Numb Promotes an Increase in Skeletal Muscle ... - Wiley Online Library

TISSUE-SPECIFIC STEM CELLS Numb Promotes an Increase in Skeletal Muscle Progenitor Cells in the Embryonic Somite AURE´LIE JORY,a ISABELLE LE ROUX,a BARBARA GAYRAUD-MOREL,a PIERRE ROCHETEAU,a MICHEL COHEN-TANNOUDJI,b ANA CUMANO,c SHAHRAGIM TAJBAKHSHa Stem Cells & Development, Department of Developmental Biology, bUnite´ de Geńe´tique fonctionelle de la Souris, cUnite´ du De´veloppement des Lymphocytes, Institut Pasteur, Centre National de la Recherche Scientifique (CNRS) URA 2578, Paris, France a

Key Words. Numb • Notch • Myf5 • Pax7 • Pax3 • Somite • Stem cell

ABSTRACT Multiple cell types arise from cells in the dermomyotome of the somite that express Pax3 and Pax7, and myogenesis is regulated by Notch signaling. The asymmetric cell fate determinant Numb is thought to promote differentiation of skeletal muscle and other lineages by negatively regulating Notch signaling. We used transgenesis to overexpress Numb spatiotemporally in Pax31/Pax71 somitic stem and progenitor cells in mouse embryos using a spatiotemporally regulated enhancer element from the Myf5 locus that can target muscle progenitor cells prior to cell commitment. Molecular analyses as well as examination of dermal and skeletal muscle cell fates in vivo show that although Numb

is thought to be associated with muscle differentiation, unexpectedly the common stem/progenitor pool size for these lineages is increased in Numb-transgenic embryos. Prospective isolation of the relevant transgenic cells and analysis by quantitative reverse-transcription polymerase chain reaction demonstrated that, in this context, canonical Notch targets are not significantly downregulated. These findings were corroborated using a Notch reporter mouse during the formation of somites and prior to lineage segregation. Thus, we propose that Numb can regulate the selfrenewal of dermal and muscle progenitors during a lineage progression. STEM CELLS 2009;27:2769–2780

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION During development and regeneration, stem cell self-renewal and differentiation are intricately regulated. Cell diversity can be achieved by asymmetric cell divisions and the selective segregation of cell determinants to one daughter cell [1]. The asymmetric cell fate determinant Numb functions in a wide range of binary cell decisions in Drosophila, yet its role in this process is unclear in vertebrates [1–3]. Although many of the proteins involved in asymmetric cell divisions are conserved between Drosophila and vertebrates, direct evidence demonstrating that these proteins function during mitosis as cell fate determinants in vertebrates is lacking. Numb is membranous and/or cortical and it is also associated with vesicles [4–6]. Asymmetric distribution of Numb was observed in mitotic neuronal progenitors in the mouse forebrain [7], mouse retinal precursors [8], isolated cortical progenitors [9], mouse skeletal muscle satellite cells [10, 11], chick neural tube, [12, 13], dermomyotome (DM) [14, 15], and mammalian hematopoietic stem cells [16, 17]. In Drosophila Numb can repress Notch signaling [18–21]. Interestingly, Numb and Numblike can bind to Notch [7, 12,

20, 22], and they act redundantly during mouse brain development to regulate the balance between progenitor cell maintenance and differentiation [12, 23, 24]. Notch signaling also plays a critical role in regulating myogenic progenitors in vertebrates [25–27]. Downstream targets of Notch include hairy and enhancer of split (Hes) or Hes-related repressor protein also called Hey/Hes/HRT/CHF/ gridlock (Herp) [28], as well as Nrarp [29]. The inhibition of Myod expression by Notch signaling in cell lines occurs via Hes1 [30] and/or Hey1 [31], and mutations in the Notch ligand Delta1 [26] or RBP-Jk [27] result in muscle precursor loss and skeletal muscle hypotrophy. Numb overexpression in vertebrates in vitro was reported to inhibit Notch activity, leading to increased differentiation [10, 32], and these regulators reciprocally negatively regulate one another [33]. However, a direct correlation between the asymmetric segregation of Numb and subsequent cell fate decisions is lacking. In Nb:Nbl double mutants, progenitor cells differentiate prematurely and neuronal progenitors are lost [23], whereas later inactivation of these genes can sometimes result in an overproliferation of neuronal progenitors [24]. Some of these discrepancies may be due to the distinct isoforms of Numb (p65, p66, p71, and p72 [34, 35];

Author contributions: A.J., I.L.R., and B.G.-M.: conception and design, collection and/or assembly of data, manuscript writing, final approval of manuscript; P.R. and A.C.: collection and/or assembly of data, final approval of manuscript; M.C.-T.: final approval of manuscript, provision of reagents; S.T.: conception and design, financial support, collection and/or assembly of data, manuscript writing, final approval of manuscript. Correspondence: Shahragim Tajbakhsh, PhD., Stem Cells & Development, Dept. of Developmental Biology, Institut Pasteur, CNRS URA 2578, 25 rue du Dr. Roux, 75724 Paris Cedex 15, France. Telephone: 33-1-40-61-35-20; Fax: 33-1-45-68-89-63; e-mail: shaht@ pasteur.fr Received February 11, 2009; accepted for publication August 16, 2009; first published online in STEM CELLS EXPRESS SepC AlphaMed Press 1066-5099/2009/$30.00/0 doi: 10.1002/stem.220 tember 25, 2009. V

STEM CELLS 2009;27:2769–2780 www.StemCells.com

The Role of Numb in Cell Fate Regulation

2770

supporting information Fig. 1) or the role of Numb in maintaining epithelial integrity [36]. Although Nb:Nbl double-null mutant embryos are lethal around embryonic day 9 (E9), Notch1 and Delta1 expression was reported to be downregulated in somites of these mutants [37]. These metameric structures generate distinct cell types including dorsal dermis, skeletal and smooth muscle, endothelial cells, and connective tissue [38]. Other studies reported that both Numb [14, 37, 39] and Notch [25-27, 40, 41] regulate cell fate and lineage progression in the somite. The paired/homeodomain genes Pax3 and Pax7, and the myogenic determination factors (MRFs) Myf5, Mrf4, and Myod, play critical roles in regulating skeletal muscle cell fate [38, 42-48]. A pseudostratified epithelium called the dermomyotome contains all of the skeletal muscle stem and muscle progenitor cells (MPCs) for the body [38]. Concomitant cell fate choices occur in the dorsal DM as delaminating cells generate dorsal dermis and the underlying differentiating myotome [49, 50]. Single-cell labeling studies in avians demonstrated that dorsal dermis and MPCs share a common ancestor [51]. Thus, the DM provides a paradigm for investigating how these binary cell fate decisions are executed. The mechanism by which the DM assures its own growth while maintaining a pool of lineage-specific stem/progenitor cells is unknown. We and others have postulated that asymmetric cell divisions promote self-renewal as well as cell diversity in the DM [14, 15, 49]. Notably, DM growth and differentiation correlate with shifts in the plane of cell divisions [51]. These observations suggest that asymmetric divisions regulate cell fate decisions in the DM. The extent to which Numb and Notch intervene in this process remains unresolved. Several studies point to a role for Numb in skeletal myogenesis: (1) it defines muscle identity in Drosophila muscle precursors [52, 53]; (2) Numb segregates asymmetrically in dividing adult mouse muscle satellite cells [10, 11]; and (3) overexpression of avian Numb in the ventral DM localizes these cells to the myotome [39]. However, the role of Numb in regulating cell fate and differentiation remains controversial [10, 11, 23, 24, 32, 33]. The misexpression of cell fate regulators such as Numb has provided important insights into their role during cell commitment [12, 35-37, 54-56]. Moreover, overexpression studies in the avian ventral DM showed that bone morphogenic protein (BMP) and Notch signaling differentially regulate endothelial, mural, and striated muscle lineages [57], whereas Numb directs DM cells to the myotome [39]. The latter observation is in keeping with some studies in postnatal satellite cells [10, 32, 33], and the central nervous system [12, 24]. However, other studies reported a role for Numb in progenitor cells [11, 23, 58]. To resolve these discrepancies, we used a transgenic approach to misexpress Numb spatiotemporally in vivo when cell fate decisions are executed. We provide evidence that Numb overexpression does not result in the downregulation of Notch activity but, surprisingly, results in upregulation of several targets, leading us to propose that Numb can modulate self-renewal and commitment in a context-dependent manner during a lineage progression.

