Adding structural information to the von Hippel–Lindau (VHL) tumor
suppressor interaction network
E. Leonardi
a,b, A. Murgia
a, S.C.E. Tosatto
b,* aDepartment of Biology, University of Padua, 35131 Padua, Italy b
Department of Pediatrics, University of Padua, 35131 Padua, Italy
a r t i c l e
i n f o
Article history:
Received 17 August 2009 Revised 21 October 2009 Accepted 26 October 2009 Available online 28 October 2009 Edited by Paul Bertone
Keywords:
Von Hippel–Lindau Hypoxia inducible factor Structural bioinformatics Linear motif
Protein interaction
a b s t r a c t
The von Hippel–Lindau (VHL) tumor suppressor gene is a protein interaction hub, controlling numerous genes implicated in tumor progression. Here we focus on structural aspects of protein interactions for a list of 35 experimentally verified protein VHL (pVHL) interactors. Using structural information and computational analysis we have located three distinct interaction interfaces (A, B, and C). Interface B is the most versatile, recognizing a refined linear motif present in 17 otherwise non-related proteins. It has been possible to distinguish compatible and exclusive interactions by relating pVHL function to interaction interfaces and subcellular localization. A novel hypothesis is presented regarding the possible function of the N-terminus as an inhibitor of pVHL function. Ó2009 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.
1. Introduction
Von Hippel–Lindau (VHL) syndrome is a dominantly inherited familial cancer syndrome with variable expression and age-depen-dent penetrance, characterized by a predisposition to develop var-ious tumors, including among others hemangioblastomas, clear cell renal carcinomas and pheochromocytomas. Predisposition to develop this variety of tumors is linked to germ line inactivation of the VHL tumor suppressor gene, coding for VHL protein (pVHL). Pathology development in VHL disease occurs after somatic inacti-vation of the remaining wild-type allele in a susceptible cell[1]. Certain classes of pVHL mutations confer different site-specific risks of cancer, suggesting pVHL to have multiple tissue-specific tumor suppressor functions[2]. How pVHL mutations cause differ-ent disease phenotypes remains incompletely understood.
It is widely accepted that pVHL functions as target recognition component of the E3 ubiquitin ligase complex targeting the
a
sub-units of hypoxia-inducible factors (HIF) for oxygen-dependent pro-teolytic degradation[3–5]. pVHL inactivation leads to stabilization of HIF1a
and HIF2a
and activation of their downstream target genes implicated in angiogenesis, cell growth, and metabolism, e.g. vascular endothelial growth factor (VEGF), platelet derived growth factor (PDGFB), transforming growth factor (TGF) anderythropoietin[6]. However, distinct pVHL-containing complexes indicate pVHL involvement in different signaling pathways, includ-ing microtubule dynamics, primary cilium maintenance, cell prolif-eration, neuronal apoptosis and extracellular matrix deposition. Several studies have recently investigated the numerous pVHL functions, providing insight into pVHL mediated signaling net-works involved in tumor formation (for recent reviews see
[1,7,8]). Experimental determination of protein interactions pro-vides growing data about new pVHL partners and, partially, the protein regions involved. However, a large amount of data remains to be confirmed.
We address the problem by combining information from exper-imental data with structural analysis[9–12]. Defining the pVHL interaction interfaces was the primary goal. As pVHL has more interaction partners than available surface, some interactions must be mutually exclusive, while others interact simultaneously[13– 15]. Structural information also offers the possibility to distinguish domain–domain from peptide–domain interactions[16]. Hypothe-ses are drawn regarding the expected interaction type in each inter-face and interaction partners are classified. Augmented with functional information we contribute to define functional sub-net-works existing in different time and space conditions, adding a dy-namic component to pVHL function. In contrast to a similar analysis recently proposed for p53[17], where the authors made large scale interaction predictions, here we base our hypotheses on a subset of experimentally validated interaction data from the literature.
0014-5793/$36.00Ó2009 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2009.10.070
*Corresponding author. Fax: +39 049 827 6260.
E-mail address:[email protected](S.C.E. Tosatto).
