Search

Viruses frequently carry out replication in specialized compartments within cells. The effect of these structures on virus replication is poorly understood. Recent research supports phase separation as a foundational principle for organization of cellular components with the potential to influence viral replication. In this review, phase separation is described in the context of formation of viral replication centers, with an emphasis on the nonsegmented negative-strand RNA viruses. Consideration is given to the interplay between phase separation and the critical processes of viral transcription and genome replication, and the role of these regions in pathogen-host interactions is discussed. Finally, critical questions that must be addressed to fully understand how phase separation influences viral replication and the viral life cycle are presented, along with information about new approaches that could be used to make important breakthroughs in this emerging field.

Intrinsically disordered regions (IDRs) are ubiquitous across all domains of life and play a range of functional roles. While folded domains are generally well-described by a single 3D structure, IDRs exist in a collection of interconverting states known as an ensemble. This structural heterogeneity means IDRs are largely absent from the PDB, contributing to a lack of computational approaches to predict ensemble conformational properties from sequence. Here we combine rational sequence design, large-scale molecular simulations, and deep learning to develop ALBATROSS, a deep learning model for predicting IDR ensemble dimensions from sequence. ALBATROSS enables the instantaneous prediction of ensemble average properties at proteome-wide scale. ALBATROSS is lightweight, easy-to-use, and accessible as both a locally installable software package and a point-and-click style interface in the cloud. We first demonstrate the applicability of our predictors by examining the generalizability of sequence-ensemble relationships in IDRs. Then, we leverage the high-throughput nature of ALBATROSS to characterize emergent biophysical behavior of IDRs within and between proteomes.

Biomolecular condensates form via phase transitions of condensate-specific biomacromolecules. Intrinsically disordered regions (IDRs) featuring the appropriate sequence grammar can contribute homotypic and heterotypic interactions to the driving forces for phase separation of multivalent proteins. At this juncture, experiments and computations have matured to the point where the concentrations of coexisting dense and dilute phases can be quantified for individual IDRs in complex milieus bothin vitroandin vivo. For a macromolecule such as a disordered protein in a solvent, the locus of points that connects concentrations of the two coexisting phases defines a phase boundary or binodal. Often, only a few points along the binodal, especially in the dense phase, are accessible for measurement. In such cases and for quantitative and comparative analysis of parameters that describe the driving forces for phase separation, it is useful to fit measured or computed binodals to well-known mean-field free energies for polymer solutions. Unfortunately, the non-linearity of the underlying free energy functions makes it challenging to put mean-field theories into practice. Here, we present FIREBALL, a suite of computational tools designed to enable efficient construction, analysis, and fitting to experimental or computed data of binodals. We show that depending on the theory being used, one can also extract information regarding coil-to-globule transitions of individual macromolecules. Here, we emphasize the ease-of-use and utility of FIREBALL using examples based on data for two different IDRs.Statement of SignificanceMacromolecular phase separation drives the assembly of membraneless bodies known as biomolecular condensates. Measurements and computer simulations can now be brought to bear to quantify how the concentrations of macromolecules in coexisting dilute and dense phases vary with changes to solution conditions. These mappings can be fit to analytical expressions for free energies of solution to extract information regarding parameters that enable comparative assessments of the balance of macromolecule-solvent interactions across different systems. However, the underlying free energies are non-linear and fitting them to actual data is non-trivial. To enable comparative numerical analyses, we introduce FIREBALL, a user-friendly suite of computational tools that allows one to generate, analyze, and fit phase diagrams and coil-to-globule transitions using well-known theories.

Positively charged repeat peptides are emerging as key players in neurodegenerative diseases. These peptides can perturb diverse cellular pathways but a unifying framework for how such promiscuous toxicity arises has remained elusive. We used mass-spectrometry-based proteomics to define the protein targets of these neurotoxic peptides and found that they all share similar sequence features that drive their aberrant condensation with these positively charged peptides. We trained a machine learning algorithm to detect such sequence features and unexpectedly discovered that this mode of toxicity is not limited to human repeat expansion disorders but has evolved countless times across the tree of life in the form of cationic antimicrobial and venom peptides. We demonstrate that an excess in positive charge is necessary and sufficient for this killer activity, which we name ‘polycation poisoning’. These findings reveal an ancient and conserved mechanism and inform ways to leverage its design rules for new generations of bioactive peptides.

