The effect of amyloidogenic peptides on bacterial aging correlates with their intrinsic aggregation propensity

The formation of aggregates by misfolded proteins is thought to be inherently toxic, affecting cell fitness. This observation has led to the suggestion that selection against protein aggregation might be a major constraint on protein evolution. The precise fitness cost associated with protein aggregation has been traditionally difficult to evaluate. Moreover, it is not known if the detrimental effect of aggregates on cell physiology is generic or depends on the specific structural features of the protein deposit. In bacteria, the accumulation of intracellular protein aggregates reduces cell reproductive ability, promoting cellular aging. Here, we exploit the cell division defects promoted by the intracellular aggregation of Alzheimer's-disease-related amyloid β peptide in bacteria to demonstrate that the fitness cost associated with protein misfolding and aggregation is connected to the protein sequence, which controls both the in vivo aggregation rates and the conformational properties of the aggregates. We also show that the deleterious impact of protein aggregation on bacterial division can be buffered by molecular chaperones, likely broadening the sequential space on which natural selection can act. Overall, the results in the present work have potential implications for the evolution of proteins and provide a robust system to experimentally model and quantify the impact of protein aggregation on cell fitness.     

Understanding the Origins of Loss of Protein Function by Analyzing the Effects of Thousands of Variants on Activity and Abundance

Understanding and predicting how amino acid substitutions affect proteins are keys to our basic understanding of protein function and evolution. Amino acid changes may affect protein function in a number of ways including direct perturbations of activity or indirect effects on protein folding and stability. We have analyzed 6,749 experimentally determined variant effects from multiplexed assays on abundance and activity in two proteins (NUDT15 and PTEN) to quantify these effects and find that a third of the variants cause loss of function, and about half of loss-of-function variants also have low cellular abundance. We analyze the structural and mechanistic origins of loss of function and use the experimental data to find residues important for enzymatic activity. We performed computational analyses of protein stability and evolutionary conservation and show how we may predict positions where variants cause loss of activity or abundance. In this way, our results link thermodynamic stability and evolutionary conservation to experimental studies of different properties of protein fitness landscapes.     

Kinetics of amyloid aggregation of mammal apomyoglobins and correlation with their amino acid sequences

In protein deposition disorders, a normally soluble protein is deposited as insoluble aggregates, referred to as amyloid. The intrinsic effects of specific mutations on the rates of protein aggregation and amyloid formation of unfolded polypeptide chains can be correlated with changes in hydrophobicity, propensity to convert alpha-helical to beta sheet conformation and charge. In this paper, we report the aggregation rates of buffalo, horse and bovine apomyoglobins. The experimental values were compared with the theoretical ones evaluated considering the amino acid differences among the sequences. Our results show that the mutations which play critical roles in the rate-determining step of apomyoglobin aggregation are those located within the N-terminal region of the molecule.     

How evolutionary pressure against protein aggregation shaped chaperone specificity

As protein aggregation is potentially lethal, control of protein conformation by molecular chaperones is essential for cellular organisms. This is especially important during protein expression and translocation, since proteins are then unfolded and therefore most susceptible to form non-native interactions. Using TANGO, a statistical mechanics algorithm to predict protein aggregation, we here analyse the aggregation propensities of 28 complete proteomes. Our results show that between 10% and 20% of the residues in these proteomes are within aggregating protein segments and that this represents a lower limit for the aggregation tendency of globular proteins. Further, we show that not only evolution strongly pressurizes aggregation downwards by minimizing the amount of strongly aggregating sequences but also by selectively capping strongly aggregating hydrophobic protein sequences with arginine, lysine and proline. These residues are strongly favoured at these positions as they function as gatekeepers that are most efficient at opposing aggregation. Finally, we demonstrate that the substrate specificity of different unrelated chaperone families is geared by these gatekeepers. Chaperones face the difficulty of having to combine substrate affinity for a broad range of hydrophobic sequences and selectivity for those hydrophobic sequences that aggregate most strongly. We show that chaperones achieve these requirements by preferentially binding hydrophobic sequences that are capped by positively charged gatekeeper residues. In other words, targeting evolutionarily selected gatekeepers allows chaperones to prioritize substrate recognition according to aggregation propensity.     

