Sebastian Maurer-Stroh

Exploring the sequence determinants of amyloid structure using position-specific scoring matrices

Protein aggregation results in β-sheet–like assemblies that adopt either a variety of amorphous morphologies or ordered amyloid-like structures. These differences in structure also reflect biological differences; amyloid and amorphous β-sheet aggregates have different chaperone affinities, accumulate in different cellular locations and are degraded by different mechanisms. Further, amyloid function depends entirely on a high intrinsic degree of order. Here we experimentally explored the sequence space of amyloid hexapeptides and used the derived data to build Waltz, a web-based tool that uses a position-specific scoring matrix to determine amyloid-forming sequences. Waltz allows users to identify and better distinguish between amyloid sequences and amorphous β-sheet aggregates and allowed us to identify amyloid-forming regions in functional amyloids.

TPR Subunits of the Anaphase-Promoting Complex Mediate Binding to the Activator Protein CDH1

BACKGROUND Chromosome segregation and mitotic exit depend on activation of the anaphase-promoting complex (APC) by the substrate adaptor proteins CDC20 and CDH1. The APC is a ubiquitin ligase composed of at least 11 subunits. The interaction of APC2 and APC11 with E2 enzymes is sufficient for ubiquitination reactions, but the functions of most other subunits are unknown. RESULTS We have biochemically characterized subcomplexes of the human APC. One subcomplex, containing APC2/11, APC1, APC4, and APC5, can assemble multiubiquitin chains but is unable to bind CDH1 and to ubiquitinate substrates. The other subcomplex contains all known APC subunits except APC2/11. This subcomplex can recruit CDH1 but fails to support any ubiquitination reaction. In vitro, the C termini of CDC20 and CDH1 bind to the closely related TPR subunits APC3 and APC7. Homology modeling predicts that these proteins are similar in structure to the peroxisomal import receptor PEX5, which binds cargo proteins via their C termini. APC activation by CDH1 depends on a conserved C-terminal motif that is also found in CDC20 and APC10. CONCLUSIONS APC1, APC4, and APC5 may connect APC2/11 with TPR subunits. TPR domains in APC3 and APC7 recruit CDH1 to the APC and may thereby bring substrates into close proximity of APC2/11 and E2 enzymes. In analogy to PEX5, the different TPR subunits of the APC might function as receptors that interact with the C termini of regulatory proteins such as CDH1, CDC20, and APC10.

Kleisins: A Superfamily of Bacterial and Eukaryotic SMC Protein Partners

We describe a superfamily of eukaryotic and prokaryotic proteins (kleisins) that includes ScpA, Scc1, Rec8, and Barren. Scc1 interacts with SMC proteins through N- and C-terminal domains to form a ring-like structure. Since these are the only domains conserved among kleisins, we suggest that ring formation with SMC proteins may define this family.

Prediction of peroxisomal targeting signal 1 containing proteins from amino acid sequence.

Peroxisomal matrix proteins have to be imported into their target organelle post-translationally. The major translocation pathway depends on a C-terminal targeting signal, termed PTS1. Our previous analysis of sequence variability in the PTS1 motif revealed that, in addition to the known C-terminal tripeptide, at least nine residues directly upstream are important for signal recognition in the PTS1-Pex5 receptor complex. The refined PTS1 motif description was implemented in a prediction tool composed of taxon-specific functions (metazoa, fungi, remaining taxa), capable of recognising potential PTS1s in query sequences. The composite score function consists of classical profile terms and additional terms penalising deviations from the derived physical property pattern over sequence segments. The prediction algorithm has been validated with a self-consistency and three different cross-validation tests. Additionally, we tested the tool on a large set of non-peroxisomal negatives, on mutation data, and compared the prediction rate to the PTS1 component of the PSORT2 program. The sensitivity of our predictor in recognising documented PTS1 signal containing proteins is close to 90% for reliable prediction. The predictor distinguishes even SKL-appended non-peroxisomally targeted proteins such as a mouse dihydrofolate reductase-SKL construct. The corresponding rate of false positives is not worse than 0.8%; thus, the tool can be applied for large-scale unsupervised sequence database annotation. A scan of public protein databases uncovered a number of yet uncharacterised proteins for which the PTS1 signal might be critical for biological function. The predicted presence of a PTS1 signal implies peroxisomal localisation in the absence of N-terminal targeting sequences such as the mitochondrial import signal.