The SARS-CoV-2 Nucleocapsid (N) protein is responsible for condensation of the viral genome. Characterizing the mechanisms controlling nucleic acid binding is a key step in understanding how condensation is realized. Here, we focus on the role of the RNA Binding Domain (RBD) and its flanking disordered N-Terminal Domain (NTD) tail, using single-molecule Förster Resonance Energy Transfer and coarse-grained simulations. We quantified contact site size and binding affinity for nucleic acids and concomitant conformational changes occurring in the disordered region. We found that the disordered NTD increases the affinity of the RBD for RNA by about 50-fold. Binding of both nonspecific and specific RNA results in a modulation of the tail configurations, which respond in an RNA length-dependent manner. Not only does the disordered NTD increase affinity for RNA, but mutations that occur in the Omicron variant modulate the interactions, indicating a functional role of the disordered tail. Finally, we found that the NTD-RBD preferentially interacts with single-stranded RNA and that the resulting protein:RNA complexes are flexible and dynamic. We speculate that this mechanism of interaction enables the Nucleocapsid protein to search the viral genome for and bind to high-affinity motifs.

In immature oocytes, Balbiani bodies are conserved membraneless condensates implicated in oocyte polarization, the organization of mitochondria, and long-term organelle and RNA storage. In Xenopus laevis, Balbiani body assembly is mediated by the protein Velo1. Velo1 contains an N-terminal prion-like domain (PLD) that is essential for Balbiani body formation. PLDs have emerged as a class of intrinsically disordered regions that can undergo various different types of intracellular phase transitions and are often associated with dynamic, liquid-like condensates. Intriguingly, the Velo1 PLD forms solid-like assemblies. Here we sought to understand why Velo1 phase behavior appears to be biophysically distinct from that of other PLD-containing proteins. Through bioinformatic analysis and coarse-grained simulations, we predict that the clustering of aromatic residues and the amino acid composition of residues between aromatics can influence condensate material properties, organization, and the driving forces for assembly. To test our predictions, we redesigned the Velo1 PLD to test the impact of targeted sequence changes in vivo. We found that the Velo1 design with evenly spaced aromatic residues shows rapid internal dynamics, as probed by fluorescent recovery after photobleaching, even when recruited into Balbiani bodies. Our results suggest that Velo1 might have been selected in evolution for distinctly clustered aromatic residues to maintain the structure of Balbiani bodies in long-lived oocytes. In general, our work identifies several tunable parameters that can be used to augment the condensate material state, offering a road map for the design of synthetic condensates.

SummaryThe pre-mRNA life cycle requires intron processing; yet, how intron processing defects influence splicing and gene expression is unclear. Here, we find TTDN1, which is frequently mutated in non-photosensitive trichothiodystrophy (NP-TTD), functionally links intron lariat processing to the spliceosome. The conserved TTDN1 C-terminal region directly binds lariat debranching enzyme DBR1, while its N-terminal intrinsically disordered region (IDR) binds the intron binding complex (IBC). The IDR forms condensatesin vitroand is needed for IBC interaction. TTDN1 loss causes significant intron lariat accumulation, as well as splicing and gene expression defects, mirroring phenotypes observed in NP-TTD patient cells.Ttdn1Δ/Δmice recapitulate intron processing defects and neurodevelopmental phenotypes seen in NP-TTD. A DBR1-IDR fusion recruits DBR1 to the IBC and circumvents the requirement for TTDN1, indicating this tethering role as its major molecular function. Collectively, our findings unveil key functional connections between lariat processing, splicing outcomes, and NP-TTD molecular pathology.

Cellular organization is determined by a combination of membrane-bound and membrane-less biomolecular assemblies that range from clusters of tens of molecules to micrometer-sized cellular bodies. Over the last decade, membrane-less assemblies have come to be referred to as biomolecular condensates, reflecting their ability to condense specific molecules with respect to the remainder of the cell. In many cases, the physics of phase transitions provides a conceptual framework and a mathematical toolkit to describe the assembly, maintenance, and dissolution of biomolecular condensates. Among the various quantitative and qualitative models applied to understand intracellular phase transitions, the stickers-and-spacers framework offers an intuitive yet rigorous means to map biomolecular sequences and structure to the driving forces needed for higher-order assembly. This chapter introduces the fundamental concepts behind the stickers-and-spacers model, considers its application to different biological systems, and discusses limitations and misconceptions around the model.