The rate of polyQ-mediated aggregation is dramatically affected by the number and location of surrounding domains

The nine polyglutamine (polyQ) neurodegenerative diseases are caused in part by a gain-of-function mechanism involving protein misfolding, the deposition of β-sheet-rich aggregates and neuronal toxicity. While previous experimental evidence suggests that the polyQ-induced misfolding mechanism is context dependent, the properties of the host protein, including the domain architecture and location of the polyQ tract, have not been investigated. Here, we use variants of a model polyQ-containing protein to systematically determine the effect of the location and number of flanking folded domains on polyQ-mediated aggregation. Our data indicate that when a pathological-length polyQ tract is present between two domains, it aggregates more slowly than the same-length tract in a terminal location within the protein. We also demonstrate that increasing the number of flanking domains decreases the polyQ protein's aggregation rate. Our experimental data, together with a bioinformatic analysis of all human proteins possessing polyQ tracts, suggest that repeat location and protein domain architecture affect the disease susceptibility of human polyQ proteins.     

Derivation of a solubility condition for proteins from an analysis of the competition between folding and aggregation

Failure in maintaining protein solubility in vivo impairs protein homeostasis and results in protein misfolding and aggregation, which are often associated with severe neurodegenerative and systemic disorders that include Alzheimer's and Parkinson's diseases and type II diabetes. In this work we formulate a model of the competition between folding and aggregation, and derive a condition on the solubility of proteins in terms of the stability of their folded states, their aggregation propensities and their degradation rates. From our model, the bistability between folding and aggregation emerges as an intrinsic aspect of protein homeostasis. The analysis of the conditions that determine such a bistability provides a rationalization of the recently observed relationship between the cellular abundance and the aggregation propensity of proteins. We then discuss how the solubility condition that we derive can help rationalise the correlation that has been reported between evolutionary rates and expression levels or proteins, as well as in vivo protein solubility and expression level measurements, and recently elucidated trends of proteome evolution.     

Association of alphaB-crystallin, a small heat shock protein, with actin: role in modulating actin filament dynamics in vivo

Disruption of cytoskeletal assembly is one of the early effects of any stress that can ultimately lead to cell death. Stabilization of cytoskeletal assembly, therefore, is a critical event that regulates cell survival under stress. alphaB-crystallin, a small heat shock protein, has been shown to associate with cytoskeletal proteins under normal and stress conditions. Earlier reports suggest that alphaB-crystallin could prevent stress-induced aggregation of actin in vitro. However, the molecular mechanisms by which alphaB-crystallin stabilizes actin filaments in vivo are not known. Using the H9C2 rat cardiomyoblast cell line as a model system, we show that upon heat stress, alphaB-crystallin preferentially partitions from the soluble cytosolic fraction to the insoluble cytoskeletal protein-rich fraction. Confocal microscopic analysis shows that alphaB-crystallin associates with actin filaments during heat stress and the extent of association increases with time. Further, immunoprecipitation experiments show that alphaB-crystallin interacts directly with actin. Treatment of heat-stressed H9C2 cells with the actin depolymerzing agent, cytochalasin B, failed to disorganize actin. We show that this association of alphaB-crystallin with actin is dependent on its phosphorylation status, as treatment of cells with MAPK inhibitors SB202190 or PD98059 results in abrogation of this association. Our results indicate that alphaB-crystallin regulates actin filament dynamics in vivo and protects cells from stress-induced death. Further, our studies suggest that the association of alphaB-crystallin with actin helps maintenance of pinocytosis, a physiological function essential for survival of cells.     

