Discovery at the VICB
New Key to Aberrant Signaling in Cancer
By: Carol A. Rouzer, VICB Communications
Published: June 14, 2016
The N-terminal region of the Notch intracellular domain triggers degradation of the protein and suppression of Notch-dependent signaling.
The Notch signaling pathway is an intercellular communications system that plays an important role in the development of multicellular organisms. Aberrant Notch signaling can also contribute to the malignant behavior of some cancers. Thus, understanding the mechanisms that control the pathway may identify targets for novel cancer therapeutics. To this end, Vanderbilt Institute of Chemical Biology member Ethan Lee and his collaborator Stacey Huppert (Cincinnati Children’s Hospital Medical Center) investigated a previously unknown region in the Notch protein that leads to its degradation. They show that mutations in this structure lead to enhanced Notch signaling, and that such activation mutations can be found in human cancer [M. R. Broadus, et al., Cell Rep., 15, 1-10].
Notch signaling takes place between two adjacent cells. The signaling cell must express a DSL (Delta/Serrate/Lag-2) type 1 transmembrane protein that binds to the Notch receptor on the recipient cell. Interaction of the two proteins leads to proteolytic cleavage of Notch, releasing the Notch intracellular domain (NICD) into the cytoplasm. NICD then travels to the nucleus where it forms a complex with other proteins, including CSL (CBF1/RBPjk/Su(H)/Lag-1), MAML (Mastermind-Like), and additional coactivators. Interaction of this complex with the promoter regions of target genes results in increased transcription (Figure 1). Termination of Notch-dependent signaling occurs following ubiquitination of the PEST domain found at the C-terminus of the NICD (Figure 2). Proteasomal degradation of the ubiquitinated NICD follows.
FIGURE 1. Overview of the Notch signaling pathway. A DSL type 1 ligand (in this case, Delta) on the signaling cell interacts with the Notch receptor on the recipient cell. Ligand binding leads to cleavage by two proteases (ADAM10 or TACE and γ-secretase), releasing the Notch intracellular domain (NICD). In the nucleus, the NICD forms a complex with other proteins, including CSL and MAM, while also triggering the release of co-repressors (Co-R). The result is a complex that promotes transcription of Notch pathway target genes. Figure adapted by permission from Macmillan Publishers Ltd. from M. R. Broadus, et al., Cell Rep., 15, 1-10. Copyright 2016 Macmillan Publishers Ltd.
FIGURE 2. Domain structures of the four human NICD paralogs. Figure reproduced under the Creative Commons Attribution 4.0 Unported License from M. R. Broadus, et al., Cell Rep., 15, 1-10. Copyright 2016, M. R. Broadus, et al.
The Lee and Huppert labs’ interest in NICD degradation began with studies of the fate of the four human NICD paralogs (hNICD1-4) in Xenopus laevus egg extracts. They had previously used this model system for the study of β-catenin degradation. The investigators found that [35S]-labeled hNICD1 was rapidly degraded by the Xenopus extract, whereas the other three hNICD paralogs were quite stable. The proteasome inhibitor MG132 blocked hNICD1 degradation, but addition of high concentrations of β-catenin had no effect. Fusion of hNICD1 to luciferase caused rapid degradation of the protein in Xenopus extracts. The degradation, which could be easily monitored by the loss of luminescence in the sample, did not occur upon fusion of any of the other three hNICD paralogs. Fusion of the four hNICD paralogs to the MYC protein led to degradation of only hNICD1-MYC in Xenopus extract, but all four fusion proteins were degraded when expressed in cultured human cells. Together, the results suggested that the hNICD1 protein is degraded in Xenopus extracts by a mechanism that is proteasome-dependent but distinct from that of β-catenin. This mechanism is not shared by the other hNICD paralogs; however all four paralogs are degraded by a different system that is functional in the intact cell.