A surge in research focused on understanding the physical principles governing the formation, properties, and function of membraneless compartments has occurred over the past decade. Compartments such as the nucleolus, stress granules, and nuclear speckles have been designated as biomolecular condensates to describe their shared property of spatially concentrating biomolecules. Although this research has historically been carried out in animal and fungal systems, recent work has begun to explore whether these same principles are relevant in plants. Effectively understanding and studying biomolecular condensates require interdisciplinary expertise that spans cell biology, biochemistry, and condensed matter physics and biophysics. As such, some involved concepts may be unfamiliar to any given individual. This review focuses on introducing concepts essential to the study of biomolecular condensates and phase separation for biologists seeking to carry out research in this area and further examines aspects of biomolecular condensates that are relevant to plant systems.

Cell-to-cell variability is shaped by transcription dynamics because genes are transcribed in bursts interspersed with inactive periods. The stochasticity of bursting means that genes transcribed in rare bursts exhibit more heterogeneity at the single cell level than genes that burst often 1, 2. Transcription starts with the binding of Transcription Factors (TFs) to specific sequence motifs where they recruit the transcription machinery 3. In some systems, individual TF binding events temporally correlate with the firing of transcriptional bursts, defining the target gene’s frequency and duration 4–6. However, in the absence of methods that assess the impact of different TFs on transcription dynamics at the same genetic loci, it remains unclear whether DNA binding kinetics are the sole determinant of bursting. Here we develop an imaging-based synthetic recruitment assay, CRISPRburst, and measure how 92 human TFs impact bursting kinetics. We show that TFs recruited to chromatin under identical conditions generate diverse bursting signatures, some TFs increasing the probability of the gene turning on while others increase the number of mRNA molecules transcribed per burst. We find that the association of TFs with specific protein partners determines their bursting output, and train a model to predict the kinetic signatures of all human TFs. These kinetic signatures can be used as a TF classification system complementary to existing families based on DNA binding domains. Additionally, kinetic signatures provide a rational framework to design synthetic activators, model transcription regulation, and understand expression heterogeneity.

Intrinsically disordered proteins and protein regions (IDRs) make up around 30% of the human proteome where they play essential roles in dictating and regulating many core biological processes. While IDRs are often studied as isolated domains, in naturally occurring proteins most IDRs are found adjacent to folded domains, where they exist as either N- or C-terminal tails or as linkers connecting two folded domains. Prior work has shown that charge properties of IDRs can influence their conformational behavior, both in isolation and in the context of folded domains. In contrast, the converse scenario is less well-explored: how do the charge properties of folded domains influence IDR conformational behavior? To answer this question, we combined a large-scale structural bioinformatics analysis with all-atom implicit solvent simulations of both rationally designed and naturally occurring proteins. Our results reveal three key takeaways. Firstly, the relative position and accessibility of charged residues across the surface of a folded domain can dictate IDR conformational behavior, overriding expectations based on net surface charge properties. Secondly, naturally occurring proteins possess multiple charge patches that are physically accessible to local IDRs. Finally, even modest changes in the local electrostatic environment of a folded domain can substantially modulate IDR-folded domain interactions. Taken together, our results suggest that folded domain surfaces can act as local determinants of IDR conformational behavior., Graphical abstract Image 1, Highlights • Intrinsically disordered regions (IDRs) are mostly found adjacent to folded domains. • Here we propose that the folded domain surface properties influence IDR behavior. • We combine all-atom simulations and sequence design of IDRs and folded domains. • IDR conformational behavior is determined by a complex combination of factors. • Folded domains can substantially alter IDR conformational biases.

SummaryMost non-membrane-bound compartments in cells that form via phase separation have complex composition. Phase separation of individual proteins that form these compartments has been well documented. However, we know relatively little about how the dynamics of individual proteins are modulated in multicomponent systems. Here, we used in vitro reconstitution and in vivo experiments to investigate how complex composition within a liquid-like ‘P granule’ compartment affects the dynamics of one of its scaffold proteins PGL-3. Using mutational analysis and biophysical perturbations, we generated PGL-3 constructs with reduced alpha-helicity that phase separate in vitro into condensates with widely varying dynamics. We use these PGL-3 constructs to show that introducing other P granule components buffers against large change of dynamics within liquid-like compartments. This dynamics-buffering effect is mediated by weak interactions among two or more components. Such dynamics-buffering may contribute to robust functional output of cellular liquid-like compartments.

