Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Aug 1;36(15):2204-2215.
doi: 10.15252/embj.201796632. Epub 2017 Jun 1.

Structural and functional dissection of the interplay between lipid and Notch binding by human Notch ligands

Affiliations

Structural and functional dissection of the interplay between lipid and Notch binding by human Notch ligands

Richard J Suckling et al. EMBO J. .

Abstract

Recent data have expanded our understanding of Notch signalling by identifying a C2 domain at the N-terminus of Notch ligands, which has both lipid- and receptor-binding properties. We present novel structures of human ligands Jagged2 and Delta-like4 and human Notch2, together with functional assays, which suggest that ligand-mediated coupling of membrane recognition and Notch binding is likely to be critical in establishing the optimal context for Notch signalling. Comparisons between the Jagged and Delta family show a huge diversity in the structures of the loops at the apex of the C2 domain implicated in membrane recognition and Jagged1 missense mutations, which affect these loops and are associated with extrahepatic biliary atresia, lead to a loss of membrane recognition, but do not alter Notch binding. Taken together, these data suggest that C2 domain binding to membranes is an important element in tuning ligand-dependent Notch signalling in different physiological contexts.

Keywords: Delta‐like; Jagged; Notch; lipid‐binding; specificity.

PubMed Disclaimer

Figures

Figure 1
Figure 1. New structures of human Notch ligands and receptors
  1. A

    Structures of human Delta‐like4 (N‐EGF3) and human Jagged2 (N‐EGF2 and N‐EGF3). Two crystal forms of Jagged2 (N‐EGF2) (green) and one crystal form of Jagged2 (N‐EGF3) (dark green) were solved, including a total of seven crystallographically independent copies of N‐EGF2, and 1 copy of N‐EGF3 (N.b. not all of EGF3 is visible in the electron density). None of the EGF domains are of the calcium binding type. All of these copies have been superposed on each other by alignment of the DSL domain, showing some flexibility in the hinge region between the C2 and DSL domains, and further flexibility in the hinge regions between the EGF domains.

  2. B, C

    Superposition of all of the known human Notch ligand structures (Jagged1, PDB ID = 4CC1 (light green) (Chillakuri et al, 2013), Delta‐like1, PDB ID = 4XBM (light blue) (Kershaw et al, 2015), Delta‐like4 (blue), Jagged2 (green)), and all of the Delta‐like4 variant‐Notch1 complex structures (red) across the DSL domain shows that these structures appear to fall into two groups (C). One group has an overall globally bent arrangement (C(ii)), which includes all of the Delta‐like4 molecules bound to Notch1 EGF11‐13, and a second group with a straighter arrangement (C(i)). Binding of ligands to a Notch receptor in a native context likely requires the ligands to be in the straighter arrangement, as the bent arrangement is incompatible with binding to EGF10 (Weisshuhn et al, 2016).

  3. D

    Crystal structure of human Notch2 (EGF11‐13), and comparison with modified human Notch1 EGF11‐13 (PDB ID = 4D0E) (Taylor et al, 2014) Phe‐478 and Phe‐516 are highlighted in Notch2 as these appear to be shielded from the solvent by the glycans on Ser‐462 and Ser‐500, respectively. Throughout all figures, where space does not allow full naming, ligands and Notch receptors are identified by initial letter and number, for example Notch1—N1 or Jagged1—J1.

Figure 2
Figure 2. Comparison of the N‐terminal C2 domains in the different Notch ligands
  1. The loops between strands 1 and 2 (calcium‐binding region 1 (CBR1)), and between 5 and 6 (CBR3) of the Notch ligand C2 domains differ in length and conformation.

  2. The loop between strands 3 and 4 (CBR2) is also different.

  3. The Jagged1 and Jagged2 C2 domains bind calcium ions (shown in red), whereas Delta‐like4 and Delta‐like1 do not; the aspartates involved in calcium binding are not conserved in Delta‐like4 and Delta‐like1. Both Jagged1 and Jagged2 contain N‐glycosylation sites in the loop between strands 5 and 6. The α(1,6)‐linked fucose on the Asn‐153 N‐glycosylation site in Jagged2 packs against Trp‐151 side chain (shown). All of these differences at the putative lipid‐binding site likely reflects the different Notch ligand lipid‐binding specificity.