To identify the location of the degron - the structure in hNICD1 that is responsible for its degradation in Xenopus extracts - the investigators constructed multiple fusion proteins containing different regions of hNICD1 and hNICD2. They tested these proteins for their susceptibility to degradation in the presence of Xenopus extract. The results demonstrated that 35 amino acids at the N-terminus of hNICD1 (the N1-box) were responsible for its instability in the extracts. Comparison of the hNICD1 and hNICD2 sequences revealed key differences in the N1-box. Based on this comparison, the investigators constructed L1755Δ, S1757M, and Q1763K mutant hNICD1 proteins. In each case, an N1-box residue of hNICD1 was converted to the corresponding residue in hNICD2. All three of these mutants exhibited increased stability in Xenopus extracts as did a triple mutant, hNICD1LSQ, that was altered at all three sites. An additional interesting finding was that deletion of the first ten N1-box amino acids to form the hNICD1NTΔ10 mutant led to increased stability, despite the fact that destabilization of hNICD2 required transfer of a longer, 35 amino acid stretch from the N-terminus of hNICD1.
Further experiments confirmed the importance of the N-terminus of hNICD1 to its stability. Fusion of the first 50 residues of hNICD1 to luciferase destabilized the protein in Xenopus extracts, whereas the unaltered luciferase protein was stable. In contrast, fusion of the first 50 residues of hNICD1 containing N1-box mutations did not destabilize luciferase. When hNICD1, hNICD1NTΔ10, and hNICD1LSQ were expressed in HEK293 cells, the mutant proteins exhibited higher steady-state levels and increased transcriptional activity than the wild-type protein. Consistently, green fluorescent protein (GFP) exhibited higher stability than GFP fused to full-length hNICD1 or the first 50 amino acids of hNICD1 when the proteins were expressed in HEK293 cells (Figure 3). Degradation of the fused GFP proteins occurred in both the nucleus and cytoplasm.
FIGURE 3. Disappearance of fluorescence from green fluorescent protein (GFP), GFP fused to hNICD1, or GFP fused to the first 50 amino acids of hNICD1 expressed in HEK293 cells. Figure reproduced under the Creative Commons Attribution 4.0 Unported License from M. R. Broadus, et al., Cell Rep., 15, 1-10. Copyright 2016, M. R. Broadus, et al.
After confirming that the murine NICD1 protein exhibited the same instability as the human protein in Xenopus extracts, the investigators explored the effects of the N1-box on the full-length mouse Notch receptor (mNotch) in zebrafish embryos. Notch signaling is important for somite formation during development. The investigators showed that injection of the mRNA for wild-type mNotch led to a disruption of somite formation due to excessive Notch signaling (Figure 4). When they injected the mRNA for mNotch bearing the N1-box triple mutation, they observed a 2-fold greater effect on somite formation than was observed with the wild-type mNotch mRNA.
FIGURE 4.Somite formation is a critical process during development. The developing somites, appearing like round ring-like segments, can be seen in the control zebrafish embryo to the left. Injection of the mRNA for mNotch1 results in a disruption in somite formation as can be seen in the embryo to the right. Figure reproduced under the Creative Commons Attribution 4.0 Unported License from M. R. Broadus, et al.,Cell Rep., 15, 1-10. Copyright 2016, M. R. Broadus, et al.
Notch signaling also suppresses neural differentiation during development. The investigators confirmed this by injecting mNotch1 mRNA into transgenic zebrafish embryos that express GFP under the control of the neurogenin 1 promoter. Neurogenin 1 is a transcription factor that promotes neural development. Transgenic embryos injected with mNotch1 mRNA exhibited decreased GFP expression, denoting inhibition of neurogenin 1 promoter activity. These effects were more pronounced when the mRNA for mNotch1 bearing the triple N1-box mutation was injected. In contrast, injection of mNotch 1 mRNA into transgenic mice that express red fluorescent protein (RFP) under the control of the promoter for her4, a Notch target gene, resulted in increased RFP expression. These results confirmed the activity of the mNotch protein in the injected embryos. Once gain, the effects were intensified when the mRNA for the mNotch1 N1-box triple mutant (mNotch1LSQ) was used (Figure 5).