The plant cell internal environment is a dynamic, intricate landscape composed of many intracellular compartments. Cells organize some cellular components through formation of biomolecular condensates-non-stoichiometric assemblies of protein and/or nucleic acids. In many cases, phase separation appears to either underly or contribute to the formation of biomolecular condensates. Many canonical membraneless compartments within animal cells form in a manner that is at least consistent with phase separation, including nucleoli, stress granules, Cajal bodies, and numerous additional bodies, regulated by developmental and environmental stimuli. In this Review, we examine the emerging roles for phase separation in plants. Further, drawing on studies carried out in other organisms, we identify cellular phenomenon in plants that might also arise via phase separation. We propose that plants make use of phase separation to a much greater extent than has been previously appreciated, implicating phase separation as an evolutionarily ancient mechanism for cellular organization.

Not too sticky There is increasing evidence for a role of liquid-liquid phase separation (LLPS) in many cellular processes. Many proteins that undergo LLPS include prionlike domains (PLDs), which are enriched in polar amino acids and often interspersed with aromatic residues. Combining experimental data with simulations, Martin et al. quantified concentrations of PLDs in coexisting dilute and dense phases as a function of temperature and show that the phase behavior is determined by the number of aromatic residues and their patterning, with uniform patterning of aromatic residues promoting LLPS and inhibiting aggregation. They developed a sticker-and-spacers model that can predict the phase behavior of PLDs on the basis of their sequence. Science , this issue p. 694

Intrinsically disordered protein regions (IDRs) pervasively engage in essential molecular functions, yet they are often poorly conserved as assessed by sequence alignment. To understand the seeming paradox of how sequence variability is compatible with function, we examined the functional determinants for a poorly conserved but essential IDR. We show that IDR function depends on two distinct but related properties: sequence- and chemical specificity. While sequence-specificity works via linear binding motifs, chemical-specificity reflects the sequence-encoded chemistry of multivalent interactions through amino acids across an IDR. Unexpectedly, an apparently essential binding motif can be removed if compensatory changes to the sequence chemistry are made, highlighting the orthogonality and interoperability of both properties. Our results provide a general framework to understand the functional constraints on IDR sequence evolution.One-Sentence SummaryInteractions driven by intrinsically disordered regions can be understood using a two-dimensional landscape that defines binding via motif-dependent and motif-independent contributions.

SummaryThe circadian clock times cellular processes to the day/night cycle via a Transcription-Translation negative Feedback Loop (TTFL). However, a mechanistic understanding of the negative arm in both the timing of the TTFL and its control of output is lacking. We posited that the formation of negative-arm protein complexes was fundamental to clock regulation stemming from the negative arm. Using a modified peptide microarray approach termed Linear motif discovery using rational design (LOCATE), we characterized the interaction of the disordered negative-arm clock protein FREQUENCY to its partner protein FREQUENCY-Interacting RNA helicase. LOCATE identified a specific Short Linear Motif (SLiM) and interaction “hotspot” as well as positively charged “islands” that mediate electrostatic interactions, suggesting a model where negative arm proteins form a “fuzzy” complex essential for clock timing and robustness. Further analysis revealed that the positively charged islands were an evolutionarily conserved feature in higher eukaryotes and contributed to proper clock function.