Figure EV1
Figure EV1. Structural comparison of human Delta‐like4 ligand with rat Delta‐like4
  1. A–C

    Superposition of human Delta‐like4 (N‐EGF3) (A) with rat Delta‐like4SLP(N‐EGF1) (grey) –Notch1 (EGF11‐13) (pink) complex structure (PDB ID = 4XL1) (Luca et al, 2015) across the DSL domain, highlights the greater angle between the C2 and DSL domains in the apo ligand structure (B). There are also differences in the loops between strands 3 and 4 (CBR2), and strands 5 and 6 (CBR3) in the C2 domain between rat and human Delta‐like4 (C). The ConSurf Server (consurf.tau.ac.il) was used to create a surface representation of the evolutionary conservation of residues in Delta‐like4 based on an alignment from zebrafish to humans using ClustalO (data not shown). This highlights that CBR2 and CBR3 are the most variable regions of the structure (C).

Figure 3
Figure 3. Comparison of the residues involved in complex formation in the different Notch receptors and Notch ligands
  1. A, B

    Comparison of the residues involved in complex formation (Luca et al, 2015) in Notch1 and Notch2 (A), and in the different human ligands (B), at site 1 (C2:EGF12) and site 2 (DSL:EGF11). N.b. The complex structure between Delta‐like4 and Notch1 was of rat Delta‐like4 (shown in panel A); human Delta‐like4 is shown in panel (B). Conservation of the residues involved in complex formation is indicated by background colour, highlighting the high conservation of the ligand‐binding sites in Notch1 and Notch2. There are a few residues in the receptor‐binding sites of the ligands that are absolutely conserved, with site 2 (DSL‐EGF11) being more variable than site 1 (C2‐EGF12).

Figure 4
Figure 4. Functional coupling between liposome and Notch binding of Notch ligands
  1. All canonical human Notch ligands (Jagged1, Jagged2, Delta‐like1 and Delta‐like4) (N‐EGF3) bind to both human Notch1 (i) and Notch2 (ii) EGF11‐13 to a similar extent as assessed by plate assay. Notch ligands were bound to nickel‐coated plates before biotinylated pre‐clustered Notch was added with NeutrAvidin‐conjugated HRP. Binding is shown at high (H, 300 nM) and low (L, 20 nM) protein concentrations with Delta‐like3, which does not bind Notch, acting as a negative control at these concentrations. All components were purified from insect cells, and three independent experiments were performed with all points in duplicate in each.

  2. All canonical human Notch ligands (Jagged1 (J1), Jagged2 (J2), Delta‐like1 (D1) and Delta‐like4 (D4)) (N‐EGF3) bound to fluorescently labelled liposomes (PC/PE/PS)—Delta‐like3 (N‐EGF1), Notch1 (EGF11‐13) and Notch2 (EGF11‐13) did not bind. Four independent experiments with a minimum of 36 replicates were performed.

  3. At low concentrations of ligand, that is, below concentrations where liposome binding can be observed, addition of Notch1 EGF11‐13 stimulated binding of Jagged1 N‐EGF3 to liposomes. This effect could be abolished by an antibody against the Notch‐binding DSL domain of Jagged1 (α‐DSL), or by substitution of residues critical for ligand binding to Notch (Leu468Ala) (LBR). Liposome binding was also abolished by substitutions that directly perturb the putative lipid‐binding site in the C2 domain of Jagged1 (Asp140Ala/Asp144Ala, C2(1) and Del1Del2Asp140Ala, C2(2)) (Chillakuri et al, 2013). Fifteen independent experiments with a minimum of 18 replicates were performed.

