Engineering Aspects of Protein Interactions and Self-Assembly

304 CHIMIA 2018, 72, No. 5 News from New Chemistry Professors iN switzerlaNd

doi:10.2533/chimia.2018.304 Chimia 72 (2018) 304–308 © Swiss Chemical Society Engineering Aspects of Protein Interactions and Self-assembly

Miriam Linsenmeier and Paolo Arosio*

Abstract: In the new Laboratory for Biochemical Engineering (LBCE) at ETH Zurich researchers combine prin- ciples 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. Keywords: Alzheimer’s disease · Amyloids · Microfluidics · Protein aggregation · Protein interactions · Protein phase separation

1. Protein Structure and changes and even to single point mutations Interactions of the protein sequence, and is not free of mistakes. In some cases, aberrant interac- Self-organization is one of the funda- tions leading to misfolding and undesired mental processes underlying the origin aggregation can occur.[1] This is the case of life. The series of cellular functions on for instance of amyloids, which represent which life depends originate at the molec- a particularly important class of aberrant ular level from interactions between bio- aggregates of peptides and proteins which molecules, namely proteins and nucleic are involved in several human disorders acids. Proteins represent a unique class of including Alzheimer’s, Parkinson’s and lyophilic colloids with heterogeneous and systemic amyloidoses.[2] Amyloids can Paolo Arosio obtained his master’s complex structures, which exhibit ampho- originate from a broad range of very dif- degree in chemical engineering from the teric polyelectrolyte nature and the simul- ferent peptides and proteins and exhibit Politecnico di Milano in 2007, and his taneous presence of hydrophobic and hy- a remarkably consistent supramolecular doctoral degree from the Swiss Federal drophilic patches. Proteins are synthesized structure, which is mostly independent Institute of Technology ETH Zurich in as linear chains of amino acids which can of the precursor protein and consists of a 2011, working under the supervision of fold into suitable secondary and tertiary diameter of few nanometers, a length of Prof. M. Morbidelli. His PhD work was 3-dimensional structures. In some cases, several microns, and a predominant cross awarded the ETH Medal. Subsequently the stability and the activity of proteins β-sheet secondary structure.[2c] This com- he carried out postdoctoral research at the require the formation of quaternary struc- mon structure is primarily due to the high Department of Chemistry at the University tures, such as complexation into oligomers number of hydrogen bonds that proteins of Cambridge, UK, with Prof. T. P. J. or assembly into large filaments. can form along the β-sheet structure. In Knowles, funded by the Swiss National Structure and intermolecular forces addition to these aspecific interactions, the Science Foundation (SNSF) and by the are two strongly interconnected aspects in thermodynamic stability of the fibrils is European Marie Curie Fellowship scheme proteins: evolution has optimized the for- further promoted by the exclusion of water for career development. In 2016 he re- mation of secondary, tertiary and quater- molecules and specific side-chain interac- turned to ETH as Assistant Professor in nary structure by controlling a delicate bal- tions within the inner core of the fibrils.[2c] Biochemical Engineering. His research ance of intramolecular and intermolecular Although amyloid structures were interests focus on understanding and con- forces, which include electrostatic, van originally discovered in association with trolling protein self-assembly and protein der Waals, solvent and hydrophobic inter- a variety of pathological conditions,[2a,c] interaction processes that underlie prob- actions. In particular, hydrophobic forces increasing evidence demonstrates that am- lems of key fundamental and practical im- play a major role in protein folding, since yloids underlie a series of physiological portance in biology and biotechnology. globular proteins show stable native struc- functions,[3] including the natural storage tures in which hydrophilic groups are sol- of peptide hormones,[4] the formation of vent-exposed, while hydrophobic patches adhesive microbe biofilms[5] and cell com- are buried intramolecularly. At the same partmentalization.[6] time, the complex surface chemistry and As biochemical engineers we are inter- structure of proteins allow these molecules ested in understanding the physical deter- *Correspondence: Prof. P. Arosio to perform specific tasks by generating minants of biomolecular assembly in living ETH Zurich highly specific intermolecular interactions organisms and their connection with func- Department of Chemistry and Applied Biosciences with other proteins and biomolecules. tional and aberrant behaviors. This activity Institute for Chemical and Bioengineering, Vladimir Prelog Weg 1, CH-8093 Zurich This subtle balance of intermolecular requires addressing a series of theoretical E-mail: [email protected] forces is highly sensitive to environmental and analytical challenges, since biological News from New Chemistry Professors iN switzerlaNd CHIMIA 2018, 72, No. 5 305