Many disordered proteins conserve essential functions in the face of extensive sequence variation, making it challenging to identify the mechanisms responsible for functional selection. Here we identify the molecular mechanism of functional selection for the disordered adenovirus early gene 1A (E1A) protein. E1A competes with host factors to bind the retinoblastoma (Rb) protein, subverting cell cycle regulation. We show that two binding motifs tethered by a hypervariable disordered linker drive picomolar affinity Rb binding and host factor displacement. Compensatory changes in amino acid sequence composition and sequence length lead to conservation of optimal tethering across a large family of E1A linkers. We refer to this compensatory mechanism as conformational buffering. We also detect coevolution of the motifs and linker, which can preserve or eliminate the tethering mechanism. Conformational buffering and motif-linker coevolution explain robust functional encoding within hypervariable disordered linkers and could underlie functional selection of many disordered protein regions., This work was supported by: Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT) Grants PICT no. 2013-1895 and no. 2017-1924 (L. B. C.), no. 2012-2550 and no. 2015-1213 (I. E. S.) and no. 2016-4605 (G. P. G.); US National Institutes of Health no. GM115556 and no. CA141244 (G. W. D.) and no. 5R01NS056114 (R. V. P.); Florida Department of Health (FLDOH) no. 20B17 (G. W. D.); US National Science Foundation no. MCB-1614766 (R. V. P.); a travel award from the USF Nexus Initiative and a Creative Scholarship Grant from the USF College of Arts and Sciences (G. W. D. and L. B. C.); Labex EpiGenMed (Investissements d’avenir) program no. ANR-10-LABX-12-01 (P. B.); French National Research Agency no. ANR-10-INBS-04-01 and no. ANR-10-INBS-05 (P. B.); Spanish Ministerio de Ciencia y Universidades MICYU-FEDER no. RTI2018-097189-C2-1 (G. F.-B.); Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET, Argentina) doctoral fellowship (N. S. G.-F., M. S. and N. A. G.), postdoctoral fellowship (J. G.), and permanent researcher (L. B. C., G. d. P.-G. and I. E. S.); Fulbright Visiting Scholar Program (N. S. G.-F.); Ministerio de Ciencia e Innovación, España, no. BES-2013-063991 and no. EEBB-I-16-11670 (S. B.-V.); Longer Life Foundation: A RGA/Washington University Collaboration (A. S. H.); HPC resources of the CALMIP supercomputing center no. 2016-P16032 (G. F.-B.); and Cluster of Scientific Computing (http://ccc.umh.es/) of the Miguel Hernández University (G. F.-B.). The synchrotron SAXS data were collected at beamline P12, operated by EMBL Hamburg at the PETRA III storage ring (DESY, Hamburg, Germany). We thank K. Perez at the Protein Expression and Purification Core Facility at EMBL (Heidelberg) for critical help with ITC experiments, and P. Aramendia for providing critical access to fluorescence spectrometry equipment at Centro de Investigaciones en Bionanociencias (CIBION, Argentina).

The rise of high-throughput experiments has transformed how scientists approach biological questions. The ubiquity of large-scale assays that can test thousands of samples in a day has necessitated the development of new computational approaches to interpret this data. Among these tools, machine learning approaches are increasingly being utilized due to their ability to infer complex nonlinear patterns from high-dimensional data. Despite their effectiveness, machine learning (and in particular deep learning) approaches are not always accessible or easy to implement for those with limited computational expertise. Here we present PARROT, a general framework for training and applying deep learning-based predictors on large protein datasets. Using an internal recurrent neural network architecture, PARROT is capable of tackling both classification and regression tasks while only requiring raw protein sequences as input. We showcase the potential uses of PARROT on three diverse machine learning tasks: predicting phosphorylation sites, predicting transcriptional activation function of peptides generated by high-throughput reporter assays, and predicting the fibrillization propensity of amyloid beta with data generated by deep mutational scanning. Through these examples, we demonstrate that PARROT is easy to use, performs comparably to state-of-the-art computational tools, and is applicable for a wide array of biological problems.

Tardigrades, also known as water bears, make up a phylum of small but extremely hardy animals, renowned for their ability to survive extreme stresses, including desiccation. How tardigrades survive desiccation is one of the enduring mysteries of animal physiology. Here we show that CAHS D, an intrinsically disordered protein belonging to a unique family of proteins possessed only by tardigrades, undergoes a liquid-to-gel phase transition in a concentration dependent manner. Unlike other gelling proteins, such as gelatin, our data support a mechanism in which gel formation of CAHS D is driven by intermolecular β-β interactions. We find that gel formation corresponds with strong coordination of water and slowing of water diffusion. The degree of water coordination correlates with the ability of CAHS D to protect lactate dehydrogenase from unfolding when dried. This implies that the mechanism for unfolding protection can be attributed to a combination of hydration and slowed molecular motion. Conversely, rapid diffusion leading to efficient molecular shielding appears to be the predominant mechanism preventing protein aggregation. Our study demonstrates that distinct mechanisms are required for holistic protection during desiccation, and that protectants, such as CAHS D, can act as molecular ‘Swiss Army Knives’ capable of providing protection through several different mechanisms simultaneously.

Xu et al. (2021) uncover a conserved viral mechanism to subvert the innate immune system whereby herpesvirus tegument proteins suppress cGAS:DNA phase separation.

