The E3 ligase TRIM1 ubiquitinates LRRK2 and controls its localization,degradation, and toxicity

Missense mutations in leucine-rich repeat kinase 2 (LRRK2) are the most common cause of familial Parkinson’s disease (PD); however, pathways regulating LRRK2 subcellular localization, function, and turnover are not fully defined. We performed quantitative mass spectrometry–based interactome studies to identify 48 novel LRRK2 interactors, including the microtubule-associated E3 ubiquitin ligase TRIM1 (tripartite motif family 1). TRIM1 recruits LRRK2 to the microtubule cytoskeleton for ubiquitination and proteasomal degradation by binding LRRK2_911–919, a nine amino acid segment within a flexible interdomain region (LRRK2_853–981), which we designate the “regulatory loop” (RL). Phosphorylation of LRRK2 Ser910/Ser935 within LRRK2 RL influences LRRK2’s association with cytoplasmic 14-3-3 versus microtubule-bound TRIM1. Association with TRIM1 modulates LRRK2’s interaction with Rab29 and prevents upregulation of LRRK2 kinase activity by Rab29 in an E3-ligase–dependent manner. Finally, TRIM1 rescues neurite outgrowth deficits caused by PD-driving mutant LRRK2 G2019S. Our data suggest that TRIM1 is a critical regulator of LRRK2, controlling its degradation, localization, binding partners, kinase activity, and cytotoxicity.     

Characterization of a Six-Subunit Holo-Elongator Complex Required for the Regulated Expression of a Group of Genes in Saccharomyces cerevisiae

The Elongator complex associated with elongating RNA polymerase II in Saccharomyces cerevisiae was originally reported to have three subunits, Elp1, Elp2, and Elp3. Using the tandem affinity purification (TAP) procedure, we have purified a six-subunit yeast Holo-Elongator complex containing three additional polypeptides, which we have named Elp4, Elp5, and Elp6. TAP tapping and subsequent purification of any one of the six subunits result in the isolation of all six components. Purification of Elongator in higher salt concentrations served to demonstrate that the complex could be separated into two subcomplexes: one consisted of Elp1, -2, and -3, and the other consisted of Elp4, -5, and -6. Deletions of the individual genes encoding the new Elongator subunits showed that only the ELP5 gene is essential for growth. Disruption of the two nonessential new Elongator-encoding genes, ELP4 and ELP6, caused the same phenotypes observed with knockouts of the original Elongator-encoding genes. Results of microarray analyses demonstrated that the gene expression profiles of strains containing deletions of genes encoding subunits of either Elongator subcomplex, in which we detected significantly altered mRNA expression levels for 96 genes, are very similar, implying that all the Elongator subunits likely function together to regulate a group of S. cerevisiae genes in vivo.     

Towards accurate imputation of quantitative genetic interactions

Recent technological breakthroughs have enabled high-throughput quantitative measurements of hundreds of thousands of genetic interactions among hundreds of genes in Saccharomyces cerevisiae. However, these assays often fail to measure the genetic interactions among up to 40% of the studied gene pairs. Here we present a novel method, which combines genetic interaction data together with diverse genomic data, to quantitatively impute these missing interactions. We also present data on almost 190,000 novel interactions.     

HIV Nef is Secreted in Exosomes and Triggers Apoptosis in Bystander CD4+ T Cells

The HIV accessory protein negative factor (Nef) is one of the earliest and most abundantly expressed viral proteins. It is also found in the serum of infected individuals (Caby MP, Lankar D, Vincendeau-Scherrer C, Raposo G, Bonnerot C. Exosomal-like vesicles are present in human blood plasma. Int Immunol 2005;17:879–887). Extracellular Nef protein has deleterious effects on CD4⁺ T cells (James CO, Huang MB, Khan M, Garcia-Barrio M, Powell MD, Bond VC. Extracellular Nef protein targets CD4⁺ T cells for apoptosis by interacting with CXCR4 surface receptors. J Virol 2004;78:3099–3109), the primary targets of HIV, and can suppress immunoglobulin class switching in bystander B cells (Qiao X, He B, Chiu A, Knowles DM, Chadburn A, Cerutti A. Human immunodeficiency virus 1 Nef suppresses CD40-dependent immunoglobulin class switching in bystander B cells. Nat Immunol 2006;7:302–310). Nevertheless, the mode of exit of Nef from infected cells remains a conundrum. We found that Nef stimulates its own export via the release of exosomes from all cells examined. Depending on its intracellular location, these Nef exosomes form at the plasma membrane, late endosomes or both compartments in Jurkat, SupT1 and primary T cells, respectively. Nef release through exosomes is conserved also during HIV-1 infection of peripheral blood lymphocytes (PBLs). Released Nef exosomes cause activation-induced cell death of resting PBLs in vitro . Thus, HIV-infected cells export Nef in bioactive vesicles, which facilitate the depletion of CD4⁺ T cells that is a hallmark of acquired immunodeficiency syndrome (AIDS).     

