Search

Drug discovery is notorious for its low success rates [5]. Despite best research efforts, the majority of drugs fail at early stages of development, even before they enter clinical trials. This phenomenon stems from the inherent complexity of biological systems and our poor understanding of human diseases. To improve that understanding, swaths of data have been generated in recent years. Still, data does not easily translate into knowledge or actionable insights. Here we explore how approaches from the recommendation system domain could help scientists comprehend the ever-growing amount of biomedical facts. The aim of these efforts is to make better drug development decisions, which ultimately result in safe and efficient treatments for patients [3]. Recommendation systems are well established in e-commerce, streaming and social media platforms, however in the biomedical domain their usage is limited to a few recent studies [1, 6, 7, 8]. Direct transfer of classic recommendation approaches to the biomedical domain is not trivial. Specifics of the problem space impose numerous challenges for a recommendation system practitioner, to name a few: Regardless of the challenges, the adoption of recommendation systems presents numerous opportunities to support and accelerate drug discovery. Even a slight increase in success rate of drug pipelines will result in a vast number of patients gaining access to safe and effective treatments. Recommendation systems could play a leading role in this process. Adding context to experimental data is one class of problems that could benefit from recommenders. In this process new data is integrated with prior evidence to produce a new hypothesis. In a typical scenario, thousands of genes need to be ranked by their relevance to a disease given new and existing data. As a case study we focused on finding out why some lung cancer patients develop resistance to treatments. Current protocol to find resistance markers starts with high-throughput genomic screens resulting in an initial list of potential gene candidates, followed by tedious manual curation by several experts to reduce the list to a manageable number for further follow-up. To find resistance markers faster and to reduce bias we built a hybrid recommendation system on top of a heterogeneous biomedical knowledge graph [2]. In the absence of continuous feedback and training data, we approached recommendations as a multi-objective optimization problem [4]. Genes were ranked by trading off diverse types of evidence that link them to potential mechanisms of resistance in lung cancer. We used a knowledge graph as the primary source of features, so that the relevance of a gene could be expressed via properties of a graph. Our hybrid feature set also included clinical and pre-clinical data as well as metrics of literature support obtained with natural language processing techniques. This hybrid approach helped to identify novel resistance mechanisms that could have been overlooked by experts due to inherent bias or limited integration of data. Most importantly, our method reduced the time required to prioritise resistance markers from months to minutes and became a standard procedure for processing genomic screens. Another class of problems exists around target identification tasks. The idea here is to find a molecular target, often a gene or a protein, that could be modulated with a drug to treat a disease. As the number of potential targets is large, the search space can be reduced using network propagation on a dedicated subgraph that captures the functional relationship between genes. This approach also requires a set of seed genes, defined based on high confidence associations with diseases. Disease preferences are then propagated through the network resulting in a preference distribution for the complete set of genes which is used to reduce the search space. In contrast to adding context to experiments, a considerable amount of training data is available to support target identification. For instance, both successful and failed clinical trials can act as a useful source of data for target identification. Such a setting warrants use of supervised recommendation systems. A supervised approach, however introduces another machine learning hurdle — trust. Since supervised models are typically ”black boxes”, their quality must be ascertained indirectly, for example using train-test split and estimating model’s performance on the test set. Such quantitative performance metrics often are of little value to a biological expert looking for relevant gene targets. Instead, experts instinctively assess model quality by checking if a list of recommendations contains a handful of expected genes [9]. To simultaneously use biologists’ intuitions as training data, while avoiding an overly optimistic trust in model output, we used an ensemble modeling approach. We partitioned training data among multiple models such that each available training gene was omitted from one model’s training data. The model was then permitted to assess this previously unseen gene in constructing its final list of recommendations, while training genes were removed from consideration. Each model therefore produced a list of recommendations based on an incomplete set of genes. A final set of recommendations was then constructed by collating each individual model’s output list. Because these output lists were constructed with biologist input through supervised training, biologists placed a higher degree of trust in the recommendations. This allowed roughly two dozen genes to be fast-tracked for manual assessment and experimental screening. In summary, accumulation of large amounts of biomedical data coupled with a need to comprehend and reason about it makes drug discovery an attractive field to apply recommendation techniques. Specifics of the problem space and complexity of biological systems call for efficient recommendation solutions that could operate in unsupervised or weakly supervised settings. At the same time, a strong emphasis on explainability is essential to gain trust of biomedical experts.