2. Materials and methods
pVHL interacting proteins and their binding regions (in terms of amino acid sequence) were extracted from the relevant literature, with sequences from UniProt[18]. Structures were downloaded from the PDB[19]where available. Fold recognition was performed using MANIFOLD[20]to identify a structural template with suffi-cient sequence similarity in cases with no experimental structure. FlexServ[21]was used to estimate pVHL flexibility. To map pVHL evolutionary sequence conservation, 35 closely related sequences identified with PSI-BLAST[22](default parameters) were realigned with ClustalW[23]. The N-terminal region of pVHL was analyzed to gain more insight on its function. SPRITZ[24]was used to predict disorder, the consensus method for secondary structure prediction
[25]and REPETITA[26]to validate the N-terminal repeat. The mul-tiple sequence alignment was drawn using ESPript[27]and struc-tures were visualized in PyMol (URL:http://www.pymol.org/).
Regions of pVHL sequence where several proteins interact were inferred to contain a binding interface. These were mapped on the pVHL structure considering all residues with high DSSP surface accessibility[28]and with high structural and functional conserva-tion scores, derived from Consurf[29]and Crescendo[30], respec-tively. For each class of interactors, the Pfam[31]and CATH[32]
classification codes were compared to search for common domains or architectures through visual inspection and structural align-ment. A potential linear motif which could mediate the interaction with pVHL was searched on the protein sequences using Jalview
[33]starting from the degenerate HIF1
a
peptide with matches re-corded and visually inspected. Different peptide definitions yielded comparable results (data not shown) and PepSite[34]was used to score the pVHL surface for candidate interaction motifs. For each interactor, the GO[35]molecular function and cellular localization allowed us to determine functions mediated by each interface re-gion with different space and time patterns.3. Results and discussion
3.1. Data retrieval
Fig. 1 shows an overview of the sequence features of pVHL. Information about sequence, domain architecture, structure and
function of interactors was collected in a list containing 35 proteins with experimentally determined pVHL interaction regions ( Supple-mentary data S1). Four molecules (CK1, CK2, GSK3 and NEDD8) are involved in a post-translational modifications of specific amino acids. The boundaries of pVHL sequence mediating the interaction has been refined for 26 interactors but remains undefined for fibro-nectin, COL4A2, HIF1AN, BCL2L11 and VHLAK. The region binding pVHL has been experimentally determined for 18 proteins and re-mains unknown for 17 interactors. Structural details are known for Elongin C (EloC), Elongin B (EloB) and HIF1
a
which are co-crystal-lized with pVHL [3–5]. For seven interacting proteins, the pVHL binding region is a particular domain: p53 DNA binding domain[36], HuR RRM1[37], Nur77 hormone receptor[38], Sp1 zinc fin-gers [39], PRKCD Kinase[40], PRKCZ PB1[41]and VHLAK KRAB-A [42] domains. The structure is known only for p53, Sp1 and PRKCZ PB1 domains, while fold recognition allowed identification of a template with sufficient sequence identity for the others. The large pVHL binding regions of VDU1/USP33, VDU2/USP20, Jade-1 and aPKCk contain distinct structural domains separated by disordered region. HIF1AN and BCL2A11 interact with pVHL through a linker region. HIF2
a
and RPB1 interact through a func-tional motif [43]. The pVHL binding region for fibronectin and COL4A2 might be located on their N-terminal domain since the interaction occurs at the intracellular site of the endoplasmic retic-ulum[6].3.2. Interface definition
Fig. 2summarizes the pVHL structural features with its
a
- andb-domains. Comparing the results obtained by Consurf (Fig. 2A) and Crescendo (Fig. 2B) it is noted that surface regions containing relevant functional residues are located in two interfaces known to interact with HIF1
a
and EloC (1LM8VH and 1LM8CV). In addition a third interface appears on the back of pVHL. Protein structures fre-quently interact with various domains using different surface areas[13]. Mapping the putative protein interacting regions on the pVHL structure for interactors listed in Supplementary data S1, the majority of interactions occur in three specific areas of pVHL that we define interfaces A, B and C (seeFig. 2C). No structure is avail-able for the pVHL N-terminus containing putative phosphorylation
Fig. 1.pVHL sequence features. Domain architecture and protein sequence alignment of pVHL. The pVHL domains are shown on the top part. Secondary structure and homologous sequences of pVHL are shown in the center.acc: accessibility level from DSSP (black = high and white = low). The protein-interaction regions are represented as lines at the bottom of the sequence alignment.
sites. In the following, we describe the characteristics of each inter-face separately, which are also summarized inFig. 3.
