Network-based elucidation of colon cancer drug resistance mechanisms by phosphoproteomic time-series analysis.

Aberrant signaling pathway activity is a hallmark of tumorigenesis and progression, which has guided targeted inhibitor design for over 30 years. Yet, adaptive resistance mechanisms, induced by rapid, context-specific signaling network rewiring, continue to challenge therapeutic efficacy. Leveraging progress in proteomic technologies and network-based methodologies, we introduce Virtual Enrichment-based Signaling Protein-activity Analysis (VESPA)-an algorithm designed to elucidate mechanisms of cell response and adaptation to drug perturbations-and use it to analyze 7-point phosphoproteomic time series from colorectal cancer cells treated with clinically-relevant inhibitors and control media. Interrogating tumor-specific enzyme/substrate interactions accurately infers kinase and phosphatase activity, based on their substrate phosphorylation state, effectively accounting for signal crosstalk and sparse phosphoproteome coverage. The analysis elucidates time-dependent signaling pathway response to each drug perturbation and, more importantly, cell adaptive response and rewiring, experimentally confirmed by CRISPR knock-out assays, suggesting broad applicability to cancer and other diseases.

Specific regulation of BACH1 by the hotspot mutant p53R175H reveals a distinct gain-of-function mechanism.

Although the gain of function (GOF) of p53 mutants is well recognized, it remains unclear whether different p53 mutants share the same cofactors to induce GOFs. In a proteomic screen, we identified BACH1 as a cellular factor that recognizes the p53 DNA-binding domain depending on its mutation status. BACH1 strongly interacts with p53R175H but fails to effectively bind wild-type p53 or other hotspot mutants in vivo for functional regulation. Notably, p53R175H acts as a repressor for ferroptosis by abrogating BACH1-mediated downregulation of SLC7A11 to enhance tumor growth; conversely, p53R175H promotes BACH1-dependent tumor metastasis by upregulating expression of pro-metastatic targets. Mechanistically, p53R175H-mediated bidirectional regulation of BACH1 function is dependent on its ability to recruit the histone demethylase LSD2 to target promoters and differentially modulate transcription. These data demonstrate that BACH1 acts as a unique partner for p53R175H in executing its specific GOFs and suggest that different p53 mutants induce their GOFs through distinct mechanisms.

Network-based elucidation of colon cancer drug resistance by phosphoproteomic time-series analysis.

Aberrant signaling pathway activity is a hallmark of tumorigenesis and progression, which has guided targeted inhibitor design for over 30 years. Yet, adaptive resistance mechanisms, induced by rapid, context-specific signaling network rewiring, continue to challenge therapeutic efficacy. By leveraging progress in proteomic technologies and network-based methodologies, over the past decade, we developed VESPA-an algorithm designed to elucidate mechanisms of cell response and adaptation to drug perturbations-and used it to analyze 7-point phosphoproteomic time series from colorectal cancer cells treated with clinically-relevant inhibitors and control media. Interrogation of tumor-specific enzyme/substrate interactions accurately inferred kinase and phosphatase activity, based on their inferred substrate phosphorylation state, effectively accounting for signal cross-talk and sparse phosphoproteome coverage. The analysis elucidated time-dependent signaling pathway response to each drug perturbation and, more importantly, cell adaptive response and rewiring that was experimentally confirmed by CRISPRko assays, suggesting broad applicability to cancer and other diseases.

Sorting of cadherin-catenin-associated proteins into individual clusters.

