Nuclear-import receptors as gatekeepers of pathological phase transitions in ALS/FTD.

Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) are fatal neurodegenerative disorders on a disease spectrum that are characterized by the cytoplasmic mislocalization and aberrant phase transitions of prion-like RNA-binding proteins (RBPs). The common accumulation of TAR DNA-binding protein-43 (TDP-43), fused in sarcoma (FUS), and other nuclear RBPs in detergent-insoluble aggregates in the cytoplasm of degenerating neurons in ALS/FTD is connected to nuclear pore dysfunction and other defects in the nucleocytoplasmic transport machinery. Recent advances suggest that beyond their canonical role in the nuclear import of protein cargoes, nuclear-import receptors (NIRs) can prevent and reverse aberrant phase transitions of TDP-43, FUS, and related prion-like RBPs and restore their nuclear localization and function. Here, we showcase the NIR family and how they recognize cargo, drive nuclear import, and chaperone prion-like RBPs linked to ALS/FTD. We also discuss the promise of enhancing NIR levels and developing potentiated NIR variants as therapeutic strategies for ALS/FTD and related neurodegenerative proteinopathies.

The interface of condensates of the hnRNPA1 low-complexity domain promotes formation of amyloid fibrils.

The maturation of liquid-like protein condensates into amyloid fibrils has been associated with several neurodegenerative diseases. However, the molecular mechanisms underlying this liquid-to-solid transition have remained largely unclear. Here we analyse the amyloid formation mediated by condensation of the low-complexity domain of hnRNPA1, a protein involved in amyotrophic lateral sclerosis. We show that phase separation and fibrillization are connected but distinct processes that are modulated by different regions of the protein sequence. By monitoring the spatial and temporal evolution of amyloid formation we demonstrate that the formation of fibrils does not occur homogeneously inside the droplets but is promoted at the interface of the condensates. We further show that coating the interface of the droplets with surfactant molecules inhibits fibril formation. Our results reveal that the interface of biomolecular condensates of hnRNPA1 promotes fibril formation, therefore suggesting interfaces as a potential novel therapeutic target against the formation of aberrant amyloids mediated by condensation.

Dynamic arrest and aging of biomolecular condensates are modulated by low-complexity domains, RNA and biochemical activity.

Biomolecular condensates require suitable control of material properties for their function. Here we apply Differential Dynamic Microscopy (DDM) to probe the material properties of an in vitro model of processing bodies consisting of out-of-equilibrium condensates formed by the DEAD-box ATPase Dhh1 in the presence of ATP and RNA. By applying this single-droplet technique we show that condensates within the same population exhibit a distribution of material properties, which are regulated on several levels. Removal of the low-complexity domains (LCDs) of the protein decreases the fluidity of the condensates. Structured RNA leads to a larger fraction of dynamically arrested condensates with respect to unstructured polyuridylic acid (polyU). Promotion of the enzymatic ATPase activity of Dhh1 reduces aging of the condensates and the formation of arrested structures, indicating that biochemical activity and material turnover can maintain fluid-like properties over time.

Programmable Zwitterionic Droplets as Biomolecular Sorters and Model of Membraneless Organelles.

Increasing evidence indicates that cells can regulate biochemical functions in time and space by generating membraneless compartments with well-defined mesoscopic properties. One important mechanism underlying this control is simple coacervation driven by associative disordered proteins that encode multivalent interactions. Inspired by these observations, programmable droplets based on simple coacervation of responsive synthetic polymers that mimic the "stickers-and-spacers" architecture of biological disordered proteins are developed. Zwitterionic polymers that undergo an enthalpy-driven liquid-liquid phase separation process and form liquid droplets that remarkably exclude most molecules are developed. Starting from this reference material, different functional groups in the zwitterionic polymer are progressively added to encode an increasing number of different intermolecular interactions. This strategy allowed the multiple emerging properties of the droplets to be controlled independently, such as stimulus-responsiveness, polarity, selective uptake of client molecules, fusion times, and miscibility. By exploiting this high programmability, a model of cellular compartmentalization is reproduced and droplets capable of confining different molecules in space without physical barriers are generated. Moreover, these biomolecular sorters are demonstrated to be able to localize, separate, and enable the detection of target molecules even within complex mixtures, opening attractive applications in bioseparation, and diagnostics.

Sequestration within biomolecular condensates inhibits Aß-42 amyloid formation.

Biomolecular condensates are emerging as an efficient strategy developed by cells to control biochemical reactions in space and time by locally modifying composition and environment. Yet, local increase in protein concentration within these compartments could promote aberrant aggregation events, including the nucleation and growth of amyloid fibrils. Understanding protein stability within the crowded and heterogeneous environment of biological condensates is therefore crucial, not only when the aggregation-prone protein is the scaffold element of the condensates but also when proteins are recruited as client molecules within the compartments. Here, we investigate the partitioning and aggregation kinetics of the amyloidogenic peptide Abeta42 (Aß-42), the peptide strongly associated with Alzheimer's disease, recruited into condensates based on low complexity domains (LCDs) derived from the DEAD-box proteins Laf1, Dbp1 and Ddx4, which are associated with biological membraneless organelles. We show that interactions between Aß-42 and the scaffold proteins promote sequestration and local increase of the peptide concentration within the condensates. Yet, heterotypic interactions within the condensates inhibit the formation of amyloid fibrils. These results demonstrate that biomolecular condensates could sequester aggregation-prone proteins and prevent aberrant aggregation events, despite the local increase in their concentration. Biomolecular condensates could therefore work not only as hot-spots of protein aggregation but also as protective reservoirs, since the heterogenous composition of the condensates could prevent the formation of ordered fibrillar aggregates.