Protein aggregation is an underappreciated mechanism that may contribute to the loss- and oncogenic-gain-of-function of mutant tumor suppressors such as p53 and axin. In the present study, we describe amyloid-like aggregation behaviour of the second most frequently mutated tumor suppressor in human cancer, PTEN. In silico analysis revealed a particularly high aggregation vulnerability for this protein, which was corroborated by in vitro aggregation assays. In cultured tumor cells, we found that under stress conditions, PTEN readily undergoes amyloid-like aggregation as a result of mutation. However, we also show that severe dysregulation of protein homeostasis may lead to aggregation of wild-type PTEN. These observations were supported by a small survey of patient-derived uterine tumor tissues, which found that more than 25% of tumors analyzed displayed wild-type PTEN aggregation. Finally, in an exploratory clinical study we found that PTEN aggregation status was correlated with a decline in clinical outcome. Our findings establish that the tumor suppressor PTEN is highly aggregation-prone and our work suggests that protein aggregation might be an underestimated but prevalent component of cancer cell biology.

Although p53 protein aggregates have been observed in cancer cell lines and tumour tissue, their impact in cancer remains largely unknown. Here, we extensively screened for p53 aggregation phenotypes in tumour biopsies, and identified nuclear inclusion bodies (nIBs) of transcriptionally inactive mutant or wild-type p53 as the most frequent aggregation-like phenotype across six different cancer types. p53-positive nIBs co-stained with nuclear aggregation markers, and shared molecular hallmarks of nIBs commonly found in neurodegenerative disorders. In cell culture, tumour-associated stress was a strong inducer of p53 aggregation and nIB formation. This was most prominent for mutant p53, but could also be observed in wild-type p53 cell lines, for which nIB formation correlated with the loss of p53's transcriptional activity. Importantly, protein aggregation also fuelled the dysregulation of the proteostasis network in the tumour cell by inducing a hyperactivated, oncogenic heat-shock response, to which tumours are commonly addicted, and by overloading the proteasomal degradation system, an observation that was most pronounced for structurally destabilized mutant p53. Patients showing tumours with p53-positive nIBs suffered from a poor clinical outcome, similar to those with loss of p53 expression, and tumour biopsies showed a differential proteostatic expression profile associated with p53-positive nIBs. p53-positive nIBs therefore highlight a malignant state of the tumour that results from the interplay between (1) the functional inactivation of p53 through mutation and/or aggregation, and (2) microenvironmental stress, a combination that catalyses proteostatic dysregulation. This study highlights several unexpected clinical, biological and therapeutically unexplored parallels between cancer and neurodegeneration. Copyright © 2017 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

In malignant cancer, excessive amounts of mutant p53 often lead to its aggregation, a feature that was recently identified as druggable. Here, we describe that the induction of a heat-shock response by Foldlin, a small-molecule tool compound, reduced the protein levels of misfolded/aggregated mutant p53 while contact mutants or wild-type p53 remained largely unaffected. Foldlin also prevented the formation of stress-induced p53 nuclear inclusion bodies. In spite of our inability to identify a specific molecular target, Foldlin also reduced protein levels of aggregating SOD1 variants. Finally, by screening a library of 778 FDA-approved compounds for their ability to reduce misfolded mutant p53, we identified the proteasome inhibitor Bortezomib with similar cellular effects as Foldlin. Overall, the induction of a cellular heat shock response seems an effective strategy to deal with pathological protein aggregation. It remains to be seen however, how this strategy can be translated to a clinical setting.