The cytoplasmic tails of classical cadherins form a multiprotein cadherin-catenin complex (CCC) that constitutes the major structural unit of adherens junctions (AJs). The CCC in AJs forms junctional clusters, "E clusters," driven by cis and trans interactions in the cadherin ectodomain and stabilized by α-catenin-actin interactions. Additional proteins are known to bind to the cytoplasmic region of the CCC. Here, we analyze how these CCC-associated proteins (CAPs) integrate into cadherin clusters and how they affect the clustering process. Using a cross-linking approach coupled with mass spectrometry, we found that the majority of CAPs, including the force-sensing protein vinculin, interact with CCCs outside of AJs. Accordingly, structural modeling shows that there is not enough space for CAPs the size of vinculin to integrate into E clusters. Using two CAPs, scribble and erbin, as examples, we provide evidence that these proteins form separate clusters, which we term "C clusters." As proof of principle, we show, by using cadherin ectodomain monoclonal antibodies (mAbs), that mAb-bound E-cadherin forms separate clusters that undergo trans interactions. Taken together, our data suggest that, in addition to its role in cell-cell adhesion, CAP-driven CCC clustering serves to organize cytoplasmic proteins into distinct domains that may synchronize signaling networks of neighboring cells within tissues.

Integrating 3D structural information into systems biology.

Systems biology is a data-heavy field that focuses on systems-wide depictions of biological phenomena necessarily sacrificing a detailed characterization of individual components. As an example, genome-wide protein interaction networks are widely used in systems biology and continuously extended and refined as new sources of evidence become available. Despite the vast amount of information about individual protein structures and protein complexes that has accumulated in the past 50 years in the Protein Data Bank, the data, computational tools, and language of structural biology are not an integral part of systems biology. However, increasing effort has been devoted to this integration, and the related literature is reviewed here. Relationships between proteins that are detected via structural similarity offer a rich source of information not available from sequence similarity, and homology modeling can be used to leverage Protein Data Bank structures to produce 3D models for a significant fraction of many proteomes. A number of structure-informed genomic and cross-species (i.e., virus-host) interactomes will be described, and the unique information they provide will be illustrated with a number of examples. Tissue- and tumor-specific interactomes have also been developed through computational strategies that exploit patient information and through genetic interactions available from increasingly sensitive screens. Strategies to integrate structural information with these alternate data sources will be described. Finally, efforts to link protein structure space with chemical compound space offer novel sources of information in drug design, off-target identification, and the identification of targets for compounds found to be effective in phenotypic screens.

Oncoprotein-specific molecular interaction maps (SigMaps) for cancer network analyses.

Tumor-specific elucidation of physical and functional oncoprotein interactions could improve tumorigenic mechanism characterization and therapeutic response prediction. Current interaction models and pathways, however, lack context specificity and are not oncoprotein specific. We introduce SigMaps as context-specific networks, comprising modulators, effectors and cognate binding-partners of a specific oncoprotein. SigMaps are reconstructed de novo by integrating diverse evidence sources-including protein structure, gene expression and mutational profiles-via the OncoSig machine learning framework. We first generated a KRAS-specific SigMap for lung adenocarcinoma, which recapitulated published KRAS biology, identified novel synthetic lethal proteins that were experimentally validated in three-dimensional spheroid models and established uncharacterized crosstalk with RAB/RHO. To show that OncoSig is generalizable, we first inferred SigMaps for the ten most mutated human oncoproteins and then for the full repertoire of 715 proteins in the COSMIC Cancer Gene Census. Taken together, these SigMaps show that the cell's regulatory and signaling architecture is highly tissue specific.

A Structure-Informed Atlas of Human-Virus Interactions.

While knowledge of protein-protein interactions (PPIs) is critical for understanding virus-host relationships, limitations on the scalability of high-throughput methods have hampered their identification beyond a number of well-studied viruses. Here, we implement an in silico computational framework (pathogen host interactome prediction using structure similarity [P-HIPSTer]) that employs structural information to predict ∼282,000 pan viral-human PPIs with an experimental validation rate of ∼76%. In addition to rediscovering known biology, P-HIPSTer has yielded a series of new findings: the discovery of shared and unique machinery employed across human-infecting viruses, a likely role for ZIKV-ESR1 interactions in modulating viral replication, the identification of PPIs that discriminate between human papilloma viruses (HPVs) with high and low oncogenic potential, and a structure-enabled history of evolutionary selective pressure imposed on the human proteome. Further, P-HIPSTer enables discovery of previously unappreciated cellular circuits that act on human-infecting viruses and provides insight into experimentally intractable viruses.

