang et al.,2005). However, no assignment of any numb protein isoform was attributed to these findings. Because numb4 expression continues after the induction of neuronal differentiation, we asked whether numb PRRs isoforms influence the maturation of cortical neurons. Using the morphological features of cortical cells to catalog the number of neuronal subtypes and the length of neurite growth, numb2 or numb4 overexpression resulted in substantially longer neurites (numb2-EGFP, 19 ± 4.3 μm; numb4-EGFP, 23.10 ± 4.1 μm; after 3 days in vitro, N ≥ 50 neurons examined) in pyramidal-like neurons, compared with those of control neurons (EGFP alone, 11.6 ± 3.1 μm, after 3 days in vitro, N ≥ 50 in three independent experiments). Of interest, there were several additional features associated with the effects of numb2 and numb4 on neurite growth. First, the higher rate in neurite growth was maintained longer in culture (Fig. 6A,B, 5 days in vitro; Fig. 6C,D, 10 days in vitro), ruling out delays in neurite growth in control cultures. Second, the enhanced neurite growth in numb2 or numb4 cultures appeared to be cell type specific, as there was no change of shape, size, and process growth in stellate-like cells (Fig. 6E,F). Finally, numb2 or numb4 overexpression led to more pyramidal shaped neurons (numb4-EGFP: 58.9 ± 3% pyramidal-like, 8.05 ± 5% stellate-like) compared with varied neuronal subtypes in control cultures (EGFP alone: 32.8 ± 3% pyramidal-like, 15.5 ± 5% stellate-like). Because this shift in neuronal subtypes was unexpected, we examined similar cultures for cell death to ensure that the presence of fewer stellate-like cells in numb2 or numb4 cultures is not an apoptosis-based phenomenon. Using live Hoechst 33580 staining to detect nuclear fragmentation, there was no significant difference between numb4-EGFP and EGFP cultures (data not shown). Together, these results suggest that the effect of numb on neurite growth (Sestan et al.,1999; Redmond et al.,2000; Shen et al.,2002) occurs in an isoform-dependent manner. Although more sensitive clonogenic assays and differentiation assays are needed, before any clear conclusions can be made.

thumbnail image

Figure 6. Numb4 enhances neurite growth in pyramidal-like neurons. A–D: Neurite growth occurs at a more enhanced rate in the numb4-enhanced green fluorescent protein (EGFP) cortical pyramidal-like neurons (B,D) than in their EGFP counterparts (A,C). A and B are 5 days in vitro; C and D are 10 days in vitro. E,F: In contrast, EGFP (E) and numb4-EGFP (F) appear to have similar effects on the size and shape of cortical stellate-like cells. Scale bar = 20 μm.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Initially, these results appear to be in opposition to the prevailing hypotheses that numb functions to maintain an undifferentiated status. However, they are only in contrast if one simply thinks of the four isoforms as functionally redundant proteins. In fact, our results are in agreement with the hypothesis that numb1 and numb3 potentiate and maintain undifferentiated neural progenitors. However, numb2 and numb4 appear to mediate either the lineage commitment or onset of neural differentiation. Together these results suggest that numb should not be thought of as a single protein but a family of stage-specific proteins with distinct functions in regulating neural development. This distinction may explain the “Numb conundrum” suggested by Yoon and Gaiano (2005). This conundrum arises from reports on Drosophila and in vitro mammalian cells, suggesting that numb inhibits Notch. However, the loss of numb function in vivo appears to lead to Notch inhibition. Two independent groups reported contradictory results concerning the generation of precocious cortical neurons in a Numb-null background (Zilian et al.,2001; Petersen et al.,2004). It is very possible that Numb-null mice produce low levels of transcripts that can be alternatively spliced and that the induction of numb4 occurs sooner and accounts for the appearance of the precocious neurons. It is also possible that numb does not serve a significant role in regulating Notch signaling events. Our current in vitro data argue against this explanation, as basal notch signaling is elevated in numb-depleted cells and based on recent reports that notch signaling levels are critical for stem cell self-renewal (Androutsellis-Theotokis et al.,2006; Guentchev and McKay,2006) and gliogenic differentiation (Furukawa et al.,2000; Morrison et al.,2000; Hitoshi et al.,2002; Grandbarbe et al.,2003). We examined the levels of Bmi and nestin (two markers of stem cells) and GFAP (glial marker) in p19 cells with augmented notch signaling levels 24 hr after numb depletion. Consistent with the observation that notch signal strength drives gliogenic differentiation, the level of GFAP mRNA expression rose precipitously in numb MO cells or cells overexpressing any single numb isoform after 24 hr of delta-Fc exposure. In contrast, cells overexpressing numb1 and 3 showed higher basal levels of BMi mRNA than numb2 cells, numb4 cells, or numb-depleted cells (manuscript in preparation). This finding is consistent with three distinct notions that activated Notch signaling is important in stem cell self-renewal and in a strength-dependent instruction for gliogenic differentiation and that numb has a potent role in augmenting the level of notch signal strength.