  4. Addition of Notch1 EGF11‐13 stimulated binding of Jagged1 and Delta‐like4 N‐EGF3 to liposomes, with Notch2 having a similar effect on Jagged2, but neither had an effect on Delta‐like1 or Delta‐like3 in terms of liposome binding. Five independent experiments with 50 replicates were performed. Data were analysed with Prism 6 or 7 (GraphPad, San Diego, CA, USA).

Data information: Comparisons between two groups were performed with a two‐tailed unpaired t‐test. Statistical differences among various groups were assessed with ordinary one‐way ANOVA by comparison to the mean of a control column. Values are presented together with the mean ± SD. **P < 0.01; ***P < 0.001; ***P < 0.0001.
Figure EV2
Figure EV2. Human Delta‐like4 and Jagged1 exhibit different binding preferences for ganglioside‐ or sphingomyelin‐rich liposomes
  1. A, B

    Delta‐like4 shows preferential binding to ganglioside‐rich liposomes compared to Jagged1 (A) whilst the converse is true for sphingomyelin‐rich liposomes (B). Seventeen and nine independent experiments were performed (respectively) with a minimum of twenty replicates in each. Data were analysed with Prism 6 or 7 (GraphPad, San Diego, CA, USA). Comparisons between two groups were performed with a two‐tailed unpaired t‐test. Statistical differences among various groups were assessed with ordinary one‐way ANOVA by comparison to the mean of a control column. Values are presented together with the mean ± SD. ****P < 0.0001.

Figure 5
Figure 5. Disease‐causing substitutions affecting C2 loops selectively alter membrane but not Notch binding
  1. A

    C2 domain of Jagged1 showing position of extrahepatic biliary atresia (EHBA) causing substitutions Asn53Asp and Lys65Met in loop regions, the position of Arg203Lys in DSL Notch‐binding site also associated with EHBA is shown.

  2. B–E

    C2 EHBA variants reduce Notch1 (data not shown) and Notch2 activation (B) C2 EHBA variants do not affect Notch binding, unlike Arg203Lys (C) but liposome binding is reduced and the Notch boosting effect is lost (D, E). At least three independent experiments with ten replicates were performed for all assays. Data were analysed with Prism 6 or 7 (GraphPad, San Diego, CA, USA). Comparisons between two groups were performed with a two‐tailed unpaired t‐test. Statistical differences among various groups were assessed with ordinary one‐way ANOVA by comparison to the mean of a control column. Values are presented together with the mean ± SD. ***P < 0.001; ***P < 0.0001.

Comment in

  • The lipid-binding side of Notch ligands.
    Shilo BZ, Sprinzak D. Shilo BZ, et al. EMBO J. 2017 Aug 1;36(15):2182-2183. doi: 10.15252/embj.201797621. Epub 2017 Jul 14. EMBO J. 2017. PMID: 28710261 Free PMC article.

Similar articles

Cited by

References

    1. Afonine PV, Grosse‐Kunstleve RW, Echols N, Headd JJ, Moriarty NW, Mustyakimov M, Terwilliger TC, Urzhumtsev A, Zwart PH, Adams PD (2012) Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallogr D Biol Crystallogr 68: 352–367 - PMC - PubMed
    1. Aricescu AR, Lu W, Jones EY (2006) A time‐ and cost‐efficient system for high‐level protein production in mammalian cells. Acta Crystallogr D Biol Crystallogr 62: 1243–1250 - PubMed
    1. Bray SJ (2016) Notch signalling in context. Nat Rev Mol Cell Biol 17: 722–735 - PubMed
    1. Chillakuri CR, Sheppard D, Ilagan MX, Holt LR, Abbott F, Liang S, Kopan R, Handford PA, Lea SM (2013) Structural analysis uncovers lipid‐binding properties of Notch ligands. Cell Rep 5: 861–867 - PMC - PubMed
    1. Corbalan‐Garcia S, Gomez‐Fernandez JC (2014) Signaling through C2 domains: more than one lipid target. Biochim Biophys Acta 1838: 1536–1547 - PubMed

Publication types

MeSH terms

LinkOut - more resources