Trans-endocytosis elicited by nectins transfers cytoplasmic cargo, including infectious material, between cells.

Here, we show that cells expressing the adherens junction protein nectin-1 capture nectin-4-containing membranes from the surface of adjacent cells in a trans-endocytosis process. We find that internalized nectin-1-nectin-4 complexes follow the endocytic pathway. The nectin-1 cytoplasmic tail controls transfer: its deletion prevents trans-endocytosis, while its exchange with the nectin-4 tail reverses transfer direction. Nectin-1-expressing cells acquire dye-labeled cytoplasmic proteins synchronously with nectin-4, a process most active during cell adhesion. Some cytoplasmic cargo remains functional after transfer, as demonstrated with encapsidated genomes of measles virus (MeV). This virus uses nectin-4, but not nectin-1, as a receptor. Epithelial cells expressing nectin-4, but not those expressing another MeV receptor in its place, can transfer infection to nectin-1-expressing primary neurons. Thus, this newly discovered process can move cytoplasmic cargo, including infectious material, from epithelial cells to neurons. We name the process nectin-elicited cytoplasm transfer (NECT). NECT-related trans-endocytosis processes may be exploited by pathogens to extend tropism. This article has an associated First Person interview with the first author of the paper.

Integration of Multi-omics Data for Gene Regulatory Network Inference and Application to Breast Cancer.

Underlying a cancer phenotype is a specific gene regulatory network that represents the complex regulatory relationships between genes. However, it remains a challenge to find cancer-related gene regulatory network because of insufficient sample sizes and complex regulatory mechanisms in which gene is influenced by not only other genes but also other biological factors. With the development of high-throughput technologies and the unprecedented wealth of multi-omics data give us a new opportunity to design machine learning method to investigate underlying gene regulatory network. In this paper, we propose an approach, which use biweight midcorrelation to measure the correlation between factors and make use of nonconvex penalty based sparse regression for gene regulatory network inference (BMNPGRN). BMNCGRN incorporates multi-omics data (including DNA methylation and copy number variation) and their interactions in gene regulatory network model. The experimental results on synthetic datasets show that BMNPGRN outperforms popular and state-of-the-art methods (including DCGRN, ARACNE and CLR) under false positive control. Furthermore, we applied BMNPGRN on breast cancer (BRCA) data from The Cancer Genome Atlas database and provided gene regulatory network.

Pathogenic IgG4 autoantibodies from endemic pemphigus foliaceus recognize a desmoglein-1 conformational epitope.

Fogo Selvagem (FS), the endemic form of pemphigus foliaceus, is mediated by pathogenic IgG4 autoantibodies against the amino-terminal extracellular cadherin domain of the desmosomal cadherin desmoglein 1 (Dsg1). Here we define the detailed epitopes of these pathogenic antibodies. Proteolytic footprinting showed that IgG4 from 95% of FS donor sera (19/20) recognized a 16-residue peptide (A129LNSMGQDLERPLELR144) from the EC1 domain of Dsg1 that overlaps the binding site for an adhesive-partner desmosomal cadherin molecule. Mutation of Dsg1 residues M133 and Q135 reduced the binding of FS IgG4 autoantibodies to Dsg1 by ∼50%. Molecular modeling identified two nearby EC1 domain residues (Q82 and V83) likely to contribute to the epitope. Mutation of these residues completely abolished the binding of FS IgG4 to Dsg1. Bead aggregation assays showed that native binding interactions between Dsg1 and desmocollin 1 (Dsc1), which underlie desmosome structure, were abolished by Fab fragments of FS IgG4. These results further define the molecular mechanism by which FS IgG4 autoantibodies interfere with desmosome structure and lead to cell-cell detachment, the hallmark of this disease.

