Characterization of Notch1 antibodies that inhibit signaling of both normal and mutated Notch1 receptors

Abstract

Background: Notch receptors normally play a key role in guiding a variety of cell fate decisions during development and differentiation of metazoan organisms. On the other hand, dysregulation of Notch1 signaling is associated with many different types of cancer as well as tumor angiogenesis, making Notch1 a potential therapeutic target.

Principal findings: Here we report the in vitro activities of inhibitory Notch1 monoclonal antibodies derived from cell-based and solid-phase screening of a phage display library. Two classes of antibodies were found, one directed against the EGF-repeat region that encompasses the ligand-binding domain (LBD), and the second directed against the activation switch of the receptor, the Notch negative regulatory region (NRR). The antibodies are selective for Notch1, inhibiting Jag2-dependent signaling by Notch1 but not by Notch 2 and 3 in reporter gene assays, with EC(50) values as low as 5+/-3 nM and 0.13+/-0.09 nM for the LBD and NRR antibodies, respectively, and fail to recognize Notch4. While more potent, NRR antibodies are incomplete antagonists of Notch1 signaling. The antagonistic activity of LBD, but not NRR, antibodies is strongly dependent on the activating ligand. Both LBD and NRR antibodies bind to Notch1 on human tumor cell lines and inhibit the expression of sentinel Notch target genes, including HES1, HES5, and DTX1. NRR antibodies also strongly inhibit ligand-independent signaling in heterologous cells transiently expressing Notch1 receptors with diverse NRR "class I" point mutations, the most common type of mutation found in human T-cell acute lymphoblastic leukemia (T-ALL). In contrast, NRR antibodies failed to antagonize Notch1 receptors bearing rare "class II" or "class III" mutations, in which amino acid insertions generate a duplicated or constitutively sensitive metalloprotease cleavage site. Signaling in T-ALL cell lines bearing class I mutations is partially refractory to inhibitory antibodies as compared to cell-penetrating gamma-secretase inhibitors.

Conclusions/significance: Antibodies that compete with Notch1 ligand binding or that bind to the negative regulatory region can act as potent inhibitors of Notch1 signaling. These antibodies may have clinical utility for conditions in which inhibition of signaling by wild-type Notch1 is desired, but are likely to be of limited value for treatment of T-ALLs associated with aberrant Notch1 activation.

Figure 1. Ligand-competition by Notch1 antibodies.

A panel of Notch1 antibodies was tested in a ligand competition assay (DELFIA) for binding to EGF repeats 1–13 of Notch1. This assay measures inhibition of binding of Eu-labeled Notch1 ECD-Fc fusion protein to immobilized DLL4. The ECD-Fc fusion comprises EGF repeats 1–13 (Ala19-Gln526) which includes the ligand binding domain (EGF repeats 11–13) but lacks the NRR. Human IgG and competing soluble DLL4 were used as negative and positive controls, respectively. The data were normalized with respect to the “no blocker” controls and curve fitted using a fixed 100% plateau, shared slopes and variable base lines. Error bars represent the standard deviation from triplicate values.

Figure 2. Inhibition of Jag2-dependent Notch signaling by Notch1 antibodies.

Notch1 reporter activity was measured in co-culture assays with T-REX-U2OS Notch1-Gal4 reporter cells and Flp-In-3T3 cells expressing human Jag2. Representative examples are shown for each of the LBD and NRR antibodies. Error bars represent the standard deviation from triplicate values. (A) Inhibition of Notch1-Gal4 signaling by LBD antibodies WC613 (open circles), WC133 (closed circles), WC179 (open triangles), and WC97 (closed triangles). (B) Inhibition of Notch1-Gal4 signaling by NRR antibodies WC75 (closed circles) and WC629 (open circles). (C) The Notch isoform specificity of Notch1 antibodies was tested in co-culture assays with T-REX-U2OS Notch-Gal4 reporter cells (human Notch1, 2, and 3) and Flp-In-3T3 cells expressing human Jag2. NRR and LBD antibodies were used at a fixed antibody concentration of 167 nM. Reporter cell lines used: hNotch1-Gal4 (black bars), hNotch2-Gal4 (lined bars), hNotch3-Gal4 (cross-hatched bars). The activity of a UAS-luciferase reporter transiently expressed in the T-REX-U2OS Notch-Gal4 cells was normalized to untreated controls. IgG isotype controls are shown. Error bars represent standard deviation. (D) The species specificity of Notch1 antibodies was tested in co-culture assays with T-REX-U2OS cells expressing wild-type mouse Notch1 and Flp-In-3T3 cells expressing human Jag2. NRR and LBD antibodies were used at a fixed concentration of 167 nM. The activity of a 4xCSL-luciferase reporter transiently expressed in the T-REX-U2OS cells was normalized against the non-specific IgG control. Error bars represent standard deviation.

