Search

Interactions between genetic perturbations and segregating loci can cause perturbations to show different phenotypic effects across genetically distinct individuals. To study these interactions on a genome scale in many individuals, we used combinatorial DNA barcode sequencing to measure the fitness effects of 8,013 CRISPRi perturbations targeting 1,721 distinct genes in 169 yeast cross progeny (or segregants). We identified 457 genes whose perturbation has different effects across segregants. Several factors caused perturbations to show variable effects, including baseline segregant fitness, the mean effect of a perturbation across segregants, and interacting loci. We mapped 248 interacting loci and found four hub loci that interact with many different perturbations. Perturbations that interact with a given hub exhibit similar epistatic relationships with the hub and show enrichment for cellular processes that may mediate these interactions. These results suggest that an individual’s response to perturbations is shaped by a network of perturbation-locus interactions that cannot be measured by approaches that examine perturbations or natural variation alone.

Determining how genetic polymorphisms enable certain fungi to persist in mammalian hosts can improve understanding of opportunistic fungal pathogenesis, a source of substantial human morbidity and mortality. We examined the genetic basis of fungal persistence in mice using a cross between a clinical isolate and the lab reference strain of the budding yeastSaccharomyces cerevisiae. Employing chromosomally-encoded barcodes, we tracked the relative abundances of 822 genotyped, haploid segregants in multiple organs over time and performed linkage mapping of their persistence in hosts. Detected loci showed a mix of general and antagonistically pleiotropic effects across organs. General loci showed similar effects across all organs, while antagonistically pleiotropic loci showed contrasting effects in the brain and the kidneys, liver, and spleen. Persistence in an organ required both generally beneficial alleles and organ-appropriate pleiotropic alleles. This genetic architecture resulted in many segregants persisting in the brain or in non-brain organs, but few segregants persisting in all organs. These results show complex combinations of genetic polymorphisms collectively cause and constrain fungal persistence in different parts of the mammalian body.

As synthetic biology expands and accelerates into real-world applications, methods for quantitatively and precisely engineering biological function become increasingly relevant. This is particularly true for applications that require programmed sensing to dynamically regulate gene expression in response to stimuli. However, few methods have been described that can engineer biological sensing with any level of quantitative precision. Here, we present two complementary methods for precision engineering of genetic sensors:in silicoselection and machine-learning-enabled forward engineering. Both methods use a large-scale genotype-phenotype dataset to identify DNA sequences that encode sensors with quantitatively specified dose response. First, we show thatin silicoselection can be used to engineer sensors with a wide range of dose-response curves. To demonstratein silicoselection for precise, multi-objective engineering, we simultaneously tune a genetic sensor’s sensitivity (EC50) and saturating output to meet quantitative specifications. In addition, we engineer sensors with inverted dose-response and specifiedEC50. Second, we demonstrate a machine-learning-enabled approach to predictively engineer genetic sensors with mutation combinations that are not present in the large-scale dataset. We show that the interpretable machine learning results can be combined with a biophysical model to engineer sensors with improved inverted dose-response curves.

Identification of adaptive targets in experimental evolution typically relies on extensive replication and genetic reconstruction. An alternative approach is to directly assay all mutations in an evolved clone by generating pools of segregants that contain random combinations of evolved mutations. Here, we apply this method to 6 Saccharomyces cerevisiae clones isolated from 4 diploid populations that were clonally evolved for 2,000 generations in rich glucose medium. Each clone contains 17–26 mutations relative to the ancestor. We derived intermediate genotypes between the founder and the evolved clones by bulk mating sporulated cultures of the evolved clones to a barcoded haploid version of the ancestor. We competed the resulting barcoded diploids en masse and quantified fitness in the experimental and alternative environments by barcode sequencing. We estimated average fitness effects of evolved mutations using barcode-based fitness assays and whole-genome sequencing for a subset of segregants. In contrast to our previous work with haploid evolved clones, we find that diploids carry fewer beneficial mutations, with modest fitness effects (up to 5.4%) in the environment in which they arose. In agreement with theoretical expectations, reconstruction experiments show that all mutations with a detectable fitness effect manifest some degree of dominance over the ancestral allele, and most are overdominant. Genotypes with lower fitness effects in alternative environments allowed us to identify conditions that drive adaptation in our system.

Identification of adaptive targets in experimental evolution typically relies on extensive replication and allele reconstructions. An alternative approach is to directly assay all mutations in an evolved clone by generating pools of segregants that contain random combinations of the evolved mutations. Here, we apply this method to 6 clones isolated from 4 diploid populations that were clonally evolved for 2,000 generations in rich glucose medium. Each clone contains ∼20-25 mutations relative to the ancestor. We derived intermediate genotypes between the founder and the evolved clones by bulk mating sporulated cultures of each evolved clone to a barcoded haploid version of the founder. We competed the barcoded segregants en masse and quantified the fitness of each barcode. We estimated average fitness effects of evolved mutations using barcode fitness and whole genome sequencing for a subset of segregants or time-course whole population whole genome sequencing. In contrast to our previous work in haploid populations, we find that diploids carry fewer evolved mutations with a detectable fitness effect (6%), contributing a modest fitness advantage (up to 5.4%). In agreement with theoretical expectations, reconstruction experiments show that all adaptive mutations manifest some degree of dominance over the ancestral allele, and most are overdominant. Competition assays under conditions that deviated from the evolutionary environment show that adaptive mutations are often pleiotropic.