Kinetic model for the coupling between allosteric transitions in GroEL and substrate protein folding and aggregation

The bacterial chaperonin GroEL and the co-chaperonin GroES assist in the folding of a number of structurally unrelated substrate proteins (SPs). In the absence of chaperonins, SP folds by the kinetic partitioning mechanism (KPM), according to which a fraction of unfolded molecules reaches the native state directly, while the remaining fraction gets trapped in a potentially aggregation-prone misfolded state. During the catalytic reaction cycle, GroEL undergoes a series of allosteric transitions (T<-->R-->R"-->T) triggered by SP capture, ATP binding and hydrolysis, and GroES binding. We developed a general kinetic model that takes into account the coupling between the rates of the allosteric transitions and the folding and aggregation of the SP. Our model, in which the GroEL allosteric rates and SP-dependent folding and aggregation rates are independently varied without prior assumption, quantitatively fits the GroEL concentration-dependent data on the yield of native ribulose bisphosphate carboxylase/oxygenase (Rubisco) as a function of time. The extracted kinetic parameters for the GroEL reaction cycle are consistent with the available values from independent experiments. In addition, we also obtained physically reasonable parameters for the kinetic steps in the reaction cycle that are difficult to measure. If experimental values for GroEL allosteric rates are used, the time-dependent changes in native-state yield at eight GroEL concentrations can be quantitatively fit using only three SP-dependent parameters. The model predicts that the differences in the efficiencies (as measured by yields of the native state) of GroEL, single-ring mutant (SR1), and variants of SR1, in the rescue of mitochondrial malate dehydrogenase, citrate synthase, and Rubisco, are related to the large variations in the allosteric transition rates. We also show that GroEL/S mutants that efficiently fold one SP at the expense of all others are due to a decrease in the rate of a key step in the reaction cycle, which implies that wild-type GroEL has evolved as a compromise between generality and specificity. We predict that, under maximum loading conditions and saturating ATP concentration, the efficiency of GroEL (using parameters for Rubisco) depends predominantly on the rate of R-->R" transition, while the equilibrium constant of the T<-->R has a small effect only. Both under sub- and superstoichiometric GroEL concentrations, enhanced efficiency is achieved by rapid turnover of the reaction cycle, which is in accord with the predictions of the iterative annealing mechanism. The effects are most dramatic at substoichiometric conditions (most relevant for in vivo situations) when SP aggregation can outcompete capture of SP by chaperonins.     

Codon Size Reduction as the Origin of the Triplet Genetic Code

The genetic code appears to be optimized in its robustness to missense errors and frameshift errors. In addition, the genetic code is near-optimal in terms of its ability to carry information in addition to the sequences of encoded proteins. As evolution has no foresight, optimality of the modern genetic code suggests that it evolved from less optimal code variants. The length of codons in the genetic code is also optimal, as three is the minimal nucleotide combination that can encode the twenty standard amino acids. The apparent impossibility of transitions between codon sizes in a discontinuous manner during evolution has resulted in an unbending view that the genetic code was always triplet. Yet, recent experimental evidence on quadruplet decoding, as well as the discovery of organisms with ambiguous and dual decoding, suggest that the possibility of the evolution of triplet decoding from living systems with non-triplet decoding merits reconsideration and further exploration. To explore this possibility we designed a mathematical model of the evolution of primitive digital coding systems which can decode nucleotide sequences into protein sequences. These coding systems can evolve their nucleotide sequences via genetic events of Darwinian evolution, such as point-mutations. The replication rates of such coding systems depend on the accuracy of the generated protein sequences. Computer simulations based on our model show that decoding systems with codons of length greater than three spontaneously evolve into predominantly triplet decoding systems. Our findings suggest a plausible scenario for the evolution of the triplet genetic code in a continuous manner. This scenario suggests an explanation of how protein synthesis could be accomplished by means of long RNA-RNA interactions prior to the emergence of the complex decoding machinery, such as the ribosome, that is required for stabilization and discrimination of otherwise weak triplet codon-anticodon interactions.     