MATERIALS

AND

METHODS

Generation of Constructs for Transgenic Mice The Numb-GFP fusion protein was reported previously and its subcellular expression pattern was validated in primary muscle cells and several cell lines [11] (data not shown). The 3XEpEGFP-IHR, 3XEpE-NbGFP-IHR, and 6XEpE-NbGFP constructs

were created by subcloning three or six tandem copies of the extended 900-base pair epaxial enhancer (EpE) [59, 60] in a pSK vector containing either the minimal Myf5 promotor and 50 untranslated region (UTR) of Myf5, or the minimal TK promotor and 50 UTR of Myf5. The different EpE/promoter combinations were then subcloned into pSK vectors containing the EGFPIRES-H2BmRFP-SV40pA, Numbp66EGFP-IRES-H2BmRFPSV40pA, or Numbp66EGFP-SV40pA fragments, respectively.

Immunohistochemistry and Imaging Embryos were harvested in 1 phosphate-buffered saline (PBS) and screened for high or low levels of live NbGFP or GFP expression under a Leica fluorescent stereomicroscope (Heerbrugg, Switzerland, http://www.leica.com). Embryos were fixed overnight in 0.5% paraformaldehyde at 4 C, and washed in PBS. Immunohistochemistry (IHC) on cryosections was carried out as described [43]. For whole-mount embryo IHC, the fixed embryos were permeabilized and blocked in 10% heat-inactivated filtered goat serum/3% bovine serum albumin/0.5% Triton X-100 and incubated in primary and secondary antibodies overnight in blocking solution. Antibody references and dilutions are described in supporting information Table 1. Images were acquired with a Zeiss Axioplan equipped with an Apotome and Axiovision software (Carl Zeiss, Jena, Germany, http://www.zeiss.com), or a Leica SPE confocal and Leica Application Suite (LAS) software. All images were assembled in Adobe Photoshop and Indesign (Adobe Systems, San Jose, CA, http://www.adobe.com). Some images were assembled as projections of successive confocal acquisitions. For quantification of cells in the DM, confocal sections were acquired with z-steps of 5 lm (for bromodeoxyuridine [BrdU] staining) and 8 lm (for Pax3 staining). Each cell population within the dermomyotome and myotome was counted on every section.

BrdU Pulse, Immunohistochemistry, and In Situ Hybridization Pregnant females were injected with a fresh solution of 10 mg BrdU/ml of saline (50 mg BrdU/kg). After a 2-hour BrdU pulse, embryos were harvested and fixed as indicated above. Fixed embryos were incubated for 1 hour at 37 C in 0.2 N HCl and neutralized several times by washing in 0.1 M borate, pH 8.5 (borax), for 15 minutes, then in 1 PBS for 30 minutes, and processed for IHC as indicated above. Embryos were incubated with primary anti-BrdU antibody (Megabase, Lincoln, NE; 1/800 for 48 hours). In situ hybridization was carried out as described in Tajbakhsh et al. [48]. The Alx4 in situ hybridization (ISH) probe is from F. Meijling, the Mrf4 probe is described in Kassar-Duchossoy et al. [42], and the mNumb ISH probe was generated from the mNumbp66 cDNA and it recognizes all isoforms of Numb.

Cell Preparations for Western Blot and Fluorescence-Activated Cell Sorting Interlimb somites of stage-matched embryos were dissected, lysed for Western blot analysis or trypsin treated, and isolated by fluorescence-activated cell sorting (FACS) using green fluorescent protein (GFP) gate. FACS was done with the total GFP population or the subfractioned GFPhigh population and the GFPlow population (10% and 60% of the total GFP population, respectively; supporting information Fig. 2). Sorted cells were collected directly in RNA lysis buffer (Qiagen RNAeasy Micro purification Kit; Hilden, Germany, http://www1.qiagen.com) or Western blot lysis buffer and processed for RNA extraction or Western blot, respectively.

RNA extraction, Reverse Transcription, and Quantitative Polymerase Chain Reaction Total RNA was extracted from cells isolated by FACS on GFP or NbGFP positivity using the Qiagen RNAeasy Micro purification

Jory, Le Roux, Gayraud-Morel et al.

2771

Kit. DNase-treated (400-600 ng; Roche, Mannheim, Germany, http://www.roche.com) RNA was processed for random-primed reverse transcription using the SuperScript II reverse transcriptase protocol of Invitrogen, (Carlsbad, CA, http://www.invitrogen.com). The cDNAs were then analyzed by real-time polymerase chain reaction (PCR) using powerSYBR Green Universal Mix or Taqman universal Master Mix and an ABI Prism 7700 (Applied Biosystems) and a StepOnePlus (Applied Biosystems). Glyceraldehyde-3-phosphate dehydrogenase reference transcript levels were used for the normalization of each target within each sample (¼ D cycle threshold [CT]). Target expression (DCT) of each NbGFP sample was normalized to each control GFP sample and averaged between at least three biological replicates for each condition using the D(DCT) method [61]. All primer sequences and references are listed in the supporting information Table 1. Custom primers were designed using the Primer3Plus online software (http://www.bioinformatics.nl/cgibin/primer3plus/primer3plus.cgi). Serial dilutions of total E10.75 wild-type (WT) embryo cDNA were used to calculate the amplification efficiency of each primer set according to the equation: E ¼ 10 1/slope. Primer dissociation experiments were performed to assure that no primer dimers or false amplicons would interfere with the results.

Statistical Relevance of Observed Differences All described differences in whole-mount ISH signal, IHC proteins levels, or cell counts were observed from pools of n 3 Tg:3XEpE-GFP-IHR or WT control embryos versus n 3 Tg:6X::3XEpE-NbGFP(IHR) embryos of somite stage-matched embryos identically treated on the same day and image processed in identical conditions.

RESULTS Asymmetric Distribution of Numb During Mitosis in Mouse Somites At E10.75, Numb mRNA is ubiquitously expressed in the embryo, with higher levels observed in numerous locations including the somites (supporting information Fig. 3A). Different antibodies specific for all Numb isoforms, but not recognizing Numblike [11] were used to detect Numb protein in somites (see supporting information Fig. 1A for antibody peptide sequence). As in the chick [14, 15], high expression of Numb was observed in myotomal cells, whereas lower levels were observed in DM cells. Numb was expressed with a predominant cortical/vesicular localization, and it appeared to preferentially accumulate on the basolateral/apical portion of these cells in interphase (supporting information Fig. 3D, 1E; data not shown). It is possible that this is due to staining in ‘‘endfeet’’ of more basally located cells, as was described in the central nervous system [36]. Strikingly, we observed that in most of the mitotic cells of the dorsal DM (91%; n ¼ 363 mitoses, n ¼ 3 embryos; supporting information Fig. 3H) Numb was distributed asymmetrically on the cortical membrane as a crescent; this crescent was oriented toward the myotome (Fig. 1A–1R; supporting information Fig. 3D–G). A Numb crescent was also observed in mitotic cells in the neural epithelium (supporting information Fig. 3B, 3B’). This distribution of Numb in mitotic cells is reminiscent of that seen in sensory organ precursor (SOP) and neuroblast divisions in Drosophila [62]. By analogy, Numb could be inherited according to the plane of cell division, preferentially to only one daughter cell. A double labeling of Numb at anaphase with the mitotic kinase AuroraB, which stains the mitotic kinetochore spindles and midbody, confirms that cortical Numb may be inherited either by the two daughter cells (Fig. 1I– www.StemCells.com

Figure 1. Distribution of Numb in the mouse dorsal somite. (A-R): Immunostaining on WT embryos for Numb, for the myotomal marker M-cadherin, and for c-tubulin to highlight the centrosomes (A–H), and for Numb and AuroraB to highlight the kinetochore spindles and midbodies (I–R). The nuclei are labeled by a Hoechst staining (A–R). Lateral confocal sections of E9.5 (A–D) whole-mount embryos (apical side of DM, dotted line). (C, D) and (G, H): High magnifications of the boxed mitotic cells in (B) and (F), respectively, show the apical cortical Numb crescent in dividing cells. (E–R): Transverse sections of E10.5 embryos showing mitosis in the dorsal DM. (M) and (R): Schematic representations of (L) and (Q), respectively. Scale bars ¼ 25 lm (A, B, E, F) 10 lm (I–L, N–Q). Abbreviations: DM, dermomyotome; E, embryonic day; Mcad, M-cadherin; cTub, c-tubulin; WT, wild type.