In diploid species, genetic loci can show additive, dominance, and epistatic effects. To characterize the contributions of these different types of genetic effects to heritable traits, we use a double barcoding system to generate and phenotype a panel of ~200,000 diploid yeast strains that can be partitioned into hundreds of interrelated families. This experiment enables the detection of thousands of epistatic loci, many whose effects vary across families. Here, we show traits are largely specified by a small number of hub loci with major additive and dominance effects, and pervasive epistasis. Genetic background commonly influences both the additive and dominance effects of loci, with multiple modifiers typically involved. The most prominent dominance modifier in our data is the mating locus, which has no effect on its own. Our findings show that the interplay between additivity, dominance, and epistasis underlies a complex genotype-to-phenotype map in diploids.

We used a double barcoding system to generate and phenotype a panel of ~200,000 diploid yeast segregants that can be partitioned into hundreds of interrelated families. This experimental design enabled the detection of thousands of genetic interactions and many loci whose effects vary across families. Traits were largely specified by a small number of hub loci with major additive and dominance effects, and pervasive epistasis. Genetic background commonly influenced both the additive and dominance effects of loci, with multiple modifiers typically involved. The most prominent dominance modifier was the mating locus, which had no effect on its own. Our findings show that the interplay between additivity, dominance, and epistasis underlies a complex genotype-to-phenotype map in diploids.One sentence summaryIn diploids, epistasis frequently modifies both additivity and dominance.

In rapidly adapting asexual populations, including many microbial pathogens and viruses, numerous mutant lineages often compete for dominance within the population1–5. These complex evolutionary dynamics determine the outcomes of adaptation, but have been difficult to observe directly. Previous studies have used whole-genome sequencing to follow molecular adaptation6–10; however, these methods have limited resolution in microbial populations. Here we introduce a renewable barcoding system to observe evolutionary dynamics at high resolution in laboratory budding yeast. We find nested patterns of interference and hitchhiking even at low frequencies. These events are driven by the continuous appearance of new mutations that modify the fates of existing lineages before they reach substantial frequencies. We observe how the distribution of fitness within the population changes over time, and find a travelling wave of adaptation that has been predicted by theory11–17. We show that clonal competition creates a dynamical ‘rich-get-richer’ effect: fitness advantages that are acquired early in evolution drive clonal expansions, which increase the chances of acquiring future mutations. However, less-fit lineages also routinely leapfrog over strains of higher fitness. Our results demonstrate that this combination of factors, which is not accounted for in existing models of evolutionary dynamics, is critical in determining the rate, predictability and molecular basis of adaptation. A renewable barcoding system reveals the evolutionary dynamics of laboratory budding yeast, showing that fitness changes over time in a travelling wave of adaptation that can fluctuate owing to leapfrogging events.

Allostery is a fundamental biophysical mechanism that underlies cellular sensing, signaling, and metabolism. Quantitative methods to characterize the genotype-phenotype relationships for allosteric proteins would provide data needed to improve engineering of biological systems, to uncover the role of allosteric mis-regulation in disease, and to develop allosterically targeted drugs1. Here we report the large-scale measurement of the genotype-phenotype landscape for an allosteric protein: the lac repressor from Escherichia coli, LacI. Using a method that combines long-read and short-read DNA sequencing, we quantitatively determine the dose-response curves for nearly 105 variants of the LacI sensor. With the resulting data, we train a deep neural network (DNN) capable of predicting the dose-response curves for additional LacI genotypes in silico. We then map the impact of amino acid substitutions on the allosteric function of LacI. Additionally, we demonstrate engineering of allosteric function with unprecedented accuracy by identifying LacI variants from the measured landscape with quantitatively specified dose-response curves. Finally, we discover sensors with previously unreported band-stop dose-response curves. Overall, our results provide the first high-coverage, quantitative view of genotype-phenotype relationships for an allosteric protein, revealing a surprising diversity of phenotypes and showing that each phenotype is accessible via multiple distinct genotypes.

To characterize how protein-protein interaction (PPI) networks change, we quantified the relative PPI abundance of 1.6 million protein pairs in the yeast Saccharomyces cerevisiae across nine growth conditions, with replication, for a total of 44 million measurements. Our multi-condition screen identified 13,764 pairwise PPIs, a threefold increase over PPIs identified in one condition. A few ‘immutable’ PPIs are present across all conditions, while most ‘mutable’ PPIs are rarely observed. Immutable PPIs aggregate into highly connected ‘core’ network modules, with most network remodeling occurring within a loosely connected ‘accessory’ module. Mutable PPIs are less likely to co-express, co-localize, and be explained by simple mass action kinetics, and more likely to contain proteins with intrinsically disordered regions, implying that environment-dependent association and binding is critical to cellular adaptation. Our results show that protein interactomes are larger than previously thought and contain highly dynamic regions that reorganize to drive or respond to cellular changes.

Here, we describe a protocol for automated library preparation for DNA barcode sequencing. The protocol uses a Hamilton NGS-STAR automated liquid handler equipped with an automated thermocycler. This protocol requires an automated liquid handler (Hamilton Robotics, NGS-STAR) equipped with an automated thermocycler.

Motivation Barcode sequencing (bar-seq) is a high-throughput, and cost effective method to assay large numbers of cell lineages or genotypes in complex cell pools. Because of its advantages, applications for bar-seq are quickly growing—from using neutral random barcodes to study the evolution of microbes or cancer, to using pseudo-barcodes, such as shRNAs or sgRNAs to simultaneously screen large numbers of cell perturbations. However, the computational pipelines for bar-seq clustering are not well developed. Available methods often yield a high frequency of under-clustering artifacts that result in spurious barcodes, or over-clustering artifacts that group distinct barcodes together. Here, we developed Bartender, an accurate clustering algorithm to detect barcodes and their abundances from raw next-generation sequencing data. Results In contrast with existing methods that cluster based on sequence similarity alone, Bartender uses a modified two-sample proportion test that also considers cluster size. This modification results in higher accuracy and lower rates of under- and over-clustering artifacts. Additionally, Bartender includes unique molecular identifier handling and a ‘multiple time point’ mode that matches barcode clusters between different clustering runs for seamless handling of time course data. Bartender is a set of simple-to-use command line tools that can be performed on a laptop at comparable run times to existing methods. Availability and implementation Bartender is available at no charge for non-commercial use at https://github.com/LaoZZZZZ/bartender-1.1. Supplementary information Supplementary data are available at Bioinformatics online.