The notion that numb inhibits Notch is based on genetic and biochemical studies during the past decade (Campos-Ortega,1996; Guo et al.,1996; Spana and Doe,1996; Verdi et al.,1999; Wakamatsu et al.,2000). However, the results presented here add a new caveat to the discussion. Each of the four numb isoforms examined in this study is equally capable of regulating Notch signaling strength whether one examines the activation of CSL or further downstream at the level of HES transcription. What causes the differences in biological readouts of numb isoforms? One possible explanation is that numb also modulates other signaling transduction pathways that work in concert with Notch to regulate proliferation and differentiation. To this end, Tokumitsu and colleagues showed that serine phosphorylation potentially mediates numb function (Tokumitsu et al.,2005). Their demonstration that serine264 is phosphorylated by Ca2+/Calmodulin-dependent kinase and that phosphorylation events recruit the docking of the 14-3-3 protein isoforms (β, γ, ϵ, θ; Tokumitsu et al.,2005) suggests the possibility that this intriguing scaffold may modulate more than just the Notch-mediated signaling events. Unpublished results from this laboratory identified IQGAP as another phosphoserine-dependent scaffold protein that connects numb to potentially other signaling cascades (J.M.V. and R.C., unpublished data). Furthermore, it is certainly possible that specific combinations of numb isoforms may differentially regulate Notch signaling strength that would not be apparent from these current studies. In support of this possibility, unpublished work from this lab demonstrates that numb isoforms have the ability to form heterodimers and homodimers both in vivo and in vitro (J.M.V. and C.K., unpublished data). It is possible that these dimers either have discrete functions or apply a novel mechanism that regulates Notch antagonism by sequestering stronger or weaker inhibitory numb isoforms. Nevertheless, these studies point to roles for specific numb isoforms outside in addition to inhibiting Notch signaling strength.

Structure–function of Numb Isoforms

A common theme among the popular models of neurogenesis (NCSCs, embryonic cortical progenitors, adult subventricular zone, adult subgranular zone and embryonal cell lines) is that the numb isoforms containing insertions into the PRR domain (PRRL: numb1, and numb3) give way to the isoforms lacking these insertions (PRRs, numb2 and numb4), as neural progenitors initiate the differentiation process. This finding is strong evidence that numb cannot be studied as a single generic protein; rather, the studies of numb function must consider the various isoforms with their specific cellular functions. It further suggests that specific numb isoforms can be used as measures of “stemness” in undifferentiated cell populations, as numb1 and numb3 are expressed mainly in undifferentiated proliferative cells. In contrast, numb2 and numb4 are expressed upon differentiation. Our studies cannot determine whether the appearance of numb2 and numb4 is causal to the differentiation process or if numb1 and numb3 are required for self-renewal. However, a unidirectional pathway in which numb4 expression precedes MASH1 expression is suggested because there is (1) cell-specific and time-dependent expression of numb isoforms, (2) the induction of MASH1, and (3) the evidence suggesting that a switch in numb expression takes place in the absence of functional MASH1.

Although immunochemical analysis of numb in wild-type cells provides information about the morphological distribution of the numb protein, it cannot address the issues related to specific numb isoforms. We resolved some of these issues by showing that cortical and P19 neurons mainly express the PTBsPRRs numb isoform. The lack of the PTB (11 amino acids) and PRR (49 amino acids) inserts (Verdi et al.,1996,1999) removes several potentially important residues from the protein that may be essential for the numb interaction with other molecules, which in turn have important biological implications.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Cell Culture