Analysis of biomolecular condensates and protein phase separation with microfluidic technology.

An increasing body of evidence shows that membraneless organelles are key components in cellular organization. These observations open a variety of outstanding questions about the physico-chemical rules underlying their assembly, disassembly and functions. Some molecular determinants of biomolecular condensates are challenging to probe and understand in complex in vivo systems. Minimalistic in vitro reconstitution approaches can fill this gap, mimicking key biological features, while maintaining sufficient simplicity to enable the analysis of fundamental aspects of biomolecular condensates. In this context, microfluidic technologies are highly attractive tools for the analysis of biomolecular phase transitions. In addition to enabling high-throughput measurements on small sample volumes, microfluidic tools provide for exquisite control of self-assembly in both time and space, leading to accurate quantitative analysis of biomolecular phase transitions. Here, with a specific focus on droplet-based microfluidics, we describe the advantages of microfluidic technology for the analysis of several aspects of phase separation. These include phase diagrams, dynamics of assembly and disassembly, rheological and surface properties, exchange of materials with the surrounding environment and the coupling between compartmentalization and biochemical reactions. We illustrate these concepts with selected examples, ranging from simple solutions of individual proteins to more complex mixtures of proteins and RNA, which represent synthetic models of biological membraneless organelles. Finally, we discuss how this technology may impact the bottom-up fabrication of synthetic artificial cells and for the development of synthetic protein materials in biotechnology.

Examination of SOD1 aggregation modulators and their effect on SOD1 enzymatic activity as a proxy for potential toxicity.

Small-molecule inhibitors of abnormal protein self-assembly are promising candidates for developing therapy against proteinopathies. Such compounds have been examined primarily as inhibitors of amyloid ß-protein (Aß), whereas testing of inhibitors of other amyloidogenic proteins has lagged behind. An important issue with screening compound libraries is that although an inhibitor suitable for therapy must be both effective and nontoxic, typical screening focuses on efficacy, whereas safety typically is tested at a later stage using cells and/or animals. In addition, typical thioflavin T (ThT)-fluorescence-based screens use the final fluorescence value as a readout, potentially missing important kinetic information. Here, we examined potential inhibitors of superoxide dismutase 1 (SOD1) using ThT-fluorescence including the different phases of fluorescence change and added a parallel screen of SOD1 activity as a potential proxy for compound toxicity. Some compounds previously reported to inhibit other amyloidogenic proteins also inhibited SOD1 aggregation at low micromolar concentrations, whereas others were ineffective. Analysis of the lag phase and exponential slope added important information that could help exclude false-positive or false-negative results. SOD1 was highly resistant to inhibition of its activity, and therefore, did not have the necessary sensitivity to serve as a proxy for examining potential toxicity.

Microfluidic Shrinking Droplet Concentrator for Analyte Detection and Phase Separation of Protein Solutions.

We develop a droplet microfluidic platform to increase the concentration of analytes in solution via reduction of the sample volume under well-defined conditions. This approach improves the detection and quantification of analytes without requiring any a priori information on their structure nor physical chemical properties. Samples are compartmentalized and processed in water-in-oil droplets that are individually stored in cylindrical microwells located on top of a microfluidic channel. The individual droplets shrink over time due to water extraction in the surrounding oil, leading to an increase in the analyte concentration up to 100,000-fold within the droplet. We demonstrate the power of this approach for detection applications by quantifying a broad range of single analytes such as small molecules, proteins, nanoparticles, exosomes, and amyloid fibrils. With this setup, we can measure pM concentrations, corresponding to zeptomole (10-21 mol) amounts encapsulated in individual droplets. We further show that the droplet concentrator device, or DroMiCo, can quantify unlabeled proteins in nM concentrations and analyze multicomponent mixtures when coupled with a prefractionation step. We illustrate this concept by detecting femtomoles (10-15 mol) of soluble protein oligomers prefractionated by size exclusion chromatography. Finally, we apply the DroMiCo to the analysis of phase diagrams of macromolecules, including synthetic polymers and proteins. Specifically, we analyze the liquid-liquid phase separation of an in vitro model of cellular membraneless compartments, composed of a phase separating protein in the presence of defined concentrations of molecular modulators such as RNA and ATP.

Dynamics of Synthetic Membraneless Organelles in Microfluidic Droplets.