systems consist of heterogeneous mixtures the generation of the soluble intermediates. of nuclei before fibril growth can occur. that often exhibit transient interactions, In analogy to traditional fields of chemical Chemical kinetic analysis has revealed that low concentration of metastable species engineering, such as polymer reaction en- each individual microscopic reaction of and complex aggregation networks. In gineering, colloidal dispersions and com- growth and secondary nucleation is actu- the following paragraphs we show how bustion processes, also in this context con- ally present from the very beginning of the concepts of chemical engineering lead cepts of reaction engineering and chemical aggregation process, and that the lag-phase to theoretical and experimental advances kinetics allow to unravel complex kinetic represents the time required for fibrils to which allow to achieve quantitative infor- schemes from a limited number of avail- amplify and reach a critical concentration mation about protein self-assembly pro- able macroscopic data (Fig. 1).[9] which is detectable by experimental ana- cesses at the molecular level. We highlight In addition to the identification of the lytical methods.[12] This concept has been the direct implications of these findings in critical microscopic steps that are most re- exploited to develop a highly sensitive as- different areas of biological and biomedi- sponsible for the generation of toxic spe- say for the quantification of low concentra- cal sciences, including the diagnosis and cies,[10] this information is fundamental to tions of amyloid fibrils.[13] therapy of Alzheimer’s disease and cell explain several aspects of the aggregation Measurements of aggregation rates can compartmentalization. We finally discuss process.[11] For instance, the application also be used to indirectly quantify specific how the developed approaches and the les- of chemical kinetic analysis has eluci- interactions between molecules by analyz- sons learned from nature allow to design dated the nature and the molecular origin ing changes in the aggregation profiles in functional materials as well as to monitor of the macroscopic lag-phase that is typi- the absence and presence of the binding and engineer the stability of proteins in cally observed in the formation of amy- partner.[14] This approach is attractive in biotechnology. loids from solutions of soluble peptides applications characterized by the presence and proteins.[12] This lag-phase could be of complex mixtures as well as transient erroneously interpreted as a waiting time and metastable interactions. One example 2. Reaction Engineering in required to reach a critical concentration is the analysis of the mechanisms of inhi- Amyloids and Alzheimer’s Disease

The formation of amyloid plaques of the peptide Abeta42 (Aβ42) or analogous peptides, which originate from the frag- a mentation of the amyloid precursor protein (APP), is a hallmark of Alzheimer’s disease (AD), a devastating form of dementia that affects more than one hundred thou- sand people in Switzerland and millions of people worldwide.[7] An increasing amount of evidence indicates that the formation of amyloid fibrils is a critical upstream pro- DNNAJB6 b proSP-C Brichos cess in the series of events leading to the loss of neuronal function.[8] Today no ef- Hsp70 fective pharmaceutical treatment exists for this disorder, a fact that reflects our cur- rent lack of understanding of the molecu- (ii) lar mechanisms underpinning the disease. ooligligoommeerr + ele ongation Despite the direct proof of the causality monomer secondary between the formation of aggregates and primary fibfibrrilil nucleation nucleation (iii) the disease is still lacking, the aggregation (i) process represents an attractive target for therapeutic intervention.