Cortical cultures were prepared as previously described (Bani-Yaghoub et al.,2006). In brief, dorsal telencephalons or cortices were isolated from time pregnant (10–17 dpc) CD1 mice (Charles River) and neonates (PN0) after CO2 inhalation and cervical dislocation. Cerebral cortices were mechanically dissected in Hanks' balanced salt solution (Invitrogen). A portion of the tissue was prepared for Western blot analysis, and the remainder was used to establish primary cortical progenitor cultures. Cortices were gently triturated and passed through 70-μm nylon cell strainers (BD Biosciences). A total of 5 × 105 cells/well (1 × 105 cells/well for neurite growth analysis) were seeded on the poly-D-Lysine (40 μg/ml) coated coverslips, using six-well plates. Cells were incubated in DMEM + 10% fetal bovine serum (HyClone) + 15 ng/ml basic fibroblast growth factor (R&D Systems) overnight at 37°C, 5% CO2. To generate cortical neurons, cells were incubated in DMEM + 0.5% fetal bovine serum (FBS) + N2 supplement (Invitrogen), starting the next morning. The medium was replenished every 2 days throughout the course of this study.

P19 cells were grown in αMEM (Invitrogen) + 10% FBS as previously described (Jones-Villeneuve et al.,1983; Rudnicki et al.,1989). For differentiation studies, cells were cultured in suspensions in nonadherent bacterial grade dishes (Fischer) in αMEM+ 10% FBS + 10−7M all-trans retinoic acid (Sigma) for 4 days, added at time of aggregation, and refreshed every other day. After 4 days of aggregation in the presence of RA, the cells were plated on adherent tissue culture dishes to complete the differentiation process. Both cortical and P19 cells were stained with different markers to determine the cell identity throughout differentiation. In some experiments, cells were counterstained with Hoechst 33580 (Sigma) to identify viable cells.

Numb Knockdown

Numb-specific antisense morpholino were applied to cells for 72 hr, using Endoporter (Gene-Tools). The 72-hr time point was empirically chosen by generating lysates of morpholino-treated P19 cells at 12-hr intervals and by performing Western analysis for the disappearance of an immunoreactive numb band. The morpholino was refreshed every 24–36 hr to ensure high concentrations of the oligomer. The sequence of the morpholino was from the −7 to +8 of the murine numb cDNA. Standard Control morpholino (Gene-Tools) and randomized numb sequence morpholinos were used as controls. For isoform-specific expression, numb knockdown cells were transduced with retroviral particles containing cDNA coding for individual numb isoforms fused to EGFP. After 24 hr, cells were sorted to enrich for EGFP-positive cells and 2,500 EGFP+ cells were plated for differentiation experiments.

Gene Overexpression

EGFP, numb-EGFP, and dominant-negative MASH1-EGFP were infected into the E13–E14 cortical progenitors and P19 cells, using Lipofectamine 2000 (Gibco) and the AP2 retroviral vectors (Galipeau et al.,1999; Bani-Yaghoub et al.,2006). Two to three stable P19 clones were generated for each numb isoform, using P19 medium supplemented with G418. Neuronal differentiation was analyzed in triplicate for each clone, using standard procedures (Bani-Yaghoub et al.,2006).

Semiquantitative RT-PCR Analysis

Total RNA was isolated, using the RNAaqueous kit (Ambion). First-strand synthesis was performed using SuperScript III (Invitrogen), according to the manufacturer's specifications. Specific primers for mouse numb isoforms, MASH1, Delta1, and glyceraldehydes-3-phosphate dehydrogenase coding regions were used at 94°C for 5 min (1 cycle), 94°C for 30 sec, 55°C for 30 sec (25–40 cycles), and 72°C for 45 sec. The amplified DNA fragments were run on 1.5% agarose gels and visualized by ethidium bromide staining. Semiquantitative RT-PCR was performed to determine the ratio of numb isoform transcripts, according to the previously published methods (Verdi and Anderson,1994).

Quantitative RT-PCR Analysis

Quantitative RT-PCR (Battelli et al.,2005) and quantification was performed according the methods of Livak and Schmittgen (2001). Primers were designed, using standard software.

Immunocytochemical Analysis

Antibodies used in this study included polyclonal numb (Dr. W. Zhong, Yale), Nestin (Dr. R. McKay, NIH), monoclonal MASH1 (Dr. D. Anderson, Caltech), βIII tubulin (TUJ1, Chemicon), microtubule associated protein 2a+2b (MAP2, Sigma), monoclonal and polyclonal rhodamine- and fluorescein-conjugated secondary antibodies (Chemicon). Cells were washed with phosphate buffered saline (PBS) and fixed with 4% paraformaldehyde. Nonspecific binding was blocked in PBS + 10% goat serum (Sigma) and 0.1% Nonidet P-40 for 30 min. Cells were incubated with the primary antibody for 1 hr at room temperature, washed with PBS three times (10 min each), and exposed to the second antibody for 1 hr at room temperature. Cells were washed in PBS again and mounted using Vectashield mounting medium (Vector Labs).