Cells can form membraneless organelles by liquid-liquid phase separation. As these organelles are highly dynamic, it is crucial to understand the kinetics of these phase transitions. Here, we use droplet-based microfluidics to mix reagents by chaotic advection and observe nucleation, growth, and coarsening in volumes comparable to cells (pL) and on timescales of seconds. We apply this platform to analyze the dynamics of synthetic organelles formed by the DEAD-box ATPase Dhh1 and RNA, which are associated with the formation of processing bodies in yeast. We show that the timescale of phase separation decreases linearly as the volume of the compartment increases. Moreover, the synthetic organelles coarsen into one single droplet via gravity-induced coalescence, which can be arrested by introducing a hydrogel matrix that mimics the cytoskeleton. This approach is an attractive platform to investigate the dynamics of compartmentalization in artificial cells.

Pat1 promotes processing body assembly by enhancing the phase separation of the DEAD-box ATPase Dhh1 and RNA.

Processing bodies (PBs) are cytoplasmic mRNP granules that assemble via liquid-liquid phase separation and are implicated in the decay or storage of mRNAs. How PB assembly is regulated in cells remains unclear. Previously, we identified the ATPase activity of the DEAD-box protein Dhh1 as a key regulator of PB dynamics and demonstrated that Not1, an activator of the Dhh1 ATPase and member of the CCR4-NOT deadenylase complex inhibits PB assembly in vivo (Mugler et al., 2016). Here, we show that the PB component Pat1 antagonizes Not1 and promotes PB assembly via its direct interaction with Dhh1. Intriguingly, in vivo PB dynamics can be recapitulated in vitro, since Pat1 enhances the phase separation of Dhh1 and RNA into liquid droplets, whereas Not1 reverses Pat1-Dhh1-RNA condensation. Overall, our results uncover a function of Pat1 in promoting the multimerization of Dhh1 on mRNA, thereby aiding the assembly of large multivalent mRNP granules that are PBs.

Engineering Aspects of Protein Interactions and Self-assembly.

In the new Laboratory for Biochemical Engineering (LBCE) at ETH Zurich researchers combine principles of chemical engineering with microfluidic technology and biophysical methods to investigate the physical determinants of biomolecular self-assembly in living organisms. In this account, we show the impact of this activity on concrete applications in biomedical sciences and biotechnology. We focus in particular on the field of protein aggregation and phase separation, and we highlight examples in the context of diagnosis and treatment of Alzheimer's disease and neurodegenerative disorders, cell compartmentalization as well as manufacturing and delivery of therapeutic proteins.

The dark matter of the cancer genome: aberrations in regulatory elements, untranslated regions, splice sites, non-coding RNA and synonymous mutations.

Cancer is a disease of the genome caused by oncogene activation and tumor suppressor gene inhibition. Deep sequencing studies including large consortia such as TCGA and ICGC identified numerous tumor-specific mutations not only in protein-coding sequences but also in non-coding sequences. Although 98% of the genome is not translated into proteins, most studies have neglected the information hidden in this "dark matter" of the genome. Malignancy-driving mutations can occur in all genetic elements outside the coding region, namely in enhancer, silencer, insulator, and promoter as well as in 5'-UTR and 3'-UTR Intron or splice site mutations can alter the splicing pattern. Moreover, cancer genomes contain mutations within non-coding RNA, such as microRNA, lncRNA, and lincRNA A synonymous mutation changes the coding region in the DNA and RNA but not the protein sequence. Importantly, oncogenes such as TERT or miR-21 as well as tumor suppressor genes such as TP53/p53, APC, BRCA1, or RB1 can be affected by these alterations. In summary, coding-independent mutations can affect gene regulation from transcription, splicing, mRNA stability to translation, and hence, this largely neglected area needs functional studies to elucidate the mechanisms underlying tumorigenesis. This review will focus on the important role and novel mechanisms of these non-coding or allegedly silent mutations in tumorigenesis.

Limited role of free TDP-43 as a diagnostic tool in neurodegenerative diseases.

TAR DNA-binding protein 43 (TDP-43) is one of the neuropathological hallmarks in amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD). It is present in patients' blood and cerebrospinal fluid (CSF); however, the source and clinical relevance of TDP-43 measurements in body fluids is uncertain. We investigated paired CSF and serum samples, blood lymphocytes, brain urea fractions and purified exosomes from CSF for TDP-43 by one- (1D), and two-dimensional (2D) Western immunoblotting (WB) and quantitative mass spectrometry (MRM) in patients with ALS, FTLD and non-neurodegenerative diseases. By means of 2D-WB we were able to demonstrate a similar isoform pattern of TDP-43 in lymphocytes, serum and CSF in contrast to that of brain urea fractions with TDP-43 pathology. We found that the TDP-43 CSF to blood concentration ratio is about 1:200. As a possible brain specific fraction we found TDP-43 in exosome preparations from CSF by immunoblot and MRM. We conclude that TDP-43 in CSF originates mainly from blood. Measurements of TDP-43 in CSF and blood are of minor importance as a diagnostic tool, but may be important for monitoring therapy effects of TDP-43 modifying drugs.