In particular, it is emerging that soluble Increasedmass transport oligomersgeneratedduringtheaggregation c (PrimaryNucleation) process could represent a source of toxicity via several possible mechanisms.[2a] The experimental characterization of Elongation k these species is primarily impaired by + their metastability and low concentrations. Fragmentation k- Moreover, the identification of the time (Secondarynucleation) evolution of their concentration is compli- Heterogeneous cated by the variety of microscopic steps of primarynucleation nucleation and growth which compose the Surface Bulk, Mechanical forces (shear) aggregation process. Indeed, in addition to primary nucleation events, amyloid fibrils Fig. 1. Reaction engineering in protein aggregation. (A) Chemical kinetics allows to extract infor- grow via elongation and several secondary mation about the microscopic mechanisms of nucleation and growth underlying the formation nucleation processes, including fibril frag- of amyloids. (B) This analysis can lead to the quantification of specific interactions in complex mixtures as well as to the identification of perturbations of individual reactions in complex kinetic mentation and monomer-dependent sec- schemes. This strategy has allowed for instance to identify the variety of microscopic mecha- [2a] ondary nucleation reactions (Fig. 1A). nisms of inhibition of molecular chaperones against the formation of amyloid fibrils.[18a] (C) The It is therefore crucial to identify the contri- same approach shed light on the combined effect of interfaces and mechanical agitation on the bution of these individual reactions to the formation of amyloid fibrils from soluble human insulin.[41] Figures reproduced with permission of overall aggregation rate and in particular to Springer Nature (ref. [18a]) and ACS Publications (ref. [41]). 306 CHIMIA 2018, 72, No. 5 News from New Chemistry Professors iN switzerlaNd

