Mechanisms of ligand-mediated inhibition in Notch signaling activity in Drosophila



The transmembrane proteins Delta and Serrate act as ligands for the signaling receptor Notch. In addition to this activating role, Delta and Serrate can also inhibit Notch signaling activity. This inhibitory effect is concentration-dependent and appears to be evolutionarily conserved. In characterizing the underlying cellular mechanisms of the ligand inhibitory effect, we can confirm that ligand-mediated inhibition of Notch signaling can occur as a cell autonomous process (cis-inhibition) and that ligand-mediated inhibition prevents a step in Notch signaling activation early enough to suppress Notch ectodomain shedding. Developmental Dynamics 239:798–805, 2010. © 2010 Wiley-Liss, Inc.


Notch signaling is a molecular event that mediates short range cell communication and is used in many contexts during cell differentiation and pattern formation. Signaling activity is triggered by an interaction between the extracellular domain of the transmembrane Notch receptor with one of two ligands, Delta or Serrate (reviewed in Schweisguth, 2004). This leads to the exposure of a metalloprotease cleavage site on the Notch extracellular domain (NECD; Parks et al., 2000; Nichols et al., 2007). Members of the ADAM family of metalloproteases then cleave Notch at this site, releasing the ectodomain (Brou et al., 2000; Lieber et al., 2002), and generating a substrate for a second, ligand-independent cleavage within the transmembrane domain (Kopan et al., 1996; Schroeter et al., 1998). This series of cleavage events releases the Notch intracellular domain (NICD), which migrates into the nucleus to elicit target gene expression (reviewed in Kovall, 2007).

The interactions between Notch and its ligands are complex, as Serrate and Delta can act both as signal activators or inhibitors in a concentration-dependent manner (Doherty et al., 1996; de Celis and Bray, 1997; Klein et al., 1997; Jacobsen et al., 1998). Studies of wing patterning in Drosophila have highlighted the significance of these dual modes of action (Micchelli et al., 1997). In late third larval instar larvae, Notch and both its ligands are expressed throughout the wing pouch. However, Notch signaling activity is only elicited at the dorsal/ventral (D/V) boundary region of the wing disc (reviewed in Irvine and Rauskolb, 2001). This pattern of activation is not consistent with the ubiquitous expression of Notch and its ligands, suggesting the existence of mechanisms that restrict the activation of Notch, one of them being ligand-mediated inhibition (Klein et al., 1997; Micchelli et al., 1997; Le Borgne et al., 2005; Glittenberg et al., 2006;).

Observations from genetic studies on ligand-mediated inhibition can be explained by the existence of ligand-mediated cis inhibition of Notch signaling (Heitzler and Simpson, 1993; Jacobsen et al., 1998; Glittenberg et al., 2006), but the experimental paradigms cannot rule out inhibition mediated by a trans interaction (inter-cellular) between the ligands. Furthermore, no mechanism has been proposed for the inhibition (Klein et al., 1997; Jacobsen et al., 1998; Perez et al., 2005; Glittenberg et al., 2006). Here we have established a cell culture system to study ligand-mediated inhibition of Notch signaling in Drosophila. Our work focuses on demonstrating that ligand-mediated inhibition can take place as a cis inhibitory effect which prevents a step before the process of shedding of the ectodomain of Notch.


Notch ligands can exert an inhibitory effect on Notch signaling activity in a concentration-dependent manner: high levels of ligand expression induce the ligand inhibitory effect, whereas when lower levels are present, only the signaling activating function is observed (Klein et al., 1997; Micchelli et al., 1997; Jacobsen et al., 1998). A clear demonstration of ligand-mediated inhibition is observed in cells mutant for both Delta and Serrate in Drosophila wing discs. They exhibit activation of Notch signaling in mutant cells at the border of the clones (Fig. 1A and Micchelli et al., 1997), mediated by ligand from surrounding wild-type cells. Clearly the ligand can mediate an inhibitory effect on signaling, but these observations do not answer the question of how does the ligand mediate inhibition? A cell autonomous cis inhibitory process could be mediated by either high levels of ligand inhibiting the Notch signal transduction process or by co-expression of Notch and ligand within a same cell, inhibiting the ligands ability to induce Notch signaling in a neighboring cell. A trans inhibitory process might arise if high levels of ligand allow ligands of neighboring cells to establish intercellular interactions between ligands. This would titrate out ligand available for inducing activation mediated by a trans interaction with the Notch receptor. In support of the ligand in trans model Delta-expressing cells are able to form cell aggregates (Fehon et al., 1990; Parks et al., 2006), indicating that a trans interaction between Delta molecules could readily occur. To determine the mechanism that lies behind ligand-mediated inhibition, we developed a cell-based assay to test the existence of the cis inhibitory process and dissect the molecular interactions taking place. Unfortunately, this assay does not allow us to examine any trans inhibitory processes.