Summary Gene expression burdens cells by consuming resources and energy. While numerous studies have investigated regulation of expression level, little is known about gene design elements that govern expression costs. Here, we ask how cells minimize production costs while maintaining a given protein expression level and whether there are gene architectures that optimize this process. We measured fitness of ∼14,000 E. coli strains, each expressing a reporter gene with a unique 5′ architecture. By comparing cost-effective and ineffective architectures, we found that cost per protein molecule could be minimized by lowering transcription levels, regulating translation speeds, and utilizing amino acids that are cheap to synthesize and that are less hydrophobic. We then examined natural E. coli genes and found that highly expressed genes have evolved more forcefully to minimize costs associated with their expression. Our study thus elucidates gene design elements that improve the economy of protein expression in natural and heterologous systems.

Summary We developed a flexible toolkit for combinatorial screening in Saccharomyces cerevisiae, which generates large libraries of cells, each uniquely barcoded to mark a combination of DNA elements. This interaction sequencing platform (iSeq 2.0) includes genomic landing pads that assemble combinations through sequential integration of plasmids or yeast mating, 15 barcoded plasmid libraries containing split selectable markers (URA3AI, KanMXAI, HphMXAI, and NatMXAI), and an array of ∼24,000 “double-barcoder” strains that can make existing yeast libraries iSeq compatible. Various DNA elements are compatible with iSeq: DNA introduced on integrating plasmids, engineered genomic modifications, or entire genetic backgrounds. DNA element libraries are modular and interchangeable, and any two libraries can be combined, making iSeq capable of performing many new combinatorial screens by short-read sequencing.

Large-scale genetic interaction (GI) screens in yeast have been invaluable for our understanding of molecular systems biology and for characterizing novel gene function. Owing in part to the high costs and long experiment times required, a preponderance of GI data has been generated in a single environmental condition. However, an unknown fraction of GIs may be specific to other conditions. Here, we developed a pooled-growth CRISPRi-based sequencing assay for GIs, CRISPRiSeq, which increases throughput such that GIs can be easily assayed across multiple growth conditions. We assayed the fitness of approximately 17,000 strains encompassing approximately 7700 pairwise interactions in five conditions and found that the additional conditions increased the number of GIs detected nearly threefold over the number detected in rich media alone. In addition, we found that condition-specific GIs are prevalent and improved the power to functionally classify genes. Finally, we found new links during respiratory growth between members of the Ras nutrient–sensing pathway and both the COG complex and a gene of unknown function. Our results highlight the potential of conditional GI screens to improve our understanding of cellular genetic networks.

Summary Standard practice for phenotyping complex cell pools is to measure the fold enrichment of genotype-specific amplicons after a period of competitive growth. Here, we show that fold-enrichment measures cannot be compared across genotype pools with different fitness distributions. We develop a method to calculate an unbiased estimate of relative fitness by tracking abundances over several time points and show how to optimize experimental protocols to minimize fitness measurement error.

New technological advances have enabled high-throughput phenotyping at the single-cell level, yet analyzing the large amount of data generated by high throughput phenotyping experiments automatically and accurately is a considerable challenge. Here we introduce Processing Images Easily (PIE), software that automatically tracks growth of microbial colonies in low-magnification brightfield images by combining adaptive object-center recognition with gradient-based object-outline recognition. PIE recognizes colony outlines very robustly and accurately across a wide range of image brightnesses, focal depths, and organisms. Beyond accurate colony recognition, PIE is designed to easily integrate with complex experiments, allowing colony tracking across multiple experimental phases and classification based on fluorescence intensity. We show that PIE can be used to accurately measure the growth rates of large numbers (>90,000) of bacterial or yeast microcolonies in a single-time-lapse experiment, allowing calculation of population-wide growth properties. Finally, PIE is able to track individual colonies across multiple experimental phases, measuring both growth and fluorescence properties of the microcolonies.Author SummaryHigh-throughput microscopy has enabled automated collection of large amounts of growth and gene-expression data in microbes. Computational methods that can precisely recognize and track organisms in images are essential to performing measurements at scale using automated microscopy. We have developed PIE, software that automatically recognizes microbial colonies in microscopy images, tracks them in imaging time-series, and performs measurements of growth and, potentially, gene expression. PIE is highly effective on low-resolution images, outperforming current state-of-the-art approaches in both speed and accuracy, and works well in microbes of varying shapes and sizes. In addition, PIE allows tracking microcolonies across arbitrary sequences of experimental phases, each collecting data in different modalities. We show that PIE allows measurement of growth and fluorescence properties in tens of thousands of microbial colonies in a single experiment, and that in turn the scale of these measurements can lead to important insights about interindividual differences in growth and stress response. PIE is available as a Python package (https://doi.org/10.5281/zenodo.4987127) with documentation currently at https://pie-image.readthedocs.io/; users can also run analysis on individual images or time-series without the need to install PIE by using our web application, currently available at http://pie.hpc.nyu.edu/.

Evolution of large asexual cell populations underlies ∼30% of deaths worldwide, including those caused by bacteria, fungi, parasites, and cancer. However, the dynamics underlying these evolutionary processes remain poorly understood because they involve many competing beneficial lineages, most of which never rise above extremely low frequencies in the population. To observe these normally hidden evolutionary dynamics, we constructed a sequencing-based ultra high-resolution lineage tracking system in Saccharomyces cerevisiae that allowed us to monitor the relative frequencies of ∼500,000 lineages simultaneously. In contrast to some expectations, we found that the spectrum of fitness effects of beneficial mutations is neither exponential nor monotonic. Early adaptation is a predictable consequence of this spectrum and is strikingly reproducible, but the initial small-effect mutations are soon outcompeted by rarer large-effect mutations that result in variability between replicates. These results suggest that early evolutionary dynamics may be deterministic for a period of time before stochastic effects become important.