While many proteins and protein regions utilize a complex repertoire of amino acids to achieve their biological function, a subset of protein sequences are enriched in a reduced set of amino acids. These so-called low complexity (LC) sequences, specifically intrinsically disordered variants of LC sequences, have been the focus of recent investigations owing to their roles in a range of biological functions, specifically phase separation. Computational studies of LC sequences have provided rich insights into their behavior both as individual proteins in dilute solutions and as the drivers and modulators of higher-order assemblies. Here, we review how simulations performed across distinct resolutions have provided different types of insights into the biological role of LC sequences.

Telomeres are nucleoprotein complexes that protect the ends of chromosomes and are essential for chromosome stability in Eukaryotes. In cells, individual telomeres form distinct globules of finite size that appear to be smaller than expected for bare DNA. Moreover, upon changes in their protein composition, telomeres can cluster to form telomere-induced-foci (TIFs) or co-localize with promyelocytic leukemia (PML) nuclear bodies. The physical basis for collapse of individual telomeres and coalescence of multiple ones remains unclear, as does the relationship between these two phenomena. By combining single-molecule measurements, optical microscopy, turbidity assays, and simulations, we show that the telomere scaffolding protein TRF2 can condense individual DNA chains and drives coalescence of multiple DNA molecules, leading to phase separation and the formation of liquid-like droplets. Addition of the TRF2 binding protein hRap1 modulates phase boundaries and tunes the specificity of solution demixing while simultaneously altering the degree of DNA compaction. Our results suggest that the condensation of single telomeres and formation of biomolecular condensates containing multiple telomeres are two different outcomes driven by the same set of molecular interactions. Moreover, binding partners, such as other telomere components, can alter those interactions to promote single-chain DNA compaction over multiple-chain phase separation.

Intrinsically disordered proteins and protein regions make up a substantial fraction of many proteomes where they play a wide variety of essential roles. A critical first step in understanding the role of disordered protein regions in biological function is to identify those disordered regions correctly. Computational methods for disorder prediction have emerged as a core set of tools to guide experiments, interpret results, and develop hypotheses. Given the multiple different predictors available, consensus scores have emerged as a popular approach to mitigate biases or limitations of any single method. Consensus scores integrate the outcome of multiple independent disorder predictors and provide a per-residue value that reflects the number of tools that predict a residue to be disordered. Although consensus scores help mitigate the inherent problems of using any single disorder predictor, they are computationally expensive to generate. They also necessitate the installation of multiple different software tools, which can be prohibitively difficult. To address this challenge, we developed a deep-learning-based predictor of consensus disorder scores. Our predictor, metapredict, utilizes a bidirectional recurrent neural network trained on the consensus disorder scores from 12 proteomes. By benchmarking metapredict using two orthogonal approaches, we found that metapredict is among the most accurate disorder predictors currently available. Metapredict is also remarkably fast, enabling proteome-scale disorder prediction in minutes. Importantly, metapredict is fully open source and is distributed as a Python package, a collection of command-line tools, and a web server, maximizing the potential practical utility of the predictor. We believe metapredict offers a convenient, accessible, accurate, and high-performance predictor for single-proteins and proteomes alike.Statement of SignificanceIntrinsically disordered regions are found across all kingdoms of life where they play a variety of essential roles. Being able to accurately and quickly identify disordered regions in proteins using just the amino acid sequence is critical for the appropriate design and interpretation of experiments. Despite this, performing large-scale disorder prediction on thousands of sequences is challenging using extant disorder predictors due to various difficulties including general installation and computational requirements. We have developed an accurate, high-performance and easy-to-use predictor of protein disorder and structure. Our predictor, metapredict, was designed for both proteome-scale analysis and individual sequence predictions alike. Metapredict is implemented as a collection of local tools and an online web server, and is appropriate for both seasoned computational biologists and novices alike.