3.3. Interface A – processing
It has been demonstrated that interaction of EloB and EloC with pVHL depends on the binding of EloC to a 10-amino acid
a
-helical sequence motif xLxxxCxxx[AILV], referred to as the BC box [44]. pVHL also interacts with Cullin-2 at the specific Cul2box, located C-terminal to the BC box, forming the E3 ubiquitin ligase complex responsible for recognition and recruitment of tar-get proteins to be degraded by the 26S proteasome [45]. Four proteins, other than EloB/C and Cul2, have been experimentally determined to interact with the pVHL
a
-domain through a spe-cific domain. The p53, Nur77, HuR and VBP1 interacting regions overlap with those of EloC. This supports the idea that the pVHLa
-domain mediates domain–domain interactions. The EloC, p53, HuR and Nur77 domains interacting with VHL have a similar two-layer sandwich architecture, with different CATH classifica-tions. Their structures are mainly composed ofb-sheets with ana
-helix that might mediate the interaction in analogy to EloC. The interaction interface between pVHL and EloC is composed of three pVHLa
-helices (H1, H2 and H3) and the H4 helix of EloC, forming hydrophobic contacts[3]. Flexibility analysis of the pVHL structure indicates thea
-domain as flexible (Fig. 2C) and suggests a conformational change upon interactor binding as described in[46].
EloC, p53, HuR, Nur77 and VBP1 are mutually exclusive and compete for binding to the
a
-domain (Fig. 3). p53, HuR and Nur77 are prevalently expressed in the nucleus, and EloC in the cytoplasm. Thea
-domain also contains the neddylation site, K159, which appears to be necessary for fibronectin binding. When neddylated, pVHL cannot interact with Cul2 and stabilizes fibro-nectin, showing processing functions depending on localization and physiological cell status[47].3.4. Interface B – substrate recognition
The pVHLb-domain (interface B), containing the well-studied HIF1
a
binding site, was found to be essential for binding 13 differ-ent proteins with variations termed B1 and B2. The interaction site for seven proteins was shown to contain the HIF1a
binding site. The region essential for VDU1 and VDU2 binding overlaps only par-tially and for HIF1AN and BCL2A11, the specific pVHLb-domain interaction site is unknown. Only 10 proteins have experimentally determined pVHL interacting regions, with a prevalence of unstructured linkers.For Sp1 the interaction is known to occur through one of three Zn finger domains and pVHL interactions with other metalloproteins have been hypothesized [48]. Among the listed Zn finger domain containing proteins, only the aPKCk binding region overlaps with its Zn finger domain. For VDU1, VDU2, PRKCZ and RPB1 the experimentally determined interacting re-gion does not contain a Zn finger. Another domain–domain interaction has been hypothesized by Zhou et al. [49] which have shown that the Jade-1 PHD module, structurally similar to the Zn finger domain, is most important for VHL-mediated stabilization. In our list, Jade-1 is the only protein containing a PHD domain. Another protein Kinase C delta (PRKCD) interacts with pVHL through the catalytic domain and there is no correla-tion between degradacorrela-tion of PRKCD and the presence of active pVHL[40].
Proteins of unknown pVHL interacting region are likely to use disordered sequence stretches. The HIF1
a
binding sites, named ODD domains, are disordered and contain a hydroxylated proline. It has been demonstrated that RPB1 interacts with pVHL through a proline[43]. Sharing a conserved motif containing this proline, it seems interface B mediates interactions occurring between the pVHL b-domain and different peptides. The known interface B interactors were systematically searched for linear motifs resem-bling the HIF1a
ODD sequence and structurally validated with Pep-Site. The results (shown inFig. 4) point to the presence of a proline box followed by a hydrophobic box with the consensus pattern [LIV]xPx(6,9)uxu, whereu is prevalently a hydrophobic residue, in a likely disordered region. This pattern agrees with the locationFig. 2.pVHL structure overview. (A) Cartoon representations of the pVHL structure (PDB code 1LM8 chain V) coloured blue (N-terminal) to red (C-terminal). The structure, shown in standard view (front) and after a 180°rotation around thez-axis (back), is maintained throughout the following panels. (B) Surface of pVHL projecting phylogenetic conservation by Consurf. The Consurf prediction is presented with magenta shading for highly conserved residues and cyan shading showing for variable residues. (C) Surface of pVHL where Crescendo residues clusters are coloured: lime green, cluster A; orange, cluster B; yellow, cluster C; red, cluster D; blue, cluster E; violet, cluster F. (D) Three pVHL interfaces mapped on structure surface of pVHL: green, interface A; violet, interface B; yellow, interface C. (E) Analysis of pVHL flexibility of each residue (x-axis) in terms of relative displacement (y-axis, in Angstroms) in the monomer (continuous line) and in complex with EloC (dashed line).