Growth heterogeneity within microbial populations has been thought to increase fitness by allowing some cells to survive an unpredictable change to a harsh environment. A new study demonstrates that heterogeneity can increase fitness even in the absence of environmental changes.

The dynamics of genetic diversity in large clonally evolving cell populations are poorly understood, despite having implications for the treatment of cancer and microbial infections. Here, we combine barcode lineage tracking, sequencing of adaptive clones and mathematical modelling of mutational dynamics to understand adaptive diversity changes during experimental evolution of Saccharomyces cerevisiae under nitrogen and carbon limitation. We find that, despite differences in beneficial mutational mechanisms and fitness effects, early adaptive genetic diversity increases predictably, driven by the expansion of many single-mutant lineages. However, a crash in adaptive diversity follows, caused by highly fit double-mutant 'jackpot' clones that are fed from exponentially growing single mutants, a process closely related to the classic Luria-Delbruck experiment. The diversity crash is likely to be a general feature of asexual evolution with clonal interference; however, both its timing and magnitude are stochastic and depend on the population size, the distribution of beneficial fitness effects and patterns of epistasis.

The low costs of array‐synthesized oligonucleotide libraries are empowering rapid advances in quantitative and synthetic biology. However, high synthesis error rates, uneven representation, and lack of access to individual oligonucleotides limit the true potential of these libraries. We have developed a cost‐effective method called Recombinase Directed Indexing (REDI), which involves integration of a complex library into yeast, site‐specific recombination to index library DNA, and next‐generation sequencing to identify desired clones. We used REDI to generate a library of ~3,300 DNA probes that exhibited > 96% purity and remarkable uniformity (> 95% of probes within twofold of the median abundance). Additionally, we created a collection of ~9,000 individually accessible CRISPR interference yeast strains for > 99% of genes required for either fermentative or respiratory growth, demonstrating the utility of REDI for rapid and cost‐effective creation of strain collections from oligonucleotide pools. Our approach is adaptable to any complex DNA library, and fundamentally changes how these libraries can be parsed, maintained, propagated, and characterized.

Several large-scale efforts have systematically catalogued protein-protein interactions (PPIs) of a cell in a single environment. However, little is known about how the protein interactome changes across environmental perturbations. Current technologies, which assay one PPI at a time, are too low throughput to make it practical to study protein interactome dynamics. Here, we develop a highly parallel protein-protein interaction sequencing (PPiSeq) platform that uses a novel double barcoding system in conjunction with the dihydrofolate reductase protein-fragment complementation assay in Saccharomyces cerevisiae. PPiSeq detects PPIs at a rate that is on par with current assays and, in contrast with current methods, quantitatively scores PPIs with enough accuracy and sensitivity to detect changes across environments. Both PPI scoring and the bulk of strain construction can be performed with cell pools, making the assay scalable and easily reproduced across environments. PPiSeq is therefore a powerful new tool for large-scale investigations of dynamic PPIs., Protein-protein interactions (PPIs) play a major role in defining biological functions. Here, the authors present PPiSeq, a method to quantitatively score dynamic PPIs in yeast combining fragment complementation assays, genomic double-barcoding, and an analytical framework to precisely call fitness from barcode lineage trajectories.

Barcode sequencing (bar-seq) is a high-throughput, and cost effective method to assay large numbers of lineages or genotypes in complex cell pools. Because of its advantages, applications for bar-seq are quickly growing – from using neutral random barcodes to study the evolution of microbes or cancer, to using pseudo-barcodes, such as shRNAs, sgRNAs, or transposon insertion libraries, to simultaneously screen large numbers of cell perturbations. However, the computational pipelines for bar-seq have not been well developed. Available methods, which use prior information and/or simple brute-force comparisons, are slow and often result in overclustering artifacts that group distinct barcodes together. Here, we developed Bartender: an ultrafast and accurate clustering algorithm to detect barcodes and their abundances from raw next-generation sequencing data. To improve speed and reduce unnecessary pairwise comparisons, Bartender employs a divide-and-conquer strategy that intelligently sorts barcode reads into distinct bins before performing comparisons. To improve accuracy and reduce over-clustering artifacts, Bartender employs a modified two-sample proportion test that uses information on both the cluster sequence distances and cluster sizes to make merging decisions. Additionally, Bartender includes a “multiple time point” mode, which matches barcode clusters between different clustering runs for seamless handling of time course data. For both simulated and real data, Bartender clusters millions of unique barcodes in a few minutes at high accuracy (>99.9%), and is ~100-fold faster than previous methods. Bartender is a set of simple-to-use command line tools that can be performed on a laptop.AvailabilityBartender is available at no charge for non-commercial use at https://github.com/LaoZZZZZ/bartender-1.1.