The accumulation of toxic protein aggregates is thought to play a key role in a range of degenerative pathologies, but it remains unclear why aggregation of polypeptides into non-native assemblies is toxic and why cellular clearance pathways offer ineffective protection. We here study the A4V mutant of SOD1, which forms toxic aggregates in motor neurons of patients with familial amyotrophic lateral sclerosis (ALS). A comparison of the location of aggregation prone regions (APRs) and Hsp70 binding sites in the denatured state of SOD1 reveals that ALS-associated mutations promote exposure of the APRs more than the strongest Hsc/Hsp70 binding site that we could detect. Mutations designed to increase the exposure of this Hsp70 interaction site in the denatured state promote aggregation but also display an increased interaction with Hsp70 chaperones. Depending on the cell type, in vitro this resulted in cellular inclusion body formation or increased clearance, accompanied with a suppression of cytotoxicity. The latter was also observed in a zebrafish model in vivo. Our results suggest that the uncontrolled accumulation of toxic SOD1A4V aggregates results from insufficient detection by the cellular surveillance network. ispartof: PROTEIN ENGINEERING DESIGN & SELECTION vol:32 issue:10 pages:443-457 ispartof: location:England status: published

Accurate prediction of the impact of genomic variation on phenotype is a major goal of computational biology and an important contributor to personalized medicine. Computational predictions can lead to a better understanding of the mechanisms underlying genetic diseases, including cancer, but their adoption requires thorough and unbiased assessment. Cystathionine-beta-synthase (CBS) is an enzyme that catalyzes the first step of the transsulfuration pathway, from homocysteine to cystathionine, and in which variations are associated with human hyperhomocysteinemia and homocystinuria. We have created a computational challenge under the CAGI framework to evaluate how well different methods can predict the phenotypic effect(s) of CBS single amino acid substitutions using a blinded experimental data set. CAGI participants were asked to predict yeast growth based on the identity of the mutations. The performance of the methods was evaluated using several metrics. The CBS challenge highlighted the difficulty of predicting the phenotype of an ex vivo system in a model organism when classification models were trained on human disease data. We also discuss the variations in difficulty of prediction for known benign and deleterious variants, as well as identify methodological and experimental constraints with lessons to be learned for future challenges. ispartof: HUMAN MUTATION vol:40 issue:9 pages:1530-1545 ispartof: location:United States status: published

Natural selection shapes protein solubility to physiological requirements and recombinant applications that require higher protein concentrations are often problematic. This raises the question whether the solubility of natural protein sequences can be improved. We here show an anti-correlation between the number of aggregation prone regions (APRs) in a protein sequence and its solubility, suggesting that mutational suppression of APRs provides a simple strategy to increase protein solubility. We show that mutations at specific positions within a protein structure can act as APR suppressors without affecting protein stability. These hot spots for protein solubility are both structure and sequence dependent but can be computationally predicted. We demonstrate this by reducing the aggregation of human α-galactosidase and protective antigen of Bacillus anthracis through mutation. Our results indicate that many proteins possess hot spots allowing to adapt protein solubility independently of structure and function., Mutations in aggregation prone regions of recombinant proteins often improve their solubility, although they might cause negative effects on their structure and function. Here, the authors identify proteins hot spots that can be exploited to optimize solubility without compromising stability.

Soluble functional proteins may transform into insoluble amyloid fibrils that deposit in a variety of tissues. Amyloid formation is a hallmark of age-related degenerative disorders. Perhaps surprisingly, amyloid fibrils can also be beneficial and are frequently exploited for diverse functional roles in organisms. Here we introduce AmyPro, an open-access database providing a comprehensive, carefully curated collection of validated amyloid fibril-forming proteins from all kingdoms of life classified into broad functional categories (http://amypro.net). In particular, AmyPro provides the boundaries of experimentally validated amyloidogenic sequence regions, short descriptions of the functional relevance of the proteins and their amyloid state, a list of the experimental techniques applied to study the amyloid state, important structural/functional/variation/mutation data transferred from UniProt, a list of relevant PDB structures categorized according to protein states, database cross-references and literature references. AmyPro greatly improves on similar currently available resources by incorporating both prions and functional amyloids in addition to pathogenic amyloids, and allows users to screen their sequences against the entire collection of validated amyloidogenic sequence fragments. By enabling further elucidation of the sequential determinants of amyloid fibril formation, we hope AmyPro will enhance the development of new methods for the precise prediction of amyloidogenic regions within proteins., SCOPUS: ar.j, info:eu-repo/semantics/published