bition of amyloid fibrils. Inhibitors can in- teract with several soluble and aggregated kT (i) (ii) protein species in the reaction mixture, χ Single phase (i) and therefore they can potentially perturb many different microscopic reactions. Depending on the specific reactions that are affected in the aggregation network, different modalities of inhibition can have dramatically different consequences on the LLPS (ii) generation of the toxic intermediates.[14] In particular, some modalities could increase Φc ΦD Φ the toxicity even if the overall aggregation Fig. 2. Thermodynamics of liquid mixtures. Protein solutions exhibit complex phase diagrams, in- [15] rate is retarded. It is clear that an effec- cluding liquid-liquid phase separation (LLPS) induced by attractive intermolecular forces (χ). LLPS tive intervention should aim at inhibiting generates a dispersed phase at high protein concentration (ϕ ) into a continuous phase at diluted D the specific steps responsible for the gen- protein concentration (ϕ ). Recent findings indicate that this process underlies the formation of c eration of the toxic oligomers rather than membraneless bodies in cells. Several mechanisms can trigger the assembly and disassembly of at generically arresting the aggregation these bodies, including interactions between proteins and RNA molecules. The images show the rate.[14] In this context, chemical kinetics phase separation of a solution of the model DEAD-box protein Dhh1 associated with the forma- is a powerful tool to identify specific in- tion of processing bodies in the absence (i) and presence (ii) of RNA.[26] Scale bar is 25 µm. teractions between the inhibitor and the several protein species present in the re- stress-induced adjustment of the cellular tion observed with amyloids, in which the action mixture, as well as to discriminate metabolism.[20–22] These membraneless dominant energetic driving force for ag- which specific microscopic reactions are bodies have been observed in the nucleus, gregation is represented by aspecific hy- affected by the presence of the inhibitor.[14] in the cytoplasm and in membrane-associ- drogen bonds along the cross-β-sheet For instance, the aggregation scheme of ated parts of the cell, and are rich in nucleic structures. This aspecificity largely ex- the Aβ42 peptide is characterized by the acids and RNA-binding proteins. Despite plains the reason why many different solu- presence of a secondary nucleation reac- the widespread appearance in many differ- ble peptides and proteins can convert into tion catalyzed by the surfaces of existing ent organisms, the mechanisms that control a similar, generic amyloid structure.[2a] In fibrils.[16] Molecules capable of inhibiting their reversible formation and disassembly contrast, different membraneless compart- this molecular process are highly efficient as well as their biophysical properties re- ments require specific compositions and in removing the toxicity associated with main largely elusive. tailored biophysical properties to perform the formation of amyloids without modify- Most of these proteins possess N- and/ specific tasks, which are encoded by spe- ing the final fibril load.[16] The application or C-terminal low-complexity domains cific interactions. of this kinetic approach is leading to the (LCDs), which are intrinsically disordered In analogy to protein folding, misregu- identification of several small molecules[17] and dynamic structures that appear to play lation of the delicate balance of interactions and antibodies[15] capable of targeting the a key role in the liquid–liquid phase transi- in LLPS can promote the transition from a nucleation processes in AD. tion underlying the formation of the bod- liquid-like state to solid, irreversible aggre- Moreover, by applying the same plat- ies.[23,24] The sequences of the LCDs con- gates.[27] This behavior has been observed form we have shown that molecular chap- tain a limited set of amino acids which are for instance with the RNA-binding protein erones, which are key components of the frequently arranged in repetitive sequence Fused in Sarcoma (FUS), which is associ- protein homeostasis network, can inhibit motifs and encode specific intermolecu- ated with Amyotrophic Lateral Sclerosis the formation of amyloids via a variety of lar interactions: for instance, glutamine (ALS), a neurodegenerative disease which molecular mechanisms, thereby proving and asparagine amino acids induce polar is characterized by the degeneration of that nature has evolved multiple strategies forces, interspersed aromatic amino ac- motor neurons in brain and spinal cord. to avoid aberrant protein aggregation[18] ids enable π-π stacking, and repetitive Wild-type FUS possesses an N-terminal (Fig. 1B). sequence motifs, such as polyQ and RG/ low-complexity domain and several RGG- RGG repeats,[23,25] promote multivalent in- rich regions which promote the formation teractions.[20] The specificity of these inter- of stable liquid–liquid phase separation. 3. Protein Liquid–Liquid Phase actions is important for the cell to carefully The substitution of a single amino acid Separation control the balance of forces required for (G156E) associated with ALS pathology the assembly and disassembly of the bod- results in the aging of the protein-rich Attractive protein–protein interac- ies. The presence of attractive interactions droplets into fibrillar, solid amyloids.[25b] tions do not result only in the formation is required to induce LLPS but at the same These findings indicate that attractive of aggregates. Protein solutions exhibit a time this attractive energy should not lead intermolecular interactions in protein solu- rich phase diagram that includes crystals, to irreversible structures that would prevent tions can promote the formation of a con- nematic phases and liquid–liquid phase the recovery of the soluble components. A tinuum of protein-rich phases from liquid separation (LLPS).[19] The latter process is picture is emerging in which cells have to solid,[28] and that this process is highly characterized by the formation of a highly developed a series of control mechanisms sensitive to small perturbations, including concentrated disperse phase within a dilute to tune the biophysical properties and the for instance minor changes in the amino continuous phase (Fig. 2). Recent findings dynamics of the protein-rich droplets, in- acid sequence induced by mutations or revealed that LLPS plays an important role cluding for instance interactions between post-translational modifications. Given the in cell compartmentalization.[20] In addi- proteins and RNA molecules,[26] post- fatal consequences that aberrant protein tion to vesicle-like organelles, cells can translational modifications[6] and binding aggregation can have in vivo, it is of fun- form dynamic membraneless organelles to ATP. [6] damental importance to identify the mo- with defined composition and widespread The high specificity of the biomolecu- lecular determinants of control functional functions, including ribosome biogenesis, lar interactions underpinning liquid phase LLPs to identify targets for intervention RNA processing and storage, as well as transition is very different from the situa- when misregulated behavior is observed. News from New Chemistry Professors iN switzerlaNd CHIMIA 2018, 72, No. 5 307

