19th-banner-rev.gif

Book of Abstracts: Albany 2007

category image Albany 2007
Conversation 15
June 19-23 2007

Synchrotron-Based Structural Proteomics of Vesicle Transport

Synchrotron X-ray protein crystallography plays critically important roles in structural investigations on macromolecules. A progress report will be presented on our systems approach for developing and operating synchrotron X-ray protein crystallography beam lines at the Photon Factory, Tsukuba. The beam lines provide state-of-the-art, user-friendly experimental environment including crystal exchange robots based on the SSRL system. Current projects include two new insertion device beam lines, one optimized for micro focus experiments with low energy SAD capabilities and another on highthroughput data acquisition for pharmaceutical industry. Our synchrotron developments are closely coupled with the efforts in structural proteomics on intracellular transport and post-translational modification of proteins. Our targets are mostly eukaryotic proteins involved in vesicle transport between the ER, the Golgi apparatus and endosomes/lysosomes, post-translational modification of newly synthesized proteins, exocytosis, endocytosis, ubiquitin-dependent protein sorting, and autophagy.

GGAs (Golgi-localizing, γ-adaptin ear domain homology, ARF-binding) are monomeric clathrin adaptor proteins that are involved in selective transport of lysosomal cargo receptors from the TGN to endosomes. GGAs consist of four functional regions: an N-terminal VHS (Vps27/Hrs/Stam) domain, a GAT (GGA and Tom1) domain, a hinge region and a C-terminal GAE (γ-adaptin ear) domain. The N-terminal VHS domains of GGAs form complexes with the cytoplasmic domains of sorting receptors by recognizing acidic-cluster-dileucine (ACLL) sequences. The X-ray structures of the GGA1 VHS domain and those in complex with the C-terminal ACLL motifs of cation-independent mannose 6-phosphate receptor (CI-MPR) [1] and BACE (β-secretase) [2] have shown that the recognition of cargo or cargo-receptor is accomplished by the combination of hydrophobic and electron static interactions as well as shape complementarity. The GAT domain binds Arf1 in its GTP form, which is responsible for tethering GGAs onto the TGN membrane. The crystal structures of the GAT domain and its complex with Arf-GTP have shown that the N-terminal half of the GAT domain is partially unfolded but adopts a helix-loop-helix structure with hydrophobic surface to interact with Switches I and II as well as the interswitch β-sheet [3,4]. The C-terminal GAE domain forms an Ig-domain [5] and binds various FXXF motifs regulating the protein-protein interactions of GGA.

The C-terminal part of the GAT domain can recognize ubiquitin molecules on two distinct sites [6]. This ubiquitin recognition appears to be involved in sorting mono-ubiquitinated cell surface receptors such as EGFR destined for degradation by the ESCRT pathway. A very similar double sided binding was observed for the GAT domain of Tom1 [7]. The first step in the ESCRT complex pathway is Hrs, a human homolog of yeast Vps27p. Its ubiquitin interacting motif (UIM) is a short alpha helical stretch which had been believed to bind one ubiquitin. The crystallographic and surface plasmon resonance analyses revealed that it can actually bind two ubiquitin molecules on either side of the helix with equal affinity [8]. Very similar interaction surfaces for ubiquitin recognition are achieved by forming almost identical binding residues by a shift of two residues along the UIM sequence. RNAi experiments on mutants indicate that this double binding is physiologically relevant and a revisit of UIM sequences revealed the existence of other double sided UIMs. The last example is the ubiquitin recognition by the GLUE domain of Eap45, a human homolog of Vps36p, of ESCRT-II complex. It has been a puzzle how Eap45 can recognize ubiquitinated receptors because it lacks the NZF domain which in case of Vps36 is responsible for ubiquitin binding. A collaborative work with H. Stenmark?s group has shown that the GLUE domain of Eap45 can not only bind ubiquitin but also PI(3,4,5)P3 [9]. The subsequent crystallographic study unraveled how it can recognize both ubiquitin, using its canonical Ile44 surface, and phosphoinositide simultaneously [10].

References and Footnotes
  1. T. Shiba et al. Nature 415, 937, 2002
  2. T. Shiba et al. Traffic 5, 437, 2004
  3. T. Shiba et al. Nature Structural Biology 10, 386, 2003
  4. M. Kawasaki et al., Current Opinion in Structural Biology 15, 681, 2005
  5. T. Nogi et al. Nature Structural Biology, 9, 527, 2002
  6. M. Kawasaki et al., Genes Cells 10, 639, 2005
  7. M. Akutsu et al., FEBS Letters 579, 5385, 2005
  8. S. Hirano et al., Nature Structural and Molecular Biology, 13, 272, 2006
  9. T. Slagsvold et al., J. Biol. Chem., 280, 19600?19606, 2005
  10. S. Hirano et al., Nature Structural and Molecular Biology, 13, 1031, 2006

Soichi Wakatsuki

Structural Biology Research Center, Photon Factory, Institute of Materials Structure Science, High Energy Accelerator Research Organization (KEK), Tsukuba, Ibaraki 305-0801, Japan

Tel & Fax: +81-29-864-5631,
Email: soichi.wakatsuki@kek.jp