Figure 1.

Ligand co-expression with Notch inhibits Notch signaling activity. A:Drosophila wing imaginal discs with Ser/Dl MARCM clones marked by green fluorescent protein (GFP) expression [A,D,G, and green in C,F,I]. Wingless expression [B,E,H and red in C,F,I] marks Notch signaling. Note ectopic Notch signaling is observed in Ser and Dl cells bordering wild-type cells. B: Analysis of ligand-mediated inhibition of Notch signaling activity in cell-based reporter assays, cells in which signaling was activated by Delta were co-transfected to express Fringe to enhance the level of Notch signaling.

A Cell Culture System to Study Ligand Inhibition

A cell-based assay of ligand induced inhibition of Notch signaling was established using Drosophila S2 cells as they do not express detectable levels of Notch or Notch ligands (Fehon et al., 1991). Notch signaling activity was assessed using a Notch signaling responsive element (Furriols and Bray, 2001, Experimental Procedures; and Perez et al., 2005).

We found that co-expression of Serrate (Ser) with Notch in reporter cells inhibited Serrate-induced Notch signaling and similarly, co-expression of Delta (Dl) with Notch in reporter cells inhibited Delta-induced Notch signaling (Fig. 1B). These results mimic the inhibitory effects seen in vivo (Micchelli et al., 1997; and Fig. 1A). To evaluate whether one ligand can inhibit the other ligand from signaling, we co-expressed Notch with one ligand in reporter cells and tested whether it could abrogate activation of signaling activity induced by the other ligand (Fig. 2A). We observe that Serrate and Notch co-expression in reporter cells inhibits Delta-induced Notch signaling and Delta and Notch co-expression in reporter cells inhibits Serrate-induced Notch signaling (Fig. 2A). These results show ligand-mediated inhibition of Notch signaling in cell culture, and demonstrate that either ligand can inhibit signaling elicited by the same ligand or the other ligand expressed in a neighboring cell.

Figure 2.

Signaling and aggregation properties of S2 cells expressing Notch and its ligands. A: Serrate and Notch co-expression in reporter cells inhibits Delta-induced Notch signaling and Delta and Notch co-expression in reporter cells inhibits Serrate-induced Notch signaling. B: The ability of cells to aggregate when expressing Notch and/or Delta was quantified using cell aggregation assays. [A] S2 cells, [B] stable S2 cell line expressing Notch (S2N), [C] stable S2 cell line expressing Dl (S2Dl), [D] S2N cells transiently transfected with Delta (S2N(Dl)), [E] S2N cells mixed with S2Dl cells (S2N+S2Dl), [F] S2N(Dl) mixed with S2Dl (S2N(Dl)+S2Dl). The image collection and aggregate quantification was made with live cell samples after 15 min of mixing. Only groups of four cells or larger were considered as aggregates.

Interactions between Notch and Notch ligands promote cell adhesion (Fehon et al., 1990; Parks et al., 2006). If co-expression of Notch and ligand inhibits an intercellular interaction, this might alter cellular aggregation properties. In the aggregation conditions tested, cells stably expressing Notch (S2N) or Delta (S2Dl) did not display homotypic aggregate formation over the low background observed for normal S2 cells (S2; Fig. 2B). The lack of aggregation in S2Dl cells, which differs from previous reports (Fehon et al., 1990; Parks et al., 2006), could be due to more stringent aggregation conditions being used. When mixing different cell populations by combining S2N cells with S2Dl cells, S2N cells co-expressing Dl with S2Dl cells or S2N cells co-expressing Dl on their own, they all formed aggregates (Fig. 2B). The fact that cells co-expressing Notch and Delta form aggregates when mixed with Delta-expressing cells but are unable to exhibit Notch signaling activation (Fig. 1B) indicates that co-expression of both Notch and Delta has no significant inhibitory effect on the Notch–Delta intercellular interaction (Fig. 2B).