Silencing c-Myc translation as a therapeutic strategy through targeting PI3Kδ and CK1ε in hematological malignancies.

Phosphoinositide 3-kinase (PI3K) and the proteasome pathway are both involved in activating the mechanistic target of rapamycin (mTOR). Because mTOR signaling is required for initiation of messenger RNA translation, we hypothesized that cotargeting the PI3K and proteasome pathways might synergistically inhibit translation of c-Myc. We found that a novel PI3K δ isoform inhibitor TGR-1202, but not the approved PI3Kδ inhibitor idelalisib, was highly synergistic with the proteasome inhibitor carfilzomib in lymphoma, leukemia, and myeloma cell lines and primary lymphoma and leukemia cells. TGR-1202 and carfilzomib (TC) synergistically inhibited phosphorylation of the eukaryotic translation initiation factor 4E (eIF4E)-binding protein 1 (4E-BP1), leading to suppression of c-Myc translation and silencing of c-Myc-dependent transcription. The synergistic cytotoxicity of TC was rescued by overexpression of eIF4E or c-Myc. TGR-1202, but not other PI3Kδ inhibitors, inhibited casein kinase-1 ε (CK1ε). Targeting CK1ε using a selective chemical inhibitor or short hairpin RNA complements the effects of idelalisib, as a single agent or in combination with carfilzomib, in repressing phosphorylation of 4E-BP1 and the protein level of c-Myc. These results suggest that TGR-1202 is a dual PI3Kδ/CK1ε inhibitor, which may in part explain the clinical activity of TGR-1202 in aggressive lymphoma not found with idelalisib. Targeting CK1ε should become an integral part of therapeutic strategies targeting translation of oncogenes such as c-Myc.

Acetylation-regulated interaction between p53 and SET reveals a widespread regulatory mode.

Although lysine acetylation is now recognized as a general protein modification for both histones and non-histone proteins, the mechanisms of acetylation-mediated actions are not completely understood. Acetylation of the C-terminal domain (CTD) of p53 (also known as TP53) was an early example of non-histone protein acetylation and its precise role remains unclear. Lysine acetylation often creates binding sites for bromodomain-containing 'reader' proteins. Here we use a proteomic screen to identify the oncoprotein SET as a major cellular factor whose binding with p53 is dependent on CTD acetylation status. SET profoundly inhibits p53 transcriptional activity in unstressed cells, but SET-mediated repression is abolished by stress-induced acetylation of p53 CTD. Moreover, loss of the interaction with SET activates p53, resulting in tumour regression in mouse xenograft models. Notably, the acidic domain of SET acts as a 'reader' for the unacetylated CTD of p53 and this mechanism of acetylation-dependent regulation is widespread in nature. For example, acetylation of p53 also modulates its interactions with similar acidic domains found in other p53 regulators including VPRBP (also known as DCAF1), DAXX and PELP1 (refs. 7, 8, 9), and computational analysis of the proteome has identified numerous proteins with the potential to serve as acidic domain readers and lysine-rich ligands. Unlike bromodomain readers, which preferentially bind the acetylated forms of their cognate ligands, the acidic domain readers specifically recognize the unacetylated forms of their ligands. Finally, the acetylation-dependent regulation of p53 was further validated in vivo by using a knock-in mouse model expressing an acetylation-mimicking form of p53. These results reveal that acidic-domain-containing factors act as a class of acetylation-dependent regulators by targeting p53 and, potentially, other proteins.

Lipids Regulate Lck Protein Activity through Their Interactions with the Lck Src Homology 2 Domain.