Figure 3. Ligand-dependence of Notch1 inhibition by LBD and NRR antibodies.

Notch1 signaling in T-REX-U2OS Notch1-Gal4 cells was stimulated by co-culture with Flp-In-3T3 cells expressing Jag2 (closed circles), Jag1 (open circles), DLL1 (closed triangles), or DLL4 (open triangles). The activity of a UAS-luciferase reporter transiently expressed in the T-REX-U2OS Notch-Gal4 cells was normalized to untreated controls. Representative examples of Notch1-Gal4 reporter inhibition are shown for LBD antibody WC613 (A) and for NRR antibody WC629 (B).

Figure 4. Notch1 antibodies inhibit Notch target gene expression in cancer cell lines.

The ability of antibodies (20 µg/ml) or GSI (5 µM) to inhibit Notch1 target gene expression (HES1, HES5, DTX1) was analyzed by quantitative real time PCR (qRT-PCR) of mRNA extracted from LS-1034 or TALL-1 cancer cell lines co-cultured with Flp-In 3T3-Jag2 cells for 22 h at 37°C. qRT-PCR was performed in triplicate with the Stratagene Mx3005P (Agilent Technologies, BioCrest Manufacturing, Cedar Creek, TX). Values were normalized on the basis of GAPDH mRNA expression. Gene expression (% mRNA remaining) normalized to Jag2-dependent signal (100%) from at least four experiments is represented (error bars indicate error standard, *p<0.05).

Figure 5. LBD and NRR antibodies bind to cancer cell lines.

Notch1 surface expression in LS-1034, TALL-1, BxPC3, and Colo_205 cancer cell lines was examined by flow cytometry (FACSCalibur, BD BioSciences, San Jose, CA) after staining of cells with the LBD antibody WC613 (green line) or NRR antibody WC75 (red line), and R-PE-conjugated anti-human IgG antibody (Jackson ImmunoResearch,, Inc., West Grove, PA). An irrelevant human IgG isotype antibody (hIgG) (blue line) was used as negative control.

Figure 6. NRR antibody WC75 inhibits ligand-independent signaling by Notch1 receptors harboring T-ALL-associated mutations.

T-REX-U2OS cells were transiently co-transfected with a 4xCSL-luciferase reporter construct and full-length Notch1 cDNA constructs encoding mutated receptors that exhibit ligand-independent activation of Notch signaling: (A) class I point mutations; (B) insertional mutations p12 (class II), Jurkat (class III) and VSV (juxtamembrane). Activity of luciferase after treatment with the NRR WC75 Notch1 antibody (10 µg/ml) was measured in cell lysates using the Bright-Glo assay kit (Promega). Reporter activity induced by the wt-Notch1 construct was used as baseline control. Error bars represent standard deviation.

Figure 7. Comparison of the effects of NRR antibody WC75 or GSI on Notch1-dependent target gene expression in DND-41 cells.

A) Following treatment of DND-41 cells for 72 hr with either DMSO (0.08%), GSI (compound E, 1 µM), control nonspecific human antibody (10 µg/ml), or NRR antibody WC75 (10 µg/ml), the expression levels of 20 genes that define a T-ALL-specific Notch1 signature were measured with a ligation-mediated amplification/fluorescent bead-based detection system. Each column represents an independent experimental replicate. Dark red indicates high gene expression and dark blue low gene expression. Notch marker gene expression is depicted as a ratio of the expression of the marker gene relative to the mean of four control genes. The summed score combines expression ratios by summing them with a sign determined by the expected direction of regulation as determined from the positive controls (GSI-treated). The weighted summed score metric is a variant of the summed score metric that combines expression ratios by summing them with a weight and sign determined by the signal-to-noise ratio of the positive control (GSI-treated) and negative controls (DMSO-treated). B) DTX1 and c-MYC expression levels assessed by qRT-PCR following 3 days of treatment of DND-41 cells with control nonspecific human antibody (IgG, 10 µg/ml), WC75 NRR-N1 antibody (10 µg/ml), or GSI (compound E, 1 µM). Expression of each transcript was determined in triplicate, and each experiment was repeated three times.

Figure 8. NRR WC75 Notch1 antibody proliferation of T-ALL cells.

Proliferation of DND-41 (L1594P/D1610V NRR-N1 mutations) and KOPT-K1 (L1601P NRR-N1 mutation) cells (2.5×10³ cells/well) was assessed in a 384-well format for up to 5 days in the presence of either the WC75 antibody (10 µg/ml) or the gamma secretase inhibitor compound E (GSI, 100 nM). Growth inhibition was measured by CellTiter-Glo® (Promega).