A cis Inhibitory Effect

To test the existence of cis inhibition, we made use of the observation that treatment of Notch-expressing cells with ethylenediaminetetraacetic acid (EDTA) results in ligand independent activation of Notch signaling (Rand et al., 2000; Krejci and Bray, 2007). EDTA acts by chelating calcium ions, which are essential for the maintenance of the bonds between the membrane-tethered and extracellular regions of the Notch protein (Rand et al., 2000; Sanchez-Irizarry et al., 2004) and also for the maintenance of the structure of the Notch LNR repeats (Sanchez-Iricarry et al., 2004). The LNR repeats structure has been reported to have a protective role to prevent ligand-independent metalloprotease-mediated cleavage (Sanchez- Irizarry et al., 2004). Use of a matrix metalloprotease inhibitor (galardin) demonstrates that EDTA activation is dependent on metalloprotease activity, suggesting that EDTA treatment results in the exposure of the Notch metalloprotease cleavage site (in both the furin cleaved and uncleaved forms of Notch) and, hence, metalloprotease cleavage and shedding of the Notch ectodomain (Fig. 3A). EDTA was seen to act only at the cell surface (see Supp. Fig. S1, which is available online).

Figure 3.

The ligand inhibitory effect is a cis effect. A–C: Cell-based reporter assay with ethylenediaminetetraacetic acid (EDTA) -induced Notch signaling activation. A: The metalloprotease inhibitor galardin was used to establish that EDTA-induced Notch signaling activity is dependent on metalloprotease activity. B,C: Signaling properties of Notch with and without ligand co-expression; expression of either Delta or Serrate resulted in inhibition of Notch signaling in a cell autonomous manner. D,E: Co-expression of ligand prevents the loss of the Notch ectodomain. *nonspecific bands. Cells were treated with EDTA or Ca-containing phosphate buffered saline (PBS; as a control for the EDTA treatment). The supernatant containing any cleaved Notch extracellular domain was discarded and the cells were then lysed in RIPA buffer. The levels of extracellular Notch in the cell lysates were assessed by Western blot. Ligand co-expression inhibits loss of extracellular Notch. The relative NECD levels with reference to the loading control were quantified using the Quantity One software (Bio-Rad). In D, a decrease NECD relative levels of 30.6% is observed in S2N cells upon EDTA treatment and no decrease is detected on S2N,Dl cells. In E, a 9.8% decrease in NECD levels in detected upon EDTA treatment in S2N cells and no decrease is detected on S2N,Ser cells. In E, an enhanced image of the full-length Notch bands is provided above the Western Blot. The degree of decrease in Notch extracellular levels was variable in Western Blot analysis of different assays of Notch-ligand co-expression but a higher decrease in cells expressing Notch alone, upon EDTA treatment, was always observed when compared with cells transfected for Notch-ligand co-expression. Two different antibody batches were used in Figure 3D,E.

When activating Notch signaling with EDTA treatment we find that co-expression of Notch and its ligands prevents EDTA induced Notch signaling activity (Fig. 3B,C). This effect is not due to titration of mediators of the metallothionein (MT) promoter activity (Fig. S2). A reduction in cell surface Notch was not detected in our experiments (Fig. 4), although we note that using a different assay others have come to a different conclusion (Perez et al., 2005). Our results demonstrate that ligand expressed in cis inhibits Notch signaling and that the ligand induced inhibition observed in vivo could be due to a cell autonomous cis inhibitory effect.

Figure 4.

Notch cell surface levels are unaffected by Delta expression. A: S2N, S2N cells transfected with Dl expression construct and SL2 cells were fixed and stained for Extracellular Notch. These cells were then analyzed using flow cytometry for N staining. SL2 cells do not express N and so act as a negative control, demonstrating the range of autofluorescence of these cells. Any fluorescence detected above this level is due to detection of cell surface Notch. B: The percentage of Notch-positive cells in both S2N populations is low, but the percentage of cells positive for Notch and the level of Notch detected is unaffected by expression of Delta at levels that significantly inhibit Notch signaling activity. This result differs from the one reported by Perez et al. (2005) where Dl co-expression with Notch is seen to reduce cell surface Notch levels. The difference might result from the experiments done in Perez et al. (2005) being carried out in cells transiently transfected for Notch expression while the ones described in this manuscript regard assays performed in stable cell line of Notch-expressing cells. C: S2N and S2N cells transfected with Dl expression construct were fixed and stained for Extracellular Notch without permeabilization. Extracellular Notch is detected only at the cell surface, and the level of Extracellular Notch detected appears unaffected by expression of Delta as analyzed by FACS.