Recently a number of aggregation disease polypeptides have been shown to spread from cell to cell thereby displaying prionoid behaviour. Studying aggregate internalisation however is often hampered by the complex kinetics of the aggregation process resulting in the concomitant uptake of aggregates of different sizes by competing mechanisms, which makes it difficult to isolate pathway-specific responses to aggregates. We designed synthetic aggregating peptides bearing different aggregation propensities with the aim of producing modes of uptake that are sufficiently distinct to differentially analyse the cellular response to internalization. We found that small acidic aggregates (≤500 nm diameter) were taken up by non-specific endocytosis as part of the fluid phase and travelled through the endosomal compartment to lysosomes. By contrast, basic bigger aggregates (>1 μm) were taken up through a mechanism dependent of cytoskeletal reorganization and membrane remodelling with the morphological hallmarks of phagocytosis. Importantly, the properties of these aggregates not only determined the mechanism of internalization but also the involvement of the proteostatic machinery (the assembly of interconnected networks that control the biogenesis, folding, trafficking and degradation of proteins) in the process: while the internalization of small acidic aggregates is HSF1 independent, the uptake of larger basic aggregates was HSF1 dependent requiring Hsp70. Our results show that the biophysical properties of aggregates determine both their mechanism of internalisation and proteostatic response. It remains to be seen whether these differences in cellular response contribute to the particular role of specific aggregated proteins in disease. ispartof: Journal of Biological Chemistry vol:290 issue:1 pages:242-58 ispartof: location:United States status: published

We investigated the origin and the effect of insertion D67D-THGERDLGPA within HIV-1 RT from a patient failing antiviral therapy. The insertion developed within the context of pre-existing NRTI and NNRTI mutations (M41L, L210W, T215Y and N348I). Concurrently, the NRTI mutations T69I and V118I and the NNRTI mutations K103N and Y181C were detected for the first time. High-level drug resistance (fold-changes≥50) and a good replication capacity (87% of wild-type) were observed, significantly higher than for the previous virus without insertion. The insertion was very similar to a region within human chromosome 17 (31/34 nucleotide identity), and had already been detected independently in a Japanese HIV-1 isolate. These results suggest that a particular sequence within human chromosome 17 is prone to horizontal gene transfer into the HIV-1 RT finger subdomain. This insertion confers selective advantage to HIV-1 by its contribution to multi-drug resistance and restoration of impaired replication capacity.

γ-Secretases are a family of intramembrane-cleaving proteases involved in various signaling pathways and diseases, including Alzheimer's disease (AD). Cells co-express differing γ-secretase complexes, including two homologous presenilins (PSENs). We examined the significance of this heterogeneity and identified a unique motif in PSEN2 that directs this γ-secretase to late endosomes/lysosomes via a phosphorylation-dependent interaction with the AP-1 adaptor complex. Accordingly, PSEN2 selectively cleaves late endosomal/lysosomal localized substrates and generates the prominent pool of intracellular Aβ that contains longer Aβ; familial AD (FAD)-associated mutations in PSEN2 increased the levels of longer Aβ further. Moreover, a subset of FAD mutants in PSEN1, normally more broadly distributed in the cell, phenocopies PSEN2 and shifts its localization to late endosomes/lysosomes. Thus, localization of γ-secretases determines substrate specificity, while FAD-causing mutations strongly enhance accumulation of aggregation-prone Aβ42 in intracellular acidic compartments. The findings reveal potentially important roles for specific intracellular, localized reactions contributing to AD pathogenesis. publisher: Elsevier articletitle: Restricted Location of PSEN2/γ-Secretase Determines Substrate Specificity and Generates an Intracellular Aβ Pool journaltitle: Cell articlelink: http://dx.doi.org/10.1016/j.cell.2016.05.020 content_type: article copyright: © 2016 Elsevier Inc. ispartof: Cell vol:166 issue:1 pages:193-208 ispartof: location:United States status: published

Hexanucleotide repeat expansions in C9orf72 are the most common cause of amyotrophic lateral sclerosis (ALS) and frontotemporal degeneration (FTD) (c9ALS/FTD). Unconventional translation of these repeats produces dipeptide repeat proteins (DPRs) that may cause neurodegeneration. We performed a modifier screen in Drosophila and discovered a critical role for importins and exportins, Ran-GTP cycle regulators, nuclear pore components and arginine methylases in mediating DPR toxicity. These findings provide evidence for an important role for nucleocytoplasmic transport in the pathogenic mechanism of c9ALS/FTD.