Western Blot Analysis

Cells were washed with PBS and lysed in ice-cold RIPA lysis buffer composed of 150 mM NaCl, 10 mM Tris, 7.2, 0.1% sodium dodecyl sulfate (SDS) 1% Triton X-100, 1% deoxycholate 5 mM ethylenediaminetetraacetic acid plus 1 mM phenylmethylsufonyl fluoride, 10 mM benzamidine, 2 μg/ml leupeptin, 100 μM sodium orthovanadate, 10 mM p-nitrophenylphosphate (inhibitors added from 100× stocks frozen at −80°C). The clarified protein lysates were separated by 8% SDS-polyacrylamide gel electrophoresis under reducing conditions and blotted onto Immobilon-P (Millipore) nitrocellulose. Membranes were blocked with TBST (10 mM Tris buffered saline pH 7.4, 0.1% Tween-20), supplemented with 5% nonfat milk and incubated with primary antibody at 4°C overnight. Membranes were then incubated with horseradish peroxidase–conjugated goat anti-rabbit or anti-mouse IgG for 1 hr at room temperature. After 3 washes in TBST, a chemiluminescence material (Amersham) was applied and the membrane was exposed to an x-ray film.

Luciferase Assay

Stable P19 lines expressing numb1–4 or EGFP alone were transiently cotransfected with 1 μg of renillia luciferase cDNA and 3 μg of either HES (Sestan et al.,1999) or CSL reporter constructs (Small et al.,2003). After transfection, cells were grown in serum-free DMEM, supplemented with 5 μM Delta-Fc– or Fc-conditioned medium for 10 min, and the cells were prepared for luciferase activity using the dual luciferase system (Promega) to normalize for transfection efficiencies. The luminescence of each sample was measured in triplicate by using a FLUOstar OPTIMA luminometer.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The authors thank Drs. D. Anderson (Caltech), R. McKay (NIH), W. Zhong (Yale), R. Mulligan (Harvard), and J. Galipeau (McGill) for providing MASH1, Nestin, and Numb antibodies; HEK 293 GPG; and the AP2 retroviral vector, respectively. J.M.V. was supported by the COBRE in Stem Cell and Regenerative Medicine and from the Medical Research Council and the Alzheimer's Society of Canada. J.R. was supported by an IGERT Predoctoral Training in Functional Genomics of Model Organisms, University of Maine, Orono, Maine.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

The Supplementary Material referred to in this article can be found at http://www.interscience.wiley.com/jpages/1058-8388/suppmat

FilenameFormatSizeDescription
jws-dvdy.21072.img1.tif1563K Supplementary Figure 1 : Sox2 and MAP2 immunohistochemistry demarcate neural progenitors and their neuronal progeny respectively.A)A section of E10 developing cerebral cortex shows the presence of Sox2 positive neural progenitors in the ventricular zone.B)A section of cerebral cortex at birth shows the MAP2 positive neurons. Cortex (Cx), Lateral ventricle (LV), Ventricular zone (VZ).C)The undifferentiated Sox2 positive P19 cells.D)The MAP2 positive P19 neurons after 8 days in vitro. Scale bar:C(15 ?m),D(20 ?m),EandF(10 ?m).
jws-dvdy.21072.img2.tif2116K Supplementary Figure 2 : numb4-EGFP is stably expressed in P19 cells.Using lipofectamine, P19 cells maintain numb4-EGFP expression in undifferentiated adherent cultures(A)and aggregates treated with RA for 4 days(B) .C)The corresponding phase contrast image of B.D-E)P19 neurons are present at days 8(D)and 12(E)in vitro, based on neurite growth. Scale bar: 20 ?m.
jws-dvdy.21072.img3.tif1111K Supplementary Figure 3 : MASH1 immunoreactivity is detected in P19 EGFP clone upon RA induction. A)Undifferentiated P19 cells do not express MASH1 prior to RA treatment.C)MASH1 is expressed in cell aggregates upon RA induction.BandD)The corresponding phase contrast images of A and C, respectively. Scale bar: 30 ?m.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.