Many aspects of photoperception by plants and microorganisms are initiated by the phytochrome (Phy) family of photoreceptors that detect light through interconversion between red light- (Pr) and far-red light-absorbing (Pfr) states. Plants synthesize a small family of Phy isoforms (PhyA to PhyE) that collectively regulate photomorphogenesis and temperature perception through redundant and unique actions. While the selective roles of these isoforms have been partially attributed to their differing abundances, expression patterns, affinities for downstream partners, and turnover rates, we show here from analysis of recombinant Arabidopsis chromoproteins that the Phy isoforms also display distinct biophysical properties. Included are a hypsochromic shift in the Pr absorption for PhyC and varying rates of Pfr to Pr thermal reversion, part of which can be attributed to the core photosensory module in each. Most strikingly, PhyB combines strong temperature dependence of thermal reversion with an order-of-magnitude faster rate to likely serve as the main physiological thermosensor, whereby thermal reversion competes with photoconversion. In addition, comparisons of Pfr occupancies for PhyA and PhyB under a range of red- and white-light fluence rates imply that low-light environments are effectively sensed by PhyA, while high-light environments, such as full sun, are effectively sensed by PhyB. Parallel analyses of the Phy isoforms from potato and maize showed that the unique features within the Arabidopsis family are conserved, thus indicating that the distinct biophysical properties among plant Phy isoforms emerged early in Phy evolution, likely to enable full interrogation of their light and temperature environments.

The rise of high-throughput experiments has transformed how scientists approach biological questions. The ubiquity of large-scale assays that can test thousands of samples in a day has necessitated the development of new computational approaches to interpret this data. Among these tools, machine learning approaches are increasingly being utilized due to their ability to infer complex non-linear patterns from high-dimensional data. Despite their effectiveness, machine learning (and in particular deep learning) approaches are not always accessible or easy to implement for those with limited computational expertise. Here we present PARROT, a general framework for training and applying deep learning-based predictors on large protein datasets. Using an internal recurrent neural network architecture, PARROT is capable of tackling both classification and regression tasks while only requiring raw protein sequences as input. We showcase the potential uses of PARROT on three diverse machine learning tasks: predicting phosphorylation sites, predicting transcriptional activation function of peptides generated by high-throughput reporter assays, and predicting the fibrillization propensity of amyloid beta with data generated by deep mutational scanning. Through these examples, we demonstrate that PARROT is easy to use, performs comparably to state-of-the-art computational tools, and is applicable for a wide array of biological problems.

It is increasingly appreciated that intracellular pH changes are important biological signals. This motivates the elucidation of molecular mechanisms of pH-sensing. We determined that a nucleocytoplasmic pH oscillation was required for the transcriptional response to carbon starvation inS. cerevisiae. The SWI/SNF chromatin remodeling complex is a key mediator of this transcriptional response. We found that a glutamine-rich low complexity sequence (QLC) in theSNF5subunit of this complex, and histidines within this sequence, were required for efficient transcriptional reprogramming during carbon starvation. Furthermore, theSNF5QLC mediated pH-dependent recruitment of SWI/SNF to a model promoterin vitro.Simulations showed that protonation of histidines within theSNF5QLC lead to conformational expansion, providing a potential biophysical mechanism for regulation of these interactions. Together, our results indicate that that pH changes are a second messenger for transcriptional reprogramming during carbon starvation, and that theSNF5QLC acts as a pH-sensor.

Many disordered proteins conserve essential functions in the face of extensive sequence variation. This makes it challenging to identify the forces responsible for functional selection. Viruses are robust model systems to investigate functional selection and they take advantage of protein disorder to acquire novel traits. Here, we combine structural and computational biophysics with evolutionary analysis to determine the molecular basis for functional selection in the intrinsically disordered adenovirus early gene 1A (E1A) protein. E1A competes with host factors to bind the retinoblastoma (Rb) protein, triggering early S-phase entry and disrupting normal cellular proliferation. We show that the ability to outcompete host factors depends on the picomolar binding affinity of E1A for Rb, which is driven by two binding motifs tethered by a hypervariable disordered linker. Binding affinity is determined by the spatial dimensions of the linker, which constrain the relative position of the two binding motifs. Despite substantial sequence variation across evolution, the linker dimensions are finely optimized through compensatory changes in amino acid sequence and sequence length, leading to conserved linker dimensions and maximal affinity. We refer to the mechanism that conserves spatial dimensions despite large-scale variations in sequence as conformational buffering. Conformational buffering explains how variable disordered proteins encode functions and could be a general mechanism for functional selection within disordered protein regions.

