Liquid Liquid Phase Separation in Cells

May 23, 2018 to May 25, 2018
Location : CECAM-HQ-EPFL, Lausanne, Switzerland
   EPFL on iPhone
   Visa requirements


  • Julian Shillcock (EPFL, Blue Brain Project, Switzerland)
  • Luis Bagatolli (Yachay Tech University, Ecuador)
  • Mikko Haataja (Princeton University, USA)
  • John H. Ipsen (The University of Southern Denmark, Denmark)




It is increasingly recognized that the cytoplasm of many cell types is not a homogeneous fluid environment. Experiments have revealed compositionally-distinct droplets within the cell’s cytoplasm (Hyman and Brangwynne, 2011, Weber and Brangwynne, 2015; Courchaine et al., 2016). Such droplets include the nucleolus, centrosomes, stress granules and RNA particles, and provide the cell with a powerful mechanism to spatially organize biochemical reactions (Banani et al., 2017). This class of condensed phases has recently been extended to include the post-synaptic density (PSD) in neuronal synapses that are important for synaptic plasticity (Zeng et al., 2016), and it has been suggested that vesicle clusters form a distinct phase inside the active zone of the pre-synaptic neuron (Milovanovic and De Camilli, 2017). Recent experiments have reconstituted condensed phase droplets in vitro (Lin et al., 2015; Feric et al., 2016) providing opportunities for detailed modelling that is not possible in vivo. These condensed phase droplets lack a bounding membrane, and rapidly exchange molecules with the surrounding cytoplasm.

The formation of condensed droplets in the cytoplasm is an example of a liquid-liquid phase separation (LLPS) that drives the formation of regions enriched in RNA and selected protein species surrounded by the remaining species-poor cytoplasm (Berry et al., 2015; Brangwynne et al., 2015). Within these domains proteins may exhibit distinct phases of matter. This is reminiscent of the liquid-ordered phase in the lipid membranes (Ipsen et al., 1987), and may reflect a functional need (Bagatolli et al., 2010; Jacobs and Frenkel 2017). In the three-dimensional realm of the cytoplasm, membrane-less fluid droplets exhibit complex dynamics. Water in the cytoplasm responds to oscillatory cellular metabolic processes (Thoke et al., 2015). The combination of metabolic coupling in the cytoplasm and active enhancement of biochemical processing in droplets form an exciting starting point for exploring how cells control their internal state.

The structure, function and modelling of condensed phase droplets are the subjects of this workshop. Expected outcomes include an improved understanding of these phases and novel methods for manipulating droplets for therapeutic purposes (Aguzzi and Altmeyer, 2016) using optogenetics (Shin et al. 2017) and microfluidic control (Taylor et al., 2016).

The purpose of this workshop is to bring together scientists with complementary skills (experimentalists, simulators and theorists) to discuss new results in the field of liquid-liquid phase transitions in the cellular cytoplasm. Since a similar workshop in 2014, there has been a surge of new exciting experimental results on coarsening kinetics of droplets and observations of multiple co-existing liquid phases (Berry 2015, Brangwynne 2015, Feric 2016) along with the development of mean-field continuum models awaiting quantitative testing (Berry 2015, Shin 2017, Jacobs and Frenkel, 2017). The increasing power of computer hardware makes it possible to study simplified models of LLPS using coarse-grained simulations. The discussion will be focussed on the challenges in developing simplified theoretical models of the relevant multicomponent experimental systems, and identifying applicable simulation techniques, as well as surveying the wide range of experiments being conducted in vivo and in vitro.

1) What can be learned from minimal experimental systems?

The cellular cytoplasm is a highly-crowded, multicomponent fluid and in order to untangle the features relevant to LLPS, simplified experimental systems are required. A range of in vitro systems that recapitulate the formation of condensed liquid droplets have been developed, and form the basis for simulations assessing the importance of crowding, fluid structure, and entropic forces on droplet formation and dynamics. Light microscopy and optogenetics provide powerful tools for probing and controlling these minimal systems.

2) Can computer simulations suggest routes for experimental manipulations using these techniques?

Membraneless liquid droplets have sizes on micron length scales, which is on the border of system sizes that can be simulated using coarse-grained, particle-based simulation techniques such as dissipative particle dynamics. Parallel coarse-grained simulation tools can simulate large numbers of molecules, but how much detail is needed to accurately reproduce experimental behaviour? Are simplified force fields sufficient to capture experimental systems that involve highly-charged polyampholytes?

3) Can theoretical models identify the relevant thermodynamic forces crucial for droplet formation and sustained existence?

Proteins that form condensed liquid droplets typically have many low complexity regions, are multiply-charged, and lack secondary structure. A combination of many weak forces are involved in their interactions. Polymers are also subject to strong entropic forces that are sensitive to the state of their aqueous environment, including salt concentration, pH, crowding, etc. Theory suggests that a multi-component interacting system may demix or remain dispersed depending on the variance of a random assignment of interaction strengths between the molecule types, and that this behavior requires only small changes in composition and interaction strength (Jacobs and Frenkel, 2017). How do we go beyond existing mean field theories to predict more complex droplet dynamics?



A. Aguzzi and M. Altmeyer, Trends in Cell Biology 26:547-558 (2016)

L. A. Bagatolli et al., Progress in Lipid Research 49:478-389 (2010)

S. F. Banani et al., Nature Reviews Molecular Cell Biology 18:285-298 (2017)

J. Berry et al., PNAS 112:E5237-E5245 (2015)

C. P. Brangwynne et al., Nature Physics 11:899-904 (2015)

E. M. Courchaine et al., EMBO Journal 35:1603-1612 (2016)

M. Feric et al., Cell 165:1-12 (2016)

A. A. Hyman and C. P. Brangwynne, Developmental Cell 21:14-16 (2011)

J. H. Ipsen et al., Biochim. Biophys. Acta 905:162-172 (1987)

W. M. Jacobs and D. Frenkel, Biophys. J. 112:683-691 (2017)

Y. Lin et al., Mol. Cell. 60:208-219 (2015)

D. Milovanovic and P. De Camilli, Neuron 93:995-1002 (2017)

Y. Shin et al., Cell 168:159-171 (2017)

N. Taylor et al., Soft Matter 12:9142-9150 (2016)

H. S. Thoke et al., PLoS ONE 10(2): e0117308 (2015)

S. C. Weber and C. P. Brangwynne, Curr. Biol. 25:641-646 (2015)

M. Zeng et al., Cell 166:1163-1175 (2016)