FIGURE 5. (Top) Transgenic zebrafish embryos that express GFP under the control of the neurogenin 1 promoter. From left to right are a control embryo, an embryo injected with the mRNA for wild-type mNotch1, and an embryo injected with the mRNA for mNotch1 bearing the inactivating triple mutation in the N1-Box. (Bottom) The treatments are the same as described above, but in this case the transgenic embryos express RFP under the control of the Notch-responsive her4 promoter. Figure reproduced under the Creative Commons Attribution 4.0 Unported License from M. R. Broadus, et al., Cell Rep., 15, 1-10. Copyright 2016, M. R. Broadus, et al.
A well-known mechanism for NICD degradation is by the proteasomal pathway following ubiquitination of the C-terminal PEST domain. The investigators showed that N1-box-mediated NICD1 degradation is independent of that pathway by mutating the responsible sequences in the PEST domain, and by expressing components of the relevant ubiquitin ligases or their dominant negative mutants. These manipulations had little effect on N1-box-mediated degradation.
CSL, one of the proteins that forms a complex with NICD in the cell nucleus, binds to NICD at its RAM domain, which overlaps the N1-box (Figure 2). This led the investigators to hypothesize that binding CSL to NICD1 might interfere with N1-box-mediated degradation. In support of that hypothesis, they found that adding recombinant CSL to Xenopus extracts inhibited degradation of wild-type hNICD1 but not a mutant hNICD1 that cannot bind CSL. This result suggests that CSL protects NICD1 from degradation triggered by the N1-box. In contrast, CSL promotes degradation triggered by the PEST domain. Consequently, mutation of hNICD1 in both the N1-box and the CSL-binding domain results in greater stability than mutation of the N1-box alone, as the combination of the two mutations suppresses degradation by both pathways. Furthermore, mutation in both the PEST and CSL-binding domains results in less stability than mutation in the PEST domain alone. In both cases, PEST-dependent degradation will be blocked, but the CSL-binding domain mutation removes the protection against N1-box-mediated degradation that is afforded by CSL.
Together the results suggest that the N1-box of NICD1 promotes degradation of NICD1 under some circumstances. As the investigators first noted that hNICD1 degradation in Xenopus extracts was suppressed by a protease inhibitor, they hypothesized that a heretofore unidentified ubiquitin ligase is involved in the degradation process. They further speculate that N1-box-mediated degradation might occur rapidly in the cytosol before NICD1 can be stabilized by CSL binding in the nucleus. It is possible that the purpose of the N1-box is to dampen stochastic flux in the levels of NICD1. Whether or not this is true, the investigators were interested in determining if the N1-box plays a role in aberrant Notch signaling that occurs in some cancers. Scanning of the COSMIC database of human tumors, they identified two mutations, R1758S and S1776C, in the hNICD1 N1-box of some tumor samples. When these mutant NICD1 proteins were expressed in HEK293 cells, they exhibited higher steady-state levels and transcriptional activity than wild-type hNICD1. Transgenic zebrafish embryos encoding either the R1758S or S1776C Notch1 mutation (generated by CRISPER-Cas9) also exhibited greater Notch signaling activity than those encoding the wild-type Notch1. This was illustrated by suppression of somite formation and alterations in the expression of genes under the control of Notch-dependent signaling that were observed in the mutant-expressing embryos (Figure 6). These results suggest that failure of N1-box-mediated degradation may lead to excessive Notch-depending signaling in human cancer. Further studies will show if the N1-box can be exploited as a target for novel cancer therapeutic agents.
FIGURE 6. (Top) Transgenic zebrafish embryos that express GFP under the control of the neurogenin 1 promoter. (Bottom) Transgenic embryos that express RFP under the control of the her4 promoter. In each case, the embryos expressed wild-type Notch1 (uninjected) or Notch1 carrying a silent mutation (Control) or the R1758S or S1776C mutations identified in human tumors. Figure reproduced under the Creative Commons Attribution 4.0 Unported License from M. R. Broadus, et al., Cell Rep., 15, 1-10. Copyright 2016, M. R. Broadus, et al.
View Cell Reports article: Identification of a Paralog-Specific Notch1 Intracellular Domain Degron