Estimating the fitness effect of an insertion sequence

Since its discovery, mobile DNA has fascinated researchers. In particular, many researchers have debated why insertion sequences persist in prokaryote genomes and populations. While some authors think that insertion sequences persist only because of occasional beneficial effects they have on their hosts, others argue that horizontal gene transfer is strong enough to overcome their generally detrimental effects. In this study, we model the long-term fate of a prokaryote cell population, of which a small proportion of cells has been infected with one insertion sequence per cell. Based on our model and the distribution of IS5, an insertion sequence for which sufficient data is available in 525 fully sequenced proteobacterial genomes, we show that the fitness cost of insertion sequences is so small that they are effectively neutral or only slightly detrimental. We also show that an insertion sequence infection can persist and reach the empirically observed distribution if the rate of horizontal gene transfer is at least as large as the fitness cost, and that this rate is well within the rates of horizontal gene transfer observed in nature. In addition, we show that the time needed to reach the observed prevalence of IS5 is unrealistically long for the fitness cost and horizontal gene transfer rate that we computed. Occasional beneficial effects may thus have played an important role in the fast spreading of insertion sequences like IS5.     

Neural networks enable efficient and accurate simulation-based inference of evolutionary parameters from adaptation dynamics

The rate of adaptive evolution depends on the rate at which beneficial mutations are introduced into a population and the fitness effects of those mutations. The rate of beneficial mutations and their expected fitness effects is often difficult to empirically quantify. As these 2 parameters determine the pace of evolutionary change in a population, the dynamics of adaptive evolution may enable inference of their values. Copy number variants (CNVs) are a pervasive source of heritable variation that can facilitate rapid adaptive evolution. Previously, we developed a locus-specific fluorescent CNV reporter to quantify CNV dynamics in evolving populations maintained in nutrient-limiting conditions using chemostats. Here, we use CNV adaptation dynamics to estimate the rate at which beneficial CNVs are introduced through de novo mutation and their fitness effects using simulation-based likelihood–free inference approaches. We tested the suitability of 2 evolutionary models: a standard Wright–Fisher model and a chemostat model. We evaluated 2 likelihood-free inference algorithms: the well-established Approximate Bayesian Computation with Sequential Monte Carlo (ABC-SMC) algorithm, and the recently developed Neural Posterior Estimation (NPE) algorithm, which applies an artificial neural network to directly estimate the posterior distribution. By systematically evaluating the suitability of different inference methods and models, we show that NPE has several advantages over ABC-SMC and that a Wright–Fisher evolutionary model suffices in most cases. Using our validated inference framework, we estimate the CNV formation rate at the GAP1 locus in the yeast Saccharomyces cerevisiae to be 10^−4.7 to 10⁻⁴ CNVs per cell division and a fitness coefficient of 0.04 to 0.1 per generation for GAP1 CNVs in glutamine-limited chemostats. We experimentally validated our inference-based estimates using 2 distinct experimental methods—barcode lineage tracking and pairwise fitness assays—which provide independent confirmation of the accuracy of our approach. Our results are consistent with a beneficial CNV supply rate that is 10-fold greater than the estimated rates of beneficial single-nucleotide mutations, explaining the outsized importance of CNVs in rapid adaptive evolution. More generally, our study demonstrates the utility of novel neural network–based likelihood–free inference methods for inferring the rates and effects of evolutionary processes from empirical data with possible applications ranging from tumor to viral evolution.

This study shows that simulation-based inference of evolutionary dynamics using neural networks can yield parameter values for fitness and mutation rate that are difficult to determine experimentally, including those of copy number variants (CNVs) during experimental adaptive evolution of yeast.     