Contains fulltext : 108708.pdf (Publisher’s version ) (Open Access) Current therapeutic approaches under development for Alzheimer disease, including gamma-secretase modulating therapy, aim at increasing the production of Abeta(1-38) and Abeta(1-40) at the cost of longer Abeta peptides. Here, we consider the aggregation of Abeta(1-38) and Abeta(1-43) in addition to Abeta(1-40) and Abeta(1-42), in particular their behavior in mixtures representing the complex in vivo Abeta pool. We demonstrate that Abeta(1-38) and Abeta(1-43) aggregate similar to Abeta(1-40) and Abeta(1-42), respectively, but display a variation in the kinetics of assembly and toxicity due to differences in short timescale conformational plasticity. In biologically relevant mixtures of Abeta, Abeta(1-38) and Abeta(1-43) significantly affect the behaviors of Abeta(1-40) and Abeta(1-42). The short timescale conformational flexibility of Abeta(1-38) is suggested to be responsible for enhancing toxicity of Abeta(1-40) while exerting a cyto-protective effect on Abeta(1-42). Our results indicate that the complex in vivo Abeta peptide array and variations thereof is critical in Alzheimer disease, which can influence the selection of current and new therapeutic strategies.

Protein aggregation is a major factor limiting the biotechnological and therapeutic application of many proteins, including enzymes and monoclonal antibodies. The molecular principles underlying aggregation are by now sufficiently understood to allow rational redesign of natural polypeptide sequences for decreased aggregation tendency, and hence potentially increased expression and solubility. Given that aggregation-prone regions (APRs) tend to contribute to the stability of the hydrophobic core or to functional sites of the protein, mutations in these regions have to be carefully selected in order not to disrupt protein structure or function. Therefore, we here provide access to an automated pipeline to identify mutations that reduce protein aggregation by reducing the intrinsic aggregation propensity of the sequence (using the TANGO algorithm), while taking care not to disrupt the thermodynamic stability of the native structure (using the empirical force-field FoldX). Moreover, by providing a plot of the intrinsic aggregation propensity score of APRs corrected by the local stability of that region in the folded structure, we allow users to prioritize those regions in the protein that are most in need of improvement through protein engineering. The method can be accessed at http://solubis.switchlab.org/. ispartof: Protein Engineering, Design & Selection vol:29 issue:8 pages:285-289 ispartof: location:England status: published

Objective: To assess the genetic contribution of TBK1, a gene implicated in amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and FTD-ALS, in Belgian FTD and ALS patient cohorts containing a significant part of genetically unresolved patients. Methods: We sequenced TBK1 in a hospital-based cohort of 482 unrelated patients with FTD and FTD-ALS and 147 patients with ALS and an extended Belgian FTD-ALS family DR158. We followed up mutation carriers by segregation studies, transcript and protein expression analysis, and immunohistochemistry. Results: We identified 11 patients carrying a loss-of-function (LOF) mutation resulting in an overall mutation frequency of 1.7% (11/629), 1.1% in patients with FTD (5/460), 3.4% in patients with ALS (5/147), and 4.5% in patients with FTD-ALS (1/22). We found 1 LOF mutation, p.Glu643del, in 6 unrelated patients segregating with disease in family DR158. Of 2 mutation carriers, brain and spinal cord was characterized by TDP-43-positive pathology. The LOF mutations including the p.Glu643del mutation led to loss of transcript or protein in blood and brain. Conclusions: TBK1 LOF mutations are the third most frequent cause of clinical FTD in the Belgian clinically based patient cohort, after C9orf72 and GRN, and the second most common cause of clinical ALS after C9orf72. These findings reinforce that FTD and ALS belong to the same disease continuum. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially.