Lymphocyte-specific protein-tyrosine kinase (Lck) plays an essential role in T cell receptor (TCR) signaling and T cell development, but its activation mechanism is not fully understood. To explore the possibility that plasma membrane (PM) lipids control TCR signaling activities of Lck, we measured the membrane binding properties of its regulatory Src homology 2 (SH2) and Src homology 3 domains. The Lck SH2 domain binds anionic PM lipids with high affinity but with low specificity. Electrostatic potential calculation, NMR analysis, and mutational studies identified the lipid-binding site of the Lck SH2 domain that includes surface-exposed basic, aromatic, and hydrophobic residues but not the phospho-Tyr binding pocket. Mutation of lipid binding residues greatly reduced the interaction of Lck with the ζ chain in the activated TCR signaling complex and its overall TCR signaling activities. These results suggest that PM lipids, including phosphatidylinositol 4,5-bisphosphate and phosphatidylinositol 3,4,5-trisphosphate, modulate interaction of Lck with its binding partners in the TCR signaling complex and its TCR signaling activities in a spatiotemporally specific manner via its SH2 domain.

Dclk1 Defines Quiescent Pancreatic Progenitors that Promote Injury-Induced Regeneration and Tumorigenesis.

The existence of adult pancreatic progenitor cells has been debated. While some favor the concept of facultative progenitors involved in homeostasis and repair, neither a location nor markers for such cells have been defined. Using genetic lineage tracing, we show that Doublecortin-like kinase-1 (Dclk1) labels a rare population of long-lived, quiescent pancreatic cells. In vitro, Dclk1+ cells proliferate readily and sustain pancreatic organoid growth. In vivo, Dclk1+ cells are necessary for pancreatic regeneration following injury and chronic inflammation. Accordingly, their loss has detrimental effects after cerulein-induced pancreatitis. Expression of mutant Kras in Dclk1+ cells does not affect their quiescence or longevity. However, experimental pancreatitis converts Kras mutant Dclk1+ cells into potent cancer-initiating cells. As a potential effector of Kras, Dclk1 contributes functionally to the pathogenesis of pancreatic cancer. Taken together, these observations indicate that Dclk1 marks quiescent pancreatic progenitors that are candidates for the origin of pancreatic cancer.

p21-activated Kinases (PAKs) Mediate the Phosphorylation of PREX2 Protein to Initiate Feedback Inhibition of Rac1 GTPase.

Phosphatidylinositol 3,4,5-trisphosphate (PIP3)-dependent Rac exchanger 2 (PREX2) is a guanine nucleotide exchange factor (GEF) for the Ras-related C3 botulinum toxin substrate 1 (Rac1) GTPase, facilitating the exchange of GDP for GTP on Rac1. GTP-bound Rac1 then activates its downstream effectors, including p21-activated kinases (PAKs). PREX2 and Rac1 are frequently mutated in cancer and have key roles within the insulin-signaling pathway. Rac1 can be inactivated by multiple mechanisms; however, negative regulation by insulin is not well understood. Here, we show that in response to being activated after insulin stimulation, Rac1 initiates its own inactivation by decreasing PREX2 GEF activity. Following PREX2-mediated activation of Rac1 by the second messengers PIP3 or Gßγ, we found that PREX2 was phosphorylated through a PAK-dependent mechanism. PAK-mediated phosphorylation of PREX2 reduced GEF activity toward Rac1 by inhibiting PREX2 binding to PIP3 and Gßγ. Cell fractionation experiments also revealed that phosphorylation prevented PREX2 from localizing to the cellular membrane. Furthermore, the onset of insulin-induced phosphorylation of PREX2 was delayed compared with AKT. Altogether, we propose that second messengers activate the Rac1 signal, which sets in motion a cascade whereby PAKs phosphorylate and negatively regulate PREX2 to decrease Rac1 activation. This type of regulation would allow for transient activation of the PREX2-Rac1 signal and may be relevant in multiple physiological processes, including diseases such as diabetes and cancer when insulin signaling is chronically activated.