1M) or by only one daughter cell (Fig. 1N–1R). Attempts to quantify the number of cells that segregate Numb during anaphase were not successful at this stage due to the multiple planes of cell division in the epaxial DM, the limited number of anaphases/telophases, and the transient expression of Numb crescent (loss during late anaphase). Since different isoforms of Numb were reported to perform distinct functions, we examined their expression by reverse-transcription (RT)-PCR and Western blot analysis. Only the phosphotyrosine binding i (PTBi) isoforms (p66, p72) were present in somites, with Nbp66 at higher levels compared with Nbp72 (supporting information Fig. 1B–1D). Nbp66, but not Nbp72, appears to be

2772

the predominant isoform in primary adult myoblasts (BGM and ST, unpublished observations). Thus, we chose Nbp66 (hereafter called Nb) for further studies.

Spatiotemporal Targeting of Transgenic NbGFP in the Dorsal Dermomyotome To assess the role of Numb in MPCs, we overexpressed this protein in stem/progenitor cells prior to the commitment of the DM cells to a dermal or myogenic cell fate. Notably, overexpression studies have been critical in determining if cell fate outcomes can be correlated with changes in the types of cell divisions. This was done extensively with Numb in Drosophila [63, 64]. To do this, we generated transgenic mouse lines that stably overexpressed a Numb-GFP fusion protein (NbGFP) [11], or a control GFP construct in the dorsal DM using the extended epaxial enhancer of Myf5 (EpE). This enhancer is necessary and sufficient to drive reporter gene expression in the dorsal (epaxial) domain of the developing somite where Myf5 is expressed initially [65, 66]. Control (Tg:3XEpE-GFP-IHR) and NbGFP transgenes (Tg:6XEpENbGFP, Tg:3XEpE-NbGFP-IHR) transgenes exhibited robust expression in Pax3þ/Pax7þ cells in the DM before cell commitment (Fig. 2A; supporting information Fig. 4G, 4J). This expression in dorsal somites occurred between E9 and E11.75 when the DM undergoes a delamination, and most cell fate decisions are being executed. Accordingly, and in support of the strategy to target DM cells prior to commitment, only a subset of the DM cells that overexpressed NbGFP also expressed Myf5 protein (supporting information Fig. 4C–4F). Importantly, NbGFP expression declines during later stages of myogenic lineage progression, and in a rostrocaudal development gradient as expected for the EpE enhancer [66] (supporting information Figures 4A, 4B and 2A, 2B). Therefore, these transgenes allowed us to spatiotemporally target expression within the dorsal DM when MPCs are initiating cell commitment. In control and NbGFP transgenics, virtually all GFPþ and NbGFPþ cells in the dorsal DM were Pax3þ by E10.5 (supporting information Fig. 4H–4L). Whole-mount in situ hybridization using antisense Numb showed that the transgenic NbGFP transcripts (n ¼ 3 embryos) were detected well before endogenous Numb transcripts (n ¼ 3 embryos) (Fig. 2D, 2E), and as expected, they were restricted to the dorsal half of the somites (Fig. 2D). The level of NbGFP overexpression was quantitated by quantitative RT-PCR (qRT-PCR) after isolation of cells by FACS from interlimb somites of E10.5 embryos into high and low GFPþ cells (supporting information Fig. 2A–2D). Consistent with the GFP protein expression levels, 2.9- and 14.6-fold higher transcripts were observed compared with the endogenous Numb in GFPlow and GFPhigh expressing cells, respectively (Fig. 2B, 2C). In keeping with these findings, Western blotting of DM cells isolated by FACS from NbGFP transgenics showed an 3 overexpression compared with endogenous Numb protein (data not shown). The subcellular localization of NbGFP was predominantly confined to the cell membrane and cortex in interphase and mitotic cells as previously reported in adult satellite cells [11] (Figure 2H, 2K, supporting information Figure 4G–4J; data not shown), whereas the control GFP was cytoplasmic (supporting information Fig. 4D, 4K, 4M; data not shown). Transgenic NbGFP overexpression resulted in the uniform membranous and submembranous distribution of NbGFP in interphase cells (Fig. 2F–2K). In mitotic cells, homogenous distribution of Nb (endogenous Nb þ NbGFP) was observed in 93% and 58% of NbGFPhigh (n ¼ 208 mitosis, n ¼ 3 embryos) and NbGFPlow (n ¼ 209 mitosis, n ¼ 3 embryos)


embryos, respectively (Fig. 2L; supporting information Fig. 2). Notably, NbGFPlow embryos were informative since the levels of NbGFP were variable. Here, mitotic cells that maintained the endogenous Numb crescent correlated with low levels of NbGFP (Fig. 2J, white arrowheads), whereas within the same confocal field, mitotic cells with higher levels of NbGFP (Fig. 2G, 2H, 2J, 2K, white arrows) lost this Numb crescent. In NbGFPhigh embryos, most of the cells had levels of NbGFP expression leading to 93% of mitotic cells with a uniform expression of Numb expression. This observation was corroborated by videomicroscopy of somite cultures expressing NbGFP, where the distribution of this protein was noted in both daughter cells as mitosis terminated (Fig. 2M– 2T; supporting information Movie 1). Therefore, high levels of NbGFP expression result in a reproducible and uniform inheritance of Numb in the two daughter cells (Fig. 2F–2K, arrow). This imbalanced inheritance pattern with respect to the wild-type may provoke phenotypic changes.

Overexpression of NbGFP Concomitantly Affects Dorsal Dermis and Myogenic Cell Fates The effects of NbGFP overexpression were assessed by examining dermal versus skeletal muscle cell fate decisions given that these cell lineages share a common ancestor (Fig. 3L) [51]. Desmin was used as a marker to identify myoblasts and postmitotic differentiated cells in the myotome. Interestingly, during the peak of activity of the EpE (E10; interlimb, 3035 somites), NbGFP overexpression in the dorsal DM resulted in a transient increase in Desmin protein levels in the epaxial myotome by whole-mount immunostainings (Fig. 3A–3D). This observation was confirmed by Western blot analysis of somites from stage-matched embryos (Fig. 3M, left panel). In more mature anterior somites, this increased expression of Desmin was less discernable due to the growth of the myotome resulting from myoblasts arising from other regions of the DM, as well as the downregulation of the EpE (Fig. 3M, right panel). One interpretation of this phenotype is that NbGFP promotes MPC commitment and differentiation. If this were true, we would expect that the shift of DM cells toward the myogenic lineage would impact on the fate of other progenitors arising from the dorsal DM. Expression of the homeodomaincontaining gene Alx4, which labels cells of the dorsal dermis [67], was assessed by whole-mount ISH. Unexpectedly, overexpression of NbGFP did not perturb the early Alx4 ISH pattern or its level of expression at E10.5 (Fig. 3E, 3F). However, determination of dorsal dermal cells occurs slightly later than that of MPCs. Indeed at E10.5, myogenic determination had already initiated in caudal somites (Fig. 3A), whereas determination of the dorsal dermis, as measured by Alx4 expression, barely commenced in interlimb somites at this stage (Fig. 3E). Interestingly, by E11.5 the Alx4þ dorsal dermis territory was expanded in NbGFP-transgenic embryos, in both dorsal and ventral directions compared with the control (Fig. 3G, 3H, arrows, asterisk). This phenotype likely arises early before the dermal progenitors left the DM since transgene expression was strongly downregulated at E11.5. Therefore, in the dorsal dermis and the myotome, it appears that an increase in the number of committed cells was observed in the NbGFP-transgenic embryos. Our observations above indicate that NbGFP overexpression yields more differentiated muscle cells. Given that myogenic lineage progression occurs within a continuum at this stage with stem and committed cells present simultaneously, we investigated the onset of myogenesis. Surprisingly, expression analysis of Desmin at an earlier stage (E9.5; Fig. 3I–3K),