Motivation: Protein aggregation is associated with a number of protein misfolding diseases and is a major concern for therapeutic proteins. Aggregation is caused by the presence of aggregation-prone regions (APRs) in the amino acid sequence of the protein. The lower the aggregation propensity of APRs and the better they are protected by native interactions within the folded structure of the protein, the more aggregation is prevented. Therefore, both the local thermodynamic stability of APRs in the native structure and their intrinsic aggregation propensity are a key parameter that needs to be optimized to prevent protein aggregation. Results: The Solubis method presented here automates the process of carefully selecting point mutations that minimize the intrinsic aggregation propensity while improving local protein stability. Availability and implementation: All information about the Solubis plugin is available at http://solubisyasara.switchlab.org/. Contact: joost.schymkowitz@switch.vib-kuleuven.be or Frederic.Rousseau@switch.vib-kuleuven.be Supplementary information: Supplementary data are available at Bioinformatics online.

Protein aggregation is sequence specific, favoring self-assembly over cross-seeding with non-homologous sequences. Still, as the majority of proteins in a proteome are aggregation prone, the high level of homogeneity of protein inclusions in vivo both during recombinant overexpression and in disease remains surprising. To investigate the selectivity of protein aggregation in a proteomic context, we here compared the selectivity of aggregation-determined interactions with antibody binding. To that purpose, we synthesized biotin-labeled peptides, corresponding to aggregation-determining sequences of the bacterial protein β-galactosidase and two human disease biomarkers: C-reactive protein and prostate-specific antigen. We analyzed the selectivity of their interactions in Escherichia coli lysate, human serum and human seminal plasma, respectively, using a Western blot-like approach in which the aggregating peptides replace the conventional antibody. We observed specific peptide accumulation in the same bands detected by antibody staining. Combined spectroscopic and mutagenic studies confirmed accumulation resulted from binding of the peptide on the identical sequence of the immobilized target protein. Further, we analyzed the sequence redundancy of aggregating sequences and found that about 90% of them are unique within their proteome. As a result, the combined specificity and low sequence redundancy of aggregating sequences therefore contribute to the observed homogeneity of protein aggregation in vivo. This suggests that these intrinsic proteomic properties naturally compartmentalize aggregation events in sequence space. In the event of physiological stress, this might benefit the ability of cells to respond to proteostatic stress by allowing chaperones to focus on specific aggregation events rather than having to face systemic proteostatic failure.

Protein aggregation is geared by aggregation-prone regions that self-associate by β-strand interactions. Charged residues and prolines are enriched at the flanks of aggregation-prone regions resulting in decreased aggregation. It is still unclear what drives the overrepresentation of these "aggregation gatekeepers", that is, whether their presence results from structural constraints determining protein stability or whether they constitute a bona fide functional class selectively maintained to control protein aggregation. As functional residues are typically conserved regardless of their cost to protein stability, we compared sequence conservation and thermodynamic cost of these residues in 2659 protein families in Escherichia coli. Across protein families, we find gatekeepers to be under strong selective conservation while at the same time representing a significant thermodynamic cost to protein structure. This finding supports the notion that aggregation gatekeepers are not structurally determined but evolutionary selected to control protein aggregation. ispartof: Journal of Molecular Biology vol:426 issue:12 pages:2405-2412 ispartof: location:Netherlands status: published

Aggregated forms of the amyloid-β peptide are hypothesized to act as the prime toxic agents in Alzheimer disease (AD). The in vivo amyloid-β peptide pool consists of both C- and N-terminally truncated or mutated peptides, and the composition thereof significantly determines AD risk. Other variations, such as biotinylation, are introduced as molecular tools to aid the understanding of disease mechanisms. Since these modifications have the potential to alter key aggregation properties of the amyloid-β peptide, we present a comparative study of the aggregation of a substantial set of the most common in vivo identified and in vitro produced amyloid-β peptides. STRUCTURED SUMMARY OF PROTEIN INTERACTIONS: Amyloid beta and Amyloid betabind by fluorescence technology (View Interaction: 1, 2, 3, 4, 5) Amyloid beta and Amyloid betabind by transmission electron microscopy (View Interaction: 1, 2) Amyloid beta and Amyloid betabind by filter binding (View Interaction: 1, 2, 3)., Journal Article, SCOPUS: ar.j, info:eu-repo/semantics/published