In analogy with the field of amyloids, the insulin, one of the most common therapeu- 5. Conclusions and Outlook application of concepts of polymer phys- tic peptides. The application of chemical ics,[21] chemical kinetics and thermody- kinetic analysis enabled us to go beyond In summary, we have shown that key namics of complex mixtures[29] can have a the empirical observation of this effect and concepts of chemical engineering, such as significant impact on activity. demonstrate that interfaces specifically reaction engineering, thermodynamics of promote heterogeneous nucleation events, complex mixtures and mass transport phe- while mechanical agitation amplifies the nomena at the micron scale, can contribute 4. From Nature to Biotechnology aggregation rate by promoting the turnover to theoretical and practical problems of of species on the surface and the fragmen- biomolecular self-assembly. Implications In the paragraphs above we have dis- tation of fibrils in bulk.[41] These findings of these studies range from biomedical cussed some examples of the rich pan- can explain some contradictory results engineering to biotechnology. Concrete orama of protein interactions and self- reported in the literature about the effect examples include the diagnosis and treat- assembly in living organisms. The lessons of shear on protein aggregation and have ment of Alzheimer’s disease, the issue of learned from biology open the possibil- practical implications in the bioprocessing cell compartmentalization and the stability ity to develop advanced protein materials of proteins. of therapeutic proteins during industrial that exhibit attractive features in terms The analytical characterization of pro- manufacturing. of function, specificity and biocompat- tein aggregates and, more generally, of bio- We finally note that the applications of ibility.[30] Protein cages,[31] functionalized colloids is another area where biochemical these approaches in biological sciences are amyloids[30a] and self-assembling pep- engineers can largely contribute. In this not limited to the field of protein aggre- tides[32] are only a few examples of the va- context novel approaches based on mi- gation. Theoretical and experimental ad- riety of materials that have been recently crofluidic technology[42] are emerging as vances based on microfluidic technology developed in laboratories in Switzerland powerful tools to improve the experimen- have the power to address a variety of other and in the world. These materials find ap- tal characterization of proteins and biocol- problems including for instance interac- plications in a broad range of fields includ- loids in both fundamental studies and prac- tions between proteins and nucleic acids, ing water treatment,[33] food science,[34] tical applications,[43] including for instance cellular chemotaxis, separation and char- vaccine delivery,[35] drug encapsulation[36] the industrial bioprocessing of therapeutic acterization of exosomes and vesicles. We and enzyme technology.[37] proteins.[44] have started to apply our methods to these In other applications of material sci- The manipulation of matter at the mi- systems, which have practical implications ences protein aggregation represents an cron scale allows to accurately control heat in other areas of biomedical sciences and undesired process that must be carefully and mass transport phenomena, which have biotechnology such as immunology, drug avoided. This is for instance the case of different features with respect to the bulk delivery, and cosmetics. therapeutic proteins, a class of drugs that and open the possibility to perform opera- have been recently developed against dis- tions which are not achievable with tradi- Acknowledgements [43] eases such as multiple sclerosis, rheuma- tional methods. Moreover, the reduced Work in our laboratory is funded by a start- toid arthritis, tumors and hemophilia.