Motif Refinement of the Peroxisomal Targeting Signal 1 and Evaluation of Taxon-specific Differences

Eukaryote peroxisomes, plant glyoxysomes and trypanosomal glycosomes belong to the microbody family of organelles that compartmentalise a variety of biochemical processes. The interaction between the PTS1 signal and its cognate receptor Pex5 initiates the major import mechanism for proteins into the matrix of these organelles. Relying on the analysis of amino acid sequence variability of known PTS1-targeted proteins and PTS1-containing peptides that interact with Pex5 in the yeast two-hybrid assay, on binding site studies of the Pex5-ligand complex crystal structure, 3D models and sequences of Pex5 proteins from various taxa, we derived the requirements for a C-terminal amino acid sequence to interact productively with Pex5. We found evidence that, at least the 12 C-terminal residues of a given substrate protein are implicated in PTS1 signal recognition. This motif can be structurally and functionally divided into three regions: (i) the C-terminal tripeptide, (ii) a region interacting with the surface of Pex5 (about four residues further upstream), and (iii) a polar, solvent-accessible and unstructured region with linker function (the remaining five residues). Specificity differences are confined to taxonomic subgroups (metazoa and fungi) and are connected with amino acid type preferences in region 1 and deviating hydrophobicity patterns in region 2.

Refinement and prediction of protein prenylation motifs

We refined the motifs for carboxy-terminal protein prenylation by analysis of known substrates for farnesyltransferase (FT), geranylgeranyltransferase I (GGT1) and geranylgeranyltransferase II (GGT2). In addition to the CaaX box for the first two enzymes, we identify a preceding linker region that appears constrained in physicochemical properties, requiring small or flexible, preferably hydrophilic, amino acids. Predictors were constructed on the basis of sequence and physical property profiles, including interpositional correlations, and are available as the Prenylation Prediction Suite (PrePS, http://mendel.imp.univie.ac.at/sat/PrePS) which also allows evaluation of evolutionary motif conservation. PrePS can predict partially overlapping substrate specificities, which is of medical importance in the case of understanding cellular action of FT inhibitors as anticancer and anti-parasite agents.

SNPeffect 4.0: on-line prediction of molecular and structural effects of protein-coding variants

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.

Progranulin genetic variability contributes to amyotrophic lateral sclerosis.

Objectives: Null mutations in progranulin (PGRN) cause ubiquitin-positive frontotemporal dementia (FTD) linked to chromosome 17q21 (FTDU-17). Here we examined PGRN genetic variability in amyotrophic lateral sclerosis (ALS), a neurodegenerative motor neuron disease that overlaps with FTD at a clinical, pathologic, and epidemiologic level. Methods: We sequenced all exons, exon-intron boundaries, and 5′ and 3′ regulatory regions of PGRN in a Belgian sample of 230 patients with ALS. The frequency of observed genetic variants was determined in 436 healthy control individuals. The contribution of eight frequent polymorphisms to ALS risk, onset age, and survival was assessed in an association study in the Belgian sample and a replication series of 308 Dutch patients with ALS and 345 Dutch controls. Results: In patients with ALS we identified 11 mutations, 5 of which were predicted to affect PGRN protein sequence or levels (four missense mutations and one 5′ regulatory variant). Moreover, common variants (rs9897526, rs34424835, and rs850713) and haplotypes were significantly associated with a reduction in age at onset and a shorter survival after onset of ALS in both the Belgian and the Dutch studies. Conclusion: PGRN acts as a modifier of the course of disease in patients with amyotrophic lateral sclerosis, through earlier onset and shorter survival.

Genetic variability in progranulin contributes to risk for clinically diagnosed Alzheimer disease

Objective: Loss-of-function mutations in the progranulin gene (PGRN) were identified in frontotemporal lobar degeneration (FTLD) with ubiquitin-immunoreactive neuronal inclusions (FTLD-U). We assessed whether PGRN also contributes to genetic risk for Alzheimer disease (AD) in an extended Belgian AD patient group (n = 779, onset age 74.7 ± 8.7 years). Methods: A mutation analysis of the PGRN coding region was performed. The effect of missense mutations was assessed using in silico predictions and protein modeling. Risk effects of common genetic variants were estimated by logistic regression analysis and gene-based haplotype association analysis. Results: We observed seven missense mutations in eight patients (1.3%). Convincing pathogenic evidence was obtained for two missense mutations, p.Cys139Arg and p.Pro451Leu, affecting PGRN protein folding and leading to loss of PGRN by degradation of the misfolded protein. In addition, we showed that PGRN haplotypes were associated with increased risk for AD. Conclusions: Our data support a role for PGRN in patients with clinically diagnosed Alzheimer disease (AD). Further, we hypothesize that at least some PGRN missense mutations might lead to loss of functional protein. Whether the underlying pathology in our cases proves to be AD, frontotemporal lobar degeneration, or a combination of the two must await further investigations.

Design of multistable RNA molecules

We show that the problem of designing RNA sequences that can fold into multiple stable secondary structures can be transformed into a combinatorial optimization problem that can be solved by means of simple heuristics. Hence it is feasible to design RNA switches with prescribed structural alternatives. We discuss the theoretical background and present an efficient tool that allows the design of various types of switches. We argue that both the general properties of the sequence structure map of RNA secondary structures and the ease with which our design tool finds bistable RNAs strongly indicates that RNA switches are easily accessible in evolution. Thus conformational switches are yet another function for which RNA can be employed.