A Complex-based Reconstruction of the Saccharomyces cerevisiae Interactome

Most cellular processes are performed by proteomic units that interact with each other. These units are often stoichiometrically stable complexes comprised of several proteins. To obtain a faithful view of the protein interactome we must view it in terms of these basic units (complexes and proteins) and the interactions between them. This study makes two contributions toward this goal. First, it provides a new algorithm for reconstruction of stable complexes from a variety of heterogeneous biological assays; our approach combines state-of-the-art machine learning methods with a novel hierarchical clustering algorithm that allows clusters to overlap. We demonstrate that our approach constructs over 40% more known complexes than other recent methods and that the complexes it produces are more biologically coherent even compared with the reference set. We provide experimental support for some of our novel predictions, identifying both a new complex involved in nutrient starvation and a new component of the eisosome complex. Second, we provide a high accuracy algorithm for the novel problem of predicting transient interactions involving complexes. We show that our complex level network, which we call ComplexNet, provides novel insights regarding the protein-protein interaction network. In particular, we reinterpret the finding that “hubs” in the network are enriched for being essential, showing instead that essential proteins tend to be clustered together in essential complexes and that these essential complexes tend to be large.Biological processes exhibit a hierarchical structure in which the basic working units, proteins, physically associate to form stoichiometrically stable complexes. Complexes interact with individual proteins or other complexes to form functional modules and pathways that carry out most cellular processes. Such higher level interactions are more transient than those within complexes and are highly dependent on temporal and spatial context. The function of each protein or complex depends on its interaction partners. Therefore, a faithful reconstruction of the entire set of complexes in the cell is essential to identifying the function of individual proteins and complexes as well as serving as a building block for understanding the higher level organization of the cell, such as the interactions of complexes and proteins within cellular pathways. Here we describe a novel method for reconstruction of complexes from a variety of biological assays and a method for predicting the network of interactions relating these core cellular units (complexes and proteins).Our reconstruction effort focuses on the yeast Saccharomyces cerevisiae. Yeast serves as the prototypical case study for the reconstruction of protein-protein interaction networks. Moreover the yeast complexes often have conserved orthologs in other organisms, including human, and are of interest in their own right. Several studies (1–4) using a variety of assays have generated high throughput data that directly measure protein-protein interactions. Most notably, two high quality data sets (3, 4) used tandem affinity purification (TAP)¹ followed by MS to provide a proteome-wide measurement of protein complexes. These data provide the basis for attempting a comprehensive reconstruction of a large fraction of the protein complexes in this organism. Indeed a number of works (5, 6) have attempted such a reconstruction. Generally speaking, all use the same general procedure: one or more data sources are used to estimate a set of affinities between pairs of proteins, essentially measuring the likelihood of that pair to participate together in a complex. These affinities induce a weighted graph whose nodes are proteins and whose edges encode the affinities. A clustering algorithm is then used to construct complexes, sets of proteins that have high affinity in the graph. Although similar at a high level, the different methods differ significantly on the design choices made for the key steps in the process.Recent works (since 2006) all focus on processing the proteome-wide TAP-MS data and using the results to define complexes. Gavin et al. (3), Collins et al. (7), and Hart et al. (5) all use probabilistic models that compare the number of interactions observed between proteins in the data versus the number expected in some null model. Collins et al. (7) and Hart et al. (5) both used all three of the available high throughput data sets (2–4) in an attempt to provide a unified interaction network. The two unified networks resulting from these studies were shown to have large overlap and to achieve comparable agreement with the set of co-complex interactions in the MIPS data set (8) that are collated from previous small scale studies. The interaction graphs resulting from the computed affinity scores are then clustered to produce a set of identified complexes. Gavin et al. (3), Hart et al. (5), and Pu et al. (6) all use a Markov clustering (MCL) (9) procedure; Collins et al. (7) use a hierarchical agglomerative clustering (HAC) procedure but do not suggest a computational procedure for using the resulting dendrogram to produce specific complex predictions.Despite the fairly high quality of these networks and the agreement between them, they still contain many false positives and negatives. False negatives can arise, for example, from the difficulty in detecting interactions involving low abundance proteins or membrane proteins or from cases where the tag added to the bait protein during TAP-MS prevents binding of the bait to its interacting partners. False positives can arise, for example, from complexes that share components or from the contaminants that bind to the bait nonspecifically after cell lysis. Therefore, the set of complexes derived from the protein-protein interaction network alone has limited accuracy. Less than 20% of the MIPS complexes (8), which are derived from reliable small scale experiments, are exactly captured by the predictions of Pu et al. (6) or by those of Hart et al. (5).In this study, we constructed a method that generates a set of complexes with higher sensitivity and coverage by integrating multiple sources of data, including mRNA gene expression data, cellular localization, and yeast