Copy number variants (CNVs) are a pervasive source of genetic variation and evolutionary potential, but the dynamics and diversity of CNVs within evolving populations remain unclear. Long-term evolution experiments in chemostats provide an ideal system for studying the molecular processes underlying CNV formation and the temporal dynamics with which they are generated, selected, and maintained. Here, we developed a fluorescent CNV reporter to detect de novo gene amplifications and deletions in individual cells. We used the CNV reporter in Saccharomyces cerevisiae to study CNV formation at the GAP1 locus, which encodes the general amino acid permease, in different nutrient-limited chemostat conditions. We find that under strong selection, GAP1 CNVs are repeatedly generated and selected during the early stages of adaptive evolution, resulting in predictable dynamics. Molecular characterization of CNV-containing lineages shows that the CNV reporter detects different classes of CNVs, including aneuploidies, nonreciprocal translocations, tandem duplications, and complex CNVs. Despite GAP1’s proximity to repeat sequences that facilitate intrachromosomal recombination, breakpoint analysis revealed that short inverted repeat sequences mediate formation of at least 50% of GAP1 CNVs. Inverted repeat sequences are also found at breakpoints at the DUR3 locus, where CNVs are selected in urea-limited chemostats. Analysis of 28 CNV breakpoints indicates that inverted repeats are typically 8 nucleotides in length and separated by 40 bases. The features of these CNVs are consistent with origin-dependent inverted-repeat amplification (ODIRA), suggesting that replication-based mechanisms of CNV formation may be a common source of gene amplification. We combined the CNV reporter with barcode lineage tracking and found that 102–104 independent CNV-containing lineages initially compete within populations, resulting in extreme clonal interference. However, only a small number (18–21) of CNV lineages ever constitute more than 1% of the CNV subpopulation, and as selection progresses, the diversity of CNV lineages declines. Our study introduces a novel means of studying CNVs in heterogeneous cell populations and provides insight into their dynamics, diversity, and formation mechanisms in the context of adaptive evolution., A fluorescent reporter system reveals that copy number variants (CNVs) are repeatedly generated and selected during the early stages of adaptive evolution, resulting in initially predictable dynamics with thousands of independent CNV-containing lineages competing within populations., Author summary Duplications and deletions of genomic sequence, known as copy number variants, are a common source of genetic diversity across all domains of life. Copy number variants play a crucial role in driving evolutionary processes but can also cause genetic disease and cancer. Although copy number variants are important drivers of diversity, adaptation, and disease, the underlying dynamics of their formation and selection are poorly understood. Copy number variants are difficult to detect, especially when present at low frequencies in heterogenous evolving populations. To overcome this challenge, we developed a novel fluorescent reporter that allows us to visualize copy number variants as they emerge and to track them throughout hundreds of generations of laboratory evolution. We show that copy number variants arise early and repeatedly, that they are diverse in size and copy number, and that they are generated at a high rate, leading to competition among cells containing different copy number variants. Molecular characterization of copy number variants indicates that many of them are likely generated by errors during DNA replication. This method is broadly applicable to studying the molecular mechanisms underlying formation of copy number variants, as well as their role in driving evolutionary processes and cancer.

Mutations affecting either the structure or regulation of the microtubule-associated protein Tau cause neuronal cell death and dementia. However, the molecular mechanisms mediating these deleterious effects remain unclear. Among the most characterized activities of Tau is the ability to regulate microtubule dynamics, known to be essential for proper cell function and viability. Here we have tested the hypothesis that Tau mutations causing neurodegeneration also alter the ability of Tau to regulate the dynamic instability behaviors of microtubules. Using in vitro microtubule dynamics assays to assess average microtubule growth rates, microtubule growth rate distributions, and catastrophe frequencies, we found that all tested mutants possessing amino acid substitutions or deletions mapping to either the repeat or interrepeat regions of Tau do indeed compromise its ability to regulate microtubule dynamics. Further mutational analyses suggest a novel mechanism of Tau regulatory action based on an "alternative core" of microtubule binding and regulatory activities composed of two repeats and the interrepeat between them. In this model, the interrepeat serves as the primary regulator of microtubule dynamics, whereas the flanking repeats serve as tethers to properly position the interrepeat on the microtubule. Importantly, since there are multiple interrepeats on each Tau molecule, there are also multiple cores on each Tau molecule, each with distinct mechanistic capabilities, thereby providing significant regulatory potential. Taken together, the data are consistent with a microtubule misregulation mechanism for Tau-mediated neuronal cell death and provide a novel mechanistic model for normal and pathological Tau action.

The microtubule-associated protein tau is implicated in the pathogenesis of many neurodegenerative diseases, including fronto-temporal dementia and parkinsonism linked to chromosome 17 (FTDP-17), in which both RNA splicing and amino acid substitution mutations in tau cause dominantly inherited early onset dementia. RNA-splicing FTDP-17 mutations alter the wild-type ∼50:50 3-repeat (3R) to 4-repeat (4R) tau isoform ratio, usually resulting in an excess of 4R tau. To examine further how splicing mutations might cause dysfunction by misregulation of microtubule dynamics, we used video microscopy to determine the in vitro behavior of individual microtubules stabilized by varying amounts of human 4R and 3R tau. At low tau:tubulin ratios (1:55 and 1:45), all 3R isoforms reduced microtubule growth rates relative to the no-tau control, whereas all 4R isoforms increased them; however, at a high tau:tubulin ratio (1:20), both 4R and 3R tau increased the growth rates. Further analysis revealed two distinct subpopulations of growing microtubules in the absence of tau. Increasing concentrations of both 4R and 3R tau resulted in an increase in the size of the faster growing subpopulation of microtubules; however, 4R tau caused a redistribution to the faster growing subpopulation at lower tau:tubulin ratios than 3R tau. This modulation of discrete growth rate subpopulations by tau suggests that tau causes a conformational shift in the microtubule resulting in altered dynamics. Quantitative and qualitative differences observed between 4R and 3R tau are consistent with a "microtubule misregulation" model in which abnormal tau isoform expression results in the inability to properly regulate microtubule dynamics, leading to neuronal death and dementia.