[38] amount of material and time required by up grant from ETH Zurich, the Swiss National The industrial production of these drugs the analysis, inherent to miniaturization, Science Foundation (205321_179055), the involves a series of steps (expression, pu- largely increase the throughput of the ETH Zurich Research Grant ETH-33 17-2, rification, shipping and storage) which can measurements. A recent example is repre- the Synapsis Foundation and the Novartis expose the proteins to different types of sented by a microfluidic diffusion platform Foundation for medical-biological Research. stresses, such as acidic, thermal and me- in which the dominance of Brownian mo- We thank Marie Kopp (ETH Zurich) for acquir- chanical treatment.[39] Such stresses can tion under laminar flow regime is exploited ing the images shown in Figure 3C. compromise the chemical and physical to accurately measure the time and spatial Received: February 28, 2018 stability of proteins and, in some cases, evolution of diffusion profiles[45] (Fig. induce aggregation. Aggregation must be 3A). This technique allows to evaluate the strictly avoided, not only because it leads size distribution of complex mixtures in [1] F. Chiti, C. M. Dobson, Annu. Rev. Biochem. to loss of valuable product, but, most im- the submicron range, as well as to quan- 2006, 75, 333. portantly, because aggregates are poten- tify specific interactions between biomol- [2] a) T. P. J. Knowles, M. Vendruscolo, C. M. Dobson, Nat. Rev. Mol. Cell Biol. 2014, 15, 384; [39] [45] tially immunogenic. The stability dur- ecules in solution. The same approach b) P. T. Lansbury, H. A. Lashuel, Nature 2006, ing manufacturing and administration is can also be applied to measure the viscos- 443, 774; c) R. Riek, D. S. Eisenberg, Nature therefore crucial for the successful appli- ity of complex solutions at zero shear by 2016, 539, 227. cation of these drugs. probing the Brownian motion of standard [3] a) D. M. Fowler, A. V. Koulov, W. E. Balch, J. W. [46] Kelly, Trends Biochem. Sci. 2007, 32, 217; b) L. P. Also in this context, the application tracers (Fig. 3B). Blanco, M. L. Evans, D. R. Smith, M. P. Badtke, of concepts of reaction engineering and Another strength of microfluidics in M. R. Chapman, Trends Microbiol. 2012, 20, 66. chemical kinetics can reveal the complex protein science is the possibility to manip- [4] S. K. Maji, M. H. Perrin, M. R. Sawaya, S. molecular mechanisms underlying aggre- ulate multiphase flows and compartmen- Jessberger, K. Vadodaria, R. A. Rissman, P. S. gation.[40] This analysis has implications in talize solutions in small reactors of pico- Singru, K. P. R. Nilsson, R. Simon, D. Schubert, D. Eisenberg, J. Rivier, P. Sawchenko, W. Vale, R. the identification of lead candidates during and nano-liter by generating water-in-oil Riek, Science 2009, 325, 328. the early-stages of the process as well as emulsions[42,47] (Fig. 3C). Droplet micro- [5] M. R. Chapman, L. S. Robinson, J. S. Pinkner, R. in the optimization of processing and for- fluidic technology is particularly suitable Roth, J. Heuser, M. Hammar, S. Normark, S. J. mulation conditions. A recent example is to create physiological-like environments Hultgren, Science 2002, 295, 851. [6] S. Saad, G. Cereghetti, Y. H. Feng, P. Picotti, M. represented by our study on the analysis and investigate phase transition, nucleation Peter, R. Dechant, Nat. Cell Biol. 2017, 19, 1202. of the synergistic effect of interfaces and events, reaction-diffusion phenomena and [7] World Alzheimer Report 2016, Alzheimer’s mechanical forces on protein aggrega- confinement effects in the absence of air- Disease International (ADI), London, 2016. tion[41] (Fig. 1C). We have shown that un- water interfaces.[48] Moreover the possibil- [8] D. J. Selkoe, J. Hardy, EMBO Mol. Med. 2016, 8, der physiological conditions hydrophobic/ ity to generate thousands of identical drop- 595. [9] G. Meisl, J. B. Kirkegaard, P. Arosio, T. C. T. hydrophilic interfaces can dramatically ac- lets in short time dramatically improves the Michaels, M. Vendruscolo, C. M. Dobson, S. celerate the formation of fibrils of human throughput of the measurements. Linse, T. P. J. Knowles, Nat. Protoc. 2016, 11, 252. 308 CHIMIA 2018, 72, No. 5 News from New Chemistry Professors iN switzerlaNd