of known pVHL interactions for HIF1
a
[4]and RPB1[50]. With few exceptions (PRKCZ, Jade-1 and perhaps VDU1, VDU2) the predicted motifs fit well into the respective structures. Furthermore, the mo-tif is present also in the two proteins (HIF1AN and BCL2L11) for which a specificb-domain interaction site were not well defined. Although experimental binding assays will be necessary to verify the prediction, our results summarize well the plasticity of the pVHL B interface.As many interface B interactors result to be mutually exclusive, the selection of interaction occurs in different ways. Domain–do-main interactions are more specific or have higher binding affini-ties than domain–peptide interactions. The different interactors have specific localization, e.g. RPB1 is nuclear while HIF1
a
is nucle-ar and cytoplasmic. Another selection criterion may be the time at which some proteins are expressed with higher concentration. It appears that pVHL through interface B can bind proteins which will be ubiquitinylated and then degraded, yet it can also bind other proteins which will be stabilized, e.g. Jade-1[49], BCL2L11[51]and microtubules (MT)[52]. Interactions with other proteins play a role in transcriptional regulation[6], e.g. Sp1, HuR, HIF1AN and perhaps VHLaK for which an interaction pVHL interface remains unassigned.
3.5. Interface C – localization
The back view of pVHLb-domain (Fig. 1) presents two regions for known interactors. TBP1 interacts with pVHL residues 136– 154 containing ab-sheet and a linker between the VHL
a
- andb -domains[53]. It is difficult to predict how TBP1 can interact with this extended region. Residues 114–138 of pVHL interact with eEF1A [54] and contain the nuclear export signal (NES), DxGxxDxxL[55], in a loop between twob-strands forming a polar interface between pVHLa
- andb-domains. Surprisingly, the NES motif is part of the polar interface partially located in front of thea
-domain. This may be explained by the flexibility ofa
-domain which likely changes conformation and increases accessibility for other proteins. During transport from nucleus to cytoplasm, pVHL is captured by a nuclear export receptor component. In the cyto-plasm, TBP1 (component of the 26S proteasome) interacts with pVHL in complex with EloC, anchoring it to the proteasome where the ubiquitinylated substrates are degraded[53]. This interface ap-pears to determine pVHL localization. Interestingly, the areas inter-acting with TBP1 and eEF1A forming interface C are coded by exon 2 of the VHL gene. Moreover amino acids 114–177 appear to be re-quired for perinuclear (ER) pVHL localization[56].Fig. 3.Overview of pVHL interactors. Pictures represent pVHL interacting partner in four different cellular compartments: nucleus, cytoplasm, endoplasmic reticulum and membrane periphery. At the center of each picture there is a cartoon representation of the crystal structure of pVHL (PDB code 1LM8, chain V) highlighting the protein interface as circles: interface A in cyan, interface B in blue, and interface C in green. When the structure of a pVHL interaction partner is known, it is represented as cartoon, with their acronym and PDB code. A percentage sequence identity is shown if the structure refers to a homologous protein. Light blue circles are used if the interaction region is unknown. A red border indicates proteins for which information about pVHL interaction regions exists but it is located in an unstructured region or in domains for which no structure is available. A grey circle indicates proteins interacting with an unknown pVHL region. KIFAP3 and CK2 interact with the VHL N-terminal sequence which is not included in the crystal structure. Phosphorylation (red balls) and neddylation sites (yellow ball) are also shown.