Evolving cellular communities, such as the gut microbiome, pathogenic infections, and cancer, consist of large populations of ~10(7)-10(14) cells. Because of their large population sizes, adaptation within these populations can be driven by many beneficial mutations that never rise above extremely low frequencies. Genome sequencing methods such as clonal, single cell, or whole population sequencing are poorly suited to detect these rare beneficial lineages, and, more generally, to characterize which mutations are most important to the population dynamics. Here, we introduce an alternative approach: high-resolution lineage tracking with DNA barcodes. In contrast to whole genome sequencing, lineage tracking can detect a beneficial mutation at an extremely low frequency within the population, and estimate its time of occurrence and fitness effect. Many lineage trajectories can be observed in parallel, allowing one to observe the population dynamics in exquisite detail. We describe some of the technical and analytical challenges to lineage tracking with DNA barcodes and discuss its applications to studies of evolution, infectious disease and cancer.

Biological systems produce phenotypes that appear to be robust to perturbation by mutations and environmental variation. Prior studies identified genes that, when impaired, reveal previously cryptic genetic variation. This result is typically interpreted as evidence that the disrupted gene normally increases robustness to mutations, as such robustness would allow cryptic variants to accumulate. However, revelation of cryptic genetic variation is not necessarily evidence that a mutationally robust state has been made less robust. Demonstrating a difference in robustness requires comparing the ability of each state (with the gene perturbed or intact) to suppress the effects of new mutations. Previous studies used strains in which the existing genetic variation had been filtered by selection. Here, we use mutation accumulation (MA) lines that have experienced minimal selection, to test the ability of histone H2A.Z (HTZ1) to increase robustness to mutations in the yeast Saccharomyces cerevisiae. HTZ1, a regulator of chromatin structure and gene expression, represents a class of genes implicated in mutational robustness. It had previously been shown to increase robustness of yeast cell morphology to fluctuations in the external or internal microenvironment. We measured morphological variation within and among 79 MA lines with and without HTZ1. Analysis of within-line variation confirms that HTZ1 increases microenvironmental robustness. Analysis of between-line variation shows the morphological effects of eliminating HTZ1 to be highly dependent on the line, which implies that HTZ1 interacts with mutations that have accumulated in the lines. However, lines without HTZ1 are, as a group, not more phenotypically diverse than lines with HTZ1 present. The presence of HTZ1, therefore, does not confer greater robustness to mutations than its absence. Our results provide experimental evidence that revelation of cryptic genetic variation cannot be assumed to be caused by loss of robustness, and therefore force reevaluation of prior claims based on that assumption., Author Summary Natural populations typically harbor much genetic variation. Some of this variation is cryptic — it does not affect observed traits except if the organism is exposed to a major environmental or genetic perturbation. One often-proposed explanation for the revelation of cryptic genetic variation is that the perturbation has made the organisms less robust to mutations, thereby revealing effects of previously neutral mutations that natural selection had allowed to accumulate. Such effects would be dependent on genetic background, as particular cryptic variants will be present in some individuals and not others. We show that a perturbation of chromatin architecture caused by mutation of the budding-yeast gene HTZ1, encoding histone H2A.Z, does alter phenotypes in a genetic background-dependent manner but does not reduce mutational robustness. Using a large set of yeast lines that accumulated mutations with minimal natural selection, we compared variation in cell morphology with and without HTZ1. The effect of eliminating HTZ1 was highly line dependent, suggesting that HTZ1 interacts extensively with genetic variation in the lines. However, HTZ1+ lines span a range of phenotypes similar to that of the corresponding HTZ1− lines. Our results therefore call into question prior studies linking revelation of cryptic genetic variation with reduced mutational robustness.

Adaptive evolution plays a large role in generating the phenotypic diversity observed in nature, yet current methods are impractical for characterizing the molecular basis and fitness effects of large numbers of individual adaptive mutations. Here we used a DNA barcoding approach to generate the genotype-to-fitness map for adaptation-driving mutations from a Saccharomyces cerevisiae population experimentally evolved by serial transfer under limiting glucose. We isolated and measured the fitness of thousands of independent adaptive clones, and sequenced the genomes of hundreds of clones. We found only two major classes of adaptive mutations: self-diploidization, and mutations in the nutrient-responsive Ras/PKA and TOR/Sch9 pathways. Our large sample size and precision of measurement allowed us to determine that there are significant differences in fitness between mutations in different genes, between different paralogs, and even between different classes of mutations within the same gene.

Organisms are subject to random changes in their external environments, as well as in their internal components. A central goal of evolutionary systems biology is to understand how living systems cope with-and in some cases exploit-this variation. Many cellular and developmental processes operate with high fidelity to produce stereotyped, irreversible outcomes despite environmental and genetic perturbation. These processes are said to be robust or insensitive to variation. Robustness can lead to single, invariant phenotypes, or it can take the form of phenotypic plasticity, in which different environmental conditions reproducibly induce distinct phenotypes. Some organisms cope with environmental variation not with robust responses but with stochastic, reversible fate decisions. In those organisms, lower robustness yields heterogeneity among individuals, which in turn serves as a bet-hedging mechanism for the population. Considering high-fidelity and bet-hedging processes together-as a robustness continuum-provides a unifying framework for analyzing and conceptualizing variation in complex evolving systems. This framework can be applied to understanding the architectures and dynamics of the regulatory networks that underlie fate decisions in microbes, plants, animals, and cancer cells.