two-hybrid data. The data integration approach was used in some early works on predicting protein-protein interactions (10, 11) and more recently by Qiu and Noble (12), but these studies focus only on predicting pairs of proteins in the same complex and not on reconstructing entire complexes. Many recent studies (13–21) have successfully integrated multiple types of data to predict functional linkage between proteins, constructing a graph whose pairwise affinity score summarizes the information from different sources of data. However, because the data integration is not trained toward predicting complexes, the high affinity pairs contain transient binding partners and even protein pairs that never interact directly but merely function in the same pathways. When these graphs are clustered, the clusters correspond to a variety of cellular entities, including pathways, functional modules, or co-expression clusters. We developed a data integration approach that is aimed directly at the problem of predicting stoichiometrically stable complexes.We used a two-phase automated procedure that we trained on a new high quality reference set that we generated from annotations in MIPS and SGD and from manual curation of the literature. In the first phase, we used boosting (22), a state-of-the-art machine learning method, to train an affinity function that is specifically aimed at predicting whether two proteins are co-complexed. Unlike most other learning methods, boosting is capable of inducing useful features by combining different aspects of the raw data, making it particularly well suited to a data integration setting. Once we generated the learned affinity graph over pairs of proteins, we predicted complexes by using a novel clustering algorithm called hierarchical agglomerative clustering with overlap (HACO). The HACO algorithm is a simple and elegant extension of HAC that addresses many of its limitations, such as the irreversible commitment to a possibly incorrect clustering decision. HACO can be applied to any setting where HAC is applied; given the enormous usefulness of HAC for the analysis of biological data sets of many different types (e.g. Refs. 7, 23, and 24), we believe that HACO may be applicable in a broad range of other tasks.To validate our approach, we tested the ability of our methods and other methods to predict reference complexes that were not used in training. By integrating multiple sources of data, we recovered more reference complexes than other state-of-the-art methods (5, 6) when applied to the same set of yeast proteins. We also validated our predicted set of complexes against external data sources that are not used in the training. In all cases, our predictions were shown to be more coherent than other methods and, in many cases, more coherent even than the set of reference complexes.A detailed examination of our predicted complexes suggests that many of them were previously known but not included in our (comprehensive) reference set, suggesting that our complexes form a valuable new set of reference complexes. In several cases, our predicted complexes were not previously characterized. We experimentally validated two of these predictions: a new component in the recently characterized eisosome complex (25), which marks the site of endocytosis in eukaryotes, and a newly characterized six-protein complex, including four phosphatases, that appears to be involved in the response to nutrient starvation and that we named the nutrient starvation complex (NSC).The complex-based view provides a new perspective on the analysis and reconstruction of the protein interaction network. In the past, Jeong et al. (26) have suggested that the degree of a protein in an interaction network is positively correlated with its essentiality and have argued that “hubs” in the network are more likely to be essential because they are involved in more interactions. Our analysis presents a complex-based alternative view: essential proteins tend to cluster together in essential complexes (5), and essential complexes tend to be large; thus, the essential hubs in the network are often members in large complexes comprised mostly of essential proteins. We also reformulate the task of reconstructing the protein interaction network. Rather than considering interactions between individual proteins (27–29), a somewhat confusing network that confounds interactions within complexes and interactions between complexes, we tackle the novel task of predicting a comprehensive protein interaction network that involves both individual proteins and larger complexes. We argue that these entities are the right building blocks in reconstructing cellular processes, providing a view of cellular interaction networks that is both easier to interpret than the complex network of interactions between individual proteins and more faithful to biological reality. Moreover a complex, which is a stable collection of many proteins that act together, provides a more robust basis for predicting interactions as we can combine signals for all its constituent proteins, reducing sensitivity to noise.To accomplish this goal, we constructed a reference set of complex-complex interactions, considering two complexes to interact if they are significantly enriched for reliable interactions between their components. We further augmented this set with a hand-curated list of established complex-complex interactions. We then used a machine learning approach to detect the “signature” of such interactions from a large set of assays that are likely to be indicative. We explored different machine learning methods and showed that a partially supervised naïve Bayes model, where we learned the model from both labeled and unlabeled interactions, provides the best performance. This model was applied both to our predicted complexes and to individual proteins, providing a new, comprehensive reconstruction of the S. cerevisiae interaction network, which can be downloaded from our project Web page.² We showed that entities that are predicted to interact are more likely to share the same functional categories. A detailed investigation of our new predicted interactions presents many that are established in the literature as well as some that are novel but consistent, presenting plausible hypotheses for further investigation.     