a 12X Size distribution

% Free Ligand 60 Complex Ligand - Deconvolution 40 Ta rget

0 1 10 5mm r(nm)

2 b Microfluidics c aqueousphase 1.8 DLS aqueous 1.6 oil phase

r oil

η 1.4

1.2 oil 1

0.8 0 20 40 60 80 100 BSAconcentration (mg/mL )

Fig. 3. Separation processes and mass transport at the micron scale. A) Microfluidic diffusion platform for the analysis of interactions and size distri- butions in the submicron range. B) The same device can be applied to measure the viscosity of protein solutions and complex mixtures. C) Images of a co-flow and a T-junction 2-D device for the generation of water-in-oil emulsions.[42,47] Images have been acquired by brightfield and fluorescence microscopy, respectively. Scale bar is 50 µm. These new methods find applications in many fields of protein science, including the analysis of protein stability during bioprocessing and formulation of therapeutic proteins.[44] Figures reproduced with permission of ACS Publications, refs [45, 46].

[10] S. I. A. Cohen, S. Linse, L. M. Luheshi, E. Brangwynne, Proc. Natl. Acad. Sci. USA 2015, [36] U. Shimanovich, I. Efimov, T. O. Mason, P. Hellstrand, D. A. White, L. Rajah, D. E. Otzen, 112, 7189. Flagmeier, A. K. Buell, A. Gedanken, S. Linse, K. M. Vendruscolo, C. M. Dobson, T. P. J. Knowles, [24] A. Molliex, J. Temirov, J. Lee, M. Coughlin, A. P. S. Akerfeldt, C. M. Dobson, D. A. Weitz, T. P. J. Proc. Natl. Acad. Sci. USA 2013, 110, 9758. Kanagaraj, H. J. Kim, T. Mittag, J. P. Taylor, Cell Knowles, ACS Nano 2015, 9, 43. [11] P. Arosio, Pharm. Res. 2017, 34, 2628. 2015, 163, 123. [37] X. M. Zhou, A. Entwistle, H. Zhang, A. P. [12] P. Arosio, T. P. J. Knowles, S. Linse, Phys. Chem. [25] a) P. Thandapani, T. R. O’Connor, T. L. Bailey, S. Jackson, T. O. Mason, U. Shimanovich, T. P. J. Chem. Phys. 2015, 17, 7606. Richard, Mol. Cell 2013, 50, 613; b) A. Patel, H. O. Knowles, A. T. Smith, E. B. Sawyer, S. Perrett, [13] P. Arosio, R. Cukalevski, B. Frohm, T. P. J. Lee, L. Jawerth, S. Maharana, M. Jahnel, M. Y. Hein, ChemCatChem 2014, 6, 1961. Knowles, S. Linse, J. Am. Chem. Soc. 2014, 136, S. Stoynov, J. Mahamid, S. Saha, T. M. Franzmann, [38] S. Aggarwal, Nat. Biotechnol. 2014, 32, 32. 219. A. Pozniakovski, I. Poser, N. Maghelli, L. A. Royer, [39] W. Jiskoot, T. W. Randolph, D. B. Volkin, C. R. [14] P. Arosio, M. Vendruscolo, C. M. Dobson, T. P. J. M. Weigert, E. W. Myers, S. Grill, D. Drechsel, A. A. Middaugh, C. Schoneich, G. Winter, W. Friess, D. Knowles, Trends Pharmacol. Sci. 2014, 35, 127. Hyman, S. Alberti, Cell 2015, 162, 1066. J. A. Crommelin, J. F. Carpenter, J. Pharm. Sci. [15] F. A. Aprile, P. Sormanni, M. Perni, P. Arosio, [26] C. F. Mugler, M. Hondele, S. Heinrich, R. 2012, 101, 946. S. Linse, T. P. J. Knowles, C. M. Dobson, M. Sachdev, P. Vallotton, A. Y. Koek, L. Y. Chan, K. [40] a) P. Arosio, S. Rima, M. Lattuada, M. Morbidelli, Vendruscolo, Sci. Adv. 2017, 3, e1700488. Weis, Elife 2016, 5, e18746. J. Phys. Chem. B 2012, 116, 7066; b) C. J. [16] S. I. A. Cohen, P. Arosio, J. Presto, F. R. Kuruden- [27] Z. Monahan, V. H. Ryan, A. M. Janke, K. A. Roberts, Biotechnol. Bioeng. 