The acquisition of new genes, via horizontal transfer or gene duplication/diversification, has been the dominant mechanism thus far implicated in the evolution of microbial pathogenicity. In contrast, the role of many other modes of evolution—such as changes in gene expression regulation—remains unknown. A transition to a pathogenic lifestyle has recently taken place in some lineages of the budding yeast Saccharomyces cerevisiae. Here we identify a module of physically interacting proteins involved in endocytosis that has experienced selective sweeps for multiple cis-regulatory mutations that down-regulate gene expression levels in a pathogenic yeast. To test if these adaptations affect virulence, we created a panel of single-allele knockout strains whose hemizygous state mimics the genes' adaptive down-regulations, and measured their virulence in a mammalian host. Despite having no growth advantage in standard laboratory conditions, nearly all of the strains were more virulent than their wild-type progenitor, suggesting that these adaptations likely played a role in the evolution of pathogenicity. Furthermore, genetic variants at these loci were associated with clinical origin across 88 diverse yeast strains, suggesting the adaptations may have contributed to the virulence of a wide range of clinical isolates. We also detected pleiotropic effects of these adaptations on a wide range of morphological traits, which appear to have been mitigated by compensatory mutations at other loci. These results suggest that cis-regulatory adaptation can occur at the level of physically interacting modules and that one such polygenic adaptation led to increased virulence during the evolution of a pathogenic yeast.

A new experimental approach reveals a bet-hedging strategy in unstressed, clonal yeast cells, whereby they adopt a range of growth states that correlate with expression of a trehalose-synthesis regulator and predict resistance to future stress., Genetically identical cells grown in the same culture display striking cell-to-cell heterogeneity in gene expression and other traits. A crucial challenge is to understand how much of this heterogeneity reflects the noise tolerance of a robust system and how much serves a biological function. In bacteria, stochastic gene expression results in cell-to-cell heterogeneity that might serve as a bet-hedging mechanism, allowing a few cells to survive through an antimicrobial treatment while others perish. Despite its clinical importance, the molecular mechanisms underlying bet hedging remain unclear. Here, we investigate the mechanisms of bet hedging in Saccharomyces cerevisiae using a new high-throughput microscopy assay that monitors variable protein expression, morphology, growth rate, and survival outcomes of tens of thousands of yeast microcolonies simultaneously. We find that clonal populations display broad distributions of growth rates and that slow growth predicts resistance to heat killing in a probabalistic manner. We identify several gene products that are likely to play a role in bet hedging and confirm that Tsl1, a trehalose-synthesis regulator, is an important component of this resistance. Tsl1 abundance correlates with growth rate and replicative age and predicts survival. Our results suggest that yeast bet hedging results from multiple epigenetic growth states determined by a combination of stochastic and deterministic factors., Author Summary Genetically identical cells grown in the same environment can display heterogeneity in their morphology, behavior, and composition of their cellular components. In some microorganisms, such cellular heterogeneity can underlie a phenomenon known as bet hedging because it enables some cells to survive in harsh environments, hence increasing the overall population fitness when environmental shifts are unpredictable. Bet hedging is likely to be an important strategy by which microbes infect humans and evade antimicrobial treatments, yet little is known of how cellular heterogeneity contributes to microbial survival. Here, we study the mechanisms underlying bet hedging in yeast. We find that populations of genetically identical yeast contain a broad distribution of growth rates and that slow growth predicts resistance to heat killing in a graded fashion. We identify several gene products that are likely to play a role in this bet-hedging strategy and confirm that Tsl1, a regulator of the production of the disaccharide trehalose, is an important component of acute stress resistance. Finally, we find that old age in cells correlates with a Tsl1-abundant, stress-resistant cell state. Our results suggest that trehalose synthesis is part of a complex and multifactorial mechanism that underlies bet hedging in yeast.

Organisms are subject to random changes in their external environments, as well as in their internal components. A central goal of evolutionary systems biology is to understand how living systems cope with—and in some cases exploit—this variation. Many cellular and developmental processes operate with high fidelity to produce stereotyped, irreversible outcomes despite environmental and genetic perturbation. These processes are said to be robust or insensitive to variation. Robustness can lead to single, invariant phenotypes, or it can take the form of phenotypic plasticity, in which different environmental conditions reproducibly induce distinct phenotypes. Some organisms cope with environmental variation not with robust responses but with stochastic, reversible fate decisions. In those organisms, lower robustness yields heterogeneity among individuals, which in turn serves as a bet-hedging mechanism for the population. Considering high-fidelity and bet-hedging processes together—as a robustness continuum—provides a unifying framework for analyzing and conceptualizing variation in complex evolving systems. This framework can be applied to understanding the architectures and dynamics of the regulatory networks that underlie fate decisions in microbes, plants, animals, and cancer cells.

Background Innumerable biological investigations require comparing collections of molecules, cells or organisms to one another with respect to one or more of their properties. Almost all of these comparisons are performed manually, which can be susceptible to inadvertent bias as well as miss subtle effects. The development and application of computer-assisted analytical and interpretive tools could help address these issues and thereby dramatically improve these investigations. Results We have developed novel computer-assisted analytical and interpretive tools and applied them to recent studies examining the ability of 3-repeat and 4-repeat tau to regulate the dynamic behavior of microtubules in vitro. More specifically, we have developed an automated and objective method to define growth, shortening and attenuation events from real time videos of dynamic microtubules, and demonstrated its validity by comparing it to manually assessed data. Additionally, we have used the same data to develop a general strategy of building different models of interest, computing appropriate dissimilarity functions to compare them, and embedding them on a two-dimensional plot for visualization and easy comparison. Application of these methods to assess microtubule growth rates and growth rate distributions established the validity of the embedding procedure and revealed non-linearity in the relationship between the tau:tubulin molar ratio and growth rate distribution. Conclusion This work addresses the need of the biological community for rigorously quantitative and generally applicable computational tools for comparative studies. The two-dimensional embedding method retains the inherent structure of the data, and yet markedly simplifies comparison between models and parameters of different samples. Most notably, even in cases where numerous parameters exist by which to compare the different samples, our embedding procedure provides a generally applicable computational strategy to detect subtle relationships between different molecules or conditions that might otherwise escape manual analyses.