Dynamic RNA Fitness Landscapes of a Group I Ribozyme during Changes to the Experimental Environment

Fitness landscapes of protein and RNA molecules can be studied experimentally using high-throughput techniques to measure the functional effects of numerous combinations of mutations. The rugged topography of these molecular fitness landscapes is important for understanding and predicting natural and experimental evolution. Mutational effects are also dependent upon environmental conditions, but the effects of environmental changes on fitness landscapes remains poorly understood. Here, we investigate the changes to the fitness landscape of a catalytic RNA molecule while changing a single environmental variable that is critical for RNA structure and function. Using high-throughput sequencing of in vitro selections, we mapped a fitness landscape of the Azoarcus group I ribozyme under eight different concentrations of magnesium ions (1–48 mM MgCl₂). The data revealed the magnesium dependence of 16,384 mutational neighbors, and from this, we investigated the magnesium induced changes to the topography of the fitness landscape. The results showed that increasing magnesium concentration improved the relative fitness of sequences at higher mutational distances while also reducing the ruggedness of the mutational trajectories on the landscape. As a result, as magnesium concentration was increased, simulated populations evolved toward higher fitness faster. Curve-fitting of the magnesium dependence of individual ribozymes demonstrated that deep sequencing of in vitro reactions can be used to evaluate the structural stability of thousands of sequences in parallel. Overall, the results highlight how environmental changes that stabilize structures can also alter the ruggedness of fitness landscapes and alter evolutionary processes.     

Structural dynamics shape the fitness window of alanine:glyoxylate aminotransferase

The conformational landscape of a protein is constantly expanded by genetic variations that have a minimal impact on the function(s) while causing subtle effects on protein structure. The wider the conformational space sampled by these variants, the higher the probabilities to adapt to changes in environmental conditions. However, the probability that a single mutation may result in a pathogenic phenotype also increases. Here we present a paradigmatic example of how protein evolution balances structural stability and dynamics to maximize protein adaptability and preserve protein fitness. We took advantage of known genetic variations of human alanine:glyoxylate aminotransferase (AGT1), which is present as a common major allelic form (AGT‐Ma) and a minor polymorphic form (AGT‐Mi) expressed in 20% of Caucasian population. By integrating crystallographic studies and molecular dynamics simulations, we show that AGT‐Ma is endowed with structurally unstable (frustrated) regions, which become disordered in AGT‐Mi. An in‐depth biochemical characterization of variants from an anticonsensus library, encompassing the frustrated regions, correlates this plasticity to a fitness window defined by AGT‐Ma and AGT‐Mi. Finally, co‐immunoprecipitation analysis suggests that structural frustration in AGT1 could favor additional functions related to protein–protein interactions. These results expand our understanding of protein structural evolution by establishing that naturally occurring genetic variations tip the balance between stability and frustration to maximize the ensemble of conformations falling within a well‐defined fitness window, thus expanding the adaptability potential of the protein.     

Expression profiling of abundant genes in pulmonary and cardiac muscle tissues of Tibetan Antelope (Pantholops hodgsonii)

The Tibetan Antelope (TA), which has lived at high altitude for millions of years, was selected as the model species of high hypoxia-tolerant adaptation. Here we constructed two cDNA libraries from lung and cardiac muscle tissues, obtained EST sequences from the libraries, and acquired extensive expression data related energy metabolism genes. Comparative analyses of synonymous (Ks) and nonsynonymous (Ka) substitution rates of nucleus-encoded mitochondrial unigenes among different species revealed that many antelope genes have undergone rapid evolution. Surfactant-associated protein A (SP-A) and surfactant-associated protein B (SP-B) genes in the AT lineage experienced accelerated evolution compared to goat and sheep, and these two genes are highly expressed in the lung tissue. This study suggests that many specific genes of lung and cardiac muscle tissues showed unique expression profiles and may undergo fast adaptive evolution in TA. These data provide useful information for studying on molecular adaptation to high-altitude in humans as well as other mammals.     

Unsupervised Inference of Protein Fitness Landscape from Deep Mutational Scan

The recent technological advances underlying the screening of large combinatorial libraries in high-throughput mutational scans deepen our understanding of adaptive protein evolution and boost its applications in protein design. Nevertheless, the large number of possible genotypes requires suitable computational methods for data analysis, the prediction of mutational effects, and the generation of optimized sequences. We describe a computational method that, trained on sequencing samples from multiple rounds of a screening experiment, provides a model of the genotype–fitness relationship. We tested the method on five large-scale mutational scans, yielding accurate predictions of the mutational effects on fitness. The inferred fitness landscape is robust to experimental and sampling noise and exhibits high generalization power in terms of broader sequence space exploration and higher fitness variant predictions. We investigate the role of epistasis and show that the inferred model provides structural information about the 3D contacts in the molecular fold.     