These observations also suggest that, when ligand is expressed at high levels, it could be inhibiting a step before the shedding of Notch ectodomain. To test this hypothesis, the levels of extracellular Notch were assessed in cells expressing either Notch or Notch and Delta before and after exposure to EDTA treatment (Fig. 3D). In the case of Notch-expressing cells, the levels of the extracellular domain of Notch detected upon EDTA treatment were lower than in untreated cells. Whereas for cells that co-express Notch and Delta, little reduction in extracellular Notch levels was apparent. Serrate behaves in analogous manner by protecting Notch from EDTA-induced ectodomain shedding (Fig. 3E). Furthermore, the ligand cis-inhibitory activity is likely to involve an interaction with Notch EGF repeats 10–12 (Supp. Figs. S3, S4), as when membrane tethered NECD (ECN) is expressed there is ectopic Notch signaling activation at the border of expression while with ECN lacking this region (ECN Δ10–12) this effect is lost (Figs. S3, S4). These dominant-negative constructs were seen to be localized at the plasma membrane region with a higher accumulation in the apical region (Fig. S5). One interpretation for how a dominant-negative construct can lead to signaling activation is that ECN binds ligand involved in mediating the cis-inhibitory effect and these cells are then responsive to the ligand expressed in neighboring wild-type cells.

These results demonstrate clearly for the first time that both Notch ligands can exert a cell autonomous cis inhibitory effect. Furthermore, this effect is mediated at least in part by preventing an activation step before Notch ecto-domain shedding and involves the ligand binding region of Notch EGF repeats 10–12.

Ax[M1] is Sensitive to cis-inhibition

Ax[M1] belongs to the Abruptex class of Notch mutations and presents an antimorphic phenotype. It has both features of Notch gain of function (e.g., reduced number of bristles) and loss of function because an increase in the dosage of Notch rescues the Ax phenotype (de Celis and Garcia-Bellido, 1994). One model put forward to explain these opposing phenotypes is that Ax molecules may have reduced inhibitory interactions with the ligands, which would be responsible for the gain of function phenotypes, while having lower signaling ability which would lead to the loss of function features (de Celis and Bray, 2000). Our observations show that Ax[M1] is equally sensitive to ligand cis inhibition while having reduced signaling ability compared with Notch in cell assays (Fig. 5). This does not agree with the in vivo gain of function features of Ax[M1], suggesting the existence of further uncharacterized factors contributing to inhibitory interactions, not addressed here, to which Ax[M1] might be less sensitive than Notch. The observed sensitivity of Ax[M1] toward cis-inhibition in cell culture goes in agreement with the results in Perez et al., 2005. In wing development, Notch signaling activity is restricted to the DV boundary region of the wing disc by ligand-mediated inhibition (Micchelli et al., 1997). Ax[M1] flies exhibit broader Notch signaling activity at the DV boundary region of the wing disc than wild-type flies, but signaling activity remains restricted to the DV boundary region (Perez et al., 2005), suggesting that Ax[M1] is susceptible to inhibitory regulation as seen for ligand cis-inhibition on Figure 5.

Figure 5.

Ax[M1] is sensitive to ligand cis inhibition. A,B: Cell-based reporter assays with ethylenediaminetetraacetic acid (EDTA) -induced Notch signaling activation shows that Ax[M1] is sensitive to both Ser and Dl cis-inhibition and that Ax[M1] signaling levels are lower than Notch signaling levels.

Role for Ligand Endocytosis in Ligand cis Inhibition

High levels of ligand cause a reduction in the levels of Notch not associated with signaling activation (Fig. 6, and Glittenberg et al., 2006). This would suggest that Notch might have enhanced endocytosis and degradation promoted by high ligand levels. Although high levels of ligand cause a reduction in Notch levels (Fig. 6), nonetheless, when producing clones mutant for Dl and Ser, the absence of ligand does not appear to affect Notch levels (Fig. 6G). The hypothesis that ligands might be affecting the trafficking and signaling properties of Notch has been explored for Serrate in Drosophila (Glittenberg et al., 2006). Here, we extend the analysis to Delta and use the cell culture system to look at the impact of Delta trafficking defects on ligand cis-inhibition. Notch ligand recycling is regulated by the monoubiquitin ligases Neuralized (Neur) and Mindbomb (Mib1) and is necessary for the ligand to functionally activate Notch signaling (Pavlopoulos et al., 2001; Itoh et al., 2003; Le Borgne and Schweisguth, 2003).