Host Cell Factors in HIV Replication: Meta-Analysis of Genome-Wide Studies

We have analyzed host cell genes linked to HIV replication that were identified in nine genome-wide studies, including three independent siRNA screens. Overlaps among the siRNA screens were very modest (<7% for any pairwise combination), and similarly, only modest overlaps were seen in pairwise comparisons with other types of genome-wide studies. Combining all genes from the genome-wide studies together with genes reported in the literature to affect HIV yields 2,410 protein-coding genes, or fully 9.5% of all human genes (though of course some of these are false positive calls). Here we report an “encyclopedia” of all overlaps between studies (available at http://www.hostpathogen.org), which yielded a more extensively corroborated set of host factors assisting HIV replication. We used these genes to calculate refined networks that specify cellular subsystems recruited by HIV to assist in replication, and present additional analysis specifying host cell genes that are attractive as potential therapeutic targets.     

A genome-wide screen identifies the evolutionarily conserved KEOPS complex as a telomere regulator

Telomere capping is the essential function of telomeres. To identify new genes involved in telomere capping, we carried out a genome-wide screen in Saccharomyces cerevisiae for suppressors of cdc13-1, an allele of the telomere-capping protein Cdc13. We report the identification of five novel suppressors, including the previously uncharacterized gene YML036W, which we name CGI121. Cgi121 is part of a conserved protein complex -- the KEOPS complex -- containing the protein kinase Bud32, the putative peptidase Kae1, and the uncharacterized protein Gon7. Deletion of CGI121 suppresses cdc13-1 via the dramatic reduction in ssDNA levels that accumulate in cdc13-1 cgi121 mutants. Deletion of BUD32 or other KEOPS components leads to short telomeres and a failure to add telomeres de novo to DNA double-strand breaks. Our results therefore indicate that the KEOPS complex promotes both telomere uncapping and telomere elongation.     

Non-degradative Ubiquitination of Protein Kinases

Growing evidence supports other regulatory roles for protein ubiquitination in addition to serving as a tag for proteasomal degradation. In contrast to other common post-translational modifications, such as phosphorylation, little is known about how non-degradative ubiquitination modulates protein structure, dynamics, and function. Due to the wealth of knowledge concerning protein kinase structure and regulation, we examined kinase ubiquitination using ubiquitin remnant immunoaffinity enrichment and quantitative mass spectrometry to identify ubiquitinated kinases and the sites of ubiquitination in Jurkat and HEK293 cells. We find that, unlike phosphorylation, ubiquitination most commonly occurs in structured domains, and on the kinase domain, ubiquitination is concentrated in regions known to be important for regulating activity. We hypothesized that ubiquitination, like other post-translational modifications, may alter the conformational equilibrium of the modified protein. We chose one human kinase, ZAP-70, to simulate using molecular dynamics with and without a monoubiquitin modification. In Jurkat cells, ZAP-70 is ubiquitinated at several sites that are not sensitive to proteasome inhibition and thus may have other regulatory roles. Our simulations show that ubiquitination influences the conformational ensemble of ZAP-70 in a site-dependent manner. When monoubiquitinated at K377, near the C-helix, the active conformation of the ZAP-70 C-helix is disrupted. In contrast, when monoubiquitinated at K476, near the kinase hinge region, an active-like ZAP-70 C-helix conformation is stabilized. These results lead to testable hypotheses that ubiquitination directly modulates kinase activity, and that ubiquitination is likely to alter structure, dynamics, and function in other protein classes as well.     

Navigating the chaperone network: an integrative map of physical and genetic interactions mediated by the hsp90 chaperone

Physical, genetic, and chemical-genetic interactions centered on the conserved chaperone Hsp90 were mapped at high resolution in yeast using systematic proteomic and genomic methods. Physical interactions were identified using genome-wide two hybrid screens combined with large-scale affinity purification of Hsp90-containing protein complexes. Genetic interactions were uncovered using synthetic genetic array technology and by a microarray-based chemical-genetic screen of a set of about 4700 viable yeast gene deletion mutants for hypersensitivity to the Hsp90 inhibitor geldanamycin. An extended network, consisting of 198 putative physical interactions and 451 putative genetic and chemical-genetic interactions, was found to connect Hsp90 to cofactors and substrates involved in a wide range of cellular functions. Two novel Hsp90 cofactors, Tah1 (YCR060W) and Pih1 (YHR034C), were also identified. These cofactors interact physically and functionally with the conserved AAA(+)-type DNA helicases Rvb1/Rvb2, which are key components of several chromatin remodeling factors, thereby linking Hsp90 to epigenetic gene regulation.