Relationship between energy distribution and fold stability: Insights from molecular dynamics simulations of native and mutant proteins

Most proteins must fold to a well-defined structure with a minimal stability to perform their function. Here we use a simple, molecular dynamics-based, energy decomposition approach to map the principal energetic interactions in a set of proteins representative of different folds. This work involves the all-atom simulation and analysis of the native structures and mutants of five different proteins representative of an all-alpha (yACPB, Protein A), all-beta (SH3), and a mixed alpha/beta fold (Proteins G and L). Given a certain structure, a native sequence and a set of mutants, we show that our model discriminates the ability of a mutation to yield a more or less stable protein, in agreement with experimental data, catching the principal energetic determinants of protein stabilization. Our approach identifies the interaction determinants responsible to define a fold and shows that mutations can either modulate the strength of pair-wise coupling between residues important for folding, or modify the profile of the principal interactions. Furthermore, we address the question of how to evaluate the fitness of a sequence to a given structure by comparing the information contained in the energy map, which recapitulates the chemistry of the sequence, to that contained in the contact map, which recapitulates the fold topology. The results show that the better fit between the energetic properties of the sequence and the fold topology corresponds to a higher stabilization of the protein. We discuss the relevance of these observations to the analysis of protein designability and to the rational evolution of new sequences.     

Basing population genetic inferences and models of molecular evolution upon desired stationary distributions of DNA or protein sequences

Models of molecular evolution tend to be overly simplistic caricatures of biology that are prone to assigning high probabilities to biologically implausible DNA or protein sequences. Here, we explore how to construct time-reversible evolutionary models that yield stationary distributions of sequences that match given target distributions. By adopting comparatively realistic target distributions,evolutionary models can be improved. Instead of focusing on estimating parameters, we concentrate on the population genetic implications of these models. Specifically, we obtain estimates of the product of effective population size and relative fitness difference of alleles. The approach is illustrated with two applications to protein-coding DNA. In the first, a codon-based evolutionary model yields a stationary distribution of sequences, which, when the sequences are translated,matches a variable-length Markov model trained on human proteins. In the second, we introduce an insertion-deletion model that describes selectively neutral evolutionary changes to DNA. We then show how to modify the neutral model so that its stationary distribution at the amino acid level can match a profile hidden Markov model, such as the one associated with the Pfam database.     

Protein evolution is faster outside the cell

Some proteins are highly conserved across all species, whereas others diverge significantly even between closely related species. Attempts have been made to correlate the rate of protein evolution to amino acid composition, protein dispensability, and the number of protein-protein interactions, but in all cases, conflicting studies have shown that the theories are hard to confirm experimentally. The only correlation that is undisputed so far is that highly/broadly expressed proteins seem to evolve at a lower rate. Consequently, it has been suggested that correlations between evolution rate and factors like protein dispensability or the number of protein-protein interactions could be just secondary effects due to differences in expression. The purpose of this study was to analyze mammalian proteins/genes with known subcellular location for variations in evolution rates. We show that proteins that are exported (extracellular proteins) evolve faster than proteins that reside inside the cell (intracellular proteins). We find weak, but significant, correlations between evolution rates and expression levels, percentage of tissues in which the proteins are expressed (expression broadness), and the number of protein interaction partners. More important, we show that the observed difference in evolution rate between extra- and intracellular proteins is largely independent of expression levels, expression broadness, and the number of protein-protein interactions. We also find that the difference is not caused by an overrepresentation of immunological proteins or disulfide bridge-containing proteins among the extracellular data set. We conclude that the subcellular location of a mammalian protein has a larger effect on its evolution rate than any of the other factors studied in this paper, including expression levels/patterns. We observe a difference in evolution rates between extracellular and intracellular proteins for a yeast data set as well and again show that it is completely independent of expression levels.