2773

Figure 2. NbGFP overexpression in the dorsal dermomyotome (DM) promotes homogeneous distribution of Numb during mitosis. (A): Scheme of different 3X:EpE-GFP-IHR control and 6X::3XEpE-NbGFP(IHR) NbGFP overexpression constructs. Three multimers of the EpE were necessary to drive robust live GFP expression in control embryos (transient F0 or F1 lines), but six copies were needed to detect live NbGFP expression in embryos harvested from stable F1 lines. (B, C): Quantitative reverse-transcription polymerase chain reaction of the relative amounts of Numb transcripts present in interlimb somites; GFPhigh and GFPlow cells isolated by fluorescence-activated cell sorting from Tg:6X::3XEpENbGFP(IHR) and compared with GFPhigh and GFPlow cells of Tg:3XEpE-GFP-IHR control embryos at E10.5 (supporting information Fig. 2, n is the number of samples analyzed with 3 embryos of comparable GFP intensity per sample). **, p < .001. (D, E): Numb whole-mount ISH on E10.75 littermate embryos (n ¼ 3 per genotype). Bracket in (D) indicates dorsal-ventral extent of the somite. Note the ISH signal for Numb in Tg:6X::3XEpE-NbGFP(IHR) is restricted to the dorsal domain of somites. (F–K): Confocal lateral sections of interlimb somites of whole-mount immunohistochemistry-stained Tg:EpE-NbGFP(IHR) (3X or 3X::6X) embryos (apical side of DM, dotted line). Note that the apical cortical crescent of Numb (arrowheads) in mitotic cells is present only in NbGFPlow embryos; cells expressing higher levels of NbGFP show a homogeneous distribution of Numb (arrows). (L): Relative quantification of Numb distribution in dorsal DM mitotic cells from NbGFPhigh embryos expressing Tg:6X::3XEpE-NbGFP(IHR) (n ¼ 3 embryos; n ¼ 208 total cells counted) and NbGFPlow embryos expressing Tg:3XEpE-NbGFP(IHR) (n ¼ 3 embryos; n ¼ 209 total cells counted). (M): Live videomicroscopy of an E9.5 interlimb 6X::3XEpE-NbGFP(IHR) somite. (N–T): Video sequence of a dividing NbGFPþ dorsal DM cell boxed in (M). Note that the overexpressed NbGFP is equally distributed to both daughter cells. Scale bar ¼ 25 lm (F–K); 50 lm (M); and 10 lm (N–T). Abbreviations: CT, cycle threshold; E, embryonic day; eGFP, enhanced green fluorescent protein; EpE, epaxial enhancer; FL, forelimb; HL; hind limb; ISH, in situ hybridization; M-Cad, M-cadherin; cTub, c-tubulin.

www.StemCells.com

2774


Figure 3. Stage-dependent effects of Numb misexpression on differentiation in the dermomyotome (DM).(A–D): Whole-mount Desmin immunohistochemistry (IHC) of E10.5 of embryos (n ¼ 4 for each genotype). (B, D): Lateral view of cumulated Z-stacks of Desmin expression in E10.5 interlimb somites. Bracket indicates dorsal myotome. Scale bar ¼ 100 lm. (E–H): Alx4 ISH on E10.5 (E, F) (n ¼ 3 for each genotype) or E11.5 (G, H) embryos (n ¼ 8 control, n ¼ 4 NbGFP; dorsal most border of somites, dotted line). White arrows and asterisk indicate increase in Alx4 in the dorsal and central dermal domains, respectively. (I–K): Whole-mount IHC of E9 embryos (n ¼ 3 for each genotype). Note that Desmin expression is delayed by three somites in the Tg:6X::3XEpE-NbGFP(IHR) embryos (somite numbers indicated, arrows; posterior most somites are younger). (L): Scheme of the lineages arising from dorsal DM progenitor cells. (M): Western blot of lysates isolated from interlimb somites of individual control (n ¼ 3 Tg:3XEpE-GFP-IHR or WT) or n ¼ 3 Tg:6X::3XEpE-NbGFP(IHR) embryos. Abbreviations: FL, forelimb; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GFP, green fluorescent protein; HL; hind limb; ISH, in situ hybridization; som, somite; WT, wild type.

showed that myogenesis was delayed in transgenic embryos overexpressing NbGFP (n ¼ 3) compared with the control (n ¼ 3). Thus the increase in differentiation observed at the later stage may not be direct. To investigate further how NbGFP overexpression affects muscle differentiation, we performed qRT-PCR analysis with Mrf4, Myod, Myogenin, and Desmin. This analysis showed significant upregulation of these markers in NbGFP overexpressed embryos (Fig. 4A; supporting information Fig. 2). Taken together, we show that NbGFP overexpression in progenitors of the dorsal DM promotes both an increase in myogenic (E10–E10.5) and dermis (E11) cell commitment, and delays the onset of differentiation. If premature differentiation takes place, the respective progenitors in the DM could be depleted. Alternatively, Numb may regulate self-renewal of progenitors in the DM, thus increasing the ancestral progeni-

tor pool size, which results in an increase in muscle as well as dermal precursors. To investigate this possibility, we examined the progenitor cell population by marker gene expression and confocal microscopy.

Numb Plays a Role in Progenitor Cell Self-Renewal in the Somite Since uncommitted stem and progenitor cells in the DM are Pax3þ and/or Pax7þ [43, 45, 68], we examined the expression of these genes in the transgenic embryos by whole-mount immunohistochemistry and RT-qPCR. Importantly, we find that the number of Pax3þ cells in the dorsal somite at E10.75 increased by 34% in the NbGFP-transgenic embryos (control GFP: n ¼ 3 embryos, n ¼ 6,863 total cells counted; NbGFP: n ¼ 3 embryos, n ¼ 8,756 total cells counted; Fig. 5A, 5C;


Figure 4. Numb misexpression in the dermomyotome increases myogenic marker expression. Quantitative reverse-transcription polymerase chain reaction of relative amounts of transcripts in interlimb somites. GFPHigh cells were isolated by fluorescence-activated cell sorting from Tg:6X::3XEpE-NbGFP(IHR) and compared with GFPHigh cells from Tg:3XEpE-GFP-IHR control embryos at E10.5 (supporting information Fig. 2; n is the number of samples analyzed with 3 embryos of comparable GFP intensity per sample). **, p < .001. Abbreviations: CT, cycle threshold; EpE, epaxial enhancer; GFP, green fluorescent protein.

data not shown). We then analyzed the proliferation index in the NbGFP and control transgenics by enumerating BrdUþ cells in the DM (Fig. 5B; data not shown). Accordingly, the numbers of BrdUþ cells as well as the total number of DM cells were elevated in NbGFP-transgenic embryos compared with the control (control GFP: n ¼ 3 embryos, n ¼ 13,101 total cells counted; NbGFP: n ¼ 3 embryos, n ¼ 14,410 total cells counted; Fig. 5A, 5B, 5D). Therefore high levels of Numb in this context result in more proliferation of DM cells. Importantly however, the percentage of BrdUþ and Pax3þ cells in the DM of control and transgenic embryos was similar, indicating that the rate of proliferation was not affected by NbGFP overexpression (Fig. 5D). Furthermore, the level of apoptosis was not overtly altered between the transgenic NbGFP and the control in the Pax3þ DM cells (data not shown). Notably, the overproliferation phenotype was confined principally to the dorsal DM (Fig. 5A, 5B) where transgene expression is most active. This observation reinforces

2775

our findings above showing that the EpE element exerts spatiotemporal expression on the transgenes preferentially in stem and progenitor cells in the DM. To complement these findings, we examined NbGFP overexpression in adult satellite cells associated with the myofiber niche where their properties are less subject to variations compared with cells grown on a culture dish [11]. NbGFP overexpression resulted in a significant increase in the total number of cells, as well as an increase in the number of differentiated cells, however the latter was not observed in all cases (supporting information Fig. 5A-C). These findings suggest that Numb overexpression can promote an increase in precursor and differentiated cells. Accordingly, cells that were NbGFP high/Myogenin negative as well as NbGFP low/ Myogenin positive were observed (supporting information Fig. 5H, 5I).