Figure 6.

High levels of ligands cause a signaling-independent reduction of Notch levels. A–C: High levels of Delta (red), in wild-type Oregon flies, colocalize with regions exhibiting a reduction in the total levels of Notch (green; α-Nicd). D: High levels of Serrate lead to a reduction in Notch levels (α-Nicd). E,F: Ectopic Notch signaling activity (Cut, red) with Serrate overexpression in the dpp domain. G–I:Drosophila wing imaginal discs with Ser/Dl MARCM clones marked by green fluorescent protein (GFP) expression (green). Wingless expression (red) marks Notch signaling, Notch levels detected with polyclonal α-Nicd (blue). There is no apparent variation in Notch levels within Ser/Dl MARCM clones except in regions with signaling activation.

S2 cells co-expressing Notch and ligand were treated with dsRNA against Mib1 and/or Neur (dsMib1 and/or dsNeur) to inhibit ligand endocytosis. The effects on ligand cis inhibition were assessed using reporter assays where Notch signaling activity was induced by treating the cells with EDTA. Inhibition of Delta endocytosis did not significantly decrease Delta inhibitory activity (Fig. 7). A similar behavior was observed for Serrate-mediated inhibition (data not shown). This could suggest that the main mechanism for ligand inhibition is the protection of Notch ectodomain shedding. However, other processes of internalization may take place for the endocytosis of the Notch-ligand complex so these results can only suggest that Neur/Mib1-mediated internalization should not play a role in the regulation of Notch signaling activity in cell culture.

Figure 7.

The role of ligand endocytosis in ligand cis inhibition. Treatment of the cells with dsRNA to Mib1, Neuralized or a combination of the two, causes a minute reduction in the inhibitory potential of Dl.


Interactions between Notch and its ligands can lead to activation or inactivation of signaling by Su(H)/CBF. While there is substantial evidence that activation results from a trans interaction between the ligands and Notch, the mechanism and topology for the inhibitory effect are less clear. Activation of Notch signaling at the edge of clones of cells lacking Delta and Serrate suggests that the effect is due to a cis inhibitory interaction between Notch and the ligands. However, given the context, this observation cannot rule out trans interactions between the ligands themselves to be responsible for inhibition. Here, we have used a cell culture system in which we can control and integrate both the signal sending and signal receiving cell to demonstrate that the inhibitory effect can take place as a cis effect, that acts mainly by inhibiting the shedding of the ecto-domain of Notch.

Other possible mechanisms would be inhibition of signaling by reduction of Notch levels caused by high levels of ligand or trafficking effects on Notch caused by the ligand. The latter model could be involved both in the reduction of Notch levels and in the protection of Notch ectodomain shedding. Under conditions of Serrate overexpression in the domain of Dpp expression, Ser reduces Notch levels in all the domain of expression (Fig. 6). However, this does not prevent Notch signaling activation on the ventral side region close to the DV boundary, showing that this effect is not enough to sustain efficient cis inhibition. The effect of ligand trafficking defects on ligand cis inhibition assessed using the cell culture system are minute, arguing against ligand trafficking having prominent role in cis inhibition.

A mechanism of ecto-domain shedding protection could rely, in principle, on a physical molecular interaction between extracellular Notch and ligand. Here, we show that a dominant-negative form of Notch (ECN) is able to modulate ligand inhibition, suggesting that the inhibitory effect might rely on interactions mediated by the extracellular region of Notch. More specifically that cis inhibition involves the Notch region of EGF repeats 10–12 because ECNΔ10–12 does not affect cis inhibition. This had already been hinted at in the work of Jakobsen et al. (1998) where ECN is reported to interfere with Dl signaling ability while ECNΔ10–12 is not. This highlights that the Notch EGF10–12 region could mediate an important interaction for cis inhibition, whether involving a cis or trans intracellular interaction is unclear. The report of Cordle et al. (2008), where structural docking models suggest that an interaction between Notch EGF repeats 10–12 and the DSL domains can occur in cis would favor the first hypothesis.

Altogether, our results support that ligand inhibition can take place as a cis-inhibitory effect acting by preventing a step before Notch ecto-domain shedding and involves an interaction mediated by Notch EGF repeats 10–12.