BackgroundBiomolecular condensates are non-stoichiometric assemblies that are characterized by their capacity to spatially concentrate biomolecules and play a key role in cellular organization. Proteins that drive the formation of biomolecular condensates frequently contain oligomerization domains and intrinsically disordered regions (IDRs), both of which can contribute multivalent interactions that drive higher-order assembly. Our understanding of the relative and temporal contribution of oligomerization domains and IDRs to the material properties of in vivo biomolecular condensates is limited. Similarly, the spatial and temporal dependence of protein oligomeric state inside condensates has been largely unexplored in vivo.MethodsIn this study, we combined quantitative microscopy with number and brightness analysis to investigate the aging, material properties, and protein oligomeric state of biomolecular condensates in vivo. Our work is focused on condensates formed by AUXIN RESPONSE FACTOR 19 (ARF19), which is a transcription factor integral to the signaling pathway for the plant hormone auxin. ARF19 contains a large central glutamine-rich IDR and a C-terminal Phox Bem1 (PB1) oligomerization domain and forms cytoplasmic condensates.ResultsOur results reveal that the IDR amino acid composition can influence the morphology and material properties of ARF19 condensates. In contrast the distribution of oligomeric species within condensates appears insensitive to the IDR composition. In addition, we identified a relationship between the abundance of higher- and lower-order oligomers within individual condensates and their apparent fluidity.ConclusionsIDR amino acid composition affects condensate morphology and material properties. In ARF condensates, altering the amino acid composition of the IDR did not greatly affect the oligomeric state of proteins within the condensate.

Summary Ribonucleoprotein condensates can exhibit diverse physical states in vitro and in vivo. Despite considerable progress, the relevance of condensate physical states for in vivo biological function remains limited. Here, we investigated the physical properties of processing bodies (P bodies) and their impact on mRNA storage in mature Drosophila oocytes. We show that the conserved DEAD-box RNA helicase Me31B forms viscous P body condensates, which adopt an arrested physical state. We demonstrate that structurally distinct proteins and protein-protein interactions, together with RNA, regulate the physical properties of P bodies. Using live imaging and in situ hybridization, we show that the arrested state and integrity of P bodies support the storage of bicoid (bcd) mRNA and that egg activation modulates P body properties, leading to the release of bcd for translation in the early embryo. Together, this work provides an example of how physical states of condensates regulate cellular function in development., Graphical abstract, Highlights • P bodies adopt an arrested physical state in the mature oocyte • Multivalent interactions and structurally distinct proteins regulate P body properties • The arrested state and integrity of P bodies support the storage of bicoid mRNA • Egg activation modulates P body properties and leads to release of bicoid mRNA, Sankaranarayanan et al. show that P bodies, conserved ribonucleoprotein condensates found in Drosophila oocytes, are regulated by specific structural features and weak multivalent interactions. In vivo, P bodies adopt an arrested physical state that is critical for the storage of the bicoid patterning mRNA until egg activation.

Biomolecular condensates that form via liquid-liquid phase separation can exhibit diverse physical states. Despite considerable progress, the relevance of condensate physical states forin vivobiological function remains limited. Here, we investigated the physical properties ofin vivoprocessing bodies (P bodies) and their impact on mRNA storage in matureDrosophilaoocytes. We show that the conserved DEAD-box RNA helicase Me31B forms P body condensates which adopt a less dynamic, arrested physical state. We demonstrate that structurally distinct proteins and hydrophobic and electrostatic interactions, together with RNA and intrinsically disordered regions, regulate the physical properties of P bodies. Finally, using live imaging, we show that the arrested state of P bodies is required to prevent the premature release ofbicoid(bcd) mRNA, a body axis determinant, and that P body dissolution leads tobcdrelease. Together, this work establishes a role for arrested states of biomolecular condensates in regulating cellular function in a developing organism.

Phase separation is emerging as a universal principle for how cells use dynamic subcompartmentalization to organize biochemical reactions in time and space1,2. Yet, whether the emergent physical properties of these biomolecular condensates are important for their biological function remains unclear. The intrinsically disordered protein PopZ forms membraneless condensates at the poles of the bacterium Caulobacter crescentus and selectively sequesters kinase-signaling cascades to regulate asymmetric cell division3–5. By dissecting the molecular grammar underlying PopZ phase separation, we find that unlike many eukaryotic examples, where unstructured regions drive condensation6,7, a structured domain of PopZ drives condensation, while conserved repulsive features of the disordered region modulate material properties. By generating rationally designed PopZ mutants, we find that the exact material properties of PopZ condensates directly determine cellular fitness, providing direct evidence for the physiological importance of the emergent properties of biomolecular condensates. Our work codifies a clear set of design principles illuminating how sequence variation in a disordered domain alters the function of a widely conserved bacterial condensate. We used these insights to repurpose PopZ as a modular platform for generating synthetic condensates of tunable function in human cells.