3.6. pVHL dynamics
pVHL contains several post-translational modification sites, determining its dynamic properties. Reversible pVHL neddylation distinguishes between HIF-related functions and stabilization of the extracellular matrix[47]. Together with the localization signals in interface C this determines to a great extent the precise function pVHL may be carrying out at any given point in time.
A more elusive pVHL regulatory mechanism is phosphorylation. Several phosphorylation sites present in or near the pVHL N-termi-nus are responsible for proper fibronectin deposition [57] and microtubule dynamics[58]. In addition, the N-terminus is entirely missing in other mammalian orthologs, suggesting a human spe-cific mechanism. Eight tandem repeats with a GxEEx pattern are contained in the N-terminus (seeFig. 1). REPETITA periodicity
anal-ysis strongly suggests that the repeat is a solenoid and covers the entire pVHL N-terminus, with two degenerate units at the N- and one at the C-terminal flanks of the GxEEx pattern. All known six-residue solenoid repeats formb-helix structures[59], where each repeat corresponds to a shortb-strand and a turn typically cen-tered on a glycine. Whether the pVHL N-terminus can form the sameb-solenoid remains however unclear due to the occurrence of adjacent charged residues which may cause electrostatic repulsion.
In any case, the putative phosphorylation sites located in the re-peat region are likely to cause a conformational switch, creating a pattern of three charged residues known to be highly correlated with intrinsic disorder[60]. The serines located close to two pro-lines (seeFig. 2) have prompted us to use Pepsite to determine hypothetical pVHL–peptide interactions. Intriguingly, the results
Fig. 4.pVHL interface B linear motifs. The putative linear motifs defining pVHL interactions at interface B are shown. (A) Multiple sequence alignment of identified linear motifs with the sequence identifier and positions on the left side. The right side contains information on the disorder (Y, yes; N, no; P, partial) and secondary structure (H, helix; E, extended/beta; and C, coil/loop) and two Pepsite predictions. PepsiteP-values, estimating the probability of achieving a similar result by chance, are complemented with the presence of the central proline in the predicted motif. Extremely low (bold) and high (italics and underlined)P-values are highlighted. The two regions corresponding to the Pepsite predictions are drawn above and two motif determinant boxes are drawn below the sequence alignment. In (B) the pVHL structure (PDB code 1LM8, chain V) is shown oriented as inFig. 1with the bound HIF1apeptide in grey spheres. The hydroxylated proline residue is shown in red and the two motif determinant boxes are shown in darker grey. A typical Pepsite prediction (BCL2L11, PepsiteP-value 0.0001) is shown in (C). Note the similar binding site with proximity to the hydroxylated proline location.
place the interaction site on interface B in correspondence with the HIF1
a
peptide (P-value = 0.024). This opens the possibility that hu-man pVHL may have evolved a signaling mechanism to deactivate binding of specific interactors through phosphorylation and con-formational rearrangement, e.g. fibronectin[57]. Experiments will be necessary to verify this hypothesis.4. Conclusions
We have presented a characterization of pVHL which attempts to summarize the known experimental interactions and address them from a structural point of view (seeFig. 3). The modular nat-ure of pVHL becomes apparent from the subdivision in three inter-action interfaces corresponding to processing, substrate recognition and localization. These highlight various protein inter-action types, namely domain–domain (interface A) and domain– peptide (interface B), with interface C being less clear. Structural characterization of the putative interface B interaction peptides yielded both a complete list of hypothetical interaction motifs and the intriguing possibility for the pVHL N-terminus to auto-in-hibit substrate recognition after phosphorylation. Our findings, rationalizing pVHL function at the molecular level through exper-imental data from the literature, can serve as a gold standard for the analysis of other putative interaction partners. They can be useful to design and interpret specific experimental interaction as-says for the 200 or more remaining putative pVHL interaction part-ners. Addressing these in different time and space points will provide the ultimate validation for this complex protein.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the online version, atdoi:10.1016/j.febslet.2009.10.070.
References
[1] Kaelin, W.G. (2007) Von Hippel–Lindau disease. Annu. Rev. Pathol. 2, 145–173. [2] Ong, K.R., Woodward, E.R., Killick, P., Lim, C., Macdonald, F. and Maher, E.R. (2007) Genotype–phenotype correlations in von Hippel–Lindau disease. Hum. Mutat. 28, 143–149.