Notch Signaling in Somites Is Not Significantly Downregulated by Numb We then investigated the mechanisms by which Numb affects the proliferation status of progenitors in the dorsal DM. Numb has been proposed to antagonize Notch signaling, therefore we determined which downstream regulators of Notch signaling are present in the somite by assaying for their expression by RT-qPCR of cells isolated by FACS directly from control transgenic embryos. At E10.5 in WT dorsal somitic cells, Nrarp, Hey1, Hey2, and Hes6 were expressed (confirmed by microarray analysis of Myf5GFP-P/þ MPCs; data not shown). We then determined whether some of these mediators were affected by overexpression of NbGFP. We prospectively isolated by FACS NbGFPhigh (n ¼ 5 samples), NbGFPlow (n ¼ 5 samples) and control GFPhigh (n ¼ 4 samples), control GFPlow (n ¼ 4 samples) cells from interlimb somites of NbGFP (n ¼ 4-5 embryos/samples) and control (n ¼ 3-4 embryos/sample) transgenic embryos at E10.5, respectively (supporting information Fig. 2). As above, we reasoned that the ‘‘high’’ and ‘‘low’’ GFPþ cells corresponded to the cells that were more ancestral in the lineage or more committed, respectively. The levels of Numb transcript in these cells were well above endogenous levels in both cases (Fig. 2B, 2C). The levels of transcripts of Notch target genes in these populations were then compared by RT-qPCR (Fig. 6A, 6B;

Figure 5. Numb overexpression in the dermomyotome (DM) results in an increase in progenitor cells. (A): Scheme showing orientation of confocal lateral scanned planes generated on whole-mount immunostained embryos used for BrdU and Pax3 quantifications. The dorsal half of identically positioned interlimb somites of stage-matched Tg:6X::3XEpE-NbGFP(IHR) embryos (n ¼ 3) and Tg:3XEpE-GFP-IHR embryos (n ¼ 3) were scanned. DM (GFPþ/Desmin) and myotome (Desminþ) cells were distinguished (total cell number ¼ total Hoechstþ nuclei within the dorsal somite in a zA to zE scan). (B, C): BrdU (B) and Pax3 (C) quantifications. (D): BrdUþ and Pax3þ cell ratios, calculated from total cell number from zA-zC. Abbreviations: BrdU, bromodeoxyuridine; EpE, epaxial enhancer; GFP, green fluorescent protein.

www.StemCells.com

2776


Figure 6. Numb overexpression does not downregulate Notch signaling during somitogenesis. (A, B): Quantitative reverse-transcription polymerase chain reaction of the relative amounts of Notch signaling targets in embryonic day (E) 10.5 interlimb NbGFPHigh cells (A) and NbGFPlow cells (B) from Tg:6X::3XEpE-NbGFP(IHR) compared with Tg:3XEpE-GFP-IHR control embryos (supporting information Fig. 2; n is the number of samples analyzed with 3 embryos of comparable GFP intensity per sample). *, p < .01, **, p < .001. (C) and (E): Confocal lateral scans with their corresponding DIC merged images ([D] and [F], respectively) of E11.5 caudal tail somites stained for Numb (C, D’), GFP (E, F’), and bromo-chloro-indolyl-galactopyranoside (D, D’, F, F’). Outer somite borders are outlined in (C–F). (D’) and (F’): High magnifications of the insets boxed in (D) and (F), respectively. (G, H): HeyL whole-mount ISH on E11.5 caudal somites in tails. (I): Immunostaining of GFP in tail of E10.75 Tg:6X::3XEpE-NbGFP(IHR) embryo. Abbreviations: CT, cycle threshold; DIC, differential contrast; EpE, epaxial enhancer; GFP, green fluorescent protein; ISH, in situ hybridization. Scale bars ¼ 50 lm.

data not shown). Interestingly, the mRNA levels of the downstream effectors of Notch signaling (Hes1, Hes6, Hey1, Hey2, HeyL, Nrarp) in GFPhigh and GFPlow cells, as well as in the intermediate (data not shown) population, were either not altered or, intriguingly, they were significantly increased in E10.5 transgenic NbGFP compared with control embryos. This was also the case for the Notch1 and Notch2 receptors. This suggests that in the context of the dorsal somite, Numb participates in stem/progenitor pool homeostasis without downregulating canonical Notch signaling. These results were further confirmed at earlier stages of somite patterning in caudal somites. Notch targets such as HeyL are implicated in the somite segmentation clock and subsequent patterning [41] and exhibit a segmented transcript distribution pattern [69] (Fig. 6G). Both control and NbGFPtransgenic lines expressed Numb at high levels in the caudal somites and presomitic mesoderm, where somite segmentation occurs (Fig. 6I; data not shown). Importantly, high levels of NbGFP expression in the presomitic mesoderm and caudal

somites did not perturb somite segmentation (Fig. 6E, 6F, 6H, 6I). Moreover, the HeyL expression pattern was not perturbed in the tail somites of the NbGFP embryos (n ¼ 8) compared with control (n ¼ 4) (Fig. 6G, 6H). To confirm these observations, NbGFP-transgenic mice were crossed with Notch activity sensor (NAS) mice [70], which act as a readout of canonical Notch signaling. NAS embryos exhibit typical Notch signaling activity as shown by 5-bromo-4-chloro-3-indolylbeta-D-galactopyranoside (X-gal) staining of the caudal lip of the budding tail somites (n ¼ 4 embryos; Fig. 6D, D’). In the NbGFP-transgenic embryos in these caudal somites, the NAS pattern was not altered (n ¼ 4 embryos). Importantly, high NbGFP-expressing cells were robustly X-galþ, reflecting high Notch signaling activity (Fig. 6E, 6F, 6F’). In addition, analysis of endogenous Numb expression in NAS embryos showed that Numb is highly expressed in almost all of the cells in the caudal somites, and that this high level of expression is compatible with high NAS expression (Fig. 6C, 6D, 6D’). High levels of Numb protein are therefore compatible with high


2777

levels of Notch signaling in transgenic as well as WT cells, indicating that in young as with more mature somites Numb expression does not inhibit Notch signaling.