Experimental Procedures

Fly Stocks

(Oregon); (w;P{UAS-NECNΔ10–122A), (w; P{UAS-NECN2T5;MKRS/TM6B), described in Lawrence et al. (2000), (w;ptc-Gal4,UAS-GFP;Gbe-Su(H)-lacZ) described in Furriols and Bray (2001), (w;;UAS-Ser) described in Speicher et al. (1994). Dl and Ser mutant clones described in Micchelli et al. (1997).

Plasmids and Constructs

pMT-Fng from Perez et al. (2005). pMT-Dl from DGRC (Drosophila Genome Resource Centre). pMT-Ser a gift from François Schweisguth. Notch reporter constructs from Perez et al. (2005).

Notch Signaling Reporter Assays

In all reporter assays 24 hr before induction of Notch expression, S2N cells (a stable cell line from the DGRC with Notch expression under the control of a metallothionein promoter) were transiently transfected with 0.025 μg of the reporter plasmid (NRE), 0.025 μg of polIII-Renilla Luciferase, and the amount of pMT-Fng, pMT-Dl, and pMT-Ser used in transfections was of 0.1 μg per well of cells. DNA amounts were adjusted to a constant value with an empty vector. All assays were carried out in 96-well plates with 3 replicate wells for each condition in which parallel transfections and treatments were performed. Galardin (Biomol International) was used at 10 μM. Luciferase activity was assessed using the Dual-Glo system from Promega. The data presented are from one experiment representative of three independent experiments. The values presented are averages of three replicates normalized to Renilla luciferase to allow corrections regarding transfection efficiency and presented as the fold change in activity compared with the level of ligand-induced Notch activity. The amount of pMT promoter was seen not to have a titrating inhibitory effect in signaling behavior (See Fig. S2).

Co-culture reporter assays.

Signaling in S2N cells was activated by co-culture with an equal amount of ligand-expressing cells (either S2-Dl cells (DGRC) or S2 cells transiently transfected with 0.1 μg of pMT-Dl or of pMT-Ser) for 48 hr. During co-culture, the cells were incubated in media containing 0.7 mM CuSO4.

EDTA-activation reporter assays.

Signaling in S2N cells (stable cell line from the DGRC) was activated by treating the cells with 2 mM EDTA, at 25°C for 30 min. Notch expression in S2N cells was induced for 36 hr with 0.7 mM CuSO4 before EDTA treatment.

RNAi assays.

Cells treated with dsRNA (double-stranded RNA) were submitted to two hits of dsRNA. They were first plated in 24-well plates, treated with a 1:4 mix of dsRNA and EC Buffer (effectene kit from Qiagen), incubated for 48h before replating in 96-well plates for transfection with a second hit of dsRNA and the reporter constructs.

Aggregation assay.

Cells were prepared with a final concentration of 2 × 106 cells/ml and mixed in an orbital shaker at 150 rpm, 25°C for 15 min.


Wing imaginal disc stainings as described in Hayward et al. (2005). For cell imaging and fluorescence-activated cell sorting (FACS) analysis, cells were fixed in 4% paraformaldehyde in phosphate buffered saline (PBS), blocked with a 2% fetal calf serum, 2% bovine serum albumin (BSA) in PBS solution, stained with antibodies diluted in blocking solution and washed with PBS. For FACS analysis cells were resuspended in PBS, filtered with a 300-μm mesh, and assessed in a Dako CyAn ADP analyzer. Antibodies used from DSHB (Developmental Studies Hybridoma Bank) are as follows: anti-NECD (C458.2), anti-DE-Cadherin (DCAD2), anti-extracellular Dl (C594.9B), anti-Cut (2B10), and anti-Wg (4D4). Anti–β-galactosidase, from Cappel and Sheep anti-Nicd from the Martinez Arias laboratory.

Western Blotting

Notch and ligand expression induced with 0.7 mM CuSO4. Protein samples of cell lysates were separated using a 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blot analysis made for Notch (anti-NECD C458.2H, DSHB) and nuclear laminin DMO (DSHB).


We thank F. Schweisguth, S.M. Cohen, the DGRC, and the DSHB for providing constructs and reagents. We thank S. Muñoz-Descalzo, C. Balayo, and T. Langdon for help with experiments and M. Milán for comments and advice. U.M.F. received a PhD fellowship from the Fundação para a Ciência e Tecnologia. This work was supported by the Wellcome Trust.