α-Catenin-mediated cadherin clustering couples cadherin and actin dynamics.

The function of the actin-binding domain of α-catenin, αABD, including its possible role in the direct anchorage of the cadherin-catenin complex to the actin cytoskeleton, has remained uncertain. We identified two point mutations on the αABD surface that interfere with αABD binding to actin and used them to probe the role of α-catenin-actin interactions in adherens junctions. We found that the junctions directly bound to actin via αABD were more dynamic than the junctions bound to actin indirectly through vinculin and that recombinant αABD interacted with cortical actin but not with actin bundles. This interaction resulted in the formation of numerous short-lived cortex-bound αABD clusters. Our data suggest that αABD clustering drives the continuous assembly of transient, actin-associated cadherin-catenin clusters whose disassembly is maintained by actin depolymerization. It appears then that such actin-dependent αABD clustering is a unique molecular mechanism mediating both integrity and reassembly of the cell-cell adhesive interface formed through weak cis- and trans-intercadherin interactions.

Nectin ectodomain structures reveal a canonical adhesive interface.

Nectins are immunoglobulin superfamily glycoproteins that mediate intercellular adhesion in many vertebrate tissues. Homophilic and heterophilic interactions between nectin family members help mediate tissue patterning. We determined the homophilic binding affinities and heterophilic specificities of all four nectins and the related protein nectin-like 5 (Necl-5) from human and mouse, revealing a range of homophilic interaction strengths and a defined heterophilic specificity pattern. To understand the molecular basis of their adhesion and specificity, we determined the crystal structures of natively glycosylated full ectodomains or adhesive fragments of all four nectins and Necl-5. All of the crystal structures revealed dimeric nectins bound through a stereotyped interface that was previously proposed to represent a cis dimer. However, conservation of this interface and the results of targeted cross-linking experiments showed that this dimer probably represents the adhesive trans interaction. The structure of the dimer provides a simple molecular explanation for the adhesive binding specificity of nectins.

Using systems and structure biology tools to dissect cellular phenotypes.

The Center for the Multiscale Analysis of Genetic Networks (MAGNet, http://magnet.c2b2.columbia.edu) was established in 2005, with the mission of providing the biomedical research community with Structural and Systems Biology algorithms and software tools for the dissection of molecular interactions and for the interaction-based elucidation of cellular phenotypes. Over the last 7 years, MAGNet investigators have developed many novel analysis methodologies, which have led to important biological discoveries, including understanding the role of the DNA shape in protein-DNA binding specificity and the discovery of genes causally related to the presentation of malignant phenotypes, including lymphoma, glioma, and melanoma. Software tools implementing these methodologies have been broadly adopted by the research community and are made freely available through geWorkbench, the Center's integrated analysis platform. Additionally, MAGNet has been instrumental in organizing and developing key conferences and meetings focused on the emerging field of systems biology and regulatory genomics, with special focus on cancer-related research.

Diversity in DNA recognition by p53 revealed by crystal structures with Hoogsteen base pairs.

p53 binds as a tetramer to DNA targets consisting of two decameric half-sites separated by a variable spacer. Here we present high-resolution crystal structures of complexes between p53 core-domain tetramers and DNA targets consisting of contiguous half-sites. In contrast to previously reported p53-DNA complexes that show standard Watson-Crick base pairs, the newly reported structures show noncanonical Hoogsteen base-pairing geometry at the central A-T doublet of each half-site. Structural and computational analyses show that the Hoogsteen geometry distinctly modulates the B-DNA helix in terms of local shape and electrostatic potential, which, together with the contiguous DNA configuration, results in enhanced protein-DNA and protein-protein interactions compared to noncontiguous half-sites. Our results suggest a mechanism relating spacer length to protein-DNA binding affinity. Our findings also expand the current understanding of protein-DNA recognition and establish the structural and chemical properties of Hoogsteen base pairs as the basis for a novel mode of sequence readout.