Single nucleotide variants (SNVs) are, together with copy number variation, the primary source of variation in the human genome and are associated with phenotypic variation such as altered response to drug treatment and susceptibility to disease. Linking structural effects of non-synonymous SNVs to functional outcomes is a major issue in structural bioinformatics. The SNPeffect database (http://snpeffect.switchlab.org) uses sequence- and structure-based bioinformatics tools to predict the effect of protein-coding SNVs on the structural phenotype of proteins. It integrates aggregation prediction (TANGO), amyloid prediction (WALTZ), chaperone-binding prediction (LIMBO) and protein stability analysis (FoldX) for structural phenotyping. Additionally, SNPeffect holds information on affected catalytic sites and a number of post-translational modifications. The database contains all known human protein variants from UniProt, but users can now also submit custom protein variants for a SNPeffect analysis, including automated structure modeling. The new meta-analysis application allows plotting correlations between phenotypic features for a user-selected set of variants. ispartof: Nucleic Acids Research vol:40 issue:D1 pages:D935-D939 ispartof: location:England status: published

Protein aggregation is a hallmark of over 30 human pathologies. In these diseases, the aggregation of one or a few specific proteins is often toxic, leading to cellular degeneration and/or organ disruption in addition to the loss-of-function resulting from protein misfolding. Although the pathophysiological consequences of these diseases are overt, the molecular dysregulations leading to aggregate toxicity are still unclear and appear to be diverse and multifactorial. The molecular mechanisms of protein aggregation and therefore the biophysical parameters favoring protein aggregation are better understood. Here we perform an in silico survey of the impact of human sequence variation on the aggregation propensity of human proteins. We find that disease-associated variations are statistically significantly enriched in mutations that increase the aggregation potential of human proteins when compared to neutral sequence variations. These findings suggest that protein aggregation might have a broader impact on human disease than generally assumed and that beyond loss-of-function, the aggregation of mutant proteins involved in cancer, immune disorders or inflammation could potentially further contribute to disease by additional burden on cellular protein homeostasis., Author Summary Protein aggregation has been recognized to contribute to the development of more than 30 human diseases such as Alzheimer and Parkinson disease. Here we have performed an in silico survey of human sequence variations to evaluate whether protein aggregation might impact human disease beyond the above-mentioned aggregation diseases. We find that human disease mutations are more likely to increase the aggregation potential of proteins than non-disease associated mutations. This survey therefore suggests the possibility that protein aggregation is a more widespread disease modifier than previously expected.

We previously showed the existence of selective pressure against protein aggregation by the enrichment of aggregation-opposing ‘gatekeeper’ residues at strategic places along the sequence of proteins. Here we analyzed the relationship between protein lifetime and protein aggregation by combining experimentally determined turnover rates, expression data, structural data and chaperone interaction data on a set of more than 500 proteins. We find that selective pressure on protein sequences against aggregation is not homogeneous but that short-living proteins on average have a higher aggregation propensity and fewer chaperone interactions than long-living proteins. We also find that short-living proteins are more often associated to deposition diseases. These findings suggest that the efficient degradation of high-turnover proteins is sufficient to preclude aggregation, but also that factors that inhibit proteasomal activity, such as physiological ageing, will primarily affect the aggregation of short-living proteins., Author Summary In order to carry out their biological function, proteins need to fold into well-defined three-dimensional structures. Protein aggregation is a process whereby proteins misfold into inactive and often toxic higher order structures, which is implied in about 30 human diseases such as Alzheimer's disease, Parkinson's disease and systemic amyloidosis. In earlier work it has been shown that although protein aggregation is an intrinsic property of polypeptide chains that cannot be entirely avoided, evolution has optimized protein sequences to minimize the risk of aggregation in a proteome. Here we show that this pressure is not uniform, but that proteins with a short lifetime have on average a higher aggregation propensity than long-living proteins. In addition, we show that high turnover proteins also make fewer interactions with chaperones. Taken together, these observations suggest that under normal physiological conditions the aggregation propensity of short-lived proteins does not represent a significant treat for the biochemistry of the cell. Presumably the strong dependence of these proteins on proteasomal degradation is sufficient to preclude the accumulation of aggregates. As proteasomal activity declines with age this would also explain why we observe a higher association of high turnover proteins with age-dependent aggregation-related diseases.