The catalog of RNAs present in dendrites represents the complete repertoire of local translation that contributes to synaptic plasticity. Most views hold that a pool of dendritic mRNAs is selectively transported to a dendritic destination. This view requires that some mRNAs in the dendrite are locally enriched relative to the cell body; however, quantitative comparisons that would support this assumption do not currently exist. These issues related to somatodendritic distribution of mRNAs also apply to the microRNAs, ∼21 nucleotide noncoding transcripts that bind to target mRNAs and either inhibit their translation or destabilize them. We combined laser capture with multiplex real-time RT (reverse transcription) PCR to quantify microRNAs in the neuritic and somatic compartments separately. The samples were standardized by RT-PCR measurements of a set of mRNAs, including known dendritic mRNAs, in these two compartments. Most neuronal miRNAs were detected in dendrites. With a few notable exceptions, most miRNAs were distributed through the somatodendritic compartment across a nearly constant gradient. Thus for lower-abundance miRNAs, the total neuronal concentration of the miRNA can remain readily detectable in the cell body but vanish from the dendrite. A very small number of miRNAs deviate from the distribution gradient across the miRNA population as relatively enriched or depleted in the dendrite.

The microtubule-associated protein tau is implicated in the pathogenesis of many neurodegenerative diseases, including fronto-temporal dementia and parkinsonism linked to chromosome 17 (FTDP-17), in which both RNA splicing and amino acid substitution mutations in tau cause dominantly inherited early onset dementia. RNA-splicing FTDP-17 mutations alter the wild-type approximately 50:50 3-repeat (3R) to 4-repeat (4R) tau isoform ratio, usually resulting in an excess of 4R tau. To examine further how splicing mutations might cause dysfunction by misregulation of microtubule dynamics, we used video microscopy to determine the in vitro behavior of individual microtubules stabilized by varying amounts of human 4R and 3R tau. At low tau:tubulin ratios (1:55 and 1:45), all 3R isoforms reduced microtubule growth rates relative to the no-tau control, whereas all 4R isoforms increased them; however, at a high tau:tubulin ratio (1:20), both 4R and 3R tau increased the growth rates. Further analysis revealed two distinct subpopulations of growing microtubules in the absence of tau. Increasing concentrations of both 4R and 3R tau resulted in an increase in the size of the faster growing subpopulation of microtubules; however, 4R tau caused a redistribution to the faster growing subpopulation at lower tau:tubulin ratios than 3R tau. This modulation of discrete growth rate subpopulations by tau suggests that tau causes a conformational shift in the microtubule resulting in altered dynamics. Quantitative and qualitative differences observed between 4R and 3R tau are consistent with a "microtubule misregulation" model in which abnormal tau isoform expression results in the inability to properly regulate microtubule dynamics, leading to neuronal death and dementia.

Regulatory and developmental systems produce phenotypes that are robust to environmental and genetic variation. A gene product that normally contributes to this robustness is termed a phenotypic capacitor. When a phenotypic capacitor fails, for example when challenged by a harsh environment or mutation, the system becomes less robust and thus produces greater phenotypic variation. A functional phenotypic capacitor provides a mechanism by which hidden polymorphism can accumulate, whereas its failure provides a mechanism by which evolutionary change might be promoted. The primary example to date of a phenotypic capacitor is Hsp90, a molecular chaperone that targets a large set of signal transduction proteins. In both Drosophila and Arabidopsis, compromised Hsp90 function results in pleiotropic phenotypic effects dependent on the underlying genotype. For some traits, Hsp90 also appears to buffer stochastic variation, yet the relationship between environmental and genetic buffering remains an important unresolved question. We previously used simulations of knockout mutations in transcriptional networks to predict that many gene products would act as phenotypic capacitors. To test this prediction, we use high-throughput morphological phenotyping of individual yeast cells from single-gene deletion strains to identify gene products that buffer environmental variation in Saccharomyces cerevisiae. We find more than 300 gene products that, when absent, increase morphological variation. Overrepresented among these capacitors are gene products that control chromosome organization and DNA integrity, RNA elongation, protein modification, cell cycle, and response to stimuli such as stress. Capacitors have a high number of synthetic-lethal interactions but knockouts of these genes do not tend to cause severe decreases in growth rate. Each capacitor can be classified based on whether or not it is encoded by a gene with a paralog in the genome. Capacitors with a duplicate are highly connected in the protein–protein interaction network and show considerable divergence in expression from their paralogs. In contrast, capacitors encoded by singleton genes are part of highly interconnected protein clusters whose other members also tend to affect phenotypic variability or fitness. These results suggest that buffering and release of variation is a widespread phenomenon that is caused by incomplete functional redundancy at multiple levels in the genetic architecture., Author Summary Most species maintain abundant genetic variation and experience a wide range of environmental conditions, yet phenotypic differences between individuals are usually small. This phenomenon, known as phenotypic robustness, presents an apparent contradiction: if biological systems are so resistant to variation, how do they diverge and adapt through evolutionary time? Here, we address this question by investigating the molecular mechanisms that underlie phenotypic robustness and how these mechanisms can be broken to produce phenotypic heterogeneity. We identify genes that contribute to phenotypic robustness in yeast by analyzing the variance of morphological phenotypes in a comprehensive collection of single-gene knockout strains. We find that ∼5% of yeast genes break phenotypic robustness when knocked out. The products of these genes tend to be involved in critical cellular processes, including maintaining DNA stability, processing RNA, modifying proteins, and responding to stressful environments. These genes tend to interact genetically with a large number of other genes, and their products tend to interact physically with a large number of other gene products. Our results suggest that loss of phenotypic robustness might be a common phenomenon during evolution that occurs when cellular networks are disrupted., A genome-wide screen inSaccharomyces cerevisiae identifies over 300 gene products that buffer environmental variation--dubbed phenotypic capacitors--and function as hubs in protein-protein and synthetic-lethal interaction networks.