[3] Stebbins, C.E., Kaelin Jr., W.G. and Pavletich, N.P. (1999) Structure of the VHL– ElonginC–ElonginB complex: implications for VHL tumor suppressor function. Science 284, 455–461.
[4] Hon, W.C. et al. (2002) Structural basis for the recognition of hydroxyproline in HIF-1 alpha by pVHL. Nature 417, 975–978.
[5] Min, J.H., Yang, H., Ivan, M., Gertler, F., Kaelin Jr., W.G. and Pavletich, N.P. (2002) Structure of an HIF-1alpha–pVHL complex: hydroxyproline recognition in signaling. Science 296, 1886–1889.
[6] Kaelin Jr., W.G. (2008) The von Hippel–Lindau tumour suppressor protein: O2 sensing and cancer. Nat. Rev. Cancer 8, 865–873.
[7] Frew, I.J. and Krek, W. (2007) Multitasking by pVHL in tumour suppression. Curr. Opin. Cell Biol. 19, 685–690.
[8] Frew, I.J. and Krek, W. (2008) pVHL: a multipurpose adaptor protein. Sci. Signal. 1, pe30.
[9] Kiel, C., Beltrao, P. and Serrano, L. (2008) Analyzing protein interaction networks using structural information. Annu. Rev. Biochem. 77, 415–441. [10] Beltrao, P., Kiel, C. and Serrano, L. (2007) Structures in systems biology. Curr.
Opin. Struct. Biol. 17, 378–384.
[11] Campagna, A., Serrano, L. and Kiel, C. (2008) Shaping dots and lines: adding modularity into protein interaction networks using structural information. FEBS Lett. 582, 1231–1236.
[12] Tuncbag, N., Kar, G., Keskin, O., Gursoy, A. and Nussinov, R. (2009) A survey of available tools and web servers for analysis of protein–protein interactions and interfaces. Brief Bioinform. 10, 217–232.
[13] Kim, P.M., Lu, L.J., Xia, Y. and Gerstein, M.B. (2006) Relating three-dimensional structures to protein networks provides evolutionary insights. Science 314, 1938–1941.
[14] Devos, D. and Russell, R.B. (2007) A more complete, complexed and structured interactome. Curr. Opin. Struct. Biol. 17, 370–377.
[15] Tuncbag, N., Gursoy, A., Guney, E., Nussinov, R. and Keskin, O. (2008) Architectures and functional coverage of protein–protein interfaces. J. Mol. Biol. 381, 785–802.
[16] Aloy, P. and Russell, R.B. (2006) Structural systems biology: modelling protein interactions. Nat. Rev. Mol. Cell Biol. 7, 188–197.
[17] Tuncbag, N., Kar, G., Gursoy, A., Keskin, O. and Nussinov, R. (2009) Towards inferring time dimensionality in protein–protein interaction networks by integrating structures: the p53 example. Mol. Biosyst. [Epub ahead of print]. [18] Bairoch, A. et al. (2005) The Universal Protein Resource (UniProt). Nucleic
Acids Res. 33, D154–D159.
[19] Bourne, P.E. et al. (2004) The distribution and query systems of the RCSB Protein Data Bank. Nucleic Acids Res. 32 Database Issue, D223–D225. [20] Bindewald, E., Cestaro, A., Hesser, J., Heiler, M. and Tosatto, S.C. (2003)
MANIFOLD: protein fold recognition based on secondary structure, sequence similarity and enzyme classification. Protein Eng. 16, 785–789.
[21] Camps, J. et al. (2009) FlexServ: an integrated tool for the analysis of protein flexibility. Bioinformatics 25, 1709–1710.
[22] Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W. and Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402.
[23] Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673–4680.
[24] Vullo, A., Bortolami, O., Pollastri, G. and Tosatto, S.C. (2006) Spritz: a server for the prediction of intrinsically disordered regions in protein sequences using kernel machines. Nucleic Acids Res. 34, W164–W168.
[25] Albrecht, M., Tosatto, S.C., Lengauer, T. and Valle, G. (2003) Simple consensus procedures are effective and sufficient in secondary structure prediction. Protein Eng. 16, 459–462.