DISCUSSION Somites generate multiple cell types with precise spatiotemporal control. The balance between the self-renewal of distinct stem cell populations and their coordinated differentiation is critical for the establishment of tissues. Although endothelial, mural, dorsal dermal, and smooth muscle cells are specified in the DM, it is thought that only skeletal myogenesis is regulated by Pax3 and Pax7 [38]. Given that all of the cells in the DM express one or both of these proteins additional regulators must coordinate cell fate decisions in the DM. These observations prompted us to investigate the role of Numb and its interplay with Notch in a precise spatiotemporal context in vivo and within a lineage progression when dermal and skeletal muscle cell fates occur concomitantly. This gain-of-function approach previously identified Numb as a critical cell fate determinant in SOP and neuroblast cells, and skeletal muscle precursors in Drosophila [63, 64], as well as pointing to a role for Notch and BMP4 signaling in cell fate decisions in the somite [39, 57]. One complication in vertebrates is the presence of a related molecule Numblike, and multiple isoforms for Numb (supporting information Fig. 1). Notably, Nbp66 is distinct from avian Numb (supporting information Fig. 1), and its subcellular distribution in the DM in mice (present study) and avians [14, 15] appears to be different. Targeting the overexpression of NbGFP to stem/progenitor cells in the DM promotes an increase in cell number without altering their rate of proliferation. These findings contrast with studies where the overexpression of Numb was reported to promote muscle differentiation in satellite cells or in myoblasts in culture [10, 32], however the precursor cells were not systematically enumerated in those studies. Although we observed a slight increase in differentiation in the somite at early stages, we reason that this is due to a disproportional increase in the stem/progenitor pool size, thereby resulting in more differentiated cells (Fig. 7). Accordingly, as the myotome is initiated, we observed that Numb overexpression delayed differentiation. Notably, the ancestral pool is not depleted by overexpressing Numb, and differentiation is neither overtly favored nor repressed. Since our aim was to target Numb expression in the ancestral population, our approach precluded us from examining the role of Numb directly in the decision to differentiate. If Numb does play a role in the latter, we propose that a transitory cell state would be susceptible to differentiate by inhibition of Notch, since Numb overexpression did not promote commitment in all of the myogenic progenitors. The stage in which Notch intervenes in vivo is not yet resolved. Genetic experiments in the mouse using compromised Delta1 activity showed that the early Pax3/Pax7 progenitors were not affected, yet a decline in this population was observed later during development [26]. Overexpression of Delta1 in chick somites suggested that Notch acts at a later stage, during myogenic commitment or differentiation [25]. Our finding that Numb is distributed asymmetrically as a crescent that can be unequally partitioned to prospective daughter cells during mitosis suggests that asymmetric cell divisions might regulate self-renewal or differentiation. Interestingly, virtually all of the dividing cells examined in the dorsal DM where MPCs are born exhibit this cortical Numb crescent oriented toward the myotome, and importantly, the overexpression of NbGFP resulted in its widespread distribuwww.StemCells.com

Figure 7. Model for cell fate regulation by Numb. In somites, binary decisions result in the generation of dermal (Alx4þ) and skeletal muscle (Pax3þ) progenitor cells, followed by their commitment. Targeted spatiotemporal overexpression of Numb in these progenitors by transgenesis results in the increase in the number of this ancestral population in the DM. As a consequence, both dermal and skeletal muscle progenitor pool sizes increase. We propose that this is a result in a shift from asymmetric to symmetric cell divisions. Interestingly, differentiation is neither inhibited nor favored. Rather, due to the increase in the ancestral pool size, more committed cells are observed in both lineages. Abbreviation: DM, dermomyotome.

tion in the dividing DM cells. This phenotype may provoke with the cell fate changes that we observed after NbGFP overexpression. This notion can be investigated in future studies by examining the role of Numb during mitosis in Nb:Nbl conditional double mutants. Experiments with adult muscle satellite cells as well as skeletal muscle C2C7 cells (V. Shinin, S. Tajbakhsh, unpublished observations) overexpressing NbGFP do not show a disproportional propensity to differentiate, indicating that as in the embryo, the role of Numb in cell differentiation is highly cell-context regulated in the adult. The finding that the myotome expresses high levels of Numb raises the possibility that Numb may play additional roles in differentiation, as it was proposed for cell-cell adhesion and recycling transmembrane receptors in differentiated neurons [36, 71] and cells in culture [4, 72], or its regulation of cell fates independent of its asymmetric localization in Drosophila [62]. Whether Numb plays a direct role in selecting cells that will differentiate, or its upregulation is a consequence of differentiation, needs further investigation. In our present analysis, the changes that we noted on Notch signaling effectors were observed in cells that had low and high overexpression levels on Numb. Other regulators such as ACBD3, a Golgi-associated protein in interphase that interacts directly with Numb and regulates its function during mitosis [73], may also intervene at one of these steps in the somite during cell fate decisions. How these proteins coregulate Notch activity is not clear. In this light, we report an increase in the number of dermal as well as muscle cells in accordance with an increase in number of their ancestral founders, whereas in the retina, Numb overexpression resulted in the production of photoreceptors at the expense of other cell types [74]. These seemingly disparate observations underscore the importance in timing or cell cycle stage in which Numb exerts its functions. Although dermal and skeletal muscle precursors share a common ancestor in somites it is not known how cell commitment occurs from a common Pax3/Pax7-positive ancestral cell in the DM. One possibility is that binary asymmetric cell divisions generate a dermal progenitor that downregulates Pax7 expression and a skeletal muscle progenitor that retains Pax7 protein. If this were to occur exclusively, it would result in the depletion of DM cells. This is not the case, as the DM


2778

increases incrementally in size. Thus, other resident DM cells must compensate for the consequent loss of cells by performing symmetric cell divisions as well. Another possibility is that individual lineages are already specified, and an asymmetric cell division gives rise to a self-renewing lineage-determined DM cell and a committed cell that leaves the DM. This scenario would not result in a net loss of cells from the DM. Again, other types of cell divisions would be necessary to account for DM growth. Thus in both scenarios, symmetric self-renewing cell divisions should operate to account for the increase in the DM size. Therefore, we speculate that a balance between symmetric and asymmetric cell divisions regulates growth and cell type specification within the DM. Our finding that Numb distributes asymmetrically during mitosis is in keeping with this notion. Furthermore, as indicated above the overexpression of Numb in the DM results in symmetrictype cell divisions, and this occurs concomitantly with in an increase in dermal and skeletal muscle progenitors, as well as in the size of the DM. However, a role for Numb in regulating DM cell adhesion is not excluded. The direct correlation of Numb asymmetry with cell fate decisions has been elusive in vertebrates, and only a few paradigms that have been amenable to address this issue have been reported [8, 75]. The widely held view that Numb antagonizes Notch signaling stems largely from work in Drosophila [18–21]. In higher vertebrates, no direct example is available in vivo within a lineage progression. By prospectively isolating only those cells overexpressing NbGFP we assess for the first time several downstream targets of Notch signaling. Importantly, these cells are in a spatiotemporal context, and a lineage progression. Previous studies showed that Notch signaling inhibits Myod expression via Hes1 [30]. Here we report that in NbGFP expressing cells (both high and low GFP) several Notch signaling effectors were either not altered, or unexpectedly significantly increased in expression. Given the role that Numb plays in endocytosis [4, 72], it is possible that the turnover of molecules that putatively inhibit Notch activity was recycled by Numb, thereby resulting in an increase in some Notch targets. Alternatively, Numb stimulates Notch targets by a mechanism that remains to be determined, or uniquely in a particular cell state. Thus, these findings argue against a dominant role in the inhibition of Notch signaling by Numb in the majority of dorsal DM cells at this stage, although other regions of the somite were reported to be dependent on Notch activity [39, 57]. Interestingly, overexpression of Numb was reported to inhibit Notch activity, but only in cells with low Notch signaling [33]. This subtle regulation, and other experimental strategies are needed to address the role of these molecules after the disappearance of the somite and the formation of muscle masses. The failure of Numb overexpression to downregulate a Notch reporter in immature somites suggests that Numb may not act by inhibiting Notch signaling in the process of boundary formation in somites [41]. Using an overexpression strategy in quail somites, it was suggested recently that Notch and BMP4 signaling regulate the fate of smooth muscle and endothelial cell fates from the ventral DM [39]. In that study, the overexpression of Numb

REFERENCES 1 2

Knoblich JA, Mechanisms of asymmetric stem cell division. Cell 2008;132:583–597. Betschinger J, Knoblich JA. Dare to be different: asymmetric cell division in Drosophila, C. elegans, vertebrates. Curr Biol 2004;14: R674–R685.

directed DM cells to the myotome, however the fate of progenitors was not examined. Our findings led us to propose an alternate model where Numb overexpression results primarily in an increase in the progenitor pool size. The modest increase in differentiated cells would be a consequence of this event, as the rate of proliferation and differentiation does not appear to be altered (Fig. 7). Interestingly, Numb deregulation is associated with a number of cancers, consistent with the notion that it can act as a tumor suppressor [76]. In addition, Numb has been reported to regulate the ubiquitin-mediated degradation of Gli1 [77], an effector of Sonic hedgehog signaling that is known to promote the emergence of MPCs and regulate their lineage progression [78]. This raises the intriguing possibility that Numb may modulate the decision to differentiate by alternatively regulating Shh and Notch signaling pathways as was proposed from studies in vitro [77].