2007, 98, 927; c) kandy, H. Biverstal, L. Dolfe, C. Dunning, X. T. Burke, S. N. Rhoads, G. H. Zerze, R. O’Meally, P. Arosio, S. Rima, M. Morbidelli, Pharm. Res. Yang, B. Frohm, M. Vendruscolo, J. Johansson, G. L. Dignon, A. E. Conicella, W. Zheng, R. B. 2013, 30, 641. C. M. Dobson, A. Fisahn, T. P. J. Knowles, S. Best, R. N. Cole, J. Mittal, F. Shewmaker, N. L. [41] F. Grigolato, C. Colombo, R. Ferrari, L. Rezab- Linse, Nat. Struct. Mol. Biol. 2015, 22, 207. Fawzi, EMBO J. 2017, 36, 2951. kova, P. Arosio, ACS Nano 2017, 11, 11358. [17] J. Habchi, S. Chia, R. Limbocker, B. Mannini, [28] A. Aguzzi, M. Altmeyer, Trends Cell Biol. 2016, [42] G. M. Whitesides, Nature 2006, 442, 368. M. Ahn, M. Perni, O. Hansson, P. Arosio, J. R. 26, 547. [43] J. Charmet, P. Arosio, T. P. J. Knowles, J. Mol. Kumita, P. K. Challa, S. I. A. Cohen, S. Linse, C. [29] M. T. Wei, S. Elbaum-Garfinkle, A. S. Holehouse, Biol. 2018, 430, 565. M. Dobson, T. P. J. Knowles, M. Vendruscolo, C. C. Chen, M. Feric, C. B. Arnold, R. D. [44] M. R. G. Kopp, P. Arosio, J. Pharm. Sci. 2018, Proc. Natl. Acad. Sci. USA 2017, 114, E200. Priestley, R. V. Pappu, C. P. Brangwynne, Nat. doi:10.1016/j.xphs.2018.01.001. [18] a) P. Arosio, T. C. T. Michaels, S. Linse, Chem. 2017, 9, 1118. [45] P. Arosio, T. Muller, L. Rajah, E. V. Yates, F. C. Mansson, C. Emanuelsson, J. Presto, J. [30] a) T. P. J. Knowles, R. Mezzenga, Adv. Mater. A. Aprile, Y. B. Zhang, S. I. A. Cohen, D. A. Johansson, M. Vendruscolo, C. M. Dobson, T. P. 2016, 28, 6546; b) G. Wei, Z. Q. Su, N. P. White, T. W. Herling, E. J. De Genst, S. Linse, J. Knowles, Nat. Commun. 2016, 7, 10948; b) C. Reynolds, P. Arosio, I. W. Hamley, E. Gazit, R. M. Vendruscolo, C. M. Dobson, T. P. J. Knowles, Mansson, P. Arosio, R. Hussein, H. H. Kampinga, Mezzenga, Chem. Soc. Rev. 2017, 46, 4661. ACS Nano 2016, 10, 333. R. M. Hashem, W. C. Boelens, C. M. Dobson, T. [31] Y. Azuma, R. Zschoche, M. Tinzl, D. Hilvert, [46] P. Arosio, K. Hu, F. A. Aprile, T. Muller, T. P. J. P. J. Knowles, S. Linse, C. Emanuelsson, J. Biol. Angew. Chem. Int. Ed. 2016, 55, 1531. Knowles, Anal. Chem. 2016, 88, 3488. Chem. 2014, 289, 31066. [32] H. G. Cui, M. J. Webber, S. I. Stupp, Biopolymers [47] a) O. J. Dressler, X. C. I. Solvas, A. J. deMello, [19] R. Piazza, Curr. Opin. Colloid Interface Sci. 2010, 94, 1. Annu. Rev. Anal. Chem. 2017, 10; b) A. S. Utada, 2000, 5, 38. [33] S. Bolisetty, R. Mezzenga, Nat. Nanotechnol. E. Lorenceau, D. R. Link, P. D. Kaplan, H. A. [20] S. F. Banani, H. O. Lee, A. A. Hyman, M. K. 2016, 11, 365. Stone, D. A. Weitz, Science 2005, 308, 537. Rosen, Nat. Rev. Mol. Cell Bio. 2017, 18, 285. [34] Y. Shen, L. Posavec, S. Bolisetty, F. M. Hilty, G. [48] T. P. J. Knowles, D. A. White, A. R. Abate, J. J. [21] C. P. Brangwynne, P. Tompa, R. V. Pappu, Nat. Nystrom, J. Kohlbrecher, M. Hilbe, A. Rossi, J. Agresti, S. I. A. Cohen, R. A. Sperling, E. J. De Phys. 2015, 11, 899. Baumgartner, M. B. Zimmermann, R. Mezzenga, Genst, C. M. Dobson, D. A. Weitz, Proc. Natl. [22] Y. Shin, C. P. Brangwynne, Science 2017, 357, Nat. Nanotechnol. 2017, 12, 642. Acad. Sci. USA 2011, 108, 14746. 6357. [35] G. A. Hudalla, T. Sun, J. Z. Gasiorowski, H. F. [23] S. Elbaum-Garfinkle, Y. Kim, K. Szczepaniak, Han, Y. F. Tian, A. S. Chong, J. H. Collier, Nat. C. C. Chen, C. R. Eckmann, S. Myong, C. P. Mater. 2014, 13, 829.