[26] Marsella, L., Sirocco, F., Trovato, A., Seno, F. and Tosatto, S.C. (2009) REPETITA: detection and discrimination of the periodicity of protein solenoid repeats by discrete Fourier transform. Bioinformatics 25, i289–i295.
[27] Gouet, P., Courcelle, E., Stuart, D.I. and Metoz, F. (1999) ESPript: analysis of multiple sequence alignments in PostScript. Bioinformatics 15, 305–308. [28] Kabsch, W. and Sander, C. (1983) Dictionary of protein secondary structure:
pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22, 2577–2637.
[29] Landau, M., Mayrose, I., Rosenberg, Y., Glaser, F., Martz, E., Pupko, T. and Ben-Tal, N. (2005) ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures. Nucleic Acids Res. 33, W299–W302. [30] Chelliah, V., Chen, L., Blundell, T.L. and Lovell, S.C. (2004) Distinguishing structural and functional restraints in evolution in order to identify interaction sites. J. Mol. Biol. 342, 1487–1504.
[31] Bateman, A. et al. (2004) The Pfam protein families database. Nucleic Acids Res. 32, D138–D141.
[32] Pearl, F.M. et al. (2003) The CATH database: an extended protein family resource for structural and functional genomics. Nucleic Acids Res. 31, 452– 455.
[33] Waterhouse, A.M., Procter, J.B., Martin, D.M., Clamp, M. and Barton, G.J. (2009) Jalview Version 2—a multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189–1191.
[34] Petsalaki, E., Stark, A., Garcia-Urdiales, E. and Russell, R.B. (2009) Accurate prediction of peptide binding sites on protein surfaces. PLoS Comput. Biol. 5, e1000335.
[35] Gene Ontology Consortium (2006). The Gene Ontology (GO) project in 2006. Nucleic Acids Res. 34, D322–D3226.
[36] Roe, J.S., Kim, H., Lee, S.M., Kim, S.T., Cho, E.J. and Youn, H.D. (2006) P53 stabilization and transactivation by a von Hippel–Lindau protein. Mol. Cell 22, 395–405.
[37] Datta, K., Mondal, S., Sinha, S., Li, J., Wang, E., Knebelmann, B., Karumanchi, S.A. and Mukhopadhyay, D. (2005) Role of elongin-binding domain of von Hippel Lindau gene product on HuR-mediated VPF/VEGF mRNA stability in renal cell carcinoma. Oncogene 24, 7850–7858.
[38] Kim, B.Y., Kim, H., Cho, E.J. and Youn, H.D. (2008) Nur77 upregulates HIF-alpha by inhibiting pVHL-mediated degradation. Exp. Mol. Med. 40, 71–83. [39] Pal, S., Claffey, K.P., Cohen, H.T. and Mukhopadhyay, D. (1998) Activation of
Sp1-mediated vascular permeability factor/vascular endothelial growth factor transcription requires specific interaction with protein kinase C zeta. J. Biol. Chem. 273, 26277–26280.
[40] Iturrioz, X., Durgan, J., Calleja, V., Larijani, B., Okuda, H., Whelan, R. and Parker, P.J. (2006) The von Hippel–Lindau tumour-suppressor protein interaction with protein kinase Cdelta. Biochem. J. 397, 109–120.
[41] Iturrioz, X. and Parker, P.J. (2007) PKCzetaII is a target for degradation through the tumour suppressor protein pVHL. FEBS Lett. 581, 1397–1402.
[42] Li, Z., Wang, D., Na, X., Schoen, S.R., Messing, E.M. and Wu, G. (2003) The VHL protein recruits a novel KRAB-A domain protein to repress HIF-1alpha transcriptional activity. EMBO J. 22, 1857–1867.
[43] Mikhaylova, O. et al. (2008) The von Hippel–Lindau tumor suppressor protein and Egl-9-Type proline hydroxylases regulate the large subunit of RNA polymerase II in response to oxidative stress. Mol. Cell Biol. 28, 2701–2717. [44] Kibel, A., Iliopoulos, O., DeCaprio, J.A. and Kaelin Jr., W.G. (1995) Binding of the
von Hippel–Lindau tumor suppressor protein to Elongin B and C. Science 269, 1444–1446.