CONCLUSION Although a role for Notch has been proposed for regulating skeletal and smooth muscle fates in the ventral somite, and muscle progenitor production at later stages, we show that muscle progenitors in the dorsal DM do not react to Numb overexpression by collectively committing to myogenesis and downregulating Notch effector molecules. Moreover, we provide direct evidence for the asymmetric distribution of Numb in dividing DM cells and we propose that Numb can stimulate self-renewal of stem/progenitor cells without significantly downregulating Notch signaling. Given the important role that Numb plays in multiple cell fate choices in metazoans, and its proposed role in epithelial cells and in carcinogenesis, our studies provide important insights into the function of Numb within a normal lineage progression in vivo in the context of cell fate choices.

ACKNOWLEDGMENTS We thank P. Flamant and G. Dumas for technical assistance, members of the lab for helpful discussions, and Pierre Charneau for providing lentivirus reagents. This work was funded by grants from the Institut Pasteur, Association Française contre les Myopathies, Association pour la recherche sur le cancer, MyoRes (European Union Framework 6 project LSHG-CT2004-511978) and EuroStemCell (European Union Framework 6 project LHSB-CT-2003-503005). AJ was supported by fellowships from the French Ministry of Education and Research, the AFM and the Pasteur-Weizmann foundation.

DISCLOSURE

OF OF

POTENTIAL CONFLICTS INTEREST

The authors indicate no potential conflicts of interest.

3 4 5

Roegiers F, Jan YN. Asymmetric cell division. Curr Opin Cell Biol 2004;16:195–205. Nishimura T, Kaibuchi K. Numb controls integrin endocytosis for directional cell migration with aPKC, PAR-3. Dev Cell 2007;13: 15–28. Dho SE, Trejo J, Siderovski DP et al. Dynamic regulation of mammalian numb by G protein-coupled receptors, protein kinase C activation: structural determinants of numb association with the cortical membrane. Mol Biol Cell 2006;17:4142–4155.


6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

23 24 25 26 27 28 29 30 31 32 33 34

Smith CA, Lau KM, Rahmani Z et al. aPKC-mediated phosphorylation regulates asymmetric membrane localization of the cell fate determinant Numb. EMBO J 2007;26:468–480. Zhong W, Feder JN, Jiang MM et al. Asymmetric localization of a mammalian numb homolog during mouse cortical neurogenesis. Neuron 1996;17:43–53. Cayouette M, Whitmore AV, Jeffery G et al. Asymmetric segregation of Numb in retinal development, the influence of the pigmented epithelium. J Neurosci 2001;21:5643–5651. Shen Q, Zhong W, Jan YN et al. Asymmetric Numb distribution is critical for asymmetric cell division of mouse cerebral cortical stem cells, neuroblasts. Development 2002;129:4843–4853. Conboy IM, Rando TA. The regulation of Notch signaling controls satellite cell activation, cell fate determination in postnatal myogenesis. Dev Cell 2002;3:397–409. Shinin V, Gayraud-Morel B, Gomes D et al. Asymmetric division, cosegregation of template DNA strands in adult muscle satellite cells. Nat Cell Biol 2006;8:677–682. Wakamatsu Y, Maynard TM, Jones SU, et al. NUMB localizes in the basal cortex of mitotic avian neuroepithelial cells, modulates neuronal differentiation by binding to NOTCH-1. Neuron 1999;23:71–81. Wakamatsu Y, Maynard TM, Weston JA. Fate determination of neural crest cells by NOTCH-mediated lateral inhibition, asymmetrical cell division during gangliogenesis. Development 2000;127:2811–2821. Venters SJ, Ordahl CP. Asymmetric cell divisions are concentrated in the dermomyotome dorsomedial lip during epaxial primary myotome morphogenesis. Anat Embryol (Berl) 2005;209:449–460. Holowacz T, Zeng L, Lassar AB. Asymmetric localization of numb in the chick somite, the influence of myogenic signals. Dev Dyn 2006; 235:633–645. Chang JT, Palanivel VR, Kinjyo I et al. Asymmetric T lymphocyte division in the initiation of adaptive immune responses. Science 2007; 315:1687–1691. Wu M, Kwon HY, Rattis F et al. Imaging hematopoietic precursor division in real time. Cell Stem Cell 2007;1:541–554. Le Borgne R, Bardin A, Schweisguth F. The roles of receptor, ligand endocytosis in regulating Notch signaling. Development 2005;132: 1751–1762. Schweisguth F. Notch signaling activity. Curr Biol 2004;14: R129–R138. Guo M, Jan LY, Jan YN. Control of daughter cell fates during asymmetric division: interaction of Numb, Notch. Neuron 1996;17:27–41. Le Borgne R. Regulation of Notch signalling by endocytosis, endosomal sorting. Curr Opin Cell Biol 2006;18:213–222. Frise E, Knoblich JA, Younger-Shepherd S et al. The Drosophila Numb protein inhibits signaling of the Notch receptor during cell-cell interaction in sensory organ lineage. Proc Natl Acad Sci U S A 1996; 93:11925–11932. Petersen PH, Zou K, Hwang JK et al. Progenitor cell maintenance requires numb, numblike during mouse neurogenesis. Nature 2002; 419:929–934. Li HS, Wang D, Shen Q et al. Inactivation of Numb, Numblike in embryonic dorsal forebrain impairs neurogenesis, disrupts cortical morphogenesis. Neuron 2003;40:1105–1118. Hirsinger E, Malapert P, Dubrulle J, et al. Notch signalling acts in postmitotic avian myogenic cells to control MyoD activation. Development 2001;128:107–116. Schuster-Gossler K, Cordes R, Gossler A. Premature myogenic differentiation, depletion of progenitor cells cause severe muscle hypotrophy in Delta1 mutants. Proc Natl Acad Sci U S A 2007;104:537–542. Vasyutina E, Lenhard DC, Wende H et al. RBP-J (Rbpsuh) is essential to maintain muscle progenitor cells, to generate satellite cells. Proc Natl Acad Sci U S A 2007;104:4443–4448. Kageyama R, Ohtsuka T, Kobayashi T. The Hes gene family: repressors, oscillators that orchestrate embryogenesis. Development 2007; 134:1243–1251. Lamar E, Deblandre G, Wettstein D et al. Nrarp is a novel intracellular component of the Notch signaling pathway. Genes Dev 2001;15: 1885–1899. Jarriault S, Brou C, Logeat F et al. Signalling downstream of activated mammalian Notch. Nature 1995;377:355–358. Sun J, Kamei CN, Layne MD et al. Regulation of myogenic terminal differentiation by the hairy-related transcription factor CHF2. J Biol Chem 2001;276:18591–18596. Kitzmann M, Bonnieu A, Duret C et al. Inhibition of Notch signaling induces myotube hypertrophy by recruiting a subpopulation of reserve cells. J Cell Physiol 2006;208:538–548. Chapman G, Liu L, Sahlgren C et al. High levels of Notch signaling down-regulate Numb, Numblike. J Cell Biol 2006;175:535–540. Dho SE, French MB, Woods SA et al. Characterization of four mammalian numb protein isoforms. Identification of cytoplasmic, membrane-associated variants of the phosphotyrosine binding domain. J Biol Chem 1999;274:33097–33104.

www.StemCells.com

2779

35 Verdi JM, Bashirullah A, Goldhawk DE et al. Distinct human NUMB isoforms regulate differentiation vs. proliferation in the neuronal lineage. Proc Natl Acad Sci U S A 1999;96:10472–10476. 36 Rasin MR, Gazula VR, Breunig JJ et al. Numb, Numbl are required for maintenance of cadherin-based adhesion, polarity of neural progenitors. Nat Neurosci 2007;10:819–827. 37 Petersen PH, Tang H, Zou K et al. The enigma of the numb-Notch relationship during mammalian embryogenesis. Dev Neurosci 2006; 28:156–168. 38 Sambasivan R, Tajbakhsh S. Skeletal muscle stem cell birth, properties. Semin Cell Dev Biol 2007;18:870–882. 39 Ben-Yair R, Kalcheim C. Notch, bone morphogenetic protein differentially act on dermomyotome cells to generate endothelium, smooth, striated muscle. J Cell Biol 2008;180:607–618. 40 Delfini MC, Hirsinger E, Pourquie O et al. Delta 1-activated notch inhibits muscle differentiation without affecting Myf5, Pax3 expression in chick limb myogenesis. Development 2000;127:5213–5224. 41 Dequeánt ML, Pourquie O. Segmental patterning of the vertebrate embryonic axis. Nat Rev Genet 2008;9:370–382. 42 Kassar-Duchossoy L, Gayraud-Morel B, Gom[grave]es D et al. Mrf4 determines skeletal muscle identity in Myf5:Myod double-mutant mice. Nature 2004;431:466–471. 43 Kassar-Duchossoy L, Giacone E, Gayraud-Morel B et al. Pax3/Pax7 mark a novel population of primitive myogenic cells during development. Genes Dev 2005;19:1426–1431. 44 Kuang S, Kuroda K, Le Grand F et al. Asymmetric self-renewal, commitment of satellite stem cells in muscle. Cell 2007;129:999–1010. 45 Relaix F, Rocancourt D, Mansouri A et al. A Pax3/Pax7-dependent population of skeletal muscle progenitor cells. Nature 2005;435: 948–953. 46 Rudnicki MA, Schneglesberg PNG, Stead RH et al. MyoD or myf-5 is required for the formation of skeletal muscle. Cell 1993;75: 1351–1359. 47 Seale P, Sabourin LA, Girgis-Gabardo A et al. Pax7 is required for the specification of myogenic satellite cells. Cell 2000;102: 777–786. 48 Tajbakhsh S, Rocancourt D, Cossu G et al. Redefining the genetic hierarchies controlling skeletal myogenesis: Pax-3, Myf-5 act upstream of MyoD. Cell 1997;89:127–138. 49 Tajbakhsh S. Stem cells to tissue: molecular, cellular, anatomical heterogeneity in skeletal muscle. Curr Opin Genet Dev 2003;13:413–422. 50 Kalcheim C, Kahane N, Cinnamon Y et al. Mechanisms of lineage segregation in the avian dermomyotome. Anat Embryol (Berl) 2006; 211(suppl 1):31–36. 51 Ben-Yair R, Kalcheim C. Lineage analysis of the avian dermomyotome sheet reveals the existence of single cells with both dermal, muscle progenitor fates. Development 2005;132:689–701. 52 Ruiz Go´mez M, Bate M. Segregation of myogenic lineages in Drosophila requires numb. Development 1997;124:4857–4866. 53 Han Z, Bodmer R. Myogenic cells fates are antagonized by Notch only in asymmetric lineages of the Drosophila heart, with or without cell division. Development 2003;130:3039–3051. 54 Toriya M, Tokunaga A, Sawamoto K et al. Distinct functions of human numb isoforms revealed by misexpression in the neural stem cell lineage in the Drosophila larval brain. Dev Neurosci 2006;28: 142–155. 55 Zhong W, Jiang MM, Weinmaster G et al. Differential expression of mammalian Numb, Numblike, Notch1 suggests distinct roles during mouse cortical neurogenesis. Development 1997;124:1887–1897. 56 Cayouette M, Barres BA, Raff M. Importance of intrinsic mechanisms in cell fate decisions in the developing rat retina. Neuron 2003;40: 897–904. 57 Sato Y, Watanabe T, Saito D et al. Notch mediates the segmental specification of angioblasts in somites, their directed migration toward the dorsal aorta in avian embryos. Dev Cell 2008;14:890–901. 58 Petersen PH, Zou K, Krauss S et al. Continuing role for mouse Numb, Numbl in maintaining progenitor cells during cortical neurogenesis. Nat Neurosci 2004;7:803–811. 59 Carvajal JJ, Cox D, Summerbell D et al. A BAC transgenic analysis of the Mrf4/Myf5 locus reveals interdigitated elements that control activation, maintenance of gene expression during muscle development. Development 2001;128:1857–1868. 60 Borello U, Berarducci B, Murphy P et al. The Wnt/beta-catenin pathway regulates Gli-mediated Myf5 expression during somitogenesis. Development 2006;133:3723–3732. 61 Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 2008;3:1101–1108. 62 Bhalerao S, Berdnik D, Torok T et al. Localization-dependent, -independent roles of numb contribute to cell-fate specification in Drosophila. Curr Biol 2005;15:1583–1590. 63 Uemura T, Shepherd S, Ackerman L et al. numb, a gene required in determination of cell fate during sensory organ formation in Drosophila embryos. Cell 1989;58:349–360.


2780

64 Rhyu MS, Jan LY, Jan YN. Asymmetric distribution of numb protein during division of the sensory organ precursor cell confers distinct fates to daughter cells [see comments]. Cell 1994;76:477–491. 65 Summerbell D, Ashby PR, Coutelle O et al. The expression of Myf5 in the developing mouse embryo is controlled by discrete, dispersed enhancers specific for particular populations of skeletal muscle precursors. Development 2000;127:3745–3757. 66 Teboul L, Hadchouel J, Daubas P et al. The early epaxial enhancer is essential for the initial expression of the skeletal muscle determination gene Myf5 but not for subsequent, multiple phases of somitic myogenesis. Development 2002;129:4571–4580. 67 Ahmed MU, Cheng L, Dietrich S. Establishment of the epaxialhypaxial boundary in the avian myotome. Dev Dyn 2006;235: 1884–1894. 68 Gros J, Manceau M, Thome V et al. A common somitic origin for embryonic muscle progenitors, satellite cells. Nature 2005;435: 954–958. 69 Leimeister C, Schumacher N, Steidl C et al. Analysis of HeyL expression in wild-type, Notch pathway mutant mouse embryos. Mech Dev 2000;98:175–178. 70 Souilhol C, Cormier S, Monet M et al. Nas transgenic mouse line allows visualization of Notch pathway activity in vivo. Genesis 2006; 44:277–286.

71 Kuo CT, Mirzadeh Z, Soriano-Navarro M et al. Postnatal deletion of Numb/Numblike reveals repair, remodeling capacity in the subventricular neurogenic niche. Cell 2006;127:1253–1264. 72 Nishimura T, Yamaguchi T, Tokunaga A et al. Role of numb in dendritic spine development with a Cdc42 GEF intersectin, EphB2. Mol Biol Cell 2006;17:1273–1285. 73 Zhou Y, Atkins JB, Rompani SB et al. The mammalian Golgi regulates numb signaling in asymmetric cell division by releasing ACBD3 during mitosis. Cell 2007;129:163–178. 74 Cayouette M, Raff M, The orientation of cell division influences cellfate choice in the developing mammalian retina. Development 2003; 130:2329–2339. 75 Cayouette M, Poggi L, Harris WA. Lineage in the vertebrate retina. Trends Neurosci 2006;29:563–570. 76 Caussinus E, Gonzalez C. Induction of tumor growth by altered stemcell asymmetric division in Drosophila melanogaster. Nat Genet 2005; 37:1125–1129. 77 Di Marcotullio L, Ferretti E, Greco A et al. Numb is a suppressor of Hedgehog signalling, targets Gli1 for Itch-dependent ubiquitination. Nat Cell Biol 2006;8:1415–1423. 78 Tajbakhsh S, Buckingham M. The birth of muscle progenitor cells in the mouse: spatiotemporal considerations. Curr Top Dev Biol 2000; 48:225–268.

See www.StemCells.com for supporting information available online.