[45] Kamura, T., Maenaka, K., Kotoshiba, S., Matsumoto, M., Kohda, D., Conaway, R.C., Conaway, J.W. and Nakayama, K.I. (2004) VHL-box and SOCS-box domains determine binding specificity for Cul2-Rbx1 and Cul5-Rbx2 modules of ubiquitin ligases. Genes Dev. 18, 3055–3065.
[46] Liu, J. and Nussinov, R. (2008) Allosteric effects in the marginally stable von Hippel–Lindau tumor suppressor protein and allostery-based rescue mutant design. Proc. Natl. Acad. Sci. USA 105, 901–906.
[47] Russell, R.C. and Ohh, M. (2008) NEDD8 acts as a ‘molecular switch’ defining the functional selectivity of VHL. EMBO Rep. 9, 486–491.
[48] Cohen, H.T. et al. (1999) An important von Hippel–Lindau tumor suppressor domain mediates Sp1-binding and self-association. Biochem. Biophys. Res. Commun. 266, 43–50.
[49] Zhou, M.I., Wang, H., Foy, R.L., Ross, J.J. and Cohen, H.T. (2004) Tumor suppressor von Hippel–Lindau (VHL) stabilization of Jade-1 protein occurs through plant homeodomains and is VHL mutation dependent. Cancer Res. 64, 1278–1286.
[50] Kuznetsova, A.V. et al. (2003) Von Hippel–Lindau protein binds hyperphosphorylated large subunit of RNA polymerase II through a proline hydroxylation motif and targets it for ubiquitination. Proc. Natl. Acad. Sci. USA 100, 2706–2711.
[51] Guo, Y., Schoell, M.C. and Freeman, R.S. (2009) The von Hippel–Lindau protein sensitizes renal carcinoma cells to apoptotic stimuli through stabilization of BIM(EL). Oncogene 28, 1864–1874.
[52] Lolkema, M.P., Mehra, N., Jorna, A.S., van Beest, M., Giles, R.H. and Voest, E.E. (2004) The von Hippel–Lindau tumor suppressor protein influences microtubule dynamics at the cell periphery. Exp. Cell Res. 301, 139–146. [53] Corn, P.G., McDonald 3rd, E.R., Herman, J.G. and El-Deiry, W.S. (2003)
Tat-binding protein-1, a component of the 26S proteasome, contributes to the E3 ubiquitin ligase function of the von Hippel–Lindau protein. Nat. Genet. 35, 229–237.
[54] Khacho, M., Mekhail, K., Pilon-Larose, K., Pause, A., Cote, J. and Lee, S. (2008) EEF1A is a novel component of the mammalian nuclear protein export machinery. Mol. Biol. Cell 19, 5296–5308.
[55] Khacho, M., Mekhail, K., Pilon-Larose, K., Payette, J. and Lee, S. (2008) Cancer-causing mutations in a novel transcription-dependent nuclear export motif of VHL abrogate oxygen-dependent degradation of hypoxia-inducible factor. Mol. Cell Biol. 28, 302–314.
[56] Schoenfeld, A.R., Davidowitz, E.J. and Burk, R.D. (2001) Endoplasmic reticulum/cytosolic localization of von Hippel–Lindau gene products is mediated by a 64-amino acid region. Int. J. Cancer 91, 457–467.
[57] Lolkema, M.P., Gervais, M.L., Snijckers, C.M., Hill, R.P., Giles, R.H., Voest, E.E. and Ohh, M. (2005) Tumor suppression by the von Hippel–Lindau protein requires phosphorylation of the acidic domain. J. Biol. Chem. 280, 22205– 22211.
[58] Hergovich, A., Lisztwan, J., Thoma, C.R., Wirbelauer, C., Barry, R.E. and Krek, W. (2006) Priming-dependent phosphorylation and regulation of the tumor suppressor pVHL by glycogen synthase kinase 3. Mol. Cell Biol. 26, 5784–5796. [59] Kajava, A.V. (2001) Review: proteins with repeated sequence–structural
prediction and modeling. J. Struct. Biol. 134, 132–144.
[60] Lise, S. and Jones, D.T. (2005) Sequence patterns associated with disordered regions in proteins. Proteins 58, 144–150.